diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,318052 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 45430, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 2.201188641866608e-05, + "grad_norm": 9.443755079884987, + "learning_rate": 7.336757153338226e-09, + "loss": 1.2788, + "step": 1 + }, + { + "epoch": 4.402377283733216e-05, + "grad_norm": 10.32259595416518, + "learning_rate": 1.4673514306676451e-08, + "loss": 1.2056, + "step": 2 + }, + { + "epoch": 6.603565925599824e-05, + "grad_norm": 1.8090913347133906, + "learning_rate": 2.2010271460014677e-08, + "loss": 0.7536, + "step": 3 + }, + { + "epoch": 8.804754567466431e-05, + "grad_norm": 10.393457773490553, + "learning_rate": 2.9347028613352903e-08, + "loss": 1.2394, + "step": 4 + }, + { + "epoch": 0.00011005943209333039, + "grad_norm": 9.162485143395154, + "learning_rate": 3.6683785766691125e-08, + "loss": 1.2157, + "step": 5 + }, + { + "epoch": 0.00013207131851199647, + "grad_norm": 9.226695874496, + "learning_rate": 4.4020542920029354e-08, + "loss": 1.2069, + "step": 6 + }, + { + "epoch": 0.00015408320493066256, + "grad_norm": 11.50290479340809, + "learning_rate": 5.1357300073367576e-08, + "loss": 1.1886, + "step": 7 + }, + { + "epoch": 0.00017609509134932863, + "grad_norm": 11.020203571240272, + "learning_rate": 5.8694057226705805e-08, + "loss": 1.1775, + "step": 8 + }, + { + "epoch": 0.00019810697776799472, + "grad_norm": 13.274475876904893, + "learning_rate": 6.603081438004402e-08, + "loss": 1.245, + "step": 9 + }, + { + "epoch": 0.00022011886418666079, + "grad_norm": 11.251683094666284, + "learning_rate": 7.336757153338225e-08, + "loss": 1.3414, + "step": 10 + }, + { + "epoch": 0.00024213075060532688, + "grad_norm": 10.199390858372784, + "learning_rate": 8.070432868672047e-08, + "loss": 1.1712, + "step": 11 + }, + { + "epoch": 0.00026414263702399294, + "grad_norm": 11.138083294223705, + "learning_rate": 8.804108584005871e-08, + "loss": 1.1716, + "step": 12 + }, + { + "epoch": 0.00028615452344265906, + "grad_norm": 10.62020324549401, + "learning_rate": 9.537784299339694e-08, + "loss": 1.2208, + "step": 13 + }, + { + "epoch": 0.0003081664098613251, + "grad_norm": 1.7882462579272778, + "learning_rate": 1.0271460014673515e-07, + "loss": 0.727, + "step": 14 + }, + { + "epoch": 0.0003301782962799912, + "grad_norm": 1.9684679533150018, + "learning_rate": 1.1005135730007338e-07, + "loss": 0.8217, + "step": 15 + }, + { + "epoch": 0.00035219018269865726, + "grad_norm": 10.587403146798803, + "learning_rate": 1.1738811445341161e-07, + "loss": 1.1876, + "step": 16 + }, + { + "epoch": 0.0003742020691173234, + "grad_norm": 11.624572032501217, + "learning_rate": 1.2472487160674983e-07, + "loss": 1.235, + "step": 17 + }, + { + "epoch": 0.00039621395553598944, + "grad_norm": 1.8712784470722021, + "learning_rate": 1.3206162876008804e-07, + "loss": 0.7769, + "step": 18 + }, + { + "epoch": 0.0004182258419546555, + "grad_norm": 10.314989317720634, + "learning_rate": 1.3939838591342628e-07, + "loss": 1.2372, + "step": 19 + }, + { + "epoch": 0.00044023772837332157, + "grad_norm": 11.865149723463569, + "learning_rate": 1.467351430667645e-07, + "loss": 1.3276, + "step": 20 + }, + { + "epoch": 0.0004622496147919877, + "grad_norm": 1.7189289096147446, + "learning_rate": 1.5407190022010272e-07, + "loss": 0.7586, + "step": 21 + }, + { + "epoch": 0.00048426150121065375, + "grad_norm": 10.873795940836086, + "learning_rate": 1.6140865737344093e-07, + "loss": 1.1596, + "step": 22 + }, + { + "epoch": 0.0005062733876293198, + "grad_norm": 9.629412972424273, + "learning_rate": 1.6874541452677917e-07, + "loss": 1.1718, + "step": 23 + }, + { + "epoch": 0.0005282852740479859, + "grad_norm": 12.093232381545937, + "learning_rate": 1.7608217168011742e-07, + "loss": 1.138, + "step": 24 + }, + { + "epoch": 0.000550297160466652, + "grad_norm": 11.537246301843545, + "learning_rate": 1.8341892883345563e-07, + "loss": 1.185, + "step": 25 + }, + { + "epoch": 0.0005723090468853181, + "grad_norm": 11.333248736309182, + "learning_rate": 1.9075568598679387e-07, + "loss": 1.3223, + "step": 26 + }, + { + "epoch": 0.0005943209333039842, + "grad_norm": 1.971160732437492, + "learning_rate": 1.9809244314013206e-07, + "loss": 0.781, + "step": 27 + }, + { + "epoch": 0.0006163328197226503, + "grad_norm": 1.8291590520516983, + "learning_rate": 2.054292002934703e-07, + "loss": 0.7536, + "step": 28 + }, + { + "epoch": 0.0006383447061413163, + "grad_norm": 11.071207278465293, + "learning_rate": 2.1276595744680852e-07, + "loss": 1.2693, + "step": 29 + }, + { + "epoch": 0.0006603565925599824, + "grad_norm": 12.505443092828417, + "learning_rate": 2.2010271460014676e-07, + "loss": 1.2466, + "step": 30 + }, + { + "epoch": 0.0006823684789786484, + "grad_norm": 11.412583392956517, + "learning_rate": 2.2743947175348498e-07, + "loss": 1.1782, + "step": 31 + }, + { + "epoch": 0.0007043803653973145, + "grad_norm": 12.498898515640175, + "learning_rate": 2.3477622890682322e-07, + "loss": 1.0968, + "step": 32 + }, + { + "epoch": 0.0007263922518159806, + "grad_norm": 1.6066690178539367, + "learning_rate": 2.4211298606016144e-07, + "loss": 0.7498, + "step": 33 + }, + { + "epoch": 0.0007484041382346468, + "grad_norm": 14.572843100434868, + "learning_rate": 2.4944974321349965e-07, + "loss": 1.1424, + "step": 34 + }, + { + "epoch": 0.0007704160246533128, + "grad_norm": 1.8148123183481277, + "learning_rate": 2.5678650036683787e-07, + "loss": 0.7864, + "step": 35 + }, + { + "epoch": 0.0007924279110719789, + "grad_norm": 10.623374232874383, + "learning_rate": 2.641232575201761e-07, + "loss": 1.1442, + "step": 36 + }, + { + "epoch": 0.0008144397974906449, + "grad_norm": 1.7444673165732205, + "learning_rate": 2.7146001467351435e-07, + "loss": 0.7584, + "step": 37 + }, + { + "epoch": 0.000836451683909311, + "grad_norm": 8.917946106900715, + "learning_rate": 2.7879677182685257e-07, + "loss": 1.0359, + "step": 38 + }, + { + "epoch": 0.0008584635703279771, + "grad_norm": 1.8243151721322868, + "learning_rate": 2.861335289801908e-07, + "loss": 0.7598, + "step": 39 + }, + { + "epoch": 0.0008804754567466431, + "grad_norm": 10.089482356384137, + "learning_rate": 2.93470286133529e-07, + "loss": 1.0946, + "step": 40 + }, + { + "epoch": 0.0009024873431653093, + "grad_norm": 10.312746194236366, + "learning_rate": 3.008070432868672e-07, + "loss": 1.0309, + "step": 41 + }, + { + "epoch": 0.0009244992295839754, + "grad_norm": 1.8825522995884347, + "learning_rate": 3.0814380044020543e-07, + "loss": 0.7897, + "step": 42 + }, + { + "epoch": 0.0009465111160026414, + "grad_norm": 9.580038139147332, + "learning_rate": 3.1548055759354365e-07, + "loss": 1.0861, + "step": 43 + }, + { + "epoch": 0.0009685230024213075, + "grad_norm": 10.884718791393711, + "learning_rate": 3.2281731474688186e-07, + "loss": 1.0538, + "step": 44 + }, + { + "epoch": 0.0009905348888399736, + "grad_norm": 8.216264293397098, + "learning_rate": 3.3015407190022013e-07, + "loss": 1.0871, + "step": 45 + }, + { + "epoch": 0.0010125467752586396, + "grad_norm": 9.603432303131271, + "learning_rate": 3.3749082905355835e-07, + "loss": 1.0464, + "step": 46 + }, + { + "epoch": 0.0010345586616773057, + "grad_norm": 1.6366657014425383, + "learning_rate": 3.4482758620689656e-07, + "loss": 0.7083, + "step": 47 + }, + { + "epoch": 0.0010565705480959718, + "grad_norm": 9.892355259322965, + "learning_rate": 3.5216434336023483e-07, + "loss": 1.1584, + "step": 48 + }, + { + "epoch": 0.0010785824345146378, + "grad_norm": 27.000178497963745, + "learning_rate": 3.5950110051357305e-07, + "loss": 1.0921, + "step": 49 + }, + { + "epoch": 0.001100594320933304, + "grad_norm": 8.370297756072365, + "learning_rate": 3.6683785766691126e-07, + "loss": 1.0619, + "step": 50 + }, + { + "epoch": 0.00112260620735197, + "grad_norm": 9.840618017278643, + "learning_rate": 3.741746148202495e-07, + "loss": 0.9894, + "step": 51 + }, + { + "epoch": 0.0011446180937706362, + "grad_norm": 6.816323215530582, + "learning_rate": 3.8151137197358775e-07, + "loss": 0.822, + "step": 52 + }, + { + "epoch": 0.0011666299801893023, + "grad_norm": 8.338946436540887, + "learning_rate": 3.8884812912692596e-07, + "loss": 0.8879, + "step": 53 + }, + { + "epoch": 0.0011886418666079684, + "grad_norm": 1.7277371716391692, + "learning_rate": 3.961848862802641e-07, + "loss": 0.7847, + "step": 54 + }, + { + "epoch": 0.0012106537530266344, + "grad_norm": 6.304818928081049, + "learning_rate": 4.0352164343360234e-07, + "loss": 0.8631, + "step": 55 + }, + { + "epoch": 0.0012326656394453005, + "grad_norm": 5.529746102131662, + "learning_rate": 4.108584005869406e-07, + "loss": 0.815, + "step": 56 + }, + { + "epoch": 0.0012546775258639666, + "grad_norm": 1.7721071346820363, + "learning_rate": 4.181951577402788e-07, + "loss": 0.7993, + "step": 57 + }, + { + "epoch": 0.0012766894122826326, + "grad_norm": 5.663246005175035, + "learning_rate": 4.2553191489361704e-07, + "loss": 0.9506, + "step": 58 + }, + { + "epoch": 0.0012987012987012987, + "grad_norm": 5.0058583236673595, + "learning_rate": 4.3286867204695526e-07, + "loss": 0.9737, + "step": 59 + }, + { + "epoch": 0.0013207131851199648, + "grad_norm": 5.072629448889672, + "learning_rate": 4.402054292002935e-07, + "loss": 0.8114, + "step": 60 + }, + { + "epoch": 0.0013427250715386308, + "grad_norm": 7.95330356664534, + "learning_rate": 4.4754218635363174e-07, + "loss": 0.819, + "step": 61 + }, + { + "epoch": 0.001364736957957297, + "grad_norm": 5.0530268605600375, + "learning_rate": 4.5487894350696996e-07, + "loss": 0.8353, + "step": 62 + }, + { + "epoch": 0.001386748844375963, + "grad_norm": 4.951500587461793, + "learning_rate": 4.622157006603082e-07, + "loss": 1.0047, + "step": 63 + }, + { + "epoch": 0.001408760730794629, + "grad_norm": 4.533626888732095, + "learning_rate": 4.6955245781364644e-07, + "loss": 0.9539, + "step": 64 + }, + { + "epoch": 0.001430772617213295, + "grad_norm": 4.694807721710902, + "learning_rate": 4.768892149669847e-07, + "loss": 1.0198, + "step": 65 + }, + { + "epoch": 0.0014527845036319612, + "grad_norm": 3.9917323893946755, + "learning_rate": 4.842259721203229e-07, + "loss": 0.7652, + "step": 66 + }, + { + "epoch": 0.0014747963900506274, + "grad_norm": 3.9971534947793628, + "learning_rate": 4.915627292736611e-07, + "loss": 0.8178, + "step": 67 + }, + { + "epoch": 0.0014968082764692935, + "grad_norm": 4.2708219368838565, + "learning_rate": 4.988994864269993e-07, + "loss": 0.7694, + "step": 68 + }, + { + "epoch": 0.0015188201628879596, + "grad_norm": 3.9658706637836, + "learning_rate": 5.062362435803375e-07, + "loss": 0.8029, + "step": 69 + }, + { + "epoch": 0.0015408320493066256, + "grad_norm": 1.764654934831639, + "learning_rate": 5.135730007336757e-07, + "loss": 0.7876, + "step": 70 + }, + { + "epoch": 0.0015628439357252917, + "grad_norm": 4.222363903178032, + "learning_rate": 5.20909757887014e-07, + "loss": 0.8802, + "step": 71 + }, + { + "epoch": 0.0015848558221439578, + "grad_norm": 3.7156758419054396, + "learning_rate": 5.282465150403522e-07, + "loss": 0.8255, + "step": 72 + }, + { + "epoch": 0.0016068677085626238, + "grad_norm": 4.835314273457954, + "learning_rate": 5.355832721936905e-07, + "loss": 0.8451, + "step": 73 + }, + { + "epoch": 0.0016288795949812899, + "grad_norm": 4.759623947462703, + "learning_rate": 5.429200293470287e-07, + "loss": 0.7905, + "step": 74 + }, + { + "epoch": 0.001650891481399956, + "grad_norm": 4.053577997462934, + "learning_rate": 5.502567865003669e-07, + "loss": 0.8643, + "step": 75 + }, + { + "epoch": 0.001672903367818622, + "grad_norm": 3.187620080082407, + "learning_rate": 5.575935436537051e-07, + "loss": 0.7563, + "step": 76 + }, + { + "epoch": 0.001694915254237288, + "grad_norm": 4.317521907564232, + "learning_rate": 5.649303008070434e-07, + "loss": 0.762, + "step": 77 + }, + { + "epoch": 0.0017169271406559542, + "grad_norm": 3.0613451679857664, + "learning_rate": 5.722670579603816e-07, + "loss": 0.6976, + "step": 78 + }, + { + "epoch": 0.0017389390270746202, + "grad_norm": 3.6667524616203493, + "learning_rate": 5.796038151137198e-07, + "loss": 0.7887, + "step": 79 + }, + { + "epoch": 0.0017609509134932863, + "grad_norm": 4.197061895793068, + "learning_rate": 5.86940572267058e-07, + "loss": 0.6974, + "step": 80 + }, + { + "epoch": 0.0017829627999119523, + "grad_norm": 6.097065553492179, + "learning_rate": 5.942773294203962e-07, + "loss": 0.8121, + "step": 81 + }, + { + "epoch": 0.0018049746863306186, + "grad_norm": 1.6643464508827965, + "learning_rate": 6.016140865737344e-07, + "loss": 0.7582, + "step": 82 + }, + { + "epoch": 0.0018269865727492847, + "grad_norm": 2.8201097532083073, + "learning_rate": 6.089508437270726e-07, + "loss": 0.7755, + "step": 83 + }, + { + "epoch": 0.0018489984591679508, + "grad_norm": 2.9664382131562674, + "learning_rate": 6.162876008804109e-07, + "loss": 0.7335, + "step": 84 + }, + { + "epoch": 0.0018710103455866168, + "grad_norm": 2.9201946262377834, + "learning_rate": 6.236243580337492e-07, + "loss": 0.7162, + "step": 85 + }, + { + "epoch": 0.0018930222320052829, + "grad_norm": 3.0198902466551423, + "learning_rate": 6.309611151870873e-07, + "loss": 0.7426, + "step": 86 + }, + { + "epoch": 0.001915034118423949, + "grad_norm": 2.548381257710728, + "learning_rate": 6.382978723404255e-07, + "loss": 0.6347, + "step": 87 + }, + { + "epoch": 0.001937046004842615, + "grad_norm": 2.313202680438072, + "learning_rate": 6.456346294937637e-07, + "loss": 0.6581, + "step": 88 + }, + { + "epoch": 0.001959057891261281, + "grad_norm": 2.5666806398294106, + "learning_rate": 6.52971386647102e-07, + "loss": 0.7146, + "step": 89 + }, + { + "epoch": 0.001981069777679947, + "grad_norm": 2.5222538539061174, + "learning_rate": 6.603081438004403e-07, + "loss": 0.6542, + "step": 90 + }, + { + "epoch": 0.002003081664098613, + "grad_norm": 3.8093071059593675, + "learning_rate": 6.676449009537785e-07, + "loss": 0.7676, + "step": 91 + }, + { + "epoch": 0.0020250935505172793, + "grad_norm": 2.820931438728749, + "learning_rate": 6.749816581071167e-07, + "loss": 0.6979, + "step": 92 + }, + { + "epoch": 0.0020471054369359453, + "grad_norm": 2.518328489513722, + "learning_rate": 6.823184152604549e-07, + "loss": 0.7212, + "step": 93 + }, + { + "epoch": 0.0020691173233546114, + "grad_norm": 2.3767828598032983, + "learning_rate": 6.896551724137931e-07, + "loss": 0.629, + "step": 94 + }, + { + "epoch": 0.0020911292097732775, + "grad_norm": 1.6133185931181362, + "learning_rate": 6.969919295671313e-07, + "loss": 0.7425, + "step": 95 + }, + { + "epoch": 0.0021131410961919435, + "grad_norm": 2.2305933376595704, + "learning_rate": 7.043286867204697e-07, + "loss": 0.6308, + "step": 96 + }, + { + "epoch": 0.0021351529826106096, + "grad_norm": 1.8487748686197516, + "learning_rate": 7.116654438738079e-07, + "loss": 0.7803, + "step": 97 + }, + { + "epoch": 0.0021571648690292757, + "grad_norm": 3.321265359794489, + "learning_rate": 7.190022010271461e-07, + "loss": 0.6485, + "step": 98 + }, + { + "epoch": 0.0021791767554479417, + "grad_norm": 3.4634244966277477, + "learning_rate": 7.263389581804843e-07, + "loss": 0.5934, + "step": 99 + }, + { + "epoch": 0.002201188641866608, + "grad_norm": 2.4208601092884443, + "learning_rate": 7.336757153338225e-07, + "loss": 0.7365, + "step": 100 + }, + { + "epoch": 0.002223200528285274, + "grad_norm": 2.2705848631619414, + "learning_rate": 7.410124724871607e-07, + "loss": 0.6561, + "step": 101 + }, + { + "epoch": 0.00224521241470394, + "grad_norm": 2.5591295712890747, + "learning_rate": 7.48349229640499e-07, + "loss": 0.7473, + "step": 102 + }, + { + "epoch": 0.002267224301122606, + "grad_norm": 2.183724308514613, + "learning_rate": 7.556859867938373e-07, + "loss": 0.7446, + "step": 103 + }, + { + "epoch": 0.0022892361875412725, + "grad_norm": 2.055233676930236, + "learning_rate": 7.630227439471755e-07, + "loss": 0.5778, + "step": 104 + }, + { + "epoch": 0.0023112480739599386, + "grad_norm": 2.6386448468216788, + "learning_rate": 7.703595011005137e-07, + "loss": 0.638, + "step": 105 + }, + { + "epoch": 0.0023332599603786046, + "grad_norm": 2.423937969270911, + "learning_rate": 7.776962582538519e-07, + "loss": 0.6335, + "step": 106 + }, + { + "epoch": 0.0023552718467972707, + "grad_norm": 2.622486150987964, + "learning_rate": 7.850330154071901e-07, + "loss": 0.6459, + "step": 107 + }, + { + "epoch": 0.0023772837332159368, + "grad_norm": 1.9070488535656602, + "learning_rate": 7.923697725605282e-07, + "loss": 0.654, + "step": 108 + }, + { + "epoch": 0.002399295619634603, + "grad_norm": 3.367927432035045, + "learning_rate": 7.997065297138665e-07, + "loss": 0.6405, + "step": 109 + }, + { + "epoch": 0.002421307506053269, + "grad_norm": 2.3317699079002145, + "learning_rate": 8.070432868672047e-07, + "loss": 0.6614, + "step": 110 + }, + { + "epoch": 0.002443319392471935, + "grad_norm": 1.9611503202345977, + "learning_rate": 8.143800440205429e-07, + "loss": 0.6036, + "step": 111 + }, + { + "epoch": 0.002465331278890601, + "grad_norm": 2.269538022798717, + "learning_rate": 8.217168011738812e-07, + "loss": 0.6308, + "step": 112 + }, + { + "epoch": 0.002487343165309267, + "grad_norm": 2.0921137284170364, + "learning_rate": 8.290535583272194e-07, + "loss": 0.6634, + "step": 113 + }, + { + "epoch": 0.002509355051727933, + "grad_norm": 2.813778932476409, + "learning_rate": 8.363903154805576e-07, + "loss": 0.7834, + "step": 114 + }, + { + "epoch": 0.002531366938146599, + "grad_norm": 2.388961633955085, + "learning_rate": 8.437270726338959e-07, + "loss": 0.6261, + "step": 115 + }, + { + "epoch": 0.0025533788245652653, + "grad_norm": 2.1906964466148886, + "learning_rate": 8.510638297872341e-07, + "loss": 0.6089, + "step": 116 + }, + { + "epoch": 0.0025753907109839313, + "grad_norm": 1.964704937004403, + "learning_rate": 8.584005869405723e-07, + "loss": 0.5437, + "step": 117 + }, + { + "epoch": 0.0025974025974025974, + "grad_norm": 2.2992566636908456, + "learning_rate": 8.657373440939105e-07, + "loss": 0.7093, + "step": 118 + }, + { + "epoch": 0.0026194144838212635, + "grad_norm": 2.6872772890708654, + "learning_rate": 8.730741012472488e-07, + "loss": 0.5883, + "step": 119 + }, + { + "epoch": 0.0026414263702399295, + "grad_norm": 1.760265646649394, + "learning_rate": 8.80410858400587e-07, + "loss": 0.7935, + "step": 120 + }, + { + "epoch": 0.0026634382566585956, + "grad_norm": 2.387052958127284, + "learning_rate": 8.877476155539253e-07, + "loss": 0.5726, + "step": 121 + }, + { + "epoch": 0.0026854501430772617, + "grad_norm": 2.4231147488507805, + "learning_rate": 8.950843727072635e-07, + "loss": 0.7089, + "step": 122 + }, + { + "epoch": 0.0027074620294959277, + "grad_norm": 1.676218806936214, + "learning_rate": 9.024211298606017e-07, + "loss": 0.7795, + "step": 123 + }, + { + "epoch": 0.002729473915914594, + "grad_norm": 1.6001822661758969, + "learning_rate": 9.097578870139399e-07, + "loss": 0.7257, + "step": 124 + }, + { + "epoch": 0.00275148580233326, + "grad_norm": 1.843623573033582, + "learning_rate": 9.170946441672781e-07, + "loss": 0.604, + "step": 125 + }, + { + "epoch": 0.002773497688751926, + "grad_norm": 2.286012855715687, + "learning_rate": 9.244314013206164e-07, + "loss": 0.5768, + "step": 126 + }, + { + "epoch": 0.002795509575170592, + "grad_norm": 2.5607765669266285, + "learning_rate": 9.317681584739547e-07, + "loss": 0.692, + "step": 127 + }, + { + "epoch": 0.002817521461589258, + "grad_norm": 2.0753974809166, + "learning_rate": 9.391049156272929e-07, + "loss": 0.6165, + "step": 128 + }, + { + "epoch": 0.002839533348007924, + "grad_norm": 1.9432366861210497, + "learning_rate": 9.464416727806311e-07, + "loss": 0.5832, + "step": 129 + }, + { + "epoch": 0.00286154523442659, + "grad_norm": 2.250496694743626, + "learning_rate": 9.537784299339693e-07, + "loss": 0.5975, + "step": 130 + }, + { + "epoch": 0.0028835571208452562, + "grad_norm": 1.9552036802642647, + "learning_rate": 9.611151870873076e-07, + "loss": 0.5509, + "step": 131 + }, + { + "epoch": 0.0029055690072639223, + "grad_norm": 2.0942205883793457, + "learning_rate": 9.684519442406457e-07, + "loss": 0.5613, + "step": 132 + }, + { + "epoch": 0.002927580893682589, + "grad_norm": 2.1152329942412282, + "learning_rate": 9.757887013939839e-07, + "loss": 0.6133, + "step": 133 + }, + { + "epoch": 0.002949592780101255, + "grad_norm": 2.042444151618426, + "learning_rate": 9.831254585473222e-07, + "loss": 0.5254, + "step": 134 + }, + { + "epoch": 0.002971604666519921, + "grad_norm": 2.113481386054546, + "learning_rate": 9.904622157006603e-07, + "loss": 0.5868, + "step": 135 + }, + { + "epoch": 0.002993616552938587, + "grad_norm": 2.456460206057515, + "learning_rate": 9.977989728539986e-07, + "loss": 0.5776, + "step": 136 + }, + { + "epoch": 0.003015628439357253, + "grad_norm": 1.607134157851857, + "learning_rate": 1.0051357300073367e-06, + "loss": 0.7832, + "step": 137 + }, + { + "epoch": 0.003037640325775919, + "grad_norm": 1.9455195624606196, + "learning_rate": 1.012472487160675e-06, + "loss": 0.5308, + "step": 138 + }, + { + "epoch": 0.003059652212194585, + "grad_norm": 1.4806440210626903, + "learning_rate": 1.0198092443140134e-06, + "loss": 0.7425, + "step": 139 + }, + { + "epoch": 0.0030816640986132513, + "grad_norm": 4.137962866061726, + "learning_rate": 1.0271460014673515e-06, + "loss": 0.6377, + "step": 140 + }, + { + "epoch": 0.0031036759850319173, + "grad_norm": 1.8651300173342333, + "learning_rate": 1.0344827586206898e-06, + "loss": 0.5754, + "step": 141 + }, + { + "epoch": 0.0031256878714505834, + "grad_norm": 2.4809284898541466, + "learning_rate": 1.041819515774028e-06, + "loss": 0.5447, + "step": 142 + }, + { + "epoch": 0.0031476997578692495, + "grad_norm": 2.205091290278845, + "learning_rate": 1.0491562729273662e-06, + "loss": 0.5537, + "step": 143 + }, + { + "epoch": 0.0031697116442879155, + "grad_norm": 2.33461549277035, + "learning_rate": 1.0564930300807043e-06, + "loss": 0.6682, + "step": 144 + }, + { + "epoch": 0.0031917235307065816, + "grad_norm": 2.278989921019053, + "learning_rate": 1.0638297872340427e-06, + "loss": 0.6543, + "step": 145 + }, + { + "epoch": 0.0032137354171252477, + "grad_norm": 1.7888236016649126, + "learning_rate": 1.071166544387381e-06, + "loss": 0.5193, + "step": 146 + }, + { + "epoch": 0.0032357473035439137, + "grad_norm": 2.0303630876723884, + "learning_rate": 1.078503301540719e-06, + "loss": 0.5936, + "step": 147 + }, + { + "epoch": 0.0032577591899625798, + "grad_norm": 1.7087405969270817, + "learning_rate": 1.0858400586940574e-06, + "loss": 0.5346, + "step": 148 + }, + { + "epoch": 0.003279771076381246, + "grad_norm": 1.4795406648214442, + "learning_rate": 1.0931768158473955e-06, + "loss": 0.7702, + "step": 149 + }, + { + "epoch": 0.003301782962799912, + "grad_norm": 2.0419167354141443, + "learning_rate": 1.1005135730007338e-06, + "loss": 0.5912, + "step": 150 + }, + { + "epoch": 0.003323794849218578, + "grad_norm": 2.079867753954542, + "learning_rate": 1.107850330154072e-06, + "loss": 0.5947, + "step": 151 + }, + { + "epoch": 0.003345806735637244, + "grad_norm": 1.369301388923201, + "learning_rate": 1.1151870873074103e-06, + "loss": 0.7868, + "step": 152 + }, + { + "epoch": 0.00336781862205591, + "grad_norm": 1.3961896596084973, + "learning_rate": 1.1225238444607486e-06, + "loss": 0.7716, + "step": 153 + }, + { + "epoch": 0.003389830508474576, + "grad_norm": 1.8924506137383468, + "learning_rate": 1.1298606016140867e-06, + "loss": 0.5795, + "step": 154 + }, + { + "epoch": 0.0034118423948932422, + "grad_norm": 1.365736575658436, + "learning_rate": 1.137197358767425e-06, + "loss": 0.7335, + "step": 155 + }, + { + "epoch": 0.0034338542813119083, + "grad_norm": 2.768658042032026, + "learning_rate": 1.1445341159207631e-06, + "loss": 0.5236, + "step": 156 + }, + { + "epoch": 0.0034558661677305744, + "grad_norm": 2.965903322261619, + "learning_rate": 1.1518708730741012e-06, + "loss": 0.5447, + "step": 157 + }, + { + "epoch": 0.0034778780541492404, + "grad_norm": 2.260110147841431, + "learning_rate": 1.1592076302274396e-06, + "loss": 0.541, + "step": 158 + }, + { + "epoch": 0.0034998899405679065, + "grad_norm": 2.062693618020467, + "learning_rate": 1.1665443873807777e-06, + "loss": 0.4976, + "step": 159 + }, + { + "epoch": 0.0035219018269865726, + "grad_norm": 2.9662888469368194, + "learning_rate": 1.173881144534116e-06, + "loss": 0.5767, + "step": 160 + }, + { + "epoch": 0.0035439137134052386, + "grad_norm": 3.109447719520694, + "learning_rate": 1.181217901687454e-06, + "loss": 0.6023, + "step": 161 + }, + { + "epoch": 0.0035659255998239047, + "grad_norm": 2.7393860423462835, + "learning_rate": 1.1885546588407924e-06, + "loss": 0.5638, + "step": 162 + }, + { + "epoch": 0.003587937486242571, + "grad_norm": 1.2098505337613201, + "learning_rate": 1.1958914159941307e-06, + "loss": 0.7376, + "step": 163 + }, + { + "epoch": 0.0036099493726612373, + "grad_norm": 2.2723517983515125, + "learning_rate": 1.2032281731474689e-06, + "loss": 0.5438, + "step": 164 + }, + { + "epoch": 0.0036319612590799033, + "grad_norm": 2.771637749947578, + "learning_rate": 1.2105649303008072e-06, + "loss": 0.5465, + "step": 165 + }, + { + "epoch": 0.0036539731454985694, + "grad_norm": 2.0600894078421024, + "learning_rate": 1.2179016874541453e-06, + "loss": 0.5832, + "step": 166 + }, + { + "epoch": 0.0036759850319172355, + "grad_norm": 2.5683859339428508, + "learning_rate": 1.2252384446074836e-06, + "loss": 0.5475, + "step": 167 + }, + { + "epoch": 0.0036979969183359015, + "grad_norm": 2.6094695985240257, + "learning_rate": 1.2325752017608217e-06, + "loss": 0.5901, + "step": 168 + }, + { + "epoch": 0.0037200088047545676, + "grad_norm": 1.979952498214073, + "learning_rate": 1.23991195891416e-06, + "loss": 0.6196, + "step": 169 + }, + { + "epoch": 0.0037420206911732336, + "grad_norm": 2.2252659876673535, + "learning_rate": 1.2472487160674984e-06, + "loss": 0.4889, + "step": 170 + }, + { + "epoch": 0.0037640325775918997, + "grad_norm": 2.0285928429321842, + "learning_rate": 1.2545854732208365e-06, + "loss": 0.4982, + "step": 171 + }, + { + "epoch": 0.0037860444640105658, + "grad_norm": 1.9434746552133237, + "learning_rate": 1.2619222303741746e-06, + "loss": 0.5505, + "step": 172 + }, + { + "epoch": 0.003808056350429232, + "grad_norm": 2.1065479598228567, + "learning_rate": 1.269258987527513e-06, + "loss": 0.6296, + "step": 173 + }, + { + "epoch": 0.003830068236847898, + "grad_norm": 1.8975711120882008, + "learning_rate": 1.276595744680851e-06, + "loss": 0.5216, + "step": 174 + }, + { + "epoch": 0.003852080123266564, + "grad_norm": 1.0740112251749245, + "learning_rate": 1.2839325018341893e-06, + "loss": 0.6888, + "step": 175 + }, + { + "epoch": 0.00387409200968523, + "grad_norm": 2.0570120183615592, + "learning_rate": 1.2912692589875274e-06, + "loss": 0.5261, + "step": 176 + }, + { + "epoch": 0.003896103896103896, + "grad_norm": 2.685593331487789, + "learning_rate": 1.298606016140866e-06, + "loss": 0.5185, + "step": 177 + }, + { + "epoch": 0.003918115782522562, + "grad_norm": 3.2497377870643946, + "learning_rate": 1.305942773294204e-06, + "loss": 0.4839, + "step": 178 + }, + { + "epoch": 0.003940127668941229, + "grad_norm": 1.8195290518501561, + "learning_rate": 1.3132795304475424e-06, + "loss": 0.5339, + "step": 179 + }, + { + "epoch": 0.003962139555359894, + "grad_norm": 2.560672870175208, + "learning_rate": 1.3206162876008805e-06, + "loss": 0.5212, + "step": 180 + }, + { + "epoch": 0.003984151441778561, + "grad_norm": 0.9957609096676734, + "learning_rate": 1.3279530447542188e-06, + "loss": 0.7198, + "step": 181 + }, + { + "epoch": 0.004006163328197226, + "grad_norm": 2.041638699996718, + "learning_rate": 1.335289801907557e-06, + "loss": 0.5999, + "step": 182 + }, + { + "epoch": 0.004028175214615893, + "grad_norm": 1.7967357547259357, + "learning_rate": 1.3426265590608953e-06, + "loss": 0.5547, + "step": 183 + }, + { + "epoch": 0.0040501871010345586, + "grad_norm": 1.0478052042206953, + "learning_rate": 1.3499633162142334e-06, + "loss": 0.7286, + "step": 184 + }, + { + "epoch": 0.004072198987453225, + "grad_norm": 1.0298924887754308, + "learning_rate": 1.3573000733675717e-06, + "loss": 0.7113, + "step": 185 + }, + { + "epoch": 0.004094210873871891, + "grad_norm": 2.013607738336974, + "learning_rate": 1.3646368305209098e-06, + "loss": 0.5986, + "step": 186 + }, + { + "epoch": 0.004116222760290557, + "grad_norm": 1.8370780446133104, + "learning_rate": 1.3719735876742481e-06, + "loss": 0.5906, + "step": 187 + }, + { + "epoch": 0.004138234646709223, + "grad_norm": 0.980127715611591, + "learning_rate": 1.3793103448275862e-06, + "loss": 0.7243, + "step": 188 + }, + { + "epoch": 0.004160246533127889, + "grad_norm": 2.653280611810966, + "learning_rate": 1.3866471019809246e-06, + "loss": 0.596, + "step": 189 + }, + { + "epoch": 0.004182258419546555, + "grad_norm": 2.4207671131686057, + "learning_rate": 1.3939838591342627e-06, + "loss": 0.5219, + "step": 190 + }, + { + "epoch": 0.0042042703059652214, + "grad_norm": 1.840098244086559, + "learning_rate": 1.4013206162876012e-06, + "loss": 0.5421, + "step": 191 + }, + { + "epoch": 0.004226282192383887, + "grad_norm": 3.848212402112636, + "learning_rate": 1.4086573734409393e-06, + "loss": 0.5965, + "step": 192 + }, + { + "epoch": 0.004248294078802554, + "grad_norm": 2.283905172289067, + "learning_rate": 1.4159941305942772e-06, + "loss": 0.5214, + "step": 193 + }, + { + "epoch": 0.004270305965221219, + "grad_norm": 0.9042430984727481, + "learning_rate": 1.4233308877476158e-06, + "loss": 0.7044, + "step": 194 + }, + { + "epoch": 0.004292317851639886, + "grad_norm": 2.6081484992104116, + "learning_rate": 1.4306676449009539e-06, + "loss": 0.4868, + "step": 195 + }, + { + "epoch": 0.004314329738058551, + "grad_norm": 0.9698446846534511, + "learning_rate": 1.4380044020542922e-06, + "loss": 0.6972, + "step": 196 + }, + { + "epoch": 0.004336341624477218, + "grad_norm": 5.2850982252563545, + "learning_rate": 1.4453411592076303e-06, + "loss": 0.5944, + "step": 197 + }, + { + "epoch": 0.0043583535108958835, + "grad_norm": 2.7183482091708844, + "learning_rate": 1.4526779163609686e-06, + "loss": 0.4953, + "step": 198 + }, + { + "epoch": 0.00438036539731455, + "grad_norm": 2.0571901872908636, + "learning_rate": 1.4600146735143067e-06, + "loss": 0.5165, + "step": 199 + }, + { + "epoch": 0.004402377283733216, + "grad_norm": 1.966277824409953, + "learning_rate": 1.467351430667645e-06, + "loss": 0.5297, + "step": 200 + }, + { + "epoch": 0.004424389170151882, + "grad_norm": 1.9822340874129638, + "learning_rate": 1.4746881878209832e-06, + "loss": 0.5132, + "step": 201 + }, + { + "epoch": 0.004446401056570548, + "grad_norm": 1.7632278792361251, + "learning_rate": 1.4820249449743215e-06, + "loss": 0.525, + "step": 202 + }, + { + "epoch": 0.004468412942989214, + "grad_norm": 1.767822893022219, + "learning_rate": 1.4893617021276596e-06, + "loss": 0.4933, + "step": 203 + }, + { + "epoch": 0.00449042482940788, + "grad_norm": 2.0046270540303066, + "learning_rate": 1.496698459280998e-06, + "loss": 0.4742, + "step": 204 + }, + { + "epoch": 0.004512436715826546, + "grad_norm": 2.3257951943565085, + "learning_rate": 1.504035216434336e-06, + "loss": 0.6185, + "step": 205 + }, + { + "epoch": 0.004534448602245212, + "grad_norm": 1.87983769956431, + "learning_rate": 1.5113719735876746e-06, + "loss": 0.5506, + "step": 206 + }, + { + "epoch": 0.0045564604886638785, + "grad_norm": 0.8602156736333176, + "learning_rate": 1.5187087307410125e-06, + "loss": 0.7219, + "step": 207 + }, + { + "epoch": 0.004578472375082545, + "grad_norm": 8.708509937967335, + "learning_rate": 1.526045487894351e-06, + "loss": 0.5086, + "step": 208 + }, + { + "epoch": 0.004600484261501211, + "grad_norm": 2.138971792262686, + "learning_rate": 1.533382245047689e-06, + "loss": 0.5226, + "step": 209 + }, + { + "epoch": 0.004622496147919877, + "grad_norm": 2.2379423322898093, + "learning_rate": 1.5407190022010274e-06, + "loss": 0.5078, + "step": 210 + }, + { + "epoch": 0.004644508034338543, + "grad_norm": 1.6303787887945507, + "learning_rate": 1.5480557593543655e-06, + "loss": 0.488, + "step": 211 + }, + { + "epoch": 0.004666519920757209, + "grad_norm": 2.5078579783615655, + "learning_rate": 1.5553925165077038e-06, + "loss": 0.5915, + "step": 212 + }, + { + "epoch": 0.004688531807175875, + "grad_norm": 0.9396086458127094, + "learning_rate": 1.562729273661042e-06, + "loss": 0.764, + "step": 213 + }, + { + "epoch": 0.004710543693594541, + "grad_norm": 1.9741526735583625, + "learning_rate": 1.5700660308143803e-06, + "loss": 0.5744, + "step": 214 + }, + { + "epoch": 0.004732555580013207, + "grad_norm": 3.1999385243988065, + "learning_rate": 1.5774027879677184e-06, + "loss": 0.5709, + "step": 215 + }, + { + "epoch": 0.0047545674664318735, + "grad_norm": 2.077591608093479, + "learning_rate": 1.5847395451210565e-06, + "loss": 0.5699, + "step": 216 + }, + { + "epoch": 0.004776579352850539, + "grad_norm": 9.313825077103676, + "learning_rate": 1.5920763022743948e-06, + "loss": 0.6395, + "step": 217 + }, + { + "epoch": 0.004798591239269206, + "grad_norm": 2.8917699927619607, + "learning_rate": 1.599413059427733e-06, + "loss": 0.5462, + "step": 218 + }, + { + "epoch": 0.004820603125687871, + "grad_norm": 0.8735026715332712, + "learning_rate": 1.6067498165810713e-06, + "loss": 0.7396, + "step": 219 + }, + { + "epoch": 0.004842615012106538, + "grad_norm": 2.3990832684898047, + "learning_rate": 1.6140865737344094e-06, + "loss": 0.6281, + "step": 220 + }, + { + "epoch": 0.004864626898525203, + "grad_norm": 2.077240569154774, + "learning_rate": 1.6214233308877477e-06, + "loss": 0.6483, + "step": 221 + }, + { + "epoch": 0.00488663878494387, + "grad_norm": 2.3938980115748936, + "learning_rate": 1.6287600880410858e-06, + "loss": 0.5753, + "step": 222 + }, + { + "epoch": 0.0049086506713625355, + "grad_norm": 1.0606632791495523, + "learning_rate": 1.6360968451944243e-06, + "loss": 0.6826, + "step": 223 + }, + { + "epoch": 0.004930662557781202, + "grad_norm": 1.8493290202720634, + "learning_rate": 1.6434336023477624e-06, + "loss": 0.501, + "step": 224 + }, + { + "epoch": 0.004952674444199868, + "grad_norm": 2.224505086256316, + "learning_rate": 1.6507703595011008e-06, + "loss": 0.5265, + "step": 225 + }, + { + "epoch": 0.004974686330618534, + "grad_norm": 1.908206978225864, + "learning_rate": 1.6581071166544389e-06, + "loss": 0.4931, + "step": 226 + }, + { + "epoch": 0.0049966982170372, + "grad_norm": 237.2762404115871, + "learning_rate": 1.6654438738077772e-06, + "loss": 0.6458, + "step": 227 + }, + { + "epoch": 0.005018710103455866, + "grad_norm": 4.601504278484779, + "learning_rate": 1.6727806309611153e-06, + "loss": 0.5612, + "step": 228 + }, + { + "epoch": 0.005040721989874532, + "grad_norm": 2.3277165484682785, + "learning_rate": 1.6801173881144536e-06, + "loss": 0.5519, + "step": 229 + }, + { + "epoch": 0.005062733876293198, + "grad_norm": 2.018585440482969, + "learning_rate": 1.6874541452677917e-06, + "loss": 0.5858, + "step": 230 + }, + { + "epoch": 0.005084745762711864, + "grad_norm": 0.8121748732797454, + "learning_rate": 1.69479090242113e-06, + "loss": 0.6889, + "step": 231 + }, + { + "epoch": 0.0051067576491305305, + "grad_norm": 2.548520870716747, + "learning_rate": 1.7021276595744682e-06, + "loss": 0.555, + "step": 232 + }, + { + "epoch": 0.005128769535549196, + "grad_norm": 0.7911885986696344, + "learning_rate": 1.7094644167278065e-06, + "loss": 0.7279, + "step": 233 + }, + { + "epoch": 0.005150781421967863, + "grad_norm": 0.8306785740925763, + "learning_rate": 1.7168011738811446e-06, + "loss": 0.722, + "step": 234 + }, + { + "epoch": 0.005172793308386528, + "grad_norm": 2.468456499717846, + "learning_rate": 1.724137931034483e-06, + "loss": 0.419, + "step": 235 + }, + { + "epoch": 0.005194805194805195, + "grad_norm": 1.9371284067561971, + "learning_rate": 1.731474688187821e-06, + "loss": 0.4952, + "step": 236 + }, + { + "epoch": 0.005216817081223861, + "grad_norm": 2.0358373714903397, + "learning_rate": 1.7388114453411596e-06, + "loss": 0.6035, + "step": 237 + }, + { + "epoch": 0.005238828967642527, + "grad_norm": 0.7697470448741319, + "learning_rate": 1.7461482024944977e-06, + "loss": 0.6958, + "step": 238 + }, + { + "epoch": 0.005260840854061193, + "grad_norm": 1.975174169609722, + "learning_rate": 1.753484959647836e-06, + "loss": 0.5652, + "step": 239 + }, + { + "epoch": 0.005282852740479859, + "grad_norm": 2.632802862906087, + "learning_rate": 1.760821716801174e-06, + "loss": 0.5654, + "step": 240 + }, + { + "epoch": 0.0053048646268985256, + "grad_norm": 2.1808667216850535, + "learning_rate": 1.7681584739545122e-06, + "loss": 0.5744, + "step": 241 + }, + { + "epoch": 0.005326876513317191, + "grad_norm": 1.7587344774503393, + "learning_rate": 1.7754952311078505e-06, + "loss": 0.4992, + "step": 242 + }, + { + "epoch": 0.005348888399735858, + "grad_norm": 1.950671472590007, + "learning_rate": 1.7828319882611886e-06, + "loss": 0.4238, + "step": 243 + }, + { + "epoch": 0.005370900286154523, + "grad_norm": 0.8008541230181889, + "learning_rate": 1.790168745414527e-06, + "loss": 0.7045, + "step": 244 + }, + { + "epoch": 0.00539291217257319, + "grad_norm": 2.592798636746794, + "learning_rate": 1.797505502567865e-06, + "loss": 0.4611, + "step": 245 + }, + { + "epoch": 0.0054149240589918555, + "grad_norm": 0.7501485182620865, + "learning_rate": 1.8048422597212034e-06, + "loss": 0.6784, + "step": 246 + }, + { + "epoch": 0.005436935945410522, + "grad_norm": 1.821303303752374, + "learning_rate": 1.8121790168745415e-06, + "loss": 0.4681, + "step": 247 + }, + { + "epoch": 0.005458947831829188, + "grad_norm": 2.4745801142574666, + "learning_rate": 1.8195157740278798e-06, + "loss": 0.5378, + "step": 248 + }, + { + "epoch": 0.005480959718247854, + "grad_norm": 3.4127594395081324, + "learning_rate": 1.826852531181218e-06, + "loss": 0.4048, + "step": 249 + }, + { + "epoch": 0.00550297160466652, + "grad_norm": 1.8370385564730005, + "learning_rate": 1.8341892883345563e-06, + "loss": 0.4545, + "step": 250 + }, + { + "epoch": 0.005524983491085186, + "grad_norm": 0.7518434676783818, + "learning_rate": 1.8415260454878944e-06, + "loss": 0.7108, + "step": 251 + }, + { + "epoch": 0.005546995377503852, + "grad_norm": 2.0307476939715103, + "learning_rate": 1.848862802641233e-06, + "loss": 0.5997, + "step": 252 + }, + { + "epoch": 0.005569007263922518, + "grad_norm": 4.641122226970733, + "learning_rate": 1.8561995597945708e-06, + "loss": 0.5479, + "step": 253 + }, + { + "epoch": 0.005591019150341184, + "grad_norm": 1.9007602395877279, + "learning_rate": 1.8635363169479093e-06, + "loss": 0.5301, + "step": 254 + }, + { + "epoch": 0.0056130310367598505, + "grad_norm": 1.7490602599827896, + "learning_rate": 1.8708730741012474e-06, + "loss": 0.4973, + "step": 255 + }, + { + "epoch": 0.005635042923178516, + "grad_norm": 2.568552959410404, + "learning_rate": 1.8782098312545858e-06, + "loss": 0.5104, + "step": 256 + }, + { + "epoch": 0.005657054809597183, + "grad_norm": 2.079118561495754, + "learning_rate": 1.8855465884079239e-06, + "loss": 0.5697, + "step": 257 + }, + { + "epoch": 0.005679066696015848, + "grad_norm": 2.0718645206976265, + "learning_rate": 1.8928833455612622e-06, + "loss": 0.5588, + "step": 258 + }, + { + "epoch": 0.005701078582434515, + "grad_norm": 2.868317732047967, + "learning_rate": 1.9002201027146003e-06, + "loss": 0.5787, + "step": 259 + }, + { + "epoch": 0.00572309046885318, + "grad_norm": 2.072184459626606, + "learning_rate": 1.9075568598679386e-06, + "loss": 0.5424, + "step": 260 + }, + { + "epoch": 0.005745102355271847, + "grad_norm": 2.3499589775720082, + "learning_rate": 1.9148936170212767e-06, + "loss": 0.4593, + "step": 261 + }, + { + "epoch": 0.0057671142416905125, + "grad_norm": 5.602829304300727, + "learning_rate": 1.9222303741746153e-06, + "loss": 0.5357, + "step": 262 + }, + { + "epoch": 0.005789126128109179, + "grad_norm": 0.7673774144724468, + "learning_rate": 1.9295671313279534e-06, + "loss": 0.6852, + "step": 263 + }, + { + "epoch": 0.005811138014527845, + "grad_norm": 7.612519213268656, + "learning_rate": 1.9369038884812915e-06, + "loss": 0.4424, + "step": 264 + }, + { + "epoch": 0.005833149900946511, + "grad_norm": 2.201245282774354, + "learning_rate": 1.9442406456346296e-06, + "loss": 0.4214, + "step": 265 + }, + { + "epoch": 0.005855161787365178, + "grad_norm": 2.1325283319613577, + "learning_rate": 1.9515774027879677e-06, + "loss": 0.5384, + "step": 266 + }, + { + "epoch": 0.005877173673783843, + "grad_norm": 2.4104140942412906, + "learning_rate": 1.9589141599413062e-06, + "loss": 0.485, + "step": 267 + }, + { + "epoch": 0.00589918556020251, + "grad_norm": 2.137269049417403, + "learning_rate": 1.9662509170946444e-06, + "loss": 0.5285, + "step": 268 + }, + { + "epoch": 0.005921197446621175, + "grad_norm": 4.539334315207091, + "learning_rate": 1.9735876742479825e-06, + "loss": 0.4838, + "step": 269 + }, + { + "epoch": 0.005943209333039842, + "grad_norm": 2.710116539674892, + "learning_rate": 1.9809244314013206e-06, + "loss": 0.5908, + "step": 270 + }, + { + "epoch": 0.0059652212194585075, + "grad_norm": 2.417572118357592, + "learning_rate": 1.988261188554659e-06, + "loss": 0.4493, + "step": 271 + }, + { + "epoch": 0.005987233105877174, + "grad_norm": 2.5333534583770536, + "learning_rate": 1.9955979457079972e-06, + "loss": 0.4341, + "step": 272 + }, + { + "epoch": 0.00600924499229584, + "grad_norm": 1.7795909770169311, + "learning_rate": 2.0029347028613353e-06, + "loss": 0.5617, + "step": 273 + }, + { + "epoch": 0.006031256878714506, + "grad_norm": 9.58493108169327, + "learning_rate": 2.0102714600146734e-06, + "loss": 0.4885, + "step": 274 + }, + { + "epoch": 0.006053268765133172, + "grad_norm": 2.577224364352579, + "learning_rate": 2.017608217168012e-06, + "loss": 0.4792, + "step": 275 + }, + { + "epoch": 0.006075280651551838, + "grad_norm": 0.7606721781024294, + "learning_rate": 2.02494497432135e-06, + "loss": 0.6791, + "step": 276 + }, + { + "epoch": 0.006097292537970504, + "grad_norm": 0.7707912574001987, + "learning_rate": 2.0322817314746886e-06, + "loss": 0.7128, + "step": 277 + }, + { + "epoch": 0.00611930442438917, + "grad_norm": 1.8927866592570675, + "learning_rate": 2.0396184886280267e-06, + "loss": 0.3997, + "step": 278 + }, + { + "epoch": 0.006141316310807836, + "grad_norm": 0.7164492337745023, + "learning_rate": 2.046955245781365e-06, + "loss": 0.672, + "step": 279 + }, + { + "epoch": 0.0061633281972265025, + "grad_norm": 2.1467531158518236, + "learning_rate": 2.054292002934703e-06, + "loss": 0.5459, + "step": 280 + }, + { + "epoch": 0.006185340083645168, + "grad_norm": 5.525508894083202, + "learning_rate": 2.0616287600880415e-06, + "loss": 0.5638, + "step": 281 + }, + { + "epoch": 0.006207351970063835, + "grad_norm": 5.282744198112085, + "learning_rate": 2.0689655172413796e-06, + "loss": 0.3662, + "step": 282 + }, + { + "epoch": 0.0062293638564825, + "grad_norm": 1.9657505644075484, + "learning_rate": 2.0763022743947177e-06, + "loss": 0.5462, + "step": 283 + }, + { + "epoch": 0.006251375742901167, + "grad_norm": 9.346091475500575, + "learning_rate": 2.083639031548056e-06, + "loss": 0.4576, + "step": 284 + }, + { + "epoch": 0.006273387629319832, + "grad_norm": 12.488759900099405, + "learning_rate": 2.0909757887013943e-06, + "loss": 0.6632, + "step": 285 + }, + { + "epoch": 0.006295399515738499, + "grad_norm": 2.53266018404855, + "learning_rate": 2.0983125458547324e-06, + "loss": 0.5257, + "step": 286 + }, + { + "epoch": 0.0063174114021571645, + "grad_norm": 1.893153476817632, + "learning_rate": 2.1056493030080706e-06, + "loss": 0.5422, + "step": 287 + }, + { + "epoch": 0.006339423288575831, + "grad_norm": 2.1383022261404125, + "learning_rate": 2.1129860601614087e-06, + "loss": 0.5459, + "step": 288 + }, + { + "epoch": 0.006361435174994497, + "grad_norm": 2.292112238501794, + "learning_rate": 2.1203228173147468e-06, + "loss": 0.5238, + "step": 289 + }, + { + "epoch": 0.006383447061413163, + "grad_norm": 2.0055012870024473, + "learning_rate": 2.1276595744680853e-06, + "loss": 0.5127, + "step": 290 + }, + { + "epoch": 0.006405458947831829, + "grad_norm": 3.015966871211799, + "learning_rate": 2.1349963316214234e-06, + "loss": 0.4986, + "step": 291 + }, + { + "epoch": 0.006427470834250495, + "grad_norm": 2.153625799230337, + "learning_rate": 2.142333088774762e-06, + "loss": 0.5265, + "step": 292 + }, + { + "epoch": 0.006449482720669161, + "grad_norm": 1.749168483004697, + "learning_rate": 2.1496698459281e-06, + "loss": 0.5172, + "step": 293 + }, + { + "epoch": 0.0064714946070878274, + "grad_norm": 0.6646778801767323, + "learning_rate": 2.157006603081438e-06, + "loss": 0.6518, + "step": 294 + }, + { + "epoch": 0.006493506493506494, + "grad_norm": 2.7617877408452487, + "learning_rate": 2.1643433602347763e-06, + "loss": 0.4213, + "step": 295 + }, + { + "epoch": 0.0065155183799251596, + "grad_norm": 5.170624116456264, + "learning_rate": 2.171680117388115e-06, + "loss": 0.5438, + "step": 296 + }, + { + "epoch": 0.006537530266343826, + "grad_norm": 2.4171812486309765, + "learning_rate": 2.179016874541453e-06, + "loss": 0.4767, + "step": 297 + }, + { + "epoch": 0.006559542152762492, + "grad_norm": 2.242446931047062, + "learning_rate": 2.186353631694791e-06, + "loss": 0.5248, + "step": 298 + }, + { + "epoch": 0.006581554039181158, + "grad_norm": 1.9806138698059337, + "learning_rate": 2.193690388848129e-06, + "loss": 0.5097, + "step": 299 + }, + { + "epoch": 0.006603565925599824, + "grad_norm": 2.5036782211128426, + "learning_rate": 2.2010271460014677e-06, + "loss": 0.5731, + "step": 300 + }, + { + "epoch": 0.00662557781201849, + "grad_norm": 0.7416182972228351, + "learning_rate": 2.2083639031548058e-06, + "loss": 0.6894, + "step": 301 + }, + { + "epoch": 0.006647589698437156, + "grad_norm": 1.8320390267243267, + "learning_rate": 2.215700660308144e-06, + "loss": 0.5499, + "step": 302 + }, + { + "epoch": 0.0066696015848558225, + "grad_norm": 3.529748769649405, + "learning_rate": 2.223037417461482e-06, + "loss": 0.5079, + "step": 303 + }, + { + "epoch": 0.006691613471274488, + "grad_norm": 2.3945632090867273, + "learning_rate": 2.2303741746148205e-06, + "loss": 0.471, + "step": 304 + }, + { + "epoch": 0.006713625357693155, + "grad_norm": 0.6898423786276648, + "learning_rate": 2.2377109317681586e-06, + "loss": 0.6572, + "step": 305 + }, + { + "epoch": 0.00673563724411182, + "grad_norm": 2.0249017829875506, + "learning_rate": 2.245047688921497e-06, + "loss": 0.5182, + "step": 306 + }, + { + "epoch": 0.006757649130530487, + "grad_norm": 2.33435699524806, + "learning_rate": 2.2523844460748353e-06, + "loss": 0.4479, + "step": 307 + }, + { + "epoch": 0.006779661016949152, + "grad_norm": 2.075647151504237, + "learning_rate": 2.2597212032281734e-06, + "loss": 0.5156, + "step": 308 + }, + { + "epoch": 0.006801672903367819, + "grad_norm": 2.1805318829786433, + "learning_rate": 2.2670579603815115e-06, + "loss": 0.4151, + "step": 309 + }, + { + "epoch": 0.0068236847897864845, + "grad_norm": 2.64988329620839, + "learning_rate": 2.27439471753485e-06, + "loss": 0.5807, + "step": 310 + }, + { + "epoch": 0.006845696676205151, + "grad_norm": 2.3073948754047247, + "learning_rate": 2.281731474688188e-06, + "loss": 0.5826, + "step": 311 + }, + { + "epoch": 0.006867708562623817, + "grad_norm": 8.080986857447485, + "learning_rate": 2.2890682318415263e-06, + "loss": 0.581, + "step": 312 + }, + { + "epoch": 0.006889720449042483, + "grad_norm": 0.7345621009832478, + "learning_rate": 2.2964049889948644e-06, + "loss": 0.7011, + "step": 313 + }, + { + "epoch": 0.006911732335461149, + "grad_norm": 2.042363707602837, + "learning_rate": 2.3037417461482025e-06, + "loss": 0.3312, + "step": 314 + }, + { + "epoch": 0.006933744221879815, + "grad_norm": 2.373158450454424, + "learning_rate": 2.311078503301541e-06, + "loss": 0.5341, + "step": 315 + }, + { + "epoch": 0.006955756108298481, + "grad_norm": 2.3446180112316046, + "learning_rate": 2.318415260454879e-06, + "loss": 0.4938, + "step": 316 + }, + { + "epoch": 0.006977767994717147, + "grad_norm": 2.349638745150719, + "learning_rate": 2.3257520176082172e-06, + "loss": 0.5736, + "step": 317 + }, + { + "epoch": 0.006999779881135813, + "grad_norm": 2.0508598567204444, + "learning_rate": 2.3330887747615553e-06, + "loss": 0.6326, + "step": 318 + }, + { + "epoch": 0.0070217917675544795, + "grad_norm": 2.3225914363819617, + "learning_rate": 2.340425531914894e-06, + "loss": 0.4427, + "step": 319 + }, + { + "epoch": 0.007043803653973145, + "grad_norm": 2.423277631259846, + "learning_rate": 2.347762289068232e-06, + "loss": 0.5381, + "step": 320 + }, + { + "epoch": 0.007065815540391812, + "grad_norm": 2.3683724275090494, + "learning_rate": 2.3550990462215705e-06, + "loss": 0.5647, + "step": 321 + }, + { + "epoch": 0.007087827426810477, + "grad_norm": 0.8355631759419746, + "learning_rate": 2.362435803374908e-06, + "loss": 0.6729, + "step": 322 + }, + { + "epoch": 0.007109839313229144, + "grad_norm": 2.7542767088562012, + "learning_rate": 2.3697725605282467e-06, + "loss": 0.4465, + "step": 323 + }, + { + "epoch": 0.007131851199647809, + "grad_norm": 2.0562496364447496, + "learning_rate": 2.377109317681585e-06, + "loss": 0.3947, + "step": 324 + }, + { + "epoch": 0.007153863086066476, + "grad_norm": 2.3140116544168, + "learning_rate": 2.3844460748349234e-06, + "loss": 0.4762, + "step": 325 + }, + { + "epoch": 0.007175874972485142, + "grad_norm": 2.01692865658088, + "learning_rate": 2.3917828319882615e-06, + "loss": 0.4676, + "step": 326 + }, + { + "epoch": 0.007197886858903808, + "grad_norm": 2.26299842800908, + "learning_rate": 2.3991195891415996e-06, + "loss": 0.4236, + "step": 327 + }, + { + "epoch": 0.0072198987453224745, + "grad_norm": 0.687001330611912, + "learning_rate": 2.4064563462949377e-06, + "loss": 0.6764, + "step": 328 + }, + { + "epoch": 0.00724191063174114, + "grad_norm": 2.5273037362004955, + "learning_rate": 2.4137931034482762e-06, + "loss": 0.5683, + "step": 329 + }, + { + "epoch": 0.007263922518159807, + "grad_norm": 2.0200611084772024, + "learning_rate": 2.4211298606016144e-06, + "loss": 0.5261, + "step": 330 + }, + { + "epoch": 0.007285934404578472, + "grad_norm": 1.957232829669494, + "learning_rate": 2.4284666177549525e-06, + "loss": 0.4881, + "step": 331 + }, + { + "epoch": 0.007307946290997139, + "grad_norm": 2.036821448781718, + "learning_rate": 2.4358033749082906e-06, + "loss": 0.4919, + "step": 332 + }, + { + "epoch": 0.007329958177415804, + "grad_norm": 0.7158754526842065, + "learning_rate": 2.443140132061629e-06, + "loss": 0.6815, + "step": 333 + }, + { + "epoch": 0.007351970063834471, + "grad_norm": 3.3536912232642258, + "learning_rate": 2.4504768892149672e-06, + "loss": 0.4279, + "step": 334 + }, + { + "epoch": 0.0073739819502531365, + "grad_norm": 2.380022278567027, + "learning_rate": 2.4578136463683053e-06, + "loss": 0.4166, + "step": 335 + }, + { + "epoch": 0.007395993836671803, + "grad_norm": 2.7893474023811886, + "learning_rate": 2.4651504035216434e-06, + "loss": 0.5522, + "step": 336 + }, + { + "epoch": 0.007418005723090469, + "grad_norm": 1.906429707707433, + "learning_rate": 2.4724871606749816e-06, + "loss": 0.5196, + "step": 337 + }, + { + "epoch": 0.007440017609509135, + "grad_norm": 0.7186571262860842, + "learning_rate": 2.47982391782832e-06, + "loss": 0.7131, + "step": 338 + }, + { + "epoch": 0.007462029495927801, + "grad_norm": 2.666783857182124, + "learning_rate": 2.487160674981658e-06, + "loss": 0.4644, + "step": 339 + }, + { + "epoch": 0.007484041382346467, + "grad_norm": 0.703013619779152, + "learning_rate": 2.4944974321349967e-06, + "loss": 0.669, + "step": 340 + }, + { + "epoch": 0.007506053268765133, + "grad_norm": 2.0879806220122004, + "learning_rate": 2.501834189288335e-06, + "loss": 0.4975, + "step": 341 + }, + { + "epoch": 0.007528065155183799, + "grad_norm": 2.164157527409554, + "learning_rate": 2.509170946441673e-06, + "loss": 0.5124, + "step": 342 + }, + { + "epoch": 0.007550077041602465, + "grad_norm": 2.521701169745473, + "learning_rate": 2.5165077035950115e-06, + "loss": 0.5002, + "step": 343 + }, + { + "epoch": 0.0075720889280211316, + "grad_norm": 2.0518364790367705, + "learning_rate": 2.523844460748349e-06, + "loss": 0.4598, + "step": 344 + }, + { + "epoch": 0.007594100814439797, + "grad_norm": 5.392422317540975, + "learning_rate": 2.5311812179016877e-06, + "loss": 0.4185, + "step": 345 + }, + { + "epoch": 0.007616112700858464, + "grad_norm": 2.8500515344899466, + "learning_rate": 2.538517975055026e-06, + "loss": 0.5152, + "step": 346 + }, + { + "epoch": 0.007638124587277129, + "grad_norm": 0.687150065254694, + "learning_rate": 2.5458547322083643e-06, + "loss": 0.6693, + "step": 347 + }, + { + "epoch": 0.007660136473695796, + "grad_norm": 1.855291399192172, + "learning_rate": 2.553191489361702e-06, + "loss": 0.4922, + "step": 348 + }, + { + "epoch": 0.0076821483601144614, + "grad_norm": 1.9204297448303278, + "learning_rate": 2.5605282465150406e-06, + "loss": 0.4904, + "step": 349 + }, + { + "epoch": 0.007704160246533128, + "grad_norm": 2.4171200518155582, + "learning_rate": 2.5678650036683787e-06, + "loss": 0.4537, + "step": 350 + }, + { + "epoch": 0.007726172132951794, + "grad_norm": 2.998132122092363, + "learning_rate": 2.575201760821717e-06, + "loss": 0.4566, + "step": 351 + }, + { + "epoch": 0.00774818401937046, + "grad_norm": 1.773844094412697, + "learning_rate": 2.582538517975055e-06, + "loss": 0.4344, + "step": 352 + }, + { + "epoch": 0.007770195905789126, + "grad_norm": 0.7034274547843036, + "learning_rate": 2.5898752751283934e-06, + "loss": 0.687, + "step": 353 + }, + { + "epoch": 0.007792207792207792, + "grad_norm": 2.6553934458227832, + "learning_rate": 2.597212032281732e-06, + "loss": 0.4667, + "step": 354 + }, + { + "epoch": 0.007814219678626458, + "grad_norm": 1.6901087625472666, + "learning_rate": 2.60454878943507e-06, + "loss": 0.4512, + "step": 355 + }, + { + "epoch": 0.007836231565045124, + "grad_norm": 3.0929098349435433, + "learning_rate": 2.611885546588408e-06, + "loss": 0.5299, + "step": 356 + }, + { + "epoch": 0.00785824345146379, + "grad_norm": 2.0320879075013325, + "learning_rate": 2.6192223037417463e-06, + "loss": 0.3954, + "step": 357 + }, + { + "epoch": 0.007880255337882457, + "grad_norm": 0.7159086434463131, + "learning_rate": 2.626559060895085e-06, + "loss": 0.6909, + "step": 358 + }, + { + "epoch": 0.007902267224301122, + "grad_norm": 2.463453917456427, + "learning_rate": 2.6338958180484225e-06, + "loss": 0.4951, + "step": 359 + }, + { + "epoch": 0.007924279110719789, + "grad_norm": 6.728560818305018, + "learning_rate": 2.641232575201761e-06, + "loss": 0.4977, + "step": 360 + }, + { + "epoch": 0.007946290997138455, + "grad_norm": 4.620160873560401, + "learning_rate": 2.648569332355099e-06, + "loss": 0.4775, + "step": 361 + }, + { + "epoch": 0.007968302883557122, + "grad_norm": 0.7860355967184924, + "learning_rate": 2.6559060895084377e-06, + "loss": 0.6624, + "step": 362 + }, + { + "epoch": 0.007990314769975786, + "grad_norm": 2.013459627326298, + "learning_rate": 2.6632428466617754e-06, + "loss": 0.5264, + "step": 363 + }, + { + "epoch": 0.008012326656394453, + "grad_norm": 2.361162286858791, + "learning_rate": 2.670579603815114e-06, + "loss": 0.5195, + "step": 364 + }, + { + "epoch": 0.00803433854281312, + "grad_norm": 2.7422386388136073, + "learning_rate": 2.677916360968452e-06, + "loss": 0.5188, + "step": 365 + }, + { + "epoch": 0.008056350429231786, + "grad_norm": 2.1790275448026915, + "learning_rate": 2.6852531181217905e-06, + "loss": 0.4709, + "step": 366 + }, + { + "epoch": 0.00807836231565045, + "grad_norm": 2.5120733375543858, + "learning_rate": 2.6925898752751282e-06, + "loss": 0.5144, + "step": 367 + }, + { + "epoch": 0.008100374202069117, + "grad_norm": 2.217082000432307, + "learning_rate": 2.6999266324284668e-06, + "loss": 0.5575, + "step": 368 + }, + { + "epoch": 0.008122386088487784, + "grad_norm": 2.456889653954885, + "learning_rate": 2.7072633895818053e-06, + "loss": 0.4313, + "step": 369 + }, + { + "epoch": 0.00814439797490645, + "grad_norm": 2.099204131447328, + "learning_rate": 2.7146001467351434e-06, + "loss": 0.5353, + "step": 370 + }, + { + "epoch": 0.008166409861325115, + "grad_norm": 2.53438028179099, + "learning_rate": 2.721936903888481e-06, + "loss": 0.4315, + "step": 371 + }, + { + "epoch": 0.008188421747743781, + "grad_norm": 2.1726046286980947, + "learning_rate": 2.7292736610418196e-06, + "loss": 0.4313, + "step": 372 + }, + { + "epoch": 0.008210433634162448, + "grad_norm": 2.2352819760063714, + "learning_rate": 2.736610418195158e-06, + "loss": 0.4594, + "step": 373 + }, + { + "epoch": 0.008232445520581114, + "grad_norm": 3.4748206422661414, + "learning_rate": 2.7439471753484963e-06, + "loss": 0.4409, + "step": 374 + }, + { + "epoch": 0.008254457406999779, + "grad_norm": 1.7366058667791893, + "learning_rate": 2.7512839325018344e-06, + "loss": 0.5108, + "step": 375 + }, + { + "epoch": 0.008276469293418446, + "grad_norm": 0.6772985915617693, + "learning_rate": 2.7586206896551725e-06, + "loss": 0.6789, + "step": 376 + }, + { + "epoch": 0.008298481179837112, + "grad_norm": 3.884540342760096, + "learning_rate": 2.765957446808511e-06, + "loss": 0.46, + "step": 377 + }, + { + "epoch": 0.008320493066255779, + "grad_norm": 2.1778941209706386, + "learning_rate": 2.773294203961849e-06, + "loss": 0.4811, + "step": 378 + }, + { + "epoch": 0.008342504952674443, + "grad_norm": 0.7085476567426933, + "learning_rate": 2.7806309611151872e-06, + "loss": 0.6977, + "step": 379 + }, + { + "epoch": 0.00836451683909311, + "grad_norm": 4.040425085240727, + "learning_rate": 2.7879677182685254e-06, + "loss": 0.5888, + "step": 380 + }, + { + "epoch": 0.008386528725511776, + "grad_norm": 2.3620780057060506, + "learning_rate": 2.795304475421864e-06, + "loss": 0.5415, + "step": 381 + }, + { + "epoch": 0.008408540611930443, + "grad_norm": 2.181474424255332, + "learning_rate": 2.8026412325752024e-06, + "loss": 0.4561, + "step": 382 + }, + { + "epoch": 0.008430552498349108, + "grad_norm": 0.6872878492800664, + "learning_rate": 2.80997798972854e-06, + "loss": 0.6891, + "step": 383 + }, + { + "epoch": 0.008452564384767774, + "grad_norm": 2.6411082545264812, + "learning_rate": 2.8173147468818786e-06, + "loss": 0.3798, + "step": 384 + }, + { + "epoch": 0.00847457627118644, + "grad_norm": 2.2620492004748907, + "learning_rate": 2.8246515040352168e-06, + "loss": 0.466, + "step": 385 + }, + { + "epoch": 0.008496588157605107, + "grad_norm": 2.0873089738652606, + "learning_rate": 2.8319882611885544e-06, + "loss": 0.5014, + "step": 386 + }, + { + "epoch": 0.008518600044023774, + "grad_norm": 2.3116087740151245, + "learning_rate": 2.839325018341893e-06, + "loss": 0.4162, + "step": 387 + }, + { + "epoch": 0.008540611930442438, + "grad_norm": 0.6898547736745562, + "learning_rate": 2.8466617754952315e-06, + "loss": 0.6632, + "step": 388 + }, + { + "epoch": 0.008562623816861105, + "grad_norm": 0.7043432583534848, + "learning_rate": 2.8539985326485696e-06, + "loss": 0.6486, + "step": 389 + }, + { + "epoch": 0.008584635703279771, + "grad_norm": 2.312464773137141, + "learning_rate": 2.8613352898019077e-06, + "loss": 0.4414, + "step": 390 + }, + { + "epoch": 0.008606647589698438, + "grad_norm": 0.6775839910331992, + "learning_rate": 2.868672046955246e-06, + "loss": 0.6914, + "step": 391 + }, + { + "epoch": 0.008628659476117103, + "grad_norm": 0.6783628482353694, + "learning_rate": 2.8760088041085844e-06, + "loss": 0.6708, + "step": 392 + }, + { + "epoch": 0.00865067136253577, + "grad_norm": 2.3857203517310444, + "learning_rate": 2.8833455612619225e-06, + "loss": 0.4973, + "step": 393 + }, + { + "epoch": 0.008672683248954436, + "grad_norm": 2.2308397023639635, + "learning_rate": 2.8906823184152606e-06, + "loss": 0.487, + "step": 394 + }, + { + "epoch": 0.008694695135373102, + "grad_norm": 4.257532077086737, + "learning_rate": 2.8980190755685987e-06, + "loss": 0.5308, + "step": 395 + }, + { + "epoch": 0.008716707021791767, + "grad_norm": 4.6615579275925665, + "learning_rate": 2.9053558327219372e-06, + "loss": 0.5394, + "step": 396 + }, + { + "epoch": 0.008738718908210433, + "grad_norm": 0.6771922506435569, + "learning_rate": 2.9126925898752758e-06, + "loss": 0.6695, + "step": 397 + }, + { + "epoch": 0.0087607307946291, + "grad_norm": 0.6693589904891113, + "learning_rate": 2.9200293470286135e-06, + "loss": 0.6967, + "step": 398 + }, + { + "epoch": 0.008782742681047766, + "grad_norm": 1.7668751075611753, + "learning_rate": 2.927366104181952e-06, + "loss": 0.5126, + "step": 399 + }, + { + "epoch": 0.008804754567466431, + "grad_norm": 2.2873059945393925, + "learning_rate": 2.93470286133529e-06, + "loss": 0.5056, + "step": 400 + }, + { + "epoch": 0.008826766453885098, + "grad_norm": 0.6773205779670173, + "learning_rate": 2.9420396184886286e-06, + "loss": 0.6929, + "step": 401 + }, + { + "epoch": 0.008848778340303764, + "grad_norm": 3.330563613248861, + "learning_rate": 2.9493763756419663e-06, + "loss": 0.5434, + "step": 402 + }, + { + "epoch": 0.00887079022672243, + "grad_norm": 2.634114315138519, + "learning_rate": 2.956713132795305e-06, + "loss": 0.4711, + "step": 403 + }, + { + "epoch": 0.008892802113141095, + "grad_norm": 2.061132086909879, + "learning_rate": 2.964049889948643e-06, + "loss": 0.4471, + "step": 404 + }, + { + "epoch": 0.008914813999559762, + "grad_norm": 4.484744585165306, + "learning_rate": 2.9713866471019815e-06, + "loss": 0.4571, + "step": 405 + }, + { + "epoch": 0.008936825885978428, + "grad_norm": 1.9372164839922061, + "learning_rate": 2.978723404255319e-06, + "loss": 0.4858, + "step": 406 + }, + { + "epoch": 0.008958837772397095, + "grad_norm": 2.165755670117837, + "learning_rate": 2.9860601614086577e-06, + "loss": 0.4675, + "step": 407 + }, + { + "epoch": 0.00898084965881576, + "grad_norm": 2.8276143315846864, + "learning_rate": 2.993396918561996e-06, + "loss": 0.4207, + "step": 408 + }, + { + "epoch": 0.009002861545234426, + "grad_norm": 2.0909144542345346, + "learning_rate": 3.000733675715334e-06, + "loss": 0.6007, + "step": 409 + }, + { + "epoch": 0.009024873431653093, + "grad_norm": 2.0741110306801605, + "learning_rate": 3.008070432868672e-06, + "loss": 0.4389, + "step": 410 + }, + { + "epoch": 0.00904688531807176, + "grad_norm": 0.6746278813393526, + "learning_rate": 3.0154071900220106e-06, + "loss": 0.6678, + "step": 411 + }, + { + "epoch": 0.009068897204490424, + "grad_norm": 2.205405475352074, + "learning_rate": 3.022743947175349e-06, + "loss": 0.4886, + "step": 412 + }, + { + "epoch": 0.00909090909090909, + "grad_norm": 1.928897974440472, + "learning_rate": 3.030080704328687e-06, + "loss": 0.4094, + "step": 413 + }, + { + "epoch": 0.009112920977327757, + "grad_norm": 3.740874091941628, + "learning_rate": 3.037417461482025e-06, + "loss": 0.4055, + "step": 414 + }, + { + "epoch": 0.009134932863746423, + "grad_norm": 1.9861258441009493, + "learning_rate": 3.0447542186353634e-06, + "loss": 0.5421, + "step": 415 + }, + { + "epoch": 0.00915694475016509, + "grad_norm": 2.1541832765382467, + "learning_rate": 3.052090975788702e-06, + "loss": 0.5201, + "step": 416 + }, + { + "epoch": 0.009178956636583755, + "grad_norm": 1.8423416176640515, + "learning_rate": 3.0594277329420397e-06, + "loss": 0.3908, + "step": 417 + }, + { + "epoch": 0.009200968523002421, + "grad_norm": 1.7454557251834886, + "learning_rate": 3.066764490095378e-06, + "loss": 0.4614, + "step": 418 + }, + { + "epoch": 0.009222980409421088, + "grad_norm": 3.068163598591084, + "learning_rate": 3.0741012472487163e-06, + "loss": 0.5041, + "step": 419 + }, + { + "epoch": 0.009244992295839754, + "grad_norm": 2.5328923673634267, + "learning_rate": 3.081438004402055e-06, + "loss": 0.4353, + "step": 420 + }, + { + "epoch": 0.009267004182258419, + "grad_norm": 2.0367760177472243, + "learning_rate": 3.0887747615553925e-06, + "loss": 0.4829, + "step": 421 + }, + { + "epoch": 0.009289016068677085, + "grad_norm": 3.0773865246458407, + "learning_rate": 3.096111518708731e-06, + "loss": 0.42, + "step": 422 + }, + { + "epoch": 0.009311027955095752, + "grad_norm": 1.7648122144082352, + "learning_rate": 3.103448275862069e-06, + "loss": 0.3878, + "step": 423 + }, + { + "epoch": 0.009333039841514418, + "grad_norm": 2.1785985733200284, + "learning_rate": 3.1107850330154077e-06, + "loss": 0.4591, + "step": 424 + }, + { + "epoch": 0.009355051727933083, + "grad_norm": 1.8396640973977725, + "learning_rate": 3.1181217901687454e-06, + "loss": 0.4657, + "step": 425 + }, + { + "epoch": 0.00937706361435175, + "grad_norm": 0.675271616703505, + "learning_rate": 3.125458547322084e-06, + "loss": 0.7013, + "step": 426 + }, + { + "epoch": 0.009399075500770416, + "grad_norm": 2.445714801807103, + "learning_rate": 3.1327953044754224e-06, + "loss": 0.4658, + "step": 427 + }, + { + "epoch": 0.009421087387189083, + "grad_norm": 1.8236797706634715, + "learning_rate": 3.1401320616287606e-06, + "loss": 0.3876, + "step": 428 + }, + { + "epoch": 0.009443099273607748, + "grad_norm": 2.0413265271038967, + "learning_rate": 3.1474688187820982e-06, + "loss": 0.4737, + "step": 429 + }, + { + "epoch": 0.009465111160026414, + "grad_norm": 2.15147054999003, + "learning_rate": 3.1548055759354368e-06, + "loss": 0.4537, + "step": 430 + }, + { + "epoch": 0.00948712304644508, + "grad_norm": 1.9986980733351212, + "learning_rate": 3.1621423330887753e-06, + "loss": 0.3371, + "step": 431 + }, + { + "epoch": 0.009509134932863747, + "grad_norm": 2.160479129198948, + "learning_rate": 3.169479090242113e-06, + "loss": 0.5237, + "step": 432 + }, + { + "epoch": 0.009531146819282412, + "grad_norm": 0.6582830018484703, + "learning_rate": 3.1768158473954515e-06, + "loss": 0.6581, + "step": 433 + }, + { + "epoch": 0.009553158705701078, + "grad_norm": 2.5741388887270915, + "learning_rate": 3.1841526045487896e-06, + "loss": 0.5439, + "step": 434 + }, + { + "epoch": 0.009575170592119745, + "grad_norm": 2.0593885158131173, + "learning_rate": 3.191489361702128e-06, + "loss": 0.4251, + "step": 435 + }, + { + "epoch": 0.009597182478538411, + "grad_norm": 3.668228810217631, + "learning_rate": 3.198826118855466e-06, + "loss": 0.4843, + "step": 436 + }, + { + "epoch": 0.009619194364957076, + "grad_norm": 0.6767383030414423, + "learning_rate": 3.2061628760088044e-06, + "loss": 0.6479, + "step": 437 + }, + { + "epoch": 0.009641206251375743, + "grad_norm": 2.864555510317399, + "learning_rate": 3.2134996331621425e-06, + "loss": 0.4777, + "step": 438 + }, + { + "epoch": 0.009663218137794409, + "grad_norm": 2.1728633683399186, + "learning_rate": 3.220836390315481e-06, + "loss": 0.4462, + "step": 439 + }, + { + "epoch": 0.009685230024213076, + "grad_norm": 2.087306871927884, + "learning_rate": 3.2281731474688187e-06, + "loss": 0.4445, + "step": 440 + }, + { + "epoch": 0.00970724191063174, + "grad_norm": 18.885901339488267, + "learning_rate": 3.2355099046221573e-06, + "loss": 0.4567, + "step": 441 + }, + { + "epoch": 0.009729253797050407, + "grad_norm": 3.4618489025587564, + "learning_rate": 3.2428466617754954e-06, + "loss": 0.408, + "step": 442 + }, + { + "epoch": 0.009751265683469073, + "grad_norm": 2.1834138011382676, + "learning_rate": 3.250183418928834e-06, + "loss": 0.3949, + "step": 443 + }, + { + "epoch": 0.00977327756988774, + "grad_norm": 0.6922623792620362, + "learning_rate": 3.2575201760821716e-06, + "loss": 0.6529, + "step": 444 + }, + { + "epoch": 0.009795289456306406, + "grad_norm": 2.0246737686307315, + "learning_rate": 3.26485693323551e-06, + "loss": 0.4292, + "step": 445 + }, + { + "epoch": 0.009817301342725071, + "grad_norm": 2.2663204149854272, + "learning_rate": 3.2721936903888487e-06, + "loss": 0.425, + "step": 446 + }, + { + "epoch": 0.009839313229143738, + "grad_norm": 2.0031457357588502, + "learning_rate": 3.2795304475421868e-06, + "loss": 0.4684, + "step": 447 + }, + { + "epoch": 0.009861325115562404, + "grad_norm": 2.23277075741281, + "learning_rate": 3.286867204695525e-06, + "loss": 0.5048, + "step": 448 + }, + { + "epoch": 0.00988333700198107, + "grad_norm": 3.6618440645948276, + "learning_rate": 3.294203961848863e-06, + "loss": 0.4375, + "step": 449 + }, + { + "epoch": 0.009905348888399735, + "grad_norm": 1.8544630772990849, + "learning_rate": 3.3015407190022015e-06, + "loss": 0.4511, + "step": 450 + }, + { + "epoch": 0.009927360774818402, + "grad_norm": 2.9274440795354164, + "learning_rate": 3.3088774761555396e-06, + "loss": 0.5209, + "step": 451 + }, + { + "epoch": 0.009949372661237068, + "grad_norm": 2.2526293520222875, + "learning_rate": 3.3162142333088777e-06, + "loss": 0.5639, + "step": 452 + }, + { + "epoch": 0.009971384547655735, + "grad_norm": 2.119944926996104, + "learning_rate": 3.323550990462216e-06, + "loss": 0.534, + "step": 453 + }, + { + "epoch": 0.0099933964340744, + "grad_norm": 1.6927576020318973, + "learning_rate": 3.3308877476155544e-06, + "loss": 0.4623, + "step": 454 + }, + { + "epoch": 0.010015408320493066, + "grad_norm": 2.1563766992897118, + "learning_rate": 3.338224504768892e-06, + "loss": 0.4355, + "step": 455 + }, + { + "epoch": 0.010037420206911733, + "grad_norm": 2.1617163214367006, + "learning_rate": 3.3455612619222306e-06, + "loss": 0.4278, + "step": 456 + }, + { + "epoch": 0.010059432093330399, + "grad_norm": 2.289572232290885, + "learning_rate": 3.3528980190755687e-06, + "loss": 0.4378, + "step": 457 + }, + { + "epoch": 0.010081443979749064, + "grad_norm": 6.929223048436192, + "learning_rate": 3.3602347762289072e-06, + "loss": 0.4959, + "step": 458 + }, + { + "epoch": 0.01010345586616773, + "grad_norm": 4.866870390246564, + "learning_rate": 3.367571533382245e-06, + "loss": 0.4357, + "step": 459 + }, + { + "epoch": 0.010125467752586397, + "grad_norm": 0.6726888643756666, + "learning_rate": 3.3749082905355835e-06, + "loss": 0.6418, + "step": 460 + }, + { + "epoch": 0.010147479639005063, + "grad_norm": 0.694624835219907, + "learning_rate": 3.382245047688922e-06, + "loss": 0.6573, + "step": 461 + }, + { + "epoch": 0.010169491525423728, + "grad_norm": 1.6540051143762258, + "learning_rate": 3.38958180484226e-06, + "loss": 0.3528, + "step": 462 + }, + { + "epoch": 0.010191503411842395, + "grad_norm": 1.8027097313788305, + "learning_rate": 3.396918561995598e-06, + "loss": 0.5697, + "step": 463 + }, + { + "epoch": 0.010213515298261061, + "grad_norm": 0.687215873168635, + "learning_rate": 3.4042553191489363e-06, + "loss": 0.6496, + "step": 464 + }, + { + "epoch": 0.010235527184679728, + "grad_norm": 2.0457789249788316, + "learning_rate": 3.411592076302275e-06, + "loss": 0.3734, + "step": 465 + }, + { + "epoch": 0.010257539071098392, + "grad_norm": 2.1230191959280638, + "learning_rate": 3.418928833455613e-06, + "loss": 0.4434, + "step": 466 + }, + { + "epoch": 0.010279550957517059, + "grad_norm": 1.7243133528562802, + "learning_rate": 3.426265590608951e-06, + "loss": 0.4392, + "step": 467 + }, + { + "epoch": 0.010301562843935725, + "grad_norm": 3.157950278575295, + "learning_rate": 3.433602347762289e-06, + "loss": 0.4472, + "step": 468 + }, + { + "epoch": 0.010323574730354392, + "grad_norm": 5.374728906910621, + "learning_rate": 3.4409391049156277e-06, + "loss": 0.5048, + "step": 469 + }, + { + "epoch": 0.010345586616773057, + "grad_norm": 2.3393558882953482, + "learning_rate": 3.448275862068966e-06, + "loss": 0.5137, + "step": 470 + }, + { + "epoch": 0.010367598503191723, + "grad_norm": 2.1249622191709983, + "learning_rate": 3.455612619222304e-06, + "loss": 0.435, + "step": 471 + }, + { + "epoch": 0.01038961038961039, + "grad_norm": 1.8970042396816633, + "learning_rate": 3.462949376375642e-06, + "loss": 0.4736, + "step": 472 + }, + { + "epoch": 0.010411622276029056, + "grad_norm": 0.6620056342312294, + "learning_rate": 3.4702861335289806e-06, + "loss": 0.6742, + "step": 473 + }, + { + "epoch": 0.010433634162447723, + "grad_norm": 2.8899779856250114, + "learning_rate": 3.477622890682319e-06, + "loss": 0.5906, + "step": 474 + }, + { + "epoch": 0.010455646048866387, + "grad_norm": 2.062110561438504, + "learning_rate": 3.484959647835657e-06, + "loss": 0.5422, + "step": 475 + }, + { + "epoch": 0.010477657935285054, + "grad_norm": 1.9714591428986237, + "learning_rate": 3.4922964049889953e-06, + "loss": 0.5028, + "step": 476 + }, + { + "epoch": 0.01049966982170372, + "grad_norm": 0.6858851722381558, + "learning_rate": 3.4996331621423334e-06, + "loss": 0.7076, + "step": 477 + }, + { + "epoch": 0.010521681708122387, + "grad_norm": 2.822464762621738, + "learning_rate": 3.506969919295672e-06, + "loss": 0.4681, + "step": 478 + }, + { + "epoch": 0.010543693594541052, + "grad_norm": 2.5707324548679455, + "learning_rate": 3.5143066764490097e-06, + "loss": 0.4557, + "step": 479 + }, + { + "epoch": 0.010565705480959718, + "grad_norm": 2.2272122456790013, + "learning_rate": 3.521643433602348e-06, + "loss": 0.4443, + "step": 480 + }, + { + "epoch": 0.010587717367378385, + "grad_norm": 1.7686240325601612, + "learning_rate": 3.5289801907556863e-06, + "loss": 0.4134, + "step": 481 + }, + { + "epoch": 0.010609729253797051, + "grad_norm": 2.633298253713975, + "learning_rate": 3.5363169479090244e-06, + "loss": 0.5444, + "step": 482 + }, + { + "epoch": 0.010631741140215716, + "grad_norm": 2.860286682370898, + "learning_rate": 3.5436537050623625e-06, + "loss": 0.427, + "step": 483 + }, + { + "epoch": 0.010653753026634382, + "grad_norm": 2.1713378608830163, + "learning_rate": 3.550990462215701e-06, + "loss": 0.4613, + "step": 484 + }, + { + "epoch": 0.010675764913053049, + "grad_norm": 2.057689354381901, + "learning_rate": 3.558327219369039e-06, + "loss": 0.4376, + "step": 485 + }, + { + "epoch": 0.010697776799471715, + "grad_norm": 2.1291723712590644, + "learning_rate": 3.5656639765223773e-06, + "loss": 0.5158, + "step": 486 + }, + { + "epoch": 0.01071978868589038, + "grad_norm": 1.9328175773876113, + "learning_rate": 3.5730007336757154e-06, + "loss": 0.4387, + "step": 487 + }, + { + "epoch": 0.010741800572309047, + "grad_norm": 0.7041037238152165, + "learning_rate": 3.580337490829054e-06, + "loss": 0.6696, + "step": 488 + }, + { + "epoch": 0.010763812458727713, + "grad_norm": 0.6779131903879964, + "learning_rate": 3.5876742479823925e-06, + "loss": 0.661, + "step": 489 + }, + { + "epoch": 0.01078582434514638, + "grad_norm": 2.259987785180117, + "learning_rate": 3.59501100513573e-06, + "loss": 0.4697, + "step": 490 + }, + { + "epoch": 0.010807836231565044, + "grad_norm": 2.0098512579372527, + "learning_rate": 3.6023477622890683e-06, + "loss": 0.5394, + "step": 491 + }, + { + "epoch": 0.010829848117983711, + "grad_norm": 2.2983548800728273, + "learning_rate": 3.6096845194424068e-06, + "loss": 0.529, + "step": 492 + }, + { + "epoch": 0.010851860004402377, + "grad_norm": 0.6621739658854875, + "learning_rate": 3.6170212765957453e-06, + "loss": 0.6614, + "step": 493 + }, + { + "epoch": 0.010873871890821044, + "grad_norm": 0.6549719219845017, + "learning_rate": 3.624358033749083e-06, + "loss": 0.6436, + "step": 494 + }, + { + "epoch": 0.010895883777239709, + "grad_norm": 2.7161074267914005, + "learning_rate": 3.6316947909024215e-06, + "loss": 0.516, + "step": 495 + }, + { + "epoch": 0.010917895663658375, + "grad_norm": 2.431605835675152, + "learning_rate": 3.6390315480557596e-06, + "loss": 0.373, + "step": 496 + }, + { + "epoch": 0.010939907550077042, + "grad_norm": 1.857593872754517, + "learning_rate": 3.646368305209098e-06, + "loss": 0.3963, + "step": 497 + }, + { + "epoch": 0.010961919436495708, + "grad_norm": 2.8391889905551158, + "learning_rate": 3.653705062362436e-06, + "loss": 0.3921, + "step": 498 + }, + { + "epoch": 0.010983931322914373, + "grad_norm": 2.128416238562181, + "learning_rate": 3.6610418195157744e-06, + "loss": 0.473, + "step": 499 + }, + { + "epoch": 0.01100594320933304, + "grad_norm": 1.9996218627075208, + "learning_rate": 3.6683785766691125e-06, + "loss": 0.532, + "step": 500 + }, + { + "epoch": 0.011027955095751706, + "grad_norm": 2.7380969718101555, + "learning_rate": 3.675715333822451e-06, + "loss": 0.3855, + "step": 501 + }, + { + "epoch": 0.011049966982170372, + "grad_norm": 2.28995147485516, + "learning_rate": 3.6830520909757887e-06, + "loss": 0.4376, + "step": 502 + }, + { + "epoch": 0.011071978868589039, + "grad_norm": 1.9129696887834748, + "learning_rate": 3.6903888481291273e-06, + "loss": 0.391, + "step": 503 + }, + { + "epoch": 0.011093990755007704, + "grad_norm": 2.469647823775814, + "learning_rate": 3.697725605282466e-06, + "loss": 0.4697, + "step": 504 + }, + { + "epoch": 0.01111600264142637, + "grad_norm": 2.093427921354592, + "learning_rate": 3.7050623624358035e-06, + "loss": 0.3861, + "step": 505 + }, + { + "epoch": 0.011138014527845037, + "grad_norm": 2.6403178461403827, + "learning_rate": 3.7123991195891416e-06, + "loss": 0.5379, + "step": 506 + }, + { + "epoch": 0.011160026414263703, + "grad_norm": 2.2807416402534066, + "learning_rate": 3.71973587674248e-06, + "loss": 0.4787, + "step": 507 + }, + { + "epoch": 0.011182038300682368, + "grad_norm": 0.6610416997348149, + "learning_rate": 3.7270726338958187e-06, + "loss": 0.6693, + "step": 508 + }, + { + "epoch": 0.011204050187101034, + "grad_norm": 2.0482251489028274, + "learning_rate": 3.7344093910491563e-06, + "loss": 0.4247, + "step": 509 + }, + { + "epoch": 0.011226062073519701, + "grad_norm": 0.6917597124144349, + "learning_rate": 3.741746148202495e-06, + "loss": 0.6595, + "step": 510 + }, + { + "epoch": 0.011248073959938367, + "grad_norm": 2.1994879515808368, + "learning_rate": 3.749082905355833e-06, + "loss": 0.5428, + "step": 511 + }, + { + "epoch": 0.011270085846357032, + "grad_norm": 1.8285870516083587, + "learning_rate": 3.7564196625091715e-06, + "loss": 0.3585, + "step": 512 + }, + { + "epoch": 0.011292097732775699, + "grad_norm": 2.6451292073233206, + "learning_rate": 3.763756419662509e-06, + "loss": 0.3611, + "step": 513 + }, + { + "epoch": 0.011314109619194365, + "grad_norm": 3.7106995147564388, + "learning_rate": 3.7710931768158477e-06, + "loss": 0.4095, + "step": 514 + }, + { + "epoch": 0.011336121505613032, + "grad_norm": 2.8522057317696703, + "learning_rate": 3.778429933969186e-06, + "loss": 0.3816, + "step": 515 + }, + { + "epoch": 0.011358133392031696, + "grad_norm": 0.6607994180271936, + "learning_rate": 3.7857666911225244e-06, + "loss": 0.6623, + "step": 516 + }, + { + "epoch": 0.011380145278450363, + "grad_norm": 2.3908680291947535, + "learning_rate": 3.793103448275862e-06, + "loss": 0.3827, + "step": 517 + }, + { + "epoch": 0.01140215716486903, + "grad_norm": 2.1697893104525736, + "learning_rate": 3.8004402054292006e-06, + "loss": 0.4225, + "step": 518 + }, + { + "epoch": 0.011424169051287696, + "grad_norm": 14.48131855641422, + "learning_rate": 3.8077769625825387e-06, + "loss": 0.4037, + "step": 519 + }, + { + "epoch": 0.01144618093770636, + "grad_norm": 0.6833099915372118, + "learning_rate": 3.815113719735877e-06, + "loss": 0.6482, + "step": 520 + }, + { + "epoch": 0.011468192824125027, + "grad_norm": 2.0376290288255436, + "learning_rate": 3.822450476889215e-06, + "loss": 0.4494, + "step": 521 + }, + { + "epoch": 0.011490204710543694, + "grad_norm": 2.1385055098238595, + "learning_rate": 3.8297872340425535e-06, + "loss": 0.4961, + "step": 522 + }, + { + "epoch": 0.01151221659696236, + "grad_norm": 1.50915006554951, + "learning_rate": 3.837123991195892e-06, + "loss": 0.3837, + "step": 523 + }, + { + "epoch": 0.011534228483381025, + "grad_norm": 2.8960459895129538, + "learning_rate": 3.8444607483492305e-06, + "loss": 0.4381, + "step": 524 + }, + { + "epoch": 0.011556240369799691, + "grad_norm": 1.9392272924875653, + "learning_rate": 3.851797505502568e-06, + "loss": 0.5087, + "step": 525 + }, + { + "epoch": 0.011578252256218358, + "grad_norm": 2.1114379266059964, + "learning_rate": 3.859134262655907e-06, + "loss": 0.4363, + "step": 526 + }, + { + "epoch": 0.011600264142637024, + "grad_norm": 2.269546695757783, + "learning_rate": 3.8664710198092444e-06, + "loss": 0.3821, + "step": 527 + }, + { + "epoch": 0.01162227602905569, + "grad_norm": 4.213717984859181, + "learning_rate": 3.873807776962583e-06, + "loss": 0.5184, + "step": 528 + }, + { + "epoch": 0.011644287915474356, + "grad_norm": 0.6810219343625664, + "learning_rate": 3.881144534115921e-06, + "loss": 0.6627, + "step": 529 + }, + { + "epoch": 0.011666299801893022, + "grad_norm": 2.000401077452117, + "learning_rate": 3.888481291269259e-06, + "loss": 0.4741, + "step": 530 + }, + { + "epoch": 0.011688311688311689, + "grad_norm": 0.6686396181788647, + "learning_rate": 3.895818048422598e-06, + "loss": 0.6596, + "step": 531 + }, + { + "epoch": 0.011710323574730355, + "grad_norm": 0.6946137228758347, + "learning_rate": 3.903154805575935e-06, + "loss": 0.7076, + "step": 532 + }, + { + "epoch": 0.01173233546114902, + "grad_norm": 2.149573010989446, + "learning_rate": 3.910491562729274e-06, + "loss": 0.4488, + "step": 533 + }, + { + "epoch": 0.011754347347567686, + "grad_norm": 3.2259719734957373, + "learning_rate": 3.9178283198826125e-06, + "loss": 0.4503, + "step": 534 + }, + { + "epoch": 0.011776359233986353, + "grad_norm": 0.655259984659527, + "learning_rate": 3.925165077035951e-06, + "loss": 0.6879, + "step": 535 + }, + { + "epoch": 0.01179837112040502, + "grad_norm": 1.761318886724906, + "learning_rate": 3.932501834189289e-06, + "loss": 0.4042, + "step": 536 + }, + { + "epoch": 0.011820383006823684, + "grad_norm": 2.0744450721842225, + "learning_rate": 3.939838591342626e-06, + "loss": 0.4397, + "step": 537 + }, + { + "epoch": 0.01184239489324235, + "grad_norm": 2.2424488962344125, + "learning_rate": 3.947175348495965e-06, + "loss": 0.5137, + "step": 538 + }, + { + "epoch": 0.011864406779661017, + "grad_norm": 0.679013495731472, + "learning_rate": 3.9545121056493035e-06, + "loss": 0.6604, + "step": 539 + }, + { + "epoch": 0.011886418666079684, + "grad_norm": 2.07800089459439, + "learning_rate": 3.961848862802641e-06, + "loss": 0.5295, + "step": 540 + }, + { + "epoch": 0.011908430552498349, + "grad_norm": 1.8463496990602302, + "learning_rate": 3.96918561995598e-06, + "loss": 0.4032, + "step": 541 + }, + { + "epoch": 0.011930442438917015, + "grad_norm": 2.9552776662349785, + "learning_rate": 3.976522377109318e-06, + "loss": 0.5347, + "step": 542 + }, + { + "epoch": 0.011952454325335682, + "grad_norm": 2.4482751288167504, + "learning_rate": 3.983859134262657e-06, + "loss": 0.4949, + "step": 543 + }, + { + "epoch": 0.011974466211754348, + "grad_norm": 3.2082951208892267, + "learning_rate": 3.9911958914159944e-06, + "loss": 0.4036, + "step": 544 + }, + { + "epoch": 0.011996478098173013, + "grad_norm": 0.6884318893454237, + "learning_rate": 3.998532648569333e-06, + "loss": 0.6799, + "step": 545 + }, + { + "epoch": 0.01201848998459168, + "grad_norm": 2.3224541219835806, + "learning_rate": 4.005869405722671e-06, + "loss": 0.374, + "step": 546 + }, + { + "epoch": 0.012040501871010346, + "grad_norm": 2.3432387691531438, + "learning_rate": 4.013206162876009e-06, + "loss": 0.4402, + "step": 547 + }, + { + "epoch": 0.012062513757429012, + "grad_norm": 2.1086543292051836, + "learning_rate": 4.020542920029347e-06, + "loss": 0.4554, + "step": 548 + }, + { + "epoch": 0.012084525643847677, + "grad_norm": 1.8283465438733886, + "learning_rate": 4.027879677182685e-06, + "loss": 0.3707, + "step": 549 + }, + { + "epoch": 0.012106537530266344, + "grad_norm": 2.264686719497075, + "learning_rate": 4.035216434336024e-06, + "loss": 0.4511, + "step": 550 + }, + { + "epoch": 0.01212854941668501, + "grad_norm": 2.157620260672143, + "learning_rate": 4.042553191489362e-06, + "loss": 0.5032, + "step": 551 + }, + { + "epoch": 0.012150561303103677, + "grad_norm": 3.0005087507903387, + "learning_rate": 4.0498899486427e-06, + "loss": 0.4036, + "step": 552 + }, + { + "epoch": 0.012172573189522341, + "grad_norm": 0.6605142719739748, + "learning_rate": 4.057226705796039e-06, + "loss": 0.6407, + "step": 553 + }, + { + "epoch": 0.012194585075941008, + "grad_norm": 3.1245287235118964, + "learning_rate": 4.064563462949377e-06, + "loss": 0.3905, + "step": 554 + }, + { + "epoch": 0.012216596962359674, + "grad_norm": 2.0415659363979515, + "learning_rate": 4.071900220102715e-06, + "loss": 0.3307, + "step": 555 + }, + { + "epoch": 0.01223860884877834, + "grad_norm": 2.07344928159212, + "learning_rate": 4.0792369772560534e-06, + "loss": 0.5107, + "step": 556 + }, + { + "epoch": 0.012260620735197006, + "grad_norm": 2.3460120281065855, + "learning_rate": 4.086573734409391e-06, + "loss": 0.4166, + "step": 557 + }, + { + "epoch": 0.012282632621615672, + "grad_norm": 1.8203808595109947, + "learning_rate": 4.09391049156273e-06, + "loss": 0.4862, + "step": 558 + }, + { + "epoch": 0.012304644508034339, + "grad_norm": 3.7680233546721627, + "learning_rate": 4.101247248716067e-06, + "loss": 0.4036, + "step": 559 + }, + { + "epoch": 0.012326656394453005, + "grad_norm": 8.258139676503088, + "learning_rate": 4.108584005869406e-06, + "loss": 0.4593, + "step": 560 + }, + { + "epoch": 0.012348668280871672, + "grad_norm": 2.2449156221831816, + "learning_rate": 4.115920763022744e-06, + "loss": 0.4665, + "step": 561 + }, + { + "epoch": 0.012370680167290336, + "grad_norm": 1.9406645233396005, + "learning_rate": 4.123257520176083e-06, + "loss": 0.5326, + "step": 562 + }, + { + "epoch": 0.012392692053709003, + "grad_norm": 2.272924815165241, + "learning_rate": 4.130594277329421e-06, + "loss": 0.4414, + "step": 563 + }, + { + "epoch": 0.01241470394012767, + "grad_norm": 2.2173441789021204, + "learning_rate": 4.137931034482759e-06, + "loss": 0.488, + "step": 564 + }, + { + "epoch": 0.012436715826546336, + "grad_norm": 2.022020914954592, + "learning_rate": 4.145267791636097e-06, + "loss": 0.5043, + "step": 565 + }, + { + "epoch": 0.012458727712965, + "grad_norm": 1.9006942446724606, + "learning_rate": 4.152604548789435e-06, + "loss": 0.5099, + "step": 566 + }, + { + "epoch": 0.012480739599383667, + "grad_norm": 2.3906750082783903, + "learning_rate": 4.159941305942773e-06, + "loss": 0.3854, + "step": 567 + }, + { + "epoch": 0.012502751485802334, + "grad_norm": 2.0040533755987036, + "learning_rate": 4.167278063096112e-06, + "loss": 0.4197, + "step": 568 + }, + { + "epoch": 0.012524763372221, + "grad_norm": 2.2524453246366494, + "learning_rate": 4.17461482024945e-06, + "loss": 0.3736, + "step": 569 + }, + { + "epoch": 0.012546775258639665, + "grad_norm": 0.6798744071806412, + "learning_rate": 4.181951577402789e-06, + "loss": 0.6917, + "step": 570 + }, + { + "epoch": 0.012568787145058331, + "grad_norm": 2.7947767411874938, + "learning_rate": 4.189288334556126e-06, + "loss": 0.4738, + "step": 571 + }, + { + "epoch": 0.012590799031476998, + "grad_norm": 2.0624397344679806, + "learning_rate": 4.196625091709465e-06, + "loss": 0.4362, + "step": 572 + }, + { + "epoch": 0.012612810917895664, + "grad_norm": 2.712375608300836, + "learning_rate": 4.203961848862803e-06, + "loss": 0.4654, + "step": 573 + }, + { + "epoch": 0.012634822804314329, + "grad_norm": 3.3353786578411873, + "learning_rate": 4.211298606016141e-06, + "loss": 0.4101, + "step": 574 + }, + { + "epoch": 0.012656834690732996, + "grad_norm": 2.4590530101247023, + "learning_rate": 4.21863536316948e-06, + "loss": 0.4781, + "step": 575 + }, + { + "epoch": 0.012678846577151662, + "grad_norm": 2.54829453855078, + "learning_rate": 4.225972120322817e-06, + "loss": 0.4798, + "step": 576 + }, + { + "epoch": 0.012700858463570329, + "grad_norm": 2.3247988250752396, + "learning_rate": 4.233308877476156e-06, + "loss": 0.4619, + "step": 577 + }, + { + "epoch": 0.012722870349988993, + "grad_norm": 0.6770568459653074, + "learning_rate": 4.2406456346294935e-06, + "loss": 0.6383, + "step": 578 + }, + { + "epoch": 0.01274488223640766, + "grad_norm": 1.9112155493704828, + "learning_rate": 4.247982391782832e-06, + "loss": 0.3292, + "step": 579 + }, + { + "epoch": 0.012766894122826326, + "grad_norm": 1.9810431475742412, + "learning_rate": 4.255319148936171e-06, + "loss": 0.5272, + "step": 580 + }, + { + "epoch": 0.012788906009244993, + "grad_norm": 1.8383454531658565, + "learning_rate": 4.262655906089509e-06, + "loss": 0.4741, + "step": 581 + }, + { + "epoch": 0.012810917895663658, + "grad_norm": 1.9961339234845301, + "learning_rate": 4.269992663242847e-06, + "loss": 0.457, + "step": 582 + }, + { + "epoch": 0.012832929782082324, + "grad_norm": 2.0167565005331185, + "learning_rate": 4.277329420396185e-06, + "loss": 0.4808, + "step": 583 + }, + { + "epoch": 0.01285494166850099, + "grad_norm": 2.565928006216234, + "learning_rate": 4.284666177549524e-06, + "loss": 0.4416, + "step": 584 + }, + { + "epoch": 0.012876953554919657, + "grad_norm": 2.180207003463902, + "learning_rate": 4.292002934702862e-06, + "loss": 0.4559, + "step": 585 + }, + { + "epoch": 0.012898965441338322, + "grad_norm": 2.557356591485481, + "learning_rate": 4.2993396918562e-06, + "loss": 0.5173, + "step": 586 + }, + { + "epoch": 0.012920977327756988, + "grad_norm": 0.6744463824062763, + "learning_rate": 4.306676449009538e-06, + "loss": 0.6576, + "step": 587 + }, + { + "epoch": 0.012942989214175655, + "grad_norm": 15.355491521368617, + "learning_rate": 4.314013206162876e-06, + "loss": 0.4206, + "step": 588 + }, + { + "epoch": 0.012965001100594321, + "grad_norm": 2.348691532808759, + "learning_rate": 4.321349963316215e-06, + "loss": 0.4556, + "step": 589 + }, + { + "epoch": 0.012987012987012988, + "grad_norm": 2.0052955733423454, + "learning_rate": 4.3286867204695526e-06, + "loss": 0.5274, + "step": 590 + }, + { + "epoch": 0.013009024873431653, + "grad_norm": 2.9186768737671356, + "learning_rate": 4.336023477622891e-06, + "loss": 0.4567, + "step": 591 + }, + { + "epoch": 0.013031036759850319, + "grad_norm": 2.088215167206346, + "learning_rate": 4.34336023477623e-06, + "loss": 0.4392, + "step": 592 + }, + { + "epoch": 0.013053048646268986, + "grad_norm": 2.0648766487557495, + "learning_rate": 4.350696991929568e-06, + "loss": 0.5007, + "step": 593 + }, + { + "epoch": 0.013075060532687652, + "grad_norm": 2.3602551889362564, + "learning_rate": 4.358033749082906e-06, + "loss": 0.3842, + "step": 594 + }, + { + "epoch": 0.013097072419106317, + "grad_norm": 2.2912742107442337, + "learning_rate": 4.3653705062362435e-06, + "loss": 0.4135, + "step": 595 + }, + { + "epoch": 0.013119084305524983, + "grad_norm": 4.363277674367627, + "learning_rate": 4.372707263389582e-06, + "loss": 0.3396, + "step": 596 + }, + { + "epoch": 0.01314109619194365, + "grad_norm": 0.699865377540117, + "learning_rate": 4.380044020542921e-06, + "loss": 0.683, + "step": 597 + }, + { + "epoch": 0.013163108078362316, + "grad_norm": 1.8243908085529819, + "learning_rate": 4.387380777696258e-06, + "loss": 0.3976, + "step": 598 + }, + { + "epoch": 0.013185119964780981, + "grad_norm": 1.9028758838448743, + "learning_rate": 4.394717534849597e-06, + "loss": 0.4024, + "step": 599 + }, + { + "epoch": 0.013207131851199648, + "grad_norm": 0.6656522952032994, + "learning_rate": 4.402054292002935e-06, + "loss": 0.6516, + "step": 600 + }, + { + "epoch": 0.013229143737618314, + "grad_norm": 2.1702800709876158, + "learning_rate": 4.409391049156273e-06, + "loss": 0.4095, + "step": 601 + }, + { + "epoch": 0.01325115562403698, + "grad_norm": 4.869836425510219, + "learning_rate": 4.4167278063096116e-06, + "loss": 0.4905, + "step": 602 + }, + { + "epoch": 0.013273167510455645, + "grad_norm": 7.205404828046775, + "learning_rate": 4.42406456346295e-06, + "loss": 0.4688, + "step": 603 + }, + { + "epoch": 0.013295179396874312, + "grad_norm": 1.9073476816268613, + "learning_rate": 4.431401320616288e-06, + "loss": 0.413, + "step": 604 + }, + { + "epoch": 0.013317191283292978, + "grad_norm": 2.175945831530418, + "learning_rate": 4.438738077769626e-06, + "loss": 0.5177, + "step": 605 + }, + { + "epoch": 0.013339203169711645, + "grad_norm": 1.938473368241419, + "learning_rate": 4.446074834922964e-06, + "loss": 0.4749, + "step": 606 + }, + { + "epoch": 0.01336121505613031, + "grad_norm": 2.5489718188056716, + "learning_rate": 4.4534115920763025e-06, + "loss": 0.4871, + "step": 607 + }, + { + "epoch": 0.013383226942548976, + "grad_norm": 2.7478041119526755, + "learning_rate": 4.460748349229641e-06, + "loss": 0.3696, + "step": 608 + }, + { + "epoch": 0.013405238828967643, + "grad_norm": 2.6136121882165195, + "learning_rate": 4.468085106382979e-06, + "loss": 0.4561, + "step": 609 + }, + { + "epoch": 0.01342725071538631, + "grad_norm": 2.2264143452383305, + "learning_rate": 4.475421863536317e-06, + "loss": 0.4831, + "step": 610 + }, + { + "epoch": 0.013449262601804974, + "grad_norm": 1.9774524214936187, + "learning_rate": 4.482758620689656e-06, + "loss": 0.4172, + "step": 611 + }, + { + "epoch": 0.01347127448822364, + "grad_norm": 2.663501948318877, + "learning_rate": 4.490095377842994e-06, + "loss": 0.3859, + "step": 612 + }, + { + "epoch": 0.013493286374642307, + "grad_norm": 2.0912265488063686, + "learning_rate": 4.497432134996332e-06, + "loss": 0.4183, + "step": 613 + }, + { + "epoch": 0.013515298261060973, + "grad_norm": 2.196518307668095, + "learning_rate": 4.504768892149671e-06, + "loss": 0.3847, + "step": 614 + }, + { + "epoch": 0.013537310147479638, + "grad_norm": 0.6690322623860633, + "learning_rate": 4.512105649303008e-06, + "loss": 0.6366, + "step": 615 + }, + { + "epoch": 0.013559322033898305, + "grad_norm": 1.7649444499473512, + "learning_rate": 4.519442406456347e-06, + "loss": 0.3714, + "step": 616 + }, + { + "epoch": 0.013581333920316971, + "grad_norm": 2.331718726605639, + "learning_rate": 4.5267791636096845e-06, + "loss": 0.4509, + "step": 617 + }, + { + "epoch": 0.013603345806735638, + "grad_norm": 0.6623859548467096, + "learning_rate": 4.534115920763023e-06, + "loss": 0.6769, + "step": 618 + }, + { + "epoch": 0.013625357693154304, + "grad_norm": 2.182175294210041, + "learning_rate": 4.5414526779163616e-06, + "loss": 0.4729, + "step": 619 + }, + { + "epoch": 0.013647369579572969, + "grad_norm": 3.0457945453733863, + "learning_rate": 4.5487894350697e-06, + "loss": 0.3169, + "step": 620 + }, + { + "epoch": 0.013669381465991635, + "grad_norm": 2.105039462648622, + "learning_rate": 4.556126192223038e-06, + "loss": 0.402, + "step": 621 + }, + { + "epoch": 0.013691393352410302, + "grad_norm": 2.214792263832393, + "learning_rate": 4.563462949376376e-06, + "loss": 0.4843, + "step": 622 + }, + { + "epoch": 0.013713405238828968, + "grad_norm": 2.731723336440893, + "learning_rate": 4.570799706529714e-06, + "loss": 0.4446, + "step": 623 + }, + { + "epoch": 0.013735417125247633, + "grad_norm": 6.666915037721983, + "learning_rate": 4.5781364636830525e-06, + "loss": 0.4823, + "step": 624 + }, + { + "epoch": 0.0137574290116663, + "grad_norm": 0.685239007990456, + "learning_rate": 4.58547322083639e-06, + "loss": 0.6431, + "step": 625 + }, + { + "epoch": 0.013779440898084966, + "grad_norm": 2.0519669705089036, + "learning_rate": 4.592809977989729e-06, + "loss": 0.3418, + "step": 626 + }, + { + "epoch": 0.013801452784503633, + "grad_norm": 0.7010659478682686, + "learning_rate": 4.600146735143067e-06, + "loss": 0.6703, + "step": 627 + }, + { + "epoch": 0.013823464670922297, + "grad_norm": 2.699959526891051, + "learning_rate": 4.607483492296405e-06, + "loss": 0.4934, + "step": 628 + }, + { + "epoch": 0.013845476557340964, + "grad_norm": 2.07753993754311, + "learning_rate": 4.6148202494497435e-06, + "loss": 0.4281, + "step": 629 + }, + { + "epoch": 0.01386748844375963, + "grad_norm": 2.221333647709962, + "learning_rate": 4.622157006603082e-06, + "loss": 0.4749, + "step": 630 + }, + { + "epoch": 0.013889500330178297, + "grad_norm": 1.9560404331034216, + "learning_rate": 4.6294937637564206e-06, + "loss": 0.4376, + "step": 631 + }, + { + "epoch": 0.013911512216596962, + "grad_norm": 2.248313136923709, + "learning_rate": 4.636830520909758e-06, + "loss": 0.3994, + "step": 632 + }, + { + "epoch": 0.013933524103015628, + "grad_norm": 1.7783835535563837, + "learning_rate": 4.644167278063097e-06, + "loss": 0.4106, + "step": 633 + }, + { + "epoch": 0.013955535989434295, + "grad_norm": 1.9933242846612667, + "learning_rate": 4.6515040352164345e-06, + "loss": 0.3556, + "step": 634 + }, + { + "epoch": 0.013977547875852961, + "grad_norm": 2.0040967851354203, + "learning_rate": 4.658840792369773e-06, + "loss": 0.5324, + "step": 635 + }, + { + "epoch": 0.013999559762271626, + "grad_norm": 0.6454580173378712, + "learning_rate": 4.666177549523111e-06, + "loss": 0.6658, + "step": 636 + }, + { + "epoch": 0.014021571648690292, + "grad_norm": 2.32748320428085, + "learning_rate": 4.673514306676449e-06, + "loss": 0.4587, + "step": 637 + }, + { + "epoch": 0.014043583535108959, + "grad_norm": 2.6389294518012236, + "learning_rate": 4.680851063829788e-06, + "loss": 0.3578, + "step": 638 + }, + { + "epoch": 0.014065595421527625, + "grad_norm": 1.9577106319184978, + "learning_rate": 4.688187820983126e-06, + "loss": 0.4405, + "step": 639 + }, + { + "epoch": 0.01408760730794629, + "grad_norm": 2.805491056685606, + "learning_rate": 4.695524578136464e-06, + "loss": 0.4485, + "step": 640 + }, + { + "epoch": 0.014109619194364957, + "grad_norm": 2.04769584964521, + "learning_rate": 4.7028613352898025e-06, + "loss": 0.3565, + "step": 641 + }, + { + "epoch": 0.014131631080783623, + "grad_norm": 2.592254759747205, + "learning_rate": 4.710198092443141e-06, + "loss": 0.4957, + "step": 642 + }, + { + "epoch": 0.01415364296720229, + "grad_norm": 0.6549061816535181, + "learning_rate": 4.717534849596479e-06, + "loss": 0.619, + "step": 643 + }, + { + "epoch": 0.014175654853620955, + "grad_norm": 1.9736353590595679, + "learning_rate": 4.724871606749816e-06, + "loss": 0.4394, + "step": 644 + }, + { + "epoch": 0.014197666740039621, + "grad_norm": 2.520686759018818, + "learning_rate": 4.732208363903155e-06, + "loss": 0.4691, + "step": 645 + }, + { + "epoch": 0.014219678626458288, + "grad_norm": 2.5423407550524244, + "learning_rate": 4.7395451210564935e-06, + "loss": 0.4409, + "step": 646 + }, + { + "epoch": 0.014241690512876954, + "grad_norm": 8.642126653955396, + "learning_rate": 4.746881878209831e-06, + "loss": 0.4255, + "step": 647 + }, + { + "epoch": 0.014263702399295619, + "grad_norm": 0.6863001926748536, + "learning_rate": 4.75421863536317e-06, + "loss": 0.6707, + "step": 648 + }, + { + "epoch": 0.014285714285714285, + "grad_norm": 3.7153693202265647, + "learning_rate": 4.761555392516508e-06, + "loss": 0.4493, + "step": 649 + }, + { + "epoch": 0.014307726172132952, + "grad_norm": 1.8612066792981967, + "learning_rate": 4.768892149669847e-06, + "loss": 0.4462, + "step": 650 + }, + { + "epoch": 0.014329738058551618, + "grad_norm": 2.304949456822317, + "learning_rate": 4.7762289068231845e-06, + "loss": 0.4432, + "step": 651 + }, + { + "epoch": 0.014351749944970285, + "grad_norm": 1.97881088300149, + "learning_rate": 4.783565663976523e-06, + "loss": 0.4156, + "step": 652 + }, + { + "epoch": 0.01437376183138895, + "grad_norm": 2.2760262084109266, + "learning_rate": 4.790902421129861e-06, + "loss": 0.4859, + "step": 653 + }, + { + "epoch": 0.014395773717807616, + "grad_norm": 1.8861320228016314, + "learning_rate": 4.798239178283199e-06, + "loss": 0.3742, + "step": 654 + }, + { + "epoch": 0.014417785604226283, + "grad_norm": 6.149330076801007, + "learning_rate": 4.805575935436537e-06, + "loss": 0.4767, + "step": 655 + }, + { + "epoch": 0.014439797490644949, + "grad_norm": 1.5962827789591147, + "learning_rate": 4.8129126925898754e-06, + "loss": 0.406, + "step": 656 + }, + { + "epoch": 0.014461809377063614, + "grad_norm": 2.631247724512857, + "learning_rate": 4.820249449743214e-06, + "loss": 0.4965, + "step": 657 + }, + { + "epoch": 0.01448382126348228, + "grad_norm": 1.717267455992544, + "learning_rate": 4.8275862068965525e-06, + "loss": 0.4844, + "step": 658 + }, + { + "epoch": 0.014505833149900947, + "grad_norm": 2.0437161551345353, + "learning_rate": 4.83492296404989e-06, + "loss": 0.3664, + "step": 659 + }, + { + "epoch": 0.014527845036319613, + "grad_norm": 2.5587891386290886, + "learning_rate": 4.842259721203229e-06, + "loss": 0.4348, + "step": 660 + }, + { + "epoch": 0.014549856922738278, + "grad_norm": 2.477967429651576, + "learning_rate": 4.849596478356567e-06, + "loss": 0.4423, + "step": 661 + }, + { + "epoch": 0.014571868809156945, + "grad_norm": 2.226770155345193, + "learning_rate": 4.856933235509905e-06, + "loss": 0.4589, + "step": 662 + }, + { + "epoch": 0.014593880695575611, + "grad_norm": 1.9886922223085737, + "learning_rate": 4.8642699926632435e-06, + "loss": 0.3781, + "step": 663 + }, + { + "epoch": 0.014615892581994278, + "grad_norm": 1.7656197769363753, + "learning_rate": 4.871606749816581e-06, + "loss": 0.3405, + "step": 664 + }, + { + "epoch": 0.014637904468412942, + "grad_norm": 2.1864661325741888, + "learning_rate": 4.87894350696992e-06, + "loss": 0.4466, + "step": 665 + }, + { + "epoch": 0.014659916354831609, + "grad_norm": 2.7952757739839367, + "learning_rate": 4.886280264123258e-06, + "loss": 0.4062, + "step": 666 + }, + { + "epoch": 0.014681928241250275, + "grad_norm": 2.493558721259452, + "learning_rate": 4.893617021276596e-06, + "loss": 0.5099, + "step": 667 + }, + { + "epoch": 0.014703940127668942, + "grad_norm": 0.7215124064081021, + "learning_rate": 4.9009537784299344e-06, + "loss": 0.6568, + "step": 668 + }, + { + "epoch": 0.014725952014087607, + "grad_norm": 3.8915277940232667, + "learning_rate": 4.908290535583273e-06, + "loss": 0.4716, + "step": 669 + }, + { + "epoch": 0.014747963900506273, + "grad_norm": 2.323615941844713, + "learning_rate": 4.915627292736611e-06, + "loss": 0.4561, + "step": 670 + }, + { + "epoch": 0.01476997578692494, + "grad_norm": 3.1959367064401443, + "learning_rate": 4.922964049889949e-06, + "loss": 0.3607, + "step": 671 + }, + { + "epoch": 0.014791987673343606, + "grad_norm": 3.8452941697712495, + "learning_rate": 4.930300807043287e-06, + "loss": 0.4218, + "step": 672 + }, + { + "epoch": 0.01481399955976227, + "grad_norm": 1.904625128342675, + "learning_rate": 4.937637564196625e-06, + "loss": 0.3759, + "step": 673 + }, + { + "epoch": 0.014836011446180937, + "grad_norm": 0.6693304093844905, + "learning_rate": 4.944974321349963e-06, + "loss": 0.6491, + "step": 674 + }, + { + "epoch": 0.014858023332599604, + "grad_norm": 4.350192825012493, + "learning_rate": 4.952311078503302e-06, + "loss": 0.4422, + "step": 675 + }, + { + "epoch": 0.01488003521901827, + "grad_norm": 2.2616584295583912, + "learning_rate": 4.95964783565664e-06, + "loss": 0.425, + "step": 676 + }, + { + "epoch": 0.014902047105436935, + "grad_norm": 0.6801998943448441, + "learning_rate": 4.966984592809979e-06, + "loss": 0.6565, + "step": 677 + }, + { + "epoch": 0.014924058991855602, + "grad_norm": 1.7806865064006945, + "learning_rate": 4.974321349963316e-06, + "loss": 0.3737, + "step": 678 + }, + { + "epoch": 0.014946070878274268, + "grad_norm": 1.7717834239180688, + "learning_rate": 4.981658107116655e-06, + "loss": 0.4235, + "step": 679 + }, + { + "epoch": 0.014968082764692935, + "grad_norm": 9.411113876396639, + "learning_rate": 4.9889948642699935e-06, + "loss": 0.4933, + "step": 680 + }, + { + "epoch": 0.014990094651111601, + "grad_norm": 0.6700520309270159, + "learning_rate": 4.996331621423331e-06, + "loss": 0.6516, + "step": 681 + }, + { + "epoch": 0.015012106537530266, + "grad_norm": 2.278412592080462, + "learning_rate": 5.00366837857667e-06, + "loss": 0.3943, + "step": 682 + }, + { + "epoch": 0.015034118423948932, + "grad_norm": 2.006863851729811, + "learning_rate": 5.011005135730008e-06, + "loss": 0.4016, + "step": 683 + }, + { + "epoch": 0.015056130310367599, + "grad_norm": 2.0101620629859105, + "learning_rate": 5.018341892883346e-06, + "loss": 0.4497, + "step": 684 + }, + { + "epoch": 0.015078142196786265, + "grad_norm": 2.037624281202618, + "learning_rate": 5.025678650036684e-06, + "loss": 0.4927, + "step": 685 + }, + { + "epoch": 0.01510015408320493, + "grad_norm": 29.273530695942824, + "learning_rate": 5.033015407190023e-06, + "loss": 0.4035, + "step": 686 + }, + { + "epoch": 0.015122165969623597, + "grad_norm": 2.670462082624328, + "learning_rate": 5.040352164343361e-06, + "loss": 0.4253, + "step": 687 + }, + { + "epoch": 0.015144177856042263, + "grad_norm": 1.9113437906791262, + "learning_rate": 5.047688921496698e-06, + "loss": 0.4458, + "step": 688 + }, + { + "epoch": 0.01516618974246093, + "grad_norm": 1.7271609164502566, + "learning_rate": 5.055025678650038e-06, + "loss": 0.4102, + "step": 689 + }, + { + "epoch": 0.015188201628879594, + "grad_norm": 2.2865181351862662, + "learning_rate": 5.062362435803375e-06, + "loss": 0.4303, + "step": 690 + }, + { + "epoch": 0.01521021351529826, + "grad_norm": 2.25919632087103, + "learning_rate": 5.069699192956714e-06, + "loss": 0.4137, + "step": 691 + }, + { + "epoch": 0.015232225401716927, + "grad_norm": 2.003229399277588, + "learning_rate": 5.077035950110052e-06, + "loss": 0.4891, + "step": 692 + }, + { + "epoch": 0.015254237288135594, + "grad_norm": 1.760530244401505, + "learning_rate": 5.084372707263389e-06, + "loss": 0.5083, + "step": 693 + }, + { + "epoch": 0.015276249174554259, + "grad_norm": 3.7217770780021477, + "learning_rate": 5.091709464416729e-06, + "loss": 0.4012, + "step": 694 + }, + { + "epoch": 0.015298261060972925, + "grad_norm": 1.935974401791528, + "learning_rate": 5.099046221570066e-06, + "loss": 0.391, + "step": 695 + }, + { + "epoch": 0.015320272947391592, + "grad_norm": 4.445640981375923, + "learning_rate": 5.106382978723404e-06, + "loss": 0.4606, + "step": 696 + }, + { + "epoch": 0.015342284833810258, + "grad_norm": 2.822055239699726, + "learning_rate": 5.1137197358767434e-06, + "loss": 0.491, + "step": 697 + }, + { + "epoch": 0.015364296720228923, + "grad_norm": 3.4595980057296987, + "learning_rate": 5.121056493030081e-06, + "loss": 0.4361, + "step": 698 + }, + { + "epoch": 0.01538630860664759, + "grad_norm": 3.339119651731267, + "learning_rate": 5.128393250183419e-06, + "loss": 0.4614, + "step": 699 + }, + { + "epoch": 0.015408320493066256, + "grad_norm": 0.7407995225948446, + "learning_rate": 5.135730007336757e-06, + "loss": 0.6577, + "step": 700 + }, + { + "epoch": 0.015430332379484922, + "grad_norm": 3.2856074032252702, + "learning_rate": 5.143066764490096e-06, + "loss": 0.3781, + "step": 701 + }, + { + "epoch": 0.015452344265903587, + "grad_norm": 2.1743341999027397, + "learning_rate": 5.150403521643434e-06, + "loss": 0.4763, + "step": 702 + }, + { + "epoch": 0.015474356152322254, + "grad_norm": 1.7282227540228394, + "learning_rate": 5.157740278796772e-06, + "loss": 0.3606, + "step": 703 + }, + { + "epoch": 0.01549636803874092, + "grad_norm": 0.6816421943564118, + "learning_rate": 5.16507703595011e-06, + "loss": 0.6616, + "step": 704 + }, + { + "epoch": 0.015518379925159587, + "grad_norm": 2.022866230982866, + "learning_rate": 5.172413793103449e-06, + "loss": 0.4656, + "step": 705 + }, + { + "epoch": 0.015540391811578251, + "grad_norm": 2.0162930297600865, + "learning_rate": 5.179750550256787e-06, + "loss": 0.497, + "step": 706 + }, + { + "epoch": 0.015562403697996918, + "grad_norm": 2.039305132787687, + "learning_rate": 5.1870873074101245e-06, + "loss": 0.4366, + "step": 707 + }, + { + "epoch": 0.015584415584415584, + "grad_norm": 2.533936555518672, + "learning_rate": 5.194424064563464e-06, + "loss": 0.4642, + "step": 708 + }, + { + "epoch": 0.015606427470834251, + "grad_norm": 0.677968850739296, + "learning_rate": 5.201760821716802e-06, + "loss": 0.6592, + "step": 709 + }, + { + "epoch": 0.015628439357252916, + "grad_norm": 3.869739799656546, + "learning_rate": 5.20909757887014e-06, + "loss": 0.3612, + "step": 710 + }, + { + "epoch": 0.015650451243671582, + "grad_norm": 2.106283747879046, + "learning_rate": 5.216434336023478e-06, + "loss": 0.3523, + "step": 711 + }, + { + "epoch": 0.01567246313009025, + "grad_norm": 3.1079781010477094, + "learning_rate": 5.223771093176816e-06, + "loss": 0.4256, + "step": 712 + }, + { + "epoch": 0.015694475016508915, + "grad_norm": 1.7471936137873036, + "learning_rate": 5.231107850330155e-06, + "loss": 0.3705, + "step": 713 + }, + { + "epoch": 0.01571648690292758, + "grad_norm": 2.4795123185339674, + "learning_rate": 5.238444607483493e-06, + "loss": 0.3514, + "step": 714 + }, + { + "epoch": 0.015738498789346248, + "grad_norm": 0.6730877654405285, + "learning_rate": 5.24578136463683e-06, + "loss": 0.6403, + "step": 715 + }, + { + "epoch": 0.015760510675764915, + "grad_norm": 3.154115611752868, + "learning_rate": 5.25311812179017e-06, + "loss": 0.3605, + "step": 716 + }, + { + "epoch": 0.015782522562183578, + "grad_norm": 2.3213013379811365, + "learning_rate": 5.260454878943507e-06, + "loss": 0.4921, + "step": 717 + }, + { + "epoch": 0.015804534448602244, + "grad_norm": 1.905174384133652, + "learning_rate": 5.267791636096845e-06, + "loss": 0.4476, + "step": 718 + }, + { + "epoch": 0.01582654633502091, + "grad_norm": 2.3986554201991144, + "learning_rate": 5.275128393250184e-06, + "loss": 0.3877, + "step": 719 + }, + { + "epoch": 0.015848558221439577, + "grad_norm": 2.6563965125770452, + "learning_rate": 5.282465150403522e-06, + "loss": 0.4765, + "step": 720 + }, + { + "epoch": 0.015870570107858244, + "grad_norm": 3.819572610747216, + "learning_rate": 5.289801907556861e-06, + "loss": 0.4552, + "step": 721 + }, + { + "epoch": 0.01589258199427691, + "grad_norm": 2.5319954275270207, + "learning_rate": 5.297138664710198e-06, + "loss": 0.4718, + "step": 722 + }, + { + "epoch": 0.015914593880695577, + "grad_norm": 1.9893208012092682, + "learning_rate": 5.304475421863536e-06, + "loss": 0.3642, + "step": 723 + }, + { + "epoch": 0.015936605767114243, + "grad_norm": 2.1601095705757953, + "learning_rate": 5.311812179016875e-06, + "loss": 0.3703, + "step": 724 + }, + { + "epoch": 0.015958617653532906, + "grad_norm": 2.533951063197356, + "learning_rate": 5.319148936170213e-06, + "loss": 0.5107, + "step": 725 + }, + { + "epoch": 0.015980629539951573, + "grad_norm": 2.2281539420286363, + "learning_rate": 5.326485693323551e-06, + "loss": 0.402, + "step": 726 + }, + { + "epoch": 0.01600264142637024, + "grad_norm": 2.3281935853666136, + "learning_rate": 5.33382245047689e-06, + "loss": 0.3397, + "step": 727 + }, + { + "epoch": 0.016024653312788906, + "grad_norm": 2.329868869274188, + "learning_rate": 5.341159207630228e-06, + "loss": 0.4553, + "step": 728 + }, + { + "epoch": 0.016046665199207572, + "grad_norm": 1.8090416496865138, + "learning_rate": 5.348495964783566e-06, + "loss": 0.378, + "step": 729 + }, + { + "epoch": 0.01606867708562624, + "grad_norm": 2.489915117099039, + "learning_rate": 5.355832721936904e-06, + "loss": 0.4349, + "step": 730 + }, + { + "epoch": 0.016090688972044905, + "grad_norm": 2.838600846660645, + "learning_rate": 5.3631694790902426e-06, + "loss": 0.4608, + "step": 731 + }, + { + "epoch": 0.01611270085846357, + "grad_norm": 2.067160167570208, + "learning_rate": 5.370506236243581e-06, + "loss": 0.4548, + "step": 732 + }, + { + "epoch": 0.016134712744882235, + "grad_norm": 2.3098938997980034, + "learning_rate": 5.377842993396919e-06, + "loss": 0.4616, + "step": 733 + }, + { + "epoch": 0.0161567246313009, + "grad_norm": 2.448345692959674, + "learning_rate": 5.3851797505502565e-06, + "loss": 0.4299, + "step": 734 + }, + { + "epoch": 0.016178736517719568, + "grad_norm": 1.8922789972868395, + "learning_rate": 5.392516507703596e-06, + "loss": 0.3225, + "step": 735 + }, + { + "epoch": 0.016200748404138234, + "grad_norm": 4.1105456650435706, + "learning_rate": 5.3998532648569335e-06, + "loss": 0.4569, + "step": 736 + }, + { + "epoch": 0.0162227602905569, + "grad_norm": 2.4005082540972755, + "learning_rate": 5.407190022010272e-06, + "loss": 0.4902, + "step": 737 + }, + { + "epoch": 0.016244772176975567, + "grad_norm": 1.9655570738883954, + "learning_rate": 5.414526779163611e-06, + "loss": 0.4169, + "step": 738 + }, + { + "epoch": 0.016266784063394234, + "grad_norm": 0.7037872063397046, + "learning_rate": 5.421863536316948e-06, + "loss": 0.6399, + "step": 739 + }, + { + "epoch": 0.0162887959498129, + "grad_norm": 2.1148568258725, + "learning_rate": 5.429200293470287e-06, + "loss": 0.4284, + "step": 740 + }, + { + "epoch": 0.016310807836231567, + "grad_norm": 2.348089282909439, + "learning_rate": 5.4365370506236245e-06, + "loss": 0.3176, + "step": 741 + }, + { + "epoch": 0.01633281972265023, + "grad_norm": 1.7555953239182025, + "learning_rate": 5.443873807776962e-06, + "loss": 0.4959, + "step": 742 + }, + { + "epoch": 0.016354831609068896, + "grad_norm": 1.9071633051813381, + "learning_rate": 5.4512105649303016e-06, + "loss": 0.5161, + "step": 743 + }, + { + "epoch": 0.016376843495487563, + "grad_norm": 1.5947369408273935, + "learning_rate": 5.458547322083639e-06, + "loss": 0.2888, + "step": 744 + }, + { + "epoch": 0.01639885538190623, + "grad_norm": 2.8495932622773896, + "learning_rate": 5.465884079236977e-06, + "loss": 0.4353, + "step": 745 + }, + { + "epoch": 0.016420867268324896, + "grad_norm": 0.6828701999619853, + "learning_rate": 5.473220836390316e-06, + "loss": 0.643, + "step": 746 + }, + { + "epoch": 0.016442879154743562, + "grad_norm": 1.8782156688487837, + "learning_rate": 5.480557593543654e-06, + "loss": 0.3909, + "step": 747 + }, + { + "epoch": 0.01646489104116223, + "grad_norm": 2.3602260320740114, + "learning_rate": 5.4878943506969925e-06, + "loss": 0.4842, + "step": 748 + }, + { + "epoch": 0.016486902927580895, + "grad_norm": 2.420903970726538, + "learning_rate": 5.49523110785033e-06, + "loss": 0.5078, + "step": 749 + }, + { + "epoch": 0.016508914813999558, + "grad_norm": 0.7177875140364159, + "learning_rate": 5.502567865003669e-06, + "loss": 0.6676, + "step": 750 + }, + { + "epoch": 0.016530926700418225, + "grad_norm": 2.338049073245788, + "learning_rate": 5.509904622157007e-06, + "loss": 0.5115, + "step": 751 + }, + { + "epoch": 0.01655293858683689, + "grad_norm": 1.8726342811894192, + "learning_rate": 5.517241379310345e-06, + "loss": 0.4783, + "step": 752 + }, + { + "epoch": 0.016574950473255558, + "grad_norm": 2.900876821641742, + "learning_rate": 5.524578136463683e-06, + "loss": 0.468, + "step": 753 + }, + { + "epoch": 0.016596962359674224, + "grad_norm": 0.6576869524774074, + "learning_rate": 5.531914893617022e-06, + "loss": 0.6423, + "step": 754 + }, + { + "epoch": 0.01661897424609289, + "grad_norm": 2.690087894277831, + "learning_rate": 5.53925165077036e-06, + "loss": 0.419, + "step": 755 + }, + { + "epoch": 0.016640986132511557, + "grad_norm": 2.0292424497999444, + "learning_rate": 5.546588407923698e-06, + "loss": 0.5034, + "step": 756 + }, + { + "epoch": 0.016662998018930224, + "grad_norm": 2.827593947988244, + "learning_rate": 5.553925165077037e-06, + "loss": 0.3691, + "step": 757 + }, + { + "epoch": 0.016685009905348887, + "grad_norm": 2.4435128819564635, + "learning_rate": 5.5612619222303745e-06, + "loss": 0.3989, + "step": 758 + }, + { + "epoch": 0.016707021791767553, + "grad_norm": 1.7713947528044143, + "learning_rate": 5.568598679383713e-06, + "loss": 0.4316, + "step": 759 + }, + { + "epoch": 0.01672903367818622, + "grad_norm": 2.468572782482417, + "learning_rate": 5.575935436537051e-06, + "loss": 0.4201, + "step": 760 + }, + { + "epoch": 0.016751045564604886, + "grad_norm": 2.461094500955994, + "learning_rate": 5.583272193690389e-06, + "loss": 0.3685, + "step": 761 + }, + { + "epoch": 0.016773057451023553, + "grad_norm": 10.071879531560759, + "learning_rate": 5.590608950843728e-06, + "loss": 0.5087, + "step": 762 + }, + { + "epoch": 0.01679506933744222, + "grad_norm": 1.6235503893478536, + "learning_rate": 5.5979457079970655e-06, + "loss": 0.3994, + "step": 763 + }, + { + "epoch": 0.016817081223860886, + "grad_norm": 2.1231150877953078, + "learning_rate": 5.605282465150405e-06, + "loss": 0.4203, + "step": 764 + }, + { + "epoch": 0.016839093110279552, + "grad_norm": 2.4901753703911713, + "learning_rate": 5.6126192223037425e-06, + "loss": 0.3564, + "step": 765 + }, + { + "epoch": 0.016861104996698215, + "grad_norm": 2.012490545035428, + "learning_rate": 5.61995597945708e-06, + "loss": 0.42, + "step": 766 + }, + { + "epoch": 0.016883116883116882, + "grad_norm": 2.0430416431315717, + "learning_rate": 5.627292736610419e-06, + "loss": 0.388, + "step": 767 + }, + { + "epoch": 0.01690512876953555, + "grad_norm": 2.201338747117045, + "learning_rate": 5.634629493763757e-06, + "loss": 0.4347, + "step": 768 + }, + { + "epoch": 0.016927140655954215, + "grad_norm": 2.6854127691195084, + "learning_rate": 5.641966250917095e-06, + "loss": 0.379, + "step": 769 + }, + { + "epoch": 0.01694915254237288, + "grad_norm": 2.5992032990087215, + "learning_rate": 5.6493030080704335e-06, + "loss": 0.5786, + "step": 770 + }, + { + "epoch": 0.016971164428791548, + "grad_norm": 0.7025746595709161, + "learning_rate": 5.656639765223771e-06, + "loss": 0.66, + "step": 771 + }, + { + "epoch": 0.016993176315210214, + "grad_norm": 1.7899002566673923, + "learning_rate": 5.663976522377109e-06, + "loss": 0.3542, + "step": 772 + }, + { + "epoch": 0.01701518820162888, + "grad_norm": 2.2289506149558185, + "learning_rate": 5.671313279530448e-06, + "loss": 0.4165, + "step": 773 + }, + { + "epoch": 0.017037200088047547, + "grad_norm": 2.0558602100771894, + "learning_rate": 5.678650036683786e-06, + "loss": 0.3453, + "step": 774 + }, + { + "epoch": 0.01705921197446621, + "grad_norm": 2.2209639664542267, + "learning_rate": 5.685986793837125e-06, + "loss": 0.3617, + "step": 775 + }, + { + "epoch": 0.017081223860884877, + "grad_norm": 1.825775397250416, + "learning_rate": 5.693323550990463e-06, + "loss": 0.4121, + "step": 776 + }, + { + "epoch": 0.017103235747303543, + "grad_norm": 1.8734505655802351, + "learning_rate": 5.700660308143801e-06, + "loss": 0.4, + "step": 777 + }, + { + "epoch": 0.01712524763372221, + "grad_norm": 0.6756616706559047, + "learning_rate": 5.707997065297139e-06, + "loss": 0.6437, + "step": 778 + }, + { + "epoch": 0.017147259520140876, + "grad_norm": 2.212932137189686, + "learning_rate": 5.715333822450477e-06, + "loss": 0.447, + "step": 779 + }, + { + "epoch": 0.017169271406559543, + "grad_norm": 2.1885002734702117, + "learning_rate": 5.7226705796038155e-06, + "loss": 0.4072, + "step": 780 + }, + { + "epoch": 0.01719128329297821, + "grad_norm": 4.015406604591243, + "learning_rate": 5.730007336757154e-06, + "loss": 0.422, + "step": 781 + }, + { + "epoch": 0.017213295179396876, + "grad_norm": 3.878427304277725, + "learning_rate": 5.737344093910492e-06, + "loss": 0.4709, + "step": 782 + }, + { + "epoch": 0.01723530706581554, + "grad_norm": 0.677751496804819, + "learning_rate": 5.744680851063831e-06, + "loss": 0.6526, + "step": 783 + }, + { + "epoch": 0.017257318952234205, + "grad_norm": 3.675042734600026, + "learning_rate": 5.752017608217169e-06, + "loss": 0.4462, + "step": 784 + }, + { + "epoch": 0.017279330838652872, + "grad_norm": 1.6302040717190753, + "learning_rate": 5.759354365370506e-06, + "loss": 0.4467, + "step": 785 + }, + { + "epoch": 0.01730134272507154, + "grad_norm": 2.025312860185753, + "learning_rate": 5.766691122523845e-06, + "loss": 0.5059, + "step": 786 + }, + { + "epoch": 0.017323354611490205, + "grad_norm": 2.006127238227051, + "learning_rate": 5.7740278796771835e-06, + "loss": 0.3659, + "step": 787 + }, + { + "epoch": 0.01734536649790887, + "grad_norm": 2.1908934703554497, + "learning_rate": 5.781364636830521e-06, + "loss": 0.4292, + "step": 788 + }, + { + "epoch": 0.017367378384327538, + "grad_norm": 0.6576693943923574, + "learning_rate": 5.78870139398386e-06, + "loss": 0.6496, + "step": 789 + }, + { + "epoch": 0.017389390270746204, + "grad_norm": 4.603909940054451, + "learning_rate": 5.796038151137197e-06, + "loss": 0.3456, + "step": 790 + }, + { + "epoch": 0.017411402157164867, + "grad_norm": 0.6672676099736435, + "learning_rate": 5.803374908290536e-06, + "loss": 0.6702, + "step": 791 + }, + { + "epoch": 0.017433414043583534, + "grad_norm": 2.3498502681904645, + "learning_rate": 5.8107116654438745e-06, + "loss": 0.4914, + "step": 792 + }, + { + "epoch": 0.0174554259300022, + "grad_norm": 1.8584302410913283, + "learning_rate": 5.818048422597212e-06, + "loss": 0.4121, + "step": 793 + }, + { + "epoch": 0.017477437816420867, + "grad_norm": 3.0045803680049756, + "learning_rate": 5.8253851797505515e-06, + "loss": 0.4059, + "step": 794 + }, + { + "epoch": 0.017499449702839533, + "grad_norm": 2.503324749583801, + "learning_rate": 5.832721936903889e-06, + "loss": 0.4559, + "step": 795 + }, + { + "epoch": 0.0175214615892582, + "grad_norm": 1.9547336898845868, + "learning_rate": 5.840058694057227e-06, + "loss": 0.4399, + "step": 796 + }, + { + "epoch": 0.017543473475676866, + "grad_norm": 0.719323909340487, + "learning_rate": 5.8473954512105654e-06, + "loss": 0.6456, + "step": 797 + }, + { + "epoch": 0.017565485362095533, + "grad_norm": 2.8121100675829998, + "learning_rate": 5.854732208363904e-06, + "loss": 0.4699, + "step": 798 + }, + { + "epoch": 0.0175874972485142, + "grad_norm": 3.255842773669828, + "learning_rate": 5.862068965517242e-06, + "loss": 0.4143, + "step": 799 + }, + { + "epoch": 0.017609509134932862, + "grad_norm": 3.291791936947551, + "learning_rate": 5.86940572267058e-06, + "loss": 0.3467, + "step": 800 + }, + { + "epoch": 0.01763152102135153, + "grad_norm": 2.0711477914601093, + "learning_rate": 5.876742479823918e-06, + "loss": 0.4817, + "step": 801 + }, + { + "epoch": 0.017653532907770195, + "grad_norm": 6.112224240664967, + "learning_rate": 5.884079236977257e-06, + "loss": 0.3597, + "step": 802 + }, + { + "epoch": 0.017675544794188862, + "grad_norm": 2.80149479705435, + "learning_rate": 5.891415994130595e-06, + "loss": 0.4145, + "step": 803 + }, + { + "epoch": 0.01769755668060753, + "grad_norm": 1.7643675889326893, + "learning_rate": 5.898752751283933e-06, + "loss": 0.4762, + "step": 804 + }, + { + "epoch": 0.017719568567026195, + "grad_norm": 0.7068958829005534, + "learning_rate": 5.906089508437271e-06, + "loss": 0.6568, + "step": 805 + }, + { + "epoch": 0.01774158045344486, + "grad_norm": 2.3564117091487033, + "learning_rate": 5.91342626559061e-06, + "loss": 0.4154, + "step": 806 + }, + { + "epoch": 0.017763592339863528, + "grad_norm": 2.51709372338935, + "learning_rate": 5.920763022743947e-06, + "loss": 0.505, + "step": 807 + }, + { + "epoch": 0.01778560422628219, + "grad_norm": 4.237194768828965, + "learning_rate": 5.928099779897286e-06, + "loss": 0.4312, + "step": 808 + }, + { + "epoch": 0.017807616112700857, + "grad_norm": 2.0554424705724226, + "learning_rate": 5.935436537050624e-06, + "loss": 0.4257, + "step": 809 + }, + { + "epoch": 0.017829627999119524, + "grad_norm": 1.8259146532283475, + "learning_rate": 5.942773294203963e-06, + "loss": 0.3984, + "step": 810 + }, + { + "epoch": 0.01785163988553819, + "grad_norm": 2.066112431798249, + "learning_rate": 5.950110051357301e-06, + "loss": 0.483, + "step": 811 + }, + { + "epoch": 0.017873651771956857, + "grad_norm": 2.1539322930308993, + "learning_rate": 5.957446808510638e-06, + "loss": 0.442, + "step": 812 + }, + { + "epoch": 0.017895663658375523, + "grad_norm": 1.8885600479308662, + "learning_rate": 5.964783565663978e-06, + "loss": 0.3646, + "step": 813 + }, + { + "epoch": 0.01791767554479419, + "grad_norm": 2.0782100868262092, + "learning_rate": 5.972120322817315e-06, + "loss": 0.4668, + "step": 814 + }, + { + "epoch": 0.017939687431212856, + "grad_norm": 2.9810941567339686, + "learning_rate": 5.979457079970653e-06, + "loss": 0.4453, + "step": 815 + }, + { + "epoch": 0.01796169931763152, + "grad_norm": 1.742865032838832, + "learning_rate": 5.986793837123992e-06, + "loss": 0.4597, + "step": 816 + }, + { + "epoch": 0.017983711204050186, + "grad_norm": 2.762379844820335, + "learning_rate": 5.99413059427733e-06, + "loss": 0.5184, + "step": 817 + }, + { + "epoch": 0.018005723090468852, + "grad_norm": 3.259179225807565, + "learning_rate": 6.001467351430668e-06, + "loss": 0.5057, + "step": 818 + }, + { + "epoch": 0.01802773497688752, + "grad_norm": 0.696749600438473, + "learning_rate": 6.008804108584006e-06, + "loss": 0.663, + "step": 819 + }, + { + "epoch": 0.018049746863306185, + "grad_norm": 2.2641587823605183, + "learning_rate": 6.016140865737344e-06, + "loss": 0.4526, + "step": 820 + }, + { + "epoch": 0.018071758749724852, + "grad_norm": 2.2600684894107994, + "learning_rate": 6.0234776228906835e-06, + "loss": 0.4291, + "step": 821 + }, + { + "epoch": 0.01809377063614352, + "grad_norm": 3.0014477975975358, + "learning_rate": 6.030814380044021e-06, + "loss": 0.459, + "step": 822 + }, + { + "epoch": 0.018115782522562185, + "grad_norm": 2.3153292908815897, + "learning_rate": 6.038151137197359e-06, + "loss": 0.3835, + "step": 823 + }, + { + "epoch": 0.018137794408980848, + "grad_norm": 2.0494329563904783, + "learning_rate": 6.045487894350698e-06, + "loss": 0.4373, + "step": 824 + }, + { + "epoch": 0.018159806295399514, + "grad_norm": 2.3425712581367253, + "learning_rate": 6.052824651504036e-06, + "loss": 0.4903, + "step": 825 + }, + { + "epoch": 0.01818181818181818, + "grad_norm": 5.66041854209767, + "learning_rate": 6.060161408657374e-06, + "loss": 0.4618, + "step": 826 + }, + { + "epoch": 0.018203830068236847, + "grad_norm": 2.017675138978397, + "learning_rate": 6.067498165810712e-06, + "loss": 0.4172, + "step": 827 + }, + { + "epoch": 0.018225841954655514, + "grad_norm": 3.0866285010554986, + "learning_rate": 6.07483492296405e-06, + "loss": 0.483, + "step": 828 + }, + { + "epoch": 0.01824785384107418, + "grad_norm": 0.6750598315122361, + "learning_rate": 6.082171680117389e-06, + "loss": 0.6442, + "step": 829 + }, + { + "epoch": 0.018269865727492847, + "grad_norm": 2.027220682782886, + "learning_rate": 6.089508437270727e-06, + "loss": 0.4455, + "step": 830 + }, + { + "epoch": 0.018291877613911513, + "grad_norm": 0.6932695424550448, + "learning_rate": 6.0968451944240646e-06, + "loss": 0.6607, + "step": 831 + }, + { + "epoch": 0.01831388950033018, + "grad_norm": 0.6745141420845101, + "learning_rate": 6.104181951577404e-06, + "loss": 0.6481, + "step": 832 + }, + { + "epoch": 0.018335901386748843, + "grad_norm": 2.1343906744000924, + "learning_rate": 6.111518708730742e-06, + "loss": 0.4445, + "step": 833 + }, + { + "epoch": 0.01835791327316751, + "grad_norm": 2.0464486978102476, + "learning_rate": 6.118855465884079e-06, + "loss": 0.3847, + "step": 834 + }, + { + "epoch": 0.018379925159586176, + "grad_norm": 2.250367914577515, + "learning_rate": 6.126192223037418e-06, + "loss": 0.4913, + "step": 835 + }, + { + "epoch": 0.018401937046004842, + "grad_norm": 2.306675956034549, + "learning_rate": 6.133528980190756e-06, + "loss": 0.4727, + "step": 836 + }, + { + "epoch": 0.01842394893242351, + "grad_norm": 1.8259019474713054, + "learning_rate": 6.140865737344094e-06, + "loss": 0.3805, + "step": 837 + }, + { + "epoch": 0.018445960818842175, + "grad_norm": 0.7197775553773584, + "learning_rate": 6.148202494497433e-06, + "loss": 0.63, + "step": 838 + }, + { + "epoch": 0.018467972705260842, + "grad_norm": 4.780395881622654, + "learning_rate": 6.15553925165077e-06, + "loss": 0.4546, + "step": 839 + }, + { + "epoch": 0.01848998459167951, + "grad_norm": 0.6846514306095495, + "learning_rate": 6.16287600880411e-06, + "loss": 0.6378, + "step": 840 + }, + { + "epoch": 0.01851199647809817, + "grad_norm": 2.0101871592730873, + "learning_rate": 6.170212765957447e-06, + "loss": 0.4702, + "step": 841 + }, + { + "epoch": 0.018534008364516838, + "grad_norm": 0.6527826038496636, + "learning_rate": 6.177549523110785e-06, + "loss": 0.6646, + "step": 842 + }, + { + "epoch": 0.018556020250935504, + "grad_norm": 2.413717749577913, + "learning_rate": 6.184886280264124e-06, + "loss": 0.4512, + "step": 843 + }, + { + "epoch": 0.01857803213735417, + "grad_norm": 2.046132808159179, + "learning_rate": 6.192223037417462e-06, + "loss": 0.4064, + "step": 844 + }, + { + "epoch": 0.018600044023772837, + "grad_norm": 0.6888898793872721, + "learning_rate": 6.1995597945708e-06, + "loss": 0.6669, + "step": 845 + }, + { + "epoch": 0.018622055910191504, + "grad_norm": 0.7110575097487952, + "learning_rate": 6.206896551724138e-06, + "loss": 0.6553, + "step": 846 + }, + { + "epoch": 0.01864406779661017, + "grad_norm": 2.3295283745379924, + "learning_rate": 6.214233308877477e-06, + "loss": 0.4391, + "step": 847 + }, + { + "epoch": 0.018666079683028837, + "grad_norm": 0.6649601249104176, + "learning_rate": 6.221570066030815e-06, + "loss": 0.634, + "step": 848 + }, + { + "epoch": 0.0186880915694475, + "grad_norm": 1.7316760410869743, + "learning_rate": 6.228906823184153e-06, + "loss": 0.4158, + "step": 849 + }, + { + "epoch": 0.018710103455866166, + "grad_norm": 1.952850720478357, + "learning_rate": 6.236243580337491e-06, + "loss": 0.4207, + "step": 850 + }, + { + "epoch": 0.018732115342284833, + "grad_norm": 1.8723447992486724, + "learning_rate": 6.24358033749083e-06, + "loss": 0.4275, + "step": 851 + }, + { + "epoch": 0.0187541272287035, + "grad_norm": 2.106062321142339, + "learning_rate": 6.250917094644168e-06, + "loss": 0.4405, + "step": 852 + }, + { + "epoch": 0.018776139115122166, + "grad_norm": 1.7867199612338744, + "learning_rate": 6.2582538517975055e-06, + "loss": 0.4179, + "step": 853 + }, + { + "epoch": 0.018798151001540832, + "grad_norm": 0.7505043281481816, + "learning_rate": 6.265590608950845e-06, + "loss": 0.6333, + "step": 854 + }, + { + "epoch": 0.0188201628879595, + "grad_norm": 2.434841878021269, + "learning_rate": 6.272927366104183e-06, + "loss": 0.4388, + "step": 855 + }, + { + "epoch": 0.018842174774378165, + "grad_norm": 1.9158166551271263, + "learning_rate": 6.280264123257521e-06, + "loss": 0.4074, + "step": 856 + }, + { + "epoch": 0.018864186660796832, + "grad_norm": 1.7800972160631974, + "learning_rate": 6.287600880410859e-06, + "loss": 0.4406, + "step": 857 + }, + { + "epoch": 0.018886198547215495, + "grad_norm": 0.6927475868961975, + "learning_rate": 6.2949376375641965e-06, + "loss": 0.6482, + "step": 858 + }, + { + "epoch": 0.01890821043363416, + "grad_norm": 1.9028762373918195, + "learning_rate": 6.302274394717536e-06, + "loss": 0.4452, + "step": 859 + }, + { + "epoch": 0.018930222320052828, + "grad_norm": 2.3284024426404764, + "learning_rate": 6.3096111518708736e-06, + "loss": 0.3427, + "step": 860 + }, + { + "epoch": 0.018952234206471495, + "grad_norm": 2.4862385819072514, + "learning_rate": 6.316947909024211e-06, + "loss": 0.4438, + "step": 861 + }, + { + "epoch": 0.01897424609289016, + "grad_norm": 2.0406136246317095, + "learning_rate": 6.324284666177551e-06, + "loss": 0.4465, + "step": 862 + }, + { + "epoch": 0.018996257979308828, + "grad_norm": 3.4008384278450565, + "learning_rate": 6.331621423330888e-06, + "loss": 0.426, + "step": 863 + }, + { + "epoch": 0.019018269865727494, + "grad_norm": 0.7180121529662267, + "learning_rate": 6.338958180484226e-06, + "loss": 0.6303, + "step": 864 + }, + { + "epoch": 0.01904028175214616, + "grad_norm": 0.7138124028790747, + "learning_rate": 6.3462949376375645e-06, + "loss": 0.6297, + "step": 865 + }, + { + "epoch": 0.019062293638564824, + "grad_norm": 0.6747545798745306, + "learning_rate": 6.353631694790903e-06, + "loss": 0.665, + "step": 866 + }, + { + "epoch": 0.01908430552498349, + "grad_norm": 0.653427254369497, + "learning_rate": 6.360968451944242e-06, + "loss": 0.6412, + "step": 867 + }, + { + "epoch": 0.019106317411402157, + "grad_norm": 1.9285951438989306, + "learning_rate": 6.368305209097579e-06, + "loss": 0.3312, + "step": 868 + }, + { + "epoch": 0.019128329297820823, + "grad_norm": 2.5298547134490197, + "learning_rate": 6.375641966250917e-06, + "loss": 0.4188, + "step": 869 + }, + { + "epoch": 0.01915034118423949, + "grad_norm": 2.447350168740994, + "learning_rate": 6.382978723404256e-06, + "loss": 0.3546, + "step": 870 + }, + { + "epoch": 0.019172353070658156, + "grad_norm": 0.8132129951954123, + "learning_rate": 6.390315480557594e-06, + "loss": 0.6433, + "step": 871 + }, + { + "epoch": 0.019194364957076823, + "grad_norm": 14.006395358853629, + "learning_rate": 6.397652237710932e-06, + "loss": 0.4846, + "step": 872 + }, + { + "epoch": 0.01921637684349549, + "grad_norm": 2.647185001281328, + "learning_rate": 6.404988994864271e-06, + "loss": 0.4407, + "step": 873 + }, + { + "epoch": 0.019238388729914152, + "grad_norm": 2.790263976989656, + "learning_rate": 6.412325752017609e-06, + "loss": 0.4231, + "step": 874 + }, + { + "epoch": 0.01926040061633282, + "grad_norm": 2.122824570462353, + "learning_rate": 6.419662509170947e-06, + "loss": 0.5627, + "step": 875 + }, + { + "epoch": 0.019282412502751485, + "grad_norm": 2.389869211005779, + "learning_rate": 6.426999266324285e-06, + "loss": 0.3935, + "step": 876 + }, + { + "epoch": 0.01930442438917015, + "grad_norm": 1.8848361198605947, + "learning_rate": 6.434336023477623e-06, + "loss": 0.4139, + "step": 877 + }, + { + "epoch": 0.019326436275588818, + "grad_norm": 1.8229016166658816, + "learning_rate": 6.441672780630962e-06, + "loss": 0.5004, + "step": 878 + }, + { + "epoch": 0.019348448162007485, + "grad_norm": 5.33721969382114, + "learning_rate": 6.4490095377843e-06, + "loss": 0.3244, + "step": 879 + }, + { + "epoch": 0.01937046004842615, + "grad_norm": 0.7246492804821133, + "learning_rate": 6.4563462949376374e-06, + "loss": 0.679, + "step": 880 + }, + { + "epoch": 0.019392471934844818, + "grad_norm": 2.175392140161578, + "learning_rate": 6.463683052090977e-06, + "loss": 0.4309, + "step": 881 + }, + { + "epoch": 0.01941448382126348, + "grad_norm": 1.726527335506735, + "learning_rate": 6.4710198092443145e-06, + "loss": 0.3453, + "step": 882 + }, + { + "epoch": 0.019436495707682147, + "grad_norm": 2.0412764706979023, + "learning_rate": 6.478356566397653e-06, + "loss": 0.4339, + "step": 883 + }, + { + "epoch": 0.019458507594100814, + "grad_norm": 2.867168883465438, + "learning_rate": 6.485693323550991e-06, + "loss": 0.4928, + "step": 884 + }, + { + "epoch": 0.01948051948051948, + "grad_norm": 3.2157223679130484, + "learning_rate": 6.493030080704329e-06, + "loss": 0.3934, + "step": 885 + }, + { + "epoch": 0.019502531366938147, + "grad_norm": 0.6738040842765701, + "learning_rate": 6.500366837857668e-06, + "loss": 0.625, + "step": 886 + }, + { + "epoch": 0.019524543253356813, + "grad_norm": 3.5670747128319227, + "learning_rate": 6.5077035950110055e-06, + "loss": 0.4642, + "step": 887 + }, + { + "epoch": 0.01954655513977548, + "grad_norm": 1.5783824795761514, + "learning_rate": 6.515040352164343e-06, + "loss": 0.3348, + "step": 888 + }, + { + "epoch": 0.019568567026194146, + "grad_norm": 2.213263477252955, + "learning_rate": 6.5223771093176826e-06, + "loss": 0.4179, + "step": 889 + }, + { + "epoch": 0.019590578912612813, + "grad_norm": 1.6873030472893908, + "learning_rate": 6.52971386647102e-06, + "loss": 0.3567, + "step": 890 + }, + { + "epoch": 0.019612590799031476, + "grad_norm": 0.6786088457591164, + "learning_rate": 6.537050623624358e-06, + "loss": 0.6389, + "step": 891 + }, + { + "epoch": 0.019634602685450142, + "grad_norm": 2.320598658658551, + "learning_rate": 6.544387380777697e-06, + "loss": 0.3783, + "step": 892 + }, + { + "epoch": 0.01965661457186881, + "grad_norm": 2.35608232955067, + "learning_rate": 6.551724137931035e-06, + "loss": 0.5824, + "step": 893 + }, + { + "epoch": 0.019678626458287475, + "grad_norm": 2.113597073146196, + "learning_rate": 6.5590608950843735e-06, + "loss": 0.4628, + "step": 894 + }, + { + "epoch": 0.01970063834470614, + "grad_norm": 1.9154110986686685, + "learning_rate": 6.566397652237711e-06, + "loss": 0.3919, + "step": 895 + }, + { + "epoch": 0.019722650231124808, + "grad_norm": 1.8465074539499433, + "learning_rate": 6.57373440939105e-06, + "loss": 0.3305, + "step": 896 + }, + { + "epoch": 0.019744662117543475, + "grad_norm": 6.839151185211937, + "learning_rate": 6.581071166544388e-06, + "loss": 0.4045, + "step": 897 + }, + { + "epoch": 0.01976667400396214, + "grad_norm": 2.1630455959455284, + "learning_rate": 6.588407923697726e-06, + "loss": 0.4355, + "step": 898 + }, + { + "epoch": 0.019788685890380804, + "grad_norm": 0.6961280034609401, + "learning_rate": 6.595744680851064e-06, + "loss": 0.6614, + "step": 899 + }, + { + "epoch": 0.01981069777679947, + "grad_norm": 1.766533237981235, + "learning_rate": 6.603081438004403e-06, + "loss": 0.3792, + "step": 900 + }, + { + "epoch": 0.019832709663218137, + "grad_norm": 1.6080154788367151, + "learning_rate": 6.610418195157741e-06, + "loss": 0.3889, + "step": 901 + }, + { + "epoch": 0.019854721549636804, + "grad_norm": 2.0527447101029375, + "learning_rate": 6.617754952311079e-06, + "loss": 0.4717, + "step": 902 + }, + { + "epoch": 0.01987673343605547, + "grad_norm": 1.95868041632486, + "learning_rate": 6.625091709464418e-06, + "loss": 0.5373, + "step": 903 + }, + { + "epoch": 0.019898745322474137, + "grad_norm": 1.9085164775061736, + "learning_rate": 6.6324284666177555e-06, + "loss": 0.4123, + "step": 904 + }, + { + "epoch": 0.019920757208892803, + "grad_norm": 0.675818559087022, + "learning_rate": 6.639765223771094e-06, + "loss": 0.654, + "step": 905 + }, + { + "epoch": 0.01994276909531147, + "grad_norm": 1.9465259921221705, + "learning_rate": 6.647101980924432e-06, + "loss": 0.4184, + "step": 906 + }, + { + "epoch": 0.019964780981730133, + "grad_norm": 2.493680868773037, + "learning_rate": 6.654438738077769e-06, + "loss": 0.4067, + "step": 907 + }, + { + "epoch": 0.0199867928681488, + "grad_norm": 2.649204365486509, + "learning_rate": 6.661775495231109e-06, + "loss": 0.3893, + "step": 908 + }, + { + "epoch": 0.020008804754567466, + "grad_norm": 2.6348212232601425, + "learning_rate": 6.6691122523844464e-06, + "loss": 0.515, + "step": 909 + }, + { + "epoch": 0.020030816640986132, + "grad_norm": 2.6285647595086843, + "learning_rate": 6.676449009537784e-06, + "loss": 0.4008, + "step": 910 + }, + { + "epoch": 0.0200528285274048, + "grad_norm": 2.243274538545867, + "learning_rate": 6.6837857666911235e-06, + "loss": 0.286, + "step": 911 + }, + { + "epoch": 0.020074840413823465, + "grad_norm": 1.921945289774911, + "learning_rate": 6.691122523844461e-06, + "loss": 0.4629, + "step": 912 + }, + { + "epoch": 0.02009685230024213, + "grad_norm": 2.4486681737236062, + "learning_rate": 6.6984592809978e-06, + "loss": 0.386, + "step": 913 + }, + { + "epoch": 0.020118864186660798, + "grad_norm": 2.0591075767685325, + "learning_rate": 6.705796038151137e-06, + "loss": 0.392, + "step": 914 + }, + { + "epoch": 0.020140876073079465, + "grad_norm": 2.363986499184395, + "learning_rate": 6.713132795304476e-06, + "loss": 0.4597, + "step": 915 + }, + { + "epoch": 0.020162887959498128, + "grad_norm": 1.9079870303782616, + "learning_rate": 6.7204695524578145e-06, + "loss": 0.4323, + "step": 916 + }, + { + "epoch": 0.020184899845916794, + "grad_norm": 2.3899687120179727, + "learning_rate": 6.727806309611152e-06, + "loss": 0.5062, + "step": 917 + }, + { + "epoch": 0.02020691173233546, + "grad_norm": 4.032107247793033, + "learning_rate": 6.73514306676449e-06, + "loss": 0.4612, + "step": 918 + }, + { + "epoch": 0.020228923618754127, + "grad_norm": 2.5065323044839363, + "learning_rate": 6.742479823917829e-06, + "loss": 0.4129, + "step": 919 + }, + { + "epoch": 0.020250935505172794, + "grad_norm": 2.537415382995791, + "learning_rate": 6.749816581071167e-06, + "loss": 0.4459, + "step": 920 + }, + { + "epoch": 0.02027294739159146, + "grad_norm": 2.374463926728743, + "learning_rate": 6.7571533382245055e-06, + "loss": 0.4737, + "step": 921 + }, + { + "epoch": 0.020294959278010127, + "grad_norm": 0.7494339909039541, + "learning_rate": 6.764490095377844e-06, + "loss": 0.6209, + "step": 922 + }, + { + "epoch": 0.020316971164428793, + "grad_norm": 2.4876678007457067, + "learning_rate": 6.771826852531182e-06, + "loss": 0.4141, + "step": 923 + }, + { + "epoch": 0.020338983050847456, + "grad_norm": 2.6396511123859945, + "learning_rate": 6.77916360968452e-06, + "loss": 0.4343, + "step": 924 + }, + { + "epoch": 0.020360994937266123, + "grad_norm": 2.745428781584802, + "learning_rate": 6.786500366837858e-06, + "loss": 0.4388, + "step": 925 + }, + { + "epoch": 0.02038300682368479, + "grad_norm": 0.6850816769592861, + "learning_rate": 6.793837123991196e-06, + "loss": 0.633, + "step": 926 + }, + { + "epoch": 0.020405018710103456, + "grad_norm": 2.739014568465521, + "learning_rate": 6.801173881144535e-06, + "loss": 0.4179, + "step": 927 + }, + { + "epoch": 0.020427030596522122, + "grad_norm": 2.0608178787346185, + "learning_rate": 6.808510638297873e-06, + "loss": 0.4631, + "step": 928 + }, + { + "epoch": 0.02044904248294079, + "grad_norm": 1.9342083359009814, + "learning_rate": 6.815847395451212e-06, + "loss": 0.3924, + "step": 929 + }, + { + "epoch": 0.020471054369359455, + "grad_norm": 1.8278861036689296, + "learning_rate": 6.82318415260455e-06, + "loss": 0.3847, + "step": 930 + }, + { + "epoch": 0.02049306625577812, + "grad_norm": 1.8973827095289126, + "learning_rate": 6.830520909757887e-06, + "loss": 0.4644, + "step": 931 + }, + { + "epoch": 0.020515078142196785, + "grad_norm": 9.94113922641944, + "learning_rate": 6.837857666911226e-06, + "loss": 0.4664, + "step": 932 + }, + { + "epoch": 0.02053709002861545, + "grad_norm": 2.043365292035432, + "learning_rate": 6.845194424064564e-06, + "loss": 0.4118, + "step": 933 + }, + { + "epoch": 0.020559101915034118, + "grad_norm": 0.76074443468251, + "learning_rate": 6.852531181217902e-06, + "loss": 0.642, + "step": 934 + }, + { + "epoch": 0.020581113801452784, + "grad_norm": 2.1990848559033154, + "learning_rate": 6.859867938371241e-06, + "loss": 0.5069, + "step": 935 + }, + { + "epoch": 0.02060312568787145, + "grad_norm": 2.7571633676723293, + "learning_rate": 6.867204695524578e-06, + "loss": 0.4557, + "step": 936 + }, + { + "epoch": 0.020625137574290117, + "grad_norm": 5.283783486279419, + "learning_rate": 6.874541452677916e-06, + "loss": 0.3787, + "step": 937 + }, + { + "epoch": 0.020647149460708784, + "grad_norm": 0.6754053595440652, + "learning_rate": 6.8818782098312554e-06, + "loss": 0.6039, + "step": 938 + }, + { + "epoch": 0.02066916134712745, + "grad_norm": 1.9771072660221125, + "learning_rate": 6.889214966984593e-06, + "loss": 0.3806, + "step": 939 + }, + { + "epoch": 0.020691173233546113, + "grad_norm": 2.0572642398697685, + "learning_rate": 6.896551724137932e-06, + "loss": 0.3713, + "step": 940 + }, + { + "epoch": 0.02071318511996478, + "grad_norm": 2.3467522559980987, + "learning_rate": 6.90388848129127e-06, + "loss": 0.4412, + "step": 941 + }, + { + "epoch": 0.020735197006383446, + "grad_norm": 2.938220500180199, + "learning_rate": 6.911225238444608e-06, + "loss": 0.3799, + "step": 942 + }, + { + "epoch": 0.020757208892802113, + "grad_norm": 0.7406272193389065, + "learning_rate": 6.918561995597946e-06, + "loss": 0.6389, + "step": 943 + }, + { + "epoch": 0.02077922077922078, + "grad_norm": 1.8010847899266762, + "learning_rate": 6.925898752751284e-06, + "loss": 0.4094, + "step": 944 + }, + { + "epoch": 0.020801232665639446, + "grad_norm": 3.6213007342766246, + "learning_rate": 6.933235509904623e-06, + "loss": 0.4735, + "step": 945 + }, + { + "epoch": 0.020823244552058112, + "grad_norm": 0.6703494861525995, + "learning_rate": 6.940572267057961e-06, + "loss": 0.643, + "step": 946 + }, + { + "epoch": 0.02084525643847678, + "grad_norm": 0.6734859150701765, + "learning_rate": 6.947909024211299e-06, + "loss": 0.6168, + "step": 947 + }, + { + "epoch": 0.020867268324895445, + "grad_norm": 3.5780833470471847, + "learning_rate": 6.955245781364638e-06, + "loss": 0.4284, + "step": 948 + }, + { + "epoch": 0.020889280211314108, + "grad_norm": 0.6766506950756207, + "learning_rate": 6.962582538517976e-06, + "loss": 0.6185, + "step": 949 + }, + { + "epoch": 0.020911292097732775, + "grad_norm": 1.7005200847798825, + "learning_rate": 6.969919295671314e-06, + "loss": 0.4652, + "step": 950 + }, + { + "epoch": 0.02093330398415144, + "grad_norm": 0.6965790207776367, + "learning_rate": 6.977256052824652e-06, + "loss": 0.6355, + "step": 951 + }, + { + "epoch": 0.020955315870570108, + "grad_norm": 2.5109561814771384, + "learning_rate": 6.984592809977991e-06, + "loss": 0.4131, + "step": 952 + }, + { + "epoch": 0.020977327756988774, + "grad_norm": 2.580820828863021, + "learning_rate": 6.991929567131328e-06, + "loss": 0.4272, + "step": 953 + }, + { + "epoch": 0.02099933964340744, + "grad_norm": 2.756629755984036, + "learning_rate": 6.999266324284667e-06, + "loss": 0.4733, + "step": 954 + }, + { + "epoch": 0.021021351529826107, + "grad_norm": 6.739115569260982, + "learning_rate": 7.0066030814380046e-06, + "loss": 0.4195, + "step": 955 + }, + { + "epoch": 0.021043363416244774, + "grad_norm": 2.806480115012723, + "learning_rate": 7.013939838591344e-06, + "loss": 0.3971, + "step": 956 + }, + { + "epoch": 0.021065375302663437, + "grad_norm": 0.6536450218261071, + "learning_rate": 7.021276595744682e-06, + "loss": 0.6216, + "step": 957 + }, + { + "epoch": 0.021087387189082103, + "grad_norm": 2.6670474393638157, + "learning_rate": 7.028613352898019e-06, + "loss": 0.4062, + "step": 958 + }, + { + "epoch": 0.02110939907550077, + "grad_norm": 4.980107570135016, + "learning_rate": 7.035950110051359e-06, + "loss": 0.3961, + "step": 959 + }, + { + "epoch": 0.021131410961919436, + "grad_norm": 2.4756147024428854, + "learning_rate": 7.043286867204696e-06, + "loss": 0.4094, + "step": 960 + }, + { + "epoch": 0.021153422848338103, + "grad_norm": 1.7501491581298854, + "learning_rate": 7.050623624358034e-06, + "loss": 0.4844, + "step": 961 + }, + { + "epoch": 0.02117543473475677, + "grad_norm": 2.859430370616331, + "learning_rate": 7.057960381511373e-06, + "loss": 0.3931, + "step": 962 + }, + { + "epoch": 0.021197446621175436, + "grad_norm": 1.8868927589788365, + "learning_rate": 7.06529713866471e-06, + "loss": 0.3528, + "step": 963 + }, + { + "epoch": 0.021219458507594102, + "grad_norm": 1.806288429379955, + "learning_rate": 7.072633895818049e-06, + "loss": 0.4269, + "step": 964 + }, + { + "epoch": 0.021241470394012765, + "grad_norm": 0.7263160972875619, + "learning_rate": 7.079970652971387e-06, + "loss": 0.6426, + "step": 965 + }, + { + "epoch": 0.021263482280431432, + "grad_norm": 2.400117074373797, + "learning_rate": 7.087307410124725e-06, + "loss": 0.4354, + "step": 966 + }, + { + "epoch": 0.021285494166850098, + "grad_norm": 2.2103597585129475, + "learning_rate": 7.0946441672780644e-06, + "loss": 0.3912, + "step": 967 + }, + { + "epoch": 0.021307506053268765, + "grad_norm": 2.16904821005984, + "learning_rate": 7.101980924431402e-06, + "loss": 0.4043, + "step": 968 + }, + { + "epoch": 0.02132951793968743, + "grad_norm": 0.6707425069484165, + "learning_rate": 7.10931768158474e-06, + "loss": 0.6429, + "step": 969 + }, + { + "epoch": 0.021351529826106098, + "grad_norm": 3.7384454772845617, + "learning_rate": 7.116654438738078e-06, + "loss": 0.4944, + "step": 970 + }, + { + "epoch": 0.021373541712524764, + "grad_norm": 3.291857890769613, + "learning_rate": 7.123991195891417e-06, + "loss": 0.4571, + "step": 971 + }, + { + "epoch": 0.02139555359894343, + "grad_norm": 1.8418710909485987, + "learning_rate": 7.1313279530447546e-06, + "loss": 0.3844, + "step": 972 + }, + { + "epoch": 0.021417565485362097, + "grad_norm": 2.06798046269282, + "learning_rate": 7.138664710198093e-06, + "loss": 0.3746, + "step": 973 + }, + { + "epoch": 0.02143957737178076, + "grad_norm": 6.697108566903209, + "learning_rate": 7.146001467351431e-06, + "loss": 0.5186, + "step": 974 + }, + { + "epoch": 0.021461589258199427, + "grad_norm": 1.8577788395253239, + "learning_rate": 7.15333822450477e-06, + "loss": 0.3909, + "step": 975 + }, + { + "epoch": 0.021483601144618093, + "grad_norm": 2.345614095396413, + "learning_rate": 7.160674981658108e-06, + "loss": 0.5107, + "step": 976 + }, + { + "epoch": 0.02150561303103676, + "grad_norm": 1.6847352001138751, + "learning_rate": 7.1680117388114455e-06, + "loss": 0.4873, + "step": 977 + }, + { + "epoch": 0.021527624917455426, + "grad_norm": 2.402970033154779, + "learning_rate": 7.175348495964785e-06, + "loss": 0.5238, + "step": 978 + }, + { + "epoch": 0.021549636803874093, + "grad_norm": 2.1701262972476356, + "learning_rate": 7.182685253118123e-06, + "loss": 0.3793, + "step": 979 + }, + { + "epoch": 0.02157164869029276, + "grad_norm": 1.9627465202255214, + "learning_rate": 7.19002201027146e-06, + "loss": 0.495, + "step": 980 + }, + { + "epoch": 0.021593660576711426, + "grad_norm": 2.120185197339168, + "learning_rate": 7.197358767424799e-06, + "loss": 0.4142, + "step": 981 + }, + { + "epoch": 0.02161567246313009, + "grad_norm": 2.7980791340239066, + "learning_rate": 7.2046955245781365e-06, + "loss": 0.3708, + "step": 982 + }, + { + "epoch": 0.021637684349548755, + "grad_norm": 2.355240812878928, + "learning_rate": 7.212032281731475e-06, + "loss": 0.4675, + "step": 983 + }, + { + "epoch": 0.021659696235967422, + "grad_norm": 1.8137603716622388, + "learning_rate": 7.2193690388848136e-06, + "loss": 0.3916, + "step": 984 + }, + { + "epoch": 0.02168170812238609, + "grad_norm": 2.485102965781405, + "learning_rate": 7.226705796038151e-06, + "loss": 0.4734, + "step": 985 + }, + { + "epoch": 0.021703720008804755, + "grad_norm": 2.4837506207485904, + "learning_rate": 7.234042553191491e-06, + "loss": 0.479, + "step": 986 + }, + { + "epoch": 0.02172573189522342, + "grad_norm": 2.589242968029375, + "learning_rate": 7.241379310344828e-06, + "loss": 0.4783, + "step": 987 + }, + { + "epoch": 0.021747743781642088, + "grad_norm": 2.47313819799128, + "learning_rate": 7.248716067498166e-06, + "loss": 0.3907, + "step": 988 + }, + { + "epoch": 0.021769755668060754, + "grad_norm": 2.396724090088687, + "learning_rate": 7.2560528246515045e-06, + "loss": 0.3847, + "step": 989 + }, + { + "epoch": 0.021791767554479417, + "grad_norm": 1.9978194408616108, + "learning_rate": 7.263389581804843e-06, + "loss": 0.4255, + "step": 990 + }, + { + "epoch": 0.021813779440898084, + "grad_norm": 0.7252469758955891, + "learning_rate": 7.270726338958181e-06, + "loss": 0.6304, + "step": 991 + }, + { + "epoch": 0.02183579132731675, + "grad_norm": 3.3533145756484206, + "learning_rate": 7.278063096111519e-06, + "loss": 0.4401, + "step": 992 + }, + { + "epoch": 0.021857803213735417, + "grad_norm": 0.6485973436295202, + "learning_rate": 7.285399853264857e-06, + "loss": 0.6148, + "step": 993 + }, + { + "epoch": 0.021879815100154083, + "grad_norm": 1.8473985133890198, + "learning_rate": 7.292736610418196e-06, + "loss": 0.3694, + "step": 994 + }, + { + "epoch": 0.02190182698657275, + "grad_norm": 0.6456920852629261, + "learning_rate": 7.300073367571534e-06, + "loss": 0.6503, + "step": 995 + }, + { + "epoch": 0.021923838872991416, + "grad_norm": 2.0821025123195787, + "learning_rate": 7.307410124724872e-06, + "loss": 0.4497, + "step": 996 + }, + { + "epoch": 0.021945850759410083, + "grad_norm": 2.137904882144331, + "learning_rate": 7.314746881878211e-06, + "loss": 0.4402, + "step": 997 + }, + { + "epoch": 0.021967862645828746, + "grad_norm": 0.7050705768888877, + "learning_rate": 7.322083639031549e-06, + "loss": 0.6412, + "step": 998 + }, + { + "epoch": 0.021989874532247412, + "grad_norm": 1.6551919720120278, + "learning_rate": 7.3294203961848865e-06, + "loss": 0.4579, + "step": 999 + }, + { + "epoch": 0.02201188641866608, + "grad_norm": 0.6840442112704752, + "learning_rate": 7.336757153338225e-06, + "loss": 0.6208, + "step": 1000 + }, + { + "epoch": 0.022033898305084745, + "grad_norm": 2.1711906954380886, + "learning_rate": 7.3440939104915636e-06, + "loss": 0.3864, + "step": 1001 + }, + { + "epoch": 0.022055910191503412, + "grad_norm": 2.1175388949817995, + "learning_rate": 7.351430667644902e-06, + "loss": 0.4637, + "step": 1002 + }, + { + "epoch": 0.02207792207792208, + "grad_norm": 1.724896558746446, + "learning_rate": 7.35876742479824e-06, + "loss": 0.3734, + "step": 1003 + }, + { + "epoch": 0.022099933964340745, + "grad_norm": 2.731099320168181, + "learning_rate": 7.3661041819515775e-06, + "loss": 0.4952, + "step": 1004 + }, + { + "epoch": 0.02212194585075941, + "grad_norm": 9.350419328063113, + "learning_rate": 7.373440939104917e-06, + "loss": 0.4401, + "step": 1005 + }, + { + "epoch": 0.022143957737178078, + "grad_norm": 3.3049484339245954, + "learning_rate": 7.3807776962582545e-06, + "loss": 0.4505, + "step": 1006 + }, + { + "epoch": 0.02216596962359674, + "grad_norm": 1.6526434317406034, + "learning_rate": 7.388114453411592e-06, + "loss": 0.392, + "step": 1007 + }, + { + "epoch": 0.022187981510015407, + "grad_norm": 2.4726306099180997, + "learning_rate": 7.395451210564932e-06, + "loss": 0.4026, + "step": 1008 + }, + { + "epoch": 0.022209993396434074, + "grad_norm": 0.6915685860456834, + "learning_rate": 7.402787967718269e-06, + "loss": 0.6447, + "step": 1009 + }, + { + "epoch": 0.02223200528285274, + "grad_norm": 2.4916442948722555, + "learning_rate": 7.410124724871607e-06, + "loss": 0.577, + "step": 1010 + }, + { + "epoch": 0.022254017169271407, + "grad_norm": 2.4280155942822668, + "learning_rate": 7.4174614820249455e-06, + "loss": 0.3522, + "step": 1011 + }, + { + "epoch": 0.022276029055690073, + "grad_norm": 2.2444239889804702, + "learning_rate": 7.424798239178283e-06, + "loss": 0.3893, + "step": 1012 + }, + { + "epoch": 0.02229804094210874, + "grad_norm": 3.422221615406555, + "learning_rate": 7.4321349963316226e-06, + "loss": 0.4822, + "step": 1013 + }, + { + "epoch": 0.022320052828527406, + "grad_norm": 0.6988048283461012, + "learning_rate": 7.43947175348496e-06, + "loss": 0.6169, + "step": 1014 + }, + { + "epoch": 0.02234206471494607, + "grad_norm": 1.8512607563754035, + "learning_rate": 7.446808510638298e-06, + "loss": 0.3931, + "step": 1015 + }, + { + "epoch": 0.022364076601364736, + "grad_norm": 2.023825293178189, + "learning_rate": 7.454145267791637e-06, + "loss": 0.4018, + "step": 1016 + }, + { + "epoch": 0.022386088487783402, + "grad_norm": 2.461799995584084, + "learning_rate": 7.461482024944975e-06, + "loss": 0.4266, + "step": 1017 + }, + { + "epoch": 0.02240810037420207, + "grad_norm": 2.3261203880514136, + "learning_rate": 7.468818782098313e-06, + "loss": 0.4825, + "step": 1018 + }, + { + "epoch": 0.022430112260620735, + "grad_norm": 2.397353500695767, + "learning_rate": 7.476155539251651e-06, + "loss": 0.4614, + "step": 1019 + }, + { + "epoch": 0.022452124147039402, + "grad_norm": 2.4606074723445395, + "learning_rate": 7.48349229640499e-06, + "loss": 0.5093, + "step": 1020 + }, + { + "epoch": 0.02247413603345807, + "grad_norm": 1.838189078538969, + "learning_rate": 7.490829053558328e-06, + "loss": 0.4687, + "step": 1021 + }, + { + "epoch": 0.022496147919876735, + "grad_norm": 2.4679139904606218, + "learning_rate": 7.498165810711666e-06, + "loss": 0.3531, + "step": 1022 + }, + { + "epoch": 0.022518159806295398, + "grad_norm": 4.747057540368027, + "learning_rate": 7.505502567865004e-06, + "loss": 0.3964, + "step": 1023 + }, + { + "epoch": 0.022540171692714064, + "grad_norm": 2.097271635908076, + "learning_rate": 7.512839325018343e-06, + "loss": 0.4288, + "step": 1024 + }, + { + "epoch": 0.02256218357913273, + "grad_norm": 2.2305215595303234, + "learning_rate": 7.520176082171681e-06, + "loss": 0.3778, + "step": 1025 + }, + { + "epoch": 0.022584195465551397, + "grad_norm": 1.685233665979835, + "learning_rate": 7.527512839325018e-06, + "loss": 0.341, + "step": 1026 + }, + { + "epoch": 0.022606207351970064, + "grad_norm": 2.116476418248822, + "learning_rate": 7.534849596478358e-06, + "loss": 0.3772, + "step": 1027 + }, + { + "epoch": 0.02262821923838873, + "grad_norm": 2.868298684862285, + "learning_rate": 7.5421863536316955e-06, + "loss": 0.3884, + "step": 1028 + }, + { + "epoch": 0.022650231124807397, + "grad_norm": 1.8825043440246054, + "learning_rate": 7.549523110785033e-06, + "loss": 0.4255, + "step": 1029 + }, + { + "epoch": 0.022672243011226063, + "grad_norm": 1.998565964491035, + "learning_rate": 7.556859867938372e-06, + "loss": 0.5264, + "step": 1030 + }, + { + "epoch": 0.022694254897644726, + "grad_norm": 1.874308504404801, + "learning_rate": 7.56419662509171e-06, + "loss": 0.5334, + "step": 1031 + }, + { + "epoch": 0.022716266784063393, + "grad_norm": 2.382521168292167, + "learning_rate": 7.571533382245049e-06, + "loss": 0.4459, + "step": 1032 + }, + { + "epoch": 0.02273827867048206, + "grad_norm": 1.713206342163148, + "learning_rate": 7.5788701393983865e-06, + "loss": 0.3858, + "step": 1033 + }, + { + "epoch": 0.022760290556900726, + "grad_norm": 2.747263180893162, + "learning_rate": 7.586206896551724e-06, + "loss": 0.4909, + "step": 1034 + }, + { + "epoch": 0.022782302443319392, + "grad_norm": 1.7552660205484873, + "learning_rate": 7.5935436537050635e-06, + "loss": 0.4047, + "step": 1035 + }, + { + "epoch": 0.02280431432973806, + "grad_norm": 1.7908242746454723, + "learning_rate": 7.600880410858401e-06, + "loss": 0.4638, + "step": 1036 + }, + { + "epoch": 0.022826326216156725, + "grad_norm": 2.171455689649461, + "learning_rate": 7.608217168011739e-06, + "loss": 0.4174, + "step": 1037 + }, + { + "epoch": 0.022848338102575392, + "grad_norm": 0.7365538675435875, + "learning_rate": 7.6155539251650774e-06, + "loss": 0.6512, + "step": 1038 + }, + { + "epoch": 0.02287034998899406, + "grad_norm": 3.333483505548332, + "learning_rate": 7.622890682318416e-06, + "loss": 0.4571, + "step": 1039 + }, + { + "epoch": 0.02289236187541272, + "grad_norm": 0.6638121886246129, + "learning_rate": 7.630227439471754e-06, + "loss": 0.5998, + "step": 1040 + }, + { + "epoch": 0.022914373761831388, + "grad_norm": 2.752635506750744, + "learning_rate": 7.637564196625093e-06, + "loss": 0.3629, + "step": 1041 + }, + { + "epoch": 0.022936385648250054, + "grad_norm": 0.6511824961383198, + "learning_rate": 7.64490095377843e-06, + "loss": 0.6278, + "step": 1042 + }, + { + "epoch": 0.02295839753466872, + "grad_norm": 2.9714197533056845, + "learning_rate": 7.652237710931768e-06, + "loss": 0.4758, + "step": 1043 + }, + { + "epoch": 0.022980409421087387, + "grad_norm": 1.8112512555021285, + "learning_rate": 7.659574468085107e-06, + "loss": 0.4478, + "step": 1044 + }, + { + "epoch": 0.023002421307506054, + "grad_norm": 1.8621017795023778, + "learning_rate": 7.666911225238445e-06, + "loss": 0.4482, + "step": 1045 + }, + { + "epoch": 0.02302443319392472, + "grad_norm": 0.7397924991801836, + "learning_rate": 7.674247982391784e-06, + "loss": 0.6585, + "step": 1046 + }, + { + "epoch": 0.023046445080343387, + "grad_norm": 4.141249448948716, + "learning_rate": 7.68158473954512e-06, + "loss": 0.455, + "step": 1047 + }, + { + "epoch": 0.02306845696676205, + "grad_norm": 1.7612153519652878, + "learning_rate": 7.688921496698461e-06, + "loss": 0.4276, + "step": 1048 + }, + { + "epoch": 0.023090468853180716, + "grad_norm": 1.5720466044091883, + "learning_rate": 7.696258253851798e-06, + "loss": 0.3629, + "step": 1049 + }, + { + "epoch": 0.023112480739599383, + "grad_norm": 7.9930846490195995, + "learning_rate": 7.703595011005136e-06, + "loss": 0.3816, + "step": 1050 + }, + { + "epoch": 0.02313449262601805, + "grad_norm": 1.4919937886971286, + "learning_rate": 7.710931768158475e-06, + "loss": 0.4275, + "step": 1051 + }, + { + "epoch": 0.023156504512436716, + "grad_norm": 1.4747795424560142, + "learning_rate": 7.718268525311814e-06, + "loss": 0.3854, + "step": 1052 + }, + { + "epoch": 0.023178516398855382, + "grad_norm": 2.0149295409311163, + "learning_rate": 7.72560528246515e-06, + "loss": 0.4332, + "step": 1053 + }, + { + "epoch": 0.02320052828527405, + "grad_norm": 2.058937369553756, + "learning_rate": 7.732942039618489e-06, + "loss": 0.4529, + "step": 1054 + }, + { + "epoch": 0.023222540171692715, + "grad_norm": 2.21790566666291, + "learning_rate": 7.740278796771827e-06, + "loss": 0.4433, + "step": 1055 + }, + { + "epoch": 0.02324455205811138, + "grad_norm": 2.788722783548533, + "learning_rate": 7.747615553925166e-06, + "loss": 0.3518, + "step": 1056 + }, + { + "epoch": 0.023266563944530045, + "grad_norm": 1.820044800105777, + "learning_rate": 7.754952311078504e-06, + "loss": 0.3838, + "step": 1057 + }, + { + "epoch": 0.02328857583094871, + "grad_norm": 1.9618236698402514, + "learning_rate": 7.762289068231841e-06, + "loss": 0.3896, + "step": 1058 + }, + { + "epoch": 0.023310587717367378, + "grad_norm": 0.7726208011377583, + "learning_rate": 7.769625825385182e-06, + "loss": 0.5906, + "step": 1059 + }, + { + "epoch": 0.023332599603786044, + "grad_norm": 0.7312716990516364, + "learning_rate": 7.776962582538518e-06, + "loss": 0.6494, + "step": 1060 + }, + { + "epoch": 0.02335461149020471, + "grad_norm": 2.017128741589902, + "learning_rate": 7.784299339691857e-06, + "loss": 0.5806, + "step": 1061 + }, + { + "epoch": 0.023376623376623377, + "grad_norm": 6.386697703183041, + "learning_rate": 7.791636096845195e-06, + "loss": 0.4588, + "step": 1062 + }, + { + "epoch": 0.023398635263042044, + "grad_norm": 4.197454420068049, + "learning_rate": 7.798972853998534e-06, + "loss": 0.4453, + "step": 1063 + }, + { + "epoch": 0.02342064714946071, + "grad_norm": 2.1024735187959362, + "learning_rate": 7.80630961115187e-06, + "loss": 0.3966, + "step": 1064 + }, + { + "epoch": 0.023442659035879373, + "grad_norm": 1.5674213847329987, + "learning_rate": 7.81364636830521e-06, + "loss": 0.392, + "step": 1065 + }, + { + "epoch": 0.02346467092229804, + "grad_norm": 5.808056809944783, + "learning_rate": 7.820983125458548e-06, + "loss": 0.3966, + "step": 1066 + }, + { + "epoch": 0.023486682808716706, + "grad_norm": 1.0800711563320875, + "learning_rate": 7.828319882611886e-06, + "loss": 0.6503, + "step": 1067 + }, + { + "epoch": 0.023508694695135373, + "grad_norm": 2.0351252334589653, + "learning_rate": 7.835656639765225e-06, + "loss": 0.391, + "step": 1068 + }, + { + "epoch": 0.02353070658155404, + "grad_norm": 0.7681340918971201, + "learning_rate": 7.842993396918562e-06, + "loss": 0.6458, + "step": 1069 + }, + { + "epoch": 0.023552718467972706, + "grad_norm": 0.6652745281091009, + "learning_rate": 7.850330154071902e-06, + "loss": 0.6037, + "step": 1070 + }, + { + "epoch": 0.023574730354391372, + "grad_norm": 1.9050132308229262, + "learning_rate": 7.857666911225239e-06, + "loss": 0.4563, + "step": 1071 + }, + { + "epoch": 0.02359674224081004, + "grad_norm": 4.2807795189860185, + "learning_rate": 7.865003668378577e-06, + "loss": 0.3618, + "step": 1072 + }, + { + "epoch": 0.023618754127228702, + "grad_norm": 2.1598420009609836, + "learning_rate": 7.872340425531916e-06, + "loss": 0.3569, + "step": 1073 + }, + { + "epoch": 0.02364076601364737, + "grad_norm": 2.0098097442775176, + "learning_rate": 7.879677182685253e-06, + "loss": 0.4034, + "step": 1074 + }, + { + "epoch": 0.023662777900066035, + "grad_norm": 1.7883798006226466, + "learning_rate": 7.887013939838593e-06, + "loss": 0.3173, + "step": 1075 + }, + { + "epoch": 0.0236847897864847, + "grad_norm": 2.1692200707677545, + "learning_rate": 7.89435069699193e-06, + "loss": 0.388, + "step": 1076 + }, + { + "epoch": 0.023706801672903368, + "grad_norm": 1.483272107923597, + "learning_rate": 7.901687454145268e-06, + "loss": 0.4054, + "step": 1077 + }, + { + "epoch": 0.023728813559322035, + "grad_norm": 2.017871416878195, + "learning_rate": 7.909024211298607e-06, + "loss": 0.487, + "step": 1078 + }, + { + "epoch": 0.0237508254457407, + "grad_norm": 1.720866233542981, + "learning_rate": 7.916360968451945e-06, + "loss": 0.3565, + "step": 1079 + }, + { + "epoch": 0.023772837332159368, + "grad_norm": 1.9176716156409352, + "learning_rate": 7.923697725605282e-06, + "loss": 0.4271, + "step": 1080 + }, + { + "epoch": 0.02379484921857803, + "grad_norm": 5.647560650877725, + "learning_rate": 7.93103448275862e-06, + "loss": 0.4033, + "step": 1081 + }, + { + "epoch": 0.023816861104996697, + "grad_norm": 2.2805409752847106, + "learning_rate": 7.93837123991196e-06, + "loss": 0.4355, + "step": 1082 + }, + { + "epoch": 0.023838872991415364, + "grad_norm": 1.9857233771866203, + "learning_rate": 7.945707997065298e-06, + "loss": 0.3911, + "step": 1083 + }, + { + "epoch": 0.02386088487783403, + "grad_norm": 2.1071789563778833, + "learning_rate": 7.953044754218636e-06, + "loss": 0.4017, + "step": 1084 + }, + { + "epoch": 0.023882896764252697, + "grad_norm": 2.6490001948501756, + "learning_rate": 7.960381511371973e-06, + "loss": 0.4031, + "step": 1085 + }, + { + "epoch": 0.023904908650671363, + "grad_norm": 2.240658596655648, + "learning_rate": 7.967718268525313e-06, + "loss": 0.4256, + "step": 1086 + }, + { + "epoch": 0.02392692053709003, + "grad_norm": 2.281662534033413, + "learning_rate": 7.97505502567865e-06, + "loss": 0.4899, + "step": 1087 + }, + { + "epoch": 0.023948932423508696, + "grad_norm": 2.7798199655120266, + "learning_rate": 7.982391782831989e-06, + "loss": 0.5003, + "step": 1088 + }, + { + "epoch": 0.02397094430992736, + "grad_norm": 2.061700073640866, + "learning_rate": 7.989728539985327e-06, + "loss": 0.4734, + "step": 1089 + }, + { + "epoch": 0.023992956196346026, + "grad_norm": 2.310720669657178, + "learning_rate": 7.997065297138666e-06, + "loss": 0.7001, + "step": 1090 + }, + { + "epoch": 0.024014968082764692, + "grad_norm": 2.575652653117501, + "learning_rate": 8.004402054292003e-06, + "loss": 0.4302, + "step": 1091 + }, + { + "epoch": 0.02403697996918336, + "grad_norm": 2.095567213958053, + "learning_rate": 8.011738811445341e-06, + "loss": 0.4795, + "step": 1092 + }, + { + "epoch": 0.024058991855602025, + "grad_norm": 1.9576235534192947, + "learning_rate": 8.01907556859868e-06, + "loss": 0.5021, + "step": 1093 + }, + { + "epoch": 0.02408100374202069, + "grad_norm": 3.814913068078046, + "learning_rate": 8.026412325752018e-06, + "loss": 0.2893, + "step": 1094 + }, + { + "epoch": 0.024103015628439358, + "grad_norm": 2.247807969754488, + "learning_rate": 8.033749082905357e-06, + "loss": 0.3182, + "step": 1095 + }, + { + "epoch": 0.024125027514858025, + "grad_norm": 2.395812798144611, + "learning_rate": 8.041085840058694e-06, + "loss": 0.3953, + "step": 1096 + }, + { + "epoch": 0.02414703940127669, + "grad_norm": 1.2688831775920077, + "learning_rate": 8.048422597212034e-06, + "loss": 0.6573, + "step": 1097 + }, + { + "epoch": 0.024169051287695354, + "grad_norm": 1.9609302210599273, + "learning_rate": 8.05575935436537e-06, + "loss": 0.4002, + "step": 1098 + }, + { + "epoch": 0.02419106317411402, + "grad_norm": 2.6281254759325496, + "learning_rate": 8.06309611151871e-06, + "loss": 0.4255, + "step": 1099 + }, + { + "epoch": 0.024213075060532687, + "grad_norm": 1.6567134600296434, + "learning_rate": 8.070432868672048e-06, + "loss": 0.3892, + "step": 1100 + }, + { + "epoch": 0.024235086946951354, + "grad_norm": 1.175452527553573, + "learning_rate": 8.077769625825386e-06, + "loss": 0.6318, + "step": 1101 + }, + { + "epoch": 0.02425709883337002, + "grad_norm": 2.150694976457689, + "learning_rate": 8.085106382978723e-06, + "loss": 0.3186, + "step": 1102 + }, + { + "epoch": 0.024279110719788687, + "grad_norm": 2.9436565797417624, + "learning_rate": 8.092443140132062e-06, + "loss": 0.4868, + "step": 1103 + }, + { + "epoch": 0.024301122606207353, + "grad_norm": 2.971239768724241, + "learning_rate": 8.0997798972854e-06, + "loss": 0.4131, + "step": 1104 + }, + { + "epoch": 0.02432313449262602, + "grad_norm": 1.829283943224936, + "learning_rate": 8.107116654438739e-06, + "loss": 0.3953, + "step": 1105 + }, + { + "epoch": 0.024345146379044683, + "grad_norm": 2.47430482143781, + "learning_rate": 8.114453411592077e-06, + "loss": 0.5254, + "step": 1106 + }, + { + "epoch": 0.02436715826546335, + "grad_norm": 3.4246240390560887, + "learning_rate": 8.121790168745414e-06, + "loss": 0.5049, + "step": 1107 + }, + { + "epoch": 0.024389170151882016, + "grad_norm": 1.8707368126884163, + "learning_rate": 8.129126925898754e-06, + "loss": 0.4006, + "step": 1108 + }, + { + "epoch": 0.024411182038300682, + "grad_norm": 1.7904238124306457, + "learning_rate": 8.136463683052091e-06, + "loss": 0.3812, + "step": 1109 + }, + { + "epoch": 0.02443319392471935, + "grad_norm": 2.2061516519152233, + "learning_rate": 8.14380044020543e-06, + "loss": 0.3817, + "step": 1110 + }, + { + "epoch": 0.024455205811138015, + "grad_norm": 2.770492173788608, + "learning_rate": 8.151137197358768e-06, + "loss": 0.4872, + "step": 1111 + }, + { + "epoch": 0.02447721769755668, + "grad_norm": 1.8312217866909606, + "learning_rate": 8.158473954512107e-06, + "loss": 0.5196, + "step": 1112 + }, + { + "epoch": 0.024499229583975348, + "grad_norm": 1.9421712687318209, + "learning_rate": 8.165810711665445e-06, + "loss": 0.4067, + "step": 1113 + }, + { + "epoch": 0.02452124147039401, + "grad_norm": 2.0840668528945456, + "learning_rate": 8.173147468818782e-06, + "loss": 0.3479, + "step": 1114 + }, + { + "epoch": 0.024543253356812678, + "grad_norm": 4.4701751632716356, + "learning_rate": 8.18048422597212e-06, + "loss": 0.3808, + "step": 1115 + }, + { + "epoch": 0.024565265243231344, + "grad_norm": 1.679750873049837, + "learning_rate": 8.18782098312546e-06, + "loss": 0.3823, + "step": 1116 + }, + { + "epoch": 0.02458727712965001, + "grad_norm": 2.5423145565397958, + "learning_rate": 8.195157740278798e-06, + "loss": 0.4394, + "step": 1117 + }, + { + "epoch": 0.024609289016068677, + "grad_norm": 2.234201074493842, + "learning_rate": 8.202494497432135e-06, + "loss": 0.4565, + "step": 1118 + }, + { + "epoch": 0.024631300902487344, + "grad_norm": 3.3833423015116035, + "learning_rate": 8.209831254585475e-06, + "loss": 0.3289, + "step": 1119 + }, + { + "epoch": 0.02465331278890601, + "grad_norm": 1.8151523055979675, + "learning_rate": 8.217168011738812e-06, + "loss": 0.4723, + "step": 1120 + }, + { + "epoch": 0.024675324675324677, + "grad_norm": 2.712180979921609, + "learning_rate": 8.22450476889215e-06, + "loss": 0.4017, + "step": 1121 + }, + { + "epoch": 0.024697336561743343, + "grad_norm": 1.6796984095935474, + "learning_rate": 8.231841526045489e-06, + "loss": 0.3608, + "step": 1122 + }, + { + "epoch": 0.024719348448162006, + "grad_norm": 1.8059964476164574, + "learning_rate": 8.239178283198827e-06, + "loss": 0.4745, + "step": 1123 + }, + { + "epoch": 0.024741360334580673, + "grad_norm": 1.7156318985427501, + "learning_rate": 8.246515040352166e-06, + "loss": 0.4059, + "step": 1124 + }, + { + "epoch": 0.02476337222099934, + "grad_norm": 1.0013078100444912, + "learning_rate": 8.253851797505503e-06, + "loss": 0.622, + "step": 1125 + }, + { + "epoch": 0.024785384107418006, + "grad_norm": 1.9864229162971876, + "learning_rate": 8.261188554658841e-06, + "loss": 0.3792, + "step": 1126 + }, + { + "epoch": 0.024807395993836672, + "grad_norm": 2.8315061433631845, + "learning_rate": 8.26852531181218e-06, + "loss": 0.4484, + "step": 1127 + }, + { + "epoch": 0.02482940788025534, + "grad_norm": 2.0029930868053714, + "learning_rate": 8.275862068965518e-06, + "loss": 0.3246, + "step": 1128 + }, + { + "epoch": 0.024851419766674005, + "grad_norm": 2.358492535771842, + "learning_rate": 8.283198826118855e-06, + "loss": 0.4456, + "step": 1129 + }, + { + "epoch": 0.02487343165309267, + "grad_norm": 2.09097913949541, + "learning_rate": 8.290535583272194e-06, + "loss": 0.3467, + "step": 1130 + }, + { + "epoch": 0.024895443539511335, + "grad_norm": 0.698973166367171, + "learning_rate": 8.297872340425532e-06, + "loss": 0.6357, + "step": 1131 + }, + { + "epoch": 0.02491745542593, + "grad_norm": 1.766961264275285, + "learning_rate": 8.30520909757887e-06, + "loss": 0.399, + "step": 1132 + }, + { + "epoch": 0.024939467312348668, + "grad_norm": 1.5779581314741822, + "learning_rate": 8.31254585473221e-06, + "loss": 0.3798, + "step": 1133 + }, + { + "epoch": 0.024961479198767334, + "grad_norm": 2.0049818126631145, + "learning_rate": 8.319882611885546e-06, + "loss": 0.4769, + "step": 1134 + }, + { + "epoch": 0.024983491085186, + "grad_norm": 1.6228755869316955, + "learning_rate": 8.327219369038886e-06, + "loss": 0.4242, + "step": 1135 + }, + { + "epoch": 0.025005502971604667, + "grad_norm": 3.1131677810742713, + "learning_rate": 8.334556126192223e-06, + "loss": 0.4241, + "step": 1136 + }, + { + "epoch": 0.025027514858023334, + "grad_norm": 1.6326896716911272, + "learning_rate": 8.341892883345562e-06, + "loss": 0.4481, + "step": 1137 + }, + { + "epoch": 0.025049526744442, + "grad_norm": 0.8156482436122307, + "learning_rate": 8.3492296404989e-06, + "loss": 0.615, + "step": 1138 + }, + { + "epoch": 0.025071538630860663, + "grad_norm": 2.1096469599842487, + "learning_rate": 8.356566397652239e-06, + "loss": 0.4574, + "step": 1139 + }, + { + "epoch": 0.02509355051727933, + "grad_norm": 1.737904871953456, + "learning_rate": 8.363903154805577e-06, + "loss": 0.5099, + "step": 1140 + }, + { + "epoch": 0.025115562403697996, + "grad_norm": 1.7489342904841327, + "learning_rate": 8.371239911958914e-06, + "loss": 0.3548, + "step": 1141 + }, + { + "epoch": 0.025137574290116663, + "grad_norm": 2.8206452082068565, + "learning_rate": 8.378576669112253e-06, + "loss": 0.3952, + "step": 1142 + }, + { + "epoch": 0.02515958617653533, + "grad_norm": 2.1884708604524854, + "learning_rate": 8.385913426265591e-06, + "loss": 0.4456, + "step": 1143 + }, + { + "epoch": 0.025181598062953996, + "grad_norm": 2.893209210011646, + "learning_rate": 8.39325018341893e-06, + "loss": 0.4566, + "step": 1144 + }, + { + "epoch": 0.025203609949372662, + "grad_norm": 2.3233670635971673, + "learning_rate": 8.400586940572267e-06, + "loss": 0.4283, + "step": 1145 + }, + { + "epoch": 0.02522562183579133, + "grad_norm": 1.9257585680311686, + "learning_rate": 8.407923697725607e-06, + "loss": 0.3943, + "step": 1146 + }, + { + "epoch": 0.02524763372220999, + "grad_norm": 1.9210444566649743, + "learning_rate": 8.415260454878944e-06, + "loss": 0.3708, + "step": 1147 + }, + { + "epoch": 0.025269645608628658, + "grad_norm": 2.108496600234457, + "learning_rate": 8.422597212032282e-06, + "loss": 0.2818, + "step": 1148 + }, + { + "epoch": 0.025291657495047325, + "grad_norm": 1.6918738744336532, + "learning_rate": 8.42993396918562e-06, + "loss": 0.3827, + "step": 1149 + }, + { + "epoch": 0.02531366938146599, + "grad_norm": 0.6957368556700135, + "learning_rate": 8.43727072633896e-06, + "loss": 0.6114, + "step": 1150 + }, + { + "epoch": 0.025335681267884658, + "grad_norm": 1.8493722386162965, + "learning_rate": 8.444607483492298e-06, + "loss": 0.3781, + "step": 1151 + }, + { + "epoch": 0.025357693154303324, + "grad_norm": 0.7003724089025277, + "learning_rate": 8.451944240645635e-06, + "loss": 0.6389, + "step": 1152 + }, + { + "epoch": 0.02537970504072199, + "grad_norm": 0.6416707248136718, + "learning_rate": 8.459280997798973e-06, + "loss": 0.6085, + "step": 1153 + }, + { + "epoch": 0.025401716927140657, + "grad_norm": 2.0666163180543595, + "learning_rate": 8.466617754952312e-06, + "loss": 0.543, + "step": 1154 + }, + { + "epoch": 0.025423728813559324, + "grad_norm": 1.9597330059434397, + "learning_rate": 8.47395451210565e-06, + "loss": 0.4657, + "step": 1155 + }, + { + "epoch": 0.025445740699977987, + "grad_norm": 2.0326994171463313, + "learning_rate": 8.481291269258987e-06, + "loss": 0.3633, + "step": 1156 + }, + { + "epoch": 0.025467752586396653, + "grad_norm": 1.7246083165930495, + "learning_rate": 8.488628026412327e-06, + "loss": 0.4734, + "step": 1157 + }, + { + "epoch": 0.02548976447281532, + "grad_norm": 1.918180099930028, + "learning_rate": 8.495964783565664e-06, + "loss": 0.5018, + "step": 1158 + }, + { + "epoch": 0.025511776359233986, + "grad_norm": 1.92203418809124, + "learning_rate": 8.503301540719003e-06, + "loss": 0.5138, + "step": 1159 + }, + { + "epoch": 0.025533788245652653, + "grad_norm": 1.8685529435971844, + "learning_rate": 8.510638297872341e-06, + "loss": 0.4973, + "step": 1160 + }, + { + "epoch": 0.02555580013207132, + "grad_norm": 0.8825774986358083, + "learning_rate": 8.51797505502568e-06, + "loss": 0.6124, + "step": 1161 + }, + { + "epoch": 0.025577812018489986, + "grad_norm": 2.047147829348496, + "learning_rate": 8.525311812179018e-06, + "loss": 0.4294, + "step": 1162 + }, + { + "epoch": 0.025599823904908652, + "grad_norm": 1.5705668148705698, + "learning_rate": 8.532648569332355e-06, + "loss": 0.4507, + "step": 1163 + }, + { + "epoch": 0.025621835791327315, + "grad_norm": 1.8169503089785752, + "learning_rate": 8.539985326485694e-06, + "loss": 0.4923, + "step": 1164 + }, + { + "epoch": 0.02564384767774598, + "grad_norm": 1.8285838221590391, + "learning_rate": 8.547322083639032e-06, + "loss": 0.395, + "step": 1165 + }, + { + "epoch": 0.025665859564164648, + "grad_norm": 1.9081509214762742, + "learning_rate": 8.55465884079237e-06, + "loss": 0.453, + "step": 1166 + }, + { + "epoch": 0.025687871450583315, + "grad_norm": 1.9157843952811973, + "learning_rate": 8.56199559794571e-06, + "loss": 0.4383, + "step": 1167 + }, + { + "epoch": 0.02570988333700198, + "grad_norm": 1.2999054047418797, + "learning_rate": 8.569332355099048e-06, + "loss": 0.2469, + "step": 1168 + }, + { + "epoch": 0.025731895223420648, + "grad_norm": 2.403588179602473, + "learning_rate": 8.576669112252385e-06, + "loss": 0.4058, + "step": 1169 + }, + { + "epoch": 0.025753907109839314, + "grad_norm": 1.6299255873330434, + "learning_rate": 8.584005869405723e-06, + "loss": 0.3618, + "step": 1170 + }, + { + "epoch": 0.02577591899625798, + "grad_norm": 1.7262464036689236, + "learning_rate": 8.591342626559062e-06, + "loss": 0.4474, + "step": 1171 + }, + { + "epoch": 0.025797930882676644, + "grad_norm": 1.9435866396758434, + "learning_rate": 8.5986793837124e-06, + "loss": 0.4507, + "step": 1172 + }, + { + "epoch": 0.02581994276909531, + "grad_norm": 3.359502390766612, + "learning_rate": 8.606016140865739e-06, + "loss": 0.4392, + "step": 1173 + }, + { + "epoch": 0.025841954655513977, + "grad_norm": 1.510777971170063, + "learning_rate": 8.613352898019076e-06, + "loss": 0.3615, + "step": 1174 + }, + { + "epoch": 0.025863966541932643, + "grad_norm": 2.1997021383994833, + "learning_rate": 8.620689655172414e-06, + "loss": 0.4137, + "step": 1175 + }, + { + "epoch": 0.02588597842835131, + "grad_norm": 0.8002794552781006, + "learning_rate": 8.628026412325753e-06, + "loss": 0.6612, + "step": 1176 + }, + { + "epoch": 0.025907990314769976, + "grad_norm": 0.7078366249702354, + "learning_rate": 8.635363169479091e-06, + "loss": 0.6126, + "step": 1177 + }, + { + "epoch": 0.025930002201188643, + "grad_norm": 2.5646860708222663, + "learning_rate": 8.64269992663243e-06, + "loss": 0.4484, + "step": 1178 + }, + { + "epoch": 0.02595201408760731, + "grad_norm": 1.8873263174818458, + "learning_rate": 8.650036683785768e-06, + "loss": 0.4049, + "step": 1179 + }, + { + "epoch": 0.025974025974025976, + "grad_norm": 2.053015354657751, + "learning_rate": 8.657373440939105e-06, + "loss": 0.4333, + "step": 1180 + }, + { + "epoch": 0.02599603786044464, + "grad_norm": 1.6505014880769122, + "learning_rate": 8.664710198092444e-06, + "loss": 0.3641, + "step": 1181 + }, + { + "epoch": 0.026018049746863305, + "grad_norm": 1.7702493591490647, + "learning_rate": 8.672046955245782e-06, + "loss": 0.373, + "step": 1182 + }, + { + "epoch": 0.026040061633281972, + "grad_norm": 1.6756955500425725, + "learning_rate": 8.679383712399119e-06, + "loss": 0.34, + "step": 1183 + }, + { + "epoch": 0.026062073519700638, + "grad_norm": 2.1270949474605922, + "learning_rate": 8.68672046955246e-06, + "loss": 0.4911, + "step": 1184 + }, + { + "epoch": 0.026084085406119305, + "grad_norm": 2.243567584791854, + "learning_rate": 8.694057226705796e-06, + "loss": 0.4319, + "step": 1185 + }, + { + "epoch": 0.02610609729253797, + "grad_norm": 2.439380137785054, + "learning_rate": 8.701393983859136e-06, + "loss": 0.4008, + "step": 1186 + }, + { + "epoch": 0.026128109178956638, + "grad_norm": 1.7348301459138207, + "learning_rate": 8.708730741012473e-06, + "loss": 0.3799, + "step": 1187 + }, + { + "epoch": 0.026150121065375304, + "grad_norm": 1.9759845239010703, + "learning_rate": 8.716067498165812e-06, + "loss": 0.4476, + "step": 1188 + }, + { + "epoch": 0.026172132951793967, + "grad_norm": 2.2173162295974342, + "learning_rate": 8.72340425531915e-06, + "loss": 0.4163, + "step": 1189 + }, + { + "epoch": 0.026194144838212634, + "grad_norm": 1.71264656618118, + "learning_rate": 8.730741012472487e-06, + "loss": 0.3823, + "step": 1190 + }, + { + "epoch": 0.0262161567246313, + "grad_norm": 15.438164748777336, + "learning_rate": 8.738077769625826e-06, + "loss": 0.3804, + "step": 1191 + }, + { + "epoch": 0.026238168611049967, + "grad_norm": 1.1444263733982944, + "learning_rate": 8.745414526779164e-06, + "loss": 0.6545, + "step": 1192 + }, + { + "epoch": 0.026260180497468633, + "grad_norm": 2.617157641842866, + "learning_rate": 8.752751283932503e-06, + "loss": 0.4289, + "step": 1193 + }, + { + "epoch": 0.0262821923838873, + "grad_norm": 0.8626692082546559, + "learning_rate": 8.760088041085841e-06, + "loss": 0.6329, + "step": 1194 + }, + { + "epoch": 0.026304204270305966, + "grad_norm": 1.920728031499012, + "learning_rate": 8.76742479823918e-06, + "loss": 0.4018, + "step": 1195 + }, + { + "epoch": 0.026326216156724633, + "grad_norm": 2.191286217451945, + "learning_rate": 8.774761555392517e-06, + "loss": 0.4775, + "step": 1196 + }, + { + "epoch": 0.026348228043143296, + "grad_norm": 1.741557683744048, + "learning_rate": 8.782098312545855e-06, + "loss": 0.4334, + "step": 1197 + }, + { + "epoch": 0.026370239929561962, + "grad_norm": 1.9021300530839917, + "learning_rate": 8.789435069699194e-06, + "loss": 0.4578, + "step": 1198 + }, + { + "epoch": 0.02639225181598063, + "grad_norm": 1.8133785162122822, + "learning_rate": 8.796771826852532e-06, + "loss": 0.4414, + "step": 1199 + }, + { + "epoch": 0.026414263702399295, + "grad_norm": 2.1970415757996467, + "learning_rate": 8.80410858400587e-06, + "loss": 0.342, + "step": 1200 + }, + { + "epoch": 0.026436275588817962, + "grad_norm": 3.3559982920508813, + "learning_rate": 8.811445341159208e-06, + "loss": 0.3668, + "step": 1201 + }, + { + "epoch": 0.02645828747523663, + "grad_norm": 2.524568237444157, + "learning_rate": 8.818782098312546e-06, + "loss": 0.4079, + "step": 1202 + }, + { + "epoch": 0.026480299361655295, + "grad_norm": 2.204044430217563, + "learning_rate": 8.826118855465885e-06, + "loss": 0.4875, + "step": 1203 + }, + { + "epoch": 0.02650231124807396, + "grad_norm": 2.0726615807694393, + "learning_rate": 8.833455612619223e-06, + "loss": 0.4106, + "step": 1204 + }, + { + "epoch": 0.026524323134492624, + "grad_norm": 1.6525823554096901, + "learning_rate": 8.840792369772562e-06, + "loss": 0.6687, + "step": 1205 + }, + { + "epoch": 0.02654633502091129, + "grad_norm": 1.828957045757644, + "learning_rate": 8.8481291269259e-06, + "loss": 0.4974, + "step": 1206 + }, + { + "epoch": 0.026568346907329957, + "grad_norm": 1.8408048118172617, + "learning_rate": 8.855465884079237e-06, + "loss": 0.4352, + "step": 1207 + }, + { + "epoch": 0.026590358793748624, + "grad_norm": 2.0177001513559265, + "learning_rate": 8.862802641232576e-06, + "loss": 0.4019, + "step": 1208 + }, + { + "epoch": 0.02661237068016729, + "grad_norm": 1.8551019550182515, + "learning_rate": 8.870139398385914e-06, + "loss": 0.4291, + "step": 1209 + }, + { + "epoch": 0.026634382566585957, + "grad_norm": 2.139492718366993, + "learning_rate": 8.877476155539253e-06, + "loss": 0.5112, + "step": 1210 + }, + { + "epoch": 0.026656394453004623, + "grad_norm": 0.763416844514447, + "learning_rate": 8.884812912692591e-06, + "loss": 0.6235, + "step": 1211 + }, + { + "epoch": 0.02667840633942329, + "grad_norm": 3.5905413878851027, + "learning_rate": 8.892149669845928e-06, + "loss": 0.3961, + "step": 1212 + }, + { + "epoch": 0.026700418225841956, + "grad_norm": 0.7814712413713238, + "learning_rate": 8.899486426999268e-06, + "loss": 0.6546, + "step": 1213 + }, + { + "epoch": 0.02672243011226062, + "grad_norm": 0.7966742419387128, + "learning_rate": 8.906823184152605e-06, + "loss": 0.6289, + "step": 1214 + }, + { + "epoch": 0.026744441998679286, + "grad_norm": 2.0401763106506414, + "learning_rate": 8.914159941305944e-06, + "loss": 0.4636, + "step": 1215 + }, + { + "epoch": 0.026766453885097952, + "grad_norm": 2.5024423764148773, + "learning_rate": 8.921496698459282e-06, + "loss": 0.4251, + "step": 1216 + }, + { + "epoch": 0.02678846577151662, + "grad_norm": 0.7301635438645285, + "learning_rate": 8.92883345561262e-06, + "loss": 0.6101, + "step": 1217 + }, + { + "epoch": 0.026810477657935285, + "grad_norm": 2.4693664362815153, + "learning_rate": 8.936170212765958e-06, + "loss": 0.3271, + "step": 1218 + }, + { + "epoch": 0.026832489544353952, + "grad_norm": 2.2309202590879114, + "learning_rate": 8.943506969919296e-06, + "loss": 0.4395, + "step": 1219 + }, + { + "epoch": 0.02685450143077262, + "grad_norm": 1.7660392946573698, + "learning_rate": 8.950843727072635e-06, + "loss": 0.345, + "step": 1220 + }, + { + "epoch": 0.026876513317191285, + "grad_norm": 1.8481088969004862, + "learning_rate": 8.958180484225973e-06, + "loss": 0.413, + "step": 1221 + }, + { + "epoch": 0.026898525203609948, + "grad_norm": 1.8196116581926542, + "learning_rate": 8.965517241379312e-06, + "loss": 0.3836, + "step": 1222 + }, + { + "epoch": 0.026920537090028614, + "grad_norm": 2.1046566179056025, + "learning_rate": 8.972853998532649e-06, + "loss": 0.4684, + "step": 1223 + }, + { + "epoch": 0.02694254897644728, + "grad_norm": 2.4757480068680358, + "learning_rate": 8.980190755685989e-06, + "loss": 0.418, + "step": 1224 + }, + { + "epoch": 0.026964560862865947, + "grad_norm": 2.003612455703262, + "learning_rate": 8.987527512839326e-06, + "loss": 0.4791, + "step": 1225 + }, + { + "epoch": 0.026986572749284614, + "grad_norm": 2.8120529569603447, + "learning_rate": 8.994864269992664e-06, + "loss": 0.4259, + "step": 1226 + }, + { + "epoch": 0.02700858463570328, + "grad_norm": 0.8861079709523181, + "learning_rate": 9.002201027146003e-06, + "loss": 0.629, + "step": 1227 + }, + { + "epoch": 0.027030596522121947, + "grad_norm": 1.806413720062956, + "learning_rate": 9.009537784299341e-06, + "loss": 0.4333, + "step": 1228 + }, + { + "epoch": 0.027052608408540613, + "grad_norm": 1.8935176868976777, + "learning_rate": 9.016874541452678e-06, + "loss": 0.4859, + "step": 1229 + }, + { + "epoch": 0.027074620294959276, + "grad_norm": 2.576359116249735, + "learning_rate": 9.024211298606017e-06, + "loss": 0.4487, + "step": 1230 + }, + { + "epoch": 0.027096632181377943, + "grad_norm": 1.6535475942024138, + "learning_rate": 9.031548055759355e-06, + "loss": 0.393, + "step": 1231 + }, + { + "epoch": 0.02711864406779661, + "grad_norm": 1.7852610417477959, + "learning_rate": 9.038884812912694e-06, + "loss": 0.438, + "step": 1232 + }, + { + "epoch": 0.027140655954215276, + "grad_norm": 1.8332897671672073, + "learning_rate": 9.046221570066032e-06, + "loss": 0.4734, + "step": 1233 + }, + { + "epoch": 0.027162667840633942, + "grad_norm": 2.064516212461937, + "learning_rate": 9.053558327219369e-06, + "loss": 0.4538, + "step": 1234 + }, + { + "epoch": 0.02718467972705261, + "grad_norm": 2.778311024701245, + "learning_rate": 9.06089508437271e-06, + "loss": 0.4089, + "step": 1235 + }, + { + "epoch": 0.027206691613471275, + "grad_norm": 2.0632650480658357, + "learning_rate": 9.068231841526046e-06, + "loss": 0.3951, + "step": 1236 + }, + { + "epoch": 0.027228703499889942, + "grad_norm": 3.085026657590634, + "learning_rate": 9.075568598679385e-06, + "loss": 0.4357, + "step": 1237 + }, + { + "epoch": 0.02725071538630861, + "grad_norm": 1.6472089717587635, + "learning_rate": 9.082905355832723e-06, + "loss": 0.3308, + "step": 1238 + }, + { + "epoch": 0.02727272727272727, + "grad_norm": 1.7105841005952214, + "learning_rate": 9.09024211298606e-06, + "loss": 0.3653, + "step": 1239 + }, + { + "epoch": 0.027294739159145938, + "grad_norm": 2.2223542903191196, + "learning_rate": 9.0975788701394e-06, + "loss": 0.3986, + "step": 1240 + }, + { + "epoch": 0.027316751045564604, + "grad_norm": 2.3557357212333923, + "learning_rate": 9.104915627292737e-06, + "loss": 0.4445, + "step": 1241 + }, + { + "epoch": 0.02733876293198327, + "grad_norm": 2.0138773117585482, + "learning_rate": 9.112252384446076e-06, + "loss": 0.4036, + "step": 1242 + }, + { + "epoch": 0.027360774818401937, + "grad_norm": 0.7158580990175372, + "learning_rate": 9.119589141599414e-06, + "loss": 0.6317, + "step": 1243 + }, + { + "epoch": 0.027382786704820604, + "grad_norm": 1.7756398838081504, + "learning_rate": 9.126925898752753e-06, + "loss": 0.4065, + "step": 1244 + }, + { + "epoch": 0.02740479859123927, + "grad_norm": 0.7013642101625396, + "learning_rate": 9.13426265590609e-06, + "loss": 0.6588, + "step": 1245 + }, + { + "epoch": 0.027426810477657937, + "grad_norm": 3.007832480472568, + "learning_rate": 9.141599413059428e-06, + "loss": 0.5228, + "step": 1246 + }, + { + "epoch": 0.0274488223640766, + "grad_norm": 0.6656526137928231, + "learning_rate": 9.148936170212767e-06, + "loss": 0.6072, + "step": 1247 + }, + { + "epoch": 0.027470834250495266, + "grad_norm": 2.2172924083153225, + "learning_rate": 9.156272927366105e-06, + "loss": 0.4835, + "step": 1248 + }, + { + "epoch": 0.027492846136913933, + "grad_norm": 1.8693818066390946, + "learning_rate": 9.163609684519444e-06, + "loss": 0.3213, + "step": 1249 + }, + { + "epoch": 0.0275148580233326, + "grad_norm": 2.874523735802841, + "learning_rate": 9.17094644167278e-06, + "loss": 0.5136, + "step": 1250 + }, + { + "epoch": 0.027536869909751266, + "grad_norm": 0.7524607539876177, + "learning_rate": 9.17828319882612e-06, + "loss": 0.6321, + "step": 1251 + }, + { + "epoch": 0.027558881796169932, + "grad_norm": 1.7234186039615664, + "learning_rate": 9.185619955979457e-06, + "loss": 0.3281, + "step": 1252 + }, + { + "epoch": 0.0275808936825886, + "grad_norm": 2.6054589552484733, + "learning_rate": 9.192956713132796e-06, + "loss": 0.4223, + "step": 1253 + }, + { + "epoch": 0.027602905569007265, + "grad_norm": 2.874133269620098, + "learning_rate": 9.200293470286135e-06, + "loss": 0.3681, + "step": 1254 + }, + { + "epoch": 0.02762491745542593, + "grad_norm": 2.5203178211991935, + "learning_rate": 9.207630227439473e-06, + "loss": 0.3904, + "step": 1255 + }, + { + "epoch": 0.027646929341844595, + "grad_norm": 1.6955299902592724, + "learning_rate": 9.21496698459281e-06, + "loss": 0.4105, + "step": 1256 + }, + { + "epoch": 0.02766894122826326, + "grad_norm": 2.6869013452421373, + "learning_rate": 9.222303741746148e-06, + "loss": 0.4976, + "step": 1257 + }, + { + "epoch": 0.027690953114681928, + "grad_norm": 3.6959810507841953, + "learning_rate": 9.229640498899487e-06, + "loss": 0.4816, + "step": 1258 + }, + { + "epoch": 0.027712965001100594, + "grad_norm": 1.8289573101975716, + "learning_rate": 9.236977256052826e-06, + "loss": 0.3779, + "step": 1259 + }, + { + "epoch": 0.02773497688751926, + "grad_norm": 1.9134341735022429, + "learning_rate": 9.244314013206164e-06, + "loss": 0.4473, + "step": 1260 + }, + { + "epoch": 0.027756988773937927, + "grad_norm": 8.430726676875384, + "learning_rate": 9.251650770359501e-06, + "loss": 0.2882, + "step": 1261 + }, + { + "epoch": 0.027779000660356594, + "grad_norm": 1.9708870503799898, + "learning_rate": 9.258987527512841e-06, + "loss": 0.4271, + "step": 1262 + }, + { + "epoch": 0.027801012546775257, + "grad_norm": 4.762738972961295, + "learning_rate": 9.266324284666178e-06, + "loss": 0.3808, + "step": 1263 + }, + { + "epoch": 0.027823024433193923, + "grad_norm": 0.7125366275669475, + "learning_rate": 9.273661041819517e-06, + "loss": 0.6417, + "step": 1264 + }, + { + "epoch": 0.02784503631961259, + "grad_norm": 2.194701459868575, + "learning_rate": 9.280997798972855e-06, + "loss": 0.478, + "step": 1265 + }, + { + "epoch": 0.027867048206031256, + "grad_norm": 4.468585135499233, + "learning_rate": 9.288334556126194e-06, + "loss": 0.5154, + "step": 1266 + }, + { + "epoch": 0.027889060092449923, + "grad_norm": 5.172602288470895, + "learning_rate": 9.295671313279532e-06, + "loss": 0.4752, + "step": 1267 + }, + { + "epoch": 0.02791107197886859, + "grad_norm": 1.7534083077055176, + "learning_rate": 9.303008070432869e-06, + "loss": 0.4095, + "step": 1268 + }, + { + "epoch": 0.027933083865287256, + "grad_norm": 1.6818801748376004, + "learning_rate": 9.310344827586207e-06, + "loss": 0.4744, + "step": 1269 + }, + { + "epoch": 0.027955095751705922, + "grad_norm": 2.1171840625525973, + "learning_rate": 9.317681584739546e-06, + "loss": 0.4965, + "step": 1270 + }, + { + "epoch": 0.02797710763812459, + "grad_norm": 2.2326416876180275, + "learning_rate": 9.325018341892885e-06, + "loss": 0.4074, + "step": 1271 + }, + { + "epoch": 0.027999119524543252, + "grad_norm": 2.135041339595782, + "learning_rate": 9.332355099046221e-06, + "loss": 0.4184, + "step": 1272 + }, + { + "epoch": 0.02802113141096192, + "grad_norm": 3.3017735164237187, + "learning_rate": 9.339691856199562e-06, + "loss": 0.4325, + "step": 1273 + }, + { + "epoch": 0.028043143297380585, + "grad_norm": 1.7563602978470232, + "learning_rate": 9.347028613352898e-06, + "loss": 0.4228, + "step": 1274 + }, + { + "epoch": 0.02806515518379925, + "grad_norm": 0.7812341626527228, + "learning_rate": 9.354365370506237e-06, + "loss": 0.6317, + "step": 1275 + }, + { + "epoch": 0.028087167070217918, + "grad_norm": 19.132107601961476, + "learning_rate": 9.361702127659576e-06, + "loss": 0.3712, + "step": 1276 + }, + { + "epoch": 0.028109178956636584, + "grad_norm": 0.6944922201059951, + "learning_rate": 9.369038884812914e-06, + "loss": 0.632, + "step": 1277 + }, + { + "epoch": 0.02813119084305525, + "grad_norm": 0.6817864651685734, + "learning_rate": 9.376375641966253e-06, + "loss": 0.6326, + "step": 1278 + }, + { + "epoch": 0.028153202729473917, + "grad_norm": 2.2184046916316977, + "learning_rate": 9.38371239911959e-06, + "loss": 0.398, + "step": 1279 + }, + { + "epoch": 0.02817521461589258, + "grad_norm": 1.9820652904787164, + "learning_rate": 9.391049156272928e-06, + "loss": 0.3621, + "step": 1280 + }, + { + "epoch": 0.028197226502311247, + "grad_norm": 2.40696876485982, + "learning_rate": 9.398385913426266e-06, + "loss": 0.3869, + "step": 1281 + }, + { + "epoch": 0.028219238388729913, + "grad_norm": 2.328164827314991, + "learning_rate": 9.405722670579605e-06, + "loss": 0.4577, + "step": 1282 + }, + { + "epoch": 0.02824125027514858, + "grad_norm": 2.1402710701102725, + "learning_rate": 9.413059427732942e-06, + "loss": 0.4296, + "step": 1283 + }, + { + "epoch": 0.028263262161567246, + "grad_norm": 2.005953625446797, + "learning_rate": 9.420396184886282e-06, + "loss": 0.4632, + "step": 1284 + }, + { + "epoch": 0.028285274047985913, + "grad_norm": 1.8252111980215104, + "learning_rate": 9.427732942039619e-06, + "loss": 0.3829, + "step": 1285 + }, + { + "epoch": 0.02830728593440458, + "grad_norm": 2.215224220533431, + "learning_rate": 9.435069699192957e-06, + "loss": 0.4031, + "step": 1286 + }, + { + "epoch": 0.028329297820823246, + "grad_norm": 4.0679533562225325, + "learning_rate": 9.442406456346296e-06, + "loss": 0.4538, + "step": 1287 + }, + { + "epoch": 0.02835130970724191, + "grad_norm": 1.9811985768637785, + "learning_rate": 9.449743213499633e-06, + "loss": 0.3105, + "step": 1288 + }, + { + "epoch": 0.028373321593660576, + "grad_norm": 1.8478145756680804, + "learning_rate": 9.457079970652973e-06, + "loss": 0.4774, + "step": 1289 + }, + { + "epoch": 0.028395333480079242, + "grad_norm": 2.0175275898515346, + "learning_rate": 9.46441672780631e-06, + "loss": 0.3999, + "step": 1290 + }, + { + "epoch": 0.02841734536649791, + "grad_norm": 2.2545260125179927, + "learning_rate": 9.471753484959648e-06, + "loss": 0.3265, + "step": 1291 + }, + { + "epoch": 0.028439357252916575, + "grad_norm": 2.1322864752560684, + "learning_rate": 9.479090242112987e-06, + "loss": 0.4412, + "step": 1292 + }, + { + "epoch": 0.02846136913933524, + "grad_norm": 2.509767065138884, + "learning_rate": 9.486426999266326e-06, + "loss": 0.3937, + "step": 1293 + }, + { + "epoch": 0.028483381025753908, + "grad_norm": 2.0614749071479457, + "learning_rate": 9.493763756419662e-06, + "loss": 0.4001, + "step": 1294 + }, + { + "epoch": 0.028505392912172575, + "grad_norm": 1.0203744696320496, + "learning_rate": 9.501100513573001e-06, + "loss": 0.6428, + "step": 1295 + }, + { + "epoch": 0.028527404798591238, + "grad_norm": 0.8608160123851282, + "learning_rate": 9.50843727072634e-06, + "loss": 0.632, + "step": 1296 + }, + { + "epoch": 0.028549416685009904, + "grad_norm": 2.6296988882195502, + "learning_rate": 9.515774027879678e-06, + "loss": 0.5419, + "step": 1297 + }, + { + "epoch": 0.02857142857142857, + "grad_norm": 2.431317440472111, + "learning_rate": 9.523110785033016e-06, + "loss": 0.4276, + "step": 1298 + }, + { + "epoch": 0.028593440457847237, + "grad_norm": 2.949361329131247, + "learning_rate": 9.530447542186353e-06, + "loss": 0.4099, + "step": 1299 + }, + { + "epoch": 0.028615452344265904, + "grad_norm": 2.294823402586512, + "learning_rate": 9.537784299339694e-06, + "loss": 0.3952, + "step": 1300 + }, + { + "epoch": 0.02863746423068457, + "grad_norm": 1.6228561956627268, + "learning_rate": 9.54512105649303e-06, + "loss": 0.3816, + "step": 1301 + }, + { + "epoch": 0.028659476117103237, + "grad_norm": 2.0960352537991307, + "learning_rate": 9.552457813646369e-06, + "loss": 0.3512, + "step": 1302 + }, + { + "epoch": 0.028681488003521903, + "grad_norm": 5.613624011651852, + "learning_rate": 9.559794570799707e-06, + "loss": 0.3733, + "step": 1303 + }, + { + "epoch": 0.02870349988994057, + "grad_norm": 2.7126451839229233, + "learning_rate": 9.567131327953046e-06, + "loss": 0.4473, + "step": 1304 + }, + { + "epoch": 0.028725511776359233, + "grad_norm": 2.4834870524747275, + "learning_rate": 9.574468085106385e-06, + "loss": 0.3827, + "step": 1305 + }, + { + "epoch": 0.0287475236627779, + "grad_norm": 1.4519719151718662, + "learning_rate": 9.581804842259721e-06, + "loss": 0.4276, + "step": 1306 + }, + { + "epoch": 0.028769535549196566, + "grad_norm": 1.929598640467336, + "learning_rate": 9.58914159941306e-06, + "loss": 0.3376, + "step": 1307 + }, + { + "epoch": 0.028791547435615232, + "grad_norm": 1.7731887766009486, + "learning_rate": 9.596478356566398e-06, + "loss": 0.5437, + "step": 1308 + }, + { + "epoch": 0.0288135593220339, + "grad_norm": 1.6816771560645227, + "learning_rate": 9.603815113719737e-06, + "loss": 0.6429, + "step": 1309 + }, + { + "epoch": 0.028835571208452565, + "grad_norm": 1.699339659686501, + "learning_rate": 9.611151870873074e-06, + "loss": 0.4803, + "step": 1310 + }, + { + "epoch": 0.02885758309487123, + "grad_norm": 1.9198167799630486, + "learning_rate": 9.618488628026414e-06, + "loss": 0.4516, + "step": 1311 + }, + { + "epoch": 0.028879594981289898, + "grad_norm": 4.14693196476176, + "learning_rate": 9.625825385179751e-06, + "loss": 0.3796, + "step": 1312 + }, + { + "epoch": 0.02890160686770856, + "grad_norm": 1.9855950960711746, + "learning_rate": 9.63316214233309e-06, + "loss": 0.3751, + "step": 1313 + }, + { + "epoch": 0.028923618754127228, + "grad_norm": 2.1253931170060847, + "learning_rate": 9.640498899486428e-06, + "loss": 0.4097, + "step": 1314 + }, + { + "epoch": 0.028945630640545894, + "grad_norm": 1.6035398253783062, + "learning_rate": 9.647835656639766e-06, + "loss": 0.355, + "step": 1315 + }, + { + "epoch": 0.02896764252696456, + "grad_norm": 1.6726251171545492, + "learning_rate": 9.655172413793105e-06, + "loss": 0.2967, + "step": 1316 + }, + { + "epoch": 0.028989654413383227, + "grad_norm": 1.7695454645773756, + "learning_rate": 9.662509170946442e-06, + "loss": 0.3769, + "step": 1317 + }, + { + "epoch": 0.029011666299801894, + "grad_norm": 2.2242148130824497, + "learning_rate": 9.66984592809978e-06, + "loss": 0.425, + "step": 1318 + }, + { + "epoch": 0.02903367818622056, + "grad_norm": 1.858535937107876, + "learning_rate": 9.677182685253119e-06, + "loss": 0.4425, + "step": 1319 + }, + { + "epoch": 0.029055690072639227, + "grad_norm": 1.1426354958167202, + "learning_rate": 9.684519442406457e-06, + "loss": 0.6271, + "step": 1320 + }, + { + "epoch": 0.02907770195905789, + "grad_norm": 1.9337888551702354, + "learning_rate": 9.691856199559794e-06, + "loss": 0.403, + "step": 1321 + }, + { + "epoch": 0.029099713845476556, + "grad_norm": 2.2560437905194513, + "learning_rate": 9.699192956713135e-06, + "loss": 0.3417, + "step": 1322 + }, + { + "epoch": 0.029121725731895223, + "grad_norm": 1.6711776047227558, + "learning_rate": 9.706529713866471e-06, + "loss": 0.3221, + "step": 1323 + }, + { + "epoch": 0.02914373761831389, + "grad_norm": 1.7147128150124415, + "learning_rate": 9.71386647101981e-06, + "loss": 0.3929, + "step": 1324 + }, + { + "epoch": 0.029165749504732556, + "grad_norm": 1.9147292187400788, + "learning_rate": 9.721203228173148e-06, + "loss": 0.3863, + "step": 1325 + }, + { + "epoch": 0.029187761391151222, + "grad_norm": 1.7043272422778502, + "learning_rate": 9.728539985326487e-06, + "loss": 0.301, + "step": 1326 + }, + { + "epoch": 0.02920977327756989, + "grad_norm": 2.027717652608504, + "learning_rate": 9.735876742479825e-06, + "loss": 0.4834, + "step": 1327 + }, + { + "epoch": 0.029231785163988555, + "grad_norm": 2.1132641644926187, + "learning_rate": 9.743213499633162e-06, + "loss": 0.3848, + "step": 1328 + }, + { + "epoch": 0.02925379705040722, + "grad_norm": 1.7351576566151818, + "learning_rate": 9.750550256786501e-06, + "loss": 0.3465, + "step": 1329 + }, + { + "epoch": 0.029275808936825885, + "grad_norm": 1.8201194056408554, + "learning_rate": 9.75788701393984e-06, + "loss": 0.3809, + "step": 1330 + }, + { + "epoch": 0.02929782082324455, + "grad_norm": 2.732391913261655, + "learning_rate": 9.765223771093178e-06, + "loss": 0.4202, + "step": 1331 + }, + { + "epoch": 0.029319832709663218, + "grad_norm": 1.6353496577590645, + "learning_rate": 9.772560528246516e-06, + "loss": 0.3679, + "step": 1332 + }, + { + "epoch": 0.029341844596081884, + "grad_norm": 3.537187320156967, + "learning_rate": 9.779897285399855e-06, + "loss": 0.4823, + "step": 1333 + }, + { + "epoch": 0.02936385648250055, + "grad_norm": 1.7054282449583789, + "learning_rate": 9.787234042553192e-06, + "loss": 0.4087, + "step": 1334 + }, + { + "epoch": 0.029385868368919217, + "grad_norm": 3.4194324220376364, + "learning_rate": 9.79457079970653e-06, + "loss": 0.3812, + "step": 1335 + }, + { + "epoch": 0.029407880255337884, + "grad_norm": 1.5985092785676516, + "learning_rate": 9.801907556859869e-06, + "loss": 0.3756, + "step": 1336 + }, + { + "epoch": 0.02942989214175655, + "grad_norm": 2.047733550787133, + "learning_rate": 9.809244314013206e-06, + "loss": 0.3379, + "step": 1337 + }, + { + "epoch": 0.029451904028175213, + "grad_norm": 1.85818244967215, + "learning_rate": 9.816581071166546e-06, + "loss": 0.4619, + "step": 1338 + }, + { + "epoch": 0.02947391591459388, + "grad_norm": 0.7594263961940994, + "learning_rate": 9.823917828319883e-06, + "loss": 0.6499, + "step": 1339 + }, + { + "epoch": 0.029495927801012546, + "grad_norm": 0.6802409814946272, + "learning_rate": 9.831254585473221e-06, + "loss": 0.6082, + "step": 1340 + }, + { + "epoch": 0.029517939687431213, + "grad_norm": 1.8077096837216862, + "learning_rate": 9.83859134262656e-06, + "loss": 0.4503, + "step": 1341 + }, + { + "epoch": 0.02953995157384988, + "grad_norm": 1.8136794413013564, + "learning_rate": 9.845928099779898e-06, + "loss": 0.4723, + "step": 1342 + }, + { + "epoch": 0.029561963460268546, + "grad_norm": 2.4536751800434513, + "learning_rate": 9.853264856933237e-06, + "loss": 0.4141, + "step": 1343 + }, + { + "epoch": 0.029583975346687212, + "grad_norm": 0.6708264127972291, + "learning_rate": 9.860601614086574e-06, + "loss": 0.6111, + "step": 1344 + }, + { + "epoch": 0.02960598723310588, + "grad_norm": 1.9853620326771777, + "learning_rate": 9.867938371239912e-06, + "loss": 0.4469, + "step": 1345 + }, + { + "epoch": 0.02962799911952454, + "grad_norm": 1.7023835739168611, + "learning_rate": 9.87527512839325e-06, + "loss": 0.3059, + "step": 1346 + }, + { + "epoch": 0.029650011005943208, + "grad_norm": 4.128862519433832, + "learning_rate": 9.88261188554659e-06, + "loss": 0.4331, + "step": 1347 + }, + { + "epoch": 0.029672022892361875, + "grad_norm": 1.7296642150510932, + "learning_rate": 9.889948642699926e-06, + "loss": 0.4234, + "step": 1348 + }, + { + "epoch": 0.02969403477878054, + "grad_norm": 4.776980824310914, + "learning_rate": 9.897285399853266e-06, + "loss": 0.3594, + "step": 1349 + }, + { + "epoch": 0.029716046665199208, + "grad_norm": 1.992118889454905, + "learning_rate": 9.904622157006603e-06, + "loss": 0.348, + "step": 1350 + }, + { + "epoch": 0.029738058551617874, + "grad_norm": 1.8696621146138932, + "learning_rate": 9.911958914159942e-06, + "loss": 0.3988, + "step": 1351 + }, + { + "epoch": 0.02976007043803654, + "grad_norm": 1.702416360823522, + "learning_rate": 9.91929567131328e-06, + "loss": 0.4593, + "step": 1352 + }, + { + "epoch": 0.029782082324455207, + "grad_norm": 2.0424093271422636, + "learning_rate": 9.926632428466619e-06, + "loss": 0.3933, + "step": 1353 + }, + { + "epoch": 0.02980409421087387, + "grad_norm": 1.778654095716652, + "learning_rate": 9.933969185619957e-06, + "loss": 0.4343, + "step": 1354 + }, + { + "epoch": 0.029826106097292537, + "grad_norm": 0.7551494794317206, + "learning_rate": 9.941305942773294e-06, + "loss": 0.6091, + "step": 1355 + }, + { + "epoch": 0.029848117983711203, + "grad_norm": 1.6395819384689652, + "learning_rate": 9.948642699926633e-06, + "loss": 0.4241, + "step": 1356 + }, + { + "epoch": 0.02987012987012987, + "grad_norm": 1.8593478631920546, + "learning_rate": 9.955979457079971e-06, + "loss": 0.3584, + "step": 1357 + }, + { + "epoch": 0.029892141756548536, + "grad_norm": 1.5747903904981118, + "learning_rate": 9.96331621423331e-06, + "loss": 0.3843, + "step": 1358 + }, + { + "epoch": 0.029914153642967203, + "grad_norm": 1.9577427087717403, + "learning_rate": 9.970652971386648e-06, + "loss": 0.4974, + "step": 1359 + }, + { + "epoch": 0.02993616552938587, + "grad_norm": 1.5607488350369652, + "learning_rate": 9.977989728539987e-06, + "loss": 0.4788, + "step": 1360 + }, + { + "epoch": 0.029958177415804536, + "grad_norm": 2.6388028023239936, + "learning_rate": 9.985326485693324e-06, + "loss": 0.5692, + "step": 1361 + }, + { + "epoch": 0.029980189302223202, + "grad_norm": 2.064240490809818, + "learning_rate": 9.992663242846662e-06, + "loss": 0.5018, + "step": 1362 + }, + { + "epoch": 0.030002201188641865, + "grad_norm": 1.9903083013590377, + "learning_rate": 1e-05, + "loss": 0.3399, + "step": 1363 + }, + { + "epoch": 0.03002421307506053, + "grad_norm": 11.184980332389785, + "learning_rate": 9.999999987293887e-06, + "loss": 0.3755, + "step": 1364 + }, + { + "epoch": 0.030046224961479198, + "grad_norm": 2.5999577985928566, + "learning_rate": 9.99999994917554e-06, + "loss": 0.3802, + "step": 1365 + }, + { + "epoch": 0.030068236847897865, + "grad_norm": 4.824184117767869, + "learning_rate": 9.999999885644966e-06, + "loss": 0.4041, + "step": 1366 + }, + { + "epoch": 0.03009024873431653, + "grad_norm": 1.8328390485756525, + "learning_rate": 9.999999796702163e-06, + "loss": 0.4412, + "step": 1367 + }, + { + "epoch": 0.030112260620735198, + "grad_norm": 1.865859390850676, + "learning_rate": 9.999999682347128e-06, + "loss": 0.422, + "step": 1368 + }, + { + "epoch": 0.030134272507153864, + "grad_norm": 3.418787628863579, + "learning_rate": 9.999999542579866e-06, + "loss": 0.3474, + "step": 1369 + }, + { + "epoch": 0.03015628439357253, + "grad_norm": 4.44922326253476, + "learning_rate": 9.999999377400376e-06, + "loss": 0.401, + "step": 1370 + }, + { + "epoch": 0.030178296279991194, + "grad_norm": 0.7173608643977205, + "learning_rate": 9.99999918680866e-06, + "loss": 0.6086, + "step": 1371 + }, + { + "epoch": 0.03020030816640986, + "grad_norm": 1.4941796923211095, + "learning_rate": 9.999998970804719e-06, + "loss": 0.3336, + "step": 1372 + }, + { + "epoch": 0.030222320052828527, + "grad_norm": 3.5087478008068924, + "learning_rate": 9.999998729388552e-06, + "loss": 0.3924, + "step": 1373 + }, + { + "epoch": 0.030244331939247193, + "grad_norm": 2.3632760084056006, + "learning_rate": 9.999998462560161e-06, + "loss": 0.3775, + "step": 1374 + }, + { + "epoch": 0.03026634382566586, + "grad_norm": 1.7745973628688707, + "learning_rate": 9.999998170319547e-06, + "loss": 0.4971, + "step": 1375 + }, + { + "epoch": 0.030288355712084526, + "grad_norm": 2.14369787658062, + "learning_rate": 9.999997852666713e-06, + "loss": 0.5038, + "step": 1376 + }, + { + "epoch": 0.030310367598503193, + "grad_norm": 1.891036465311999, + "learning_rate": 9.99999750960166e-06, + "loss": 0.3641, + "step": 1377 + }, + { + "epoch": 0.03033237948492186, + "grad_norm": 1.935436539979598, + "learning_rate": 9.99999714112439e-06, + "loss": 0.3933, + "step": 1378 + }, + { + "epoch": 0.030354391371340522, + "grad_norm": 0.6799716802911351, + "learning_rate": 9.999996747234904e-06, + "loss": 0.6255, + "step": 1379 + }, + { + "epoch": 0.03037640325775919, + "grad_norm": 1.7953460614628496, + "learning_rate": 9.999996327933205e-06, + "loss": 0.3988, + "step": 1380 + }, + { + "epoch": 0.030398415144177855, + "grad_norm": 1.9022387610692173, + "learning_rate": 9.999995883219293e-06, + "loss": 0.4206, + "step": 1381 + }, + { + "epoch": 0.03042042703059652, + "grad_norm": 1.7057267073793752, + "learning_rate": 9.999995413093175e-06, + "loss": 0.364, + "step": 1382 + }, + { + "epoch": 0.030442438917015188, + "grad_norm": 2.018419293873119, + "learning_rate": 9.999994917554846e-06, + "loss": 0.5225, + "step": 1383 + }, + { + "epoch": 0.030464450803433855, + "grad_norm": 1.9607601256462708, + "learning_rate": 9.999994396604315e-06, + "loss": 0.4062, + "step": 1384 + }, + { + "epoch": 0.03048646268985252, + "grad_norm": 2.011032382567715, + "learning_rate": 9.999993850241584e-06, + "loss": 0.4491, + "step": 1385 + }, + { + "epoch": 0.030508474576271188, + "grad_norm": 2.273223783174927, + "learning_rate": 9.999993278466654e-06, + "loss": 0.4035, + "step": 1386 + }, + { + "epoch": 0.030530486462689854, + "grad_norm": 1.921438222895838, + "learning_rate": 9.999992681279525e-06, + "loss": 0.4493, + "step": 1387 + }, + { + "epoch": 0.030552498349108517, + "grad_norm": 1.9003824814576578, + "learning_rate": 9.999992058680205e-06, + "loss": 0.4642, + "step": 1388 + }, + { + "epoch": 0.030574510235527184, + "grad_norm": 2.587762924986724, + "learning_rate": 9.999991410668694e-06, + "loss": 0.3794, + "step": 1389 + }, + { + "epoch": 0.03059652212194585, + "grad_norm": 2.6424150547530734, + "learning_rate": 9.999990737244999e-06, + "loss": 0.4699, + "step": 1390 + }, + { + "epoch": 0.030618534008364517, + "grad_norm": 2.5396656585741737, + "learning_rate": 9.999990038409119e-06, + "loss": 0.3439, + "step": 1391 + }, + { + "epoch": 0.030640545894783183, + "grad_norm": 1.9624318062938482, + "learning_rate": 9.999989314161062e-06, + "loss": 0.4669, + "step": 1392 + }, + { + "epoch": 0.03066255778120185, + "grad_norm": 2.9166744535371616, + "learning_rate": 9.999988564500826e-06, + "loss": 0.4322, + "step": 1393 + }, + { + "epoch": 0.030684569667620516, + "grad_norm": 2.7112355636641596, + "learning_rate": 9.99998778942842e-06, + "loss": 0.4922, + "step": 1394 + }, + { + "epoch": 0.030706581554039183, + "grad_norm": 2.1323045701934666, + "learning_rate": 9.999986988943847e-06, + "loss": 0.461, + "step": 1395 + }, + { + "epoch": 0.030728593440457846, + "grad_norm": 2.4080339565191675, + "learning_rate": 9.99998616304711e-06, + "loss": 0.4559, + "step": 1396 + }, + { + "epoch": 0.030750605326876512, + "grad_norm": 2.0958995635278166, + "learning_rate": 9.99998531173821e-06, + "loss": 0.4344, + "step": 1397 + }, + { + "epoch": 0.03077261721329518, + "grad_norm": 2.00903604579013, + "learning_rate": 9.999984435017155e-06, + "loss": 0.3504, + "step": 1398 + }, + { + "epoch": 0.030794629099713845, + "grad_norm": 2.0929537575866166, + "learning_rate": 9.999983532883953e-06, + "loss": 0.4084, + "step": 1399 + }, + { + "epoch": 0.030816640986132512, + "grad_norm": 2.2728152688724763, + "learning_rate": 9.999982605338603e-06, + "loss": 0.4544, + "step": 1400 + }, + { + "epoch": 0.030838652872551178, + "grad_norm": 2.9264963279819485, + "learning_rate": 9.99998165238111e-06, + "loss": 0.4208, + "step": 1401 + }, + { + "epoch": 0.030860664758969845, + "grad_norm": 2.3774220248732605, + "learning_rate": 9.99998067401148e-06, + "loss": 0.4577, + "step": 1402 + }, + { + "epoch": 0.03088267664538851, + "grad_norm": 1.7831588306757777, + "learning_rate": 9.999979670229719e-06, + "loss": 0.3792, + "step": 1403 + }, + { + "epoch": 0.030904688531807174, + "grad_norm": 1.7883583094716953, + "learning_rate": 9.999978641035831e-06, + "loss": 0.3992, + "step": 1404 + }, + { + "epoch": 0.03092670041822584, + "grad_norm": 2.441434347663062, + "learning_rate": 9.999977586429822e-06, + "loss": 0.3984, + "step": 1405 + }, + { + "epoch": 0.030948712304644507, + "grad_norm": 1.9857857532062335, + "learning_rate": 9.999976506411697e-06, + "loss": 0.4259, + "step": 1406 + }, + { + "epoch": 0.030970724191063174, + "grad_norm": 1.7196778191152458, + "learning_rate": 9.999975400981461e-06, + "loss": 0.3968, + "step": 1407 + }, + { + "epoch": 0.03099273607748184, + "grad_norm": 4.024534523359646, + "learning_rate": 9.999974270139118e-06, + "loss": 0.4426, + "step": 1408 + }, + { + "epoch": 0.031014747963900507, + "grad_norm": 3.1653723639871845, + "learning_rate": 9.99997311388468e-06, + "loss": 0.3463, + "step": 1409 + }, + { + "epoch": 0.031036759850319173, + "grad_norm": 0.7353493562870689, + "learning_rate": 9.999971932218146e-06, + "loss": 0.6139, + "step": 1410 + }, + { + "epoch": 0.03105877173673784, + "grad_norm": 2.1952832733385947, + "learning_rate": 9.999970725139525e-06, + "loss": 0.4388, + "step": 1411 + }, + { + "epoch": 0.031080783623156503, + "grad_norm": 3.3057494992522654, + "learning_rate": 9.999969492648822e-06, + "loss": 0.3456, + "step": 1412 + }, + { + "epoch": 0.03110279550957517, + "grad_norm": 0.7267186496522792, + "learning_rate": 9.999968234746045e-06, + "loss": 0.65, + "step": 1413 + }, + { + "epoch": 0.031124807395993836, + "grad_norm": 0.6524969626527133, + "learning_rate": 9.999966951431197e-06, + "loss": 0.6079, + "step": 1414 + }, + { + "epoch": 0.031146819282412502, + "grad_norm": 2.0537572510741944, + "learning_rate": 9.99996564270429e-06, + "loss": 0.4196, + "step": 1415 + }, + { + "epoch": 0.03116883116883117, + "grad_norm": 1.7950139302641386, + "learning_rate": 9.999964308565326e-06, + "loss": 0.4613, + "step": 1416 + }, + { + "epoch": 0.031190843055249835, + "grad_norm": 2.4106550930176516, + "learning_rate": 9.999962949014313e-06, + "loss": 0.3671, + "step": 1417 + }, + { + "epoch": 0.031212854941668502, + "grad_norm": 2.035604237854595, + "learning_rate": 9.99996156405126e-06, + "loss": 0.4744, + "step": 1418 + }, + { + "epoch": 0.03123486682808717, + "grad_norm": 0.8021990932310172, + "learning_rate": 9.999960153676171e-06, + "loss": 0.6454, + "step": 1419 + }, + { + "epoch": 0.03125687871450583, + "grad_norm": 2.801860555498594, + "learning_rate": 9.999958717889055e-06, + "loss": 0.4482, + "step": 1420 + }, + { + "epoch": 0.0312788906009245, + "grad_norm": 0.7399500825269918, + "learning_rate": 9.999957256689918e-06, + "loss": 0.6362, + "step": 1421 + }, + { + "epoch": 0.031300902487343164, + "grad_norm": 2.218610769954249, + "learning_rate": 9.999955770078768e-06, + "loss": 0.454, + "step": 1422 + }, + { + "epoch": 0.031322914373761834, + "grad_norm": 0.6607990964368382, + "learning_rate": 9.999954258055613e-06, + "loss": 0.6564, + "step": 1423 + }, + { + "epoch": 0.0313449262601805, + "grad_norm": 1.9631168974710642, + "learning_rate": 9.999952720620462e-06, + "loss": 0.3688, + "step": 1424 + }, + { + "epoch": 0.03136693814659916, + "grad_norm": 2.408686328037689, + "learning_rate": 9.99995115777332e-06, + "loss": 0.3887, + "step": 1425 + }, + { + "epoch": 0.03138895003301783, + "grad_norm": 0.6633296976883366, + "learning_rate": 9.999949569514196e-06, + "loss": 0.5806, + "step": 1426 + }, + { + "epoch": 0.03141096191943649, + "grad_norm": 0.692034035366471, + "learning_rate": 9.999947955843099e-06, + "loss": 0.6437, + "step": 1427 + }, + { + "epoch": 0.03143297380585516, + "grad_norm": 0.6729096270684335, + "learning_rate": 9.999946316760034e-06, + "loss": 0.6236, + "step": 1428 + }, + { + "epoch": 0.031454985692273826, + "grad_norm": 1.9541575338517432, + "learning_rate": 9.999944652265015e-06, + "loss": 0.4626, + "step": 1429 + }, + { + "epoch": 0.031476997578692496, + "grad_norm": 1.8135274139415798, + "learning_rate": 9.999942962358045e-06, + "loss": 0.2651, + "step": 1430 + }, + { + "epoch": 0.03149900946511116, + "grad_norm": 2.1889963590057326, + "learning_rate": 9.999941247039135e-06, + "loss": 0.3363, + "step": 1431 + }, + { + "epoch": 0.03152102135152983, + "grad_norm": 0.7443424289544788, + "learning_rate": 9.999939506308297e-06, + "loss": 0.6631, + "step": 1432 + }, + { + "epoch": 0.03154303323794849, + "grad_norm": 1.831476404570037, + "learning_rate": 9.999937740165531e-06, + "loss": 0.3941, + "step": 1433 + }, + { + "epoch": 0.031565045124367155, + "grad_norm": 2.0927874770539536, + "learning_rate": 9.999935948610856e-06, + "loss": 0.4285, + "step": 1434 + }, + { + "epoch": 0.031587057010785825, + "grad_norm": 3.7114442897514377, + "learning_rate": 9.999934131644274e-06, + "loss": 0.5446, + "step": 1435 + }, + { + "epoch": 0.03160906889720449, + "grad_norm": 1.8003302954776645, + "learning_rate": 9.999932289265797e-06, + "loss": 0.4463, + "step": 1436 + }, + { + "epoch": 0.03163108078362316, + "grad_norm": 1.9816625017379164, + "learning_rate": 9.999930421475435e-06, + "loss": 0.4418, + "step": 1437 + }, + { + "epoch": 0.03165309267004182, + "grad_norm": 1.7740996089052918, + "learning_rate": 9.999928528273194e-06, + "loss": 0.3361, + "step": 1438 + }, + { + "epoch": 0.03167510455646049, + "grad_norm": 1.7583658084844969, + "learning_rate": 9.999926609659089e-06, + "loss": 0.4293, + "step": 1439 + }, + { + "epoch": 0.031697116442879154, + "grad_norm": 2.434746115726488, + "learning_rate": 9.999924665633125e-06, + "loss": 0.3943, + "step": 1440 + }, + { + "epoch": 0.03171912832929782, + "grad_norm": 2.2417983925429583, + "learning_rate": 9.999922696195315e-06, + "loss": 0.3723, + "step": 1441 + }, + { + "epoch": 0.03174114021571649, + "grad_norm": 1.5340766795695342, + "learning_rate": 9.999920701345668e-06, + "loss": 0.4396, + "step": 1442 + }, + { + "epoch": 0.03176315210213515, + "grad_norm": 1.6099887259271106, + "learning_rate": 9.999918681084193e-06, + "loss": 0.4141, + "step": 1443 + }, + { + "epoch": 0.03178516398855382, + "grad_norm": 5.262837498666287, + "learning_rate": 9.999916635410903e-06, + "loss": 0.479, + "step": 1444 + }, + { + "epoch": 0.03180717587497248, + "grad_norm": 4.181734260814097, + "learning_rate": 9.999914564325805e-06, + "loss": 0.3833, + "step": 1445 + }, + { + "epoch": 0.03182918776139115, + "grad_norm": 6.722429591505467, + "learning_rate": 9.999912467828911e-06, + "loss": 0.4572, + "step": 1446 + }, + { + "epoch": 0.031851199647809816, + "grad_norm": 1.8786227390613839, + "learning_rate": 9.999910345920232e-06, + "loss": 0.4079, + "step": 1447 + }, + { + "epoch": 0.031873211534228486, + "grad_norm": 3.6764812062924803, + "learning_rate": 9.999908198599779e-06, + "loss": 0.4099, + "step": 1448 + }, + { + "epoch": 0.03189522342064715, + "grad_norm": 2.303697766907342, + "learning_rate": 9.999906025867563e-06, + "loss": 0.365, + "step": 1449 + }, + { + "epoch": 0.03191723530706581, + "grad_norm": 1.9283318598137393, + "learning_rate": 9.999903827723593e-06, + "loss": 0.4236, + "step": 1450 + }, + { + "epoch": 0.03193924719348448, + "grad_norm": 4.158002328230038, + "learning_rate": 9.999901604167885e-06, + "loss": 0.4568, + "step": 1451 + }, + { + "epoch": 0.031961259079903145, + "grad_norm": 1.5124391999411788, + "learning_rate": 9.999899355200444e-06, + "loss": 0.4576, + "step": 1452 + }, + { + "epoch": 0.031983270966321815, + "grad_norm": 2.0923780111040653, + "learning_rate": 9.999897080821286e-06, + "loss": 0.4327, + "step": 1453 + }, + { + "epoch": 0.03200528285274048, + "grad_norm": 1.7883202252034054, + "learning_rate": 9.99989478103042e-06, + "loss": 0.4193, + "step": 1454 + }, + { + "epoch": 0.03202729473915915, + "grad_norm": 1.85680194872166, + "learning_rate": 9.999892455827859e-06, + "loss": 0.3902, + "step": 1455 + }, + { + "epoch": 0.03204930662557781, + "grad_norm": 1.7272499707868718, + "learning_rate": 9.999890105213615e-06, + "loss": 0.3397, + "step": 1456 + }, + { + "epoch": 0.03207131851199648, + "grad_norm": 0.806234670671301, + "learning_rate": 9.999887729187698e-06, + "loss": 0.5956, + "step": 1457 + }, + { + "epoch": 0.032093330398415144, + "grad_norm": 0.7237684010929504, + "learning_rate": 9.999885327750123e-06, + "loss": 0.6215, + "step": 1458 + }, + { + "epoch": 0.03211534228483381, + "grad_norm": 1.6012324876872452, + "learning_rate": 9.9998829009009e-06, + "loss": 0.348, + "step": 1459 + }, + { + "epoch": 0.03213735417125248, + "grad_norm": 24.122996374890384, + "learning_rate": 9.999880448640043e-06, + "loss": 0.4418, + "step": 1460 + }, + { + "epoch": 0.03215936605767114, + "grad_norm": 1.43947318946806, + "learning_rate": 9.999877970967562e-06, + "loss": 0.4215, + "step": 1461 + }, + { + "epoch": 0.03218137794408981, + "grad_norm": 0.7402585439694405, + "learning_rate": 9.999875467883474e-06, + "loss": 0.6072, + "step": 1462 + }, + { + "epoch": 0.03220338983050847, + "grad_norm": 1.7748647145979315, + "learning_rate": 9.999872939387786e-06, + "loss": 0.3602, + "step": 1463 + }, + { + "epoch": 0.03222540171692714, + "grad_norm": 4.076111169083641, + "learning_rate": 9.999870385480515e-06, + "loss": 0.3745, + "step": 1464 + }, + { + "epoch": 0.032247413603345806, + "grad_norm": 0.7555277429476029, + "learning_rate": 9.999867806161672e-06, + "loss": 0.6366, + "step": 1465 + }, + { + "epoch": 0.03226942548976447, + "grad_norm": 2.08219035257268, + "learning_rate": 9.999865201431272e-06, + "loss": 0.4358, + "step": 1466 + }, + { + "epoch": 0.03229143737618314, + "grad_norm": 0.7496771796603289, + "learning_rate": 9.999862571289325e-06, + "loss": 0.6337, + "step": 1467 + }, + { + "epoch": 0.0323134492626018, + "grad_norm": 1.9676768336096893, + "learning_rate": 9.999859915735847e-06, + "loss": 0.4107, + "step": 1468 + }, + { + "epoch": 0.03233546114902047, + "grad_norm": 2.1180611649704097, + "learning_rate": 9.999857234770853e-06, + "loss": 0.444, + "step": 1469 + }, + { + "epoch": 0.032357473035439135, + "grad_norm": 2.035041707016048, + "learning_rate": 9.999854528394353e-06, + "loss": 0.3764, + "step": 1470 + }, + { + "epoch": 0.032379484921857805, + "grad_norm": 2.7486947634486385, + "learning_rate": 9.999851796606362e-06, + "loss": 0.3985, + "step": 1471 + }, + { + "epoch": 0.03240149680827647, + "grad_norm": 1.4814910173796478, + "learning_rate": 9.999849039406894e-06, + "loss": 0.3726, + "step": 1472 + }, + { + "epoch": 0.03242350869469514, + "grad_norm": 11.110395572370583, + "learning_rate": 9.999846256795964e-06, + "loss": 0.4884, + "step": 1473 + }, + { + "epoch": 0.0324455205811138, + "grad_norm": 1.613610970478598, + "learning_rate": 9.999843448773585e-06, + "loss": 0.352, + "step": 1474 + }, + { + "epoch": 0.032467532467532464, + "grad_norm": 2.1838089484998333, + "learning_rate": 9.999840615339771e-06, + "loss": 0.4318, + "step": 1475 + }, + { + "epoch": 0.032489544353951134, + "grad_norm": 1.813345918007503, + "learning_rate": 9.999837756494539e-06, + "loss": 0.4089, + "step": 1476 + }, + { + "epoch": 0.0325115562403698, + "grad_norm": 2.8540450973434175, + "learning_rate": 9.9998348722379e-06, + "loss": 0.311, + "step": 1477 + }, + { + "epoch": 0.03253356812678847, + "grad_norm": 1.912844796866063, + "learning_rate": 9.999831962569871e-06, + "loss": 0.3881, + "step": 1478 + }, + { + "epoch": 0.03255558001320713, + "grad_norm": 1.9800042826960236, + "learning_rate": 9.999829027490464e-06, + "loss": 0.3942, + "step": 1479 + }, + { + "epoch": 0.0325775918996258, + "grad_norm": 1.697186972354486, + "learning_rate": 9.999826066999698e-06, + "loss": 0.3664, + "step": 1480 + }, + { + "epoch": 0.03259960378604446, + "grad_norm": 2.0853510028410622, + "learning_rate": 9.999823081097586e-06, + "loss": 0.4631, + "step": 1481 + }, + { + "epoch": 0.03262161567246313, + "grad_norm": 1.611384578602328, + "learning_rate": 9.999820069784143e-06, + "loss": 0.4569, + "step": 1482 + }, + { + "epoch": 0.032643627558881796, + "grad_norm": 2.681142815401733, + "learning_rate": 9.999817033059383e-06, + "loss": 0.3542, + "step": 1483 + }, + { + "epoch": 0.03266563944530046, + "grad_norm": 1.7205783594470798, + "learning_rate": 9.999813970923323e-06, + "loss": 0.3271, + "step": 1484 + }, + { + "epoch": 0.03268765133171913, + "grad_norm": 1.062637090935274, + "learning_rate": 9.99981088337598e-06, + "loss": 0.6264, + "step": 1485 + }, + { + "epoch": 0.03270966321813779, + "grad_norm": 1.9608736034593401, + "learning_rate": 9.999807770417367e-06, + "loss": 0.3441, + "step": 1486 + }, + { + "epoch": 0.03273167510455646, + "grad_norm": 2.036226403682019, + "learning_rate": 9.999804632047501e-06, + "loss": 0.3295, + "step": 1487 + }, + { + "epoch": 0.032753686990975125, + "grad_norm": 2.1320782123402795, + "learning_rate": 9.999801468266396e-06, + "loss": 0.4048, + "step": 1488 + }, + { + "epoch": 0.032775698877393795, + "grad_norm": 1.4916331759017087, + "learning_rate": 9.999798279074073e-06, + "loss": 0.3523, + "step": 1489 + }, + { + "epoch": 0.03279771076381246, + "grad_norm": 0.7362896181625683, + "learning_rate": 9.999795064470544e-06, + "loss": 0.6332, + "step": 1490 + }, + { + "epoch": 0.03281972265023112, + "grad_norm": 2.2449859502876044, + "learning_rate": 9.999791824455827e-06, + "loss": 0.4151, + "step": 1491 + }, + { + "epoch": 0.03284173453664979, + "grad_norm": 1.7247003617109042, + "learning_rate": 9.999788559029937e-06, + "loss": 0.2721, + "step": 1492 + }, + { + "epoch": 0.032863746423068455, + "grad_norm": 2.128631055329259, + "learning_rate": 9.999785268192892e-06, + "loss": 0.4775, + "step": 1493 + }, + { + "epoch": 0.032885758309487124, + "grad_norm": 2.210857270481957, + "learning_rate": 9.99978195194471e-06, + "loss": 0.4784, + "step": 1494 + }, + { + "epoch": 0.03290777019590579, + "grad_norm": 2.0149391412815794, + "learning_rate": 9.999778610285404e-06, + "loss": 0.4631, + "step": 1495 + }, + { + "epoch": 0.03292978208232446, + "grad_norm": 6.1945599537211855, + "learning_rate": 9.999775243214993e-06, + "loss": 0.3872, + "step": 1496 + }, + { + "epoch": 0.03295179396874312, + "grad_norm": 2.881056778913017, + "learning_rate": 9.999771850733495e-06, + "loss": 0.3841, + "step": 1497 + }, + { + "epoch": 0.03297380585516179, + "grad_norm": 2.0182387016659833, + "learning_rate": 9.999768432840927e-06, + "loss": 0.382, + "step": 1498 + }, + { + "epoch": 0.032995817741580453, + "grad_norm": 1.690015122651707, + "learning_rate": 9.999764989537306e-06, + "loss": 0.3699, + "step": 1499 + }, + { + "epoch": 0.033017829627999117, + "grad_norm": 0.7867924315356846, + "learning_rate": 9.999761520822648e-06, + "loss": 0.634, + "step": 1500 + }, + { + "epoch": 0.033039841514417786, + "grad_norm": 3.9590073483560126, + "learning_rate": 9.999758026696973e-06, + "loss": 0.4003, + "step": 1501 + }, + { + "epoch": 0.03306185340083645, + "grad_norm": 1.9818681573182206, + "learning_rate": 9.999754507160297e-06, + "loss": 0.4572, + "step": 1502 + }, + { + "epoch": 0.03308386528725512, + "grad_norm": 0.6713477210422208, + "learning_rate": 9.99975096221264e-06, + "loss": 0.5796, + "step": 1503 + }, + { + "epoch": 0.03310587717367378, + "grad_norm": 2.4305400427048527, + "learning_rate": 9.999747391854018e-06, + "loss": 0.4514, + "step": 1504 + }, + { + "epoch": 0.03312788906009245, + "grad_norm": 1.7996573713585187, + "learning_rate": 9.999743796084449e-06, + "loss": 0.4021, + "step": 1505 + }, + { + "epoch": 0.033149900946511116, + "grad_norm": 1.7789062069594723, + "learning_rate": 9.999740174903953e-06, + "loss": 0.4221, + "step": 1506 + }, + { + "epoch": 0.033171912832929785, + "grad_norm": 1.674257351521181, + "learning_rate": 9.999736528312547e-06, + "loss": 0.4402, + "step": 1507 + }, + { + "epoch": 0.03319392471934845, + "grad_norm": 3.096956123332705, + "learning_rate": 9.99973285631025e-06, + "loss": 0.3967, + "step": 1508 + }, + { + "epoch": 0.03321593660576711, + "grad_norm": 0.841815532765584, + "learning_rate": 9.99972915889708e-06, + "loss": 0.6107, + "step": 1509 + }, + { + "epoch": 0.03323794849218578, + "grad_norm": 3.0665073898015027, + "learning_rate": 9.999725436073058e-06, + "loss": 0.4468, + "step": 1510 + }, + { + "epoch": 0.033259960378604445, + "grad_norm": 2.3996887933968827, + "learning_rate": 9.999721687838202e-06, + "loss": 0.3853, + "step": 1511 + }, + { + "epoch": 0.033281972265023115, + "grad_norm": 1.969145891101938, + "learning_rate": 9.999717914192528e-06, + "loss": 0.3942, + "step": 1512 + }, + { + "epoch": 0.03330398415144178, + "grad_norm": 2.1962562520061586, + "learning_rate": 9.999714115136059e-06, + "loss": 0.4421, + "step": 1513 + }, + { + "epoch": 0.03332599603786045, + "grad_norm": 1.7383158343080003, + "learning_rate": 9.999710290668813e-06, + "loss": 0.3791, + "step": 1514 + }, + { + "epoch": 0.03334800792427911, + "grad_norm": 1.8434891153929758, + "learning_rate": 9.99970644079081e-06, + "loss": 0.3867, + "step": 1515 + }, + { + "epoch": 0.033370019810697774, + "grad_norm": 0.6916871368028679, + "learning_rate": 9.999702565502068e-06, + "loss": 0.5954, + "step": 1516 + }, + { + "epoch": 0.033392031697116444, + "grad_norm": 1.9287057979358133, + "learning_rate": 9.999698664802606e-06, + "loss": 0.3463, + "step": 1517 + }, + { + "epoch": 0.03341404358353511, + "grad_norm": 1.914538449064333, + "learning_rate": 9.999694738692446e-06, + "loss": 0.438, + "step": 1518 + }, + { + "epoch": 0.03343605546995378, + "grad_norm": 1.6912061385667385, + "learning_rate": 9.99969078717161e-06, + "loss": 0.4317, + "step": 1519 + }, + { + "epoch": 0.03345806735637244, + "grad_norm": 0.7359207025264212, + "learning_rate": 9.999686810240115e-06, + "loss": 0.6117, + "step": 1520 + }, + { + "epoch": 0.03348007924279111, + "grad_norm": 2.0057839373859925, + "learning_rate": 9.99968280789798e-06, + "loss": 0.5245, + "step": 1521 + }, + { + "epoch": 0.03350209112920977, + "grad_norm": 2.103726656391364, + "learning_rate": 9.999678780145227e-06, + "loss": 0.4437, + "step": 1522 + }, + { + "epoch": 0.03352410301562844, + "grad_norm": 3.0821881427094158, + "learning_rate": 9.999674726981879e-06, + "loss": 0.4213, + "step": 1523 + }, + { + "epoch": 0.033546114902047106, + "grad_norm": 2.7811537359732936, + "learning_rate": 9.999670648407951e-06, + "loss": 0.4628, + "step": 1524 + }, + { + "epoch": 0.03356812678846577, + "grad_norm": 1.7688233338907653, + "learning_rate": 9.999666544423468e-06, + "loss": 0.2011, + "step": 1525 + }, + { + "epoch": 0.03359013867488444, + "grad_norm": 2.0442435901114586, + "learning_rate": 9.999662415028451e-06, + "loss": 0.3677, + "step": 1526 + }, + { + "epoch": 0.0336121505613031, + "grad_norm": 2.374043266929755, + "learning_rate": 9.999658260222918e-06, + "loss": 0.3776, + "step": 1527 + }, + { + "epoch": 0.03363416244772177, + "grad_norm": 1.6292689750718192, + "learning_rate": 9.999654080006893e-06, + "loss": 0.3701, + "step": 1528 + }, + { + "epoch": 0.033656174334140435, + "grad_norm": 1.939869278855266, + "learning_rate": 9.999649874380394e-06, + "loss": 0.4094, + "step": 1529 + }, + { + "epoch": 0.033678186220559105, + "grad_norm": 1.8602462769805168, + "learning_rate": 9.999645643343448e-06, + "loss": 0.4248, + "step": 1530 + }, + { + "epoch": 0.03370019810697777, + "grad_norm": 11.254500556816746, + "learning_rate": 9.99964138689607e-06, + "loss": 0.3395, + "step": 1531 + }, + { + "epoch": 0.03372220999339643, + "grad_norm": 10.832697295921188, + "learning_rate": 9.999637105038287e-06, + "loss": 0.3258, + "step": 1532 + }, + { + "epoch": 0.0337442218798151, + "grad_norm": 1.9775353826622748, + "learning_rate": 9.999632797770115e-06, + "loss": 0.4038, + "step": 1533 + }, + { + "epoch": 0.033766233766233764, + "grad_norm": 2.024906842699565, + "learning_rate": 9.999628465091581e-06, + "loss": 0.3594, + "step": 1534 + }, + { + "epoch": 0.033788245652652434, + "grad_norm": 4.336148446905888, + "learning_rate": 9.999624107002705e-06, + "loss": 0.4428, + "step": 1535 + }, + { + "epoch": 0.0338102575390711, + "grad_norm": 2.2601855480918625, + "learning_rate": 9.99961972350351e-06, + "loss": 0.4332, + "step": 1536 + }, + { + "epoch": 0.03383226942548977, + "grad_norm": 1.999001347289373, + "learning_rate": 9.999615314594018e-06, + "loss": 0.3605, + "step": 1537 + }, + { + "epoch": 0.03385428131190843, + "grad_norm": 2.5410809423180014, + "learning_rate": 9.999610880274251e-06, + "loss": 0.4232, + "step": 1538 + }, + { + "epoch": 0.0338762931983271, + "grad_norm": 2.9236190169197602, + "learning_rate": 9.99960642054423e-06, + "loss": 0.3878, + "step": 1539 + }, + { + "epoch": 0.03389830508474576, + "grad_norm": 0.7365851879534067, + "learning_rate": 9.99960193540398e-06, + "loss": 0.6216, + "step": 1540 + }, + { + "epoch": 0.033920316971164426, + "grad_norm": 1.805484845911334, + "learning_rate": 9.999597424853524e-06, + "loss": 0.4242, + "step": 1541 + }, + { + "epoch": 0.033942328857583096, + "grad_norm": 1.6705948588233057, + "learning_rate": 9.999592888892883e-06, + "loss": 0.394, + "step": 1542 + }, + { + "epoch": 0.03396434074400176, + "grad_norm": 3.028160499542413, + "learning_rate": 9.999588327522083e-06, + "loss": 0.3526, + "step": 1543 + }, + { + "epoch": 0.03398635263042043, + "grad_norm": 1.885184888569619, + "learning_rate": 9.999583740741143e-06, + "loss": 0.4687, + "step": 1544 + }, + { + "epoch": 0.03400836451683909, + "grad_norm": 2.133484396360501, + "learning_rate": 9.99957912855009e-06, + "loss": 0.386, + "step": 1545 + }, + { + "epoch": 0.03403037640325776, + "grad_norm": 1.5617642442403485, + "learning_rate": 9.999574490948945e-06, + "loss": 0.4061, + "step": 1546 + }, + { + "epoch": 0.034052388289676425, + "grad_norm": 1.438828061570263, + "learning_rate": 9.999569827937731e-06, + "loss": 0.3304, + "step": 1547 + }, + { + "epoch": 0.034074400176095095, + "grad_norm": 0.7132062087133528, + "learning_rate": 9.999565139516476e-06, + "loss": 0.6391, + "step": 1548 + }, + { + "epoch": 0.03409641206251376, + "grad_norm": 1.8231833068710932, + "learning_rate": 9.9995604256852e-06, + "loss": 0.435, + "step": 1549 + }, + { + "epoch": 0.03411842394893242, + "grad_norm": 1.813447825208378, + "learning_rate": 9.999555686443929e-06, + "loss": 0.432, + "step": 1550 + }, + { + "epoch": 0.03414043583535109, + "grad_norm": 4.722587145642004, + "learning_rate": 9.999550921792685e-06, + "loss": 0.3651, + "step": 1551 + }, + { + "epoch": 0.034162447721769754, + "grad_norm": 4.005377615337494, + "learning_rate": 9.999546131731495e-06, + "loss": 0.4603, + "step": 1552 + }, + { + "epoch": 0.034184459608188424, + "grad_norm": 2.04919615254606, + "learning_rate": 9.99954131626038e-06, + "loss": 0.4695, + "step": 1553 + }, + { + "epoch": 0.03420647149460709, + "grad_norm": 2.485947765785158, + "learning_rate": 9.999536475379366e-06, + "loss": 0.3968, + "step": 1554 + }, + { + "epoch": 0.03422848338102576, + "grad_norm": 2.512544637129882, + "learning_rate": 9.999531609088478e-06, + "loss": 0.4311, + "step": 1555 + }, + { + "epoch": 0.03425049526744442, + "grad_norm": 1.9966219768993496, + "learning_rate": 9.999526717387742e-06, + "loss": 0.3507, + "step": 1556 + }, + { + "epoch": 0.03427250715386308, + "grad_norm": 0.683983779121949, + "learning_rate": 9.99952180027718e-06, + "loss": 0.6271, + "step": 1557 + }, + { + "epoch": 0.03429451904028175, + "grad_norm": 2.2337491827222857, + "learning_rate": 9.999516857756818e-06, + "loss": 0.3629, + "step": 1558 + }, + { + "epoch": 0.034316530926700416, + "grad_norm": 3.4128115182692538, + "learning_rate": 9.999511889826683e-06, + "loss": 0.4422, + "step": 1559 + }, + { + "epoch": 0.034338542813119086, + "grad_norm": 0.6816630138172773, + "learning_rate": 9.999506896486798e-06, + "loss": 0.6305, + "step": 1560 + }, + { + "epoch": 0.03436055469953775, + "grad_norm": 3.3322220827443783, + "learning_rate": 9.999501877737188e-06, + "loss": 0.3297, + "step": 1561 + }, + { + "epoch": 0.03438256658595642, + "grad_norm": 3.177387540620388, + "learning_rate": 9.99949683357788e-06, + "loss": 0.4043, + "step": 1562 + }, + { + "epoch": 0.03440457847237508, + "grad_norm": 1.56000599985702, + "learning_rate": 9.9994917640089e-06, + "loss": 0.3921, + "step": 1563 + }, + { + "epoch": 0.03442659035879375, + "grad_norm": 3.0196068713612907, + "learning_rate": 9.999486669030274e-06, + "loss": 0.4327, + "step": 1564 + }, + { + "epoch": 0.034448602245212415, + "grad_norm": 2.7305452418357943, + "learning_rate": 9.999481548642025e-06, + "loss": 0.3921, + "step": 1565 + }, + { + "epoch": 0.03447061413163108, + "grad_norm": 1.6273989406442366, + "learning_rate": 9.999476402844181e-06, + "loss": 0.3723, + "step": 1566 + }, + { + "epoch": 0.03449262601804975, + "grad_norm": 2.506224682323829, + "learning_rate": 9.99947123163677e-06, + "loss": 0.352, + "step": 1567 + }, + { + "epoch": 0.03451463790446841, + "grad_norm": 2.097721538359455, + "learning_rate": 9.999466035019815e-06, + "loss": 0.3926, + "step": 1568 + }, + { + "epoch": 0.03453664979088708, + "grad_norm": 2.917773382315619, + "learning_rate": 9.999460812993343e-06, + "loss": 0.3849, + "step": 1569 + }, + { + "epoch": 0.034558661677305744, + "grad_norm": 3.0890795634733608, + "learning_rate": 9.999455565557383e-06, + "loss": 0.4229, + "step": 1570 + }, + { + "epoch": 0.034580673563724414, + "grad_norm": 3.1863337632087747, + "learning_rate": 9.999450292711959e-06, + "loss": 0.4659, + "step": 1571 + }, + { + "epoch": 0.03460268545014308, + "grad_norm": 2.1519054340524955, + "learning_rate": 9.999444994457099e-06, + "loss": 0.4647, + "step": 1572 + }, + { + "epoch": 0.03462469733656175, + "grad_norm": 2.433745845817816, + "learning_rate": 9.99943967079283e-06, + "loss": 0.3794, + "step": 1573 + }, + { + "epoch": 0.03464670922298041, + "grad_norm": 3.7211397024343458, + "learning_rate": 9.999434321719179e-06, + "loss": 0.4286, + "step": 1574 + }, + { + "epoch": 0.03466872110939907, + "grad_norm": 3.5846535645599387, + "learning_rate": 9.999428947236172e-06, + "loss": 0.3915, + "step": 1575 + }, + { + "epoch": 0.03469073299581774, + "grad_norm": 2.5675463628489505, + "learning_rate": 9.999423547343838e-06, + "loss": 0.4569, + "step": 1576 + }, + { + "epoch": 0.034712744882236406, + "grad_norm": 1.816507336596639, + "learning_rate": 9.999418122042203e-06, + "loss": 0.4341, + "step": 1577 + }, + { + "epoch": 0.034734756768655076, + "grad_norm": 1.8391783199617329, + "learning_rate": 9.999412671331296e-06, + "loss": 0.4807, + "step": 1578 + }, + { + "epoch": 0.03475676865507374, + "grad_norm": 0.7367601372239643, + "learning_rate": 9.999407195211144e-06, + "loss": 0.6062, + "step": 1579 + }, + { + "epoch": 0.03477878054149241, + "grad_norm": 2.3894432865724657, + "learning_rate": 9.999401693681775e-06, + "loss": 0.3911, + "step": 1580 + }, + { + "epoch": 0.03480079242791107, + "grad_norm": 1.7882626936219166, + "learning_rate": 9.999396166743217e-06, + "loss": 0.4204, + "step": 1581 + }, + { + "epoch": 0.034822804314329735, + "grad_norm": 0.66026168284321, + "learning_rate": 9.999390614395496e-06, + "loss": 0.6109, + "step": 1582 + }, + { + "epoch": 0.034844816200748405, + "grad_norm": 2.3774351630907367, + "learning_rate": 9.999385036638644e-06, + "loss": 0.4895, + "step": 1583 + }, + { + "epoch": 0.03486682808716707, + "grad_norm": 1.7630346700225472, + "learning_rate": 9.999379433472687e-06, + "loss": 0.3263, + "step": 1584 + }, + { + "epoch": 0.03488883997358574, + "grad_norm": 0.6880115753953585, + "learning_rate": 9.999373804897655e-06, + "loss": 0.6019, + "step": 1585 + }, + { + "epoch": 0.0349108518600044, + "grad_norm": 0.6721225253190976, + "learning_rate": 9.999368150913574e-06, + "loss": 0.624, + "step": 1586 + }, + { + "epoch": 0.03493286374642307, + "grad_norm": 2.3347695742827144, + "learning_rate": 9.999362471520474e-06, + "loss": 0.4023, + "step": 1587 + }, + { + "epoch": 0.034954875632841734, + "grad_norm": 2.735155004847663, + "learning_rate": 9.999356766718385e-06, + "loss": 0.4543, + "step": 1588 + }, + { + "epoch": 0.034976887519260404, + "grad_norm": 4.202343225227171, + "learning_rate": 9.999351036507334e-06, + "loss": 0.4139, + "step": 1589 + }, + { + "epoch": 0.03499889940567907, + "grad_norm": 0.694271674372504, + "learning_rate": 9.999345280887354e-06, + "loss": 0.6193, + "step": 1590 + }, + { + "epoch": 0.03502091129209773, + "grad_norm": 3.047597357583647, + "learning_rate": 9.99933949985847e-06, + "loss": 0.4005, + "step": 1591 + }, + { + "epoch": 0.0350429231785164, + "grad_norm": 2.2982063079807777, + "learning_rate": 9.999333693420711e-06, + "loss": 0.5128, + "step": 1592 + }, + { + "epoch": 0.03506493506493506, + "grad_norm": 1.9210597268385665, + "learning_rate": 9.99932786157411e-06, + "loss": 0.4523, + "step": 1593 + }, + { + "epoch": 0.03508694695135373, + "grad_norm": 2.039330793904767, + "learning_rate": 9.999322004318695e-06, + "loss": 0.3723, + "step": 1594 + }, + { + "epoch": 0.035108958837772396, + "grad_norm": 3.297782605873397, + "learning_rate": 9.999316121654497e-06, + "loss": 0.3859, + "step": 1595 + }, + { + "epoch": 0.035130970724191066, + "grad_norm": 1.7183734056747721, + "learning_rate": 9.999310213581543e-06, + "loss": 0.3763, + "step": 1596 + }, + { + "epoch": 0.03515298261060973, + "grad_norm": 1.792825864053922, + "learning_rate": 9.999304280099866e-06, + "loss": 0.4273, + "step": 1597 + }, + { + "epoch": 0.0351749944970284, + "grad_norm": 2.084819381647813, + "learning_rate": 9.999298321209494e-06, + "loss": 0.5258, + "step": 1598 + }, + { + "epoch": 0.03519700638344706, + "grad_norm": 2.593660041602836, + "learning_rate": 9.999292336910457e-06, + "loss": 0.3696, + "step": 1599 + }, + { + "epoch": 0.035219018269865725, + "grad_norm": 1.8036097833565834, + "learning_rate": 9.99928632720279e-06, + "loss": 0.4379, + "step": 1600 + }, + { + "epoch": 0.035241030156284395, + "grad_norm": 0.7359182665389122, + "learning_rate": 9.999280292086518e-06, + "loss": 0.5979, + "step": 1601 + }, + { + "epoch": 0.03526304204270306, + "grad_norm": 2.053399772021395, + "learning_rate": 9.999274231561674e-06, + "loss": 0.3819, + "step": 1602 + }, + { + "epoch": 0.03528505392912173, + "grad_norm": 0.683247835878328, + "learning_rate": 9.999268145628288e-06, + "loss": 0.617, + "step": 1603 + }, + { + "epoch": 0.03530706581554039, + "grad_norm": 1.9807155711363207, + "learning_rate": 9.999262034286391e-06, + "loss": 0.4469, + "step": 1604 + }, + { + "epoch": 0.03532907770195906, + "grad_norm": 2.2702834935896523, + "learning_rate": 9.999255897536016e-06, + "loss": 0.4496, + "step": 1605 + }, + { + "epoch": 0.035351089588377724, + "grad_norm": 2.8290254533998866, + "learning_rate": 9.999249735377193e-06, + "loss": 0.4073, + "step": 1606 + }, + { + "epoch": 0.03537310147479639, + "grad_norm": 2.197351869789155, + "learning_rate": 9.999243547809953e-06, + "loss": 0.3979, + "step": 1607 + }, + { + "epoch": 0.03539511336121506, + "grad_norm": 3.1234552310486574, + "learning_rate": 9.999237334834327e-06, + "loss": 0.4082, + "step": 1608 + }, + { + "epoch": 0.03541712524763372, + "grad_norm": 4.934616622128041, + "learning_rate": 9.999231096450348e-06, + "loss": 0.4542, + "step": 1609 + }, + { + "epoch": 0.03543913713405239, + "grad_norm": 18.394318547249352, + "learning_rate": 9.999224832658046e-06, + "loss": 0.4445, + "step": 1610 + }, + { + "epoch": 0.03546114902047105, + "grad_norm": 0.8784405232347912, + "learning_rate": 9.999218543457454e-06, + "loss": 0.6269, + "step": 1611 + }, + { + "epoch": 0.03548316090688972, + "grad_norm": 1.9019522277244478, + "learning_rate": 9.999212228848602e-06, + "loss": 0.3108, + "step": 1612 + }, + { + "epoch": 0.035505172793308386, + "grad_norm": 3.1732648605866665, + "learning_rate": 9.999205888831527e-06, + "loss": 0.4293, + "step": 1613 + }, + { + "epoch": 0.035527184679727056, + "grad_norm": 1.826233759155285, + "learning_rate": 9.999199523406256e-06, + "loss": 0.4502, + "step": 1614 + }, + { + "epoch": 0.03554919656614572, + "grad_norm": 2.1130714412457134, + "learning_rate": 9.999193132572823e-06, + "loss": 0.4544, + "step": 1615 + }, + { + "epoch": 0.03557120845256438, + "grad_norm": 3.2231380415794124, + "learning_rate": 9.999186716331261e-06, + "loss": 0.3444, + "step": 1616 + }, + { + "epoch": 0.03559322033898305, + "grad_norm": 2.183175833583381, + "learning_rate": 9.999180274681603e-06, + "loss": 0.3649, + "step": 1617 + }, + { + "epoch": 0.035615232225401715, + "grad_norm": 2.285565433229778, + "learning_rate": 9.99917380762388e-06, + "loss": 0.4208, + "step": 1618 + }, + { + "epoch": 0.035637244111820385, + "grad_norm": 2.567840213609749, + "learning_rate": 9.999167315158128e-06, + "loss": 0.4528, + "step": 1619 + }, + { + "epoch": 0.03565925599823905, + "grad_norm": 1.9807464869273308, + "learning_rate": 9.999160797284378e-06, + "loss": 0.3851, + "step": 1620 + }, + { + "epoch": 0.03568126788465772, + "grad_norm": 2.6008992291024837, + "learning_rate": 9.999154254002662e-06, + "loss": 0.3783, + "step": 1621 + }, + { + "epoch": 0.03570327977107638, + "grad_norm": 3.276007493439764, + "learning_rate": 9.999147685313013e-06, + "loss": 0.3698, + "step": 1622 + }, + { + "epoch": 0.03572529165749505, + "grad_norm": 2.193825622947121, + "learning_rate": 9.999141091215468e-06, + "loss": 0.3542, + "step": 1623 + }, + { + "epoch": 0.035747303543913714, + "grad_norm": 2.459097879442177, + "learning_rate": 9.999134471710057e-06, + "loss": 0.4326, + "step": 1624 + }, + { + "epoch": 0.03576931543033238, + "grad_norm": 4.556290429879397, + "learning_rate": 9.999127826796816e-06, + "loss": 0.3691, + "step": 1625 + }, + { + "epoch": 0.03579132731675105, + "grad_norm": 0.9668559338984687, + "learning_rate": 9.999121156475777e-06, + "loss": 0.5922, + "step": 1626 + }, + { + "epoch": 0.03581333920316971, + "grad_norm": 2.343945122130847, + "learning_rate": 9.999114460746976e-06, + "loss": 0.3808, + "step": 1627 + }, + { + "epoch": 0.03583535108958838, + "grad_norm": 5.330624307383106, + "learning_rate": 9.999107739610443e-06, + "loss": 0.3073, + "step": 1628 + }, + { + "epoch": 0.03585736297600704, + "grad_norm": 2.719720581803954, + "learning_rate": 9.999100993066217e-06, + "loss": 0.4287, + "step": 1629 + }, + { + "epoch": 0.03587937486242571, + "grad_norm": 2.524467827893151, + "learning_rate": 9.99909422111433e-06, + "loss": 0.398, + "step": 1630 + }, + { + "epoch": 0.035901386748844376, + "grad_norm": 4.6900786307027955, + "learning_rate": 9.999087423754815e-06, + "loss": 0.3449, + "step": 1631 + }, + { + "epoch": 0.03592339863526304, + "grad_norm": 2.58276431085882, + "learning_rate": 9.99908060098771e-06, + "loss": 0.4213, + "step": 1632 + }, + { + "epoch": 0.03594541052168171, + "grad_norm": 1.6328327347301743, + "learning_rate": 9.999073752813046e-06, + "loss": 0.3303, + "step": 1633 + }, + { + "epoch": 0.03596742240810037, + "grad_norm": 0.7441448583210359, + "learning_rate": 9.99906687923086e-06, + "loss": 0.6014, + "step": 1634 + }, + { + "epoch": 0.03598943429451904, + "grad_norm": 2.5730804766065654, + "learning_rate": 9.999059980241187e-06, + "loss": 0.3721, + "step": 1635 + }, + { + "epoch": 0.036011446180937705, + "grad_norm": 2.2517830479509726, + "learning_rate": 9.999053055844061e-06, + "loss": 0.3592, + "step": 1636 + }, + { + "epoch": 0.036033458067356375, + "grad_norm": 2.422058580550241, + "learning_rate": 9.999046106039519e-06, + "loss": 0.3544, + "step": 1637 + }, + { + "epoch": 0.03605546995377504, + "grad_norm": 2.046587439324385, + "learning_rate": 9.999039130827593e-06, + "loss": 0.4302, + "step": 1638 + }, + { + "epoch": 0.03607748184019371, + "grad_norm": 1.7650070091940444, + "learning_rate": 9.999032130208322e-06, + "loss": 0.3877, + "step": 1639 + }, + { + "epoch": 0.03609949372661237, + "grad_norm": 0.7240513602597942, + "learning_rate": 9.99902510418174e-06, + "loss": 0.615, + "step": 1640 + }, + { + "epoch": 0.036121505613031034, + "grad_norm": 0.6801898247998961, + "learning_rate": 9.999018052747882e-06, + "loss": 0.6115, + "step": 1641 + }, + { + "epoch": 0.036143517499449704, + "grad_norm": 0.6719425566584905, + "learning_rate": 9.999010975906785e-06, + "loss": 0.593, + "step": 1642 + }, + { + "epoch": 0.03616552938586837, + "grad_norm": 1.4678548215616438, + "learning_rate": 9.999003873658485e-06, + "loss": 0.3632, + "step": 1643 + }, + { + "epoch": 0.03618754127228704, + "grad_norm": 4.506953044619028, + "learning_rate": 9.998996746003016e-06, + "loss": 0.4191, + "step": 1644 + }, + { + "epoch": 0.0362095531587057, + "grad_norm": 1.8335027828107993, + "learning_rate": 9.998989592940417e-06, + "loss": 0.4053, + "step": 1645 + }, + { + "epoch": 0.03623156504512437, + "grad_norm": 1.6617434995808003, + "learning_rate": 9.998982414470724e-06, + "loss": 0.3019, + "step": 1646 + }, + { + "epoch": 0.03625357693154303, + "grad_norm": 0.7909796976327115, + "learning_rate": 9.998975210593973e-06, + "loss": 0.6281, + "step": 1647 + }, + { + "epoch": 0.036275588817961696, + "grad_norm": 1.9244331623007742, + "learning_rate": 9.998967981310199e-06, + "loss": 0.3204, + "step": 1648 + }, + { + "epoch": 0.036297600704380366, + "grad_norm": 2.1985327075893943, + "learning_rate": 9.99896072661944e-06, + "loss": 0.4923, + "step": 1649 + }, + { + "epoch": 0.03631961259079903, + "grad_norm": 2.1037335405175495, + "learning_rate": 9.998953446521734e-06, + "loss": 0.4373, + "step": 1650 + }, + { + "epoch": 0.0363416244772177, + "grad_norm": 1.9363878154608904, + "learning_rate": 9.998946141017117e-06, + "loss": 0.36, + "step": 1651 + }, + { + "epoch": 0.03636363636363636, + "grad_norm": 0.6875035115424946, + "learning_rate": 9.998938810105627e-06, + "loss": 0.6486, + "step": 1652 + }, + { + "epoch": 0.03638564825005503, + "grad_norm": 0.6690972943836462, + "learning_rate": 9.998931453787298e-06, + "loss": 0.6186, + "step": 1653 + }, + { + "epoch": 0.036407660136473695, + "grad_norm": 0.6563605414803201, + "learning_rate": 9.99892407206217e-06, + "loss": 0.6226, + "step": 1654 + }, + { + "epoch": 0.036429672022892365, + "grad_norm": 2.5907280757366435, + "learning_rate": 9.998916664930283e-06, + "loss": 0.3987, + "step": 1655 + }, + { + "epoch": 0.03645168390931103, + "grad_norm": 2.3973482862397937, + "learning_rate": 9.99890923239167e-06, + "loss": 0.3121, + "step": 1656 + }, + { + "epoch": 0.03647369579572969, + "grad_norm": 2.329614920025456, + "learning_rate": 9.99890177444637e-06, + "loss": 0.3859, + "step": 1657 + }, + { + "epoch": 0.03649570768214836, + "grad_norm": 1.6079278303870572, + "learning_rate": 9.998894291094423e-06, + "loss": 0.387, + "step": 1658 + }, + { + "epoch": 0.036517719568567024, + "grad_norm": 0.7170686871168203, + "learning_rate": 9.998886782335866e-06, + "loss": 0.5801, + "step": 1659 + }, + { + "epoch": 0.036539731454985694, + "grad_norm": 1.6570020422980392, + "learning_rate": 9.998879248170736e-06, + "loss": 0.4042, + "step": 1660 + }, + { + "epoch": 0.03656174334140436, + "grad_norm": 0.7295063038343272, + "learning_rate": 9.998871688599072e-06, + "loss": 0.5778, + "step": 1661 + }, + { + "epoch": 0.03658375522782303, + "grad_norm": 1.7427565825027926, + "learning_rate": 9.998864103620912e-06, + "loss": 0.3749, + "step": 1662 + }, + { + "epoch": 0.03660576711424169, + "grad_norm": 3.6920726951309653, + "learning_rate": 9.998856493236296e-06, + "loss": 0.4208, + "step": 1663 + }, + { + "epoch": 0.03662777900066036, + "grad_norm": 1.7022450431120597, + "learning_rate": 9.998848857445262e-06, + "loss": 0.3576, + "step": 1664 + }, + { + "epoch": 0.03664979088707902, + "grad_norm": 1.8434442764699024, + "learning_rate": 9.998841196247848e-06, + "loss": 0.3671, + "step": 1665 + }, + { + "epoch": 0.036671802773497686, + "grad_norm": 0.7117650217110525, + "learning_rate": 9.998833509644092e-06, + "loss": 0.6152, + "step": 1666 + }, + { + "epoch": 0.036693814659916356, + "grad_norm": 2.5232397763316925, + "learning_rate": 9.998825797634037e-06, + "loss": 0.4164, + "step": 1667 + }, + { + "epoch": 0.03671582654633502, + "grad_norm": 2.0503672260122947, + "learning_rate": 9.99881806021772e-06, + "loss": 0.36, + "step": 1668 + }, + { + "epoch": 0.03673783843275369, + "grad_norm": 3.579825704915343, + "learning_rate": 9.998810297395178e-06, + "loss": 0.4186, + "step": 1669 + }, + { + "epoch": 0.03675985031917235, + "grad_norm": 1.9930018941253245, + "learning_rate": 9.998802509166452e-06, + "loss": 0.3943, + "step": 1670 + }, + { + "epoch": 0.03678186220559102, + "grad_norm": 1.7981230205101517, + "learning_rate": 9.998794695531584e-06, + "loss": 0.3936, + "step": 1671 + }, + { + "epoch": 0.036803874092009685, + "grad_norm": 0.7238316518830028, + "learning_rate": 9.998786856490612e-06, + "loss": 0.5825, + "step": 1672 + }, + { + "epoch": 0.03682588597842835, + "grad_norm": 1.549571916091774, + "learning_rate": 9.998778992043575e-06, + "loss": 0.3203, + "step": 1673 + }, + { + "epoch": 0.03684789786484702, + "grad_norm": 1.8496475209722605, + "learning_rate": 9.998771102190512e-06, + "loss": 0.3466, + "step": 1674 + }, + { + "epoch": 0.03686990975126568, + "grad_norm": 2.818040488294093, + "learning_rate": 9.998763186931468e-06, + "loss": 0.3935, + "step": 1675 + }, + { + "epoch": 0.03689192163768435, + "grad_norm": 3.383592796122553, + "learning_rate": 9.998755246266478e-06, + "loss": 0.4782, + "step": 1676 + }, + { + "epoch": 0.036913933524103014, + "grad_norm": 0.6897442439902116, + "learning_rate": 9.998747280195585e-06, + "loss": 0.6198, + "step": 1677 + }, + { + "epoch": 0.036935945410521684, + "grad_norm": 2.5232150175298527, + "learning_rate": 9.998739288718828e-06, + "loss": 0.3491, + "step": 1678 + }, + { + "epoch": 0.03695795729694035, + "grad_norm": 2.032607910651698, + "learning_rate": 9.998731271836249e-06, + "loss": 0.3953, + "step": 1679 + }, + { + "epoch": 0.03697996918335902, + "grad_norm": 3.065836909481969, + "learning_rate": 9.998723229547888e-06, + "loss": 0.4049, + "step": 1680 + }, + { + "epoch": 0.03700198106977768, + "grad_norm": 1.7270053858083707, + "learning_rate": 9.998715161853787e-06, + "loss": 0.393, + "step": 1681 + }, + { + "epoch": 0.03702399295619634, + "grad_norm": 1.9644430494281935, + "learning_rate": 9.998707068753985e-06, + "loss": 0.3449, + "step": 1682 + }, + { + "epoch": 0.03704600484261501, + "grad_norm": 1.873240838268799, + "learning_rate": 9.998698950248524e-06, + "loss": 0.3761, + "step": 1683 + }, + { + "epoch": 0.037068016729033676, + "grad_norm": 3.0884615762129877, + "learning_rate": 9.998690806337446e-06, + "loss": 0.4832, + "step": 1684 + }, + { + "epoch": 0.037090028615452346, + "grad_norm": 1.7226909736639373, + "learning_rate": 9.998682637020793e-06, + "loss": 0.4259, + "step": 1685 + }, + { + "epoch": 0.03711204050187101, + "grad_norm": 0.696491129324499, + "learning_rate": 9.998674442298604e-06, + "loss": 0.5817, + "step": 1686 + }, + { + "epoch": 0.03713405238828968, + "grad_norm": 1.6466458420072478, + "learning_rate": 9.998666222170921e-06, + "loss": 0.384, + "step": 1687 + }, + { + "epoch": 0.03715606427470834, + "grad_norm": 2.1392193959151884, + "learning_rate": 9.99865797663779e-06, + "loss": 0.3722, + "step": 1688 + }, + { + "epoch": 0.03717807616112701, + "grad_norm": 0.6576670478503942, + "learning_rate": 9.998649705699246e-06, + "loss": 0.6152, + "step": 1689 + }, + { + "epoch": 0.037200088047545675, + "grad_norm": 2.1300540061030637, + "learning_rate": 9.998641409355336e-06, + "loss": 0.4892, + "step": 1690 + }, + { + "epoch": 0.03722209993396434, + "grad_norm": 17.390781523165636, + "learning_rate": 9.998633087606102e-06, + "loss": 0.4134, + "step": 1691 + }, + { + "epoch": 0.03724411182038301, + "grad_norm": 10.929950242901558, + "learning_rate": 9.998624740451583e-06, + "loss": 0.3509, + "step": 1692 + }, + { + "epoch": 0.03726612370680167, + "grad_norm": 1.7926898146497376, + "learning_rate": 9.998616367891827e-06, + "loss": 0.3991, + "step": 1693 + }, + { + "epoch": 0.03728813559322034, + "grad_norm": 1.6617542406728594, + "learning_rate": 9.99860796992687e-06, + "loss": 0.4407, + "step": 1694 + }, + { + "epoch": 0.037310147479639004, + "grad_norm": 2.1722973593718415, + "learning_rate": 9.998599546556759e-06, + "loss": 0.3841, + "step": 1695 + }, + { + "epoch": 0.037332159366057674, + "grad_norm": 2.6860330650858844, + "learning_rate": 9.998591097781536e-06, + "loss": 0.3977, + "step": 1696 + }, + { + "epoch": 0.03735417125247634, + "grad_norm": 0.7351548713875303, + "learning_rate": 9.998582623601245e-06, + "loss": 0.6032, + "step": 1697 + }, + { + "epoch": 0.037376183138895, + "grad_norm": 0.7026224546457674, + "learning_rate": 9.998574124015925e-06, + "loss": 0.6358, + "step": 1698 + }, + { + "epoch": 0.03739819502531367, + "grad_norm": 0.6700980916788284, + "learning_rate": 9.998565599025622e-06, + "loss": 0.6231, + "step": 1699 + }, + { + "epoch": 0.03742020691173233, + "grad_norm": 2.573667225453592, + "learning_rate": 9.998557048630381e-06, + "loss": 0.3789, + "step": 1700 + }, + { + "epoch": 0.037442218798151, + "grad_norm": 2.147201025239279, + "learning_rate": 9.998548472830243e-06, + "loss": 0.3662, + "step": 1701 + }, + { + "epoch": 0.037464230684569666, + "grad_norm": 3.985808635091544, + "learning_rate": 9.998539871625252e-06, + "loss": 0.4908, + "step": 1702 + }, + { + "epoch": 0.037486242570988336, + "grad_norm": 4.089535639286889, + "learning_rate": 9.998531245015451e-06, + "loss": 0.3824, + "step": 1703 + }, + { + "epoch": 0.037508254457407, + "grad_norm": 1.7531944210868735, + "learning_rate": 9.998522593000886e-06, + "loss": 0.3476, + "step": 1704 + }, + { + "epoch": 0.03753026634382567, + "grad_norm": 2.835094806858525, + "learning_rate": 9.9985139155816e-06, + "loss": 0.4079, + "step": 1705 + }, + { + "epoch": 0.03755227823024433, + "grad_norm": 1.8023951204458986, + "learning_rate": 9.998505212757637e-06, + "loss": 0.3714, + "step": 1706 + }, + { + "epoch": 0.037574290116662995, + "grad_norm": 2.6623806491638873, + "learning_rate": 9.99849648452904e-06, + "loss": 0.3657, + "step": 1707 + }, + { + "epoch": 0.037596302003081665, + "grad_norm": 2.4262004444101612, + "learning_rate": 9.998487730895855e-06, + "loss": 0.3832, + "step": 1708 + }, + { + "epoch": 0.03761831388950033, + "grad_norm": 10.760859690558126, + "learning_rate": 9.998478951858125e-06, + "loss": 0.4041, + "step": 1709 + }, + { + "epoch": 0.037640325775919, + "grad_norm": 3.4495678661563547, + "learning_rate": 9.998470147415897e-06, + "loss": 0.3697, + "step": 1710 + }, + { + "epoch": 0.03766233766233766, + "grad_norm": 1.0503423750095713, + "learning_rate": 9.998461317569214e-06, + "loss": 0.6238, + "step": 1711 + }, + { + "epoch": 0.03768434954875633, + "grad_norm": 2.039522211548707, + "learning_rate": 9.998452462318122e-06, + "loss": 0.4408, + "step": 1712 + }, + { + "epoch": 0.037706361435174994, + "grad_norm": 1.9445555191746318, + "learning_rate": 9.998443581662665e-06, + "loss": 0.4927, + "step": 1713 + }, + { + "epoch": 0.037728373321593664, + "grad_norm": 2.0780489061283087, + "learning_rate": 9.998434675602888e-06, + "loss": 0.5064, + "step": 1714 + }, + { + "epoch": 0.03775038520801233, + "grad_norm": 2.0185094170885334, + "learning_rate": 9.998425744138835e-06, + "loss": 0.3802, + "step": 1715 + }, + { + "epoch": 0.03777239709443099, + "grad_norm": 0.7461984011474961, + "learning_rate": 9.998416787270556e-06, + "loss": 0.6255, + "step": 1716 + }, + { + "epoch": 0.03779440898084966, + "grad_norm": 3.643331864961918, + "learning_rate": 9.99840780499809e-06, + "loss": 0.4027, + "step": 1717 + }, + { + "epoch": 0.03781642086726832, + "grad_norm": 0.7516772683249976, + "learning_rate": 9.998398797321488e-06, + "loss": 0.5725, + "step": 1718 + }, + { + "epoch": 0.03783843275368699, + "grad_norm": 0.6932451112377703, + "learning_rate": 9.998389764240795e-06, + "loss": 0.614, + "step": 1719 + }, + { + "epoch": 0.037860444640105656, + "grad_norm": 2.8787267172691258, + "learning_rate": 9.998380705756054e-06, + "loss": 0.3378, + "step": 1720 + }, + { + "epoch": 0.037882456526524326, + "grad_norm": 1.909569490149495, + "learning_rate": 9.998371621867314e-06, + "loss": 0.3939, + "step": 1721 + }, + { + "epoch": 0.03790446841294299, + "grad_norm": 2.0362277303865053, + "learning_rate": 9.998362512574619e-06, + "loss": 0.3961, + "step": 1722 + }, + { + "epoch": 0.03792648029936165, + "grad_norm": 1.8398434913797008, + "learning_rate": 9.998353377878019e-06, + "loss": 0.399, + "step": 1723 + }, + { + "epoch": 0.03794849218578032, + "grad_norm": 2.732804518117676, + "learning_rate": 9.998344217777555e-06, + "loss": 0.4037, + "step": 1724 + }, + { + "epoch": 0.037970504072198985, + "grad_norm": 1.6785103428577137, + "learning_rate": 9.998335032273278e-06, + "loss": 0.3675, + "step": 1725 + }, + { + "epoch": 0.037992515958617655, + "grad_norm": 2.5834236546980067, + "learning_rate": 9.998325821365232e-06, + "loss": 0.3885, + "step": 1726 + }, + { + "epoch": 0.03801452784503632, + "grad_norm": 2.559857295238446, + "learning_rate": 9.998316585053466e-06, + "loss": 0.4658, + "step": 1727 + }, + { + "epoch": 0.03803653973145499, + "grad_norm": 2.3807657742886703, + "learning_rate": 9.998307323338025e-06, + "loss": 0.3996, + "step": 1728 + }, + { + "epoch": 0.03805855161787365, + "grad_norm": 2.7691177410898358, + "learning_rate": 9.998298036218956e-06, + "loss": 0.4393, + "step": 1729 + }, + { + "epoch": 0.03808056350429232, + "grad_norm": 2.323130597038375, + "learning_rate": 9.99828872369631e-06, + "loss": 0.4216, + "step": 1730 + }, + { + "epoch": 0.038102575390710984, + "grad_norm": 2.2465456074600367, + "learning_rate": 9.998279385770129e-06, + "loss": 0.465, + "step": 1731 + }, + { + "epoch": 0.03812458727712965, + "grad_norm": 2.137908360060173, + "learning_rate": 9.998270022440463e-06, + "loss": 0.3711, + "step": 1732 + }, + { + "epoch": 0.03814659916354832, + "grad_norm": 1.9964886493035896, + "learning_rate": 9.99826063370736e-06, + "loss": 0.3664, + "step": 1733 + }, + { + "epoch": 0.03816861104996698, + "grad_norm": 2.449498120077218, + "learning_rate": 9.998251219570867e-06, + "loss": 0.3924, + "step": 1734 + }, + { + "epoch": 0.03819062293638565, + "grad_norm": 4.055746343400832, + "learning_rate": 9.99824178003103e-06, + "loss": 0.4525, + "step": 1735 + }, + { + "epoch": 0.03821263482280431, + "grad_norm": 2.227061139381523, + "learning_rate": 9.998232315087904e-06, + "loss": 0.3983, + "step": 1736 + }, + { + "epoch": 0.03823464670922298, + "grad_norm": 1.2987909569685534, + "learning_rate": 9.998222824741527e-06, + "loss": 0.6293, + "step": 1737 + }, + { + "epoch": 0.038256658595641646, + "grad_norm": 2.545515469413742, + "learning_rate": 9.998213308991955e-06, + "loss": 0.3898, + "step": 1738 + }, + { + "epoch": 0.038278670482060316, + "grad_norm": 3.331490156229853, + "learning_rate": 9.998203767839234e-06, + "loss": 0.4147, + "step": 1739 + }, + { + "epoch": 0.03830068236847898, + "grad_norm": 2.663508178898255, + "learning_rate": 9.998194201283411e-06, + "loss": 0.3881, + "step": 1740 + }, + { + "epoch": 0.03832269425489764, + "grad_norm": 0.7163153265630416, + "learning_rate": 9.998184609324536e-06, + "loss": 0.6066, + "step": 1741 + }, + { + "epoch": 0.03834470614131631, + "grad_norm": 3.0987483094860084, + "learning_rate": 9.998174991962657e-06, + "loss": 0.4087, + "step": 1742 + }, + { + "epoch": 0.038366718027734975, + "grad_norm": 3.6306835089204834, + "learning_rate": 9.998165349197824e-06, + "loss": 0.4106, + "step": 1743 + }, + { + "epoch": 0.038388729914153645, + "grad_norm": 2.1208206764002235, + "learning_rate": 9.998155681030086e-06, + "loss": 0.3514, + "step": 1744 + }, + { + "epoch": 0.03841074180057231, + "grad_norm": 2.074715987419014, + "learning_rate": 9.998145987459492e-06, + "loss": 0.3967, + "step": 1745 + }, + { + "epoch": 0.03843275368699098, + "grad_norm": 3.136150049193577, + "learning_rate": 9.99813626848609e-06, + "loss": 0.3305, + "step": 1746 + }, + { + "epoch": 0.03845476557340964, + "grad_norm": 1.6880703403710333, + "learning_rate": 9.99812652410993e-06, + "loss": 0.3539, + "step": 1747 + }, + { + "epoch": 0.038476777459828304, + "grad_norm": 2.553831330523572, + "learning_rate": 9.998116754331062e-06, + "loss": 0.5161, + "step": 1748 + }, + { + "epoch": 0.038498789346246974, + "grad_norm": 1.7455024245531483, + "learning_rate": 9.998106959149536e-06, + "loss": 0.3153, + "step": 1749 + }, + { + "epoch": 0.03852080123266564, + "grad_norm": 2.549535112290723, + "learning_rate": 9.9980971385654e-06, + "loss": 0.5305, + "step": 1750 + }, + { + "epoch": 0.03854281311908431, + "grad_norm": 6.936149453429348, + "learning_rate": 9.998087292578707e-06, + "loss": 0.3497, + "step": 1751 + }, + { + "epoch": 0.03856482500550297, + "grad_norm": 1.1314346870942307, + "learning_rate": 9.998077421189503e-06, + "loss": 0.6353, + "step": 1752 + }, + { + "epoch": 0.03858683689192164, + "grad_norm": 2.0423723995900733, + "learning_rate": 9.998067524397842e-06, + "loss": 0.393, + "step": 1753 + }, + { + "epoch": 0.0386088487783403, + "grad_norm": 2.4291004255759936, + "learning_rate": 9.998057602203772e-06, + "loss": 0.3821, + "step": 1754 + }, + { + "epoch": 0.03863086066475897, + "grad_norm": 2.6203842670519695, + "learning_rate": 9.998047654607343e-06, + "loss": 0.3853, + "step": 1755 + }, + { + "epoch": 0.038652872551177636, + "grad_norm": 2.5216543295524976, + "learning_rate": 9.998037681608608e-06, + "loss": 0.3814, + "step": 1756 + }, + { + "epoch": 0.0386748844375963, + "grad_norm": 1.8120813013019008, + "learning_rate": 9.998027683207616e-06, + "loss": 0.3786, + "step": 1757 + }, + { + "epoch": 0.03869689632401497, + "grad_norm": 2.3312330796426393, + "learning_rate": 9.998017659404417e-06, + "loss": 0.449, + "step": 1758 + }, + { + "epoch": 0.03871890821043363, + "grad_norm": 2.2607752312938096, + "learning_rate": 9.998007610199065e-06, + "loss": 0.3653, + "step": 1759 + }, + { + "epoch": 0.0387409200968523, + "grad_norm": 3.458001351021414, + "learning_rate": 9.99799753559161e-06, + "loss": 0.5051, + "step": 1760 + }, + { + "epoch": 0.038762931983270965, + "grad_norm": 0.8335987959294199, + "learning_rate": 9.9979874355821e-06, + "loss": 0.5864, + "step": 1761 + }, + { + "epoch": 0.038784943869689635, + "grad_norm": 0.8291331906404907, + "learning_rate": 9.997977310170589e-06, + "loss": 0.6296, + "step": 1762 + }, + { + "epoch": 0.0388069557561083, + "grad_norm": 1.9914523034467164, + "learning_rate": 9.99796715935713e-06, + "loss": 0.4859, + "step": 1763 + }, + { + "epoch": 0.03882896764252696, + "grad_norm": 2.150418034294403, + "learning_rate": 9.99795698314177e-06, + "loss": 0.3781, + "step": 1764 + }, + { + "epoch": 0.03885097952894563, + "grad_norm": 2.680888528563441, + "learning_rate": 9.997946781524565e-06, + "loss": 0.4336, + "step": 1765 + }, + { + "epoch": 0.038872991415364294, + "grad_norm": 0.7209848754949291, + "learning_rate": 9.997936554505565e-06, + "loss": 0.6162, + "step": 1766 + }, + { + "epoch": 0.038895003301782964, + "grad_norm": 3.7500443213627497, + "learning_rate": 9.997926302084821e-06, + "loss": 0.3294, + "step": 1767 + }, + { + "epoch": 0.03891701518820163, + "grad_norm": 2.7594382682666976, + "learning_rate": 9.997916024262388e-06, + "loss": 0.4021, + "step": 1768 + }, + { + "epoch": 0.0389390270746203, + "grad_norm": 3.0597771459591177, + "learning_rate": 9.997905721038318e-06, + "loss": 0.3615, + "step": 1769 + }, + { + "epoch": 0.03896103896103896, + "grad_norm": 2.4443368664246075, + "learning_rate": 9.997895392412659e-06, + "loss": 0.4007, + "step": 1770 + }, + { + "epoch": 0.03898305084745763, + "grad_norm": 1.7964872478004013, + "learning_rate": 9.997885038385468e-06, + "loss": 0.3368, + "step": 1771 + }, + { + "epoch": 0.03900506273387629, + "grad_norm": 2.189513013661093, + "learning_rate": 9.997874658956796e-06, + "loss": 0.405, + "step": 1772 + }, + { + "epoch": 0.039027074620294956, + "grad_norm": 2.7942490923387377, + "learning_rate": 9.997864254126697e-06, + "loss": 0.3661, + "step": 1773 + }, + { + "epoch": 0.039049086506713626, + "grad_norm": 2.1700286524353896, + "learning_rate": 9.99785382389522e-06, + "loss": 0.3498, + "step": 1774 + }, + { + "epoch": 0.03907109839313229, + "grad_norm": 2.107406151896017, + "learning_rate": 9.997843368262423e-06, + "loss": 0.4889, + "step": 1775 + }, + { + "epoch": 0.03909311027955096, + "grad_norm": 2.7391918406921305, + "learning_rate": 9.997832887228357e-06, + "loss": 0.4334, + "step": 1776 + }, + { + "epoch": 0.03911512216596962, + "grad_norm": 4.038749365823912, + "learning_rate": 9.997822380793075e-06, + "loss": 0.3034, + "step": 1777 + }, + { + "epoch": 0.03913713405238829, + "grad_norm": 2.1520288809893686, + "learning_rate": 9.99781184895663e-06, + "loss": 0.5136, + "step": 1778 + }, + { + "epoch": 0.039159145938806955, + "grad_norm": 2.242087668425538, + "learning_rate": 9.997801291719076e-06, + "loss": 0.4633, + "step": 1779 + }, + { + "epoch": 0.039181157825225625, + "grad_norm": 2.6226865516134104, + "learning_rate": 9.997790709080466e-06, + "loss": 0.4259, + "step": 1780 + }, + { + "epoch": 0.03920316971164429, + "grad_norm": 2.392289683899932, + "learning_rate": 9.997780101040856e-06, + "loss": 0.3855, + "step": 1781 + }, + { + "epoch": 0.03922518159806295, + "grad_norm": 2.1353019263080206, + "learning_rate": 9.997769467600298e-06, + "loss": 0.42, + "step": 1782 + }, + { + "epoch": 0.03924719348448162, + "grad_norm": 3.726303095143431, + "learning_rate": 9.997758808758847e-06, + "loss": 0.5119, + "step": 1783 + }, + { + "epoch": 0.039269205370900284, + "grad_norm": 2.551660203300772, + "learning_rate": 9.997748124516556e-06, + "loss": 0.4873, + "step": 1784 + }, + { + "epoch": 0.039291217257318954, + "grad_norm": 2.825838471589817, + "learning_rate": 9.997737414873479e-06, + "loss": 0.4427, + "step": 1785 + }, + { + "epoch": 0.03931322914373762, + "grad_norm": 3.3884753460609067, + "learning_rate": 9.997726679829672e-06, + "loss": 0.3592, + "step": 1786 + }, + { + "epoch": 0.03933524103015629, + "grad_norm": 0.9627974936215736, + "learning_rate": 9.997715919385189e-06, + "loss": 0.6005, + "step": 1787 + }, + { + "epoch": 0.03935725291657495, + "grad_norm": 3.2189473054718136, + "learning_rate": 9.997705133540085e-06, + "loss": 0.3948, + "step": 1788 + }, + { + "epoch": 0.03937926480299361, + "grad_norm": 2.0962283992591875, + "learning_rate": 9.997694322294416e-06, + "loss": 0.3525, + "step": 1789 + }, + { + "epoch": 0.03940127668941228, + "grad_norm": 0.7187193689735588, + "learning_rate": 9.997683485648234e-06, + "loss": 0.5879, + "step": 1790 + }, + { + "epoch": 0.039423288575830946, + "grad_norm": 3.199640167811173, + "learning_rate": 9.997672623601594e-06, + "loss": 0.4172, + "step": 1791 + }, + { + "epoch": 0.039445300462249616, + "grad_norm": 2.9957780395799474, + "learning_rate": 9.997661736154555e-06, + "loss": 0.3755, + "step": 1792 + }, + { + "epoch": 0.03946731234866828, + "grad_norm": 2.6525843588530753, + "learning_rate": 9.997650823307169e-06, + "loss": 0.3888, + "step": 1793 + }, + { + "epoch": 0.03948932423508695, + "grad_norm": 2.279808114648385, + "learning_rate": 9.997639885059493e-06, + "loss": 0.396, + "step": 1794 + }, + { + "epoch": 0.03951133612150561, + "grad_norm": 2.714067973555696, + "learning_rate": 9.997628921411582e-06, + "loss": 0.4175, + "step": 1795 + }, + { + "epoch": 0.03953334800792428, + "grad_norm": 1.9333471412392997, + "learning_rate": 9.99761793236349e-06, + "loss": 0.5266, + "step": 1796 + }, + { + "epoch": 0.039555359894342945, + "grad_norm": 0.8432054863568674, + "learning_rate": 9.997606917915277e-06, + "loss": 0.5973, + "step": 1797 + }, + { + "epoch": 0.03957737178076161, + "grad_norm": 2.2308275148547305, + "learning_rate": 9.997595878066996e-06, + "loss": 0.4439, + "step": 1798 + }, + { + "epoch": 0.03959938366718028, + "grad_norm": 2.6718370571630574, + "learning_rate": 9.997584812818703e-06, + "loss": 0.3954, + "step": 1799 + }, + { + "epoch": 0.03962139555359894, + "grad_norm": 3.2824008299735166, + "learning_rate": 9.997573722170457e-06, + "loss": 0.4604, + "step": 1800 + }, + { + "epoch": 0.03964340744001761, + "grad_norm": 2.5128796673174745, + "learning_rate": 9.997562606122311e-06, + "loss": 0.4607, + "step": 1801 + }, + { + "epoch": 0.039665419326436274, + "grad_norm": 2.0074826063296385, + "learning_rate": 9.997551464674323e-06, + "loss": 0.3917, + "step": 1802 + }, + { + "epoch": 0.039687431212854944, + "grad_norm": 1.8497338834384387, + "learning_rate": 9.997540297826549e-06, + "loss": 0.4127, + "step": 1803 + }, + { + "epoch": 0.03970944309927361, + "grad_norm": 3.1506121014600077, + "learning_rate": 9.997529105579047e-06, + "loss": 0.336, + "step": 1804 + }, + { + "epoch": 0.03973145498569228, + "grad_norm": 2.3630910257670243, + "learning_rate": 9.99751788793187e-06, + "loss": 0.4155, + "step": 1805 + }, + { + "epoch": 0.03975346687211094, + "grad_norm": 0.7353612526254664, + "learning_rate": 9.997506644885082e-06, + "loss": 0.6128, + "step": 1806 + }, + { + "epoch": 0.0397754787585296, + "grad_norm": 1.994756682069636, + "learning_rate": 9.997495376438737e-06, + "loss": 0.4093, + "step": 1807 + }, + { + "epoch": 0.03979749064494827, + "grad_norm": 0.7009160985366969, + "learning_rate": 9.997484082592888e-06, + "loss": 0.6448, + "step": 1808 + }, + { + "epoch": 0.039819502531366936, + "grad_norm": 2.5614379877461984, + "learning_rate": 9.997472763347597e-06, + "loss": 0.3951, + "step": 1809 + }, + { + "epoch": 0.039841514417785606, + "grad_norm": 2.2613531473025548, + "learning_rate": 9.997461418702923e-06, + "loss": 0.3862, + "step": 1810 + }, + { + "epoch": 0.03986352630420427, + "grad_norm": 0.6639392757078172, + "learning_rate": 9.99745004865892e-06, + "loss": 0.6071, + "step": 1811 + }, + { + "epoch": 0.03988553819062294, + "grad_norm": 1.9255430378650134, + "learning_rate": 9.997438653215645e-06, + "loss": 0.5, + "step": 1812 + }, + { + "epoch": 0.0399075500770416, + "grad_norm": 2.480819558296067, + "learning_rate": 9.99742723237316e-06, + "loss": 0.4108, + "step": 1813 + }, + { + "epoch": 0.039929561963460265, + "grad_norm": 2.1014245261371785, + "learning_rate": 9.997415786131518e-06, + "loss": 0.3199, + "step": 1814 + }, + { + "epoch": 0.039951573849878935, + "grad_norm": 0.693200071736629, + "learning_rate": 9.997404314490783e-06, + "loss": 0.6091, + "step": 1815 + }, + { + "epoch": 0.0399735857362976, + "grad_norm": 2.266112338306848, + "learning_rate": 9.99739281745101e-06, + "loss": 0.356, + "step": 1816 + }, + { + "epoch": 0.03999559762271627, + "grad_norm": 10.955300226783116, + "learning_rate": 9.997381295012256e-06, + "loss": 0.4933, + "step": 1817 + }, + { + "epoch": 0.04001760950913493, + "grad_norm": 7.070929905803723, + "learning_rate": 9.997369747174583e-06, + "loss": 0.4163, + "step": 1818 + }, + { + "epoch": 0.0400396213955536, + "grad_norm": 2.034254803371337, + "learning_rate": 9.997358173938048e-06, + "loss": 0.3667, + "step": 1819 + }, + { + "epoch": 0.040061633281972264, + "grad_norm": 2.2568091031093553, + "learning_rate": 9.997346575302708e-06, + "loss": 0.3489, + "step": 1820 + }, + { + "epoch": 0.040083645168390934, + "grad_norm": 2.4547436983625675, + "learning_rate": 9.997334951268627e-06, + "loss": 0.3998, + "step": 1821 + }, + { + "epoch": 0.0401056570548096, + "grad_norm": 27.673418768555734, + "learning_rate": 9.997323301835859e-06, + "loss": 0.4484, + "step": 1822 + }, + { + "epoch": 0.04012766894122826, + "grad_norm": 5.4367353503966385, + "learning_rate": 9.997311627004464e-06, + "loss": 0.3705, + "step": 1823 + }, + { + "epoch": 0.04014968082764693, + "grad_norm": 2.734102429293642, + "learning_rate": 9.997299926774505e-06, + "loss": 0.2957, + "step": 1824 + }, + { + "epoch": 0.04017169271406559, + "grad_norm": 2.366065957458242, + "learning_rate": 9.997288201146038e-06, + "loss": 0.4673, + "step": 1825 + }, + { + "epoch": 0.04019370460048426, + "grad_norm": 2.2390438867795, + "learning_rate": 9.997276450119122e-06, + "loss": 0.4182, + "step": 1826 + }, + { + "epoch": 0.040215716486902926, + "grad_norm": 2.3102508049910986, + "learning_rate": 9.99726467369382e-06, + "loss": 0.487, + "step": 1827 + }, + { + "epoch": 0.040237728373321596, + "grad_norm": 1.8533244334583525, + "learning_rate": 9.997252871870188e-06, + "loss": 0.4314, + "step": 1828 + }, + { + "epoch": 0.04025974025974026, + "grad_norm": 3.3775136738411025, + "learning_rate": 9.997241044648291e-06, + "loss": 0.4816, + "step": 1829 + }, + { + "epoch": 0.04028175214615893, + "grad_norm": 4.390778731106698, + "learning_rate": 9.997229192028184e-06, + "loss": 0.3538, + "step": 1830 + }, + { + "epoch": 0.04030376403257759, + "grad_norm": 2.1844670777004898, + "learning_rate": 9.997217314009931e-06, + "loss": 0.3355, + "step": 1831 + }, + { + "epoch": 0.040325775918996255, + "grad_norm": 1.8001094957208617, + "learning_rate": 9.997205410593589e-06, + "loss": 0.3939, + "step": 1832 + }, + { + "epoch": 0.040347787805414925, + "grad_norm": 7.558226161634157, + "learning_rate": 9.997193481779222e-06, + "loss": 0.4335, + "step": 1833 + }, + { + "epoch": 0.04036979969183359, + "grad_norm": 0.7689599602778957, + "learning_rate": 9.997181527566888e-06, + "loss": 0.5999, + "step": 1834 + }, + { + "epoch": 0.04039181157825226, + "grad_norm": 3.25091947637301, + "learning_rate": 9.997169547956649e-06, + "loss": 0.333, + "step": 1835 + }, + { + "epoch": 0.04041382346467092, + "grad_norm": 4.311083669321607, + "learning_rate": 9.997157542948565e-06, + "loss": 0.3666, + "step": 1836 + }, + { + "epoch": 0.04043583535108959, + "grad_norm": 3.0030813564466214, + "learning_rate": 9.997145512542699e-06, + "loss": 0.4136, + "step": 1837 + }, + { + "epoch": 0.040457847237508254, + "grad_norm": 1.6058940606198215, + "learning_rate": 9.99713345673911e-06, + "loss": 0.3313, + "step": 1838 + }, + { + "epoch": 0.04047985912392692, + "grad_norm": 4.582897475806606, + "learning_rate": 9.99712137553786e-06, + "loss": 0.4353, + "step": 1839 + }, + { + "epoch": 0.04050187101034559, + "grad_norm": 3.0927977711278127, + "learning_rate": 9.99710926893901e-06, + "loss": 0.43, + "step": 1840 + }, + { + "epoch": 0.04052388289676425, + "grad_norm": 6.339958607069154, + "learning_rate": 9.99709713694262e-06, + "loss": 0.4023, + "step": 1841 + }, + { + "epoch": 0.04054589478318292, + "grad_norm": 1.6340271846754457, + "learning_rate": 9.997084979548757e-06, + "loss": 0.3522, + "step": 1842 + }, + { + "epoch": 0.04056790666960158, + "grad_norm": 2.233221229056008, + "learning_rate": 9.997072796757477e-06, + "loss": 0.4406, + "step": 1843 + }, + { + "epoch": 0.04058991855602025, + "grad_norm": 1.9393942397757569, + "learning_rate": 9.997060588568845e-06, + "loss": 0.3706, + "step": 1844 + }, + { + "epoch": 0.040611930442438916, + "grad_norm": 1.6709266016975433, + "learning_rate": 9.997048354982923e-06, + "loss": 0.3258, + "step": 1845 + }, + { + "epoch": 0.040633942328857586, + "grad_norm": 1.872157753508948, + "learning_rate": 9.997036095999773e-06, + "loss": 0.3873, + "step": 1846 + }, + { + "epoch": 0.04065595421527625, + "grad_norm": 3.036714688716851, + "learning_rate": 9.997023811619456e-06, + "loss": 0.4306, + "step": 1847 + }, + { + "epoch": 0.04067796610169491, + "grad_norm": 2.538722016907998, + "learning_rate": 9.997011501842033e-06, + "loss": 0.3977, + "step": 1848 + }, + { + "epoch": 0.04069997798811358, + "grad_norm": 0.8001213801418047, + "learning_rate": 9.996999166667571e-06, + "loss": 0.6208, + "step": 1849 + }, + { + "epoch": 0.040721989874532245, + "grad_norm": 2.467765786329154, + "learning_rate": 9.996986806096132e-06, + "loss": 0.4049, + "step": 1850 + }, + { + "epoch": 0.040744001760950915, + "grad_norm": 2.1430526668001506, + "learning_rate": 9.996974420127776e-06, + "loss": 0.3827, + "step": 1851 + }, + { + "epoch": 0.04076601364736958, + "grad_norm": 2.2556099652233184, + "learning_rate": 9.996962008762566e-06, + "loss": 0.392, + "step": 1852 + }, + { + "epoch": 0.04078802553378825, + "grad_norm": 5.477535922820726, + "learning_rate": 9.996949572000566e-06, + "loss": 0.4065, + "step": 1853 + }, + { + "epoch": 0.04081003742020691, + "grad_norm": 0.6847196743450311, + "learning_rate": 9.996937109841842e-06, + "loss": 0.6175, + "step": 1854 + }, + { + "epoch": 0.040832049306625574, + "grad_norm": 5.741550215912977, + "learning_rate": 9.996924622286454e-06, + "loss": 0.3512, + "step": 1855 + }, + { + "epoch": 0.040854061193044244, + "grad_norm": 0.6539887073669518, + "learning_rate": 9.996912109334466e-06, + "loss": 0.5908, + "step": 1856 + }, + { + "epoch": 0.04087607307946291, + "grad_norm": 2.1795102888645483, + "learning_rate": 9.996899570985943e-06, + "loss": 0.3822, + "step": 1857 + }, + { + "epoch": 0.04089808496588158, + "grad_norm": 6.373673132101706, + "learning_rate": 9.996887007240948e-06, + "loss": 0.4087, + "step": 1858 + }, + { + "epoch": 0.04092009685230024, + "grad_norm": 8.718989287493228, + "learning_rate": 9.996874418099542e-06, + "loss": 0.4063, + "step": 1859 + }, + { + "epoch": 0.04094210873871891, + "grad_norm": 2.4981028580280134, + "learning_rate": 9.996861803561794e-06, + "loss": 0.396, + "step": 1860 + }, + { + "epoch": 0.04096412062513757, + "grad_norm": 1.5664639360943091, + "learning_rate": 9.996849163627764e-06, + "loss": 0.3716, + "step": 1861 + }, + { + "epoch": 0.04098613251155624, + "grad_norm": 2.187786761343516, + "learning_rate": 9.99683649829752e-06, + "loss": 0.3439, + "step": 1862 + }, + { + "epoch": 0.041008144397974906, + "grad_norm": 2.3254996209995062, + "learning_rate": 9.996823807571122e-06, + "loss": 0.5138, + "step": 1863 + }, + { + "epoch": 0.04103015628439357, + "grad_norm": 2.923786543888465, + "learning_rate": 9.996811091448639e-06, + "loss": 0.4665, + "step": 1864 + }, + { + "epoch": 0.04105216817081224, + "grad_norm": 2.1549760404693004, + "learning_rate": 9.99679834993013e-06, + "loss": 0.3582, + "step": 1865 + }, + { + "epoch": 0.0410741800572309, + "grad_norm": 1.8271706685723164, + "learning_rate": 9.996785583015667e-06, + "loss": 0.3508, + "step": 1866 + }, + { + "epoch": 0.04109619194364957, + "grad_norm": 2.463738063625103, + "learning_rate": 9.99677279070531e-06, + "loss": 0.3946, + "step": 1867 + }, + { + "epoch": 0.041118203830068235, + "grad_norm": 2.980800274393335, + "learning_rate": 9.996759972999123e-06, + "loss": 0.3937, + "step": 1868 + }, + { + "epoch": 0.041140215716486905, + "grad_norm": 3.8200653365535504, + "learning_rate": 9.996747129897174e-06, + "loss": 0.371, + "step": 1869 + }, + { + "epoch": 0.04116222760290557, + "grad_norm": 4.532210733397917, + "learning_rate": 9.996734261399528e-06, + "loss": 0.3814, + "step": 1870 + }, + { + "epoch": 0.04118423948932424, + "grad_norm": 2.9333803527290163, + "learning_rate": 9.99672136750625e-06, + "loss": 0.3465, + "step": 1871 + }, + { + "epoch": 0.0412062513757429, + "grad_norm": 2.234510748805676, + "learning_rate": 9.996708448217405e-06, + "loss": 0.4159, + "step": 1872 + }, + { + "epoch": 0.041228263262161564, + "grad_norm": 1.8425035259626497, + "learning_rate": 9.99669550353306e-06, + "loss": 0.3109, + "step": 1873 + }, + { + "epoch": 0.041250275148580234, + "grad_norm": 0.7443837726960446, + "learning_rate": 9.996682533453279e-06, + "loss": 0.5977, + "step": 1874 + }, + { + "epoch": 0.0412722870349989, + "grad_norm": 2.574525218267642, + "learning_rate": 9.99666953797813e-06, + "loss": 0.3822, + "step": 1875 + }, + { + "epoch": 0.04129429892141757, + "grad_norm": 0.692315994071297, + "learning_rate": 9.996656517107675e-06, + "loss": 0.6032, + "step": 1876 + }, + { + "epoch": 0.04131631080783623, + "grad_norm": 12.197904218110862, + "learning_rate": 9.996643470841986e-06, + "loss": 0.3704, + "step": 1877 + }, + { + "epoch": 0.0413383226942549, + "grad_norm": 2.323294396644287, + "learning_rate": 9.996630399181124e-06, + "loss": 0.392, + "step": 1878 + }, + { + "epoch": 0.04136033458067356, + "grad_norm": 0.6681714212004093, + "learning_rate": 9.99661730212516e-06, + "loss": 0.5914, + "step": 1879 + }, + { + "epoch": 0.041382346467092226, + "grad_norm": 3.6588511263140084, + "learning_rate": 9.996604179674156e-06, + "loss": 0.389, + "step": 1880 + }, + { + "epoch": 0.041404358353510896, + "grad_norm": 2.5362915510463795, + "learning_rate": 9.996591031828182e-06, + "loss": 0.3996, + "step": 1881 + }, + { + "epoch": 0.04142637023992956, + "grad_norm": 3.2183344209829943, + "learning_rate": 9.996577858587306e-06, + "loss": 0.4467, + "step": 1882 + }, + { + "epoch": 0.04144838212634823, + "grad_norm": 2.187979492139034, + "learning_rate": 9.996564659951589e-06, + "loss": 0.3582, + "step": 1883 + }, + { + "epoch": 0.04147039401276689, + "grad_norm": 2.694252838998066, + "learning_rate": 9.996551435921105e-06, + "loss": 0.3826, + "step": 1884 + }, + { + "epoch": 0.04149240589918556, + "grad_norm": 2.2185945061454326, + "learning_rate": 9.996538186495917e-06, + "loss": 0.4327, + "step": 1885 + }, + { + "epoch": 0.041514417785604225, + "grad_norm": 0.7290980401184175, + "learning_rate": 9.996524911676093e-06, + "loss": 0.6283, + "step": 1886 + }, + { + "epoch": 0.041536429672022895, + "grad_norm": 0.6958673417619363, + "learning_rate": 9.9965116114617e-06, + "loss": 0.6214, + "step": 1887 + }, + { + "epoch": 0.04155844155844156, + "grad_norm": 2.4225335805636448, + "learning_rate": 9.996498285852808e-06, + "loss": 0.4633, + "step": 1888 + }, + { + "epoch": 0.04158045344486022, + "grad_norm": 3.2279010805267734, + "learning_rate": 9.996484934849482e-06, + "loss": 0.3568, + "step": 1889 + }, + { + "epoch": 0.04160246533127889, + "grad_norm": 1.926127343887242, + "learning_rate": 9.996471558451792e-06, + "loss": 0.3783, + "step": 1890 + }, + { + "epoch": 0.041624477217697554, + "grad_norm": 2.1290114425607807, + "learning_rate": 9.996458156659806e-06, + "loss": 0.366, + "step": 1891 + }, + { + "epoch": 0.041646489104116224, + "grad_norm": 1.964822987492441, + "learning_rate": 9.99644472947359e-06, + "loss": 0.4606, + "step": 1892 + }, + { + "epoch": 0.04166850099053489, + "grad_norm": 0.8415798169954282, + "learning_rate": 9.996431276893213e-06, + "loss": 0.5968, + "step": 1893 + }, + { + "epoch": 0.04169051287695356, + "grad_norm": 0.7433723565378625, + "learning_rate": 9.996417798918745e-06, + "loss": 0.6015, + "step": 1894 + }, + { + "epoch": 0.04171252476337222, + "grad_norm": 1.9190082157214319, + "learning_rate": 9.996404295550253e-06, + "loss": 0.3945, + "step": 1895 + }, + { + "epoch": 0.04173453664979089, + "grad_norm": 2.3925419420812117, + "learning_rate": 9.996390766787804e-06, + "loss": 0.4449, + "step": 1896 + }, + { + "epoch": 0.04175654853620955, + "grad_norm": 2.7359335975482724, + "learning_rate": 9.99637721263147e-06, + "loss": 0.3645, + "step": 1897 + }, + { + "epoch": 0.041778560422628216, + "grad_norm": 0.7878026633523452, + "learning_rate": 9.996363633081319e-06, + "loss": 0.5984, + "step": 1898 + }, + { + "epoch": 0.041800572309046886, + "grad_norm": 2.2809064875974547, + "learning_rate": 9.996350028137421e-06, + "loss": 0.3888, + "step": 1899 + }, + { + "epoch": 0.04182258419546555, + "grad_norm": 13.591315264277531, + "learning_rate": 9.996336397799841e-06, + "loss": 0.4166, + "step": 1900 + }, + { + "epoch": 0.04184459608188422, + "grad_norm": 3.4586216150686795, + "learning_rate": 9.996322742068651e-06, + "loss": 0.329, + "step": 1901 + }, + { + "epoch": 0.04186660796830288, + "grad_norm": 1.6649176992667654, + "learning_rate": 9.996309060943921e-06, + "loss": 0.3475, + "step": 1902 + }, + { + "epoch": 0.04188861985472155, + "grad_norm": 0.7905072135694328, + "learning_rate": 9.996295354425722e-06, + "loss": 0.5843, + "step": 1903 + }, + { + "epoch": 0.041910631741140215, + "grad_norm": 2.624041789608382, + "learning_rate": 9.99628162251412e-06, + "loss": 0.3773, + "step": 1904 + }, + { + "epoch": 0.04193264362755888, + "grad_norm": 1.6499918499563875, + "learning_rate": 9.996267865209186e-06, + "loss": 0.4033, + "step": 1905 + }, + { + "epoch": 0.04195465551397755, + "grad_norm": 4.7588946082973935, + "learning_rate": 9.996254082510989e-06, + "loss": 0.2872, + "step": 1906 + }, + { + "epoch": 0.04197666740039621, + "grad_norm": 3.803468662999869, + "learning_rate": 9.996240274419603e-06, + "loss": 0.407, + "step": 1907 + }, + { + "epoch": 0.04199867928681488, + "grad_norm": 1.866171216360038, + "learning_rate": 9.996226440935095e-06, + "loss": 0.3828, + "step": 1908 + }, + { + "epoch": 0.042020691173233544, + "grad_norm": 1.795680874379633, + "learning_rate": 9.996212582057535e-06, + "loss": 0.4035, + "step": 1909 + }, + { + "epoch": 0.042042703059652214, + "grad_norm": 2.4717687664766133, + "learning_rate": 9.996198697786994e-06, + "loss": 0.3981, + "step": 1910 + }, + { + "epoch": 0.04206471494607088, + "grad_norm": 2.4078901103708374, + "learning_rate": 9.996184788123544e-06, + "loss": 0.3829, + "step": 1911 + }, + { + "epoch": 0.04208672683248955, + "grad_norm": 3.201445322257545, + "learning_rate": 9.996170853067255e-06, + "loss": 0.3795, + "step": 1912 + }, + { + "epoch": 0.04210873871890821, + "grad_norm": 1.7189408990438857, + "learning_rate": 9.996156892618196e-06, + "loss": 0.3828, + "step": 1913 + }, + { + "epoch": 0.042130750605326873, + "grad_norm": 1.8709507120608668, + "learning_rate": 9.99614290677644e-06, + "loss": 0.4926, + "step": 1914 + }, + { + "epoch": 0.04215276249174554, + "grad_norm": 2.2411645666297084, + "learning_rate": 9.996128895542057e-06, + "loss": 0.3907, + "step": 1915 + }, + { + "epoch": 0.042174774378164206, + "grad_norm": 1.8274324574392264, + "learning_rate": 9.996114858915118e-06, + "loss": 0.2809, + "step": 1916 + }, + { + "epoch": 0.042196786264582876, + "grad_norm": 2.270681046185438, + "learning_rate": 9.996100796895696e-06, + "loss": 0.4484, + "step": 1917 + }, + { + "epoch": 0.04221879815100154, + "grad_norm": 1.817067387150196, + "learning_rate": 9.99608670948386e-06, + "loss": 0.3602, + "step": 1918 + }, + { + "epoch": 0.04224081003742021, + "grad_norm": 3.1655578215557223, + "learning_rate": 9.996072596679684e-06, + "loss": 0.3811, + "step": 1919 + }, + { + "epoch": 0.04226282192383887, + "grad_norm": 2.8046149983817013, + "learning_rate": 9.99605845848324e-06, + "loss": 0.4275, + "step": 1920 + }, + { + "epoch": 0.04228483381025754, + "grad_norm": 2.03636988186871, + "learning_rate": 9.996044294894598e-06, + "loss": 0.4464, + "step": 1921 + }, + { + "epoch": 0.042306845696676205, + "grad_norm": 3.13902455204832, + "learning_rate": 9.996030105913829e-06, + "loss": 0.3661, + "step": 1922 + }, + { + "epoch": 0.04232885758309487, + "grad_norm": 2.4240515081416563, + "learning_rate": 9.996015891541008e-06, + "loss": 0.4182, + "step": 1923 + }, + { + "epoch": 0.04235086946951354, + "grad_norm": 1.1299293129307646, + "learning_rate": 9.996001651776206e-06, + "loss": 0.6113, + "step": 1924 + }, + { + "epoch": 0.0423728813559322, + "grad_norm": 2.9921793070552325, + "learning_rate": 9.995987386619496e-06, + "loss": 0.3893, + "step": 1925 + }, + { + "epoch": 0.04239489324235087, + "grad_norm": 1.5802450165413977, + "learning_rate": 9.995973096070948e-06, + "loss": 0.3257, + "step": 1926 + }, + { + "epoch": 0.042416905128769535, + "grad_norm": 2.471166629638865, + "learning_rate": 9.995958780130638e-06, + "loss": 0.4357, + "step": 1927 + }, + { + "epoch": 0.042438917015188204, + "grad_norm": 1.7834297386143398, + "learning_rate": 9.995944438798638e-06, + "loss": 0.4761, + "step": 1928 + }, + { + "epoch": 0.04246092890160687, + "grad_norm": 1.777524088423046, + "learning_rate": 9.995930072075017e-06, + "loss": 0.4023, + "step": 1929 + }, + { + "epoch": 0.04248294078802553, + "grad_norm": 2.6571801098528183, + "learning_rate": 9.995915679959854e-06, + "loss": 0.3418, + "step": 1930 + }, + { + "epoch": 0.0425049526744442, + "grad_norm": 2.644646124779811, + "learning_rate": 9.995901262453218e-06, + "loss": 0.4955, + "step": 1931 + }, + { + "epoch": 0.042526964560862864, + "grad_norm": 3.013916514666715, + "learning_rate": 9.995886819555184e-06, + "loss": 0.4438, + "step": 1932 + }, + { + "epoch": 0.042548976447281533, + "grad_norm": 2.241476980394224, + "learning_rate": 9.995872351265822e-06, + "loss": 0.3542, + "step": 1933 + }, + { + "epoch": 0.042570988333700197, + "grad_norm": 1.9963947828332822, + "learning_rate": 9.995857857585214e-06, + "loss": 0.3743, + "step": 1934 + }, + { + "epoch": 0.042593000220118866, + "grad_norm": 2.1440174650714137, + "learning_rate": 9.995843338513425e-06, + "loss": 0.3786, + "step": 1935 + }, + { + "epoch": 0.04261501210653753, + "grad_norm": 2.453742206439528, + "learning_rate": 9.99582879405053e-06, + "loss": 0.4226, + "step": 1936 + }, + { + "epoch": 0.0426370239929562, + "grad_norm": 1.150457875769351, + "learning_rate": 9.995814224196609e-06, + "loss": 0.5977, + "step": 1937 + }, + { + "epoch": 0.04265903587937486, + "grad_norm": 0.9986643262004392, + "learning_rate": 9.99579962895173e-06, + "loss": 0.6159, + "step": 1938 + }, + { + "epoch": 0.042681047765793526, + "grad_norm": 1.6602978977334877, + "learning_rate": 9.995785008315969e-06, + "loss": 0.3716, + "step": 1939 + }, + { + "epoch": 0.042703059652212196, + "grad_norm": 3.5481026324336615, + "learning_rate": 9.995770362289401e-06, + "loss": 0.3873, + "step": 1940 + }, + { + "epoch": 0.04272507153863086, + "grad_norm": 1.8190824760666295, + "learning_rate": 9.995755690872098e-06, + "loss": 0.3632, + "step": 1941 + }, + { + "epoch": 0.04274708342504953, + "grad_norm": 0.9090802101946291, + "learning_rate": 9.995740994064137e-06, + "loss": 0.6007, + "step": 1942 + }, + { + "epoch": 0.04276909531146819, + "grad_norm": 1.6605429979640016, + "learning_rate": 9.995726271865593e-06, + "loss": 0.4096, + "step": 1943 + }, + { + "epoch": 0.04279110719788686, + "grad_norm": 1.021272713044115, + "learning_rate": 9.995711524276542e-06, + "loss": 0.577, + "step": 1944 + }, + { + "epoch": 0.042813119084305525, + "grad_norm": 20.191880782669426, + "learning_rate": 9.995696751297054e-06, + "loss": 0.4017, + "step": 1945 + }, + { + "epoch": 0.042835130970724195, + "grad_norm": 2.5602036804028274, + "learning_rate": 9.995681952927207e-06, + "loss": 0.4887, + "step": 1946 + }, + { + "epoch": 0.04285714285714286, + "grad_norm": 1.7095077145501167, + "learning_rate": 9.995667129167078e-06, + "loss": 0.3176, + "step": 1947 + }, + { + "epoch": 0.04287915474356152, + "grad_norm": 1.9116570035446985, + "learning_rate": 9.99565228001674e-06, + "loss": 0.4398, + "step": 1948 + }, + { + "epoch": 0.04290116662998019, + "grad_norm": 2.5480803900851705, + "learning_rate": 9.995637405476267e-06, + "loss": 0.3965, + "step": 1949 + }, + { + "epoch": 0.042923178516398854, + "grad_norm": 2.577387404158289, + "learning_rate": 9.995622505545738e-06, + "loss": 0.4659, + "step": 1950 + }, + { + "epoch": 0.042945190402817524, + "grad_norm": 0.792080565358374, + "learning_rate": 9.995607580225227e-06, + "loss": 0.6148, + "step": 1951 + }, + { + "epoch": 0.04296720228923619, + "grad_norm": 2.2041908440687386, + "learning_rate": 9.99559262951481e-06, + "loss": 0.458, + "step": 1952 + }, + { + "epoch": 0.04298921417565486, + "grad_norm": 2.0133497846309334, + "learning_rate": 9.995577653414563e-06, + "loss": 0.4019, + "step": 1953 + }, + { + "epoch": 0.04301122606207352, + "grad_norm": 1.8803046494062938, + "learning_rate": 9.995562651924563e-06, + "loss": 0.2974, + "step": 1954 + }, + { + "epoch": 0.04303323794849218, + "grad_norm": 1.9061809202948954, + "learning_rate": 9.995547625044887e-06, + "loss": 0.4076, + "step": 1955 + }, + { + "epoch": 0.04305524983491085, + "grad_norm": 2.1456283861762957, + "learning_rate": 9.995532572775609e-06, + "loss": 0.3653, + "step": 1956 + }, + { + "epoch": 0.043077261721329516, + "grad_norm": 1.6313028508909713, + "learning_rate": 9.995517495116805e-06, + "loss": 0.3424, + "step": 1957 + }, + { + "epoch": 0.043099273607748186, + "grad_norm": 1.940521418845058, + "learning_rate": 9.995502392068553e-06, + "loss": 0.3698, + "step": 1958 + }, + { + "epoch": 0.04312128549416685, + "grad_norm": 2.4005177477296984, + "learning_rate": 9.99548726363093e-06, + "loss": 0.4001, + "step": 1959 + }, + { + "epoch": 0.04314329738058552, + "grad_norm": 4.071029823847707, + "learning_rate": 9.995472109804014e-06, + "loss": 0.4563, + "step": 1960 + }, + { + "epoch": 0.04316530926700418, + "grad_norm": 1.7692019634804743, + "learning_rate": 9.99545693058788e-06, + "loss": 0.4022, + "step": 1961 + }, + { + "epoch": 0.04318732115342285, + "grad_norm": 2.2296575080639114, + "learning_rate": 9.995441725982604e-06, + "loss": 0.4062, + "step": 1962 + }, + { + "epoch": 0.043209333039841515, + "grad_norm": 3.4218353911158745, + "learning_rate": 9.995426495988267e-06, + "loss": 0.4377, + "step": 1963 + }, + { + "epoch": 0.04323134492626018, + "grad_norm": 2.3721133474278564, + "learning_rate": 9.995411240604946e-06, + "loss": 0.4095, + "step": 1964 + }, + { + "epoch": 0.04325335681267885, + "grad_norm": 0.8154777712216386, + "learning_rate": 9.995395959832715e-06, + "loss": 0.6254, + "step": 1965 + }, + { + "epoch": 0.04327536869909751, + "grad_norm": 7.10308336001453, + "learning_rate": 9.995380653671654e-06, + "loss": 0.4033, + "step": 1966 + }, + { + "epoch": 0.04329738058551618, + "grad_norm": 5.579970549884728, + "learning_rate": 9.995365322121841e-06, + "loss": 0.4006, + "step": 1967 + }, + { + "epoch": 0.043319392471934844, + "grad_norm": 2.089615622815762, + "learning_rate": 9.995349965183352e-06, + "loss": 0.2973, + "step": 1968 + }, + { + "epoch": 0.043341404358353514, + "grad_norm": 0.6880310178927104, + "learning_rate": 9.995334582856268e-06, + "loss": 0.585, + "step": 1969 + }, + { + "epoch": 0.04336341624477218, + "grad_norm": 2.0835113348268974, + "learning_rate": 9.995319175140665e-06, + "loss": 0.344, + "step": 1970 + }, + { + "epoch": 0.04338542813119084, + "grad_norm": 1.5532573855262894, + "learning_rate": 9.995303742036622e-06, + "loss": 0.3451, + "step": 1971 + }, + { + "epoch": 0.04340744001760951, + "grad_norm": 2.1327244463576442, + "learning_rate": 9.995288283544218e-06, + "loss": 0.4497, + "step": 1972 + }, + { + "epoch": 0.04342945190402817, + "grad_norm": 2.1709142883338983, + "learning_rate": 9.995272799663532e-06, + "loss": 0.3689, + "step": 1973 + }, + { + "epoch": 0.04345146379044684, + "grad_norm": 2.1591051858283143, + "learning_rate": 9.995257290394639e-06, + "loss": 0.3811, + "step": 1974 + }, + { + "epoch": 0.043473475676865506, + "grad_norm": 1.9505646607684641, + "learning_rate": 9.995241755737621e-06, + "loss": 0.3661, + "step": 1975 + }, + { + "epoch": 0.043495487563284176, + "grad_norm": 2.119482634855179, + "learning_rate": 9.995226195692558e-06, + "loss": 0.3756, + "step": 1976 + }, + { + "epoch": 0.04351749944970284, + "grad_norm": 1.7868703553979777, + "learning_rate": 9.995210610259526e-06, + "loss": 0.3773, + "step": 1977 + }, + { + "epoch": 0.04353951133612151, + "grad_norm": 2.3589284694696087, + "learning_rate": 9.995194999438607e-06, + "loss": 0.5503, + "step": 1978 + }, + { + "epoch": 0.04356152322254017, + "grad_norm": 2.056260925349671, + "learning_rate": 9.995179363229879e-06, + "loss": 0.4531, + "step": 1979 + }, + { + "epoch": 0.043583535108958835, + "grad_norm": 1.9803277323715194, + "learning_rate": 9.99516370163342e-06, + "loss": 0.3716, + "step": 1980 + }, + { + "epoch": 0.043605546995377505, + "grad_norm": 0.8147644090903364, + "learning_rate": 9.995148014649314e-06, + "loss": 0.6094, + "step": 1981 + }, + { + "epoch": 0.04362755888179617, + "grad_norm": 1.8236560021795727, + "learning_rate": 9.995132302277635e-06, + "loss": 0.3708, + "step": 1982 + }, + { + "epoch": 0.04364957076821484, + "grad_norm": 1.9307318975156873, + "learning_rate": 9.995116564518468e-06, + "loss": 0.3194, + "step": 1983 + }, + { + "epoch": 0.0436715826546335, + "grad_norm": 1.927381069540463, + "learning_rate": 9.99510080137189e-06, + "loss": 0.361, + "step": 1984 + }, + { + "epoch": 0.04369359454105217, + "grad_norm": 2.4759456824317665, + "learning_rate": 9.995085012837981e-06, + "loss": 0.3296, + "step": 1985 + }, + { + "epoch": 0.043715606427470834, + "grad_norm": 3.4333423513723034, + "learning_rate": 9.995069198916823e-06, + "loss": 0.3404, + "step": 1986 + }, + { + "epoch": 0.043737618313889504, + "grad_norm": 1.691698769839726, + "learning_rate": 9.995053359608495e-06, + "loss": 0.3412, + "step": 1987 + }, + { + "epoch": 0.04375963020030817, + "grad_norm": 3.355717551163306, + "learning_rate": 9.995037494913079e-06, + "loss": 0.3576, + "step": 1988 + }, + { + "epoch": 0.04378164208672683, + "grad_norm": 2.1354357974804365, + "learning_rate": 9.995021604830653e-06, + "loss": 0.4606, + "step": 1989 + }, + { + "epoch": 0.0438036539731455, + "grad_norm": 2.9850251310832703, + "learning_rate": 9.9950056893613e-06, + "loss": 0.344, + "step": 1990 + }, + { + "epoch": 0.04382566585956416, + "grad_norm": 13.157288560530704, + "learning_rate": 9.994989748505099e-06, + "loss": 0.2953, + "step": 1991 + }, + { + "epoch": 0.04384767774598283, + "grad_norm": 1.6879358148574581, + "learning_rate": 9.994973782262135e-06, + "loss": 0.4227, + "step": 1992 + }, + { + "epoch": 0.043869689632401496, + "grad_norm": 1.881000725729099, + "learning_rate": 9.994957790632485e-06, + "loss": 0.4863, + "step": 1993 + }, + { + "epoch": 0.043891701518820166, + "grad_norm": 1.9124586115661146, + "learning_rate": 9.994941773616231e-06, + "loss": 0.4583, + "step": 1994 + }, + { + "epoch": 0.04391371340523883, + "grad_norm": 0.7263700971885574, + "learning_rate": 9.994925731213457e-06, + "loss": 0.5861, + "step": 1995 + }, + { + "epoch": 0.04393572529165749, + "grad_norm": 1.7398212326826636, + "learning_rate": 9.994909663424242e-06, + "loss": 0.3873, + "step": 1996 + }, + { + "epoch": 0.04395773717807616, + "grad_norm": 1.9862099324453524, + "learning_rate": 9.994893570248668e-06, + "loss": 0.4017, + "step": 1997 + }, + { + "epoch": 0.043979749064494825, + "grad_norm": 2.204063795780504, + "learning_rate": 9.994877451686814e-06, + "loss": 0.37, + "step": 1998 + }, + { + "epoch": 0.044001760950913495, + "grad_norm": 1.8563416266676878, + "learning_rate": 9.994861307738769e-06, + "loss": 0.3907, + "step": 1999 + }, + { + "epoch": 0.04402377283733216, + "grad_norm": 1.8642146201179826, + "learning_rate": 9.99484513840461e-06, + "loss": 0.4217, + "step": 2000 + }, + { + "epoch": 0.04404578472375083, + "grad_norm": 2.441563062006074, + "learning_rate": 9.99482894368442e-06, + "loss": 0.4469, + "step": 2001 + }, + { + "epoch": 0.04406779661016949, + "grad_norm": 0.6593980825754897, + "learning_rate": 9.994812723578282e-06, + "loss": 0.6224, + "step": 2002 + }, + { + "epoch": 0.04408980849658816, + "grad_norm": 6.806315405770816, + "learning_rate": 9.994796478086277e-06, + "loss": 0.3971, + "step": 2003 + }, + { + "epoch": 0.044111820383006824, + "grad_norm": 18.65455995629888, + "learning_rate": 9.99478020720849e-06, + "loss": 0.4265, + "step": 2004 + }, + { + "epoch": 0.04413383226942549, + "grad_norm": 1.8718032885458948, + "learning_rate": 9.994763910945002e-06, + "loss": 0.3903, + "step": 2005 + }, + { + "epoch": 0.04415584415584416, + "grad_norm": 2.1832247129087614, + "learning_rate": 9.994747589295895e-06, + "loss": 0.3609, + "step": 2006 + }, + { + "epoch": 0.04417785604226282, + "grad_norm": 2.233903480312773, + "learning_rate": 9.994731242261253e-06, + "loss": 0.3024, + "step": 2007 + }, + { + "epoch": 0.04419986792868149, + "grad_norm": 1.9447341399441371, + "learning_rate": 9.99471486984116e-06, + "loss": 0.4247, + "step": 2008 + }, + { + "epoch": 0.04422187981510015, + "grad_norm": 1.903971193773395, + "learning_rate": 9.994698472035699e-06, + "loss": 0.3818, + "step": 2009 + }, + { + "epoch": 0.04424389170151882, + "grad_norm": 2.052877039070874, + "learning_rate": 9.994682048844951e-06, + "loss": 0.3633, + "step": 2010 + }, + { + "epoch": 0.044265903587937486, + "grad_norm": 3.8078937655503418, + "learning_rate": 9.994665600269002e-06, + "loss": 0.4931, + "step": 2011 + }, + { + "epoch": 0.044287915474356156, + "grad_norm": 2.172228520749762, + "learning_rate": 9.994649126307935e-06, + "loss": 0.3306, + "step": 2012 + }, + { + "epoch": 0.04430992736077482, + "grad_norm": 0.7375659879999413, + "learning_rate": 9.994632626961833e-06, + "loss": 0.6055, + "step": 2013 + }, + { + "epoch": 0.04433193924719348, + "grad_norm": 2.6247431476736427, + "learning_rate": 9.994616102230781e-06, + "loss": 0.3459, + "step": 2014 + }, + { + "epoch": 0.04435395113361215, + "grad_norm": 2.4580113564270376, + "learning_rate": 9.99459955211486e-06, + "loss": 0.5289, + "step": 2015 + }, + { + "epoch": 0.044375963020030815, + "grad_norm": 2.072471306808831, + "learning_rate": 9.994582976614158e-06, + "loss": 0.3609, + "step": 2016 + }, + { + "epoch": 0.044397974906449485, + "grad_norm": 0.6718578118914174, + "learning_rate": 9.994566375728758e-06, + "loss": 0.624, + "step": 2017 + }, + { + "epoch": 0.04441998679286815, + "grad_norm": 1.8097688145146675, + "learning_rate": 9.994549749458744e-06, + "loss": 0.4074, + "step": 2018 + }, + { + "epoch": 0.04444199867928682, + "grad_norm": 3.98978053190249, + "learning_rate": 9.994533097804202e-06, + "loss": 0.3761, + "step": 2019 + }, + { + "epoch": 0.04446401056570548, + "grad_norm": 1.6238058183419557, + "learning_rate": 9.994516420765212e-06, + "loss": 0.3899, + "step": 2020 + }, + { + "epoch": 0.044486022452124144, + "grad_norm": 0.6645436589152892, + "learning_rate": 9.994499718341864e-06, + "loss": 0.6, + "step": 2021 + }, + { + "epoch": 0.044508034338542814, + "grad_norm": 0.6877827756587513, + "learning_rate": 9.99448299053424e-06, + "loss": 0.6164, + "step": 2022 + }, + { + "epoch": 0.04453004622496148, + "grad_norm": 18.737548518131486, + "learning_rate": 9.994466237342426e-06, + "loss": 0.384, + "step": 2023 + }, + { + "epoch": 0.04455205811138015, + "grad_norm": 2.1890851920003347, + "learning_rate": 9.994449458766508e-06, + "loss": 0.3632, + "step": 2024 + }, + { + "epoch": 0.04457406999779881, + "grad_norm": 3.1152476154750737, + "learning_rate": 9.99443265480657e-06, + "loss": 0.3569, + "step": 2025 + }, + { + "epoch": 0.04459608188421748, + "grad_norm": 1.7785889316922647, + "learning_rate": 9.994415825462696e-06, + "loss": 0.4832, + "step": 2026 + }, + { + "epoch": 0.04461809377063614, + "grad_norm": 2.1604997865966147, + "learning_rate": 9.994398970734974e-06, + "loss": 0.4383, + "step": 2027 + }, + { + "epoch": 0.04464010565705481, + "grad_norm": 2.061150257320768, + "learning_rate": 9.994382090623489e-06, + "loss": 0.4157, + "step": 2028 + }, + { + "epoch": 0.044662117543473476, + "grad_norm": 2.4137434782366434, + "learning_rate": 9.994365185128327e-06, + "loss": 0.4513, + "step": 2029 + }, + { + "epoch": 0.04468412942989214, + "grad_norm": 0.6644248654393287, + "learning_rate": 9.994348254249573e-06, + "loss": 0.5934, + "step": 2030 + }, + { + "epoch": 0.04470614131631081, + "grad_norm": 0.6453575985558663, + "learning_rate": 9.994331297987314e-06, + "loss": 0.6044, + "step": 2031 + }, + { + "epoch": 0.04472815320272947, + "grad_norm": 2.3326129632897414, + "learning_rate": 9.994314316341637e-06, + "loss": 0.3857, + "step": 2032 + }, + { + "epoch": 0.04475016508914814, + "grad_norm": 2.0887014446281174, + "learning_rate": 9.994297309312626e-06, + "loss": 0.3681, + "step": 2033 + }, + { + "epoch": 0.044772176975566805, + "grad_norm": 2.466549756595247, + "learning_rate": 9.994280276900368e-06, + "loss": 0.3701, + "step": 2034 + }, + { + "epoch": 0.044794188861985475, + "grad_norm": 1.9960433234693145, + "learning_rate": 9.99426321910495e-06, + "loss": 0.3814, + "step": 2035 + }, + { + "epoch": 0.04481620074840414, + "grad_norm": 2.1945498829700116, + "learning_rate": 9.99424613592646e-06, + "loss": 0.3701, + "step": 2036 + }, + { + "epoch": 0.04483821263482281, + "grad_norm": 2.591626841377076, + "learning_rate": 9.994229027364983e-06, + "loss": 0.3318, + "step": 2037 + }, + { + "epoch": 0.04486022452124147, + "grad_norm": 5.441213343444078, + "learning_rate": 9.994211893420607e-06, + "loss": 0.4001, + "step": 2038 + }, + { + "epoch": 0.044882236407660134, + "grad_norm": 0.6992384736413554, + "learning_rate": 9.994194734093418e-06, + "loss": 0.6235, + "step": 2039 + }, + { + "epoch": 0.044904248294078804, + "grad_norm": 0.6643239979546633, + "learning_rate": 9.994177549383503e-06, + "loss": 0.607, + "step": 2040 + }, + { + "epoch": 0.04492626018049747, + "grad_norm": 3.5739434132128136, + "learning_rate": 9.994160339290952e-06, + "loss": 0.3927, + "step": 2041 + }, + { + "epoch": 0.04494827206691614, + "grad_norm": 0.6691579381934416, + "learning_rate": 9.99414310381585e-06, + "loss": 0.6055, + "step": 2042 + }, + { + "epoch": 0.0449702839533348, + "grad_norm": 1.9897269199160221, + "learning_rate": 9.994125842958287e-06, + "loss": 0.3779, + "step": 2043 + }, + { + "epoch": 0.04499229583975347, + "grad_norm": 1.6279032027439388, + "learning_rate": 9.994108556718347e-06, + "loss": 0.4279, + "step": 2044 + }, + { + "epoch": 0.04501430772617213, + "grad_norm": 2.116963521083424, + "learning_rate": 9.99409124509612e-06, + "loss": 0.3549, + "step": 2045 + }, + { + "epoch": 0.045036319612590796, + "grad_norm": 2.216304223400271, + "learning_rate": 9.994073908091694e-06, + "loss": 0.4055, + "step": 2046 + }, + { + "epoch": 0.045058331499009466, + "grad_norm": 3.248171363613951, + "learning_rate": 9.994056545705158e-06, + "loss": 0.4153, + "step": 2047 + }, + { + "epoch": 0.04508034338542813, + "grad_norm": 2.0471321093647528, + "learning_rate": 9.994039157936597e-06, + "loss": 0.3816, + "step": 2048 + }, + { + "epoch": 0.0451023552718468, + "grad_norm": 0.7029757440697131, + "learning_rate": 9.994021744786105e-06, + "loss": 0.5789, + "step": 2049 + }, + { + "epoch": 0.04512436715826546, + "grad_norm": 3.109765611897399, + "learning_rate": 9.994004306253763e-06, + "loss": 0.4104, + "step": 2050 + }, + { + "epoch": 0.04514637904468413, + "grad_norm": 2.3311367174597786, + "learning_rate": 9.993986842339668e-06, + "loss": 0.4351, + "step": 2051 + }, + { + "epoch": 0.045168390931102795, + "grad_norm": 4.019074901199281, + "learning_rate": 9.993969353043901e-06, + "loss": 0.4681, + "step": 2052 + }, + { + "epoch": 0.045190402817521465, + "grad_norm": 1.986491161066665, + "learning_rate": 9.993951838366556e-06, + "loss": 0.3558, + "step": 2053 + }, + { + "epoch": 0.04521241470394013, + "grad_norm": 1.5552134182199429, + "learning_rate": 9.993934298307718e-06, + "loss": 0.4529, + "step": 2054 + }, + { + "epoch": 0.04523442659035879, + "grad_norm": 2.162305088246212, + "learning_rate": 9.993916732867483e-06, + "loss": 0.2972, + "step": 2055 + }, + { + "epoch": 0.04525643847677746, + "grad_norm": 2.5680518685988254, + "learning_rate": 9.993899142045933e-06, + "loss": 0.4264, + "step": 2056 + }, + { + "epoch": 0.045278450363196124, + "grad_norm": 1.999467943083765, + "learning_rate": 9.99388152584316e-06, + "loss": 0.3233, + "step": 2057 + }, + { + "epoch": 0.045300462249614794, + "grad_norm": 1.8571430248845604, + "learning_rate": 9.993863884259256e-06, + "loss": 0.448, + "step": 2058 + }, + { + "epoch": 0.04532247413603346, + "grad_norm": 2.2151488022524113, + "learning_rate": 9.993846217294307e-06, + "loss": 0.4264, + "step": 2059 + }, + { + "epoch": 0.04534448602245213, + "grad_norm": 2.096119894102263, + "learning_rate": 9.993828524948404e-06, + "loss": 0.3953, + "step": 2060 + }, + { + "epoch": 0.04536649790887079, + "grad_norm": 2.4259664639350365, + "learning_rate": 9.993810807221636e-06, + "loss": 0.3843, + "step": 2061 + }, + { + "epoch": 0.04538850979528945, + "grad_norm": 1.7241821727439821, + "learning_rate": 9.993793064114096e-06, + "loss": 0.3985, + "step": 2062 + }, + { + "epoch": 0.04541052168170812, + "grad_norm": 2.90805584068316, + "learning_rate": 9.993775295625872e-06, + "loss": 0.4417, + "step": 2063 + }, + { + "epoch": 0.045432533568126786, + "grad_norm": 2.3834802040176446, + "learning_rate": 9.993757501757052e-06, + "loss": 0.3709, + "step": 2064 + }, + { + "epoch": 0.045454545454545456, + "grad_norm": 4.471349874209077, + "learning_rate": 9.993739682507733e-06, + "loss": 0.3546, + "step": 2065 + }, + { + "epoch": 0.04547655734096412, + "grad_norm": 1.86106273846894, + "learning_rate": 9.993721837878001e-06, + "loss": 0.403, + "step": 2066 + }, + { + "epoch": 0.04549856922738279, + "grad_norm": 6.00881241456269, + "learning_rate": 9.993703967867946e-06, + "loss": 0.4031, + "step": 2067 + }, + { + "epoch": 0.04552058111380145, + "grad_norm": 2.1417817447433514, + "learning_rate": 9.993686072477659e-06, + "loss": 0.4785, + "step": 2068 + }, + { + "epoch": 0.04554259300022012, + "grad_norm": 1.769316456630498, + "learning_rate": 9.993668151707234e-06, + "loss": 0.3741, + "step": 2069 + }, + { + "epoch": 0.045564604886638785, + "grad_norm": 6.693153737023305, + "learning_rate": 9.993650205556762e-06, + "loss": 0.3759, + "step": 2070 + }, + { + "epoch": 0.04558661677305745, + "grad_norm": 1.7238451692255297, + "learning_rate": 9.99363223402633e-06, + "loss": 0.3717, + "step": 2071 + }, + { + "epoch": 0.04560862865947612, + "grad_norm": 2.4308301940042565, + "learning_rate": 9.993614237116032e-06, + "loss": 0.4148, + "step": 2072 + }, + { + "epoch": 0.04563064054589478, + "grad_norm": 1.5724939189187135, + "learning_rate": 9.993596214825959e-06, + "loss": 0.3735, + "step": 2073 + }, + { + "epoch": 0.04565265243231345, + "grad_norm": 2.3712899518134547, + "learning_rate": 9.993578167156204e-06, + "loss": 0.42, + "step": 2074 + }, + { + "epoch": 0.045674664318732114, + "grad_norm": 0.7480288159640079, + "learning_rate": 9.993560094106856e-06, + "loss": 0.5999, + "step": 2075 + }, + { + "epoch": 0.045696676205150784, + "grad_norm": 1.83400642804707, + "learning_rate": 9.99354199567801e-06, + "loss": 0.4287, + "step": 2076 + }, + { + "epoch": 0.04571868809156945, + "grad_norm": 2.4545487341004892, + "learning_rate": 9.993523871869755e-06, + "loss": 0.5009, + "step": 2077 + }, + { + "epoch": 0.04574069997798812, + "grad_norm": 0.6717040740204219, + "learning_rate": 9.993505722682185e-06, + "loss": 0.5934, + "step": 2078 + }, + { + "epoch": 0.04576271186440678, + "grad_norm": 2.072954450593836, + "learning_rate": 9.993487548115392e-06, + "loss": 0.4252, + "step": 2079 + }, + { + "epoch": 0.04578472375082544, + "grad_norm": 2.363333475310388, + "learning_rate": 9.99346934816947e-06, + "loss": 0.4481, + "step": 2080 + }, + { + "epoch": 0.04580673563724411, + "grad_norm": 2.144422862361454, + "learning_rate": 9.993451122844507e-06, + "loss": 0.425, + "step": 2081 + }, + { + "epoch": 0.045828747523662776, + "grad_norm": 2.099761589302513, + "learning_rate": 9.993432872140599e-06, + "loss": 0.3095, + "step": 2082 + }, + { + "epoch": 0.045850759410081446, + "grad_norm": 2.3132560145110435, + "learning_rate": 9.993414596057837e-06, + "loss": 0.3674, + "step": 2083 + }, + { + "epoch": 0.04587277129650011, + "grad_norm": 2.038038557198818, + "learning_rate": 9.993396294596316e-06, + "loss": 0.3994, + "step": 2084 + }, + { + "epoch": 0.04589478318291878, + "grad_norm": 1.9576138061502502, + "learning_rate": 9.993377967756127e-06, + "loss": 0.4868, + "step": 2085 + }, + { + "epoch": 0.04591679506933744, + "grad_norm": 2.743000699544406, + "learning_rate": 9.993359615537365e-06, + "loss": 0.4322, + "step": 2086 + }, + { + "epoch": 0.045938806955756105, + "grad_norm": 2.3501189165483742, + "learning_rate": 9.993341237940122e-06, + "loss": 0.4009, + "step": 2087 + }, + { + "epoch": 0.045960818842174775, + "grad_norm": 2.5463581220960894, + "learning_rate": 9.993322834964491e-06, + "loss": 0.3496, + "step": 2088 + }, + { + "epoch": 0.04598283072859344, + "grad_norm": 2.2140738819279284, + "learning_rate": 9.993304406610567e-06, + "loss": 0.3654, + "step": 2089 + }, + { + "epoch": 0.04600484261501211, + "grad_norm": 6.511138716854843, + "learning_rate": 9.993285952878444e-06, + "loss": 0.2994, + "step": 2090 + }, + { + "epoch": 0.04602685450143077, + "grad_norm": 2.3530432308128244, + "learning_rate": 9.993267473768214e-06, + "loss": 0.394, + "step": 2091 + }, + { + "epoch": 0.04604886638784944, + "grad_norm": 0.7903349493261544, + "learning_rate": 9.993248969279972e-06, + "loss": 0.5915, + "step": 2092 + }, + { + "epoch": 0.046070878274268104, + "grad_norm": 0.8015009282310072, + "learning_rate": 9.99323043941381e-06, + "loss": 0.6117, + "step": 2093 + }, + { + "epoch": 0.046092890160686774, + "grad_norm": 2.736856118911189, + "learning_rate": 9.993211884169824e-06, + "loss": 0.411, + "step": 2094 + }, + { + "epoch": 0.04611490204710544, + "grad_norm": 0.6910195260571571, + "learning_rate": 9.99319330354811e-06, + "loss": 0.6043, + "step": 2095 + }, + { + "epoch": 0.0461369139335241, + "grad_norm": 8.425729718975425, + "learning_rate": 9.993174697548759e-06, + "loss": 0.3818, + "step": 2096 + }, + { + "epoch": 0.04615892581994277, + "grad_norm": 0.7371089334218313, + "learning_rate": 9.993156066171868e-06, + "loss": 0.5881, + "step": 2097 + }, + { + "epoch": 0.04618093770636143, + "grad_norm": 3.122057001672273, + "learning_rate": 9.993137409417531e-06, + "loss": 0.4069, + "step": 2098 + }, + { + "epoch": 0.0462029495927801, + "grad_norm": 3.6646319774292437, + "learning_rate": 9.993118727285841e-06, + "loss": 0.3374, + "step": 2099 + }, + { + "epoch": 0.046224961479198766, + "grad_norm": 2.289530225087562, + "learning_rate": 9.993100019776897e-06, + "loss": 0.3724, + "step": 2100 + }, + { + "epoch": 0.046246973365617436, + "grad_norm": 5.363526595252649, + "learning_rate": 9.993081286890792e-06, + "loss": 0.3313, + "step": 2101 + }, + { + "epoch": 0.0462689852520361, + "grad_norm": 1.6394079496866116, + "learning_rate": 9.99306252862762e-06, + "loss": 0.3707, + "step": 2102 + }, + { + "epoch": 0.04629099713845477, + "grad_norm": 2.7019977384589775, + "learning_rate": 9.993043744987476e-06, + "loss": 0.4058, + "step": 2103 + }, + { + "epoch": 0.04631300902487343, + "grad_norm": 1.5621628026080667, + "learning_rate": 9.993024935970458e-06, + "loss": 0.3824, + "step": 2104 + }, + { + "epoch": 0.046335020911292095, + "grad_norm": 2.3600214807583964, + "learning_rate": 9.99300610157666e-06, + "loss": 0.3413, + "step": 2105 + }, + { + "epoch": 0.046357032797710765, + "grad_norm": 2.406468158417333, + "learning_rate": 9.992987241806178e-06, + "loss": 0.3149, + "step": 2106 + }, + { + "epoch": 0.04637904468412943, + "grad_norm": 1.8667433502152917, + "learning_rate": 9.992968356659108e-06, + "loss": 0.3275, + "step": 2107 + }, + { + "epoch": 0.0464010565705481, + "grad_norm": 2.326805608718692, + "learning_rate": 9.992949446135546e-06, + "loss": 0.3572, + "step": 2108 + }, + { + "epoch": 0.04642306845696676, + "grad_norm": 2.6118806264041, + "learning_rate": 9.992930510235587e-06, + "loss": 0.4619, + "step": 2109 + }, + { + "epoch": 0.04644508034338543, + "grad_norm": 2.213680302978941, + "learning_rate": 9.99291154895933e-06, + "loss": 0.3916, + "step": 2110 + }, + { + "epoch": 0.046467092229804094, + "grad_norm": 1.4320089903221145, + "learning_rate": 9.992892562306868e-06, + "loss": 0.3561, + "step": 2111 + }, + { + "epoch": 0.04648910411622276, + "grad_norm": 1.9064282927052847, + "learning_rate": 9.9928735502783e-06, + "loss": 0.339, + "step": 2112 + }, + { + "epoch": 0.04651111600264143, + "grad_norm": 0.976423097008631, + "learning_rate": 9.992854512873723e-06, + "loss": 0.6285, + "step": 2113 + }, + { + "epoch": 0.04653312788906009, + "grad_norm": 1.7878464291598664, + "learning_rate": 9.99283545009323e-06, + "loss": 0.3922, + "step": 2114 + }, + { + "epoch": 0.04655513977547876, + "grad_norm": 2.480135221317665, + "learning_rate": 9.992816361936922e-06, + "loss": 0.4374, + "step": 2115 + }, + { + "epoch": 0.04657715166189742, + "grad_norm": 2.1611829648457044, + "learning_rate": 9.992797248404892e-06, + "loss": 0.3563, + "step": 2116 + }, + { + "epoch": 0.04659916354831609, + "grad_norm": 0.6473161773281215, + "learning_rate": 9.992778109497243e-06, + "loss": 0.5932, + "step": 2117 + }, + { + "epoch": 0.046621175434734756, + "grad_norm": 2.738354347844776, + "learning_rate": 9.992758945214068e-06, + "loss": 0.4309, + "step": 2118 + }, + { + "epoch": 0.046643187321153426, + "grad_norm": 2.4845997428573865, + "learning_rate": 9.992739755555463e-06, + "loss": 0.386, + "step": 2119 + }, + { + "epoch": 0.04666519920757209, + "grad_norm": 1.7769942918671513, + "learning_rate": 9.992720540521529e-06, + "loss": 0.3729, + "step": 2120 + }, + { + "epoch": 0.04668721109399075, + "grad_norm": 2.441942631542764, + "learning_rate": 9.992701300112362e-06, + "loss": 0.3929, + "step": 2121 + }, + { + "epoch": 0.04670922298040942, + "grad_norm": 1.8703650991623457, + "learning_rate": 9.99268203432806e-06, + "loss": 0.435, + "step": 2122 + }, + { + "epoch": 0.046731234866828085, + "grad_norm": 1.6860913874287409, + "learning_rate": 9.992662743168723e-06, + "loss": 0.3613, + "step": 2123 + }, + { + "epoch": 0.046753246753246755, + "grad_norm": 1.848726464613004, + "learning_rate": 9.992643426634445e-06, + "loss": 0.4099, + "step": 2124 + }, + { + "epoch": 0.04677525863966542, + "grad_norm": 2.1182562478454927, + "learning_rate": 9.992624084725328e-06, + "loss": 0.4708, + "step": 2125 + }, + { + "epoch": 0.04679727052608409, + "grad_norm": 2.5662547312423745, + "learning_rate": 9.992604717441468e-06, + "loss": 0.3284, + "step": 2126 + }, + { + "epoch": 0.04681928241250275, + "grad_norm": 2.4773729073209707, + "learning_rate": 9.992585324782965e-06, + "loss": 0.4074, + "step": 2127 + }, + { + "epoch": 0.04684129429892142, + "grad_norm": 2.0104475458931934, + "learning_rate": 9.992565906749914e-06, + "loss": 0.4371, + "step": 2128 + }, + { + "epoch": 0.046863306185340084, + "grad_norm": 2.2250041901454987, + "learning_rate": 9.992546463342418e-06, + "loss": 0.3432, + "step": 2129 + }, + { + "epoch": 0.04688531807175875, + "grad_norm": 1.5750409196518438, + "learning_rate": 9.992526994560575e-06, + "loss": 0.3168, + "step": 2130 + }, + { + "epoch": 0.04690732995817742, + "grad_norm": 1.9155193289857853, + "learning_rate": 9.992507500404482e-06, + "loss": 0.3652, + "step": 2131 + }, + { + "epoch": 0.04692934184459608, + "grad_norm": 1.9472004527198392, + "learning_rate": 9.992487980874238e-06, + "loss": 0.3878, + "step": 2132 + }, + { + "epoch": 0.04695135373101475, + "grad_norm": 2.14469873792108, + "learning_rate": 9.992468435969945e-06, + "loss": 0.4394, + "step": 2133 + }, + { + "epoch": 0.04697336561743341, + "grad_norm": 1.5521267907110103, + "learning_rate": 9.992448865691701e-06, + "loss": 0.3082, + "step": 2134 + }, + { + "epoch": 0.04699537750385208, + "grad_norm": 2.172435601220313, + "learning_rate": 9.992429270039605e-06, + "loss": 0.4466, + "step": 2135 + }, + { + "epoch": 0.047017389390270746, + "grad_norm": 2.0763020878395793, + "learning_rate": 9.992409649013755e-06, + "loss": 0.3628, + "step": 2136 + }, + { + "epoch": 0.04703940127668941, + "grad_norm": 1.6060258434409291, + "learning_rate": 9.992390002614254e-06, + "loss": 0.3795, + "step": 2137 + }, + { + "epoch": 0.04706141316310808, + "grad_norm": 3.535297902711836, + "learning_rate": 9.9923703308412e-06, + "loss": 0.3685, + "step": 2138 + }, + { + "epoch": 0.04708342504952674, + "grad_norm": 2.4314279107846635, + "learning_rate": 9.992350633694693e-06, + "loss": 0.4126, + "step": 2139 + }, + { + "epoch": 0.04710543693594541, + "grad_norm": 0.8360868845405253, + "learning_rate": 9.992330911174835e-06, + "loss": 0.5935, + "step": 2140 + }, + { + "epoch": 0.047127448822364075, + "grad_norm": 3.5360071765562826, + "learning_rate": 9.992311163281722e-06, + "loss": 0.369, + "step": 2141 + }, + { + "epoch": 0.047149460708782745, + "grad_norm": 12.077178444242358, + "learning_rate": 9.99229139001546e-06, + "loss": 0.3557, + "step": 2142 + }, + { + "epoch": 0.04717147259520141, + "grad_norm": 2.1285348980979633, + "learning_rate": 9.992271591376144e-06, + "loss": 0.4151, + "step": 2143 + }, + { + "epoch": 0.04719348448162008, + "grad_norm": 2.5261471051535254, + "learning_rate": 9.992251767363878e-06, + "loss": 0.3815, + "step": 2144 + }, + { + "epoch": 0.04721549636803874, + "grad_norm": 1.965053676330806, + "learning_rate": 9.992231917978764e-06, + "loss": 0.3868, + "step": 2145 + }, + { + "epoch": 0.047237508254457404, + "grad_norm": 2.134623326301483, + "learning_rate": 9.992212043220898e-06, + "loss": 0.4154, + "step": 2146 + }, + { + "epoch": 0.047259520140876074, + "grad_norm": 1.835781551059857, + "learning_rate": 9.992192143090384e-06, + "loss": 0.4458, + "step": 2147 + }, + { + "epoch": 0.04728153202729474, + "grad_norm": 1.9668875566823572, + "learning_rate": 9.992172217587323e-06, + "loss": 0.3775, + "step": 2148 + }, + { + "epoch": 0.04730354391371341, + "grad_norm": 0.6815145247791128, + "learning_rate": 9.99215226671182e-06, + "loss": 0.5805, + "step": 2149 + }, + { + "epoch": 0.04732555580013207, + "grad_norm": 3.539554576287895, + "learning_rate": 9.992132290463967e-06, + "loss": 0.4348, + "step": 2150 + }, + { + "epoch": 0.04734756768655074, + "grad_norm": 1.8229892614739076, + "learning_rate": 9.992112288843874e-06, + "loss": 0.4067, + "step": 2151 + }, + { + "epoch": 0.0473695795729694, + "grad_norm": 2.40388150357875, + "learning_rate": 9.99209226185164e-06, + "loss": 0.4492, + "step": 2152 + }, + { + "epoch": 0.04739159145938807, + "grad_norm": 2.178493869580918, + "learning_rate": 9.992072209487366e-06, + "loss": 0.4384, + "step": 2153 + }, + { + "epoch": 0.047413603345806736, + "grad_norm": 0.6493714029157889, + "learning_rate": 9.992052131751154e-06, + "loss": 0.5549, + "step": 2154 + }, + { + "epoch": 0.0474356152322254, + "grad_norm": 2.0463830102347567, + "learning_rate": 9.992032028643108e-06, + "loss": 0.4642, + "step": 2155 + }, + { + "epoch": 0.04745762711864407, + "grad_norm": 1.677477907448246, + "learning_rate": 9.992011900163328e-06, + "loss": 0.4321, + "step": 2156 + }, + { + "epoch": 0.04747963900506273, + "grad_norm": 2.080074703391062, + "learning_rate": 9.991991746311916e-06, + "loss": 0.3793, + "step": 2157 + }, + { + "epoch": 0.0475016508914814, + "grad_norm": 2.6342764721274285, + "learning_rate": 9.991971567088976e-06, + "loss": 0.4433, + "step": 2158 + }, + { + "epoch": 0.047523662777900065, + "grad_norm": 1.774077079393641, + "learning_rate": 9.99195136249461e-06, + "loss": 0.439, + "step": 2159 + }, + { + "epoch": 0.047545674664318735, + "grad_norm": 1.6916018980286935, + "learning_rate": 9.991931132528923e-06, + "loss": 0.4015, + "step": 2160 + }, + { + "epoch": 0.0475676865507374, + "grad_norm": 3.123014939869576, + "learning_rate": 9.991910877192012e-06, + "loss": 0.3666, + "step": 2161 + }, + { + "epoch": 0.04758969843715606, + "grad_norm": 2.227655562942716, + "learning_rate": 9.991890596483985e-06, + "loss": 0.385, + "step": 2162 + }, + { + "epoch": 0.04761171032357473, + "grad_norm": 1.7008754979461393, + "learning_rate": 9.991870290404944e-06, + "loss": 0.4293, + "step": 2163 + }, + { + "epoch": 0.047633722209993394, + "grad_norm": 1.783622754889394, + "learning_rate": 9.99184995895499e-06, + "loss": 0.3064, + "step": 2164 + }, + { + "epoch": 0.047655734096412064, + "grad_norm": 1.8265528823343566, + "learning_rate": 9.99182960213423e-06, + "loss": 0.4273, + "step": 2165 + }, + { + "epoch": 0.04767774598283073, + "grad_norm": 2.396387581920044, + "learning_rate": 9.991809219942765e-06, + "loss": 0.4339, + "step": 2166 + }, + { + "epoch": 0.0476997578692494, + "grad_norm": 0.7208214902907879, + "learning_rate": 9.991788812380701e-06, + "loss": 0.5985, + "step": 2167 + }, + { + "epoch": 0.04772176975566806, + "grad_norm": 0.6746797679388954, + "learning_rate": 9.991768379448137e-06, + "loss": 0.5896, + "step": 2168 + }, + { + "epoch": 0.04774378164208673, + "grad_norm": 1.5283202142250192, + "learning_rate": 9.99174792114518e-06, + "loss": 0.3068, + "step": 2169 + }, + { + "epoch": 0.04776579352850539, + "grad_norm": 1.8845973329844097, + "learning_rate": 9.991727437471934e-06, + "loss": 0.4789, + "step": 2170 + }, + { + "epoch": 0.047787805414924056, + "grad_norm": 2.2319829743425763, + "learning_rate": 9.991706928428505e-06, + "loss": 0.3825, + "step": 2171 + }, + { + "epoch": 0.047809817301342726, + "grad_norm": 2.981697444635023, + "learning_rate": 9.991686394014992e-06, + "loss": 0.3796, + "step": 2172 + }, + { + "epoch": 0.04783182918776139, + "grad_norm": 1.5763726984033564, + "learning_rate": 9.991665834231503e-06, + "loss": 0.4525, + "step": 2173 + }, + { + "epoch": 0.04785384107418006, + "grad_norm": 2.155861165528192, + "learning_rate": 9.991645249078143e-06, + "loss": 0.4159, + "step": 2174 + }, + { + "epoch": 0.04787585296059872, + "grad_norm": 1.7816026719936375, + "learning_rate": 9.991624638555015e-06, + "loss": 0.4048, + "step": 2175 + }, + { + "epoch": 0.04789786484701739, + "grad_norm": 1.6884064547823443, + "learning_rate": 9.991604002662225e-06, + "loss": 0.319, + "step": 2176 + }, + { + "epoch": 0.047919876733436055, + "grad_norm": 2.034792673419789, + "learning_rate": 9.991583341399875e-06, + "loss": 0.3296, + "step": 2177 + }, + { + "epoch": 0.04794188861985472, + "grad_norm": 0.8200690683364704, + "learning_rate": 9.991562654768075e-06, + "loss": 0.6031, + "step": 2178 + }, + { + "epoch": 0.04796390050627339, + "grad_norm": 1.7000020220903596, + "learning_rate": 9.991541942766925e-06, + "loss": 0.4259, + "step": 2179 + }, + { + "epoch": 0.04798591239269205, + "grad_norm": 2.5731783291095742, + "learning_rate": 9.991521205396533e-06, + "loss": 0.4451, + "step": 2180 + }, + { + "epoch": 0.04800792427911072, + "grad_norm": 0.69847269026698, + "learning_rate": 9.991500442657003e-06, + "loss": 0.6083, + "step": 2181 + }, + { + "epoch": 0.048029936165529384, + "grad_norm": 0.6607527394924749, + "learning_rate": 9.991479654548444e-06, + "loss": 0.5935, + "step": 2182 + }, + { + "epoch": 0.048051948051948054, + "grad_norm": 1.6643939767645435, + "learning_rate": 9.991458841070958e-06, + "loss": 0.4191, + "step": 2183 + }, + { + "epoch": 0.04807395993836672, + "grad_norm": 3.527480159787202, + "learning_rate": 9.991438002224652e-06, + "loss": 0.418, + "step": 2184 + }, + { + "epoch": 0.04809597182478539, + "grad_norm": 1.6400626741954707, + "learning_rate": 9.991417138009631e-06, + "loss": 0.3531, + "step": 2185 + }, + { + "epoch": 0.04811798371120405, + "grad_norm": 1.7259027768593294, + "learning_rate": 9.991396248426003e-06, + "loss": 0.3946, + "step": 2186 + }, + { + "epoch": 0.04813999559762271, + "grad_norm": 2.021377220309678, + "learning_rate": 9.991375333473872e-06, + "loss": 0.3786, + "step": 2187 + }, + { + "epoch": 0.04816200748404138, + "grad_norm": 3.4234355042194045, + "learning_rate": 9.991354393153346e-06, + "loss": 0.4459, + "step": 2188 + }, + { + "epoch": 0.048184019370460046, + "grad_norm": 1.7567918257342054, + "learning_rate": 9.991333427464532e-06, + "loss": 0.3921, + "step": 2189 + }, + { + "epoch": 0.048206031256878716, + "grad_norm": 0.9164395853523934, + "learning_rate": 9.991312436407532e-06, + "loss": 0.586, + "step": 2190 + }, + { + "epoch": 0.04822804314329738, + "grad_norm": 0.861378013037446, + "learning_rate": 9.991291419982458e-06, + "loss": 0.601, + "step": 2191 + }, + { + "epoch": 0.04825005502971605, + "grad_norm": 1.7607872390143484, + "learning_rate": 9.991270378189415e-06, + "loss": 0.4151, + "step": 2192 + }, + { + "epoch": 0.04827206691613471, + "grad_norm": 0.6918921442583302, + "learning_rate": 9.99124931102851e-06, + "loss": 0.615, + "step": 2193 + }, + { + "epoch": 0.04829407880255338, + "grad_norm": 2.0598959307155265, + "learning_rate": 9.991228218499848e-06, + "loss": 0.3164, + "step": 2194 + }, + { + "epoch": 0.048316090688972045, + "grad_norm": 2.092014829752797, + "learning_rate": 9.991207100603542e-06, + "loss": 0.3368, + "step": 2195 + }, + { + "epoch": 0.04833810257539071, + "grad_norm": 2.066940953051557, + "learning_rate": 9.991185957339692e-06, + "loss": 0.377, + "step": 2196 + }, + { + "epoch": 0.04836011446180938, + "grad_norm": 2.0762770173361442, + "learning_rate": 9.99116478870841e-06, + "loss": 0.3728, + "step": 2197 + }, + { + "epoch": 0.04838212634822804, + "grad_norm": 1.6019242071303863, + "learning_rate": 9.991143594709802e-06, + "loss": 0.4255, + "step": 2198 + }, + { + "epoch": 0.04840413823464671, + "grad_norm": 2.1203046759267985, + "learning_rate": 9.991122375343975e-06, + "loss": 0.4296, + "step": 2199 + }, + { + "epoch": 0.048426150121065374, + "grad_norm": 12.900973133813565, + "learning_rate": 9.99110113061104e-06, + "loss": 0.3547, + "step": 2200 + }, + { + "epoch": 0.048448162007484044, + "grad_norm": 1.1662026882636523, + "learning_rate": 9.991079860511102e-06, + "loss": 0.631, + "step": 2201 + }, + { + "epoch": 0.04847017389390271, + "grad_norm": 1.6448097424280599, + "learning_rate": 9.99105856504427e-06, + "loss": 0.3688, + "step": 2202 + }, + { + "epoch": 0.04849218578032137, + "grad_norm": 1.5869653637204522, + "learning_rate": 9.991037244210653e-06, + "loss": 0.3863, + "step": 2203 + }, + { + "epoch": 0.04851419766674004, + "grad_norm": 1.832786618240704, + "learning_rate": 9.991015898010359e-06, + "loss": 0.4699, + "step": 2204 + }, + { + "epoch": 0.0485362095531587, + "grad_norm": 0.7481261325514188, + "learning_rate": 9.990994526443494e-06, + "loss": 0.6086, + "step": 2205 + }, + { + "epoch": 0.04855822143957737, + "grad_norm": 0.6746476813615719, + "learning_rate": 9.990973129510171e-06, + "loss": 0.5996, + "step": 2206 + }, + { + "epoch": 0.048580233325996036, + "grad_norm": 2.0339239633720045, + "learning_rate": 9.990951707210495e-06, + "loss": 0.4064, + "step": 2207 + }, + { + "epoch": 0.048602245212414706, + "grad_norm": 2.4138828556283007, + "learning_rate": 9.990930259544578e-06, + "loss": 0.3699, + "step": 2208 + }, + { + "epoch": 0.04862425709883337, + "grad_norm": 0.7165810968017717, + "learning_rate": 9.990908786512526e-06, + "loss": 0.5858, + "step": 2209 + }, + { + "epoch": 0.04864626898525204, + "grad_norm": 29.012930470888943, + "learning_rate": 9.990887288114451e-06, + "loss": 0.4181, + "step": 2210 + }, + { + "epoch": 0.0486682808716707, + "grad_norm": 1.8338368574151267, + "learning_rate": 9.990865764350458e-06, + "loss": 0.4492, + "step": 2211 + }, + { + "epoch": 0.048690292758089365, + "grad_norm": 2.153964727385539, + "learning_rate": 9.990844215220662e-06, + "loss": 0.3946, + "step": 2212 + }, + { + "epoch": 0.048712304644508035, + "grad_norm": 0.8409522210071454, + "learning_rate": 9.990822640725169e-06, + "loss": 0.6048, + "step": 2213 + }, + { + "epoch": 0.0487343165309267, + "grad_norm": 2.0043182031886797, + "learning_rate": 9.990801040864088e-06, + "loss": 0.4546, + "step": 2214 + }, + { + "epoch": 0.04875632841734537, + "grad_norm": 1.7017663866351538, + "learning_rate": 9.990779415637531e-06, + "loss": 0.4193, + "step": 2215 + }, + { + "epoch": 0.04877834030376403, + "grad_norm": 2.0076619825021296, + "learning_rate": 9.990757765045608e-06, + "loss": 0.3715, + "step": 2216 + }, + { + "epoch": 0.0488003521901827, + "grad_norm": 1.762299902656607, + "learning_rate": 9.990736089088428e-06, + "loss": 0.3694, + "step": 2217 + }, + { + "epoch": 0.048822364076601364, + "grad_norm": 2.7224272732110673, + "learning_rate": 9.9907143877661e-06, + "loss": 0.4029, + "step": 2218 + }, + { + "epoch": 0.048844375963020034, + "grad_norm": 3.5075489353115366, + "learning_rate": 9.990692661078736e-06, + "loss": 0.3825, + "step": 2219 + }, + { + "epoch": 0.0488663878494387, + "grad_norm": 1.7835002826884685, + "learning_rate": 9.990670909026447e-06, + "loss": 0.3539, + "step": 2220 + }, + { + "epoch": 0.04888839973585736, + "grad_norm": 3.1811693706395383, + "learning_rate": 9.990649131609341e-06, + "loss": 0.3607, + "step": 2221 + }, + { + "epoch": 0.04891041162227603, + "grad_norm": 2.5094328355492737, + "learning_rate": 9.99062732882753e-06, + "loss": 0.508, + "step": 2222 + }, + { + "epoch": 0.04893242350869469, + "grad_norm": 2.189103146561263, + "learning_rate": 9.990605500681127e-06, + "loss": 0.3597, + "step": 2223 + }, + { + "epoch": 0.04895443539511336, + "grad_norm": 1.4200584291893823, + "learning_rate": 9.99058364717024e-06, + "loss": 0.3789, + "step": 2224 + }, + { + "epoch": 0.048976447281532026, + "grad_norm": 2.357088456972578, + "learning_rate": 9.990561768294983e-06, + "loss": 0.3783, + "step": 2225 + }, + { + "epoch": 0.048998459167950696, + "grad_norm": 1.7670196784204235, + "learning_rate": 9.990539864055462e-06, + "loss": 0.3756, + "step": 2226 + }, + { + "epoch": 0.04902047105436936, + "grad_norm": 2.0694297112138025, + "learning_rate": 9.990517934451792e-06, + "loss": 0.4441, + "step": 2227 + }, + { + "epoch": 0.04904248294078802, + "grad_norm": 0.751275432076838, + "learning_rate": 9.990495979484085e-06, + "loss": 0.6064, + "step": 2228 + }, + { + "epoch": 0.04906449482720669, + "grad_norm": 2.0980889844510995, + "learning_rate": 9.990473999152451e-06, + "loss": 0.4451, + "step": 2229 + }, + { + "epoch": 0.049086506713625355, + "grad_norm": 2.1370206729559142, + "learning_rate": 9.990451993457003e-06, + "loss": 0.4376, + "step": 2230 + }, + { + "epoch": 0.049108518600044025, + "grad_norm": 2.0575267649845888, + "learning_rate": 9.990429962397852e-06, + "loss": 0.4165, + "step": 2231 + }, + { + "epoch": 0.04913053048646269, + "grad_norm": 2.948627850594981, + "learning_rate": 9.99040790597511e-06, + "loss": 0.4669, + "step": 2232 + }, + { + "epoch": 0.04915254237288136, + "grad_norm": 1.584019610036924, + "learning_rate": 9.99038582418889e-06, + "loss": 0.3527, + "step": 2233 + }, + { + "epoch": 0.04917455425930002, + "grad_norm": 1.823635738683376, + "learning_rate": 9.990363717039301e-06, + "loss": 0.4652, + "step": 2234 + }, + { + "epoch": 0.04919656614571869, + "grad_norm": 0.7036797837821656, + "learning_rate": 9.990341584526461e-06, + "loss": 0.6074, + "step": 2235 + }, + { + "epoch": 0.049218578032137354, + "grad_norm": 1.6357115113100211, + "learning_rate": 9.990319426650477e-06, + "loss": 0.39, + "step": 2236 + }, + { + "epoch": 0.04924058991855602, + "grad_norm": 1.7330229661921621, + "learning_rate": 9.990297243411467e-06, + "loss": 0.3511, + "step": 2237 + }, + { + "epoch": 0.04926260180497469, + "grad_norm": 2.083645804318804, + "learning_rate": 9.990275034809537e-06, + "loss": 0.3706, + "step": 2238 + }, + { + "epoch": 0.04928461369139335, + "grad_norm": 0.6669045840171614, + "learning_rate": 9.990252800844805e-06, + "loss": 0.6134, + "step": 2239 + }, + { + "epoch": 0.04930662557781202, + "grad_norm": 1.779948713295624, + "learning_rate": 9.990230541517382e-06, + "loss": 0.4289, + "step": 2240 + }, + { + "epoch": 0.04932863746423068, + "grad_norm": 1.9345677574882834, + "learning_rate": 9.990208256827383e-06, + "loss": 0.3535, + "step": 2241 + }, + { + "epoch": 0.04935064935064935, + "grad_norm": 2.236609127299859, + "learning_rate": 9.99018594677492e-06, + "loss": 0.3655, + "step": 2242 + }, + { + "epoch": 0.049372661237068016, + "grad_norm": 1.8083226732860282, + "learning_rate": 9.990163611360106e-06, + "loss": 0.3512, + "step": 2243 + }, + { + "epoch": 0.049394673123486686, + "grad_norm": 0.6796451681578599, + "learning_rate": 9.990141250583054e-06, + "loss": 0.6077, + "step": 2244 + }, + { + "epoch": 0.04941668500990535, + "grad_norm": 1.6347905968443297, + "learning_rate": 9.99011886444388e-06, + "loss": 0.3629, + "step": 2245 + }, + { + "epoch": 0.04943869689632401, + "grad_norm": 2.67349649094502, + "learning_rate": 9.990096452942693e-06, + "loss": 0.4354, + "step": 2246 + }, + { + "epoch": 0.04946070878274268, + "grad_norm": 1.8325064302540643, + "learning_rate": 9.990074016079614e-06, + "loss": 0.4318, + "step": 2247 + }, + { + "epoch": 0.049482720669161345, + "grad_norm": 1.889746730900341, + "learning_rate": 9.990051553854751e-06, + "loss": 0.41, + "step": 2248 + }, + { + "epoch": 0.049504732555580015, + "grad_norm": 1.9376653393688326, + "learning_rate": 9.990029066268221e-06, + "loss": 0.4089, + "step": 2249 + }, + { + "epoch": 0.04952674444199868, + "grad_norm": 0.7063683126642671, + "learning_rate": 9.990006553320137e-06, + "loss": 0.5754, + "step": 2250 + }, + { + "epoch": 0.04954875632841735, + "grad_norm": 1.8418483018189777, + "learning_rate": 9.989984015010615e-06, + "loss": 0.3759, + "step": 2251 + }, + { + "epoch": 0.04957076821483601, + "grad_norm": 1.4456057575958865, + "learning_rate": 9.989961451339769e-06, + "loss": 0.2603, + "step": 2252 + }, + { + "epoch": 0.049592780101254674, + "grad_norm": 1.6979820346683832, + "learning_rate": 9.989938862307712e-06, + "loss": 0.2821, + "step": 2253 + }, + { + "epoch": 0.049614791987673344, + "grad_norm": 1.9328832700010132, + "learning_rate": 9.989916247914561e-06, + "loss": 0.4502, + "step": 2254 + }, + { + "epoch": 0.04963680387409201, + "grad_norm": 1.6430993912047358, + "learning_rate": 9.98989360816043e-06, + "loss": 0.3581, + "step": 2255 + }, + { + "epoch": 0.04965881576051068, + "grad_norm": 3.123579484993593, + "learning_rate": 9.989870943045435e-06, + "loss": 0.3227, + "step": 2256 + }, + { + "epoch": 0.04968082764692934, + "grad_norm": 1.931131715671466, + "learning_rate": 9.989848252569689e-06, + "loss": 0.3967, + "step": 2257 + }, + { + "epoch": 0.04970283953334801, + "grad_norm": 1.5032365982130416, + "learning_rate": 9.989825536733309e-06, + "loss": 0.4382, + "step": 2258 + }, + { + "epoch": 0.04972485141976667, + "grad_norm": 1.816537257662962, + "learning_rate": 9.98980279553641e-06, + "loss": 0.3923, + "step": 2259 + }, + { + "epoch": 0.04974686330618534, + "grad_norm": 1.7291538376509776, + "learning_rate": 9.989780028979107e-06, + "loss": 0.4223, + "step": 2260 + }, + { + "epoch": 0.049768875192604006, + "grad_norm": 1.7054630626000586, + "learning_rate": 9.989757237061518e-06, + "loss": 0.5351, + "step": 2261 + }, + { + "epoch": 0.04979088707902267, + "grad_norm": 1.5871679163520152, + "learning_rate": 9.989734419783757e-06, + "loss": 0.3829, + "step": 2262 + }, + { + "epoch": 0.04981289896544134, + "grad_norm": 1.9011904067882388, + "learning_rate": 9.989711577145938e-06, + "loss": 0.4652, + "step": 2263 + }, + { + "epoch": 0.04983491085186, + "grad_norm": 1.597387453516145, + "learning_rate": 9.989688709148181e-06, + "loss": 0.3616, + "step": 2264 + }, + { + "epoch": 0.04985692273827867, + "grad_norm": 1.4004424577994616, + "learning_rate": 9.9896658157906e-06, + "loss": 0.3987, + "step": 2265 + }, + { + "epoch": 0.049878934624697335, + "grad_norm": 2.429626104021366, + "learning_rate": 9.989642897073313e-06, + "loss": 0.3913, + "step": 2266 + }, + { + "epoch": 0.049900946511116005, + "grad_norm": 1.9505825516633835, + "learning_rate": 9.989619952996434e-06, + "loss": 0.3935, + "step": 2267 + }, + { + "epoch": 0.04992295839753467, + "grad_norm": 1.8594007715252867, + "learning_rate": 9.989596983560082e-06, + "loss": 0.3694, + "step": 2268 + }, + { + "epoch": 0.04994497028395334, + "grad_norm": 0.6897334404939744, + "learning_rate": 9.989573988764371e-06, + "loss": 0.6081, + "step": 2269 + }, + { + "epoch": 0.049966982170372, + "grad_norm": 1.5686428604350402, + "learning_rate": 9.989550968609422e-06, + "loss": 0.4085, + "step": 2270 + }, + { + "epoch": 0.049988994056790664, + "grad_norm": 0.7045613781513397, + "learning_rate": 9.989527923095348e-06, + "loss": 0.6072, + "step": 2271 + }, + { + "epoch": 0.050011005943209334, + "grad_norm": 3.5239140535642153, + "learning_rate": 9.98950485222227e-06, + "loss": 0.4441, + "step": 2272 + }, + { + "epoch": 0.050033017829628, + "grad_norm": 1.8042389140998853, + "learning_rate": 9.9894817559903e-06, + "loss": 0.4382, + "step": 2273 + }, + { + "epoch": 0.05005502971604667, + "grad_norm": 2.0270378770687905, + "learning_rate": 9.98945863439956e-06, + "loss": 0.409, + "step": 2274 + }, + { + "epoch": 0.05007704160246533, + "grad_norm": 1.8183023582278282, + "learning_rate": 9.989435487450165e-06, + "loss": 0.4332, + "step": 2275 + }, + { + "epoch": 0.050099053488884, + "grad_norm": 0.702051138670208, + "learning_rate": 9.989412315142232e-06, + "loss": 0.6034, + "step": 2276 + }, + { + "epoch": 0.05012106537530266, + "grad_norm": 1.8874449326675342, + "learning_rate": 9.989389117475883e-06, + "loss": 0.3848, + "step": 2277 + }, + { + "epoch": 0.050143077261721326, + "grad_norm": 1.478143685242001, + "learning_rate": 9.989365894451233e-06, + "loss": 0.3444, + "step": 2278 + }, + { + "epoch": 0.050165089148139996, + "grad_norm": 0.6809132899599225, + "learning_rate": 9.989342646068398e-06, + "loss": 0.582, + "step": 2279 + }, + { + "epoch": 0.05018710103455866, + "grad_norm": 0.6881879570702455, + "learning_rate": 9.989319372327503e-06, + "loss": 0.6069, + "step": 2280 + }, + { + "epoch": 0.05020911292097733, + "grad_norm": 1.6421749229261675, + "learning_rate": 9.989296073228656e-06, + "loss": 0.4152, + "step": 2281 + }, + { + "epoch": 0.05023112480739599, + "grad_norm": 1.6290621435692147, + "learning_rate": 9.989272748771985e-06, + "loss": 0.3985, + "step": 2282 + }, + { + "epoch": 0.05025313669381466, + "grad_norm": 1.4481973098922938, + "learning_rate": 9.989249398957602e-06, + "loss": 0.3529, + "step": 2283 + }, + { + "epoch": 0.050275148580233325, + "grad_norm": 1.5711061581955668, + "learning_rate": 9.989226023785629e-06, + "loss": 0.3633, + "step": 2284 + }, + { + "epoch": 0.050297160466651995, + "grad_norm": 1.296792390851866, + "learning_rate": 9.989202623256185e-06, + "loss": 0.2937, + "step": 2285 + }, + { + "epoch": 0.05031917235307066, + "grad_norm": 2.5991616979639214, + "learning_rate": 9.989179197369387e-06, + "loss": 0.4347, + "step": 2286 + }, + { + "epoch": 0.05034118423948932, + "grad_norm": 1.982189824464136, + "learning_rate": 9.989155746125356e-06, + "loss": 0.453, + "step": 2287 + }, + { + "epoch": 0.05036319612590799, + "grad_norm": 1.8542529182571375, + "learning_rate": 9.989132269524212e-06, + "loss": 0.3304, + "step": 2288 + }, + { + "epoch": 0.050385208012326654, + "grad_norm": 1.8580347142737097, + "learning_rate": 9.989108767566069e-06, + "loss": 0.444, + "step": 2289 + }, + { + "epoch": 0.050407219898745324, + "grad_norm": 1.8397913273536937, + "learning_rate": 9.989085240251051e-06, + "loss": 0.3658, + "step": 2290 + }, + { + "epoch": 0.05042923178516399, + "grad_norm": 1.4730784388767124, + "learning_rate": 9.989061687579278e-06, + "loss": 0.401, + "step": 2291 + }, + { + "epoch": 0.05045124367158266, + "grad_norm": 2.4556975949493727, + "learning_rate": 9.989038109550867e-06, + "loss": 0.2989, + "step": 2292 + }, + { + "epoch": 0.05047325555800132, + "grad_norm": 2.2562228562997895, + "learning_rate": 9.989014506165939e-06, + "loss": 0.4487, + "step": 2293 + }, + { + "epoch": 0.05049526744441998, + "grad_norm": 1.6933232904210764, + "learning_rate": 9.988990877424616e-06, + "loss": 0.3326, + "step": 2294 + }, + { + "epoch": 0.05051727933083865, + "grad_norm": 0.9258187371287752, + "learning_rate": 9.988967223327015e-06, + "loss": 0.6037, + "step": 2295 + }, + { + "epoch": 0.050539291217257316, + "grad_norm": 0.8741520221730722, + "learning_rate": 9.988943543873258e-06, + "loss": 0.595, + "step": 2296 + }, + { + "epoch": 0.050561303103675986, + "grad_norm": 1.6156695684666258, + "learning_rate": 9.988919839063462e-06, + "loss": 0.4092, + "step": 2297 + }, + { + "epoch": 0.05058331499009465, + "grad_norm": 0.6865625274642446, + "learning_rate": 9.988896108897753e-06, + "loss": 0.6119, + "step": 2298 + }, + { + "epoch": 0.05060532687651332, + "grad_norm": 1.7549863194520607, + "learning_rate": 9.988872353376249e-06, + "loss": 0.4302, + "step": 2299 + }, + { + "epoch": 0.05062733876293198, + "grad_norm": 1.960561102076394, + "learning_rate": 9.98884857249907e-06, + "loss": 0.3948, + "step": 2300 + }, + { + "epoch": 0.05064935064935065, + "grad_norm": 5.143432515905921, + "learning_rate": 9.988824766266337e-06, + "loss": 0.4289, + "step": 2301 + }, + { + "epoch": 0.050671362535769315, + "grad_norm": 1.6501003509461105, + "learning_rate": 9.98880093467817e-06, + "loss": 0.3828, + "step": 2302 + }, + { + "epoch": 0.05069337442218798, + "grad_norm": 2.1667996787012616, + "learning_rate": 9.988777077734694e-06, + "loss": 0.4028, + "step": 2303 + }, + { + "epoch": 0.05071538630860665, + "grad_norm": 1.9423777248518992, + "learning_rate": 9.988753195436029e-06, + "loss": 0.4241, + "step": 2304 + }, + { + "epoch": 0.05073739819502531, + "grad_norm": 1.2658189525724066, + "learning_rate": 9.988729287782292e-06, + "loss": 0.6248, + "step": 2305 + }, + { + "epoch": 0.05075941008144398, + "grad_norm": 1.7139934922190299, + "learning_rate": 9.98870535477361e-06, + "loss": 0.4281, + "step": 2306 + }, + { + "epoch": 0.050781421967862644, + "grad_norm": 2.1815112657394584, + "learning_rate": 9.9886813964101e-06, + "loss": 0.4275, + "step": 2307 + }, + { + "epoch": 0.050803433854281314, + "grad_norm": 0.8235323771650584, + "learning_rate": 9.98865741269189e-06, + "loss": 0.5885, + "step": 2308 + }, + { + "epoch": 0.05082544574069998, + "grad_norm": 2.066089422880247, + "learning_rate": 9.988633403619095e-06, + "loss": 0.3013, + "step": 2309 + }, + { + "epoch": 0.05084745762711865, + "grad_norm": 2.6153515002438454, + "learning_rate": 9.98860936919184e-06, + "loss": 0.3091, + "step": 2310 + }, + { + "epoch": 0.05086946951353731, + "grad_norm": 2.0183014040198466, + "learning_rate": 9.988585309410248e-06, + "loss": 0.3692, + "step": 2311 + }, + { + "epoch": 0.05089148139995597, + "grad_norm": 1.559056072072537, + "learning_rate": 9.98856122427444e-06, + "loss": 0.3423, + "step": 2312 + }, + { + "epoch": 0.05091349328637464, + "grad_norm": 2.0329078786053123, + "learning_rate": 9.98853711378454e-06, + "loss": 0.3806, + "step": 2313 + }, + { + "epoch": 0.050935505172793306, + "grad_norm": 2.2596871757560932, + "learning_rate": 9.98851297794067e-06, + "loss": 0.3847, + "step": 2314 + }, + { + "epoch": 0.050957517059211976, + "grad_norm": 1.7400476835190142, + "learning_rate": 9.988488816742949e-06, + "loss": 0.4232, + "step": 2315 + }, + { + "epoch": 0.05097952894563064, + "grad_norm": 1.8770897391043428, + "learning_rate": 9.988464630191506e-06, + "loss": 0.4123, + "step": 2316 + }, + { + "epoch": 0.05100154083204931, + "grad_norm": 2.1327006277037186, + "learning_rate": 9.988440418286459e-06, + "loss": 0.3578, + "step": 2317 + }, + { + "epoch": 0.05102355271846797, + "grad_norm": 1.9660989903540984, + "learning_rate": 9.988416181027934e-06, + "loss": 0.3551, + "step": 2318 + }, + { + "epoch": 0.051045564604886635, + "grad_norm": 3.190831130288953, + "learning_rate": 9.988391918416052e-06, + "loss": 0.361, + "step": 2319 + }, + { + "epoch": 0.051067576491305305, + "grad_norm": 5.166409510195183, + "learning_rate": 9.988367630450938e-06, + "loss": 0.3314, + "step": 2320 + }, + { + "epoch": 0.05108958837772397, + "grad_norm": 1.747246549663651, + "learning_rate": 9.988343317132715e-06, + "loss": 0.4145, + "step": 2321 + }, + { + "epoch": 0.05111160026414264, + "grad_norm": 3.294914716927106, + "learning_rate": 9.988318978461506e-06, + "loss": 0.5045, + "step": 2322 + }, + { + "epoch": 0.0511336121505613, + "grad_norm": 1.707240900453473, + "learning_rate": 9.988294614437435e-06, + "loss": 0.3722, + "step": 2323 + }, + { + "epoch": 0.05115562403697997, + "grad_norm": 1.7609518127288315, + "learning_rate": 9.988270225060627e-06, + "loss": 0.4247, + "step": 2324 + }, + { + "epoch": 0.051177635923398634, + "grad_norm": 1.2161131410123736, + "learning_rate": 9.988245810331203e-06, + "loss": 0.6404, + "step": 2325 + }, + { + "epoch": 0.051199647809817304, + "grad_norm": 2.3476113903968665, + "learning_rate": 9.988221370249292e-06, + "loss": 0.3739, + "step": 2326 + }, + { + "epoch": 0.05122165969623597, + "grad_norm": 2.203846529914715, + "learning_rate": 9.988196904815014e-06, + "loss": 0.4488, + "step": 2327 + }, + { + "epoch": 0.05124367158265463, + "grad_norm": 2.2529934196783534, + "learning_rate": 9.988172414028492e-06, + "loss": 0.3525, + "step": 2328 + }, + { + "epoch": 0.0512656834690733, + "grad_norm": 1.7501329917539845, + "learning_rate": 9.988147897889854e-06, + "loss": 0.3989, + "step": 2329 + }, + { + "epoch": 0.05128769535549196, + "grad_norm": 2.093146795501962, + "learning_rate": 9.988123356399226e-06, + "loss": 0.4171, + "step": 2330 + }, + { + "epoch": 0.05130970724191063, + "grad_norm": 2.366288500337015, + "learning_rate": 9.988098789556728e-06, + "loss": 0.4277, + "step": 2331 + }, + { + "epoch": 0.051331719128329296, + "grad_norm": 1.8052910100690718, + "learning_rate": 9.988074197362487e-06, + "loss": 0.4282, + "step": 2332 + }, + { + "epoch": 0.051353731014747966, + "grad_norm": 1.7889701531451692, + "learning_rate": 9.98804957981663e-06, + "loss": 0.3769, + "step": 2333 + }, + { + "epoch": 0.05137574290116663, + "grad_norm": 0.8220186119008288, + "learning_rate": 9.98802493691928e-06, + "loss": 0.6016, + "step": 2334 + }, + { + "epoch": 0.0513977547875853, + "grad_norm": 1.8601717042432375, + "learning_rate": 9.988000268670562e-06, + "loss": 0.3398, + "step": 2335 + }, + { + "epoch": 0.05141976667400396, + "grad_norm": 1.8258978002718689, + "learning_rate": 9.9879755750706e-06, + "loss": 0.3828, + "step": 2336 + }, + { + "epoch": 0.051441778560422625, + "grad_norm": 1.7467583646262184, + "learning_rate": 9.987950856119523e-06, + "loss": 0.3389, + "step": 2337 + }, + { + "epoch": 0.051463790446841295, + "grad_norm": 1.985048373398138, + "learning_rate": 9.987926111817457e-06, + "loss": 0.406, + "step": 2338 + }, + { + "epoch": 0.05148580233325996, + "grad_norm": 1.6356951685178893, + "learning_rate": 9.987901342164525e-06, + "loss": 0.4135, + "step": 2339 + }, + { + "epoch": 0.05150781421967863, + "grad_norm": 1.7517712665589757, + "learning_rate": 9.987876547160852e-06, + "loss": 0.3399, + "step": 2340 + }, + { + "epoch": 0.05152982610609729, + "grad_norm": 2.887753509301198, + "learning_rate": 9.987851726806567e-06, + "loss": 0.395, + "step": 2341 + }, + { + "epoch": 0.05155183799251596, + "grad_norm": 1.93552608313866, + "learning_rate": 9.987826881101794e-06, + "loss": 0.4364, + "step": 2342 + }, + { + "epoch": 0.051573849878934624, + "grad_norm": 1.7264671580608062, + "learning_rate": 9.987802010046661e-06, + "loss": 0.503, + "step": 2343 + }, + { + "epoch": 0.05159586176535329, + "grad_norm": 0.7117306014451957, + "learning_rate": 9.987777113641293e-06, + "loss": 0.5614, + "step": 2344 + }, + { + "epoch": 0.05161787365177196, + "grad_norm": 1.480887630364379, + "learning_rate": 9.987752191885817e-06, + "loss": 0.3761, + "step": 2345 + }, + { + "epoch": 0.05163988553819062, + "grad_norm": 1.9282529079818573, + "learning_rate": 9.987727244780362e-06, + "loss": 0.3579, + "step": 2346 + }, + { + "epoch": 0.05166189742460929, + "grad_norm": 1.563991931682252, + "learning_rate": 9.98770227232505e-06, + "loss": 0.3837, + "step": 2347 + }, + { + "epoch": 0.051683909311027953, + "grad_norm": 2.011745580828666, + "learning_rate": 9.987677274520012e-06, + "loss": 0.4254, + "step": 2348 + }, + { + "epoch": 0.05170592119744662, + "grad_norm": 2.8755700818943306, + "learning_rate": 9.987652251365372e-06, + "loss": 0.3203, + "step": 2349 + }, + { + "epoch": 0.051727933083865286, + "grad_norm": 2.661954935838649, + "learning_rate": 9.987627202861258e-06, + "loss": 0.3837, + "step": 2350 + }, + { + "epoch": 0.051749944970283956, + "grad_norm": 1.9972581109323961, + "learning_rate": 9.9876021290078e-06, + "loss": 0.3236, + "step": 2351 + }, + { + "epoch": 0.05177195685670262, + "grad_norm": 2.2182623540171353, + "learning_rate": 9.987577029805122e-06, + "loss": 0.3858, + "step": 2352 + }, + { + "epoch": 0.05179396874312128, + "grad_norm": 2.2085927541592865, + "learning_rate": 9.987551905253354e-06, + "loss": 0.4316, + "step": 2353 + }, + { + "epoch": 0.05181598062953995, + "grad_norm": 2.123779460759335, + "learning_rate": 9.987526755352623e-06, + "loss": 0.3865, + "step": 2354 + }, + { + "epoch": 0.051837992515958616, + "grad_norm": 1.761642677346713, + "learning_rate": 9.987501580103057e-06, + "loss": 0.3826, + "step": 2355 + }, + { + "epoch": 0.051860004402377285, + "grad_norm": 0.6931496731230338, + "learning_rate": 9.987476379504782e-06, + "loss": 0.583, + "step": 2356 + }, + { + "epoch": 0.05188201628879595, + "grad_norm": 1.3711918478959673, + "learning_rate": 9.987451153557927e-06, + "loss": 0.3574, + "step": 2357 + }, + { + "epoch": 0.05190402817521462, + "grad_norm": 1.9539563277282723, + "learning_rate": 9.987425902262621e-06, + "loss": 0.4105, + "step": 2358 + }, + { + "epoch": 0.05192604006163328, + "grad_norm": 1.6807084842753917, + "learning_rate": 9.987400625618994e-06, + "loss": 0.3641, + "step": 2359 + }, + { + "epoch": 0.05194805194805195, + "grad_norm": 0.6514836541605116, + "learning_rate": 9.987375323627173e-06, + "loss": 0.5678, + "step": 2360 + }, + { + "epoch": 0.051970063834470615, + "grad_norm": 2.1537863253100644, + "learning_rate": 9.987349996287284e-06, + "loss": 0.4324, + "step": 2361 + }, + { + "epoch": 0.05199207572088928, + "grad_norm": 1.604197572808468, + "learning_rate": 9.98732464359946e-06, + "loss": 0.3371, + "step": 2362 + }, + { + "epoch": 0.05201408760730795, + "grad_norm": 1.8151176316776125, + "learning_rate": 9.987299265563825e-06, + "loss": 0.4122, + "step": 2363 + }, + { + "epoch": 0.05203609949372661, + "grad_norm": 1.869327819839773, + "learning_rate": 9.987273862180512e-06, + "loss": 0.3522, + "step": 2364 + }, + { + "epoch": 0.05205811138014528, + "grad_norm": 1.771071117926203, + "learning_rate": 9.98724843344965e-06, + "loss": 0.469, + "step": 2365 + }, + { + "epoch": 0.052080123266563944, + "grad_norm": 1.6519494935742693, + "learning_rate": 9.987222979371364e-06, + "loss": 0.3364, + "step": 2366 + }, + { + "epoch": 0.052102135152982613, + "grad_norm": 2.782218100262278, + "learning_rate": 9.98719749994579e-06, + "loss": 0.393, + "step": 2367 + }, + { + "epoch": 0.052124147039401277, + "grad_norm": 2.463330691567033, + "learning_rate": 9.987171995173052e-06, + "loss": 0.3835, + "step": 2368 + }, + { + "epoch": 0.05214615892581994, + "grad_norm": 1.9135183020061057, + "learning_rate": 9.987146465053283e-06, + "loss": 0.4304, + "step": 2369 + }, + { + "epoch": 0.05216817081223861, + "grad_norm": 1.867955158934396, + "learning_rate": 9.98712090958661e-06, + "loss": 0.2744, + "step": 2370 + }, + { + "epoch": 0.05219018269865727, + "grad_norm": 1.7730051196380712, + "learning_rate": 9.987095328773165e-06, + "loss": 0.4057, + "step": 2371 + }, + { + "epoch": 0.05221219458507594, + "grad_norm": 2.186962821288047, + "learning_rate": 9.987069722613077e-06, + "loss": 0.4578, + "step": 2372 + }, + { + "epoch": 0.052234206471494606, + "grad_norm": 1.8259446477698869, + "learning_rate": 9.987044091106474e-06, + "loss": 0.3912, + "step": 2373 + }, + { + "epoch": 0.052256218357913276, + "grad_norm": 0.7632867929705581, + "learning_rate": 9.987018434253492e-06, + "loss": 0.5771, + "step": 2374 + }, + { + "epoch": 0.05227823024433194, + "grad_norm": 2.4796017667118444, + "learning_rate": 9.986992752054256e-06, + "loss": 0.3894, + "step": 2375 + }, + { + "epoch": 0.05230024213075061, + "grad_norm": 1.872406930633513, + "learning_rate": 9.9869670445089e-06, + "loss": 0.3791, + "step": 2376 + }, + { + "epoch": 0.05232225401716927, + "grad_norm": 1.8635015196136593, + "learning_rate": 9.986941311617553e-06, + "loss": 0.3742, + "step": 2377 + }, + { + "epoch": 0.052344265903587935, + "grad_norm": 1.7397950111870344, + "learning_rate": 9.986915553380347e-06, + "loss": 0.4242, + "step": 2378 + }, + { + "epoch": 0.052366277790006605, + "grad_norm": 0.667884392351584, + "learning_rate": 9.98688976979741e-06, + "loss": 0.6204, + "step": 2379 + }, + { + "epoch": 0.05238828967642527, + "grad_norm": 1.495720407233055, + "learning_rate": 9.986863960868876e-06, + "loss": 0.3582, + "step": 2380 + }, + { + "epoch": 0.05241030156284394, + "grad_norm": 1.6506298991183292, + "learning_rate": 9.986838126594874e-06, + "loss": 0.3958, + "step": 2381 + }, + { + "epoch": 0.0524323134492626, + "grad_norm": 1.8550217349637583, + "learning_rate": 9.986812266975537e-06, + "loss": 0.4972, + "step": 2382 + }, + { + "epoch": 0.05245432533568127, + "grad_norm": 1.5730696484194329, + "learning_rate": 9.986786382010996e-06, + "loss": 0.3504, + "step": 2383 + }, + { + "epoch": 0.052476337222099934, + "grad_norm": 0.6697729632253957, + "learning_rate": 9.986760471701384e-06, + "loss": 0.6192, + "step": 2384 + }, + { + "epoch": 0.0524983491085186, + "grad_norm": 1.721065104433048, + "learning_rate": 9.986734536046829e-06, + "loss": 0.5018, + "step": 2385 + }, + { + "epoch": 0.05252036099493727, + "grad_norm": 2.0071254096819113, + "learning_rate": 9.986708575047465e-06, + "loss": 0.4025, + "step": 2386 + }, + { + "epoch": 0.05254237288135593, + "grad_norm": 2.1017329543398224, + "learning_rate": 9.986682588703424e-06, + "loss": 0.4322, + "step": 2387 + }, + { + "epoch": 0.0525643847677746, + "grad_norm": 2.8967846774954333, + "learning_rate": 9.986656577014839e-06, + "loss": 0.4269, + "step": 2388 + }, + { + "epoch": 0.05258639665419326, + "grad_norm": 2.0131982523245675, + "learning_rate": 9.986630539981841e-06, + "loss": 0.4363, + "step": 2389 + }, + { + "epoch": 0.05260840854061193, + "grad_norm": 1.9577576993172605, + "learning_rate": 9.986604477604562e-06, + "loss": 0.4103, + "step": 2390 + }, + { + "epoch": 0.052630420427030596, + "grad_norm": 1.9209567743095464, + "learning_rate": 9.986578389883135e-06, + "loss": 0.384, + "step": 2391 + }, + { + "epoch": 0.052652432313449266, + "grad_norm": 1.5833750347574154, + "learning_rate": 9.986552276817693e-06, + "loss": 0.3291, + "step": 2392 + }, + { + "epoch": 0.05267444419986793, + "grad_norm": 0.6836252718219401, + "learning_rate": 9.986526138408368e-06, + "loss": 0.5879, + "step": 2393 + }, + { + "epoch": 0.05269645608628659, + "grad_norm": 1.653376180384074, + "learning_rate": 9.986499974655292e-06, + "loss": 0.4694, + "step": 2394 + }, + { + "epoch": 0.05271846797270526, + "grad_norm": 1.6943382635051532, + "learning_rate": 9.9864737855586e-06, + "loss": 0.3983, + "step": 2395 + }, + { + "epoch": 0.052740479859123925, + "grad_norm": 1.6874368103210415, + "learning_rate": 9.986447571118424e-06, + "loss": 0.4569, + "step": 2396 + }, + { + "epoch": 0.052762491745542595, + "grad_norm": 0.6802475190477957, + "learning_rate": 9.986421331334898e-06, + "loss": 0.6256, + "step": 2397 + }, + { + "epoch": 0.05278450363196126, + "grad_norm": 1.9199928280140177, + "learning_rate": 9.986395066208155e-06, + "loss": 0.3173, + "step": 2398 + }, + { + "epoch": 0.05280651551837993, + "grad_norm": 0.6436650407596552, + "learning_rate": 9.986368775738329e-06, + "loss": 0.5864, + "step": 2399 + }, + { + "epoch": 0.05282852740479859, + "grad_norm": 1.7724387262460488, + "learning_rate": 9.98634245992555e-06, + "loss": 0.4075, + "step": 2400 + }, + { + "epoch": 0.05285053929121726, + "grad_norm": 0.6223880819733106, + "learning_rate": 9.986316118769957e-06, + "loss": 0.5895, + "step": 2401 + }, + { + "epoch": 0.052872551177635924, + "grad_norm": 2.0566370187052496, + "learning_rate": 9.986289752271681e-06, + "loss": 0.3465, + "step": 2402 + }, + { + "epoch": 0.05289456306405459, + "grad_norm": 1.901965317765137, + "learning_rate": 9.986263360430857e-06, + "loss": 0.372, + "step": 2403 + }, + { + "epoch": 0.05291657495047326, + "grad_norm": 1.385308540864602, + "learning_rate": 9.986236943247617e-06, + "loss": 0.2958, + "step": 2404 + }, + { + "epoch": 0.05293858683689192, + "grad_norm": 2.4954319876940594, + "learning_rate": 9.986210500722097e-06, + "loss": 0.4646, + "step": 2405 + }, + { + "epoch": 0.05296059872331059, + "grad_norm": 1.70193936201346, + "learning_rate": 9.986184032854433e-06, + "loss": 0.4059, + "step": 2406 + }, + { + "epoch": 0.05298261060972925, + "grad_norm": 0.6845236511819742, + "learning_rate": 9.986157539644759e-06, + "loss": 0.6299, + "step": 2407 + }, + { + "epoch": 0.05300462249614792, + "grad_norm": 2.078748030024662, + "learning_rate": 9.986131021093206e-06, + "loss": 0.4646, + "step": 2408 + }, + { + "epoch": 0.053026634382566586, + "grad_norm": 2.5707391425032964, + "learning_rate": 9.98610447719991e-06, + "loss": 0.3986, + "step": 2409 + }, + { + "epoch": 0.05304864626898525, + "grad_norm": 1.412817764566251, + "learning_rate": 9.98607790796501e-06, + "loss": 0.2872, + "step": 2410 + }, + { + "epoch": 0.05307065815540392, + "grad_norm": 2.2052873974139886, + "learning_rate": 9.986051313388638e-06, + "loss": 0.4039, + "step": 2411 + }, + { + "epoch": 0.05309267004182258, + "grad_norm": 2.304437035033203, + "learning_rate": 9.986024693470929e-06, + "loss": 0.4013, + "step": 2412 + }, + { + "epoch": 0.05311468192824125, + "grad_norm": 1.4189454021799839, + "learning_rate": 9.985998048212018e-06, + "loss": 0.385, + "step": 2413 + }, + { + "epoch": 0.053136693814659915, + "grad_norm": 1.5296462323225444, + "learning_rate": 9.985971377612041e-06, + "loss": 0.3703, + "step": 2414 + }, + { + "epoch": 0.053158705701078585, + "grad_norm": 1.688313393789264, + "learning_rate": 9.985944681671134e-06, + "loss": 0.447, + "step": 2415 + }, + { + "epoch": 0.05318071758749725, + "grad_norm": 2.157867116473004, + "learning_rate": 9.985917960389432e-06, + "loss": 0.4624, + "step": 2416 + }, + { + "epoch": 0.05320272947391592, + "grad_norm": 2.147542542666589, + "learning_rate": 9.985891213767073e-06, + "loss": 0.4066, + "step": 2417 + }, + { + "epoch": 0.05322474136033458, + "grad_norm": 1.501137975343944, + "learning_rate": 9.98586444180419e-06, + "loss": 0.3737, + "step": 2418 + }, + { + "epoch": 0.053246753246753244, + "grad_norm": 1.6754706910622086, + "learning_rate": 9.98583764450092e-06, + "loss": 0.3312, + "step": 2419 + }, + { + "epoch": 0.053268765133171914, + "grad_norm": 1.5213720587133048, + "learning_rate": 9.985810821857398e-06, + "loss": 0.397, + "step": 2420 + }, + { + "epoch": 0.05329077701959058, + "grad_norm": 1.725212089616995, + "learning_rate": 9.985783973873763e-06, + "loss": 0.399, + "step": 2421 + }, + { + "epoch": 0.05331278890600925, + "grad_norm": 0.6667496107030929, + "learning_rate": 9.98575710055015e-06, + "loss": 0.5971, + "step": 2422 + }, + { + "epoch": 0.05333480079242791, + "grad_norm": 2.1435997467041292, + "learning_rate": 9.985730201886697e-06, + "loss": 0.3825, + "step": 2423 + }, + { + "epoch": 0.05335681267884658, + "grad_norm": 1.9307035164749393, + "learning_rate": 9.985703277883537e-06, + "loss": 0.364, + "step": 2424 + }, + { + "epoch": 0.05337882456526524, + "grad_norm": 2.2026482077237697, + "learning_rate": 9.98567632854081e-06, + "loss": 0.3857, + "step": 2425 + }, + { + "epoch": 0.05340083645168391, + "grad_norm": 2.037189710123145, + "learning_rate": 9.985649353858653e-06, + "loss": 0.5214, + "step": 2426 + }, + { + "epoch": 0.053422848338102576, + "grad_norm": 2.0600937849549124, + "learning_rate": 9.985622353837202e-06, + "loss": 0.3697, + "step": 2427 + }, + { + "epoch": 0.05344486022452124, + "grad_norm": 2.452771085538898, + "learning_rate": 9.985595328476595e-06, + "loss": 0.3409, + "step": 2428 + }, + { + "epoch": 0.05346687211093991, + "grad_norm": 2.4327940318310164, + "learning_rate": 9.985568277776968e-06, + "loss": 0.4108, + "step": 2429 + }, + { + "epoch": 0.05348888399735857, + "grad_norm": 1.902922359693479, + "learning_rate": 9.98554120173846e-06, + "loss": 0.4849, + "step": 2430 + }, + { + "epoch": 0.05351089588377724, + "grad_norm": 0.6544995849723195, + "learning_rate": 9.985514100361208e-06, + "loss": 0.6044, + "step": 2431 + }, + { + "epoch": 0.053532907770195905, + "grad_norm": 2.66253860425898, + "learning_rate": 9.985486973645348e-06, + "loss": 0.393, + "step": 2432 + }, + { + "epoch": 0.053554919656614575, + "grad_norm": 1.804601274746238, + "learning_rate": 9.98545982159102e-06, + "loss": 0.4121, + "step": 2433 + }, + { + "epoch": 0.05357693154303324, + "grad_norm": 2.2975989751798944, + "learning_rate": 9.985432644198365e-06, + "loss": 0.3691, + "step": 2434 + }, + { + "epoch": 0.0535989434294519, + "grad_norm": 1.7356942301518483, + "learning_rate": 9.985405441467515e-06, + "loss": 0.3484, + "step": 2435 + }, + { + "epoch": 0.05362095531587057, + "grad_norm": 1.5137599924549319, + "learning_rate": 9.985378213398612e-06, + "loss": 0.2906, + "step": 2436 + }, + { + "epoch": 0.053642967202289234, + "grad_norm": 1.71197646352191, + "learning_rate": 9.985350959991793e-06, + "loss": 0.386, + "step": 2437 + }, + { + "epoch": 0.053664979088707904, + "grad_norm": 1.5332065111368893, + "learning_rate": 9.985323681247197e-06, + "loss": 0.471, + "step": 2438 + }, + { + "epoch": 0.05368699097512657, + "grad_norm": 1.540945130839626, + "learning_rate": 9.985296377164962e-06, + "loss": 0.4584, + "step": 2439 + }, + { + "epoch": 0.05370900286154524, + "grad_norm": 1.903944793875449, + "learning_rate": 9.985269047745227e-06, + "loss": 0.4863, + "step": 2440 + }, + { + "epoch": 0.0537310147479639, + "grad_norm": 2.938513553936154, + "learning_rate": 9.985241692988133e-06, + "loss": 0.4877, + "step": 2441 + }, + { + "epoch": 0.05375302663438257, + "grad_norm": 2.074782543079239, + "learning_rate": 9.985214312893814e-06, + "loss": 0.3762, + "step": 2442 + }, + { + "epoch": 0.05377503852080123, + "grad_norm": 5.029278221580713, + "learning_rate": 9.985186907462415e-06, + "loss": 0.4524, + "step": 2443 + }, + { + "epoch": 0.053797050407219896, + "grad_norm": 1.6811570405947593, + "learning_rate": 9.985159476694073e-06, + "loss": 0.4264, + "step": 2444 + }, + { + "epoch": 0.053819062293638566, + "grad_norm": 1.7120376171630896, + "learning_rate": 9.985132020588926e-06, + "loss": 0.4681, + "step": 2445 + }, + { + "epoch": 0.05384107418005723, + "grad_norm": 1.7129287312062573, + "learning_rate": 9.985104539147114e-06, + "loss": 0.4477, + "step": 2446 + }, + { + "epoch": 0.0538630860664759, + "grad_norm": 3.217441491108833, + "learning_rate": 9.985077032368779e-06, + "loss": 0.3373, + "step": 2447 + }, + { + "epoch": 0.05388509795289456, + "grad_norm": 1.5836006683199928, + "learning_rate": 9.985049500254058e-06, + "loss": 0.3393, + "step": 2448 + }, + { + "epoch": 0.05390710983931323, + "grad_norm": 0.7283765869112389, + "learning_rate": 9.985021942803093e-06, + "loss": 0.5884, + "step": 2449 + }, + { + "epoch": 0.053929121725731895, + "grad_norm": 1.641110409322585, + "learning_rate": 9.984994360016022e-06, + "loss": 0.37, + "step": 2450 + }, + { + "epoch": 0.053951133612150565, + "grad_norm": 1.9540280117712534, + "learning_rate": 9.984966751892986e-06, + "loss": 0.4108, + "step": 2451 + }, + { + "epoch": 0.05397314549856923, + "grad_norm": 2.1417052681668167, + "learning_rate": 9.984939118434125e-06, + "loss": 0.3972, + "step": 2452 + }, + { + "epoch": 0.05399515738498789, + "grad_norm": 1.9620366853581979, + "learning_rate": 9.984911459639583e-06, + "loss": 0.428, + "step": 2453 + }, + { + "epoch": 0.05401716927140656, + "grad_norm": 1.8043049148108758, + "learning_rate": 9.984883775509495e-06, + "loss": 0.3989, + "step": 2454 + }, + { + "epoch": 0.054039181157825224, + "grad_norm": 2.0173815073631207, + "learning_rate": 9.984856066044007e-06, + "loss": 0.4477, + "step": 2455 + }, + { + "epoch": 0.054061193044243894, + "grad_norm": 0.6503198558016696, + "learning_rate": 9.984828331243255e-06, + "loss": 0.5633, + "step": 2456 + }, + { + "epoch": 0.05408320493066256, + "grad_norm": 1.5486416878470408, + "learning_rate": 9.984800571107382e-06, + "loss": 0.4182, + "step": 2457 + }, + { + "epoch": 0.05410521681708123, + "grad_norm": 1.5766605378290373, + "learning_rate": 9.98477278563653e-06, + "loss": 0.3714, + "step": 2458 + }, + { + "epoch": 0.05412722870349989, + "grad_norm": 2.725186380811474, + "learning_rate": 9.984744974830839e-06, + "loss": 0.4676, + "step": 2459 + }, + { + "epoch": 0.05414924058991855, + "grad_norm": 9.035645736820065, + "learning_rate": 9.984717138690452e-06, + "loss": 0.4244, + "step": 2460 + }, + { + "epoch": 0.05417125247633722, + "grad_norm": 2.9541184592264744, + "learning_rate": 9.984689277215507e-06, + "loss": 0.404, + "step": 2461 + }, + { + "epoch": 0.054193264362755886, + "grad_norm": 1.720015128052549, + "learning_rate": 9.98466139040615e-06, + "loss": 0.3807, + "step": 2462 + }, + { + "epoch": 0.054215276249174556, + "grad_norm": 2.5604389422829876, + "learning_rate": 9.98463347826252e-06, + "loss": 0.3448, + "step": 2463 + }, + { + "epoch": 0.05423728813559322, + "grad_norm": 1.5781866056329592, + "learning_rate": 9.984605540784758e-06, + "loss": 0.3626, + "step": 2464 + }, + { + "epoch": 0.05425930002201189, + "grad_norm": 2.144467169878931, + "learning_rate": 9.984577577973009e-06, + "loss": 0.3277, + "step": 2465 + }, + { + "epoch": 0.05428131190843055, + "grad_norm": 2.3115283895262646, + "learning_rate": 9.984549589827413e-06, + "loss": 0.3798, + "step": 2466 + }, + { + "epoch": 0.05430332379484922, + "grad_norm": 2.62443491996762, + "learning_rate": 9.984521576348112e-06, + "loss": 0.4329, + "step": 2467 + }, + { + "epoch": 0.054325335681267885, + "grad_norm": 2.172058077232556, + "learning_rate": 9.984493537535252e-06, + "loss": 0.5035, + "step": 2468 + }, + { + "epoch": 0.05434734756768655, + "grad_norm": 1.9522047121864157, + "learning_rate": 9.98446547338897e-06, + "loss": 0.4506, + "step": 2469 + }, + { + "epoch": 0.05436935945410522, + "grad_norm": 1.8364320824182674, + "learning_rate": 9.984437383909413e-06, + "loss": 0.3396, + "step": 2470 + }, + { + "epoch": 0.05439137134052388, + "grad_norm": 2.2401085156271727, + "learning_rate": 9.984409269096721e-06, + "loss": 0.4732, + "step": 2471 + }, + { + "epoch": 0.05441338322694255, + "grad_norm": 2.0711730507740533, + "learning_rate": 9.984381128951039e-06, + "loss": 0.4464, + "step": 2472 + }, + { + "epoch": 0.054435395113361214, + "grad_norm": 1.6534688998964655, + "learning_rate": 9.984352963472507e-06, + "loss": 0.3856, + "step": 2473 + }, + { + "epoch": 0.054457406999779884, + "grad_norm": 0.70523698870893, + "learning_rate": 9.984324772661272e-06, + "loss": 0.6223, + "step": 2474 + }, + { + "epoch": 0.05447941888619855, + "grad_norm": 2.875330657262733, + "learning_rate": 9.984296556517475e-06, + "loss": 0.421, + "step": 2475 + }, + { + "epoch": 0.05450143077261722, + "grad_norm": 0.6540977848799262, + "learning_rate": 9.98426831504126e-06, + "loss": 0.5884, + "step": 2476 + }, + { + "epoch": 0.05452344265903588, + "grad_norm": 0.6568512214245552, + "learning_rate": 9.98424004823277e-06, + "loss": 0.5991, + "step": 2477 + }, + { + "epoch": 0.05454545454545454, + "grad_norm": 2.1240832552260267, + "learning_rate": 9.984211756092148e-06, + "loss": 0.3754, + "step": 2478 + }, + { + "epoch": 0.05456746643187321, + "grad_norm": 2.059275545393163, + "learning_rate": 9.98418343861954e-06, + "loss": 0.4017, + "step": 2479 + }, + { + "epoch": 0.054589478318291876, + "grad_norm": 2.1278179079035295, + "learning_rate": 9.984155095815088e-06, + "loss": 0.4317, + "step": 2480 + }, + { + "epoch": 0.054611490204710546, + "grad_norm": 1.6511919738423826, + "learning_rate": 9.98412672767894e-06, + "loss": 0.4342, + "step": 2481 + }, + { + "epoch": 0.05463350209112921, + "grad_norm": 2.4089665030597964, + "learning_rate": 9.984098334211235e-06, + "loss": 0.3781, + "step": 2482 + }, + { + "epoch": 0.05465551397754788, + "grad_norm": 2.0498635348502092, + "learning_rate": 9.984069915412118e-06, + "loss": 0.4312, + "step": 2483 + }, + { + "epoch": 0.05467752586396654, + "grad_norm": 2.214689595780471, + "learning_rate": 9.984041471281736e-06, + "loss": 0.3817, + "step": 2484 + }, + { + "epoch": 0.054699537750385205, + "grad_norm": 2.1973267359124473, + "learning_rate": 9.984013001820232e-06, + "loss": 0.3736, + "step": 2485 + }, + { + "epoch": 0.054721549636803875, + "grad_norm": 0.8892193194235505, + "learning_rate": 9.98398450702775e-06, + "loss": 0.5733, + "step": 2486 + }, + { + "epoch": 0.05474356152322254, + "grad_norm": 1.6545110046988456, + "learning_rate": 9.983955986904436e-06, + "loss": 0.3197, + "step": 2487 + }, + { + "epoch": 0.05476557340964121, + "grad_norm": 1.547805370926858, + "learning_rate": 9.983927441450436e-06, + "loss": 0.4148, + "step": 2488 + }, + { + "epoch": 0.05478758529605987, + "grad_norm": 1.82610228988636, + "learning_rate": 9.983898870665894e-06, + "loss": 0.344, + "step": 2489 + }, + { + "epoch": 0.05480959718247854, + "grad_norm": 2.7645919265333045, + "learning_rate": 9.983870274550953e-06, + "loss": 0.4818, + "step": 2490 + }, + { + "epoch": 0.054831609068897204, + "grad_norm": 2.1433737741600254, + "learning_rate": 9.983841653105763e-06, + "loss": 0.4105, + "step": 2491 + }, + { + "epoch": 0.054853620955315874, + "grad_norm": 3.2315370180547887, + "learning_rate": 9.983813006330465e-06, + "loss": 0.4802, + "step": 2492 + }, + { + "epoch": 0.05487563284173454, + "grad_norm": 2.021608564313409, + "learning_rate": 9.983784334225207e-06, + "loss": 0.38, + "step": 2493 + }, + { + "epoch": 0.0548976447281532, + "grad_norm": 0.7045586720393119, + "learning_rate": 9.983755636790135e-06, + "loss": 0.5916, + "step": 2494 + }, + { + "epoch": 0.05491965661457187, + "grad_norm": 1.742956748262076, + "learning_rate": 9.983726914025393e-06, + "loss": 0.3562, + "step": 2495 + }, + { + "epoch": 0.05494166850099053, + "grad_norm": 2.621691470018183, + "learning_rate": 9.983698165931129e-06, + "loss": 0.4202, + "step": 2496 + }, + { + "epoch": 0.0549636803874092, + "grad_norm": 0.6742962318832232, + "learning_rate": 9.983669392507486e-06, + "loss": 0.6, + "step": 2497 + }, + { + "epoch": 0.054985692273827866, + "grad_norm": 1.9876527100603716, + "learning_rate": 9.983640593754614e-06, + "loss": 0.3786, + "step": 2498 + }, + { + "epoch": 0.055007704160246536, + "grad_norm": 1.7006557366051878, + "learning_rate": 9.983611769672658e-06, + "loss": 0.3048, + "step": 2499 + }, + { + "epoch": 0.0550297160466652, + "grad_norm": 1.5995457390941699, + "learning_rate": 9.983582920261762e-06, + "loss": 0.3119, + "step": 2500 + }, + { + "epoch": 0.05505172793308386, + "grad_norm": 0.6485436668626745, + "learning_rate": 9.98355404552208e-06, + "loss": 0.5896, + "step": 2501 + }, + { + "epoch": 0.05507373981950253, + "grad_norm": 0.6742062005637549, + "learning_rate": 9.983525145453748e-06, + "loss": 0.5888, + "step": 2502 + }, + { + "epoch": 0.055095751705921195, + "grad_norm": 0.6655717353155012, + "learning_rate": 9.983496220056921e-06, + "loss": 0.5962, + "step": 2503 + }, + { + "epoch": 0.055117763592339865, + "grad_norm": 1.3427426851902537, + "learning_rate": 9.983467269331744e-06, + "loss": 0.3029, + "step": 2504 + }, + { + "epoch": 0.05513977547875853, + "grad_norm": 1.9463952303413827, + "learning_rate": 9.983438293278362e-06, + "loss": 0.4018, + "step": 2505 + }, + { + "epoch": 0.0551617873651772, + "grad_norm": 2.3197780030633037, + "learning_rate": 9.983409291896926e-06, + "loss": 0.4701, + "step": 2506 + }, + { + "epoch": 0.05518379925159586, + "grad_norm": 2.1055361462293596, + "learning_rate": 9.98338026518758e-06, + "loss": 0.3492, + "step": 2507 + }, + { + "epoch": 0.05520581113801453, + "grad_norm": 1.5988782064291809, + "learning_rate": 9.983351213150472e-06, + "loss": 0.3648, + "step": 2508 + }, + { + "epoch": 0.055227823024433194, + "grad_norm": 2.4604499810100884, + "learning_rate": 9.983322135785751e-06, + "loss": 0.4116, + "step": 2509 + }, + { + "epoch": 0.05524983491085186, + "grad_norm": 2.9367929046727164, + "learning_rate": 9.983293033093566e-06, + "loss": 0.4708, + "step": 2510 + }, + { + "epoch": 0.05527184679727053, + "grad_norm": 2.036230790625574, + "learning_rate": 9.98326390507406e-06, + "loss": 0.3798, + "step": 2511 + }, + { + "epoch": 0.05529385868368919, + "grad_norm": 0.8112788038923134, + "learning_rate": 9.98323475172739e-06, + "loss": 0.5724, + "step": 2512 + }, + { + "epoch": 0.05531587057010786, + "grad_norm": 3.2012373400776837, + "learning_rate": 9.983205573053693e-06, + "loss": 0.4143, + "step": 2513 + }, + { + "epoch": 0.05533788245652652, + "grad_norm": 2.8281491004585795, + "learning_rate": 9.983176369053125e-06, + "loss": 0.4597, + "step": 2514 + }, + { + "epoch": 0.05535989434294519, + "grad_norm": 1.83197181987867, + "learning_rate": 9.983147139725831e-06, + "loss": 0.3988, + "step": 2515 + }, + { + "epoch": 0.055381906229363856, + "grad_norm": 1.595924322415039, + "learning_rate": 9.98311788507196e-06, + "loss": 0.4284, + "step": 2516 + }, + { + "epoch": 0.055403918115782526, + "grad_norm": 2.074419091920886, + "learning_rate": 9.983088605091664e-06, + "loss": 0.4648, + "step": 2517 + }, + { + "epoch": 0.05542593000220119, + "grad_norm": 0.6860393404284983, + "learning_rate": 9.983059299785088e-06, + "loss": 0.5808, + "step": 2518 + }, + { + "epoch": 0.05544794188861985, + "grad_norm": 1.7419799304526893, + "learning_rate": 9.983029969152381e-06, + "loss": 0.3925, + "step": 2519 + }, + { + "epoch": 0.05546995377503852, + "grad_norm": 1.610850702570643, + "learning_rate": 9.983000613193694e-06, + "loss": 0.4258, + "step": 2520 + }, + { + "epoch": 0.055491965661457185, + "grad_norm": 4.002801054818885, + "learning_rate": 9.982971231909176e-06, + "loss": 0.363, + "step": 2521 + }, + { + "epoch": 0.055513977547875855, + "grad_norm": 2.0236793193722633, + "learning_rate": 9.982941825298974e-06, + "loss": 0.369, + "step": 2522 + }, + { + "epoch": 0.05553598943429452, + "grad_norm": 1.4839855620388431, + "learning_rate": 9.98291239336324e-06, + "loss": 0.4265, + "step": 2523 + }, + { + "epoch": 0.05555800132071319, + "grad_norm": 1.6488003667408972, + "learning_rate": 9.982882936102124e-06, + "loss": 0.3833, + "step": 2524 + }, + { + "epoch": 0.05558001320713185, + "grad_norm": 1.8749118528849966, + "learning_rate": 9.982853453515772e-06, + "loss": 0.372, + "step": 2525 + }, + { + "epoch": 0.055602025093550514, + "grad_norm": 1.64876710969349, + "learning_rate": 9.982823945604338e-06, + "loss": 0.4059, + "step": 2526 + }, + { + "epoch": 0.055624036979969184, + "grad_norm": 1.9518342694750683, + "learning_rate": 9.98279441236797e-06, + "loss": 0.3425, + "step": 2527 + }, + { + "epoch": 0.05564604886638785, + "grad_norm": 2.0027188108172966, + "learning_rate": 9.98276485380682e-06, + "loss": 0.3374, + "step": 2528 + }, + { + "epoch": 0.05566806075280652, + "grad_norm": 1.8230442929364805, + "learning_rate": 9.982735269921036e-06, + "loss": 0.4798, + "step": 2529 + }, + { + "epoch": 0.05569007263922518, + "grad_norm": 1.6548514188481038, + "learning_rate": 9.982705660710768e-06, + "loss": 0.3637, + "step": 2530 + }, + { + "epoch": 0.05571208452564385, + "grad_norm": 1.930437834905811, + "learning_rate": 9.982676026176169e-06, + "loss": 0.372, + "step": 2531 + }, + { + "epoch": 0.05573409641206251, + "grad_norm": 1.5584243247341134, + "learning_rate": 9.982646366317386e-06, + "loss": 0.4637, + "step": 2532 + }, + { + "epoch": 0.05575610829848118, + "grad_norm": 4.031625214339225, + "learning_rate": 9.982616681134572e-06, + "loss": 0.5243, + "step": 2533 + }, + { + "epoch": 0.055778120184899846, + "grad_norm": 1.8141715732612698, + "learning_rate": 9.982586970627878e-06, + "loss": 0.4384, + "step": 2534 + }, + { + "epoch": 0.05580013207131851, + "grad_norm": 1.8798213646191395, + "learning_rate": 9.982557234797456e-06, + "loss": 0.3867, + "step": 2535 + }, + { + "epoch": 0.05582214395773718, + "grad_norm": 1.7468255092279563, + "learning_rate": 9.982527473643455e-06, + "loss": 0.3616, + "step": 2536 + }, + { + "epoch": 0.05584415584415584, + "grad_norm": 1.5576405941231248, + "learning_rate": 9.982497687166027e-06, + "loss": 0.3587, + "step": 2537 + }, + { + "epoch": 0.05586616773057451, + "grad_norm": 1.8216467184722778, + "learning_rate": 9.982467875365323e-06, + "loss": 0.3927, + "step": 2538 + }, + { + "epoch": 0.055888179616993175, + "grad_norm": 1.6054750320103397, + "learning_rate": 9.982438038241496e-06, + "loss": 0.351, + "step": 2539 + }, + { + "epoch": 0.055910191503411845, + "grad_norm": 1.858922754443318, + "learning_rate": 9.982408175794697e-06, + "loss": 0.3818, + "step": 2540 + }, + { + "epoch": 0.05593220338983051, + "grad_norm": 1.5481137677887307, + "learning_rate": 9.982378288025076e-06, + "loss": 0.4121, + "step": 2541 + }, + { + "epoch": 0.05595421527624918, + "grad_norm": 1.871423832432397, + "learning_rate": 9.982348374932787e-06, + "loss": 0.4609, + "step": 2542 + }, + { + "epoch": 0.05597622716266784, + "grad_norm": 0.8584100226209075, + "learning_rate": 9.982318436517982e-06, + "loss": 0.6117, + "step": 2543 + }, + { + "epoch": 0.055998239049086504, + "grad_norm": 0.7378611137677403, + "learning_rate": 9.982288472780813e-06, + "loss": 0.5892, + "step": 2544 + }, + { + "epoch": 0.056020250935505174, + "grad_norm": 1.561792340473195, + "learning_rate": 9.98225848372143e-06, + "loss": 0.3054, + "step": 2545 + }, + { + "epoch": 0.05604226282192384, + "grad_norm": 2.1575389938999856, + "learning_rate": 9.982228469339988e-06, + "loss": 0.415, + "step": 2546 + }, + { + "epoch": 0.05606427470834251, + "grad_norm": 0.7259803106291879, + "learning_rate": 9.982198429636638e-06, + "loss": 0.5696, + "step": 2547 + }, + { + "epoch": 0.05608628659476117, + "grad_norm": 1.700082406270839, + "learning_rate": 9.982168364611536e-06, + "loss": 0.3535, + "step": 2548 + }, + { + "epoch": 0.05610829848117984, + "grad_norm": 1.4518776359622427, + "learning_rate": 9.98213827426483e-06, + "loss": 0.3515, + "step": 2549 + }, + { + "epoch": 0.0561303103675985, + "grad_norm": 4.304684249645128, + "learning_rate": 9.982108158596674e-06, + "loss": 0.371, + "step": 2550 + }, + { + "epoch": 0.056152322254017166, + "grad_norm": 3.609232573224378, + "learning_rate": 9.982078017607225e-06, + "loss": 0.3248, + "step": 2551 + }, + { + "epoch": 0.056174334140435836, + "grad_norm": 2.285851921985717, + "learning_rate": 9.982047851296632e-06, + "loss": 0.4182, + "step": 2552 + }, + { + "epoch": 0.0561963460268545, + "grad_norm": 2.9356248397590097, + "learning_rate": 9.982017659665051e-06, + "loss": 0.3553, + "step": 2553 + }, + { + "epoch": 0.05621835791327317, + "grad_norm": 0.8847245223243596, + "learning_rate": 9.981987442712634e-06, + "loss": 0.5899, + "step": 2554 + }, + { + "epoch": 0.05624036979969183, + "grad_norm": 1.8144533580796582, + "learning_rate": 9.981957200439533e-06, + "loss": 0.3679, + "step": 2555 + }, + { + "epoch": 0.0562623816861105, + "grad_norm": 0.6630827313994989, + "learning_rate": 9.981926932845904e-06, + "loss": 0.5803, + "step": 2556 + }, + { + "epoch": 0.056284393572529165, + "grad_norm": 2.3381235063318844, + "learning_rate": 9.981896639931903e-06, + "loss": 0.4923, + "step": 2557 + }, + { + "epoch": 0.056306405458947835, + "grad_norm": 1.8034641029640623, + "learning_rate": 9.98186632169768e-06, + "loss": 0.334, + "step": 2558 + }, + { + "epoch": 0.0563284173453665, + "grad_norm": 2.774284699167372, + "learning_rate": 9.981835978143389e-06, + "loss": 0.4565, + "step": 2559 + }, + { + "epoch": 0.05635042923178516, + "grad_norm": 2.1401590041732628, + "learning_rate": 9.981805609269186e-06, + "loss": 0.4763, + "step": 2560 + }, + { + "epoch": 0.05637244111820383, + "grad_norm": 1.5468982529620505, + "learning_rate": 9.981775215075224e-06, + "loss": 0.3729, + "step": 2561 + }, + { + "epoch": 0.056394453004622494, + "grad_norm": 1.9990637649227705, + "learning_rate": 9.98174479556166e-06, + "loss": 0.395, + "step": 2562 + }, + { + "epoch": 0.056416464891041164, + "grad_norm": 1.7712523300540226, + "learning_rate": 9.981714350728647e-06, + "loss": 0.3526, + "step": 2563 + }, + { + "epoch": 0.05643847677745983, + "grad_norm": 1.9563742504965504, + "learning_rate": 9.98168388057634e-06, + "loss": 0.4455, + "step": 2564 + }, + { + "epoch": 0.0564604886638785, + "grad_norm": 2.6303350736492437, + "learning_rate": 9.981653385104893e-06, + "loss": 0.3754, + "step": 2565 + }, + { + "epoch": 0.05648250055029716, + "grad_norm": 1.8367370234915505, + "learning_rate": 9.981622864314462e-06, + "loss": 0.5174, + "step": 2566 + }, + { + "epoch": 0.05650451243671583, + "grad_norm": 1.4074963639700695, + "learning_rate": 9.9815923182052e-06, + "loss": 0.3963, + "step": 2567 + }, + { + "epoch": 0.05652652432313449, + "grad_norm": 1.9135282754397114, + "learning_rate": 9.981561746777266e-06, + "loss": 0.3663, + "step": 2568 + }, + { + "epoch": 0.056548536209553156, + "grad_norm": 1.7138580968300936, + "learning_rate": 9.981531150030811e-06, + "loss": 0.4415, + "step": 2569 + }, + { + "epoch": 0.056570548095971826, + "grad_norm": 1.457470703730696, + "learning_rate": 9.981500527965998e-06, + "loss": 0.2826, + "step": 2570 + }, + { + "epoch": 0.05659255998239049, + "grad_norm": 1.8775464994773776, + "learning_rate": 9.981469880582973e-06, + "loss": 0.3822, + "step": 2571 + }, + { + "epoch": 0.05661457186880916, + "grad_norm": 1.4922748608403387, + "learning_rate": 9.981439207881898e-06, + "loss": 0.4061, + "step": 2572 + }, + { + "epoch": 0.05663658375522782, + "grad_norm": 2.8961635915566033, + "learning_rate": 9.981408509862927e-06, + "loss": 0.3887, + "step": 2573 + }, + { + "epoch": 0.05665859564164649, + "grad_norm": 1.976934445903846, + "learning_rate": 9.981377786526214e-06, + "loss": 0.3933, + "step": 2574 + }, + { + "epoch": 0.056680607528065155, + "grad_norm": 1.2867265387121094, + "learning_rate": 9.98134703787192e-06, + "loss": 0.6303, + "step": 2575 + }, + { + "epoch": 0.05670261941448382, + "grad_norm": 1.7458761536522676, + "learning_rate": 9.981316263900197e-06, + "loss": 0.3461, + "step": 2576 + }, + { + "epoch": 0.05672463130090249, + "grad_norm": 3.665144812796755, + "learning_rate": 9.981285464611203e-06, + "loss": 0.4487, + "step": 2577 + }, + { + "epoch": 0.05674664318732115, + "grad_norm": 2.571046465613561, + "learning_rate": 9.981254640005096e-06, + "loss": 0.4566, + "step": 2578 + }, + { + "epoch": 0.05676865507373982, + "grad_norm": 1.519443061071245, + "learning_rate": 9.981223790082031e-06, + "loss": 0.479, + "step": 2579 + }, + { + "epoch": 0.056790666960158484, + "grad_norm": 1.5179534386936415, + "learning_rate": 9.981192914842163e-06, + "loss": 0.3307, + "step": 2580 + }, + { + "epoch": 0.056812678846577154, + "grad_norm": 1.6672496842226, + "learning_rate": 9.981162014285654e-06, + "loss": 0.4171, + "step": 2581 + }, + { + "epoch": 0.05683469073299582, + "grad_norm": 1.768292939245663, + "learning_rate": 9.981131088412657e-06, + "loss": 0.3356, + "step": 2582 + }, + { + "epoch": 0.05685670261941449, + "grad_norm": 1.5295369257881613, + "learning_rate": 9.981100137223329e-06, + "loss": 0.3495, + "step": 2583 + }, + { + "epoch": 0.05687871450583315, + "grad_norm": 2.097698326280131, + "learning_rate": 9.981069160717829e-06, + "loss": 0.3592, + "step": 2584 + }, + { + "epoch": 0.05690072639225181, + "grad_norm": 2.0524986358534254, + "learning_rate": 9.981038158896315e-06, + "loss": 0.4248, + "step": 2585 + }, + { + "epoch": 0.05692273827867048, + "grad_norm": 1.4092032494038844, + "learning_rate": 9.981007131758942e-06, + "loss": 0.6199, + "step": 2586 + }, + { + "epoch": 0.056944750165089146, + "grad_norm": 1.7400080726985343, + "learning_rate": 9.980976079305868e-06, + "loss": 0.4125, + "step": 2587 + }, + { + "epoch": 0.056966762051507816, + "grad_norm": 1.7849741701971358, + "learning_rate": 9.980945001537254e-06, + "loss": 0.3695, + "step": 2588 + }, + { + "epoch": 0.05698877393792648, + "grad_norm": 1.3061286092431503, + "learning_rate": 9.980913898453256e-06, + "loss": 0.3984, + "step": 2589 + }, + { + "epoch": 0.05701078582434515, + "grad_norm": 1.9140813768825569, + "learning_rate": 9.980882770054033e-06, + "loss": 0.4191, + "step": 2590 + }, + { + "epoch": 0.05703279771076381, + "grad_norm": 2.1648870240266853, + "learning_rate": 9.98085161633974e-06, + "loss": 0.4, + "step": 2591 + }, + { + "epoch": 0.057054809597182475, + "grad_norm": 0.7714617584024349, + "learning_rate": 9.980820437310538e-06, + "loss": 0.606, + "step": 2592 + }, + { + "epoch": 0.057076821483601145, + "grad_norm": 1.693177047437145, + "learning_rate": 9.980789232966586e-06, + "loss": 0.3345, + "step": 2593 + }, + { + "epoch": 0.05709883337001981, + "grad_norm": 2.235178371430091, + "learning_rate": 9.98075800330804e-06, + "loss": 0.4344, + "step": 2594 + }, + { + "epoch": 0.05712084525643848, + "grad_norm": 1.6625176624399698, + "learning_rate": 9.980726748335061e-06, + "loss": 0.3735, + "step": 2595 + }, + { + "epoch": 0.05714285714285714, + "grad_norm": 1.8818585730403363, + "learning_rate": 9.980695468047807e-06, + "loss": 0.3677, + "step": 2596 + }, + { + "epoch": 0.05716486902927581, + "grad_norm": 1.6383561943229352, + "learning_rate": 9.980664162446439e-06, + "loss": 0.35, + "step": 2597 + }, + { + "epoch": 0.057186880915694474, + "grad_norm": 1.5374249149343624, + "learning_rate": 9.980632831531113e-06, + "loss": 0.4151, + "step": 2598 + }, + { + "epoch": 0.057208892802113144, + "grad_norm": 1.827570082662261, + "learning_rate": 9.98060147530199e-06, + "loss": 0.4084, + "step": 2599 + }, + { + "epoch": 0.05723090468853181, + "grad_norm": 2.0375402246704457, + "learning_rate": 9.980570093759227e-06, + "loss": 0.4767, + "step": 2600 + }, + { + "epoch": 0.05725291657495047, + "grad_norm": 2.0576784535420267, + "learning_rate": 9.980538686902987e-06, + "loss": 0.4152, + "step": 2601 + }, + { + "epoch": 0.05727492846136914, + "grad_norm": 2.078625037750025, + "learning_rate": 9.980507254733429e-06, + "loss": 0.4537, + "step": 2602 + }, + { + "epoch": 0.0572969403477878, + "grad_norm": 1.860072890928157, + "learning_rate": 9.98047579725071e-06, + "loss": 0.3453, + "step": 2603 + }, + { + "epoch": 0.05731895223420647, + "grad_norm": 2.574866987926942, + "learning_rate": 9.980444314454992e-06, + "loss": 0.4046, + "step": 2604 + }, + { + "epoch": 0.057340964120625136, + "grad_norm": 2.450723146058566, + "learning_rate": 9.980412806346434e-06, + "loss": 0.3601, + "step": 2605 + }, + { + "epoch": 0.057362976007043806, + "grad_norm": 0.7886662493385947, + "learning_rate": 9.9803812729252e-06, + "loss": 0.5804, + "step": 2606 + }, + { + "epoch": 0.05738498789346247, + "grad_norm": 1.7339378109243198, + "learning_rate": 9.980349714191443e-06, + "loss": 0.4235, + "step": 2607 + }, + { + "epoch": 0.05740699977988114, + "grad_norm": 2.3350155742808716, + "learning_rate": 9.98031813014533e-06, + "loss": 0.4106, + "step": 2608 + }, + { + "epoch": 0.0574290116662998, + "grad_norm": 1.8066057546238126, + "learning_rate": 9.980286520787017e-06, + "loss": 0.4575, + "step": 2609 + }, + { + "epoch": 0.057451023552718465, + "grad_norm": 1.4748119145357308, + "learning_rate": 9.980254886116668e-06, + "loss": 0.4363, + "step": 2610 + }, + { + "epoch": 0.057473035439137135, + "grad_norm": 3.021069654265591, + "learning_rate": 9.980223226134442e-06, + "loss": 0.4149, + "step": 2611 + }, + { + "epoch": 0.0574950473255558, + "grad_norm": 2.580060817648554, + "learning_rate": 9.9801915408405e-06, + "loss": 0.4141, + "step": 2612 + }, + { + "epoch": 0.05751705921197447, + "grad_norm": 0.6909485535613238, + "learning_rate": 9.980159830235006e-06, + "loss": 0.5747, + "step": 2613 + }, + { + "epoch": 0.05753907109839313, + "grad_norm": 0.663604122998996, + "learning_rate": 9.980128094318116e-06, + "loss": 0.5631, + "step": 2614 + }, + { + "epoch": 0.0575610829848118, + "grad_norm": 4.425158956673436, + "learning_rate": 9.980096333089994e-06, + "loss": 0.3676, + "step": 2615 + }, + { + "epoch": 0.057583094871230464, + "grad_norm": 1.5915201426807064, + "learning_rate": 9.9800645465508e-06, + "loss": 0.4035, + "step": 2616 + }, + { + "epoch": 0.05760510675764913, + "grad_norm": 1.5759613522475884, + "learning_rate": 9.9800327347007e-06, + "loss": 0.458, + "step": 2617 + }, + { + "epoch": 0.0576271186440678, + "grad_norm": 1.7707085861858203, + "learning_rate": 9.98000089753985e-06, + "loss": 0.3588, + "step": 2618 + }, + { + "epoch": 0.05764913053048646, + "grad_norm": 0.6885730488478382, + "learning_rate": 9.979969035068415e-06, + "loss": 0.5804, + "step": 2619 + }, + { + "epoch": 0.05767114241690513, + "grad_norm": 1.730540687957025, + "learning_rate": 9.979937147286556e-06, + "loss": 0.3572, + "step": 2620 + }, + { + "epoch": 0.05769315430332379, + "grad_norm": 1.7791512848355053, + "learning_rate": 9.979905234194436e-06, + "loss": 0.3657, + "step": 2621 + }, + { + "epoch": 0.05771516618974246, + "grad_norm": 1.5205532258887413, + "learning_rate": 9.979873295792217e-06, + "loss": 0.375, + "step": 2622 + }, + { + "epoch": 0.057737178076161126, + "grad_norm": 1.837226928266029, + "learning_rate": 9.97984133208006e-06, + "loss": 0.3605, + "step": 2623 + }, + { + "epoch": 0.057759189962579796, + "grad_norm": 2.140510314310169, + "learning_rate": 9.979809343058127e-06, + "loss": 0.4232, + "step": 2624 + }, + { + "epoch": 0.05778120184899846, + "grad_norm": 1.9173515654746565, + "learning_rate": 9.979777328726583e-06, + "loss": 0.4072, + "step": 2625 + }, + { + "epoch": 0.05780321373541712, + "grad_norm": 1.7086104417095305, + "learning_rate": 9.979745289085589e-06, + "loss": 0.3402, + "step": 2626 + }, + { + "epoch": 0.05782522562183579, + "grad_norm": 1.9657147893180162, + "learning_rate": 9.97971322413531e-06, + "loss": 0.3478, + "step": 2627 + }, + { + "epoch": 0.057847237508254455, + "grad_norm": 1.57990219787966, + "learning_rate": 9.979681133875907e-06, + "loss": 0.3489, + "step": 2628 + }, + { + "epoch": 0.057869249394673125, + "grad_norm": 5.678636110239321, + "learning_rate": 9.979649018307542e-06, + "loss": 0.3422, + "step": 2629 + }, + { + "epoch": 0.05789126128109179, + "grad_norm": 1.5836304489927544, + "learning_rate": 9.979616877430382e-06, + "loss": 0.4003, + "step": 2630 + }, + { + "epoch": 0.05791327316751046, + "grad_norm": 1.7870207136642813, + "learning_rate": 9.979584711244585e-06, + "loss": 0.4057, + "step": 2631 + }, + { + "epoch": 0.05793528505392912, + "grad_norm": 1.7660302416629494, + "learning_rate": 9.97955251975032e-06, + "loss": 0.3747, + "step": 2632 + }, + { + "epoch": 0.05795729694034779, + "grad_norm": 1.5782681227068576, + "learning_rate": 9.979520302947747e-06, + "loss": 0.3776, + "step": 2633 + }, + { + "epoch": 0.057979308826766454, + "grad_norm": 2.025143571487994, + "learning_rate": 9.979488060837032e-06, + "loss": 0.4602, + "step": 2634 + }, + { + "epoch": 0.05800132071318512, + "grad_norm": 1.47038687907434, + "learning_rate": 9.979455793418339e-06, + "loss": 0.37, + "step": 2635 + }, + { + "epoch": 0.05802333259960379, + "grad_norm": 0.7425593394311467, + "learning_rate": 9.979423500691827e-06, + "loss": 0.5985, + "step": 2636 + }, + { + "epoch": 0.05804534448602245, + "grad_norm": 61.408165394371785, + "learning_rate": 9.979391182657664e-06, + "loss": 0.4348, + "step": 2637 + }, + { + "epoch": 0.05806735637244112, + "grad_norm": 2.3789551423838833, + "learning_rate": 9.979358839316018e-06, + "loss": 0.3884, + "step": 2638 + }, + { + "epoch": 0.05808936825885978, + "grad_norm": 2.5676908825882605, + "learning_rate": 9.979326470667046e-06, + "loss": 0.4376, + "step": 2639 + }, + { + "epoch": 0.05811138014527845, + "grad_norm": 2.156987375071832, + "learning_rate": 9.979294076710919e-06, + "loss": 0.3634, + "step": 2640 + }, + { + "epoch": 0.058133392031697116, + "grad_norm": 1.822347108105653, + "learning_rate": 9.979261657447796e-06, + "loss": 0.3741, + "step": 2641 + }, + { + "epoch": 0.05815540391811578, + "grad_norm": 4.0137674938110015, + "learning_rate": 9.979229212877845e-06, + "loss": 0.426, + "step": 2642 + }, + { + "epoch": 0.05817741580453445, + "grad_norm": 2.5917678666777157, + "learning_rate": 9.979196743001231e-06, + "loss": 0.4975, + "step": 2643 + }, + { + "epoch": 0.05819942769095311, + "grad_norm": 1.7360134491120556, + "learning_rate": 9.97916424781812e-06, + "loss": 0.4698, + "step": 2644 + }, + { + "epoch": 0.05822143957737178, + "grad_norm": 1.816426087799993, + "learning_rate": 9.979131727328673e-06, + "loss": 0.3922, + "step": 2645 + }, + { + "epoch": 0.058243451463790445, + "grad_norm": 1.48175530663867, + "learning_rate": 9.979099181533058e-06, + "loss": 0.399, + "step": 2646 + }, + { + "epoch": 0.058265463350209115, + "grad_norm": 1.7110081066909448, + "learning_rate": 9.97906661043144e-06, + "loss": 0.3274, + "step": 2647 + }, + { + "epoch": 0.05828747523662778, + "grad_norm": 1.7611982819574021, + "learning_rate": 9.979034014023986e-06, + "loss": 0.3803, + "step": 2648 + }, + { + "epoch": 0.05830948712304645, + "grad_norm": 1.3891967492487343, + "learning_rate": 9.979001392310862e-06, + "loss": 0.3587, + "step": 2649 + }, + { + "epoch": 0.05833149900946511, + "grad_norm": 2.6463128679302383, + "learning_rate": 9.978968745292228e-06, + "loss": 0.3496, + "step": 2650 + }, + { + "epoch": 0.058353510895883774, + "grad_norm": 2.7064393350980533, + "learning_rate": 9.978936072968257e-06, + "loss": 0.4577, + "step": 2651 + }, + { + "epoch": 0.058375522782302444, + "grad_norm": 4.22806451372453, + "learning_rate": 9.97890337533911e-06, + "loss": 0.4451, + "step": 2652 + }, + { + "epoch": 0.05839753466872111, + "grad_norm": 2.5824738918978243, + "learning_rate": 9.978870652404959e-06, + "loss": 0.3473, + "step": 2653 + }, + { + "epoch": 0.05841954655513978, + "grad_norm": 1.6415823406572747, + "learning_rate": 9.978837904165963e-06, + "loss": 0.3856, + "step": 2654 + }, + { + "epoch": 0.05844155844155844, + "grad_norm": 2.6407327613595273, + "learning_rate": 9.978805130622295e-06, + "loss": 0.4765, + "step": 2655 + }, + { + "epoch": 0.05846357032797711, + "grad_norm": 2.4073452290142434, + "learning_rate": 9.978772331774117e-06, + "loss": 0.3844, + "step": 2656 + }, + { + "epoch": 0.05848558221439577, + "grad_norm": 1.7141997155704813, + "learning_rate": 9.978739507621597e-06, + "loss": 0.4863, + "step": 2657 + }, + { + "epoch": 0.05850759410081444, + "grad_norm": 1.9030074272591972, + "learning_rate": 9.978706658164905e-06, + "loss": 0.3851, + "step": 2658 + }, + { + "epoch": 0.058529605987233106, + "grad_norm": 2.210020706096756, + "learning_rate": 9.978673783404203e-06, + "loss": 0.3304, + "step": 2659 + }, + { + "epoch": 0.05855161787365177, + "grad_norm": 1.7938392738150144, + "learning_rate": 9.97864088333966e-06, + "loss": 0.3196, + "step": 2660 + }, + { + "epoch": 0.05857362976007044, + "grad_norm": 5.321457818373994, + "learning_rate": 9.978607957971445e-06, + "loss": 0.3412, + "step": 2661 + }, + { + "epoch": 0.0585956416464891, + "grad_norm": 3.56642855896372, + "learning_rate": 9.978575007299723e-06, + "loss": 0.4322, + "step": 2662 + }, + { + "epoch": 0.05861765353290777, + "grad_norm": 2.0624743749800696, + "learning_rate": 9.978542031324662e-06, + "loss": 0.3972, + "step": 2663 + }, + { + "epoch": 0.058639665419326435, + "grad_norm": 2.3682223512690435, + "learning_rate": 9.978509030046431e-06, + "loss": 0.4009, + "step": 2664 + }, + { + "epoch": 0.058661677305745105, + "grad_norm": 2.5967921082173717, + "learning_rate": 9.978476003465194e-06, + "loss": 0.4238, + "step": 2665 + }, + { + "epoch": 0.05868368919216377, + "grad_norm": 2.4821972885930386, + "learning_rate": 9.978442951581124e-06, + "loss": 0.3625, + "step": 2666 + }, + { + "epoch": 0.05870570107858243, + "grad_norm": 3.2962008619925394, + "learning_rate": 9.978409874394385e-06, + "loss": 0.3209, + "step": 2667 + }, + { + "epoch": 0.0587277129650011, + "grad_norm": 2.1304400986023886, + "learning_rate": 9.978376771905147e-06, + "loss": 0.3841, + "step": 2668 + }, + { + "epoch": 0.058749724851419764, + "grad_norm": 1.7092107890264343, + "learning_rate": 9.978343644113579e-06, + "loss": 0.4213, + "step": 2669 + }, + { + "epoch": 0.058771736737838434, + "grad_norm": 2.01450086796852, + "learning_rate": 9.978310491019848e-06, + "loss": 0.3845, + "step": 2670 + }, + { + "epoch": 0.0587937486242571, + "grad_norm": 2.0082193488573052, + "learning_rate": 9.978277312624122e-06, + "loss": 0.4187, + "step": 2671 + }, + { + "epoch": 0.05881576051067577, + "grad_norm": 1.5631481011861466, + "learning_rate": 9.978244108926571e-06, + "loss": 0.3549, + "step": 2672 + }, + { + "epoch": 0.05883777239709443, + "grad_norm": 1.5537798759507242, + "learning_rate": 9.978210879927361e-06, + "loss": 0.4023, + "step": 2673 + }, + { + "epoch": 0.0588597842835131, + "grad_norm": 2.478332762154289, + "learning_rate": 9.978177625626666e-06, + "loss": 0.3768, + "step": 2674 + }, + { + "epoch": 0.05888179616993176, + "grad_norm": 2.624206731142201, + "learning_rate": 9.97814434602465e-06, + "loss": 0.4464, + "step": 2675 + }, + { + "epoch": 0.058903808056350426, + "grad_norm": 1.7383853747506015, + "learning_rate": 9.978111041121483e-06, + "loss": 0.4145, + "step": 2676 + }, + { + "epoch": 0.058925819942769096, + "grad_norm": 0.8887979536658991, + "learning_rate": 9.978077710917338e-06, + "loss": 0.5899, + "step": 2677 + }, + { + "epoch": 0.05894783182918776, + "grad_norm": 1.9715509971247698, + "learning_rate": 9.97804435541238e-06, + "loss": 0.3694, + "step": 2678 + }, + { + "epoch": 0.05896984371560643, + "grad_norm": 0.681338654916998, + "learning_rate": 9.978010974606782e-06, + "loss": 0.5776, + "step": 2679 + }, + { + "epoch": 0.05899185560202509, + "grad_norm": 2.1138427829547304, + "learning_rate": 9.97797756850071e-06, + "loss": 0.3536, + "step": 2680 + }, + { + "epoch": 0.05901386748844376, + "grad_norm": 1.6268622791167293, + "learning_rate": 9.977944137094338e-06, + "loss": 0.3971, + "step": 2681 + }, + { + "epoch": 0.059035879374862425, + "grad_norm": 1.5458110767945632, + "learning_rate": 9.977910680387833e-06, + "loss": 0.4327, + "step": 2682 + }, + { + "epoch": 0.059057891261281095, + "grad_norm": 1.6576868135657938, + "learning_rate": 9.977877198381364e-06, + "loss": 0.321, + "step": 2683 + }, + { + "epoch": 0.05907990314769976, + "grad_norm": 6.402484390608937, + "learning_rate": 9.977843691075104e-06, + "loss": 0.3345, + "step": 2684 + }, + { + "epoch": 0.05910191503411842, + "grad_norm": 2.1187173466657736, + "learning_rate": 9.977810158469223e-06, + "loss": 0.34, + "step": 2685 + }, + { + "epoch": 0.05912392692053709, + "grad_norm": 2.0508589126500603, + "learning_rate": 9.977776600563889e-06, + "loss": 0.3387, + "step": 2686 + }, + { + "epoch": 0.059145938806955754, + "grad_norm": 2.6829895636361805, + "learning_rate": 9.977743017359275e-06, + "loss": 0.3492, + "step": 2687 + }, + { + "epoch": 0.059167950693374424, + "grad_norm": 2.7109677860762957, + "learning_rate": 9.977709408855552e-06, + "loss": 0.3596, + "step": 2688 + }, + { + "epoch": 0.05918996257979309, + "grad_norm": 2.808104110117117, + "learning_rate": 9.977675775052888e-06, + "loss": 0.3811, + "step": 2689 + }, + { + "epoch": 0.05921197446621176, + "grad_norm": 2.7040902170326744, + "learning_rate": 9.977642115951456e-06, + "loss": 0.3377, + "step": 2690 + }, + { + "epoch": 0.05923398635263042, + "grad_norm": 1.2977326368063966, + "learning_rate": 9.977608431551426e-06, + "loss": 0.6234, + "step": 2691 + }, + { + "epoch": 0.05925599823904908, + "grad_norm": 1.9026831330983525, + "learning_rate": 9.977574721852972e-06, + "loss": 0.4251, + "step": 2692 + }, + { + "epoch": 0.05927801012546775, + "grad_norm": 1.8510108945945012, + "learning_rate": 9.97754098685626e-06, + "loss": 0.437, + "step": 2693 + }, + { + "epoch": 0.059300022011886416, + "grad_norm": 1.9334209884987525, + "learning_rate": 9.977507226561467e-06, + "loss": 0.4122, + "step": 2694 + }, + { + "epoch": 0.059322033898305086, + "grad_norm": 1.8083914639748195, + "learning_rate": 9.977473440968763e-06, + "loss": 0.3809, + "step": 2695 + }, + { + "epoch": 0.05934404578472375, + "grad_norm": 16.425478840700702, + "learning_rate": 9.977439630078316e-06, + "loss": 0.4788, + "step": 2696 + }, + { + "epoch": 0.05936605767114242, + "grad_norm": 2.6240326193504075, + "learning_rate": 9.977405793890304e-06, + "loss": 0.3419, + "step": 2697 + }, + { + "epoch": 0.05938806955756108, + "grad_norm": 2.107985034450695, + "learning_rate": 9.977371932404894e-06, + "loss": 0.3876, + "step": 2698 + }, + { + "epoch": 0.05941008144397975, + "grad_norm": 2.5017602914543837, + "learning_rate": 9.977338045622261e-06, + "loss": 0.4237, + "step": 2699 + }, + { + "epoch": 0.059432093330398415, + "grad_norm": 1.6102648874001066, + "learning_rate": 9.977304133542574e-06, + "loss": 0.2763, + "step": 2700 + }, + { + "epoch": 0.05945410521681708, + "grad_norm": 2.9212442619826446, + "learning_rate": 9.977270196166008e-06, + "loss": 0.3595, + "step": 2701 + }, + { + "epoch": 0.05947611710323575, + "grad_norm": 1.7634334805939866, + "learning_rate": 9.977236233492736e-06, + "loss": 0.4584, + "step": 2702 + }, + { + "epoch": 0.05949812898965441, + "grad_norm": 2.225731309400079, + "learning_rate": 9.97720224552293e-06, + "loss": 0.3416, + "step": 2703 + }, + { + "epoch": 0.05952014087607308, + "grad_norm": 1.3754900255748397, + "learning_rate": 9.97716823225676e-06, + "loss": 0.295, + "step": 2704 + }, + { + "epoch": 0.059542152762491744, + "grad_norm": 1.6211671279906001, + "learning_rate": 9.977134193694405e-06, + "loss": 0.4234, + "step": 2705 + }, + { + "epoch": 0.059564164648910414, + "grad_norm": 2.1682144950953286, + "learning_rate": 9.97710012983603e-06, + "loss": 0.3642, + "step": 2706 + }, + { + "epoch": 0.05958617653532908, + "grad_norm": 1.0624813361079668, + "learning_rate": 9.977066040681815e-06, + "loss": 0.6025, + "step": 2707 + }, + { + "epoch": 0.05960818842174774, + "grad_norm": 1.3914530209402176, + "learning_rate": 9.97703192623193e-06, + "loss": 0.3508, + "step": 2708 + }, + { + "epoch": 0.05963020030816641, + "grad_norm": 1.4670560955059586, + "learning_rate": 9.976997786486549e-06, + "loss": 0.3398, + "step": 2709 + }, + { + "epoch": 0.05965221219458507, + "grad_norm": 2.0841771017360062, + "learning_rate": 9.976963621445844e-06, + "loss": 0.3945, + "step": 2710 + }, + { + "epoch": 0.05967422408100374, + "grad_norm": 0.6535904118220938, + "learning_rate": 9.976929431109992e-06, + "loss": 0.5745, + "step": 2711 + }, + { + "epoch": 0.059696235967422406, + "grad_norm": 1.7142026955190117, + "learning_rate": 9.976895215479163e-06, + "loss": 0.4106, + "step": 2712 + }, + { + "epoch": 0.059718247853841076, + "grad_norm": 1.7780118170061188, + "learning_rate": 9.976860974553534e-06, + "loss": 0.3297, + "step": 2713 + }, + { + "epoch": 0.05974025974025974, + "grad_norm": 2.1660415735789633, + "learning_rate": 9.976826708333278e-06, + "loss": 0.506, + "step": 2714 + }, + { + "epoch": 0.05976227162667841, + "grad_norm": 3.480836484712446, + "learning_rate": 9.976792416818568e-06, + "loss": 0.3465, + "step": 2715 + }, + { + "epoch": 0.05978428351309707, + "grad_norm": 2.031946659745916, + "learning_rate": 9.97675810000958e-06, + "loss": 0.4364, + "step": 2716 + }, + { + "epoch": 0.059806295399515735, + "grad_norm": 2.093901230354398, + "learning_rate": 9.976723757906487e-06, + "loss": 0.4271, + "step": 2717 + }, + { + "epoch": 0.059828307285934405, + "grad_norm": 1.7356784919507728, + "learning_rate": 9.976689390509464e-06, + "loss": 0.4067, + "step": 2718 + }, + { + "epoch": 0.05985031917235307, + "grad_norm": 1.6385427860689674, + "learning_rate": 9.976654997818686e-06, + "loss": 0.3738, + "step": 2719 + }, + { + "epoch": 0.05987233105877174, + "grad_norm": 2.2021293132051922, + "learning_rate": 9.976620579834329e-06, + "loss": 0.3866, + "step": 2720 + }, + { + "epoch": 0.0598943429451904, + "grad_norm": 1.3064029585556445, + "learning_rate": 9.976586136556565e-06, + "loss": 0.3862, + "step": 2721 + }, + { + "epoch": 0.05991635483160907, + "grad_norm": 1.75411300076758, + "learning_rate": 9.976551667985572e-06, + "loss": 0.4034, + "step": 2722 + }, + { + "epoch": 0.059938366718027734, + "grad_norm": 1.8445115124650224, + "learning_rate": 9.976517174121522e-06, + "loss": 0.4232, + "step": 2723 + }, + { + "epoch": 0.059960378604446404, + "grad_norm": 2.1008447808107293, + "learning_rate": 9.976482654964594e-06, + "loss": 0.3296, + "step": 2724 + }, + { + "epoch": 0.05998239049086507, + "grad_norm": 1.538837725962342, + "learning_rate": 9.97644811051496e-06, + "loss": 0.4096, + "step": 2725 + }, + { + "epoch": 0.06000440237728373, + "grad_norm": 1.798799179456045, + "learning_rate": 9.976413540772798e-06, + "loss": 0.3312, + "step": 2726 + }, + { + "epoch": 0.0600264142637024, + "grad_norm": 1.7611418325871324, + "learning_rate": 9.976378945738282e-06, + "loss": 0.3129, + "step": 2727 + }, + { + "epoch": 0.06004842615012106, + "grad_norm": 2.127026424862592, + "learning_rate": 9.976344325411591e-06, + "loss": 0.382, + "step": 2728 + }, + { + "epoch": 0.06007043803653973, + "grad_norm": 1.4564148849857352, + "learning_rate": 9.976309679792896e-06, + "loss": 0.3317, + "step": 2729 + }, + { + "epoch": 0.060092449922958396, + "grad_norm": 1.6738226201717477, + "learning_rate": 9.976275008882378e-06, + "loss": 0.3192, + "step": 2730 + }, + { + "epoch": 0.060114461809377066, + "grad_norm": 1.7628187934382995, + "learning_rate": 9.97624031268021e-06, + "loss": 0.3463, + "step": 2731 + }, + { + "epoch": 0.06013647369579573, + "grad_norm": 2.5852377758940017, + "learning_rate": 9.976205591186569e-06, + "loss": 0.3099, + "step": 2732 + }, + { + "epoch": 0.06015848558221439, + "grad_norm": 1.668554767377942, + "learning_rate": 9.976170844401632e-06, + "loss": 0.4687, + "step": 2733 + }, + { + "epoch": 0.06018049746863306, + "grad_norm": 1.6220310840810708, + "learning_rate": 9.976136072325575e-06, + "loss": 0.3991, + "step": 2734 + }, + { + "epoch": 0.060202509355051725, + "grad_norm": 2.372943712018614, + "learning_rate": 9.976101274958574e-06, + "loss": 0.4734, + "step": 2735 + }, + { + "epoch": 0.060224521241470395, + "grad_norm": 1.6170569530451866, + "learning_rate": 9.976066452300808e-06, + "loss": 0.298, + "step": 2736 + }, + { + "epoch": 0.06024653312788906, + "grad_norm": 1.7878970353802273, + "learning_rate": 9.976031604352454e-06, + "loss": 0.4026, + "step": 2737 + }, + { + "epoch": 0.06026854501430773, + "grad_norm": 2.380475742793439, + "learning_rate": 9.975996731113686e-06, + "loss": 0.4653, + "step": 2738 + }, + { + "epoch": 0.06029055690072639, + "grad_norm": 1.7979243537353395, + "learning_rate": 9.975961832584683e-06, + "loss": 0.4472, + "step": 2739 + }, + { + "epoch": 0.06031256878714506, + "grad_norm": 1.6703332446421537, + "learning_rate": 9.975926908765625e-06, + "loss": 0.4072, + "step": 2740 + }, + { + "epoch": 0.060334580673563724, + "grad_norm": 1.9744461762645578, + "learning_rate": 9.975891959656687e-06, + "loss": 0.4448, + "step": 2741 + }, + { + "epoch": 0.06035659255998239, + "grad_norm": 3.5695335677608657, + "learning_rate": 9.975856985258045e-06, + "loss": 0.3496, + "step": 2742 + }, + { + "epoch": 0.06037860444640106, + "grad_norm": 1.675503774874559, + "learning_rate": 9.97582198556988e-06, + "loss": 0.3068, + "step": 2743 + }, + { + "epoch": 0.06040061633281972, + "grad_norm": 4.523319172727671, + "learning_rate": 9.975786960592366e-06, + "loss": 0.3862, + "step": 2744 + }, + { + "epoch": 0.06042262821923839, + "grad_norm": 1.7791268090848122, + "learning_rate": 9.975751910325686e-06, + "loss": 0.3984, + "step": 2745 + }, + { + "epoch": 0.06044464010565705, + "grad_norm": 3.2071836777172336, + "learning_rate": 9.975716834770015e-06, + "loss": 0.4988, + "step": 2746 + }, + { + "epoch": 0.06046665199207572, + "grad_norm": 1.62103693677764, + "learning_rate": 9.97568173392553e-06, + "loss": 0.3325, + "step": 2747 + }, + { + "epoch": 0.060488663878494386, + "grad_norm": 1.2056228586949682, + "learning_rate": 9.975646607792414e-06, + "loss": 0.6053, + "step": 2748 + }, + { + "epoch": 0.060510675764913056, + "grad_norm": 1.9811593450175538, + "learning_rate": 9.97561145637084e-06, + "loss": 0.3698, + "step": 2749 + }, + { + "epoch": 0.06053268765133172, + "grad_norm": 5.7533960911846975, + "learning_rate": 9.97557627966099e-06, + "loss": 0.3995, + "step": 2750 + }, + { + "epoch": 0.06055469953775038, + "grad_norm": 2.009084142665422, + "learning_rate": 9.975541077663043e-06, + "loss": 0.4004, + "step": 2751 + }, + { + "epoch": 0.06057671142416905, + "grad_norm": 2.036036645452174, + "learning_rate": 9.975505850377179e-06, + "loss": 0.3453, + "step": 2752 + }, + { + "epoch": 0.060598723310587715, + "grad_norm": 2.757955083797331, + "learning_rate": 9.975470597803571e-06, + "loss": 0.5422, + "step": 2753 + }, + { + "epoch": 0.060620735197006385, + "grad_norm": 1.7522427972367125, + "learning_rate": 9.975435319942406e-06, + "loss": 0.4369, + "step": 2754 + }, + { + "epoch": 0.06064274708342505, + "grad_norm": 3.32295946754796, + "learning_rate": 9.975400016793857e-06, + "loss": 0.3985, + "step": 2755 + }, + { + "epoch": 0.06066475896984372, + "grad_norm": 3.885362818608051, + "learning_rate": 9.975364688358107e-06, + "loss": 0.3882, + "step": 2756 + }, + { + "epoch": 0.06068677085626238, + "grad_norm": 1.4378664303546496, + "learning_rate": 9.975329334635335e-06, + "loss": 0.284, + "step": 2757 + }, + { + "epoch": 0.060708782742681044, + "grad_norm": 1.7155784141861612, + "learning_rate": 9.97529395562572e-06, + "loss": 0.3888, + "step": 2758 + }, + { + "epoch": 0.060730794629099714, + "grad_norm": 1.40019218749554, + "learning_rate": 9.97525855132944e-06, + "loss": 0.3873, + "step": 2759 + }, + { + "epoch": 0.06075280651551838, + "grad_norm": 1.792223255208145, + "learning_rate": 9.97522312174668e-06, + "loss": 0.31, + "step": 2760 + }, + { + "epoch": 0.06077481840193705, + "grad_norm": 1.826628551841873, + "learning_rate": 9.975187666877617e-06, + "loss": 0.4472, + "step": 2761 + }, + { + "epoch": 0.06079683028835571, + "grad_norm": 0.9975180820131585, + "learning_rate": 9.975152186722429e-06, + "loss": 0.6018, + "step": 2762 + }, + { + "epoch": 0.06081884217477438, + "grad_norm": 1.6416259067356045, + "learning_rate": 9.9751166812813e-06, + "loss": 0.4006, + "step": 2763 + }, + { + "epoch": 0.06084085406119304, + "grad_norm": 2.6270851746641557, + "learning_rate": 9.97508115055441e-06, + "loss": 0.431, + "step": 2764 + }, + { + "epoch": 0.06086286594761171, + "grad_norm": 2.114478938097615, + "learning_rate": 9.975045594541936e-06, + "loss": 0.3487, + "step": 2765 + }, + { + "epoch": 0.060884877834030376, + "grad_norm": 1.4450168689072158, + "learning_rate": 9.975010013244064e-06, + "loss": 0.375, + "step": 2766 + }, + { + "epoch": 0.06090688972044904, + "grad_norm": 1.3900211694080034, + "learning_rate": 9.974974406660971e-06, + "loss": 0.3362, + "step": 2767 + }, + { + "epoch": 0.06092890160686771, + "grad_norm": 3.8469548656663264, + "learning_rate": 9.97493877479284e-06, + "loss": 0.363, + "step": 2768 + }, + { + "epoch": 0.06095091349328637, + "grad_norm": 0.6445847782004598, + "learning_rate": 9.97490311763985e-06, + "loss": 0.5869, + "step": 2769 + }, + { + "epoch": 0.06097292537970504, + "grad_norm": 2.1231891610038645, + "learning_rate": 9.974867435202184e-06, + "loss": 0.5604, + "step": 2770 + }, + { + "epoch": 0.060994937266123705, + "grad_norm": 2.096703876137983, + "learning_rate": 9.974831727480023e-06, + "loss": 0.4125, + "step": 2771 + }, + { + "epoch": 0.061016949152542375, + "grad_norm": 2.1475019357905762, + "learning_rate": 9.974795994473548e-06, + "loss": 0.3925, + "step": 2772 + }, + { + "epoch": 0.06103896103896104, + "grad_norm": 5.31504899043628, + "learning_rate": 9.97476023618294e-06, + "loss": 0.4051, + "step": 2773 + }, + { + "epoch": 0.06106097292537971, + "grad_norm": 1.6798118203229466, + "learning_rate": 9.974724452608385e-06, + "loss": 0.4266, + "step": 2774 + }, + { + "epoch": 0.06108298481179837, + "grad_norm": 2.1900894529483965, + "learning_rate": 9.97468864375006e-06, + "loss": 0.4227, + "step": 2775 + }, + { + "epoch": 0.061104996698217035, + "grad_norm": 1.9137209513371793, + "learning_rate": 9.974652809608147e-06, + "loss": 0.3652, + "step": 2776 + }, + { + "epoch": 0.061127008584635704, + "grad_norm": 1.6340049815537856, + "learning_rate": 9.97461695018283e-06, + "loss": 0.3577, + "step": 2777 + }, + { + "epoch": 0.06114902047105437, + "grad_norm": 2.229248080694313, + "learning_rate": 9.97458106547429e-06, + "loss": 0.4379, + "step": 2778 + }, + { + "epoch": 0.06117103235747304, + "grad_norm": 3.193562304373662, + "learning_rate": 9.974545155482713e-06, + "loss": 0.4003, + "step": 2779 + }, + { + "epoch": 0.0611930442438917, + "grad_norm": 1.432065415330071, + "learning_rate": 9.974509220208279e-06, + "loss": 0.3553, + "step": 2780 + }, + { + "epoch": 0.06121505613031037, + "grad_norm": 2.1175065759435188, + "learning_rate": 9.974473259651169e-06, + "loss": 0.4224, + "step": 2781 + }, + { + "epoch": 0.061237068016729033, + "grad_norm": 1.7308522161946749, + "learning_rate": 9.974437273811567e-06, + "loss": 0.3107, + "step": 2782 + }, + { + "epoch": 0.061259079903147697, + "grad_norm": 1.4587856998974327, + "learning_rate": 9.974401262689657e-06, + "loss": 0.3383, + "step": 2783 + }, + { + "epoch": 0.061281091789566366, + "grad_norm": 1.5311532202197773, + "learning_rate": 9.97436522628562e-06, + "loss": 0.4033, + "step": 2784 + }, + { + "epoch": 0.06130310367598503, + "grad_norm": 2.4125602402778132, + "learning_rate": 9.974329164599642e-06, + "loss": 0.4217, + "step": 2785 + }, + { + "epoch": 0.0613251155624037, + "grad_norm": 1.4350631213372114, + "learning_rate": 9.974293077631902e-06, + "loss": 0.2957, + "step": 2786 + }, + { + "epoch": 0.06134712744882236, + "grad_norm": 2.442554052250134, + "learning_rate": 9.974256965382588e-06, + "loss": 0.4236, + "step": 2787 + }, + { + "epoch": 0.06136913933524103, + "grad_norm": 1.8155614197348005, + "learning_rate": 9.974220827851881e-06, + "loss": 0.3833, + "step": 2788 + }, + { + "epoch": 0.061391151221659696, + "grad_norm": 1.668084351352378, + "learning_rate": 9.974184665039965e-06, + "loss": 0.3882, + "step": 2789 + }, + { + "epoch": 0.061413163108078365, + "grad_norm": 2.9150018326741933, + "learning_rate": 9.974148476947024e-06, + "loss": 0.3637, + "step": 2790 + }, + { + "epoch": 0.06143517499449703, + "grad_norm": 1.4650002444822228, + "learning_rate": 9.97411226357324e-06, + "loss": 0.4387, + "step": 2791 + }, + { + "epoch": 0.06145718688091569, + "grad_norm": 1.7955277506889913, + "learning_rate": 9.974076024918802e-06, + "loss": 0.4354, + "step": 2792 + }, + { + "epoch": 0.06147919876733436, + "grad_norm": 2.2290834985838788, + "learning_rate": 9.974039760983891e-06, + "loss": 0.4116, + "step": 2793 + }, + { + "epoch": 0.061501210653753025, + "grad_norm": 0.8221626333443756, + "learning_rate": 9.97400347176869e-06, + "loss": 0.6027, + "step": 2794 + }, + { + "epoch": 0.061523222540171695, + "grad_norm": 1.5612436638357916, + "learning_rate": 9.973967157273386e-06, + "loss": 0.3202, + "step": 2795 + }, + { + "epoch": 0.06154523442659036, + "grad_norm": 0.7138926528099169, + "learning_rate": 9.97393081749816e-06, + "loss": 0.6048, + "step": 2796 + }, + { + "epoch": 0.06156724631300903, + "grad_norm": 1.5562518698545946, + "learning_rate": 9.973894452443202e-06, + "loss": 0.399, + "step": 2797 + }, + { + "epoch": 0.06158925819942769, + "grad_norm": 1.5919327314746081, + "learning_rate": 9.973858062108694e-06, + "loss": 0.3094, + "step": 2798 + }, + { + "epoch": 0.06161127008584636, + "grad_norm": 0.7299287890731432, + "learning_rate": 9.973821646494819e-06, + "loss": 0.5677, + "step": 2799 + }, + { + "epoch": 0.061633281972265024, + "grad_norm": 0.7283957485105121, + "learning_rate": 9.973785205601764e-06, + "loss": 0.621, + "step": 2800 + }, + { + "epoch": 0.06165529385868369, + "grad_norm": 1.7051792543829045, + "learning_rate": 9.973748739429716e-06, + "loss": 0.3433, + "step": 2801 + }, + { + "epoch": 0.061677305745102357, + "grad_norm": 1.4961334700528077, + "learning_rate": 9.973712247978858e-06, + "loss": 0.3102, + "step": 2802 + }, + { + "epoch": 0.06169931763152102, + "grad_norm": 0.7234084154933775, + "learning_rate": 9.973675731249377e-06, + "loss": 0.6127, + "step": 2803 + }, + { + "epoch": 0.06172132951793969, + "grad_norm": 1.8042010173983274, + "learning_rate": 9.973639189241455e-06, + "loss": 0.3659, + "step": 2804 + }, + { + "epoch": 0.06174334140435835, + "grad_norm": 2.3084570820246375, + "learning_rate": 9.973602621955283e-06, + "loss": 0.3991, + "step": 2805 + }, + { + "epoch": 0.06176535329077702, + "grad_norm": 3.111367284701426, + "learning_rate": 9.973566029391041e-06, + "loss": 0.4483, + "step": 2806 + }, + { + "epoch": 0.061787365177195686, + "grad_norm": 1.7190213045467617, + "learning_rate": 9.973529411548921e-06, + "loss": 0.3609, + "step": 2807 + }, + { + "epoch": 0.06180937706361435, + "grad_norm": 1.5632943601120552, + "learning_rate": 9.973492768429105e-06, + "loss": 0.3692, + "step": 2808 + }, + { + "epoch": 0.06183138895003302, + "grad_norm": 1.5947101563690331, + "learning_rate": 9.97345610003178e-06, + "loss": 0.3489, + "step": 2809 + }, + { + "epoch": 0.06185340083645168, + "grad_norm": 3.1383745521831767, + "learning_rate": 9.973419406357135e-06, + "loss": 0.3936, + "step": 2810 + }, + { + "epoch": 0.06187541272287035, + "grad_norm": 1.834695088182512, + "learning_rate": 9.973382687405352e-06, + "loss": 0.4126, + "step": 2811 + }, + { + "epoch": 0.061897424609289015, + "grad_norm": 1.451320085020713, + "learning_rate": 9.973345943176623e-06, + "loss": 0.3891, + "step": 2812 + }, + { + "epoch": 0.061919436495707685, + "grad_norm": 2.058012605661944, + "learning_rate": 9.973309173671129e-06, + "loss": 0.4126, + "step": 2813 + }, + { + "epoch": 0.06194144838212635, + "grad_norm": 1.693701553533326, + "learning_rate": 9.97327237888906e-06, + "loss": 0.3797, + "step": 2814 + }, + { + "epoch": 0.06196346026854502, + "grad_norm": 1.9948157512039804, + "learning_rate": 9.973235558830602e-06, + "loss": 0.4549, + "step": 2815 + }, + { + "epoch": 0.06198547215496368, + "grad_norm": 1.8552205606987828, + "learning_rate": 9.973198713495946e-06, + "loss": 0.4169, + "step": 2816 + }, + { + "epoch": 0.062007484041382344, + "grad_norm": 1.7844016612627374, + "learning_rate": 9.973161842885272e-06, + "loss": 0.418, + "step": 2817 + }, + { + "epoch": 0.062029495927801014, + "grad_norm": 1.8080330529390034, + "learning_rate": 9.973124946998773e-06, + "loss": 0.3698, + "step": 2818 + }, + { + "epoch": 0.06205150781421968, + "grad_norm": 0.8088415596595495, + "learning_rate": 9.973088025836638e-06, + "loss": 0.5756, + "step": 2819 + }, + { + "epoch": 0.06207351970063835, + "grad_norm": 1.4287271736780351, + "learning_rate": 9.973051079399048e-06, + "loss": 0.3544, + "step": 2820 + }, + { + "epoch": 0.06209553158705701, + "grad_norm": 1.9500719620427989, + "learning_rate": 9.973014107686195e-06, + "loss": 0.4205, + "step": 2821 + }, + { + "epoch": 0.06211754347347568, + "grad_norm": 2.1518855384541555, + "learning_rate": 9.972977110698265e-06, + "loss": 0.3938, + "step": 2822 + }, + { + "epoch": 0.06213955535989434, + "grad_norm": 2.6950034679893773, + "learning_rate": 9.97294008843545e-06, + "loss": 0.2882, + "step": 2823 + }, + { + "epoch": 0.062161567246313006, + "grad_norm": 3.311219086605014, + "learning_rate": 9.972903040897934e-06, + "loss": 0.307, + "step": 2824 + }, + { + "epoch": 0.062183579132731676, + "grad_norm": 1.907597530079572, + "learning_rate": 9.972865968085907e-06, + "loss": 0.3949, + "step": 2825 + }, + { + "epoch": 0.06220559101915034, + "grad_norm": 2.078184487946816, + "learning_rate": 9.972828869999558e-06, + "loss": 0.3336, + "step": 2826 + }, + { + "epoch": 0.06222760290556901, + "grad_norm": 2.4008045705153096, + "learning_rate": 9.972791746639073e-06, + "loss": 0.4535, + "step": 2827 + }, + { + "epoch": 0.06224961479198767, + "grad_norm": 1.4976328209778482, + "learning_rate": 9.972754598004644e-06, + "loss": 0.2566, + "step": 2828 + }, + { + "epoch": 0.06227162667840634, + "grad_norm": 0.6879970618661614, + "learning_rate": 9.972717424096456e-06, + "loss": 0.5841, + "step": 2829 + }, + { + "epoch": 0.062293638564825005, + "grad_norm": 1.9631251623022563, + "learning_rate": 9.972680224914702e-06, + "loss": 0.3213, + "step": 2830 + }, + { + "epoch": 0.062315650451243675, + "grad_norm": 1.5143069099520126, + "learning_rate": 9.97264300045957e-06, + "loss": 0.3383, + "step": 2831 + }, + { + "epoch": 0.06233766233766234, + "grad_norm": 1.7191142663451846, + "learning_rate": 9.972605750731248e-06, + "loss": 0.3571, + "step": 2832 + }, + { + "epoch": 0.062359674224081, + "grad_norm": 1.4096588215335628, + "learning_rate": 9.972568475729925e-06, + "loss": 0.4933, + "step": 2833 + }, + { + "epoch": 0.06238168611049967, + "grad_norm": 7.168372299668961, + "learning_rate": 9.97253117545579e-06, + "loss": 0.3247, + "step": 2834 + }, + { + "epoch": 0.062403697996918334, + "grad_norm": 1.5416315557600242, + "learning_rate": 9.972493849909035e-06, + "loss": 0.3246, + "step": 2835 + }, + { + "epoch": 0.062425709883337004, + "grad_norm": 1.5297106920579566, + "learning_rate": 9.972456499089849e-06, + "loss": 0.4655, + "step": 2836 + }, + { + "epoch": 0.06244772176975567, + "grad_norm": 1.683127958021478, + "learning_rate": 9.97241912299842e-06, + "loss": 0.4669, + "step": 2837 + }, + { + "epoch": 0.06246973365617434, + "grad_norm": 1.9688092805609294, + "learning_rate": 9.97238172163494e-06, + "loss": 0.3936, + "step": 2838 + }, + { + "epoch": 0.062491745542593, + "grad_norm": 2.211943584370897, + "learning_rate": 9.972344294999598e-06, + "loss": 0.503, + "step": 2839 + }, + { + "epoch": 0.06251375742901166, + "grad_norm": 1.5797085945314464, + "learning_rate": 9.972306843092586e-06, + "loss": 0.3513, + "step": 2840 + }, + { + "epoch": 0.06253576931543033, + "grad_norm": 1.4609347698453359, + "learning_rate": 9.972269365914092e-06, + "loss": 0.3511, + "step": 2841 + }, + { + "epoch": 0.062557781201849, + "grad_norm": 1.986877670176865, + "learning_rate": 9.972231863464308e-06, + "loss": 0.4157, + "step": 2842 + }, + { + "epoch": 0.06257979308826767, + "grad_norm": 1.4206671729527653, + "learning_rate": 9.972194335743422e-06, + "loss": 0.3454, + "step": 2843 + }, + { + "epoch": 0.06260180497468633, + "grad_norm": 1.6346696846396807, + "learning_rate": 9.97215678275163e-06, + "loss": 0.4123, + "step": 2844 + }, + { + "epoch": 0.06262381686110499, + "grad_norm": 1.76731467523463, + "learning_rate": 9.972119204489116e-06, + "loss": 0.4579, + "step": 2845 + }, + { + "epoch": 0.06264582874752367, + "grad_norm": 1.8120404106991184, + "learning_rate": 9.972081600956077e-06, + "loss": 0.401, + "step": 2846 + }, + { + "epoch": 0.06266784063394233, + "grad_norm": 1.8838626652500114, + "learning_rate": 9.9720439721527e-06, + "loss": 0.3794, + "step": 2847 + }, + { + "epoch": 0.062689852520361, + "grad_norm": 4.091596208355047, + "learning_rate": 9.97200631807918e-06, + "loss": 0.3167, + "step": 2848 + }, + { + "epoch": 0.06271186440677966, + "grad_norm": 2.772810908735078, + "learning_rate": 9.971968638735704e-06, + "loss": 0.3724, + "step": 2849 + }, + { + "epoch": 0.06273387629319832, + "grad_norm": 1.9428715220586148, + "learning_rate": 9.971930934122467e-06, + "loss": 0.3647, + "step": 2850 + }, + { + "epoch": 0.062755888179617, + "grad_norm": 2.572934225409613, + "learning_rate": 9.97189320423966e-06, + "loss": 0.4135, + "step": 2851 + }, + { + "epoch": 0.06277790006603566, + "grad_norm": 1.4433911368719516, + "learning_rate": 9.971855449087474e-06, + "loss": 0.4063, + "step": 2852 + }, + { + "epoch": 0.06279991195245432, + "grad_norm": 0.6947182457793895, + "learning_rate": 9.9718176686661e-06, + "loss": 0.5989, + "step": 2853 + }, + { + "epoch": 0.06282192383887299, + "grad_norm": 1.8384660676632458, + "learning_rate": 9.971779862975731e-06, + "loss": 0.4193, + "step": 2854 + }, + { + "epoch": 0.06284393572529166, + "grad_norm": 1.3552281693170138, + "learning_rate": 9.971742032016561e-06, + "loss": 0.3345, + "step": 2855 + }, + { + "epoch": 0.06286594761171033, + "grad_norm": 2.066011980718705, + "learning_rate": 9.971704175788778e-06, + "loss": 0.3636, + "step": 2856 + }, + { + "epoch": 0.06288795949812899, + "grad_norm": 0.6676081236763122, + "learning_rate": 9.971666294292578e-06, + "loss": 0.5925, + "step": 2857 + }, + { + "epoch": 0.06290997138454765, + "grad_norm": 1.773518803184834, + "learning_rate": 9.971628387528154e-06, + "loss": 0.4761, + "step": 2858 + }, + { + "epoch": 0.06293198327096632, + "grad_norm": 1.9776929448855156, + "learning_rate": 9.971590455495697e-06, + "loss": 0.4378, + "step": 2859 + }, + { + "epoch": 0.06295399515738499, + "grad_norm": 1.2392147307920174, + "learning_rate": 9.971552498195398e-06, + "loss": 0.32, + "step": 2860 + }, + { + "epoch": 0.06297600704380366, + "grad_norm": 0.6287573841873993, + "learning_rate": 9.971514515627454e-06, + "loss": 0.5543, + "step": 2861 + }, + { + "epoch": 0.06299801893022232, + "grad_norm": 1.3369248754087173, + "learning_rate": 9.971476507792054e-06, + "loss": 0.4237, + "step": 2862 + }, + { + "epoch": 0.06302003081664098, + "grad_norm": 1.6319846770734012, + "learning_rate": 9.971438474689395e-06, + "loss": 0.4373, + "step": 2863 + }, + { + "epoch": 0.06304204270305966, + "grad_norm": 1.223377561609134, + "learning_rate": 9.971400416319666e-06, + "loss": 0.3555, + "step": 2864 + }, + { + "epoch": 0.06306405458947832, + "grad_norm": 0.7028324962544716, + "learning_rate": 9.971362332683063e-06, + "loss": 0.5998, + "step": 2865 + }, + { + "epoch": 0.06308606647589698, + "grad_norm": 0.6657655890354589, + "learning_rate": 9.971324223779782e-06, + "loss": 0.6056, + "step": 2866 + }, + { + "epoch": 0.06310807836231565, + "grad_norm": 0.6589813440650235, + "learning_rate": 9.971286089610012e-06, + "loss": 0.6062, + "step": 2867 + }, + { + "epoch": 0.06313009024873431, + "grad_norm": 1.4886558070057148, + "learning_rate": 9.971247930173948e-06, + "loss": 0.4165, + "step": 2868 + }, + { + "epoch": 0.06315210213515299, + "grad_norm": 1.563033075579979, + "learning_rate": 9.971209745471786e-06, + "loss": 0.3907, + "step": 2869 + }, + { + "epoch": 0.06317411402157165, + "grad_norm": 1.4792561371414912, + "learning_rate": 9.971171535503719e-06, + "loss": 0.3889, + "step": 2870 + }, + { + "epoch": 0.06319612590799031, + "grad_norm": 1.6580933129769164, + "learning_rate": 9.971133300269941e-06, + "loss": 0.3436, + "step": 2871 + }, + { + "epoch": 0.06321813779440898, + "grad_norm": 1.614975955084317, + "learning_rate": 9.971095039770646e-06, + "loss": 0.3246, + "step": 2872 + }, + { + "epoch": 0.06324014968082765, + "grad_norm": 1.2205004309254754, + "learning_rate": 9.971056754006029e-06, + "loss": 0.3283, + "step": 2873 + }, + { + "epoch": 0.06326216156724632, + "grad_norm": 1.540590700683975, + "learning_rate": 9.971018442976284e-06, + "loss": 0.3436, + "step": 2874 + }, + { + "epoch": 0.06328417345366498, + "grad_norm": 2.0527517275477045, + "learning_rate": 9.970980106681608e-06, + "loss": 0.3889, + "step": 2875 + }, + { + "epoch": 0.06330618534008364, + "grad_norm": 1.6264287325767504, + "learning_rate": 9.970941745122192e-06, + "loss": 0.3524, + "step": 2876 + }, + { + "epoch": 0.0633281972265023, + "grad_norm": 1.9560151273125057, + "learning_rate": 9.970903358298233e-06, + "loss": 0.5558, + "step": 2877 + }, + { + "epoch": 0.06335020911292098, + "grad_norm": 1.6225187960546714, + "learning_rate": 9.970864946209925e-06, + "loss": 0.4434, + "step": 2878 + }, + { + "epoch": 0.06337222099933965, + "grad_norm": 1.5868019478858413, + "learning_rate": 9.970826508857468e-06, + "loss": 0.3539, + "step": 2879 + }, + { + "epoch": 0.06339423288575831, + "grad_norm": 1.2856583868804972, + "learning_rate": 9.970788046241052e-06, + "loss": 0.3359, + "step": 2880 + }, + { + "epoch": 0.06341624477217697, + "grad_norm": 1.9394657243079552, + "learning_rate": 9.970749558360873e-06, + "loss": 0.4154, + "step": 2881 + }, + { + "epoch": 0.06343825665859563, + "grad_norm": 1.5349300039805147, + "learning_rate": 9.970711045217127e-06, + "loss": 0.3856, + "step": 2882 + }, + { + "epoch": 0.06346026854501431, + "grad_norm": 1.6139528683802393, + "learning_rate": 9.970672506810013e-06, + "loss": 0.4704, + "step": 2883 + }, + { + "epoch": 0.06348228043143297, + "grad_norm": 1.704510322721273, + "learning_rate": 9.970633943139722e-06, + "loss": 0.3733, + "step": 2884 + }, + { + "epoch": 0.06350429231785164, + "grad_norm": 3.4811751701779716, + "learning_rate": 9.970595354206453e-06, + "loss": 0.3992, + "step": 2885 + }, + { + "epoch": 0.0635263042042703, + "grad_norm": 2.195102768340707, + "learning_rate": 9.970556740010402e-06, + "loss": 0.3593, + "step": 2886 + }, + { + "epoch": 0.06354831609068898, + "grad_norm": 2.9863325884097165, + "learning_rate": 9.970518100551765e-06, + "loss": 0.4437, + "step": 2887 + }, + { + "epoch": 0.06357032797710764, + "grad_norm": 1.8365888558830827, + "learning_rate": 9.970479435830738e-06, + "loss": 0.5025, + "step": 2888 + }, + { + "epoch": 0.0635923398635263, + "grad_norm": 1.4461372536476826, + "learning_rate": 9.970440745847518e-06, + "loss": 0.3461, + "step": 2889 + }, + { + "epoch": 0.06361435174994497, + "grad_norm": 0.9464458901996844, + "learning_rate": 9.9704020306023e-06, + "loss": 0.6283, + "step": 2890 + }, + { + "epoch": 0.06363636363636363, + "grad_norm": 7.2809710256558695, + "learning_rate": 9.970363290095282e-06, + "loss": 0.3012, + "step": 2891 + }, + { + "epoch": 0.0636583755227823, + "grad_norm": 1.7923446148496671, + "learning_rate": 9.970324524326663e-06, + "loss": 0.4194, + "step": 2892 + }, + { + "epoch": 0.06368038740920097, + "grad_norm": 1.425737378893241, + "learning_rate": 9.970285733296637e-06, + "loss": 0.4075, + "step": 2893 + }, + { + "epoch": 0.06370239929561963, + "grad_norm": 0.6706508205036525, + "learning_rate": 9.9702469170054e-06, + "loss": 0.5904, + "step": 2894 + }, + { + "epoch": 0.0637244111820383, + "grad_norm": 1.3788941843459088, + "learning_rate": 9.970208075453153e-06, + "loss": 0.4249, + "step": 2895 + }, + { + "epoch": 0.06374642306845697, + "grad_norm": 1.7811509057706312, + "learning_rate": 9.970169208640092e-06, + "loss": 0.4032, + "step": 2896 + }, + { + "epoch": 0.06376843495487564, + "grad_norm": 1.7329898494306548, + "learning_rate": 9.970130316566413e-06, + "loss": 0.3516, + "step": 2897 + }, + { + "epoch": 0.0637904468412943, + "grad_norm": 1.5847584935965993, + "learning_rate": 9.970091399232317e-06, + "loss": 0.3543, + "step": 2898 + }, + { + "epoch": 0.06381245872771296, + "grad_norm": 3.1119221019090717, + "learning_rate": 9.970052456637998e-06, + "loss": 0.4189, + "step": 2899 + }, + { + "epoch": 0.06383447061413162, + "grad_norm": 1.406329214362476, + "learning_rate": 9.970013488783657e-06, + "loss": 0.462, + "step": 2900 + }, + { + "epoch": 0.0638564825005503, + "grad_norm": 2.051050182737118, + "learning_rate": 9.96997449566949e-06, + "loss": 0.3836, + "step": 2901 + }, + { + "epoch": 0.06387849438696896, + "grad_norm": 1.968779346396818, + "learning_rate": 9.969935477295697e-06, + "loss": 0.3928, + "step": 2902 + }, + { + "epoch": 0.06390050627338763, + "grad_norm": 1.3920960396460516, + "learning_rate": 9.969896433662473e-06, + "loss": 0.3707, + "step": 2903 + }, + { + "epoch": 0.06392251815980629, + "grad_norm": 1.5154041880464095, + "learning_rate": 9.96985736477002e-06, + "loss": 0.3593, + "step": 2904 + }, + { + "epoch": 0.06394453004622497, + "grad_norm": 1.3335638366841476, + "learning_rate": 9.969818270618536e-06, + "loss": 0.402, + "step": 2905 + }, + { + "epoch": 0.06396654193264363, + "grad_norm": 3.0567251214217332, + "learning_rate": 9.969779151208219e-06, + "loss": 0.4286, + "step": 2906 + }, + { + "epoch": 0.0639885538190623, + "grad_norm": 1.0278151726544582, + "learning_rate": 9.969740006539267e-06, + "loss": 0.6245, + "step": 2907 + }, + { + "epoch": 0.06401056570548096, + "grad_norm": 0.8279686580735818, + "learning_rate": 9.96970083661188e-06, + "loss": 0.5898, + "step": 2908 + }, + { + "epoch": 0.06403257759189962, + "grad_norm": 0.7242616947462195, + "learning_rate": 9.969661641426255e-06, + "loss": 0.6097, + "step": 2909 + }, + { + "epoch": 0.0640545894783183, + "grad_norm": 1.6322996078244885, + "learning_rate": 9.969622420982597e-06, + "loss": 0.3636, + "step": 2910 + }, + { + "epoch": 0.06407660136473696, + "grad_norm": 0.745497956473502, + "learning_rate": 9.969583175281099e-06, + "loss": 0.5756, + "step": 2911 + }, + { + "epoch": 0.06409861325115562, + "grad_norm": 1.9570810846484752, + "learning_rate": 9.969543904321962e-06, + "loss": 0.4131, + "step": 2912 + }, + { + "epoch": 0.06412062513757429, + "grad_norm": 0.8966105748829923, + "learning_rate": 9.969504608105388e-06, + "loss": 0.599, + "step": 2913 + }, + { + "epoch": 0.06414263702399296, + "grad_norm": 0.8311371826244248, + "learning_rate": 9.969465286631573e-06, + "loss": 0.5815, + "step": 2914 + }, + { + "epoch": 0.06416464891041163, + "grad_norm": 1.5952037817774147, + "learning_rate": 9.969425939900721e-06, + "loss": 0.384, + "step": 2915 + }, + { + "epoch": 0.06418666079683029, + "grad_norm": 3.795184314529443, + "learning_rate": 9.96938656791303e-06, + "loss": 0.4052, + "step": 2916 + }, + { + "epoch": 0.06420867268324895, + "grad_norm": 1.802320260236647, + "learning_rate": 9.969347170668699e-06, + "loss": 0.3883, + "step": 2917 + }, + { + "epoch": 0.06423068456966761, + "grad_norm": 1.7013456411361696, + "learning_rate": 9.96930774816793e-06, + "loss": 0.385, + "step": 2918 + }, + { + "epoch": 0.06425269645608629, + "grad_norm": 1.7294326912555837, + "learning_rate": 9.969268300410921e-06, + "loss": 0.3978, + "step": 2919 + }, + { + "epoch": 0.06427470834250495, + "grad_norm": 1.2960150366754826, + "learning_rate": 9.969228827397876e-06, + "loss": 0.3659, + "step": 2920 + }, + { + "epoch": 0.06429672022892362, + "grad_norm": 0.8535448844338298, + "learning_rate": 9.969189329128993e-06, + "loss": 0.6248, + "step": 2921 + }, + { + "epoch": 0.06431873211534228, + "grad_norm": 1.7586369343764454, + "learning_rate": 9.969149805604472e-06, + "loss": 0.4491, + "step": 2922 + }, + { + "epoch": 0.06434074400176096, + "grad_norm": 1.9029871920453059, + "learning_rate": 9.96911025682452e-06, + "loss": 0.43, + "step": 2923 + }, + { + "epoch": 0.06436275588817962, + "grad_norm": 0.7281797360859952, + "learning_rate": 9.969070682789327e-06, + "loss": 0.6097, + "step": 2924 + }, + { + "epoch": 0.06438476777459828, + "grad_norm": 1.4657693929134359, + "learning_rate": 9.969031083499106e-06, + "loss": 0.2973, + "step": 2925 + }, + { + "epoch": 0.06440677966101695, + "grad_norm": 2.1750219704721436, + "learning_rate": 9.968991458954051e-06, + "loss": 0.5057, + "step": 2926 + }, + { + "epoch": 0.06442879154743561, + "grad_norm": 1.482456395033747, + "learning_rate": 9.968951809154364e-06, + "loss": 0.305, + "step": 2927 + }, + { + "epoch": 0.06445080343385429, + "grad_norm": 2.703694723690597, + "learning_rate": 9.968912134100249e-06, + "loss": 0.4163, + "step": 2928 + }, + { + "epoch": 0.06447281532027295, + "grad_norm": 1.4475874771453303, + "learning_rate": 9.968872433791905e-06, + "loss": 0.4285, + "step": 2929 + }, + { + "epoch": 0.06449482720669161, + "grad_norm": 1.8956359204502224, + "learning_rate": 9.968832708229537e-06, + "loss": 0.3862, + "step": 2930 + }, + { + "epoch": 0.06451683909311028, + "grad_norm": 1.6581827783054068, + "learning_rate": 9.968792957413344e-06, + "loss": 0.263, + "step": 2931 + }, + { + "epoch": 0.06453885097952894, + "grad_norm": 1.5390503547265153, + "learning_rate": 9.968753181343528e-06, + "loss": 0.3905, + "step": 2932 + }, + { + "epoch": 0.06456086286594762, + "grad_norm": 3.416851470733378, + "learning_rate": 9.968713380020294e-06, + "loss": 0.4901, + "step": 2933 + }, + { + "epoch": 0.06458287475236628, + "grad_norm": 1.6901185179766178, + "learning_rate": 9.968673553443841e-06, + "loss": 0.448, + "step": 2934 + }, + { + "epoch": 0.06460488663878494, + "grad_norm": 1.5945199721858934, + "learning_rate": 9.968633701614375e-06, + "loss": 0.2976, + "step": 2935 + }, + { + "epoch": 0.0646268985252036, + "grad_norm": 1.7820476577171742, + "learning_rate": 9.968593824532095e-06, + "loss": 0.3538, + "step": 2936 + }, + { + "epoch": 0.06464891041162228, + "grad_norm": 1.4669031014017713, + "learning_rate": 9.968553922197205e-06, + "loss": 0.2842, + "step": 2937 + }, + { + "epoch": 0.06467092229804094, + "grad_norm": 0.9154023645219308, + "learning_rate": 9.968513994609909e-06, + "loss": 0.5785, + "step": 2938 + }, + { + "epoch": 0.06469293418445961, + "grad_norm": 0.8110310347393644, + "learning_rate": 9.968474041770408e-06, + "loss": 0.5873, + "step": 2939 + }, + { + "epoch": 0.06471494607087827, + "grad_norm": 1.4131129811830634, + "learning_rate": 9.968434063678906e-06, + "loss": 0.4437, + "step": 2940 + }, + { + "epoch": 0.06473695795729693, + "grad_norm": 1.6827860991102461, + "learning_rate": 9.968394060335607e-06, + "loss": 0.3333, + "step": 2941 + }, + { + "epoch": 0.06475896984371561, + "grad_norm": 1.6999120381437056, + "learning_rate": 9.968354031740712e-06, + "loss": 0.3791, + "step": 2942 + }, + { + "epoch": 0.06478098173013427, + "grad_norm": 1.7015114946930714, + "learning_rate": 9.968313977894426e-06, + "loss": 0.3649, + "step": 2943 + }, + { + "epoch": 0.06480299361655294, + "grad_norm": 1.6619995951253899, + "learning_rate": 9.968273898796952e-06, + "loss": 0.3773, + "step": 2944 + }, + { + "epoch": 0.0648250055029716, + "grad_norm": 2.0970101644318997, + "learning_rate": 9.968233794448496e-06, + "loss": 0.4086, + "step": 2945 + }, + { + "epoch": 0.06484701738939028, + "grad_norm": 1.6974116536012125, + "learning_rate": 9.968193664849259e-06, + "loss": 0.315, + "step": 2946 + }, + { + "epoch": 0.06486902927580894, + "grad_norm": 1.6054127225591728, + "learning_rate": 9.968153509999446e-06, + "loss": 0.3574, + "step": 2947 + }, + { + "epoch": 0.0648910411622276, + "grad_norm": 1.1789024736535552, + "learning_rate": 9.968113329899261e-06, + "loss": 0.6328, + "step": 2948 + }, + { + "epoch": 0.06491305304864627, + "grad_norm": 1.6971739647465238, + "learning_rate": 9.968073124548907e-06, + "loss": 0.3748, + "step": 2949 + }, + { + "epoch": 0.06493506493506493, + "grad_norm": 1.6073757396280604, + "learning_rate": 9.968032893948592e-06, + "loss": 0.3534, + "step": 2950 + }, + { + "epoch": 0.0649570768214836, + "grad_norm": 1.791268179697355, + "learning_rate": 9.967992638098517e-06, + "loss": 0.2817, + "step": 2951 + }, + { + "epoch": 0.06497908870790227, + "grad_norm": 2.755130710636671, + "learning_rate": 9.967952356998885e-06, + "loss": 0.5025, + "step": 2952 + }, + { + "epoch": 0.06500110059432093, + "grad_norm": 1.7380872984038702, + "learning_rate": 9.967912050649907e-06, + "loss": 0.3535, + "step": 2953 + }, + { + "epoch": 0.0650231124807396, + "grad_norm": 1.5276925420559575, + "learning_rate": 9.967871719051782e-06, + "loss": 0.3798, + "step": 2954 + }, + { + "epoch": 0.06504512436715827, + "grad_norm": 1.3989604827811195, + "learning_rate": 9.967831362204718e-06, + "loss": 0.3957, + "step": 2955 + }, + { + "epoch": 0.06506713625357693, + "grad_norm": 2.27599390646345, + "learning_rate": 9.96779098010892e-06, + "loss": 0.3959, + "step": 2956 + }, + { + "epoch": 0.0650891481399956, + "grad_norm": 1.8855080968897264, + "learning_rate": 9.967750572764592e-06, + "loss": 0.4151, + "step": 2957 + }, + { + "epoch": 0.06511116002641426, + "grad_norm": 0.8650354714168104, + "learning_rate": 9.96771014017194e-06, + "loss": 0.6154, + "step": 2958 + }, + { + "epoch": 0.06513317191283292, + "grad_norm": 1.443411566197045, + "learning_rate": 9.96766968233117e-06, + "loss": 0.3841, + "step": 2959 + }, + { + "epoch": 0.0651551837992516, + "grad_norm": 2.0591873915481087, + "learning_rate": 9.967629199242485e-06, + "loss": 0.4114, + "step": 2960 + }, + { + "epoch": 0.06517719568567026, + "grad_norm": 0.6824850897058831, + "learning_rate": 9.967588690906094e-06, + "loss": 0.5902, + "step": 2961 + }, + { + "epoch": 0.06519920757208893, + "grad_norm": 1.2539955162739125, + "learning_rate": 9.967548157322203e-06, + "loss": 0.3045, + "step": 2962 + }, + { + "epoch": 0.06522121945850759, + "grad_norm": 0.6312740897518465, + "learning_rate": 9.967507598491014e-06, + "loss": 0.6038, + "step": 2963 + }, + { + "epoch": 0.06524323134492627, + "grad_norm": 4.451672386443938, + "learning_rate": 9.967467014412737e-06, + "loss": 0.3965, + "step": 2964 + }, + { + "epoch": 0.06526524323134493, + "grad_norm": 1.4398770992734213, + "learning_rate": 9.967426405087576e-06, + "loss": 0.3396, + "step": 2965 + }, + { + "epoch": 0.06528725511776359, + "grad_norm": 1.556608224363515, + "learning_rate": 9.967385770515741e-06, + "loss": 0.3636, + "step": 2966 + }, + { + "epoch": 0.06530926700418226, + "grad_norm": 1.6872363462669657, + "learning_rate": 9.967345110697435e-06, + "loss": 0.4103, + "step": 2967 + }, + { + "epoch": 0.06533127889060092, + "grad_norm": 0.8293840076107488, + "learning_rate": 9.967304425632865e-06, + "loss": 0.5742, + "step": 2968 + }, + { + "epoch": 0.0653532907770196, + "grad_norm": 0.8181829124423107, + "learning_rate": 9.96726371532224e-06, + "loss": 0.5783, + "step": 2969 + }, + { + "epoch": 0.06537530266343826, + "grad_norm": 1.7465494559184365, + "learning_rate": 9.967222979765763e-06, + "loss": 0.384, + "step": 2970 + }, + { + "epoch": 0.06539731454985692, + "grad_norm": 2.023793155634355, + "learning_rate": 9.967182218963645e-06, + "loss": 0.4319, + "step": 2971 + }, + { + "epoch": 0.06541932643627558, + "grad_norm": 2.0625040653028344, + "learning_rate": 9.967141432916092e-06, + "loss": 0.3242, + "step": 2972 + }, + { + "epoch": 0.06544133832269425, + "grad_norm": 0.7209364192968182, + "learning_rate": 9.967100621623309e-06, + "loss": 0.5902, + "step": 2973 + }, + { + "epoch": 0.06546335020911292, + "grad_norm": 3.426067668376198, + "learning_rate": 9.967059785085505e-06, + "loss": 0.3418, + "step": 2974 + }, + { + "epoch": 0.06548536209553159, + "grad_norm": 1.60626258305054, + "learning_rate": 9.96701892330289e-06, + "loss": 0.4001, + "step": 2975 + }, + { + "epoch": 0.06550737398195025, + "grad_norm": 2.702568256100641, + "learning_rate": 9.966978036275668e-06, + "loss": 0.3372, + "step": 2976 + }, + { + "epoch": 0.06552938586836891, + "grad_norm": 0.7714431179747998, + "learning_rate": 9.966937124004049e-06, + "loss": 0.601, + "step": 2977 + }, + { + "epoch": 0.06555139775478759, + "grad_norm": 1.4817676186507682, + "learning_rate": 9.966896186488239e-06, + "loss": 0.3453, + "step": 2978 + }, + { + "epoch": 0.06557340964120625, + "grad_norm": 2.6415905371841375, + "learning_rate": 9.966855223728448e-06, + "loss": 0.3702, + "step": 2979 + }, + { + "epoch": 0.06559542152762492, + "grad_norm": 1.6640743025106886, + "learning_rate": 9.966814235724884e-06, + "loss": 0.3617, + "step": 2980 + }, + { + "epoch": 0.06561743341404358, + "grad_norm": 1.5867260799038612, + "learning_rate": 9.966773222477754e-06, + "loss": 0.4179, + "step": 2981 + }, + { + "epoch": 0.06563944530046224, + "grad_norm": 1.6763094235349347, + "learning_rate": 9.966732183987269e-06, + "loss": 0.456, + "step": 2982 + }, + { + "epoch": 0.06566145718688092, + "grad_norm": 2.1752379765157475, + "learning_rate": 9.966691120253634e-06, + "loss": 0.386, + "step": 2983 + }, + { + "epoch": 0.06568346907329958, + "grad_norm": 2.5840729462571246, + "learning_rate": 9.966650031277061e-06, + "loss": 0.3799, + "step": 2984 + }, + { + "epoch": 0.06570548095971825, + "grad_norm": 1.5056467654463488, + "learning_rate": 9.966608917057756e-06, + "loss": 0.3983, + "step": 2985 + }, + { + "epoch": 0.06572749284613691, + "grad_norm": 1.3951554141862716, + "learning_rate": 9.96656777759593e-06, + "loss": 0.2341, + "step": 2986 + }, + { + "epoch": 0.06574950473255559, + "grad_norm": 3.667495100220499, + "learning_rate": 9.96652661289179e-06, + "loss": 0.4027, + "step": 2987 + }, + { + "epoch": 0.06577151661897425, + "grad_norm": 2.0081141744700135, + "learning_rate": 9.966485422945549e-06, + "loss": 0.4667, + "step": 2988 + }, + { + "epoch": 0.06579352850539291, + "grad_norm": 2.0670271899862023, + "learning_rate": 9.966444207757413e-06, + "loss": 0.3368, + "step": 2989 + }, + { + "epoch": 0.06581554039181158, + "grad_norm": 1.3431119740020898, + "learning_rate": 9.966402967327591e-06, + "loss": 0.3757, + "step": 2990 + }, + { + "epoch": 0.06583755227823024, + "grad_norm": 2.2860310109384208, + "learning_rate": 9.966361701656295e-06, + "loss": 0.3564, + "step": 2991 + }, + { + "epoch": 0.06585956416464891, + "grad_norm": 0.7864824723300045, + "learning_rate": 9.966320410743735e-06, + "loss": 0.5803, + "step": 2992 + }, + { + "epoch": 0.06588157605106758, + "grad_norm": 1.4883724814851107, + "learning_rate": 9.966279094590118e-06, + "loss": 0.3236, + "step": 2993 + }, + { + "epoch": 0.06590358793748624, + "grad_norm": 1.7190098572984684, + "learning_rate": 9.966237753195655e-06, + "loss": 0.3724, + "step": 2994 + }, + { + "epoch": 0.0659255998239049, + "grad_norm": 1.7765768127060584, + "learning_rate": 9.966196386560559e-06, + "loss": 0.3295, + "step": 2995 + }, + { + "epoch": 0.06594761171032358, + "grad_norm": 1.877268318288621, + "learning_rate": 9.966154994685037e-06, + "loss": 0.4365, + "step": 2996 + }, + { + "epoch": 0.06596962359674224, + "grad_norm": 1.5248357462766409, + "learning_rate": 9.9661135775693e-06, + "loss": 0.419, + "step": 2997 + }, + { + "epoch": 0.06599163548316091, + "grad_norm": 2.6091135589441796, + "learning_rate": 9.96607213521356e-06, + "loss": 0.4064, + "step": 2998 + }, + { + "epoch": 0.06601364736957957, + "grad_norm": 1.3302928344975027, + "learning_rate": 9.966030667618024e-06, + "loss": 0.3242, + "step": 2999 + }, + { + "epoch": 0.06603565925599823, + "grad_norm": 2.906863941259508, + "learning_rate": 9.965989174782908e-06, + "loss": 0.3278, + "step": 3000 + }, + { + "epoch": 0.06605767114241691, + "grad_norm": 1.5731864508903761, + "learning_rate": 9.96594765670842e-06, + "loss": 0.3761, + "step": 3001 + }, + { + "epoch": 0.06607968302883557, + "grad_norm": 1.345073382100886, + "learning_rate": 9.96590611339477e-06, + "loss": 0.4068, + "step": 3002 + }, + { + "epoch": 0.06610169491525424, + "grad_norm": 1.4628023632985714, + "learning_rate": 9.96586454484217e-06, + "loss": 0.3416, + "step": 3003 + }, + { + "epoch": 0.0661237068016729, + "grad_norm": 2.824036272534656, + "learning_rate": 9.965822951050832e-06, + "loss": 0.4797, + "step": 3004 + }, + { + "epoch": 0.06614571868809158, + "grad_norm": 0.7477034688053364, + "learning_rate": 9.965781332020966e-06, + "loss": 0.5605, + "step": 3005 + }, + { + "epoch": 0.06616773057451024, + "grad_norm": 0.6785186255265921, + "learning_rate": 9.965739687752786e-06, + "loss": 0.5995, + "step": 3006 + }, + { + "epoch": 0.0661897424609289, + "grad_norm": 2.3661209916440895, + "learning_rate": 9.9656980182465e-06, + "loss": 0.4493, + "step": 3007 + }, + { + "epoch": 0.06621175434734757, + "grad_norm": 7.653033383274193, + "learning_rate": 9.965656323502325e-06, + "loss": 0.3968, + "step": 3008 + }, + { + "epoch": 0.06623376623376623, + "grad_norm": 4.2787125958772005, + "learning_rate": 9.965614603520468e-06, + "loss": 0.4504, + "step": 3009 + }, + { + "epoch": 0.0662557781201849, + "grad_norm": 1.7867391744260517, + "learning_rate": 9.965572858301142e-06, + "loss": 0.3543, + "step": 3010 + }, + { + "epoch": 0.06627779000660357, + "grad_norm": 0.6901212946882038, + "learning_rate": 9.965531087844562e-06, + "loss": 0.5747, + "step": 3011 + }, + { + "epoch": 0.06629980189302223, + "grad_norm": 1.7160616948236282, + "learning_rate": 9.965489292150936e-06, + "loss": 0.4323, + "step": 3012 + }, + { + "epoch": 0.0663218137794409, + "grad_norm": 1.7365374592066791, + "learning_rate": 9.965447471220481e-06, + "loss": 0.3699, + "step": 3013 + }, + { + "epoch": 0.06634382566585957, + "grad_norm": 2.336844178145809, + "learning_rate": 9.965405625053406e-06, + "loss": 0.313, + "step": 3014 + }, + { + "epoch": 0.06636583755227823, + "grad_norm": 1.6893767357035652, + "learning_rate": 9.965363753649926e-06, + "loss": 0.3803, + "step": 3015 + }, + { + "epoch": 0.0663878494386969, + "grad_norm": 2.6913541118943822, + "learning_rate": 9.965321857010253e-06, + "loss": 0.4224, + "step": 3016 + }, + { + "epoch": 0.06640986132511556, + "grad_norm": 2.7899828464223733, + "learning_rate": 9.965279935134598e-06, + "loss": 0.3578, + "step": 3017 + }, + { + "epoch": 0.06643187321153422, + "grad_norm": 1.7108729494459691, + "learning_rate": 9.965237988023177e-06, + "loss": 0.2912, + "step": 3018 + }, + { + "epoch": 0.0664538850979529, + "grad_norm": 2.066597533572751, + "learning_rate": 9.965196015676201e-06, + "loss": 0.4518, + "step": 3019 + }, + { + "epoch": 0.06647589698437156, + "grad_norm": 1.4957648743143312, + "learning_rate": 9.965154018093886e-06, + "loss": 0.4336, + "step": 3020 + }, + { + "epoch": 0.06649790887079023, + "grad_norm": 1.4282413036905737, + "learning_rate": 9.965111995276441e-06, + "loss": 0.3564, + "step": 3021 + }, + { + "epoch": 0.06651992075720889, + "grad_norm": 1.432884717134926, + "learning_rate": 9.965069947224086e-06, + "loss": 0.4009, + "step": 3022 + }, + { + "epoch": 0.06654193264362755, + "grad_norm": 5.101366441540157, + "learning_rate": 9.965027873937029e-06, + "loss": 0.4996, + "step": 3023 + }, + { + "epoch": 0.06656394453004623, + "grad_norm": 1.9469576012400125, + "learning_rate": 9.964985775415486e-06, + "loss": 0.4423, + "step": 3024 + }, + { + "epoch": 0.06658595641646489, + "grad_norm": 2.0165040271157806, + "learning_rate": 9.964943651659671e-06, + "loss": 0.4682, + "step": 3025 + }, + { + "epoch": 0.06660796830288356, + "grad_norm": 2.1774974942329606, + "learning_rate": 9.964901502669798e-06, + "loss": 0.3772, + "step": 3026 + }, + { + "epoch": 0.06662998018930222, + "grad_norm": 0.7419515750553118, + "learning_rate": 9.96485932844608e-06, + "loss": 0.6093, + "step": 3027 + }, + { + "epoch": 0.0666519920757209, + "grad_norm": 0.7068905466307824, + "learning_rate": 9.964817128988733e-06, + "loss": 0.5997, + "step": 3028 + }, + { + "epoch": 0.06667400396213956, + "grad_norm": 1.5317003509822005, + "learning_rate": 9.964774904297972e-06, + "loss": 0.4207, + "step": 3029 + }, + { + "epoch": 0.06669601584855822, + "grad_norm": 2.8952469692755836, + "learning_rate": 9.96473265437401e-06, + "loss": 0.3593, + "step": 3030 + }, + { + "epoch": 0.06671802773497688, + "grad_norm": 1.7991230140146128, + "learning_rate": 9.964690379217063e-06, + "loss": 0.3872, + "step": 3031 + }, + { + "epoch": 0.06674003962139555, + "grad_norm": 1.4038404096784154, + "learning_rate": 9.964648078827343e-06, + "loss": 0.3851, + "step": 3032 + }, + { + "epoch": 0.06676205150781422, + "grad_norm": 1.4375338001226508, + "learning_rate": 9.96460575320507e-06, + "loss": 0.3687, + "step": 3033 + }, + { + "epoch": 0.06678406339423289, + "grad_norm": 0.7273069815935675, + "learning_rate": 9.964563402350455e-06, + "loss": 0.6131, + "step": 3034 + }, + { + "epoch": 0.06680607528065155, + "grad_norm": 1.7272692391518398, + "learning_rate": 9.964521026263715e-06, + "loss": 0.3778, + "step": 3035 + }, + { + "epoch": 0.06682808716707021, + "grad_norm": 1.685200219489452, + "learning_rate": 9.964478624945065e-06, + "loss": 0.3431, + "step": 3036 + }, + { + "epoch": 0.06685009905348889, + "grad_norm": 3.657864771638454, + "learning_rate": 9.964436198394719e-06, + "loss": 0.3772, + "step": 3037 + }, + { + "epoch": 0.06687211093990755, + "grad_norm": 1.7553973655145194, + "learning_rate": 9.964393746612895e-06, + "loss": 0.4111, + "step": 3038 + }, + { + "epoch": 0.06689412282632622, + "grad_norm": 1.6332123562461718, + "learning_rate": 9.964351269599808e-06, + "loss": 0.2565, + "step": 3039 + }, + { + "epoch": 0.06691613471274488, + "grad_norm": 1.4942524892082671, + "learning_rate": 9.964308767355672e-06, + "loss": 0.4015, + "step": 3040 + }, + { + "epoch": 0.06693814659916354, + "grad_norm": 0.6499221908832898, + "learning_rate": 9.964266239880707e-06, + "loss": 0.5566, + "step": 3041 + }, + { + "epoch": 0.06696015848558222, + "grad_norm": 1.5315372068696085, + "learning_rate": 9.964223687175126e-06, + "loss": 0.4043, + "step": 3042 + }, + { + "epoch": 0.06698217037200088, + "grad_norm": 1.6200566350730852, + "learning_rate": 9.964181109239146e-06, + "loss": 0.3858, + "step": 3043 + }, + { + "epoch": 0.06700418225841955, + "grad_norm": 1.8736485388071376, + "learning_rate": 9.964138506072983e-06, + "loss": 0.4302, + "step": 3044 + }, + { + "epoch": 0.06702619414483821, + "grad_norm": 1.61006162500728, + "learning_rate": 9.964095877676854e-06, + "loss": 0.3466, + "step": 3045 + }, + { + "epoch": 0.06704820603125689, + "grad_norm": 1.683350859828072, + "learning_rate": 9.964053224050978e-06, + "loss": 0.3659, + "step": 3046 + }, + { + "epoch": 0.06707021791767555, + "grad_norm": 2.464918102417413, + "learning_rate": 9.964010545195567e-06, + "loss": 0.486, + "step": 3047 + }, + { + "epoch": 0.06709222980409421, + "grad_norm": 1.4883670630416905, + "learning_rate": 9.963967841110841e-06, + "loss": 0.4179, + "step": 3048 + }, + { + "epoch": 0.06711424169051287, + "grad_norm": 0.7137839377083143, + "learning_rate": 9.963925111797017e-06, + "loss": 0.6036, + "step": 3049 + }, + { + "epoch": 0.06713625357693154, + "grad_norm": 1.7257521178390476, + "learning_rate": 9.963882357254311e-06, + "loss": 0.4021, + "step": 3050 + }, + { + "epoch": 0.06715826546335021, + "grad_norm": 1.6359718781203232, + "learning_rate": 9.963839577482941e-06, + "loss": 0.4881, + "step": 3051 + }, + { + "epoch": 0.06718027734976888, + "grad_norm": 1.7996962918104225, + "learning_rate": 9.963796772483125e-06, + "loss": 0.4429, + "step": 3052 + }, + { + "epoch": 0.06720228923618754, + "grad_norm": 0.6595563748163102, + "learning_rate": 9.963753942255079e-06, + "loss": 0.5574, + "step": 3053 + }, + { + "epoch": 0.0672243011226062, + "grad_norm": 1.8192379308747912, + "learning_rate": 9.963711086799021e-06, + "loss": 0.3757, + "step": 3054 + }, + { + "epoch": 0.06724631300902488, + "grad_norm": 1.7853321430347724, + "learning_rate": 9.963668206115171e-06, + "loss": 0.403, + "step": 3055 + }, + { + "epoch": 0.06726832489544354, + "grad_norm": 1.8168072962580906, + "learning_rate": 9.963625300203746e-06, + "loss": 0.4263, + "step": 3056 + }, + { + "epoch": 0.0672903367818622, + "grad_norm": 1.7433739254791991, + "learning_rate": 9.963582369064962e-06, + "loss": 0.4102, + "step": 3057 + }, + { + "epoch": 0.06731234866828087, + "grad_norm": 0.6826643211698901, + "learning_rate": 9.963539412699037e-06, + "loss": 0.5967, + "step": 3058 + }, + { + "epoch": 0.06733436055469953, + "grad_norm": 0.6595790031041705, + "learning_rate": 9.963496431106193e-06, + "loss": 0.5698, + "step": 3059 + }, + { + "epoch": 0.06735637244111821, + "grad_norm": 3.1963181327857515, + "learning_rate": 9.963453424286647e-06, + "loss": 0.3783, + "step": 3060 + }, + { + "epoch": 0.06737838432753687, + "grad_norm": 1.514228503825723, + "learning_rate": 9.963410392240616e-06, + "loss": 0.3732, + "step": 3061 + }, + { + "epoch": 0.06740039621395554, + "grad_norm": 1.9736917931646325, + "learning_rate": 9.963367334968318e-06, + "loss": 0.3638, + "step": 3062 + }, + { + "epoch": 0.0674224081003742, + "grad_norm": 1.383325099858887, + "learning_rate": 9.963324252469975e-06, + "loss": 0.2826, + "step": 3063 + }, + { + "epoch": 0.06744441998679286, + "grad_norm": 2.249972467954859, + "learning_rate": 9.963281144745804e-06, + "loss": 0.4028, + "step": 3064 + }, + { + "epoch": 0.06746643187321154, + "grad_norm": 1.9329660282230459, + "learning_rate": 9.963238011796025e-06, + "loss": 0.3542, + "step": 3065 + }, + { + "epoch": 0.0674884437596302, + "grad_norm": 2.104294272538937, + "learning_rate": 9.963194853620856e-06, + "loss": 0.3796, + "step": 3066 + }, + { + "epoch": 0.06751045564604886, + "grad_norm": 1.4181667310165669, + "learning_rate": 9.963151670220519e-06, + "loss": 0.2577, + "step": 3067 + }, + { + "epoch": 0.06753246753246753, + "grad_norm": 1.6109124801665515, + "learning_rate": 9.963108461595231e-06, + "loss": 0.4129, + "step": 3068 + }, + { + "epoch": 0.0675544794188862, + "grad_norm": 2.5117586760682533, + "learning_rate": 9.96306522774521e-06, + "loss": 0.4163, + "step": 3069 + }, + { + "epoch": 0.06757649130530487, + "grad_norm": 2.012894226412358, + "learning_rate": 9.96302196867068e-06, + "loss": 0.4466, + "step": 3070 + }, + { + "epoch": 0.06759850319172353, + "grad_norm": 1.5722055168341509, + "learning_rate": 9.962978684371858e-06, + "loss": 0.4665, + "step": 3071 + }, + { + "epoch": 0.0676205150781422, + "grad_norm": 2.348843165368727, + "learning_rate": 9.962935374848966e-06, + "loss": 0.2989, + "step": 3072 + }, + { + "epoch": 0.06764252696456086, + "grad_norm": 1.7430575045442385, + "learning_rate": 9.962892040102221e-06, + "loss": 0.3305, + "step": 3073 + }, + { + "epoch": 0.06766453885097953, + "grad_norm": 1.5628568907543667, + "learning_rate": 9.962848680131848e-06, + "loss": 0.2969, + "step": 3074 + }, + { + "epoch": 0.0676865507373982, + "grad_norm": 1.656987765656548, + "learning_rate": 9.96280529493806e-06, + "loss": 0.355, + "step": 3075 + }, + { + "epoch": 0.06770856262381686, + "grad_norm": 1.5824898474793119, + "learning_rate": 9.962761884521086e-06, + "loss": 0.4109, + "step": 3076 + }, + { + "epoch": 0.06773057451023552, + "grad_norm": 2.216151699315149, + "learning_rate": 9.96271844888114e-06, + "loss": 0.3829, + "step": 3077 + }, + { + "epoch": 0.0677525863966542, + "grad_norm": 0.8460344719969777, + "learning_rate": 9.962674988018448e-06, + "loss": 0.5817, + "step": 3078 + }, + { + "epoch": 0.06777459828307286, + "grad_norm": 0.7144756543052228, + "learning_rate": 9.962631501933227e-06, + "loss": 0.5606, + "step": 3079 + }, + { + "epoch": 0.06779661016949153, + "grad_norm": 1.6398672464808839, + "learning_rate": 9.962587990625699e-06, + "loss": 0.4182, + "step": 3080 + }, + { + "epoch": 0.06781862205591019, + "grad_norm": 2.257636065560109, + "learning_rate": 9.962544454096086e-06, + "loss": 0.3249, + "step": 3081 + }, + { + "epoch": 0.06784063394232885, + "grad_norm": 1.6431788771727478, + "learning_rate": 9.962500892344607e-06, + "loss": 0.358, + "step": 3082 + }, + { + "epoch": 0.06786264582874753, + "grad_norm": 2.15188444104973, + "learning_rate": 9.962457305371487e-06, + "loss": 0.3455, + "step": 3083 + }, + { + "epoch": 0.06788465771516619, + "grad_norm": 1.6883563581901093, + "learning_rate": 9.962413693176946e-06, + "loss": 0.4271, + "step": 3084 + }, + { + "epoch": 0.06790666960158485, + "grad_norm": 1.7924079154492554, + "learning_rate": 9.962370055761203e-06, + "loss": 0.3951, + "step": 3085 + }, + { + "epoch": 0.06792868148800352, + "grad_norm": 1.3892571448772173, + "learning_rate": 9.962326393124485e-06, + "loss": 0.3783, + "step": 3086 + }, + { + "epoch": 0.0679506933744222, + "grad_norm": 1.5408006575172908, + "learning_rate": 9.96228270526701e-06, + "loss": 0.33, + "step": 3087 + }, + { + "epoch": 0.06797270526084086, + "grad_norm": 2.1334593089730487, + "learning_rate": 9.962238992189e-06, + "loss": 0.4614, + "step": 3088 + }, + { + "epoch": 0.06799471714725952, + "grad_norm": 3.338788452623844, + "learning_rate": 9.962195253890678e-06, + "loss": 0.3402, + "step": 3089 + }, + { + "epoch": 0.06801672903367818, + "grad_norm": 1.9129994652242135, + "learning_rate": 9.962151490372268e-06, + "loss": 0.3614, + "step": 3090 + }, + { + "epoch": 0.06803874092009685, + "grad_norm": 1.628589627207251, + "learning_rate": 9.96210770163399e-06, + "loss": 0.3497, + "step": 3091 + }, + { + "epoch": 0.06806075280651552, + "grad_norm": 1.9656024355473734, + "learning_rate": 9.96206388767607e-06, + "loss": 0.3647, + "step": 3092 + }, + { + "epoch": 0.06808276469293419, + "grad_norm": 6.644449606909439, + "learning_rate": 9.962020048498723e-06, + "loss": 0.3584, + "step": 3093 + }, + { + "epoch": 0.06810477657935285, + "grad_norm": 1.8804573907928928, + "learning_rate": 9.961976184102182e-06, + "loss": 0.3866, + "step": 3094 + }, + { + "epoch": 0.06812678846577151, + "grad_norm": 1.9530124168064253, + "learning_rate": 9.961932294486662e-06, + "loss": 0.4645, + "step": 3095 + }, + { + "epoch": 0.06814880035219019, + "grad_norm": 1.8120728742419268, + "learning_rate": 9.961888379652392e-06, + "loss": 0.3656, + "step": 3096 + }, + { + "epoch": 0.06817081223860885, + "grad_norm": 1.7722538266084082, + "learning_rate": 9.961844439599588e-06, + "loss": 0.3447, + "step": 3097 + }, + { + "epoch": 0.06819282412502752, + "grad_norm": 1.7175793686970262, + "learning_rate": 9.961800474328481e-06, + "loss": 0.3566, + "step": 3098 + }, + { + "epoch": 0.06821483601144618, + "grad_norm": 1.4409087627634605, + "learning_rate": 9.961756483839288e-06, + "loss": 0.3625, + "step": 3099 + }, + { + "epoch": 0.06823684789786484, + "grad_norm": 1.8922020196434033, + "learning_rate": 9.96171246813224e-06, + "loss": 0.2876, + "step": 3100 + }, + { + "epoch": 0.06825885978428352, + "grad_norm": 1.380103323784661, + "learning_rate": 9.96166842720755e-06, + "loss": 0.6192, + "step": 3101 + }, + { + "epoch": 0.06828087167070218, + "grad_norm": 1.4067360245856542, + "learning_rate": 9.961624361065455e-06, + "loss": 0.4042, + "step": 3102 + }, + { + "epoch": 0.06830288355712084, + "grad_norm": 2.341036588693321, + "learning_rate": 9.961580269706166e-06, + "loss": 0.4029, + "step": 3103 + }, + { + "epoch": 0.06832489544353951, + "grad_norm": 1.6930111295495063, + "learning_rate": 9.961536153129916e-06, + "loss": 0.3997, + "step": 3104 + }, + { + "epoch": 0.06834690732995818, + "grad_norm": 1.3060465909541987, + "learning_rate": 9.961492011336927e-06, + "loss": 0.3899, + "step": 3105 + }, + { + "epoch": 0.06836891921637685, + "grad_norm": 5.704457902937552, + "learning_rate": 9.961447844327422e-06, + "loss": 0.4252, + "step": 3106 + }, + { + "epoch": 0.06839093110279551, + "grad_norm": 1.653392133709017, + "learning_rate": 9.961403652101625e-06, + "loss": 0.3413, + "step": 3107 + }, + { + "epoch": 0.06841294298921417, + "grad_norm": 0.9284495920059321, + "learning_rate": 9.961359434659763e-06, + "loss": 0.5756, + "step": 3108 + }, + { + "epoch": 0.06843495487563284, + "grad_norm": 2.958324416675198, + "learning_rate": 9.961315192002059e-06, + "loss": 0.3876, + "step": 3109 + }, + { + "epoch": 0.06845696676205151, + "grad_norm": 1.5513092692917843, + "learning_rate": 9.961270924128738e-06, + "loss": 0.4602, + "step": 3110 + }, + { + "epoch": 0.06847897864847018, + "grad_norm": 2.3789141433289176, + "learning_rate": 9.961226631040027e-06, + "loss": 0.265, + "step": 3111 + }, + { + "epoch": 0.06850099053488884, + "grad_norm": 2.7434989825985414, + "learning_rate": 9.961182312736148e-06, + "loss": 0.4504, + "step": 3112 + }, + { + "epoch": 0.0685230024213075, + "grad_norm": 1.675071360299008, + "learning_rate": 9.96113796921733e-06, + "loss": 0.4534, + "step": 3113 + }, + { + "epoch": 0.06854501430772617, + "grad_norm": 1.6885145517914335, + "learning_rate": 9.961093600483794e-06, + "loss": 0.3647, + "step": 3114 + }, + { + "epoch": 0.06856702619414484, + "grad_norm": 1.841892950155224, + "learning_rate": 9.961049206535768e-06, + "loss": 0.3409, + "step": 3115 + }, + { + "epoch": 0.0685890380805635, + "grad_norm": 2.295299622953022, + "learning_rate": 9.961004787373477e-06, + "loss": 0.3622, + "step": 3116 + }, + { + "epoch": 0.06861104996698217, + "grad_norm": 0.7966638587748035, + "learning_rate": 9.960960342997146e-06, + "loss": 0.6232, + "step": 3117 + }, + { + "epoch": 0.06863306185340083, + "grad_norm": 1.868213315203806, + "learning_rate": 9.960915873407004e-06, + "loss": 0.4313, + "step": 3118 + }, + { + "epoch": 0.06865507373981951, + "grad_norm": 0.663435387083987, + "learning_rate": 9.960871378603274e-06, + "loss": 0.5541, + "step": 3119 + }, + { + "epoch": 0.06867708562623817, + "grad_norm": 2.3632593055141755, + "learning_rate": 9.960826858586183e-06, + "loss": 0.5435, + "step": 3120 + }, + { + "epoch": 0.06869909751265683, + "grad_norm": 1.664443916873168, + "learning_rate": 9.960782313355958e-06, + "loss": 0.4294, + "step": 3121 + }, + { + "epoch": 0.0687211093990755, + "grad_norm": 1.8151692012077794, + "learning_rate": 9.960737742912824e-06, + "loss": 0.3951, + "step": 3122 + }, + { + "epoch": 0.06874312128549416, + "grad_norm": 1.7476466070427963, + "learning_rate": 9.960693147257008e-06, + "loss": 0.4185, + "step": 3123 + }, + { + "epoch": 0.06876513317191284, + "grad_norm": 0.727118688850306, + "learning_rate": 9.960648526388737e-06, + "loss": 0.5715, + "step": 3124 + }, + { + "epoch": 0.0687871450583315, + "grad_norm": 1.5410038755569702, + "learning_rate": 9.960603880308237e-06, + "loss": 0.3322, + "step": 3125 + }, + { + "epoch": 0.06880915694475016, + "grad_norm": 1.6863461637350938, + "learning_rate": 9.960559209015738e-06, + "loss": 0.3518, + "step": 3126 + }, + { + "epoch": 0.06883116883116883, + "grad_norm": 2.0021918875184914, + "learning_rate": 9.960514512511463e-06, + "loss": 0.3117, + "step": 3127 + }, + { + "epoch": 0.0688531807175875, + "grad_norm": 1.7568888998244154, + "learning_rate": 9.960469790795642e-06, + "loss": 0.3836, + "step": 3128 + }, + { + "epoch": 0.06887519260400617, + "grad_norm": 1.930125209850217, + "learning_rate": 9.9604250438685e-06, + "loss": 0.4498, + "step": 3129 + }, + { + "epoch": 0.06889720449042483, + "grad_norm": 1.5816216117451003, + "learning_rate": 9.960380271730264e-06, + "loss": 0.366, + "step": 3130 + }, + { + "epoch": 0.06891921637684349, + "grad_norm": 1.5773494248404354, + "learning_rate": 9.960335474381166e-06, + "loss": 0.4283, + "step": 3131 + }, + { + "epoch": 0.06894122826326216, + "grad_norm": 0.7114385701024747, + "learning_rate": 9.960290651821429e-06, + "loss": 0.5983, + "step": 3132 + }, + { + "epoch": 0.06896324014968083, + "grad_norm": 2.0930039789627144, + "learning_rate": 9.960245804051283e-06, + "loss": 0.4502, + "step": 3133 + }, + { + "epoch": 0.0689852520360995, + "grad_norm": 1.5520303423836455, + "learning_rate": 9.960200931070955e-06, + "loss": 0.3346, + "step": 3134 + }, + { + "epoch": 0.06900726392251816, + "grad_norm": 0.6947142959550098, + "learning_rate": 9.960156032880675e-06, + "loss": 0.59, + "step": 3135 + }, + { + "epoch": 0.06902927580893682, + "grad_norm": 1.4903471447402779, + "learning_rate": 9.960111109480668e-06, + "loss": 0.4327, + "step": 3136 + }, + { + "epoch": 0.0690512876953555, + "grad_norm": 1.846158906341071, + "learning_rate": 9.960066160871164e-06, + "loss": 0.3911, + "step": 3137 + }, + { + "epoch": 0.06907329958177416, + "grad_norm": 1.585827358471488, + "learning_rate": 9.960021187052393e-06, + "loss": 0.3885, + "step": 3138 + }, + { + "epoch": 0.06909531146819282, + "grad_norm": 1.5527221093373362, + "learning_rate": 9.95997618802458e-06, + "loss": 0.3881, + "step": 3139 + }, + { + "epoch": 0.06911732335461149, + "grad_norm": 1.720113133322067, + "learning_rate": 9.959931163787958e-06, + "loss": 0.4261, + "step": 3140 + }, + { + "epoch": 0.06913933524103015, + "grad_norm": 1.3401900000225975, + "learning_rate": 9.959886114342754e-06, + "loss": 0.4211, + "step": 3141 + }, + { + "epoch": 0.06916134712744883, + "grad_norm": 1.5362115495338289, + "learning_rate": 9.959841039689194e-06, + "loss": 0.4616, + "step": 3142 + }, + { + "epoch": 0.06918335901386749, + "grad_norm": 1.7361764739754357, + "learning_rate": 9.959795939827513e-06, + "loss": 0.4337, + "step": 3143 + }, + { + "epoch": 0.06920537090028615, + "grad_norm": 2.357441855010394, + "learning_rate": 9.959750814757933e-06, + "loss": 0.4042, + "step": 3144 + }, + { + "epoch": 0.06922738278670482, + "grad_norm": 0.7328184841320041, + "learning_rate": 9.95970566448069e-06, + "loss": 0.5859, + "step": 3145 + }, + { + "epoch": 0.0692493946731235, + "grad_norm": 0.717142127022777, + "learning_rate": 9.95966048899601e-06, + "loss": 0.583, + "step": 3146 + }, + { + "epoch": 0.06927140655954216, + "grad_norm": 2.2011172098726157, + "learning_rate": 9.959615288304123e-06, + "loss": 0.3459, + "step": 3147 + }, + { + "epoch": 0.06929341844596082, + "grad_norm": 2.988881540229902, + "learning_rate": 9.959570062405259e-06, + "loss": 0.3771, + "step": 3148 + }, + { + "epoch": 0.06931543033237948, + "grad_norm": 2.2986927853142394, + "learning_rate": 9.959524811299648e-06, + "loss": 0.3964, + "step": 3149 + }, + { + "epoch": 0.06933744221879815, + "grad_norm": 2.7113094294947495, + "learning_rate": 9.959479534987522e-06, + "loss": 0.4136, + "step": 3150 + }, + { + "epoch": 0.06935945410521682, + "grad_norm": 1.6910191834919466, + "learning_rate": 9.959434233469107e-06, + "loss": 0.3649, + "step": 3151 + }, + { + "epoch": 0.06938146599163549, + "grad_norm": 0.7170474473073758, + "learning_rate": 9.959388906744635e-06, + "loss": 0.5978, + "step": 3152 + }, + { + "epoch": 0.06940347787805415, + "grad_norm": 1.6634107975447263, + "learning_rate": 9.959343554814337e-06, + "loss": 0.3927, + "step": 3153 + }, + { + "epoch": 0.06942548976447281, + "grad_norm": 1.7684840808440936, + "learning_rate": 9.959298177678443e-06, + "loss": 0.3876, + "step": 3154 + }, + { + "epoch": 0.06944750165089149, + "grad_norm": 8.307427032136859, + "learning_rate": 9.959252775337185e-06, + "loss": 0.3392, + "step": 3155 + }, + { + "epoch": 0.06946951353731015, + "grad_norm": 7.4854497268329965, + "learning_rate": 9.95920734779079e-06, + "loss": 0.3014, + "step": 3156 + }, + { + "epoch": 0.06949152542372881, + "grad_norm": 1.795522760701882, + "learning_rate": 9.959161895039493e-06, + "loss": 0.3512, + "step": 3157 + }, + { + "epoch": 0.06951353731014748, + "grad_norm": 1.5586886507471838, + "learning_rate": 9.959116417083522e-06, + "loss": 0.3364, + "step": 3158 + }, + { + "epoch": 0.06953554919656614, + "grad_norm": 1.6796447770692235, + "learning_rate": 9.959070913923113e-06, + "loss": 0.4208, + "step": 3159 + }, + { + "epoch": 0.06955756108298482, + "grad_norm": 1.4458471809529476, + "learning_rate": 9.959025385558492e-06, + "loss": 0.3153, + "step": 3160 + }, + { + "epoch": 0.06957957296940348, + "grad_norm": 2.0612361827254873, + "learning_rate": 9.95897983198989e-06, + "loss": 0.4189, + "step": 3161 + }, + { + "epoch": 0.06960158485582214, + "grad_norm": 0.7205581586242623, + "learning_rate": 9.958934253217542e-06, + "loss": 0.5763, + "step": 3162 + }, + { + "epoch": 0.0696235967422408, + "grad_norm": 0.7061101070819178, + "learning_rate": 9.95888864924168e-06, + "loss": 0.5864, + "step": 3163 + }, + { + "epoch": 0.06964560862865947, + "grad_norm": 2.20505216490052, + "learning_rate": 9.958843020062534e-06, + "loss": 0.4117, + "step": 3164 + }, + { + "epoch": 0.06966762051507815, + "grad_norm": 2.0822893200341923, + "learning_rate": 9.958797365680334e-06, + "loss": 0.4392, + "step": 3165 + }, + { + "epoch": 0.06968963240149681, + "grad_norm": 0.6368276175082375, + "learning_rate": 9.958751686095316e-06, + "loss": 0.5706, + "step": 3166 + }, + { + "epoch": 0.06971164428791547, + "grad_norm": 2.295476164852415, + "learning_rate": 9.958705981307709e-06, + "loss": 0.3556, + "step": 3167 + }, + { + "epoch": 0.06973365617433414, + "grad_norm": 1.408329974475926, + "learning_rate": 9.958660251317747e-06, + "loss": 0.3713, + "step": 3168 + }, + { + "epoch": 0.06975566806075281, + "grad_norm": 0.6637734851070399, + "learning_rate": 9.958614496125662e-06, + "loss": 0.577, + "step": 3169 + }, + { + "epoch": 0.06977767994717148, + "grad_norm": 1.423654755635763, + "learning_rate": 9.958568715731687e-06, + "loss": 0.3673, + "step": 3170 + }, + { + "epoch": 0.06979969183359014, + "grad_norm": 1.8417327891934472, + "learning_rate": 9.958522910136053e-06, + "loss": 0.3594, + "step": 3171 + }, + { + "epoch": 0.0698217037200088, + "grad_norm": 1.8938196252954589, + "learning_rate": 9.958477079338994e-06, + "loss": 0.3574, + "step": 3172 + }, + { + "epoch": 0.06984371560642746, + "grad_norm": 1.840409902639635, + "learning_rate": 9.958431223340746e-06, + "loss": 0.4064, + "step": 3173 + }, + { + "epoch": 0.06986572749284614, + "grad_norm": 1.55225499209935, + "learning_rate": 9.958385342141537e-06, + "loss": 0.4553, + "step": 3174 + }, + { + "epoch": 0.0698877393792648, + "grad_norm": 1.7187111262292503, + "learning_rate": 9.958339435741601e-06, + "loss": 0.3912, + "step": 3175 + }, + { + "epoch": 0.06990975126568347, + "grad_norm": 1.5790645796395657, + "learning_rate": 9.958293504141173e-06, + "loss": 0.3768, + "step": 3176 + }, + { + "epoch": 0.06993176315210213, + "grad_norm": 4.720371130706193, + "learning_rate": 9.958247547340487e-06, + "loss": 0.3248, + "step": 3177 + }, + { + "epoch": 0.06995377503852081, + "grad_norm": 1.795773049520134, + "learning_rate": 9.958201565339774e-06, + "loss": 0.3279, + "step": 3178 + }, + { + "epoch": 0.06997578692493947, + "grad_norm": 2.606493546119422, + "learning_rate": 9.958155558139271e-06, + "loss": 0.3668, + "step": 3179 + }, + { + "epoch": 0.06999779881135813, + "grad_norm": 3.009182579573621, + "learning_rate": 9.958109525739208e-06, + "loss": 0.3767, + "step": 3180 + }, + { + "epoch": 0.0700198106977768, + "grad_norm": 3.1811957416206047, + "learning_rate": 9.958063468139822e-06, + "loss": 0.3272, + "step": 3181 + }, + { + "epoch": 0.07004182258419546, + "grad_norm": 0.7158972486682963, + "learning_rate": 9.958017385341348e-06, + "loss": 0.5865, + "step": 3182 + }, + { + "epoch": 0.07006383447061414, + "grad_norm": 1.7392842579700578, + "learning_rate": 9.957971277344016e-06, + "loss": 0.3436, + "step": 3183 + }, + { + "epoch": 0.0700858463570328, + "grad_norm": 1.938909437414825, + "learning_rate": 9.957925144148064e-06, + "loss": 0.3716, + "step": 3184 + }, + { + "epoch": 0.07010785824345146, + "grad_norm": 1.3985710378501592, + "learning_rate": 9.957878985753723e-06, + "loss": 0.3119, + "step": 3185 + }, + { + "epoch": 0.07012987012987013, + "grad_norm": 2.357361477088496, + "learning_rate": 9.957832802161232e-06, + "loss": 0.3742, + "step": 3186 + }, + { + "epoch": 0.0701518820162888, + "grad_norm": 3.671754750051423, + "learning_rate": 9.957786593370821e-06, + "loss": 0.401, + "step": 3187 + }, + { + "epoch": 0.07017389390270747, + "grad_norm": 2.8264581971756098, + "learning_rate": 9.957740359382731e-06, + "loss": 0.4733, + "step": 3188 + }, + { + "epoch": 0.07019590578912613, + "grad_norm": 0.7172098304582245, + "learning_rate": 9.95769410019719e-06, + "loss": 0.6111, + "step": 3189 + }, + { + "epoch": 0.07021791767554479, + "grad_norm": 1.6460595628825638, + "learning_rate": 9.957647815814438e-06, + "loss": 0.4319, + "step": 3190 + }, + { + "epoch": 0.07023992956196345, + "grad_norm": 2.626070173603464, + "learning_rate": 9.957601506234708e-06, + "loss": 0.4218, + "step": 3191 + }, + { + "epoch": 0.07026194144838213, + "grad_norm": 4.17626525517506, + "learning_rate": 9.957555171458235e-06, + "loss": 0.4194, + "step": 3192 + }, + { + "epoch": 0.0702839533348008, + "grad_norm": 2.0123068335600687, + "learning_rate": 9.957508811485257e-06, + "loss": 0.3584, + "step": 3193 + }, + { + "epoch": 0.07030596522121946, + "grad_norm": 8.16274936886683, + "learning_rate": 9.957462426316007e-06, + "loss": 0.4179, + "step": 3194 + }, + { + "epoch": 0.07032797710763812, + "grad_norm": 1.8089877345566645, + "learning_rate": 9.957416015950724e-06, + "loss": 0.3941, + "step": 3195 + }, + { + "epoch": 0.0703499889940568, + "grad_norm": 1.9621311302220106, + "learning_rate": 9.95736958038964e-06, + "loss": 0.4713, + "step": 3196 + }, + { + "epoch": 0.07037200088047546, + "grad_norm": 1.5734294514443248, + "learning_rate": 9.957323119632994e-06, + "loss": 0.3808, + "step": 3197 + }, + { + "epoch": 0.07039401276689412, + "grad_norm": 0.6900206013418173, + "learning_rate": 9.95727663368102e-06, + "loss": 0.5981, + "step": 3198 + }, + { + "epoch": 0.07041602465331279, + "grad_norm": 3.718337175083057, + "learning_rate": 9.957230122533953e-06, + "loss": 0.3696, + "step": 3199 + }, + { + "epoch": 0.07043803653973145, + "grad_norm": 0.6510496113552633, + "learning_rate": 9.957183586192035e-06, + "loss": 0.58, + "step": 3200 + }, + { + "epoch": 0.07046004842615013, + "grad_norm": 2.750970643064179, + "learning_rate": 9.957137024655497e-06, + "loss": 0.3502, + "step": 3201 + }, + { + "epoch": 0.07048206031256879, + "grad_norm": 2.684915179342636, + "learning_rate": 9.95709043792458e-06, + "loss": 0.3172, + "step": 3202 + }, + { + "epoch": 0.07050407219898745, + "grad_norm": 0.6480541965717915, + "learning_rate": 9.957043825999516e-06, + "loss": 0.5922, + "step": 3203 + }, + { + "epoch": 0.07052608408540612, + "grad_norm": 2.0057717131705304, + "learning_rate": 9.956997188880544e-06, + "loss": 0.3352, + "step": 3204 + }, + { + "epoch": 0.07054809597182478, + "grad_norm": 1.679336816788645, + "learning_rate": 9.956950526567903e-06, + "loss": 0.3792, + "step": 3205 + }, + { + "epoch": 0.07057010785824346, + "grad_norm": 2.456547791855945, + "learning_rate": 9.95690383906183e-06, + "loss": 0.4201, + "step": 3206 + }, + { + "epoch": 0.07059211974466212, + "grad_norm": 3.1722127338419224, + "learning_rate": 9.956857126362557e-06, + "loss": 0.3468, + "step": 3207 + }, + { + "epoch": 0.07061413163108078, + "grad_norm": 1.7553006396039559, + "learning_rate": 9.956810388470328e-06, + "loss": 0.4051, + "step": 3208 + }, + { + "epoch": 0.07063614351749944, + "grad_norm": 1.7215865661641505, + "learning_rate": 9.956763625385377e-06, + "loss": 0.3383, + "step": 3209 + }, + { + "epoch": 0.07065815540391812, + "grad_norm": 7.903646121962174, + "learning_rate": 9.956716837107943e-06, + "loss": 0.3605, + "step": 3210 + }, + { + "epoch": 0.07068016729033678, + "grad_norm": 0.6805230593085224, + "learning_rate": 9.956670023638262e-06, + "loss": 0.5699, + "step": 3211 + }, + { + "epoch": 0.07070217917675545, + "grad_norm": 2.1531724838545276, + "learning_rate": 9.956623184976575e-06, + "loss": 0.3497, + "step": 3212 + }, + { + "epoch": 0.07072419106317411, + "grad_norm": 1.8637987485002814, + "learning_rate": 9.956576321123118e-06, + "loss": 0.4162, + "step": 3213 + }, + { + "epoch": 0.07074620294959277, + "grad_norm": 1.6675310206150762, + "learning_rate": 9.956529432078128e-06, + "loss": 0.3556, + "step": 3214 + }, + { + "epoch": 0.07076821483601145, + "grad_norm": 0.6848231853595287, + "learning_rate": 9.956482517841845e-06, + "loss": 0.6078, + "step": 3215 + }, + { + "epoch": 0.07079022672243011, + "grad_norm": 0.6352083831053379, + "learning_rate": 9.956435578414508e-06, + "loss": 0.5741, + "step": 3216 + }, + { + "epoch": 0.07081223860884878, + "grad_norm": 2.6700020526860953, + "learning_rate": 9.956388613796354e-06, + "loss": 0.3243, + "step": 3217 + }, + { + "epoch": 0.07083425049526744, + "grad_norm": 3.573901873083517, + "learning_rate": 9.956341623987622e-06, + "loss": 0.3472, + "step": 3218 + }, + { + "epoch": 0.07085626238168612, + "grad_norm": 2.135932689720572, + "learning_rate": 9.956294608988552e-06, + "loss": 0.3907, + "step": 3219 + }, + { + "epoch": 0.07087827426810478, + "grad_norm": 1.86871870451779, + "learning_rate": 9.956247568799383e-06, + "loss": 0.4097, + "step": 3220 + }, + { + "epoch": 0.07090028615452344, + "grad_norm": 0.681559726257985, + "learning_rate": 9.956200503420352e-06, + "loss": 0.6089, + "step": 3221 + }, + { + "epoch": 0.0709222980409421, + "grad_norm": 2.072318768190809, + "learning_rate": 9.956153412851701e-06, + "loss": 0.3337, + "step": 3222 + }, + { + "epoch": 0.07094430992736077, + "grad_norm": 1.7021289403789155, + "learning_rate": 9.956106297093668e-06, + "loss": 0.4866, + "step": 3223 + }, + { + "epoch": 0.07096632181377945, + "grad_norm": 2.215574361068442, + "learning_rate": 9.956059156146491e-06, + "loss": 0.4683, + "step": 3224 + }, + { + "epoch": 0.07098833370019811, + "grad_norm": 1.851360662564937, + "learning_rate": 9.95601199001041e-06, + "loss": 0.384, + "step": 3225 + }, + { + "epoch": 0.07101034558661677, + "grad_norm": 0.6427382666598563, + "learning_rate": 9.955964798685668e-06, + "loss": 0.5651, + "step": 3226 + }, + { + "epoch": 0.07103235747303543, + "grad_norm": 0.6907234164931909, + "learning_rate": 9.955917582172501e-06, + "loss": 0.5951, + "step": 3227 + }, + { + "epoch": 0.07105436935945411, + "grad_norm": 1.8874979813473414, + "learning_rate": 9.955870340471151e-06, + "loss": 0.3507, + "step": 3228 + }, + { + "epoch": 0.07107638124587277, + "grad_norm": 2.2311492157788333, + "learning_rate": 9.955823073581857e-06, + "loss": 0.3733, + "step": 3229 + }, + { + "epoch": 0.07109839313229144, + "grad_norm": 2.0870603732878115, + "learning_rate": 9.955775781504861e-06, + "loss": 0.3518, + "step": 3230 + }, + { + "epoch": 0.0711204050187101, + "grad_norm": 2.654151691538057, + "learning_rate": 9.9557284642404e-06, + "loss": 0.375, + "step": 3231 + }, + { + "epoch": 0.07114241690512876, + "grad_norm": 1.7499326567428621, + "learning_rate": 9.955681121788718e-06, + "loss": 0.2979, + "step": 3232 + }, + { + "epoch": 0.07116442879154744, + "grad_norm": 1.841579891745201, + "learning_rate": 9.955633754150056e-06, + "loss": 0.3677, + "step": 3233 + }, + { + "epoch": 0.0711864406779661, + "grad_norm": 4.267069439552565, + "learning_rate": 9.95558636132465e-06, + "loss": 0.4301, + "step": 3234 + }, + { + "epoch": 0.07120845256438477, + "grad_norm": 1.5115953050159041, + "learning_rate": 9.955538943312745e-06, + "loss": 0.2879, + "step": 3235 + }, + { + "epoch": 0.07123046445080343, + "grad_norm": 2.1158979899632193, + "learning_rate": 9.95549150011458e-06, + "loss": 0.4084, + "step": 3236 + }, + { + "epoch": 0.0712524763372221, + "grad_norm": 1.937652531632013, + "learning_rate": 9.9554440317304e-06, + "loss": 0.3896, + "step": 3237 + }, + { + "epoch": 0.07127448822364077, + "grad_norm": 1.8532821489844393, + "learning_rate": 9.95539653816044e-06, + "loss": 0.3827, + "step": 3238 + }, + { + "epoch": 0.07129650011005943, + "grad_norm": 1.9070477670065273, + "learning_rate": 9.955349019404946e-06, + "loss": 0.4054, + "step": 3239 + }, + { + "epoch": 0.0713185119964781, + "grad_norm": 1.487001100327053, + "learning_rate": 9.955301475464158e-06, + "loss": 0.391, + "step": 3240 + }, + { + "epoch": 0.07134052388289676, + "grad_norm": 2.19091361061578, + "learning_rate": 9.955253906338317e-06, + "loss": 0.4839, + "step": 3241 + }, + { + "epoch": 0.07136253576931544, + "grad_norm": 1.8686498931026965, + "learning_rate": 9.955206312027667e-06, + "loss": 0.3016, + "step": 3242 + }, + { + "epoch": 0.0713845476557341, + "grad_norm": 1.5292935294042471, + "learning_rate": 9.955158692532448e-06, + "loss": 0.4279, + "step": 3243 + }, + { + "epoch": 0.07140655954215276, + "grad_norm": 2.0195778795958605, + "learning_rate": 9.955111047852902e-06, + "loss": 0.4362, + "step": 3244 + }, + { + "epoch": 0.07142857142857142, + "grad_norm": 2.470408086712498, + "learning_rate": 9.955063377989272e-06, + "loss": 0.3353, + "step": 3245 + }, + { + "epoch": 0.0714505833149901, + "grad_norm": 1.5595100123347867, + "learning_rate": 9.955015682941798e-06, + "loss": 0.3832, + "step": 3246 + }, + { + "epoch": 0.07147259520140876, + "grad_norm": 6.315182161477845, + "learning_rate": 9.954967962710726e-06, + "loss": 0.2642, + "step": 3247 + }, + { + "epoch": 0.07149460708782743, + "grad_norm": 1.6856334526044143, + "learning_rate": 9.954920217296297e-06, + "loss": 0.4272, + "step": 3248 + }, + { + "epoch": 0.07151661897424609, + "grad_norm": 11.375756310505757, + "learning_rate": 9.954872446698753e-06, + "loss": 0.3286, + "step": 3249 + }, + { + "epoch": 0.07153863086066475, + "grad_norm": 1.693364874677239, + "learning_rate": 9.954824650918337e-06, + "loss": 0.3618, + "step": 3250 + }, + { + "epoch": 0.07156064274708343, + "grad_norm": 1.2810150463558905, + "learning_rate": 9.954776829955292e-06, + "loss": 0.3136, + "step": 3251 + }, + { + "epoch": 0.0715826546335021, + "grad_norm": 14.822077804995187, + "learning_rate": 9.954728983809862e-06, + "loss": 0.3546, + "step": 3252 + }, + { + "epoch": 0.07160466651992076, + "grad_norm": 1.7609902570966613, + "learning_rate": 9.95468111248229e-06, + "loss": 0.4141, + "step": 3253 + }, + { + "epoch": 0.07162667840633942, + "grad_norm": 1.8696566408475246, + "learning_rate": 9.954633215972818e-06, + "loss": 0.3707, + "step": 3254 + }, + { + "epoch": 0.07164869029275808, + "grad_norm": 1.7817390885413051, + "learning_rate": 9.95458529428169e-06, + "loss": 0.3679, + "step": 3255 + }, + { + "epoch": 0.07167070217917676, + "grad_norm": 0.8378880640021344, + "learning_rate": 9.954537347409151e-06, + "loss": 0.5697, + "step": 3256 + }, + { + "epoch": 0.07169271406559542, + "grad_norm": 5.529081624301917, + "learning_rate": 9.954489375355443e-06, + "loss": 0.3456, + "step": 3257 + }, + { + "epoch": 0.07171472595201409, + "grad_norm": 0.6901719407813631, + "learning_rate": 9.95444137812081e-06, + "loss": 0.6046, + "step": 3258 + }, + { + "epoch": 0.07173673783843275, + "grad_norm": 1.834985248361653, + "learning_rate": 9.954393355705496e-06, + "loss": 0.3219, + "step": 3259 + }, + { + "epoch": 0.07175874972485143, + "grad_norm": 1.4556376676495002, + "learning_rate": 9.954345308109746e-06, + "loss": 0.3094, + "step": 3260 + }, + { + "epoch": 0.07178076161127009, + "grad_norm": 1.5839532474996034, + "learning_rate": 9.9542972353338e-06, + "loss": 0.3685, + "step": 3261 + }, + { + "epoch": 0.07180277349768875, + "grad_norm": 1.5279526751041324, + "learning_rate": 9.95424913737791e-06, + "loss": 0.4161, + "step": 3262 + }, + { + "epoch": 0.07182478538410741, + "grad_norm": 1.6033586747224573, + "learning_rate": 9.954201014242313e-06, + "loss": 0.4014, + "step": 3263 + }, + { + "epoch": 0.07184679727052608, + "grad_norm": 1.9240912553657081, + "learning_rate": 9.95415286592726e-06, + "loss": 0.3582, + "step": 3264 + }, + { + "epoch": 0.07186880915694475, + "grad_norm": 0.9598845851210125, + "learning_rate": 9.954104692432991e-06, + "loss": 0.5839, + "step": 3265 + }, + { + "epoch": 0.07189082104336342, + "grad_norm": 1.27663054252349, + "learning_rate": 9.954056493759752e-06, + "loss": 0.3558, + "step": 3266 + }, + { + "epoch": 0.07191283292978208, + "grad_norm": 1.8555108163514695, + "learning_rate": 9.954008269907787e-06, + "loss": 0.3969, + "step": 3267 + }, + { + "epoch": 0.07193484481620074, + "grad_norm": 1.761033405160394, + "learning_rate": 9.953960020877345e-06, + "loss": 0.4076, + "step": 3268 + }, + { + "epoch": 0.07195685670261942, + "grad_norm": 2.1379476032878415, + "learning_rate": 9.953911746668667e-06, + "loss": 0.4913, + "step": 3269 + }, + { + "epoch": 0.07197886858903808, + "grad_norm": 1.5960728195206009, + "learning_rate": 9.953863447282002e-06, + "loss": 0.4014, + "step": 3270 + }, + { + "epoch": 0.07200088047545675, + "grad_norm": 2.0790720052772387, + "learning_rate": 9.953815122717592e-06, + "loss": 0.3908, + "step": 3271 + }, + { + "epoch": 0.07202289236187541, + "grad_norm": 1.6916735244411651, + "learning_rate": 9.953766772975684e-06, + "loss": 0.439, + "step": 3272 + }, + { + "epoch": 0.07204490424829407, + "grad_norm": 1.8164133770696247, + "learning_rate": 9.953718398056525e-06, + "loss": 0.3141, + "step": 3273 + }, + { + "epoch": 0.07206691613471275, + "grad_norm": 1.805843167081566, + "learning_rate": 9.953669997960359e-06, + "loss": 0.3716, + "step": 3274 + }, + { + "epoch": 0.07208892802113141, + "grad_norm": 1.614566171647637, + "learning_rate": 9.95362157268743e-06, + "loss": 0.3782, + "step": 3275 + }, + { + "epoch": 0.07211093990755008, + "grad_norm": 1.928124847977361, + "learning_rate": 9.95357312223799e-06, + "loss": 0.4504, + "step": 3276 + }, + { + "epoch": 0.07213295179396874, + "grad_norm": 1.4276563714936885, + "learning_rate": 9.95352464661228e-06, + "loss": 0.3617, + "step": 3277 + }, + { + "epoch": 0.07215496368038742, + "grad_norm": 0.8430700355068582, + "learning_rate": 9.95347614581055e-06, + "loss": 0.5674, + "step": 3278 + }, + { + "epoch": 0.07217697556680608, + "grad_norm": 1.4390690981956031, + "learning_rate": 9.953427619833045e-06, + "loss": 0.3658, + "step": 3279 + }, + { + "epoch": 0.07219898745322474, + "grad_norm": 2.9170881816591545, + "learning_rate": 9.953379068680011e-06, + "loss": 0.4948, + "step": 3280 + }, + { + "epoch": 0.0722209993396434, + "grad_norm": 0.6785069258183061, + "learning_rate": 9.953330492351696e-06, + "loss": 0.5954, + "step": 3281 + }, + { + "epoch": 0.07224301122606207, + "grad_norm": 0.6311976042000387, + "learning_rate": 9.953281890848345e-06, + "loss": 0.5613, + "step": 3282 + }, + { + "epoch": 0.07226502311248074, + "grad_norm": 0.6552448151410701, + "learning_rate": 9.953233264170206e-06, + "loss": 0.5786, + "step": 3283 + }, + { + "epoch": 0.07228703499889941, + "grad_norm": 1.854482392605474, + "learning_rate": 9.953184612317528e-06, + "loss": 0.3424, + "step": 3284 + }, + { + "epoch": 0.07230904688531807, + "grad_norm": 3.5249697688550583, + "learning_rate": 9.953135935290557e-06, + "loss": 0.3808, + "step": 3285 + }, + { + "epoch": 0.07233105877173673, + "grad_norm": 3.9958155528194608, + "learning_rate": 9.953087233089538e-06, + "loss": 0.3048, + "step": 3286 + }, + { + "epoch": 0.07235307065815541, + "grad_norm": 2.084791969671556, + "learning_rate": 9.953038505714721e-06, + "loss": 0.4035, + "step": 3287 + }, + { + "epoch": 0.07237508254457407, + "grad_norm": 2.602348507489142, + "learning_rate": 9.952989753166353e-06, + "loss": 0.3898, + "step": 3288 + }, + { + "epoch": 0.07239709443099274, + "grad_norm": 1.595523083209297, + "learning_rate": 9.952940975444682e-06, + "loss": 0.4229, + "step": 3289 + }, + { + "epoch": 0.0724191063174114, + "grad_norm": 1.5584335136009075, + "learning_rate": 9.952892172549956e-06, + "loss": 0.274, + "step": 3290 + }, + { + "epoch": 0.07244111820383006, + "grad_norm": 0.8981044258426908, + "learning_rate": 9.952843344482425e-06, + "loss": 0.586, + "step": 3291 + }, + { + "epoch": 0.07246313009024874, + "grad_norm": 1.6845172345344166, + "learning_rate": 9.952794491242332e-06, + "loss": 0.3803, + "step": 3292 + }, + { + "epoch": 0.0724851419766674, + "grad_norm": 1.5221570039442784, + "learning_rate": 9.95274561282993e-06, + "loss": 0.3481, + "step": 3293 + }, + { + "epoch": 0.07250715386308607, + "grad_norm": 1.7715449514970012, + "learning_rate": 9.952696709245467e-06, + "loss": 0.4026, + "step": 3294 + }, + { + "epoch": 0.07252916574950473, + "grad_norm": 1.9696298834479071, + "learning_rate": 9.952647780489189e-06, + "loss": 0.487, + "step": 3295 + }, + { + "epoch": 0.07255117763592339, + "grad_norm": 1.9949339148514484, + "learning_rate": 9.952598826561345e-06, + "loss": 0.3944, + "step": 3296 + }, + { + "epoch": 0.07257318952234207, + "grad_norm": 2.530458912850564, + "learning_rate": 9.952549847462187e-06, + "loss": 0.3685, + "step": 3297 + }, + { + "epoch": 0.07259520140876073, + "grad_norm": 0.7056079612626275, + "learning_rate": 9.95250084319196e-06, + "loss": 0.5728, + "step": 3298 + }, + { + "epoch": 0.0726172132951794, + "grad_norm": 2.0467879604640693, + "learning_rate": 9.952451813750915e-06, + "loss": 0.3923, + "step": 3299 + }, + { + "epoch": 0.07263922518159806, + "grad_norm": 1.761433663710871, + "learning_rate": 9.952402759139302e-06, + "loss": 0.3922, + "step": 3300 + }, + { + "epoch": 0.07266123706801673, + "grad_norm": 1.5040060551033023, + "learning_rate": 9.952353679357369e-06, + "loss": 0.3742, + "step": 3301 + }, + { + "epoch": 0.0726832489544354, + "grad_norm": 1.4333102984308936, + "learning_rate": 9.952304574405365e-06, + "loss": 0.3225, + "step": 3302 + }, + { + "epoch": 0.07270526084085406, + "grad_norm": 1.7817276525555004, + "learning_rate": 9.952255444283541e-06, + "loss": 0.3713, + "step": 3303 + }, + { + "epoch": 0.07272727272727272, + "grad_norm": 0.6937550183732188, + "learning_rate": 9.952206288992145e-06, + "loss": 0.5623, + "step": 3304 + }, + { + "epoch": 0.07274928461369139, + "grad_norm": 1.6138991723955154, + "learning_rate": 9.95215710853143e-06, + "loss": 0.3948, + "step": 3305 + }, + { + "epoch": 0.07277129650011006, + "grad_norm": 2.039041778360482, + "learning_rate": 9.952107902901643e-06, + "loss": 0.4071, + "step": 3306 + }, + { + "epoch": 0.07279330838652873, + "grad_norm": 1.8977482107283707, + "learning_rate": 9.952058672103035e-06, + "loss": 0.3455, + "step": 3307 + }, + { + "epoch": 0.07281532027294739, + "grad_norm": 4.006131002187012, + "learning_rate": 9.952009416135856e-06, + "loss": 0.445, + "step": 3308 + }, + { + "epoch": 0.07283733215936605, + "grad_norm": 1.720642293691826, + "learning_rate": 9.951960135000356e-06, + "loss": 0.4173, + "step": 3309 + }, + { + "epoch": 0.07285934404578473, + "grad_norm": 1.6711884146885432, + "learning_rate": 9.951910828696786e-06, + "loss": 0.3695, + "step": 3310 + }, + { + "epoch": 0.07288135593220339, + "grad_norm": 0.6606941143457096, + "learning_rate": 9.951861497225396e-06, + "loss": 0.5694, + "step": 3311 + }, + { + "epoch": 0.07290336781862206, + "grad_norm": 1.5835945705284211, + "learning_rate": 9.951812140586439e-06, + "loss": 0.3226, + "step": 3312 + }, + { + "epoch": 0.07292537970504072, + "grad_norm": 2.814658942578322, + "learning_rate": 9.951762758780164e-06, + "loss": 0.3159, + "step": 3313 + }, + { + "epoch": 0.07294739159145938, + "grad_norm": 1.5769717618081545, + "learning_rate": 9.95171335180682e-06, + "loss": 0.4421, + "step": 3314 + }, + { + "epoch": 0.07296940347787806, + "grad_norm": 1.434099453613434, + "learning_rate": 9.951663919666662e-06, + "loss": 0.4304, + "step": 3315 + }, + { + "epoch": 0.07299141536429672, + "grad_norm": 2.263991859415658, + "learning_rate": 9.95161446235994e-06, + "loss": 0.4579, + "step": 3316 + }, + { + "epoch": 0.07301342725071538, + "grad_norm": 1.7990454177166697, + "learning_rate": 9.951564979886903e-06, + "loss": 0.2961, + "step": 3317 + }, + { + "epoch": 0.07303543913713405, + "grad_norm": 1.7722643353109584, + "learning_rate": 9.951515472247806e-06, + "loss": 0.3529, + "step": 3318 + }, + { + "epoch": 0.07305745102355272, + "grad_norm": 1.9759322800876316, + "learning_rate": 9.951465939442897e-06, + "loss": 0.3658, + "step": 3319 + }, + { + "epoch": 0.07307946290997139, + "grad_norm": 0.6646267698767708, + "learning_rate": 9.951416381472431e-06, + "loss": 0.5604, + "step": 3320 + }, + { + "epoch": 0.07310147479639005, + "grad_norm": 0.6779913969794565, + "learning_rate": 9.951366798336659e-06, + "loss": 0.6196, + "step": 3321 + }, + { + "epoch": 0.07312348668280871, + "grad_norm": 1.529550796454282, + "learning_rate": 9.951317190035832e-06, + "loss": 0.4005, + "step": 3322 + }, + { + "epoch": 0.07314549856922738, + "grad_norm": 0.6331886555706354, + "learning_rate": 9.951267556570202e-06, + "loss": 0.5739, + "step": 3323 + }, + { + "epoch": 0.07316751045564605, + "grad_norm": 2.1695509913831157, + "learning_rate": 9.951217897940022e-06, + "loss": 0.4357, + "step": 3324 + }, + { + "epoch": 0.07318952234206472, + "grad_norm": 2.014831496305223, + "learning_rate": 9.951168214145546e-06, + "loss": 0.4407, + "step": 3325 + }, + { + "epoch": 0.07321153422848338, + "grad_norm": 1.8148071317947023, + "learning_rate": 9.951118505187022e-06, + "loss": 0.5043, + "step": 3326 + }, + { + "epoch": 0.07323354611490204, + "grad_norm": 2.5295620971072683, + "learning_rate": 9.951068771064708e-06, + "loss": 0.4363, + "step": 3327 + }, + { + "epoch": 0.07325555800132072, + "grad_norm": 1.4967008992904538, + "learning_rate": 9.951019011778852e-06, + "loss": 0.2893, + "step": 3328 + }, + { + "epoch": 0.07327756988773938, + "grad_norm": 2.1874151341818355, + "learning_rate": 9.950969227329711e-06, + "loss": 0.3417, + "step": 3329 + }, + { + "epoch": 0.07329958177415805, + "grad_norm": 0.7368213529269562, + "learning_rate": 9.950919417717537e-06, + "loss": 0.5743, + "step": 3330 + }, + { + "epoch": 0.07332159366057671, + "grad_norm": 2.1432911182495964, + "learning_rate": 9.950869582942582e-06, + "loss": 0.4371, + "step": 3331 + }, + { + "epoch": 0.07334360554699537, + "grad_norm": 2.696003694734215, + "learning_rate": 9.9508197230051e-06, + "loss": 0.407, + "step": 3332 + }, + { + "epoch": 0.07336561743341405, + "grad_norm": 1.5540495061056654, + "learning_rate": 9.95076983790534e-06, + "loss": 0.4148, + "step": 3333 + }, + { + "epoch": 0.07338762931983271, + "grad_norm": 1.575770330022431, + "learning_rate": 9.950719927643563e-06, + "loss": 0.3231, + "step": 3334 + }, + { + "epoch": 0.07340964120625137, + "grad_norm": 1.551448221937318, + "learning_rate": 9.950669992220018e-06, + "loss": 0.3345, + "step": 3335 + }, + { + "epoch": 0.07343165309267004, + "grad_norm": 1.4472487609126934, + "learning_rate": 9.950620031634962e-06, + "loss": 0.3299, + "step": 3336 + }, + { + "epoch": 0.07345366497908871, + "grad_norm": 2.922075395856201, + "learning_rate": 9.950570045888646e-06, + "loss": 0.4182, + "step": 3337 + }, + { + "epoch": 0.07347567686550738, + "grad_norm": 5.213043500538733, + "learning_rate": 9.950520034981324e-06, + "loss": 0.3559, + "step": 3338 + }, + { + "epoch": 0.07349768875192604, + "grad_norm": 7.659258354658082, + "learning_rate": 9.950469998913252e-06, + "loss": 0.4133, + "step": 3339 + }, + { + "epoch": 0.0735197006383447, + "grad_norm": 1.665297910090088, + "learning_rate": 9.950419937684684e-06, + "loss": 0.2916, + "step": 3340 + }, + { + "epoch": 0.07354171252476337, + "grad_norm": 1.9583003409143585, + "learning_rate": 9.950369851295873e-06, + "loss": 0.3492, + "step": 3341 + }, + { + "epoch": 0.07356372441118204, + "grad_norm": 2.6248640724470795, + "learning_rate": 9.950319739747075e-06, + "loss": 0.4375, + "step": 3342 + }, + { + "epoch": 0.0735857362976007, + "grad_norm": 1.7422467158396064, + "learning_rate": 9.950269603038544e-06, + "loss": 0.3738, + "step": 3343 + }, + { + "epoch": 0.07360774818401937, + "grad_norm": 1.9261882845929799, + "learning_rate": 9.950219441170535e-06, + "loss": 0.4621, + "step": 3344 + }, + { + "epoch": 0.07362976007043803, + "grad_norm": 1.8147225570341328, + "learning_rate": 9.950169254143303e-06, + "loss": 0.4565, + "step": 3345 + }, + { + "epoch": 0.0736517719568567, + "grad_norm": 2.1661574379864943, + "learning_rate": 9.950119041957103e-06, + "loss": 0.4731, + "step": 3346 + }, + { + "epoch": 0.07367378384327537, + "grad_norm": 2.241388102008437, + "learning_rate": 9.95006880461219e-06, + "loss": 0.3668, + "step": 3347 + }, + { + "epoch": 0.07369579572969404, + "grad_norm": 0.6952260281275744, + "learning_rate": 9.950018542108818e-06, + "loss": 0.572, + "step": 3348 + }, + { + "epoch": 0.0737178076161127, + "grad_norm": 1.5565341330966058, + "learning_rate": 9.949968254447248e-06, + "loss": 0.3978, + "step": 3349 + }, + { + "epoch": 0.07373981950253136, + "grad_norm": 1.432722466959655, + "learning_rate": 9.94991794162773e-06, + "loss": 0.352, + "step": 3350 + }, + { + "epoch": 0.07376183138895004, + "grad_norm": 2.5949339093663895, + "learning_rate": 9.94986760365052e-06, + "loss": 0.3487, + "step": 3351 + }, + { + "epoch": 0.0737838432753687, + "grad_norm": 2.1255522637113056, + "learning_rate": 9.949817240515875e-06, + "loss": 0.3528, + "step": 3352 + }, + { + "epoch": 0.07380585516178736, + "grad_norm": 1.6365647472361011, + "learning_rate": 9.949766852224053e-06, + "loss": 0.4362, + "step": 3353 + }, + { + "epoch": 0.07382786704820603, + "grad_norm": 0.6859447284091644, + "learning_rate": 9.949716438775306e-06, + "loss": 0.5959, + "step": 3354 + }, + { + "epoch": 0.07384987893462469, + "grad_norm": 2.620200042536338, + "learning_rate": 9.949666000169894e-06, + "loss": 0.3728, + "step": 3355 + }, + { + "epoch": 0.07387189082104337, + "grad_norm": 2.0803407985464077, + "learning_rate": 9.949615536408071e-06, + "loss": 0.4104, + "step": 3356 + }, + { + "epoch": 0.07389390270746203, + "grad_norm": 1.8637257399139004, + "learning_rate": 9.949565047490094e-06, + "loss": 0.4396, + "step": 3357 + }, + { + "epoch": 0.0739159145938807, + "grad_norm": 1.7421478650805489, + "learning_rate": 9.94951453341622e-06, + "loss": 0.477, + "step": 3358 + }, + { + "epoch": 0.07393792648029936, + "grad_norm": 1.6505432386192869, + "learning_rate": 9.949463994186706e-06, + "loss": 0.318, + "step": 3359 + }, + { + "epoch": 0.07395993836671803, + "grad_norm": 1.7555179977802051, + "learning_rate": 9.94941342980181e-06, + "loss": 0.3122, + "step": 3360 + }, + { + "epoch": 0.0739819502531367, + "grad_norm": 1.4757320861797751, + "learning_rate": 9.949362840261786e-06, + "loss": 0.4823, + "step": 3361 + }, + { + "epoch": 0.07400396213955536, + "grad_norm": 1.6991633522171405, + "learning_rate": 9.949312225566892e-06, + "loss": 0.4042, + "step": 3362 + }, + { + "epoch": 0.07402597402597402, + "grad_norm": 1.6228124092697587, + "learning_rate": 9.949261585717384e-06, + "loss": 0.4127, + "step": 3363 + }, + { + "epoch": 0.07404798591239269, + "grad_norm": 2.5124793381562207, + "learning_rate": 9.949210920713525e-06, + "loss": 0.4165, + "step": 3364 + }, + { + "epoch": 0.07406999779881136, + "grad_norm": 1.4496191070055031, + "learning_rate": 9.949160230555567e-06, + "loss": 0.3369, + "step": 3365 + }, + { + "epoch": 0.07409200968523003, + "grad_norm": 4.512825925935653, + "learning_rate": 9.94910951524377e-06, + "loss": 0.3826, + "step": 3366 + }, + { + "epoch": 0.07411402157164869, + "grad_norm": 1.5025834926117805, + "learning_rate": 9.94905877477839e-06, + "loss": 0.2914, + "step": 3367 + }, + { + "epoch": 0.07413603345806735, + "grad_norm": 1.77722329883272, + "learning_rate": 9.949008009159687e-06, + "loss": 0.4349, + "step": 3368 + }, + { + "epoch": 0.07415804534448603, + "grad_norm": 2.161398281945458, + "learning_rate": 9.948957218387916e-06, + "loss": 0.3438, + "step": 3369 + }, + { + "epoch": 0.07418005723090469, + "grad_norm": 1.3273736718014255, + "learning_rate": 9.948906402463338e-06, + "loss": 0.3535, + "step": 3370 + }, + { + "epoch": 0.07420206911732335, + "grad_norm": 1.5136512562688025, + "learning_rate": 9.948855561386213e-06, + "loss": 0.3325, + "step": 3371 + }, + { + "epoch": 0.07422408100374202, + "grad_norm": 0.730714856970239, + "learning_rate": 9.948804695156794e-06, + "loss": 0.5787, + "step": 3372 + }, + { + "epoch": 0.07424609289016068, + "grad_norm": 1.9816616967284604, + "learning_rate": 9.948753803775342e-06, + "loss": 0.4089, + "step": 3373 + }, + { + "epoch": 0.07426810477657936, + "grad_norm": 0.6703020087083754, + "learning_rate": 9.948702887242119e-06, + "loss": 0.5689, + "step": 3374 + }, + { + "epoch": 0.07429011666299802, + "grad_norm": 1.5282152600477013, + "learning_rate": 9.948651945557378e-06, + "loss": 0.3526, + "step": 3375 + }, + { + "epoch": 0.07431212854941668, + "grad_norm": 1.6536461642284022, + "learning_rate": 9.948600978721381e-06, + "loss": 0.3956, + "step": 3376 + }, + { + "epoch": 0.07433414043583535, + "grad_norm": 1.3472613689176487, + "learning_rate": 9.948549986734386e-06, + "loss": 0.3326, + "step": 3377 + }, + { + "epoch": 0.07435615232225402, + "grad_norm": 2.0354467122558106, + "learning_rate": 9.948498969596656e-06, + "loss": 0.3806, + "step": 3378 + }, + { + "epoch": 0.07437816420867269, + "grad_norm": 4.290501101290931, + "learning_rate": 9.948447927308445e-06, + "loss": 0.4155, + "step": 3379 + }, + { + "epoch": 0.07440017609509135, + "grad_norm": 1.5313861727396725, + "learning_rate": 9.948396859870013e-06, + "loss": 0.3544, + "step": 3380 + }, + { + "epoch": 0.07442218798151001, + "grad_norm": 2.1330774858385713, + "learning_rate": 9.948345767281625e-06, + "loss": 0.3419, + "step": 3381 + }, + { + "epoch": 0.07444419986792868, + "grad_norm": 2.7101705923210235, + "learning_rate": 9.948294649543532e-06, + "loss": 0.325, + "step": 3382 + }, + { + "epoch": 0.07446621175434735, + "grad_norm": 2.6878380698167876, + "learning_rate": 9.948243506656002e-06, + "loss": 0.4853, + "step": 3383 + }, + { + "epoch": 0.07448822364076602, + "grad_norm": 1.9118933780861207, + "learning_rate": 9.948192338619292e-06, + "loss": 0.375, + "step": 3384 + }, + { + "epoch": 0.07451023552718468, + "grad_norm": 6.479426661224424, + "learning_rate": 9.948141145433659e-06, + "loss": 0.4069, + "step": 3385 + }, + { + "epoch": 0.07453224741360334, + "grad_norm": 2.1099804583064015, + "learning_rate": 9.948089927099367e-06, + "loss": 0.4102, + "step": 3386 + }, + { + "epoch": 0.074554259300022, + "grad_norm": 2.1795749203067714, + "learning_rate": 9.948038683616676e-06, + "loss": 0.4539, + "step": 3387 + }, + { + "epoch": 0.07457627118644068, + "grad_norm": 1.8299513303525166, + "learning_rate": 9.947987414985845e-06, + "loss": 0.4933, + "step": 3388 + }, + { + "epoch": 0.07459828307285934, + "grad_norm": 1.6041612726013974, + "learning_rate": 9.947936121207134e-06, + "loss": 0.4269, + "step": 3389 + }, + { + "epoch": 0.07462029495927801, + "grad_norm": 0.9120979847979014, + "learning_rate": 9.947884802280805e-06, + "loss": 0.5847, + "step": 3390 + }, + { + "epoch": 0.07464230684569667, + "grad_norm": 1.710469588917276, + "learning_rate": 9.94783345820712e-06, + "loss": 0.4132, + "step": 3391 + }, + { + "epoch": 0.07466431873211535, + "grad_norm": 1.6528794677531575, + "learning_rate": 9.947782088986338e-06, + "loss": 0.332, + "step": 3392 + }, + { + "epoch": 0.07468633061853401, + "grad_norm": 2.372308519598118, + "learning_rate": 9.94773069461872e-06, + "loss": 0.243, + "step": 3393 + }, + { + "epoch": 0.07470834250495267, + "grad_norm": 2.3532710384441744, + "learning_rate": 9.947679275104526e-06, + "loss": 0.3305, + "step": 3394 + }, + { + "epoch": 0.07473035439137134, + "grad_norm": 1.6243241341588648, + "learning_rate": 9.947627830444022e-06, + "loss": 0.3794, + "step": 3395 + }, + { + "epoch": 0.07475236627779, + "grad_norm": 4.426102419719305, + "learning_rate": 9.947576360637468e-06, + "loss": 0.3738, + "step": 3396 + }, + { + "epoch": 0.07477437816420868, + "grad_norm": 1.553632357153817, + "learning_rate": 9.94752486568512e-06, + "loss": 0.4037, + "step": 3397 + }, + { + "epoch": 0.07479639005062734, + "grad_norm": 1.5377635815605595, + "learning_rate": 9.947473345587247e-06, + "loss": 0.4154, + "step": 3398 + }, + { + "epoch": 0.074818401937046, + "grad_norm": 0.7891865663742634, + "learning_rate": 9.947421800344108e-06, + "loss": 0.5987, + "step": 3399 + }, + { + "epoch": 0.07484041382346467, + "grad_norm": 5.838386139164648, + "learning_rate": 9.947370229955964e-06, + "loss": 0.3703, + "step": 3400 + }, + { + "epoch": 0.07486242570988334, + "grad_norm": 2.154337811750131, + "learning_rate": 9.947318634423078e-06, + "loss": 0.3862, + "step": 3401 + }, + { + "epoch": 0.074884437596302, + "grad_norm": 1.468412591562013, + "learning_rate": 9.947267013745712e-06, + "loss": 0.424, + "step": 3402 + }, + { + "epoch": 0.07490644948272067, + "grad_norm": 2.241918502825961, + "learning_rate": 9.947215367924127e-06, + "loss": 0.3693, + "step": 3403 + }, + { + "epoch": 0.07492846136913933, + "grad_norm": 1.5696696482853447, + "learning_rate": 9.94716369695859e-06, + "loss": 0.3725, + "step": 3404 + }, + { + "epoch": 0.074950473255558, + "grad_norm": 3.654855521153362, + "learning_rate": 9.947112000849358e-06, + "loss": 0.3518, + "step": 3405 + }, + { + "epoch": 0.07497248514197667, + "grad_norm": 12.375480918439132, + "learning_rate": 9.947060279596698e-06, + "loss": 0.4275, + "step": 3406 + }, + { + "epoch": 0.07499449702839533, + "grad_norm": 1.581653916689031, + "learning_rate": 9.94700853320087e-06, + "loss": 0.3358, + "step": 3407 + }, + { + "epoch": 0.075016508914814, + "grad_norm": 1.8250108289979823, + "learning_rate": 9.94695676166214e-06, + "loss": 0.4421, + "step": 3408 + }, + { + "epoch": 0.07503852080123266, + "grad_norm": 1.5723239501831376, + "learning_rate": 9.946904964980766e-06, + "loss": 0.3312, + "step": 3409 + }, + { + "epoch": 0.07506053268765134, + "grad_norm": 2.0124321174893285, + "learning_rate": 9.946853143157017e-06, + "loss": 0.4313, + "step": 3410 + }, + { + "epoch": 0.07508254457407, + "grad_norm": 5.512450853803659, + "learning_rate": 9.946801296191154e-06, + "loss": 0.3598, + "step": 3411 + }, + { + "epoch": 0.07510455646048866, + "grad_norm": 8.533544502041188, + "learning_rate": 9.94674942408344e-06, + "loss": 0.3436, + "step": 3412 + }, + { + "epoch": 0.07512656834690733, + "grad_norm": 2.1111838206101523, + "learning_rate": 9.94669752683414e-06, + "loss": 0.3613, + "step": 3413 + }, + { + "epoch": 0.07514858023332599, + "grad_norm": 1.5204200864675779, + "learning_rate": 9.946645604443516e-06, + "loss": 0.4195, + "step": 3414 + }, + { + "epoch": 0.07517059211974467, + "grad_norm": 0.7171042402571804, + "learning_rate": 9.946593656911833e-06, + "loss": 0.5914, + "step": 3415 + }, + { + "epoch": 0.07519260400616333, + "grad_norm": 1.6173196642135546, + "learning_rate": 9.946541684239355e-06, + "loss": 0.3297, + "step": 3416 + }, + { + "epoch": 0.07521461589258199, + "grad_norm": 3.9701817895375724, + "learning_rate": 9.946489686426344e-06, + "loss": 0.3259, + "step": 3417 + }, + { + "epoch": 0.07523662777900066, + "grad_norm": 2.40454926856797, + "learning_rate": 9.94643766347307e-06, + "loss": 0.3471, + "step": 3418 + }, + { + "epoch": 0.07525863966541933, + "grad_norm": 2.3340638055067555, + "learning_rate": 9.94638561537979e-06, + "loss": 0.3695, + "step": 3419 + }, + { + "epoch": 0.075280651551838, + "grad_norm": 1.3333718685374483, + "learning_rate": 9.946333542146774e-06, + "loss": 0.3935, + "step": 3420 + }, + { + "epoch": 0.07530266343825666, + "grad_norm": 3.391034011554932, + "learning_rate": 9.946281443774284e-06, + "loss": 0.4217, + "step": 3421 + }, + { + "epoch": 0.07532467532467532, + "grad_norm": 1.9380603522123554, + "learning_rate": 9.946229320262586e-06, + "loss": 0.4454, + "step": 3422 + }, + { + "epoch": 0.07534668721109399, + "grad_norm": 1.6279451507978628, + "learning_rate": 9.946177171611947e-06, + "loss": 0.3487, + "step": 3423 + }, + { + "epoch": 0.07536869909751266, + "grad_norm": 1.9648642449788252, + "learning_rate": 9.946124997822625e-06, + "loss": 0.4401, + "step": 3424 + }, + { + "epoch": 0.07539071098393132, + "grad_norm": 2.2695555849255906, + "learning_rate": 9.946072798894892e-06, + "loss": 0.4212, + "step": 3425 + }, + { + "epoch": 0.07541272287034999, + "grad_norm": 1.67844876326996, + "learning_rate": 9.946020574829012e-06, + "loss": 0.4473, + "step": 3426 + }, + { + "epoch": 0.07543473475676865, + "grad_norm": 1.8559140298184633, + "learning_rate": 9.945968325625249e-06, + "loss": 0.4028, + "step": 3427 + }, + { + "epoch": 0.07545674664318733, + "grad_norm": 2.509736844725678, + "learning_rate": 9.945916051283865e-06, + "loss": 0.3897, + "step": 3428 + }, + { + "epoch": 0.07547875852960599, + "grad_norm": 3.100233894988693, + "learning_rate": 9.945863751805136e-06, + "loss": 0.3957, + "step": 3429 + }, + { + "epoch": 0.07550077041602465, + "grad_norm": 1.532730645990779, + "learning_rate": 9.945811427189318e-06, + "loss": 0.3512, + "step": 3430 + }, + { + "epoch": 0.07552278230244332, + "grad_norm": 0.732574989172478, + "learning_rate": 9.94575907743668e-06, + "loss": 0.6263, + "step": 3431 + }, + { + "epoch": 0.07554479418886198, + "grad_norm": 0.6842643831341525, + "learning_rate": 9.94570670254749e-06, + "loss": 0.5869, + "step": 3432 + }, + { + "epoch": 0.07556680607528066, + "grad_norm": 1.682774276639536, + "learning_rate": 9.945654302522013e-06, + "loss": 0.3177, + "step": 3433 + }, + { + "epoch": 0.07558881796169932, + "grad_norm": 1.4769225915282453, + "learning_rate": 9.945601877360514e-06, + "loss": 0.3789, + "step": 3434 + }, + { + "epoch": 0.07561082984811798, + "grad_norm": 0.6509902190664121, + "learning_rate": 9.945549427063259e-06, + "loss": 0.6071, + "step": 3435 + }, + { + "epoch": 0.07563284173453665, + "grad_norm": 2.1443342577593274, + "learning_rate": 9.945496951630517e-06, + "loss": 0.4894, + "step": 3436 + }, + { + "epoch": 0.07565485362095531, + "grad_norm": 1.7765232797681734, + "learning_rate": 9.945444451062556e-06, + "loss": 0.3567, + "step": 3437 + }, + { + "epoch": 0.07567686550737399, + "grad_norm": 1.6282075539879313, + "learning_rate": 9.945391925359637e-06, + "loss": 0.317, + "step": 3438 + }, + { + "epoch": 0.07569887739379265, + "grad_norm": 3.355725433346894, + "learning_rate": 9.945339374522034e-06, + "loss": 0.3476, + "step": 3439 + }, + { + "epoch": 0.07572088928021131, + "grad_norm": 1.743243555147682, + "learning_rate": 9.945286798550009e-06, + "loss": 0.4511, + "step": 3440 + }, + { + "epoch": 0.07574290116662998, + "grad_norm": 1.8341153952285878, + "learning_rate": 9.945234197443831e-06, + "loss": 0.4188, + "step": 3441 + }, + { + "epoch": 0.07576491305304865, + "grad_norm": 2.6860003324726427, + "learning_rate": 9.945181571203767e-06, + "loss": 0.3433, + "step": 3442 + }, + { + "epoch": 0.07578692493946732, + "grad_norm": 1.9451240237683816, + "learning_rate": 9.945128919830085e-06, + "loss": 0.3838, + "step": 3443 + }, + { + "epoch": 0.07580893682588598, + "grad_norm": 1.686519655027294, + "learning_rate": 9.945076243323052e-06, + "loss": 0.4063, + "step": 3444 + }, + { + "epoch": 0.07583094871230464, + "grad_norm": 1.5368998539480436, + "learning_rate": 9.945023541682935e-06, + "loss": 0.343, + "step": 3445 + }, + { + "epoch": 0.0758529605987233, + "grad_norm": 2.1021904006020398, + "learning_rate": 9.944970814910004e-06, + "loss": 0.3708, + "step": 3446 + }, + { + "epoch": 0.07587497248514198, + "grad_norm": 1.7877603680012348, + "learning_rate": 9.944918063004526e-06, + "loss": 0.4365, + "step": 3447 + }, + { + "epoch": 0.07589698437156064, + "grad_norm": 2.3744269573258676, + "learning_rate": 9.944865285966767e-06, + "loss": 0.3636, + "step": 3448 + }, + { + "epoch": 0.07591899625797931, + "grad_norm": 1.8410558498016114, + "learning_rate": 9.944812483797e-06, + "loss": 0.4601, + "step": 3449 + }, + { + "epoch": 0.07594100814439797, + "grad_norm": 1.3274920990823966, + "learning_rate": 9.944759656495489e-06, + "loss": 0.3191, + "step": 3450 + }, + { + "epoch": 0.07596302003081665, + "grad_norm": 1.7534551133183751, + "learning_rate": 9.944706804062505e-06, + "loss": 0.4012, + "step": 3451 + }, + { + "epoch": 0.07598503191723531, + "grad_norm": 1.9793919517937644, + "learning_rate": 9.944653926498314e-06, + "loss": 0.4905, + "step": 3452 + }, + { + "epoch": 0.07600704380365397, + "grad_norm": 2.200794775425972, + "learning_rate": 9.944601023803187e-06, + "loss": 0.4339, + "step": 3453 + }, + { + "epoch": 0.07602905569007264, + "grad_norm": 1.5907013274531292, + "learning_rate": 9.944548095977394e-06, + "loss": 0.4093, + "step": 3454 + }, + { + "epoch": 0.0760510675764913, + "grad_norm": 3.1957259012388333, + "learning_rate": 9.944495143021199e-06, + "loss": 0.4152, + "step": 3455 + }, + { + "epoch": 0.07607307946290998, + "grad_norm": 2.681903691372959, + "learning_rate": 9.944442164934876e-06, + "loss": 0.4113, + "step": 3456 + }, + { + "epoch": 0.07609509134932864, + "grad_norm": 2.617632286199877, + "learning_rate": 9.944389161718692e-06, + "loss": 0.3934, + "step": 3457 + }, + { + "epoch": 0.0761171032357473, + "grad_norm": 1.77349128585592, + "learning_rate": 9.944336133372918e-06, + "loss": 0.4509, + "step": 3458 + }, + { + "epoch": 0.07613911512216597, + "grad_norm": 1.8427947228174402, + "learning_rate": 9.944283079897824e-06, + "loss": 0.4278, + "step": 3459 + }, + { + "epoch": 0.07616112700858464, + "grad_norm": 1.804152754843352, + "learning_rate": 9.944230001293676e-06, + "loss": 0.3568, + "step": 3460 + }, + { + "epoch": 0.0761831388950033, + "grad_norm": 1.442158227184099, + "learning_rate": 9.944176897560744e-06, + "loss": 0.3726, + "step": 3461 + }, + { + "epoch": 0.07620515078142197, + "grad_norm": 2.0198709966103627, + "learning_rate": 9.944123768699303e-06, + "loss": 0.4189, + "step": 3462 + }, + { + "epoch": 0.07622716266784063, + "grad_norm": 2.6762373064476797, + "learning_rate": 9.94407061470962e-06, + "loss": 0.3759, + "step": 3463 + }, + { + "epoch": 0.0762491745542593, + "grad_norm": 1.7433188997002487, + "learning_rate": 9.944017435591963e-06, + "loss": 0.4146, + "step": 3464 + }, + { + "epoch": 0.07627118644067797, + "grad_norm": 1.0007531080087706, + "learning_rate": 9.943964231346605e-06, + "loss": 0.5922, + "step": 3465 + }, + { + "epoch": 0.07629319832709663, + "grad_norm": 1.801207577010257, + "learning_rate": 9.943911001973815e-06, + "loss": 0.4319, + "step": 3466 + }, + { + "epoch": 0.0763152102135153, + "grad_norm": 2.6018266075461858, + "learning_rate": 9.943857747473866e-06, + "loss": 0.4431, + "step": 3467 + }, + { + "epoch": 0.07633722209993396, + "grad_norm": 1.480958111793476, + "learning_rate": 9.943804467847028e-06, + "loss": 0.3802, + "step": 3468 + }, + { + "epoch": 0.07635923398635264, + "grad_norm": 3.6047089446945675, + "learning_rate": 9.943751163093569e-06, + "loss": 0.3682, + "step": 3469 + }, + { + "epoch": 0.0763812458727713, + "grad_norm": 2.320935418992986, + "learning_rate": 9.94369783321376e-06, + "loss": 0.3897, + "step": 3470 + }, + { + "epoch": 0.07640325775918996, + "grad_norm": 2.887255749791489, + "learning_rate": 9.943644478207876e-06, + "loss": 0.4145, + "step": 3471 + }, + { + "epoch": 0.07642526964560863, + "grad_norm": 2.2188704188670534, + "learning_rate": 9.943591098076184e-06, + "loss": 0.3683, + "step": 3472 + }, + { + "epoch": 0.07644728153202729, + "grad_norm": 2.767065882064332, + "learning_rate": 9.94353769281896e-06, + "loss": 0.3921, + "step": 3473 + }, + { + "epoch": 0.07646929341844597, + "grad_norm": 1.782501388464197, + "learning_rate": 9.943484262436469e-06, + "loss": 0.4234, + "step": 3474 + }, + { + "epoch": 0.07649130530486463, + "grad_norm": 2.043048183153255, + "learning_rate": 9.943430806928989e-06, + "loss": 0.3347, + "step": 3475 + }, + { + "epoch": 0.07651331719128329, + "grad_norm": 1.149071627442975, + "learning_rate": 9.943377326296787e-06, + "loss": 0.6121, + "step": 3476 + }, + { + "epoch": 0.07653532907770196, + "grad_norm": 1.7284001212045903, + "learning_rate": 9.943323820540137e-06, + "loss": 0.4044, + "step": 3477 + }, + { + "epoch": 0.07655734096412063, + "grad_norm": 8.097691655104171, + "learning_rate": 9.94327028965931e-06, + "loss": 0.4669, + "step": 3478 + }, + { + "epoch": 0.0765793528505393, + "grad_norm": 0.7860321698184233, + "learning_rate": 9.94321673365458e-06, + "loss": 0.6005, + "step": 3479 + }, + { + "epoch": 0.07660136473695796, + "grad_norm": 5.728539332677127, + "learning_rate": 9.943163152526218e-06, + "loss": 0.3883, + "step": 3480 + }, + { + "epoch": 0.07662337662337662, + "grad_norm": 1.7932398808838608, + "learning_rate": 9.943109546274495e-06, + "loss": 0.4219, + "step": 3481 + }, + { + "epoch": 0.07664538850979528, + "grad_norm": 2.437739906667994, + "learning_rate": 9.943055914899683e-06, + "loss": 0.3215, + "step": 3482 + }, + { + "epoch": 0.07666740039621396, + "grad_norm": 0.8008485731699775, + "learning_rate": 9.943002258402061e-06, + "loss": 0.5504, + "step": 3483 + }, + { + "epoch": 0.07668941228263262, + "grad_norm": 1.9192378987008378, + "learning_rate": 9.942948576781892e-06, + "loss": 0.4468, + "step": 3484 + }, + { + "epoch": 0.07671142416905129, + "grad_norm": 1.5202027347552092, + "learning_rate": 9.942894870039456e-06, + "loss": 0.3355, + "step": 3485 + }, + { + "epoch": 0.07673343605546995, + "grad_norm": 0.8233768898028023, + "learning_rate": 9.942841138175024e-06, + "loss": 0.5786, + "step": 3486 + }, + { + "epoch": 0.07675544794188861, + "grad_norm": 2.9908774433151017, + "learning_rate": 9.942787381188869e-06, + "loss": 0.2967, + "step": 3487 + }, + { + "epoch": 0.07677745982830729, + "grad_norm": 4.837390760089506, + "learning_rate": 9.942733599081263e-06, + "loss": 0.428, + "step": 3488 + }, + { + "epoch": 0.07679947171472595, + "grad_norm": 2.74620879760065, + "learning_rate": 9.94267979185248e-06, + "loss": 0.4041, + "step": 3489 + }, + { + "epoch": 0.07682148360114462, + "grad_norm": 1.828966845991808, + "learning_rate": 9.942625959502794e-06, + "loss": 0.3952, + "step": 3490 + }, + { + "epoch": 0.07684349548756328, + "grad_norm": 3.3463577936208586, + "learning_rate": 9.94257210203248e-06, + "loss": 0.4033, + "step": 3491 + }, + { + "epoch": 0.07686550737398196, + "grad_norm": 0.7330453655698057, + "learning_rate": 9.942518219441808e-06, + "loss": 0.5563, + "step": 3492 + }, + { + "epoch": 0.07688751926040062, + "grad_norm": 2.968691766730939, + "learning_rate": 9.942464311731055e-06, + "loss": 0.3898, + "step": 3493 + }, + { + "epoch": 0.07690953114681928, + "grad_norm": 1.5255608106134375, + "learning_rate": 9.942410378900493e-06, + "loss": 0.3517, + "step": 3494 + }, + { + "epoch": 0.07693154303323795, + "grad_norm": 1.4058568932048003, + "learning_rate": 9.942356420950399e-06, + "loss": 0.3608, + "step": 3495 + }, + { + "epoch": 0.07695355491965661, + "grad_norm": 2.448768213647094, + "learning_rate": 9.942302437881045e-06, + "loss": 0.3523, + "step": 3496 + }, + { + "epoch": 0.07697556680607529, + "grad_norm": 2.4560455205292016, + "learning_rate": 9.942248429692703e-06, + "loss": 0.3421, + "step": 3497 + }, + { + "epoch": 0.07699757869249395, + "grad_norm": 1.5131062451912014, + "learning_rate": 9.942194396385652e-06, + "loss": 0.3506, + "step": 3498 + }, + { + "epoch": 0.07701959057891261, + "grad_norm": 2.080135230938836, + "learning_rate": 9.942140337960166e-06, + "loss": 0.4359, + "step": 3499 + }, + { + "epoch": 0.07704160246533127, + "grad_norm": 2.3953534682546813, + "learning_rate": 9.942086254416516e-06, + "loss": 0.3318, + "step": 3500 + }, + { + "epoch": 0.07706361435174995, + "grad_norm": 0.6314652144679709, + "learning_rate": 9.942032145754982e-06, + "loss": 0.5701, + "step": 3501 + }, + { + "epoch": 0.07708562623816861, + "grad_norm": 0.6570048419934067, + "learning_rate": 9.941978011975835e-06, + "loss": 0.6154, + "step": 3502 + }, + { + "epoch": 0.07710763812458728, + "grad_norm": 1.641132874865518, + "learning_rate": 9.941923853079351e-06, + "loss": 0.3401, + "step": 3503 + }, + { + "epoch": 0.07712965001100594, + "grad_norm": 4.08648891433332, + "learning_rate": 9.941869669065806e-06, + "loss": 0.3902, + "step": 3504 + }, + { + "epoch": 0.0771516618974246, + "grad_norm": 1.5261035478082245, + "learning_rate": 9.941815459935478e-06, + "loss": 0.3159, + "step": 3505 + }, + { + "epoch": 0.07717367378384328, + "grad_norm": 1.8261160032652992, + "learning_rate": 9.941761225688637e-06, + "loss": 0.3715, + "step": 3506 + }, + { + "epoch": 0.07719568567026194, + "grad_norm": 1.554903192381167, + "learning_rate": 9.941706966325562e-06, + "loss": 0.2977, + "step": 3507 + }, + { + "epoch": 0.0772176975566806, + "grad_norm": 1.8712702596740496, + "learning_rate": 9.941652681846527e-06, + "loss": 0.521, + "step": 3508 + }, + { + "epoch": 0.07723970944309927, + "grad_norm": 1.9257112234166776, + "learning_rate": 9.941598372251811e-06, + "loss": 0.4084, + "step": 3509 + }, + { + "epoch": 0.07726172132951795, + "grad_norm": 2.1849656900570777, + "learning_rate": 9.941544037541688e-06, + "loss": 0.3476, + "step": 3510 + }, + { + "epoch": 0.07728373321593661, + "grad_norm": 2.2910134873261883, + "learning_rate": 9.94148967771643e-06, + "loss": 0.3947, + "step": 3511 + }, + { + "epoch": 0.07730574510235527, + "grad_norm": 2.768073762644742, + "learning_rate": 9.941435292776322e-06, + "loss": 0.3737, + "step": 3512 + }, + { + "epoch": 0.07732775698877394, + "grad_norm": 2.2979774772184203, + "learning_rate": 9.941380882721637e-06, + "loss": 0.4248, + "step": 3513 + }, + { + "epoch": 0.0773497688751926, + "grad_norm": 0.7073277470720553, + "learning_rate": 9.941326447552647e-06, + "loss": 0.5646, + "step": 3514 + }, + { + "epoch": 0.07737178076161128, + "grad_norm": 1.641586477002389, + "learning_rate": 9.941271987269634e-06, + "loss": 0.448, + "step": 3515 + }, + { + "epoch": 0.07739379264802994, + "grad_norm": 2.887552665505363, + "learning_rate": 9.941217501872871e-06, + "loss": 0.3736, + "step": 3516 + }, + { + "epoch": 0.0774158045344486, + "grad_norm": 1.8727672724324733, + "learning_rate": 9.941162991362638e-06, + "loss": 0.3519, + "step": 3517 + }, + { + "epoch": 0.07743781642086726, + "grad_norm": 2.1049955454796487, + "learning_rate": 9.941108455739212e-06, + "loss": 0.4141, + "step": 3518 + }, + { + "epoch": 0.07745982830728594, + "grad_norm": 1.5809243760447549, + "learning_rate": 9.941053895002868e-06, + "loss": 0.3688, + "step": 3519 + }, + { + "epoch": 0.0774818401937046, + "grad_norm": 2.3968658236306717, + "learning_rate": 9.940999309153884e-06, + "loss": 0.3342, + "step": 3520 + }, + { + "epoch": 0.07750385208012327, + "grad_norm": 2.861790977076157, + "learning_rate": 9.940944698192537e-06, + "loss": 0.3797, + "step": 3521 + }, + { + "epoch": 0.07752586396654193, + "grad_norm": 1.8173699032452133, + "learning_rate": 9.940890062119107e-06, + "loss": 0.416, + "step": 3522 + }, + { + "epoch": 0.0775478758529606, + "grad_norm": 2.247725410216292, + "learning_rate": 9.94083540093387e-06, + "loss": 0.3265, + "step": 3523 + }, + { + "epoch": 0.07756988773937927, + "grad_norm": 0.6875607485156836, + "learning_rate": 9.940780714637102e-06, + "loss": 0.6301, + "step": 3524 + }, + { + "epoch": 0.07759189962579793, + "grad_norm": 1.8744095362612843, + "learning_rate": 9.940726003229083e-06, + "loss": 0.3668, + "step": 3525 + }, + { + "epoch": 0.0776139115122166, + "grad_norm": 2.2747866851864615, + "learning_rate": 9.940671266710092e-06, + "loss": 0.3155, + "step": 3526 + }, + { + "epoch": 0.07763592339863526, + "grad_norm": 0.644949799063826, + "learning_rate": 9.940616505080407e-06, + "loss": 0.6008, + "step": 3527 + }, + { + "epoch": 0.07765793528505392, + "grad_norm": 2.060442293264464, + "learning_rate": 9.940561718340304e-06, + "loss": 0.3622, + "step": 3528 + }, + { + "epoch": 0.0776799471714726, + "grad_norm": 1.973569574926799, + "learning_rate": 9.940506906490061e-06, + "loss": 0.3837, + "step": 3529 + }, + { + "epoch": 0.07770195905789126, + "grad_norm": 2.783888526272308, + "learning_rate": 9.940452069529963e-06, + "loss": 0.363, + "step": 3530 + }, + { + "epoch": 0.07772397094430993, + "grad_norm": 1.95028735372745, + "learning_rate": 9.940397207460279e-06, + "loss": 0.4161, + "step": 3531 + }, + { + "epoch": 0.07774598283072859, + "grad_norm": 2.291502928686395, + "learning_rate": 9.940342320281296e-06, + "loss": 0.318, + "step": 3532 + }, + { + "epoch": 0.07776799471714727, + "grad_norm": 1.5035228630479265, + "learning_rate": 9.94028740799329e-06, + "loss": 0.382, + "step": 3533 + }, + { + "epoch": 0.07779000660356593, + "grad_norm": 18.513042240515176, + "learning_rate": 9.940232470596538e-06, + "loss": 0.4366, + "step": 3534 + }, + { + "epoch": 0.07781201848998459, + "grad_norm": 1.4032130099021987, + "learning_rate": 9.940177508091321e-06, + "loss": 0.3008, + "step": 3535 + }, + { + "epoch": 0.07783403037640325, + "grad_norm": 1.654040806230774, + "learning_rate": 9.940122520477923e-06, + "loss": 0.3789, + "step": 3536 + }, + { + "epoch": 0.07785604226282192, + "grad_norm": 1.7878682372823527, + "learning_rate": 9.940067507756614e-06, + "loss": 0.4487, + "step": 3537 + }, + { + "epoch": 0.0778780541492406, + "grad_norm": 2.008742690771669, + "learning_rate": 9.94001246992768e-06, + "loss": 0.3907, + "step": 3538 + }, + { + "epoch": 0.07790006603565926, + "grad_norm": 2.24723540408078, + "learning_rate": 9.9399574069914e-06, + "loss": 0.465, + "step": 3539 + }, + { + "epoch": 0.07792207792207792, + "grad_norm": 2.0934148892936446, + "learning_rate": 9.939902318948054e-06, + "loss": 0.4203, + "step": 3540 + }, + { + "epoch": 0.07794408980849658, + "grad_norm": 0.7647974383582818, + "learning_rate": 9.93984720579792e-06, + "loss": 0.6273, + "step": 3541 + }, + { + "epoch": 0.07796610169491526, + "grad_norm": 1.6872931602395949, + "learning_rate": 9.939792067541278e-06, + "loss": 0.332, + "step": 3542 + }, + { + "epoch": 0.07798811358133392, + "grad_norm": 2.029778749892507, + "learning_rate": 9.93973690417841e-06, + "loss": 0.3597, + "step": 3543 + }, + { + "epoch": 0.07801012546775259, + "grad_norm": 3.07636923929464, + "learning_rate": 9.939681715709597e-06, + "loss": 0.3847, + "step": 3544 + }, + { + "epoch": 0.07803213735417125, + "grad_norm": 1.5039051164114254, + "learning_rate": 9.939626502135117e-06, + "loss": 0.3896, + "step": 3545 + }, + { + "epoch": 0.07805414924058991, + "grad_norm": 0.6304141516097813, + "learning_rate": 9.939571263455254e-06, + "loss": 0.5881, + "step": 3546 + }, + { + "epoch": 0.07807616112700859, + "grad_norm": 1.4342087704011632, + "learning_rate": 9.939515999670286e-06, + "loss": 0.3868, + "step": 3547 + }, + { + "epoch": 0.07809817301342725, + "grad_norm": 1.5552645690483842, + "learning_rate": 9.939460710780495e-06, + "loss": 0.3847, + "step": 3548 + }, + { + "epoch": 0.07812018489984592, + "grad_norm": 1.4715318320564144, + "learning_rate": 9.93940539678616e-06, + "loss": 0.3719, + "step": 3549 + }, + { + "epoch": 0.07814219678626458, + "grad_norm": 1.5211852568980428, + "learning_rate": 9.939350057687565e-06, + "loss": 0.4139, + "step": 3550 + }, + { + "epoch": 0.07816420867268326, + "grad_norm": 2.014871784707209, + "learning_rate": 9.93929469348499e-06, + "loss": 0.3783, + "step": 3551 + }, + { + "epoch": 0.07818622055910192, + "grad_norm": 0.6734504942387279, + "learning_rate": 9.939239304178716e-06, + "loss": 0.5905, + "step": 3552 + }, + { + "epoch": 0.07820823244552058, + "grad_norm": 0.6850793852304728, + "learning_rate": 9.939183889769025e-06, + "loss": 0.5781, + "step": 3553 + }, + { + "epoch": 0.07823024433193924, + "grad_norm": 1.5442982337683184, + "learning_rate": 9.939128450256199e-06, + "loss": 0.3731, + "step": 3554 + }, + { + "epoch": 0.07825225621835791, + "grad_norm": 1.5680359017104406, + "learning_rate": 9.939072985640519e-06, + "loss": 0.304, + "step": 3555 + }, + { + "epoch": 0.07827426810477658, + "grad_norm": 1.6434345117855784, + "learning_rate": 9.939017495922267e-06, + "loss": 0.4051, + "step": 3556 + }, + { + "epoch": 0.07829627999119525, + "grad_norm": 2.2801353066962244, + "learning_rate": 9.938961981101725e-06, + "loss": 0.3717, + "step": 3557 + }, + { + "epoch": 0.07831829187761391, + "grad_norm": 1.8678240440090617, + "learning_rate": 9.938906441179174e-06, + "loss": 0.2984, + "step": 3558 + }, + { + "epoch": 0.07834030376403257, + "grad_norm": 1.645741920663737, + "learning_rate": 9.9388508761549e-06, + "loss": 0.3494, + "step": 3559 + }, + { + "epoch": 0.07836231565045125, + "grad_norm": 1.4764919701434995, + "learning_rate": 9.938795286029182e-06, + "loss": 0.3553, + "step": 3560 + }, + { + "epoch": 0.07838432753686991, + "grad_norm": 1.4824872586315536, + "learning_rate": 9.938739670802305e-06, + "loss": 0.3855, + "step": 3561 + }, + { + "epoch": 0.07840633942328858, + "grad_norm": 2.0779473829361206, + "learning_rate": 9.938684030474548e-06, + "loss": 0.3687, + "step": 3562 + }, + { + "epoch": 0.07842835130970724, + "grad_norm": 0.8339468634499243, + "learning_rate": 9.938628365046197e-06, + "loss": 0.6095, + "step": 3563 + }, + { + "epoch": 0.0784503631961259, + "grad_norm": 2.018197392934252, + "learning_rate": 9.938572674517535e-06, + "loss": 0.3415, + "step": 3564 + }, + { + "epoch": 0.07847237508254458, + "grad_norm": 2.1198591729645435, + "learning_rate": 9.938516958888844e-06, + "loss": 0.3144, + "step": 3565 + }, + { + "epoch": 0.07849438696896324, + "grad_norm": 1.5796525123952188, + "learning_rate": 9.938461218160405e-06, + "loss": 0.4168, + "step": 3566 + }, + { + "epoch": 0.0785163988553819, + "grad_norm": 2.271017072219929, + "learning_rate": 9.938405452332504e-06, + "loss": 0.3854, + "step": 3567 + }, + { + "epoch": 0.07853841074180057, + "grad_norm": 1.6682327764548812, + "learning_rate": 9.938349661405426e-06, + "loss": 0.3398, + "step": 3568 + }, + { + "epoch": 0.07856042262821925, + "grad_norm": 0.6634039433163609, + "learning_rate": 9.938293845379451e-06, + "loss": 0.5644, + "step": 3569 + }, + { + "epoch": 0.07858243451463791, + "grad_norm": 1.4851812175042438, + "learning_rate": 9.938238004254865e-06, + "loss": 0.3373, + "step": 3570 + }, + { + "epoch": 0.07860444640105657, + "grad_norm": 1.6892694290784485, + "learning_rate": 9.938182138031949e-06, + "loss": 0.3835, + "step": 3571 + }, + { + "epoch": 0.07862645828747523, + "grad_norm": 0.6760642450124749, + "learning_rate": 9.93812624671099e-06, + "loss": 0.5947, + "step": 3572 + }, + { + "epoch": 0.0786484701738939, + "grad_norm": 1.8558017179485713, + "learning_rate": 9.938070330292272e-06, + "loss": 0.4445, + "step": 3573 + }, + { + "epoch": 0.07867048206031257, + "grad_norm": 0.6504781915738727, + "learning_rate": 9.938014388776077e-06, + "loss": 0.5752, + "step": 3574 + }, + { + "epoch": 0.07869249394673124, + "grad_norm": 2.015860706126626, + "learning_rate": 9.93795842216269e-06, + "loss": 0.4265, + "step": 3575 + }, + { + "epoch": 0.0787145058331499, + "grad_norm": 1.362122467170608, + "learning_rate": 9.937902430452397e-06, + "loss": 0.3919, + "step": 3576 + }, + { + "epoch": 0.07873651771956856, + "grad_norm": 0.6405820043853546, + "learning_rate": 9.937846413645481e-06, + "loss": 0.5507, + "step": 3577 + }, + { + "epoch": 0.07875852960598723, + "grad_norm": 0.6422898836792024, + "learning_rate": 9.937790371742226e-06, + "loss": 0.5749, + "step": 3578 + }, + { + "epoch": 0.0787805414924059, + "grad_norm": 2.175706174193613, + "learning_rate": 9.937734304742921e-06, + "loss": 0.4491, + "step": 3579 + }, + { + "epoch": 0.07880255337882457, + "grad_norm": 1.9578820475916536, + "learning_rate": 9.937678212647846e-06, + "loss": 0.3909, + "step": 3580 + }, + { + "epoch": 0.07882456526524323, + "grad_norm": 4.441245387162005, + "learning_rate": 9.937622095457287e-06, + "loss": 0.3935, + "step": 3581 + }, + { + "epoch": 0.07884657715166189, + "grad_norm": 2.364479157966138, + "learning_rate": 9.937565953171533e-06, + "loss": 0.4149, + "step": 3582 + }, + { + "epoch": 0.07886858903808057, + "grad_norm": 1.712139792055428, + "learning_rate": 9.937509785790865e-06, + "loss": 0.3823, + "step": 3583 + }, + { + "epoch": 0.07889060092449923, + "grad_norm": 2.5735919281319943, + "learning_rate": 9.937453593315569e-06, + "loss": 0.4543, + "step": 3584 + }, + { + "epoch": 0.0789126128109179, + "grad_norm": 1.7241162731838078, + "learning_rate": 9.937397375745934e-06, + "loss": 0.3012, + "step": 3585 + }, + { + "epoch": 0.07893462469733656, + "grad_norm": 0.6511387082937728, + "learning_rate": 9.937341133082242e-06, + "loss": 0.5578, + "step": 3586 + }, + { + "epoch": 0.07895663658375522, + "grad_norm": 0.669630979298875, + "learning_rate": 9.937284865324781e-06, + "loss": 0.584, + "step": 3587 + }, + { + "epoch": 0.0789786484701739, + "grad_norm": 1.6446964654522902, + "learning_rate": 9.937228572473836e-06, + "loss": 0.3352, + "step": 3588 + }, + { + "epoch": 0.07900066035659256, + "grad_norm": 2.0022719287313557, + "learning_rate": 9.937172254529693e-06, + "loss": 0.319, + "step": 3589 + }, + { + "epoch": 0.07902267224301122, + "grad_norm": 2.2878726145456434, + "learning_rate": 9.937115911492639e-06, + "loss": 0.3147, + "step": 3590 + }, + { + "epoch": 0.07904468412942989, + "grad_norm": 4.196077444418339, + "learning_rate": 9.937059543362958e-06, + "loss": 0.422, + "step": 3591 + }, + { + "epoch": 0.07906669601584856, + "grad_norm": 1.9705596916789498, + "learning_rate": 9.93700315014094e-06, + "loss": 0.4872, + "step": 3592 + }, + { + "epoch": 0.07908870790226723, + "grad_norm": 1.4878207445870453, + "learning_rate": 9.93694673182687e-06, + "loss": 0.4629, + "step": 3593 + }, + { + "epoch": 0.07911071978868589, + "grad_norm": 4.04914324479365, + "learning_rate": 9.936890288421035e-06, + "loss": 0.4093, + "step": 3594 + }, + { + "epoch": 0.07913273167510455, + "grad_norm": 2.183392301481257, + "learning_rate": 9.936833819923722e-06, + "loss": 0.4643, + "step": 3595 + }, + { + "epoch": 0.07915474356152322, + "grad_norm": 1.821578717692201, + "learning_rate": 9.936777326335216e-06, + "loss": 0.4222, + "step": 3596 + }, + { + "epoch": 0.0791767554479419, + "grad_norm": 3.139570292716593, + "learning_rate": 9.936720807655807e-06, + "loss": 0.4686, + "step": 3597 + }, + { + "epoch": 0.07919876733436056, + "grad_norm": 1.9563859118298659, + "learning_rate": 9.936664263885782e-06, + "loss": 0.4722, + "step": 3598 + }, + { + "epoch": 0.07922077922077922, + "grad_norm": 0.754966640838158, + "learning_rate": 9.936607695025426e-06, + "loss": 0.5845, + "step": 3599 + }, + { + "epoch": 0.07924279110719788, + "grad_norm": 0.7282910689665344, + "learning_rate": 9.936551101075029e-06, + "loss": 0.5862, + "step": 3600 + }, + { + "epoch": 0.07926480299361656, + "grad_norm": 2.016808732044164, + "learning_rate": 9.936494482034875e-06, + "loss": 0.3736, + "step": 3601 + }, + { + "epoch": 0.07928681488003522, + "grad_norm": 1.43577027322466, + "learning_rate": 9.936437837905257e-06, + "loss": 0.315, + "step": 3602 + }, + { + "epoch": 0.07930882676645389, + "grad_norm": 1.998082345229954, + "learning_rate": 9.936381168686459e-06, + "loss": 0.3472, + "step": 3603 + }, + { + "epoch": 0.07933083865287255, + "grad_norm": 1.6015109121448625, + "learning_rate": 9.93632447437877e-06, + "loss": 0.4704, + "step": 3604 + }, + { + "epoch": 0.07935285053929121, + "grad_norm": 1.6677943784682723, + "learning_rate": 9.936267754982478e-06, + "loss": 0.3499, + "step": 3605 + }, + { + "epoch": 0.07937486242570989, + "grad_norm": 1.5031806898334419, + "learning_rate": 9.936211010497874e-06, + "loss": 0.3248, + "step": 3606 + }, + { + "epoch": 0.07939687431212855, + "grad_norm": 1.5192843008873516, + "learning_rate": 9.936154240925243e-06, + "loss": 0.3144, + "step": 3607 + }, + { + "epoch": 0.07941888619854721, + "grad_norm": 1.0000689589480563, + "learning_rate": 9.936097446264873e-06, + "loss": 0.5841, + "step": 3608 + }, + { + "epoch": 0.07944089808496588, + "grad_norm": 3.9765911929556257, + "learning_rate": 9.936040626517054e-06, + "loss": 0.4367, + "step": 3609 + }, + { + "epoch": 0.07946290997138455, + "grad_norm": 2.313832302376944, + "learning_rate": 9.935983781682076e-06, + "loss": 0.3762, + "step": 3610 + }, + { + "epoch": 0.07948492185780322, + "grad_norm": 1.6803729520327007, + "learning_rate": 9.935926911760227e-06, + "loss": 0.4123, + "step": 3611 + }, + { + "epoch": 0.07950693374422188, + "grad_norm": 0.6619972663743319, + "learning_rate": 9.935870016751798e-06, + "loss": 0.6112, + "step": 3612 + }, + { + "epoch": 0.07952894563064054, + "grad_norm": 1.441401966145798, + "learning_rate": 9.935813096657074e-06, + "loss": 0.313, + "step": 3613 + }, + { + "epoch": 0.0795509575170592, + "grad_norm": 1.8344446966949208, + "learning_rate": 9.935756151476346e-06, + "loss": 0.4264, + "step": 3614 + }, + { + "epoch": 0.07957296940347788, + "grad_norm": 1.8016805378576932, + "learning_rate": 9.935699181209905e-06, + "loss": 0.3318, + "step": 3615 + }, + { + "epoch": 0.07959498128989655, + "grad_norm": 1.3777016794081884, + "learning_rate": 9.93564218585804e-06, + "loss": 0.3823, + "step": 3616 + }, + { + "epoch": 0.07961699317631521, + "grad_norm": 0.7731240133367655, + "learning_rate": 9.935585165421038e-06, + "loss": 0.5795, + "step": 3617 + }, + { + "epoch": 0.07963900506273387, + "grad_norm": 1.832359972112448, + "learning_rate": 9.93552811989919e-06, + "loss": 0.4662, + "step": 3618 + }, + { + "epoch": 0.07966101694915254, + "grad_norm": 1.5578326086257546, + "learning_rate": 9.935471049292788e-06, + "loss": 0.4352, + "step": 3619 + }, + { + "epoch": 0.07968302883557121, + "grad_norm": 1.5136633001862565, + "learning_rate": 9.935413953602122e-06, + "loss": 0.2975, + "step": 3620 + }, + { + "epoch": 0.07970504072198988, + "grad_norm": 1.6652532729879972, + "learning_rate": 9.93535683282748e-06, + "loss": 0.46, + "step": 3621 + }, + { + "epoch": 0.07972705260840854, + "grad_norm": 1.9576036283338036, + "learning_rate": 9.935299686969155e-06, + "loss": 0.42, + "step": 3622 + }, + { + "epoch": 0.0797490644948272, + "grad_norm": 1.4486226118295142, + "learning_rate": 9.935242516027434e-06, + "loss": 0.3858, + "step": 3623 + }, + { + "epoch": 0.07977107638124588, + "grad_norm": 1.6017388844444913, + "learning_rate": 9.93518532000261e-06, + "loss": 0.3236, + "step": 3624 + }, + { + "epoch": 0.07979308826766454, + "grad_norm": 1.410318243472537, + "learning_rate": 9.935128098894973e-06, + "loss": 0.3817, + "step": 3625 + }, + { + "epoch": 0.0798151001540832, + "grad_norm": 0.6695197862797013, + "learning_rate": 9.935070852704814e-06, + "loss": 0.5726, + "step": 3626 + }, + { + "epoch": 0.07983711204050187, + "grad_norm": 1.5474226198432297, + "learning_rate": 9.935013581432424e-06, + "loss": 0.2584, + "step": 3627 + }, + { + "epoch": 0.07985912392692053, + "grad_norm": 0.6301533191532643, + "learning_rate": 9.934956285078092e-06, + "loss": 0.5742, + "step": 3628 + }, + { + "epoch": 0.07988113581333921, + "grad_norm": 1.7749386860142822, + "learning_rate": 9.934898963642113e-06, + "loss": 0.4533, + "step": 3629 + }, + { + "epoch": 0.07990314769975787, + "grad_norm": 1.729669588210487, + "learning_rate": 9.934841617124777e-06, + "loss": 0.4461, + "step": 3630 + }, + { + "epoch": 0.07992515958617653, + "grad_norm": 3.260793205872838, + "learning_rate": 9.934784245526374e-06, + "loss": 0.4052, + "step": 3631 + }, + { + "epoch": 0.0799471714725952, + "grad_norm": 1.8821924524511673, + "learning_rate": 9.934726848847196e-06, + "loss": 0.3538, + "step": 3632 + }, + { + "epoch": 0.07996918335901387, + "grad_norm": 1.768435179086859, + "learning_rate": 9.934669427087536e-06, + "loss": 0.4282, + "step": 3633 + }, + { + "epoch": 0.07999119524543254, + "grad_norm": 2.5989592674678534, + "learning_rate": 9.934611980247684e-06, + "loss": 0.4505, + "step": 3634 + }, + { + "epoch": 0.0800132071318512, + "grad_norm": 0.720820869388489, + "learning_rate": 9.934554508327933e-06, + "loss": 0.5708, + "step": 3635 + }, + { + "epoch": 0.08003521901826986, + "grad_norm": 1.4835841316805902, + "learning_rate": 9.934497011328577e-06, + "loss": 0.354, + "step": 3636 + }, + { + "epoch": 0.08005723090468853, + "grad_norm": 1.4446988944049313, + "learning_rate": 9.934439489249906e-06, + "loss": 0.3507, + "step": 3637 + }, + { + "epoch": 0.0800792427911072, + "grad_norm": 1.5688565693174372, + "learning_rate": 9.93438194209221e-06, + "loss": 0.4106, + "step": 3638 + }, + { + "epoch": 0.08010125467752587, + "grad_norm": 1.9535022977500343, + "learning_rate": 9.934324369855788e-06, + "loss": 0.3341, + "step": 3639 + }, + { + "epoch": 0.08012326656394453, + "grad_norm": 0.6761425302241101, + "learning_rate": 9.934266772540925e-06, + "loss": 0.5947, + "step": 3640 + }, + { + "epoch": 0.08014527845036319, + "grad_norm": 2.2844128755080617, + "learning_rate": 9.93420915014792e-06, + "loss": 0.4025, + "step": 3641 + }, + { + "epoch": 0.08016729033678187, + "grad_norm": 1.4409960603425238, + "learning_rate": 9.934151502677062e-06, + "loss": 0.3613, + "step": 3642 + }, + { + "epoch": 0.08018930222320053, + "grad_norm": 1.5469967837009582, + "learning_rate": 9.934093830128648e-06, + "loss": 0.3844, + "step": 3643 + }, + { + "epoch": 0.0802113141096192, + "grad_norm": 1.510305799180945, + "learning_rate": 9.934036132502965e-06, + "loss": 0.369, + "step": 3644 + }, + { + "epoch": 0.08023332599603786, + "grad_norm": 1.7075432526619092, + "learning_rate": 9.933978409800311e-06, + "loss": 0.3595, + "step": 3645 + }, + { + "epoch": 0.08025533788245652, + "grad_norm": 0.6754370501593886, + "learning_rate": 9.93392066202098e-06, + "loss": 0.5764, + "step": 3646 + }, + { + "epoch": 0.0802773497688752, + "grad_norm": 2.053144931730013, + "learning_rate": 9.933862889165261e-06, + "loss": 0.4067, + "step": 3647 + }, + { + "epoch": 0.08029936165529386, + "grad_norm": 1.6694221842112746, + "learning_rate": 9.933805091233451e-06, + "loss": 0.3728, + "step": 3648 + }, + { + "epoch": 0.08032137354171252, + "grad_norm": 0.630031715756448, + "learning_rate": 9.933747268225844e-06, + "loss": 0.5626, + "step": 3649 + }, + { + "epoch": 0.08034338542813119, + "grad_norm": 0.6111093906730306, + "learning_rate": 9.93368942014273e-06, + "loss": 0.581, + "step": 3650 + }, + { + "epoch": 0.08036539731454986, + "grad_norm": 2.734337328059369, + "learning_rate": 9.93363154698441e-06, + "loss": 0.4052, + "step": 3651 + }, + { + "epoch": 0.08038740920096853, + "grad_norm": 1.8220407622826036, + "learning_rate": 9.933573648751171e-06, + "loss": 0.3246, + "step": 3652 + }, + { + "epoch": 0.08040942108738719, + "grad_norm": 0.6494249038458568, + "learning_rate": 9.933515725443313e-06, + "loss": 0.5949, + "step": 3653 + }, + { + "epoch": 0.08043143297380585, + "grad_norm": 1.2749302675753287, + "learning_rate": 9.933457777061126e-06, + "loss": 0.4127, + "step": 3654 + }, + { + "epoch": 0.08045344486022452, + "grad_norm": 2.1206903049574404, + "learning_rate": 9.933399803604905e-06, + "loss": 0.4115, + "step": 3655 + }, + { + "epoch": 0.08047545674664319, + "grad_norm": 1.6843567906226222, + "learning_rate": 9.933341805074947e-06, + "loss": 0.3979, + "step": 3656 + }, + { + "epoch": 0.08049746863306186, + "grad_norm": 1.607633561128785, + "learning_rate": 9.933283781471546e-06, + "loss": 0.3121, + "step": 3657 + }, + { + "epoch": 0.08051948051948052, + "grad_norm": 1.733746604403302, + "learning_rate": 9.933225732794996e-06, + "loss": 0.4539, + "step": 3658 + }, + { + "epoch": 0.08054149240589918, + "grad_norm": 0.6802817918338635, + "learning_rate": 9.933167659045592e-06, + "loss": 0.5905, + "step": 3659 + }, + { + "epoch": 0.08056350429231786, + "grad_norm": 1.6143415023897272, + "learning_rate": 9.93310956022363e-06, + "loss": 0.3451, + "step": 3660 + }, + { + "epoch": 0.08058551617873652, + "grad_norm": 1.5260185938735051, + "learning_rate": 9.933051436329406e-06, + "loss": 0.3606, + "step": 3661 + }, + { + "epoch": 0.08060752806515518, + "grad_norm": 1.8788096319029581, + "learning_rate": 9.932993287363214e-06, + "loss": 0.3954, + "step": 3662 + }, + { + "epoch": 0.08062953995157385, + "grad_norm": 1.2870988431446284, + "learning_rate": 9.93293511332535e-06, + "loss": 0.3699, + "step": 3663 + }, + { + "epoch": 0.08065155183799251, + "grad_norm": 1.495044161346804, + "learning_rate": 9.93287691421611e-06, + "loss": 0.4172, + "step": 3664 + }, + { + "epoch": 0.08067356372441119, + "grad_norm": 1.9575183750417418, + "learning_rate": 9.932818690035789e-06, + "loss": 0.4118, + "step": 3665 + }, + { + "epoch": 0.08069557561082985, + "grad_norm": 1.6626163751910243, + "learning_rate": 9.932760440784682e-06, + "loss": 0.4014, + "step": 3666 + }, + { + "epoch": 0.08071758749724851, + "grad_norm": 1.8172270983301957, + "learning_rate": 9.932702166463087e-06, + "loss": 0.3649, + "step": 3667 + }, + { + "epoch": 0.08073959938366718, + "grad_norm": 1.6053385901650172, + "learning_rate": 9.9326438670713e-06, + "loss": 0.3056, + "step": 3668 + }, + { + "epoch": 0.08076161127008584, + "grad_norm": 1.604561844804861, + "learning_rate": 9.932585542609617e-06, + "loss": 0.3708, + "step": 3669 + }, + { + "epoch": 0.08078362315650452, + "grad_norm": 1.62568753303361, + "learning_rate": 9.932527193078334e-06, + "loss": 0.3575, + "step": 3670 + }, + { + "epoch": 0.08080563504292318, + "grad_norm": 0.6929982630499506, + "learning_rate": 9.932468818477748e-06, + "loss": 0.6019, + "step": 3671 + }, + { + "epoch": 0.08082764692934184, + "grad_norm": 3.5219574404292633, + "learning_rate": 9.932410418808155e-06, + "loss": 0.3916, + "step": 3672 + }, + { + "epoch": 0.0808496588157605, + "grad_norm": 2.2463065262200503, + "learning_rate": 9.932351994069853e-06, + "loss": 0.3589, + "step": 3673 + }, + { + "epoch": 0.08087167070217918, + "grad_norm": 1.6332375381681186, + "learning_rate": 9.932293544263138e-06, + "loss": 0.3926, + "step": 3674 + }, + { + "epoch": 0.08089368258859785, + "grad_norm": 1.4419979390830733, + "learning_rate": 9.932235069388307e-06, + "loss": 0.3679, + "step": 3675 + }, + { + "epoch": 0.08091569447501651, + "grad_norm": 0.6660771571052362, + "learning_rate": 9.932176569445658e-06, + "loss": 0.5983, + "step": 3676 + }, + { + "epoch": 0.08093770636143517, + "grad_norm": 3.00339089967232, + "learning_rate": 9.932118044435488e-06, + "loss": 0.333, + "step": 3677 + }, + { + "epoch": 0.08095971824785383, + "grad_norm": 1.7380903918916455, + "learning_rate": 9.932059494358095e-06, + "loss": 0.3925, + "step": 3678 + }, + { + "epoch": 0.08098173013427251, + "grad_norm": 0.6664973842457423, + "learning_rate": 9.932000919213774e-06, + "loss": 0.5658, + "step": 3679 + }, + { + "epoch": 0.08100374202069117, + "grad_norm": 1.5963207623624482, + "learning_rate": 9.931942319002825e-06, + "loss": 0.4295, + "step": 3680 + }, + { + "epoch": 0.08102575390710984, + "grad_norm": 2.4716622088220923, + "learning_rate": 9.931883693725546e-06, + "loss": 0.4123, + "step": 3681 + }, + { + "epoch": 0.0810477657935285, + "grad_norm": 1.4422885416740676, + "learning_rate": 9.931825043382234e-06, + "loss": 0.3138, + "step": 3682 + }, + { + "epoch": 0.08106977767994718, + "grad_norm": 1.9777779397773543, + "learning_rate": 9.931766367973188e-06, + "loss": 0.4578, + "step": 3683 + }, + { + "epoch": 0.08109178956636584, + "grad_norm": 2.5946377915494403, + "learning_rate": 9.931707667498706e-06, + "loss": 0.4287, + "step": 3684 + }, + { + "epoch": 0.0811138014527845, + "grad_norm": 1.5738046036158455, + "learning_rate": 9.931648941959087e-06, + "loss": 0.3335, + "step": 3685 + }, + { + "epoch": 0.08113581333920317, + "grad_norm": 0.7609348918118454, + "learning_rate": 9.931590191354627e-06, + "loss": 0.5972, + "step": 3686 + }, + { + "epoch": 0.08115782522562183, + "grad_norm": 1.5817033019298157, + "learning_rate": 9.931531415685625e-06, + "loss": 0.4593, + "step": 3687 + }, + { + "epoch": 0.0811798371120405, + "grad_norm": 1.7513120255186265, + "learning_rate": 9.931472614952382e-06, + "loss": 0.4024, + "step": 3688 + }, + { + "epoch": 0.08120184899845917, + "grad_norm": 1.375769155827506, + "learning_rate": 9.931413789155196e-06, + "loss": 0.2771, + "step": 3689 + }, + { + "epoch": 0.08122386088487783, + "grad_norm": 1.6114321826686984, + "learning_rate": 9.931354938294366e-06, + "loss": 0.4714, + "step": 3690 + }, + { + "epoch": 0.0812458727712965, + "grad_norm": 1.4091226967388302, + "learning_rate": 9.931296062370189e-06, + "loss": 0.2999, + "step": 3691 + }, + { + "epoch": 0.08126788465771517, + "grad_norm": 1.6691236511219874, + "learning_rate": 9.931237161382967e-06, + "loss": 0.3969, + "step": 3692 + }, + { + "epoch": 0.08128989654413384, + "grad_norm": 1.4443232812419582, + "learning_rate": 9.931178235332999e-06, + "loss": 0.3308, + "step": 3693 + }, + { + "epoch": 0.0813119084305525, + "grad_norm": 1.7327245511868368, + "learning_rate": 9.931119284220583e-06, + "loss": 0.3797, + "step": 3694 + }, + { + "epoch": 0.08133392031697116, + "grad_norm": 2.203859066581201, + "learning_rate": 9.93106030804602e-06, + "loss": 0.3772, + "step": 3695 + }, + { + "epoch": 0.08135593220338982, + "grad_norm": 1.3988042097468094, + "learning_rate": 9.931001306809609e-06, + "loss": 0.3156, + "step": 3696 + }, + { + "epoch": 0.0813779440898085, + "grad_norm": 1.3814440814236113, + "learning_rate": 9.93094228051165e-06, + "loss": 0.3499, + "step": 3697 + }, + { + "epoch": 0.08139995597622716, + "grad_norm": 1.682082663459182, + "learning_rate": 9.930883229152443e-06, + "loss": 0.4284, + "step": 3698 + }, + { + "epoch": 0.08142196786264583, + "grad_norm": 0.7201252519104764, + "learning_rate": 9.930824152732287e-06, + "loss": 0.5858, + "step": 3699 + }, + { + "epoch": 0.08144397974906449, + "grad_norm": 0.6696289563184348, + "learning_rate": 9.930765051251486e-06, + "loss": 0.5561, + "step": 3700 + }, + { + "epoch": 0.08146599163548317, + "grad_norm": 1.537988305436491, + "learning_rate": 9.930705924710336e-06, + "loss": 0.3749, + "step": 3701 + }, + { + "epoch": 0.08148800352190183, + "grad_norm": 1.876396873230783, + "learning_rate": 9.93064677310914e-06, + "loss": 0.4421, + "step": 3702 + }, + { + "epoch": 0.0815100154083205, + "grad_norm": 1.565863061656114, + "learning_rate": 9.930587596448198e-06, + "loss": 0.4215, + "step": 3703 + }, + { + "epoch": 0.08153202729473916, + "grad_norm": 0.7355469591945082, + "learning_rate": 9.93052839472781e-06, + "loss": 0.6113, + "step": 3704 + }, + { + "epoch": 0.08155403918115782, + "grad_norm": 2.5916217218093944, + "learning_rate": 9.930469167948278e-06, + "loss": 0.424, + "step": 3705 + }, + { + "epoch": 0.0815760510675765, + "grad_norm": 2.166446341498243, + "learning_rate": 9.930409916109904e-06, + "loss": 0.5536, + "step": 3706 + }, + { + "epoch": 0.08159806295399516, + "grad_norm": 2.2274606261138876, + "learning_rate": 9.930350639212987e-06, + "loss": 0.3389, + "step": 3707 + }, + { + "epoch": 0.08162007484041382, + "grad_norm": 0.681974363842489, + "learning_rate": 9.930291337257827e-06, + "loss": 0.5832, + "step": 3708 + }, + { + "epoch": 0.08164208672683249, + "grad_norm": 1.3839482567930896, + "learning_rate": 9.930232010244729e-06, + "loss": 0.4483, + "step": 3709 + }, + { + "epoch": 0.08166409861325115, + "grad_norm": 0.6498141084670958, + "learning_rate": 9.930172658173994e-06, + "loss": 0.5765, + "step": 3710 + }, + { + "epoch": 0.08168611049966983, + "grad_norm": 0.6453751807812848, + "learning_rate": 9.93011328104592e-06, + "loss": 0.5506, + "step": 3711 + }, + { + "epoch": 0.08170812238608849, + "grad_norm": 1.2922632313021054, + "learning_rate": 9.930053878860815e-06, + "loss": 0.3574, + "step": 3712 + }, + { + "epoch": 0.08173013427250715, + "grad_norm": 1.798539153159574, + "learning_rate": 9.929994451618975e-06, + "loss": 0.3545, + "step": 3713 + }, + { + "epoch": 0.08175214615892581, + "grad_norm": 1.3093559335146046, + "learning_rate": 9.929934999320705e-06, + "loss": 0.3966, + "step": 3714 + }, + { + "epoch": 0.08177415804534449, + "grad_norm": 1.5598933658944871, + "learning_rate": 9.929875521966306e-06, + "loss": 0.4867, + "step": 3715 + }, + { + "epoch": 0.08179616993176315, + "grad_norm": 1.4806202151750527, + "learning_rate": 9.929816019556083e-06, + "loss": 0.3124, + "step": 3716 + }, + { + "epoch": 0.08181818181818182, + "grad_norm": 1.6819051357791082, + "learning_rate": 9.929756492090332e-06, + "loss": 0.3681, + "step": 3717 + }, + { + "epoch": 0.08184019370460048, + "grad_norm": 1.573239394855753, + "learning_rate": 9.929696939569364e-06, + "loss": 0.391, + "step": 3718 + }, + { + "epoch": 0.08186220559101914, + "grad_norm": 1.1913367413739397, + "learning_rate": 9.929637361993474e-06, + "loss": 0.3348, + "step": 3719 + }, + { + "epoch": 0.08188421747743782, + "grad_norm": 1.932136738301298, + "learning_rate": 9.92957775936297e-06, + "loss": 0.4149, + "step": 3720 + }, + { + "epoch": 0.08190622936385648, + "grad_norm": 1.3196805865797818, + "learning_rate": 9.929518131678153e-06, + "loss": 0.4739, + "step": 3721 + }, + { + "epoch": 0.08192824125027515, + "grad_norm": 1.9523720738876778, + "learning_rate": 9.929458478939327e-06, + "loss": 0.4169, + "step": 3722 + }, + { + "epoch": 0.08195025313669381, + "grad_norm": 1.6536742238569715, + "learning_rate": 9.929398801146794e-06, + "loss": 0.3806, + "step": 3723 + }, + { + "epoch": 0.08197226502311249, + "grad_norm": 1.5584588325638171, + "learning_rate": 9.929339098300857e-06, + "loss": 0.3401, + "step": 3724 + }, + { + "epoch": 0.08199427690953115, + "grad_norm": 1.4391535162642486, + "learning_rate": 9.929279370401822e-06, + "loss": 0.3786, + "step": 3725 + }, + { + "epoch": 0.08201628879594981, + "grad_norm": 4.0140496504678485, + "learning_rate": 9.929219617449988e-06, + "loss": 0.4323, + "step": 3726 + }, + { + "epoch": 0.08203830068236848, + "grad_norm": 1.7752677970057837, + "learning_rate": 9.929159839445662e-06, + "loss": 0.4714, + "step": 3727 + }, + { + "epoch": 0.08206031256878714, + "grad_norm": 0.9610462073933739, + "learning_rate": 9.929100036389148e-06, + "loss": 0.5865, + "step": 3728 + }, + { + "epoch": 0.08208232445520582, + "grad_norm": 1.5512543674713337, + "learning_rate": 9.929040208280749e-06, + "loss": 0.3816, + "step": 3729 + }, + { + "epoch": 0.08210433634162448, + "grad_norm": 1.3668921335755526, + "learning_rate": 9.92898035512077e-06, + "loss": 0.3284, + "step": 3730 + }, + { + "epoch": 0.08212634822804314, + "grad_norm": 1.6557651654111503, + "learning_rate": 9.928920476909514e-06, + "loss": 0.3876, + "step": 3731 + }, + { + "epoch": 0.0821483601144618, + "grad_norm": 2.2485551086881057, + "learning_rate": 9.928860573647285e-06, + "loss": 0.3479, + "step": 3732 + }, + { + "epoch": 0.08217037200088048, + "grad_norm": 2.434765451522161, + "learning_rate": 9.92880064533439e-06, + "loss": 0.3952, + "step": 3733 + }, + { + "epoch": 0.08219238388729914, + "grad_norm": 1.459109856296722, + "learning_rate": 9.928740691971128e-06, + "loss": 0.3371, + "step": 3734 + }, + { + "epoch": 0.08221439577371781, + "grad_norm": 1.3333616757697526, + "learning_rate": 9.928680713557811e-06, + "loss": 0.3545, + "step": 3735 + }, + { + "epoch": 0.08223640766013647, + "grad_norm": 0.7999330329714889, + "learning_rate": 9.92862071009474e-06, + "loss": 0.5748, + "step": 3736 + }, + { + "epoch": 0.08225841954655513, + "grad_norm": 1.3671278338070838, + "learning_rate": 9.92856068158222e-06, + "loss": 0.3306, + "step": 3737 + }, + { + "epoch": 0.08228043143297381, + "grad_norm": 1.5787155749062558, + "learning_rate": 9.928500628020556e-06, + "loss": 0.4117, + "step": 3738 + }, + { + "epoch": 0.08230244331939247, + "grad_norm": 1.4354751035753957, + "learning_rate": 9.928440549410053e-06, + "loss": 0.4031, + "step": 3739 + }, + { + "epoch": 0.08232445520581114, + "grad_norm": 2.0313157847298773, + "learning_rate": 9.928380445751017e-06, + "loss": 0.3806, + "step": 3740 + }, + { + "epoch": 0.0823464670922298, + "grad_norm": 1.6435811448066755, + "learning_rate": 9.928320317043755e-06, + "loss": 0.3634, + "step": 3741 + }, + { + "epoch": 0.08236847897864848, + "grad_norm": 1.9751695231905948, + "learning_rate": 9.92826016328857e-06, + "loss": 0.4614, + "step": 3742 + }, + { + "epoch": 0.08239049086506714, + "grad_norm": 2.034363675465334, + "learning_rate": 9.928199984485771e-06, + "loss": 0.409, + "step": 3743 + }, + { + "epoch": 0.0824125027514858, + "grad_norm": 1.326023280419985, + "learning_rate": 9.92813978063566e-06, + "loss": 0.3898, + "step": 3744 + }, + { + "epoch": 0.08243451463790447, + "grad_norm": 1.4405953307458748, + "learning_rate": 9.928079551738542e-06, + "loss": 0.355, + "step": 3745 + }, + { + "epoch": 0.08245652652432313, + "grad_norm": 0.6598973069531814, + "learning_rate": 9.928019297794728e-06, + "loss": 0.5754, + "step": 3746 + }, + { + "epoch": 0.0824785384107418, + "grad_norm": 0.6542423148206908, + "learning_rate": 9.927959018804523e-06, + "loss": 0.5747, + "step": 3747 + }, + { + "epoch": 0.08250055029716047, + "grad_norm": 1.660971090224079, + "learning_rate": 9.927898714768232e-06, + "loss": 0.4897, + "step": 3748 + }, + { + "epoch": 0.08252256218357913, + "grad_norm": 1.5838152473686786, + "learning_rate": 9.927838385686162e-06, + "loss": 0.3773, + "step": 3749 + }, + { + "epoch": 0.0825445740699978, + "grad_norm": 2.0215568963185757, + "learning_rate": 9.927778031558616e-06, + "loss": 0.473, + "step": 3750 + }, + { + "epoch": 0.08256658595641647, + "grad_norm": 1.379390499253276, + "learning_rate": 9.927717652385909e-06, + "loss": 0.3715, + "step": 3751 + }, + { + "epoch": 0.08258859784283513, + "grad_norm": 0.6780107233630973, + "learning_rate": 9.927657248168342e-06, + "loss": 0.5611, + "step": 3752 + }, + { + "epoch": 0.0826106097292538, + "grad_norm": 1.2059025438981994, + "learning_rate": 9.92759681890622e-06, + "loss": 0.3016, + "step": 3753 + }, + { + "epoch": 0.08263262161567246, + "grad_norm": 1.3649284987990569, + "learning_rate": 9.927536364599856e-06, + "loss": 0.3078, + "step": 3754 + }, + { + "epoch": 0.08265463350209112, + "grad_norm": 1.9874233471654235, + "learning_rate": 9.927475885249554e-06, + "loss": 0.3434, + "step": 3755 + }, + { + "epoch": 0.0826766453885098, + "grad_norm": 2.1339222331787693, + "learning_rate": 9.92741538085562e-06, + "loss": 0.3833, + "step": 3756 + }, + { + "epoch": 0.08269865727492846, + "grad_norm": 0.6859029302879005, + "learning_rate": 9.927354851418365e-06, + "loss": 0.5711, + "step": 3757 + }, + { + "epoch": 0.08272066916134713, + "grad_norm": 2.0324629767286657, + "learning_rate": 9.927294296938095e-06, + "loss": 0.4371, + "step": 3758 + }, + { + "epoch": 0.08274268104776579, + "grad_norm": 3.6740008244869324, + "learning_rate": 9.92723371741512e-06, + "loss": 0.4114, + "step": 3759 + }, + { + "epoch": 0.08276469293418445, + "grad_norm": 1.7022556803164768, + "learning_rate": 9.927173112849742e-06, + "loss": 0.4009, + "step": 3760 + }, + { + "epoch": 0.08278670482060313, + "grad_norm": 0.6601013249778628, + "learning_rate": 9.927112483242274e-06, + "loss": 0.5801, + "step": 3761 + }, + { + "epoch": 0.08280871670702179, + "grad_norm": 1.967355380844371, + "learning_rate": 9.927051828593023e-06, + "loss": 0.4398, + "step": 3762 + }, + { + "epoch": 0.08283072859344046, + "grad_norm": 1.853165249067634, + "learning_rate": 9.926991148902296e-06, + "loss": 0.3967, + "step": 3763 + }, + { + "epoch": 0.08285274047985912, + "grad_norm": 1.4858381339077047, + "learning_rate": 9.926930444170403e-06, + "loss": 0.3522, + "step": 3764 + }, + { + "epoch": 0.0828747523662778, + "grad_norm": 0.6481946028369597, + "learning_rate": 9.926869714397652e-06, + "loss": 0.5519, + "step": 3765 + }, + { + "epoch": 0.08289676425269646, + "grad_norm": 3.6025115647361643, + "learning_rate": 9.926808959584352e-06, + "loss": 0.336, + "step": 3766 + }, + { + "epoch": 0.08291877613911512, + "grad_norm": 1.4858400023949085, + "learning_rate": 9.926748179730812e-06, + "loss": 0.3757, + "step": 3767 + }, + { + "epoch": 0.08294078802553378, + "grad_norm": 1.7494082877497341, + "learning_rate": 9.92668737483734e-06, + "loss": 0.3549, + "step": 3768 + }, + { + "epoch": 0.08296279991195245, + "grad_norm": 1.534098617680511, + "learning_rate": 9.926626544904245e-06, + "loss": 0.3325, + "step": 3769 + }, + { + "epoch": 0.08298481179837112, + "grad_norm": 0.6288384175006467, + "learning_rate": 9.926565689931836e-06, + "loss": 0.5689, + "step": 3770 + }, + { + "epoch": 0.08300682368478979, + "grad_norm": 1.217553184773016, + "learning_rate": 9.926504809920424e-06, + "loss": 0.2981, + "step": 3771 + }, + { + "epoch": 0.08302883557120845, + "grad_norm": 3.38471571420602, + "learning_rate": 9.926443904870317e-06, + "loss": 0.3896, + "step": 3772 + }, + { + "epoch": 0.08305084745762711, + "grad_norm": 3.546386241575343, + "learning_rate": 9.926382974781824e-06, + "loss": 0.4345, + "step": 3773 + }, + { + "epoch": 0.08307285934404579, + "grad_norm": 1.92554410850294, + "learning_rate": 9.926322019655258e-06, + "loss": 0.3469, + "step": 3774 + }, + { + "epoch": 0.08309487123046445, + "grad_norm": 1.5439905352310672, + "learning_rate": 9.926261039490922e-06, + "loss": 0.4388, + "step": 3775 + }, + { + "epoch": 0.08311688311688312, + "grad_norm": 1.4675021935525525, + "learning_rate": 9.926200034289134e-06, + "loss": 0.3652, + "step": 3776 + }, + { + "epoch": 0.08313889500330178, + "grad_norm": 2.104156103254286, + "learning_rate": 9.9261390040502e-06, + "loss": 0.321, + "step": 3777 + }, + { + "epoch": 0.08316090688972044, + "grad_norm": 1.4865049720995305, + "learning_rate": 9.926077948774428e-06, + "loss": 0.4308, + "step": 3778 + }, + { + "epoch": 0.08318291877613912, + "grad_norm": 1.966650823285053, + "learning_rate": 9.926016868462132e-06, + "loss": 0.4025, + "step": 3779 + }, + { + "epoch": 0.08320493066255778, + "grad_norm": 1.5077271264362875, + "learning_rate": 9.925955763113622e-06, + "loss": 0.3552, + "step": 3780 + }, + { + "epoch": 0.08322694254897645, + "grad_norm": 2.1555419945358065, + "learning_rate": 9.925894632729207e-06, + "loss": 0.3206, + "step": 3781 + }, + { + "epoch": 0.08324895443539511, + "grad_norm": 1.4397659377069512, + "learning_rate": 9.925833477309198e-06, + "loss": 0.3573, + "step": 3782 + }, + { + "epoch": 0.08327096632181379, + "grad_norm": 2.0317496248488904, + "learning_rate": 9.925772296853907e-06, + "loss": 0.3576, + "step": 3783 + }, + { + "epoch": 0.08329297820823245, + "grad_norm": 1.5610606173327057, + "learning_rate": 9.925711091363644e-06, + "loss": 0.5093, + "step": 3784 + }, + { + "epoch": 0.08331499009465111, + "grad_norm": 1.631792883246824, + "learning_rate": 9.925649860838721e-06, + "loss": 0.3588, + "step": 3785 + }, + { + "epoch": 0.08333700198106977, + "grad_norm": 1.516000038742008, + "learning_rate": 9.925588605279448e-06, + "loss": 0.362, + "step": 3786 + }, + { + "epoch": 0.08335901386748844, + "grad_norm": 1.4377184416375632, + "learning_rate": 9.925527324686137e-06, + "loss": 0.4495, + "step": 3787 + }, + { + "epoch": 0.08338102575390711, + "grad_norm": 1.793421997732967, + "learning_rate": 9.925466019059097e-06, + "loss": 0.4244, + "step": 3788 + }, + { + "epoch": 0.08340303764032578, + "grad_norm": 1.397991046424098, + "learning_rate": 9.925404688398646e-06, + "loss": 0.3202, + "step": 3789 + }, + { + "epoch": 0.08342504952674444, + "grad_norm": 0.7166770321269826, + "learning_rate": 9.925343332705088e-06, + "loss": 0.5639, + "step": 3790 + }, + { + "epoch": 0.0834470614131631, + "grad_norm": 1.9897551163450289, + "learning_rate": 9.92528195197874e-06, + "loss": 0.3814, + "step": 3791 + }, + { + "epoch": 0.08346907329958178, + "grad_norm": 1.6191982780234395, + "learning_rate": 9.925220546219914e-06, + "loss": 0.4536, + "step": 3792 + }, + { + "epoch": 0.08349108518600044, + "grad_norm": 1.5833731878383004, + "learning_rate": 9.925159115428917e-06, + "loss": 0.3683, + "step": 3793 + }, + { + "epoch": 0.0835130970724191, + "grad_norm": 2.578923999291331, + "learning_rate": 9.925097659606068e-06, + "loss": 0.4275, + "step": 3794 + }, + { + "epoch": 0.08353510895883777, + "grad_norm": 1.3583073847753904, + "learning_rate": 9.925036178751674e-06, + "loss": 0.444, + "step": 3795 + }, + { + "epoch": 0.08355712084525643, + "grad_norm": 1.3790360658089358, + "learning_rate": 9.92497467286605e-06, + "loss": 0.3526, + "step": 3796 + }, + { + "epoch": 0.08357913273167511, + "grad_norm": 1.791703744763121, + "learning_rate": 9.924913141949508e-06, + "loss": 0.4518, + "step": 3797 + }, + { + "epoch": 0.08360114461809377, + "grad_norm": 2.808242196577142, + "learning_rate": 9.924851586002362e-06, + "loss": 0.3914, + "step": 3798 + }, + { + "epoch": 0.08362315650451244, + "grad_norm": 1.9338054351731497, + "learning_rate": 9.924790005024923e-06, + "loss": 0.3719, + "step": 3799 + }, + { + "epoch": 0.0836451683909311, + "grad_norm": 1.3986112765101075, + "learning_rate": 9.924728399017504e-06, + "loss": 0.2841, + "step": 3800 + }, + { + "epoch": 0.08366718027734978, + "grad_norm": 1.8519949014606407, + "learning_rate": 9.924666767980418e-06, + "loss": 0.4351, + "step": 3801 + }, + { + "epoch": 0.08368919216376844, + "grad_norm": 1.44543709162605, + "learning_rate": 9.924605111913981e-06, + "loss": 0.3762, + "step": 3802 + }, + { + "epoch": 0.0837112040501871, + "grad_norm": 0.7086309600469987, + "learning_rate": 9.924543430818503e-06, + "loss": 0.5471, + "step": 3803 + }, + { + "epoch": 0.08373321593660576, + "grad_norm": 1.6570759035985385, + "learning_rate": 9.924481724694299e-06, + "loss": 0.328, + "step": 3804 + }, + { + "epoch": 0.08375522782302443, + "grad_norm": 1.644016469699709, + "learning_rate": 9.924419993541683e-06, + "loss": 0.3675, + "step": 3805 + }, + { + "epoch": 0.0837772397094431, + "grad_norm": 0.6623925832779132, + "learning_rate": 9.92435823736097e-06, + "loss": 0.5909, + "step": 3806 + }, + { + "epoch": 0.08379925159586177, + "grad_norm": 1.3906582607440539, + "learning_rate": 9.924296456152468e-06, + "loss": 0.3283, + "step": 3807 + }, + { + "epoch": 0.08382126348228043, + "grad_norm": 1.816213215191303, + "learning_rate": 9.924234649916499e-06, + "loss": 0.5199, + "step": 3808 + }, + { + "epoch": 0.0838432753686991, + "grad_norm": 1.8023178481312385, + "learning_rate": 9.924172818653372e-06, + "loss": 0.3383, + "step": 3809 + }, + { + "epoch": 0.08386528725511776, + "grad_norm": 1.9905617643033349, + "learning_rate": 9.924110962363401e-06, + "loss": 0.3787, + "step": 3810 + }, + { + "epoch": 0.08388729914153643, + "grad_norm": 1.6750687038085725, + "learning_rate": 9.924049081046902e-06, + "loss": 0.3137, + "step": 3811 + }, + { + "epoch": 0.0839093110279551, + "grad_norm": 1.3700147853647382, + "learning_rate": 9.923987174704192e-06, + "loss": 0.3707, + "step": 3812 + }, + { + "epoch": 0.08393132291437376, + "grad_norm": 0.8267393989332875, + "learning_rate": 9.923925243335581e-06, + "loss": 0.5868, + "step": 3813 + }, + { + "epoch": 0.08395333480079242, + "grad_norm": 1.4914005487919955, + "learning_rate": 9.923863286941386e-06, + "loss": 0.4171, + "step": 3814 + }, + { + "epoch": 0.0839753466872111, + "grad_norm": 1.4791381579330383, + "learning_rate": 9.923801305521924e-06, + "loss": 0.3728, + "step": 3815 + }, + { + "epoch": 0.08399735857362976, + "grad_norm": 2.3703854242414484, + "learning_rate": 9.923739299077504e-06, + "loss": 0.4557, + "step": 3816 + }, + { + "epoch": 0.08401937046004843, + "grad_norm": 1.4702225078147577, + "learning_rate": 9.923677267608447e-06, + "loss": 0.3572, + "step": 3817 + }, + { + "epoch": 0.08404138234646709, + "grad_norm": 1.992547655847881, + "learning_rate": 9.923615211115065e-06, + "loss": 0.3466, + "step": 3818 + }, + { + "epoch": 0.08406339423288575, + "grad_norm": 1.7249948445985, + "learning_rate": 9.923553129597673e-06, + "loss": 0.3893, + "step": 3819 + }, + { + "epoch": 0.08408540611930443, + "grad_norm": 2.209340530943458, + "learning_rate": 9.923491023056592e-06, + "loss": 0.3528, + "step": 3820 + }, + { + "epoch": 0.08410741800572309, + "grad_norm": 1.7755676287648592, + "learning_rate": 9.92342889149213e-06, + "loss": 0.4249, + "step": 3821 + }, + { + "epoch": 0.08412942989214175, + "grad_norm": 1.6585431015915424, + "learning_rate": 9.923366734904609e-06, + "loss": 0.3683, + "step": 3822 + }, + { + "epoch": 0.08415144177856042, + "grad_norm": 1.5028162744227216, + "learning_rate": 9.923304553294342e-06, + "loss": 0.403, + "step": 3823 + }, + { + "epoch": 0.0841734536649791, + "grad_norm": 1.4384632882564858, + "learning_rate": 9.923242346661646e-06, + "loss": 0.4442, + "step": 3824 + }, + { + "epoch": 0.08419546555139776, + "grad_norm": 1.2314085061757307, + "learning_rate": 9.923180115006834e-06, + "loss": 0.3173, + "step": 3825 + }, + { + "epoch": 0.08421747743781642, + "grad_norm": 2.0980352791513153, + "learning_rate": 9.923117858330227e-06, + "loss": 0.4327, + "step": 3826 + }, + { + "epoch": 0.08423948932423508, + "grad_norm": 0.7420797373822186, + "learning_rate": 9.92305557663214e-06, + "loss": 0.5618, + "step": 3827 + }, + { + "epoch": 0.08426150121065375, + "grad_norm": 1.8951857379185568, + "learning_rate": 9.922993269912887e-06, + "loss": 0.341, + "step": 3828 + }, + { + "epoch": 0.08428351309707242, + "grad_norm": 1.7647545249078553, + "learning_rate": 9.922930938172786e-06, + "loss": 0.3827, + "step": 3829 + }, + { + "epoch": 0.08430552498349109, + "grad_norm": 3.10087262320339, + "learning_rate": 9.922868581412155e-06, + "loss": 0.3035, + "step": 3830 + }, + { + "epoch": 0.08432753686990975, + "grad_norm": 1.4163470370658924, + "learning_rate": 9.92280619963131e-06, + "loss": 0.2985, + "step": 3831 + }, + { + "epoch": 0.08434954875632841, + "grad_norm": 1.5001630560196535, + "learning_rate": 9.922743792830569e-06, + "loss": 0.3657, + "step": 3832 + }, + { + "epoch": 0.08437156064274709, + "grad_norm": 2.085329952728956, + "learning_rate": 9.922681361010249e-06, + "loss": 0.4119, + "step": 3833 + }, + { + "epoch": 0.08439357252916575, + "grad_norm": 1.6970190317218925, + "learning_rate": 9.922618904170666e-06, + "loss": 0.3289, + "step": 3834 + }, + { + "epoch": 0.08441558441558442, + "grad_norm": 1.5930564275719665, + "learning_rate": 9.922556422312137e-06, + "loss": 0.4693, + "step": 3835 + }, + { + "epoch": 0.08443759630200308, + "grad_norm": 1.5034400231815395, + "learning_rate": 9.922493915434982e-06, + "loss": 0.2546, + "step": 3836 + }, + { + "epoch": 0.08445960818842174, + "grad_norm": 1.5115388411983832, + "learning_rate": 9.922431383539517e-06, + "loss": 0.3842, + "step": 3837 + }, + { + "epoch": 0.08448162007484042, + "grad_norm": 1.483834041068255, + "learning_rate": 9.922368826626061e-06, + "loss": 0.413, + "step": 3838 + }, + { + "epoch": 0.08450363196125908, + "grad_norm": 1.2866562641915347, + "learning_rate": 9.92230624469493e-06, + "loss": 0.3093, + "step": 3839 + }, + { + "epoch": 0.08452564384767774, + "grad_norm": 1.7459588692540788, + "learning_rate": 9.922243637746445e-06, + "loss": 0.4523, + "step": 3840 + }, + { + "epoch": 0.08454765573409641, + "grad_norm": 1.5630756120157365, + "learning_rate": 9.922181005780921e-06, + "loss": 0.4891, + "step": 3841 + }, + { + "epoch": 0.08456966762051508, + "grad_norm": 0.7542093175995201, + "learning_rate": 9.922118348798678e-06, + "loss": 0.5524, + "step": 3842 + }, + { + "epoch": 0.08459167950693375, + "grad_norm": 2.2281398027382, + "learning_rate": 9.922055666800035e-06, + "loss": 0.3712, + "step": 3843 + }, + { + "epoch": 0.08461369139335241, + "grad_norm": 1.8088152302469211, + "learning_rate": 9.92199295978531e-06, + "loss": 0.3978, + "step": 3844 + }, + { + "epoch": 0.08463570327977107, + "grad_norm": 1.5519755593344862, + "learning_rate": 9.92193022775482e-06, + "loss": 0.4052, + "step": 3845 + }, + { + "epoch": 0.08465771516618974, + "grad_norm": 1.6463270917104855, + "learning_rate": 9.921867470708886e-06, + "loss": 0.4784, + "step": 3846 + }, + { + "epoch": 0.08467972705260841, + "grad_norm": 0.6585545854845469, + "learning_rate": 9.921804688647825e-06, + "loss": 0.5707, + "step": 3847 + }, + { + "epoch": 0.08470173893902708, + "grad_norm": 1.7207797894217547, + "learning_rate": 9.92174188157196e-06, + "loss": 0.3427, + "step": 3848 + }, + { + "epoch": 0.08472375082544574, + "grad_norm": 1.4191225679237067, + "learning_rate": 9.921679049481606e-06, + "loss": 0.3648, + "step": 3849 + }, + { + "epoch": 0.0847457627118644, + "grad_norm": 1.8754847340154726, + "learning_rate": 9.921616192377083e-06, + "loss": 0.362, + "step": 3850 + }, + { + "epoch": 0.08476777459828307, + "grad_norm": 1.4840141895840244, + "learning_rate": 9.921553310258714e-06, + "loss": 0.4191, + "step": 3851 + }, + { + "epoch": 0.08478978648470174, + "grad_norm": 1.55681258673431, + "learning_rate": 9.921490403126814e-06, + "loss": 0.388, + "step": 3852 + }, + { + "epoch": 0.0848117983711204, + "grad_norm": 1.3350793395568588, + "learning_rate": 9.921427470981706e-06, + "loss": 0.3186, + "step": 3853 + }, + { + "epoch": 0.08483381025753907, + "grad_norm": 1.8875623924147282, + "learning_rate": 9.921364513823707e-06, + "loss": 0.4974, + "step": 3854 + }, + { + "epoch": 0.08485582214395773, + "grad_norm": 1.2347808915474692, + "learning_rate": 9.92130153165314e-06, + "loss": 0.3313, + "step": 3855 + }, + { + "epoch": 0.08487783403037641, + "grad_norm": 1.659983219854059, + "learning_rate": 9.921238524470322e-06, + "loss": 0.3756, + "step": 3856 + }, + { + "epoch": 0.08489984591679507, + "grad_norm": 0.7359904312675289, + "learning_rate": 9.921175492275577e-06, + "loss": 0.5866, + "step": 3857 + }, + { + "epoch": 0.08492185780321374, + "grad_norm": 1.4159367220912513, + "learning_rate": 9.921112435069222e-06, + "loss": 0.3678, + "step": 3858 + }, + { + "epoch": 0.0849438696896324, + "grad_norm": 0.6511006264214495, + "learning_rate": 9.921049352851579e-06, + "loss": 0.5654, + "step": 3859 + }, + { + "epoch": 0.08496588157605106, + "grad_norm": 1.5303940213194502, + "learning_rate": 9.920986245622969e-06, + "loss": 0.4045, + "step": 3860 + }, + { + "epoch": 0.08498789346246974, + "grad_norm": 1.5446975612686247, + "learning_rate": 9.920923113383711e-06, + "loss": 0.3075, + "step": 3861 + }, + { + "epoch": 0.0850099053488884, + "grad_norm": 2.6658556518722865, + "learning_rate": 9.920859956134128e-06, + "loss": 0.3347, + "step": 3862 + }, + { + "epoch": 0.08503191723530706, + "grad_norm": 1.4273315231981352, + "learning_rate": 9.92079677387454e-06, + "loss": 0.346, + "step": 3863 + }, + { + "epoch": 0.08505392912172573, + "grad_norm": 1.6983763982538103, + "learning_rate": 9.920733566605267e-06, + "loss": 0.4119, + "step": 3864 + }, + { + "epoch": 0.0850759410081444, + "grad_norm": 1.654682589284345, + "learning_rate": 9.920670334326631e-06, + "loss": 0.3717, + "step": 3865 + }, + { + "epoch": 0.08509795289456307, + "grad_norm": 2.1627816520691123, + "learning_rate": 9.920607077038955e-06, + "loss": 0.336, + "step": 3866 + }, + { + "epoch": 0.08511996478098173, + "grad_norm": 1.5244248199002106, + "learning_rate": 9.920543794742558e-06, + "loss": 0.3647, + "step": 3867 + }, + { + "epoch": 0.08514197666740039, + "grad_norm": 2.005400310164856, + "learning_rate": 9.920480487437762e-06, + "loss": 0.4185, + "step": 3868 + }, + { + "epoch": 0.08516398855381906, + "grad_norm": 1.4267917115590267, + "learning_rate": 9.920417155124891e-06, + "loss": 0.4767, + "step": 3869 + }, + { + "epoch": 0.08518600044023773, + "grad_norm": 1.3184265507286719, + "learning_rate": 9.920353797804267e-06, + "loss": 0.3417, + "step": 3870 + }, + { + "epoch": 0.0852080123266564, + "grad_norm": 2.1275944593091505, + "learning_rate": 9.920290415476208e-06, + "loss": 0.4648, + "step": 3871 + }, + { + "epoch": 0.08523002421307506, + "grad_norm": 1.7410077388235474, + "learning_rate": 9.92022700814104e-06, + "loss": 0.4415, + "step": 3872 + }, + { + "epoch": 0.08525203609949372, + "grad_norm": 0.9303814830931849, + "learning_rate": 9.920163575799082e-06, + "loss": 0.5931, + "step": 3873 + }, + { + "epoch": 0.0852740479859124, + "grad_norm": 1.4959995680082396, + "learning_rate": 9.920100118450661e-06, + "loss": 0.3858, + "step": 3874 + }, + { + "epoch": 0.08529605987233106, + "grad_norm": 1.642356636145428, + "learning_rate": 9.920036636096095e-06, + "loss": 0.3454, + "step": 3875 + }, + { + "epoch": 0.08531807175874973, + "grad_norm": 1.2485023300332778, + "learning_rate": 9.919973128735709e-06, + "loss": 0.3033, + "step": 3876 + }, + { + "epoch": 0.08534008364516839, + "grad_norm": 0.6949543809853689, + "learning_rate": 9.919909596369824e-06, + "loss": 0.5827, + "step": 3877 + }, + { + "epoch": 0.08536209553158705, + "grad_norm": 1.5127217279924496, + "learning_rate": 9.919846038998766e-06, + "loss": 0.3437, + "step": 3878 + }, + { + "epoch": 0.08538410741800573, + "grad_norm": 3.256586762021155, + "learning_rate": 9.919782456622855e-06, + "loss": 0.4315, + "step": 3879 + }, + { + "epoch": 0.08540611930442439, + "grad_norm": 1.6204477215644877, + "learning_rate": 9.919718849242414e-06, + "loss": 0.3542, + "step": 3880 + }, + { + "epoch": 0.08542813119084305, + "grad_norm": 1.4655054071137676, + "learning_rate": 9.91965521685777e-06, + "loss": 0.4488, + "step": 3881 + }, + { + "epoch": 0.08545014307726172, + "grad_norm": 1.8272211627084882, + "learning_rate": 9.919591559469242e-06, + "loss": 0.4127, + "step": 3882 + }, + { + "epoch": 0.0854721549636804, + "grad_norm": 1.9662485358017832, + "learning_rate": 9.919527877077158e-06, + "loss": 0.4323, + "step": 3883 + }, + { + "epoch": 0.08549416685009906, + "grad_norm": 1.5590901071072334, + "learning_rate": 9.919464169681837e-06, + "loss": 0.3261, + "step": 3884 + }, + { + "epoch": 0.08551617873651772, + "grad_norm": 2.2722161683109636, + "learning_rate": 9.919400437283605e-06, + "loss": 0.3928, + "step": 3885 + }, + { + "epoch": 0.08553819062293638, + "grad_norm": 1.4105291439822079, + "learning_rate": 9.919336679882788e-06, + "loss": 0.3694, + "step": 3886 + }, + { + "epoch": 0.08556020250935505, + "grad_norm": 4.9481425377738795, + "learning_rate": 9.919272897479706e-06, + "loss": 0.4329, + "step": 3887 + }, + { + "epoch": 0.08558221439577372, + "grad_norm": 1.5732115607319233, + "learning_rate": 9.919209090074684e-06, + "loss": 0.4041, + "step": 3888 + }, + { + "epoch": 0.08560422628219239, + "grad_norm": 3.1160766442614847, + "learning_rate": 9.919145257668049e-06, + "loss": 0.3908, + "step": 3889 + }, + { + "epoch": 0.08562623816861105, + "grad_norm": 1.674401943375083, + "learning_rate": 9.919081400260125e-06, + "loss": 0.3604, + "step": 3890 + }, + { + "epoch": 0.08564825005502971, + "grad_norm": 1.3261161855211137, + "learning_rate": 9.919017517851232e-06, + "loss": 0.3141, + "step": 3891 + }, + { + "epoch": 0.08567026194144839, + "grad_norm": 0.897360773715379, + "learning_rate": 9.918953610441701e-06, + "loss": 0.6152, + "step": 3892 + }, + { + "epoch": 0.08569227382786705, + "grad_norm": 1.6431516698198356, + "learning_rate": 9.918889678031852e-06, + "loss": 0.3805, + "step": 3893 + }, + { + "epoch": 0.08571428571428572, + "grad_norm": 0.7286726413937074, + "learning_rate": 9.918825720622013e-06, + "loss": 0.5904, + "step": 3894 + }, + { + "epoch": 0.08573629760070438, + "grad_norm": 1.4457470614912349, + "learning_rate": 9.918761738212508e-06, + "loss": 0.3714, + "step": 3895 + }, + { + "epoch": 0.08575830948712304, + "grad_norm": 1.2765857335323978, + "learning_rate": 9.918697730803661e-06, + "loss": 0.3864, + "step": 3896 + }, + { + "epoch": 0.08578032137354172, + "grad_norm": 1.6151250358052764, + "learning_rate": 9.9186336983958e-06, + "loss": 0.4535, + "step": 3897 + }, + { + "epoch": 0.08580233325996038, + "grad_norm": 2.2439165581602136, + "learning_rate": 9.918569640989246e-06, + "loss": 0.4011, + "step": 3898 + }, + { + "epoch": 0.08582434514637904, + "grad_norm": 1.8221727853163818, + "learning_rate": 9.918505558584327e-06, + "loss": 0.4275, + "step": 3899 + }, + { + "epoch": 0.08584635703279771, + "grad_norm": 2.112004919973427, + "learning_rate": 9.918441451181373e-06, + "loss": 0.3709, + "step": 3900 + }, + { + "epoch": 0.08586836891921637, + "grad_norm": 1.4768064004691448, + "learning_rate": 9.918377318780703e-06, + "loss": 0.3303, + "step": 3901 + }, + { + "epoch": 0.08589038080563505, + "grad_norm": 1.7275900552742192, + "learning_rate": 9.918313161382646e-06, + "loss": 0.3208, + "step": 3902 + }, + { + "epoch": 0.08591239269205371, + "grad_norm": 1.3435279691888433, + "learning_rate": 9.918248978987528e-06, + "loss": 0.2976, + "step": 3903 + }, + { + "epoch": 0.08593440457847237, + "grad_norm": 1.7789929989082058, + "learning_rate": 9.918184771595675e-06, + "loss": 0.3971, + "step": 3904 + }, + { + "epoch": 0.08595641646489104, + "grad_norm": 1.4780628363551584, + "learning_rate": 9.918120539207416e-06, + "loss": 0.3834, + "step": 3905 + }, + { + "epoch": 0.08597842835130971, + "grad_norm": 1.53465534007511, + "learning_rate": 9.918056281823071e-06, + "loss": 0.4094, + "step": 3906 + }, + { + "epoch": 0.08600044023772838, + "grad_norm": 1.3649606606353923, + "learning_rate": 9.917991999442973e-06, + "loss": 0.371, + "step": 3907 + }, + { + "epoch": 0.08602245212414704, + "grad_norm": 2.8086027380644527, + "learning_rate": 9.917927692067445e-06, + "loss": 0.4654, + "step": 3908 + }, + { + "epoch": 0.0860444640105657, + "grad_norm": 1.6905034434231976, + "learning_rate": 9.917863359696816e-06, + "loss": 0.4419, + "step": 3909 + }, + { + "epoch": 0.08606647589698437, + "grad_norm": 3.965142171513075, + "learning_rate": 9.917799002331412e-06, + "loss": 0.3633, + "step": 3910 + }, + { + "epoch": 0.08608848778340304, + "grad_norm": 1.5458223800468445, + "learning_rate": 9.917734619971559e-06, + "loss": 0.3614, + "step": 3911 + }, + { + "epoch": 0.0861104996698217, + "grad_norm": 1.7609017308190136, + "learning_rate": 9.917670212617587e-06, + "loss": 0.3911, + "step": 3912 + }, + { + "epoch": 0.08613251155624037, + "grad_norm": 1.402953972899419, + "learning_rate": 9.917605780269819e-06, + "loss": 0.6143, + "step": 3913 + }, + { + "epoch": 0.08615452344265903, + "grad_norm": 1.7788510611533188, + "learning_rate": 9.917541322928587e-06, + "loss": 0.3961, + "step": 3914 + }, + { + "epoch": 0.08617653532907771, + "grad_norm": 2.671756691667426, + "learning_rate": 9.917476840594215e-06, + "loss": 0.4043, + "step": 3915 + }, + { + "epoch": 0.08619854721549637, + "grad_norm": 1.2708799962132997, + "learning_rate": 9.917412333267036e-06, + "loss": 0.2692, + "step": 3916 + }, + { + "epoch": 0.08622055910191503, + "grad_norm": 13.182112011587263, + "learning_rate": 9.917347800947371e-06, + "loss": 0.3217, + "step": 3917 + }, + { + "epoch": 0.0862425709883337, + "grad_norm": 0.752073629523651, + "learning_rate": 9.917283243635552e-06, + "loss": 0.584, + "step": 3918 + }, + { + "epoch": 0.08626458287475236, + "grad_norm": 1.6768176074965024, + "learning_rate": 9.917218661331907e-06, + "loss": 0.4398, + "step": 3919 + }, + { + "epoch": 0.08628659476117104, + "grad_norm": 1.2271603484573017, + "learning_rate": 9.917154054036765e-06, + "loss": 0.2923, + "step": 3920 + }, + { + "epoch": 0.0863086066475897, + "grad_norm": 2.0055026541692467, + "learning_rate": 9.917089421750449e-06, + "loss": 0.3778, + "step": 3921 + }, + { + "epoch": 0.08633061853400836, + "grad_norm": 0.8674210254806435, + "learning_rate": 9.917024764473294e-06, + "loss": 0.5974, + "step": 3922 + }, + { + "epoch": 0.08635263042042703, + "grad_norm": 2.28062722021652, + "learning_rate": 9.916960082205627e-06, + "loss": 0.4252, + "step": 3923 + }, + { + "epoch": 0.0863746423068457, + "grad_norm": 1.3680817806400618, + "learning_rate": 9.916895374947773e-06, + "loss": 0.4328, + "step": 3924 + }, + { + "epoch": 0.08639665419326437, + "grad_norm": 0.7430307796006327, + "learning_rate": 9.916830642700066e-06, + "loss": 0.5853, + "step": 3925 + }, + { + "epoch": 0.08641866607968303, + "grad_norm": 1.6754670602735118, + "learning_rate": 9.916765885462831e-06, + "loss": 0.3801, + "step": 3926 + }, + { + "epoch": 0.08644067796610169, + "grad_norm": 1.6444456977005115, + "learning_rate": 9.916701103236402e-06, + "loss": 0.4227, + "step": 3927 + }, + { + "epoch": 0.08646268985252036, + "grad_norm": 1.4410211889044757, + "learning_rate": 9.916636296021103e-06, + "loss": 0.3056, + "step": 3928 + }, + { + "epoch": 0.08648470173893903, + "grad_norm": 1.8224148610513857, + "learning_rate": 9.916571463817264e-06, + "loss": 0.3393, + "step": 3929 + }, + { + "epoch": 0.0865067136253577, + "grad_norm": 1.7758248796244396, + "learning_rate": 9.916506606625217e-06, + "loss": 0.3786, + "step": 3930 + }, + { + "epoch": 0.08652872551177636, + "grad_norm": 1.5343415173441397, + "learning_rate": 9.916441724445292e-06, + "loss": 0.3121, + "step": 3931 + }, + { + "epoch": 0.08655073739819502, + "grad_norm": 1.8803089106680886, + "learning_rate": 9.916376817277816e-06, + "loss": 0.3134, + "step": 3932 + }, + { + "epoch": 0.0865727492846137, + "grad_norm": 1.3527405627860327, + "learning_rate": 9.91631188512312e-06, + "loss": 0.3412, + "step": 3933 + }, + { + "epoch": 0.08659476117103236, + "grad_norm": 2.324139136773947, + "learning_rate": 9.916246927981536e-06, + "loss": 0.4117, + "step": 3934 + }, + { + "epoch": 0.08661677305745102, + "grad_norm": 1.5731350762237943, + "learning_rate": 9.916181945853391e-06, + "loss": 0.3255, + "step": 3935 + }, + { + "epoch": 0.08663878494386969, + "grad_norm": 1.5608173212786838, + "learning_rate": 9.916116938739018e-06, + "loss": 0.428, + "step": 3936 + }, + { + "epoch": 0.08666079683028835, + "grad_norm": 1.2204508318657443, + "learning_rate": 9.916051906638745e-06, + "loss": 0.2816, + "step": 3937 + }, + { + "epoch": 0.08668280871670703, + "grad_norm": 1.6188645934235455, + "learning_rate": 9.915986849552904e-06, + "loss": 0.3993, + "step": 3938 + }, + { + "epoch": 0.08670482060312569, + "grad_norm": 1.5595705039672738, + "learning_rate": 9.915921767481824e-06, + "loss": 0.4634, + "step": 3939 + }, + { + "epoch": 0.08672683248954435, + "grad_norm": 1.6430873707944127, + "learning_rate": 9.915856660425838e-06, + "loss": 0.3534, + "step": 3940 + }, + { + "epoch": 0.08674884437596302, + "grad_norm": 1.7089977139152701, + "learning_rate": 9.915791528385274e-06, + "loss": 0.3232, + "step": 3941 + }, + { + "epoch": 0.08677085626238168, + "grad_norm": 2.5284571274287497, + "learning_rate": 9.915726371360467e-06, + "loss": 0.4081, + "step": 3942 + }, + { + "epoch": 0.08679286814880036, + "grad_norm": 1.4495211693045436, + "learning_rate": 9.915661189351746e-06, + "loss": 0.4558, + "step": 3943 + }, + { + "epoch": 0.08681488003521902, + "grad_norm": 1.7251662101202205, + "learning_rate": 9.915595982359442e-06, + "loss": 0.4374, + "step": 3944 + }, + { + "epoch": 0.08683689192163768, + "grad_norm": 1.1128483491413828, + "learning_rate": 9.915530750383888e-06, + "loss": 0.6047, + "step": 3945 + }, + { + "epoch": 0.08685890380805635, + "grad_norm": 2.309381851114229, + "learning_rate": 9.915465493425414e-06, + "loss": 0.3799, + "step": 3946 + }, + { + "epoch": 0.08688091569447502, + "grad_norm": 1.5233613288460146, + "learning_rate": 9.91540021148435e-06, + "loss": 0.3271, + "step": 3947 + }, + { + "epoch": 0.08690292758089369, + "grad_norm": 1.3948573624928564, + "learning_rate": 9.91533490456103e-06, + "loss": 0.422, + "step": 3948 + }, + { + "epoch": 0.08692493946731235, + "grad_norm": 3.4046824663089934, + "learning_rate": 9.915269572655789e-06, + "loss": 0.3904, + "step": 3949 + }, + { + "epoch": 0.08694695135373101, + "grad_norm": 2.120291582942575, + "learning_rate": 9.915204215768954e-06, + "loss": 0.4276, + "step": 3950 + }, + { + "epoch": 0.08696896324014967, + "grad_norm": 1.7413205136029526, + "learning_rate": 9.915138833900858e-06, + "loss": 0.3661, + "step": 3951 + }, + { + "epoch": 0.08699097512656835, + "grad_norm": 1.8497249362878851, + "learning_rate": 9.915073427051835e-06, + "loss": 0.4024, + "step": 3952 + }, + { + "epoch": 0.08701298701298701, + "grad_norm": 1.3451002406484214, + "learning_rate": 9.915007995222217e-06, + "loss": 0.3862, + "step": 3953 + }, + { + "epoch": 0.08703499889940568, + "grad_norm": 1.6276685785370475, + "learning_rate": 9.914942538412336e-06, + "loss": 0.3617, + "step": 3954 + }, + { + "epoch": 0.08705701078582434, + "grad_norm": 0.9034392113092535, + "learning_rate": 9.914877056622524e-06, + "loss": 0.582, + "step": 3955 + }, + { + "epoch": 0.08707902267224302, + "grad_norm": 1.8800937501922557, + "learning_rate": 9.914811549853115e-06, + "loss": 0.3768, + "step": 3956 + }, + { + "epoch": 0.08710103455866168, + "grad_norm": 0.7358352292759505, + "learning_rate": 9.914746018104443e-06, + "loss": 0.5721, + "step": 3957 + }, + { + "epoch": 0.08712304644508034, + "grad_norm": 1.4785147455100653, + "learning_rate": 9.914680461376838e-06, + "loss": 0.4167, + "step": 3958 + }, + { + "epoch": 0.087145058331499, + "grad_norm": 1.883577074651031, + "learning_rate": 9.914614879670634e-06, + "loss": 0.4564, + "step": 3959 + }, + { + "epoch": 0.08716707021791767, + "grad_norm": 1.7476936735932944, + "learning_rate": 9.914549272986166e-06, + "loss": 0.4006, + "step": 3960 + }, + { + "epoch": 0.08718908210433635, + "grad_norm": 2.1848494072775226, + "learning_rate": 9.914483641323769e-06, + "loss": 0.3164, + "step": 3961 + }, + { + "epoch": 0.08721109399075501, + "grad_norm": 1.4410877237568094, + "learning_rate": 9.914417984683773e-06, + "loss": 0.3431, + "step": 3962 + }, + { + "epoch": 0.08723310587717367, + "grad_norm": 1.8301934778492912, + "learning_rate": 9.914352303066513e-06, + "loss": 0.4449, + "step": 3963 + }, + { + "epoch": 0.08725511776359234, + "grad_norm": 1.8778929799105883, + "learning_rate": 9.914286596472322e-06, + "loss": 0.4273, + "step": 3964 + }, + { + "epoch": 0.08727712965001101, + "grad_norm": 1.6013047763158395, + "learning_rate": 9.914220864901533e-06, + "loss": 0.3455, + "step": 3965 + }, + { + "epoch": 0.08729914153642968, + "grad_norm": 1.6526329377465936, + "learning_rate": 9.914155108354485e-06, + "loss": 0.3591, + "step": 3966 + }, + { + "epoch": 0.08732115342284834, + "grad_norm": 1.782228824427305, + "learning_rate": 9.914089326831507e-06, + "loss": 0.3418, + "step": 3967 + }, + { + "epoch": 0.087343165309267, + "grad_norm": 2.3240455517961114, + "learning_rate": 9.914023520332936e-06, + "loss": 0.407, + "step": 3968 + }, + { + "epoch": 0.08736517719568566, + "grad_norm": 1.146290462837505, + "learning_rate": 9.913957688859107e-06, + "loss": 0.6294, + "step": 3969 + }, + { + "epoch": 0.08738718908210434, + "grad_norm": 1.383259584962667, + "learning_rate": 9.913891832410352e-06, + "loss": 0.3842, + "step": 3970 + }, + { + "epoch": 0.087409200968523, + "grad_norm": 1.44239330824787, + "learning_rate": 9.913825950987005e-06, + "loss": 0.4015, + "step": 3971 + }, + { + "epoch": 0.08743121285494167, + "grad_norm": 1.592888492486526, + "learning_rate": 9.913760044589407e-06, + "loss": 0.3789, + "step": 3972 + }, + { + "epoch": 0.08745322474136033, + "grad_norm": 1.650939183366858, + "learning_rate": 9.913694113217886e-06, + "loss": 0.3863, + "step": 3973 + }, + { + "epoch": 0.08747523662777901, + "grad_norm": 1.2559733487067215, + "learning_rate": 9.913628156872781e-06, + "loss": 0.292, + "step": 3974 + }, + { + "epoch": 0.08749724851419767, + "grad_norm": 3.6445689083590276, + "learning_rate": 9.913562175554425e-06, + "loss": 0.3692, + "step": 3975 + }, + { + "epoch": 0.08751926040061633, + "grad_norm": 1.6028526131016163, + "learning_rate": 9.913496169263156e-06, + "loss": 0.4713, + "step": 3976 + }, + { + "epoch": 0.087541272287035, + "grad_norm": 1.6077298865261935, + "learning_rate": 9.913430137999307e-06, + "loss": 0.3984, + "step": 3977 + }, + { + "epoch": 0.08756328417345366, + "grad_norm": 1.9804827917355514, + "learning_rate": 9.913364081763215e-06, + "loss": 0.3993, + "step": 3978 + }, + { + "epoch": 0.08758529605987234, + "grad_norm": 0.8822985207158557, + "learning_rate": 9.913298000555217e-06, + "loss": 0.5832, + "step": 3979 + }, + { + "epoch": 0.087607307946291, + "grad_norm": 14.037677879827589, + "learning_rate": 9.913231894375645e-06, + "loss": 0.4009, + "step": 3980 + }, + { + "epoch": 0.08762931983270966, + "grad_norm": 1.840553792857903, + "learning_rate": 9.913165763224837e-06, + "loss": 0.3317, + "step": 3981 + }, + { + "epoch": 0.08765133171912833, + "grad_norm": 1.939555574589522, + "learning_rate": 9.913099607103129e-06, + "loss": 0.2788, + "step": 3982 + }, + { + "epoch": 0.087673343605547, + "grad_norm": 1.7508435797658444, + "learning_rate": 9.91303342601086e-06, + "loss": 0.3973, + "step": 3983 + }, + { + "epoch": 0.08769535549196567, + "grad_norm": 2.194134978579596, + "learning_rate": 9.912967219948362e-06, + "loss": 0.4035, + "step": 3984 + }, + { + "epoch": 0.08771736737838433, + "grad_norm": 0.6893614279819971, + "learning_rate": 9.912900988915973e-06, + "loss": 0.5513, + "step": 3985 + }, + { + "epoch": 0.08773937926480299, + "grad_norm": 1.631692627514827, + "learning_rate": 9.91283473291403e-06, + "loss": 0.3541, + "step": 3986 + }, + { + "epoch": 0.08776139115122165, + "grad_norm": 2.0148576520580392, + "learning_rate": 9.91276845194287e-06, + "loss": 0.4122, + "step": 3987 + }, + { + "epoch": 0.08778340303764033, + "grad_norm": 1.6268322030396163, + "learning_rate": 9.91270214600283e-06, + "loss": 0.3704, + "step": 3988 + }, + { + "epoch": 0.087805414924059, + "grad_norm": 1.5002074247977883, + "learning_rate": 9.912635815094247e-06, + "loss": 0.3398, + "step": 3989 + }, + { + "epoch": 0.08782742681047766, + "grad_norm": 2.0577370654688583, + "learning_rate": 9.912569459217458e-06, + "loss": 0.3415, + "step": 3990 + }, + { + "epoch": 0.08784943869689632, + "grad_norm": 2.8371523182826595, + "learning_rate": 9.912503078372798e-06, + "loss": 0.41, + "step": 3991 + }, + { + "epoch": 0.08787145058331498, + "grad_norm": 1.601851325367492, + "learning_rate": 9.912436672560608e-06, + "loss": 0.3932, + "step": 3992 + }, + { + "epoch": 0.08789346246973366, + "grad_norm": 1.6546588030126246, + "learning_rate": 9.912370241781223e-06, + "loss": 0.3007, + "step": 3993 + }, + { + "epoch": 0.08791547435615232, + "grad_norm": 1.2325915364296425, + "learning_rate": 9.91230378603498e-06, + "loss": 0.3238, + "step": 3994 + }, + { + "epoch": 0.08793748624257099, + "grad_norm": 1.9596759643326458, + "learning_rate": 9.91223730532222e-06, + "loss": 0.4284, + "step": 3995 + }, + { + "epoch": 0.08795949812898965, + "grad_norm": 1.5177306996525524, + "learning_rate": 9.912170799643281e-06, + "loss": 0.437, + "step": 3996 + }, + { + "epoch": 0.08798151001540833, + "grad_norm": 1.3966779837054044, + "learning_rate": 9.912104268998496e-06, + "loss": 0.4024, + "step": 3997 + }, + { + "epoch": 0.08800352190182699, + "grad_norm": 1.7132716808130783, + "learning_rate": 9.912037713388208e-06, + "loss": 0.3454, + "step": 3998 + }, + { + "epoch": 0.08802553378824565, + "grad_norm": 3.003551165984273, + "learning_rate": 9.911971132812752e-06, + "loss": 0.3709, + "step": 3999 + }, + { + "epoch": 0.08804754567466432, + "grad_norm": 1.8627602581864968, + "learning_rate": 9.911904527272468e-06, + "loss": 0.3095, + "step": 4000 + }, + { + "epoch": 0.08806955756108298, + "grad_norm": 1.7475482202426813, + "learning_rate": 9.911837896767696e-06, + "loss": 0.4325, + "step": 4001 + }, + { + "epoch": 0.08809156944750166, + "grad_norm": 1.869198748951564, + "learning_rate": 9.911771241298773e-06, + "loss": 0.3701, + "step": 4002 + }, + { + "epoch": 0.08811358133392032, + "grad_norm": 7.668253794642832, + "learning_rate": 9.911704560866035e-06, + "loss": 0.4116, + "step": 4003 + }, + { + "epoch": 0.08813559322033898, + "grad_norm": 1.265455577749812, + "learning_rate": 9.911637855469827e-06, + "loss": 0.2657, + "step": 4004 + }, + { + "epoch": 0.08815760510675764, + "grad_norm": 1.454630225347315, + "learning_rate": 9.911571125110484e-06, + "loss": 0.3299, + "step": 4005 + }, + { + "epoch": 0.08817961699317632, + "grad_norm": 1.6434580028027888, + "learning_rate": 9.911504369788345e-06, + "loss": 0.3697, + "step": 4006 + }, + { + "epoch": 0.08820162887959498, + "grad_norm": 1.777083532658385, + "learning_rate": 9.911437589503751e-06, + "loss": 0.3313, + "step": 4007 + }, + { + "epoch": 0.08822364076601365, + "grad_norm": 2.0041570756716682, + "learning_rate": 9.91137078425704e-06, + "loss": 0.446, + "step": 4008 + }, + { + "epoch": 0.08824565265243231, + "grad_norm": 1.6217608035925346, + "learning_rate": 9.911303954048553e-06, + "loss": 0.4117, + "step": 4009 + }, + { + "epoch": 0.08826766453885097, + "grad_norm": 0.8848857784601691, + "learning_rate": 9.911237098878627e-06, + "loss": 0.5912, + "step": 4010 + }, + { + "epoch": 0.08828967642526965, + "grad_norm": 0.7673930167094503, + "learning_rate": 9.911170218747606e-06, + "loss": 0.5912, + "step": 4011 + }, + { + "epoch": 0.08831168831168831, + "grad_norm": 1.3982933861007822, + "learning_rate": 9.911103313655825e-06, + "loss": 0.3238, + "step": 4012 + }, + { + "epoch": 0.08833370019810698, + "grad_norm": 2.31527405552755, + "learning_rate": 9.91103638360363e-06, + "loss": 0.4869, + "step": 4013 + }, + { + "epoch": 0.08835571208452564, + "grad_norm": 1.4730424556173627, + "learning_rate": 9.910969428591355e-06, + "loss": 0.4269, + "step": 4014 + }, + { + "epoch": 0.08837772397094432, + "grad_norm": 2.486727579664422, + "learning_rate": 9.910902448619343e-06, + "loss": 0.3349, + "step": 4015 + }, + { + "epoch": 0.08839973585736298, + "grad_norm": 1.4605274259416383, + "learning_rate": 9.910835443687936e-06, + "loss": 0.3772, + "step": 4016 + }, + { + "epoch": 0.08842174774378164, + "grad_norm": 1.5726240215772902, + "learning_rate": 9.910768413797471e-06, + "loss": 0.447, + "step": 4017 + }, + { + "epoch": 0.0884437596302003, + "grad_norm": 1.4075707030722695, + "learning_rate": 9.910701358948291e-06, + "loss": 0.2995, + "step": 4018 + }, + { + "epoch": 0.08846577151661897, + "grad_norm": 0.9803641487743556, + "learning_rate": 9.910634279140737e-06, + "loss": 0.5678, + "step": 4019 + }, + { + "epoch": 0.08848778340303765, + "grad_norm": 1.6593508621434392, + "learning_rate": 9.91056717437515e-06, + "loss": 0.3022, + "step": 4020 + }, + { + "epoch": 0.08850979528945631, + "grad_norm": 1.7720142375122236, + "learning_rate": 9.91050004465187e-06, + "loss": 0.4154, + "step": 4021 + }, + { + "epoch": 0.08853180717587497, + "grad_norm": 1.597923676816879, + "learning_rate": 9.91043288997124e-06, + "loss": 0.3861, + "step": 4022 + }, + { + "epoch": 0.08855381906229363, + "grad_norm": 2.1680519595860073, + "learning_rate": 9.910365710333597e-06, + "loss": 0.3948, + "step": 4023 + }, + { + "epoch": 0.08857583094871231, + "grad_norm": 1.6405809506392994, + "learning_rate": 9.910298505739287e-06, + "loss": 0.2758, + "step": 4024 + }, + { + "epoch": 0.08859784283513097, + "grad_norm": 1.5371861565620877, + "learning_rate": 9.91023127618865e-06, + "loss": 0.3522, + "step": 4025 + }, + { + "epoch": 0.08861985472154964, + "grad_norm": 1.6283112643686606, + "learning_rate": 9.910164021682028e-06, + "loss": 0.3164, + "step": 4026 + }, + { + "epoch": 0.0886418666079683, + "grad_norm": 1.610397539047137, + "learning_rate": 9.910096742219761e-06, + "loss": 0.3301, + "step": 4027 + }, + { + "epoch": 0.08866387849438696, + "grad_norm": 1.7136496045688858, + "learning_rate": 9.910029437802194e-06, + "loss": 0.3871, + "step": 4028 + }, + { + "epoch": 0.08868589038080564, + "grad_norm": 1.474233738795821, + "learning_rate": 9.909962108429666e-06, + "loss": 0.3688, + "step": 4029 + }, + { + "epoch": 0.0887079022672243, + "grad_norm": 1.372660901764328, + "learning_rate": 9.909894754102521e-06, + "loss": 0.3567, + "step": 4030 + }, + { + "epoch": 0.08872991415364297, + "grad_norm": 2.4126702830212836, + "learning_rate": 9.909827374821102e-06, + "loss": 0.3358, + "step": 4031 + }, + { + "epoch": 0.08875192604006163, + "grad_norm": 2.4927611094492472, + "learning_rate": 9.909759970585751e-06, + "loss": 0.3201, + "step": 4032 + }, + { + "epoch": 0.08877393792648029, + "grad_norm": 5.088690126837452, + "learning_rate": 9.909692541396806e-06, + "loss": 0.2692, + "step": 4033 + }, + { + "epoch": 0.08879594981289897, + "grad_norm": 2.3433042541723923, + "learning_rate": 9.909625087254618e-06, + "loss": 0.3853, + "step": 4034 + }, + { + "epoch": 0.08881796169931763, + "grad_norm": 2.560064130304038, + "learning_rate": 9.909557608159524e-06, + "loss": 0.2648, + "step": 4035 + }, + { + "epoch": 0.0888399735857363, + "grad_norm": 1.3377770205797794, + "learning_rate": 9.90949010411187e-06, + "loss": 0.4161, + "step": 4036 + }, + { + "epoch": 0.08886198547215496, + "grad_norm": 1.8979930604380921, + "learning_rate": 9.909422575111995e-06, + "loss": 0.3213, + "step": 4037 + }, + { + "epoch": 0.08888399735857364, + "grad_norm": 1.418602885562994, + "learning_rate": 9.909355021160246e-06, + "loss": 0.3586, + "step": 4038 + }, + { + "epoch": 0.0889060092449923, + "grad_norm": 1.6756500453159848, + "learning_rate": 9.909287442256965e-06, + "loss": 0.4135, + "step": 4039 + }, + { + "epoch": 0.08892802113141096, + "grad_norm": 0.7679592150140176, + "learning_rate": 9.909219838402497e-06, + "loss": 0.5763, + "step": 4040 + }, + { + "epoch": 0.08895003301782962, + "grad_norm": 3.041382779615065, + "learning_rate": 9.909152209597182e-06, + "loss": 0.4692, + "step": 4041 + }, + { + "epoch": 0.08897204490424829, + "grad_norm": 1.8660716863168156, + "learning_rate": 9.909084555841368e-06, + "loss": 0.3769, + "step": 4042 + }, + { + "epoch": 0.08899405679066696, + "grad_norm": 1.6470147391084222, + "learning_rate": 9.909016877135395e-06, + "loss": 0.3781, + "step": 4043 + }, + { + "epoch": 0.08901606867708563, + "grad_norm": 1.6484474258587694, + "learning_rate": 9.90894917347961e-06, + "loss": 0.4347, + "step": 4044 + }, + { + "epoch": 0.08903808056350429, + "grad_norm": 1.3932035662371758, + "learning_rate": 9.908881444874354e-06, + "loss": 0.3488, + "step": 4045 + }, + { + "epoch": 0.08906009244992295, + "grad_norm": 1.9619108663423748, + "learning_rate": 9.908813691319975e-06, + "loss": 0.4367, + "step": 4046 + }, + { + "epoch": 0.08908210433634163, + "grad_norm": 1.5224359988644056, + "learning_rate": 9.908745912816814e-06, + "loss": 0.3985, + "step": 4047 + }, + { + "epoch": 0.0891041162227603, + "grad_norm": 0.6564178146919337, + "learning_rate": 9.908678109365218e-06, + "loss": 0.5806, + "step": 4048 + }, + { + "epoch": 0.08912612810917896, + "grad_norm": 1.316034946230516, + "learning_rate": 9.908610280965532e-06, + "loss": 0.3178, + "step": 4049 + }, + { + "epoch": 0.08914813999559762, + "grad_norm": 1.9442502064326985, + "learning_rate": 9.908542427618097e-06, + "loss": 0.3606, + "step": 4050 + }, + { + "epoch": 0.08917015188201628, + "grad_norm": 2.284606098945343, + "learning_rate": 9.908474549323259e-06, + "loss": 0.4203, + "step": 4051 + }, + { + "epoch": 0.08919216376843496, + "grad_norm": 0.6549702931342987, + "learning_rate": 9.908406646081366e-06, + "loss": 0.5872, + "step": 4052 + }, + { + "epoch": 0.08921417565485362, + "grad_norm": 1.8965692207674165, + "learning_rate": 9.908338717892761e-06, + "loss": 0.3884, + "step": 4053 + }, + { + "epoch": 0.08923618754127229, + "grad_norm": 1.4748336396130377, + "learning_rate": 9.90827076475779e-06, + "loss": 0.3921, + "step": 4054 + }, + { + "epoch": 0.08925819942769095, + "grad_norm": 2.2471712612533223, + "learning_rate": 9.908202786676798e-06, + "loss": 0.3465, + "step": 4055 + }, + { + "epoch": 0.08928021131410963, + "grad_norm": 1.4964562662802035, + "learning_rate": 9.908134783650128e-06, + "loss": 0.3467, + "step": 4056 + }, + { + "epoch": 0.08930222320052829, + "grad_norm": 1.5116800247436353, + "learning_rate": 9.90806675567813e-06, + "loss": 0.444, + "step": 4057 + }, + { + "epoch": 0.08932423508694695, + "grad_norm": 0.6831010812563681, + "learning_rate": 9.907998702761148e-06, + "loss": 0.5776, + "step": 4058 + }, + { + "epoch": 0.08934624697336561, + "grad_norm": 1.703891531148532, + "learning_rate": 9.907930624899527e-06, + "loss": 0.3923, + "step": 4059 + }, + { + "epoch": 0.08936825885978428, + "grad_norm": 1.9550782463100473, + "learning_rate": 9.907862522093612e-06, + "loss": 0.4769, + "step": 4060 + }, + { + "epoch": 0.08939027074620295, + "grad_norm": 1.4536457689408424, + "learning_rate": 9.907794394343753e-06, + "loss": 0.3507, + "step": 4061 + }, + { + "epoch": 0.08941228263262162, + "grad_norm": 1.705966294196739, + "learning_rate": 9.907726241650293e-06, + "loss": 0.3674, + "step": 4062 + }, + { + "epoch": 0.08943429451904028, + "grad_norm": 0.6610184256913747, + "learning_rate": 9.907658064013578e-06, + "loss": 0.5899, + "step": 4063 + }, + { + "epoch": 0.08945630640545894, + "grad_norm": 1.4808670074358221, + "learning_rate": 9.907589861433957e-06, + "loss": 0.4128, + "step": 4064 + }, + { + "epoch": 0.08947831829187762, + "grad_norm": 2.052573533017279, + "learning_rate": 9.907521633911776e-06, + "loss": 0.2615, + "step": 4065 + }, + { + "epoch": 0.08950033017829628, + "grad_norm": 1.6933595373453576, + "learning_rate": 9.90745338144738e-06, + "loss": 0.3447, + "step": 4066 + }, + { + "epoch": 0.08952234206471495, + "grad_norm": 1.6003957368214248, + "learning_rate": 9.907385104041118e-06, + "loss": 0.3858, + "step": 4067 + }, + { + "epoch": 0.08954435395113361, + "grad_norm": 1.51733386793898, + "learning_rate": 9.907316801693337e-06, + "loss": 0.3757, + "step": 4068 + }, + { + "epoch": 0.08956636583755227, + "grad_norm": 0.6350233050785751, + "learning_rate": 9.907248474404382e-06, + "loss": 0.5845, + "step": 4069 + }, + { + "epoch": 0.08958837772397095, + "grad_norm": 1.9954214208894276, + "learning_rate": 9.907180122174602e-06, + "loss": 0.3178, + "step": 4070 + }, + { + "epoch": 0.08961038961038961, + "grad_norm": 1.2279681901587884, + "learning_rate": 9.907111745004345e-06, + "loss": 0.3521, + "step": 4071 + }, + { + "epoch": 0.08963240149680828, + "grad_norm": 1.499618109062525, + "learning_rate": 9.907043342893956e-06, + "loss": 0.3665, + "step": 4072 + }, + { + "epoch": 0.08965441338322694, + "grad_norm": 0.6575938113447665, + "learning_rate": 9.906974915843784e-06, + "loss": 0.5775, + "step": 4073 + }, + { + "epoch": 0.08967642526964562, + "grad_norm": 1.3852939572412368, + "learning_rate": 9.906906463854178e-06, + "loss": 0.2878, + "step": 4074 + }, + { + "epoch": 0.08969843715606428, + "grad_norm": 2.9511631279453625, + "learning_rate": 9.906837986925484e-06, + "loss": 0.3497, + "step": 4075 + }, + { + "epoch": 0.08972044904248294, + "grad_norm": 1.5809951770745077, + "learning_rate": 9.90676948505805e-06, + "loss": 0.4198, + "step": 4076 + }, + { + "epoch": 0.0897424609289016, + "grad_norm": 4.103975933680601, + "learning_rate": 9.906700958252227e-06, + "loss": 0.4303, + "step": 4077 + }, + { + "epoch": 0.08976447281532027, + "grad_norm": 1.637799062938848, + "learning_rate": 9.90663240650836e-06, + "loss": 0.3163, + "step": 4078 + }, + { + "epoch": 0.08978648470173894, + "grad_norm": 2.0127368266080556, + "learning_rate": 9.906563829826799e-06, + "loss": 0.3513, + "step": 4079 + }, + { + "epoch": 0.08980849658815761, + "grad_norm": 1.630934561197398, + "learning_rate": 9.906495228207893e-06, + "loss": 0.3988, + "step": 4080 + }, + { + "epoch": 0.08983050847457627, + "grad_norm": 1.6013168525703236, + "learning_rate": 9.906426601651988e-06, + "loss": 0.3736, + "step": 4081 + }, + { + "epoch": 0.08985252036099493, + "grad_norm": 1.8429525834997211, + "learning_rate": 9.906357950159437e-06, + "loss": 0.3213, + "step": 4082 + }, + { + "epoch": 0.0898745322474136, + "grad_norm": 1.3516334308155813, + "learning_rate": 9.906289273730584e-06, + "loss": 0.3805, + "step": 4083 + }, + { + "epoch": 0.08989654413383227, + "grad_norm": 1.8316648453430766, + "learning_rate": 9.906220572365783e-06, + "loss": 0.338, + "step": 4084 + }, + { + "epoch": 0.08991855602025094, + "grad_norm": 0.6960688154656379, + "learning_rate": 9.90615184606538e-06, + "loss": 0.5652, + "step": 4085 + }, + { + "epoch": 0.0899405679066696, + "grad_norm": 1.8567619662989863, + "learning_rate": 9.906083094829724e-06, + "loss": 0.4013, + "step": 4086 + }, + { + "epoch": 0.08996257979308826, + "grad_norm": 1.9373830512133727, + "learning_rate": 9.906014318659168e-06, + "loss": 0.4274, + "step": 4087 + }, + { + "epoch": 0.08998459167950694, + "grad_norm": 1.5317015807563494, + "learning_rate": 9.905945517554057e-06, + "loss": 0.3486, + "step": 4088 + }, + { + "epoch": 0.0900066035659256, + "grad_norm": 5.076307791729852, + "learning_rate": 9.905876691514742e-06, + "loss": 0.4036, + "step": 4089 + }, + { + "epoch": 0.09002861545234427, + "grad_norm": 13.2213231880001, + "learning_rate": 9.905807840541577e-06, + "loss": 0.4701, + "step": 4090 + }, + { + "epoch": 0.09005062733876293, + "grad_norm": 1.824780278319761, + "learning_rate": 9.905738964634905e-06, + "loss": 0.4015, + "step": 4091 + }, + { + "epoch": 0.09007263922518159, + "grad_norm": 8.210306274763719, + "learning_rate": 9.905670063795081e-06, + "loss": 0.3338, + "step": 4092 + }, + { + "epoch": 0.09009465111160027, + "grad_norm": 2.0589813727370996, + "learning_rate": 9.905601138022455e-06, + "loss": 0.4459, + "step": 4093 + }, + { + "epoch": 0.09011666299801893, + "grad_norm": 1.6360288001429604, + "learning_rate": 9.905532187317375e-06, + "loss": 0.4033, + "step": 4094 + }, + { + "epoch": 0.0901386748844376, + "grad_norm": 1.3834560021437405, + "learning_rate": 9.905463211680192e-06, + "loss": 0.3365, + "step": 4095 + }, + { + "epoch": 0.09016068677085626, + "grad_norm": 1.9844479964014048, + "learning_rate": 9.905394211111256e-06, + "loss": 0.4043, + "step": 4096 + }, + { + "epoch": 0.09018269865727493, + "grad_norm": 1.933383604596995, + "learning_rate": 9.905325185610921e-06, + "loss": 0.3109, + "step": 4097 + }, + { + "epoch": 0.0902047105436936, + "grad_norm": 0.647244379868464, + "learning_rate": 9.905256135179536e-06, + "loss": 0.598, + "step": 4098 + }, + { + "epoch": 0.09022672243011226, + "grad_norm": 1.717756431548549, + "learning_rate": 9.905187059817449e-06, + "loss": 0.355, + "step": 4099 + }, + { + "epoch": 0.09024873431653092, + "grad_norm": 1.7940857868656168, + "learning_rate": 9.905117959525015e-06, + "loss": 0.4371, + "step": 4100 + }, + { + "epoch": 0.09027074620294959, + "grad_norm": 0.6277376408456891, + "learning_rate": 9.905048834302585e-06, + "loss": 0.5636, + "step": 4101 + }, + { + "epoch": 0.09029275808936826, + "grad_norm": 1.3246257243947568, + "learning_rate": 9.904979684150507e-06, + "loss": 0.3404, + "step": 4102 + }, + { + "epoch": 0.09031476997578693, + "grad_norm": 2.5634849192472626, + "learning_rate": 9.904910509069134e-06, + "loss": 0.3565, + "step": 4103 + }, + { + "epoch": 0.09033678186220559, + "grad_norm": 4.897804951221928, + "learning_rate": 9.90484130905882e-06, + "loss": 0.4797, + "step": 4104 + }, + { + "epoch": 0.09035879374862425, + "grad_norm": 5.305089442888751, + "learning_rate": 9.904772084119915e-06, + "loss": 0.373, + "step": 4105 + }, + { + "epoch": 0.09038080563504293, + "grad_norm": 0.632082076486161, + "learning_rate": 9.90470283425277e-06, + "loss": 0.5466, + "step": 4106 + }, + { + "epoch": 0.09040281752146159, + "grad_norm": 1.4713126292206595, + "learning_rate": 9.904633559457737e-06, + "loss": 0.3263, + "step": 4107 + }, + { + "epoch": 0.09042482940788026, + "grad_norm": 2.0264974903941284, + "learning_rate": 9.90456425973517e-06, + "loss": 0.4514, + "step": 4108 + }, + { + "epoch": 0.09044684129429892, + "grad_norm": 1.6294806640624755, + "learning_rate": 9.904494935085419e-06, + "loss": 0.3207, + "step": 4109 + }, + { + "epoch": 0.09046885318071758, + "grad_norm": 1.739140425538713, + "learning_rate": 9.904425585508836e-06, + "loss": 0.3772, + "step": 4110 + }, + { + "epoch": 0.09049086506713626, + "grad_norm": 1.4077890120594143, + "learning_rate": 9.904356211005776e-06, + "loss": 0.3932, + "step": 4111 + }, + { + "epoch": 0.09051287695355492, + "grad_norm": 1.457316224794754, + "learning_rate": 9.904286811576591e-06, + "loss": 0.3282, + "step": 4112 + }, + { + "epoch": 0.09053488883997358, + "grad_norm": 1.7153858942467275, + "learning_rate": 9.904217387221633e-06, + "loss": 0.4714, + "step": 4113 + }, + { + "epoch": 0.09055690072639225, + "grad_norm": 1.6951322849076764, + "learning_rate": 9.904147937941256e-06, + "loss": 0.3623, + "step": 4114 + }, + { + "epoch": 0.09057891261281092, + "grad_norm": 1.8064199101882916, + "learning_rate": 9.904078463735809e-06, + "loss": 0.3165, + "step": 4115 + }, + { + "epoch": 0.09060092449922959, + "grad_norm": 1.4837529232001192, + "learning_rate": 9.90400896460565e-06, + "loss": 0.3535, + "step": 4116 + }, + { + "epoch": 0.09062293638564825, + "grad_norm": 1.5918711668629593, + "learning_rate": 9.90393944055113e-06, + "loss": 0.3884, + "step": 4117 + }, + { + "epoch": 0.09064494827206691, + "grad_norm": 1.5317190389695028, + "learning_rate": 9.903869891572601e-06, + "loss": 0.4466, + "step": 4118 + }, + { + "epoch": 0.09066696015848558, + "grad_norm": 1.5189385824265367, + "learning_rate": 9.90380031767042e-06, + "loss": 0.386, + "step": 4119 + }, + { + "epoch": 0.09068897204490425, + "grad_norm": 1.8122328099666267, + "learning_rate": 9.903730718844938e-06, + "loss": 0.2988, + "step": 4120 + }, + { + "epoch": 0.09071098393132292, + "grad_norm": 1.621947510609229, + "learning_rate": 9.903661095096509e-06, + "loss": 0.3246, + "step": 4121 + }, + { + "epoch": 0.09073299581774158, + "grad_norm": 1.9954340411012281, + "learning_rate": 9.903591446425488e-06, + "loss": 0.3892, + "step": 4122 + }, + { + "epoch": 0.09075500770416024, + "grad_norm": 1.9567303322008252, + "learning_rate": 9.903521772832226e-06, + "loss": 0.3721, + "step": 4123 + }, + { + "epoch": 0.0907770195905789, + "grad_norm": 0.649744089253316, + "learning_rate": 9.903452074317081e-06, + "loss": 0.5625, + "step": 4124 + }, + { + "epoch": 0.09079903147699758, + "grad_norm": 1.7006022975678827, + "learning_rate": 9.903382350880407e-06, + "loss": 0.3952, + "step": 4125 + }, + { + "epoch": 0.09082104336341625, + "grad_norm": 1.4994768283090598, + "learning_rate": 9.903312602522555e-06, + "loss": 0.459, + "step": 4126 + }, + { + "epoch": 0.09084305524983491, + "grad_norm": 3.175993164094074, + "learning_rate": 9.903242829243882e-06, + "loss": 0.4179, + "step": 4127 + }, + { + "epoch": 0.09086506713625357, + "grad_norm": 2.0397949104368744, + "learning_rate": 9.90317303104474e-06, + "loss": 0.3425, + "step": 4128 + }, + { + "epoch": 0.09088707902267225, + "grad_norm": 1.444805332348054, + "learning_rate": 9.903103207925489e-06, + "loss": 0.2953, + "step": 4129 + }, + { + "epoch": 0.09090909090909091, + "grad_norm": 1.8703570706296244, + "learning_rate": 9.903033359886478e-06, + "loss": 0.4388, + "step": 4130 + }, + { + "epoch": 0.09093110279550957, + "grad_norm": 1.3748858548163145, + "learning_rate": 9.902963486928066e-06, + "loss": 0.3598, + "step": 4131 + }, + { + "epoch": 0.09095311468192824, + "grad_norm": 3.855303003747126, + "learning_rate": 9.902893589050607e-06, + "loss": 0.3938, + "step": 4132 + }, + { + "epoch": 0.0909751265683469, + "grad_norm": 1.5669012631588648, + "learning_rate": 9.902823666254455e-06, + "loss": 0.4826, + "step": 4133 + }, + { + "epoch": 0.09099713845476558, + "grad_norm": 1.5330976472446853, + "learning_rate": 9.902753718539968e-06, + "loss": 0.4756, + "step": 4134 + }, + { + "epoch": 0.09101915034118424, + "grad_norm": 1.388323507843803, + "learning_rate": 9.902683745907499e-06, + "loss": 0.3871, + "step": 4135 + }, + { + "epoch": 0.0910411622276029, + "grad_norm": 1.7228211841783023, + "learning_rate": 9.902613748357403e-06, + "loss": 0.374, + "step": 4136 + }, + { + "epoch": 0.09106317411402157, + "grad_norm": 2.010500807058215, + "learning_rate": 9.90254372589004e-06, + "loss": 0.3626, + "step": 4137 + }, + { + "epoch": 0.09108518600044024, + "grad_norm": 2.4129119756957067, + "learning_rate": 9.902473678505763e-06, + "loss": 0.4362, + "step": 4138 + }, + { + "epoch": 0.0911071978868589, + "grad_norm": 0.6965854603931878, + "learning_rate": 9.902403606204926e-06, + "loss": 0.5729, + "step": 4139 + }, + { + "epoch": 0.09112920977327757, + "grad_norm": 0.6482576490853392, + "learning_rate": 9.902333508987889e-06, + "loss": 0.571, + "step": 4140 + }, + { + "epoch": 0.09115122165969623, + "grad_norm": 0.6248552857023727, + "learning_rate": 9.902263386855006e-06, + "loss": 0.5724, + "step": 4141 + }, + { + "epoch": 0.0911732335461149, + "grad_norm": 1.492300482825472, + "learning_rate": 9.902193239806634e-06, + "loss": 0.3727, + "step": 4142 + }, + { + "epoch": 0.09119524543253357, + "grad_norm": 6.549995051947381, + "learning_rate": 9.90212306784313e-06, + "loss": 0.3576, + "step": 4143 + }, + { + "epoch": 0.09121725731895224, + "grad_norm": 2.3240906358048643, + "learning_rate": 9.90205287096485e-06, + "loss": 0.3376, + "step": 4144 + }, + { + "epoch": 0.0912392692053709, + "grad_norm": 3.7291443682426295, + "learning_rate": 9.901982649172151e-06, + "loss": 0.3393, + "step": 4145 + }, + { + "epoch": 0.09126128109178956, + "grad_norm": 2.144136080649643, + "learning_rate": 9.90191240246539e-06, + "loss": 0.4085, + "step": 4146 + }, + { + "epoch": 0.09128329297820824, + "grad_norm": 1.6725274087487174, + "learning_rate": 9.901842130844924e-06, + "loss": 0.3374, + "step": 4147 + }, + { + "epoch": 0.0913053048646269, + "grad_norm": 2.3105398496410894, + "learning_rate": 9.901771834311108e-06, + "loss": 0.3985, + "step": 4148 + }, + { + "epoch": 0.09132731675104556, + "grad_norm": 2.5447723730805762, + "learning_rate": 9.901701512864304e-06, + "loss": 0.4943, + "step": 4149 + }, + { + "epoch": 0.09134932863746423, + "grad_norm": 2.1202876851766166, + "learning_rate": 9.901631166504865e-06, + "loss": 0.4673, + "step": 4150 + }, + { + "epoch": 0.09137134052388289, + "grad_norm": 1.935830979180387, + "learning_rate": 9.901560795233151e-06, + "loss": 0.3749, + "step": 4151 + }, + { + "epoch": 0.09139335241030157, + "grad_norm": 2.732655809278713, + "learning_rate": 9.901490399049518e-06, + "loss": 0.3812, + "step": 4152 + }, + { + "epoch": 0.09141536429672023, + "grad_norm": 0.9061943167692753, + "learning_rate": 9.901419977954325e-06, + "loss": 0.5574, + "step": 4153 + }, + { + "epoch": 0.0914373761831389, + "grad_norm": 1.7590042640714278, + "learning_rate": 9.90134953194793e-06, + "loss": 0.3074, + "step": 4154 + }, + { + "epoch": 0.09145938806955756, + "grad_norm": 0.7396441388597478, + "learning_rate": 9.901279061030691e-06, + "loss": 0.5961, + "step": 4155 + }, + { + "epoch": 0.09148139995597623, + "grad_norm": 1.7387899345790119, + "learning_rate": 9.901208565202964e-06, + "loss": 0.4888, + "step": 4156 + }, + { + "epoch": 0.0915034118423949, + "grad_norm": 1.3617382014751687, + "learning_rate": 9.901138044465111e-06, + "loss": 0.3171, + "step": 4157 + }, + { + "epoch": 0.09152542372881356, + "grad_norm": 1.6161603543185745, + "learning_rate": 9.901067498817486e-06, + "loss": 0.3529, + "step": 4158 + }, + { + "epoch": 0.09154743561523222, + "grad_norm": 1.7262011828935646, + "learning_rate": 9.900996928260454e-06, + "loss": 0.4916, + "step": 4159 + }, + { + "epoch": 0.09156944750165089, + "grad_norm": 1.8890264150819398, + "learning_rate": 9.900926332794368e-06, + "loss": 0.4243, + "step": 4160 + }, + { + "epoch": 0.09159145938806956, + "grad_norm": 1.9211673742289461, + "learning_rate": 9.900855712419587e-06, + "loss": 0.4021, + "step": 4161 + }, + { + "epoch": 0.09161347127448823, + "grad_norm": 1.6877541937677554, + "learning_rate": 9.900785067136473e-06, + "loss": 0.4382, + "step": 4162 + }, + { + "epoch": 0.09163548316090689, + "grad_norm": 1.8348644212802527, + "learning_rate": 9.900714396945382e-06, + "loss": 0.3371, + "step": 4163 + }, + { + "epoch": 0.09165749504732555, + "grad_norm": 1.0542168903230826, + "learning_rate": 9.900643701846676e-06, + "loss": 0.5775, + "step": 4164 + }, + { + "epoch": 0.09167950693374423, + "grad_norm": 1.233639575106887, + "learning_rate": 9.900572981840712e-06, + "loss": 0.251, + "step": 4165 + }, + { + "epoch": 0.09170151882016289, + "grad_norm": 1.5100107233555706, + "learning_rate": 9.900502236927851e-06, + "loss": 0.3917, + "step": 4166 + }, + { + "epoch": 0.09172353070658155, + "grad_norm": 1.703831305809837, + "learning_rate": 9.900431467108452e-06, + "loss": 0.3925, + "step": 4167 + }, + { + "epoch": 0.09174554259300022, + "grad_norm": 1.4223252492289822, + "learning_rate": 9.900360672382875e-06, + "loss": 0.2742, + "step": 4168 + }, + { + "epoch": 0.09176755447941888, + "grad_norm": 1.6414504520006754, + "learning_rate": 9.900289852751479e-06, + "loss": 0.4115, + "step": 4169 + }, + { + "epoch": 0.09178956636583756, + "grad_norm": 1.5099943894032684, + "learning_rate": 9.900219008214625e-06, + "loss": 0.3453, + "step": 4170 + }, + { + "epoch": 0.09181157825225622, + "grad_norm": 0.7430301894547156, + "learning_rate": 9.900148138772672e-06, + "loss": 0.5669, + "step": 4171 + }, + { + "epoch": 0.09183359013867488, + "grad_norm": 0.7246025629294616, + "learning_rate": 9.900077244425979e-06, + "loss": 0.5813, + "step": 4172 + }, + { + "epoch": 0.09185560202509355, + "grad_norm": 2.7762011239557056, + "learning_rate": 9.90000632517491e-06, + "loss": 0.4218, + "step": 4173 + }, + { + "epoch": 0.09187761391151221, + "grad_norm": 2.5124820274361226, + "learning_rate": 9.899935381019823e-06, + "loss": 0.4406, + "step": 4174 + }, + { + "epoch": 0.09189962579793089, + "grad_norm": 1.9431274613762215, + "learning_rate": 9.899864411961077e-06, + "loss": 0.3764, + "step": 4175 + }, + { + "epoch": 0.09192163768434955, + "grad_norm": 3.177508544164446, + "learning_rate": 9.899793417999037e-06, + "loss": 0.4868, + "step": 4176 + }, + { + "epoch": 0.09194364957076821, + "grad_norm": 3.308218293835187, + "learning_rate": 9.899722399134062e-06, + "loss": 0.3263, + "step": 4177 + }, + { + "epoch": 0.09196566145718688, + "grad_norm": 1.5498596798719895, + "learning_rate": 9.89965135536651e-06, + "loss": 0.3598, + "step": 4178 + }, + { + "epoch": 0.09198767334360555, + "grad_norm": 1.591952112061146, + "learning_rate": 9.899580286696745e-06, + "loss": 0.4383, + "step": 4179 + }, + { + "epoch": 0.09200968523002422, + "grad_norm": 1.2498619597802674, + "learning_rate": 9.89950919312513e-06, + "loss": 0.292, + "step": 4180 + }, + { + "epoch": 0.09203169711644288, + "grad_norm": 2.0367180918639556, + "learning_rate": 9.899438074652022e-06, + "loss": 0.4449, + "step": 4181 + }, + { + "epoch": 0.09205370900286154, + "grad_norm": 1.115591369908098, + "learning_rate": 9.899366931277784e-06, + "loss": 0.5648, + "step": 4182 + }, + { + "epoch": 0.0920757208892802, + "grad_norm": 1.9889719253199436, + "learning_rate": 9.899295763002781e-06, + "loss": 0.4918, + "step": 4183 + }, + { + "epoch": 0.09209773277569888, + "grad_norm": 1.7529828748546192, + "learning_rate": 9.89922456982737e-06, + "loss": 0.3356, + "step": 4184 + }, + { + "epoch": 0.09211974466211754, + "grad_norm": 5.91895866285259, + "learning_rate": 9.899153351751914e-06, + "loss": 0.4456, + "step": 4185 + }, + { + "epoch": 0.09214175654853621, + "grad_norm": 1.8830868794087474, + "learning_rate": 9.899082108776776e-06, + "loss": 0.2913, + "step": 4186 + }, + { + "epoch": 0.09216376843495487, + "grad_norm": 1.9222108365864559, + "learning_rate": 9.899010840902318e-06, + "loss": 0.3473, + "step": 4187 + }, + { + "epoch": 0.09218578032137355, + "grad_norm": 1.746828849812745, + "learning_rate": 9.8989395481289e-06, + "loss": 0.3454, + "step": 4188 + }, + { + "epoch": 0.09220779220779221, + "grad_norm": 1.7861946587495245, + "learning_rate": 9.898868230456889e-06, + "loss": 0.3647, + "step": 4189 + }, + { + "epoch": 0.09222980409421087, + "grad_norm": 1.705344528408941, + "learning_rate": 9.898796887886644e-06, + "loss": 0.3524, + "step": 4190 + }, + { + "epoch": 0.09225181598062954, + "grad_norm": 0.9866478346129123, + "learning_rate": 9.89872552041853e-06, + "loss": 0.5867, + "step": 4191 + }, + { + "epoch": 0.0922738278670482, + "grad_norm": 1.257576245336448, + "learning_rate": 9.898654128052904e-06, + "loss": 0.3178, + "step": 4192 + }, + { + "epoch": 0.09229583975346688, + "grad_norm": 2.0755756725614933, + "learning_rate": 9.898582710790135e-06, + "loss": 0.4552, + "step": 4193 + }, + { + "epoch": 0.09231785163988554, + "grad_norm": 2.327442930323803, + "learning_rate": 9.898511268630582e-06, + "loss": 0.3785, + "step": 4194 + }, + { + "epoch": 0.0923398635263042, + "grad_norm": 2.1439946165918142, + "learning_rate": 9.898439801574613e-06, + "loss": 0.4057, + "step": 4195 + }, + { + "epoch": 0.09236187541272287, + "grad_norm": 2.671446346994435, + "learning_rate": 9.898368309622586e-06, + "loss": 0.453, + "step": 4196 + }, + { + "epoch": 0.09238388729914154, + "grad_norm": 1.9812048670824225, + "learning_rate": 9.898296792774868e-06, + "loss": 0.4237, + "step": 4197 + }, + { + "epoch": 0.0924058991855602, + "grad_norm": 0.7054070984742024, + "learning_rate": 9.898225251031819e-06, + "loss": 0.5665, + "step": 4198 + }, + { + "epoch": 0.09242791107197887, + "grad_norm": 1.918218344263317, + "learning_rate": 9.898153684393806e-06, + "loss": 0.3792, + "step": 4199 + }, + { + "epoch": 0.09244992295839753, + "grad_norm": 1.2827454592391572, + "learning_rate": 9.898082092861193e-06, + "loss": 0.2699, + "step": 4200 + }, + { + "epoch": 0.0924719348448162, + "grad_norm": 1.8199202202377498, + "learning_rate": 9.898010476434338e-06, + "loss": 0.4393, + "step": 4201 + }, + { + "epoch": 0.09249394673123487, + "grad_norm": 1.7354962630347295, + "learning_rate": 9.897938835113613e-06, + "loss": 0.3944, + "step": 4202 + }, + { + "epoch": 0.09251595861765353, + "grad_norm": 1.4880574969627016, + "learning_rate": 9.897867168899376e-06, + "loss": 0.3986, + "step": 4203 + }, + { + "epoch": 0.0925379705040722, + "grad_norm": 1.9761403196758467, + "learning_rate": 9.897795477791994e-06, + "loss": 0.3869, + "step": 4204 + }, + { + "epoch": 0.09255998239049086, + "grad_norm": 2.1103526303036135, + "learning_rate": 9.89772376179183e-06, + "loss": 0.439, + "step": 4205 + }, + { + "epoch": 0.09258199427690954, + "grad_norm": 1.3891676213115651, + "learning_rate": 9.897652020899253e-06, + "loss": 0.3947, + "step": 4206 + }, + { + "epoch": 0.0926040061633282, + "grad_norm": 1.487289412041378, + "learning_rate": 9.897580255114621e-06, + "loss": 0.4021, + "step": 4207 + }, + { + "epoch": 0.09262601804974686, + "grad_norm": 1.4404479808766226, + "learning_rate": 9.897508464438302e-06, + "loss": 0.3434, + "step": 4208 + }, + { + "epoch": 0.09264802993616553, + "grad_norm": 2.238381637009854, + "learning_rate": 9.897436648870661e-06, + "loss": 0.4411, + "step": 4209 + }, + { + "epoch": 0.09267004182258419, + "grad_norm": 5.3130785919912515, + "learning_rate": 9.897364808412063e-06, + "loss": 0.4631, + "step": 4210 + }, + { + "epoch": 0.09269205370900287, + "grad_norm": 1.895817214189801, + "learning_rate": 9.897292943062872e-06, + "loss": 0.3364, + "step": 4211 + }, + { + "epoch": 0.09271406559542153, + "grad_norm": 1.5227247047843264, + "learning_rate": 9.897221052823456e-06, + "loss": 0.3656, + "step": 4212 + }, + { + "epoch": 0.09273607748184019, + "grad_norm": 1.3351849059077772, + "learning_rate": 9.897149137694175e-06, + "loss": 0.3352, + "step": 4213 + }, + { + "epoch": 0.09275808936825886, + "grad_norm": 1.684380497314817, + "learning_rate": 9.8970771976754e-06, + "loss": 0.4486, + "step": 4214 + }, + { + "epoch": 0.09278010125467753, + "grad_norm": 1.7253823183785453, + "learning_rate": 9.897005232767495e-06, + "loss": 0.4026, + "step": 4215 + }, + { + "epoch": 0.0928021131410962, + "grad_norm": 1.7670674100254362, + "learning_rate": 9.896933242970824e-06, + "loss": 0.3502, + "step": 4216 + }, + { + "epoch": 0.09282412502751486, + "grad_norm": 0.7580292557172996, + "learning_rate": 9.896861228285754e-06, + "loss": 0.5734, + "step": 4217 + }, + { + "epoch": 0.09284613691393352, + "grad_norm": 1.6277844792316023, + "learning_rate": 9.896789188712654e-06, + "loss": 0.3788, + "step": 4218 + }, + { + "epoch": 0.09286814880035218, + "grad_norm": 1.8315829019335264, + "learning_rate": 9.896717124251884e-06, + "loss": 0.5153, + "step": 4219 + }, + { + "epoch": 0.09289016068677086, + "grad_norm": 2.1209463218822693, + "learning_rate": 9.896645034903816e-06, + "loss": 0.4104, + "step": 4220 + }, + { + "epoch": 0.09291217257318952, + "grad_norm": 5.6348639628463735, + "learning_rate": 9.896572920668812e-06, + "loss": 0.3637, + "step": 4221 + }, + { + "epoch": 0.09293418445960819, + "grad_norm": 4.69702742236573, + "learning_rate": 9.89650078154724e-06, + "loss": 0.4816, + "step": 4222 + }, + { + "epoch": 0.09295619634602685, + "grad_norm": 2.077322141464251, + "learning_rate": 9.896428617539469e-06, + "loss": 0.3682, + "step": 4223 + }, + { + "epoch": 0.09297820823244551, + "grad_norm": 0.6474628666626703, + "learning_rate": 9.896356428645863e-06, + "loss": 0.5628, + "step": 4224 + }, + { + "epoch": 0.09300022011886419, + "grad_norm": 1.716524596253899, + "learning_rate": 9.896284214866792e-06, + "loss": 0.3877, + "step": 4225 + }, + { + "epoch": 0.09302223200528285, + "grad_norm": 1.743340168970171, + "learning_rate": 9.896211976202618e-06, + "loss": 0.3622, + "step": 4226 + }, + { + "epoch": 0.09304424389170152, + "grad_norm": 1.4661908996824158, + "learning_rate": 9.896139712653711e-06, + "loss": 0.4291, + "step": 4227 + }, + { + "epoch": 0.09306625577812018, + "grad_norm": 1.682340136685676, + "learning_rate": 9.896067424220439e-06, + "loss": 0.3411, + "step": 4228 + }, + { + "epoch": 0.09308826766453886, + "grad_norm": 0.6809841666371134, + "learning_rate": 9.89599511090317e-06, + "loss": 0.5581, + "step": 4229 + }, + { + "epoch": 0.09311027955095752, + "grad_norm": 1.9201490669680459, + "learning_rate": 9.895922772702269e-06, + "loss": 0.3868, + "step": 4230 + }, + { + "epoch": 0.09313229143737618, + "grad_norm": 1.4739900859320836, + "learning_rate": 9.895850409618104e-06, + "loss": 0.387, + "step": 4231 + }, + { + "epoch": 0.09315430332379485, + "grad_norm": 1.500187987714469, + "learning_rate": 9.895778021651044e-06, + "loss": 0.2741, + "step": 4232 + }, + { + "epoch": 0.09317631521021351, + "grad_norm": 2.3456466033228778, + "learning_rate": 9.895705608801457e-06, + "loss": 0.4426, + "step": 4233 + }, + { + "epoch": 0.09319832709663219, + "grad_norm": 1.6019404626630527, + "learning_rate": 9.89563317106971e-06, + "loss": 0.2894, + "step": 4234 + }, + { + "epoch": 0.09322033898305085, + "grad_norm": 1.464231408452608, + "learning_rate": 9.895560708456173e-06, + "loss": 0.3191, + "step": 4235 + }, + { + "epoch": 0.09324235086946951, + "grad_norm": 1.4712191672387631, + "learning_rate": 9.895488220961212e-06, + "loss": 0.3812, + "step": 4236 + }, + { + "epoch": 0.09326436275588817, + "grad_norm": 1.6019286185987682, + "learning_rate": 9.895415708585196e-06, + "loss": 0.3757, + "step": 4237 + }, + { + "epoch": 0.09328637464230685, + "grad_norm": 1.9372617421315854, + "learning_rate": 9.895343171328495e-06, + "loss": 0.3813, + "step": 4238 + }, + { + "epoch": 0.09330838652872551, + "grad_norm": 1.7291745245232726, + "learning_rate": 9.895270609191476e-06, + "loss": 0.4273, + "step": 4239 + }, + { + "epoch": 0.09333039841514418, + "grad_norm": 1.5819008494919522, + "learning_rate": 9.895198022174509e-06, + "loss": 0.3201, + "step": 4240 + }, + { + "epoch": 0.09335241030156284, + "grad_norm": 2.2345858684881508, + "learning_rate": 9.895125410277961e-06, + "loss": 0.3135, + "step": 4241 + }, + { + "epoch": 0.0933744221879815, + "grad_norm": 1.58565670991234, + "learning_rate": 9.895052773502203e-06, + "loss": 0.3677, + "step": 4242 + }, + { + "epoch": 0.09339643407440018, + "grad_norm": 1.660455102258174, + "learning_rate": 9.894980111847604e-06, + "loss": 0.3804, + "step": 4243 + }, + { + "epoch": 0.09341844596081884, + "grad_norm": 0.7196613393445253, + "learning_rate": 9.894907425314533e-06, + "loss": 0.562, + "step": 4244 + }, + { + "epoch": 0.0934404578472375, + "grad_norm": 2.6564357109307353, + "learning_rate": 9.894834713903358e-06, + "loss": 0.2862, + "step": 4245 + }, + { + "epoch": 0.09346246973365617, + "grad_norm": 2.785671255553793, + "learning_rate": 9.89476197761445e-06, + "loss": 0.3857, + "step": 4246 + }, + { + "epoch": 0.09348448162007485, + "grad_norm": 2.06476279074637, + "learning_rate": 9.89468921644818e-06, + "loss": 0.4264, + "step": 4247 + }, + { + "epoch": 0.09350649350649351, + "grad_norm": 1.9463244444507448, + "learning_rate": 9.894616430404915e-06, + "loss": 0.4242, + "step": 4248 + }, + { + "epoch": 0.09352850539291217, + "grad_norm": 1.2598887026129122, + "learning_rate": 9.894543619485028e-06, + "loss": 0.4044, + "step": 4249 + }, + { + "epoch": 0.09355051727933084, + "grad_norm": 1.9840105246299666, + "learning_rate": 9.894470783688885e-06, + "loss": 0.4001, + "step": 4250 + }, + { + "epoch": 0.0935725291657495, + "grad_norm": 0.6749703878634161, + "learning_rate": 9.894397923016861e-06, + "loss": 0.5673, + "step": 4251 + }, + { + "epoch": 0.09359454105216818, + "grad_norm": 1.616034484657778, + "learning_rate": 9.894325037469322e-06, + "loss": 0.403, + "step": 4252 + }, + { + "epoch": 0.09361655293858684, + "grad_norm": 1.5692465563866245, + "learning_rate": 9.89425212704664e-06, + "loss": 0.3788, + "step": 4253 + }, + { + "epoch": 0.0936385648250055, + "grad_norm": 0.6644223089938034, + "learning_rate": 9.894179191749188e-06, + "loss": 0.5778, + "step": 4254 + }, + { + "epoch": 0.09366057671142416, + "grad_norm": 11.544242509787885, + "learning_rate": 9.894106231577331e-06, + "loss": 0.4432, + "step": 4255 + }, + { + "epoch": 0.09368258859784284, + "grad_norm": 0.6497444115682264, + "learning_rate": 9.894033246531446e-06, + "loss": 0.5794, + "step": 4256 + }, + { + "epoch": 0.0937046004842615, + "grad_norm": 1.7710302448976571, + "learning_rate": 9.893960236611899e-06, + "loss": 0.3894, + "step": 4257 + }, + { + "epoch": 0.09372661237068017, + "grad_norm": 0.6387389550959246, + "learning_rate": 9.893887201819066e-06, + "loss": 0.5795, + "step": 4258 + }, + { + "epoch": 0.09374862425709883, + "grad_norm": 1.19600344219279, + "learning_rate": 9.893814142153315e-06, + "loss": 0.304, + "step": 4259 + }, + { + "epoch": 0.0937706361435175, + "grad_norm": 1.789832144551236, + "learning_rate": 9.893741057615015e-06, + "loss": 0.4267, + "step": 4260 + }, + { + "epoch": 0.09379264802993617, + "grad_norm": 0.6554558945512857, + "learning_rate": 9.893667948204542e-06, + "loss": 0.5783, + "step": 4261 + }, + { + "epoch": 0.09381465991635483, + "grad_norm": 0.6408352507920827, + "learning_rate": 9.893594813922267e-06, + "loss": 0.5566, + "step": 4262 + }, + { + "epoch": 0.0938366718027735, + "grad_norm": 2.2463697673967546, + "learning_rate": 9.893521654768559e-06, + "loss": 0.3889, + "step": 4263 + }, + { + "epoch": 0.09385868368919216, + "grad_norm": 6.645306671358854, + "learning_rate": 9.893448470743792e-06, + "loss": 0.4176, + "step": 4264 + }, + { + "epoch": 0.09388069557561082, + "grad_norm": 1.4303987508975051, + "learning_rate": 9.893375261848339e-06, + "loss": 0.3157, + "step": 4265 + }, + { + "epoch": 0.0939027074620295, + "grad_norm": 1.3692621994695422, + "learning_rate": 9.893302028082567e-06, + "loss": 0.3529, + "step": 4266 + }, + { + "epoch": 0.09392471934844816, + "grad_norm": 1.5870934527832048, + "learning_rate": 9.893228769446854e-06, + "loss": 0.3394, + "step": 4267 + }, + { + "epoch": 0.09394673123486683, + "grad_norm": 1.7292683069979784, + "learning_rate": 9.893155485941571e-06, + "loss": 0.3354, + "step": 4268 + }, + { + "epoch": 0.09396874312128549, + "grad_norm": 2.0090509903744556, + "learning_rate": 9.893082177567087e-06, + "loss": 0.3851, + "step": 4269 + }, + { + "epoch": 0.09399075500770417, + "grad_norm": 1.6175935173815668, + "learning_rate": 9.893008844323777e-06, + "loss": 0.3936, + "step": 4270 + }, + { + "epoch": 0.09401276689412283, + "grad_norm": 1.7435525538962287, + "learning_rate": 9.892935486212015e-06, + "loss": 0.3498, + "step": 4271 + }, + { + "epoch": 0.09403477878054149, + "grad_norm": 1.6671947716271116, + "learning_rate": 9.892862103232172e-06, + "loss": 0.3413, + "step": 4272 + }, + { + "epoch": 0.09405679066696016, + "grad_norm": 2.5622399614713975, + "learning_rate": 9.892788695384622e-06, + "loss": 0.3982, + "step": 4273 + }, + { + "epoch": 0.09407880255337882, + "grad_norm": 1.5534425753375238, + "learning_rate": 9.892715262669737e-06, + "loss": 0.4798, + "step": 4274 + }, + { + "epoch": 0.0941008144397975, + "grad_norm": 1.7191583673558017, + "learning_rate": 9.89264180508789e-06, + "loss": 0.3386, + "step": 4275 + }, + { + "epoch": 0.09412282632621616, + "grad_norm": 1.9674089725990882, + "learning_rate": 9.892568322639457e-06, + "loss": 0.4141, + "step": 4276 + }, + { + "epoch": 0.09414483821263482, + "grad_norm": 2.0071982294868445, + "learning_rate": 9.892494815324808e-06, + "loss": 0.3421, + "step": 4277 + }, + { + "epoch": 0.09416685009905348, + "grad_norm": 0.7200295821065323, + "learning_rate": 9.89242128314432e-06, + "loss": 0.5684, + "step": 4278 + }, + { + "epoch": 0.09418886198547216, + "grad_norm": 1.836168078466519, + "learning_rate": 9.892347726098364e-06, + "loss": 0.257, + "step": 4279 + }, + { + "epoch": 0.09421087387189082, + "grad_norm": 1.6196277047959318, + "learning_rate": 9.892274144187314e-06, + "loss": 0.3558, + "step": 4280 + }, + { + "epoch": 0.09423288575830949, + "grad_norm": 1.435021065411634, + "learning_rate": 9.892200537411548e-06, + "loss": 0.3562, + "step": 4281 + }, + { + "epoch": 0.09425489764472815, + "grad_norm": 1.4774925465211408, + "learning_rate": 9.892126905771434e-06, + "loss": 0.335, + "step": 4282 + }, + { + "epoch": 0.09427690953114681, + "grad_norm": 1.4876174488573857, + "learning_rate": 9.892053249267349e-06, + "loss": 0.2942, + "step": 4283 + }, + { + "epoch": 0.09429892141756549, + "grad_norm": 1.763249869948978, + "learning_rate": 9.891979567899667e-06, + "loss": 0.4357, + "step": 4284 + }, + { + "epoch": 0.09432093330398415, + "grad_norm": 1.8374304743411687, + "learning_rate": 9.891905861668765e-06, + "loss": 0.368, + "step": 4285 + }, + { + "epoch": 0.09434294519040282, + "grad_norm": 2.041930489830845, + "learning_rate": 9.891832130575014e-06, + "loss": 0.4223, + "step": 4286 + }, + { + "epoch": 0.09436495707682148, + "grad_norm": 1.5922181130763182, + "learning_rate": 9.891758374618792e-06, + "loss": 0.3251, + "step": 4287 + }, + { + "epoch": 0.09438696896324016, + "grad_norm": 1.5202057142193885, + "learning_rate": 9.89168459380047e-06, + "loss": 0.2615, + "step": 4288 + }, + { + "epoch": 0.09440898084965882, + "grad_norm": 1.993914903758684, + "learning_rate": 9.891610788120428e-06, + "loss": 0.3557, + "step": 4289 + }, + { + "epoch": 0.09443099273607748, + "grad_norm": 0.6686856070185486, + "learning_rate": 9.891536957579036e-06, + "loss": 0.5709, + "step": 4290 + }, + { + "epoch": 0.09445300462249615, + "grad_norm": 1.44722000389738, + "learning_rate": 9.89146310217667e-06, + "loss": 0.4154, + "step": 4291 + }, + { + "epoch": 0.09447501650891481, + "grad_norm": 1.9403852412600036, + "learning_rate": 9.89138922191371e-06, + "loss": 0.3243, + "step": 4292 + }, + { + "epoch": 0.09449702839533348, + "grad_norm": 1.4275742570920247, + "learning_rate": 9.891315316790526e-06, + "loss": 0.4096, + "step": 4293 + }, + { + "epoch": 0.09451904028175215, + "grad_norm": 2.145045970217373, + "learning_rate": 9.891241386807497e-06, + "loss": 0.3222, + "step": 4294 + }, + { + "epoch": 0.09454105216817081, + "grad_norm": 2.169314957803069, + "learning_rate": 9.891167431964998e-06, + "loss": 0.3673, + "step": 4295 + }, + { + "epoch": 0.09456306405458947, + "grad_norm": 1.401189887495612, + "learning_rate": 9.891093452263405e-06, + "loss": 0.318, + "step": 4296 + }, + { + "epoch": 0.09458507594100815, + "grad_norm": 1.4759184000218306, + "learning_rate": 9.891019447703092e-06, + "loss": 0.2869, + "step": 4297 + }, + { + "epoch": 0.09460708782742681, + "grad_norm": 1.3674044545494612, + "learning_rate": 9.89094541828444e-06, + "loss": 0.29, + "step": 4298 + }, + { + "epoch": 0.09462909971384548, + "grad_norm": 1.4394667149834992, + "learning_rate": 9.89087136400782e-06, + "loss": 0.3707, + "step": 4299 + }, + { + "epoch": 0.09465111160026414, + "grad_norm": 1.776938618369293, + "learning_rate": 9.890797284873608e-06, + "loss": 0.3104, + "step": 4300 + }, + { + "epoch": 0.0946731234866828, + "grad_norm": 1.5100795506835598, + "learning_rate": 9.890723180882185e-06, + "loss": 0.3618, + "step": 4301 + }, + { + "epoch": 0.09469513537310148, + "grad_norm": 0.7517238892600195, + "learning_rate": 9.890649052033926e-06, + "loss": 0.5818, + "step": 4302 + }, + { + "epoch": 0.09471714725952014, + "grad_norm": 0.6997459098820789, + "learning_rate": 9.890574898329205e-06, + "loss": 0.5712, + "step": 4303 + }, + { + "epoch": 0.0947391591459388, + "grad_norm": 1.578467626611199, + "learning_rate": 9.890500719768403e-06, + "loss": 0.3394, + "step": 4304 + }, + { + "epoch": 0.09476117103235747, + "grad_norm": 0.6673683098379509, + "learning_rate": 9.890426516351897e-06, + "loss": 0.5935, + "step": 4305 + }, + { + "epoch": 0.09478318291877615, + "grad_norm": 1.490053545769976, + "learning_rate": 9.890352288080059e-06, + "loss": 0.3784, + "step": 4306 + }, + { + "epoch": 0.09480519480519481, + "grad_norm": 1.8096148129627752, + "learning_rate": 9.890278034953271e-06, + "loss": 0.3537, + "step": 4307 + }, + { + "epoch": 0.09482720669161347, + "grad_norm": 1.4488439296747218, + "learning_rate": 9.890203756971907e-06, + "loss": 0.385, + "step": 4308 + }, + { + "epoch": 0.09484921857803214, + "grad_norm": 1.44032537094444, + "learning_rate": 9.890129454136348e-06, + "loss": 0.3774, + "step": 4309 + }, + { + "epoch": 0.0948712304644508, + "grad_norm": 0.8040997367755841, + "learning_rate": 9.890055126446971e-06, + "loss": 0.5794, + "step": 4310 + }, + { + "epoch": 0.09489324235086948, + "grad_norm": 1.526206284685184, + "learning_rate": 9.889980773904152e-06, + "loss": 0.3523, + "step": 4311 + }, + { + "epoch": 0.09491525423728814, + "grad_norm": 1.466176853731094, + "learning_rate": 9.889906396508269e-06, + "loss": 0.3237, + "step": 4312 + }, + { + "epoch": 0.0949372661237068, + "grad_norm": 1.4288123875069292, + "learning_rate": 9.8898319942597e-06, + "loss": 0.4037, + "step": 4313 + }, + { + "epoch": 0.09495927801012546, + "grad_norm": 1.4613825690082987, + "learning_rate": 9.889757567158827e-06, + "loss": 0.4101, + "step": 4314 + }, + { + "epoch": 0.09498128989654413, + "grad_norm": 4.300786358175422, + "learning_rate": 9.889683115206023e-06, + "loss": 0.3934, + "step": 4315 + }, + { + "epoch": 0.0950033017829628, + "grad_norm": 0.6740387215145096, + "learning_rate": 9.88960863840167e-06, + "loss": 0.5535, + "step": 4316 + }, + { + "epoch": 0.09502531366938147, + "grad_norm": 1.4573702714644667, + "learning_rate": 9.889534136746141e-06, + "loss": 0.3101, + "step": 4317 + }, + { + "epoch": 0.09504732555580013, + "grad_norm": 1.3806788804424417, + "learning_rate": 9.889459610239822e-06, + "loss": 0.2883, + "step": 4318 + }, + { + "epoch": 0.0950693374422188, + "grad_norm": 0.6457733603544444, + "learning_rate": 9.889385058883089e-06, + "loss": 0.5949, + "step": 4319 + }, + { + "epoch": 0.09509134932863747, + "grad_norm": 1.7097962118803358, + "learning_rate": 9.88931048267632e-06, + "loss": 0.3756, + "step": 4320 + }, + { + "epoch": 0.09511336121505613, + "grad_norm": 1.3909231136252314, + "learning_rate": 9.889235881619894e-06, + "loss": 0.352, + "step": 4321 + }, + { + "epoch": 0.0951353731014748, + "grad_norm": 1.5858969430715122, + "learning_rate": 9.88916125571419e-06, + "loss": 0.433, + "step": 4322 + }, + { + "epoch": 0.09515738498789346, + "grad_norm": 1.6904806309622877, + "learning_rate": 9.889086604959588e-06, + "loss": 0.4407, + "step": 4323 + }, + { + "epoch": 0.09517939687431212, + "grad_norm": 1.565622075206355, + "learning_rate": 9.889011929356467e-06, + "loss": 0.4511, + "step": 4324 + }, + { + "epoch": 0.0952014087607308, + "grad_norm": 1.6300272935051552, + "learning_rate": 9.888937228905208e-06, + "loss": 0.4629, + "step": 4325 + }, + { + "epoch": 0.09522342064714946, + "grad_norm": 1.4847978368328496, + "learning_rate": 9.888862503606189e-06, + "loss": 0.378, + "step": 4326 + }, + { + "epoch": 0.09524543253356813, + "grad_norm": 1.4543439094585962, + "learning_rate": 9.88878775345979e-06, + "loss": 0.3386, + "step": 4327 + }, + { + "epoch": 0.09526744441998679, + "grad_norm": 0.6951518746560709, + "learning_rate": 9.888712978466391e-06, + "loss": 0.5736, + "step": 4328 + }, + { + "epoch": 0.09528945630640547, + "grad_norm": 1.3088380368894532, + "learning_rate": 9.888638178626373e-06, + "loss": 0.3287, + "step": 4329 + }, + { + "epoch": 0.09531146819282413, + "grad_norm": 0.6619852279314625, + "learning_rate": 9.888563353940114e-06, + "loss": 0.56, + "step": 4330 + }, + { + "epoch": 0.09533348007924279, + "grad_norm": 1.5762168913098962, + "learning_rate": 9.888488504407999e-06, + "loss": 0.4351, + "step": 4331 + }, + { + "epoch": 0.09535549196566145, + "grad_norm": 1.399655631985289, + "learning_rate": 9.888413630030403e-06, + "loss": 0.3996, + "step": 4332 + }, + { + "epoch": 0.09537750385208012, + "grad_norm": 1.8535565377947576, + "learning_rate": 9.888338730807706e-06, + "loss": 0.3495, + "step": 4333 + }, + { + "epoch": 0.0953995157384988, + "grad_norm": 1.343022867139691, + "learning_rate": 9.888263806740295e-06, + "loss": 0.3129, + "step": 4334 + }, + { + "epoch": 0.09542152762491746, + "grad_norm": 1.4335891336669235, + "learning_rate": 9.888188857828545e-06, + "loss": 0.3868, + "step": 4335 + }, + { + "epoch": 0.09544353951133612, + "grad_norm": 1.4669667373516093, + "learning_rate": 9.88811388407284e-06, + "loss": 0.4391, + "step": 4336 + }, + { + "epoch": 0.09546555139775478, + "grad_norm": 1.6417422251914473, + "learning_rate": 9.88803888547356e-06, + "loss": 0.3995, + "step": 4337 + }, + { + "epoch": 0.09548756328417346, + "grad_norm": 2.4070552330289354, + "learning_rate": 9.887963862031087e-06, + "loss": 0.3711, + "step": 4338 + }, + { + "epoch": 0.09550957517059212, + "grad_norm": 1.6424098619845457, + "learning_rate": 9.8878888137458e-06, + "loss": 0.3753, + "step": 4339 + }, + { + "epoch": 0.09553158705701079, + "grad_norm": 1.3944412899291592, + "learning_rate": 9.887813740618083e-06, + "loss": 0.3115, + "step": 4340 + }, + { + "epoch": 0.09555359894342945, + "grad_norm": 2.064435678757254, + "learning_rate": 9.887738642648316e-06, + "loss": 0.3544, + "step": 4341 + }, + { + "epoch": 0.09557561082984811, + "grad_norm": 3.928505227492317, + "learning_rate": 9.88766351983688e-06, + "loss": 0.4003, + "step": 4342 + }, + { + "epoch": 0.09559762271626679, + "grad_norm": 1.949613992289495, + "learning_rate": 9.88758837218416e-06, + "loss": 0.4411, + "step": 4343 + }, + { + "epoch": 0.09561963460268545, + "grad_norm": 46.231588869939024, + "learning_rate": 9.887513199690537e-06, + "loss": 0.3789, + "step": 4344 + }, + { + "epoch": 0.09564164648910412, + "grad_norm": 1.5428986339793516, + "learning_rate": 9.887438002356392e-06, + "loss": 0.3748, + "step": 4345 + }, + { + "epoch": 0.09566365837552278, + "grad_norm": 1.635036941609803, + "learning_rate": 9.887362780182104e-06, + "loss": 0.4319, + "step": 4346 + }, + { + "epoch": 0.09568567026194146, + "grad_norm": 1.7492340981257408, + "learning_rate": 9.88728753316806e-06, + "loss": 0.3601, + "step": 4347 + }, + { + "epoch": 0.09570768214836012, + "grad_norm": 1.5857622410721546, + "learning_rate": 9.887212261314642e-06, + "loss": 0.3836, + "step": 4348 + }, + { + "epoch": 0.09572969403477878, + "grad_norm": 1.3948776133496434, + "learning_rate": 9.88713696462223e-06, + "loss": 0.3544, + "step": 4349 + }, + { + "epoch": 0.09575170592119744, + "grad_norm": 1.5526339107762483, + "learning_rate": 9.887061643091209e-06, + "loss": 0.4593, + "step": 4350 + }, + { + "epoch": 0.09577371780761611, + "grad_norm": 1.7997047319469175, + "learning_rate": 9.88698629672196e-06, + "loss": 0.3664, + "step": 4351 + }, + { + "epoch": 0.09579572969403478, + "grad_norm": 1.3061532421313142, + "learning_rate": 9.88691092551487e-06, + "loss": 0.4216, + "step": 4352 + }, + { + "epoch": 0.09581774158045345, + "grad_norm": 1.1708564951515048, + "learning_rate": 9.886835529470316e-06, + "loss": 0.3014, + "step": 4353 + }, + { + "epoch": 0.09583975346687211, + "grad_norm": 1.8007286238949718, + "learning_rate": 9.886760108588686e-06, + "loss": 0.3205, + "step": 4354 + }, + { + "epoch": 0.09586176535329077, + "grad_norm": 1.5186383061120008, + "learning_rate": 9.88668466287036e-06, + "loss": 0.2812, + "step": 4355 + }, + { + "epoch": 0.09588377723970944, + "grad_norm": 1.5981298764117589, + "learning_rate": 9.886609192315723e-06, + "loss": 0.2975, + "step": 4356 + }, + { + "epoch": 0.09590578912612811, + "grad_norm": 0.9367078991155818, + "learning_rate": 9.88653369692516e-06, + "loss": 0.5803, + "step": 4357 + }, + { + "epoch": 0.09592780101254678, + "grad_norm": 0.8581600043220934, + "learning_rate": 9.886458176699055e-06, + "loss": 0.6084, + "step": 4358 + }, + { + "epoch": 0.09594981289896544, + "grad_norm": 0.6760645817902865, + "learning_rate": 9.886382631637786e-06, + "loss": 0.5603, + "step": 4359 + }, + { + "epoch": 0.0959718247853841, + "grad_norm": 1.7757852194267598, + "learning_rate": 9.886307061741744e-06, + "loss": 0.3052, + "step": 4360 + }, + { + "epoch": 0.09599383667180278, + "grad_norm": 1.3562016584392338, + "learning_rate": 9.886231467011309e-06, + "loss": 0.3675, + "step": 4361 + }, + { + "epoch": 0.09601584855822144, + "grad_norm": 2.036326156956959, + "learning_rate": 9.886155847446866e-06, + "loss": 0.3737, + "step": 4362 + }, + { + "epoch": 0.0960378604446401, + "grad_norm": 1.3408496168654445, + "learning_rate": 9.8860802030488e-06, + "loss": 0.3127, + "step": 4363 + }, + { + "epoch": 0.09605987233105877, + "grad_norm": 1.6199687030760748, + "learning_rate": 9.886004533817496e-06, + "loss": 0.3858, + "step": 4364 + }, + { + "epoch": 0.09608188421747743, + "grad_norm": 1.6452564720688154, + "learning_rate": 9.885928839753337e-06, + "loss": 0.3437, + "step": 4365 + }, + { + "epoch": 0.09610389610389611, + "grad_norm": 1.4702450343620808, + "learning_rate": 9.885853120856708e-06, + "loss": 0.378, + "step": 4366 + }, + { + "epoch": 0.09612590799031477, + "grad_norm": 1.436019429404039, + "learning_rate": 9.885777377127996e-06, + "loss": 0.6262, + "step": 4367 + }, + { + "epoch": 0.09614791987673343, + "grad_norm": 1.5354549309081944, + "learning_rate": 9.885701608567582e-06, + "loss": 0.3238, + "step": 4368 + }, + { + "epoch": 0.0961699317631521, + "grad_norm": 1.6178708458721156, + "learning_rate": 9.885625815175854e-06, + "loss": 0.33, + "step": 4369 + }, + { + "epoch": 0.09619194364957077, + "grad_norm": 1.6275933793867419, + "learning_rate": 9.8855499969532e-06, + "loss": 0.3244, + "step": 4370 + }, + { + "epoch": 0.09621395553598944, + "grad_norm": 1.8825869488987277, + "learning_rate": 9.885474153899997e-06, + "loss": 0.4243, + "step": 4371 + }, + { + "epoch": 0.0962359674224081, + "grad_norm": 1.5614339378980713, + "learning_rate": 9.885398286016636e-06, + "loss": 0.3325, + "step": 4372 + }, + { + "epoch": 0.09625797930882676, + "grad_norm": 1.8162253276321838, + "learning_rate": 9.885322393303505e-06, + "loss": 0.3467, + "step": 4373 + }, + { + "epoch": 0.09627999119524543, + "grad_norm": 1.5889794295643376, + "learning_rate": 9.885246475760984e-06, + "loss": 0.4649, + "step": 4374 + }, + { + "epoch": 0.0963020030816641, + "grad_norm": 1.2527826244929225, + "learning_rate": 9.885170533389464e-06, + "loss": 0.325, + "step": 4375 + }, + { + "epoch": 0.09632401496808277, + "grad_norm": 1.2766677540546412, + "learning_rate": 9.885094566189326e-06, + "loss": 0.3322, + "step": 4376 + }, + { + "epoch": 0.09634602685450143, + "grad_norm": 1.5334378189015923, + "learning_rate": 9.88501857416096e-06, + "loss": 0.3619, + "step": 4377 + }, + { + "epoch": 0.09636803874092009, + "grad_norm": 1.6500366888738514, + "learning_rate": 9.884942557304749e-06, + "loss": 0.3764, + "step": 4378 + }, + { + "epoch": 0.09639005062733877, + "grad_norm": 1.5035812677463534, + "learning_rate": 9.884866515621082e-06, + "loss": 0.3364, + "step": 4379 + }, + { + "epoch": 0.09641206251375743, + "grad_norm": 1.6350310129843624, + "learning_rate": 9.884790449110347e-06, + "loss": 0.4363, + "step": 4380 + }, + { + "epoch": 0.0964340744001761, + "grad_norm": 1.6379950298045034, + "learning_rate": 9.884714357772926e-06, + "loss": 0.3353, + "step": 4381 + }, + { + "epoch": 0.09645608628659476, + "grad_norm": 1.9119362976550276, + "learning_rate": 9.884638241609208e-06, + "loss": 0.386, + "step": 4382 + }, + { + "epoch": 0.09647809817301342, + "grad_norm": 1.9190507043231624, + "learning_rate": 9.884562100619582e-06, + "loss": 0.37, + "step": 4383 + }, + { + "epoch": 0.0965001100594321, + "grad_norm": 1.5187663537317784, + "learning_rate": 9.88448593480443e-06, + "loss": 0.3259, + "step": 4384 + }, + { + "epoch": 0.09652212194585076, + "grad_norm": 1.3742794907940812, + "learning_rate": 9.884409744164143e-06, + "loss": 0.2726, + "step": 4385 + }, + { + "epoch": 0.09654413383226942, + "grad_norm": 1.3569033872226588, + "learning_rate": 9.884333528699107e-06, + "loss": 0.3553, + "step": 4386 + }, + { + "epoch": 0.09656614571868809, + "grad_norm": 1.7270304182915004, + "learning_rate": 9.88425728840971e-06, + "loss": 0.3566, + "step": 4387 + }, + { + "epoch": 0.09658815760510676, + "grad_norm": 3.2349157909541497, + "learning_rate": 9.88418102329634e-06, + "loss": 0.3913, + "step": 4388 + }, + { + "epoch": 0.09661016949152543, + "grad_norm": 0.7858520923683604, + "learning_rate": 9.884104733359381e-06, + "loss": 0.5623, + "step": 4389 + }, + { + "epoch": 0.09663218137794409, + "grad_norm": 0.7268465774932734, + "learning_rate": 9.884028418599226e-06, + "loss": 0.5675, + "step": 4390 + }, + { + "epoch": 0.09665419326436275, + "grad_norm": 1.8715038191047473, + "learning_rate": 9.883952079016257e-06, + "loss": 0.4732, + "step": 4391 + }, + { + "epoch": 0.09667620515078142, + "grad_norm": 1.3326673539157852, + "learning_rate": 9.88387571461087e-06, + "loss": 0.3778, + "step": 4392 + }, + { + "epoch": 0.0966982170372001, + "grad_norm": 6.03509029204385, + "learning_rate": 9.883799325383443e-06, + "loss": 0.4063, + "step": 4393 + }, + { + "epoch": 0.09672022892361876, + "grad_norm": 1.3621962185956875, + "learning_rate": 9.883722911334373e-06, + "loss": 0.3168, + "step": 4394 + }, + { + "epoch": 0.09674224081003742, + "grad_norm": 1.8005675207824279, + "learning_rate": 9.883646472464043e-06, + "loss": 0.3569, + "step": 4395 + }, + { + "epoch": 0.09676425269645608, + "grad_norm": 1.7373041844278754, + "learning_rate": 9.883570008772847e-06, + "loss": 0.2833, + "step": 4396 + }, + { + "epoch": 0.09678626458287476, + "grad_norm": 0.8381254790876955, + "learning_rate": 9.883493520261167e-06, + "loss": 0.5816, + "step": 4397 + }, + { + "epoch": 0.09680827646929342, + "grad_norm": 1.5075720541129924, + "learning_rate": 9.883417006929394e-06, + "loss": 0.3207, + "step": 4398 + }, + { + "epoch": 0.09683028835571209, + "grad_norm": 1.523889539484221, + "learning_rate": 9.88334046877792e-06, + "loss": 0.3587, + "step": 4399 + }, + { + "epoch": 0.09685230024213075, + "grad_norm": 1.5003137585250257, + "learning_rate": 9.883263905807131e-06, + "loss": 0.3852, + "step": 4400 + }, + { + "epoch": 0.09687431212854941, + "grad_norm": 1.3653858064820776, + "learning_rate": 9.883187318017417e-06, + "loss": 0.43, + "step": 4401 + }, + { + "epoch": 0.09689632401496809, + "grad_norm": 1.703278586302595, + "learning_rate": 9.883110705409167e-06, + "loss": 0.4865, + "step": 4402 + }, + { + "epoch": 0.09691833590138675, + "grad_norm": 1.6721032766755435, + "learning_rate": 9.883034067982771e-06, + "loss": 0.412, + "step": 4403 + }, + { + "epoch": 0.09694034778780541, + "grad_norm": 1.8586464809049879, + "learning_rate": 9.882957405738617e-06, + "loss": 0.3995, + "step": 4404 + }, + { + "epoch": 0.09696235967422408, + "grad_norm": 0.748961231594859, + "learning_rate": 9.882880718677097e-06, + "loss": 0.5733, + "step": 4405 + }, + { + "epoch": 0.09698437156064274, + "grad_norm": 2.0399721657081034, + "learning_rate": 9.882804006798597e-06, + "loss": 0.441, + "step": 4406 + }, + { + "epoch": 0.09700638344706142, + "grad_norm": 1.641351199914096, + "learning_rate": 9.882727270103511e-06, + "loss": 0.3809, + "step": 4407 + }, + { + "epoch": 0.09702839533348008, + "grad_norm": 1.7637747843182399, + "learning_rate": 9.882650508592227e-06, + "loss": 0.354, + "step": 4408 + }, + { + "epoch": 0.09705040721989874, + "grad_norm": 1.5938771413887844, + "learning_rate": 9.882573722265133e-06, + "loss": 0.4112, + "step": 4409 + }, + { + "epoch": 0.0970724191063174, + "grad_norm": 1.3866030553879416, + "learning_rate": 9.882496911122626e-06, + "loss": 0.3319, + "step": 4410 + }, + { + "epoch": 0.09709443099273608, + "grad_norm": 1.5978863906437466, + "learning_rate": 9.882420075165089e-06, + "loss": 0.2969, + "step": 4411 + }, + { + "epoch": 0.09711644287915475, + "grad_norm": 0.6778745301704573, + "learning_rate": 9.882343214392917e-06, + "loss": 0.5711, + "step": 4412 + }, + { + "epoch": 0.09713845476557341, + "grad_norm": 1.736058562797053, + "learning_rate": 9.882266328806498e-06, + "loss": 0.3769, + "step": 4413 + }, + { + "epoch": 0.09716046665199207, + "grad_norm": 1.4600270554845909, + "learning_rate": 9.882189418406226e-06, + "loss": 0.3631, + "step": 4414 + }, + { + "epoch": 0.09718247853841074, + "grad_norm": 1.851312218125637, + "learning_rate": 9.882112483192487e-06, + "loss": 0.3479, + "step": 4415 + }, + { + "epoch": 0.09720449042482941, + "grad_norm": 1.954782202038178, + "learning_rate": 9.882035523165677e-06, + "loss": 0.3502, + "step": 4416 + }, + { + "epoch": 0.09722650231124808, + "grad_norm": 0.6528665432504944, + "learning_rate": 9.881958538326184e-06, + "loss": 0.5751, + "step": 4417 + }, + { + "epoch": 0.09724851419766674, + "grad_norm": 0.623226988613942, + "learning_rate": 9.8818815286744e-06, + "loss": 0.548, + "step": 4418 + }, + { + "epoch": 0.0972705260840854, + "grad_norm": 1.4862027835839338, + "learning_rate": 9.881804494210718e-06, + "loss": 0.3432, + "step": 4419 + }, + { + "epoch": 0.09729253797050408, + "grad_norm": 1.5822507939256287, + "learning_rate": 9.88172743493553e-06, + "loss": 0.2864, + "step": 4420 + }, + { + "epoch": 0.09731454985692274, + "grad_norm": 0.6375018572388973, + "learning_rate": 9.881650350849221e-06, + "loss": 0.5566, + "step": 4421 + }, + { + "epoch": 0.0973365617433414, + "grad_norm": 1.6720439632165671, + "learning_rate": 9.88157324195219e-06, + "loss": 0.3747, + "step": 4422 + }, + { + "epoch": 0.09735857362976007, + "grad_norm": 1.5858349866243102, + "learning_rate": 9.881496108244826e-06, + "loss": 0.3491, + "step": 4423 + }, + { + "epoch": 0.09738058551617873, + "grad_norm": 1.5753587193552034, + "learning_rate": 9.881418949727524e-06, + "loss": 0.3855, + "step": 4424 + }, + { + "epoch": 0.09740259740259741, + "grad_norm": 1.4513049909058309, + "learning_rate": 9.881341766400671e-06, + "loss": 0.3852, + "step": 4425 + }, + { + "epoch": 0.09742460928901607, + "grad_norm": 0.6672785413299511, + "learning_rate": 9.881264558264663e-06, + "loss": 0.5871, + "step": 4426 + }, + { + "epoch": 0.09744662117543473, + "grad_norm": 0.6714115454567048, + "learning_rate": 9.881187325319892e-06, + "loss": 0.5983, + "step": 4427 + }, + { + "epoch": 0.0974686330618534, + "grad_norm": 1.4597980857262283, + "learning_rate": 9.881110067566748e-06, + "loss": 0.3524, + "step": 4428 + }, + { + "epoch": 0.09749064494827207, + "grad_norm": 0.628316889773677, + "learning_rate": 9.881032785005627e-06, + "loss": 0.5638, + "step": 4429 + }, + { + "epoch": 0.09751265683469074, + "grad_norm": 1.8696162932620783, + "learning_rate": 9.880955477636921e-06, + "loss": 0.3302, + "step": 4430 + }, + { + "epoch": 0.0975346687211094, + "grad_norm": 1.554411675231399, + "learning_rate": 9.880878145461021e-06, + "loss": 0.3523, + "step": 4431 + }, + { + "epoch": 0.09755668060752806, + "grad_norm": 0.6298943961855893, + "learning_rate": 9.880800788478322e-06, + "loss": 0.5748, + "step": 4432 + }, + { + "epoch": 0.09757869249394673, + "grad_norm": 3.220634120618275, + "learning_rate": 9.880723406689217e-06, + "loss": 0.389, + "step": 4433 + }, + { + "epoch": 0.0976007043803654, + "grad_norm": 0.6519464405838223, + "learning_rate": 9.880646000094098e-06, + "loss": 0.5585, + "step": 4434 + }, + { + "epoch": 0.09762271626678407, + "grad_norm": 0.6254712140792877, + "learning_rate": 9.880568568693358e-06, + "loss": 0.519, + "step": 4435 + }, + { + "epoch": 0.09764472815320273, + "grad_norm": 1.6355285505434447, + "learning_rate": 9.880491112487393e-06, + "loss": 0.3515, + "step": 4436 + }, + { + "epoch": 0.09766674003962139, + "grad_norm": 0.6357201550472744, + "learning_rate": 9.880413631476598e-06, + "loss": 0.5596, + "step": 4437 + }, + { + "epoch": 0.09768875192604007, + "grad_norm": 1.9762917643651081, + "learning_rate": 9.88033612566136e-06, + "loss": 0.4029, + "step": 4438 + }, + { + "epoch": 0.09771076381245873, + "grad_norm": 1.4655696642790803, + "learning_rate": 9.88025859504208e-06, + "loss": 0.3735, + "step": 4439 + }, + { + "epoch": 0.0977327756988774, + "grad_norm": 1.4623797020357885, + "learning_rate": 9.880181039619148e-06, + "loss": 0.3553, + "step": 4440 + }, + { + "epoch": 0.09775478758529606, + "grad_norm": 1.427776957526089, + "learning_rate": 9.880103459392958e-06, + "loss": 0.3725, + "step": 4441 + }, + { + "epoch": 0.09777679947171472, + "grad_norm": 1.7030678560658241, + "learning_rate": 9.880025854363908e-06, + "loss": 0.4639, + "step": 4442 + }, + { + "epoch": 0.0977988113581334, + "grad_norm": 1.4959229237095113, + "learning_rate": 9.879948224532388e-06, + "loss": 0.3579, + "step": 4443 + }, + { + "epoch": 0.09782082324455206, + "grad_norm": 1.6506036199476517, + "learning_rate": 9.879870569898796e-06, + "loss": 0.4821, + "step": 4444 + }, + { + "epoch": 0.09784283513097072, + "grad_norm": 0.6869157754749168, + "learning_rate": 9.879792890463524e-06, + "loss": 0.5761, + "step": 4445 + }, + { + "epoch": 0.09786484701738939, + "grad_norm": 1.5444272026299455, + "learning_rate": 9.87971518622697e-06, + "loss": 0.4998, + "step": 4446 + }, + { + "epoch": 0.09788685890380805, + "grad_norm": 2.218626854489033, + "learning_rate": 9.879637457189525e-06, + "loss": 0.4306, + "step": 4447 + }, + { + "epoch": 0.09790887079022673, + "grad_norm": 1.3482331957527363, + "learning_rate": 9.879559703351588e-06, + "loss": 0.3422, + "step": 4448 + }, + { + "epoch": 0.09793088267664539, + "grad_norm": 1.8301287538618733, + "learning_rate": 9.879481924713552e-06, + "loss": 0.4102, + "step": 4449 + }, + { + "epoch": 0.09795289456306405, + "grad_norm": 1.1481685339730667, + "learning_rate": 9.87940412127581e-06, + "loss": 0.2427, + "step": 4450 + }, + { + "epoch": 0.09797490644948272, + "grad_norm": 2.4078517430349486, + "learning_rate": 9.879326293038763e-06, + "loss": 0.447, + "step": 4451 + }, + { + "epoch": 0.09799691833590139, + "grad_norm": 1.7845121807575268, + "learning_rate": 9.879248440002803e-06, + "loss": 0.4223, + "step": 4452 + }, + { + "epoch": 0.09801893022232006, + "grad_norm": 0.6808767972590986, + "learning_rate": 9.879170562168325e-06, + "loss": 0.5687, + "step": 4453 + }, + { + "epoch": 0.09804094210873872, + "grad_norm": 1.765872682837465, + "learning_rate": 9.879092659535728e-06, + "loss": 0.3871, + "step": 4454 + }, + { + "epoch": 0.09806295399515738, + "grad_norm": 1.5335365482092815, + "learning_rate": 9.879014732105403e-06, + "loss": 0.3295, + "step": 4455 + }, + { + "epoch": 0.09808496588157604, + "grad_norm": 1.265847071237765, + "learning_rate": 9.87893677987775e-06, + "loss": 0.3244, + "step": 4456 + }, + { + "epoch": 0.09810697776799472, + "grad_norm": 1.6520434178350556, + "learning_rate": 9.878858802853167e-06, + "loss": 0.3647, + "step": 4457 + }, + { + "epoch": 0.09812898965441338, + "grad_norm": 1.3615442867117638, + "learning_rate": 9.878780801032044e-06, + "loss": 0.2845, + "step": 4458 + }, + { + "epoch": 0.09815100154083205, + "grad_norm": 2.366488111629593, + "learning_rate": 9.878702774414783e-06, + "loss": 0.3723, + "step": 4459 + }, + { + "epoch": 0.09817301342725071, + "grad_norm": 1.541710537397436, + "learning_rate": 9.878624723001777e-06, + "loss": 0.4549, + "step": 4460 + }, + { + "epoch": 0.09819502531366939, + "grad_norm": 1.8063538442937856, + "learning_rate": 9.878546646793425e-06, + "loss": 0.379, + "step": 4461 + }, + { + "epoch": 0.09821703720008805, + "grad_norm": 1.8412609814161573, + "learning_rate": 9.878468545790123e-06, + "loss": 0.4834, + "step": 4462 + }, + { + "epoch": 0.09823904908650671, + "grad_norm": 1.3954197397116046, + "learning_rate": 9.878390419992269e-06, + "loss": 0.3667, + "step": 4463 + }, + { + "epoch": 0.09826106097292538, + "grad_norm": 1.5534198764106766, + "learning_rate": 9.878312269400258e-06, + "loss": 0.4364, + "step": 4464 + }, + { + "epoch": 0.09828307285934404, + "grad_norm": 1.9906438730840992, + "learning_rate": 9.878234094014488e-06, + "loss": 0.2991, + "step": 4465 + }, + { + "epoch": 0.09830508474576272, + "grad_norm": 1.3006862275603535, + "learning_rate": 9.878155893835358e-06, + "loss": 0.2669, + "step": 4466 + }, + { + "epoch": 0.09832709663218138, + "grad_norm": 1.4783249824307356, + "learning_rate": 9.878077668863261e-06, + "loss": 0.3647, + "step": 4467 + }, + { + "epoch": 0.09834910851860004, + "grad_norm": 1.640130717239968, + "learning_rate": 9.8779994190986e-06, + "loss": 0.4331, + "step": 4468 + }, + { + "epoch": 0.0983711204050187, + "grad_norm": 2.325151388100391, + "learning_rate": 9.877921144541771e-06, + "loss": 0.363, + "step": 4469 + }, + { + "epoch": 0.09839313229143738, + "grad_norm": 0.6688463871010891, + "learning_rate": 9.87784284519317e-06, + "loss": 0.5613, + "step": 4470 + }, + { + "epoch": 0.09841514417785605, + "grad_norm": 2.400000915863412, + "learning_rate": 9.877764521053197e-06, + "loss": 0.3232, + "step": 4471 + }, + { + "epoch": 0.09843715606427471, + "grad_norm": 1.6279017678284342, + "learning_rate": 9.877686172122247e-06, + "loss": 0.3491, + "step": 4472 + }, + { + "epoch": 0.09845916795069337, + "grad_norm": 1.3323271493807753, + "learning_rate": 9.877607798400723e-06, + "loss": 0.35, + "step": 4473 + }, + { + "epoch": 0.09848117983711203, + "grad_norm": 1.4392565661542516, + "learning_rate": 9.877529399889019e-06, + "loss": 0.2837, + "step": 4474 + }, + { + "epoch": 0.09850319172353071, + "grad_norm": 1.6776042738294181, + "learning_rate": 9.877450976587534e-06, + "loss": 0.3942, + "step": 4475 + }, + { + "epoch": 0.09852520360994937, + "grad_norm": 1.4944021564112626, + "learning_rate": 9.877372528496671e-06, + "loss": 0.3439, + "step": 4476 + }, + { + "epoch": 0.09854721549636804, + "grad_norm": 2.1381166866563284, + "learning_rate": 9.877294055616823e-06, + "loss": 0.4079, + "step": 4477 + }, + { + "epoch": 0.0985692273827867, + "grad_norm": 2.0288760101561945, + "learning_rate": 9.877215557948392e-06, + "loss": 0.3497, + "step": 4478 + }, + { + "epoch": 0.09859123926920538, + "grad_norm": 1.6407059261253787, + "learning_rate": 9.877137035491777e-06, + "loss": 0.4058, + "step": 4479 + }, + { + "epoch": 0.09861325115562404, + "grad_norm": 1.5593190918784818, + "learning_rate": 9.877058488247376e-06, + "loss": 0.3724, + "step": 4480 + }, + { + "epoch": 0.0986352630420427, + "grad_norm": 0.6988536625971329, + "learning_rate": 9.876979916215588e-06, + "loss": 0.56, + "step": 4481 + }, + { + "epoch": 0.09865727492846137, + "grad_norm": 2.0062261171562934, + "learning_rate": 9.876901319396812e-06, + "loss": 0.367, + "step": 4482 + }, + { + "epoch": 0.09867928681488003, + "grad_norm": 0.6444066976579779, + "learning_rate": 9.876822697791451e-06, + "loss": 0.5569, + "step": 4483 + }, + { + "epoch": 0.0987012987012987, + "grad_norm": 1.2399956047223735, + "learning_rate": 9.876744051399899e-06, + "loss": 0.3184, + "step": 4484 + }, + { + "epoch": 0.09872331058771737, + "grad_norm": 1.3872634396839112, + "learning_rate": 9.87666538022256e-06, + "loss": 0.3863, + "step": 4485 + }, + { + "epoch": 0.09874532247413603, + "grad_norm": 2.108659678109013, + "learning_rate": 9.876586684259832e-06, + "loss": 0.4002, + "step": 4486 + }, + { + "epoch": 0.0987673343605547, + "grad_norm": 2.8963636400631607, + "learning_rate": 9.876507963512116e-06, + "loss": 0.3105, + "step": 4487 + }, + { + "epoch": 0.09878934624697337, + "grad_norm": 1.6920949275707509, + "learning_rate": 9.876429217979811e-06, + "loss": 0.3266, + "step": 4488 + }, + { + "epoch": 0.09881135813339204, + "grad_norm": 1.6713126148043451, + "learning_rate": 9.876350447663319e-06, + "loss": 0.3356, + "step": 4489 + }, + { + "epoch": 0.0988333700198107, + "grad_norm": 0.6887585455069276, + "learning_rate": 9.876271652563039e-06, + "loss": 0.5696, + "step": 4490 + }, + { + "epoch": 0.09885538190622936, + "grad_norm": 0.6493467416650299, + "learning_rate": 9.87619283267937e-06, + "loss": 0.5646, + "step": 4491 + }, + { + "epoch": 0.09887739379264802, + "grad_norm": 2.3418472307960077, + "learning_rate": 9.876113988012714e-06, + "loss": 0.3407, + "step": 4492 + }, + { + "epoch": 0.0988994056790667, + "grad_norm": 0.662957951129983, + "learning_rate": 9.876035118563475e-06, + "loss": 0.5948, + "step": 4493 + }, + { + "epoch": 0.09892141756548536, + "grad_norm": 1.743759951948791, + "learning_rate": 9.875956224332047e-06, + "loss": 0.3813, + "step": 4494 + }, + { + "epoch": 0.09894342945190403, + "grad_norm": 1.458414076820986, + "learning_rate": 9.875877305318837e-06, + "loss": 0.3757, + "step": 4495 + }, + { + "epoch": 0.09896544133832269, + "grad_norm": 1.2930392247434925, + "learning_rate": 9.875798361524242e-06, + "loss": 0.2953, + "step": 4496 + }, + { + "epoch": 0.09898745322474135, + "grad_norm": 0.7191981938068961, + "learning_rate": 9.875719392948667e-06, + "loss": 0.5862, + "step": 4497 + }, + { + "epoch": 0.09900946511116003, + "grad_norm": 1.6768884969000224, + "learning_rate": 9.875640399592511e-06, + "loss": 0.3823, + "step": 4498 + }, + { + "epoch": 0.0990314769975787, + "grad_norm": 1.7485724382475254, + "learning_rate": 9.875561381456176e-06, + "loss": 0.3859, + "step": 4499 + }, + { + "epoch": 0.09905348888399736, + "grad_norm": 1.4359619358405988, + "learning_rate": 9.875482338540063e-06, + "loss": 0.3463, + "step": 4500 + }, + { + "epoch": 0.09907550077041602, + "grad_norm": 1.5372475232754406, + "learning_rate": 9.875403270844575e-06, + "loss": 0.4326, + "step": 4501 + }, + { + "epoch": 0.0990975126568347, + "grad_norm": 1.9419384356040756, + "learning_rate": 9.875324178370111e-06, + "loss": 0.3422, + "step": 4502 + }, + { + "epoch": 0.09911952454325336, + "grad_norm": 1.7182228625592713, + "learning_rate": 9.875245061117078e-06, + "loss": 0.3713, + "step": 4503 + }, + { + "epoch": 0.09914153642967202, + "grad_norm": 1.232416889391913, + "learning_rate": 9.875165919085872e-06, + "loss": 0.2597, + "step": 4504 + }, + { + "epoch": 0.09916354831609069, + "grad_norm": 1.7507596884413377, + "learning_rate": 9.875086752276901e-06, + "loss": 0.451, + "step": 4505 + }, + { + "epoch": 0.09918556020250935, + "grad_norm": 1.3917972921087223, + "learning_rate": 9.875007560690565e-06, + "loss": 0.4147, + "step": 4506 + }, + { + "epoch": 0.09920757208892803, + "grad_norm": 1.5563471602779837, + "learning_rate": 9.874928344327264e-06, + "loss": 0.3777, + "step": 4507 + }, + { + "epoch": 0.09922958397534669, + "grad_norm": 1.314766781781188, + "learning_rate": 9.874849103187403e-06, + "loss": 0.31, + "step": 4508 + }, + { + "epoch": 0.09925159586176535, + "grad_norm": 0.6816524509241739, + "learning_rate": 9.874769837271386e-06, + "loss": 0.5829, + "step": 4509 + }, + { + "epoch": 0.09927360774818401, + "grad_norm": 1.2113784032035497, + "learning_rate": 9.874690546579615e-06, + "loss": 0.2919, + "step": 4510 + }, + { + "epoch": 0.09929561963460269, + "grad_norm": 1.1875826313217215, + "learning_rate": 9.874611231112491e-06, + "loss": 0.318, + "step": 4511 + }, + { + "epoch": 0.09931763152102135, + "grad_norm": 0.6819767575176074, + "learning_rate": 9.874531890870419e-06, + "loss": 0.5573, + "step": 4512 + }, + { + "epoch": 0.09933964340744002, + "grad_norm": 2.6063175366834246, + "learning_rate": 9.874452525853802e-06, + "loss": 0.4269, + "step": 4513 + }, + { + "epoch": 0.09936165529385868, + "grad_norm": 1.7544666889318126, + "learning_rate": 9.874373136063042e-06, + "loss": 0.3611, + "step": 4514 + }, + { + "epoch": 0.09938366718027734, + "grad_norm": 0.6631663316349516, + "learning_rate": 9.874293721498547e-06, + "loss": 0.5808, + "step": 4515 + }, + { + "epoch": 0.09940567906669602, + "grad_norm": 1.8843431872437195, + "learning_rate": 9.874214282160715e-06, + "loss": 0.3787, + "step": 4516 + }, + { + "epoch": 0.09942769095311468, + "grad_norm": 1.376561842432452, + "learning_rate": 9.874134818049952e-06, + "loss": 0.3137, + "step": 4517 + }, + { + "epoch": 0.09944970283953335, + "grad_norm": 1.4840622519909752, + "learning_rate": 9.87405532916666e-06, + "loss": 0.3943, + "step": 4518 + }, + { + "epoch": 0.09947171472595201, + "grad_norm": 1.5387439041828985, + "learning_rate": 9.873975815511249e-06, + "loss": 0.382, + "step": 4519 + }, + { + "epoch": 0.09949372661237069, + "grad_norm": 1.3333352915172625, + "learning_rate": 9.873896277084116e-06, + "loss": 0.3092, + "step": 4520 + }, + { + "epoch": 0.09951573849878935, + "grad_norm": 1.3203940302781851, + "learning_rate": 9.87381671388567e-06, + "loss": 0.2864, + "step": 4521 + }, + { + "epoch": 0.09953775038520801, + "grad_norm": 1.456807295301269, + "learning_rate": 9.873737125916314e-06, + "loss": 0.4533, + "step": 4522 + }, + { + "epoch": 0.09955976227162668, + "grad_norm": 1.4414885775595785, + "learning_rate": 9.873657513176452e-06, + "loss": 0.4483, + "step": 4523 + }, + { + "epoch": 0.09958177415804534, + "grad_norm": 1.999171908819476, + "learning_rate": 9.873577875666486e-06, + "loss": 0.3331, + "step": 4524 + }, + { + "epoch": 0.09960378604446402, + "grad_norm": 2.0290974184535266, + "learning_rate": 9.873498213386826e-06, + "loss": 0.3784, + "step": 4525 + }, + { + "epoch": 0.09962579793088268, + "grad_norm": 1.7840088458480359, + "learning_rate": 9.873418526337875e-06, + "loss": 0.3262, + "step": 4526 + }, + { + "epoch": 0.09964780981730134, + "grad_norm": 1.9939849923995931, + "learning_rate": 9.873338814520036e-06, + "loss": 0.4202, + "step": 4527 + }, + { + "epoch": 0.09966982170372, + "grad_norm": 2.2499875336141986, + "learning_rate": 9.873259077933716e-06, + "loss": 0.5096, + "step": 4528 + }, + { + "epoch": 0.09969183359013868, + "grad_norm": 1.6812654857833167, + "learning_rate": 9.87317931657932e-06, + "loss": 0.378, + "step": 4529 + }, + { + "epoch": 0.09971384547655734, + "grad_norm": 1.5428985924949388, + "learning_rate": 9.873099530457252e-06, + "loss": 0.4795, + "step": 4530 + }, + { + "epoch": 0.09973585736297601, + "grad_norm": 1.4260583591563054, + "learning_rate": 9.873019719567921e-06, + "loss": 0.3521, + "step": 4531 + }, + { + "epoch": 0.09975786924939467, + "grad_norm": 1.8650469383606871, + "learning_rate": 9.87293988391173e-06, + "loss": 0.3722, + "step": 4532 + }, + { + "epoch": 0.09977988113581333, + "grad_norm": 1.4410167344120524, + "learning_rate": 9.872860023489083e-06, + "loss": 0.4056, + "step": 4533 + }, + { + "epoch": 0.09980189302223201, + "grad_norm": 0.7670203955477929, + "learning_rate": 9.87278013830039e-06, + "loss": 0.5702, + "step": 4534 + }, + { + "epoch": 0.09982390490865067, + "grad_norm": 1.4832657455839664, + "learning_rate": 9.872700228346055e-06, + "loss": 0.3676, + "step": 4535 + }, + { + "epoch": 0.09984591679506934, + "grad_norm": 1.5354393661876808, + "learning_rate": 9.872620293626482e-06, + "loss": 0.3453, + "step": 4536 + }, + { + "epoch": 0.099867928681488, + "grad_norm": 1.6612389780359496, + "learning_rate": 9.872540334142082e-06, + "loss": 0.4707, + "step": 4537 + }, + { + "epoch": 0.09988994056790668, + "grad_norm": 1.6642810952072775, + "learning_rate": 9.872460349893259e-06, + "loss": 0.5039, + "step": 4538 + }, + { + "epoch": 0.09991195245432534, + "grad_norm": 1.6173021347135652, + "learning_rate": 9.872380340880416e-06, + "loss": 0.3228, + "step": 4539 + }, + { + "epoch": 0.099933964340744, + "grad_norm": 1.7738916339824056, + "learning_rate": 9.872300307103966e-06, + "loss": 0.4316, + "step": 4540 + }, + { + "epoch": 0.09995597622716267, + "grad_norm": 1.3152731235231765, + "learning_rate": 9.872220248564312e-06, + "loss": 0.3602, + "step": 4541 + }, + { + "epoch": 0.09997798811358133, + "grad_norm": 1.3736042379603994, + "learning_rate": 9.872140165261862e-06, + "loss": 0.3271, + "step": 4542 + }, + { + "epoch": 0.1, + "grad_norm": 1.2650096936467115, + "learning_rate": 9.872060057197022e-06, + "loss": 0.3229, + "step": 4543 + }, + { + "epoch": 0.10002201188641867, + "grad_norm": 1.5795805343796923, + "learning_rate": 9.871979924370198e-06, + "loss": 0.3889, + "step": 4544 + }, + { + "epoch": 0.10004402377283733, + "grad_norm": 1.5158074752087662, + "learning_rate": 9.871899766781802e-06, + "loss": 0.3422, + "step": 4545 + }, + { + "epoch": 0.100066035659256, + "grad_norm": 2.3761044079782616, + "learning_rate": 9.871819584432237e-06, + "loss": 0.4498, + "step": 4546 + }, + { + "epoch": 0.10008804754567466, + "grad_norm": 1.8283881469424064, + "learning_rate": 9.871739377321913e-06, + "loss": 0.4312, + "step": 4547 + }, + { + "epoch": 0.10011005943209333, + "grad_norm": 0.7973065695546268, + "learning_rate": 9.871659145451236e-06, + "loss": 0.6024, + "step": 4548 + }, + { + "epoch": 0.100132071318512, + "grad_norm": 0.7296658172555838, + "learning_rate": 9.871578888820615e-06, + "loss": 0.5515, + "step": 4549 + }, + { + "epoch": 0.10015408320493066, + "grad_norm": 2.809963227515393, + "learning_rate": 9.871498607430457e-06, + "loss": 0.399, + "step": 4550 + }, + { + "epoch": 0.10017609509134932, + "grad_norm": 1.6406366054415107, + "learning_rate": 9.871418301281169e-06, + "loss": 0.4424, + "step": 4551 + }, + { + "epoch": 0.100198106977768, + "grad_norm": 0.6702841219792938, + "learning_rate": 9.87133797037316e-06, + "loss": 0.5671, + "step": 4552 + }, + { + "epoch": 0.10022011886418666, + "grad_norm": 1.4367126230645952, + "learning_rate": 9.871257614706842e-06, + "loss": 0.2976, + "step": 4553 + }, + { + "epoch": 0.10024213075060533, + "grad_norm": 0.6996069966962267, + "learning_rate": 9.871177234282618e-06, + "loss": 0.5533, + "step": 4554 + }, + { + "epoch": 0.10026414263702399, + "grad_norm": 2.044733525732044, + "learning_rate": 9.871096829100899e-06, + "loss": 0.3896, + "step": 4555 + }, + { + "epoch": 0.10028615452344265, + "grad_norm": 1.3889253096592091, + "learning_rate": 9.871016399162094e-06, + "loss": 0.3215, + "step": 4556 + }, + { + "epoch": 0.10030816640986133, + "grad_norm": 1.6955183489751755, + "learning_rate": 9.87093594446661e-06, + "loss": 0.3812, + "step": 4557 + }, + { + "epoch": 0.10033017829627999, + "grad_norm": 1.3201573396962563, + "learning_rate": 9.870855465014856e-06, + "loss": 0.3797, + "step": 4558 + }, + { + "epoch": 0.10035219018269866, + "grad_norm": 0.677856437001108, + "learning_rate": 9.870774960807245e-06, + "loss": 0.5579, + "step": 4559 + }, + { + "epoch": 0.10037420206911732, + "grad_norm": 1.3298461176332503, + "learning_rate": 9.870694431844182e-06, + "loss": 0.326, + "step": 4560 + }, + { + "epoch": 0.100396213955536, + "grad_norm": 1.47086180178091, + "learning_rate": 9.870613878126077e-06, + "loss": 0.3866, + "step": 4561 + }, + { + "epoch": 0.10041822584195466, + "grad_norm": 1.5412353735749111, + "learning_rate": 9.870533299653341e-06, + "loss": 0.327, + "step": 4562 + }, + { + "epoch": 0.10044023772837332, + "grad_norm": 2.891162983544448, + "learning_rate": 9.870452696426381e-06, + "loss": 0.3821, + "step": 4563 + }, + { + "epoch": 0.10046224961479198, + "grad_norm": 1.6725236210835492, + "learning_rate": 9.87037206844561e-06, + "loss": 0.4035, + "step": 4564 + }, + { + "epoch": 0.10048426150121065, + "grad_norm": 1.4615186868347108, + "learning_rate": 9.870291415711436e-06, + "loss": 0.4037, + "step": 4565 + }, + { + "epoch": 0.10050627338762932, + "grad_norm": 1.3022804112282453, + "learning_rate": 9.870210738224267e-06, + "loss": 0.448, + "step": 4566 + }, + { + "epoch": 0.10052828527404799, + "grad_norm": 1.4101541751160498, + "learning_rate": 9.870130035984514e-06, + "loss": 0.3067, + "step": 4567 + }, + { + "epoch": 0.10055029716046665, + "grad_norm": 1.4727049151570835, + "learning_rate": 9.87004930899259e-06, + "loss": 0.4552, + "step": 4568 + }, + { + "epoch": 0.10057230904688531, + "grad_norm": 1.3472002996427672, + "learning_rate": 9.869968557248903e-06, + "loss": 0.3027, + "step": 4569 + }, + { + "epoch": 0.10059432093330399, + "grad_norm": 1.9620139434478947, + "learning_rate": 9.869887780753862e-06, + "loss": 0.372, + "step": 4570 + }, + { + "epoch": 0.10061633281972265, + "grad_norm": 1.8075082046937079, + "learning_rate": 9.869806979507882e-06, + "loss": 0.4186, + "step": 4571 + }, + { + "epoch": 0.10063834470614132, + "grad_norm": 1.4171991523757925, + "learning_rate": 9.86972615351137e-06, + "loss": 0.3041, + "step": 4572 + }, + { + "epoch": 0.10066035659255998, + "grad_norm": 0.6980880560322985, + "learning_rate": 9.869645302764737e-06, + "loss": 0.5415, + "step": 4573 + }, + { + "epoch": 0.10068236847897864, + "grad_norm": 1.5461967058386006, + "learning_rate": 9.869564427268394e-06, + "loss": 0.4013, + "step": 4574 + }, + { + "epoch": 0.10070438036539732, + "grad_norm": 1.5159021994005852, + "learning_rate": 9.869483527022754e-06, + "loss": 0.3407, + "step": 4575 + }, + { + "epoch": 0.10072639225181598, + "grad_norm": 1.4238027809305116, + "learning_rate": 9.869402602028228e-06, + "loss": 0.2974, + "step": 4576 + }, + { + "epoch": 0.10074840413823465, + "grad_norm": 1.6871351363077514, + "learning_rate": 9.869321652285224e-06, + "loss": 0.4459, + "step": 4577 + }, + { + "epoch": 0.10077041602465331, + "grad_norm": 2.027447230737631, + "learning_rate": 9.869240677794156e-06, + "loss": 0.391, + "step": 4578 + }, + { + "epoch": 0.10079242791107199, + "grad_norm": 1.425438736813441, + "learning_rate": 9.869159678555436e-06, + "loss": 0.4144, + "step": 4579 + }, + { + "epoch": 0.10081443979749065, + "grad_norm": 1.7064452860042032, + "learning_rate": 9.869078654569475e-06, + "loss": 0.3644, + "step": 4580 + }, + { + "epoch": 0.10083645168390931, + "grad_norm": 1.4959154688063108, + "learning_rate": 9.868997605836683e-06, + "loss": 0.3057, + "step": 4581 + }, + { + "epoch": 0.10085846357032797, + "grad_norm": 1.4231067567903508, + "learning_rate": 9.868916532357475e-06, + "loss": 0.3795, + "step": 4582 + }, + { + "epoch": 0.10088047545674664, + "grad_norm": 1.4230799644268466, + "learning_rate": 9.86883543413226e-06, + "loss": 0.2575, + "step": 4583 + }, + { + "epoch": 0.10090248734316531, + "grad_norm": 1.5463616077451583, + "learning_rate": 9.868754311161453e-06, + "loss": 0.3239, + "step": 4584 + }, + { + "epoch": 0.10092449922958398, + "grad_norm": 2.1918294428549245, + "learning_rate": 9.868673163445465e-06, + "loss": 0.3592, + "step": 4585 + }, + { + "epoch": 0.10094651111600264, + "grad_norm": 1.3280564254823164, + "learning_rate": 9.868591990984709e-06, + "loss": 0.3383, + "step": 4586 + }, + { + "epoch": 0.1009685230024213, + "grad_norm": 0.7919573576196259, + "learning_rate": 9.868510793779596e-06, + "loss": 0.5975, + "step": 4587 + }, + { + "epoch": 0.10099053488883997, + "grad_norm": 1.5183530866655373, + "learning_rate": 9.86842957183054e-06, + "loss": 0.4536, + "step": 4588 + }, + { + "epoch": 0.10101254677525864, + "grad_norm": 2.0375703634473528, + "learning_rate": 9.868348325137954e-06, + "loss": 0.2836, + "step": 4589 + }, + { + "epoch": 0.1010345586616773, + "grad_norm": 1.6681774133823968, + "learning_rate": 9.86826705370225e-06, + "loss": 0.2638, + "step": 4590 + }, + { + "epoch": 0.10105657054809597, + "grad_norm": 1.5754569485944858, + "learning_rate": 9.868185757523842e-06, + "loss": 0.4235, + "step": 4591 + }, + { + "epoch": 0.10107858243451463, + "grad_norm": 1.6122054406298247, + "learning_rate": 9.868104436603143e-06, + "loss": 0.3337, + "step": 4592 + }, + { + "epoch": 0.10110059432093331, + "grad_norm": 0.6664933086385847, + "learning_rate": 9.868023090940565e-06, + "loss": 0.5657, + "step": 4593 + }, + { + "epoch": 0.10112260620735197, + "grad_norm": 1.6540140852561114, + "learning_rate": 9.867941720536523e-06, + "loss": 0.3254, + "step": 4594 + }, + { + "epoch": 0.10114461809377064, + "grad_norm": 2.0069574724308406, + "learning_rate": 9.86786032539143e-06, + "loss": 0.4186, + "step": 4595 + }, + { + "epoch": 0.1011666299801893, + "grad_norm": 1.4055092700767413, + "learning_rate": 9.867778905505699e-06, + "loss": 0.4112, + "step": 4596 + }, + { + "epoch": 0.10118864186660796, + "grad_norm": 2.302271737762081, + "learning_rate": 9.867697460879746e-06, + "loss": 0.3621, + "step": 4597 + }, + { + "epoch": 0.10121065375302664, + "grad_norm": 1.9480142638173945, + "learning_rate": 9.86761599151398e-06, + "loss": 0.342, + "step": 4598 + }, + { + "epoch": 0.1012326656394453, + "grad_norm": 0.7188784860438344, + "learning_rate": 9.867534497408822e-06, + "loss": 0.5821, + "step": 4599 + }, + { + "epoch": 0.10125467752586396, + "grad_norm": 1.8208551566875943, + "learning_rate": 9.867452978564682e-06, + "loss": 0.392, + "step": 4600 + }, + { + "epoch": 0.10127668941228263, + "grad_norm": 1.8512562163422142, + "learning_rate": 9.867371434981976e-06, + "loss": 0.3812, + "step": 4601 + }, + { + "epoch": 0.1012987012987013, + "grad_norm": 1.6500851379918833, + "learning_rate": 9.867289866661116e-06, + "loss": 0.4019, + "step": 4602 + }, + { + "epoch": 0.10132071318511997, + "grad_norm": 2.601493385618311, + "learning_rate": 9.867208273602517e-06, + "loss": 0.4011, + "step": 4603 + }, + { + "epoch": 0.10134272507153863, + "grad_norm": 2.2906970798964554, + "learning_rate": 9.867126655806596e-06, + "loss": 0.348, + "step": 4604 + }, + { + "epoch": 0.1013647369579573, + "grad_norm": 2.3373682818860004, + "learning_rate": 9.867045013273767e-06, + "loss": 0.4835, + "step": 4605 + }, + { + "epoch": 0.10138674884437596, + "grad_norm": 1.6424592450567033, + "learning_rate": 9.866963346004444e-06, + "loss": 0.3367, + "step": 4606 + }, + { + "epoch": 0.10140876073079463, + "grad_norm": 0.6540365173621543, + "learning_rate": 9.866881653999043e-06, + "loss": 0.5609, + "step": 4607 + }, + { + "epoch": 0.1014307726172133, + "grad_norm": 1.88726451829329, + "learning_rate": 9.866799937257978e-06, + "loss": 0.4731, + "step": 4608 + }, + { + "epoch": 0.10145278450363196, + "grad_norm": 3.400413788195294, + "learning_rate": 9.866718195781667e-06, + "loss": 0.3193, + "step": 4609 + }, + { + "epoch": 0.10147479639005062, + "grad_norm": 1.523751247834615, + "learning_rate": 9.86663642957052e-06, + "loss": 0.3765, + "step": 4610 + }, + { + "epoch": 0.1014968082764693, + "grad_norm": 2.558019675077382, + "learning_rate": 9.866554638624958e-06, + "loss": 0.3634, + "step": 4611 + }, + { + "epoch": 0.10151882016288796, + "grad_norm": 0.6463295569695338, + "learning_rate": 9.866472822945395e-06, + "loss": 0.5654, + "step": 4612 + }, + { + "epoch": 0.10154083204930663, + "grad_norm": 1.2879765732788964, + "learning_rate": 9.866390982532246e-06, + "loss": 0.3719, + "step": 4613 + }, + { + "epoch": 0.10156284393572529, + "grad_norm": 0.6455777663642648, + "learning_rate": 9.866309117385928e-06, + "loss": 0.571, + "step": 4614 + }, + { + "epoch": 0.10158485582214395, + "grad_norm": 1.9758616564514595, + "learning_rate": 9.866227227506857e-06, + "loss": 0.3886, + "step": 4615 + }, + { + "epoch": 0.10160686770856263, + "grad_norm": 1.8709149203868431, + "learning_rate": 9.866145312895448e-06, + "loss": 0.4006, + "step": 4616 + }, + { + "epoch": 0.10162887959498129, + "grad_norm": 1.5842386236304673, + "learning_rate": 9.86606337355212e-06, + "loss": 0.3754, + "step": 4617 + }, + { + "epoch": 0.10165089148139995, + "grad_norm": 1.5848119286251792, + "learning_rate": 9.865981409477285e-06, + "loss": 0.3227, + "step": 4618 + }, + { + "epoch": 0.10167290336781862, + "grad_norm": 0.6630861915689399, + "learning_rate": 9.865899420671363e-06, + "loss": 0.5707, + "step": 4619 + }, + { + "epoch": 0.1016949152542373, + "grad_norm": 1.2331205304920807, + "learning_rate": 9.86581740713477e-06, + "loss": 0.355, + "step": 4620 + }, + { + "epoch": 0.10171692714065596, + "grad_norm": 2.1854000281883335, + "learning_rate": 9.865735368867924e-06, + "loss": 0.3637, + "step": 4621 + }, + { + "epoch": 0.10173893902707462, + "grad_norm": 1.5387622634647307, + "learning_rate": 9.865653305871238e-06, + "loss": 0.3902, + "step": 4622 + }, + { + "epoch": 0.10176095091349328, + "grad_norm": 0.6684526406419334, + "learning_rate": 9.865571218145133e-06, + "loss": 0.5868, + "step": 4623 + }, + { + "epoch": 0.10178296279991195, + "grad_norm": 1.4647494532948577, + "learning_rate": 9.865489105690027e-06, + "loss": 0.4102, + "step": 4624 + }, + { + "epoch": 0.10180497468633062, + "grad_norm": 0.6236273696636847, + "learning_rate": 9.865406968506332e-06, + "loss": 0.5713, + "step": 4625 + }, + { + "epoch": 0.10182698657274929, + "grad_norm": 2.0035149365930316, + "learning_rate": 9.865324806594472e-06, + "loss": 0.4269, + "step": 4626 + }, + { + "epoch": 0.10184899845916795, + "grad_norm": 1.841957192167263, + "learning_rate": 9.865242619954859e-06, + "loss": 0.303, + "step": 4627 + }, + { + "epoch": 0.10187101034558661, + "grad_norm": 1.6229974048036704, + "learning_rate": 9.865160408587913e-06, + "loss": 0.3713, + "step": 4628 + }, + { + "epoch": 0.10189302223200529, + "grad_norm": 2.06513679272515, + "learning_rate": 9.865078172494053e-06, + "loss": 0.4138, + "step": 4629 + }, + { + "epoch": 0.10191503411842395, + "grad_norm": 1.697595393186847, + "learning_rate": 9.864995911673697e-06, + "loss": 0.3647, + "step": 4630 + }, + { + "epoch": 0.10193704600484262, + "grad_norm": 0.6527893954069641, + "learning_rate": 9.86491362612726e-06, + "loss": 0.5622, + "step": 4631 + }, + { + "epoch": 0.10195905789126128, + "grad_norm": 1.378790082894011, + "learning_rate": 9.864831315855163e-06, + "loss": 0.3455, + "step": 4632 + }, + { + "epoch": 0.10198106977767994, + "grad_norm": 1.6982829410533684, + "learning_rate": 9.864748980857822e-06, + "loss": 0.3231, + "step": 4633 + }, + { + "epoch": 0.10200308166409862, + "grad_norm": 1.5496821120478736, + "learning_rate": 9.864666621135658e-06, + "loss": 0.3991, + "step": 4634 + }, + { + "epoch": 0.10202509355051728, + "grad_norm": 0.6329436659353767, + "learning_rate": 9.864584236689088e-06, + "loss": 0.5655, + "step": 4635 + }, + { + "epoch": 0.10204710543693594, + "grad_norm": 1.6797385117833086, + "learning_rate": 9.864501827518533e-06, + "loss": 0.3893, + "step": 4636 + }, + { + "epoch": 0.10206911732335461, + "grad_norm": 0.638102074562945, + "learning_rate": 9.864419393624408e-06, + "loss": 0.5603, + "step": 4637 + }, + { + "epoch": 0.10209112920977327, + "grad_norm": 0.6385401369845803, + "learning_rate": 9.864336935007134e-06, + "loss": 0.5688, + "step": 4638 + }, + { + "epoch": 0.10211314109619195, + "grad_norm": 1.487096516155226, + "learning_rate": 9.864254451667133e-06, + "loss": 0.3964, + "step": 4639 + }, + { + "epoch": 0.10213515298261061, + "grad_norm": 1.3570507245871621, + "learning_rate": 9.864171943604819e-06, + "loss": 0.45, + "step": 4640 + }, + { + "epoch": 0.10215716486902927, + "grad_norm": 2.275962687533415, + "learning_rate": 9.864089410820614e-06, + "loss": 0.3189, + "step": 4641 + }, + { + "epoch": 0.10217917675544794, + "grad_norm": 1.5809248356690382, + "learning_rate": 9.864006853314937e-06, + "loss": 0.4261, + "step": 4642 + }, + { + "epoch": 0.10220118864186661, + "grad_norm": 2.00827766388376, + "learning_rate": 9.863924271088208e-06, + "loss": 0.3583, + "step": 4643 + }, + { + "epoch": 0.10222320052828528, + "grad_norm": 0.6358244703687506, + "learning_rate": 9.863841664140847e-06, + "loss": 0.5618, + "step": 4644 + }, + { + "epoch": 0.10224521241470394, + "grad_norm": 0.661907300593839, + "learning_rate": 9.863759032473273e-06, + "loss": 0.578, + "step": 4645 + }, + { + "epoch": 0.1022672243011226, + "grad_norm": 1.7518764238330093, + "learning_rate": 9.863676376085904e-06, + "loss": 0.46, + "step": 4646 + }, + { + "epoch": 0.10228923618754127, + "grad_norm": 0.6546820437884686, + "learning_rate": 9.863593694979165e-06, + "loss": 0.5842, + "step": 4647 + }, + { + "epoch": 0.10231124807395994, + "grad_norm": 0.6249297927574989, + "learning_rate": 9.863510989153474e-06, + "loss": 0.5558, + "step": 4648 + }, + { + "epoch": 0.1023332599603786, + "grad_norm": 0.6421254687570559, + "learning_rate": 9.86342825860925e-06, + "loss": 0.5458, + "step": 4649 + }, + { + "epoch": 0.10235527184679727, + "grad_norm": 0.6498829318388402, + "learning_rate": 9.863345503346916e-06, + "loss": 0.5421, + "step": 4650 + }, + { + "epoch": 0.10237728373321593, + "grad_norm": 1.462919976016678, + "learning_rate": 9.86326272336689e-06, + "loss": 0.3897, + "step": 4651 + }, + { + "epoch": 0.10239929561963461, + "grad_norm": 1.7479713161140051, + "learning_rate": 9.863179918669591e-06, + "loss": 0.4044, + "step": 4652 + }, + { + "epoch": 0.10242130750605327, + "grad_norm": 1.4427611634916053, + "learning_rate": 9.863097089255445e-06, + "loss": 0.3117, + "step": 4653 + }, + { + "epoch": 0.10244331939247193, + "grad_norm": 1.6535948946496863, + "learning_rate": 9.86301423512487e-06, + "loss": 0.365, + "step": 4654 + }, + { + "epoch": 0.1024653312788906, + "grad_norm": 1.7283056110314219, + "learning_rate": 9.862931356278288e-06, + "loss": 0.3806, + "step": 4655 + }, + { + "epoch": 0.10248734316530926, + "grad_norm": 2.148810221867711, + "learning_rate": 9.862848452716121e-06, + "loss": 0.5149, + "step": 4656 + }, + { + "epoch": 0.10250935505172794, + "grad_norm": 2.561188019038365, + "learning_rate": 9.862765524438788e-06, + "loss": 0.3437, + "step": 4657 + }, + { + "epoch": 0.1025313669381466, + "grad_norm": 1.8012874457863801, + "learning_rate": 9.862682571446714e-06, + "loss": 0.3786, + "step": 4658 + }, + { + "epoch": 0.10255337882456526, + "grad_norm": 1.6749115599834339, + "learning_rate": 9.862599593740316e-06, + "loss": 0.4599, + "step": 4659 + }, + { + "epoch": 0.10257539071098393, + "grad_norm": 1.6761440751914347, + "learning_rate": 9.862516591320019e-06, + "loss": 0.3786, + "step": 4660 + }, + { + "epoch": 0.1025974025974026, + "grad_norm": 1.47423683623061, + "learning_rate": 9.862433564186243e-06, + "loss": 0.3765, + "step": 4661 + }, + { + "epoch": 0.10261941448382127, + "grad_norm": 0.7647884061935051, + "learning_rate": 9.862350512339412e-06, + "loss": 0.5901, + "step": 4662 + }, + { + "epoch": 0.10264142637023993, + "grad_norm": 1.6885330223968091, + "learning_rate": 9.862267435779947e-06, + "loss": 0.3488, + "step": 4663 + }, + { + "epoch": 0.10266343825665859, + "grad_norm": 2.9330941168699414, + "learning_rate": 9.86218433450827e-06, + "loss": 0.444, + "step": 4664 + }, + { + "epoch": 0.10268545014307726, + "grad_norm": 1.53263501580122, + "learning_rate": 9.862101208524801e-06, + "loss": 0.421, + "step": 4665 + }, + { + "epoch": 0.10270746202949593, + "grad_norm": 1.497643592474996, + "learning_rate": 9.862018057829969e-06, + "loss": 0.3527, + "step": 4666 + }, + { + "epoch": 0.1027294739159146, + "grad_norm": 2.434065881656715, + "learning_rate": 9.86193488242419e-06, + "loss": 0.3093, + "step": 4667 + }, + { + "epoch": 0.10275148580233326, + "grad_norm": 0.6746902728181311, + "learning_rate": 9.86185168230789e-06, + "loss": 0.5452, + "step": 4668 + }, + { + "epoch": 0.10277349768875192, + "grad_norm": 0.699853650411778, + "learning_rate": 9.861768457481494e-06, + "loss": 0.5684, + "step": 4669 + }, + { + "epoch": 0.1027955095751706, + "grad_norm": 1.690466518882752, + "learning_rate": 9.86168520794542e-06, + "loss": 0.346, + "step": 4670 + }, + { + "epoch": 0.10281752146158926, + "grad_norm": 0.6572838091191342, + "learning_rate": 9.861601933700093e-06, + "loss": 0.5653, + "step": 4671 + }, + { + "epoch": 0.10283953334800792, + "grad_norm": 1.7730932382888231, + "learning_rate": 9.861518634745936e-06, + "loss": 0.379, + "step": 4672 + }, + { + "epoch": 0.10286154523442659, + "grad_norm": 1.3717092103496071, + "learning_rate": 9.861435311083376e-06, + "loss": 0.3044, + "step": 4673 + }, + { + "epoch": 0.10288355712084525, + "grad_norm": 2.4987122047858623, + "learning_rate": 9.861351962712831e-06, + "loss": 0.4346, + "step": 4674 + }, + { + "epoch": 0.10290556900726393, + "grad_norm": 1.4265838433706164, + "learning_rate": 9.861268589634728e-06, + "loss": 0.3573, + "step": 4675 + }, + { + "epoch": 0.10292758089368259, + "grad_norm": 1.5272018967284957, + "learning_rate": 9.86118519184949e-06, + "loss": 0.4122, + "step": 4676 + }, + { + "epoch": 0.10294959278010125, + "grad_norm": 2.0761793774122825, + "learning_rate": 9.861101769357541e-06, + "loss": 0.3661, + "step": 4677 + }, + { + "epoch": 0.10297160466651992, + "grad_norm": 2.0166005094053934, + "learning_rate": 9.861018322159303e-06, + "loss": 0.3415, + "step": 4678 + }, + { + "epoch": 0.10299361655293858, + "grad_norm": 1.6707159515316155, + "learning_rate": 9.860934850255203e-06, + "loss": 0.4112, + "step": 4679 + }, + { + "epoch": 0.10301562843935726, + "grad_norm": 1.5613979756959993, + "learning_rate": 9.860851353645664e-06, + "loss": 0.387, + "step": 4680 + }, + { + "epoch": 0.10303764032577592, + "grad_norm": 2.650082750760426, + "learning_rate": 9.86076783233111e-06, + "loss": 0.4092, + "step": 4681 + }, + { + "epoch": 0.10305965221219458, + "grad_norm": 0.8558340684572289, + "learning_rate": 9.860684286311968e-06, + "loss": 0.5828, + "step": 4682 + }, + { + "epoch": 0.10308166409861325, + "grad_norm": 1.5516752414286021, + "learning_rate": 9.860600715588657e-06, + "loss": 0.3903, + "step": 4683 + }, + { + "epoch": 0.10310367598503192, + "grad_norm": 1.5767728285452736, + "learning_rate": 9.860517120161608e-06, + "loss": 0.4091, + "step": 4684 + }, + { + "epoch": 0.10312568787145059, + "grad_norm": 2.6185799984303197, + "learning_rate": 9.860433500031242e-06, + "loss": 0.4293, + "step": 4685 + }, + { + "epoch": 0.10314769975786925, + "grad_norm": 0.6735783585408879, + "learning_rate": 9.860349855197986e-06, + "loss": 0.5776, + "step": 4686 + }, + { + "epoch": 0.10316971164428791, + "grad_norm": 0.6612010686795943, + "learning_rate": 9.860266185662264e-06, + "loss": 0.5735, + "step": 4687 + }, + { + "epoch": 0.10319172353070657, + "grad_norm": 1.964438122259984, + "learning_rate": 9.860182491424501e-06, + "loss": 0.3906, + "step": 4688 + }, + { + "epoch": 0.10321373541712525, + "grad_norm": 1.5006527712451574, + "learning_rate": 9.860098772485122e-06, + "loss": 0.2456, + "step": 4689 + }, + { + "epoch": 0.10323574730354391, + "grad_norm": 1.6765309921502902, + "learning_rate": 9.860015028844555e-06, + "loss": 0.3961, + "step": 4690 + }, + { + "epoch": 0.10325775918996258, + "grad_norm": 1.9402122931061725, + "learning_rate": 9.859931260503225e-06, + "loss": 0.3525, + "step": 4691 + }, + { + "epoch": 0.10327977107638124, + "grad_norm": 1.51985112119042, + "learning_rate": 9.859847467461555e-06, + "loss": 0.377, + "step": 4692 + }, + { + "epoch": 0.10330178296279992, + "grad_norm": 1.653277321850585, + "learning_rate": 9.859763649719972e-06, + "loss": 0.359, + "step": 4693 + }, + { + "epoch": 0.10332379484921858, + "grad_norm": 2.036082596996399, + "learning_rate": 9.859679807278903e-06, + "loss": 0.3248, + "step": 4694 + }, + { + "epoch": 0.10334580673563724, + "grad_norm": 2.1759506107354474, + "learning_rate": 9.859595940138772e-06, + "loss": 0.4073, + "step": 4695 + }, + { + "epoch": 0.10336781862205591, + "grad_norm": 1.539619711988651, + "learning_rate": 9.859512048300011e-06, + "loss": 0.4114, + "step": 4696 + }, + { + "epoch": 0.10338983050847457, + "grad_norm": 2.5248035087260052, + "learning_rate": 9.85942813176304e-06, + "loss": 0.3254, + "step": 4697 + }, + { + "epoch": 0.10341184239489325, + "grad_norm": 2.151879638207733, + "learning_rate": 9.859344190528289e-06, + "loss": 0.3981, + "step": 4698 + }, + { + "epoch": 0.10343385428131191, + "grad_norm": 1.2174623686147272, + "learning_rate": 9.859260224596183e-06, + "loss": 0.3158, + "step": 4699 + }, + { + "epoch": 0.10345586616773057, + "grad_norm": 0.7453757518513745, + "learning_rate": 9.859176233967149e-06, + "loss": 0.566, + "step": 4700 + }, + { + "epoch": 0.10347787805414924, + "grad_norm": 1.632914422850276, + "learning_rate": 9.859092218641614e-06, + "loss": 0.3585, + "step": 4701 + }, + { + "epoch": 0.10349988994056791, + "grad_norm": 1.9297074881095346, + "learning_rate": 9.859008178620006e-06, + "loss": 0.358, + "step": 4702 + }, + { + "epoch": 0.10352190182698658, + "grad_norm": 1.7873160640776582, + "learning_rate": 9.85892411390275e-06, + "loss": 0.3628, + "step": 4703 + }, + { + "epoch": 0.10354391371340524, + "grad_norm": 8.00483767938981, + "learning_rate": 9.858840024490274e-06, + "loss": 0.3607, + "step": 4704 + }, + { + "epoch": 0.1035659255998239, + "grad_norm": 0.679267498750873, + "learning_rate": 9.858755910383009e-06, + "loss": 0.5799, + "step": 4705 + }, + { + "epoch": 0.10358793748624257, + "grad_norm": 1.9514460243342526, + "learning_rate": 9.858671771581377e-06, + "loss": 0.3553, + "step": 4706 + }, + { + "epoch": 0.10360994937266124, + "grad_norm": 1.6215945132518612, + "learning_rate": 9.858587608085809e-06, + "loss": 0.3254, + "step": 4707 + }, + { + "epoch": 0.1036319612590799, + "grad_norm": 0.6405966277924262, + "learning_rate": 9.85850341989673e-06, + "loss": 0.5709, + "step": 4708 + }, + { + "epoch": 0.10365397314549857, + "grad_norm": 1.3352001242881897, + "learning_rate": 9.858419207014572e-06, + "loss": 0.2924, + "step": 4709 + }, + { + "epoch": 0.10367598503191723, + "grad_norm": 3.0698325174297794, + "learning_rate": 9.858334969439759e-06, + "loss": 0.3224, + "step": 4710 + }, + { + "epoch": 0.10369799691833591, + "grad_norm": 1.238069914443932, + "learning_rate": 9.85825070717272e-06, + "loss": 0.3087, + "step": 4711 + }, + { + "epoch": 0.10372000880475457, + "grad_norm": 1.5345317846381195, + "learning_rate": 9.858166420213886e-06, + "loss": 0.4256, + "step": 4712 + }, + { + "epoch": 0.10374202069117323, + "grad_norm": 0.6450597115025702, + "learning_rate": 9.858082108563683e-06, + "loss": 0.563, + "step": 4713 + }, + { + "epoch": 0.1037640325775919, + "grad_norm": 1.5638569818576509, + "learning_rate": 9.857997772222541e-06, + "loss": 0.4434, + "step": 4714 + }, + { + "epoch": 0.10378604446401056, + "grad_norm": 1.875120163121548, + "learning_rate": 9.857913411190885e-06, + "loss": 0.3837, + "step": 4715 + }, + { + "epoch": 0.10380805635042924, + "grad_norm": 2.659065326016836, + "learning_rate": 9.857829025469149e-06, + "loss": 0.4622, + "step": 4716 + }, + { + "epoch": 0.1038300682368479, + "grad_norm": 1.4275468717222666, + "learning_rate": 9.857744615057758e-06, + "loss": 0.3333, + "step": 4717 + }, + { + "epoch": 0.10385208012326656, + "grad_norm": 1.619031046351624, + "learning_rate": 9.857660179957143e-06, + "loss": 0.4018, + "step": 4718 + }, + { + "epoch": 0.10387409200968523, + "grad_norm": 1.4831096242295463, + "learning_rate": 9.85757572016773e-06, + "loss": 0.3518, + "step": 4719 + }, + { + "epoch": 0.1038961038961039, + "grad_norm": 1.677742430422998, + "learning_rate": 9.857491235689954e-06, + "loss": 0.4542, + "step": 4720 + }, + { + "epoch": 0.10391811578252257, + "grad_norm": 1.53820047246517, + "learning_rate": 9.85740672652424e-06, + "loss": 0.3494, + "step": 4721 + }, + { + "epoch": 0.10394012766894123, + "grad_norm": 1.5040445620652347, + "learning_rate": 9.857322192671017e-06, + "loss": 0.3548, + "step": 4722 + }, + { + "epoch": 0.10396213955535989, + "grad_norm": 0.720828021414722, + "learning_rate": 9.857237634130717e-06, + "loss": 0.566, + "step": 4723 + }, + { + "epoch": 0.10398415144177856, + "grad_norm": 1.8689664479488977, + "learning_rate": 9.857153050903769e-06, + "loss": 0.3799, + "step": 4724 + }, + { + "epoch": 0.10400616332819723, + "grad_norm": 1.6066433794881714, + "learning_rate": 9.857068442990603e-06, + "loss": 0.3507, + "step": 4725 + }, + { + "epoch": 0.1040281752146159, + "grad_norm": 2.381054390534869, + "learning_rate": 9.856983810391648e-06, + "loss": 0.4195, + "step": 4726 + }, + { + "epoch": 0.10405018710103456, + "grad_norm": 1.87913971044407, + "learning_rate": 9.856899153107335e-06, + "loss": 0.4237, + "step": 4727 + }, + { + "epoch": 0.10407219898745322, + "grad_norm": 1.4582333940301522, + "learning_rate": 9.856814471138096e-06, + "loss": 0.37, + "step": 4728 + }, + { + "epoch": 0.10409421087387188, + "grad_norm": 1.5400860229858713, + "learning_rate": 9.856729764484358e-06, + "loss": 0.418, + "step": 4729 + }, + { + "epoch": 0.10411622276029056, + "grad_norm": 2.3127996326480034, + "learning_rate": 9.856645033146555e-06, + "loss": 0.427, + "step": 4730 + }, + { + "epoch": 0.10413823464670922, + "grad_norm": 2.0042426427440025, + "learning_rate": 9.856560277125115e-06, + "loss": 0.3519, + "step": 4731 + }, + { + "epoch": 0.10416024653312789, + "grad_norm": 0.7217275324619307, + "learning_rate": 9.856475496420468e-06, + "loss": 0.5563, + "step": 4732 + }, + { + "epoch": 0.10418225841954655, + "grad_norm": 1.6841191221390286, + "learning_rate": 9.856390691033048e-06, + "loss": 0.4365, + "step": 4733 + }, + { + "epoch": 0.10420427030596523, + "grad_norm": 1.552730416680124, + "learning_rate": 9.856305860963284e-06, + "loss": 0.3663, + "step": 4734 + }, + { + "epoch": 0.10422628219238389, + "grad_norm": 2.5446476296565836, + "learning_rate": 9.856221006211607e-06, + "loss": 0.3476, + "step": 4735 + }, + { + "epoch": 0.10424829407880255, + "grad_norm": 1.8209592036532418, + "learning_rate": 9.85613612677845e-06, + "loss": 0.3885, + "step": 4736 + }, + { + "epoch": 0.10427030596522122, + "grad_norm": 2.0127240788806526, + "learning_rate": 9.856051222664245e-06, + "loss": 0.2944, + "step": 4737 + }, + { + "epoch": 0.10429231785163988, + "grad_norm": 0.6500578231117707, + "learning_rate": 9.85596629386942e-06, + "loss": 0.5438, + "step": 4738 + }, + { + "epoch": 0.10431432973805856, + "grad_norm": 1.568234999445955, + "learning_rate": 9.855881340394408e-06, + "loss": 0.3855, + "step": 4739 + }, + { + "epoch": 0.10433634162447722, + "grad_norm": 5.3643394631009285, + "learning_rate": 9.85579636223964e-06, + "loss": 0.4641, + "step": 4740 + }, + { + "epoch": 0.10435835351089588, + "grad_norm": 1.5427101075834244, + "learning_rate": 9.855711359405552e-06, + "loss": 0.4085, + "step": 4741 + }, + { + "epoch": 0.10438036539731455, + "grad_norm": 1.4714160068023105, + "learning_rate": 9.855626331892571e-06, + "loss": 0.2943, + "step": 4742 + }, + { + "epoch": 0.10440237728373322, + "grad_norm": 1.383495755266476, + "learning_rate": 9.855541279701133e-06, + "loss": 0.4309, + "step": 4743 + }, + { + "epoch": 0.10442438917015189, + "grad_norm": 1.648916531620318, + "learning_rate": 9.855456202831669e-06, + "loss": 0.3801, + "step": 4744 + }, + { + "epoch": 0.10444640105657055, + "grad_norm": 1.5829080693828101, + "learning_rate": 9.85537110128461e-06, + "loss": 0.305, + "step": 4745 + }, + { + "epoch": 0.10446841294298921, + "grad_norm": 1.6465105843135308, + "learning_rate": 9.85528597506039e-06, + "loss": 0.5532, + "step": 4746 + }, + { + "epoch": 0.10449042482940787, + "grad_norm": 1.4270370144075872, + "learning_rate": 9.85520082415944e-06, + "loss": 0.275, + "step": 4747 + }, + { + "epoch": 0.10451243671582655, + "grad_norm": 1.6404970689906964, + "learning_rate": 9.855115648582194e-06, + "loss": 0.3629, + "step": 4748 + }, + { + "epoch": 0.10453444860224521, + "grad_norm": 1.9675116173895444, + "learning_rate": 9.855030448329084e-06, + "loss": 0.3649, + "step": 4749 + }, + { + "epoch": 0.10455646048866388, + "grad_norm": 0.6871974340742804, + "learning_rate": 9.854945223400546e-06, + "loss": 0.5457, + "step": 4750 + }, + { + "epoch": 0.10457847237508254, + "grad_norm": 0.6805192039884853, + "learning_rate": 9.85485997379701e-06, + "loss": 0.5778, + "step": 4751 + }, + { + "epoch": 0.10460048426150122, + "grad_norm": 1.5588816936004728, + "learning_rate": 9.85477469951891e-06, + "loss": 0.399, + "step": 4752 + }, + { + "epoch": 0.10462249614791988, + "grad_norm": 1.8159470149792132, + "learning_rate": 9.85468940056668e-06, + "loss": 0.3988, + "step": 4753 + }, + { + "epoch": 0.10464450803433854, + "grad_norm": 1.3235683706341943, + "learning_rate": 9.854604076940754e-06, + "loss": 0.412, + "step": 4754 + }, + { + "epoch": 0.1046665199207572, + "grad_norm": 1.5325234012779514, + "learning_rate": 9.854518728641564e-06, + "loss": 0.4183, + "step": 4755 + }, + { + "epoch": 0.10468853180717587, + "grad_norm": 1.7117919741513399, + "learning_rate": 9.854433355669545e-06, + "loss": 0.3621, + "step": 4756 + }, + { + "epoch": 0.10471054369359455, + "grad_norm": 2.58780344057055, + "learning_rate": 9.85434795802513e-06, + "loss": 0.4002, + "step": 4757 + }, + { + "epoch": 0.10473255558001321, + "grad_norm": 3.011607616662976, + "learning_rate": 9.854262535708755e-06, + "loss": 0.3661, + "step": 4758 + }, + { + "epoch": 0.10475456746643187, + "grad_norm": 2.4159952478685702, + "learning_rate": 9.854177088720852e-06, + "loss": 0.4395, + "step": 4759 + }, + { + "epoch": 0.10477657935285054, + "grad_norm": 1.5585644427884653, + "learning_rate": 9.854091617061854e-06, + "loss": 0.4398, + "step": 4760 + }, + { + "epoch": 0.10479859123926921, + "grad_norm": 0.8584609914284321, + "learning_rate": 9.854006120732201e-06, + "loss": 0.5837, + "step": 4761 + }, + { + "epoch": 0.10482060312568788, + "grad_norm": 1.791013127982391, + "learning_rate": 9.85392059973232e-06, + "loss": 0.3583, + "step": 4762 + }, + { + "epoch": 0.10484261501210654, + "grad_norm": 2.117797451631637, + "learning_rate": 9.853835054062653e-06, + "loss": 0.4529, + "step": 4763 + }, + { + "epoch": 0.1048646268985252, + "grad_norm": 1.6145261866754974, + "learning_rate": 9.853749483723629e-06, + "loss": 0.4233, + "step": 4764 + }, + { + "epoch": 0.10488663878494386, + "grad_norm": 1.5565544607127249, + "learning_rate": 9.853663888715688e-06, + "loss": 0.2863, + "step": 4765 + }, + { + "epoch": 0.10490865067136254, + "grad_norm": 1.3648284559430457, + "learning_rate": 9.85357826903926e-06, + "loss": 0.3974, + "step": 4766 + }, + { + "epoch": 0.1049306625577812, + "grad_norm": 2.372509924555831, + "learning_rate": 9.853492624694782e-06, + "loss": 0.4117, + "step": 4767 + }, + { + "epoch": 0.10495267444419987, + "grad_norm": 1.395688763329789, + "learning_rate": 9.853406955682691e-06, + "loss": 0.3795, + "step": 4768 + }, + { + "epoch": 0.10497468633061853, + "grad_norm": 1.6622785968705938, + "learning_rate": 9.853321262003422e-06, + "loss": 0.3701, + "step": 4769 + }, + { + "epoch": 0.1049966982170372, + "grad_norm": 1.4748478030808985, + "learning_rate": 9.85323554365741e-06, + "loss": 0.3023, + "step": 4770 + }, + { + "epoch": 0.10501871010345587, + "grad_norm": 1.1658565773388607, + "learning_rate": 9.853149800645088e-06, + "loss": 0.2782, + "step": 4771 + }, + { + "epoch": 0.10504072198987453, + "grad_norm": 0.7560593431901858, + "learning_rate": 9.853064032966895e-06, + "loss": 0.5921, + "step": 4772 + }, + { + "epoch": 0.1050627338762932, + "grad_norm": 1.680189149924243, + "learning_rate": 9.852978240623267e-06, + "loss": 0.2921, + "step": 4773 + }, + { + "epoch": 0.10508474576271186, + "grad_norm": 1.6164119282335987, + "learning_rate": 9.852892423614638e-06, + "loss": 0.3085, + "step": 4774 + }, + { + "epoch": 0.10510675764913054, + "grad_norm": 1.2140006043400045, + "learning_rate": 9.852806581941445e-06, + "loss": 0.284, + "step": 4775 + }, + { + "epoch": 0.1051287695355492, + "grad_norm": 1.921155366149712, + "learning_rate": 9.852720715604126e-06, + "loss": 0.4003, + "step": 4776 + }, + { + "epoch": 0.10515078142196786, + "grad_norm": 2.202840520867312, + "learning_rate": 9.852634824603116e-06, + "loss": 0.3667, + "step": 4777 + }, + { + "epoch": 0.10517279330838653, + "grad_norm": 2.395963767376237, + "learning_rate": 9.85254890893885e-06, + "loss": 0.3415, + "step": 4778 + }, + { + "epoch": 0.10519480519480519, + "grad_norm": 0.679381670946793, + "learning_rate": 9.852462968611767e-06, + "loss": 0.5464, + "step": 4779 + }, + { + "epoch": 0.10521681708122387, + "grad_norm": 2.7302349615821213, + "learning_rate": 9.852377003622302e-06, + "loss": 0.3669, + "step": 4780 + }, + { + "epoch": 0.10523882896764253, + "grad_norm": 2.971858798747269, + "learning_rate": 9.852291013970895e-06, + "loss": 0.4094, + "step": 4781 + }, + { + "epoch": 0.10526084085406119, + "grad_norm": 2.2044319155054364, + "learning_rate": 9.852204999657979e-06, + "loss": 0.4327, + "step": 4782 + }, + { + "epoch": 0.10528285274047985, + "grad_norm": 1.9503629564832412, + "learning_rate": 9.852118960683992e-06, + "loss": 0.3793, + "step": 4783 + }, + { + "epoch": 0.10530486462689853, + "grad_norm": 1.4788915279720378, + "learning_rate": 9.852032897049374e-06, + "loss": 0.4035, + "step": 4784 + }, + { + "epoch": 0.1053268765133172, + "grad_norm": 1.6581655684897294, + "learning_rate": 9.851946808754562e-06, + "loss": 0.3107, + "step": 4785 + }, + { + "epoch": 0.10534888839973586, + "grad_norm": 2.488686416202198, + "learning_rate": 9.85186069579999e-06, + "loss": 0.4291, + "step": 4786 + }, + { + "epoch": 0.10537090028615452, + "grad_norm": 1.3555034862210926, + "learning_rate": 9.8517745581861e-06, + "loss": 0.3159, + "step": 4787 + }, + { + "epoch": 0.10539291217257318, + "grad_norm": 0.735141011347643, + "learning_rate": 9.851688395913326e-06, + "loss": 0.5974, + "step": 4788 + }, + { + "epoch": 0.10541492405899186, + "grad_norm": 1.278846890832756, + "learning_rate": 9.851602208982108e-06, + "loss": 0.2923, + "step": 4789 + }, + { + "epoch": 0.10543693594541052, + "grad_norm": 1.6150106212276394, + "learning_rate": 9.851515997392884e-06, + "loss": 0.3716, + "step": 4790 + }, + { + "epoch": 0.10545894783182919, + "grad_norm": 0.662195329152308, + "learning_rate": 9.851429761146091e-06, + "loss": 0.539, + "step": 4791 + }, + { + "epoch": 0.10548095971824785, + "grad_norm": 9.06562822340073, + "learning_rate": 9.851343500242169e-06, + "loss": 0.2889, + "step": 4792 + }, + { + "epoch": 0.10550297160466653, + "grad_norm": 0.6621495120666268, + "learning_rate": 9.851257214681557e-06, + "loss": 0.5895, + "step": 4793 + }, + { + "epoch": 0.10552498349108519, + "grad_norm": 1.2980209319462206, + "learning_rate": 9.851170904464691e-06, + "loss": 0.3295, + "step": 4794 + }, + { + "epoch": 0.10554699537750385, + "grad_norm": 1.5106545741908102, + "learning_rate": 9.851084569592009e-06, + "loss": 0.4961, + "step": 4795 + }, + { + "epoch": 0.10556900726392252, + "grad_norm": 1.7759126728267836, + "learning_rate": 9.850998210063953e-06, + "loss": 0.4056, + "step": 4796 + }, + { + "epoch": 0.10559101915034118, + "grad_norm": 1.3720230601556935, + "learning_rate": 9.850911825880962e-06, + "loss": 0.318, + "step": 4797 + }, + { + "epoch": 0.10561303103675986, + "grad_norm": 1.6080673836340287, + "learning_rate": 9.850825417043472e-06, + "loss": 0.3873, + "step": 4798 + }, + { + "epoch": 0.10563504292317852, + "grad_norm": 3.014761041384984, + "learning_rate": 9.850738983551923e-06, + "loss": 0.3923, + "step": 4799 + }, + { + "epoch": 0.10565705480959718, + "grad_norm": 2.064622701088485, + "learning_rate": 9.850652525406757e-06, + "loss": 0.468, + "step": 4800 + }, + { + "epoch": 0.10567906669601584, + "grad_norm": 1.8632382553161075, + "learning_rate": 9.850566042608409e-06, + "loss": 0.3955, + "step": 4801 + }, + { + "epoch": 0.10570107858243452, + "grad_norm": 5.259218287922594, + "learning_rate": 9.850479535157323e-06, + "loss": 0.2933, + "step": 4802 + }, + { + "epoch": 0.10572309046885318, + "grad_norm": 1.9865608227449743, + "learning_rate": 9.850393003053936e-06, + "loss": 0.4082, + "step": 4803 + }, + { + "epoch": 0.10574510235527185, + "grad_norm": 1.8048749009162168, + "learning_rate": 9.850306446298688e-06, + "loss": 0.4241, + "step": 4804 + }, + { + "epoch": 0.10576711424169051, + "grad_norm": 1.6851782250641576, + "learning_rate": 9.850219864892018e-06, + "loss": 0.3272, + "step": 4805 + }, + { + "epoch": 0.10578912612810917, + "grad_norm": 1.9451652327929774, + "learning_rate": 9.850133258834369e-06, + "loss": 0.3424, + "step": 4806 + }, + { + "epoch": 0.10581113801452785, + "grad_norm": 0.738005647832766, + "learning_rate": 9.850046628126179e-06, + "loss": 0.5488, + "step": 4807 + }, + { + "epoch": 0.10583314990094651, + "grad_norm": 1.7323229557850925, + "learning_rate": 9.84995997276789e-06, + "loss": 0.3536, + "step": 4808 + }, + { + "epoch": 0.10585516178736518, + "grad_norm": 1.5735776692618395, + "learning_rate": 9.84987329275994e-06, + "loss": 0.3733, + "step": 4809 + }, + { + "epoch": 0.10587717367378384, + "grad_norm": 2.1837095568244504, + "learning_rate": 9.849786588102772e-06, + "loss": 0.3903, + "step": 4810 + }, + { + "epoch": 0.10589918556020252, + "grad_norm": 0.6572324760701718, + "learning_rate": 9.849699858796826e-06, + "loss": 0.5541, + "step": 4811 + }, + { + "epoch": 0.10592119744662118, + "grad_norm": 1.740852662403551, + "learning_rate": 9.84961310484254e-06, + "loss": 0.3813, + "step": 4812 + }, + { + "epoch": 0.10594320933303984, + "grad_norm": 2.2130644646835034, + "learning_rate": 9.84952632624036e-06, + "loss": 0.3783, + "step": 4813 + }, + { + "epoch": 0.1059652212194585, + "grad_norm": 1.792820051524619, + "learning_rate": 9.849439522990721e-06, + "loss": 0.4291, + "step": 4814 + }, + { + "epoch": 0.10598723310587717, + "grad_norm": 3.6596200826426015, + "learning_rate": 9.849352695094069e-06, + "loss": 0.4451, + "step": 4815 + }, + { + "epoch": 0.10600924499229585, + "grad_norm": 1.3652078981247608, + "learning_rate": 9.849265842550844e-06, + "loss": 0.3397, + "step": 4816 + }, + { + "epoch": 0.10603125687871451, + "grad_norm": 1.6682915816086303, + "learning_rate": 9.849178965361489e-06, + "loss": 0.3559, + "step": 4817 + }, + { + "epoch": 0.10605326876513317, + "grad_norm": 2.231380225098674, + "learning_rate": 9.84909206352644e-06, + "loss": 0.3672, + "step": 4818 + }, + { + "epoch": 0.10607528065155183, + "grad_norm": 1.5545993351614422, + "learning_rate": 9.849005137046143e-06, + "loss": 0.3655, + "step": 4819 + }, + { + "epoch": 0.1060972925379705, + "grad_norm": 1.3693534418293019, + "learning_rate": 9.84891818592104e-06, + "loss": 0.3455, + "step": 4820 + }, + { + "epoch": 0.10611930442438917, + "grad_norm": 1.427019951177354, + "learning_rate": 9.848831210151574e-06, + "loss": 0.3862, + "step": 4821 + }, + { + "epoch": 0.10614131631080784, + "grad_norm": 0.768083642411944, + "learning_rate": 9.848744209738183e-06, + "loss": 0.5779, + "step": 4822 + }, + { + "epoch": 0.1061633281972265, + "grad_norm": 1.5994280551765083, + "learning_rate": 9.848657184681311e-06, + "loss": 0.4302, + "step": 4823 + }, + { + "epoch": 0.10618534008364516, + "grad_norm": 0.6762909530984255, + "learning_rate": 9.848570134981404e-06, + "loss": 0.5402, + "step": 4824 + }, + { + "epoch": 0.10620735197006384, + "grad_norm": 1.5918064686986102, + "learning_rate": 9.848483060638896e-06, + "loss": 0.3397, + "step": 4825 + }, + { + "epoch": 0.1062293638564825, + "grad_norm": 1.0731234161050862, + "learning_rate": 9.848395961654239e-06, + "loss": 0.339, + "step": 4826 + }, + { + "epoch": 0.10625137574290117, + "grad_norm": 1.290906669209281, + "learning_rate": 9.84830883802787e-06, + "loss": 0.3339, + "step": 4827 + }, + { + "epoch": 0.10627338762931983, + "grad_norm": 1.4260069036631815, + "learning_rate": 9.848221689760231e-06, + "loss": 0.3673, + "step": 4828 + }, + { + "epoch": 0.10629539951573849, + "grad_norm": 0.9356576993223741, + "learning_rate": 9.848134516851769e-06, + "loss": 0.5853, + "step": 4829 + }, + { + "epoch": 0.10631741140215717, + "grad_norm": 0.8671280777557224, + "learning_rate": 9.848047319302925e-06, + "loss": 0.5915, + "step": 4830 + }, + { + "epoch": 0.10633942328857583, + "grad_norm": 2.1692877338250884, + "learning_rate": 9.84796009711414e-06, + "loss": 0.3102, + "step": 4831 + }, + { + "epoch": 0.1063614351749945, + "grad_norm": 0.6445827710203142, + "learning_rate": 9.847872850285862e-06, + "loss": 0.5614, + "step": 4832 + }, + { + "epoch": 0.10638344706141316, + "grad_norm": 1.511466675546201, + "learning_rate": 9.847785578818531e-06, + "loss": 0.3813, + "step": 4833 + }, + { + "epoch": 0.10640545894783184, + "grad_norm": 0.8024811072832938, + "learning_rate": 9.847698282712593e-06, + "loss": 0.5797, + "step": 4834 + }, + { + "epoch": 0.1064274708342505, + "grad_norm": 1.5265030136631625, + "learning_rate": 9.847610961968487e-06, + "loss": 0.4537, + "step": 4835 + }, + { + "epoch": 0.10644948272066916, + "grad_norm": 0.755227122799156, + "learning_rate": 9.847523616586663e-06, + "loss": 0.5772, + "step": 4836 + }, + { + "epoch": 0.10647149460708782, + "grad_norm": 1.6544729976083508, + "learning_rate": 9.84743624656756e-06, + "loss": 0.3643, + "step": 4837 + }, + { + "epoch": 0.10649350649350649, + "grad_norm": 1.579101153367562, + "learning_rate": 9.847348851911626e-06, + "loss": 0.2942, + "step": 4838 + }, + { + "epoch": 0.10651551837992516, + "grad_norm": 3.074779186021604, + "learning_rate": 9.847261432619301e-06, + "loss": 0.4161, + "step": 4839 + }, + { + "epoch": 0.10653753026634383, + "grad_norm": 1.6184394094790142, + "learning_rate": 9.847173988691033e-06, + "loss": 0.2982, + "step": 4840 + }, + { + "epoch": 0.10655954215276249, + "grad_norm": 0.7021562387831104, + "learning_rate": 9.847086520127265e-06, + "loss": 0.5786, + "step": 4841 + }, + { + "epoch": 0.10658155403918115, + "grad_norm": 2.1077769003261726, + "learning_rate": 9.84699902692844e-06, + "loss": 0.4047, + "step": 4842 + }, + { + "epoch": 0.10660356592559983, + "grad_norm": 1.6079885817772, + "learning_rate": 9.846911509095005e-06, + "loss": 0.4206, + "step": 4843 + }, + { + "epoch": 0.1066255778120185, + "grad_norm": 0.6568698961463321, + "learning_rate": 9.846823966627404e-06, + "loss": 0.5645, + "step": 4844 + }, + { + "epoch": 0.10664758969843716, + "grad_norm": 0.6273148263309125, + "learning_rate": 9.846736399526082e-06, + "loss": 0.559, + "step": 4845 + }, + { + "epoch": 0.10666960158485582, + "grad_norm": 1.6933164130505247, + "learning_rate": 9.846648807791484e-06, + "loss": 0.3561, + "step": 4846 + }, + { + "epoch": 0.10669161347127448, + "grad_norm": 1.534460283163443, + "learning_rate": 9.846561191424054e-06, + "loss": 0.3993, + "step": 4847 + }, + { + "epoch": 0.10671362535769316, + "grad_norm": 1.476169173313852, + "learning_rate": 9.846473550424241e-06, + "loss": 0.3871, + "step": 4848 + }, + { + "epoch": 0.10673563724411182, + "grad_norm": 1.886299596387814, + "learning_rate": 9.846385884792484e-06, + "loss": 0.3805, + "step": 4849 + }, + { + "epoch": 0.10675764913053049, + "grad_norm": 2.30073398264363, + "learning_rate": 9.846298194529236e-06, + "loss": 0.4088, + "step": 4850 + }, + { + "epoch": 0.10677966101694915, + "grad_norm": 1.7058716857409089, + "learning_rate": 9.846210479634937e-06, + "loss": 0.3433, + "step": 4851 + }, + { + "epoch": 0.10680167290336783, + "grad_norm": 2.89994936840609, + "learning_rate": 9.846122740110034e-06, + "loss": 0.4098, + "step": 4852 + }, + { + "epoch": 0.10682368478978649, + "grad_norm": 18.633423403601608, + "learning_rate": 9.846034975954977e-06, + "loss": 0.3975, + "step": 4853 + }, + { + "epoch": 0.10684569667620515, + "grad_norm": 1.4722052324282955, + "learning_rate": 9.845947187170206e-06, + "loss": 0.3289, + "step": 4854 + }, + { + "epoch": 0.10686770856262381, + "grad_norm": 1.8578831298594238, + "learning_rate": 9.84585937375617e-06, + "loss": 0.3059, + "step": 4855 + }, + { + "epoch": 0.10688972044904248, + "grad_norm": 1.72017785325137, + "learning_rate": 9.845771535713318e-06, + "loss": 0.4335, + "step": 4856 + }, + { + "epoch": 0.10691173233546115, + "grad_norm": 2.128748955868515, + "learning_rate": 9.845683673042092e-06, + "loss": 0.4597, + "step": 4857 + }, + { + "epoch": 0.10693374422187982, + "grad_norm": 1.6684965336090694, + "learning_rate": 9.84559578574294e-06, + "loss": 0.2886, + "step": 4858 + }, + { + "epoch": 0.10695575610829848, + "grad_norm": 1.4853441035310582, + "learning_rate": 9.845507873816309e-06, + "loss": 0.3881, + "step": 4859 + }, + { + "epoch": 0.10697776799471714, + "grad_norm": 1.4823720027883933, + "learning_rate": 9.845419937262645e-06, + "loss": 0.3703, + "step": 4860 + }, + { + "epoch": 0.1069997798811358, + "grad_norm": 2.3078909249659487, + "learning_rate": 9.845331976082398e-06, + "loss": 0.4054, + "step": 4861 + }, + { + "epoch": 0.10702179176755448, + "grad_norm": 3.7039150704649826, + "learning_rate": 9.845243990276012e-06, + "loss": 0.3751, + "step": 4862 + }, + { + "epoch": 0.10704380365397315, + "grad_norm": 0.8549182645994936, + "learning_rate": 9.845155979843936e-06, + "loss": 0.5873, + "step": 4863 + }, + { + "epoch": 0.10706581554039181, + "grad_norm": 1.8407964776441181, + "learning_rate": 9.845067944786616e-06, + "loss": 0.4511, + "step": 4864 + }, + { + "epoch": 0.10708782742681047, + "grad_norm": 3.5862210952569766, + "learning_rate": 9.844979885104497e-06, + "loss": 0.4085, + "step": 4865 + }, + { + "epoch": 0.10710983931322915, + "grad_norm": 1.6229841599881296, + "learning_rate": 9.844891800798032e-06, + "loss": 0.4618, + "step": 4866 + }, + { + "epoch": 0.10713185119964781, + "grad_norm": 0.6510778102510059, + "learning_rate": 9.844803691867666e-06, + "loss": 0.5512, + "step": 4867 + }, + { + "epoch": 0.10715386308606648, + "grad_norm": 1.836682719739985, + "learning_rate": 9.844715558313846e-06, + "loss": 0.3842, + "step": 4868 + }, + { + "epoch": 0.10717587497248514, + "grad_norm": 1.8742554556764446, + "learning_rate": 9.84462740013702e-06, + "loss": 0.4458, + "step": 4869 + }, + { + "epoch": 0.1071978868589038, + "grad_norm": 2.300835047889961, + "learning_rate": 9.844539217337638e-06, + "loss": 0.3668, + "step": 4870 + }, + { + "epoch": 0.10721989874532248, + "grad_norm": 1.269334065685149, + "learning_rate": 9.844451009916147e-06, + "loss": 0.3326, + "step": 4871 + }, + { + "epoch": 0.10724191063174114, + "grad_norm": 1.8834808575515072, + "learning_rate": 9.844362777872993e-06, + "loss": 0.3577, + "step": 4872 + }, + { + "epoch": 0.1072639225181598, + "grad_norm": 1.9583671522292634, + "learning_rate": 9.84427452120863e-06, + "loss": 0.4187, + "step": 4873 + }, + { + "epoch": 0.10728593440457847, + "grad_norm": 1.917917301756045, + "learning_rate": 9.844186239923502e-06, + "loss": 0.3558, + "step": 4874 + }, + { + "epoch": 0.10730794629099714, + "grad_norm": 1.7289390636871689, + "learning_rate": 9.844097934018057e-06, + "loss": 0.3979, + "step": 4875 + }, + { + "epoch": 0.10732995817741581, + "grad_norm": 0.7358139232720843, + "learning_rate": 9.844009603492747e-06, + "loss": 0.5581, + "step": 4876 + }, + { + "epoch": 0.10735197006383447, + "grad_norm": 1.6503886739096982, + "learning_rate": 9.843921248348022e-06, + "loss": 0.3758, + "step": 4877 + }, + { + "epoch": 0.10737398195025313, + "grad_norm": 1.7428151412049109, + "learning_rate": 9.843832868584326e-06, + "loss": 0.4366, + "step": 4878 + }, + { + "epoch": 0.1073959938366718, + "grad_norm": 1.4775164498866862, + "learning_rate": 9.843744464202113e-06, + "loss": 0.3715, + "step": 4879 + }, + { + "epoch": 0.10741800572309047, + "grad_norm": 1.6938130871251604, + "learning_rate": 9.84365603520183e-06, + "loss": 0.4239, + "step": 4880 + }, + { + "epoch": 0.10744001760950914, + "grad_norm": 0.651031315961643, + "learning_rate": 9.843567581583924e-06, + "loss": 0.5621, + "step": 4881 + }, + { + "epoch": 0.1074620294959278, + "grad_norm": 1.4335071549602048, + "learning_rate": 9.84347910334885e-06, + "loss": 0.3148, + "step": 4882 + }, + { + "epoch": 0.10748404138234646, + "grad_norm": 1.286351824100802, + "learning_rate": 9.843390600497054e-06, + "loss": 0.1923, + "step": 4883 + }, + { + "epoch": 0.10750605326876514, + "grad_norm": 2.6010216783100555, + "learning_rate": 9.843302073028987e-06, + "loss": 0.3764, + "step": 4884 + }, + { + "epoch": 0.1075280651551838, + "grad_norm": 1.5626359967383687, + "learning_rate": 9.843213520945098e-06, + "loss": 0.3292, + "step": 4885 + }, + { + "epoch": 0.10755007704160247, + "grad_norm": 1.5298111829216199, + "learning_rate": 9.843124944245839e-06, + "loss": 0.3194, + "step": 4886 + }, + { + "epoch": 0.10757208892802113, + "grad_norm": 1.7586457229698287, + "learning_rate": 9.843036342931657e-06, + "loss": 0.4248, + "step": 4887 + }, + { + "epoch": 0.10759410081443979, + "grad_norm": 1.7399042359120906, + "learning_rate": 9.842947717003006e-06, + "loss": 0.4003, + "step": 4888 + }, + { + "epoch": 0.10761611270085847, + "grad_norm": 2.172683806915401, + "learning_rate": 9.842859066460336e-06, + "loss": 0.3448, + "step": 4889 + }, + { + "epoch": 0.10763812458727713, + "grad_norm": 1.5290770497531523, + "learning_rate": 9.842770391304093e-06, + "loss": 0.2868, + "step": 4890 + }, + { + "epoch": 0.1076601364736958, + "grad_norm": 0.7123545826641868, + "learning_rate": 9.842681691534731e-06, + "loss": 0.5601, + "step": 4891 + }, + { + "epoch": 0.10768214836011446, + "grad_norm": 1.5519606742677285, + "learning_rate": 9.842592967152705e-06, + "loss": 0.4307, + "step": 4892 + }, + { + "epoch": 0.10770416024653313, + "grad_norm": 46.15324657160236, + "learning_rate": 9.842504218158457e-06, + "loss": 0.3572, + "step": 4893 + }, + { + "epoch": 0.1077261721329518, + "grad_norm": 1.7419218732627453, + "learning_rate": 9.842415444552444e-06, + "loss": 0.3751, + "step": 4894 + }, + { + "epoch": 0.10774818401937046, + "grad_norm": 2.033996229170613, + "learning_rate": 9.842326646335118e-06, + "loss": 0.5012, + "step": 4895 + }, + { + "epoch": 0.10777019590578912, + "grad_norm": 1.2290018690077011, + "learning_rate": 9.842237823506926e-06, + "loss": 0.2429, + "step": 4896 + }, + { + "epoch": 0.10779220779220779, + "grad_norm": 2.030871733663118, + "learning_rate": 9.842148976068323e-06, + "loss": 0.3857, + "step": 4897 + }, + { + "epoch": 0.10781421967862646, + "grad_norm": 1.705623953798263, + "learning_rate": 9.842060104019758e-06, + "loss": 0.3498, + "step": 4898 + }, + { + "epoch": 0.10783623156504513, + "grad_norm": 1.8454372669660715, + "learning_rate": 9.841971207361684e-06, + "loss": 0.3641, + "step": 4899 + }, + { + "epoch": 0.10785824345146379, + "grad_norm": 2.0692608709110556, + "learning_rate": 9.841882286094555e-06, + "loss": 0.2959, + "step": 4900 + }, + { + "epoch": 0.10788025533788245, + "grad_norm": 1.66912560682362, + "learning_rate": 9.841793340218819e-06, + "loss": 0.2859, + "step": 4901 + }, + { + "epoch": 0.10790226722430113, + "grad_norm": 0.7788562919926596, + "learning_rate": 9.841704369734928e-06, + "loss": 0.5465, + "step": 4902 + }, + { + "epoch": 0.10792427911071979, + "grad_norm": 0.6799238216230938, + "learning_rate": 9.841615374643338e-06, + "loss": 0.5479, + "step": 4903 + }, + { + "epoch": 0.10794629099713846, + "grad_norm": 2.6800509062556688, + "learning_rate": 9.8415263549445e-06, + "loss": 0.3618, + "step": 4904 + }, + { + "epoch": 0.10796830288355712, + "grad_norm": 1.6395697558049085, + "learning_rate": 9.841437310638864e-06, + "loss": 0.3357, + "step": 4905 + }, + { + "epoch": 0.10799031476997578, + "grad_norm": 1.5580239544500991, + "learning_rate": 9.841348241726886e-06, + "loss": 0.441, + "step": 4906 + }, + { + "epoch": 0.10801232665639446, + "grad_norm": 2.1837220998090103, + "learning_rate": 9.841259148209015e-06, + "loss": 0.4168, + "step": 4907 + }, + { + "epoch": 0.10803433854281312, + "grad_norm": 2.2177050339284525, + "learning_rate": 9.841170030085707e-06, + "loss": 0.4647, + "step": 4908 + }, + { + "epoch": 0.10805635042923178, + "grad_norm": 1.5643240225185429, + "learning_rate": 9.841080887357413e-06, + "loss": 0.3857, + "step": 4909 + }, + { + "epoch": 0.10807836231565045, + "grad_norm": 1.7717158079967388, + "learning_rate": 9.840991720024587e-06, + "loss": 0.406, + "step": 4910 + }, + { + "epoch": 0.10810037420206911, + "grad_norm": 1.0456717146701444, + "learning_rate": 9.840902528087681e-06, + "loss": 0.5932, + "step": 4911 + }, + { + "epoch": 0.10812238608848779, + "grad_norm": 1.5968862489438558, + "learning_rate": 9.84081331154715e-06, + "loss": 0.3475, + "step": 4912 + }, + { + "epoch": 0.10814439797490645, + "grad_norm": 1.9091384699485914, + "learning_rate": 9.840724070403448e-06, + "loss": 0.4399, + "step": 4913 + }, + { + "epoch": 0.10816640986132511, + "grad_norm": 2.3411426123637886, + "learning_rate": 9.840634804657026e-06, + "loss": 0.4581, + "step": 4914 + }, + { + "epoch": 0.10818842174774378, + "grad_norm": 1.6645633469836645, + "learning_rate": 9.84054551430834e-06, + "loss": 0.3545, + "step": 4915 + }, + { + "epoch": 0.10821043363416245, + "grad_norm": 1.5080434275624333, + "learning_rate": 9.840456199357842e-06, + "loss": 0.3672, + "step": 4916 + }, + { + "epoch": 0.10823244552058112, + "grad_norm": 0.7062631884592271, + "learning_rate": 9.840366859805986e-06, + "loss": 0.5504, + "step": 4917 + }, + { + "epoch": 0.10825445740699978, + "grad_norm": 2.065037088234113, + "learning_rate": 9.840277495653227e-06, + "loss": 0.4192, + "step": 4918 + }, + { + "epoch": 0.10827646929341844, + "grad_norm": 0.6947067898342881, + "learning_rate": 9.84018810690002e-06, + "loss": 0.5676, + "step": 4919 + }, + { + "epoch": 0.1082984811798371, + "grad_norm": 0.6487544065955344, + "learning_rate": 9.840098693546815e-06, + "loss": 0.5624, + "step": 4920 + }, + { + "epoch": 0.10832049306625578, + "grad_norm": 1.458009128590888, + "learning_rate": 9.840009255594074e-06, + "loss": 0.3841, + "step": 4921 + }, + { + "epoch": 0.10834250495267445, + "grad_norm": 1.6459830908161677, + "learning_rate": 9.839919793042245e-06, + "loss": 0.418, + "step": 4922 + }, + { + "epoch": 0.10836451683909311, + "grad_norm": 0.7418248409306836, + "learning_rate": 9.839830305891785e-06, + "loss": 0.566, + "step": 4923 + }, + { + "epoch": 0.10838652872551177, + "grad_norm": 0.7031892732214498, + "learning_rate": 9.839740794143149e-06, + "loss": 0.5528, + "step": 4924 + }, + { + "epoch": 0.10840854061193045, + "grad_norm": 0.6561950495476512, + "learning_rate": 9.839651257796793e-06, + "loss": 0.5457, + "step": 4925 + }, + { + "epoch": 0.10843055249834911, + "grad_norm": 2.2688860665284714, + "learning_rate": 9.839561696853169e-06, + "loss": 0.433, + "step": 4926 + }, + { + "epoch": 0.10845256438476777, + "grad_norm": 1.8314442275875293, + "learning_rate": 9.839472111312734e-06, + "loss": 0.4585, + "step": 4927 + }, + { + "epoch": 0.10847457627118644, + "grad_norm": 1.7323748934544696, + "learning_rate": 9.839382501175944e-06, + "loss": 0.3203, + "step": 4928 + }, + { + "epoch": 0.1084965881576051, + "grad_norm": 1.5253374806176248, + "learning_rate": 9.839292866443253e-06, + "loss": 0.3813, + "step": 4929 + }, + { + "epoch": 0.10851860004402378, + "grad_norm": 11.737213794529538, + "learning_rate": 9.839203207115118e-06, + "loss": 0.3943, + "step": 4930 + }, + { + "epoch": 0.10854061193044244, + "grad_norm": 1.897689767233536, + "learning_rate": 9.839113523191995e-06, + "loss": 0.3808, + "step": 4931 + }, + { + "epoch": 0.1085626238168611, + "grad_norm": 2.201837415656009, + "learning_rate": 9.839023814674337e-06, + "loss": 0.4282, + "step": 4932 + }, + { + "epoch": 0.10858463570327977, + "grad_norm": 1.180252574970148, + "learning_rate": 9.838934081562601e-06, + "loss": 0.5555, + "step": 4933 + }, + { + "epoch": 0.10860664758969844, + "grad_norm": 1.9754126703613233, + "learning_rate": 9.838844323857247e-06, + "loss": 0.3738, + "step": 4934 + }, + { + "epoch": 0.1086286594761171, + "grad_norm": 0.7740092089691001, + "learning_rate": 9.838754541558725e-06, + "loss": 0.5681, + "step": 4935 + }, + { + "epoch": 0.10865067136253577, + "grad_norm": 0.6332843079428402, + "learning_rate": 9.838664734667496e-06, + "loss": 0.5529, + "step": 4936 + }, + { + "epoch": 0.10867268324895443, + "grad_norm": 2.7192168344724235, + "learning_rate": 9.838574903184013e-06, + "loss": 0.3841, + "step": 4937 + }, + { + "epoch": 0.1086946951353731, + "grad_norm": 1.7584760055463253, + "learning_rate": 9.838485047108735e-06, + "loss": 0.3735, + "step": 4938 + }, + { + "epoch": 0.10871670702179177, + "grad_norm": 1.6399963963653175, + "learning_rate": 9.838395166442118e-06, + "loss": 0.421, + "step": 4939 + }, + { + "epoch": 0.10873871890821044, + "grad_norm": 1.1911305820304383, + "learning_rate": 9.838305261184619e-06, + "loss": 0.5727, + "step": 4940 + }, + { + "epoch": 0.1087607307946291, + "grad_norm": 1.857999640210833, + "learning_rate": 9.838215331336696e-06, + "loss": 0.39, + "step": 4941 + }, + { + "epoch": 0.10878274268104776, + "grad_norm": 1.8442302150767247, + "learning_rate": 9.838125376898802e-06, + "loss": 0.3249, + "step": 4942 + }, + { + "epoch": 0.10880475456746644, + "grad_norm": 2.813028588505112, + "learning_rate": 9.8380353978714e-06, + "loss": 0.4399, + "step": 4943 + }, + { + "epoch": 0.1088267664538851, + "grad_norm": 1.7582389534041623, + "learning_rate": 9.837945394254942e-06, + "loss": 0.3365, + "step": 4944 + }, + { + "epoch": 0.10884877834030376, + "grad_norm": 1.8217971815216618, + "learning_rate": 9.837855366049887e-06, + "loss": 0.3262, + "step": 4945 + }, + { + "epoch": 0.10887079022672243, + "grad_norm": 1.7552613864387212, + "learning_rate": 9.837765313256696e-06, + "loss": 0.3935, + "step": 4946 + }, + { + "epoch": 0.10889280211314109, + "grad_norm": 2.1598249973637773, + "learning_rate": 9.837675235875822e-06, + "loss": 0.3917, + "step": 4947 + }, + { + "epoch": 0.10891481399955977, + "grad_norm": 1.5162265389733984, + "learning_rate": 9.837585133907726e-06, + "loss": 0.4545, + "step": 4948 + }, + { + "epoch": 0.10893682588597843, + "grad_norm": 3.9498122620594702, + "learning_rate": 9.837495007352864e-06, + "loss": 0.3516, + "step": 4949 + }, + { + "epoch": 0.1089588377723971, + "grad_norm": 2.3251763181049436, + "learning_rate": 9.837404856211695e-06, + "loss": 0.3487, + "step": 4950 + }, + { + "epoch": 0.10898084965881576, + "grad_norm": 1.5312575991297137, + "learning_rate": 9.837314680484677e-06, + "loss": 0.375, + "step": 4951 + }, + { + "epoch": 0.10900286154523443, + "grad_norm": 0.7614273441537556, + "learning_rate": 9.837224480172268e-06, + "loss": 0.5633, + "step": 4952 + }, + { + "epoch": 0.1090248734316531, + "grad_norm": 2.289098872153149, + "learning_rate": 9.837134255274928e-06, + "loss": 0.4894, + "step": 4953 + }, + { + "epoch": 0.10904688531807176, + "grad_norm": 1.6839375188857852, + "learning_rate": 9.837044005793112e-06, + "loss": 0.4906, + "step": 4954 + }, + { + "epoch": 0.10906889720449042, + "grad_norm": 0.6998269373459133, + "learning_rate": 9.836953731727283e-06, + "loss": 0.5728, + "step": 4955 + }, + { + "epoch": 0.10909090909090909, + "grad_norm": 2.098152986471452, + "learning_rate": 9.836863433077896e-06, + "loss": 0.4268, + "step": 4956 + }, + { + "epoch": 0.10911292097732776, + "grad_norm": 0.6546078569174687, + "learning_rate": 9.836773109845413e-06, + "loss": 0.5614, + "step": 4957 + }, + { + "epoch": 0.10913493286374643, + "grad_norm": 2.3412188029985406, + "learning_rate": 9.836682762030292e-06, + "loss": 0.4709, + "step": 4958 + }, + { + "epoch": 0.10915694475016509, + "grad_norm": 2.566403182779939, + "learning_rate": 9.836592389632992e-06, + "loss": 0.3615, + "step": 4959 + }, + { + "epoch": 0.10917895663658375, + "grad_norm": 1.5177102389702293, + "learning_rate": 9.836501992653971e-06, + "loss": 0.3239, + "step": 4960 + }, + { + "epoch": 0.10920096852300241, + "grad_norm": 0.6868398626898102, + "learning_rate": 9.836411571093692e-06, + "loss": 0.5329, + "step": 4961 + }, + { + "epoch": 0.10922298040942109, + "grad_norm": 2.2065342520462945, + "learning_rate": 9.83632112495261e-06, + "loss": 0.3652, + "step": 4962 + }, + { + "epoch": 0.10924499229583975, + "grad_norm": 1.6471816809936042, + "learning_rate": 9.836230654231189e-06, + "loss": 0.3881, + "step": 4963 + }, + { + "epoch": 0.10926700418225842, + "grad_norm": 1.6064692684862822, + "learning_rate": 9.836140158929886e-06, + "loss": 0.3601, + "step": 4964 + }, + { + "epoch": 0.10928901606867708, + "grad_norm": 1.8504806230917155, + "learning_rate": 9.83604963904916e-06, + "loss": 0.3536, + "step": 4965 + }, + { + "epoch": 0.10931102795509576, + "grad_norm": 0.6930505031952963, + "learning_rate": 9.835959094589474e-06, + "loss": 0.5212, + "step": 4966 + }, + { + "epoch": 0.10933303984151442, + "grad_norm": 1.8364526244379573, + "learning_rate": 9.835868525551288e-06, + "loss": 0.3889, + "step": 4967 + }, + { + "epoch": 0.10935505172793308, + "grad_norm": 1.839461923858242, + "learning_rate": 9.83577793193506e-06, + "loss": 0.3034, + "step": 4968 + }, + { + "epoch": 0.10937706361435175, + "grad_norm": 1.7591757738218219, + "learning_rate": 9.835687313741252e-06, + "loss": 0.3904, + "step": 4969 + }, + { + "epoch": 0.10939907550077041, + "grad_norm": 1.8976883375034193, + "learning_rate": 9.835596670970324e-06, + "loss": 0.2838, + "step": 4970 + }, + { + "epoch": 0.10942108738718909, + "grad_norm": 1.590312990730291, + "learning_rate": 9.835506003622739e-06, + "loss": 0.3423, + "step": 4971 + }, + { + "epoch": 0.10944309927360775, + "grad_norm": 2.004958347670583, + "learning_rate": 9.835415311698953e-06, + "loss": 0.3909, + "step": 4972 + }, + { + "epoch": 0.10946511116002641, + "grad_norm": 1.7650914659283758, + "learning_rate": 9.835324595199431e-06, + "loss": 0.3534, + "step": 4973 + }, + { + "epoch": 0.10948712304644508, + "grad_norm": 3.035713956626387, + "learning_rate": 9.835233854124633e-06, + "loss": 0.4188, + "step": 4974 + }, + { + "epoch": 0.10950913493286375, + "grad_norm": 2.320206893124125, + "learning_rate": 9.83514308847502e-06, + "loss": 0.3748, + "step": 4975 + }, + { + "epoch": 0.10953114681928242, + "grad_norm": 1.5377667671993818, + "learning_rate": 9.835052298251053e-06, + "loss": 0.3177, + "step": 4976 + }, + { + "epoch": 0.10955315870570108, + "grad_norm": 1.4135847669919923, + "learning_rate": 9.834961483453194e-06, + "loss": 0.3785, + "step": 4977 + }, + { + "epoch": 0.10957517059211974, + "grad_norm": 2.142697486004674, + "learning_rate": 9.834870644081904e-06, + "loss": 0.3272, + "step": 4978 + }, + { + "epoch": 0.1095971824785384, + "grad_norm": 2.345746662349815, + "learning_rate": 9.834779780137645e-06, + "loss": 0.3088, + "step": 4979 + }, + { + "epoch": 0.10961919436495708, + "grad_norm": 1.747527712792351, + "learning_rate": 9.834688891620878e-06, + "loss": 0.3506, + "step": 4980 + }, + { + "epoch": 0.10964120625137574, + "grad_norm": 1.7245446541495786, + "learning_rate": 9.834597978532067e-06, + "loss": 0.4145, + "step": 4981 + }, + { + "epoch": 0.10966321813779441, + "grad_norm": 1.3848027637425464, + "learning_rate": 9.834507040871672e-06, + "loss": 0.2648, + "step": 4982 + }, + { + "epoch": 0.10968523002421307, + "grad_norm": 2.0384254208480184, + "learning_rate": 9.834416078640157e-06, + "loss": 0.3534, + "step": 4983 + }, + { + "epoch": 0.10970724191063175, + "grad_norm": 0.7655239317103391, + "learning_rate": 9.834325091837982e-06, + "loss": 0.5873, + "step": 4984 + }, + { + "epoch": 0.10972925379705041, + "grad_norm": 2.554501541280453, + "learning_rate": 9.834234080465612e-06, + "loss": 0.4814, + "step": 4985 + }, + { + "epoch": 0.10975126568346907, + "grad_norm": 2.487510356695542, + "learning_rate": 9.834143044523508e-06, + "loss": 0.3409, + "step": 4986 + }, + { + "epoch": 0.10977327756988774, + "grad_norm": 0.6707195407821878, + "learning_rate": 9.834051984012131e-06, + "loss": 0.5838, + "step": 4987 + }, + { + "epoch": 0.1097952894563064, + "grad_norm": 0.6695810916310369, + "learning_rate": 9.833960898931948e-06, + "loss": 0.5632, + "step": 4988 + }, + { + "epoch": 0.10981730134272508, + "grad_norm": 1.48351925909887, + "learning_rate": 9.83386978928342e-06, + "loss": 0.2402, + "step": 4989 + }, + { + "epoch": 0.10983931322914374, + "grad_norm": 1.7242294471032191, + "learning_rate": 9.833778655067008e-06, + "loss": 0.3216, + "step": 4990 + }, + { + "epoch": 0.1098613251155624, + "grad_norm": 0.6498550306901403, + "learning_rate": 9.833687496283177e-06, + "loss": 0.5698, + "step": 4991 + }, + { + "epoch": 0.10988333700198107, + "grad_norm": 1.9174834171543238, + "learning_rate": 9.833596312932391e-06, + "loss": 0.3117, + "step": 4992 + }, + { + "epoch": 0.10990534888839974, + "grad_norm": 1.7975700451657326, + "learning_rate": 9.833505105015113e-06, + "loss": 0.3642, + "step": 4993 + }, + { + "epoch": 0.1099273607748184, + "grad_norm": 1.9034393113551678, + "learning_rate": 9.833413872531807e-06, + "loss": 0.3814, + "step": 4994 + }, + { + "epoch": 0.10994937266123707, + "grad_norm": 1.6486355773174044, + "learning_rate": 9.833322615482935e-06, + "loss": 0.402, + "step": 4995 + }, + { + "epoch": 0.10997138454765573, + "grad_norm": 1.648486605913668, + "learning_rate": 9.83323133386896e-06, + "loss": 0.3736, + "step": 4996 + }, + { + "epoch": 0.1099933964340744, + "grad_norm": 0.7196993902688213, + "learning_rate": 9.833140027690351e-06, + "loss": 0.5788, + "step": 4997 + }, + { + "epoch": 0.11001540832049307, + "grad_norm": 1.3776175068651968, + "learning_rate": 9.833048696947567e-06, + "loss": 0.3136, + "step": 4998 + }, + { + "epoch": 0.11003742020691173, + "grad_norm": 18.58946130271579, + "learning_rate": 9.832957341641074e-06, + "loss": 0.3869, + "step": 4999 + }, + { + "epoch": 0.1100594320933304, + "grad_norm": 2.858970882478387, + "learning_rate": 9.832865961771337e-06, + "loss": 0.3219, + "step": 5000 + }, + { + "epoch": 0.11008144397974906, + "grad_norm": 1.6856153595583345, + "learning_rate": 9.832774557338818e-06, + "loss": 0.3169, + "step": 5001 + }, + { + "epoch": 0.11010345586616772, + "grad_norm": 4.095965800445151, + "learning_rate": 9.832683128343982e-06, + "loss": 0.3529, + "step": 5002 + }, + { + "epoch": 0.1101254677525864, + "grad_norm": 1.447523372188846, + "learning_rate": 9.832591674787298e-06, + "loss": 0.3579, + "step": 5003 + }, + { + "epoch": 0.11014747963900506, + "grad_norm": 1.7090961608538882, + "learning_rate": 9.832500196669229e-06, + "loss": 0.4244, + "step": 5004 + }, + { + "epoch": 0.11016949152542373, + "grad_norm": 1.6595183002692255, + "learning_rate": 9.832408693990234e-06, + "loss": 0.425, + "step": 5005 + }, + { + "epoch": 0.11019150341184239, + "grad_norm": 1.9452212802995457, + "learning_rate": 9.832317166750786e-06, + "loss": 0.3924, + "step": 5006 + }, + { + "epoch": 0.11021351529826107, + "grad_norm": 1.919683215595296, + "learning_rate": 9.832225614951345e-06, + "loss": 0.3546, + "step": 5007 + }, + { + "epoch": 0.11023552718467973, + "grad_norm": 1.4230954818471502, + "learning_rate": 9.83213403859238e-06, + "loss": 0.2988, + "step": 5008 + }, + { + "epoch": 0.11025753907109839, + "grad_norm": 1.6734090442167056, + "learning_rate": 9.832042437674353e-06, + "loss": 0.2839, + "step": 5009 + }, + { + "epoch": 0.11027955095751706, + "grad_norm": 0.7042120685356018, + "learning_rate": 9.831950812197733e-06, + "loss": 0.5569, + "step": 5010 + }, + { + "epoch": 0.11030156284393572, + "grad_norm": 0.6725146551037555, + "learning_rate": 9.831859162162982e-06, + "loss": 0.5623, + "step": 5011 + }, + { + "epoch": 0.1103235747303544, + "grad_norm": 1.7146785608860675, + "learning_rate": 9.83176748757057e-06, + "loss": 0.3891, + "step": 5012 + }, + { + "epoch": 0.11034558661677306, + "grad_norm": 1.5836759289410611, + "learning_rate": 9.831675788420959e-06, + "loss": 0.3346, + "step": 5013 + }, + { + "epoch": 0.11036759850319172, + "grad_norm": 1.762103491821811, + "learning_rate": 9.831584064714615e-06, + "loss": 0.3974, + "step": 5014 + }, + { + "epoch": 0.11038961038961038, + "grad_norm": 3.018012591765383, + "learning_rate": 9.831492316452008e-06, + "loss": 0.467, + "step": 5015 + }, + { + "epoch": 0.11041162227602906, + "grad_norm": 1.599139728839219, + "learning_rate": 9.831400543633603e-06, + "loss": 0.4661, + "step": 5016 + }, + { + "epoch": 0.11043363416244772, + "grad_norm": 1.4143866485814527, + "learning_rate": 9.831308746259862e-06, + "loss": 0.3272, + "step": 5017 + }, + { + "epoch": 0.11045564604886639, + "grad_norm": 1.715642112689873, + "learning_rate": 9.831216924331258e-06, + "loss": 0.2894, + "step": 5018 + }, + { + "epoch": 0.11047765793528505, + "grad_norm": 1.895720637844832, + "learning_rate": 9.831125077848254e-06, + "loss": 0.4331, + "step": 5019 + }, + { + "epoch": 0.11049966982170371, + "grad_norm": 1.407590483284715, + "learning_rate": 9.831033206811319e-06, + "loss": 0.3959, + "step": 5020 + }, + { + "epoch": 0.11052168170812239, + "grad_norm": 1.6816477756049186, + "learning_rate": 9.830941311220916e-06, + "loss": 0.4465, + "step": 5021 + }, + { + "epoch": 0.11054369359454105, + "grad_norm": 0.7680208027674434, + "learning_rate": 9.830849391077517e-06, + "loss": 0.5361, + "step": 5022 + }, + { + "epoch": 0.11056570548095972, + "grad_norm": 1.4740856847334516, + "learning_rate": 9.830757446381584e-06, + "loss": 0.365, + "step": 5023 + }, + { + "epoch": 0.11058771736737838, + "grad_norm": 1.6089522671156755, + "learning_rate": 9.83066547713359e-06, + "loss": 0.4098, + "step": 5024 + }, + { + "epoch": 0.11060972925379706, + "grad_norm": 0.6403731667450522, + "learning_rate": 9.830573483333998e-06, + "loss": 0.5429, + "step": 5025 + }, + { + "epoch": 0.11063174114021572, + "grad_norm": 2.9441708290535873, + "learning_rate": 9.830481464983276e-06, + "loss": 0.2974, + "step": 5026 + }, + { + "epoch": 0.11065375302663438, + "grad_norm": 2.211892333146506, + "learning_rate": 9.830389422081896e-06, + "loss": 0.3343, + "step": 5027 + }, + { + "epoch": 0.11067576491305305, + "grad_norm": 1.8593423738431307, + "learning_rate": 9.83029735463032e-06, + "loss": 0.3654, + "step": 5028 + }, + { + "epoch": 0.11069777679947171, + "grad_norm": 1.9429546998872618, + "learning_rate": 9.83020526262902e-06, + "loss": 0.3879, + "step": 5029 + }, + { + "epoch": 0.11071978868589039, + "grad_norm": 0.7106833950380208, + "learning_rate": 9.830113146078462e-06, + "loss": 0.5564, + "step": 5030 + }, + { + "epoch": 0.11074180057230905, + "grad_norm": 1.993470551453326, + "learning_rate": 9.830021004979114e-06, + "loss": 0.4345, + "step": 5031 + }, + { + "epoch": 0.11076381245872771, + "grad_norm": 2.5004036304678, + "learning_rate": 9.829928839331445e-06, + "loss": 0.3987, + "step": 5032 + }, + { + "epoch": 0.11078582434514637, + "grad_norm": 1.632089985253193, + "learning_rate": 9.829836649135924e-06, + "loss": 0.363, + "step": 5033 + }, + { + "epoch": 0.11080783623156505, + "grad_norm": 0.6398103261765866, + "learning_rate": 9.829744434393017e-06, + "loss": 0.5524, + "step": 5034 + }, + { + "epoch": 0.11082984811798371, + "grad_norm": 1.7782607456120374, + "learning_rate": 9.829652195103198e-06, + "loss": 0.3434, + "step": 5035 + }, + { + "epoch": 0.11085186000440238, + "grad_norm": 1.3124443427442414, + "learning_rate": 9.829559931266931e-06, + "loss": 0.3233, + "step": 5036 + }, + { + "epoch": 0.11087387189082104, + "grad_norm": 3.163349152215003, + "learning_rate": 9.829467642884688e-06, + "loss": 0.4347, + "step": 5037 + }, + { + "epoch": 0.1108958837772397, + "grad_norm": 2.2530869736815906, + "learning_rate": 9.829375329956932e-06, + "loss": 0.4058, + "step": 5038 + }, + { + "epoch": 0.11091789566365838, + "grad_norm": 1.565453858428988, + "learning_rate": 9.82928299248414e-06, + "loss": 0.3398, + "step": 5039 + }, + { + "epoch": 0.11093990755007704, + "grad_norm": 0.6619909529577512, + "learning_rate": 9.829190630466777e-06, + "loss": 0.546, + "step": 5040 + }, + { + "epoch": 0.1109619194364957, + "grad_norm": 1.8321459131176399, + "learning_rate": 9.829098243905314e-06, + "loss": 0.3887, + "step": 5041 + }, + { + "epoch": 0.11098393132291437, + "grad_norm": 1.9938446474121372, + "learning_rate": 9.82900583280022e-06, + "loss": 0.3324, + "step": 5042 + }, + { + "epoch": 0.11100594320933305, + "grad_norm": 1.4803654936409922, + "learning_rate": 9.828913397151964e-06, + "loss": 0.3378, + "step": 5043 + }, + { + "epoch": 0.11102795509575171, + "grad_norm": 1.8763777325680875, + "learning_rate": 9.828820936961015e-06, + "loss": 0.415, + "step": 5044 + }, + { + "epoch": 0.11104996698217037, + "grad_norm": 1.7029161074920904, + "learning_rate": 9.828728452227848e-06, + "loss": 0.3776, + "step": 5045 + }, + { + "epoch": 0.11107197886858904, + "grad_norm": 0.6213654144758995, + "learning_rate": 9.828635942952925e-06, + "loss": 0.5119, + "step": 5046 + }, + { + "epoch": 0.1110939907550077, + "grad_norm": 1.6161061602717357, + "learning_rate": 9.828543409136724e-06, + "loss": 0.3894, + "step": 5047 + }, + { + "epoch": 0.11111600264142638, + "grad_norm": 2.291031446398102, + "learning_rate": 9.82845085077971e-06, + "loss": 0.4973, + "step": 5048 + }, + { + "epoch": 0.11113801452784504, + "grad_norm": 1.6575083976625449, + "learning_rate": 9.828358267882354e-06, + "loss": 0.296, + "step": 5049 + }, + { + "epoch": 0.1111600264142637, + "grad_norm": 1.5271800655477448, + "learning_rate": 9.82826566044513e-06, + "loss": 0.3135, + "step": 5050 + }, + { + "epoch": 0.11118203830068236, + "grad_norm": 0.6399874399647824, + "learning_rate": 9.828173028468506e-06, + "loss": 0.5378, + "step": 5051 + }, + { + "epoch": 0.11120405018710103, + "grad_norm": 2.116496937003946, + "learning_rate": 9.828080371952952e-06, + "loss": 0.3368, + "step": 5052 + }, + { + "epoch": 0.1112260620735197, + "grad_norm": 1.407531522642259, + "learning_rate": 9.827987690898941e-06, + "loss": 0.3234, + "step": 5053 + }, + { + "epoch": 0.11124807395993837, + "grad_norm": 1.4454408294553764, + "learning_rate": 9.827894985306946e-06, + "loss": 0.3499, + "step": 5054 + }, + { + "epoch": 0.11127008584635703, + "grad_norm": 1.725519226856455, + "learning_rate": 9.827802255177432e-06, + "loss": 0.4207, + "step": 5055 + }, + { + "epoch": 0.1112920977327757, + "grad_norm": 1.6127734188643432, + "learning_rate": 9.827709500510873e-06, + "loss": 0.3857, + "step": 5056 + }, + { + "epoch": 0.11131410961919437, + "grad_norm": 1.5680553153528187, + "learning_rate": 9.827616721307743e-06, + "loss": 0.2608, + "step": 5057 + }, + { + "epoch": 0.11133612150561303, + "grad_norm": 1.5215113218636778, + "learning_rate": 9.827523917568511e-06, + "loss": 0.3322, + "step": 5058 + }, + { + "epoch": 0.1113581333920317, + "grad_norm": 0.6833519051213948, + "learning_rate": 9.827431089293649e-06, + "loss": 0.5768, + "step": 5059 + }, + { + "epoch": 0.11138014527845036, + "grad_norm": 3.017077783567174, + "learning_rate": 9.82733823648363e-06, + "loss": 0.4448, + "step": 5060 + }, + { + "epoch": 0.11140215716486902, + "grad_norm": 1.7541968808328614, + "learning_rate": 9.827245359138925e-06, + "loss": 0.2704, + "step": 5061 + }, + { + "epoch": 0.1114241690512877, + "grad_norm": 2.3155287138243006, + "learning_rate": 9.827152457260005e-06, + "loss": 0.4291, + "step": 5062 + }, + { + "epoch": 0.11144618093770636, + "grad_norm": 2.047228178337009, + "learning_rate": 9.827059530847344e-06, + "loss": 0.343, + "step": 5063 + }, + { + "epoch": 0.11146819282412503, + "grad_norm": 2.124946451717386, + "learning_rate": 9.826966579901413e-06, + "loss": 0.4645, + "step": 5064 + }, + { + "epoch": 0.11149020471054369, + "grad_norm": 5.416009994026476, + "learning_rate": 9.826873604422686e-06, + "loss": 0.2927, + "step": 5065 + }, + { + "epoch": 0.11151221659696237, + "grad_norm": 1.6385803007965922, + "learning_rate": 9.826780604411634e-06, + "loss": 0.3926, + "step": 5066 + }, + { + "epoch": 0.11153422848338103, + "grad_norm": 1.8638309452750335, + "learning_rate": 9.82668757986873e-06, + "loss": 0.3321, + "step": 5067 + }, + { + "epoch": 0.11155624036979969, + "grad_norm": 6.205295085270952, + "learning_rate": 9.826594530794447e-06, + "loss": 0.3422, + "step": 5068 + }, + { + "epoch": 0.11157825225621835, + "grad_norm": 1.6018072877046314, + "learning_rate": 9.82650145718926e-06, + "loss": 0.4208, + "step": 5069 + }, + { + "epoch": 0.11160026414263702, + "grad_norm": 1.7972733798563403, + "learning_rate": 9.826408359053636e-06, + "loss": 0.4462, + "step": 5070 + }, + { + "epoch": 0.1116222760290557, + "grad_norm": 1.7471446668098067, + "learning_rate": 9.826315236388055e-06, + "loss": 0.3833, + "step": 5071 + }, + { + "epoch": 0.11164428791547436, + "grad_norm": 1.7965781454868979, + "learning_rate": 9.826222089192985e-06, + "loss": 0.381, + "step": 5072 + }, + { + "epoch": 0.11166629980189302, + "grad_norm": 0.7013038297287345, + "learning_rate": 9.826128917468904e-06, + "loss": 0.5573, + "step": 5073 + }, + { + "epoch": 0.11168831168831168, + "grad_norm": 1.9131049722592903, + "learning_rate": 9.826035721216282e-06, + "loss": 0.3408, + "step": 5074 + }, + { + "epoch": 0.11171032357473036, + "grad_norm": 13.959922982498561, + "learning_rate": 9.825942500435595e-06, + "loss": 0.2844, + "step": 5075 + }, + { + "epoch": 0.11173233546114902, + "grad_norm": 0.6695871192294461, + "learning_rate": 9.825849255127315e-06, + "loss": 0.5515, + "step": 5076 + }, + { + "epoch": 0.11175434734756769, + "grad_norm": 1.7148759374357185, + "learning_rate": 9.825755985291916e-06, + "loss": 0.4357, + "step": 5077 + }, + { + "epoch": 0.11177635923398635, + "grad_norm": 4.413499087950748, + "learning_rate": 9.825662690929873e-06, + "loss": 0.4859, + "step": 5078 + }, + { + "epoch": 0.11179837112040501, + "grad_norm": 1.7240968805896344, + "learning_rate": 9.825569372041661e-06, + "loss": 0.3898, + "step": 5079 + }, + { + "epoch": 0.11182038300682369, + "grad_norm": 1.874997900765935, + "learning_rate": 9.825476028627754e-06, + "loss": 0.2778, + "step": 5080 + }, + { + "epoch": 0.11184239489324235, + "grad_norm": 0.6964657926426031, + "learning_rate": 9.825382660688624e-06, + "loss": 0.5565, + "step": 5081 + }, + { + "epoch": 0.11186440677966102, + "grad_norm": 1.4461417226634623, + "learning_rate": 9.825289268224747e-06, + "loss": 0.2859, + "step": 5082 + }, + { + "epoch": 0.11188641866607968, + "grad_norm": 1.4926160763804932, + "learning_rate": 9.825195851236599e-06, + "loss": 0.3072, + "step": 5083 + }, + { + "epoch": 0.11190843055249836, + "grad_norm": 1.6136273051059318, + "learning_rate": 9.825102409724652e-06, + "loss": 0.3217, + "step": 5084 + }, + { + "epoch": 0.11193044243891702, + "grad_norm": 0.6837064822072935, + "learning_rate": 9.825008943689383e-06, + "loss": 0.6028, + "step": 5085 + }, + { + "epoch": 0.11195245432533568, + "grad_norm": 2.8386390667640473, + "learning_rate": 9.824915453131267e-06, + "loss": 0.4361, + "step": 5086 + }, + { + "epoch": 0.11197446621175434, + "grad_norm": 1.7274827816300544, + "learning_rate": 9.82482193805078e-06, + "loss": 0.3545, + "step": 5087 + }, + { + "epoch": 0.11199647809817301, + "grad_norm": 1.4708649215706, + "learning_rate": 9.824728398448394e-06, + "loss": 0.3269, + "step": 5088 + }, + { + "epoch": 0.11201848998459168, + "grad_norm": 1.7632468102715897, + "learning_rate": 9.824634834324587e-06, + "loss": 0.4238, + "step": 5089 + }, + { + "epoch": 0.11204050187101035, + "grad_norm": 3.206882644154306, + "learning_rate": 9.824541245679832e-06, + "loss": 0.416, + "step": 5090 + }, + { + "epoch": 0.11206251375742901, + "grad_norm": 0.6726394491673768, + "learning_rate": 9.82444763251461e-06, + "loss": 0.5722, + "step": 5091 + }, + { + "epoch": 0.11208452564384767, + "grad_norm": 2.1979752747518955, + "learning_rate": 9.824353994829392e-06, + "loss": 0.3748, + "step": 5092 + }, + { + "epoch": 0.11210653753026634, + "grad_norm": 2.0467268820306157, + "learning_rate": 9.824260332624654e-06, + "loss": 0.4827, + "step": 5093 + }, + { + "epoch": 0.11212854941668501, + "grad_norm": 2.852570136749105, + "learning_rate": 9.824166645900875e-06, + "loss": 0.3478, + "step": 5094 + }, + { + "epoch": 0.11215056130310368, + "grad_norm": 1.6211610364422049, + "learning_rate": 9.824072934658529e-06, + "loss": 0.3019, + "step": 5095 + }, + { + "epoch": 0.11217257318952234, + "grad_norm": 1.8796016998731029, + "learning_rate": 9.823979198898093e-06, + "loss": 0.3423, + "step": 5096 + }, + { + "epoch": 0.112194585075941, + "grad_norm": 1.7521129699221953, + "learning_rate": 9.823885438620042e-06, + "loss": 0.435, + "step": 5097 + }, + { + "epoch": 0.11221659696235968, + "grad_norm": 0.6369712000153172, + "learning_rate": 9.823791653824855e-06, + "loss": 0.5381, + "step": 5098 + }, + { + "epoch": 0.11223860884877834, + "grad_norm": 1.761015809151668, + "learning_rate": 9.823697844513008e-06, + "loss": 0.313, + "step": 5099 + }, + { + "epoch": 0.112260620735197, + "grad_norm": 1.9856629558733645, + "learning_rate": 9.823604010684975e-06, + "loss": 0.4009, + "step": 5100 + }, + { + "epoch": 0.11228263262161567, + "grad_norm": 2.0710664865024353, + "learning_rate": 9.823510152341236e-06, + "loss": 0.4451, + "step": 5101 + }, + { + "epoch": 0.11230464450803433, + "grad_norm": 1.394262316602849, + "learning_rate": 9.823416269482267e-06, + "loss": 0.4283, + "step": 5102 + }, + { + "epoch": 0.11232665639445301, + "grad_norm": 1.4269278262920548, + "learning_rate": 9.823322362108545e-06, + "loss": 0.3349, + "step": 5103 + }, + { + "epoch": 0.11234866828087167, + "grad_norm": 2.3532646675136335, + "learning_rate": 9.823228430220547e-06, + "loss": 0.4122, + "step": 5104 + }, + { + "epoch": 0.11237068016729033, + "grad_norm": 1.56483321488098, + "learning_rate": 9.823134473818752e-06, + "loss": 0.356, + "step": 5105 + }, + { + "epoch": 0.112392692053709, + "grad_norm": 1.9559099830743505, + "learning_rate": 9.823040492903636e-06, + "loss": 0.3948, + "step": 5106 + }, + { + "epoch": 0.11241470394012767, + "grad_norm": 2.1577779551705185, + "learning_rate": 9.822946487475677e-06, + "loss": 0.3614, + "step": 5107 + }, + { + "epoch": 0.11243671582654634, + "grad_norm": 2.353336429064073, + "learning_rate": 9.822852457535352e-06, + "loss": 0.3692, + "step": 5108 + }, + { + "epoch": 0.112458727712965, + "grad_norm": 1.723463879845016, + "learning_rate": 9.82275840308314e-06, + "loss": 0.3779, + "step": 5109 + }, + { + "epoch": 0.11248073959938366, + "grad_norm": 1.5662091374541784, + "learning_rate": 9.822664324119522e-06, + "loss": 0.4181, + "step": 5110 + }, + { + "epoch": 0.11250275148580233, + "grad_norm": 1.5535278114906348, + "learning_rate": 9.822570220644969e-06, + "loss": 0.3521, + "step": 5111 + }, + { + "epoch": 0.112524763372221, + "grad_norm": 1.5586683767558416, + "learning_rate": 9.822476092659965e-06, + "loss": 0.4134, + "step": 5112 + }, + { + "epoch": 0.11254677525863967, + "grad_norm": 0.6938390768087224, + "learning_rate": 9.822381940164986e-06, + "loss": 0.558, + "step": 5113 + }, + { + "epoch": 0.11256878714505833, + "grad_norm": 1.56728519198419, + "learning_rate": 9.822287763160512e-06, + "loss": 0.3518, + "step": 5114 + }, + { + "epoch": 0.11259079903147699, + "grad_norm": 1.6543286025433304, + "learning_rate": 9.82219356164702e-06, + "loss": 0.3852, + "step": 5115 + }, + { + "epoch": 0.11261281091789567, + "grad_norm": 1.453502691352753, + "learning_rate": 9.82209933562499e-06, + "loss": 0.3405, + "step": 5116 + }, + { + "epoch": 0.11263482280431433, + "grad_norm": 3.204542985406141, + "learning_rate": 9.8220050850949e-06, + "loss": 0.3187, + "step": 5117 + }, + { + "epoch": 0.112656834690733, + "grad_norm": 1.814275542509429, + "learning_rate": 9.82191081005723e-06, + "loss": 0.4942, + "step": 5118 + }, + { + "epoch": 0.11267884657715166, + "grad_norm": 1.6954469162922388, + "learning_rate": 9.821816510512458e-06, + "loss": 0.4084, + "step": 5119 + }, + { + "epoch": 0.11270085846357032, + "grad_norm": 0.706020317845356, + "learning_rate": 9.821722186461064e-06, + "loss": 0.5718, + "step": 5120 + }, + { + "epoch": 0.112722870349989, + "grad_norm": 1.661612708411446, + "learning_rate": 9.82162783790353e-06, + "loss": 0.3387, + "step": 5121 + }, + { + "epoch": 0.11274488223640766, + "grad_norm": 2.7922942703271714, + "learning_rate": 9.821533464840329e-06, + "loss": 0.4476, + "step": 5122 + }, + { + "epoch": 0.11276689412282632, + "grad_norm": 1.6236216175685232, + "learning_rate": 9.821439067271945e-06, + "loss": 0.3706, + "step": 5123 + }, + { + "epoch": 0.11278890600924499, + "grad_norm": 2.262363149780442, + "learning_rate": 9.821344645198857e-06, + "loss": 0.3886, + "step": 5124 + }, + { + "epoch": 0.11281091789566366, + "grad_norm": 1.9496349632717378, + "learning_rate": 9.821250198621549e-06, + "loss": 0.4363, + "step": 5125 + }, + { + "epoch": 0.11283292978208233, + "grad_norm": 2.154694979802222, + "learning_rate": 9.821155727540492e-06, + "loss": 0.3795, + "step": 5126 + }, + { + "epoch": 0.11285494166850099, + "grad_norm": 6.0221322868906695, + "learning_rate": 9.821061231956174e-06, + "loss": 0.3704, + "step": 5127 + }, + { + "epoch": 0.11287695355491965, + "grad_norm": 1.6459708425913455, + "learning_rate": 9.820966711869072e-06, + "loss": 0.3972, + "step": 5128 + }, + { + "epoch": 0.11289896544133832, + "grad_norm": 1.609956057354065, + "learning_rate": 9.820872167279669e-06, + "loss": 0.288, + "step": 5129 + }, + { + "epoch": 0.112920977327757, + "grad_norm": 1.56831839317567, + "learning_rate": 9.82077759818844e-06, + "loss": 0.3672, + "step": 5130 + }, + { + "epoch": 0.11294298921417566, + "grad_norm": 1.4214347144152764, + "learning_rate": 9.820683004595873e-06, + "loss": 0.3459, + "step": 5131 + }, + { + "epoch": 0.11296500110059432, + "grad_norm": 1.6034950338299252, + "learning_rate": 9.820588386502441e-06, + "loss": 0.4334, + "step": 5132 + }, + { + "epoch": 0.11298701298701298, + "grad_norm": 1.9556936559526525, + "learning_rate": 9.82049374390863e-06, + "loss": 0.4346, + "step": 5133 + }, + { + "epoch": 0.11300902487343166, + "grad_norm": 4.92286553799732, + "learning_rate": 9.820399076814922e-06, + "loss": 0.324, + "step": 5134 + }, + { + "epoch": 0.11303103675985032, + "grad_norm": 2.3611127886870285, + "learning_rate": 9.820304385221793e-06, + "loss": 0.4119, + "step": 5135 + }, + { + "epoch": 0.11305304864626899, + "grad_norm": 0.7216220800235105, + "learning_rate": 9.820209669129727e-06, + "loss": 0.5781, + "step": 5136 + }, + { + "epoch": 0.11307506053268765, + "grad_norm": 1.479083777885514, + "learning_rate": 9.820114928539209e-06, + "loss": 0.3223, + "step": 5137 + }, + { + "epoch": 0.11309707241910631, + "grad_norm": 3.17164747091067, + "learning_rate": 9.820020163450713e-06, + "loss": 0.3604, + "step": 5138 + }, + { + "epoch": 0.11311908430552499, + "grad_norm": 1.5138410679893572, + "learning_rate": 9.819925373864728e-06, + "loss": 0.4146, + "step": 5139 + }, + { + "epoch": 0.11314109619194365, + "grad_norm": 1.8956162195009894, + "learning_rate": 9.819830559781731e-06, + "loss": 0.3937, + "step": 5140 + }, + { + "epoch": 0.11316310807836232, + "grad_norm": 1.6265758443924638, + "learning_rate": 9.819735721202206e-06, + "loss": 0.3154, + "step": 5141 + }, + { + "epoch": 0.11318511996478098, + "grad_norm": 2.114341980364171, + "learning_rate": 9.819640858126634e-06, + "loss": 0.4111, + "step": 5142 + }, + { + "epoch": 0.11320713185119964, + "grad_norm": 1.5707253155485372, + "learning_rate": 9.819545970555497e-06, + "loss": 0.2808, + "step": 5143 + }, + { + "epoch": 0.11322914373761832, + "grad_norm": 2.004039039279695, + "learning_rate": 9.819451058489279e-06, + "loss": 0.4793, + "step": 5144 + }, + { + "epoch": 0.11325115562403698, + "grad_norm": 1.440065889441846, + "learning_rate": 9.81935612192846e-06, + "loss": 0.3882, + "step": 5145 + }, + { + "epoch": 0.11327316751045564, + "grad_norm": 2.15141897255197, + "learning_rate": 9.819261160873526e-06, + "loss": 0.3537, + "step": 5146 + }, + { + "epoch": 0.11329517939687431, + "grad_norm": 3.527283467772354, + "learning_rate": 9.819166175324954e-06, + "loss": 0.3956, + "step": 5147 + }, + { + "epoch": 0.11331719128329298, + "grad_norm": 0.694720233984797, + "learning_rate": 9.819071165283233e-06, + "loss": 0.5627, + "step": 5148 + }, + { + "epoch": 0.11333920316971165, + "grad_norm": 0.6577072371957903, + "learning_rate": 9.81897613074884e-06, + "loss": 0.5746, + "step": 5149 + }, + { + "epoch": 0.11336121505613031, + "grad_norm": 1.5664353286448343, + "learning_rate": 9.818881071722264e-06, + "loss": 0.4355, + "step": 5150 + }, + { + "epoch": 0.11338322694254897, + "grad_norm": 2.093303492771817, + "learning_rate": 9.818785988203984e-06, + "loss": 0.3251, + "step": 5151 + }, + { + "epoch": 0.11340523882896764, + "grad_norm": 1.7430146695066906, + "learning_rate": 9.818690880194484e-06, + "loss": 0.3074, + "step": 5152 + }, + { + "epoch": 0.11342725071538631, + "grad_norm": 1.9464079759967117, + "learning_rate": 9.818595747694249e-06, + "loss": 0.4867, + "step": 5153 + }, + { + "epoch": 0.11344926260180498, + "grad_norm": 0.7235832784949492, + "learning_rate": 9.81850059070376e-06, + "loss": 0.5614, + "step": 5154 + }, + { + "epoch": 0.11347127448822364, + "grad_norm": 2.7518819711428604, + "learning_rate": 9.818405409223502e-06, + "loss": 0.3587, + "step": 5155 + }, + { + "epoch": 0.1134932863746423, + "grad_norm": 2.030351634796833, + "learning_rate": 9.818310203253958e-06, + "loss": 0.3961, + "step": 5156 + }, + { + "epoch": 0.11351529826106098, + "grad_norm": 1.9411453828503225, + "learning_rate": 9.818214972795614e-06, + "loss": 0.4635, + "step": 5157 + }, + { + "epoch": 0.11353731014747964, + "grad_norm": 0.6752553171740312, + "learning_rate": 9.818119717848951e-06, + "loss": 0.5719, + "step": 5158 + }, + { + "epoch": 0.1135593220338983, + "grad_norm": 2.066519849281895, + "learning_rate": 9.818024438414456e-06, + "loss": 0.4569, + "step": 5159 + }, + { + "epoch": 0.11358133392031697, + "grad_norm": 5.234510971531731, + "learning_rate": 9.81792913449261e-06, + "loss": 0.4462, + "step": 5160 + }, + { + "epoch": 0.11360334580673563, + "grad_norm": 1.4653297939948877, + "learning_rate": 9.817833806083901e-06, + "loss": 0.3807, + "step": 5161 + }, + { + "epoch": 0.11362535769315431, + "grad_norm": 1.8873490562221475, + "learning_rate": 9.817738453188812e-06, + "loss": 0.4596, + "step": 5162 + }, + { + "epoch": 0.11364736957957297, + "grad_norm": 1.602890729576666, + "learning_rate": 9.817643075807826e-06, + "loss": 0.4192, + "step": 5163 + }, + { + "epoch": 0.11366938146599163, + "grad_norm": 3.071564698771749, + "learning_rate": 9.81754767394143e-06, + "loss": 0.3451, + "step": 5164 + }, + { + "epoch": 0.1136913933524103, + "grad_norm": 0.7680003972724626, + "learning_rate": 9.817452247590108e-06, + "loss": 0.5792, + "step": 5165 + }, + { + "epoch": 0.11371340523882897, + "grad_norm": 1.5123560468057224, + "learning_rate": 9.817356796754345e-06, + "loss": 0.4226, + "step": 5166 + }, + { + "epoch": 0.11373541712524764, + "grad_norm": 1.8376111375581312, + "learning_rate": 9.817261321434626e-06, + "loss": 0.4418, + "step": 5167 + }, + { + "epoch": 0.1137574290116663, + "grad_norm": 1.472189261881807, + "learning_rate": 9.817165821631436e-06, + "loss": 0.359, + "step": 5168 + }, + { + "epoch": 0.11377944089808496, + "grad_norm": 1.1528345720896993, + "learning_rate": 9.817070297345262e-06, + "loss": 0.4042, + "step": 5169 + }, + { + "epoch": 0.11380145278450363, + "grad_norm": 1.7066397358783545, + "learning_rate": 9.816974748576589e-06, + "loss": 0.2937, + "step": 5170 + }, + { + "epoch": 0.1138234646709223, + "grad_norm": 1.6227781982205283, + "learning_rate": 9.8168791753259e-06, + "loss": 0.3587, + "step": 5171 + }, + { + "epoch": 0.11384547655734097, + "grad_norm": 1.6258841346628135, + "learning_rate": 9.816783577593682e-06, + "loss": 0.423, + "step": 5172 + }, + { + "epoch": 0.11386748844375963, + "grad_norm": 2.267470914498364, + "learning_rate": 9.816687955380425e-06, + "loss": 0.4033, + "step": 5173 + }, + { + "epoch": 0.11388950033017829, + "grad_norm": 2.2318024382006545, + "learning_rate": 9.816592308686607e-06, + "loss": 0.3972, + "step": 5174 + }, + { + "epoch": 0.11391151221659697, + "grad_norm": 1.377041615913359, + "learning_rate": 9.816496637512721e-06, + "loss": 0.3385, + "step": 5175 + }, + { + "epoch": 0.11393352410301563, + "grad_norm": 0.6838659293634584, + "learning_rate": 9.816400941859251e-06, + "loss": 0.5595, + "step": 5176 + }, + { + "epoch": 0.1139555359894343, + "grad_norm": 1.4112315923019978, + "learning_rate": 9.816305221726683e-06, + "loss": 0.3586, + "step": 5177 + }, + { + "epoch": 0.11397754787585296, + "grad_norm": 1.3063656812895057, + "learning_rate": 9.816209477115504e-06, + "loss": 0.3942, + "step": 5178 + }, + { + "epoch": 0.11399955976227162, + "grad_norm": 1.4653216085982343, + "learning_rate": 9.8161137080262e-06, + "loss": 0.3171, + "step": 5179 + }, + { + "epoch": 0.1140215716486903, + "grad_norm": 1.5728387212287103, + "learning_rate": 9.816017914459257e-06, + "loss": 0.3406, + "step": 5180 + }, + { + "epoch": 0.11404358353510896, + "grad_norm": 1.85708078080103, + "learning_rate": 9.815922096415164e-06, + "loss": 0.2845, + "step": 5181 + }, + { + "epoch": 0.11406559542152762, + "grad_norm": 1.3586399613186055, + "learning_rate": 9.815826253894407e-06, + "loss": 0.303, + "step": 5182 + }, + { + "epoch": 0.11408760730794629, + "grad_norm": 1.6509412067934424, + "learning_rate": 9.815730386897473e-06, + "loss": 0.428, + "step": 5183 + }, + { + "epoch": 0.11410961919436495, + "grad_norm": 0.7014732322142735, + "learning_rate": 9.81563449542485e-06, + "loss": 0.5715, + "step": 5184 + }, + { + "epoch": 0.11413163108078363, + "grad_norm": 1.6960576463108126, + "learning_rate": 9.815538579477025e-06, + "loss": 0.4719, + "step": 5185 + }, + { + "epoch": 0.11415364296720229, + "grad_norm": 2.3817008738150687, + "learning_rate": 9.815442639054484e-06, + "loss": 0.3342, + "step": 5186 + }, + { + "epoch": 0.11417565485362095, + "grad_norm": 2.1904690506069335, + "learning_rate": 9.815346674157715e-06, + "loss": 0.369, + "step": 5187 + }, + { + "epoch": 0.11419766674003962, + "grad_norm": 1.287902363431786, + "learning_rate": 9.815250684787207e-06, + "loss": 0.3515, + "step": 5188 + }, + { + "epoch": 0.11421967862645829, + "grad_norm": 0.7167598069909658, + "learning_rate": 9.815154670943447e-06, + "loss": 0.5949, + "step": 5189 + }, + { + "epoch": 0.11424169051287696, + "grad_norm": 1.7562083967001159, + "learning_rate": 9.815058632626926e-06, + "loss": 0.3246, + "step": 5190 + }, + { + "epoch": 0.11426370239929562, + "grad_norm": 0.7076767312074429, + "learning_rate": 9.814962569838127e-06, + "loss": 0.5747, + "step": 5191 + }, + { + "epoch": 0.11428571428571428, + "grad_norm": 1.2870034821939906, + "learning_rate": 9.81486648257754e-06, + "loss": 0.3341, + "step": 5192 + }, + { + "epoch": 0.11430772617213295, + "grad_norm": 1.384186337207811, + "learning_rate": 9.814770370845657e-06, + "loss": 0.3295, + "step": 5193 + }, + { + "epoch": 0.11432973805855162, + "grad_norm": 1.4490803633583165, + "learning_rate": 9.814674234642961e-06, + "loss": 0.3823, + "step": 5194 + }, + { + "epoch": 0.11435174994497029, + "grad_norm": 9.639195384564106, + "learning_rate": 9.814578073969945e-06, + "loss": 0.3725, + "step": 5195 + }, + { + "epoch": 0.11437376183138895, + "grad_norm": 1.9919526095153277, + "learning_rate": 9.814481888827095e-06, + "loss": 0.3681, + "step": 5196 + }, + { + "epoch": 0.11439577371780761, + "grad_norm": 2.5502875340038247, + "learning_rate": 9.8143856792149e-06, + "loss": 0.4461, + "step": 5197 + }, + { + "epoch": 0.11441778560422629, + "grad_norm": 1.3554030240890322, + "learning_rate": 9.81428944513385e-06, + "loss": 0.4349, + "step": 5198 + }, + { + "epoch": 0.11443979749064495, + "grad_norm": 0.9565449649659318, + "learning_rate": 9.814193186584434e-06, + "loss": 0.5636, + "step": 5199 + }, + { + "epoch": 0.11446180937706361, + "grad_norm": 3.945323228606013, + "learning_rate": 9.81409690356714e-06, + "loss": 0.4166, + "step": 5200 + }, + { + "epoch": 0.11448382126348228, + "grad_norm": 2.1155925327513123, + "learning_rate": 9.81400059608246e-06, + "loss": 0.3767, + "step": 5201 + }, + { + "epoch": 0.11450583314990094, + "grad_norm": 2.0148100643651867, + "learning_rate": 9.813904264130882e-06, + "loss": 0.3734, + "step": 5202 + }, + { + "epoch": 0.11452784503631962, + "grad_norm": 1.5960184639343733, + "learning_rate": 9.813807907712895e-06, + "loss": 0.434, + "step": 5203 + }, + { + "epoch": 0.11454985692273828, + "grad_norm": 3.4843797413177935, + "learning_rate": 9.813711526828988e-06, + "loss": 0.3917, + "step": 5204 + }, + { + "epoch": 0.11457186880915694, + "grad_norm": 2.037948842254551, + "learning_rate": 9.813615121479654e-06, + "loss": 0.3997, + "step": 5205 + }, + { + "epoch": 0.1145938806955756, + "grad_norm": 1.1864757645762338, + "learning_rate": 9.81351869166538e-06, + "loss": 0.2611, + "step": 5206 + }, + { + "epoch": 0.11461589258199428, + "grad_norm": 1.7089966393145872, + "learning_rate": 9.813422237386657e-06, + "loss": 0.3595, + "step": 5207 + }, + { + "epoch": 0.11463790446841295, + "grad_norm": 0.7079050410101704, + "learning_rate": 9.813325758643976e-06, + "loss": 0.5711, + "step": 5208 + }, + { + "epoch": 0.11465991635483161, + "grad_norm": 2.2000252777884275, + "learning_rate": 9.813229255437826e-06, + "loss": 0.4358, + "step": 5209 + }, + { + "epoch": 0.11468192824125027, + "grad_norm": 1.7457415867318544, + "learning_rate": 9.813132727768698e-06, + "loss": 0.3699, + "step": 5210 + }, + { + "epoch": 0.11470394012766894, + "grad_norm": 1.9638941466389854, + "learning_rate": 9.813036175637083e-06, + "loss": 0.3505, + "step": 5211 + }, + { + "epoch": 0.11472595201408761, + "grad_norm": 2.2674360139388967, + "learning_rate": 9.812939599043471e-06, + "loss": 0.387, + "step": 5212 + }, + { + "epoch": 0.11474796390050628, + "grad_norm": 8.554025315636459, + "learning_rate": 9.812842997988354e-06, + "loss": 0.3405, + "step": 5213 + }, + { + "epoch": 0.11476997578692494, + "grad_norm": 1.7563717641552425, + "learning_rate": 9.812746372472224e-06, + "loss": 0.3448, + "step": 5214 + }, + { + "epoch": 0.1147919876733436, + "grad_norm": 2.2062597452316832, + "learning_rate": 9.812649722495569e-06, + "loss": 0.3941, + "step": 5215 + }, + { + "epoch": 0.11481399955976228, + "grad_norm": 1.7017958594883393, + "learning_rate": 9.81255304805888e-06, + "loss": 0.4117, + "step": 5216 + }, + { + "epoch": 0.11483601144618094, + "grad_norm": 1.4115463581265615, + "learning_rate": 9.812456349162652e-06, + "loss": 0.3677, + "step": 5217 + }, + { + "epoch": 0.1148580233325996, + "grad_norm": 1.6203741100440263, + "learning_rate": 9.812359625807373e-06, + "loss": 0.4058, + "step": 5218 + }, + { + "epoch": 0.11488003521901827, + "grad_norm": 1.7905419109736302, + "learning_rate": 9.812262877993537e-06, + "loss": 0.3926, + "step": 5219 + }, + { + "epoch": 0.11490204710543693, + "grad_norm": 1.3979217094113487, + "learning_rate": 9.812166105721634e-06, + "loss": 0.3941, + "step": 5220 + }, + { + "epoch": 0.11492405899185561, + "grad_norm": 2.070653330386513, + "learning_rate": 9.812069308992156e-06, + "loss": 0.3564, + "step": 5221 + }, + { + "epoch": 0.11494607087827427, + "grad_norm": 1.8500765229462433, + "learning_rate": 9.811972487805597e-06, + "loss": 0.3921, + "step": 5222 + }, + { + "epoch": 0.11496808276469293, + "grad_norm": 1.5411453884398416, + "learning_rate": 9.811875642162447e-06, + "loss": 0.4547, + "step": 5223 + }, + { + "epoch": 0.1149900946511116, + "grad_norm": 1.4751026781777334, + "learning_rate": 9.811778772063199e-06, + "loss": 0.3544, + "step": 5224 + }, + { + "epoch": 0.11501210653753027, + "grad_norm": 1.3654793902264428, + "learning_rate": 9.811681877508345e-06, + "loss": 0.3939, + "step": 5225 + }, + { + "epoch": 0.11503411842394894, + "grad_norm": 2.429639346898513, + "learning_rate": 9.811584958498379e-06, + "loss": 0.3726, + "step": 5226 + }, + { + "epoch": 0.1150561303103676, + "grad_norm": 1.7336070851323613, + "learning_rate": 9.81148801503379e-06, + "loss": 0.3967, + "step": 5227 + }, + { + "epoch": 0.11507814219678626, + "grad_norm": 2.163297403683483, + "learning_rate": 9.811391047115074e-06, + "loss": 0.4353, + "step": 5228 + }, + { + "epoch": 0.11510015408320493, + "grad_norm": 1.5825601031342027, + "learning_rate": 9.811294054742722e-06, + "loss": 0.3516, + "step": 5229 + }, + { + "epoch": 0.1151221659696236, + "grad_norm": 0.7842431890738828, + "learning_rate": 9.811197037917229e-06, + "loss": 0.59, + "step": 5230 + }, + { + "epoch": 0.11514417785604227, + "grad_norm": 1.3635523952661308, + "learning_rate": 9.811099996639087e-06, + "loss": 0.3421, + "step": 5231 + }, + { + "epoch": 0.11516618974246093, + "grad_norm": 0.6445091980995463, + "learning_rate": 9.811002930908788e-06, + "loss": 0.5551, + "step": 5232 + }, + { + "epoch": 0.11518820162887959, + "grad_norm": 1.619943792279935, + "learning_rate": 9.810905840726826e-06, + "loss": 0.4104, + "step": 5233 + }, + { + "epoch": 0.11521021351529825, + "grad_norm": 1.599028200599019, + "learning_rate": 9.810808726093694e-06, + "loss": 0.3045, + "step": 5234 + }, + { + "epoch": 0.11523222540171693, + "grad_norm": 2.5121661184029196, + "learning_rate": 9.81071158700989e-06, + "loss": 0.4246, + "step": 5235 + }, + { + "epoch": 0.1152542372881356, + "grad_norm": 1.3688327770153665, + "learning_rate": 9.8106144234759e-06, + "loss": 0.421, + "step": 5236 + }, + { + "epoch": 0.11527624917455426, + "grad_norm": 0.8265089228195035, + "learning_rate": 9.810517235492224e-06, + "loss": 0.5721, + "step": 5237 + }, + { + "epoch": 0.11529826106097292, + "grad_norm": 1.7032782725096003, + "learning_rate": 9.810420023059352e-06, + "loss": 0.3709, + "step": 5238 + }, + { + "epoch": 0.1153202729473916, + "grad_norm": 1.3891487147900878, + "learning_rate": 9.810322786177782e-06, + "loss": 0.4034, + "step": 5239 + }, + { + "epoch": 0.11534228483381026, + "grad_norm": 1.3453457896060224, + "learning_rate": 9.810225524848006e-06, + "loss": 0.3995, + "step": 5240 + }, + { + "epoch": 0.11536429672022892, + "grad_norm": 1.5283348457698183, + "learning_rate": 9.810128239070517e-06, + "loss": 0.3765, + "step": 5241 + }, + { + "epoch": 0.11538630860664759, + "grad_norm": 1.4035244742911883, + "learning_rate": 9.810030928845813e-06, + "loss": 0.4173, + "step": 5242 + }, + { + "epoch": 0.11540832049306625, + "grad_norm": 1.3085753269333986, + "learning_rate": 9.809933594174386e-06, + "loss": 0.3302, + "step": 5243 + }, + { + "epoch": 0.11543033237948493, + "grad_norm": 1.4194861079241938, + "learning_rate": 9.80983623505673e-06, + "loss": 0.3567, + "step": 5244 + }, + { + "epoch": 0.11545234426590359, + "grad_norm": 2.305368746972907, + "learning_rate": 9.809738851493342e-06, + "loss": 0.3299, + "step": 5245 + }, + { + "epoch": 0.11547435615232225, + "grad_norm": 1.6166216471825943, + "learning_rate": 9.809641443484715e-06, + "loss": 0.3735, + "step": 5246 + }, + { + "epoch": 0.11549636803874092, + "grad_norm": 1.3922438625395213, + "learning_rate": 9.809544011031347e-06, + "loss": 0.3276, + "step": 5247 + }, + { + "epoch": 0.11551837992515959, + "grad_norm": 1.7695741362008472, + "learning_rate": 9.809446554133728e-06, + "loss": 0.3896, + "step": 5248 + }, + { + "epoch": 0.11554039181157826, + "grad_norm": 1.2661336027535965, + "learning_rate": 9.809349072792358e-06, + "loss": 0.3418, + "step": 5249 + }, + { + "epoch": 0.11556240369799692, + "grad_norm": 1.809702315627451, + "learning_rate": 9.809251567007733e-06, + "loss": 0.3553, + "step": 5250 + }, + { + "epoch": 0.11558441558441558, + "grad_norm": 0.6856778590328935, + "learning_rate": 9.809154036780344e-06, + "loss": 0.5383, + "step": 5251 + }, + { + "epoch": 0.11560642747083424, + "grad_norm": 1.664437072910841, + "learning_rate": 9.80905648211069e-06, + "loss": 0.3552, + "step": 5252 + }, + { + "epoch": 0.11562843935725292, + "grad_norm": 1.1995938933182515, + "learning_rate": 9.808958902999266e-06, + "loss": 0.3151, + "step": 5253 + }, + { + "epoch": 0.11565045124367158, + "grad_norm": 0.6553168572263649, + "learning_rate": 9.808861299446567e-06, + "loss": 0.5723, + "step": 5254 + }, + { + "epoch": 0.11567246313009025, + "grad_norm": 1.5138359328961126, + "learning_rate": 9.808763671453092e-06, + "loss": 0.357, + "step": 5255 + }, + { + "epoch": 0.11569447501650891, + "grad_norm": 2.708480265138946, + "learning_rate": 9.808666019019336e-06, + "loss": 0.3855, + "step": 5256 + }, + { + "epoch": 0.11571648690292759, + "grad_norm": 1.4750707985430251, + "learning_rate": 9.808568342145791e-06, + "loss": 0.3512, + "step": 5257 + }, + { + "epoch": 0.11573849878934625, + "grad_norm": 10.23417252195191, + "learning_rate": 9.808470640832962e-06, + "loss": 0.3741, + "step": 5258 + }, + { + "epoch": 0.11576051067576491, + "grad_norm": 1.5986043767233995, + "learning_rate": 9.808372915081337e-06, + "loss": 0.2733, + "step": 5259 + }, + { + "epoch": 0.11578252256218358, + "grad_norm": 1.6491408645545393, + "learning_rate": 9.808275164891416e-06, + "loss": 0.4029, + "step": 5260 + }, + { + "epoch": 0.11580453444860224, + "grad_norm": 1.7914683916429797, + "learning_rate": 9.808177390263697e-06, + "loss": 0.3522, + "step": 5261 + }, + { + "epoch": 0.11582654633502092, + "grad_norm": 1.3846473623318032, + "learning_rate": 9.808079591198676e-06, + "loss": 0.411, + "step": 5262 + }, + { + "epoch": 0.11584855822143958, + "grad_norm": 0.7570048772898605, + "learning_rate": 9.807981767696851e-06, + "loss": 0.5538, + "step": 5263 + }, + { + "epoch": 0.11587057010785824, + "grad_norm": 0.7275701499200558, + "learning_rate": 9.807883919758717e-06, + "loss": 0.5506, + "step": 5264 + }, + { + "epoch": 0.1158925819942769, + "grad_norm": 0.6355767415959047, + "learning_rate": 9.807786047384773e-06, + "loss": 0.5582, + "step": 5265 + }, + { + "epoch": 0.11591459388069558, + "grad_norm": 2.5146775269286983, + "learning_rate": 9.807688150575514e-06, + "loss": 0.3436, + "step": 5266 + }, + { + "epoch": 0.11593660576711425, + "grad_norm": 1.9730655404144262, + "learning_rate": 9.807590229331441e-06, + "loss": 0.3484, + "step": 5267 + }, + { + "epoch": 0.11595861765353291, + "grad_norm": 1.435453475188163, + "learning_rate": 9.80749228365305e-06, + "loss": 0.3693, + "step": 5268 + }, + { + "epoch": 0.11598062953995157, + "grad_norm": 0.8589406533084816, + "learning_rate": 9.807394313540842e-06, + "loss": 0.5518, + "step": 5269 + }, + { + "epoch": 0.11600264142637023, + "grad_norm": 1.6682570591198858, + "learning_rate": 9.807296318995307e-06, + "loss": 0.4032, + "step": 5270 + }, + { + "epoch": 0.11602465331278891, + "grad_norm": 2.0408597128891297, + "learning_rate": 9.807198300016952e-06, + "loss": 0.4047, + "step": 5271 + }, + { + "epoch": 0.11604666519920757, + "grad_norm": 1.4140243701778143, + "learning_rate": 9.80710025660627e-06, + "loss": 0.2773, + "step": 5272 + }, + { + "epoch": 0.11606867708562624, + "grad_norm": 1.345257749977033, + "learning_rate": 9.807002188763759e-06, + "loss": 0.3146, + "step": 5273 + }, + { + "epoch": 0.1160906889720449, + "grad_norm": 1.9823708092796126, + "learning_rate": 9.80690409648992e-06, + "loss": 0.3969, + "step": 5274 + }, + { + "epoch": 0.11611270085846358, + "grad_norm": 2.7296833364542885, + "learning_rate": 9.806805979785249e-06, + "loss": 0.3989, + "step": 5275 + }, + { + "epoch": 0.11613471274488224, + "grad_norm": 2.2897242722879767, + "learning_rate": 9.80670783865025e-06, + "loss": 0.3573, + "step": 5276 + }, + { + "epoch": 0.1161567246313009, + "grad_norm": 0.7359501433245903, + "learning_rate": 9.806609673085414e-06, + "loss": 0.5417, + "step": 5277 + }, + { + "epoch": 0.11617873651771957, + "grad_norm": 1.3475606418157522, + "learning_rate": 9.806511483091245e-06, + "loss": 0.3305, + "step": 5278 + }, + { + "epoch": 0.11620074840413823, + "grad_norm": 0.692252180371567, + "learning_rate": 9.806413268668243e-06, + "loss": 0.5615, + "step": 5279 + }, + { + "epoch": 0.1162227602905569, + "grad_norm": 1.457895007787851, + "learning_rate": 9.806315029816902e-06, + "loss": 0.3774, + "step": 5280 + }, + { + "epoch": 0.11624477217697557, + "grad_norm": 0.6489518620586193, + "learning_rate": 9.806216766537726e-06, + "loss": 0.5567, + "step": 5281 + }, + { + "epoch": 0.11626678406339423, + "grad_norm": 1.4511605475616, + "learning_rate": 9.806118478831213e-06, + "loss": 0.405, + "step": 5282 + }, + { + "epoch": 0.1162887959498129, + "grad_norm": 1.9409813995015326, + "learning_rate": 9.80602016669786e-06, + "loss": 0.2729, + "step": 5283 + }, + { + "epoch": 0.11631080783623156, + "grad_norm": 1.5293723966151913, + "learning_rate": 9.805921830138172e-06, + "loss": 0.3573, + "step": 5284 + }, + { + "epoch": 0.11633281972265024, + "grad_norm": 1.4686088591777673, + "learning_rate": 9.805823469152643e-06, + "loss": 0.3531, + "step": 5285 + }, + { + "epoch": 0.1163548316090689, + "grad_norm": 0.8095216624362179, + "learning_rate": 9.805725083741777e-06, + "loss": 0.5626, + "step": 5286 + }, + { + "epoch": 0.11637684349548756, + "grad_norm": 0.7276493857745938, + "learning_rate": 9.805626673906073e-06, + "loss": 0.5817, + "step": 5287 + }, + { + "epoch": 0.11639885538190622, + "grad_norm": 1.6114914138277543, + "learning_rate": 9.80552823964603e-06, + "loss": 0.4226, + "step": 5288 + }, + { + "epoch": 0.1164208672683249, + "grad_norm": 1.7347196971904069, + "learning_rate": 9.80542978096215e-06, + "loss": 0.3298, + "step": 5289 + }, + { + "epoch": 0.11644287915474356, + "grad_norm": 3.1042225445289495, + "learning_rate": 9.805331297854932e-06, + "loss": 0.3051, + "step": 5290 + }, + { + "epoch": 0.11646489104116223, + "grad_norm": 0.7888561137825243, + "learning_rate": 9.805232790324877e-06, + "loss": 0.5598, + "step": 5291 + }, + { + "epoch": 0.11648690292758089, + "grad_norm": 0.7788005151366041, + "learning_rate": 9.805134258372486e-06, + "loss": 0.5456, + "step": 5292 + }, + { + "epoch": 0.11650891481399955, + "grad_norm": 1.7178872111066998, + "learning_rate": 9.80503570199826e-06, + "loss": 0.4048, + "step": 5293 + }, + { + "epoch": 0.11653092670041823, + "grad_norm": 1.7288922740902397, + "learning_rate": 9.8049371212027e-06, + "loss": 0.3866, + "step": 5294 + }, + { + "epoch": 0.1165529385868369, + "grad_norm": 1.2932791978723759, + "learning_rate": 9.804838515986305e-06, + "loss": 0.3783, + "step": 5295 + }, + { + "epoch": 0.11657495047325556, + "grad_norm": 1.5266764182485126, + "learning_rate": 9.804739886349579e-06, + "loss": 0.3788, + "step": 5296 + }, + { + "epoch": 0.11659696235967422, + "grad_norm": 1.8410793222349686, + "learning_rate": 9.804641232293022e-06, + "loss": 0.3184, + "step": 5297 + }, + { + "epoch": 0.1166189742460929, + "grad_norm": 1.4196703624048532, + "learning_rate": 9.804542553817133e-06, + "loss": 0.3905, + "step": 5298 + }, + { + "epoch": 0.11664098613251156, + "grad_norm": 1.7608203046452966, + "learning_rate": 9.804443850922419e-06, + "loss": 0.3887, + "step": 5299 + }, + { + "epoch": 0.11666299801893022, + "grad_norm": 2.1579606828736186, + "learning_rate": 9.804345123609376e-06, + "loss": 0.3778, + "step": 5300 + }, + { + "epoch": 0.11668500990534889, + "grad_norm": 1.4273401683446125, + "learning_rate": 9.80424637187851e-06, + "loss": 0.4255, + "step": 5301 + }, + { + "epoch": 0.11670702179176755, + "grad_norm": 1.4913834226682852, + "learning_rate": 9.80414759573032e-06, + "loss": 0.4136, + "step": 5302 + }, + { + "epoch": 0.11672903367818623, + "grad_norm": 3.56234407643793, + "learning_rate": 9.80404879516531e-06, + "loss": 0.4048, + "step": 5303 + }, + { + "epoch": 0.11675104556460489, + "grad_norm": 1.603713090482001, + "learning_rate": 9.80394997018398e-06, + "loss": 0.3411, + "step": 5304 + }, + { + "epoch": 0.11677305745102355, + "grad_norm": 6.173068154314863, + "learning_rate": 9.803851120786834e-06, + "loss": 0.3845, + "step": 5305 + }, + { + "epoch": 0.11679506933744221, + "grad_norm": 1.741639398025067, + "learning_rate": 9.803752246974377e-06, + "loss": 0.3409, + "step": 5306 + }, + { + "epoch": 0.11681708122386089, + "grad_norm": 1.4185718812997623, + "learning_rate": 9.803653348747104e-06, + "loss": 0.3304, + "step": 5307 + }, + { + "epoch": 0.11683909311027955, + "grad_norm": 2.0625865752439028, + "learning_rate": 9.803554426105524e-06, + "loss": 0.398, + "step": 5308 + }, + { + "epoch": 0.11686110499669822, + "grad_norm": 1.3532746332301857, + "learning_rate": 9.80345547905014e-06, + "loss": 0.3282, + "step": 5309 + }, + { + "epoch": 0.11688311688311688, + "grad_norm": 1.768250994332981, + "learning_rate": 9.803356507581449e-06, + "loss": 0.3405, + "step": 5310 + }, + { + "epoch": 0.11690512876953554, + "grad_norm": 3.143044605709416, + "learning_rate": 9.803257511699961e-06, + "loss": 0.3773, + "step": 5311 + }, + { + "epoch": 0.11692714065595422, + "grad_norm": 1.0708122238685485, + "learning_rate": 9.803158491406174e-06, + "loss": 0.5543, + "step": 5312 + }, + { + "epoch": 0.11694915254237288, + "grad_norm": 2.2995297741555327, + "learning_rate": 9.803059446700596e-06, + "loss": 0.3977, + "step": 5313 + }, + { + "epoch": 0.11697116442879155, + "grad_norm": 0.7186232268570807, + "learning_rate": 9.802960377583725e-06, + "loss": 0.5783, + "step": 5314 + }, + { + "epoch": 0.11699317631521021, + "grad_norm": 1.697491245880726, + "learning_rate": 9.802861284056069e-06, + "loss": 0.3512, + "step": 5315 + }, + { + "epoch": 0.11701518820162889, + "grad_norm": 1.3260335257277305, + "learning_rate": 9.80276216611813e-06, + "loss": 0.3569, + "step": 5316 + }, + { + "epoch": 0.11703720008804755, + "grad_norm": 1.4404775383714552, + "learning_rate": 9.802663023770408e-06, + "loss": 0.3857, + "step": 5317 + }, + { + "epoch": 0.11705921197446621, + "grad_norm": 2.164918261584977, + "learning_rate": 9.802563857013413e-06, + "loss": 0.3901, + "step": 5318 + }, + { + "epoch": 0.11708122386088488, + "grad_norm": 1.7685263379311633, + "learning_rate": 9.80246466584765e-06, + "loss": 0.3674, + "step": 5319 + }, + { + "epoch": 0.11710323574730354, + "grad_norm": 1.7629384084088517, + "learning_rate": 9.802365450273615e-06, + "loss": 0.3677, + "step": 5320 + }, + { + "epoch": 0.11712524763372222, + "grad_norm": 1.366812975293896, + "learning_rate": 9.802266210291817e-06, + "loss": 0.3576, + "step": 5321 + }, + { + "epoch": 0.11714725952014088, + "grad_norm": 1.4211464958628506, + "learning_rate": 9.802166945902762e-06, + "loss": 0.3808, + "step": 5322 + }, + { + "epoch": 0.11716927140655954, + "grad_norm": 1.8297347433893796, + "learning_rate": 9.802067657106951e-06, + "loss": 0.6062, + "step": 5323 + }, + { + "epoch": 0.1171912832929782, + "grad_norm": 2.7620148096823858, + "learning_rate": 9.801968343904892e-06, + "loss": 0.3955, + "step": 5324 + }, + { + "epoch": 0.11721329517939687, + "grad_norm": 1.6374050192903185, + "learning_rate": 9.801869006297087e-06, + "loss": 0.3037, + "step": 5325 + }, + { + "epoch": 0.11723530706581554, + "grad_norm": 1.7895017178174322, + "learning_rate": 9.801769644284043e-06, + "loss": 0.3513, + "step": 5326 + }, + { + "epoch": 0.11725731895223421, + "grad_norm": 0.8197155838296573, + "learning_rate": 9.801670257866265e-06, + "loss": 0.5708, + "step": 5327 + }, + { + "epoch": 0.11727933083865287, + "grad_norm": 1.85157002685392, + "learning_rate": 9.801570847044254e-06, + "loss": 0.4299, + "step": 5328 + }, + { + "epoch": 0.11730134272507153, + "grad_norm": 2.7804493260613974, + "learning_rate": 9.80147141181852e-06, + "loss": 0.3981, + "step": 5329 + }, + { + "epoch": 0.11732335461149021, + "grad_norm": 2.0022472095465926, + "learning_rate": 9.801371952189567e-06, + "loss": 0.4764, + "step": 5330 + }, + { + "epoch": 0.11734536649790887, + "grad_norm": 1.3021910575828455, + "learning_rate": 9.8012724681579e-06, + "loss": 0.3374, + "step": 5331 + }, + { + "epoch": 0.11736737838432754, + "grad_norm": 1.6298706016835651, + "learning_rate": 9.801172959724025e-06, + "loss": 0.3416, + "step": 5332 + }, + { + "epoch": 0.1173893902707462, + "grad_norm": 1.699698017776326, + "learning_rate": 9.801073426888447e-06, + "loss": 0.4137, + "step": 5333 + }, + { + "epoch": 0.11741140215716486, + "grad_norm": 2.011652620609305, + "learning_rate": 9.800973869651674e-06, + "loss": 0.4396, + "step": 5334 + }, + { + "epoch": 0.11743341404358354, + "grad_norm": 1.6212714545140987, + "learning_rate": 9.800874288014211e-06, + "loss": 0.3907, + "step": 5335 + }, + { + "epoch": 0.1174554259300022, + "grad_norm": 1.594008899339636, + "learning_rate": 9.800774681976561e-06, + "loss": 0.3938, + "step": 5336 + }, + { + "epoch": 0.11747743781642087, + "grad_norm": 1.5588320938893616, + "learning_rate": 9.800675051539235e-06, + "loss": 0.3655, + "step": 5337 + }, + { + "epoch": 0.11749944970283953, + "grad_norm": 1.6224085774582442, + "learning_rate": 9.800575396702737e-06, + "loss": 0.3887, + "step": 5338 + }, + { + "epoch": 0.1175214615892582, + "grad_norm": 1.6026769919296433, + "learning_rate": 9.800475717467573e-06, + "loss": 0.4132, + "step": 5339 + }, + { + "epoch": 0.11754347347567687, + "grad_norm": 1.6301611092461126, + "learning_rate": 9.80037601383425e-06, + "loss": 0.3928, + "step": 5340 + }, + { + "epoch": 0.11756548536209553, + "grad_norm": 1.482831856021753, + "learning_rate": 9.800276285803275e-06, + "loss": 0.388, + "step": 5341 + }, + { + "epoch": 0.1175874972485142, + "grad_norm": 1.7337432536159556, + "learning_rate": 9.800176533375155e-06, + "loss": 0.3591, + "step": 5342 + }, + { + "epoch": 0.11760950913493286, + "grad_norm": 1.1258861084430456, + "learning_rate": 9.800076756550398e-06, + "loss": 0.6027, + "step": 5343 + }, + { + "epoch": 0.11763152102135153, + "grad_norm": 1.6050391483464055, + "learning_rate": 9.799976955329508e-06, + "loss": 0.3341, + "step": 5344 + }, + { + "epoch": 0.1176535329077702, + "grad_norm": 1.6110591449518936, + "learning_rate": 9.799877129712995e-06, + "loss": 0.392, + "step": 5345 + }, + { + "epoch": 0.11767554479418886, + "grad_norm": 1.4088795172740434, + "learning_rate": 9.799777279701368e-06, + "loss": 0.3565, + "step": 5346 + }, + { + "epoch": 0.11769755668060752, + "grad_norm": 1.5418568952895442, + "learning_rate": 9.79967740529513e-06, + "loss": 0.3764, + "step": 5347 + }, + { + "epoch": 0.1177195685670262, + "grad_norm": 7.080070969252912, + "learning_rate": 9.799577506494791e-06, + "loss": 0.3065, + "step": 5348 + }, + { + "epoch": 0.11774158045344486, + "grad_norm": 1.488207586220796, + "learning_rate": 9.799477583300859e-06, + "loss": 0.339, + "step": 5349 + }, + { + "epoch": 0.11776359233986353, + "grad_norm": 1.5591099245662496, + "learning_rate": 9.79937763571384e-06, + "loss": 0.4426, + "step": 5350 + }, + { + "epoch": 0.11778560422628219, + "grad_norm": 1.6494846192762809, + "learning_rate": 9.799277663734245e-06, + "loss": 0.3561, + "step": 5351 + }, + { + "epoch": 0.11780761611270085, + "grad_norm": 1.422493894571587, + "learning_rate": 9.79917766736258e-06, + "loss": 0.4239, + "step": 5352 + }, + { + "epoch": 0.11782962799911953, + "grad_norm": 1.7014005752486638, + "learning_rate": 9.799077646599352e-06, + "loss": 0.3902, + "step": 5353 + }, + { + "epoch": 0.11785163988553819, + "grad_norm": 1.5456510666286105, + "learning_rate": 9.798977601445073e-06, + "loss": 0.4212, + "step": 5354 + }, + { + "epoch": 0.11787365177195686, + "grad_norm": 1.3349073359415908, + "learning_rate": 9.798877531900249e-06, + "loss": 0.4194, + "step": 5355 + }, + { + "epoch": 0.11789566365837552, + "grad_norm": 2.347415039573865, + "learning_rate": 9.798777437965387e-06, + "loss": 0.4015, + "step": 5356 + }, + { + "epoch": 0.1179176755447942, + "grad_norm": 0.7654409830956894, + "learning_rate": 9.798677319640998e-06, + "loss": 0.5775, + "step": 5357 + }, + { + "epoch": 0.11793968743121286, + "grad_norm": 4.867255261473943, + "learning_rate": 9.798577176927593e-06, + "loss": 0.3721, + "step": 5358 + }, + { + "epoch": 0.11796169931763152, + "grad_norm": 2.0878323965311085, + "learning_rate": 9.798477009825677e-06, + "loss": 0.3809, + "step": 5359 + }, + { + "epoch": 0.11798371120405018, + "grad_norm": 1.508770677701525, + "learning_rate": 9.798376818335763e-06, + "loss": 0.2915, + "step": 5360 + }, + { + "epoch": 0.11800572309046885, + "grad_norm": 1.9634187786497093, + "learning_rate": 9.798276602458354e-06, + "loss": 0.4278, + "step": 5361 + }, + { + "epoch": 0.11802773497688752, + "grad_norm": 1.5403327361874195, + "learning_rate": 9.798176362193965e-06, + "loss": 0.4196, + "step": 5362 + }, + { + "epoch": 0.11804974686330619, + "grad_norm": 2.200179980264402, + "learning_rate": 9.798076097543104e-06, + "loss": 0.4957, + "step": 5363 + }, + { + "epoch": 0.11807175874972485, + "grad_norm": 1.544512014052504, + "learning_rate": 9.79797580850628e-06, + "loss": 0.4025, + "step": 5364 + }, + { + "epoch": 0.11809377063614351, + "grad_norm": 1.8701943424275442, + "learning_rate": 9.797875495084003e-06, + "loss": 0.3804, + "step": 5365 + }, + { + "epoch": 0.11811578252256219, + "grad_norm": 1.701364933411452, + "learning_rate": 9.797775157276782e-06, + "loss": 0.3881, + "step": 5366 + }, + { + "epoch": 0.11813779440898085, + "grad_norm": 1.7814125025677885, + "learning_rate": 9.79767479508513e-06, + "loss": 0.3474, + "step": 5367 + }, + { + "epoch": 0.11815980629539952, + "grad_norm": 1.8361765517827802, + "learning_rate": 9.797574408509553e-06, + "loss": 0.3686, + "step": 5368 + }, + { + "epoch": 0.11818181818181818, + "grad_norm": 1.3312400004867064, + "learning_rate": 9.797473997550563e-06, + "loss": 0.2835, + "step": 5369 + }, + { + "epoch": 0.11820383006823684, + "grad_norm": 1.6132441606235448, + "learning_rate": 9.797373562208673e-06, + "loss": 0.4129, + "step": 5370 + }, + { + "epoch": 0.11822584195465552, + "grad_norm": 1.4544670245920057, + "learning_rate": 9.797273102484388e-06, + "loss": 0.3565, + "step": 5371 + }, + { + "epoch": 0.11824785384107418, + "grad_norm": 1.6551062148472853, + "learning_rate": 9.797172618378223e-06, + "loss": 0.3875, + "step": 5372 + }, + { + "epoch": 0.11826986572749285, + "grad_norm": 2.2213340481805313, + "learning_rate": 9.797072109890687e-06, + "loss": 0.3735, + "step": 5373 + }, + { + "epoch": 0.11829187761391151, + "grad_norm": 0.7133712834653793, + "learning_rate": 9.79697157702229e-06, + "loss": 0.5649, + "step": 5374 + }, + { + "epoch": 0.11831388950033017, + "grad_norm": 2.55562759020622, + "learning_rate": 9.796871019773544e-06, + "loss": 0.416, + "step": 5375 + }, + { + "epoch": 0.11833590138674885, + "grad_norm": 0.6502788753661816, + "learning_rate": 9.79677043814496e-06, + "loss": 0.5538, + "step": 5376 + }, + { + "epoch": 0.11835791327316751, + "grad_norm": 1.5482503335268352, + "learning_rate": 9.796669832137049e-06, + "loss": 0.3772, + "step": 5377 + }, + { + "epoch": 0.11837992515958617, + "grad_norm": 1.762334393336996, + "learning_rate": 9.796569201750323e-06, + "loss": 0.3071, + "step": 5378 + }, + { + "epoch": 0.11840193704600484, + "grad_norm": 1.224433508322814, + "learning_rate": 9.796468546985294e-06, + "loss": 0.3195, + "step": 5379 + }, + { + "epoch": 0.11842394893242351, + "grad_norm": 1.58466266593824, + "learning_rate": 9.796367867842471e-06, + "loss": 0.3405, + "step": 5380 + }, + { + "epoch": 0.11844596081884218, + "grad_norm": 1.3863820789358707, + "learning_rate": 9.796267164322368e-06, + "loss": 0.372, + "step": 5381 + }, + { + "epoch": 0.11846797270526084, + "grad_norm": 0.6539370736588301, + "learning_rate": 9.796166436425496e-06, + "loss": 0.5348, + "step": 5382 + }, + { + "epoch": 0.1184899845916795, + "grad_norm": 1.420015052678749, + "learning_rate": 9.796065684152367e-06, + "loss": 0.3515, + "step": 5383 + }, + { + "epoch": 0.11851199647809817, + "grad_norm": 1.691471578081738, + "learning_rate": 9.795964907503494e-06, + "loss": 0.3537, + "step": 5384 + }, + { + "epoch": 0.11853400836451684, + "grad_norm": 1.436778086939591, + "learning_rate": 9.795864106479387e-06, + "loss": 0.3647, + "step": 5385 + }, + { + "epoch": 0.1185560202509355, + "grad_norm": 1.4254360169265456, + "learning_rate": 9.795763281080561e-06, + "loss": 0.356, + "step": 5386 + }, + { + "epoch": 0.11857803213735417, + "grad_norm": 1.5716813326152803, + "learning_rate": 9.795662431307526e-06, + "loss": 0.3673, + "step": 5387 + }, + { + "epoch": 0.11860004402377283, + "grad_norm": 1.8357401752913263, + "learning_rate": 9.795561557160795e-06, + "loss": 0.408, + "step": 5388 + }, + { + "epoch": 0.11862205591019151, + "grad_norm": 1.9726767916252206, + "learning_rate": 9.795460658640882e-06, + "loss": 0.3879, + "step": 5389 + }, + { + "epoch": 0.11864406779661017, + "grad_norm": 1.6807364467106378, + "learning_rate": 9.795359735748299e-06, + "loss": 0.4092, + "step": 5390 + }, + { + "epoch": 0.11866607968302884, + "grad_norm": 0.6718532737578855, + "learning_rate": 9.79525878848356e-06, + "loss": 0.5583, + "step": 5391 + }, + { + "epoch": 0.1186880915694475, + "grad_norm": 1.4133148779588947, + "learning_rate": 9.795157816847175e-06, + "loss": 0.2937, + "step": 5392 + }, + { + "epoch": 0.11871010345586616, + "grad_norm": 1.8566248875242084, + "learning_rate": 9.795056820839661e-06, + "loss": 0.3713, + "step": 5393 + }, + { + "epoch": 0.11873211534228484, + "grad_norm": 1.713755868890237, + "learning_rate": 9.794955800461528e-06, + "loss": 0.4263, + "step": 5394 + }, + { + "epoch": 0.1187541272287035, + "grad_norm": 1.8249644077801956, + "learning_rate": 9.794854755713293e-06, + "loss": 0.3788, + "step": 5395 + }, + { + "epoch": 0.11877613911512216, + "grad_norm": 1.3457672964578913, + "learning_rate": 9.794753686595466e-06, + "loss": 0.2392, + "step": 5396 + }, + { + "epoch": 0.11879815100154083, + "grad_norm": 1.6490622826387047, + "learning_rate": 9.794652593108563e-06, + "loss": 0.371, + "step": 5397 + }, + { + "epoch": 0.1188201628879595, + "grad_norm": 0.69458410641538, + "learning_rate": 9.794551475253097e-06, + "loss": 0.5588, + "step": 5398 + }, + { + "epoch": 0.11884217477437817, + "grad_norm": 1.4856136035173975, + "learning_rate": 9.794450333029581e-06, + "loss": 0.3458, + "step": 5399 + }, + { + "epoch": 0.11886418666079683, + "grad_norm": 1.7866452634273273, + "learning_rate": 9.794349166438533e-06, + "loss": 0.4155, + "step": 5400 + }, + { + "epoch": 0.1188861985472155, + "grad_norm": 1.6439829742030503, + "learning_rate": 9.794247975480461e-06, + "loss": 0.3453, + "step": 5401 + }, + { + "epoch": 0.11890821043363416, + "grad_norm": 1.9565188041252666, + "learning_rate": 9.794146760155884e-06, + "loss": 0.4548, + "step": 5402 + }, + { + "epoch": 0.11893022232005283, + "grad_norm": 1.4114973403378586, + "learning_rate": 9.794045520465315e-06, + "loss": 0.3311, + "step": 5403 + }, + { + "epoch": 0.1189522342064715, + "grad_norm": 1.7322845453217002, + "learning_rate": 9.793944256409267e-06, + "loss": 0.3277, + "step": 5404 + }, + { + "epoch": 0.11897424609289016, + "grad_norm": 1.9267395411977986, + "learning_rate": 9.793842967988257e-06, + "loss": 0.3668, + "step": 5405 + }, + { + "epoch": 0.11899625797930882, + "grad_norm": 1.2953746676309372, + "learning_rate": 9.7937416552028e-06, + "loss": 0.3016, + "step": 5406 + }, + { + "epoch": 0.1190182698657275, + "grad_norm": 1.7372485577615686, + "learning_rate": 9.793640318053408e-06, + "loss": 0.478, + "step": 5407 + }, + { + "epoch": 0.11904028175214616, + "grad_norm": 1.5394317333892376, + "learning_rate": 9.793538956540598e-06, + "loss": 0.3555, + "step": 5408 + }, + { + "epoch": 0.11906229363856483, + "grad_norm": 0.6728623044069636, + "learning_rate": 9.793437570664887e-06, + "loss": 0.5517, + "step": 5409 + }, + { + "epoch": 0.11908430552498349, + "grad_norm": 1.4954662858666306, + "learning_rate": 9.793336160426787e-06, + "loss": 0.4386, + "step": 5410 + }, + { + "epoch": 0.11910631741140215, + "grad_norm": 1.8146053928927506, + "learning_rate": 9.793234725826814e-06, + "loss": 0.3796, + "step": 5411 + }, + { + "epoch": 0.11912832929782083, + "grad_norm": 1.5993532997102289, + "learning_rate": 9.793133266865485e-06, + "loss": 0.3828, + "step": 5412 + }, + { + "epoch": 0.11915034118423949, + "grad_norm": 1.6871739543594495, + "learning_rate": 9.793031783543316e-06, + "loss": 0.4053, + "step": 5413 + }, + { + "epoch": 0.11917235307065815, + "grad_norm": 1.5446104317278169, + "learning_rate": 9.79293027586082e-06, + "loss": 0.3676, + "step": 5414 + }, + { + "epoch": 0.11919436495707682, + "grad_norm": 1.6062622588465285, + "learning_rate": 9.792828743818517e-06, + "loss": 0.3636, + "step": 5415 + }, + { + "epoch": 0.11921637684349548, + "grad_norm": 1.6067630698425948, + "learning_rate": 9.792727187416918e-06, + "loss": 0.3668, + "step": 5416 + }, + { + "epoch": 0.11923838872991416, + "grad_norm": 1.2610842591789213, + "learning_rate": 9.792625606656542e-06, + "loss": 0.4168, + "step": 5417 + }, + { + "epoch": 0.11926040061633282, + "grad_norm": 1.7160631869154173, + "learning_rate": 9.792524001537906e-06, + "loss": 0.3765, + "step": 5418 + }, + { + "epoch": 0.11928241250275148, + "grad_norm": 3.1878340968944836, + "learning_rate": 9.792422372061526e-06, + "loss": 0.4183, + "step": 5419 + }, + { + "epoch": 0.11930442438917015, + "grad_norm": 2.064349124379395, + "learning_rate": 9.792320718227917e-06, + "loss": 0.2996, + "step": 5420 + }, + { + "epoch": 0.11932643627558882, + "grad_norm": 1.5051952677791733, + "learning_rate": 9.792219040037599e-06, + "loss": 0.3208, + "step": 5421 + }, + { + "epoch": 0.11934844816200749, + "grad_norm": 4.276595424637522, + "learning_rate": 9.792117337491085e-06, + "loss": 0.3644, + "step": 5422 + }, + { + "epoch": 0.11937046004842615, + "grad_norm": 1.7886479734242202, + "learning_rate": 9.792015610588892e-06, + "loss": 0.4059, + "step": 5423 + }, + { + "epoch": 0.11939247193484481, + "grad_norm": 1.969665765710699, + "learning_rate": 9.79191385933154e-06, + "loss": 0.4247, + "step": 5424 + }, + { + "epoch": 0.11941448382126348, + "grad_norm": 1.4451700821348419, + "learning_rate": 9.791812083719543e-06, + "loss": 0.3592, + "step": 5425 + }, + { + "epoch": 0.11943649570768215, + "grad_norm": 0.7399323206799525, + "learning_rate": 9.79171028375342e-06, + "loss": 0.591, + "step": 5426 + }, + { + "epoch": 0.11945850759410082, + "grad_norm": 1.7595179534345438, + "learning_rate": 9.791608459433691e-06, + "loss": 0.3246, + "step": 5427 + }, + { + "epoch": 0.11948051948051948, + "grad_norm": 0.6333402880333274, + "learning_rate": 9.791506610760866e-06, + "loss": 0.5563, + "step": 5428 + }, + { + "epoch": 0.11950253136693814, + "grad_norm": 2.280476213848903, + "learning_rate": 9.79140473773547e-06, + "loss": 0.4853, + "step": 5429 + }, + { + "epoch": 0.11952454325335682, + "grad_norm": 1.3833290016667619, + "learning_rate": 9.791302840358019e-06, + "loss": 0.3519, + "step": 5430 + }, + { + "epoch": 0.11954655513977548, + "grad_norm": 1.5278264932140804, + "learning_rate": 9.79120091862903e-06, + "loss": 0.3981, + "step": 5431 + }, + { + "epoch": 0.11956856702619414, + "grad_norm": 1.660136701145227, + "learning_rate": 9.791098972549018e-06, + "loss": 0.394, + "step": 5432 + }, + { + "epoch": 0.11959057891261281, + "grad_norm": 2.4086455041753276, + "learning_rate": 9.790997002118507e-06, + "loss": 0.3024, + "step": 5433 + }, + { + "epoch": 0.11961259079903147, + "grad_norm": 1.7167903845320567, + "learning_rate": 9.79089500733801e-06, + "loss": 0.3814, + "step": 5434 + }, + { + "epoch": 0.11963460268545015, + "grad_norm": 1.345170063671517, + "learning_rate": 9.790792988208052e-06, + "loss": 0.3244, + "step": 5435 + }, + { + "epoch": 0.11965661457186881, + "grad_norm": 1.6334627802923247, + "learning_rate": 9.790690944729143e-06, + "loss": 0.4409, + "step": 5436 + }, + { + "epoch": 0.11967862645828747, + "grad_norm": 1.5334577552291142, + "learning_rate": 9.790588876901807e-06, + "loss": 0.3153, + "step": 5437 + }, + { + "epoch": 0.11970063834470614, + "grad_norm": 1.5530255626212024, + "learning_rate": 9.790486784726563e-06, + "loss": 0.3685, + "step": 5438 + }, + { + "epoch": 0.11972265023112481, + "grad_norm": 1.790369997943802, + "learning_rate": 9.790384668203928e-06, + "loss": 0.3172, + "step": 5439 + }, + { + "epoch": 0.11974466211754348, + "grad_norm": 4.344217484551098, + "learning_rate": 9.790282527334421e-06, + "loss": 0.3744, + "step": 5440 + }, + { + "epoch": 0.11976667400396214, + "grad_norm": 1.7019453452673627, + "learning_rate": 9.790180362118562e-06, + "loss": 0.3676, + "step": 5441 + }, + { + "epoch": 0.1197886858903808, + "grad_norm": 1.6599341637761138, + "learning_rate": 9.790078172556869e-06, + "loss": 0.4491, + "step": 5442 + }, + { + "epoch": 0.11981069777679947, + "grad_norm": 1.4023759959721191, + "learning_rate": 9.789975958649863e-06, + "loss": 0.3382, + "step": 5443 + }, + { + "epoch": 0.11983270966321814, + "grad_norm": 1.6123140248886543, + "learning_rate": 9.789873720398063e-06, + "loss": 0.3687, + "step": 5444 + }, + { + "epoch": 0.1198547215496368, + "grad_norm": 2.045158528033621, + "learning_rate": 9.789771457801989e-06, + "loss": 0.3698, + "step": 5445 + }, + { + "epoch": 0.11987673343605547, + "grad_norm": 1.404228171029643, + "learning_rate": 9.789669170862159e-06, + "loss": 0.289, + "step": 5446 + }, + { + "epoch": 0.11989874532247413, + "grad_norm": 1.8121208307308543, + "learning_rate": 9.789566859579093e-06, + "loss": 0.3483, + "step": 5447 + }, + { + "epoch": 0.11992075720889281, + "grad_norm": 0.7598213377075493, + "learning_rate": 9.789464523953314e-06, + "loss": 0.5429, + "step": 5448 + }, + { + "epoch": 0.11994276909531147, + "grad_norm": 2.0286396615386932, + "learning_rate": 9.78936216398534e-06, + "loss": 0.3993, + "step": 5449 + }, + { + "epoch": 0.11996478098173013, + "grad_norm": 0.6474668644328404, + "learning_rate": 9.78925977967569e-06, + "loss": 0.5626, + "step": 5450 + }, + { + "epoch": 0.1199867928681488, + "grad_norm": 1.6426906932809786, + "learning_rate": 9.789157371024888e-06, + "loss": 0.3348, + "step": 5451 + }, + { + "epoch": 0.12000880475456746, + "grad_norm": 1.5253386494754424, + "learning_rate": 9.78905493803345e-06, + "loss": 0.3721, + "step": 5452 + }, + { + "epoch": 0.12003081664098614, + "grad_norm": 1.5389465082280442, + "learning_rate": 9.788952480701902e-06, + "loss": 0.3473, + "step": 5453 + }, + { + "epoch": 0.1200528285274048, + "grad_norm": 1.7020704562827256, + "learning_rate": 9.788849999030759e-06, + "loss": 0.3983, + "step": 5454 + }, + { + "epoch": 0.12007484041382346, + "grad_norm": 1.9745210414017775, + "learning_rate": 9.788747493020543e-06, + "loss": 0.39, + "step": 5455 + }, + { + "epoch": 0.12009685230024213, + "grad_norm": 2.4134380360761694, + "learning_rate": 9.788644962671778e-06, + "loss": 0.4932, + "step": 5456 + }, + { + "epoch": 0.1201188641866608, + "grad_norm": 1.4593486400816067, + "learning_rate": 9.788542407984984e-06, + "loss": 0.3459, + "step": 5457 + }, + { + "epoch": 0.12014087607307947, + "grad_norm": 1.7030685803528491, + "learning_rate": 9.78843982896068e-06, + "loss": 0.3526, + "step": 5458 + }, + { + "epoch": 0.12016288795949813, + "grad_norm": 1.5140639038613943, + "learning_rate": 9.78833722559939e-06, + "loss": 0.3509, + "step": 5459 + }, + { + "epoch": 0.12018489984591679, + "grad_norm": 0.9931670162244075, + "learning_rate": 9.788234597901637e-06, + "loss": 0.552, + "step": 5460 + }, + { + "epoch": 0.12020691173233546, + "grad_norm": 2.5892726425127033, + "learning_rate": 9.788131945867938e-06, + "loss": 0.4618, + "step": 5461 + }, + { + "epoch": 0.12022892361875413, + "grad_norm": 1.5650623688189136, + "learning_rate": 9.788029269498817e-06, + "loss": 0.3495, + "step": 5462 + }, + { + "epoch": 0.1202509355051728, + "grad_norm": 3.205732413398955, + "learning_rate": 9.787926568794795e-06, + "loss": 0.5427, + "step": 5463 + }, + { + "epoch": 0.12027294739159146, + "grad_norm": 1.979185685332324, + "learning_rate": 9.787823843756396e-06, + "loss": 0.3831, + "step": 5464 + }, + { + "epoch": 0.12029495927801012, + "grad_norm": 1.944570379173569, + "learning_rate": 9.787721094384141e-06, + "loss": 0.3794, + "step": 5465 + }, + { + "epoch": 0.12031697116442878, + "grad_norm": 1.9681195704849332, + "learning_rate": 9.787618320678551e-06, + "loss": 0.3617, + "step": 5466 + }, + { + "epoch": 0.12033898305084746, + "grad_norm": 2.340140381503634, + "learning_rate": 9.78751552264015e-06, + "loss": 0.3328, + "step": 5467 + }, + { + "epoch": 0.12036099493726612, + "grad_norm": 1.55209694660997, + "learning_rate": 9.78741270026946e-06, + "loss": 0.4341, + "step": 5468 + }, + { + "epoch": 0.12038300682368479, + "grad_norm": 1.6796537311622544, + "learning_rate": 9.787309853567002e-06, + "loss": 0.409, + "step": 5469 + }, + { + "epoch": 0.12040501871010345, + "grad_norm": 1.9565804500488402, + "learning_rate": 9.7872069825333e-06, + "loss": 0.3488, + "step": 5470 + }, + { + "epoch": 0.12042703059652213, + "grad_norm": 1.846667668832839, + "learning_rate": 9.78710408716888e-06, + "loss": 0.31, + "step": 5471 + }, + { + "epoch": 0.12044904248294079, + "grad_norm": 1.4618996555933144, + "learning_rate": 9.787001167474259e-06, + "loss": 0.3381, + "step": 5472 + }, + { + "epoch": 0.12047105436935945, + "grad_norm": 2.043265967109054, + "learning_rate": 9.786898223449964e-06, + "loss": 0.3042, + "step": 5473 + }, + { + "epoch": 0.12049306625577812, + "grad_norm": 1.966268311278586, + "learning_rate": 9.786795255096517e-06, + "loss": 0.2862, + "step": 5474 + }, + { + "epoch": 0.12051507814219678, + "grad_norm": 1.9459395087803193, + "learning_rate": 9.786692262414441e-06, + "loss": 0.3787, + "step": 5475 + }, + { + "epoch": 0.12053709002861546, + "grad_norm": 0.914109546568577, + "learning_rate": 9.786589245404261e-06, + "loss": 0.6068, + "step": 5476 + }, + { + "epoch": 0.12055910191503412, + "grad_norm": 14.34600132557383, + "learning_rate": 9.786486204066499e-06, + "loss": 0.4121, + "step": 5477 + }, + { + "epoch": 0.12058111380145278, + "grad_norm": 0.7411685658528744, + "learning_rate": 9.786383138401679e-06, + "loss": 0.5389, + "step": 5478 + }, + { + "epoch": 0.12060312568787145, + "grad_norm": 1.6090532702032563, + "learning_rate": 9.786280048410327e-06, + "loss": 0.2571, + "step": 5479 + }, + { + "epoch": 0.12062513757429012, + "grad_norm": 1.9250425512605098, + "learning_rate": 9.786176934092964e-06, + "loss": 0.3897, + "step": 5480 + }, + { + "epoch": 0.12064714946070879, + "grad_norm": 1.7307393508349347, + "learning_rate": 9.786073795450113e-06, + "loss": 0.3942, + "step": 5481 + }, + { + "epoch": 0.12066916134712745, + "grad_norm": 1.870701277909409, + "learning_rate": 9.785970632482302e-06, + "loss": 0.3336, + "step": 5482 + }, + { + "epoch": 0.12069117323354611, + "grad_norm": 1.7045387265201135, + "learning_rate": 9.785867445190053e-06, + "loss": 0.4118, + "step": 5483 + }, + { + "epoch": 0.12071318511996477, + "grad_norm": 1.8948905625672172, + "learning_rate": 9.785764233573894e-06, + "loss": 0.4475, + "step": 5484 + }, + { + "epoch": 0.12073519700638345, + "grad_norm": 2.7693721711698847, + "learning_rate": 9.785660997634344e-06, + "loss": 0.3696, + "step": 5485 + }, + { + "epoch": 0.12075720889280211, + "grad_norm": 1.8955772287841288, + "learning_rate": 9.78555773737193e-06, + "loss": 0.4095, + "step": 5486 + }, + { + "epoch": 0.12077922077922078, + "grad_norm": 1.436043679632332, + "learning_rate": 9.785454452787178e-06, + "loss": 0.322, + "step": 5487 + }, + { + "epoch": 0.12080123266563944, + "grad_norm": 1.1676235490627154, + "learning_rate": 9.785351143880613e-06, + "loss": 0.3347, + "step": 5488 + }, + { + "epoch": 0.12082324455205812, + "grad_norm": 2.327281203361457, + "learning_rate": 9.785247810652758e-06, + "loss": 0.3319, + "step": 5489 + }, + { + "epoch": 0.12084525643847678, + "grad_norm": 2.3374438530413992, + "learning_rate": 9.78514445310414e-06, + "loss": 0.3629, + "step": 5490 + }, + { + "epoch": 0.12086726832489544, + "grad_norm": 1.5741489506672603, + "learning_rate": 9.785041071235282e-06, + "loss": 0.4013, + "step": 5491 + }, + { + "epoch": 0.1208892802113141, + "grad_norm": 1.7630257140406154, + "learning_rate": 9.784937665046714e-06, + "loss": 0.3488, + "step": 5492 + }, + { + "epoch": 0.12091129209773277, + "grad_norm": 1.165898340818819, + "learning_rate": 9.784834234538955e-06, + "loss": 0.2832, + "step": 5493 + }, + { + "epoch": 0.12093330398415145, + "grad_norm": 1.4608496474951642, + "learning_rate": 9.784730779712537e-06, + "loss": 0.4622, + "step": 5494 + }, + { + "epoch": 0.12095531587057011, + "grad_norm": 1.4368019451429848, + "learning_rate": 9.784627300567983e-06, + "loss": 0.4294, + "step": 5495 + }, + { + "epoch": 0.12097732775698877, + "grad_norm": 1.2255471466133276, + "learning_rate": 9.784523797105819e-06, + "loss": 0.6037, + "step": 5496 + }, + { + "epoch": 0.12099933964340744, + "grad_norm": 1.6571957416209693, + "learning_rate": 9.78442026932657e-06, + "loss": 0.3917, + "step": 5497 + }, + { + "epoch": 0.12102135152982611, + "grad_norm": 1.8815800869826649, + "learning_rate": 9.784316717230765e-06, + "loss": 0.3764, + "step": 5498 + }, + { + "epoch": 0.12104336341624478, + "grad_norm": 1.9690652377722142, + "learning_rate": 9.784213140818925e-06, + "loss": 0.3604, + "step": 5499 + }, + { + "epoch": 0.12106537530266344, + "grad_norm": 1.469792483927316, + "learning_rate": 9.784109540091585e-06, + "loss": 0.3554, + "step": 5500 + }, + { + "epoch": 0.1210873871890821, + "grad_norm": 1.3458227033991432, + "learning_rate": 9.784005915049263e-06, + "loss": 0.3618, + "step": 5501 + }, + { + "epoch": 0.12110939907550076, + "grad_norm": 1.7849322595624517, + "learning_rate": 9.78390226569249e-06, + "loss": 0.3408, + "step": 5502 + }, + { + "epoch": 0.12113141096191944, + "grad_norm": 1.4046809816742216, + "learning_rate": 9.783798592021792e-06, + "loss": 0.3082, + "step": 5503 + }, + { + "epoch": 0.1211534228483381, + "grad_norm": 3.9211100325371784, + "learning_rate": 9.783694894037697e-06, + "loss": 0.4072, + "step": 5504 + }, + { + "epoch": 0.12117543473475677, + "grad_norm": 1.5780625978628797, + "learning_rate": 9.783591171740729e-06, + "loss": 0.3114, + "step": 5505 + }, + { + "epoch": 0.12119744662117543, + "grad_norm": 1.6912214967511499, + "learning_rate": 9.783487425131418e-06, + "loss": 0.4119, + "step": 5506 + }, + { + "epoch": 0.1212194585075941, + "grad_norm": 1.6153656931102658, + "learning_rate": 9.783383654210292e-06, + "loss": 0.3704, + "step": 5507 + }, + { + "epoch": 0.12124147039401277, + "grad_norm": 1.7732860408710536, + "learning_rate": 9.783279858977875e-06, + "loss": 0.3634, + "step": 5508 + }, + { + "epoch": 0.12126348228043143, + "grad_norm": 1.5384615319421908, + "learning_rate": 9.783176039434697e-06, + "loss": 0.3035, + "step": 5509 + }, + { + "epoch": 0.1212854941668501, + "grad_norm": 1.517490942069978, + "learning_rate": 9.783072195581284e-06, + "loss": 0.3096, + "step": 5510 + }, + { + "epoch": 0.12130750605326876, + "grad_norm": 2.072207122445589, + "learning_rate": 9.782968327418165e-06, + "loss": 0.4468, + "step": 5511 + }, + { + "epoch": 0.12132951793968744, + "grad_norm": 1.0008351813204004, + "learning_rate": 9.78286443494587e-06, + "loss": 0.568, + "step": 5512 + }, + { + "epoch": 0.1213515298261061, + "grad_norm": 1.5564598837151318, + "learning_rate": 9.782760518164922e-06, + "loss": 0.2903, + "step": 5513 + }, + { + "epoch": 0.12137354171252476, + "grad_norm": 1.7471292750820027, + "learning_rate": 9.782656577075853e-06, + "loss": 0.3821, + "step": 5514 + }, + { + "epoch": 0.12139555359894343, + "grad_norm": 1.6323625653942901, + "learning_rate": 9.78255261167919e-06, + "loss": 0.3328, + "step": 5515 + }, + { + "epoch": 0.12141756548536209, + "grad_norm": 1.961321886086374, + "learning_rate": 9.782448621975462e-06, + "loss": 0.4858, + "step": 5516 + }, + { + "epoch": 0.12143957737178077, + "grad_norm": 0.6726365121338304, + "learning_rate": 9.782344607965196e-06, + "loss": 0.5679, + "step": 5517 + }, + { + "epoch": 0.12146158925819943, + "grad_norm": 1.585281445819245, + "learning_rate": 9.782240569648922e-06, + "loss": 0.4091, + "step": 5518 + }, + { + "epoch": 0.12148360114461809, + "grad_norm": 2.869986821952932, + "learning_rate": 9.782136507027169e-06, + "loss": 0.3738, + "step": 5519 + }, + { + "epoch": 0.12150561303103675, + "grad_norm": 0.656017353539156, + "learning_rate": 9.782032420100466e-06, + "loss": 0.552, + "step": 5520 + }, + { + "epoch": 0.12152762491745543, + "grad_norm": 1.305277851073219, + "learning_rate": 9.781928308869341e-06, + "loss": 0.3446, + "step": 5521 + }, + { + "epoch": 0.1215496368038741, + "grad_norm": 1.5252903307983077, + "learning_rate": 9.781824173334323e-06, + "loss": 0.319, + "step": 5522 + }, + { + "epoch": 0.12157164869029276, + "grad_norm": 1.237086165798652, + "learning_rate": 9.781720013495942e-06, + "loss": 0.321, + "step": 5523 + }, + { + "epoch": 0.12159366057671142, + "grad_norm": 1.9962252796557325, + "learning_rate": 9.781615829354727e-06, + "loss": 0.4437, + "step": 5524 + }, + { + "epoch": 0.12161567246313008, + "grad_norm": 0.7480839259716411, + "learning_rate": 9.781511620911208e-06, + "loss": 0.5551, + "step": 5525 + }, + { + "epoch": 0.12163768434954876, + "grad_norm": 1.318452177161577, + "learning_rate": 9.781407388165913e-06, + "loss": 0.3475, + "step": 5526 + }, + { + "epoch": 0.12165969623596742, + "grad_norm": 1.52615013460836, + "learning_rate": 9.781303131119374e-06, + "loss": 0.4204, + "step": 5527 + }, + { + "epoch": 0.12168170812238609, + "grad_norm": 0.6780459776817056, + "learning_rate": 9.781198849772121e-06, + "loss": 0.5495, + "step": 5528 + }, + { + "epoch": 0.12170372000880475, + "grad_norm": 1.4790688619411116, + "learning_rate": 9.781094544124682e-06, + "loss": 0.4325, + "step": 5529 + }, + { + "epoch": 0.12172573189522343, + "grad_norm": 0.6308573598978017, + "learning_rate": 9.78099021417759e-06, + "loss": 0.5439, + "step": 5530 + }, + { + "epoch": 0.12174774378164209, + "grad_norm": 1.5518706387372776, + "learning_rate": 9.78088585993137e-06, + "loss": 0.3765, + "step": 5531 + }, + { + "epoch": 0.12176975566806075, + "grad_norm": 2.1432791994544678, + "learning_rate": 9.780781481386558e-06, + "loss": 0.4246, + "step": 5532 + }, + { + "epoch": 0.12179176755447942, + "grad_norm": 1.484592047423357, + "learning_rate": 9.780677078543681e-06, + "loss": 0.3228, + "step": 5533 + }, + { + "epoch": 0.12181377944089808, + "grad_norm": 0.667002361368967, + "learning_rate": 9.780572651403272e-06, + "loss": 0.5856, + "step": 5534 + }, + { + "epoch": 0.12183579132731676, + "grad_norm": 1.3681848405214005, + "learning_rate": 9.78046819996586e-06, + "loss": 0.3879, + "step": 5535 + }, + { + "epoch": 0.12185780321373542, + "grad_norm": 2.447880096059983, + "learning_rate": 9.780363724231977e-06, + "loss": 0.4341, + "step": 5536 + }, + { + "epoch": 0.12187981510015408, + "grad_norm": 1.5229837219720341, + "learning_rate": 9.780259224202153e-06, + "loss": 0.2945, + "step": 5537 + }, + { + "epoch": 0.12190182698657274, + "grad_norm": 1.5316200843458745, + "learning_rate": 9.78015469987692e-06, + "loss": 0.4133, + "step": 5538 + }, + { + "epoch": 0.12192383887299142, + "grad_norm": 1.2330213750842818, + "learning_rate": 9.780050151256808e-06, + "loss": 0.3118, + "step": 5539 + }, + { + "epoch": 0.12194585075941008, + "grad_norm": 1.6022686243053477, + "learning_rate": 9.779945578342349e-06, + "loss": 0.4769, + "step": 5540 + }, + { + "epoch": 0.12196786264582875, + "grad_norm": 1.2673467899163167, + "learning_rate": 9.779840981134074e-06, + "loss": 0.3232, + "step": 5541 + }, + { + "epoch": 0.12198987453224741, + "grad_norm": 1.7134894639567617, + "learning_rate": 9.779736359632518e-06, + "loss": 0.4005, + "step": 5542 + }, + { + "epoch": 0.12201188641866607, + "grad_norm": 1.442612925506877, + "learning_rate": 9.779631713838207e-06, + "loss": 0.2959, + "step": 5543 + }, + { + "epoch": 0.12203389830508475, + "grad_norm": 1.6557843098947675, + "learning_rate": 9.779527043751678e-06, + "loss": 0.4517, + "step": 5544 + }, + { + "epoch": 0.12205591019150341, + "grad_norm": 3.6931482624469343, + "learning_rate": 9.779422349373459e-06, + "loss": 0.4273, + "step": 5545 + }, + { + "epoch": 0.12207792207792208, + "grad_norm": 1.7510361961036085, + "learning_rate": 9.779317630704086e-06, + "loss": 0.4117, + "step": 5546 + }, + { + "epoch": 0.12209993396434074, + "grad_norm": 1.4351861487297424, + "learning_rate": 9.779212887744089e-06, + "loss": 0.2507, + "step": 5547 + }, + { + "epoch": 0.12212194585075942, + "grad_norm": 1.8996421230656564, + "learning_rate": 9.779108120494e-06, + "loss": 0.269, + "step": 5548 + }, + { + "epoch": 0.12214395773717808, + "grad_norm": 2.081110687735983, + "learning_rate": 9.779003328954351e-06, + "loss": 0.3906, + "step": 5549 + }, + { + "epoch": 0.12216596962359674, + "grad_norm": 1.67096818567598, + "learning_rate": 9.778898513125676e-06, + "loss": 0.4011, + "step": 5550 + }, + { + "epoch": 0.1221879815100154, + "grad_norm": 2.722094302735293, + "learning_rate": 9.778793673008508e-06, + "loss": 0.3634, + "step": 5551 + }, + { + "epoch": 0.12220999339643407, + "grad_norm": 0.6793409320735652, + "learning_rate": 9.778688808603378e-06, + "loss": 0.5652, + "step": 5552 + }, + { + "epoch": 0.12223200528285275, + "grad_norm": 1.7415534782598205, + "learning_rate": 9.778583919910822e-06, + "loss": 0.4373, + "step": 5553 + }, + { + "epoch": 0.12225401716927141, + "grad_norm": 2.460634262755268, + "learning_rate": 9.77847900693137e-06, + "loss": 0.3609, + "step": 5554 + }, + { + "epoch": 0.12227602905569007, + "grad_norm": 0.6288229961012467, + "learning_rate": 9.778374069665558e-06, + "loss": 0.5606, + "step": 5555 + }, + { + "epoch": 0.12229804094210873, + "grad_norm": 1.514336469864021, + "learning_rate": 9.778269108113914e-06, + "loss": 0.3804, + "step": 5556 + }, + { + "epoch": 0.1223200528285274, + "grad_norm": 1.3467034521459413, + "learning_rate": 9.778164122276979e-06, + "loss": 0.2733, + "step": 5557 + }, + { + "epoch": 0.12234206471494607, + "grad_norm": 2.8059678966736366, + "learning_rate": 9.77805911215528e-06, + "loss": 0.4324, + "step": 5558 + }, + { + "epoch": 0.12236407660136474, + "grad_norm": 0.6747382929217282, + "learning_rate": 9.777954077749355e-06, + "loss": 0.5685, + "step": 5559 + }, + { + "epoch": 0.1223860884877834, + "grad_norm": 4.5678769276911675, + "learning_rate": 9.777849019059738e-06, + "loss": 0.4365, + "step": 5560 + }, + { + "epoch": 0.12240810037420206, + "grad_norm": 0.644603794798911, + "learning_rate": 9.777743936086958e-06, + "loss": 0.5503, + "step": 5561 + }, + { + "epoch": 0.12243011226062074, + "grad_norm": 0.6959059383208355, + "learning_rate": 9.777638828831554e-06, + "loss": 0.5781, + "step": 5562 + }, + { + "epoch": 0.1224521241470394, + "grad_norm": 3.3821456170086552, + "learning_rate": 9.777533697294059e-06, + "loss": 0.4253, + "step": 5563 + }, + { + "epoch": 0.12247413603345807, + "grad_norm": 1.981150468360166, + "learning_rate": 9.777428541475006e-06, + "loss": 0.2966, + "step": 5564 + }, + { + "epoch": 0.12249614791987673, + "grad_norm": 1.8375045449525524, + "learning_rate": 9.777323361374931e-06, + "loss": 0.3012, + "step": 5565 + }, + { + "epoch": 0.12251815980629539, + "grad_norm": 1.4214074209901981, + "learning_rate": 9.777218156994368e-06, + "loss": 0.3801, + "step": 5566 + }, + { + "epoch": 0.12254017169271407, + "grad_norm": 1.817708745893952, + "learning_rate": 9.77711292833385e-06, + "loss": 0.4232, + "step": 5567 + }, + { + "epoch": 0.12256218357913273, + "grad_norm": 2.1522810860949075, + "learning_rate": 9.777007675393915e-06, + "loss": 0.3829, + "step": 5568 + }, + { + "epoch": 0.1225841954655514, + "grad_norm": 1.5893889058987696, + "learning_rate": 9.776902398175096e-06, + "loss": 0.3132, + "step": 5569 + }, + { + "epoch": 0.12260620735197006, + "grad_norm": 1.5522452799790758, + "learning_rate": 9.77679709667793e-06, + "loss": 0.3106, + "step": 5570 + }, + { + "epoch": 0.12262821923838874, + "grad_norm": 1.4765994857969502, + "learning_rate": 9.776691770902948e-06, + "loss": 0.3661, + "step": 5571 + }, + { + "epoch": 0.1226502311248074, + "grad_norm": 1.8260744480493758, + "learning_rate": 9.776586420850689e-06, + "loss": 0.302, + "step": 5572 + }, + { + "epoch": 0.12267224301122606, + "grad_norm": 1.6573491568152237, + "learning_rate": 9.776481046521689e-06, + "loss": 0.355, + "step": 5573 + }, + { + "epoch": 0.12269425489764473, + "grad_norm": 1.6842594739406571, + "learning_rate": 9.776375647916479e-06, + "loss": 0.3234, + "step": 5574 + }, + { + "epoch": 0.12271626678406339, + "grad_norm": 1.6211675237502596, + "learning_rate": 9.7762702250356e-06, + "loss": 0.3238, + "step": 5575 + }, + { + "epoch": 0.12273827867048206, + "grad_norm": 1.949224211713794, + "learning_rate": 9.776164777879585e-06, + "loss": 0.4215, + "step": 5576 + }, + { + "epoch": 0.12276029055690073, + "grad_norm": 1.7133846491566342, + "learning_rate": 9.776059306448969e-06, + "loss": 0.3408, + "step": 5577 + }, + { + "epoch": 0.12278230244331939, + "grad_norm": 1.4215021211101426, + "learning_rate": 9.775953810744291e-06, + "loss": 0.3561, + "step": 5578 + }, + { + "epoch": 0.12280431432973805, + "grad_norm": 1.1662055578045725, + "learning_rate": 9.775848290766086e-06, + "loss": 0.3101, + "step": 5579 + }, + { + "epoch": 0.12282632621615673, + "grad_norm": 1.7740752482515143, + "learning_rate": 9.775742746514889e-06, + "loss": 0.404, + "step": 5580 + }, + { + "epoch": 0.1228483381025754, + "grad_norm": 1.6780531193636274, + "learning_rate": 9.775637177991236e-06, + "loss": 0.3739, + "step": 5581 + }, + { + "epoch": 0.12287034998899406, + "grad_norm": 1.3735933719495677, + "learning_rate": 9.775531585195666e-06, + "loss": 0.3675, + "step": 5582 + }, + { + "epoch": 0.12289236187541272, + "grad_norm": 3.0834559339648697, + "learning_rate": 9.775425968128716e-06, + "loss": 0.4344, + "step": 5583 + }, + { + "epoch": 0.12291437376183138, + "grad_norm": 4.911674289367969, + "learning_rate": 9.775320326790918e-06, + "loss": 0.3978, + "step": 5584 + }, + { + "epoch": 0.12293638564825006, + "grad_norm": 1.7649494261735272, + "learning_rate": 9.775214661182815e-06, + "loss": 0.4101, + "step": 5585 + }, + { + "epoch": 0.12295839753466872, + "grad_norm": 2.5406852741483283, + "learning_rate": 9.775108971304941e-06, + "loss": 0.32, + "step": 5586 + }, + { + "epoch": 0.12298040942108739, + "grad_norm": 1.515200357747415, + "learning_rate": 9.775003257157833e-06, + "loss": 0.3381, + "step": 5587 + }, + { + "epoch": 0.12300242130750605, + "grad_norm": 0.8110267455087017, + "learning_rate": 9.774897518742029e-06, + "loss": 0.5702, + "step": 5588 + }, + { + "epoch": 0.12302443319392473, + "grad_norm": 1.1305026397546745, + "learning_rate": 9.774791756058063e-06, + "loss": 0.2574, + "step": 5589 + }, + { + "epoch": 0.12304644508034339, + "grad_norm": 0.7291572197527515, + "learning_rate": 9.77468596910648e-06, + "loss": 0.5538, + "step": 5590 + }, + { + "epoch": 0.12306845696676205, + "grad_norm": 1.5360205264713684, + "learning_rate": 9.774580157887812e-06, + "loss": 0.4226, + "step": 5591 + }, + { + "epoch": 0.12309046885318072, + "grad_norm": 1.6716707115909104, + "learning_rate": 9.774474322402598e-06, + "loss": 0.323, + "step": 5592 + }, + { + "epoch": 0.12311248073959938, + "grad_norm": 1.4625626473457072, + "learning_rate": 9.774368462651376e-06, + "loss": 0.2868, + "step": 5593 + }, + { + "epoch": 0.12313449262601806, + "grad_norm": 1.601538818916845, + "learning_rate": 9.774262578634683e-06, + "loss": 0.3687, + "step": 5594 + }, + { + "epoch": 0.12315650451243672, + "grad_norm": 0.7299298292566582, + "learning_rate": 9.77415667035306e-06, + "loss": 0.5543, + "step": 5595 + }, + { + "epoch": 0.12317851639885538, + "grad_norm": 1.765603660618812, + "learning_rate": 9.774050737807044e-06, + "loss": 0.3404, + "step": 5596 + }, + { + "epoch": 0.12320052828527404, + "grad_norm": 0.722874519524946, + "learning_rate": 9.773944780997171e-06, + "loss": 0.557, + "step": 5597 + }, + { + "epoch": 0.12322254017169272, + "grad_norm": 1.4270298184791645, + "learning_rate": 9.773838799923982e-06, + "loss": 0.3063, + "step": 5598 + }, + { + "epoch": 0.12324455205811138, + "grad_norm": 0.6566065977816811, + "learning_rate": 9.773732794588015e-06, + "loss": 0.5619, + "step": 5599 + }, + { + "epoch": 0.12326656394453005, + "grad_norm": 1.5982391214599991, + "learning_rate": 9.773626764989807e-06, + "loss": 0.3433, + "step": 5600 + }, + { + "epoch": 0.12328857583094871, + "grad_norm": 0.6407987290881689, + "learning_rate": 9.773520711129903e-06, + "loss": 0.565, + "step": 5601 + }, + { + "epoch": 0.12331058771736737, + "grad_norm": 0.680939878648113, + "learning_rate": 9.773414633008833e-06, + "loss": 0.5621, + "step": 5602 + }, + { + "epoch": 0.12333259960378605, + "grad_norm": 1.868500459911778, + "learning_rate": 9.773308530627144e-06, + "loss": 0.3226, + "step": 5603 + }, + { + "epoch": 0.12335461149020471, + "grad_norm": 1.7431443311159287, + "learning_rate": 9.773202403985371e-06, + "loss": 0.2765, + "step": 5604 + }, + { + "epoch": 0.12337662337662338, + "grad_norm": 1.5334920149360118, + "learning_rate": 9.773096253084055e-06, + "loss": 0.3528, + "step": 5605 + }, + { + "epoch": 0.12339863526304204, + "grad_norm": 0.6759205445391115, + "learning_rate": 9.772990077923735e-06, + "loss": 0.5892, + "step": 5606 + }, + { + "epoch": 0.1234206471494607, + "grad_norm": 1.405045688923036, + "learning_rate": 9.772883878504949e-06, + "loss": 0.3812, + "step": 5607 + }, + { + "epoch": 0.12344265903587938, + "grad_norm": 0.6302531844598328, + "learning_rate": 9.77277765482824e-06, + "loss": 0.5241, + "step": 5608 + }, + { + "epoch": 0.12346467092229804, + "grad_norm": 1.605753089936535, + "learning_rate": 9.772671406894146e-06, + "loss": 0.316, + "step": 5609 + }, + { + "epoch": 0.1234866828087167, + "grad_norm": 0.6423328129404655, + "learning_rate": 9.772565134703205e-06, + "loss": 0.5741, + "step": 5610 + }, + { + "epoch": 0.12350869469513537, + "grad_norm": 1.5512950570194715, + "learning_rate": 9.772458838255963e-06, + "loss": 0.3858, + "step": 5611 + }, + { + "epoch": 0.12353070658155405, + "grad_norm": 2.3032890931738264, + "learning_rate": 9.772352517552954e-06, + "loss": 0.4095, + "step": 5612 + }, + { + "epoch": 0.12355271846797271, + "grad_norm": 1.501390310554208, + "learning_rate": 9.772246172594722e-06, + "loss": 0.3517, + "step": 5613 + }, + { + "epoch": 0.12357473035439137, + "grad_norm": 2.488073229184473, + "learning_rate": 9.772139803381807e-06, + "loss": 0.4375, + "step": 5614 + }, + { + "epoch": 0.12359674224081003, + "grad_norm": 1.502771854498393, + "learning_rate": 9.772033409914747e-06, + "loss": 0.3832, + "step": 5615 + }, + { + "epoch": 0.1236187541272287, + "grad_norm": 1.5497400788949502, + "learning_rate": 9.771926992194088e-06, + "loss": 0.3122, + "step": 5616 + }, + { + "epoch": 0.12364076601364737, + "grad_norm": 1.683424557202718, + "learning_rate": 9.771820550220366e-06, + "loss": 0.3825, + "step": 5617 + }, + { + "epoch": 0.12366277790006604, + "grad_norm": 2.5438667041803487, + "learning_rate": 9.771714083994123e-06, + "loss": 0.3265, + "step": 5618 + }, + { + "epoch": 0.1236847897864847, + "grad_norm": 1.5231412148567203, + "learning_rate": 9.771607593515902e-06, + "loss": 0.3636, + "step": 5619 + }, + { + "epoch": 0.12370680167290336, + "grad_norm": 1.4972456215593812, + "learning_rate": 9.771501078786242e-06, + "loss": 0.32, + "step": 5620 + }, + { + "epoch": 0.12372881355932204, + "grad_norm": 1.610418342022616, + "learning_rate": 9.771394539805686e-06, + "loss": 0.3618, + "step": 5621 + }, + { + "epoch": 0.1237508254457407, + "grad_norm": 17.557983698651903, + "learning_rate": 9.771287976574775e-06, + "loss": 0.4429, + "step": 5622 + }, + { + "epoch": 0.12377283733215937, + "grad_norm": 1.578531660420293, + "learning_rate": 9.77118138909405e-06, + "loss": 0.3957, + "step": 5623 + }, + { + "epoch": 0.12379484921857803, + "grad_norm": 1.5209785057009662, + "learning_rate": 9.771074777364053e-06, + "loss": 0.3925, + "step": 5624 + }, + { + "epoch": 0.12381686110499669, + "grad_norm": 1.493084475584351, + "learning_rate": 9.770968141385326e-06, + "loss": 0.2766, + "step": 5625 + }, + { + "epoch": 0.12383887299141537, + "grad_norm": 0.7459579136969784, + "learning_rate": 9.77086148115841e-06, + "loss": 0.5653, + "step": 5626 + }, + { + "epoch": 0.12386088487783403, + "grad_norm": 1.6730120254297505, + "learning_rate": 9.77075479668385e-06, + "loss": 0.3872, + "step": 5627 + }, + { + "epoch": 0.1238828967642527, + "grad_norm": 1.6202501585715339, + "learning_rate": 9.770648087962187e-06, + "loss": 0.4445, + "step": 5628 + }, + { + "epoch": 0.12390490865067136, + "grad_norm": 1.507421331500341, + "learning_rate": 9.77054135499396e-06, + "loss": 0.3585, + "step": 5629 + }, + { + "epoch": 0.12392692053709004, + "grad_norm": 1.4591129980205717, + "learning_rate": 9.770434597779716e-06, + "loss": 0.4197, + "step": 5630 + }, + { + "epoch": 0.1239489324235087, + "grad_norm": 1.9945441427367276, + "learning_rate": 9.770327816319996e-06, + "loss": 0.4654, + "step": 5631 + }, + { + "epoch": 0.12397094430992736, + "grad_norm": 1.4737676152735022, + "learning_rate": 9.770221010615341e-06, + "loss": 0.3941, + "step": 5632 + }, + { + "epoch": 0.12399295619634602, + "grad_norm": 0.659999475841352, + "learning_rate": 9.770114180666296e-06, + "loss": 0.5844, + "step": 5633 + }, + { + "epoch": 0.12401496808276469, + "grad_norm": 2.066574021353124, + "learning_rate": 9.770007326473403e-06, + "loss": 0.3946, + "step": 5634 + }, + { + "epoch": 0.12403697996918336, + "grad_norm": 1.349141877674219, + "learning_rate": 9.769900448037204e-06, + "loss": 0.3008, + "step": 5635 + }, + { + "epoch": 0.12405899185560203, + "grad_norm": 0.6156747631025755, + "learning_rate": 9.769793545358246e-06, + "loss": 0.5487, + "step": 5636 + }, + { + "epoch": 0.12408100374202069, + "grad_norm": 2.965136017163705, + "learning_rate": 9.769686618437068e-06, + "loss": 0.3139, + "step": 5637 + }, + { + "epoch": 0.12410301562843935, + "grad_norm": 1.496762220072545, + "learning_rate": 9.769579667274216e-06, + "loss": 0.3575, + "step": 5638 + }, + { + "epoch": 0.12412502751485803, + "grad_norm": 0.6425430769057371, + "learning_rate": 9.769472691870234e-06, + "loss": 0.5575, + "step": 5639 + }, + { + "epoch": 0.1241470394012767, + "grad_norm": 1.5413422663983456, + "learning_rate": 9.769365692225663e-06, + "loss": 0.3295, + "step": 5640 + }, + { + "epoch": 0.12416905128769536, + "grad_norm": 1.8066912699277704, + "learning_rate": 9.769258668341049e-06, + "loss": 0.2909, + "step": 5641 + }, + { + "epoch": 0.12419106317411402, + "grad_norm": 2.105213009785167, + "learning_rate": 9.769151620216935e-06, + "loss": 0.3632, + "step": 5642 + }, + { + "epoch": 0.12421307506053268, + "grad_norm": 0.6680665883997118, + "learning_rate": 9.769044547853864e-06, + "loss": 0.5584, + "step": 5643 + }, + { + "epoch": 0.12423508694695136, + "grad_norm": 1.6966437961373566, + "learning_rate": 9.768937451252383e-06, + "loss": 0.4148, + "step": 5644 + }, + { + "epoch": 0.12425709883337002, + "grad_norm": 1.485685227733843, + "learning_rate": 9.768830330413035e-06, + "loss": 0.4643, + "step": 5645 + }, + { + "epoch": 0.12427911071978869, + "grad_norm": 1.8788391826528619, + "learning_rate": 9.768723185336363e-06, + "loss": 0.4971, + "step": 5646 + }, + { + "epoch": 0.12430112260620735, + "grad_norm": 1.9838615476734573, + "learning_rate": 9.768616016022914e-06, + "loss": 0.3379, + "step": 5647 + }, + { + "epoch": 0.12432313449262601, + "grad_norm": 1.772882477154418, + "learning_rate": 9.768508822473232e-06, + "loss": 0.3683, + "step": 5648 + }, + { + "epoch": 0.12434514637904469, + "grad_norm": 1.5422277882447426, + "learning_rate": 9.76840160468786e-06, + "loss": 0.3926, + "step": 5649 + }, + { + "epoch": 0.12436715826546335, + "grad_norm": 2.122271933820104, + "learning_rate": 9.768294362667346e-06, + "loss": 0.4231, + "step": 5650 + }, + { + "epoch": 0.12438917015188201, + "grad_norm": 1.456482391774272, + "learning_rate": 9.768187096412231e-06, + "loss": 0.3697, + "step": 5651 + }, + { + "epoch": 0.12441118203830068, + "grad_norm": 1.470689577012232, + "learning_rate": 9.768079805923065e-06, + "loss": 0.3256, + "step": 5652 + }, + { + "epoch": 0.12443319392471935, + "grad_norm": 1.8028450682572392, + "learning_rate": 9.76797249120039e-06, + "loss": 0.4458, + "step": 5653 + }, + { + "epoch": 0.12445520581113802, + "grad_norm": 0.6795975314554081, + "learning_rate": 9.767865152244751e-06, + "loss": 0.5485, + "step": 5654 + }, + { + "epoch": 0.12447721769755668, + "grad_norm": 1.446514086589443, + "learning_rate": 9.767757789056695e-06, + "loss": 0.2798, + "step": 5655 + }, + { + "epoch": 0.12449922958397534, + "grad_norm": 1.5566887244908325, + "learning_rate": 9.767650401636769e-06, + "loss": 0.355, + "step": 5656 + }, + { + "epoch": 0.124521241470394, + "grad_norm": 1.4953055772929662, + "learning_rate": 9.767542989985516e-06, + "loss": 0.3391, + "step": 5657 + }, + { + "epoch": 0.12454325335681268, + "grad_norm": 0.6607547442273104, + "learning_rate": 9.767435554103484e-06, + "loss": 0.6015, + "step": 5658 + }, + { + "epoch": 0.12456526524323135, + "grad_norm": 1.5907659750959093, + "learning_rate": 9.767328093991218e-06, + "loss": 0.3451, + "step": 5659 + }, + { + "epoch": 0.12458727712965001, + "grad_norm": 1.299150234422941, + "learning_rate": 9.767220609649263e-06, + "loss": 0.2882, + "step": 5660 + }, + { + "epoch": 0.12460928901606867, + "grad_norm": 1.3697094641152956, + "learning_rate": 9.767113101078168e-06, + "loss": 0.3694, + "step": 5661 + }, + { + "epoch": 0.12463130090248735, + "grad_norm": 1.6608012248219772, + "learning_rate": 9.767005568278479e-06, + "loss": 0.3105, + "step": 5662 + }, + { + "epoch": 0.12465331278890601, + "grad_norm": 1.4349072395448645, + "learning_rate": 9.766898011250738e-06, + "loss": 0.3656, + "step": 5663 + }, + { + "epoch": 0.12467532467532468, + "grad_norm": 1.544012408021618, + "learning_rate": 9.766790429995498e-06, + "loss": 0.4258, + "step": 5664 + }, + { + "epoch": 0.12469733656174334, + "grad_norm": 0.6575401916107193, + "learning_rate": 9.766682824513302e-06, + "loss": 0.5696, + "step": 5665 + }, + { + "epoch": 0.124719348448162, + "grad_norm": 1.4580751915155854, + "learning_rate": 9.7665751948047e-06, + "loss": 0.3072, + "step": 5666 + }, + { + "epoch": 0.12474136033458068, + "grad_norm": 0.6477822827320284, + "learning_rate": 9.766467540870234e-06, + "loss": 0.5594, + "step": 5667 + }, + { + "epoch": 0.12476337222099934, + "grad_norm": 0.6434245042161608, + "learning_rate": 9.766359862710454e-06, + "loss": 0.5869, + "step": 5668 + }, + { + "epoch": 0.124785384107418, + "grad_norm": 1.7386669042804834, + "learning_rate": 9.76625216032591e-06, + "loss": 0.3392, + "step": 5669 + }, + { + "epoch": 0.12480739599383667, + "grad_norm": 1.294625812458043, + "learning_rate": 9.766144433717146e-06, + "loss": 0.3605, + "step": 5670 + }, + { + "epoch": 0.12482940788025534, + "grad_norm": 1.4579324122492934, + "learning_rate": 9.766036682884709e-06, + "loss": 0.3363, + "step": 5671 + }, + { + "epoch": 0.12485141976667401, + "grad_norm": 1.776143239693949, + "learning_rate": 9.765928907829148e-06, + "loss": 0.5054, + "step": 5672 + }, + { + "epoch": 0.12487343165309267, + "grad_norm": 1.517503829427776, + "learning_rate": 9.76582110855101e-06, + "loss": 0.4411, + "step": 5673 + }, + { + "epoch": 0.12489544353951133, + "grad_norm": 0.7010234558899837, + "learning_rate": 9.765713285050847e-06, + "loss": 0.5699, + "step": 5674 + }, + { + "epoch": 0.12491745542593, + "grad_norm": 1.5737667711500194, + "learning_rate": 9.7656054373292e-06, + "loss": 0.3216, + "step": 5675 + }, + { + "epoch": 0.12493946731234867, + "grad_norm": 1.2981445164257253, + "learning_rate": 9.765497565386622e-06, + "loss": 0.359, + "step": 5676 + }, + { + "epoch": 0.12496147919876734, + "grad_norm": 1.5713516246120347, + "learning_rate": 9.76538966922366e-06, + "loss": 0.4084, + "step": 5677 + }, + { + "epoch": 0.124983491085186, + "grad_norm": 1.3506745653254022, + "learning_rate": 9.76528174884086e-06, + "loss": 0.4056, + "step": 5678 + }, + { + "epoch": 0.12500550297160468, + "grad_norm": 3.3074003598807127, + "learning_rate": 9.765173804238776e-06, + "loss": 0.3173, + "step": 5679 + }, + { + "epoch": 0.12502751485802333, + "grad_norm": 6.496638621487652, + "learning_rate": 9.76506583541795e-06, + "loss": 0.3358, + "step": 5680 + }, + { + "epoch": 0.125049526744442, + "grad_norm": 1.4300482306657414, + "learning_rate": 9.764957842378938e-06, + "loss": 0.3703, + "step": 5681 + }, + { + "epoch": 0.12507153863086065, + "grad_norm": 1.847299112459602, + "learning_rate": 9.764849825122283e-06, + "loss": 0.3883, + "step": 5682 + }, + { + "epoch": 0.12509355051727933, + "grad_norm": 0.663244078045561, + "learning_rate": 9.764741783648535e-06, + "loss": 0.565, + "step": 5683 + }, + { + "epoch": 0.125115562403698, + "grad_norm": 1.5574928640306474, + "learning_rate": 9.764633717958247e-06, + "loss": 0.3586, + "step": 5684 + }, + { + "epoch": 0.12513757429011665, + "grad_norm": 1.5716369656041644, + "learning_rate": 9.764525628051963e-06, + "loss": 0.3716, + "step": 5685 + }, + { + "epoch": 0.12515958617653533, + "grad_norm": 0.6530480652635591, + "learning_rate": 9.764417513930235e-06, + "loss": 0.5666, + "step": 5686 + }, + { + "epoch": 0.125181598062954, + "grad_norm": 2.803662415235532, + "learning_rate": 9.764309375593613e-06, + "loss": 0.3837, + "step": 5687 + }, + { + "epoch": 0.12520360994937266, + "grad_norm": 0.6411635407771306, + "learning_rate": 9.764201213042645e-06, + "loss": 0.585, + "step": 5688 + }, + { + "epoch": 0.12522562183579133, + "grad_norm": 1.631198649017802, + "learning_rate": 9.764093026277882e-06, + "loss": 0.3841, + "step": 5689 + }, + { + "epoch": 0.12524763372220998, + "grad_norm": 1.537530731307137, + "learning_rate": 9.763984815299874e-06, + "loss": 0.3224, + "step": 5690 + }, + { + "epoch": 0.12526964560862866, + "grad_norm": 1.8673682806279532, + "learning_rate": 9.76387658010917e-06, + "loss": 0.3887, + "step": 5691 + }, + { + "epoch": 0.12529165749504734, + "grad_norm": 1.8450935670210433, + "learning_rate": 9.763768320706319e-06, + "loss": 0.3203, + "step": 5692 + }, + { + "epoch": 0.125313669381466, + "grad_norm": 1.5749597861491211, + "learning_rate": 9.763660037091874e-06, + "loss": 0.4073, + "step": 5693 + }, + { + "epoch": 0.12533568126788466, + "grad_norm": 1.4045391723445175, + "learning_rate": 9.763551729266386e-06, + "loss": 0.3705, + "step": 5694 + }, + { + "epoch": 0.1253576931543033, + "grad_norm": 1.3673490933197332, + "learning_rate": 9.7634433972304e-06, + "loss": 0.4081, + "step": 5695 + }, + { + "epoch": 0.125379705040722, + "grad_norm": 1.4571701553599596, + "learning_rate": 9.763335040984471e-06, + "loss": 0.3228, + "step": 5696 + }, + { + "epoch": 0.12540171692714067, + "grad_norm": 0.6920540668156344, + "learning_rate": 9.76322666052915e-06, + "loss": 0.5247, + "step": 5697 + }, + { + "epoch": 0.12542372881355932, + "grad_norm": 0.6613130362978542, + "learning_rate": 9.763118255864985e-06, + "loss": 0.5528, + "step": 5698 + }, + { + "epoch": 0.125445740699978, + "grad_norm": 2.7994994406667453, + "learning_rate": 9.763009826992531e-06, + "loss": 0.3615, + "step": 5699 + }, + { + "epoch": 0.12546775258639664, + "grad_norm": 1.781078539292267, + "learning_rate": 9.762901373912334e-06, + "loss": 0.4043, + "step": 5700 + }, + { + "epoch": 0.12548976447281532, + "grad_norm": 1.7988470528017375, + "learning_rate": 9.76279289662495e-06, + "loss": 0.3075, + "step": 5701 + }, + { + "epoch": 0.125511776359234, + "grad_norm": 0.7046976646720804, + "learning_rate": 9.762684395130928e-06, + "loss": 0.5499, + "step": 5702 + }, + { + "epoch": 0.12553378824565264, + "grad_norm": 2.8002501700858144, + "learning_rate": 9.76257586943082e-06, + "loss": 0.3925, + "step": 5703 + }, + { + "epoch": 0.12555580013207132, + "grad_norm": 2.194440512665179, + "learning_rate": 9.762467319525175e-06, + "loss": 0.3782, + "step": 5704 + }, + { + "epoch": 0.12557781201849, + "grad_norm": 1.6632299801608112, + "learning_rate": 9.762358745414549e-06, + "loss": 0.4263, + "step": 5705 + }, + { + "epoch": 0.12559982390490865, + "grad_norm": 2.163312899247279, + "learning_rate": 9.76225014709949e-06, + "loss": 0.3849, + "step": 5706 + }, + { + "epoch": 0.12562183579132732, + "grad_norm": 1.8408913126416901, + "learning_rate": 9.762141524580553e-06, + "loss": 0.3712, + "step": 5707 + }, + { + "epoch": 0.12564384767774597, + "grad_norm": 1.8723752312233264, + "learning_rate": 9.762032877858289e-06, + "loss": 0.3392, + "step": 5708 + }, + { + "epoch": 0.12566585956416465, + "grad_norm": 2.087067315194015, + "learning_rate": 9.76192420693325e-06, + "loss": 0.4091, + "step": 5709 + }, + { + "epoch": 0.12568787145058333, + "grad_norm": 1.560205837530191, + "learning_rate": 9.761815511805987e-06, + "loss": 0.379, + "step": 5710 + }, + { + "epoch": 0.12570988333700198, + "grad_norm": 3.5209226390337207, + "learning_rate": 9.761706792477054e-06, + "loss": 0.3416, + "step": 5711 + }, + { + "epoch": 0.12573189522342065, + "grad_norm": 2.01064515168582, + "learning_rate": 9.761598048947004e-06, + "loss": 0.3361, + "step": 5712 + }, + { + "epoch": 0.1257539071098393, + "grad_norm": 0.6709176176978688, + "learning_rate": 9.76148928121639e-06, + "loss": 0.5474, + "step": 5713 + }, + { + "epoch": 0.12577591899625798, + "grad_norm": 2.2899643803101393, + "learning_rate": 9.761380489285763e-06, + "loss": 0.4372, + "step": 5714 + }, + { + "epoch": 0.12579793088267666, + "grad_norm": 2.2861514131608427, + "learning_rate": 9.761271673155677e-06, + "loss": 0.3906, + "step": 5715 + }, + { + "epoch": 0.1258199427690953, + "grad_norm": 0.6249015023563278, + "learning_rate": 9.761162832826683e-06, + "loss": 0.5412, + "step": 5716 + }, + { + "epoch": 0.12584195465551398, + "grad_norm": 1.8742614003311795, + "learning_rate": 9.76105396829934e-06, + "loss": 0.409, + "step": 5717 + }, + { + "epoch": 0.12586396654193263, + "grad_norm": 0.6521919211852374, + "learning_rate": 9.760945079574192e-06, + "loss": 0.5626, + "step": 5718 + }, + { + "epoch": 0.1258859784283513, + "grad_norm": 1.3324766540660413, + "learning_rate": 9.760836166651801e-06, + "loss": 0.2755, + "step": 5719 + }, + { + "epoch": 0.12590799031476999, + "grad_norm": 3.189652695105734, + "learning_rate": 9.760727229532717e-06, + "loss": 0.3775, + "step": 5720 + }, + { + "epoch": 0.12593000220118863, + "grad_norm": 1.4756989100325628, + "learning_rate": 9.760618268217493e-06, + "loss": 0.2965, + "step": 5721 + }, + { + "epoch": 0.1259520140876073, + "grad_norm": 0.6388472526993532, + "learning_rate": 9.760509282706683e-06, + "loss": 0.5623, + "step": 5722 + }, + { + "epoch": 0.12597402597402596, + "grad_norm": 2.2920176304238877, + "learning_rate": 9.760400273000842e-06, + "loss": 0.3691, + "step": 5723 + }, + { + "epoch": 0.12599603786044464, + "grad_norm": 2.248687869763867, + "learning_rate": 9.760291239100524e-06, + "loss": 0.3827, + "step": 5724 + }, + { + "epoch": 0.12601804974686331, + "grad_norm": 1.8922297670438721, + "learning_rate": 9.760182181006283e-06, + "loss": 0.3969, + "step": 5725 + }, + { + "epoch": 0.12604006163328196, + "grad_norm": 1.402957261388701, + "learning_rate": 9.760073098718674e-06, + "loss": 0.3685, + "step": 5726 + }, + { + "epoch": 0.12606207351970064, + "grad_norm": 1.2911887678999023, + "learning_rate": 9.759963992238249e-06, + "loss": 0.3915, + "step": 5727 + }, + { + "epoch": 0.12608408540611932, + "grad_norm": 1.3338063547123393, + "learning_rate": 9.759854861565565e-06, + "loss": 0.3511, + "step": 5728 + }, + { + "epoch": 0.12610609729253797, + "grad_norm": 2.0627841859697993, + "learning_rate": 9.759745706701174e-06, + "loss": 0.2984, + "step": 5729 + }, + { + "epoch": 0.12612810917895664, + "grad_norm": 1.3806906747470742, + "learning_rate": 9.759636527645633e-06, + "loss": 0.3553, + "step": 5730 + }, + { + "epoch": 0.1261501210653753, + "grad_norm": 1.6490850445338792, + "learning_rate": 9.759527324399496e-06, + "loss": 0.3706, + "step": 5731 + }, + { + "epoch": 0.12617213295179397, + "grad_norm": 2.294136623295731, + "learning_rate": 9.75941809696332e-06, + "loss": 0.3796, + "step": 5732 + }, + { + "epoch": 0.12619414483821265, + "grad_norm": 1.485073888410764, + "learning_rate": 9.759308845337658e-06, + "loss": 0.2558, + "step": 5733 + }, + { + "epoch": 0.1262161567246313, + "grad_norm": 1.7514941760062719, + "learning_rate": 9.759199569523065e-06, + "loss": 0.3463, + "step": 5734 + }, + { + "epoch": 0.12623816861104997, + "grad_norm": 1.7333507949493179, + "learning_rate": 9.759090269520097e-06, + "loss": 0.4043, + "step": 5735 + }, + { + "epoch": 0.12626018049746862, + "grad_norm": 1.635492778433208, + "learning_rate": 9.758980945329311e-06, + "loss": 0.4015, + "step": 5736 + }, + { + "epoch": 0.1262821923838873, + "grad_norm": 1.3159178800394178, + "learning_rate": 9.75887159695126e-06, + "loss": 0.2412, + "step": 5737 + }, + { + "epoch": 0.12630420427030598, + "grad_norm": 1.615472363285036, + "learning_rate": 9.758762224386502e-06, + "loss": 0.4719, + "step": 5738 + }, + { + "epoch": 0.12632621615672462, + "grad_norm": 1.4538528102023895, + "learning_rate": 9.75865282763559e-06, + "loss": 0.3508, + "step": 5739 + }, + { + "epoch": 0.1263482280431433, + "grad_norm": 1.1166515201384715, + "learning_rate": 9.758543406699084e-06, + "loss": 0.2672, + "step": 5740 + }, + { + "epoch": 0.12637023992956195, + "grad_norm": 2.5990892322169805, + "learning_rate": 9.758433961577536e-06, + "loss": 0.2858, + "step": 5741 + }, + { + "epoch": 0.12639225181598063, + "grad_norm": 1.5026422431701976, + "learning_rate": 9.758324492271505e-06, + "loss": 0.3658, + "step": 5742 + }, + { + "epoch": 0.1264142637023993, + "grad_norm": 1.5093567056195611, + "learning_rate": 9.758214998781548e-06, + "loss": 0.3841, + "step": 5743 + }, + { + "epoch": 0.12643627558881795, + "grad_norm": 2.1854335878989963, + "learning_rate": 9.75810548110822e-06, + "loss": 0.3513, + "step": 5744 + }, + { + "epoch": 0.12645828747523663, + "grad_norm": 1.5605295667124697, + "learning_rate": 9.757995939252076e-06, + "loss": 0.3276, + "step": 5745 + }, + { + "epoch": 0.1264802993616553, + "grad_norm": 1.3156089569857135, + "learning_rate": 9.757886373213676e-06, + "loss": 0.3624, + "step": 5746 + }, + { + "epoch": 0.12650231124807396, + "grad_norm": 1.5273931963587386, + "learning_rate": 9.757776782993575e-06, + "loss": 0.3787, + "step": 5747 + }, + { + "epoch": 0.12652432313449263, + "grad_norm": 1.4905731351902112, + "learning_rate": 9.75766716859233e-06, + "loss": 0.4573, + "step": 5748 + }, + { + "epoch": 0.12654633502091128, + "grad_norm": 0.7038278359212865, + "learning_rate": 9.7575575300105e-06, + "loss": 0.5748, + "step": 5749 + }, + { + "epoch": 0.12656834690732996, + "grad_norm": 0.6583289690854822, + "learning_rate": 9.757447867248638e-06, + "loss": 0.5646, + "step": 5750 + }, + { + "epoch": 0.12659035879374864, + "grad_norm": 1.422855869199825, + "learning_rate": 9.757338180307305e-06, + "loss": 0.3488, + "step": 5751 + }, + { + "epoch": 0.12661237068016729, + "grad_norm": 1.4455280622783895, + "learning_rate": 9.757228469187059e-06, + "loss": 0.3168, + "step": 5752 + }, + { + "epoch": 0.12663438256658596, + "grad_norm": 1.482656484535185, + "learning_rate": 9.757118733888456e-06, + "loss": 0.3513, + "step": 5753 + }, + { + "epoch": 0.1266563944530046, + "grad_norm": 2.879289640226069, + "learning_rate": 9.757008974412052e-06, + "loss": 0.4312, + "step": 5754 + }, + { + "epoch": 0.1266784063394233, + "grad_norm": 1.55049053073492, + "learning_rate": 9.756899190758408e-06, + "loss": 0.4046, + "step": 5755 + }, + { + "epoch": 0.12670041822584197, + "grad_norm": 1.3843727240873762, + "learning_rate": 9.75678938292808e-06, + "loss": 0.3806, + "step": 5756 + }, + { + "epoch": 0.12672243011226061, + "grad_norm": 1.6746832080092495, + "learning_rate": 9.756679550921627e-06, + "loss": 0.4005, + "step": 5757 + }, + { + "epoch": 0.1267444419986793, + "grad_norm": 1.2403676527047953, + "learning_rate": 9.756569694739607e-06, + "loss": 0.3941, + "step": 5758 + }, + { + "epoch": 0.12676645388509794, + "grad_norm": 1.744507279677766, + "learning_rate": 9.75645981438258e-06, + "loss": 0.3568, + "step": 5759 + }, + { + "epoch": 0.12678846577151662, + "grad_norm": 1.3335649978102573, + "learning_rate": 9.7563499098511e-06, + "loss": 0.3587, + "step": 5760 + }, + { + "epoch": 0.1268104776579353, + "grad_norm": 1.5906486329670708, + "learning_rate": 9.75623998114573e-06, + "loss": 0.3128, + "step": 5761 + }, + { + "epoch": 0.12683248954435394, + "grad_norm": 1.3005970892380585, + "learning_rate": 9.756130028267026e-06, + "loss": 0.2599, + "step": 5762 + }, + { + "epoch": 0.12685450143077262, + "grad_norm": 1.516364239640252, + "learning_rate": 9.756020051215548e-06, + "loss": 0.3869, + "step": 5763 + }, + { + "epoch": 0.12687651331719127, + "grad_norm": 7.01602870617298, + "learning_rate": 9.755910049991857e-06, + "loss": 0.4011, + "step": 5764 + }, + { + "epoch": 0.12689852520360995, + "grad_norm": 0.7111634656828483, + "learning_rate": 9.755800024596508e-06, + "loss": 0.5666, + "step": 5765 + }, + { + "epoch": 0.12692053709002862, + "grad_norm": 0.6856314904911625, + "learning_rate": 9.755689975030063e-06, + "loss": 0.5509, + "step": 5766 + }, + { + "epoch": 0.12694254897644727, + "grad_norm": 1.7455994913038706, + "learning_rate": 9.75557990129308e-06, + "loss": 0.3118, + "step": 5767 + }, + { + "epoch": 0.12696456086286595, + "grad_norm": 1.3841436518710224, + "learning_rate": 9.755469803386119e-06, + "loss": 0.3779, + "step": 5768 + }, + { + "epoch": 0.12698657274928463, + "grad_norm": 1.5427361584631363, + "learning_rate": 9.75535968130974e-06, + "loss": 0.3108, + "step": 5769 + }, + { + "epoch": 0.12700858463570328, + "grad_norm": 1.2603296292997428, + "learning_rate": 9.7552495350645e-06, + "loss": 0.3306, + "step": 5770 + }, + { + "epoch": 0.12703059652212195, + "grad_norm": 1.6486320510625496, + "learning_rate": 9.755139364650963e-06, + "loss": 0.3882, + "step": 5771 + }, + { + "epoch": 0.1270526084085406, + "grad_norm": 1.4095646009816865, + "learning_rate": 9.755029170069686e-06, + "loss": 0.4214, + "step": 5772 + }, + { + "epoch": 0.12707462029495928, + "grad_norm": 1.4210842125496785, + "learning_rate": 9.75491895132123e-06, + "loss": 0.4097, + "step": 5773 + }, + { + "epoch": 0.12709663218137796, + "grad_norm": 1.9821588886395507, + "learning_rate": 9.754808708406157e-06, + "loss": 0.4049, + "step": 5774 + }, + { + "epoch": 0.1271186440677966, + "grad_norm": 0.896908201065695, + "learning_rate": 9.754698441325024e-06, + "loss": 0.5711, + "step": 5775 + }, + { + "epoch": 0.12714065595421528, + "grad_norm": 1.5053677347344943, + "learning_rate": 9.75458815007839e-06, + "loss": 0.3832, + "step": 5776 + }, + { + "epoch": 0.12716266784063393, + "grad_norm": 0.7391960848703916, + "learning_rate": 9.754477834666822e-06, + "loss": 0.5479, + "step": 5777 + }, + { + "epoch": 0.1271846797270526, + "grad_norm": 1.9895610924279539, + "learning_rate": 9.754367495090877e-06, + "loss": 0.3483, + "step": 5778 + }, + { + "epoch": 0.12720669161347128, + "grad_norm": 1.3822423183350927, + "learning_rate": 9.754257131351115e-06, + "loss": 0.329, + "step": 5779 + }, + { + "epoch": 0.12722870349988993, + "grad_norm": 1.2201430711748595, + "learning_rate": 9.754146743448098e-06, + "loss": 0.272, + "step": 5780 + }, + { + "epoch": 0.1272507153863086, + "grad_norm": 1.4522288452707386, + "learning_rate": 9.754036331382386e-06, + "loss": 0.3899, + "step": 5781 + }, + { + "epoch": 0.12727272727272726, + "grad_norm": 1.719892768548406, + "learning_rate": 9.753925895154542e-06, + "loss": 0.3592, + "step": 5782 + }, + { + "epoch": 0.12729473915914594, + "grad_norm": 1.509572030187656, + "learning_rate": 9.753815434765126e-06, + "loss": 0.3965, + "step": 5783 + }, + { + "epoch": 0.1273167510455646, + "grad_norm": 0.9901435801154406, + "learning_rate": 9.7537049502147e-06, + "loss": 0.5813, + "step": 5784 + }, + { + "epoch": 0.12733876293198326, + "grad_norm": 1.5997526438263994, + "learning_rate": 9.753594441503826e-06, + "loss": 0.3713, + "step": 5785 + }, + { + "epoch": 0.12736077481840194, + "grad_norm": 1.4397294162736602, + "learning_rate": 9.753483908633063e-06, + "loss": 0.3721, + "step": 5786 + }, + { + "epoch": 0.12738278670482062, + "grad_norm": 1.9099582226525476, + "learning_rate": 9.753373351602977e-06, + "loss": 0.4507, + "step": 5787 + }, + { + "epoch": 0.12740479859123927, + "grad_norm": 1.4942652005587862, + "learning_rate": 9.753262770414124e-06, + "loss": 0.3417, + "step": 5788 + }, + { + "epoch": 0.12742681047765794, + "grad_norm": 1.287589661009163, + "learning_rate": 9.753152165067072e-06, + "loss": 0.3057, + "step": 5789 + }, + { + "epoch": 0.1274488223640766, + "grad_norm": 1.9887847287047478, + "learning_rate": 9.753041535562382e-06, + "loss": 0.3617, + "step": 5790 + }, + { + "epoch": 0.12747083425049527, + "grad_norm": 0.7129538549497777, + "learning_rate": 9.752930881900613e-06, + "loss": 0.5595, + "step": 5791 + }, + { + "epoch": 0.12749284613691395, + "grad_norm": 0.7059610965414186, + "learning_rate": 9.75282020408233e-06, + "loss": 0.5341, + "step": 5792 + }, + { + "epoch": 0.1275148580233326, + "grad_norm": 1.4651547527486157, + "learning_rate": 9.752709502108094e-06, + "loss": 0.3324, + "step": 5793 + }, + { + "epoch": 0.12753686990975127, + "grad_norm": 0.649649002893017, + "learning_rate": 9.752598775978469e-06, + "loss": 0.5728, + "step": 5794 + }, + { + "epoch": 0.12755888179616992, + "grad_norm": 0.6379958997432923, + "learning_rate": 9.752488025694018e-06, + "loss": 0.528, + "step": 5795 + }, + { + "epoch": 0.1275808936825886, + "grad_norm": 1.3927608105779978, + "learning_rate": 9.752377251255302e-06, + "loss": 0.3518, + "step": 5796 + }, + { + "epoch": 0.12760290556900727, + "grad_norm": 1.6958505241725155, + "learning_rate": 9.752266452662886e-06, + "loss": 0.2815, + "step": 5797 + }, + { + "epoch": 0.12762491745542592, + "grad_norm": 1.3169987675869315, + "learning_rate": 9.752155629917332e-06, + "loss": 0.3309, + "step": 5798 + }, + { + "epoch": 0.1276469293418446, + "grad_norm": 1.5587069562840723, + "learning_rate": 9.752044783019206e-06, + "loss": 0.3609, + "step": 5799 + }, + { + "epoch": 0.12766894122826325, + "grad_norm": 1.7176512528488626, + "learning_rate": 9.751933911969065e-06, + "loss": 0.4081, + "step": 5800 + }, + { + "epoch": 0.12769095311468193, + "grad_norm": 0.8213508675185062, + "learning_rate": 9.751823016767478e-06, + "loss": 0.5478, + "step": 5801 + }, + { + "epoch": 0.1277129650011006, + "grad_norm": 2.849555887624825, + "learning_rate": 9.751712097415007e-06, + "loss": 0.3455, + "step": 5802 + }, + { + "epoch": 0.12773497688751925, + "grad_norm": 0.7530923857113742, + "learning_rate": 9.751601153912217e-06, + "loss": 0.5433, + "step": 5803 + }, + { + "epoch": 0.12775698877393793, + "grad_norm": 0.6943526375901704, + "learning_rate": 9.75149018625967e-06, + "loss": 0.5522, + "step": 5804 + }, + { + "epoch": 0.1277790006603566, + "grad_norm": 1.5644318775159105, + "learning_rate": 9.75137919445793e-06, + "loss": 0.3316, + "step": 5805 + }, + { + "epoch": 0.12780101254677526, + "grad_norm": 1.329649779534291, + "learning_rate": 9.751268178507562e-06, + "loss": 0.3254, + "step": 5806 + }, + { + "epoch": 0.12782302443319393, + "grad_norm": 1.28937647345513, + "learning_rate": 9.75115713840913e-06, + "loss": 0.3976, + "step": 5807 + }, + { + "epoch": 0.12784503631961258, + "grad_norm": 1.263948885803583, + "learning_rate": 9.751046074163199e-06, + "loss": 0.363, + "step": 5808 + }, + { + "epoch": 0.12786704820603126, + "grad_norm": 2.728801455951631, + "learning_rate": 9.750934985770334e-06, + "loss": 0.3808, + "step": 5809 + }, + { + "epoch": 0.12788906009244994, + "grad_norm": 1.6430193051526312, + "learning_rate": 9.750823873231095e-06, + "loss": 0.4468, + "step": 5810 + }, + { + "epoch": 0.12791107197886858, + "grad_norm": 1.612174519026883, + "learning_rate": 9.750712736546052e-06, + "loss": 0.3697, + "step": 5811 + }, + { + "epoch": 0.12793308386528726, + "grad_norm": 1.4287921861135402, + "learning_rate": 9.750601575715767e-06, + "loss": 0.3451, + "step": 5812 + }, + { + "epoch": 0.1279550957517059, + "grad_norm": 1.33438987228357, + "learning_rate": 9.750490390740808e-06, + "loss": 0.3228, + "step": 5813 + }, + { + "epoch": 0.1279771076381246, + "grad_norm": 2.062967169298506, + "learning_rate": 9.750379181621736e-06, + "loss": 0.3614, + "step": 5814 + }, + { + "epoch": 0.12799911952454326, + "grad_norm": 1.7132173600319984, + "learning_rate": 9.75026794835912e-06, + "loss": 0.3577, + "step": 5815 + }, + { + "epoch": 0.1280211314109619, + "grad_norm": 1.4798481689374852, + "learning_rate": 9.750156690953523e-06, + "loss": 0.3358, + "step": 5816 + }, + { + "epoch": 0.1280431432973806, + "grad_norm": 1.5925680493594563, + "learning_rate": 9.75004540940551e-06, + "loss": 0.3539, + "step": 5817 + }, + { + "epoch": 0.12806515518379924, + "grad_norm": 2.1046500305755083, + "learning_rate": 9.749934103715648e-06, + "loss": 0.422, + "step": 5818 + }, + { + "epoch": 0.12808716707021792, + "grad_norm": 1.640637258326101, + "learning_rate": 9.749822773884503e-06, + "loss": 0.3549, + "step": 5819 + }, + { + "epoch": 0.1281091789566366, + "grad_norm": 1.4162672297982755, + "learning_rate": 9.74971141991264e-06, + "loss": 0.2645, + "step": 5820 + }, + { + "epoch": 0.12813119084305524, + "grad_norm": 1.678107399736304, + "learning_rate": 9.749600041800626e-06, + "loss": 0.4724, + "step": 5821 + }, + { + "epoch": 0.12815320272947392, + "grad_norm": 1.4929878751002366, + "learning_rate": 9.749488639549026e-06, + "loss": 0.3898, + "step": 5822 + }, + { + "epoch": 0.12817521461589257, + "grad_norm": 1.6522012493337566, + "learning_rate": 9.749377213158405e-06, + "loss": 0.2657, + "step": 5823 + }, + { + "epoch": 0.12819722650231125, + "grad_norm": 1.5493254604748845, + "learning_rate": 9.749265762629332e-06, + "loss": 0.4021, + "step": 5824 + }, + { + "epoch": 0.12821923838872992, + "grad_norm": 1.9527968022500743, + "learning_rate": 9.749154287962371e-06, + "loss": 0.3897, + "step": 5825 + }, + { + "epoch": 0.12824125027514857, + "grad_norm": 2.2073366220609847, + "learning_rate": 9.74904278915809e-06, + "loss": 0.489, + "step": 5826 + }, + { + "epoch": 0.12826326216156725, + "grad_norm": 1.5036522780957087, + "learning_rate": 9.748931266217057e-06, + "loss": 0.5862, + "step": 5827 + }, + { + "epoch": 0.12828527404798593, + "grad_norm": 1.3621546580655532, + "learning_rate": 9.748819719139835e-06, + "loss": 0.4023, + "step": 5828 + }, + { + "epoch": 0.12830728593440457, + "grad_norm": 1.362633391254685, + "learning_rate": 9.748708147926995e-06, + "loss": 0.3657, + "step": 5829 + }, + { + "epoch": 0.12832929782082325, + "grad_norm": 1.5005438830565418, + "learning_rate": 9.748596552579102e-06, + "loss": 0.2667, + "step": 5830 + }, + { + "epoch": 0.1283513097072419, + "grad_norm": 0.719795490775031, + "learning_rate": 9.748484933096722e-06, + "loss": 0.586, + "step": 5831 + }, + { + "epoch": 0.12837332159366058, + "grad_norm": 0.6737660517719322, + "learning_rate": 9.748373289480424e-06, + "loss": 0.5773, + "step": 5832 + }, + { + "epoch": 0.12839533348007925, + "grad_norm": 1.5113297847741929, + "learning_rate": 9.748261621730775e-06, + "loss": 0.2444, + "step": 5833 + }, + { + "epoch": 0.1284173453664979, + "grad_norm": 1.4235702534759433, + "learning_rate": 9.748149929848344e-06, + "loss": 0.3517, + "step": 5834 + }, + { + "epoch": 0.12843935725291658, + "grad_norm": 1.6531192187069408, + "learning_rate": 9.748038213833697e-06, + "loss": 0.3722, + "step": 5835 + }, + { + "epoch": 0.12846136913933523, + "grad_norm": 0.9689643416174533, + "learning_rate": 9.747926473687402e-06, + "loss": 0.5839, + "step": 5836 + }, + { + "epoch": 0.1284833810257539, + "grad_norm": 0.8642182691260781, + "learning_rate": 9.747814709410026e-06, + "loss": 0.5366, + "step": 5837 + }, + { + "epoch": 0.12850539291217258, + "grad_norm": 1.3715910264461926, + "learning_rate": 9.747702921002138e-06, + "loss": 0.356, + "step": 5838 + }, + { + "epoch": 0.12852740479859123, + "grad_norm": 1.7131210637992096, + "learning_rate": 9.747591108464307e-06, + "loss": 0.4006, + "step": 5839 + }, + { + "epoch": 0.1285494166850099, + "grad_norm": 1.4258149503806246, + "learning_rate": 9.747479271797101e-06, + "loss": 0.3314, + "step": 5840 + }, + { + "epoch": 0.12857142857142856, + "grad_norm": 2.046114353490297, + "learning_rate": 9.747367411001088e-06, + "loss": 0.3853, + "step": 5841 + }, + { + "epoch": 0.12859344045784724, + "grad_norm": 0.855829578470081, + "learning_rate": 9.747255526076835e-06, + "loss": 0.5541, + "step": 5842 + }, + { + "epoch": 0.1286154523442659, + "grad_norm": 1.71131942085861, + "learning_rate": 9.747143617024914e-06, + "loss": 0.4223, + "step": 5843 + }, + { + "epoch": 0.12863746423068456, + "grad_norm": 1.451302091819602, + "learning_rate": 9.74703168384589e-06, + "loss": 0.3181, + "step": 5844 + }, + { + "epoch": 0.12865947611710324, + "grad_norm": 1.6529399877047761, + "learning_rate": 9.746919726540334e-06, + "loss": 0.3981, + "step": 5845 + }, + { + "epoch": 0.12868148800352192, + "grad_norm": 1.39663792422827, + "learning_rate": 9.746807745108816e-06, + "loss": 0.2938, + "step": 5846 + }, + { + "epoch": 0.12870349988994056, + "grad_norm": 1.6615540697076794, + "learning_rate": 9.746695739551904e-06, + "loss": 0.3396, + "step": 5847 + }, + { + "epoch": 0.12872551177635924, + "grad_norm": 2.3820214374077784, + "learning_rate": 9.746583709870165e-06, + "loss": 0.4136, + "step": 5848 + }, + { + "epoch": 0.1287475236627779, + "grad_norm": 0.7645858634472451, + "learning_rate": 9.746471656064172e-06, + "loss": 0.5792, + "step": 5849 + }, + { + "epoch": 0.12876953554919657, + "grad_norm": 1.5791533474055741, + "learning_rate": 9.746359578134493e-06, + "loss": 0.3773, + "step": 5850 + }, + { + "epoch": 0.12879154743561524, + "grad_norm": 0.7225834242959545, + "learning_rate": 9.746247476081699e-06, + "loss": 0.5573, + "step": 5851 + }, + { + "epoch": 0.1288135593220339, + "grad_norm": 1.3250952843278538, + "learning_rate": 9.746135349906357e-06, + "loss": 0.3401, + "step": 5852 + }, + { + "epoch": 0.12883557120845257, + "grad_norm": 1.7057734719169653, + "learning_rate": 9.746023199609037e-06, + "loss": 0.368, + "step": 5853 + }, + { + "epoch": 0.12885758309487122, + "grad_norm": 1.3598544415630107, + "learning_rate": 9.74591102519031e-06, + "loss": 0.3765, + "step": 5854 + }, + { + "epoch": 0.1288795949812899, + "grad_norm": 1.2795003911567264, + "learning_rate": 9.74579882665075e-06, + "loss": 0.3761, + "step": 5855 + }, + { + "epoch": 0.12890160686770857, + "grad_norm": 0.6895191073857438, + "learning_rate": 9.745686603990921e-06, + "loss": 0.5611, + "step": 5856 + }, + { + "epoch": 0.12892361875412722, + "grad_norm": 1.4909878729912855, + "learning_rate": 9.745574357211398e-06, + "loss": 0.4201, + "step": 5857 + }, + { + "epoch": 0.1289456306405459, + "grad_norm": 1.490453278582154, + "learning_rate": 9.745462086312748e-06, + "loss": 0.33, + "step": 5858 + }, + { + "epoch": 0.12896764252696455, + "grad_norm": 0.719580610455634, + "learning_rate": 9.745349791295544e-06, + "loss": 0.5639, + "step": 5859 + }, + { + "epoch": 0.12898965441338323, + "grad_norm": 1.958856617311571, + "learning_rate": 9.745237472160355e-06, + "loss": 0.4046, + "step": 5860 + }, + { + "epoch": 0.1290116662998019, + "grad_norm": 0.6760045265686011, + "learning_rate": 9.745125128907753e-06, + "loss": 0.5544, + "step": 5861 + }, + { + "epoch": 0.12903367818622055, + "grad_norm": 1.4086402829378808, + "learning_rate": 9.745012761538309e-06, + "loss": 0.3462, + "step": 5862 + }, + { + "epoch": 0.12905569007263923, + "grad_norm": 1.5968498663479287, + "learning_rate": 9.744900370052593e-06, + "loss": 0.5067, + "step": 5863 + }, + { + "epoch": 0.12907770195905788, + "grad_norm": 1.1560288329804396, + "learning_rate": 9.744787954451178e-06, + "loss": 0.3181, + "step": 5864 + }, + { + "epoch": 0.12909971384547655, + "grad_norm": 1.4750399082197443, + "learning_rate": 9.744675514734634e-06, + "loss": 0.3333, + "step": 5865 + }, + { + "epoch": 0.12912172573189523, + "grad_norm": 1.5408241888079162, + "learning_rate": 9.744563050903533e-06, + "loss": 0.3562, + "step": 5866 + }, + { + "epoch": 0.12914373761831388, + "grad_norm": 0.7728671177090651, + "learning_rate": 9.744450562958447e-06, + "loss": 0.5493, + "step": 5867 + }, + { + "epoch": 0.12916574950473256, + "grad_norm": 1.2103217484483078, + "learning_rate": 9.744338050899946e-06, + "loss": 0.3647, + "step": 5868 + }, + { + "epoch": 0.12918776139115123, + "grad_norm": 1.3704532505071794, + "learning_rate": 9.744225514728604e-06, + "loss": 0.4105, + "step": 5869 + }, + { + "epoch": 0.12920977327756988, + "grad_norm": 1.6039610275655296, + "learning_rate": 9.74411295444499e-06, + "loss": 0.3617, + "step": 5870 + }, + { + "epoch": 0.12923178516398856, + "grad_norm": 0.6635776902087165, + "learning_rate": 9.74400037004968e-06, + "loss": 0.5273, + "step": 5871 + }, + { + "epoch": 0.1292537970504072, + "grad_norm": 1.2066924714999354, + "learning_rate": 9.743887761543245e-06, + "loss": 0.3818, + "step": 5872 + }, + { + "epoch": 0.1292758089368259, + "grad_norm": 1.7520730018345927, + "learning_rate": 9.743775128926254e-06, + "loss": 0.3285, + "step": 5873 + }, + { + "epoch": 0.12929782082324456, + "grad_norm": 1.7249921481778427, + "learning_rate": 9.743662472199285e-06, + "loss": 0.3601, + "step": 5874 + }, + { + "epoch": 0.1293198327096632, + "grad_norm": 2.4157655827213356, + "learning_rate": 9.743549791362904e-06, + "loss": 0.4444, + "step": 5875 + }, + { + "epoch": 0.1293418445960819, + "grad_norm": 1.8146425424476433, + "learning_rate": 9.74343708641769e-06, + "loss": 0.4165, + "step": 5876 + }, + { + "epoch": 0.12936385648250054, + "grad_norm": 1.5917545284755128, + "learning_rate": 9.743324357364213e-06, + "loss": 0.2685, + "step": 5877 + }, + { + "epoch": 0.12938586836891922, + "grad_norm": 1.5221635817605574, + "learning_rate": 9.743211604203045e-06, + "loss": 0.426, + "step": 5878 + }, + { + "epoch": 0.1294078802553379, + "grad_norm": 1.0880046387373516, + "learning_rate": 9.74309882693476e-06, + "loss": 0.2828, + "step": 5879 + }, + { + "epoch": 0.12942989214175654, + "grad_norm": 1.1768814439672002, + "learning_rate": 9.74298602555993e-06, + "loss": 0.3717, + "step": 5880 + }, + { + "epoch": 0.12945190402817522, + "grad_norm": 1.6317909691829406, + "learning_rate": 9.742873200079132e-06, + "loss": 0.3164, + "step": 5881 + }, + { + "epoch": 0.12947391591459387, + "grad_norm": 1.4279647254479058, + "learning_rate": 9.742760350492936e-06, + "loss": 0.3852, + "step": 5882 + }, + { + "epoch": 0.12949592780101254, + "grad_norm": 0.6713902381582532, + "learning_rate": 9.742647476801918e-06, + "loss": 0.5458, + "step": 5883 + }, + { + "epoch": 0.12951793968743122, + "grad_norm": 2.043479732018537, + "learning_rate": 9.742534579006649e-06, + "loss": 0.4648, + "step": 5884 + }, + { + "epoch": 0.12953995157384987, + "grad_norm": 1.644937313386467, + "learning_rate": 9.742421657107704e-06, + "loss": 0.2954, + "step": 5885 + }, + { + "epoch": 0.12956196346026855, + "grad_norm": 1.6562216056013552, + "learning_rate": 9.742308711105657e-06, + "loss": 0.3821, + "step": 5886 + }, + { + "epoch": 0.12958397534668722, + "grad_norm": 1.3860579573836669, + "learning_rate": 9.742195741001081e-06, + "loss": 0.3468, + "step": 5887 + }, + { + "epoch": 0.12960598723310587, + "grad_norm": 2.3179837573260147, + "learning_rate": 9.742082746794553e-06, + "loss": 0.3436, + "step": 5888 + }, + { + "epoch": 0.12962799911952455, + "grad_norm": 1.9129872769425411, + "learning_rate": 9.741969728486644e-06, + "loss": 0.3941, + "step": 5889 + }, + { + "epoch": 0.1296500110059432, + "grad_norm": 1.3636289558533559, + "learning_rate": 9.741856686077933e-06, + "loss": 0.3261, + "step": 5890 + }, + { + "epoch": 0.12967202289236188, + "grad_norm": 1.8866463851631616, + "learning_rate": 9.741743619568988e-06, + "loss": 0.3712, + "step": 5891 + }, + { + "epoch": 0.12969403477878055, + "grad_norm": 1.2495929263276175, + "learning_rate": 9.741630528960389e-06, + "loss": 0.3386, + "step": 5892 + }, + { + "epoch": 0.1297160466651992, + "grad_norm": 1.1475305751348421, + "learning_rate": 9.741517414252706e-06, + "loss": 0.2292, + "step": 5893 + }, + { + "epoch": 0.12973805855161788, + "grad_norm": 1.7053145571348434, + "learning_rate": 9.74140427544652e-06, + "loss": 0.3846, + "step": 5894 + }, + { + "epoch": 0.12976007043803653, + "grad_norm": 1.72192625561242, + "learning_rate": 9.741291112542401e-06, + "loss": 0.4183, + "step": 5895 + }, + { + "epoch": 0.1297820823244552, + "grad_norm": 1.6918386853330967, + "learning_rate": 9.741177925540926e-06, + "loss": 0.3731, + "step": 5896 + }, + { + "epoch": 0.12980409421087388, + "grad_norm": 1.9586740727174368, + "learning_rate": 9.741064714442669e-06, + "loss": 0.4374, + "step": 5897 + }, + { + "epoch": 0.12982610609729253, + "grad_norm": 1.4553409632796108, + "learning_rate": 9.740951479248209e-06, + "loss": 0.4572, + "step": 5898 + }, + { + "epoch": 0.1298481179837112, + "grad_norm": 1.3685483769910267, + "learning_rate": 9.740838219958118e-06, + "loss": 0.278, + "step": 5899 + }, + { + "epoch": 0.12987012987012986, + "grad_norm": 1.259650509633382, + "learning_rate": 9.740724936572972e-06, + "loss": 0.2983, + "step": 5900 + }, + { + "epoch": 0.12989214175654853, + "grad_norm": 0.734122210245128, + "learning_rate": 9.740611629093348e-06, + "loss": 0.572, + "step": 5901 + }, + { + "epoch": 0.1299141536429672, + "grad_norm": 1.6460890806611759, + "learning_rate": 9.740498297519822e-06, + "loss": 0.3239, + "step": 5902 + }, + { + "epoch": 0.12993616552938586, + "grad_norm": 1.2261106023845854, + "learning_rate": 9.74038494185297e-06, + "loss": 0.2803, + "step": 5903 + }, + { + "epoch": 0.12995817741580454, + "grad_norm": 0.646868326721972, + "learning_rate": 9.740271562093366e-06, + "loss": 0.5481, + "step": 5904 + }, + { + "epoch": 0.1299801893022232, + "grad_norm": 1.4613238467258858, + "learning_rate": 9.740158158241586e-06, + "loss": 0.3244, + "step": 5905 + }, + { + "epoch": 0.13000220118864186, + "grad_norm": 1.6407660585431318, + "learning_rate": 9.74004473029821e-06, + "loss": 0.3669, + "step": 5906 + }, + { + "epoch": 0.13002421307506054, + "grad_norm": 0.6713957004741418, + "learning_rate": 9.739931278263813e-06, + "loss": 0.5584, + "step": 5907 + }, + { + "epoch": 0.1300462249614792, + "grad_norm": 1.2278950052646, + "learning_rate": 9.73981780213897e-06, + "loss": 0.3611, + "step": 5908 + }, + { + "epoch": 0.13006823684789787, + "grad_norm": 0.6729383553770262, + "learning_rate": 9.73970430192426e-06, + "loss": 0.5534, + "step": 5909 + }, + { + "epoch": 0.13009024873431654, + "grad_norm": 1.5598842902938117, + "learning_rate": 9.739590777620259e-06, + "loss": 0.3795, + "step": 5910 + }, + { + "epoch": 0.1301122606207352, + "grad_norm": 1.4197291646089552, + "learning_rate": 9.739477229227544e-06, + "loss": 0.3013, + "step": 5911 + }, + { + "epoch": 0.13013427250715387, + "grad_norm": 3.9903402604597797, + "learning_rate": 9.73936365674669e-06, + "loss": 0.4014, + "step": 5912 + }, + { + "epoch": 0.13015628439357252, + "grad_norm": 1.6879189622930055, + "learning_rate": 9.739250060178277e-06, + "loss": 0.3043, + "step": 5913 + }, + { + "epoch": 0.1301782962799912, + "grad_norm": 2.0052913517070845, + "learning_rate": 9.739136439522882e-06, + "loss": 0.3763, + "step": 5914 + }, + { + "epoch": 0.13020030816640987, + "grad_norm": 1.2612399757195594, + "learning_rate": 9.739022794781082e-06, + "loss": 0.3548, + "step": 5915 + }, + { + "epoch": 0.13022232005282852, + "grad_norm": 0.6411514207240672, + "learning_rate": 9.738909125953452e-06, + "loss": 0.5575, + "step": 5916 + }, + { + "epoch": 0.1302443319392472, + "grad_norm": 1.6224003436865269, + "learning_rate": 9.738795433040575e-06, + "loss": 0.3838, + "step": 5917 + }, + { + "epoch": 0.13026634382566585, + "grad_norm": 0.6571878263522154, + "learning_rate": 9.738681716043026e-06, + "loss": 0.5652, + "step": 5918 + }, + { + "epoch": 0.13028835571208452, + "grad_norm": 2.480306514774733, + "learning_rate": 9.738567974961382e-06, + "loss": 0.3543, + "step": 5919 + }, + { + "epoch": 0.1303103675985032, + "grad_norm": 1.5258242667436213, + "learning_rate": 9.738454209796221e-06, + "loss": 0.2872, + "step": 5920 + }, + { + "epoch": 0.13033237948492185, + "grad_norm": 1.41036671423933, + "learning_rate": 9.738340420548125e-06, + "loss": 0.3938, + "step": 5921 + }, + { + "epoch": 0.13035439137134053, + "grad_norm": 0.6625302961398603, + "learning_rate": 9.738226607217668e-06, + "loss": 0.533, + "step": 5922 + }, + { + "epoch": 0.13037640325775918, + "grad_norm": 1.7221642116375973, + "learning_rate": 9.73811276980543e-06, + "loss": 0.3443, + "step": 5923 + }, + { + "epoch": 0.13039841514417785, + "grad_norm": 1.7180063785153665, + "learning_rate": 9.73799890831199e-06, + "loss": 0.3128, + "step": 5924 + }, + { + "epoch": 0.13042042703059653, + "grad_norm": 2.036246909822814, + "learning_rate": 9.737885022737926e-06, + "loss": 0.369, + "step": 5925 + }, + { + "epoch": 0.13044243891701518, + "grad_norm": 3.5633973890173003, + "learning_rate": 9.737771113083819e-06, + "loss": 0.3563, + "step": 5926 + }, + { + "epoch": 0.13046445080343386, + "grad_norm": 1.9716195810455441, + "learning_rate": 9.737657179350244e-06, + "loss": 0.3487, + "step": 5927 + }, + { + "epoch": 0.13048646268985253, + "grad_norm": 1.7097937055839258, + "learning_rate": 9.737543221537782e-06, + "loss": 0.3621, + "step": 5928 + }, + { + "epoch": 0.13050847457627118, + "grad_norm": 1.8958933925200294, + "learning_rate": 9.737429239647012e-06, + "loss": 0.3245, + "step": 5929 + }, + { + "epoch": 0.13053048646268986, + "grad_norm": 1.3725897713134811, + "learning_rate": 9.737315233678514e-06, + "loss": 0.3421, + "step": 5930 + }, + { + "epoch": 0.1305524983491085, + "grad_norm": 1.6560290338645531, + "learning_rate": 9.737201203632868e-06, + "loss": 0.415, + "step": 5931 + }, + { + "epoch": 0.13057451023552719, + "grad_norm": 1.5408924047705406, + "learning_rate": 9.737087149510653e-06, + "loss": 0.3781, + "step": 5932 + }, + { + "epoch": 0.13059652212194586, + "grad_norm": 1.7060831239481555, + "learning_rate": 9.736973071312447e-06, + "loss": 0.4088, + "step": 5933 + }, + { + "epoch": 0.1306185340083645, + "grad_norm": 2.6597576308910096, + "learning_rate": 9.736858969038834e-06, + "loss": 0.3354, + "step": 5934 + }, + { + "epoch": 0.1306405458947832, + "grad_norm": 0.7314621992032226, + "learning_rate": 9.736744842690389e-06, + "loss": 0.5587, + "step": 5935 + }, + { + "epoch": 0.13066255778120184, + "grad_norm": 1.317130336391094, + "learning_rate": 9.736630692267693e-06, + "loss": 0.3597, + "step": 5936 + }, + { + "epoch": 0.13068456966762051, + "grad_norm": 1.2315337159962636, + "learning_rate": 9.736516517771328e-06, + "loss": 0.354, + "step": 5937 + }, + { + "epoch": 0.1307065815540392, + "grad_norm": 1.5281769545572024, + "learning_rate": 9.736402319201873e-06, + "loss": 0.3667, + "step": 5938 + }, + { + "epoch": 0.13072859344045784, + "grad_norm": 0.6500704660416117, + "learning_rate": 9.736288096559909e-06, + "loss": 0.556, + "step": 5939 + }, + { + "epoch": 0.13075060532687652, + "grad_norm": 1.6442004269176127, + "learning_rate": 9.736173849846018e-06, + "loss": 0.3554, + "step": 5940 + }, + { + "epoch": 0.13077261721329517, + "grad_norm": 0.6423306595612993, + "learning_rate": 9.736059579060779e-06, + "loss": 0.549, + "step": 5941 + }, + { + "epoch": 0.13079462909971384, + "grad_norm": 1.732855010541684, + "learning_rate": 9.73594528420477e-06, + "loss": 0.3222, + "step": 5942 + }, + { + "epoch": 0.13081664098613252, + "grad_norm": 1.9815301366862605, + "learning_rate": 9.735830965278579e-06, + "loss": 0.4922, + "step": 5943 + }, + { + "epoch": 0.13083865287255117, + "grad_norm": 1.3877201237235615, + "learning_rate": 9.73571662228278e-06, + "loss": 0.3243, + "step": 5944 + }, + { + "epoch": 0.13086066475896985, + "grad_norm": 2.875918191445142, + "learning_rate": 9.735602255217957e-06, + "loss": 0.463, + "step": 5945 + }, + { + "epoch": 0.1308826766453885, + "grad_norm": 1.6330534815826991, + "learning_rate": 9.735487864084693e-06, + "loss": 0.3549, + "step": 5946 + }, + { + "epoch": 0.13090468853180717, + "grad_norm": 2.941482234938923, + "learning_rate": 9.735373448883565e-06, + "loss": 0.3782, + "step": 5947 + }, + { + "epoch": 0.13092670041822585, + "grad_norm": 1.49075039144338, + "learning_rate": 9.735259009615158e-06, + "loss": 0.3716, + "step": 5948 + }, + { + "epoch": 0.1309487123046445, + "grad_norm": 0.6759953157163275, + "learning_rate": 9.735144546280053e-06, + "loss": 0.5543, + "step": 5949 + }, + { + "epoch": 0.13097072419106318, + "grad_norm": 4.220550929627762, + "learning_rate": 9.735030058878832e-06, + "loss": 0.2656, + "step": 5950 + }, + { + "epoch": 0.13099273607748185, + "grad_norm": 1.7141437303801408, + "learning_rate": 9.734915547412076e-06, + "loss": 0.3476, + "step": 5951 + }, + { + "epoch": 0.1310147479639005, + "grad_norm": 1.612688846240683, + "learning_rate": 9.734801011880367e-06, + "loss": 0.3407, + "step": 5952 + }, + { + "epoch": 0.13103675985031918, + "grad_norm": 2.639135251419945, + "learning_rate": 9.734686452284288e-06, + "loss": 0.349, + "step": 5953 + }, + { + "epoch": 0.13105877173673783, + "grad_norm": 3.679533106512397, + "learning_rate": 9.73457186862442e-06, + "loss": 0.4126, + "step": 5954 + }, + { + "epoch": 0.1310807836231565, + "grad_norm": 1.6731151728687774, + "learning_rate": 9.734457260901347e-06, + "loss": 0.3555, + "step": 5955 + }, + { + "epoch": 0.13110279550957518, + "grad_norm": 1.5490930626924486, + "learning_rate": 9.73434262911565e-06, + "loss": 0.3661, + "step": 5956 + }, + { + "epoch": 0.13112480739599383, + "grad_norm": 1.4368551281725028, + "learning_rate": 9.734227973267912e-06, + "loss": 0.3845, + "step": 5957 + }, + { + "epoch": 0.1311468192824125, + "grad_norm": 1.246815531047142, + "learning_rate": 9.734113293358717e-06, + "loss": 0.2669, + "step": 5958 + }, + { + "epoch": 0.13116883116883116, + "grad_norm": 2.157100164271806, + "learning_rate": 9.733998589388645e-06, + "loss": 0.4305, + "step": 5959 + }, + { + "epoch": 0.13119084305524983, + "grad_norm": 1.5827781369219491, + "learning_rate": 9.733883861358282e-06, + "loss": 0.361, + "step": 5960 + }, + { + "epoch": 0.1312128549416685, + "grad_norm": 1.6253876623576566, + "learning_rate": 9.733769109268209e-06, + "loss": 0.3705, + "step": 5961 + }, + { + "epoch": 0.13123486682808716, + "grad_norm": 1.5025604115054798, + "learning_rate": 9.73365433311901e-06, + "loss": 0.323, + "step": 5962 + }, + { + "epoch": 0.13125687871450584, + "grad_norm": 1.40718303108885, + "learning_rate": 9.733539532911271e-06, + "loss": 0.3526, + "step": 5963 + }, + { + "epoch": 0.13127889060092449, + "grad_norm": 2.018409605952446, + "learning_rate": 9.733424708645571e-06, + "loss": 0.3497, + "step": 5964 + }, + { + "epoch": 0.13130090248734316, + "grad_norm": 1.3834895926915158, + "learning_rate": 9.733309860322494e-06, + "loss": 0.4039, + "step": 5965 + }, + { + "epoch": 0.13132291437376184, + "grad_norm": 1.4096192935256604, + "learning_rate": 9.733194987942627e-06, + "loss": 0.3857, + "step": 5966 + }, + { + "epoch": 0.1313449262601805, + "grad_norm": 1.4172148007405498, + "learning_rate": 9.733080091506552e-06, + "loss": 0.3336, + "step": 5967 + }, + { + "epoch": 0.13136693814659917, + "grad_norm": 0.6688990541353124, + "learning_rate": 9.732965171014852e-06, + "loss": 0.5465, + "step": 5968 + }, + { + "epoch": 0.13138895003301784, + "grad_norm": 2.0559524889018284, + "learning_rate": 9.732850226468114e-06, + "loss": 0.3693, + "step": 5969 + }, + { + "epoch": 0.1314109619194365, + "grad_norm": 1.3899403496074167, + "learning_rate": 9.732735257866919e-06, + "loss": 0.3577, + "step": 5970 + }, + { + "epoch": 0.13143297380585517, + "grad_norm": 1.524081107212449, + "learning_rate": 9.732620265211852e-06, + "loss": 0.3486, + "step": 5971 + }, + { + "epoch": 0.13145498569227382, + "grad_norm": 0.6091489227313107, + "learning_rate": 9.7325052485035e-06, + "loss": 0.5168, + "step": 5972 + }, + { + "epoch": 0.1314769975786925, + "grad_norm": 1.772312670344652, + "learning_rate": 9.732390207742441e-06, + "loss": 0.4219, + "step": 5973 + }, + { + "epoch": 0.13149900946511117, + "grad_norm": 1.4243274649234097, + "learning_rate": 9.732275142929268e-06, + "loss": 0.3568, + "step": 5974 + }, + { + "epoch": 0.13152102135152982, + "grad_norm": 1.7574615935411044, + "learning_rate": 9.732160054064562e-06, + "loss": 0.3915, + "step": 5975 + }, + { + "epoch": 0.1315430332379485, + "grad_norm": 1.4426285924165363, + "learning_rate": 9.732044941148907e-06, + "loss": 0.3926, + "step": 5976 + }, + { + "epoch": 0.13156504512436715, + "grad_norm": 2.469294142605935, + "learning_rate": 9.731929804182889e-06, + "loss": 0.41, + "step": 5977 + }, + { + "epoch": 0.13158705701078582, + "grad_norm": 1.3114607885526584, + "learning_rate": 9.731814643167095e-06, + "loss": 0.3685, + "step": 5978 + }, + { + "epoch": 0.1316090688972045, + "grad_norm": 2.1455841476286697, + "learning_rate": 9.731699458102105e-06, + "loss": 0.352, + "step": 5979 + }, + { + "epoch": 0.13163108078362315, + "grad_norm": 1.4905602381719643, + "learning_rate": 9.731584248988511e-06, + "loss": 0.3223, + "step": 5980 + }, + { + "epoch": 0.13165309267004183, + "grad_norm": 1.3257263801654202, + "learning_rate": 9.731469015826895e-06, + "loss": 0.3863, + "step": 5981 + }, + { + "epoch": 0.13167510455646048, + "grad_norm": 0.6697203071691165, + "learning_rate": 9.731353758617843e-06, + "loss": 0.5474, + "step": 5982 + }, + { + "epoch": 0.13169711644287915, + "grad_norm": 1.5514801220483923, + "learning_rate": 9.73123847736194e-06, + "loss": 0.3997, + "step": 5983 + }, + { + "epoch": 0.13171912832929783, + "grad_norm": 1.574747118519315, + "learning_rate": 9.731123172059773e-06, + "loss": 0.3095, + "step": 5984 + }, + { + "epoch": 0.13174114021571648, + "grad_norm": 1.3938664694835683, + "learning_rate": 9.731007842711929e-06, + "loss": 0.3648, + "step": 5985 + }, + { + "epoch": 0.13176315210213516, + "grad_norm": 1.743498081154604, + "learning_rate": 9.730892489318992e-06, + "loss": 0.3364, + "step": 5986 + }, + { + "epoch": 0.13178516398855383, + "grad_norm": 1.4977785225948854, + "learning_rate": 9.73077711188155e-06, + "loss": 0.3388, + "step": 5987 + }, + { + "epoch": 0.13180717587497248, + "grad_norm": 1.7394672874143937, + "learning_rate": 9.730661710400188e-06, + "loss": 0.4645, + "step": 5988 + }, + { + "epoch": 0.13182918776139116, + "grad_norm": 2.046011416305206, + "learning_rate": 9.730546284875495e-06, + "loss": 0.354, + "step": 5989 + }, + { + "epoch": 0.1318511996478098, + "grad_norm": 1.6088741931616748, + "learning_rate": 9.730430835308056e-06, + "loss": 0.4311, + "step": 5990 + }, + { + "epoch": 0.13187321153422848, + "grad_norm": 1.6822015139856075, + "learning_rate": 9.730315361698454e-06, + "loss": 0.3055, + "step": 5991 + }, + { + "epoch": 0.13189522342064716, + "grad_norm": 1.305346680474915, + "learning_rate": 9.730199864047282e-06, + "loss": 0.3164, + "step": 5992 + }, + { + "epoch": 0.1319172353070658, + "grad_norm": 1.9848287842935743, + "learning_rate": 9.730084342355127e-06, + "loss": 0.5052, + "step": 5993 + }, + { + "epoch": 0.1319392471934845, + "grad_norm": 1.5800179015555809, + "learning_rate": 9.72996879662257e-06, + "loss": 0.3408, + "step": 5994 + }, + { + "epoch": 0.13196125907990314, + "grad_norm": 1.6027580943071422, + "learning_rate": 9.729853226850204e-06, + "loss": 0.4165, + "step": 5995 + }, + { + "epoch": 0.13198327096632181, + "grad_norm": 0.6647165061023376, + "learning_rate": 9.729737633038614e-06, + "loss": 0.5406, + "step": 5996 + }, + { + "epoch": 0.1320052828527405, + "grad_norm": 1.7870001514762055, + "learning_rate": 9.729622015188386e-06, + "loss": 0.3183, + "step": 5997 + }, + { + "epoch": 0.13202729473915914, + "grad_norm": 1.2268164319678792, + "learning_rate": 9.72950637330011e-06, + "loss": 0.3989, + "step": 5998 + }, + { + "epoch": 0.13204930662557782, + "grad_norm": 1.646019237075447, + "learning_rate": 9.729390707374375e-06, + "loss": 0.4542, + "step": 5999 + }, + { + "epoch": 0.13207131851199647, + "grad_norm": 1.717911504275973, + "learning_rate": 9.729275017411765e-06, + "loss": 0.4436, + "step": 6000 + }, + { + "epoch": 0.13209333039841514, + "grad_norm": 1.4351438799026093, + "learning_rate": 9.729159303412871e-06, + "loss": 0.3938, + "step": 6001 + }, + { + "epoch": 0.13211534228483382, + "grad_norm": 1.3014472607367624, + "learning_rate": 9.72904356537828e-06, + "loss": 0.3286, + "step": 6002 + }, + { + "epoch": 0.13213735417125247, + "grad_norm": 1.1281018307504842, + "learning_rate": 9.728927803308581e-06, + "loss": 0.3792, + "step": 6003 + }, + { + "epoch": 0.13215936605767115, + "grad_norm": 1.8191654095463472, + "learning_rate": 9.728812017204363e-06, + "loss": 0.3446, + "step": 6004 + }, + { + "epoch": 0.1321813779440898, + "grad_norm": 0.6951576544174424, + "learning_rate": 9.728696207066211e-06, + "loss": 0.5444, + "step": 6005 + }, + { + "epoch": 0.13220338983050847, + "grad_norm": 0.6746990698149418, + "learning_rate": 9.728580372894716e-06, + "loss": 0.5612, + "step": 6006 + }, + { + "epoch": 0.13222540171692715, + "grad_norm": 0.6290036757558893, + "learning_rate": 9.728464514690468e-06, + "loss": 0.5714, + "step": 6007 + }, + { + "epoch": 0.1322474136033458, + "grad_norm": 1.3629448210829913, + "learning_rate": 9.728348632454053e-06, + "loss": 0.2804, + "step": 6008 + }, + { + "epoch": 0.13226942548976448, + "grad_norm": 1.5132813492524912, + "learning_rate": 9.72823272618606e-06, + "loss": 0.4305, + "step": 6009 + }, + { + "epoch": 0.13229143737618315, + "grad_norm": 1.6134199450382123, + "learning_rate": 9.728116795887082e-06, + "loss": 0.3259, + "step": 6010 + }, + { + "epoch": 0.1323134492626018, + "grad_norm": 1.4225393417760555, + "learning_rate": 9.728000841557705e-06, + "loss": 0.3843, + "step": 6011 + }, + { + "epoch": 0.13233546114902048, + "grad_norm": 1.6942034347287949, + "learning_rate": 9.727884863198519e-06, + "loss": 0.351, + "step": 6012 + }, + { + "epoch": 0.13235747303543913, + "grad_norm": 2.540557607258255, + "learning_rate": 9.727768860810112e-06, + "loss": 0.3449, + "step": 6013 + }, + { + "epoch": 0.1323794849218578, + "grad_norm": 1.5065922491847068, + "learning_rate": 9.727652834393078e-06, + "loss": 0.3274, + "step": 6014 + }, + { + "epoch": 0.13240149680827648, + "grad_norm": 2.129647324702184, + "learning_rate": 9.727536783948e-06, + "loss": 0.3155, + "step": 6015 + }, + { + "epoch": 0.13242350869469513, + "grad_norm": 1.3157019931590936, + "learning_rate": 9.727420709475475e-06, + "loss": 0.3154, + "step": 6016 + }, + { + "epoch": 0.1324455205811138, + "grad_norm": 1.1019787989018819, + "learning_rate": 9.727304610976086e-06, + "loss": 0.5698, + "step": 6017 + }, + { + "epoch": 0.13246753246753246, + "grad_norm": 0.8723955170041742, + "learning_rate": 9.72718848845043e-06, + "loss": 0.5572, + "step": 6018 + }, + { + "epoch": 0.13248954435395113, + "grad_norm": 0.6792210417036555, + "learning_rate": 9.727072341899092e-06, + "loss": 0.557, + "step": 6019 + }, + { + "epoch": 0.1325115562403698, + "grad_norm": 1.4439448108176907, + "learning_rate": 9.726956171322664e-06, + "loss": 0.4003, + "step": 6020 + }, + { + "epoch": 0.13253356812678846, + "grad_norm": 1.6124464603971558, + "learning_rate": 9.726839976721736e-06, + "loss": 0.3242, + "step": 6021 + }, + { + "epoch": 0.13255558001320714, + "grad_norm": 1.6539693854283555, + "learning_rate": 9.7267237580969e-06, + "loss": 0.3652, + "step": 6022 + }, + { + "epoch": 0.13257759189962579, + "grad_norm": 1.4645857990050601, + "learning_rate": 9.726607515448744e-06, + "loss": 0.3636, + "step": 6023 + }, + { + "epoch": 0.13259960378604446, + "grad_norm": 1.5572450937757643, + "learning_rate": 9.726491248777862e-06, + "loss": 0.3905, + "step": 6024 + }, + { + "epoch": 0.13262161567246314, + "grad_norm": 1.3564854248729976, + "learning_rate": 9.726374958084844e-06, + "loss": 0.3168, + "step": 6025 + }, + { + "epoch": 0.1326436275588818, + "grad_norm": 1.7938077186979928, + "learning_rate": 9.726258643370279e-06, + "loss": 0.3708, + "step": 6026 + }, + { + "epoch": 0.13266563944530047, + "grad_norm": 1.4894575610349035, + "learning_rate": 9.72614230463476e-06, + "loss": 0.4067, + "step": 6027 + }, + { + "epoch": 0.13268765133171914, + "grad_norm": 2.166586212906468, + "learning_rate": 9.726025941878879e-06, + "loss": 0.6402, + "step": 6028 + }, + { + "epoch": 0.1327096632181378, + "grad_norm": 1.3969820632026666, + "learning_rate": 9.725909555103226e-06, + "loss": 0.5726, + "step": 6029 + }, + { + "epoch": 0.13273167510455647, + "grad_norm": 1.428456794765985, + "learning_rate": 9.72579314430839e-06, + "loss": 0.3958, + "step": 6030 + }, + { + "epoch": 0.13275368699097512, + "grad_norm": 0.8360256523092467, + "learning_rate": 9.72567670949497e-06, + "loss": 0.5675, + "step": 6031 + }, + { + "epoch": 0.1327756988773938, + "grad_norm": 1.6073266116632783, + "learning_rate": 9.72556025066355e-06, + "loss": 0.395, + "step": 6032 + }, + { + "epoch": 0.13279771076381247, + "grad_norm": 1.7143888735584871, + "learning_rate": 9.725443767814726e-06, + "loss": 0.3676, + "step": 6033 + }, + { + "epoch": 0.13281972265023112, + "grad_norm": 1.6981595259712061, + "learning_rate": 9.72532726094909e-06, + "loss": 0.3697, + "step": 6034 + }, + { + "epoch": 0.1328417345366498, + "grad_norm": 1.395353142118217, + "learning_rate": 9.725210730067232e-06, + "loss": 0.4158, + "step": 6035 + }, + { + "epoch": 0.13286374642306845, + "grad_norm": 1.6809609490711361, + "learning_rate": 9.725094175169747e-06, + "loss": 0.3796, + "step": 6036 + }, + { + "epoch": 0.13288575830948712, + "grad_norm": 2.2829079942084385, + "learning_rate": 9.724977596257223e-06, + "loss": 0.6362, + "step": 6037 + }, + { + "epoch": 0.1329077701959058, + "grad_norm": 1.241690361933407, + "learning_rate": 9.724860993330258e-06, + "loss": 0.3831, + "step": 6038 + }, + { + "epoch": 0.13292978208232445, + "grad_norm": 1.3024978596774004, + "learning_rate": 9.724744366389443e-06, + "loss": 0.3277, + "step": 6039 + }, + { + "epoch": 0.13295179396874313, + "grad_norm": 1.4327875604190625, + "learning_rate": 9.724627715435368e-06, + "loss": 0.3749, + "step": 6040 + }, + { + "epoch": 0.13297380585516178, + "grad_norm": 0.950681455354147, + "learning_rate": 9.724511040468629e-06, + "loss": 0.5758, + "step": 6041 + }, + { + "epoch": 0.13299581774158045, + "grad_norm": 1.4360697178439688, + "learning_rate": 9.724394341489816e-06, + "loss": 0.2975, + "step": 6042 + }, + { + "epoch": 0.13301782962799913, + "grad_norm": 2.33710173194457, + "learning_rate": 9.724277618499524e-06, + "loss": 0.3053, + "step": 6043 + }, + { + "epoch": 0.13303984151441778, + "grad_norm": 1.0044924317238442, + "learning_rate": 9.724160871498347e-06, + "loss": 0.5837, + "step": 6044 + }, + { + "epoch": 0.13306185340083646, + "grad_norm": 1.5977665234233263, + "learning_rate": 9.724044100486879e-06, + "loss": 0.3557, + "step": 6045 + }, + { + "epoch": 0.1330838652872551, + "grad_norm": 1.6542307055849452, + "learning_rate": 9.723927305465708e-06, + "loss": 0.3617, + "step": 6046 + }, + { + "epoch": 0.13310587717367378, + "grad_norm": 1.2941345101302213, + "learning_rate": 9.723810486435435e-06, + "loss": 0.3439, + "step": 6047 + }, + { + "epoch": 0.13312788906009246, + "grad_norm": 1.6246505731085943, + "learning_rate": 9.723693643396647e-06, + "loss": 0.4574, + "step": 6048 + }, + { + "epoch": 0.1331499009465111, + "grad_norm": 1.332632671637085, + "learning_rate": 9.723576776349944e-06, + "loss": 0.2989, + "step": 6049 + }, + { + "epoch": 0.13317191283292978, + "grad_norm": 1.236659977890513, + "learning_rate": 9.723459885295916e-06, + "loss": 0.3503, + "step": 6050 + }, + { + "epoch": 0.13319392471934846, + "grad_norm": 1.5161013048161338, + "learning_rate": 9.723342970235157e-06, + "loss": 0.3532, + "step": 6051 + }, + { + "epoch": 0.1332159366057671, + "grad_norm": 3.8105890292455684, + "learning_rate": 9.723226031168264e-06, + "loss": 0.3467, + "step": 6052 + }, + { + "epoch": 0.1332379484921858, + "grad_norm": 3.3456708717790895, + "learning_rate": 9.723109068095828e-06, + "loss": 0.3671, + "step": 6053 + }, + { + "epoch": 0.13325996037860444, + "grad_norm": 1.586732542957801, + "learning_rate": 9.722992081018447e-06, + "loss": 0.337, + "step": 6054 + }, + { + "epoch": 0.1332819722650231, + "grad_norm": 1.2471077925140925, + "learning_rate": 9.722875069936713e-06, + "loss": 0.582, + "step": 6055 + }, + { + "epoch": 0.1333039841514418, + "grad_norm": 1.4928663696776867, + "learning_rate": 9.722758034851221e-06, + "loss": 0.4135, + "step": 6056 + }, + { + "epoch": 0.13332599603786044, + "grad_norm": 2.4022996775447463, + "learning_rate": 9.722640975762567e-06, + "loss": 0.415, + "step": 6057 + }, + { + "epoch": 0.13334800792427912, + "grad_norm": 1.3382117531871205, + "learning_rate": 9.722523892671345e-06, + "loss": 0.426, + "step": 6058 + }, + { + "epoch": 0.13337001981069777, + "grad_norm": 2.1057236550900265, + "learning_rate": 9.72240678557815e-06, + "loss": 0.4067, + "step": 6059 + }, + { + "epoch": 0.13339203169711644, + "grad_norm": 0.7535253906505799, + "learning_rate": 9.722289654483578e-06, + "loss": 0.5703, + "step": 6060 + }, + { + "epoch": 0.13341404358353512, + "grad_norm": 2.1321979675312734, + "learning_rate": 9.722172499388223e-06, + "loss": 0.3277, + "step": 6061 + }, + { + "epoch": 0.13343605546995377, + "grad_norm": 1.2741249044819218, + "learning_rate": 9.722055320292683e-06, + "loss": 0.3261, + "step": 6062 + }, + { + "epoch": 0.13345806735637245, + "grad_norm": 1.2456082727228397, + "learning_rate": 9.721938117197549e-06, + "loss": 0.3058, + "step": 6063 + }, + { + "epoch": 0.1334800792427911, + "grad_norm": 1.632599926623638, + "learning_rate": 9.721820890103421e-06, + "loss": 0.2843, + "step": 6064 + }, + { + "epoch": 0.13350209112920977, + "grad_norm": 1.5760309496941938, + "learning_rate": 9.721703639010893e-06, + "loss": 0.2631, + "step": 6065 + }, + { + "epoch": 0.13352410301562845, + "grad_norm": 1.3015025171362917, + "learning_rate": 9.72158636392056e-06, + "loss": 0.2873, + "step": 6066 + }, + { + "epoch": 0.1335461149020471, + "grad_norm": 2.634224726037409, + "learning_rate": 9.721469064833022e-06, + "loss": 0.4003, + "step": 6067 + }, + { + "epoch": 0.13356812678846577, + "grad_norm": 1.2371766336452834, + "learning_rate": 9.721351741748872e-06, + "loss": 0.307, + "step": 6068 + }, + { + "epoch": 0.13359013867488445, + "grad_norm": 1.9721589883488226, + "learning_rate": 9.721234394668704e-06, + "loss": 0.2557, + "step": 6069 + }, + { + "epoch": 0.1336121505613031, + "grad_norm": 0.9046569011362576, + "learning_rate": 9.72111702359312e-06, + "loss": 0.5966, + "step": 6070 + }, + { + "epoch": 0.13363416244772178, + "grad_norm": 0.7992342946494967, + "learning_rate": 9.720999628522712e-06, + "loss": 0.5713, + "step": 6071 + }, + { + "epoch": 0.13365617433414043, + "grad_norm": 1.6017264517437104, + "learning_rate": 9.720882209458078e-06, + "loss": 0.4581, + "step": 6072 + }, + { + "epoch": 0.1336781862205591, + "grad_norm": 1.7637254365034076, + "learning_rate": 9.720764766399815e-06, + "loss": 0.3538, + "step": 6073 + }, + { + "epoch": 0.13370019810697778, + "grad_norm": 1.330869245000906, + "learning_rate": 9.72064729934852e-06, + "loss": 0.324, + "step": 6074 + }, + { + "epoch": 0.13372220999339643, + "grad_norm": 0.689728265461838, + "learning_rate": 9.72052980830479e-06, + "loss": 0.5369, + "step": 6075 + }, + { + "epoch": 0.1337442218798151, + "grad_norm": 0.7410872129097177, + "learning_rate": 9.720412293269224e-06, + "loss": 0.5575, + "step": 6076 + }, + { + "epoch": 0.13376623376623376, + "grad_norm": 0.7088812514613065, + "learning_rate": 9.720294754242415e-06, + "loss": 0.5591, + "step": 6077 + }, + { + "epoch": 0.13378824565265243, + "grad_norm": 1.6102820111873992, + "learning_rate": 9.720177191224965e-06, + "loss": 0.4207, + "step": 6078 + }, + { + "epoch": 0.1338102575390711, + "grad_norm": 1.4790967358346474, + "learning_rate": 9.720059604217466e-06, + "loss": 0.3859, + "step": 6079 + }, + { + "epoch": 0.13383226942548976, + "grad_norm": 1.3843397760023795, + "learning_rate": 9.719941993220521e-06, + "loss": 0.3302, + "step": 6080 + }, + { + "epoch": 0.13385428131190844, + "grad_norm": 1.4226964143911633, + "learning_rate": 9.719824358234727e-06, + "loss": 0.3974, + "step": 6081 + }, + { + "epoch": 0.13387629319832708, + "grad_norm": 1.6832238234647077, + "learning_rate": 9.719706699260678e-06, + "loss": 0.2779, + "step": 6082 + }, + { + "epoch": 0.13389830508474576, + "grad_norm": 1.4665524856520782, + "learning_rate": 9.719589016298975e-06, + "loss": 0.3855, + "step": 6083 + }, + { + "epoch": 0.13392031697116444, + "grad_norm": 1.5883637942373081, + "learning_rate": 9.719471309350215e-06, + "loss": 0.3939, + "step": 6084 + }, + { + "epoch": 0.1339423288575831, + "grad_norm": 1.6145172799383147, + "learning_rate": 9.719353578414998e-06, + "loss": 0.3388, + "step": 6085 + }, + { + "epoch": 0.13396434074400176, + "grad_norm": 1.3884628945000281, + "learning_rate": 9.719235823493923e-06, + "loss": 0.3503, + "step": 6086 + }, + { + "epoch": 0.1339863526304204, + "grad_norm": 1.3381114040678974, + "learning_rate": 9.719118044587583e-06, + "loss": 0.3926, + "step": 6087 + }, + { + "epoch": 0.1340083645168391, + "grad_norm": 1.308089136073661, + "learning_rate": 9.719000241696584e-06, + "loss": 0.3891, + "step": 6088 + }, + { + "epoch": 0.13403037640325777, + "grad_norm": 1.4059955103530972, + "learning_rate": 9.718882414821518e-06, + "loss": 0.3299, + "step": 6089 + }, + { + "epoch": 0.13405238828967642, + "grad_norm": 1.0962647562343362, + "learning_rate": 9.718764563962989e-06, + "loss": 0.5692, + "step": 6090 + }, + { + "epoch": 0.1340744001760951, + "grad_norm": 1.7096788086359178, + "learning_rate": 9.71864668912159e-06, + "loss": 0.3079, + "step": 6091 + }, + { + "epoch": 0.13409641206251377, + "grad_norm": 1.553697487124331, + "learning_rate": 9.718528790297928e-06, + "loss": 0.4276, + "step": 6092 + }, + { + "epoch": 0.13411842394893242, + "grad_norm": 2.1600272582429256, + "learning_rate": 9.718410867492598e-06, + "loss": 0.3102, + "step": 6093 + }, + { + "epoch": 0.1341404358353511, + "grad_norm": 1.6095978799142068, + "learning_rate": 9.718292920706199e-06, + "loss": 0.3703, + "step": 6094 + }, + { + "epoch": 0.13416244772176975, + "grad_norm": 1.300110924427689, + "learning_rate": 9.718174949939329e-06, + "loss": 0.3318, + "step": 6095 + }, + { + "epoch": 0.13418445960818842, + "grad_norm": 1.5786308720838693, + "learning_rate": 9.718056955192592e-06, + "loss": 0.3565, + "step": 6096 + }, + { + "epoch": 0.1342064714946071, + "grad_norm": 1.3835124091193776, + "learning_rate": 9.717938936466585e-06, + "loss": 0.3824, + "step": 6097 + }, + { + "epoch": 0.13422848338102575, + "grad_norm": 1.2772235203115332, + "learning_rate": 9.717820893761907e-06, + "loss": 0.278, + "step": 6098 + }, + { + "epoch": 0.13425049526744443, + "grad_norm": 0.7302509536254108, + "learning_rate": 9.717702827079159e-06, + "loss": 0.5267, + "step": 6099 + }, + { + "epoch": 0.13427250715386307, + "grad_norm": 1.5295165356618878, + "learning_rate": 9.717584736418941e-06, + "loss": 0.3876, + "step": 6100 + }, + { + "epoch": 0.13429451904028175, + "grad_norm": 1.44274568183694, + "learning_rate": 9.717466621781853e-06, + "loss": 0.3583, + "step": 6101 + }, + { + "epoch": 0.13431653092670043, + "grad_norm": 1.930334817362995, + "learning_rate": 9.717348483168497e-06, + "loss": 0.3717, + "step": 6102 + }, + { + "epoch": 0.13433854281311908, + "grad_norm": 0.6436434499053236, + "learning_rate": 9.717230320579471e-06, + "loss": 0.5834, + "step": 6103 + }, + { + "epoch": 0.13436055469953775, + "grad_norm": 1.689289331924816, + "learning_rate": 9.717112134015376e-06, + "loss": 0.3866, + "step": 6104 + }, + { + "epoch": 0.1343825665859564, + "grad_norm": 1.7799820586778057, + "learning_rate": 9.716993923476815e-06, + "loss": 0.3913, + "step": 6105 + }, + { + "epoch": 0.13440457847237508, + "grad_norm": 1.4299268801368041, + "learning_rate": 9.716875688964387e-06, + "loss": 0.4071, + "step": 6106 + }, + { + "epoch": 0.13442659035879376, + "grad_norm": 1.551434190402731, + "learning_rate": 9.716757430478692e-06, + "loss": 0.4023, + "step": 6107 + }, + { + "epoch": 0.1344486022452124, + "grad_norm": 0.6518952080228726, + "learning_rate": 9.716639148020332e-06, + "loss": 0.5281, + "step": 6108 + }, + { + "epoch": 0.13447061413163108, + "grad_norm": 0.6619623981254554, + "learning_rate": 9.716520841589907e-06, + "loss": 0.571, + "step": 6109 + }, + { + "epoch": 0.13449262601804976, + "grad_norm": 2.9990486386155037, + "learning_rate": 9.71640251118802e-06, + "loss": 0.3654, + "step": 6110 + }, + { + "epoch": 0.1345146379044684, + "grad_norm": 1.6259218994205558, + "learning_rate": 9.716284156815274e-06, + "loss": 0.402, + "step": 6111 + }, + { + "epoch": 0.1345366497908871, + "grad_norm": 1.6125933059349045, + "learning_rate": 9.716165778472267e-06, + "loss": 0.3741, + "step": 6112 + }, + { + "epoch": 0.13455866167730574, + "grad_norm": 0.6206342450935789, + "learning_rate": 9.716047376159602e-06, + "loss": 0.5685, + "step": 6113 + }, + { + "epoch": 0.1345806735637244, + "grad_norm": 0.6074425163381946, + "learning_rate": 9.715928949877881e-06, + "loss": 0.5562, + "step": 6114 + }, + { + "epoch": 0.1346026854501431, + "grad_norm": 1.51650788993933, + "learning_rate": 9.715810499627705e-06, + "loss": 0.4043, + "step": 6115 + }, + { + "epoch": 0.13462469733656174, + "grad_norm": 0.6175094715047085, + "learning_rate": 9.715692025409678e-06, + "loss": 0.571, + "step": 6116 + }, + { + "epoch": 0.13464670922298042, + "grad_norm": 1.5020941554074962, + "learning_rate": 9.7155735272244e-06, + "loss": 0.3224, + "step": 6117 + }, + { + "epoch": 0.13466872110939906, + "grad_norm": 1.4750276624943162, + "learning_rate": 9.715455005072474e-06, + "loss": 0.4345, + "step": 6118 + }, + { + "epoch": 0.13469073299581774, + "grad_norm": 1.2622544355780083, + "learning_rate": 9.715336458954503e-06, + "loss": 0.4604, + "step": 6119 + }, + { + "epoch": 0.13471274488223642, + "grad_norm": 1.7740665631167336, + "learning_rate": 9.715217888871089e-06, + "loss": 0.3868, + "step": 6120 + }, + { + "epoch": 0.13473475676865507, + "grad_norm": 1.6124543491644687, + "learning_rate": 9.715099294822835e-06, + "loss": 0.3947, + "step": 6121 + }, + { + "epoch": 0.13475676865507374, + "grad_norm": 1.5280248175115587, + "learning_rate": 9.714980676810341e-06, + "loss": 0.3751, + "step": 6122 + }, + { + "epoch": 0.1347787805414924, + "grad_norm": 1.426627131339646, + "learning_rate": 9.714862034834215e-06, + "loss": 0.4281, + "step": 6123 + }, + { + "epoch": 0.13480079242791107, + "grad_norm": 1.460805069728267, + "learning_rate": 9.714743368895057e-06, + "loss": 0.3934, + "step": 6124 + }, + { + "epoch": 0.13482280431432975, + "grad_norm": 5.438570703361439, + "learning_rate": 9.71462467899347e-06, + "loss": 0.491, + "step": 6125 + }, + { + "epoch": 0.1348448162007484, + "grad_norm": 1.3516626263626677, + "learning_rate": 9.714505965130057e-06, + "loss": 0.3388, + "step": 6126 + }, + { + "epoch": 0.13486682808716707, + "grad_norm": 1.3773124771243535, + "learning_rate": 9.714387227305422e-06, + "loss": 0.3676, + "step": 6127 + }, + { + "epoch": 0.13488883997358572, + "grad_norm": 1.6199644941742857, + "learning_rate": 9.714268465520169e-06, + "loss": 0.4207, + "step": 6128 + }, + { + "epoch": 0.1349108518600044, + "grad_norm": 1.4976393697728847, + "learning_rate": 9.7141496797749e-06, + "loss": 0.3598, + "step": 6129 + }, + { + "epoch": 0.13493286374642308, + "grad_norm": 1.5876866605440005, + "learning_rate": 9.714030870070222e-06, + "loss": 0.4007, + "step": 6130 + }, + { + "epoch": 0.13495487563284173, + "grad_norm": 1.466229731235677, + "learning_rate": 9.713912036406735e-06, + "loss": 0.3652, + "step": 6131 + }, + { + "epoch": 0.1349768875192604, + "grad_norm": 2.222188802851021, + "learning_rate": 9.713793178785044e-06, + "loss": 0.4933, + "step": 6132 + }, + { + "epoch": 0.13499889940567908, + "grad_norm": 1.519771349237781, + "learning_rate": 9.713674297205754e-06, + "loss": 0.3804, + "step": 6133 + }, + { + "epoch": 0.13502091129209773, + "grad_norm": 0.7358059542962724, + "learning_rate": 9.71355539166947e-06, + "loss": 0.5527, + "step": 6134 + }, + { + "epoch": 0.1350429231785164, + "grad_norm": 2.0915693314411836, + "learning_rate": 9.713436462176793e-06, + "loss": 0.4301, + "step": 6135 + }, + { + "epoch": 0.13506493506493505, + "grad_norm": 1.2114580585413415, + "learning_rate": 9.71331750872833e-06, + "loss": 0.3785, + "step": 6136 + }, + { + "epoch": 0.13508694695135373, + "grad_norm": 1.1450247645173723, + "learning_rate": 9.713198531324688e-06, + "loss": 0.3099, + "step": 6137 + }, + { + "epoch": 0.1351089588377724, + "grad_norm": 1.6605043485123163, + "learning_rate": 9.713079529966466e-06, + "loss": 0.3765, + "step": 6138 + }, + { + "epoch": 0.13513097072419106, + "grad_norm": 1.6104871119255728, + "learning_rate": 9.712960504654274e-06, + "loss": 0.3732, + "step": 6139 + }, + { + "epoch": 0.13515298261060973, + "grad_norm": 1.2540280659373222, + "learning_rate": 9.712841455388712e-06, + "loss": 0.3398, + "step": 6140 + }, + { + "epoch": 0.13517499449702838, + "grad_norm": 1.2240413157688599, + "learning_rate": 9.712722382170391e-06, + "loss": 0.3514, + "step": 6141 + }, + { + "epoch": 0.13519700638344706, + "grad_norm": 1.9258570472608436, + "learning_rate": 9.712603284999911e-06, + "loss": 0.3451, + "step": 6142 + }, + { + "epoch": 0.13521901826986574, + "grad_norm": 1.3850600049600708, + "learning_rate": 9.71248416387788e-06, + "loss": 0.2686, + "step": 6143 + }, + { + "epoch": 0.1352410301562844, + "grad_norm": 1.5329219034859427, + "learning_rate": 9.712365018804902e-06, + "loss": 0.4548, + "step": 6144 + }, + { + "epoch": 0.13526304204270306, + "grad_norm": 1.4701502653969656, + "learning_rate": 9.712245849781582e-06, + "loss": 0.3858, + "step": 6145 + }, + { + "epoch": 0.1352850539291217, + "grad_norm": 1.610481777113093, + "learning_rate": 9.71212665680853e-06, + "loss": 0.3825, + "step": 6146 + }, + { + "epoch": 0.1353070658155404, + "grad_norm": 1.3222508758180573, + "learning_rate": 9.712007439886346e-06, + "loss": 0.3171, + "step": 6147 + }, + { + "epoch": 0.13532907770195907, + "grad_norm": 1.6146232939986354, + "learning_rate": 9.71188819901564e-06, + "loss": 0.475, + "step": 6148 + }, + { + "epoch": 0.13535108958837772, + "grad_norm": 1.5315590304737252, + "learning_rate": 9.711768934197015e-06, + "loss": 0.3805, + "step": 6149 + }, + { + "epoch": 0.1353731014747964, + "grad_norm": 1.8004664049008634, + "learning_rate": 9.71164964543108e-06, + "loss": 0.2991, + "step": 6150 + }, + { + "epoch": 0.13539511336121507, + "grad_norm": 1.706709137011453, + "learning_rate": 9.71153033271844e-06, + "loss": 0.2892, + "step": 6151 + }, + { + "epoch": 0.13541712524763372, + "grad_norm": 1.457007083322515, + "learning_rate": 9.711410996059701e-06, + "loss": 0.3018, + "step": 6152 + }, + { + "epoch": 0.1354391371340524, + "grad_norm": 0.740132164103192, + "learning_rate": 9.71129163545547e-06, + "loss": 0.5213, + "step": 6153 + }, + { + "epoch": 0.13546114902047104, + "grad_norm": 1.4143291410165808, + "learning_rate": 9.711172250906354e-06, + "loss": 0.3728, + "step": 6154 + }, + { + "epoch": 0.13548316090688972, + "grad_norm": 1.4985497802129988, + "learning_rate": 9.71105284241296e-06, + "loss": 0.337, + "step": 6155 + }, + { + "epoch": 0.1355051727933084, + "grad_norm": 1.9157272471952362, + "learning_rate": 9.710933409975894e-06, + "loss": 0.3026, + "step": 6156 + }, + { + "epoch": 0.13552718467972705, + "grad_norm": 1.4581584633877533, + "learning_rate": 9.710813953595763e-06, + "loss": 0.4372, + "step": 6157 + }, + { + "epoch": 0.13554919656614572, + "grad_norm": 8.930282787404842, + "learning_rate": 9.710694473273173e-06, + "loss": 0.3116, + "step": 6158 + }, + { + "epoch": 0.13557120845256437, + "grad_norm": 1.3236964077441709, + "learning_rate": 9.710574969008735e-06, + "loss": 0.3605, + "step": 6159 + }, + { + "epoch": 0.13559322033898305, + "grad_norm": 1.8223501302320626, + "learning_rate": 9.710455440803052e-06, + "loss": 0.4023, + "step": 6160 + }, + { + "epoch": 0.13561523222540173, + "grad_norm": 0.7280689062730408, + "learning_rate": 9.710335888656734e-06, + "loss": 0.5504, + "step": 6161 + }, + { + "epoch": 0.13563724411182038, + "grad_norm": 1.269851252829662, + "learning_rate": 9.710216312570388e-06, + "loss": 0.3922, + "step": 6162 + }, + { + "epoch": 0.13565925599823905, + "grad_norm": 1.9507009257410464, + "learning_rate": 9.710096712544623e-06, + "loss": 0.3785, + "step": 6163 + }, + { + "epoch": 0.1356812678846577, + "grad_norm": 1.6312615799345582, + "learning_rate": 9.709977088580047e-06, + "loss": 0.4226, + "step": 6164 + }, + { + "epoch": 0.13570327977107638, + "grad_norm": 2.282333378866842, + "learning_rate": 9.709857440677264e-06, + "loss": 0.3543, + "step": 6165 + }, + { + "epoch": 0.13572529165749506, + "grad_norm": 2.240184459750591, + "learning_rate": 9.709737768836886e-06, + "loss": 0.4149, + "step": 6166 + }, + { + "epoch": 0.1357473035439137, + "grad_norm": 1.4868457003045867, + "learning_rate": 9.70961807305952e-06, + "loss": 0.3209, + "step": 6167 + }, + { + "epoch": 0.13576931543033238, + "grad_norm": 3.214987845764161, + "learning_rate": 9.709498353345774e-06, + "loss": 0.3774, + "step": 6168 + }, + { + "epoch": 0.13579132731675106, + "grad_norm": 1.2543088879147928, + "learning_rate": 9.709378609696256e-06, + "loss": 0.3904, + "step": 6169 + }, + { + "epoch": 0.1358133392031697, + "grad_norm": 1.2855907151148676, + "learning_rate": 9.709258842111578e-06, + "loss": 0.3603, + "step": 6170 + }, + { + "epoch": 0.13583535108958839, + "grad_norm": 1.8989118856809946, + "learning_rate": 9.709139050592344e-06, + "loss": 0.3604, + "step": 6171 + }, + { + "epoch": 0.13585736297600703, + "grad_norm": 1.4349954820546913, + "learning_rate": 9.709019235139165e-06, + "loss": 0.4003, + "step": 6172 + }, + { + "epoch": 0.1358793748624257, + "grad_norm": 0.6787910175424363, + "learning_rate": 9.708899395752651e-06, + "loss": 0.5339, + "step": 6173 + }, + { + "epoch": 0.1359013867488444, + "grad_norm": 2.0202246928063228, + "learning_rate": 9.70877953243341e-06, + "loss": 0.3708, + "step": 6174 + }, + { + "epoch": 0.13592339863526304, + "grad_norm": 2.567388023599687, + "learning_rate": 9.708659645182052e-06, + "loss": 0.3593, + "step": 6175 + }, + { + "epoch": 0.13594541052168171, + "grad_norm": 1.4472684103922488, + "learning_rate": 9.708539733999184e-06, + "loss": 0.341, + "step": 6176 + }, + { + "epoch": 0.13596742240810036, + "grad_norm": 1.7223381104490003, + "learning_rate": 9.708419798885418e-06, + "loss": 0.3756, + "step": 6177 + }, + { + "epoch": 0.13598943429451904, + "grad_norm": 0.6107130035390529, + "learning_rate": 9.708299839841362e-06, + "loss": 0.5523, + "step": 6178 + }, + { + "epoch": 0.13601144618093772, + "grad_norm": 2.0385764193843525, + "learning_rate": 9.708179856867625e-06, + "loss": 0.4048, + "step": 6179 + }, + { + "epoch": 0.13603345806735637, + "grad_norm": 1.5131408628133507, + "learning_rate": 9.70805984996482e-06, + "loss": 0.3693, + "step": 6180 + }, + { + "epoch": 0.13605546995377504, + "grad_norm": 2.510690354870823, + "learning_rate": 9.707939819133555e-06, + "loss": 0.3531, + "step": 6181 + }, + { + "epoch": 0.1360774818401937, + "grad_norm": 1.4465473652080274, + "learning_rate": 9.70781976437444e-06, + "loss": 0.3399, + "step": 6182 + }, + { + "epoch": 0.13609949372661237, + "grad_norm": 1.5650175457866624, + "learning_rate": 9.707699685688083e-06, + "loss": 0.3701, + "step": 6183 + }, + { + "epoch": 0.13612150561303105, + "grad_norm": 0.6553694449621734, + "learning_rate": 9.7075795830751e-06, + "loss": 0.5756, + "step": 6184 + }, + { + "epoch": 0.1361435174994497, + "grad_norm": 1.5611110200395504, + "learning_rate": 9.707459456536094e-06, + "loss": 0.4762, + "step": 6185 + }, + { + "epoch": 0.13616552938586837, + "grad_norm": 1.5560945054155566, + "learning_rate": 9.707339306071684e-06, + "loss": 0.3678, + "step": 6186 + }, + { + "epoch": 0.13618754127228702, + "grad_norm": 1.2286600530480638, + "learning_rate": 9.707219131682472e-06, + "loss": 0.3039, + "step": 6187 + }, + { + "epoch": 0.1362095531587057, + "grad_norm": 1.5811561725086238, + "learning_rate": 9.707098933369074e-06, + "loss": 0.4072, + "step": 6188 + }, + { + "epoch": 0.13623156504512438, + "grad_norm": 0.6360737975234277, + "learning_rate": 9.706978711132102e-06, + "loss": 0.5297, + "step": 6189 + }, + { + "epoch": 0.13625357693154302, + "grad_norm": 1.9194174858120245, + "learning_rate": 9.706858464972164e-06, + "loss": 0.415, + "step": 6190 + }, + { + "epoch": 0.1362755888179617, + "grad_norm": 0.6555135276587347, + "learning_rate": 9.70673819488987e-06, + "loss": 0.5521, + "step": 6191 + }, + { + "epoch": 0.13629760070438038, + "grad_norm": 0.6346916411191432, + "learning_rate": 9.706617900885836e-06, + "loss": 0.529, + "step": 6192 + }, + { + "epoch": 0.13631961259079903, + "grad_norm": 1.8992414900139858, + "learning_rate": 9.706497582960668e-06, + "loss": 0.3635, + "step": 6193 + }, + { + "epoch": 0.1363416244772177, + "grad_norm": 1.2901616394503579, + "learning_rate": 9.70637724111498e-06, + "loss": 0.2711, + "step": 6194 + }, + { + "epoch": 0.13636363636363635, + "grad_norm": 0.6407554496815459, + "learning_rate": 9.706256875349386e-06, + "loss": 0.5579, + "step": 6195 + }, + { + "epoch": 0.13638564825005503, + "grad_norm": 1.4193430485724172, + "learning_rate": 9.706136485664494e-06, + "loss": 0.3544, + "step": 6196 + }, + { + "epoch": 0.1364076601364737, + "grad_norm": 1.5595592643623406, + "learning_rate": 9.706016072060918e-06, + "loss": 0.3305, + "step": 6197 + }, + { + "epoch": 0.13642967202289236, + "grad_norm": 1.3263202046595588, + "learning_rate": 9.705895634539268e-06, + "loss": 0.3051, + "step": 6198 + }, + { + "epoch": 0.13645168390931103, + "grad_norm": 1.2824127506751082, + "learning_rate": 9.70577517310016e-06, + "loss": 0.3888, + "step": 6199 + }, + { + "epoch": 0.13647369579572968, + "grad_norm": 1.4073552289965465, + "learning_rate": 9.705654687744202e-06, + "loss": 0.3286, + "step": 6200 + }, + { + "epoch": 0.13649570768214836, + "grad_norm": 2.1343670197801328, + "learning_rate": 9.705534178472009e-06, + "loss": 0.4018, + "step": 6201 + }, + { + "epoch": 0.13651771956856704, + "grad_norm": 1.2632487534532082, + "learning_rate": 9.705413645284191e-06, + "loss": 0.3388, + "step": 6202 + }, + { + "epoch": 0.13653973145498569, + "grad_norm": 1.3499748539016514, + "learning_rate": 9.705293088181364e-06, + "loss": 0.3115, + "step": 6203 + }, + { + "epoch": 0.13656174334140436, + "grad_norm": 1.1810676006150027, + "learning_rate": 9.705172507164138e-06, + "loss": 0.3418, + "step": 6204 + }, + { + "epoch": 0.136583755227823, + "grad_norm": 1.1715550185309016, + "learning_rate": 9.705051902233126e-06, + "loss": 0.364, + "step": 6205 + }, + { + "epoch": 0.1366057671142417, + "grad_norm": 1.4932755682349346, + "learning_rate": 9.704931273388944e-06, + "loss": 0.3177, + "step": 6206 + }, + { + "epoch": 0.13662777900066037, + "grad_norm": 1.7546712855980895, + "learning_rate": 9.7048106206322e-06, + "loss": 0.3606, + "step": 6207 + }, + { + "epoch": 0.13664979088707901, + "grad_norm": 1.2409069391817351, + "learning_rate": 9.704689943963512e-06, + "loss": 0.3632, + "step": 6208 + }, + { + "epoch": 0.1366718027734977, + "grad_norm": 1.2477779313894934, + "learning_rate": 9.704569243383491e-06, + "loss": 0.3039, + "step": 6209 + }, + { + "epoch": 0.13669381465991637, + "grad_norm": 1.5272175314950795, + "learning_rate": 9.70444851889275e-06, + "loss": 0.3394, + "step": 6210 + }, + { + "epoch": 0.13671582654633502, + "grad_norm": 1.8610188550011235, + "learning_rate": 9.704327770491904e-06, + "loss": 0.3334, + "step": 6211 + }, + { + "epoch": 0.1367378384327537, + "grad_norm": 1.2499718333437608, + "learning_rate": 9.704206998181566e-06, + "loss": 0.3582, + "step": 6212 + }, + { + "epoch": 0.13675985031917234, + "grad_norm": 0.6638804958092341, + "learning_rate": 9.70408620196235e-06, + "loss": 0.5512, + "step": 6213 + }, + { + "epoch": 0.13678186220559102, + "grad_norm": 1.4720057851524484, + "learning_rate": 9.703965381834872e-06, + "loss": 0.3469, + "step": 6214 + }, + { + "epoch": 0.1368038740920097, + "grad_norm": 1.6304741446546869, + "learning_rate": 9.70384453779974e-06, + "loss": 0.3511, + "step": 6215 + }, + { + "epoch": 0.13682588597842835, + "grad_norm": 2.0228365254962912, + "learning_rate": 9.703723669857574e-06, + "loss": 0.426, + "step": 6216 + }, + { + "epoch": 0.13684789786484702, + "grad_norm": 1.713689504839497, + "learning_rate": 9.703602778008986e-06, + "loss": 0.3582, + "step": 6217 + }, + { + "epoch": 0.13686990975126567, + "grad_norm": 2.567716043288672, + "learning_rate": 9.703481862254591e-06, + "loss": 0.379, + "step": 6218 + }, + { + "epoch": 0.13689192163768435, + "grad_norm": 1.2846816100132523, + "learning_rate": 9.703360922595005e-06, + "loss": 0.383, + "step": 6219 + }, + { + "epoch": 0.13691393352410303, + "grad_norm": 0.647059043035995, + "learning_rate": 9.703239959030839e-06, + "loss": 0.5539, + "step": 6220 + }, + { + "epoch": 0.13693594541052168, + "grad_norm": 1.323337883146388, + "learning_rate": 9.703118971562711e-06, + "loss": 0.3817, + "step": 6221 + }, + { + "epoch": 0.13695795729694035, + "grad_norm": 1.6868230865217235, + "learning_rate": 9.702997960191233e-06, + "loss": 0.518, + "step": 6222 + }, + { + "epoch": 0.136979969183359, + "grad_norm": 0.6459765499438992, + "learning_rate": 9.702876924917023e-06, + "loss": 0.5866, + "step": 6223 + }, + { + "epoch": 0.13700198106977768, + "grad_norm": 1.3984030032587251, + "learning_rate": 9.702755865740695e-06, + "loss": 0.3261, + "step": 6224 + }, + { + "epoch": 0.13702399295619636, + "grad_norm": 1.658742121489759, + "learning_rate": 9.702634782662864e-06, + "loss": 0.3592, + "step": 6225 + }, + { + "epoch": 0.137046004842615, + "grad_norm": 1.490969315650288, + "learning_rate": 9.702513675684145e-06, + "loss": 0.3881, + "step": 6226 + }, + { + "epoch": 0.13706801672903368, + "grad_norm": 1.349878790514218, + "learning_rate": 9.702392544805154e-06, + "loss": 0.2654, + "step": 6227 + }, + { + "epoch": 0.13709002861545233, + "grad_norm": 2.4030587933367125, + "learning_rate": 9.702271390026508e-06, + "loss": 0.3721, + "step": 6228 + }, + { + "epoch": 0.137112040501871, + "grad_norm": 1.6576075464170898, + "learning_rate": 9.702150211348822e-06, + "loss": 0.3706, + "step": 6229 + }, + { + "epoch": 0.13713405238828968, + "grad_norm": 1.4516145687747706, + "learning_rate": 9.702029008772709e-06, + "loss": 0.3571, + "step": 6230 + }, + { + "epoch": 0.13715606427470833, + "grad_norm": 1.5861991548751564, + "learning_rate": 9.70190778229879e-06, + "loss": 0.3823, + "step": 6231 + }, + { + "epoch": 0.137178076161127, + "grad_norm": 1.429856611493325, + "learning_rate": 9.701786531927676e-06, + "loss": 0.3565, + "step": 6232 + }, + { + "epoch": 0.1372000880475457, + "grad_norm": 0.6929624156459325, + "learning_rate": 9.701665257659987e-06, + "loss": 0.5813, + "step": 6233 + }, + { + "epoch": 0.13722209993396434, + "grad_norm": 1.4074299705803621, + "learning_rate": 9.701543959496339e-06, + "loss": 0.3745, + "step": 6234 + }, + { + "epoch": 0.137244111820383, + "grad_norm": 1.498897387018413, + "learning_rate": 9.701422637437345e-06, + "loss": 0.4745, + "step": 6235 + }, + { + "epoch": 0.13726612370680166, + "grad_norm": 0.6479342100960339, + "learning_rate": 9.701301291483626e-06, + "loss": 0.569, + "step": 6236 + }, + { + "epoch": 0.13728813559322034, + "grad_norm": 1.3680738005706126, + "learning_rate": 9.7011799216358e-06, + "loss": 0.3938, + "step": 6237 + }, + { + "epoch": 0.13731014747963902, + "grad_norm": 1.5877983028175506, + "learning_rate": 9.701058527894476e-06, + "loss": 0.3696, + "step": 6238 + }, + { + "epoch": 0.13733215936605767, + "grad_norm": 1.244721781612647, + "learning_rate": 9.700937110260278e-06, + "loss": 0.3598, + "step": 6239 + }, + { + "epoch": 0.13735417125247634, + "grad_norm": 2.0445014296777266, + "learning_rate": 9.700815668733821e-06, + "loss": 0.4137, + "step": 6240 + }, + { + "epoch": 0.137376183138895, + "grad_norm": 1.91692391339002, + "learning_rate": 9.700694203315721e-06, + "loss": 0.3285, + "step": 6241 + }, + { + "epoch": 0.13739819502531367, + "grad_norm": 1.437555295386401, + "learning_rate": 9.700572714006599e-06, + "loss": 0.344, + "step": 6242 + }, + { + "epoch": 0.13742020691173235, + "grad_norm": 2.296707906880865, + "learning_rate": 9.700451200807069e-06, + "loss": 0.326, + "step": 6243 + }, + { + "epoch": 0.137442218798151, + "grad_norm": 1.2229259512656059, + "learning_rate": 9.700329663717747e-06, + "loss": 0.3788, + "step": 6244 + }, + { + "epoch": 0.13746423068456967, + "grad_norm": 2.375189188284282, + "learning_rate": 9.700208102739256e-06, + "loss": 0.3766, + "step": 6245 + }, + { + "epoch": 0.13748624257098832, + "grad_norm": 1.373415370820493, + "learning_rate": 9.700086517872209e-06, + "loss": 0.2726, + "step": 6246 + }, + { + "epoch": 0.137508254457407, + "grad_norm": 2.539663310237124, + "learning_rate": 9.699964909117228e-06, + "loss": 0.2933, + "step": 6247 + }, + { + "epoch": 0.13753026634382567, + "grad_norm": 1.4547307038018407, + "learning_rate": 9.699843276474928e-06, + "loss": 0.2971, + "step": 6248 + }, + { + "epoch": 0.13755227823024432, + "grad_norm": 1.4631840148295203, + "learning_rate": 9.699721619945929e-06, + "loss": 0.3115, + "step": 6249 + }, + { + "epoch": 0.137574290116663, + "grad_norm": 1.5202394874184748, + "learning_rate": 9.699599939530848e-06, + "loss": 0.3883, + "step": 6250 + }, + { + "epoch": 0.13759630200308168, + "grad_norm": 2.1021420720492348, + "learning_rate": 9.699478235230305e-06, + "loss": 0.2479, + "step": 6251 + }, + { + "epoch": 0.13761831388950033, + "grad_norm": 2.3849453453691156, + "learning_rate": 9.699356507044915e-06, + "loss": 0.387, + "step": 6252 + }, + { + "epoch": 0.137640325775919, + "grad_norm": 1.5627127996316086, + "learning_rate": 9.699234754975302e-06, + "loss": 0.4322, + "step": 6253 + }, + { + "epoch": 0.13766233766233765, + "grad_norm": 1.404597777420919, + "learning_rate": 9.699112979022081e-06, + "loss": 0.3279, + "step": 6254 + }, + { + "epoch": 0.13768434954875633, + "grad_norm": 1.2955416163228637, + "learning_rate": 9.698991179185873e-06, + "loss": 0.3335, + "step": 6255 + }, + { + "epoch": 0.137706361435175, + "grad_norm": 5.456709021784786, + "learning_rate": 9.698869355467295e-06, + "loss": 0.3909, + "step": 6256 + }, + { + "epoch": 0.13772837332159366, + "grad_norm": 2.2367750473654584, + "learning_rate": 9.698747507866967e-06, + "loss": 0.2637, + "step": 6257 + }, + { + "epoch": 0.13775038520801233, + "grad_norm": 1.3604650708749917, + "learning_rate": 9.698625636385508e-06, + "loss": 0.3576, + "step": 6258 + }, + { + "epoch": 0.13777239709443098, + "grad_norm": 0.7192013828307271, + "learning_rate": 9.69850374102354e-06, + "loss": 0.5963, + "step": 6259 + }, + { + "epoch": 0.13779440898084966, + "grad_norm": 1.7883145287445754, + "learning_rate": 9.698381821781677e-06, + "loss": 0.4246, + "step": 6260 + }, + { + "epoch": 0.13781642086726834, + "grad_norm": 1.4266318945696506, + "learning_rate": 9.698259878660545e-06, + "loss": 0.2448, + "step": 6261 + }, + { + "epoch": 0.13783843275368698, + "grad_norm": 1.2493356144838819, + "learning_rate": 9.69813791166076e-06, + "loss": 0.321, + "step": 6262 + }, + { + "epoch": 0.13786044464010566, + "grad_norm": 1.7612006185256461, + "learning_rate": 9.698015920782943e-06, + "loss": 0.4174, + "step": 6263 + }, + { + "epoch": 0.1378824565265243, + "grad_norm": 1.2286319538512263, + "learning_rate": 9.697893906027713e-06, + "loss": 0.3139, + "step": 6264 + }, + { + "epoch": 0.137904468412943, + "grad_norm": 17.26141382179547, + "learning_rate": 9.697771867395692e-06, + "loss": 0.3982, + "step": 6265 + }, + { + "epoch": 0.13792648029936166, + "grad_norm": 0.6358145773716644, + "learning_rate": 9.697649804887497e-06, + "loss": 0.5463, + "step": 6266 + }, + { + "epoch": 0.1379484921857803, + "grad_norm": 1.9788271194880807, + "learning_rate": 9.697527718503752e-06, + "loss": 0.3259, + "step": 6267 + }, + { + "epoch": 0.137970504072199, + "grad_norm": 1.6103155002187863, + "learning_rate": 9.697405608245076e-06, + "loss": 0.4639, + "step": 6268 + }, + { + "epoch": 0.13799251595861764, + "grad_norm": 1.5069135389420842, + "learning_rate": 9.697283474112089e-06, + "loss": 0.3206, + "step": 6269 + }, + { + "epoch": 0.13801452784503632, + "grad_norm": 1.4231356054791902, + "learning_rate": 9.697161316105412e-06, + "loss": 0.4119, + "step": 6270 + }, + { + "epoch": 0.138036539731455, + "grad_norm": 2.185931178554237, + "learning_rate": 9.697039134225668e-06, + "loss": 0.4019, + "step": 6271 + }, + { + "epoch": 0.13805855161787364, + "grad_norm": 1.6450482395595394, + "learning_rate": 9.696916928473473e-06, + "loss": 0.3508, + "step": 6272 + }, + { + "epoch": 0.13808056350429232, + "grad_norm": 1.8505485492250562, + "learning_rate": 9.696794698849455e-06, + "loss": 0.3104, + "step": 6273 + }, + { + "epoch": 0.138102575390711, + "grad_norm": 2.5965195642970453, + "learning_rate": 9.69667244535423e-06, + "loss": 0.3853, + "step": 6274 + }, + { + "epoch": 0.13812458727712965, + "grad_norm": 3.1376610100806768, + "learning_rate": 9.69655016798842e-06, + "loss": 0.3115, + "step": 6275 + }, + { + "epoch": 0.13814659916354832, + "grad_norm": 1.4536066349149657, + "learning_rate": 9.69642786675265e-06, + "loss": 0.3068, + "step": 6276 + }, + { + "epoch": 0.13816861104996697, + "grad_norm": 1.706368746914262, + "learning_rate": 9.696305541647535e-06, + "loss": 0.4577, + "step": 6277 + }, + { + "epoch": 0.13819062293638565, + "grad_norm": 1.6057517537673562, + "learning_rate": 9.696183192673703e-06, + "loss": 0.3028, + "step": 6278 + }, + { + "epoch": 0.13821263482280433, + "grad_norm": 1.7491700449258658, + "learning_rate": 9.696060819831772e-06, + "loss": 0.3737, + "step": 6279 + }, + { + "epoch": 0.13823464670922297, + "grad_norm": 1.5016890159349674, + "learning_rate": 9.695938423122367e-06, + "loss": 0.3538, + "step": 6280 + }, + { + "epoch": 0.13825665859564165, + "grad_norm": 1.5071399509507057, + "learning_rate": 9.695816002546108e-06, + "loss": 0.392, + "step": 6281 + }, + { + "epoch": 0.1382786704820603, + "grad_norm": 1.464633571885256, + "learning_rate": 9.695693558103617e-06, + "loss": 0.3798, + "step": 6282 + }, + { + "epoch": 0.13830068236847898, + "grad_norm": 1.8763876548857596, + "learning_rate": 9.695571089795518e-06, + "loss": 0.348, + "step": 6283 + }, + { + "epoch": 0.13832269425489765, + "grad_norm": 1.3203257682564127, + "learning_rate": 9.695448597622432e-06, + "loss": 0.4217, + "step": 6284 + }, + { + "epoch": 0.1383447061413163, + "grad_norm": 1.5459544137826307, + "learning_rate": 9.695326081584982e-06, + "loss": 0.3789, + "step": 6285 + }, + { + "epoch": 0.13836671802773498, + "grad_norm": 1.733197818636718, + "learning_rate": 9.69520354168379e-06, + "loss": 0.3791, + "step": 6286 + }, + { + "epoch": 0.13838872991415363, + "grad_norm": 0.6890509452164055, + "learning_rate": 9.69508097791948e-06, + "loss": 0.5304, + "step": 6287 + }, + { + "epoch": 0.1384107418005723, + "grad_norm": 1.6725502888533508, + "learning_rate": 9.694958390292677e-06, + "loss": 0.3399, + "step": 6288 + }, + { + "epoch": 0.13843275368699098, + "grad_norm": 1.5246655031529976, + "learning_rate": 9.694835778803998e-06, + "loss": 0.364, + "step": 6289 + }, + { + "epoch": 0.13845476557340963, + "grad_norm": 4.602175686623926, + "learning_rate": 9.69471314345407e-06, + "loss": 0.3783, + "step": 6290 + }, + { + "epoch": 0.1384767774598283, + "grad_norm": 0.6655469681233755, + "learning_rate": 9.694590484243516e-06, + "loss": 0.5763, + "step": 6291 + }, + { + "epoch": 0.138498789346247, + "grad_norm": 1.600277091902184, + "learning_rate": 9.69446780117296e-06, + "loss": 0.4303, + "step": 6292 + }, + { + "epoch": 0.13852080123266564, + "grad_norm": 2.019534985704921, + "learning_rate": 9.694345094243025e-06, + "loss": 0.3876, + "step": 6293 + }, + { + "epoch": 0.1385428131190843, + "grad_norm": 1.2232940208002832, + "learning_rate": 9.694222363454336e-06, + "loss": 0.2768, + "step": 6294 + }, + { + "epoch": 0.13856482500550296, + "grad_norm": 1.3881876595304494, + "learning_rate": 9.694099608807512e-06, + "loss": 0.3759, + "step": 6295 + }, + { + "epoch": 0.13858683689192164, + "grad_norm": 0.6516506378560878, + "learning_rate": 9.693976830303183e-06, + "loss": 0.5418, + "step": 6296 + }, + { + "epoch": 0.13860884877834032, + "grad_norm": 1.531031515116001, + "learning_rate": 9.693854027941968e-06, + "loss": 0.3031, + "step": 6297 + }, + { + "epoch": 0.13863086066475896, + "grad_norm": 1.7178754296526275, + "learning_rate": 9.693731201724494e-06, + "loss": 0.4359, + "step": 6298 + }, + { + "epoch": 0.13865287255117764, + "grad_norm": 0.6497478255602964, + "learning_rate": 9.693608351651385e-06, + "loss": 0.5663, + "step": 6299 + }, + { + "epoch": 0.1386748844375963, + "grad_norm": 2.420766086694704, + "learning_rate": 9.693485477723264e-06, + "loss": 0.3536, + "step": 6300 + }, + { + "epoch": 0.13869689632401497, + "grad_norm": 1.7657953898646792, + "learning_rate": 9.693362579940757e-06, + "loss": 0.4154, + "step": 6301 + }, + { + "epoch": 0.13871890821043364, + "grad_norm": 1.5257196804073396, + "learning_rate": 9.693239658304488e-06, + "loss": 0.3467, + "step": 6302 + }, + { + "epoch": 0.1387409200968523, + "grad_norm": 1.5129204145729576, + "learning_rate": 9.693116712815082e-06, + "loss": 0.3174, + "step": 6303 + }, + { + "epoch": 0.13876293198327097, + "grad_norm": 4.372117996227521, + "learning_rate": 9.692993743473165e-06, + "loss": 0.3933, + "step": 6304 + }, + { + "epoch": 0.13878494386968962, + "grad_norm": 1.2560387012466923, + "learning_rate": 9.692870750279358e-06, + "loss": 0.2497, + "step": 6305 + }, + { + "epoch": 0.1388069557561083, + "grad_norm": 1.7054466144696963, + "learning_rate": 9.69274773323429e-06, + "loss": 0.3291, + "step": 6306 + }, + { + "epoch": 0.13882896764252697, + "grad_norm": 1.5584045092793135, + "learning_rate": 9.692624692338587e-06, + "loss": 0.2657, + "step": 6307 + }, + { + "epoch": 0.13885097952894562, + "grad_norm": 1.7512847885875071, + "learning_rate": 9.692501627592869e-06, + "loss": 0.3634, + "step": 6308 + }, + { + "epoch": 0.1388729914153643, + "grad_norm": 2.0850493014272553, + "learning_rate": 9.692378538997766e-06, + "loss": 0.4, + "step": 6309 + }, + { + "epoch": 0.13889500330178298, + "grad_norm": 1.3935621408162462, + "learning_rate": 9.692255426553902e-06, + "loss": 0.3461, + "step": 6310 + }, + { + "epoch": 0.13891701518820163, + "grad_norm": 1.6715968129364467, + "learning_rate": 9.692132290261903e-06, + "loss": 0.3974, + "step": 6311 + }, + { + "epoch": 0.1389390270746203, + "grad_norm": 1.3894311972941895, + "learning_rate": 9.692009130122397e-06, + "loss": 0.2897, + "step": 6312 + }, + { + "epoch": 0.13896103896103895, + "grad_norm": 1.5865196429384405, + "learning_rate": 9.691885946136007e-06, + "loss": 0.3788, + "step": 6313 + }, + { + "epoch": 0.13898305084745763, + "grad_norm": 1.4266819448356365, + "learning_rate": 9.69176273830336e-06, + "loss": 0.3208, + "step": 6314 + }, + { + "epoch": 0.1390050627338763, + "grad_norm": 1.767514708368969, + "learning_rate": 9.691639506625082e-06, + "loss": 0.4057, + "step": 6315 + }, + { + "epoch": 0.13902707462029495, + "grad_norm": 1.6291647622329932, + "learning_rate": 9.691516251101798e-06, + "loss": 0.3483, + "step": 6316 + }, + { + "epoch": 0.13904908650671363, + "grad_norm": 1.6589502259473463, + "learning_rate": 9.691392971734139e-06, + "loss": 0.3634, + "step": 6317 + }, + { + "epoch": 0.13907109839313228, + "grad_norm": 1.4127259740141707, + "learning_rate": 9.691269668522726e-06, + "loss": 0.3168, + "step": 6318 + }, + { + "epoch": 0.13909311027955096, + "grad_norm": 1.3757830488035676, + "learning_rate": 9.691146341468189e-06, + "loss": 0.3676, + "step": 6319 + }, + { + "epoch": 0.13911512216596963, + "grad_norm": 1.6690818731068073, + "learning_rate": 9.691022990571153e-06, + "loss": 0.3781, + "step": 6320 + }, + { + "epoch": 0.13913713405238828, + "grad_norm": 3.0735227959149416, + "learning_rate": 9.690899615832248e-06, + "loss": 0.307, + "step": 6321 + }, + { + "epoch": 0.13915914593880696, + "grad_norm": 1.9395315043932748, + "learning_rate": 9.690776217252097e-06, + "loss": 0.4005, + "step": 6322 + }, + { + "epoch": 0.1391811578252256, + "grad_norm": 1.3529036569882684, + "learning_rate": 9.69065279483133e-06, + "loss": 0.3484, + "step": 6323 + }, + { + "epoch": 0.1392031697116443, + "grad_norm": 1.6809440043573571, + "learning_rate": 9.690529348570572e-06, + "loss": 0.4107, + "step": 6324 + }, + { + "epoch": 0.13922518159806296, + "grad_norm": 0.7336638742932117, + "learning_rate": 9.690405878470455e-06, + "loss": 0.5771, + "step": 6325 + }, + { + "epoch": 0.1392471934844816, + "grad_norm": 1.8521773735507683, + "learning_rate": 9.690282384531601e-06, + "loss": 0.3535, + "step": 6326 + }, + { + "epoch": 0.1392692053709003, + "grad_norm": 8.60327151324875, + "learning_rate": 9.69015886675464e-06, + "loss": 0.3885, + "step": 6327 + }, + { + "epoch": 0.13929121725731894, + "grad_norm": 1.6790182036102235, + "learning_rate": 9.6900353251402e-06, + "loss": 0.4314, + "step": 6328 + }, + { + "epoch": 0.13931322914373762, + "grad_norm": 1.7560239038910495, + "learning_rate": 9.689911759688909e-06, + "loss": 0.4118, + "step": 6329 + }, + { + "epoch": 0.1393352410301563, + "grad_norm": 1.6803074678300298, + "learning_rate": 9.689788170401395e-06, + "loss": 0.3795, + "step": 6330 + }, + { + "epoch": 0.13935725291657494, + "grad_norm": 3.654251500107717, + "learning_rate": 9.689664557278285e-06, + "loss": 0.3962, + "step": 6331 + }, + { + "epoch": 0.13937926480299362, + "grad_norm": 1.4992415391729925, + "learning_rate": 9.689540920320208e-06, + "loss": 0.3628, + "step": 6332 + }, + { + "epoch": 0.1394012766894123, + "grad_norm": 2.280134615524777, + "learning_rate": 9.689417259527794e-06, + "loss": 0.3981, + "step": 6333 + }, + { + "epoch": 0.13942328857583094, + "grad_norm": 1.3717128212505596, + "learning_rate": 9.689293574901668e-06, + "loss": 0.4195, + "step": 6334 + }, + { + "epoch": 0.13944530046224962, + "grad_norm": 2.217934259452451, + "learning_rate": 9.689169866442462e-06, + "loss": 0.3246, + "step": 6335 + }, + { + "epoch": 0.13946731234866827, + "grad_norm": 1.5651468758360745, + "learning_rate": 9.689046134150802e-06, + "loss": 0.374, + "step": 6336 + }, + { + "epoch": 0.13948932423508695, + "grad_norm": 0.7076605796958761, + "learning_rate": 9.68892237802732e-06, + "loss": 0.5627, + "step": 6337 + }, + { + "epoch": 0.13951133612150562, + "grad_norm": 2.2335002179873884, + "learning_rate": 9.688798598072641e-06, + "loss": 0.4494, + "step": 6338 + }, + { + "epoch": 0.13953334800792427, + "grad_norm": 2.2556051204864866, + "learning_rate": 9.688674794287399e-06, + "loss": 0.3847, + "step": 6339 + }, + { + "epoch": 0.13955535989434295, + "grad_norm": 1.5166383143207938, + "learning_rate": 9.688550966672216e-06, + "loss": 0.3267, + "step": 6340 + }, + { + "epoch": 0.1395773717807616, + "grad_norm": 1.354806648374436, + "learning_rate": 9.68842711522773e-06, + "loss": 0.3387, + "step": 6341 + }, + { + "epoch": 0.13959938366718028, + "grad_norm": 1.7556222562535482, + "learning_rate": 9.688303239954564e-06, + "loss": 0.4244, + "step": 6342 + }, + { + "epoch": 0.13962139555359895, + "grad_norm": 1.7040206310716977, + "learning_rate": 9.688179340853352e-06, + "loss": 0.3418, + "step": 6343 + }, + { + "epoch": 0.1396434074400176, + "grad_norm": 1.749451703274177, + "learning_rate": 9.68805541792472e-06, + "loss": 0.3472, + "step": 6344 + }, + { + "epoch": 0.13966541932643628, + "grad_norm": 1.3211379535190022, + "learning_rate": 9.6879314711693e-06, + "loss": 0.4344, + "step": 6345 + }, + { + "epoch": 0.13968743121285493, + "grad_norm": 1.667827600079329, + "learning_rate": 9.68780750058772e-06, + "loss": 0.4182, + "step": 6346 + }, + { + "epoch": 0.1397094430992736, + "grad_norm": 2.2062521112163376, + "learning_rate": 9.687683506180613e-06, + "loss": 0.3177, + "step": 6347 + }, + { + "epoch": 0.13973145498569228, + "grad_norm": 1.793348092235633, + "learning_rate": 9.687559487948608e-06, + "loss": 0.3615, + "step": 6348 + }, + { + "epoch": 0.13975346687211093, + "grad_norm": 0.6665505580379303, + "learning_rate": 9.687435445892335e-06, + "loss": 0.5851, + "step": 6349 + }, + { + "epoch": 0.1397754787585296, + "grad_norm": 1.8463229173402966, + "learning_rate": 9.687311380012422e-06, + "loss": 0.29, + "step": 6350 + }, + { + "epoch": 0.13979749064494829, + "grad_norm": 0.6259174737204576, + "learning_rate": 9.687187290309505e-06, + "loss": 0.5397, + "step": 6351 + }, + { + "epoch": 0.13981950253136693, + "grad_norm": 1.4006744157429256, + "learning_rate": 9.68706317678421e-06, + "loss": 0.3668, + "step": 6352 + }, + { + "epoch": 0.1398415144177856, + "grad_norm": 1.3644780097814457, + "learning_rate": 9.68693903943717e-06, + "loss": 0.4523, + "step": 6353 + }, + { + "epoch": 0.13986352630420426, + "grad_norm": 1.5740261978035612, + "learning_rate": 9.686814878269015e-06, + "loss": 0.298, + "step": 6354 + }, + { + "epoch": 0.13988553819062294, + "grad_norm": 4.750873432792325, + "learning_rate": 9.686690693280376e-06, + "loss": 0.3031, + "step": 6355 + }, + { + "epoch": 0.13990755007704161, + "grad_norm": 1.5335560581083874, + "learning_rate": 9.686566484471887e-06, + "loss": 0.3345, + "step": 6356 + }, + { + "epoch": 0.13992956196346026, + "grad_norm": 1.548473500416518, + "learning_rate": 9.686442251844175e-06, + "loss": 0.3879, + "step": 6357 + }, + { + "epoch": 0.13995157384987894, + "grad_norm": 2.1212492315768774, + "learning_rate": 9.686317995397872e-06, + "loss": 0.3708, + "step": 6358 + }, + { + "epoch": 0.1399735857362976, + "grad_norm": 2.5679252002242956, + "learning_rate": 9.686193715133612e-06, + "loss": 0.3312, + "step": 6359 + }, + { + "epoch": 0.13999559762271627, + "grad_norm": 1.9817236727677057, + "learning_rate": 9.686069411052025e-06, + "loss": 0.4094, + "step": 6360 + }, + { + "epoch": 0.14001760950913494, + "grad_norm": 1.391691570941263, + "learning_rate": 9.685945083153744e-06, + "loss": 0.3791, + "step": 6361 + }, + { + "epoch": 0.1400396213955536, + "grad_norm": 1.7730842575837398, + "learning_rate": 9.6858207314394e-06, + "loss": 0.3675, + "step": 6362 + }, + { + "epoch": 0.14006163328197227, + "grad_norm": 2.5024193478745516, + "learning_rate": 9.685696355909624e-06, + "loss": 0.3268, + "step": 6363 + }, + { + "epoch": 0.14008364516839092, + "grad_norm": 1.4808743290057125, + "learning_rate": 9.685571956565053e-06, + "loss": 0.2998, + "step": 6364 + }, + { + "epoch": 0.1401056570548096, + "grad_norm": 0.7664314545469505, + "learning_rate": 9.68544753340631e-06, + "loss": 0.5555, + "step": 6365 + }, + { + "epoch": 0.14012766894122827, + "grad_norm": 1.7518632479071683, + "learning_rate": 9.685323086434037e-06, + "loss": 0.4276, + "step": 6366 + }, + { + "epoch": 0.14014968082764692, + "grad_norm": 1.6223835495504664, + "learning_rate": 9.68519861564886e-06, + "loss": 0.3425, + "step": 6367 + }, + { + "epoch": 0.1401716927140656, + "grad_norm": 2.920542201031232, + "learning_rate": 9.685074121051415e-06, + "loss": 0.3501, + "step": 6368 + }, + { + "epoch": 0.14019370460048425, + "grad_norm": 3.1272343272609016, + "learning_rate": 9.684949602642334e-06, + "loss": 0.376, + "step": 6369 + }, + { + "epoch": 0.14021571648690292, + "grad_norm": 1.4363726450839511, + "learning_rate": 9.68482506042225e-06, + "loss": 0.3791, + "step": 6370 + }, + { + "epoch": 0.1402377283733216, + "grad_norm": 1.6207195996977375, + "learning_rate": 9.684700494391795e-06, + "loss": 0.4678, + "step": 6371 + }, + { + "epoch": 0.14025974025974025, + "grad_norm": 0.6615480003847936, + "learning_rate": 9.6845759045516e-06, + "loss": 0.5619, + "step": 6372 + }, + { + "epoch": 0.14028175214615893, + "grad_norm": 0.6548178235272819, + "learning_rate": 9.684451290902303e-06, + "loss": 0.5728, + "step": 6373 + }, + { + "epoch": 0.1403037640325776, + "grad_norm": 3.304695303087279, + "learning_rate": 9.684326653444537e-06, + "loss": 0.3144, + "step": 6374 + }, + { + "epoch": 0.14032577591899625, + "grad_norm": 0.6230518104433688, + "learning_rate": 9.684201992178931e-06, + "loss": 0.539, + "step": 6375 + }, + { + "epoch": 0.14034778780541493, + "grad_norm": 1.5118871645800733, + "learning_rate": 9.684077307106123e-06, + "loss": 0.4253, + "step": 6376 + }, + { + "epoch": 0.14036979969183358, + "grad_norm": 1.8565199738262277, + "learning_rate": 9.683952598226744e-06, + "loss": 0.3294, + "step": 6377 + }, + { + "epoch": 0.14039181157825226, + "grad_norm": 1.5105079203396041, + "learning_rate": 9.68382786554143e-06, + "loss": 0.4337, + "step": 6378 + }, + { + "epoch": 0.14041382346467093, + "grad_norm": 1.7574161076506696, + "learning_rate": 9.683703109050813e-06, + "loss": 0.3294, + "step": 6379 + }, + { + "epoch": 0.14043583535108958, + "grad_norm": 1.8199914791414142, + "learning_rate": 9.683578328755528e-06, + "loss": 0.3727, + "step": 6380 + }, + { + "epoch": 0.14045784723750826, + "grad_norm": 1.3803456333070974, + "learning_rate": 9.683453524656208e-06, + "loss": 0.3174, + "step": 6381 + }, + { + "epoch": 0.1404798591239269, + "grad_norm": 2.1595078449127394, + "learning_rate": 9.68332869675349e-06, + "loss": 0.4038, + "step": 6382 + }, + { + "epoch": 0.14050187101034559, + "grad_norm": 2.2649255687712824, + "learning_rate": 9.683203845048004e-06, + "loss": 0.3433, + "step": 6383 + }, + { + "epoch": 0.14052388289676426, + "grad_norm": 1.1897575032054426, + "learning_rate": 9.68307896954039e-06, + "loss": 0.3486, + "step": 6384 + }, + { + "epoch": 0.1405458947831829, + "grad_norm": 1.768264590252483, + "learning_rate": 9.682954070231279e-06, + "loss": 0.3709, + "step": 6385 + }, + { + "epoch": 0.1405679066696016, + "grad_norm": 1.3810140314984618, + "learning_rate": 9.682829147121306e-06, + "loss": 0.293, + "step": 6386 + }, + { + "epoch": 0.14058991855602024, + "grad_norm": 1.2295623377618872, + "learning_rate": 9.682704200211108e-06, + "loss": 0.3256, + "step": 6387 + }, + { + "epoch": 0.14061193044243891, + "grad_norm": 0.9267716685372447, + "learning_rate": 9.68257922950132e-06, + "loss": 0.5538, + "step": 6388 + }, + { + "epoch": 0.1406339423288576, + "grad_norm": 4.582653520688142, + "learning_rate": 9.682454234992573e-06, + "loss": 0.3933, + "step": 6389 + }, + { + "epoch": 0.14065595421527624, + "grad_norm": 2.208805040069444, + "learning_rate": 9.682329216685505e-06, + "loss": 0.3595, + "step": 6390 + }, + { + "epoch": 0.14067796610169492, + "grad_norm": 1.5187797486112646, + "learning_rate": 9.682204174580754e-06, + "loss": 0.3706, + "step": 6391 + }, + { + "epoch": 0.1406999779881136, + "grad_norm": 2.058379635471691, + "learning_rate": 9.682079108678951e-06, + "loss": 0.3409, + "step": 6392 + }, + { + "epoch": 0.14072198987453224, + "grad_norm": 1.7413477861568358, + "learning_rate": 9.681954018980735e-06, + "loss": 0.3952, + "step": 6393 + }, + { + "epoch": 0.14074400176095092, + "grad_norm": 1.3474503675548597, + "learning_rate": 9.68182890548674e-06, + "loss": 0.3779, + "step": 6394 + }, + { + "epoch": 0.14076601364736957, + "grad_norm": 1.3937563561982687, + "learning_rate": 9.681703768197601e-06, + "loss": 0.3018, + "step": 6395 + }, + { + "epoch": 0.14078802553378825, + "grad_norm": 2.673232754210146, + "learning_rate": 9.681578607113957e-06, + "loss": 0.3144, + "step": 6396 + }, + { + "epoch": 0.14081003742020692, + "grad_norm": 1.5215193604959503, + "learning_rate": 9.681453422236443e-06, + "loss": 0.3631, + "step": 6397 + }, + { + "epoch": 0.14083204930662557, + "grad_norm": 1.5338503730582256, + "learning_rate": 9.681328213565693e-06, + "loss": 0.3166, + "step": 6398 + }, + { + "epoch": 0.14085406119304425, + "grad_norm": 1.416833880368768, + "learning_rate": 9.681202981102347e-06, + "loss": 0.3458, + "step": 6399 + }, + { + "epoch": 0.1408760730794629, + "grad_norm": 1.8977820427019498, + "learning_rate": 9.681077724847038e-06, + "loss": 0.3374, + "step": 6400 + }, + { + "epoch": 0.14089808496588158, + "grad_norm": 3.8232295564127283, + "learning_rate": 9.680952444800404e-06, + "loss": 0.4572, + "step": 6401 + }, + { + "epoch": 0.14092009685230025, + "grad_norm": 1.721246855092868, + "learning_rate": 9.680827140963084e-06, + "loss": 0.349, + "step": 6402 + }, + { + "epoch": 0.1409421087387189, + "grad_norm": 0.9931232468472887, + "learning_rate": 9.680701813335711e-06, + "loss": 0.5893, + "step": 6403 + }, + { + "epoch": 0.14096412062513758, + "grad_norm": 1.4555601834755274, + "learning_rate": 9.680576461918924e-06, + "loss": 0.2536, + "step": 6404 + }, + { + "epoch": 0.14098613251155623, + "grad_norm": 1.8260840608933855, + "learning_rate": 9.680451086713359e-06, + "loss": 0.2863, + "step": 6405 + }, + { + "epoch": 0.1410081443979749, + "grad_norm": 1.5136802707106873, + "learning_rate": 9.680325687719656e-06, + "loss": 0.4122, + "step": 6406 + }, + { + "epoch": 0.14103015628439358, + "grad_norm": 2.866975322142896, + "learning_rate": 9.68020026493845e-06, + "loss": 0.4024, + "step": 6407 + }, + { + "epoch": 0.14105216817081223, + "grad_norm": 4.9846289939127, + "learning_rate": 9.680074818370377e-06, + "loss": 0.3533, + "step": 6408 + }, + { + "epoch": 0.1410741800572309, + "grad_norm": 0.6487565784879291, + "learning_rate": 9.679949348016078e-06, + "loss": 0.5556, + "step": 6409 + }, + { + "epoch": 0.14109619194364956, + "grad_norm": 4.402004851368414, + "learning_rate": 9.679823853876189e-06, + "loss": 0.3826, + "step": 6410 + }, + { + "epoch": 0.14111820383006823, + "grad_norm": 2.378134638103349, + "learning_rate": 9.679698335951347e-06, + "loss": 0.3692, + "step": 6411 + }, + { + "epoch": 0.1411402157164869, + "grad_norm": 5.277521720614087, + "learning_rate": 9.679572794242191e-06, + "loss": 0.3945, + "step": 6412 + }, + { + "epoch": 0.14116222760290556, + "grad_norm": 0.7375639544329915, + "learning_rate": 9.67944722874936e-06, + "loss": 0.587, + "step": 6413 + }, + { + "epoch": 0.14118423948932424, + "grad_norm": 12.664238090210887, + "learning_rate": 9.67932163947349e-06, + "loss": 0.4177, + "step": 6414 + }, + { + "epoch": 0.1412062513757429, + "grad_norm": 8.983765289414139, + "learning_rate": 9.67919602641522e-06, + "loss": 0.4404, + "step": 6415 + }, + { + "epoch": 0.14122826326216156, + "grad_norm": 3.047977174432533, + "learning_rate": 9.67907038957519e-06, + "loss": 0.4278, + "step": 6416 + }, + { + "epoch": 0.14125027514858024, + "grad_norm": 6.050239006375301, + "learning_rate": 9.678944728954037e-06, + "loss": 0.5009, + "step": 6417 + }, + { + "epoch": 0.1412722870349989, + "grad_norm": 2.064904700483039, + "learning_rate": 9.678819044552401e-06, + "loss": 0.3193, + "step": 6418 + }, + { + "epoch": 0.14129429892141757, + "grad_norm": 1.3678538947271741, + "learning_rate": 9.678693336370917e-06, + "loss": 0.3116, + "step": 6419 + }, + { + "epoch": 0.14131631080783624, + "grad_norm": 1.7293637820254903, + "learning_rate": 9.67856760441023e-06, + "loss": 0.363, + "step": 6420 + }, + { + "epoch": 0.1413383226942549, + "grad_norm": 2.4228707408937735, + "learning_rate": 9.678441848670974e-06, + "loss": 0.4369, + "step": 6421 + }, + { + "epoch": 0.14136033458067357, + "grad_norm": 1.383339964217631, + "learning_rate": 9.67831606915379e-06, + "loss": 0.2773, + "step": 6422 + }, + { + "epoch": 0.14138234646709222, + "grad_norm": 2.04362639543305, + "learning_rate": 9.678190265859318e-06, + "loss": 0.3403, + "step": 6423 + }, + { + "epoch": 0.1414043583535109, + "grad_norm": 2.0712357423218704, + "learning_rate": 9.678064438788194e-06, + "loss": 0.2703, + "step": 6424 + }, + { + "epoch": 0.14142637023992957, + "grad_norm": 1.2472679214873201, + "learning_rate": 9.677938587941063e-06, + "loss": 0.367, + "step": 6425 + }, + { + "epoch": 0.14144838212634822, + "grad_norm": 1.364991952916733, + "learning_rate": 9.677812713318561e-06, + "loss": 0.3621, + "step": 6426 + }, + { + "epoch": 0.1414703940127669, + "grad_norm": 1.404794759581008, + "learning_rate": 9.677686814921327e-06, + "loss": 0.3575, + "step": 6427 + }, + { + "epoch": 0.14149240589918555, + "grad_norm": 3.540783217962288, + "learning_rate": 9.677560892750005e-06, + "loss": 0.366, + "step": 6428 + }, + { + "epoch": 0.14151441778560422, + "grad_norm": 2.17403825071494, + "learning_rate": 9.677434946805229e-06, + "loss": 0.3531, + "step": 6429 + }, + { + "epoch": 0.1415364296720229, + "grad_norm": 2.0513760441996354, + "learning_rate": 9.677308977087645e-06, + "loss": 0.4297, + "step": 6430 + }, + { + "epoch": 0.14155844155844155, + "grad_norm": 1.8392540545975187, + "learning_rate": 9.677182983597888e-06, + "loss": 0.4054, + "step": 6431 + }, + { + "epoch": 0.14158045344486023, + "grad_norm": 4.3919920035851385, + "learning_rate": 9.677056966336605e-06, + "loss": 0.4343, + "step": 6432 + }, + { + "epoch": 0.1416024653312789, + "grad_norm": 3.80420924737039, + "learning_rate": 9.676930925304429e-06, + "loss": 0.38, + "step": 6433 + }, + { + "epoch": 0.14162447721769755, + "grad_norm": 1.628179759825793, + "learning_rate": 9.676804860502006e-06, + "loss": 0.3368, + "step": 6434 + }, + { + "epoch": 0.14164648910411623, + "grad_norm": 1.657854552047005, + "learning_rate": 9.676678771929973e-06, + "loss": 0.4049, + "step": 6435 + }, + { + "epoch": 0.14166850099053488, + "grad_norm": 2.038436273497819, + "learning_rate": 9.676552659588974e-06, + "loss": 0.4053, + "step": 6436 + }, + { + "epoch": 0.14169051287695356, + "grad_norm": 1.7205510006037494, + "learning_rate": 9.676426523479647e-06, + "loss": 0.3885, + "step": 6437 + }, + { + "epoch": 0.14171252476337223, + "grad_norm": 0.670188578670234, + "learning_rate": 9.676300363602636e-06, + "loss": 0.5323, + "step": 6438 + }, + { + "epoch": 0.14173453664979088, + "grad_norm": 1.4907243513676876, + "learning_rate": 9.676174179958582e-06, + "loss": 0.3106, + "step": 6439 + }, + { + "epoch": 0.14175654853620956, + "grad_norm": 1.5686945051843106, + "learning_rate": 9.676047972548123e-06, + "loss": 0.3737, + "step": 6440 + }, + { + "epoch": 0.1417785604226282, + "grad_norm": 2.124843258103795, + "learning_rate": 9.675921741371903e-06, + "loss": 0.4563, + "step": 6441 + }, + { + "epoch": 0.14180057230904689, + "grad_norm": 1.6846663854548758, + "learning_rate": 9.675795486430563e-06, + "loss": 0.3323, + "step": 6442 + }, + { + "epoch": 0.14182258419546556, + "grad_norm": 2.113479691136304, + "learning_rate": 9.675669207724747e-06, + "loss": 0.285, + "step": 6443 + }, + { + "epoch": 0.1418445960818842, + "grad_norm": 0.6512223321433322, + "learning_rate": 9.675542905255092e-06, + "loss": 0.5569, + "step": 6444 + }, + { + "epoch": 0.1418666079683029, + "grad_norm": 3.4941858377861297, + "learning_rate": 9.675416579022244e-06, + "loss": 0.2829, + "step": 6445 + }, + { + "epoch": 0.14188861985472154, + "grad_norm": 2.6312161084580628, + "learning_rate": 9.675290229026843e-06, + "loss": 0.3061, + "step": 6446 + }, + { + "epoch": 0.14191063174114021, + "grad_norm": 1.7097204246664972, + "learning_rate": 9.675163855269533e-06, + "loss": 0.411, + "step": 6447 + }, + { + "epoch": 0.1419326436275589, + "grad_norm": 2.3876226972535615, + "learning_rate": 9.675037457750956e-06, + "loss": 0.3952, + "step": 6448 + }, + { + "epoch": 0.14195465551397754, + "grad_norm": 18.50985751186438, + "learning_rate": 9.674911036471751e-06, + "loss": 0.3267, + "step": 6449 + }, + { + "epoch": 0.14197666740039622, + "grad_norm": 0.6149776753624883, + "learning_rate": 9.674784591432563e-06, + "loss": 0.5172, + "step": 6450 + }, + { + "epoch": 0.14199867928681487, + "grad_norm": 1.631143901991113, + "learning_rate": 9.674658122634036e-06, + "loss": 0.3523, + "step": 6451 + }, + { + "epoch": 0.14202069117323354, + "grad_norm": 1.895038421978862, + "learning_rate": 9.674531630076812e-06, + "loss": 0.3417, + "step": 6452 + }, + { + "epoch": 0.14204270305965222, + "grad_norm": 1.5566381152698452, + "learning_rate": 9.674405113761531e-06, + "loss": 0.2931, + "step": 6453 + }, + { + "epoch": 0.14206471494607087, + "grad_norm": 1.7741577939231006, + "learning_rate": 9.67427857368884e-06, + "loss": 0.3468, + "step": 6454 + }, + { + "epoch": 0.14208672683248955, + "grad_norm": 1.6428068104235816, + "learning_rate": 9.67415200985938e-06, + "loss": 0.3208, + "step": 6455 + }, + { + "epoch": 0.14210873871890822, + "grad_norm": 0.6552463875864076, + "learning_rate": 9.674025422273796e-06, + "loss": 0.527, + "step": 6456 + }, + { + "epoch": 0.14213075060532687, + "grad_norm": 1.8908716638937257, + "learning_rate": 9.673898810932728e-06, + "loss": 0.3193, + "step": 6457 + }, + { + "epoch": 0.14215276249174555, + "grad_norm": 1.5419988767029558, + "learning_rate": 9.673772175836824e-06, + "loss": 0.3559, + "step": 6458 + }, + { + "epoch": 0.1421747743781642, + "grad_norm": 1.8007944675600749, + "learning_rate": 9.673645516986725e-06, + "loss": 0.3774, + "step": 6459 + }, + { + "epoch": 0.14219678626458288, + "grad_norm": 2.4227020626283755, + "learning_rate": 9.673518834383073e-06, + "loss": 0.2455, + "step": 6460 + }, + { + "epoch": 0.14221879815100155, + "grad_norm": 2.0561321885090478, + "learning_rate": 9.673392128026515e-06, + "loss": 0.3328, + "step": 6461 + }, + { + "epoch": 0.1422408100374202, + "grad_norm": 1.4700942344432013, + "learning_rate": 9.673265397917693e-06, + "loss": 0.403, + "step": 6462 + }, + { + "epoch": 0.14226282192383888, + "grad_norm": 2.4195061289885276, + "learning_rate": 9.673138644057254e-06, + "loss": 0.319, + "step": 6463 + }, + { + "epoch": 0.14228483381025753, + "grad_norm": 1.651211802961559, + "learning_rate": 9.673011866445838e-06, + "loss": 0.305, + "step": 6464 + }, + { + "epoch": 0.1423068456966762, + "grad_norm": 2.1009782372420145, + "learning_rate": 9.672885065084091e-06, + "loss": 0.2949, + "step": 6465 + }, + { + "epoch": 0.14232885758309488, + "grad_norm": 14.031312348323254, + "learning_rate": 9.672758239972662e-06, + "loss": 0.3529, + "step": 6466 + }, + { + "epoch": 0.14235086946951353, + "grad_norm": 0.6606331659820655, + "learning_rate": 9.672631391112189e-06, + "loss": 0.5362, + "step": 6467 + }, + { + "epoch": 0.1423728813559322, + "grad_norm": 0.6730699147439657, + "learning_rate": 9.672504518503318e-06, + "loss": 0.5535, + "step": 6468 + }, + { + "epoch": 0.14239489324235086, + "grad_norm": 2.0221050728404144, + "learning_rate": 9.672377622146695e-06, + "loss": 0.352, + "step": 6469 + }, + { + "epoch": 0.14241690512876953, + "grad_norm": 1.298352809401153, + "learning_rate": 9.672250702042967e-06, + "loss": 0.309, + "step": 6470 + }, + { + "epoch": 0.1424389170151882, + "grad_norm": 1.2272893085517347, + "learning_rate": 9.672123758192775e-06, + "loss": 0.3745, + "step": 6471 + }, + { + "epoch": 0.14246092890160686, + "grad_norm": 1.7229748594347407, + "learning_rate": 9.671996790596767e-06, + "loss": 0.4704, + "step": 6472 + }, + { + "epoch": 0.14248294078802554, + "grad_norm": 1.5920687565286546, + "learning_rate": 9.67186979925559e-06, + "loss": 0.4308, + "step": 6473 + }, + { + "epoch": 0.1425049526744442, + "grad_norm": 1.5897451089467007, + "learning_rate": 9.671742784169885e-06, + "loss": 0.4148, + "step": 6474 + }, + { + "epoch": 0.14252696456086286, + "grad_norm": 2.6409826641332304, + "learning_rate": 9.671615745340298e-06, + "loss": 0.3368, + "step": 6475 + }, + { + "epoch": 0.14254897644728154, + "grad_norm": 1.7465968518372401, + "learning_rate": 9.671488682767478e-06, + "loss": 0.3772, + "step": 6476 + }, + { + "epoch": 0.1425709883337002, + "grad_norm": 0.7261524966003384, + "learning_rate": 9.671361596452067e-06, + "loss": 0.5202, + "step": 6477 + }, + { + "epoch": 0.14259300022011887, + "grad_norm": 1.3233668579333953, + "learning_rate": 9.671234486394715e-06, + "loss": 0.3512, + "step": 6478 + }, + { + "epoch": 0.14261501210653754, + "grad_norm": 1.423551507950562, + "learning_rate": 9.671107352596066e-06, + "loss": 0.4006, + "step": 6479 + }, + { + "epoch": 0.1426370239929562, + "grad_norm": 5.3825630632892345, + "learning_rate": 9.670980195056764e-06, + "loss": 0.3489, + "step": 6480 + }, + { + "epoch": 0.14265903587937487, + "grad_norm": 0.6517691721851606, + "learning_rate": 9.670853013777459e-06, + "loss": 0.5358, + "step": 6481 + }, + { + "epoch": 0.14268104776579352, + "grad_norm": 1.7257095160596323, + "learning_rate": 9.670725808758795e-06, + "loss": 0.391, + "step": 6482 + }, + { + "epoch": 0.1427030596522122, + "grad_norm": 1.671647405519596, + "learning_rate": 9.67059858000142e-06, + "loss": 0.4334, + "step": 6483 + }, + { + "epoch": 0.14272507153863087, + "grad_norm": 1.6251762791772821, + "learning_rate": 9.67047132750598e-06, + "loss": 0.3814, + "step": 6484 + }, + { + "epoch": 0.14274708342504952, + "grad_norm": 1.6056458391325963, + "learning_rate": 9.67034405127312e-06, + "loss": 0.3839, + "step": 6485 + }, + { + "epoch": 0.1427690953114682, + "grad_norm": 1.7607393309342043, + "learning_rate": 9.670216751303489e-06, + "loss": 0.3099, + "step": 6486 + }, + { + "epoch": 0.14279110719788685, + "grad_norm": 1.2609453871102672, + "learning_rate": 9.670089427597735e-06, + "loss": 0.3863, + "step": 6487 + }, + { + "epoch": 0.14281311908430552, + "grad_norm": 0.7018189738595606, + "learning_rate": 9.669962080156503e-06, + "loss": 0.5293, + "step": 6488 + }, + { + "epoch": 0.1428351309707242, + "grad_norm": 6.710167120469957, + "learning_rate": 9.669834708980441e-06, + "loss": 0.3397, + "step": 6489 + }, + { + "epoch": 0.14285714285714285, + "grad_norm": 1.731833937496662, + "learning_rate": 9.669707314070195e-06, + "loss": 0.3241, + "step": 6490 + }, + { + "epoch": 0.14287915474356153, + "grad_norm": 0.6615304464261866, + "learning_rate": 9.669579895426413e-06, + "loss": 0.5555, + "step": 6491 + }, + { + "epoch": 0.1429011666299802, + "grad_norm": 0.6404635003301039, + "learning_rate": 9.669452453049746e-06, + "loss": 0.5617, + "step": 6492 + }, + { + "epoch": 0.14292317851639885, + "grad_norm": 1.8929062035897417, + "learning_rate": 9.669324986940838e-06, + "loss": 0.3613, + "step": 6493 + }, + { + "epoch": 0.14294519040281753, + "grad_norm": 1.5841183004373491, + "learning_rate": 9.669197497100337e-06, + "loss": 0.321, + "step": 6494 + }, + { + "epoch": 0.14296720228923618, + "grad_norm": 1.763252332104535, + "learning_rate": 9.669069983528892e-06, + "loss": 0.3845, + "step": 6495 + }, + { + "epoch": 0.14298921417565486, + "grad_norm": 20.65235612778663, + "learning_rate": 9.668942446227151e-06, + "loss": 0.3303, + "step": 6496 + }, + { + "epoch": 0.14301122606207353, + "grad_norm": 1.636643745933655, + "learning_rate": 9.668814885195764e-06, + "loss": 0.3711, + "step": 6497 + }, + { + "epoch": 0.14303323794849218, + "grad_norm": 3.174841968785866, + "learning_rate": 9.668687300435375e-06, + "loss": 0.419, + "step": 6498 + }, + { + "epoch": 0.14305524983491086, + "grad_norm": 1.218592930857863, + "learning_rate": 9.668559691946634e-06, + "loss": 0.3993, + "step": 6499 + }, + { + "epoch": 0.1430772617213295, + "grad_norm": 1.69155742843498, + "learning_rate": 9.668432059730193e-06, + "loss": 0.386, + "step": 6500 + }, + { + "epoch": 0.14309927360774818, + "grad_norm": 2.0831250085110535, + "learning_rate": 9.668304403786697e-06, + "loss": 0.3085, + "step": 6501 + }, + { + "epoch": 0.14312128549416686, + "grad_norm": 1.6853595663676397, + "learning_rate": 9.668176724116796e-06, + "loss": 0.3532, + "step": 6502 + }, + { + "epoch": 0.1431432973805855, + "grad_norm": 0.7114581530456487, + "learning_rate": 9.66804902072114e-06, + "loss": 0.5409, + "step": 6503 + }, + { + "epoch": 0.1431653092670042, + "grad_norm": 2.1742109440123625, + "learning_rate": 9.667921293600375e-06, + "loss": 0.3627, + "step": 6504 + }, + { + "epoch": 0.14318732115342284, + "grad_norm": 0.6868421254475636, + "learning_rate": 9.667793542755153e-06, + "loss": 0.5619, + "step": 6505 + }, + { + "epoch": 0.1432093330398415, + "grad_norm": 1.3998872575634353, + "learning_rate": 9.667665768186121e-06, + "loss": 0.3068, + "step": 6506 + }, + { + "epoch": 0.1432313449262602, + "grad_norm": 2.8011029158849037, + "learning_rate": 9.66753796989393e-06, + "loss": 0.4738, + "step": 6507 + }, + { + "epoch": 0.14325335681267884, + "grad_norm": 1.6863159980227271, + "learning_rate": 9.66741014787923e-06, + "loss": 0.3849, + "step": 6508 + }, + { + "epoch": 0.14327536869909752, + "grad_norm": 1.6659601435309026, + "learning_rate": 9.66728230214267e-06, + "loss": 0.3299, + "step": 6509 + }, + { + "epoch": 0.14329738058551617, + "grad_norm": 2.0217011752089165, + "learning_rate": 9.6671544326849e-06, + "loss": 0.3852, + "step": 6510 + }, + { + "epoch": 0.14331939247193484, + "grad_norm": 1.370144672009803, + "learning_rate": 9.66702653950657e-06, + "loss": 0.3555, + "step": 6511 + }, + { + "epoch": 0.14334140435835352, + "grad_norm": 17.983973777490434, + "learning_rate": 9.666898622608329e-06, + "loss": 0.4265, + "step": 6512 + }, + { + "epoch": 0.14336341624477217, + "grad_norm": 0.8846166170377022, + "learning_rate": 9.666770681990827e-06, + "loss": 0.551, + "step": 6513 + }, + { + "epoch": 0.14338542813119085, + "grad_norm": 1.4720660859019683, + "learning_rate": 9.666642717654716e-06, + "loss": 0.3345, + "step": 6514 + }, + { + "epoch": 0.14340744001760952, + "grad_norm": 1.4252328390185287, + "learning_rate": 9.666514729600643e-06, + "loss": 0.416, + "step": 6515 + }, + { + "epoch": 0.14342945190402817, + "grad_norm": 1.520383164648851, + "learning_rate": 9.666386717829263e-06, + "loss": 0.3839, + "step": 6516 + }, + { + "epoch": 0.14345146379044685, + "grad_norm": 0.6582047405352226, + "learning_rate": 9.666258682341224e-06, + "loss": 0.5321, + "step": 6517 + }, + { + "epoch": 0.1434734756768655, + "grad_norm": 1.5313651936829529, + "learning_rate": 9.666130623137178e-06, + "loss": 0.3564, + "step": 6518 + }, + { + "epoch": 0.14349548756328417, + "grad_norm": 0.6966420305367769, + "learning_rate": 9.666002540217774e-06, + "loss": 0.5607, + "step": 6519 + }, + { + "epoch": 0.14351749944970285, + "grad_norm": 1.3659918860253595, + "learning_rate": 9.665874433583664e-06, + "loss": 0.3269, + "step": 6520 + }, + { + "epoch": 0.1435395113361215, + "grad_norm": 2.278791373679212, + "learning_rate": 9.665746303235498e-06, + "loss": 0.362, + "step": 6521 + }, + { + "epoch": 0.14356152322254018, + "grad_norm": 6.703934244520737, + "learning_rate": 9.66561814917393e-06, + "loss": 0.3698, + "step": 6522 + }, + { + "epoch": 0.14358353510895883, + "grad_norm": 2.8016936579586424, + "learning_rate": 9.66548997139961e-06, + "loss": 0.3578, + "step": 6523 + }, + { + "epoch": 0.1436055469953775, + "grad_norm": 1.7294341159108426, + "learning_rate": 9.665361769913187e-06, + "loss": 0.4619, + "step": 6524 + }, + { + "epoch": 0.14362755888179618, + "grad_norm": 1.4015796752138134, + "learning_rate": 9.665233544715317e-06, + "loss": 0.3537, + "step": 6525 + }, + { + "epoch": 0.14364957076821483, + "grad_norm": 6.428227464409705, + "learning_rate": 9.665105295806647e-06, + "loss": 0.3813, + "step": 6526 + }, + { + "epoch": 0.1436715826546335, + "grad_norm": 2.53755839763968, + "learning_rate": 9.664977023187834e-06, + "loss": 0.4435, + "step": 6527 + }, + { + "epoch": 0.14369359454105216, + "grad_norm": 2.09131655699948, + "learning_rate": 9.664848726859525e-06, + "loss": 0.3909, + "step": 6528 + }, + { + "epoch": 0.14371560642747083, + "grad_norm": 1.6956428333896152, + "learning_rate": 9.664720406822375e-06, + "loss": 0.4193, + "step": 6529 + }, + { + "epoch": 0.1437376183138895, + "grad_norm": 1.9992454457685322, + "learning_rate": 9.664592063077035e-06, + "loss": 0.3458, + "step": 6530 + }, + { + "epoch": 0.14375963020030816, + "grad_norm": 7.162699147915779, + "learning_rate": 9.664463695624157e-06, + "loss": 0.3909, + "step": 6531 + }, + { + "epoch": 0.14378164208672684, + "grad_norm": 2.2544193853226173, + "learning_rate": 9.664335304464396e-06, + "loss": 0.3075, + "step": 6532 + }, + { + "epoch": 0.1438036539731455, + "grad_norm": 2.771360805083484, + "learning_rate": 9.664206889598402e-06, + "loss": 0.2723, + "step": 6533 + }, + { + "epoch": 0.14382566585956416, + "grad_norm": 1.4571488769145398, + "learning_rate": 9.664078451026827e-06, + "loss": 0.276, + "step": 6534 + }, + { + "epoch": 0.14384767774598284, + "grad_norm": 1.940684550406298, + "learning_rate": 9.663949988750326e-06, + "loss": 0.3346, + "step": 6535 + }, + { + "epoch": 0.1438696896324015, + "grad_norm": 0.8831941264120715, + "learning_rate": 9.66382150276955e-06, + "loss": 0.5525, + "step": 6536 + }, + { + "epoch": 0.14389170151882016, + "grad_norm": 1.8370070724518672, + "learning_rate": 9.663692993085155e-06, + "loss": 0.3955, + "step": 6537 + }, + { + "epoch": 0.14391371340523884, + "grad_norm": 1.3321126227240647, + "learning_rate": 9.663564459697793e-06, + "loss": 0.3706, + "step": 6538 + }, + { + "epoch": 0.1439357252916575, + "grad_norm": 1.6131219851136671, + "learning_rate": 9.663435902608113e-06, + "loss": 0.3007, + "step": 6539 + }, + { + "epoch": 0.14395773717807617, + "grad_norm": 2.025029482600824, + "learning_rate": 9.663307321816774e-06, + "loss": 0.4098, + "step": 6540 + }, + { + "epoch": 0.14397974906449482, + "grad_norm": 1.5795426148316891, + "learning_rate": 9.663178717324424e-06, + "loss": 0.3669, + "step": 6541 + }, + { + "epoch": 0.1440017609509135, + "grad_norm": 0.7183604548612533, + "learning_rate": 9.663050089131724e-06, + "loss": 0.5709, + "step": 6542 + }, + { + "epoch": 0.14402377283733217, + "grad_norm": 1.3776207828352087, + "learning_rate": 9.662921437239322e-06, + "loss": 0.3706, + "step": 6543 + }, + { + "epoch": 0.14404578472375082, + "grad_norm": 3.1727580422812003, + "learning_rate": 9.662792761647874e-06, + "loss": 0.4098, + "step": 6544 + }, + { + "epoch": 0.1440677966101695, + "grad_norm": 1.558463625229693, + "learning_rate": 9.662664062358033e-06, + "loss": 0.3662, + "step": 6545 + }, + { + "epoch": 0.14408980849658815, + "grad_norm": 1.5210819522916992, + "learning_rate": 9.662535339370454e-06, + "loss": 0.384, + "step": 6546 + }, + { + "epoch": 0.14411182038300682, + "grad_norm": 1.4434760567470604, + "learning_rate": 9.66240659268579e-06, + "loss": 0.3317, + "step": 6547 + }, + { + "epoch": 0.1441338322694255, + "grad_norm": 1.4982690909176912, + "learning_rate": 9.662277822304697e-06, + "loss": 0.3856, + "step": 6548 + }, + { + "epoch": 0.14415584415584415, + "grad_norm": 1.566777167609157, + "learning_rate": 9.66214902822783e-06, + "loss": 0.3524, + "step": 6549 + }, + { + "epoch": 0.14417785604226283, + "grad_norm": 2.0050771553795026, + "learning_rate": 9.66202021045584e-06, + "loss": 0.3841, + "step": 6550 + }, + { + "epoch": 0.14419986792868147, + "grad_norm": 1.527513257790596, + "learning_rate": 9.661891368989385e-06, + "loss": 0.3384, + "step": 6551 + }, + { + "epoch": 0.14422187981510015, + "grad_norm": 1.2219927795707461, + "learning_rate": 9.661762503829118e-06, + "loss": 0.3359, + "step": 6552 + }, + { + "epoch": 0.14424389170151883, + "grad_norm": 2.658451808986449, + "learning_rate": 9.661633614975696e-06, + "loss": 0.4086, + "step": 6553 + }, + { + "epoch": 0.14426590358793748, + "grad_norm": 0.8059789542307332, + "learning_rate": 9.661504702429771e-06, + "loss": 0.5471, + "step": 6554 + }, + { + "epoch": 0.14428791547435615, + "grad_norm": 1.1930067341709738, + "learning_rate": 9.661375766192003e-06, + "loss": 0.3215, + "step": 6555 + }, + { + "epoch": 0.14430992736077483, + "grad_norm": 1.4777440701115894, + "learning_rate": 9.661246806263042e-06, + "loss": 0.419, + "step": 6556 + }, + { + "epoch": 0.14433193924719348, + "grad_norm": 0.6769727658312925, + "learning_rate": 9.661117822643546e-06, + "loss": 0.5886, + "step": 6557 + }, + { + "epoch": 0.14435395113361216, + "grad_norm": 1.4632113243825993, + "learning_rate": 9.66098881533417e-06, + "loss": 0.4375, + "step": 6558 + }, + { + "epoch": 0.1443759630200308, + "grad_norm": 0.6654531974877826, + "learning_rate": 9.66085978433557e-06, + "loss": 0.565, + "step": 6559 + }, + { + "epoch": 0.14439797490644948, + "grad_norm": 0.6866528587953946, + "learning_rate": 9.660730729648403e-06, + "loss": 0.5777, + "step": 6560 + }, + { + "epoch": 0.14441998679286816, + "grad_norm": 1.506846165881265, + "learning_rate": 9.660601651273323e-06, + "loss": 0.3166, + "step": 6561 + }, + { + "epoch": 0.1444419986792868, + "grad_norm": 1.6121493759586591, + "learning_rate": 9.660472549210986e-06, + "loss": 0.4194, + "step": 6562 + }, + { + "epoch": 0.1444640105657055, + "grad_norm": 1.3915292552850635, + "learning_rate": 9.66034342346205e-06, + "loss": 0.3019, + "step": 6563 + }, + { + "epoch": 0.14448602245212414, + "grad_norm": 1.894829851151075, + "learning_rate": 9.66021427402717e-06, + "loss": 0.4547, + "step": 6564 + }, + { + "epoch": 0.1445080343385428, + "grad_norm": 1.6415234502864826, + "learning_rate": 9.660085100907003e-06, + "loss": 0.4424, + "step": 6565 + }, + { + "epoch": 0.1445300462249615, + "grad_norm": 1.385837015974905, + "learning_rate": 9.659955904102205e-06, + "loss": 0.3563, + "step": 6566 + }, + { + "epoch": 0.14455205811138014, + "grad_norm": 1.643524877825314, + "learning_rate": 9.659826683613432e-06, + "loss": 0.2721, + "step": 6567 + }, + { + "epoch": 0.14457406999779882, + "grad_norm": 1.8043690971580488, + "learning_rate": 9.65969743944134e-06, + "loss": 0.4019, + "step": 6568 + }, + { + "epoch": 0.14459608188421746, + "grad_norm": 1.7878303635510087, + "learning_rate": 9.65956817158659e-06, + "loss": 0.2962, + "step": 6569 + }, + { + "epoch": 0.14461809377063614, + "grad_norm": 0.7350903056688838, + "learning_rate": 9.659438880049834e-06, + "loss": 0.5546, + "step": 6570 + }, + { + "epoch": 0.14464010565705482, + "grad_norm": 1.1796345108929571, + "learning_rate": 9.659309564831734e-06, + "loss": 0.2908, + "step": 6571 + }, + { + "epoch": 0.14466211754347347, + "grad_norm": 2.518611692125163, + "learning_rate": 9.659180225932943e-06, + "loss": 0.3169, + "step": 6572 + }, + { + "epoch": 0.14468412942989214, + "grad_norm": 1.8906078345995632, + "learning_rate": 9.65905086335412e-06, + "loss": 0.3353, + "step": 6573 + }, + { + "epoch": 0.14470614131631082, + "grad_norm": 1.4637402861252011, + "learning_rate": 9.658921477095923e-06, + "loss": 0.3449, + "step": 6574 + }, + { + "epoch": 0.14472815320272947, + "grad_norm": 1.4819889308829826, + "learning_rate": 9.658792067159009e-06, + "loss": 0.4057, + "step": 6575 + }, + { + "epoch": 0.14475016508914815, + "grad_norm": 1.8244488035700748, + "learning_rate": 9.658662633544036e-06, + "loss": 0.429, + "step": 6576 + }, + { + "epoch": 0.1447721769755668, + "grad_norm": 0.6875392058496321, + "learning_rate": 9.65853317625166e-06, + "loss": 0.562, + "step": 6577 + }, + { + "epoch": 0.14479418886198547, + "grad_norm": 0.6642055273240602, + "learning_rate": 9.658403695282543e-06, + "loss": 0.5547, + "step": 6578 + }, + { + "epoch": 0.14481620074840415, + "grad_norm": 4.681538666627291, + "learning_rate": 9.65827419063734e-06, + "loss": 0.4491, + "step": 6579 + }, + { + "epoch": 0.1448382126348228, + "grad_norm": 0.632951069000322, + "learning_rate": 9.65814466231671e-06, + "loss": 0.5725, + "step": 6580 + }, + { + "epoch": 0.14486022452124148, + "grad_norm": 1.6039764786721356, + "learning_rate": 9.65801511032131e-06, + "loss": 0.4114, + "step": 6581 + }, + { + "epoch": 0.14488223640766013, + "grad_norm": 3.1454081471486894, + "learning_rate": 9.657885534651803e-06, + "loss": 0.3387, + "step": 6582 + }, + { + "epoch": 0.1449042482940788, + "grad_norm": 1.6337899351976948, + "learning_rate": 9.657755935308842e-06, + "loss": 0.2969, + "step": 6583 + }, + { + "epoch": 0.14492626018049748, + "grad_norm": 1.3360434145273266, + "learning_rate": 9.657626312293087e-06, + "loss": 0.3681, + "step": 6584 + }, + { + "epoch": 0.14494827206691613, + "grad_norm": 1.620415545316982, + "learning_rate": 9.657496665605198e-06, + "loss": 0.3901, + "step": 6585 + }, + { + "epoch": 0.1449702839533348, + "grad_norm": 1.599371759784329, + "learning_rate": 9.657366995245834e-06, + "loss": 0.4398, + "step": 6586 + }, + { + "epoch": 0.14499229583975345, + "grad_norm": 1.4884129660750869, + "learning_rate": 9.657237301215654e-06, + "loss": 0.411, + "step": 6587 + }, + { + "epoch": 0.14501430772617213, + "grad_norm": 2.963551636488963, + "learning_rate": 9.657107583515317e-06, + "loss": 0.402, + "step": 6588 + }, + { + "epoch": 0.1450363196125908, + "grad_norm": 14.556070252608306, + "learning_rate": 9.656977842145482e-06, + "loss": 0.3155, + "step": 6589 + }, + { + "epoch": 0.14505833149900946, + "grad_norm": 1.6568930925932162, + "learning_rate": 9.656848077106809e-06, + "loss": 0.3732, + "step": 6590 + }, + { + "epoch": 0.14508034338542813, + "grad_norm": 1.7770547579319074, + "learning_rate": 9.656718288399956e-06, + "loss": 0.3604, + "step": 6591 + }, + { + "epoch": 0.14510235527184678, + "grad_norm": 1.751237371104948, + "learning_rate": 9.656588476025583e-06, + "loss": 0.4594, + "step": 6592 + }, + { + "epoch": 0.14512436715826546, + "grad_norm": 0.8629430523655048, + "learning_rate": 9.656458639984352e-06, + "loss": 0.5667, + "step": 6593 + }, + { + "epoch": 0.14514637904468414, + "grad_norm": 1.5614833405732047, + "learning_rate": 9.656328780276921e-06, + "loss": 0.4122, + "step": 6594 + }, + { + "epoch": 0.1451683909311028, + "grad_norm": 1.8440740185661284, + "learning_rate": 9.65619889690395e-06, + "loss": 0.314, + "step": 6595 + }, + { + "epoch": 0.14519040281752146, + "grad_norm": 1.4420275880853166, + "learning_rate": 9.6560689898661e-06, + "loss": 0.3447, + "step": 6596 + }, + { + "epoch": 0.14521241470394014, + "grad_norm": 1.5641125816353485, + "learning_rate": 9.655939059164032e-06, + "loss": 0.4354, + "step": 6597 + }, + { + "epoch": 0.1452344265903588, + "grad_norm": 0.6983768949389734, + "learning_rate": 9.655809104798402e-06, + "loss": 0.5793, + "step": 6598 + }, + { + "epoch": 0.14525643847677747, + "grad_norm": 1.4707506318841284, + "learning_rate": 9.655679126769875e-06, + "loss": 0.353, + "step": 6599 + }, + { + "epoch": 0.14527845036319612, + "grad_norm": 1.3117800198674623, + "learning_rate": 9.65554912507911e-06, + "loss": 0.3573, + "step": 6600 + }, + { + "epoch": 0.1453004622496148, + "grad_norm": 1.842434781110855, + "learning_rate": 9.65541909972677e-06, + "loss": 0.2912, + "step": 6601 + }, + { + "epoch": 0.14532247413603347, + "grad_norm": 0.6690592757826528, + "learning_rate": 9.655289050713513e-06, + "loss": 0.5366, + "step": 6602 + }, + { + "epoch": 0.14534448602245212, + "grad_norm": 1.477235685135315, + "learning_rate": 9.655158978039998e-06, + "loss": 0.4069, + "step": 6603 + }, + { + "epoch": 0.1453664979088708, + "grad_norm": 1.630012190103945, + "learning_rate": 9.655028881706891e-06, + "loss": 0.3183, + "step": 6604 + }, + { + "epoch": 0.14538850979528944, + "grad_norm": 1.557380488109653, + "learning_rate": 9.65489876171485e-06, + "loss": 0.4376, + "step": 6605 + }, + { + "epoch": 0.14541052168170812, + "grad_norm": 1.7093303500294141, + "learning_rate": 9.654768618064538e-06, + "loss": 0.3381, + "step": 6606 + }, + { + "epoch": 0.1454325335681268, + "grad_norm": 0.6275783440261017, + "learning_rate": 9.654638450756617e-06, + "loss": 0.5537, + "step": 6607 + }, + { + "epoch": 0.14545454545454545, + "grad_norm": 1.4887619530858673, + "learning_rate": 9.654508259791746e-06, + "loss": 0.3364, + "step": 6608 + }, + { + "epoch": 0.14547655734096412, + "grad_norm": 1.8159643669647396, + "learning_rate": 9.654378045170588e-06, + "loss": 0.4665, + "step": 6609 + }, + { + "epoch": 0.14549856922738277, + "grad_norm": 1.5109911792102695, + "learning_rate": 9.654247806893804e-06, + "loss": 0.2878, + "step": 6610 + }, + { + "epoch": 0.14552058111380145, + "grad_norm": 0.6646197145754347, + "learning_rate": 9.654117544962058e-06, + "loss": 0.5683, + "step": 6611 + }, + { + "epoch": 0.14554259300022013, + "grad_norm": 3.6597616168468, + "learning_rate": 9.653987259376012e-06, + "loss": 0.4429, + "step": 6612 + }, + { + "epoch": 0.14556460488663878, + "grad_norm": 0.6245796507748395, + "learning_rate": 9.653856950136326e-06, + "loss": 0.522, + "step": 6613 + }, + { + "epoch": 0.14558661677305745, + "grad_norm": 1.3856141193638514, + "learning_rate": 9.653726617243662e-06, + "loss": 0.3106, + "step": 6614 + }, + { + "epoch": 0.14560862865947613, + "grad_norm": 1.7740397403999124, + "learning_rate": 9.653596260698685e-06, + "loss": 0.4205, + "step": 6615 + }, + { + "epoch": 0.14563064054589478, + "grad_norm": 1.7081980464534674, + "learning_rate": 9.653465880502056e-06, + "loss": 0.3349, + "step": 6616 + }, + { + "epoch": 0.14565265243231346, + "grad_norm": 1.7655657467764845, + "learning_rate": 9.653335476654439e-06, + "loss": 0.3584, + "step": 6617 + }, + { + "epoch": 0.1456746643187321, + "grad_norm": 1.5290379783210328, + "learning_rate": 9.653205049156493e-06, + "loss": 0.2859, + "step": 6618 + }, + { + "epoch": 0.14569667620515078, + "grad_norm": 1.7161917508631972, + "learning_rate": 9.653074598008886e-06, + "loss": 0.3954, + "step": 6619 + }, + { + "epoch": 0.14571868809156946, + "grad_norm": 1.4030431854744272, + "learning_rate": 9.652944123212278e-06, + "loss": 0.3027, + "step": 6620 + }, + { + "epoch": 0.1457406999779881, + "grad_norm": 1.6448868956748777, + "learning_rate": 9.652813624767332e-06, + "loss": 0.322, + "step": 6621 + }, + { + "epoch": 0.14576271186440679, + "grad_norm": 1.4084288543824757, + "learning_rate": 9.652683102674712e-06, + "loss": 0.2861, + "step": 6622 + }, + { + "epoch": 0.14578472375082543, + "grad_norm": 1.3557540004070474, + "learning_rate": 9.652552556935081e-06, + "loss": 0.2928, + "step": 6623 + }, + { + "epoch": 0.1458067356372441, + "grad_norm": 1.827276589618058, + "learning_rate": 9.652421987549105e-06, + "loss": 0.4608, + "step": 6624 + }, + { + "epoch": 0.1458287475236628, + "grad_norm": 1.9250109815521386, + "learning_rate": 9.652291394517443e-06, + "loss": 0.3159, + "step": 6625 + }, + { + "epoch": 0.14585075941008144, + "grad_norm": 2.645115884968228, + "learning_rate": 9.652160777840762e-06, + "loss": 0.3586, + "step": 6626 + }, + { + "epoch": 0.14587277129650011, + "grad_norm": 1.5183485224087845, + "learning_rate": 9.652030137519726e-06, + "loss": 0.3305, + "step": 6627 + }, + { + "epoch": 0.14589478318291876, + "grad_norm": 1.715072732534483, + "learning_rate": 9.651899473554995e-06, + "loss": 0.2654, + "step": 6628 + }, + { + "epoch": 0.14591679506933744, + "grad_norm": 1.9058002163570171, + "learning_rate": 9.65176878594724e-06, + "loss": 0.3607, + "step": 6629 + }, + { + "epoch": 0.14593880695575612, + "grad_norm": 1.8057390498849966, + "learning_rate": 9.651638074697118e-06, + "loss": 0.3462, + "step": 6630 + }, + { + "epoch": 0.14596081884217477, + "grad_norm": 1.952105088092289, + "learning_rate": 9.651507339805298e-06, + "loss": 0.379, + "step": 6631 + }, + { + "epoch": 0.14598283072859344, + "grad_norm": 1.497751889594792, + "learning_rate": 9.651376581272443e-06, + "loss": 0.4239, + "step": 6632 + }, + { + "epoch": 0.14600484261501212, + "grad_norm": 1.9035531431528805, + "learning_rate": 9.651245799099219e-06, + "loss": 0.3037, + "step": 6633 + }, + { + "epoch": 0.14602685450143077, + "grad_norm": 0.6979169132378399, + "learning_rate": 9.651114993286287e-06, + "loss": 0.5414, + "step": 6634 + }, + { + "epoch": 0.14604886638784945, + "grad_norm": 1.5891899579172597, + "learning_rate": 9.650984163834316e-06, + "loss": 0.3049, + "step": 6635 + }, + { + "epoch": 0.1460708782742681, + "grad_norm": 0.6494065871197369, + "learning_rate": 9.650853310743966e-06, + "loss": 0.5267, + "step": 6636 + }, + { + "epoch": 0.14609289016068677, + "grad_norm": 1.3538611278780697, + "learning_rate": 9.650722434015909e-06, + "loss": 0.2714, + "step": 6637 + }, + { + "epoch": 0.14611490204710545, + "grad_norm": 0.6361048511919073, + "learning_rate": 9.650591533650803e-06, + "loss": 0.5634, + "step": 6638 + }, + { + "epoch": 0.1461369139335241, + "grad_norm": 1.4053041314592623, + "learning_rate": 9.650460609649318e-06, + "loss": 0.2823, + "step": 6639 + }, + { + "epoch": 0.14615892581994278, + "grad_norm": 2.0606657059929967, + "learning_rate": 9.65032966201212e-06, + "loss": 0.3619, + "step": 6640 + }, + { + "epoch": 0.14618093770636142, + "grad_norm": 2.0317412781241453, + "learning_rate": 9.650198690739872e-06, + "loss": 0.4043, + "step": 6641 + }, + { + "epoch": 0.1462029495927801, + "grad_norm": 3.093871693288146, + "learning_rate": 9.650067695833239e-06, + "loss": 0.437, + "step": 6642 + }, + { + "epoch": 0.14622496147919878, + "grad_norm": 0.6806592784703145, + "learning_rate": 9.649936677292887e-06, + "loss": 0.5781, + "step": 6643 + }, + { + "epoch": 0.14624697336561743, + "grad_norm": 1.2115436642414856, + "learning_rate": 9.649805635119484e-06, + "loss": 0.2601, + "step": 6644 + }, + { + "epoch": 0.1462689852520361, + "grad_norm": 1.790095681811507, + "learning_rate": 9.649674569313694e-06, + "loss": 0.37, + "step": 6645 + }, + { + "epoch": 0.14629099713845475, + "grad_norm": 1.7062400294857463, + "learning_rate": 9.649543479876185e-06, + "loss": 0.4506, + "step": 6646 + }, + { + "epoch": 0.14631300902487343, + "grad_norm": 3.1646292726833942, + "learning_rate": 9.649412366807622e-06, + "loss": 0.3965, + "step": 6647 + }, + { + "epoch": 0.1463350209112921, + "grad_norm": 8.811029251684731, + "learning_rate": 9.649281230108673e-06, + "loss": 0.3136, + "step": 6648 + }, + { + "epoch": 0.14635703279771076, + "grad_norm": 1.382616056791319, + "learning_rate": 9.649150069780003e-06, + "loss": 0.3754, + "step": 6649 + }, + { + "epoch": 0.14637904468412943, + "grad_norm": 1.3571556785587002, + "learning_rate": 9.649018885822277e-06, + "loss": 0.3156, + "step": 6650 + }, + { + "epoch": 0.14640105657054808, + "grad_norm": 1.2904314410347824, + "learning_rate": 9.648887678236163e-06, + "loss": 0.3303, + "step": 6651 + }, + { + "epoch": 0.14642306845696676, + "grad_norm": 1.510891415523315, + "learning_rate": 9.648756447022329e-06, + "loss": 0.4609, + "step": 6652 + }, + { + "epoch": 0.14644508034338544, + "grad_norm": 2.022260623479401, + "learning_rate": 9.648625192181441e-06, + "loss": 0.3284, + "step": 6653 + }, + { + "epoch": 0.14646709222980409, + "grad_norm": 1.9489440534686449, + "learning_rate": 9.64849391371417e-06, + "loss": 0.3638, + "step": 6654 + }, + { + "epoch": 0.14648910411622276, + "grad_norm": 1.5616992713139892, + "learning_rate": 9.648362611621177e-06, + "loss": 0.3853, + "step": 6655 + }, + { + "epoch": 0.14651111600264144, + "grad_norm": 1.7213554041124894, + "learning_rate": 9.648231285903132e-06, + "loss": 0.4072, + "step": 6656 + }, + { + "epoch": 0.1465331278890601, + "grad_norm": 1.9185330798202316, + "learning_rate": 9.648099936560702e-06, + "loss": 0.4607, + "step": 6657 + }, + { + "epoch": 0.14655513977547877, + "grad_norm": 1.6620690142511882, + "learning_rate": 9.647968563594555e-06, + "loss": 0.3797, + "step": 6658 + }, + { + "epoch": 0.14657715166189741, + "grad_norm": 0.6926300640151932, + "learning_rate": 9.647837167005358e-06, + "loss": 0.5507, + "step": 6659 + }, + { + "epoch": 0.1465991635483161, + "grad_norm": 1.6282696355004755, + "learning_rate": 9.64770574679378e-06, + "loss": 0.2788, + "step": 6660 + }, + { + "epoch": 0.14662117543473477, + "grad_norm": 1.4426885482346854, + "learning_rate": 9.647574302960491e-06, + "loss": 0.3411, + "step": 6661 + }, + { + "epoch": 0.14664318732115342, + "grad_norm": 1.6746568439795337, + "learning_rate": 9.647442835506154e-06, + "loss": 0.3391, + "step": 6662 + }, + { + "epoch": 0.1466651992075721, + "grad_norm": 1.812415740628587, + "learning_rate": 9.64731134443144e-06, + "loss": 0.3165, + "step": 6663 + }, + { + "epoch": 0.14668721109399074, + "grad_norm": 1.4027677410320667, + "learning_rate": 9.647179829737018e-06, + "loss": 0.3261, + "step": 6664 + }, + { + "epoch": 0.14670922298040942, + "grad_norm": 1.4295192292735086, + "learning_rate": 9.647048291423555e-06, + "loss": 0.3331, + "step": 6665 + }, + { + "epoch": 0.1467312348668281, + "grad_norm": 1.9673559426273368, + "learning_rate": 9.64691672949172e-06, + "loss": 0.4527, + "step": 6666 + }, + { + "epoch": 0.14675324675324675, + "grad_norm": 1.395047153640992, + "learning_rate": 9.64678514394218e-06, + "loss": 0.3142, + "step": 6667 + }, + { + "epoch": 0.14677525863966542, + "grad_norm": 2.631996054409601, + "learning_rate": 9.646653534775607e-06, + "loss": 0.4485, + "step": 6668 + }, + { + "epoch": 0.14679727052608407, + "grad_norm": 3.0456495688343272, + "learning_rate": 9.646521901992667e-06, + "loss": 0.3361, + "step": 6669 + }, + { + "epoch": 0.14681928241250275, + "grad_norm": 1.653548358495391, + "learning_rate": 9.646390245594032e-06, + "loss": 0.3185, + "step": 6670 + }, + { + "epoch": 0.14684129429892143, + "grad_norm": 1.3234027643567454, + "learning_rate": 9.646258565580368e-06, + "loss": 0.3304, + "step": 6671 + }, + { + "epoch": 0.14686330618534008, + "grad_norm": 0.6363926991429147, + "learning_rate": 9.646126861952347e-06, + "loss": 0.5495, + "step": 6672 + }, + { + "epoch": 0.14688531807175875, + "grad_norm": 1.5024107205283614, + "learning_rate": 9.645995134710636e-06, + "loss": 0.3775, + "step": 6673 + }, + { + "epoch": 0.14690732995817743, + "grad_norm": 1.1902299343129275, + "learning_rate": 9.645863383855905e-06, + "loss": 0.2776, + "step": 6674 + }, + { + "epoch": 0.14692934184459608, + "grad_norm": 1.8554332371515132, + "learning_rate": 9.645731609388826e-06, + "loss": 0.3709, + "step": 6675 + }, + { + "epoch": 0.14695135373101476, + "grad_norm": 1.5361652068640765, + "learning_rate": 9.645599811310066e-06, + "loss": 0.4357, + "step": 6676 + }, + { + "epoch": 0.1469733656174334, + "grad_norm": 1.560257657488654, + "learning_rate": 9.645467989620296e-06, + "loss": 0.3266, + "step": 6677 + }, + { + "epoch": 0.14699537750385208, + "grad_norm": 1.8288578926928103, + "learning_rate": 9.645336144320185e-06, + "loss": 0.3336, + "step": 6678 + }, + { + "epoch": 0.14701738939027076, + "grad_norm": 4.093910317900051, + "learning_rate": 9.645204275410404e-06, + "loss": 0.3792, + "step": 6679 + }, + { + "epoch": 0.1470394012766894, + "grad_norm": 0.6669040030384865, + "learning_rate": 9.645072382891623e-06, + "loss": 0.5772, + "step": 6680 + }, + { + "epoch": 0.14706141316310808, + "grad_norm": 1.6474972781823423, + "learning_rate": 9.644940466764515e-06, + "loss": 0.3652, + "step": 6681 + }, + { + "epoch": 0.14708342504952673, + "grad_norm": 1.80863417306001, + "learning_rate": 9.644808527029744e-06, + "loss": 0.356, + "step": 6682 + }, + { + "epoch": 0.1471054369359454, + "grad_norm": 1.6554501545646423, + "learning_rate": 9.644676563687986e-06, + "loss": 0.382, + "step": 6683 + }, + { + "epoch": 0.1471274488223641, + "grad_norm": 1.9562359824379802, + "learning_rate": 9.64454457673991e-06, + "loss": 0.3629, + "step": 6684 + }, + { + "epoch": 0.14714946070878274, + "grad_norm": 1.325540614809849, + "learning_rate": 9.644412566186188e-06, + "loss": 0.3289, + "step": 6685 + }, + { + "epoch": 0.1471714725952014, + "grad_norm": 1.43284976816953, + "learning_rate": 9.644280532027488e-06, + "loss": 0.3499, + "step": 6686 + }, + { + "epoch": 0.14719348448162006, + "grad_norm": 1.7436270597752817, + "learning_rate": 9.644148474264483e-06, + "loss": 0.4304, + "step": 6687 + }, + { + "epoch": 0.14721549636803874, + "grad_norm": 1.3096590020652594, + "learning_rate": 9.644016392897845e-06, + "loss": 0.3276, + "step": 6688 + }, + { + "epoch": 0.14723750825445742, + "grad_norm": 1.856683053782768, + "learning_rate": 9.643884287928244e-06, + "loss": 0.3097, + "step": 6689 + }, + { + "epoch": 0.14725952014087607, + "grad_norm": 0.6431647104607979, + "learning_rate": 9.643752159356353e-06, + "loss": 0.5449, + "step": 6690 + }, + { + "epoch": 0.14728153202729474, + "grad_norm": 1.6415166543547273, + "learning_rate": 9.643620007182841e-06, + "loss": 0.3813, + "step": 6691 + }, + { + "epoch": 0.1473035439137134, + "grad_norm": 1.481134595712156, + "learning_rate": 9.643487831408382e-06, + "loss": 0.3714, + "step": 6692 + }, + { + "epoch": 0.14732555580013207, + "grad_norm": 0.6358500234715374, + "learning_rate": 9.643355632033646e-06, + "loss": 0.5546, + "step": 6693 + }, + { + "epoch": 0.14734756768655075, + "grad_norm": 1.793588812752368, + "learning_rate": 9.643223409059308e-06, + "loss": 0.3204, + "step": 6694 + }, + { + "epoch": 0.1473695795729694, + "grad_norm": 1.622069874842982, + "learning_rate": 9.643091162486035e-06, + "loss": 0.4759, + "step": 6695 + }, + { + "epoch": 0.14739159145938807, + "grad_norm": 4.5712137519186085, + "learning_rate": 9.642958892314503e-06, + "loss": 0.3388, + "step": 6696 + }, + { + "epoch": 0.14741360334580675, + "grad_norm": 1.608365763906123, + "learning_rate": 9.642826598545381e-06, + "loss": 0.2875, + "step": 6697 + }, + { + "epoch": 0.1474356152322254, + "grad_norm": 1.5185776215686102, + "learning_rate": 9.642694281179345e-06, + "loss": 0.3842, + "step": 6698 + }, + { + "epoch": 0.14745762711864407, + "grad_norm": 0.70829455569837, + "learning_rate": 9.642561940217068e-06, + "loss": 0.57, + "step": 6699 + }, + { + "epoch": 0.14747963900506272, + "grad_norm": 2.8865524167014787, + "learning_rate": 9.642429575659217e-06, + "loss": 0.3798, + "step": 6700 + }, + { + "epoch": 0.1475016508914814, + "grad_norm": 0.6312960672182912, + "learning_rate": 9.642297187506472e-06, + "loss": 0.5391, + "step": 6701 + }, + { + "epoch": 0.14752366277790008, + "grad_norm": 1.5401980757377374, + "learning_rate": 9.6421647757595e-06, + "loss": 0.4095, + "step": 6702 + }, + { + "epoch": 0.14754567466431873, + "grad_norm": 1.4233482637593429, + "learning_rate": 9.642032340418975e-06, + "loss": 0.4042, + "step": 6703 + }, + { + "epoch": 0.1475676865507374, + "grad_norm": 0.6219614162636989, + "learning_rate": 9.641899881485572e-06, + "loss": 0.5435, + "step": 6704 + }, + { + "epoch": 0.14758969843715605, + "grad_norm": 2.615598162795616, + "learning_rate": 9.641767398959965e-06, + "loss": 0.4147, + "step": 6705 + }, + { + "epoch": 0.14761171032357473, + "grad_norm": 1.5132529747855457, + "learning_rate": 9.641634892842824e-06, + "loss": 0.4199, + "step": 6706 + }, + { + "epoch": 0.1476337222099934, + "grad_norm": 1.444951034971567, + "learning_rate": 9.641502363134825e-06, + "loss": 0.4096, + "step": 6707 + }, + { + "epoch": 0.14765573409641206, + "grad_norm": 21.293461937368523, + "learning_rate": 9.641369809836641e-06, + "loss": 0.331, + "step": 6708 + }, + { + "epoch": 0.14767774598283073, + "grad_norm": 0.6784973905066889, + "learning_rate": 9.641237232948946e-06, + "loss": 0.5497, + "step": 6709 + }, + { + "epoch": 0.14769975786924938, + "grad_norm": 1.5216519892492175, + "learning_rate": 9.641104632472413e-06, + "loss": 0.3968, + "step": 6710 + }, + { + "epoch": 0.14772176975566806, + "grad_norm": 0.629591691795819, + "learning_rate": 9.640972008407715e-06, + "loss": 0.5461, + "step": 6711 + }, + { + "epoch": 0.14774378164208674, + "grad_norm": 0.6154373838612999, + "learning_rate": 9.640839360755528e-06, + "loss": 0.5283, + "step": 6712 + }, + { + "epoch": 0.14776579352850538, + "grad_norm": 1.457035890975328, + "learning_rate": 9.640706689516526e-06, + "loss": 0.3015, + "step": 6713 + }, + { + "epoch": 0.14778780541492406, + "grad_norm": 1.6712735773775804, + "learning_rate": 9.640573994691383e-06, + "loss": 0.3734, + "step": 6714 + }, + { + "epoch": 0.14780981730134274, + "grad_norm": 1.5242607100319023, + "learning_rate": 9.640441276280772e-06, + "loss": 0.3503, + "step": 6715 + }, + { + "epoch": 0.1478318291877614, + "grad_norm": 1.5440069932173572, + "learning_rate": 9.64030853428537e-06, + "loss": 0.3793, + "step": 6716 + }, + { + "epoch": 0.14785384107418006, + "grad_norm": 1.2069699400409484, + "learning_rate": 9.64017576870585e-06, + "loss": 0.2642, + "step": 6717 + }, + { + "epoch": 0.1478758529605987, + "grad_norm": 1.5532490634210454, + "learning_rate": 9.640042979542887e-06, + "loss": 0.3213, + "step": 6718 + }, + { + "epoch": 0.1478978648470174, + "grad_norm": 1.755073148053067, + "learning_rate": 9.639910166797157e-06, + "loss": 0.3603, + "step": 6719 + }, + { + "epoch": 0.14791987673343607, + "grad_norm": 1.5252606147140297, + "learning_rate": 9.639777330469333e-06, + "loss": 0.4048, + "step": 6720 + }, + { + "epoch": 0.14794188861985472, + "grad_norm": 0.7522896406511003, + "learning_rate": 9.639644470560091e-06, + "loss": 0.5456, + "step": 6721 + }, + { + "epoch": 0.1479639005062734, + "grad_norm": 1.958539118750168, + "learning_rate": 9.639511587070108e-06, + "loss": 0.4258, + "step": 6722 + }, + { + "epoch": 0.14798591239269204, + "grad_norm": 1.4158045154914933, + "learning_rate": 9.639378680000055e-06, + "loss": 0.4055, + "step": 6723 + }, + { + "epoch": 0.14800792427911072, + "grad_norm": 1.2439143832090862, + "learning_rate": 9.639245749350611e-06, + "loss": 0.3315, + "step": 6724 + }, + { + "epoch": 0.1480299361655294, + "grad_norm": 1.5567791302779819, + "learning_rate": 9.639112795122454e-06, + "loss": 0.3047, + "step": 6725 + }, + { + "epoch": 0.14805194805194805, + "grad_norm": 2.320102627409547, + "learning_rate": 9.638979817316254e-06, + "loss": 0.3469, + "step": 6726 + }, + { + "epoch": 0.14807395993836672, + "grad_norm": 0.6808663273677104, + "learning_rate": 9.63884681593269e-06, + "loss": 0.5444, + "step": 6727 + }, + { + "epoch": 0.14809597182478537, + "grad_norm": 1.1553856765400872, + "learning_rate": 9.638713790972438e-06, + "loss": 0.3331, + "step": 6728 + }, + { + "epoch": 0.14811798371120405, + "grad_norm": 1.227426830570628, + "learning_rate": 9.638580742436174e-06, + "loss": 0.3442, + "step": 6729 + }, + { + "epoch": 0.14813999559762273, + "grad_norm": 0.6376091332276914, + "learning_rate": 9.638447670324574e-06, + "loss": 0.5356, + "step": 6730 + }, + { + "epoch": 0.14816200748404137, + "grad_norm": 1.3798205219779378, + "learning_rate": 9.638314574638312e-06, + "loss": 0.4335, + "step": 6731 + }, + { + "epoch": 0.14818401937046005, + "grad_norm": 0.6446109844729053, + "learning_rate": 9.638181455378069e-06, + "loss": 0.5638, + "step": 6732 + }, + { + "epoch": 0.1482060312568787, + "grad_norm": 1.2243908804076533, + "learning_rate": 9.638048312544517e-06, + "loss": 0.3989, + "step": 6733 + }, + { + "epoch": 0.14822804314329738, + "grad_norm": 0.6463157190370244, + "learning_rate": 9.637915146138334e-06, + "loss": 0.5443, + "step": 6734 + }, + { + "epoch": 0.14825005502971605, + "grad_norm": 0.6421884997577101, + "learning_rate": 9.6377819561602e-06, + "loss": 0.5539, + "step": 6735 + }, + { + "epoch": 0.1482720669161347, + "grad_norm": 0.6124727992129863, + "learning_rate": 9.63764874261079e-06, + "loss": 0.5542, + "step": 6736 + }, + { + "epoch": 0.14829407880255338, + "grad_norm": 0.6357254981457577, + "learning_rate": 9.63751550549078e-06, + "loss": 0.5664, + "step": 6737 + }, + { + "epoch": 0.14831609068897206, + "grad_norm": 4.078118682824261, + "learning_rate": 9.637382244800847e-06, + "loss": 0.2297, + "step": 6738 + }, + { + "epoch": 0.1483381025753907, + "grad_norm": 1.346427922251939, + "learning_rate": 9.63724896054167e-06, + "loss": 0.3259, + "step": 6739 + }, + { + "epoch": 0.14836011446180938, + "grad_norm": 2.362460942847075, + "learning_rate": 9.637115652713923e-06, + "loss": 0.2732, + "step": 6740 + }, + { + "epoch": 0.14838212634822803, + "grad_norm": 0.6293629111698408, + "learning_rate": 9.63698232131829e-06, + "loss": 0.5253, + "step": 6741 + }, + { + "epoch": 0.1484041382346467, + "grad_norm": 1.5476863007309176, + "learning_rate": 9.636848966355443e-06, + "loss": 0.4248, + "step": 6742 + }, + { + "epoch": 0.1484261501210654, + "grad_norm": 1.6351434800885956, + "learning_rate": 9.63671558782606e-06, + "loss": 0.438, + "step": 6743 + }, + { + "epoch": 0.14844816200748404, + "grad_norm": 1.5394822810478768, + "learning_rate": 9.636582185730823e-06, + "loss": 0.3237, + "step": 6744 + }, + { + "epoch": 0.1484701738939027, + "grad_norm": 1.4267572741855115, + "learning_rate": 9.636448760070407e-06, + "loss": 0.2697, + "step": 6745 + }, + { + "epoch": 0.14849218578032136, + "grad_norm": 0.6332096682425844, + "learning_rate": 9.636315310845488e-06, + "loss": 0.5474, + "step": 6746 + }, + { + "epoch": 0.14851419766674004, + "grad_norm": 1.33424265889133, + "learning_rate": 9.636181838056748e-06, + "loss": 0.3315, + "step": 6747 + }, + { + "epoch": 0.14853620955315872, + "grad_norm": 2.15029155517265, + "learning_rate": 9.636048341704865e-06, + "loss": 0.3234, + "step": 6748 + }, + { + "epoch": 0.14855822143957736, + "grad_norm": 2.4834111333246622, + "learning_rate": 9.635914821790515e-06, + "loss": 0.3562, + "step": 6749 + }, + { + "epoch": 0.14858023332599604, + "grad_norm": 1.3622886851660794, + "learning_rate": 9.63578127831438e-06, + "loss": 0.3752, + "step": 6750 + }, + { + "epoch": 0.1486022452124147, + "grad_norm": 1.6629344024049055, + "learning_rate": 9.635647711277137e-06, + "loss": 0.3205, + "step": 6751 + }, + { + "epoch": 0.14862425709883337, + "grad_norm": 1.5499933504684953, + "learning_rate": 9.635514120679463e-06, + "loss": 0.3545, + "step": 6752 + }, + { + "epoch": 0.14864626898525204, + "grad_norm": 0.6724643007357596, + "learning_rate": 9.63538050652204e-06, + "loss": 0.5603, + "step": 6753 + }, + { + "epoch": 0.1486682808716707, + "grad_norm": 1.5325039810727668, + "learning_rate": 9.635246868805546e-06, + "loss": 0.3619, + "step": 6754 + }, + { + "epoch": 0.14869029275808937, + "grad_norm": 1.2744967935743934, + "learning_rate": 9.635113207530659e-06, + "loss": 0.3155, + "step": 6755 + }, + { + "epoch": 0.14871230464450805, + "grad_norm": 0.624245488281203, + "learning_rate": 9.63497952269806e-06, + "loss": 0.5454, + "step": 6756 + }, + { + "epoch": 0.1487343165309267, + "grad_norm": 1.1497776067882435, + "learning_rate": 9.634845814308427e-06, + "loss": 0.3093, + "step": 6757 + }, + { + "epoch": 0.14875632841734537, + "grad_norm": 0.6263426570559908, + "learning_rate": 9.63471208236244e-06, + "loss": 0.5576, + "step": 6758 + }, + { + "epoch": 0.14877834030376402, + "grad_norm": 0.6306803794195066, + "learning_rate": 9.63457832686078e-06, + "loss": 0.5529, + "step": 6759 + }, + { + "epoch": 0.1488003521901827, + "grad_norm": 2.3559835591070075, + "learning_rate": 9.634444547804128e-06, + "loss": 0.3923, + "step": 6760 + }, + { + "epoch": 0.14882236407660138, + "grad_norm": 1.2246321698136464, + "learning_rate": 9.634310745193159e-06, + "loss": 0.3156, + "step": 6761 + }, + { + "epoch": 0.14884437596302003, + "grad_norm": 1.080027849354549, + "learning_rate": 9.634176919028557e-06, + "loss": 0.2826, + "step": 6762 + }, + { + "epoch": 0.1488663878494387, + "grad_norm": 1.646891044973146, + "learning_rate": 9.634043069310999e-06, + "loss": 0.3841, + "step": 6763 + }, + { + "epoch": 0.14888839973585735, + "grad_norm": 2.7772432247012304, + "learning_rate": 9.633909196041168e-06, + "loss": 0.3182, + "step": 6764 + }, + { + "epoch": 0.14891041162227603, + "grad_norm": 11.230534940470452, + "learning_rate": 9.633775299219746e-06, + "loss": 0.3486, + "step": 6765 + }, + { + "epoch": 0.1489324235086947, + "grad_norm": 2.1652304269180394, + "learning_rate": 9.63364137884741e-06, + "loss": 0.4499, + "step": 6766 + }, + { + "epoch": 0.14895443539511335, + "grad_norm": 1.4133756447818873, + "learning_rate": 9.63350743492484e-06, + "loss": 0.3536, + "step": 6767 + }, + { + "epoch": 0.14897644728153203, + "grad_norm": 1.3581043253922698, + "learning_rate": 9.633373467452721e-06, + "loss": 0.3728, + "step": 6768 + }, + { + "epoch": 0.14899845916795068, + "grad_norm": 1.387068369407949, + "learning_rate": 9.633239476431731e-06, + "loss": 0.2969, + "step": 6769 + }, + { + "epoch": 0.14902047105436936, + "grad_norm": 1.3515392022069446, + "learning_rate": 9.633105461862551e-06, + "loss": 0.394, + "step": 6770 + }, + { + "epoch": 0.14904248294078803, + "grad_norm": 1.6285594272212138, + "learning_rate": 9.632971423745863e-06, + "loss": 0.3716, + "step": 6771 + }, + { + "epoch": 0.14906449482720668, + "grad_norm": 1.3179754629893141, + "learning_rate": 9.632837362082348e-06, + "loss": 0.3229, + "step": 6772 + }, + { + "epoch": 0.14908650671362536, + "grad_norm": 1.5362352710184215, + "learning_rate": 9.632703276872685e-06, + "loss": 0.4066, + "step": 6773 + }, + { + "epoch": 0.149108518600044, + "grad_norm": 1.5377038901008717, + "learning_rate": 9.63256916811756e-06, + "loss": 0.254, + "step": 6774 + }, + { + "epoch": 0.1491305304864627, + "grad_norm": 1.7913536198099866, + "learning_rate": 9.632435035817654e-06, + "loss": 0.3405, + "step": 6775 + }, + { + "epoch": 0.14915254237288136, + "grad_norm": 1.724684703068157, + "learning_rate": 9.632300879973644e-06, + "loss": 0.4586, + "step": 6776 + }, + { + "epoch": 0.1491745542593, + "grad_norm": 1.2377550699476974, + "learning_rate": 9.632166700586215e-06, + "loss": 0.3776, + "step": 6777 + }, + { + "epoch": 0.1491965661457187, + "grad_norm": 4.065430106757302, + "learning_rate": 9.632032497656049e-06, + "loss": 0.3717, + "step": 6778 + }, + { + "epoch": 0.14921857803213737, + "grad_norm": 2.7314279146500224, + "learning_rate": 9.63189827118383e-06, + "loss": 0.3202, + "step": 6779 + }, + { + "epoch": 0.14924058991855602, + "grad_norm": 1.4703173079885274, + "learning_rate": 9.631764021170236e-06, + "loss": 0.3327, + "step": 6780 + }, + { + "epoch": 0.1492626018049747, + "grad_norm": 1.742844254695312, + "learning_rate": 9.631629747615952e-06, + "loss": 0.3477, + "step": 6781 + }, + { + "epoch": 0.14928461369139334, + "grad_norm": 1.5584750868122925, + "learning_rate": 9.63149545052166e-06, + "loss": 0.3994, + "step": 6782 + }, + { + "epoch": 0.14930662557781202, + "grad_norm": 2.118018131308362, + "learning_rate": 9.631361129888041e-06, + "loss": 0.365, + "step": 6783 + }, + { + "epoch": 0.1493286374642307, + "grad_norm": 1.1567776293404728, + "learning_rate": 9.631226785715781e-06, + "loss": 0.3428, + "step": 6784 + }, + { + "epoch": 0.14935064935064934, + "grad_norm": 1.6921099580793237, + "learning_rate": 9.63109241800556e-06, + "loss": 0.4105, + "step": 6785 + }, + { + "epoch": 0.14937266123706802, + "grad_norm": 0.7615149703441044, + "learning_rate": 9.630958026758062e-06, + "loss": 0.5213, + "step": 6786 + }, + { + "epoch": 0.14939467312348667, + "grad_norm": 2.3064234412840907, + "learning_rate": 9.630823611973971e-06, + "loss": 0.356, + "step": 6787 + }, + { + "epoch": 0.14941668500990535, + "grad_norm": 1.266680403514596, + "learning_rate": 9.630689173653967e-06, + "loss": 0.3331, + "step": 6788 + }, + { + "epoch": 0.14943869689632402, + "grad_norm": 0.6561269013187735, + "learning_rate": 9.630554711798736e-06, + "loss": 0.5456, + "step": 6789 + }, + { + "epoch": 0.14946070878274267, + "grad_norm": 1.3324605547394108, + "learning_rate": 9.630420226408962e-06, + "loss": 0.3004, + "step": 6790 + }, + { + "epoch": 0.14948272066916135, + "grad_norm": 1.9638228365731798, + "learning_rate": 9.630285717485327e-06, + "loss": 0.3308, + "step": 6791 + }, + { + "epoch": 0.14950473255558, + "grad_norm": 2.345186525767214, + "learning_rate": 9.630151185028514e-06, + "loss": 0.3618, + "step": 6792 + }, + { + "epoch": 0.14952674444199868, + "grad_norm": 1.6273130044831388, + "learning_rate": 9.630016629039206e-06, + "loss": 0.4487, + "step": 6793 + }, + { + "epoch": 0.14954875632841735, + "grad_norm": 0.6900902670310209, + "learning_rate": 9.62988204951809e-06, + "loss": 0.5234, + "step": 6794 + }, + { + "epoch": 0.149570768214836, + "grad_norm": 3.0700121068818413, + "learning_rate": 9.62974744646585e-06, + "loss": 0.3698, + "step": 6795 + }, + { + "epoch": 0.14959278010125468, + "grad_norm": 0.6611344626195689, + "learning_rate": 9.629612819883166e-06, + "loss": 0.5318, + "step": 6796 + }, + { + "epoch": 0.14961479198767336, + "grad_norm": 1.4891272687859822, + "learning_rate": 9.629478169770725e-06, + "loss": 0.3607, + "step": 6797 + }, + { + "epoch": 0.149636803874092, + "grad_norm": 1.8351806509511408, + "learning_rate": 9.62934349612921e-06, + "loss": 0.4278, + "step": 6798 + }, + { + "epoch": 0.14965881576051068, + "grad_norm": 1.5314849744400683, + "learning_rate": 9.62920879895931e-06, + "loss": 0.2927, + "step": 6799 + }, + { + "epoch": 0.14968082764692933, + "grad_norm": 2.144745761083234, + "learning_rate": 9.629074078261704e-06, + "loss": 0.4494, + "step": 6800 + }, + { + "epoch": 0.149702839533348, + "grad_norm": 2.141775210197939, + "learning_rate": 9.628939334037079e-06, + "loss": 0.4043, + "step": 6801 + }, + { + "epoch": 0.14972485141976669, + "grad_norm": 1.2785866852363548, + "learning_rate": 9.62880456628612e-06, + "loss": 0.3961, + "step": 6802 + }, + { + "epoch": 0.14974686330618533, + "grad_norm": 13.156449549102463, + "learning_rate": 9.628669775009514e-06, + "loss": 0.4304, + "step": 6803 + }, + { + "epoch": 0.149768875192604, + "grad_norm": 1.4216347088625825, + "learning_rate": 9.62853496020794e-06, + "loss": 0.3311, + "step": 6804 + }, + { + "epoch": 0.14979088707902266, + "grad_norm": 4.743633349558932, + "learning_rate": 9.62840012188209e-06, + "loss": 0.2963, + "step": 6805 + }, + { + "epoch": 0.14981289896544134, + "grad_norm": 2.006071119165584, + "learning_rate": 9.628265260032645e-06, + "loss": 0.4113, + "step": 6806 + }, + { + "epoch": 0.14983491085186001, + "grad_norm": 3.133841707142009, + "learning_rate": 9.628130374660291e-06, + "loss": 0.3772, + "step": 6807 + }, + { + "epoch": 0.14985692273827866, + "grad_norm": 0.8510562119614572, + "learning_rate": 9.627995465765714e-06, + "loss": 0.5704, + "step": 6808 + }, + { + "epoch": 0.14987893462469734, + "grad_norm": 1.2835678784715578, + "learning_rate": 9.627860533349601e-06, + "loss": 0.2941, + "step": 6809 + }, + { + "epoch": 0.149900946511116, + "grad_norm": 2.0805435616192587, + "learning_rate": 9.627725577412637e-06, + "loss": 0.4106, + "step": 6810 + }, + { + "epoch": 0.14992295839753467, + "grad_norm": 0.6509121144375617, + "learning_rate": 9.627590597955508e-06, + "loss": 0.5404, + "step": 6811 + }, + { + "epoch": 0.14994497028395334, + "grad_norm": 1.9401956500760151, + "learning_rate": 9.627455594978898e-06, + "loss": 0.2866, + "step": 6812 + }, + { + "epoch": 0.149966982170372, + "grad_norm": 1.7247256381702551, + "learning_rate": 9.627320568483496e-06, + "loss": 0.3309, + "step": 6813 + }, + { + "epoch": 0.14998899405679067, + "grad_norm": 0.6607335405406223, + "learning_rate": 9.627185518469986e-06, + "loss": 0.5339, + "step": 6814 + }, + { + "epoch": 0.15001100594320935, + "grad_norm": 2.256606390694557, + "learning_rate": 9.627050444939055e-06, + "loss": 0.4674, + "step": 6815 + }, + { + "epoch": 0.150033017829628, + "grad_norm": 0.6905891443448492, + "learning_rate": 9.626915347891392e-06, + "loss": 0.5614, + "step": 6816 + }, + { + "epoch": 0.15005502971604667, + "grad_norm": 1.553976418944621, + "learning_rate": 9.62678022732768e-06, + "loss": 0.3115, + "step": 6817 + }, + { + "epoch": 0.15007704160246532, + "grad_norm": 1.2680626021660795, + "learning_rate": 9.626645083248607e-06, + "loss": 0.2876, + "step": 6818 + }, + { + "epoch": 0.150099053488884, + "grad_norm": 1.8854018418922402, + "learning_rate": 9.62650991565486e-06, + "loss": 0.3693, + "step": 6819 + }, + { + "epoch": 0.15012106537530268, + "grad_norm": 2.9308529911586136, + "learning_rate": 9.626374724547127e-06, + "loss": 0.2815, + "step": 6820 + }, + { + "epoch": 0.15014307726172132, + "grad_norm": 1.1103619653075347, + "learning_rate": 9.626239509926094e-06, + "loss": 0.2376, + "step": 6821 + }, + { + "epoch": 0.15016508914814, + "grad_norm": 0.6291470879473077, + "learning_rate": 9.62610427179245e-06, + "loss": 0.5416, + "step": 6822 + }, + { + "epoch": 0.15018710103455865, + "grad_norm": 1.4970377287933696, + "learning_rate": 9.625969010146878e-06, + "loss": 0.3923, + "step": 6823 + }, + { + "epoch": 0.15020911292097733, + "grad_norm": 1.5227390016047888, + "learning_rate": 9.62583372499007e-06, + "loss": 0.3241, + "step": 6824 + }, + { + "epoch": 0.150231124807396, + "grad_norm": 1.3258316336657283, + "learning_rate": 9.62569841632271e-06, + "loss": 0.3713, + "step": 6825 + }, + { + "epoch": 0.15025313669381465, + "grad_norm": 1.3353377344803015, + "learning_rate": 9.62556308414549e-06, + "loss": 0.3467, + "step": 6826 + }, + { + "epoch": 0.15027514858023333, + "grad_norm": 1.2854510078855859, + "learning_rate": 9.625427728459094e-06, + "loss": 0.3219, + "step": 6827 + }, + { + "epoch": 0.15029716046665198, + "grad_norm": 1.6630598915087507, + "learning_rate": 9.625292349264213e-06, + "loss": 0.3068, + "step": 6828 + }, + { + "epoch": 0.15031917235307066, + "grad_norm": 1.7170301996214599, + "learning_rate": 9.625156946561531e-06, + "loss": 0.4346, + "step": 6829 + }, + { + "epoch": 0.15034118423948933, + "grad_norm": 1.7871897774573526, + "learning_rate": 9.62502152035174e-06, + "loss": 0.403, + "step": 6830 + }, + { + "epoch": 0.15036319612590798, + "grad_norm": 1.330302311694526, + "learning_rate": 9.624886070635526e-06, + "loss": 0.3523, + "step": 6831 + }, + { + "epoch": 0.15038520801232666, + "grad_norm": 1.214460998652205, + "learning_rate": 9.624750597413577e-06, + "loss": 0.3399, + "step": 6832 + }, + { + "epoch": 0.1504072198987453, + "grad_norm": 1.7091176733928684, + "learning_rate": 9.624615100686585e-06, + "loss": 0.3119, + "step": 6833 + }, + { + "epoch": 0.15042923178516399, + "grad_norm": 1.5235354964732932, + "learning_rate": 9.624479580455236e-06, + "loss": 0.3288, + "step": 6834 + }, + { + "epoch": 0.15045124367158266, + "grad_norm": 1.52509267893818, + "learning_rate": 9.624344036720219e-06, + "loss": 0.3979, + "step": 6835 + }, + { + "epoch": 0.1504732555580013, + "grad_norm": 1.6840813091765585, + "learning_rate": 9.624208469482223e-06, + "loss": 0.3825, + "step": 6836 + }, + { + "epoch": 0.15049526744442, + "grad_norm": 1.6047125432918536, + "learning_rate": 9.624072878741937e-06, + "loss": 0.3804, + "step": 6837 + }, + { + "epoch": 0.15051727933083867, + "grad_norm": 4.962193245487453, + "learning_rate": 9.62393726450005e-06, + "loss": 0.2893, + "step": 6838 + }, + { + "epoch": 0.15053929121725731, + "grad_norm": 1.3486402736248977, + "learning_rate": 9.623801626757252e-06, + "loss": 0.2748, + "step": 6839 + }, + { + "epoch": 0.150561303103676, + "grad_norm": 1.421216152386257, + "learning_rate": 9.62366596551423e-06, + "loss": 0.3172, + "step": 6840 + }, + { + "epoch": 0.15058331499009464, + "grad_norm": 2.1180913796726206, + "learning_rate": 9.623530280771678e-06, + "loss": 0.4071, + "step": 6841 + }, + { + "epoch": 0.15060532687651332, + "grad_norm": 1.4064806386481308, + "learning_rate": 9.623394572530283e-06, + "loss": 0.3484, + "step": 6842 + }, + { + "epoch": 0.150627338762932, + "grad_norm": 2.409787365061177, + "learning_rate": 9.623258840790733e-06, + "loss": 0.3542, + "step": 6843 + }, + { + "epoch": 0.15064935064935064, + "grad_norm": 1.5519380888276595, + "learning_rate": 9.62312308555372e-06, + "loss": 0.3678, + "step": 6844 + }, + { + "epoch": 0.15067136253576932, + "grad_norm": 2.4111660141764824, + "learning_rate": 9.622987306819933e-06, + "loss": 0.4332, + "step": 6845 + }, + { + "epoch": 0.15069337442218797, + "grad_norm": 2.231432577701035, + "learning_rate": 9.622851504590064e-06, + "loss": 0.3886, + "step": 6846 + }, + { + "epoch": 0.15071538630860665, + "grad_norm": 2.691232811897657, + "learning_rate": 9.622715678864801e-06, + "loss": 0.3272, + "step": 6847 + }, + { + "epoch": 0.15073739819502532, + "grad_norm": 3.990985468310219, + "learning_rate": 9.622579829644836e-06, + "loss": 0.3598, + "step": 6848 + }, + { + "epoch": 0.15075941008144397, + "grad_norm": 1.4826702509612015, + "learning_rate": 9.622443956930856e-06, + "loss": 0.3923, + "step": 6849 + }, + { + "epoch": 0.15078142196786265, + "grad_norm": 2.529867536973247, + "learning_rate": 9.622308060723558e-06, + "loss": 0.3621, + "step": 6850 + }, + { + "epoch": 0.1508034338542813, + "grad_norm": 2.621265344210551, + "learning_rate": 9.622172141023626e-06, + "loss": 0.3823, + "step": 6851 + }, + { + "epoch": 0.15082544574069998, + "grad_norm": 1.4930835366520314, + "learning_rate": 9.622036197831755e-06, + "loss": 0.3536, + "step": 6852 + }, + { + "epoch": 0.15084745762711865, + "grad_norm": 1.952585925797861, + "learning_rate": 9.621900231148634e-06, + "loss": 0.3745, + "step": 6853 + }, + { + "epoch": 0.1508694695135373, + "grad_norm": 1.7658266971028769, + "learning_rate": 9.621764240974954e-06, + "loss": 0.3429, + "step": 6854 + }, + { + "epoch": 0.15089148139995598, + "grad_norm": 1.6901158858553564, + "learning_rate": 9.621628227311406e-06, + "loss": 0.345, + "step": 6855 + }, + { + "epoch": 0.15091349328637466, + "grad_norm": 1.4293473890667872, + "learning_rate": 9.621492190158684e-06, + "loss": 0.3819, + "step": 6856 + }, + { + "epoch": 0.1509355051727933, + "grad_norm": 0.7501844303501808, + "learning_rate": 9.621356129517476e-06, + "loss": 0.5642, + "step": 6857 + }, + { + "epoch": 0.15095751705921198, + "grad_norm": 1.3437451626028767, + "learning_rate": 9.621220045388477e-06, + "loss": 0.3316, + "step": 6858 + }, + { + "epoch": 0.15097952894563063, + "grad_norm": 1.1090955328848642, + "learning_rate": 9.621083937772375e-06, + "loss": 0.3786, + "step": 6859 + }, + { + "epoch": 0.1510015408320493, + "grad_norm": 1.4890306778079794, + "learning_rate": 9.620947806669864e-06, + "loss": 0.3684, + "step": 6860 + }, + { + "epoch": 0.15102355271846798, + "grad_norm": 1.4428084238712908, + "learning_rate": 9.620811652081634e-06, + "loss": 0.3818, + "step": 6861 + }, + { + "epoch": 0.15104556460488663, + "grad_norm": 1.430500743430371, + "learning_rate": 9.620675474008381e-06, + "loss": 0.4007, + "step": 6862 + }, + { + "epoch": 0.1510675764913053, + "grad_norm": 1.3870186300872778, + "learning_rate": 9.620539272450792e-06, + "loss": 0.3765, + "step": 6863 + }, + { + "epoch": 0.15108958837772396, + "grad_norm": 1.6165222372557477, + "learning_rate": 9.62040304740956e-06, + "loss": 0.4193, + "step": 6864 + }, + { + "epoch": 0.15111160026414264, + "grad_norm": 2.1354833100360513, + "learning_rate": 9.620266798885382e-06, + "loss": 0.3532, + "step": 6865 + }, + { + "epoch": 0.15113361215056131, + "grad_norm": 1.647185189280764, + "learning_rate": 9.620130526878949e-06, + "loss": 0.3806, + "step": 6866 + }, + { + "epoch": 0.15115562403697996, + "grad_norm": 1.5025859983718493, + "learning_rate": 9.619994231390947e-06, + "loss": 0.4032, + "step": 6867 + }, + { + "epoch": 0.15117763592339864, + "grad_norm": 1.787711618963391, + "learning_rate": 9.619857912422076e-06, + "loss": 0.422, + "step": 6868 + }, + { + "epoch": 0.1511996478098173, + "grad_norm": 1.4198275573861947, + "learning_rate": 9.619721569973028e-06, + "loss": 0.3431, + "step": 6869 + }, + { + "epoch": 0.15122165969623597, + "grad_norm": 0.7196718095595629, + "learning_rate": 9.619585204044493e-06, + "loss": 0.5511, + "step": 6870 + }, + { + "epoch": 0.15124367158265464, + "grad_norm": 0.6785550623566271, + "learning_rate": 9.619448814637167e-06, + "loss": 0.5389, + "step": 6871 + }, + { + "epoch": 0.1512656834690733, + "grad_norm": 1.6435102715502194, + "learning_rate": 9.61931240175174e-06, + "loss": 0.3298, + "step": 6872 + }, + { + "epoch": 0.15128769535549197, + "grad_norm": 2.313993730679731, + "learning_rate": 9.61917596538891e-06, + "loss": 0.3498, + "step": 6873 + }, + { + "epoch": 0.15130970724191062, + "grad_norm": 2.0765028348240326, + "learning_rate": 9.619039505549364e-06, + "loss": 0.4383, + "step": 6874 + }, + { + "epoch": 0.1513317191283293, + "grad_norm": 1.276566372312974, + "learning_rate": 9.6189030222338e-06, + "loss": 0.3841, + "step": 6875 + }, + { + "epoch": 0.15135373101474797, + "grad_norm": 1.5379829199726143, + "learning_rate": 9.618766515442912e-06, + "loss": 0.4297, + "step": 6876 + }, + { + "epoch": 0.15137574290116662, + "grad_norm": 2.0216197686366173, + "learning_rate": 9.618629985177393e-06, + "loss": 0.3165, + "step": 6877 + }, + { + "epoch": 0.1513977547875853, + "grad_norm": 2.074281979126712, + "learning_rate": 9.618493431437935e-06, + "loss": 0.3941, + "step": 6878 + }, + { + "epoch": 0.15141976667400397, + "grad_norm": 1.5978368678158037, + "learning_rate": 9.618356854225233e-06, + "loss": 0.3299, + "step": 6879 + }, + { + "epoch": 0.15144177856042262, + "grad_norm": 2.2117074590007486, + "learning_rate": 9.618220253539982e-06, + "loss": 0.3463, + "step": 6880 + }, + { + "epoch": 0.1514637904468413, + "grad_norm": 1.7086130234748622, + "learning_rate": 9.618083629382877e-06, + "loss": 0.2404, + "step": 6881 + }, + { + "epoch": 0.15148580233325995, + "grad_norm": 1.3551795516480496, + "learning_rate": 9.61794698175461e-06, + "loss": 0.37, + "step": 6882 + }, + { + "epoch": 0.15150781421967863, + "grad_norm": 2.589169577503135, + "learning_rate": 9.617810310655878e-06, + "loss": 0.3352, + "step": 6883 + }, + { + "epoch": 0.1515298261060973, + "grad_norm": 1.493449323614, + "learning_rate": 9.617673616087373e-06, + "loss": 0.3692, + "step": 6884 + }, + { + "epoch": 0.15155183799251595, + "grad_norm": 1.2339541094397695, + "learning_rate": 9.617536898049793e-06, + "loss": 0.3951, + "step": 6885 + }, + { + "epoch": 0.15157384987893463, + "grad_norm": 0.9899953828126626, + "learning_rate": 9.617400156543829e-06, + "loss": 0.5447, + "step": 6886 + }, + { + "epoch": 0.15159586176535328, + "grad_norm": 1.502737292482869, + "learning_rate": 9.61726339157018e-06, + "loss": 0.3057, + "step": 6887 + }, + { + "epoch": 0.15161787365177196, + "grad_norm": 4.022066268789141, + "learning_rate": 9.617126603129537e-06, + "loss": 0.3316, + "step": 6888 + }, + { + "epoch": 0.15163988553819063, + "grad_norm": 1.3236072328520483, + "learning_rate": 9.616989791222598e-06, + "loss": 0.3392, + "step": 6889 + }, + { + "epoch": 0.15166189742460928, + "grad_norm": 1.2891953663960107, + "learning_rate": 9.616852955850058e-06, + "loss": 0.2893, + "step": 6890 + }, + { + "epoch": 0.15168390931102796, + "grad_norm": 1.513770465474619, + "learning_rate": 9.616716097012612e-06, + "loss": 0.3267, + "step": 6891 + }, + { + "epoch": 0.1517059211974466, + "grad_norm": 7.300939975484151, + "learning_rate": 9.616579214710956e-06, + "loss": 0.3515, + "step": 6892 + }, + { + "epoch": 0.15172793308386529, + "grad_norm": 1.5308275139689547, + "learning_rate": 9.616442308945784e-06, + "loss": 0.3341, + "step": 6893 + }, + { + "epoch": 0.15174994497028396, + "grad_norm": 7.0732318808104, + "learning_rate": 9.616305379717796e-06, + "loss": 0.4438, + "step": 6894 + }, + { + "epoch": 0.1517719568567026, + "grad_norm": 3.2439284642479507, + "learning_rate": 9.616168427027683e-06, + "loss": 0.407, + "step": 6895 + }, + { + "epoch": 0.1517939687431213, + "grad_norm": 3.3456767951648616, + "learning_rate": 9.616031450876143e-06, + "loss": 0.3741, + "step": 6896 + }, + { + "epoch": 0.15181598062953996, + "grad_norm": 2.2775468813809074, + "learning_rate": 9.615894451263873e-06, + "loss": 0.3548, + "step": 6897 + }, + { + "epoch": 0.15183799251595861, + "grad_norm": 1.7733931823280862, + "learning_rate": 9.61575742819157e-06, + "loss": 0.3484, + "step": 6898 + }, + { + "epoch": 0.1518600044023773, + "grad_norm": 1.7970026809049817, + "learning_rate": 9.615620381659927e-06, + "loss": 0.3732, + "step": 6899 + }, + { + "epoch": 0.15188201628879594, + "grad_norm": 1.0057020700181243, + "learning_rate": 9.615483311669642e-06, + "loss": 0.5461, + "step": 6900 + }, + { + "epoch": 0.15190402817521462, + "grad_norm": 2.6761034154130763, + "learning_rate": 9.615346218221413e-06, + "loss": 0.3466, + "step": 6901 + }, + { + "epoch": 0.1519260400616333, + "grad_norm": 1.3974187892551049, + "learning_rate": 9.615209101315937e-06, + "loss": 0.3116, + "step": 6902 + }, + { + "epoch": 0.15194805194805194, + "grad_norm": 4.093092681081768, + "learning_rate": 9.61507196095391e-06, + "loss": 0.3128, + "step": 6903 + }, + { + "epoch": 0.15197006383447062, + "grad_norm": 2.354929703470651, + "learning_rate": 9.614934797136028e-06, + "loss": 0.2878, + "step": 6904 + }, + { + "epoch": 0.15199207572088927, + "grad_norm": 1.540577040960864, + "learning_rate": 9.61479760986299e-06, + "loss": 0.3366, + "step": 6905 + }, + { + "epoch": 0.15201408760730795, + "grad_norm": 1.7460471330409875, + "learning_rate": 9.614660399135491e-06, + "loss": 0.472, + "step": 6906 + }, + { + "epoch": 0.15203609949372662, + "grad_norm": 2.7430224616226795, + "learning_rate": 9.61452316495423e-06, + "loss": 0.4394, + "step": 6907 + }, + { + "epoch": 0.15205811138014527, + "grad_norm": 1.5757217028027366, + "learning_rate": 9.614385907319905e-06, + "loss": 0.368, + "step": 6908 + }, + { + "epoch": 0.15208012326656395, + "grad_norm": 1.484178318085775, + "learning_rate": 9.614248626233213e-06, + "loss": 0.4058, + "step": 6909 + }, + { + "epoch": 0.1521021351529826, + "grad_norm": 0.7643603235958215, + "learning_rate": 9.61411132169485e-06, + "loss": 0.5793, + "step": 6910 + }, + { + "epoch": 0.15212414703940128, + "grad_norm": 1.5234975528851382, + "learning_rate": 9.613973993705517e-06, + "loss": 0.3321, + "step": 6911 + }, + { + "epoch": 0.15214615892581995, + "grad_norm": 0.7004640437311287, + "learning_rate": 9.613836642265908e-06, + "loss": 0.5411, + "step": 6912 + }, + { + "epoch": 0.1521681708122386, + "grad_norm": 1.6839686057935612, + "learning_rate": 9.613699267376725e-06, + "loss": 0.4078, + "step": 6913 + }, + { + "epoch": 0.15219018269865728, + "grad_norm": 2.0427632206956554, + "learning_rate": 9.613561869038664e-06, + "loss": 0.3633, + "step": 6914 + }, + { + "epoch": 0.15221219458507593, + "grad_norm": 1.804090826439803, + "learning_rate": 9.613424447252425e-06, + "loss": 0.3792, + "step": 6915 + }, + { + "epoch": 0.1522342064714946, + "grad_norm": 1.8671327057791505, + "learning_rate": 9.613287002018702e-06, + "loss": 0.4413, + "step": 6916 + }, + { + "epoch": 0.15225621835791328, + "grad_norm": 1.7820228533367588, + "learning_rate": 9.613149533338202e-06, + "loss": 0.3556, + "step": 6917 + }, + { + "epoch": 0.15227823024433193, + "grad_norm": 1.719551512590578, + "learning_rate": 9.613012041211616e-06, + "loss": 0.328, + "step": 6918 + }, + { + "epoch": 0.1523002421307506, + "grad_norm": 1.797657254400337, + "learning_rate": 9.612874525639644e-06, + "loss": 0.3672, + "step": 6919 + }, + { + "epoch": 0.15232225401716928, + "grad_norm": 0.737390596265584, + "learning_rate": 9.612736986622988e-06, + "loss": 0.5127, + "step": 6920 + }, + { + "epoch": 0.15234426590358793, + "grad_norm": 1.3972447884889712, + "learning_rate": 9.612599424162344e-06, + "loss": 0.3452, + "step": 6921 + }, + { + "epoch": 0.1523662777900066, + "grad_norm": 1.5212354097539427, + "learning_rate": 9.612461838258414e-06, + "loss": 0.3485, + "step": 6922 + }, + { + "epoch": 0.15238828967642526, + "grad_norm": 1.496202041582887, + "learning_rate": 9.612324228911894e-06, + "loss": 0.3734, + "step": 6923 + }, + { + "epoch": 0.15241030156284394, + "grad_norm": 0.6769956549536132, + "learning_rate": 9.612186596123487e-06, + "loss": 0.5244, + "step": 6924 + }, + { + "epoch": 0.1524323134492626, + "grad_norm": 1.5003675916896206, + "learning_rate": 9.61204893989389e-06, + "loss": 0.3262, + "step": 6925 + }, + { + "epoch": 0.15245432533568126, + "grad_norm": 1.3756037187002834, + "learning_rate": 9.611911260223802e-06, + "loss": 0.2617, + "step": 6926 + }, + { + "epoch": 0.15247633722209994, + "grad_norm": 1.7001663589350282, + "learning_rate": 9.611773557113925e-06, + "loss": 0.4148, + "step": 6927 + }, + { + "epoch": 0.1524983491085186, + "grad_norm": 1.5210433695138088, + "learning_rate": 9.611635830564958e-06, + "loss": 0.3878, + "step": 6928 + }, + { + "epoch": 0.15252036099493727, + "grad_norm": 3.2111889684652546, + "learning_rate": 9.6114980805776e-06, + "loss": 0.3186, + "step": 6929 + }, + { + "epoch": 0.15254237288135594, + "grad_norm": 1.9832696571322461, + "learning_rate": 9.611360307152552e-06, + "loss": 0.3579, + "step": 6930 + }, + { + "epoch": 0.1525643847677746, + "grad_norm": 0.7072032115400436, + "learning_rate": 9.611222510290515e-06, + "loss": 0.5484, + "step": 6931 + }, + { + "epoch": 0.15258639665419327, + "grad_norm": 1.3816769826186597, + "learning_rate": 9.611084689992188e-06, + "loss": 0.3002, + "step": 6932 + }, + { + "epoch": 0.15260840854061192, + "grad_norm": 1.996127163040656, + "learning_rate": 9.610946846258274e-06, + "loss": 0.4361, + "step": 6933 + }, + { + "epoch": 0.1526304204270306, + "grad_norm": 3.067459523010324, + "learning_rate": 9.61080897908947e-06, + "loss": 0.3139, + "step": 6934 + }, + { + "epoch": 0.15265243231344927, + "grad_norm": 1.569009228420586, + "learning_rate": 9.610671088486478e-06, + "loss": 0.3401, + "step": 6935 + }, + { + "epoch": 0.15267444419986792, + "grad_norm": 1.670401716979331, + "learning_rate": 9.610533174449999e-06, + "loss": 0.3558, + "step": 6936 + }, + { + "epoch": 0.1526964560862866, + "grad_norm": 1.749062977889206, + "learning_rate": 9.610395236980733e-06, + "loss": 0.3869, + "step": 6937 + }, + { + "epoch": 0.15271846797270527, + "grad_norm": 1.5313921346490327, + "learning_rate": 9.610257276079384e-06, + "loss": 0.3602, + "step": 6938 + }, + { + "epoch": 0.15274047985912392, + "grad_norm": 1.735142860480784, + "learning_rate": 9.610119291746651e-06, + "loss": 0.3863, + "step": 6939 + }, + { + "epoch": 0.1527624917455426, + "grad_norm": 2.246442558053512, + "learning_rate": 9.609981283983234e-06, + "loss": 0.3432, + "step": 6940 + }, + { + "epoch": 0.15278450363196125, + "grad_norm": 1.5755463220476893, + "learning_rate": 9.60984325278984e-06, + "loss": 0.3529, + "step": 6941 + }, + { + "epoch": 0.15280651551837993, + "grad_norm": 1.46988747983869, + "learning_rate": 9.609705198167164e-06, + "loss": 0.2715, + "step": 6942 + }, + { + "epoch": 0.1528285274047986, + "grad_norm": 1.7401246386139375, + "learning_rate": 9.609567120115908e-06, + "loss": 0.4284, + "step": 6943 + }, + { + "epoch": 0.15285053929121725, + "grad_norm": 1.237016693398892, + "learning_rate": 9.609429018636779e-06, + "loss": 0.3322, + "step": 6944 + }, + { + "epoch": 0.15287255117763593, + "grad_norm": 1.5850111610590343, + "learning_rate": 9.609290893730474e-06, + "loss": 0.3858, + "step": 6945 + }, + { + "epoch": 0.15289456306405458, + "grad_norm": 1.5520656562148214, + "learning_rate": 9.6091527453977e-06, + "loss": 0.3793, + "step": 6946 + }, + { + "epoch": 0.15291657495047326, + "grad_norm": 1.4462152445036205, + "learning_rate": 9.609014573639152e-06, + "loss": 0.4269, + "step": 6947 + }, + { + "epoch": 0.15293858683689193, + "grad_norm": 1.3666973716868631, + "learning_rate": 9.608876378455539e-06, + "loss": 0.317, + "step": 6948 + }, + { + "epoch": 0.15296059872331058, + "grad_norm": 1.4565040251960815, + "learning_rate": 9.608738159847559e-06, + "loss": 0.3349, + "step": 6949 + }, + { + "epoch": 0.15298261060972926, + "grad_norm": 1.656386246183873, + "learning_rate": 9.608599917815917e-06, + "loss": 0.377, + "step": 6950 + }, + { + "epoch": 0.1530046224961479, + "grad_norm": 1.8806997962412042, + "learning_rate": 9.608461652361314e-06, + "loss": 0.3048, + "step": 6951 + }, + { + "epoch": 0.15302663438256658, + "grad_norm": 1.950074018705657, + "learning_rate": 9.608323363484456e-06, + "loss": 0.3624, + "step": 6952 + }, + { + "epoch": 0.15304864626898526, + "grad_norm": 2.2447265660104976, + "learning_rate": 9.60818505118604e-06, + "loss": 0.3812, + "step": 6953 + }, + { + "epoch": 0.1530706581554039, + "grad_norm": 1.69939278194056, + "learning_rate": 9.608046715466773e-06, + "loss": 0.3664, + "step": 6954 + }, + { + "epoch": 0.1530926700418226, + "grad_norm": 1.6177259222895835, + "learning_rate": 9.607908356327357e-06, + "loss": 0.367, + "step": 6955 + }, + { + "epoch": 0.15311468192824126, + "grad_norm": 1.5215284776438283, + "learning_rate": 9.607769973768497e-06, + "loss": 0.3451, + "step": 6956 + }, + { + "epoch": 0.1531366938146599, + "grad_norm": 1.543971396987712, + "learning_rate": 9.607631567790895e-06, + "loss": 0.3558, + "step": 6957 + }, + { + "epoch": 0.1531587057010786, + "grad_norm": 1.6231533800281868, + "learning_rate": 9.607493138395255e-06, + "loss": 0.3855, + "step": 6958 + }, + { + "epoch": 0.15318071758749724, + "grad_norm": 0.6509890141485202, + "learning_rate": 9.607354685582277e-06, + "loss": 0.5381, + "step": 6959 + }, + { + "epoch": 0.15320272947391592, + "grad_norm": 1.5350187588425435, + "learning_rate": 9.60721620935267e-06, + "loss": 0.3707, + "step": 6960 + }, + { + "epoch": 0.1532247413603346, + "grad_norm": 1.888756141483737, + "learning_rate": 9.607077709707133e-06, + "loss": 0.3494, + "step": 6961 + }, + { + "epoch": 0.15324675324675324, + "grad_norm": 1.5652923509315662, + "learning_rate": 9.606939186646374e-06, + "loss": 0.3019, + "step": 6962 + }, + { + "epoch": 0.15326876513317192, + "grad_norm": 1.5663373528123876, + "learning_rate": 9.606800640171095e-06, + "loss": 0.3, + "step": 6963 + }, + { + "epoch": 0.15329077701959057, + "grad_norm": 1.2709743317719597, + "learning_rate": 9.606662070282e-06, + "loss": 0.324, + "step": 6964 + }, + { + "epoch": 0.15331278890600925, + "grad_norm": 2.517561642286681, + "learning_rate": 9.606523476979794e-06, + "loss": 0.3427, + "step": 6965 + }, + { + "epoch": 0.15333480079242792, + "grad_norm": 1.5187944602880152, + "learning_rate": 9.606384860265182e-06, + "loss": 0.3401, + "step": 6966 + }, + { + "epoch": 0.15335681267884657, + "grad_norm": 1.6834959758914103, + "learning_rate": 9.606246220138868e-06, + "loss": 0.4201, + "step": 6967 + }, + { + "epoch": 0.15337882456526525, + "grad_norm": 1.7148957306456172, + "learning_rate": 9.606107556601554e-06, + "loss": 0.347, + "step": 6968 + }, + { + "epoch": 0.1534008364516839, + "grad_norm": 1.9484142756543639, + "learning_rate": 9.605968869653948e-06, + "loss": 0.3152, + "step": 6969 + }, + { + "epoch": 0.15342284833810257, + "grad_norm": 1.3918314245178167, + "learning_rate": 9.605830159296754e-06, + "loss": 0.3145, + "step": 6970 + }, + { + "epoch": 0.15344486022452125, + "grad_norm": 0.6416748206650412, + "learning_rate": 9.605691425530678e-06, + "loss": 0.547, + "step": 6971 + }, + { + "epoch": 0.1534668721109399, + "grad_norm": 1.3764149768027933, + "learning_rate": 9.605552668356423e-06, + "loss": 0.3088, + "step": 6972 + }, + { + "epoch": 0.15348888399735858, + "grad_norm": 1.7668195172579477, + "learning_rate": 9.605413887774695e-06, + "loss": 0.3253, + "step": 6973 + }, + { + "epoch": 0.15351089588377723, + "grad_norm": 2.790588784309047, + "learning_rate": 9.605275083786198e-06, + "loss": 0.4365, + "step": 6974 + }, + { + "epoch": 0.1535329077701959, + "grad_norm": 0.645662450622514, + "learning_rate": 9.605136256391641e-06, + "loss": 0.5711, + "step": 6975 + }, + { + "epoch": 0.15355491965661458, + "grad_norm": 1.9721554273602004, + "learning_rate": 9.604997405591729e-06, + "loss": 0.372, + "step": 6976 + }, + { + "epoch": 0.15357693154303323, + "grad_norm": 3.109860354886347, + "learning_rate": 9.604858531387165e-06, + "loss": 0.4622, + "step": 6977 + }, + { + "epoch": 0.1535989434294519, + "grad_norm": 4.0345510144197005, + "learning_rate": 9.604719633778654e-06, + "loss": 0.3558, + "step": 6978 + }, + { + "epoch": 0.15362095531587058, + "grad_norm": 2.2050935709479433, + "learning_rate": 9.604580712766907e-06, + "loss": 0.356, + "step": 6979 + }, + { + "epoch": 0.15364296720228923, + "grad_norm": 2.7881713745236976, + "learning_rate": 9.604441768352626e-06, + "loss": 0.3222, + "step": 6980 + }, + { + "epoch": 0.1536649790887079, + "grad_norm": 2.889819913580422, + "learning_rate": 9.604302800536518e-06, + "loss": 0.4191, + "step": 6981 + }, + { + "epoch": 0.15368699097512656, + "grad_norm": 3.0221188796638185, + "learning_rate": 9.604163809319289e-06, + "loss": 0.4398, + "step": 6982 + }, + { + "epoch": 0.15370900286154524, + "grad_norm": 1.4575093167378643, + "learning_rate": 9.604024794701648e-06, + "loss": 0.4147, + "step": 6983 + }, + { + "epoch": 0.1537310147479639, + "grad_norm": 1.423740211424239, + "learning_rate": 9.603885756684297e-06, + "loss": 0.3312, + "step": 6984 + }, + { + "epoch": 0.15375302663438256, + "grad_norm": 1.3776927229928768, + "learning_rate": 9.603746695267947e-06, + "loss": 0.3477, + "step": 6985 + }, + { + "epoch": 0.15377503852080124, + "grad_norm": 1.6917672442573837, + "learning_rate": 9.603607610453303e-06, + "loss": 0.4237, + "step": 6986 + }, + { + "epoch": 0.1537970504072199, + "grad_norm": 1.2749868924560421, + "learning_rate": 9.60346850224107e-06, + "loss": 0.3333, + "step": 6987 + }, + { + "epoch": 0.15381906229363856, + "grad_norm": 1.8990643050680511, + "learning_rate": 9.603329370631959e-06, + "loss": 0.3295, + "step": 6988 + }, + { + "epoch": 0.15384107418005724, + "grad_norm": 1.5932593064029217, + "learning_rate": 9.603190215626674e-06, + "loss": 0.3697, + "step": 6989 + }, + { + "epoch": 0.1538630860664759, + "grad_norm": 1.5451197754319281, + "learning_rate": 9.603051037225923e-06, + "loss": 0.3931, + "step": 6990 + }, + { + "epoch": 0.15388509795289457, + "grad_norm": 2.2438078972890927, + "learning_rate": 9.602911835430415e-06, + "loss": 0.3836, + "step": 6991 + }, + { + "epoch": 0.15390710983931322, + "grad_norm": 1.6861800152727686, + "learning_rate": 9.602772610240854e-06, + "loss": 0.3219, + "step": 6992 + }, + { + "epoch": 0.1539291217257319, + "grad_norm": 1.4094920991056994, + "learning_rate": 9.602633361657951e-06, + "loss": 0.3383, + "step": 6993 + }, + { + "epoch": 0.15395113361215057, + "grad_norm": 1.6405179533317449, + "learning_rate": 9.602494089682412e-06, + "loss": 0.365, + "step": 6994 + }, + { + "epoch": 0.15397314549856922, + "grad_norm": 1.7887946771092773, + "learning_rate": 9.602354794314947e-06, + "loss": 0.343, + "step": 6995 + }, + { + "epoch": 0.1539951573849879, + "grad_norm": 1.6280110600330635, + "learning_rate": 9.60221547555626e-06, + "loss": 0.3239, + "step": 6996 + }, + { + "epoch": 0.15401716927140657, + "grad_norm": 1.4008187849676133, + "learning_rate": 9.602076133407063e-06, + "loss": 0.3946, + "step": 6997 + }, + { + "epoch": 0.15403918115782522, + "grad_norm": 1.4985494760055147, + "learning_rate": 9.601936767868061e-06, + "loss": 0.327, + "step": 6998 + }, + { + "epoch": 0.1540611930442439, + "grad_norm": 2.8352405567899286, + "learning_rate": 9.601797378939964e-06, + "loss": 0.3879, + "step": 6999 + }, + { + "epoch": 0.15408320493066255, + "grad_norm": 2.0615296178188838, + "learning_rate": 9.601657966623482e-06, + "loss": 0.4272, + "step": 7000 + }, + { + "epoch": 0.15410521681708123, + "grad_norm": 1.310758359015201, + "learning_rate": 9.60151853091932e-06, + "loss": 0.2943, + "step": 7001 + }, + { + "epoch": 0.1541272287034999, + "grad_norm": 2.5457436644444136, + "learning_rate": 9.601379071828189e-06, + "loss": 0.3151, + "step": 7002 + }, + { + "epoch": 0.15414924058991855, + "grad_norm": 1.7629732439853472, + "learning_rate": 9.601239589350797e-06, + "loss": 0.2599, + "step": 7003 + }, + { + "epoch": 0.15417125247633723, + "grad_norm": 1.5468371989441034, + "learning_rate": 9.601100083487854e-06, + "loss": 0.3425, + "step": 7004 + }, + { + "epoch": 0.15419326436275588, + "grad_norm": 0.6910256671270218, + "learning_rate": 9.600960554240067e-06, + "loss": 0.5752, + "step": 7005 + }, + { + "epoch": 0.15421527624917455, + "grad_norm": 1.6242930572586152, + "learning_rate": 9.600821001608147e-06, + "loss": 0.4234, + "step": 7006 + }, + { + "epoch": 0.15423728813559323, + "grad_norm": 0.6655014399100104, + "learning_rate": 9.600681425592803e-06, + "loss": 0.5493, + "step": 7007 + }, + { + "epoch": 0.15425930002201188, + "grad_norm": 1.8460141292081094, + "learning_rate": 9.600541826194744e-06, + "loss": 0.4042, + "step": 7008 + }, + { + "epoch": 0.15428131190843056, + "grad_norm": 1.6115588509556638, + "learning_rate": 9.600402203414679e-06, + "loss": 0.321, + "step": 7009 + }, + { + "epoch": 0.1543033237948492, + "grad_norm": 1.5979302808098468, + "learning_rate": 9.60026255725332e-06, + "loss": 0.3677, + "step": 7010 + }, + { + "epoch": 0.15432533568126788, + "grad_norm": 1.5859868041273195, + "learning_rate": 9.600122887711372e-06, + "loss": 0.2707, + "step": 7011 + }, + { + "epoch": 0.15434734756768656, + "grad_norm": 1.9616154153031011, + "learning_rate": 9.599983194789548e-06, + "loss": 0.3246, + "step": 7012 + }, + { + "epoch": 0.1543693594541052, + "grad_norm": 0.6680581679949051, + "learning_rate": 9.59984347848856e-06, + "loss": 0.5385, + "step": 7013 + }, + { + "epoch": 0.1543913713405239, + "grad_norm": 1.5433238543092112, + "learning_rate": 9.599703738809113e-06, + "loss": 0.3088, + "step": 7014 + }, + { + "epoch": 0.15441338322694254, + "grad_norm": 2.6737074007576958, + "learning_rate": 9.599563975751924e-06, + "loss": 0.4595, + "step": 7015 + }, + { + "epoch": 0.1544353951133612, + "grad_norm": 1.3452678389114356, + "learning_rate": 9.599424189317697e-06, + "loss": 0.3325, + "step": 7016 + }, + { + "epoch": 0.1544574069997799, + "grad_norm": 3.8640385646631086, + "learning_rate": 9.599284379507145e-06, + "loss": 0.4246, + "step": 7017 + }, + { + "epoch": 0.15447941888619854, + "grad_norm": 1.7921217447670699, + "learning_rate": 9.599144546320977e-06, + "loss": 0.4875, + "step": 7018 + }, + { + "epoch": 0.15450143077261722, + "grad_norm": 1.8236267744359502, + "learning_rate": 9.599004689759908e-06, + "loss": 0.3712, + "step": 7019 + }, + { + "epoch": 0.1545234426590359, + "grad_norm": 1.9384073231386105, + "learning_rate": 9.598864809824644e-06, + "loss": 0.4271, + "step": 7020 + }, + { + "epoch": 0.15454545454545454, + "grad_norm": 0.6562273975986606, + "learning_rate": 9.5987249065159e-06, + "loss": 0.56, + "step": 7021 + }, + { + "epoch": 0.15456746643187322, + "grad_norm": 1.8363850332010112, + "learning_rate": 9.598584979834382e-06, + "loss": 0.33, + "step": 7022 + }, + { + "epoch": 0.15458947831829187, + "grad_norm": 0.65337336043412, + "learning_rate": 9.598445029780807e-06, + "loss": 0.5615, + "step": 7023 + }, + { + "epoch": 0.15461149020471054, + "grad_norm": 9.793945228225935, + "learning_rate": 9.598305056355882e-06, + "loss": 0.2825, + "step": 7024 + }, + { + "epoch": 0.15463350209112922, + "grad_norm": 1.7116087219620635, + "learning_rate": 9.59816505956032e-06, + "loss": 0.3951, + "step": 7025 + }, + { + "epoch": 0.15465551397754787, + "grad_norm": 1.5056310685206293, + "learning_rate": 9.598025039394832e-06, + "loss": 0.3579, + "step": 7026 + }, + { + "epoch": 0.15467752586396655, + "grad_norm": 1.9361606547635042, + "learning_rate": 9.597884995860129e-06, + "loss": 0.2847, + "step": 7027 + }, + { + "epoch": 0.1546995377503852, + "grad_norm": 2.194381877622716, + "learning_rate": 9.597744928956925e-06, + "loss": 0.4016, + "step": 7028 + }, + { + "epoch": 0.15472154963680387, + "grad_norm": 1.6163326133067584, + "learning_rate": 9.59760483868593e-06, + "loss": 0.3899, + "step": 7029 + }, + { + "epoch": 0.15474356152322255, + "grad_norm": 0.6657938196082028, + "learning_rate": 9.597464725047859e-06, + "loss": 0.581, + "step": 7030 + }, + { + "epoch": 0.1547655734096412, + "grad_norm": 1.950373252837081, + "learning_rate": 9.59732458804342e-06, + "loss": 0.3805, + "step": 7031 + }, + { + "epoch": 0.15478758529605988, + "grad_norm": 1.5274871488315207, + "learning_rate": 9.597184427673325e-06, + "loss": 0.3808, + "step": 7032 + }, + { + "epoch": 0.15480959718247853, + "grad_norm": 1.1838137574720504, + "learning_rate": 9.597044243938291e-06, + "loss": 0.3221, + "step": 7033 + }, + { + "epoch": 0.1548316090688972, + "grad_norm": 2.1332620794774653, + "learning_rate": 9.596904036839025e-06, + "loss": 0.3484, + "step": 7034 + }, + { + "epoch": 0.15485362095531588, + "grad_norm": 0.6554492743074503, + "learning_rate": 9.596763806376246e-06, + "loss": 0.5587, + "step": 7035 + }, + { + "epoch": 0.15487563284173453, + "grad_norm": 0.6582038985741264, + "learning_rate": 9.59662355255066e-06, + "loss": 0.5397, + "step": 7036 + }, + { + "epoch": 0.1548976447281532, + "grad_norm": 1.8488638517884344, + "learning_rate": 9.596483275362985e-06, + "loss": 0.3433, + "step": 7037 + }, + { + "epoch": 0.15491965661457188, + "grad_norm": 1.3661995850742532, + "learning_rate": 9.596342974813932e-06, + "loss": 0.3909, + "step": 7038 + }, + { + "epoch": 0.15494166850099053, + "grad_norm": 1.8920612091872122, + "learning_rate": 9.596202650904213e-06, + "loss": 0.2687, + "step": 7039 + }, + { + "epoch": 0.1549636803874092, + "grad_norm": 1.445560464685703, + "learning_rate": 9.596062303634543e-06, + "loss": 0.4039, + "step": 7040 + }, + { + "epoch": 0.15498569227382786, + "grad_norm": 1.3251512526741391, + "learning_rate": 9.595921933005632e-06, + "loss": 0.3874, + "step": 7041 + }, + { + "epoch": 0.15500770416024653, + "grad_norm": 1.7701775092561827, + "learning_rate": 9.595781539018198e-06, + "loss": 0.3777, + "step": 7042 + }, + { + "epoch": 0.1550297160466652, + "grad_norm": 1.269602965685289, + "learning_rate": 9.595641121672955e-06, + "loss": 0.2738, + "step": 7043 + }, + { + "epoch": 0.15505172793308386, + "grad_norm": 0.6572500837502395, + "learning_rate": 9.59550068097061e-06, + "loss": 0.5621, + "step": 7044 + }, + { + "epoch": 0.15507373981950254, + "grad_norm": 1.6169049559712345, + "learning_rate": 9.595360216911882e-06, + "loss": 0.3273, + "step": 7045 + }, + { + "epoch": 0.1550957517059212, + "grad_norm": 0.6303826689971919, + "learning_rate": 9.595219729497484e-06, + "loss": 0.5395, + "step": 7046 + }, + { + "epoch": 0.15511776359233986, + "grad_norm": 0.6435283916772191, + "learning_rate": 9.595079218728129e-06, + "loss": 0.5601, + "step": 7047 + }, + { + "epoch": 0.15513977547875854, + "grad_norm": 1.1881052548136901, + "learning_rate": 9.594938684604533e-06, + "loss": 0.3135, + "step": 7048 + }, + { + "epoch": 0.1551617873651772, + "grad_norm": 2.338718080380448, + "learning_rate": 9.594798127127409e-06, + "loss": 0.3026, + "step": 7049 + }, + { + "epoch": 0.15518379925159587, + "grad_norm": 1.6307510134778733, + "learning_rate": 9.594657546297473e-06, + "loss": 0.3133, + "step": 7050 + }, + { + "epoch": 0.15520581113801452, + "grad_norm": 1.7678043526815208, + "learning_rate": 9.594516942115436e-06, + "loss": 0.3116, + "step": 7051 + }, + { + "epoch": 0.1552278230244332, + "grad_norm": 4.271767594161978, + "learning_rate": 9.594376314582016e-06, + "loss": 0.3986, + "step": 7052 + }, + { + "epoch": 0.15524983491085187, + "grad_norm": 1.5565668658700338, + "learning_rate": 9.594235663697925e-06, + "loss": 0.3421, + "step": 7053 + }, + { + "epoch": 0.15527184679727052, + "grad_norm": 1.589266359675178, + "learning_rate": 9.59409498946388e-06, + "loss": 0.4876, + "step": 7054 + }, + { + "epoch": 0.1552938586836892, + "grad_norm": 1.5167028262962325, + "learning_rate": 9.593954291880597e-06, + "loss": 0.3102, + "step": 7055 + }, + { + "epoch": 0.15531587057010784, + "grad_norm": 1.2545788130094755, + "learning_rate": 9.593813570948787e-06, + "loss": 0.2845, + "step": 7056 + }, + { + "epoch": 0.15533788245652652, + "grad_norm": 0.6940749602091426, + "learning_rate": 9.593672826669167e-06, + "loss": 0.5689, + "step": 7057 + }, + { + "epoch": 0.1553598943429452, + "grad_norm": 1.8407337056973405, + "learning_rate": 9.593532059042453e-06, + "loss": 0.296, + "step": 7058 + }, + { + "epoch": 0.15538190622936385, + "grad_norm": 1.3198649426215412, + "learning_rate": 9.593391268069362e-06, + "loss": 0.2918, + "step": 7059 + }, + { + "epoch": 0.15540391811578252, + "grad_norm": 0.6440281442526453, + "learning_rate": 9.593250453750608e-06, + "loss": 0.5488, + "step": 7060 + }, + { + "epoch": 0.1554259300022012, + "grad_norm": 1.6699130357205725, + "learning_rate": 9.593109616086906e-06, + "loss": 0.258, + "step": 7061 + }, + { + "epoch": 0.15544794188861985, + "grad_norm": 1.7167884783652827, + "learning_rate": 9.592968755078971e-06, + "loss": 0.4501, + "step": 7062 + }, + { + "epoch": 0.15546995377503853, + "grad_norm": 1.688396572404498, + "learning_rate": 9.592827870727521e-06, + "loss": 0.3245, + "step": 7063 + }, + { + "epoch": 0.15549196566145718, + "grad_norm": 2.1902478734476962, + "learning_rate": 9.592686963033273e-06, + "loss": 0.3704, + "step": 7064 + }, + { + "epoch": 0.15551397754787585, + "grad_norm": 1.5103457331091081, + "learning_rate": 9.59254603199694e-06, + "loss": 0.4352, + "step": 7065 + }, + { + "epoch": 0.15553598943429453, + "grad_norm": 0.6365234739128237, + "learning_rate": 9.59240507761924e-06, + "loss": 0.5056, + "step": 7066 + }, + { + "epoch": 0.15555800132071318, + "grad_norm": 3.0526449385449075, + "learning_rate": 9.592264099900888e-06, + "loss": 0.3332, + "step": 7067 + }, + { + "epoch": 0.15558001320713186, + "grad_norm": 1.597771776832449, + "learning_rate": 9.592123098842603e-06, + "loss": 0.3385, + "step": 7068 + }, + { + "epoch": 0.1556020250935505, + "grad_norm": 0.6421899515532054, + "learning_rate": 9.5919820744451e-06, + "loss": 0.5561, + "step": 7069 + }, + { + "epoch": 0.15562403697996918, + "grad_norm": 2.1028058851377254, + "learning_rate": 9.591841026709096e-06, + "loss": 0.3707, + "step": 7070 + }, + { + "epoch": 0.15564604886638786, + "grad_norm": 1.2406910143907002, + "learning_rate": 9.591699955635308e-06, + "loss": 0.3693, + "step": 7071 + }, + { + "epoch": 0.1556680607528065, + "grad_norm": 1.331171634821467, + "learning_rate": 9.591558861224452e-06, + "loss": 0.3752, + "step": 7072 + }, + { + "epoch": 0.15569007263922519, + "grad_norm": 1.4481930036431863, + "learning_rate": 9.591417743477248e-06, + "loss": 0.4343, + "step": 7073 + }, + { + "epoch": 0.15571208452564383, + "grad_norm": 1.2865458377573227, + "learning_rate": 9.591276602394408e-06, + "loss": 0.414, + "step": 7074 + }, + { + "epoch": 0.1557340964120625, + "grad_norm": 1.372722354730096, + "learning_rate": 9.591135437976655e-06, + "loss": 0.3168, + "step": 7075 + }, + { + "epoch": 0.1557561082984812, + "grad_norm": 0.6546175422873893, + "learning_rate": 9.590994250224704e-06, + "loss": 0.571, + "step": 7076 + }, + { + "epoch": 0.15577812018489984, + "grad_norm": 8.109849127748285, + "learning_rate": 9.590853039139271e-06, + "loss": 0.3482, + "step": 7077 + }, + { + "epoch": 0.15580013207131851, + "grad_norm": 1.336032761565594, + "learning_rate": 9.590711804721076e-06, + "loss": 0.2855, + "step": 7078 + }, + { + "epoch": 0.1558221439577372, + "grad_norm": 1.1884462391702664, + "learning_rate": 9.590570546970837e-06, + "loss": 0.2839, + "step": 7079 + }, + { + "epoch": 0.15584415584415584, + "grad_norm": 0.6259618206986509, + "learning_rate": 9.59042926588927e-06, + "loss": 0.5377, + "step": 7080 + }, + { + "epoch": 0.15586616773057452, + "grad_norm": 0.6111220711535521, + "learning_rate": 9.590287961477093e-06, + "loss": 0.5608, + "step": 7081 + }, + { + "epoch": 0.15588817961699317, + "grad_norm": 1.2811431841492202, + "learning_rate": 9.590146633735027e-06, + "loss": 0.3755, + "step": 7082 + }, + { + "epoch": 0.15591019150341184, + "grad_norm": 1.9391392365534863, + "learning_rate": 9.59000528266379e-06, + "loss": 0.3479, + "step": 7083 + }, + { + "epoch": 0.15593220338983052, + "grad_norm": 1.6530802283653483, + "learning_rate": 9.589863908264097e-06, + "loss": 0.3908, + "step": 7084 + }, + { + "epoch": 0.15595421527624917, + "grad_norm": 1.6104111874747444, + "learning_rate": 9.589722510536667e-06, + "loss": 0.3754, + "step": 7085 + }, + { + "epoch": 0.15597622716266785, + "grad_norm": 1.2468838061756111, + "learning_rate": 9.589581089482223e-06, + "loss": 0.3509, + "step": 7086 + }, + { + "epoch": 0.1559982390490865, + "grad_norm": 1.40081645792867, + "learning_rate": 9.589439645101479e-06, + "loss": 0.3825, + "step": 7087 + }, + { + "epoch": 0.15602025093550517, + "grad_norm": 1.7027871067593214, + "learning_rate": 9.589298177395156e-06, + "loss": 0.4634, + "step": 7088 + }, + { + "epoch": 0.15604226282192385, + "grad_norm": 0.668067968806326, + "learning_rate": 9.589156686363975e-06, + "loss": 0.5356, + "step": 7089 + }, + { + "epoch": 0.1560642747083425, + "grad_norm": 1.3481567503071559, + "learning_rate": 9.589015172008651e-06, + "loss": 0.3917, + "step": 7090 + }, + { + "epoch": 0.15608628659476118, + "grad_norm": 1.7008980323841363, + "learning_rate": 9.588873634329905e-06, + "loss": 0.3356, + "step": 7091 + }, + { + "epoch": 0.15610829848117982, + "grad_norm": 1.470225940437892, + "learning_rate": 9.588732073328456e-06, + "loss": 0.2961, + "step": 7092 + }, + { + "epoch": 0.1561303103675985, + "grad_norm": 1.3872991769135528, + "learning_rate": 9.588590489005025e-06, + "loss": 0.4543, + "step": 7093 + }, + { + "epoch": 0.15615232225401718, + "grad_norm": 1.22659413898353, + "learning_rate": 9.58844888136033e-06, + "loss": 0.3077, + "step": 7094 + }, + { + "epoch": 0.15617433414043583, + "grad_norm": 2.848597763170482, + "learning_rate": 9.588307250395093e-06, + "loss": 0.374, + "step": 7095 + }, + { + "epoch": 0.1561963460268545, + "grad_norm": 1.8701172425118082, + "learning_rate": 9.588165596110031e-06, + "loss": 0.3932, + "step": 7096 + }, + { + "epoch": 0.15621835791327315, + "grad_norm": 5.0887959281012325, + "learning_rate": 9.588023918505865e-06, + "loss": 0.4676, + "step": 7097 + }, + { + "epoch": 0.15624036979969183, + "grad_norm": 1.687274091805599, + "learning_rate": 9.587882217583314e-06, + "loss": 0.4386, + "step": 7098 + }, + { + "epoch": 0.1562623816861105, + "grad_norm": 1.7933198003122515, + "learning_rate": 9.587740493343099e-06, + "loss": 0.3215, + "step": 7099 + }, + { + "epoch": 0.15628439357252916, + "grad_norm": 2.499300463553214, + "learning_rate": 9.587598745785943e-06, + "loss": 0.3418, + "step": 7100 + }, + { + "epoch": 0.15630640545894783, + "grad_norm": 1.7177748256533958, + "learning_rate": 9.587456974912563e-06, + "loss": 0.3722, + "step": 7101 + }, + { + "epoch": 0.1563284173453665, + "grad_norm": 0.6591776705295134, + "learning_rate": 9.58731518072368e-06, + "loss": 0.5138, + "step": 7102 + }, + { + "epoch": 0.15635042923178516, + "grad_norm": 1.358607473624912, + "learning_rate": 9.587173363220015e-06, + "loss": 0.3961, + "step": 7103 + }, + { + "epoch": 0.15637244111820384, + "grad_norm": 1.5072638507675564, + "learning_rate": 9.58703152240229e-06, + "loss": 0.3506, + "step": 7104 + }, + { + "epoch": 0.15639445300462249, + "grad_norm": 1.6404405242310123, + "learning_rate": 9.586889658271226e-06, + "loss": 0.3876, + "step": 7105 + }, + { + "epoch": 0.15641646489104116, + "grad_norm": 1.5002517496164682, + "learning_rate": 9.58674777082754e-06, + "loss": 0.351, + "step": 7106 + }, + { + "epoch": 0.15643847677745984, + "grad_norm": 1.4536393063215196, + "learning_rate": 9.586605860071958e-06, + "loss": 0.3669, + "step": 7107 + }, + { + "epoch": 0.1564604886638785, + "grad_norm": 1.3869737879635555, + "learning_rate": 9.5864639260052e-06, + "loss": 0.3652, + "step": 7108 + }, + { + "epoch": 0.15648250055029717, + "grad_norm": 1.3251195076662512, + "learning_rate": 9.586321968627985e-06, + "loss": 0.3047, + "step": 7109 + }, + { + "epoch": 0.15650451243671581, + "grad_norm": 0.6331090720804097, + "learning_rate": 9.586179987941038e-06, + "loss": 0.5226, + "step": 7110 + }, + { + "epoch": 0.1565265243231345, + "grad_norm": 1.7847989173881835, + "learning_rate": 9.586037983945077e-06, + "loss": 0.3699, + "step": 7111 + }, + { + "epoch": 0.15654853620955317, + "grad_norm": 1.4482657372719985, + "learning_rate": 9.585895956640825e-06, + "loss": 0.3516, + "step": 7112 + }, + { + "epoch": 0.15657054809597182, + "grad_norm": 1.5446310359896378, + "learning_rate": 9.585753906029008e-06, + "loss": 0.3189, + "step": 7113 + }, + { + "epoch": 0.1565925599823905, + "grad_norm": 1.700809194655614, + "learning_rate": 9.585611832110341e-06, + "loss": 0.3348, + "step": 7114 + }, + { + "epoch": 0.15661457186880914, + "grad_norm": 3.8880556411716456, + "learning_rate": 9.585469734885551e-06, + "loss": 0.4243, + "step": 7115 + }, + { + "epoch": 0.15663658375522782, + "grad_norm": 1.5930009218803909, + "learning_rate": 9.585327614355358e-06, + "loss": 0.448, + "step": 7116 + }, + { + "epoch": 0.1566585956416465, + "grad_norm": 1.2834603062655716, + "learning_rate": 9.585185470520483e-06, + "loss": 0.3042, + "step": 7117 + }, + { + "epoch": 0.15668060752806515, + "grad_norm": 0.658836807135423, + "learning_rate": 9.585043303381653e-06, + "loss": 0.5435, + "step": 7118 + }, + { + "epoch": 0.15670261941448382, + "grad_norm": 21.37576472043176, + "learning_rate": 9.584901112939588e-06, + "loss": 0.3103, + "step": 7119 + }, + { + "epoch": 0.1567246313009025, + "grad_norm": 1.201364617531798, + "learning_rate": 9.58475889919501e-06, + "loss": 0.343, + "step": 7120 + }, + { + "epoch": 0.15674664318732115, + "grad_norm": 1.6693157380891874, + "learning_rate": 9.584616662148642e-06, + "loss": 0.395, + "step": 7121 + }, + { + "epoch": 0.15676865507373983, + "grad_norm": 1.6673606066682773, + "learning_rate": 9.584474401801209e-06, + "loss": 0.334, + "step": 7122 + }, + { + "epoch": 0.15679066696015848, + "grad_norm": 1.6035812421820828, + "learning_rate": 9.58433211815343e-06, + "loss": 0.3374, + "step": 7123 + }, + { + "epoch": 0.15681267884657715, + "grad_norm": 2.0587567748561653, + "learning_rate": 9.584189811206032e-06, + "loss": 0.2379, + "step": 7124 + }, + { + "epoch": 0.15683469073299583, + "grad_norm": 1.8753324671553804, + "learning_rate": 9.584047480959734e-06, + "loss": 0.364, + "step": 7125 + }, + { + "epoch": 0.15685670261941448, + "grad_norm": 1.9240692373128119, + "learning_rate": 9.583905127415267e-06, + "loss": 0.4602, + "step": 7126 + }, + { + "epoch": 0.15687871450583316, + "grad_norm": 1.4869496931258324, + "learning_rate": 9.583762750573346e-06, + "loss": 0.3699, + "step": 7127 + }, + { + "epoch": 0.1569007263922518, + "grad_norm": 1.6043497266116473, + "learning_rate": 9.5836203504347e-06, + "loss": 0.3598, + "step": 7128 + }, + { + "epoch": 0.15692273827867048, + "grad_norm": 1.9285398111039678, + "learning_rate": 9.583477927000052e-06, + "loss": 0.4341, + "step": 7129 + }, + { + "epoch": 0.15694475016508916, + "grad_norm": 1.6881606174943786, + "learning_rate": 9.583335480270122e-06, + "loss": 0.3239, + "step": 7130 + }, + { + "epoch": 0.1569667620515078, + "grad_norm": 2.0959120900668515, + "learning_rate": 9.583193010245639e-06, + "loss": 0.3332, + "step": 7131 + }, + { + "epoch": 0.15698877393792648, + "grad_norm": 1.6291753617474514, + "learning_rate": 9.583050516927324e-06, + "loss": 0.3246, + "step": 7132 + }, + { + "epoch": 0.15701078582434513, + "grad_norm": 1.7886387045371666, + "learning_rate": 9.582908000315902e-06, + "loss": 0.3661, + "step": 7133 + }, + { + "epoch": 0.1570327977107638, + "grad_norm": 1.3491288972363344, + "learning_rate": 9.582765460412097e-06, + "loss": 0.3158, + "step": 7134 + }, + { + "epoch": 0.1570548095971825, + "grad_norm": 6.838673733624926, + "learning_rate": 9.582622897216636e-06, + "loss": 0.4038, + "step": 7135 + }, + { + "epoch": 0.15707682148360114, + "grad_norm": 0.7049719000732141, + "learning_rate": 9.58248031073024e-06, + "loss": 0.5391, + "step": 7136 + }, + { + "epoch": 0.1570988333700198, + "grad_norm": 1.4701190382513838, + "learning_rate": 9.582337700953636e-06, + "loss": 0.3521, + "step": 7137 + }, + { + "epoch": 0.1571208452564385, + "grad_norm": 1.4207899111760824, + "learning_rate": 9.582195067887547e-06, + "loss": 0.4256, + "step": 7138 + }, + { + "epoch": 0.15714285714285714, + "grad_norm": 1.3073849765297083, + "learning_rate": 9.5820524115327e-06, + "loss": 0.3094, + "step": 7139 + }, + { + "epoch": 0.15716486902927582, + "grad_norm": 1.535346880645884, + "learning_rate": 9.581909731889818e-06, + "loss": 0.4314, + "step": 7140 + }, + { + "epoch": 0.15718688091569447, + "grad_norm": 0.6404858793459877, + "learning_rate": 9.581767028959628e-06, + "loss": 0.5609, + "step": 7141 + }, + { + "epoch": 0.15720889280211314, + "grad_norm": 1.5475050955003389, + "learning_rate": 9.581624302742854e-06, + "loss": 0.4327, + "step": 7142 + }, + { + "epoch": 0.15723090468853182, + "grad_norm": 1.8655921089462368, + "learning_rate": 9.581481553240221e-06, + "loss": 0.3727, + "step": 7143 + }, + { + "epoch": 0.15725291657495047, + "grad_norm": 1.4832306208956807, + "learning_rate": 9.581338780452458e-06, + "loss": 0.3743, + "step": 7144 + }, + { + "epoch": 0.15727492846136915, + "grad_norm": 1.67213670722387, + "learning_rate": 9.581195984380285e-06, + "loss": 0.2542, + "step": 7145 + }, + { + "epoch": 0.1572969403477878, + "grad_norm": 0.6392071081060714, + "learning_rate": 9.581053165024433e-06, + "loss": 0.5479, + "step": 7146 + }, + { + "epoch": 0.15731895223420647, + "grad_norm": 6.761075195589089, + "learning_rate": 9.580910322385624e-06, + "loss": 0.3538, + "step": 7147 + }, + { + "epoch": 0.15734096412062515, + "grad_norm": 1.5187016604440866, + "learning_rate": 9.580767456464586e-06, + "loss": 0.3864, + "step": 7148 + }, + { + "epoch": 0.1573629760070438, + "grad_norm": 5.53548503755474, + "learning_rate": 9.580624567262046e-06, + "loss": 0.3838, + "step": 7149 + }, + { + "epoch": 0.15738498789346247, + "grad_norm": 1.3526430359997579, + "learning_rate": 9.580481654778729e-06, + "loss": 0.3694, + "step": 7150 + }, + { + "epoch": 0.15740699977988112, + "grad_norm": 1.788671362366797, + "learning_rate": 9.580338719015358e-06, + "loss": 0.3786, + "step": 7151 + }, + { + "epoch": 0.1574290116662998, + "grad_norm": 4.900925391819355, + "learning_rate": 9.580195759972666e-06, + "loss": 0.2628, + "step": 7152 + }, + { + "epoch": 0.15745102355271848, + "grad_norm": 1.3457194961289103, + "learning_rate": 9.580052777651374e-06, + "loss": 0.4024, + "step": 7153 + }, + { + "epoch": 0.15747303543913713, + "grad_norm": 0.6372673647336543, + "learning_rate": 9.579909772052214e-06, + "loss": 0.5304, + "step": 7154 + }, + { + "epoch": 0.1574950473255558, + "grad_norm": 3.385503671300908, + "learning_rate": 9.579766743175906e-06, + "loss": 0.3893, + "step": 7155 + }, + { + "epoch": 0.15751705921197445, + "grad_norm": 1.6209314082957418, + "learning_rate": 9.579623691023184e-06, + "loss": 0.359, + "step": 7156 + }, + { + "epoch": 0.15753907109839313, + "grad_norm": 1.458960816613384, + "learning_rate": 9.579480615594771e-06, + "loss": 0.2964, + "step": 7157 + }, + { + "epoch": 0.1575610829848118, + "grad_norm": 1.8353814047186054, + "learning_rate": 9.579337516891393e-06, + "loss": 0.3912, + "step": 7158 + }, + { + "epoch": 0.15758309487123046, + "grad_norm": 0.6613982609467006, + "learning_rate": 9.579194394913782e-06, + "loss": 0.557, + "step": 7159 + }, + { + "epoch": 0.15760510675764913, + "grad_norm": 0.641378090174485, + "learning_rate": 9.579051249662664e-06, + "loss": 0.5514, + "step": 7160 + }, + { + "epoch": 0.1576271186440678, + "grad_norm": 2.1671102681577237, + "learning_rate": 9.57890808113876e-06, + "loss": 0.3438, + "step": 7161 + }, + { + "epoch": 0.15764913053048646, + "grad_norm": 1.9639374570786323, + "learning_rate": 9.578764889342806e-06, + "loss": 0.2841, + "step": 7162 + }, + { + "epoch": 0.15767114241690514, + "grad_norm": 1.8675838718243902, + "learning_rate": 9.578621674275529e-06, + "loss": 0.3126, + "step": 7163 + }, + { + "epoch": 0.15769315430332378, + "grad_norm": 1.3831229150687254, + "learning_rate": 9.57847843593765e-06, + "loss": 0.3285, + "step": 7164 + }, + { + "epoch": 0.15771516618974246, + "grad_norm": 1.6300695147659312, + "learning_rate": 9.578335174329904e-06, + "loss": 0.3344, + "step": 7165 + }, + { + "epoch": 0.15773717807616114, + "grad_norm": 1.4992710441192016, + "learning_rate": 9.578191889453018e-06, + "loss": 0.3695, + "step": 7166 + }, + { + "epoch": 0.1577591899625798, + "grad_norm": 1.6393792419516453, + "learning_rate": 9.578048581307716e-06, + "loss": 0.3888, + "step": 7167 + }, + { + "epoch": 0.15778120184899846, + "grad_norm": 11.997777053614977, + "learning_rate": 9.57790524989473e-06, + "loss": 0.3017, + "step": 7168 + }, + { + "epoch": 0.1578032137354171, + "grad_norm": 2.398110670511312, + "learning_rate": 9.577761895214788e-06, + "loss": 0.3789, + "step": 7169 + }, + { + "epoch": 0.1578252256218358, + "grad_norm": 2.0735299966770175, + "learning_rate": 9.577618517268618e-06, + "loss": 0.333, + "step": 7170 + }, + { + "epoch": 0.15784723750825447, + "grad_norm": 1.7684954505503474, + "learning_rate": 9.577475116056949e-06, + "loss": 0.3885, + "step": 7171 + }, + { + "epoch": 0.15786924939467312, + "grad_norm": 1.598392235316388, + "learning_rate": 9.577331691580511e-06, + "loss": 0.3821, + "step": 7172 + }, + { + "epoch": 0.1578912612810918, + "grad_norm": 0.7648801909194102, + "learning_rate": 9.57718824384003e-06, + "loss": 0.519, + "step": 7173 + }, + { + "epoch": 0.15791327316751044, + "grad_norm": 1.3572828943963238, + "learning_rate": 9.57704477283624e-06, + "loss": 0.2622, + "step": 7174 + }, + { + "epoch": 0.15793528505392912, + "grad_norm": 2.361076894500765, + "learning_rate": 9.576901278569864e-06, + "loss": 0.3251, + "step": 7175 + }, + { + "epoch": 0.1579572969403478, + "grad_norm": 1.9373523457036625, + "learning_rate": 9.576757761041636e-06, + "loss": 0.3752, + "step": 7176 + }, + { + "epoch": 0.15797930882676645, + "grad_norm": 2.034763722238881, + "learning_rate": 9.576614220252282e-06, + "loss": 0.3767, + "step": 7177 + }, + { + "epoch": 0.15800132071318512, + "grad_norm": 2.770067555278954, + "learning_rate": 9.576470656202535e-06, + "loss": 0.2844, + "step": 7178 + }, + { + "epoch": 0.1580233325996038, + "grad_norm": 11.172744786040223, + "learning_rate": 9.57632706889312e-06, + "loss": 0.3528, + "step": 7179 + }, + { + "epoch": 0.15804534448602245, + "grad_norm": 1.841944668570989, + "learning_rate": 9.576183458324774e-06, + "loss": 0.4324, + "step": 7180 + }, + { + "epoch": 0.15806735637244113, + "grad_norm": 1.9241104071153654, + "learning_rate": 9.57603982449822e-06, + "loss": 0.4243, + "step": 7181 + }, + { + "epoch": 0.15808936825885977, + "grad_norm": 1.5196612144593453, + "learning_rate": 9.575896167414192e-06, + "loss": 0.2861, + "step": 7182 + }, + { + "epoch": 0.15811138014527845, + "grad_norm": 1.505177859984505, + "learning_rate": 9.575752487073417e-06, + "loss": 0.3336, + "step": 7183 + }, + { + "epoch": 0.15813339203169713, + "grad_norm": 1.8883952912117699, + "learning_rate": 9.575608783476628e-06, + "loss": 0.3644, + "step": 7184 + }, + { + "epoch": 0.15815540391811578, + "grad_norm": 2.1729605524185205, + "learning_rate": 9.575465056624555e-06, + "loss": 0.4291, + "step": 7185 + }, + { + "epoch": 0.15817741580453445, + "grad_norm": 1.6234709133716325, + "learning_rate": 9.575321306517929e-06, + "loss": 0.4375, + "step": 7186 + }, + { + "epoch": 0.1581994276909531, + "grad_norm": 2.5318189420644646, + "learning_rate": 9.575177533157475e-06, + "loss": 0.3364, + "step": 7187 + }, + { + "epoch": 0.15822143957737178, + "grad_norm": 1.6750598116305224, + "learning_rate": 9.575033736543933e-06, + "loss": 0.3807, + "step": 7188 + }, + { + "epoch": 0.15824345146379046, + "grad_norm": 1.9305689262751293, + "learning_rate": 9.574889916678028e-06, + "loss": 0.3915, + "step": 7189 + }, + { + "epoch": 0.1582654633502091, + "grad_norm": 6.755918391263008, + "learning_rate": 9.57474607356049e-06, + "loss": 0.3963, + "step": 7190 + }, + { + "epoch": 0.15828747523662778, + "grad_norm": 1.4921236429261948, + "learning_rate": 9.574602207192052e-06, + "loss": 0.3456, + "step": 7191 + }, + { + "epoch": 0.15830948712304643, + "grad_norm": 1.436705467911802, + "learning_rate": 9.574458317573448e-06, + "loss": 0.3144, + "step": 7192 + }, + { + "epoch": 0.1583314990094651, + "grad_norm": 0.7088768758326478, + "learning_rate": 9.574314404705404e-06, + "loss": 0.5628, + "step": 7193 + }, + { + "epoch": 0.1583535108958838, + "grad_norm": 2.1131304133074673, + "learning_rate": 9.574170468588655e-06, + "loss": 0.4308, + "step": 7194 + }, + { + "epoch": 0.15837552278230244, + "grad_norm": 1.258935428718504, + "learning_rate": 9.574026509223932e-06, + "loss": 0.2745, + "step": 7195 + }, + { + "epoch": 0.1583975346687211, + "grad_norm": 1.2344057964207238, + "learning_rate": 9.573882526611965e-06, + "loss": 0.3406, + "step": 7196 + }, + { + "epoch": 0.15841954655513976, + "grad_norm": 1.3151751686929551, + "learning_rate": 9.573738520753488e-06, + "loss": 0.2838, + "step": 7197 + }, + { + "epoch": 0.15844155844155844, + "grad_norm": 1.570620500695371, + "learning_rate": 9.573594491649232e-06, + "loss": 0.2617, + "step": 7198 + }, + { + "epoch": 0.15846357032797712, + "grad_norm": 1.315793422949603, + "learning_rate": 9.573450439299927e-06, + "loss": 0.3329, + "step": 7199 + }, + { + "epoch": 0.15848558221439576, + "grad_norm": 5.8190880431899625, + "learning_rate": 9.57330636370631e-06, + "loss": 0.3097, + "step": 7200 + }, + { + "epoch": 0.15850759410081444, + "grad_norm": 0.6515433351491134, + "learning_rate": 9.573162264869108e-06, + "loss": 0.5293, + "step": 7201 + }, + { + "epoch": 0.15852960598723312, + "grad_norm": 1.766569526108004, + "learning_rate": 9.573018142789057e-06, + "loss": 0.3513, + "step": 7202 + }, + { + "epoch": 0.15855161787365177, + "grad_norm": 2.160295329724734, + "learning_rate": 9.572873997466887e-06, + "loss": 0.3108, + "step": 7203 + }, + { + "epoch": 0.15857362976007044, + "grad_norm": 1.7118232606838528, + "learning_rate": 9.572729828903334e-06, + "loss": 0.4565, + "step": 7204 + }, + { + "epoch": 0.1585956416464891, + "grad_norm": 2.0475415735881617, + "learning_rate": 9.572585637099126e-06, + "loss": 0.371, + "step": 7205 + }, + { + "epoch": 0.15861765353290777, + "grad_norm": 0.6506315595429483, + "learning_rate": 9.572441422054998e-06, + "loss": 0.5345, + "step": 7206 + }, + { + "epoch": 0.15863966541932645, + "grad_norm": 0.6284131891152606, + "learning_rate": 9.572297183771685e-06, + "loss": 0.5408, + "step": 7207 + }, + { + "epoch": 0.1586616773057451, + "grad_norm": 1.3013261159012404, + "learning_rate": 9.572152922249918e-06, + "loss": 0.3422, + "step": 7208 + }, + { + "epoch": 0.15868368919216377, + "grad_norm": 2.4449278166744097, + "learning_rate": 9.572008637490431e-06, + "loss": 0.4551, + "step": 7209 + }, + { + "epoch": 0.15870570107858242, + "grad_norm": 1.9705326372954521, + "learning_rate": 9.571864329493957e-06, + "loss": 0.3706, + "step": 7210 + }, + { + "epoch": 0.1587277129650011, + "grad_norm": 3.2778478739936183, + "learning_rate": 9.57171999826123e-06, + "loss": 0.4239, + "step": 7211 + }, + { + "epoch": 0.15874972485141978, + "grad_norm": 1.4762722650560975, + "learning_rate": 9.57157564379298e-06, + "loss": 0.3396, + "step": 7212 + }, + { + "epoch": 0.15877173673783843, + "grad_norm": 0.701065939855173, + "learning_rate": 9.571431266089945e-06, + "loss": 0.5622, + "step": 7213 + }, + { + "epoch": 0.1587937486242571, + "grad_norm": 0.6740955627951324, + "learning_rate": 9.57128686515286e-06, + "loss": 0.522, + "step": 7214 + }, + { + "epoch": 0.15881576051067575, + "grad_norm": 1.337848344655196, + "learning_rate": 9.571142440982453e-06, + "loss": 0.3704, + "step": 7215 + }, + { + "epoch": 0.15883777239709443, + "grad_norm": 5.282068089197605, + "learning_rate": 9.570997993579463e-06, + "loss": 0.3603, + "step": 7216 + }, + { + "epoch": 0.1588597842835131, + "grad_norm": 2.0161043227384274, + "learning_rate": 9.570853522944621e-06, + "loss": 0.4475, + "step": 7217 + }, + { + "epoch": 0.15888179616993175, + "grad_norm": 1.4995647651108754, + "learning_rate": 9.570709029078664e-06, + "loss": 0.326, + "step": 7218 + }, + { + "epoch": 0.15890380805635043, + "grad_norm": 1.4571441404533751, + "learning_rate": 9.570564511982325e-06, + "loss": 0.409, + "step": 7219 + }, + { + "epoch": 0.1589258199427691, + "grad_norm": 0.7425539491831783, + "learning_rate": 9.57041997165634e-06, + "loss": 0.5528, + "step": 7220 + }, + { + "epoch": 0.15894783182918776, + "grad_norm": 1.4393595679385096, + "learning_rate": 9.57027540810144e-06, + "loss": 0.3793, + "step": 7221 + }, + { + "epoch": 0.15896984371560643, + "grad_norm": 1.5355498545733999, + "learning_rate": 9.570130821318362e-06, + "loss": 0.3953, + "step": 7222 + }, + { + "epoch": 0.15899185560202508, + "grad_norm": 1.384162494195069, + "learning_rate": 9.569986211307842e-06, + "loss": 0.382, + "step": 7223 + }, + { + "epoch": 0.15901386748844376, + "grad_norm": 2.1651712855421357, + "learning_rate": 9.569841578070613e-06, + "loss": 0.4223, + "step": 7224 + }, + { + "epoch": 0.15903587937486244, + "grad_norm": 2.4020623332102495, + "learning_rate": 9.569696921607412e-06, + "loss": 0.3575, + "step": 7225 + }, + { + "epoch": 0.1590578912612811, + "grad_norm": 1.5627018334896121, + "learning_rate": 9.569552241918972e-06, + "loss": 0.3603, + "step": 7226 + }, + { + "epoch": 0.15907990314769976, + "grad_norm": 1.9132787848106423, + "learning_rate": 9.56940753900603e-06, + "loss": 0.3261, + "step": 7227 + }, + { + "epoch": 0.1591019150341184, + "grad_norm": 1.5802488807924002, + "learning_rate": 9.56926281286932e-06, + "loss": 0.4492, + "step": 7228 + }, + { + "epoch": 0.1591239269205371, + "grad_norm": 1.6498968387317896, + "learning_rate": 9.569118063509579e-06, + "loss": 0.3244, + "step": 7229 + }, + { + "epoch": 0.15914593880695577, + "grad_norm": 1.4331336927048166, + "learning_rate": 9.568973290927542e-06, + "loss": 0.299, + "step": 7230 + }, + { + "epoch": 0.15916795069337442, + "grad_norm": 0.7424465479051745, + "learning_rate": 9.568828495123945e-06, + "loss": 0.5402, + "step": 7231 + }, + { + "epoch": 0.1591899625797931, + "grad_norm": 2.001451344341508, + "learning_rate": 9.568683676099524e-06, + "loss": 0.3604, + "step": 7232 + }, + { + "epoch": 0.15921197446621174, + "grad_norm": 2.3373834420490374, + "learning_rate": 9.568538833855016e-06, + "loss": 0.3396, + "step": 7233 + }, + { + "epoch": 0.15923398635263042, + "grad_norm": 1.5776762400766662, + "learning_rate": 9.568393968391155e-06, + "loss": 0.4049, + "step": 7234 + }, + { + "epoch": 0.1592559982390491, + "grad_norm": 0.6424456656662889, + "learning_rate": 9.568249079708679e-06, + "loss": 0.5599, + "step": 7235 + }, + { + "epoch": 0.15927801012546774, + "grad_norm": 1.4126142231157812, + "learning_rate": 9.568104167808321e-06, + "loss": 0.3394, + "step": 7236 + }, + { + "epoch": 0.15930002201188642, + "grad_norm": 1.5868215387102516, + "learning_rate": 9.567959232690821e-06, + "loss": 0.3171, + "step": 7237 + }, + { + "epoch": 0.15932203389830507, + "grad_norm": 1.797991246717854, + "learning_rate": 9.567814274356918e-06, + "loss": 0.3597, + "step": 7238 + }, + { + "epoch": 0.15934404578472375, + "grad_norm": 0.6888500989424242, + "learning_rate": 9.567669292807342e-06, + "loss": 0.554, + "step": 7239 + }, + { + "epoch": 0.15936605767114242, + "grad_norm": 0.6740927115143274, + "learning_rate": 9.567524288042836e-06, + "loss": 0.5591, + "step": 7240 + }, + { + "epoch": 0.15938806955756107, + "grad_norm": 41.20598528497671, + "learning_rate": 9.567379260064131e-06, + "loss": 0.3179, + "step": 7241 + }, + { + "epoch": 0.15941008144397975, + "grad_norm": 1.7863969141023486, + "learning_rate": 9.567234208871972e-06, + "loss": 0.3409, + "step": 7242 + }, + { + "epoch": 0.15943209333039843, + "grad_norm": 1.4530879522834084, + "learning_rate": 9.567089134467089e-06, + "loss": 0.3721, + "step": 7243 + }, + { + "epoch": 0.15945410521681708, + "grad_norm": 1.9931977929482552, + "learning_rate": 9.566944036850221e-06, + "loss": 0.3171, + "step": 7244 + }, + { + "epoch": 0.15947611710323575, + "grad_norm": 1.7660706273341675, + "learning_rate": 9.566798916022108e-06, + "loss": 0.4434, + "step": 7245 + }, + { + "epoch": 0.1594981289896544, + "grad_norm": 1.5153158072839374, + "learning_rate": 9.566653771983488e-06, + "loss": 0.3873, + "step": 7246 + }, + { + "epoch": 0.15952014087607308, + "grad_norm": 1.4916167048203897, + "learning_rate": 9.566508604735093e-06, + "loss": 0.3899, + "step": 7247 + }, + { + "epoch": 0.15954215276249176, + "grad_norm": 2.133766764665687, + "learning_rate": 9.566363414277666e-06, + "loss": 0.4692, + "step": 7248 + }, + { + "epoch": 0.1595641646489104, + "grad_norm": 0.8234124412899113, + "learning_rate": 9.566218200611945e-06, + "loss": 0.5747, + "step": 7249 + }, + { + "epoch": 0.15958617653532908, + "grad_norm": 0.7087280412695379, + "learning_rate": 9.566072963738664e-06, + "loss": 0.5353, + "step": 7250 + }, + { + "epoch": 0.15960818842174773, + "grad_norm": 2.1407593631240855, + "learning_rate": 9.565927703658565e-06, + "loss": 0.3877, + "step": 7251 + }, + { + "epoch": 0.1596302003081664, + "grad_norm": 1.5310403908473589, + "learning_rate": 9.565782420372385e-06, + "loss": 0.3459, + "step": 7252 + }, + { + "epoch": 0.15965221219458509, + "grad_norm": 1.7813016152120313, + "learning_rate": 9.565637113880862e-06, + "loss": 0.3324, + "step": 7253 + }, + { + "epoch": 0.15967422408100373, + "grad_norm": 0.798586657354933, + "learning_rate": 9.565491784184734e-06, + "loss": 0.5781, + "step": 7254 + }, + { + "epoch": 0.1596962359674224, + "grad_norm": 3.0560808540936706, + "learning_rate": 9.565346431284744e-06, + "loss": 0.3947, + "step": 7255 + }, + { + "epoch": 0.15971824785384106, + "grad_norm": 1.4668707660163671, + "learning_rate": 9.565201055181624e-06, + "loss": 0.3148, + "step": 7256 + }, + { + "epoch": 0.15974025974025974, + "grad_norm": 0.7612831596475287, + "learning_rate": 9.565055655876117e-06, + "loss": 0.5471, + "step": 7257 + }, + { + "epoch": 0.15976227162667841, + "grad_norm": 1.921007458917989, + "learning_rate": 9.56491023336896e-06, + "loss": 0.3981, + "step": 7258 + }, + { + "epoch": 0.15978428351309706, + "grad_norm": 0.6482224716295931, + "learning_rate": 9.564764787660895e-06, + "loss": 0.5523, + "step": 7259 + }, + { + "epoch": 0.15980629539951574, + "grad_norm": 2.7158234893464432, + "learning_rate": 9.564619318752659e-06, + "loss": 0.4126, + "step": 7260 + }, + { + "epoch": 0.15982830728593442, + "grad_norm": 3.223308466312972, + "learning_rate": 9.56447382664499e-06, + "loss": 0.3531, + "step": 7261 + }, + { + "epoch": 0.15985031917235307, + "grad_norm": 1.921561386620844, + "learning_rate": 9.564328311338631e-06, + "loss": 0.4157, + "step": 7262 + }, + { + "epoch": 0.15987233105877174, + "grad_norm": 1.87246415994649, + "learning_rate": 9.56418277283432e-06, + "loss": 0.3488, + "step": 7263 + }, + { + "epoch": 0.1598943429451904, + "grad_norm": 1.9043118547450295, + "learning_rate": 9.564037211132794e-06, + "loss": 0.3657, + "step": 7264 + }, + { + "epoch": 0.15991635483160907, + "grad_norm": 1.4475787550654402, + "learning_rate": 9.563891626234797e-06, + "loss": 0.3129, + "step": 7265 + }, + { + "epoch": 0.15993836671802775, + "grad_norm": 0.979843465449059, + "learning_rate": 9.56374601814107e-06, + "loss": 0.545, + "step": 7266 + }, + { + "epoch": 0.1599603786044464, + "grad_norm": 1.7285334183344665, + "learning_rate": 9.563600386852345e-06, + "loss": 0.3087, + "step": 7267 + }, + { + "epoch": 0.15998239049086507, + "grad_norm": 6.3977598308808545, + "learning_rate": 9.56345473236937e-06, + "loss": 0.313, + "step": 7268 + }, + { + "epoch": 0.16000440237728372, + "grad_norm": 3.6305851635315256, + "learning_rate": 9.563309054692884e-06, + "loss": 0.2452, + "step": 7269 + }, + { + "epoch": 0.1600264142637024, + "grad_norm": 0.6271159388798662, + "learning_rate": 9.563163353823623e-06, + "loss": 0.5464, + "step": 7270 + }, + { + "epoch": 0.16004842615012108, + "grad_norm": 1.2969282970506517, + "learning_rate": 9.563017629762333e-06, + "loss": 0.318, + "step": 7271 + }, + { + "epoch": 0.16007043803653973, + "grad_norm": 2.9650286206648584, + "learning_rate": 9.562871882509752e-06, + "loss": 0.4719, + "step": 7272 + }, + { + "epoch": 0.1600924499229584, + "grad_norm": 2.1036558215924352, + "learning_rate": 9.56272611206662e-06, + "loss": 0.3949, + "step": 7273 + }, + { + "epoch": 0.16011446180937705, + "grad_norm": 0.7226035837572682, + "learning_rate": 9.562580318433679e-06, + "loss": 0.5348, + "step": 7274 + }, + { + "epoch": 0.16013647369579573, + "grad_norm": 1.5401382249812965, + "learning_rate": 9.562434501611671e-06, + "loss": 0.4256, + "step": 7275 + }, + { + "epoch": 0.1601584855822144, + "grad_norm": 1.51074756820327, + "learning_rate": 9.562288661601335e-06, + "loss": 0.3167, + "step": 7276 + }, + { + "epoch": 0.16018049746863305, + "grad_norm": 1.3777758208571305, + "learning_rate": 9.562142798403413e-06, + "loss": 0.3073, + "step": 7277 + }, + { + "epoch": 0.16020250935505173, + "grad_norm": 1.7274301687958793, + "learning_rate": 9.561996912018646e-06, + "loss": 0.3497, + "step": 7278 + }, + { + "epoch": 0.1602245212414704, + "grad_norm": 1.5425410361056093, + "learning_rate": 9.561851002447778e-06, + "loss": 0.3477, + "step": 7279 + }, + { + "epoch": 0.16024653312788906, + "grad_norm": 2.0226925853288336, + "learning_rate": 9.561705069691547e-06, + "loss": 0.3886, + "step": 7280 + }, + { + "epoch": 0.16026854501430773, + "grad_norm": 2.1335911368446667, + "learning_rate": 9.561559113750698e-06, + "loss": 0.3306, + "step": 7281 + }, + { + "epoch": 0.16029055690072638, + "grad_norm": 5.809645280001609, + "learning_rate": 9.56141313462597e-06, + "loss": 0.4274, + "step": 7282 + }, + { + "epoch": 0.16031256878714506, + "grad_norm": 1.577690394095497, + "learning_rate": 9.561267132318106e-06, + "loss": 0.2956, + "step": 7283 + }, + { + "epoch": 0.16033458067356374, + "grad_norm": 13.068350794412602, + "learning_rate": 9.561121106827846e-06, + "loss": 0.3772, + "step": 7284 + }, + { + "epoch": 0.16035659255998239, + "grad_norm": 1.3863096107352602, + "learning_rate": 9.560975058155937e-06, + "loss": 0.2553, + "step": 7285 + }, + { + "epoch": 0.16037860444640106, + "grad_norm": 0.6980046636097892, + "learning_rate": 9.560828986303117e-06, + "loss": 0.5694, + "step": 7286 + }, + { + "epoch": 0.1604006163328197, + "grad_norm": 2.3946552557534773, + "learning_rate": 9.560682891270131e-06, + "loss": 0.319, + "step": 7287 + }, + { + "epoch": 0.1604226282192384, + "grad_norm": 2.024347034021711, + "learning_rate": 9.56053677305772e-06, + "loss": 0.3267, + "step": 7288 + }, + { + "epoch": 0.16044464010565707, + "grad_norm": 2.0624510909146605, + "learning_rate": 9.560390631666628e-06, + "loss": 0.4013, + "step": 7289 + }, + { + "epoch": 0.16046665199207572, + "grad_norm": 3.078718476381135, + "learning_rate": 9.560244467097595e-06, + "loss": 0.3863, + "step": 7290 + }, + { + "epoch": 0.1604886638784944, + "grad_norm": 1.7757763465030008, + "learning_rate": 9.560098279351366e-06, + "loss": 0.2591, + "step": 7291 + }, + { + "epoch": 0.16051067576491304, + "grad_norm": 1.6080830721705104, + "learning_rate": 9.559952068428684e-06, + "loss": 0.3738, + "step": 7292 + }, + { + "epoch": 0.16053268765133172, + "grad_norm": 1.6713319180817714, + "learning_rate": 9.559805834330292e-06, + "loss": 0.3083, + "step": 7293 + }, + { + "epoch": 0.1605546995377504, + "grad_norm": 1.6759581470478724, + "learning_rate": 9.559659577056934e-06, + "loss": 0.4435, + "step": 7294 + }, + { + "epoch": 0.16057671142416904, + "grad_norm": 2.072017419812815, + "learning_rate": 9.55951329660935e-06, + "loss": 0.2888, + "step": 7295 + }, + { + "epoch": 0.16059872331058772, + "grad_norm": 0.660223724620244, + "learning_rate": 9.559366992988288e-06, + "loss": 0.5562, + "step": 7296 + }, + { + "epoch": 0.16062073519700637, + "grad_norm": 1.4023619110689514, + "learning_rate": 9.559220666194489e-06, + "loss": 0.312, + "step": 7297 + }, + { + "epoch": 0.16064274708342505, + "grad_norm": 1.411804790187141, + "learning_rate": 9.559074316228696e-06, + "loss": 0.2674, + "step": 7298 + }, + { + "epoch": 0.16066475896984372, + "grad_norm": 1.6322860920801945, + "learning_rate": 9.558927943091655e-06, + "loss": 0.3339, + "step": 7299 + }, + { + "epoch": 0.16068677085626237, + "grad_norm": 1.3200024158795478, + "learning_rate": 9.558781546784108e-06, + "loss": 0.3115, + "step": 7300 + }, + { + "epoch": 0.16070878274268105, + "grad_norm": 1.7668220272006112, + "learning_rate": 9.558635127306802e-06, + "loss": 0.4073, + "step": 7301 + }, + { + "epoch": 0.16073079462909973, + "grad_norm": 1.4220522352021547, + "learning_rate": 9.558488684660476e-06, + "loss": 0.3239, + "step": 7302 + }, + { + "epoch": 0.16075280651551838, + "grad_norm": 1.6705692905594935, + "learning_rate": 9.55834221884588e-06, + "loss": 0.2591, + "step": 7303 + }, + { + "epoch": 0.16077481840193705, + "grad_norm": 1.5487958493939293, + "learning_rate": 9.558195729863753e-06, + "loss": 0.3395, + "step": 7304 + }, + { + "epoch": 0.1607968302883557, + "grad_norm": 1.5996986271526175, + "learning_rate": 9.558049217714845e-06, + "loss": 0.4198, + "step": 7305 + }, + { + "epoch": 0.16081884217477438, + "grad_norm": 0.6762410875982207, + "learning_rate": 9.557902682399896e-06, + "loss": 0.5573, + "step": 7306 + }, + { + "epoch": 0.16084085406119306, + "grad_norm": 1.7808290166006613, + "learning_rate": 9.557756123919653e-06, + "loss": 0.4143, + "step": 7307 + }, + { + "epoch": 0.1608628659476117, + "grad_norm": 1.9572475001493201, + "learning_rate": 9.557609542274861e-06, + "loss": 0.3845, + "step": 7308 + }, + { + "epoch": 0.16088487783403038, + "grad_norm": 1.4521940420901507, + "learning_rate": 9.557462937466265e-06, + "loss": 0.298, + "step": 7309 + }, + { + "epoch": 0.16090688972044903, + "grad_norm": 2.7346129432225985, + "learning_rate": 9.557316309494609e-06, + "loss": 0.4237, + "step": 7310 + }, + { + "epoch": 0.1609289016068677, + "grad_norm": 1.4740292886184043, + "learning_rate": 9.557169658360638e-06, + "loss": 0.3219, + "step": 7311 + }, + { + "epoch": 0.16095091349328638, + "grad_norm": 1.7794590922491438, + "learning_rate": 9.557022984065101e-06, + "loss": 0.4217, + "step": 7312 + }, + { + "epoch": 0.16097292537970503, + "grad_norm": 1.6850207916987772, + "learning_rate": 9.556876286608739e-06, + "loss": 0.2931, + "step": 7313 + }, + { + "epoch": 0.1609949372661237, + "grad_norm": 1.375428151655123, + "learning_rate": 9.5567295659923e-06, + "loss": 0.3429, + "step": 7314 + }, + { + "epoch": 0.16101694915254236, + "grad_norm": 1.3604182695297478, + "learning_rate": 9.556582822216527e-06, + "loss": 0.3352, + "step": 7315 + }, + { + "epoch": 0.16103896103896104, + "grad_norm": 1.4817580434858066, + "learning_rate": 9.55643605528217e-06, + "loss": 0.2608, + "step": 7316 + }, + { + "epoch": 0.16106097292537971, + "grad_norm": 2.7083790278829, + "learning_rate": 9.556289265189972e-06, + "loss": 0.3762, + "step": 7317 + }, + { + "epoch": 0.16108298481179836, + "grad_norm": 0.7266378516559985, + "learning_rate": 9.55614245194068e-06, + "loss": 0.5573, + "step": 7318 + }, + { + "epoch": 0.16110499669821704, + "grad_norm": 2.2174312096544964, + "learning_rate": 9.555995615535039e-06, + "loss": 0.4068, + "step": 7319 + }, + { + "epoch": 0.16112700858463572, + "grad_norm": 1.7483132381780233, + "learning_rate": 9.555848755973797e-06, + "loss": 0.3805, + "step": 7320 + }, + { + "epoch": 0.16114902047105437, + "grad_norm": 1.4574066389412228, + "learning_rate": 9.555701873257699e-06, + "loss": 0.377, + "step": 7321 + }, + { + "epoch": 0.16117103235747304, + "grad_norm": 1.340496034435133, + "learning_rate": 9.555554967387492e-06, + "loss": 0.1948, + "step": 7322 + }, + { + "epoch": 0.1611930442438917, + "grad_norm": 0.6433828270181726, + "learning_rate": 9.555408038363925e-06, + "loss": 0.5418, + "step": 7323 + }, + { + "epoch": 0.16121505613031037, + "grad_norm": 1.26241339881205, + "learning_rate": 9.555261086187742e-06, + "loss": 0.2755, + "step": 7324 + }, + { + "epoch": 0.16123706801672905, + "grad_norm": 2.463246523278917, + "learning_rate": 9.555114110859688e-06, + "loss": 0.4242, + "step": 7325 + }, + { + "epoch": 0.1612590799031477, + "grad_norm": 2.224666951335217, + "learning_rate": 9.554967112380515e-06, + "loss": 0.343, + "step": 7326 + }, + { + "epoch": 0.16128109178956637, + "grad_norm": 2.0195408885322648, + "learning_rate": 9.554820090750968e-06, + "loss": 0.3487, + "step": 7327 + }, + { + "epoch": 0.16130310367598502, + "grad_norm": 2.0763567578704456, + "learning_rate": 9.554673045971791e-06, + "loss": 0.3588, + "step": 7328 + }, + { + "epoch": 0.1613251155624037, + "grad_norm": 1.6286027140857644, + "learning_rate": 9.554525978043736e-06, + "loss": 0.3645, + "step": 7329 + }, + { + "epoch": 0.16134712744882238, + "grad_norm": 1.6187598310907425, + "learning_rate": 9.554378886967548e-06, + "loss": 0.3146, + "step": 7330 + }, + { + "epoch": 0.16136913933524102, + "grad_norm": 1.3667118202662152, + "learning_rate": 9.554231772743974e-06, + "loss": 0.3353, + "step": 7331 + }, + { + "epoch": 0.1613911512216597, + "grad_norm": 1.773473527036326, + "learning_rate": 9.554084635373764e-06, + "loss": 0.354, + "step": 7332 + }, + { + "epoch": 0.16141316310807835, + "grad_norm": 0.6494246075175935, + "learning_rate": 9.553937474857666e-06, + "loss": 0.5254, + "step": 7333 + }, + { + "epoch": 0.16143517499449703, + "grad_norm": 0.6806753167018258, + "learning_rate": 9.553790291196425e-06, + "loss": 0.5603, + "step": 7334 + }, + { + "epoch": 0.1614571868809157, + "grad_norm": 1.6206681928109874, + "learning_rate": 9.55364308439079e-06, + "loss": 0.341, + "step": 7335 + }, + { + "epoch": 0.16147919876733435, + "grad_norm": 1.4780654687491237, + "learning_rate": 9.553495854441513e-06, + "loss": 0.336, + "step": 7336 + }, + { + "epoch": 0.16150121065375303, + "grad_norm": 1.721210676862557, + "learning_rate": 9.553348601349337e-06, + "loss": 0.3319, + "step": 7337 + }, + { + "epoch": 0.16152322254017168, + "grad_norm": 2.077422550108891, + "learning_rate": 9.553201325115014e-06, + "loss": 0.3419, + "step": 7338 + }, + { + "epoch": 0.16154523442659036, + "grad_norm": 1.6843954931971847, + "learning_rate": 9.553054025739289e-06, + "loss": 0.323, + "step": 7339 + }, + { + "epoch": 0.16156724631300903, + "grad_norm": 1.450622238112069, + "learning_rate": 9.552906703222914e-06, + "loss": 0.381, + "step": 7340 + }, + { + "epoch": 0.16158925819942768, + "grad_norm": 1.6259397325876488, + "learning_rate": 9.552759357566636e-06, + "loss": 0.4312, + "step": 7341 + }, + { + "epoch": 0.16161127008584636, + "grad_norm": 2.451843169798326, + "learning_rate": 9.552611988771205e-06, + "loss": 0.3977, + "step": 7342 + }, + { + "epoch": 0.16163328197226504, + "grad_norm": 5.693965740201175, + "learning_rate": 9.552464596837367e-06, + "loss": 0.3344, + "step": 7343 + }, + { + "epoch": 0.16165529385868369, + "grad_norm": 1.5154205606998652, + "learning_rate": 9.552317181765876e-06, + "loss": 0.4137, + "step": 7344 + }, + { + "epoch": 0.16167730574510236, + "grad_norm": 3.8886841335108557, + "learning_rate": 9.552169743557478e-06, + "loss": 0.374, + "step": 7345 + }, + { + "epoch": 0.161699317631521, + "grad_norm": 0.7599272408122684, + "learning_rate": 9.552022282212922e-06, + "loss": 0.5366, + "step": 7346 + }, + { + "epoch": 0.1617213295179397, + "grad_norm": 1.4070510325185475, + "learning_rate": 9.551874797732961e-06, + "loss": 0.2445, + "step": 7347 + }, + { + "epoch": 0.16174334140435837, + "grad_norm": 3.8424676586071884, + "learning_rate": 9.55172729011834e-06, + "loss": 0.3467, + "step": 7348 + }, + { + "epoch": 0.16176535329077701, + "grad_norm": 1.5136425518121668, + "learning_rate": 9.55157975936981e-06, + "loss": 0.3248, + "step": 7349 + }, + { + "epoch": 0.1617873651771957, + "grad_norm": 2.0750639340161188, + "learning_rate": 9.551432205488125e-06, + "loss": 0.3604, + "step": 7350 + }, + { + "epoch": 0.16180937706361434, + "grad_norm": 1.4448134755916995, + "learning_rate": 9.55128462847403e-06, + "loss": 0.423, + "step": 7351 + }, + { + "epoch": 0.16183138895003302, + "grad_norm": 0.6761808566260515, + "learning_rate": 9.551137028328275e-06, + "loss": 0.5561, + "step": 7352 + }, + { + "epoch": 0.1618534008364517, + "grad_norm": 0.6774747575537484, + "learning_rate": 9.550989405051613e-06, + "loss": 0.5603, + "step": 7353 + }, + { + "epoch": 0.16187541272287034, + "grad_norm": 0.6376565235794174, + "learning_rate": 9.550841758644794e-06, + "loss": 0.5254, + "step": 7354 + }, + { + "epoch": 0.16189742460928902, + "grad_norm": 5.528821116684129, + "learning_rate": 9.550694089108566e-06, + "loss": 0.3462, + "step": 7355 + }, + { + "epoch": 0.16191943649570767, + "grad_norm": 1.7597947892354449, + "learning_rate": 9.550546396443681e-06, + "loss": 0.3479, + "step": 7356 + }, + { + "epoch": 0.16194144838212635, + "grad_norm": 3.048495287909846, + "learning_rate": 9.55039868065089e-06, + "loss": 0.3993, + "step": 7357 + }, + { + "epoch": 0.16196346026854502, + "grad_norm": 1.5304620400788844, + "learning_rate": 9.550250941730943e-06, + "loss": 0.3769, + "step": 7358 + }, + { + "epoch": 0.16198547215496367, + "grad_norm": 2.1292246932650047, + "learning_rate": 9.55010317968459e-06, + "loss": 0.3622, + "step": 7359 + }, + { + "epoch": 0.16200748404138235, + "grad_norm": 2.16890664402361, + "learning_rate": 9.549955394512587e-06, + "loss": 0.4053, + "step": 7360 + }, + { + "epoch": 0.16202949592780103, + "grad_norm": 2.1409908255196726, + "learning_rate": 9.549807586215679e-06, + "loss": 0.3315, + "step": 7361 + }, + { + "epoch": 0.16205150781421968, + "grad_norm": 0.8747517504194715, + "learning_rate": 9.54965975479462e-06, + "loss": 0.566, + "step": 7362 + }, + { + "epoch": 0.16207351970063835, + "grad_norm": 1.707543623714608, + "learning_rate": 9.54951190025016e-06, + "loss": 0.3829, + "step": 7363 + }, + { + "epoch": 0.162095531587057, + "grad_norm": 1.663158283992101, + "learning_rate": 9.549364022583052e-06, + "loss": 0.3454, + "step": 7364 + }, + { + "epoch": 0.16211754347347568, + "grad_norm": 1.8706248885990284, + "learning_rate": 9.549216121794046e-06, + "loss": 0.3576, + "step": 7365 + }, + { + "epoch": 0.16213955535989436, + "grad_norm": 1.5803412753079864, + "learning_rate": 9.549068197883897e-06, + "loss": 0.339, + "step": 7366 + }, + { + "epoch": 0.162161567246313, + "grad_norm": 2.5410703966065453, + "learning_rate": 9.548920250853353e-06, + "loss": 0.3591, + "step": 7367 + }, + { + "epoch": 0.16218357913273168, + "grad_norm": 2.209190312143896, + "learning_rate": 9.548772280703167e-06, + "loss": 0.3349, + "step": 7368 + }, + { + "epoch": 0.16220559101915033, + "grad_norm": 3.0833389606069384, + "learning_rate": 9.548624287434093e-06, + "loss": 0.3466, + "step": 7369 + }, + { + "epoch": 0.162227602905569, + "grad_norm": 1.5799133612860272, + "learning_rate": 9.54847627104688e-06, + "loss": 0.3317, + "step": 7370 + }, + { + "epoch": 0.16224961479198768, + "grad_norm": 1.3655669370483225, + "learning_rate": 9.548328231542283e-06, + "loss": 0.3182, + "step": 7371 + }, + { + "epoch": 0.16227162667840633, + "grad_norm": 1.5871631340668975, + "learning_rate": 9.548180168921053e-06, + "loss": 0.3724, + "step": 7372 + }, + { + "epoch": 0.162293638564825, + "grad_norm": 1.9906338593385753, + "learning_rate": 9.548032083183943e-06, + "loss": 0.4591, + "step": 7373 + }, + { + "epoch": 0.16231565045124366, + "grad_norm": 2.5778983715100856, + "learning_rate": 9.547883974331703e-06, + "loss": 0.3964, + "step": 7374 + }, + { + "epoch": 0.16233766233766234, + "grad_norm": 3.0266974645856854, + "learning_rate": 9.547735842365092e-06, + "loss": 0.3487, + "step": 7375 + }, + { + "epoch": 0.162359674224081, + "grad_norm": 0.8450066690245873, + "learning_rate": 9.547587687284859e-06, + "loss": 0.5746, + "step": 7376 + }, + { + "epoch": 0.16238168611049966, + "grad_norm": 1.663488530356009, + "learning_rate": 9.547439509091754e-06, + "loss": 0.3922, + "step": 7377 + }, + { + "epoch": 0.16240369799691834, + "grad_norm": 0.6675032895478571, + "learning_rate": 9.547291307786534e-06, + "loss": 0.5287, + "step": 7378 + }, + { + "epoch": 0.162425709883337, + "grad_norm": 1.581275505391469, + "learning_rate": 9.547143083369954e-06, + "loss": 0.368, + "step": 7379 + }, + { + "epoch": 0.16244772176975567, + "grad_norm": 1.6724919054534026, + "learning_rate": 9.546994835842762e-06, + "loss": 0.3569, + "step": 7380 + }, + { + "epoch": 0.16246973365617434, + "grad_norm": 1.9504558823251372, + "learning_rate": 9.546846565205716e-06, + "loss": 0.3239, + "step": 7381 + }, + { + "epoch": 0.162491745542593, + "grad_norm": 1.4685624569803344, + "learning_rate": 9.546698271459568e-06, + "loss": 0.366, + "step": 7382 + }, + { + "epoch": 0.16251375742901167, + "grad_norm": 3.231144919715985, + "learning_rate": 9.54654995460507e-06, + "loss": 0.4009, + "step": 7383 + }, + { + "epoch": 0.16253576931543035, + "grad_norm": 2.503823155668504, + "learning_rate": 9.546401614642977e-06, + "loss": 0.3779, + "step": 7384 + }, + { + "epoch": 0.162557781201849, + "grad_norm": 1.016774210127703, + "learning_rate": 9.546253251574044e-06, + "loss": 0.5357, + "step": 7385 + }, + { + "epoch": 0.16257979308826767, + "grad_norm": 1.675247280485062, + "learning_rate": 9.546104865399025e-06, + "loss": 0.3717, + "step": 7386 + }, + { + "epoch": 0.16260180497468632, + "grad_norm": 1.6217079304799864, + "learning_rate": 9.545956456118673e-06, + "loss": 0.2301, + "step": 7387 + }, + { + "epoch": 0.162623816861105, + "grad_norm": 1.4424097361645325, + "learning_rate": 9.545808023733742e-06, + "loss": 0.3618, + "step": 7388 + }, + { + "epoch": 0.16264582874752367, + "grad_norm": 1.4832878690606848, + "learning_rate": 9.545659568244989e-06, + "loss": 0.3798, + "step": 7389 + }, + { + "epoch": 0.16266784063394232, + "grad_norm": 0.6803811738070924, + "learning_rate": 9.545511089653164e-06, + "loss": 0.5707, + "step": 7390 + }, + { + "epoch": 0.162689852520361, + "grad_norm": 1.5372626067339246, + "learning_rate": 9.545362587959026e-06, + "loss": 0.3345, + "step": 7391 + }, + { + "epoch": 0.16271186440677965, + "grad_norm": 2.2277447281314076, + "learning_rate": 9.545214063163328e-06, + "loss": 0.3733, + "step": 7392 + }, + { + "epoch": 0.16273387629319833, + "grad_norm": 1.8659902150411434, + "learning_rate": 9.545065515266825e-06, + "loss": 0.3176, + "step": 7393 + }, + { + "epoch": 0.162755888179617, + "grad_norm": 1.673105352122768, + "learning_rate": 9.54491694427027e-06, + "loss": 0.3522, + "step": 7394 + }, + { + "epoch": 0.16277790006603565, + "grad_norm": 2.4822291591077907, + "learning_rate": 9.544768350174423e-06, + "loss": 0.3388, + "step": 7395 + }, + { + "epoch": 0.16279991195245433, + "grad_norm": 0.8261233476708354, + "learning_rate": 9.544619732980034e-06, + "loss": 0.5187, + "step": 7396 + }, + { + "epoch": 0.16282192383887298, + "grad_norm": 1.6975618486505726, + "learning_rate": 9.544471092687861e-06, + "loss": 0.4066, + "step": 7397 + }, + { + "epoch": 0.16284393572529166, + "grad_norm": 1.432067669330511, + "learning_rate": 9.54432242929866e-06, + "loss": 0.3051, + "step": 7398 + }, + { + "epoch": 0.16286594761171033, + "grad_norm": 1.7657493514162768, + "learning_rate": 9.544173742813185e-06, + "loss": 0.4036, + "step": 7399 + }, + { + "epoch": 0.16288795949812898, + "grad_norm": 1.47870358624138, + "learning_rate": 9.544025033232191e-06, + "loss": 0.3423, + "step": 7400 + }, + { + "epoch": 0.16290997138454766, + "grad_norm": 0.6703169602998709, + "learning_rate": 9.543876300556437e-06, + "loss": 0.546, + "step": 7401 + }, + { + "epoch": 0.16293198327096634, + "grad_norm": 1.4043420939872093, + "learning_rate": 9.543727544786676e-06, + "loss": 0.3014, + "step": 7402 + }, + { + "epoch": 0.16295399515738498, + "grad_norm": 31.786857323200095, + "learning_rate": 9.543578765923665e-06, + "loss": 0.393, + "step": 7403 + }, + { + "epoch": 0.16297600704380366, + "grad_norm": 3.6626150912892137, + "learning_rate": 9.54342996396816e-06, + "loss": 0.3761, + "step": 7404 + }, + { + "epoch": 0.1629980189302223, + "grad_norm": 3.3450871306067804, + "learning_rate": 9.543281138920919e-06, + "loss": 0.402, + "step": 7405 + }, + { + "epoch": 0.163020030816641, + "grad_norm": 2.5765372420992727, + "learning_rate": 9.543132290782697e-06, + "loss": 0.3544, + "step": 7406 + }, + { + "epoch": 0.16304204270305966, + "grad_norm": 2.302243189285229, + "learning_rate": 9.542983419554248e-06, + "loss": 0.5041, + "step": 7407 + }, + { + "epoch": 0.1630640545894783, + "grad_norm": 1.8983030430567471, + "learning_rate": 9.542834525236332e-06, + "loss": 0.2983, + "step": 7408 + }, + { + "epoch": 0.163086066475897, + "grad_norm": 3.3331018782313495, + "learning_rate": 9.542685607829706e-06, + "loss": 0.4059, + "step": 7409 + }, + { + "epoch": 0.16310807836231564, + "grad_norm": 1.2484181541151285, + "learning_rate": 9.542536667335125e-06, + "loss": 0.2918, + "step": 7410 + }, + { + "epoch": 0.16313009024873432, + "grad_norm": 1.8019714739636377, + "learning_rate": 9.542387703753346e-06, + "loss": 0.34, + "step": 7411 + }, + { + "epoch": 0.163152102135153, + "grad_norm": 1.5735243994261816, + "learning_rate": 9.542238717085129e-06, + "loss": 0.4569, + "step": 7412 + }, + { + "epoch": 0.16317411402157164, + "grad_norm": 2.2119455232680925, + "learning_rate": 9.542089707331225e-06, + "loss": 0.4106, + "step": 7413 + }, + { + "epoch": 0.16319612590799032, + "grad_norm": 1.5733541451706192, + "learning_rate": 9.541940674492399e-06, + "loss": 0.3907, + "step": 7414 + }, + { + "epoch": 0.16321813779440897, + "grad_norm": 2.0439328436637125, + "learning_rate": 9.541791618569402e-06, + "loss": 0.2824, + "step": 7415 + }, + { + "epoch": 0.16324014968082765, + "grad_norm": 2.3739187322661266, + "learning_rate": 9.541642539562995e-06, + "loss": 0.4162, + "step": 7416 + }, + { + "epoch": 0.16326216156724632, + "grad_norm": 2.096247389137367, + "learning_rate": 9.541493437473936e-06, + "loss": 0.2458, + "step": 7417 + }, + { + "epoch": 0.16328417345366497, + "grad_norm": 1.8321479288769849, + "learning_rate": 9.54134431230298e-06, + "loss": 0.3228, + "step": 7418 + }, + { + "epoch": 0.16330618534008365, + "grad_norm": 1.3498446127246102, + "learning_rate": 9.541195164050889e-06, + "loss": 0.3228, + "step": 7419 + }, + { + "epoch": 0.1633281972265023, + "grad_norm": 2.1125433515715337, + "learning_rate": 9.541045992718417e-06, + "loss": 0.3093, + "step": 7420 + }, + { + "epoch": 0.16335020911292097, + "grad_norm": 1.3230648181210571, + "learning_rate": 9.540896798306324e-06, + "loss": 0.303, + "step": 7421 + }, + { + "epoch": 0.16337222099933965, + "grad_norm": 1.364954880694317, + "learning_rate": 9.540747580815371e-06, + "loss": 0.2872, + "step": 7422 + }, + { + "epoch": 0.1633942328857583, + "grad_norm": 1.6721292044212626, + "learning_rate": 9.54059834024631e-06, + "loss": 0.3998, + "step": 7423 + }, + { + "epoch": 0.16341624477217698, + "grad_norm": 1.642976857807322, + "learning_rate": 9.540449076599905e-06, + "loss": 0.3598, + "step": 7424 + }, + { + "epoch": 0.16343825665859565, + "grad_norm": 2.0127820435706476, + "learning_rate": 9.540299789876912e-06, + "loss": 0.3894, + "step": 7425 + }, + { + "epoch": 0.1634602685450143, + "grad_norm": 0.8047003968134828, + "learning_rate": 9.54015048007809e-06, + "loss": 0.5427, + "step": 7426 + }, + { + "epoch": 0.16348228043143298, + "grad_norm": 1.4469944140184616, + "learning_rate": 9.540001147204199e-06, + "loss": 0.3097, + "step": 7427 + }, + { + "epoch": 0.16350429231785163, + "grad_norm": 2.2185521789231024, + "learning_rate": 9.539851791255996e-06, + "loss": 0.4046, + "step": 7428 + }, + { + "epoch": 0.1635263042042703, + "grad_norm": 1.8785828381880745, + "learning_rate": 9.539702412234244e-06, + "loss": 0.3994, + "step": 7429 + }, + { + "epoch": 0.16354831609068898, + "grad_norm": 0.6437332399683782, + "learning_rate": 9.539553010139697e-06, + "loss": 0.5689, + "step": 7430 + }, + { + "epoch": 0.16357032797710763, + "grad_norm": 1.6305477470139798, + "learning_rate": 9.539403584973119e-06, + "loss": 0.3818, + "step": 7431 + }, + { + "epoch": 0.1635923398635263, + "grad_norm": 1.4805079613631482, + "learning_rate": 9.539254136735266e-06, + "loss": 0.3989, + "step": 7432 + }, + { + "epoch": 0.16361435174994496, + "grad_norm": 0.6939456240531778, + "learning_rate": 9.539104665426899e-06, + "loss": 0.5304, + "step": 7433 + }, + { + "epoch": 0.16363636363636364, + "grad_norm": 0.671127837316332, + "learning_rate": 9.538955171048778e-06, + "loss": 0.5327, + "step": 7434 + }, + { + "epoch": 0.1636583755227823, + "grad_norm": 1.5181754194027675, + "learning_rate": 9.538805653601662e-06, + "loss": 0.3652, + "step": 7435 + }, + { + "epoch": 0.16368038740920096, + "grad_norm": 0.6348605924554909, + "learning_rate": 9.53865611308631e-06, + "loss": 0.5753, + "step": 7436 + }, + { + "epoch": 0.16370239929561964, + "grad_norm": 1.9468850352418081, + "learning_rate": 9.538506549503485e-06, + "loss": 0.3201, + "step": 7437 + }, + { + "epoch": 0.1637244111820383, + "grad_norm": 1.5120181010039713, + "learning_rate": 9.538356962853947e-06, + "loss": 0.2803, + "step": 7438 + }, + { + "epoch": 0.16374642306845696, + "grad_norm": 1.622700870012215, + "learning_rate": 9.538207353138453e-06, + "loss": 0.4209, + "step": 7439 + }, + { + "epoch": 0.16376843495487564, + "grad_norm": 1.8642557916385198, + "learning_rate": 9.538057720357766e-06, + "loss": 0.3531, + "step": 7440 + }, + { + "epoch": 0.1637904468412943, + "grad_norm": 2.1217771041566436, + "learning_rate": 9.537908064512642e-06, + "loss": 0.3828, + "step": 7441 + }, + { + "epoch": 0.16381245872771297, + "grad_norm": 1.5944118249745873, + "learning_rate": 9.537758385603849e-06, + "loss": 0.3843, + "step": 7442 + }, + { + "epoch": 0.16383447061413164, + "grad_norm": 0.8523409747197555, + "learning_rate": 9.537608683632143e-06, + "loss": 0.5529, + "step": 7443 + }, + { + "epoch": 0.1638564825005503, + "grad_norm": 1.686282161389938, + "learning_rate": 9.537458958598287e-06, + "loss": 0.3025, + "step": 7444 + }, + { + "epoch": 0.16387849438696897, + "grad_norm": 2.0469324729928733, + "learning_rate": 9.537309210503038e-06, + "loss": 0.4191, + "step": 7445 + }, + { + "epoch": 0.16390050627338762, + "grad_norm": 1.4564641323300374, + "learning_rate": 9.537159439347164e-06, + "loss": 0.3552, + "step": 7446 + }, + { + "epoch": 0.1639225181598063, + "grad_norm": 1.623898423754565, + "learning_rate": 9.53700964513142e-06, + "loss": 0.4505, + "step": 7447 + }, + { + "epoch": 0.16394453004622497, + "grad_norm": 1.6591028056476156, + "learning_rate": 9.536859827856566e-06, + "loss": 0.3737, + "step": 7448 + }, + { + "epoch": 0.16396654193264362, + "grad_norm": 1.4106845027464516, + "learning_rate": 9.536709987523371e-06, + "loss": 0.3452, + "step": 7449 + }, + { + "epoch": 0.1639885538190623, + "grad_norm": 1.8601031207725827, + "learning_rate": 9.536560124132592e-06, + "loss": 0.3041, + "step": 7450 + }, + { + "epoch": 0.16401056570548095, + "grad_norm": 1.7312060849175464, + "learning_rate": 9.53641023768499e-06, + "loss": 0.3459, + "step": 7451 + }, + { + "epoch": 0.16403257759189963, + "grad_norm": 1.8715591426325366, + "learning_rate": 9.536260328181327e-06, + "loss": 0.4383, + "step": 7452 + }, + { + "epoch": 0.1640545894783183, + "grad_norm": 1.802548736671447, + "learning_rate": 9.536110395622368e-06, + "loss": 0.3574, + "step": 7453 + }, + { + "epoch": 0.16407660136473695, + "grad_norm": 1.9685980027058263, + "learning_rate": 9.535960440008873e-06, + "loss": 0.3261, + "step": 7454 + }, + { + "epoch": 0.16409861325115563, + "grad_norm": 1.6066470835416562, + "learning_rate": 9.535810461341602e-06, + "loss": 0.2786, + "step": 7455 + }, + { + "epoch": 0.16412062513757428, + "grad_norm": 0.8063776849883866, + "learning_rate": 9.53566045962132e-06, + "loss": 0.5562, + "step": 7456 + }, + { + "epoch": 0.16414263702399295, + "grad_norm": 2.1732860315800093, + "learning_rate": 9.53551043484879e-06, + "loss": 0.3128, + "step": 7457 + }, + { + "epoch": 0.16416464891041163, + "grad_norm": 1.4419867042365442, + "learning_rate": 9.53536038702477e-06, + "loss": 0.4375, + "step": 7458 + }, + { + "epoch": 0.16418666079683028, + "grad_norm": 1.6578096007793768, + "learning_rate": 9.535210316150027e-06, + "loss": 0.2825, + "step": 7459 + }, + { + "epoch": 0.16420867268324896, + "grad_norm": 1.3611683563181474, + "learning_rate": 9.535060222225325e-06, + "loss": 0.3697, + "step": 7460 + }, + { + "epoch": 0.16423068456966763, + "grad_norm": 1.4445575166147158, + "learning_rate": 9.534910105251421e-06, + "loss": 0.3078, + "step": 7461 + }, + { + "epoch": 0.16425269645608628, + "grad_norm": 1.3398918264199469, + "learning_rate": 9.534759965229084e-06, + "loss": 0.3196, + "step": 7462 + }, + { + "epoch": 0.16427470834250496, + "grad_norm": 1.777473153518185, + "learning_rate": 9.534609802159071e-06, + "loss": 0.4362, + "step": 7463 + }, + { + "epoch": 0.1642967202289236, + "grad_norm": 1.9428009260380121, + "learning_rate": 9.534459616042151e-06, + "loss": 0.3882, + "step": 7464 + }, + { + "epoch": 0.1643187321153423, + "grad_norm": 1.3386724057939245, + "learning_rate": 9.534309406879084e-06, + "loss": 0.3303, + "step": 7465 + }, + { + "epoch": 0.16434074400176096, + "grad_norm": 2.5491965714289995, + "learning_rate": 9.534159174670634e-06, + "loss": 0.3571, + "step": 7466 + }, + { + "epoch": 0.1643627558881796, + "grad_norm": 2.948174187503492, + "learning_rate": 9.534008919417565e-06, + "loss": 0.2643, + "step": 7467 + }, + { + "epoch": 0.1643847677745983, + "grad_norm": 0.7236784019674668, + "learning_rate": 9.53385864112064e-06, + "loss": 0.5621, + "step": 7468 + }, + { + "epoch": 0.16440677966101694, + "grad_norm": 0.6655713513274865, + "learning_rate": 9.533708339780623e-06, + "loss": 0.5325, + "step": 7469 + }, + { + "epoch": 0.16442879154743562, + "grad_norm": 1.551906749008795, + "learning_rate": 9.53355801539828e-06, + "loss": 0.356, + "step": 7470 + }, + { + "epoch": 0.1644508034338543, + "grad_norm": 1.5811947962319586, + "learning_rate": 9.53340766797437e-06, + "loss": 0.4133, + "step": 7471 + }, + { + "epoch": 0.16447281532027294, + "grad_norm": 1.65746010106263, + "learning_rate": 9.533257297509663e-06, + "loss": 0.4067, + "step": 7472 + }, + { + "epoch": 0.16449482720669162, + "grad_norm": 0.6821313750068816, + "learning_rate": 9.53310690400492e-06, + "loss": 0.5414, + "step": 7473 + }, + { + "epoch": 0.16451683909311027, + "grad_norm": 0.7114750762918227, + "learning_rate": 9.532956487460905e-06, + "loss": 0.5349, + "step": 7474 + }, + { + "epoch": 0.16453885097952894, + "grad_norm": 1.443419333819892, + "learning_rate": 9.532806047878384e-06, + "loss": 0.364, + "step": 7475 + }, + { + "epoch": 0.16456086286594762, + "grad_norm": 2.7977155292384976, + "learning_rate": 9.532655585258121e-06, + "loss": 0.3869, + "step": 7476 + }, + { + "epoch": 0.16458287475236627, + "grad_norm": 1.9687416977170182, + "learning_rate": 9.532505099600882e-06, + "loss": 0.336, + "step": 7477 + }, + { + "epoch": 0.16460488663878495, + "grad_norm": 1.8168857067927204, + "learning_rate": 9.53235459090743e-06, + "loss": 0.3738, + "step": 7478 + }, + { + "epoch": 0.1646268985252036, + "grad_norm": 1.682181282352791, + "learning_rate": 9.532204059178527e-06, + "loss": 0.3064, + "step": 7479 + }, + { + "epoch": 0.16464891041162227, + "grad_norm": 0.6498327438651095, + "learning_rate": 9.532053504414945e-06, + "loss": 0.5478, + "step": 7480 + }, + { + "epoch": 0.16467092229804095, + "grad_norm": 2.042846725266276, + "learning_rate": 9.531902926617444e-06, + "loss": 0.3324, + "step": 7481 + }, + { + "epoch": 0.1646929341844596, + "grad_norm": 1.5318053114579895, + "learning_rate": 9.531752325786793e-06, + "loss": 0.3962, + "step": 7482 + }, + { + "epoch": 0.16471494607087828, + "grad_norm": 1.6434773593431324, + "learning_rate": 9.531601701923753e-06, + "loss": 0.3567, + "step": 7483 + }, + { + "epoch": 0.16473695795729695, + "grad_norm": 1.4995545597402227, + "learning_rate": 9.531451055029096e-06, + "loss": 0.3962, + "step": 7484 + }, + { + "epoch": 0.1647589698437156, + "grad_norm": 1.9446006859199403, + "learning_rate": 9.531300385103578e-06, + "loss": 0.2318, + "step": 7485 + }, + { + "epoch": 0.16478098173013428, + "grad_norm": 1.5303758692579708, + "learning_rate": 9.531149692147975e-06, + "loss": 0.3033, + "step": 7486 + }, + { + "epoch": 0.16480299361655293, + "grad_norm": 1.6200187484299926, + "learning_rate": 9.530998976163047e-06, + "loss": 0.3112, + "step": 7487 + }, + { + "epoch": 0.1648250055029716, + "grad_norm": 1.7085404235139356, + "learning_rate": 9.530848237149562e-06, + "loss": 0.421, + "step": 7488 + }, + { + "epoch": 0.16484701738939028, + "grad_norm": 0.708484423764682, + "learning_rate": 9.530697475108283e-06, + "loss": 0.5376, + "step": 7489 + }, + { + "epoch": 0.16486902927580893, + "grad_norm": 0.6281002619268102, + "learning_rate": 9.530546690039982e-06, + "loss": 0.5295, + "step": 7490 + }, + { + "epoch": 0.1648910411622276, + "grad_norm": 1.4457582596375629, + "learning_rate": 9.53039588194542e-06, + "loss": 0.3144, + "step": 7491 + }, + { + "epoch": 0.16491305304864626, + "grad_norm": 1.4024292667305815, + "learning_rate": 9.530245050825365e-06, + "loss": 0.3615, + "step": 7492 + }, + { + "epoch": 0.16493506493506493, + "grad_norm": 1.7851957650653012, + "learning_rate": 9.530094196680585e-06, + "loss": 0.4168, + "step": 7493 + }, + { + "epoch": 0.1649570768214836, + "grad_norm": 2.781478939604787, + "learning_rate": 9.529943319511847e-06, + "loss": 0.2577, + "step": 7494 + }, + { + "epoch": 0.16497908870790226, + "grad_norm": 1.9793079584731403, + "learning_rate": 9.529792419319915e-06, + "loss": 0.3189, + "step": 7495 + }, + { + "epoch": 0.16500110059432094, + "grad_norm": 0.694067939538095, + "learning_rate": 9.529641496105557e-06, + "loss": 0.5593, + "step": 7496 + }, + { + "epoch": 0.1650231124807396, + "grad_norm": 1.818414851365028, + "learning_rate": 9.529490549869542e-06, + "loss": 0.4611, + "step": 7497 + }, + { + "epoch": 0.16504512436715826, + "grad_norm": 1.624245585738015, + "learning_rate": 9.529339580612636e-06, + "loss": 0.3495, + "step": 7498 + }, + { + "epoch": 0.16506713625357694, + "grad_norm": 1.4280253890836045, + "learning_rate": 9.529188588335604e-06, + "loss": 0.2362, + "step": 7499 + }, + { + "epoch": 0.1650891481399956, + "grad_norm": 3.3671038771043063, + "learning_rate": 9.529037573039218e-06, + "loss": 0.303, + "step": 7500 + }, + { + "epoch": 0.16511116002641427, + "grad_norm": 0.6384438973951813, + "learning_rate": 9.528886534724242e-06, + "loss": 0.5291, + "step": 7501 + }, + { + "epoch": 0.16513317191283294, + "grad_norm": 0.666745013391237, + "learning_rate": 9.528735473391445e-06, + "loss": 0.5407, + "step": 7502 + }, + { + "epoch": 0.1651551837992516, + "grad_norm": 2.497450756530545, + "learning_rate": 9.528584389041594e-06, + "loss": 0.3439, + "step": 7503 + }, + { + "epoch": 0.16517719568567027, + "grad_norm": 1.505710526247059, + "learning_rate": 9.528433281675457e-06, + "loss": 0.3351, + "step": 7504 + }, + { + "epoch": 0.16519920757208892, + "grad_norm": 1.5591254594316124, + "learning_rate": 9.528282151293804e-06, + "loss": 0.2829, + "step": 7505 + }, + { + "epoch": 0.1652212194585076, + "grad_norm": 0.6806318181229126, + "learning_rate": 9.528130997897401e-06, + "loss": 0.5606, + "step": 7506 + }, + { + "epoch": 0.16524323134492627, + "grad_norm": 1.9713792090692515, + "learning_rate": 9.527979821487015e-06, + "loss": 0.3206, + "step": 7507 + }, + { + "epoch": 0.16526524323134492, + "grad_norm": 0.6504132960403045, + "learning_rate": 9.527828622063418e-06, + "loss": 0.5287, + "step": 7508 + }, + { + "epoch": 0.1652872551177636, + "grad_norm": 5.980856723911117, + "learning_rate": 9.527677399627375e-06, + "loss": 0.3021, + "step": 7509 + }, + { + "epoch": 0.16530926700418225, + "grad_norm": 0.6370759744926456, + "learning_rate": 9.527526154179655e-06, + "loss": 0.5433, + "step": 7510 + }, + { + "epoch": 0.16533127889060092, + "grad_norm": 2.261054683641221, + "learning_rate": 9.52737488572103e-06, + "loss": 0.3583, + "step": 7511 + }, + { + "epoch": 0.1653532907770196, + "grad_norm": 2.0683883431362986, + "learning_rate": 9.527223594252267e-06, + "loss": 0.3591, + "step": 7512 + }, + { + "epoch": 0.16537530266343825, + "grad_norm": 7.816343668582655, + "learning_rate": 9.527072279774133e-06, + "loss": 0.4093, + "step": 7513 + }, + { + "epoch": 0.16539731454985693, + "grad_norm": 0.6771338120578136, + "learning_rate": 9.5269209422874e-06, + "loss": 0.5623, + "step": 7514 + }, + { + "epoch": 0.16541932643627558, + "grad_norm": 1.3922112947589267, + "learning_rate": 9.526769581792834e-06, + "loss": 0.3467, + "step": 7515 + }, + { + "epoch": 0.16544133832269425, + "grad_norm": 2.2097284276943223, + "learning_rate": 9.526618198291206e-06, + "loss": 0.3245, + "step": 7516 + }, + { + "epoch": 0.16546335020911293, + "grad_norm": 1.9015670912622153, + "learning_rate": 9.526466791783287e-06, + "loss": 0.455, + "step": 7517 + }, + { + "epoch": 0.16548536209553158, + "grad_norm": 1.3399621384408953, + "learning_rate": 9.526315362269842e-06, + "loss": 0.3866, + "step": 7518 + }, + { + "epoch": 0.16550737398195026, + "grad_norm": 2.476211512568027, + "learning_rate": 9.526163909751647e-06, + "loss": 0.3975, + "step": 7519 + }, + { + "epoch": 0.1655293858683689, + "grad_norm": 1.587131267340443, + "learning_rate": 9.526012434229466e-06, + "loss": 0.351, + "step": 7520 + }, + { + "epoch": 0.16555139775478758, + "grad_norm": 2.2876257039543235, + "learning_rate": 9.525860935704072e-06, + "loss": 0.3374, + "step": 7521 + }, + { + "epoch": 0.16557340964120626, + "grad_norm": 3.138520426679337, + "learning_rate": 9.525709414176233e-06, + "loss": 0.3433, + "step": 7522 + }, + { + "epoch": 0.1655954215276249, + "grad_norm": 2.1390954982897297, + "learning_rate": 9.525557869646723e-06, + "loss": 0.3313, + "step": 7523 + }, + { + "epoch": 0.16561743341404359, + "grad_norm": 2.404380946122192, + "learning_rate": 9.525406302116307e-06, + "loss": 0.3869, + "step": 7524 + }, + { + "epoch": 0.16563944530046226, + "grad_norm": 3.227966638249937, + "learning_rate": 9.525254711585758e-06, + "loss": 0.4082, + "step": 7525 + }, + { + "epoch": 0.1656614571868809, + "grad_norm": 1.6197276850165632, + "learning_rate": 9.525103098055847e-06, + "loss": 0.3988, + "step": 7526 + }, + { + "epoch": 0.1656834690732996, + "grad_norm": 3.1994450824615734, + "learning_rate": 9.524951461527344e-06, + "loss": 0.38, + "step": 7527 + }, + { + "epoch": 0.16570548095971824, + "grad_norm": 2.0768626812312885, + "learning_rate": 9.52479980200102e-06, + "loss": 0.3428, + "step": 7528 + }, + { + "epoch": 0.16572749284613691, + "grad_norm": 1.9893543531510032, + "learning_rate": 9.524648119477643e-06, + "loss": 0.3396, + "step": 7529 + }, + { + "epoch": 0.1657495047325556, + "grad_norm": 2.4269912322537786, + "learning_rate": 9.524496413957987e-06, + "loss": 0.284, + "step": 7530 + }, + { + "epoch": 0.16577151661897424, + "grad_norm": 1.98783803218628, + "learning_rate": 9.524344685442824e-06, + "loss": 0.3819, + "step": 7531 + }, + { + "epoch": 0.16579352850539292, + "grad_norm": 2.142874090438626, + "learning_rate": 9.524192933932921e-06, + "loss": 0.3748, + "step": 7532 + }, + { + "epoch": 0.16581554039181157, + "grad_norm": 1.6374890529410429, + "learning_rate": 9.524041159429052e-06, + "loss": 0.4172, + "step": 7533 + }, + { + "epoch": 0.16583755227823024, + "grad_norm": 1.603944937229655, + "learning_rate": 9.52388936193199e-06, + "loss": 0.392, + "step": 7534 + }, + { + "epoch": 0.16585956416464892, + "grad_norm": 2.0767433453350215, + "learning_rate": 9.523737541442504e-06, + "loss": 0.323, + "step": 7535 + }, + { + "epoch": 0.16588157605106757, + "grad_norm": 1.994844619189368, + "learning_rate": 9.523585697961364e-06, + "loss": 0.4503, + "step": 7536 + }, + { + "epoch": 0.16590358793748625, + "grad_norm": 1.4383646410838653, + "learning_rate": 9.523433831489347e-06, + "loss": 0.3523, + "step": 7537 + }, + { + "epoch": 0.1659255998239049, + "grad_norm": 0.6771847058159612, + "learning_rate": 9.52328194202722e-06, + "loss": 0.5709, + "step": 7538 + }, + { + "epoch": 0.16594761171032357, + "grad_norm": 1.6028756268059254, + "learning_rate": 9.523130029575755e-06, + "loss": 0.3657, + "step": 7539 + }, + { + "epoch": 0.16596962359674225, + "grad_norm": 4.087527085543681, + "learning_rate": 9.522978094135728e-06, + "loss": 0.3527, + "step": 7540 + }, + { + "epoch": 0.1659916354831609, + "grad_norm": 1.3222150991788821, + "learning_rate": 9.52282613570791e-06, + "loss": 0.2812, + "step": 7541 + }, + { + "epoch": 0.16601364736957958, + "grad_norm": 2.259342386796803, + "learning_rate": 9.52267415429307e-06, + "loss": 0.4366, + "step": 7542 + }, + { + "epoch": 0.16603565925599825, + "grad_norm": 0.6633178938588609, + "learning_rate": 9.522522149891985e-06, + "loss": 0.5384, + "step": 7543 + }, + { + "epoch": 0.1660576711424169, + "grad_norm": 1.3818541452786155, + "learning_rate": 9.522370122505425e-06, + "loss": 0.3803, + "step": 7544 + }, + { + "epoch": 0.16607968302883558, + "grad_norm": 1.6207367711166591, + "learning_rate": 9.522218072134162e-06, + "loss": 0.3774, + "step": 7545 + }, + { + "epoch": 0.16610169491525423, + "grad_norm": 5.788098774831431, + "learning_rate": 9.522065998778968e-06, + "loss": 0.3922, + "step": 7546 + }, + { + "epoch": 0.1661237068016729, + "grad_norm": 1.8251840583651218, + "learning_rate": 9.52191390244062e-06, + "loss": 0.3433, + "step": 7547 + }, + { + "epoch": 0.16614571868809158, + "grad_norm": 2.0339098064424337, + "learning_rate": 9.521761783119889e-06, + "loss": 0.412, + "step": 7548 + }, + { + "epoch": 0.16616773057451023, + "grad_norm": 1.7335854552029406, + "learning_rate": 9.521609640817548e-06, + "loss": 0.3111, + "step": 7549 + }, + { + "epoch": 0.1661897424609289, + "grad_norm": 2.724156944422121, + "learning_rate": 9.521457475534369e-06, + "loss": 0.381, + "step": 7550 + }, + { + "epoch": 0.16621175434734756, + "grad_norm": 1.9228868067828346, + "learning_rate": 9.521305287271128e-06, + "loss": 0.3765, + "step": 7551 + }, + { + "epoch": 0.16623376623376623, + "grad_norm": 1.3595884578989443, + "learning_rate": 9.521153076028596e-06, + "loss": 0.2755, + "step": 7552 + }, + { + "epoch": 0.1662557781201849, + "grad_norm": 1.6182086246293654, + "learning_rate": 9.521000841807546e-06, + "loss": 0.2734, + "step": 7553 + }, + { + "epoch": 0.16627779000660356, + "grad_norm": 1.4591456909954243, + "learning_rate": 9.520848584608755e-06, + "loss": 0.3474, + "step": 7554 + }, + { + "epoch": 0.16629980189302224, + "grad_norm": 3.748391967124075, + "learning_rate": 9.520696304432995e-06, + "loss": 0.4107, + "step": 7555 + }, + { + "epoch": 0.16632181377944089, + "grad_norm": 4.1213309111854475, + "learning_rate": 9.520544001281041e-06, + "loss": 0.3813, + "step": 7556 + }, + { + "epoch": 0.16634382566585956, + "grad_norm": 1.5264350152232822, + "learning_rate": 9.520391675153664e-06, + "loss": 0.2358, + "step": 7557 + }, + { + "epoch": 0.16636583755227824, + "grad_norm": 1.6677461365343151, + "learning_rate": 9.52023932605164e-06, + "loss": 0.4213, + "step": 7558 + }, + { + "epoch": 0.1663878494386969, + "grad_norm": 3.617065276761783, + "learning_rate": 9.520086953975747e-06, + "loss": 0.2864, + "step": 7559 + }, + { + "epoch": 0.16640986132511557, + "grad_norm": 1.725181329259536, + "learning_rate": 9.519934558926755e-06, + "loss": 0.3629, + "step": 7560 + }, + { + "epoch": 0.16643187321153421, + "grad_norm": 2.1033686919388046, + "learning_rate": 9.519782140905437e-06, + "loss": 0.41, + "step": 7561 + }, + { + "epoch": 0.1664538850979529, + "grad_norm": 0.8256670220175334, + "learning_rate": 9.519629699912574e-06, + "loss": 0.5367, + "step": 7562 + }, + { + "epoch": 0.16647589698437157, + "grad_norm": 1.5321784827952731, + "learning_rate": 9.519477235948935e-06, + "loss": 0.4565, + "step": 7563 + }, + { + "epoch": 0.16649790887079022, + "grad_norm": 1.3827040971998064, + "learning_rate": 9.519324749015297e-06, + "loss": 0.369, + "step": 7564 + }, + { + "epoch": 0.1665199207572089, + "grad_norm": 1.8363046811720527, + "learning_rate": 9.519172239112436e-06, + "loss": 0.3087, + "step": 7565 + }, + { + "epoch": 0.16654193264362757, + "grad_norm": 0.6536467818522509, + "learning_rate": 9.519019706241123e-06, + "loss": 0.5452, + "step": 7566 + }, + { + "epoch": 0.16656394453004622, + "grad_norm": 1.975257697578898, + "learning_rate": 9.51886715040214e-06, + "loss": 0.2661, + "step": 7567 + }, + { + "epoch": 0.1665859564164649, + "grad_norm": 1.5436400436956534, + "learning_rate": 9.518714571596257e-06, + "loss": 0.3943, + "step": 7568 + }, + { + "epoch": 0.16660796830288355, + "grad_norm": 0.621276467950537, + "learning_rate": 9.518561969824251e-06, + "loss": 0.5339, + "step": 7569 + }, + { + "epoch": 0.16662998018930222, + "grad_norm": 3.855407554902105, + "learning_rate": 9.5184093450869e-06, + "loss": 0.3993, + "step": 7570 + }, + { + "epoch": 0.1666519920757209, + "grad_norm": 1.6295843146432116, + "learning_rate": 9.518256697384973e-06, + "loss": 0.3862, + "step": 7571 + }, + { + "epoch": 0.16667400396213955, + "grad_norm": 1.4896709232326941, + "learning_rate": 9.518104026719253e-06, + "loss": 0.3899, + "step": 7572 + }, + { + "epoch": 0.16669601584855823, + "grad_norm": 2.021325852237335, + "learning_rate": 9.517951333090515e-06, + "loss": 0.3453, + "step": 7573 + }, + { + "epoch": 0.16671802773497688, + "grad_norm": 2.16192531282776, + "learning_rate": 9.517798616499529e-06, + "loss": 0.3651, + "step": 7574 + }, + { + "epoch": 0.16674003962139555, + "grad_norm": 1.7082552415637078, + "learning_rate": 9.51764587694708e-06, + "loss": 0.3249, + "step": 7575 + }, + { + "epoch": 0.16676205150781423, + "grad_norm": 2.0349705350005496, + "learning_rate": 9.517493114433935e-06, + "loss": 0.2971, + "step": 7576 + }, + { + "epoch": 0.16678406339423288, + "grad_norm": 1.4390706372805084, + "learning_rate": 9.517340328960879e-06, + "loss": 0.348, + "step": 7577 + }, + { + "epoch": 0.16680607528065156, + "grad_norm": 0.677061301706888, + "learning_rate": 9.517187520528682e-06, + "loss": 0.5829, + "step": 7578 + }, + { + "epoch": 0.1668280871670702, + "grad_norm": 1.8867760879915543, + "learning_rate": 9.517034689138125e-06, + "loss": 0.3169, + "step": 7579 + }, + { + "epoch": 0.16685009905348888, + "grad_norm": 2.2984255956144, + "learning_rate": 9.516881834789982e-06, + "loss": 0.2986, + "step": 7580 + }, + { + "epoch": 0.16687211093990756, + "grad_norm": 1.2945530359591977, + "learning_rate": 9.516728957485032e-06, + "loss": 0.3013, + "step": 7581 + }, + { + "epoch": 0.1668941228263262, + "grad_norm": 2.2309520090247497, + "learning_rate": 9.516576057224049e-06, + "loss": 0.4049, + "step": 7582 + }, + { + "epoch": 0.16691613471274488, + "grad_norm": 1.5298942682166279, + "learning_rate": 9.516423134007814e-06, + "loss": 0.3754, + "step": 7583 + }, + { + "epoch": 0.16693814659916356, + "grad_norm": 1.3217094899366135, + "learning_rate": 9.516270187837102e-06, + "loss": 0.2727, + "step": 7584 + }, + { + "epoch": 0.1669601584855822, + "grad_norm": 6.974597520658152, + "learning_rate": 9.51611721871269e-06, + "loss": 0.4227, + "step": 7585 + }, + { + "epoch": 0.1669821703720009, + "grad_norm": 1.3738154684550192, + "learning_rate": 9.515964226635357e-06, + "loss": 0.2949, + "step": 7586 + }, + { + "epoch": 0.16700418225841954, + "grad_norm": 1.679378924276962, + "learning_rate": 9.515811211605879e-06, + "loss": 0.3653, + "step": 7587 + }, + { + "epoch": 0.1670261941448382, + "grad_norm": 1.94517227037071, + "learning_rate": 9.515658173625035e-06, + "loss": 0.3841, + "step": 7588 + }, + { + "epoch": 0.1670482060312569, + "grad_norm": 1.6594960946629194, + "learning_rate": 9.515505112693601e-06, + "loss": 0.3671, + "step": 7589 + }, + { + "epoch": 0.16707021791767554, + "grad_norm": 1.6524765922796842, + "learning_rate": 9.515352028812357e-06, + "loss": 0.446, + "step": 7590 + }, + { + "epoch": 0.16709222980409422, + "grad_norm": 2.1013238970627937, + "learning_rate": 9.515198921982078e-06, + "loss": 0.3478, + "step": 7591 + }, + { + "epoch": 0.16711424169051287, + "grad_norm": 3.0485462653682305, + "learning_rate": 9.515045792203545e-06, + "loss": 0.3442, + "step": 7592 + }, + { + "epoch": 0.16713625357693154, + "grad_norm": 1.7643591219443822, + "learning_rate": 9.514892639477538e-06, + "loss": 0.4627, + "step": 7593 + }, + { + "epoch": 0.16715826546335022, + "grad_norm": 1.5724645838344593, + "learning_rate": 9.51473946380483e-06, + "loss": 0.3583, + "step": 7594 + }, + { + "epoch": 0.16718027734976887, + "grad_norm": 2.9515817162540388, + "learning_rate": 9.514586265186204e-06, + "loss": 0.3628, + "step": 7595 + }, + { + "epoch": 0.16720228923618755, + "grad_norm": 0.7005676346295929, + "learning_rate": 9.514433043622438e-06, + "loss": 0.5449, + "step": 7596 + }, + { + "epoch": 0.1672243011226062, + "grad_norm": 1.412218530745132, + "learning_rate": 9.514279799114307e-06, + "loss": 0.3827, + "step": 7597 + }, + { + "epoch": 0.16724631300902487, + "grad_norm": 1.941342690210329, + "learning_rate": 9.514126531662595e-06, + "loss": 0.3588, + "step": 7598 + }, + { + "epoch": 0.16726832489544355, + "grad_norm": 1.3839165064039294, + "learning_rate": 9.513973241268077e-06, + "loss": 0.3325, + "step": 7599 + }, + { + "epoch": 0.1672903367818622, + "grad_norm": 1.6277808006169674, + "learning_rate": 9.513819927931535e-06, + "loss": 0.3189, + "step": 7600 + }, + { + "epoch": 0.16731234866828087, + "grad_norm": 1.754879127725012, + "learning_rate": 9.513666591653746e-06, + "loss": 0.388, + "step": 7601 + }, + { + "epoch": 0.16733436055469955, + "grad_norm": 1.480292043699826, + "learning_rate": 9.51351323243549e-06, + "loss": 0.3416, + "step": 7602 + }, + { + "epoch": 0.1673563724411182, + "grad_norm": 1.4877041294511506, + "learning_rate": 9.513359850277548e-06, + "loss": 0.3818, + "step": 7603 + }, + { + "epoch": 0.16737838432753688, + "grad_norm": 1.350203856698514, + "learning_rate": 9.513206445180695e-06, + "loss": 0.2796, + "step": 7604 + }, + { + "epoch": 0.16740039621395553, + "grad_norm": 1.746352700750205, + "learning_rate": 9.513053017145717e-06, + "loss": 0.3492, + "step": 7605 + }, + { + "epoch": 0.1674224081003742, + "grad_norm": 1.7494694790911023, + "learning_rate": 9.51289956617339e-06, + "loss": 0.3439, + "step": 7606 + }, + { + "epoch": 0.16744441998679288, + "grad_norm": 1.399534335886712, + "learning_rate": 9.512746092264496e-06, + "loss": 0.3307, + "step": 7607 + }, + { + "epoch": 0.16746643187321153, + "grad_norm": 24.019202568517976, + "learning_rate": 9.512592595419812e-06, + "loss": 0.2835, + "step": 7608 + }, + { + "epoch": 0.1674884437596302, + "grad_norm": 1.846904753573619, + "learning_rate": 9.512439075640119e-06, + "loss": 0.3168, + "step": 7609 + }, + { + "epoch": 0.16751045564604886, + "grad_norm": 1.8984042099885454, + "learning_rate": 9.5122855329262e-06, + "loss": 0.2796, + "step": 7610 + }, + { + "epoch": 0.16753246753246753, + "grad_norm": 1.6704193488154775, + "learning_rate": 9.512131967278832e-06, + "loss": 0.3671, + "step": 7611 + }, + { + "epoch": 0.1675544794188862, + "grad_norm": 2.8668701018417537, + "learning_rate": 9.511978378698798e-06, + "loss": 0.4211, + "step": 7612 + }, + { + "epoch": 0.16757649130530486, + "grad_norm": 0.7197041551166374, + "learning_rate": 9.511824767186877e-06, + "loss": 0.5382, + "step": 7613 + }, + { + "epoch": 0.16759850319172354, + "grad_norm": 2.2955528228184243, + "learning_rate": 9.511671132743849e-06, + "loss": 0.3558, + "step": 7614 + }, + { + "epoch": 0.16762051507814218, + "grad_norm": 1.5539567962240257, + "learning_rate": 9.511517475370498e-06, + "loss": 0.3414, + "step": 7615 + }, + { + "epoch": 0.16764252696456086, + "grad_norm": 1.8972131112783035, + "learning_rate": 9.5113637950676e-06, + "loss": 0.3433, + "step": 7616 + }, + { + "epoch": 0.16766453885097954, + "grad_norm": 0.6550677592771639, + "learning_rate": 9.511210091835941e-06, + "loss": 0.5506, + "step": 7617 + }, + { + "epoch": 0.1676865507373982, + "grad_norm": 1.6678957127648333, + "learning_rate": 9.511056365676301e-06, + "loss": 0.3239, + "step": 7618 + }, + { + "epoch": 0.16770856262381686, + "grad_norm": 2.1687276326332734, + "learning_rate": 9.51090261658946e-06, + "loss": 0.3907, + "step": 7619 + }, + { + "epoch": 0.16773057451023551, + "grad_norm": 1.1565287136734614, + "learning_rate": 9.5107488445762e-06, + "loss": 0.2438, + "step": 7620 + }, + { + "epoch": 0.1677525863966542, + "grad_norm": 1.8294121038497702, + "learning_rate": 9.510595049637303e-06, + "loss": 0.3751, + "step": 7621 + }, + { + "epoch": 0.16777459828307287, + "grad_norm": 2.1650066058407544, + "learning_rate": 9.51044123177355e-06, + "loss": 0.4436, + "step": 7622 + }, + { + "epoch": 0.16779661016949152, + "grad_norm": 2.052605258232392, + "learning_rate": 9.510287390985721e-06, + "loss": 0.3933, + "step": 7623 + }, + { + "epoch": 0.1678186220559102, + "grad_norm": 0.735558119758196, + "learning_rate": 9.5101335272746e-06, + "loss": 0.5472, + "step": 7624 + }, + { + "epoch": 0.16784063394232887, + "grad_norm": 1.9428485833574134, + "learning_rate": 9.509979640640972e-06, + "loss": 0.4607, + "step": 7625 + }, + { + "epoch": 0.16786264582874752, + "grad_norm": 1.5901684837613492, + "learning_rate": 9.509825731085614e-06, + "loss": 0.332, + "step": 7626 + }, + { + "epoch": 0.1678846577151662, + "grad_norm": 1.6466959918055937, + "learning_rate": 9.50967179860931e-06, + "loss": 0.4489, + "step": 7627 + }, + { + "epoch": 0.16790666960158485, + "grad_norm": 1.5696077463432354, + "learning_rate": 9.509517843212843e-06, + "loss": 0.4136, + "step": 7628 + }, + { + "epoch": 0.16792868148800352, + "grad_norm": 3.913792919937396, + "learning_rate": 9.509363864896994e-06, + "loss": 0.3297, + "step": 7629 + }, + { + "epoch": 0.1679506933744222, + "grad_norm": 1.7313237989297363, + "learning_rate": 9.509209863662547e-06, + "loss": 0.4078, + "step": 7630 + }, + { + "epoch": 0.16797270526084085, + "grad_norm": 6.165403304952929, + "learning_rate": 9.509055839510285e-06, + "loss": 0.3423, + "step": 7631 + }, + { + "epoch": 0.16799471714725953, + "grad_norm": 2.0438837511369607, + "learning_rate": 9.50890179244099e-06, + "loss": 0.3785, + "step": 7632 + }, + { + "epoch": 0.16801672903367817, + "grad_norm": 1.561040946575447, + "learning_rate": 9.508747722455446e-06, + "loss": 0.368, + "step": 7633 + }, + { + "epoch": 0.16803874092009685, + "grad_norm": 2.889281383964944, + "learning_rate": 9.508593629554433e-06, + "loss": 0.337, + "step": 7634 + }, + { + "epoch": 0.16806075280651553, + "grad_norm": 1.4581799774983772, + "learning_rate": 9.508439513738739e-06, + "loss": 0.37, + "step": 7635 + }, + { + "epoch": 0.16808276469293418, + "grad_norm": 0.6771163028117424, + "learning_rate": 9.508285375009142e-06, + "loss": 0.5221, + "step": 7636 + }, + { + "epoch": 0.16810477657935285, + "grad_norm": 3.0019749792760075, + "learning_rate": 9.50813121336643e-06, + "loss": 0.2919, + "step": 7637 + }, + { + "epoch": 0.1681267884657715, + "grad_norm": 1.9552384048994589, + "learning_rate": 9.507977028811382e-06, + "loss": 0.3698, + "step": 7638 + }, + { + "epoch": 0.16814880035219018, + "grad_norm": 2.1071261608246097, + "learning_rate": 9.507822821344785e-06, + "loss": 0.3434, + "step": 7639 + }, + { + "epoch": 0.16817081223860886, + "grad_norm": 2.161543558233318, + "learning_rate": 9.507668590967423e-06, + "loss": 0.357, + "step": 7640 + }, + { + "epoch": 0.1681928241250275, + "grad_norm": 1.6790452831918754, + "learning_rate": 9.507514337680078e-06, + "loss": 0.3122, + "step": 7641 + }, + { + "epoch": 0.16821483601144618, + "grad_norm": 1.32083521572461, + "learning_rate": 9.507360061483535e-06, + "loss": 0.3261, + "step": 7642 + }, + { + "epoch": 0.16823684789786486, + "grad_norm": 0.6808381584724525, + "learning_rate": 9.507205762378577e-06, + "loss": 0.5411, + "step": 7643 + }, + { + "epoch": 0.1682588597842835, + "grad_norm": 1.64107244684042, + "learning_rate": 9.50705144036599e-06, + "loss": 0.3691, + "step": 7644 + }, + { + "epoch": 0.1682808716707022, + "grad_norm": 2.271285958038637, + "learning_rate": 9.506897095446558e-06, + "loss": 0.2979, + "step": 7645 + }, + { + "epoch": 0.16830288355712084, + "grad_norm": 3.144306755695857, + "learning_rate": 9.506742727621063e-06, + "loss": 0.3541, + "step": 7646 + }, + { + "epoch": 0.1683248954435395, + "grad_norm": 1.7969988591216521, + "learning_rate": 9.506588336890294e-06, + "loss": 0.3388, + "step": 7647 + }, + { + "epoch": 0.1683469073299582, + "grad_norm": 1.495293504118872, + "learning_rate": 9.506433923255031e-06, + "loss": 0.3421, + "step": 7648 + }, + { + "epoch": 0.16836891921637684, + "grad_norm": 2.0900327207207483, + "learning_rate": 9.506279486716062e-06, + "loss": 0.3191, + "step": 7649 + }, + { + "epoch": 0.16839093110279552, + "grad_norm": 1.5522021745436119, + "learning_rate": 9.506125027274168e-06, + "loss": 0.4269, + "step": 7650 + }, + { + "epoch": 0.16841294298921416, + "grad_norm": 0.6864421094727555, + "learning_rate": 9.50597054493014e-06, + "loss": 0.5644, + "step": 7651 + }, + { + "epoch": 0.16843495487563284, + "grad_norm": 1.44641510187082, + "learning_rate": 9.50581603968476e-06, + "loss": 0.2881, + "step": 7652 + }, + { + "epoch": 0.16845696676205152, + "grad_norm": 1.5155475634154711, + "learning_rate": 9.50566151153881e-06, + "loss": 0.3508, + "step": 7653 + }, + { + "epoch": 0.16847897864847017, + "grad_norm": 1.6701738219667726, + "learning_rate": 9.505506960493082e-06, + "loss": 0.3675, + "step": 7654 + }, + { + "epoch": 0.16850099053488884, + "grad_norm": 1.7037556549252901, + "learning_rate": 9.505352386548357e-06, + "loss": 0.3679, + "step": 7655 + }, + { + "epoch": 0.1685230024213075, + "grad_norm": 1.480579263549985, + "learning_rate": 9.50519778970542e-06, + "loss": 0.37, + "step": 7656 + }, + { + "epoch": 0.16854501430772617, + "grad_norm": 0.6103098425132062, + "learning_rate": 9.505043169965061e-06, + "loss": 0.5458, + "step": 7657 + }, + { + "epoch": 0.16856702619414485, + "grad_norm": 1.396821336712218, + "learning_rate": 9.504888527328062e-06, + "loss": 0.3816, + "step": 7658 + }, + { + "epoch": 0.1685890380805635, + "grad_norm": 5.135734781313229, + "learning_rate": 9.504733861795211e-06, + "loss": 0.3824, + "step": 7659 + }, + { + "epoch": 0.16861104996698217, + "grad_norm": 1.4044053908050174, + "learning_rate": 9.504579173367293e-06, + "loss": 0.3704, + "step": 7660 + }, + { + "epoch": 0.16863306185340082, + "grad_norm": 2.4034162555886183, + "learning_rate": 9.504424462045094e-06, + "loss": 0.4377, + "step": 7661 + }, + { + "epoch": 0.1686550737398195, + "grad_norm": 1.8236617143834057, + "learning_rate": 9.504269727829403e-06, + "loss": 0.413, + "step": 7662 + }, + { + "epoch": 0.16867708562623818, + "grad_norm": 1.9767610518496075, + "learning_rate": 9.504114970721e-06, + "loss": 0.395, + "step": 7663 + }, + { + "epoch": 0.16869909751265683, + "grad_norm": 1.1853488641004022, + "learning_rate": 9.50396019072068e-06, + "loss": 0.2208, + "step": 7664 + }, + { + "epoch": 0.1687211093990755, + "grad_norm": 2.096882208700831, + "learning_rate": 9.503805387829223e-06, + "loss": 0.3699, + "step": 7665 + }, + { + "epoch": 0.16874312128549418, + "grad_norm": 0.7319508514696961, + "learning_rate": 9.50365056204742e-06, + "loss": 0.521, + "step": 7666 + }, + { + "epoch": 0.16876513317191283, + "grad_norm": 2.426769657022523, + "learning_rate": 9.503495713376057e-06, + "loss": 0.3421, + "step": 7667 + }, + { + "epoch": 0.1687871450583315, + "grad_norm": 1.6274560310836497, + "learning_rate": 9.503340841815918e-06, + "loss": 0.3969, + "step": 7668 + }, + { + "epoch": 0.16880915694475015, + "grad_norm": 2.012063628340158, + "learning_rate": 9.503185947367793e-06, + "loss": 0.3619, + "step": 7669 + }, + { + "epoch": 0.16883116883116883, + "grad_norm": 1.3201741015239004, + "learning_rate": 9.503031030032468e-06, + "loss": 0.3577, + "step": 7670 + }, + { + "epoch": 0.1688531807175875, + "grad_norm": 1.8139867652043506, + "learning_rate": 9.50287608981073e-06, + "loss": 0.3686, + "step": 7671 + }, + { + "epoch": 0.16887519260400616, + "grad_norm": 1.6827649969218499, + "learning_rate": 9.50272112670337e-06, + "loss": 0.4158, + "step": 7672 + }, + { + "epoch": 0.16889720449042483, + "grad_norm": 1.3847388144667858, + "learning_rate": 9.502566140711171e-06, + "loss": 0.3868, + "step": 7673 + }, + { + "epoch": 0.16891921637684348, + "grad_norm": 0.7173322839544941, + "learning_rate": 9.502411131834923e-06, + "loss": 0.5463, + "step": 7674 + }, + { + "epoch": 0.16894122826326216, + "grad_norm": 0.651637609325439, + "learning_rate": 9.502256100075415e-06, + "loss": 0.5525, + "step": 7675 + }, + { + "epoch": 0.16896324014968084, + "grad_norm": 0.6556731418614056, + "learning_rate": 9.502101045433432e-06, + "loss": 0.5652, + "step": 7676 + }, + { + "epoch": 0.1689852520360995, + "grad_norm": 1.4377043558837896, + "learning_rate": 9.501945967909763e-06, + "loss": 0.4039, + "step": 7677 + }, + { + "epoch": 0.16900726392251816, + "grad_norm": 0.6299979296561612, + "learning_rate": 9.501790867505199e-06, + "loss": 0.5428, + "step": 7678 + }, + { + "epoch": 0.1690292758089368, + "grad_norm": 1.8775023563773319, + "learning_rate": 9.501635744220523e-06, + "loss": 0.3773, + "step": 7679 + }, + { + "epoch": 0.1690512876953555, + "grad_norm": 2.2708078860540652, + "learning_rate": 9.501480598056527e-06, + "loss": 0.2755, + "step": 7680 + }, + { + "epoch": 0.16907329958177417, + "grad_norm": 1.5977449052092174, + "learning_rate": 9.501325429014e-06, + "loss": 0.3933, + "step": 7681 + }, + { + "epoch": 0.16909531146819282, + "grad_norm": 1.2332487462094597, + "learning_rate": 9.501170237093729e-06, + "loss": 0.3137, + "step": 7682 + }, + { + "epoch": 0.1691173233546115, + "grad_norm": 1.266766965805448, + "learning_rate": 9.501015022296503e-06, + "loss": 0.3145, + "step": 7683 + }, + { + "epoch": 0.16913933524103017, + "grad_norm": 1.6174856474476793, + "learning_rate": 9.500859784623113e-06, + "loss": 0.3508, + "step": 7684 + }, + { + "epoch": 0.16916134712744882, + "grad_norm": 1.4899589718198922, + "learning_rate": 9.500704524074344e-06, + "loss": 0.4192, + "step": 7685 + }, + { + "epoch": 0.1691833590138675, + "grad_norm": 1.6975119801808407, + "learning_rate": 9.500549240650989e-06, + "loss": 0.3307, + "step": 7686 + }, + { + "epoch": 0.16920537090028615, + "grad_norm": 1.3600493928281534, + "learning_rate": 9.500393934353834e-06, + "loss": 0.4262, + "step": 7687 + }, + { + "epoch": 0.16922738278670482, + "grad_norm": 1.5146702617125956, + "learning_rate": 9.50023860518367e-06, + "loss": 0.3519, + "step": 7688 + }, + { + "epoch": 0.1692493946731235, + "grad_norm": 2.7239322712705065, + "learning_rate": 9.500083253141286e-06, + "loss": 0.3541, + "step": 7689 + }, + { + "epoch": 0.16927140655954215, + "grad_norm": 0.9729400833685558, + "learning_rate": 9.499927878227472e-06, + "loss": 0.5522, + "step": 7690 + }, + { + "epoch": 0.16929341844596082, + "grad_norm": 1.5101390097220624, + "learning_rate": 9.499772480443019e-06, + "loss": 0.3929, + "step": 7691 + }, + { + "epoch": 0.16931543033237947, + "grad_norm": 1.3379262504707503, + "learning_rate": 9.499617059788713e-06, + "loss": 0.2506, + "step": 7692 + }, + { + "epoch": 0.16933744221879815, + "grad_norm": 2.48969796104482, + "learning_rate": 9.499461616265346e-06, + "loss": 0.2845, + "step": 7693 + }, + { + "epoch": 0.16935945410521683, + "grad_norm": 0.6738787057970165, + "learning_rate": 9.49930614987371e-06, + "loss": 0.5476, + "step": 7694 + }, + { + "epoch": 0.16938146599163548, + "grad_norm": 0.6605007258259122, + "learning_rate": 9.499150660614594e-06, + "loss": 0.5705, + "step": 7695 + }, + { + "epoch": 0.16940347787805415, + "grad_norm": 0.6408434813459903, + "learning_rate": 9.498995148488785e-06, + "loss": 0.5163, + "step": 7696 + }, + { + "epoch": 0.1694254897644728, + "grad_norm": 9.112304783854784, + "learning_rate": 9.498839613497078e-06, + "loss": 0.2987, + "step": 7697 + }, + { + "epoch": 0.16944750165089148, + "grad_norm": 1.7390371646148657, + "learning_rate": 9.498684055640262e-06, + "loss": 0.3768, + "step": 7698 + }, + { + "epoch": 0.16946951353731016, + "grad_norm": 1.5776151349728111, + "learning_rate": 9.498528474919126e-06, + "loss": 0.3372, + "step": 7699 + }, + { + "epoch": 0.1694915254237288, + "grad_norm": 1.444920611957428, + "learning_rate": 9.498372871334462e-06, + "loss": 0.3874, + "step": 7700 + }, + { + "epoch": 0.16951353731014748, + "grad_norm": 1.587910212567783, + "learning_rate": 9.49821724488706e-06, + "loss": 0.4142, + "step": 7701 + }, + { + "epoch": 0.16953554919656613, + "grad_norm": 0.6886660413276783, + "learning_rate": 9.498061595577715e-06, + "loss": 0.5365, + "step": 7702 + }, + { + "epoch": 0.1695575610829848, + "grad_norm": 1.7424554336439, + "learning_rate": 9.49790592340721e-06, + "loss": 0.4088, + "step": 7703 + }, + { + "epoch": 0.16957957296940349, + "grad_norm": 0.6930376600498802, + "learning_rate": 9.497750228376346e-06, + "loss": 0.5315, + "step": 7704 + }, + { + "epoch": 0.16960158485582214, + "grad_norm": 0.6464525327855801, + "learning_rate": 9.497594510485906e-06, + "loss": 0.5291, + "step": 7705 + }, + { + "epoch": 0.1696235967422408, + "grad_norm": 2.603986215519123, + "learning_rate": 9.497438769736685e-06, + "loss": 0.2948, + "step": 7706 + }, + { + "epoch": 0.1696456086286595, + "grad_norm": 0.6668087096866736, + "learning_rate": 9.497283006129474e-06, + "loss": 0.5314, + "step": 7707 + }, + { + "epoch": 0.16966762051507814, + "grad_norm": 1.5724807703414239, + "learning_rate": 9.497127219665066e-06, + "loss": 0.2396, + "step": 7708 + }, + { + "epoch": 0.16968963240149681, + "grad_norm": 1.504536001913497, + "learning_rate": 9.49697141034425e-06, + "loss": 0.3103, + "step": 7709 + }, + { + "epoch": 0.16971164428791546, + "grad_norm": 1.7440813159820923, + "learning_rate": 9.496815578167822e-06, + "loss": 0.4315, + "step": 7710 + }, + { + "epoch": 0.16973365617433414, + "grad_norm": 1.3607255624063237, + "learning_rate": 9.49665972313657e-06, + "loss": 0.418, + "step": 7711 + }, + { + "epoch": 0.16975566806075282, + "grad_norm": 1.8425177654909541, + "learning_rate": 9.49650384525129e-06, + "loss": 0.2775, + "step": 7712 + }, + { + "epoch": 0.16977767994717147, + "grad_norm": 1.3621709087808236, + "learning_rate": 9.49634794451277e-06, + "loss": 0.2954, + "step": 7713 + }, + { + "epoch": 0.16979969183359014, + "grad_norm": 1.69325566306121, + "learning_rate": 9.496192020921804e-06, + "loss": 0.2863, + "step": 7714 + }, + { + "epoch": 0.1698217037200088, + "grad_norm": 2.050142792323312, + "learning_rate": 9.496036074479184e-06, + "loss": 0.2754, + "step": 7715 + }, + { + "epoch": 0.16984371560642747, + "grad_norm": 1.9162377535775867, + "learning_rate": 9.495880105185705e-06, + "loss": 0.328, + "step": 7716 + }, + { + "epoch": 0.16986572749284615, + "grad_norm": 1.426523772129377, + "learning_rate": 9.495724113042158e-06, + "loss": 0.3421, + "step": 7717 + }, + { + "epoch": 0.1698877393792648, + "grad_norm": 2.5505281771034336, + "learning_rate": 9.495568098049336e-06, + "loss": 0.4216, + "step": 7718 + }, + { + "epoch": 0.16990975126568347, + "grad_norm": 2.7087255291301124, + "learning_rate": 9.495412060208032e-06, + "loss": 0.4189, + "step": 7719 + }, + { + "epoch": 0.16993176315210212, + "grad_norm": 2.95658261758375, + "learning_rate": 9.495255999519038e-06, + "loss": 0.4346, + "step": 7720 + }, + { + "epoch": 0.1699537750385208, + "grad_norm": 1.6862532127956358, + "learning_rate": 9.495099915983147e-06, + "loss": 0.3666, + "step": 7721 + }, + { + "epoch": 0.16997578692493948, + "grad_norm": 0.8163920532331915, + "learning_rate": 9.494943809601156e-06, + "loss": 0.5471, + "step": 7722 + }, + { + "epoch": 0.16999779881135813, + "grad_norm": 2.3681455131659703, + "learning_rate": 9.494787680373854e-06, + "loss": 0.3931, + "step": 7723 + }, + { + "epoch": 0.1700198106977768, + "grad_norm": 1.7222458500231181, + "learning_rate": 9.494631528302037e-06, + "loss": 0.2812, + "step": 7724 + }, + { + "epoch": 0.17004182258419548, + "grad_norm": 0.6923844905644703, + "learning_rate": 9.494475353386498e-06, + "loss": 0.5491, + "step": 7725 + }, + { + "epoch": 0.17006383447061413, + "grad_norm": 0.6263848400398612, + "learning_rate": 9.49431915562803e-06, + "loss": 0.5181, + "step": 7726 + }, + { + "epoch": 0.1700858463570328, + "grad_norm": 2.6787820690645354, + "learning_rate": 9.494162935027428e-06, + "loss": 0.3422, + "step": 7727 + }, + { + "epoch": 0.17010785824345145, + "grad_norm": 1.8023341501385433, + "learning_rate": 9.494006691585485e-06, + "loss": 0.4201, + "step": 7728 + }, + { + "epoch": 0.17012987012987013, + "grad_norm": 0.6676291539059295, + "learning_rate": 9.493850425302997e-06, + "loss": 0.5251, + "step": 7729 + }, + { + "epoch": 0.1701518820162888, + "grad_norm": 1.4042902106534747, + "learning_rate": 9.493694136180755e-06, + "loss": 0.3055, + "step": 7730 + }, + { + "epoch": 0.17017389390270746, + "grad_norm": 1.5306507817359933, + "learning_rate": 9.493537824219556e-06, + "loss": 0.4636, + "step": 7731 + }, + { + "epoch": 0.17019590578912613, + "grad_norm": 2.010557592608827, + "learning_rate": 9.493381489420195e-06, + "loss": 0.3738, + "step": 7732 + }, + { + "epoch": 0.17021791767554478, + "grad_norm": 3.2550386215369396, + "learning_rate": 9.493225131783464e-06, + "loss": 0.2872, + "step": 7733 + }, + { + "epoch": 0.17023992956196346, + "grad_norm": 2.289811359978017, + "learning_rate": 9.493068751310157e-06, + "loss": 0.3971, + "step": 7734 + }, + { + "epoch": 0.17026194144838214, + "grad_norm": 2.067135756394397, + "learning_rate": 9.492912348001074e-06, + "loss": 0.3025, + "step": 7735 + }, + { + "epoch": 0.17028395333480079, + "grad_norm": 0.7104350874610366, + "learning_rate": 9.492755921857004e-06, + "loss": 0.5308, + "step": 7736 + }, + { + "epoch": 0.17030596522121946, + "grad_norm": 1.9729178306528403, + "learning_rate": 9.492599472878745e-06, + "loss": 0.3768, + "step": 7737 + }, + { + "epoch": 0.1703279771076381, + "grad_norm": 1.6298658164363231, + "learning_rate": 9.492443001067094e-06, + "loss": 0.3601, + "step": 7738 + }, + { + "epoch": 0.1703499889940568, + "grad_norm": 0.6473681547226117, + "learning_rate": 9.492286506422841e-06, + "loss": 0.5298, + "step": 7739 + }, + { + "epoch": 0.17037200088047547, + "grad_norm": 1.64894912565475, + "learning_rate": 9.492129988946785e-06, + "loss": 0.3276, + "step": 7740 + }, + { + "epoch": 0.17039401276689412, + "grad_norm": 1.6952451383683578, + "learning_rate": 9.491973448639723e-06, + "loss": 0.3663, + "step": 7741 + }, + { + "epoch": 0.1704160246533128, + "grad_norm": 1.6769104186689163, + "learning_rate": 9.491816885502444e-06, + "loss": 0.3453, + "step": 7742 + }, + { + "epoch": 0.17043803653973144, + "grad_norm": 1.849281335996857, + "learning_rate": 9.491660299535753e-06, + "loss": 0.3963, + "step": 7743 + }, + { + "epoch": 0.17046004842615012, + "grad_norm": 1.4783696453948387, + "learning_rate": 9.491503690740437e-06, + "loss": 0.3556, + "step": 7744 + }, + { + "epoch": 0.1704820603125688, + "grad_norm": 1.4886251579861278, + "learning_rate": 9.4913470591173e-06, + "loss": 0.3647, + "step": 7745 + }, + { + "epoch": 0.17050407219898744, + "grad_norm": 1.482213359110978, + "learning_rate": 9.491190404667132e-06, + "loss": 0.4529, + "step": 7746 + }, + { + "epoch": 0.17052608408540612, + "grad_norm": 1.5380382305149318, + "learning_rate": 9.49103372739073e-06, + "loss": 0.3258, + "step": 7747 + }, + { + "epoch": 0.1705480959718248, + "grad_norm": 1.7668975703097454, + "learning_rate": 9.490877027288894e-06, + "loss": 0.3777, + "step": 7748 + }, + { + "epoch": 0.17057010785824345, + "grad_norm": 1.3855756383535227, + "learning_rate": 9.490720304362416e-06, + "loss": 0.3484, + "step": 7749 + }, + { + "epoch": 0.17059211974466212, + "grad_norm": 1.9629091922566422, + "learning_rate": 9.490563558612095e-06, + "loss": 0.3845, + "step": 7750 + }, + { + "epoch": 0.17061413163108077, + "grad_norm": 1.6413588851043248, + "learning_rate": 9.490406790038729e-06, + "loss": 0.3896, + "step": 7751 + }, + { + "epoch": 0.17063614351749945, + "grad_norm": 1.9758194269003524, + "learning_rate": 9.49024999864311e-06, + "loss": 0.3384, + "step": 7752 + }, + { + "epoch": 0.17065815540391813, + "grad_norm": 1.9913709387560783, + "learning_rate": 9.490093184426039e-06, + "loss": 0.4287, + "step": 7753 + }, + { + "epoch": 0.17068016729033678, + "grad_norm": 1.5921175311776545, + "learning_rate": 9.489936347388312e-06, + "loss": 0.2561, + "step": 7754 + }, + { + "epoch": 0.17070217917675545, + "grad_norm": 1.6923384726968695, + "learning_rate": 9.489779487530725e-06, + "loss": 0.3327, + "step": 7755 + }, + { + "epoch": 0.1707241910631741, + "grad_norm": 1.5957860928776093, + "learning_rate": 9.489622604854078e-06, + "loss": 0.3755, + "step": 7756 + }, + { + "epoch": 0.17074620294959278, + "grad_norm": 0.7244320982544035, + "learning_rate": 9.489465699359165e-06, + "loss": 0.5523, + "step": 7757 + }, + { + "epoch": 0.17076821483601146, + "grad_norm": 1.3948563649727348, + "learning_rate": 9.489308771046786e-06, + "loss": 0.3139, + "step": 7758 + }, + { + "epoch": 0.1707902267224301, + "grad_norm": 1.5568143996701098, + "learning_rate": 9.489151819917737e-06, + "loss": 0.4382, + "step": 7759 + }, + { + "epoch": 0.17081223860884878, + "grad_norm": 1.5709447245694714, + "learning_rate": 9.488994845972816e-06, + "loss": 0.4123, + "step": 7760 + }, + { + "epoch": 0.17083425049526743, + "grad_norm": 0.6285227518527345, + "learning_rate": 9.488837849212821e-06, + "loss": 0.5229, + "step": 7761 + }, + { + "epoch": 0.1708562623816861, + "grad_norm": 1.430811870530717, + "learning_rate": 9.48868082963855e-06, + "loss": 0.3644, + "step": 7762 + }, + { + "epoch": 0.17087827426810479, + "grad_norm": 1.3188139030088377, + "learning_rate": 9.488523787250802e-06, + "loss": 0.3322, + "step": 7763 + }, + { + "epoch": 0.17090028615452343, + "grad_norm": 3.3194991398205778, + "learning_rate": 9.488366722050372e-06, + "loss": 0.3826, + "step": 7764 + }, + { + "epoch": 0.1709222980409421, + "grad_norm": 2.150131525303651, + "learning_rate": 9.488209634038064e-06, + "loss": 0.4137, + "step": 7765 + }, + { + "epoch": 0.1709443099273608, + "grad_norm": 1.7544082693365624, + "learning_rate": 9.488052523214671e-06, + "loss": 0.3607, + "step": 7766 + }, + { + "epoch": 0.17096632181377944, + "grad_norm": 0.6900925763205066, + "learning_rate": 9.487895389580992e-06, + "loss": 0.5247, + "step": 7767 + }, + { + "epoch": 0.17098833370019811, + "grad_norm": 2.6442603346402778, + "learning_rate": 9.487738233137829e-06, + "loss": 0.3393, + "step": 7768 + }, + { + "epoch": 0.17101034558661676, + "grad_norm": 4.250198640962888, + "learning_rate": 9.487581053885976e-06, + "loss": 0.3535, + "step": 7769 + }, + { + "epoch": 0.17103235747303544, + "grad_norm": 1.6031486712389622, + "learning_rate": 9.487423851826237e-06, + "loss": 0.3542, + "step": 7770 + }, + { + "epoch": 0.17105436935945412, + "grad_norm": 2.9105570184704073, + "learning_rate": 9.487266626959408e-06, + "loss": 0.4508, + "step": 7771 + }, + { + "epoch": 0.17107638124587277, + "grad_norm": 1.4973019416647186, + "learning_rate": 9.487109379286287e-06, + "loss": 0.2914, + "step": 7772 + }, + { + "epoch": 0.17109839313229144, + "grad_norm": 0.647582952833382, + "learning_rate": 9.486952108807678e-06, + "loss": 0.5631, + "step": 7773 + }, + { + "epoch": 0.1711204050187101, + "grad_norm": 1.669587710922027, + "learning_rate": 9.486794815524376e-06, + "loss": 0.3362, + "step": 7774 + }, + { + "epoch": 0.17114241690512877, + "grad_norm": 0.6551663051954291, + "learning_rate": 9.48663749943718e-06, + "loss": 0.5545, + "step": 7775 + }, + { + "epoch": 0.17116442879154745, + "grad_norm": 1.8548929873805038, + "learning_rate": 9.486480160546893e-06, + "loss": 0.4475, + "step": 7776 + }, + { + "epoch": 0.1711864406779661, + "grad_norm": 3.1436523891815815, + "learning_rate": 9.486322798854312e-06, + "loss": 0.3364, + "step": 7777 + }, + { + "epoch": 0.17120845256438477, + "grad_norm": 1.9282573361728705, + "learning_rate": 9.486165414360237e-06, + "loss": 0.3684, + "step": 7778 + }, + { + "epoch": 0.17123046445080342, + "grad_norm": 0.6333687436583096, + "learning_rate": 9.48600800706547e-06, + "loss": 0.5407, + "step": 7779 + }, + { + "epoch": 0.1712524763372221, + "grad_norm": 2.670911277169102, + "learning_rate": 9.485850576970807e-06, + "loss": 0.3624, + "step": 7780 + }, + { + "epoch": 0.17127448822364078, + "grad_norm": 8.588877332443452, + "learning_rate": 9.485693124077053e-06, + "loss": 0.3166, + "step": 7781 + }, + { + "epoch": 0.17129650011005942, + "grad_norm": 1.756784215193337, + "learning_rate": 9.485535648385005e-06, + "loss": 0.3722, + "step": 7782 + }, + { + "epoch": 0.1713185119964781, + "grad_norm": 1.5095913623870068, + "learning_rate": 9.485378149895464e-06, + "loss": 0.4077, + "step": 7783 + }, + { + "epoch": 0.17134052388289678, + "grad_norm": 1.4497734784922158, + "learning_rate": 9.48522062860923e-06, + "loss": 0.335, + "step": 7784 + }, + { + "epoch": 0.17136253576931543, + "grad_norm": 1.3086119058965306, + "learning_rate": 9.485063084527106e-06, + "loss": 0.4475, + "step": 7785 + }, + { + "epoch": 0.1713845476557341, + "grad_norm": 1.7335824926869927, + "learning_rate": 9.484905517649891e-06, + "loss": 0.3249, + "step": 7786 + }, + { + "epoch": 0.17140655954215275, + "grad_norm": 1.147897652784291, + "learning_rate": 9.484747927978383e-06, + "loss": 0.2917, + "step": 7787 + }, + { + "epoch": 0.17142857142857143, + "grad_norm": 2.032797172244804, + "learning_rate": 9.48459031551339e-06, + "loss": 0.4074, + "step": 7788 + }, + { + "epoch": 0.1714505833149901, + "grad_norm": 1.5104416907039833, + "learning_rate": 9.484432680255707e-06, + "loss": 0.2806, + "step": 7789 + }, + { + "epoch": 0.17147259520140876, + "grad_norm": 1.898479838616492, + "learning_rate": 9.484275022206136e-06, + "loss": 0.391, + "step": 7790 + }, + { + "epoch": 0.17149460708782743, + "grad_norm": 1.7805014525889793, + "learning_rate": 9.48411734136548e-06, + "loss": 0.3189, + "step": 7791 + }, + { + "epoch": 0.17151661897424608, + "grad_norm": 1.785980789951198, + "learning_rate": 9.483959637734539e-06, + "loss": 0.3936, + "step": 7792 + }, + { + "epoch": 0.17153863086066476, + "grad_norm": 1.414323226590481, + "learning_rate": 9.483801911314117e-06, + "loss": 0.3477, + "step": 7793 + }, + { + "epoch": 0.17156064274708344, + "grad_norm": 1.4485876160374795, + "learning_rate": 9.483644162105012e-06, + "loss": 0.3738, + "step": 7794 + }, + { + "epoch": 0.17158265463350209, + "grad_norm": 1.566268466781159, + "learning_rate": 9.483486390108029e-06, + "loss": 0.2808, + "step": 7795 + }, + { + "epoch": 0.17160466651992076, + "grad_norm": 1.353971593732898, + "learning_rate": 9.483328595323969e-06, + "loss": 0.2761, + "step": 7796 + }, + { + "epoch": 0.1716266784063394, + "grad_norm": 1.5821248410697553, + "learning_rate": 9.48317077775363e-06, + "loss": 0.352, + "step": 7797 + }, + { + "epoch": 0.1716486902927581, + "grad_norm": 0.7671805707544795, + "learning_rate": 9.48301293739782e-06, + "loss": 0.5504, + "step": 7798 + }, + { + "epoch": 0.17167070217917677, + "grad_norm": 0.6910743476839343, + "learning_rate": 9.48285507425734e-06, + "loss": 0.5371, + "step": 7799 + }, + { + "epoch": 0.17169271406559541, + "grad_norm": 1.4361462266052207, + "learning_rate": 9.482697188332988e-06, + "loss": 0.321, + "step": 7800 + }, + { + "epoch": 0.1717147259520141, + "grad_norm": 1.922208468769767, + "learning_rate": 9.482539279625572e-06, + "loss": 0.3298, + "step": 7801 + }, + { + "epoch": 0.17173673783843274, + "grad_norm": 4.380998510282887, + "learning_rate": 9.482381348135892e-06, + "loss": 0.3249, + "step": 7802 + }, + { + "epoch": 0.17175874972485142, + "grad_norm": 1.257428598341016, + "learning_rate": 9.48222339386475e-06, + "loss": 0.3454, + "step": 7803 + }, + { + "epoch": 0.1717807616112701, + "grad_norm": 1.6772912029138023, + "learning_rate": 9.48206541681295e-06, + "loss": 0.3801, + "step": 7804 + }, + { + "epoch": 0.17180277349768874, + "grad_norm": 1.6540894284839922, + "learning_rate": 9.481907416981295e-06, + "loss": 0.376, + "step": 7805 + }, + { + "epoch": 0.17182478538410742, + "grad_norm": 1.7071302103794186, + "learning_rate": 9.481749394370585e-06, + "loss": 0.4314, + "step": 7806 + }, + { + "epoch": 0.1718467972705261, + "grad_norm": 5.855045341782327, + "learning_rate": 9.481591348981628e-06, + "loss": 0.3417, + "step": 7807 + }, + { + "epoch": 0.17186880915694475, + "grad_norm": 1.5126725534031993, + "learning_rate": 9.481433280815224e-06, + "loss": 0.4275, + "step": 7808 + }, + { + "epoch": 0.17189082104336342, + "grad_norm": 0.7993019550370777, + "learning_rate": 9.48127518987218e-06, + "loss": 0.5237, + "step": 7809 + }, + { + "epoch": 0.17191283292978207, + "grad_norm": 1.7676652848763765, + "learning_rate": 9.481117076153293e-06, + "loss": 0.3556, + "step": 7810 + }, + { + "epoch": 0.17193484481620075, + "grad_norm": 0.7132100640539396, + "learning_rate": 9.480958939659371e-06, + "loss": 0.529, + "step": 7811 + }, + { + "epoch": 0.17195685670261943, + "grad_norm": 1.5373373546030231, + "learning_rate": 9.48080078039122e-06, + "loss": 0.3264, + "step": 7812 + }, + { + "epoch": 0.17197886858903808, + "grad_norm": 1.7754718876330746, + "learning_rate": 9.480642598349637e-06, + "loss": 0.3397, + "step": 7813 + }, + { + "epoch": 0.17200088047545675, + "grad_norm": 1.9877647128362514, + "learning_rate": 9.480484393535432e-06, + "loss": 0.3657, + "step": 7814 + }, + { + "epoch": 0.1720228923618754, + "grad_norm": 1.7797101828856188, + "learning_rate": 9.480326165949407e-06, + "loss": 0.3714, + "step": 7815 + }, + { + "epoch": 0.17204490424829408, + "grad_norm": 0.6951100222720731, + "learning_rate": 9.480167915592366e-06, + "loss": 0.516, + "step": 7816 + }, + { + "epoch": 0.17206691613471276, + "grad_norm": 1.4043651337162033, + "learning_rate": 9.480009642465114e-06, + "loss": 0.3723, + "step": 7817 + }, + { + "epoch": 0.1720889280211314, + "grad_norm": 1.2906407814449732, + "learning_rate": 9.479851346568453e-06, + "loss": 0.3284, + "step": 7818 + }, + { + "epoch": 0.17211093990755008, + "grad_norm": 1.1212842246500807, + "learning_rate": 9.479693027903192e-06, + "loss": 0.2956, + "step": 7819 + }, + { + "epoch": 0.17213295179396873, + "grad_norm": 1.4958917387624127, + "learning_rate": 9.479534686470132e-06, + "loss": 0.2882, + "step": 7820 + }, + { + "epoch": 0.1721549636803874, + "grad_norm": 1.394640149057457, + "learning_rate": 9.479376322270077e-06, + "loss": 0.3064, + "step": 7821 + }, + { + "epoch": 0.17217697556680608, + "grad_norm": 1.8484697425235592, + "learning_rate": 9.479217935303835e-06, + "loss": 0.3009, + "step": 7822 + }, + { + "epoch": 0.17219898745322473, + "grad_norm": 0.7046401658661129, + "learning_rate": 9.479059525572209e-06, + "loss": 0.5411, + "step": 7823 + }, + { + "epoch": 0.1722209993396434, + "grad_norm": 1.6201027338198024, + "learning_rate": 9.478901093076006e-06, + "loss": 0.358, + "step": 7824 + }, + { + "epoch": 0.1722430112260621, + "grad_norm": 1.702469388174621, + "learning_rate": 9.478742637816029e-06, + "loss": 0.3914, + "step": 7825 + }, + { + "epoch": 0.17226502311248074, + "grad_norm": 1.4088983554440269, + "learning_rate": 9.478584159793085e-06, + "loss": 0.3465, + "step": 7826 + }, + { + "epoch": 0.1722870349988994, + "grad_norm": 1.5708781791703974, + "learning_rate": 9.47842565900798e-06, + "loss": 0.3208, + "step": 7827 + }, + { + "epoch": 0.17230904688531806, + "grad_norm": 1.5542165489241049, + "learning_rate": 9.478267135461515e-06, + "loss": 0.3628, + "step": 7828 + }, + { + "epoch": 0.17233105877173674, + "grad_norm": 1.6121082508174678, + "learning_rate": 9.478108589154502e-06, + "loss": 0.3753, + "step": 7829 + }, + { + "epoch": 0.17235307065815542, + "grad_norm": 1.4450183079258074, + "learning_rate": 9.477950020087743e-06, + "loss": 0.3127, + "step": 7830 + }, + { + "epoch": 0.17237508254457407, + "grad_norm": 1.4269825749279252, + "learning_rate": 9.477791428262045e-06, + "loss": 0.3914, + "step": 7831 + }, + { + "epoch": 0.17239709443099274, + "grad_norm": 1.5514913400774937, + "learning_rate": 9.477632813678214e-06, + "loss": 0.3109, + "step": 7832 + }, + { + "epoch": 0.1724191063174114, + "grad_norm": 1.4732275782680715, + "learning_rate": 9.477474176337056e-06, + "loss": 0.2883, + "step": 7833 + }, + { + "epoch": 0.17244111820383007, + "grad_norm": 0.7509992987996985, + "learning_rate": 9.477315516239376e-06, + "loss": 0.5343, + "step": 7834 + }, + { + "epoch": 0.17246313009024875, + "grad_norm": 1.385970568467349, + "learning_rate": 9.477156833385982e-06, + "loss": 0.3641, + "step": 7835 + }, + { + "epoch": 0.1724851419766674, + "grad_norm": 1.5385152068878178, + "learning_rate": 9.476998127777682e-06, + "loss": 0.3009, + "step": 7836 + }, + { + "epoch": 0.17250715386308607, + "grad_norm": 1.7868860518374245, + "learning_rate": 9.476839399415278e-06, + "loss": 0.449, + "step": 7837 + }, + { + "epoch": 0.17252916574950472, + "grad_norm": 1.33373237246193, + "learning_rate": 9.47668064829958e-06, + "loss": 0.3952, + "step": 7838 + }, + { + "epoch": 0.1725511776359234, + "grad_norm": 1.914427256588901, + "learning_rate": 9.476521874431394e-06, + "loss": 0.3843, + "step": 7839 + }, + { + "epoch": 0.17257318952234207, + "grad_norm": 1.1669116039720644, + "learning_rate": 9.47636307781153e-06, + "loss": 0.3567, + "step": 7840 + }, + { + "epoch": 0.17259520140876072, + "grad_norm": 0.7138317602707633, + "learning_rate": 9.47620425844079e-06, + "loss": 0.5537, + "step": 7841 + }, + { + "epoch": 0.1726172132951794, + "grad_norm": 1.8086475270547704, + "learning_rate": 9.476045416319984e-06, + "loss": 0.4417, + "step": 7842 + }, + { + "epoch": 0.17263922518159805, + "grad_norm": 1.678484301341565, + "learning_rate": 9.475886551449918e-06, + "loss": 0.4405, + "step": 7843 + }, + { + "epoch": 0.17266123706801673, + "grad_norm": 1.6304693768568457, + "learning_rate": 9.475727663831402e-06, + "loss": 0.3156, + "step": 7844 + }, + { + "epoch": 0.1726832489544354, + "grad_norm": 2.1373827858053587, + "learning_rate": 9.475568753465241e-06, + "loss": 0.3674, + "step": 7845 + }, + { + "epoch": 0.17270526084085405, + "grad_norm": 1.8852011657747474, + "learning_rate": 9.475409820352243e-06, + "loss": 0.3123, + "step": 7846 + }, + { + "epoch": 0.17272727272727273, + "grad_norm": 2.4532927347702818, + "learning_rate": 9.475250864493216e-06, + "loss": 0.3381, + "step": 7847 + }, + { + "epoch": 0.1727492846136914, + "grad_norm": 0.6421240311866111, + "learning_rate": 9.475091885888968e-06, + "loss": 0.5597, + "step": 7848 + }, + { + "epoch": 0.17277129650011006, + "grad_norm": 4.047384192294004, + "learning_rate": 9.474932884540309e-06, + "loss": 0.4036, + "step": 7849 + }, + { + "epoch": 0.17279330838652873, + "grad_norm": 1.5491849969744576, + "learning_rate": 9.474773860448044e-06, + "loss": 0.3238, + "step": 7850 + }, + { + "epoch": 0.17281532027294738, + "grad_norm": 1.3448169802000458, + "learning_rate": 9.474614813612982e-06, + "loss": 0.3184, + "step": 7851 + }, + { + "epoch": 0.17283733215936606, + "grad_norm": 0.6580802939723869, + "learning_rate": 9.47445574403593e-06, + "loss": 0.5378, + "step": 7852 + }, + { + "epoch": 0.17285934404578474, + "grad_norm": 1.565627408252098, + "learning_rate": 9.474296651717701e-06, + "loss": 0.4561, + "step": 7853 + }, + { + "epoch": 0.17288135593220338, + "grad_norm": 1.7552135477382995, + "learning_rate": 9.4741375366591e-06, + "loss": 0.3918, + "step": 7854 + }, + { + "epoch": 0.17290336781862206, + "grad_norm": 1.691085888518486, + "learning_rate": 9.473978398860936e-06, + "loss": 0.2592, + "step": 7855 + }, + { + "epoch": 0.1729253797050407, + "grad_norm": 1.333622201623066, + "learning_rate": 9.47381923832402e-06, + "loss": 0.3674, + "step": 7856 + }, + { + "epoch": 0.1729473915914594, + "grad_norm": 1.6272737589525805, + "learning_rate": 9.473660055049158e-06, + "loss": 0.3573, + "step": 7857 + }, + { + "epoch": 0.17296940347787806, + "grad_norm": 1.218972370886337, + "learning_rate": 9.473500849037161e-06, + "loss": 0.2517, + "step": 7858 + }, + { + "epoch": 0.1729914153642967, + "grad_norm": 1.3845637464948486, + "learning_rate": 9.473341620288836e-06, + "loss": 0.4275, + "step": 7859 + }, + { + "epoch": 0.1730134272507154, + "grad_norm": 2.0375499409414473, + "learning_rate": 9.473182368804996e-06, + "loss": 0.3783, + "step": 7860 + }, + { + "epoch": 0.17303543913713404, + "grad_norm": 1.4118071378198003, + "learning_rate": 9.473023094586447e-06, + "loss": 0.3401, + "step": 7861 + }, + { + "epoch": 0.17305745102355272, + "grad_norm": 1.4215363545073965, + "learning_rate": 9.472863797633998e-06, + "loss": 0.3687, + "step": 7862 + }, + { + "epoch": 0.1730794629099714, + "grad_norm": 1.8294652709598909, + "learning_rate": 9.472704477948462e-06, + "loss": 0.429, + "step": 7863 + }, + { + "epoch": 0.17310147479639004, + "grad_norm": 1.4459981263061823, + "learning_rate": 9.472545135530646e-06, + "loss": 0.4397, + "step": 7864 + }, + { + "epoch": 0.17312348668280872, + "grad_norm": 0.7099383951392592, + "learning_rate": 9.472385770381363e-06, + "loss": 0.5389, + "step": 7865 + }, + { + "epoch": 0.1731454985692274, + "grad_norm": 1.2648373803297819, + "learning_rate": 9.472226382501419e-06, + "loss": 0.4375, + "step": 7866 + }, + { + "epoch": 0.17316751045564605, + "grad_norm": 0.6496874353040605, + "learning_rate": 9.472066971891625e-06, + "loss": 0.5294, + "step": 7867 + }, + { + "epoch": 0.17318952234206472, + "grad_norm": 1.6690479037132782, + "learning_rate": 9.471907538552793e-06, + "loss": 0.3339, + "step": 7868 + }, + { + "epoch": 0.17321153422848337, + "grad_norm": 1.2149183306184115, + "learning_rate": 9.471748082485733e-06, + "loss": 0.3719, + "step": 7869 + }, + { + "epoch": 0.17323354611490205, + "grad_norm": 0.6368996848439141, + "learning_rate": 9.471588603691254e-06, + "loss": 0.52, + "step": 7870 + }, + { + "epoch": 0.17325555800132073, + "grad_norm": 2.0149449753618414, + "learning_rate": 9.47142910217017e-06, + "loss": 0.3822, + "step": 7871 + }, + { + "epoch": 0.17327756988773937, + "grad_norm": 0.6321971815303887, + "learning_rate": 9.471269577923284e-06, + "loss": 0.5316, + "step": 7872 + }, + { + "epoch": 0.17329958177415805, + "grad_norm": 1.5683480520101858, + "learning_rate": 9.471110030951415e-06, + "loss": 0.3698, + "step": 7873 + }, + { + "epoch": 0.1733215936605767, + "grad_norm": 0.6479073639322013, + "learning_rate": 9.47095046125537e-06, + "loss": 0.5634, + "step": 7874 + }, + { + "epoch": 0.17334360554699538, + "grad_norm": 0.6482773556910891, + "learning_rate": 9.47079086883596e-06, + "loss": 0.5563, + "step": 7875 + }, + { + "epoch": 0.17336561743341405, + "grad_norm": 1.9429521198985809, + "learning_rate": 9.470631253693998e-06, + "loss": 0.4223, + "step": 7876 + }, + { + "epoch": 0.1733876293198327, + "grad_norm": 2.786780535987926, + "learning_rate": 9.470471615830292e-06, + "loss": 0.4055, + "step": 7877 + }, + { + "epoch": 0.17340964120625138, + "grad_norm": 1.4787282499827512, + "learning_rate": 9.470311955245658e-06, + "loss": 0.2898, + "step": 7878 + }, + { + "epoch": 0.17343165309267003, + "grad_norm": 1.5215477198716778, + "learning_rate": 9.470152271940903e-06, + "loss": 0.4233, + "step": 7879 + }, + { + "epoch": 0.1734536649790887, + "grad_norm": 1.5559984284295423, + "learning_rate": 9.469992565916841e-06, + "loss": 0.2992, + "step": 7880 + }, + { + "epoch": 0.17347567686550738, + "grad_norm": 1.560965366405977, + "learning_rate": 9.469832837174285e-06, + "loss": 0.2904, + "step": 7881 + }, + { + "epoch": 0.17349768875192603, + "grad_norm": 1.8904607723700786, + "learning_rate": 9.469673085714042e-06, + "loss": 0.2918, + "step": 7882 + }, + { + "epoch": 0.1735197006383447, + "grad_norm": 0.8307591943852985, + "learning_rate": 9.469513311536928e-06, + "loss": 0.5327, + "step": 7883 + }, + { + "epoch": 0.17354171252476336, + "grad_norm": 1.7862296657092536, + "learning_rate": 9.469353514643754e-06, + "loss": 0.2945, + "step": 7884 + }, + { + "epoch": 0.17356372441118204, + "grad_norm": 1.3071546883245342, + "learning_rate": 9.469193695035331e-06, + "loss": 0.3558, + "step": 7885 + }, + { + "epoch": 0.1735857362976007, + "grad_norm": 0.6600469458708464, + "learning_rate": 9.469033852712473e-06, + "loss": 0.5444, + "step": 7886 + }, + { + "epoch": 0.17360774818401936, + "grad_norm": 1.5305164478721127, + "learning_rate": 9.468873987675993e-06, + "loss": 0.3744, + "step": 7887 + }, + { + "epoch": 0.17362976007043804, + "grad_norm": 1.3712907699898225, + "learning_rate": 9.4687140999267e-06, + "loss": 0.3458, + "step": 7888 + }, + { + "epoch": 0.17365177195685672, + "grad_norm": 1.6160192706191077, + "learning_rate": 9.46855418946541e-06, + "loss": 0.3596, + "step": 7889 + }, + { + "epoch": 0.17367378384327536, + "grad_norm": 1.2917318311803043, + "learning_rate": 9.468394256292935e-06, + "loss": 0.3897, + "step": 7890 + }, + { + "epoch": 0.17369579572969404, + "grad_norm": 1.467941151991228, + "learning_rate": 9.468234300410086e-06, + "loss": 0.2879, + "step": 7891 + }, + { + "epoch": 0.1737178076161127, + "grad_norm": 1.603690158205172, + "learning_rate": 9.46807432181768e-06, + "loss": 0.43, + "step": 7892 + }, + { + "epoch": 0.17373981950253137, + "grad_norm": 0.8212350260209075, + "learning_rate": 9.467914320516525e-06, + "loss": 0.5337, + "step": 7893 + }, + { + "epoch": 0.17376183138895004, + "grad_norm": 1.7710934293053646, + "learning_rate": 9.467754296507438e-06, + "loss": 0.433, + "step": 7894 + }, + { + "epoch": 0.1737838432753687, + "grad_norm": 1.7610864839275844, + "learning_rate": 9.46759424979123e-06, + "loss": 0.3823, + "step": 7895 + }, + { + "epoch": 0.17380585516178737, + "grad_norm": 1.2255233274582162, + "learning_rate": 9.467434180368715e-06, + "loss": 0.3308, + "step": 7896 + }, + { + "epoch": 0.17382786704820602, + "grad_norm": 1.4697950721728408, + "learning_rate": 9.467274088240708e-06, + "loss": 0.2667, + "step": 7897 + }, + { + "epoch": 0.1738498789346247, + "grad_norm": 1.8722234511391371, + "learning_rate": 9.467113973408022e-06, + "loss": 0.3111, + "step": 7898 + }, + { + "epoch": 0.17387189082104337, + "grad_norm": 1.3314555080777533, + "learning_rate": 9.466953835871471e-06, + "loss": 0.29, + "step": 7899 + }, + { + "epoch": 0.17389390270746202, + "grad_norm": 1.54350613952111, + "learning_rate": 9.466793675631866e-06, + "loss": 0.3458, + "step": 7900 + }, + { + "epoch": 0.1739159145938807, + "grad_norm": 1.341625567175772, + "learning_rate": 9.466633492690024e-06, + "loss": 0.3026, + "step": 7901 + }, + { + "epoch": 0.17393792648029935, + "grad_norm": 1.3045838242759977, + "learning_rate": 9.46647328704676e-06, + "loss": 0.2924, + "step": 7902 + }, + { + "epoch": 0.17395993836671803, + "grad_norm": 1.4660654033097313, + "learning_rate": 9.466313058702885e-06, + "loss": 0.4166, + "step": 7903 + }, + { + "epoch": 0.1739819502531367, + "grad_norm": 1.2651689797774515, + "learning_rate": 9.466152807659215e-06, + "loss": 0.2971, + "step": 7904 + }, + { + "epoch": 0.17400396213955535, + "grad_norm": 0.6929920538122488, + "learning_rate": 9.465992533916566e-06, + "loss": 0.5403, + "step": 7905 + }, + { + "epoch": 0.17402597402597403, + "grad_norm": 1.1581597655950504, + "learning_rate": 9.465832237475749e-06, + "loss": 0.2818, + "step": 7906 + }, + { + "epoch": 0.1740479859123927, + "grad_norm": 1.6711137279256405, + "learning_rate": 9.465671918337582e-06, + "loss": 0.4768, + "step": 7907 + }, + { + "epoch": 0.17406999779881135, + "grad_norm": 1.7606505156849743, + "learning_rate": 9.46551157650288e-06, + "loss": 0.3457, + "step": 7908 + }, + { + "epoch": 0.17409200968523003, + "grad_norm": 1.5465515130062004, + "learning_rate": 9.465351211972453e-06, + "loss": 0.382, + "step": 7909 + }, + { + "epoch": 0.17411402157164868, + "grad_norm": 1.1798243923332632, + "learning_rate": 9.465190824747122e-06, + "loss": 0.3387, + "step": 7910 + }, + { + "epoch": 0.17413603345806736, + "grad_norm": 1.3221841503644773, + "learning_rate": 9.4650304148277e-06, + "loss": 0.3133, + "step": 7911 + }, + { + "epoch": 0.17415804534448603, + "grad_norm": 1.8139915574536045, + "learning_rate": 9.464869982215001e-06, + "loss": 0.4148, + "step": 7912 + }, + { + "epoch": 0.17418005723090468, + "grad_norm": 0.6973723408456333, + "learning_rate": 9.46470952690984e-06, + "loss": 0.5486, + "step": 7913 + }, + { + "epoch": 0.17420206911732336, + "grad_norm": 1.72903071796262, + "learning_rate": 9.464549048913037e-06, + "loss": 0.3793, + "step": 7914 + }, + { + "epoch": 0.174224081003742, + "grad_norm": 1.3970268465128495, + "learning_rate": 9.464388548225403e-06, + "loss": 0.3639, + "step": 7915 + }, + { + "epoch": 0.1742460928901607, + "grad_norm": 1.4745213793873928, + "learning_rate": 9.464228024847755e-06, + "loss": 0.3689, + "step": 7916 + }, + { + "epoch": 0.17426810477657936, + "grad_norm": 4.221045525726565, + "learning_rate": 9.464067478780909e-06, + "loss": 0.3456, + "step": 7917 + }, + { + "epoch": 0.174290116662998, + "grad_norm": 1.384545422948151, + "learning_rate": 9.46390691002568e-06, + "loss": 0.3465, + "step": 7918 + }, + { + "epoch": 0.1743121285494167, + "grad_norm": 1.8022411826960332, + "learning_rate": 9.463746318582888e-06, + "loss": 0.3146, + "step": 7919 + }, + { + "epoch": 0.17433414043583534, + "grad_norm": 2.9217648624331622, + "learning_rate": 9.463585704453342e-06, + "loss": 0.3256, + "step": 7920 + }, + { + "epoch": 0.17435615232225402, + "grad_norm": 1.4679852799492796, + "learning_rate": 9.463425067637866e-06, + "loss": 0.4118, + "step": 7921 + }, + { + "epoch": 0.1743781642086727, + "grad_norm": 0.6578527131048003, + "learning_rate": 9.463264408137272e-06, + "loss": 0.5624, + "step": 7922 + }, + { + "epoch": 0.17440017609509134, + "grad_norm": 1.7696112205371106, + "learning_rate": 9.463103725952378e-06, + "loss": 0.3282, + "step": 7923 + }, + { + "epoch": 0.17442218798151002, + "grad_norm": 2.277522192926198, + "learning_rate": 9.462943021084002e-06, + "loss": 0.3427, + "step": 7924 + }, + { + "epoch": 0.17444419986792867, + "grad_norm": 1.5468857912688176, + "learning_rate": 9.462782293532957e-06, + "loss": 0.3383, + "step": 7925 + }, + { + "epoch": 0.17446621175434734, + "grad_norm": 1.3125562856915733, + "learning_rate": 9.462621543300062e-06, + "loss": 0.3646, + "step": 7926 + }, + { + "epoch": 0.17448822364076602, + "grad_norm": 1.4802437771219923, + "learning_rate": 9.462460770386132e-06, + "loss": 0.3322, + "step": 7927 + }, + { + "epoch": 0.17451023552718467, + "grad_norm": 1.6071760960003654, + "learning_rate": 9.46229997479199e-06, + "loss": 0.2913, + "step": 7928 + }, + { + "epoch": 0.17453224741360335, + "grad_norm": 1.3659905734341156, + "learning_rate": 9.462139156518447e-06, + "loss": 0.2858, + "step": 7929 + }, + { + "epoch": 0.17455425930002202, + "grad_norm": 1.3972755745519285, + "learning_rate": 9.461978315566322e-06, + "loss": 0.3459, + "step": 7930 + }, + { + "epoch": 0.17457627118644067, + "grad_norm": 1.4971550216409597, + "learning_rate": 9.461817451936434e-06, + "loss": 0.3049, + "step": 7931 + }, + { + "epoch": 0.17459828307285935, + "grad_norm": 2.140210236687067, + "learning_rate": 9.4616565656296e-06, + "loss": 0.3051, + "step": 7932 + }, + { + "epoch": 0.174620294959278, + "grad_norm": 1.9892298834473543, + "learning_rate": 9.461495656646638e-06, + "loss": 0.3067, + "step": 7933 + }, + { + "epoch": 0.17464230684569668, + "grad_norm": 0.6926950806418324, + "learning_rate": 9.461334724988364e-06, + "loss": 0.5416, + "step": 7934 + }, + { + "epoch": 0.17466431873211535, + "grad_norm": 1.9257619319401837, + "learning_rate": 9.461173770655598e-06, + "loss": 0.3598, + "step": 7935 + }, + { + "epoch": 0.174686330618534, + "grad_norm": 1.7390096677651266, + "learning_rate": 9.461012793649157e-06, + "loss": 0.3656, + "step": 7936 + }, + { + "epoch": 0.17470834250495268, + "grad_norm": 1.3108838963019365, + "learning_rate": 9.460851793969859e-06, + "loss": 0.378, + "step": 7937 + }, + { + "epoch": 0.17473035439137133, + "grad_norm": 1.6475600278125013, + "learning_rate": 9.460690771618523e-06, + "loss": 0.3357, + "step": 7938 + }, + { + "epoch": 0.17475236627779, + "grad_norm": 1.2622166280827014, + "learning_rate": 9.460529726595966e-06, + "loss": 0.3967, + "step": 7939 + }, + { + "epoch": 0.17477437816420868, + "grad_norm": 1.5859948468551706, + "learning_rate": 9.460368658903007e-06, + "loss": 0.3238, + "step": 7940 + }, + { + "epoch": 0.17479639005062733, + "grad_norm": 2.694487832249705, + "learning_rate": 9.460207568540466e-06, + "loss": 0.3939, + "step": 7941 + }, + { + "epoch": 0.174818401937046, + "grad_norm": 1.5006972283121833, + "learning_rate": 9.460046455509162e-06, + "loss": 0.2908, + "step": 7942 + }, + { + "epoch": 0.17484041382346466, + "grad_norm": 1.6905573201918793, + "learning_rate": 9.459885319809911e-06, + "loss": 0.3267, + "step": 7943 + }, + { + "epoch": 0.17486242570988333, + "grad_norm": 1.3562424017505212, + "learning_rate": 9.459724161443535e-06, + "loss": 0.3252, + "step": 7944 + }, + { + "epoch": 0.174884437596302, + "grad_norm": 1.5493320849204135, + "learning_rate": 9.45956298041085e-06, + "loss": 0.2818, + "step": 7945 + }, + { + "epoch": 0.17490644948272066, + "grad_norm": 0.693034251950916, + "learning_rate": 9.45940177671268e-06, + "loss": 0.5341, + "step": 7946 + }, + { + "epoch": 0.17492846136913934, + "grad_norm": 0.622519514865005, + "learning_rate": 9.459240550349837e-06, + "loss": 0.5009, + "step": 7947 + }, + { + "epoch": 0.17495047325555801, + "grad_norm": 1.6922154375445466, + "learning_rate": 9.459079301323148e-06, + "loss": 0.4914, + "step": 7948 + }, + { + "epoch": 0.17497248514197666, + "grad_norm": 1.8920228469810185, + "learning_rate": 9.458918029633428e-06, + "loss": 0.4019, + "step": 7949 + }, + { + "epoch": 0.17499449702839534, + "grad_norm": 1.2027749756525954, + "learning_rate": 9.458756735281498e-06, + "loss": 0.3358, + "step": 7950 + }, + { + "epoch": 0.175016508914814, + "grad_norm": 1.374762795414854, + "learning_rate": 9.458595418268177e-06, + "loss": 0.3524, + "step": 7951 + }, + { + "epoch": 0.17503852080123267, + "grad_norm": 1.6866149338203997, + "learning_rate": 9.458434078594285e-06, + "loss": 0.3326, + "step": 7952 + }, + { + "epoch": 0.17506053268765134, + "grad_norm": 2.3068693456099916, + "learning_rate": 9.458272716260644e-06, + "loss": 0.3883, + "step": 7953 + }, + { + "epoch": 0.17508254457407, + "grad_norm": 2.2754399383743578, + "learning_rate": 9.458111331268072e-06, + "loss": 0.3867, + "step": 7954 + }, + { + "epoch": 0.17510455646048867, + "grad_norm": 1.5010396671138941, + "learning_rate": 9.457949923617391e-06, + "loss": 0.3496, + "step": 7955 + }, + { + "epoch": 0.17512656834690732, + "grad_norm": 1.6543947328509434, + "learning_rate": 9.457788493309419e-06, + "loss": 0.3512, + "step": 7956 + }, + { + "epoch": 0.175148580233326, + "grad_norm": 1.4029059100351582, + "learning_rate": 9.457627040344977e-06, + "loss": 0.4826, + "step": 7957 + }, + { + "epoch": 0.17517059211974467, + "grad_norm": 4.065872727459805, + "learning_rate": 9.457465564724887e-06, + "loss": 0.364, + "step": 7958 + }, + { + "epoch": 0.17519260400616332, + "grad_norm": 1.4471983636381172, + "learning_rate": 9.457304066449968e-06, + "loss": 0.4037, + "step": 7959 + }, + { + "epoch": 0.175214615892582, + "grad_norm": 1.5980307436297259, + "learning_rate": 9.457142545521042e-06, + "loss": 0.4304, + "step": 7960 + }, + { + "epoch": 0.17523662777900065, + "grad_norm": 1.4375216889422504, + "learning_rate": 9.456981001938932e-06, + "loss": 0.3142, + "step": 7961 + }, + { + "epoch": 0.17525863966541932, + "grad_norm": 2.4000463159974306, + "learning_rate": 9.456819435704454e-06, + "loss": 0.3086, + "step": 7962 + }, + { + "epoch": 0.175280651551838, + "grad_norm": 1.208872209890037, + "learning_rate": 9.456657846818434e-06, + "loss": 0.274, + "step": 7963 + }, + { + "epoch": 0.17530266343825665, + "grad_norm": 1.4491971644442838, + "learning_rate": 9.45649623528169e-06, + "loss": 0.336, + "step": 7964 + }, + { + "epoch": 0.17532467532467533, + "grad_norm": 1.6724294932490946, + "learning_rate": 9.456334601095043e-06, + "loss": 0.2915, + "step": 7965 + }, + { + "epoch": 0.175346687211094, + "grad_norm": 1.6366090350353677, + "learning_rate": 9.456172944259317e-06, + "loss": 0.3162, + "step": 7966 + }, + { + "epoch": 0.17536869909751265, + "grad_norm": 1.494541321621431, + "learning_rate": 9.456011264775334e-06, + "loss": 0.4421, + "step": 7967 + }, + { + "epoch": 0.17539071098393133, + "grad_norm": 1.4938193634082615, + "learning_rate": 9.455849562643913e-06, + "loss": 0.3565, + "step": 7968 + }, + { + "epoch": 0.17541272287034998, + "grad_norm": 1.4143953562390246, + "learning_rate": 9.455687837865877e-06, + "loss": 0.3121, + "step": 7969 + }, + { + "epoch": 0.17543473475676866, + "grad_norm": 0.9123655383840354, + "learning_rate": 9.455526090442049e-06, + "loss": 0.5621, + "step": 7970 + }, + { + "epoch": 0.17545674664318733, + "grad_norm": 1.8193808429025502, + "learning_rate": 9.455364320373249e-06, + "loss": 0.362, + "step": 7971 + }, + { + "epoch": 0.17547875852960598, + "grad_norm": 1.3196280195945302, + "learning_rate": 9.4552025276603e-06, + "loss": 0.3557, + "step": 7972 + }, + { + "epoch": 0.17550077041602466, + "grad_norm": 1.4462351084062468, + "learning_rate": 9.455040712304027e-06, + "loss": 0.341, + "step": 7973 + }, + { + "epoch": 0.1755227823024433, + "grad_norm": 1.9724971371493112, + "learning_rate": 9.45487887430525e-06, + "loss": 0.4593, + "step": 7974 + }, + { + "epoch": 0.17554479418886199, + "grad_norm": 1.3625623503873459, + "learning_rate": 9.454717013664789e-06, + "loss": 0.3602, + "step": 7975 + }, + { + "epoch": 0.17556680607528066, + "grad_norm": 2.676464754831476, + "learning_rate": 9.454555130383471e-06, + "loss": 0.393, + "step": 7976 + }, + { + "epoch": 0.1755888179616993, + "grad_norm": 1.4376855464185574, + "learning_rate": 9.454393224462116e-06, + "loss": 0.4129, + "step": 7977 + }, + { + "epoch": 0.175610829848118, + "grad_norm": 0.7184402591506381, + "learning_rate": 9.454231295901548e-06, + "loss": 0.5566, + "step": 7978 + }, + { + "epoch": 0.17563284173453664, + "grad_norm": 1.1693105629771092, + "learning_rate": 9.45406934470259e-06, + "loss": 0.3351, + "step": 7979 + }, + { + "epoch": 0.17565485362095531, + "grad_norm": 1.5858209099454519, + "learning_rate": 9.453907370866066e-06, + "loss": 0.3421, + "step": 7980 + }, + { + "epoch": 0.175676865507374, + "grad_norm": 1.9550648388101912, + "learning_rate": 9.453745374392798e-06, + "loss": 0.4045, + "step": 7981 + }, + { + "epoch": 0.17569887739379264, + "grad_norm": 1.4652553812095432, + "learning_rate": 9.45358335528361e-06, + "loss": 0.3396, + "step": 7982 + }, + { + "epoch": 0.17572088928021132, + "grad_norm": 2.8470786641106587, + "learning_rate": 9.453421313539324e-06, + "loss": 0.4471, + "step": 7983 + }, + { + "epoch": 0.17574290116662997, + "grad_norm": 6.453909671726106, + "learning_rate": 9.453259249160765e-06, + "loss": 0.4325, + "step": 7984 + }, + { + "epoch": 0.17576491305304864, + "grad_norm": 3.866559954389981, + "learning_rate": 9.453097162148757e-06, + "loss": 0.3395, + "step": 7985 + }, + { + "epoch": 0.17578692493946732, + "grad_norm": 1.4203986674422149, + "learning_rate": 9.452935052504121e-06, + "loss": 0.4445, + "step": 7986 + }, + { + "epoch": 0.17580893682588597, + "grad_norm": 1.856627141879867, + "learning_rate": 9.452772920227687e-06, + "loss": 0.3546, + "step": 7987 + }, + { + "epoch": 0.17583094871230465, + "grad_norm": 1.9645105357957813, + "learning_rate": 9.45261076532027e-06, + "loss": 0.4223, + "step": 7988 + }, + { + "epoch": 0.17585296059872332, + "grad_norm": 1.8036999149531423, + "learning_rate": 9.452448587782704e-06, + "loss": 0.4297, + "step": 7989 + }, + { + "epoch": 0.17587497248514197, + "grad_norm": 2.076351742293778, + "learning_rate": 9.452286387615805e-06, + "loss": 0.2953, + "step": 7990 + }, + { + "epoch": 0.17589698437156065, + "grad_norm": 1.1687510602528777, + "learning_rate": 9.452124164820402e-06, + "loss": 0.3684, + "step": 7991 + }, + { + "epoch": 0.1759189962579793, + "grad_norm": 1.5897970065808924, + "learning_rate": 9.451961919397319e-06, + "loss": 0.3718, + "step": 7992 + }, + { + "epoch": 0.17594100814439798, + "grad_norm": 1.42359374942858, + "learning_rate": 9.451799651347378e-06, + "loss": 0.3548, + "step": 7993 + }, + { + "epoch": 0.17596302003081665, + "grad_norm": 1.474523156312123, + "learning_rate": 9.451637360671407e-06, + "loss": 0.3491, + "step": 7994 + }, + { + "epoch": 0.1759850319172353, + "grad_norm": 0.6744752672318598, + "learning_rate": 9.451475047370229e-06, + "loss": 0.5718, + "step": 7995 + }, + { + "epoch": 0.17600704380365398, + "grad_norm": 1.9182916190116606, + "learning_rate": 9.451312711444668e-06, + "loss": 0.3315, + "step": 7996 + }, + { + "epoch": 0.17602905569007263, + "grad_norm": 0.6786722161631873, + "learning_rate": 9.451150352895552e-06, + "loss": 0.5686, + "step": 7997 + }, + { + "epoch": 0.1760510675764913, + "grad_norm": 1.3601419562278851, + "learning_rate": 9.450987971723704e-06, + "loss": 0.3385, + "step": 7998 + }, + { + "epoch": 0.17607307946290998, + "grad_norm": 1.358865679113696, + "learning_rate": 9.450825567929953e-06, + "loss": 0.3245, + "step": 7999 + }, + { + "epoch": 0.17609509134932863, + "grad_norm": 1.244083491534059, + "learning_rate": 9.450663141515117e-06, + "loss": 0.3047, + "step": 8000 + }, + { + "epoch": 0.1761171032357473, + "grad_norm": 0.6422734353667257, + "learning_rate": 9.45050069248003e-06, + "loss": 0.5266, + "step": 8001 + }, + { + "epoch": 0.17613911512216596, + "grad_norm": 5.3879566867096695, + "learning_rate": 9.45033822082551e-06, + "loss": 0.3021, + "step": 8002 + }, + { + "epoch": 0.17616112700858463, + "grad_norm": 0.6462646748609233, + "learning_rate": 9.450175726552387e-06, + "loss": 0.5236, + "step": 8003 + }, + { + "epoch": 0.1761831388950033, + "grad_norm": 1.3163186433733083, + "learning_rate": 9.450013209661486e-06, + "loss": 0.27, + "step": 8004 + }, + { + "epoch": 0.17620515078142196, + "grad_norm": 1.7709879798157484, + "learning_rate": 9.449850670153636e-06, + "loss": 0.2766, + "step": 8005 + }, + { + "epoch": 0.17622716266784064, + "grad_norm": 1.475539809413448, + "learning_rate": 9.449688108029656e-06, + "loss": 0.3984, + "step": 8006 + }, + { + "epoch": 0.1762491745542593, + "grad_norm": 1.3060013852541104, + "learning_rate": 9.44952552329038e-06, + "loss": 0.3295, + "step": 8007 + }, + { + "epoch": 0.17627118644067796, + "grad_norm": 0.684547913503104, + "learning_rate": 9.449362915936629e-06, + "loss": 0.5318, + "step": 8008 + }, + { + "epoch": 0.17629319832709664, + "grad_norm": 0.6585231252840567, + "learning_rate": 9.44920028596923e-06, + "loss": 0.5582, + "step": 8009 + }, + { + "epoch": 0.1763152102135153, + "grad_norm": 1.7116735157941663, + "learning_rate": 9.449037633389012e-06, + "loss": 0.3985, + "step": 8010 + }, + { + "epoch": 0.17633722209993397, + "grad_norm": 1.5311527228977537, + "learning_rate": 9.448874958196802e-06, + "loss": 0.3439, + "step": 8011 + }, + { + "epoch": 0.17635923398635264, + "grad_norm": 1.3660135182431337, + "learning_rate": 9.448712260393423e-06, + "loss": 0.3566, + "step": 8012 + }, + { + "epoch": 0.1763812458727713, + "grad_norm": 0.6354953233542758, + "learning_rate": 9.448549539979704e-06, + "loss": 0.5573, + "step": 8013 + }, + { + "epoch": 0.17640325775918997, + "grad_norm": 0.6357605193991015, + "learning_rate": 9.448386796956472e-06, + "loss": 0.5355, + "step": 8014 + }, + { + "epoch": 0.17642526964560862, + "grad_norm": 1.4168576988206665, + "learning_rate": 9.448224031324555e-06, + "loss": 0.3388, + "step": 8015 + }, + { + "epoch": 0.1764472815320273, + "grad_norm": 0.6085019536381822, + "learning_rate": 9.44806124308478e-06, + "loss": 0.5216, + "step": 8016 + }, + { + "epoch": 0.17646929341844597, + "grad_norm": 1.4468233707468794, + "learning_rate": 9.447898432237975e-06, + "loss": 0.384, + "step": 8017 + }, + { + "epoch": 0.17649130530486462, + "grad_norm": 1.405776238906281, + "learning_rate": 9.447735598784963e-06, + "loss": 0.3036, + "step": 8018 + }, + { + "epoch": 0.1765133171912833, + "grad_norm": 2.2054936318214793, + "learning_rate": 9.447572742726578e-06, + "loss": 0.334, + "step": 8019 + }, + { + "epoch": 0.17653532907770195, + "grad_norm": 1.3100700408350001, + "learning_rate": 9.447409864063644e-06, + "loss": 0.4081, + "step": 8020 + }, + { + "epoch": 0.17655734096412062, + "grad_norm": 1.5079719629524928, + "learning_rate": 9.447246962796988e-06, + "loss": 0.4057, + "step": 8021 + }, + { + "epoch": 0.1765793528505393, + "grad_norm": 1.1054313008579728, + "learning_rate": 9.447084038927441e-06, + "loss": 0.3178, + "step": 8022 + }, + { + "epoch": 0.17660136473695795, + "grad_norm": 1.8975808032199661, + "learning_rate": 9.446921092455828e-06, + "loss": 0.421, + "step": 8023 + }, + { + "epoch": 0.17662337662337663, + "grad_norm": 1.2853697970828608, + "learning_rate": 9.446758123382979e-06, + "loss": 0.3629, + "step": 8024 + }, + { + "epoch": 0.17664538850979528, + "grad_norm": 1.4668879974536233, + "learning_rate": 9.446595131709724e-06, + "loss": 0.3751, + "step": 8025 + }, + { + "epoch": 0.17666740039621395, + "grad_norm": 1.333134085889507, + "learning_rate": 9.446432117436887e-06, + "loss": 0.29, + "step": 8026 + }, + { + "epoch": 0.17668941228263263, + "grad_norm": 0.7145885396554541, + "learning_rate": 9.4462690805653e-06, + "loss": 0.5501, + "step": 8027 + }, + { + "epoch": 0.17671142416905128, + "grad_norm": 1.3010358242306606, + "learning_rate": 9.446106021095788e-06, + "loss": 0.3067, + "step": 8028 + }, + { + "epoch": 0.17673343605546996, + "grad_norm": 1.3551998626766828, + "learning_rate": 9.445942939029184e-06, + "loss": 0.3855, + "step": 8029 + }, + { + "epoch": 0.17675544794188863, + "grad_norm": 1.628414880845458, + "learning_rate": 9.445779834366316e-06, + "loss": 0.3662, + "step": 8030 + }, + { + "epoch": 0.17677745982830728, + "grad_norm": 1.7298959107534948, + "learning_rate": 9.445616707108012e-06, + "loss": 0.3204, + "step": 8031 + }, + { + "epoch": 0.17679947171472596, + "grad_norm": 1.295687161135026, + "learning_rate": 9.4454535572551e-06, + "loss": 0.3537, + "step": 8032 + }, + { + "epoch": 0.1768214836011446, + "grad_norm": 1.5730255738175125, + "learning_rate": 9.44529038480841e-06, + "loss": 0.4048, + "step": 8033 + }, + { + "epoch": 0.17684349548756328, + "grad_norm": 1.3812531260884, + "learning_rate": 9.445127189768773e-06, + "loss": 0.3994, + "step": 8034 + }, + { + "epoch": 0.17686550737398196, + "grad_norm": 1.590220983111435, + "learning_rate": 9.444963972137016e-06, + "loss": 0.3769, + "step": 8035 + }, + { + "epoch": 0.1768875192604006, + "grad_norm": 1.5835172272773501, + "learning_rate": 9.44480073191397e-06, + "loss": 0.3336, + "step": 8036 + }, + { + "epoch": 0.1769095311468193, + "grad_norm": 1.410260822461982, + "learning_rate": 9.444637469100464e-06, + "loss": 0.3265, + "step": 8037 + }, + { + "epoch": 0.17693154303323794, + "grad_norm": 1.559940352865225, + "learning_rate": 9.444474183697328e-06, + "loss": 0.4875, + "step": 8038 + }, + { + "epoch": 0.1769535549196566, + "grad_norm": 2.2701953879734935, + "learning_rate": 9.444310875705391e-06, + "loss": 0.33, + "step": 8039 + }, + { + "epoch": 0.1769755668060753, + "grad_norm": 1.7676901099149163, + "learning_rate": 9.444147545125485e-06, + "loss": 0.3669, + "step": 8040 + }, + { + "epoch": 0.17699757869249394, + "grad_norm": 2.0594575583718786, + "learning_rate": 9.443984191958438e-06, + "loss": 0.3571, + "step": 8041 + }, + { + "epoch": 0.17701959057891262, + "grad_norm": 2.743606708767103, + "learning_rate": 9.443820816205082e-06, + "loss": 0.363, + "step": 8042 + }, + { + "epoch": 0.17704160246533127, + "grad_norm": 1.3829219383565492, + "learning_rate": 9.443657417866247e-06, + "loss": 0.3819, + "step": 8043 + }, + { + "epoch": 0.17706361435174994, + "grad_norm": 1.3207317262746918, + "learning_rate": 9.443493996942762e-06, + "loss": 0.3611, + "step": 8044 + }, + { + "epoch": 0.17708562623816862, + "grad_norm": 1.2752497431807233, + "learning_rate": 9.443330553435461e-06, + "loss": 0.3389, + "step": 8045 + }, + { + "epoch": 0.17710763812458727, + "grad_norm": 1.4043045872249496, + "learning_rate": 9.443167087345171e-06, + "loss": 0.3915, + "step": 8046 + }, + { + "epoch": 0.17712965001100595, + "grad_norm": 1.1978969102391899, + "learning_rate": 9.443003598672723e-06, + "loss": 0.3183, + "step": 8047 + }, + { + "epoch": 0.17715166189742462, + "grad_norm": 1.5326197668666168, + "learning_rate": 9.44284008741895e-06, + "loss": 0.314, + "step": 8048 + }, + { + "epoch": 0.17717367378384327, + "grad_norm": 1.2459442338553244, + "learning_rate": 9.442676553584682e-06, + "loss": 0.3855, + "step": 8049 + }, + { + "epoch": 0.17719568567026195, + "grad_norm": 1.4663466397562963, + "learning_rate": 9.442512997170751e-06, + "loss": 0.4083, + "step": 8050 + }, + { + "epoch": 0.1772176975566806, + "grad_norm": 1.5075998040847791, + "learning_rate": 9.442349418177988e-06, + "loss": 0.5113, + "step": 8051 + }, + { + "epoch": 0.17723970944309927, + "grad_norm": 1.822001340909613, + "learning_rate": 9.442185816607222e-06, + "loss": 0.2855, + "step": 8052 + }, + { + "epoch": 0.17726172132951795, + "grad_norm": 0.7437509831660596, + "learning_rate": 9.442022192459288e-06, + "loss": 0.5441, + "step": 8053 + }, + { + "epoch": 0.1772837332159366, + "grad_norm": 0.7006060052378776, + "learning_rate": 9.441858545735014e-06, + "loss": 0.5543, + "step": 8054 + }, + { + "epoch": 0.17730574510235528, + "grad_norm": 1.469178725603231, + "learning_rate": 9.441694876435236e-06, + "loss": 0.3324, + "step": 8055 + }, + { + "epoch": 0.17732775698877393, + "grad_norm": 1.7533496988262054, + "learning_rate": 9.441531184560782e-06, + "loss": 0.307, + "step": 8056 + }, + { + "epoch": 0.1773497688751926, + "grad_norm": 1.702529509966896, + "learning_rate": 9.441367470112485e-06, + "loss": 0.3281, + "step": 8057 + }, + { + "epoch": 0.17737178076161128, + "grad_norm": 1.5154379538464595, + "learning_rate": 9.44120373309118e-06, + "loss": 0.3707, + "step": 8058 + }, + { + "epoch": 0.17739379264802993, + "grad_norm": 1.5973025807601628, + "learning_rate": 9.441039973497697e-06, + "loss": 0.3892, + "step": 8059 + }, + { + "epoch": 0.1774158045344486, + "grad_norm": 2.325953936161364, + "learning_rate": 9.440876191332866e-06, + "loss": 0.3321, + "step": 8060 + }, + { + "epoch": 0.17743781642086726, + "grad_norm": 2.957699366683063, + "learning_rate": 9.440712386597522e-06, + "loss": 0.3508, + "step": 8061 + }, + { + "epoch": 0.17745982830728593, + "grad_norm": 1.339257381876101, + "learning_rate": 9.440548559292498e-06, + "loss": 0.2242, + "step": 8062 + }, + { + "epoch": 0.1774818401937046, + "grad_norm": 1.6928173617148048, + "learning_rate": 9.440384709418625e-06, + "loss": 0.3115, + "step": 8063 + }, + { + "epoch": 0.17750385208012326, + "grad_norm": 1.460994965448748, + "learning_rate": 9.440220836976734e-06, + "loss": 0.4064, + "step": 8064 + }, + { + "epoch": 0.17752586396654194, + "grad_norm": 1.2613602726171644, + "learning_rate": 9.440056941967664e-06, + "loss": 0.5648, + "step": 8065 + }, + { + "epoch": 0.17754787585296058, + "grad_norm": 1.5195285115773627, + "learning_rate": 9.439893024392243e-06, + "loss": 0.3485, + "step": 8066 + }, + { + "epoch": 0.17756988773937926, + "grad_norm": 1.6328267009000301, + "learning_rate": 9.439729084251306e-06, + "loss": 0.3388, + "step": 8067 + }, + { + "epoch": 0.17759189962579794, + "grad_norm": 1.6866232359512106, + "learning_rate": 9.439565121545683e-06, + "loss": 0.3876, + "step": 8068 + }, + { + "epoch": 0.1776139115122166, + "grad_norm": 0.6460202845423827, + "learning_rate": 9.439401136276213e-06, + "loss": 0.5533, + "step": 8069 + }, + { + "epoch": 0.17763592339863526, + "grad_norm": 1.870938873448558, + "learning_rate": 9.439237128443726e-06, + "loss": 0.3714, + "step": 8070 + }, + { + "epoch": 0.17765793528505394, + "grad_norm": 1.7226089374253561, + "learning_rate": 9.439073098049055e-06, + "loss": 0.3785, + "step": 8071 + }, + { + "epoch": 0.1776799471714726, + "grad_norm": 1.689911464176852, + "learning_rate": 9.438909045093034e-06, + "loss": 0.4268, + "step": 8072 + }, + { + "epoch": 0.17770195905789127, + "grad_norm": 1.8096020988122856, + "learning_rate": 9.438744969576499e-06, + "loss": 0.3454, + "step": 8073 + }, + { + "epoch": 0.17772397094430992, + "grad_norm": 2.030630715152709, + "learning_rate": 9.438580871500282e-06, + "loss": 0.3205, + "step": 8074 + }, + { + "epoch": 0.1777459828307286, + "grad_norm": 2.5528127430051577, + "learning_rate": 9.438416750865215e-06, + "loss": 0.4452, + "step": 8075 + }, + { + "epoch": 0.17776799471714727, + "grad_norm": 1.6495486380586366, + "learning_rate": 9.438252607672139e-06, + "loss": 0.3078, + "step": 8076 + }, + { + "epoch": 0.17779000660356592, + "grad_norm": 1.2625495121607333, + "learning_rate": 9.43808844192188e-06, + "loss": 0.2317, + "step": 8077 + }, + { + "epoch": 0.1778120184899846, + "grad_norm": 0.9936650699618244, + "learning_rate": 9.437924253615278e-06, + "loss": 0.5638, + "step": 8078 + }, + { + "epoch": 0.17783403037640325, + "grad_norm": 1.522808145071241, + "learning_rate": 9.437760042753165e-06, + "loss": 0.3779, + "step": 8079 + }, + { + "epoch": 0.17785604226282192, + "grad_norm": 1.7130581846934787, + "learning_rate": 9.437595809336375e-06, + "loss": 0.3129, + "step": 8080 + }, + { + "epoch": 0.1778780541492406, + "grad_norm": 6.6907211964328726, + "learning_rate": 9.437431553365745e-06, + "loss": 0.3513, + "step": 8081 + }, + { + "epoch": 0.17790006603565925, + "grad_norm": 7.912962594346851, + "learning_rate": 9.437267274842108e-06, + "loss": 0.464, + "step": 8082 + }, + { + "epoch": 0.17792207792207793, + "grad_norm": 2.6681611769012186, + "learning_rate": 9.437102973766302e-06, + "loss": 0.3308, + "step": 8083 + }, + { + "epoch": 0.17794408980849657, + "grad_norm": 1.5121986748311644, + "learning_rate": 9.436938650139158e-06, + "loss": 0.3097, + "step": 8084 + }, + { + "epoch": 0.17796610169491525, + "grad_norm": 7.282973444446293, + "learning_rate": 9.436774303961512e-06, + "loss": 0.3236, + "step": 8085 + }, + { + "epoch": 0.17798811358133393, + "grad_norm": 2.5729853226148256, + "learning_rate": 9.436609935234202e-06, + "loss": 0.3077, + "step": 8086 + }, + { + "epoch": 0.17801012546775258, + "grad_norm": 0.6750026014418056, + "learning_rate": 9.43644554395806e-06, + "loss": 0.5485, + "step": 8087 + }, + { + "epoch": 0.17803213735417125, + "grad_norm": 1.3654835455506358, + "learning_rate": 9.436281130133923e-06, + "loss": 0.408, + "step": 8088 + }, + { + "epoch": 0.17805414924058993, + "grad_norm": 1.4706107926930754, + "learning_rate": 9.436116693762627e-06, + "loss": 0.351, + "step": 8089 + }, + { + "epoch": 0.17807616112700858, + "grad_norm": 2.311667397475236, + "learning_rate": 9.435952234845007e-06, + "loss": 0.4106, + "step": 8090 + }, + { + "epoch": 0.17809817301342726, + "grad_norm": 1.653322715928849, + "learning_rate": 9.4357877533819e-06, + "loss": 0.3532, + "step": 8091 + }, + { + "epoch": 0.1781201848998459, + "grad_norm": 2.6276284819696905, + "learning_rate": 9.435623249374141e-06, + "loss": 0.359, + "step": 8092 + }, + { + "epoch": 0.17814219678626458, + "grad_norm": 0.6222048307403437, + "learning_rate": 9.435458722822567e-06, + "loss": 0.5119, + "step": 8093 + }, + { + "epoch": 0.17816420867268326, + "grad_norm": 1.529620317315872, + "learning_rate": 9.435294173728014e-06, + "loss": 0.3122, + "step": 8094 + }, + { + "epoch": 0.1781862205591019, + "grad_norm": 1.3161779397596338, + "learning_rate": 9.435129602091315e-06, + "loss": 0.29, + "step": 8095 + }, + { + "epoch": 0.1782082324455206, + "grad_norm": 1.2765910537785568, + "learning_rate": 9.434965007913312e-06, + "loss": 0.267, + "step": 8096 + }, + { + "epoch": 0.17823024433193924, + "grad_norm": 1.3860908896301996, + "learning_rate": 9.434800391194836e-06, + "loss": 0.3145, + "step": 8097 + }, + { + "epoch": 0.1782522562183579, + "grad_norm": 0.680156966289689, + "learning_rate": 9.43463575193673e-06, + "loss": 0.5521, + "step": 8098 + }, + { + "epoch": 0.1782742681047766, + "grad_norm": 0.6330307068794754, + "learning_rate": 9.434471090139825e-06, + "loss": 0.5369, + "step": 8099 + }, + { + "epoch": 0.17829627999119524, + "grad_norm": 1.5421234076403378, + "learning_rate": 9.43430640580496e-06, + "loss": 0.4132, + "step": 8100 + }, + { + "epoch": 0.17831829187761392, + "grad_norm": 1.1614959934625693, + "learning_rate": 9.434141698932972e-06, + "loss": 0.2994, + "step": 8101 + }, + { + "epoch": 0.17834030376403257, + "grad_norm": 1.3928127414703397, + "learning_rate": 9.433976969524698e-06, + "loss": 0.2541, + "step": 8102 + }, + { + "epoch": 0.17836231565045124, + "grad_norm": 1.3620238827486333, + "learning_rate": 9.433812217580977e-06, + "loss": 0.2732, + "step": 8103 + }, + { + "epoch": 0.17838432753686992, + "grad_norm": 0.669574048536287, + "learning_rate": 9.433647443102644e-06, + "loss": 0.5326, + "step": 8104 + }, + { + "epoch": 0.17840633942328857, + "grad_norm": 1.298009530195013, + "learning_rate": 9.433482646090536e-06, + "loss": 0.4168, + "step": 8105 + }, + { + "epoch": 0.17842835130970724, + "grad_norm": 1.360393110829146, + "learning_rate": 9.433317826545493e-06, + "loss": 0.3817, + "step": 8106 + }, + { + "epoch": 0.17845036319612592, + "grad_norm": 1.1267345613336395, + "learning_rate": 9.433152984468353e-06, + "loss": 0.3015, + "step": 8107 + }, + { + "epoch": 0.17847237508254457, + "grad_norm": 1.59443431911818, + "learning_rate": 9.432988119859951e-06, + "loss": 0.2529, + "step": 8108 + }, + { + "epoch": 0.17849438696896325, + "grad_norm": 1.5285508025659889, + "learning_rate": 9.432823232721125e-06, + "loss": 0.3635, + "step": 8109 + }, + { + "epoch": 0.1785163988553819, + "grad_norm": 4.234977347622446, + "learning_rate": 9.432658323052715e-06, + "loss": 0.486, + "step": 8110 + }, + { + "epoch": 0.17853841074180057, + "grad_norm": 1.3155230525202681, + "learning_rate": 9.43249339085556e-06, + "loss": 0.4032, + "step": 8111 + }, + { + "epoch": 0.17856042262821925, + "grad_norm": 2.2001373426060815, + "learning_rate": 9.432328436130493e-06, + "loss": 0.3425, + "step": 8112 + }, + { + "epoch": 0.1785824345146379, + "grad_norm": 1.5350933342492845, + "learning_rate": 9.432163458878359e-06, + "loss": 0.3672, + "step": 8113 + }, + { + "epoch": 0.17860444640105658, + "grad_norm": 1.3017830363700609, + "learning_rate": 9.431998459099994e-06, + "loss": 0.378, + "step": 8114 + }, + { + "epoch": 0.17862645828747523, + "grad_norm": 1.5013590146049403, + "learning_rate": 9.431833436796234e-06, + "loss": 0.2543, + "step": 8115 + }, + { + "epoch": 0.1786484701738939, + "grad_norm": 1.4690803080596593, + "learning_rate": 9.43166839196792e-06, + "loss": 0.3411, + "step": 8116 + }, + { + "epoch": 0.17867048206031258, + "grad_norm": 2.1869046717798852, + "learning_rate": 9.431503324615891e-06, + "loss": 0.2852, + "step": 8117 + }, + { + "epoch": 0.17869249394673123, + "grad_norm": 0.7336609964163191, + "learning_rate": 9.431338234740986e-06, + "loss": 0.5319, + "step": 8118 + }, + { + "epoch": 0.1787145058331499, + "grad_norm": 1.4395688396984727, + "learning_rate": 9.431173122344045e-06, + "loss": 0.3197, + "step": 8119 + }, + { + "epoch": 0.17873651771956856, + "grad_norm": 0.6544402884461854, + "learning_rate": 9.431007987425905e-06, + "loss": 0.5612, + "step": 8120 + }, + { + "epoch": 0.17875852960598723, + "grad_norm": 2.3085885095760896, + "learning_rate": 9.430842829987403e-06, + "loss": 0.3767, + "step": 8121 + }, + { + "epoch": 0.1787805414924059, + "grad_norm": 1.5258585662367945, + "learning_rate": 9.430677650029385e-06, + "loss": 0.3639, + "step": 8122 + }, + { + "epoch": 0.17880255337882456, + "grad_norm": 1.7195369234399667, + "learning_rate": 9.430512447552687e-06, + "loss": 0.3802, + "step": 8123 + }, + { + "epoch": 0.17882456526524323, + "grad_norm": 0.6319993357018807, + "learning_rate": 9.430347222558148e-06, + "loss": 0.5443, + "step": 8124 + }, + { + "epoch": 0.17884657715166188, + "grad_norm": 1.3689327184397002, + "learning_rate": 9.430181975046607e-06, + "loss": 0.3641, + "step": 8125 + }, + { + "epoch": 0.17886858903808056, + "grad_norm": 1.6697287404985537, + "learning_rate": 9.430016705018907e-06, + "loss": 0.4232, + "step": 8126 + }, + { + "epoch": 0.17889060092449924, + "grad_norm": 1.5742791009522956, + "learning_rate": 9.429851412475886e-06, + "loss": 0.3038, + "step": 8127 + }, + { + "epoch": 0.1789126128109179, + "grad_norm": 1.5903134334324132, + "learning_rate": 9.429686097418383e-06, + "loss": 0.3073, + "step": 8128 + }, + { + "epoch": 0.17893462469733656, + "grad_norm": 1.438039418079565, + "learning_rate": 9.429520759847241e-06, + "loss": 0.4165, + "step": 8129 + }, + { + "epoch": 0.17895663658375524, + "grad_norm": 1.5411792041317653, + "learning_rate": 9.429355399763298e-06, + "loss": 0.2663, + "step": 8130 + }, + { + "epoch": 0.1789786484701739, + "grad_norm": 1.2751214181475095, + "learning_rate": 9.429190017167395e-06, + "loss": 0.3266, + "step": 8131 + }, + { + "epoch": 0.17900066035659257, + "grad_norm": 1.2832722772625238, + "learning_rate": 9.429024612060375e-06, + "loss": 0.3148, + "step": 8132 + }, + { + "epoch": 0.17902267224301122, + "grad_norm": 0.7576225768848713, + "learning_rate": 9.428859184443076e-06, + "loss": 0.5529, + "step": 8133 + }, + { + "epoch": 0.1790446841294299, + "grad_norm": 1.5428185244486634, + "learning_rate": 9.428693734316338e-06, + "loss": 0.3672, + "step": 8134 + }, + { + "epoch": 0.17906669601584857, + "grad_norm": 1.4508157740758612, + "learning_rate": 9.428528261681003e-06, + "loss": 0.3335, + "step": 8135 + }, + { + "epoch": 0.17908870790226722, + "grad_norm": 1.6839509127465955, + "learning_rate": 9.428362766537913e-06, + "loss": 0.3421, + "step": 8136 + }, + { + "epoch": 0.1791107197886859, + "grad_norm": 0.6272716148819735, + "learning_rate": 9.428197248887908e-06, + "loss": 0.5538, + "step": 8137 + }, + { + "epoch": 0.17913273167510455, + "grad_norm": 0.6265077895130857, + "learning_rate": 9.42803170873183e-06, + "loss": 0.5136, + "step": 8138 + }, + { + "epoch": 0.17915474356152322, + "grad_norm": 1.6039520611640887, + "learning_rate": 9.427866146070519e-06, + "loss": 0.3677, + "step": 8139 + }, + { + "epoch": 0.1791767554479419, + "grad_norm": 1.4967433153012948, + "learning_rate": 9.427700560904819e-06, + "loss": 0.4707, + "step": 8140 + }, + { + "epoch": 0.17919876733436055, + "grad_norm": 0.6518110615752443, + "learning_rate": 9.427534953235568e-06, + "loss": 0.5381, + "step": 8141 + }, + { + "epoch": 0.17922077922077922, + "grad_norm": 1.2778017824227315, + "learning_rate": 9.427369323063612e-06, + "loss": 0.3721, + "step": 8142 + }, + { + "epoch": 0.17924279110719787, + "grad_norm": 0.6472830622300005, + "learning_rate": 9.427203670389788e-06, + "loss": 0.5298, + "step": 8143 + }, + { + "epoch": 0.17926480299361655, + "grad_norm": 1.724133681770954, + "learning_rate": 9.427037995214942e-06, + "loss": 0.4429, + "step": 8144 + }, + { + "epoch": 0.17928681488003523, + "grad_norm": 1.7816264235563164, + "learning_rate": 9.426872297539912e-06, + "loss": 0.2955, + "step": 8145 + }, + { + "epoch": 0.17930882676645388, + "grad_norm": 0.6417856990586588, + "learning_rate": 9.426706577365545e-06, + "loss": 0.5527, + "step": 8146 + }, + { + "epoch": 0.17933083865287255, + "grad_norm": 1.3656436573456043, + "learning_rate": 9.426540834692681e-06, + "loss": 0.3473, + "step": 8147 + }, + { + "epoch": 0.17935285053929123, + "grad_norm": 1.4852709093686058, + "learning_rate": 9.42637506952216e-06, + "loss": 0.4778, + "step": 8148 + }, + { + "epoch": 0.17937486242570988, + "grad_norm": 1.7509934570438463, + "learning_rate": 9.426209281854829e-06, + "loss": 0.3632, + "step": 8149 + }, + { + "epoch": 0.17939687431212856, + "grad_norm": 0.6132842605926412, + "learning_rate": 9.426043471691527e-06, + "loss": 0.5437, + "step": 8150 + }, + { + "epoch": 0.1794188861985472, + "grad_norm": 1.5207512388359543, + "learning_rate": 9.425877639033098e-06, + "loss": 0.3916, + "step": 8151 + }, + { + "epoch": 0.17944089808496588, + "grad_norm": 1.4104775228455422, + "learning_rate": 9.425711783880385e-06, + "loss": 0.3656, + "step": 8152 + }, + { + "epoch": 0.17946290997138456, + "grad_norm": 1.410086935939977, + "learning_rate": 9.425545906234233e-06, + "loss": 0.359, + "step": 8153 + }, + { + "epoch": 0.1794849218578032, + "grad_norm": 1.2641032054217303, + "learning_rate": 9.42538000609548e-06, + "loss": 0.3848, + "step": 8154 + }, + { + "epoch": 0.17950693374422189, + "grad_norm": 1.526215066293386, + "learning_rate": 9.425214083464972e-06, + "loss": 0.4069, + "step": 8155 + }, + { + "epoch": 0.17952894563064054, + "grad_norm": 0.6901866264087085, + "learning_rate": 9.425048138343553e-06, + "loss": 0.5906, + "step": 8156 + }, + { + "epoch": 0.1795509575170592, + "grad_norm": 1.7374947242996281, + "learning_rate": 9.424882170732067e-06, + "loss": 0.4041, + "step": 8157 + }, + { + "epoch": 0.1795729694034779, + "grad_norm": 1.3547984604852084, + "learning_rate": 9.424716180631355e-06, + "loss": 0.3788, + "step": 8158 + }, + { + "epoch": 0.17959498128989654, + "grad_norm": 1.7406289892941305, + "learning_rate": 9.424550168042261e-06, + "loss": 0.3041, + "step": 8159 + }, + { + "epoch": 0.17961699317631521, + "grad_norm": 2.1545463972598364, + "learning_rate": 9.424384132965631e-06, + "loss": 0.4161, + "step": 8160 + }, + { + "epoch": 0.17963900506273386, + "grad_norm": 0.649124562195144, + "learning_rate": 9.424218075402307e-06, + "loss": 0.536, + "step": 8161 + }, + { + "epoch": 0.17966101694915254, + "grad_norm": 0.6401840371418351, + "learning_rate": 9.424051995353134e-06, + "loss": 0.5265, + "step": 8162 + }, + { + "epoch": 0.17968302883557122, + "grad_norm": 1.4233861839548474, + "learning_rate": 9.423885892818954e-06, + "loss": 0.3392, + "step": 8163 + }, + { + "epoch": 0.17970504072198987, + "grad_norm": 1.8939948567952387, + "learning_rate": 9.423719767800614e-06, + "loss": 0.341, + "step": 8164 + }, + { + "epoch": 0.17972705260840854, + "grad_norm": 0.6092440168027989, + "learning_rate": 9.423553620298957e-06, + "loss": 0.5139, + "step": 8165 + }, + { + "epoch": 0.1797490644948272, + "grad_norm": 1.2925612083933573, + "learning_rate": 9.423387450314829e-06, + "loss": 0.324, + "step": 8166 + }, + { + "epoch": 0.17977107638124587, + "grad_norm": 1.6415585980735807, + "learning_rate": 9.423221257849071e-06, + "loss": 0.393, + "step": 8167 + }, + { + "epoch": 0.17979308826766455, + "grad_norm": 1.4316937883401957, + "learning_rate": 9.423055042902532e-06, + "loss": 0.3587, + "step": 8168 + }, + { + "epoch": 0.1798151001540832, + "grad_norm": 1.52874018018792, + "learning_rate": 9.422888805476052e-06, + "loss": 0.3993, + "step": 8169 + }, + { + "epoch": 0.17983711204050187, + "grad_norm": 1.3864702463204006, + "learning_rate": 9.42272254557048e-06, + "loss": 0.3323, + "step": 8170 + }, + { + "epoch": 0.17985912392692055, + "grad_norm": 1.27374123644285, + "learning_rate": 9.422556263186659e-06, + "loss": 0.3149, + "step": 8171 + }, + { + "epoch": 0.1798811358133392, + "grad_norm": 2.033045098102272, + "learning_rate": 9.422389958325433e-06, + "loss": 0.457, + "step": 8172 + }, + { + "epoch": 0.17990314769975788, + "grad_norm": 1.6332239434088065, + "learning_rate": 9.42222363098765e-06, + "loss": 0.3811, + "step": 8173 + }, + { + "epoch": 0.17992515958617653, + "grad_norm": 2.6355823807616403, + "learning_rate": 9.422057281174155e-06, + "loss": 0.3512, + "step": 8174 + }, + { + "epoch": 0.1799471714725952, + "grad_norm": 1.3552162860508963, + "learning_rate": 9.421890908885793e-06, + "loss": 0.2788, + "step": 8175 + }, + { + "epoch": 0.17996918335901388, + "grad_norm": 1.572010018757032, + "learning_rate": 9.421724514123407e-06, + "loss": 0.418, + "step": 8176 + }, + { + "epoch": 0.17999119524543253, + "grad_norm": 1.3511734656219905, + "learning_rate": 9.421558096887847e-06, + "loss": 0.4244, + "step": 8177 + }, + { + "epoch": 0.1800132071318512, + "grad_norm": 1.3825682887797937, + "learning_rate": 9.421391657179955e-06, + "loss": 0.3298, + "step": 8178 + }, + { + "epoch": 0.18003521901826985, + "grad_norm": 1.3989811698127683, + "learning_rate": 9.42122519500058e-06, + "loss": 0.2992, + "step": 8179 + }, + { + "epoch": 0.18005723090468853, + "grad_norm": 0.7334641376733521, + "learning_rate": 9.421058710350566e-06, + "loss": 0.5427, + "step": 8180 + }, + { + "epoch": 0.1800792427911072, + "grad_norm": 1.0293449974246458, + "learning_rate": 9.42089220323076e-06, + "loss": 0.3295, + "step": 8181 + }, + { + "epoch": 0.18010125467752586, + "grad_norm": 0.6482908155612999, + "learning_rate": 9.420725673642007e-06, + "loss": 0.5326, + "step": 8182 + }, + { + "epoch": 0.18012326656394453, + "grad_norm": 1.7069129482574115, + "learning_rate": 9.420559121585156e-06, + "loss": 0.4368, + "step": 8183 + }, + { + "epoch": 0.18014527845036318, + "grad_norm": 1.390824609231277, + "learning_rate": 9.42039254706105e-06, + "loss": 0.4678, + "step": 8184 + }, + { + "epoch": 0.18016729033678186, + "grad_norm": 1.6899188883278347, + "learning_rate": 9.420225950070538e-06, + "loss": 0.3939, + "step": 8185 + }, + { + "epoch": 0.18018930222320054, + "grad_norm": 1.8231077926544674, + "learning_rate": 9.420059330614468e-06, + "loss": 0.4145, + "step": 8186 + }, + { + "epoch": 0.1802113141096192, + "grad_norm": 1.9581682723718488, + "learning_rate": 9.419892688693684e-06, + "loss": 0.34, + "step": 8187 + }, + { + "epoch": 0.18023332599603786, + "grad_norm": 0.7090273767757618, + "learning_rate": 9.419726024309034e-06, + "loss": 0.5428, + "step": 8188 + }, + { + "epoch": 0.18025533788245654, + "grad_norm": 1.9068890250712953, + "learning_rate": 9.419559337461365e-06, + "loss": 0.3387, + "step": 8189 + }, + { + "epoch": 0.1802773497688752, + "grad_norm": 0.6563051312401039, + "learning_rate": 9.419392628151524e-06, + "loss": 0.5479, + "step": 8190 + }, + { + "epoch": 0.18029936165529387, + "grad_norm": 2.5513293499254557, + "learning_rate": 9.41922589638036e-06, + "loss": 0.3847, + "step": 8191 + }, + { + "epoch": 0.18032137354171252, + "grad_norm": 1.2097159063027527, + "learning_rate": 9.419059142148717e-06, + "loss": 0.3308, + "step": 8192 + }, + { + "epoch": 0.1803433854281312, + "grad_norm": 1.3357470991737332, + "learning_rate": 9.418892365457446e-06, + "loss": 0.3478, + "step": 8193 + }, + { + "epoch": 0.18036539731454987, + "grad_norm": 1.2815196597892407, + "learning_rate": 9.418725566307392e-06, + "loss": 0.3179, + "step": 8194 + }, + { + "epoch": 0.18038740920096852, + "grad_norm": 1.5074475397139864, + "learning_rate": 9.418558744699404e-06, + "loss": 0.3661, + "step": 8195 + }, + { + "epoch": 0.1804094210873872, + "grad_norm": 1.409274782716209, + "learning_rate": 9.418391900634332e-06, + "loss": 0.3722, + "step": 8196 + }, + { + "epoch": 0.18043143297380584, + "grad_norm": 1.2480037352250426, + "learning_rate": 9.41822503411302e-06, + "loss": 0.369, + "step": 8197 + }, + { + "epoch": 0.18045344486022452, + "grad_norm": 2.3618165701342706, + "learning_rate": 9.418058145136316e-06, + "loss": 0.3048, + "step": 8198 + }, + { + "epoch": 0.1804754567466432, + "grad_norm": 2.3487511261932257, + "learning_rate": 9.417891233705073e-06, + "loss": 0.3142, + "step": 8199 + }, + { + "epoch": 0.18049746863306185, + "grad_norm": 1.5996110739828031, + "learning_rate": 9.417724299820133e-06, + "loss": 0.3902, + "step": 8200 + }, + { + "epoch": 0.18051948051948052, + "grad_norm": 1.4650950360090818, + "learning_rate": 9.417557343482352e-06, + "loss": 0.3441, + "step": 8201 + }, + { + "epoch": 0.18054149240589917, + "grad_norm": 1.681939989742797, + "learning_rate": 9.41739036469257e-06, + "loss": 0.4456, + "step": 8202 + }, + { + "epoch": 0.18056350429231785, + "grad_norm": 1.3774507609294164, + "learning_rate": 9.417223363451643e-06, + "loss": 0.3351, + "step": 8203 + }, + { + "epoch": 0.18058551617873653, + "grad_norm": 1.930378978195838, + "learning_rate": 9.417056339760415e-06, + "loss": 0.2591, + "step": 8204 + }, + { + "epoch": 0.18060752806515518, + "grad_norm": 0.7773777186637353, + "learning_rate": 9.416889293619737e-06, + "loss": 0.554, + "step": 8205 + }, + { + "epoch": 0.18062953995157385, + "grad_norm": 1.490275832742489, + "learning_rate": 9.416722225030458e-06, + "loss": 0.3833, + "step": 8206 + }, + { + "epoch": 0.1806515518379925, + "grad_norm": 1.5304608979049605, + "learning_rate": 9.416555133993425e-06, + "loss": 0.3826, + "step": 8207 + }, + { + "epoch": 0.18067356372441118, + "grad_norm": 0.6851881348487716, + "learning_rate": 9.416388020509492e-06, + "loss": 0.5459, + "step": 8208 + }, + { + "epoch": 0.18069557561082986, + "grad_norm": 1.4990931499335423, + "learning_rate": 9.416220884579503e-06, + "loss": 0.3345, + "step": 8209 + }, + { + "epoch": 0.1807175874972485, + "grad_norm": 1.779123201373564, + "learning_rate": 9.416053726204312e-06, + "loss": 0.4344, + "step": 8210 + }, + { + "epoch": 0.18073959938366718, + "grad_norm": 1.9580004733522431, + "learning_rate": 9.415886545384764e-06, + "loss": 0.3433, + "step": 8211 + }, + { + "epoch": 0.18076161127008586, + "grad_norm": 1.4374845585246008, + "learning_rate": 9.415719342121712e-06, + "loss": 0.2996, + "step": 8212 + }, + { + "epoch": 0.1807836231565045, + "grad_norm": 1.5161708734162813, + "learning_rate": 9.415552116416005e-06, + "loss": 0.3758, + "step": 8213 + }, + { + "epoch": 0.18080563504292319, + "grad_norm": 1.7891142781396416, + "learning_rate": 9.415384868268493e-06, + "loss": 0.3503, + "step": 8214 + }, + { + "epoch": 0.18082764692934183, + "grad_norm": 1.1797157726740706, + "learning_rate": 9.415217597680026e-06, + "loss": 0.3013, + "step": 8215 + }, + { + "epoch": 0.1808496588157605, + "grad_norm": 0.7906249686109846, + "learning_rate": 9.415050304651455e-06, + "loss": 0.5779, + "step": 8216 + }, + { + "epoch": 0.1808716707021792, + "grad_norm": 2.1317000827091226, + "learning_rate": 9.414882989183625e-06, + "loss": 0.2893, + "step": 8217 + }, + { + "epoch": 0.18089368258859784, + "grad_norm": 1.5994611297401637, + "learning_rate": 9.414715651277395e-06, + "loss": 0.3701, + "step": 8218 + }, + { + "epoch": 0.18091569447501651, + "grad_norm": 1.570708759699205, + "learning_rate": 9.414548290933609e-06, + "loss": 0.3457, + "step": 8219 + }, + { + "epoch": 0.18093770636143516, + "grad_norm": 0.6637258443375228, + "learning_rate": 9.41438090815312e-06, + "loss": 0.5595, + "step": 8220 + }, + { + "epoch": 0.18095971824785384, + "grad_norm": 0.6400671536714599, + "learning_rate": 9.414213502936777e-06, + "loss": 0.5308, + "step": 8221 + }, + { + "epoch": 0.18098173013427252, + "grad_norm": 1.66958116905742, + "learning_rate": 9.414046075285433e-06, + "loss": 0.4323, + "step": 8222 + }, + { + "epoch": 0.18100374202069117, + "grad_norm": 1.5342142923702484, + "learning_rate": 9.41387862519994e-06, + "loss": 0.3333, + "step": 8223 + }, + { + "epoch": 0.18102575390710984, + "grad_norm": 0.7208356706648246, + "learning_rate": 9.413711152681144e-06, + "loss": 0.5553, + "step": 8224 + }, + { + "epoch": 0.1810477657935285, + "grad_norm": 2.015855794167602, + "learning_rate": 9.413543657729903e-06, + "loss": 0.4435, + "step": 8225 + }, + { + "epoch": 0.18106977767994717, + "grad_norm": 1.2270914072732386, + "learning_rate": 9.41337614034706e-06, + "loss": 0.2918, + "step": 8226 + }, + { + "epoch": 0.18109178956636585, + "grad_norm": 2.432696716575601, + "learning_rate": 9.413208600533475e-06, + "loss": 0.3926, + "step": 8227 + }, + { + "epoch": 0.1811138014527845, + "grad_norm": 1.5964209973604897, + "learning_rate": 9.413041038289994e-06, + "loss": 0.4412, + "step": 8228 + }, + { + "epoch": 0.18113581333920317, + "grad_norm": 1.1832763494779401, + "learning_rate": 9.41287345361747e-06, + "loss": 0.2581, + "step": 8229 + }, + { + "epoch": 0.18115782522562185, + "grad_norm": 0.7085992835680348, + "learning_rate": 9.412705846516756e-06, + "loss": 0.5516, + "step": 8230 + }, + { + "epoch": 0.1811798371120405, + "grad_norm": 1.5781264612622963, + "learning_rate": 9.412538216988703e-06, + "loss": 0.3588, + "step": 8231 + }, + { + "epoch": 0.18120184899845918, + "grad_norm": 1.8352730321535697, + "learning_rate": 9.41237056503416e-06, + "loss": 0.3415, + "step": 8232 + }, + { + "epoch": 0.18122386088487782, + "grad_norm": 1.7470958236996912, + "learning_rate": 9.412202890653984e-06, + "loss": 0.4197, + "step": 8233 + }, + { + "epoch": 0.1812458727712965, + "grad_norm": 0.7002965417218872, + "learning_rate": 9.412035193849024e-06, + "loss": 0.5253, + "step": 8234 + }, + { + "epoch": 0.18126788465771518, + "grad_norm": 1.4366624622548227, + "learning_rate": 9.411867474620135e-06, + "loss": 0.3481, + "step": 8235 + }, + { + "epoch": 0.18128989654413383, + "grad_norm": 0.6564175469817753, + "learning_rate": 9.411699732968167e-06, + "loss": 0.5538, + "step": 8236 + }, + { + "epoch": 0.1813119084305525, + "grad_norm": 2.2694020637873105, + "learning_rate": 9.411531968893972e-06, + "loss": 0.3336, + "step": 8237 + }, + { + "epoch": 0.18133392031697115, + "grad_norm": 1.0004767321714936, + "learning_rate": 9.411364182398405e-06, + "loss": 0.2712, + "step": 8238 + }, + { + "epoch": 0.18135593220338983, + "grad_norm": 5.470069326830297, + "learning_rate": 9.411196373482318e-06, + "loss": 0.31, + "step": 8239 + }, + { + "epoch": 0.1813779440898085, + "grad_norm": 1.4397036147834028, + "learning_rate": 9.411028542146564e-06, + "loss": 0.3373, + "step": 8240 + }, + { + "epoch": 0.18139995597622716, + "grad_norm": 3.315309989875267, + "learning_rate": 9.410860688391995e-06, + "loss": 0.4344, + "step": 8241 + }, + { + "epoch": 0.18142196786264583, + "grad_norm": 1.3054541295199416, + "learning_rate": 9.410692812219464e-06, + "loss": 0.4143, + "step": 8242 + }, + { + "epoch": 0.18144397974906448, + "grad_norm": 1.4043379137118381, + "learning_rate": 9.410524913629825e-06, + "loss": 0.3983, + "step": 8243 + }, + { + "epoch": 0.18146599163548316, + "grad_norm": 1.5539149889905053, + "learning_rate": 9.410356992623932e-06, + "loss": 0.3909, + "step": 8244 + }, + { + "epoch": 0.18148800352190184, + "grad_norm": 1.9974593170833328, + "learning_rate": 9.410189049202636e-06, + "loss": 0.298, + "step": 8245 + }, + { + "epoch": 0.18151001540832049, + "grad_norm": 1.6873360754902371, + "learning_rate": 9.410021083366794e-06, + "loss": 0.3122, + "step": 8246 + }, + { + "epoch": 0.18153202729473916, + "grad_norm": 1.2596487673124939, + "learning_rate": 9.409853095117259e-06, + "loss": 0.3621, + "step": 8247 + }, + { + "epoch": 0.1815540391811578, + "grad_norm": 1.4848617276325904, + "learning_rate": 9.40968508445488e-06, + "loss": 0.3139, + "step": 8248 + }, + { + "epoch": 0.1815760510675765, + "grad_norm": 1.175375150425756, + "learning_rate": 9.409517051380517e-06, + "loss": 0.2959, + "step": 8249 + }, + { + "epoch": 0.18159806295399517, + "grad_norm": 0.9804827967444474, + "learning_rate": 9.409348995895023e-06, + "loss": 0.588, + "step": 8250 + }, + { + "epoch": 0.18162007484041381, + "grad_norm": 1.3712072035059937, + "learning_rate": 9.409180917999248e-06, + "loss": 0.3156, + "step": 8251 + }, + { + "epoch": 0.1816420867268325, + "grad_norm": 1.3280267336789373, + "learning_rate": 9.40901281769405e-06, + "loss": 0.3004, + "step": 8252 + }, + { + "epoch": 0.18166409861325117, + "grad_norm": 0.6597904210218137, + "learning_rate": 9.408844694980283e-06, + "loss": 0.5567, + "step": 8253 + }, + { + "epoch": 0.18168611049966982, + "grad_norm": 0.6245663455344201, + "learning_rate": 9.408676549858801e-06, + "loss": 0.5285, + "step": 8254 + }, + { + "epoch": 0.1817081223860885, + "grad_norm": 1.776382336109812, + "learning_rate": 9.408508382330457e-06, + "loss": 0.3491, + "step": 8255 + }, + { + "epoch": 0.18173013427250714, + "grad_norm": 1.2432313903629548, + "learning_rate": 9.408340192396108e-06, + "loss": 0.2968, + "step": 8256 + }, + { + "epoch": 0.18175214615892582, + "grad_norm": 2.2385195083192273, + "learning_rate": 9.408171980056608e-06, + "loss": 0.3281, + "step": 8257 + }, + { + "epoch": 0.1817741580453445, + "grad_norm": 1.529455328873763, + "learning_rate": 9.408003745312813e-06, + "loss": 0.352, + "step": 8258 + }, + { + "epoch": 0.18179616993176315, + "grad_norm": 1.3260759497065104, + "learning_rate": 9.407835488165575e-06, + "loss": 0.2608, + "step": 8259 + }, + { + "epoch": 0.18181818181818182, + "grad_norm": 1.3033708281324725, + "learning_rate": 9.407667208615753e-06, + "loss": 0.4123, + "step": 8260 + }, + { + "epoch": 0.18184019370460047, + "grad_norm": 0.9038795809444267, + "learning_rate": 9.4074989066642e-06, + "loss": 0.5722, + "step": 8261 + }, + { + "epoch": 0.18186220559101915, + "grad_norm": 1.4534400623640167, + "learning_rate": 9.40733058231177e-06, + "loss": 0.3806, + "step": 8262 + }, + { + "epoch": 0.18188421747743783, + "grad_norm": 1.763484433731073, + "learning_rate": 9.407162235559322e-06, + "loss": 0.3501, + "step": 8263 + }, + { + "epoch": 0.18190622936385648, + "grad_norm": 1.5757100119130778, + "learning_rate": 9.40699386640771e-06, + "loss": 0.3517, + "step": 8264 + }, + { + "epoch": 0.18192824125027515, + "grad_norm": 1.4549505613194413, + "learning_rate": 9.406825474857788e-06, + "loss": 0.3676, + "step": 8265 + }, + { + "epoch": 0.1819502531366938, + "grad_norm": 0.6716505517779157, + "learning_rate": 9.406657060910414e-06, + "loss": 0.5542, + "step": 8266 + }, + { + "epoch": 0.18197226502311248, + "grad_norm": 0.6298630030333555, + "learning_rate": 9.406488624566443e-06, + "loss": 0.5453, + "step": 8267 + }, + { + "epoch": 0.18199427690953116, + "grad_norm": 1.488953821575799, + "learning_rate": 9.406320165826732e-06, + "loss": 0.3664, + "step": 8268 + }, + { + "epoch": 0.1820162887959498, + "grad_norm": 1.9301937788419525, + "learning_rate": 9.406151684692138e-06, + "loss": 0.345, + "step": 8269 + }, + { + "epoch": 0.18203830068236848, + "grad_norm": 1.5831305069400603, + "learning_rate": 9.405983181163513e-06, + "loss": 0.3196, + "step": 8270 + }, + { + "epoch": 0.18206031256878716, + "grad_norm": 0.6433151722062516, + "learning_rate": 9.405814655241719e-06, + "loss": 0.5274, + "step": 8271 + }, + { + "epoch": 0.1820823244552058, + "grad_norm": 1.235645824188852, + "learning_rate": 9.405646106927609e-06, + "loss": 0.2991, + "step": 8272 + }, + { + "epoch": 0.18210433634162448, + "grad_norm": 1.4257533217254577, + "learning_rate": 9.405477536222041e-06, + "loss": 0.3488, + "step": 8273 + }, + { + "epoch": 0.18212634822804313, + "grad_norm": 1.347676668611, + "learning_rate": 9.40530894312587e-06, + "loss": 0.2765, + "step": 8274 + }, + { + "epoch": 0.1821483601144618, + "grad_norm": 1.197812164928728, + "learning_rate": 9.405140327639956e-06, + "loss": 0.2587, + "step": 8275 + }, + { + "epoch": 0.1821703720008805, + "grad_norm": 1.4277838917936432, + "learning_rate": 9.404971689765153e-06, + "loss": 0.3302, + "step": 8276 + }, + { + "epoch": 0.18219238388729914, + "grad_norm": 1.4397994486972785, + "learning_rate": 9.404803029502319e-06, + "loss": 0.387, + "step": 8277 + }, + { + "epoch": 0.1822143957737178, + "grad_norm": 0.722755287104698, + "learning_rate": 9.404634346852312e-06, + "loss": 0.5562, + "step": 8278 + }, + { + "epoch": 0.18223640766013646, + "grad_norm": 1.3240407083370451, + "learning_rate": 9.404465641815987e-06, + "loss": 0.338, + "step": 8279 + }, + { + "epoch": 0.18225841954655514, + "grad_norm": 1.4842886572880294, + "learning_rate": 9.404296914394206e-06, + "loss": 0.4365, + "step": 8280 + }, + { + "epoch": 0.18228043143297382, + "grad_norm": 1.4704716099995117, + "learning_rate": 9.404128164587821e-06, + "loss": 0.3559, + "step": 8281 + }, + { + "epoch": 0.18230244331939247, + "grad_norm": 1.3671547046565906, + "learning_rate": 9.403959392397694e-06, + "loss": 0.3811, + "step": 8282 + }, + { + "epoch": 0.18232445520581114, + "grad_norm": 1.3144525796158302, + "learning_rate": 9.40379059782468e-06, + "loss": 0.3006, + "step": 8283 + }, + { + "epoch": 0.1823464670922298, + "grad_norm": 1.5155064224927655, + "learning_rate": 9.40362178086964e-06, + "loss": 0.3032, + "step": 8284 + }, + { + "epoch": 0.18236847897864847, + "grad_norm": 2.3033909469062737, + "learning_rate": 9.40345294153343e-06, + "loss": 0.4189, + "step": 8285 + }, + { + "epoch": 0.18239049086506715, + "grad_norm": 1.3522051648251967, + "learning_rate": 9.403284079816905e-06, + "loss": 0.3589, + "step": 8286 + }, + { + "epoch": 0.1824125027514858, + "grad_norm": 1.5476455481347513, + "learning_rate": 9.40311519572093e-06, + "loss": 0.414, + "step": 8287 + }, + { + "epoch": 0.18243451463790447, + "grad_norm": 1.7294713984347985, + "learning_rate": 9.402946289246356e-06, + "loss": 0.3666, + "step": 8288 + }, + { + "epoch": 0.18245652652432315, + "grad_norm": 1.6291512228648506, + "learning_rate": 9.402777360394048e-06, + "loss": 0.3986, + "step": 8289 + }, + { + "epoch": 0.1824785384107418, + "grad_norm": 1.8909184417135376, + "learning_rate": 9.402608409164861e-06, + "loss": 0.3705, + "step": 8290 + }, + { + "epoch": 0.18250055029716047, + "grad_norm": 1.4134059999217874, + "learning_rate": 9.402439435559654e-06, + "loss": 0.3338, + "step": 8291 + }, + { + "epoch": 0.18252256218357912, + "grad_norm": 1.7481931137861064, + "learning_rate": 9.402270439579286e-06, + "loss": 0.5126, + "step": 8292 + }, + { + "epoch": 0.1825445740699978, + "grad_norm": 1.5948017718926564, + "learning_rate": 9.402101421224618e-06, + "loss": 0.3831, + "step": 8293 + }, + { + "epoch": 0.18256658595641648, + "grad_norm": 0.6691545341512916, + "learning_rate": 9.401932380496505e-06, + "loss": 0.5621, + "step": 8294 + }, + { + "epoch": 0.18258859784283513, + "grad_norm": 1.4834132997982834, + "learning_rate": 9.40176331739581e-06, + "loss": 0.2689, + "step": 8295 + }, + { + "epoch": 0.1826106097292538, + "grad_norm": 0.6297419479779854, + "learning_rate": 9.40159423192339e-06, + "loss": 0.5028, + "step": 8296 + }, + { + "epoch": 0.18263262161567245, + "grad_norm": 0.6220603317375932, + "learning_rate": 9.401425124080106e-06, + "loss": 0.5418, + "step": 8297 + }, + { + "epoch": 0.18265463350209113, + "grad_norm": 1.342626332224305, + "learning_rate": 9.401255993866813e-06, + "loss": 0.3698, + "step": 8298 + }, + { + "epoch": 0.1826766453885098, + "grad_norm": 1.189140685352855, + "learning_rate": 9.401086841284377e-06, + "loss": 0.3226, + "step": 8299 + }, + { + "epoch": 0.18269865727492846, + "grad_norm": 1.6473572954208637, + "learning_rate": 9.400917666333655e-06, + "loss": 0.4469, + "step": 8300 + }, + { + "epoch": 0.18272066916134713, + "grad_norm": 1.3072109315924427, + "learning_rate": 9.400748469015507e-06, + "loss": 0.4098, + "step": 8301 + }, + { + "epoch": 0.18274268104776578, + "grad_norm": 0.6626838047883499, + "learning_rate": 9.40057924933079e-06, + "loss": 0.5559, + "step": 8302 + }, + { + "epoch": 0.18276469293418446, + "grad_norm": 1.1294116292560226, + "learning_rate": 9.400410007280366e-06, + "loss": 0.3651, + "step": 8303 + }, + { + "epoch": 0.18278670482060314, + "grad_norm": 0.6748779198907865, + "learning_rate": 9.4002407428651e-06, + "loss": 0.5838, + "step": 8304 + }, + { + "epoch": 0.18280871670702178, + "grad_norm": 1.3724693143431612, + "learning_rate": 9.400071456085845e-06, + "loss": 0.4562, + "step": 8305 + }, + { + "epoch": 0.18283072859344046, + "grad_norm": 1.529110419812115, + "learning_rate": 9.399902146943464e-06, + "loss": 0.3334, + "step": 8306 + }, + { + "epoch": 0.1828527404798591, + "grad_norm": 1.372888267990687, + "learning_rate": 9.399732815438818e-06, + "loss": 0.3685, + "step": 8307 + }, + { + "epoch": 0.1828747523662778, + "grad_norm": 1.443053158287593, + "learning_rate": 9.399563461572769e-06, + "loss": 0.3815, + "step": 8308 + }, + { + "epoch": 0.18289676425269646, + "grad_norm": 0.6582904375163843, + "learning_rate": 9.399394085346175e-06, + "loss": 0.5142, + "step": 8309 + }, + { + "epoch": 0.1829187761391151, + "grad_norm": 1.5269084259552936, + "learning_rate": 9.399224686759898e-06, + "loss": 0.3559, + "step": 8310 + }, + { + "epoch": 0.1829407880255338, + "grad_norm": 0.6253738850720774, + "learning_rate": 9.3990552658148e-06, + "loss": 0.5273, + "step": 8311 + }, + { + "epoch": 0.18296279991195247, + "grad_norm": 1.408744114827763, + "learning_rate": 9.398885822511741e-06, + "loss": 0.3236, + "step": 8312 + }, + { + "epoch": 0.18298481179837112, + "grad_norm": 2.6442018911078513, + "learning_rate": 9.39871635685158e-06, + "loss": 0.2916, + "step": 8313 + }, + { + "epoch": 0.1830068236847898, + "grad_norm": 0.6507492735937085, + "learning_rate": 9.398546868835183e-06, + "loss": 0.5371, + "step": 8314 + }, + { + "epoch": 0.18302883557120844, + "grad_norm": 1.7999487672540704, + "learning_rate": 9.398377358463408e-06, + "loss": 0.3587, + "step": 8315 + }, + { + "epoch": 0.18305084745762712, + "grad_norm": 1.617619417348229, + "learning_rate": 9.398207825737118e-06, + "loss": 0.2963, + "step": 8316 + }, + { + "epoch": 0.1830728593440458, + "grad_norm": 1.2904146828035667, + "learning_rate": 9.398038270657175e-06, + "loss": 0.3124, + "step": 8317 + }, + { + "epoch": 0.18309487123046445, + "grad_norm": 1.8792248098927677, + "learning_rate": 9.397868693224438e-06, + "loss": 0.3395, + "step": 8318 + }, + { + "epoch": 0.18311688311688312, + "grad_norm": 0.644704446678407, + "learning_rate": 9.397699093439773e-06, + "loss": 0.5235, + "step": 8319 + }, + { + "epoch": 0.18313889500330177, + "grad_norm": 0.659264083395479, + "learning_rate": 9.397529471304038e-06, + "loss": 0.5308, + "step": 8320 + }, + { + "epoch": 0.18316090688972045, + "grad_norm": 0.6443639941392194, + "learning_rate": 9.397359826818098e-06, + "loss": 0.5217, + "step": 8321 + }, + { + "epoch": 0.18318291877613913, + "grad_norm": 2.198801496266274, + "learning_rate": 9.397190159982814e-06, + "loss": 0.3253, + "step": 8322 + }, + { + "epoch": 0.18320493066255777, + "grad_norm": 1.6337145339396464, + "learning_rate": 9.397020470799047e-06, + "loss": 0.3793, + "step": 8323 + }, + { + "epoch": 0.18322694254897645, + "grad_norm": 1.7033851768307868, + "learning_rate": 9.396850759267662e-06, + "loss": 0.3744, + "step": 8324 + }, + { + "epoch": 0.1832489544353951, + "grad_norm": 0.681904538333848, + "learning_rate": 9.39668102538952e-06, + "loss": 0.533, + "step": 8325 + }, + { + "epoch": 0.18327096632181378, + "grad_norm": 1.505730323208555, + "learning_rate": 9.396511269165486e-06, + "loss": 0.3188, + "step": 8326 + }, + { + "epoch": 0.18329297820823245, + "grad_norm": 1.6438171792958716, + "learning_rate": 9.396341490596418e-06, + "loss": 0.3101, + "step": 8327 + }, + { + "epoch": 0.1833149900946511, + "grad_norm": 21.157602318399857, + "learning_rate": 9.396171689683183e-06, + "loss": 0.3534, + "step": 8328 + }, + { + "epoch": 0.18333700198106978, + "grad_norm": 2.7172535863961897, + "learning_rate": 9.396001866426643e-06, + "loss": 0.3546, + "step": 8329 + }, + { + "epoch": 0.18335901386748846, + "grad_norm": 1.3798652915315972, + "learning_rate": 9.39583202082766e-06, + "loss": 0.3605, + "step": 8330 + }, + { + "epoch": 0.1833810257539071, + "grad_norm": 1.428892861805491, + "learning_rate": 9.395662152887099e-06, + "loss": 0.3136, + "step": 8331 + }, + { + "epoch": 0.18340303764032578, + "grad_norm": 1.998274564010884, + "learning_rate": 9.395492262605821e-06, + "loss": 0.3332, + "step": 8332 + }, + { + "epoch": 0.18342504952674443, + "grad_norm": 1.4473098684848869, + "learning_rate": 9.395322349984693e-06, + "loss": 0.3343, + "step": 8333 + }, + { + "epoch": 0.1834470614131631, + "grad_norm": 10.3297873903873, + "learning_rate": 9.395152415024575e-06, + "loss": 0.3498, + "step": 8334 + }, + { + "epoch": 0.1834690732995818, + "grad_norm": 1.2885307309252911, + "learning_rate": 9.394982457726332e-06, + "loss": 0.3293, + "step": 8335 + }, + { + "epoch": 0.18349108518600044, + "grad_norm": 1.767693838154064, + "learning_rate": 9.394812478090828e-06, + "loss": 0.4098, + "step": 8336 + }, + { + "epoch": 0.1835130970724191, + "grad_norm": 1.1237560654111043, + "learning_rate": 9.394642476118927e-06, + "loss": 0.3817, + "step": 8337 + }, + { + "epoch": 0.18353510895883776, + "grad_norm": 0.674903664708887, + "learning_rate": 9.394472451811494e-06, + "loss": 0.5444, + "step": 8338 + }, + { + "epoch": 0.18355712084525644, + "grad_norm": 1.2595235595766694, + "learning_rate": 9.39430240516939e-06, + "loss": 0.2487, + "step": 8339 + }, + { + "epoch": 0.18357913273167512, + "grad_norm": 1.6738560006095928, + "learning_rate": 9.394132336193482e-06, + "loss": 0.2653, + "step": 8340 + }, + { + "epoch": 0.18360114461809376, + "grad_norm": 2.2065925680022973, + "learning_rate": 9.393962244884634e-06, + "loss": 0.3353, + "step": 8341 + }, + { + "epoch": 0.18362315650451244, + "grad_norm": 2.180140360539683, + "learning_rate": 9.393792131243711e-06, + "loss": 0.3219, + "step": 8342 + }, + { + "epoch": 0.1836451683909311, + "grad_norm": 0.6423367346143397, + "learning_rate": 9.393621995271575e-06, + "loss": 0.5533, + "step": 8343 + }, + { + "epoch": 0.18366718027734977, + "grad_norm": 1.583047800822285, + "learning_rate": 9.393451836969095e-06, + "loss": 0.2445, + "step": 8344 + }, + { + "epoch": 0.18368919216376844, + "grad_norm": 1.4372938273201563, + "learning_rate": 9.39328165633713e-06, + "loss": 0.3838, + "step": 8345 + }, + { + "epoch": 0.1837112040501871, + "grad_norm": 1.3670902511475163, + "learning_rate": 9.393111453376548e-06, + "loss": 0.4005, + "step": 8346 + }, + { + "epoch": 0.18373321593660577, + "grad_norm": 1.5749777042491273, + "learning_rate": 9.392941228088217e-06, + "loss": 0.355, + "step": 8347 + }, + { + "epoch": 0.18375522782302442, + "grad_norm": 1.5281702247222722, + "learning_rate": 9.392770980472999e-06, + "loss": 0.2453, + "step": 8348 + }, + { + "epoch": 0.1837772397094431, + "grad_norm": 1.455581657976121, + "learning_rate": 9.392600710531759e-06, + "loss": 0.4963, + "step": 8349 + }, + { + "epoch": 0.18379925159586177, + "grad_norm": 1.4482302463113745, + "learning_rate": 9.392430418265361e-06, + "loss": 0.3585, + "step": 8350 + }, + { + "epoch": 0.18382126348228042, + "grad_norm": 1.5906980397052042, + "learning_rate": 9.392260103674675e-06, + "loss": 0.331, + "step": 8351 + }, + { + "epoch": 0.1838432753686991, + "grad_norm": 1.7949250716877838, + "learning_rate": 9.392089766760564e-06, + "loss": 0.4171, + "step": 8352 + }, + { + "epoch": 0.18386528725511778, + "grad_norm": 2.5511112674754806, + "learning_rate": 9.391919407523893e-06, + "loss": 0.3414, + "step": 8353 + }, + { + "epoch": 0.18388729914153643, + "grad_norm": 1.445796235698342, + "learning_rate": 9.391749025965528e-06, + "loss": 0.4173, + "step": 8354 + }, + { + "epoch": 0.1839093110279551, + "grad_norm": 0.6934375933155582, + "learning_rate": 9.391578622086338e-06, + "loss": 0.5641, + "step": 8355 + }, + { + "epoch": 0.18393132291437375, + "grad_norm": 1.5976531274719847, + "learning_rate": 9.391408195887185e-06, + "loss": 0.4397, + "step": 8356 + }, + { + "epoch": 0.18395333480079243, + "grad_norm": 12.378486657949532, + "learning_rate": 9.391237747368937e-06, + "loss": 0.3348, + "step": 8357 + }, + { + "epoch": 0.1839753466872111, + "grad_norm": 1.995621156546692, + "learning_rate": 9.391067276532461e-06, + "loss": 0.4606, + "step": 8358 + }, + { + "epoch": 0.18399735857362975, + "grad_norm": 1.0619857604498044, + "learning_rate": 9.390896783378622e-06, + "loss": 0.2286, + "step": 8359 + }, + { + "epoch": 0.18401937046004843, + "grad_norm": 0.6407092135523137, + "learning_rate": 9.390726267908287e-06, + "loss": 0.5461, + "step": 8360 + }, + { + "epoch": 0.18404138234646708, + "grad_norm": 1.455708493421221, + "learning_rate": 9.390555730122324e-06, + "loss": 0.3559, + "step": 8361 + }, + { + "epoch": 0.18406339423288576, + "grad_norm": 1.2366897261409338, + "learning_rate": 9.3903851700216e-06, + "loss": 0.4078, + "step": 8362 + }, + { + "epoch": 0.18408540611930443, + "grad_norm": 1.648158847899077, + "learning_rate": 9.390214587606976e-06, + "loss": 0.3993, + "step": 8363 + }, + { + "epoch": 0.18410741800572308, + "grad_norm": 1.3852348124161034, + "learning_rate": 9.390043982879326e-06, + "loss": 0.3087, + "step": 8364 + }, + { + "epoch": 0.18412942989214176, + "grad_norm": 1.4206127314397516, + "learning_rate": 9.389873355839514e-06, + "loss": 0.4203, + "step": 8365 + }, + { + "epoch": 0.1841514417785604, + "grad_norm": 8.726510385648771, + "learning_rate": 9.389702706488407e-06, + "loss": 0.3699, + "step": 8366 + }, + { + "epoch": 0.1841734536649791, + "grad_norm": 2.1231387012077607, + "learning_rate": 9.389532034826874e-06, + "loss": 0.325, + "step": 8367 + }, + { + "epoch": 0.18419546555139776, + "grad_norm": 1.3262499990487824, + "learning_rate": 9.38936134085578e-06, + "loss": 0.3144, + "step": 8368 + }, + { + "epoch": 0.1842174774378164, + "grad_norm": 2.8059507860306714, + "learning_rate": 9.389190624575997e-06, + "loss": 0.3604, + "step": 8369 + }, + { + "epoch": 0.1842394893242351, + "grad_norm": 1.6529654642599962, + "learning_rate": 9.389019885988387e-06, + "loss": 0.3708, + "step": 8370 + }, + { + "epoch": 0.18426150121065377, + "grad_norm": 4.343118299539356, + "learning_rate": 9.38884912509382e-06, + "loss": 0.4267, + "step": 8371 + }, + { + "epoch": 0.18428351309707242, + "grad_norm": 1.6844586925078426, + "learning_rate": 9.388678341893165e-06, + "loss": 0.3845, + "step": 8372 + }, + { + "epoch": 0.1843055249834911, + "grad_norm": 1.4427598313225445, + "learning_rate": 9.38850753638729e-06, + "loss": 0.4422, + "step": 8373 + }, + { + "epoch": 0.18432753686990974, + "grad_norm": 1.5024906655000771, + "learning_rate": 9.388336708577063e-06, + "loss": 0.3308, + "step": 8374 + }, + { + "epoch": 0.18434954875632842, + "grad_norm": 1.770875433033196, + "learning_rate": 9.388165858463349e-06, + "loss": 0.3672, + "step": 8375 + }, + { + "epoch": 0.1843715606427471, + "grad_norm": 0.6666349742696489, + "learning_rate": 9.387994986047021e-06, + "loss": 0.5228, + "step": 8376 + }, + { + "epoch": 0.18439357252916574, + "grad_norm": 1.435701104239968, + "learning_rate": 9.387824091328945e-06, + "loss": 0.3556, + "step": 8377 + }, + { + "epoch": 0.18441558441558442, + "grad_norm": 1.4270494438817365, + "learning_rate": 9.38765317430999e-06, + "loss": 0.3237, + "step": 8378 + }, + { + "epoch": 0.18443759630200307, + "grad_norm": 0.6695409514950372, + "learning_rate": 9.387482234991024e-06, + "loss": 0.5454, + "step": 8379 + }, + { + "epoch": 0.18445960818842175, + "grad_norm": 1.3910387100174362, + "learning_rate": 9.387311273372917e-06, + "loss": 0.3753, + "step": 8380 + }, + { + "epoch": 0.18448162007484042, + "grad_norm": 1.3897030926949572, + "learning_rate": 9.387140289456538e-06, + "loss": 0.3876, + "step": 8381 + }, + { + "epoch": 0.18450363196125907, + "grad_norm": 1.5085063814971051, + "learning_rate": 9.386969283242754e-06, + "loss": 0.4258, + "step": 8382 + }, + { + "epoch": 0.18452564384767775, + "grad_norm": 1.187436435996445, + "learning_rate": 9.386798254732437e-06, + "loss": 0.2991, + "step": 8383 + }, + { + "epoch": 0.1845476557340964, + "grad_norm": 1.366512561170462, + "learning_rate": 9.386627203926453e-06, + "loss": 0.3399, + "step": 8384 + }, + { + "epoch": 0.18456966762051508, + "grad_norm": 0.7012609657148433, + "learning_rate": 9.386456130825672e-06, + "loss": 0.5233, + "step": 8385 + }, + { + "epoch": 0.18459167950693375, + "grad_norm": 1.485706704527933, + "learning_rate": 9.386285035430967e-06, + "loss": 0.372, + "step": 8386 + }, + { + "epoch": 0.1846136913933524, + "grad_norm": 1.8757736676382533, + "learning_rate": 9.386113917743206e-06, + "loss": 0.3278, + "step": 8387 + }, + { + "epoch": 0.18463570327977108, + "grad_norm": 1.6385304178086035, + "learning_rate": 9.385942777763254e-06, + "loss": 0.3744, + "step": 8388 + }, + { + "epoch": 0.18465771516618973, + "grad_norm": 1.7962860342674836, + "learning_rate": 9.385771615491988e-06, + "loss": 0.3587, + "step": 8389 + }, + { + "epoch": 0.1846797270526084, + "grad_norm": 1.3305847793361893, + "learning_rate": 9.385600430930272e-06, + "loss": 0.2936, + "step": 8390 + }, + { + "epoch": 0.18470173893902708, + "grad_norm": 0.6756471939857239, + "learning_rate": 9.38542922407898e-06, + "loss": 0.552, + "step": 8391 + }, + { + "epoch": 0.18472375082544573, + "grad_norm": 0.680427799693615, + "learning_rate": 9.385257994938981e-06, + "loss": 0.5362, + "step": 8392 + }, + { + "epoch": 0.1847457627118644, + "grad_norm": 3.0799110265306333, + "learning_rate": 9.385086743511145e-06, + "loss": 0.3461, + "step": 8393 + }, + { + "epoch": 0.18476777459828309, + "grad_norm": 1.3544132387663017, + "learning_rate": 9.384915469796341e-06, + "loss": 0.3563, + "step": 8394 + }, + { + "epoch": 0.18478978648470173, + "grad_norm": 0.6377164626789433, + "learning_rate": 9.384744173795442e-06, + "loss": 0.5123, + "step": 8395 + }, + { + "epoch": 0.1848117983711204, + "grad_norm": 0.6214331723643453, + "learning_rate": 9.384572855509318e-06, + "loss": 0.5204, + "step": 8396 + }, + { + "epoch": 0.18483381025753906, + "grad_norm": 1.5108591140716008, + "learning_rate": 9.384401514938837e-06, + "loss": 0.3733, + "step": 8397 + }, + { + "epoch": 0.18485582214395774, + "grad_norm": 2.0374354212311085, + "learning_rate": 9.384230152084875e-06, + "loss": 0.3655, + "step": 8398 + }, + { + "epoch": 0.18487783403037641, + "grad_norm": 1.850302483211712, + "learning_rate": 9.384058766948297e-06, + "loss": 0.4064, + "step": 8399 + }, + { + "epoch": 0.18489984591679506, + "grad_norm": 1.5307028554547555, + "learning_rate": 9.38388735952998e-06, + "loss": 0.4391, + "step": 8400 + }, + { + "epoch": 0.18492185780321374, + "grad_norm": 0.6493764380230915, + "learning_rate": 9.38371592983079e-06, + "loss": 0.5499, + "step": 8401 + }, + { + "epoch": 0.1849438696896324, + "grad_norm": 1.5304048957099874, + "learning_rate": 9.3835444778516e-06, + "loss": 0.3533, + "step": 8402 + }, + { + "epoch": 0.18496588157605107, + "grad_norm": 0.6445469642552241, + "learning_rate": 9.383373003593283e-06, + "loss": 0.5599, + "step": 8403 + }, + { + "epoch": 0.18498789346246974, + "grad_norm": 2.167020586091913, + "learning_rate": 9.383201507056709e-06, + "loss": 0.3382, + "step": 8404 + }, + { + "epoch": 0.1850099053488884, + "grad_norm": 1.241806624149709, + "learning_rate": 9.383029988242751e-06, + "loss": 0.3404, + "step": 8405 + }, + { + "epoch": 0.18503191723530707, + "grad_norm": 0.6515892910671777, + "learning_rate": 9.382858447152279e-06, + "loss": 0.547, + "step": 8406 + }, + { + "epoch": 0.18505392912172572, + "grad_norm": 1.559479377009075, + "learning_rate": 9.382686883786165e-06, + "loss": 0.4376, + "step": 8407 + }, + { + "epoch": 0.1850759410081444, + "grad_norm": 3.220196988522475, + "learning_rate": 9.382515298145282e-06, + "loss": 0.4525, + "step": 8408 + }, + { + "epoch": 0.18509795289456307, + "grad_norm": 1.9079970193488722, + "learning_rate": 9.382343690230502e-06, + "loss": 0.2725, + "step": 8409 + }, + { + "epoch": 0.18511996478098172, + "grad_norm": 1.4913457500603031, + "learning_rate": 9.382172060042696e-06, + "loss": 0.3656, + "step": 8410 + }, + { + "epoch": 0.1851419766674004, + "grad_norm": 1.5672775867838364, + "learning_rate": 9.382000407582738e-06, + "loss": 0.4287, + "step": 8411 + }, + { + "epoch": 0.18516398855381908, + "grad_norm": 1.3821690168714102, + "learning_rate": 9.381828732851499e-06, + "loss": 0.3957, + "step": 8412 + }, + { + "epoch": 0.18518600044023772, + "grad_norm": 1.3830458570472615, + "learning_rate": 9.381657035849852e-06, + "loss": 0.2847, + "step": 8413 + }, + { + "epoch": 0.1852080123266564, + "grad_norm": 5.178812786549735, + "learning_rate": 9.38148531657867e-06, + "loss": 0.4282, + "step": 8414 + }, + { + "epoch": 0.18523002421307505, + "grad_norm": 3.62764567918215, + "learning_rate": 9.381313575038825e-06, + "loss": 0.2074, + "step": 8415 + }, + { + "epoch": 0.18525203609949373, + "grad_norm": 1.46658238592099, + "learning_rate": 9.381141811231188e-06, + "loss": 0.4025, + "step": 8416 + }, + { + "epoch": 0.1852740479859124, + "grad_norm": 1.6349695334265326, + "learning_rate": 9.380970025156636e-06, + "loss": 0.3387, + "step": 8417 + }, + { + "epoch": 0.18529605987233105, + "grad_norm": 4.029935423870324, + "learning_rate": 9.380798216816043e-06, + "loss": 0.317, + "step": 8418 + }, + { + "epoch": 0.18531807175874973, + "grad_norm": 1.3286828453955206, + "learning_rate": 9.380626386210277e-06, + "loss": 0.3574, + "step": 8419 + }, + { + "epoch": 0.18534008364516838, + "grad_norm": 1.3063890305308141, + "learning_rate": 9.380454533340215e-06, + "loss": 0.3186, + "step": 8420 + }, + { + "epoch": 0.18536209553158706, + "grad_norm": 1.5604163146991064, + "learning_rate": 9.38028265820673e-06, + "loss": 0.4551, + "step": 8421 + }, + { + "epoch": 0.18538410741800573, + "grad_norm": 1.5322686100702656, + "learning_rate": 9.380110760810693e-06, + "loss": 0.4791, + "step": 8422 + }, + { + "epoch": 0.18540611930442438, + "grad_norm": 0.6840967208170357, + "learning_rate": 9.37993884115298e-06, + "loss": 0.5488, + "step": 8423 + }, + { + "epoch": 0.18542813119084306, + "grad_norm": 1.3107269937646477, + "learning_rate": 9.379766899234464e-06, + "loss": 0.2841, + "step": 8424 + }, + { + "epoch": 0.1854501430772617, + "grad_norm": 1.572977090386486, + "learning_rate": 9.37959493505602e-06, + "loss": 0.3383, + "step": 8425 + }, + { + "epoch": 0.18547215496368039, + "grad_norm": 1.2884053687845314, + "learning_rate": 9.37942294861852e-06, + "loss": 0.3417, + "step": 8426 + }, + { + "epoch": 0.18549416685009906, + "grad_norm": 1.494524519517373, + "learning_rate": 9.379250939922841e-06, + "loss": 0.3679, + "step": 8427 + }, + { + "epoch": 0.1855161787365177, + "grad_norm": 1.840750675672919, + "learning_rate": 9.379078908969854e-06, + "loss": 0.3914, + "step": 8428 + }, + { + "epoch": 0.1855381906229364, + "grad_norm": 1.861434495002563, + "learning_rate": 9.378906855760437e-06, + "loss": 0.2808, + "step": 8429 + }, + { + "epoch": 0.18556020250935507, + "grad_norm": 1.5251421910981842, + "learning_rate": 9.37873478029546e-06, + "loss": 0.293, + "step": 8430 + }, + { + "epoch": 0.18558221439577371, + "grad_norm": 2.312560288563436, + "learning_rate": 9.3785626825758e-06, + "loss": 0.4476, + "step": 8431 + }, + { + "epoch": 0.1856042262821924, + "grad_norm": 1.458008468759834, + "learning_rate": 9.378390562602334e-06, + "loss": 0.379, + "step": 8432 + }, + { + "epoch": 0.18562623816861104, + "grad_norm": 1.305032144324584, + "learning_rate": 9.378218420375932e-06, + "loss": 0.3979, + "step": 8433 + }, + { + "epoch": 0.18564825005502972, + "grad_norm": 1.4766623424831942, + "learning_rate": 9.37804625589747e-06, + "loss": 0.3963, + "step": 8434 + }, + { + "epoch": 0.1856702619414484, + "grad_norm": 0.680107690082704, + "learning_rate": 9.377874069167827e-06, + "loss": 0.5806, + "step": 8435 + }, + { + "epoch": 0.18569227382786704, + "grad_norm": 1.2563648391608246, + "learning_rate": 9.377701860187875e-06, + "loss": 0.2742, + "step": 8436 + }, + { + "epoch": 0.18571428571428572, + "grad_norm": 1.6943329799116162, + "learning_rate": 9.37752962895849e-06, + "loss": 0.2982, + "step": 8437 + }, + { + "epoch": 0.18573629760070437, + "grad_norm": 1.6094046156836244, + "learning_rate": 9.377357375480544e-06, + "loss": 0.3482, + "step": 8438 + }, + { + "epoch": 0.18575830948712305, + "grad_norm": 1.3564213768566695, + "learning_rate": 9.377185099754919e-06, + "loss": 0.3521, + "step": 8439 + }, + { + "epoch": 0.18578032137354172, + "grad_norm": 2.7340906828276923, + "learning_rate": 9.377012801782485e-06, + "loss": 0.3952, + "step": 8440 + }, + { + "epoch": 0.18580233325996037, + "grad_norm": 0.6511728193129123, + "learning_rate": 9.376840481564122e-06, + "loss": 0.5591, + "step": 8441 + }, + { + "epoch": 0.18582434514637905, + "grad_norm": 0.6493469978785007, + "learning_rate": 9.376668139100701e-06, + "loss": 0.5302, + "step": 8442 + }, + { + "epoch": 0.1858463570327977, + "grad_norm": 1.4775869705792894, + "learning_rate": 9.376495774393101e-06, + "loss": 0.3504, + "step": 8443 + }, + { + "epoch": 0.18586836891921638, + "grad_norm": 1.5503968874327307, + "learning_rate": 9.3763233874422e-06, + "loss": 0.3266, + "step": 8444 + }, + { + "epoch": 0.18589038080563505, + "grad_norm": 1.9201503764096592, + "learning_rate": 9.376150978248868e-06, + "loss": 0.4214, + "step": 8445 + }, + { + "epoch": 0.1859123926920537, + "grad_norm": 1.6187570398680853, + "learning_rate": 9.375978546813986e-06, + "loss": 0.3219, + "step": 8446 + }, + { + "epoch": 0.18593440457847238, + "grad_norm": 1.1891268581213437, + "learning_rate": 9.37580609313843e-06, + "loss": 0.3154, + "step": 8447 + }, + { + "epoch": 0.18595641646489103, + "grad_norm": 1.8127431652842796, + "learning_rate": 9.375633617223075e-06, + "loss": 0.3476, + "step": 8448 + }, + { + "epoch": 0.1859784283513097, + "grad_norm": 1.5099650300994587, + "learning_rate": 9.3754611190688e-06, + "loss": 0.3519, + "step": 8449 + }, + { + "epoch": 0.18600044023772838, + "grad_norm": 3.3317568454600286, + "learning_rate": 9.375288598676479e-06, + "loss": 0.2858, + "step": 8450 + }, + { + "epoch": 0.18602245212414703, + "grad_norm": 1.6592334202856622, + "learning_rate": 9.37511605604699e-06, + "loss": 0.4248, + "step": 8451 + }, + { + "epoch": 0.1860444640105657, + "grad_norm": 1.557844962139916, + "learning_rate": 9.37494349118121e-06, + "loss": 0.3628, + "step": 8452 + }, + { + "epoch": 0.18606647589698438, + "grad_norm": 0.8382927379492567, + "learning_rate": 9.374770904080015e-06, + "loss": 0.5308, + "step": 8453 + }, + { + "epoch": 0.18608848778340303, + "grad_norm": 1.888801082903721, + "learning_rate": 9.374598294744284e-06, + "loss": 0.3574, + "step": 8454 + }, + { + "epoch": 0.1861104996698217, + "grad_norm": 2.1775277043296617, + "learning_rate": 9.374425663174893e-06, + "loss": 0.3508, + "step": 8455 + }, + { + "epoch": 0.18613251155624036, + "grad_norm": 1.8789912187348075, + "learning_rate": 9.37425300937272e-06, + "loss": 0.2646, + "step": 8456 + }, + { + "epoch": 0.18615452344265904, + "grad_norm": 1.3735473727747758, + "learning_rate": 9.374080333338642e-06, + "loss": 0.2833, + "step": 8457 + }, + { + "epoch": 0.1861765353290777, + "grad_norm": 1.4149964364853835, + "learning_rate": 9.373907635073536e-06, + "loss": 0.324, + "step": 8458 + }, + { + "epoch": 0.18619854721549636, + "grad_norm": 1.5199213630107504, + "learning_rate": 9.373734914578281e-06, + "loss": 0.3443, + "step": 8459 + }, + { + "epoch": 0.18622055910191504, + "grad_norm": 1.2650106651474353, + "learning_rate": 9.373562171853756e-06, + "loss": 0.2947, + "step": 8460 + }, + { + "epoch": 0.1862425709883337, + "grad_norm": 1.2457380565579372, + "learning_rate": 9.373389406900835e-06, + "loss": 0.4053, + "step": 8461 + }, + { + "epoch": 0.18626458287475237, + "grad_norm": 1.356251525594943, + "learning_rate": 9.3732166197204e-06, + "loss": 0.3415, + "step": 8462 + }, + { + "epoch": 0.18628659476117104, + "grad_norm": 1.3816473678041574, + "learning_rate": 9.373043810313328e-06, + "loss": 0.3517, + "step": 8463 + }, + { + "epoch": 0.1863086066475897, + "grad_norm": 1.608304515717855, + "learning_rate": 9.372870978680498e-06, + "loss": 0.3245, + "step": 8464 + }, + { + "epoch": 0.18633061853400837, + "grad_norm": 1.4998360183204325, + "learning_rate": 9.372698124822784e-06, + "loss": 0.4336, + "step": 8465 + }, + { + "epoch": 0.18635263042042702, + "grad_norm": 1.5651100867558227, + "learning_rate": 9.37252524874107e-06, + "loss": 0.3596, + "step": 8466 + }, + { + "epoch": 0.1863746423068457, + "grad_norm": 1.7081390257004057, + "learning_rate": 9.372352350436232e-06, + "loss": 0.4241, + "step": 8467 + }, + { + "epoch": 0.18639665419326437, + "grad_norm": 1.3616100000054925, + "learning_rate": 9.372179429909149e-06, + "loss": 0.3014, + "step": 8468 + }, + { + "epoch": 0.18641866607968302, + "grad_norm": 0.8124576084985361, + "learning_rate": 9.3720064871607e-06, + "loss": 0.5323, + "step": 8469 + }, + { + "epoch": 0.1864406779661017, + "grad_norm": 1.3528335237539906, + "learning_rate": 9.371833522191764e-06, + "loss": 0.4253, + "step": 8470 + }, + { + "epoch": 0.18646268985252037, + "grad_norm": 1.429592308247378, + "learning_rate": 9.37166053500322e-06, + "loss": 0.3993, + "step": 8471 + }, + { + "epoch": 0.18648470173893902, + "grad_norm": 1.855256077355165, + "learning_rate": 9.371487525595947e-06, + "loss": 0.3677, + "step": 8472 + }, + { + "epoch": 0.1865067136253577, + "grad_norm": 0.6442645155500516, + "learning_rate": 9.371314493970827e-06, + "loss": 0.5484, + "step": 8473 + }, + { + "epoch": 0.18652872551177635, + "grad_norm": 0.6105925590722553, + "learning_rate": 9.371141440128734e-06, + "loss": 0.5247, + "step": 8474 + }, + { + "epoch": 0.18655073739819503, + "grad_norm": 1.240550716064032, + "learning_rate": 9.370968364070554e-06, + "loss": 0.307, + "step": 8475 + }, + { + "epoch": 0.1865727492846137, + "grad_norm": 1.3238436880626947, + "learning_rate": 9.370795265797162e-06, + "loss": 0.3519, + "step": 8476 + }, + { + "epoch": 0.18659476117103235, + "grad_norm": 0.6247126798386221, + "learning_rate": 9.370622145309437e-06, + "loss": 0.5285, + "step": 8477 + }, + { + "epoch": 0.18661677305745103, + "grad_norm": 0.6419184030461852, + "learning_rate": 9.370449002608263e-06, + "loss": 0.5344, + "step": 8478 + }, + { + "epoch": 0.18663878494386968, + "grad_norm": 1.3165468207410211, + "learning_rate": 9.370275837694516e-06, + "loss": 0.3235, + "step": 8479 + }, + { + "epoch": 0.18666079683028836, + "grad_norm": 1.8040687359748926, + "learning_rate": 9.37010265056908e-06, + "loss": 0.33, + "step": 8480 + }, + { + "epoch": 0.18668280871670703, + "grad_norm": 1.208794636230406, + "learning_rate": 9.369929441232833e-06, + "loss": 0.3227, + "step": 8481 + }, + { + "epoch": 0.18670482060312568, + "grad_norm": 1.26947206602366, + "learning_rate": 9.369756209686654e-06, + "loss": 0.3904, + "step": 8482 + }, + { + "epoch": 0.18672683248954436, + "grad_norm": 2.0129027602923153, + "learning_rate": 9.369582955931426e-06, + "loss": 0.3346, + "step": 8483 + }, + { + "epoch": 0.186748844375963, + "grad_norm": 0.6787653017542931, + "learning_rate": 9.36940967996803e-06, + "loss": 0.5531, + "step": 8484 + }, + { + "epoch": 0.18677085626238168, + "grad_norm": 1.9396452267686433, + "learning_rate": 9.369236381797343e-06, + "loss": 0.4405, + "step": 8485 + }, + { + "epoch": 0.18679286814880036, + "grad_norm": 1.2408888190782505, + "learning_rate": 9.369063061420248e-06, + "loss": 0.3682, + "step": 8486 + }, + { + "epoch": 0.186814880035219, + "grad_norm": 1.3978995088404338, + "learning_rate": 9.368889718837627e-06, + "loss": 0.2922, + "step": 8487 + }, + { + "epoch": 0.1868368919216377, + "grad_norm": 1.4295948774605205, + "learning_rate": 9.36871635405036e-06, + "loss": 0.3667, + "step": 8488 + }, + { + "epoch": 0.18685890380805634, + "grad_norm": 0.6778034954461996, + "learning_rate": 9.368542967059327e-06, + "loss": 0.5506, + "step": 8489 + }, + { + "epoch": 0.186880915694475, + "grad_norm": 1.5332678192648703, + "learning_rate": 9.368369557865411e-06, + "loss": 0.4002, + "step": 8490 + }, + { + "epoch": 0.1869029275808937, + "grad_norm": 1.7417731008509467, + "learning_rate": 9.36819612646949e-06, + "loss": 0.3682, + "step": 8491 + }, + { + "epoch": 0.18692493946731234, + "grad_norm": 1.3461967649434763, + "learning_rate": 9.368022672872452e-06, + "loss": 0.3434, + "step": 8492 + }, + { + "epoch": 0.18694695135373102, + "grad_norm": 2.052350674644209, + "learning_rate": 9.367849197075172e-06, + "loss": 0.4142, + "step": 8493 + }, + { + "epoch": 0.1869689632401497, + "grad_norm": 1.2287825564534407, + "learning_rate": 9.367675699078536e-06, + "loss": 0.3386, + "step": 8494 + }, + { + "epoch": 0.18699097512656834, + "grad_norm": 1.5007575160829045, + "learning_rate": 9.367502178883423e-06, + "loss": 0.4385, + "step": 8495 + }, + { + "epoch": 0.18701298701298702, + "grad_norm": 1.3345058646244499, + "learning_rate": 9.367328636490715e-06, + "loss": 0.3344, + "step": 8496 + }, + { + "epoch": 0.18703499889940567, + "grad_norm": 0.6703935738800977, + "learning_rate": 9.367155071901297e-06, + "loss": 0.5643, + "step": 8497 + }, + { + "epoch": 0.18705701078582435, + "grad_norm": 13.879000957964424, + "learning_rate": 9.366981485116049e-06, + "loss": 0.3712, + "step": 8498 + }, + { + "epoch": 0.18707902267224302, + "grad_norm": 0.6444908645450224, + "learning_rate": 9.366807876135851e-06, + "loss": 0.5297, + "step": 8499 + }, + { + "epoch": 0.18710103455866167, + "grad_norm": 1.4062176574579575, + "learning_rate": 9.36663424496159e-06, + "loss": 0.3777, + "step": 8500 + }, + { + "epoch": 0.18712304644508035, + "grad_norm": 1.376038032110213, + "learning_rate": 9.366460591594146e-06, + "loss": 0.3666, + "step": 8501 + }, + { + "epoch": 0.187145058331499, + "grad_norm": 1.679026714961997, + "learning_rate": 9.366286916034401e-06, + "loss": 0.3728, + "step": 8502 + }, + { + "epoch": 0.18716707021791767, + "grad_norm": 1.2936266202852964, + "learning_rate": 9.366113218283239e-06, + "loss": 0.355, + "step": 8503 + }, + { + "epoch": 0.18718908210433635, + "grad_norm": 0.6516987072925787, + "learning_rate": 9.365939498341541e-06, + "loss": 0.5369, + "step": 8504 + }, + { + "epoch": 0.187211093990755, + "grad_norm": 1.4356014539325639, + "learning_rate": 9.365765756210192e-06, + "loss": 0.3254, + "step": 8505 + }, + { + "epoch": 0.18723310587717368, + "grad_norm": 1.7636019669955083, + "learning_rate": 9.365591991890075e-06, + "loss": 0.3263, + "step": 8506 + }, + { + "epoch": 0.18725511776359233, + "grad_norm": 1.4217268654185935, + "learning_rate": 9.36541820538207e-06, + "loss": 0.3395, + "step": 8507 + }, + { + "epoch": 0.187277129650011, + "grad_norm": 1.828209746297241, + "learning_rate": 9.365244396687066e-06, + "loss": 0.3499, + "step": 8508 + }, + { + "epoch": 0.18729914153642968, + "grad_norm": 8.415965291126408, + "learning_rate": 9.365070565805941e-06, + "loss": 0.3092, + "step": 8509 + }, + { + "epoch": 0.18732115342284833, + "grad_norm": 0.6655026581306575, + "learning_rate": 9.364896712739582e-06, + "loss": 0.5624, + "step": 8510 + }, + { + "epoch": 0.187343165309267, + "grad_norm": 1.639739671939411, + "learning_rate": 9.36472283748887e-06, + "loss": 0.3724, + "step": 8511 + }, + { + "epoch": 0.18736517719568568, + "grad_norm": 1.0728793273041168, + "learning_rate": 9.36454894005469e-06, + "loss": 0.3301, + "step": 8512 + }, + { + "epoch": 0.18738718908210433, + "grad_norm": 1.677552120300024, + "learning_rate": 9.364375020437925e-06, + "loss": 0.2633, + "step": 8513 + }, + { + "epoch": 0.187409200968523, + "grad_norm": 2.5422030062054013, + "learning_rate": 9.36420107863946e-06, + "loss": 0.4225, + "step": 8514 + }, + { + "epoch": 0.18743121285494166, + "grad_norm": 1.5631238119231803, + "learning_rate": 9.36402711466018e-06, + "loss": 0.3204, + "step": 8515 + }, + { + "epoch": 0.18745322474136034, + "grad_norm": 1.431867105571597, + "learning_rate": 9.363853128500966e-06, + "loss": 0.2483, + "step": 8516 + }, + { + "epoch": 0.187475236627779, + "grad_norm": 1.4763644063950518, + "learning_rate": 9.363679120162703e-06, + "loss": 0.4384, + "step": 8517 + }, + { + "epoch": 0.18749724851419766, + "grad_norm": 3.227984442169558, + "learning_rate": 9.363505089646279e-06, + "loss": 0.3902, + "step": 8518 + }, + { + "epoch": 0.18751926040061634, + "grad_norm": 1.3436718289216807, + "learning_rate": 9.363331036952575e-06, + "loss": 0.3758, + "step": 8519 + }, + { + "epoch": 0.187541272287035, + "grad_norm": 1.387377217114598, + "learning_rate": 9.363156962082478e-06, + "loss": 0.4481, + "step": 8520 + }, + { + "epoch": 0.18756328417345366, + "grad_norm": 1.2984899202354314, + "learning_rate": 9.362982865036869e-06, + "loss": 0.3089, + "step": 8521 + }, + { + "epoch": 0.18758529605987234, + "grad_norm": 1.243567406720828, + "learning_rate": 9.362808745816637e-06, + "loss": 0.354, + "step": 8522 + }, + { + "epoch": 0.187607307946291, + "grad_norm": 1.6913747990230414, + "learning_rate": 9.362634604422662e-06, + "loss": 0.3678, + "step": 8523 + }, + { + "epoch": 0.18762931983270967, + "grad_norm": 1.2651727748955008, + "learning_rate": 9.362460440855836e-06, + "loss": 0.3329, + "step": 8524 + }, + { + "epoch": 0.18765133171912832, + "grad_norm": 0.6622946796156678, + "learning_rate": 9.362286255117039e-06, + "loss": 0.5251, + "step": 8525 + }, + { + "epoch": 0.187673343605547, + "grad_norm": 0.6462679140052593, + "learning_rate": 9.362112047207156e-06, + "loss": 0.5105, + "step": 8526 + }, + { + "epoch": 0.18769535549196567, + "grad_norm": 1.8657711514474535, + "learning_rate": 9.361937817127076e-06, + "loss": 0.4148, + "step": 8527 + }, + { + "epoch": 0.18771736737838432, + "grad_norm": 1.7296088039975541, + "learning_rate": 9.361763564877681e-06, + "loss": 0.2642, + "step": 8528 + }, + { + "epoch": 0.187739379264803, + "grad_norm": 1.6506151637718571, + "learning_rate": 9.36158929045986e-06, + "loss": 0.3573, + "step": 8529 + }, + { + "epoch": 0.18776139115122165, + "grad_norm": 1.5375403945471435, + "learning_rate": 9.361414993874496e-06, + "loss": 0.302, + "step": 8530 + }, + { + "epoch": 0.18778340303764032, + "grad_norm": 3.6666245687110015, + "learning_rate": 9.361240675122475e-06, + "loss": 0.3886, + "step": 8531 + }, + { + "epoch": 0.187805414924059, + "grad_norm": 4.225716599080679, + "learning_rate": 9.361066334204686e-06, + "loss": 0.3539, + "step": 8532 + }, + { + "epoch": 0.18782742681047765, + "grad_norm": 1.4739578851432906, + "learning_rate": 9.36089197112201e-06, + "loss": 0.2597, + "step": 8533 + }, + { + "epoch": 0.18784943869689633, + "grad_norm": 1.3474605108890225, + "learning_rate": 9.360717585875337e-06, + "loss": 0.4054, + "step": 8534 + }, + { + "epoch": 0.187871450583315, + "grad_norm": 3.915980541660762, + "learning_rate": 9.360543178465551e-06, + "loss": 0.386, + "step": 8535 + }, + { + "epoch": 0.18789346246973365, + "grad_norm": 1.78673609900067, + "learning_rate": 9.360368748893542e-06, + "loss": 0.3934, + "step": 8536 + }, + { + "epoch": 0.18791547435615233, + "grad_norm": 2.0397662445802873, + "learning_rate": 9.360194297160194e-06, + "loss": 0.4001, + "step": 8537 + }, + { + "epoch": 0.18793748624257098, + "grad_norm": 0.7063341766144181, + "learning_rate": 9.360019823266392e-06, + "loss": 0.5248, + "step": 8538 + }, + { + "epoch": 0.18795949812898965, + "grad_norm": 1.3320382079151072, + "learning_rate": 9.359845327213027e-06, + "loss": 0.3617, + "step": 8539 + }, + { + "epoch": 0.18798151001540833, + "grad_norm": 0.6660380907145385, + "learning_rate": 9.359670809000982e-06, + "loss": 0.5191, + "step": 8540 + }, + { + "epoch": 0.18800352190182698, + "grad_norm": 1.5083956615843963, + "learning_rate": 9.359496268631146e-06, + "loss": 0.3335, + "step": 8541 + }, + { + "epoch": 0.18802553378824566, + "grad_norm": 1.664139736274087, + "learning_rate": 9.359321706104405e-06, + "loss": 0.3755, + "step": 8542 + }, + { + "epoch": 0.1880475456746643, + "grad_norm": 2.293911528656337, + "learning_rate": 9.359147121421648e-06, + "loss": 0.2766, + "step": 8543 + }, + { + "epoch": 0.18806955756108298, + "grad_norm": 1.8539609604934935, + "learning_rate": 9.358972514583759e-06, + "loss": 0.3854, + "step": 8544 + }, + { + "epoch": 0.18809156944750166, + "grad_norm": 1.3607130017447053, + "learning_rate": 9.35879788559163e-06, + "loss": 0.359, + "step": 8545 + }, + { + "epoch": 0.1881135813339203, + "grad_norm": 1.5027914530394024, + "learning_rate": 9.358623234446145e-06, + "loss": 0.39, + "step": 8546 + }, + { + "epoch": 0.188135593220339, + "grad_norm": 1.7306442233800725, + "learning_rate": 9.358448561148192e-06, + "loss": 0.3418, + "step": 8547 + }, + { + "epoch": 0.18815760510675764, + "grad_norm": 1.6663790507185492, + "learning_rate": 9.358273865698661e-06, + "loss": 0.272, + "step": 8548 + }, + { + "epoch": 0.1881796169931763, + "grad_norm": 1.5258788981187295, + "learning_rate": 9.358099148098436e-06, + "loss": 0.3527, + "step": 8549 + }, + { + "epoch": 0.188201628879595, + "grad_norm": 1.5735824139065844, + "learning_rate": 9.357924408348409e-06, + "loss": 0.3568, + "step": 8550 + }, + { + "epoch": 0.18822364076601364, + "grad_norm": 0.8323796740491868, + "learning_rate": 9.357749646449468e-06, + "loss": 0.5191, + "step": 8551 + }, + { + "epoch": 0.18824565265243232, + "grad_norm": 0.7599658750597269, + "learning_rate": 9.357574862402496e-06, + "loss": 0.5486, + "step": 8552 + }, + { + "epoch": 0.188267664538851, + "grad_norm": 1.8717094942644292, + "learning_rate": 9.357400056208388e-06, + "loss": 0.3661, + "step": 8553 + }, + { + "epoch": 0.18828967642526964, + "grad_norm": 1.3396940198493776, + "learning_rate": 9.357225227868028e-06, + "loss": 0.3183, + "step": 8554 + }, + { + "epoch": 0.18831168831168832, + "grad_norm": 1.4960359621234367, + "learning_rate": 9.357050377382306e-06, + "loss": 0.3491, + "step": 8555 + }, + { + "epoch": 0.18833370019810697, + "grad_norm": 1.6973095203768707, + "learning_rate": 9.356875504752112e-06, + "loss": 0.3692, + "step": 8556 + }, + { + "epoch": 0.18835571208452564, + "grad_norm": 12.106213535212275, + "learning_rate": 9.356700609978334e-06, + "loss": 0.3911, + "step": 8557 + }, + { + "epoch": 0.18837772397094432, + "grad_norm": 1.531190907588552, + "learning_rate": 9.356525693061857e-06, + "loss": 0.4247, + "step": 8558 + }, + { + "epoch": 0.18839973585736297, + "grad_norm": 1.6143119533918255, + "learning_rate": 9.356350754003575e-06, + "loss": 0.3951, + "step": 8559 + }, + { + "epoch": 0.18842174774378165, + "grad_norm": 1.5924273963132343, + "learning_rate": 9.356175792804377e-06, + "loss": 0.4168, + "step": 8560 + }, + { + "epoch": 0.1884437596302003, + "grad_norm": 1.5754304524772564, + "learning_rate": 9.356000809465149e-06, + "loss": 0.3257, + "step": 8561 + }, + { + "epoch": 0.18846577151661897, + "grad_norm": 1.4137023066517094, + "learning_rate": 9.355825803986783e-06, + "loss": 0.3349, + "step": 8562 + }, + { + "epoch": 0.18848778340303765, + "grad_norm": 1.1240608002974735, + "learning_rate": 9.355650776370168e-06, + "loss": 0.5622, + "step": 8563 + }, + { + "epoch": 0.1885097952894563, + "grad_norm": 1.4201613310101693, + "learning_rate": 9.355475726616193e-06, + "loss": 0.3041, + "step": 8564 + }, + { + "epoch": 0.18853180717587498, + "grad_norm": 1.6215971917268208, + "learning_rate": 9.355300654725745e-06, + "loss": 0.4651, + "step": 8565 + }, + { + "epoch": 0.18855381906229363, + "grad_norm": 0.7944738248265156, + "learning_rate": 9.355125560699719e-06, + "loss": 0.5592, + "step": 8566 + }, + { + "epoch": 0.1885758309487123, + "grad_norm": 1.7171283343379173, + "learning_rate": 9.354950444539002e-06, + "loss": 0.329, + "step": 8567 + }, + { + "epoch": 0.18859784283513098, + "grad_norm": 2.03228507835585, + "learning_rate": 9.354775306244485e-06, + "loss": 0.3811, + "step": 8568 + }, + { + "epoch": 0.18861985472154963, + "grad_norm": 1.2739535924815593, + "learning_rate": 9.354600145817056e-06, + "loss": 0.3047, + "step": 8569 + }, + { + "epoch": 0.1886418666079683, + "grad_norm": 1.8891924552459525, + "learning_rate": 9.35442496325761e-06, + "loss": 0.4003, + "step": 8570 + }, + { + "epoch": 0.18866387849438696, + "grad_norm": 1.4433826423927276, + "learning_rate": 9.35424975856703e-06, + "loss": 0.3508, + "step": 8571 + }, + { + "epoch": 0.18868589038080563, + "grad_norm": 0.8596433187387881, + "learning_rate": 9.354074531746213e-06, + "loss": 0.5412, + "step": 8572 + }, + { + "epoch": 0.1887079022672243, + "grad_norm": 1.3397398190976777, + "learning_rate": 9.353899282796047e-06, + "loss": 0.3919, + "step": 8573 + }, + { + "epoch": 0.18872991415364296, + "grad_norm": 1.4164231716867193, + "learning_rate": 9.353724011717424e-06, + "loss": 0.4054, + "step": 8574 + }, + { + "epoch": 0.18875192604006163, + "grad_norm": 2.6139046830836365, + "learning_rate": 9.353548718511231e-06, + "loss": 0.4261, + "step": 8575 + }, + { + "epoch": 0.1887739379264803, + "grad_norm": 1.6225707756598835, + "learning_rate": 9.353373403178366e-06, + "loss": 0.4755, + "step": 8576 + }, + { + "epoch": 0.18879594981289896, + "grad_norm": 1.9318182524963847, + "learning_rate": 9.353198065719713e-06, + "loss": 0.3454, + "step": 8577 + }, + { + "epoch": 0.18881796169931764, + "grad_norm": 1.3564838042375607, + "learning_rate": 9.353022706136166e-06, + "loss": 0.4117, + "step": 8578 + }, + { + "epoch": 0.1888399735857363, + "grad_norm": 1.9736714419645416, + "learning_rate": 9.352847324428616e-06, + "loss": 0.2702, + "step": 8579 + }, + { + "epoch": 0.18886198547215496, + "grad_norm": 1.5504633828424788, + "learning_rate": 9.352671920597954e-06, + "loss": 0.4199, + "step": 8580 + }, + { + "epoch": 0.18888399735857364, + "grad_norm": 1.8049732145858544, + "learning_rate": 9.352496494645072e-06, + "loss": 0.31, + "step": 8581 + }, + { + "epoch": 0.1889060092449923, + "grad_norm": 1.4361334392599696, + "learning_rate": 9.352321046570862e-06, + "loss": 0.3238, + "step": 8582 + }, + { + "epoch": 0.18892802113141097, + "grad_norm": 1.461479381206366, + "learning_rate": 9.352145576376216e-06, + "loss": 0.2938, + "step": 8583 + }, + { + "epoch": 0.18895003301782962, + "grad_norm": 1.3197623228298032, + "learning_rate": 9.351970084062024e-06, + "loss": 0.2355, + "step": 8584 + }, + { + "epoch": 0.1889720449042483, + "grad_norm": 2.632555102886421, + "learning_rate": 9.35179456962918e-06, + "loss": 0.3549, + "step": 8585 + }, + { + "epoch": 0.18899405679066697, + "grad_norm": 1.6340565534225275, + "learning_rate": 9.351619033078574e-06, + "loss": 0.3639, + "step": 8586 + }, + { + "epoch": 0.18901606867708562, + "grad_norm": 1.2404466769370264, + "learning_rate": 9.3514434744111e-06, + "loss": 0.3093, + "step": 8587 + }, + { + "epoch": 0.1890380805635043, + "grad_norm": 0.7694237706706706, + "learning_rate": 9.35126789362765e-06, + "loss": 0.5475, + "step": 8588 + }, + { + "epoch": 0.18906009244992295, + "grad_norm": 0.7007779962610159, + "learning_rate": 9.351092290729113e-06, + "loss": 0.5143, + "step": 8589 + }, + { + "epoch": 0.18908210433634162, + "grad_norm": 1.2514517482420335, + "learning_rate": 9.350916665716388e-06, + "loss": 0.3443, + "step": 8590 + }, + { + "epoch": 0.1891041162227603, + "grad_norm": 2.2334288552385155, + "learning_rate": 9.350741018590361e-06, + "loss": 0.3407, + "step": 8591 + }, + { + "epoch": 0.18912612810917895, + "grad_norm": 0.6436953284564929, + "learning_rate": 9.35056534935193e-06, + "loss": 0.5737, + "step": 8592 + }, + { + "epoch": 0.18914813999559763, + "grad_norm": 2.511060215527502, + "learning_rate": 9.350389658001984e-06, + "loss": 0.329, + "step": 8593 + }, + { + "epoch": 0.1891701518820163, + "grad_norm": 2.1314367548188433, + "learning_rate": 9.350213944541418e-06, + "loss": 0.3578, + "step": 8594 + }, + { + "epoch": 0.18919216376843495, + "grad_norm": 1.5789365722127002, + "learning_rate": 9.350038208971123e-06, + "loss": 0.3194, + "step": 8595 + }, + { + "epoch": 0.18921417565485363, + "grad_norm": 2.2853445056657513, + "learning_rate": 9.349862451291996e-06, + "loss": 0.4002, + "step": 8596 + }, + { + "epoch": 0.18923618754127228, + "grad_norm": 1.3581160540134445, + "learning_rate": 9.349686671504927e-06, + "loss": 0.3098, + "step": 8597 + }, + { + "epoch": 0.18925819942769095, + "grad_norm": 3.787905403294578, + "learning_rate": 9.34951086961081e-06, + "loss": 0.2761, + "step": 8598 + }, + { + "epoch": 0.18928021131410963, + "grad_norm": 0.777931467340491, + "learning_rate": 9.349335045610538e-06, + "loss": 0.57, + "step": 8599 + }, + { + "epoch": 0.18930222320052828, + "grad_norm": 2.955409174040946, + "learning_rate": 9.349159199505007e-06, + "loss": 0.2934, + "step": 8600 + }, + { + "epoch": 0.18932423508694696, + "grad_norm": 1.4914050198634736, + "learning_rate": 9.34898333129511e-06, + "loss": 0.4001, + "step": 8601 + }, + { + "epoch": 0.1893462469733656, + "grad_norm": 1.7070301242282635, + "learning_rate": 9.348807440981738e-06, + "loss": 0.3103, + "step": 8602 + }, + { + "epoch": 0.18936825885978428, + "grad_norm": 1.5137776604592637, + "learning_rate": 9.348631528565786e-06, + "loss": 0.3779, + "step": 8603 + }, + { + "epoch": 0.18939027074620296, + "grad_norm": 1.455474657233538, + "learning_rate": 9.348455594048152e-06, + "loss": 0.2635, + "step": 8604 + }, + { + "epoch": 0.1894122826326216, + "grad_norm": 1.2498036092341682, + "learning_rate": 9.348279637429726e-06, + "loss": 0.2985, + "step": 8605 + }, + { + "epoch": 0.18943429451904029, + "grad_norm": 1.4264793780587883, + "learning_rate": 9.348103658711403e-06, + "loss": 0.3101, + "step": 8606 + }, + { + "epoch": 0.18945630640545894, + "grad_norm": 1.4209070276233395, + "learning_rate": 9.347927657894078e-06, + "loss": 0.3343, + "step": 8607 + }, + { + "epoch": 0.1894783182918776, + "grad_norm": 1.5507104710788944, + "learning_rate": 9.347751634978645e-06, + "loss": 0.3776, + "step": 8608 + }, + { + "epoch": 0.1895003301782963, + "grad_norm": 1.4253013834033825, + "learning_rate": 9.347575589966e-06, + "loss": 0.3188, + "step": 8609 + }, + { + "epoch": 0.18952234206471494, + "grad_norm": 1.6144435488654967, + "learning_rate": 9.347399522857037e-06, + "loss": 0.3095, + "step": 8610 + }, + { + "epoch": 0.18954435395113362, + "grad_norm": 6.802215261198756, + "learning_rate": 9.34722343365265e-06, + "loss": 0.3459, + "step": 8611 + }, + { + "epoch": 0.1895663658375523, + "grad_norm": 1.5690243746869055, + "learning_rate": 9.347047322353735e-06, + "loss": 0.387, + "step": 8612 + }, + { + "epoch": 0.18958837772397094, + "grad_norm": 1.7896979276673897, + "learning_rate": 9.346871188961186e-06, + "loss": 0.3368, + "step": 8613 + }, + { + "epoch": 0.18961038961038962, + "grad_norm": 1.1898362737651267, + "learning_rate": 9.346695033475899e-06, + "loss": 0.3804, + "step": 8614 + }, + { + "epoch": 0.18963240149680827, + "grad_norm": 1.6868271992614858, + "learning_rate": 9.34651885589877e-06, + "loss": 0.2565, + "step": 8615 + }, + { + "epoch": 0.18965441338322694, + "grad_norm": 1.3733931602593892, + "learning_rate": 9.346342656230692e-06, + "loss": 0.3913, + "step": 8616 + }, + { + "epoch": 0.18967642526964562, + "grad_norm": 3.2367097343548354, + "learning_rate": 9.346166434472563e-06, + "loss": 0.3307, + "step": 8617 + }, + { + "epoch": 0.18969843715606427, + "grad_norm": 1.4514988918656735, + "learning_rate": 9.345990190625277e-06, + "loss": 0.3691, + "step": 8618 + }, + { + "epoch": 0.18972044904248295, + "grad_norm": 0.7462670884569519, + "learning_rate": 9.345813924689733e-06, + "loss": 0.554, + "step": 8619 + }, + { + "epoch": 0.1897424609289016, + "grad_norm": 1.599019651790855, + "learning_rate": 9.345637636666823e-06, + "loss": 0.3886, + "step": 8620 + }, + { + "epoch": 0.18976447281532027, + "grad_norm": 1.5869868049873483, + "learning_rate": 9.345461326557443e-06, + "loss": 0.3837, + "step": 8621 + }, + { + "epoch": 0.18978648470173895, + "grad_norm": 1.495962596468387, + "learning_rate": 9.345284994362492e-06, + "loss": 0.3694, + "step": 8622 + }, + { + "epoch": 0.1898084965881576, + "grad_norm": 1.4310795341553368, + "learning_rate": 9.345108640082862e-06, + "loss": 0.4056, + "step": 8623 + }, + { + "epoch": 0.18983050847457628, + "grad_norm": 1.9263481415344244, + "learning_rate": 9.344932263719454e-06, + "loss": 0.2733, + "step": 8624 + }, + { + "epoch": 0.18985252036099493, + "grad_norm": 0.6452832654502235, + "learning_rate": 9.344755865273162e-06, + "loss": 0.5614, + "step": 8625 + }, + { + "epoch": 0.1898745322474136, + "grad_norm": 3.22110060735182, + "learning_rate": 9.344579444744884e-06, + "loss": 0.3316, + "step": 8626 + }, + { + "epoch": 0.18989654413383228, + "grad_norm": 0.6429378498944962, + "learning_rate": 9.344403002135515e-06, + "loss": 0.5743, + "step": 8627 + }, + { + "epoch": 0.18991855602025093, + "grad_norm": 1.3156833340338192, + "learning_rate": 9.344226537445953e-06, + "loss": 0.3554, + "step": 8628 + }, + { + "epoch": 0.1899405679066696, + "grad_norm": 0.6218586151816633, + "learning_rate": 9.344050050677093e-06, + "loss": 0.545, + "step": 8629 + }, + { + "epoch": 0.18996257979308825, + "grad_norm": 1.1805676795789692, + "learning_rate": 9.343873541829834e-06, + "loss": 0.2929, + "step": 8630 + }, + { + "epoch": 0.18998459167950693, + "grad_norm": 1.902721018861412, + "learning_rate": 9.34369701090507e-06, + "loss": 0.3519, + "step": 8631 + }, + { + "epoch": 0.1900066035659256, + "grad_norm": 1.7351488706521487, + "learning_rate": 9.343520457903702e-06, + "loss": 0.3762, + "step": 8632 + }, + { + "epoch": 0.19002861545234426, + "grad_norm": 0.6234671298766632, + "learning_rate": 9.343343882826625e-06, + "loss": 0.5278, + "step": 8633 + }, + { + "epoch": 0.19005062733876293, + "grad_norm": 1.4910701698005477, + "learning_rate": 9.343167285674738e-06, + "loss": 0.5017, + "step": 8634 + }, + { + "epoch": 0.1900726392251816, + "grad_norm": 1.89704519854822, + "learning_rate": 9.342990666448938e-06, + "loss": 0.4588, + "step": 8635 + }, + { + "epoch": 0.19009465111160026, + "grad_norm": 3.407065234652475, + "learning_rate": 9.342814025150122e-06, + "loss": 0.4842, + "step": 8636 + }, + { + "epoch": 0.19011666299801894, + "grad_norm": 1.418589717473873, + "learning_rate": 9.342637361779187e-06, + "loss": 0.3055, + "step": 8637 + }, + { + "epoch": 0.1901386748844376, + "grad_norm": 2.1737208557194183, + "learning_rate": 9.342460676337034e-06, + "loss": 0.4138, + "step": 8638 + }, + { + "epoch": 0.19016068677085626, + "grad_norm": 1.3706990185824146, + "learning_rate": 9.34228396882456e-06, + "loss": 0.3329, + "step": 8639 + }, + { + "epoch": 0.19018269865727494, + "grad_norm": 1.425086114251673, + "learning_rate": 9.34210723924266e-06, + "loss": 0.321, + "step": 8640 + }, + { + "epoch": 0.1902047105436936, + "grad_norm": 1.4626886574829645, + "learning_rate": 9.341930487592234e-06, + "loss": 0.3666, + "step": 8641 + }, + { + "epoch": 0.19022672243011227, + "grad_norm": 1.1714986548255797, + "learning_rate": 9.341753713874181e-06, + "loss": 0.341, + "step": 8642 + }, + { + "epoch": 0.19024873431653092, + "grad_norm": 1.4422030153126406, + "learning_rate": 9.3415769180894e-06, + "loss": 0.3575, + "step": 8643 + }, + { + "epoch": 0.1902707462029496, + "grad_norm": 1.5852745029361122, + "learning_rate": 9.341400100238789e-06, + "loss": 0.3576, + "step": 8644 + }, + { + "epoch": 0.19029275808936827, + "grad_norm": 1.3713278767977988, + "learning_rate": 9.341223260323246e-06, + "loss": 0.3802, + "step": 8645 + }, + { + "epoch": 0.19031476997578692, + "grad_norm": 9.057313825308775, + "learning_rate": 9.34104639834367e-06, + "loss": 0.4501, + "step": 8646 + }, + { + "epoch": 0.1903367818622056, + "grad_norm": 1.9108832935428617, + "learning_rate": 9.34086951430096e-06, + "loss": 0.333, + "step": 8647 + }, + { + "epoch": 0.19035879374862424, + "grad_norm": 1.9864334185254395, + "learning_rate": 9.340692608196014e-06, + "loss": 0.2724, + "step": 8648 + }, + { + "epoch": 0.19038080563504292, + "grad_norm": 0.6979506457302885, + "learning_rate": 9.340515680029733e-06, + "loss": 0.5617, + "step": 8649 + }, + { + "epoch": 0.1904028175214616, + "grad_norm": 1.7062003723695391, + "learning_rate": 9.340338729803016e-06, + "loss": 0.3852, + "step": 8650 + }, + { + "epoch": 0.19042482940788025, + "grad_norm": 1.761500279667701, + "learning_rate": 9.340161757516761e-06, + "loss": 0.3793, + "step": 8651 + }, + { + "epoch": 0.19044684129429892, + "grad_norm": 0.6516819073199173, + "learning_rate": 9.33998476317187e-06, + "loss": 0.5488, + "step": 8652 + }, + { + "epoch": 0.1904688531807176, + "grad_norm": 1.9864653257397662, + "learning_rate": 9.339807746769238e-06, + "loss": 0.2706, + "step": 8653 + }, + { + "epoch": 0.19049086506713625, + "grad_norm": 2.2212518848600715, + "learning_rate": 9.339630708309769e-06, + "loss": 0.3468, + "step": 8654 + }, + { + "epoch": 0.19051287695355493, + "grad_norm": 1.3306952834325476, + "learning_rate": 9.339453647794361e-06, + "loss": 0.3085, + "step": 8655 + }, + { + "epoch": 0.19053488883997358, + "grad_norm": 2.137885126168228, + "learning_rate": 9.339276565223913e-06, + "loss": 0.3264, + "step": 8656 + }, + { + "epoch": 0.19055690072639225, + "grad_norm": 1.5284979223044293, + "learning_rate": 9.339099460599328e-06, + "loss": 0.3987, + "step": 8657 + }, + { + "epoch": 0.19057891261281093, + "grad_norm": 0.688700271889409, + "learning_rate": 9.338922333921502e-06, + "loss": 0.5729, + "step": 8658 + }, + { + "epoch": 0.19060092449922958, + "grad_norm": 1.6221157759960638, + "learning_rate": 9.33874518519134e-06, + "loss": 0.3516, + "step": 8659 + }, + { + "epoch": 0.19062293638564826, + "grad_norm": 1.6787861552196626, + "learning_rate": 9.33856801440974e-06, + "loss": 0.3588, + "step": 8660 + }, + { + "epoch": 0.1906449482720669, + "grad_norm": 1.563043774590231, + "learning_rate": 9.3383908215776e-06, + "loss": 0.3881, + "step": 8661 + }, + { + "epoch": 0.19066696015848558, + "grad_norm": 1.526713246914618, + "learning_rate": 9.338213606695823e-06, + "loss": 0.3196, + "step": 8662 + }, + { + "epoch": 0.19068897204490426, + "grad_norm": 1.237975115598497, + "learning_rate": 9.33803636976531e-06, + "loss": 0.2782, + "step": 8663 + }, + { + "epoch": 0.1907109839313229, + "grad_norm": 0.6358604044547609, + "learning_rate": 9.337859110786961e-06, + "loss": 0.5364, + "step": 8664 + }, + { + "epoch": 0.19073299581774159, + "grad_norm": 1.7597968111676128, + "learning_rate": 9.33768182976168e-06, + "loss": 0.4202, + "step": 8665 + }, + { + "epoch": 0.19075500770416023, + "grad_norm": 1.4628974959684824, + "learning_rate": 9.337504526690361e-06, + "loss": 0.3737, + "step": 8666 + }, + { + "epoch": 0.1907770195905789, + "grad_norm": 1.4548514484247705, + "learning_rate": 9.337327201573914e-06, + "loss": 0.3442, + "step": 8667 + }, + { + "epoch": 0.1907990314769976, + "grad_norm": 1.4326130672522532, + "learning_rate": 9.337149854413232e-06, + "loss": 0.4156, + "step": 8668 + }, + { + "epoch": 0.19082104336341624, + "grad_norm": 1.6781220197528597, + "learning_rate": 9.336972485209221e-06, + "loss": 0.372, + "step": 8669 + }, + { + "epoch": 0.19084305524983491, + "grad_norm": 1.666968054773562, + "learning_rate": 9.336795093962783e-06, + "loss": 0.3425, + "step": 8670 + }, + { + "epoch": 0.19086506713625356, + "grad_norm": 6.336423640385115, + "learning_rate": 9.336617680674817e-06, + "loss": 0.2905, + "step": 8671 + }, + { + "epoch": 0.19088707902267224, + "grad_norm": 1.6532297947915102, + "learning_rate": 9.336440245346226e-06, + "loss": 0.3633, + "step": 8672 + }, + { + "epoch": 0.19090909090909092, + "grad_norm": 1.1136404109029978, + "learning_rate": 9.33626278797791e-06, + "loss": 0.2837, + "step": 8673 + }, + { + "epoch": 0.19093110279550957, + "grad_norm": 1.3781729515431405, + "learning_rate": 9.336085308570775e-06, + "loss": 0.253, + "step": 8674 + }, + { + "epoch": 0.19095311468192824, + "grad_norm": 1.7463591411641382, + "learning_rate": 9.335907807125719e-06, + "loss": 0.3254, + "step": 8675 + }, + { + "epoch": 0.19097512656834692, + "grad_norm": 1.7772057753283026, + "learning_rate": 9.335730283643644e-06, + "loss": 0.3302, + "step": 8676 + }, + { + "epoch": 0.19099713845476557, + "grad_norm": 1.4988092599599192, + "learning_rate": 9.335552738125457e-06, + "loss": 0.2934, + "step": 8677 + }, + { + "epoch": 0.19101915034118425, + "grad_norm": 1.5361691246275069, + "learning_rate": 9.335375170572056e-06, + "loss": 0.3664, + "step": 8678 + }, + { + "epoch": 0.1910411622276029, + "grad_norm": 1.7768147918426473, + "learning_rate": 9.335197580984346e-06, + "loss": 0.3274, + "step": 8679 + }, + { + "epoch": 0.19106317411402157, + "grad_norm": 1.7376413751706012, + "learning_rate": 9.335019969363226e-06, + "loss": 0.3906, + "step": 8680 + }, + { + "epoch": 0.19108518600044025, + "grad_norm": 1.7032523710084935, + "learning_rate": 9.334842335709602e-06, + "loss": 0.3648, + "step": 8681 + }, + { + "epoch": 0.1911071978868589, + "grad_norm": 1.9927411778302033, + "learning_rate": 9.334664680024376e-06, + "loss": 0.3118, + "step": 8682 + }, + { + "epoch": 0.19112920977327758, + "grad_norm": 1.5524099224533108, + "learning_rate": 9.334487002308452e-06, + "loss": 0.3802, + "step": 8683 + }, + { + "epoch": 0.19115122165969622, + "grad_norm": 0.6616825083093553, + "learning_rate": 9.33430930256273e-06, + "loss": 0.5322, + "step": 8684 + }, + { + "epoch": 0.1911732335461149, + "grad_norm": 1.385080791074549, + "learning_rate": 9.334131580788116e-06, + "loss": 0.3676, + "step": 8685 + }, + { + "epoch": 0.19119524543253358, + "grad_norm": 1.659963028302166, + "learning_rate": 9.333953836985512e-06, + "loss": 0.3787, + "step": 8686 + }, + { + "epoch": 0.19121725731895223, + "grad_norm": 1.316119527585999, + "learning_rate": 9.33377607115582e-06, + "loss": 0.3056, + "step": 8687 + }, + { + "epoch": 0.1912392692053709, + "grad_norm": 1.4123665253726143, + "learning_rate": 9.333598283299947e-06, + "loss": 0.2956, + "step": 8688 + }, + { + "epoch": 0.19126128109178955, + "grad_norm": 1.6729372564960396, + "learning_rate": 9.333420473418795e-06, + "loss": 0.3795, + "step": 8689 + }, + { + "epoch": 0.19128329297820823, + "grad_norm": 0.6167562540433099, + "learning_rate": 9.333242641513268e-06, + "loss": 0.541, + "step": 8690 + }, + { + "epoch": 0.1913053048646269, + "grad_norm": 0.6466017868154511, + "learning_rate": 9.333064787584265e-06, + "loss": 0.5364, + "step": 8691 + }, + { + "epoch": 0.19132731675104556, + "grad_norm": 1.6255214602663133, + "learning_rate": 9.3328869116327e-06, + "loss": 0.2842, + "step": 8692 + }, + { + "epoch": 0.19134932863746423, + "grad_norm": 1.2764038970561644, + "learning_rate": 9.332709013659466e-06, + "loss": 0.3253, + "step": 8693 + }, + { + "epoch": 0.1913713405238829, + "grad_norm": 1.4780197334770877, + "learning_rate": 9.332531093665475e-06, + "loss": 0.3068, + "step": 8694 + }, + { + "epoch": 0.19139335241030156, + "grad_norm": 1.7055507110800685, + "learning_rate": 9.332353151651629e-06, + "loss": 0.3822, + "step": 8695 + }, + { + "epoch": 0.19141536429672024, + "grad_norm": 1.613748868743717, + "learning_rate": 9.33217518761883e-06, + "loss": 0.4841, + "step": 8696 + }, + { + "epoch": 0.19143737618313889, + "grad_norm": 5.287468627909282, + "learning_rate": 9.331997201567987e-06, + "loss": 0.3228, + "step": 8697 + }, + { + "epoch": 0.19145938806955756, + "grad_norm": 0.6576170593231662, + "learning_rate": 9.3318191935e-06, + "loss": 0.5486, + "step": 8698 + }, + { + "epoch": 0.19148139995597624, + "grad_norm": 1.3041097161322626, + "learning_rate": 9.331641163415777e-06, + "loss": 0.3676, + "step": 8699 + }, + { + "epoch": 0.1915034118423949, + "grad_norm": 0.6525042980686824, + "learning_rate": 9.331463111316222e-06, + "loss": 0.5457, + "step": 8700 + }, + { + "epoch": 0.19152542372881357, + "grad_norm": 1.3168542863959964, + "learning_rate": 9.331285037202238e-06, + "loss": 0.36, + "step": 8701 + }, + { + "epoch": 0.19154743561523221, + "grad_norm": 1.6141862073718716, + "learning_rate": 9.331106941074732e-06, + "loss": 0.3421, + "step": 8702 + }, + { + "epoch": 0.1915694475016509, + "grad_norm": 4.290865353701841, + "learning_rate": 9.330928822934608e-06, + "loss": 0.334, + "step": 8703 + }, + { + "epoch": 0.19159145938806957, + "grad_norm": 0.6292177740706962, + "learning_rate": 9.330750682782775e-06, + "loss": 0.5369, + "step": 8704 + }, + { + "epoch": 0.19161347127448822, + "grad_norm": 1.799002582343769, + "learning_rate": 9.330572520620134e-06, + "loss": 0.4363, + "step": 8705 + }, + { + "epoch": 0.1916354831609069, + "grad_norm": 0.6418362658209806, + "learning_rate": 9.33039433644759e-06, + "loss": 0.5604, + "step": 8706 + }, + { + "epoch": 0.19165749504732554, + "grad_norm": 1.6901362204767907, + "learning_rate": 9.330216130266052e-06, + "loss": 0.3959, + "step": 8707 + }, + { + "epoch": 0.19167950693374422, + "grad_norm": 1.519795283661259, + "learning_rate": 9.330037902076426e-06, + "loss": 0.3856, + "step": 8708 + }, + { + "epoch": 0.1917015188201629, + "grad_norm": 1.8897728309578177, + "learning_rate": 9.329859651879614e-06, + "loss": 0.3084, + "step": 8709 + }, + { + "epoch": 0.19172353070658155, + "grad_norm": 1.4346433984276958, + "learning_rate": 9.329681379676524e-06, + "loss": 0.4204, + "step": 8710 + }, + { + "epoch": 0.19174554259300022, + "grad_norm": 1.7228609594797595, + "learning_rate": 9.329503085468063e-06, + "loss": 0.3711, + "step": 8711 + }, + { + "epoch": 0.19176755447941887, + "grad_norm": 1.9125656372954436, + "learning_rate": 9.329324769255137e-06, + "loss": 0.4167, + "step": 8712 + }, + { + "epoch": 0.19178956636583755, + "grad_norm": 1.4737158741331107, + "learning_rate": 9.32914643103865e-06, + "loss": 0.329, + "step": 8713 + }, + { + "epoch": 0.19181157825225623, + "grad_norm": 0.6355269064017802, + "learning_rate": 9.328968070819512e-06, + "loss": 0.5293, + "step": 8714 + }, + { + "epoch": 0.19183359013867488, + "grad_norm": 1.6037064268501204, + "learning_rate": 9.328789688598626e-06, + "loss": 0.342, + "step": 8715 + }, + { + "epoch": 0.19185560202509355, + "grad_norm": 9.150530404912024, + "learning_rate": 9.3286112843769e-06, + "loss": 0.4156, + "step": 8716 + }, + { + "epoch": 0.19187761391151223, + "grad_norm": 1.4080664053668384, + "learning_rate": 9.328432858155242e-06, + "loss": 0.4034, + "step": 8717 + }, + { + "epoch": 0.19189962579793088, + "grad_norm": 1.4350493681993155, + "learning_rate": 9.328254409934555e-06, + "loss": 0.3059, + "step": 8718 + }, + { + "epoch": 0.19192163768434956, + "grad_norm": 1.2566296558091472, + "learning_rate": 9.328075939715753e-06, + "loss": 0.3616, + "step": 8719 + }, + { + "epoch": 0.1919436495707682, + "grad_norm": 1.5475888105869133, + "learning_rate": 9.327897447499737e-06, + "loss": 0.3489, + "step": 8720 + }, + { + "epoch": 0.19196566145718688, + "grad_norm": 1.5051960953306622, + "learning_rate": 9.327718933287415e-06, + "loss": 0.3281, + "step": 8721 + }, + { + "epoch": 0.19198767334360556, + "grad_norm": 1.6507588204473091, + "learning_rate": 9.327540397079694e-06, + "loss": 0.4716, + "step": 8722 + }, + { + "epoch": 0.1920096852300242, + "grad_norm": 1.3436225768249932, + "learning_rate": 9.327361838877485e-06, + "loss": 0.281, + "step": 8723 + }, + { + "epoch": 0.19203169711644288, + "grad_norm": 1.6807652668466966, + "learning_rate": 9.327183258681692e-06, + "loss": 0.4171, + "step": 8724 + }, + { + "epoch": 0.19205370900286153, + "grad_norm": 1.333339325367706, + "learning_rate": 9.327004656493223e-06, + "loss": 0.3691, + "step": 8725 + }, + { + "epoch": 0.1920757208892802, + "grad_norm": 0.6797135509256041, + "learning_rate": 9.326826032312987e-06, + "loss": 0.531, + "step": 8726 + }, + { + "epoch": 0.1920977327756989, + "grad_norm": 1.490378923392607, + "learning_rate": 9.326647386141892e-06, + "loss": 0.3669, + "step": 8727 + }, + { + "epoch": 0.19211974466211754, + "grad_norm": 1.4544104641341433, + "learning_rate": 9.326468717980844e-06, + "loss": 0.2775, + "step": 8728 + }, + { + "epoch": 0.1921417565485362, + "grad_norm": 1.4363855381423003, + "learning_rate": 9.326290027830753e-06, + "loss": 0.3143, + "step": 8729 + }, + { + "epoch": 0.19216376843495486, + "grad_norm": 2.197390330283762, + "learning_rate": 9.326111315692526e-06, + "loss": 0.3409, + "step": 8730 + }, + { + "epoch": 0.19218578032137354, + "grad_norm": 8.113052620005307, + "learning_rate": 9.325932581567073e-06, + "loss": 0.3538, + "step": 8731 + }, + { + "epoch": 0.19220779220779222, + "grad_norm": 1.4346398069761621, + "learning_rate": 9.325753825455299e-06, + "loss": 0.3345, + "step": 8732 + }, + { + "epoch": 0.19222980409421087, + "grad_norm": 1.2068994364353507, + "learning_rate": 9.325575047358118e-06, + "loss": 0.2754, + "step": 8733 + }, + { + "epoch": 0.19225181598062954, + "grad_norm": 1.2516324111290444, + "learning_rate": 9.325396247276432e-06, + "loss": 0.3716, + "step": 8734 + }, + { + "epoch": 0.19227382786704822, + "grad_norm": 1.283704044335511, + "learning_rate": 9.325217425211154e-06, + "loss": 0.3802, + "step": 8735 + }, + { + "epoch": 0.19229583975346687, + "grad_norm": 1.3541527551568593, + "learning_rate": 9.325038581163193e-06, + "loss": 0.373, + "step": 8736 + }, + { + "epoch": 0.19231785163988555, + "grad_norm": 1.5250207748163644, + "learning_rate": 9.324859715133457e-06, + "loss": 0.3689, + "step": 8737 + }, + { + "epoch": 0.1923398635263042, + "grad_norm": 0.6576926999829975, + "learning_rate": 9.324680827122854e-06, + "loss": 0.5238, + "step": 8738 + }, + { + "epoch": 0.19236187541272287, + "grad_norm": 1.5403051637339509, + "learning_rate": 9.324501917132294e-06, + "loss": 0.3585, + "step": 8739 + }, + { + "epoch": 0.19238388729914155, + "grad_norm": 4.299722646354386, + "learning_rate": 9.324322985162688e-06, + "loss": 0.3994, + "step": 8740 + }, + { + "epoch": 0.1924058991855602, + "grad_norm": 1.399823821867308, + "learning_rate": 9.324144031214943e-06, + "loss": 0.3185, + "step": 8741 + }, + { + "epoch": 0.19242791107197887, + "grad_norm": 2.1762785347011397, + "learning_rate": 9.323965055289969e-06, + "loss": 0.364, + "step": 8742 + }, + { + "epoch": 0.19244992295839752, + "grad_norm": 1.2828594628878003, + "learning_rate": 9.323786057388678e-06, + "loss": 0.3206, + "step": 8743 + }, + { + "epoch": 0.1924719348448162, + "grad_norm": 1.2590227866628492, + "learning_rate": 9.323607037511975e-06, + "loss": 0.3177, + "step": 8744 + }, + { + "epoch": 0.19249394673123488, + "grad_norm": 1.4901265876590448, + "learning_rate": 9.323427995660775e-06, + "loss": 0.3303, + "step": 8745 + }, + { + "epoch": 0.19251595861765353, + "grad_norm": 0.666785202903706, + "learning_rate": 9.323248931835984e-06, + "loss": 0.5432, + "step": 8746 + }, + { + "epoch": 0.1925379705040722, + "grad_norm": 1.2394522195261661, + "learning_rate": 9.323069846038513e-06, + "loss": 0.3205, + "step": 8747 + }, + { + "epoch": 0.19255998239049085, + "grad_norm": 1.4670669183371916, + "learning_rate": 9.322890738269275e-06, + "loss": 0.2965, + "step": 8748 + }, + { + "epoch": 0.19258199427690953, + "grad_norm": 1.226892557981118, + "learning_rate": 9.322711608529177e-06, + "loss": 0.3028, + "step": 8749 + }, + { + "epoch": 0.1926040061633282, + "grad_norm": 1.208138283257749, + "learning_rate": 9.322532456819131e-06, + "loss": 0.3502, + "step": 8750 + }, + { + "epoch": 0.19262601804974686, + "grad_norm": 1.794228414454892, + "learning_rate": 9.322353283140047e-06, + "loss": 0.3557, + "step": 8751 + }, + { + "epoch": 0.19264802993616553, + "grad_norm": 0.6617712199503681, + "learning_rate": 9.322174087492836e-06, + "loss": 0.5203, + "step": 8752 + }, + { + "epoch": 0.1926700418225842, + "grad_norm": 1.287499760236792, + "learning_rate": 9.321994869878409e-06, + "loss": 0.4589, + "step": 8753 + }, + { + "epoch": 0.19269205370900286, + "grad_norm": 1.706526486342281, + "learning_rate": 9.321815630297674e-06, + "loss": 0.3086, + "step": 8754 + }, + { + "epoch": 0.19271406559542154, + "grad_norm": 1.7564162098945464, + "learning_rate": 9.321636368751547e-06, + "loss": 0.5005, + "step": 8755 + }, + { + "epoch": 0.19273607748184018, + "grad_norm": 1.4602065309957886, + "learning_rate": 9.321457085240937e-06, + "loss": 0.4319, + "step": 8756 + }, + { + "epoch": 0.19275808936825886, + "grad_norm": 1.765821224396469, + "learning_rate": 9.321277779766751e-06, + "loss": 0.2955, + "step": 8757 + }, + { + "epoch": 0.19278010125467754, + "grad_norm": 1.4883639189089557, + "learning_rate": 9.321098452329908e-06, + "loss": 0.4267, + "step": 8758 + }, + { + "epoch": 0.1928021131410962, + "grad_norm": 1.327103967450587, + "learning_rate": 9.320919102931314e-06, + "loss": 0.3287, + "step": 8759 + }, + { + "epoch": 0.19282412502751486, + "grad_norm": 1.4628857359608833, + "learning_rate": 9.320739731571883e-06, + "loss": 0.3237, + "step": 8760 + }, + { + "epoch": 0.1928461369139335, + "grad_norm": 1.4636925880654426, + "learning_rate": 9.320560338252522e-06, + "loss": 0.3462, + "step": 8761 + }, + { + "epoch": 0.1928681488003522, + "grad_norm": 1.4126911651665766, + "learning_rate": 9.320380922974149e-06, + "loss": 0.3542, + "step": 8762 + }, + { + "epoch": 0.19289016068677087, + "grad_norm": 1.2137917461572008, + "learning_rate": 9.320201485737673e-06, + "loss": 0.2975, + "step": 8763 + }, + { + "epoch": 0.19291217257318952, + "grad_norm": 1.5947335220071568, + "learning_rate": 9.320022026544006e-06, + "loss": 0.3494, + "step": 8764 + }, + { + "epoch": 0.1929341844596082, + "grad_norm": 1.4199499738852421, + "learning_rate": 9.31984254539406e-06, + "loss": 0.428, + "step": 8765 + }, + { + "epoch": 0.19295619634602684, + "grad_norm": 0.6981180600378546, + "learning_rate": 9.319663042288747e-06, + "loss": 0.5596, + "step": 8766 + }, + { + "epoch": 0.19297820823244552, + "grad_norm": 1.5088843357205706, + "learning_rate": 9.319483517228982e-06, + "loss": 0.374, + "step": 8767 + }, + { + "epoch": 0.1930002201188642, + "grad_norm": 1.1060374826710901, + "learning_rate": 9.319303970215672e-06, + "loss": 0.2958, + "step": 8768 + }, + { + "epoch": 0.19302223200528285, + "grad_norm": 0.6429330244918204, + "learning_rate": 9.319124401249735e-06, + "loss": 0.5445, + "step": 8769 + }, + { + "epoch": 0.19304424389170152, + "grad_norm": 1.1259764772345044, + "learning_rate": 9.31894481033208e-06, + "loss": 0.3377, + "step": 8770 + }, + { + "epoch": 0.19306625577812017, + "grad_norm": 1.3577861694201234, + "learning_rate": 9.318765197463625e-06, + "loss": 0.4284, + "step": 8771 + }, + { + "epoch": 0.19308826766453885, + "grad_norm": 1.9695570472873714, + "learning_rate": 9.318585562645276e-06, + "loss": 0.3495, + "step": 8772 + }, + { + "epoch": 0.19311027955095753, + "grad_norm": 2.0192223588201608, + "learning_rate": 9.318405905877948e-06, + "loss": 0.2567, + "step": 8773 + }, + { + "epoch": 0.19313229143737617, + "grad_norm": 2.238020353174853, + "learning_rate": 9.318226227162555e-06, + "loss": 0.3935, + "step": 8774 + }, + { + "epoch": 0.19315430332379485, + "grad_norm": 1.563155308130905, + "learning_rate": 9.318046526500014e-06, + "loss": 0.3635, + "step": 8775 + }, + { + "epoch": 0.19317631521021353, + "grad_norm": 4.886561572069851, + "learning_rate": 9.317866803891231e-06, + "loss": 0.4819, + "step": 8776 + }, + { + "epoch": 0.19319832709663218, + "grad_norm": 0.6770326316624196, + "learning_rate": 9.317687059337125e-06, + "loss": 0.5263, + "step": 8777 + }, + { + "epoch": 0.19322033898305085, + "grad_norm": 2.4465122083616606, + "learning_rate": 9.317507292838605e-06, + "loss": 0.3854, + "step": 8778 + }, + { + "epoch": 0.1932423508694695, + "grad_norm": 1.4329253462939546, + "learning_rate": 9.317327504396589e-06, + "loss": 0.3868, + "step": 8779 + }, + { + "epoch": 0.19326436275588818, + "grad_norm": 1.6413900490508375, + "learning_rate": 9.31714769401199e-06, + "loss": 0.2942, + "step": 8780 + }, + { + "epoch": 0.19328637464230686, + "grad_norm": 1.5283844671186178, + "learning_rate": 9.31696786168572e-06, + "loss": 0.4264, + "step": 8781 + }, + { + "epoch": 0.1933083865287255, + "grad_norm": 1.9194635760945362, + "learning_rate": 9.316788007418695e-06, + "loss": 0.449, + "step": 8782 + }, + { + "epoch": 0.19333039841514418, + "grad_norm": 1.7169353775786873, + "learning_rate": 9.316608131211826e-06, + "loss": 0.4058, + "step": 8783 + }, + { + "epoch": 0.19335241030156283, + "grad_norm": 1.3771165753391308, + "learning_rate": 9.316428233066031e-06, + "loss": 0.3552, + "step": 8784 + }, + { + "epoch": 0.1933744221879815, + "grad_norm": 2.4819932025506724, + "learning_rate": 9.316248312982222e-06, + "loss": 0.4061, + "step": 8785 + }, + { + "epoch": 0.1933964340744002, + "grad_norm": 0.6733574269931438, + "learning_rate": 9.316068370961313e-06, + "loss": 0.556, + "step": 8786 + }, + { + "epoch": 0.19341844596081884, + "grad_norm": 1.2587415625567993, + "learning_rate": 9.31588840700422e-06, + "loss": 0.3263, + "step": 8787 + }, + { + "epoch": 0.1934404578472375, + "grad_norm": 0.6266774535719486, + "learning_rate": 9.315708421111859e-06, + "loss": 0.5298, + "step": 8788 + }, + { + "epoch": 0.19346246973365616, + "grad_norm": 1.7417191215897276, + "learning_rate": 9.315528413285141e-06, + "loss": 0.4495, + "step": 8789 + }, + { + "epoch": 0.19348448162007484, + "grad_norm": 1.3881564617293123, + "learning_rate": 9.315348383524984e-06, + "loss": 0.4082, + "step": 8790 + }, + { + "epoch": 0.19350649350649352, + "grad_norm": 2.2882554400811896, + "learning_rate": 9.3151683318323e-06, + "loss": 0.2907, + "step": 8791 + }, + { + "epoch": 0.19352850539291216, + "grad_norm": 1.784218087544064, + "learning_rate": 9.314988258208008e-06, + "loss": 0.2883, + "step": 8792 + }, + { + "epoch": 0.19355051727933084, + "grad_norm": 1.5230208578892335, + "learning_rate": 9.314808162653019e-06, + "loss": 0.2451, + "step": 8793 + }, + { + "epoch": 0.19357252916574952, + "grad_norm": 0.6997923837448454, + "learning_rate": 9.314628045168253e-06, + "loss": 0.5396, + "step": 8794 + }, + { + "epoch": 0.19359454105216817, + "grad_norm": 1.5021131268171186, + "learning_rate": 9.31444790575462e-06, + "loss": 0.3306, + "step": 8795 + }, + { + "epoch": 0.19361655293858684, + "grad_norm": 1.4496646910601116, + "learning_rate": 9.314267744413042e-06, + "loss": 0.4202, + "step": 8796 + }, + { + "epoch": 0.1936385648250055, + "grad_norm": 1.7132387309719956, + "learning_rate": 9.314087561144429e-06, + "loss": 0.372, + "step": 8797 + }, + { + "epoch": 0.19366057671142417, + "grad_norm": 1.3909009774906251, + "learning_rate": 9.313907355949699e-06, + "loss": 0.3164, + "step": 8798 + }, + { + "epoch": 0.19368258859784285, + "grad_norm": 1.220724475064422, + "learning_rate": 9.313727128829767e-06, + "loss": 0.3723, + "step": 8799 + }, + { + "epoch": 0.1937046004842615, + "grad_norm": 1.5268674685480894, + "learning_rate": 9.31354687978555e-06, + "loss": 0.3545, + "step": 8800 + }, + { + "epoch": 0.19372661237068017, + "grad_norm": 1.2102244388788486, + "learning_rate": 9.313366608817965e-06, + "loss": 0.3288, + "step": 8801 + }, + { + "epoch": 0.19374862425709882, + "grad_norm": 0.640068043358285, + "learning_rate": 9.313186315927925e-06, + "loss": 0.5476, + "step": 8802 + }, + { + "epoch": 0.1937706361435175, + "grad_norm": 2.310523741746313, + "learning_rate": 9.31300600111635e-06, + "loss": 0.3667, + "step": 8803 + }, + { + "epoch": 0.19379264802993618, + "grad_norm": 1.5214670176791663, + "learning_rate": 9.312825664384155e-06, + "loss": 0.3125, + "step": 8804 + }, + { + "epoch": 0.19381465991635483, + "grad_norm": 1.3806144804361142, + "learning_rate": 9.312645305732256e-06, + "loss": 0.2752, + "step": 8805 + }, + { + "epoch": 0.1938366718027735, + "grad_norm": 1.2240274438600074, + "learning_rate": 9.312464925161571e-06, + "loss": 0.329, + "step": 8806 + }, + { + "epoch": 0.19385868368919215, + "grad_norm": 0.6240331832716204, + "learning_rate": 9.312284522673014e-06, + "loss": 0.5473, + "step": 8807 + }, + { + "epoch": 0.19388069557561083, + "grad_norm": 1.3239195941874897, + "learning_rate": 9.312104098267505e-06, + "loss": 0.3958, + "step": 8808 + }, + { + "epoch": 0.1939027074620295, + "grad_norm": 1.3548699132332849, + "learning_rate": 9.31192365194596e-06, + "loss": 0.3302, + "step": 8809 + }, + { + "epoch": 0.19392471934844815, + "grad_norm": 0.6301897577042879, + "learning_rate": 9.311743183709295e-06, + "loss": 0.5137, + "step": 8810 + }, + { + "epoch": 0.19394673123486683, + "grad_norm": 1.3587658996469796, + "learning_rate": 9.311562693558428e-06, + "loss": 0.2711, + "step": 8811 + }, + { + "epoch": 0.19396874312128548, + "grad_norm": 1.9737313649644648, + "learning_rate": 9.311382181494278e-06, + "loss": 0.36, + "step": 8812 + }, + { + "epoch": 0.19399075500770416, + "grad_norm": 1.2600010023947128, + "learning_rate": 9.311201647517758e-06, + "loss": 0.3066, + "step": 8813 + }, + { + "epoch": 0.19401276689412283, + "grad_norm": 0.5990429514713279, + "learning_rate": 9.31102109162979e-06, + "loss": 0.5259, + "step": 8814 + }, + { + "epoch": 0.19403477878054148, + "grad_norm": 1.7113080031730334, + "learning_rate": 9.310840513831289e-06, + "loss": 0.3479, + "step": 8815 + }, + { + "epoch": 0.19405679066696016, + "grad_norm": 1.0897320907527683, + "learning_rate": 9.310659914123174e-06, + "loss": 0.2464, + "step": 8816 + }, + { + "epoch": 0.19407880255337884, + "grad_norm": 1.652371886987868, + "learning_rate": 9.310479292506365e-06, + "loss": 0.3382, + "step": 8817 + }, + { + "epoch": 0.1941008144397975, + "grad_norm": 1.2474067877382389, + "learning_rate": 9.310298648981776e-06, + "loss": 0.3288, + "step": 8818 + }, + { + "epoch": 0.19412282632621616, + "grad_norm": 1.399809636081313, + "learning_rate": 9.310117983550327e-06, + "loss": 0.3247, + "step": 8819 + }, + { + "epoch": 0.1941448382126348, + "grad_norm": 1.4142468588351342, + "learning_rate": 9.309937296212936e-06, + "loss": 0.3981, + "step": 8820 + }, + { + "epoch": 0.1941668500990535, + "grad_norm": 0.6400767388464638, + "learning_rate": 9.309756586970522e-06, + "loss": 0.5373, + "step": 8821 + }, + { + "epoch": 0.19418886198547217, + "grad_norm": 1.1469915811997362, + "learning_rate": 9.309575855824002e-06, + "loss": 0.336, + "step": 8822 + }, + { + "epoch": 0.19421087387189082, + "grad_norm": 1.9695635591788712, + "learning_rate": 9.309395102774294e-06, + "loss": 0.4044, + "step": 8823 + }, + { + "epoch": 0.1942328857583095, + "grad_norm": 1.6711799326996934, + "learning_rate": 9.30921432782232e-06, + "loss": 0.3199, + "step": 8824 + }, + { + "epoch": 0.19425489764472814, + "grad_norm": 0.6242818320756076, + "learning_rate": 9.309033530968997e-06, + "loss": 0.5393, + "step": 8825 + }, + { + "epoch": 0.19427690953114682, + "grad_norm": 1.257358698806986, + "learning_rate": 9.308852712215243e-06, + "loss": 0.3007, + "step": 8826 + }, + { + "epoch": 0.1942989214175655, + "grad_norm": 1.3622250996463423, + "learning_rate": 9.308671871561978e-06, + "loss": 0.3729, + "step": 8827 + }, + { + "epoch": 0.19432093330398414, + "grad_norm": 1.381646014223393, + "learning_rate": 9.308491009010121e-06, + "loss": 0.3473, + "step": 8828 + }, + { + "epoch": 0.19434294519040282, + "grad_norm": 1.5641504251060834, + "learning_rate": 9.30831012456059e-06, + "loss": 0.3598, + "step": 8829 + }, + { + "epoch": 0.19436495707682147, + "grad_norm": 1.6759426755456976, + "learning_rate": 9.308129218214308e-06, + "loss": 0.3886, + "step": 8830 + }, + { + "epoch": 0.19438696896324015, + "grad_norm": 1.2515163323150784, + "learning_rate": 9.30794828997219e-06, + "loss": 0.3091, + "step": 8831 + }, + { + "epoch": 0.19440898084965882, + "grad_norm": 1.3858675569222967, + "learning_rate": 9.307767339835157e-06, + "loss": 0.3309, + "step": 8832 + }, + { + "epoch": 0.19443099273607747, + "grad_norm": 0.6602477918249916, + "learning_rate": 9.307586367804129e-06, + "loss": 0.5209, + "step": 8833 + }, + { + "epoch": 0.19445300462249615, + "grad_norm": 2.8342437739960022, + "learning_rate": 9.307405373880025e-06, + "loss": 0.3447, + "step": 8834 + }, + { + "epoch": 0.19447501650891483, + "grad_norm": 0.6461566270728913, + "learning_rate": 9.307224358063768e-06, + "loss": 0.5367, + "step": 8835 + }, + { + "epoch": 0.19449702839533348, + "grad_norm": 1.3819961554122908, + "learning_rate": 9.307043320356275e-06, + "loss": 0.3745, + "step": 8836 + }, + { + "epoch": 0.19451904028175215, + "grad_norm": 0.6288402545176892, + "learning_rate": 9.306862260758466e-06, + "loss": 0.5327, + "step": 8837 + }, + { + "epoch": 0.1945410521681708, + "grad_norm": 2.488228834896023, + "learning_rate": 9.306681179271264e-06, + "loss": 0.3531, + "step": 8838 + }, + { + "epoch": 0.19456306405458948, + "grad_norm": 1.1245041796720519, + "learning_rate": 9.306500075895585e-06, + "loss": 0.3776, + "step": 8839 + }, + { + "epoch": 0.19458507594100816, + "grad_norm": 1.5884133102377103, + "learning_rate": 9.306318950632353e-06, + "loss": 0.4494, + "step": 8840 + }, + { + "epoch": 0.1946070878274268, + "grad_norm": 1.5009686993777862, + "learning_rate": 9.306137803482487e-06, + "loss": 0.4133, + "step": 8841 + }, + { + "epoch": 0.19462909971384548, + "grad_norm": 1.54784212707822, + "learning_rate": 9.305956634446909e-06, + "loss": 0.3257, + "step": 8842 + }, + { + "epoch": 0.19465111160026413, + "grad_norm": 3.777160694937442, + "learning_rate": 9.305775443526538e-06, + "loss": 0.3016, + "step": 8843 + }, + { + "epoch": 0.1946731234866828, + "grad_norm": 0.6579448215104179, + "learning_rate": 9.305594230722296e-06, + "loss": 0.5335, + "step": 8844 + }, + { + "epoch": 0.19469513537310149, + "grad_norm": 4.010687086052679, + "learning_rate": 9.305412996035104e-06, + "loss": 0.3202, + "step": 8845 + }, + { + "epoch": 0.19471714725952013, + "grad_norm": 1.3655689335190253, + "learning_rate": 9.305231739465884e-06, + "loss": 0.331, + "step": 8846 + }, + { + "epoch": 0.1947391591459388, + "grad_norm": 1.5751853674366114, + "learning_rate": 9.305050461015554e-06, + "loss": 0.3781, + "step": 8847 + }, + { + "epoch": 0.19476117103235746, + "grad_norm": 1.9566500323768885, + "learning_rate": 9.304869160685037e-06, + "loss": 0.4037, + "step": 8848 + }, + { + "epoch": 0.19478318291877614, + "grad_norm": 0.6410283738666829, + "learning_rate": 9.304687838475257e-06, + "loss": 0.5426, + "step": 8849 + }, + { + "epoch": 0.19480519480519481, + "grad_norm": 1.5014547548693244, + "learning_rate": 9.304506494387133e-06, + "loss": 0.229, + "step": 8850 + }, + { + "epoch": 0.19482720669161346, + "grad_norm": 1.635375105177387, + "learning_rate": 9.304325128421586e-06, + "loss": 0.4415, + "step": 8851 + }, + { + "epoch": 0.19484921857803214, + "grad_norm": 1.8448427797919287, + "learning_rate": 9.30414374057954e-06, + "loss": 0.3362, + "step": 8852 + }, + { + "epoch": 0.1948712304644508, + "grad_norm": 1.6460032308997272, + "learning_rate": 9.303962330861915e-06, + "loss": 0.3264, + "step": 8853 + }, + { + "epoch": 0.19489324235086947, + "grad_norm": 2.098812042577103, + "learning_rate": 9.303780899269634e-06, + "loss": 0.3802, + "step": 8854 + }, + { + "epoch": 0.19491525423728814, + "grad_norm": 1.6403857721069135, + "learning_rate": 9.303599445803619e-06, + "loss": 0.3359, + "step": 8855 + }, + { + "epoch": 0.1949372661237068, + "grad_norm": 0.6258455385830568, + "learning_rate": 9.303417970464794e-06, + "loss": 0.532, + "step": 8856 + }, + { + "epoch": 0.19495927801012547, + "grad_norm": 7.131205067260459, + "learning_rate": 9.303236473254078e-06, + "loss": 0.4072, + "step": 8857 + }, + { + "epoch": 0.19498128989654415, + "grad_norm": 1.9574962781080816, + "learning_rate": 9.303054954172395e-06, + "loss": 0.3784, + "step": 8858 + }, + { + "epoch": 0.1950033017829628, + "grad_norm": 0.6430477186047734, + "learning_rate": 9.302873413220666e-06, + "loss": 0.5353, + "step": 8859 + }, + { + "epoch": 0.19502531366938147, + "grad_norm": 1.3844145478528853, + "learning_rate": 9.302691850399817e-06, + "loss": 0.3704, + "step": 8860 + }, + { + "epoch": 0.19504732555580012, + "grad_norm": 0.6045193576762627, + "learning_rate": 9.30251026571077e-06, + "loss": 0.4964, + "step": 8861 + }, + { + "epoch": 0.1950693374422188, + "grad_norm": 2.050011101371972, + "learning_rate": 9.302328659154445e-06, + "loss": 0.3878, + "step": 8862 + }, + { + "epoch": 0.19509134932863748, + "grad_norm": 1.6992246585206188, + "learning_rate": 9.302147030731767e-06, + "loss": 0.365, + "step": 8863 + }, + { + "epoch": 0.19511336121505612, + "grad_norm": 1.4834305523052753, + "learning_rate": 9.301965380443658e-06, + "loss": 0.4259, + "step": 8864 + }, + { + "epoch": 0.1951353731014748, + "grad_norm": 2.5055911819430507, + "learning_rate": 9.301783708291044e-06, + "loss": 0.3906, + "step": 8865 + }, + { + "epoch": 0.19515738498789345, + "grad_norm": 0.6080269640685351, + "learning_rate": 9.301602014274846e-06, + "loss": 0.5081, + "step": 8866 + }, + { + "epoch": 0.19517939687431213, + "grad_norm": 0.639934012138489, + "learning_rate": 9.301420298395987e-06, + "loss": 0.5232, + "step": 8867 + }, + { + "epoch": 0.1952014087607308, + "grad_norm": 0.6139782063818199, + "learning_rate": 9.301238560655393e-06, + "loss": 0.5324, + "step": 8868 + }, + { + "epoch": 0.19522342064714945, + "grad_norm": 1.5195498696957683, + "learning_rate": 9.301056801053987e-06, + "loss": 0.2434, + "step": 8869 + }, + { + "epoch": 0.19524543253356813, + "grad_norm": 1.3829427873491662, + "learning_rate": 9.30087501959269e-06, + "loss": 0.3666, + "step": 8870 + }, + { + "epoch": 0.19526744441998678, + "grad_norm": 1.7199835594639756, + "learning_rate": 9.30069321627243e-06, + "loss": 0.2991, + "step": 8871 + }, + { + "epoch": 0.19528945630640546, + "grad_norm": 1.4195165842711133, + "learning_rate": 9.300511391094126e-06, + "loss": 0.3434, + "step": 8872 + }, + { + "epoch": 0.19531146819282413, + "grad_norm": 0.6736365844279746, + "learning_rate": 9.300329544058705e-06, + "loss": 0.5739, + "step": 8873 + }, + { + "epoch": 0.19533348007924278, + "grad_norm": 0.6400458576747807, + "learning_rate": 9.300147675167095e-06, + "loss": 0.5105, + "step": 8874 + }, + { + "epoch": 0.19535549196566146, + "grad_norm": 1.2926353290797992, + "learning_rate": 9.299965784420214e-06, + "loss": 0.3648, + "step": 8875 + }, + { + "epoch": 0.19537750385208014, + "grad_norm": 2.164744524219867, + "learning_rate": 9.29978387181899e-06, + "loss": 0.4248, + "step": 8876 + }, + { + "epoch": 0.19539951573849879, + "grad_norm": 1.9342728487070628, + "learning_rate": 9.299601937364347e-06, + "loss": 0.3143, + "step": 8877 + }, + { + "epoch": 0.19542152762491746, + "grad_norm": 1.4887739372850235, + "learning_rate": 9.299419981057207e-06, + "loss": 0.3917, + "step": 8878 + }, + { + "epoch": 0.1954435395113361, + "grad_norm": 2.0568086955469367, + "learning_rate": 9.299238002898499e-06, + "loss": 0.3927, + "step": 8879 + }, + { + "epoch": 0.1954655513977548, + "grad_norm": 1.2192949883599822, + "learning_rate": 9.299056002889145e-06, + "loss": 0.3596, + "step": 8880 + }, + { + "epoch": 0.19548756328417347, + "grad_norm": 5.27027845779004, + "learning_rate": 9.298873981030071e-06, + "loss": 0.4004, + "step": 8881 + }, + { + "epoch": 0.19550957517059211, + "grad_norm": 1.2216690669644545, + "learning_rate": 9.298691937322202e-06, + "loss": 0.3258, + "step": 8882 + }, + { + "epoch": 0.1955315870570108, + "grad_norm": 1.3718761975174416, + "learning_rate": 9.298509871766464e-06, + "loss": 0.2967, + "step": 8883 + }, + { + "epoch": 0.19555359894342944, + "grad_norm": 1.702405097275711, + "learning_rate": 9.29832778436378e-06, + "loss": 0.3566, + "step": 8884 + }, + { + "epoch": 0.19557561082984812, + "grad_norm": 1.227433450071216, + "learning_rate": 9.29814567511508e-06, + "loss": 0.3266, + "step": 8885 + }, + { + "epoch": 0.1955976227162668, + "grad_norm": 1.4271710378706977, + "learning_rate": 9.297963544021286e-06, + "loss": 0.4286, + "step": 8886 + }, + { + "epoch": 0.19561963460268544, + "grad_norm": 1.5119363492467075, + "learning_rate": 9.297781391083322e-06, + "loss": 0.3122, + "step": 8887 + }, + { + "epoch": 0.19564164648910412, + "grad_norm": 1.3801212264375244, + "learning_rate": 9.297599216302117e-06, + "loss": 0.4394, + "step": 8888 + }, + { + "epoch": 0.19566365837552277, + "grad_norm": 0.81880098766656, + "learning_rate": 9.297417019678598e-06, + "loss": 0.5514, + "step": 8889 + }, + { + "epoch": 0.19568567026194145, + "grad_norm": 1.7120303362597422, + "learning_rate": 9.297234801213689e-06, + "loss": 0.3559, + "step": 8890 + }, + { + "epoch": 0.19570768214836012, + "grad_norm": 1.6548843462088767, + "learning_rate": 9.297052560908313e-06, + "loss": 0.3197, + "step": 8891 + }, + { + "epoch": 0.19572969403477877, + "grad_norm": 1.5646432605649554, + "learning_rate": 9.296870298763401e-06, + "loss": 0.3659, + "step": 8892 + }, + { + "epoch": 0.19575170592119745, + "grad_norm": 0.6461199059478553, + "learning_rate": 9.296688014779877e-06, + "loss": 0.5616, + "step": 8893 + }, + { + "epoch": 0.1957737178076161, + "grad_norm": 1.1122967931439358, + "learning_rate": 9.296505708958669e-06, + "loss": 0.301, + "step": 8894 + }, + { + "epoch": 0.19579572969403478, + "grad_norm": 1.7207644914564562, + "learning_rate": 9.296323381300703e-06, + "loss": 0.3216, + "step": 8895 + }, + { + "epoch": 0.19581774158045345, + "grad_norm": 1.384238954845201, + "learning_rate": 9.296141031806905e-06, + "loss": 0.4308, + "step": 8896 + }, + { + "epoch": 0.1958397534668721, + "grad_norm": 1.8834657383513476, + "learning_rate": 9.295958660478203e-06, + "loss": 0.3487, + "step": 8897 + }, + { + "epoch": 0.19586176535329078, + "grad_norm": 1.6880569638576328, + "learning_rate": 9.295776267315522e-06, + "loss": 0.3513, + "step": 8898 + }, + { + "epoch": 0.19588377723970946, + "grad_norm": 1.6647310713800243, + "learning_rate": 9.29559385231979e-06, + "loss": 0.3139, + "step": 8899 + }, + { + "epoch": 0.1959057891261281, + "grad_norm": 1.6895715533603612, + "learning_rate": 9.295411415491933e-06, + "loss": 0.4055, + "step": 8900 + }, + { + "epoch": 0.19592780101254678, + "grad_norm": 1.3364294435258852, + "learning_rate": 9.295228956832883e-06, + "loss": 0.3019, + "step": 8901 + }, + { + "epoch": 0.19594981289896543, + "grad_norm": 3.424460356043358, + "learning_rate": 9.29504647634356e-06, + "loss": 0.4473, + "step": 8902 + }, + { + "epoch": 0.1959718247853841, + "grad_norm": 2.0221095773260456, + "learning_rate": 9.294863974024897e-06, + "loss": 0.3503, + "step": 8903 + }, + { + "epoch": 0.19599383667180278, + "grad_norm": 0.9428257232051663, + "learning_rate": 9.294681449877818e-06, + "loss": 0.5421, + "step": 8904 + }, + { + "epoch": 0.19601584855822143, + "grad_norm": 1.3021982010825244, + "learning_rate": 9.294498903903254e-06, + "loss": 0.2157, + "step": 8905 + }, + { + "epoch": 0.1960378604446401, + "grad_norm": 1.6599852573684892, + "learning_rate": 9.294316336102132e-06, + "loss": 0.4235, + "step": 8906 + }, + { + "epoch": 0.19605987233105876, + "grad_norm": 1.2686478557393481, + "learning_rate": 9.294133746475376e-06, + "loss": 0.3422, + "step": 8907 + }, + { + "epoch": 0.19608188421747744, + "grad_norm": 1.5615874910049494, + "learning_rate": 9.29395113502392e-06, + "loss": 0.3337, + "step": 8908 + }, + { + "epoch": 0.1961038961038961, + "grad_norm": 1.4360072765935872, + "learning_rate": 9.293768501748687e-06, + "loss": 0.2734, + "step": 8909 + }, + { + "epoch": 0.19612590799031476, + "grad_norm": 3.452354997922297, + "learning_rate": 9.29358584665061e-06, + "loss": 0.3431, + "step": 8910 + }, + { + "epoch": 0.19614791987673344, + "grad_norm": 1.1751703380010075, + "learning_rate": 9.293403169730613e-06, + "loss": 0.2989, + "step": 8911 + }, + { + "epoch": 0.1961699317631521, + "grad_norm": 1.9155366281776405, + "learning_rate": 9.293220470989625e-06, + "loss": 0.3317, + "step": 8912 + }, + { + "epoch": 0.19619194364957077, + "grad_norm": 1.9996869951738, + "learning_rate": 9.293037750428579e-06, + "loss": 0.4285, + "step": 8913 + }, + { + "epoch": 0.19621395553598944, + "grad_norm": 1.654456618957549, + "learning_rate": 9.292855008048398e-06, + "loss": 0.3918, + "step": 8914 + }, + { + "epoch": 0.1962359674224081, + "grad_norm": 1.8672404860889078, + "learning_rate": 9.292672243850015e-06, + "loss": 0.3399, + "step": 8915 + }, + { + "epoch": 0.19625797930882677, + "grad_norm": 1.7680437526136559, + "learning_rate": 9.292489457834355e-06, + "loss": 0.3433, + "step": 8916 + }, + { + "epoch": 0.19627999119524545, + "grad_norm": 1.5603781728578936, + "learning_rate": 9.292306650002349e-06, + "loss": 0.3118, + "step": 8917 + }, + { + "epoch": 0.1963020030816641, + "grad_norm": 1.2486879004756144, + "learning_rate": 9.292123820354928e-06, + "loss": 0.3182, + "step": 8918 + }, + { + "epoch": 0.19632401496808277, + "grad_norm": 1.5855959970780058, + "learning_rate": 9.291940968893019e-06, + "loss": 0.3569, + "step": 8919 + }, + { + "epoch": 0.19634602685450142, + "grad_norm": 2.046117115594034, + "learning_rate": 9.29175809561755e-06, + "loss": 0.4408, + "step": 8920 + }, + { + "epoch": 0.1963680387409201, + "grad_norm": 1.5642879425343506, + "learning_rate": 9.291575200529454e-06, + "loss": 0.3946, + "step": 8921 + }, + { + "epoch": 0.19639005062733877, + "grad_norm": 0.9896407029412657, + "learning_rate": 9.291392283629657e-06, + "loss": 0.538, + "step": 8922 + }, + { + "epoch": 0.19641206251375742, + "grad_norm": 1.550986877428781, + "learning_rate": 9.291209344919092e-06, + "loss": 0.3881, + "step": 8923 + }, + { + "epoch": 0.1964340744001761, + "grad_norm": 1.28795401350412, + "learning_rate": 9.291026384398687e-06, + "loss": 0.3061, + "step": 8924 + }, + { + "epoch": 0.19645608628659475, + "grad_norm": 0.6792754520442714, + "learning_rate": 9.290843402069371e-06, + "loss": 0.5599, + "step": 8925 + }, + { + "epoch": 0.19647809817301343, + "grad_norm": 1.4069412793825238, + "learning_rate": 9.290660397932077e-06, + "loss": 0.3558, + "step": 8926 + }, + { + "epoch": 0.1965001100594321, + "grad_norm": 1.3139046214328318, + "learning_rate": 9.290477371987731e-06, + "loss": 0.3102, + "step": 8927 + }, + { + "epoch": 0.19652212194585075, + "grad_norm": 2.008769203199727, + "learning_rate": 9.290294324237268e-06, + "loss": 0.4554, + "step": 8928 + }, + { + "epoch": 0.19654413383226943, + "grad_norm": 1.7654271316966872, + "learning_rate": 9.290111254681612e-06, + "loss": 0.3866, + "step": 8929 + }, + { + "epoch": 0.19656614571868808, + "grad_norm": 1.5620552155387004, + "learning_rate": 9.2899281633217e-06, + "loss": 0.4299, + "step": 8930 + }, + { + "epoch": 0.19658815760510676, + "grad_norm": 0.9789578308079073, + "learning_rate": 9.28974505015846e-06, + "loss": 0.5777, + "step": 8931 + }, + { + "epoch": 0.19661016949152543, + "grad_norm": 1.4873299826601556, + "learning_rate": 9.289561915192821e-06, + "loss": 0.3164, + "step": 8932 + }, + { + "epoch": 0.19663218137794408, + "grad_norm": 5.148696542432325, + "learning_rate": 9.289378758425716e-06, + "loss": 0.3859, + "step": 8933 + }, + { + "epoch": 0.19665419326436276, + "grad_norm": 1.5691538557852924, + "learning_rate": 9.289195579858073e-06, + "loss": 0.3078, + "step": 8934 + }, + { + "epoch": 0.19667620515078144, + "grad_norm": 2.1532228486395932, + "learning_rate": 9.289012379490826e-06, + "loss": 0.3501, + "step": 8935 + }, + { + "epoch": 0.19669821703720008, + "grad_norm": 1.5013427480531636, + "learning_rate": 9.288829157324907e-06, + "loss": 0.3672, + "step": 8936 + }, + { + "epoch": 0.19672022892361876, + "grad_norm": 1.3981835353783307, + "learning_rate": 9.288645913361245e-06, + "loss": 0.3254, + "step": 8937 + }, + { + "epoch": 0.1967422408100374, + "grad_norm": 2.2039135034333213, + "learning_rate": 9.28846264760077e-06, + "loss": 0.3362, + "step": 8938 + }, + { + "epoch": 0.1967642526964561, + "grad_norm": 2.6969458720999127, + "learning_rate": 9.288279360044417e-06, + "loss": 0.4322, + "step": 8939 + }, + { + "epoch": 0.19678626458287476, + "grad_norm": 0.7022696408116947, + "learning_rate": 9.288096050693113e-06, + "loss": 0.5352, + "step": 8940 + }, + { + "epoch": 0.19680827646929341, + "grad_norm": 1.5453258774765413, + "learning_rate": 9.287912719547793e-06, + "loss": 0.3946, + "step": 8941 + }, + { + "epoch": 0.1968302883557121, + "grad_norm": 1.3944721227313723, + "learning_rate": 9.28772936660939e-06, + "loss": 0.3133, + "step": 8942 + }, + { + "epoch": 0.19685230024213074, + "grad_norm": 1.386816092242444, + "learning_rate": 9.287545991878832e-06, + "loss": 0.3553, + "step": 8943 + }, + { + "epoch": 0.19687431212854942, + "grad_norm": 1.7361288679128142, + "learning_rate": 9.287362595357054e-06, + "loss": 0.3825, + "step": 8944 + }, + { + "epoch": 0.1968963240149681, + "grad_norm": 1.691291686463515, + "learning_rate": 9.287179177044988e-06, + "loss": 0.3724, + "step": 8945 + }, + { + "epoch": 0.19691833590138674, + "grad_norm": 1.5703139971075049, + "learning_rate": 9.286995736943564e-06, + "loss": 0.3373, + "step": 8946 + }, + { + "epoch": 0.19694034778780542, + "grad_norm": 1.3176366738752219, + "learning_rate": 9.286812275053716e-06, + "loss": 0.3344, + "step": 8947 + }, + { + "epoch": 0.19696235967422407, + "grad_norm": 1.8989768561663718, + "learning_rate": 9.286628791376376e-06, + "loss": 0.3577, + "step": 8948 + }, + { + "epoch": 0.19698437156064275, + "grad_norm": 2.116314747518626, + "learning_rate": 9.286445285912476e-06, + "loss": 0.3485, + "step": 8949 + }, + { + "epoch": 0.19700638344706142, + "grad_norm": 1.3688032790382179, + "learning_rate": 9.28626175866295e-06, + "loss": 0.3487, + "step": 8950 + }, + { + "epoch": 0.19702839533348007, + "grad_norm": 1.4948247996613884, + "learning_rate": 9.28607820962873e-06, + "loss": 0.378, + "step": 8951 + }, + { + "epoch": 0.19705040721989875, + "grad_norm": 1.8201653763598435, + "learning_rate": 9.285894638810748e-06, + "loss": 0.3379, + "step": 8952 + }, + { + "epoch": 0.1970724191063174, + "grad_norm": 0.6643517310256566, + "learning_rate": 9.28571104620994e-06, + "loss": 0.5585, + "step": 8953 + }, + { + "epoch": 0.19709443099273607, + "grad_norm": 1.292065549066447, + "learning_rate": 9.285527431827236e-06, + "loss": 0.2301, + "step": 8954 + }, + { + "epoch": 0.19711644287915475, + "grad_norm": 0.6705190746389165, + "learning_rate": 9.285343795663569e-06, + "loss": 0.5282, + "step": 8955 + }, + { + "epoch": 0.1971384547655734, + "grad_norm": 1.3495706876713673, + "learning_rate": 9.285160137719872e-06, + "loss": 0.4028, + "step": 8956 + }, + { + "epoch": 0.19716046665199208, + "grad_norm": 1.4762008399692619, + "learning_rate": 9.284976457997084e-06, + "loss": 0.3645, + "step": 8957 + }, + { + "epoch": 0.19718247853841075, + "grad_norm": 2.104869540280251, + "learning_rate": 9.284792756496133e-06, + "loss": 0.336, + "step": 8958 + }, + { + "epoch": 0.1972044904248294, + "grad_norm": 0.632200537196637, + "learning_rate": 9.284609033217953e-06, + "loss": 0.5413, + "step": 8959 + }, + { + "epoch": 0.19722650231124808, + "grad_norm": 1.5927987445454492, + "learning_rate": 9.28442528816348e-06, + "loss": 0.3679, + "step": 8960 + }, + { + "epoch": 0.19724851419766673, + "grad_norm": 1.513920833916814, + "learning_rate": 9.284241521333646e-06, + "loss": 0.3394, + "step": 8961 + }, + { + "epoch": 0.1972705260840854, + "grad_norm": 1.9579828005183502, + "learning_rate": 9.284057732729387e-06, + "loss": 0.3352, + "step": 8962 + }, + { + "epoch": 0.19729253797050408, + "grad_norm": 1.6652918722344472, + "learning_rate": 9.283873922351636e-06, + "loss": 0.3481, + "step": 8963 + }, + { + "epoch": 0.19731454985692273, + "grad_norm": 2.2178472515131786, + "learning_rate": 9.283690090201324e-06, + "loss": 0.3984, + "step": 8964 + }, + { + "epoch": 0.1973365617433414, + "grad_norm": 1.541189871427686, + "learning_rate": 9.283506236279392e-06, + "loss": 0.4137, + "step": 8965 + }, + { + "epoch": 0.19735857362976006, + "grad_norm": 1.5715955657996654, + "learning_rate": 9.283322360586769e-06, + "loss": 0.3761, + "step": 8966 + }, + { + "epoch": 0.19738058551617874, + "grad_norm": 1.2887190712621501, + "learning_rate": 9.283138463124393e-06, + "loss": 0.2237, + "step": 8967 + }, + { + "epoch": 0.1974025974025974, + "grad_norm": 1.9944408338671633, + "learning_rate": 9.282954543893194e-06, + "loss": 0.3022, + "step": 8968 + }, + { + "epoch": 0.19742460928901606, + "grad_norm": 2.0352091539747663, + "learning_rate": 9.282770602894111e-06, + "loss": 0.3279, + "step": 8969 + }, + { + "epoch": 0.19744662117543474, + "grad_norm": 1.2464549556320341, + "learning_rate": 9.282586640128079e-06, + "loss": 0.3252, + "step": 8970 + }, + { + "epoch": 0.1974686330618534, + "grad_norm": 1.3245533834597163, + "learning_rate": 9.28240265559603e-06, + "loss": 0.3135, + "step": 8971 + }, + { + "epoch": 0.19749064494827206, + "grad_norm": 1.5660650539513554, + "learning_rate": 9.282218649298903e-06, + "loss": 0.3894, + "step": 8972 + }, + { + "epoch": 0.19751265683469074, + "grad_norm": 1.8044576142230078, + "learning_rate": 9.282034621237628e-06, + "loss": 0.3504, + "step": 8973 + }, + { + "epoch": 0.1975346687211094, + "grad_norm": 1.3986042337696751, + "learning_rate": 9.281850571413144e-06, + "loss": 0.3426, + "step": 8974 + }, + { + "epoch": 0.19755668060752807, + "grad_norm": 0.7859081422147206, + "learning_rate": 9.281666499826387e-06, + "loss": 0.5495, + "step": 8975 + }, + { + "epoch": 0.19757869249394674, + "grad_norm": 1.5016952458315964, + "learning_rate": 9.28148240647829e-06, + "loss": 0.3127, + "step": 8976 + }, + { + "epoch": 0.1976007043803654, + "grad_norm": 1.580481394359308, + "learning_rate": 9.281298291369791e-06, + "loss": 0.4039, + "step": 8977 + }, + { + "epoch": 0.19762271626678407, + "grad_norm": 1.3798917543984976, + "learning_rate": 9.281114154501823e-06, + "loss": 0.3716, + "step": 8978 + }, + { + "epoch": 0.19764472815320272, + "grad_norm": 1.2566870642390342, + "learning_rate": 9.280929995875324e-06, + "loss": 0.264, + "step": 8979 + }, + { + "epoch": 0.1976667400396214, + "grad_norm": 1.9099611247921373, + "learning_rate": 9.28074581549123e-06, + "loss": 0.387, + "step": 8980 + }, + { + "epoch": 0.19768875192604007, + "grad_norm": 0.6507109190107602, + "learning_rate": 9.280561613350476e-06, + "loss": 0.5421, + "step": 8981 + }, + { + "epoch": 0.19771076381245872, + "grad_norm": 1.1579142355030525, + "learning_rate": 9.280377389453998e-06, + "loss": 0.2252, + "step": 8982 + }, + { + "epoch": 0.1977327756988774, + "grad_norm": 1.3188636543553887, + "learning_rate": 9.280193143802736e-06, + "loss": 0.4152, + "step": 8983 + }, + { + "epoch": 0.19775478758529605, + "grad_norm": 1.3852096898165562, + "learning_rate": 9.280008876397619e-06, + "loss": 0.3405, + "step": 8984 + }, + { + "epoch": 0.19777679947171473, + "grad_norm": 1.4695200652004914, + "learning_rate": 9.27982458723959e-06, + "loss": 0.422, + "step": 8985 + }, + { + "epoch": 0.1977988113581334, + "grad_norm": 1.5389548713173327, + "learning_rate": 9.279640276329584e-06, + "loss": 0.3604, + "step": 8986 + }, + { + "epoch": 0.19782082324455205, + "grad_norm": 1.539457653579003, + "learning_rate": 9.279455943668537e-06, + "loss": 0.4002, + "step": 8987 + }, + { + "epoch": 0.19784283513097073, + "grad_norm": 2.0819611851710884, + "learning_rate": 9.279271589257387e-06, + "loss": 0.4072, + "step": 8988 + }, + { + "epoch": 0.19786484701738938, + "grad_norm": 0.6114880409563723, + "learning_rate": 9.279087213097069e-06, + "loss": 0.5039, + "step": 8989 + }, + { + "epoch": 0.19788685890380805, + "grad_norm": 1.3508529665949782, + "learning_rate": 9.278902815188522e-06, + "loss": 0.3619, + "step": 8990 + }, + { + "epoch": 0.19790887079022673, + "grad_norm": 0.6344798177642114, + "learning_rate": 9.278718395532683e-06, + "loss": 0.5572, + "step": 8991 + }, + { + "epoch": 0.19793088267664538, + "grad_norm": 0.6450896838105765, + "learning_rate": 9.278533954130488e-06, + "loss": 0.5439, + "step": 8992 + }, + { + "epoch": 0.19795289456306406, + "grad_norm": 3.148001232580412, + "learning_rate": 9.278349490982874e-06, + "loss": 0.4153, + "step": 8993 + }, + { + "epoch": 0.1979749064494827, + "grad_norm": 1.1945765514741167, + "learning_rate": 9.278165006090781e-06, + "loss": 0.3624, + "step": 8994 + }, + { + "epoch": 0.19799691833590138, + "grad_norm": 1.7467093889301843, + "learning_rate": 9.277980499455146e-06, + "loss": 0.3619, + "step": 8995 + }, + { + "epoch": 0.19801893022232006, + "grad_norm": 0.6066763399673748, + "learning_rate": 9.277795971076903e-06, + "loss": 0.5242, + "step": 8996 + }, + { + "epoch": 0.1980409421087387, + "grad_norm": 1.8353632958978048, + "learning_rate": 9.277611420956995e-06, + "loss": 0.3681, + "step": 8997 + }, + { + "epoch": 0.1980629539951574, + "grad_norm": 1.2632946210800404, + "learning_rate": 9.27742684909636e-06, + "loss": 0.2564, + "step": 8998 + }, + { + "epoch": 0.19808496588157606, + "grad_norm": 1.7239763661973977, + "learning_rate": 9.277242255495931e-06, + "loss": 0.3531, + "step": 8999 + }, + { + "epoch": 0.1981069777679947, + "grad_norm": 1.1934380859191143, + "learning_rate": 9.27705764015665e-06, + "loss": 0.3185, + "step": 9000 + }, + { + "epoch": 0.1981289896544134, + "grad_norm": 2.137059598689563, + "learning_rate": 9.276873003079455e-06, + "loss": 0.3303, + "step": 9001 + }, + { + "epoch": 0.19815100154083204, + "grad_norm": 1.7044593541325537, + "learning_rate": 9.276688344265284e-06, + "loss": 0.3925, + "step": 9002 + }, + { + "epoch": 0.19817301342725072, + "grad_norm": 1.8619175529483458, + "learning_rate": 9.276503663715075e-06, + "loss": 0.3476, + "step": 9003 + }, + { + "epoch": 0.1981950253136694, + "grad_norm": 0.6574845061297384, + "learning_rate": 9.276318961429767e-06, + "loss": 0.5501, + "step": 9004 + }, + { + "epoch": 0.19821703720008804, + "grad_norm": 1.3565710680618088, + "learning_rate": 9.276134237410298e-06, + "loss": 0.4281, + "step": 9005 + }, + { + "epoch": 0.19823904908650672, + "grad_norm": 1.5518239846182424, + "learning_rate": 9.275949491657607e-06, + "loss": 0.3521, + "step": 9006 + }, + { + "epoch": 0.19826106097292537, + "grad_norm": 1.3471407319558875, + "learning_rate": 9.275764724172636e-06, + "loss": 0.3539, + "step": 9007 + }, + { + "epoch": 0.19828307285934405, + "grad_norm": 1.2663274724395586, + "learning_rate": 9.27557993495632e-06, + "loss": 0.4016, + "step": 9008 + }, + { + "epoch": 0.19830508474576272, + "grad_norm": 1.877085384802255, + "learning_rate": 9.2753951240096e-06, + "loss": 0.508, + "step": 9009 + }, + { + "epoch": 0.19832709663218137, + "grad_norm": 1.789989257591703, + "learning_rate": 9.275210291333416e-06, + "loss": 0.3981, + "step": 9010 + }, + { + "epoch": 0.19834910851860005, + "grad_norm": 0.6577798161778337, + "learning_rate": 9.275025436928705e-06, + "loss": 0.5862, + "step": 9011 + }, + { + "epoch": 0.1983711204050187, + "grad_norm": 1.4553715161151595, + "learning_rate": 9.274840560796409e-06, + "loss": 0.3372, + "step": 9012 + }, + { + "epoch": 0.19839313229143737, + "grad_norm": 2.266071977434196, + "learning_rate": 9.274655662937466e-06, + "loss": 0.4396, + "step": 9013 + }, + { + "epoch": 0.19841514417785605, + "grad_norm": 2.2775553212658197, + "learning_rate": 9.274470743352816e-06, + "loss": 0.3765, + "step": 9014 + }, + { + "epoch": 0.1984371560642747, + "grad_norm": 1.2949222768150424, + "learning_rate": 9.274285802043401e-06, + "loss": 0.3423, + "step": 9015 + }, + { + "epoch": 0.19845916795069338, + "grad_norm": 1.592262190018209, + "learning_rate": 9.274100839010158e-06, + "loss": 0.2464, + "step": 9016 + }, + { + "epoch": 0.19848117983711205, + "grad_norm": 1.2214730443871684, + "learning_rate": 9.273915854254029e-06, + "loss": 0.3519, + "step": 9017 + }, + { + "epoch": 0.1985031917235307, + "grad_norm": 3.3174353706043913, + "learning_rate": 9.273730847775952e-06, + "loss": 0.3242, + "step": 9018 + }, + { + "epoch": 0.19852520360994938, + "grad_norm": 1.5113020712922183, + "learning_rate": 9.27354581957687e-06, + "loss": 0.37, + "step": 9019 + }, + { + "epoch": 0.19854721549636803, + "grad_norm": 1.580184032896393, + "learning_rate": 9.273360769657721e-06, + "loss": 0.3682, + "step": 9020 + }, + { + "epoch": 0.1985692273827867, + "grad_norm": 1.3031259067807057, + "learning_rate": 9.273175698019448e-06, + "loss": 0.3725, + "step": 9021 + }, + { + "epoch": 0.19859123926920538, + "grad_norm": 2.072561233878429, + "learning_rate": 9.272990604662987e-06, + "loss": 0.2838, + "step": 9022 + }, + { + "epoch": 0.19861325115562403, + "grad_norm": 0.6811686961955766, + "learning_rate": 9.272805489589286e-06, + "loss": 0.5455, + "step": 9023 + }, + { + "epoch": 0.1986352630420427, + "grad_norm": 1.2180770269110908, + "learning_rate": 9.27262035279928e-06, + "loss": 0.3218, + "step": 9024 + }, + { + "epoch": 0.19865727492846136, + "grad_norm": 3.8016316482638133, + "learning_rate": 9.272435194293912e-06, + "loss": 0.4088, + "step": 9025 + }, + { + "epoch": 0.19867928681488004, + "grad_norm": 2.1282785690646335, + "learning_rate": 9.272250014074122e-06, + "loss": 0.3789, + "step": 9026 + }, + { + "epoch": 0.1987012987012987, + "grad_norm": 0.6606509925569392, + "learning_rate": 9.272064812140853e-06, + "loss": 0.5247, + "step": 9027 + }, + { + "epoch": 0.19872331058771736, + "grad_norm": 1.3496064372421095, + "learning_rate": 9.271879588495045e-06, + "loss": 0.4169, + "step": 9028 + }, + { + "epoch": 0.19874532247413604, + "grad_norm": 1.1472169554602174, + "learning_rate": 9.27169434313764e-06, + "loss": 0.2757, + "step": 9029 + }, + { + "epoch": 0.1987673343605547, + "grad_norm": 1.2905143770298733, + "learning_rate": 9.271509076069578e-06, + "loss": 0.2801, + "step": 9030 + }, + { + "epoch": 0.19878934624697336, + "grad_norm": 1.3509190542313108, + "learning_rate": 9.271323787291802e-06, + "loss": 0.2644, + "step": 9031 + }, + { + "epoch": 0.19881135813339204, + "grad_norm": 0.6417106027122508, + "learning_rate": 9.271138476805253e-06, + "loss": 0.5221, + "step": 9032 + }, + { + "epoch": 0.1988333700198107, + "grad_norm": 1.7491305715256422, + "learning_rate": 9.270953144610875e-06, + "loss": 0.4791, + "step": 9033 + }, + { + "epoch": 0.19885538190622937, + "grad_norm": 1.4089408852999095, + "learning_rate": 9.270767790709607e-06, + "loss": 0.2909, + "step": 9034 + }, + { + "epoch": 0.19887739379264802, + "grad_norm": 1.5906068848485846, + "learning_rate": 9.270582415102392e-06, + "loss": 0.3323, + "step": 9035 + }, + { + "epoch": 0.1988994056790667, + "grad_norm": 1.715615814668397, + "learning_rate": 9.270397017790174e-06, + "loss": 0.4042, + "step": 9036 + }, + { + "epoch": 0.19892141756548537, + "grad_norm": 1.8404420677089801, + "learning_rate": 9.270211598773892e-06, + "loss": 0.4144, + "step": 9037 + }, + { + "epoch": 0.19894342945190402, + "grad_norm": 1.5355403897049653, + "learning_rate": 9.27002615805449e-06, + "loss": 0.3912, + "step": 9038 + }, + { + "epoch": 0.1989654413383227, + "grad_norm": 1.2773407013056455, + "learning_rate": 9.269840695632913e-06, + "loss": 0.3949, + "step": 9039 + }, + { + "epoch": 0.19898745322474137, + "grad_norm": 2.2824549756065093, + "learning_rate": 9.269655211510097e-06, + "loss": 0.3631, + "step": 9040 + }, + { + "epoch": 0.19900946511116002, + "grad_norm": 1.6892906744031366, + "learning_rate": 9.269469705686992e-06, + "loss": 0.2802, + "step": 9041 + }, + { + "epoch": 0.1990314769975787, + "grad_norm": 1.4561873091868804, + "learning_rate": 9.269284178164538e-06, + "loss": 0.361, + "step": 9042 + }, + { + "epoch": 0.19905348888399735, + "grad_norm": 1.3022420752730988, + "learning_rate": 9.269098628943676e-06, + "loss": 0.3329, + "step": 9043 + }, + { + "epoch": 0.19907550077041603, + "grad_norm": 0.6948185288908476, + "learning_rate": 9.268913058025353e-06, + "loss": 0.5393, + "step": 9044 + }, + { + "epoch": 0.1990975126568347, + "grad_norm": 1.2311590170591016, + "learning_rate": 9.268727465410507e-06, + "loss": 0.4447, + "step": 9045 + }, + { + "epoch": 0.19911952454325335, + "grad_norm": 1.30196979977363, + "learning_rate": 9.268541851100086e-06, + "loss": 0.3647, + "step": 9046 + }, + { + "epoch": 0.19914153642967203, + "grad_norm": 1.2050507627308211, + "learning_rate": 9.26835621509503e-06, + "loss": 0.3509, + "step": 9047 + }, + { + "epoch": 0.19916354831609068, + "grad_norm": 1.4230195183782854, + "learning_rate": 9.268170557396285e-06, + "loss": 0.319, + "step": 9048 + }, + { + "epoch": 0.19918556020250935, + "grad_norm": 2.8685842842384153, + "learning_rate": 9.267984878004795e-06, + "loss": 0.293, + "step": 9049 + }, + { + "epoch": 0.19920757208892803, + "grad_norm": 1.3481974905216527, + "learning_rate": 9.2677991769215e-06, + "loss": 0.361, + "step": 9050 + }, + { + "epoch": 0.19922958397534668, + "grad_norm": 2.1355804765486686, + "learning_rate": 9.267613454147346e-06, + "loss": 0.326, + "step": 9051 + }, + { + "epoch": 0.19925159586176536, + "grad_norm": 1.6611619684677423, + "learning_rate": 9.26742770968328e-06, + "loss": 0.383, + "step": 9052 + }, + { + "epoch": 0.199273607748184, + "grad_norm": 2.018812185588809, + "learning_rate": 9.26724194353024e-06, + "loss": 0.2907, + "step": 9053 + }, + { + "epoch": 0.19929561963460268, + "grad_norm": 0.6579165234022263, + "learning_rate": 9.267056155689173e-06, + "loss": 0.526, + "step": 9054 + }, + { + "epoch": 0.19931763152102136, + "grad_norm": 1.510170732149609, + "learning_rate": 9.266870346161026e-06, + "loss": 0.2808, + "step": 9055 + }, + { + "epoch": 0.19933964340744, + "grad_norm": 1.8713295242796122, + "learning_rate": 9.266684514946738e-06, + "loss": 0.3206, + "step": 9056 + }, + { + "epoch": 0.19936165529385869, + "grad_norm": 1.7719942247250837, + "learning_rate": 9.266498662047258e-06, + "loss": 0.3432, + "step": 9057 + }, + { + "epoch": 0.19938366718027736, + "grad_norm": 1.6771211374574098, + "learning_rate": 9.26631278746353e-06, + "loss": 0.4046, + "step": 9058 + }, + { + "epoch": 0.199405679066696, + "grad_norm": 1.1348991107097832, + "learning_rate": 9.266126891196493e-06, + "loss": 0.3602, + "step": 9059 + }, + { + "epoch": 0.1994276909531147, + "grad_norm": 1.8493923724521792, + "learning_rate": 9.2659409732471e-06, + "loss": 0.3891, + "step": 9060 + }, + { + "epoch": 0.19944970283953334, + "grad_norm": 1.2060951928003492, + "learning_rate": 9.265755033616292e-06, + "loss": 0.2366, + "step": 9061 + }, + { + "epoch": 0.19947171472595202, + "grad_norm": 1.4542095815976628, + "learning_rate": 9.265569072305016e-06, + "loss": 0.3343, + "step": 9062 + }, + { + "epoch": 0.1994937266123707, + "grad_norm": 1.580244823539149, + "learning_rate": 9.265383089314213e-06, + "loss": 0.4006, + "step": 9063 + }, + { + "epoch": 0.19951573849878934, + "grad_norm": 0.6409181170315861, + "learning_rate": 9.265197084644832e-06, + "loss": 0.5364, + "step": 9064 + }, + { + "epoch": 0.19953775038520802, + "grad_norm": 1.377242697989331, + "learning_rate": 9.265011058297815e-06, + "loss": 0.4263, + "step": 9065 + }, + { + "epoch": 0.19955976227162667, + "grad_norm": 1.8328335968673615, + "learning_rate": 9.264825010274111e-06, + "loss": 0.3194, + "step": 9066 + }, + { + "epoch": 0.19958177415804534, + "grad_norm": 1.751434593533325, + "learning_rate": 9.264638940574665e-06, + "loss": 0.2633, + "step": 9067 + }, + { + "epoch": 0.19960378604446402, + "grad_norm": 1.484256795865039, + "learning_rate": 9.264452849200421e-06, + "loss": 0.3946, + "step": 9068 + }, + { + "epoch": 0.19962579793088267, + "grad_norm": 0.589717334646394, + "learning_rate": 9.264266736152325e-06, + "loss": 0.5237, + "step": 9069 + }, + { + "epoch": 0.19964780981730135, + "grad_norm": 1.6959477980887976, + "learning_rate": 9.264080601431325e-06, + "loss": 0.4206, + "step": 9070 + }, + { + "epoch": 0.19966982170372, + "grad_norm": 2.9561037640423415, + "learning_rate": 9.263894445038364e-06, + "loss": 0.3839, + "step": 9071 + }, + { + "epoch": 0.19969183359013867, + "grad_norm": 0.6266901136926903, + "learning_rate": 9.263708266974391e-06, + "loss": 0.5387, + "step": 9072 + }, + { + "epoch": 0.19971384547655735, + "grad_norm": 2.5246060817354268, + "learning_rate": 9.263522067240349e-06, + "loss": 0.3173, + "step": 9073 + }, + { + "epoch": 0.199735857362976, + "grad_norm": 1.1210825918329088, + "learning_rate": 9.26333584583719e-06, + "loss": 0.3319, + "step": 9074 + }, + { + "epoch": 0.19975786924939468, + "grad_norm": 0.6443125797198981, + "learning_rate": 9.263149602765854e-06, + "loss": 0.5565, + "step": 9075 + }, + { + "epoch": 0.19977988113581335, + "grad_norm": 0.625701521850689, + "learning_rate": 9.26296333802729e-06, + "loss": 0.5748, + "step": 9076 + }, + { + "epoch": 0.199801893022232, + "grad_norm": 1.268167290883575, + "learning_rate": 9.262777051622444e-06, + "loss": 0.3238, + "step": 9077 + }, + { + "epoch": 0.19982390490865068, + "grad_norm": 0.6214453247700552, + "learning_rate": 9.262590743552266e-06, + "loss": 0.5415, + "step": 9078 + }, + { + "epoch": 0.19984591679506933, + "grad_norm": 1.9117442201073056, + "learning_rate": 9.262404413817701e-06, + "loss": 0.3274, + "step": 9079 + }, + { + "epoch": 0.199867928681488, + "grad_norm": 0.6339695226393471, + "learning_rate": 9.262218062419694e-06, + "loss": 0.5418, + "step": 9080 + }, + { + "epoch": 0.19988994056790668, + "grad_norm": 1.6462870188066918, + "learning_rate": 9.262031689359194e-06, + "loss": 0.4097, + "step": 9081 + }, + { + "epoch": 0.19991195245432533, + "grad_norm": 0.5929931284284773, + "learning_rate": 9.261845294637148e-06, + "loss": 0.5003, + "step": 9082 + }, + { + "epoch": 0.199933964340744, + "grad_norm": 1.4395638923340794, + "learning_rate": 9.261658878254505e-06, + "loss": 0.3635, + "step": 9083 + }, + { + "epoch": 0.19995597622716266, + "grad_norm": 1.5430540036183014, + "learning_rate": 9.261472440212209e-06, + "loss": 0.2748, + "step": 9084 + }, + { + "epoch": 0.19997798811358133, + "grad_norm": 1.4149521355912833, + "learning_rate": 9.261285980511209e-06, + "loss": 0.4364, + "step": 9085 + }, + { + "epoch": 0.2, + "grad_norm": 1.804867084207571, + "learning_rate": 9.261099499152454e-06, + "loss": 0.3406, + "step": 9086 + }, + { + "epoch": 0.20002201188641866, + "grad_norm": 1.70063562998413, + "learning_rate": 9.260912996136893e-06, + "loss": 0.4058, + "step": 9087 + }, + { + "epoch": 0.20004402377283734, + "grad_norm": 3.6771923477478916, + "learning_rate": 9.26072647146547e-06, + "loss": 0.3864, + "step": 9088 + }, + { + "epoch": 0.200066035659256, + "grad_norm": 1.6848970826442158, + "learning_rate": 9.260539925139133e-06, + "loss": 0.3811, + "step": 9089 + }, + { + "epoch": 0.20008804754567466, + "grad_norm": 1.3108458450778198, + "learning_rate": 9.260353357158835e-06, + "loss": 0.4434, + "step": 9090 + }, + { + "epoch": 0.20011005943209334, + "grad_norm": 1.2128185537884435, + "learning_rate": 9.260166767525519e-06, + "loss": 0.3625, + "step": 9091 + }, + { + "epoch": 0.200132071318512, + "grad_norm": 0.6561812457014788, + "learning_rate": 9.259980156240136e-06, + "loss": 0.5437, + "step": 9092 + }, + { + "epoch": 0.20015408320493067, + "grad_norm": 2.2416561884549147, + "learning_rate": 9.259793523303633e-06, + "loss": 0.3193, + "step": 9093 + }, + { + "epoch": 0.20017609509134932, + "grad_norm": 1.5635665639967775, + "learning_rate": 9.25960686871696e-06, + "loss": 0.4335, + "step": 9094 + }, + { + "epoch": 0.200198106977768, + "grad_norm": 2.4894362534436887, + "learning_rate": 9.259420192481066e-06, + "loss": 0.3463, + "step": 9095 + }, + { + "epoch": 0.20022011886418667, + "grad_norm": 2.36390998038675, + "learning_rate": 9.259233494596898e-06, + "loss": 0.4442, + "step": 9096 + }, + { + "epoch": 0.20024213075060532, + "grad_norm": 1.5633708442687406, + "learning_rate": 9.259046775065406e-06, + "loss": 0.4487, + "step": 9097 + }, + { + "epoch": 0.200264142637024, + "grad_norm": 5.916776529858731, + "learning_rate": 9.258860033887538e-06, + "loss": 0.2999, + "step": 9098 + }, + { + "epoch": 0.20028615452344267, + "grad_norm": 0.6317346055535212, + "learning_rate": 9.258673271064244e-06, + "loss": 0.5219, + "step": 9099 + }, + { + "epoch": 0.20030816640986132, + "grad_norm": 1.3304959960440053, + "learning_rate": 9.258486486596472e-06, + "loss": 0.3608, + "step": 9100 + }, + { + "epoch": 0.20033017829628, + "grad_norm": 1.3098646703888683, + "learning_rate": 9.258299680485175e-06, + "loss": 0.3169, + "step": 9101 + }, + { + "epoch": 0.20035219018269865, + "grad_norm": 1.7703936230430353, + "learning_rate": 9.258112852731299e-06, + "loss": 0.3726, + "step": 9102 + }, + { + "epoch": 0.20037420206911732, + "grad_norm": 1.2233294072556171, + "learning_rate": 9.257926003335791e-06, + "loss": 0.3048, + "step": 9103 + }, + { + "epoch": 0.200396213955536, + "grad_norm": 1.8957992459323558, + "learning_rate": 9.257739132299609e-06, + "loss": 0.3497, + "step": 9104 + }, + { + "epoch": 0.20041822584195465, + "grad_norm": 0.6497253281398583, + "learning_rate": 9.257552239623694e-06, + "loss": 0.5119, + "step": 9105 + }, + { + "epoch": 0.20044023772837333, + "grad_norm": 1.5623653958150452, + "learning_rate": 9.257365325309e-06, + "loss": 0.4502, + "step": 9106 + }, + { + "epoch": 0.20046224961479198, + "grad_norm": 0.6353269619236214, + "learning_rate": 9.257178389356478e-06, + "loss": 0.532, + "step": 9107 + }, + { + "epoch": 0.20048426150121065, + "grad_norm": 1.5015893883598568, + "learning_rate": 9.256991431767076e-06, + "loss": 0.3103, + "step": 9108 + }, + { + "epoch": 0.20050627338762933, + "grad_norm": 1.683838561167028, + "learning_rate": 9.256804452541745e-06, + "loss": 0.3895, + "step": 9109 + }, + { + "epoch": 0.20052828527404798, + "grad_norm": 0.6109663291477663, + "learning_rate": 9.256617451681436e-06, + "loss": 0.5014, + "step": 9110 + }, + { + "epoch": 0.20055029716046666, + "grad_norm": 1.569161396025052, + "learning_rate": 9.256430429187097e-06, + "loss": 0.3538, + "step": 9111 + }, + { + "epoch": 0.2005723090468853, + "grad_norm": 1.4310776072432534, + "learning_rate": 9.25624338505968e-06, + "loss": 0.3363, + "step": 9112 + }, + { + "epoch": 0.20059432093330398, + "grad_norm": 0.6399813003693068, + "learning_rate": 9.256056319300137e-06, + "loss": 0.4942, + "step": 9113 + }, + { + "epoch": 0.20061633281972266, + "grad_norm": 1.6275148948073086, + "learning_rate": 9.255869231909417e-06, + "loss": 0.308, + "step": 9114 + }, + { + "epoch": 0.2006383447061413, + "grad_norm": 1.433026546986731, + "learning_rate": 9.255682122888473e-06, + "loss": 0.3716, + "step": 9115 + }, + { + "epoch": 0.20066035659255999, + "grad_norm": 0.6643162292458143, + "learning_rate": 9.255494992238252e-06, + "loss": 0.561, + "step": 9116 + }, + { + "epoch": 0.20068236847897866, + "grad_norm": 0.6364174356171018, + "learning_rate": 9.255307839959709e-06, + "loss": 0.5377, + "step": 9117 + }, + { + "epoch": 0.2007043803653973, + "grad_norm": 0.5860752593966667, + "learning_rate": 9.255120666053792e-06, + "loss": 0.5035, + "step": 9118 + }, + { + "epoch": 0.200726392251816, + "grad_norm": 1.5263493035793456, + "learning_rate": 9.254933470521456e-06, + "loss": 0.4461, + "step": 9119 + }, + { + "epoch": 0.20074840413823464, + "grad_norm": 0.6725933837400399, + "learning_rate": 9.254746253363649e-06, + "loss": 0.5196, + "step": 9120 + }, + { + "epoch": 0.20077041602465331, + "grad_norm": 1.502268709259023, + "learning_rate": 9.254559014581324e-06, + "loss": 0.4303, + "step": 9121 + }, + { + "epoch": 0.200792427911072, + "grad_norm": 1.308635687499001, + "learning_rate": 9.254371754175434e-06, + "loss": 0.3777, + "step": 9122 + }, + { + "epoch": 0.20081443979749064, + "grad_norm": 1.8534354575682541, + "learning_rate": 9.254184472146927e-06, + "loss": 0.2865, + "step": 9123 + }, + { + "epoch": 0.20083645168390932, + "grad_norm": 1.431998729854522, + "learning_rate": 9.253997168496757e-06, + "loss": 0.3814, + "step": 9124 + }, + { + "epoch": 0.20085846357032797, + "grad_norm": 1.7041185480982517, + "learning_rate": 9.253809843225879e-06, + "loss": 0.3359, + "step": 9125 + }, + { + "epoch": 0.20088047545674664, + "grad_norm": 1.6673389582152407, + "learning_rate": 9.25362249633524e-06, + "loss": 0.4779, + "step": 9126 + }, + { + "epoch": 0.20090248734316532, + "grad_norm": 1.3631518158677978, + "learning_rate": 9.253435127825794e-06, + "loss": 0.2801, + "step": 9127 + }, + { + "epoch": 0.20092449922958397, + "grad_norm": 1.5059534989197583, + "learning_rate": 9.253247737698493e-06, + "loss": 0.3201, + "step": 9128 + }, + { + "epoch": 0.20094651111600265, + "grad_norm": 5.272410305305054, + "learning_rate": 9.253060325954292e-06, + "loss": 0.3799, + "step": 9129 + }, + { + "epoch": 0.2009685230024213, + "grad_norm": 0.7065646185963118, + "learning_rate": 9.25287289259414e-06, + "loss": 0.5192, + "step": 9130 + }, + { + "epoch": 0.20099053488883997, + "grad_norm": 1.4217742094055739, + "learning_rate": 9.252685437618992e-06, + "loss": 0.3497, + "step": 9131 + }, + { + "epoch": 0.20101254677525865, + "grad_norm": 1.2575205971644565, + "learning_rate": 9.2524979610298e-06, + "loss": 0.3497, + "step": 9132 + }, + { + "epoch": 0.2010345586616773, + "grad_norm": 1.3919216138600445, + "learning_rate": 9.252310462827517e-06, + "loss": 0.3396, + "step": 9133 + }, + { + "epoch": 0.20105657054809598, + "grad_norm": 1.293784921885733, + "learning_rate": 9.252122943013094e-06, + "loss": 0.3437, + "step": 9134 + }, + { + "epoch": 0.20107858243451462, + "grad_norm": 0.6472435034432309, + "learning_rate": 9.251935401587486e-06, + "loss": 0.5379, + "step": 9135 + }, + { + "epoch": 0.2011005943209333, + "grad_norm": 1.3874398830477612, + "learning_rate": 9.251747838551647e-06, + "loss": 0.3804, + "step": 9136 + }, + { + "epoch": 0.20112260620735198, + "grad_norm": 4.379857117100071, + "learning_rate": 9.25156025390653e-06, + "loss": 0.3608, + "step": 9137 + }, + { + "epoch": 0.20114461809377063, + "grad_norm": 1.7080768450072847, + "learning_rate": 9.251372647653085e-06, + "loss": 0.353, + "step": 9138 + }, + { + "epoch": 0.2011666299801893, + "grad_norm": 1.1161090437629084, + "learning_rate": 9.251185019792271e-06, + "loss": 0.3519, + "step": 9139 + }, + { + "epoch": 0.20118864186660798, + "grad_norm": 0.6774600747463994, + "learning_rate": 9.250997370325036e-06, + "loss": 0.5489, + "step": 9140 + }, + { + "epoch": 0.20121065375302663, + "grad_norm": 0.6550165550580466, + "learning_rate": 9.250809699252336e-06, + "loss": 0.5492, + "step": 9141 + }, + { + "epoch": 0.2012326656394453, + "grad_norm": 1.8660915815549108, + "learning_rate": 9.250622006575128e-06, + "loss": 0.337, + "step": 9142 + }, + { + "epoch": 0.20125467752586396, + "grad_norm": 0.6384772107717475, + "learning_rate": 9.250434292294362e-06, + "loss": 0.5655, + "step": 9143 + }, + { + "epoch": 0.20127668941228263, + "grad_norm": 1.2852048129154365, + "learning_rate": 9.25024655641099e-06, + "loss": 0.3222, + "step": 9144 + }, + { + "epoch": 0.2012987012987013, + "grad_norm": 2.8119545654566105, + "learning_rate": 9.250058798925973e-06, + "loss": 0.3821, + "step": 9145 + }, + { + "epoch": 0.20132071318511996, + "grad_norm": 1.733171852093502, + "learning_rate": 9.249871019840261e-06, + "loss": 0.3763, + "step": 9146 + }, + { + "epoch": 0.20134272507153864, + "grad_norm": 0.6769639849504094, + "learning_rate": 9.249683219154809e-06, + "loss": 0.5249, + "step": 9147 + }, + { + "epoch": 0.20136473695795729, + "grad_norm": 1.2141614121488384, + "learning_rate": 9.249495396870572e-06, + "loss": 0.3083, + "step": 9148 + }, + { + "epoch": 0.20138674884437596, + "grad_norm": 1.3093748417770665, + "learning_rate": 9.249307552988503e-06, + "loss": 0.3736, + "step": 9149 + }, + { + "epoch": 0.20140876073079464, + "grad_norm": 1.3557833283979157, + "learning_rate": 9.249119687509558e-06, + "loss": 0.3439, + "step": 9150 + }, + { + "epoch": 0.2014307726172133, + "grad_norm": 1.9642626693481073, + "learning_rate": 9.24893180043469e-06, + "loss": 0.365, + "step": 9151 + }, + { + "epoch": 0.20145278450363197, + "grad_norm": 0.6799899527497926, + "learning_rate": 9.248743891764858e-06, + "loss": 0.506, + "step": 9152 + }, + { + "epoch": 0.20147479639005061, + "grad_norm": 1.656421187126208, + "learning_rate": 9.248555961501014e-06, + "loss": 0.3411, + "step": 9153 + }, + { + "epoch": 0.2014968082764693, + "grad_norm": 1.414417446425667, + "learning_rate": 9.248368009644112e-06, + "loss": 0.3669, + "step": 9154 + }, + { + "epoch": 0.20151882016288797, + "grad_norm": 2.8289676562741994, + "learning_rate": 9.248180036195111e-06, + "loss": 0.2108, + "step": 9155 + }, + { + "epoch": 0.20154083204930662, + "grad_norm": 1.3437460723808743, + "learning_rate": 9.247992041154964e-06, + "loss": 0.2984, + "step": 9156 + }, + { + "epoch": 0.2015628439357253, + "grad_norm": 2.257149935979003, + "learning_rate": 9.247804024524627e-06, + "loss": 0.3933, + "step": 9157 + }, + { + "epoch": 0.20158485582214397, + "grad_norm": 1.2631136137157382, + "learning_rate": 9.247615986305054e-06, + "loss": 0.305, + "step": 9158 + }, + { + "epoch": 0.20160686770856262, + "grad_norm": 1.7079490439095535, + "learning_rate": 9.247427926497204e-06, + "loss": 0.4309, + "step": 9159 + }, + { + "epoch": 0.2016288795949813, + "grad_norm": 1.365760596476314, + "learning_rate": 9.247239845102029e-06, + "loss": 0.3396, + "step": 9160 + }, + { + "epoch": 0.20165089148139995, + "grad_norm": 1.742350612202036, + "learning_rate": 9.247051742120487e-06, + "loss": 0.4081, + "step": 9161 + }, + { + "epoch": 0.20167290336781862, + "grad_norm": 2.2203702310049813, + "learning_rate": 9.246863617553534e-06, + "loss": 0.3713, + "step": 9162 + }, + { + "epoch": 0.2016949152542373, + "grad_norm": 1.798101189771966, + "learning_rate": 9.246675471402127e-06, + "loss": 0.3465, + "step": 9163 + }, + { + "epoch": 0.20171692714065595, + "grad_norm": 1.5712214730359646, + "learning_rate": 9.24648730366722e-06, + "loss": 0.3455, + "step": 9164 + }, + { + "epoch": 0.20173893902707463, + "grad_norm": 1.5824197909321582, + "learning_rate": 9.246299114349772e-06, + "loss": 0.394, + "step": 9165 + }, + { + "epoch": 0.20176095091349328, + "grad_norm": 0.6832008089357187, + "learning_rate": 9.246110903450736e-06, + "loss": 0.564, + "step": 9166 + }, + { + "epoch": 0.20178296279991195, + "grad_norm": 0.6590049849971142, + "learning_rate": 9.245922670971072e-06, + "loss": 0.5217, + "step": 9167 + }, + { + "epoch": 0.20180497468633063, + "grad_norm": 1.5851987046556282, + "learning_rate": 9.245734416911734e-06, + "loss": 0.415, + "step": 9168 + }, + { + "epoch": 0.20182698657274928, + "grad_norm": 5.225677155046207, + "learning_rate": 9.245546141273682e-06, + "loss": 0.3336, + "step": 9169 + }, + { + "epoch": 0.20184899845916796, + "grad_norm": 0.601678211322812, + "learning_rate": 9.24535784405787e-06, + "loss": 0.5141, + "step": 9170 + }, + { + "epoch": 0.2018710103455866, + "grad_norm": 2.4447535828998994, + "learning_rate": 9.245169525265255e-06, + "loss": 0.3853, + "step": 9171 + }, + { + "epoch": 0.20189302223200528, + "grad_norm": 5.160196634218705, + "learning_rate": 9.244981184896796e-06, + "loss": 0.4025, + "step": 9172 + }, + { + "epoch": 0.20191503411842396, + "grad_norm": 1.410187132186902, + "learning_rate": 9.244792822953449e-06, + "loss": 0.3119, + "step": 9173 + }, + { + "epoch": 0.2019370460048426, + "grad_norm": 2.8471019786509033, + "learning_rate": 9.24460443943617e-06, + "loss": 0.3355, + "step": 9174 + }, + { + "epoch": 0.20195905789126128, + "grad_norm": 1.168658201935611, + "learning_rate": 9.244416034345922e-06, + "loss": 0.2797, + "step": 9175 + }, + { + "epoch": 0.20198106977767993, + "grad_norm": 2.300164885949647, + "learning_rate": 9.244227607683655e-06, + "loss": 0.3407, + "step": 9176 + }, + { + "epoch": 0.2020030816640986, + "grad_norm": 1.4301023760170835, + "learning_rate": 9.244039159450334e-06, + "loss": 0.3304, + "step": 9177 + }, + { + "epoch": 0.2020250935505173, + "grad_norm": 1.6046574130422897, + "learning_rate": 9.243850689646909e-06, + "loss": 0.3904, + "step": 9178 + }, + { + "epoch": 0.20204710543693594, + "grad_norm": 1.292618229420118, + "learning_rate": 9.243662198274345e-06, + "loss": 0.3331, + "step": 9179 + }, + { + "epoch": 0.2020691173233546, + "grad_norm": 0.8236844964350778, + "learning_rate": 9.243473685333595e-06, + "loss": 0.5477, + "step": 9180 + }, + { + "epoch": 0.2020911292097733, + "grad_norm": 1.3317390666425486, + "learning_rate": 9.24328515082562e-06, + "loss": 0.3838, + "step": 9181 + }, + { + "epoch": 0.20211314109619194, + "grad_norm": 1.4573067398573238, + "learning_rate": 9.243096594751377e-06, + "loss": 0.3597, + "step": 9182 + }, + { + "epoch": 0.20213515298261062, + "grad_norm": 1.533910956452817, + "learning_rate": 9.242908017111827e-06, + "loss": 0.3634, + "step": 9183 + }, + { + "epoch": 0.20215716486902927, + "grad_norm": 1.39133157018367, + "learning_rate": 9.242719417907922e-06, + "loss": 0.3184, + "step": 9184 + }, + { + "epoch": 0.20217917675544794, + "grad_norm": 1.2797608430975471, + "learning_rate": 9.242530797140627e-06, + "loss": 0.2888, + "step": 9185 + }, + { + "epoch": 0.20220118864186662, + "grad_norm": 1.3929913623741608, + "learning_rate": 9.242342154810898e-06, + "loss": 0.2786, + "step": 9186 + }, + { + "epoch": 0.20222320052828527, + "grad_norm": 1.428004522225961, + "learning_rate": 9.242153490919694e-06, + "loss": 0.3646, + "step": 9187 + }, + { + "epoch": 0.20224521241470395, + "grad_norm": 1.2578376346635993, + "learning_rate": 9.241964805467974e-06, + "loss": 0.3672, + "step": 9188 + }, + { + "epoch": 0.2022672243011226, + "grad_norm": 1.5198857712702916, + "learning_rate": 9.241776098456697e-06, + "loss": 0.332, + "step": 9189 + }, + { + "epoch": 0.20228923618754127, + "grad_norm": 1.5834063812004937, + "learning_rate": 9.241587369886822e-06, + "loss": 0.3374, + "step": 9190 + }, + { + "epoch": 0.20231124807395995, + "grad_norm": 3.1584539214754916, + "learning_rate": 9.241398619759307e-06, + "loss": 0.3715, + "step": 9191 + }, + { + "epoch": 0.2023332599603786, + "grad_norm": 1.219382309358059, + "learning_rate": 9.241209848075115e-06, + "loss": 0.2715, + "step": 9192 + }, + { + "epoch": 0.20235527184679727, + "grad_norm": 0.7309569051871795, + "learning_rate": 9.2410210548352e-06, + "loss": 0.5317, + "step": 9193 + }, + { + "epoch": 0.20237728373321592, + "grad_norm": 4.483538980880892, + "learning_rate": 9.240832240040525e-06, + "loss": 0.4222, + "step": 9194 + }, + { + "epoch": 0.2023992956196346, + "grad_norm": 0.6589013776738234, + "learning_rate": 9.24064340369205e-06, + "loss": 0.5361, + "step": 9195 + }, + { + "epoch": 0.20242130750605328, + "grad_norm": 1.6139934472299073, + "learning_rate": 9.240454545790733e-06, + "loss": 0.3864, + "step": 9196 + }, + { + "epoch": 0.20244331939247193, + "grad_norm": 1.7524583087457837, + "learning_rate": 9.240265666337536e-06, + "loss": 0.2512, + "step": 9197 + }, + { + "epoch": 0.2024653312788906, + "grad_norm": 1.3674416769804483, + "learning_rate": 9.240076765333417e-06, + "loss": 0.3032, + "step": 9198 + }, + { + "epoch": 0.20248734316530928, + "grad_norm": 19.931081494981466, + "learning_rate": 9.239887842779338e-06, + "loss": 0.4022, + "step": 9199 + }, + { + "epoch": 0.20250935505172793, + "grad_norm": 1.4739982230820403, + "learning_rate": 9.239698898676257e-06, + "loss": 0.3277, + "step": 9200 + }, + { + "epoch": 0.2025313669381466, + "grad_norm": 1.5715499948399414, + "learning_rate": 9.239509933025135e-06, + "loss": 0.3517, + "step": 9201 + }, + { + "epoch": 0.20255337882456526, + "grad_norm": 1.195779829353383, + "learning_rate": 9.239320945826931e-06, + "loss": 0.2822, + "step": 9202 + }, + { + "epoch": 0.20257539071098393, + "grad_norm": 1.1742394880459854, + "learning_rate": 9.23913193708261e-06, + "loss": 0.3282, + "step": 9203 + }, + { + "epoch": 0.2025974025974026, + "grad_norm": 1.4981794892106302, + "learning_rate": 9.238942906793128e-06, + "loss": 0.3841, + "step": 9204 + }, + { + "epoch": 0.20261941448382126, + "grad_norm": 1.6697507063967232, + "learning_rate": 9.238753854959448e-06, + "loss": 0.3599, + "step": 9205 + }, + { + "epoch": 0.20264142637023994, + "grad_norm": 2.46723379696156, + "learning_rate": 9.23856478158253e-06, + "loss": 0.2029, + "step": 9206 + }, + { + "epoch": 0.20266343825665858, + "grad_norm": 1.8988000438046824, + "learning_rate": 9.238375686663337e-06, + "loss": 0.3562, + "step": 9207 + }, + { + "epoch": 0.20268545014307726, + "grad_norm": 1.8235782181519127, + "learning_rate": 9.238186570202828e-06, + "loss": 0.3252, + "step": 9208 + }, + { + "epoch": 0.20270746202949594, + "grad_norm": 1.0931370443221518, + "learning_rate": 9.237997432201962e-06, + "loss": 0.5623, + "step": 9209 + }, + { + "epoch": 0.2027294739159146, + "grad_norm": 2.7233695404659306, + "learning_rate": 9.237808272661705e-06, + "loss": 0.3927, + "step": 9210 + }, + { + "epoch": 0.20275148580233326, + "grad_norm": 2.6449049054304608, + "learning_rate": 9.237619091583015e-06, + "loss": 0.3823, + "step": 9211 + }, + { + "epoch": 0.2027734976887519, + "grad_norm": 0.6514454304741413, + "learning_rate": 9.237429888966855e-06, + "loss": 0.5427, + "step": 9212 + }, + { + "epoch": 0.2027955095751706, + "grad_norm": 1.4016406425496328, + "learning_rate": 9.23724066481419e-06, + "loss": 0.3804, + "step": 9213 + }, + { + "epoch": 0.20281752146158927, + "grad_norm": 1.3503353527712492, + "learning_rate": 9.237051419125973e-06, + "loss": 0.3875, + "step": 9214 + }, + { + "epoch": 0.20283953334800792, + "grad_norm": 1.5682329055822577, + "learning_rate": 9.236862151903173e-06, + "loss": 0.3387, + "step": 9215 + }, + { + "epoch": 0.2028615452344266, + "grad_norm": 2.408239084750186, + "learning_rate": 9.236672863146749e-06, + "loss": 0.3834, + "step": 9216 + }, + { + "epoch": 0.20288355712084524, + "grad_norm": 1.5107761280745473, + "learning_rate": 9.236483552857665e-06, + "loss": 0.4034, + "step": 9217 + }, + { + "epoch": 0.20290556900726392, + "grad_norm": 1.0331852565725588, + "learning_rate": 9.236294221036882e-06, + "loss": 0.5304, + "step": 9218 + }, + { + "epoch": 0.2029275808936826, + "grad_norm": 1.6834462607520189, + "learning_rate": 9.236104867685363e-06, + "loss": 0.3745, + "step": 9219 + }, + { + "epoch": 0.20294959278010125, + "grad_norm": 1.3754811109023042, + "learning_rate": 9.235915492804069e-06, + "loss": 0.3432, + "step": 9220 + }, + { + "epoch": 0.20297160466651992, + "grad_norm": 0.7198243100429609, + "learning_rate": 9.235726096393962e-06, + "loss": 0.532, + "step": 9221 + }, + { + "epoch": 0.2029936165529386, + "grad_norm": 1.977163347236111, + "learning_rate": 9.235536678456008e-06, + "loss": 0.3535, + "step": 9222 + }, + { + "epoch": 0.20301562843935725, + "grad_norm": 1.4611123539551767, + "learning_rate": 9.235347238991166e-06, + "loss": 0.3483, + "step": 9223 + }, + { + "epoch": 0.20303764032577593, + "grad_norm": 1.3523774234764372, + "learning_rate": 9.235157778000399e-06, + "loss": 0.3385, + "step": 9224 + }, + { + "epoch": 0.20305965221219457, + "grad_norm": 1.2954348451824966, + "learning_rate": 9.234968295484673e-06, + "loss": 0.3178, + "step": 9225 + }, + { + "epoch": 0.20308166409861325, + "grad_norm": 1.45833777227205, + "learning_rate": 9.23477879144495e-06, + "loss": 0.356, + "step": 9226 + }, + { + "epoch": 0.20310367598503193, + "grad_norm": 0.8821048163149342, + "learning_rate": 9.234589265882192e-06, + "loss": 0.5548, + "step": 9227 + }, + { + "epoch": 0.20312568787145058, + "grad_norm": 0.8308237512546884, + "learning_rate": 9.234399718797363e-06, + "loss": 0.564, + "step": 9228 + }, + { + "epoch": 0.20314769975786925, + "grad_norm": 0.7270307407349493, + "learning_rate": 9.234210150191423e-06, + "loss": 0.5219, + "step": 9229 + }, + { + "epoch": 0.2031697116442879, + "grad_norm": 1.2977366906115764, + "learning_rate": 9.23402056006534e-06, + "loss": 0.3011, + "step": 9230 + }, + { + "epoch": 0.20319172353070658, + "grad_norm": 1.2360080281523325, + "learning_rate": 9.233830948420079e-06, + "loss": 0.3227, + "step": 9231 + }, + { + "epoch": 0.20321373541712526, + "grad_norm": 1.7301733889832742, + "learning_rate": 9.233641315256598e-06, + "loss": 0.3865, + "step": 9232 + }, + { + "epoch": 0.2032357473035439, + "grad_norm": 0.9309577682444641, + "learning_rate": 9.233451660575864e-06, + "loss": 0.5638, + "step": 9233 + }, + { + "epoch": 0.20325775918996258, + "grad_norm": 1.2500460052946756, + "learning_rate": 9.23326198437884e-06, + "loss": 0.3667, + "step": 9234 + }, + { + "epoch": 0.20327977107638123, + "grad_norm": 1.436333005322422, + "learning_rate": 9.233072286666491e-06, + "loss": 0.3252, + "step": 9235 + }, + { + "epoch": 0.2033017829627999, + "grad_norm": 1.7326635563968207, + "learning_rate": 9.232882567439782e-06, + "loss": 0.4093, + "step": 9236 + }, + { + "epoch": 0.2033237948492186, + "grad_norm": 0.9082226186319126, + "learning_rate": 9.232692826699674e-06, + "loss": 0.5523, + "step": 9237 + }, + { + "epoch": 0.20334580673563724, + "grad_norm": 1.5493088438445803, + "learning_rate": 9.232503064447134e-06, + "loss": 0.367, + "step": 9238 + }, + { + "epoch": 0.2033678186220559, + "grad_norm": 1.2943998057026938, + "learning_rate": 9.232313280683126e-06, + "loss": 0.3091, + "step": 9239 + }, + { + "epoch": 0.2033898305084746, + "grad_norm": 0.676625058723299, + "learning_rate": 9.232123475408614e-06, + "loss": 0.5495, + "step": 9240 + }, + { + "epoch": 0.20341184239489324, + "grad_norm": 2.4652998897396987, + "learning_rate": 9.231933648624562e-06, + "loss": 0.397, + "step": 9241 + }, + { + "epoch": 0.20343385428131192, + "grad_norm": 1.395589798615387, + "learning_rate": 9.231743800331936e-06, + "loss": 0.3917, + "step": 9242 + }, + { + "epoch": 0.20345586616773056, + "grad_norm": 1.259886547695398, + "learning_rate": 9.231553930531701e-06, + "loss": 0.3617, + "step": 9243 + }, + { + "epoch": 0.20347787805414924, + "grad_norm": 2.2963018564753233, + "learning_rate": 9.231364039224821e-06, + "loss": 0.4149, + "step": 9244 + }, + { + "epoch": 0.20349988994056792, + "grad_norm": 1.4035123624974797, + "learning_rate": 9.231174126412264e-06, + "loss": 0.3924, + "step": 9245 + }, + { + "epoch": 0.20352190182698657, + "grad_norm": 2.4785143523296496, + "learning_rate": 9.230984192094992e-06, + "loss": 0.366, + "step": 9246 + }, + { + "epoch": 0.20354391371340524, + "grad_norm": 1.5977828498891984, + "learning_rate": 9.23079423627397e-06, + "loss": 0.422, + "step": 9247 + }, + { + "epoch": 0.2035659255998239, + "grad_norm": 1.4700163655303415, + "learning_rate": 9.230604258950164e-06, + "loss": 0.3827, + "step": 9248 + }, + { + "epoch": 0.20358793748624257, + "grad_norm": 4.638973011336575, + "learning_rate": 9.230414260124542e-06, + "loss": 0.3642, + "step": 9249 + }, + { + "epoch": 0.20360994937266125, + "grad_norm": 0.9358442065024332, + "learning_rate": 9.230224239798067e-06, + "loss": 0.5414, + "step": 9250 + }, + { + "epoch": 0.2036319612590799, + "grad_norm": 0.8481452903204783, + "learning_rate": 9.230034197971706e-06, + "loss": 0.58, + "step": 9251 + }, + { + "epoch": 0.20365397314549857, + "grad_norm": 1.5276328336049314, + "learning_rate": 9.229844134646425e-06, + "loss": 0.4135, + "step": 9252 + }, + { + "epoch": 0.20367598503191722, + "grad_norm": 1.3618808493556476, + "learning_rate": 9.22965404982319e-06, + "loss": 0.2915, + "step": 9253 + }, + { + "epoch": 0.2036979969183359, + "grad_norm": 1.2928210946829253, + "learning_rate": 9.229463943502965e-06, + "loss": 0.2807, + "step": 9254 + }, + { + "epoch": 0.20372000880475458, + "grad_norm": 2.637381556246247, + "learning_rate": 9.229273815686719e-06, + "loss": 0.3051, + "step": 9255 + }, + { + "epoch": 0.20374202069117323, + "grad_norm": 1.6152735963344949, + "learning_rate": 9.229083666375417e-06, + "loss": 0.367, + "step": 9256 + }, + { + "epoch": 0.2037640325775919, + "grad_norm": 1.3642604889296817, + "learning_rate": 9.228893495570026e-06, + "loss": 0.3059, + "step": 9257 + }, + { + "epoch": 0.20378604446401058, + "grad_norm": 1.170092346280321, + "learning_rate": 9.22870330327151e-06, + "loss": 0.2769, + "step": 9258 + }, + { + "epoch": 0.20380805635042923, + "grad_norm": 1.90465939193954, + "learning_rate": 9.228513089480839e-06, + "loss": 0.3776, + "step": 9259 + }, + { + "epoch": 0.2038300682368479, + "grad_norm": 1.5572873100511564, + "learning_rate": 9.22832285419898e-06, + "loss": 0.376, + "step": 9260 + }, + { + "epoch": 0.20385208012326655, + "grad_norm": 1.4038510452808572, + "learning_rate": 9.228132597426898e-06, + "loss": 0.5681, + "step": 9261 + }, + { + "epoch": 0.20387409200968523, + "grad_norm": 1.3814413235318432, + "learning_rate": 9.227942319165559e-06, + "loss": 0.2917, + "step": 9262 + }, + { + "epoch": 0.2038961038961039, + "grad_norm": 2.1141171784543924, + "learning_rate": 9.227752019415932e-06, + "loss": 0.4342, + "step": 9263 + }, + { + "epoch": 0.20391811578252256, + "grad_norm": 1.5442354795575088, + "learning_rate": 9.227561698178983e-06, + "loss": 0.366, + "step": 9264 + }, + { + "epoch": 0.20394012766894123, + "grad_norm": 1.5450468011491778, + "learning_rate": 9.22737135545568e-06, + "loss": 0.3787, + "step": 9265 + }, + { + "epoch": 0.20396213955535988, + "grad_norm": 1.4941528768433334, + "learning_rate": 9.227180991246992e-06, + "loss": 0.398, + "step": 9266 + }, + { + "epoch": 0.20398415144177856, + "grad_norm": 1.4600377963192452, + "learning_rate": 9.226990605553882e-06, + "loss": 0.3596, + "step": 9267 + }, + { + "epoch": 0.20400616332819724, + "grad_norm": 1.4251611099933694, + "learning_rate": 9.226800198377323e-06, + "loss": 0.3481, + "step": 9268 + }, + { + "epoch": 0.2040281752146159, + "grad_norm": 1.423249565343289, + "learning_rate": 9.226609769718279e-06, + "loss": 0.413, + "step": 9269 + }, + { + "epoch": 0.20405018710103456, + "grad_norm": 1.3703918119287766, + "learning_rate": 9.22641931957772e-06, + "loss": 0.2837, + "step": 9270 + }, + { + "epoch": 0.2040721989874532, + "grad_norm": 1.8521848988035625, + "learning_rate": 9.226228847956613e-06, + "loss": 0.4236, + "step": 9271 + }, + { + "epoch": 0.2040942108738719, + "grad_norm": 0.8017280965219435, + "learning_rate": 9.226038354855924e-06, + "loss": 0.5246, + "step": 9272 + }, + { + "epoch": 0.20411622276029057, + "grad_norm": 1.4115134984052014, + "learning_rate": 9.225847840276625e-06, + "loss": 0.3379, + "step": 9273 + }, + { + "epoch": 0.20413823464670922, + "grad_norm": 1.0848684583902053, + "learning_rate": 9.225657304219682e-06, + "loss": 0.2687, + "step": 9274 + }, + { + "epoch": 0.2041602465331279, + "grad_norm": 4.187771951869323, + "learning_rate": 9.225466746686062e-06, + "loss": 0.3342, + "step": 9275 + }, + { + "epoch": 0.20418225841954654, + "grad_norm": 1.0772257882258338, + "learning_rate": 9.225276167676737e-06, + "loss": 0.2921, + "step": 9276 + }, + { + "epoch": 0.20420427030596522, + "grad_norm": 2.7222228450124364, + "learning_rate": 9.225085567192675e-06, + "loss": 0.3791, + "step": 9277 + }, + { + "epoch": 0.2042262821923839, + "grad_norm": 1.540363590677834, + "learning_rate": 9.224894945234844e-06, + "loss": 0.3648, + "step": 9278 + }, + { + "epoch": 0.20424829407880254, + "grad_norm": 1.3949652124008995, + "learning_rate": 9.22470430180421e-06, + "loss": 0.3949, + "step": 9279 + }, + { + "epoch": 0.20427030596522122, + "grad_norm": 1.9424704664364276, + "learning_rate": 9.224513636901747e-06, + "loss": 0.328, + "step": 9280 + }, + { + "epoch": 0.2042923178516399, + "grad_norm": 1.4953093222426843, + "learning_rate": 9.224322950528421e-06, + "loss": 0.3809, + "step": 9281 + }, + { + "epoch": 0.20431432973805855, + "grad_norm": 1.3017086195847523, + "learning_rate": 9.2241322426852e-06, + "loss": 0.2761, + "step": 9282 + }, + { + "epoch": 0.20433634162447722, + "grad_norm": 1.5276175528604945, + "learning_rate": 9.223941513373056e-06, + "loss": 0.4161, + "step": 9283 + }, + { + "epoch": 0.20435835351089587, + "grad_norm": 1.7824758514521637, + "learning_rate": 9.223750762592957e-06, + "loss": 0.3367, + "step": 9284 + }, + { + "epoch": 0.20438036539731455, + "grad_norm": 1.6082591902841343, + "learning_rate": 9.223559990345874e-06, + "loss": 0.3806, + "step": 9285 + }, + { + "epoch": 0.20440237728373323, + "grad_norm": 1.5988312199247783, + "learning_rate": 9.223369196632774e-06, + "loss": 0.3719, + "step": 9286 + }, + { + "epoch": 0.20442438917015188, + "grad_norm": 1.2213123960942482, + "learning_rate": 9.22317838145463e-06, + "loss": 0.4061, + "step": 9287 + }, + { + "epoch": 0.20444640105657055, + "grad_norm": 0.6492023705920619, + "learning_rate": 9.222987544812407e-06, + "loss": 0.5335, + "step": 9288 + }, + { + "epoch": 0.2044684129429892, + "grad_norm": 9.33879029333829, + "learning_rate": 9.222796686707081e-06, + "loss": 0.3288, + "step": 9289 + }, + { + "epoch": 0.20449042482940788, + "grad_norm": 2.517648011078084, + "learning_rate": 9.222605807139616e-06, + "loss": 0.296, + "step": 9290 + }, + { + "epoch": 0.20451243671582656, + "grad_norm": 1.6636042402824223, + "learning_rate": 9.222414906110986e-06, + "loss": 0.3781, + "step": 9291 + }, + { + "epoch": 0.2045344486022452, + "grad_norm": 1.500883884654763, + "learning_rate": 9.22222398362216e-06, + "loss": 0.3476, + "step": 9292 + }, + { + "epoch": 0.20455646048866388, + "grad_norm": 1.7251995307877928, + "learning_rate": 9.22203303967411e-06, + "loss": 0.4437, + "step": 9293 + }, + { + "epoch": 0.20457847237508253, + "grad_norm": 1.2943920863171614, + "learning_rate": 9.221842074267803e-06, + "loss": 0.3701, + "step": 9294 + }, + { + "epoch": 0.2046004842615012, + "grad_norm": 0.6398546799498204, + "learning_rate": 9.221651087404211e-06, + "loss": 0.5267, + "step": 9295 + }, + { + "epoch": 0.20462249614791989, + "grad_norm": 0.6220823379546349, + "learning_rate": 9.221460079084307e-06, + "loss": 0.5513, + "step": 9296 + }, + { + "epoch": 0.20464450803433853, + "grad_norm": 0.6413019878397755, + "learning_rate": 9.22126904930906e-06, + "loss": 0.554, + "step": 9297 + }, + { + "epoch": 0.2046665199207572, + "grad_norm": 1.1586639115263375, + "learning_rate": 9.221077998079441e-06, + "loss": 0.344, + "step": 9298 + }, + { + "epoch": 0.2046885318071759, + "grad_norm": 1.2413009315825385, + "learning_rate": 9.22088692539642e-06, + "loss": 0.3006, + "step": 9299 + }, + { + "epoch": 0.20471054369359454, + "grad_norm": 1.9846232434888722, + "learning_rate": 9.220695831260971e-06, + "loss": 0.4016, + "step": 9300 + }, + { + "epoch": 0.20473255558001321, + "grad_norm": 1.2535091487778047, + "learning_rate": 9.220504715674062e-06, + "loss": 0.367, + "step": 9301 + }, + { + "epoch": 0.20475456746643186, + "grad_norm": 0.638033530825221, + "learning_rate": 9.220313578636664e-06, + "loss": 0.4882, + "step": 9302 + }, + { + "epoch": 0.20477657935285054, + "grad_norm": 1.69190444673368, + "learning_rate": 9.220122420149753e-06, + "loss": 0.3397, + "step": 9303 + }, + { + "epoch": 0.20479859123926922, + "grad_norm": 2.023766036601725, + "learning_rate": 9.219931240214296e-06, + "loss": 0.3768, + "step": 9304 + }, + { + "epoch": 0.20482060312568787, + "grad_norm": 1.5240014589661501, + "learning_rate": 9.219740038831267e-06, + "loss": 0.4114, + "step": 9305 + }, + { + "epoch": 0.20484261501210654, + "grad_norm": 2.363168751012225, + "learning_rate": 9.219548816001636e-06, + "loss": 0.2394, + "step": 9306 + }, + { + "epoch": 0.2048646268985252, + "grad_norm": 2.914677764980897, + "learning_rate": 9.219357571726378e-06, + "loss": 0.3418, + "step": 9307 + }, + { + "epoch": 0.20488663878494387, + "grad_norm": 1.316458353249728, + "learning_rate": 9.21916630600646e-06, + "loss": 0.2992, + "step": 9308 + }, + { + "epoch": 0.20490865067136255, + "grad_norm": 1.5702470976807592, + "learning_rate": 9.218975018842859e-06, + "loss": 0.3282, + "step": 9309 + }, + { + "epoch": 0.2049306625577812, + "grad_norm": 1.3638385506654707, + "learning_rate": 9.218783710236546e-06, + "loss": 0.3364, + "step": 9310 + }, + { + "epoch": 0.20495267444419987, + "grad_norm": 1.7131949607834405, + "learning_rate": 9.21859238018849e-06, + "loss": 0.3651, + "step": 9311 + }, + { + "epoch": 0.20497468633061852, + "grad_norm": 1.2686081617665452, + "learning_rate": 9.21840102869967e-06, + "loss": 0.323, + "step": 9312 + }, + { + "epoch": 0.2049966982170372, + "grad_norm": 2.2713364602040422, + "learning_rate": 9.218209655771048e-06, + "loss": 0.3628, + "step": 9313 + }, + { + "epoch": 0.20501871010345588, + "grad_norm": 1.8901303785932813, + "learning_rate": 9.218018261403608e-06, + "loss": 0.4091, + "step": 9314 + }, + { + "epoch": 0.20504072198987452, + "grad_norm": 1.5291148279201487, + "learning_rate": 9.217826845598317e-06, + "loss": 0.358, + "step": 9315 + }, + { + "epoch": 0.2050627338762932, + "grad_norm": 1.7483031461874599, + "learning_rate": 9.21763540835615e-06, + "loss": 0.3184, + "step": 9316 + }, + { + "epoch": 0.20508474576271185, + "grad_norm": 2.259707344027754, + "learning_rate": 9.217443949678077e-06, + "loss": 0.4101, + "step": 9317 + }, + { + "epoch": 0.20510675764913053, + "grad_norm": 1.8497132428748535, + "learning_rate": 9.217252469565072e-06, + "loss": 0.3951, + "step": 9318 + }, + { + "epoch": 0.2051287695355492, + "grad_norm": 2.030321125759106, + "learning_rate": 9.217060968018111e-06, + "loss": 0.3082, + "step": 9319 + }, + { + "epoch": 0.20515078142196785, + "grad_norm": 1.401678784195127, + "learning_rate": 9.216869445038165e-06, + "loss": 0.3609, + "step": 9320 + }, + { + "epoch": 0.20517279330838653, + "grad_norm": 1.1917104999635424, + "learning_rate": 9.216677900626207e-06, + "loss": 0.2882, + "step": 9321 + }, + { + "epoch": 0.2051948051948052, + "grad_norm": 1.5649323511419226, + "learning_rate": 9.216486334783212e-06, + "loss": 0.3369, + "step": 9322 + }, + { + "epoch": 0.20521681708122386, + "grad_norm": 1.4577907894628352, + "learning_rate": 9.216294747510151e-06, + "loss": 0.3344, + "step": 9323 + }, + { + "epoch": 0.20523882896764253, + "grad_norm": 1.413033048609211, + "learning_rate": 9.216103138808002e-06, + "loss": 0.4091, + "step": 9324 + }, + { + "epoch": 0.20526084085406118, + "grad_norm": 4.127402798136952, + "learning_rate": 9.215911508677735e-06, + "loss": 0.3278, + "step": 9325 + }, + { + "epoch": 0.20528285274047986, + "grad_norm": 0.7089546520267933, + "learning_rate": 9.215719857120326e-06, + "loss": 0.5293, + "step": 9326 + }, + { + "epoch": 0.20530486462689854, + "grad_norm": 1.3136228653073958, + "learning_rate": 9.215528184136747e-06, + "loss": 0.3619, + "step": 9327 + }, + { + "epoch": 0.20532687651331719, + "grad_norm": 1.9644108695749105, + "learning_rate": 9.215336489727975e-06, + "loss": 0.4136, + "step": 9328 + }, + { + "epoch": 0.20534888839973586, + "grad_norm": 0.6508033399223441, + "learning_rate": 9.215144773894982e-06, + "loss": 0.5335, + "step": 9329 + }, + { + "epoch": 0.2053709002861545, + "grad_norm": 2.02281277883726, + "learning_rate": 9.214953036638746e-06, + "loss": 0.3359, + "step": 9330 + }, + { + "epoch": 0.2053929121725732, + "grad_norm": 6.5274587370570325, + "learning_rate": 9.214761277960235e-06, + "loss": 0.3299, + "step": 9331 + }, + { + "epoch": 0.20541492405899187, + "grad_norm": 1.6379856527946428, + "learning_rate": 9.214569497860428e-06, + "loss": 0.3304, + "step": 9332 + }, + { + "epoch": 0.20543693594541051, + "grad_norm": 3.2207667911159183, + "learning_rate": 9.214377696340302e-06, + "loss": 0.3194, + "step": 9333 + }, + { + "epoch": 0.2054589478318292, + "grad_norm": 1.611874564272194, + "learning_rate": 9.214185873400827e-06, + "loss": 0.3923, + "step": 9334 + }, + { + "epoch": 0.20548095971824784, + "grad_norm": 0.7178759223299275, + "learning_rate": 9.213994029042978e-06, + "loss": 0.5222, + "step": 9335 + }, + { + "epoch": 0.20550297160466652, + "grad_norm": 1.3576484723111621, + "learning_rate": 9.213802163267734e-06, + "loss": 0.3207, + "step": 9336 + }, + { + "epoch": 0.2055249834910852, + "grad_norm": 1.487307185593689, + "learning_rate": 9.213610276076068e-06, + "loss": 0.3814, + "step": 9337 + }, + { + "epoch": 0.20554699537750384, + "grad_norm": 1.5375525339629115, + "learning_rate": 9.213418367468954e-06, + "loss": 0.3217, + "step": 9338 + }, + { + "epoch": 0.20556900726392252, + "grad_norm": 2.0159730650212704, + "learning_rate": 9.21322643744737e-06, + "loss": 0.3461, + "step": 9339 + }, + { + "epoch": 0.2055910191503412, + "grad_norm": 1.391607509917943, + "learning_rate": 9.213034486012288e-06, + "loss": 0.3612, + "step": 9340 + }, + { + "epoch": 0.20561303103675985, + "grad_norm": 0.6357893072518251, + "learning_rate": 9.212842513164688e-06, + "loss": 0.5284, + "step": 9341 + }, + { + "epoch": 0.20563504292317852, + "grad_norm": 1.402929214748522, + "learning_rate": 9.212650518905544e-06, + "loss": 0.3411, + "step": 9342 + }, + { + "epoch": 0.20565705480959717, + "grad_norm": 2.0876964158813576, + "learning_rate": 9.212458503235829e-06, + "loss": 0.3366, + "step": 9343 + }, + { + "epoch": 0.20567906669601585, + "grad_norm": 1.8901269324306584, + "learning_rate": 9.21226646615652e-06, + "loss": 0.4268, + "step": 9344 + }, + { + "epoch": 0.20570107858243453, + "grad_norm": 1.4926700167315765, + "learning_rate": 9.212074407668596e-06, + "loss": 0.3914, + "step": 9345 + }, + { + "epoch": 0.20572309046885318, + "grad_norm": 2.4819498099115815, + "learning_rate": 9.211882327773033e-06, + "loss": 0.2926, + "step": 9346 + }, + { + "epoch": 0.20574510235527185, + "grad_norm": 1.318016523024782, + "learning_rate": 9.211690226470802e-06, + "loss": 0.2842, + "step": 9347 + }, + { + "epoch": 0.2057671142416905, + "grad_norm": 1.7384836730912192, + "learning_rate": 9.211498103762886e-06, + "loss": 0.4002, + "step": 9348 + }, + { + "epoch": 0.20578912612810918, + "grad_norm": 1.956961933673367, + "learning_rate": 9.211305959650257e-06, + "loss": 0.3819, + "step": 9349 + }, + { + "epoch": 0.20581113801452786, + "grad_norm": 1.4233893068584373, + "learning_rate": 9.211113794133891e-06, + "loss": 0.3637, + "step": 9350 + }, + { + "epoch": 0.2058331499009465, + "grad_norm": 2.1419633552222344, + "learning_rate": 9.210921607214768e-06, + "loss": 0.3708, + "step": 9351 + }, + { + "epoch": 0.20585516178736518, + "grad_norm": 1.7489786531582256, + "learning_rate": 9.210729398893864e-06, + "loss": 0.3041, + "step": 9352 + }, + { + "epoch": 0.20587717367378383, + "grad_norm": 1.3294454129025162, + "learning_rate": 9.210537169172153e-06, + "loss": 0.2727, + "step": 9353 + }, + { + "epoch": 0.2058991855602025, + "grad_norm": 0.6869022598761162, + "learning_rate": 9.210344918050615e-06, + "loss": 0.5522, + "step": 9354 + }, + { + "epoch": 0.20592119744662118, + "grad_norm": 1.1813901035732834, + "learning_rate": 9.210152645530229e-06, + "loss": 0.2618, + "step": 9355 + }, + { + "epoch": 0.20594320933303983, + "grad_norm": 0.6489249356052654, + "learning_rate": 9.209960351611966e-06, + "loss": 0.5329, + "step": 9356 + }, + { + "epoch": 0.2059652212194585, + "grad_norm": 0.651203358184879, + "learning_rate": 9.209768036296807e-06, + "loss": 0.5674, + "step": 9357 + }, + { + "epoch": 0.20598723310587716, + "grad_norm": 1.4278620454891076, + "learning_rate": 9.209575699585732e-06, + "loss": 0.3491, + "step": 9358 + }, + { + "epoch": 0.20600924499229584, + "grad_norm": 1.7532131019969006, + "learning_rate": 9.209383341479714e-06, + "loss": 0.3488, + "step": 9359 + }, + { + "epoch": 0.2060312568787145, + "grad_norm": 2.321651883615816, + "learning_rate": 9.209190961979732e-06, + "loss": 0.3112, + "step": 9360 + }, + { + "epoch": 0.20605326876513316, + "grad_norm": 1.4080231624334334, + "learning_rate": 9.208998561086765e-06, + "loss": 0.2979, + "step": 9361 + }, + { + "epoch": 0.20607528065155184, + "grad_norm": 4.0652632279432215, + "learning_rate": 9.208806138801789e-06, + "loss": 0.284, + "step": 9362 + }, + { + "epoch": 0.20609729253797052, + "grad_norm": 1.8201857863941049, + "learning_rate": 9.208613695125783e-06, + "loss": 0.3808, + "step": 9363 + }, + { + "epoch": 0.20611930442438917, + "grad_norm": 2.401796057333509, + "learning_rate": 9.208421230059726e-06, + "loss": 0.3548, + "step": 9364 + }, + { + "epoch": 0.20614131631080784, + "grad_norm": 1.7346868910755109, + "learning_rate": 9.208228743604596e-06, + "loss": 0.3583, + "step": 9365 + }, + { + "epoch": 0.2061633281972265, + "grad_norm": 1.2220693147650161, + "learning_rate": 9.208036235761369e-06, + "loss": 0.3269, + "step": 9366 + }, + { + "epoch": 0.20618534008364517, + "grad_norm": 1.4509957511594596, + "learning_rate": 9.207843706531026e-06, + "loss": 0.3457, + "step": 9367 + }, + { + "epoch": 0.20620735197006385, + "grad_norm": 1.2991704126484784, + "learning_rate": 9.207651155914545e-06, + "loss": 0.304, + "step": 9368 + }, + { + "epoch": 0.2062293638564825, + "grad_norm": 1.5843220898225396, + "learning_rate": 9.207458583912902e-06, + "loss": 0.3549, + "step": 9369 + }, + { + "epoch": 0.20625137574290117, + "grad_norm": 1.5786028524467912, + "learning_rate": 9.20726599052708e-06, + "loss": 0.4002, + "step": 9370 + }, + { + "epoch": 0.20627338762931982, + "grad_norm": 1.0284088378502454, + "learning_rate": 9.207073375758056e-06, + "loss": 0.2854, + "step": 9371 + }, + { + "epoch": 0.2062953995157385, + "grad_norm": 1.6069792053321332, + "learning_rate": 9.206880739606806e-06, + "loss": 0.36, + "step": 9372 + }, + { + "epoch": 0.20631741140215717, + "grad_norm": 1.9938202482207774, + "learning_rate": 9.206688082074314e-06, + "loss": 0.3562, + "step": 9373 + }, + { + "epoch": 0.20633942328857582, + "grad_norm": 1.3505874267653017, + "learning_rate": 9.206495403161557e-06, + "loss": 0.3224, + "step": 9374 + }, + { + "epoch": 0.2063614351749945, + "grad_norm": 1.5706023735005716, + "learning_rate": 9.206302702869514e-06, + "loss": 0.3596, + "step": 9375 + }, + { + "epoch": 0.20638344706141315, + "grad_norm": 1.305208190223973, + "learning_rate": 9.206109981199162e-06, + "loss": 0.3767, + "step": 9376 + }, + { + "epoch": 0.20640545894783183, + "grad_norm": 1.7850367633227286, + "learning_rate": 9.205917238151488e-06, + "loss": 0.3131, + "step": 9377 + }, + { + "epoch": 0.2064274708342505, + "grad_norm": 1.195744832825265, + "learning_rate": 9.205724473727463e-06, + "loss": 0.3255, + "step": 9378 + }, + { + "epoch": 0.20644948272066915, + "grad_norm": 1.3790864825023665, + "learning_rate": 9.20553168792807e-06, + "loss": 0.2941, + "step": 9379 + }, + { + "epoch": 0.20647149460708783, + "grad_norm": 1.832365052993027, + "learning_rate": 9.205338880754292e-06, + "loss": 0.377, + "step": 9380 + }, + { + "epoch": 0.2064935064935065, + "grad_norm": 1.4775539565917857, + "learning_rate": 9.205146052207106e-06, + "loss": 0.3337, + "step": 9381 + }, + { + "epoch": 0.20651551837992516, + "grad_norm": 1.8447844507874402, + "learning_rate": 9.20495320228749e-06, + "loss": 0.3843, + "step": 9382 + }, + { + "epoch": 0.20653753026634383, + "grad_norm": 1.5823467092459573, + "learning_rate": 9.204760330996428e-06, + "loss": 0.3614, + "step": 9383 + }, + { + "epoch": 0.20655954215276248, + "grad_norm": 1.541298297199467, + "learning_rate": 9.204567438334896e-06, + "loss": 0.2499, + "step": 9384 + }, + { + "epoch": 0.20658155403918116, + "grad_norm": 3.2365161702990948, + "learning_rate": 9.20437452430388e-06, + "loss": 0.5155, + "step": 9385 + }, + { + "epoch": 0.20660356592559984, + "grad_norm": 1.6554334446766958, + "learning_rate": 9.204181588904358e-06, + "loss": 0.3206, + "step": 9386 + }, + { + "epoch": 0.20662557781201848, + "grad_norm": 1.5734285203945404, + "learning_rate": 9.203988632137308e-06, + "loss": 0.3784, + "step": 9387 + }, + { + "epoch": 0.20664758969843716, + "grad_norm": 3.0405990283871187, + "learning_rate": 9.203795654003713e-06, + "loss": 0.399, + "step": 9388 + }, + { + "epoch": 0.2066696015848558, + "grad_norm": 1.6483830301783413, + "learning_rate": 9.203602654504554e-06, + "loss": 0.4524, + "step": 9389 + }, + { + "epoch": 0.2066916134712745, + "grad_norm": 0.8945774606995195, + "learning_rate": 9.203409633640811e-06, + "loss": 0.5232, + "step": 9390 + }, + { + "epoch": 0.20671362535769316, + "grad_norm": 1.4501380005576974, + "learning_rate": 9.203216591413466e-06, + "loss": 0.3353, + "step": 9391 + }, + { + "epoch": 0.20673563724411181, + "grad_norm": 1.7087017024796214, + "learning_rate": 9.2030235278235e-06, + "loss": 0.3489, + "step": 9392 + }, + { + "epoch": 0.2067576491305305, + "grad_norm": 3.5631804866384575, + "learning_rate": 9.202830442871893e-06, + "loss": 0.2603, + "step": 9393 + }, + { + "epoch": 0.20677966101694914, + "grad_norm": 1.0037140356077596, + "learning_rate": 9.202637336559628e-06, + "loss": 0.2836, + "step": 9394 + }, + { + "epoch": 0.20680167290336782, + "grad_norm": 1.2698835144625527, + "learning_rate": 9.202444208887683e-06, + "loss": 0.2567, + "step": 9395 + }, + { + "epoch": 0.2068236847897865, + "grad_norm": 1.3680776187052432, + "learning_rate": 9.202251059857046e-06, + "loss": 0.2849, + "step": 9396 + }, + { + "epoch": 0.20684569667620514, + "grad_norm": 0.6637468004387584, + "learning_rate": 9.202057889468692e-06, + "loss": 0.5354, + "step": 9397 + }, + { + "epoch": 0.20686770856262382, + "grad_norm": 1.7133280962069042, + "learning_rate": 9.201864697723608e-06, + "loss": 0.3247, + "step": 9398 + }, + { + "epoch": 0.2068897204490425, + "grad_norm": 1.7458125631789176, + "learning_rate": 9.201671484622771e-06, + "loss": 0.3133, + "step": 9399 + }, + { + "epoch": 0.20691173233546115, + "grad_norm": 1.4890026175867035, + "learning_rate": 9.201478250167167e-06, + "loss": 0.3303, + "step": 9400 + }, + { + "epoch": 0.20693374422187982, + "grad_norm": 1.8225215009896156, + "learning_rate": 9.201284994357776e-06, + "loss": 0.4471, + "step": 9401 + }, + { + "epoch": 0.20695575610829847, + "grad_norm": 1.2906777763382566, + "learning_rate": 9.201091717195581e-06, + "loss": 0.2752, + "step": 9402 + }, + { + "epoch": 0.20697776799471715, + "grad_norm": 1.4627820391516855, + "learning_rate": 9.200898418681565e-06, + "loss": 0.3354, + "step": 9403 + }, + { + "epoch": 0.20699977988113583, + "grad_norm": 0.6963504865182977, + "learning_rate": 9.200705098816709e-06, + "loss": 0.5274, + "step": 9404 + }, + { + "epoch": 0.20702179176755447, + "grad_norm": 1.4550566206070876, + "learning_rate": 9.200511757601997e-06, + "loss": 0.3437, + "step": 9405 + }, + { + "epoch": 0.20704380365397315, + "grad_norm": 1.443074615353124, + "learning_rate": 9.20031839503841e-06, + "loss": 0.2868, + "step": 9406 + }, + { + "epoch": 0.2070658155403918, + "grad_norm": 1.1735668769286975, + "learning_rate": 9.20012501112693e-06, + "loss": 0.2279, + "step": 9407 + }, + { + "epoch": 0.20708782742681048, + "grad_norm": 1.5681082204075605, + "learning_rate": 9.199931605868543e-06, + "loss": 0.3357, + "step": 9408 + }, + { + "epoch": 0.20710983931322915, + "grad_norm": 1.64368691866944, + "learning_rate": 9.19973817926423e-06, + "loss": 0.4277, + "step": 9409 + }, + { + "epoch": 0.2071318511996478, + "grad_norm": 1.916839147783192, + "learning_rate": 9.199544731314976e-06, + "loss": 0.2519, + "step": 9410 + }, + { + "epoch": 0.20715386308606648, + "grad_norm": 1.2578786197432545, + "learning_rate": 9.19935126202176e-06, + "loss": 0.3094, + "step": 9411 + }, + { + "epoch": 0.20717587497248513, + "grad_norm": 1.4070479662635673, + "learning_rate": 9.199157771385569e-06, + "loss": 0.4055, + "step": 9412 + }, + { + "epoch": 0.2071978868589038, + "grad_norm": 1.4103029754340755, + "learning_rate": 9.198964259407385e-06, + "loss": 0.3439, + "step": 9413 + }, + { + "epoch": 0.20721989874532248, + "grad_norm": 1.9108791484976173, + "learning_rate": 9.198770726088193e-06, + "loss": 0.3499, + "step": 9414 + }, + { + "epoch": 0.20724191063174113, + "grad_norm": 1.4886310611533202, + "learning_rate": 9.198577171428975e-06, + "loss": 0.349, + "step": 9415 + }, + { + "epoch": 0.2072639225181598, + "grad_norm": 1.2256494539518126, + "learning_rate": 9.198383595430713e-06, + "loss": 0.3844, + "step": 9416 + }, + { + "epoch": 0.20728593440457846, + "grad_norm": 0.6546911532042246, + "learning_rate": 9.198189998094395e-06, + "loss": 0.529, + "step": 9417 + }, + { + "epoch": 0.20730794629099714, + "grad_norm": 1.3230559131492534, + "learning_rate": 9.197996379421004e-06, + "loss": 0.2735, + "step": 9418 + }, + { + "epoch": 0.2073299581774158, + "grad_norm": 1.8410608363005816, + "learning_rate": 9.197802739411522e-06, + "loss": 0.4897, + "step": 9419 + }, + { + "epoch": 0.20735197006383446, + "grad_norm": 0.6076510027196323, + "learning_rate": 9.197609078066933e-06, + "loss": 0.5213, + "step": 9420 + }, + { + "epoch": 0.20737398195025314, + "grad_norm": 1.4692798379812049, + "learning_rate": 9.197415395388224e-06, + "loss": 0.3608, + "step": 9421 + }, + { + "epoch": 0.20739599383667182, + "grad_norm": 1.7011449355143544, + "learning_rate": 9.197221691376377e-06, + "loss": 0.399, + "step": 9422 + }, + { + "epoch": 0.20741800572309047, + "grad_norm": 4.414377172361622, + "learning_rate": 9.19702796603238e-06, + "loss": 0.4071, + "step": 9423 + }, + { + "epoch": 0.20744001760950914, + "grad_norm": 1.265539530618677, + "learning_rate": 9.196834219357211e-06, + "loss": 0.3182, + "step": 9424 + }, + { + "epoch": 0.2074620294959278, + "grad_norm": 1.6332057658143437, + "learning_rate": 9.196640451351862e-06, + "loss": 0.4213, + "step": 9425 + }, + { + "epoch": 0.20748404138234647, + "grad_norm": 1.3872565759899684, + "learning_rate": 9.196446662017312e-06, + "loss": 0.3394, + "step": 9426 + }, + { + "epoch": 0.20750605326876514, + "grad_norm": 1.5203170883507813, + "learning_rate": 9.19625285135455e-06, + "loss": 0.3529, + "step": 9427 + }, + { + "epoch": 0.2075280651551838, + "grad_norm": 1.5664500605679885, + "learning_rate": 9.196059019364558e-06, + "loss": 0.3589, + "step": 9428 + }, + { + "epoch": 0.20755007704160247, + "grad_norm": 1.3612887708763717, + "learning_rate": 9.195865166048323e-06, + "loss": 0.3144, + "step": 9429 + }, + { + "epoch": 0.20757208892802112, + "grad_norm": 1.2363119515470609, + "learning_rate": 9.195671291406832e-06, + "loss": 0.3136, + "step": 9430 + }, + { + "epoch": 0.2075941008144398, + "grad_norm": 1.309351580812487, + "learning_rate": 9.195477395441066e-06, + "loss": 0.3015, + "step": 9431 + }, + { + "epoch": 0.20761611270085847, + "grad_norm": 1.7925117161883037, + "learning_rate": 9.195283478152013e-06, + "loss": 0.4322, + "step": 9432 + }, + { + "epoch": 0.20763812458727712, + "grad_norm": 0.7158315001728297, + "learning_rate": 9.195089539540657e-06, + "loss": 0.5698, + "step": 9433 + }, + { + "epoch": 0.2076601364736958, + "grad_norm": 1.4229737822907438, + "learning_rate": 9.194895579607987e-06, + "loss": 0.3082, + "step": 9434 + }, + { + "epoch": 0.20768214836011445, + "grad_norm": 1.4117019954429832, + "learning_rate": 9.194701598354985e-06, + "loss": 0.3381, + "step": 9435 + }, + { + "epoch": 0.20770416024653313, + "grad_norm": 1.3909135864558748, + "learning_rate": 9.19450759578264e-06, + "loss": 0.3175, + "step": 9436 + }, + { + "epoch": 0.2077261721329518, + "grad_norm": 1.4909558903271234, + "learning_rate": 9.194313571891935e-06, + "loss": 0.3308, + "step": 9437 + }, + { + "epoch": 0.20774818401937045, + "grad_norm": 1.4204717217148364, + "learning_rate": 9.19411952668386e-06, + "loss": 0.438, + "step": 9438 + }, + { + "epoch": 0.20777019590578913, + "grad_norm": 1.6512341098907382, + "learning_rate": 9.193925460159397e-06, + "loss": 0.2887, + "step": 9439 + }, + { + "epoch": 0.2077922077922078, + "grad_norm": 0.7244393826653713, + "learning_rate": 9.193731372319533e-06, + "loss": 0.5675, + "step": 9440 + }, + { + "epoch": 0.20781421967862646, + "grad_norm": 1.2580246351862208, + "learning_rate": 9.19353726316526e-06, + "loss": 0.3184, + "step": 9441 + }, + { + "epoch": 0.20783623156504513, + "grad_norm": 0.6628950917256842, + "learning_rate": 9.193343132697557e-06, + "loss": 0.5479, + "step": 9442 + }, + { + "epoch": 0.20785824345146378, + "grad_norm": 1.865207186657422, + "learning_rate": 9.193148980917414e-06, + "loss": 0.3703, + "step": 9443 + }, + { + "epoch": 0.20788025533788246, + "grad_norm": 1.3928961700357698, + "learning_rate": 9.19295480782582e-06, + "loss": 0.3327, + "step": 9444 + }, + { + "epoch": 0.20790226722430113, + "grad_norm": 0.6249949129909069, + "learning_rate": 9.192760613423758e-06, + "loss": 0.5408, + "step": 9445 + }, + { + "epoch": 0.20792427911071978, + "grad_norm": 1.435532752523152, + "learning_rate": 9.192566397712216e-06, + "loss": 0.3027, + "step": 9446 + }, + { + "epoch": 0.20794629099713846, + "grad_norm": 1.8171216925418519, + "learning_rate": 9.192372160692183e-06, + "loss": 0.3266, + "step": 9447 + }, + { + "epoch": 0.2079683028835571, + "grad_norm": 1.3773813279071887, + "learning_rate": 9.192177902364643e-06, + "loss": 0.3662, + "step": 9448 + }, + { + "epoch": 0.2079903147699758, + "grad_norm": 0.6454218713328843, + "learning_rate": 9.191983622730587e-06, + "loss": 0.5262, + "step": 9449 + }, + { + "epoch": 0.20801232665639446, + "grad_norm": 1.5905175579967241, + "learning_rate": 9.191789321790999e-06, + "loss": 0.4119, + "step": 9450 + }, + { + "epoch": 0.2080343385428131, + "grad_norm": 1.6177298336418444, + "learning_rate": 9.19159499954687e-06, + "loss": 0.3702, + "step": 9451 + }, + { + "epoch": 0.2080563504292318, + "grad_norm": 0.6210491063316622, + "learning_rate": 9.191400655999182e-06, + "loss": 0.531, + "step": 9452 + }, + { + "epoch": 0.20807836231565044, + "grad_norm": 1.22108342708159, + "learning_rate": 9.191206291148932e-06, + "loss": 0.3802, + "step": 9453 + }, + { + "epoch": 0.20810037420206912, + "grad_norm": 1.4539710322304145, + "learning_rate": 9.191011904997098e-06, + "loss": 0.3871, + "step": 9454 + }, + { + "epoch": 0.2081223860884878, + "grad_norm": 0.6437178875328303, + "learning_rate": 9.190817497544673e-06, + "loss": 0.5472, + "step": 9455 + }, + { + "epoch": 0.20814439797490644, + "grad_norm": 1.473524841507212, + "learning_rate": 9.190623068792644e-06, + "loss": 0.3385, + "step": 9456 + }, + { + "epoch": 0.20816640986132512, + "grad_norm": 1.462754522391871, + "learning_rate": 9.190428618742e-06, + "loss": 0.3029, + "step": 9457 + }, + { + "epoch": 0.20818842174774377, + "grad_norm": 1.647399033334574, + "learning_rate": 9.190234147393731e-06, + "loss": 0.2946, + "step": 9458 + }, + { + "epoch": 0.20821043363416245, + "grad_norm": 0.6258097377927286, + "learning_rate": 9.19003965474882e-06, + "loss": 0.5301, + "step": 9459 + }, + { + "epoch": 0.20823244552058112, + "grad_norm": 1.9301776348667712, + "learning_rate": 9.18984514080826e-06, + "loss": 0.2926, + "step": 9460 + }, + { + "epoch": 0.20825445740699977, + "grad_norm": 1.3813188107605054, + "learning_rate": 9.189650605573039e-06, + "loss": 0.3802, + "step": 9461 + }, + { + "epoch": 0.20827646929341845, + "grad_norm": 1.234888349686249, + "learning_rate": 9.189456049044143e-06, + "loss": 0.3166, + "step": 9462 + }, + { + "epoch": 0.20829848117983712, + "grad_norm": 1.3850415670094118, + "learning_rate": 9.189261471222564e-06, + "loss": 0.405, + "step": 9463 + }, + { + "epoch": 0.20832049306625577, + "grad_norm": 0.6320106512445142, + "learning_rate": 9.18906687210929e-06, + "loss": 0.5277, + "step": 9464 + }, + { + "epoch": 0.20834250495267445, + "grad_norm": 1.415132693944839, + "learning_rate": 9.18887225170531e-06, + "loss": 0.3012, + "step": 9465 + }, + { + "epoch": 0.2083645168390931, + "grad_norm": 2.052451810760035, + "learning_rate": 9.188677610011612e-06, + "loss": 0.2928, + "step": 9466 + }, + { + "epoch": 0.20838652872551178, + "grad_norm": 1.7649435955911852, + "learning_rate": 9.188482947029187e-06, + "loss": 0.3088, + "step": 9467 + }, + { + "epoch": 0.20840854061193045, + "grad_norm": 2.9587243007544135, + "learning_rate": 9.188288262759023e-06, + "loss": 0.3177, + "step": 9468 + }, + { + "epoch": 0.2084305524983491, + "grad_norm": 1.6380609304790912, + "learning_rate": 9.188093557202108e-06, + "loss": 0.3311, + "step": 9469 + }, + { + "epoch": 0.20845256438476778, + "grad_norm": 1.752310717305698, + "learning_rate": 9.187898830359438e-06, + "loss": 0.4606, + "step": 9470 + }, + { + "epoch": 0.20847457627118643, + "grad_norm": 1.2583915419307146, + "learning_rate": 9.187704082231996e-06, + "loss": 0.3396, + "step": 9471 + }, + { + "epoch": 0.2084965881576051, + "grad_norm": 1.5401913932632783, + "learning_rate": 9.187509312820775e-06, + "loss": 0.3187, + "step": 9472 + }, + { + "epoch": 0.20851860004402378, + "grad_norm": 1.6985989878302228, + "learning_rate": 9.187314522126763e-06, + "loss": 0.3842, + "step": 9473 + }, + { + "epoch": 0.20854061193044243, + "grad_norm": 1.3892036607187046, + "learning_rate": 9.18711971015095e-06, + "loss": 0.4083, + "step": 9474 + }, + { + "epoch": 0.2085626238168611, + "grad_norm": 1.4376331264118267, + "learning_rate": 9.186924876894327e-06, + "loss": 0.3278, + "step": 9475 + }, + { + "epoch": 0.20858463570327976, + "grad_norm": 1.5428100977451493, + "learning_rate": 9.186730022357887e-06, + "loss": 0.3265, + "step": 9476 + }, + { + "epoch": 0.20860664758969844, + "grad_norm": 2.4807690413963397, + "learning_rate": 9.186535146542617e-06, + "loss": 0.3259, + "step": 9477 + }, + { + "epoch": 0.2086286594761171, + "grad_norm": 1.4976841906758058, + "learning_rate": 9.186340249449507e-06, + "loss": 0.3652, + "step": 9478 + }, + { + "epoch": 0.20865067136253576, + "grad_norm": 1.467416459790832, + "learning_rate": 9.186145331079548e-06, + "loss": 0.4391, + "step": 9479 + }, + { + "epoch": 0.20867268324895444, + "grad_norm": 1.330705021648295, + "learning_rate": 9.185950391433733e-06, + "loss": 0.291, + "step": 9480 + }, + { + "epoch": 0.20869469513537312, + "grad_norm": 1.1804804894452672, + "learning_rate": 9.18575543051305e-06, + "loss": 0.2781, + "step": 9481 + }, + { + "epoch": 0.20871670702179176, + "grad_norm": 1.4929823846252679, + "learning_rate": 9.185560448318492e-06, + "loss": 0.4279, + "step": 9482 + }, + { + "epoch": 0.20873871890821044, + "grad_norm": 1.3810812973225388, + "learning_rate": 9.18536544485105e-06, + "loss": 0.3035, + "step": 9483 + }, + { + "epoch": 0.2087607307946291, + "grad_norm": 1.221156051853346, + "learning_rate": 9.18517042011171e-06, + "loss": 0.2873, + "step": 9484 + }, + { + "epoch": 0.20878274268104777, + "grad_norm": 1.9131099558358453, + "learning_rate": 9.184975374101472e-06, + "loss": 0.3503, + "step": 9485 + }, + { + "epoch": 0.20880475456746644, + "grad_norm": 1.406831471276833, + "learning_rate": 9.18478030682132e-06, + "loss": 0.372, + "step": 9486 + }, + { + "epoch": 0.2088267664538851, + "grad_norm": 1.7564671957004412, + "learning_rate": 9.184585218272248e-06, + "loss": 0.3698, + "step": 9487 + }, + { + "epoch": 0.20884877834030377, + "grad_norm": 1.3162703306101233, + "learning_rate": 9.18439010845525e-06, + "loss": 0.3863, + "step": 9488 + }, + { + "epoch": 0.20887079022672242, + "grad_norm": 1.3525436067144045, + "learning_rate": 9.184194977371312e-06, + "loss": 0.3257, + "step": 9489 + }, + { + "epoch": 0.2088928021131411, + "grad_norm": 1.9398581169954958, + "learning_rate": 9.18399982502143e-06, + "loss": 0.3486, + "step": 9490 + }, + { + "epoch": 0.20891481399955977, + "grad_norm": 0.7000178384251136, + "learning_rate": 9.183804651406596e-06, + "loss": 0.5475, + "step": 9491 + }, + { + "epoch": 0.20893682588597842, + "grad_norm": 2.4880866006168194, + "learning_rate": 9.1836094565278e-06, + "loss": 0.3653, + "step": 9492 + }, + { + "epoch": 0.2089588377723971, + "grad_norm": 1.8196605295409611, + "learning_rate": 9.183414240386034e-06, + "loss": 0.3645, + "step": 9493 + }, + { + "epoch": 0.20898084965881575, + "grad_norm": 0.6448000260192395, + "learning_rate": 9.18321900298229e-06, + "loss": 0.5286, + "step": 9494 + }, + { + "epoch": 0.20900286154523443, + "grad_norm": 1.5980225322873163, + "learning_rate": 9.183023744317564e-06, + "loss": 0.4351, + "step": 9495 + }, + { + "epoch": 0.2090248734316531, + "grad_norm": 2.9044709431488362, + "learning_rate": 9.182828464392845e-06, + "loss": 0.346, + "step": 9496 + }, + { + "epoch": 0.20904688531807175, + "grad_norm": 1.6553278075353726, + "learning_rate": 9.182633163209125e-06, + "loss": 0.3856, + "step": 9497 + }, + { + "epoch": 0.20906889720449043, + "grad_norm": 1.4457431638601355, + "learning_rate": 9.182437840767399e-06, + "loss": 0.3711, + "step": 9498 + }, + { + "epoch": 0.20909090909090908, + "grad_norm": 1.4910371666905808, + "learning_rate": 9.182242497068657e-06, + "loss": 0.3692, + "step": 9499 + }, + { + "epoch": 0.20911292097732775, + "grad_norm": 0.6480160026541628, + "learning_rate": 9.182047132113894e-06, + "loss": 0.5112, + "step": 9500 + }, + { + "epoch": 0.20913493286374643, + "grad_norm": 1.272721740848787, + "learning_rate": 9.181851745904102e-06, + "loss": 0.3044, + "step": 9501 + }, + { + "epoch": 0.20915694475016508, + "grad_norm": 0.6372564920130153, + "learning_rate": 9.181656338440275e-06, + "loss": 0.5363, + "step": 9502 + }, + { + "epoch": 0.20917895663658376, + "grad_norm": 1.5571313560318005, + "learning_rate": 9.181460909723405e-06, + "loss": 0.4028, + "step": 9503 + }, + { + "epoch": 0.20920096852300243, + "grad_norm": 1.338941783521037, + "learning_rate": 9.181265459754485e-06, + "loss": 0.3485, + "step": 9504 + }, + { + "epoch": 0.20922298040942108, + "grad_norm": 1.2617923420171588, + "learning_rate": 9.181069988534512e-06, + "loss": 0.2942, + "step": 9505 + }, + { + "epoch": 0.20924499229583976, + "grad_norm": 1.29191470815119, + "learning_rate": 9.180874496064474e-06, + "loss": 0.3679, + "step": 9506 + }, + { + "epoch": 0.2092670041822584, + "grad_norm": 1.4220149178479125, + "learning_rate": 9.180678982345366e-06, + "loss": 0.3601, + "step": 9507 + }, + { + "epoch": 0.2092890160686771, + "grad_norm": 0.6424920341997207, + "learning_rate": 9.180483447378184e-06, + "loss": 0.5012, + "step": 9508 + }, + { + "epoch": 0.20931102795509576, + "grad_norm": 1.819236020644557, + "learning_rate": 9.180287891163921e-06, + "loss": 0.3477, + "step": 9509 + }, + { + "epoch": 0.2093330398415144, + "grad_norm": 1.4923156357562588, + "learning_rate": 9.18009231370357e-06, + "loss": 0.2974, + "step": 9510 + }, + { + "epoch": 0.2093550517279331, + "grad_norm": 1.9579782027956927, + "learning_rate": 9.179896714998126e-06, + "loss": 0.3891, + "step": 9511 + }, + { + "epoch": 0.20937706361435174, + "grad_norm": 1.340466742155588, + "learning_rate": 9.179701095048583e-06, + "loss": 0.3715, + "step": 9512 + }, + { + "epoch": 0.20939907550077042, + "grad_norm": 1.5950519663147713, + "learning_rate": 9.179505453855933e-06, + "loss": 0.3511, + "step": 9513 + }, + { + "epoch": 0.2094210873871891, + "grad_norm": 2.021375241347487, + "learning_rate": 9.179309791421174e-06, + "loss": 0.33, + "step": 9514 + }, + { + "epoch": 0.20944309927360774, + "grad_norm": 1.372531818559332, + "learning_rate": 9.179114107745298e-06, + "loss": 0.3296, + "step": 9515 + }, + { + "epoch": 0.20946511116002642, + "grad_norm": 1.8287324042081052, + "learning_rate": 9.1789184028293e-06, + "loss": 0.4106, + "step": 9516 + }, + { + "epoch": 0.20948712304644507, + "grad_norm": 1.3612459638561052, + "learning_rate": 9.178722676674175e-06, + "loss": 0.454, + "step": 9517 + }, + { + "epoch": 0.20950913493286374, + "grad_norm": 1.4835023153260387, + "learning_rate": 9.178526929280917e-06, + "loss": 0.3873, + "step": 9518 + }, + { + "epoch": 0.20953114681928242, + "grad_norm": 1.5204565364194476, + "learning_rate": 9.178331160650522e-06, + "loss": 0.368, + "step": 9519 + }, + { + "epoch": 0.20955315870570107, + "grad_norm": 0.6385066237127509, + "learning_rate": 9.178135370783985e-06, + "loss": 0.514, + "step": 9520 + }, + { + "epoch": 0.20957517059211975, + "grad_norm": 1.4193587001070662, + "learning_rate": 9.1779395596823e-06, + "loss": 0.3232, + "step": 9521 + }, + { + "epoch": 0.20959718247853842, + "grad_norm": 1.1482615364724418, + "learning_rate": 9.177743727346461e-06, + "loss": 0.3124, + "step": 9522 + }, + { + "epoch": 0.20961919436495707, + "grad_norm": 1.5624141067633817, + "learning_rate": 9.177547873777469e-06, + "loss": 0.3042, + "step": 9523 + }, + { + "epoch": 0.20964120625137575, + "grad_norm": 1.213120401462538, + "learning_rate": 9.177351998976312e-06, + "loss": 0.3678, + "step": 9524 + }, + { + "epoch": 0.2096632181377944, + "grad_norm": 1.2945197747725927, + "learning_rate": 9.177156102943989e-06, + "loss": 0.374, + "step": 9525 + }, + { + "epoch": 0.20968523002421308, + "grad_norm": 0.6241036585087687, + "learning_rate": 9.176960185681498e-06, + "loss": 0.5442, + "step": 9526 + }, + { + "epoch": 0.20970724191063175, + "grad_norm": 1.6846889055136722, + "learning_rate": 9.176764247189831e-06, + "loss": 0.3872, + "step": 9527 + }, + { + "epoch": 0.2097292537970504, + "grad_norm": 1.8057672423881455, + "learning_rate": 9.176568287469984e-06, + "loss": 0.341, + "step": 9528 + }, + { + "epoch": 0.20975126568346908, + "grad_norm": 1.6685647265729155, + "learning_rate": 9.176372306522956e-06, + "loss": 0.2816, + "step": 9529 + }, + { + "epoch": 0.20977327756988773, + "grad_norm": 1.4718564018528488, + "learning_rate": 9.176176304349742e-06, + "loss": 0.3182, + "step": 9530 + }, + { + "epoch": 0.2097952894563064, + "grad_norm": 1.5938462249994683, + "learning_rate": 9.175980280951335e-06, + "loss": 0.3221, + "step": 9531 + }, + { + "epoch": 0.20981730134272508, + "grad_norm": 1.8781417218490368, + "learning_rate": 9.175784236328735e-06, + "loss": 0.2374, + "step": 9532 + }, + { + "epoch": 0.20983931322914373, + "grad_norm": 0.638202305641667, + "learning_rate": 9.175588170482937e-06, + "loss": 0.5118, + "step": 9533 + }, + { + "epoch": 0.2098613251155624, + "grad_norm": 1.4000909943799444, + "learning_rate": 9.175392083414937e-06, + "loss": 0.3244, + "step": 9534 + }, + { + "epoch": 0.20988333700198106, + "grad_norm": 0.6495084069617619, + "learning_rate": 9.175195975125733e-06, + "loss": 0.5308, + "step": 9535 + }, + { + "epoch": 0.20990534888839973, + "grad_norm": 1.5359497165662175, + "learning_rate": 9.174999845616322e-06, + "loss": 0.3298, + "step": 9536 + }, + { + "epoch": 0.2099273607748184, + "grad_norm": 1.3447073199924529, + "learning_rate": 9.174803694887697e-06, + "loss": 0.3774, + "step": 9537 + }, + { + "epoch": 0.20994937266123706, + "grad_norm": 0.6206273558992798, + "learning_rate": 9.17460752294086e-06, + "loss": 0.5377, + "step": 9538 + }, + { + "epoch": 0.20997138454765574, + "grad_norm": 2.4979330058348452, + "learning_rate": 9.174411329776803e-06, + "loss": 0.3845, + "step": 9539 + }, + { + "epoch": 0.2099933964340744, + "grad_norm": 1.2174728827480337, + "learning_rate": 9.174215115396528e-06, + "loss": 0.2842, + "step": 9540 + }, + { + "epoch": 0.21001540832049306, + "grad_norm": 0.6370758294233374, + "learning_rate": 9.174018879801028e-06, + "loss": 0.5197, + "step": 9541 + }, + { + "epoch": 0.21003742020691174, + "grad_norm": 2.4293858582582155, + "learning_rate": 9.173822622991305e-06, + "loss": 0.3901, + "step": 9542 + }, + { + "epoch": 0.2100594320933304, + "grad_norm": 1.3926651946615476, + "learning_rate": 9.173626344968351e-06, + "loss": 0.3466, + "step": 9543 + }, + { + "epoch": 0.21008144397974907, + "grad_norm": 2.1077988099442555, + "learning_rate": 9.173430045733169e-06, + "loss": 0.3119, + "step": 9544 + }, + { + "epoch": 0.21010345586616774, + "grad_norm": 1.7541917210295594, + "learning_rate": 9.173233725286753e-06, + "loss": 0.409, + "step": 9545 + }, + { + "epoch": 0.2101254677525864, + "grad_norm": 1.480882898473344, + "learning_rate": 9.173037383630102e-06, + "loss": 0.3332, + "step": 9546 + }, + { + "epoch": 0.21014747963900507, + "grad_norm": 2.270070366084351, + "learning_rate": 9.172841020764213e-06, + "loss": 0.3445, + "step": 9547 + }, + { + "epoch": 0.21016949152542372, + "grad_norm": 1.778000188262566, + "learning_rate": 9.172644636690085e-06, + "loss": 0.3984, + "step": 9548 + }, + { + "epoch": 0.2101915034118424, + "grad_norm": 1.2716116146353729, + "learning_rate": 9.172448231408717e-06, + "loss": 0.3492, + "step": 9549 + }, + { + "epoch": 0.21021351529826107, + "grad_norm": 1.4874717936888626, + "learning_rate": 9.172251804921105e-06, + "loss": 0.3299, + "step": 9550 + }, + { + "epoch": 0.21023552718467972, + "grad_norm": 1.2455333843451522, + "learning_rate": 9.17205535722825e-06, + "loss": 0.2834, + "step": 9551 + }, + { + "epoch": 0.2102575390710984, + "grad_norm": 1.441758129395771, + "learning_rate": 9.171858888331146e-06, + "loss": 0.3163, + "step": 9552 + }, + { + "epoch": 0.21027955095751705, + "grad_norm": 2.272120044135521, + "learning_rate": 9.171662398230799e-06, + "loss": 0.3527, + "step": 9553 + }, + { + "epoch": 0.21030156284393572, + "grad_norm": 1.215360030470612, + "learning_rate": 9.171465886928199e-06, + "loss": 0.2546, + "step": 9554 + }, + { + "epoch": 0.2103235747303544, + "grad_norm": 1.4281865986522593, + "learning_rate": 9.17126935442435e-06, + "loss": 0.3728, + "step": 9555 + }, + { + "epoch": 0.21034558661677305, + "grad_norm": 1.8025368261365775, + "learning_rate": 9.17107280072025e-06, + "loss": 0.4964, + "step": 9556 + }, + { + "epoch": 0.21036759850319173, + "grad_norm": 1.2552023816765396, + "learning_rate": 9.170876225816898e-06, + "loss": 0.3153, + "step": 9557 + }, + { + "epoch": 0.21038961038961038, + "grad_norm": 7.612722619248723, + "learning_rate": 9.17067962971529e-06, + "loss": 0.2583, + "step": 9558 + }, + { + "epoch": 0.21041162227602905, + "grad_norm": 0.6793176112500363, + "learning_rate": 9.17048301241643e-06, + "loss": 0.5126, + "step": 9559 + }, + { + "epoch": 0.21043363416244773, + "grad_norm": 0.6949293334619062, + "learning_rate": 9.170286373921318e-06, + "loss": 0.542, + "step": 9560 + }, + { + "epoch": 0.21045564604886638, + "grad_norm": 1.5204299399803127, + "learning_rate": 9.170089714230948e-06, + "loss": 0.3294, + "step": 9561 + }, + { + "epoch": 0.21047765793528506, + "grad_norm": 1.6935333951475566, + "learning_rate": 9.169893033346323e-06, + "loss": 0.3694, + "step": 9562 + }, + { + "epoch": 0.21049966982170373, + "grad_norm": 0.6659484179739279, + "learning_rate": 9.16969633126844e-06, + "loss": 0.57, + "step": 9563 + }, + { + "epoch": 0.21052168170812238, + "grad_norm": 1.395836799612915, + "learning_rate": 9.169499607998303e-06, + "loss": 0.2918, + "step": 9564 + }, + { + "epoch": 0.21054369359454106, + "grad_norm": 1.331883385381064, + "learning_rate": 9.169302863536906e-06, + "loss": 0.3023, + "step": 9565 + }, + { + "epoch": 0.2105657054809597, + "grad_norm": 1.314301958869252, + "learning_rate": 9.169106097885255e-06, + "loss": 0.3066, + "step": 9566 + }, + { + "epoch": 0.21058771736737839, + "grad_norm": 1.658636158430524, + "learning_rate": 9.168909311044346e-06, + "loss": 0.404, + "step": 9567 + }, + { + "epoch": 0.21060972925379706, + "grad_norm": 1.5737054750063664, + "learning_rate": 9.168712503015182e-06, + "loss": 0.4316, + "step": 9568 + }, + { + "epoch": 0.2106317411402157, + "grad_norm": 1.3756119794954769, + "learning_rate": 9.16851567379876e-06, + "loss": 0.3432, + "step": 9569 + }, + { + "epoch": 0.2106537530266344, + "grad_norm": 1.6278303538786467, + "learning_rate": 9.168318823396082e-06, + "loss": 0.4467, + "step": 9570 + }, + { + "epoch": 0.21067576491305304, + "grad_norm": 2.0190615390760063, + "learning_rate": 9.168121951808152e-06, + "loss": 0.4318, + "step": 9571 + }, + { + "epoch": 0.21069777679947171, + "grad_norm": 1.9404914531050368, + "learning_rate": 9.167925059035964e-06, + "loss": 0.3701, + "step": 9572 + }, + { + "epoch": 0.2107197886858904, + "grad_norm": 1.4038987974376893, + "learning_rate": 9.167728145080522e-06, + "loss": 0.3966, + "step": 9573 + }, + { + "epoch": 0.21074180057230904, + "grad_norm": 1.6232485646143653, + "learning_rate": 9.167531209942828e-06, + "loss": 0.4234, + "step": 9574 + }, + { + "epoch": 0.21076381245872772, + "grad_norm": 3.8286731505418485, + "learning_rate": 9.167334253623881e-06, + "loss": 0.3689, + "step": 9575 + }, + { + "epoch": 0.21078582434514637, + "grad_norm": 1.2467898578950212, + "learning_rate": 9.167137276124682e-06, + "loss": 0.3538, + "step": 9576 + }, + { + "epoch": 0.21080783623156504, + "grad_norm": 1.818246048345907, + "learning_rate": 9.166940277446233e-06, + "loss": 0.3882, + "step": 9577 + }, + { + "epoch": 0.21082984811798372, + "grad_norm": 2.1233354734384604, + "learning_rate": 9.166743257589537e-06, + "loss": 0.3745, + "step": 9578 + }, + { + "epoch": 0.21085186000440237, + "grad_norm": 1.5446245303535764, + "learning_rate": 9.166546216555591e-06, + "loss": 0.3941, + "step": 9579 + }, + { + "epoch": 0.21087387189082105, + "grad_norm": 2.0498709593379436, + "learning_rate": 9.1663491543454e-06, + "loss": 0.2829, + "step": 9580 + }, + { + "epoch": 0.21089588377723972, + "grad_norm": 1.558034112700759, + "learning_rate": 9.166152070959964e-06, + "loss": 0.342, + "step": 9581 + }, + { + "epoch": 0.21091789566365837, + "grad_norm": 2.214988390067768, + "learning_rate": 9.165954966400284e-06, + "loss": 0.2764, + "step": 9582 + }, + { + "epoch": 0.21093990755007705, + "grad_norm": 1.4850363072972548, + "learning_rate": 9.165757840667364e-06, + "loss": 0.2915, + "step": 9583 + }, + { + "epoch": 0.2109619194364957, + "grad_norm": 1.4259682049650158, + "learning_rate": 9.165560693762205e-06, + "loss": 0.3383, + "step": 9584 + }, + { + "epoch": 0.21098393132291438, + "grad_norm": 1.4474341109080913, + "learning_rate": 9.165363525685807e-06, + "loss": 0.4173, + "step": 9585 + }, + { + "epoch": 0.21100594320933305, + "grad_norm": 1.376960253365758, + "learning_rate": 9.165166336439175e-06, + "loss": 0.3485, + "step": 9586 + }, + { + "epoch": 0.2110279550957517, + "grad_norm": 1.4927180617989693, + "learning_rate": 9.164969126023312e-06, + "loss": 0.4045, + "step": 9587 + }, + { + "epoch": 0.21104996698217038, + "grad_norm": 1.4175973056774298, + "learning_rate": 9.164771894439214e-06, + "loss": 0.2557, + "step": 9588 + }, + { + "epoch": 0.21107197886858903, + "grad_norm": 1.4639821855095603, + "learning_rate": 9.16457464168789e-06, + "loss": 0.3382, + "step": 9589 + }, + { + "epoch": 0.2110939907550077, + "grad_norm": 2.8814777680316364, + "learning_rate": 9.16437736777034e-06, + "loss": 0.4213, + "step": 9590 + }, + { + "epoch": 0.21111600264142638, + "grad_norm": 1.3996393233765871, + "learning_rate": 9.164180072687566e-06, + "loss": 0.3138, + "step": 9591 + }, + { + "epoch": 0.21113801452784503, + "grad_norm": 1.5476522080958315, + "learning_rate": 9.163982756440573e-06, + "loss": 0.3254, + "step": 9592 + }, + { + "epoch": 0.2111600264142637, + "grad_norm": 1.5476098882006817, + "learning_rate": 9.163785419030363e-06, + "loss": 0.3324, + "step": 9593 + }, + { + "epoch": 0.21118203830068236, + "grad_norm": 1.189509216280003, + "learning_rate": 9.163588060457936e-06, + "loss": 0.295, + "step": 9594 + }, + { + "epoch": 0.21120405018710103, + "grad_norm": 1.9234784614208869, + "learning_rate": 9.163390680724298e-06, + "loss": 0.4403, + "step": 9595 + }, + { + "epoch": 0.2112260620735197, + "grad_norm": 1.2699822011246207, + "learning_rate": 9.163193279830453e-06, + "loss": 0.2664, + "step": 9596 + }, + { + "epoch": 0.21124807395993836, + "grad_norm": 0.7604236152009431, + "learning_rate": 9.162995857777402e-06, + "loss": 0.5412, + "step": 9597 + }, + { + "epoch": 0.21127008584635704, + "grad_norm": 0.7219841904519293, + "learning_rate": 9.16279841456615e-06, + "loss": 0.5581, + "step": 9598 + }, + { + "epoch": 0.21129209773277569, + "grad_norm": 1.5768362473444297, + "learning_rate": 9.1626009501977e-06, + "loss": 0.3412, + "step": 9599 + }, + { + "epoch": 0.21131410961919436, + "grad_norm": 1.2345763903506661, + "learning_rate": 9.162403464673055e-06, + "loss": 0.3023, + "step": 9600 + }, + { + "epoch": 0.21133612150561304, + "grad_norm": 2.159447241225908, + "learning_rate": 9.162205957993217e-06, + "loss": 0.405, + "step": 9601 + }, + { + "epoch": 0.2113581333920317, + "grad_norm": 1.3877988953163711, + "learning_rate": 9.162008430159196e-06, + "loss": 0.3523, + "step": 9602 + }, + { + "epoch": 0.21138014527845037, + "grad_norm": 1.7592878217121957, + "learning_rate": 9.161810881171988e-06, + "loss": 0.3983, + "step": 9603 + }, + { + "epoch": 0.21140215716486904, + "grad_norm": 1.5299811254889055, + "learning_rate": 9.161613311032603e-06, + "loss": 0.3659, + "step": 9604 + }, + { + "epoch": 0.2114241690512877, + "grad_norm": 1.1898082771405785, + "learning_rate": 9.161415719742043e-06, + "loss": 0.3612, + "step": 9605 + }, + { + "epoch": 0.21144618093770637, + "grad_norm": 1.4076460327596139, + "learning_rate": 9.161218107301311e-06, + "loss": 0.3637, + "step": 9606 + }, + { + "epoch": 0.21146819282412502, + "grad_norm": 1.3822925619687194, + "learning_rate": 9.161020473711415e-06, + "loss": 0.3345, + "step": 9607 + }, + { + "epoch": 0.2114902047105437, + "grad_norm": 1.3945102767061668, + "learning_rate": 9.160822818973354e-06, + "loss": 0.3043, + "step": 9608 + }, + { + "epoch": 0.21151221659696237, + "grad_norm": 0.9916897027687406, + "learning_rate": 9.160625143088137e-06, + "loss": 0.5512, + "step": 9609 + }, + { + "epoch": 0.21153422848338102, + "grad_norm": 1.3331248408614178, + "learning_rate": 9.16042744605677e-06, + "loss": 0.3456, + "step": 9610 + }, + { + "epoch": 0.2115562403697997, + "grad_norm": 35.614759318066014, + "learning_rate": 9.160229727880252e-06, + "loss": 0.3826, + "step": 9611 + }, + { + "epoch": 0.21157825225621835, + "grad_norm": 0.6658361275626327, + "learning_rate": 9.16003198855959e-06, + "loss": 0.5132, + "step": 9612 + }, + { + "epoch": 0.21160026414263702, + "grad_norm": 9.174536037977347, + "learning_rate": 9.159834228095793e-06, + "loss": 0.2989, + "step": 9613 + }, + { + "epoch": 0.2116222760290557, + "grad_norm": 1.3182245703900066, + "learning_rate": 9.159636446489862e-06, + "loss": 0.3733, + "step": 9614 + }, + { + "epoch": 0.21164428791547435, + "grad_norm": 1.5507895336718194, + "learning_rate": 9.159438643742802e-06, + "loss": 0.3013, + "step": 9615 + }, + { + "epoch": 0.21166629980189303, + "grad_norm": 1.5569352881621943, + "learning_rate": 9.159240819855622e-06, + "loss": 0.335, + "step": 9616 + }, + { + "epoch": 0.21168831168831168, + "grad_norm": 1.3046158130272978, + "learning_rate": 9.159042974829322e-06, + "loss": 0.3473, + "step": 9617 + }, + { + "epoch": 0.21171032357473035, + "grad_norm": 1.4352772162498533, + "learning_rate": 9.158845108664914e-06, + "loss": 0.3916, + "step": 9618 + }, + { + "epoch": 0.21173233546114903, + "grad_norm": 0.8164368885877687, + "learning_rate": 9.158647221363398e-06, + "loss": 0.5456, + "step": 9619 + }, + { + "epoch": 0.21175434734756768, + "grad_norm": 1.3329663033461592, + "learning_rate": 9.158449312925783e-06, + "loss": 0.3457, + "step": 9620 + }, + { + "epoch": 0.21177635923398636, + "grad_norm": 2.6670645330811658, + "learning_rate": 9.158251383353074e-06, + "loss": 0.3902, + "step": 9621 + }, + { + "epoch": 0.21179837112040503, + "grad_norm": 0.7012422444489923, + "learning_rate": 9.158053432646275e-06, + "loss": 0.5467, + "step": 9622 + }, + { + "epoch": 0.21182038300682368, + "grad_norm": 1.2024977020767589, + "learning_rate": 9.157855460806396e-06, + "loss": 0.3059, + "step": 9623 + }, + { + "epoch": 0.21184239489324236, + "grad_norm": 1.541164398629723, + "learning_rate": 9.157657467834441e-06, + "loss": 0.354, + "step": 9624 + }, + { + "epoch": 0.211864406779661, + "grad_norm": 0.6311947063205903, + "learning_rate": 9.157459453731416e-06, + "loss": 0.5087, + "step": 9625 + }, + { + "epoch": 0.21188641866607968, + "grad_norm": 1.4637098799752513, + "learning_rate": 9.157261418498326e-06, + "loss": 0.3437, + "step": 9626 + }, + { + "epoch": 0.21190843055249836, + "grad_norm": 1.408149995213884, + "learning_rate": 9.157063362136182e-06, + "loss": 0.3302, + "step": 9627 + }, + { + "epoch": 0.211930442438917, + "grad_norm": 1.3732089376335512, + "learning_rate": 9.156865284645987e-06, + "loss": 0.4218, + "step": 9628 + }, + { + "epoch": 0.2119524543253357, + "grad_norm": 1.1332638124550067, + "learning_rate": 9.156667186028748e-06, + "loss": 0.2871, + "step": 9629 + }, + { + "epoch": 0.21197446621175434, + "grad_norm": 1.2992380735437943, + "learning_rate": 9.156469066285474e-06, + "loss": 0.3583, + "step": 9630 + }, + { + "epoch": 0.211996478098173, + "grad_norm": 1.0282602679022577, + "learning_rate": 9.156270925417168e-06, + "loss": 0.3164, + "step": 9631 + }, + { + "epoch": 0.2120184899845917, + "grad_norm": 1.1827412009058786, + "learning_rate": 9.156072763424841e-06, + "loss": 0.3452, + "step": 9632 + }, + { + "epoch": 0.21204050187101034, + "grad_norm": 1.4686183663617354, + "learning_rate": 9.155874580309498e-06, + "loss": 0.3195, + "step": 9633 + }, + { + "epoch": 0.21206251375742902, + "grad_norm": 0.8113576433674153, + "learning_rate": 9.155676376072146e-06, + "loss": 0.5557, + "step": 9634 + }, + { + "epoch": 0.21208452564384767, + "grad_norm": 1.157396412564593, + "learning_rate": 9.155478150713794e-06, + "loss": 0.2472, + "step": 9635 + }, + { + "epoch": 0.21210653753026634, + "grad_norm": 1.2403599285645486, + "learning_rate": 9.15527990423545e-06, + "loss": 0.3563, + "step": 9636 + }, + { + "epoch": 0.21212854941668502, + "grad_norm": 1.2308002767094757, + "learning_rate": 9.15508163663812e-06, + "loss": 0.2828, + "step": 9637 + }, + { + "epoch": 0.21215056130310367, + "grad_norm": 1.3255802278015256, + "learning_rate": 9.15488334792281e-06, + "loss": 0.3932, + "step": 9638 + }, + { + "epoch": 0.21217257318952235, + "grad_norm": 5.861542850776495, + "learning_rate": 9.15468503809053e-06, + "loss": 0.4316, + "step": 9639 + }, + { + "epoch": 0.212194585075941, + "grad_norm": 1.312134650031135, + "learning_rate": 9.154486707142288e-06, + "loss": 0.2981, + "step": 9640 + }, + { + "epoch": 0.21221659696235967, + "grad_norm": 1.9401303223801345, + "learning_rate": 9.154288355079093e-06, + "loss": 0.349, + "step": 9641 + }, + { + "epoch": 0.21223860884877835, + "grad_norm": 1.5573108353849754, + "learning_rate": 9.15408998190195e-06, + "loss": 0.4549, + "step": 9642 + }, + { + "epoch": 0.212260620735197, + "grad_norm": 3.1828906210125565, + "learning_rate": 9.15389158761187e-06, + "loss": 0.2958, + "step": 9643 + }, + { + "epoch": 0.21228263262161567, + "grad_norm": 1.608623120004181, + "learning_rate": 9.15369317220986e-06, + "loss": 0.2651, + "step": 9644 + }, + { + "epoch": 0.21230464450803435, + "grad_norm": 1.341107906859451, + "learning_rate": 9.153494735696927e-06, + "loss": 0.3125, + "step": 9645 + }, + { + "epoch": 0.212326656394453, + "grad_norm": 2.8630843796519225, + "learning_rate": 9.153296278074083e-06, + "loss": 0.3304, + "step": 9646 + }, + { + "epoch": 0.21234866828087168, + "grad_norm": 1.3018369972176715, + "learning_rate": 9.153097799342335e-06, + "loss": 0.3468, + "step": 9647 + }, + { + "epoch": 0.21237068016729033, + "grad_norm": 1.441907418353457, + "learning_rate": 9.15289929950269e-06, + "loss": 0.3586, + "step": 9648 + }, + { + "epoch": 0.212392692053709, + "grad_norm": 1.3214132105062175, + "learning_rate": 9.15270077855616e-06, + "loss": 0.3242, + "step": 9649 + }, + { + "epoch": 0.21241470394012768, + "grad_norm": 1.723454692524685, + "learning_rate": 9.152502236503753e-06, + "loss": 0.2759, + "step": 9650 + }, + { + "epoch": 0.21243671582654633, + "grad_norm": 1.227782666388801, + "learning_rate": 9.152303673346477e-06, + "loss": 0.3171, + "step": 9651 + }, + { + "epoch": 0.212458727712965, + "grad_norm": 1.2341703412367186, + "learning_rate": 9.152105089085342e-06, + "loss": 0.3505, + "step": 9652 + }, + { + "epoch": 0.21248073959938366, + "grad_norm": 2.6826794170063146, + "learning_rate": 9.151906483721356e-06, + "loss": 0.3281, + "step": 9653 + }, + { + "epoch": 0.21250275148580233, + "grad_norm": 1.3611843669858967, + "learning_rate": 9.15170785725553e-06, + "loss": 0.3994, + "step": 9654 + }, + { + "epoch": 0.212524763372221, + "grad_norm": 1.656185087576986, + "learning_rate": 9.151509209688874e-06, + "loss": 0.3607, + "step": 9655 + }, + { + "epoch": 0.21254677525863966, + "grad_norm": 1.4128880269384765, + "learning_rate": 9.151310541022394e-06, + "loss": 0.3847, + "step": 9656 + }, + { + "epoch": 0.21256878714505834, + "grad_norm": 1.8998831858289624, + "learning_rate": 9.151111851257104e-06, + "loss": 0.2708, + "step": 9657 + }, + { + "epoch": 0.21259079903147698, + "grad_norm": 1.6480276709184847, + "learning_rate": 9.150913140394012e-06, + "loss": 0.3881, + "step": 9658 + }, + { + "epoch": 0.21261281091789566, + "grad_norm": 1.3035319506016358, + "learning_rate": 9.150714408434127e-06, + "loss": 0.3014, + "step": 9659 + }, + { + "epoch": 0.21263482280431434, + "grad_norm": 1.541882520720632, + "learning_rate": 9.150515655378462e-06, + "loss": 0.4108, + "step": 9660 + }, + { + "epoch": 0.212656834690733, + "grad_norm": 1.2794933746581028, + "learning_rate": 9.150316881228025e-06, + "loss": 0.3379, + "step": 9661 + }, + { + "epoch": 0.21267884657715166, + "grad_norm": 1.29223960992489, + "learning_rate": 9.150118085983826e-06, + "loss": 0.3689, + "step": 9662 + }, + { + "epoch": 0.21270085846357034, + "grad_norm": 1.3629109192942932, + "learning_rate": 9.149919269646875e-06, + "loss": 0.2758, + "step": 9663 + }, + { + "epoch": 0.212722870349989, + "grad_norm": 0.8123446816105099, + "learning_rate": 9.149720432218185e-06, + "loss": 0.5075, + "step": 9664 + }, + { + "epoch": 0.21274488223640767, + "grad_norm": 1.3709786736685172, + "learning_rate": 9.149521573698763e-06, + "loss": 0.3268, + "step": 9665 + }, + { + "epoch": 0.21276689412282632, + "grad_norm": 1.7215412264092593, + "learning_rate": 9.149322694089623e-06, + "loss": 0.3875, + "step": 9666 + }, + { + "epoch": 0.212788906009245, + "grad_norm": 1.2561117173514775, + "learning_rate": 9.149123793391774e-06, + "loss": 0.3758, + "step": 9667 + }, + { + "epoch": 0.21281091789566367, + "grad_norm": 1.2802898753141314, + "learning_rate": 9.148924871606229e-06, + "loss": 0.3364, + "step": 9668 + }, + { + "epoch": 0.21283292978208232, + "grad_norm": 1.3284066584281518, + "learning_rate": 9.148725928733994e-06, + "loss": 0.3695, + "step": 9669 + }, + { + "epoch": 0.212854941668501, + "grad_norm": 1.3242874407340486, + "learning_rate": 9.148526964776085e-06, + "loss": 0.314, + "step": 9670 + }, + { + "epoch": 0.21287695355491965, + "grad_norm": 1.840167886975421, + "learning_rate": 9.148327979733514e-06, + "loss": 0.372, + "step": 9671 + }, + { + "epoch": 0.21289896544133832, + "grad_norm": 1.228343628445202, + "learning_rate": 9.148128973607286e-06, + "loss": 0.3083, + "step": 9672 + }, + { + "epoch": 0.212920977327757, + "grad_norm": 1.1805839869907349, + "learning_rate": 9.14792994639842e-06, + "loss": 0.3389, + "step": 9673 + }, + { + "epoch": 0.21294298921417565, + "grad_norm": 1.841978183818336, + "learning_rate": 9.14773089810792e-06, + "loss": 0.5522, + "step": 9674 + }, + { + "epoch": 0.21296500110059433, + "grad_norm": 3.4372751909323638, + "learning_rate": 9.147531828736803e-06, + "loss": 0.2723, + "step": 9675 + }, + { + "epoch": 0.21298701298701297, + "grad_norm": 0.7158817309196788, + "learning_rate": 9.147332738286082e-06, + "loss": 0.5215, + "step": 9676 + }, + { + "epoch": 0.21300902487343165, + "grad_norm": 1.61284942783872, + "learning_rate": 9.147133626756765e-06, + "loss": 0.3483, + "step": 9677 + }, + { + "epoch": 0.21303103675985033, + "grad_norm": 3.155016642408302, + "learning_rate": 9.146934494149865e-06, + "loss": 0.3834, + "step": 9678 + }, + { + "epoch": 0.21305304864626898, + "grad_norm": 1.3180971523869789, + "learning_rate": 9.146735340466393e-06, + "loss": 0.3134, + "step": 9679 + }, + { + "epoch": 0.21307506053268765, + "grad_norm": 1.475232341263232, + "learning_rate": 9.146536165707364e-06, + "loss": 0.3713, + "step": 9680 + }, + { + "epoch": 0.2130970724191063, + "grad_norm": 1.3800113141597132, + "learning_rate": 9.14633696987379e-06, + "loss": 0.3543, + "step": 9681 + }, + { + "epoch": 0.21311908430552498, + "grad_norm": 0.6360457484018563, + "learning_rate": 9.146137752966679e-06, + "loss": 0.5564, + "step": 9682 + }, + { + "epoch": 0.21314109619194366, + "grad_norm": 1.4392403479816076, + "learning_rate": 9.145938514987049e-06, + "loss": 0.4103, + "step": 9683 + }, + { + "epoch": 0.2131631080783623, + "grad_norm": 1.7880723524701312, + "learning_rate": 9.14573925593591e-06, + "loss": 0.2736, + "step": 9684 + }, + { + "epoch": 0.21318511996478098, + "grad_norm": 0.6599015093709736, + "learning_rate": 9.145539975814275e-06, + "loss": 0.5266, + "step": 9685 + }, + { + "epoch": 0.21320713185119966, + "grad_norm": 1.5034245693352648, + "learning_rate": 9.145340674623157e-06, + "loss": 0.3294, + "step": 9686 + }, + { + "epoch": 0.2132291437376183, + "grad_norm": 1.4442127132925708, + "learning_rate": 9.145141352363569e-06, + "loss": 0.3991, + "step": 9687 + }, + { + "epoch": 0.213251155624037, + "grad_norm": 7.605495798872255, + "learning_rate": 9.144942009036521e-06, + "loss": 0.4122, + "step": 9688 + }, + { + "epoch": 0.21327316751045564, + "grad_norm": 1.6346060019874513, + "learning_rate": 9.144742644643033e-06, + "loss": 0.3035, + "step": 9689 + }, + { + "epoch": 0.2132951793968743, + "grad_norm": 1.3635825450041614, + "learning_rate": 9.14454325918411e-06, + "loss": 0.2859, + "step": 9690 + }, + { + "epoch": 0.213317191283293, + "grad_norm": 1.687575111056665, + "learning_rate": 9.144343852660774e-06, + "loss": 0.3622, + "step": 9691 + }, + { + "epoch": 0.21333920316971164, + "grad_norm": 1.456507492857144, + "learning_rate": 9.144144425074033e-06, + "loss": 0.4423, + "step": 9692 + }, + { + "epoch": 0.21336121505613032, + "grad_norm": 1.3619070390229815, + "learning_rate": 9.143944976424902e-06, + "loss": 0.381, + "step": 9693 + }, + { + "epoch": 0.21338322694254896, + "grad_norm": 0.6554880923076977, + "learning_rate": 9.143745506714393e-06, + "loss": 0.5475, + "step": 9694 + }, + { + "epoch": 0.21340523882896764, + "grad_norm": 1.4904005484234248, + "learning_rate": 9.14354601594352e-06, + "loss": 0.4454, + "step": 9695 + }, + { + "epoch": 0.21342725071538632, + "grad_norm": 0.6432076066832588, + "learning_rate": 9.1433465041133e-06, + "loss": 0.5466, + "step": 9696 + }, + { + "epoch": 0.21344926260180497, + "grad_norm": 2.414816138500843, + "learning_rate": 9.143146971224745e-06, + "loss": 0.2946, + "step": 9697 + }, + { + "epoch": 0.21347127448822364, + "grad_norm": 1.3277138991449458, + "learning_rate": 9.14294741727887e-06, + "loss": 0.3522, + "step": 9698 + }, + { + "epoch": 0.2134932863746423, + "grad_norm": 1.5149765410982954, + "learning_rate": 9.142747842276686e-06, + "loss": 0.4255, + "step": 9699 + }, + { + "epoch": 0.21351529826106097, + "grad_norm": 1.9858498440759935, + "learning_rate": 9.142548246219212e-06, + "loss": 0.3501, + "step": 9700 + }, + { + "epoch": 0.21353731014747965, + "grad_norm": 0.6568090289624573, + "learning_rate": 9.142348629107459e-06, + "loss": 0.5499, + "step": 9701 + }, + { + "epoch": 0.2135593220338983, + "grad_norm": 1.916301452397306, + "learning_rate": 9.142148990942444e-06, + "loss": 0.3899, + "step": 9702 + }, + { + "epoch": 0.21358133392031697, + "grad_norm": 1.677627385179865, + "learning_rate": 9.141949331725179e-06, + "loss": 0.5323, + "step": 9703 + }, + { + "epoch": 0.21360334580673565, + "grad_norm": 1.6907686313687331, + "learning_rate": 9.14174965145668e-06, + "loss": 0.3251, + "step": 9704 + }, + { + "epoch": 0.2136253576931543, + "grad_norm": 1.2069788039575886, + "learning_rate": 9.141549950137963e-06, + "loss": 0.4411, + "step": 9705 + }, + { + "epoch": 0.21364736957957298, + "grad_norm": 1.6760510659281518, + "learning_rate": 9.141350227770041e-06, + "loss": 0.3708, + "step": 9706 + }, + { + "epoch": 0.21366938146599163, + "grad_norm": 1.6314536295498059, + "learning_rate": 9.14115048435393e-06, + "loss": 0.3096, + "step": 9707 + }, + { + "epoch": 0.2136913933524103, + "grad_norm": 1.7402109132460954, + "learning_rate": 9.140950719890647e-06, + "loss": 0.435, + "step": 9708 + }, + { + "epoch": 0.21371340523882898, + "grad_norm": 1.5918184755758178, + "learning_rate": 9.140750934381205e-06, + "loss": 0.3224, + "step": 9709 + }, + { + "epoch": 0.21373541712524763, + "grad_norm": 4.020333636714144, + "learning_rate": 9.140551127826617e-06, + "loss": 0.344, + "step": 9710 + }, + { + "epoch": 0.2137574290116663, + "grad_norm": 1.3897162257975564, + "learning_rate": 9.140351300227905e-06, + "loss": 0.2822, + "step": 9711 + }, + { + "epoch": 0.21377944089808495, + "grad_norm": 0.6582162133667724, + "learning_rate": 9.140151451586079e-06, + "loss": 0.5189, + "step": 9712 + }, + { + "epoch": 0.21380145278450363, + "grad_norm": 0.6584387126532947, + "learning_rate": 9.139951581902158e-06, + "loss": 0.529, + "step": 9713 + }, + { + "epoch": 0.2138234646709223, + "grad_norm": 1.5446421832130686, + "learning_rate": 9.139751691177155e-06, + "loss": 0.382, + "step": 9714 + }, + { + "epoch": 0.21384547655734096, + "grad_norm": 2.897124307643082, + "learning_rate": 9.139551779412089e-06, + "loss": 0.3337, + "step": 9715 + }, + { + "epoch": 0.21386748844375963, + "grad_norm": 1.517379623620946, + "learning_rate": 9.139351846607974e-06, + "loss": 0.3423, + "step": 9716 + }, + { + "epoch": 0.21388950033017828, + "grad_norm": 0.6493579403084303, + "learning_rate": 9.139151892765828e-06, + "loss": 0.56, + "step": 9717 + }, + { + "epoch": 0.21391151221659696, + "grad_norm": 1.838650741446947, + "learning_rate": 9.138951917886665e-06, + "loss": 0.3128, + "step": 9718 + }, + { + "epoch": 0.21393352410301564, + "grad_norm": 1.3313525294697919, + "learning_rate": 9.138751921971501e-06, + "loss": 0.2935, + "step": 9719 + }, + { + "epoch": 0.2139555359894343, + "grad_norm": 0.6453641025139647, + "learning_rate": 9.138551905021355e-06, + "loss": 0.4873, + "step": 9720 + }, + { + "epoch": 0.21397754787585296, + "grad_norm": 1.7511461659357257, + "learning_rate": 9.138351867037243e-06, + "loss": 0.4239, + "step": 9721 + }, + { + "epoch": 0.2139995597622716, + "grad_norm": 1.2435773201547033, + "learning_rate": 9.138151808020181e-06, + "loss": 0.2838, + "step": 9722 + }, + { + "epoch": 0.2140215716486903, + "grad_norm": 0.6194603228726363, + "learning_rate": 9.137951727971185e-06, + "loss": 0.53, + "step": 9723 + }, + { + "epoch": 0.21404358353510897, + "grad_norm": 1.836612235511851, + "learning_rate": 9.137751626891271e-06, + "loss": 0.3733, + "step": 9724 + }, + { + "epoch": 0.21406559542152762, + "grad_norm": 1.673069345730593, + "learning_rate": 9.137551504781461e-06, + "loss": 0.3025, + "step": 9725 + }, + { + "epoch": 0.2140876073079463, + "grad_norm": 1.983967509556946, + "learning_rate": 9.137351361642768e-06, + "loss": 0.303, + "step": 9726 + }, + { + "epoch": 0.21410961919436497, + "grad_norm": 2.5741919621283342, + "learning_rate": 9.137151197476208e-06, + "loss": 0.3398, + "step": 9727 + }, + { + "epoch": 0.21413163108078362, + "grad_norm": 3.6154680026291675, + "learning_rate": 9.136951012282803e-06, + "loss": 0.3577, + "step": 9728 + }, + { + "epoch": 0.2141536429672023, + "grad_norm": 1.3180016251107518, + "learning_rate": 9.136750806063566e-06, + "loss": 0.3276, + "step": 9729 + }, + { + "epoch": 0.21417565485362094, + "grad_norm": 1.4561098059497852, + "learning_rate": 9.136550578819519e-06, + "loss": 0.3931, + "step": 9730 + }, + { + "epoch": 0.21419766674003962, + "grad_norm": 0.6554031092383248, + "learning_rate": 9.136350330551675e-06, + "loss": 0.5617, + "step": 9731 + }, + { + "epoch": 0.2142196786264583, + "grad_norm": 2.28931347142354, + "learning_rate": 9.136150061261054e-06, + "loss": 0.4362, + "step": 9732 + }, + { + "epoch": 0.21424169051287695, + "grad_norm": 1.1803444066512907, + "learning_rate": 9.135949770948673e-06, + "loss": 0.288, + "step": 9733 + }, + { + "epoch": 0.21426370239929562, + "grad_norm": 1.5482084072860232, + "learning_rate": 9.135749459615551e-06, + "loss": 0.3645, + "step": 9734 + }, + { + "epoch": 0.21428571428571427, + "grad_norm": 1.3823164753822919, + "learning_rate": 9.135549127262706e-06, + "loss": 0.4361, + "step": 9735 + }, + { + "epoch": 0.21430772617213295, + "grad_norm": 1.2555536144147519, + "learning_rate": 9.135348773891154e-06, + "loss": 0.3961, + "step": 9736 + }, + { + "epoch": 0.21432973805855163, + "grad_norm": 1.577607961133582, + "learning_rate": 9.135148399501919e-06, + "loss": 0.3293, + "step": 9737 + }, + { + "epoch": 0.21435174994497028, + "grad_norm": 1.9365778703396352, + "learning_rate": 9.134948004096013e-06, + "loss": 0.4749, + "step": 9738 + }, + { + "epoch": 0.21437376183138895, + "grad_norm": 0.6331140024820885, + "learning_rate": 9.134747587674457e-06, + "loss": 0.5411, + "step": 9739 + }, + { + "epoch": 0.2143957737178076, + "grad_norm": 1.6667840766468422, + "learning_rate": 9.134547150238272e-06, + "loss": 0.2957, + "step": 9740 + }, + { + "epoch": 0.21441778560422628, + "grad_norm": 3.2092501793698127, + "learning_rate": 9.13434669178847e-06, + "loss": 0.3982, + "step": 9741 + }, + { + "epoch": 0.21443979749064496, + "grad_norm": 1.4495150848704883, + "learning_rate": 9.134146212326076e-06, + "loss": 0.3879, + "step": 9742 + }, + { + "epoch": 0.2144618093770636, + "grad_norm": 1.4358127395014486, + "learning_rate": 9.133945711852109e-06, + "loss": 0.3413, + "step": 9743 + }, + { + "epoch": 0.21448382126348228, + "grad_norm": 0.6248315168481774, + "learning_rate": 9.133745190367584e-06, + "loss": 0.5012, + "step": 9744 + }, + { + "epoch": 0.21450583314990096, + "grad_norm": 4.430272114842945, + "learning_rate": 9.133544647873523e-06, + "loss": 0.2826, + "step": 9745 + }, + { + "epoch": 0.2145278450363196, + "grad_norm": 1.3364657625378236, + "learning_rate": 9.133344084370943e-06, + "loss": 0.3468, + "step": 9746 + }, + { + "epoch": 0.21454985692273829, + "grad_norm": 0.9400176865492448, + "learning_rate": 9.133143499860866e-06, + "loss": 0.2401, + "step": 9747 + }, + { + "epoch": 0.21457186880915693, + "grad_norm": 1.2805724532290716, + "learning_rate": 9.132942894344311e-06, + "loss": 0.3611, + "step": 9748 + }, + { + "epoch": 0.2145938806955756, + "grad_norm": 0.6164067203991503, + "learning_rate": 9.132742267822297e-06, + "loss": 0.4971, + "step": 9749 + }, + { + "epoch": 0.2146158925819943, + "grad_norm": 1.3110112743239741, + "learning_rate": 9.132541620295842e-06, + "loss": 0.3067, + "step": 9750 + }, + { + "epoch": 0.21463790446841294, + "grad_norm": 1.5207810112566877, + "learning_rate": 9.132340951765967e-06, + "loss": 0.2949, + "step": 9751 + }, + { + "epoch": 0.21465991635483161, + "grad_norm": 1.3398200881251312, + "learning_rate": 9.132140262233695e-06, + "loss": 0.4788, + "step": 9752 + }, + { + "epoch": 0.21468192824125026, + "grad_norm": 1.4580527243552222, + "learning_rate": 9.131939551700041e-06, + "loss": 0.3831, + "step": 9753 + }, + { + "epoch": 0.21470394012766894, + "grad_norm": 1.4181768976663682, + "learning_rate": 9.131738820166028e-06, + "loss": 0.3232, + "step": 9754 + }, + { + "epoch": 0.21472595201408762, + "grad_norm": 1.9301670438650627, + "learning_rate": 9.131538067632675e-06, + "loss": 0.393, + "step": 9755 + }, + { + "epoch": 0.21474796390050627, + "grad_norm": 1.5816290505175596, + "learning_rate": 9.131337294101003e-06, + "loss": 0.3871, + "step": 9756 + }, + { + "epoch": 0.21476997578692494, + "grad_norm": 1.4754879098299543, + "learning_rate": 9.131136499572033e-06, + "loss": 0.413, + "step": 9757 + }, + { + "epoch": 0.2147919876733436, + "grad_norm": 1.3333087150467107, + "learning_rate": 9.130935684046785e-06, + "loss": 0.3268, + "step": 9758 + }, + { + "epoch": 0.21481399955976227, + "grad_norm": 1.5783247678766101, + "learning_rate": 9.130734847526276e-06, + "loss": 0.3618, + "step": 9759 + }, + { + "epoch": 0.21483601144618095, + "grad_norm": 2.8449012874615174, + "learning_rate": 9.130533990011535e-06, + "loss": 0.3646, + "step": 9760 + }, + { + "epoch": 0.2148580233325996, + "grad_norm": 0.6565936157418719, + "learning_rate": 9.130333111503575e-06, + "loss": 0.5362, + "step": 9761 + }, + { + "epoch": 0.21488003521901827, + "grad_norm": 1.3270866646583452, + "learning_rate": 9.130132212003422e-06, + "loss": 0.3774, + "step": 9762 + }, + { + "epoch": 0.21490204710543695, + "grad_norm": 2.088764880786577, + "learning_rate": 9.129931291512093e-06, + "loss": 0.4073, + "step": 9763 + }, + { + "epoch": 0.2149240589918556, + "grad_norm": 1.3999586037822798, + "learning_rate": 9.129730350030613e-06, + "loss": 0.3012, + "step": 9764 + }, + { + "epoch": 0.21494607087827428, + "grad_norm": 1.2216548003190792, + "learning_rate": 9.129529387559998e-06, + "loss": 0.3157, + "step": 9765 + }, + { + "epoch": 0.21496808276469292, + "grad_norm": 1.6779566720919723, + "learning_rate": 9.129328404101276e-06, + "loss": 0.2919, + "step": 9766 + }, + { + "epoch": 0.2149900946511116, + "grad_norm": 1.224971413562187, + "learning_rate": 9.129127399655465e-06, + "loss": 0.3007, + "step": 9767 + }, + { + "epoch": 0.21501210653753028, + "grad_norm": 1.2304400360220191, + "learning_rate": 9.128926374223585e-06, + "loss": 0.2425, + "step": 9768 + }, + { + "epoch": 0.21503411842394893, + "grad_norm": 2.0182517026372144, + "learning_rate": 9.128725327806663e-06, + "loss": 0.373, + "step": 9769 + }, + { + "epoch": 0.2150561303103676, + "grad_norm": 0.6136736108146941, + "learning_rate": 9.128524260405714e-06, + "loss": 0.5059, + "step": 9770 + }, + { + "epoch": 0.21507814219678625, + "grad_norm": 1.7230261352530394, + "learning_rate": 9.128323172021765e-06, + "loss": 0.3629, + "step": 9771 + }, + { + "epoch": 0.21510015408320493, + "grad_norm": 1.5571268026214897, + "learning_rate": 9.128122062655835e-06, + "loss": 0.2966, + "step": 9772 + }, + { + "epoch": 0.2151221659696236, + "grad_norm": 2.2366606804933693, + "learning_rate": 9.127920932308949e-06, + "loss": 0.3674, + "step": 9773 + }, + { + "epoch": 0.21514417785604226, + "grad_norm": 0.6399084535068538, + "learning_rate": 9.127719780982127e-06, + "loss": 0.5579, + "step": 9774 + }, + { + "epoch": 0.21516618974246093, + "grad_norm": 1.7951086809197456, + "learning_rate": 9.127518608676393e-06, + "loss": 0.3186, + "step": 9775 + }, + { + "epoch": 0.21518820162887958, + "grad_norm": 1.2517686503952967, + "learning_rate": 9.127317415392768e-06, + "loss": 0.2614, + "step": 9776 + }, + { + "epoch": 0.21521021351529826, + "grad_norm": 1.3094192962204037, + "learning_rate": 9.127116201132276e-06, + "loss": 0.305, + "step": 9777 + }, + { + "epoch": 0.21523222540171694, + "grad_norm": 2.489520020058935, + "learning_rate": 9.126914965895936e-06, + "loss": 0.4455, + "step": 9778 + }, + { + "epoch": 0.21525423728813559, + "grad_norm": 1.4348036119476488, + "learning_rate": 9.126713709684775e-06, + "loss": 0.3272, + "step": 9779 + }, + { + "epoch": 0.21527624917455426, + "grad_norm": 1.3085564397020129, + "learning_rate": 9.126512432499816e-06, + "loss": 0.3981, + "step": 9780 + }, + { + "epoch": 0.2152982610609729, + "grad_norm": 1.1428131346735708, + "learning_rate": 9.126311134342078e-06, + "loss": 0.3372, + "step": 9781 + }, + { + "epoch": 0.2153202729473916, + "grad_norm": 1.1944237883775042, + "learning_rate": 9.126109815212588e-06, + "loss": 0.3376, + "step": 9782 + }, + { + "epoch": 0.21534228483381027, + "grad_norm": 1.5201598947184345, + "learning_rate": 9.125908475112366e-06, + "loss": 0.3402, + "step": 9783 + }, + { + "epoch": 0.21536429672022891, + "grad_norm": 1.442305860514431, + "learning_rate": 9.125707114042438e-06, + "loss": 0.401, + "step": 9784 + }, + { + "epoch": 0.2153863086066476, + "grad_norm": 1.386189566006503, + "learning_rate": 9.125505732003826e-06, + "loss": 0.3167, + "step": 9785 + }, + { + "epoch": 0.21540832049306627, + "grad_norm": 0.6468816198736241, + "learning_rate": 9.125304328997554e-06, + "loss": 0.5101, + "step": 9786 + }, + { + "epoch": 0.21543033237948492, + "grad_norm": 1.2936082831004443, + "learning_rate": 9.125102905024645e-06, + "loss": 0.3661, + "step": 9787 + }, + { + "epoch": 0.2154523442659036, + "grad_norm": 1.6866200240352778, + "learning_rate": 9.124901460086124e-06, + "loss": 0.4207, + "step": 9788 + }, + { + "epoch": 0.21547435615232224, + "grad_norm": 1.3281311127125326, + "learning_rate": 9.124699994183016e-06, + "loss": 0.2841, + "step": 9789 + }, + { + "epoch": 0.21549636803874092, + "grad_norm": 1.997958704080281, + "learning_rate": 9.12449850731634e-06, + "loss": 0.3825, + "step": 9790 + }, + { + "epoch": 0.2155183799251596, + "grad_norm": 0.6362082573003979, + "learning_rate": 9.124296999487122e-06, + "loss": 0.5255, + "step": 9791 + }, + { + "epoch": 0.21554039181157825, + "grad_norm": 2.288766135577811, + "learning_rate": 9.12409547069639e-06, + "loss": 0.3366, + "step": 9792 + }, + { + "epoch": 0.21556240369799692, + "grad_norm": 2.389259410727541, + "learning_rate": 9.123893920945164e-06, + "loss": 0.3989, + "step": 9793 + }, + { + "epoch": 0.21558441558441557, + "grad_norm": 1.9852234576558294, + "learning_rate": 9.12369235023447e-06, + "loss": 0.4291, + "step": 9794 + }, + { + "epoch": 0.21560642747083425, + "grad_norm": 2.73775848680398, + "learning_rate": 9.123490758565332e-06, + "loss": 0.3403, + "step": 9795 + }, + { + "epoch": 0.21562843935725293, + "grad_norm": 1.5771456781603062, + "learning_rate": 9.123289145938774e-06, + "loss": 0.3366, + "step": 9796 + }, + { + "epoch": 0.21565045124367158, + "grad_norm": 1.2073170222836456, + "learning_rate": 9.123087512355824e-06, + "loss": 0.308, + "step": 9797 + }, + { + "epoch": 0.21567246313009025, + "grad_norm": 2.889027324361165, + "learning_rate": 9.122885857817502e-06, + "loss": 0.329, + "step": 9798 + }, + { + "epoch": 0.2156944750165089, + "grad_norm": 4.764223261995341, + "learning_rate": 9.122684182324837e-06, + "loss": 0.3855, + "step": 9799 + }, + { + "epoch": 0.21571648690292758, + "grad_norm": 1.6635010654180353, + "learning_rate": 9.12248248587885e-06, + "loss": 0.4586, + "step": 9800 + }, + { + "epoch": 0.21573849878934626, + "grad_norm": 1.118810118513695, + "learning_rate": 9.12228076848057e-06, + "loss": 0.2732, + "step": 9801 + }, + { + "epoch": 0.2157605106757649, + "grad_norm": 1.3183350731909826, + "learning_rate": 9.122079030131021e-06, + "loss": 0.2972, + "step": 9802 + }, + { + "epoch": 0.21578252256218358, + "grad_norm": 2.060616237878675, + "learning_rate": 9.121877270831228e-06, + "loss": 0.407, + "step": 9803 + }, + { + "epoch": 0.21580453444860226, + "grad_norm": 0.6676912493481527, + "learning_rate": 9.121675490582214e-06, + "loss": 0.531, + "step": 9804 + }, + { + "epoch": 0.2158265463350209, + "grad_norm": 2.690195726101676, + "learning_rate": 9.121473689385008e-06, + "loss": 0.3395, + "step": 9805 + }, + { + "epoch": 0.21584855822143958, + "grad_norm": 1.3993827150184814, + "learning_rate": 9.121271867240635e-06, + "loss": 0.3429, + "step": 9806 + }, + { + "epoch": 0.21587057010785823, + "grad_norm": 0.6094019556878665, + "learning_rate": 9.12107002415012e-06, + "loss": 0.4899, + "step": 9807 + }, + { + "epoch": 0.2158925819942769, + "grad_norm": 2.1019959774917845, + "learning_rate": 9.120868160114489e-06, + "loss": 0.362, + "step": 9808 + }, + { + "epoch": 0.2159145938806956, + "grad_norm": 1.072612317348178, + "learning_rate": 9.12066627513477e-06, + "loss": 0.3147, + "step": 9809 + }, + { + "epoch": 0.21593660576711424, + "grad_norm": 1.5851780598612693, + "learning_rate": 9.120464369211984e-06, + "loss": 0.3737, + "step": 9810 + }, + { + "epoch": 0.2159586176535329, + "grad_norm": 1.6454928168104093, + "learning_rate": 9.12026244234716e-06, + "loss": 0.4119, + "step": 9811 + }, + { + "epoch": 0.21598062953995156, + "grad_norm": 3.2215644139780246, + "learning_rate": 9.120060494541328e-06, + "loss": 0.37, + "step": 9812 + }, + { + "epoch": 0.21600264142637024, + "grad_norm": 1.3259457763743987, + "learning_rate": 9.119858525795508e-06, + "loss": 0.3854, + "step": 9813 + }, + { + "epoch": 0.21602465331278892, + "grad_norm": 0.6485921511560591, + "learning_rate": 9.11965653611073e-06, + "loss": 0.5191, + "step": 9814 + }, + { + "epoch": 0.21604666519920757, + "grad_norm": 1.8963679081289169, + "learning_rate": 9.119454525488021e-06, + "loss": 0.3848, + "step": 9815 + }, + { + "epoch": 0.21606867708562624, + "grad_norm": 1.31418695099902, + "learning_rate": 9.119252493928406e-06, + "loss": 0.2992, + "step": 9816 + }, + { + "epoch": 0.2160906889720449, + "grad_norm": 1.6785374106587305, + "learning_rate": 9.119050441432913e-06, + "loss": 0.3032, + "step": 9817 + }, + { + "epoch": 0.21611270085846357, + "grad_norm": 1.6966269580778175, + "learning_rate": 9.118848368002569e-06, + "loss": 0.3459, + "step": 9818 + }, + { + "epoch": 0.21613471274488225, + "grad_norm": 1.4990202139255677, + "learning_rate": 9.1186462736384e-06, + "loss": 0.4077, + "step": 9819 + }, + { + "epoch": 0.2161567246313009, + "grad_norm": 0.6366674667678034, + "learning_rate": 9.118444158341433e-06, + "loss": 0.5731, + "step": 9820 + }, + { + "epoch": 0.21617873651771957, + "grad_norm": 1.2794883364435932, + "learning_rate": 9.118242022112698e-06, + "loss": 0.3268, + "step": 9821 + }, + { + "epoch": 0.21620074840413822, + "grad_norm": 1.4625683980919861, + "learning_rate": 9.118039864953218e-06, + "loss": 0.3893, + "step": 9822 + }, + { + "epoch": 0.2162227602905569, + "grad_norm": 0.5979154386692614, + "learning_rate": 9.117837686864023e-06, + "loss": 0.4962, + "step": 9823 + }, + { + "epoch": 0.21624477217697557, + "grad_norm": 1.7120846712913924, + "learning_rate": 9.117635487846142e-06, + "loss": 0.2521, + "step": 9824 + }, + { + "epoch": 0.21626678406339422, + "grad_norm": 1.820476421437261, + "learning_rate": 9.117433267900599e-06, + "loss": 0.3607, + "step": 9825 + }, + { + "epoch": 0.2162887959498129, + "grad_norm": 1.765715880756943, + "learning_rate": 9.117231027028424e-06, + "loss": 0.3296, + "step": 9826 + }, + { + "epoch": 0.21631080783623158, + "grad_norm": 1.3419335796335858, + "learning_rate": 9.117028765230645e-06, + "loss": 0.2993, + "step": 9827 + }, + { + "epoch": 0.21633281972265023, + "grad_norm": 1.3044042697585891, + "learning_rate": 9.11682648250829e-06, + "loss": 0.2789, + "step": 9828 + }, + { + "epoch": 0.2163548316090689, + "grad_norm": 1.863217641091164, + "learning_rate": 9.116624178862387e-06, + "loss": 0.408, + "step": 9829 + }, + { + "epoch": 0.21637684349548755, + "grad_norm": 1.9588833121807616, + "learning_rate": 9.116421854293961e-06, + "loss": 0.2984, + "step": 9830 + }, + { + "epoch": 0.21639885538190623, + "grad_norm": 2.1577661967976436, + "learning_rate": 9.116219508804048e-06, + "loss": 0.3165, + "step": 9831 + }, + { + "epoch": 0.2164208672683249, + "grad_norm": 1.662619919541817, + "learning_rate": 9.116017142393667e-06, + "loss": 0.3932, + "step": 9832 + }, + { + "epoch": 0.21644287915474356, + "grad_norm": 3.006930130004191, + "learning_rate": 9.115814755063853e-06, + "loss": 0.4831, + "step": 9833 + }, + { + "epoch": 0.21646489104116223, + "grad_norm": 0.6626029472841312, + "learning_rate": 9.115612346815634e-06, + "loss": 0.5486, + "step": 9834 + }, + { + "epoch": 0.21648690292758088, + "grad_norm": 1.6225973820767565, + "learning_rate": 9.115409917650035e-06, + "loss": 0.3705, + "step": 9835 + }, + { + "epoch": 0.21650891481399956, + "grad_norm": 1.440261042525771, + "learning_rate": 9.115207467568089e-06, + "loss": 0.3789, + "step": 9836 + }, + { + "epoch": 0.21653092670041824, + "grad_norm": 0.6265867180068141, + "learning_rate": 9.115004996570821e-06, + "loss": 0.5367, + "step": 9837 + }, + { + "epoch": 0.21655293858683689, + "grad_norm": 1.6081103389248603, + "learning_rate": 9.114802504659265e-06, + "loss": 0.2487, + "step": 9838 + }, + { + "epoch": 0.21657495047325556, + "grad_norm": 0.6259560973000119, + "learning_rate": 9.114599991834446e-06, + "loss": 0.5368, + "step": 9839 + }, + { + "epoch": 0.2165969623596742, + "grad_norm": 1.3615600345004972, + "learning_rate": 9.114397458097395e-06, + "loss": 0.3882, + "step": 9840 + }, + { + "epoch": 0.2166189742460929, + "grad_norm": 1.4313143820587293, + "learning_rate": 9.11419490344914e-06, + "loss": 0.2713, + "step": 9841 + }, + { + "epoch": 0.21664098613251156, + "grad_norm": 0.6215510461165294, + "learning_rate": 9.113992327890714e-06, + "loss": 0.5502, + "step": 9842 + }, + { + "epoch": 0.21666299801893021, + "grad_norm": 1.6709823911029025, + "learning_rate": 9.113789731423142e-06, + "loss": 0.2864, + "step": 9843 + }, + { + "epoch": 0.2166850099053489, + "grad_norm": 1.1517948960503996, + "learning_rate": 9.113587114047457e-06, + "loss": 0.3276, + "step": 9844 + }, + { + "epoch": 0.21670702179176757, + "grad_norm": 1.2930307847510876, + "learning_rate": 9.113384475764686e-06, + "loss": 0.3656, + "step": 9845 + }, + { + "epoch": 0.21672903367818622, + "grad_norm": 1.5327360923934328, + "learning_rate": 9.113181816575862e-06, + "loss": 0.3574, + "step": 9846 + }, + { + "epoch": 0.2167510455646049, + "grad_norm": 1.4204991978190713, + "learning_rate": 9.112979136482014e-06, + "loss": 0.3762, + "step": 9847 + }, + { + "epoch": 0.21677305745102354, + "grad_norm": 2.6603845035802585, + "learning_rate": 9.11277643548417e-06, + "loss": 0.3274, + "step": 9848 + }, + { + "epoch": 0.21679506933744222, + "grad_norm": 1.400494408714307, + "learning_rate": 9.112573713583361e-06, + "loss": 0.3168, + "step": 9849 + }, + { + "epoch": 0.2168170812238609, + "grad_norm": 1.4679999866590359, + "learning_rate": 9.11237097078062e-06, + "loss": 0.2824, + "step": 9850 + }, + { + "epoch": 0.21683909311027955, + "grad_norm": 3.676499690487234, + "learning_rate": 9.112168207076975e-06, + "loss": 0.3768, + "step": 9851 + }, + { + "epoch": 0.21686110499669822, + "grad_norm": 0.6517722051512111, + "learning_rate": 9.111965422473457e-06, + "loss": 0.53, + "step": 9852 + }, + { + "epoch": 0.21688311688311687, + "grad_norm": 1.5104897748655344, + "learning_rate": 9.111762616971097e-06, + "loss": 0.3495, + "step": 9853 + }, + { + "epoch": 0.21690512876953555, + "grad_norm": 1.2803729388629608, + "learning_rate": 9.111559790570926e-06, + "loss": 0.2889, + "step": 9854 + }, + { + "epoch": 0.21692714065595423, + "grad_norm": 0.6648082600858, + "learning_rate": 9.111356943273973e-06, + "loss": 0.5066, + "step": 9855 + }, + { + "epoch": 0.21694915254237288, + "grad_norm": 1.5302010671128592, + "learning_rate": 9.11115407508127e-06, + "loss": 0.3142, + "step": 9856 + }, + { + "epoch": 0.21697116442879155, + "grad_norm": 1.6465385278869686, + "learning_rate": 9.11095118599385e-06, + "loss": 0.3184, + "step": 9857 + }, + { + "epoch": 0.2169931763152102, + "grad_norm": 1.3011763453783827, + "learning_rate": 9.110748276012742e-06, + "loss": 0.3474, + "step": 9858 + }, + { + "epoch": 0.21701518820162888, + "grad_norm": 1.432425655710964, + "learning_rate": 9.110545345138977e-06, + "loss": 0.2829, + "step": 9859 + }, + { + "epoch": 0.21703720008804755, + "grad_norm": 1.9274843014102003, + "learning_rate": 9.11034239337359e-06, + "loss": 0.392, + "step": 9860 + }, + { + "epoch": 0.2170592119744662, + "grad_norm": 0.7212331080324902, + "learning_rate": 9.110139420717607e-06, + "loss": 0.5401, + "step": 9861 + }, + { + "epoch": 0.21708122386088488, + "grad_norm": 1.3050260114206118, + "learning_rate": 9.109936427172062e-06, + "loss": 0.2948, + "step": 9862 + }, + { + "epoch": 0.21710323574730353, + "grad_norm": 2.2505329026234535, + "learning_rate": 9.109733412737988e-06, + "loss": 0.2999, + "step": 9863 + }, + { + "epoch": 0.2171252476337222, + "grad_norm": 2.3297904539868304, + "learning_rate": 9.109530377416416e-06, + "loss": 0.3926, + "step": 9864 + }, + { + "epoch": 0.21714725952014088, + "grad_norm": 1.5414231002007588, + "learning_rate": 9.109327321208377e-06, + "loss": 0.3803, + "step": 9865 + }, + { + "epoch": 0.21716927140655953, + "grad_norm": 2.23554799626547, + "learning_rate": 9.109124244114903e-06, + "loss": 0.3681, + "step": 9866 + }, + { + "epoch": 0.2171912832929782, + "grad_norm": 2.2697868348496133, + "learning_rate": 9.108921146137028e-06, + "loss": 0.3561, + "step": 9867 + }, + { + "epoch": 0.2172132951793969, + "grad_norm": 1.4771653756159497, + "learning_rate": 9.108718027275785e-06, + "loss": 0.3727, + "step": 9868 + }, + { + "epoch": 0.21723530706581554, + "grad_norm": 1.5055011905845308, + "learning_rate": 9.108514887532202e-06, + "loss": 0.3925, + "step": 9869 + }, + { + "epoch": 0.2172573189522342, + "grad_norm": 3.481063059336405, + "learning_rate": 9.108311726907315e-06, + "loss": 0.3395, + "step": 9870 + }, + { + "epoch": 0.21727933083865286, + "grad_norm": 2.341603139058064, + "learning_rate": 9.108108545402155e-06, + "loss": 0.46, + "step": 9871 + }, + { + "epoch": 0.21730134272507154, + "grad_norm": 1.1908012621511184, + "learning_rate": 9.107905343017754e-06, + "loss": 0.2996, + "step": 9872 + }, + { + "epoch": 0.21732335461149022, + "grad_norm": 1.9590743322470838, + "learning_rate": 9.107702119755146e-06, + "loss": 0.281, + "step": 9873 + }, + { + "epoch": 0.21734536649790887, + "grad_norm": 0.6825813216953005, + "learning_rate": 9.107498875615366e-06, + "loss": 0.5338, + "step": 9874 + }, + { + "epoch": 0.21736737838432754, + "grad_norm": 1.2020050958519197, + "learning_rate": 9.107295610599443e-06, + "loss": 0.3272, + "step": 9875 + }, + { + "epoch": 0.2173893902707462, + "grad_norm": 1.6626104920246703, + "learning_rate": 9.107092324708411e-06, + "loss": 0.4334, + "step": 9876 + }, + { + "epoch": 0.21741140215716487, + "grad_norm": 1.4378806631997012, + "learning_rate": 9.106889017943306e-06, + "loss": 0.2934, + "step": 9877 + }, + { + "epoch": 0.21743341404358354, + "grad_norm": 1.4119429270032289, + "learning_rate": 9.106685690305158e-06, + "loss": 0.4162, + "step": 9878 + }, + { + "epoch": 0.2174554259300022, + "grad_norm": 1.8617320303025078, + "learning_rate": 9.106482341795002e-06, + "loss": 0.3707, + "step": 9879 + }, + { + "epoch": 0.21747743781642087, + "grad_norm": 1.5901566165517784, + "learning_rate": 9.10627897241387e-06, + "loss": 0.4024, + "step": 9880 + }, + { + "epoch": 0.21749944970283952, + "grad_norm": 1.6445010343069182, + "learning_rate": 9.1060755821628e-06, + "loss": 0.4128, + "step": 9881 + }, + { + "epoch": 0.2175214615892582, + "grad_norm": 1.5239350858804273, + "learning_rate": 9.10587217104282e-06, + "loss": 0.3812, + "step": 9882 + }, + { + "epoch": 0.21754347347567687, + "grad_norm": 1.7886653480881751, + "learning_rate": 9.105668739054967e-06, + "loss": 0.3814, + "step": 9883 + }, + { + "epoch": 0.21756548536209552, + "grad_norm": 3.3049095394365007, + "learning_rate": 9.105465286200275e-06, + "loss": 0.3928, + "step": 9884 + }, + { + "epoch": 0.2175874972485142, + "grad_norm": 1.6613645283189573, + "learning_rate": 9.105261812479775e-06, + "loss": 0.3266, + "step": 9885 + }, + { + "epoch": 0.21760950913493288, + "grad_norm": 3.546173108160611, + "learning_rate": 9.105058317894506e-06, + "loss": 0.3276, + "step": 9886 + }, + { + "epoch": 0.21763152102135153, + "grad_norm": 1.6205618721545412, + "learning_rate": 9.1048548024455e-06, + "loss": 0.4197, + "step": 9887 + }, + { + "epoch": 0.2176535329077702, + "grad_norm": 1.2518404343927505, + "learning_rate": 9.104651266133788e-06, + "loss": 0.2886, + "step": 9888 + }, + { + "epoch": 0.21767554479418885, + "grad_norm": 0.7029870855001673, + "learning_rate": 9.10444770896041e-06, + "loss": 0.5466, + "step": 9889 + }, + { + "epoch": 0.21769755668060753, + "grad_norm": 1.6340279580332184, + "learning_rate": 9.104244130926398e-06, + "loss": 0.4565, + "step": 9890 + }, + { + "epoch": 0.2177195685670262, + "grad_norm": 1.5606902279803556, + "learning_rate": 9.104040532032786e-06, + "loss": 0.3899, + "step": 9891 + }, + { + "epoch": 0.21774158045344486, + "grad_norm": 1.5482423190169152, + "learning_rate": 9.103836912280611e-06, + "loss": 0.3066, + "step": 9892 + }, + { + "epoch": 0.21776359233986353, + "grad_norm": 1.506705225912775, + "learning_rate": 9.103633271670904e-06, + "loss": 0.3439, + "step": 9893 + }, + { + "epoch": 0.21778560422628218, + "grad_norm": 1.536141713844928, + "learning_rate": 9.103429610204705e-06, + "loss": 0.3313, + "step": 9894 + }, + { + "epoch": 0.21780761611270086, + "grad_norm": 1.8107119479174716, + "learning_rate": 9.103225927883043e-06, + "loss": 0.4371, + "step": 9895 + }, + { + "epoch": 0.21782962799911953, + "grad_norm": 1.9758666277686736, + "learning_rate": 9.10302222470696e-06, + "loss": 0.4889, + "step": 9896 + }, + { + "epoch": 0.21785163988553818, + "grad_norm": 1.9662215796369233, + "learning_rate": 9.102818500677487e-06, + "loss": 0.3369, + "step": 9897 + }, + { + "epoch": 0.21787365177195686, + "grad_norm": 1.407579361531681, + "learning_rate": 9.10261475579566e-06, + "loss": 0.4714, + "step": 9898 + }, + { + "epoch": 0.2178956636583755, + "grad_norm": 1.755857800005374, + "learning_rate": 9.102410990062516e-06, + "loss": 0.3064, + "step": 9899 + }, + { + "epoch": 0.2179176755447942, + "grad_norm": 3.0875785578170447, + "learning_rate": 9.10220720347909e-06, + "loss": 0.2888, + "step": 9900 + }, + { + "epoch": 0.21793968743121286, + "grad_norm": 1.7192414229699051, + "learning_rate": 9.102003396046416e-06, + "loss": 0.3158, + "step": 9901 + }, + { + "epoch": 0.2179616993176315, + "grad_norm": 1.363371243316691, + "learning_rate": 9.101799567765531e-06, + "loss": 0.4348, + "step": 9902 + }, + { + "epoch": 0.2179837112040502, + "grad_norm": 1.530717406740793, + "learning_rate": 9.101595718637473e-06, + "loss": 0.3374, + "step": 9903 + }, + { + "epoch": 0.21800572309046887, + "grad_norm": 0.6993190563227296, + "learning_rate": 9.101391848663275e-06, + "loss": 0.5511, + "step": 9904 + }, + { + "epoch": 0.21802773497688752, + "grad_norm": 1.5492274045008294, + "learning_rate": 9.101187957843974e-06, + "loss": 0.5134, + "step": 9905 + }, + { + "epoch": 0.2180497468633062, + "grad_norm": 1.6473126689238975, + "learning_rate": 9.100984046180608e-06, + "loss": 0.378, + "step": 9906 + }, + { + "epoch": 0.21807175874972484, + "grad_norm": 1.468231825746575, + "learning_rate": 9.10078011367421e-06, + "loss": 0.387, + "step": 9907 + }, + { + "epoch": 0.21809377063614352, + "grad_norm": 1.3251095849796057, + "learning_rate": 9.100576160325819e-06, + "loss": 0.3296, + "step": 9908 + }, + { + "epoch": 0.2181157825225622, + "grad_norm": 0.6390804050720481, + "learning_rate": 9.100372186136473e-06, + "loss": 0.5456, + "step": 9909 + }, + { + "epoch": 0.21813779440898085, + "grad_norm": 2.011601283122005, + "learning_rate": 9.100168191107203e-06, + "loss": 0.4449, + "step": 9910 + }, + { + "epoch": 0.21815980629539952, + "grad_norm": 1.2790135873635826, + "learning_rate": 9.099964175239052e-06, + "loss": 0.2936, + "step": 9911 + }, + { + "epoch": 0.21818181818181817, + "grad_norm": 1.3173708735432048, + "learning_rate": 9.099760138533056e-06, + "loss": 0.2361, + "step": 9912 + }, + { + "epoch": 0.21820383006823685, + "grad_norm": 2.0563728900864873, + "learning_rate": 9.099556080990248e-06, + "loss": 0.3856, + "step": 9913 + }, + { + "epoch": 0.21822584195465553, + "grad_norm": 1.3059356081684734, + "learning_rate": 9.099352002611669e-06, + "loss": 0.3593, + "step": 9914 + }, + { + "epoch": 0.21824785384107417, + "grad_norm": 1.3648115684312043, + "learning_rate": 9.099147903398352e-06, + "loss": 0.3961, + "step": 9915 + }, + { + "epoch": 0.21826986572749285, + "grad_norm": 2.02986087119297, + "learning_rate": 9.09894378335134e-06, + "loss": 0.364, + "step": 9916 + }, + { + "epoch": 0.2182918776139115, + "grad_norm": 1.7293578115156338, + "learning_rate": 9.098739642471666e-06, + "loss": 0.3831, + "step": 9917 + }, + { + "epoch": 0.21831388950033018, + "grad_norm": 1.3048368662957026, + "learning_rate": 9.09853548076037e-06, + "loss": 0.3798, + "step": 9918 + }, + { + "epoch": 0.21833590138674885, + "grad_norm": 1.44331673185262, + "learning_rate": 9.098331298218488e-06, + "loss": 0.4, + "step": 9919 + }, + { + "epoch": 0.2183579132731675, + "grad_norm": 1.5729269974954239, + "learning_rate": 9.098127094847059e-06, + "loss": 0.3353, + "step": 9920 + }, + { + "epoch": 0.21837992515958618, + "grad_norm": 1.5045362021023956, + "learning_rate": 9.09792287064712e-06, + "loss": 0.3935, + "step": 9921 + }, + { + "epoch": 0.21840193704600483, + "grad_norm": 1.8845663695630035, + "learning_rate": 9.097718625619709e-06, + "loss": 0.3525, + "step": 9922 + }, + { + "epoch": 0.2184239489324235, + "grad_norm": 2.80075331022918, + "learning_rate": 9.097514359765864e-06, + "loss": 0.363, + "step": 9923 + }, + { + "epoch": 0.21844596081884218, + "grad_norm": 1.6417249961552232, + "learning_rate": 9.097310073086625e-06, + "loss": 0.3647, + "step": 9924 + }, + { + "epoch": 0.21846797270526083, + "grad_norm": 0.7526302719237353, + "learning_rate": 9.097105765583028e-06, + "loss": 0.5476, + "step": 9925 + }, + { + "epoch": 0.2184899845916795, + "grad_norm": 1.3270638058660118, + "learning_rate": 9.09690143725611e-06, + "loss": 0.3348, + "step": 9926 + }, + { + "epoch": 0.21851199647809819, + "grad_norm": 1.702942327411488, + "learning_rate": 9.096697088106914e-06, + "loss": 0.3277, + "step": 9927 + }, + { + "epoch": 0.21853400836451684, + "grad_norm": 1.3728168434349088, + "learning_rate": 9.096492718136477e-06, + "loss": 0.3164, + "step": 9928 + }, + { + "epoch": 0.2185560202509355, + "grad_norm": 1.534905040155741, + "learning_rate": 9.096288327345835e-06, + "loss": 0.314, + "step": 9929 + }, + { + "epoch": 0.21857803213735416, + "grad_norm": 1.3108552613537316, + "learning_rate": 9.09608391573603e-06, + "loss": 0.3842, + "step": 9930 + }, + { + "epoch": 0.21860004402377284, + "grad_norm": 1.7687343142593008, + "learning_rate": 9.095879483308099e-06, + "loss": 0.4094, + "step": 9931 + }, + { + "epoch": 0.21862205591019152, + "grad_norm": 2.146988697792926, + "learning_rate": 9.095675030063081e-06, + "loss": 0.3286, + "step": 9932 + }, + { + "epoch": 0.21864406779661016, + "grad_norm": 2.222928242219604, + "learning_rate": 9.095470556002018e-06, + "loss": 0.3213, + "step": 9933 + }, + { + "epoch": 0.21866607968302884, + "grad_norm": 1.6544761363114189, + "learning_rate": 9.095266061125944e-06, + "loss": 0.4332, + "step": 9934 + }, + { + "epoch": 0.2186880915694475, + "grad_norm": 1.591420534703079, + "learning_rate": 9.095061545435904e-06, + "loss": 0.2997, + "step": 9935 + }, + { + "epoch": 0.21871010345586617, + "grad_norm": 0.7625689617706111, + "learning_rate": 9.094857008932934e-06, + "loss": 0.5286, + "step": 9936 + }, + { + "epoch": 0.21873211534228484, + "grad_norm": 0.6895588510722558, + "learning_rate": 9.094652451618074e-06, + "loss": 0.537, + "step": 9937 + }, + { + "epoch": 0.2187541272287035, + "grad_norm": 1.4099374226530808, + "learning_rate": 9.094447873492362e-06, + "loss": 0.2969, + "step": 9938 + }, + { + "epoch": 0.21877613911512217, + "grad_norm": 0.6330816099058373, + "learning_rate": 9.094243274556842e-06, + "loss": 0.5448, + "step": 9939 + }, + { + "epoch": 0.21879815100154082, + "grad_norm": 1.226293912768309, + "learning_rate": 9.094038654812552e-06, + "loss": 0.281, + "step": 9940 + }, + { + "epoch": 0.2188201628879595, + "grad_norm": 1.7247079039901472, + "learning_rate": 9.09383401426053e-06, + "loss": 0.3483, + "step": 9941 + }, + { + "epoch": 0.21884217477437817, + "grad_norm": 15.39795198923616, + "learning_rate": 9.093629352901818e-06, + "loss": 0.2966, + "step": 9942 + }, + { + "epoch": 0.21886418666079682, + "grad_norm": 0.8444294959317471, + "learning_rate": 9.093424670737456e-06, + "loss": 0.5442, + "step": 9943 + }, + { + "epoch": 0.2188861985472155, + "grad_norm": 1.650739262672568, + "learning_rate": 9.093219967768485e-06, + "loss": 0.3309, + "step": 9944 + }, + { + "epoch": 0.21890821043363418, + "grad_norm": 1.4203184052432272, + "learning_rate": 9.093015243995943e-06, + "loss": 0.3144, + "step": 9945 + }, + { + "epoch": 0.21893022232005283, + "grad_norm": 1.521455304626575, + "learning_rate": 9.092810499420874e-06, + "loss": 0.3588, + "step": 9946 + }, + { + "epoch": 0.2189522342064715, + "grad_norm": 1.4266225533592736, + "learning_rate": 9.092605734044316e-06, + "loss": 0.3678, + "step": 9947 + }, + { + "epoch": 0.21897424609289015, + "grad_norm": 1.430382347611383, + "learning_rate": 9.09240094786731e-06, + "loss": 0.2639, + "step": 9948 + }, + { + "epoch": 0.21899625797930883, + "grad_norm": 1.3999646863033628, + "learning_rate": 9.092196140890896e-06, + "loss": 0.3027, + "step": 9949 + }, + { + "epoch": 0.2190182698657275, + "grad_norm": 1.8426749817081276, + "learning_rate": 9.091991313116118e-06, + "loss": 0.3698, + "step": 9950 + }, + { + "epoch": 0.21904028175214615, + "grad_norm": 1.3614295748518277, + "learning_rate": 9.091786464544014e-06, + "loss": 0.4154, + "step": 9951 + }, + { + "epoch": 0.21906229363856483, + "grad_norm": 1.641424893448001, + "learning_rate": 9.091581595175624e-06, + "loss": 0.3042, + "step": 9952 + }, + { + "epoch": 0.21908430552498348, + "grad_norm": 1.3664140110796135, + "learning_rate": 9.091376705011995e-06, + "loss": 0.3628, + "step": 9953 + }, + { + "epoch": 0.21910631741140216, + "grad_norm": 10.557791212446004, + "learning_rate": 9.091171794054163e-06, + "loss": 0.4051, + "step": 9954 + }, + { + "epoch": 0.21912832929782083, + "grad_norm": 0.7107349697669125, + "learning_rate": 9.090966862303173e-06, + "loss": 0.5186, + "step": 9955 + }, + { + "epoch": 0.21915034118423948, + "grad_norm": 1.2887341931913694, + "learning_rate": 9.090761909760061e-06, + "loss": 0.3231, + "step": 9956 + }, + { + "epoch": 0.21917235307065816, + "grad_norm": 3.606464366500457, + "learning_rate": 9.090556936425876e-06, + "loss": 0.4037, + "step": 9957 + }, + { + "epoch": 0.2191943649570768, + "grad_norm": 1.3402843920846221, + "learning_rate": 9.090351942301654e-06, + "loss": 0.354, + "step": 9958 + }, + { + "epoch": 0.2192163768434955, + "grad_norm": 1.7708445748868331, + "learning_rate": 9.090146927388441e-06, + "loss": 0.3556, + "step": 9959 + }, + { + "epoch": 0.21923838872991416, + "grad_norm": 4.075424904214933, + "learning_rate": 9.089941891687277e-06, + "loss": 0.4277, + "step": 9960 + }, + { + "epoch": 0.2192604006163328, + "grad_norm": 0.6751662707164544, + "learning_rate": 9.089736835199203e-06, + "loss": 0.5396, + "step": 9961 + }, + { + "epoch": 0.2192824125027515, + "grad_norm": 1.5557217641810552, + "learning_rate": 9.089531757925262e-06, + "loss": 0.3719, + "step": 9962 + }, + { + "epoch": 0.21930442438917014, + "grad_norm": 1.9069857292252355, + "learning_rate": 9.089326659866497e-06, + "loss": 0.3816, + "step": 9963 + }, + { + "epoch": 0.21932643627558882, + "grad_norm": 1.2457467840468461, + "learning_rate": 9.089121541023952e-06, + "loss": 0.2949, + "step": 9964 + }, + { + "epoch": 0.2193484481620075, + "grad_norm": 0.647154916658394, + "learning_rate": 9.088916401398665e-06, + "loss": 0.5325, + "step": 9965 + }, + { + "epoch": 0.21937046004842614, + "grad_norm": 0.6006506326938764, + "learning_rate": 9.088711240991682e-06, + "loss": 0.5283, + "step": 9966 + }, + { + "epoch": 0.21939247193484482, + "grad_norm": 0.629476972083667, + "learning_rate": 9.088506059804046e-06, + "loss": 0.5462, + "step": 9967 + }, + { + "epoch": 0.2194144838212635, + "grad_norm": 1.6821230915486078, + "learning_rate": 9.088300857836796e-06, + "loss": 0.4273, + "step": 9968 + }, + { + "epoch": 0.21943649570768214, + "grad_norm": 1.867290509510484, + "learning_rate": 9.08809563509098e-06, + "loss": 0.3512, + "step": 9969 + }, + { + "epoch": 0.21945850759410082, + "grad_norm": 1.2075847419181625, + "learning_rate": 9.087890391567637e-06, + "loss": 0.3377, + "step": 9970 + }, + { + "epoch": 0.21948051948051947, + "grad_norm": 1.7587891424547513, + "learning_rate": 9.087685127267813e-06, + "loss": 0.2846, + "step": 9971 + }, + { + "epoch": 0.21950253136693815, + "grad_norm": 2.2721914131719814, + "learning_rate": 9.08747984219255e-06, + "loss": 0.3726, + "step": 9972 + }, + { + "epoch": 0.21952454325335682, + "grad_norm": 3.5279712723714436, + "learning_rate": 9.087274536342891e-06, + "loss": 0.4428, + "step": 9973 + }, + { + "epoch": 0.21954655513977547, + "grad_norm": 1.7094130878956262, + "learning_rate": 9.087069209719879e-06, + "loss": 0.3317, + "step": 9974 + }, + { + "epoch": 0.21956856702619415, + "grad_norm": 1.383252692762858, + "learning_rate": 9.08686386232456e-06, + "loss": 0.3235, + "step": 9975 + }, + { + "epoch": 0.2195905789126128, + "grad_norm": 1.9680160608336223, + "learning_rate": 9.086658494157975e-06, + "loss": 0.3922, + "step": 9976 + }, + { + "epoch": 0.21961259079903148, + "grad_norm": 1.9562819533124194, + "learning_rate": 9.086453105221169e-06, + "loss": 0.3018, + "step": 9977 + }, + { + "epoch": 0.21963460268545015, + "grad_norm": 2.0293548032670086, + "learning_rate": 9.086247695515185e-06, + "loss": 0.3407, + "step": 9978 + }, + { + "epoch": 0.2196566145718688, + "grad_norm": 0.6671061312811954, + "learning_rate": 9.086042265041069e-06, + "loss": 0.5462, + "step": 9979 + }, + { + "epoch": 0.21967862645828748, + "grad_norm": 1.9439029935754297, + "learning_rate": 9.085836813799864e-06, + "loss": 0.403, + "step": 9980 + }, + { + "epoch": 0.21970063834470613, + "grad_norm": 1.549552921549335, + "learning_rate": 9.085631341792614e-06, + "loss": 0.3301, + "step": 9981 + }, + { + "epoch": 0.2197226502311248, + "grad_norm": 2.017062687903129, + "learning_rate": 9.085425849020362e-06, + "loss": 0.4293, + "step": 9982 + }, + { + "epoch": 0.21974466211754348, + "grad_norm": 1.4062108388321821, + "learning_rate": 9.085220335484153e-06, + "loss": 0.4082, + "step": 9983 + }, + { + "epoch": 0.21976667400396213, + "grad_norm": 1.720181005660655, + "learning_rate": 9.085014801185034e-06, + "loss": 0.3116, + "step": 9984 + }, + { + "epoch": 0.2197886858903808, + "grad_norm": 1.3147298998043273, + "learning_rate": 9.084809246124047e-06, + "loss": 0.2702, + "step": 9985 + }, + { + "epoch": 0.21981069777679949, + "grad_norm": 1.2981970206799336, + "learning_rate": 9.084603670302238e-06, + "loss": 0.2834, + "step": 9986 + }, + { + "epoch": 0.21983270966321813, + "grad_norm": 1.3681432061735384, + "learning_rate": 9.08439807372065e-06, + "loss": 0.3277, + "step": 9987 + }, + { + "epoch": 0.2198547215496368, + "grad_norm": 0.6658389705510387, + "learning_rate": 9.084192456380332e-06, + "loss": 0.5273, + "step": 9988 + }, + { + "epoch": 0.21987673343605546, + "grad_norm": 1.4753775532568378, + "learning_rate": 9.083986818282325e-06, + "loss": 0.3616, + "step": 9989 + }, + { + "epoch": 0.21989874532247414, + "grad_norm": 2.0197274103807423, + "learning_rate": 9.083781159427673e-06, + "loss": 0.3796, + "step": 9990 + }, + { + "epoch": 0.21992075720889281, + "grad_norm": 1.5227031977681023, + "learning_rate": 9.083575479817428e-06, + "loss": 0.2852, + "step": 9991 + }, + { + "epoch": 0.21994276909531146, + "grad_norm": 1.7656543967585323, + "learning_rate": 9.083369779452629e-06, + "loss": 0.4664, + "step": 9992 + }, + { + "epoch": 0.21996478098173014, + "grad_norm": 2.8772502758664507, + "learning_rate": 9.083164058334323e-06, + "loss": 0.3768, + "step": 9993 + }, + { + "epoch": 0.2199867928681488, + "grad_norm": 1.6625897993890568, + "learning_rate": 9.082958316463557e-06, + "loss": 0.3806, + "step": 9994 + }, + { + "epoch": 0.22000880475456747, + "grad_norm": 1.9292785008914846, + "learning_rate": 9.082752553841375e-06, + "loss": 0.3494, + "step": 9995 + }, + { + "epoch": 0.22003081664098614, + "grad_norm": 1.856588217003316, + "learning_rate": 9.082546770468823e-06, + "loss": 0.3296, + "step": 9996 + }, + { + "epoch": 0.2200528285274048, + "grad_norm": 5.637837080766503, + "learning_rate": 9.082340966346949e-06, + "loss": 0.383, + "step": 9997 + }, + { + "epoch": 0.22007484041382347, + "grad_norm": 1.0867675300672304, + "learning_rate": 9.082135141476798e-06, + "loss": 0.3113, + "step": 9998 + }, + { + "epoch": 0.22009685230024212, + "grad_norm": 2.5595979899247094, + "learning_rate": 9.081929295859414e-06, + "loss": 0.341, + "step": 9999 + }, + { + "epoch": 0.2201188641866608, + "grad_norm": 1.9161870544053115, + "learning_rate": 9.081723429495846e-06, + "loss": 0.3919, + "step": 10000 + }, + { + "epoch": 0.22014087607307947, + "grad_norm": 1.4762608840539835, + "learning_rate": 9.081517542387137e-06, + "loss": 0.2729, + "step": 10001 + }, + { + "epoch": 0.22016288795949812, + "grad_norm": 0.6988127031809532, + "learning_rate": 9.081311634534337e-06, + "loss": 0.5139, + "step": 10002 + }, + { + "epoch": 0.2201848998459168, + "grad_norm": 1.5442656143078797, + "learning_rate": 9.08110570593849e-06, + "loss": 0.2992, + "step": 10003 + }, + { + "epoch": 0.22020691173233545, + "grad_norm": 7.627541648297324, + "learning_rate": 9.080899756600646e-06, + "loss": 0.2077, + "step": 10004 + }, + { + "epoch": 0.22022892361875412, + "grad_norm": 1.4080638061150828, + "learning_rate": 9.080693786521845e-06, + "loss": 0.4804, + "step": 10005 + }, + { + "epoch": 0.2202509355051728, + "grad_norm": 4.009114717801659, + "learning_rate": 9.080487795703142e-06, + "loss": 0.3795, + "step": 10006 + }, + { + "epoch": 0.22027294739159145, + "grad_norm": 3.9052958371377877, + "learning_rate": 9.080281784145579e-06, + "loss": 0.3652, + "step": 10007 + }, + { + "epoch": 0.22029495927801013, + "grad_norm": 2.8253741476753627, + "learning_rate": 9.080075751850203e-06, + "loss": 0.3827, + "step": 10008 + }, + { + "epoch": 0.2203169711644288, + "grad_norm": 0.6680617684596498, + "learning_rate": 9.079869698818063e-06, + "loss": 0.5596, + "step": 10009 + }, + { + "epoch": 0.22033898305084745, + "grad_norm": 1.3357580725796738, + "learning_rate": 9.079663625050205e-06, + "loss": 0.2987, + "step": 10010 + }, + { + "epoch": 0.22036099493726613, + "grad_norm": 2.143175270715665, + "learning_rate": 9.079457530547678e-06, + "loss": 0.3371, + "step": 10011 + }, + { + "epoch": 0.22038300682368478, + "grad_norm": 1.666009187705872, + "learning_rate": 9.079251415311529e-06, + "loss": 0.3104, + "step": 10012 + }, + { + "epoch": 0.22040501871010346, + "grad_norm": 1.4889775990021987, + "learning_rate": 9.079045279342803e-06, + "loss": 0.3117, + "step": 10013 + }, + { + "epoch": 0.22042703059652213, + "grad_norm": 1.620458210047551, + "learning_rate": 9.078839122642551e-06, + "loss": 0.4191, + "step": 10014 + }, + { + "epoch": 0.22044904248294078, + "grad_norm": 1.872018701947339, + "learning_rate": 9.078632945211819e-06, + "loss": 0.408, + "step": 10015 + }, + { + "epoch": 0.22047105436935946, + "grad_norm": 2.2390599684156984, + "learning_rate": 9.078426747051654e-06, + "loss": 0.3246, + "step": 10016 + }, + { + "epoch": 0.2204930662557781, + "grad_norm": 0.6291694148492557, + "learning_rate": 9.078220528163107e-06, + "loss": 0.5346, + "step": 10017 + }, + { + "epoch": 0.22051507814219679, + "grad_norm": 1.4699212658908019, + "learning_rate": 9.078014288547223e-06, + "loss": 0.3199, + "step": 10018 + }, + { + "epoch": 0.22053709002861546, + "grad_norm": 1.637790142182876, + "learning_rate": 9.077808028205052e-06, + "loss": 0.3331, + "step": 10019 + }, + { + "epoch": 0.2205591019150341, + "grad_norm": 1.4583507140014813, + "learning_rate": 9.077601747137641e-06, + "loss": 0.3977, + "step": 10020 + }, + { + "epoch": 0.2205811138014528, + "grad_norm": 1.6458520550268052, + "learning_rate": 9.077395445346041e-06, + "loss": 0.434, + "step": 10021 + }, + { + "epoch": 0.22060312568787144, + "grad_norm": 1.7105401445405908, + "learning_rate": 9.077189122831299e-06, + "loss": 0.304, + "step": 10022 + }, + { + "epoch": 0.22062513757429011, + "grad_norm": 2.7318885813284153, + "learning_rate": 9.076982779594461e-06, + "loss": 0.3871, + "step": 10023 + }, + { + "epoch": 0.2206471494607088, + "grad_norm": 1.4093388962713356, + "learning_rate": 9.076776415636579e-06, + "loss": 0.4299, + "step": 10024 + }, + { + "epoch": 0.22066916134712744, + "grad_norm": 0.6742445892819396, + "learning_rate": 9.076570030958702e-06, + "loss": 0.562, + "step": 10025 + }, + { + "epoch": 0.22069117323354612, + "grad_norm": 1.6118808072735304, + "learning_rate": 9.076363625561877e-06, + "loss": 0.3999, + "step": 10026 + }, + { + "epoch": 0.2207131851199648, + "grad_norm": 1.651159649874218, + "learning_rate": 9.076157199447155e-06, + "loss": 0.2486, + "step": 10027 + }, + { + "epoch": 0.22073519700638344, + "grad_norm": 2.6276467012223503, + "learning_rate": 9.075950752615583e-06, + "loss": 0.3696, + "step": 10028 + }, + { + "epoch": 0.22075720889280212, + "grad_norm": 0.6042112279241693, + "learning_rate": 9.07574428506821e-06, + "loss": 0.5432, + "step": 10029 + }, + { + "epoch": 0.22077922077922077, + "grad_norm": 1.3140925850679426, + "learning_rate": 9.075537796806092e-06, + "loss": 0.3194, + "step": 10030 + }, + { + "epoch": 0.22080123266563945, + "grad_norm": 0.5992998346132583, + "learning_rate": 9.075331287830268e-06, + "loss": 0.5303, + "step": 10031 + }, + { + "epoch": 0.22082324455205812, + "grad_norm": 1.7773221738772342, + "learning_rate": 9.075124758141794e-06, + "loss": 0.3606, + "step": 10032 + }, + { + "epoch": 0.22084525643847677, + "grad_norm": 1.2871430879246661, + "learning_rate": 9.07491820774172e-06, + "loss": 0.3076, + "step": 10033 + }, + { + "epoch": 0.22086726832489545, + "grad_norm": 1.5649448933932626, + "learning_rate": 9.074711636631093e-06, + "loss": 0.3742, + "step": 10034 + }, + { + "epoch": 0.2208892802113141, + "grad_norm": 0.6260983743557058, + "learning_rate": 9.074505044810965e-06, + "loss": 0.5512, + "step": 10035 + }, + { + "epoch": 0.22091129209773278, + "grad_norm": 1.6646839026844351, + "learning_rate": 9.074298432282382e-06, + "loss": 0.298, + "step": 10036 + }, + { + "epoch": 0.22093330398415145, + "grad_norm": 1.3498127477663477, + "learning_rate": 9.0740917990464e-06, + "loss": 0.3102, + "step": 10037 + }, + { + "epoch": 0.2209553158705701, + "grad_norm": 1.7129415896769213, + "learning_rate": 9.073885145104066e-06, + "loss": 0.2746, + "step": 10038 + }, + { + "epoch": 0.22097732775698878, + "grad_norm": 1.9021866602849886, + "learning_rate": 9.073678470456431e-06, + "loss": 0.3346, + "step": 10039 + }, + { + "epoch": 0.22099933964340743, + "grad_norm": 1.2371635464454593, + "learning_rate": 9.073471775104544e-06, + "loss": 0.2605, + "step": 10040 + }, + { + "epoch": 0.2210213515298261, + "grad_norm": 0.6252286401290681, + "learning_rate": 9.073265059049456e-06, + "loss": 0.5214, + "step": 10041 + }, + { + "epoch": 0.22104336341624478, + "grad_norm": 3.2506079987652123, + "learning_rate": 9.07305832229222e-06, + "loss": 0.3465, + "step": 10042 + }, + { + "epoch": 0.22106537530266343, + "grad_norm": 2.4966685391888572, + "learning_rate": 9.072851564833884e-06, + "loss": 0.4005, + "step": 10043 + }, + { + "epoch": 0.2210873871890821, + "grad_norm": 1.391114945841452, + "learning_rate": 9.072644786675501e-06, + "loss": 0.2974, + "step": 10044 + }, + { + "epoch": 0.22110939907550076, + "grad_norm": 1.5602532584788877, + "learning_rate": 9.072437987818118e-06, + "loss": 0.3901, + "step": 10045 + }, + { + "epoch": 0.22113141096191943, + "grad_norm": 2.266085606009571, + "learning_rate": 9.072231168262793e-06, + "loss": 0.3982, + "step": 10046 + }, + { + "epoch": 0.2211534228483381, + "grad_norm": 1.9158945419494484, + "learning_rate": 9.072024328010569e-06, + "loss": 0.3331, + "step": 10047 + }, + { + "epoch": 0.22117543473475676, + "grad_norm": 97.21193728997652, + "learning_rate": 9.071817467062503e-06, + "loss": 0.309, + "step": 10048 + }, + { + "epoch": 0.22119744662117544, + "grad_norm": 2.295839523482496, + "learning_rate": 9.071610585419644e-06, + "loss": 0.2808, + "step": 10049 + }, + { + "epoch": 0.2212194585075941, + "grad_norm": 1.3791064453025688, + "learning_rate": 9.071403683083045e-06, + "loss": 0.2796, + "step": 10050 + }, + { + "epoch": 0.22124147039401276, + "grad_norm": 2.351642256608084, + "learning_rate": 9.071196760053756e-06, + "loss": 0.4691, + "step": 10051 + }, + { + "epoch": 0.22126348228043144, + "grad_norm": 1.3832260532937364, + "learning_rate": 9.07098981633283e-06, + "loss": 0.3053, + "step": 10052 + }, + { + "epoch": 0.2212854941668501, + "grad_norm": 1.4471809032288607, + "learning_rate": 9.070782851921315e-06, + "loss": 0.3545, + "step": 10053 + }, + { + "epoch": 0.22130750605326877, + "grad_norm": 1.2716252899036775, + "learning_rate": 9.07057586682027e-06, + "loss": 0.2813, + "step": 10054 + }, + { + "epoch": 0.22132951793968744, + "grad_norm": 2.7350313403464397, + "learning_rate": 9.07036886103074e-06, + "loss": 0.2959, + "step": 10055 + }, + { + "epoch": 0.2213515298261061, + "grad_norm": 2.2150359001799718, + "learning_rate": 9.070161834553782e-06, + "loss": 0.3143, + "step": 10056 + }, + { + "epoch": 0.22137354171252477, + "grad_norm": 1.9091602157020613, + "learning_rate": 9.069954787390444e-06, + "loss": 0.3547, + "step": 10057 + }, + { + "epoch": 0.22139555359894342, + "grad_norm": 3.1214870297831383, + "learning_rate": 9.069747719541784e-06, + "loss": 0.2762, + "step": 10058 + }, + { + "epoch": 0.2214175654853621, + "grad_norm": 1.7981284116095182, + "learning_rate": 9.069540631008849e-06, + "loss": 0.4376, + "step": 10059 + }, + { + "epoch": 0.22143957737178077, + "grad_norm": 1.539162392507486, + "learning_rate": 9.069333521792692e-06, + "loss": 0.3837, + "step": 10060 + }, + { + "epoch": 0.22146158925819942, + "grad_norm": 1.5424341423994412, + "learning_rate": 9.06912639189437e-06, + "loss": 0.3736, + "step": 10061 + }, + { + "epoch": 0.2214836011446181, + "grad_norm": 1.3997118052483377, + "learning_rate": 9.068919241314929e-06, + "loss": 0.321, + "step": 10062 + }, + { + "epoch": 0.22150561303103675, + "grad_norm": 0.6538339118882642, + "learning_rate": 9.068712070055428e-06, + "loss": 0.5338, + "step": 10063 + }, + { + "epoch": 0.22152762491745542, + "grad_norm": 1.6768458248431029, + "learning_rate": 9.068504878116917e-06, + "loss": 0.3145, + "step": 10064 + }, + { + "epoch": 0.2215496368038741, + "grad_norm": 0.6124427287026585, + "learning_rate": 9.06829766550045e-06, + "loss": 0.5108, + "step": 10065 + }, + { + "epoch": 0.22157164869029275, + "grad_norm": 0.634671221047924, + "learning_rate": 9.068090432207079e-06, + "loss": 0.5173, + "step": 10066 + }, + { + "epoch": 0.22159366057671143, + "grad_norm": 3.3733959404369345, + "learning_rate": 9.06788317823786e-06, + "loss": 0.4236, + "step": 10067 + }, + { + "epoch": 0.2216156724631301, + "grad_norm": 3.2646766645461045, + "learning_rate": 9.067675903593843e-06, + "loss": 0.3387, + "step": 10068 + }, + { + "epoch": 0.22163768434954875, + "grad_norm": 1.3361965324684213, + "learning_rate": 9.067468608276082e-06, + "loss": 0.3609, + "step": 10069 + }, + { + "epoch": 0.22165969623596743, + "grad_norm": 2.1331757861458653, + "learning_rate": 9.067261292285632e-06, + "loss": 0.4493, + "step": 10070 + }, + { + "epoch": 0.22168170812238608, + "grad_norm": 2.1360205892253754, + "learning_rate": 9.067053955623548e-06, + "loss": 0.3277, + "step": 10071 + }, + { + "epoch": 0.22170372000880476, + "grad_norm": 0.6543982012481867, + "learning_rate": 9.066846598290879e-06, + "loss": 0.5205, + "step": 10072 + }, + { + "epoch": 0.22172573189522343, + "grad_norm": 1.5079875904562596, + "learning_rate": 9.066639220288682e-06, + "loss": 0.3216, + "step": 10073 + }, + { + "epoch": 0.22174774378164208, + "grad_norm": 0.6653219670613505, + "learning_rate": 9.066431821618013e-06, + "loss": 0.5457, + "step": 10074 + }, + { + "epoch": 0.22176975566806076, + "grad_norm": 3.250217981402518, + "learning_rate": 9.066224402279921e-06, + "loss": 0.2755, + "step": 10075 + }, + { + "epoch": 0.2217917675544794, + "grad_norm": 1.530169372276316, + "learning_rate": 9.066016962275466e-06, + "loss": 0.3347, + "step": 10076 + }, + { + "epoch": 0.22181377944089808, + "grad_norm": 1.6082179796314464, + "learning_rate": 9.065809501605696e-06, + "loss": 0.3467, + "step": 10077 + }, + { + "epoch": 0.22183579132731676, + "grad_norm": 1.5237379208728241, + "learning_rate": 9.06560202027167e-06, + "loss": 0.5321, + "step": 10078 + }, + { + "epoch": 0.2218578032137354, + "grad_norm": 2.7744148963166517, + "learning_rate": 9.065394518274444e-06, + "loss": 0.2933, + "step": 10079 + }, + { + "epoch": 0.2218798151001541, + "grad_norm": 2.136210228156009, + "learning_rate": 9.065186995615067e-06, + "loss": 0.4243, + "step": 10080 + }, + { + "epoch": 0.22190182698657274, + "grad_norm": 0.692753536526863, + "learning_rate": 9.064979452294597e-06, + "loss": 0.5671, + "step": 10081 + }, + { + "epoch": 0.2219238388729914, + "grad_norm": 1.903458976897811, + "learning_rate": 9.064771888314088e-06, + "loss": 0.4584, + "step": 10082 + }, + { + "epoch": 0.2219458507594101, + "grad_norm": 1.6025212684835783, + "learning_rate": 9.064564303674596e-06, + "loss": 0.3996, + "step": 10083 + }, + { + "epoch": 0.22196786264582874, + "grad_norm": 1.4483922355904937, + "learning_rate": 9.064356698377175e-06, + "loss": 0.3619, + "step": 10084 + }, + { + "epoch": 0.22198987453224742, + "grad_norm": 1.583157464716975, + "learning_rate": 9.064149072422882e-06, + "loss": 0.3758, + "step": 10085 + }, + { + "epoch": 0.2220118864186661, + "grad_norm": 2.4383768953329845, + "learning_rate": 9.063941425812768e-06, + "loss": 0.3862, + "step": 10086 + }, + { + "epoch": 0.22203389830508474, + "grad_norm": 1.4356735509183385, + "learning_rate": 9.063733758547892e-06, + "loss": 0.361, + "step": 10087 + }, + { + "epoch": 0.22205591019150342, + "grad_norm": 0.648116407880609, + "learning_rate": 9.063526070629309e-06, + "loss": 0.5352, + "step": 10088 + }, + { + "epoch": 0.22207792207792207, + "grad_norm": 1.2970047090632637, + "learning_rate": 9.063318362058075e-06, + "loss": 0.2777, + "step": 10089 + }, + { + "epoch": 0.22209993396434075, + "grad_norm": 1.3655028938825504, + "learning_rate": 9.063110632835244e-06, + "loss": 0.3183, + "step": 10090 + }, + { + "epoch": 0.22212194585075942, + "grad_norm": 0.6394459191502432, + "learning_rate": 9.062902882961872e-06, + "loss": 0.5186, + "step": 10091 + }, + { + "epoch": 0.22214395773717807, + "grad_norm": 1.807023770063429, + "learning_rate": 9.062695112439014e-06, + "loss": 0.3557, + "step": 10092 + }, + { + "epoch": 0.22216596962359675, + "grad_norm": 1.1972218554339973, + "learning_rate": 9.062487321267731e-06, + "loss": 0.3183, + "step": 10093 + }, + { + "epoch": 0.2221879815100154, + "grad_norm": 1.476997681107885, + "learning_rate": 9.062279509449075e-06, + "loss": 0.3469, + "step": 10094 + }, + { + "epoch": 0.22220999339643407, + "grad_norm": 1.4761827477498854, + "learning_rate": 9.062071676984102e-06, + "loss": 0.2877, + "step": 10095 + }, + { + "epoch": 0.22223200528285275, + "grad_norm": 0.6277225892864858, + "learning_rate": 9.061863823873866e-06, + "loss": 0.5001, + "step": 10096 + }, + { + "epoch": 0.2222540171692714, + "grad_norm": 1.6009453726351146, + "learning_rate": 9.06165595011943e-06, + "loss": 0.3616, + "step": 10097 + }, + { + "epoch": 0.22227602905569008, + "grad_norm": 2.104759622492241, + "learning_rate": 9.061448055721846e-06, + "loss": 0.3083, + "step": 10098 + }, + { + "epoch": 0.22229804094210873, + "grad_norm": 1.514181762745054, + "learning_rate": 9.06124014068217e-06, + "loss": 0.3245, + "step": 10099 + }, + { + "epoch": 0.2223200528285274, + "grad_norm": 0.6514706854440839, + "learning_rate": 9.061032205001463e-06, + "loss": 0.5323, + "step": 10100 + }, + { + "epoch": 0.22234206471494608, + "grad_norm": 2.964714326259931, + "learning_rate": 9.060824248680778e-06, + "loss": 0.3736, + "step": 10101 + }, + { + "epoch": 0.22236407660136473, + "grad_norm": 1.3244785983529088, + "learning_rate": 9.060616271721173e-06, + "loss": 0.3653, + "step": 10102 + }, + { + "epoch": 0.2223860884877834, + "grad_norm": 2.617189157909431, + "learning_rate": 9.060408274123705e-06, + "loss": 0.3527, + "step": 10103 + }, + { + "epoch": 0.22240810037420206, + "grad_norm": 2.03004661630838, + "learning_rate": 9.060200255889431e-06, + "loss": 0.3878, + "step": 10104 + }, + { + "epoch": 0.22243011226062073, + "grad_norm": 0.6351106485904559, + "learning_rate": 9.059992217019407e-06, + "loss": 0.5211, + "step": 10105 + }, + { + "epoch": 0.2224521241470394, + "grad_norm": 0.6312951747596277, + "learning_rate": 9.059784157514695e-06, + "loss": 0.541, + "step": 10106 + }, + { + "epoch": 0.22247413603345806, + "grad_norm": 1.3341362861967676, + "learning_rate": 9.059576077376346e-06, + "loss": 0.2786, + "step": 10107 + }, + { + "epoch": 0.22249614791987674, + "grad_norm": 3.4470631129935607, + "learning_rate": 9.059367976605422e-06, + "loss": 0.3593, + "step": 10108 + }, + { + "epoch": 0.2225181598062954, + "grad_norm": 1.6643372624205353, + "learning_rate": 9.059159855202979e-06, + "loss": 0.2859, + "step": 10109 + }, + { + "epoch": 0.22254017169271406, + "grad_norm": 1.4009976751680953, + "learning_rate": 9.058951713170076e-06, + "loss": 0.3341, + "step": 10110 + }, + { + "epoch": 0.22256218357913274, + "grad_norm": 0.6820377803901284, + "learning_rate": 9.058743550507769e-06, + "loss": 0.5349, + "step": 10111 + }, + { + "epoch": 0.2225841954655514, + "grad_norm": 1.5028191325544031, + "learning_rate": 9.058535367217118e-06, + "loss": 0.2954, + "step": 10112 + }, + { + "epoch": 0.22260620735197006, + "grad_norm": 0.6553679201244799, + "learning_rate": 9.05832716329918e-06, + "loss": 0.5579, + "step": 10113 + }, + { + "epoch": 0.22262821923838874, + "grad_norm": 1.5311270668416779, + "learning_rate": 9.058118938755014e-06, + "loss": 0.2765, + "step": 10114 + }, + { + "epoch": 0.2226502311248074, + "grad_norm": 1.997500349469286, + "learning_rate": 9.057910693585676e-06, + "loss": 0.3021, + "step": 10115 + }, + { + "epoch": 0.22267224301122607, + "grad_norm": 1.5170754974397205, + "learning_rate": 9.057702427792227e-06, + "loss": 0.3333, + "step": 10116 + }, + { + "epoch": 0.22269425489764472, + "grad_norm": 0.6341683486418225, + "learning_rate": 9.057494141375724e-06, + "loss": 0.5353, + "step": 10117 + }, + { + "epoch": 0.2227162667840634, + "grad_norm": 1.3637780126033534, + "learning_rate": 9.057285834337226e-06, + "loss": 0.3284, + "step": 10118 + }, + { + "epoch": 0.22273827867048207, + "grad_norm": 2.3476362347277777, + "learning_rate": 9.057077506677791e-06, + "loss": 0.3366, + "step": 10119 + }, + { + "epoch": 0.22276029055690072, + "grad_norm": 2.7618344012967895, + "learning_rate": 9.05686915839848e-06, + "loss": 0.4525, + "step": 10120 + }, + { + "epoch": 0.2227823024433194, + "grad_norm": 1.5946624484354166, + "learning_rate": 9.056660789500352e-06, + "loss": 0.3903, + "step": 10121 + }, + { + "epoch": 0.22280431432973805, + "grad_norm": 0.6398314427384993, + "learning_rate": 9.056452399984461e-06, + "loss": 0.5313, + "step": 10122 + }, + { + "epoch": 0.22282632621615672, + "grad_norm": 1.664320854107258, + "learning_rate": 9.056243989851871e-06, + "loss": 0.4531, + "step": 10123 + }, + { + "epoch": 0.2228483381025754, + "grad_norm": 2.078146590964168, + "learning_rate": 9.056035559103641e-06, + "loss": 0.3941, + "step": 10124 + }, + { + "epoch": 0.22287034998899405, + "grad_norm": 1.5983928039883386, + "learning_rate": 9.055827107740829e-06, + "loss": 0.2922, + "step": 10125 + }, + { + "epoch": 0.22289236187541273, + "grad_norm": 1.5375807633520904, + "learning_rate": 9.055618635764493e-06, + "loss": 0.404, + "step": 10126 + }, + { + "epoch": 0.2229143737618314, + "grad_norm": 0.617159182530267, + "learning_rate": 9.055410143175696e-06, + "loss": 0.4991, + "step": 10127 + }, + { + "epoch": 0.22293638564825005, + "grad_norm": 0.6137845339970803, + "learning_rate": 9.055201629975495e-06, + "loss": 0.5253, + "step": 10128 + }, + { + "epoch": 0.22295839753466873, + "grad_norm": 1.5842952323378847, + "learning_rate": 9.054993096164951e-06, + "loss": 0.3545, + "step": 10129 + }, + { + "epoch": 0.22298040942108738, + "grad_norm": 1.8351207673705638, + "learning_rate": 9.054784541745124e-06, + "loss": 0.3933, + "step": 10130 + }, + { + "epoch": 0.22300242130750605, + "grad_norm": 5.20426709094203, + "learning_rate": 9.054575966717073e-06, + "loss": 0.2353, + "step": 10131 + }, + { + "epoch": 0.22302443319392473, + "grad_norm": 1.6526818709425437, + "learning_rate": 9.054367371081859e-06, + "loss": 0.4031, + "step": 10132 + }, + { + "epoch": 0.22304644508034338, + "grad_norm": 2.392967188133186, + "learning_rate": 9.05415875484054e-06, + "loss": 0.3425, + "step": 10133 + }, + { + "epoch": 0.22306845696676206, + "grad_norm": 0.6080855759519272, + "learning_rate": 9.05395011799418e-06, + "loss": 0.5143, + "step": 10134 + }, + { + "epoch": 0.2230904688531807, + "grad_norm": 0.6372088855588453, + "learning_rate": 9.053741460543836e-06, + "loss": 0.5405, + "step": 10135 + }, + { + "epoch": 0.22311248073959938, + "grad_norm": 36.46187804489783, + "learning_rate": 9.053532782490571e-06, + "loss": 0.3165, + "step": 10136 + }, + { + "epoch": 0.22313449262601806, + "grad_norm": 1.4705409034671895, + "learning_rate": 9.053324083835444e-06, + "loss": 0.3183, + "step": 10137 + }, + { + "epoch": 0.2231565045124367, + "grad_norm": 2.262266558445703, + "learning_rate": 9.053115364579515e-06, + "loss": 0.3319, + "step": 10138 + }, + { + "epoch": 0.2231785163988554, + "grad_norm": 2.199743494583059, + "learning_rate": 9.052906624723847e-06, + "loss": 0.3932, + "step": 10139 + }, + { + "epoch": 0.22320052828527404, + "grad_norm": 1.5880402056224685, + "learning_rate": 9.052697864269501e-06, + "loss": 0.3125, + "step": 10140 + }, + { + "epoch": 0.2232225401716927, + "grad_norm": 1.2113467903735007, + "learning_rate": 9.052489083217535e-06, + "loss": 0.2632, + "step": 10141 + }, + { + "epoch": 0.2232445520581114, + "grad_norm": 7.986189699280344, + "learning_rate": 9.052280281569014e-06, + "loss": 0.3068, + "step": 10142 + }, + { + "epoch": 0.22326656394453004, + "grad_norm": 2.4597050435349206, + "learning_rate": 9.052071459324996e-06, + "loss": 0.4531, + "step": 10143 + }, + { + "epoch": 0.22328857583094872, + "grad_norm": 0.7006453006952831, + "learning_rate": 9.051862616486544e-06, + "loss": 0.5182, + "step": 10144 + }, + { + "epoch": 0.22331058771736736, + "grad_norm": 1.4652501743823796, + "learning_rate": 9.051653753054718e-06, + "loss": 0.2192, + "step": 10145 + }, + { + "epoch": 0.22333259960378604, + "grad_norm": 1.3898572812615233, + "learning_rate": 9.05144486903058e-06, + "loss": 0.2982, + "step": 10146 + }, + { + "epoch": 0.22335461149020472, + "grad_norm": 1.4214951872728885, + "learning_rate": 9.051235964415195e-06, + "loss": 0.3045, + "step": 10147 + }, + { + "epoch": 0.22337662337662337, + "grad_norm": 1.3298669241115675, + "learning_rate": 9.05102703920962e-06, + "loss": 0.2286, + "step": 10148 + }, + { + "epoch": 0.22339863526304204, + "grad_norm": 2.1005179748887937, + "learning_rate": 9.05081809341492e-06, + "loss": 0.3543, + "step": 10149 + }, + { + "epoch": 0.22342064714946072, + "grad_norm": 1.480981901305324, + "learning_rate": 9.050609127032154e-06, + "loss": 0.3378, + "step": 10150 + }, + { + "epoch": 0.22344265903587937, + "grad_norm": 1.5650290859897193, + "learning_rate": 9.050400140062387e-06, + "loss": 0.3691, + "step": 10151 + }, + { + "epoch": 0.22346467092229805, + "grad_norm": 2.017261827324005, + "learning_rate": 9.050191132506679e-06, + "loss": 0.365, + "step": 10152 + }, + { + "epoch": 0.2234866828087167, + "grad_norm": 1.2120920240424446, + "learning_rate": 9.049982104366094e-06, + "loss": 0.2474, + "step": 10153 + }, + { + "epoch": 0.22350869469513537, + "grad_norm": 2.377176057491611, + "learning_rate": 9.049773055641693e-06, + "loss": 0.3266, + "step": 10154 + }, + { + "epoch": 0.22353070658155405, + "grad_norm": 1.7114271130352103, + "learning_rate": 9.04956398633454e-06, + "loss": 0.3496, + "step": 10155 + }, + { + "epoch": 0.2235527184679727, + "grad_norm": 1.480816160765158, + "learning_rate": 9.049354896445696e-06, + "loss": 0.3198, + "step": 10156 + }, + { + "epoch": 0.22357473035439138, + "grad_norm": 1.87571018832228, + "learning_rate": 9.049145785976225e-06, + "loss": 0.3617, + "step": 10157 + }, + { + "epoch": 0.22359674224081003, + "grad_norm": 2.0797563252012274, + "learning_rate": 9.048936654927189e-06, + "loss": 0.4222, + "step": 10158 + }, + { + "epoch": 0.2236187541272287, + "grad_norm": 1.2804943128698751, + "learning_rate": 9.04872750329965e-06, + "loss": 0.2868, + "step": 10159 + }, + { + "epoch": 0.22364076601364738, + "grad_norm": 1.357664777263253, + "learning_rate": 9.048518331094673e-06, + "loss": 0.2665, + "step": 10160 + }, + { + "epoch": 0.22366277790006603, + "grad_norm": 0.708725945396796, + "learning_rate": 9.04830913831332e-06, + "loss": 0.5216, + "step": 10161 + }, + { + "epoch": 0.2236847897864847, + "grad_norm": 1.4180054758354232, + "learning_rate": 9.048099924956654e-06, + "loss": 0.2809, + "step": 10162 + }, + { + "epoch": 0.22370680167290335, + "grad_norm": 0.6566783873031139, + "learning_rate": 9.04789069102574e-06, + "loss": 0.5398, + "step": 10163 + }, + { + "epoch": 0.22372881355932203, + "grad_norm": 1.9849126889062885, + "learning_rate": 9.047681436521637e-06, + "loss": 0.3333, + "step": 10164 + }, + { + "epoch": 0.2237508254457407, + "grad_norm": 2.057334175893536, + "learning_rate": 9.047472161445416e-06, + "loss": 0.2934, + "step": 10165 + }, + { + "epoch": 0.22377283733215936, + "grad_norm": 1.4759464827865085, + "learning_rate": 9.047262865798134e-06, + "loss": 0.3147, + "step": 10166 + }, + { + "epoch": 0.22379484921857803, + "grad_norm": 1.7398080122840534, + "learning_rate": 9.047053549580855e-06, + "loss": 0.3847, + "step": 10167 + }, + { + "epoch": 0.2238168611049967, + "grad_norm": 1.8300048349963693, + "learning_rate": 9.046844212794648e-06, + "loss": 0.329, + "step": 10168 + }, + { + "epoch": 0.22383887299141536, + "grad_norm": 0.7900154249777686, + "learning_rate": 9.046634855440573e-06, + "loss": 0.5303, + "step": 10169 + }, + { + "epoch": 0.22386088487783404, + "grad_norm": 1.7053032048290147, + "learning_rate": 9.046425477519693e-06, + "loss": 0.3558, + "step": 10170 + }, + { + "epoch": 0.2238828967642527, + "grad_norm": 1.5447151215108175, + "learning_rate": 9.046216079033077e-06, + "loss": 0.3351, + "step": 10171 + }, + { + "epoch": 0.22390490865067136, + "grad_norm": 1.2470104701397873, + "learning_rate": 9.046006659981783e-06, + "loss": 0.3539, + "step": 10172 + }, + { + "epoch": 0.22392692053709004, + "grad_norm": 2.509886310441578, + "learning_rate": 9.04579722036688e-06, + "loss": 0.3198, + "step": 10173 + }, + { + "epoch": 0.2239489324235087, + "grad_norm": 1.3845176200701625, + "learning_rate": 9.045587760189432e-06, + "loss": 0.3999, + "step": 10174 + }, + { + "epoch": 0.22397094430992737, + "grad_norm": 1.4504207017909272, + "learning_rate": 9.045378279450502e-06, + "loss": 0.3501, + "step": 10175 + }, + { + "epoch": 0.22399295619634602, + "grad_norm": 1.3432067767981173, + "learning_rate": 9.045168778151154e-06, + "loss": 0.3465, + "step": 10176 + }, + { + "epoch": 0.2240149680827647, + "grad_norm": 1.7008398951267845, + "learning_rate": 9.044959256292456e-06, + "loss": 0.187, + "step": 10177 + }, + { + "epoch": 0.22403697996918337, + "grad_norm": 0.6987839846968671, + "learning_rate": 9.04474971387547e-06, + "loss": 0.5388, + "step": 10178 + }, + { + "epoch": 0.22405899185560202, + "grad_norm": 1.5449750143237355, + "learning_rate": 9.044540150901261e-06, + "loss": 0.312, + "step": 10179 + }, + { + "epoch": 0.2240810037420207, + "grad_norm": 0.6429571096875845, + "learning_rate": 9.044330567370897e-06, + "loss": 0.5233, + "step": 10180 + }, + { + "epoch": 0.22410301562843934, + "grad_norm": 1.5477614175675805, + "learning_rate": 9.044120963285441e-06, + "loss": 0.3842, + "step": 10181 + }, + { + "epoch": 0.22412502751485802, + "grad_norm": 1.718936948913595, + "learning_rate": 9.043911338645958e-06, + "loss": 0.3394, + "step": 10182 + }, + { + "epoch": 0.2241470394012767, + "grad_norm": 2.608516223360202, + "learning_rate": 9.043701693453513e-06, + "loss": 0.4367, + "step": 10183 + }, + { + "epoch": 0.22416905128769535, + "grad_norm": 1.7647311070368639, + "learning_rate": 9.043492027709172e-06, + "loss": 0.3465, + "step": 10184 + }, + { + "epoch": 0.22419106317411402, + "grad_norm": 1.3706151906338377, + "learning_rate": 9.043282341414003e-06, + "loss": 0.2844, + "step": 10185 + }, + { + "epoch": 0.22421307506053267, + "grad_norm": 1.283582524920325, + "learning_rate": 9.04307263456907e-06, + "loss": 0.2885, + "step": 10186 + }, + { + "epoch": 0.22423508694695135, + "grad_norm": 1.5307946447163356, + "learning_rate": 9.042862907175438e-06, + "loss": 0.3416, + "step": 10187 + }, + { + "epoch": 0.22425709883337003, + "grad_norm": 34.64722937971199, + "learning_rate": 9.042653159234173e-06, + "loss": 0.3586, + "step": 10188 + }, + { + "epoch": 0.22427911071978868, + "grad_norm": 1.4349558898862966, + "learning_rate": 9.042443390746342e-06, + "loss": 0.3468, + "step": 10189 + }, + { + "epoch": 0.22430112260620735, + "grad_norm": 2.3023096303248622, + "learning_rate": 9.04223360171301e-06, + "loss": 0.3419, + "step": 10190 + }, + { + "epoch": 0.22432313449262603, + "grad_norm": 1.6572524237737334, + "learning_rate": 9.042023792135245e-06, + "loss": 0.3349, + "step": 10191 + }, + { + "epoch": 0.22434514637904468, + "grad_norm": 1.5893269612321401, + "learning_rate": 9.041813962014112e-06, + "loss": 0.2696, + "step": 10192 + }, + { + "epoch": 0.22436715826546336, + "grad_norm": 2.088671018992908, + "learning_rate": 9.041604111350678e-06, + "loss": 0.4174, + "step": 10193 + }, + { + "epoch": 0.224389170151882, + "grad_norm": 0.7701723169349325, + "learning_rate": 9.041394240146009e-06, + "loss": 0.5268, + "step": 10194 + }, + { + "epoch": 0.22441118203830068, + "grad_norm": 1.2535585183851143, + "learning_rate": 9.04118434840117e-06, + "loss": 0.3839, + "step": 10195 + }, + { + "epoch": 0.22443319392471936, + "grad_norm": 3.443485655988886, + "learning_rate": 9.040974436117233e-06, + "loss": 0.4574, + "step": 10196 + }, + { + "epoch": 0.224455205811138, + "grad_norm": 1.6531501985819292, + "learning_rate": 9.04076450329526e-06, + "loss": 0.3919, + "step": 10197 + }, + { + "epoch": 0.22447721769755669, + "grad_norm": 1.6164674116431177, + "learning_rate": 9.040554549936321e-06, + "loss": 0.3048, + "step": 10198 + }, + { + "epoch": 0.22449922958397533, + "grad_norm": 2.6710326578130803, + "learning_rate": 9.040344576041479e-06, + "loss": 0.2799, + "step": 10199 + }, + { + "epoch": 0.224521241470394, + "grad_norm": 1.396040654308244, + "learning_rate": 9.040134581611805e-06, + "loss": 0.3716, + "step": 10200 + }, + { + "epoch": 0.2245432533568127, + "grad_norm": 1.535587979030425, + "learning_rate": 9.039924566648366e-06, + "loss": 0.2819, + "step": 10201 + }, + { + "epoch": 0.22456526524323134, + "grad_norm": 1.2321840798689767, + "learning_rate": 9.039714531152226e-06, + "loss": 0.2885, + "step": 10202 + }, + { + "epoch": 0.22458727712965001, + "grad_norm": 1.2464238667828564, + "learning_rate": 9.039504475124458e-06, + "loss": 0.3484, + "step": 10203 + }, + { + "epoch": 0.22460928901606866, + "grad_norm": 1.4317844453230024, + "learning_rate": 9.039294398566126e-06, + "loss": 0.3435, + "step": 10204 + }, + { + "epoch": 0.22463130090248734, + "grad_norm": 1.5750438423630893, + "learning_rate": 9.039084301478295e-06, + "loss": 0.4713, + "step": 10205 + }, + { + "epoch": 0.22465331278890602, + "grad_norm": 1.5809064359567901, + "learning_rate": 9.038874183862042e-06, + "loss": 0.4046, + "step": 10206 + }, + { + "epoch": 0.22467532467532467, + "grad_norm": 1.5982407485148455, + "learning_rate": 9.038664045718424e-06, + "loss": 0.3161, + "step": 10207 + }, + { + "epoch": 0.22469733656174334, + "grad_norm": 1.210354611936307, + "learning_rate": 9.038453887048515e-06, + "loss": 0.3397, + "step": 10208 + }, + { + "epoch": 0.22471934844816202, + "grad_norm": 1.2730345066219968, + "learning_rate": 9.038243707853384e-06, + "loss": 0.3846, + "step": 10209 + }, + { + "epoch": 0.22474136033458067, + "grad_norm": 3.09537192222246, + "learning_rate": 9.038033508134097e-06, + "loss": 0.4191, + "step": 10210 + }, + { + "epoch": 0.22476337222099935, + "grad_norm": 5.528577325407171, + "learning_rate": 9.037823287891721e-06, + "loss": 0.4121, + "step": 10211 + }, + { + "epoch": 0.224785384107418, + "grad_norm": 1.3324471858593347, + "learning_rate": 9.037613047127328e-06, + "loss": 0.3521, + "step": 10212 + }, + { + "epoch": 0.22480739599383667, + "grad_norm": 1.1587217183783778, + "learning_rate": 9.037402785841982e-06, + "loss": 0.2703, + "step": 10213 + }, + { + "epoch": 0.22482940788025535, + "grad_norm": 1.5739747279211553, + "learning_rate": 9.037192504036757e-06, + "loss": 0.3751, + "step": 10214 + }, + { + "epoch": 0.224851419766674, + "grad_norm": 1.575330468784836, + "learning_rate": 9.036982201712718e-06, + "loss": 0.428, + "step": 10215 + }, + { + "epoch": 0.22487343165309268, + "grad_norm": 1.7000599462366133, + "learning_rate": 9.036771878870934e-06, + "loss": 0.4046, + "step": 10216 + }, + { + "epoch": 0.22489544353951132, + "grad_norm": 0.7911963216682838, + "learning_rate": 9.036561535512475e-06, + "loss": 0.5246, + "step": 10217 + }, + { + "epoch": 0.22491745542593, + "grad_norm": 1.3342912867261536, + "learning_rate": 9.036351171638413e-06, + "loss": 0.3248, + "step": 10218 + }, + { + "epoch": 0.22493946731234868, + "grad_norm": 1.2554046947887652, + "learning_rate": 9.036140787249812e-06, + "loss": 0.3043, + "step": 10219 + }, + { + "epoch": 0.22496147919876733, + "grad_norm": 1.9203019385316196, + "learning_rate": 9.035930382347741e-06, + "loss": 0.4065, + "step": 10220 + }, + { + "epoch": 0.224983491085186, + "grad_norm": 1.7057015105038145, + "learning_rate": 9.035719956933276e-06, + "loss": 0.3825, + "step": 10221 + }, + { + "epoch": 0.22500550297160465, + "grad_norm": 1.37268894719713, + "learning_rate": 9.03550951100748e-06, + "loss": 0.3924, + "step": 10222 + }, + { + "epoch": 0.22502751485802333, + "grad_norm": 1.6754720244841028, + "learning_rate": 9.035299044571425e-06, + "loss": 0.3787, + "step": 10223 + }, + { + "epoch": 0.225049526744442, + "grad_norm": 0.647530687567779, + "learning_rate": 9.03508855762618e-06, + "loss": 0.5214, + "step": 10224 + }, + { + "epoch": 0.22507153863086066, + "grad_norm": 1.2262255852304702, + "learning_rate": 9.034878050172815e-06, + "loss": 0.3415, + "step": 10225 + }, + { + "epoch": 0.22509355051727933, + "grad_norm": 0.6250362773156141, + "learning_rate": 9.034667522212402e-06, + "loss": 0.5196, + "step": 10226 + }, + { + "epoch": 0.225115562403698, + "grad_norm": 1.7200130068404966, + "learning_rate": 9.034456973746007e-06, + "loss": 0.2595, + "step": 10227 + }, + { + "epoch": 0.22513757429011666, + "grad_norm": 4.017436717324413, + "learning_rate": 9.034246404774704e-06, + "loss": 0.3824, + "step": 10228 + }, + { + "epoch": 0.22515958617653534, + "grad_norm": 1.720690404317123, + "learning_rate": 9.03403581529956e-06, + "loss": 0.3545, + "step": 10229 + }, + { + "epoch": 0.22518159806295399, + "grad_norm": 1.2843600163622375, + "learning_rate": 9.033825205321647e-06, + "loss": 0.3715, + "step": 10230 + }, + { + "epoch": 0.22520360994937266, + "grad_norm": 2.5467544076134887, + "learning_rate": 9.033614574842036e-06, + "loss": 0.3136, + "step": 10231 + }, + { + "epoch": 0.22522562183579134, + "grad_norm": 2.4462670951330363, + "learning_rate": 9.033403923861797e-06, + "loss": 0.3805, + "step": 10232 + }, + { + "epoch": 0.22524763372221, + "grad_norm": 0.6470908202973527, + "learning_rate": 9.033193252381999e-06, + "loss": 0.5486, + "step": 10233 + }, + { + "epoch": 0.22526964560862867, + "grad_norm": 1.5789517503584574, + "learning_rate": 9.032982560403714e-06, + "loss": 0.3361, + "step": 10234 + }, + { + "epoch": 0.22529165749504731, + "grad_norm": 2.3731718473387766, + "learning_rate": 9.032771847928015e-06, + "loss": 0.3903, + "step": 10235 + }, + { + "epoch": 0.225313669381466, + "grad_norm": 0.627032762231124, + "learning_rate": 9.032561114955968e-06, + "loss": 0.5003, + "step": 10236 + }, + { + "epoch": 0.22533568126788467, + "grad_norm": 1.8371680598297295, + "learning_rate": 9.032350361488649e-06, + "loss": 0.2974, + "step": 10237 + }, + { + "epoch": 0.22535769315430332, + "grad_norm": 1.6212203440552928, + "learning_rate": 9.032139587527126e-06, + "loss": 0.4355, + "step": 10238 + }, + { + "epoch": 0.225379705040722, + "grad_norm": 3.0831984961206333, + "learning_rate": 9.03192879307247e-06, + "loss": 0.3339, + "step": 10239 + }, + { + "epoch": 0.22540171692714064, + "grad_norm": 4.661403842985619, + "learning_rate": 9.031717978125757e-06, + "loss": 0.3178, + "step": 10240 + }, + { + "epoch": 0.22542372881355932, + "grad_norm": 1.3903523896528507, + "learning_rate": 9.031507142688052e-06, + "loss": 0.29, + "step": 10241 + }, + { + "epoch": 0.225445740699978, + "grad_norm": 1.2973430194224083, + "learning_rate": 9.031296286760431e-06, + "loss": 0.3053, + "step": 10242 + }, + { + "epoch": 0.22546775258639665, + "grad_norm": 1.5691980116079824, + "learning_rate": 9.031085410343962e-06, + "loss": 0.2963, + "step": 10243 + }, + { + "epoch": 0.22548976447281532, + "grad_norm": 1.4390178992651634, + "learning_rate": 9.030874513439723e-06, + "loss": 0.4168, + "step": 10244 + }, + { + "epoch": 0.22551177635923397, + "grad_norm": 1.7592947391731755, + "learning_rate": 9.03066359604878e-06, + "loss": 0.3752, + "step": 10245 + }, + { + "epoch": 0.22553378824565265, + "grad_norm": 0.9135253769164018, + "learning_rate": 9.030452658172205e-06, + "loss": 0.5166, + "step": 10246 + }, + { + "epoch": 0.22555580013207133, + "grad_norm": 1.482334879764151, + "learning_rate": 9.030241699811073e-06, + "loss": 0.4179, + "step": 10247 + }, + { + "epoch": 0.22557781201848998, + "grad_norm": 1.9485364074920242, + "learning_rate": 9.030030720966456e-06, + "loss": 0.3607, + "step": 10248 + }, + { + "epoch": 0.22559982390490865, + "grad_norm": 1.6374536003227815, + "learning_rate": 9.029819721639426e-06, + "loss": 0.4111, + "step": 10249 + }, + { + "epoch": 0.22562183579132733, + "grad_norm": 1.8366290587913234, + "learning_rate": 9.029608701831054e-06, + "loss": 0.3294, + "step": 10250 + }, + { + "epoch": 0.22564384767774598, + "grad_norm": 2.0686099328009155, + "learning_rate": 9.029397661542412e-06, + "loss": 0.387, + "step": 10251 + }, + { + "epoch": 0.22566585956416466, + "grad_norm": 1.9040325918290544, + "learning_rate": 9.029186600774576e-06, + "loss": 0.3726, + "step": 10252 + }, + { + "epoch": 0.2256878714505833, + "grad_norm": 0.6601313083787024, + "learning_rate": 9.028975519528615e-06, + "loss": 0.5185, + "step": 10253 + }, + { + "epoch": 0.22570988333700198, + "grad_norm": 1.4463999778326067, + "learning_rate": 9.028764417805605e-06, + "loss": 0.3883, + "step": 10254 + }, + { + "epoch": 0.22573189522342066, + "grad_norm": 1.3611255472737336, + "learning_rate": 9.028553295606616e-06, + "loss": 0.3062, + "step": 10255 + }, + { + "epoch": 0.2257539071098393, + "grad_norm": 1.407761363233479, + "learning_rate": 9.028342152932723e-06, + "loss": 0.4297, + "step": 10256 + }, + { + "epoch": 0.22577591899625798, + "grad_norm": 2.2445325818003217, + "learning_rate": 9.028130989785e-06, + "loss": 0.2566, + "step": 10257 + }, + { + "epoch": 0.22579793088267663, + "grad_norm": 0.6909588401838821, + "learning_rate": 9.027919806164516e-06, + "loss": 0.5388, + "step": 10258 + }, + { + "epoch": 0.2258199427690953, + "grad_norm": 1.8599525298235766, + "learning_rate": 9.027708602072349e-06, + "loss": 0.3719, + "step": 10259 + }, + { + "epoch": 0.225841954655514, + "grad_norm": 1.5491522762448617, + "learning_rate": 9.027497377509569e-06, + "loss": 0.3486, + "step": 10260 + }, + { + "epoch": 0.22586396654193264, + "grad_norm": 1.444914799304083, + "learning_rate": 9.027286132477251e-06, + "loss": 0.3215, + "step": 10261 + }, + { + "epoch": 0.22588597842835131, + "grad_norm": 1.5442052442082976, + "learning_rate": 9.027074866976471e-06, + "loss": 0.2795, + "step": 10262 + }, + { + "epoch": 0.22590799031476996, + "grad_norm": 1.8952979716648783, + "learning_rate": 9.026863581008298e-06, + "loss": 0.3768, + "step": 10263 + }, + { + "epoch": 0.22593000220118864, + "grad_norm": 1.4627123823980623, + "learning_rate": 9.026652274573809e-06, + "loss": 0.3991, + "step": 10264 + }, + { + "epoch": 0.22595201408760732, + "grad_norm": 0.6423299819820781, + "learning_rate": 9.026440947674078e-06, + "loss": 0.5272, + "step": 10265 + }, + { + "epoch": 0.22597402597402597, + "grad_norm": 0.6020366918755712, + "learning_rate": 9.026229600310177e-06, + "loss": 0.5315, + "step": 10266 + }, + { + "epoch": 0.22599603786044464, + "grad_norm": 1.633608429168982, + "learning_rate": 9.026018232483182e-06, + "loss": 0.3195, + "step": 10267 + }, + { + "epoch": 0.22601804974686332, + "grad_norm": 1.514144372359958, + "learning_rate": 9.025806844194168e-06, + "loss": 0.4151, + "step": 10268 + }, + { + "epoch": 0.22604006163328197, + "grad_norm": 2.121547847064109, + "learning_rate": 9.025595435444205e-06, + "loss": 0.3903, + "step": 10269 + }, + { + "epoch": 0.22606207351970065, + "grad_norm": 2.0176782626293908, + "learning_rate": 9.025384006234372e-06, + "loss": 0.4084, + "step": 10270 + }, + { + "epoch": 0.2260840854061193, + "grad_norm": 1.623198470828948, + "learning_rate": 9.025172556565742e-06, + "loss": 0.3676, + "step": 10271 + }, + { + "epoch": 0.22610609729253797, + "grad_norm": 1.5437021940143856, + "learning_rate": 9.02496108643939e-06, + "loss": 0.3768, + "step": 10272 + }, + { + "epoch": 0.22612810917895665, + "grad_norm": 1.2598837954082158, + "learning_rate": 9.024749595856392e-06, + "loss": 0.2925, + "step": 10273 + }, + { + "epoch": 0.2261501210653753, + "grad_norm": 0.6469106828503614, + "learning_rate": 9.02453808481782e-06, + "loss": 0.5205, + "step": 10274 + }, + { + "epoch": 0.22617213295179397, + "grad_norm": 1.6490120276353317, + "learning_rate": 9.024326553324751e-06, + "loss": 0.3683, + "step": 10275 + }, + { + "epoch": 0.22619414483821262, + "grad_norm": 1.6964853252618766, + "learning_rate": 9.024115001378258e-06, + "loss": 0.3149, + "step": 10276 + }, + { + "epoch": 0.2262161567246313, + "grad_norm": 1.7963536477353128, + "learning_rate": 9.02390342897942e-06, + "loss": 0.3082, + "step": 10277 + }, + { + "epoch": 0.22623816861104998, + "grad_norm": 1.607740408614371, + "learning_rate": 9.023691836129307e-06, + "loss": 0.3584, + "step": 10278 + }, + { + "epoch": 0.22626018049746863, + "grad_norm": 1.4984341852070875, + "learning_rate": 9.023480222829e-06, + "loss": 0.363, + "step": 10279 + }, + { + "epoch": 0.2262821923838873, + "grad_norm": 1.3340199172411344, + "learning_rate": 9.023268589079573e-06, + "loss": 0.379, + "step": 10280 + }, + { + "epoch": 0.22630420427030595, + "grad_norm": 2.348945138389483, + "learning_rate": 9.023056934882097e-06, + "loss": 0.3308, + "step": 10281 + }, + { + "epoch": 0.22632621615672463, + "grad_norm": 1.3774749072316315, + "learning_rate": 9.022845260237653e-06, + "loss": 0.2668, + "step": 10282 + }, + { + "epoch": 0.2263482280431433, + "grad_norm": 1.6410532706701026, + "learning_rate": 9.022633565147315e-06, + "loss": 0.322, + "step": 10283 + }, + { + "epoch": 0.22637023992956196, + "grad_norm": 0.62754710712029, + "learning_rate": 9.022421849612162e-06, + "loss": 0.5028, + "step": 10284 + }, + { + "epoch": 0.22639225181598063, + "grad_norm": 0.6213887029231623, + "learning_rate": 9.022210113633265e-06, + "loss": 0.5306, + "step": 10285 + }, + { + "epoch": 0.22641426370239928, + "grad_norm": 5.449408746352993, + "learning_rate": 9.0219983572117e-06, + "loss": 0.3203, + "step": 10286 + }, + { + "epoch": 0.22643627558881796, + "grad_norm": 1.469338224582721, + "learning_rate": 9.021786580348548e-06, + "loss": 0.3752, + "step": 10287 + }, + { + "epoch": 0.22645828747523664, + "grad_norm": 2.0056497797030706, + "learning_rate": 9.021574783044882e-06, + "loss": 0.351, + "step": 10288 + }, + { + "epoch": 0.22648029936165529, + "grad_norm": 1.6360500103300863, + "learning_rate": 9.021362965301782e-06, + "loss": 0.2608, + "step": 10289 + }, + { + "epoch": 0.22650231124807396, + "grad_norm": 0.6681151267362512, + "learning_rate": 9.021151127120318e-06, + "loss": 0.5365, + "step": 10290 + }, + { + "epoch": 0.22652432313449264, + "grad_norm": 1.7578485087781917, + "learning_rate": 9.020939268501572e-06, + "loss": 0.4685, + "step": 10291 + }, + { + "epoch": 0.2265463350209113, + "grad_norm": 1.4122839090597523, + "learning_rate": 9.020727389446618e-06, + "loss": 0.3509, + "step": 10292 + }, + { + "epoch": 0.22656834690732996, + "grad_norm": 1.7495917949420245, + "learning_rate": 9.020515489956536e-06, + "loss": 0.4126, + "step": 10293 + }, + { + "epoch": 0.22659035879374861, + "grad_norm": 2.4202126056543873, + "learning_rate": 9.0203035700324e-06, + "loss": 0.3565, + "step": 10294 + }, + { + "epoch": 0.2266123706801673, + "grad_norm": 2.2345153261487347, + "learning_rate": 9.02009162967529e-06, + "loss": 0.3651, + "step": 10295 + }, + { + "epoch": 0.22663438256658597, + "grad_norm": 1.9008619455980906, + "learning_rate": 9.019879668886278e-06, + "loss": 0.3215, + "step": 10296 + }, + { + "epoch": 0.22665639445300462, + "grad_norm": 1.3432398651202557, + "learning_rate": 9.019667687666447e-06, + "loss": 0.2916, + "step": 10297 + }, + { + "epoch": 0.2266784063394233, + "grad_norm": 2.930815751225026, + "learning_rate": 9.01945568601687e-06, + "loss": 0.3023, + "step": 10298 + }, + { + "epoch": 0.22670041822584194, + "grad_norm": 1.9002561027662335, + "learning_rate": 9.019243663938628e-06, + "loss": 0.2915, + "step": 10299 + }, + { + "epoch": 0.22672243011226062, + "grad_norm": 1.4462663379880774, + "learning_rate": 9.019031621432797e-06, + "loss": 0.2996, + "step": 10300 + }, + { + "epoch": 0.2267444419986793, + "grad_norm": 1.4506172292015072, + "learning_rate": 9.018819558500455e-06, + "loss": 0.3269, + "step": 10301 + }, + { + "epoch": 0.22676645388509795, + "grad_norm": 0.6174894908841208, + "learning_rate": 9.01860747514268e-06, + "loss": 0.5187, + "step": 10302 + }, + { + "epoch": 0.22678846577151662, + "grad_norm": 1.3413874240507286, + "learning_rate": 9.018395371360548e-06, + "loss": 0.3776, + "step": 10303 + }, + { + "epoch": 0.22681047765793527, + "grad_norm": 2.198571472506527, + "learning_rate": 9.018183247155139e-06, + "loss": 0.3624, + "step": 10304 + }, + { + "epoch": 0.22683248954435395, + "grad_norm": 1.6530552368588465, + "learning_rate": 9.017971102527532e-06, + "loss": 0.4064, + "step": 10305 + }, + { + "epoch": 0.22685450143077263, + "grad_norm": 1.4724926944266692, + "learning_rate": 9.017758937478802e-06, + "loss": 0.2915, + "step": 10306 + }, + { + "epoch": 0.22687651331719128, + "grad_norm": 2.134544703404669, + "learning_rate": 9.017546752010029e-06, + "loss": 0.362, + "step": 10307 + }, + { + "epoch": 0.22689852520360995, + "grad_norm": 1.3056193589050364, + "learning_rate": 9.017334546122293e-06, + "loss": 0.4188, + "step": 10308 + }, + { + "epoch": 0.22692053709002863, + "grad_norm": 1.2180923440691056, + "learning_rate": 9.01712231981667e-06, + "loss": 0.2303, + "step": 10309 + }, + { + "epoch": 0.22694254897644728, + "grad_norm": 1.1350626704512408, + "learning_rate": 9.016910073094241e-06, + "loss": 0.2884, + "step": 10310 + }, + { + "epoch": 0.22696456086286595, + "grad_norm": 1.4041660571392949, + "learning_rate": 9.01669780595608e-06, + "loss": 0.3016, + "step": 10311 + }, + { + "epoch": 0.2269865727492846, + "grad_norm": 1.351444998547086, + "learning_rate": 9.016485518403273e-06, + "loss": 0.3724, + "step": 10312 + }, + { + "epoch": 0.22700858463570328, + "grad_norm": 3.177753150756828, + "learning_rate": 9.016273210436894e-06, + "loss": 0.3071, + "step": 10313 + }, + { + "epoch": 0.22703059652212196, + "grad_norm": 2.8817756440872917, + "learning_rate": 9.016060882058023e-06, + "loss": 0.3105, + "step": 10314 + }, + { + "epoch": 0.2270526084085406, + "grad_norm": 0.7798425878136416, + "learning_rate": 9.01584853326774e-06, + "loss": 0.5413, + "step": 10315 + }, + { + "epoch": 0.22707462029495928, + "grad_norm": 1.586694185946974, + "learning_rate": 9.015636164067121e-06, + "loss": 0.2941, + "step": 10316 + }, + { + "epoch": 0.22709663218137793, + "grad_norm": 2.8677282788708163, + "learning_rate": 9.015423774457251e-06, + "loss": 0.3682, + "step": 10317 + }, + { + "epoch": 0.2271186440677966, + "grad_norm": 1.7022061242978452, + "learning_rate": 9.015211364439205e-06, + "loss": 0.4581, + "step": 10318 + }, + { + "epoch": 0.2271406559542153, + "grad_norm": 2.9268982138186193, + "learning_rate": 9.014998934014065e-06, + "loss": 0.2926, + "step": 10319 + }, + { + "epoch": 0.22716266784063394, + "grad_norm": 2.070674651591062, + "learning_rate": 9.01478648318291e-06, + "loss": 0.3371, + "step": 10320 + }, + { + "epoch": 0.2271846797270526, + "grad_norm": 1.4961760853447543, + "learning_rate": 9.014574011946818e-06, + "loss": 0.3124, + "step": 10321 + }, + { + "epoch": 0.22720669161347126, + "grad_norm": 1.7877843112521459, + "learning_rate": 9.01436152030687e-06, + "loss": 0.2808, + "step": 10322 + }, + { + "epoch": 0.22722870349988994, + "grad_norm": 0.6259040471324914, + "learning_rate": 9.014149008264148e-06, + "loss": 0.5657, + "step": 10323 + }, + { + "epoch": 0.22725071538630862, + "grad_norm": 0.6317427163095256, + "learning_rate": 9.013936475819728e-06, + "loss": 0.5257, + "step": 10324 + }, + { + "epoch": 0.22727272727272727, + "grad_norm": 1.8082708134682437, + "learning_rate": 9.013723922974695e-06, + "loss": 0.3451, + "step": 10325 + }, + { + "epoch": 0.22729473915914594, + "grad_norm": 1.7227729009974435, + "learning_rate": 9.013511349730127e-06, + "loss": 0.3463, + "step": 10326 + }, + { + "epoch": 0.2273167510455646, + "grad_norm": 2.229781840707171, + "learning_rate": 9.013298756087103e-06, + "loss": 0.2976, + "step": 10327 + }, + { + "epoch": 0.22733876293198327, + "grad_norm": 1.4230544952956552, + "learning_rate": 9.013086142046705e-06, + "loss": 0.3363, + "step": 10328 + }, + { + "epoch": 0.22736077481840195, + "grad_norm": 7.447854234439473, + "learning_rate": 9.012873507610014e-06, + "loss": 0.3714, + "step": 10329 + }, + { + "epoch": 0.2273827867048206, + "grad_norm": 1.7485202384664726, + "learning_rate": 9.012660852778109e-06, + "loss": 0.3151, + "step": 10330 + }, + { + "epoch": 0.22740479859123927, + "grad_norm": 1.3486303107013233, + "learning_rate": 9.012448177552072e-06, + "loss": 0.3975, + "step": 10331 + }, + { + "epoch": 0.22742681047765795, + "grad_norm": 3.9726107609350634, + "learning_rate": 9.012235481932984e-06, + "loss": 0.3413, + "step": 10332 + }, + { + "epoch": 0.2274488223640766, + "grad_norm": 1.3338541242124764, + "learning_rate": 9.012022765921926e-06, + "loss": 0.3634, + "step": 10333 + }, + { + "epoch": 0.22747083425049527, + "grad_norm": 1.7792322847757773, + "learning_rate": 9.011810029519979e-06, + "loss": 0.3497, + "step": 10334 + }, + { + "epoch": 0.22749284613691392, + "grad_norm": 2.0826748470584286, + "learning_rate": 9.011597272728223e-06, + "loss": 0.3136, + "step": 10335 + }, + { + "epoch": 0.2275148580233326, + "grad_norm": 0.7041767438692452, + "learning_rate": 9.011384495547741e-06, + "loss": 0.557, + "step": 10336 + }, + { + "epoch": 0.22753686990975128, + "grad_norm": 1.4336346469176988, + "learning_rate": 9.011171697979614e-06, + "loss": 0.4636, + "step": 10337 + }, + { + "epoch": 0.22755888179616993, + "grad_norm": 0.6552410783014438, + "learning_rate": 9.010958880024925e-06, + "loss": 0.5404, + "step": 10338 + }, + { + "epoch": 0.2275808936825886, + "grad_norm": 3.1407586190160517, + "learning_rate": 9.010746041684752e-06, + "loss": 0.4229, + "step": 10339 + }, + { + "epoch": 0.22760290556900725, + "grad_norm": 2.309112164482387, + "learning_rate": 9.010533182960179e-06, + "loss": 0.3218, + "step": 10340 + }, + { + "epoch": 0.22762491745542593, + "grad_norm": 0.644972989594588, + "learning_rate": 9.010320303852288e-06, + "loss": 0.5302, + "step": 10341 + }, + { + "epoch": 0.2276469293418446, + "grad_norm": 1.6309567728845473, + "learning_rate": 9.010107404362161e-06, + "loss": 0.3382, + "step": 10342 + }, + { + "epoch": 0.22766894122826326, + "grad_norm": 1.1214202931073824, + "learning_rate": 9.009894484490879e-06, + "loss": 0.3259, + "step": 10343 + }, + { + "epoch": 0.22769095311468193, + "grad_norm": 2.5233474445825554, + "learning_rate": 9.009681544239525e-06, + "loss": 0.3434, + "step": 10344 + }, + { + "epoch": 0.22771296500110058, + "grad_norm": 0.6462696398413199, + "learning_rate": 9.00946858360918e-06, + "loss": 0.5383, + "step": 10345 + }, + { + "epoch": 0.22773497688751926, + "grad_norm": 0.6259006943851462, + "learning_rate": 9.009255602600928e-06, + "loss": 0.517, + "step": 10346 + }, + { + "epoch": 0.22775698877393794, + "grad_norm": 1.3070092497191492, + "learning_rate": 9.009042601215852e-06, + "loss": 0.3737, + "step": 10347 + }, + { + "epoch": 0.22777900066035658, + "grad_norm": 1.497157100705043, + "learning_rate": 9.008829579455031e-06, + "loss": 0.4002, + "step": 10348 + }, + { + "epoch": 0.22780101254677526, + "grad_norm": 1.4417480877334172, + "learning_rate": 9.008616537319552e-06, + "loss": 0.33, + "step": 10349 + }, + { + "epoch": 0.22782302443319394, + "grad_norm": 1.2834082578221933, + "learning_rate": 9.008403474810497e-06, + "loss": 0.3498, + "step": 10350 + }, + { + "epoch": 0.2278450363196126, + "grad_norm": 0.6621110763738638, + "learning_rate": 9.008190391928945e-06, + "loss": 0.5299, + "step": 10351 + }, + { + "epoch": 0.22786704820603126, + "grad_norm": 1.6718653373740273, + "learning_rate": 9.007977288675984e-06, + "loss": 0.4285, + "step": 10352 + }, + { + "epoch": 0.2278890600924499, + "grad_norm": 0.6612839432444624, + "learning_rate": 9.007764165052693e-06, + "loss": 0.5304, + "step": 10353 + }, + { + "epoch": 0.2279110719788686, + "grad_norm": 0.6375553809152169, + "learning_rate": 9.007551021060157e-06, + "loss": 0.5363, + "step": 10354 + }, + { + "epoch": 0.22793308386528727, + "grad_norm": 2.0877885528933238, + "learning_rate": 9.00733785669946e-06, + "loss": 0.4198, + "step": 10355 + }, + { + "epoch": 0.22795509575170592, + "grad_norm": 1.8065057637398607, + "learning_rate": 9.007124671971686e-06, + "loss": 0.3008, + "step": 10356 + }, + { + "epoch": 0.2279771076381246, + "grad_norm": 2.2695748026093767, + "learning_rate": 9.006911466877915e-06, + "loss": 0.3084, + "step": 10357 + }, + { + "epoch": 0.22799911952454324, + "grad_norm": 1.423646784605304, + "learning_rate": 9.006698241419234e-06, + "loss": 0.3227, + "step": 10358 + }, + { + "epoch": 0.22802113141096192, + "grad_norm": 0.6808247579696691, + "learning_rate": 9.006484995596725e-06, + "loss": 0.5363, + "step": 10359 + }, + { + "epoch": 0.2280431432973806, + "grad_norm": 1.311841819447382, + "learning_rate": 9.006271729411473e-06, + "loss": 0.3688, + "step": 10360 + }, + { + "epoch": 0.22806515518379925, + "grad_norm": 0.6733050360619773, + "learning_rate": 9.00605844286456e-06, + "loss": 0.5278, + "step": 10361 + }, + { + "epoch": 0.22808716707021792, + "grad_norm": 1.2886951420860802, + "learning_rate": 9.005845135957073e-06, + "loss": 0.2742, + "step": 10362 + }, + { + "epoch": 0.22810917895663657, + "grad_norm": 2.631628755092202, + "learning_rate": 9.005631808690092e-06, + "loss": 0.3182, + "step": 10363 + }, + { + "epoch": 0.22813119084305525, + "grad_norm": 1.6482649869420487, + "learning_rate": 9.005418461064705e-06, + "loss": 0.3209, + "step": 10364 + }, + { + "epoch": 0.22815320272947393, + "grad_norm": 1.7513307631449901, + "learning_rate": 9.005205093081997e-06, + "loss": 0.3782, + "step": 10365 + }, + { + "epoch": 0.22817521461589257, + "grad_norm": 2.0629409177506273, + "learning_rate": 9.004991704743047e-06, + "loss": 0.3813, + "step": 10366 + }, + { + "epoch": 0.22819722650231125, + "grad_norm": 0.6283905727466651, + "learning_rate": 9.004778296048944e-06, + "loss": 0.5237, + "step": 10367 + }, + { + "epoch": 0.2282192383887299, + "grad_norm": 1.7808640226444339, + "learning_rate": 9.004564867000772e-06, + "loss": 0.3239, + "step": 10368 + }, + { + "epoch": 0.22824125027514858, + "grad_norm": 1.2743843696354338, + "learning_rate": 9.004351417599615e-06, + "loss": 0.3586, + "step": 10369 + }, + { + "epoch": 0.22826326216156725, + "grad_norm": 1.3449753588791786, + "learning_rate": 9.004137947846558e-06, + "loss": 0.4122, + "step": 10370 + }, + { + "epoch": 0.2282852740479859, + "grad_norm": 1.4970120714628645, + "learning_rate": 9.003924457742686e-06, + "loss": 0.4881, + "step": 10371 + }, + { + "epoch": 0.22830728593440458, + "grad_norm": 1.5394194075923229, + "learning_rate": 9.003710947289084e-06, + "loss": 0.3256, + "step": 10372 + }, + { + "epoch": 0.22832929782082326, + "grad_norm": 1.140092640935586, + "learning_rate": 9.003497416486837e-06, + "loss": 0.3382, + "step": 10373 + }, + { + "epoch": 0.2283513097072419, + "grad_norm": 0.6445946133502883, + "learning_rate": 9.003283865337032e-06, + "loss": 0.5467, + "step": 10374 + }, + { + "epoch": 0.22837332159366058, + "grad_norm": 1.7010117596773144, + "learning_rate": 9.00307029384075e-06, + "loss": 0.3545, + "step": 10375 + }, + { + "epoch": 0.22839533348007923, + "grad_norm": 2.0249008667819512, + "learning_rate": 9.002856701999081e-06, + "loss": 0.3075, + "step": 10376 + }, + { + "epoch": 0.2284173453664979, + "grad_norm": 2.06615601393581, + "learning_rate": 9.002643089813111e-06, + "loss": 0.426, + "step": 10377 + }, + { + "epoch": 0.22843935725291659, + "grad_norm": 0.6173693665647307, + "learning_rate": 9.002429457283921e-06, + "loss": 0.5503, + "step": 10378 + }, + { + "epoch": 0.22846136913933524, + "grad_norm": 2.7190886734159587, + "learning_rate": 9.0022158044126e-06, + "loss": 0.3852, + "step": 10379 + }, + { + "epoch": 0.2284833810257539, + "grad_norm": 0.628922759333093, + "learning_rate": 9.002002131200233e-06, + "loss": 0.5252, + "step": 10380 + }, + { + "epoch": 0.22850539291217256, + "grad_norm": 1.3244615365396297, + "learning_rate": 9.001788437647907e-06, + "loss": 0.2906, + "step": 10381 + }, + { + "epoch": 0.22852740479859124, + "grad_norm": 1.468133041353, + "learning_rate": 9.001574723756706e-06, + "loss": 0.3867, + "step": 10382 + }, + { + "epoch": 0.22854941668500992, + "grad_norm": 1.3532022252847786, + "learning_rate": 9.001360989527718e-06, + "loss": 0.3915, + "step": 10383 + }, + { + "epoch": 0.22857142857142856, + "grad_norm": 1.3550191213246414, + "learning_rate": 9.00114723496203e-06, + "loss": 0.3395, + "step": 10384 + }, + { + "epoch": 0.22859344045784724, + "grad_norm": 0.6088193687821398, + "learning_rate": 9.000933460060725e-06, + "loss": 0.4844, + "step": 10385 + }, + { + "epoch": 0.2286154523442659, + "grad_norm": 1.5666945715876557, + "learning_rate": 9.000719664824893e-06, + "loss": 0.3178, + "step": 10386 + }, + { + "epoch": 0.22863746423068457, + "grad_norm": 1.3413153811613128, + "learning_rate": 9.000505849255621e-06, + "loss": 0.4143, + "step": 10387 + }, + { + "epoch": 0.22865947611710324, + "grad_norm": 1.6545730892605448, + "learning_rate": 9.000292013353992e-06, + "loss": 0.3449, + "step": 10388 + }, + { + "epoch": 0.2286814880035219, + "grad_norm": 1.2388122701562656, + "learning_rate": 9.000078157121095e-06, + "loss": 0.3299, + "step": 10389 + }, + { + "epoch": 0.22870349988994057, + "grad_norm": 1.448317323439713, + "learning_rate": 8.999864280558017e-06, + "loss": 0.3712, + "step": 10390 + }, + { + "epoch": 0.22872551177635925, + "grad_norm": 0.6479047289261514, + "learning_rate": 8.999650383665843e-06, + "loss": 0.5264, + "step": 10391 + }, + { + "epoch": 0.2287475236627779, + "grad_norm": 1.841122655302073, + "learning_rate": 8.999436466445664e-06, + "loss": 0.4293, + "step": 10392 + }, + { + "epoch": 0.22876953554919657, + "grad_norm": 1.5800168970016717, + "learning_rate": 8.999222528898565e-06, + "loss": 0.3343, + "step": 10393 + }, + { + "epoch": 0.22879154743561522, + "grad_norm": 1.6380104360138084, + "learning_rate": 8.999008571025633e-06, + "loss": 0.3623, + "step": 10394 + }, + { + "epoch": 0.2288135593220339, + "grad_norm": 3.6372763560631474, + "learning_rate": 8.998794592827954e-06, + "loss": 0.378, + "step": 10395 + }, + { + "epoch": 0.22883557120845258, + "grad_norm": 1.4297425194195488, + "learning_rate": 8.998580594306621e-06, + "loss": 0.2504, + "step": 10396 + }, + { + "epoch": 0.22885758309487123, + "grad_norm": 1.2863249683368025, + "learning_rate": 8.998366575462716e-06, + "loss": 0.3665, + "step": 10397 + }, + { + "epoch": 0.2288795949812899, + "grad_norm": 1.5907920935374769, + "learning_rate": 8.998152536297328e-06, + "loss": 0.3674, + "step": 10398 + }, + { + "epoch": 0.22890160686770855, + "grad_norm": 1.4072951929864759, + "learning_rate": 8.997938476811547e-06, + "loss": 0.3578, + "step": 10399 + }, + { + "epoch": 0.22892361875412723, + "grad_norm": 0.6262351971772847, + "learning_rate": 8.99772439700646e-06, + "loss": 0.5118, + "step": 10400 + }, + { + "epoch": 0.2289456306405459, + "grad_norm": 1.6635978402280218, + "learning_rate": 8.997510296883151e-06, + "loss": 0.3884, + "step": 10401 + }, + { + "epoch": 0.22896764252696455, + "grad_norm": 1.5254496184400061, + "learning_rate": 8.997296176442716e-06, + "loss": 0.3038, + "step": 10402 + }, + { + "epoch": 0.22898965441338323, + "grad_norm": 1.319599532755729, + "learning_rate": 8.997082035686238e-06, + "loss": 0.3326, + "step": 10403 + }, + { + "epoch": 0.22901166629980188, + "grad_norm": 1.402076873170403, + "learning_rate": 8.996867874614805e-06, + "loss": 0.4083, + "step": 10404 + }, + { + "epoch": 0.22903367818622056, + "grad_norm": 1.3095990468782215, + "learning_rate": 8.996653693229507e-06, + "loss": 0.2945, + "step": 10405 + }, + { + "epoch": 0.22905569007263923, + "grad_norm": 2.063857391065026, + "learning_rate": 8.996439491531433e-06, + "loss": 0.3492, + "step": 10406 + }, + { + "epoch": 0.22907770195905788, + "grad_norm": 1.445912564052324, + "learning_rate": 8.99622526952167e-06, + "loss": 0.273, + "step": 10407 + }, + { + "epoch": 0.22909971384547656, + "grad_norm": 1.7308637230075392, + "learning_rate": 8.99601102720131e-06, + "loss": 0.4987, + "step": 10408 + }, + { + "epoch": 0.22912172573189524, + "grad_norm": 1.930208674022149, + "learning_rate": 8.995796764571437e-06, + "loss": 0.4281, + "step": 10409 + }, + { + "epoch": 0.2291437376183139, + "grad_norm": 1.4835194665350055, + "learning_rate": 8.995582481633145e-06, + "loss": 0.4245, + "step": 10410 + }, + { + "epoch": 0.22916574950473256, + "grad_norm": 1.4500099560805286, + "learning_rate": 8.995368178387518e-06, + "loss": 0.2912, + "step": 10411 + }, + { + "epoch": 0.2291877613911512, + "grad_norm": 1.482098455519464, + "learning_rate": 8.995153854835652e-06, + "loss": 0.325, + "step": 10412 + }, + { + "epoch": 0.2292097732775699, + "grad_norm": 2.17506462083438, + "learning_rate": 8.994939510978628e-06, + "loss": 0.3824, + "step": 10413 + }, + { + "epoch": 0.22923178516398857, + "grad_norm": 1.6212191680974934, + "learning_rate": 8.994725146817543e-06, + "loss": 0.3326, + "step": 10414 + }, + { + "epoch": 0.22925379705040722, + "grad_norm": 9.058839208109475, + "learning_rate": 8.99451076235348e-06, + "loss": 0.3287, + "step": 10415 + }, + { + "epoch": 0.2292758089368259, + "grad_norm": 1.5223244863537158, + "learning_rate": 8.994296357587532e-06, + "loss": 0.279, + "step": 10416 + }, + { + "epoch": 0.22929782082324454, + "grad_norm": 1.506192015624056, + "learning_rate": 8.994081932520792e-06, + "loss": 0.3865, + "step": 10417 + }, + { + "epoch": 0.22931983270966322, + "grad_norm": 1.1426206270916814, + "learning_rate": 8.993867487154344e-06, + "loss": 0.2254, + "step": 10418 + }, + { + "epoch": 0.2293418445960819, + "grad_norm": 2.044078856520081, + "learning_rate": 8.99365302148928e-06, + "loss": 0.3397, + "step": 10419 + }, + { + "epoch": 0.22936385648250054, + "grad_norm": 1.2956952131515966, + "learning_rate": 8.99343853552669e-06, + "loss": 0.4022, + "step": 10420 + }, + { + "epoch": 0.22938586836891922, + "grad_norm": 0.6699009870009999, + "learning_rate": 8.993224029267664e-06, + "loss": 0.5339, + "step": 10421 + }, + { + "epoch": 0.22940788025533787, + "grad_norm": 2.1039547210375944, + "learning_rate": 8.993009502713292e-06, + "loss": 0.4505, + "step": 10422 + }, + { + "epoch": 0.22942989214175655, + "grad_norm": 1.1755014952715044, + "learning_rate": 8.992794955864668e-06, + "loss": 0.2592, + "step": 10423 + }, + { + "epoch": 0.22945190402817522, + "grad_norm": 1.4668002069886035, + "learning_rate": 8.992580388722875e-06, + "loss": 0.4081, + "step": 10424 + }, + { + "epoch": 0.22947391591459387, + "grad_norm": 1.3951465464439101, + "learning_rate": 8.992365801289009e-06, + "loss": 0.4957, + "step": 10425 + }, + { + "epoch": 0.22949592780101255, + "grad_norm": 2.559352041050224, + "learning_rate": 8.99215119356416e-06, + "loss": 0.3613, + "step": 10426 + }, + { + "epoch": 0.2295179396874312, + "grad_norm": 1.2115797940636162, + "learning_rate": 8.99193656554942e-06, + "loss": 0.3793, + "step": 10427 + }, + { + "epoch": 0.22953995157384988, + "grad_norm": 1.9920410278984153, + "learning_rate": 8.991721917245875e-06, + "loss": 0.3243, + "step": 10428 + }, + { + "epoch": 0.22956196346026855, + "grad_norm": 1.2815443286808204, + "learning_rate": 8.99150724865462e-06, + "loss": 0.3156, + "step": 10429 + }, + { + "epoch": 0.2295839753466872, + "grad_norm": 2.4476144462558107, + "learning_rate": 8.991292559776744e-06, + "loss": 0.3321, + "step": 10430 + }, + { + "epoch": 0.22960598723310588, + "grad_norm": 1.8238181803051936, + "learning_rate": 8.991077850613338e-06, + "loss": 0.4097, + "step": 10431 + }, + { + "epoch": 0.22962799911952456, + "grad_norm": 1.6989768581155202, + "learning_rate": 8.990863121165498e-06, + "loss": 0.2872, + "step": 10432 + }, + { + "epoch": 0.2296500110059432, + "grad_norm": 2.5972568469396378, + "learning_rate": 8.99064837143431e-06, + "loss": 0.3922, + "step": 10433 + }, + { + "epoch": 0.22967202289236188, + "grad_norm": 0.6360004664144837, + "learning_rate": 8.990433601420867e-06, + "loss": 0.5021, + "step": 10434 + }, + { + "epoch": 0.22969403477878053, + "grad_norm": 1.371460315460061, + "learning_rate": 8.990218811126258e-06, + "loss": 0.3254, + "step": 10435 + }, + { + "epoch": 0.2297160466651992, + "grad_norm": 2.841574050647046, + "learning_rate": 8.990004000551582e-06, + "loss": 0.3818, + "step": 10436 + }, + { + "epoch": 0.22973805855161789, + "grad_norm": 0.6542277601506421, + "learning_rate": 8.989789169697923e-06, + "loss": 0.5279, + "step": 10437 + }, + { + "epoch": 0.22976007043803653, + "grad_norm": 0.634190362607248, + "learning_rate": 8.989574318566376e-06, + "loss": 0.5463, + "step": 10438 + }, + { + "epoch": 0.2297820823244552, + "grad_norm": 0.6176936522029663, + "learning_rate": 8.989359447158034e-06, + "loss": 0.5062, + "step": 10439 + }, + { + "epoch": 0.22980409421087386, + "grad_norm": 1.4358204012782092, + "learning_rate": 8.989144555473987e-06, + "loss": 0.3951, + "step": 10440 + }, + { + "epoch": 0.22982610609729254, + "grad_norm": 1.8268560849037916, + "learning_rate": 8.98892964351533e-06, + "loss": 0.4011, + "step": 10441 + }, + { + "epoch": 0.22984811798371121, + "grad_norm": 1.4311109532389588, + "learning_rate": 8.988714711283153e-06, + "loss": 0.3638, + "step": 10442 + }, + { + "epoch": 0.22987012987012986, + "grad_norm": 1.9302237923898473, + "learning_rate": 8.988499758778548e-06, + "loss": 0.4156, + "step": 10443 + }, + { + "epoch": 0.22989214175654854, + "grad_norm": 1.652228653618515, + "learning_rate": 8.988284786002609e-06, + "loss": 0.2649, + "step": 10444 + }, + { + "epoch": 0.2299141536429672, + "grad_norm": 0.6495372820845475, + "learning_rate": 8.988069792956426e-06, + "loss": 0.5101, + "step": 10445 + }, + { + "epoch": 0.22993616552938587, + "grad_norm": 1.7740596070604626, + "learning_rate": 8.987854779641095e-06, + "loss": 0.3152, + "step": 10446 + }, + { + "epoch": 0.22995817741580454, + "grad_norm": 1.4862019797294201, + "learning_rate": 8.987639746057708e-06, + "loss": 0.3972, + "step": 10447 + }, + { + "epoch": 0.2299801893022232, + "grad_norm": 1.3930524430287765, + "learning_rate": 8.987424692207356e-06, + "loss": 0.3618, + "step": 10448 + }, + { + "epoch": 0.23000220118864187, + "grad_norm": 2.0808785998274253, + "learning_rate": 8.987209618091133e-06, + "loss": 0.3711, + "step": 10449 + }, + { + "epoch": 0.23002421307506055, + "grad_norm": 1.6712685952000625, + "learning_rate": 8.986994523710133e-06, + "loss": 0.2912, + "step": 10450 + }, + { + "epoch": 0.2300462249614792, + "grad_norm": 1.6634248736809425, + "learning_rate": 8.986779409065449e-06, + "loss": 0.3279, + "step": 10451 + }, + { + "epoch": 0.23006823684789787, + "grad_norm": 1.3023310372871961, + "learning_rate": 8.986564274158174e-06, + "loss": 0.2425, + "step": 10452 + }, + { + "epoch": 0.23009024873431652, + "grad_norm": 0.6384282066083928, + "learning_rate": 8.986349118989401e-06, + "loss": 0.5408, + "step": 10453 + }, + { + "epoch": 0.2301122606207352, + "grad_norm": 0.6552096557462026, + "learning_rate": 8.986133943560223e-06, + "loss": 0.5452, + "step": 10454 + }, + { + "epoch": 0.23013427250715388, + "grad_norm": 0.6365888051007599, + "learning_rate": 8.985918747871736e-06, + "loss": 0.5531, + "step": 10455 + }, + { + "epoch": 0.23015628439357252, + "grad_norm": 1.3663675317999318, + "learning_rate": 8.985703531925032e-06, + "loss": 0.3537, + "step": 10456 + }, + { + "epoch": 0.2301782962799912, + "grad_norm": 1.7095160320686134, + "learning_rate": 8.985488295721204e-06, + "loss": 0.3822, + "step": 10457 + }, + { + "epoch": 0.23020030816640985, + "grad_norm": 2.000200774669888, + "learning_rate": 8.985273039261346e-06, + "loss": 0.4448, + "step": 10458 + }, + { + "epoch": 0.23022232005282853, + "grad_norm": 0.6437592001070604, + "learning_rate": 8.985057762546554e-06, + "loss": 0.5177, + "step": 10459 + }, + { + "epoch": 0.2302443319392472, + "grad_norm": 1.222995302149792, + "learning_rate": 8.984842465577921e-06, + "loss": 0.3466, + "step": 10460 + }, + { + "epoch": 0.23026634382566585, + "grad_norm": 0.6404291330040747, + "learning_rate": 8.984627148356541e-06, + "loss": 0.5034, + "step": 10461 + }, + { + "epoch": 0.23028835571208453, + "grad_norm": 1.7373239066887047, + "learning_rate": 8.98441181088351e-06, + "loss": 0.3816, + "step": 10462 + }, + { + "epoch": 0.23031036759850318, + "grad_norm": 1.7431544783174728, + "learning_rate": 8.98419645315992e-06, + "loss": 0.3529, + "step": 10463 + }, + { + "epoch": 0.23033237948492186, + "grad_norm": 1.3516544915026663, + "learning_rate": 8.983981075186866e-06, + "loss": 0.29, + "step": 10464 + }, + { + "epoch": 0.23035439137134053, + "grad_norm": 1.7890457352256661, + "learning_rate": 8.983765676965446e-06, + "loss": 0.3026, + "step": 10465 + }, + { + "epoch": 0.23037640325775918, + "grad_norm": 6.564146754751451, + "learning_rate": 8.983550258496749e-06, + "loss": 0.4058, + "step": 10466 + }, + { + "epoch": 0.23039841514417786, + "grad_norm": 1.4926923523119953, + "learning_rate": 8.983334819781875e-06, + "loss": 0.3254, + "step": 10467 + }, + { + "epoch": 0.2304204270305965, + "grad_norm": 1.961968734789958, + "learning_rate": 8.983119360821915e-06, + "loss": 0.3136, + "step": 10468 + }, + { + "epoch": 0.23044243891701519, + "grad_norm": 1.561486513502824, + "learning_rate": 8.982903881617966e-06, + "loss": 0.3497, + "step": 10469 + }, + { + "epoch": 0.23046445080343386, + "grad_norm": 1.4471283158558008, + "learning_rate": 8.982688382171125e-06, + "loss": 0.4, + "step": 10470 + }, + { + "epoch": 0.2304864626898525, + "grad_norm": 1.7767006183218654, + "learning_rate": 8.982472862482484e-06, + "loss": 0.3444, + "step": 10471 + }, + { + "epoch": 0.2305084745762712, + "grad_norm": 1.297462523105736, + "learning_rate": 8.982257322553139e-06, + "loss": 0.2366, + "step": 10472 + }, + { + "epoch": 0.23053048646268987, + "grad_norm": 1.7805505916261624, + "learning_rate": 8.982041762384188e-06, + "loss": 0.322, + "step": 10473 + }, + { + "epoch": 0.23055249834910851, + "grad_norm": 2.2229424332123275, + "learning_rate": 8.981826181976723e-06, + "loss": 0.3403, + "step": 10474 + }, + { + "epoch": 0.2305745102355272, + "grad_norm": 1.5902907302211864, + "learning_rate": 8.98161058133184e-06, + "loss": 0.3773, + "step": 10475 + }, + { + "epoch": 0.23059652212194584, + "grad_norm": 1.9038660318621405, + "learning_rate": 8.98139496045064e-06, + "loss": 0.4269, + "step": 10476 + }, + { + "epoch": 0.23061853400836452, + "grad_norm": 1.4775239169177778, + "learning_rate": 8.981179319334214e-06, + "loss": 0.3796, + "step": 10477 + }, + { + "epoch": 0.2306405458947832, + "grad_norm": 1.2952415731515436, + "learning_rate": 8.980963657983658e-06, + "loss": 0.3279, + "step": 10478 + }, + { + "epoch": 0.23066255778120184, + "grad_norm": 2.0209666157519313, + "learning_rate": 8.98074797640007e-06, + "loss": 0.3793, + "step": 10479 + }, + { + "epoch": 0.23068456966762052, + "grad_norm": 1.3269526351834344, + "learning_rate": 8.980532274584543e-06, + "loss": 0.3777, + "step": 10480 + }, + { + "epoch": 0.23070658155403917, + "grad_norm": 1.4701926997393944, + "learning_rate": 8.980316552538179e-06, + "loss": 0.3783, + "step": 10481 + }, + { + "epoch": 0.23072859344045785, + "grad_norm": 1.4675112101989385, + "learning_rate": 8.980100810262069e-06, + "loss": 0.3665, + "step": 10482 + }, + { + "epoch": 0.23075060532687652, + "grad_norm": 1.9095327750890747, + "learning_rate": 8.979885047757313e-06, + "loss": 0.3425, + "step": 10483 + }, + { + "epoch": 0.23077261721329517, + "grad_norm": 1.5242488178468478, + "learning_rate": 8.979669265025003e-06, + "loss": 0.3215, + "step": 10484 + }, + { + "epoch": 0.23079462909971385, + "grad_norm": 1.2921122622976149, + "learning_rate": 8.979453462066241e-06, + "loss": 0.2445, + "step": 10485 + }, + { + "epoch": 0.2308166409861325, + "grad_norm": 0.731525423726377, + "learning_rate": 8.979237638882122e-06, + "loss": 0.5172, + "step": 10486 + }, + { + "epoch": 0.23083865287255118, + "grad_norm": 2.1938994418826665, + "learning_rate": 8.97902179547374e-06, + "loss": 0.4201, + "step": 10487 + }, + { + "epoch": 0.23086066475896985, + "grad_norm": 1.4774299723826267, + "learning_rate": 8.978805931842199e-06, + "loss": 0.3837, + "step": 10488 + }, + { + "epoch": 0.2308826766453885, + "grad_norm": 1.5042908003981337, + "learning_rate": 8.978590047988589e-06, + "loss": 0.4049, + "step": 10489 + }, + { + "epoch": 0.23090468853180718, + "grad_norm": 2.156972010223784, + "learning_rate": 8.97837414391401e-06, + "loss": 0.3949, + "step": 10490 + }, + { + "epoch": 0.23092670041822586, + "grad_norm": 0.6530685601430606, + "learning_rate": 8.97815821961956e-06, + "loss": 0.5381, + "step": 10491 + }, + { + "epoch": 0.2309487123046445, + "grad_norm": 29.69382425430776, + "learning_rate": 8.977942275106334e-06, + "loss": 0.4294, + "step": 10492 + }, + { + "epoch": 0.23097072419106318, + "grad_norm": 1.4945707695228783, + "learning_rate": 8.977726310375432e-06, + "loss": 0.3196, + "step": 10493 + }, + { + "epoch": 0.23099273607748183, + "grad_norm": 0.6638453045620101, + "learning_rate": 8.97751032542795e-06, + "loss": 0.536, + "step": 10494 + }, + { + "epoch": 0.2310147479639005, + "grad_norm": 1.5618493227906207, + "learning_rate": 8.977294320264987e-06, + "loss": 0.3404, + "step": 10495 + }, + { + "epoch": 0.23103675985031918, + "grad_norm": 0.6218304848145031, + "learning_rate": 8.977078294887642e-06, + "loss": 0.4975, + "step": 10496 + }, + { + "epoch": 0.23105877173673783, + "grad_norm": 0.6266848179033511, + "learning_rate": 8.976862249297009e-06, + "loss": 0.5228, + "step": 10497 + }, + { + "epoch": 0.2310807836231565, + "grad_norm": 1.859522276641514, + "learning_rate": 8.976646183494188e-06, + "loss": 0.4289, + "step": 10498 + }, + { + "epoch": 0.23110279550957516, + "grad_norm": 3.355654415322963, + "learning_rate": 8.97643009748028e-06, + "loss": 0.3934, + "step": 10499 + }, + { + "epoch": 0.23112480739599384, + "grad_norm": 1.4150734033193508, + "learning_rate": 8.97621399125638e-06, + "loss": 0.2603, + "step": 10500 + }, + { + "epoch": 0.2311468192824125, + "grad_norm": 0.6479396114209344, + "learning_rate": 8.975997864823587e-06, + "loss": 0.5118, + "step": 10501 + }, + { + "epoch": 0.23116883116883116, + "grad_norm": 1.9579320139835241, + "learning_rate": 8.975781718182998e-06, + "loss": 0.377, + "step": 10502 + }, + { + "epoch": 0.23119084305524984, + "grad_norm": 1.368093686591837, + "learning_rate": 8.975565551335714e-06, + "loss": 0.3462, + "step": 10503 + }, + { + "epoch": 0.2312128549416685, + "grad_norm": 1.8042778106107946, + "learning_rate": 8.975349364282832e-06, + "loss": 0.3089, + "step": 10504 + }, + { + "epoch": 0.23123486682808717, + "grad_norm": 1.6906277772819893, + "learning_rate": 8.975133157025453e-06, + "loss": 0.3263, + "step": 10505 + }, + { + "epoch": 0.23125687871450584, + "grad_norm": 1.3749046040148067, + "learning_rate": 8.974916929564673e-06, + "loss": 0.3404, + "step": 10506 + }, + { + "epoch": 0.2312788906009245, + "grad_norm": 0.6867345526596905, + "learning_rate": 8.974700681901593e-06, + "loss": 0.539, + "step": 10507 + }, + { + "epoch": 0.23130090248734317, + "grad_norm": 1.4276883705850383, + "learning_rate": 8.974484414037312e-06, + "loss": 0.3425, + "step": 10508 + }, + { + "epoch": 0.23132291437376182, + "grad_norm": 1.249386184435922, + "learning_rate": 8.974268125972928e-06, + "loss": 0.2931, + "step": 10509 + }, + { + "epoch": 0.2313449262601805, + "grad_norm": 1.803201408553178, + "learning_rate": 8.97405181770954e-06, + "loss": 0.3678, + "step": 10510 + }, + { + "epoch": 0.23136693814659917, + "grad_norm": 0.6325494169204887, + "learning_rate": 8.973835489248247e-06, + "loss": 0.5383, + "step": 10511 + }, + { + "epoch": 0.23138895003301782, + "grad_norm": 2.0497743430737434, + "learning_rate": 8.973619140590153e-06, + "loss": 0.3812, + "step": 10512 + }, + { + "epoch": 0.2314109619194365, + "grad_norm": 1.2770203269783518, + "learning_rate": 8.973402771736353e-06, + "loss": 0.3775, + "step": 10513 + }, + { + "epoch": 0.23143297380585517, + "grad_norm": 1.157854200009956, + "learning_rate": 8.973186382687946e-06, + "loss": 0.311, + "step": 10514 + }, + { + "epoch": 0.23145498569227382, + "grad_norm": 1.5432107802378066, + "learning_rate": 8.972969973446037e-06, + "loss": 0.3863, + "step": 10515 + }, + { + "epoch": 0.2314769975786925, + "grad_norm": 1.6657121944945492, + "learning_rate": 8.972753544011721e-06, + "loss": 0.4222, + "step": 10516 + }, + { + "epoch": 0.23149900946511115, + "grad_norm": 2.914181229192683, + "learning_rate": 8.9725370943861e-06, + "loss": 0.363, + "step": 10517 + }, + { + "epoch": 0.23152102135152983, + "grad_norm": 1.3783412815038671, + "learning_rate": 8.972320624570272e-06, + "loss": 0.28, + "step": 10518 + }, + { + "epoch": 0.2315430332379485, + "grad_norm": 1.2385299367602334, + "learning_rate": 8.97210413456534e-06, + "loss": 0.316, + "step": 10519 + }, + { + "epoch": 0.23156504512436715, + "grad_norm": 0.6353320492913686, + "learning_rate": 8.971887624372403e-06, + "loss": 0.5143, + "step": 10520 + }, + { + "epoch": 0.23158705701078583, + "grad_norm": 1.4459051495330966, + "learning_rate": 8.971671093992563e-06, + "loss": 0.3747, + "step": 10521 + }, + { + "epoch": 0.23160906889720448, + "grad_norm": 1.517774577988209, + "learning_rate": 8.971454543426916e-06, + "loss": 0.2681, + "step": 10522 + }, + { + "epoch": 0.23163108078362316, + "grad_norm": 0.6531296059753289, + "learning_rate": 8.971237972676569e-06, + "loss": 0.5434, + "step": 10523 + }, + { + "epoch": 0.23165309267004183, + "grad_norm": 0.614572272660015, + "learning_rate": 8.971021381742618e-06, + "loss": 0.506, + "step": 10524 + }, + { + "epoch": 0.23167510455646048, + "grad_norm": 1.2601327265646334, + "learning_rate": 8.970804770626164e-06, + "loss": 0.2949, + "step": 10525 + }, + { + "epoch": 0.23169711644287916, + "grad_norm": 1.6117889279929003, + "learning_rate": 8.970588139328311e-06, + "loss": 0.3545, + "step": 10526 + }, + { + "epoch": 0.2317191283292978, + "grad_norm": 0.6206794263746771, + "learning_rate": 8.970371487850158e-06, + "loss": 0.5123, + "step": 10527 + }, + { + "epoch": 0.23174114021571648, + "grad_norm": 1.5846464258567798, + "learning_rate": 8.970154816192805e-06, + "loss": 0.2486, + "step": 10528 + }, + { + "epoch": 0.23176315210213516, + "grad_norm": 1.2507231790392808, + "learning_rate": 8.969938124357357e-06, + "loss": 0.3315, + "step": 10529 + }, + { + "epoch": 0.2317851639885538, + "grad_norm": 1.6954732931962548, + "learning_rate": 8.96972141234491e-06, + "loss": 0.3975, + "step": 10530 + }, + { + "epoch": 0.2318071758749725, + "grad_norm": 1.3520809930725226, + "learning_rate": 8.96950468015657e-06, + "loss": 0.3777, + "step": 10531 + }, + { + "epoch": 0.23182918776139116, + "grad_norm": 0.6432689614959304, + "learning_rate": 8.969287927793436e-06, + "loss": 0.5283, + "step": 10532 + }, + { + "epoch": 0.2318511996478098, + "grad_norm": 1.668984607647036, + "learning_rate": 8.96907115525661e-06, + "loss": 0.3877, + "step": 10533 + }, + { + "epoch": 0.2318732115342285, + "grad_norm": 1.5334750989656039, + "learning_rate": 8.968854362547195e-06, + "loss": 0.4043, + "step": 10534 + }, + { + "epoch": 0.23189522342064714, + "grad_norm": 4.322678554866996, + "learning_rate": 8.96863754966629e-06, + "loss": 0.3516, + "step": 10535 + }, + { + "epoch": 0.23191723530706582, + "grad_norm": 1.9818495346780194, + "learning_rate": 8.968420716615e-06, + "loss": 0.2824, + "step": 10536 + }, + { + "epoch": 0.2319392471934845, + "grad_norm": 1.5300909801902078, + "learning_rate": 8.968203863394427e-06, + "loss": 0.1987, + "step": 10537 + }, + { + "epoch": 0.23196125907990314, + "grad_norm": 1.2263211725621255, + "learning_rate": 8.967986990005671e-06, + "loss": 0.3285, + "step": 10538 + }, + { + "epoch": 0.23198327096632182, + "grad_norm": 1.1604858851354143, + "learning_rate": 8.967770096449838e-06, + "loss": 0.2756, + "step": 10539 + }, + { + "epoch": 0.23200528285274047, + "grad_norm": 1.1519519073658782, + "learning_rate": 8.967553182728024e-06, + "loss": 0.3141, + "step": 10540 + }, + { + "epoch": 0.23202729473915915, + "grad_norm": 1.5965503841719575, + "learning_rate": 8.967336248841336e-06, + "loss": 0.3433, + "step": 10541 + }, + { + "epoch": 0.23204930662557782, + "grad_norm": 1.6742026561720909, + "learning_rate": 8.967119294790879e-06, + "loss": 0.3913, + "step": 10542 + }, + { + "epoch": 0.23207131851199647, + "grad_norm": 1.627961931911714, + "learning_rate": 8.966902320577748e-06, + "loss": 0.3742, + "step": 10543 + }, + { + "epoch": 0.23209333039841515, + "grad_norm": 1.4845251623774767, + "learning_rate": 8.966685326203053e-06, + "loss": 0.3196, + "step": 10544 + }, + { + "epoch": 0.2321153422848338, + "grad_norm": 1.5309750466888508, + "learning_rate": 8.966468311667893e-06, + "loss": 0.4035, + "step": 10545 + }, + { + "epoch": 0.23213735417125247, + "grad_norm": 1.4396811018583906, + "learning_rate": 8.96625127697337e-06, + "loss": 0.2588, + "step": 10546 + }, + { + "epoch": 0.23215936605767115, + "grad_norm": 1.3315397175125043, + "learning_rate": 8.966034222120592e-06, + "loss": 0.3093, + "step": 10547 + }, + { + "epoch": 0.2321813779440898, + "grad_norm": 1.4606645651712973, + "learning_rate": 8.965817147110657e-06, + "loss": 0.3794, + "step": 10548 + }, + { + "epoch": 0.23220338983050848, + "grad_norm": 1.491251108170987, + "learning_rate": 8.965600051944671e-06, + "loss": 0.4032, + "step": 10549 + }, + { + "epoch": 0.23222540171692715, + "grad_norm": 0.7142225861403123, + "learning_rate": 8.965382936623739e-06, + "loss": 0.547, + "step": 10550 + }, + { + "epoch": 0.2322474136033458, + "grad_norm": 2.0578385409658715, + "learning_rate": 8.96516580114896e-06, + "loss": 0.3718, + "step": 10551 + }, + { + "epoch": 0.23226942548976448, + "grad_norm": 2.521887562988583, + "learning_rate": 8.96494864552144e-06, + "loss": 0.3546, + "step": 10552 + }, + { + "epoch": 0.23229143737618313, + "grad_norm": 1.425017891122385, + "learning_rate": 8.964731469742283e-06, + "loss": 0.4875, + "step": 10553 + }, + { + "epoch": 0.2323134492626018, + "grad_norm": 1.647863990226379, + "learning_rate": 8.964514273812592e-06, + "loss": 0.3298, + "step": 10554 + }, + { + "epoch": 0.23233546114902048, + "grad_norm": 1.776179020672015, + "learning_rate": 8.964297057733473e-06, + "loss": 0.3405, + "step": 10555 + }, + { + "epoch": 0.23235747303543913, + "grad_norm": 1.6408947038865944, + "learning_rate": 8.964079821506026e-06, + "loss": 0.501, + "step": 10556 + }, + { + "epoch": 0.2323794849218578, + "grad_norm": 1.6975876003134862, + "learning_rate": 8.963862565131358e-06, + "loss": 0.343, + "step": 10557 + }, + { + "epoch": 0.23240149680827646, + "grad_norm": 0.639442063305006, + "learning_rate": 8.963645288610573e-06, + "loss": 0.5088, + "step": 10558 + }, + { + "epoch": 0.23242350869469514, + "grad_norm": 2.2854644872862813, + "learning_rate": 8.963427991944774e-06, + "loss": 0.3476, + "step": 10559 + }, + { + "epoch": 0.2324455205811138, + "grad_norm": 1.6840656459721364, + "learning_rate": 8.963210675135068e-06, + "loss": 0.359, + "step": 10560 + }, + { + "epoch": 0.23246753246753246, + "grad_norm": 2.353119458416678, + "learning_rate": 8.962993338182556e-06, + "loss": 0.2894, + "step": 10561 + }, + { + "epoch": 0.23248954435395114, + "grad_norm": 0.6448244500216557, + "learning_rate": 8.962775981088346e-06, + "loss": 0.5087, + "step": 10562 + }, + { + "epoch": 0.2325115562403698, + "grad_norm": 1.605005428657247, + "learning_rate": 8.962558603853537e-06, + "loss": 0.3287, + "step": 10563 + }, + { + "epoch": 0.23253356812678846, + "grad_norm": 1.4703374361441557, + "learning_rate": 8.962341206479242e-06, + "loss": 0.3801, + "step": 10564 + }, + { + "epoch": 0.23255558001320714, + "grad_norm": 1.931882761213841, + "learning_rate": 8.96212378896656e-06, + "loss": 0.2743, + "step": 10565 + }, + { + "epoch": 0.2325775918996258, + "grad_norm": 4.704475231464584, + "learning_rate": 8.961906351316598e-06, + "loss": 0.4218, + "step": 10566 + }, + { + "epoch": 0.23259960378604447, + "grad_norm": 1.6037179044022158, + "learning_rate": 8.96168889353046e-06, + "loss": 0.3679, + "step": 10567 + }, + { + "epoch": 0.23262161567246312, + "grad_norm": 1.221158666306104, + "learning_rate": 8.961471415609253e-06, + "loss": 0.3579, + "step": 10568 + }, + { + "epoch": 0.2326436275588818, + "grad_norm": 2.1926163385552075, + "learning_rate": 8.961253917554083e-06, + "loss": 0.4009, + "step": 10569 + }, + { + "epoch": 0.23266563944530047, + "grad_norm": 1.353517275418285, + "learning_rate": 8.961036399366053e-06, + "loss": 0.3024, + "step": 10570 + }, + { + "epoch": 0.23268765133171912, + "grad_norm": 1.4432008324106855, + "learning_rate": 8.960818861046267e-06, + "loss": 0.3936, + "step": 10571 + }, + { + "epoch": 0.2327096632181378, + "grad_norm": 1.6270765286136326, + "learning_rate": 8.960601302595835e-06, + "loss": 0.3522, + "step": 10572 + }, + { + "epoch": 0.23273167510455647, + "grad_norm": 1.4111049170602759, + "learning_rate": 8.96038372401586e-06, + "loss": 0.3992, + "step": 10573 + }, + { + "epoch": 0.23275368699097512, + "grad_norm": 0.7263170950300439, + "learning_rate": 8.960166125307448e-06, + "loss": 0.506, + "step": 10574 + }, + { + "epoch": 0.2327756988773938, + "grad_norm": 2.21144789826136, + "learning_rate": 8.959948506471708e-06, + "loss": 0.3228, + "step": 10575 + }, + { + "epoch": 0.23279771076381245, + "grad_norm": 1.3383900848321486, + "learning_rate": 8.95973086750974e-06, + "loss": 0.3497, + "step": 10576 + }, + { + "epoch": 0.23281972265023113, + "grad_norm": 2.386412561783656, + "learning_rate": 8.959513208422655e-06, + "loss": 0.5405, + "step": 10577 + }, + { + "epoch": 0.2328417345366498, + "grad_norm": 2.329540366560333, + "learning_rate": 8.95929552921156e-06, + "loss": 0.3817, + "step": 10578 + }, + { + "epoch": 0.23286374642306845, + "grad_norm": 1.3787085911014225, + "learning_rate": 8.959077829877556e-06, + "loss": 0.3113, + "step": 10579 + }, + { + "epoch": 0.23288575830948713, + "grad_norm": 1.5730384713995276, + "learning_rate": 8.958860110421755e-06, + "loss": 0.3369, + "step": 10580 + }, + { + "epoch": 0.23290777019590578, + "grad_norm": 1.6969666083306207, + "learning_rate": 8.95864237084526e-06, + "loss": 0.4272, + "step": 10581 + }, + { + "epoch": 0.23292978208232445, + "grad_norm": 1.5047251101316104, + "learning_rate": 8.958424611149179e-06, + "loss": 0.3666, + "step": 10582 + }, + { + "epoch": 0.23295179396874313, + "grad_norm": 0.6537562153725155, + "learning_rate": 8.95820683133462e-06, + "loss": 0.5181, + "step": 10583 + }, + { + "epoch": 0.23297380585516178, + "grad_norm": 1.3811994176818418, + "learning_rate": 8.957989031402686e-06, + "loss": 0.3535, + "step": 10584 + }, + { + "epoch": 0.23299581774158046, + "grad_norm": 1.3681424605878418, + "learning_rate": 8.957771211354488e-06, + "loss": 0.3213, + "step": 10585 + }, + { + "epoch": 0.2330178296279991, + "grad_norm": 2.2146021710930177, + "learning_rate": 8.957553371191132e-06, + "loss": 0.354, + "step": 10586 + }, + { + "epoch": 0.23303984151441778, + "grad_norm": 1.0824691188494266, + "learning_rate": 8.957335510913723e-06, + "loss": 0.2684, + "step": 10587 + }, + { + "epoch": 0.23306185340083646, + "grad_norm": 1.2378485225752545, + "learning_rate": 8.957117630523371e-06, + "loss": 0.3346, + "step": 10588 + }, + { + "epoch": 0.2330838652872551, + "grad_norm": 1.9101497311755706, + "learning_rate": 8.956899730021183e-06, + "loss": 0.3551, + "step": 10589 + }, + { + "epoch": 0.2331058771736738, + "grad_norm": 0.63732462969929, + "learning_rate": 8.956681809408264e-06, + "loss": 0.5133, + "step": 10590 + }, + { + "epoch": 0.23312788906009246, + "grad_norm": 0.6251465739065735, + "learning_rate": 8.956463868685724e-06, + "loss": 0.5325, + "step": 10591 + }, + { + "epoch": 0.2331499009465111, + "grad_norm": 2.3736529005067086, + "learning_rate": 8.95624590785467e-06, + "loss": 0.2919, + "step": 10592 + }, + { + "epoch": 0.2331719128329298, + "grad_norm": 1.672020242558704, + "learning_rate": 8.95602792691621e-06, + "loss": 0.3634, + "step": 10593 + }, + { + "epoch": 0.23319392471934844, + "grad_norm": 1.3261562470461286, + "learning_rate": 8.955809925871453e-06, + "loss": 0.3693, + "step": 10594 + }, + { + "epoch": 0.23321593660576712, + "grad_norm": 1.3305193841795009, + "learning_rate": 8.955591904721503e-06, + "loss": 0.2949, + "step": 10595 + }, + { + "epoch": 0.2332379484921858, + "grad_norm": 0.6403531411530955, + "learning_rate": 8.955373863467472e-06, + "loss": 0.5451, + "step": 10596 + }, + { + "epoch": 0.23325996037860444, + "grad_norm": 1.5543979530671905, + "learning_rate": 8.955155802110467e-06, + "loss": 0.3689, + "step": 10597 + }, + { + "epoch": 0.23328197226502312, + "grad_norm": 1.6792146883014174, + "learning_rate": 8.954937720651595e-06, + "loss": 0.3333, + "step": 10598 + }, + { + "epoch": 0.23330398415144177, + "grad_norm": 1.6040553088499148, + "learning_rate": 8.954719619091968e-06, + "loss": 0.3906, + "step": 10599 + }, + { + "epoch": 0.23332599603786044, + "grad_norm": 3.1100412691911643, + "learning_rate": 8.95450149743269e-06, + "loss": 0.4024, + "step": 10600 + }, + { + "epoch": 0.23334800792427912, + "grad_norm": 0.6395655320480995, + "learning_rate": 8.95428335567487e-06, + "loss": 0.5589, + "step": 10601 + }, + { + "epoch": 0.23337001981069777, + "grad_norm": 1.3528579675211534, + "learning_rate": 8.954065193819623e-06, + "loss": 0.3826, + "step": 10602 + }, + { + "epoch": 0.23339203169711645, + "grad_norm": 1.5784229169124078, + "learning_rate": 8.95384701186805e-06, + "loss": 0.3119, + "step": 10603 + }, + { + "epoch": 0.2334140435835351, + "grad_norm": 1.3800953954344912, + "learning_rate": 8.953628809821264e-06, + "loss": 0.2809, + "step": 10604 + }, + { + "epoch": 0.23343605546995377, + "grad_norm": 1.4441236080774236, + "learning_rate": 8.953410587680372e-06, + "loss": 0.3991, + "step": 10605 + }, + { + "epoch": 0.23345806735637245, + "grad_norm": 1.5998417372568003, + "learning_rate": 8.953192345446484e-06, + "loss": 0.3056, + "step": 10606 + }, + { + "epoch": 0.2334800792427911, + "grad_norm": 1.4343389830855655, + "learning_rate": 8.952974083120711e-06, + "loss": 0.3589, + "step": 10607 + }, + { + "epoch": 0.23350209112920978, + "grad_norm": 1.9157972632754965, + "learning_rate": 8.952755800704159e-06, + "loss": 0.4018, + "step": 10608 + }, + { + "epoch": 0.23352410301562843, + "grad_norm": 5.461923075668931, + "learning_rate": 8.95253749819794e-06, + "loss": 0.2619, + "step": 10609 + }, + { + "epoch": 0.2335461149020471, + "grad_norm": 1.4172616640027944, + "learning_rate": 8.952319175603162e-06, + "loss": 0.3024, + "step": 10610 + }, + { + "epoch": 0.23356812678846578, + "grad_norm": 1.3420129048510803, + "learning_rate": 8.952100832920936e-06, + "loss": 0.3691, + "step": 10611 + }, + { + "epoch": 0.23359013867488443, + "grad_norm": 0.6120248098255296, + "learning_rate": 8.95188247015237e-06, + "loss": 0.5145, + "step": 10612 + }, + { + "epoch": 0.2336121505613031, + "grad_norm": 1.4477284337459637, + "learning_rate": 8.951664087298576e-06, + "loss": 0.3677, + "step": 10613 + }, + { + "epoch": 0.23363416244772178, + "grad_norm": 1.5496022682780386, + "learning_rate": 8.951445684360661e-06, + "loss": 0.3446, + "step": 10614 + }, + { + "epoch": 0.23365617433414043, + "grad_norm": 1.612227900300026, + "learning_rate": 8.951227261339736e-06, + "loss": 0.3696, + "step": 10615 + }, + { + "epoch": 0.2336781862205591, + "grad_norm": 0.631348908329977, + "learning_rate": 8.951008818236915e-06, + "loss": 0.5322, + "step": 10616 + }, + { + "epoch": 0.23370019810697776, + "grad_norm": 1.6889005019997114, + "learning_rate": 8.950790355053303e-06, + "loss": 0.3776, + "step": 10617 + }, + { + "epoch": 0.23372220999339643, + "grad_norm": 1.3466287824972552, + "learning_rate": 8.950571871790012e-06, + "loss": 0.3652, + "step": 10618 + }, + { + "epoch": 0.2337442218798151, + "grad_norm": 0.6188796056499624, + "learning_rate": 8.950353368448153e-06, + "loss": 0.5212, + "step": 10619 + }, + { + "epoch": 0.23376623376623376, + "grad_norm": 0.6327052511388161, + "learning_rate": 8.950134845028836e-06, + "loss": 0.5582, + "step": 10620 + }, + { + "epoch": 0.23378824565265244, + "grad_norm": 2.1225996282731874, + "learning_rate": 8.949916301533173e-06, + "loss": 0.3952, + "step": 10621 + }, + { + "epoch": 0.2338102575390711, + "grad_norm": 1.4708864285555139, + "learning_rate": 8.949697737962273e-06, + "loss": 0.2609, + "step": 10622 + }, + { + "epoch": 0.23383226942548976, + "grad_norm": 0.6347648077426936, + "learning_rate": 8.949479154317247e-06, + "loss": 0.5365, + "step": 10623 + }, + { + "epoch": 0.23385428131190844, + "grad_norm": 1.330925116494025, + "learning_rate": 8.949260550599207e-06, + "loss": 0.4175, + "step": 10624 + }, + { + "epoch": 0.2338762931983271, + "grad_norm": 1.2406435230245447, + "learning_rate": 8.949041926809264e-06, + "loss": 0.3378, + "step": 10625 + }, + { + "epoch": 0.23389830508474577, + "grad_norm": 1.4315668573997535, + "learning_rate": 8.948823282948526e-06, + "loss": 0.3541, + "step": 10626 + }, + { + "epoch": 0.23392031697116442, + "grad_norm": 5.535084949243879, + "learning_rate": 8.948604619018109e-06, + "loss": 0.3601, + "step": 10627 + }, + { + "epoch": 0.2339423288575831, + "grad_norm": 2.2711399839242414, + "learning_rate": 8.948385935019123e-06, + "loss": 0.4537, + "step": 10628 + }, + { + "epoch": 0.23396434074400177, + "grad_norm": 0.6611084897388508, + "learning_rate": 8.948167230952678e-06, + "loss": 0.5454, + "step": 10629 + }, + { + "epoch": 0.23398635263042042, + "grad_norm": 2.255999184167663, + "learning_rate": 8.947948506819886e-06, + "loss": 0.3526, + "step": 10630 + }, + { + "epoch": 0.2340083645168391, + "grad_norm": 1.3453795499171066, + "learning_rate": 8.947729762621857e-06, + "loss": 0.3721, + "step": 10631 + }, + { + "epoch": 0.23403037640325777, + "grad_norm": 1.3434439868320036, + "learning_rate": 8.947510998359708e-06, + "loss": 0.3234, + "step": 10632 + }, + { + "epoch": 0.23405238828967642, + "grad_norm": 1.3406700043061595, + "learning_rate": 8.947292214034544e-06, + "loss": 0.2796, + "step": 10633 + }, + { + "epoch": 0.2340744001760951, + "grad_norm": 1.221920002568755, + "learning_rate": 8.947073409647482e-06, + "loss": 0.3587, + "step": 10634 + }, + { + "epoch": 0.23409641206251375, + "grad_norm": 1.3971690537595864, + "learning_rate": 8.946854585199634e-06, + "loss": 0.2517, + "step": 10635 + }, + { + "epoch": 0.23411842394893242, + "grad_norm": 2.1092270625190697, + "learning_rate": 8.946635740692109e-06, + "loss": 0.3756, + "step": 10636 + }, + { + "epoch": 0.2341404358353511, + "grad_norm": 2.072506810758764, + "learning_rate": 8.94641687612602e-06, + "loss": 0.3441, + "step": 10637 + }, + { + "epoch": 0.23416244772176975, + "grad_norm": 1.369523323018717, + "learning_rate": 8.946197991502483e-06, + "loss": 0.3369, + "step": 10638 + }, + { + "epoch": 0.23418445960818843, + "grad_norm": 2.552627162119198, + "learning_rate": 8.945979086822606e-06, + "loss": 0.4032, + "step": 10639 + }, + { + "epoch": 0.23420647149460708, + "grad_norm": 1.5666408877636293, + "learning_rate": 8.945760162087502e-06, + "loss": 0.3081, + "step": 10640 + }, + { + "epoch": 0.23422848338102575, + "grad_norm": 1.8673256443354513, + "learning_rate": 8.945541217298288e-06, + "loss": 0.3247, + "step": 10641 + }, + { + "epoch": 0.23425049526744443, + "grad_norm": 1.7524957579884934, + "learning_rate": 8.945322252456072e-06, + "loss": 0.4647, + "step": 10642 + }, + { + "epoch": 0.23427250715386308, + "grad_norm": 1.6294516983765681, + "learning_rate": 8.94510326756197e-06, + "loss": 0.3086, + "step": 10643 + }, + { + "epoch": 0.23429451904028176, + "grad_norm": 1.486234874637244, + "learning_rate": 8.944884262617091e-06, + "loss": 0.3113, + "step": 10644 + }, + { + "epoch": 0.2343165309267004, + "grad_norm": 1.6707069412595352, + "learning_rate": 8.944665237622552e-06, + "loss": 0.3442, + "step": 10645 + }, + { + "epoch": 0.23433854281311908, + "grad_norm": 2.679453658338526, + "learning_rate": 8.944446192579466e-06, + "loss": 0.3256, + "step": 10646 + }, + { + "epoch": 0.23436055469953776, + "grad_norm": 3.259373328394795, + "learning_rate": 8.944227127488945e-06, + "loss": 0.3597, + "step": 10647 + }, + { + "epoch": 0.2343825665859564, + "grad_norm": 1.2569471610851988, + "learning_rate": 8.944008042352101e-06, + "loss": 0.3199, + "step": 10648 + }, + { + "epoch": 0.23440457847237509, + "grad_norm": 1.6176115912149174, + "learning_rate": 8.943788937170051e-06, + "loss": 0.2729, + "step": 10649 + }, + { + "epoch": 0.23442659035879373, + "grad_norm": 1.266147702498455, + "learning_rate": 8.943569811943905e-06, + "loss": 0.2721, + "step": 10650 + }, + { + "epoch": 0.2344486022452124, + "grad_norm": 1.2187728609934865, + "learning_rate": 8.943350666674779e-06, + "loss": 0.3554, + "step": 10651 + }, + { + "epoch": 0.2344706141316311, + "grad_norm": 1.2148698289940854, + "learning_rate": 8.943131501363786e-06, + "loss": 0.3183, + "step": 10652 + }, + { + "epoch": 0.23449262601804974, + "grad_norm": 1.3864078292125337, + "learning_rate": 8.94291231601204e-06, + "loss": 0.3395, + "step": 10653 + }, + { + "epoch": 0.23451463790446841, + "grad_norm": 1.8833644963161087, + "learning_rate": 8.942693110620655e-06, + "loss": 0.2636, + "step": 10654 + }, + { + "epoch": 0.2345366497908871, + "grad_norm": 1.21376545762658, + "learning_rate": 8.942473885190745e-06, + "loss": 0.3405, + "step": 10655 + }, + { + "epoch": 0.23455866167730574, + "grad_norm": 1.5083697579251463, + "learning_rate": 8.942254639723426e-06, + "loss": 0.4063, + "step": 10656 + }, + { + "epoch": 0.23458067356372442, + "grad_norm": 1.755541117562679, + "learning_rate": 8.94203537421981e-06, + "loss": 0.3508, + "step": 10657 + }, + { + "epoch": 0.23460268545014307, + "grad_norm": 1.3972311979885843, + "learning_rate": 8.94181608868101e-06, + "loss": 0.4248, + "step": 10658 + }, + { + "epoch": 0.23462469733656174, + "grad_norm": 1.747003408104774, + "learning_rate": 8.941596783108145e-06, + "loss": 0.4608, + "step": 10659 + }, + { + "epoch": 0.23464670922298042, + "grad_norm": 1.6982002355780152, + "learning_rate": 8.941377457502325e-06, + "loss": 0.2983, + "step": 10660 + }, + { + "epoch": 0.23466872110939907, + "grad_norm": 3.002192652358751, + "learning_rate": 8.941158111864667e-06, + "loss": 0.2928, + "step": 10661 + }, + { + "epoch": 0.23469073299581775, + "grad_norm": 0.6448942846696831, + "learning_rate": 8.940938746196289e-06, + "loss": 0.5404, + "step": 10662 + }, + { + "epoch": 0.2347127448822364, + "grad_norm": 1.7438334940122948, + "learning_rate": 8.940719360498297e-06, + "loss": 0.2528, + "step": 10663 + }, + { + "epoch": 0.23473475676865507, + "grad_norm": 1.5233525007120632, + "learning_rate": 8.940499954771815e-06, + "loss": 0.4034, + "step": 10664 + }, + { + "epoch": 0.23475676865507375, + "grad_norm": 1.3342746645727481, + "learning_rate": 8.940280529017954e-06, + "loss": 0.251, + "step": 10665 + }, + { + "epoch": 0.2347787805414924, + "grad_norm": 2.197864423353109, + "learning_rate": 8.940061083237832e-06, + "loss": 0.4291, + "step": 10666 + }, + { + "epoch": 0.23480079242791108, + "grad_norm": 0.6689025072576841, + "learning_rate": 8.93984161743256e-06, + "loss": 0.5238, + "step": 10667 + }, + { + "epoch": 0.23482280431432973, + "grad_norm": 1.5945487665652487, + "learning_rate": 8.939622131603257e-06, + "loss": 0.4147, + "step": 10668 + }, + { + "epoch": 0.2348448162007484, + "grad_norm": 1.2040752232451601, + "learning_rate": 8.939402625751036e-06, + "loss": 0.2709, + "step": 10669 + }, + { + "epoch": 0.23486682808716708, + "grad_norm": 1.4233754910436698, + "learning_rate": 8.939183099877015e-06, + "loss": 0.3306, + "step": 10670 + }, + { + "epoch": 0.23488883997358573, + "grad_norm": 0.6229867370529432, + "learning_rate": 8.938963553982306e-06, + "loss": 0.4969, + "step": 10671 + }, + { + "epoch": 0.2349108518600044, + "grad_norm": 2.4596328964725274, + "learning_rate": 8.93874398806803e-06, + "loss": 0.4162, + "step": 10672 + }, + { + "epoch": 0.23493286374642308, + "grad_norm": 3.189325656582724, + "learning_rate": 8.9385244021353e-06, + "loss": 0.3331, + "step": 10673 + }, + { + "epoch": 0.23495487563284173, + "grad_norm": 1.4219217799161143, + "learning_rate": 8.938304796185232e-06, + "loss": 0.3864, + "step": 10674 + }, + { + "epoch": 0.2349768875192604, + "grad_norm": 1.3762600203996032, + "learning_rate": 8.938085170218942e-06, + "loss": 0.3469, + "step": 10675 + }, + { + "epoch": 0.23499889940567906, + "grad_norm": 1.4014679408080297, + "learning_rate": 8.937865524237548e-06, + "loss": 0.2762, + "step": 10676 + }, + { + "epoch": 0.23502091129209773, + "grad_norm": 2.2502635300170897, + "learning_rate": 8.937645858242163e-06, + "loss": 0.3362, + "step": 10677 + }, + { + "epoch": 0.2350429231785164, + "grad_norm": 1.522013904824152, + "learning_rate": 8.937426172233907e-06, + "loss": 0.3891, + "step": 10678 + }, + { + "epoch": 0.23506493506493506, + "grad_norm": 2.2303556375352604, + "learning_rate": 8.937206466213895e-06, + "loss": 0.352, + "step": 10679 + }, + { + "epoch": 0.23508694695135374, + "grad_norm": 1.4976826659698026, + "learning_rate": 8.936986740183244e-06, + "loss": 0.404, + "step": 10680 + }, + { + "epoch": 0.23510895883777239, + "grad_norm": 1.497782674262614, + "learning_rate": 8.936766994143069e-06, + "loss": 0.2264, + "step": 10681 + }, + { + "epoch": 0.23513097072419106, + "grad_norm": 1.9123068358573805, + "learning_rate": 8.93654722809449e-06, + "loss": 0.4556, + "step": 10682 + }, + { + "epoch": 0.23515298261060974, + "grad_norm": 13.042715417368168, + "learning_rate": 8.936327442038621e-06, + "loss": 0.3674, + "step": 10683 + }, + { + "epoch": 0.2351749944970284, + "grad_norm": 1.2625297033367047, + "learning_rate": 8.93610763597658e-06, + "loss": 0.3783, + "step": 10684 + }, + { + "epoch": 0.23519700638344707, + "grad_norm": 1.5727857898732902, + "learning_rate": 8.935887809909487e-06, + "loss": 0.493, + "step": 10685 + }, + { + "epoch": 0.23521901826986572, + "grad_norm": 1.3420335283224736, + "learning_rate": 8.935667963838455e-06, + "loss": 0.2731, + "step": 10686 + }, + { + "epoch": 0.2352410301562844, + "grad_norm": 0.6476549967757417, + "learning_rate": 8.935448097764603e-06, + "loss": 0.504, + "step": 10687 + }, + { + "epoch": 0.23526304204270307, + "grad_norm": 0.6741435650409888, + "learning_rate": 8.935228211689048e-06, + "loss": 0.5111, + "step": 10688 + }, + { + "epoch": 0.23528505392912172, + "grad_norm": 1.43232283687766, + "learning_rate": 8.93500830561291e-06, + "loss": 0.4187, + "step": 10689 + }, + { + "epoch": 0.2353070658155404, + "grad_norm": 1.5487817784621638, + "learning_rate": 8.934788379537302e-06, + "loss": 0.4356, + "step": 10690 + }, + { + "epoch": 0.23532907770195904, + "grad_norm": 1.718360761976197, + "learning_rate": 8.934568433463346e-06, + "loss": 0.3433, + "step": 10691 + }, + { + "epoch": 0.23535108958837772, + "grad_norm": 0.6266170539177185, + "learning_rate": 8.934348467392158e-06, + "loss": 0.5273, + "step": 10692 + }, + { + "epoch": 0.2353731014747964, + "grad_norm": 0.6245307608437444, + "learning_rate": 8.934128481324856e-06, + "loss": 0.5044, + "step": 10693 + }, + { + "epoch": 0.23539511336121505, + "grad_norm": 1.4197832064931255, + "learning_rate": 8.933908475262558e-06, + "loss": 0.392, + "step": 10694 + }, + { + "epoch": 0.23541712524763372, + "grad_norm": 0.6137232936957647, + "learning_rate": 8.933688449206384e-06, + "loss": 0.5139, + "step": 10695 + }, + { + "epoch": 0.2354391371340524, + "grad_norm": 1.6761285591399082, + "learning_rate": 8.93346840315745e-06, + "loss": 0.3763, + "step": 10696 + }, + { + "epoch": 0.23546114902047105, + "grad_norm": 1.3641525760714188, + "learning_rate": 8.933248337116876e-06, + "loss": 0.3388, + "step": 10697 + }, + { + "epoch": 0.23548316090688973, + "grad_norm": 1.5543070677300133, + "learning_rate": 8.933028251085777e-06, + "loss": 0.3443, + "step": 10698 + }, + { + "epoch": 0.23550517279330838, + "grad_norm": 1.6202917231769487, + "learning_rate": 8.932808145065277e-06, + "loss": 0.3114, + "step": 10699 + }, + { + "epoch": 0.23552718467972705, + "grad_norm": 7.078663349124451, + "learning_rate": 8.932588019056492e-06, + "loss": 0.2701, + "step": 10700 + }, + { + "epoch": 0.23554919656614573, + "grad_norm": 3.5150674883502693, + "learning_rate": 8.932367873060541e-06, + "loss": 0.3727, + "step": 10701 + }, + { + "epoch": 0.23557120845256438, + "grad_norm": 1.8108046210703026, + "learning_rate": 8.93214770707854e-06, + "loss": 0.3712, + "step": 10702 + }, + { + "epoch": 0.23559322033898306, + "grad_norm": 1.4625893695106715, + "learning_rate": 8.931927521111612e-06, + "loss": 0.3761, + "step": 10703 + }, + { + "epoch": 0.2356152322254017, + "grad_norm": 0.6720928760754845, + "learning_rate": 8.931707315160875e-06, + "loss": 0.4918, + "step": 10704 + }, + { + "epoch": 0.23563724411182038, + "grad_norm": 1.4109310611971693, + "learning_rate": 8.931487089227449e-06, + "loss": 0.3391, + "step": 10705 + }, + { + "epoch": 0.23565925599823906, + "grad_norm": 1.5099261173451306, + "learning_rate": 8.931266843312449e-06, + "loss": 0.2778, + "step": 10706 + }, + { + "epoch": 0.2356812678846577, + "grad_norm": 1.0894766841567134, + "learning_rate": 8.931046577416999e-06, + "loss": 0.236, + "step": 10707 + }, + { + "epoch": 0.23570327977107638, + "grad_norm": 1.3437155072827816, + "learning_rate": 8.930826291542219e-06, + "loss": 0.2871, + "step": 10708 + }, + { + "epoch": 0.23572529165749503, + "grad_norm": 0.6475980276906071, + "learning_rate": 8.930605985689223e-06, + "loss": 0.5276, + "step": 10709 + }, + { + "epoch": 0.2357473035439137, + "grad_norm": 3.174980429514313, + "learning_rate": 8.930385659859136e-06, + "loss": 0.3601, + "step": 10710 + }, + { + "epoch": 0.2357693154303324, + "grad_norm": 1.5732840657624407, + "learning_rate": 8.930165314053077e-06, + "loss": 0.3903, + "step": 10711 + }, + { + "epoch": 0.23579132731675104, + "grad_norm": 1.6432806111481564, + "learning_rate": 8.929944948272164e-06, + "loss": 0.3488, + "step": 10712 + }, + { + "epoch": 0.23581333920316971, + "grad_norm": 1.3666042124612572, + "learning_rate": 8.929724562517517e-06, + "loss": 0.3497, + "step": 10713 + }, + { + "epoch": 0.2358353510895884, + "grad_norm": 1.9465942900524833, + "learning_rate": 8.929504156790258e-06, + "loss": 0.3985, + "step": 10714 + }, + { + "epoch": 0.23585736297600704, + "grad_norm": 1.4159493604960716, + "learning_rate": 8.929283731091507e-06, + "loss": 0.3663, + "step": 10715 + }, + { + "epoch": 0.23587937486242572, + "grad_norm": 1.519242530645553, + "learning_rate": 8.929063285422384e-06, + "loss": 0.3696, + "step": 10716 + }, + { + "epoch": 0.23590138674884437, + "grad_norm": 0.6506751073068595, + "learning_rate": 8.928842819784007e-06, + "loss": 0.5199, + "step": 10717 + }, + { + "epoch": 0.23592339863526304, + "grad_norm": 1.1881608546620919, + "learning_rate": 8.9286223341775e-06, + "loss": 0.314, + "step": 10718 + }, + { + "epoch": 0.23594541052168172, + "grad_norm": 1.8300658315087985, + "learning_rate": 8.92840182860398e-06, + "loss": 0.3251, + "step": 10719 + }, + { + "epoch": 0.23596742240810037, + "grad_norm": 1.7590606437598377, + "learning_rate": 8.928181303064572e-06, + "loss": 0.4336, + "step": 10720 + }, + { + "epoch": 0.23598943429451905, + "grad_norm": 1.2152575656177733, + "learning_rate": 8.927960757560395e-06, + "loss": 0.3663, + "step": 10721 + }, + { + "epoch": 0.2360114461809377, + "grad_norm": 1.4937256869704478, + "learning_rate": 8.927740192092568e-06, + "loss": 0.3814, + "step": 10722 + }, + { + "epoch": 0.23603345806735637, + "grad_norm": 1.812819394528639, + "learning_rate": 8.927519606662214e-06, + "loss": 0.3584, + "step": 10723 + }, + { + "epoch": 0.23605546995377505, + "grad_norm": 1.9990644404899836, + "learning_rate": 8.927299001270454e-06, + "loss": 0.3367, + "step": 10724 + }, + { + "epoch": 0.2360774818401937, + "grad_norm": 2.140211722258421, + "learning_rate": 8.927078375918407e-06, + "loss": 0.3047, + "step": 10725 + }, + { + "epoch": 0.23609949372661237, + "grad_norm": 1.1929617283168106, + "learning_rate": 8.9268577306072e-06, + "loss": 0.2185, + "step": 10726 + }, + { + "epoch": 0.23612150561303102, + "grad_norm": 1.2447116428645537, + "learning_rate": 8.926637065337946e-06, + "loss": 0.3604, + "step": 10727 + }, + { + "epoch": 0.2361435174994497, + "grad_norm": 1.2416988665750999, + "learning_rate": 8.926416380111773e-06, + "loss": 0.3222, + "step": 10728 + }, + { + "epoch": 0.23616552938586838, + "grad_norm": 1.641483385491852, + "learning_rate": 8.926195674929802e-06, + "loss": 0.3896, + "step": 10729 + }, + { + "epoch": 0.23618754127228703, + "grad_norm": 1.3866849259491514, + "learning_rate": 8.925974949793153e-06, + "loss": 0.3201, + "step": 10730 + }, + { + "epoch": 0.2362095531587057, + "grad_norm": 1.7146821055247756, + "learning_rate": 8.925754204702948e-06, + "loss": 0.3853, + "step": 10731 + }, + { + "epoch": 0.23623156504512438, + "grad_norm": 1.660328409181126, + "learning_rate": 8.925533439660308e-06, + "loss": 0.248, + "step": 10732 + }, + { + "epoch": 0.23625357693154303, + "grad_norm": 1.7561828668037878, + "learning_rate": 8.925312654666358e-06, + "loss": 0.385, + "step": 10733 + }, + { + "epoch": 0.2362755888179617, + "grad_norm": 1.1060380462820283, + "learning_rate": 8.925091849722217e-06, + "loss": 0.3422, + "step": 10734 + }, + { + "epoch": 0.23629760070438036, + "grad_norm": 1.5286786343645062, + "learning_rate": 8.924871024829008e-06, + "loss": 0.3799, + "step": 10735 + }, + { + "epoch": 0.23631961259079903, + "grad_norm": 1.1397571754887734, + "learning_rate": 8.924650179987855e-06, + "loss": 0.3234, + "step": 10736 + }, + { + "epoch": 0.2363416244772177, + "grad_norm": 1.3901047172548717, + "learning_rate": 8.92442931519988e-06, + "loss": 0.3371, + "step": 10737 + }, + { + "epoch": 0.23636363636363636, + "grad_norm": 10.089196946000854, + "learning_rate": 8.924208430466203e-06, + "loss": 0.2945, + "step": 10738 + }, + { + "epoch": 0.23638564825005504, + "grad_norm": 1.7223373746318844, + "learning_rate": 8.923987525787952e-06, + "loss": 0.3348, + "step": 10739 + }, + { + "epoch": 0.23640766013647369, + "grad_norm": 1.6120095929156566, + "learning_rate": 8.923766601166243e-06, + "loss": 0.3686, + "step": 10740 + }, + { + "epoch": 0.23642967202289236, + "grad_norm": 1.2966581816328737, + "learning_rate": 8.923545656602202e-06, + "loss": 0.3859, + "step": 10741 + }, + { + "epoch": 0.23645168390931104, + "grad_norm": 0.6921379854878261, + "learning_rate": 8.923324692096954e-06, + "loss": 0.5208, + "step": 10742 + }, + { + "epoch": 0.2364736957957297, + "grad_norm": 1.7470517748517382, + "learning_rate": 8.923103707651618e-06, + "loss": 0.2623, + "step": 10743 + }, + { + "epoch": 0.23649570768214837, + "grad_norm": 1.6693211737191946, + "learning_rate": 8.922882703267322e-06, + "loss": 0.3486, + "step": 10744 + }, + { + "epoch": 0.23651771956856701, + "grad_norm": 1.2424790637543455, + "learning_rate": 8.922661678945183e-06, + "loss": 0.3726, + "step": 10745 + }, + { + "epoch": 0.2365397314549857, + "grad_norm": 2.174207380229056, + "learning_rate": 8.922440634686331e-06, + "loss": 0.3879, + "step": 10746 + }, + { + "epoch": 0.23656174334140437, + "grad_norm": 1.3403405582063093, + "learning_rate": 8.922219570491885e-06, + "loss": 0.3297, + "step": 10747 + }, + { + "epoch": 0.23658375522782302, + "grad_norm": 1.50399286790863, + "learning_rate": 8.921998486362969e-06, + "loss": 0.2789, + "step": 10748 + }, + { + "epoch": 0.2366057671142417, + "grad_norm": 0.6795648218000765, + "learning_rate": 8.921777382300708e-06, + "loss": 0.537, + "step": 10749 + }, + { + "epoch": 0.23662777900066034, + "grad_norm": 1.6551528916967269, + "learning_rate": 8.921556258306225e-06, + "loss": 0.293, + "step": 10750 + }, + { + "epoch": 0.23664979088707902, + "grad_norm": 1.353927396965847, + "learning_rate": 8.921335114380645e-06, + "loss": 0.3476, + "step": 10751 + }, + { + "epoch": 0.2366718027734977, + "grad_norm": 4.331479327935845, + "learning_rate": 8.921113950525088e-06, + "loss": 0.3153, + "step": 10752 + }, + { + "epoch": 0.23669381465991635, + "grad_norm": 1.1755447231607323, + "learning_rate": 8.920892766740686e-06, + "loss": 0.2676, + "step": 10753 + }, + { + "epoch": 0.23671582654633502, + "grad_norm": 1.1380080721103403, + "learning_rate": 8.920671563028555e-06, + "loss": 0.3768, + "step": 10754 + }, + { + "epoch": 0.2367378384327537, + "grad_norm": 1.572293425215645, + "learning_rate": 8.920450339389822e-06, + "loss": 0.3758, + "step": 10755 + }, + { + "epoch": 0.23675985031917235, + "grad_norm": 1.9854703289871807, + "learning_rate": 8.920229095825614e-06, + "loss": 0.314, + "step": 10756 + }, + { + "epoch": 0.23678186220559103, + "grad_norm": 1.1459187019038153, + "learning_rate": 8.920007832337051e-06, + "loss": 0.3383, + "step": 10757 + }, + { + "epoch": 0.23680387409200968, + "grad_norm": 1.428929229302515, + "learning_rate": 8.91978654892526e-06, + "loss": 0.3015, + "step": 10758 + }, + { + "epoch": 0.23682588597842835, + "grad_norm": 0.6581817453416695, + "learning_rate": 8.919565245591368e-06, + "loss": 0.5386, + "step": 10759 + }, + { + "epoch": 0.23684789786484703, + "grad_norm": 1.5463561294929602, + "learning_rate": 8.919343922336494e-06, + "loss": 0.3812, + "step": 10760 + }, + { + "epoch": 0.23686990975126568, + "grad_norm": 1.3707140345684496, + "learning_rate": 8.919122579161767e-06, + "loss": 0.3604, + "step": 10761 + }, + { + "epoch": 0.23689192163768436, + "grad_norm": 1.680306660368641, + "learning_rate": 8.918901216068313e-06, + "loss": 0.4103, + "step": 10762 + }, + { + "epoch": 0.236913933524103, + "grad_norm": 1.5447371205289833, + "learning_rate": 8.918679833057253e-06, + "loss": 0.2854, + "step": 10763 + }, + { + "epoch": 0.23693594541052168, + "grad_norm": 1.303822704995447, + "learning_rate": 8.918458430129717e-06, + "loss": 0.3472, + "step": 10764 + }, + { + "epoch": 0.23695795729694036, + "grad_norm": 1.2522940568239918, + "learning_rate": 8.918237007286825e-06, + "loss": 0.3563, + "step": 10765 + }, + { + "epoch": 0.236979969183359, + "grad_norm": 0.640636098768021, + "learning_rate": 8.918015564529705e-06, + "loss": 0.5254, + "step": 10766 + }, + { + "epoch": 0.23700198106977768, + "grad_norm": 1.5849897537469473, + "learning_rate": 8.917794101859484e-06, + "loss": 0.2687, + "step": 10767 + }, + { + "epoch": 0.23702399295619633, + "grad_norm": 1.4155947009921466, + "learning_rate": 8.917572619277285e-06, + "loss": 0.2919, + "step": 10768 + }, + { + "epoch": 0.237046004842615, + "grad_norm": 1.1609979202865992, + "learning_rate": 8.917351116784234e-06, + "loss": 0.3727, + "step": 10769 + }, + { + "epoch": 0.2370680167290337, + "grad_norm": 0.605977239058947, + "learning_rate": 8.917129594381458e-06, + "loss": 0.5171, + "step": 10770 + }, + { + "epoch": 0.23709002861545234, + "grad_norm": 1.2375732213551065, + "learning_rate": 8.916908052070083e-06, + "loss": 0.2621, + "step": 10771 + }, + { + "epoch": 0.237112040501871, + "grad_norm": 1.5975209134670327, + "learning_rate": 8.916686489851234e-06, + "loss": 0.3192, + "step": 10772 + }, + { + "epoch": 0.2371340523882897, + "grad_norm": 0.6158876183139923, + "learning_rate": 8.916464907726038e-06, + "loss": 0.5342, + "step": 10773 + }, + { + "epoch": 0.23715606427470834, + "grad_norm": 2.5476933634812737, + "learning_rate": 8.91624330569562e-06, + "loss": 0.3459, + "step": 10774 + }, + { + "epoch": 0.23717807616112702, + "grad_norm": 0.6099256908615253, + "learning_rate": 8.916021683761104e-06, + "loss": 0.519, + "step": 10775 + }, + { + "epoch": 0.23720008804754567, + "grad_norm": 2.3196431372546376, + "learning_rate": 8.915800041923623e-06, + "loss": 0.3729, + "step": 10776 + }, + { + "epoch": 0.23722209993396434, + "grad_norm": 1.4433199166760509, + "learning_rate": 8.915578380184297e-06, + "loss": 0.3175, + "step": 10777 + }, + { + "epoch": 0.23724411182038302, + "grad_norm": 1.1663120543478411, + "learning_rate": 8.915356698544257e-06, + "loss": 0.3656, + "step": 10778 + }, + { + "epoch": 0.23726612370680167, + "grad_norm": 4.364578140908825, + "learning_rate": 8.915134997004628e-06, + "loss": 0.2887, + "step": 10779 + }, + { + "epoch": 0.23728813559322035, + "grad_norm": 1.277601784500529, + "learning_rate": 8.914913275566537e-06, + "loss": 0.4081, + "step": 10780 + }, + { + "epoch": 0.237310147479639, + "grad_norm": 2.2858955926225217, + "learning_rate": 8.91469153423111e-06, + "loss": 0.4105, + "step": 10781 + }, + { + "epoch": 0.23733215936605767, + "grad_norm": 1.5187455273555164, + "learning_rate": 8.914469772999475e-06, + "loss": 0.3161, + "step": 10782 + }, + { + "epoch": 0.23735417125247635, + "grad_norm": 1.4261974052239128, + "learning_rate": 8.914247991872758e-06, + "loss": 0.3577, + "step": 10783 + }, + { + "epoch": 0.237376183138895, + "grad_norm": 1.2956121645437946, + "learning_rate": 8.914026190852087e-06, + "loss": 0.4152, + "step": 10784 + }, + { + "epoch": 0.23739819502531367, + "grad_norm": 0.6750224216575755, + "learning_rate": 8.91380436993859e-06, + "loss": 0.5327, + "step": 10785 + }, + { + "epoch": 0.23742020691173232, + "grad_norm": 1.4238310811665398, + "learning_rate": 8.913582529133393e-06, + "loss": 0.3039, + "step": 10786 + }, + { + "epoch": 0.237442218798151, + "grad_norm": 0.6388676386324591, + "learning_rate": 8.913360668437625e-06, + "loss": 0.5471, + "step": 10787 + }, + { + "epoch": 0.23746423068456968, + "grad_norm": 1.2307096908984636, + "learning_rate": 8.91313878785241e-06, + "loss": 0.3185, + "step": 10788 + }, + { + "epoch": 0.23748624257098833, + "grad_norm": 2.710217905572443, + "learning_rate": 8.912916887378881e-06, + "loss": 0.3261, + "step": 10789 + }, + { + "epoch": 0.237508254457407, + "grad_norm": 1.979725886984201, + "learning_rate": 8.912694967018163e-06, + "loss": 0.3994, + "step": 10790 + }, + { + "epoch": 0.23753026634382565, + "grad_norm": 1.719772279759934, + "learning_rate": 8.912473026771382e-06, + "loss": 0.35, + "step": 10791 + }, + { + "epoch": 0.23755227823024433, + "grad_norm": 1.272589778552982, + "learning_rate": 8.91225106663967e-06, + "loss": 0.3456, + "step": 10792 + }, + { + "epoch": 0.237574290116663, + "grad_norm": 0.6310360540155436, + "learning_rate": 8.912029086624153e-06, + "loss": 0.5171, + "step": 10793 + }, + { + "epoch": 0.23759630200308166, + "grad_norm": 0.6184011003220273, + "learning_rate": 8.91180708672596e-06, + "loss": 0.5286, + "step": 10794 + }, + { + "epoch": 0.23761831388950033, + "grad_norm": 1.2485958271253308, + "learning_rate": 8.911585066946216e-06, + "loss": 0.318, + "step": 10795 + }, + { + "epoch": 0.237640325775919, + "grad_norm": 2.1051231144261866, + "learning_rate": 8.911363027286055e-06, + "loss": 0.3268, + "step": 10796 + }, + { + "epoch": 0.23766233766233766, + "grad_norm": 1.4448018253344839, + "learning_rate": 8.9111409677466e-06, + "loss": 0.3137, + "step": 10797 + }, + { + "epoch": 0.23768434954875634, + "grad_norm": 1.4523500562824891, + "learning_rate": 8.910918888328984e-06, + "loss": 0.3392, + "step": 10798 + }, + { + "epoch": 0.23770636143517498, + "grad_norm": 1.305401229565338, + "learning_rate": 8.910696789034335e-06, + "loss": 0.3437, + "step": 10799 + }, + { + "epoch": 0.23772837332159366, + "grad_norm": 1.3193275822116117, + "learning_rate": 8.910474669863778e-06, + "loss": 0.3565, + "step": 10800 + }, + { + "epoch": 0.23775038520801234, + "grad_norm": 0.6608403760482919, + "learning_rate": 8.910252530818447e-06, + "loss": 0.5407, + "step": 10801 + }, + { + "epoch": 0.237772397094431, + "grad_norm": 1.5353190069039808, + "learning_rate": 8.910030371899466e-06, + "loss": 0.4264, + "step": 10802 + }, + { + "epoch": 0.23779440898084966, + "grad_norm": 1.622364324661473, + "learning_rate": 8.909808193107968e-06, + "loss": 0.4067, + "step": 10803 + }, + { + "epoch": 0.2378164208672683, + "grad_norm": 1.275426649065883, + "learning_rate": 8.909585994445082e-06, + "loss": 0.2943, + "step": 10804 + }, + { + "epoch": 0.237838432753687, + "grad_norm": 0.6447351031231797, + "learning_rate": 8.909363775911934e-06, + "loss": 0.5131, + "step": 10805 + }, + { + "epoch": 0.23786044464010567, + "grad_norm": 1.6050594788100119, + "learning_rate": 8.909141537509657e-06, + "loss": 0.2624, + "step": 10806 + }, + { + "epoch": 0.23788245652652432, + "grad_norm": 1.555888113907045, + "learning_rate": 8.90891927923938e-06, + "loss": 0.3449, + "step": 10807 + }, + { + "epoch": 0.237904468412943, + "grad_norm": 1.8100874454255078, + "learning_rate": 8.90869700110223e-06, + "loss": 0.3634, + "step": 10808 + }, + { + "epoch": 0.23792648029936164, + "grad_norm": 0.6578130253690858, + "learning_rate": 8.908474703099339e-06, + "loss": 0.5453, + "step": 10809 + }, + { + "epoch": 0.23794849218578032, + "grad_norm": 1.5697951704539814, + "learning_rate": 8.908252385231837e-06, + "loss": 0.3326, + "step": 10810 + }, + { + "epoch": 0.237970504072199, + "grad_norm": 2.509946050894452, + "learning_rate": 8.908030047500853e-06, + "loss": 0.3183, + "step": 10811 + }, + { + "epoch": 0.23799251595861765, + "grad_norm": 1.3839343805805675, + "learning_rate": 8.907807689907518e-06, + "loss": 0.341, + "step": 10812 + }, + { + "epoch": 0.23801452784503632, + "grad_norm": 0.6621375599567321, + "learning_rate": 8.90758531245296e-06, + "loss": 0.5358, + "step": 10813 + }, + { + "epoch": 0.238036539731455, + "grad_norm": 0.6263878960635396, + "learning_rate": 8.907362915138311e-06, + "loss": 0.5194, + "step": 10814 + }, + { + "epoch": 0.23805855161787365, + "grad_norm": 1.4465653556505063, + "learning_rate": 8.9071404979647e-06, + "loss": 0.3455, + "step": 10815 + }, + { + "epoch": 0.23808056350429233, + "grad_norm": 1.2771026414440956, + "learning_rate": 8.906918060933262e-06, + "loss": 0.3179, + "step": 10816 + }, + { + "epoch": 0.23810257539071097, + "grad_norm": 1.519258454830207, + "learning_rate": 8.906695604045121e-06, + "loss": 0.4162, + "step": 10817 + }, + { + "epoch": 0.23812458727712965, + "grad_norm": 1.400614459747452, + "learning_rate": 8.906473127301411e-06, + "loss": 0.3327, + "step": 10818 + }, + { + "epoch": 0.23814659916354833, + "grad_norm": 0.6392550306631353, + "learning_rate": 8.906250630703261e-06, + "loss": 0.537, + "step": 10819 + }, + { + "epoch": 0.23816861104996698, + "grad_norm": 1.6124034943141878, + "learning_rate": 8.906028114251806e-06, + "loss": 0.3875, + "step": 10820 + }, + { + "epoch": 0.23819062293638565, + "grad_norm": 1.2447998316493538, + "learning_rate": 8.905805577948172e-06, + "loss": 0.3652, + "step": 10821 + }, + { + "epoch": 0.2382126348228043, + "grad_norm": 0.6342100801797993, + "learning_rate": 8.905583021793492e-06, + "loss": 0.5265, + "step": 10822 + }, + { + "epoch": 0.23823464670922298, + "grad_norm": 1.448562703415936, + "learning_rate": 8.905360445788898e-06, + "loss": 0.3963, + "step": 10823 + }, + { + "epoch": 0.23825665859564166, + "grad_norm": 1.4976375078757487, + "learning_rate": 8.905137849935519e-06, + "loss": 0.327, + "step": 10824 + }, + { + "epoch": 0.2382786704820603, + "grad_norm": 1.3949007939852802, + "learning_rate": 8.904915234234488e-06, + "loss": 0.3887, + "step": 10825 + }, + { + "epoch": 0.23830068236847898, + "grad_norm": 1.5054217127874454, + "learning_rate": 8.904692598686937e-06, + "loss": 0.3139, + "step": 10826 + }, + { + "epoch": 0.23832269425489763, + "grad_norm": 1.8235900678684476, + "learning_rate": 8.904469943293996e-06, + "loss": 0.3026, + "step": 10827 + }, + { + "epoch": 0.2383447061413163, + "grad_norm": 0.6770412774940692, + "learning_rate": 8.904247268056799e-06, + "loss": 0.5393, + "step": 10828 + }, + { + "epoch": 0.238366718027735, + "grad_norm": 1.3646149508685195, + "learning_rate": 8.904024572976473e-06, + "loss": 0.3534, + "step": 10829 + }, + { + "epoch": 0.23838872991415364, + "grad_norm": 1.4684153192732714, + "learning_rate": 8.903801858054156e-06, + "loss": 0.4703, + "step": 10830 + }, + { + "epoch": 0.2384107418005723, + "grad_norm": 1.1743101625083752, + "learning_rate": 8.903579123290975e-06, + "loss": 0.278, + "step": 10831 + }, + { + "epoch": 0.23843275368699096, + "grad_norm": 1.640164807134789, + "learning_rate": 8.903356368688063e-06, + "loss": 0.3773, + "step": 10832 + }, + { + "epoch": 0.23845476557340964, + "grad_norm": 1.4320427587387194, + "learning_rate": 8.903133594246555e-06, + "loss": 0.4111, + "step": 10833 + }, + { + "epoch": 0.23847677745982832, + "grad_norm": 1.3570865805502794, + "learning_rate": 8.90291079996758e-06, + "loss": 0.4515, + "step": 10834 + }, + { + "epoch": 0.23849878934624696, + "grad_norm": 1.4705108067247503, + "learning_rate": 8.902687985852272e-06, + "loss": 0.3237, + "step": 10835 + }, + { + "epoch": 0.23852080123266564, + "grad_norm": 1.3461684055810124, + "learning_rate": 8.902465151901762e-06, + "loss": 0.4003, + "step": 10836 + }, + { + "epoch": 0.23854281311908432, + "grad_norm": 1.2518003279082177, + "learning_rate": 8.902242298117185e-06, + "loss": 0.3811, + "step": 10837 + }, + { + "epoch": 0.23856482500550297, + "grad_norm": 0.6613650064263421, + "learning_rate": 8.902019424499673e-06, + "loss": 0.5452, + "step": 10838 + }, + { + "epoch": 0.23858683689192164, + "grad_norm": 0.6170991222304799, + "learning_rate": 8.901796531050356e-06, + "loss": 0.5163, + "step": 10839 + }, + { + "epoch": 0.2386088487783403, + "grad_norm": 1.429699861320187, + "learning_rate": 8.901573617770369e-06, + "loss": 0.3211, + "step": 10840 + }, + { + "epoch": 0.23863086066475897, + "grad_norm": 1.8028165439234094, + "learning_rate": 8.901350684660845e-06, + "loss": 0.4774, + "step": 10841 + }, + { + "epoch": 0.23865287255117765, + "grad_norm": 1.320411939400682, + "learning_rate": 8.901127731722917e-06, + "loss": 0.3803, + "step": 10842 + }, + { + "epoch": 0.2386748844375963, + "grad_norm": 1.378209885738619, + "learning_rate": 8.900904758957718e-06, + "loss": 0.2736, + "step": 10843 + }, + { + "epoch": 0.23869689632401497, + "grad_norm": 1.9588923087077137, + "learning_rate": 8.900681766366382e-06, + "loss": 0.3134, + "step": 10844 + }, + { + "epoch": 0.23871890821043362, + "grad_norm": 0.6575048012737148, + "learning_rate": 8.90045875395004e-06, + "loss": 0.526, + "step": 10845 + }, + { + "epoch": 0.2387409200968523, + "grad_norm": 1.2636914711420468, + "learning_rate": 8.900235721709827e-06, + "loss": 0.3766, + "step": 10846 + }, + { + "epoch": 0.23876293198327098, + "grad_norm": 1.4952023389852738, + "learning_rate": 8.900012669646876e-06, + "loss": 0.3356, + "step": 10847 + }, + { + "epoch": 0.23878494386968963, + "grad_norm": 1.2401762368683433, + "learning_rate": 8.899789597762322e-06, + "loss": 0.3303, + "step": 10848 + }, + { + "epoch": 0.2388069557561083, + "grad_norm": 1.3001404763398718, + "learning_rate": 8.899566506057299e-06, + "loss": 0.3454, + "step": 10849 + }, + { + "epoch": 0.23882896764252695, + "grad_norm": 1.1221952154248875, + "learning_rate": 8.899343394532938e-06, + "loss": 0.3431, + "step": 10850 + }, + { + "epoch": 0.23885097952894563, + "grad_norm": 1.0729036223571267, + "learning_rate": 8.899120263190376e-06, + "loss": 0.3374, + "step": 10851 + }, + { + "epoch": 0.2388729914153643, + "grad_norm": 1.4388691043508464, + "learning_rate": 8.898897112030745e-06, + "loss": 0.3564, + "step": 10852 + }, + { + "epoch": 0.23889500330178295, + "grad_norm": 1.6228022480190059, + "learning_rate": 8.898673941055181e-06, + "loss": 0.3335, + "step": 10853 + }, + { + "epoch": 0.23891701518820163, + "grad_norm": 4.7305915099157785, + "learning_rate": 8.898450750264816e-06, + "loss": 0.2803, + "step": 10854 + }, + { + "epoch": 0.2389390270746203, + "grad_norm": 1.4886376820566463, + "learning_rate": 8.898227539660787e-06, + "loss": 0.3035, + "step": 10855 + }, + { + "epoch": 0.23896103896103896, + "grad_norm": 1.7121241091969959, + "learning_rate": 8.898004309244225e-06, + "loss": 0.279, + "step": 10856 + }, + { + "epoch": 0.23898305084745763, + "grad_norm": 4.208233157165637, + "learning_rate": 8.897781059016268e-06, + "loss": 0.4039, + "step": 10857 + }, + { + "epoch": 0.23900506273387628, + "grad_norm": 1.2464868406475555, + "learning_rate": 8.897557788978048e-06, + "loss": 0.3662, + "step": 10858 + }, + { + "epoch": 0.23902707462029496, + "grad_norm": 0.711060574410266, + "learning_rate": 8.897334499130702e-06, + "loss": 0.5186, + "step": 10859 + }, + { + "epoch": 0.23904908650671364, + "grad_norm": 1.3385979771035963, + "learning_rate": 8.897111189475363e-06, + "loss": 0.3105, + "step": 10860 + }, + { + "epoch": 0.2390710983931323, + "grad_norm": 1.4513033275126797, + "learning_rate": 8.896887860013167e-06, + "loss": 0.3708, + "step": 10861 + }, + { + "epoch": 0.23909311027955096, + "grad_norm": 1.3615556864126783, + "learning_rate": 8.89666451074525e-06, + "loss": 0.3286, + "step": 10862 + }, + { + "epoch": 0.2391151221659696, + "grad_norm": 1.3391666300291405, + "learning_rate": 8.896441141672744e-06, + "loss": 0.3371, + "step": 10863 + }, + { + "epoch": 0.2391371340523883, + "grad_norm": 1.3266461811250134, + "learning_rate": 8.896217752796787e-06, + "loss": 0.3443, + "step": 10864 + }, + { + "epoch": 0.23915914593880697, + "grad_norm": 1.3511850575699178, + "learning_rate": 8.895994344118515e-06, + "loss": 0.2863, + "step": 10865 + }, + { + "epoch": 0.23918115782522562, + "grad_norm": 0.663081275591996, + "learning_rate": 8.89577091563906e-06, + "loss": 0.5195, + "step": 10866 + }, + { + "epoch": 0.2392031697116443, + "grad_norm": 1.2971998164199174, + "learning_rate": 8.895547467359561e-06, + "loss": 0.4134, + "step": 10867 + }, + { + "epoch": 0.23922518159806294, + "grad_norm": 2.2128029581086683, + "learning_rate": 8.895323999281152e-06, + "loss": 0.3307, + "step": 10868 + }, + { + "epoch": 0.23924719348448162, + "grad_norm": 1.5241629464423405, + "learning_rate": 8.895100511404969e-06, + "loss": 0.3844, + "step": 10869 + }, + { + "epoch": 0.2392692053709003, + "grad_norm": 0.6464385523879812, + "learning_rate": 8.894877003732148e-06, + "loss": 0.542, + "step": 10870 + }, + { + "epoch": 0.23929121725731894, + "grad_norm": 3.8623268157366533, + "learning_rate": 8.894653476263825e-06, + "loss": 0.3319, + "step": 10871 + }, + { + "epoch": 0.23931322914373762, + "grad_norm": 1.475987968820491, + "learning_rate": 8.894429929001135e-06, + "loss": 0.3511, + "step": 10872 + }, + { + "epoch": 0.2393352410301563, + "grad_norm": 0.66203292991518, + "learning_rate": 8.894206361945217e-06, + "loss": 0.5486, + "step": 10873 + }, + { + "epoch": 0.23935725291657495, + "grad_norm": 3.7694942555607436, + "learning_rate": 8.893982775097203e-06, + "loss": 0.2103, + "step": 10874 + }, + { + "epoch": 0.23937926480299362, + "grad_norm": 0.8251264816754582, + "learning_rate": 8.893759168458233e-06, + "loss": 0.5332, + "step": 10875 + }, + { + "epoch": 0.23940127668941227, + "grad_norm": 1.4472638726225235, + "learning_rate": 8.893535542029444e-06, + "loss": 0.2622, + "step": 10876 + }, + { + "epoch": 0.23942328857583095, + "grad_norm": 1.246882502862271, + "learning_rate": 8.893311895811967e-06, + "loss": 0.3295, + "step": 10877 + }, + { + "epoch": 0.23944530046224963, + "grad_norm": 1.7187429363564444, + "learning_rate": 8.893088229806945e-06, + "loss": 0.3594, + "step": 10878 + }, + { + "epoch": 0.23946731234866828, + "grad_norm": 1.6088778780652493, + "learning_rate": 8.892864544015512e-06, + "loss": 0.3423, + "step": 10879 + }, + { + "epoch": 0.23948932423508695, + "grad_norm": 2.3171874257726217, + "learning_rate": 8.892640838438807e-06, + "loss": 0.4043, + "step": 10880 + }, + { + "epoch": 0.2395113361215056, + "grad_norm": 1.502862742094338, + "learning_rate": 8.892417113077963e-06, + "loss": 0.3177, + "step": 10881 + }, + { + "epoch": 0.23953334800792428, + "grad_norm": 2.201386744712159, + "learning_rate": 8.892193367934118e-06, + "loss": 0.3168, + "step": 10882 + }, + { + "epoch": 0.23955535989434296, + "grad_norm": 0.6923859560113231, + "learning_rate": 8.891969603008412e-06, + "loss": 0.4948, + "step": 10883 + }, + { + "epoch": 0.2395773717807616, + "grad_norm": 2.596029740839805, + "learning_rate": 8.89174581830198e-06, + "loss": 0.3916, + "step": 10884 + }, + { + "epoch": 0.23959938366718028, + "grad_norm": 1.4655534212758448, + "learning_rate": 8.89152201381596e-06, + "loss": 0.3275, + "step": 10885 + }, + { + "epoch": 0.23962139555359893, + "grad_norm": 3.951311060767158, + "learning_rate": 8.89129818955149e-06, + "loss": 0.3563, + "step": 10886 + }, + { + "epoch": 0.2396434074400176, + "grad_norm": 1.3573190616756463, + "learning_rate": 8.891074345509705e-06, + "loss": 0.3882, + "step": 10887 + }, + { + "epoch": 0.23966541932643629, + "grad_norm": 1.9019691325593653, + "learning_rate": 8.890850481691748e-06, + "loss": 0.3856, + "step": 10888 + }, + { + "epoch": 0.23968743121285493, + "grad_norm": 1.338788475006653, + "learning_rate": 8.890626598098752e-06, + "loss": 0.358, + "step": 10889 + }, + { + "epoch": 0.2397094430992736, + "grad_norm": 1.43789732798259, + "learning_rate": 8.890402694731855e-06, + "loss": 0.3282, + "step": 10890 + }, + { + "epoch": 0.23973145498569226, + "grad_norm": 0.6210629754317887, + "learning_rate": 8.890178771592198e-06, + "loss": 0.5289, + "step": 10891 + }, + { + "epoch": 0.23975346687211094, + "grad_norm": 0.5987688108492177, + "learning_rate": 8.889954828680918e-06, + "loss": 0.4973, + "step": 10892 + }, + { + "epoch": 0.23977547875852961, + "grad_norm": 0.6058378014279944, + "learning_rate": 8.889730865999151e-06, + "loss": 0.5198, + "step": 10893 + }, + { + "epoch": 0.23979749064494826, + "grad_norm": 1.2328278350126702, + "learning_rate": 8.889506883548038e-06, + "loss": 0.3067, + "step": 10894 + }, + { + "epoch": 0.23981950253136694, + "grad_norm": 1.2378008301250583, + "learning_rate": 8.889282881328714e-06, + "loss": 0.3441, + "step": 10895 + }, + { + "epoch": 0.23984151441778562, + "grad_norm": 1.3530089349994148, + "learning_rate": 8.889058859342323e-06, + "loss": 0.359, + "step": 10896 + }, + { + "epoch": 0.23986352630420427, + "grad_norm": 1.3025851161103876, + "learning_rate": 8.888834817589999e-06, + "loss": 0.3078, + "step": 10897 + }, + { + "epoch": 0.23988553819062294, + "grad_norm": 1.5016826518996345, + "learning_rate": 8.888610756072882e-06, + "loss": 0.3295, + "step": 10898 + }, + { + "epoch": 0.2399075500770416, + "grad_norm": 1.4086907200186167, + "learning_rate": 8.888386674792109e-06, + "loss": 0.2692, + "step": 10899 + }, + { + "epoch": 0.23992956196346027, + "grad_norm": 1.1982540462953044, + "learning_rate": 8.888162573748822e-06, + "loss": 0.2268, + "step": 10900 + }, + { + "epoch": 0.23995157384987895, + "grad_norm": 1.2741677920141405, + "learning_rate": 8.88793845294416e-06, + "loss": 0.3303, + "step": 10901 + }, + { + "epoch": 0.2399735857362976, + "grad_norm": 1.3611577853635717, + "learning_rate": 8.887714312379259e-06, + "loss": 0.4036, + "step": 10902 + }, + { + "epoch": 0.23999559762271627, + "grad_norm": 1.378108138232602, + "learning_rate": 8.88749015205526e-06, + "loss": 0.3239, + "step": 10903 + }, + { + "epoch": 0.24001760950913492, + "grad_norm": 1.426763817694869, + "learning_rate": 8.887265971973302e-06, + "loss": 0.3771, + "step": 10904 + }, + { + "epoch": 0.2400396213955536, + "grad_norm": 1.7295787026146565, + "learning_rate": 8.887041772134527e-06, + "loss": 0.4409, + "step": 10905 + }, + { + "epoch": 0.24006163328197228, + "grad_norm": 1.934445378241475, + "learning_rate": 8.886817552540069e-06, + "loss": 0.3669, + "step": 10906 + }, + { + "epoch": 0.24008364516839092, + "grad_norm": 1.5108067029507979, + "learning_rate": 8.886593313191072e-06, + "loss": 0.3632, + "step": 10907 + }, + { + "epoch": 0.2401056570548096, + "grad_norm": 1.3451602442256236, + "learning_rate": 8.886369054088673e-06, + "loss": 0.3391, + "step": 10908 + }, + { + "epoch": 0.24012766894122825, + "grad_norm": 1.212846378458706, + "learning_rate": 8.886144775234016e-06, + "loss": 0.2812, + "step": 10909 + }, + { + "epoch": 0.24014968082764693, + "grad_norm": 1.2430730788538777, + "learning_rate": 8.885920476628235e-06, + "loss": 0.2983, + "step": 10910 + }, + { + "epoch": 0.2401716927140656, + "grad_norm": 2.0421845551840994, + "learning_rate": 8.885696158272475e-06, + "loss": 0.443, + "step": 10911 + }, + { + "epoch": 0.24019370460048425, + "grad_norm": 1.5601980601341197, + "learning_rate": 8.885471820167873e-06, + "loss": 0.3467, + "step": 10912 + }, + { + "epoch": 0.24021571648690293, + "grad_norm": 0.7845399233481412, + "learning_rate": 8.88524746231557e-06, + "loss": 0.4985, + "step": 10913 + }, + { + "epoch": 0.2402377283733216, + "grad_norm": 1.2541073532464255, + "learning_rate": 8.885023084716707e-06, + "loss": 0.2735, + "step": 10914 + }, + { + "epoch": 0.24025974025974026, + "grad_norm": 2.213904364514413, + "learning_rate": 8.884798687372425e-06, + "loss": 0.3873, + "step": 10915 + }, + { + "epoch": 0.24028175214615893, + "grad_norm": 1.3239904863935494, + "learning_rate": 8.884574270283862e-06, + "loss": 0.3047, + "step": 10916 + }, + { + "epoch": 0.24030376403257758, + "grad_norm": 1.3611251167843066, + "learning_rate": 8.88434983345216e-06, + "loss": 0.4432, + "step": 10917 + }, + { + "epoch": 0.24032577591899626, + "grad_norm": 1.5029973341064444, + "learning_rate": 8.88412537687846e-06, + "loss": 0.3396, + "step": 10918 + }, + { + "epoch": 0.24034778780541494, + "grad_norm": 1.6971210330245543, + "learning_rate": 8.883900900563903e-06, + "loss": 0.4073, + "step": 10919 + }, + { + "epoch": 0.24036979969183359, + "grad_norm": 1.4106383923302712, + "learning_rate": 8.883676404509627e-06, + "loss": 0.4201, + "step": 10920 + }, + { + "epoch": 0.24039181157825226, + "grad_norm": 1.5351378436198717, + "learning_rate": 8.883451888716778e-06, + "loss": 0.3702, + "step": 10921 + }, + { + "epoch": 0.2404138234646709, + "grad_norm": 0.6842045851149274, + "learning_rate": 8.883227353186495e-06, + "loss": 0.5427, + "step": 10922 + }, + { + "epoch": 0.2404358353510896, + "grad_norm": 1.3625542352119178, + "learning_rate": 8.883002797919915e-06, + "loss": 0.3635, + "step": 10923 + }, + { + "epoch": 0.24045784723750827, + "grad_norm": 1.3728878636284143, + "learning_rate": 8.882778222918186e-06, + "loss": 0.3504, + "step": 10924 + }, + { + "epoch": 0.24047985912392691, + "grad_norm": 1.217785291866031, + "learning_rate": 8.882553628182446e-06, + "loss": 0.3065, + "step": 10925 + }, + { + "epoch": 0.2405018710103456, + "grad_norm": 1.9389301651442714, + "learning_rate": 8.882329013713835e-06, + "loss": 0.3764, + "step": 10926 + }, + { + "epoch": 0.24052388289676424, + "grad_norm": 1.2345923056417232, + "learning_rate": 8.882104379513498e-06, + "loss": 0.3799, + "step": 10927 + }, + { + "epoch": 0.24054589478318292, + "grad_norm": 1.0773760942961539, + "learning_rate": 8.881879725582575e-06, + "loss": 0.2568, + "step": 10928 + }, + { + "epoch": 0.2405679066696016, + "grad_norm": 2.3552351679272694, + "learning_rate": 8.881655051922207e-06, + "loss": 0.444, + "step": 10929 + }, + { + "epoch": 0.24058991855602024, + "grad_norm": 1.2850036718796813, + "learning_rate": 8.881430358533539e-06, + "loss": 0.4122, + "step": 10930 + }, + { + "epoch": 0.24061193044243892, + "grad_norm": 1.939849003089862, + "learning_rate": 8.881205645417707e-06, + "loss": 0.2629, + "step": 10931 + }, + { + "epoch": 0.24063394232885757, + "grad_norm": 1.720832550857095, + "learning_rate": 8.880980912575859e-06, + "loss": 0.311, + "step": 10932 + }, + { + "epoch": 0.24065595421527625, + "grad_norm": 0.6546285251299048, + "learning_rate": 8.880756160009136e-06, + "loss": 0.5119, + "step": 10933 + }, + { + "epoch": 0.24067796610169492, + "grad_norm": 1.4104038686215463, + "learning_rate": 8.880531387718678e-06, + "loss": 0.335, + "step": 10934 + }, + { + "epoch": 0.24069997798811357, + "grad_norm": 1.41044212040722, + "learning_rate": 8.880306595705629e-06, + "loss": 0.3487, + "step": 10935 + }, + { + "epoch": 0.24072198987453225, + "grad_norm": 1.7033251583682707, + "learning_rate": 8.88008178397113e-06, + "loss": 0.3167, + "step": 10936 + }, + { + "epoch": 0.24074400176095093, + "grad_norm": 1.5240402210668396, + "learning_rate": 8.879856952516328e-06, + "loss": 0.3914, + "step": 10937 + }, + { + "epoch": 0.24076601364736958, + "grad_norm": 1.3340232491065318, + "learning_rate": 8.879632101342359e-06, + "loss": 0.4098, + "step": 10938 + }, + { + "epoch": 0.24078802553378825, + "grad_norm": 0.6265965559486724, + "learning_rate": 8.879407230450373e-06, + "loss": 0.4927, + "step": 10939 + }, + { + "epoch": 0.2408100374202069, + "grad_norm": 1.6308327149719146, + "learning_rate": 8.879182339841506e-06, + "loss": 0.4376, + "step": 10940 + }, + { + "epoch": 0.24083204930662558, + "grad_norm": 3.7636903251207423, + "learning_rate": 8.878957429516904e-06, + "loss": 0.323, + "step": 10941 + }, + { + "epoch": 0.24085406119304426, + "grad_norm": 1.9013423073738844, + "learning_rate": 8.878732499477712e-06, + "loss": 0.3073, + "step": 10942 + }, + { + "epoch": 0.2408760730794629, + "grad_norm": 1.2620941836739803, + "learning_rate": 8.878507549725071e-06, + "loss": 0.333, + "step": 10943 + }, + { + "epoch": 0.24089808496588158, + "grad_norm": 0.606739011724734, + "learning_rate": 8.878282580260124e-06, + "loss": 0.5243, + "step": 10944 + }, + { + "epoch": 0.24092009685230023, + "grad_norm": 0.6648412288637203, + "learning_rate": 8.878057591084017e-06, + "loss": 0.539, + "step": 10945 + }, + { + "epoch": 0.2409421087387189, + "grad_norm": 1.1761115981041728, + "learning_rate": 8.877832582197889e-06, + "loss": 0.2844, + "step": 10946 + }, + { + "epoch": 0.24096412062513758, + "grad_norm": 1.3592594450135211, + "learning_rate": 8.87760755360289e-06, + "loss": 0.2821, + "step": 10947 + }, + { + "epoch": 0.24098613251155623, + "grad_norm": 2.1611115821074742, + "learning_rate": 8.877382505300156e-06, + "loss": 0.3513, + "step": 10948 + }, + { + "epoch": 0.2410081443979749, + "grad_norm": 1.450203485902474, + "learning_rate": 8.877157437290838e-06, + "loss": 0.4147, + "step": 10949 + }, + { + "epoch": 0.24103015628439356, + "grad_norm": 1.4718499287085849, + "learning_rate": 8.876932349576075e-06, + "loss": 0.4008, + "step": 10950 + }, + { + "epoch": 0.24105216817081224, + "grad_norm": 0.6500603388364035, + "learning_rate": 8.876707242157013e-06, + "loss": 0.5502, + "step": 10951 + }, + { + "epoch": 0.2410741800572309, + "grad_norm": 1.5159495967338854, + "learning_rate": 8.876482115034795e-06, + "loss": 0.4152, + "step": 10952 + }, + { + "epoch": 0.24109619194364956, + "grad_norm": 1.2056205293378206, + "learning_rate": 8.876256968210566e-06, + "loss": 0.2869, + "step": 10953 + }, + { + "epoch": 0.24111820383006824, + "grad_norm": 1.3556604872273548, + "learning_rate": 8.87603180168547e-06, + "loss": 0.337, + "step": 10954 + }, + { + "epoch": 0.24114021571648692, + "grad_norm": 1.6567535533578157, + "learning_rate": 8.875806615460654e-06, + "loss": 0.3376, + "step": 10955 + }, + { + "epoch": 0.24116222760290557, + "grad_norm": 1.4123745105752636, + "learning_rate": 8.875581409537257e-06, + "loss": 0.3889, + "step": 10956 + }, + { + "epoch": 0.24118423948932424, + "grad_norm": 0.6558826515577855, + "learning_rate": 8.875356183916428e-06, + "loss": 0.5176, + "step": 10957 + }, + { + "epoch": 0.2412062513757429, + "grad_norm": 1.3345933152037668, + "learning_rate": 8.87513093859931e-06, + "loss": 0.3118, + "step": 10958 + }, + { + "epoch": 0.24122826326216157, + "grad_norm": 0.6326958729603696, + "learning_rate": 8.87490567358705e-06, + "loss": 0.5504, + "step": 10959 + }, + { + "epoch": 0.24125027514858025, + "grad_norm": 0.594645306028328, + "learning_rate": 8.87468038888079e-06, + "loss": 0.4767, + "step": 10960 + }, + { + "epoch": 0.2412722870349989, + "grad_norm": 1.4432679798784074, + "learning_rate": 8.874455084481675e-06, + "loss": 0.3541, + "step": 10961 + }, + { + "epoch": 0.24129429892141757, + "grad_norm": 1.4060540084612208, + "learning_rate": 8.874229760390853e-06, + "loss": 0.3103, + "step": 10962 + }, + { + "epoch": 0.24131631080783622, + "grad_norm": 1.340037786159879, + "learning_rate": 8.874004416609466e-06, + "loss": 0.3933, + "step": 10963 + }, + { + "epoch": 0.2413383226942549, + "grad_norm": 1.7933658027997392, + "learning_rate": 8.873779053138663e-06, + "loss": 0.4233, + "step": 10964 + }, + { + "epoch": 0.24136033458067357, + "grad_norm": 3.505490474093115, + "learning_rate": 8.873553669979584e-06, + "loss": 0.2997, + "step": 10965 + }, + { + "epoch": 0.24138234646709222, + "grad_norm": 1.325201723364471, + "learning_rate": 8.873328267133378e-06, + "loss": 0.3432, + "step": 10966 + }, + { + "epoch": 0.2414043583535109, + "grad_norm": 1.554186501497474, + "learning_rate": 8.87310284460119e-06, + "loss": 0.3749, + "step": 10967 + }, + { + "epoch": 0.24142637023992955, + "grad_norm": 1.69588689888807, + "learning_rate": 8.872877402384169e-06, + "loss": 0.3617, + "step": 10968 + }, + { + "epoch": 0.24144838212634823, + "grad_norm": 1.3870002425230434, + "learning_rate": 8.872651940483456e-06, + "loss": 0.3726, + "step": 10969 + }, + { + "epoch": 0.2414703940127669, + "grad_norm": 1.2131357497984627, + "learning_rate": 8.872426458900199e-06, + "loss": 0.3267, + "step": 10970 + }, + { + "epoch": 0.24149240589918555, + "grad_norm": 1.885167278647154, + "learning_rate": 8.872200957635541e-06, + "loss": 0.4231, + "step": 10971 + }, + { + "epoch": 0.24151441778560423, + "grad_norm": 0.7651028398994659, + "learning_rate": 8.871975436690634e-06, + "loss": 0.5156, + "step": 10972 + }, + { + "epoch": 0.24153642967202288, + "grad_norm": 1.282443958357992, + "learning_rate": 8.87174989606662e-06, + "loss": 0.3566, + "step": 10973 + }, + { + "epoch": 0.24155844155844156, + "grad_norm": 1.48010523094787, + "learning_rate": 8.871524335764645e-06, + "loss": 0.3221, + "step": 10974 + }, + { + "epoch": 0.24158045344486023, + "grad_norm": 1.1200170948565562, + "learning_rate": 8.871298755785858e-06, + "loss": 0.2637, + "step": 10975 + }, + { + "epoch": 0.24160246533127888, + "grad_norm": 2.1659012066519736, + "learning_rate": 8.871073156131403e-06, + "loss": 0.293, + "step": 10976 + }, + { + "epoch": 0.24162447721769756, + "grad_norm": 0.6503452851161547, + "learning_rate": 8.870847536802428e-06, + "loss": 0.5518, + "step": 10977 + }, + { + "epoch": 0.24164648910411624, + "grad_norm": 1.4286267338277308, + "learning_rate": 8.870621897800079e-06, + "loss": 0.3264, + "step": 10978 + }, + { + "epoch": 0.24166850099053488, + "grad_norm": 1.5784290464746515, + "learning_rate": 8.870396239125503e-06, + "loss": 0.3374, + "step": 10979 + }, + { + "epoch": 0.24169051287695356, + "grad_norm": 0.6334191088450732, + "learning_rate": 8.870170560779846e-06, + "loss": 0.516, + "step": 10980 + }, + { + "epoch": 0.2417125247633722, + "grad_norm": 1.1279307567850108, + "learning_rate": 8.86994486276426e-06, + "loss": 0.3912, + "step": 10981 + }, + { + "epoch": 0.2417345366497909, + "grad_norm": 1.4040179141819402, + "learning_rate": 8.869719145079884e-06, + "loss": 0.3942, + "step": 10982 + }, + { + "epoch": 0.24175654853620956, + "grad_norm": 1.4930770759263678, + "learning_rate": 8.869493407727872e-06, + "loss": 0.4402, + "step": 10983 + }, + { + "epoch": 0.2417785604226282, + "grad_norm": 1.3721687643213332, + "learning_rate": 8.869267650709368e-06, + "loss": 0.3542, + "step": 10984 + }, + { + "epoch": 0.2418005723090469, + "grad_norm": 2.6112375557396903, + "learning_rate": 8.86904187402552e-06, + "loss": 0.2954, + "step": 10985 + }, + { + "epoch": 0.24182258419546554, + "grad_norm": 2.3628610535629284, + "learning_rate": 8.868816077677476e-06, + "loss": 0.3832, + "step": 10986 + }, + { + "epoch": 0.24184459608188422, + "grad_norm": 0.6564804057955697, + "learning_rate": 8.868590261666383e-06, + "loss": 0.4966, + "step": 10987 + }, + { + "epoch": 0.2418666079683029, + "grad_norm": 0.646675151348874, + "learning_rate": 8.868364425993387e-06, + "loss": 0.5198, + "step": 10988 + }, + { + "epoch": 0.24188861985472154, + "grad_norm": 2.516663411322828, + "learning_rate": 8.86813857065964e-06, + "loss": 0.3366, + "step": 10989 + }, + { + "epoch": 0.24191063174114022, + "grad_norm": 1.4732341322642504, + "learning_rate": 8.867912695666287e-06, + "loss": 0.3717, + "step": 10990 + }, + { + "epoch": 0.24193264362755887, + "grad_norm": 1.5269698075802582, + "learning_rate": 8.867686801014475e-06, + "loss": 0.3953, + "step": 10991 + }, + { + "epoch": 0.24195465551397755, + "grad_norm": 0.6419081826400239, + "learning_rate": 8.867460886705357e-06, + "loss": 0.5228, + "step": 10992 + }, + { + "epoch": 0.24197666740039622, + "grad_norm": 0.6296960563454905, + "learning_rate": 8.867234952740074e-06, + "loss": 0.5442, + "step": 10993 + }, + { + "epoch": 0.24199867928681487, + "grad_norm": 1.5532680339258145, + "learning_rate": 8.86700899911978e-06, + "loss": 0.3996, + "step": 10994 + }, + { + "epoch": 0.24202069117323355, + "grad_norm": 0.6243291497162826, + "learning_rate": 8.866783025845623e-06, + "loss": 0.4961, + "step": 10995 + }, + { + "epoch": 0.24204270305965223, + "grad_norm": 1.300747633055381, + "learning_rate": 8.866557032918748e-06, + "loss": 0.3529, + "step": 10996 + }, + { + "epoch": 0.24206471494607087, + "grad_norm": 1.5038544515218304, + "learning_rate": 8.866331020340305e-06, + "loss": 0.3152, + "step": 10997 + }, + { + "epoch": 0.24208672683248955, + "grad_norm": 1.2506957761324242, + "learning_rate": 8.866104988111445e-06, + "loss": 0.2886, + "step": 10998 + }, + { + "epoch": 0.2421087387189082, + "grad_norm": 1.266036713010963, + "learning_rate": 8.865878936233313e-06, + "loss": 0.2968, + "step": 10999 + }, + { + "epoch": 0.24213075060532688, + "grad_norm": 1.4713630270796076, + "learning_rate": 8.865652864707061e-06, + "loss": 0.3629, + "step": 11000 + }, + { + "epoch": 0.24215276249174555, + "grad_norm": 1.2506273087700974, + "learning_rate": 8.865426773533838e-06, + "loss": 0.2722, + "step": 11001 + }, + { + "epoch": 0.2421747743781642, + "grad_norm": 1.288360943261788, + "learning_rate": 8.865200662714793e-06, + "loss": 0.2991, + "step": 11002 + }, + { + "epoch": 0.24219678626458288, + "grad_norm": 6.4717939477454225, + "learning_rate": 8.864974532251072e-06, + "loss": 0.3456, + "step": 11003 + }, + { + "epoch": 0.24221879815100153, + "grad_norm": 1.2690559317870673, + "learning_rate": 8.864748382143828e-06, + "loss": 0.395, + "step": 11004 + }, + { + "epoch": 0.2422408100374202, + "grad_norm": 1.3546996777774973, + "learning_rate": 8.86452221239421e-06, + "loss": 0.3558, + "step": 11005 + }, + { + "epoch": 0.24226282192383888, + "grad_norm": 1.3402490464436696, + "learning_rate": 8.864296023003363e-06, + "loss": 0.3687, + "step": 11006 + }, + { + "epoch": 0.24228483381025753, + "grad_norm": 2.049571371104541, + "learning_rate": 8.864069813972442e-06, + "loss": 0.3797, + "step": 11007 + }, + { + "epoch": 0.2423068456966762, + "grad_norm": 1.3386873780569795, + "learning_rate": 8.863843585302596e-06, + "loss": 0.2563, + "step": 11008 + }, + { + "epoch": 0.24232885758309486, + "grad_norm": 0.6899628635356074, + "learning_rate": 8.863617336994973e-06, + "loss": 0.5429, + "step": 11009 + }, + { + "epoch": 0.24235086946951354, + "grad_norm": 1.2235689419299907, + "learning_rate": 8.863391069050721e-06, + "loss": 0.2943, + "step": 11010 + }, + { + "epoch": 0.2423728813559322, + "grad_norm": 0.6542175291978591, + "learning_rate": 8.863164781470995e-06, + "loss": 0.5208, + "step": 11011 + }, + { + "epoch": 0.24239489324235086, + "grad_norm": 1.2905007773472668, + "learning_rate": 8.86293847425694e-06, + "loss": 0.3348, + "step": 11012 + }, + { + "epoch": 0.24241690512876954, + "grad_norm": 1.2798022951299655, + "learning_rate": 8.862712147409713e-06, + "loss": 0.3193, + "step": 11013 + }, + { + "epoch": 0.2424389170151882, + "grad_norm": 1.2816685299927992, + "learning_rate": 8.862485800930458e-06, + "loss": 0.4416, + "step": 11014 + }, + { + "epoch": 0.24246092890160686, + "grad_norm": 1.1358797142464536, + "learning_rate": 8.862259434820326e-06, + "loss": 0.3338, + "step": 11015 + }, + { + "epoch": 0.24248294078802554, + "grad_norm": 1.5239963712424918, + "learning_rate": 8.86203304908047e-06, + "loss": 0.3675, + "step": 11016 + }, + { + "epoch": 0.2425049526744442, + "grad_norm": 1.7219916595113125, + "learning_rate": 8.861806643712041e-06, + "loss": 0.3847, + "step": 11017 + }, + { + "epoch": 0.24252696456086287, + "grad_norm": 1.3945163249882024, + "learning_rate": 8.861580218716186e-06, + "loss": 0.315, + "step": 11018 + }, + { + "epoch": 0.24254897644728154, + "grad_norm": 1.1020703432049566, + "learning_rate": 8.861353774094059e-06, + "loss": 0.2519, + "step": 11019 + }, + { + "epoch": 0.2425709883337002, + "grad_norm": 1.5207519939939886, + "learning_rate": 8.86112730984681e-06, + "loss": 0.3877, + "step": 11020 + }, + { + "epoch": 0.24259300022011887, + "grad_norm": 1.2013217224223658, + "learning_rate": 8.86090082597559e-06, + "loss": 0.2383, + "step": 11021 + }, + { + "epoch": 0.24261501210653752, + "grad_norm": 1.3222648504508487, + "learning_rate": 8.860674322481549e-06, + "loss": 0.3604, + "step": 11022 + }, + { + "epoch": 0.2426370239929562, + "grad_norm": 1.7099796209473386, + "learning_rate": 8.86044779936584e-06, + "loss": 0.5112, + "step": 11023 + }, + { + "epoch": 0.24265903587937487, + "grad_norm": 1.2298413104671955, + "learning_rate": 8.860221256629613e-06, + "loss": 0.3025, + "step": 11024 + }, + { + "epoch": 0.24268104776579352, + "grad_norm": 1.5553051302994545, + "learning_rate": 8.85999469427402e-06, + "loss": 0.3238, + "step": 11025 + }, + { + "epoch": 0.2427030596522122, + "grad_norm": 1.5792199955062307, + "learning_rate": 8.859768112300213e-06, + "loss": 0.3401, + "step": 11026 + }, + { + "epoch": 0.24272507153863085, + "grad_norm": 1.6127872622956894, + "learning_rate": 8.859541510709341e-06, + "loss": 0.344, + "step": 11027 + }, + { + "epoch": 0.24274708342504953, + "grad_norm": 1.3267688153168533, + "learning_rate": 8.859314889502561e-06, + "loss": 0.3452, + "step": 11028 + }, + { + "epoch": 0.2427690953114682, + "grad_norm": 1.7781392431200613, + "learning_rate": 8.85908824868102e-06, + "loss": 0.3739, + "step": 11029 + }, + { + "epoch": 0.24279110719788685, + "grad_norm": 2.609130271563595, + "learning_rate": 8.85886158824587e-06, + "loss": 0.3526, + "step": 11030 + }, + { + "epoch": 0.24281311908430553, + "grad_norm": 2.350129522006211, + "learning_rate": 8.858634908198265e-06, + "loss": 0.3225, + "step": 11031 + }, + { + "epoch": 0.24283513097072418, + "grad_norm": 1.3333889976515951, + "learning_rate": 8.858408208539357e-06, + "loss": 0.3634, + "step": 11032 + }, + { + "epoch": 0.24285714285714285, + "grad_norm": 1.4565977946997761, + "learning_rate": 8.858181489270297e-06, + "loss": 0.317, + "step": 11033 + }, + { + "epoch": 0.24287915474356153, + "grad_norm": 1.4381466809565628, + "learning_rate": 8.857954750392239e-06, + "loss": 0.3228, + "step": 11034 + }, + { + "epoch": 0.24290116662998018, + "grad_norm": 1.2440330610396928, + "learning_rate": 8.857727991906332e-06, + "loss": 0.264, + "step": 11035 + }, + { + "epoch": 0.24292317851639886, + "grad_norm": 1.7440520286614758, + "learning_rate": 8.857501213813732e-06, + "loss": 0.4261, + "step": 11036 + }, + { + "epoch": 0.24294519040281753, + "grad_norm": 1.1691687727583469, + "learning_rate": 8.85727441611559e-06, + "loss": 0.2562, + "step": 11037 + }, + { + "epoch": 0.24296720228923618, + "grad_norm": 1.2194846546537021, + "learning_rate": 8.85704759881306e-06, + "loss": 0.4163, + "step": 11038 + }, + { + "epoch": 0.24298921417565486, + "grad_norm": 1.994934865609955, + "learning_rate": 8.856820761907294e-06, + "loss": 0.2702, + "step": 11039 + }, + { + "epoch": 0.2430112260620735, + "grad_norm": 1.2748962876046996, + "learning_rate": 8.856593905399445e-06, + "loss": 0.3006, + "step": 11040 + }, + { + "epoch": 0.2430332379484922, + "grad_norm": 1.2664442941633551, + "learning_rate": 8.856367029290664e-06, + "loss": 0.2782, + "step": 11041 + }, + { + "epoch": 0.24305524983491086, + "grad_norm": 1.7846330787778157, + "learning_rate": 8.856140133582104e-06, + "loss": 0.3746, + "step": 11042 + }, + { + "epoch": 0.2430772617213295, + "grad_norm": 1.2613210291989232, + "learning_rate": 8.855913218274924e-06, + "loss": 0.3305, + "step": 11043 + }, + { + "epoch": 0.2430992736077482, + "grad_norm": 1.3932995224514084, + "learning_rate": 8.855686283370272e-06, + "loss": 0.266, + "step": 11044 + }, + { + "epoch": 0.24312128549416684, + "grad_norm": 0.7629197201779456, + "learning_rate": 8.855459328869303e-06, + "loss": 0.5305, + "step": 11045 + }, + { + "epoch": 0.24314329738058552, + "grad_norm": 1.274805714655843, + "learning_rate": 8.855232354773169e-06, + "loss": 0.3413, + "step": 11046 + }, + { + "epoch": 0.2431653092670042, + "grad_norm": 1.5002967294784533, + "learning_rate": 8.855005361083025e-06, + "loss": 0.4164, + "step": 11047 + }, + { + "epoch": 0.24318732115342284, + "grad_norm": 1.1846043949370295, + "learning_rate": 8.854778347800024e-06, + "loss": 0.3038, + "step": 11048 + }, + { + "epoch": 0.24320933303984152, + "grad_norm": 0.6260261726871221, + "learning_rate": 8.854551314925321e-06, + "loss": 0.5019, + "step": 11049 + }, + { + "epoch": 0.24323134492626017, + "grad_norm": 1.459926070554808, + "learning_rate": 8.854324262460067e-06, + "loss": 0.3374, + "step": 11050 + }, + { + "epoch": 0.24325335681267884, + "grad_norm": 1.4868960609908997, + "learning_rate": 8.854097190405421e-06, + "loss": 0.3092, + "step": 11051 + }, + { + "epoch": 0.24327536869909752, + "grad_norm": 1.809311542444734, + "learning_rate": 8.853870098762532e-06, + "loss": 0.369, + "step": 11052 + }, + { + "epoch": 0.24329738058551617, + "grad_norm": 1.4856664475439716, + "learning_rate": 8.853642987532558e-06, + "loss": 0.3323, + "step": 11053 + }, + { + "epoch": 0.24331939247193485, + "grad_norm": 1.5225976207589926, + "learning_rate": 8.85341585671665e-06, + "loss": 0.3121, + "step": 11054 + }, + { + "epoch": 0.24334140435835352, + "grad_norm": 1.421960320820224, + "learning_rate": 8.853188706315964e-06, + "loss": 0.4095, + "step": 11055 + }, + { + "epoch": 0.24336341624477217, + "grad_norm": 1.4796597704066001, + "learning_rate": 8.852961536331655e-06, + "loss": 0.3023, + "step": 11056 + }, + { + "epoch": 0.24338542813119085, + "grad_norm": 1.2654186379449495, + "learning_rate": 8.852734346764877e-06, + "loss": 0.3643, + "step": 11057 + }, + { + "epoch": 0.2434074400176095, + "grad_norm": 1.646880244997173, + "learning_rate": 8.852507137616785e-06, + "loss": 0.414, + "step": 11058 + }, + { + "epoch": 0.24342945190402818, + "grad_norm": 2.614662235754809, + "learning_rate": 8.852279908888532e-06, + "loss": 0.3739, + "step": 11059 + }, + { + "epoch": 0.24345146379044685, + "grad_norm": 0.7447707041624351, + "learning_rate": 8.852052660581276e-06, + "loss": 0.518, + "step": 11060 + }, + { + "epoch": 0.2434734756768655, + "grad_norm": 1.5078003007617304, + "learning_rate": 8.851825392696169e-06, + "loss": 0.2574, + "step": 11061 + }, + { + "epoch": 0.24349548756328418, + "grad_norm": 1.811493392654241, + "learning_rate": 8.851598105234368e-06, + "loss": 0.4122, + "step": 11062 + }, + { + "epoch": 0.24351749944970283, + "grad_norm": 1.4302345324672678, + "learning_rate": 8.85137079819703e-06, + "loss": 0.3602, + "step": 11063 + }, + { + "epoch": 0.2435395113361215, + "grad_norm": 0.6423376709782541, + "learning_rate": 8.851143471585304e-06, + "loss": 0.5165, + "step": 11064 + }, + { + "epoch": 0.24356152322254018, + "grad_norm": 5.064827782722244, + "learning_rate": 8.85091612540035e-06, + "loss": 0.3688, + "step": 11065 + }, + { + "epoch": 0.24358353510895883, + "grad_norm": 0.6491787225899964, + "learning_rate": 8.850688759643325e-06, + "loss": 0.5181, + "step": 11066 + }, + { + "epoch": 0.2436055469953775, + "grad_norm": 2.0374514602922953, + "learning_rate": 8.85046137431538e-06, + "loss": 0.3603, + "step": 11067 + }, + { + "epoch": 0.24362755888179616, + "grad_norm": 1.340219184931333, + "learning_rate": 8.850233969417674e-06, + "loss": 0.3307, + "step": 11068 + }, + { + "epoch": 0.24364957076821483, + "grad_norm": 1.6182655305025015, + "learning_rate": 8.850006544951362e-06, + "loss": 0.3341, + "step": 11069 + }, + { + "epoch": 0.2436715826546335, + "grad_norm": 1.5737665849501634, + "learning_rate": 8.8497791009176e-06, + "loss": 0.3314, + "step": 11070 + }, + { + "epoch": 0.24369359454105216, + "grad_norm": 1.4959792262969343, + "learning_rate": 8.849551637317542e-06, + "loss": 0.3261, + "step": 11071 + }, + { + "epoch": 0.24371560642747084, + "grad_norm": 0.6996342670239942, + "learning_rate": 8.849324154152345e-06, + "loss": 0.5211, + "step": 11072 + }, + { + "epoch": 0.2437376183138895, + "grad_norm": 1.4064794871302073, + "learning_rate": 8.849096651423169e-06, + "loss": 0.357, + "step": 11073 + }, + { + "epoch": 0.24375963020030816, + "grad_norm": 0.653329133188625, + "learning_rate": 8.848869129131163e-06, + "loss": 0.5501, + "step": 11074 + }, + { + "epoch": 0.24378164208672684, + "grad_norm": 1.2234095611789035, + "learning_rate": 8.84864158727749e-06, + "loss": 0.345, + "step": 11075 + }, + { + "epoch": 0.2438036539731455, + "grad_norm": 3.1234304777498894, + "learning_rate": 8.848414025863304e-06, + "loss": 0.3824, + "step": 11076 + }, + { + "epoch": 0.24382566585956417, + "grad_norm": 1.405820107772932, + "learning_rate": 8.848186444889759e-06, + "loss": 0.3646, + "step": 11077 + }, + { + "epoch": 0.24384767774598284, + "grad_norm": 1.977383808229968, + "learning_rate": 8.847958844358018e-06, + "loss": 0.3142, + "step": 11078 + }, + { + "epoch": 0.2438696896324015, + "grad_norm": 1.4682236650476257, + "learning_rate": 8.847731224269232e-06, + "loss": 0.3268, + "step": 11079 + }, + { + "epoch": 0.24389170151882017, + "grad_norm": 1.6171029434608566, + "learning_rate": 8.847503584624559e-06, + "loss": 0.3588, + "step": 11080 + }, + { + "epoch": 0.24391371340523882, + "grad_norm": 1.345580967489453, + "learning_rate": 8.847275925425158e-06, + "loss": 0.358, + "step": 11081 + }, + { + "epoch": 0.2439357252916575, + "grad_norm": 1.4151906374391736, + "learning_rate": 8.847048246672183e-06, + "loss": 0.3803, + "step": 11082 + }, + { + "epoch": 0.24395773717807617, + "grad_norm": 0.74210426992169, + "learning_rate": 8.846820548366795e-06, + "loss": 0.5454, + "step": 11083 + }, + { + "epoch": 0.24397974906449482, + "grad_norm": 0.6869412249953144, + "learning_rate": 8.846592830510147e-06, + "loss": 0.4934, + "step": 11084 + }, + { + "epoch": 0.2440017609509135, + "grad_norm": 0.6059444953263676, + "learning_rate": 8.846365093103402e-06, + "loss": 0.5198, + "step": 11085 + }, + { + "epoch": 0.24402377283733215, + "grad_norm": 0.6022913116847496, + "learning_rate": 8.846137336147711e-06, + "loss": 0.5133, + "step": 11086 + }, + { + "epoch": 0.24404578472375082, + "grad_norm": 1.709327793547513, + "learning_rate": 8.845909559644237e-06, + "loss": 0.3668, + "step": 11087 + }, + { + "epoch": 0.2440677966101695, + "grad_norm": 4.464617336668517, + "learning_rate": 8.845681763594134e-06, + "loss": 0.367, + "step": 11088 + }, + { + "epoch": 0.24408980849658815, + "grad_norm": 1.4152015318426268, + "learning_rate": 8.845453947998561e-06, + "loss": 0.3478, + "step": 11089 + }, + { + "epoch": 0.24411182038300683, + "grad_norm": 1.7389593520634494, + "learning_rate": 8.845226112858676e-06, + "loss": 0.4133, + "step": 11090 + }, + { + "epoch": 0.24413383226942548, + "grad_norm": 7.487021483564886, + "learning_rate": 8.844998258175636e-06, + "loss": 0.3545, + "step": 11091 + }, + { + "epoch": 0.24415584415584415, + "grad_norm": 1.0008429432352055, + "learning_rate": 8.8447703839506e-06, + "loss": 0.5296, + "step": 11092 + }, + { + "epoch": 0.24417785604226283, + "grad_norm": 0.8456976390783195, + "learning_rate": 8.844542490184729e-06, + "loss": 0.5199, + "step": 11093 + }, + { + "epoch": 0.24419986792868148, + "grad_norm": 1.4922739732180232, + "learning_rate": 8.844314576879176e-06, + "loss": 0.3866, + "step": 11094 + }, + { + "epoch": 0.24422187981510016, + "grad_norm": 1.2907392365724306, + "learning_rate": 8.844086644035102e-06, + "loss": 0.2863, + "step": 11095 + }, + { + "epoch": 0.24424389170151883, + "grad_norm": 1.3247953123426093, + "learning_rate": 8.843858691653666e-06, + "loss": 0.3696, + "step": 11096 + }, + { + "epoch": 0.24426590358793748, + "grad_norm": 1.5589565025041876, + "learning_rate": 8.843630719736023e-06, + "loss": 0.3878, + "step": 11097 + }, + { + "epoch": 0.24428791547435616, + "grad_norm": 1.5400204412045602, + "learning_rate": 8.843402728283337e-06, + "loss": 0.4287, + "step": 11098 + }, + { + "epoch": 0.2443099273607748, + "grad_norm": 1.6517495305826568, + "learning_rate": 8.843174717296762e-06, + "loss": 0.4069, + "step": 11099 + }, + { + "epoch": 0.24433193924719349, + "grad_norm": 1.4495032914731323, + "learning_rate": 8.842946686777462e-06, + "loss": 0.3675, + "step": 11100 + }, + { + "epoch": 0.24435395113361216, + "grad_norm": 1.5526377658563528, + "learning_rate": 8.842718636726591e-06, + "loss": 0.2395, + "step": 11101 + }, + { + "epoch": 0.2443759630200308, + "grad_norm": 1.1115585754139796, + "learning_rate": 8.84249056714531e-06, + "loss": 0.5305, + "step": 11102 + }, + { + "epoch": 0.2443979749064495, + "grad_norm": 1.2401029260506071, + "learning_rate": 8.84226247803478e-06, + "loss": 0.3468, + "step": 11103 + }, + { + "epoch": 0.24441998679286814, + "grad_norm": 2.170113026243009, + "learning_rate": 8.842034369396155e-06, + "loss": 0.3603, + "step": 11104 + }, + { + "epoch": 0.24444199867928681, + "grad_norm": 1.4307905323927443, + "learning_rate": 8.841806241230601e-06, + "loss": 0.3404, + "step": 11105 + }, + { + "epoch": 0.2444640105657055, + "grad_norm": 1.8401191929194205, + "learning_rate": 8.841578093539273e-06, + "loss": 0.3851, + "step": 11106 + }, + { + "epoch": 0.24448602245212414, + "grad_norm": 0.7328528842001732, + "learning_rate": 8.84134992632333e-06, + "loss": 0.5023, + "step": 11107 + }, + { + "epoch": 0.24450803433854282, + "grad_norm": 1.5557636044550818, + "learning_rate": 8.841121739583936e-06, + "loss": 0.3721, + "step": 11108 + }, + { + "epoch": 0.24453004622496147, + "grad_norm": 1.375402275553965, + "learning_rate": 8.840893533322247e-06, + "loss": 0.3021, + "step": 11109 + }, + { + "epoch": 0.24455205811138014, + "grad_norm": 0.642635961165999, + "learning_rate": 8.840665307539424e-06, + "loss": 0.5154, + "step": 11110 + }, + { + "epoch": 0.24457406999779882, + "grad_norm": 3.2606471131926185, + "learning_rate": 8.840437062236627e-06, + "loss": 0.3422, + "step": 11111 + }, + { + "epoch": 0.24459608188421747, + "grad_norm": 1.9166832741763438, + "learning_rate": 8.840208797415015e-06, + "loss": 0.3981, + "step": 11112 + }, + { + "epoch": 0.24461809377063615, + "grad_norm": 2.6165268718118053, + "learning_rate": 8.83998051307575e-06, + "loss": 0.2361, + "step": 11113 + }, + { + "epoch": 0.2446401056570548, + "grad_norm": 1.6573411662752953, + "learning_rate": 8.839752209219992e-06, + "loss": 0.3809, + "step": 11114 + }, + { + "epoch": 0.24466211754347347, + "grad_norm": 1.64842051975773, + "learning_rate": 8.8395238858489e-06, + "loss": 0.3011, + "step": 11115 + }, + { + "epoch": 0.24468412942989215, + "grad_norm": 3.457486980577854, + "learning_rate": 8.839295542963635e-06, + "loss": 0.2949, + "step": 11116 + }, + { + "epoch": 0.2447061413163108, + "grad_norm": 1.2220696654831043, + "learning_rate": 8.839067180565359e-06, + "loss": 0.3821, + "step": 11117 + }, + { + "epoch": 0.24472815320272948, + "grad_norm": 1.9500421454009653, + "learning_rate": 8.83883879865523e-06, + "loss": 0.3528, + "step": 11118 + }, + { + "epoch": 0.24475016508914815, + "grad_norm": 1.397019886563965, + "learning_rate": 8.83861039723441e-06, + "loss": 0.3173, + "step": 11119 + }, + { + "epoch": 0.2447721769755668, + "grad_norm": 1.732606618652005, + "learning_rate": 8.838381976304061e-06, + "loss": 0.3785, + "step": 11120 + }, + { + "epoch": 0.24479418886198548, + "grad_norm": 1.4412231719659774, + "learning_rate": 8.83815353586534e-06, + "loss": 0.4494, + "step": 11121 + }, + { + "epoch": 0.24481620074840413, + "grad_norm": 1.4912541190747177, + "learning_rate": 8.837925075919413e-06, + "loss": 0.3399, + "step": 11122 + }, + { + "epoch": 0.2448382126348228, + "grad_norm": 1.5484194329142276, + "learning_rate": 8.83769659646744e-06, + "loss": 0.3224, + "step": 11123 + }, + { + "epoch": 0.24486022452124148, + "grad_norm": 1.0583166614291883, + "learning_rate": 8.837468097510579e-06, + "loss": 0.2696, + "step": 11124 + }, + { + "epoch": 0.24488223640766013, + "grad_norm": 0.8789013252434311, + "learning_rate": 8.837239579049996e-06, + "loss": 0.5556, + "step": 11125 + }, + { + "epoch": 0.2449042482940788, + "grad_norm": 1.2956490950131607, + "learning_rate": 8.837011041086848e-06, + "loss": 0.4334, + "step": 11126 + }, + { + "epoch": 0.24492626018049746, + "grad_norm": 1.232039463992923, + "learning_rate": 8.8367824836223e-06, + "loss": 0.3356, + "step": 11127 + }, + { + "epoch": 0.24494827206691613, + "grad_norm": 1.6855760407889009, + "learning_rate": 8.83655390665751e-06, + "loss": 0.4666, + "step": 11128 + }, + { + "epoch": 0.2449702839533348, + "grad_norm": 1.5872835127491614, + "learning_rate": 8.836325310193644e-06, + "loss": 0.3903, + "step": 11129 + }, + { + "epoch": 0.24499229583975346, + "grad_norm": 0.6450418877813149, + "learning_rate": 8.836096694231862e-06, + "loss": 0.5168, + "step": 11130 + }, + { + "epoch": 0.24501430772617214, + "grad_norm": 1.5399185754156623, + "learning_rate": 8.835868058773323e-06, + "loss": 0.4074, + "step": 11131 + }, + { + "epoch": 0.24503631961259079, + "grad_norm": 1.3460923826336355, + "learning_rate": 8.835639403819194e-06, + "loss": 0.2919, + "step": 11132 + }, + { + "epoch": 0.24505833149900946, + "grad_norm": 1.8579842557097734, + "learning_rate": 8.835410729370634e-06, + "loss": 0.4208, + "step": 11133 + }, + { + "epoch": 0.24508034338542814, + "grad_norm": 1.2281410383577838, + "learning_rate": 8.835182035428807e-06, + "loss": 0.3146, + "step": 11134 + }, + { + "epoch": 0.2451023552718468, + "grad_norm": 1.6465575245758293, + "learning_rate": 8.834953321994873e-06, + "loss": 0.3441, + "step": 11135 + }, + { + "epoch": 0.24512436715826547, + "grad_norm": 1.4319612375069046, + "learning_rate": 8.834724589069996e-06, + "loss": 0.4331, + "step": 11136 + }, + { + "epoch": 0.24514637904468414, + "grad_norm": 0.706193551392818, + "learning_rate": 8.834495836655338e-06, + "loss": 0.5369, + "step": 11137 + }, + { + "epoch": 0.2451683909311028, + "grad_norm": 1.604723198386076, + "learning_rate": 8.834267064752063e-06, + "loss": 0.3517, + "step": 11138 + }, + { + "epoch": 0.24519040281752147, + "grad_norm": 3.48125855443837, + "learning_rate": 8.834038273361331e-06, + "loss": 0.3303, + "step": 11139 + }, + { + "epoch": 0.24521241470394012, + "grad_norm": 1.404822926178535, + "learning_rate": 8.833809462484309e-06, + "loss": 0.3452, + "step": 11140 + }, + { + "epoch": 0.2452344265903588, + "grad_norm": 1.754323743509034, + "learning_rate": 8.833580632122155e-06, + "loss": 0.3859, + "step": 11141 + }, + { + "epoch": 0.24525643847677747, + "grad_norm": 1.444598701113704, + "learning_rate": 8.833351782276037e-06, + "loss": 0.3503, + "step": 11142 + }, + { + "epoch": 0.24527845036319612, + "grad_norm": 1.8432232657806285, + "learning_rate": 8.833122912947114e-06, + "loss": 0.3611, + "step": 11143 + }, + { + "epoch": 0.2453004622496148, + "grad_norm": 2.822497454429182, + "learning_rate": 8.83289402413655e-06, + "loss": 0.3306, + "step": 11144 + }, + { + "epoch": 0.24532247413603345, + "grad_norm": 1.5621365336086315, + "learning_rate": 8.832665115845508e-06, + "loss": 0.3929, + "step": 11145 + }, + { + "epoch": 0.24534448602245212, + "grad_norm": 1.7924258560868676, + "learning_rate": 8.832436188075155e-06, + "loss": 0.4283, + "step": 11146 + }, + { + "epoch": 0.2453664979088708, + "grad_norm": 1.0729928647324165, + "learning_rate": 8.832207240826652e-06, + "loss": 0.2845, + "step": 11147 + }, + { + "epoch": 0.24538850979528945, + "grad_norm": 1.8274316369595682, + "learning_rate": 8.831978274101162e-06, + "loss": 0.3645, + "step": 11148 + }, + { + "epoch": 0.24541052168170813, + "grad_norm": 1.1612846172261808, + "learning_rate": 8.831749287899849e-06, + "loss": 0.3473, + "step": 11149 + }, + { + "epoch": 0.24543253356812678, + "grad_norm": 1.3103617952645927, + "learning_rate": 8.831520282223876e-06, + "loss": 0.3492, + "step": 11150 + }, + { + "epoch": 0.24545454545454545, + "grad_norm": 1.7844118476899185, + "learning_rate": 8.83129125707441e-06, + "loss": 0.297, + "step": 11151 + }, + { + "epoch": 0.24547655734096413, + "grad_norm": 2.4056324231172215, + "learning_rate": 8.831062212452613e-06, + "loss": 0.3497, + "step": 11152 + }, + { + "epoch": 0.24549856922738278, + "grad_norm": 1.5035576764160496, + "learning_rate": 8.830833148359647e-06, + "loss": 0.4268, + "step": 11153 + }, + { + "epoch": 0.24552058111380146, + "grad_norm": 1.4766629860272156, + "learning_rate": 8.83060406479668e-06, + "loss": 0.3605, + "step": 11154 + }, + { + "epoch": 0.2455425930002201, + "grad_norm": 1.943835324349351, + "learning_rate": 8.830374961764874e-06, + "loss": 0.3695, + "step": 11155 + }, + { + "epoch": 0.24556460488663878, + "grad_norm": 1.5244076883366013, + "learning_rate": 8.830145839265394e-06, + "loss": 0.4307, + "step": 11156 + }, + { + "epoch": 0.24558661677305746, + "grad_norm": 2.1349980897532315, + "learning_rate": 8.829916697299405e-06, + "loss": 0.3489, + "step": 11157 + }, + { + "epoch": 0.2456086286594761, + "grad_norm": 2.162429291808508, + "learning_rate": 8.829687535868073e-06, + "loss": 0.369, + "step": 11158 + }, + { + "epoch": 0.24563064054589479, + "grad_norm": 1.5627165644607308, + "learning_rate": 8.829458354972558e-06, + "loss": 0.2567, + "step": 11159 + }, + { + "epoch": 0.24565265243231346, + "grad_norm": 1.3923851593330792, + "learning_rate": 8.82922915461403e-06, + "loss": 0.3252, + "step": 11160 + }, + { + "epoch": 0.2456746643187321, + "grad_norm": 1.2458176662809786, + "learning_rate": 8.828999934793648e-06, + "loss": 0.3769, + "step": 11161 + }, + { + "epoch": 0.2456966762051508, + "grad_norm": 1.2883220203764514, + "learning_rate": 8.828770695512582e-06, + "loss": 0.4073, + "step": 11162 + }, + { + "epoch": 0.24571868809156944, + "grad_norm": 1.60741745198716, + "learning_rate": 8.828541436771998e-06, + "loss": 0.3337, + "step": 11163 + }, + { + "epoch": 0.24574069997798811, + "grad_norm": 1.2002387350449712, + "learning_rate": 8.828312158573058e-06, + "loss": 0.3522, + "step": 11164 + }, + { + "epoch": 0.2457627118644068, + "grad_norm": 1.4935776971967891, + "learning_rate": 8.828082860916927e-06, + "loss": 0.4319, + "step": 11165 + }, + { + "epoch": 0.24578472375082544, + "grad_norm": 2.265150199185351, + "learning_rate": 8.827853543804771e-06, + "loss": 0.3561, + "step": 11166 + }, + { + "epoch": 0.24580673563724412, + "grad_norm": 0.636750566231096, + "learning_rate": 8.827624207237756e-06, + "loss": 0.4957, + "step": 11167 + }, + { + "epoch": 0.24582874752366277, + "grad_norm": 1.801894352772622, + "learning_rate": 8.827394851217048e-06, + "loss": 0.4293, + "step": 11168 + }, + { + "epoch": 0.24585075941008144, + "grad_norm": 0.6369189452476852, + "learning_rate": 8.827165475743814e-06, + "loss": 0.5263, + "step": 11169 + }, + { + "epoch": 0.24587277129650012, + "grad_norm": 1.422424768934899, + "learning_rate": 8.826936080819217e-06, + "loss": 0.3403, + "step": 11170 + }, + { + "epoch": 0.24589478318291877, + "grad_norm": 1.4906542377364607, + "learning_rate": 8.826706666444425e-06, + "loss": 0.4105, + "step": 11171 + }, + { + "epoch": 0.24591679506933745, + "grad_norm": 1.5454081625446359, + "learning_rate": 8.8264772326206e-06, + "loss": 0.3407, + "step": 11172 + }, + { + "epoch": 0.2459388069557561, + "grad_norm": 1.7995420948961591, + "learning_rate": 8.826247779348913e-06, + "loss": 0.377, + "step": 11173 + }, + { + "epoch": 0.24596081884217477, + "grad_norm": 1.379326742794316, + "learning_rate": 8.826018306630529e-06, + "loss": 0.2509, + "step": 11174 + }, + { + "epoch": 0.24598283072859345, + "grad_norm": 1.3206322937899493, + "learning_rate": 8.82578881446661e-06, + "loss": 0.3562, + "step": 11175 + }, + { + "epoch": 0.2460048426150121, + "grad_norm": 1.632726087409948, + "learning_rate": 8.825559302858331e-06, + "loss": 0.3374, + "step": 11176 + }, + { + "epoch": 0.24602685450143078, + "grad_norm": 1.1868240245344468, + "learning_rate": 8.82532977180685e-06, + "loss": 0.3586, + "step": 11177 + }, + { + "epoch": 0.24604886638784945, + "grad_norm": 1.5810182146818688, + "learning_rate": 8.825100221313339e-06, + "loss": 0.4276, + "step": 11178 + }, + { + "epoch": 0.2460708782742681, + "grad_norm": 0.6529787582933385, + "learning_rate": 8.82487065137896e-06, + "loss": 0.5151, + "step": 11179 + }, + { + "epoch": 0.24609289016068678, + "grad_norm": 1.4079575249440326, + "learning_rate": 8.824641062004884e-06, + "loss": 0.3618, + "step": 11180 + }, + { + "epoch": 0.24611490204710543, + "grad_norm": 1.7150040206876411, + "learning_rate": 8.824411453192276e-06, + "loss": 0.4221, + "step": 11181 + }, + { + "epoch": 0.2461369139335241, + "grad_norm": 1.4464136932269054, + "learning_rate": 8.824181824942303e-06, + "loss": 0.3669, + "step": 11182 + }, + { + "epoch": 0.24615892581994278, + "grad_norm": 1.6015383364822025, + "learning_rate": 8.823952177256134e-06, + "loss": 0.3508, + "step": 11183 + }, + { + "epoch": 0.24618093770636143, + "grad_norm": 1.5460765704008106, + "learning_rate": 8.823722510134934e-06, + "loss": 0.397, + "step": 11184 + }, + { + "epoch": 0.2462029495927801, + "grad_norm": 1.8224116407906614, + "learning_rate": 8.82349282357987e-06, + "loss": 0.3819, + "step": 11185 + }, + { + "epoch": 0.24622496147919876, + "grad_norm": 1.6539054028669504, + "learning_rate": 8.82326311759211e-06, + "loss": 0.3998, + "step": 11186 + }, + { + "epoch": 0.24624697336561743, + "grad_norm": 7.015009067214053, + "learning_rate": 8.823033392172821e-06, + "loss": 0.3486, + "step": 11187 + }, + { + "epoch": 0.2462689852520361, + "grad_norm": 1.6525615259115534, + "learning_rate": 8.822803647323173e-06, + "loss": 0.4231, + "step": 11188 + }, + { + "epoch": 0.24629099713845476, + "grad_norm": 1.7296911226278486, + "learning_rate": 8.82257388304433e-06, + "loss": 0.3761, + "step": 11189 + }, + { + "epoch": 0.24631300902487344, + "grad_norm": 1.7197084577163233, + "learning_rate": 8.822344099337462e-06, + "loss": 0.3815, + "step": 11190 + }, + { + "epoch": 0.24633502091129209, + "grad_norm": 0.6565011009441805, + "learning_rate": 8.822114296203737e-06, + "loss": 0.5273, + "step": 11191 + }, + { + "epoch": 0.24635703279771076, + "grad_norm": 0.6425791072018079, + "learning_rate": 8.821884473644323e-06, + "loss": 0.5369, + "step": 11192 + }, + { + "epoch": 0.24637904468412944, + "grad_norm": 1.2052053372891278, + "learning_rate": 8.821654631660385e-06, + "loss": 0.373, + "step": 11193 + }, + { + "epoch": 0.2464010565705481, + "grad_norm": 2.144542087551035, + "learning_rate": 8.821424770253097e-06, + "loss": 0.303, + "step": 11194 + }, + { + "epoch": 0.24642306845696677, + "grad_norm": 1.1956385499220044, + "learning_rate": 8.821194889423624e-06, + "loss": 0.248, + "step": 11195 + }, + { + "epoch": 0.24644508034338544, + "grad_norm": 1.1647716451812768, + "learning_rate": 8.82096498917313e-06, + "loss": 0.2646, + "step": 11196 + }, + { + "epoch": 0.2464670922298041, + "grad_norm": 1.395135784870565, + "learning_rate": 8.820735069502791e-06, + "loss": 0.2584, + "step": 11197 + }, + { + "epoch": 0.24648910411622277, + "grad_norm": 1.6708747948890885, + "learning_rate": 8.820505130413772e-06, + "loss": 0.324, + "step": 11198 + }, + { + "epoch": 0.24651111600264142, + "grad_norm": 0.6171752411670239, + "learning_rate": 8.820275171907242e-06, + "loss": 0.5088, + "step": 11199 + }, + { + "epoch": 0.2465331278890601, + "grad_norm": 1.9536752305927383, + "learning_rate": 8.820045193984369e-06, + "loss": 0.2649, + "step": 11200 + }, + { + "epoch": 0.24655513977547877, + "grad_norm": 1.4754077779166799, + "learning_rate": 8.819815196646323e-06, + "loss": 0.3464, + "step": 11201 + }, + { + "epoch": 0.24657715166189742, + "grad_norm": 1.3062620122338764, + "learning_rate": 8.819585179894272e-06, + "loss": 0.2945, + "step": 11202 + }, + { + "epoch": 0.2465991635483161, + "grad_norm": 1.3835022313554386, + "learning_rate": 8.819355143729386e-06, + "loss": 0.3578, + "step": 11203 + }, + { + "epoch": 0.24662117543473475, + "grad_norm": 1.2885522063806725, + "learning_rate": 8.819125088152833e-06, + "loss": 0.3532, + "step": 11204 + }, + { + "epoch": 0.24664318732115342, + "grad_norm": 0.6370797952353344, + "learning_rate": 8.818895013165785e-06, + "loss": 0.5235, + "step": 11205 + }, + { + "epoch": 0.2466651992075721, + "grad_norm": 1.827288700757685, + "learning_rate": 8.818664918769407e-06, + "loss": 0.3165, + "step": 11206 + }, + { + "epoch": 0.24668721109399075, + "grad_norm": 1.3432470796749938, + "learning_rate": 8.818434804964871e-06, + "loss": 0.3783, + "step": 11207 + }, + { + "epoch": 0.24670922298040943, + "grad_norm": 1.3623348258170709, + "learning_rate": 8.818204671753347e-06, + "loss": 0.3734, + "step": 11208 + }, + { + "epoch": 0.24673123486682808, + "grad_norm": 1.3866753065739676, + "learning_rate": 8.817974519136004e-06, + "loss": 0.3182, + "step": 11209 + }, + { + "epoch": 0.24675324675324675, + "grad_norm": 1.3640644906284125, + "learning_rate": 8.81774434711401e-06, + "loss": 0.3497, + "step": 11210 + }, + { + "epoch": 0.24677525863966543, + "grad_norm": 0.6122244524889183, + "learning_rate": 8.817514155688538e-06, + "loss": 0.5163, + "step": 11211 + }, + { + "epoch": 0.24679727052608408, + "grad_norm": 0.62543824466142, + "learning_rate": 8.817283944860758e-06, + "loss": 0.5445, + "step": 11212 + }, + { + "epoch": 0.24681928241250276, + "grad_norm": 1.2405230056068417, + "learning_rate": 8.817053714631835e-06, + "loss": 0.2777, + "step": 11213 + }, + { + "epoch": 0.2468412942989214, + "grad_norm": 1.2451989652313609, + "learning_rate": 8.816823465002944e-06, + "loss": 0.2923, + "step": 11214 + }, + { + "epoch": 0.24686330618534008, + "grad_norm": 0.6166075023492162, + "learning_rate": 8.816593195975254e-06, + "loss": 0.5051, + "step": 11215 + }, + { + "epoch": 0.24688531807175876, + "grad_norm": 0.6268831346941928, + "learning_rate": 8.816362907549935e-06, + "loss": 0.5143, + "step": 11216 + }, + { + "epoch": 0.2469073299581774, + "grad_norm": 2.4694573875010217, + "learning_rate": 8.816132599728157e-06, + "loss": 0.3486, + "step": 11217 + }, + { + "epoch": 0.24692934184459608, + "grad_norm": 1.3293564090505388, + "learning_rate": 8.815902272511092e-06, + "loss": 0.384, + "step": 11218 + }, + { + "epoch": 0.24695135373101476, + "grad_norm": 1.5072372981818218, + "learning_rate": 8.815671925899909e-06, + "loss": 0.2916, + "step": 11219 + }, + { + "epoch": 0.2469733656174334, + "grad_norm": 4.335864431921401, + "learning_rate": 8.81544155989578e-06, + "loss": 0.3563, + "step": 11220 + }, + { + "epoch": 0.2469953775038521, + "grad_norm": 0.6624294392627881, + "learning_rate": 8.815211174499876e-06, + "loss": 0.5043, + "step": 11221 + }, + { + "epoch": 0.24701738939027074, + "grad_norm": 0.6204845131543083, + "learning_rate": 8.814980769713366e-06, + "loss": 0.4808, + "step": 11222 + }, + { + "epoch": 0.2470394012766894, + "grad_norm": 1.460512527663112, + "learning_rate": 8.814750345537422e-06, + "loss": 0.2498, + "step": 11223 + }, + { + "epoch": 0.2470614131631081, + "grad_norm": 1.8042835442142853, + "learning_rate": 8.814519901973216e-06, + "loss": 0.3124, + "step": 11224 + }, + { + "epoch": 0.24708342504952674, + "grad_norm": 1.4593338128478808, + "learning_rate": 8.814289439021918e-06, + "loss": 0.3437, + "step": 11225 + }, + { + "epoch": 0.24710543693594542, + "grad_norm": 1.8594011176897496, + "learning_rate": 8.8140589566847e-06, + "loss": 0.2921, + "step": 11226 + }, + { + "epoch": 0.24712744882236407, + "grad_norm": 1.3543385395734855, + "learning_rate": 8.813828454962735e-06, + "loss": 0.2919, + "step": 11227 + }, + { + "epoch": 0.24714946070878274, + "grad_norm": 0.6454147451026447, + "learning_rate": 8.81359793385719e-06, + "loss": 0.5119, + "step": 11228 + }, + { + "epoch": 0.24717147259520142, + "grad_norm": 1.3351530392729327, + "learning_rate": 8.81336739336924e-06, + "loss": 0.3705, + "step": 11229 + }, + { + "epoch": 0.24719348448162007, + "grad_norm": 1.2860248691509963, + "learning_rate": 8.813136833500058e-06, + "loss": 0.3292, + "step": 11230 + }, + { + "epoch": 0.24721549636803875, + "grad_norm": 1.5238664866953748, + "learning_rate": 8.812906254250813e-06, + "loss": 0.4219, + "step": 11231 + }, + { + "epoch": 0.2472375082544574, + "grad_norm": 1.2347313827453863, + "learning_rate": 8.812675655622676e-06, + "loss": 0.3625, + "step": 11232 + }, + { + "epoch": 0.24725952014087607, + "grad_norm": 0.6590280499016801, + "learning_rate": 8.812445037616823e-06, + "loss": 0.5054, + "step": 11233 + }, + { + "epoch": 0.24728153202729475, + "grad_norm": 1.4162774634564284, + "learning_rate": 8.812214400234423e-06, + "loss": 0.2817, + "step": 11234 + }, + { + "epoch": 0.2473035439137134, + "grad_norm": 1.4969090468664181, + "learning_rate": 8.811983743476649e-06, + "loss": 0.3191, + "step": 11235 + }, + { + "epoch": 0.24732555580013207, + "grad_norm": 1.5512662601720077, + "learning_rate": 8.811753067344673e-06, + "loss": 0.4112, + "step": 11236 + }, + { + "epoch": 0.24734756768655075, + "grad_norm": 0.6363730883402181, + "learning_rate": 8.81152237183967e-06, + "loss": 0.5335, + "step": 11237 + }, + { + "epoch": 0.2473695795729694, + "grad_norm": 0.6247103947151681, + "learning_rate": 8.811291656962808e-06, + "loss": 0.5244, + "step": 11238 + }, + { + "epoch": 0.24739159145938808, + "grad_norm": 1.3727182751390636, + "learning_rate": 8.811060922715261e-06, + "loss": 0.3447, + "step": 11239 + }, + { + "epoch": 0.24741360334580673, + "grad_norm": 1.5928896395880363, + "learning_rate": 8.810830169098206e-06, + "loss": 0.2818, + "step": 11240 + }, + { + "epoch": 0.2474356152322254, + "grad_norm": 1.2607551110271067, + "learning_rate": 8.81059939611281e-06, + "loss": 0.2543, + "step": 11241 + }, + { + "epoch": 0.24745762711864408, + "grad_norm": 1.0451879700061897, + "learning_rate": 8.81036860376025e-06, + "loss": 0.2951, + "step": 11242 + }, + { + "epoch": 0.24747963900506273, + "grad_norm": 1.3071652281102717, + "learning_rate": 8.810137792041695e-06, + "loss": 0.2814, + "step": 11243 + }, + { + "epoch": 0.2475016508914814, + "grad_norm": 1.815077954572266, + "learning_rate": 8.809906960958323e-06, + "loss": 0.3352, + "step": 11244 + }, + { + "epoch": 0.24752366277790006, + "grad_norm": 2.117646353390732, + "learning_rate": 8.809676110511303e-06, + "loss": 0.3691, + "step": 11245 + }, + { + "epoch": 0.24754567466431873, + "grad_norm": 1.3588544773057518, + "learning_rate": 8.80944524070181e-06, + "loss": 0.3351, + "step": 11246 + }, + { + "epoch": 0.2475676865507374, + "grad_norm": 1.5561095821573465, + "learning_rate": 8.809214351531016e-06, + "loss": 0.378, + "step": 11247 + }, + { + "epoch": 0.24758969843715606, + "grad_norm": 2.1420891522834107, + "learning_rate": 8.808983443000098e-06, + "loss": 0.3833, + "step": 11248 + }, + { + "epoch": 0.24761171032357474, + "grad_norm": 1.7528337306636415, + "learning_rate": 8.808752515110225e-06, + "loss": 0.3699, + "step": 11249 + }, + { + "epoch": 0.24763372220999338, + "grad_norm": 1.4228102474349853, + "learning_rate": 8.808521567862576e-06, + "loss": 0.3281, + "step": 11250 + }, + { + "epoch": 0.24765573409641206, + "grad_norm": 1.73466810233721, + "learning_rate": 8.80829060125832e-06, + "loss": 0.4361, + "step": 11251 + }, + { + "epoch": 0.24767774598283074, + "grad_norm": 1.8250793413442805, + "learning_rate": 8.808059615298632e-06, + "loss": 0.3684, + "step": 11252 + }, + { + "epoch": 0.2476997578692494, + "grad_norm": 1.4480982590511768, + "learning_rate": 8.807828609984687e-06, + "loss": 0.3456, + "step": 11253 + }, + { + "epoch": 0.24772176975566806, + "grad_norm": 1.3639362680206595, + "learning_rate": 8.80759758531766e-06, + "loss": 0.3568, + "step": 11254 + }, + { + "epoch": 0.2477437816420867, + "grad_norm": 1.53651336004419, + "learning_rate": 8.807366541298722e-06, + "loss": 0.386, + "step": 11255 + }, + { + "epoch": 0.2477657935285054, + "grad_norm": 1.4582335373851019, + "learning_rate": 8.80713547792905e-06, + "loss": 0.2345, + "step": 11256 + }, + { + "epoch": 0.24778780541492407, + "grad_norm": 0.7319843269511096, + "learning_rate": 8.806904395209816e-06, + "loss": 0.5116, + "step": 11257 + }, + { + "epoch": 0.24780981730134272, + "grad_norm": 2.5354939328580706, + "learning_rate": 8.806673293142199e-06, + "loss": 0.2976, + "step": 11258 + }, + { + "epoch": 0.2478318291877614, + "grad_norm": 1.4938004229212958, + "learning_rate": 8.806442171727369e-06, + "loss": 0.3064, + "step": 11259 + }, + { + "epoch": 0.24785384107418007, + "grad_norm": 1.4627674917464581, + "learning_rate": 8.806211030966502e-06, + "loss": 0.4338, + "step": 11260 + }, + { + "epoch": 0.24787585296059872, + "grad_norm": 4.148372361353215, + "learning_rate": 8.805979870860773e-06, + "loss": 0.3479, + "step": 11261 + }, + { + "epoch": 0.2478978648470174, + "grad_norm": 0.6593590476872894, + "learning_rate": 8.805748691411356e-06, + "loss": 0.5352, + "step": 11262 + }, + { + "epoch": 0.24791987673343605, + "grad_norm": 1.4625553796237152, + "learning_rate": 8.805517492619428e-06, + "loss": 0.3682, + "step": 11263 + }, + { + "epoch": 0.24794188861985472, + "grad_norm": 1.5602223018243275, + "learning_rate": 8.805286274486162e-06, + "loss": 0.3625, + "step": 11264 + }, + { + "epoch": 0.2479639005062734, + "grad_norm": 1.933626871385363, + "learning_rate": 8.805055037012733e-06, + "loss": 0.3641, + "step": 11265 + }, + { + "epoch": 0.24798591239269205, + "grad_norm": 1.4647052494562915, + "learning_rate": 8.804823780200318e-06, + "loss": 0.3426, + "step": 11266 + }, + { + "epoch": 0.24800792427911073, + "grad_norm": 1.2585729689468959, + "learning_rate": 8.804592504050093e-06, + "loss": 0.3073, + "step": 11267 + }, + { + "epoch": 0.24802993616552937, + "grad_norm": 1.672866669227463, + "learning_rate": 8.80436120856323e-06, + "loss": 0.3145, + "step": 11268 + }, + { + "epoch": 0.24805194805194805, + "grad_norm": 1.289194797029786, + "learning_rate": 8.804129893740907e-06, + "loss": 0.3124, + "step": 11269 + }, + { + "epoch": 0.24807395993836673, + "grad_norm": 1.5830197527205425, + "learning_rate": 8.8038985595843e-06, + "loss": 0.3759, + "step": 11270 + }, + { + "epoch": 0.24809597182478538, + "grad_norm": 0.6327252169806521, + "learning_rate": 8.803667206094583e-06, + "loss": 0.5087, + "step": 11271 + }, + { + "epoch": 0.24811798371120405, + "grad_norm": 1.4418648721800322, + "learning_rate": 8.803435833272934e-06, + "loss": 0.3242, + "step": 11272 + }, + { + "epoch": 0.2481399955976227, + "grad_norm": 1.5121593193960723, + "learning_rate": 8.803204441120527e-06, + "loss": 0.404, + "step": 11273 + }, + { + "epoch": 0.24816200748404138, + "grad_norm": 1.397316109534966, + "learning_rate": 8.80297302963854e-06, + "loss": 0.3757, + "step": 11274 + }, + { + "epoch": 0.24818401937046006, + "grad_norm": 1.733906462950233, + "learning_rate": 8.802741598828147e-06, + "loss": 0.386, + "step": 11275 + }, + { + "epoch": 0.2482060312568787, + "grad_norm": 1.7384074192598677, + "learning_rate": 8.802510148690523e-06, + "loss": 0.3881, + "step": 11276 + }, + { + "epoch": 0.24822804314329738, + "grad_norm": 0.602381132228102, + "learning_rate": 8.802278679226849e-06, + "loss": 0.5052, + "step": 11277 + }, + { + "epoch": 0.24825005502971606, + "grad_norm": 1.5491890645029343, + "learning_rate": 8.802047190438299e-06, + "loss": 0.3079, + "step": 11278 + }, + { + "epoch": 0.2482720669161347, + "grad_norm": 1.5708939237844535, + "learning_rate": 8.801815682326048e-06, + "loss": 0.3072, + "step": 11279 + }, + { + "epoch": 0.2482940788025534, + "grad_norm": 0.6477481692535595, + "learning_rate": 8.801584154891274e-06, + "loss": 0.5211, + "step": 11280 + }, + { + "epoch": 0.24831609068897204, + "grad_norm": 0.6314673687614227, + "learning_rate": 8.801352608135154e-06, + "loss": 0.5031, + "step": 11281 + }, + { + "epoch": 0.2483381025753907, + "grad_norm": 1.929150291050067, + "learning_rate": 8.801121042058866e-06, + "loss": 0.3176, + "step": 11282 + }, + { + "epoch": 0.2483601144618094, + "grad_norm": 2.8021940015262676, + "learning_rate": 8.800889456663584e-06, + "loss": 0.3915, + "step": 11283 + }, + { + "epoch": 0.24838212634822804, + "grad_norm": 1.280906686582417, + "learning_rate": 8.800657851950488e-06, + "loss": 0.2731, + "step": 11284 + }, + { + "epoch": 0.24840413823464672, + "grad_norm": 1.128917166679347, + "learning_rate": 8.800426227920753e-06, + "loss": 0.3209, + "step": 11285 + }, + { + "epoch": 0.24842615012106536, + "grad_norm": 1.4158750892498029, + "learning_rate": 8.800194584575556e-06, + "loss": 0.3166, + "step": 11286 + }, + { + "epoch": 0.24844816200748404, + "grad_norm": 1.267549404010825, + "learning_rate": 8.799962921916074e-06, + "loss": 0.3995, + "step": 11287 + }, + { + "epoch": 0.24847017389390272, + "grad_norm": 1.1555924767228432, + "learning_rate": 8.799731239943488e-06, + "loss": 0.3605, + "step": 11288 + }, + { + "epoch": 0.24849218578032137, + "grad_norm": 1.4235557598746025, + "learning_rate": 8.799499538658973e-06, + "loss": 0.4239, + "step": 11289 + }, + { + "epoch": 0.24851419766674004, + "grad_norm": 0.6659052804236495, + "learning_rate": 8.799267818063706e-06, + "loss": 0.5211, + "step": 11290 + }, + { + "epoch": 0.2485362095531587, + "grad_norm": 1.413092194458568, + "learning_rate": 8.799036078158864e-06, + "loss": 0.3899, + "step": 11291 + }, + { + "epoch": 0.24855822143957737, + "grad_norm": 1.3427136748112398, + "learning_rate": 8.798804318945629e-06, + "loss": 0.3957, + "step": 11292 + }, + { + "epoch": 0.24858023332599605, + "grad_norm": 1.1798768823869759, + "learning_rate": 8.798572540425175e-06, + "loss": 0.3157, + "step": 11293 + }, + { + "epoch": 0.2486022452124147, + "grad_norm": 1.5855178098276173, + "learning_rate": 8.798340742598682e-06, + "loss": 0.3425, + "step": 11294 + }, + { + "epoch": 0.24862425709883337, + "grad_norm": 3.089476960618388, + "learning_rate": 8.798108925467326e-06, + "loss": 0.3656, + "step": 11295 + }, + { + "epoch": 0.24864626898525202, + "grad_norm": 7.856789992955878, + "learning_rate": 8.797877089032287e-06, + "loss": 0.2647, + "step": 11296 + }, + { + "epoch": 0.2486682808716707, + "grad_norm": 1.2730946910601484, + "learning_rate": 8.797645233294742e-06, + "loss": 0.3293, + "step": 11297 + }, + { + "epoch": 0.24869029275808938, + "grad_norm": 0.6368225835390973, + "learning_rate": 8.79741335825587e-06, + "loss": 0.5124, + "step": 11298 + }, + { + "epoch": 0.24871230464450803, + "grad_norm": 2.486535447790954, + "learning_rate": 8.797181463916851e-06, + "loss": 0.4337, + "step": 11299 + }, + { + "epoch": 0.2487343165309267, + "grad_norm": 0.6073589512871854, + "learning_rate": 8.796949550278862e-06, + "loss": 0.4958, + "step": 11300 + }, + { + "epoch": 0.24875632841734538, + "grad_norm": 0.6175731551118164, + "learning_rate": 8.79671761734308e-06, + "loss": 0.5227, + "step": 11301 + }, + { + "epoch": 0.24877834030376403, + "grad_norm": 0.6041455435962554, + "learning_rate": 8.796485665110687e-06, + "loss": 0.4803, + "step": 11302 + }, + { + "epoch": 0.2488003521901827, + "grad_norm": 1.3426817421913113, + "learning_rate": 8.796253693582862e-06, + "loss": 0.3288, + "step": 11303 + }, + { + "epoch": 0.24882236407660135, + "grad_norm": 1.1660931446372163, + "learning_rate": 8.796021702760779e-06, + "loss": 0.3053, + "step": 11304 + }, + { + "epoch": 0.24884437596302003, + "grad_norm": 1.498943841468921, + "learning_rate": 8.795789692645624e-06, + "loss": 0.3848, + "step": 11305 + }, + { + "epoch": 0.2488663878494387, + "grad_norm": 0.6237692522445343, + "learning_rate": 8.795557663238571e-06, + "loss": 0.5019, + "step": 11306 + }, + { + "epoch": 0.24888839973585736, + "grad_norm": 1.3807389217387918, + "learning_rate": 8.795325614540803e-06, + "loss": 0.3822, + "step": 11307 + }, + { + "epoch": 0.24891041162227603, + "grad_norm": 0.6317918290071391, + "learning_rate": 8.795093546553495e-06, + "loss": 0.5068, + "step": 11308 + }, + { + "epoch": 0.24893242350869468, + "grad_norm": 3.5039870401385347, + "learning_rate": 8.79486145927783e-06, + "loss": 0.3701, + "step": 11309 + }, + { + "epoch": 0.24895443539511336, + "grad_norm": 2.0377869703395235, + "learning_rate": 8.794629352714987e-06, + "loss": 0.4248, + "step": 11310 + }, + { + "epoch": 0.24897644728153204, + "grad_norm": 1.432362614309978, + "learning_rate": 8.794397226866145e-06, + "loss": 0.3524, + "step": 11311 + }, + { + "epoch": 0.2489984591679507, + "grad_norm": 1.3993238009544566, + "learning_rate": 8.794165081732485e-06, + "loss": 0.375, + "step": 11312 + }, + { + "epoch": 0.24902047105436936, + "grad_norm": 1.8803871712402822, + "learning_rate": 8.793932917315184e-06, + "loss": 0.3794, + "step": 11313 + }, + { + "epoch": 0.249042482940788, + "grad_norm": 1.3728595055400867, + "learning_rate": 8.793700733615424e-06, + "loss": 0.3147, + "step": 11314 + }, + { + "epoch": 0.2490644948272067, + "grad_norm": 1.1799052171134317, + "learning_rate": 8.793468530634385e-06, + "loss": 0.342, + "step": 11315 + }, + { + "epoch": 0.24908650671362537, + "grad_norm": 1.4466758481758086, + "learning_rate": 8.793236308373247e-06, + "loss": 0.336, + "step": 11316 + }, + { + "epoch": 0.24910851860004402, + "grad_norm": 1.9974309790664497, + "learning_rate": 8.793004066833191e-06, + "loss": 0.2774, + "step": 11317 + }, + { + "epoch": 0.2491305304864627, + "grad_norm": 3.46769651408518, + "learning_rate": 8.792771806015396e-06, + "loss": 0.3103, + "step": 11318 + }, + { + "epoch": 0.24915254237288137, + "grad_norm": 1.5069358139686753, + "learning_rate": 8.792539525921042e-06, + "loss": 0.3994, + "step": 11319 + }, + { + "epoch": 0.24917455425930002, + "grad_norm": 1.4069636338066782, + "learning_rate": 8.792307226551313e-06, + "loss": 0.304, + "step": 11320 + }, + { + "epoch": 0.2491965661457187, + "grad_norm": 0.7194169055939519, + "learning_rate": 8.792074907907386e-06, + "loss": 0.5371, + "step": 11321 + }, + { + "epoch": 0.24921857803213734, + "grad_norm": 2.915199177983485, + "learning_rate": 8.791842569990443e-06, + "loss": 0.3588, + "step": 11322 + }, + { + "epoch": 0.24924058991855602, + "grad_norm": 1.5303801802256447, + "learning_rate": 8.791610212801664e-06, + "loss": 0.3508, + "step": 11323 + }, + { + "epoch": 0.2492626018049747, + "grad_norm": 0.6470639132356394, + "learning_rate": 8.791377836342232e-06, + "loss": 0.5501, + "step": 11324 + }, + { + "epoch": 0.24928461369139335, + "grad_norm": 1.234704186392743, + "learning_rate": 8.791145440613326e-06, + "loss": 0.3142, + "step": 11325 + }, + { + "epoch": 0.24930662557781202, + "grad_norm": 1.3642306184736581, + "learning_rate": 8.790913025616128e-06, + "loss": 0.3976, + "step": 11326 + }, + { + "epoch": 0.24932863746423067, + "grad_norm": 0.6563838307039085, + "learning_rate": 8.79068059135182e-06, + "loss": 0.5131, + "step": 11327 + }, + { + "epoch": 0.24935064935064935, + "grad_norm": 1.4968372607179419, + "learning_rate": 8.790448137821582e-06, + "loss": 0.354, + "step": 11328 + }, + { + "epoch": 0.24937266123706803, + "grad_norm": 0.645080627147253, + "learning_rate": 8.790215665026596e-06, + "loss": 0.5171, + "step": 11329 + }, + { + "epoch": 0.24939467312348668, + "grad_norm": 1.2969607601096962, + "learning_rate": 8.789983172968044e-06, + "loss": 0.3974, + "step": 11330 + }, + { + "epoch": 0.24941668500990535, + "grad_norm": 1.482955556629653, + "learning_rate": 8.789750661647107e-06, + "loss": 0.3701, + "step": 11331 + }, + { + "epoch": 0.249438696896324, + "grad_norm": 1.3734722501645151, + "learning_rate": 8.789518131064968e-06, + "loss": 0.308, + "step": 11332 + }, + { + "epoch": 0.24946070878274268, + "grad_norm": 2.0875108992176186, + "learning_rate": 8.789285581222804e-06, + "loss": 0.3517, + "step": 11333 + }, + { + "epoch": 0.24948272066916136, + "grad_norm": 1.2959110541812844, + "learning_rate": 8.789053012121804e-06, + "loss": 0.3636, + "step": 11334 + }, + { + "epoch": 0.24950473255558, + "grad_norm": 1.3388425341018113, + "learning_rate": 8.788820423763144e-06, + "loss": 0.3254, + "step": 11335 + }, + { + "epoch": 0.24952674444199868, + "grad_norm": 0.6894446406973853, + "learning_rate": 8.78858781614801e-06, + "loss": 0.5367, + "step": 11336 + }, + { + "epoch": 0.24954875632841733, + "grad_norm": 0.6639544958980367, + "learning_rate": 8.788355189277585e-06, + "loss": 0.5375, + "step": 11337 + }, + { + "epoch": 0.249570768214836, + "grad_norm": 1.3211019346223158, + "learning_rate": 8.788122543153047e-06, + "loss": 0.3726, + "step": 11338 + }, + { + "epoch": 0.24959278010125469, + "grad_norm": 1.8463374372660537, + "learning_rate": 8.787889877775582e-06, + "loss": 0.4261, + "step": 11339 + }, + { + "epoch": 0.24961479198767333, + "grad_norm": 1.4256422081019848, + "learning_rate": 8.78765719314637e-06, + "loss": 0.4416, + "step": 11340 + }, + { + "epoch": 0.249636803874092, + "grad_norm": 1.7958594804917958, + "learning_rate": 8.787424489266594e-06, + "loss": 0.4079, + "step": 11341 + }, + { + "epoch": 0.2496588157605107, + "grad_norm": 1.2611280337579174, + "learning_rate": 8.78719176613744e-06, + "loss": 0.2967, + "step": 11342 + }, + { + "epoch": 0.24968082764692934, + "grad_norm": 2.4750917411875735, + "learning_rate": 8.786959023760086e-06, + "loss": 0.3992, + "step": 11343 + }, + { + "epoch": 0.24970283953334801, + "grad_norm": 1.4534511567166837, + "learning_rate": 8.786726262135719e-06, + "loss": 0.3739, + "step": 11344 + }, + { + "epoch": 0.24972485141976666, + "grad_norm": 1.2867856179527537, + "learning_rate": 8.786493481265519e-06, + "loss": 0.3214, + "step": 11345 + }, + { + "epoch": 0.24974686330618534, + "grad_norm": 1.3733808156068634, + "learning_rate": 8.786260681150672e-06, + "loss": 0.3526, + "step": 11346 + }, + { + "epoch": 0.24976887519260402, + "grad_norm": 0.8140578222334386, + "learning_rate": 8.786027861792357e-06, + "loss": 0.5404, + "step": 11347 + }, + { + "epoch": 0.24979088707902267, + "grad_norm": 2.14017410377149, + "learning_rate": 8.785795023191762e-06, + "loss": 0.3532, + "step": 11348 + }, + { + "epoch": 0.24981289896544134, + "grad_norm": 1.6749384217208754, + "learning_rate": 8.785562165350066e-06, + "loss": 0.4935, + "step": 11349 + }, + { + "epoch": 0.24983491085186, + "grad_norm": 0.6731575195972423, + "learning_rate": 8.785329288268456e-06, + "loss": 0.5126, + "step": 11350 + }, + { + "epoch": 0.24985692273827867, + "grad_norm": 0.6316500796484102, + "learning_rate": 8.785096391948115e-06, + "loss": 0.5479, + "step": 11351 + }, + { + "epoch": 0.24987893462469735, + "grad_norm": 1.275991202670063, + "learning_rate": 8.784863476390223e-06, + "loss": 0.3079, + "step": 11352 + }, + { + "epoch": 0.249900946511116, + "grad_norm": 5.335419010400723, + "learning_rate": 8.784630541595971e-06, + "loss": 0.3702, + "step": 11353 + }, + { + "epoch": 0.24992295839753467, + "grad_norm": 4.001724815644281, + "learning_rate": 8.784397587566535e-06, + "loss": 0.3113, + "step": 11354 + }, + { + "epoch": 0.24994497028395332, + "grad_norm": 0.7798692176180546, + "learning_rate": 8.784164614303104e-06, + "loss": 0.5044, + "step": 11355 + }, + { + "epoch": 0.249966982170372, + "grad_norm": 1.6714660476970453, + "learning_rate": 8.783931621806862e-06, + "loss": 0.3913, + "step": 11356 + }, + { + "epoch": 0.24998899405679068, + "grad_norm": 1.3854755732420103, + "learning_rate": 8.78369861007899e-06, + "loss": 0.351, + "step": 11357 + }, + { + "epoch": 0.25001100594320935, + "grad_norm": 3.410559414084295, + "learning_rate": 8.783465579120674e-06, + "loss": 0.3609, + "step": 11358 + }, + { + "epoch": 0.250033017829628, + "grad_norm": 17.577510858544166, + "learning_rate": 8.783232528933099e-06, + "loss": 0.3646, + "step": 11359 + }, + { + "epoch": 0.25005502971604665, + "grad_norm": 1.9832523569541385, + "learning_rate": 8.78299945951745e-06, + "loss": 0.3672, + "step": 11360 + }, + { + "epoch": 0.25007704160246536, + "grad_norm": 1.5417928803516927, + "learning_rate": 8.78276637087491e-06, + "loss": 0.3946, + "step": 11361 + }, + { + "epoch": 0.250099053488884, + "grad_norm": 0.7082650309626831, + "learning_rate": 8.782533263006662e-06, + "loss": 0.5281, + "step": 11362 + }, + { + "epoch": 0.25012106537530265, + "grad_norm": 1.4841198873350672, + "learning_rate": 8.782300135913894e-06, + "loss": 0.3907, + "step": 11363 + }, + { + "epoch": 0.2501430772617213, + "grad_norm": 0.6828559900019162, + "learning_rate": 8.78206698959779e-06, + "loss": 0.5177, + "step": 11364 + }, + { + "epoch": 0.25016508914814, + "grad_norm": 0.6155228402387287, + "learning_rate": 8.781833824059536e-06, + "loss": 0.521, + "step": 11365 + }, + { + "epoch": 0.25018710103455866, + "grad_norm": 1.8759224994778618, + "learning_rate": 8.781600639300313e-06, + "loss": 0.3099, + "step": 11366 + }, + { + "epoch": 0.2502091129209773, + "grad_norm": 2.2547530594534786, + "learning_rate": 8.781367435321313e-06, + "loss": 0.3986, + "step": 11367 + }, + { + "epoch": 0.250231124807396, + "grad_norm": 1.3928514613767484, + "learning_rate": 8.781134212123714e-06, + "loss": 0.2851, + "step": 11368 + }, + { + "epoch": 0.25025313669381466, + "grad_norm": 1.4839216953225869, + "learning_rate": 8.780900969708705e-06, + "loss": 0.363, + "step": 11369 + }, + { + "epoch": 0.2502751485802333, + "grad_norm": 1.7184546822417637, + "learning_rate": 8.78066770807747e-06, + "loss": 0.3591, + "step": 11370 + }, + { + "epoch": 0.250297160466652, + "grad_norm": 1.7226685283633572, + "learning_rate": 8.780434427231197e-06, + "loss": 0.389, + "step": 11371 + }, + { + "epoch": 0.25031917235307066, + "grad_norm": 1.1619957486888706, + "learning_rate": 8.780201127171069e-06, + "loss": 0.2876, + "step": 11372 + }, + { + "epoch": 0.2503411842394893, + "grad_norm": 1.2451689090197071, + "learning_rate": 8.779967807898275e-06, + "loss": 0.2813, + "step": 11373 + }, + { + "epoch": 0.250363196125908, + "grad_norm": 1.2726711526162344, + "learning_rate": 8.779734469413997e-06, + "loss": 0.3266, + "step": 11374 + }, + { + "epoch": 0.25038520801232667, + "grad_norm": 1.4787782450156088, + "learning_rate": 8.779501111719423e-06, + "loss": 0.3319, + "step": 11375 + }, + { + "epoch": 0.2504072198987453, + "grad_norm": 1.4923064605333296, + "learning_rate": 8.779267734815738e-06, + "loss": 0.3458, + "step": 11376 + }, + { + "epoch": 0.25042923178516396, + "grad_norm": 1.331482930596254, + "learning_rate": 8.779034338704128e-06, + "loss": 0.35, + "step": 11377 + }, + { + "epoch": 0.25045124367158267, + "grad_norm": 1.7900260852362915, + "learning_rate": 8.778800923385783e-06, + "loss": 0.3302, + "step": 11378 + }, + { + "epoch": 0.2504732555580013, + "grad_norm": 3.0446925654149855, + "learning_rate": 8.778567488861885e-06, + "loss": 0.3263, + "step": 11379 + }, + { + "epoch": 0.25049526744441997, + "grad_norm": 1.4986677613896262, + "learning_rate": 8.77833403513362e-06, + "loss": 0.3285, + "step": 11380 + }, + { + "epoch": 0.25051727933083867, + "grad_norm": 1.8095794114508224, + "learning_rate": 8.778100562202179e-06, + "loss": 0.3675, + "step": 11381 + }, + { + "epoch": 0.2505392912172573, + "grad_norm": 1.2551879223785178, + "learning_rate": 8.777867070068745e-06, + "loss": 0.5387, + "step": 11382 + }, + { + "epoch": 0.25056130310367597, + "grad_norm": 1.3371833589636437, + "learning_rate": 8.777633558734503e-06, + "loss": 0.4028, + "step": 11383 + }, + { + "epoch": 0.2505833149900947, + "grad_norm": 1.2580210701562284, + "learning_rate": 8.777400028200646e-06, + "loss": 0.2852, + "step": 11384 + }, + { + "epoch": 0.2506053268765133, + "grad_norm": 1.2131023336905733, + "learning_rate": 8.777166478468356e-06, + "loss": 0.35, + "step": 11385 + }, + { + "epoch": 0.250627338762932, + "grad_norm": 1.4696517108998648, + "learning_rate": 8.77693290953882e-06, + "loss": 0.3074, + "step": 11386 + }, + { + "epoch": 0.2506493506493506, + "grad_norm": 1.4371205471763713, + "learning_rate": 8.776699321413227e-06, + "loss": 0.3564, + "step": 11387 + }, + { + "epoch": 0.2506713625357693, + "grad_norm": 2.0690380881532273, + "learning_rate": 8.776465714092765e-06, + "loss": 0.4059, + "step": 11388 + }, + { + "epoch": 0.250693374422188, + "grad_norm": 1.4493896260433587, + "learning_rate": 8.77623208757862e-06, + "loss": 0.362, + "step": 11389 + }, + { + "epoch": 0.2507153863086066, + "grad_norm": 1.3718567818539438, + "learning_rate": 8.775998441871977e-06, + "loss": 0.2948, + "step": 11390 + }, + { + "epoch": 0.25073739819502533, + "grad_norm": 1.332116328441596, + "learning_rate": 8.775764776974027e-06, + "loss": 0.3098, + "step": 11391 + }, + { + "epoch": 0.250759410081444, + "grad_norm": 1.3167578259935715, + "learning_rate": 8.775531092885956e-06, + "loss": 0.3452, + "step": 11392 + }, + { + "epoch": 0.2507814219678626, + "grad_norm": 1.649375710477534, + "learning_rate": 8.775297389608953e-06, + "loss": 0.4064, + "step": 11393 + }, + { + "epoch": 0.25080343385428133, + "grad_norm": 1.0693559087077458, + "learning_rate": 8.775063667144204e-06, + "loss": 0.4081, + "step": 11394 + }, + { + "epoch": 0.2508254457407, + "grad_norm": 1.505330423208299, + "learning_rate": 8.774829925492898e-06, + "loss": 0.3849, + "step": 11395 + }, + { + "epoch": 0.25084745762711863, + "grad_norm": 1.373794281846837, + "learning_rate": 8.774596164656222e-06, + "loss": 0.3469, + "step": 11396 + }, + { + "epoch": 0.25086946951353734, + "grad_norm": 1.243989693272714, + "learning_rate": 8.774362384635365e-06, + "loss": 0.2827, + "step": 11397 + }, + { + "epoch": 0.250891481399956, + "grad_norm": 2.116683951795214, + "learning_rate": 8.774128585431517e-06, + "loss": 0.3559, + "step": 11398 + }, + { + "epoch": 0.25091349328637463, + "grad_norm": 1.2315283051953079, + "learning_rate": 8.77389476704586e-06, + "loss": 0.361, + "step": 11399 + }, + { + "epoch": 0.2509355051727933, + "grad_norm": 1.1914964908705026, + "learning_rate": 8.77366092947959e-06, + "loss": 0.3556, + "step": 11400 + }, + { + "epoch": 0.250957517059212, + "grad_norm": 1.4949327900071232, + "learning_rate": 8.773427072733893e-06, + "loss": 0.2734, + "step": 11401 + }, + { + "epoch": 0.25097952894563064, + "grad_norm": 1.3915328986898827, + "learning_rate": 8.773193196809954e-06, + "loss": 0.3474, + "step": 11402 + }, + { + "epoch": 0.2510015408320493, + "grad_norm": 1.2600251206677826, + "learning_rate": 8.772959301708965e-06, + "loss": 0.2999, + "step": 11403 + }, + { + "epoch": 0.251023552718468, + "grad_norm": 1.8425548648260612, + "learning_rate": 8.772725387432114e-06, + "loss": 0.2942, + "step": 11404 + }, + { + "epoch": 0.25104556460488664, + "grad_norm": 1.5221588000981463, + "learning_rate": 8.77249145398059e-06, + "loss": 0.312, + "step": 11405 + }, + { + "epoch": 0.2510675764913053, + "grad_norm": 1.5454912649209789, + "learning_rate": 8.772257501355582e-06, + "loss": 0.3332, + "step": 11406 + }, + { + "epoch": 0.251089588377724, + "grad_norm": 0.7526152083861398, + "learning_rate": 8.772023529558278e-06, + "loss": 0.5375, + "step": 11407 + }, + { + "epoch": 0.25111160026414264, + "grad_norm": 0.7839439285526023, + "learning_rate": 8.77178953858987e-06, + "loss": 0.5654, + "step": 11408 + }, + { + "epoch": 0.2511336121505613, + "grad_norm": 2.7287796229857255, + "learning_rate": 8.771555528451543e-06, + "loss": 0.2847, + "step": 11409 + }, + { + "epoch": 0.25115562403698, + "grad_norm": 1.3034524328586288, + "learning_rate": 8.77132149914449e-06, + "loss": 0.3147, + "step": 11410 + }, + { + "epoch": 0.25117763592339865, + "grad_norm": 1.5224104245958314, + "learning_rate": 8.7710874506699e-06, + "loss": 0.3961, + "step": 11411 + }, + { + "epoch": 0.2511996478098173, + "grad_norm": 1.9402791309257787, + "learning_rate": 8.77085338302896e-06, + "loss": 0.4233, + "step": 11412 + }, + { + "epoch": 0.25122165969623594, + "grad_norm": 3.080112803361017, + "learning_rate": 8.770619296222862e-06, + "loss": 0.3957, + "step": 11413 + }, + { + "epoch": 0.25124367158265465, + "grad_norm": 0.793654382184575, + "learning_rate": 8.770385190252794e-06, + "loss": 0.558, + "step": 11414 + }, + { + "epoch": 0.2512656834690733, + "grad_norm": 1.6324785113257925, + "learning_rate": 8.770151065119948e-06, + "loss": 0.3243, + "step": 11415 + }, + { + "epoch": 0.25128769535549195, + "grad_norm": 0.745998539488706, + "learning_rate": 8.769916920825512e-06, + "loss": 0.5341, + "step": 11416 + }, + { + "epoch": 0.25130970724191065, + "grad_norm": 1.4121774644350076, + "learning_rate": 8.769682757370677e-06, + "loss": 0.4179, + "step": 11417 + }, + { + "epoch": 0.2513317191283293, + "grad_norm": 1.823483843019801, + "learning_rate": 8.769448574756633e-06, + "loss": 0.3873, + "step": 11418 + }, + { + "epoch": 0.25135373101474795, + "grad_norm": 1.7623871891607195, + "learning_rate": 8.76921437298457e-06, + "loss": 0.3295, + "step": 11419 + }, + { + "epoch": 0.25137574290116665, + "grad_norm": 1.2272001513082602, + "learning_rate": 8.768980152055679e-06, + "loss": 0.3136, + "step": 11420 + }, + { + "epoch": 0.2513977547875853, + "grad_norm": 1.3883572703290614, + "learning_rate": 8.768745911971151e-06, + "loss": 0.3881, + "step": 11421 + }, + { + "epoch": 0.25141976667400395, + "grad_norm": 1.3755488427876315, + "learning_rate": 8.768511652732172e-06, + "loss": 0.3701, + "step": 11422 + }, + { + "epoch": 0.2514417785604226, + "grad_norm": 0.7081175930605942, + "learning_rate": 8.768277374339939e-06, + "loss": 0.5152, + "step": 11423 + }, + { + "epoch": 0.2514637904468413, + "grad_norm": 1.2155539012723688, + "learning_rate": 8.768043076795638e-06, + "loss": 0.267, + "step": 11424 + }, + { + "epoch": 0.25148580233325996, + "grad_norm": 0.7001344821274001, + "learning_rate": 8.767808760100463e-06, + "loss": 0.5456, + "step": 11425 + }, + { + "epoch": 0.2515078142196786, + "grad_norm": 2.1536329229440594, + "learning_rate": 8.7675744242556e-06, + "loss": 0.4079, + "step": 11426 + }, + { + "epoch": 0.2515298261060973, + "grad_norm": 1.417884022938002, + "learning_rate": 8.767340069262248e-06, + "loss": 0.3521, + "step": 11427 + }, + { + "epoch": 0.25155183799251596, + "grad_norm": 1.4497559445449082, + "learning_rate": 8.767105695121591e-06, + "loss": 0.3431, + "step": 11428 + }, + { + "epoch": 0.2515738498789346, + "grad_norm": 1.3882346153299538, + "learning_rate": 8.766871301834824e-06, + "loss": 0.3306, + "step": 11429 + }, + { + "epoch": 0.2515958617653533, + "grad_norm": 1.2325098322252677, + "learning_rate": 8.766636889403135e-06, + "loss": 0.3308, + "step": 11430 + }, + { + "epoch": 0.25161787365177196, + "grad_norm": 1.4051292806266162, + "learning_rate": 8.766402457827717e-06, + "loss": 0.3313, + "step": 11431 + }, + { + "epoch": 0.2516398855381906, + "grad_norm": 1.5460556397675296, + "learning_rate": 8.766168007109762e-06, + "loss": 0.401, + "step": 11432 + }, + { + "epoch": 0.2516618974246093, + "grad_norm": 1.8048281011104073, + "learning_rate": 8.765933537250464e-06, + "loss": 0.4166, + "step": 11433 + }, + { + "epoch": 0.25168390931102796, + "grad_norm": 1.7004923964112684, + "learning_rate": 8.76569904825101e-06, + "loss": 0.3132, + "step": 11434 + }, + { + "epoch": 0.2517059211974466, + "grad_norm": 0.7282617092204257, + "learning_rate": 8.765464540112594e-06, + "loss": 0.5253, + "step": 11435 + }, + { + "epoch": 0.25172793308386526, + "grad_norm": 0.6890305384663271, + "learning_rate": 8.765230012836409e-06, + "loss": 0.5211, + "step": 11436 + }, + { + "epoch": 0.25174994497028397, + "grad_norm": 0.6497437807351067, + "learning_rate": 8.764995466423645e-06, + "loss": 0.5375, + "step": 11437 + }, + { + "epoch": 0.2517719568567026, + "grad_norm": 0.6203974494359166, + "learning_rate": 8.764760900875494e-06, + "loss": 0.5215, + "step": 11438 + }, + { + "epoch": 0.25179396874312127, + "grad_norm": 1.29830494779816, + "learning_rate": 8.764526316193152e-06, + "loss": 0.4021, + "step": 11439 + }, + { + "epoch": 0.25181598062953997, + "grad_norm": 1.5123274801225122, + "learning_rate": 8.764291712377804e-06, + "loss": 0.2956, + "step": 11440 + }, + { + "epoch": 0.2518379925159586, + "grad_norm": 0.7317588220821067, + "learning_rate": 8.76405708943065e-06, + "loss": 0.5502, + "step": 11441 + }, + { + "epoch": 0.25186000440237727, + "grad_norm": 0.9904019128556545, + "learning_rate": 8.763822447352878e-06, + "loss": 0.2723, + "step": 11442 + }, + { + "epoch": 0.251882016288796, + "grad_norm": 1.6757298874612265, + "learning_rate": 8.76358778614568e-06, + "loss": 0.3995, + "step": 11443 + }, + { + "epoch": 0.2519040281752146, + "grad_norm": 1.3111872708053809, + "learning_rate": 8.763353105810253e-06, + "loss": 0.2692, + "step": 11444 + }, + { + "epoch": 0.25192604006163327, + "grad_norm": 0.7576756804665663, + "learning_rate": 8.763118406347785e-06, + "loss": 0.5239, + "step": 11445 + }, + { + "epoch": 0.2519480519480519, + "grad_norm": 1.5646049466324827, + "learning_rate": 8.762883687759471e-06, + "loss": 0.3096, + "step": 11446 + }, + { + "epoch": 0.2519700638344706, + "grad_norm": 1.408597456439842, + "learning_rate": 8.762648950046503e-06, + "loss": 0.4471, + "step": 11447 + }, + { + "epoch": 0.2519920757208893, + "grad_norm": 1.5096045549950976, + "learning_rate": 8.762414193210078e-06, + "loss": 0.311, + "step": 11448 + }, + { + "epoch": 0.2520140876073079, + "grad_norm": 1.3529405920761821, + "learning_rate": 8.762179417251384e-06, + "loss": 0.3527, + "step": 11449 + }, + { + "epoch": 0.25203609949372663, + "grad_norm": 2.073924385892425, + "learning_rate": 8.761944622171616e-06, + "loss": 0.3376, + "step": 11450 + }, + { + "epoch": 0.2520581113801453, + "grad_norm": 1.3713193679437499, + "learning_rate": 8.761709807971968e-06, + "loss": 0.3689, + "step": 11451 + }, + { + "epoch": 0.2520801232665639, + "grad_norm": 1.5115410303846455, + "learning_rate": 8.761474974653632e-06, + "loss": 0.3131, + "step": 11452 + }, + { + "epoch": 0.25210213515298263, + "grad_norm": 0.7226458675189993, + "learning_rate": 8.761240122217804e-06, + "loss": 0.5451, + "step": 11453 + }, + { + "epoch": 0.2521241470394013, + "grad_norm": 1.442890168693017, + "learning_rate": 8.761005250665675e-06, + "loss": 0.3144, + "step": 11454 + }, + { + "epoch": 0.25214615892581993, + "grad_norm": 1.1503615813475296, + "learning_rate": 8.76077035999844e-06, + "loss": 0.287, + "step": 11455 + }, + { + "epoch": 0.25216817081223863, + "grad_norm": 1.4111337186919455, + "learning_rate": 8.760535450217293e-06, + "loss": 0.3595, + "step": 11456 + }, + { + "epoch": 0.2521901826986573, + "grad_norm": 2.209444919201772, + "learning_rate": 8.760300521323429e-06, + "loss": 0.317, + "step": 11457 + }, + { + "epoch": 0.25221219458507593, + "grad_norm": 1.6809383469314412, + "learning_rate": 8.760065573318039e-06, + "loss": 0.3723, + "step": 11458 + }, + { + "epoch": 0.2522342064714946, + "grad_norm": 1.560989520747391, + "learning_rate": 8.759830606202318e-06, + "loss": 0.3019, + "step": 11459 + }, + { + "epoch": 0.2522562183579133, + "grad_norm": 0.6696741312067581, + "learning_rate": 8.759595619977463e-06, + "loss": 0.5488, + "step": 11460 + }, + { + "epoch": 0.25227823024433194, + "grad_norm": 1.7001453765480696, + "learning_rate": 8.759360614644666e-06, + "loss": 0.2798, + "step": 11461 + }, + { + "epoch": 0.2523002421307506, + "grad_norm": 0.6376797035374283, + "learning_rate": 8.75912559020512e-06, + "loss": 0.5265, + "step": 11462 + }, + { + "epoch": 0.2523222540171693, + "grad_norm": 1.8360419436169884, + "learning_rate": 8.758890546660022e-06, + "loss": 0.3501, + "step": 11463 + }, + { + "epoch": 0.25234426590358794, + "grad_norm": 1.4581766858250997, + "learning_rate": 8.758655484010565e-06, + "loss": 0.3325, + "step": 11464 + }, + { + "epoch": 0.2523662777900066, + "grad_norm": 0.626210978567362, + "learning_rate": 8.758420402257946e-06, + "loss": 0.5329, + "step": 11465 + }, + { + "epoch": 0.2523882896764253, + "grad_norm": 1.421304238281193, + "learning_rate": 8.758185301403356e-06, + "loss": 0.3848, + "step": 11466 + }, + { + "epoch": 0.25241030156284394, + "grad_norm": 1.2200661017610361, + "learning_rate": 8.757950181447994e-06, + "loss": 0.3796, + "step": 11467 + }, + { + "epoch": 0.2524323134492626, + "grad_norm": 1.2446379325020271, + "learning_rate": 8.757715042393052e-06, + "loss": 0.2763, + "step": 11468 + }, + { + "epoch": 0.2524543253356813, + "grad_norm": 2.0543676220794245, + "learning_rate": 8.757479884239726e-06, + "loss": 0.3971, + "step": 11469 + }, + { + "epoch": 0.25247633722209994, + "grad_norm": 1.4400132341481897, + "learning_rate": 8.757244706989212e-06, + "loss": 0.2502, + "step": 11470 + }, + { + "epoch": 0.2524983491085186, + "grad_norm": 0.7028882846096334, + "learning_rate": 8.757009510642705e-06, + "loss": 0.556, + "step": 11471 + }, + { + "epoch": 0.25252036099493724, + "grad_norm": 1.16445905376123, + "learning_rate": 8.7567742952014e-06, + "loss": 0.3597, + "step": 11472 + }, + { + "epoch": 0.25254237288135595, + "grad_norm": 0.6489745882430018, + "learning_rate": 8.756539060666492e-06, + "loss": 0.5287, + "step": 11473 + }, + { + "epoch": 0.2525643847677746, + "grad_norm": 0.6149660549466683, + "learning_rate": 8.756303807039176e-06, + "loss": 0.5097, + "step": 11474 + }, + { + "epoch": 0.25258639665419325, + "grad_norm": 1.3126287460358717, + "learning_rate": 8.75606853432065e-06, + "loss": 0.3609, + "step": 11475 + }, + { + "epoch": 0.25260840854061195, + "grad_norm": 0.6192324982700281, + "learning_rate": 8.755833242512108e-06, + "loss": 0.51, + "step": 11476 + }, + { + "epoch": 0.2526304204270306, + "grad_norm": 1.3979760061707764, + "learning_rate": 8.755597931614748e-06, + "loss": 0.3522, + "step": 11477 + }, + { + "epoch": 0.25265243231344925, + "grad_norm": 1.5468454378676968, + "learning_rate": 8.755362601629762e-06, + "loss": 0.3416, + "step": 11478 + }, + { + "epoch": 0.25267444419986795, + "grad_norm": 1.4439892586321732, + "learning_rate": 8.75512725255835e-06, + "loss": 0.3468, + "step": 11479 + }, + { + "epoch": 0.2526964560862866, + "grad_norm": 1.5396159274072505, + "learning_rate": 8.754891884401705e-06, + "loss": 0.4348, + "step": 11480 + }, + { + "epoch": 0.25271846797270525, + "grad_norm": 3.5083454921250934, + "learning_rate": 8.754656497161025e-06, + "loss": 0.2905, + "step": 11481 + }, + { + "epoch": 0.2527404798591239, + "grad_norm": 1.1835401152873304, + "learning_rate": 8.754421090837507e-06, + "loss": 0.2861, + "step": 11482 + }, + { + "epoch": 0.2527624917455426, + "grad_norm": 1.2885369824216193, + "learning_rate": 8.754185665432345e-06, + "loss": 0.2812, + "step": 11483 + }, + { + "epoch": 0.25278450363196125, + "grad_norm": 0.656659345131359, + "learning_rate": 8.753950220946738e-06, + "loss": 0.5241, + "step": 11484 + }, + { + "epoch": 0.2528065155183799, + "grad_norm": 1.2830232011025078, + "learning_rate": 8.753714757381881e-06, + "loss": 0.366, + "step": 11485 + }, + { + "epoch": 0.2528285274047986, + "grad_norm": 2.003737890316181, + "learning_rate": 8.753479274738973e-06, + "loss": 0.2966, + "step": 11486 + }, + { + "epoch": 0.25285053929121726, + "grad_norm": 1.2167697422566437, + "learning_rate": 8.753243773019209e-06, + "loss": 0.3672, + "step": 11487 + }, + { + "epoch": 0.2528725511776359, + "grad_norm": 1.1913936207553169, + "learning_rate": 8.753008252223785e-06, + "loss": 0.2611, + "step": 11488 + }, + { + "epoch": 0.2528945630640546, + "grad_norm": 1.872542478992494, + "learning_rate": 8.7527727123539e-06, + "loss": 0.403, + "step": 11489 + }, + { + "epoch": 0.25291657495047326, + "grad_norm": 0.6157353476461256, + "learning_rate": 8.75253715341075e-06, + "loss": 0.5029, + "step": 11490 + }, + { + "epoch": 0.2529385868368919, + "grad_norm": 1.44502987396935, + "learning_rate": 8.752301575395531e-06, + "loss": 0.3737, + "step": 11491 + }, + { + "epoch": 0.2529605987233106, + "grad_norm": 1.3249807689548814, + "learning_rate": 8.752065978309444e-06, + "loss": 0.3251, + "step": 11492 + }, + { + "epoch": 0.25298261060972926, + "grad_norm": 1.4857141547430848, + "learning_rate": 8.751830362153682e-06, + "loss": 0.3526, + "step": 11493 + }, + { + "epoch": 0.2530046224961479, + "grad_norm": 0.6247908639949605, + "learning_rate": 8.751594726929447e-06, + "loss": 0.5099, + "step": 11494 + }, + { + "epoch": 0.25302663438256656, + "grad_norm": 1.2540299795193508, + "learning_rate": 8.751359072637932e-06, + "loss": 0.3496, + "step": 11495 + }, + { + "epoch": 0.25304864626898527, + "grad_norm": 1.490997830970009, + "learning_rate": 8.75112339928034e-06, + "loss": 0.4612, + "step": 11496 + }, + { + "epoch": 0.2530706581554039, + "grad_norm": 0.6172108502609962, + "learning_rate": 8.750887706857864e-06, + "loss": 0.5269, + "step": 11497 + }, + { + "epoch": 0.25309267004182256, + "grad_norm": 1.2169355306407017, + "learning_rate": 8.750651995371705e-06, + "loss": 0.344, + "step": 11498 + }, + { + "epoch": 0.25311468192824127, + "grad_norm": 1.8929983449955246, + "learning_rate": 8.75041626482306e-06, + "loss": 0.3562, + "step": 11499 + }, + { + "epoch": 0.2531366938146599, + "grad_norm": 1.3912684361426224, + "learning_rate": 8.750180515213123e-06, + "loss": 0.3439, + "step": 11500 + }, + { + "epoch": 0.25315870570107857, + "grad_norm": 10.405218517821769, + "learning_rate": 8.749944746543099e-06, + "loss": 0.2925, + "step": 11501 + }, + { + "epoch": 0.2531807175874973, + "grad_norm": 1.2967041932655003, + "learning_rate": 8.749708958814183e-06, + "loss": 0.3046, + "step": 11502 + }, + { + "epoch": 0.2532027294739159, + "grad_norm": 1.304622514398217, + "learning_rate": 8.749473152027573e-06, + "loss": 0.2854, + "step": 11503 + }, + { + "epoch": 0.25322474136033457, + "grad_norm": 1.3233504382892347, + "learning_rate": 8.749237326184469e-06, + "loss": 0.418, + "step": 11504 + }, + { + "epoch": 0.2532467532467532, + "grad_norm": 1.4458784780254692, + "learning_rate": 8.749001481286069e-06, + "loss": 0.3178, + "step": 11505 + }, + { + "epoch": 0.2532687651331719, + "grad_norm": 2.02141929395381, + "learning_rate": 8.74876561733357e-06, + "loss": 0.3128, + "step": 11506 + }, + { + "epoch": 0.2532907770195906, + "grad_norm": 1.204631415575363, + "learning_rate": 8.748529734328173e-06, + "loss": 0.2774, + "step": 11507 + }, + { + "epoch": 0.2533127889060092, + "grad_norm": 1.2915434546330942, + "learning_rate": 8.748293832271075e-06, + "loss": 0.313, + "step": 11508 + }, + { + "epoch": 0.2533348007924279, + "grad_norm": 1.5124046733052945, + "learning_rate": 8.748057911163476e-06, + "loss": 0.3322, + "step": 11509 + }, + { + "epoch": 0.2533568126788466, + "grad_norm": 0.6402230906061853, + "learning_rate": 8.747821971006577e-06, + "loss": 0.5201, + "step": 11510 + }, + { + "epoch": 0.2533788245652652, + "grad_norm": 1.924298854986301, + "learning_rate": 8.747586011801574e-06, + "loss": 0.2679, + "step": 11511 + }, + { + "epoch": 0.25340083645168393, + "grad_norm": 1.3765585372936222, + "learning_rate": 8.747350033549667e-06, + "loss": 0.3318, + "step": 11512 + }, + { + "epoch": 0.2534228483381026, + "grad_norm": 1.137328077347322, + "learning_rate": 8.747114036252056e-06, + "loss": 0.3216, + "step": 11513 + }, + { + "epoch": 0.25344486022452123, + "grad_norm": 1.4011409723531212, + "learning_rate": 8.74687801990994e-06, + "loss": 0.4403, + "step": 11514 + }, + { + "epoch": 0.25346687211093993, + "grad_norm": 1.3014187670258661, + "learning_rate": 8.74664198452452e-06, + "loss": 0.368, + "step": 11515 + }, + { + "epoch": 0.2534888839973586, + "grad_norm": 1.710297741405078, + "learning_rate": 8.746405930096993e-06, + "loss": 0.323, + "step": 11516 + }, + { + "epoch": 0.25351089588377723, + "grad_norm": 1.2022409845285176, + "learning_rate": 8.746169856628561e-06, + "loss": 0.276, + "step": 11517 + }, + { + "epoch": 0.2535329077701959, + "grad_norm": 1.3320164944468427, + "learning_rate": 8.745933764120421e-06, + "loss": 0.2782, + "step": 11518 + }, + { + "epoch": 0.2535549196566146, + "grad_norm": 1.2037551918712557, + "learning_rate": 8.74569765257378e-06, + "loss": 0.3267, + "step": 11519 + }, + { + "epoch": 0.25357693154303323, + "grad_norm": 1.3775190915159583, + "learning_rate": 8.745461521989828e-06, + "loss": 0.3846, + "step": 11520 + }, + { + "epoch": 0.2535989434294519, + "grad_norm": 1.7995807323028108, + "learning_rate": 8.74522537236977e-06, + "loss": 0.3094, + "step": 11521 + }, + { + "epoch": 0.2536209553158706, + "grad_norm": 1.4079641039837798, + "learning_rate": 8.744989203714809e-06, + "loss": 0.4276, + "step": 11522 + }, + { + "epoch": 0.25364296720228924, + "grad_norm": 0.6544692125614988, + "learning_rate": 8.744753016026142e-06, + "loss": 0.5113, + "step": 11523 + }, + { + "epoch": 0.2536649790887079, + "grad_norm": 1.138975726988033, + "learning_rate": 8.744516809304968e-06, + "loss": 0.3063, + "step": 11524 + }, + { + "epoch": 0.2536869909751266, + "grad_norm": 1.2008310312046349, + "learning_rate": 8.74428058355249e-06, + "loss": 0.2871, + "step": 11525 + }, + { + "epoch": 0.25370900286154524, + "grad_norm": 1.47725354784324, + "learning_rate": 8.74404433876991e-06, + "loss": 0.3294, + "step": 11526 + }, + { + "epoch": 0.2537310147479639, + "grad_norm": 2.9843941214225516, + "learning_rate": 8.743808074958424e-06, + "loss": 0.4133, + "step": 11527 + }, + { + "epoch": 0.25375302663438254, + "grad_norm": 1.5251369924105214, + "learning_rate": 8.743571792119238e-06, + "loss": 0.3308, + "step": 11528 + }, + { + "epoch": 0.25377503852080124, + "grad_norm": 16.270321261031995, + "learning_rate": 8.74333549025355e-06, + "loss": 0.3538, + "step": 11529 + }, + { + "epoch": 0.2537970504072199, + "grad_norm": 1.4348239536480467, + "learning_rate": 8.74309916936256e-06, + "loss": 0.4635, + "step": 11530 + }, + { + "epoch": 0.25381906229363854, + "grad_norm": 1.4217218540790382, + "learning_rate": 8.742862829447471e-06, + "loss": 0.3725, + "step": 11531 + }, + { + "epoch": 0.25384107418005725, + "grad_norm": 1.4759139379086703, + "learning_rate": 8.742626470509485e-06, + "loss": 0.4123, + "step": 11532 + }, + { + "epoch": 0.2538630860664759, + "grad_norm": 1.681350905774467, + "learning_rate": 8.742390092549802e-06, + "loss": 0.3179, + "step": 11533 + }, + { + "epoch": 0.25388509795289455, + "grad_norm": 1.104518741862773, + "learning_rate": 8.742153695569621e-06, + "loss": 0.3008, + "step": 11534 + }, + { + "epoch": 0.25390710983931325, + "grad_norm": 1.9100760435602955, + "learning_rate": 8.74191727957015e-06, + "loss": 0.4052, + "step": 11535 + }, + { + "epoch": 0.2539291217257319, + "grad_norm": 1.6955660828787713, + "learning_rate": 8.741680844552582e-06, + "loss": 0.3082, + "step": 11536 + }, + { + "epoch": 0.25395113361215055, + "grad_norm": 1.2153576302241913, + "learning_rate": 8.741444390518126e-06, + "loss": 0.3393, + "step": 11537 + }, + { + "epoch": 0.25397314549856925, + "grad_norm": 1.5834875905136299, + "learning_rate": 8.74120791746798e-06, + "loss": 0.3508, + "step": 11538 + }, + { + "epoch": 0.2539951573849879, + "grad_norm": 1.684868120224257, + "learning_rate": 8.740971425403347e-06, + "loss": 0.2712, + "step": 11539 + }, + { + "epoch": 0.25401716927140655, + "grad_norm": 1.2157937665838432, + "learning_rate": 8.740734914325428e-06, + "loss": 0.3134, + "step": 11540 + }, + { + "epoch": 0.2540391811578252, + "grad_norm": 1.2815179480401229, + "learning_rate": 8.740498384235426e-06, + "loss": 0.3309, + "step": 11541 + }, + { + "epoch": 0.2540611930442439, + "grad_norm": 0.7154335249350755, + "learning_rate": 8.740261835134543e-06, + "loss": 0.5255, + "step": 11542 + }, + { + "epoch": 0.25408320493066255, + "grad_norm": 1.3566717573478453, + "learning_rate": 8.740025267023982e-06, + "loss": 0.3375, + "step": 11543 + }, + { + "epoch": 0.2541052168170812, + "grad_norm": 1.4397423797422582, + "learning_rate": 8.739788679904944e-06, + "loss": 0.4007, + "step": 11544 + }, + { + "epoch": 0.2541272287034999, + "grad_norm": 0.6227328735113866, + "learning_rate": 8.73955207377863e-06, + "loss": 0.498, + "step": 11545 + }, + { + "epoch": 0.25414924058991856, + "grad_norm": 1.6468925385944915, + "learning_rate": 8.739315448646248e-06, + "loss": 0.373, + "step": 11546 + }, + { + "epoch": 0.2541712524763372, + "grad_norm": 1.3435868599693506, + "learning_rate": 8.739078804508994e-06, + "loss": 0.3142, + "step": 11547 + }, + { + "epoch": 0.2541932643627559, + "grad_norm": 1.4790979453381172, + "learning_rate": 8.738842141368076e-06, + "loss": 0.3357, + "step": 11548 + }, + { + "epoch": 0.25421527624917456, + "grad_norm": 1.4016174556617291, + "learning_rate": 8.738605459224694e-06, + "loss": 0.308, + "step": 11549 + }, + { + "epoch": 0.2542372881355932, + "grad_norm": 1.4182620555672198, + "learning_rate": 8.738368758080051e-06, + "loss": 0.4078, + "step": 11550 + }, + { + "epoch": 0.2542593000220119, + "grad_norm": 1.7044626908313867, + "learning_rate": 8.73813203793535e-06, + "loss": 0.3603, + "step": 11551 + }, + { + "epoch": 0.25428131190843056, + "grad_norm": 1.9477875707340875, + "learning_rate": 8.737895298791797e-06, + "loss": 0.4929, + "step": 11552 + }, + { + "epoch": 0.2543033237948492, + "grad_norm": 1.3470069912984963, + "learning_rate": 8.737658540650591e-06, + "loss": 0.3549, + "step": 11553 + }, + { + "epoch": 0.25432533568126786, + "grad_norm": 2.36253913304245, + "learning_rate": 8.737421763512938e-06, + "loss": 0.434, + "step": 11554 + }, + { + "epoch": 0.25434734756768657, + "grad_norm": 1.8808068544835108, + "learning_rate": 8.73718496738004e-06, + "loss": 0.3325, + "step": 11555 + }, + { + "epoch": 0.2543693594541052, + "grad_norm": 0.7176720783986281, + "learning_rate": 8.736948152253102e-06, + "loss": 0.5202, + "step": 11556 + }, + { + "epoch": 0.25439137134052386, + "grad_norm": 0.6525180791735513, + "learning_rate": 8.736711318133326e-06, + "loss": 0.5113, + "step": 11557 + }, + { + "epoch": 0.25441338322694257, + "grad_norm": 1.9885077888836413, + "learning_rate": 8.736474465021919e-06, + "loss": 0.4451, + "step": 11558 + }, + { + "epoch": 0.2544353951133612, + "grad_norm": 1.3695400613787898, + "learning_rate": 8.736237592920079e-06, + "loss": 0.4026, + "step": 11559 + }, + { + "epoch": 0.25445740699977987, + "grad_norm": 0.6418393824598998, + "learning_rate": 8.736000701829014e-06, + "loss": 0.5326, + "step": 11560 + }, + { + "epoch": 0.25447941888619857, + "grad_norm": 0.6519548301482334, + "learning_rate": 8.735763791749926e-06, + "loss": 0.5114, + "step": 11561 + }, + { + "epoch": 0.2545014307726172, + "grad_norm": 1.2554960582472545, + "learning_rate": 8.735526862684023e-06, + "loss": 0.2463, + "step": 11562 + }, + { + "epoch": 0.25452344265903587, + "grad_norm": 1.3355741326606676, + "learning_rate": 8.735289914632506e-06, + "loss": 0.3143, + "step": 11563 + }, + { + "epoch": 0.2545454545454545, + "grad_norm": 1.3933340301939754, + "learning_rate": 8.735052947596579e-06, + "loss": 0.3652, + "step": 11564 + }, + { + "epoch": 0.2545674664318732, + "grad_norm": 1.2769936152068664, + "learning_rate": 8.734815961577446e-06, + "loss": 0.3295, + "step": 11565 + }, + { + "epoch": 0.2545894783182919, + "grad_norm": 1.2226122825711216, + "learning_rate": 8.734578956576312e-06, + "loss": 0.2637, + "step": 11566 + }, + { + "epoch": 0.2546114902047105, + "grad_norm": 1.515755372097521, + "learning_rate": 8.734341932594384e-06, + "loss": 0.2581, + "step": 11567 + }, + { + "epoch": 0.2546335020911292, + "grad_norm": 2.5870052469274034, + "learning_rate": 8.734104889632864e-06, + "loss": 0.4173, + "step": 11568 + }, + { + "epoch": 0.2546555139775479, + "grad_norm": 4.164218147735478, + "learning_rate": 8.733867827692958e-06, + "loss": 0.3177, + "step": 11569 + }, + { + "epoch": 0.2546775258639665, + "grad_norm": 1.2428151662376679, + "learning_rate": 8.73363074677587e-06, + "loss": 0.353, + "step": 11570 + }, + { + "epoch": 0.25469953775038523, + "grad_norm": 1.2013613650802497, + "learning_rate": 8.733393646882807e-06, + "loss": 0.3365, + "step": 11571 + }, + { + "epoch": 0.2547215496368039, + "grad_norm": 1.2842607566730808, + "learning_rate": 8.733156528014968e-06, + "loss": 0.331, + "step": 11572 + }, + { + "epoch": 0.25474356152322253, + "grad_norm": 1.4978930458406956, + "learning_rate": 8.732919390173566e-06, + "loss": 0.2711, + "step": 11573 + }, + { + "epoch": 0.25476557340964123, + "grad_norm": 1.595321078828892, + "learning_rate": 8.732682233359802e-06, + "loss": 0.3967, + "step": 11574 + }, + { + "epoch": 0.2547875852960599, + "grad_norm": 1.3178877888157332, + "learning_rate": 8.73244505757488e-06, + "loss": 0.4213, + "step": 11575 + }, + { + "epoch": 0.25480959718247853, + "grad_norm": 1.5380372441430905, + "learning_rate": 8.732207862820012e-06, + "loss": 0.3245, + "step": 11576 + }, + { + "epoch": 0.2548316090688972, + "grad_norm": 0.7888409981166515, + "learning_rate": 8.731970649096396e-06, + "loss": 0.557, + "step": 11577 + }, + { + "epoch": 0.2548536209553159, + "grad_norm": 0.7201303004365541, + "learning_rate": 8.73173341640524e-06, + "loss": 0.5232, + "step": 11578 + }, + { + "epoch": 0.25487563284173453, + "grad_norm": 1.6936199634664268, + "learning_rate": 8.731496164747752e-06, + "loss": 0.3528, + "step": 11579 + }, + { + "epoch": 0.2548976447281532, + "grad_norm": 1.5050170243864223, + "learning_rate": 8.731258894125137e-06, + "loss": 0.3226, + "step": 11580 + }, + { + "epoch": 0.2549196566145719, + "grad_norm": 1.4156328842554504, + "learning_rate": 8.731021604538598e-06, + "loss": 0.3404, + "step": 11581 + }, + { + "epoch": 0.25494166850099054, + "grad_norm": 1.3712131861848484, + "learning_rate": 8.730784295989344e-06, + "loss": 0.4169, + "step": 11582 + }, + { + "epoch": 0.2549636803874092, + "grad_norm": 1.2292793315633161, + "learning_rate": 8.73054696847858e-06, + "loss": 0.2561, + "step": 11583 + }, + { + "epoch": 0.2549856922738279, + "grad_norm": 2.5886652905553746, + "learning_rate": 8.730309622007511e-06, + "loss": 0.3557, + "step": 11584 + }, + { + "epoch": 0.25500770416024654, + "grad_norm": 1.2782861610019074, + "learning_rate": 8.730072256577347e-06, + "loss": 0.3693, + "step": 11585 + }, + { + "epoch": 0.2550297160466652, + "grad_norm": 0.8374866897527611, + "learning_rate": 8.72983487218929e-06, + "loss": 0.5548, + "step": 11586 + }, + { + "epoch": 0.25505172793308384, + "grad_norm": 1.3992474274851125, + "learning_rate": 8.729597468844551e-06, + "loss": 0.3794, + "step": 11587 + }, + { + "epoch": 0.25507373981950254, + "grad_norm": 1.2183910614067026, + "learning_rate": 8.729360046544332e-06, + "loss": 0.417, + "step": 11588 + }, + { + "epoch": 0.2550957517059212, + "grad_norm": 1.337153209645378, + "learning_rate": 8.729122605289845e-06, + "loss": 0.3454, + "step": 11589 + }, + { + "epoch": 0.25511776359233984, + "grad_norm": 1.220774075225007, + "learning_rate": 8.728885145082293e-06, + "loss": 0.3217, + "step": 11590 + }, + { + "epoch": 0.25513977547875855, + "grad_norm": 1.6680875836435909, + "learning_rate": 8.72864766592288e-06, + "loss": 0.2777, + "step": 11591 + }, + { + "epoch": 0.2551617873651772, + "grad_norm": 0.6786730920240637, + "learning_rate": 8.728410167812819e-06, + "loss": 0.504, + "step": 11592 + }, + { + "epoch": 0.25518379925159584, + "grad_norm": 1.4133922564173877, + "learning_rate": 8.728172650753317e-06, + "loss": 0.3388, + "step": 11593 + }, + { + "epoch": 0.25520581113801455, + "grad_norm": 1.899920001641976, + "learning_rate": 8.727935114745575e-06, + "loss": 0.3762, + "step": 11594 + }, + { + "epoch": 0.2552278230244332, + "grad_norm": 1.3975267385792223, + "learning_rate": 8.727697559790805e-06, + "loss": 0.3765, + "step": 11595 + }, + { + "epoch": 0.25524983491085185, + "grad_norm": 1.794831677492021, + "learning_rate": 8.727459985890215e-06, + "loss": 0.3805, + "step": 11596 + }, + { + "epoch": 0.25527184679727055, + "grad_norm": 1.8442068912344567, + "learning_rate": 8.727222393045009e-06, + "loss": 0.3724, + "step": 11597 + }, + { + "epoch": 0.2552938586836892, + "grad_norm": 1.4670280921594157, + "learning_rate": 8.726984781256398e-06, + "loss": 0.3032, + "step": 11598 + }, + { + "epoch": 0.25531587057010785, + "grad_norm": 1.3796461173128827, + "learning_rate": 8.726747150525586e-06, + "loss": 0.3142, + "step": 11599 + }, + { + "epoch": 0.2553378824565265, + "grad_norm": 1.5219049239740727, + "learning_rate": 8.726509500853783e-06, + "loss": 0.3018, + "step": 11600 + }, + { + "epoch": 0.2553598943429452, + "grad_norm": 1.749784627796653, + "learning_rate": 8.726271832242199e-06, + "loss": 0.3746, + "step": 11601 + }, + { + "epoch": 0.25538190622936385, + "grad_norm": 1.3869401352742268, + "learning_rate": 8.726034144692037e-06, + "loss": 0.3848, + "step": 11602 + }, + { + "epoch": 0.2554039181157825, + "grad_norm": 2.635545643466241, + "learning_rate": 8.72579643820451e-06, + "loss": 0.3089, + "step": 11603 + }, + { + "epoch": 0.2554259300022012, + "grad_norm": 1.4364238691051592, + "learning_rate": 8.725558712780822e-06, + "loss": 0.3001, + "step": 11604 + }, + { + "epoch": 0.25544794188861986, + "grad_norm": 1.2400091055717426, + "learning_rate": 8.725320968422183e-06, + "loss": 0.2984, + "step": 11605 + }, + { + "epoch": 0.2554699537750385, + "grad_norm": 1.1068936643090033, + "learning_rate": 8.725083205129803e-06, + "loss": 0.3519, + "step": 11606 + }, + { + "epoch": 0.2554919656614572, + "grad_norm": 1.1655716291275113, + "learning_rate": 8.724845422904887e-06, + "loss": 0.2741, + "step": 11607 + }, + { + "epoch": 0.25551397754787586, + "grad_norm": 1.4629681072271645, + "learning_rate": 8.724607621748645e-06, + "loss": 0.2755, + "step": 11608 + }, + { + "epoch": 0.2555359894342945, + "grad_norm": 0.7455564321020859, + "learning_rate": 8.724369801662287e-06, + "loss": 0.5489, + "step": 11609 + }, + { + "epoch": 0.2555580013207132, + "grad_norm": 1.6174338542351812, + "learning_rate": 8.72413196264702e-06, + "loss": 0.3529, + "step": 11610 + }, + { + "epoch": 0.25558001320713186, + "grad_norm": 1.3882103325873596, + "learning_rate": 8.723894104704053e-06, + "loss": 0.3146, + "step": 11611 + }, + { + "epoch": 0.2556020250935505, + "grad_norm": 1.6454336901455673, + "learning_rate": 8.723656227834596e-06, + "loss": 0.3846, + "step": 11612 + }, + { + "epoch": 0.25562403697996916, + "grad_norm": 1.4560365559025317, + "learning_rate": 8.723418332039857e-06, + "loss": 0.3092, + "step": 11613 + }, + { + "epoch": 0.25564604886638786, + "grad_norm": 1.525460250425856, + "learning_rate": 8.723180417321046e-06, + "loss": 0.3886, + "step": 11614 + }, + { + "epoch": 0.2556680607528065, + "grad_norm": 0.630254039431861, + "learning_rate": 8.722942483679371e-06, + "loss": 0.5223, + "step": 11615 + }, + { + "epoch": 0.25569007263922516, + "grad_norm": 2.3007959834053, + "learning_rate": 8.722704531116043e-06, + "loss": 0.3133, + "step": 11616 + }, + { + "epoch": 0.25571208452564387, + "grad_norm": 1.614360192546002, + "learning_rate": 8.722466559632268e-06, + "loss": 0.2753, + "step": 11617 + }, + { + "epoch": 0.2557340964120625, + "grad_norm": 1.4710005219212114, + "learning_rate": 8.722228569229257e-06, + "loss": 0.3665, + "step": 11618 + }, + { + "epoch": 0.25575610829848117, + "grad_norm": 1.391507458064226, + "learning_rate": 8.721990559908222e-06, + "loss": 0.387, + "step": 11619 + }, + { + "epoch": 0.25577812018489987, + "grad_norm": 1.4145141969396609, + "learning_rate": 8.72175253167037e-06, + "loss": 0.3382, + "step": 11620 + }, + { + "epoch": 0.2558001320713185, + "grad_norm": 0.6877231569972113, + "learning_rate": 8.721514484516914e-06, + "loss": 0.5308, + "step": 11621 + }, + { + "epoch": 0.25582214395773717, + "grad_norm": 1.326010997405747, + "learning_rate": 8.72127641844906e-06, + "loss": 0.3022, + "step": 11622 + }, + { + "epoch": 0.2558441558441558, + "grad_norm": 1.5596352910853568, + "learning_rate": 8.721038333468019e-06, + "loss": 0.3486, + "step": 11623 + }, + { + "epoch": 0.2558661677305745, + "grad_norm": 1.362382626230372, + "learning_rate": 8.720800229575004e-06, + "loss": 0.3837, + "step": 11624 + }, + { + "epoch": 0.25588817961699317, + "grad_norm": 1.2357613707123574, + "learning_rate": 8.72056210677122e-06, + "loss": 0.3434, + "step": 11625 + }, + { + "epoch": 0.2559101915034118, + "grad_norm": 1.3640610732954792, + "learning_rate": 8.72032396505788e-06, + "loss": 0.3318, + "step": 11626 + }, + { + "epoch": 0.2559322033898305, + "grad_norm": 0.6083423673981376, + "learning_rate": 8.720085804436194e-06, + "loss": 0.5196, + "step": 11627 + }, + { + "epoch": 0.2559542152762492, + "grad_norm": 0.6210718469161103, + "learning_rate": 8.719847624907374e-06, + "loss": 0.5311, + "step": 11628 + }, + { + "epoch": 0.2559762271626678, + "grad_norm": 1.9327415317325685, + "learning_rate": 8.719609426472629e-06, + "loss": 0.3772, + "step": 11629 + }, + { + "epoch": 0.25599823904908653, + "grad_norm": 1.3060088097998737, + "learning_rate": 8.71937120913317e-06, + "loss": 0.2992, + "step": 11630 + }, + { + "epoch": 0.2560202509355052, + "grad_norm": 1.8054803075659294, + "learning_rate": 8.719132972890207e-06, + "loss": 0.4253, + "step": 11631 + }, + { + "epoch": 0.2560422628219238, + "grad_norm": 1.51123942858246, + "learning_rate": 8.718894717744953e-06, + "loss": 0.3319, + "step": 11632 + }, + { + "epoch": 0.25606427470834253, + "grad_norm": 1.725442849505897, + "learning_rate": 8.718656443698617e-06, + "loss": 0.3248, + "step": 11633 + }, + { + "epoch": 0.2560862865947612, + "grad_norm": 1.280048238153828, + "learning_rate": 8.718418150752409e-06, + "loss": 0.3876, + "step": 11634 + }, + { + "epoch": 0.25610829848117983, + "grad_norm": 0.640215297321096, + "learning_rate": 8.718179838907543e-06, + "loss": 0.5109, + "step": 11635 + }, + { + "epoch": 0.2561303103675985, + "grad_norm": 0.6525776805522163, + "learning_rate": 8.717941508165227e-06, + "loss": 0.5268, + "step": 11636 + }, + { + "epoch": 0.2561523222540172, + "grad_norm": 3.0477723769793457, + "learning_rate": 8.717703158526676e-06, + "loss": 0.33, + "step": 11637 + }, + { + "epoch": 0.25617433414043583, + "grad_norm": 1.4145424012925762, + "learning_rate": 8.717464789993097e-06, + "loss": 0.2726, + "step": 11638 + }, + { + "epoch": 0.2561963460268545, + "grad_norm": 1.7895037287364204, + "learning_rate": 8.717226402565706e-06, + "loss": 0.3095, + "step": 11639 + }, + { + "epoch": 0.2562183579132732, + "grad_norm": 1.3790013105653849, + "learning_rate": 8.716987996245711e-06, + "loss": 0.3059, + "step": 11640 + }, + { + "epoch": 0.25624036979969184, + "grad_norm": 1.3478140241995291, + "learning_rate": 8.716749571034327e-06, + "loss": 0.4311, + "step": 11641 + }, + { + "epoch": 0.2562623816861105, + "grad_norm": 0.6039646300205771, + "learning_rate": 8.716511126932763e-06, + "loss": 0.5185, + "step": 11642 + }, + { + "epoch": 0.2562843935725292, + "grad_norm": 1.3650541024850376, + "learning_rate": 8.71627266394223e-06, + "loss": 0.3108, + "step": 11643 + }, + { + "epoch": 0.25630640545894784, + "grad_norm": 1.5879801901655772, + "learning_rate": 8.716034182063946e-06, + "loss": 0.349, + "step": 11644 + }, + { + "epoch": 0.2563284173453665, + "grad_norm": 1.5945059388688902, + "learning_rate": 8.715795681299115e-06, + "loss": 0.2175, + "step": 11645 + }, + { + "epoch": 0.25635042923178514, + "grad_norm": 2.19572284913651, + "learning_rate": 8.715557161648956e-06, + "loss": 0.3835, + "step": 11646 + }, + { + "epoch": 0.25637244111820384, + "grad_norm": 1.4256669187368112, + "learning_rate": 8.715318623114675e-06, + "loss": 0.4025, + "step": 11647 + }, + { + "epoch": 0.2563944530046225, + "grad_norm": 1.5829211707318755, + "learning_rate": 8.715080065697489e-06, + "loss": 0.3423, + "step": 11648 + }, + { + "epoch": 0.25641646489104114, + "grad_norm": 1.597731710255643, + "learning_rate": 8.71484148939861e-06, + "loss": 0.3225, + "step": 11649 + }, + { + "epoch": 0.25643847677745985, + "grad_norm": 2.478292114820379, + "learning_rate": 8.71460289421925e-06, + "loss": 0.3955, + "step": 11650 + }, + { + "epoch": 0.2564604886638785, + "grad_norm": 1.2839981541845933, + "learning_rate": 8.71436428016062e-06, + "loss": 0.2747, + "step": 11651 + }, + { + "epoch": 0.25648250055029714, + "grad_norm": 1.3988627998577632, + "learning_rate": 8.714125647223937e-06, + "loss": 0.3708, + "step": 11652 + }, + { + "epoch": 0.25650451243671585, + "grad_norm": 1.182293250022676, + "learning_rate": 8.713886995410407e-06, + "loss": 0.2773, + "step": 11653 + }, + { + "epoch": 0.2565265243231345, + "grad_norm": 1.2369687373274678, + "learning_rate": 8.71364832472125e-06, + "loss": 0.3177, + "step": 11654 + }, + { + "epoch": 0.25654853620955315, + "grad_norm": 2.2709312795685483, + "learning_rate": 8.713409635157674e-06, + "loss": 0.3151, + "step": 11655 + }, + { + "epoch": 0.25657054809597185, + "grad_norm": 1.6678444627769242, + "learning_rate": 8.713170926720895e-06, + "loss": 0.517, + "step": 11656 + }, + { + "epoch": 0.2565925599823905, + "grad_norm": 1.304022634912106, + "learning_rate": 8.712932199412124e-06, + "loss": 0.3148, + "step": 11657 + }, + { + "epoch": 0.25661457186880915, + "grad_norm": 2.978884495369667, + "learning_rate": 8.712693453232576e-06, + "loss": 0.4877, + "step": 11658 + }, + { + "epoch": 0.2566365837552278, + "grad_norm": 0.6659654100806169, + "learning_rate": 8.712454688183464e-06, + "loss": 0.5136, + "step": 11659 + }, + { + "epoch": 0.2566585956416465, + "grad_norm": 0.6529546455890706, + "learning_rate": 8.712215904266002e-06, + "loss": 0.5166, + "step": 11660 + }, + { + "epoch": 0.25668060752806515, + "grad_norm": 1.7304753684910956, + "learning_rate": 8.7119771014814e-06, + "loss": 0.3587, + "step": 11661 + }, + { + "epoch": 0.2567026194144838, + "grad_norm": 1.6075836891311293, + "learning_rate": 8.711738279830879e-06, + "loss": 0.3334, + "step": 11662 + }, + { + "epoch": 0.2567246313009025, + "grad_norm": 3.4215031692014817, + "learning_rate": 8.711499439315647e-06, + "loss": 0.4042, + "step": 11663 + }, + { + "epoch": 0.25674664318732116, + "grad_norm": 2.237154528383435, + "learning_rate": 8.71126057993692e-06, + "loss": 0.3879, + "step": 11664 + }, + { + "epoch": 0.2567686550737398, + "grad_norm": 1.2256235378328364, + "learning_rate": 8.711021701695911e-06, + "loss": 0.3249, + "step": 11665 + }, + { + "epoch": 0.2567906669601585, + "grad_norm": 1.0945163357293834, + "learning_rate": 8.710782804593834e-06, + "loss": 0.2987, + "step": 11666 + }, + { + "epoch": 0.25681267884657716, + "grad_norm": 0.6654736057332545, + "learning_rate": 8.710543888631904e-06, + "loss": 0.5256, + "step": 11667 + }, + { + "epoch": 0.2568346907329958, + "grad_norm": 1.7562230016905762, + "learning_rate": 8.710304953811336e-06, + "loss": 0.3376, + "step": 11668 + }, + { + "epoch": 0.25685670261941446, + "grad_norm": 1.7721510383397006, + "learning_rate": 8.710066000133343e-06, + "loss": 0.3276, + "step": 11669 + }, + { + "epoch": 0.25687871450583316, + "grad_norm": 3.985070387842577, + "learning_rate": 8.709827027599139e-06, + "loss": 0.3468, + "step": 11670 + }, + { + "epoch": 0.2569007263922518, + "grad_norm": 1.5584385772580203, + "learning_rate": 8.709588036209941e-06, + "loss": 0.3551, + "step": 11671 + }, + { + "epoch": 0.25692273827867046, + "grad_norm": 1.4883509732952735, + "learning_rate": 8.70934902596696e-06, + "loss": 0.3452, + "step": 11672 + }, + { + "epoch": 0.25694475016508916, + "grad_norm": 1.8699102777041734, + "learning_rate": 8.709109996871414e-06, + "loss": 0.2824, + "step": 11673 + }, + { + "epoch": 0.2569667620515078, + "grad_norm": 1.626502669482284, + "learning_rate": 8.708870948924518e-06, + "loss": 0.332, + "step": 11674 + }, + { + "epoch": 0.25698877393792646, + "grad_norm": 1.2181512496280302, + "learning_rate": 8.708631882127484e-06, + "loss": 0.3021, + "step": 11675 + }, + { + "epoch": 0.25701078582434517, + "grad_norm": 1.3701912743186038, + "learning_rate": 8.708392796481529e-06, + "loss": 0.2787, + "step": 11676 + }, + { + "epoch": 0.2570327977107638, + "grad_norm": 1.744332009429105, + "learning_rate": 8.708153691987866e-06, + "loss": 0.3365, + "step": 11677 + }, + { + "epoch": 0.25705480959718247, + "grad_norm": 1.3723677823925693, + "learning_rate": 8.707914568647716e-06, + "loss": 0.4482, + "step": 11678 + }, + { + "epoch": 0.25707682148360117, + "grad_norm": 1.6388794129969724, + "learning_rate": 8.707675426462286e-06, + "loss": 0.4156, + "step": 11679 + }, + { + "epoch": 0.2570988333700198, + "grad_norm": 1.6616541054080787, + "learning_rate": 8.707436265432798e-06, + "loss": 0.3717, + "step": 11680 + }, + { + "epoch": 0.25712084525643847, + "grad_norm": 1.4950521701744321, + "learning_rate": 8.707197085560465e-06, + "loss": 0.3284, + "step": 11681 + }, + { + "epoch": 0.2571428571428571, + "grad_norm": 1.6713741782779783, + "learning_rate": 8.706957886846503e-06, + "loss": 0.3512, + "step": 11682 + }, + { + "epoch": 0.2571648690292758, + "grad_norm": 1.3217419506305443, + "learning_rate": 8.70671866929213e-06, + "loss": 0.3358, + "step": 11683 + }, + { + "epoch": 0.25718688091569447, + "grad_norm": 1.6108954753491236, + "learning_rate": 8.706479432898556e-06, + "loss": 0.4259, + "step": 11684 + }, + { + "epoch": 0.2572088928021131, + "grad_norm": 1.832993305117282, + "learning_rate": 8.706240177667003e-06, + "loss": 0.3301, + "step": 11685 + }, + { + "epoch": 0.2572309046885318, + "grad_norm": 1.568356063563383, + "learning_rate": 8.706000903598681e-06, + "loss": 0.4093, + "step": 11686 + }, + { + "epoch": 0.2572529165749505, + "grad_norm": 0.6910313920209742, + "learning_rate": 8.705761610694813e-06, + "loss": 0.5203, + "step": 11687 + }, + { + "epoch": 0.2572749284613691, + "grad_norm": 1.6348356228621572, + "learning_rate": 8.70552229895661e-06, + "loss": 0.398, + "step": 11688 + }, + { + "epoch": 0.25729694034778783, + "grad_norm": 1.31408817933813, + "learning_rate": 8.70528296838529e-06, + "loss": 0.3369, + "step": 11689 + }, + { + "epoch": 0.2573189522342065, + "grad_norm": 1.6927465061835838, + "learning_rate": 8.70504361898207e-06, + "loss": 0.3836, + "step": 11690 + }, + { + "epoch": 0.2573409641206251, + "grad_norm": 0.638921854388472, + "learning_rate": 8.704804250748164e-06, + "loss": 0.5122, + "step": 11691 + }, + { + "epoch": 0.25736297600704383, + "grad_norm": 1.4243354837691569, + "learning_rate": 8.704564863684793e-06, + "loss": 0.2947, + "step": 11692 + }, + { + "epoch": 0.2573849878934625, + "grad_norm": 1.376889891701379, + "learning_rate": 8.70432545779317e-06, + "loss": 0.2581, + "step": 11693 + }, + { + "epoch": 0.25740699977988113, + "grad_norm": 1.1728302184002364, + "learning_rate": 8.704086033074513e-06, + "loss": 0.3923, + "step": 11694 + }, + { + "epoch": 0.2574290116662998, + "grad_norm": 1.7717018885613367, + "learning_rate": 8.703846589530038e-06, + "loss": 0.3883, + "step": 11695 + }, + { + "epoch": 0.2574510235527185, + "grad_norm": 0.6403615644457147, + "learning_rate": 8.703607127160963e-06, + "loss": 0.5215, + "step": 11696 + }, + { + "epoch": 0.25747303543913713, + "grad_norm": 1.7160547616038186, + "learning_rate": 8.703367645968505e-06, + "loss": 0.2773, + "step": 11697 + }, + { + "epoch": 0.2574950473255558, + "grad_norm": 0.6327851864257504, + "learning_rate": 8.703128145953882e-06, + "loss": 0.5365, + "step": 11698 + }, + { + "epoch": 0.2575170592119745, + "grad_norm": 1.4774582645473648, + "learning_rate": 8.702888627118308e-06, + "loss": 0.3443, + "step": 11699 + }, + { + "epoch": 0.25753907109839314, + "grad_norm": 1.2416509948104208, + "learning_rate": 8.702649089463003e-06, + "loss": 0.3379, + "step": 11700 + }, + { + "epoch": 0.2575610829848118, + "grad_norm": 1.754250644077933, + "learning_rate": 8.702409532989185e-06, + "loss": 0.3838, + "step": 11701 + }, + { + "epoch": 0.2575830948712305, + "grad_norm": 1.5715358104580828, + "learning_rate": 8.702169957698069e-06, + "loss": 0.379, + "step": 11702 + }, + { + "epoch": 0.25760510675764914, + "grad_norm": 0.6389981295236745, + "learning_rate": 8.701930363590876e-06, + "loss": 0.5046, + "step": 11703 + }, + { + "epoch": 0.2576271186440678, + "grad_norm": 1.8754714842935845, + "learning_rate": 8.701690750668819e-06, + "loss": 0.3769, + "step": 11704 + }, + { + "epoch": 0.25764913053048644, + "grad_norm": 1.7186035124676555, + "learning_rate": 8.701451118933122e-06, + "loss": 0.3963, + "step": 11705 + }, + { + "epoch": 0.25767114241690514, + "grad_norm": 1.607069885966448, + "learning_rate": 8.701211468384995e-06, + "loss": 0.2674, + "step": 11706 + }, + { + "epoch": 0.2576931543033238, + "grad_norm": 0.6339757435739111, + "learning_rate": 8.700971799025663e-06, + "loss": 0.5196, + "step": 11707 + }, + { + "epoch": 0.25771516618974244, + "grad_norm": 1.8095304453403278, + "learning_rate": 8.700732110856342e-06, + "loss": 0.3721, + "step": 11708 + }, + { + "epoch": 0.25773717807616114, + "grad_norm": 0.6155370882434571, + "learning_rate": 8.70049240387825e-06, + "loss": 0.5126, + "step": 11709 + }, + { + "epoch": 0.2577591899625798, + "grad_norm": 0.6108333358968915, + "learning_rate": 8.700252678092606e-06, + "loss": 0.5063, + "step": 11710 + }, + { + "epoch": 0.25778120184899844, + "grad_norm": 1.708719655482932, + "learning_rate": 8.700012933500626e-06, + "loss": 0.3091, + "step": 11711 + }, + { + "epoch": 0.25780321373541715, + "grad_norm": 1.4234633874647196, + "learning_rate": 8.69977317010353e-06, + "loss": 0.4344, + "step": 11712 + }, + { + "epoch": 0.2578252256218358, + "grad_norm": 1.6101243612660672, + "learning_rate": 8.699533387902537e-06, + "loss": 0.3733, + "step": 11713 + }, + { + "epoch": 0.25784723750825445, + "grad_norm": 1.9257892949440734, + "learning_rate": 8.699293586898866e-06, + "loss": 0.3727, + "step": 11714 + }, + { + "epoch": 0.25786924939467315, + "grad_norm": 0.6251539832653553, + "learning_rate": 8.699053767093734e-06, + "loss": 0.5194, + "step": 11715 + }, + { + "epoch": 0.2578912612810918, + "grad_norm": 2.2156943308053747, + "learning_rate": 8.698813928488361e-06, + "loss": 0.2761, + "step": 11716 + }, + { + "epoch": 0.25791327316751045, + "grad_norm": 1.6247700763323378, + "learning_rate": 8.698574071083966e-06, + "loss": 0.3303, + "step": 11717 + }, + { + "epoch": 0.2579352850539291, + "grad_norm": 1.761704589451432, + "learning_rate": 8.698334194881768e-06, + "loss": 0.3135, + "step": 11718 + }, + { + "epoch": 0.2579572969403478, + "grad_norm": 1.3319253717122983, + "learning_rate": 8.698094299882986e-06, + "loss": 0.4296, + "step": 11719 + }, + { + "epoch": 0.25797930882676645, + "grad_norm": 1.918231367153947, + "learning_rate": 8.69785438608884e-06, + "loss": 0.3929, + "step": 11720 + }, + { + "epoch": 0.2580013207131851, + "grad_norm": 1.2755047397226629, + "learning_rate": 8.697614453500548e-06, + "loss": 0.3615, + "step": 11721 + }, + { + "epoch": 0.2580233325996038, + "grad_norm": 1.982945898496777, + "learning_rate": 8.697374502119332e-06, + "loss": 0.4034, + "step": 11722 + }, + { + "epoch": 0.25804534448602245, + "grad_norm": 1.6322758707286138, + "learning_rate": 8.697134531946408e-06, + "loss": 0.3229, + "step": 11723 + }, + { + "epoch": 0.2580673563724411, + "grad_norm": 1.3952299209229087, + "learning_rate": 8.696894542982997e-06, + "loss": 0.3373, + "step": 11724 + }, + { + "epoch": 0.2580893682588598, + "grad_norm": 1.9818224772022937, + "learning_rate": 8.69665453523032e-06, + "loss": 0.3493, + "step": 11725 + }, + { + "epoch": 0.25811138014527846, + "grad_norm": 1.9278716528804205, + "learning_rate": 8.696414508689594e-06, + "loss": 0.2932, + "step": 11726 + }, + { + "epoch": 0.2581333920316971, + "grad_norm": 1.3761394061107652, + "learning_rate": 8.696174463362044e-06, + "loss": 0.317, + "step": 11727 + }, + { + "epoch": 0.25815540391811576, + "grad_norm": 1.3889792570270585, + "learning_rate": 8.695934399248884e-06, + "loss": 0.3268, + "step": 11728 + }, + { + "epoch": 0.25817741580453446, + "grad_norm": 1.792733425510024, + "learning_rate": 8.695694316351338e-06, + "loss": 0.3056, + "step": 11729 + }, + { + "epoch": 0.2581994276909531, + "grad_norm": 1.3241862327529976, + "learning_rate": 8.695454214670625e-06, + "loss": 0.3143, + "step": 11730 + }, + { + "epoch": 0.25822143957737176, + "grad_norm": 1.3407715808206098, + "learning_rate": 8.695214094207966e-06, + "loss": 0.3649, + "step": 11731 + }, + { + "epoch": 0.25824345146379046, + "grad_norm": 1.3752647321704297, + "learning_rate": 8.69497395496458e-06, + "loss": 0.365, + "step": 11732 + }, + { + "epoch": 0.2582654633502091, + "grad_norm": 1.074191941082112, + "learning_rate": 8.69473379694169e-06, + "loss": 0.2658, + "step": 11733 + }, + { + "epoch": 0.25828747523662776, + "grad_norm": 1.4771477602431107, + "learning_rate": 8.694493620140514e-06, + "loss": 0.3113, + "step": 11734 + }, + { + "epoch": 0.25830948712304647, + "grad_norm": 1.2743880188766643, + "learning_rate": 8.694253424562274e-06, + "loss": 0.3648, + "step": 11735 + }, + { + "epoch": 0.2583314990094651, + "grad_norm": 1.5462207771095613, + "learning_rate": 8.69401321020819e-06, + "loss": 0.4314, + "step": 11736 + }, + { + "epoch": 0.25835351089588376, + "grad_norm": 1.2299852442444232, + "learning_rate": 8.693772977079482e-06, + "loss": 0.2748, + "step": 11737 + }, + { + "epoch": 0.25837552278230247, + "grad_norm": 1.6204511331461708, + "learning_rate": 8.693532725177375e-06, + "loss": 0.3687, + "step": 11738 + }, + { + "epoch": 0.2583975346687211, + "grad_norm": 1.7857802328602952, + "learning_rate": 8.693292454503085e-06, + "loss": 0.3531, + "step": 11739 + }, + { + "epoch": 0.25841954655513977, + "grad_norm": 1.5440255258380389, + "learning_rate": 8.693052165057836e-06, + "loss": 0.3655, + "step": 11740 + }, + { + "epoch": 0.2584415584415584, + "grad_norm": 1.4505044981472757, + "learning_rate": 8.692811856842848e-06, + "loss": 0.3544, + "step": 11741 + }, + { + "epoch": 0.2584635703279771, + "grad_norm": 1.4486540204416478, + "learning_rate": 8.692571529859345e-06, + "loss": 0.3977, + "step": 11742 + }, + { + "epoch": 0.25848558221439577, + "grad_norm": 1.614118948784364, + "learning_rate": 8.692331184108547e-06, + "loss": 0.3831, + "step": 11743 + }, + { + "epoch": 0.2585075941008144, + "grad_norm": 2.3608642619921105, + "learning_rate": 8.692090819591672e-06, + "loss": 0.3613, + "step": 11744 + }, + { + "epoch": 0.2585296059872331, + "grad_norm": 1.3525238561986654, + "learning_rate": 8.691850436309948e-06, + "loss": 0.3971, + "step": 11745 + }, + { + "epoch": 0.2585516178736518, + "grad_norm": 1.2053885404905227, + "learning_rate": 8.69161003426459e-06, + "loss": 0.3459, + "step": 11746 + }, + { + "epoch": 0.2585736297600704, + "grad_norm": 1.595104622264367, + "learning_rate": 8.691369613456825e-06, + "loss": 0.4161, + "step": 11747 + }, + { + "epoch": 0.2585956416464891, + "grad_norm": 1.3214428283898214, + "learning_rate": 8.691129173887876e-06, + "loss": 0.3052, + "step": 11748 + }, + { + "epoch": 0.2586176535329078, + "grad_norm": 1.2884250562885045, + "learning_rate": 8.690888715558959e-06, + "loss": 0.3254, + "step": 11749 + }, + { + "epoch": 0.2586396654193264, + "grad_norm": 1.513073321120535, + "learning_rate": 8.6906482384713e-06, + "loss": 0.3151, + "step": 11750 + }, + { + "epoch": 0.2586616773057451, + "grad_norm": 0.6518337949453727, + "learning_rate": 8.690407742626121e-06, + "loss": 0.5083, + "step": 11751 + }, + { + "epoch": 0.2586836891921638, + "grad_norm": 1.5313756437821298, + "learning_rate": 8.690167228024644e-06, + "loss": 0.4045, + "step": 11752 + }, + { + "epoch": 0.25870570107858243, + "grad_norm": 1.0925783102471007, + "learning_rate": 8.68992669466809e-06, + "loss": 0.3683, + "step": 11753 + }, + { + "epoch": 0.2587277129650011, + "grad_norm": 1.4441915490576644, + "learning_rate": 8.689686142557686e-06, + "loss": 0.3829, + "step": 11754 + }, + { + "epoch": 0.2587497248514198, + "grad_norm": 1.4392049967280993, + "learning_rate": 8.689445571694649e-06, + "loss": 0.3751, + "step": 11755 + }, + { + "epoch": 0.25877173673783843, + "grad_norm": 1.2754505861191288, + "learning_rate": 8.689204982080202e-06, + "loss": 0.248, + "step": 11756 + }, + { + "epoch": 0.2587937486242571, + "grad_norm": 1.374154050459132, + "learning_rate": 8.688964373715572e-06, + "loss": 0.4072, + "step": 11757 + }, + { + "epoch": 0.2588157605106758, + "grad_norm": 1.5121681161657283, + "learning_rate": 8.68872374660198e-06, + "loss": 0.335, + "step": 11758 + }, + { + "epoch": 0.25883777239709443, + "grad_norm": 0.8809171685212228, + "learning_rate": 8.688483100740647e-06, + "loss": 0.2483, + "step": 11759 + }, + { + "epoch": 0.2588597842835131, + "grad_norm": 1.6084915416917478, + "learning_rate": 8.688242436132802e-06, + "loss": 0.4377, + "step": 11760 + }, + { + "epoch": 0.2588817961699318, + "grad_norm": 0.6483631755030579, + "learning_rate": 8.688001752779661e-06, + "loss": 0.5404, + "step": 11761 + }, + { + "epoch": 0.25890380805635044, + "grad_norm": 1.387267962296714, + "learning_rate": 8.687761050682449e-06, + "loss": 0.3066, + "step": 11762 + }, + { + "epoch": 0.2589258199427691, + "grad_norm": 1.2591232674381603, + "learning_rate": 8.687520329842391e-06, + "loss": 0.2974, + "step": 11763 + }, + { + "epoch": 0.25894783182918774, + "grad_norm": 1.29758076794026, + "learning_rate": 8.687279590260711e-06, + "loss": 0.3137, + "step": 11764 + }, + { + "epoch": 0.25896984371560644, + "grad_norm": 1.5563801751573598, + "learning_rate": 8.687038831938633e-06, + "loss": 0.3594, + "step": 11765 + }, + { + "epoch": 0.2589918556020251, + "grad_norm": 0.5986091179021249, + "learning_rate": 8.686798054877377e-06, + "loss": 0.4807, + "step": 11766 + }, + { + "epoch": 0.25901386748844374, + "grad_norm": 0.6169044639220602, + "learning_rate": 8.68655725907817e-06, + "loss": 0.5047, + "step": 11767 + }, + { + "epoch": 0.25903587937486244, + "grad_norm": 1.5682810757212948, + "learning_rate": 8.686316444542232e-06, + "loss": 0.2896, + "step": 11768 + }, + { + "epoch": 0.2590578912612811, + "grad_norm": 1.3780735241136595, + "learning_rate": 8.686075611270794e-06, + "loss": 0.3615, + "step": 11769 + }, + { + "epoch": 0.25907990314769974, + "grad_norm": 1.6402223300844694, + "learning_rate": 8.685834759265073e-06, + "loss": 0.4132, + "step": 11770 + }, + { + "epoch": 0.25910191503411845, + "grad_norm": 1.402567190827925, + "learning_rate": 8.685593888526297e-06, + "loss": 0.4261, + "step": 11771 + }, + { + "epoch": 0.2591239269205371, + "grad_norm": 1.5392545906914137, + "learning_rate": 8.685352999055688e-06, + "loss": 0.3706, + "step": 11772 + }, + { + "epoch": 0.25914593880695574, + "grad_norm": 1.3975905014909769, + "learning_rate": 8.685112090854472e-06, + "loss": 0.3935, + "step": 11773 + }, + { + "epoch": 0.25916795069337445, + "grad_norm": 1.3334481540046772, + "learning_rate": 8.684871163923873e-06, + "loss": 0.3389, + "step": 11774 + }, + { + "epoch": 0.2591899625797931, + "grad_norm": 1.3377230789314318, + "learning_rate": 8.684630218265115e-06, + "loss": 0.2253, + "step": 11775 + }, + { + "epoch": 0.25921197446621175, + "grad_norm": 1.4507226009367242, + "learning_rate": 8.684389253879423e-06, + "loss": 0.2991, + "step": 11776 + }, + { + "epoch": 0.2592339863526304, + "grad_norm": 1.568058073385748, + "learning_rate": 8.68414827076802e-06, + "loss": 0.2985, + "step": 11777 + }, + { + "epoch": 0.2592559982390491, + "grad_norm": 2.5435785051184316, + "learning_rate": 8.683907268932135e-06, + "loss": 0.3311, + "step": 11778 + }, + { + "epoch": 0.25927801012546775, + "grad_norm": 0.6152044746073495, + "learning_rate": 8.683666248372988e-06, + "loss": 0.4877, + "step": 11779 + }, + { + "epoch": 0.2593000220118864, + "grad_norm": 1.527963924641489, + "learning_rate": 8.683425209091807e-06, + "loss": 0.3764, + "step": 11780 + }, + { + "epoch": 0.2593220338983051, + "grad_norm": 1.4424899293466942, + "learning_rate": 8.683184151089817e-06, + "loss": 0.4285, + "step": 11781 + }, + { + "epoch": 0.25934404578472375, + "grad_norm": 0.6260776766479044, + "learning_rate": 8.682943074368241e-06, + "loss": 0.5002, + "step": 11782 + }, + { + "epoch": 0.2593660576711424, + "grad_norm": 0.6211040402501141, + "learning_rate": 8.682701978928307e-06, + "loss": 0.5242, + "step": 11783 + }, + { + "epoch": 0.2593880695575611, + "grad_norm": 1.214704666956915, + "learning_rate": 8.682460864771238e-06, + "loss": 0.3156, + "step": 11784 + }, + { + "epoch": 0.25941008144397976, + "grad_norm": 1.4025870268198264, + "learning_rate": 8.68221973189826e-06, + "loss": 0.4237, + "step": 11785 + }, + { + "epoch": 0.2594320933303984, + "grad_norm": 1.1451747179621463, + "learning_rate": 8.6819785803106e-06, + "loss": 0.3135, + "step": 11786 + }, + { + "epoch": 0.25945410521681705, + "grad_norm": 1.2528897522622207, + "learning_rate": 8.681737410009483e-06, + "loss": 0.365, + "step": 11787 + }, + { + "epoch": 0.25947611710323576, + "grad_norm": 1.5625842270041963, + "learning_rate": 8.681496220996133e-06, + "loss": 0.313, + "step": 11788 + }, + { + "epoch": 0.2594981289896544, + "grad_norm": 0.6527453222085144, + "learning_rate": 8.681255013271778e-06, + "loss": 0.5423, + "step": 11789 + }, + { + "epoch": 0.25952014087607306, + "grad_norm": 4.88263406110633, + "learning_rate": 8.681013786837644e-06, + "loss": 0.4519, + "step": 11790 + }, + { + "epoch": 0.25954215276249176, + "grad_norm": 1.7288185411957468, + "learning_rate": 8.680772541694954e-06, + "loss": 0.4024, + "step": 11791 + }, + { + "epoch": 0.2595641646489104, + "grad_norm": 2.794749791442729, + "learning_rate": 8.680531277844937e-06, + "loss": 0.2928, + "step": 11792 + }, + { + "epoch": 0.25958617653532906, + "grad_norm": 0.641719476813273, + "learning_rate": 8.68028999528882e-06, + "loss": 0.5205, + "step": 11793 + }, + { + "epoch": 0.25960818842174777, + "grad_norm": 1.1824726384222255, + "learning_rate": 8.680048694027827e-06, + "loss": 0.395, + "step": 11794 + }, + { + "epoch": 0.2596302003081664, + "grad_norm": 1.3642003523363333, + "learning_rate": 8.679807374063185e-06, + "loss": 0.3351, + "step": 11795 + }, + { + "epoch": 0.25965221219458506, + "grad_norm": 1.9030969095318409, + "learning_rate": 8.67956603539612e-06, + "loss": 0.3394, + "step": 11796 + }, + { + "epoch": 0.25967422408100377, + "grad_norm": 1.5128542643183172, + "learning_rate": 8.67932467802786e-06, + "loss": 0.364, + "step": 11797 + }, + { + "epoch": 0.2596962359674224, + "grad_norm": 1.294779902207174, + "learning_rate": 8.67908330195963e-06, + "loss": 0.3282, + "step": 11798 + }, + { + "epoch": 0.25971824785384107, + "grad_norm": 0.6400820876179454, + "learning_rate": 8.67884190719266e-06, + "loss": 0.5041, + "step": 11799 + }, + { + "epoch": 0.2597402597402597, + "grad_norm": 1.1936466445410738, + "learning_rate": 8.678600493728173e-06, + "loss": 0.33, + "step": 11800 + }, + { + "epoch": 0.2597622716266784, + "grad_norm": 2.222877821340332, + "learning_rate": 8.678359061567399e-06, + "loss": 0.3001, + "step": 11801 + }, + { + "epoch": 0.25978428351309707, + "grad_norm": 1.4897921632010094, + "learning_rate": 8.67811761071156e-06, + "loss": 0.3779, + "step": 11802 + }, + { + "epoch": 0.2598062953995157, + "grad_norm": 1.218265693815542, + "learning_rate": 8.677876141161891e-06, + "loss": 0.3125, + "step": 11803 + }, + { + "epoch": 0.2598283072859344, + "grad_norm": 1.3833645255757818, + "learning_rate": 8.677634652919614e-06, + "loss": 0.3176, + "step": 11804 + }, + { + "epoch": 0.2598503191723531, + "grad_norm": 1.5724154233424255, + "learning_rate": 8.677393145985956e-06, + "loss": 0.3903, + "step": 11805 + }, + { + "epoch": 0.2598723310587717, + "grad_norm": 1.6112862831818866, + "learning_rate": 8.677151620362146e-06, + "loss": 0.3142, + "step": 11806 + }, + { + "epoch": 0.2598943429451904, + "grad_norm": 1.7284654945383766, + "learning_rate": 8.676910076049412e-06, + "loss": 0.3494, + "step": 11807 + }, + { + "epoch": 0.2599163548316091, + "grad_norm": 1.2974016101615582, + "learning_rate": 8.676668513048983e-06, + "loss": 0.3648, + "step": 11808 + }, + { + "epoch": 0.2599383667180277, + "grad_norm": 1.562886968667747, + "learning_rate": 8.67642693136208e-06, + "loss": 0.3579, + "step": 11809 + }, + { + "epoch": 0.2599603786044464, + "grad_norm": 1.5959216100032185, + "learning_rate": 8.676185330989941e-06, + "loss": 0.3606, + "step": 11810 + }, + { + "epoch": 0.2599823904908651, + "grad_norm": 2.338805933707548, + "learning_rate": 8.675943711933784e-06, + "loss": 0.3505, + "step": 11811 + }, + { + "epoch": 0.2600044023772837, + "grad_norm": 0.6872814848592419, + "learning_rate": 8.675702074194844e-06, + "loss": 0.5353, + "step": 11812 + }, + { + "epoch": 0.2600264142637024, + "grad_norm": 1.3494902374637938, + "learning_rate": 8.675460417774347e-06, + "loss": 0.251, + "step": 11813 + }, + { + "epoch": 0.2600484261501211, + "grad_norm": 1.4960303626140878, + "learning_rate": 8.67521874267352e-06, + "loss": 0.3569, + "step": 11814 + }, + { + "epoch": 0.26007043803653973, + "grad_norm": 1.483164719053286, + "learning_rate": 8.67497704889359e-06, + "loss": 0.3701, + "step": 11815 + }, + { + "epoch": 0.2600924499229584, + "grad_norm": 1.2845069327311858, + "learning_rate": 8.674735336435792e-06, + "loss": 0.338, + "step": 11816 + }, + { + "epoch": 0.2601144618093771, + "grad_norm": 1.6485782800261453, + "learning_rate": 8.674493605301349e-06, + "loss": 0.3528, + "step": 11817 + }, + { + "epoch": 0.26013647369579573, + "grad_norm": 1.6431087134682283, + "learning_rate": 8.67425185549149e-06, + "loss": 0.3964, + "step": 11818 + }, + { + "epoch": 0.2601584855822144, + "grad_norm": 1.1192217631256325, + "learning_rate": 8.674010087007442e-06, + "loss": 0.3879, + "step": 11819 + }, + { + "epoch": 0.2601804974686331, + "grad_norm": 1.3074412772500068, + "learning_rate": 8.673768299850438e-06, + "loss": 0.3033, + "step": 11820 + }, + { + "epoch": 0.26020250935505174, + "grad_norm": 25.550694855030027, + "learning_rate": 8.673526494021704e-06, + "loss": 0.3777, + "step": 11821 + }, + { + "epoch": 0.2602245212414704, + "grad_norm": 1.4247385596011921, + "learning_rate": 8.673284669522472e-06, + "loss": 0.2864, + "step": 11822 + }, + { + "epoch": 0.26024653312788903, + "grad_norm": 1.3046800727893235, + "learning_rate": 8.673042826353967e-06, + "loss": 0.335, + "step": 11823 + }, + { + "epoch": 0.26026854501430774, + "grad_norm": 1.3241313548747895, + "learning_rate": 8.672800964517421e-06, + "loss": 0.34, + "step": 11824 + }, + { + "epoch": 0.2602905569007264, + "grad_norm": 0.6762696362431089, + "learning_rate": 8.672559084014062e-06, + "loss": 0.5279, + "step": 11825 + }, + { + "epoch": 0.26031256878714504, + "grad_norm": 1.6263612686650002, + "learning_rate": 8.67231718484512e-06, + "loss": 0.3976, + "step": 11826 + }, + { + "epoch": 0.26033458067356374, + "grad_norm": 0.636238479422393, + "learning_rate": 8.672075267011823e-06, + "loss": 0.4936, + "step": 11827 + }, + { + "epoch": 0.2603565925599824, + "grad_norm": 1.30400407137935, + "learning_rate": 8.671833330515404e-06, + "loss": 0.3132, + "step": 11828 + }, + { + "epoch": 0.26037860444640104, + "grad_norm": 1.398739166438459, + "learning_rate": 8.67159137535709e-06, + "loss": 0.3885, + "step": 11829 + }, + { + "epoch": 0.26040061633281975, + "grad_norm": 1.3425637383991356, + "learning_rate": 8.67134940153811e-06, + "loss": 0.3656, + "step": 11830 + }, + { + "epoch": 0.2604226282192384, + "grad_norm": 1.4234764921522125, + "learning_rate": 8.671107409059695e-06, + "loss": 0.2966, + "step": 11831 + }, + { + "epoch": 0.26044464010565704, + "grad_norm": 1.5439881663674317, + "learning_rate": 8.670865397923074e-06, + "loss": 0.4127, + "step": 11832 + }, + { + "epoch": 0.26046665199207575, + "grad_norm": 1.2226485572040617, + "learning_rate": 8.67062336812948e-06, + "loss": 0.2395, + "step": 11833 + }, + { + "epoch": 0.2604886638784944, + "grad_norm": 0.6805488962869288, + "learning_rate": 8.670381319680138e-06, + "loss": 0.5216, + "step": 11834 + }, + { + "epoch": 0.26051067576491305, + "grad_norm": 0.6449818506593321, + "learning_rate": 8.670139252576284e-06, + "loss": 0.5086, + "step": 11835 + }, + { + "epoch": 0.2605326876513317, + "grad_norm": 1.3924353606142637, + "learning_rate": 8.669897166819143e-06, + "loss": 0.3641, + "step": 11836 + }, + { + "epoch": 0.2605546995377504, + "grad_norm": 2.136405283964852, + "learning_rate": 8.66965506240995e-06, + "loss": 0.499, + "step": 11837 + }, + { + "epoch": 0.26057671142416905, + "grad_norm": 1.5089406023517264, + "learning_rate": 8.66941293934993e-06, + "loss": 0.4338, + "step": 11838 + }, + { + "epoch": 0.2605987233105877, + "grad_norm": 1.380512123525447, + "learning_rate": 8.669170797640321e-06, + "loss": 0.2981, + "step": 11839 + }, + { + "epoch": 0.2606207351970064, + "grad_norm": 1.5939008651180377, + "learning_rate": 8.668928637282347e-06, + "loss": 0.258, + "step": 11840 + }, + { + "epoch": 0.26064274708342505, + "grad_norm": 0.6920610589603419, + "learning_rate": 8.668686458277241e-06, + "loss": 0.5338, + "step": 11841 + }, + { + "epoch": 0.2606647589698437, + "grad_norm": 0.6498949932144821, + "learning_rate": 8.668444260626236e-06, + "loss": 0.5334, + "step": 11842 + }, + { + "epoch": 0.2606867708562624, + "grad_norm": 1.3508140646144782, + "learning_rate": 8.668202044330558e-06, + "loss": 0.3191, + "step": 11843 + }, + { + "epoch": 0.26070878274268106, + "grad_norm": 1.151306493194056, + "learning_rate": 8.667959809391444e-06, + "loss": 0.3855, + "step": 11844 + }, + { + "epoch": 0.2607307946290997, + "grad_norm": 1.288837652288815, + "learning_rate": 8.667717555810123e-06, + "loss": 0.3511, + "step": 11845 + }, + { + "epoch": 0.26075280651551835, + "grad_norm": 1.533986246063883, + "learning_rate": 8.667475283587823e-06, + "loss": 0.2804, + "step": 11846 + }, + { + "epoch": 0.26077481840193706, + "grad_norm": 2.023219185385223, + "learning_rate": 8.667232992725779e-06, + "loss": 0.3946, + "step": 11847 + }, + { + "epoch": 0.2607968302883557, + "grad_norm": 1.4767727100581862, + "learning_rate": 8.66699068322522e-06, + "loss": 0.3795, + "step": 11848 + }, + { + "epoch": 0.26081884217477436, + "grad_norm": 1.7739596887510816, + "learning_rate": 8.66674835508738e-06, + "loss": 0.3716, + "step": 11849 + }, + { + "epoch": 0.26084085406119306, + "grad_norm": 1.2114831291180328, + "learning_rate": 8.66650600831349e-06, + "loss": 0.3372, + "step": 11850 + }, + { + "epoch": 0.2608628659476117, + "grad_norm": 1.195046228968159, + "learning_rate": 8.666263642904779e-06, + "loss": 0.3056, + "step": 11851 + }, + { + "epoch": 0.26088487783403036, + "grad_norm": 1.4941362802309057, + "learning_rate": 8.666021258862485e-06, + "loss": 0.3932, + "step": 11852 + }, + { + "epoch": 0.26090688972044906, + "grad_norm": 1.2102122217486293, + "learning_rate": 8.665778856187833e-06, + "loss": 0.2949, + "step": 11853 + }, + { + "epoch": 0.2609289016068677, + "grad_norm": 1.3853165519947568, + "learning_rate": 8.665536434882059e-06, + "loss": 0.3274, + "step": 11854 + }, + { + "epoch": 0.26095091349328636, + "grad_norm": 1.4269737460350809, + "learning_rate": 8.665293994946391e-06, + "loss": 0.3125, + "step": 11855 + }, + { + "epoch": 0.26097292537970507, + "grad_norm": 1.1714385886567864, + "learning_rate": 8.665051536382068e-06, + "loss": 0.3044, + "step": 11856 + }, + { + "epoch": 0.2609949372661237, + "grad_norm": 1.274913683932168, + "learning_rate": 8.664809059190315e-06, + "loss": 0.3115, + "step": 11857 + }, + { + "epoch": 0.26101694915254237, + "grad_norm": 1.1458860845338328, + "learning_rate": 8.66456656337237e-06, + "loss": 0.3175, + "step": 11858 + }, + { + "epoch": 0.261038961038961, + "grad_norm": 1.4174227008727618, + "learning_rate": 8.664324048929463e-06, + "loss": 0.3865, + "step": 11859 + }, + { + "epoch": 0.2610609729253797, + "grad_norm": 1.5611898488318825, + "learning_rate": 8.664081515862828e-06, + "loss": 0.3959, + "step": 11860 + }, + { + "epoch": 0.26108298481179837, + "grad_norm": 2.4226637387348426, + "learning_rate": 8.663838964173695e-06, + "loss": 0.4081, + "step": 11861 + }, + { + "epoch": 0.261104996698217, + "grad_norm": 0.8679004606224986, + "learning_rate": 8.663596393863299e-06, + "loss": 0.5442, + "step": 11862 + }, + { + "epoch": 0.2611270085846357, + "grad_norm": 1.3132319358969358, + "learning_rate": 8.663353804932871e-06, + "loss": 0.3516, + "step": 11863 + }, + { + "epoch": 0.26114902047105437, + "grad_norm": 0.6977674331270928, + "learning_rate": 8.663111197383644e-06, + "loss": 0.5253, + "step": 11864 + }, + { + "epoch": 0.261171032357473, + "grad_norm": 1.5086412061803787, + "learning_rate": 8.662868571216855e-06, + "loss": 0.2926, + "step": 11865 + }, + { + "epoch": 0.2611930442438917, + "grad_norm": 1.9451357722678, + "learning_rate": 8.662625926433732e-06, + "loss": 0.354, + "step": 11866 + }, + { + "epoch": 0.2612150561303104, + "grad_norm": 1.9104418492130777, + "learning_rate": 8.662383263035511e-06, + "loss": 0.3095, + "step": 11867 + }, + { + "epoch": 0.261237068016729, + "grad_norm": 1.232481166994814, + "learning_rate": 8.662140581023427e-06, + "loss": 0.2735, + "step": 11868 + }, + { + "epoch": 0.2612590799031477, + "grad_norm": 1.2772390111337495, + "learning_rate": 8.66189788039871e-06, + "loss": 0.3241, + "step": 11869 + }, + { + "epoch": 0.2612810917895664, + "grad_norm": 1.3159805937257747, + "learning_rate": 8.661655161162593e-06, + "loss": 0.3148, + "step": 11870 + }, + { + "epoch": 0.261303103675985, + "grad_norm": 1.4944690548187083, + "learning_rate": 8.661412423316314e-06, + "loss": 0.3514, + "step": 11871 + }, + { + "epoch": 0.2613251155624037, + "grad_norm": 1.341288721754239, + "learning_rate": 8.6611696668611e-06, + "loss": 0.3385, + "step": 11872 + }, + { + "epoch": 0.2613471274488224, + "grad_norm": 0.98614109243656, + "learning_rate": 8.660926891798193e-06, + "loss": 0.5707, + "step": 11873 + }, + { + "epoch": 0.26136913933524103, + "grad_norm": 1.5724729502416592, + "learning_rate": 8.66068409812882e-06, + "loss": 0.4489, + "step": 11874 + }, + { + "epoch": 0.2613911512216597, + "grad_norm": 1.6973871531144722, + "learning_rate": 8.660441285854219e-06, + "loss": 0.3392, + "step": 11875 + }, + { + "epoch": 0.2614131631080784, + "grad_norm": 1.912695896688355, + "learning_rate": 8.660198454975623e-06, + "loss": 0.3601, + "step": 11876 + }, + { + "epoch": 0.26143517499449703, + "grad_norm": 1.3532759476318745, + "learning_rate": 8.659955605494266e-06, + "loss": 0.3974, + "step": 11877 + }, + { + "epoch": 0.2614571868809157, + "grad_norm": 1.7231943992692267, + "learning_rate": 8.65971273741138e-06, + "loss": 0.4962, + "step": 11878 + }, + { + "epoch": 0.2614791987673344, + "grad_norm": 1.3013751782432217, + "learning_rate": 8.659469850728206e-06, + "loss": 0.3509, + "step": 11879 + }, + { + "epoch": 0.26150121065375304, + "grad_norm": 0.6570481767201064, + "learning_rate": 8.659226945445971e-06, + "loss": 0.5267, + "step": 11880 + }, + { + "epoch": 0.2615232225401717, + "grad_norm": 1.3991789022848176, + "learning_rate": 8.658984021565912e-06, + "loss": 0.2813, + "step": 11881 + }, + { + "epoch": 0.26154523442659033, + "grad_norm": 1.4464036879431084, + "learning_rate": 8.658741079089265e-06, + "loss": 0.3584, + "step": 11882 + }, + { + "epoch": 0.26156724631300904, + "grad_norm": 1.9632668090978738, + "learning_rate": 8.658498118017265e-06, + "loss": 0.373, + "step": 11883 + }, + { + "epoch": 0.2615892581994277, + "grad_norm": 1.086895382011049, + "learning_rate": 8.658255138351145e-06, + "loss": 0.2773, + "step": 11884 + }, + { + "epoch": 0.26161127008584634, + "grad_norm": 1.4266337810286045, + "learning_rate": 8.65801214009214e-06, + "loss": 0.3598, + "step": 11885 + }, + { + "epoch": 0.26163328197226504, + "grad_norm": 1.5419527550987966, + "learning_rate": 8.657769123241487e-06, + "loss": 0.3444, + "step": 11886 + }, + { + "epoch": 0.2616552938586837, + "grad_norm": 1.3267734023882034, + "learning_rate": 8.657526087800419e-06, + "loss": 0.28, + "step": 11887 + }, + { + "epoch": 0.26167730574510234, + "grad_norm": 2.467497667121345, + "learning_rate": 8.657283033770175e-06, + "loss": 0.3294, + "step": 11888 + }, + { + "epoch": 0.26169931763152104, + "grad_norm": 1.3600449565763042, + "learning_rate": 8.657039961151984e-06, + "loss": 0.2911, + "step": 11889 + }, + { + "epoch": 0.2617213295179397, + "grad_norm": 1.444890971226781, + "learning_rate": 8.656796869947086e-06, + "loss": 0.3192, + "step": 11890 + }, + { + "epoch": 0.26174334140435834, + "grad_norm": 1.9164261913287015, + "learning_rate": 8.656553760156714e-06, + "loss": 0.3505, + "step": 11891 + }, + { + "epoch": 0.261765353290777, + "grad_norm": 1.7523548810826077, + "learning_rate": 8.656310631782108e-06, + "loss": 0.2467, + "step": 11892 + }, + { + "epoch": 0.2617873651771957, + "grad_norm": 1.845964435500405, + "learning_rate": 8.656067484824498e-06, + "loss": 0.4077, + "step": 11893 + }, + { + "epoch": 0.26180937706361435, + "grad_norm": 1.2048169348348943, + "learning_rate": 8.655824319285122e-06, + "loss": 0.3169, + "step": 11894 + }, + { + "epoch": 0.261831388950033, + "grad_norm": 1.390080870618934, + "learning_rate": 8.655581135165218e-06, + "loss": 0.3536, + "step": 11895 + }, + { + "epoch": 0.2618534008364517, + "grad_norm": 1.7178470622519728, + "learning_rate": 8.65533793246602e-06, + "loss": 0.2684, + "step": 11896 + }, + { + "epoch": 0.26187541272287035, + "grad_norm": 1.5242220369863957, + "learning_rate": 8.655094711188764e-06, + "loss": 0.3614, + "step": 11897 + }, + { + "epoch": 0.261897424609289, + "grad_norm": 1.3417471150863218, + "learning_rate": 8.654851471334687e-06, + "loss": 0.407, + "step": 11898 + }, + { + "epoch": 0.2619194364957077, + "grad_norm": 1.637398601192649, + "learning_rate": 8.654608212905023e-06, + "loss": 0.3368, + "step": 11899 + }, + { + "epoch": 0.26194144838212635, + "grad_norm": 1.514650899889873, + "learning_rate": 8.654364935901011e-06, + "loss": 0.3405, + "step": 11900 + }, + { + "epoch": 0.261963460268545, + "grad_norm": 1.1373149048650406, + "learning_rate": 8.654121640323887e-06, + "loss": 0.3016, + "step": 11901 + }, + { + "epoch": 0.2619854721549637, + "grad_norm": 1.3711772949879992, + "learning_rate": 8.653878326174887e-06, + "loss": 0.3166, + "step": 11902 + }, + { + "epoch": 0.26200748404138235, + "grad_norm": 1.7289603664466997, + "learning_rate": 8.653634993455248e-06, + "loss": 0.2534, + "step": 11903 + }, + { + "epoch": 0.262029495927801, + "grad_norm": 0.7920048084075936, + "learning_rate": 8.653391642166204e-06, + "loss": 0.505, + "step": 11904 + }, + { + "epoch": 0.26205150781421965, + "grad_norm": 1.779490523439072, + "learning_rate": 8.653148272308998e-06, + "loss": 0.4029, + "step": 11905 + }, + { + "epoch": 0.26207351970063836, + "grad_norm": 1.3396942045884694, + "learning_rate": 8.652904883884861e-06, + "loss": 0.322, + "step": 11906 + }, + { + "epoch": 0.262095531587057, + "grad_norm": 1.3601763657626698, + "learning_rate": 8.652661476895032e-06, + "loss": 0.309, + "step": 11907 + }, + { + "epoch": 0.26211754347347566, + "grad_norm": 1.5675541459337456, + "learning_rate": 8.652418051340747e-06, + "loss": 0.291, + "step": 11908 + }, + { + "epoch": 0.26213955535989436, + "grad_norm": 0.6568894271800534, + "learning_rate": 8.652174607223246e-06, + "loss": 0.539, + "step": 11909 + }, + { + "epoch": 0.262161567246313, + "grad_norm": 1.5040277659825183, + "learning_rate": 8.651931144543766e-06, + "loss": 0.3472, + "step": 11910 + }, + { + "epoch": 0.26218357913273166, + "grad_norm": 1.3032693627792509, + "learning_rate": 8.651687663303542e-06, + "loss": 0.3261, + "step": 11911 + }, + { + "epoch": 0.26220559101915036, + "grad_norm": 1.5192451550178354, + "learning_rate": 8.651444163503812e-06, + "loss": 0.26, + "step": 11912 + }, + { + "epoch": 0.262227602905569, + "grad_norm": 1.6693997895242916, + "learning_rate": 8.651200645145813e-06, + "loss": 0.331, + "step": 11913 + }, + { + "epoch": 0.26224961479198766, + "grad_norm": 1.704728132886378, + "learning_rate": 8.650957108230787e-06, + "loss": 0.4138, + "step": 11914 + }, + { + "epoch": 0.26227162667840637, + "grad_norm": 1.5725259246856094, + "learning_rate": 8.650713552759966e-06, + "loss": 0.4359, + "step": 11915 + }, + { + "epoch": 0.262293638564825, + "grad_norm": 0.7578469324557383, + "learning_rate": 8.650469978734593e-06, + "loss": 0.5634, + "step": 11916 + }, + { + "epoch": 0.26231565045124366, + "grad_norm": 1.1582433435417467, + "learning_rate": 8.650226386155902e-06, + "loss": 0.3707, + "step": 11917 + }, + { + "epoch": 0.2623376623376623, + "grad_norm": 1.3452521105942896, + "learning_rate": 8.649982775025132e-06, + "loss": 0.3108, + "step": 11918 + }, + { + "epoch": 0.262359674224081, + "grad_norm": 1.6126679023446542, + "learning_rate": 8.649739145343521e-06, + "loss": 0.358, + "step": 11919 + }, + { + "epoch": 0.26238168611049967, + "grad_norm": 1.529403838487069, + "learning_rate": 8.64949549711231e-06, + "loss": 0.3168, + "step": 11920 + }, + { + "epoch": 0.2624036979969183, + "grad_norm": 1.8671666771757136, + "learning_rate": 8.649251830332735e-06, + "loss": 0.2911, + "step": 11921 + }, + { + "epoch": 0.262425709883337, + "grad_norm": 1.3025947817837162, + "learning_rate": 8.649008145006033e-06, + "loss": 0.4021, + "step": 11922 + }, + { + "epoch": 0.26244772176975567, + "grad_norm": 1.7413818463208308, + "learning_rate": 8.648764441133443e-06, + "loss": 0.3672, + "step": 11923 + }, + { + "epoch": 0.2624697336561743, + "grad_norm": 1.4220543891926698, + "learning_rate": 8.648520718716207e-06, + "loss": 0.3297, + "step": 11924 + }, + { + "epoch": 0.262491745542593, + "grad_norm": 1.78323790967411, + "learning_rate": 8.648276977755562e-06, + "loss": 0.4595, + "step": 11925 + }, + { + "epoch": 0.2625137574290117, + "grad_norm": 0.6636231220200778, + "learning_rate": 8.648033218252745e-06, + "loss": 0.5322, + "step": 11926 + }, + { + "epoch": 0.2625357693154303, + "grad_norm": 1.3542832470042978, + "learning_rate": 8.647789440208996e-06, + "loss": 0.3678, + "step": 11927 + }, + { + "epoch": 0.26255778120184897, + "grad_norm": 1.3186279206773546, + "learning_rate": 8.647545643625555e-06, + "loss": 0.3888, + "step": 11928 + }, + { + "epoch": 0.2625797930882677, + "grad_norm": 1.2096436276624367, + "learning_rate": 8.647301828503658e-06, + "loss": 0.2912, + "step": 11929 + }, + { + "epoch": 0.2626018049746863, + "grad_norm": 2.55473145287368, + "learning_rate": 8.647057994844548e-06, + "loss": 0.3644, + "step": 11930 + }, + { + "epoch": 0.262623816861105, + "grad_norm": 1.7242422014065548, + "learning_rate": 8.646814142649462e-06, + "loss": 0.426, + "step": 11931 + }, + { + "epoch": 0.2626458287475237, + "grad_norm": 1.488230595123231, + "learning_rate": 8.64657027191964e-06, + "loss": 0.2973, + "step": 11932 + }, + { + "epoch": 0.26266784063394233, + "grad_norm": 1.3659172850367436, + "learning_rate": 8.646326382656323e-06, + "loss": 0.405, + "step": 11933 + }, + { + "epoch": 0.262689852520361, + "grad_norm": 1.127646472646064, + "learning_rate": 8.646082474860747e-06, + "loss": 0.2762, + "step": 11934 + }, + { + "epoch": 0.2627118644067797, + "grad_norm": 1.169448555330834, + "learning_rate": 8.645838548534156e-06, + "loss": 0.1787, + "step": 11935 + }, + { + "epoch": 0.26273387629319833, + "grad_norm": 1.5118910896293363, + "learning_rate": 8.645594603677785e-06, + "loss": 0.2718, + "step": 11936 + }, + { + "epoch": 0.262755888179617, + "grad_norm": 2.306655796131699, + "learning_rate": 8.645350640292878e-06, + "loss": 0.3342, + "step": 11937 + }, + { + "epoch": 0.2627779000660357, + "grad_norm": 0.6352040029344971, + "learning_rate": 8.64510665838067e-06, + "loss": 0.5067, + "step": 11938 + }, + { + "epoch": 0.26279991195245433, + "grad_norm": 1.3275694818219352, + "learning_rate": 8.644862657942408e-06, + "loss": 0.343, + "step": 11939 + }, + { + "epoch": 0.262821923838873, + "grad_norm": 1.4857060288246238, + "learning_rate": 8.644618638979327e-06, + "loss": 0.2483, + "step": 11940 + }, + { + "epoch": 0.26284393572529163, + "grad_norm": 1.6256689874315575, + "learning_rate": 8.644374601492669e-06, + "loss": 0.4149, + "step": 11941 + }, + { + "epoch": 0.26286594761171034, + "grad_norm": 1.6547433141394552, + "learning_rate": 8.644130545483673e-06, + "loss": 0.2902, + "step": 11942 + }, + { + "epoch": 0.262887959498129, + "grad_norm": 1.3062497173806993, + "learning_rate": 8.64388647095358e-06, + "loss": 0.3356, + "step": 11943 + }, + { + "epoch": 0.26290997138454764, + "grad_norm": 1.527785828441825, + "learning_rate": 8.643642377903632e-06, + "loss": 0.3435, + "step": 11944 + }, + { + "epoch": 0.26293198327096634, + "grad_norm": 1.282502536263777, + "learning_rate": 8.643398266335067e-06, + "loss": 0.4014, + "step": 11945 + }, + { + "epoch": 0.262953995157385, + "grad_norm": 3.8962183954086558, + "learning_rate": 8.643154136249127e-06, + "loss": 0.3486, + "step": 11946 + }, + { + "epoch": 0.26297600704380364, + "grad_norm": 1.4935533057243398, + "learning_rate": 8.642909987647055e-06, + "loss": 0.3285, + "step": 11947 + }, + { + "epoch": 0.26299801893022234, + "grad_norm": 1.576975129991015, + "learning_rate": 8.642665820530088e-06, + "loss": 0.3373, + "step": 11948 + }, + { + "epoch": 0.263020030816641, + "grad_norm": 1.2802260083212458, + "learning_rate": 8.642421634899468e-06, + "loss": 0.2639, + "step": 11949 + }, + { + "epoch": 0.26304204270305964, + "grad_norm": 1.2271818978014026, + "learning_rate": 8.642177430756437e-06, + "loss": 0.3303, + "step": 11950 + }, + { + "epoch": 0.2630640545894783, + "grad_norm": 1.506985350751099, + "learning_rate": 8.641933208102237e-06, + "loss": 0.3464, + "step": 11951 + }, + { + "epoch": 0.263086066475897, + "grad_norm": 0.6205026849361359, + "learning_rate": 8.641688966938106e-06, + "loss": 0.5009, + "step": 11952 + }, + { + "epoch": 0.26310807836231564, + "grad_norm": 1.2501596217238042, + "learning_rate": 8.64144470726529e-06, + "loss": 0.3263, + "step": 11953 + }, + { + "epoch": 0.2631300902487343, + "grad_norm": 1.1826100236048311, + "learning_rate": 8.641200429085025e-06, + "loss": 0.3174, + "step": 11954 + }, + { + "epoch": 0.263152102135153, + "grad_norm": 1.6644729497181334, + "learning_rate": 8.640956132398558e-06, + "loss": 0.2881, + "step": 11955 + }, + { + "epoch": 0.26317411402157165, + "grad_norm": 0.6328575761360047, + "learning_rate": 8.640711817207127e-06, + "loss": 0.5034, + "step": 11956 + }, + { + "epoch": 0.2631961259079903, + "grad_norm": 1.881149610275643, + "learning_rate": 8.640467483511974e-06, + "loss": 0.3012, + "step": 11957 + }, + { + "epoch": 0.263218137794409, + "grad_norm": 1.5384144524296248, + "learning_rate": 8.640223131314342e-06, + "loss": 0.3723, + "step": 11958 + }, + { + "epoch": 0.26324014968082765, + "grad_norm": 1.1878924162886797, + "learning_rate": 8.639978760615473e-06, + "loss": 0.2857, + "step": 11959 + }, + { + "epoch": 0.2632621615672463, + "grad_norm": 1.3539126154751007, + "learning_rate": 8.63973437141661e-06, + "loss": 0.4202, + "step": 11960 + }, + { + "epoch": 0.263284173453665, + "grad_norm": 1.32064319078182, + "learning_rate": 8.63948996371899e-06, + "loss": 0.3974, + "step": 11961 + }, + { + "epoch": 0.26330618534008365, + "grad_norm": 1.4101536176922895, + "learning_rate": 8.63924553752386e-06, + "loss": 0.3424, + "step": 11962 + }, + { + "epoch": 0.2633281972265023, + "grad_norm": 1.4881970183916702, + "learning_rate": 8.639001092832462e-06, + "loss": 0.4118, + "step": 11963 + }, + { + "epoch": 0.26335020911292095, + "grad_norm": 1.683839851537344, + "learning_rate": 8.638756629646038e-06, + "loss": 0.3074, + "step": 11964 + }, + { + "epoch": 0.26337222099933966, + "grad_norm": 0.6666103232981648, + "learning_rate": 8.638512147965829e-06, + "loss": 0.509, + "step": 11965 + }, + { + "epoch": 0.2633942328857583, + "grad_norm": 1.6063016725311654, + "learning_rate": 8.638267647793078e-06, + "loss": 0.3896, + "step": 11966 + }, + { + "epoch": 0.26341624477217696, + "grad_norm": 6.23325890081688, + "learning_rate": 8.638023129129028e-06, + "loss": 0.3499, + "step": 11967 + }, + { + "epoch": 0.26343825665859566, + "grad_norm": 1.3793400145405923, + "learning_rate": 8.637778591974922e-06, + "loss": 0.3374, + "step": 11968 + }, + { + "epoch": 0.2634602685450143, + "grad_norm": 1.6888121042845834, + "learning_rate": 8.637534036332004e-06, + "loss": 0.3564, + "step": 11969 + }, + { + "epoch": 0.26348228043143296, + "grad_norm": 0.6278004751044355, + "learning_rate": 8.637289462201515e-06, + "loss": 0.5153, + "step": 11970 + }, + { + "epoch": 0.26350429231785166, + "grad_norm": 1.8755758921798211, + "learning_rate": 8.637044869584697e-06, + "loss": 0.3506, + "step": 11971 + }, + { + "epoch": 0.2635263042042703, + "grad_norm": 2.0634754038793885, + "learning_rate": 8.636800258482795e-06, + "loss": 0.3049, + "step": 11972 + }, + { + "epoch": 0.26354831609068896, + "grad_norm": 1.7082888446336029, + "learning_rate": 8.636555628897054e-06, + "loss": 0.3311, + "step": 11973 + }, + { + "epoch": 0.26357032797710767, + "grad_norm": 1.4416732110463994, + "learning_rate": 8.636310980828714e-06, + "loss": 0.2947, + "step": 11974 + }, + { + "epoch": 0.2635923398635263, + "grad_norm": 1.3490298661563802, + "learning_rate": 8.636066314279021e-06, + "loss": 0.3806, + "step": 11975 + }, + { + "epoch": 0.26361435174994496, + "grad_norm": 1.5646850790465971, + "learning_rate": 8.635821629249217e-06, + "loss": 0.3912, + "step": 11976 + }, + { + "epoch": 0.2636363636363636, + "grad_norm": 1.5046114636468433, + "learning_rate": 8.635576925740545e-06, + "loss": 0.3448, + "step": 11977 + }, + { + "epoch": 0.2636583755227823, + "grad_norm": 1.3804625127883923, + "learning_rate": 8.635332203754249e-06, + "loss": 0.2518, + "step": 11978 + }, + { + "epoch": 0.26368038740920097, + "grad_norm": 1.5626639909673226, + "learning_rate": 8.635087463291575e-06, + "loss": 0.2903, + "step": 11979 + }, + { + "epoch": 0.2637023992956196, + "grad_norm": 1.1366765829166399, + "learning_rate": 8.634842704353765e-06, + "loss": 0.3632, + "step": 11980 + }, + { + "epoch": 0.2637244111820383, + "grad_norm": 1.4223672675014338, + "learning_rate": 8.634597926942061e-06, + "loss": 0.3422, + "step": 11981 + }, + { + "epoch": 0.26374642306845697, + "grad_norm": 1.3818159813567914, + "learning_rate": 8.634353131057711e-06, + "loss": 0.3725, + "step": 11982 + }, + { + "epoch": 0.2637684349548756, + "grad_norm": 1.0713056712956526, + "learning_rate": 8.634108316701958e-06, + "loss": 0.272, + "step": 11983 + }, + { + "epoch": 0.2637904468412943, + "grad_norm": 1.5636019196543958, + "learning_rate": 8.633863483876044e-06, + "loss": 0.334, + "step": 11984 + }, + { + "epoch": 0.263812458727713, + "grad_norm": 1.1822905190193527, + "learning_rate": 8.633618632581218e-06, + "loss": 0.2627, + "step": 11985 + }, + { + "epoch": 0.2638344706141316, + "grad_norm": 0.6429292760204489, + "learning_rate": 8.633373762818718e-06, + "loss": 0.5233, + "step": 11986 + }, + { + "epoch": 0.26385648250055027, + "grad_norm": 1.4963882176073016, + "learning_rate": 8.633128874589794e-06, + "loss": 0.2718, + "step": 11987 + }, + { + "epoch": 0.263878494386969, + "grad_norm": 0.609910637391298, + "learning_rate": 8.632883967895689e-06, + "loss": 0.4995, + "step": 11988 + }, + { + "epoch": 0.2639005062733876, + "grad_norm": 1.446820087277642, + "learning_rate": 8.632639042737645e-06, + "loss": 0.4725, + "step": 11989 + }, + { + "epoch": 0.2639225181598063, + "grad_norm": 0.6339550149548167, + "learning_rate": 8.632394099116911e-06, + "loss": 0.5257, + "step": 11990 + }, + { + "epoch": 0.263944530046225, + "grad_norm": 1.4577509779025293, + "learning_rate": 8.63214913703473e-06, + "loss": 0.3124, + "step": 11991 + }, + { + "epoch": 0.26396654193264363, + "grad_norm": 1.5480490809935983, + "learning_rate": 8.631904156492346e-06, + "loss": 0.3428, + "step": 11992 + }, + { + "epoch": 0.2639885538190623, + "grad_norm": 0.6169321887791664, + "learning_rate": 8.631659157491006e-06, + "loss": 0.5181, + "step": 11993 + }, + { + "epoch": 0.264010565705481, + "grad_norm": 1.4181133145698601, + "learning_rate": 8.631414140031954e-06, + "loss": 0.3736, + "step": 11994 + }, + { + "epoch": 0.26403257759189963, + "grad_norm": 1.4377331252078211, + "learning_rate": 8.631169104116438e-06, + "loss": 0.3125, + "step": 11995 + }, + { + "epoch": 0.2640545894783183, + "grad_norm": 2.200834094755469, + "learning_rate": 8.630924049745699e-06, + "loss": 0.3377, + "step": 11996 + }, + { + "epoch": 0.264076601364737, + "grad_norm": 1.8091874286928393, + "learning_rate": 8.630678976920984e-06, + "loss": 0.3871, + "step": 11997 + }, + { + "epoch": 0.26409861325115563, + "grad_norm": 1.601622344425137, + "learning_rate": 8.630433885643541e-06, + "loss": 0.3342, + "step": 11998 + }, + { + "epoch": 0.2641206251375743, + "grad_norm": 1.8516721818322475, + "learning_rate": 8.630188775914612e-06, + "loss": 0.3811, + "step": 11999 + }, + { + "epoch": 0.26414263702399293, + "grad_norm": 1.2448480844323917, + "learning_rate": 8.629943647735445e-06, + "loss": 0.2573, + "step": 12000 + }, + { + "epoch": 0.26416464891041164, + "grad_norm": 1.4123160521504354, + "learning_rate": 8.629698501107286e-06, + "loss": 0.4398, + "step": 12001 + }, + { + "epoch": 0.2641866607968303, + "grad_norm": 1.7135317860469368, + "learning_rate": 8.62945333603138e-06, + "loss": 0.3887, + "step": 12002 + }, + { + "epoch": 0.26420867268324894, + "grad_norm": 1.7657406329080854, + "learning_rate": 8.629208152508972e-06, + "loss": 0.423, + "step": 12003 + }, + { + "epoch": 0.26423068456966764, + "grad_norm": 2.6770131353076474, + "learning_rate": 8.628962950541311e-06, + "loss": 0.3434, + "step": 12004 + }, + { + "epoch": 0.2642526964560863, + "grad_norm": 1.5630963090352583, + "learning_rate": 8.628717730129642e-06, + "loss": 0.4452, + "step": 12005 + }, + { + "epoch": 0.26427470834250494, + "grad_norm": 2.71656636137896, + "learning_rate": 8.62847249127521e-06, + "loss": 0.3112, + "step": 12006 + }, + { + "epoch": 0.26429672022892364, + "grad_norm": 1.6231573263862358, + "learning_rate": 8.628227233979262e-06, + "loss": 0.3886, + "step": 12007 + }, + { + "epoch": 0.2643187321153423, + "grad_norm": 1.697757958344283, + "learning_rate": 8.627981958243046e-06, + "loss": 0.4441, + "step": 12008 + }, + { + "epoch": 0.26434074400176094, + "grad_norm": 2.1644533944142794, + "learning_rate": 8.627736664067809e-06, + "loss": 0.2522, + "step": 12009 + }, + { + "epoch": 0.2643627558881796, + "grad_norm": 2.082871879252096, + "learning_rate": 8.627491351454795e-06, + "loss": 0.3877, + "step": 12010 + }, + { + "epoch": 0.2643847677745983, + "grad_norm": 0.7085815228939284, + "learning_rate": 8.627246020405251e-06, + "loss": 0.5277, + "step": 12011 + }, + { + "epoch": 0.26440677966101694, + "grad_norm": 0.6591694353886651, + "learning_rate": 8.627000670920426e-06, + "loss": 0.5067, + "step": 12012 + }, + { + "epoch": 0.2644287915474356, + "grad_norm": 1.4505566980468338, + "learning_rate": 8.626755303001566e-06, + "loss": 0.4463, + "step": 12013 + }, + { + "epoch": 0.2644508034338543, + "grad_norm": 4.72394191551076, + "learning_rate": 8.626509916649917e-06, + "loss": 0.4037, + "step": 12014 + }, + { + "epoch": 0.26447281532027295, + "grad_norm": 0.6126224162745123, + "learning_rate": 8.626264511866727e-06, + "loss": 0.5184, + "step": 12015 + }, + { + "epoch": 0.2644948272066916, + "grad_norm": 1.4302881415824125, + "learning_rate": 8.626019088653245e-06, + "loss": 0.2847, + "step": 12016 + }, + { + "epoch": 0.2645168390931103, + "grad_norm": 1.5917770977178929, + "learning_rate": 8.625773647010715e-06, + "loss": 0.3334, + "step": 12017 + }, + { + "epoch": 0.26453885097952895, + "grad_norm": 1.374699087854273, + "learning_rate": 8.625528186940389e-06, + "loss": 0.368, + "step": 12018 + }, + { + "epoch": 0.2645608628659476, + "grad_norm": 1.6539320243226057, + "learning_rate": 8.625282708443508e-06, + "loss": 0.349, + "step": 12019 + }, + { + "epoch": 0.2645828747523663, + "grad_norm": 1.210064007372197, + "learning_rate": 8.625037211521326e-06, + "loss": 0.2675, + "step": 12020 + }, + { + "epoch": 0.26460488663878495, + "grad_norm": 1.3292244163009173, + "learning_rate": 8.624791696175088e-06, + "loss": 0.3296, + "step": 12021 + }, + { + "epoch": 0.2646268985252036, + "grad_norm": 0.7171753873596473, + "learning_rate": 8.62454616240604e-06, + "loss": 0.5346, + "step": 12022 + }, + { + "epoch": 0.26464891041162225, + "grad_norm": 1.4953252239423058, + "learning_rate": 8.624300610215432e-06, + "loss": 0.3396, + "step": 12023 + }, + { + "epoch": 0.26467092229804096, + "grad_norm": 1.469295202639179, + "learning_rate": 8.624055039604512e-06, + "loss": 0.3552, + "step": 12024 + }, + { + "epoch": 0.2646929341844596, + "grad_norm": 2.065119803161818, + "learning_rate": 8.62380945057453e-06, + "loss": 0.3191, + "step": 12025 + }, + { + "epoch": 0.26471494607087825, + "grad_norm": 1.3739049796094083, + "learning_rate": 8.623563843126732e-06, + "loss": 0.244, + "step": 12026 + }, + { + "epoch": 0.26473695795729696, + "grad_norm": 2.2140290980548105, + "learning_rate": 8.623318217262364e-06, + "loss": 0.3684, + "step": 12027 + }, + { + "epoch": 0.2647589698437156, + "grad_norm": 1.4131485545958138, + "learning_rate": 8.623072572982678e-06, + "loss": 0.326, + "step": 12028 + }, + { + "epoch": 0.26478098173013426, + "grad_norm": 0.689393009927016, + "learning_rate": 8.622826910288922e-06, + "loss": 0.5137, + "step": 12029 + }, + { + "epoch": 0.26480299361655296, + "grad_norm": 1.3058338126031195, + "learning_rate": 8.622581229182342e-06, + "loss": 0.3884, + "step": 12030 + }, + { + "epoch": 0.2648250055029716, + "grad_norm": 1.609344315652717, + "learning_rate": 8.622335529664189e-06, + "loss": 0.3584, + "step": 12031 + }, + { + "epoch": 0.26484701738939026, + "grad_norm": 0.6312838738208421, + "learning_rate": 8.622089811735712e-06, + "loss": 0.5293, + "step": 12032 + }, + { + "epoch": 0.2648690292758089, + "grad_norm": 1.7230438664721863, + "learning_rate": 8.62184407539816e-06, + "loss": 0.3524, + "step": 12033 + }, + { + "epoch": 0.2648910411622276, + "grad_norm": 0.606772343652573, + "learning_rate": 8.621598320652779e-06, + "loss": 0.5047, + "step": 12034 + }, + { + "epoch": 0.26491305304864626, + "grad_norm": 0.6214379399231004, + "learning_rate": 8.62135254750082e-06, + "loss": 0.5314, + "step": 12035 + }, + { + "epoch": 0.2649350649350649, + "grad_norm": 1.4578572213067698, + "learning_rate": 8.621106755943533e-06, + "loss": 0.4022, + "step": 12036 + }, + { + "epoch": 0.2649570768214836, + "grad_norm": 1.317405329633284, + "learning_rate": 8.620860945982169e-06, + "loss": 0.344, + "step": 12037 + }, + { + "epoch": 0.26497908870790227, + "grad_norm": 0.6413823123960181, + "learning_rate": 8.62061511761797e-06, + "loss": 0.5023, + "step": 12038 + }, + { + "epoch": 0.2650011005943209, + "grad_norm": 1.7255826761483433, + "learning_rate": 8.620369270852193e-06, + "loss": 0.4567, + "step": 12039 + }, + { + "epoch": 0.2650231124807396, + "grad_norm": 1.1600551324398785, + "learning_rate": 8.620123405686083e-06, + "loss": 0.3261, + "step": 12040 + }, + { + "epoch": 0.26504512436715827, + "grad_norm": 0.6381474870930799, + "learning_rate": 8.619877522120893e-06, + "loss": 0.5085, + "step": 12041 + }, + { + "epoch": 0.2650671362535769, + "grad_norm": 1.6030045698643358, + "learning_rate": 8.61963162015787e-06, + "loss": 0.2916, + "step": 12042 + }, + { + "epoch": 0.2650891481399956, + "grad_norm": 1.2035939811661327, + "learning_rate": 8.619385699798263e-06, + "loss": 0.3874, + "step": 12043 + }, + { + "epoch": 0.26511116002641427, + "grad_norm": 1.8818168758787628, + "learning_rate": 8.619139761043326e-06, + "loss": 0.3039, + "step": 12044 + }, + { + "epoch": 0.2651331719128329, + "grad_norm": 1.2649046064252913, + "learning_rate": 8.618893803894307e-06, + "loss": 0.3838, + "step": 12045 + }, + { + "epoch": 0.26515518379925157, + "grad_norm": 1.4928283327295326, + "learning_rate": 8.618647828352455e-06, + "loss": 0.396, + "step": 12046 + }, + { + "epoch": 0.2651771956856703, + "grad_norm": 0.6908256762293289, + "learning_rate": 8.618401834419019e-06, + "loss": 0.538, + "step": 12047 + }, + { + "epoch": 0.2651992075720889, + "grad_norm": 1.6959012523898798, + "learning_rate": 8.618155822095254e-06, + "loss": 0.3023, + "step": 12048 + }, + { + "epoch": 0.2652212194585076, + "grad_norm": 1.5726624764800559, + "learning_rate": 8.617909791382405e-06, + "loss": 0.3927, + "step": 12049 + }, + { + "epoch": 0.2652432313449263, + "grad_norm": 1.4371808255244638, + "learning_rate": 8.617663742281725e-06, + "loss": 0.3952, + "step": 12050 + }, + { + "epoch": 0.2652652432313449, + "grad_norm": 1.8469911119855813, + "learning_rate": 8.617417674794465e-06, + "loss": 0.3973, + "step": 12051 + }, + { + "epoch": 0.2652872551177636, + "grad_norm": 1.4712992163219685, + "learning_rate": 8.617171588921876e-06, + "loss": 0.3333, + "step": 12052 + }, + { + "epoch": 0.2653092670041823, + "grad_norm": 0.6060274962600302, + "learning_rate": 8.616925484665206e-06, + "loss": 0.4903, + "step": 12053 + }, + { + "epoch": 0.26533127889060093, + "grad_norm": 1.058624415484062, + "learning_rate": 8.616679362025708e-06, + "loss": 0.2607, + "step": 12054 + }, + { + "epoch": 0.2653532907770196, + "grad_norm": 1.3592211660132327, + "learning_rate": 8.616433221004633e-06, + "loss": 0.3258, + "step": 12055 + }, + { + "epoch": 0.2653753026634383, + "grad_norm": 1.6472130724681506, + "learning_rate": 8.616187061603229e-06, + "loss": 0.3066, + "step": 12056 + }, + { + "epoch": 0.26539731454985693, + "grad_norm": 1.296505814148808, + "learning_rate": 8.615940883822752e-06, + "loss": 0.4788, + "step": 12057 + }, + { + "epoch": 0.2654193264362756, + "grad_norm": 0.6090867583102066, + "learning_rate": 8.61569468766445e-06, + "loss": 0.4992, + "step": 12058 + }, + { + "epoch": 0.26544133832269423, + "grad_norm": 1.6910851814163137, + "learning_rate": 8.615448473129575e-06, + "loss": 0.3768, + "step": 12059 + }, + { + "epoch": 0.26546335020911294, + "grad_norm": 1.4547054288545136, + "learning_rate": 8.615202240219378e-06, + "loss": 0.3156, + "step": 12060 + }, + { + "epoch": 0.2654853620955316, + "grad_norm": 1.338669867451744, + "learning_rate": 8.614955988935111e-06, + "loss": 0.3258, + "step": 12061 + }, + { + "epoch": 0.26550737398195023, + "grad_norm": 2.0360427504392047, + "learning_rate": 8.614709719278024e-06, + "loss": 0.3428, + "step": 12062 + }, + { + "epoch": 0.26552938586836894, + "grad_norm": 1.946258297065889, + "learning_rate": 8.614463431249372e-06, + "loss": 0.3856, + "step": 12063 + }, + { + "epoch": 0.2655513977547876, + "grad_norm": 1.3039198692334364, + "learning_rate": 8.614217124850404e-06, + "loss": 0.3387, + "step": 12064 + }, + { + "epoch": 0.26557340964120624, + "grad_norm": 1.5225161376363885, + "learning_rate": 8.613970800082374e-06, + "loss": 0.3708, + "step": 12065 + }, + { + "epoch": 0.26559542152762494, + "grad_norm": 0.6424797793366858, + "learning_rate": 8.61372445694653e-06, + "loss": 0.5185, + "step": 12066 + }, + { + "epoch": 0.2656174334140436, + "grad_norm": 1.6399942193459403, + "learning_rate": 8.61347809544413e-06, + "loss": 0.4149, + "step": 12067 + }, + { + "epoch": 0.26563944530046224, + "grad_norm": 1.2453379791872143, + "learning_rate": 8.61323171557642e-06, + "loss": 0.3226, + "step": 12068 + }, + { + "epoch": 0.2656614571868809, + "grad_norm": 1.7128078958489923, + "learning_rate": 8.612985317344654e-06, + "loss": 0.3461, + "step": 12069 + }, + { + "epoch": 0.2656834690732996, + "grad_norm": 1.2919795983235416, + "learning_rate": 8.612738900750086e-06, + "loss": 0.2813, + "step": 12070 + }, + { + "epoch": 0.26570548095971824, + "grad_norm": 0.6019419973135351, + "learning_rate": 8.612492465793969e-06, + "loss": 0.5276, + "step": 12071 + }, + { + "epoch": 0.2657274928461369, + "grad_norm": 1.9581959318068722, + "learning_rate": 8.612246012477554e-06, + "loss": 0.3661, + "step": 12072 + }, + { + "epoch": 0.2657495047325556, + "grad_norm": 0.6173340058487247, + "learning_rate": 8.611999540802092e-06, + "loss": 0.4941, + "step": 12073 + }, + { + "epoch": 0.26577151661897425, + "grad_norm": 3.0247135091692443, + "learning_rate": 8.611753050768837e-06, + "loss": 0.291, + "step": 12074 + }, + { + "epoch": 0.2657935285053929, + "grad_norm": 1.7010983405515203, + "learning_rate": 8.611506542379044e-06, + "loss": 0.3318, + "step": 12075 + }, + { + "epoch": 0.2658155403918116, + "grad_norm": 1.2551416481177566, + "learning_rate": 8.611260015633964e-06, + "loss": 0.3712, + "step": 12076 + }, + { + "epoch": 0.26583755227823025, + "grad_norm": 1.2505358925609806, + "learning_rate": 8.611013470534847e-06, + "loss": 0.3199, + "step": 12077 + }, + { + "epoch": 0.2658595641646489, + "grad_norm": 1.2507891031334974, + "learning_rate": 8.610766907082953e-06, + "loss": 0.3077, + "step": 12078 + }, + { + "epoch": 0.2658815760510676, + "grad_norm": 0.6215603905119684, + "learning_rate": 8.610520325279528e-06, + "loss": 0.5233, + "step": 12079 + }, + { + "epoch": 0.26590358793748625, + "grad_norm": 1.3035422589000527, + "learning_rate": 8.61027372512583e-06, + "loss": 0.2822, + "step": 12080 + }, + { + "epoch": 0.2659255998239049, + "grad_norm": 1.3136979099005544, + "learning_rate": 8.610027106623111e-06, + "loss": 0.2292, + "step": 12081 + }, + { + "epoch": 0.26594761171032355, + "grad_norm": 1.4832274828132044, + "learning_rate": 8.609780469772623e-06, + "loss": 0.3264, + "step": 12082 + }, + { + "epoch": 0.26596962359674226, + "grad_norm": 3.8570421701451267, + "learning_rate": 8.609533814575622e-06, + "loss": 0.4372, + "step": 12083 + }, + { + "epoch": 0.2659916354831609, + "grad_norm": 1.4905297424399706, + "learning_rate": 8.609287141033358e-06, + "loss": 0.4617, + "step": 12084 + }, + { + "epoch": 0.26601364736957955, + "grad_norm": 1.9262013665137812, + "learning_rate": 8.609040449147089e-06, + "loss": 0.3396, + "step": 12085 + }, + { + "epoch": 0.26603565925599826, + "grad_norm": 1.9675975673337405, + "learning_rate": 8.608793738918067e-06, + "loss": 0.4221, + "step": 12086 + }, + { + "epoch": 0.2660576711424169, + "grad_norm": 1.3688727829005838, + "learning_rate": 8.608547010347543e-06, + "loss": 0.2936, + "step": 12087 + }, + { + "epoch": 0.26607968302883556, + "grad_norm": 1.5653876967598213, + "learning_rate": 8.608300263436776e-06, + "loss": 0.3049, + "step": 12088 + }, + { + "epoch": 0.26610169491525426, + "grad_norm": 2.2810476571522194, + "learning_rate": 8.608053498187017e-06, + "loss": 0.4036, + "step": 12089 + }, + { + "epoch": 0.2661237068016729, + "grad_norm": 1.3137472815061033, + "learning_rate": 8.607806714599521e-06, + "loss": 0.3112, + "step": 12090 + }, + { + "epoch": 0.26614571868809156, + "grad_norm": 1.7183345679609585, + "learning_rate": 8.607559912675542e-06, + "loss": 0.3287, + "step": 12091 + }, + { + "epoch": 0.2661677305745102, + "grad_norm": 1.3522129249654518, + "learning_rate": 8.607313092416335e-06, + "loss": 0.2906, + "step": 12092 + }, + { + "epoch": 0.2661897424609289, + "grad_norm": 1.4172385084289933, + "learning_rate": 8.607066253823153e-06, + "loss": 0.3561, + "step": 12093 + }, + { + "epoch": 0.26621175434734756, + "grad_norm": 2.2644670403590013, + "learning_rate": 8.606819396897253e-06, + "loss": 0.3791, + "step": 12094 + }, + { + "epoch": 0.2662337662337662, + "grad_norm": 1.5042008605489552, + "learning_rate": 8.606572521639887e-06, + "loss": 0.4211, + "step": 12095 + }, + { + "epoch": 0.2662557781201849, + "grad_norm": 2.3396174248954997, + "learning_rate": 8.60632562805231e-06, + "loss": 0.3954, + "step": 12096 + }, + { + "epoch": 0.26627779000660357, + "grad_norm": 2.742697009820785, + "learning_rate": 8.60607871613578e-06, + "loss": 0.3165, + "step": 12097 + }, + { + "epoch": 0.2662998018930222, + "grad_norm": 1.4164465371036707, + "learning_rate": 8.605831785891549e-06, + "loss": 0.2638, + "step": 12098 + }, + { + "epoch": 0.2663218137794409, + "grad_norm": 1.687424021213188, + "learning_rate": 8.60558483732087e-06, + "loss": 0.3389, + "step": 12099 + }, + { + "epoch": 0.26634382566585957, + "grad_norm": 1.564659939914018, + "learning_rate": 8.605337870425004e-06, + "loss": 0.2921, + "step": 12100 + }, + { + "epoch": 0.2663658375522782, + "grad_norm": 1.3095265839932202, + "learning_rate": 8.605090885205201e-06, + "loss": 0.3434, + "step": 12101 + }, + { + "epoch": 0.2663878494386969, + "grad_norm": 1.2284438194644451, + "learning_rate": 8.604843881662719e-06, + "loss": 0.3381, + "step": 12102 + }, + { + "epoch": 0.26640986132511557, + "grad_norm": 1.3310955470574597, + "learning_rate": 8.604596859798813e-06, + "loss": 0.2712, + "step": 12103 + }, + { + "epoch": 0.2664318732115342, + "grad_norm": 1.4638143151431, + "learning_rate": 8.604349819614737e-06, + "loss": 0.3535, + "step": 12104 + }, + { + "epoch": 0.26645388509795287, + "grad_norm": 1.5941121202937254, + "learning_rate": 8.604102761111748e-06, + "loss": 0.4373, + "step": 12105 + }, + { + "epoch": 0.2664758969843716, + "grad_norm": 1.664989367960705, + "learning_rate": 8.603855684291102e-06, + "loss": 0.3532, + "step": 12106 + }, + { + "epoch": 0.2664979088707902, + "grad_norm": 1.4694642173581878, + "learning_rate": 8.603608589154053e-06, + "loss": 0.4059, + "step": 12107 + }, + { + "epoch": 0.2665199207572089, + "grad_norm": 1.636602421849589, + "learning_rate": 8.60336147570186e-06, + "loss": 0.2812, + "step": 12108 + }, + { + "epoch": 0.2665419326436276, + "grad_norm": 1.4929144164329506, + "learning_rate": 8.603114343935774e-06, + "loss": 0.35, + "step": 12109 + }, + { + "epoch": 0.2665639445300462, + "grad_norm": 1.2692184956384063, + "learning_rate": 8.602867193857054e-06, + "loss": 0.3242, + "step": 12110 + }, + { + "epoch": 0.2665859564164649, + "grad_norm": 1.3371720096133533, + "learning_rate": 8.602620025466958e-06, + "loss": 0.4386, + "step": 12111 + }, + { + "epoch": 0.2666079683028836, + "grad_norm": 1.224994922427359, + "learning_rate": 8.60237283876674e-06, + "loss": 0.3179, + "step": 12112 + }, + { + "epoch": 0.26662998018930223, + "grad_norm": 1.4020401649350758, + "learning_rate": 8.602125633757656e-06, + "loss": 0.3811, + "step": 12113 + }, + { + "epoch": 0.2666519920757209, + "grad_norm": 1.5555469802796777, + "learning_rate": 8.601878410440963e-06, + "loss": 0.4309, + "step": 12114 + }, + { + "epoch": 0.2666740039621396, + "grad_norm": 1.1779321442193347, + "learning_rate": 8.601631168817917e-06, + "loss": 0.2664, + "step": 12115 + }, + { + "epoch": 0.26669601584855823, + "grad_norm": 0.6722060710373821, + "learning_rate": 8.601383908889775e-06, + "loss": 0.533, + "step": 12116 + }, + { + "epoch": 0.2667180277349769, + "grad_norm": 1.1453205446101804, + "learning_rate": 8.601136630657793e-06, + "loss": 0.2698, + "step": 12117 + }, + { + "epoch": 0.26674003962139553, + "grad_norm": 2.583823069434546, + "learning_rate": 8.60088933412323e-06, + "loss": 0.2947, + "step": 12118 + }, + { + "epoch": 0.26676205150781424, + "grad_norm": 1.1741604719841339, + "learning_rate": 8.60064201928734e-06, + "loss": 0.2818, + "step": 12119 + }, + { + "epoch": 0.2667840633942329, + "grad_norm": 2.422551364860319, + "learning_rate": 8.600394686151384e-06, + "loss": 0.3731, + "step": 12120 + }, + { + "epoch": 0.26680607528065153, + "grad_norm": 1.4120807034990273, + "learning_rate": 8.600147334716614e-06, + "loss": 0.381, + "step": 12121 + }, + { + "epoch": 0.26682808716707024, + "grad_norm": 1.2258530167835497, + "learning_rate": 8.599899964984289e-06, + "loss": 0.2762, + "step": 12122 + }, + { + "epoch": 0.2668500990534889, + "grad_norm": 1.8539478421372222, + "learning_rate": 8.599652576955669e-06, + "loss": 0.3428, + "step": 12123 + }, + { + "epoch": 0.26687211093990754, + "grad_norm": 1.98215055570124, + "learning_rate": 8.599405170632007e-06, + "loss": 0.3696, + "step": 12124 + }, + { + "epoch": 0.26689412282632624, + "grad_norm": 3.19865894453685, + "learning_rate": 8.599157746014564e-06, + "loss": 0.3502, + "step": 12125 + }, + { + "epoch": 0.2669161347127449, + "grad_norm": 1.6052428763238162, + "learning_rate": 8.598910303104595e-06, + "loss": 0.364, + "step": 12126 + }, + { + "epoch": 0.26693814659916354, + "grad_norm": 1.3030639215222246, + "learning_rate": 8.598662841903359e-06, + "loss": 0.2731, + "step": 12127 + }, + { + "epoch": 0.2669601584855822, + "grad_norm": 1.3335749723069028, + "learning_rate": 8.598415362412113e-06, + "loss": 0.325, + "step": 12128 + }, + { + "epoch": 0.2669821703720009, + "grad_norm": 5.051074837635476, + "learning_rate": 8.598167864632117e-06, + "loss": 0.393, + "step": 12129 + }, + { + "epoch": 0.26700418225841954, + "grad_norm": 1.419065803250348, + "learning_rate": 8.597920348564626e-06, + "loss": 0.2812, + "step": 12130 + }, + { + "epoch": 0.2670261941448382, + "grad_norm": 1.6343572076647375, + "learning_rate": 8.597672814210897e-06, + "loss": 0.3425, + "step": 12131 + }, + { + "epoch": 0.2670482060312569, + "grad_norm": 0.7199131118397484, + "learning_rate": 8.597425261572194e-06, + "loss": 0.5214, + "step": 12132 + }, + { + "epoch": 0.26707021791767555, + "grad_norm": 1.406584172372602, + "learning_rate": 8.597177690649769e-06, + "loss": 0.3653, + "step": 12133 + }, + { + "epoch": 0.2670922298040942, + "grad_norm": 1.7900762582661789, + "learning_rate": 8.596930101444882e-06, + "loss": 0.3844, + "step": 12134 + }, + { + "epoch": 0.2671142416905129, + "grad_norm": 1.487279110666471, + "learning_rate": 8.596682493958794e-06, + "loss": 0.4034, + "step": 12135 + }, + { + "epoch": 0.26713625357693155, + "grad_norm": 1.4616090608670365, + "learning_rate": 8.59643486819276e-06, + "loss": 0.3812, + "step": 12136 + }, + { + "epoch": 0.2671582654633502, + "grad_norm": 1.2417654320796223, + "learning_rate": 8.59618722414804e-06, + "loss": 0.3329, + "step": 12137 + }, + { + "epoch": 0.2671802773497689, + "grad_norm": 0.6284898366398718, + "learning_rate": 8.595939561825894e-06, + "loss": 0.511, + "step": 12138 + }, + { + "epoch": 0.26720228923618755, + "grad_norm": 0.6458705814010616, + "learning_rate": 8.595691881227578e-06, + "loss": 0.5205, + "step": 12139 + }, + { + "epoch": 0.2672243011226062, + "grad_norm": 1.3146374264462957, + "learning_rate": 8.595444182354351e-06, + "loss": 0.3358, + "step": 12140 + }, + { + "epoch": 0.26724631300902485, + "grad_norm": 1.2407617691006048, + "learning_rate": 8.595196465207475e-06, + "loss": 0.3687, + "step": 12141 + }, + { + "epoch": 0.26726832489544355, + "grad_norm": 1.6388855121015835, + "learning_rate": 8.594948729788206e-06, + "loss": 0.3393, + "step": 12142 + }, + { + "epoch": 0.2672903367818622, + "grad_norm": 1.3452302791863273, + "learning_rate": 8.594700976097805e-06, + "loss": 0.3862, + "step": 12143 + }, + { + "epoch": 0.26731234866828085, + "grad_norm": 2.195805039618778, + "learning_rate": 8.59445320413753e-06, + "loss": 0.3089, + "step": 12144 + }, + { + "epoch": 0.26733436055469956, + "grad_norm": 1.416575146549733, + "learning_rate": 8.59420541390864e-06, + "loss": 0.3575, + "step": 12145 + }, + { + "epoch": 0.2673563724411182, + "grad_norm": 1.4399726208619157, + "learning_rate": 8.593957605412398e-06, + "loss": 0.3324, + "step": 12146 + }, + { + "epoch": 0.26737838432753686, + "grad_norm": 1.1932295083599664, + "learning_rate": 8.593709778650056e-06, + "loss": 0.2981, + "step": 12147 + }, + { + "epoch": 0.26740039621395556, + "grad_norm": 1.7248464838993292, + "learning_rate": 8.593461933622882e-06, + "loss": 0.4004, + "step": 12148 + }, + { + "epoch": 0.2674224081003742, + "grad_norm": 2.2592359056918934, + "learning_rate": 8.59321407033213e-06, + "loss": 0.3577, + "step": 12149 + }, + { + "epoch": 0.26744441998679286, + "grad_norm": 2.0425553421761453, + "learning_rate": 8.59296618877906e-06, + "loss": 0.35, + "step": 12150 + }, + { + "epoch": 0.2674664318732115, + "grad_norm": 1.7349600723118954, + "learning_rate": 8.592718288964935e-06, + "loss": 0.4107, + "step": 12151 + }, + { + "epoch": 0.2674884437596302, + "grad_norm": 1.8026331969500624, + "learning_rate": 8.592470370891014e-06, + "loss": 0.3391, + "step": 12152 + }, + { + "epoch": 0.26751045564604886, + "grad_norm": 1.40759380025347, + "learning_rate": 8.592222434558555e-06, + "loss": 0.3706, + "step": 12153 + }, + { + "epoch": 0.2675324675324675, + "grad_norm": 1.6945604980241022, + "learning_rate": 8.591974479968818e-06, + "loss": 0.3879, + "step": 12154 + }, + { + "epoch": 0.2675544794188862, + "grad_norm": 1.277749779501271, + "learning_rate": 8.591726507123068e-06, + "loss": 0.3144, + "step": 12155 + }, + { + "epoch": 0.26757649130530486, + "grad_norm": 1.1343667784364682, + "learning_rate": 8.59147851602256e-06, + "loss": 0.2463, + "step": 12156 + }, + { + "epoch": 0.2675985031917235, + "grad_norm": 1.944632346598789, + "learning_rate": 8.591230506668557e-06, + "loss": 0.3299, + "step": 12157 + }, + { + "epoch": 0.2676205150781422, + "grad_norm": 1.4389889730823942, + "learning_rate": 8.590982479062318e-06, + "loss": 0.4005, + "step": 12158 + }, + { + "epoch": 0.26764252696456087, + "grad_norm": 1.4538760888606836, + "learning_rate": 8.590734433205104e-06, + "loss": 0.3455, + "step": 12159 + }, + { + "epoch": 0.2676645388509795, + "grad_norm": 1.4088579177468923, + "learning_rate": 8.590486369098178e-06, + "loss": 0.2678, + "step": 12160 + }, + { + "epoch": 0.2676865507373982, + "grad_norm": 0.7096575370658912, + "learning_rate": 8.590238286742798e-06, + "loss": 0.538, + "step": 12161 + }, + { + "epoch": 0.26770856262381687, + "grad_norm": 1.8221504474919066, + "learning_rate": 8.589990186140225e-06, + "loss": 0.3288, + "step": 12162 + }, + { + "epoch": 0.2677305745102355, + "grad_norm": 0.6371820290723081, + "learning_rate": 8.589742067291721e-06, + "loss": 0.499, + "step": 12163 + }, + { + "epoch": 0.26775258639665417, + "grad_norm": 1.3986364870726016, + "learning_rate": 8.589493930198547e-06, + "loss": 0.295, + "step": 12164 + }, + { + "epoch": 0.2677745982830729, + "grad_norm": 0.645042042989047, + "learning_rate": 8.589245774861965e-06, + "loss": 0.5513, + "step": 12165 + }, + { + "epoch": 0.2677966101694915, + "grad_norm": 1.4978689092959965, + "learning_rate": 8.588997601283233e-06, + "loss": 0.3027, + "step": 12166 + }, + { + "epoch": 0.26781862205591017, + "grad_norm": 1.3992349234445889, + "learning_rate": 8.588749409463615e-06, + "loss": 0.3049, + "step": 12167 + }, + { + "epoch": 0.2678406339423289, + "grad_norm": 1.6324062749909052, + "learning_rate": 8.588501199404373e-06, + "loss": 0.4548, + "step": 12168 + }, + { + "epoch": 0.2678626458287475, + "grad_norm": 0.6831479466431305, + "learning_rate": 8.588252971106768e-06, + "loss": 0.5391, + "step": 12169 + }, + { + "epoch": 0.2678846577151662, + "grad_norm": 1.4695122362545263, + "learning_rate": 8.588004724572058e-06, + "loss": 0.3921, + "step": 12170 + }, + { + "epoch": 0.2679066696015849, + "grad_norm": 1.1409628596012606, + "learning_rate": 8.58775645980151e-06, + "loss": 0.3764, + "step": 12171 + }, + { + "epoch": 0.26792868148800353, + "grad_norm": 10.768390622136724, + "learning_rate": 8.587508176796382e-06, + "loss": 0.2671, + "step": 12172 + }, + { + "epoch": 0.2679506933744222, + "grad_norm": 1.6836133341389647, + "learning_rate": 8.587259875557938e-06, + "loss": 0.4375, + "step": 12173 + }, + { + "epoch": 0.2679727052608408, + "grad_norm": 1.5121438460568044, + "learning_rate": 8.587011556087441e-06, + "loss": 0.3479, + "step": 12174 + }, + { + "epoch": 0.26799471714725953, + "grad_norm": 1.6635104224407447, + "learning_rate": 8.586763218386149e-06, + "loss": 0.2768, + "step": 12175 + }, + { + "epoch": 0.2680167290336782, + "grad_norm": 1.2926508110973556, + "learning_rate": 8.586514862455328e-06, + "loss": 0.2918, + "step": 12176 + }, + { + "epoch": 0.26803874092009683, + "grad_norm": 0.654592883015987, + "learning_rate": 8.586266488296237e-06, + "loss": 0.551, + "step": 12177 + }, + { + "epoch": 0.26806075280651553, + "grad_norm": 1.2860997269948293, + "learning_rate": 8.586018095910143e-06, + "loss": 0.337, + "step": 12178 + }, + { + "epoch": 0.2680827646929342, + "grad_norm": 0.6273975092192032, + "learning_rate": 8.585769685298305e-06, + "loss": 0.5195, + "step": 12179 + }, + { + "epoch": 0.26810477657935283, + "grad_norm": 2.123198465529158, + "learning_rate": 8.585521256461985e-06, + "loss": 0.3047, + "step": 12180 + }, + { + "epoch": 0.26812678846577154, + "grad_norm": 0.6167783983759407, + "learning_rate": 8.585272809402447e-06, + "loss": 0.4898, + "step": 12181 + }, + { + "epoch": 0.2681488003521902, + "grad_norm": 1.6401498034352884, + "learning_rate": 8.585024344120954e-06, + "loss": 0.3573, + "step": 12182 + }, + { + "epoch": 0.26817081223860884, + "grad_norm": 1.732502338427586, + "learning_rate": 8.584775860618767e-06, + "loss": 0.3834, + "step": 12183 + }, + { + "epoch": 0.26819282412502754, + "grad_norm": 0.6577580684900701, + "learning_rate": 8.584527358897153e-06, + "loss": 0.5195, + "step": 12184 + }, + { + "epoch": 0.2682148360114462, + "grad_norm": 1.469138974543317, + "learning_rate": 8.584278838957369e-06, + "loss": 0.3589, + "step": 12185 + }, + { + "epoch": 0.26823684789786484, + "grad_norm": 1.3959704673394486, + "learning_rate": 8.584030300800683e-06, + "loss": 0.3225, + "step": 12186 + }, + { + "epoch": 0.2682588597842835, + "grad_norm": 6.219000389727037, + "learning_rate": 8.583781744428357e-06, + "loss": 0.3679, + "step": 12187 + }, + { + "epoch": 0.2682808716707022, + "grad_norm": 1.4455967746113025, + "learning_rate": 8.583533169841653e-06, + "loss": 0.3637, + "step": 12188 + }, + { + "epoch": 0.26830288355712084, + "grad_norm": 1.5076664081547406, + "learning_rate": 8.583284577041835e-06, + "loss": 0.3209, + "step": 12189 + }, + { + "epoch": 0.2683248954435395, + "grad_norm": 1.4266788022498493, + "learning_rate": 8.583035966030166e-06, + "loss": 0.3824, + "step": 12190 + }, + { + "epoch": 0.2683469073299582, + "grad_norm": 1.3685709901280698, + "learning_rate": 8.582787336807911e-06, + "loss": 0.3526, + "step": 12191 + }, + { + "epoch": 0.26836891921637684, + "grad_norm": 1.1875939640488917, + "learning_rate": 8.582538689376335e-06, + "loss": 0.2814, + "step": 12192 + }, + { + "epoch": 0.2683909311027955, + "grad_norm": 0.7152703707111987, + "learning_rate": 8.582290023736696e-06, + "loss": 0.5249, + "step": 12193 + }, + { + "epoch": 0.2684129429892142, + "grad_norm": 1.3264783636300614, + "learning_rate": 8.582041339890263e-06, + "loss": 0.3748, + "step": 12194 + }, + { + "epoch": 0.26843495487563285, + "grad_norm": 1.5166756315411178, + "learning_rate": 8.581792637838298e-06, + "loss": 0.3507, + "step": 12195 + }, + { + "epoch": 0.2684569667620515, + "grad_norm": 0.6332679245178355, + "learning_rate": 8.581543917582064e-06, + "loss": 0.5248, + "step": 12196 + }, + { + "epoch": 0.2684789786484702, + "grad_norm": 1.5298329513034876, + "learning_rate": 8.581295179122828e-06, + "loss": 0.297, + "step": 12197 + }, + { + "epoch": 0.26850099053488885, + "grad_norm": 2.5536485738056607, + "learning_rate": 8.581046422461852e-06, + "loss": 0.3537, + "step": 12198 + }, + { + "epoch": 0.2685230024213075, + "grad_norm": 1.5176001725318018, + "learning_rate": 8.580797647600401e-06, + "loss": 0.3344, + "step": 12199 + }, + { + "epoch": 0.26854501430772615, + "grad_norm": 2.412843928800443, + "learning_rate": 8.58054885453974e-06, + "loss": 0.3555, + "step": 12200 + }, + { + "epoch": 0.26856702619414485, + "grad_norm": 1.430182902781078, + "learning_rate": 8.580300043281132e-06, + "loss": 0.378, + "step": 12201 + }, + { + "epoch": 0.2685890380805635, + "grad_norm": 1.1598120200681405, + "learning_rate": 8.580051213825841e-06, + "loss": 0.3141, + "step": 12202 + }, + { + "epoch": 0.26861104996698215, + "grad_norm": 1.4024236162827473, + "learning_rate": 8.579802366175136e-06, + "loss": 0.2968, + "step": 12203 + }, + { + "epoch": 0.26863306185340086, + "grad_norm": 1.4144422783127428, + "learning_rate": 8.579553500330276e-06, + "loss": 0.3277, + "step": 12204 + }, + { + "epoch": 0.2686550737398195, + "grad_norm": 3.1416928238014474, + "learning_rate": 8.57930461629253e-06, + "loss": 0.3173, + "step": 12205 + }, + { + "epoch": 0.26867708562623815, + "grad_norm": 1.388126507844789, + "learning_rate": 8.57905571406316e-06, + "loss": 0.3437, + "step": 12206 + }, + { + "epoch": 0.26869909751265686, + "grad_norm": 1.7356653913721205, + "learning_rate": 8.578806793643435e-06, + "loss": 0.3412, + "step": 12207 + }, + { + "epoch": 0.2687211093990755, + "grad_norm": 1.4332987491723628, + "learning_rate": 8.578557855034615e-06, + "loss": 0.3546, + "step": 12208 + }, + { + "epoch": 0.26874312128549416, + "grad_norm": 1.477152291920005, + "learning_rate": 8.57830889823797e-06, + "loss": 0.379, + "step": 12209 + }, + { + "epoch": 0.2687651331719128, + "grad_norm": 1.3421310609946975, + "learning_rate": 8.578059923254763e-06, + "loss": 0.363, + "step": 12210 + }, + { + "epoch": 0.2687871450583315, + "grad_norm": 0.6823426611437993, + "learning_rate": 8.57781093008626e-06, + "loss": 0.5192, + "step": 12211 + }, + { + "epoch": 0.26880915694475016, + "grad_norm": 1.745248226031043, + "learning_rate": 8.577561918733724e-06, + "loss": 0.3482, + "step": 12212 + }, + { + "epoch": 0.2688311688311688, + "grad_norm": 1.2240807662874915, + "learning_rate": 8.577312889198423e-06, + "loss": 0.3383, + "step": 12213 + }, + { + "epoch": 0.2688531807175875, + "grad_norm": 1.3933524936012163, + "learning_rate": 8.577063841481626e-06, + "loss": 0.3595, + "step": 12214 + }, + { + "epoch": 0.26887519260400616, + "grad_norm": 0.6086016807439111, + "learning_rate": 8.576814775584592e-06, + "loss": 0.5048, + "step": 12215 + }, + { + "epoch": 0.2688972044904248, + "grad_norm": 1.4603758580455384, + "learning_rate": 8.576565691508589e-06, + "loss": 0.3321, + "step": 12216 + }, + { + "epoch": 0.2689192163768435, + "grad_norm": 2.272254292706165, + "learning_rate": 8.576316589254887e-06, + "loss": 0.3917, + "step": 12217 + }, + { + "epoch": 0.26894122826326217, + "grad_norm": 1.2788150382974048, + "learning_rate": 8.576067468824748e-06, + "loss": 0.3017, + "step": 12218 + }, + { + "epoch": 0.2689632401496808, + "grad_norm": 1.6264566187245202, + "learning_rate": 8.575818330219438e-06, + "loss": 0.3787, + "step": 12219 + }, + { + "epoch": 0.2689852520360995, + "grad_norm": 1.4244188174536414, + "learning_rate": 8.575569173440223e-06, + "loss": 0.2917, + "step": 12220 + }, + { + "epoch": 0.26900726392251817, + "grad_norm": 1.2378291835781998, + "learning_rate": 8.575319998488374e-06, + "loss": 0.3426, + "step": 12221 + }, + { + "epoch": 0.2690292758089368, + "grad_norm": 1.7035436466357263, + "learning_rate": 8.575070805365154e-06, + "loss": 0.3994, + "step": 12222 + }, + { + "epoch": 0.26905128769535547, + "grad_norm": 1.2824108817254942, + "learning_rate": 8.574821594071827e-06, + "loss": 0.3799, + "step": 12223 + }, + { + "epoch": 0.2690732995817742, + "grad_norm": 2.245641471590263, + "learning_rate": 8.574572364609663e-06, + "loss": 0.3334, + "step": 12224 + }, + { + "epoch": 0.2690953114681928, + "grad_norm": 2.0770255064665633, + "learning_rate": 8.57432311697993e-06, + "loss": 0.2876, + "step": 12225 + }, + { + "epoch": 0.26911732335461147, + "grad_norm": 0.6497246518945443, + "learning_rate": 8.57407385118389e-06, + "loss": 0.5225, + "step": 12226 + }, + { + "epoch": 0.2691393352410302, + "grad_norm": 0.6480969041309714, + "learning_rate": 8.573824567222814e-06, + "loss": 0.525, + "step": 12227 + }, + { + "epoch": 0.2691613471274488, + "grad_norm": 1.5233159707956563, + "learning_rate": 8.573575265097969e-06, + "loss": 0.3013, + "step": 12228 + }, + { + "epoch": 0.2691833590138675, + "grad_norm": 2.384791905603764, + "learning_rate": 8.573325944810619e-06, + "loss": 0.2471, + "step": 12229 + }, + { + "epoch": 0.2692053709002862, + "grad_norm": 1.4816400166927624, + "learning_rate": 8.573076606362032e-06, + "loss": 0.3675, + "step": 12230 + }, + { + "epoch": 0.2692273827867048, + "grad_norm": 0.6153320744049462, + "learning_rate": 8.572827249753478e-06, + "loss": 0.5384, + "step": 12231 + }, + { + "epoch": 0.2692493946731235, + "grad_norm": 0.6158416510140364, + "learning_rate": 8.57257787498622e-06, + "loss": 0.521, + "step": 12232 + }, + { + "epoch": 0.2692714065595421, + "grad_norm": 0.6004884520489455, + "learning_rate": 8.572328482061532e-06, + "loss": 0.5007, + "step": 12233 + }, + { + "epoch": 0.26929341844596083, + "grad_norm": 0.6228550265254775, + "learning_rate": 8.572079070980673e-06, + "loss": 0.5402, + "step": 12234 + }, + { + "epoch": 0.2693154303323795, + "grad_norm": 1.231337833537501, + "learning_rate": 8.571829641744918e-06, + "loss": 0.3239, + "step": 12235 + }, + { + "epoch": 0.26933744221879813, + "grad_norm": 1.2326001456623132, + "learning_rate": 8.57158019435553e-06, + "loss": 0.2614, + "step": 12236 + }, + { + "epoch": 0.26935945410521683, + "grad_norm": 0.6245695379083906, + "learning_rate": 8.571330728813778e-06, + "loss": 0.515, + "step": 12237 + }, + { + "epoch": 0.2693814659916355, + "grad_norm": 1.4389236264779834, + "learning_rate": 8.571081245120931e-06, + "loss": 0.3203, + "step": 12238 + }, + { + "epoch": 0.26940347787805413, + "grad_norm": 1.4999738453715692, + "learning_rate": 8.570831743278256e-06, + "loss": 0.3038, + "step": 12239 + }, + { + "epoch": 0.26942548976447284, + "grad_norm": 1.3515415891534652, + "learning_rate": 8.570582223287022e-06, + "loss": 0.3022, + "step": 12240 + }, + { + "epoch": 0.2694475016508915, + "grad_norm": 1.4843918326290384, + "learning_rate": 8.570332685148497e-06, + "loss": 0.3582, + "step": 12241 + }, + { + "epoch": 0.26946951353731013, + "grad_norm": 3.315485960540202, + "learning_rate": 8.570083128863948e-06, + "loss": 0.3679, + "step": 12242 + }, + { + "epoch": 0.26949152542372884, + "grad_norm": 1.5482270493837382, + "learning_rate": 8.569833554434646e-06, + "loss": 0.3394, + "step": 12243 + }, + { + "epoch": 0.2695135373101475, + "grad_norm": 0.6439582553938166, + "learning_rate": 8.569583961861857e-06, + "loss": 0.5145, + "step": 12244 + }, + { + "epoch": 0.26953554919656614, + "grad_norm": 1.098664845999722, + "learning_rate": 8.569334351146848e-06, + "loss": 0.2912, + "step": 12245 + }, + { + "epoch": 0.2695575610829848, + "grad_norm": 1.399578453412901, + "learning_rate": 8.569084722290892e-06, + "loss": 0.3901, + "step": 12246 + }, + { + "epoch": 0.2695795729694035, + "grad_norm": 1.5470199868962162, + "learning_rate": 8.568835075295255e-06, + "loss": 0.3227, + "step": 12247 + }, + { + "epoch": 0.26960158485582214, + "grad_norm": 1.6241388769644536, + "learning_rate": 8.568585410161207e-06, + "loss": 0.2781, + "step": 12248 + }, + { + "epoch": 0.2696235967422408, + "grad_norm": 1.6548716122114455, + "learning_rate": 8.568335726890014e-06, + "loss": 0.3656, + "step": 12249 + }, + { + "epoch": 0.2696456086286595, + "grad_norm": 1.1813200466050335, + "learning_rate": 8.568086025482951e-06, + "loss": 0.3245, + "step": 12250 + }, + { + "epoch": 0.26966762051507814, + "grad_norm": 1.9844551132701909, + "learning_rate": 8.56783630594128e-06, + "loss": 0.4237, + "step": 12251 + }, + { + "epoch": 0.2696896324014968, + "grad_norm": 1.2732458539840927, + "learning_rate": 8.567586568266275e-06, + "loss": 0.3183, + "step": 12252 + }, + { + "epoch": 0.2697116442879155, + "grad_norm": 0.6457324527567802, + "learning_rate": 8.567336812459206e-06, + "loss": 0.5098, + "step": 12253 + }, + { + "epoch": 0.26973365617433415, + "grad_norm": 2.4479964748992584, + "learning_rate": 8.567087038521336e-06, + "loss": 0.2842, + "step": 12254 + }, + { + "epoch": 0.2697556680607528, + "grad_norm": 1.8756055857111271, + "learning_rate": 8.566837246453942e-06, + "loss": 0.3736, + "step": 12255 + }, + { + "epoch": 0.26977767994717144, + "grad_norm": 3.190346514809463, + "learning_rate": 8.566587436258288e-06, + "loss": 0.3156, + "step": 12256 + }, + { + "epoch": 0.26979969183359015, + "grad_norm": 2.498166721271067, + "learning_rate": 8.566337607935647e-06, + "loss": 0.4011, + "step": 12257 + }, + { + "epoch": 0.2698217037200088, + "grad_norm": 1.4795448887264087, + "learning_rate": 8.566087761487289e-06, + "loss": 0.2778, + "step": 12258 + }, + { + "epoch": 0.26984371560642745, + "grad_norm": 0.636548981663465, + "learning_rate": 8.56583789691448e-06, + "loss": 0.5202, + "step": 12259 + }, + { + "epoch": 0.26986572749284615, + "grad_norm": 1.3831825635406256, + "learning_rate": 8.565588014218494e-06, + "loss": 0.4258, + "step": 12260 + }, + { + "epoch": 0.2698877393792648, + "grad_norm": 1.9696580971608995, + "learning_rate": 8.565338113400598e-06, + "loss": 0.3055, + "step": 12261 + }, + { + "epoch": 0.26990975126568345, + "grad_norm": 1.6539642311346128, + "learning_rate": 8.565088194462065e-06, + "loss": 0.4692, + "step": 12262 + }, + { + "epoch": 0.26993176315210216, + "grad_norm": 1.473673288961429, + "learning_rate": 8.564838257404165e-06, + "loss": 0.2881, + "step": 12263 + }, + { + "epoch": 0.2699537750385208, + "grad_norm": 1.3270644759742103, + "learning_rate": 8.564588302228166e-06, + "loss": 0.4114, + "step": 12264 + }, + { + "epoch": 0.26997578692493945, + "grad_norm": 1.4861703412605645, + "learning_rate": 8.564338328935337e-06, + "loss": 0.3222, + "step": 12265 + }, + { + "epoch": 0.26999779881135816, + "grad_norm": 0.6383357462927565, + "learning_rate": 8.564088337526953e-06, + "loss": 0.5017, + "step": 12266 + }, + { + "epoch": 0.2700198106977768, + "grad_norm": 1.6818083968106097, + "learning_rate": 8.563838328004281e-06, + "loss": 0.3008, + "step": 12267 + }, + { + "epoch": 0.27004182258419546, + "grad_norm": 1.3546290804223398, + "learning_rate": 8.563588300368596e-06, + "loss": 0.3138, + "step": 12268 + }, + { + "epoch": 0.2700638344706141, + "grad_norm": 1.3175440773816367, + "learning_rate": 8.563338254621162e-06, + "loss": 0.3275, + "step": 12269 + }, + { + "epoch": 0.2700858463570328, + "grad_norm": 1.3080626280455374, + "learning_rate": 8.563088190763256e-06, + "loss": 0.3211, + "step": 12270 + }, + { + "epoch": 0.27010785824345146, + "grad_norm": 1.31342249775132, + "learning_rate": 8.562838108796147e-06, + "loss": 0.3602, + "step": 12271 + }, + { + "epoch": 0.2701298701298701, + "grad_norm": 1.5848046799941076, + "learning_rate": 8.562588008721104e-06, + "loss": 0.3407, + "step": 12272 + }, + { + "epoch": 0.2701518820162888, + "grad_norm": 1.7527076239322594, + "learning_rate": 8.562337890539401e-06, + "loss": 0.3821, + "step": 12273 + }, + { + "epoch": 0.27017389390270746, + "grad_norm": 0.6264374987283912, + "learning_rate": 8.562087754252309e-06, + "loss": 0.5285, + "step": 12274 + }, + { + "epoch": 0.2701959057891261, + "grad_norm": 2.2258940585086457, + "learning_rate": 8.561837599861096e-06, + "loss": 0.3516, + "step": 12275 + }, + { + "epoch": 0.2702179176755448, + "grad_norm": 0.6583017976081745, + "learning_rate": 8.561587427367037e-06, + "loss": 0.5415, + "step": 12276 + }, + { + "epoch": 0.27023992956196347, + "grad_norm": 2.012094587241264, + "learning_rate": 8.561337236771401e-06, + "loss": 0.3907, + "step": 12277 + }, + { + "epoch": 0.2702619414483821, + "grad_norm": 1.869823457858634, + "learning_rate": 8.561087028075462e-06, + "loss": 0.3889, + "step": 12278 + }, + { + "epoch": 0.2702839533348008, + "grad_norm": 1.3859050280641632, + "learning_rate": 8.56083680128049e-06, + "loss": 0.3076, + "step": 12279 + }, + { + "epoch": 0.27030596522121947, + "grad_norm": 0.6277079359749019, + "learning_rate": 8.560586556387758e-06, + "loss": 0.5255, + "step": 12280 + }, + { + "epoch": 0.2703279771076381, + "grad_norm": 1.5702128067408585, + "learning_rate": 8.560336293398535e-06, + "loss": 0.2979, + "step": 12281 + }, + { + "epoch": 0.27034998899405677, + "grad_norm": 0.5856990713024046, + "learning_rate": 8.560086012314097e-06, + "loss": 0.4971, + "step": 12282 + }, + { + "epoch": 0.27037200088047547, + "grad_norm": 1.2122430185248971, + "learning_rate": 8.559835713135713e-06, + "loss": 0.3175, + "step": 12283 + }, + { + "epoch": 0.2703940127668941, + "grad_norm": 0.6121792898742952, + "learning_rate": 8.559585395864657e-06, + "loss": 0.4999, + "step": 12284 + }, + { + "epoch": 0.27041602465331277, + "grad_norm": 1.3713772563361686, + "learning_rate": 8.5593350605022e-06, + "loss": 0.2743, + "step": 12285 + }, + { + "epoch": 0.2704380365397315, + "grad_norm": 1.4599949578552804, + "learning_rate": 8.559084707049615e-06, + "loss": 0.465, + "step": 12286 + }, + { + "epoch": 0.2704600484261501, + "grad_norm": 1.8958980857212044, + "learning_rate": 8.558834335508173e-06, + "loss": 0.3916, + "step": 12287 + }, + { + "epoch": 0.2704820603125688, + "grad_norm": 1.789030200043744, + "learning_rate": 8.55858394587915e-06, + "loss": 0.3214, + "step": 12288 + }, + { + "epoch": 0.2705040721989875, + "grad_norm": 0.6474117671313967, + "learning_rate": 8.558333538163814e-06, + "loss": 0.5116, + "step": 12289 + }, + { + "epoch": 0.2705260840854061, + "grad_norm": 1.7605250903616376, + "learning_rate": 8.55808311236344e-06, + "loss": 0.4508, + "step": 12290 + }, + { + "epoch": 0.2705480959718248, + "grad_norm": 1.6533837456119682, + "learning_rate": 8.5578326684793e-06, + "loss": 0.318, + "step": 12291 + }, + { + "epoch": 0.2705701078582434, + "grad_norm": 1.8909492535522, + "learning_rate": 8.557582206512668e-06, + "loss": 0.3908, + "step": 12292 + }, + { + "epoch": 0.27059211974466213, + "grad_norm": 2.8690543193438973, + "learning_rate": 8.557331726464819e-06, + "loss": 0.2901, + "step": 12293 + }, + { + "epoch": 0.2706141316310808, + "grad_norm": 1.4246409958379216, + "learning_rate": 8.557081228337022e-06, + "loss": 0.3833, + "step": 12294 + }, + { + "epoch": 0.27063614351749943, + "grad_norm": 1.3613620439572074, + "learning_rate": 8.556830712130551e-06, + "loss": 0.3496, + "step": 12295 + }, + { + "epoch": 0.27065815540391813, + "grad_norm": 1.8061745364772517, + "learning_rate": 8.55658017784668e-06, + "loss": 0.2935, + "step": 12296 + }, + { + "epoch": 0.2706801672903368, + "grad_norm": 1.3839301653376488, + "learning_rate": 8.556329625486683e-06, + "loss": 0.304, + "step": 12297 + }, + { + "epoch": 0.27070217917675543, + "grad_norm": 1.4359090121167943, + "learning_rate": 8.556079055051833e-06, + "loss": 0.3771, + "step": 12298 + }, + { + "epoch": 0.27072419106317414, + "grad_norm": 1.6290389901650346, + "learning_rate": 8.555828466543403e-06, + "loss": 0.3795, + "step": 12299 + }, + { + "epoch": 0.2707462029495928, + "grad_norm": 1.4960693891616028, + "learning_rate": 8.555577859962665e-06, + "loss": 0.2792, + "step": 12300 + }, + { + "epoch": 0.27076821483601143, + "grad_norm": 1.7700361431529539, + "learning_rate": 8.555327235310896e-06, + "loss": 0.4193, + "step": 12301 + }, + { + "epoch": 0.27079022672243014, + "grad_norm": 1.9121096774958855, + "learning_rate": 8.55507659258937e-06, + "loss": 0.3419, + "step": 12302 + }, + { + "epoch": 0.2708122386088488, + "grad_norm": 2.1426396471804927, + "learning_rate": 8.554825931799356e-06, + "loss": 0.3638, + "step": 12303 + }, + { + "epoch": 0.27083425049526744, + "grad_norm": 2.238813689602138, + "learning_rate": 8.554575252942132e-06, + "loss": 0.3372, + "step": 12304 + }, + { + "epoch": 0.2708562623816861, + "grad_norm": 1.6599233925191192, + "learning_rate": 8.554324556018972e-06, + "loss": 0.3395, + "step": 12305 + }, + { + "epoch": 0.2708782742681048, + "grad_norm": 1.5343408258117401, + "learning_rate": 8.554073841031147e-06, + "loss": 0.3735, + "step": 12306 + }, + { + "epoch": 0.27090028615452344, + "grad_norm": 1.6770767012492749, + "learning_rate": 8.553823107979937e-06, + "loss": 0.3627, + "step": 12307 + }, + { + "epoch": 0.2709222980409421, + "grad_norm": 1.5038271805806402, + "learning_rate": 8.553572356866611e-06, + "loss": 0.359, + "step": 12308 + }, + { + "epoch": 0.2709443099273608, + "grad_norm": 1.9778724232356353, + "learning_rate": 8.553321587692444e-06, + "loss": 0.399, + "step": 12309 + }, + { + "epoch": 0.27096632181377944, + "grad_norm": 1.4650225882058991, + "learning_rate": 8.553070800458714e-06, + "loss": 0.2792, + "step": 12310 + }, + { + "epoch": 0.2709883337001981, + "grad_norm": 1.3221470277474283, + "learning_rate": 8.552819995166695e-06, + "loss": 0.2218, + "step": 12311 + }, + { + "epoch": 0.2710103455866168, + "grad_norm": 1.3661483334334108, + "learning_rate": 8.552569171817657e-06, + "loss": 0.3275, + "step": 12312 + }, + { + "epoch": 0.27103235747303545, + "grad_norm": 1.4225391822906859, + "learning_rate": 8.552318330412879e-06, + "loss": 0.3508, + "step": 12313 + }, + { + "epoch": 0.2710543693594541, + "grad_norm": 0.6392029599584707, + "learning_rate": 8.552067470953636e-06, + "loss": 0.5084, + "step": 12314 + }, + { + "epoch": 0.27107638124587274, + "grad_norm": 1.7195876598288766, + "learning_rate": 8.5518165934412e-06, + "loss": 0.2533, + "step": 12315 + }, + { + "epoch": 0.27109839313229145, + "grad_norm": 1.639194363810119, + "learning_rate": 8.55156569787685e-06, + "loss": 0.2915, + "step": 12316 + }, + { + "epoch": 0.2711204050187101, + "grad_norm": 2.087752868940656, + "learning_rate": 8.551314784261857e-06, + "loss": 0.4109, + "step": 12317 + }, + { + "epoch": 0.27114241690512875, + "grad_norm": 2.4538225953044197, + "learning_rate": 8.551063852597499e-06, + "loss": 0.329, + "step": 12318 + }, + { + "epoch": 0.27116442879154745, + "grad_norm": 1.2453361816402195, + "learning_rate": 8.55081290288505e-06, + "loss": 0.2793, + "step": 12319 + }, + { + "epoch": 0.2711864406779661, + "grad_norm": 1.8518661103281726, + "learning_rate": 8.550561935125787e-06, + "loss": 0.3882, + "step": 12320 + }, + { + "epoch": 0.27120845256438475, + "grad_norm": 1.180633452924583, + "learning_rate": 8.550310949320986e-06, + "loss": 0.4169, + "step": 12321 + }, + { + "epoch": 0.27123046445080345, + "grad_norm": 1.377784891167337, + "learning_rate": 8.55005994547192e-06, + "loss": 0.3385, + "step": 12322 + }, + { + "epoch": 0.2712524763372221, + "grad_norm": 0.6269252890019388, + "learning_rate": 8.549808923579866e-06, + "loss": 0.5194, + "step": 12323 + }, + { + "epoch": 0.27127448822364075, + "grad_norm": 1.6230561434569193, + "learning_rate": 8.5495578836461e-06, + "loss": 0.3095, + "step": 12324 + }, + { + "epoch": 0.27129650011005946, + "grad_norm": 0.6339974094542661, + "learning_rate": 8.549306825671897e-06, + "loss": 0.5374, + "step": 12325 + }, + { + "epoch": 0.2713185119964781, + "grad_norm": 1.3930774268762787, + "learning_rate": 8.549055749658535e-06, + "loss": 0.3913, + "step": 12326 + }, + { + "epoch": 0.27134052388289676, + "grad_norm": 1.3649616686566615, + "learning_rate": 8.548804655607287e-06, + "loss": 0.361, + "step": 12327 + }, + { + "epoch": 0.2713625357693154, + "grad_norm": 1.2388100896634477, + "learning_rate": 8.54855354351943e-06, + "loss": 0.3403, + "step": 12328 + }, + { + "epoch": 0.2713845476557341, + "grad_norm": 1.8885285465862156, + "learning_rate": 8.548302413396244e-06, + "loss": 0.3968, + "step": 12329 + }, + { + "epoch": 0.27140655954215276, + "grad_norm": 1.650890313833358, + "learning_rate": 8.548051265239002e-06, + "loss": 0.3187, + "step": 12330 + }, + { + "epoch": 0.2714285714285714, + "grad_norm": 2.7387218418233563, + "learning_rate": 8.54780009904898e-06, + "loss": 0.3309, + "step": 12331 + }, + { + "epoch": 0.2714505833149901, + "grad_norm": 1.511055976929964, + "learning_rate": 8.547548914827457e-06, + "loss": 0.4289, + "step": 12332 + }, + { + "epoch": 0.27147259520140876, + "grad_norm": 2.478413550782303, + "learning_rate": 8.547297712575707e-06, + "loss": 0.3386, + "step": 12333 + }, + { + "epoch": 0.2714946070878274, + "grad_norm": 2.58486633807626, + "learning_rate": 8.547046492295007e-06, + "loss": 0.3985, + "step": 12334 + }, + { + "epoch": 0.2715166189742461, + "grad_norm": 0.6377487805634257, + "learning_rate": 8.546795253986638e-06, + "loss": 0.5264, + "step": 12335 + }, + { + "epoch": 0.27153863086066476, + "grad_norm": 1.187328798491796, + "learning_rate": 8.546543997651871e-06, + "loss": 0.329, + "step": 12336 + }, + { + "epoch": 0.2715606427470834, + "grad_norm": 2.311473775820997, + "learning_rate": 8.546292723291984e-06, + "loss": 0.3152, + "step": 12337 + }, + { + "epoch": 0.2715826546335021, + "grad_norm": 1.62560108892525, + "learning_rate": 8.546041430908259e-06, + "loss": 0.3952, + "step": 12338 + }, + { + "epoch": 0.27160466651992077, + "grad_norm": 1.3277499284825505, + "learning_rate": 8.54579012050197e-06, + "loss": 0.275, + "step": 12339 + }, + { + "epoch": 0.2716266784063394, + "grad_norm": 1.9220802874178018, + "learning_rate": 8.545538792074392e-06, + "loss": 0.3365, + "step": 12340 + }, + { + "epoch": 0.27164869029275807, + "grad_norm": 0.6217734039114243, + "learning_rate": 8.545287445626804e-06, + "loss": 0.5167, + "step": 12341 + }, + { + "epoch": 0.27167070217917677, + "grad_norm": 1.453581699862335, + "learning_rate": 8.545036081160485e-06, + "loss": 0.3145, + "step": 12342 + }, + { + "epoch": 0.2716927140655954, + "grad_norm": 1.637079019062717, + "learning_rate": 8.544784698676712e-06, + "loss": 0.3027, + "step": 12343 + }, + { + "epoch": 0.27171472595201407, + "grad_norm": 4.019696325314197, + "learning_rate": 8.544533298176763e-06, + "loss": 0.3103, + "step": 12344 + }, + { + "epoch": 0.2717367378384328, + "grad_norm": 0.662776314501672, + "learning_rate": 8.544281879661914e-06, + "loss": 0.5159, + "step": 12345 + }, + { + "epoch": 0.2717587497248514, + "grad_norm": 1.7652982343426964, + "learning_rate": 8.544030443133442e-06, + "loss": 0.3457, + "step": 12346 + }, + { + "epoch": 0.27178076161127007, + "grad_norm": 1.6974854209919374, + "learning_rate": 8.543778988592628e-06, + "loss": 0.4189, + "step": 12347 + }, + { + "epoch": 0.2718027734976888, + "grad_norm": 1.503072084635637, + "learning_rate": 8.54352751604075e-06, + "loss": 0.3651, + "step": 12348 + }, + { + "epoch": 0.2718247853841074, + "grad_norm": 2.2978131352036613, + "learning_rate": 8.54327602547908e-06, + "loss": 0.3048, + "step": 12349 + }, + { + "epoch": 0.2718467972705261, + "grad_norm": 1.7001511396121456, + "learning_rate": 8.543024516908907e-06, + "loss": 0.3236, + "step": 12350 + }, + { + "epoch": 0.2718688091569447, + "grad_norm": 1.9469027551018392, + "learning_rate": 8.5427729903315e-06, + "loss": 0.3292, + "step": 12351 + }, + { + "epoch": 0.27189082104336343, + "grad_norm": 1.3693636432765557, + "learning_rate": 8.542521445748142e-06, + "loss": 0.2964, + "step": 12352 + }, + { + "epoch": 0.2719128329297821, + "grad_norm": 1.439951144717356, + "learning_rate": 8.542269883160108e-06, + "loss": 0.2696, + "step": 12353 + }, + { + "epoch": 0.2719348448162007, + "grad_norm": 1.424068789079127, + "learning_rate": 8.54201830256868e-06, + "loss": 0.3712, + "step": 12354 + }, + { + "epoch": 0.27195685670261943, + "grad_norm": 1.6257524176261524, + "learning_rate": 8.541766703975134e-06, + "loss": 0.3369, + "step": 12355 + }, + { + "epoch": 0.2719788685890381, + "grad_norm": 1.2650842378504379, + "learning_rate": 8.54151508738075e-06, + "loss": 0.2502, + "step": 12356 + }, + { + "epoch": 0.27200088047545673, + "grad_norm": 1.5973331948359775, + "learning_rate": 8.541263452786808e-06, + "loss": 0.3535, + "step": 12357 + }, + { + "epoch": 0.27202289236187543, + "grad_norm": 0.652634873532717, + "learning_rate": 8.541011800194585e-06, + "loss": 0.5142, + "step": 12358 + }, + { + "epoch": 0.2720449042482941, + "grad_norm": 1.460506733929732, + "learning_rate": 8.540760129605362e-06, + "loss": 0.2848, + "step": 12359 + }, + { + "epoch": 0.27206691613471273, + "grad_norm": 2.2831406747637746, + "learning_rate": 8.540508441020414e-06, + "loss": 0.3552, + "step": 12360 + }, + { + "epoch": 0.27208892802113144, + "grad_norm": 1.737689147190722, + "learning_rate": 8.540256734441025e-06, + "loss": 0.3364, + "step": 12361 + }, + { + "epoch": 0.2721109399075501, + "grad_norm": 1.4532673702808303, + "learning_rate": 8.540005009868472e-06, + "loss": 0.321, + "step": 12362 + }, + { + "epoch": 0.27213295179396874, + "grad_norm": 2.0784916114670264, + "learning_rate": 8.539753267304034e-06, + "loss": 0.3533, + "step": 12363 + }, + { + "epoch": 0.2721549636803874, + "grad_norm": 1.2417245380592514, + "learning_rate": 8.539501506748992e-06, + "loss": 0.3882, + "step": 12364 + }, + { + "epoch": 0.2721769755668061, + "grad_norm": 1.3710804916198676, + "learning_rate": 8.539249728204623e-06, + "loss": 0.3712, + "step": 12365 + }, + { + "epoch": 0.27219898745322474, + "grad_norm": 2.177010958888188, + "learning_rate": 8.53899793167221e-06, + "loss": 0.3127, + "step": 12366 + }, + { + "epoch": 0.2722209993396434, + "grad_norm": 1.5183885896635758, + "learning_rate": 8.538746117153031e-06, + "loss": 0.3259, + "step": 12367 + }, + { + "epoch": 0.2722430112260621, + "grad_norm": 0.6175859148013334, + "learning_rate": 8.538494284648366e-06, + "loss": 0.5083, + "step": 12368 + }, + { + "epoch": 0.27226502311248074, + "grad_norm": 0.6291640418081894, + "learning_rate": 8.538242434159494e-06, + "loss": 0.5412, + "step": 12369 + }, + { + "epoch": 0.2722870349988994, + "grad_norm": 1.4268456848861208, + "learning_rate": 8.537990565687697e-06, + "loss": 0.3425, + "step": 12370 + }, + { + "epoch": 0.2723090468853181, + "grad_norm": 2.1648900803994464, + "learning_rate": 8.537738679234254e-06, + "loss": 0.3422, + "step": 12371 + }, + { + "epoch": 0.27233105877173674, + "grad_norm": 1.1638721954232771, + "learning_rate": 8.537486774800443e-06, + "loss": 0.2973, + "step": 12372 + }, + { + "epoch": 0.2723530706581554, + "grad_norm": 1.5146410816514562, + "learning_rate": 8.537234852387549e-06, + "loss": 0.3328, + "step": 12373 + }, + { + "epoch": 0.27237508254457404, + "grad_norm": 1.3440258326958587, + "learning_rate": 8.536982911996848e-06, + "loss": 0.2843, + "step": 12374 + }, + { + "epoch": 0.27239709443099275, + "grad_norm": 1.6260737412061133, + "learning_rate": 8.536730953629622e-06, + "loss": 0.3587, + "step": 12375 + }, + { + "epoch": 0.2724191063174114, + "grad_norm": 1.5765673979962933, + "learning_rate": 8.536478977287153e-06, + "loss": 0.3127, + "step": 12376 + }, + { + "epoch": 0.27244111820383005, + "grad_norm": 0.6321354911603793, + "learning_rate": 8.536226982970721e-06, + "loss": 0.5362, + "step": 12377 + }, + { + "epoch": 0.27246313009024875, + "grad_norm": 2.243702482801433, + "learning_rate": 8.535974970681605e-06, + "loss": 0.3437, + "step": 12378 + }, + { + "epoch": 0.2724851419766674, + "grad_norm": 0.6237963551311573, + "learning_rate": 8.535722940421088e-06, + "loss": 0.5065, + "step": 12379 + }, + { + "epoch": 0.27250715386308605, + "grad_norm": 1.5930737638656443, + "learning_rate": 8.53547089219045e-06, + "loss": 0.3787, + "step": 12380 + }, + { + "epoch": 0.27252916574950475, + "grad_norm": 1.4660222366719684, + "learning_rate": 8.535218825990972e-06, + "loss": 0.3506, + "step": 12381 + }, + { + "epoch": 0.2725511776359234, + "grad_norm": 1.8076383690091569, + "learning_rate": 8.534966741823934e-06, + "loss": 0.3054, + "step": 12382 + }, + { + "epoch": 0.27257318952234205, + "grad_norm": 1.1762764577795337, + "learning_rate": 8.53471463969062e-06, + "loss": 0.3508, + "step": 12383 + }, + { + "epoch": 0.27259520140876076, + "grad_norm": 1.2236851657521588, + "learning_rate": 8.534462519592311e-06, + "loss": 0.2745, + "step": 12384 + }, + { + "epoch": 0.2726172132951794, + "grad_norm": 1.4442810883170158, + "learning_rate": 8.534210381530283e-06, + "loss": 0.4016, + "step": 12385 + }, + { + "epoch": 0.27263922518159805, + "grad_norm": 1.6186702259394727, + "learning_rate": 8.533958225505824e-06, + "loss": 0.3195, + "step": 12386 + }, + { + "epoch": 0.2726612370680167, + "grad_norm": 1.626734563424143, + "learning_rate": 8.533706051520212e-06, + "loss": 0.2799, + "step": 12387 + }, + { + "epoch": 0.2726832489544354, + "grad_norm": 1.4856441164080807, + "learning_rate": 8.533453859574732e-06, + "loss": 0.4458, + "step": 12388 + }, + { + "epoch": 0.27270526084085406, + "grad_norm": 2.369007806453543, + "learning_rate": 8.533201649670661e-06, + "loss": 0.3691, + "step": 12389 + }, + { + "epoch": 0.2727272727272727, + "grad_norm": 2.522630636174459, + "learning_rate": 8.532949421809284e-06, + "loss": 0.3422, + "step": 12390 + }, + { + "epoch": 0.2727492846136914, + "grad_norm": 1.5815734364962397, + "learning_rate": 8.532697175991885e-06, + "loss": 0.4487, + "step": 12391 + }, + { + "epoch": 0.27277129650011006, + "grad_norm": 2.3215742282189766, + "learning_rate": 8.532444912219739e-06, + "loss": 0.402, + "step": 12392 + }, + { + "epoch": 0.2727933083865287, + "grad_norm": 1.4613065835888028, + "learning_rate": 8.532192630494135e-06, + "loss": 0.3429, + "step": 12393 + }, + { + "epoch": 0.2728153202729474, + "grad_norm": 1.457355001008704, + "learning_rate": 8.531940330816353e-06, + "loss": 0.3497, + "step": 12394 + }, + { + "epoch": 0.27283733215936606, + "grad_norm": 1.4144008576385874, + "learning_rate": 8.531688013187673e-06, + "loss": 0.3478, + "step": 12395 + }, + { + "epoch": 0.2728593440457847, + "grad_norm": 1.3113873088381087, + "learning_rate": 8.531435677609382e-06, + "loss": 0.1742, + "step": 12396 + }, + { + "epoch": 0.27288135593220336, + "grad_norm": 1.3986075680509125, + "learning_rate": 8.531183324082757e-06, + "loss": 0.3609, + "step": 12397 + }, + { + "epoch": 0.27290336781862207, + "grad_norm": 1.9624882983113872, + "learning_rate": 8.530930952609086e-06, + "loss": 0.2635, + "step": 12398 + }, + { + "epoch": 0.2729253797050407, + "grad_norm": 0.6759603173055813, + "learning_rate": 8.530678563189648e-06, + "loss": 0.5007, + "step": 12399 + }, + { + "epoch": 0.27294739159145937, + "grad_norm": 0.6421901193966019, + "learning_rate": 8.530426155825726e-06, + "loss": 0.5085, + "step": 12400 + }, + { + "epoch": 0.27296940347787807, + "grad_norm": 1.996082544053495, + "learning_rate": 8.530173730518603e-06, + "loss": 0.365, + "step": 12401 + }, + { + "epoch": 0.2729914153642967, + "grad_norm": 1.2905146334522588, + "learning_rate": 8.529921287269564e-06, + "loss": 0.2611, + "step": 12402 + }, + { + "epoch": 0.27301342725071537, + "grad_norm": 0.6499577942129549, + "learning_rate": 8.529668826079891e-06, + "loss": 0.5351, + "step": 12403 + }, + { + "epoch": 0.2730354391371341, + "grad_norm": 1.2997217889047252, + "learning_rate": 8.529416346950865e-06, + "loss": 0.308, + "step": 12404 + }, + { + "epoch": 0.2730574510235527, + "grad_norm": 1.1573783569355935, + "learning_rate": 8.529163849883772e-06, + "loss": 0.3741, + "step": 12405 + }, + { + "epoch": 0.27307946290997137, + "grad_norm": 1.1848820734298737, + "learning_rate": 8.528911334879896e-06, + "loss": 0.2766, + "step": 12406 + }, + { + "epoch": 0.2731014747963901, + "grad_norm": 1.356681903663859, + "learning_rate": 8.528658801940517e-06, + "loss": 0.3687, + "step": 12407 + }, + { + "epoch": 0.2731234866828087, + "grad_norm": 1.2741289439761632, + "learning_rate": 8.52840625106692e-06, + "loss": 0.2925, + "step": 12408 + }, + { + "epoch": 0.2731454985692274, + "grad_norm": 1.5393255497403202, + "learning_rate": 8.528153682260388e-06, + "loss": 0.4524, + "step": 12409 + }, + { + "epoch": 0.273167510455646, + "grad_norm": 0.6953777842678571, + "learning_rate": 8.527901095522206e-06, + "loss": 0.4988, + "step": 12410 + }, + { + "epoch": 0.27318952234206473, + "grad_norm": 1.3770077754394972, + "learning_rate": 8.527648490853659e-06, + "loss": 0.3133, + "step": 12411 + }, + { + "epoch": 0.2732115342284834, + "grad_norm": 1.650446926596922, + "learning_rate": 8.527395868256027e-06, + "loss": 0.3711, + "step": 12412 + }, + { + "epoch": 0.273233546114902, + "grad_norm": 1.4471901029425565, + "learning_rate": 8.527143227730596e-06, + "loss": 0.3999, + "step": 12413 + }, + { + "epoch": 0.27325555800132073, + "grad_norm": 2.223748170430963, + "learning_rate": 8.52689056927865e-06, + "loss": 0.3106, + "step": 12414 + }, + { + "epoch": 0.2732775698877394, + "grad_norm": 2.602972545893025, + "learning_rate": 8.526637892901476e-06, + "loss": 0.2743, + "step": 12415 + }, + { + "epoch": 0.27329958177415803, + "grad_norm": 0.6824547740709188, + "learning_rate": 8.526385198600352e-06, + "loss": 0.5361, + "step": 12416 + }, + { + "epoch": 0.27332159366057673, + "grad_norm": 3.176771505559685, + "learning_rate": 8.526132486376566e-06, + "loss": 0.3665, + "step": 12417 + }, + { + "epoch": 0.2733436055469954, + "grad_norm": 2.294068852139408, + "learning_rate": 8.525879756231402e-06, + "loss": 0.2599, + "step": 12418 + }, + { + "epoch": 0.27336561743341403, + "grad_norm": 0.635745493473651, + "learning_rate": 8.525627008166144e-06, + "loss": 0.4919, + "step": 12419 + }, + { + "epoch": 0.27338762931983274, + "grad_norm": 9.215164425977534, + "learning_rate": 8.525374242182078e-06, + "loss": 0.351, + "step": 12420 + }, + { + "epoch": 0.2734096412062514, + "grad_norm": 1.459597302287269, + "learning_rate": 8.525121458280488e-06, + "loss": 0.3626, + "step": 12421 + }, + { + "epoch": 0.27343165309267004, + "grad_norm": 1.2646077212380396, + "learning_rate": 8.524868656462658e-06, + "loss": 0.3376, + "step": 12422 + }, + { + "epoch": 0.2734536649790887, + "grad_norm": 0.6247550096196334, + "learning_rate": 8.524615836729873e-06, + "loss": 0.5082, + "step": 12423 + }, + { + "epoch": 0.2734756768655074, + "grad_norm": 1.9169937360392555, + "learning_rate": 8.52436299908342e-06, + "loss": 0.3812, + "step": 12424 + }, + { + "epoch": 0.27349768875192604, + "grad_norm": 1.33056425770169, + "learning_rate": 8.52411014352458e-06, + "loss": 0.2989, + "step": 12425 + }, + { + "epoch": 0.2735197006383447, + "grad_norm": 1.9472425961344368, + "learning_rate": 8.523857270054643e-06, + "loss": 0.3225, + "step": 12426 + }, + { + "epoch": 0.2735417125247634, + "grad_norm": 2.14762096801492, + "learning_rate": 8.52360437867489e-06, + "loss": 0.3397, + "step": 12427 + }, + { + "epoch": 0.27356372441118204, + "grad_norm": 2.155572947585643, + "learning_rate": 8.523351469386607e-06, + "loss": 0.3111, + "step": 12428 + }, + { + "epoch": 0.2735857362976007, + "grad_norm": 0.683928447518629, + "learning_rate": 8.523098542191083e-06, + "loss": 0.4996, + "step": 12429 + }, + { + "epoch": 0.2736077481840194, + "grad_norm": 1.371728816549023, + "learning_rate": 8.5228455970896e-06, + "loss": 0.3068, + "step": 12430 + }, + { + "epoch": 0.27362976007043804, + "grad_norm": 1.8186550104503063, + "learning_rate": 8.522592634083442e-06, + "loss": 0.3027, + "step": 12431 + }, + { + "epoch": 0.2736517719568567, + "grad_norm": 0.6207548187194268, + "learning_rate": 8.5223396531739e-06, + "loss": 0.5063, + "step": 12432 + }, + { + "epoch": 0.27367378384327534, + "grad_norm": 1.661956567545133, + "learning_rate": 8.522086654362257e-06, + "loss": 0.3437, + "step": 12433 + }, + { + "epoch": 0.27369579572969405, + "grad_norm": 1.5349740814472816, + "learning_rate": 8.521833637649796e-06, + "loss": 0.3885, + "step": 12434 + }, + { + "epoch": 0.2737178076161127, + "grad_norm": 1.2408419791309289, + "learning_rate": 8.521580603037807e-06, + "loss": 0.3462, + "step": 12435 + }, + { + "epoch": 0.27373981950253135, + "grad_norm": 1.6460901994135668, + "learning_rate": 8.521327550527576e-06, + "loss": 0.2998, + "step": 12436 + }, + { + "epoch": 0.27376183138895005, + "grad_norm": 2.7527845387526653, + "learning_rate": 8.521074480120386e-06, + "loss": 0.3711, + "step": 12437 + }, + { + "epoch": 0.2737838432753687, + "grad_norm": 1.2097842050636827, + "learning_rate": 8.520821391817527e-06, + "loss": 0.3147, + "step": 12438 + }, + { + "epoch": 0.27380585516178735, + "grad_norm": 2.8103287855839847, + "learning_rate": 8.520568285620282e-06, + "loss": 0.3488, + "step": 12439 + }, + { + "epoch": 0.27382786704820605, + "grad_norm": 1.7892769301976423, + "learning_rate": 8.520315161529938e-06, + "loss": 0.3044, + "step": 12440 + }, + { + "epoch": 0.2738498789346247, + "grad_norm": 1.6988808668911155, + "learning_rate": 8.520062019547783e-06, + "loss": 0.3299, + "step": 12441 + }, + { + "epoch": 0.27387189082104335, + "grad_norm": 1.5680525357416601, + "learning_rate": 8.519808859675103e-06, + "loss": 0.4006, + "step": 12442 + }, + { + "epoch": 0.27389390270746206, + "grad_norm": 1.4620406560166264, + "learning_rate": 8.519555681913186e-06, + "loss": 0.3639, + "step": 12443 + }, + { + "epoch": 0.2739159145938807, + "grad_norm": 1.6987082920088288, + "learning_rate": 8.519302486263316e-06, + "loss": 0.3551, + "step": 12444 + }, + { + "epoch": 0.27393792648029935, + "grad_norm": 1.6587082825987354, + "learning_rate": 8.51904927272678e-06, + "loss": 0.3681, + "step": 12445 + }, + { + "epoch": 0.273959938366718, + "grad_norm": 1.3779962816881033, + "learning_rate": 8.518796041304868e-06, + "loss": 0.3127, + "step": 12446 + }, + { + "epoch": 0.2739819502531367, + "grad_norm": 1.2158391695218156, + "learning_rate": 8.518542791998863e-06, + "loss": 0.2975, + "step": 12447 + }, + { + "epoch": 0.27400396213955536, + "grad_norm": 1.3956199253255632, + "learning_rate": 8.518289524810057e-06, + "loss": 0.3499, + "step": 12448 + }, + { + "epoch": 0.274025974025974, + "grad_norm": 1.404274800643788, + "learning_rate": 8.51803623973973e-06, + "loss": 0.3673, + "step": 12449 + }, + { + "epoch": 0.2740479859123927, + "grad_norm": 1.7972548424435089, + "learning_rate": 8.517782936789179e-06, + "loss": 0.4221, + "step": 12450 + }, + { + "epoch": 0.27406999779881136, + "grad_norm": 1.5667758507135068, + "learning_rate": 8.517529615959683e-06, + "loss": 0.2931, + "step": 12451 + }, + { + "epoch": 0.27409200968523, + "grad_norm": 2.0082871836931457, + "learning_rate": 8.517276277252534e-06, + "loss": 0.4487, + "step": 12452 + }, + { + "epoch": 0.2741140215716487, + "grad_norm": 0.7239001210942642, + "learning_rate": 8.517022920669016e-06, + "loss": 0.5571, + "step": 12453 + }, + { + "epoch": 0.27413603345806736, + "grad_norm": 1.3280907119156218, + "learning_rate": 8.51676954621042e-06, + "loss": 0.3468, + "step": 12454 + }, + { + "epoch": 0.274158045344486, + "grad_norm": 0.6451965009524298, + "learning_rate": 8.516516153878034e-06, + "loss": 0.5421, + "step": 12455 + }, + { + "epoch": 0.27418005723090466, + "grad_norm": 1.6613290093805384, + "learning_rate": 8.516262743673143e-06, + "loss": 0.3413, + "step": 12456 + }, + { + "epoch": 0.27420206911732337, + "grad_norm": 1.212920481894171, + "learning_rate": 8.516009315597038e-06, + "loss": 0.241, + "step": 12457 + }, + { + "epoch": 0.274224081003742, + "grad_norm": 1.7733609827206454, + "learning_rate": 8.515755869651003e-06, + "loss": 0.314, + "step": 12458 + }, + { + "epoch": 0.27424609289016066, + "grad_norm": 1.6115339227892762, + "learning_rate": 8.515502405836332e-06, + "loss": 0.3151, + "step": 12459 + }, + { + "epoch": 0.27426810477657937, + "grad_norm": 1.6910996169055685, + "learning_rate": 8.515248924154307e-06, + "loss": 0.3081, + "step": 12460 + }, + { + "epoch": 0.274290116662998, + "grad_norm": 1.1583545064085081, + "learning_rate": 8.514995424606219e-06, + "loss": 0.3144, + "step": 12461 + }, + { + "epoch": 0.27431212854941667, + "grad_norm": 1.66102186426039, + "learning_rate": 8.514741907193358e-06, + "loss": 0.3169, + "step": 12462 + }, + { + "epoch": 0.27433414043583537, + "grad_norm": 1.6247681388522313, + "learning_rate": 8.51448837191701e-06, + "loss": 0.3513, + "step": 12463 + }, + { + "epoch": 0.274356152322254, + "grad_norm": 1.4137089509611311, + "learning_rate": 8.514234818778464e-06, + "loss": 0.2768, + "step": 12464 + }, + { + "epoch": 0.27437816420867267, + "grad_norm": 1.7429741866677724, + "learning_rate": 8.513981247779012e-06, + "loss": 0.4031, + "step": 12465 + }, + { + "epoch": 0.2744001760950914, + "grad_norm": 1.8982775886800451, + "learning_rate": 8.513727658919938e-06, + "loss": 0.3346, + "step": 12466 + }, + { + "epoch": 0.27442218798151, + "grad_norm": 1.1152414728669717, + "learning_rate": 8.513474052202533e-06, + "loss": 0.2632, + "step": 12467 + }, + { + "epoch": 0.2744441998679287, + "grad_norm": 1.1987181616362663, + "learning_rate": 8.513220427628087e-06, + "loss": 0.2812, + "step": 12468 + }, + { + "epoch": 0.2744662117543473, + "grad_norm": 0.8258484790244507, + "learning_rate": 8.512966785197885e-06, + "loss": 0.5357, + "step": 12469 + }, + { + "epoch": 0.274488223640766, + "grad_norm": 1.427913801819772, + "learning_rate": 8.512713124913222e-06, + "loss": 0.276, + "step": 12470 + }, + { + "epoch": 0.2745102355271847, + "grad_norm": 1.5135132791457662, + "learning_rate": 8.512459446775383e-06, + "loss": 0.2959, + "step": 12471 + }, + { + "epoch": 0.2745322474136033, + "grad_norm": 1.7033247538108343, + "learning_rate": 8.512205750785659e-06, + "loss": 0.3455, + "step": 12472 + }, + { + "epoch": 0.27455425930002203, + "grad_norm": 1.5700423287849976, + "learning_rate": 8.511952036945337e-06, + "loss": 0.3295, + "step": 12473 + }, + { + "epoch": 0.2745762711864407, + "grad_norm": 2.0340547484219136, + "learning_rate": 8.51169830525571e-06, + "loss": 0.3121, + "step": 12474 + }, + { + "epoch": 0.27459828307285933, + "grad_norm": 0.6384396289185638, + "learning_rate": 8.511444555718067e-06, + "loss": 0.5159, + "step": 12475 + }, + { + "epoch": 0.27462029495927803, + "grad_norm": 1.4462343371632305, + "learning_rate": 8.511190788333695e-06, + "loss": 0.3861, + "step": 12476 + }, + { + "epoch": 0.2746423068456967, + "grad_norm": 1.2375445690152478, + "learning_rate": 8.510937003103886e-06, + "loss": 0.3285, + "step": 12477 + }, + { + "epoch": 0.27466431873211533, + "grad_norm": 0.6723196646705901, + "learning_rate": 8.510683200029927e-06, + "loss": 0.5313, + "step": 12478 + }, + { + "epoch": 0.27468633061853404, + "grad_norm": 1.969691516249108, + "learning_rate": 8.510429379113114e-06, + "loss": 0.2824, + "step": 12479 + }, + { + "epoch": 0.2747083425049527, + "grad_norm": 0.6315865958668391, + "learning_rate": 8.51017554035473e-06, + "loss": 0.5168, + "step": 12480 + }, + { + "epoch": 0.27473035439137133, + "grad_norm": 0.6203803797741763, + "learning_rate": 8.50992168375607e-06, + "loss": 0.5167, + "step": 12481 + }, + { + "epoch": 0.27475236627779, + "grad_norm": 0.6015044763064561, + "learning_rate": 8.509667809318423e-06, + "loss": 0.5194, + "step": 12482 + }, + { + "epoch": 0.2747743781642087, + "grad_norm": 1.721715446218046, + "learning_rate": 8.509413917043077e-06, + "loss": 0.3835, + "step": 12483 + }, + { + "epoch": 0.27479639005062734, + "grad_norm": 1.3854703164612066, + "learning_rate": 8.509160006931327e-06, + "loss": 0.3174, + "step": 12484 + }, + { + "epoch": 0.274818401937046, + "grad_norm": 2.003884680373299, + "learning_rate": 8.508906078984457e-06, + "loss": 0.2486, + "step": 12485 + }, + { + "epoch": 0.2748404138234647, + "grad_norm": 0.6768317353377169, + "learning_rate": 8.508652133203763e-06, + "loss": 0.5155, + "step": 12486 + }, + { + "epoch": 0.27486242570988334, + "grad_norm": 1.7397953443749994, + "learning_rate": 8.508398169590535e-06, + "loss": 0.3458, + "step": 12487 + }, + { + "epoch": 0.274884437596302, + "grad_norm": 1.6845161178944628, + "learning_rate": 8.508144188146063e-06, + "loss": 0.3829, + "step": 12488 + }, + { + "epoch": 0.2749064494827207, + "grad_norm": 0.640645388505286, + "learning_rate": 8.507890188871637e-06, + "loss": 0.5153, + "step": 12489 + }, + { + "epoch": 0.27492846136913934, + "grad_norm": 1.5888548402812015, + "learning_rate": 8.507636171768546e-06, + "loss": 0.3641, + "step": 12490 + }, + { + "epoch": 0.274950473255558, + "grad_norm": 0.6180044721332197, + "learning_rate": 8.507382136838086e-06, + "loss": 0.5035, + "step": 12491 + }, + { + "epoch": 0.27497248514197664, + "grad_norm": 0.6388256988405778, + "learning_rate": 8.507128084081546e-06, + "loss": 0.4905, + "step": 12492 + }, + { + "epoch": 0.27499449702839535, + "grad_norm": 1.6050147714437772, + "learning_rate": 8.506874013500215e-06, + "loss": 0.4478, + "step": 12493 + }, + { + "epoch": 0.275016508914814, + "grad_norm": 4.869919905088719, + "learning_rate": 8.506619925095388e-06, + "loss": 0.4717, + "step": 12494 + }, + { + "epoch": 0.27503852080123264, + "grad_norm": 1.5529050015968433, + "learning_rate": 8.506365818868354e-06, + "loss": 0.3506, + "step": 12495 + }, + { + "epoch": 0.27506053268765135, + "grad_norm": 1.6575061865110696, + "learning_rate": 8.506111694820403e-06, + "loss": 0.3904, + "step": 12496 + }, + { + "epoch": 0.27508254457407, + "grad_norm": 1.4044517250371216, + "learning_rate": 8.505857552952832e-06, + "loss": 0.3534, + "step": 12497 + }, + { + "epoch": 0.27510455646048865, + "grad_norm": 7.167524142975824, + "learning_rate": 8.505603393266927e-06, + "loss": 0.3673, + "step": 12498 + }, + { + "epoch": 0.27512656834690735, + "grad_norm": 1.4245176779228548, + "learning_rate": 8.505349215763983e-06, + "loss": 0.2763, + "step": 12499 + }, + { + "epoch": 0.275148580233326, + "grad_norm": 1.2356783494724253, + "learning_rate": 8.50509502044529e-06, + "loss": 0.3181, + "step": 12500 + }, + { + "epoch": 0.27517059211974465, + "grad_norm": 1.3460432258212771, + "learning_rate": 8.50484080731214e-06, + "loss": 0.2886, + "step": 12501 + }, + { + "epoch": 0.27519260400616335, + "grad_norm": 2.133370033188857, + "learning_rate": 8.504586576365829e-06, + "loss": 0.2903, + "step": 12502 + }, + { + "epoch": 0.275214615892582, + "grad_norm": 1.6982299099188447, + "learning_rate": 8.504332327607643e-06, + "loss": 0.3775, + "step": 12503 + }, + { + "epoch": 0.27523662777900065, + "grad_norm": 1.4978433779622249, + "learning_rate": 8.504078061038878e-06, + "loss": 0.3217, + "step": 12504 + }, + { + "epoch": 0.2752586396654193, + "grad_norm": 0.7159363826964052, + "learning_rate": 8.503823776660826e-06, + "loss": 0.5127, + "step": 12505 + }, + { + "epoch": 0.275280651551838, + "grad_norm": 2.3354899212227087, + "learning_rate": 8.503569474474778e-06, + "loss": 0.3358, + "step": 12506 + }, + { + "epoch": 0.27530266343825666, + "grad_norm": 4.032353903713827, + "learning_rate": 8.503315154482028e-06, + "loss": 0.4017, + "step": 12507 + }, + { + "epoch": 0.2753246753246753, + "grad_norm": 1.3653270705117535, + "learning_rate": 8.503060816683867e-06, + "loss": 0.3291, + "step": 12508 + }, + { + "epoch": 0.275346687211094, + "grad_norm": 0.6574123449161717, + "learning_rate": 8.50280646108159e-06, + "loss": 0.5186, + "step": 12509 + }, + { + "epoch": 0.27536869909751266, + "grad_norm": 1.4722272654618231, + "learning_rate": 8.502552087676486e-06, + "loss": 0.4388, + "step": 12510 + }, + { + "epoch": 0.2753907109839313, + "grad_norm": 1.3631675632316458, + "learning_rate": 8.502297696469853e-06, + "loss": 0.2697, + "step": 12511 + }, + { + "epoch": 0.27541272287035, + "grad_norm": 0.6316507470269769, + "learning_rate": 8.502043287462979e-06, + "loss": 0.5118, + "step": 12512 + }, + { + "epoch": 0.27543473475676866, + "grad_norm": 1.6559458713359638, + "learning_rate": 8.501788860657158e-06, + "loss": 0.3996, + "step": 12513 + }, + { + "epoch": 0.2754567466431873, + "grad_norm": 1.7608429475758414, + "learning_rate": 8.501534416053688e-06, + "loss": 0.365, + "step": 12514 + }, + { + "epoch": 0.27547875852960596, + "grad_norm": 1.3142580390237035, + "learning_rate": 8.501279953653857e-06, + "loss": 0.29, + "step": 12515 + }, + { + "epoch": 0.27550077041602467, + "grad_norm": 1.627403874957312, + "learning_rate": 8.501025473458956e-06, + "loss": 0.3123, + "step": 12516 + }, + { + "epoch": 0.2755227823024433, + "grad_norm": 1.586320996263122, + "learning_rate": 8.500770975470286e-06, + "loss": 0.3373, + "step": 12517 + }, + { + "epoch": 0.27554479418886196, + "grad_norm": 2.3783502369160647, + "learning_rate": 8.500516459689134e-06, + "loss": 0.372, + "step": 12518 + }, + { + "epoch": 0.27556680607528067, + "grad_norm": 1.2927618266589056, + "learning_rate": 8.500261926116797e-06, + "loss": 0.3438, + "step": 12519 + }, + { + "epoch": 0.2755888179616993, + "grad_norm": 1.3166073871155415, + "learning_rate": 8.500007374754568e-06, + "loss": 0.2994, + "step": 12520 + }, + { + "epoch": 0.27561082984811797, + "grad_norm": 1.3548176530382987, + "learning_rate": 8.49975280560374e-06, + "loss": 0.2986, + "step": 12521 + }, + { + "epoch": 0.27563284173453667, + "grad_norm": 1.5391038673092792, + "learning_rate": 8.49949821866561e-06, + "loss": 0.2777, + "step": 12522 + }, + { + "epoch": 0.2756548536209553, + "grad_norm": 1.1236272429781424, + "learning_rate": 8.499243613941467e-06, + "loss": 0.3315, + "step": 12523 + }, + { + "epoch": 0.27567686550737397, + "grad_norm": 2.5744444357672434, + "learning_rate": 8.498988991432607e-06, + "loss": 0.3298, + "step": 12524 + }, + { + "epoch": 0.2756988773937927, + "grad_norm": 1.5100885719939796, + "learning_rate": 8.498734351140326e-06, + "loss": 0.4133, + "step": 12525 + }, + { + "epoch": 0.2757208892802113, + "grad_norm": 0.811444161220863, + "learning_rate": 8.498479693065914e-06, + "loss": 0.525, + "step": 12526 + }, + { + "epoch": 0.27574290116663, + "grad_norm": 1.8067438722230704, + "learning_rate": 8.498225017210669e-06, + "loss": 0.3767, + "step": 12527 + }, + { + "epoch": 0.2757649130530486, + "grad_norm": 1.3371928059216855, + "learning_rate": 8.497970323575883e-06, + "loss": 0.2954, + "step": 12528 + }, + { + "epoch": 0.2757869249394673, + "grad_norm": 0.65547222342734, + "learning_rate": 8.497715612162855e-06, + "loss": 0.5483, + "step": 12529 + }, + { + "epoch": 0.275808936825886, + "grad_norm": 1.860463602447538, + "learning_rate": 8.497460882972874e-06, + "loss": 0.4043, + "step": 12530 + }, + { + "epoch": 0.2758309487123046, + "grad_norm": 1.7318341475766459, + "learning_rate": 8.497206136007237e-06, + "loss": 0.3566, + "step": 12531 + }, + { + "epoch": 0.27585296059872333, + "grad_norm": 1.5376413499454338, + "learning_rate": 8.496951371267237e-06, + "loss": 0.3835, + "step": 12532 + }, + { + "epoch": 0.275874972485142, + "grad_norm": 0.666577058221183, + "learning_rate": 8.496696588754174e-06, + "loss": 0.5046, + "step": 12533 + }, + { + "epoch": 0.2758969843715606, + "grad_norm": 1.3038240290311112, + "learning_rate": 8.496441788469338e-06, + "loss": 0.4747, + "step": 12534 + }, + { + "epoch": 0.27591899625797933, + "grad_norm": 1.3668144180896846, + "learning_rate": 8.496186970414025e-06, + "loss": 0.3113, + "step": 12535 + }, + { + "epoch": 0.275941008144398, + "grad_norm": 1.3447194400473808, + "learning_rate": 8.495932134589529e-06, + "loss": 0.3104, + "step": 12536 + }, + { + "epoch": 0.27596302003081663, + "grad_norm": 0.6781195919919507, + "learning_rate": 8.495677280997147e-06, + "loss": 0.4987, + "step": 12537 + }, + { + "epoch": 0.2759850319172353, + "grad_norm": 1.2967518168125818, + "learning_rate": 8.495422409638176e-06, + "loss": 0.3809, + "step": 12538 + }, + { + "epoch": 0.276007043803654, + "grad_norm": 1.3793930934606413, + "learning_rate": 8.495167520513909e-06, + "loss": 0.4661, + "step": 12539 + }, + { + "epoch": 0.27602905569007263, + "grad_norm": 0.6453559562433628, + "learning_rate": 8.494912613625641e-06, + "loss": 0.4995, + "step": 12540 + }, + { + "epoch": 0.2760510675764913, + "grad_norm": 1.3345208165191076, + "learning_rate": 8.494657688974667e-06, + "loss": 0.3385, + "step": 12541 + }, + { + "epoch": 0.27607307946291, + "grad_norm": 1.5913315892821311, + "learning_rate": 8.494402746562287e-06, + "loss": 0.2437, + "step": 12542 + }, + { + "epoch": 0.27609509134932864, + "grad_norm": 1.7626684134758233, + "learning_rate": 8.494147786389792e-06, + "loss": 0.393, + "step": 12543 + }, + { + "epoch": 0.2761171032357473, + "grad_norm": 1.7489584457485288, + "learning_rate": 8.493892808458479e-06, + "loss": 0.417, + "step": 12544 + }, + { + "epoch": 0.276139115122166, + "grad_norm": 1.6290461667115177, + "learning_rate": 8.493637812769645e-06, + "loss": 0.3449, + "step": 12545 + }, + { + "epoch": 0.27616112700858464, + "grad_norm": 1.4573252482774965, + "learning_rate": 8.493382799324585e-06, + "loss": 0.3416, + "step": 12546 + }, + { + "epoch": 0.2761831388950033, + "grad_norm": 1.4980407852735176, + "learning_rate": 8.493127768124597e-06, + "loss": 0.3475, + "step": 12547 + }, + { + "epoch": 0.276205150781422, + "grad_norm": 6.654434402274824, + "learning_rate": 8.492872719170976e-06, + "loss": 0.291, + "step": 12548 + }, + { + "epoch": 0.27622716266784064, + "grad_norm": 1.346380096685918, + "learning_rate": 8.492617652465014e-06, + "loss": 0.3468, + "step": 12549 + }, + { + "epoch": 0.2762491745542593, + "grad_norm": 0.7744015792845558, + "learning_rate": 8.492362568008015e-06, + "loss": 0.5071, + "step": 12550 + }, + { + "epoch": 0.27627118644067794, + "grad_norm": 1.815269059031827, + "learning_rate": 8.49210746580127e-06, + "loss": 0.3532, + "step": 12551 + }, + { + "epoch": 0.27629319832709665, + "grad_norm": 8.32886218964436, + "learning_rate": 8.49185234584608e-06, + "loss": 0.3665, + "step": 12552 + }, + { + "epoch": 0.2763152102135153, + "grad_norm": 0.6520339449247939, + "learning_rate": 8.491597208143736e-06, + "loss": 0.4936, + "step": 12553 + }, + { + "epoch": 0.27633722209993394, + "grad_norm": 0.6358989652510915, + "learning_rate": 8.491342052695539e-06, + "loss": 0.5295, + "step": 12554 + }, + { + "epoch": 0.27635923398635265, + "grad_norm": 1.3746279300229427, + "learning_rate": 8.491086879502784e-06, + "loss": 0.2417, + "step": 12555 + }, + { + "epoch": 0.2763812458727713, + "grad_norm": 1.262742636817663, + "learning_rate": 8.490831688566769e-06, + "loss": 0.3498, + "step": 12556 + }, + { + "epoch": 0.27640325775918995, + "grad_norm": 1.2871517511305341, + "learning_rate": 8.490576479888789e-06, + "loss": 0.2866, + "step": 12557 + }, + { + "epoch": 0.27642526964560865, + "grad_norm": 1.3056712979625988, + "learning_rate": 8.490321253470144e-06, + "loss": 0.339, + "step": 12558 + }, + { + "epoch": 0.2764472815320273, + "grad_norm": 0.7629532383379387, + "learning_rate": 8.490066009312128e-06, + "loss": 0.5115, + "step": 12559 + }, + { + "epoch": 0.27646929341844595, + "grad_norm": 1.3081004905102778, + "learning_rate": 8.489810747416042e-06, + "loss": 0.2993, + "step": 12560 + }, + { + "epoch": 0.27649130530486465, + "grad_norm": 0.6866561545208106, + "learning_rate": 8.489555467783182e-06, + "loss": 0.5061, + "step": 12561 + }, + { + "epoch": 0.2765133171912833, + "grad_norm": 1.4387280325209109, + "learning_rate": 8.489300170414842e-06, + "loss": 0.3141, + "step": 12562 + }, + { + "epoch": 0.27653532907770195, + "grad_norm": 1.3980242267575917, + "learning_rate": 8.489044855312323e-06, + "loss": 0.4016, + "step": 12563 + }, + { + "epoch": 0.2765573409641206, + "grad_norm": 0.6223533360479078, + "learning_rate": 8.488789522476923e-06, + "loss": 0.5268, + "step": 12564 + }, + { + "epoch": 0.2765793528505393, + "grad_norm": 1.5306384364960623, + "learning_rate": 8.488534171909938e-06, + "loss": 0.2717, + "step": 12565 + }, + { + "epoch": 0.27660136473695796, + "grad_norm": 1.577595522304825, + "learning_rate": 8.488278803612666e-06, + "loss": 0.3657, + "step": 12566 + }, + { + "epoch": 0.2766233766233766, + "grad_norm": 2.220405406087533, + "learning_rate": 8.488023417586406e-06, + "loss": 0.3379, + "step": 12567 + }, + { + "epoch": 0.2766453885097953, + "grad_norm": 1.0377914020806611, + "learning_rate": 8.487768013832454e-06, + "loss": 0.2535, + "step": 12568 + }, + { + "epoch": 0.27666740039621396, + "grad_norm": 0.7727083891753239, + "learning_rate": 8.48751259235211e-06, + "loss": 0.5255, + "step": 12569 + }, + { + "epoch": 0.2766894122826326, + "grad_norm": 3.139534169594745, + "learning_rate": 8.487257153146672e-06, + "loss": 0.2876, + "step": 12570 + }, + { + "epoch": 0.2767114241690513, + "grad_norm": 1.2513341978723311, + "learning_rate": 8.487001696217438e-06, + "loss": 0.2923, + "step": 12571 + }, + { + "epoch": 0.27673343605546996, + "grad_norm": 2.0100663650396426, + "learning_rate": 8.486746221565706e-06, + "loss": 0.3153, + "step": 12572 + }, + { + "epoch": 0.2767554479418886, + "grad_norm": 2.9988810926536433, + "learning_rate": 8.486490729192773e-06, + "loss": 0.393, + "step": 12573 + }, + { + "epoch": 0.27677745982830726, + "grad_norm": 1.263002797643862, + "learning_rate": 8.486235219099942e-06, + "loss": 0.3141, + "step": 12574 + }, + { + "epoch": 0.27679947171472596, + "grad_norm": 1.4849456845291953, + "learning_rate": 8.485979691288507e-06, + "loss": 0.3728, + "step": 12575 + }, + { + "epoch": 0.2768214836011446, + "grad_norm": 0.6349203591150611, + "learning_rate": 8.485724145759769e-06, + "loss": 0.5168, + "step": 12576 + }, + { + "epoch": 0.27684349548756326, + "grad_norm": 2.2363991306430173, + "learning_rate": 8.485468582515023e-06, + "loss": 0.3813, + "step": 12577 + }, + { + "epoch": 0.27686550737398197, + "grad_norm": 1.5502535771163912, + "learning_rate": 8.485213001555576e-06, + "loss": 0.3082, + "step": 12578 + }, + { + "epoch": 0.2768875192604006, + "grad_norm": 1.956338016891969, + "learning_rate": 8.48495740288272e-06, + "loss": 0.4194, + "step": 12579 + }, + { + "epoch": 0.27690953114681927, + "grad_norm": 0.6385316905586521, + "learning_rate": 8.484701786497755e-06, + "loss": 0.5361, + "step": 12580 + }, + { + "epoch": 0.27693154303323797, + "grad_norm": 1.5318004060191661, + "learning_rate": 8.484446152401983e-06, + "loss": 0.3695, + "step": 12581 + }, + { + "epoch": 0.2769535549196566, + "grad_norm": 1.6188160531703137, + "learning_rate": 8.4841905005967e-06, + "loss": 0.3916, + "step": 12582 + }, + { + "epoch": 0.27697556680607527, + "grad_norm": 1.0822726976043437, + "learning_rate": 8.483934831083208e-06, + "loss": 0.2984, + "step": 12583 + }, + { + "epoch": 0.276997578692494, + "grad_norm": 1.5865929954350773, + "learning_rate": 8.483679143862806e-06, + "loss": 0.3662, + "step": 12584 + }, + { + "epoch": 0.2770195905789126, + "grad_norm": 1.5213920646240378, + "learning_rate": 8.483423438936792e-06, + "loss": 0.3061, + "step": 12585 + }, + { + "epoch": 0.27704160246533127, + "grad_norm": 3.3458235032581847, + "learning_rate": 8.483167716306466e-06, + "loss": 0.3581, + "step": 12586 + }, + { + "epoch": 0.2770636143517499, + "grad_norm": 1.1072643234876054, + "learning_rate": 8.482911975973129e-06, + "loss": 0.2597, + "step": 12587 + }, + { + "epoch": 0.2770856262381686, + "grad_norm": 1.46476182418134, + "learning_rate": 8.48265621793808e-06, + "loss": 0.2947, + "step": 12588 + }, + { + "epoch": 0.2771076381245873, + "grad_norm": 1.1858448741934835, + "learning_rate": 8.482400442202617e-06, + "loss": 0.3271, + "step": 12589 + }, + { + "epoch": 0.2771296500110059, + "grad_norm": 2.3887703117086385, + "learning_rate": 8.482144648768042e-06, + "loss": 0.4088, + "step": 12590 + }, + { + "epoch": 0.27715166189742463, + "grad_norm": 0.6348873708240202, + "learning_rate": 8.481888837635657e-06, + "loss": 0.523, + "step": 12591 + }, + { + "epoch": 0.2771736737838433, + "grad_norm": 1.236823420006802, + "learning_rate": 8.481633008806758e-06, + "loss": 0.4332, + "step": 12592 + }, + { + "epoch": 0.2771956856702619, + "grad_norm": 2.1961834233457336, + "learning_rate": 8.48137716228265e-06, + "loss": 0.3448, + "step": 12593 + }, + { + "epoch": 0.27721769755668063, + "grad_norm": 1.238977951980596, + "learning_rate": 8.481121298064628e-06, + "loss": 0.2653, + "step": 12594 + }, + { + "epoch": 0.2772397094430993, + "grad_norm": 1.458445120445271, + "learning_rate": 8.480865416153993e-06, + "loss": 0.301, + "step": 12595 + }, + { + "epoch": 0.27726172132951793, + "grad_norm": 2.482260593316978, + "learning_rate": 8.480609516552051e-06, + "loss": 0.3502, + "step": 12596 + }, + { + "epoch": 0.2772837332159366, + "grad_norm": 0.6241090386917983, + "learning_rate": 8.480353599260098e-06, + "loss": 0.5103, + "step": 12597 + }, + { + "epoch": 0.2773057451023553, + "grad_norm": 1.323064566437669, + "learning_rate": 8.480097664279437e-06, + "loss": 0.3114, + "step": 12598 + }, + { + "epoch": 0.27732775698877393, + "grad_norm": 1.2340131698791792, + "learning_rate": 8.479841711611365e-06, + "loss": 0.3235, + "step": 12599 + }, + { + "epoch": 0.2773497688751926, + "grad_norm": 1.52418254278376, + "learning_rate": 8.479585741257185e-06, + "loss": 0.3409, + "step": 12600 + }, + { + "epoch": 0.2773717807616113, + "grad_norm": 0.6335991969263316, + "learning_rate": 8.479329753218202e-06, + "loss": 0.5143, + "step": 12601 + }, + { + "epoch": 0.27739379264802994, + "grad_norm": 1.7685784294698317, + "learning_rate": 8.479073747495711e-06, + "loss": 0.3591, + "step": 12602 + }, + { + "epoch": 0.2774158045344486, + "grad_norm": 1.2924415233804678, + "learning_rate": 8.478817724091015e-06, + "loss": 0.372, + "step": 12603 + }, + { + "epoch": 0.2774378164208673, + "grad_norm": 1.5932055644418013, + "learning_rate": 8.478561683005417e-06, + "loss": 0.3588, + "step": 12604 + }, + { + "epoch": 0.27745982830728594, + "grad_norm": 0.6129497247023248, + "learning_rate": 8.478305624240216e-06, + "loss": 0.5025, + "step": 12605 + }, + { + "epoch": 0.2774818401937046, + "grad_norm": 1.708453982010648, + "learning_rate": 8.478049547796715e-06, + "loss": 0.2674, + "step": 12606 + }, + { + "epoch": 0.2775038520801233, + "grad_norm": 1.320243640522734, + "learning_rate": 8.477793453676215e-06, + "loss": 0.3558, + "step": 12607 + }, + { + "epoch": 0.27752586396654194, + "grad_norm": 0.6194244201016731, + "learning_rate": 8.477537341880018e-06, + "loss": 0.5135, + "step": 12608 + }, + { + "epoch": 0.2775478758529606, + "grad_norm": 1.7678312782488785, + "learning_rate": 8.477281212409425e-06, + "loss": 0.3525, + "step": 12609 + }, + { + "epoch": 0.27756988773937924, + "grad_norm": 1.3569210755628283, + "learning_rate": 8.477025065265735e-06, + "loss": 0.2969, + "step": 12610 + }, + { + "epoch": 0.27759189962579794, + "grad_norm": 1.5015134654318258, + "learning_rate": 8.476768900450256e-06, + "loss": 0.3021, + "step": 12611 + }, + { + "epoch": 0.2776139115122166, + "grad_norm": 1.8029116352218815, + "learning_rate": 8.476512717964286e-06, + "loss": 0.4762, + "step": 12612 + }, + { + "epoch": 0.27763592339863524, + "grad_norm": 1.8871849905185485, + "learning_rate": 8.47625651780913e-06, + "loss": 0.3136, + "step": 12613 + }, + { + "epoch": 0.27765793528505395, + "grad_norm": 1.3716797219798609, + "learning_rate": 8.476000299986085e-06, + "loss": 0.28, + "step": 12614 + }, + { + "epoch": 0.2776799471714726, + "grad_norm": 0.6276598663771344, + "learning_rate": 8.475744064496458e-06, + "loss": 0.5128, + "step": 12615 + }, + { + "epoch": 0.27770195905789125, + "grad_norm": 1.39120874913165, + "learning_rate": 8.475487811341547e-06, + "loss": 0.3123, + "step": 12616 + }, + { + "epoch": 0.27772397094430995, + "grad_norm": 1.3489121118557703, + "learning_rate": 8.475231540522659e-06, + "loss": 0.3634, + "step": 12617 + }, + { + "epoch": 0.2777459828307286, + "grad_norm": 1.1172211124827445, + "learning_rate": 8.474975252041094e-06, + "loss": 0.297, + "step": 12618 + }, + { + "epoch": 0.27776799471714725, + "grad_norm": 0.6159011591378315, + "learning_rate": 8.474718945898153e-06, + "loss": 0.5315, + "step": 12619 + }, + { + "epoch": 0.27779000660356595, + "grad_norm": 1.4261182588141568, + "learning_rate": 8.474462622095143e-06, + "loss": 0.3317, + "step": 12620 + }, + { + "epoch": 0.2778120184899846, + "grad_norm": 1.414563747742023, + "learning_rate": 8.474206280633363e-06, + "loss": 0.3653, + "step": 12621 + }, + { + "epoch": 0.27783403037640325, + "grad_norm": 1.5473015775953556, + "learning_rate": 8.473949921514117e-06, + "loss": 0.3467, + "step": 12622 + }, + { + "epoch": 0.2778560422628219, + "grad_norm": 0.6328123149172861, + "learning_rate": 8.473693544738709e-06, + "loss": 0.532, + "step": 12623 + }, + { + "epoch": 0.2778780541492406, + "grad_norm": 1.2229785771808703, + "learning_rate": 8.473437150308441e-06, + "loss": 0.3298, + "step": 12624 + }, + { + "epoch": 0.27790006603565925, + "grad_norm": 3.3632405271938124, + "learning_rate": 8.473180738224615e-06, + "loss": 0.2882, + "step": 12625 + }, + { + "epoch": 0.2779220779220779, + "grad_norm": 1.4331942332903829, + "learning_rate": 8.472924308488538e-06, + "loss": 0.4206, + "step": 12626 + }, + { + "epoch": 0.2779440898084966, + "grad_norm": 0.6113840906062, + "learning_rate": 8.472667861101506e-06, + "loss": 0.5015, + "step": 12627 + }, + { + "epoch": 0.27796610169491526, + "grad_norm": 1.2452757795879135, + "learning_rate": 8.472411396064831e-06, + "loss": 0.3101, + "step": 12628 + }, + { + "epoch": 0.2779881135813339, + "grad_norm": 1.3130883520205712, + "learning_rate": 8.472154913379812e-06, + "loss": 0.352, + "step": 12629 + }, + { + "epoch": 0.2780101254677526, + "grad_norm": 1.528763531827459, + "learning_rate": 8.471898413047753e-06, + "loss": 0.2864, + "step": 12630 + }, + { + "epoch": 0.27803213735417126, + "grad_norm": 1.5135840893149308, + "learning_rate": 8.471641895069957e-06, + "loss": 0.3337, + "step": 12631 + }, + { + "epoch": 0.2780541492405899, + "grad_norm": 1.4860678603843593, + "learning_rate": 8.471385359447729e-06, + "loss": 0.3165, + "step": 12632 + }, + { + "epoch": 0.27807616112700856, + "grad_norm": 1.2983659788025996, + "learning_rate": 8.471128806182372e-06, + "loss": 0.374, + "step": 12633 + }, + { + "epoch": 0.27809817301342726, + "grad_norm": 2.6722449989705215, + "learning_rate": 8.470872235275192e-06, + "loss": 0.3584, + "step": 12634 + }, + { + "epoch": 0.2781201848998459, + "grad_norm": 0.6372559833827326, + "learning_rate": 8.47061564672749e-06, + "loss": 0.5089, + "step": 12635 + }, + { + "epoch": 0.27814219678626456, + "grad_norm": 1.3183137723166969, + "learning_rate": 8.47035904054057e-06, + "loss": 0.3143, + "step": 12636 + }, + { + "epoch": 0.27816420867268327, + "grad_norm": 1.5754094506305867, + "learning_rate": 8.470102416715737e-06, + "loss": 0.3133, + "step": 12637 + }, + { + "epoch": 0.2781862205591019, + "grad_norm": 1.9871914935366277, + "learning_rate": 8.469845775254299e-06, + "loss": 0.3488, + "step": 12638 + }, + { + "epoch": 0.27820823244552056, + "grad_norm": 1.368888991689368, + "learning_rate": 8.469589116157554e-06, + "loss": 0.3014, + "step": 12639 + }, + { + "epoch": 0.27823024433193927, + "grad_norm": 1.2970519623560166, + "learning_rate": 8.469332439426813e-06, + "loss": 0.3639, + "step": 12640 + }, + { + "epoch": 0.2782522562183579, + "grad_norm": 1.666581320710369, + "learning_rate": 8.469075745063375e-06, + "loss": 0.4464, + "step": 12641 + }, + { + "epoch": 0.27827426810477657, + "grad_norm": 1.8666364271331568, + "learning_rate": 8.468819033068547e-06, + "loss": 0.4733, + "step": 12642 + }, + { + "epoch": 0.2782962799911953, + "grad_norm": 1.4894846837773594, + "learning_rate": 8.468562303443634e-06, + "loss": 0.3184, + "step": 12643 + }, + { + "epoch": 0.2783182918776139, + "grad_norm": 2.1945869746537556, + "learning_rate": 8.46830555618994e-06, + "loss": 0.3162, + "step": 12644 + }, + { + "epoch": 0.27834030376403257, + "grad_norm": 4.940163682929887, + "learning_rate": 8.468048791308771e-06, + "loss": 0.3659, + "step": 12645 + }, + { + "epoch": 0.2783623156504512, + "grad_norm": 1.7790428443146271, + "learning_rate": 8.46779200880143e-06, + "loss": 0.3858, + "step": 12646 + }, + { + "epoch": 0.2783843275368699, + "grad_norm": 2.430915640549967, + "learning_rate": 8.467535208669223e-06, + "loss": 0.2977, + "step": 12647 + }, + { + "epoch": 0.2784063394232886, + "grad_norm": 1.4569740388851837, + "learning_rate": 8.467278390913458e-06, + "loss": 0.4132, + "step": 12648 + }, + { + "epoch": 0.2784283513097072, + "grad_norm": 1.5924418918042123, + "learning_rate": 8.467021555535435e-06, + "loss": 0.3736, + "step": 12649 + }, + { + "epoch": 0.2784503631961259, + "grad_norm": 1.6670498588932223, + "learning_rate": 8.466764702536463e-06, + "loss": 0.3078, + "step": 12650 + }, + { + "epoch": 0.2784723750825446, + "grad_norm": 1.6298519110728225, + "learning_rate": 8.466507831917848e-06, + "loss": 0.3365, + "step": 12651 + }, + { + "epoch": 0.2784943869689632, + "grad_norm": 1.517218654377713, + "learning_rate": 8.466250943680894e-06, + "loss": 0.3502, + "step": 12652 + }, + { + "epoch": 0.27851639885538193, + "grad_norm": 1.1502133570306403, + "learning_rate": 8.465994037826906e-06, + "loss": 0.3178, + "step": 12653 + }, + { + "epoch": 0.2785384107418006, + "grad_norm": 2.05635406252078, + "learning_rate": 8.465737114357189e-06, + "loss": 0.398, + "step": 12654 + }, + { + "epoch": 0.27856042262821923, + "grad_norm": 0.6641888189134457, + "learning_rate": 8.46548017327305e-06, + "loss": 0.526, + "step": 12655 + }, + { + "epoch": 0.2785824345146379, + "grad_norm": 1.6389122025819465, + "learning_rate": 8.465223214575796e-06, + "loss": 0.3696, + "step": 12656 + }, + { + "epoch": 0.2786044464010566, + "grad_norm": 1.9654203294228407, + "learning_rate": 8.464966238266734e-06, + "loss": 0.3277, + "step": 12657 + }, + { + "epoch": 0.27862645828747523, + "grad_norm": 1.6435614585166047, + "learning_rate": 8.464709244347167e-06, + "loss": 0.2868, + "step": 12658 + }, + { + "epoch": 0.2786484701738939, + "grad_norm": 1.3909680457421458, + "learning_rate": 8.464452232818402e-06, + "loss": 0.3472, + "step": 12659 + }, + { + "epoch": 0.2786704820603126, + "grad_norm": 1.4024010760330425, + "learning_rate": 8.464195203681746e-06, + "loss": 0.3414, + "step": 12660 + }, + { + "epoch": 0.27869249394673123, + "grad_norm": 1.639856748663, + "learning_rate": 8.463938156938504e-06, + "loss": 0.3166, + "step": 12661 + }, + { + "epoch": 0.2787145058331499, + "grad_norm": 1.2052662900851339, + "learning_rate": 8.463681092589984e-06, + "loss": 0.3139, + "step": 12662 + }, + { + "epoch": 0.2787365177195686, + "grad_norm": 0.6539464887944364, + "learning_rate": 8.46342401063749e-06, + "loss": 0.5161, + "step": 12663 + }, + { + "epoch": 0.27875852960598724, + "grad_norm": 0.6327518787861444, + "learning_rate": 8.463166911082334e-06, + "loss": 0.5047, + "step": 12664 + }, + { + "epoch": 0.2787805414924059, + "grad_norm": 2.610848036849555, + "learning_rate": 8.462909793925817e-06, + "loss": 0.4068, + "step": 12665 + }, + { + "epoch": 0.2788025533788246, + "grad_norm": 2.216637554099845, + "learning_rate": 8.462652659169249e-06, + "loss": 0.2863, + "step": 12666 + }, + { + "epoch": 0.27882456526524324, + "grad_norm": 0.6360401685589477, + "learning_rate": 8.462395506813934e-06, + "loss": 0.5245, + "step": 12667 + }, + { + "epoch": 0.2788465771516619, + "grad_norm": 2.704369901746538, + "learning_rate": 8.46213833686118e-06, + "loss": 0.3827, + "step": 12668 + }, + { + "epoch": 0.27886858903808054, + "grad_norm": 1.2683188649957555, + "learning_rate": 8.461881149312299e-06, + "loss": 0.27, + "step": 12669 + }, + { + "epoch": 0.27889060092449924, + "grad_norm": 2.5291707219248036, + "learning_rate": 8.461623944168592e-06, + "loss": 0.3998, + "step": 12670 + }, + { + "epoch": 0.2789126128109179, + "grad_norm": 1.626463237685182, + "learning_rate": 8.461366721431367e-06, + "loss": 0.4554, + "step": 12671 + }, + { + "epoch": 0.27893462469733654, + "grad_norm": 1.5158791267511418, + "learning_rate": 8.461109481101933e-06, + "loss": 0.2973, + "step": 12672 + }, + { + "epoch": 0.27895663658375525, + "grad_norm": 1.7640875241554304, + "learning_rate": 8.460852223181597e-06, + "loss": 0.3232, + "step": 12673 + }, + { + "epoch": 0.2789786484701739, + "grad_norm": 1.5045974603649752, + "learning_rate": 8.460594947671666e-06, + "loss": 0.3492, + "step": 12674 + }, + { + "epoch": 0.27900066035659254, + "grad_norm": 1.6023717029732718, + "learning_rate": 8.460337654573449e-06, + "loss": 0.3685, + "step": 12675 + }, + { + "epoch": 0.27902267224301125, + "grad_norm": 1.667829634094666, + "learning_rate": 8.460080343888252e-06, + "loss": 0.3843, + "step": 12676 + }, + { + "epoch": 0.2790446841294299, + "grad_norm": 0.6600425377911987, + "learning_rate": 8.459823015617383e-06, + "loss": 0.5275, + "step": 12677 + }, + { + "epoch": 0.27906669601584855, + "grad_norm": 1.6217055986402549, + "learning_rate": 8.459565669762152e-06, + "loss": 0.3536, + "step": 12678 + }, + { + "epoch": 0.2790887079022672, + "grad_norm": 1.2815944272017012, + "learning_rate": 8.459308306323864e-06, + "loss": 0.3519, + "step": 12679 + }, + { + "epoch": 0.2791107197886859, + "grad_norm": 1.498702795850672, + "learning_rate": 8.45905092530383e-06, + "loss": 0.2967, + "step": 12680 + }, + { + "epoch": 0.27913273167510455, + "grad_norm": 1.7984404281437805, + "learning_rate": 8.458793526703354e-06, + "loss": 0.3529, + "step": 12681 + }, + { + "epoch": 0.2791547435615232, + "grad_norm": 1.5027719422540167, + "learning_rate": 8.458536110523748e-06, + "loss": 0.3848, + "step": 12682 + }, + { + "epoch": 0.2791767554479419, + "grad_norm": 1.5462166901027732, + "learning_rate": 8.458278676766318e-06, + "loss": 0.348, + "step": 12683 + }, + { + "epoch": 0.27919876733436055, + "grad_norm": 1.5950493118597175, + "learning_rate": 8.458021225432374e-06, + "loss": 0.2753, + "step": 12684 + }, + { + "epoch": 0.2792207792207792, + "grad_norm": 1.3433358234495834, + "learning_rate": 8.457763756523222e-06, + "loss": 0.2807, + "step": 12685 + }, + { + "epoch": 0.2792427911071979, + "grad_norm": 1.5984066618835981, + "learning_rate": 8.457506270040175e-06, + "loss": 0.3846, + "step": 12686 + }, + { + "epoch": 0.27926480299361656, + "grad_norm": 1.499952418474923, + "learning_rate": 8.457248765984538e-06, + "loss": 0.3067, + "step": 12687 + }, + { + "epoch": 0.2792868148800352, + "grad_norm": 1.3189690456856176, + "learning_rate": 8.456991244357622e-06, + "loss": 0.3067, + "step": 12688 + }, + { + "epoch": 0.2793088267664539, + "grad_norm": 1.4142827844884232, + "learning_rate": 8.456733705160734e-06, + "loss": 0.3866, + "step": 12689 + }, + { + "epoch": 0.27933083865287256, + "grad_norm": 2.0591078740901816, + "learning_rate": 8.456476148395184e-06, + "loss": 0.356, + "step": 12690 + }, + { + "epoch": 0.2793528505392912, + "grad_norm": 1.748068154603611, + "learning_rate": 8.456218574062278e-06, + "loss": 0.3259, + "step": 12691 + }, + { + "epoch": 0.27937486242570986, + "grad_norm": 1.9039813327364004, + "learning_rate": 8.45596098216333e-06, + "loss": 0.39, + "step": 12692 + }, + { + "epoch": 0.27939687431212856, + "grad_norm": 1.5815422476827745, + "learning_rate": 8.455703372699646e-06, + "loss": 0.3907, + "step": 12693 + }, + { + "epoch": 0.2794188861985472, + "grad_norm": 2.4595001655618867, + "learning_rate": 8.455445745672536e-06, + "loss": 0.4109, + "step": 12694 + }, + { + "epoch": 0.27944089808496586, + "grad_norm": 1.6014973986160865, + "learning_rate": 8.45518810108331e-06, + "loss": 0.4503, + "step": 12695 + }, + { + "epoch": 0.27946290997138457, + "grad_norm": 2.0865805128262322, + "learning_rate": 8.454930438933278e-06, + "loss": 0.4323, + "step": 12696 + }, + { + "epoch": 0.2794849218578032, + "grad_norm": 1.3233692041203144, + "learning_rate": 8.454672759223748e-06, + "loss": 0.3082, + "step": 12697 + }, + { + "epoch": 0.27950693374422186, + "grad_norm": 1.407101818269914, + "learning_rate": 8.45441506195603e-06, + "loss": 0.4082, + "step": 12698 + }, + { + "epoch": 0.27952894563064057, + "grad_norm": 2.05055221936198, + "learning_rate": 8.454157347131432e-06, + "loss": 0.2898, + "step": 12699 + }, + { + "epoch": 0.2795509575170592, + "grad_norm": 2.889872878099958, + "learning_rate": 8.45389961475127e-06, + "loss": 0.3314, + "step": 12700 + }, + { + "epoch": 0.27957296940347787, + "grad_norm": 2.8271607311958302, + "learning_rate": 8.453641864816843e-06, + "loss": 0.2998, + "step": 12701 + }, + { + "epoch": 0.27959498128989657, + "grad_norm": 1.1916474782132116, + "learning_rate": 8.453384097329473e-06, + "loss": 0.3748, + "step": 12702 + }, + { + "epoch": 0.2796169931763152, + "grad_norm": 1.6035717990376057, + "learning_rate": 8.453126312290465e-06, + "loss": 0.4082, + "step": 12703 + }, + { + "epoch": 0.27963900506273387, + "grad_norm": 1.8308018309999303, + "learning_rate": 8.452868509701126e-06, + "loss": 0.3778, + "step": 12704 + }, + { + "epoch": 0.2796610169491525, + "grad_norm": 1.5068280016787499, + "learning_rate": 8.452610689562773e-06, + "loss": 0.3401, + "step": 12705 + }, + { + "epoch": 0.2796830288355712, + "grad_norm": 4.01993937034146, + "learning_rate": 8.45235285187671e-06, + "loss": 0.4174, + "step": 12706 + }, + { + "epoch": 0.2797050407219899, + "grad_norm": 1.121679372093789, + "learning_rate": 8.452094996644248e-06, + "loss": 0.2398, + "step": 12707 + }, + { + "epoch": 0.2797270526084085, + "grad_norm": 0.701206804048419, + "learning_rate": 8.451837123866704e-06, + "loss": 0.5318, + "step": 12708 + }, + { + "epoch": 0.2797490644948272, + "grad_norm": 1.3334535260000462, + "learning_rate": 8.451579233545381e-06, + "loss": 0.3356, + "step": 12709 + }, + { + "epoch": 0.2797710763812459, + "grad_norm": 1.3621059155096469, + "learning_rate": 8.451321325681594e-06, + "loss": 0.3286, + "step": 12710 + }, + { + "epoch": 0.2797930882676645, + "grad_norm": 0.6415996731079231, + "learning_rate": 8.451063400276654e-06, + "loss": 0.5074, + "step": 12711 + }, + { + "epoch": 0.27981510015408323, + "grad_norm": 1.2570019986486427, + "learning_rate": 8.450805457331868e-06, + "loss": 0.4028, + "step": 12712 + }, + { + "epoch": 0.2798371120405019, + "grad_norm": 1.7525772394576553, + "learning_rate": 8.450547496848551e-06, + "loss": 0.3587, + "step": 12713 + }, + { + "epoch": 0.2798591239269205, + "grad_norm": 1.493732987597008, + "learning_rate": 8.450289518828012e-06, + "loss": 0.3116, + "step": 12714 + }, + { + "epoch": 0.2798811358133392, + "grad_norm": 1.2279862886743287, + "learning_rate": 8.450031523271564e-06, + "loss": 0.3059, + "step": 12715 + }, + { + "epoch": 0.2799031476997579, + "grad_norm": 3.199588015080751, + "learning_rate": 8.449773510180517e-06, + "loss": 0.3746, + "step": 12716 + }, + { + "epoch": 0.27992515958617653, + "grad_norm": 2.182262107216779, + "learning_rate": 8.44951547955618e-06, + "loss": 0.3072, + "step": 12717 + }, + { + "epoch": 0.2799471714725952, + "grad_norm": 1.2306400823625618, + "learning_rate": 8.449257431399868e-06, + "loss": 0.3918, + "step": 12718 + }, + { + "epoch": 0.2799691833590139, + "grad_norm": 1.336964822885135, + "learning_rate": 8.448999365712893e-06, + "loss": 0.4115, + "step": 12719 + }, + { + "epoch": 0.27999119524543253, + "grad_norm": 0.8028707455824752, + "learning_rate": 8.448741282496563e-06, + "loss": 0.5121, + "step": 12720 + }, + { + "epoch": 0.2800132071318512, + "grad_norm": 1.6492547460254505, + "learning_rate": 8.448483181752194e-06, + "loss": 0.3603, + "step": 12721 + }, + { + "epoch": 0.2800352190182699, + "grad_norm": 1.348414470575644, + "learning_rate": 8.448225063481093e-06, + "loss": 0.4264, + "step": 12722 + }, + { + "epoch": 0.28005723090468854, + "grad_norm": 0.6950779805911641, + "learning_rate": 8.447966927684575e-06, + "loss": 0.5441, + "step": 12723 + }, + { + "epoch": 0.2800792427911072, + "grad_norm": 0.6624481549570048, + "learning_rate": 8.44770877436395e-06, + "loss": 0.5284, + "step": 12724 + }, + { + "epoch": 0.2801012546775259, + "grad_norm": 1.758183026995525, + "learning_rate": 8.447450603520533e-06, + "loss": 0.4078, + "step": 12725 + }, + { + "epoch": 0.28012326656394454, + "grad_norm": 1.199940921554576, + "learning_rate": 8.447192415155634e-06, + "loss": 0.2686, + "step": 12726 + }, + { + "epoch": 0.2801452784503632, + "grad_norm": 1.6028692893491534, + "learning_rate": 8.446934209270565e-06, + "loss": 0.3428, + "step": 12727 + }, + { + "epoch": 0.28016729033678184, + "grad_norm": 1.9591403165195065, + "learning_rate": 8.44667598586664e-06, + "loss": 0.353, + "step": 12728 + }, + { + "epoch": 0.28018930222320054, + "grad_norm": 0.7162465473012102, + "learning_rate": 8.44641774494517e-06, + "loss": 0.4808, + "step": 12729 + }, + { + "epoch": 0.2802113141096192, + "grad_norm": 1.5148274776587727, + "learning_rate": 8.446159486507468e-06, + "loss": 0.3889, + "step": 12730 + }, + { + "epoch": 0.28023332599603784, + "grad_norm": 1.1768584904506671, + "learning_rate": 8.445901210554846e-06, + "loss": 0.3631, + "step": 12731 + }, + { + "epoch": 0.28025533788245655, + "grad_norm": 1.2467322620254377, + "learning_rate": 8.445642917088617e-06, + "loss": 0.3574, + "step": 12732 + }, + { + "epoch": 0.2802773497688752, + "grad_norm": 1.6762108400565772, + "learning_rate": 8.445384606110094e-06, + "loss": 0.4055, + "step": 12733 + }, + { + "epoch": 0.28029936165529384, + "grad_norm": 1.3355003997747454, + "learning_rate": 8.445126277620592e-06, + "loss": 0.2825, + "step": 12734 + }, + { + "epoch": 0.28032137354171255, + "grad_norm": 0.6535895346657337, + "learning_rate": 8.444867931621419e-06, + "loss": 0.5191, + "step": 12735 + }, + { + "epoch": 0.2803433854281312, + "grad_norm": 1.5935068491525175, + "learning_rate": 8.444609568113892e-06, + "loss": 0.2788, + "step": 12736 + }, + { + "epoch": 0.28036539731454985, + "grad_norm": 1.5613382241387292, + "learning_rate": 8.444351187099321e-06, + "loss": 0.2703, + "step": 12737 + }, + { + "epoch": 0.2803874092009685, + "grad_norm": 1.8530354749691893, + "learning_rate": 8.444092788579023e-06, + "loss": 0.4126, + "step": 12738 + }, + { + "epoch": 0.2804094210873872, + "grad_norm": 2.060395359609846, + "learning_rate": 8.44383437255431e-06, + "loss": 0.3591, + "step": 12739 + }, + { + "epoch": 0.28043143297380585, + "grad_norm": 1.6713553644077488, + "learning_rate": 8.443575939026492e-06, + "loss": 0.3427, + "step": 12740 + }, + { + "epoch": 0.2804534448602245, + "grad_norm": 1.6137039470141845, + "learning_rate": 8.443317487996886e-06, + "loss": 0.4245, + "step": 12741 + }, + { + "epoch": 0.2804754567466432, + "grad_norm": 1.7689007958005991, + "learning_rate": 8.443059019466806e-06, + "loss": 0.3045, + "step": 12742 + }, + { + "epoch": 0.28049746863306185, + "grad_norm": 0.6600522717183377, + "learning_rate": 8.442800533437564e-06, + "loss": 0.5211, + "step": 12743 + }, + { + "epoch": 0.2805194805194805, + "grad_norm": 1.3621238012384622, + "learning_rate": 8.442542029910474e-06, + "loss": 0.281, + "step": 12744 + }, + { + "epoch": 0.2805414924058992, + "grad_norm": 1.7523109909265533, + "learning_rate": 8.44228350888685e-06, + "loss": 0.3087, + "step": 12745 + }, + { + "epoch": 0.28056350429231786, + "grad_norm": 1.318487156275829, + "learning_rate": 8.442024970368006e-06, + "loss": 0.2946, + "step": 12746 + }, + { + "epoch": 0.2805855161787365, + "grad_norm": 0.6243163737294155, + "learning_rate": 8.441766414355256e-06, + "loss": 0.4964, + "step": 12747 + }, + { + "epoch": 0.2806075280651552, + "grad_norm": 1.4942055290473382, + "learning_rate": 8.441507840849915e-06, + "loss": 0.3154, + "step": 12748 + }, + { + "epoch": 0.28062953995157386, + "grad_norm": 1.706346218403052, + "learning_rate": 8.441249249853293e-06, + "loss": 0.3195, + "step": 12749 + }, + { + "epoch": 0.2806515518379925, + "grad_norm": 1.2185179215118123, + "learning_rate": 8.440990641366709e-06, + "loss": 0.2216, + "step": 12750 + }, + { + "epoch": 0.28067356372441116, + "grad_norm": 1.3761685442897196, + "learning_rate": 8.440732015391478e-06, + "loss": 0.3114, + "step": 12751 + }, + { + "epoch": 0.28069557561082986, + "grad_norm": 1.3027361725321043, + "learning_rate": 8.44047337192891e-06, + "loss": 0.3287, + "step": 12752 + }, + { + "epoch": 0.2807175874972485, + "grad_norm": 1.2577469238794337, + "learning_rate": 8.44021471098032e-06, + "loss": 0.3215, + "step": 12753 + }, + { + "epoch": 0.28073959938366716, + "grad_norm": 1.584395892627963, + "learning_rate": 8.43995603254703e-06, + "loss": 0.4225, + "step": 12754 + }, + { + "epoch": 0.28076161127008586, + "grad_norm": 1.6994908624354452, + "learning_rate": 8.439697336630344e-06, + "loss": 0.2803, + "step": 12755 + }, + { + "epoch": 0.2807836231565045, + "grad_norm": 0.6482227872633048, + "learning_rate": 8.439438623231586e-06, + "loss": 0.5072, + "step": 12756 + }, + { + "epoch": 0.28080563504292316, + "grad_norm": 1.9250297548003017, + "learning_rate": 8.439179892352064e-06, + "loss": 0.3053, + "step": 12757 + }, + { + "epoch": 0.28082764692934187, + "grad_norm": 1.9994279081771988, + "learning_rate": 8.438921143993096e-06, + "loss": 0.3144, + "step": 12758 + }, + { + "epoch": 0.2808496588157605, + "grad_norm": 1.4740969140964777, + "learning_rate": 8.438662378155998e-06, + "loss": 0.238, + "step": 12759 + }, + { + "epoch": 0.28087167070217917, + "grad_norm": 0.6158806014692846, + "learning_rate": 8.438403594842084e-06, + "loss": 0.5097, + "step": 12760 + }, + { + "epoch": 0.28089368258859787, + "grad_norm": 0.6171853429477281, + "learning_rate": 8.43814479405267e-06, + "loss": 0.5093, + "step": 12761 + }, + { + "epoch": 0.2809156944750165, + "grad_norm": 1.6810137360407005, + "learning_rate": 8.437885975789069e-06, + "loss": 0.4015, + "step": 12762 + }, + { + "epoch": 0.28093770636143517, + "grad_norm": 1.4379684642411605, + "learning_rate": 8.437627140052599e-06, + "loss": 0.4368, + "step": 12763 + }, + { + "epoch": 0.2809597182478538, + "grad_norm": 1.5733223124394315, + "learning_rate": 8.437368286844574e-06, + "loss": 0.3669, + "step": 12764 + }, + { + "epoch": 0.2809817301342725, + "grad_norm": 2.216121861943495, + "learning_rate": 8.437109416166312e-06, + "loss": 0.3025, + "step": 12765 + }, + { + "epoch": 0.28100374202069117, + "grad_norm": 3.7854645855640756, + "learning_rate": 8.436850528019127e-06, + "loss": 0.3432, + "step": 12766 + }, + { + "epoch": 0.2810257539071098, + "grad_norm": 1.9086318651664012, + "learning_rate": 8.436591622404331e-06, + "loss": 0.3547, + "step": 12767 + }, + { + "epoch": 0.2810477657935285, + "grad_norm": 0.6697701862701887, + "learning_rate": 8.436332699323247e-06, + "loss": 0.5062, + "step": 12768 + }, + { + "epoch": 0.2810697776799472, + "grad_norm": 1.5402889254362078, + "learning_rate": 8.436073758777188e-06, + "loss": 0.3201, + "step": 12769 + }, + { + "epoch": 0.2810917895663658, + "grad_norm": 0.6694171238471688, + "learning_rate": 8.435814800767467e-06, + "loss": 0.538, + "step": 12770 + }, + { + "epoch": 0.28111380145278453, + "grad_norm": 1.195538422612242, + "learning_rate": 8.435555825295405e-06, + "loss": 0.4327, + "step": 12771 + }, + { + "epoch": 0.2811358133392032, + "grad_norm": 1.6611343437403496, + "learning_rate": 8.435296832362316e-06, + "loss": 0.3261, + "step": 12772 + }, + { + "epoch": 0.2811578252256218, + "grad_norm": 2.0938737082192986, + "learning_rate": 8.435037821969514e-06, + "loss": 0.4127, + "step": 12773 + }, + { + "epoch": 0.2811798371120405, + "grad_norm": 2.122097510976876, + "learning_rate": 8.434778794118319e-06, + "loss": 0.3941, + "step": 12774 + }, + { + "epoch": 0.2812018489984592, + "grad_norm": 1.4239058218335605, + "learning_rate": 8.434519748810048e-06, + "loss": 0.3001, + "step": 12775 + }, + { + "epoch": 0.28122386088487783, + "grad_norm": 1.746001963763224, + "learning_rate": 8.434260686046015e-06, + "loss": 0.3713, + "step": 12776 + }, + { + "epoch": 0.2812458727712965, + "grad_norm": 1.5457056212084164, + "learning_rate": 8.434001605827536e-06, + "loss": 0.2939, + "step": 12777 + }, + { + "epoch": 0.2812678846577152, + "grad_norm": 1.2065079278314559, + "learning_rate": 8.43374250815593e-06, + "loss": 0.3409, + "step": 12778 + }, + { + "epoch": 0.28128989654413383, + "grad_norm": 1.5059172902264706, + "learning_rate": 8.433483393032513e-06, + "loss": 0.4522, + "step": 12779 + }, + { + "epoch": 0.2813119084305525, + "grad_norm": 2.256709856609366, + "learning_rate": 8.433224260458601e-06, + "loss": 0.3207, + "step": 12780 + }, + { + "epoch": 0.2813339203169712, + "grad_norm": 1.7112079014552928, + "learning_rate": 8.432965110435513e-06, + "loss": 0.2578, + "step": 12781 + }, + { + "epoch": 0.28135593220338984, + "grad_norm": 2.585518202886196, + "learning_rate": 8.432705942964565e-06, + "loss": 0.2957, + "step": 12782 + }, + { + "epoch": 0.2813779440898085, + "grad_norm": 0.827957085040001, + "learning_rate": 8.432446758047074e-06, + "loss": 0.5487, + "step": 12783 + }, + { + "epoch": 0.2813999559762272, + "grad_norm": 2.929936724935035, + "learning_rate": 8.432187555684359e-06, + "loss": 0.3611, + "step": 12784 + }, + { + "epoch": 0.28142196786264584, + "grad_norm": 4.098810365057966, + "learning_rate": 8.431928335877737e-06, + "loss": 0.3096, + "step": 12785 + }, + { + "epoch": 0.2814439797490645, + "grad_norm": 1.5056055316969792, + "learning_rate": 8.431669098628523e-06, + "loss": 0.3324, + "step": 12786 + }, + { + "epoch": 0.28146599163548314, + "grad_norm": 1.3645546029450135, + "learning_rate": 8.431409843938034e-06, + "loss": 0.3449, + "step": 12787 + }, + { + "epoch": 0.28148800352190184, + "grad_norm": 1.6366839650657319, + "learning_rate": 8.431150571807592e-06, + "loss": 0.4174, + "step": 12788 + }, + { + "epoch": 0.2815100154083205, + "grad_norm": 1.630513988068391, + "learning_rate": 8.430891282238513e-06, + "loss": 0.3774, + "step": 12789 + }, + { + "epoch": 0.28153202729473914, + "grad_norm": 1.6465713990798574, + "learning_rate": 8.430631975232112e-06, + "loss": 0.3906, + "step": 12790 + }, + { + "epoch": 0.28155403918115784, + "grad_norm": 2.2214910846107303, + "learning_rate": 8.430372650789711e-06, + "loss": 0.4044, + "step": 12791 + }, + { + "epoch": 0.2815760510675765, + "grad_norm": 1.4291429075278095, + "learning_rate": 8.430113308912626e-06, + "loss": 0.2992, + "step": 12792 + }, + { + "epoch": 0.28159806295399514, + "grad_norm": 1.8852402331185938, + "learning_rate": 8.429853949602176e-06, + "loss": 0.4083, + "step": 12793 + }, + { + "epoch": 0.28162007484041385, + "grad_norm": 1.3674653326510549, + "learning_rate": 8.429594572859677e-06, + "loss": 0.2698, + "step": 12794 + }, + { + "epoch": 0.2816420867268325, + "grad_norm": 1.6332256797745706, + "learning_rate": 8.429335178686448e-06, + "loss": 0.3715, + "step": 12795 + }, + { + "epoch": 0.28166409861325115, + "grad_norm": 2.0811917541426643, + "learning_rate": 8.42907576708381e-06, + "loss": 0.3414, + "step": 12796 + }, + { + "epoch": 0.2816861104996698, + "grad_norm": 2.420909892432789, + "learning_rate": 8.42881633805308e-06, + "loss": 0.3119, + "step": 12797 + }, + { + "epoch": 0.2817081223860885, + "grad_norm": 0.7562079312317553, + "learning_rate": 8.428556891595575e-06, + "loss": 0.4878, + "step": 12798 + }, + { + "epoch": 0.28173013427250715, + "grad_norm": 1.3035055564961695, + "learning_rate": 8.428297427712614e-06, + "loss": 0.3817, + "step": 12799 + }, + { + "epoch": 0.2817521461589258, + "grad_norm": 1.8547370160473466, + "learning_rate": 8.428037946405517e-06, + "loss": 0.4039, + "step": 12800 + }, + { + "epoch": 0.2817741580453445, + "grad_norm": 0.6330215208795701, + "learning_rate": 8.427778447675602e-06, + "loss": 0.5143, + "step": 12801 + }, + { + "epoch": 0.28179616993176315, + "grad_norm": 1.5968368868333775, + "learning_rate": 8.427518931524188e-06, + "loss": 0.3267, + "step": 12802 + }, + { + "epoch": 0.2818181818181818, + "grad_norm": 1.3243241865689557, + "learning_rate": 8.427259397952595e-06, + "loss": 0.3274, + "step": 12803 + }, + { + "epoch": 0.2818401937046005, + "grad_norm": 2.4392012717842397, + "learning_rate": 8.42699984696214e-06, + "loss": 0.3519, + "step": 12804 + }, + { + "epoch": 0.28186220559101915, + "grad_norm": 1.5923537872532412, + "learning_rate": 8.426740278554143e-06, + "loss": 0.3556, + "step": 12805 + }, + { + "epoch": 0.2818842174774378, + "grad_norm": 1.4257622133082417, + "learning_rate": 8.426480692729926e-06, + "loss": 0.3374, + "step": 12806 + }, + { + "epoch": 0.2819062293638565, + "grad_norm": 1.5218022705736816, + "learning_rate": 8.426221089490804e-06, + "loss": 0.3995, + "step": 12807 + }, + { + "epoch": 0.28192824125027516, + "grad_norm": 1.7009058153644805, + "learning_rate": 8.4259614688381e-06, + "loss": 0.4348, + "step": 12808 + }, + { + "epoch": 0.2819502531366938, + "grad_norm": 4.598860010399086, + "learning_rate": 8.425701830773129e-06, + "loss": 0.3555, + "step": 12809 + }, + { + "epoch": 0.28197226502311246, + "grad_norm": 3.0304976902890144, + "learning_rate": 8.425442175297215e-06, + "loss": 0.3058, + "step": 12810 + }, + { + "epoch": 0.28199427690953116, + "grad_norm": 1.4744710004494554, + "learning_rate": 8.425182502411676e-06, + "loss": 0.3474, + "step": 12811 + }, + { + "epoch": 0.2820162887959498, + "grad_norm": 3.2525111601230394, + "learning_rate": 8.424922812117832e-06, + "loss": 0.4056, + "step": 12812 + }, + { + "epoch": 0.28203830068236846, + "grad_norm": 0.8368374575100934, + "learning_rate": 8.424663104417002e-06, + "loss": 0.499, + "step": 12813 + }, + { + "epoch": 0.28206031256878716, + "grad_norm": 20.278120886891074, + "learning_rate": 8.424403379310507e-06, + "loss": 0.4254, + "step": 12814 + }, + { + "epoch": 0.2820823244552058, + "grad_norm": 0.6565352673563912, + "learning_rate": 8.424143636799668e-06, + "loss": 0.5072, + "step": 12815 + }, + { + "epoch": 0.28210433634162446, + "grad_norm": 0.6286244808376954, + "learning_rate": 8.423883876885801e-06, + "loss": 0.5318, + "step": 12816 + }, + { + "epoch": 0.28212634822804317, + "grad_norm": 1.2517686875155252, + "learning_rate": 8.423624099570231e-06, + "loss": 0.3335, + "step": 12817 + }, + { + "epoch": 0.2821483601144618, + "grad_norm": 2.452824787297891, + "learning_rate": 8.423364304854276e-06, + "loss": 0.3053, + "step": 12818 + }, + { + "epoch": 0.28217037200088047, + "grad_norm": 1.634019019261833, + "learning_rate": 8.423104492739257e-06, + "loss": 0.3438, + "step": 12819 + }, + { + "epoch": 0.2821923838872991, + "grad_norm": 1.3807153166691695, + "learning_rate": 8.422844663226495e-06, + "loss": 0.3007, + "step": 12820 + }, + { + "epoch": 0.2822143957737178, + "grad_norm": 1.7095003826252217, + "learning_rate": 8.422584816317307e-06, + "loss": 0.357, + "step": 12821 + }, + { + "epoch": 0.28223640766013647, + "grad_norm": 1.573458930030216, + "learning_rate": 8.42232495201302e-06, + "loss": 0.3266, + "step": 12822 + }, + { + "epoch": 0.2822584195465551, + "grad_norm": 1.8324377856124419, + "learning_rate": 8.422065070314949e-06, + "loss": 0.3934, + "step": 12823 + }, + { + "epoch": 0.2822804314329738, + "grad_norm": 2.040137841638289, + "learning_rate": 8.421805171224417e-06, + "loss": 0.385, + "step": 12824 + }, + { + "epoch": 0.28230244331939247, + "grad_norm": 1.1303581556095403, + "learning_rate": 8.421545254742745e-06, + "loss": 0.5543, + "step": 12825 + }, + { + "epoch": 0.2823244552058111, + "grad_norm": 1.5142154630900226, + "learning_rate": 8.421285320871255e-06, + "loss": 0.3141, + "step": 12826 + }, + { + "epoch": 0.2823464670922298, + "grad_norm": 1.4292044067896064, + "learning_rate": 8.421025369611266e-06, + "loss": 0.3233, + "step": 12827 + }, + { + "epoch": 0.2823684789786485, + "grad_norm": 0.702612553632038, + "learning_rate": 8.420765400964101e-06, + "loss": 0.5383, + "step": 12828 + }, + { + "epoch": 0.2823904908650671, + "grad_norm": 1.5042024530303804, + "learning_rate": 8.42050541493108e-06, + "loss": 0.3886, + "step": 12829 + }, + { + "epoch": 0.2824125027514858, + "grad_norm": 2.736476172775828, + "learning_rate": 8.420245411513525e-06, + "loss": 0.3981, + "step": 12830 + }, + { + "epoch": 0.2824345146379045, + "grad_norm": 0.6972348890342149, + "learning_rate": 8.419985390712757e-06, + "loss": 0.5201, + "step": 12831 + }, + { + "epoch": 0.2824565265243231, + "grad_norm": 1.5087928477893757, + "learning_rate": 8.419725352530098e-06, + "loss": 0.2748, + "step": 12832 + }, + { + "epoch": 0.2824785384107418, + "grad_norm": 1.8718659468162904, + "learning_rate": 8.41946529696687e-06, + "loss": 0.3349, + "step": 12833 + }, + { + "epoch": 0.2825005502971605, + "grad_norm": 1.365271217551292, + "learning_rate": 8.419205224024393e-06, + "loss": 0.33, + "step": 12834 + }, + { + "epoch": 0.28252256218357913, + "grad_norm": 1.5852170097453777, + "learning_rate": 8.418945133703993e-06, + "loss": 0.2975, + "step": 12835 + }, + { + "epoch": 0.2825445740699978, + "grad_norm": 1.2406948480949433, + "learning_rate": 8.418685026006986e-06, + "loss": 0.2837, + "step": 12836 + }, + { + "epoch": 0.2825665859564165, + "grad_norm": 1.4767397276718028, + "learning_rate": 8.418424900934698e-06, + "loss": 0.3183, + "step": 12837 + }, + { + "epoch": 0.28258859784283513, + "grad_norm": 0.8613883543878051, + "learning_rate": 8.418164758488451e-06, + "loss": 0.5404, + "step": 12838 + }, + { + "epoch": 0.2826106097292538, + "grad_norm": 0.8199500756433596, + "learning_rate": 8.417904598669564e-06, + "loss": 0.5197, + "step": 12839 + }, + { + "epoch": 0.2826326216156725, + "grad_norm": 1.4434844231483899, + "learning_rate": 8.417644421479364e-06, + "loss": 0.2962, + "step": 12840 + }, + { + "epoch": 0.28265463350209113, + "grad_norm": 1.3238119091449208, + "learning_rate": 8.417384226919167e-06, + "loss": 0.264, + "step": 12841 + }, + { + "epoch": 0.2826766453885098, + "grad_norm": 0.6331880288071724, + "learning_rate": 8.417124014990303e-06, + "loss": 0.5292, + "step": 12842 + }, + { + "epoch": 0.2826986572749285, + "grad_norm": 1.2712079633074538, + "learning_rate": 8.416863785694089e-06, + "loss": 0.3457, + "step": 12843 + }, + { + "epoch": 0.28272066916134714, + "grad_norm": 1.9891914436481954, + "learning_rate": 8.416603539031849e-06, + "loss": 0.3204, + "step": 12844 + }, + { + "epoch": 0.2827426810477658, + "grad_norm": 0.6870660843286688, + "learning_rate": 8.416343275004907e-06, + "loss": 0.5123, + "step": 12845 + }, + { + "epoch": 0.28276469293418444, + "grad_norm": 1.3994472339662327, + "learning_rate": 8.416082993614582e-06, + "loss": 0.344, + "step": 12846 + }, + { + "epoch": 0.28278670482060314, + "grad_norm": 1.3864785488979752, + "learning_rate": 8.415822694862203e-06, + "loss": 0.3215, + "step": 12847 + }, + { + "epoch": 0.2828087167070218, + "grad_norm": 0.7078261637294186, + "learning_rate": 8.415562378749089e-06, + "loss": 0.5151, + "step": 12848 + }, + { + "epoch": 0.28283072859344044, + "grad_norm": 0.6993506419138489, + "learning_rate": 8.415302045276563e-06, + "loss": 0.5357, + "step": 12849 + }, + { + "epoch": 0.28285274047985914, + "grad_norm": 1.7057871204744206, + "learning_rate": 8.415041694445947e-06, + "loss": 0.2511, + "step": 12850 + }, + { + "epoch": 0.2828747523662778, + "grad_norm": 1.9597353799898272, + "learning_rate": 8.414781326258568e-06, + "loss": 0.3401, + "step": 12851 + }, + { + "epoch": 0.28289676425269644, + "grad_norm": 1.8993661205210133, + "learning_rate": 8.414520940715747e-06, + "loss": 0.4247, + "step": 12852 + }, + { + "epoch": 0.28291877613911515, + "grad_norm": 1.4723554656714488, + "learning_rate": 8.414260537818808e-06, + "loss": 0.3442, + "step": 12853 + }, + { + "epoch": 0.2829407880255338, + "grad_norm": 1.4883243041029055, + "learning_rate": 8.414000117569074e-06, + "loss": 0.3253, + "step": 12854 + }, + { + "epoch": 0.28296279991195245, + "grad_norm": 1.421861289451979, + "learning_rate": 8.413739679967867e-06, + "loss": 0.2961, + "step": 12855 + }, + { + "epoch": 0.2829848117983711, + "grad_norm": 0.7663821855762066, + "learning_rate": 8.413479225016514e-06, + "loss": 0.5471, + "step": 12856 + }, + { + "epoch": 0.2830068236847898, + "grad_norm": 1.2857814427899446, + "learning_rate": 8.413218752716337e-06, + "loss": 0.3739, + "step": 12857 + }, + { + "epoch": 0.28302883557120845, + "grad_norm": 1.349461746809032, + "learning_rate": 8.412958263068661e-06, + "loss": 0.2909, + "step": 12858 + }, + { + "epoch": 0.2830508474576271, + "grad_norm": 1.6361514749495458, + "learning_rate": 8.412697756074807e-06, + "loss": 0.3651, + "step": 12859 + }, + { + "epoch": 0.2830728593440458, + "grad_norm": 1.4675729259304207, + "learning_rate": 8.4124372317361e-06, + "loss": 0.3096, + "step": 12860 + }, + { + "epoch": 0.28309487123046445, + "grad_norm": 0.658504083500781, + "learning_rate": 8.412176690053868e-06, + "loss": 0.5105, + "step": 12861 + }, + { + "epoch": 0.2831168831168831, + "grad_norm": 2.0248259057461166, + "learning_rate": 8.411916131029431e-06, + "loss": 0.336, + "step": 12862 + }, + { + "epoch": 0.2831388950033018, + "grad_norm": 1.531709920128978, + "learning_rate": 8.411655554664114e-06, + "loss": 0.2922, + "step": 12863 + }, + { + "epoch": 0.28316090688972045, + "grad_norm": 1.4534697248048338, + "learning_rate": 8.411394960959242e-06, + "loss": 0.3449, + "step": 12864 + }, + { + "epoch": 0.2831829187761391, + "grad_norm": 1.4977029936450121, + "learning_rate": 8.411134349916139e-06, + "loss": 0.3269, + "step": 12865 + }, + { + "epoch": 0.2832049306625578, + "grad_norm": 4.532018874594686, + "learning_rate": 8.410873721536131e-06, + "loss": 0.354, + "step": 12866 + }, + { + "epoch": 0.28322694254897646, + "grad_norm": 1.268000783590558, + "learning_rate": 8.41061307582054e-06, + "loss": 0.3358, + "step": 12867 + }, + { + "epoch": 0.2832489544353951, + "grad_norm": 2.2595756219158534, + "learning_rate": 8.410352412770692e-06, + "loss": 0.3688, + "step": 12868 + }, + { + "epoch": 0.28327096632181376, + "grad_norm": 2.8690481898127698, + "learning_rate": 8.410091732387913e-06, + "loss": 0.3052, + "step": 12869 + }, + { + "epoch": 0.28329297820823246, + "grad_norm": 0.7103552715526261, + "learning_rate": 8.409831034673528e-06, + "loss": 0.5174, + "step": 12870 + }, + { + "epoch": 0.2833149900946511, + "grad_norm": 1.5500598444663929, + "learning_rate": 8.40957031962886e-06, + "loss": 0.3571, + "step": 12871 + }, + { + "epoch": 0.28333700198106976, + "grad_norm": 1.536644892948702, + "learning_rate": 8.409309587255232e-06, + "loss": 0.3864, + "step": 12872 + }, + { + "epoch": 0.28335901386748846, + "grad_norm": 1.6029636812966175, + "learning_rate": 8.409048837553976e-06, + "loss": 0.37, + "step": 12873 + }, + { + "epoch": 0.2833810257539071, + "grad_norm": 1.5575940577253145, + "learning_rate": 8.408788070526413e-06, + "loss": 0.3519, + "step": 12874 + }, + { + "epoch": 0.28340303764032576, + "grad_norm": 1.5899324159213775, + "learning_rate": 8.408527286173867e-06, + "loss": 0.4541, + "step": 12875 + }, + { + "epoch": 0.28342504952674447, + "grad_norm": 0.6093335484936576, + "learning_rate": 8.408266484497664e-06, + "loss": 0.5083, + "step": 12876 + }, + { + "epoch": 0.2834470614131631, + "grad_norm": 1.1789822974031963, + "learning_rate": 8.408005665499133e-06, + "loss": 0.287, + "step": 12877 + }, + { + "epoch": 0.28346907329958176, + "grad_norm": 1.3708066812921336, + "learning_rate": 8.407744829179596e-06, + "loss": 0.3006, + "step": 12878 + }, + { + "epoch": 0.2834910851860004, + "grad_norm": 0.6225940708920897, + "learning_rate": 8.40748397554038e-06, + "loss": 0.5255, + "step": 12879 + }, + { + "epoch": 0.2835130970724191, + "grad_norm": 1.2735249029349431, + "learning_rate": 8.407223104582812e-06, + "loss": 0.3237, + "step": 12880 + }, + { + "epoch": 0.28353510895883777, + "grad_norm": 1.5096382806533983, + "learning_rate": 8.406962216308215e-06, + "loss": 0.3651, + "step": 12881 + }, + { + "epoch": 0.2835571208452564, + "grad_norm": 1.2400847680462748, + "learning_rate": 8.406701310717917e-06, + "loss": 0.3516, + "step": 12882 + }, + { + "epoch": 0.2835791327316751, + "grad_norm": 1.759619963115768, + "learning_rate": 8.406440387813245e-06, + "loss": 0.3313, + "step": 12883 + }, + { + "epoch": 0.28360114461809377, + "grad_norm": 1.9600270891386133, + "learning_rate": 8.406179447595522e-06, + "loss": 0.3778, + "step": 12884 + }, + { + "epoch": 0.2836231565045124, + "grad_norm": 1.5538032306783953, + "learning_rate": 8.405918490066075e-06, + "loss": 0.3681, + "step": 12885 + }, + { + "epoch": 0.2836451683909311, + "grad_norm": 1.5416369408050627, + "learning_rate": 8.40565751522623e-06, + "loss": 0.2999, + "step": 12886 + }, + { + "epoch": 0.2836671802773498, + "grad_norm": 28.488813081904237, + "learning_rate": 8.405396523077317e-06, + "loss": 0.3302, + "step": 12887 + }, + { + "epoch": 0.2836891921637684, + "grad_norm": 1.423843622864229, + "learning_rate": 8.405135513620661e-06, + "loss": 0.3345, + "step": 12888 + }, + { + "epoch": 0.2837112040501871, + "grad_norm": 1.6327901807031124, + "learning_rate": 8.404874486857586e-06, + "loss": 0.3015, + "step": 12889 + }, + { + "epoch": 0.2837332159366058, + "grad_norm": 1.6245690527770742, + "learning_rate": 8.40461344278942e-06, + "loss": 0.3389, + "step": 12890 + }, + { + "epoch": 0.2837552278230244, + "grad_norm": 0.6262778360624829, + "learning_rate": 8.40435238141749e-06, + "loss": 0.5407, + "step": 12891 + }, + { + "epoch": 0.2837772397094431, + "grad_norm": 1.2753805024139098, + "learning_rate": 8.404091302743125e-06, + "loss": 0.3263, + "step": 12892 + }, + { + "epoch": 0.2837992515958618, + "grad_norm": 1.4813750740158285, + "learning_rate": 8.403830206767647e-06, + "loss": 0.2895, + "step": 12893 + }, + { + "epoch": 0.28382126348228043, + "grad_norm": 1.7962588827382249, + "learning_rate": 8.403569093492387e-06, + "loss": 0.4046, + "step": 12894 + }, + { + "epoch": 0.2838432753686991, + "grad_norm": 0.641614482824557, + "learning_rate": 8.40330796291867e-06, + "loss": 0.5288, + "step": 12895 + }, + { + "epoch": 0.2838652872551178, + "grad_norm": 1.6381905648380168, + "learning_rate": 8.403046815047824e-06, + "loss": 0.3461, + "step": 12896 + }, + { + "epoch": 0.28388729914153643, + "grad_norm": 2.1702317678575143, + "learning_rate": 8.402785649881178e-06, + "loss": 0.3682, + "step": 12897 + }, + { + "epoch": 0.2839093110279551, + "grad_norm": 0.62842042476524, + "learning_rate": 8.402524467420057e-06, + "loss": 0.5331, + "step": 12898 + }, + { + "epoch": 0.2839313229143738, + "grad_norm": 1.618336816038043, + "learning_rate": 8.402263267665787e-06, + "loss": 0.2921, + "step": 12899 + }, + { + "epoch": 0.28395333480079243, + "grad_norm": 1.6771977011150625, + "learning_rate": 8.4020020506197e-06, + "loss": 0.3087, + "step": 12900 + }, + { + "epoch": 0.2839753466872111, + "grad_norm": 2.1188087224174303, + "learning_rate": 8.401740816283121e-06, + "loss": 0.2666, + "step": 12901 + }, + { + "epoch": 0.28399735857362973, + "grad_norm": 1.6389461825631795, + "learning_rate": 8.401479564657377e-06, + "loss": 0.4536, + "step": 12902 + }, + { + "epoch": 0.28401937046004844, + "grad_norm": 7.480374935971879, + "learning_rate": 8.401218295743798e-06, + "loss": 0.309, + "step": 12903 + }, + { + "epoch": 0.2840413823464671, + "grad_norm": 1.486435161565446, + "learning_rate": 8.400957009543709e-06, + "loss": 0.3229, + "step": 12904 + }, + { + "epoch": 0.28406339423288574, + "grad_norm": 1.8875068026890713, + "learning_rate": 8.400695706058439e-06, + "loss": 0.3155, + "step": 12905 + }, + { + "epoch": 0.28408540611930444, + "grad_norm": 0.6308431628851593, + "learning_rate": 8.40043438528932e-06, + "loss": 0.5037, + "step": 12906 + }, + { + "epoch": 0.2841074180057231, + "grad_norm": 2.0166978435818845, + "learning_rate": 8.400173047237675e-06, + "loss": 0.314, + "step": 12907 + }, + { + "epoch": 0.28412942989214174, + "grad_norm": 3.2638414793332453, + "learning_rate": 8.399911691904834e-06, + "loss": 0.3381, + "step": 12908 + }, + { + "epoch": 0.28415144177856044, + "grad_norm": 2.0032989611983947, + "learning_rate": 8.399650319292124e-06, + "loss": 0.3371, + "step": 12909 + }, + { + "epoch": 0.2841734536649791, + "grad_norm": 1.399294282317434, + "learning_rate": 8.399388929400876e-06, + "loss": 0.3848, + "step": 12910 + }, + { + "epoch": 0.28419546555139774, + "grad_norm": 3.410463741959677, + "learning_rate": 8.399127522232417e-06, + "loss": 0.3881, + "step": 12911 + }, + { + "epoch": 0.28421747743781645, + "grad_norm": 1.6854031772639175, + "learning_rate": 8.398866097788077e-06, + "loss": 0.3347, + "step": 12912 + }, + { + "epoch": 0.2842394893242351, + "grad_norm": 1.3210340818841644, + "learning_rate": 8.398604656069182e-06, + "loss": 0.3475, + "step": 12913 + }, + { + "epoch": 0.28426150121065374, + "grad_norm": 1.4983293518349587, + "learning_rate": 8.398343197077063e-06, + "loss": 0.3496, + "step": 12914 + }, + { + "epoch": 0.2842835130970724, + "grad_norm": 1.360295608328931, + "learning_rate": 8.398081720813049e-06, + "loss": 0.372, + "step": 12915 + }, + { + "epoch": 0.2843055249834911, + "grad_norm": 0.6550966371572738, + "learning_rate": 8.397820227278468e-06, + "loss": 0.5216, + "step": 12916 + }, + { + "epoch": 0.28432753686990975, + "grad_norm": 1.8587073849469982, + "learning_rate": 8.397558716474649e-06, + "loss": 0.2973, + "step": 12917 + }, + { + "epoch": 0.2843495487563284, + "grad_norm": 1.6011129137590088, + "learning_rate": 8.39729718840292e-06, + "loss": 0.4165, + "step": 12918 + }, + { + "epoch": 0.2843715606427471, + "grad_norm": 1.082063025726161, + "learning_rate": 8.397035643064613e-06, + "loss": 0.3563, + "step": 12919 + }, + { + "epoch": 0.28439357252916575, + "grad_norm": 1.6051887719855782, + "learning_rate": 8.396774080461055e-06, + "loss": 0.392, + "step": 12920 + }, + { + "epoch": 0.2844155844155844, + "grad_norm": 1.4799140672773037, + "learning_rate": 8.396512500593576e-06, + "loss": 0.3223, + "step": 12921 + }, + { + "epoch": 0.2844375963020031, + "grad_norm": 0.6605976106571044, + "learning_rate": 8.396250903463505e-06, + "loss": 0.5212, + "step": 12922 + }, + { + "epoch": 0.28445960818842175, + "grad_norm": 0.6295695621055639, + "learning_rate": 8.395989289072172e-06, + "loss": 0.4914, + "step": 12923 + }, + { + "epoch": 0.2844816200748404, + "grad_norm": 0.6112521399667746, + "learning_rate": 8.395727657420909e-06, + "loss": 0.5067, + "step": 12924 + }, + { + "epoch": 0.2845036319612591, + "grad_norm": 1.4736928063104673, + "learning_rate": 8.395466008511041e-06, + "loss": 0.3527, + "step": 12925 + }, + { + "epoch": 0.28452564384767776, + "grad_norm": 1.5710364951665001, + "learning_rate": 8.395204342343902e-06, + "loss": 0.3444, + "step": 12926 + }, + { + "epoch": 0.2845476557340964, + "grad_norm": 1.5790121726172022, + "learning_rate": 8.394942658920819e-06, + "loss": 0.3472, + "step": 12927 + }, + { + "epoch": 0.28456966762051505, + "grad_norm": 1.7421273803937445, + "learning_rate": 8.394680958243125e-06, + "loss": 0.2981, + "step": 12928 + }, + { + "epoch": 0.28459167950693376, + "grad_norm": 7.858455606341616, + "learning_rate": 8.394419240312149e-06, + "loss": 0.2767, + "step": 12929 + }, + { + "epoch": 0.2846136913933524, + "grad_norm": 1.3384650589047464, + "learning_rate": 8.394157505129217e-06, + "loss": 0.285, + "step": 12930 + }, + { + "epoch": 0.28463570327977106, + "grad_norm": 1.6020311515934584, + "learning_rate": 8.393895752695665e-06, + "loss": 0.361, + "step": 12931 + }, + { + "epoch": 0.28465771516618976, + "grad_norm": 2.134624430635148, + "learning_rate": 8.393633983012819e-06, + "loss": 0.3369, + "step": 12932 + }, + { + "epoch": 0.2846797270526084, + "grad_norm": 3.7253649095162813, + "learning_rate": 8.393372196082012e-06, + "loss": 0.3346, + "step": 12933 + }, + { + "epoch": 0.28470173893902706, + "grad_norm": 0.7977632442551115, + "learning_rate": 8.393110391904577e-06, + "loss": 0.5257, + "step": 12934 + }, + { + "epoch": 0.28472375082544576, + "grad_norm": 1.8327520138887623, + "learning_rate": 8.39284857048184e-06, + "loss": 0.3793, + "step": 12935 + }, + { + "epoch": 0.2847457627118644, + "grad_norm": 1.7968072744809398, + "learning_rate": 8.392586731815131e-06, + "loss": 0.2858, + "step": 12936 + }, + { + "epoch": 0.28476777459828306, + "grad_norm": 1.6825062601782166, + "learning_rate": 8.392324875905786e-06, + "loss": 0.2692, + "step": 12937 + }, + { + "epoch": 0.2847897864847017, + "grad_norm": 1.6293838271901755, + "learning_rate": 8.392063002755133e-06, + "loss": 0.2879, + "step": 12938 + }, + { + "epoch": 0.2848117983711204, + "grad_norm": 1.332482731926294, + "learning_rate": 8.391801112364502e-06, + "loss": 0.2853, + "step": 12939 + }, + { + "epoch": 0.28483381025753907, + "grad_norm": 2.0257912795489923, + "learning_rate": 8.391539204735223e-06, + "loss": 0.2825, + "step": 12940 + }, + { + "epoch": 0.2848558221439577, + "grad_norm": 1.5228114584274974, + "learning_rate": 8.391277279868631e-06, + "loss": 0.3205, + "step": 12941 + }, + { + "epoch": 0.2848778340303764, + "grad_norm": 1.825520236457616, + "learning_rate": 8.391015337766056e-06, + "loss": 0.3149, + "step": 12942 + }, + { + "epoch": 0.28489984591679507, + "grad_norm": 1.5145080132384592, + "learning_rate": 8.39075337842883e-06, + "loss": 0.3287, + "step": 12943 + }, + { + "epoch": 0.2849218578032137, + "grad_norm": 1.7957260739531757, + "learning_rate": 8.390491401858279e-06, + "loss": 0.3166, + "step": 12944 + }, + { + "epoch": 0.2849438696896324, + "grad_norm": 1.792550224949473, + "learning_rate": 8.39022940805574e-06, + "loss": 0.3144, + "step": 12945 + }, + { + "epoch": 0.28496588157605107, + "grad_norm": 0.713184887094533, + "learning_rate": 8.389967397022542e-06, + "loss": 0.5042, + "step": 12946 + }, + { + "epoch": 0.2849878934624697, + "grad_norm": 1.7469724035368284, + "learning_rate": 8.389705368760021e-06, + "loss": 0.4413, + "step": 12947 + }, + { + "epoch": 0.2850099053488884, + "grad_norm": 1.4904185853500929, + "learning_rate": 8.389443323269503e-06, + "loss": 0.3315, + "step": 12948 + }, + { + "epoch": 0.2850319172353071, + "grad_norm": 1.6062366834898223, + "learning_rate": 8.389181260552323e-06, + "loss": 0.2664, + "step": 12949 + }, + { + "epoch": 0.2850539291217257, + "grad_norm": 1.8317659521384757, + "learning_rate": 8.388919180609812e-06, + "loss": 0.4001, + "step": 12950 + }, + { + "epoch": 0.2850759410081444, + "grad_norm": 0.6446630218338167, + "learning_rate": 8.388657083443302e-06, + "loss": 0.5276, + "step": 12951 + }, + { + "epoch": 0.2850979528945631, + "grad_norm": 1.4317851835392894, + "learning_rate": 8.388394969054126e-06, + "loss": 0.3119, + "step": 12952 + }, + { + "epoch": 0.2851199647809817, + "grad_norm": 1.9561060714587426, + "learning_rate": 8.388132837443616e-06, + "loss": 0.2814, + "step": 12953 + }, + { + "epoch": 0.2851419766674004, + "grad_norm": 1.6294230034955552, + "learning_rate": 8.387870688613103e-06, + "loss": 0.2789, + "step": 12954 + }, + { + "epoch": 0.2851639885538191, + "grad_norm": 1.5259964203128376, + "learning_rate": 8.38760852256392e-06, + "loss": 0.366, + "step": 12955 + }, + { + "epoch": 0.28518600044023773, + "grad_norm": 0.6551297491502118, + "learning_rate": 8.3873463392974e-06, + "loss": 0.5308, + "step": 12956 + }, + { + "epoch": 0.2852080123266564, + "grad_norm": 4.877444823703843, + "learning_rate": 8.387084138814874e-06, + "loss": 0.4212, + "step": 12957 + }, + { + "epoch": 0.2852300242130751, + "grad_norm": 1.5071911565495304, + "learning_rate": 8.386821921117676e-06, + "loss": 0.359, + "step": 12958 + }, + { + "epoch": 0.28525203609949373, + "grad_norm": 1.8452384532972936, + "learning_rate": 8.386559686207138e-06, + "loss": 0.3719, + "step": 12959 + }, + { + "epoch": 0.2852740479859124, + "grad_norm": 1.9540988683565164, + "learning_rate": 8.386297434084595e-06, + "loss": 0.3822, + "step": 12960 + }, + { + "epoch": 0.28529605987233103, + "grad_norm": 2.2782001783609056, + "learning_rate": 8.386035164751377e-06, + "loss": 0.3716, + "step": 12961 + }, + { + "epoch": 0.28531807175874974, + "grad_norm": 1.9326444638752716, + "learning_rate": 8.385772878208819e-06, + "loss": 0.3347, + "step": 12962 + }, + { + "epoch": 0.2853400836451684, + "grad_norm": 3.429255603687335, + "learning_rate": 8.385510574458252e-06, + "loss": 0.3778, + "step": 12963 + }, + { + "epoch": 0.28536209553158703, + "grad_norm": 1.5252189158981613, + "learning_rate": 8.38524825350101e-06, + "loss": 0.3363, + "step": 12964 + }, + { + "epoch": 0.28538410741800574, + "grad_norm": 1.5969871999153356, + "learning_rate": 8.384985915338429e-06, + "loss": 0.312, + "step": 12965 + }, + { + "epoch": 0.2854061193044244, + "grad_norm": 2.1503630872180945, + "learning_rate": 8.384723559971837e-06, + "loss": 0.3766, + "step": 12966 + }, + { + "epoch": 0.28542813119084304, + "grad_norm": 2.2527197991811265, + "learning_rate": 8.384461187402571e-06, + "loss": 0.3398, + "step": 12967 + }, + { + "epoch": 0.28545014307726174, + "grad_norm": 2.2613271784075515, + "learning_rate": 8.384198797631964e-06, + "loss": 0.3487, + "step": 12968 + }, + { + "epoch": 0.2854721549636804, + "grad_norm": 2.1047509913070894, + "learning_rate": 8.38393639066135e-06, + "loss": 0.3543, + "step": 12969 + }, + { + "epoch": 0.28549416685009904, + "grad_norm": 1.7568107459201483, + "learning_rate": 8.38367396649206e-06, + "loss": 0.3429, + "step": 12970 + }, + { + "epoch": 0.28551617873651775, + "grad_norm": 1.9719743370682137, + "learning_rate": 8.383411525125431e-06, + "loss": 0.3619, + "step": 12971 + }, + { + "epoch": 0.2855381906229364, + "grad_norm": 3.3312753816092635, + "learning_rate": 8.383149066562797e-06, + "loss": 0.3599, + "step": 12972 + }, + { + "epoch": 0.28556020250935504, + "grad_norm": 2.6222625867676035, + "learning_rate": 8.38288659080549e-06, + "loss": 0.4223, + "step": 12973 + }, + { + "epoch": 0.2855822143957737, + "grad_norm": 2.9420054876478456, + "learning_rate": 8.382624097854842e-06, + "loss": 0.3623, + "step": 12974 + }, + { + "epoch": 0.2856042262821924, + "grad_norm": 2.115531016801053, + "learning_rate": 8.382361587712194e-06, + "loss": 0.3953, + "step": 12975 + }, + { + "epoch": 0.28562623816861105, + "grad_norm": 1.724772598697432, + "learning_rate": 8.382099060378873e-06, + "loss": 0.2938, + "step": 12976 + }, + { + "epoch": 0.2856482500550297, + "grad_norm": 2.023353688438777, + "learning_rate": 8.381836515856214e-06, + "loss": 0.3955, + "step": 12977 + }, + { + "epoch": 0.2856702619414484, + "grad_norm": 2.1834350248241305, + "learning_rate": 8.381573954145556e-06, + "loss": 0.3966, + "step": 12978 + }, + { + "epoch": 0.28569227382786705, + "grad_norm": 2.22913803096788, + "learning_rate": 8.38131137524823e-06, + "loss": 0.452, + "step": 12979 + }, + { + "epoch": 0.2857142857142857, + "grad_norm": 2.169251568608082, + "learning_rate": 8.381048779165574e-06, + "loss": 0.3119, + "step": 12980 + }, + { + "epoch": 0.2857362976007044, + "grad_norm": 1.9940797267006454, + "learning_rate": 8.380786165898917e-06, + "loss": 0.3256, + "step": 12981 + }, + { + "epoch": 0.28575830948712305, + "grad_norm": 0.6994579836086681, + "learning_rate": 8.380523535449599e-06, + "loss": 0.5348, + "step": 12982 + }, + { + "epoch": 0.2857803213735417, + "grad_norm": 1.6933913157351788, + "learning_rate": 8.380260887818951e-06, + "loss": 0.331, + "step": 12983 + }, + { + "epoch": 0.2858023332599604, + "grad_norm": 1.82500601918532, + "learning_rate": 8.379998223008312e-06, + "loss": 0.3274, + "step": 12984 + }, + { + "epoch": 0.28582434514637906, + "grad_norm": 1.6697540105250666, + "learning_rate": 8.379735541019011e-06, + "loss": 0.2805, + "step": 12985 + }, + { + "epoch": 0.2858463570327977, + "grad_norm": 1.6424602246652422, + "learning_rate": 8.37947284185239e-06, + "loss": 0.3004, + "step": 12986 + }, + { + "epoch": 0.28586836891921635, + "grad_norm": 1.7790124805366805, + "learning_rate": 8.379210125509778e-06, + "loss": 0.2976, + "step": 12987 + }, + { + "epoch": 0.28589038080563506, + "grad_norm": 2.045744048239977, + "learning_rate": 8.378947391992515e-06, + "loss": 0.3748, + "step": 12988 + }, + { + "epoch": 0.2859123926920537, + "grad_norm": 2.0392977072509813, + "learning_rate": 8.378684641301932e-06, + "loss": 0.2565, + "step": 12989 + }, + { + "epoch": 0.28593440457847236, + "grad_norm": 4.823074457349405, + "learning_rate": 8.378421873439369e-06, + "loss": 0.2907, + "step": 12990 + }, + { + "epoch": 0.28595641646489106, + "grad_norm": 1.854576740403096, + "learning_rate": 8.378159088406159e-06, + "loss": 0.4114, + "step": 12991 + }, + { + "epoch": 0.2859784283513097, + "grad_norm": 3.5057730732660746, + "learning_rate": 8.377896286203638e-06, + "loss": 0.3269, + "step": 12992 + }, + { + "epoch": 0.28600044023772836, + "grad_norm": 2.5728545925032344, + "learning_rate": 8.37763346683314e-06, + "loss": 0.3052, + "step": 12993 + }, + { + "epoch": 0.28602245212414706, + "grad_norm": 1.7836602913356232, + "learning_rate": 8.377370630296002e-06, + "loss": 0.439, + "step": 12994 + }, + { + "epoch": 0.2860444640105657, + "grad_norm": 3.7324378434014576, + "learning_rate": 8.377107776593563e-06, + "loss": 0.3189, + "step": 12995 + }, + { + "epoch": 0.28606647589698436, + "grad_norm": 1.4425882323717427, + "learning_rate": 8.376844905727153e-06, + "loss": 0.3412, + "step": 12996 + }, + { + "epoch": 0.286088487783403, + "grad_norm": 1.7344719923266456, + "learning_rate": 8.376582017698111e-06, + "loss": 0.3492, + "step": 12997 + }, + { + "epoch": 0.2861104996698217, + "grad_norm": 2.1292498594820124, + "learning_rate": 8.376319112507776e-06, + "loss": 0.3878, + "step": 12998 + }, + { + "epoch": 0.28613251155624037, + "grad_norm": 1.4406429052491223, + "learning_rate": 8.37605619015748e-06, + "loss": 0.2533, + "step": 12999 + }, + { + "epoch": 0.286154523442659, + "grad_norm": 0.662158835414709, + "learning_rate": 8.37579325064856e-06, + "loss": 0.5138, + "step": 13000 + }, + { + "epoch": 0.2861765353290777, + "grad_norm": 1.568188298956813, + "learning_rate": 8.375530293982353e-06, + "loss": 0.3311, + "step": 13001 + }, + { + "epoch": 0.28619854721549637, + "grad_norm": 1.380371151233675, + "learning_rate": 8.375267320160196e-06, + "loss": 0.3138, + "step": 13002 + }, + { + "epoch": 0.286220559101915, + "grad_norm": 1.7738547190410228, + "learning_rate": 8.375004329183425e-06, + "loss": 0.3328, + "step": 13003 + }, + { + "epoch": 0.2862425709883337, + "grad_norm": 1.5412348286491802, + "learning_rate": 8.374741321053377e-06, + "loss": 0.2998, + "step": 13004 + }, + { + "epoch": 0.28626458287475237, + "grad_norm": 1.951802535926351, + "learning_rate": 8.374478295771388e-06, + "loss": 0.3001, + "step": 13005 + }, + { + "epoch": 0.286286594761171, + "grad_norm": 1.4575039496243403, + "learning_rate": 8.374215253338796e-06, + "loss": 0.3303, + "step": 13006 + }, + { + "epoch": 0.2863086066475897, + "grad_norm": 2.162599628942179, + "learning_rate": 8.373952193756935e-06, + "loss": 0.3177, + "step": 13007 + }, + { + "epoch": 0.2863306185340084, + "grad_norm": 1.5222074262435317, + "learning_rate": 8.373689117027147e-06, + "loss": 0.3045, + "step": 13008 + }, + { + "epoch": 0.286352630420427, + "grad_norm": 2.084531643869095, + "learning_rate": 8.373426023150765e-06, + "loss": 0.2832, + "step": 13009 + }, + { + "epoch": 0.2863746423068457, + "grad_norm": 0.6369679649533222, + "learning_rate": 8.373162912129127e-06, + "loss": 0.5276, + "step": 13010 + }, + { + "epoch": 0.2863966541932644, + "grad_norm": 1.7122260658921147, + "learning_rate": 8.37289978396357e-06, + "loss": 0.3104, + "step": 13011 + }, + { + "epoch": 0.286418666079683, + "grad_norm": 1.9276294805752547, + "learning_rate": 8.372636638655434e-06, + "loss": 0.205, + "step": 13012 + }, + { + "epoch": 0.2864406779661017, + "grad_norm": 1.9190017295542077, + "learning_rate": 8.372373476206052e-06, + "loss": 0.3294, + "step": 13013 + }, + { + "epoch": 0.2864626898525204, + "grad_norm": 0.6242808735184647, + "learning_rate": 8.372110296616765e-06, + "loss": 0.4977, + "step": 13014 + }, + { + "epoch": 0.28648470173893903, + "grad_norm": 2.5014734109187753, + "learning_rate": 8.37184709988891e-06, + "loss": 0.3269, + "step": 13015 + }, + { + "epoch": 0.2865067136253577, + "grad_norm": 2.416353465928437, + "learning_rate": 8.371583886023823e-06, + "loss": 0.2932, + "step": 13016 + }, + { + "epoch": 0.2865287255117764, + "grad_norm": 1.60817345605226, + "learning_rate": 8.371320655022843e-06, + "loss": 0.4193, + "step": 13017 + }, + { + "epoch": 0.28655073739819503, + "grad_norm": 1.3227686065899598, + "learning_rate": 8.37105740688731e-06, + "loss": 0.3283, + "step": 13018 + }, + { + "epoch": 0.2865727492846137, + "grad_norm": 1.4362608536373582, + "learning_rate": 8.370794141618557e-06, + "loss": 0.3116, + "step": 13019 + }, + { + "epoch": 0.28659476117103233, + "grad_norm": 0.6361432838662318, + "learning_rate": 8.370530859217927e-06, + "loss": 0.503, + "step": 13020 + }, + { + "epoch": 0.28661677305745104, + "grad_norm": 1.4587675123704784, + "learning_rate": 8.370267559686754e-06, + "loss": 0.3106, + "step": 13021 + }, + { + "epoch": 0.2866387849438697, + "grad_norm": 1.854358736045502, + "learning_rate": 8.370004243026379e-06, + "loss": 0.2919, + "step": 13022 + }, + { + "epoch": 0.28666079683028833, + "grad_norm": 1.4379971563027332, + "learning_rate": 8.36974090923814e-06, + "loss": 0.3184, + "step": 13023 + }, + { + "epoch": 0.28668280871670704, + "grad_norm": 1.233765899570096, + "learning_rate": 8.369477558323374e-06, + "loss": 0.3286, + "step": 13024 + }, + { + "epoch": 0.2867048206031257, + "grad_norm": 7.065814468230558, + "learning_rate": 8.36921419028342e-06, + "loss": 0.3309, + "step": 13025 + }, + { + "epoch": 0.28672683248954434, + "grad_norm": 2.5318662523679274, + "learning_rate": 8.368950805119616e-06, + "loss": 0.3929, + "step": 13026 + }, + { + "epoch": 0.28674884437596304, + "grad_norm": 2.1484431018024717, + "learning_rate": 8.368687402833302e-06, + "loss": 0.4797, + "step": 13027 + }, + { + "epoch": 0.2867708562623817, + "grad_norm": 1.7506849246236982, + "learning_rate": 8.368423983425817e-06, + "loss": 0.3507, + "step": 13028 + }, + { + "epoch": 0.28679286814880034, + "grad_norm": 1.748851452578592, + "learning_rate": 8.368160546898499e-06, + "loss": 0.3913, + "step": 13029 + }, + { + "epoch": 0.28681488003521904, + "grad_norm": 1.7511489169983705, + "learning_rate": 8.367897093252686e-06, + "loss": 0.3931, + "step": 13030 + }, + { + "epoch": 0.2868368919216377, + "grad_norm": 1.307333419395354, + "learning_rate": 8.367633622489718e-06, + "loss": 0.2352, + "step": 13031 + }, + { + "epoch": 0.28685890380805634, + "grad_norm": 0.6612266087138384, + "learning_rate": 8.367370134610935e-06, + "loss": 0.5271, + "step": 13032 + }, + { + "epoch": 0.286880915694475, + "grad_norm": 1.9548366811584417, + "learning_rate": 8.367106629617673e-06, + "loss": 0.3693, + "step": 13033 + }, + { + "epoch": 0.2869029275808937, + "grad_norm": 7.404828722827004, + "learning_rate": 8.366843107511275e-06, + "loss": 0.3216, + "step": 13034 + }, + { + "epoch": 0.28692493946731235, + "grad_norm": 2.744561525681749, + "learning_rate": 8.366579568293078e-06, + "loss": 0.4154, + "step": 13035 + }, + { + "epoch": 0.286946951353731, + "grad_norm": 0.6291013885254271, + "learning_rate": 8.366316011964422e-06, + "loss": 0.5128, + "step": 13036 + }, + { + "epoch": 0.2869689632401497, + "grad_norm": 0.6313131309808109, + "learning_rate": 8.366052438526648e-06, + "loss": 0.511, + "step": 13037 + }, + { + "epoch": 0.28699097512656835, + "grad_norm": 2.2844373545715038, + "learning_rate": 8.365788847981093e-06, + "loss": 0.3818, + "step": 13038 + }, + { + "epoch": 0.287012987012987, + "grad_norm": 0.632907205218349, + "learning_rate": 8.365525240329098e-06, + "loss": 0.5168, + "step": 13039 + }, + { + "epoch": 0.2870349988994057, + "grad_norm": 1.711351549190098, + "learning_rate": 8.365261615572002e-06, + "loss": 0.344, + "step": 13040 + }, + { + "epoch": 0.28705701078582435, + "grad_norm": 1.15879067627263, + "learning_rate": 8.364997973711147e-06, + "loss": 0.2552, + "step": 13041 + }, + { + "epoch": 0.287079022672243, + "grad_norm": 1.257292210467871, + "learning_rate": 8.36473431474787e-06, + "loss": 0.3256, + "step": 13042 + }, + { + "epoch": 0.28710103455866165, + "grad_norm": 2.4585905676216955, + "learning_rate": 8.364470638683512e-06, + "loss": 0.3223, + "step": 13043 + }, + { + "epoch": 0.28712304644508035, + "grad_norm": 1.3578254374001688, + "learning_rate": 8.364206945519415e-06, + "loss": 0.3114, + "step": 13044 + }, + { + "epoch": 0.287145058331499, + "grad_norm": 0.6288863574848095, + "learning_rate": 8.363943235256919e-06, + "loss": 0.5105, + "step": 13045 + }, + { + "epoch": 0.28716707021791765, + "grad_norm": 1.4316636974598467, + "learning_rate": 8.36367950789736e-06, + "loss": 0.3252, + "step": 13046 + }, + { + "epoch": 0.28718908210433636, + "grad_norm": 0.6311795449326113, + "learning_rate": 8.363415763442084e-06, + "loss": 0.5281, + "step": 13047 + }, + { + "epoch": 0.287211093990755, + "grad_norm": 0.6040857178824801, + "learning_rate": 8.363152001892428e-06, + "loss": 0.5102, + "step": 13048 + }, + { + "epoch": 0.28723310587717366, + "grad_norm": 2.122907073206102, + "learning_rate": 8.362888223249734e-06, + "loss": 0.3552, + "step": 13049 + }, + { + "epoch": 0.28725511776359236, + "grad_norm": 1.7201255275036873, + "learning_rate": 8.362624427515342e-06, + "loss": 0.3558, + "step": 13050 + }, + { + "epoch": 0.287277129650011, + "grad_norm": 1.2812270542188364, + "learning_rate": 8.362360614690592e-06, + "loss": 0.4186, + "step": 13051 + }, + { + "epoch": 0.28729914153642966, + "grad_norm": 1.3382443853067347, + "learning_rate": 8.362096784776827e-06, + "loss": 0.3234, + "step": 13052 + }, + { + "epoch": 0.28732115342284836, + "grad_norm": 1.4327309738576608, + "learning_rate": 8.361832937775386e-06, + "loss": 0.3253, + "step": 13053 + }, + { + "epoch": 0.287343165309267, + "grad_norm": 1.543465458624538, + "learning_rate": 8.361569073687609e-06, + "loss": 0.3445, + "step": 13054 + }, + { + "epoch": 0.28736517719568566, + "grad_norm": 1.4352729452546686, + "learning_rate": 8.361305192514842e-06, + "loss": 0.4245, + "step": 13055 + }, + { + "epoch": 0.2873871890821043, + "grad_norm": 1.9775938632911814, + "learning_rate": 8.361041294258421e-06, + "loss": 0.333, + "step": 13056 + }, + { + "epoch": 0.287409200968523, + "grad_norm": 1.5718337935588806, + "learning_rate": 8.36077737891969e-06, + "loss": 0.3703, + "step": 13057 + }, + { + "epoch": 0.28743121285494166, + "grad_norm": 1.4203717840196017, + "learning_rate": 8.360513446499987e-06, + "loss": 0.3736, + "step": 13058 + }, + { + "epoch": 0.2874532247413603, + "grad_norm": 1.3583411807422088, + "learning_rate": 8.360249497000658e-06, + "loss": 0.3783, + "step": 13059 + }, + { + "epoch": 0.287475236627779, + "grad_norm": 1.5692038327259334, + "learning_rate": 8.359985530423042e-06, + "loss": 0.4001, + "step": 13060 + }, + { + "epoch": 0.28749724851419767, + "grad_norm": 1.3050176938981213, + "learning_rate": 8.35972154676848e-06, + "loss": 0.3562, + "step": 13061 + }, + { + "epoch": 0.2875192604006163, + "grad_norm": 0.7864499819852061, + "learning_rate": 8.359457546038317e-06, + "loss": 0.5317, + "step": 13062 + }, + { + "epoch": 0.287541272287035, + "grad_norm": 1.4870618427653852, + "learning_rate": 8.35919352823389e-06, + "loss": 0.3039, + "step": 13063 + }, + { + "epoch": 0.28756328417345367, + "grad_norm": 1.2548092130483715, + "learning_rate": 8.358929493356544e-06, + "loss": 0.3334, + "step": 13064 + }, + { + "epoch": 0.2875852960598723, + "grad_norm": 1.0158864928538727, + "learning_rate": 8.358665441407621e-06, + "loss": 0.2322, + "step": 13065 + }, + { + "epoch": 0.287607307946291, + "grad_norm": 0.6709175128653028, + "learning_rate": 8.358401372388461e-06, + "loss": 0.5257, + "step": 13066 + }, + { + "epoch": 0.2876293198327097, + "grad_norm": 4.060047327621179, + "learning_rate": 8.358137286300407e-06, + "loss": 0.3531, + "step": 13067 + }, + { + "epoch": 0.2876513317191283, + "grad_norm": 1.888959506924991, + "learning_rate": 8.357873183144803e-06, + "loss": 0.2536, + "step": 13068 + }, + { + "epoch": 0.28767334360554697, + "grad_norm": 1.809759447315894, + "learning_rate": 8.357609062922988e-06, + "loss": 0.3364, + "step": 13069 + }, + { + "epoch": 0.2876953554919657, + "grad_norm": 0.651935082029873, + "learning_rate": 8.357344925636308e-06, + "loss": 0.5242, + "step": 13070 + }, + { + "epoch": 0.2877173673783843, + "grad_norm": 1.6670059352754183, + "learning_rate": 8.357080771286103e-06, + "loss": 0.3642, + "step": 13071 + }, + { + "epoch": 0.287739379264803, + "grad_norm": 1.440843450329284, + "learning_rate": 8.356816599873715e-06, + "loss": 0.3269, + "step": 13072 + }, + { + "epoch": 0.2877613911512217, + "grad_norm": 2.4463292499708116, + "learning_rate": 8.35655241140049e-06, + "loss": 0.3589, + "step": 13073 + }, + { + "epoch": 0.28778340303764033, + "grad_norm": 3.431972414370633, + "learning_rate": 8.356288205867767e-06, + "loss": 0.3337, + "step": 13074 + }, + { + "epoch": 0.287805414924059, + "grad_norm": 1.858520181531193, + "learning_rate": 8.356023983276891e-06, + "loss": 0.3903, + "step": 13075 + }, + { + "epoch": 0.2878274268104777, + "grad_norm": 3.5983568984784884, + "learning_rate": 8.355759743629205e-06, + "loss": 0.2851, + "step": 13076 + }, + { + "epoch": 0.28784943869689633, + "grad_norm": 1.764216748232772, + "learning_rate": 8.35549548692605e-06, + "loss": 0.4139, + "step": 13077 + }, + { + "epoch": 0.287871450583315, + "grad_norm": 0.6501185817033801, + "learning_rate": 8.355231213168771e-06, + "loss": 0.522, + "step": 13078 + }, + { + "epoch": 0.28789346246973363, + "grad_norm": 1.937432540789804, + "learning_rate": 8.35496692235871e-06, + "loss": 0.3718, + "step": 13079 + }, + { + "epoch": 0.28791547435615233, + "grad_norm": 2.0431832671967296, + "learning_rate": 8.354702614497213e-06, + "loss": 0.3372, + "step": 13080 + }, + { + "epoch": 0.287937486242571, + "grad_norm": 0.6285480106545525, + "learning_rate": 8.354438289585618e-06, + "loss": 0.5314, + "step": 13081 + }, + { + "epoch": 0.28795949812898963, + "grad_norm": 1.6363362408329467, + "learning_rate": 8.354173947625272e-06, + "loss": 0.434, + "step": 13082 + }, + { + "epoch": 0.28798151001540834, + "grad_norm": 1.8771740697000106, + "learning_rate": 8.35390958861752e-06, + "loss": 0.2667, + "step": 13083 + }, + { + "epoch": 0.288003521901827, + "grad_norm": 0.6207531402565699, + "learning_rate": 8.353645212563701e-06, + "loss": 0.5236, + "step": 13084 + }, + { + "epoch": 0.28802553378824564, + "grad_norm": 1.799755380162308, + "learning_rate": 8.353380819465163e-06, + "loss": 0.3385, + "step": 13085 + }, + { + "epoch": 0.28804754567466434, + "grad_norm": 1.2191979861034767, + "learning_rate": 8.353116409323247e-06, + "loss": 0.4215, + "step": 13086 + }, + { + "epoch": 0.288069557561083, + "grad_norm": 1.8926740184703175, + "learning_rate": 8.352851982139299e-06, + "loss": 0.3369, + "step": 13087 + }, + { + "epoch": 0.28809156944750164, + "grad_norm": 1.5058112652548454, + "learning_rate": 8.352587537914661e-06, + "loss": 0.3628, + "step": 13088 + }, + { + "epoch": 0.28811358133392034, + "grad_norm": 1.5790376216654858, + "learning_rate": 8.352323076650679e-06, + "loss": 0.3607, + "step": 13089 + }, + { + "epoch": 0.288135593220339, + "grad_norm": 2.4767736114485825, + "learning_rate": 8.352058598348695e-06, + "loss": 0.4084, + "step": 13090 + }, + { + "epoch": 0.28815760510675764, + "grad_norm": 1.6371997028471397, + "learning_rate": 8.351794103010054e-06, + "loss": 0.2949, + "step": 13091 + }, + { + "epoch": 0.2881796169931763, + "grad_norm": 3.1593178057186115, + "learning_rate": 8.3515295906361e-06, + "loss": 0.3793, + "step": 13092 + }, + { + "epoch": 0.288201628879595, + "grad_norm": 1.2913016903980556, + "learning_rate": 8.35126506122818e-06, + "loss": 0.2799, + "step": 13093 + }, + { + "epoch": 0.28822364076601364, + "grad_norm": 2.7127197235280436, + "learning_rate": 8.351000514787633e-06, + "loss": 0.4156, + "step": 13094 + }, + { + "epoch": 0.2882456526524323, + "grad_norm": 1.5179757359493136, + "learning_rate": 8.350735951315809e-06, + "loss": 0.2721, + "step": 13095 + }, + { + "epoch": 0.288267664538851, + "grad_norm": 1.496802985507487, + "learning_rate": 8.35047137081405e-06, + "loss": 0.3695, + "step": 13096 + }, + { + "epoch": 0.28828967642526965, + "grad_norm": 1.9386714427169311, + "learning_rate": 8.350206773283701e-06, + "loss": 0.264, + "step": 13097 + }, + { + "epoch": 0.2883116883116883, + "grad_norm": 2.0772810419485266, + "learning_rate": 8.349942158726108e-06, + "loss": 0.2869, + "step": 13098 + }, + { + "epoch": 0.288333700198107, + "grad_norm": 2.469194234539552, + "learning_rate": 8.349677527142613e-06, + "loss": 0.3708, + "step": 13099 + }, + { + "epoch": 0.28835571208452565, + "grad_norm": 1.6597854351600227, + "learning_rate": 8.349412878534565e-06, + "loss": 0.4019, + "step": 13100 + }, + { + "epoch": 0.2883777239709443, + "grad_norm": 1.5561524723681395, + "learning_rate": 8.349148212903304e-06, + "loss": 0.3835, + "step": 13101 + }, + { + "epoch": 0.28839973585736295, + "grad_norm": 1.9132460303471404, + "learning_rate": 8.34888353025018e-06, + "loss": 0.3572, + "step": 13102 + }, + { + "epoch": 0.28842174774378165, + "grad_norm": 1.4591220616452407, + "learning_rate": 8.348618830576536e-06, + "loss": 0.3585, + "step": 13103 + }, + { + "epoch": 0.2884437596302003, + "grad_norm": 1.588480128444486, + "learning_rate": 8.348354113883716e-06, + "loss": 0.3463, + "step": 13104 + }, + { + "epoch": 0.28846577151661895, + "grad_norm": 0.7379518129464622, + "learning_rate": 8.348089380173068e-06, + "loss": 0.4956, + "step": 13105 + }, + { + "epoch": 0.28848778340303766, + "grad_norm": 1.6099691492443993, + "learning_rate": 8.347824629445937e-06, + "loss": 0.3091, + "step": 13106 + }, + { + "epoch": 0.2885097952894563, + "grad_norm": 1.9303696697979482, + "learning_rate": 8.347559861703669e-06, + "loss": 0.3609, + "step": 13107 + }, + { + "epoch": 0.28853180717587495, + "grad_norm": 3.675807411910425, + "learning_rate": 8.347295076947609e-06, + "loss": 0.2909, + "step": 13108 + }, + { + "epoch": 0.28855381906229366, + "grad_norm": 1.9494669518140013, + "learning_rate": 8.3470302751791e-06, + "loss": 0.3051, + "step": 13109 + }, + { + "epoch": 0.2885758309487123, + "grad_norm": 1.8612304960710175, + "learning_rate": 8.34676545639949e-06, + "loss": 0.4368, + "step": 13110 + }, + { + "epoch": 0.28859784283513096, + "grad_norm": 1.6405484823433543, + "learning_rate": 8.346500620610128e-06, + "loss": 0.3557, + "step": 13111 + }, + { + "epoch": 0.28861985472154966, + "grad_norm": 1.657296722612335, + "learning_rate": 8.346235767812357e-06, + "loss": 0.319, + "step": 13112 + }, + { + "epoch": 0.2886418666079683, + "grad_norm": 1.5552830429509925, + "learning_rate": 8.345970898007521e-06, + "loss": 0.3599, + "step": 13113 + }, + { + "epoch": 0.28866387849438696, + "grad_norm": 1.228824576824448, + "learning_rate": 8.34570601119697e-06, + "loss": 0.366, + "step": 13114 + }, + { + "epoch": 0.2886858903808056, + "grad_norm": 1.4271015661539475, + "learning_rate": 8.345441107382048e-06, + "loss": 0.3781, + "step": 13115 + }, + { + "epoch": 0.2887079022672243, + "grad_norm": 1.228709959227179, + "learning_rate": 8.345176186564103e-06, + "loss": 0.2885, + "step": 13116 + }, + { + "epoch": 0.28872991415364296, + "grad_norm": 1.1835928817468893, + "learning_rate": 8.34491124874448e-06, + "loss": 0.2846, + "step": 13117 + }, + { + "epoch": 0.2887519260400616, + "grad_norm": 1.6944683267583975, + "learning_rate": 8.344646293924528e-06, + "loss": 0.3674, + "step": 13118 + }, + { + "epoch": 0.2887739379264803, + "grad_norm": 2.701693117014847, + "learning_rate": 8.344381322105588e-06, + "loss": 0.3329, + "step": 13119 + }, + { + "epoch": 0.28879594981289897, + "grad_norm": 1.4007803274320336, + "learning_rate": 8.344116333289014e-06, + "loss": 0.3351, + "step": 13120 + }, + { + "epoch": 0.2888179616993176, + "grad_norm": 1.3970200426644606, + "learning_rate": 8.343851327476149e-06, + "loss": 0.3038, + "step": 13121 + }, + { + "epoch": 0.2888399735857363, + "grad_norm": 1.894731904173759, + "learning_rate": 8.343586304668338e-06, + "loss": 0.3679, + "step": 13122 + }, + { + "epoch": 0.28886198547215497, + "grad_norm": 3.7668876673763227, + "learning_rate": 8.343321264866933e-06, + "loss": 0.3136, + "step": 13123 + }, + { + "epoch": 0.2888839973585736, + "grad_norm": 1.408529871168106, + "learning_rate": 8.343056208073277e-06, + "loss": 0.3966, + "step": 13124 + }, + { + "epoch": 0.2889060092449923, + "grad_norm": 1.5992153459286385, + "learning_rate": 8.342791134288716e-06, + "loss": 0.3767, + "step": 13125 + }, + { + "epoch": 0.288928021131411, + "grad_norm": 3.798242936945453, + "learning_rate": 8.342526043514603e-06, + "loss": 0.3043, + "step": 13126 + }, + { + "epoch": 0.2889500330178296, + "grad_norm": 0.6926187715934906, + "learning_rate": 8.34226093575228e-06, + "loss": 0.5208, + "step": 13127 + }, + { + "epoch": 0.28897204490424827, + "grad_norm": 1.576378527893186, + "learning_rate": 8.341995811003097e-06, + "loss": 0.2619, + "step": 13128 + }, + { + "epoch": 0.288994056790667, + "grad_norm": 1.5303393024912442, + "learning_rate": 8.3417306692684e-06, + "loss": 0.3566, + "step": 13129 + }, + { + "epoch": 0.2890160686770856, + "grad_norm": 0.6192880193526203, + "learning_rate": 8.341465510549538e-06, + "loss": 0.5003, + "step": 13130 + }, + { + "epoch": 0.2890380805635043, + "grad_norm": 0.6218427255698941, + "learning_rate": 8.341200334847856e-06, + "loss": 0.5238, + "step": 13131 + }, + { + "epoch": 0.289060092449923, + "grad_norm": 0.6117538950726242, + "learning_rate": 8.340935142164706e-06, + "loss": 0.5061, + "step": 13132 + }, + { + "epoch": 0.2890821043363416, + "grad_norm": 1.8236657461407755, + "learning_rate": 8.340669932501433e-06, + "loss": 0.2959, + "step": 13133 + }, + { + "epoch": 0.2891041162227603, + "grad_norm": 1.538195415255759, + "learning_rate": 8.340404705859385e-06, + "loss": 0.3541, + "step": 13134 + }, + { + "epoch": 0.289126128109179, + "grad_norm": 1.22307049819576, + "learning_rate": 8.340139462239911e-06, + "loss": 0.227, + "step": 13135 + }, + { + "epoch": 0.28914813999559763, + "grad_norm": 1.3066177821783984, + "learning_rate": 8.339874201644359e-06, + "loss": 0.347, + "step": 13136 + }, + { + "epoch": 0.2891701518820163, + "grad_norm": 1.531576669737048, + "learning_rate": 8.339608924074074e-06, + "loss": 0.3718, + "step": 13137 + }, + { + "epoch": 0.28919216376843493, + "grad_norm": 1.784045685323352, + "learning_rate": 8.339343629530408e-06, + "loss": 0.285, + "step": 13138 + }, + { + "epoch": 0.28921417565485363, + "grad_norm": 1.8927469510784132, + "learning_rate": 8.339078318014711e-06, + "loss": 0.3775, + "step": 13139 + }, + { + "epoch": 0.2892361875412723, + "grad_norm": 1.6015891273914888, + "learning_rate": 8.338812989528326e-06, + "loss": 0.3048, + "step": 13140 + }, + { + "epoch": 0.28925819942769093, + "grad_norm": 0.8161604733391609, + "learning_rate": 8.338547644072603e-06, + "loss": 0.503, + "step": 13141 + }, + { + "epoch": 0.28928021131410964, + "grad_norm": 1.8240577299060938, + "learning_rate": 8.338282281648896e-06, + "loss": 0.3138, + "step": 13142 + }, + { + "epoch": 0.2893022232005283, + "grad_norm": 3.4975643097184683, + "learning_rate": 8.338016902258546e-06, + "loss": 0.2685, + "step": 13143 + }, + { + "epoch": 0.28932423508694693, + "grad_norm": 2.3856631415727994, + "learning_rate": 8.337751505902907e-06, + "loss": 0.3576, + "step": 13144 + }, + { + "epoch": 0.28934624697336564, + "grad_norm": 1.2754239108452232, + "learning_rate": 8.337486092583326e-06, + "loss": 0.3783, + "step": 13145 + }, + { + "epoch": 0.2893682588597843, + "grad_norm": 1.1577407518053144, + "learning_rate": 8.337220662301152e-06, + "loss": 0.3803, + "step": 13146 + }, + { + "epoch": 0.28939027074620294, + "grad_norm": 1.9699950760139586, + "learning_rate": 8.336955215057734e-06, + "loss": 0.3702, + "step": 13147 + }, + { + "epoch": 0.28941228263262164, + "grad_norm": 1.2485323660925105, + "learning_rate": 8.33668975085442e-06, + "loss": 0.304, + "step": 13148 + }, + { + "epoch": 0.2894342945190403, + "grad_norm": 1.468241769634749, + "learning_rate": 8.336424269692562e-06, + "loss": 0.315, + "step": 13149 + }, + { + "epoch": 0.28945630640545894, + "grad_norm": 2.4541471010102054, + "learning_rate": 8.336158771573508e-06, + "loss": 0.3066, + "step": 13150 + }, + { + "epoch": 0.2894783182918776, + "grad_norm": 0.6762003238270413, + "learning_rate": 8.335893256498607e-06, + "loss": 0.5028, + "step": 13151 + }, + { + "epoch": 0.2895003301782963, + "grad_norm": 1.3430297585918136, + "learning_rate": 8.335627724469208e-06, + "loss": 0.3131, + "step": 13152 + }, + { + "epoch": 0.28952234206471494, + "grad_norm": 1.107766739823402, + "learning_rate": 8.33536217548666e-06, + "loss": 0.3145, + "step": 13153 + }, + { + "epoch": 0.2895443539511336, + "grad_norm": 1.8280283309125358, + "learning_rate": 8.335096609552317e-06, + "loss": 0.3763, + "step": 13154 + }, + { + "epoch": 0.2895663658375523, + "grad_norm": 1.249746722968987, + "learning_rate": 8.334831026667523e-06, + "loss": 0.2782, + "step": 13155 + }, + { + "epoch": 0.28958837772397095, + "grad_norm": 1.3597547164468076, + "learning_rate": 8.334565426833632e-06, + "loss": 0.3154, + "step": 13156 + }, + { + "epoch": 0.2896103896103896, + "grad_norm": 1.7429986142670642, + "learning_rate": 8.334299810051992e-06, + "loss": 0.3264, + "step": 13157 + }, + { + "epoch": 0.2896324014968083, + "grad_norm": 1.9139862989727532, + "learning_rate": 8.334034176323952e-06, + "loss": 0.4138, + "step": 13158 + }, + { + "epoch": 0.28965441338322695, + "grad_norm": 0.5989418505964766, + "learning_rate": 8.333768525650863e-06, + "loss": 0.4993, + "step": 13159 + }, + { + "epoch": 0.2896764252696456, + "grad_norm": 0.6194771891218517, + "learning_rate": 8.333502858034077e-06, + "loss": 0.5021, + "step": 13160 + }, + { + "epoch": 0.28969843715606425, + "grad_norm": 1.6867731214579018, + "learning_rate": 8.333237173474943e-06, + "loss": 0.2954, + "step": 13161 + }, + { + "epoch": 0.28972044904248295, + "grad_norm": 1.7834916064667108, + "learning_rate": 8.332971471974809e-06, + "loss": 0.3459, + "step": 13162 + }, + { + "epoch": 0.2897424609289016, + "grad_norm": 0.6092767477737018, + "learning_rate": 8.332705753535029e-06, + "loss": 0.4985, + "step": 13163 + }, + { + "epoch": 0.28976447281532025, + "grad_norm": 0.6255559081444261, + "learning_rate": 8.33244001815695e-06, + "loss": 0.5265, + "step": 13164 + }, + { + "epoch": 0.28978648470173896, + "grad_norm": 1.7383905483727702, + "learning_rate": 8.332174265841928e-06, + "loss": 0.3286, + "step": 13165 + }, + { + "epoch": 0.2898084965881576, + "grad_norm": 0.6110510585045204, + "learning_rate": 8.331908496591308e-06, + "loss": 0.5061, + "step": 13166 + }, + { + "epoch": 0.28983050847457625, + "grad_norm": 16.55118387474271, + "learning_rate": 8.331642710406443e-06, + "loss": 0.3049, + "step": 13167 + }, + { + "epoch": 0.28985252036099496, + "grad_norm": 1.2355986376567427, + "learning_rate": 8.331376907288683e-06, + "loss": 0.3995, + "step": 13168 + }, + { + "epoch": 0.2898745322474136, + "grad_norm": 1.701398406729482, + "learning_rate": 8.331111087239382e-06, + "loss": 0.3875, + "step": 13169 + }, + { + "epoch": 0.28989654413383226, + "grad_norm": 1.4640693615903126, + "learning_rate": 8.330845250259887e-06, + "loss": 0.3824, + "step": 13170 + }, + { + "epoch": 0.28991855602025096, + "grad_norm": 1.6233716079459624, + "learning_rate": 8.33057939635155e-06, + "loss": 0.3201, + "step": 13171 + }, + { + "epoch": 0.2899405679066696, + "grad_norm": 1.2619187350434227, + "learning_rate": 8.330313525515725e-06, + "loss": 0.334, + "step": 13172 + }, + { + "epoch": 0.28996257979308826, + "grad_norm": 2.169412413317025, + "learning_rate": 8.330047637753759e-06, + "loss": 0.4062, + "step": 13173 + }, + { + "epoch": 0.2899845916795069, + "grad_norm": 5.9891442649072895, + "learning_rate": 8.329781733067008e-06, + "loss": 0.2702, + "step": 13174 + }, + { + "epoch": 0.2900066035659256, + "grad_norm": 1.2047145773245627, + "learning_rate": 8.329515811456821e-06, + "loss": 0.2735, + "step": 13175 + }, + { + "epoch": 0.29002861545234426, + "grad_norm": 1.6939784783968632, + "learning_rate": 8.32924987292455e-06, + "loss": 0.4033, + "step": 13176 + }, + { + "epoch": 0.2900506273387629, + "grad_norm": 0.72816427259359, + "learning_rate": 8.328983917471544e-06, + "loss": 0.5314, + "step": 13177 + }, + { + "epoch": 0.2900726392251816, + "grad_norm": 5.388360178929375, + "learning_rate": 8.328717945099158e-06, + "loss": 0.2883, + "step": 13178 + }, + { + "epoch": 0.29009465111160027, + "grad_norm": 2.1491892530340193, + "learning_rate": 8.328451955808743e-06, + "loss": 0.3642, + "step": 13179 + }, + { + "epoch": 0.2901166629980189, + "grad_norm": 2.333065273820236, + "learning_rate": 8.32818594960165e-06, + "loss": 0.3463, + "step": 13180 + }, + { + "epoch": 0.2901386748844376, + "grad_norm": 1.4221150003499319, + "learning_rate": 8.327919926479233e-06, + "loss": 0.3135, + "step": 13181 + }, + { + "epoch": 0.29016068677085627, + "grad_norm": 1.7614398535739844, + "learning_rate": 8.327653886442842e-06, + "loss": 0.3724, + "step": 13182 + }, + { + "epoch": 0.2901826986572749, + "grad_norm": 2.6071573144028855, + "learning_rate": 8.32738782949383e-06, + "loss": 0.3366, + "step": 13183 + }, + { + "epoch": 0.29020471054369357, + "grad_norm": 0.6061705984191912, + "learning_rate": 8.327121755633548e-06, + "loss": 0.5109, + "step": 13184 + }, + { + "epoch": 0.29022672243011227, + "grad_norm": 1.1787637232264858, + "learning_rate": 8.326855664863348e-06, + "loss": 0.29, + "step": 13185 + }, + { + "epoch": 0.2902487343165309, + "grad_norm": 1.5150285789957296, + "learning_rate": 8.326589557184587e-06, + "loss": 0.2835, + "step": 13186 + }, + { + "epoch": 0.29027074620294957, + "grad_norm": 2.501147269846244, + "learning_rate": 8.326323432598613e-06, + "loss": 0.3348, + "step": 13187 + }, + { + "epoch": 0.2902927580893683, + "grad_norm": 1.6202853569242186, + "learning_rate": 8.32605729110678e-06, + "loss": 0.3724, + "step": 13188 + }, + { + "epoch": 0.2903147699757869, + "grad_norm": 1.5470068928373963, + "learning_rate": 8.325791132710438e-06, + "loss": 0.2932, + "step": 13189 + }, + { + "epoch": 0.2903367818622056, + "grad_norm": 2.0863781757634983, + "learning_rate": 8.325524957410945e-06, + "loss": 0.3949, + "step": 13190 + }, + { + "epoch": 0.2903587937486243, + "grad_norm": 3.86806598499719, + "learning_rate": 8.325258765209649e-06, + "loss": 0.3541, + "step": 13191 + }, + { + "epoch": 0.2903808056350429, + "grad_norm": 1.5631410175174396, + "learning_rate": 8.324992556107907e-06, + "loss": 0.3894, + "step": 13192 + }, + { + "epoch": 0.2904028175214616, + "grad_norm": 2.313453021447313, + "learning_rate": 8.324726330107069e-06, + "loss": 0.3017, + "step": 13193 + }, + { + "epoch": 0.2904248294078803, + "grad_norm": 1.7979277435422947, + "learning_rate": 8.324460087208488e-06, + "loss": 0.3997, + "step": 13194 + }, + { + "epoch": 0.29044684129429893, + "grad_norm": 0.6212461963201541, + "learning_rate": 8.324193827413518e-06, + "loss": 0.5061, + "step": 13195 + }, + { + "epoch": 0.2904688531807176, + "grad_norm": 0.636611023702669, + "learning_rate": 8.323927550723515e-06, + "loss": 0.5103, + "step": 13196 + }, + { + "epoch": 0.29049086506713623, + "grad_norm": 1.7548102128490959, + "learning_rate": 8.323661257139827e-06, + "loss": 0.254, + "step": 13197 + }, + { + "epoch": 0.29051287695355493, + "grad_norm": 1.3989025540907867, + "learning_rate": 8.32339494666381e-06, + "loss": 0.3499, + "step": 13198 + }, + { + "epoch": 0.2905348888399736, + "grad_norm": 2.464057517586786, + "learning_rate": 8.323128619296818e-06, + "loss": 0.3491, + "step": 13199 + }, + { + "epoch": 0.29055690072639223, + "grad_norm": 2.5203672474332226, + "learning_rate": 8.322862275040204e-06, + "loss": 0.2862, + "step": 13200 + }, + { + "epoch": 0.29057891261281094, + "grad_norm": 1.623129955189401, + "learning_rate": 8.322595913895323e-06, + "loss": 0.377, + "step": 13201 + }, + { + "epoch": 0.2906009244992296, + "grad_norm": 0.6155289816677174, + "learning_rate": 8.322329535863526e-06, + "loss": 0.4886, + "step": 13202 + }, + { + "epoch": 0.29062293638564823, + "grad_norm": 1.4052358458957446, + "learning_rate": 8.32206314094617e-06, + "loss": 0.3843, + "step": 13203 + }, + { + "epoch": 0.29064494827206694, + "grad_norm": 1.165551669887036, + "learning_rate": 8.321796729144607e-06, + "loss": 0.3408, + "step": 13204 + }, + { + "epoch": 0.2906669601584856, + "grad_norm": 1.3723379806081468, + "learning_rate": 8.321530300460191e-06, + "loss": 0.3272, + "step": 13205 + }, + { + "epoch": 0.29068897204490424, + "grad_norm": 1.635137695070417, + "learning_rate": 8.321263854894279e-06, + "loss": 0.416, + "step": 13206 + }, + { + "epoch": 0.29071098393132294, + "grad_norm": 3.774929416693362, + "learning_rate": 8.32099739244822e-06, + "loss": 0.3597, + "step": 13207 + }, + { + "epoch": 0.2907329958177416, + "grad_norm": 2.034714997038035, + "learning_rate": 8.32073091312337e-06, + "loss": 0.4461, + "step": 13208 + }, + { + "epoch": 0.29075500770416024, + "grad_norm": 1.4336985336669885, + "learning_rate": 8.320464416921088e-06, + "loss": 0.3426, + "step": 13209 + }, + { + "epoch": 0.2907770195905789, + "grad_norm": 0.6526435435157633, + "learning_rate": 8.320197903842722e-06, + "loss": 0.5251, + "step": 13210 + }, + { + "epoch": 0.2907990314769976, + "grad_norm": 0.6400316280228198, + "learning_rate": 8.319931373889631e-06, + "loss": 0.5607, + "step": 13211 + }, + { + "epoch": 0.29082104336341624, + "grad_norm": 1.4539302034439847, + "learning_rate": 8.319664827063167e-06, + "loss": 0.3052, + "step": 13212 + }, + { + "epoch": 0.2908430552498349, + "grad_norm": 1.650446252512127, + "learning_rate": 8.319398263364686e-06, + "loss": 0.3067, + "step": 13213 + }, + { + "epoch": 0.2908650671362536, + "grad_norm": 1.3430207975957542, + "learning_rate": 8.319131682795544e-06, + "loss": 0.3482, + "step": 13214 + }, + { + "epoch": 0.29088707902267225, + "grad_norm": 6.419924247965126, + "learning_rate": 8.318865085357093e-06, + "loss": 0.3861, + "step": 13215 + }, + { + "epoch": 0.2909090909090909, + "grad_norm": 1.334788632471925, + "learning_rate": 8.31859847105069e-06, + "loss": 0.2576, + "step": 13216 + }, + { + "epoch": 0.2909311027955096, + "grad_norm": 0.6510004405448775, + "learning_rate": 8.318331839877688e-06, + "loss": 0.5511, + "step": 13217 + }, + { + "epoch": 0.29095311468192825, + "grad_norm": 1.147216701650633, + "learning_rate": 8.318065191839446e-06, + "loss": 0.3449, + "step": 13218 + }, + { + "epoch": 0.2909751265683469, + "grad_norm": 1.3403515102081445, + "learning_rate": 8.317798526937316e-06, + "loss": 0.2767, + "step": 13219 + }, + { + "epoch": 0.29099713845476555, + "grad_norm": 0.6217264414640254, + "learning_rate": 8.317531845172652e-06, + "loss": 0.5265, + "step": 13220 + }, + { + "epoch": 0.29101915034118425, + "grad_norm": 1.7981498153566748, + "learning_rate": 8.317265146546814e-06, + "loss": 0.406, + "step": 13221 + }, + { + "epoch": 0.2910411622276029, + "grad_norm": 1.344471083493654, + "learning_rate": 8.316998431061153e-06, + "loss": 0.3065, + "step": 13222 + }, + { + "epoch": 0.29106317411402155, + "grad_norm": 2.411831986905808, + "learning_rate": 8.316731698717027e-06, + "loss": 0.3125, + "step": 13223 + }, + { + "epoch": 0.29108518600044025, + "grad_norm": 1.3620365611997487, + "learning_rate": 8.316464949515793e-06, + "loss": 0.3615, + "step": 13224 + }, + { + "epoch": 0.2911071978868589, + "grad_norm": 1.5702977542279752, + "learning_rate": 8.316198183458803e-06, + "loss": 0.3333, + "step": 13225 + }, + { + "epoch": 0.29112920977327755, + "grad_norm": 1.3908486431172802, + "learning_rate": 8.315931400547414e-06, + "loss": 0.4026, + "step": 13226 + }, + { + "epoch": 0.29115122165969626, + "grad_norm": 1.750774855709437, + "learning_rate": 8.315664600782986e-06, + "loss": 0.395, + "step": 13227 + }, + { + "epoch": 0.2911732335461149, + "grad_norm": 0.633402976141863, + "learning_rate": 8.31539778416687e-06, + "loss": 0.5279, + "step": 13228 + }, + { + "epoch": 0.29119524543253356, + "grad_norm": 0.6419625553837516, + "learning_rate": 8.315130950700423e-06, + "loss": 0.5346, + "step": 13229 + }, + { + "epoch": 0.29121725731895226, + "grad_norm": 1.355150759415752, + "learning_rate": 8.314864100385002e-06, + "loss": 0.294, + "step": 13230 + }, + { + "epoch": 0.2912392692053709, + "grad_norm": 2.1963019731995788, + "learning_rate": 8.314597233221964e-06, + "loss": 0.3546, + "step": 13231 + }, + { + "epoch": 0.29126128109178956, + "grad_norm": 1.608393944372832, + "learning_rate": 8.314330349212665e-06, + "loss": 0.3135, + "step": 13232 + }, + { + "epoch": 0.2912832929782082, + "grad_norm": 0.665979727242875, + "learning_rate": 8.314063448358459e-06, + "loss": 0.5676, + "step": 13233 + }, + { + "epoch": 0.2913053048646269, + "grad_norm": 1.348859966017169, + "learning_rate": 8.313796530660706e-06, + "loss": 0.246, + "step": 13234 + }, + { + "epoch": 0.29132731675104556, + "grad_norm": 1.579930485549974, + "learning_rate": 8.31352959612076e-06, + "loss": 0.3556, + "step": 13235 + }, + { + "epoch": 0.2913493286374642, + "grad_norm": 1.103302752881034, + "learning_rate": 8.313262644739978e-06, + "loss": 0.3295, + "step": 13236 + }, + { + "epoch": 0.2913713405238829, + "grad_norm": 0.6224934855392874, + "learning_rate": 8.312995676519719e-06, + "loss": 0.4965, + "step": 13237 + }, + { + "epoch": 0.29139335241030156, + "grad_norm": 1.4613270386678157, + "learning_rate": 8.31272869146134e-06, + "loss": 0.2541, + "step": 13238 + }, + { + "epoch": 0.2914153642967202, + "grad_norm": 0.6120978936757068, + "learning_rate": 8.312461689566194e-06, + "loss": 0.5237, + "step": 13239 + }, + { + "epoch": 0.2914373761831389, + "grad_norm": 1.1421620755376523, + "learning_rate": 8.31219467083564e-06, + "loss": 0.3886, + "step": 13240 + }, + { + "epoch": 0.29145938806955757, + "grad_norm": 0.6198657403428671, + "learning_rate": 8.311927635271036e-06, + "loss": 0.4958, + "step": 13241 + }, + { + "epoch": 0.2914813999559762, + "grad_norm": 4.039244290422335, + "learning_rate": 8.311660582873738e-06, + "loss": 0.2606, + "step": 13242 + }, + { + "epoch": 0.29150341184239487, + "grad_norm": 1.4892172617974169, + "learning_rate": 8.311393513645105e-06, + "loss": 0.4002, + "step": 13243 + }, + { + "epoch": 0.29152542372881357, + "grad_norm": 0.6300399336448831, + "learning_rate": 8.311126427586493e-06, + "loss": 0.5309, + "step": 13244 + }, + { + "epoch": 0.2915474356152322, + "grad_norm": 1.2785413103142433, + "learning_rate": 8.31085932469926e-06, + "loss": 0.2285, + "step": 13245 + }, + { + "epoch": 0.29156944750165087, + "grad_norm": 1.4818097424869312, + "learning_rate": 8.310592204984763e-06, + "loss": 0.2702, + "step": 13246 + }, + { + "epoch": 0.2915914593880696, + "grad_norm": 0.661823550629487, + "learning_rate": 8.31032506844436e-06, + "loss": 0.5393, + "step": 13247 + }, + { + "epoch": 0.2916134712744882, + "grad_norm": 0.6141649635760886, + "learning_rate": 8.310057915079406e-06, + "loss": 0.5124, + "step": 13248 + }, + { + "epoch": 0.29163548316090687, + "grad_norm": 1.505124115310741, + "learning_rate": 8.309790744891264e-06, + "loss": 0.3345, + "step": 13249 + }, + { + "epoch": 0.2916574950473256, + "grad_norm": 1.8302171895610644, + "learning_rate": 8.30952355788129e-06, + "loss": 0.2469, + "step": 13250 + }, + { + "epoch": 0.2916795069337442, + "grad_norm": 1.658837759556784, + "learning_rate": 8.309256354050839e-06, + "loss": 0.3937, + "step": 13251 + }, + { + "epoch": 0.2917015188201629, + "grad_norm": 1.185310836840129, + "learning_rate": 8.308989133401273e-06, + "loss": 0.3327, + "step": 13252 + }, + { + "epoch": 0.2917235307065816, + "grad_norm": 1.6140801265664062, + "learning_rate": 8.308721895933946e-06, + "loss": 0.2955, + "step": 13253 + }, + { + "epoch": 0.29174554259300023, + "grad_norm": 1.4965451314483977, + "learning_rate": 8.308454641650222e-06, + "loss": 0.3446, + "step": 13254 + }, + { + "epoch": 0.2917675544794189, + "grad_norm": 0.7159336359580095, + "learning_rate": 8.308187370551453e-06, + "loss": 0.5269, + "step": 13255 + }, + { + "epoch": 0.2917895663658375, + "grad_norm": 2.004921190216992, + "learning_rate": 8.307920082639001e-06, + "loss": 0.4052, + "step": 13256 + }, + { + "epoch": 0.29181157825225623, + "grad_norm": 0.6932812816901899, + "learning_rate": 8.307652777914224e-06, + "loss": 0.5057, + "step": 13257 + }, + { + "epoch": 0.2918335901386749, + "grad_norm": 2.9821343444247033, + "learning_rate": 8.30738545637848e-06, + "loss": 0.3689, + "step": 13258 + }, + { + "epoch": 0.29185560202509353, + "grad_norm": 1.5188629844842394, + "learning_rate": 8.307118118033128e-06, + "loss": 0.3153, + "step": 13259 + }, + { + "epoch": 0.29187761391151223, + "grad_norm": 1.6622405042081596, + "learning_rate": 8.306850762879528e-06, + "loss": 0.3988, + "step": 13260 + }, + { + "epoch": 0.2918996257979309, + "grad_norm": 3.416687778588936, + "learning_rate": 8.306583390919037e-06, + "loss": 0.3584, + "step": 13261 + }, + { + "epoch": 0.29192163768434953, + "grad_norm": 0.6706517587391183, + "learning_rate": 8.306316002153013e-06, + "loss": 0.4973, + "step": 13262 + }, + { + "epoch": 0.29194364957076824, + "grad_norm": 1.80188828312245, + "learning_rate": 8.306048596582817e-06, + "loss": 0.3573, + "step": 13263 + }, + { + "epoch": 0.2919656614571869, + "grad_norm": 1.3766464900659106, + "learning_rate": 8.305781174209809e-06, + "loss": 0.403, + "step": 13264 + }, + { + "epoch": 0.29198767334360554, + "grad_norm": 1.8897029875356623, + "learning_rate": 8.305513735035345e-06, + "loss": 0.318, + "step": 13265 + }, + { + "epoch": 0.29200968523002424, + "grad_norm": 1.3066475122403292, + "learning_rate": 8.305246279060785e-06, + "loss": 0.2633, + "step": 13266 + }, + { + "epoch": 0.2920316971164429, + "grad_norm": 2.052730468679266, + "learning_rate": 8.304978806287492e-06, + "loss": 0.3212, + "step": 13267 + }, + { + "epoch": 0.29205370900286154, + "grad_norm": 1.392360823993167, + "learning_rate": 8.304711316716822e-06, + "loss": 0.3585, + "step": 13268 + }, + { + "epoch": 0.2920757208892802, + "grad_norm": 1.702265590603298, + "learning_rate": 8.304443810350133e-06, + "loss": 0.3283, + "step": 13269 + }, + { + "epoch": 0.2920977327756989, + "grad_norm": 0.6211567114727334, + "learning_rate": 8.30417628718879e-06, + "loss": 0.5275, + "step": 13270 + }, + { + "epoch": 0.29211974466211754, + "grad_norm": 1.5217728160221269, + "learning_rate": 8.303908747234147e-06, + "loss": 0.2962, + "step": 13271 + }, + { + "epoch": 0.2921417565485362, + "grad_norm": 2.059078503862934, + "learning_rate": 8.303641190487566e-06, + "loss": 0.2792, + "step": 13272 + }, + { + "epoch": 0.2921637684349549, + "grad_norm": 1.345663441658966, + "learning_rate": 8.303373616950408e-06, + "loss": 0.3101, + "step": 13273 + }, + { + "epoch": 0.29218578032137354, + "grad_norm": 0.6406851680231052, + "learning_rate": 8.303106026624031e-06, + "loss": 0.5164, + "step": 13274 + }, + { + "epoch": 0.2922077922077922, + "grad_norm": 1.7870888148532165, + "learning_rate": 8.302838419509798e-06, + "loss": 0.3972, + "step": 13275 + }, + { + "epoch": 0.2922298040942109, + "grad_norm": 0.6317804813798411, + "learning_rate": 8.302570795609066e-06, + "loss": 0.5138, + "step": 13276 + }, + { + "epoch": 0.29225181598062955, + "grad_norm": 1.4701251197872065, + "learning_rate": 8.302303154923195e-06, + "loss": 0.2508, + "step": 13277 + }, + { + "epoch": 0.2922738278670482, + "grad_norm": 2.434621572478423, + "learning_rate": 8.302035497453546e-06, + "loss": 0.3744, + "step": 13278 + }, + { + "epoch": 0.29229583975346685, + "grad_norm": 1.5490483372749013, + "learning_rate": 8.301767823201483e-06, + "loss": 0.2898, + "step": 13279 + }, + { + "epoch": 0.29231785163988555, + "grad_norm": 1.2950890838379503, + "learning_rate": 8.301500132168361e-06, + "loss": 0.3438, + "step": 13280 + }, + { + "epoch": 0.2923398635263042, + "grad_norm": 2.7808234923159856, + "learning_rate": 8.301232424355543e-06, + "loss": 0.3654, + "step": 13281 + }, + { + "epoch": 0.29236187541272285, + "grad_norm": 1.4228765885215344, + "learning_rate": 8.300964699764391e-06, + "loss": 0.3488, + "step": 13282 + }, + { + "epoch": 0.29238388729914155, + "grad_norm": 1.2826218370262443, + "learning_rate": 8.30069695839626e-06, + "loss": 0.3113, + "step": 13283 + }, + { + "epoch": 0.2924058991855602, + "grad_norm": 2.5462849730055366, + "learning_rate": 8.30042920025252e-06, + "loss": 0.3009, + "step": 13284 + }, + { + "epoch": 0.29242791107197885, + "grad_norm": 1.4723370320376863, + "learning_rate": 8.300161425334525e-06, + "loss": 0.3733, + "step": 13285 + }, + { + "epoch": 0.29244992295839756, + "grad_norm": 1.4247098377334, + "learning_rate": 8.299893633643636e-06, + "loss": 0.2904, + "step": 13286 + }, + { + "epoch": 0.2924719348448162, + "grad_norm": 0.6761678756569806, + "learning_rate": 8.299625825181218e-06, + "loss": 0.5139, + "step": 13287 + }, + { + "epoch": 0.29249394673123486, + "grad_norm": 1.5298218050691057, + "learning_rate": 8.29935799994863e-06, + "loss": 0.383, + "step": 13288 + }, + { + "epoch": 0.29251595861765356, + "grad_norm": 0.6390371119489917, + "learning_rate": 8.299090157947232e-06, + "loss": 0.4843, + "step": 13289 + }, + { + "epoch": 0.2925379705040722, + "grad_norm": 2.1619939566346877, + "learning_rate": 8.298822299178386e-06, + "loss": 0.3914, + "step": 13290 + }, + { + "epoch": 0.29255998239049086, + "grad_norm": 1.5736843770599065, + "learning_rate": 8.298554423643455e-06, + "loss": 0.3726, + "step": 13291 + }, + { + "epoch": 0.2925819942769095, + "grad_norm": 1.3738491259085068, + "learning_rate": 8.298286531343798e-06, + "loss": 0.3263, + "step": 13292 + }, + { + "epoch": 0.2926040061633282, + "grad_norm": 0.6532649298033338, + "learning_rate": 8.298018622280778e-06, + "loss": 0.5331, + "step": 13293 + }, + { + "epoch": 0.29262601804974686, + "grad_norm": 0.6256915428996445, + "learning_rate": 8.297750696455757e-06, + "loss": 0.5141, + "step": 13294 + }, + { + "epoch": 0.2926480299361655, + "grad_norm": 1.3210227423312577, + "learning_rate": 8.297482753870094e-06, + "loss": 0.3375, + "step": 13295 + }, + { + "epoch": 0.2926700418225842, + "grad_norm": 2.196232760322461, + "learning_rate": 8.297214794525155e-06, + "loss": 0.2817, + "step": 13296 + }, + { + "epoch": 0.29269205370900286, + "grad_norm": 1.2008806719804686, + "learning_rate": 8.2969468184223e-06, + "loss": 0.3269, + "step": 13297 + }, + { + "epoch": 0.2927140655954215, + "grad_norm": 1.5503306679254234, + "learning_rate": 8.296678825562891e-06, + "loss": 0.3718, + "step": 13298 + }, + { + "epoch": 0.2927360774818402, + "grad_norm": 1.686940393614224, + "learning_rate": 8.296410815948288e-06, + "loss": 0.4585, + "step": 13299 + }, + { + "epoch": 0.29275808936825887, + "grad_norm": 4.649273323667639, + "learning_rate": 8.296142789579856e-06, + "loss": 0.317, + "step": 13300 + }, + { + "epoch": 0.2927801012546775, + "grad_norm": 1.5031716200047027, + "learning_rate": 8.295874746458956e-06, + "loss": 0.4544, + "step": 13301 + }, + { + "epoch": 0.29280211314109617, + "grad_norm": 1.735975936966261, + "learning_rate": 8.295606686586952e-06, + "loss": 0.4255, + "step": 13302 + }, + { + "epoch": 0.29282412502751487, + "grad_norm": 1.5876664050295004, + "learning_rate": 8.295338609965204e-06, + "loss": 0.3295, + "step": 13303 + }, + { + "epoch": 0.2928461369139335, + "grad_norm": 1.4444104733417318, + "learning_rate": 8.295070516595073e-06, + "loss": 0.3747, + "step": 13304 + }, + { + "epoch": 0.29286814880035217, + "grad_norm": 1.304771540277803, + "learning_rate": 8.294802406477928e-06, + "loss": 0.3817, + "step": 13305 + }, + { + "epoch": 0.2928901606867709, + "grad_norm": 1.3893500210550067, + "learning_rate": 8.294534279615127e-06, + "loss": 0.323, + "step": 13306 + }, + { + "epoch": 0.2929121725731895, + "grad_norm": 0.7073288098064948, + "learning_rate": 8.294266136008031e-06, + "loss": 0.5737, + "step": 13307 + }, + { + "epoch": 0.29293418445960817, + "grad_norm": 0.6436456678425527, + "learning_rate": 8.293997975658006e-06, + "loss": 0.5034, + "step": 13308 + }, + { + "epoch": 0.2929561963460269, + "grad_norm": 1.2355910251358824, + "learning_rate": 8.293729798566414e-06, + "loss": 0.2961, + "step": 13309 + }, + { + "epoch": 0.2929782082324455, + "grad_norm": 1.2925623046169237, + "learning_rate": 8.29346160473462e-06, + "loss": 0.2882, + "step": 13310 + }, + { + "epoch": 0.2930002201188642, + "grad_norm": 1.348870424045961, + "learning_rate": 8.293193394163985e-06, + "loss": 0.2817, + "step": 13311 + }, + { + "epoch": 0.2930222320052829, + "grad_norm": 1.6248153084790724, + "learning_rate": 8.29292516685587e-06, + "loss": 0.2918, + "step": 13312 + }, + { + "epoch": 0.29304424389170153, + "grad_norm": 1.5384941559866254, + "learning_rate": 8.292656922811641e-06, + "loss": 0.4756, + "step": 13313 + }, + { + "epoch": 0.2930662557781202, + "grad_norm": 1.3772367015614875, + "learning_rate": 8.292388662032662e-06, + "loss": 0.3417, + "step": 13314 + }, + { + "epoch": 0.2930882676645388, + "grad_norm": 1.4084829715564064, + "learning_rate": 8.292120384520298e-06, + "loss": 0.3476, + "step": 13315 + }, + { + "epoch": 0.29311027955095753, + "grad_norm": 0.6767680351211911, + "learning_rate": 8.291852090275906e-06, + "loss": 0.5042, + "step": 13316 + }, + { + "epoch": 0.2931322914373762, + "grad_norm": 1.1346051699267874, + "learning_rate": 8.291583779300854e-06, + "loss": 0.2777, + "step": 13317 + }, + { + "epoch": 0.29315430332379483, + "grad_norm": 1.4900554457525041, + "learning_rate": 8.291315451596506e-06, + "loss": 0.3236, + "step": 13318 + }, + { + "epoch": 0.29317631521021353, + "grad_norm": 0.6636634394802017, + "learning_rate": 8.291047107164223e-06, + "loss": 0.5047, + "step": 13319 + }, + { + "epoch": 0.2931983270966322, + "grad_norm": 1.6547313793644853, + "learning_rate": 8.290778746005373e-06, + "loss": 0.3299, + "step": 13320 + }, + { + "epoch": 0.29322033898305083, + "grad_norm": 1.1554828514417084, + "learning_rate": 8.290510368121317e-06, + "loss": 0.2911, + "step": 13321 + }, + { + "epoch": 0.29324235086946954, + "grad_norm": 0.6212667346175708, + "learning_rate": 8.29024197351342e-06, + "loss": 0.5256, + "step": 13322 + }, + { + "epoch": 0.2932643627558882, + "grad_norm": 1.1697035048370676, + "learning_rate": 8.289973562183046e-06, + "loss": 0.3394, + "step": 13323 + }, + { + "epoch": 0.29328637464230684, + "grad_norm": 1.3764421042095245, + "learning_rate": 8.289705134131558e-06, + "loss": 0.3235, + "step": 13324 + }, + { + "epoch": 0.2933083865287255, + "grad_norm": 1.6029953883909596, + "learning_rate": 8.289436689360321e-06, + "loss": 0.3674, + "step": 13325 + }, + { + "epoch": 0.2933303984151442, + "grad_norm": 1.1519118113783486, + "learning_rate": 8.289168227870702e-06, + "loss": 0.2839, + "step": 13326 + }, + { + "epoch": 0.29335241030156284, + "grad_norm": 1.3603159699321237, + "learning_rate": 8.28889974966406e-06, + "loss": 0.3586, + "step": 13327 + }, + { + "epoch": 0.2933744221879815, + "grad_norm": 1.3963344243070777, + "learning_rate": 8.288631254741764e-06, + "loss": 0.3582, + "step": 13328 + }, + { + "epoch": 0.2933964340744002, + "grad_norm": 1.7669205456785093, + "learning_rate": 8.288362743105179e-06, + "loss": 0.3606, + "step": 13329 + }, + { + "epoch": 0.29341844596081884, + "grad_norm": 1.8939743827565312, + "learning_rate": 8.288094214755666e-06, + "loss": 0.3865, + "step": 13330 + }, + { + "epoch": 0.2934404578472375, + "grad_norm": 0.7433038760720854, + "learning_rate": 8.287825669694591e-06, + "loss": 0.5059, + "step": 13331 + }, + { + "epoch": 0.2934624697336562, + "grad_norm": 0.6962064579941447, + "learning_rate": 8.287557107923321e-06, + "loss": 0.5161, + "step": 13332 + }, + { + "epoch": 0.29348448162007484, + "grad_norm": 1.5404231072231402, + "learning_rate": 8.287288529443219e-06, + "loss": 0.3114, + "step": 13333 + }, + { + "epoch": 0.2935064935064935, + "grad_norm": 0.6489824397077419, + "learning_rate": 8.287019934255649e-06, + "loss": 0.5241, + "step": 13334 + }, + { + "epoch": 0.2935285053929122, + "grad_norm": 1.9852663161780213, + "learning_rate": 8.28675132236198e-06, + "loss": 0.3669, + "step": 13335 + }, + { + "epoch": 0.29355051727933085, + "grad_norm": 1.747617385427233, + "learning_rate": 8.286482693763574e-06, + "loss": 0.2781, + "step": 13336 + }, + { + "epoch": 0.2935725291657495, + "grad_norm": 1.4697621190256922, + "learning_rate": 8.286214048461796e-06, + "loss": 0.3883, + "step": 13337 + }, + { + "epoch": 0.29359454105216815, + "grad_norm": 1.1934695432277242, + "learning_rate": 8.285945386458014e-06, + "loss": 0.2803, + "step": 13338 + }, + { + "epoch": 0.29361655293858685, + "grad_norm": 1.9238386030023167, + "learning_rate": 8.285676707753592e-06, + "loss": 0.3101, + "step": 13339 + }, + { + "epoch": 0.2936385648250055, + "grad_norm": 0.8162385227520911, + "learning_rate": 8.285408012349895e-06, + "loss": 0.5119, + "step": 13340 + }, + { + "epoch": 0.29366057671142415, + "grad_norm": 2.6122287098178734, + "learning_rate": 8.285139300248289e-06, + "loss": 0.3916, + "step": 13341 + }, + { + "epoch": 0.29368258859784285, + "grad_norm": 1.5880505355151615, + "learning_rate": 8.284870571450138e-06, + "loss": 0.3716, + "step": 13342 + }, + { + "epoch": 0.2937046004842615, + "grad_norm": 1.453615160063748, + "learning_rate": 8.284601825956812e-06, + "loss": 0.4042, + "step": 13343 + }, + { + "epoch": 0.29372661237068015, + "grad_norm": 6.675200937997858, + "learning_rate": 8.284333063769677e-06, + "loss": 0.3104, + "step": 13344 + }, + { + "epoch": 0.29374862425709886, + "grad_norm": 1.7203783655602596, + "learning_rate": 8.284064284890091e-06, + "loss": 0.3378, + "step": 13345 + }, + { + "epoch": 0.2937706361435175, + "grad_norm": 1.3217858798491735, + "learning_rate": 8.283795489319429e-06, + "loss": 0.296, + "step": 13346 + }, + { + "epoch": 0.29379264802993615, + "grad_norm": 1.540470845922288, + "learning_rate": 8.283526677059053e-06, + "loss": 0.3955, + "step": 13347 + }, + { + "epoch": 0.29381465991635486, + "grad_norm": 2.5113382253084673, + "learning_rate": 8.28325784811033e-06, + "loss": 0.4116, + "step": 13348 + }, + { + "epoch": 0.2938366718027735, + "grad_norm": 1.7167910728698588, + "learning_rate": 8.282989002474625e-06, + "loss": 0.332, + "step": 13349 + }, + { + "epoch": 0.29385868368919216, + "grad_norm": 1.494427282588557, + "learning_rate": 8.282720140153308e-06, + "loss": 0.2713, + "step": 13350 + }, + { + "epoch": 0.2938806955756108, + "grad_norm": 1.384368460825863, + "learning_rate": 8.282451261147742e-06, + "loss": 0.3493, + "step": 13351 + }, + { + "epoch": 0.2939027074620295, + "grad_norm": 1.3760131232058028, + "learning_rate": 8.282182365459295e-06, + "loss": 0.348, + "step": 13352 + }, + { + "epoch": 0.29392471934844816, + "grad_norm": 2.054402398233729, + "learning_rate": 8.281913453089333e-06, + "loss": 0.3791, + "step": 13353 + }, + { + "epoch": 0.2939467312348668, + "grad_norm": 1.7188314907942257, + "learning_rate": 8.281644524039222e-06, + "loss": 0.3394, + "step": 13354 + }, + { + "epoch": 0.2939687431212855, + "grad_norm": 1.442155008367115, + "learning_rate": 8.281375578310332e-06, + "loss": 0.3273, + "step": 13355 + }, + { + "epoch": 0.29399075500770416, + "grad_norm": 1.5009989165241666, + "learning_rate": 8.281106615904027e-06, + "loss": 0.3061, + "step": 13356 + }, + { + "epoch": 0.2940127668941228, + "grad_norm": 1.5158385544456046, + "learning_rate": 8.280837636821674e-06, + "loss": 0.3093, + "step": 13357 + }, + { + "epoch": 0.2940347787805415, + "grad_norm": 1.3847632904193397, + "learning_rate": 8.28056864106464e-06, + "loss": 0.3158, + "step": 13358 + }, + { + "epoch": 0.29405679066696017, + "grad_norm": 1.5970555349292495, + "learning_rate": 8.280299628634296e-06, + "loss": 0.362, + "step": 13359 + }, + { + "epoch": 0.2940788025533788, + "grad_norm": 1.3814195488517271, + "learning_rate": 8.280030599532003e-06, + "loss": 0.3381, + "step": 13360 + }, + { + "epoch": 0.29410081443979746, + "grad_norm": 1.508991025824131, + "learning_rate": 8.279761553759134e-06, + "loss": 0.3717, + "step": 13361 + }, + { + "epoch": 0.29412282632621617, + "grad_norm": 1.2690532691106808, + "learning_rate": 8.279492491317052e-06, + "loss": 0.3868, + "step": 13362 + }, + { + "epoch": 0.2941448382126348, + "grad_norm": 1.181502417606005, + "learning_rate": 8.279223412207128e-06, + "loss": 0.2991, + "step": 13363 + }, + { + "epoch": 0.29416685009905347, + "grad_norm": 1.398236812585427, + "learning_rate": 8.278954316430727e-06, + "loss": 0.3297, + "step": 13364 + }, + { + "epoch": 0.29418886198547217, + "grad_norm": 1.5954741464836986, + "learning_rate": 8.278685203989218e-06, + "loss": 0.3017, + "step": 13365 + }, + { + "epoch": 0.2942108738718908, + "grad_norm": 3.796351218425517, + "learning_rate": 8.27841607488397e-06, + "loss": 0.3439, + "step": 13366 + }, + { + "epoch": 0.29423288575830947, + "grad_norm": 1.9731066412777294, + "learning_rate": 8.278146929116347e-06, + "loss": 0.3318, + "step": 13367 + }, + { + "epoch": 0.2942548976447282, + "grad_norm": 1.4518125765230956, + "learning_rate": 8.277877766687721e-06, + "loss": 0.3976, + "step": 13368 + }, + { + "epoch": 0.2942769095311468, + "grad_norm": 1.8412606244236198, + "learning_rate": 8.277608587599458e-06, + "loss": 0.4255, + "step": 13369 + }, + { + "epoch": 0.2942989214175655, + "grad_norm": 1.4036774852713376, + "learning_rate": 8.277339391852925e-06, + "loss": 0.3494, + "step": 13370 + }, + { + "epoch": 0.2943209333039842, + "grad_norm": 1.445493858020333, + "learning_rate": 8.277070179449492e-06, + "loss": 0.3846, + "step": 13371 + }, + { + "epoch": 0.2943429451904028, + "grad_norm": 0.7223003858082367, + "learning_rate": 8.276800950390527e-06, + "loss": 0.5229, + "step": 13372 + }, + { + "epoch": 0.2943649570768215, + "grad_norm": 1.4107373161378944, + "learning_rate": 8.276531704677398e-06, + "loss": 0.4032, + "step": 13373 + }, + { + "epoch": 0.2943869689632401, + "grad_norm": 1.3377394064471524, + "learning_rate": 8.276262442311474e-06, + "loss": 0.2709, + "step": 13374 + }, + { + "epoch": 0.29440898084965883, + "grad_norm": 1.623518542578191, + "learning_rate": 8.275993163294121e-06, + "loss": 0.4008, + "step": 13375 + }, + { + "epoch": 0.2944309927360775, + "grad_norm": 0.6293288126826642, + "learning_rate": 8.275723867626713e-06, + "loss": 0.4999, + "step": 13376 + }, + { + "epoch": 0.29445300462249613, + "grad_norm": 1.3158460169809483, + "learning_rate": 8.27545455531061e-06, + "loss": 0.3333, + "step": 13377 + }, + { + "epoch": 0.29447501650891483, + "grad_norm": 3.4065017451181787, + "learning_rate": 8.27518522634719e-06, + "loss": 0.295, + "step": 13378 + }, + { + "epoch": 0.2944970283953335, + "grad_norm": 1.3772914737864588, + "learning_rate": 8.27491588073782e-06, + "loss": 0.4189, + "step": 13379 + }, + { + "epoch": 0.29451904028175213, + "grad_norm": 1.6725613671390838, + "learning_rate": 8.274646518483864e-06, + "loss": 0.3667, + "step": 13380 + }, + { + "epoch": 0.29454105216817084, + "grad_norm": 1.6685202381699142, + "learning_rate": 8.274377139586693e-06, + "loss": 0.4066, + "step": 13381 + }, + { + "epoch": 0.2945630640545895, + "grad_norm": 1.1550904222137526, + "learning_rate": 8.274107744047678e-06, + "loss": 0.2986, + "step": 13382 + }, + { + "epoch": 0.29458507594100813, + "grad_norm": 0.6534202276614589, + "learning_rate": 8.273838331868187e-06, + "loss": 0.5068, + "step": 13383 + }, + { + "epoch": 0.2946070878274268, + "grad_norm": 1.0954411506055846, + "learning_rate": 8.27356890304959e-06, + "loss": 0.2563, + "step": 13384 + }, + { + "epoch": 0.2946290997138455, + "grad_norm": 1.4560831028461194, + "learning_rate": 8.273299457593256e-06, + "loss": 0.3955, + "step": 13385 + }, + { + "epoch": 0.29465111160026414, + "grad_norm": 0.6071297428910741, + "learning_rate": 8.273029995500551e-06, + "loss": 0.4912, + "step": 13386 + }, + { + "epoch": 0.2946731234866828, + "grad_norm": 1.1071821986905377, + "learning_rate": 8.272760516772852e-06, + "loss": 0.2646, + "step": 13387 + }, + { + "epoch": 0.2946951353731015, + "grad_norm": 0.6100574480564666, + "learning_rate": 8.272491021411524e-06, + "loss": 0.4919, + "step": 13388 + }, + { + "epoch": 0.29471714725952014, + "grad_norm": 1.4902750081432905, + "learning_rate": 8.272221509417935e-06, + "loss": 0.4207, + "step": 13389 + }, + { + "epoch": 0.2947391591459388, + "grad_norm": 0.6134986064922127, + "learning_rate": 8.271951980793459e-06, + "loss": 0.5103, + "step": 13390 + }, + { + "epoch": 0.2947611710323575, + "grad_norm": 1.3513799639431734, + "learning_rate": 8.27168243553946e-06, + "loss": 0.3173, + "step": 13391 + }, + { + "epoch": 0.29478318291877614, + "grad_norm": 2.0534930008256094, + "learning_rate": 8.271412873657313e-06, + "loss": 0.4258, + "step": 13392 + }, + { + "epoch": 0.2948051948051948, + "grad_norm": 1.58015313456421, + "learning_rate": 8.271143295148389e-06, + "loss": 0.3988, + "step": 13393 + }, + { + "epoch": 0.2948272066916135, + "grad_norm": 1.4236822462261753, + "learning_rate": 8.270873700014055e-06, + "loss": 0.4424, + "step": 13394 + }, + { + "epoch": 0.29484921857803215, + "grad_norm": 0.63962191028814, + "learning_rate": 8.27060408825568e-06, + "loss": 0.5138, + "step": 13395 + }, + { + "epoch": 0.2948712304644508, + "grad_norm": 2.839443699593596, + "learning_rate": 8.270334459874639e-06, + "loss": 0.3317, + "step": 13396 + }, + { + "epoch": 0.29489324235086944, + "grad_norm": 1.397695923833855, + "learning_rate": 8.270064814872298e-06, + "loss": 0.3039, + "step": 13397 + }, + { + "epoch": 0.29491525423728815, + "grad_norm": 1.2497624988562215, + "learning_rate": 8.26979515325003e-06, + "loss": 0.3488, + "step": 13398 + }, + { + "epoch": 0.2949372661237068, + "grad_norm": 1.5559034133967298, + "learning_rate": 8.269525475009203e-06, + "loss": 0.4045, + "step": 13399 + }, + { + "epoch": 0.29495927801012545, + "grad_norm": 1.7339570724236586, + "learning_rate": 8.269255780151191e-06, + "loss": 0.2465, + "step": 13400 + }, + { + "epoch": 0.29498128989654415, + "grad_norm": 1.3208190233685595, + "learning_rate": 8.268986068677362e-06, + "loss": 0.3696, + "step": 13401 + }, + { + "epoch": 0.2950033017829628, + "grad_norm": 1.520191935935746, + "learning_rate": 8.26871634058909e-06, + "loss": 0.4473, + "step": 13402 + }, + { + "epoch": 0.29502531366938145, + "grad_norm": 1.4776401632850058, + "learning_rate": 8.26844659588774e-06, + "loss": 0.3174, + "step": 13403 + }, + { + "epoch": 0.29504732555580016, + "grad_norm": 1.417941268381305, + "learning_rate": 8.268176834574691e-06, + "loss": 0.279, + "step": 13404 + }, + { + "epoch": 0.2950693374422188, + "grad_norm": 1.354853318286711, + "learning_rate": 8.267907056651305e-06, + "loss": 0.419, + "step": 13405 + }, + { + "epoch": 0.29509134932863745, + "grad_norm": 0.6223306240978694, + "learning_rate": 8.267637262118962e-06, + "loss": 0.4978, + "step": 13406 + }, + { + "epoch": 0.29511336121505616, + "grad_norm": 1.510943522000868, + "learning_rate": 8.267367450979026e-06, + "loss": 0.3054, + "step": 13407 + }, + { + "epoch": 0.2951353731014748, + "grad_norm": 1.0088926217742449, + "learning_rate": 8.267097623232874e-06, + "loss": 0.2491, + "step": 13408 + }, + { + "epoch": 0.29515738498789346, + "grad_norm": 3.242333221845448, + "learning_rate": 8.266827778881871e-06, + "loss": 0.3365, + "step": 13409 + }, + { + "epoch": 0.2951793968743121, + "grad_norm": 1.2419038380566876, + "learning_rate": 8.266557917927396e-06, + "loss": 0.4429, + "step": 13410 + }, + { + "epoch": 0.2952014087607308, + "grad_norm": 0.6340170034391893, + "learning_rate": 8.266288040370815e-06, + "loss": 0.5134, + "step": 13411 + }, + { + "epoch": 0.29522342064714946, + "grad_norm": 1.3150305734393226, + "learning_rate": 8.266018146213502e-06, + "loss": 0.3264, + "step": 13412 + }, + { + "epoch": 0.2952454325335681, + "grad_norm": 1.2610343757496627, + "learning_rate": 8.265748235456827e-06, + "loss": 0.3399, + "step": 13413 + }, + { + "epoch": 0.2952674444199868, + "grad_norm": 0.6194120666461783, + "learning_rate": 8.265478308102162e-06, + "loss": 0.51, + "step": 13414 + }, + { + "epoch": 0.29528945630640546, + "grad_norm": 1.6706744375808888, + "learning_rate": 8.265208364150882e-06, + "loss": 0.4279, + "step": 13415 + }, + { + "epoch": 0.2953114681928241, + "grad_norm": 1.3474044836581274, + "learning_rate": 8.264938403604357e-06, + "loss": 0.3627, + "step": 13416 + }, + { + "epoch": 0.2953334800792428, + "grad_norm": 1.498600754342386, + "learning_rate": 8.264668426463957e-06, + "loss": 0.3308, + "step": 13417 + }, + { + "epoch": 0.29535549196566147, + "grad_norm": 7.7470344195095375, + "learning_rate": 8.264398432731055e-06, + "loss": 0.318, + "step": 13418 + }, + { + "epoch": 0.2953775038520801, + "grad_norm": 4.210566456938356, + "learning_rate": 8.264128422407027e-06, + "loss": 0.3485, + "step": 13419 + }, + { + "epoch": 0.29539951573849876, + "grad_norm": 0.6493754356987596, + "learning_rate": 8.263858395493243e-06, + "loss": 0.5372, + "step": 13420 + }, + { + "epoch": 0.29542152762491747, + "grad_norm": 0.6192182164772688, + "learning_rate": 8.263588351991072e-06, + "loss": 0.5018, + "step": 13421 + }, + { + "epoch": 0.2954435395113361, + "grad_norm": 1.2037651384096115, + "learning_rate": 8.263318291901892e-06, + "loss": 0.2458, + "step": 13422 + }, + { + "epoch": 0.29546555139775477, + "grad_norm": 1.3647989879560845, + "learning_rate": 8.26304821522707e-06, + "loss": 0.2928, + "step": 13423 + }, + { + "epoch": 0.29548756328417347, + "grad_norm": 1.3632050062166645, + "learning_rate": 8.262778121967985e-06, + "loss": 0.4271, + "step": 13424 + }, + { + "epoch": 0.2955095751705921, + "grad_norm": 2.045699845151138, + "learning_rate": 8.262508012126005e-06, + "loss": 0.4148, + "step": 13425 + }, + { + "epoch": 0.29553158705701077, + "grad_norm": 1.4572859457227716, + "learning_rate": 8.262237885702503e-06, + "loss": 0.2832, + "step": 13426 + }, + { + "epoch": 0.2955535989434295, + "grad_norm": 1.8577437079975538, + "learning_rate": 8.261967742698854e-06, + "loss": 0.36, + "step": 13427 + }, + { + "epoch": 0.2955756108298481, + "grad_norm": 1.3999495548179846, + "learning_rate": 8.26169758311643e-06, + "loss": 0.2991, + "step": 13428 + }, + { + "epoch": 0.2955976227162668, + "grad_norm": 2.4753726845821538, + "learning_rate": 8.261427406956604e-06, + "loss": 0.2629, + "step": 13429 + }, + { + "epoch": 0.2956196346026855, + "grad_norm": 1.6425941746994441, + "learning_rate": 8.26115721422075e-06, + "loss": 0.2984, + "step": 13430 + }, + { + "epoch": 0.2956416464891041, + "grad_norm": 1.6165363340268857, + "learning_rate": 8.26088700491024e-06, + "loss": 0.394, + "step": 13431 + }, + { + "epoch": 0.2956636583755228, + "grad_norm": 1.2055758191472767, + "learning_rate": 8.260616779026446e-06, + "loss": 0.3362, + "step": 13432 + }, + { + "epoch": 0.2956856702619414, + "grad_norm": 1.456842366845082, + "learning_rate": 8.260346536570747e-06, + "loss": 0.3445, + "step": 13433 + }, + { + "epoch": 0.29570768214836013, + "grad_norm": 1.8139013626817355, + "learning_rate": 8.260076277544511e-06, + "loss": 0.3038, + "step": 13434 + }, + { + "epoch": 0.2957296940347788, + "grad_norm": 1.5184494639163533, + "learning_rate": 8.259806001949112e-06, + "loss": 0.3864, + "step": 13435 + }, + { + "epoch": 0.2957517059211974, + "grad_norm": 0.6822978556436429, + "learning_rate": 8.259535709785926e-06, + "loss": 0.53, + "step": 13436 + }, + { + "epoch": 0.29577371780761613, + "grad_norm": 2.831372739609773, + "learning_rate": 8.259265401056326e-06, + "loss": 0.4101, + "step": 13437 + }, + { + "epoch": 0.2957957296940348, + "grad_norm": 1.5753560477018327, + "learning_rate": 8.258995075761686e-06, + "loss": 0.348, + "step": 13438 + }, + { + "epoch": 0.29581774158045343, + "grad_norm": 1.718005262988419, + "learning_rate": 8.258724733903378e-06, + "loss": 0.4955, + "step": 13439 + }, + { + "epoch": 0.29583975346687214, + "grad_norm": 1.6757509566043447, + "learning_rate": 8.258454375482778e-06, + "loss": 0.4101, + "step": 13440 + }, + { + "epoch": 0.2958617653532908, + "grad_norm": 1.2807693577531483, + "learning_rate": 8.25818400050126e-06, + "loss": 0.3328, + "step": 13441 + }, + { + "epoch": 0.29588377723970943, + "grad_norm": 0.6180136684466132, + "learning_rate": 8.257913608960197e-06, + "loss": 0.5028, + "step": 13442 + }, + { + "epoch": 0.2959057891261281, + "grad_norm": 0.6214341996413815, + "learning_rate": 8.257643200860964e-06, + "loss": 0.5066, + "step": 13443 + }, + { + "epoch": 0.2959278010125468, + "grad_norm": 2.154268081712223, + "learning_rate": 8.257372776204936e-06, + "loss": 0.309, + "step": 13444 + }, + { + "epoch": 0.29594981289896544, + "grad_norm": 3.3921767166988013, + "learning_rate": 8.257102334993487e-06, + "loss": 0.2919, + "step": 13445 + }, + { + "epoch": 0.2959718247853841, + "grad_norm": 1.4159914431477887, + "learning_rate": 8.256831877227991e-06, + "loss": 0.3373, + "step": 13446 + }, + { + "epoch": 0.2959938366718028, + "grad_norm": 1.419902020063715, + "learning_rate": 8.256561402909824e-06, + "loss": 0.3309, + "step": 13447 + }, + { + "epoch": 0.29601584855822144, + "grad_norm": 1.7912514182234125, + "learning_rate": 8.256290912040358e-06, + "loss": 0.3297, + "step": 13448 + }, + { + "epoch": 0.2960378604446401, + "grad_norm": 1.4151820039631167, + "learning_rate": 8.25602040462097e-06, + "loss": 0.2858, + "step": 13449 + }, + { + "epoch": 0.2960598723310588, + "grad_norm": 1.483218086193354, + "learning_rate": 8.255749880653032e-06, + "loss": 0.3095, + "step": 13450 + }, + { + "epoch": 0.29608188421747744, + "grad_norm": 1.1463365369972063, + "learning_rate": 8.255479340137922e-06, + "loss": 0.2628, + "step": 13451 + }, + { + "epoch": 0.2961038961038961, + "grad_norm": 0.7085471859466995, + "learning_rate": 8.255208783077015e-06, + "loss": 0.5145, + "step": 13452 + }, + { + "epoch": 0.2961259079903148, + "grad_norm": 1.5600273913571288, + "learning_rate": 8.254938209471685e-06, + "loss": 0.3148, + "step": 13453 + }, + { + "epoch": 0.29614791987673345, + "grad_norm": 0.6778603111781087, + "learning_rate": 8.254667619323307e-06, + "loss": 0.5066, + "step": 13454 + }, + { + "epoch": 0.2961699317631521, + "grad_norm": 1.3253683850880031, + "learning_rate": 8.254397012633256e-06, + "loss": 0.3443, + "step": 13455 + }, + { + "epoch": 0.29619194364957074, + "grad_norm": 1.5152491090088587, + "learning_rate": 8.254126389402908e-06, + "loss": 0.2934, + "step": 13456 + }, + { + "epoch": 0.29621395553598945, + "grad_norm": 1.8471974030280025, + "learning_rate": 8.25385574963364e-06, + "loss": 0.3435, + "step": 13457 + }, + { + "epoch": 0.2962359674224081, + "grad_norm": 1.3230609333654753, + "learning_rate": 8.253585093326824e-06, + "loss": 0.2981, + "step": 13458 + }, + { + "epoch": 0.29625797930882675, + "grad_norm": 1.102069520057713, + "learning_rate": 8.253314420483836e-06, + "loss": 0.2807, + "step": 13459 + }, + { + "epoch": 0.29627999119524545, + "grad_norm": 1.4963487364064656, + "learning_rate": 8.253043731106056e-06, + "loss": 0.341, + "step": 13460 + }, + { + "epoch": 0.2963020030816641, + "grad_norm": 0.7078694695638272, + "learning_rate": 8.252773025194856e-06, + "loss": 0.5299, + "step": 13461 + }, + { + "epoch": 0.29632401496808275, + "grad_norm": 2.045724041327339, + "learning_rate": 8.252502302751613e-06, + "loss": 0.3368, + "step": 13462 + }, + { + "epoch": 0.29634602685450145, + "grad_norm": 1.0013273282068482, + "learning_rate": 8.252231563777701e-06, + "loss": 0.2308, + "step": 13463 + }, + { + "epoch": 0.2963680387409201, + "grad_norm": 1.2720588667781658, + "learning_rate": 8.251960808274499e-06, + "loss": 0.3235, + "step": 13464 + }, + { + "epoch": 0.29639005062733875, + "grad_norm": 2.13909325853893, + "learning_rate": 8.251690036243381e-06, + "loss": 0.319, + "step": 13465 + }, + { + "epoch": 0.2964120625137574, + "grad_norm": 1.8302815957876535, + "learning_rate": 8.251419247685724e-06, + "loss": 0.3526, + "step": 13466 + }, + { + "epoch": 0.2964340744001761, + "grad_norm": 1.9545050555880936, + "learning_rate": 8.251148442602905e-06, + "loss": 0.4809, + "step": 13467 + }, + { + "epoch": 0.29645608628659476, + "grad_norm": 2.3018640176496987, + "learning_rate": 8.2508776209963e-06, + "loss": 0.3024, + "step": 13468 + }, + { + "epoch": 0.2964780981730134, + "grad_norm": 12.416823258903413, + "learning_rate": 8.250606782867283e-06, + "loss": 0.274, + "step": 13469 + }, + { + "epoch": 0.2965001100594321, + "grad_norm": 1.4925462156366904, + "learning_rate": 8.250335928217234e-06, + "loss": 0.493, + "step": 13470 + }, + { + "epoch": 0.29652212194585076, + "grad_norm": 1.2869893813537367, + "learning_rate": 8.250065057047528e-06, + "loss": 0.4051, + "step": 13471 + }, + { + "epoch": 0.2965441338322694, + "grad_norm": 4.051107719031931, + "learning_rate": 8.249794169359542e-06, + "loss": 0.3364, + "step": 13472 + }, + { + "epoch": 0.2965661457186881, + "grad_norm": 2.4677576909970056, + "learning_rate": 8.249523265154653e-06, + "loss": 0.3744, + "step": 13473 + }, + { + "epoch": 0.29658815760510676, + "grad_norm": 1.3852923592194168, + "learning_rate": 8.249252344434235e-06, + "loss": 0.3135, + "step": 13474 + }, + { + "epoch": 0.2966101694915254, + "grad_norm": 1.6263838751431432, + "learning_rate": 8.24898140719967e-06, + "loss": 0.2854, + "step": 13475 + }, + { + "epoch": 0.2966321813779441, + "grad_norm": 1.481701294351258, + "learning_rate": 8.248710453452331e-06, + "loss": 0.3658, + "step": 13476 + }, + { + "epoch": 0.29665419326436276, + "grad_norm": 1.7954126627143125, + "learning_rate": 8.248439483193598e-06, + "loss": 0.3947, + "step": 13477 + }, + { + "epoch": 0.2966762051507814, + "grad_norm": 2.1827489147490744, + "learning_rate": 8.248168496424844e-06, + "loss": 0.2733, + "step": 13478 + }, + { + "epoch": 0.29669821703720006, + "grad_norm": 0.6658482018477325, + "learning_rate": 8.247897493147452e-06, + "loss": 0.5073, + "step": 13479 + }, + { + "epoch": 0.29672022892361877, + "grad_norm": 1.7085921929083543, + "learning_rate": 8.247626473362795e-06, + "loss": 0.3556, + "step": 13480 + }, + { + "epoch": 0.2967422408100374, + "grad_norm": 0.6381098007935683, + "learning_rate": 8.24735543707225e-06, + "loss": 0.5233, + "step": 13481 + }, + { + "epoch": 0.29676425269645607, + "grad_norm": 0.634703910848576, + "learning_rate": 8.2470843842772e-06, + "loss": 0.5198, + "step": 13482 + }, + { + "epoch": 0.29678626458287477, + "grad_norm": 1.4575892419364451, + "learning_rate": 8.246813314979016e-06, + "loss": 0.2807, + "step": 13483 + }, + { + "epoch": 0.2968082764692934, + "grad_norm": 2.8824271329406415, + "learning_rate": 8.24654222917908e-06, + "loss": 0.376, + "step": 13484 + }, + { + "epoch": 0.29683028835571207, + "grad_norm": 2.40561863500293, + "learning_rate": 8.24627112687877e-06, + "loss": 0.4282, + "step": 13485 + }, + { + "epoch": 0.2968523002421308, + "grad_norm": 1.2977319736497723, + "learning_rate": 8.246000008079458e-06, + "loss": 0.3094, + "step": 13486 + }, + { + "epoch": 0.2968743121285494, + "grad_norm": 2.6869581245869485, + "learning_rate": 8.245728872782528e-06, + "loss": 0.3058, + "step": 13487 + }, + { + "epoch": 0.29689632401496807, + "grad_norm": 1.9165236059676727, + "learning_rate": 8.245457720989359e-06, + "loss": 0.3447, + "step": 13488 + }, + { + "epoch": 0.2969183359013868, + "grad_norm": 1.7695126962069803, + "learning_rate": 8.245186552701323e-06, + "loss": 0.2713, + "step": 13489 + }, + { + "epoch": 0.2969403477878054, + "grad_norm": 2.6256423182942563, + "learning_rate": 8.244915367919802e-06, + "loss": 0.358, + "step": 13490 + }, + { + "epoch": 0.2969623596742241, + "grad_norm": 2.5539741706808257, + "learning_rate": 8.244644166646174e-06, + "loss": 0.3167, + "step": 13491 + }, + { + "epoch": 0.2969843715606427, + "grad_norm": 1.5328493620802521, + "learning_rate": 8.244372948881819e-06, + "loss": 0.2919, + "step": 13492 + }, + { + "epoch": 0.29700638344706143, + "grad_norm": 3.1604108250924376, + "learning_rate": 8.244101714628112e-06, + "loss": 0.4107, + "step": 13493 + }, + { + "epoch": 0.2970283953334801, + "grad_norm": 3.341073817509667, + "learning_rate": 8.243830463886433e-06, + "loss": 0.2951, + "step": 13494 + }, + { + "epoch": 0.2970504072198987, + "grad_norm": 1.6188201120455292, + "learning_rate": 8.243559196658161e-06, + "loss": 0.3155, + "step": 13495 + }, + { + "epoch": 0.29707241910631743, + "grad_norm": 0.7850255573232048, + "learning_rate": 8.243287912944676e-06, + "loss": 0.5206, + "step": 13496 + }, + { + "epoch": 0.2970944309927361, + "grad_norm": 1.7636874198501413, + "learning_rate": 8.243016612747353e-06, + "loss": 0.2896, + "step": 13497 + }, + { + "epoch": 0.29711644287915473, + "grad_norm": 2.25306325934983, + "learning_rate": 8.242745296067573e-06, + "loss": 0.363, + "step": 13498 + }, + { + "epoch": 0.29713845476557343, + "grad_norm": 3.2228936539595376, + "learning_rate": 8.242473962906716e-06, + "loss": 0.3514, + "step": 13499 + }, + { + "epoch": 0.2971604666519921, + "grad_norm": 0.6895000033726887, + "learning_rate": 8.242202613266161e-06, + "loss": 0.5146, + "step": 13500 + }, + { + "epoch": 0.29718247853841073, + "grad_norm": 2.029804418057535, + "learning_rate": 8.241931247147285e-06, + "loss": 0.2676, + "step": 13501 + }, + { + "epoch": 0.2972044904248294, + "grad_norm": 1.5663134106184453, + "learning_rate": 8.24165986455147e-06, + "loss": 0.4207, + "step": 13502 + }, + { + "epoch": 0.2972265023112481, + "grad_norm": 0.6227445440764312, + "learning_rate": 8.241388465480093e-06, + "loss": 0.4943, + "step": 13503 + }, + { + "epoch": 0.29724851419766674, + "grad_norm": 0.6415428635737166, + "learning_rate": 8.241117049934535e-06, + "loss": 0.5133, + "step": 13504 + }, + { + "epoch": 0.2972705260840854, + "grad_norm": 4.438200236970508, + "learning_rate": 8.240845617916172e-06, + "loss": 0.3301, + "step": 13505 + }, + { + "epoch": 0.2972925379705041, + "grad_norm": 2.7402016883092952, + "learning_rate": 8.240574169426389e-06, + "loss": 0.3914, + "step": 13506 + }, + { + "epoch": 0.29731454985692274, + "grad_norm": 1.8405422042110118, + "learning_rate": 8.24030270446656e-06, + "loss": 0.319, + "step": 13507 + }, + { + "epoch": 0.2973365617433414, + "grad_norm": 1.9509262850758138, + "learning_rate": 8.24003122303807e-06, + "loss": 0.3454, + "step": 13508 + }, + { + "epoch": 0.2973585736297601, + "grad_norm": 2.3295764794309872, + "learning_rate": 8.239759725142294e-06, + "loss": 0.4111, + "step": 13509 + }, + { + "epoch": 0.29738058551617874, + "grad_norm": 2.7978558246643037, + "learning_rate": 8.239488210780615e-06, + "loss": 0.377, + "step": 13510 + }, + { + "epoch": 0.2974025974025974, + "grad_norm": 1.9956079640099045, + "learning_rate": 8.239216679954414e-06, + "loss": 0.3696, + "step": 13511 + }, + { + "epoch": 0.2974246092890161, + "grad_norm": 1.7708737476965037, + "learning_rate": 8.238945132665066e-06, + "loss": 0.3638, + "step": 13512 + }, + { + "epoch": 0.29744662117543474, + "grad_norm": 0.6904395496196256, + "learning_rate": 8.238673568913956e-06, + "loss": 0.5273, + "step": 13513 + }, + { + "epoch": 0.2974686330618534, + "grad_norm": 3.2490642484871937, + "learning_rate": 8.238401988702462e-06, + "loss": 0.3526, + "step": 13514 + }, + { + "epoch": 0.29749064494827204, + "grad_norm": 1.627600089893524, + "learning_rate": 8.238130392031965e-06, + "loss": 0.2874, + "step": 13515 + }, + { + "epoch": 0.29751265683469075, + "grad_norm": 2.145104407549204, + "learning_rate": 8.237858778903844e-06, + "loss": 0.3897, + "step": 13516 + }, + { + "epoch": 0.2975346687211094, + "grad_norm": 0.6583425710226403, + "learning_rate": 8.237587149319481e-06, + "loss": 0.5112, + "step": 13517 + }, + { + "epoch": 0.29755668060752805, + "grad_norm": 1.5165588779186439, + "learning_rate": 8.237315503280256e-06, + "loss": 0.3807, + "step": 13518 + }, + { + "epoch": 0.29757869249394675, + "grad_norm": 0.6285540592029215, + "learning_rate": 8.237043840787552e-06, + "loss": 0.5133, + "step": 13519 + }, + { + "epoch": 0.2976007043803654, + "grad_norm": 1.7649309019228525, + "learning_rate": 8.236772161842745e-06, + "loss": 0.3166, + "step": 13520 + }, + { + "epoch": 0.29762271626678405, + "grad_norm": 2.6856593979902956, + "learning_rate": 8.236500466447219e-06, + "loss": 0.3328, + "step": 13521 + }, + { + "epoch": 0.29764472815320275, + "grad_norm": 5.629229277394204, + "learning_rate": 8.236228754602353e-06, + "loss": 0.3228, + "step": 13522 + }, + { + "epoch": 0.2976667400396214, + "grad_norm": 1.662746257234648, + "learning_rate": 8.23595702630953e-06, + "loss": 0.2051, + "step": 13523 + }, + { + "epoch": 0.29768875192604005, + "grad_norm": 0.6158680378889663, + "learning_rate": 8.23568528157013e-06, + "loss": 0.4957, + "step": 13524 + }, + { + "epoch": 0.2977107638124587, + "grad_norm": 1.578816880523049, + "learning_rate": 8.235413520385536e-06, + "loss": 0.2899, + "step": 13525 + }, + { + "epoch": 0.2977327756988774, + "grad_norm": 0.6551012391440708, + "learning_rate": 8.235141742757124e-06, + "loss": 0.5269, + "step": 13526 + }, + { + "epoch": 0.29775478758529605, + "grad_norm": 1.779453660034939, + "learning_rate": 8.234869948686281e-06, + "loss": 0.2112, + "step": 13527 + }, + { + "epoch": 0.2977767994717147, + "grad_norm": 1.7599996830440914, + "learning_rate": 8.234598138174386e-06, + "loss": 0.3665, + "step": 13528 + }, + { + "epoch": 0.2977988113581334, + "grad_norm": 2.4354570389984183, + "learning_rate": 8.234326311222822e-06, + "loss": 0.2316, + "step": 13529 + }, + { + "epoch": 0.29782082324455206, + "grad_norm": 1.2731750646059388, + "learning_rate": 8.234054467832965e-06, + "loss": 0.2972, + "step": 13530 + }, + { + "epoch": 0.2978428351309707, + "grad_norm": 0.6110548870905751, + "learning_rate": 8.233782608006204e-06, + "loss": 0.5264, + "step": 13531 + }, + { + "epoch": 0.2978648470173894, + "grad_norm": 1.8005704850426993, + "learning_rate": 8.233510731743917e-06, + "loss": 0.3202, + "step": 13532 + }, + { + "epoch": 0.29788685890380806, + "grad_norm": 1.7436678885310055, + "learning_rate": 8.233238839047487e-06, + "loss": 0.3279, + "step": 13533 + }, + { + "epoch": 0.2979088707902267, + "grad_norm": 1.5989679769631726, + "learning_rate": 8.232966929918293e-06, + "loss": 0.4169, + "step": 13534 + }, + { + "epoch": 0.2979308826766454, + "grad_norm": 0.6101254969555635, + "learning_rate": 8.23269500435772e-06, + "loss": 0.5101, + "step": 13535 + }, + { + "epoch": 0.29795289456306406, + "grad_norm": 1.7207608039961397, + "learning_rate": 8.23242306236715e-06, + "loss": 0.3745, + "step": 13536 + }, + { + "epoch": 0.2979749064494827, + "grad_norm": 0.6185167237188509, + "learning_rate": 8.232151103947963e-06, + "loss": 0.5186, + "step": 13537 + }, + { + "epoch": 0.29799691833590136, + "grad_norm": 1.6344196876481625, + "learning_rate": 8.231879129101544e-06, + "loss": 0.3335, + "step": 13538 + }, + { + "epoch": 0.29801893022232007, + "grad_norm": 2.525702055266434, + "learning_rate": 8.231607137829274e-06, + "loss": 0.342, + "step": 13539 + }, + { + "epoch": 0.2980409421087387, + "grad_norm": 1.453202055238866, + "learning_rate": 8.231335130132534e-06, + "loss": 0.3973, + "step": 13540 + }, + { + "epoch": 0.29806295399515736, + "grad_norm": 1.3769554171508516, + "learning_rate": 8.231063106012706e-06, + "loss": 0.2836, + "step": 13541 + }, + { + "epoch": 0.29808496588157607, + "grad_norm": 0.6126245229540255, + "learning_rate": 8.230791065471177e-06, + "loss": 0.5084, + "step": 13542 + }, + { + "epoch": 0.2981069777679947, + "grad_norm": 0.609472539031067, + "learning_rate": 8.230519008509327e-06, + "loss": 0.5288, + "step": 13543 + }, + { + "epoch": 0.29812898965441337, + "grad_norm": 1.5829323542534308, + "learning_rate": 8.230246935128537e-06, + "loss": 0.3087, + "step": 13544 + }, + { + "epoch": 0.2981510015408321, + "grad_norm": 1.664368580325436, + "learning_rate": 8.22997484533019e-06, + "loss": 0.3549, + "step": 13545 + }, + { + "epoch": 0.2981730134272507, + "grad_norm": 7.763582207448079, + "learning_rate": 8.229702739115673e-06, + "loss": 0.2883, + "step": 13546 + }, + { + "epoch": 0.29819502531366937, + "grad_norm": 1.7525347855956792, + "learning_rate": 8.229430616486368e-06, + "loss": 0.2125, + "step": 13547 + }, + { + "epoch": 0.298217037200088, + "grad_norm": 3.557026750142003, + "learning_rate": 8.22915847744365e-06, + "loss": 0.3475, + "step": 13548 + }, + { + "epoch": 0.2982390490865067, + "grad_norm": 2.1632065924845363, + "learning_rate": 8.228886321988914e-06, + "loss": 0.3486, + "step": 13549 + }, + { + "epoch": 0.2982610609729254, + "grad_norm": 0.6360168179302842, + "learning_rate": 8.228614150123535e-06, + "loss": 0.5017, + "step": 13550 + }, + { + "epoch": 0.298283072859344, + "grad_norm": 1.8622151778644904, + "learning_rate": 8.228341961848898e-06, + "loss": 0.3695, + "step": 13551 + }, + { + "epoch": 0.2983050847457627, + "grad_norm": 1.7922194342758424, + "learning_rate": 8.22806975716639e-06, + "loss": 0.4005, + "step": 13552 + }, + { + "epoch": 0.2983270966321814, + "grad_norm": 1.815162033167171, + "learning_rate": 8.22779753607739e-06, + "loss": 0.3056, + "step": 13553 + }, + { + "epoch": 0.2983491085186, + "grad_norm": 1.8155977364408735, + "learning_rate": 8.227525298583283e-06, + "loss": 0.3204, + "step": 13554 + }, + { + "epoch": 0.29837112040501873, + "grad_norm": 2.0051280901220574, + "learning_rate": 8.227253044685453e-06, + "loss": 0.2987, + "step": 13555 + }, + { + "epoch": 0.2983931322914374, + "grad_norm": 1.622391536513579, + "learning_rate": 8.226980774385284e-06, + "loss": 0.3541, + "step": 13556 + }, + { + "epoch": 0.29841514417785603, + "grad_norm": 1.7912752167251933, + "learning_rate": 8.226708487684162e-06, + "loss": 0.3575, + "step": 13557 + }, + { + "epoch": 0.29843715606427473, + "grad_norm": 1.697057680533489, + "learning_rate": 8.226436184583465e-06, + "loss": 0.4082, + "step": 13558 + }, + { + "epoch": 0.2984591679506934, + "grad_norm": 1.7104974855122252, + "learning_rate": 8.22616386508458e-06, + "loss": 0.3194, + "step": 13559 + }, + { + "epoch": 0.29848117983711203, + "grad_norm": 1.8792830627777628, + "learning_rate": 8.225891529188893e-06, + "loss": 0.3442, + "step": 13560 + }, + { + "epoch": 0.2985031917235307, + "grad_norm": 6.755782447645154, + "learning_rate": 8.225619176897786e-06, + "loss": 0.3869, + "step": 13561 + }, + { + "epoch": 0.2985252036099494, + "grad_norm": 2.2070923266794997, + "learning_rate": 8.225346808212644e-06, + "loss": 0.3085, + "step": 13562 + }, + { + "epoch": 0.29854721549636803, + "grad_norm": 2.645191459671456, + "learning_rate": 8.22507442313485e-06, + "loss": 0.3298, + "step": 13563 + }, + { + "epoch": 0.2985692273827867, + "grad_norm": 1.2295206890793005, + "learning_rate": 8.22480202166579e-06, + "loss": 0.3208, + "step": 13564 + }, + { + "epoch": 0.2985912392692054, + "grad_norm": 1.3396449032714277, + "learning_rate": 8.224529603806849e-06, + "loss": 0.2629, + "step": 13565 + }, + { + "epoch": 0.29861325115562404, + "grad_norm": 2.0737576735307504, + "learning_rate": 8.224257169559409e-06, + "loss": 0.4231, + "step": 13566 + }, + { + "epoch": 0.2986352630420427, + "grad_norm": 1.489187958647533, + "learning_rate": 8.223984718924856e-06, + "loss": 0.3423, + "step": 13567 + }, + { + "epoch": 0.2986572749284614, + "grad_norm": 11.377524688556303, + "learning_rate": 8.223712251904575e-06, + "loss": 0.279, + "step": 13568 + }, + { + "epoch": 0.29867928681488004, + "grad_norm": 2.3608301875359623, + "learning_rate": 8.22343976849995e-06, + "loss": 0.414, + "step": 13569 + }, + { + "epoch": 0.2987012987012987, + "grad_norm": 1.6465493524115893, + "learning_rate": 8.223167268712367e-06, + "loss": 0.3952, + "step": 13570 + }, + { + "epoch": 0.2987233105877174, + "grad_norm": 1.2305933100814186, + "learning_rate": 8.22289475254321e-06, + "loss": 0.2858, + "step": 13571 + }, + { + "epoch": 0.29874532247413604, + "grad_norm": 1.6852584174579273, + "learning_rate": 8.222622219993865e-06, + "loss": 0.3041, + "step": 13572 + }, + { + "epoch": 0.2987673343605547, + "grad_norm": 2.2637690121230305, + "learning_rate": 8.222349671065715e-06, + "loss": 0.3056, + "step": 13573 + }, + { + "epoch": 0.29878934624697334, + "grad_norm": 0.6548523307657682, + "learning_rate": 8.222077105760148e-06, + "loss": 0.5164, + "step": 13574 + }, + { + "epoch": 0.29881135813339205, + "grad_norm": 1.796958607830082, + "learning_rate": 8.22180452407855e-06, + "loss": 0.3701, + "step": 13575 + }, + { + "epoch": 0.2988333700198107, + "grad_norm": 1.1211154022533802, + "learning_rate": 8.2215319260223e-06, + "loss": 0.2443, + "step": 13576 + }, + { + "epoch": 0.29885538190622934, + "grad_norm": 1.439105595429023, + "learning_rate": 8.22125931159279e-06, + "loss": 0.2774, + "step": 13577 + }, + { + "epoch": 0.29887739379264805, + "grad_norm": 2.8997552695133617, + "learning_rate": 8.220986680791405e-06, + "loss": 0.3091, + "step": 13578 + }, + { + "epoch": 0.2988994056790667, + "grad_norm": 1.8022873092243967, + "learning_rate": 8.220714033619528e-06, + "loss": 0.3723, + "step": 13579 + }, + { + "epoch": 0.29892141756548535, + "grad_norm": 0.6182004492717916, + "learning_rate": 8.220441370078546e-06, + "loss": 0.509, + "step": 13580 + }, + { + "epoch": 0.29894342945190405, + "grad_norm": 2.1850439329120284, + "learning_rate": 8.220168690169844e-06, + "loss": 0.3293, + "step": 13581 + }, + { + "epoch": 0.2989654413383227, + "grad_norm": 1.516978949444276, + "learning_rate": 8.219895993894808e-06, + "loss": 0.3321, + "step": 13582 + }, + { + "epoch": 0.29898745322474135, + "grad_norm": 1.5671547061548774, + "learning_rate": 8.219623281254825e-06, + "loss": 0.3857, + "step": 13583 + }, + { + "epoch": 0.29900946511116, + "grad_norm": 1.676187181825604, + "learning_rate": 8.219350552251281e-06, + "loss": 0.2874, + "step": 13584 + }, + { + "epoch": 0.2990314769975787, + "grad_norm": 1.9318250651954085, + "learning_rate": 8.219077806885562e-06, + "loss": 0.3521, + "step": 13585 + }, + { + "epoch": 0.29905348888399735, + "grad_norm": 1.681032744103794, + "learning_rate": 8.218805045159054e-06, + "loss": 0.4399, + "step": 13586 + }, + { + "epoch": 0.299075500770416, + "grad_norm": 1.2029209969887864, + "learning_rate": 8.218532267073141e-06, + "loss": 0.2742, + "step": 13587 + }, + { + "epoch": 0.2990975126568347, + "grad_norm": 1.7873144466216127, + "learning_rate": 8.218259472629213e-06, + "loss": 0.3834, + "step": 13588 + }, + { + "epoch": 0.29911952454325336, + "grad_norm": 0.6211136315030239, + "learning_rate": 8.217986661828655e-06, + "loss": 0.4988, + "step": 13589 + }, + { + "epoch": 0.299141536429672, + "grad_norm": 9.0128282078178, + "learning_rate": 8.217713834672854e-06, + "loss": 0.364, + "step": 13590 + }, + { + "epoch": 0.2991635483160907, + "grad_norm": 0.6160121498843191, + "learning_rate": 8.217440991163194e-06, + "loss": 0.4903, + "step": 13591 + }, + { + "epoch": 0.29918556020250936, + "grad_norm": 2.466546611916817, + "learning_rate": 8.217168131301065e-06, + "loss": 0.4568, + "step": 13592 + }, + { + "epoch": 0.299207572088928, + "grad_norm": 0.6073957082867986, + "learning_rate": 8.216895255087854e-06, + "loss": 0.5016, + "step": 13593 + }, + { + "epoch": 0.2992295839753467, + "grad_norm": 2.6651924108239458, + "learning_rate": 8.216622362524947e-06, + "loss": 0.2889, + "step": 13594 + }, + { + "epoch": 0.29925159586176536, + "grad_norm": 0.610702803045058, + "learning_rate": 8.216349453613728e-06, + "loss": 0.4924, + "step": 13595 + }, + { + "epoch": 0.299273607748184, + "grad_norm": 0.6232807372896884, + "learning_rate": 8.216076528355588e-06, + "loss": 0.4942, + "step": 13596 + }, + { + "epoch": 0.29929561963460266, + "grad_norm": 1.57402785310068, + "learning_rate": 8.215803586751911e-06, + "loss": 0.3264, + "step": 13597 + }, + { + "epoch": 0.29931763152102137, + "grad_norm": 1.4104256554702985, + "learning_rate": 8.215530628804087e-06, + "loss": 0.301, + "step": 13598 + }, + { + "epoch": 0.29933964340744, + "grad_norm": 1.8019266072379951, + "learning_rate": 8.215257654513503e-06, + "loss": 0.285, + "step": 13599 + }, + { + "epoch": 0.29936165529385866, + "grad_norm": 2.6531639497066926, + "learning_rate": 8.214984663881544e-06, + "loss": 0.3727, + "step": 13600 + }, + { + "epoch": 0.29938366718027737, + "grad_norm": 3.3947843761916796, + "learning_rate": 8.2147116569096e-06, + "loss": 0.4342, + "step": 13601 + }, + { + "epoch": 0.299405679066696, + "grad_norm": 1.8225226830961718, + "learning_rate": 8.214438633599058e-06, + "loss": 0.3104, + "step": 13602 + }, + { + "epoch": 0.29942769095311467, + "grad_norm": 0.6160518132609237, + "learning_rate": 8.214165593951304e-06, + "loss": 0.5045, + "step": 13603 + }, + { + "epoch": 0.29944970283953337, + "grad_norm": 2.3539772909161014, + "learning_rate": 8.213892537967727e-06, + "loss": 0.2894, + "step": 13604 + }, + { + "epoch": 0.299471714725952, + "grad_norm": 0.6407853524334832, + "learning_rate": 8.213619465649716e-06, + "loss": 0.495, + "step": 13605 + }, + { + "epoch": 0.29949372661237067, + "grad_norm": 2.2555402024651023, + "learning_rate": 8.213346376998656e-06, + "loss": 0.2741, + "step": 13606 + }, + { + "epoch": 0.2995157384987893, + "grad_norm": 1.722213928144923, + "learning_rate": 8.21307327201594e-06, + "loss": 0.3328, + "step": 13607 + }, + { + "epoch": 0.299537750385208, + "grad_norm": 2.2925045104122037, + "learning_rate": 8.212800150702949e-06, + "loss": 0.3504, + "step": 13608 + }, + { + "epoch": 0.2995597622716267, + "grad_norm": 1.652870191131541, + "learning_rate": 8.212527013061074e-06, + "loss": 0.3891, + "step": 13609 + }, + { + "epoch": 0.2995817741580453, + "grad_norm": 1.2863470132667576, + "learning_rate": 8.212253859091705e-06, + "loss": 0.3075, + "step": 13610 + }, + { + "epoch": 0.299603786044464, + "grad_norm": 5.76484281348305, + "learning_rate": 8.211980688796231e-06, + "loss": 0.4681, + "step": 13611 + }, + { + "epoch": 0.2996257979308827, + "grad_norm": 2.2761503027393144, + "learning_rate": 8.211707502176038e-06, + "loss": 0.3969, + "step": 13612 + }, + { + "epoch": 0.2996478098173013, + "grad_norm": 0.6518512816311427, + "learning_rate": 8.211434299232512e-06, + "loss": 0.4871, + "step": 13613 + }, + { + "epoch": 0.29966982170372003, + "grad_norm": 1.9732271608201353, + "learning_rate": 8.211161079967048e-06, + "loss": 0.376, + "step": 13614 + }, + { + "epoch": 0.2996918335901387, + "grad_norm": 1.3860155794395046, + "learning_rate": 8.21088784438103e-06, + "loss": 0.3562, + "step": 13615 + }, + { + "epoch": 0.29971384547655733, + "grad_norm": 0.6362632567501243, + "learning_rate": 8.210614592475847e-06, + "loss": 0.5072, + "step": 13616 + }, + { + "epoch": 0.29973585736297603, + "grad_norm": 2.2812469269564337, + "learning_rate": 8.210341324252891e-06, + "loss": 0.2834, + "step": 13617 + }, + { + "epoch": 0.2997578692493947, + "grad_norm": 1.2980774422478316, + "learning_rate": 8.210068039713545e-06, + "loss": 0.3696, + "step": 13618 + }, + { + "epoch": 0.29977988113581333, + "grad_norm": 1.7870473642565112, + "learning_rate": 8.209794738859204e-06, + "loss": 0.2736, + "step": 13619 + }, + { + "epoch": 0.299801893022232, + "grad_norm": 1.259649246652406, + "learning_rate": 8.209521421691254e-06, + "loss": 0.4326, + "step": 13620 + }, + { + "epoch": 0.2998239049086507, + "grad_norm": 0.6417016227016358, + "learning_rate": 8.209248088211086e-06, + "loss": 0.5299, + "step": 13621 + }, + { + "epoch": 0.29984591679506933, + "grad_norm": 1.9514452842755354, + "learning_rate": 8.208974738420085e-06, + "loss": 0.4295, + "step": 13622 + }, + { + "epoch": 0.299867928681488, + "grad_norm": 1.256711898950055, + "learning_rate": 8.208701372319645e-06, + "loss": 0.2711, + "step": 13623 + }, + { + "epoch": 0.2998899405679067, + "grad_norm": 1.7043223664969467, + "learning_rate": 8.208427989911152e-06, + "loss": 0.359, + "step": 13624 + }, + { + "epoch": 0.29991195245432534, + "grad_norm": 1.578535479247014, + "learning_rate": 8.208154591196e-06, + "loss": 0.3095, + "step": 13625 + }, + { + "epoch": 0.299933964340744, + "grad_norm": 1.855060253723743, + "learning_rate": 8.207881176175573e-06, + "loss": 0.2336, + "step": 13626 + }, + { + "epoch": 0.2999559762271627, + "grad_norm": 0.6447574190934285, + "learning_rate": 8.207607744851263e-06, + "loss": 0.515, + "step": 13627 + }, + { + "epoch": 0.29997798811358134, + "grad_norm": 1.448611246752271, + "learning_rate": 8.20733429722446e-06, + "loss": 0.3288, + "step": 13628 + }, + { + "epoch": 0.3, + "grad_norm": 1.9174184636574094, + "learning_rate": 8.207060833296555e-06, + "loss": 0.3138, + "step": 13629 + }, + { + "epoch": 0.3000220118864187, + "grad_norm": 1.3500050618317316, + "learning_rate": 8.206787353068935e-06, + "loss": 0.2829, + "step": 13630 + }, + { + "epoch": 0.30004402377283734, + "grad_norm": 1.8622017562402033, + "learning_rate": 8.206513856542992e-06, + "loss": 0.2447, + "step": 13631 + }, + { + "epoch": 0.300066035659256, + "grad_norm": 2.353063214802477, + "learning_rate": 8.206240343720116e-06, + "loss": 0.3665, + "step": 13632 + }, + { + "epoch": 0.30008804754567464, + "grad_norm": 1.5876292027464365, + "learning_rate": 8.205966814601695e-06, + "loss": 0.4335, + "step": 13633 + }, + { + "epoch": 0.30011005943209335, + "grad_norm": 0.6545688931164512, + "learning_rate": 8.205693269189123e-06, + "loss": 0.4887, + "step": 13634 + }, + { + "epoch": 0.300132071318512, + "grad_norm": 1.3354154352570422, + "learning_rate": 8.205419707483786e-06, + "loss": 0.349, + "step": 13635 + }, + { + "epoch": 0.30015408320493064, + "grad_norm": 1.1962253821688442, + "learning_rate": 8.205146129487077e-06, + "loss": 0.2659, + "step": 13636 + }, + { + "epoch": 0.30017609509134935, + "grad_norm": 0.623995368903305, + "learning_rate": 8.204872535200385e-06, + "loss": 0.5056, + "step": 13637 + }, + { + "epoch": 0.300198106977768, + "grad_norm": 1.2954415326240374, + "learning_rate": 8.204598924625103e-06, + "loss": 0.347, + "step": 13638 + }, + { + "epoch": 0.30022011886418665, + "grad_norm": 3.313062953520085, + "learning_rate": 8.20432529776262e-06, + "loss": 0.3626, + "step": 13639 + }, + { + "epoch": 0.30024213075060535, + "grad_norm": 1.318788874681453, + "learning_rate": 8.204051654614325e-06, + "loss": 0.3083, + "step": 13640 + }, + { + "epoch": 0.300264142637024, + "grad_norm": 1.3772363136460812, + "learning_rate": 8.203777995181612e-06, + "loss": 0.3467, + "step": 13641 + }, + { + "epoch": 0.30028615452344265, + "grad_norm": 1.9152849492377215, + "learning_rate": 8.203504319465869e-06, + "loss": 0.387, + "step": 13642 + }, + { + "epoch": 0.3003081664098613, + "grad_norm": 1.4079542649874515, + "learning_rate": 8.20323062746849e-06, + "loss": 0.3187, + "step": 13643 + }, + { + "epoch": 0.30033017829628, + "grad_norm": 1.5701023021308562, + "learning_rate": 8.202956919190863e-06, + "loss": 0.3824, + "step": 13644 + }, + { + "epoch": 0.30035219018269865, + "grad_norm": 1.9894449686682056, + "learning_rate": 8.20268319463438e-06, + "loss": 0.2982, + "step": 13645 + }, + { + "epoch": 0.3003742020691173, + "grad_norm": 1.455764391637709, + "learning_rate": 8.202409453800433e-06, + "loss": 0.289, + "step": 13646 + }, + { + "epoch": 0.300396213955536, + "grad_norm": 0.6550418943726636, + "learning_rate": 8.202135696690415e-06, + "loss": 0.495, + "step": 13647 + }, + { + "epoch": 0.30041822584195466, + "grad_norm": 1.8159223329132863, + "learning_rate": 8.201861923305711e-06, + "loss": 0.4139, + "step": 13648 + }, + { + "epoch": 0.3004402377283733, + "grad_norm": 1.7704035960762807, + "learning_rate": 8.20158813364772e-06, + "loss": 0.297, + "step": 13649 + }, + { + "epoch": 0.300462249614792, + "grad_norm": 2.017837554179133, + "learning_rate": 8.201314327717829e-06, + "loss": 0.3622, + "step": 13650 + }, + { + "epoch": 0.30048426150121066, + "grad_norm": 2.08567655398171, + "learning_rate": 8.201040505517431e-06, + "loss": 0.5568, + "step": 13651 + }, + { + "epoch": 0.3005062733876293, + "grad_norm": 1.984609535838269, + "learning_rate": 8.200766667047917e-06, + "loss": 0.3409, + "step": 13652 + }, + { + "epoch": 0.300528285274048, + "grad_norm": 1.3226595680647604, + "learning_rate": 8.200492812310678e-06, + "loss": 0.2739, + "step": 13653 + }, + { + "epoch": 0.30055029716046666, + "grad_norm": 1.434987955475709, + "learning_rate": 8.20021894130711e-06, + "loss": 0.4719, + "step": 13654 + }, + { + "epoch": 0.3005723090468853, + "grad_norm": 0.6233460980594231, + "learning_rate": 8.199945054038599e-06, + "loss": 0.4992, + "step": 13655 + }, + { + "epoch": 0.30059432093330396, + "grad_norm": 1.7713273474623579, + "learning_rate": 8.199671150506543e-06, + "loss": 0.4026, + "step": 13656 + }, + { + "epoch": 0.30061633281972266, + "grad_norm": 1.4927871933239696, + "learning_rate": 8.199397230712329e-06, + "loss": 0.3935, + "step": 13657 + }, + { + "epoch": 0.3006383447061413, + "grad_norm": 3.064441744316796, + "learning_rate": 8.199123294657352e-06, + "loss": 0.4026, + "step": 13658 + }, + { + "epoch": 0.30066035659255996, + "grad_norm": 1.2929726972279802, + "learning_rate": 8.198849342343003e-06, + "loss": 0.2934, + "step": 13659 + }, + { + "epoch": 0.30068236847897867, + "grad_norm": 1.7178892231862073, + "learning_rate": 8.198575373770676e-06, + "loss": 0.4006, + "step": 13660 + }, + { + "epoch": 0.3007043803653973, + "grad_norm": 1.2020520649735256, + "learning_rate": 8.198301388941761e-06, + "loss": 0.3785, + "step": 13661 + }, + { + "epoch": 0.30072639225181597, + "grad_norm": 1.2743775831774404, + "learning_rate": 8.198027387857653e-06, + "loss": 0.2463, + "step": 13662 + }, + { + "epoch": 0.30074840413823467, + "grad_norm": 0.6467579662130463, + "learning_rate": 8.197753370519743e-06, + "loss": 0.48, + "step": 13663 + }, + { + "epoch": 0.3007704160246533, + "grad_norm": 1.8373563382918938, + "learning_rate": 8.197479336929423e-06, + "loss": 0.3063, + "step": 13664 + }, + { + "epoch": 0.30079242791107197, + "grad_norm": 1.6053478907194556, + "learning_rate": 8.197205287088091e-06, + "loss": 0.3094, + "step": 13665 + }, + { + "epoch": 0.3008144397974906, + "grad_norm": 1.362842630835454, + "learning_rate": 8.196931220997131e-06, + "loss": 0.2911, + "step": 13666 + }, + { + "epoch": 0.3008364516839093, + "grad_norm": 1.234481799536907, + "learning_rate": 8.196657138657941e-06, + "loss": 0.3627, + "step": 13667 + }, + { + "epoch": 0.30085846357032797, + "grad_norm": 1.3954484691967521, + "learning_rate": 8.196383040071916e-06, + "loss": 0.2977, + "step": 13668 + }, + { + "epoch": 0.3008804754567466, + "grad_norm": 2.4195269816681724, + "learning_rate": 8.196108925240448e-06, + "loss": 0.3002, + "step": 13669 + }, + { + "epoch": 0.3009024873431653, + "grad_norm": 1.6017555812153614, + "learning_rate": 8.195834794164925e-06, + "loss": 0.3037, + "step": 13670 + }, + { + "epoch": 0.300924499229584, + "grad_norm": 0.6342082075255794, + "learning_rate": 8.195560646846748e-06, + "loss": 0.4748, + "step": 13671 + }, + { + "epoch": 0.3009465111160026, + "grad_norm": 1.823146330450735, + "learning_rate": 8.195286483287304e-06, + "loss": 0.3753, + "step": 13672 + }, + { + "epoch": 0.30096852300242133, + "grad_norm": 1.299865678152396, + "learning_rate": 8.195012303487988e-06, + "loss": 0.3941, + "step": 13673 + }, + { + "epoch": 0.30099053488884, + "grad_norm": 1.5405044718557033, + "learning_rate": 8.194738107450197e-06, + "loss": 0.3166, + "step": 13674 + }, + { + "epoch": 0.3010125467752586, + "grad_norm": 1.944376414200885, + "learning_rate": 8.194463895175321e-06, + "loss": 0.3013, + "step": 13675 + }, + { + "epoch": 0.30103455866167733, + "grad_norm": 2.5680417263155317, + "learning_rate": 8.194189666664756e-06, + "loss": 0.3491, + "step": 13676 + }, + { + "epoch": 0.301056570548096, + "grad_norm": 2.270456849548989, + "learning_rate": 8.193915421919892e-06, + "loss": 0.3082, + "step": 13677 + }, + { + "epoch": 0.30107858243451463, + "grad_norm": 11.985373637528248, + "learning_rate": 8.193641160942126e-06, + "loss": 0.3095, + "step": 13678 + }, + { + "epoch": 0.3011005943209333, + "grad_norm": 1.8853123700205783, + "learning_rate": 8.193366883732853e-06, + "loss": 0.3031, + "step": 13679 + }, + { + "epoch": 0.301122606207352, + "grad_norm": 1.6080994403937048, + "learning_rate": 8.193092590293464e-06, + "loss": 0.3309, + "step": 13680 + }, + { + "epoch": 0.30114461809377063, + "grad_norm": 1.2414996141307688, + "learning_rate": 8.192818280625353e-06, + "loss": 0.2546, + "step": 13681 + }, + { + "epoch": 0.3011666299801893, + "grad_norm": 1.328559408361698, + "learning_rate": 8.192543954729916e-06, + "loss": 0.3516, + "step": 13682 + }, + { + "epoch": 0.301188641866608, + "grad_norm": 1.2751333394192015, + "learning_rate": 8.192269612608547e-06, + "loss": 0.3345, + "step": 13683 + }, + { + "epoch": 0.30121065375302664, + "grad_norm": 2.5630824397924057, + "learning_rate": 8.191995254262642e-06, + "loss": 0.3732, + "step": 13684 + }, + { + "epoch": 0.3012326656394453, + "grad_norm": 0.645206569604096, + "learning_rate": 8.191720879693592e-06, + "loss": 0.4727, + "step": 13685 + }, + { + "epoch": 0.301254677525864, + "grad_norm": 1.3059315640071043, + "learning_rate": 8.191446488902792e-06, + "loss": 0.4045, + "step": 13686 + }, + { + "epoch": 0.30127668941228264, + "grad_norm": 2.398858493447469, + "learning_rate": 8.191172081891639e-06, + "loss": 0.4043, + "step": 13687 + }, + { + "epoch": 0.3012987012987013, + "grad_norm": 2.4447379667222195, + "learning_rate": 8.190897658661526e-06, + "loss": 0.4127, + "step": 13688 + }, + { + "epoch": 0.30132071318511994, + "grad_norm": 5.529302240502822, + "learning_rate": 8.190623219213847e-06, + "loss": 0.3762, + "step": 13689 + }, + { + "epoch": 0.30134272507153864, + "grad_norm": 1.523668258940484, + "learning_rate": 8.190348763549998e-06, + "loss": 0.2495, + "step": 13690 + }, + { + "epoch": 0.3013647369579573, + "grad_norm": 1.5300203012796918, + "learning_rate": 8.190074291671374e-06, + "loss": 0.2644, + "step": 13691 + }, + { + "epoch": 0.30138674884437594, + "grad_norm": 1.526817592319795, + "learning_rate": 8.189799803579369e-06, + "loss": 0.349, + "step": 13692 + }, + { + "epoch": 0.30140876073079464, + "grad_norm": 1.4242767547978492, + "learning_rate": 8.18952529927538e-06, + "loss": 0.2844, + "step": 13693 + }, + { + "epoch": 0.3014307726172133, + "grad_norm": 1.3430581817004343, + "learning_rate": 8.1892507787608e-06, + "loss": 0.327, + "step": 13694 + }, + { + "epoch": 0.30145278450363194, + "grad_norm": 1.902407087475894, + "learning_rate": 8.188976242037024e-06, + "loss": 0.3837, + "step": 13695 + }, + { + "epoch": 0.30147479639005065, + "grad_norm": 2.1157022799779197, + "learning_rate": 8.188701689105451e-06, + "loss": 0.3214, + "step": 13696 + }, + { + "epoch": 0.3014968082764693, + "grad_norm": 1.911998349742254, + "learning_rate": 8.188427119967472e-06, + "loss": 0.2791, + "step": 13697 + }, + { + "epoch": 0.30151882016288795, + "grad_norm": 1.8243967735785434, + "learning_rate": 8.188152534624485e-06, + "loss": 0.4099, + "step": 13698 + }, + { + "epoch": 0.30154083204930665, + "grad_norm": 2.464556005246787, + "learning_rate": 8.187877933077883e-06, + "loss": 0.4691, + "step": 13699 + }, + { + "epoch": 0.3015628439357253, + "grad_norm": 1.6854043432108021, + "learning_rate": 8.187603315329066e-06, + "loss": 0.3458, + "step": 13700 + }, + { + "epoch": 0.30158485582214395, + "grad_norm": 1.5569466280789779, + "learning_rate": 8.187328681379426e-06, + "loss": 0.4272, + "step": 13701 + }, + { + "epoch": 0.3016068677085626, + "grad_norm": 1.2162239862293818, + "learning_rate": 8.187054031230361e-06, + "loss": 0.2349, + "step": 13702 + }, + { + "epoch": 0.3016288795949813, + "grad_norm": 1.6959692802836241, + "learning_rate": 8.186779364883264e-06, + "loss": 0.3608, + "step": 13703 + }, + { + "epoch": 0.30165089148139995, + "grad_norm": 4.061748476031059, + "learning_rate": 8.186504682339535e-06, + "loss": 0.3266, + "step": 13704 + }, + { + "epoch": 0.3016729033678186, + "grad_norm": 1.2460963715894686, + "learning_rate": 8.186229983600566e-06, + "loss": 0.2916, + "step": 13705 + }, + { + "epoch": 0.3016949152542373, + "grad_norm": 1.3028766091927897, + "learning_rate": 8.185955268667758e-06, + "loss": 0.4182, + "step": 13706 + }, + { + "epoch": 0.30171692714065595, + "grad_norm": 1.4632064618291325, + "learning_rate": 8.185680537542501e-06, + "loss": 0.3594, + "step": 13707 + }, + { + "epoch": 0.3017389390270746, + "grad_norm": 1.4379891762578851, + "learning_rate": 8.185405790226198e-06, + "loss": 0.341, + "step": 13708 + }, + { + "epoch": 0.3017609509134933, + "grad_norm": 1.8172671834443495, + "learning_rate": 8.18513102672024e-06, + "loss": 0.3669, + "step": 13709 + }, + { + "epoch": 0.30178296279991196, + "grad_norm": 2.184144784229234, + "learning_rate": 8.184856247026025e-06, + "loss": 0.3266, + "step": 13710 + }, + { + "epoch": 0.3018049746863306, + "grad_norm": 1.8401735568053021, + "learning_rate": 8.184581451144953e-06, + "loss": 0.3557, + "step": 13711 + }, + { + "epoch": 0.3018269865727493, + "grad_norm": 1.839155654240508, + "learning_rate": 8.184306639078414e-06, + "loss": 0.3451, + "step": 13712 + }, + { + "epoch": 0.30184899845916796, + "grad_norm": 2.666285802725001, + "learning_rate": 8.184031810827811e-06, + "loss": 0.2849, + "step": 13713 + }, + { + "epoch": 0.3018710103455866, + "grad_norm": 2.0352582448436274, + "learning_rate": 8.183756966394537e-06, + "loss": 0.2985, + "step": 13714 + }, + { + "epoch": 0.30189302223200526, + "grad_norm": 0.6589289669783625, + "learning_rate": 8.183482105779991e-06, + "loss": 0.53, + "step": 13715 + }, + { + "epoch": 0.30191503411842396, + "grad_norm": 1.4697352463512656, + "learning_rate": 8.183207228985568e-06, + "loss": 0.2909, + "step": 13716 + }, + { + "epoch": 0.3019370460048426, + "grad_norm": 1.7597895202188611, + "learning_rate": 8.182932336012666e-06, + "loss": 0.2975, + "step": 13717 + }, + { + "epoch": 0.30195905789126126, + "grad_norm": 1.3189097865553847, + "learning_rate": 8.182657426862684e-06, + "loss": 0.3389, + "step": 13718 + }, + { + "epoch": 0.30198106977767997, + "grad_norm": 1.6433378762564979, + "learning_rate": 8.182382501537014e-06, + "loss": 0.4495, + "step": 13719 + }, + { + "epoch": 0.3020030816640986, + "grad_norm": 1.4033665536374054, + "learning_rate": 8.18210756003706e-06, + "loss": 0.341, + "step": 13720 + }, + { + "epoch": 0.30202509355051727, + "grad_norm": 2.007183159372799, + "learning_rate": 8.181832602364215e-06, + "loss": 0.3275, + "step": 13721 + }, + { + "epoch": 0.30204710543693597, + "grad_norm": 1.7648026858711785, + "learning_rate": 8.181557628519879e-06, + "loss": 0.408, + "step": 13722 + }, + { + "epoch": 0.3020691173233546, + "grad_norm": 1.385317353990623, + "learning_rate": 8.181282638505446e-06, + "loss": 0.3452, + "step": 13723 + }, + { + "epoch": 0.30209112920977327, + "grad_norm": 2.1606702506398148, + "learning_rate": 8.181007632322317e-06, + "loss": 0.4134, + "step": 13724 + }, + { + "epoch": 0.3021131410961919, + "grad_norm": 1.8539153457362936, + "learning_rate": 8.180732609971888e-06, + "loss": 0.4271, + "step": 13725 + }, + { + "epoch": 0.3021351529826106, + "grad_norm": 1.3015055221053045, + "learning_rate": 8.180457571455557e-06, + "loss": 0.2987, + "step": 13726 + }, + { + "epoch": 0.30215716486902927, + "grad_norm": 1.6177949558448717, + "learning_rate": 8.180182516774722e-06, + "loss": 0.3352, + "step": 13727 + }, + { + "epoch": 0.3021791767554479, + "grad_norm": 1.8754192584541212, + "learning_rate": 8.179907445930782e-06, + "loss": 0.3781, + "step": 13728 + }, + { + "epoch": 0.3022011886418666, + "grad_norm": 1.5086030549724483, + "learning_rate": 8.179632358925134e-06, + "loss": 0.3679, + "step": 13729 + }, + { + "epoch": 0.3022232005282853, + "grad_norm": 1.6961762709073331, + "learning_rate": 8.179357255759175e-06, + "loss": 0.3938, + "step": 13730 + }, + { + "epoch": 0.3022452124147039, + "grad_norm": 1.663461653380182, + "learning_rate": 8.179082136434304e-06, + "loss": 0.361, + "step": 13731 + }, + { + "epoch": 0.30226722430112263, + "grad_norm": 1.2870537845559409, + "learning_rate": 8.178807000951922e-06, + "loss": 0.3186, + "step": 13732 + }, + { + "epoch": 0.3022892361875413, + "grad_norm": 1.3099501231750321, + "learning_rate": 8.178531849313422e-06, + "loss": 0.2489, + "step": 13733 + }, + { + "epoch": 0.3023112480739599, + "grad_norm": 1.9992002162249618, + "learning_rate": 8.17825668152021e-06, + "loss": 0.3962, + "step": 13734 + }, + { + "epoch": 0.30233325996037863, + "grad_norm": 0.6778998764348569, + "learning_rate": 8.177981497573676e-06, + "loss": 0.5158, + "step": 13735 + }, + { + "epoch": 0.3023552718467973, + "grad_norm": 29.166745801663776, + "learning_rate": 8.177706297475223e-06, + "loss": 0.3134, + "step": 13736 + }, + { + "epoch": 0.30237728373321593, + "grad_norm": 0.6250515945841622, + "learning_rate": 8.17743108122625e-06, + "loss": 0.5327, + "step": 13737 + }, + { + "epoch": 0.3023992956196346, + "grad_norm": 1.3964065002887427, + "learning_rate": 8.177155848828156e-06, + "loss": 0.3747, + "step": 13738 + }, + { + "epoch": 0.3024213075060533, + "grad_norm": 1.8588729479754573, + "learning_rate": 8.17688060028234e-06, + "loss": 0.3522, + "step": 13739 + }, + { + "epoch": 0.30244331939247193, + "grad_norm": 1.357951225851102, + "learning_rate": 8.176605335590196e-06, + "loss": 0.2689, + "step": 13740 + }, + { + "epoch": 0.3024653312788906, + "grad_norm": 1.6723812192022083, + "learning_rate": 8.17633005475313e-06, + "loss": 0.3712, + "step": 13741 + }, + { + "epoch": 0.3024873431653093, + "grad_norm": 1.556606890016012, + "learning_rate": 8.176054757772536e-06, + "loss": 0.263, + "step": 13742 + }, + { + "epoch": 0.30250935505172794, + "grad_norm": 1.4718195566175571, + "learning_rate": 8.175779444649817e-06, + "loss": 0.3563, + "step": 13743 + }, + { + "epoch": 0.3025313669381466, + "grad_norm": 1.6031677613921158, + "learning_rate": 8.175504115386368e-06, + "loss": 0.3638, + "step": 13744 + }, + { + "epoch": 0.3025533788245653, + "grad_norm": 1.1606069419808789, + "learning_rate": 8.175228769983592e-06, + "loss": 0.2492, + "step": 13745 + }, + { + "epoch": 0.30257539071098394, + "grad_norm": 0.787069847600699, + "learning_rate": 8.174953408442889e-06, + "loss": 0.5209, + "step": 13746 + }, + { + "epoch": 0.3025974025974026, + "grad_norm": 1.6293818901354677, + "learning_rate": 8.174678030765654e-06, + "loss": 0.4852, + "step": 13747 + }, + { + "epoch": 0.30261941448382124, + "grad_norm": 0.6533405144315105, + "learning_rate": 8.174402636953292e-06, + "loss": 0.4892, + "step": 13748 + }, + { + "epoch": 0.30264142637023994, + "grad_norm": 1.6367978700040082, + "learning_rate": 8.174127227007197e-06, + "loss": 0.3986, + "step": 13749 + }, + { + "epoch": 0.3026634382566586, + "grad_norm": 1.5264354343619653, + "learning_rate": 8.173851800928774e-06, + "loss": 0.3819, + "step": 13750 + }, + { + "epoch": 0.30268545014307724, + "grad_norm": 1.425302421030772, + "learning_rate": 8.173576358719419e-06, + "loss": 0.3542, + "step": 13751 + }, + { + "epoch": 0.30270746202949594, + "grad_norm": 1.5467544110992657, + "learning_rate": 8.173300900380534e-06, + "loss": 0.3759, + "step": 13752 + }, + { + "epoch": 0.3027294739159146, + "grad_norm": 0.7529237120305846, + "learning_rate": 8.173025425913519e-06, + "loss": 0.5226, + "step": 13753 + }, + { + "epoch": 0.30275148580233324, + "grad_norm": 1.2508275326923586, + "learning_rate": 8.172749935319772e-06, + "loss": 0.2735, + "step": 13754 + }, + { + "epoch": 0.30277349768875195, + "grad_norm": 4.859264532786211, + "learning_rate": 8.172474428600698e-06, + "loss": 0.3076, + "step": 13755 + }, + { + "epoch": 0.3027955095751706, + "grad_norm": 1.568231996995364, + "learning_rate": 8.172198905757691e-06, + "loss": 0.4159, + "step": 13756 + }, + { + "epoch": 0.30281752146158925, + "grad_norm": 0.6954201882239508, + "learning_rate": 8.171923366792156e-06, + "loss": 0.5015, + "step": 13757 + }, + { + "epoch": 0.30283953334800795, + "grad_norm": 1.2635030826494489, + "learning_rate": 8.17164781170549e-06, + "loss": 0.2854, + "step": 13758 + }, + { + "epoch": 0.3028615452344266, + "grad_norm": 1.361728184153739, + "learning_rate": 8.171372240499098e-06, + "loss": 0.4155, + "step": 13759 + }, + { + "epoch": 0.30288355712084525, + "grad_norm": 1.6082471930661932, + "learning_rate": 8.171096653174374e-06, + "loss": 0.3817, + "step": 13760 + }, + { + "epoch": 0.3029055690072639, + "grad_norm": 0.6149386722697774, + "learning_rate": 8.170821049732727e-06, + "loss": 0.5051, + "step": 13761 + }, + { + "epoch": 0.3029275808936826, + "grad_norm": 1.4083027216860171, + "learning_rate": 8.170545430175549e-06, + "loss": 0.3394, + "step": 13762 + }, + { + "epoch": 0.30294959278010125, + "grad_norm": 1.2478157087517119, + "learning_rate": 8.170269794504247e-06, + "loss": 0.3075, + "step": 13763 + }, + { + "epoch": 0.3029716046665199, + "grad_norm": 1.2758979482973456, + "learning_rate": 8.169994142720218e-06, + "loss": 0.3231, + "step": 13764 + }, + { + "epoch": 0.3029936165529386, + "grad_norm": 1.7148863345872727, + "learning_rate": 8.169718474824868e-06, + "loss": 0.4594, + "step": 13765 + }, + { + "epoch": 0.30301562843935725, + "grad_norm": 0.684532244634393, + "learning_rate": 8.169442790819596e-06, + "loss": 0.5059, + "step": 13766 + }, + { + "epoch": 0.3030376403257759, + "grad_norm": 0.666601397751782, + "learning_rate": 8.169167090705798e-06, + "loss": 0.54, + "step": 13767 + }, + { + "epoch": 0.3030596522121946, + "grad_norm": 1.6967917782305373, + "learning_rate": 8.168891374484882e-06, + "loss": 0.3271, + "step": 13768 + }, + { + "epoch": 0.30308166409861326, + "grad_norm": 1.3866579172041604, + "learning_rate": 8.168615642158243e-06, + "loss": 0.3281, + "step": 13769 + }, + { + "epoch": 0.3031036759850319, + "grad_norm": 1.5909499869355628, + "learning_rate": 8.168339893727291e-06, + "loss": 0.3619, + "step": 13770 + }, + { + "epoch": 0.3031256878714506, + "grad_norm": 1.7647355323609402, + "learning_rate": 8.168064129193421e-06, + "loss": 0.422, + "step": 13771 + }, + { + "epoch": 0.30314769975786926, + "grad_norm": 1.7599138192747534, + "learning_rate": 8.167788348558034e-06, + "loss": 0.2963, + "step": 13772 + }, + { + "epoch": 0.3031697116442879, + "grad_norm": 3.3787517911982756, + "learning_rate": 8.167512551822535e-06, + "loss": 0.3738, + "step": 13773 + }, + { + "epoch": 0.30319172353070656, + "grad_norm": 0.7316972637514717, + "learning_rate": 8.167236738988325e-06, + "loss": 0.5061, + "step": 13774 + }, + { + "epoch": 0.30321373541712526, + "grad_norm": 1.426973030171086, + "learning_rate": 8.166960910056805e-06, + "loss": 0.3875, + "step": 13775 + }, + { + "epoch": 0.3032357473035439, + "grad_norm": 0.6887585493951577, + "learning_rate": 8.166685065029376e-06, + "loss": 0.5349, + "step": 13776 + }, + { + "epoch": 0.30325775918996256, + "grad_norm": 1.280504988390693, + "learning_rate": 8.166409203907444e-06, + "loss": 0.3633, + "step": 13777 + }, + { + "epoch": 0.30327977107638127, + "grad_norm": 0.5994744540625593, + "learning_rate": 8.166133326692405e-06, + "loss": 0.4955, + "step": 13778 + }, + { + "epoch": 0.3033017829627999, + "grad_norm": 1.3737463981309888, + "learning_rate": 8.165857433385666e-06, + "loss": 0.2593, + "step": 13779 + }, + { + "epoch": 0.30332379484921856, + "grad_norm": 1.551398515925954, + "learning_rate": 8.165581523988627e-06, + "loss": 0.4258, + "step": 13780 + }, + { + "epoch": 0.30334580673563727, + "grad_norm": 1.8514376348137285, + "learning_rate": 8.165305598502691e-06, + "loss": 0.3934, + "step": 13781 + }, + { + "epoch": 0.3033678186220559, + "grad_norm": 1.474425063890465, + "learning_rate": 8.16502965692926e-06, + "loss": 0.3581, + "step": 13782 + }, + { + "epoch": 0.30338983050847457, + "grad_norm": 1.801066035117556, + "learning_rate": 8.164753699269737e-06, + "loss": 0.4683, + "step": 13783 + }, + { + "epoch": 0.3034118423948932, + "grad_norm": 1.2949865344194846, + "learning_rate": 8.164477725525524e-06, + "loss": 0.2692, + "step": 13784 + }, + { + "epoch": 0.3034338542813119, + "grad_norm": 9.417592562964437, + "learning_rate": 8.164201735698023e-06, + "loss": 0.2359, + "step": 13785 + }, + { + "epoch": 0.30345586616773057, + "grad_norm": 1.3703361538830257, + "learning_rate": 8.16392572978864e-06, + "loss": 0.4205, + "step": 13786 + }, + { + "epoch": 0.3034778780541492, + "grad_norm": 1.4908193412956063, + "learning_rate": 8.163649707798772e-06, + "loss": 0.4242, + "step": 13787 + }, + { + "epoch": 0.3034998899405679, + "grad_norm": 1.7186806641571042, + "learning_rate": 8.163373669729829e-06, + "loss": 0.3209, + "step": 13788 + }, + { + "epoch": 0.3035219018269866, + "grad_norm": 1.0732885749419512, + "learning_rate": 8.163097615583207e-06, + "loss": 0.2608, + "step": 13789 + }, + { + "epoch": 0.3035439137134052, + "grad_norm": 1.2892541017230477, + "learning_rate": 8.162821545360313e-06, + "loss": 0.2944, + "step": 13790 + }, + { + "epoch": 0.3035659255998239, + "grad_norm": 1.7479337762305205, + "learning_rate": 8.162545459062548e-06, + "loss": 0.2658, + "step": 13791 + }, + { + "epoch": 0.3035879374862426, + "grad_norm": 1.7204581209514918, + "learning_rate": 8.162269356691319e-06, + "loss": 0.3203, + "step": 13792 + }, + { + "epoch": 0.3036099493726612, + "grad_norm": 2.0102913977218524, + "learning_rate": 8.161993238248025e-06, + "loss": 0.3643, + "step": 13793 + }, + { + "epoch": 0.30363196125907993, + "grad_norm": 1.6388775001583897, + "learning_rate": 8.16171710373407e-06, + "loss": 0.3566, + "step": 13794 + }, + { + "epoch": 0.3036539731454986, + "grad_norm": 1.3485517408249668, + "learning_rate": 8.161440953150862e-06, + "loss": 0.3251, + "step": 13795 + }, + { + "epoch": 0.30367598503191723, + "grad_norm": 1.7841618709833027, + "learning_rate": 8.161164786499798e-06, + "loss": 0.3687, + "step": 13796 + }, + { + "epoch": 0.3036979969183359, + "grad_norm": 1.580245785612483, + "learning_rate": 8.160888603782287e-06, + "loss": 0.3569, + "step": 13797 + }, + { + "epoch": 0.3037200088047546, + "grad_norm": 1.510922164080235, + "learning_rate": 8.160612404999728e-06, + "loss": 0.3847, + "step": 13798 + }, + { + "epoch": 0.30374202069117323, + "grad_norm": 1.6379287646617209, + "learning_rate": 8.160336190153529e-06, + "loss": 0.3577, + "step": 13799 + }, + { + "epoch": 0.3037640325775919, + "grad_norm": 1.390557105991813, + "learning_rate": 8.16005995924509e-06, + "loss": 0.2561, + "step": 13800 + }, + { + "epoch": 0.3037860444640106, + "grad_norm": 2.5173115054369775, + "learning_rate": 8.159783712275817e-06, + "loss": 0.409, + "step": 13801 + }, + { + "epoch": 0.30380805635042923, + "grad_norm": 1.750453186975513, + "learning_rate": 8.159507449247115e-06, + "loss": 0.4012, + "step": 13802 + }, + { + "epoch": 0.3038300682368479, + "grad_norm": 1.6521445532398165, + "learning_rate": 8.159231170160387e-06, + "loss": 0.3326, + "step": 13803 + }, + { + "epoch": 0.3038520801232666, + "grad_norm": 1.3262347481153047, + "learning_rate": 8.158954875017037e-06, + "loss": 0.2904, + "step": 13804 + }, + { + "epoch": 0.30387409200968524, + "grad_norm": 1.3916302452801108, + "learning_rate": 8.158678563818469e-06, + "loss": 0.3225, + "step": 13805 + }, + { + "epoch": 0.3038961038961039, + "grad_norm": 1.4055792358388823, + "learning_rate": 8.158402236566087e-06, + "loss": 0.3147, + "step": 13806 + }, + { + "epoch": 0.30391811578252254, + "grad_norm": 1.0277914640962473, + "learning_rate": 8.158125893261297e-06, + "loss": 0.547, + "step": 13807 + }, + { + "epoch": 0.30394012766894124, + "grad_norm": 1.38915839379006, + "learning_rate": 8.157849533905505e-06, + "loss": 0.4024, + "step": 13808 + }, + { + "epoch": 0.3039621395553599, + "grad_norm": 1.9733680110140406, + "learning_rate": 8.15757315850011e-06, + "loss": 0.4086, + "step": 13809 + }, + { + "epoch": 0.30398415144177854, + "grad_norm": 1.7927400017175958, + "learning_rate": 8.15729676704652e-06, + "loss": 0.3206, + "step": 13810 + }, + { + "epoch": 0.30400616332819724, + "grad_norm": 1.2608004785228313, + "learning_rate": 8.157020359546141e-06, + "loss": 0.2758, + "step": 13811 + }, + { + "epoch": 0.3040281752146159, + "grad_norm": 1.1768778872979244, + "learning_rate": 8.156743936000376e-06, + "loss": 0.3525, + "step": 13812 + }, + { + "epoch": 0.30405018710103454, + "grad_norm": 1.274776772783523, + "learning_rate": 8.15646749641063e-06, + "loss": 0.2918, + "step": 13813 + }, + { + "epoch": 0.30407219898745325, + "grad_norm": 1.3805318581959323, + "learning_rate": 8.156191040778308e-06, + "loss": 0.3427, + "step": 13814 + }, + { + "epoch": 0.3040942108738719, + "grad_norm": 1.5930270043903834, + "learning_rate": 8.155914569104818e-06, + "loss": 0.3597, + "step": 13815 + }, + { + "epoch": 0.30411622276029054, + "grad_norm": 2.358129507204255, + "learning_rate": 8.15563808139156e-06, + "loss": 0.4086, + "step": 13816 + }, + { + "epoch": 0.30413823464670925, + "grad_norm": 1.2164949026747147, + "learning_rate": 8.155361577639942e-06, + "loss": 0.3005, + "step": 13817 + }, + { + "epoch": 0.3041602465331279, + "grad_norm": 2.1929528449590987, + "learning_rate": 8.15508505785137e-06, + "loss": 0.4442, + "step": 13818 + }, + { + "epoch": 0.30418225841954655, + "grad_norm": 4.888446586428601, + "learning_rate": 8.154808522027246e-06, + "loss": 0.3042, + "step": 13819 + }, + { + "epoch": 0.3042042703059652, + "grad_norm": 1.58872845376221, + "learning_rate": 8.154531970168982e-06, + "loss": 0.356, + "step": 13820 + }, + { + "epoch": 0.3042262821923839, + "grad_norm": 2.2475316138441763, + "learning_rate": 8.154255402277977e-06, + "loss": 0.3036, + "step": 13821 + }, + { + "epoch": 0.30424829407880255, + "grad_norm": 0.873720869608486, + "learning_rate": 8.15397881835564e-06, + "loss": 0.5081, + "step": 13822 + }, + { + "epoch": 0.3042703059652212, + "grad_norm": 0.8289917579586707, + "learning_rate": 8.153702218403376e-06, + "loss": 0.5233, + "step": 13823 + }, + { + "epoch": 0.3042923178516399, + "grad_norm": 1.5356680620353214, + "learning_rate": 8.15342560242259e-06, + "loss": 0.2708, + "step": 13824 + }, + { + "epoch": 0.30431432973805855, + "grad_norm": 2.022045392006791, + "learning_rate": 8.15314897041469e-06, + "loss": 0.3732, + "step": 13825 + }, + { + "epoch": 0.3043363416244772, + "grad_norm": 0.6645730240206003, + "learning_rate": 8.15287232238108e-06, + "loss": 0.5137, + "step": 13826 + }, + { + "epoch": 0.3043583535108959, + "grad_norm": 1.8493595091803403, + "learning_rate": 8.152595658323167e-06, + "loss": 0.3796, + "step": 13827 + }, + { + "epoch": 0.30438036539731456, + "grad_norm": 1.522515752733989, + "learning_rate": 8.152318978242355e-06, + "loss": 0.3335, + "step": 13828 + }, + { + "epoch": 0.3044023772837332, + "grad_norm": 2.026510689201525, + "learning_rate": 8.152042282140053e-06, + "loss": 0.3602, + "step": 13829 + }, + { + "epoch": 0.30442438917015185, + "grad_norm": 1.5466293520284338, + "learning_rate": 8.151765570017667e-06, + "loss": 0.3347, + "step": 13830 + }, + { + "epoch": 0.30444640105657056, + "grad_norm": 5.029023275443027, + "learning_rate": 8.151488841876602e-06, + "loss": 0.3562, + "step": 13831 + }, + { + "epoch": 0.3044684129429892, + "grad_norm": 0.914493594332317, + "learning_rate": 8.151212097718266e-06, + "loss": 0.5475, + "step": 13832 + }, + { + "epoch": 0.30449042482940786, + "grad_norm": 3.1816149104001656, + "learning_rate": 8.150935337544064e-06, + "loss": 0.3119, + "step": 13833 + }, + { + "epoch": 0.30451243671582656, + "grad_norm": 2.5276787513089047, + "learning_rate": 8.150658561355402e-06, + "loss": 0.3811, + "step": 13834 + }, + { + "epoch": 0.3045344486022452, + "grad_norm": 1.4939048989563695, + "learning_rate": 8.150381769153691e-06, + "loss": 0.3419, + "step": 13835 + }, + { + "epoch": 0.30455646048866386, + "grad_norm": 0.6502710291191304, + "learning_rate": 8.150104960940331e-06, + "loss": 0.4932, + "step": 13836 + }, + { + "epoch": 0.30457847237508257, + "grad_norm": 1.2017697855794083, + "learning_rate": 8.149828136716735e-06, + "loss": 0.3806, + "step": 13837 + }, + { + "epoch": 0.3046004842615012, + "grad_norm": 1.3332070985574456, + "learning_rate": 8.149551296484307e-06, + "loss": 0.3649, + "step": 13838 + }, + { + "epoch": 0.30462249614791986, + "grad_norm": 4.991000563979073, + "learning_rate": 8.149274440244455e-06, + "loss": 0.3249, + "step": 13839 + }, + { + "epoch": 0.30464450803433857, + "grad_norm": 1.6344987923559144, + "learning_rate": 8.148997567998585e-06, + "loss": 0.3276, + "step": 13840 + }, + { + "epoch": 0.3046665199207572, + "grad_norm": 0.7695770344375269, + "learning_rate": 8.148720679748106e-06, + "loss": 0.5382, + "step": 13841 + }, + { + "epoch": 0.30468853180717587, + "grad_norm": 1.5868822553582718, + "learning_rate": 8.14844377549442e-06, + "loss": 0.3032, + "step": 13842 + }, + { + "epoch": 0.3047105436935945, + "grad_norm": 0.7203910729523223, + "learning_rate": 8.148166855238945e-06, + "loss": 0.5343, + "step": 13843 + }, + { + "epoch": 0.3047325555800132, + "grad_norm": 1.4272049765195163, + "learning_rate": 8.147889918983079e-06, + "loss": 0.3717, + "step": 13844 + }, + { + "epoch": 0.30475456746643187, + "grad_norm": 1.6071242661443892, + "learning_rate": 8.147612966728231e-06, + "loss": 0.3276, + "step": 13845 + }, + { + "epoch": 0.3047765793528505, + "grad_norm": 1.8521935857186937, + "learning_rate": 8.147335998475811e-06, + "loss": 0.3586, + "step": 13846 + }, + { + "epoch": 0.3047985912392692, + "grad_norm": 1.3065478438145797, + "learning_rate": 8.147059014227226e-06, + "loss": 0.2234, + "step": 13847 + }, + { + "epoch": 0.3048206031256879, + "grad_norm": 0.6006073400051365, + "learning_rate": 8.146782013983883e-06, + "loss": 0.486, + "step": 13848 + }, + { + "epoch": 0.3048426150121065, + "grad_norm": 1.954006391807672, + "learning_rate": 8.146504997747192e-06, + "loss": 0.3989, + "step": 13849 + }, + { + "epoch": 0.3048646268985252, + "grad_norm": 1.6936699248901708, + "learning_rate": 8.146227965518556e-06, + "loss": 0.3216, + "step": 13850 + }, + { + "epoch": 0.3048866387849439, + "grad_norm": 1.503217645870855, + "learning_rate": 8.145950917299388e-06, + "loss": 0.2976, + "step": 13851 + }, + { + "epoch": 0.3049086506713625, + "grad_norm": 1.3540545275792575, + "learning_rate": 8.145673853091095e-06, + "loss": 0.397, + "step": 13852 + }, + { + "epoch": 0.30493066255778123, + "grad_norm": 1.4562985253742509, + "learning_rate": 8.145396772895083e-06, + "loss": 0.342, + "step": 13853 + }, + { + "epoch": 0.3049526744441999, + "grad_norm": 1.4442380470008738, + "learning_rate": 8.145119676712763e-06, + "loss": 0.3173, + "step": 13854 + }, + { + "epoch": 0.3049746863306185, + "grad_norm": 1.5584859636525128, + "learning_rate": 8.14484256454554e-06, + "loss": 0.3537, + "step": 13855 + }, + { + "epoch": 0.3049966982170372, + "grad_norm": 0.6840141931293867, + "learning_rate": 8.144565436394826e-06, + "loss": 0.5133, + "step": 13856 + }, + { + "epoch": 0.3050187101034559, + "grad_norm": 2.2005942343955827, + "learning_rate": 8.14428829226203e-06, + "loss": 0.3286, + "step": 13857 + }, + { + "epoch": 0.30504072198987453, + "grad_norm": 1.7831436190285428, + "learning_rate": 8.144011132148555e-06, + "loss": 0.3721, + "step": 13858 + }, + { + "epoch": 0.3050627338762932, + "grad_norm": 2.5913471793415344, + "learning_rate": 8.143733956055815e-06, + "loss": 0.2506, + "step": 13859 + }, + { + "epoch": 0.3050847457627119, + "grad_norm": 1.4096358566283063, + "learning_rate": 8.143456763985216e-06, + "loss": 0.3598, + "step": 13860 + }, + { + "epoch": 0.30510675764913053, + "grad_norm": 1.1668355547158702, + "learning_rate": 8.143179555938168e-06, + "loss": 0.2932, + "step": 13861 + }, + { + "epoch": 0.3051287695355492, + "grad_norm": 1.172829850110687, + "learning_rate": 8.142902331916081e-06, + "loss": 0.2918, + "step": 13862 + }, + { + "epoch": 0.3051507814219679, + "grad_norm": 0.6572206990332484, + "learning_rate": 8.142625091920362e-06, + "loss": 0.5207, + "step": 13863 + }, + { + "epoch": 0.30517279330838654, + "grad_norm": 1.4361694567072338, + "learning_rate": 8.142347835952418e-06, + "loss": 0.3135, + "step": 13864 + }, + { + "epoch": 0.3051948051948052, + "grad_norm": 2.3908725424111403, + "learning_rate": 8.142070564013665e-06, + "loss": 0.3254, + "step": 13865 + }, + { + "epoch": 0.30521681708122383, + "grad_norm": 1.5266568321092437, + "learning_rate": 8.141793276105503e-06, + "loss": 0.2809, + "step": 13866 + }, + { + "epoch": 0.30523882896764254, + "grad_norm": 1.2132570006867416, + "learning_rate": 8.141515972229351e-06, + "loss": 0.2552, + "step": 13867 + }, + { + "epoch": 0.3052608408540612, + "grad_norm": 0.6324526152654433, + "learning_rate": 8.141238652386612e-06, + "loss": 0.4999, + "step": 13868 + }, + { + "epoch": 0.30528285274047984, + "grad_norm": 1.3109563350531033, + "learning_rate": 8.140961316578695e-06, + "loss": 0.2217, + "step": 13869 + }, + { + "epoch": 0.30530486462689854, + "grad_norm": 1.800937491317299, + "learning_rate": 8.140683964807015e-06, + "loss": 0.2838, + "step": 13870 + }, + { + "epoch": 0.3053268765133172, + "grad_norm": 2.047853244576276, + "learning_rate": 8.140406597072976e-06, + "loss": 0.2983, + "step": 13871 + }, + { + "epoch": 0.30534888839973584, + "grad_norm": 1.517349666886333, + "learning_rate": 8.140129213377992e-06, + "loss": 0.3545, + "step": 13872 + }, + { + "epoch": 0.30537090028615455, + "grad_norm": 1.6301443174603842, + "learning_rate": 8.139851813723469e-06, + "loss": 0.3389, + "step": 13873 + }, + { + "epoch": 0.3053929121725732, + "grad_norm": 1.3023488340663691, + "learning_rate": 8.13957439811082e-06, + "loss": 0.3961, + "step": 13874 + }, + { + "epoch": 0.30541492405899184, + "grad_norm": 1.3406337390448086, + "learning_rate": 8.139296966541452e-06, + "loss": 0.4333, + "step": 13875 + }, + { + "epoch": 0.30543693594541055, + "grad_norm": 1.3113423493308223, + "learning_rate": 8.139019519016777e-06, + "loss": 0.3584, + "step": 13876 + }, + { + "epoch": 0.3054589478318292, + "grad_norm": 1.896878674806202, + "learning_rate": 8.138742055538206e-06, + "loss": 0.3815, + "step": 13877 + }, + { + "epoch": 0.30548095971824785, + "grad_norm": 1.7114329659064227, + "learning_rate": 8.138464576107147e-06, + "loss": 0.3077, + "step": 13878 + }, + { + "epoch": 0.3055029716046665, + "grad_norm": 1.2408411965058186, + "learning_rate": 8.138187080725012e-06, + "loss": 0.2741, + "step": 13879 + }, + { + "epoch": 0.3055249834910852, + "grad_norm": 1.1079321477203226, + "learning_rate": 8.13790956939321e-06, + "loss": 0.3274, + "step": 13880 + }, + { + "epoch": 0.30554699537750385, + "grad_norm": 1.35479556655464, + "learning_rate": 8.137632042113152e-06, + "loss": 0.2035, + "step": 13881 + }, + { + "epoch": 0.3055690072639225, + "grad_norm": 1.450689533403816, + "learning_rate": 8.137354498886248e-06, + "loss": 0.3523, + "step": 13882 + }, + { + "epoch": 0.3055910191503412, + "grad_norm": 0.6831664367111027, + "learning_rate": 8.13707693971391e-06, + "loss": 0.5401, + "step": 13883 + }, + { + "epoch": 0.30561303103675985, + "grad_norm": 0.6514576108857638, + "learning_rate": 8.136799364597548e-06, + "loss": 0.5248, + "step": 13884 + }, + { + "epoch": 0.3056350429231785, + "grad_norm": 1.7122527395686065, + "learning_rate": 8.136521773538573e-06, + "loss": 0.3631, + "step": 13885 + }, + { + "epoch": 0.3056570548095972, + "grad_norm": 1.4651414317160414, + "learning_rate": 8.136244166538394e-06, + "loss": 0.3411, + "step": 13886 + }, + { + "epoch": 0.30567906669601586, + "grad_norm": 1.465068534541304, + "learning_rate": 8.135966543598424e-06, + "loss": 0.2974, + "step": 13887 + }, + { + "epoch": 0.3057010785824345, + "grad_norm": 1.9126450420509418, + "learning_rate": 8.135688904720073e-06, + "loss": 0.3957, + "step": 13888 + }, + { + "epoch": 0.30572309046885315, + "grad_norm": 0.6594651244169007, + "learning_rate": 8.135411249904754e-06, + "loss": 0.5218, + "step": 13889 + }, + { + "epoch": 0.30574510235527186, + "grad_norm": 1.64906064771977, + "learning_rate": 8.135133579153875e-06, + "loss": 0.3118, + "step": 13890 + }, + { + "epoch": 0.3057671142416905, + "grad_norm": 1.4957889294533235, + "learning_rate": 8.13485589246885e-06, + "loss": 0.2364, + "step": 13891 + }, + { + "epoch": 0.30578912612810916, + "grad_norm": 1.7787074324901324, + "learning_rate": 8.134578189851086e-06, + "loss": 0.3978, + "step": 13892 + }, + { + "epoch": 0.30581113801452786, + "grad_norm": 1.7907243481252293, + "learning_rate": 8.134300471302001e-06, + "loss": 0.3772, + "step": 13893 + }, + { + "epoch": 0.3058331499009465, + "grad_norm": 1.370741133899101, + "learning_rate": 8.134022736823002e-06, + "loss": 0.2765, + "step": 13894 + }, + { + "epoch": 0.30585516178736516, + "grad_norm": 1.3128324880017652, + "learning_rate": 8.133744986415502e-06, + "loss": 0.366, + "step": 13895 + }, + { + "epoch": 0.30587717367378386, + "grad_norm": 1.4017009341174702, + "learning_rate": 8.13346722008091e-06, + "loss": 0.3869, + "step": 13896 + }, + { + "epoch": 0.3058991855602025, + "grad_norm": 1.8078763612017525, + "learning_rate": 8.133189437820643e-06, + "loss": 0.4612, + "step": 13897 + }, + { + "epoch": 0.30592119744662116, + "grad_norm": 1.2579519499971024, + "learning_rate": 8.13291163963611e-06, + "loss": 0.2935, + "step": 13898 + }, + { + "epoch": 0.30594320933303987, + "grad_norm": 1.8702513356884167, + "learning_rate": 8.132633825528721e-06, + "loss": 0.3012, + "step": 13899 + }, + { + "epoch": 0.3059652212194585, + "grad_norm": 1.3474044762428399, + "learning_rate": 8.13235599549989e-06, + "loss": 0.2885, + "step": 13900 + }, + { + "epoch": 0.30598723310587717, + "grad_norm": 1.319305501865463, + "learning_rate": 8.132078149551029e-06, + "loss": 0.3425, + "step": 13901 + }, + { + "epoch": 0.3060092449922958, + "grad_norm": 3.1526157311267045, + "learning_rate": 8.13180028768355e-06, + "loss": 0.3976, + "step": 13902 + }, + { + "epoch": 0.3060312568787145, + "grad_norm": 1.270131055338173, + "learning_rate": 8.131522409898866e-06, + "loss": 0.2974, + "step": 13903 + }, + { + "epoch": 0.30605326876513317, + "grad_norm": 1.4719689462450687, + "learning_rate": 8.131244516198386e-06, + "loss": 0.3564, + "step": 13904 + }, + { + "epoch": 0.3060752806515518, + "grad_norm": 1.8217906984583048, + "learning_rate": 8.130966606583525e-06, + "loss": 0.4187, + "step": 13905 + }, + { + "epoch": 0.3060972925379705, + "grad_norm": 0.6367740362241123, + "learning_rate": 8.130688681055698e-06, + "loss": 0.4963, + "step": 13906 + }, + { + "epoch": 0.30611930442438917, + "grad_norm": 0.668227043647672, + "learning_rate": 8.130410739616313e-06, + "loss": 0.5093, + "step": 13907 + }, + { + "epoch": 0.3061413163108078, + "grad_norm": 1.926532850228748, + "learning_rate": 8.130132782266785e-06, + "loss": 0.3002, + "step": 13908 + }, + { + "epoch": 0.3061633281972265, + "grad_norm": 2.5716051514759317, + "learning_rate": 8.129854809008526e-06, + "loss": 0.3329, + "step": 13909 + }, + { + "epoch": 0.3061853400836452, + "grad_norm": 1.3724076183591944, + "learning_rate": 8.129576819842948e-06, + "loss": 0.2648, + "step": 13910 + }, + { + "epoch": 0.3062073519700638, + "grad_norm": 1.546935435077708, + "learning_rate": 8.129298814771466e-06, + "loss": 0.2411, + "step": 13911 + }, + { + "epoch": 0.30622936385648253, + "grad_norm": 1.3845219147895886, + "learning_rate": 8.129020793795491e-06, + "loss": 0.3811, + "step": 13912 + }, + { + "epoch": 0.3062513757429012, + "grad_norm": 1.5889860802243678, + "learning_rate": 8.128742756916437e-06, + "loss": 0.3123, + "step": 13913 + }, + { + "epoch": 0.3062733876293198, + "grad_norm": 1.4193052289060892, + "learning_rate": 8.128464704135718e-06, + "loss": 0.3055, + "step": 13914 + }, + { + "epoch": 0.3062953995157385, + "grad_norm": 1.373358874684361, + "learning_rate": 8.128186635454745e-06, + "loss": 0.5115, + "step": 13915 + }, + { + "epoch": 0.3063174114021572, + "grad_norm": 0.6495907361954707, + "learning_rate": 8.12790855087493e-06, + "loss": 0.516, + "step": 13916 + }, + { + "epoch": 0.30633942328857583, + "grad_norm": 2.083533070950678, + "learning_rate": 8.127630450397691e-06, + "loss": 0.3908, + "step": 13917 + }, + { + "epoch": 0.3063614351749945, + "grad_norm": 0.627634219031728, + "learning_rate": 8.127352334024439e-06, + "loss": 0.51, + "step": 13918 + }, + { + "epoch": 0.3063834470614132, + "grad_norm": 1.6334163899504517, + "learning_rate": 8.127074201756588e-06, + "loss": 0.4283, + "step": 13919 + }, + { + "epoch": 0.30640545894783183, + "grad_norm": 1.417905238671411, + "learning_rate": 8.126796053595551e-06, + "loss": 0.3309, + "step": 13920 + }, + { + "epoch": 0.3064274708342505, + "grad_norm": 0.5931872214204307, + "learning_rate": 8.126517889542741e-06, + "loss": 0.5123, + "step": 13921 + }, + { + "epoch": 0.3064494827206692, + "grad_norm": 1.4484983416085664, + "learning_rate": 8.126239709599573e-06, + "loss": 0.3055, + "step": 13922 + }, + { + "epoch": 0.30647149460708784, + "grad_norm": 1.789136159415327, + "learning_rate": 8.125961513767461e-06, + "loss": 0.3749, + "step": 13923 + }, + { + "epoch": 0.3064935064935065, + "grad_norm": 1.5586689488582965, + "learning_rate": 8.125683302047818e-06, + "loss": 0.3433, + "step": 13924 + }, + { + "epoch": 0.30651551837992513, + "grad_norm": 1.760972342476651, + "learning_rate": 8.125405074442059e-06, + "loss": 0.3267, + "step": 13925 + }, + { + "epoch": 0.30653753026634384, + "grad_norm": 1.0660278812377864, + "learning_rate": 8.125126830951597e-06, + "loss": 0.3449, + "step": 13926 + }, + { + "epoch": 0.3065595421527625, + "grad_norm": 1.5075331726476868, + "learning_rate": 8.124848571577847e-06, + "loss": 0.3057, + "step": 13927 + }, + { + "epoch": 0.30658155403918114, + "grad_norm": 2.545256834753914, + "learning_rate": 8.124570296322221e-06, + "loss": 0.3058, + "step": 13928 + }, + { + "epoch": 0.30660356592559984, + "grad_norm": 1.3604805199636876, + "learning_rate": 8.124292005186136e-06, + "loss": 0.3495, + "step": 13929 + }, + { + "epoch": 0.3066255778120185, + "grad_norm": 3.1047560048651897, + "learning_rate": 8.124013698171007e-06, + "loss": 0.378, + "step": 13930 + }, + { + "epoch": 0.30664758969843714, + "grad_norm": 2.000245635138034, + "learning_rate": 8.123735375278246e-06, + "loss": 0.3236, + "step": 13931 + }, + { + "epoch": 0.30666960158485584, + "grad_norm": 1.2993695355575638, + "learning_rate": 8.123457036509268e-06, + "loss": 0.2675, + "step": 13932 + }, + { + "epoch": 0.3066916134712745, + "grad_norm": 1.4998915087539257, + "learning_rate": 8.123178681865489e-06, + "loss": 0.3704, + "step": 13933 + }, + { + "epoch": 0.30671362535769314, + "grad_norm": 1.4054967990814649, + "learning_rate": 8.122900311348323e-06, + "loss": 0.3182, + "step": 13934 + }, + { + "epoch": 0.30673563724411185, + "grad_norm": 1.3443433016588158, + "learning_rate": 8.122621924959184e-06, + "loss": 0.323, + "step": 13935 + }, + { + "epoch": 0.3067576491305305, + "grad_norm": 1.2783191220123156, + "learning_rate": 8.122343522699488e-06, + "loss": 0.3278, + "step": 13936 + }, + { + "epoch": 0.30677966101694915, + "grad_norm": 1.5877810089326094, + "learning_rate": 8.12206510457065e-06, + "loss": 0.3986, + "step": 13937 + }, + { + "epoch": 0.3068016729033678, + "grad_norm": 1.5111972002093046, + "learning_rate": 8.121786670574084e-06, + "loss": 0.3062, + "step": 13938 + }, + { + "epoch": 0.3068236847897865, + "grad_norm": 1.6604403258808635, + "learning_rate": 8.121508220711208e-06, + "loss": 0.3508, + "step": 13939 + }, + { + "epoch": 0.30684569667620515, + "grad_norm": 1.7859043038603393, + "learning_rate": 8.121229754983431e-06, + "loss": 0.3459, + "step": 13940 + }, + { + "epoch": 0.3068677085626238, + "grad_norm": 1.3241438953631723, + "learning_rate": 8.120951273392174e-06, + "loss": 0.3188, + "step": 13941 + }, + { + "epoch": 0.3068897204490425, + "grad_norm": 1.2946299776562347, + "learning_rate": 8.120672775938853e-06, + "loss": 0.2631, + "step": 13942 + }, + { + "epoch": 0.30691173233546115, + "grad_norm": 0.6935557460025875, + "learning_rate": 8.12039426262488e-06, + "loss": 0.5346, + "step": 13943 + }, + { + "epoch": 0.3069337442218798, + "grad_norm": 0.687105106422559, + "learning_rate": 8.12011573345167e-06, + "loss": 0.5349, + "step": 13944 + }, + { + "epoch": 0.3069557561082985, + "grad_norm": 1.4020264139783682, + "learning_rate": 8.11983718842064e-06, + "loss": 0.3174, + "step": 13945 + }, + { + "epoch": 0.30697776799471715, + "grad_norm": 1.7204910810160452, + "learning_rate": 8.119558627533208e-06, + "loss": 0.3951, + "step": 13946 + }, + { + "epoch": 0.3069997798811358, + "grad_norm": 4.539387535630764, + "learning_rate": 8.119280050790788e-06, + "loss": 0.4076, + "step": 13947 + }, + { + "epoch": 0.30702179176755445, + "grad_norm": 1.3237049789329094, + "learning_rate": 8.119001458194793e-06, + "loss": 0.3645, + "step": 13948 + }, + { + "epoch": 0.30704380365397316, + "grad_norm": 1.7977376737240054, + "learning_rate": 8.118722849746644e-06, + "loss": 0.3721, + "step": 13949 + }, + { + "epoch": 0.3070658155403918, + "grad_norm": 1.2666520172555362, + "learning_rate": 8.118444225447752e-06, + "loss": 0.3384, + "step": 13950 + }, + { + "epoch": 0.30708782742681046, + "grad_norm": 0.6831675416120438, + "learning_rate": 8.118165585299537e-06, + "loss": 0.4873, + "step": 13951 + }, + { + "epoch": 0.30710983931322916, + "grad_norm": 2.4153120001519524, + "learning_rate": 8.117886929303413e-06, + "loss": 0.3271, + "step": 13952 + }, + { + "epoch": 0.3071318511996478, + "grad_norm": 1.5333696609476046, + "learning_rate": 8.117608257460798e-06, + "loss": 0.3624, + "step": 13953 + }, + { + "epoch": 0.30715386308606646, + "grad_norm": 1.4612739193284996, + "learning_rate": 8.117329569773107e-06, + "loss": 0.3674, + "step": 13954 + }, + { + "epoch": 0.30717587497248516, + "grad_norm": 1.926526749704205, + "learning_rate": 8.117050866241756e-06, + "loss": 0.3674, + "step": 13955 + }, + { + "epoch": 0.3071978868589038, + "grad_norm": 5.689011547690239, + "learning_rate": 8.116772146868164e-06, + "loss": 0.3312, + "step": 13956 + }, + { + "epoch": 0.30721989874532246, + "grad_norm": 1.4103773506363277, + "learning_rate": 8.116493411653745e-06, + "loss": 0.368, + "step": 13957 + }, + { + "epoch": 0.30724191063174117, + "grad_norm": 1.6433761569424434, + "learning_rate": 8.116214660599915e-06, + "loss": 0.288, + "step": 13958 + }, + { + "epoch": 0.3072639225181598, + "grad_norm": 0.6594047441375045, + "learning_rate": 8.115935893708093e-06, + "loss": 0.5234, + "step": 13959 + }, + { + "epoch": 0.30728593440457846, + "grad_norm": 7.314825711819749, + "learning_rate": 8.115657110979697e-06, + "loss": 0.3196, + "step": 13960 + }, + { + "epoch": 0.3073079462909971, + "grad_norm": 1.3691594113269696, + "learning_rate": 8.115378312416141e-06, + "loss": 0.3605, + "step": 13961 + }, + { + "epoch": 0.3073299581774158, + "grad_norm": 1.2542620242371487, + "learning_rate": 8.11509949801884e-06, + "loss": 0.2983, + "step": 13962 + }, + { + "epoch": 0.30735197006383447, + "grad_norm": 2.04647300987219, + "learning_rate": 8.114820667789219e-06, + "loss": 0.377, + "step": 13963 + }, + { + "epoch": 0.3073739819502531, + "grad_norm": 1.7552324921695588, + "learning_rate": 8.114541821728687e-06, + "loss": 0.38, + "step": 13964 + }, + { + "epoch": 0.3073959938366718, + "grad_norm": 1.579756998750364, + "learning_rate": 8.114262959838663e-06, + "loss": 0.4092, + "step": 13965 + }, + { + "epoch": 0.30741800572309047, + "grad_norm": 2.3249835874852542, + "learning_rate": 8.11398408212057e-06, + "loss": 0.3949, + "step": 13966 + }, + { + "epoch": 0.3074400176095091, + "grad_norm": 1.4500007981917358, + "learning_rate": 8.113705188575818e-06, + "loss": 0.3941, + "step": 13967 + }, + { + "epoch": 0.3074620294959278, + "grad_norm": 0.6880377806567053, + "learning_rate": 8.113426279205828e-06, + "loss": 0.5233, + "step": 13968 + }, + { + "epoch": 0.3074840413823465, + "grad_norm": 0.6676692458451552, + "learning_rate": 8.113147354012017e-06, + "loss": 0.534, + "step": 13969 + }, + { + "epoch": 0.3075060532687651, + "grad_norm": 0.6142495227056917, + "learning_rate": 8.112868412995802e-06, + "loss": 0.5266, + "step": 13970 + }, + { + "epoch": 0.30752806515518377, + "grad_norm": 0.6183507284863193, + "learning_rate": 8.112589456158602e-06, + "loss": 0.5196, + "step": 13971 + }, + { + "epoch": 0.3075500770416025, + "grad_norm": 1.2807627393629164, + "learning_rate": 8.112310483501833e-06, + "loss": 0.2886, + "step": 13972 + }, + { + "epoch": 0.3075720889280211, + "grad_norm": 1.3732770847454316, + "learning_rate": 8.112031495026914e-06, + "loss": 0.393, + "step": 13973 + }, + { + "epoch": 0.3075941008144398, + "grad_norm": 1.7626973994167592, + "learning_rate": 8.111752490735265e-06, + "loss": 0.3743, + "step": 13974 + }, + { + "epoch": 0.3076161127008585, + "grad_norm": 1.578876054855037, + "learning_rate": 8.111473470628302e-06, + "loss": 0.4082, + "step": 13975 + }, + { + "epoch": 0.30763812458727713, + "grad_norm": 1.440911334110304, + "learning_rate": 8.11119443470744e-06, + "loss": 0.3719, + "step": 13976 + }, + { + "epoch": 0.3076601364736958, + "grad_norm": 1.737970968345461, + "learning_rate": 8.110915382974104e-06, + "loss": 0.4317, + "step": 13977 + }, + { + "epoch": 0.3076821483601145, + "grad_norm": 1.4029272258867105, + "learning_rate": 8.110636315429708e-06, + "loss": 0.3419, + "step": 13978 + }, + { + "epoch": 0.30770416024653313, + "grad_norm": 1.6416168810262728, + "learning_rate": 8.110357232075669e-06, + "loss": 0.275, + "step": 13979 + }, + { + "epoch": 0.3077261721329518, + "grad_norm": 0.8872171900655017, + "learning_rate": 8.110078132913409e-06, + "loss": 0.5092, + "step": 13980 + }, + { + "epoch": 0.3077481840193705, + "grad_norm": 1.2492873748039492, + "learning_rate": 8.109799017944343e-06, + "loss": 0.3402, + "step": 13981 + }, + { + "epoch": 0.30777019590578913, + "grad_norm": 2.1356914329338155, + "learning_rate": 8.109519887169893e-06, + "loss": 0.402, + "step": 13982 + }, + { + "epoch": 0.3077922077922078, + "grad_norm": 1.976346184245175, + "learning_rate": 8.109240740591473e-06, + "loss": 0.2693, + "step": 13983 + }, + { + "epoch": 0.30781421967862643, + "grad_norm": 2.159024119667551, + "learning_rate": 8.108961578210509e-06, + "loss": 0.3422, + "step": 13984 + }, + { + "epoch": 0.30783623156504514, + "grad_norm": 1.3161567770070512, + "learning_rate": 8.108682400028413e-06, + "loss": 0.3357, + "step": 13985 + }, + { + "epoch": 0.3078582434514638, + "grad_norm": 0.6235059153788782, + "learning_rate": 8.108403206046607e-06, + "loss": 0.4957, + "step": 13986 + }, + { + "epoch": 0.30788025533788244, + "grad_norm": 1.5158054305951818, + "learning_rate": 8.10812399626651e-06, + "loss": 0.3674, + "step": 13987 + }, + { + "epoch": 0.30790226722430114, + "grad_norm": 0.6353595724261092, + "learning_rate": 8.10784477068954e-06, + "loss": 0.4833, + "step": 13988 + }, + { + "epoch": 0.3079242791107198, + "grad_norm": 1.6466708977176214, + "learning_rate": 8.107565529317115e-06, + "loss": 0.2684, + "step": 13989 + }, + { + "epoch": 0.30794629099713844, + "grad_norm": 0.6363712460426951, + "learning_rate": 8.107286272150659e-06, + "loss": 0.521, + "step": 13990 + }, + { + "epoch": 0.30796830288355714, + "grad_norm": 1.308864276778399, + "learning_rate": 8.107006999191588e-06, + "loss": 0.2557, + "step": 13991 + }, + { + "epoch": 0.3079903147699758, + "grad_norm": 2.25263123660629, + "learning_rate": 8.10672771044132e-06, + "loss": 0.4409, + "step": 13992 + }, + { + "epoch": 0.30801232665639444, + "grad_norm": 1.4688858420335011, + "learning_rate": 8.106448405901279e-06, + "loss": 0.3728, + "step": 13993 + }, + { + "epoch": 0.30803433854281315, + "grad_norm": 0.6481614192708312, + "learning_rate": 8.10616908557288e-06, + "loss": 0.515, + "step": 13994 + }, + { + "epoch": 0.3080563504292318, + "grad_norm": 1.5048524890775534, + "learning_rate": 8.105889749457543e-06, + "loss": 0.2939, + "step": 13995 + }, + { + "epoch": 0.30807836231565044, + "grad_norm": 1.4685831017385735, + "learning_rate": 8.105610397556689e-06, + "loss": 0.3162, + "step": 13996 + }, + { + "epoch": 0.3081003742020691, + "grad_norm": 0.6616767833900613, + "learning_rate": 8.10533102987174e-06, + "loss": 0.5433, + "step": 13997 + }, + { + "epoch": 0.3081223860884878, + "grad_norm": 1.7406610494012742, + "learning_rate": 8.105051646404112e-06, + "loss": 0.3347, + "step": 13998 + }, + { + "epoch": 0.30814439797490645, + "grad_norm": 0.6132449642173133, + "learning_rate": 8.104772247155228e-06, + "loss": 0.5181, + "step": 13999 + }, + { + "epoch": 0.3081664098613251, + "grad_norm": 1.9047468767577749, + "learning_rate": 8.104492832126506e-06, + "loss": 0.3401, + "step": 14000 + }, + { + "epoch": 0.3081884217477438, + "grad_norm": 1.9135991443253229, + "learning_rate": 8.104213401319365e-06, + "loss": 0.3629, + "step": 14001 + }, + { + "epoch": 0.30821043363416245, + "grad_norm": 1.4897245291061356, + "learning_rate": 8.10393395473523e-06, + "loss": 0.2942, + "step": 14002 + }, + { + "epoch": 0.3082324455205811, + "grad_norm": 2.7467796864175944, + "learning_rate": 8.103654492375516e-06, + "loss": 0.3257, + "step": 14003 + }, + { + "epoch": 0.3082544574069998, + "grad_norm": 1.7533382866900626, + "learning_rate": 8.103375014241647e-06, + "loss": 0.3431, + "step": 14004 + }, + { + "epoch": 0.30827646929341845, + "grad_norm": 2.0223931557728694, + "learning_rate": 8.103095520335042e-06, + "loss": 0.4143, + "step": 14005 + }, + { + "epoch": 0.3082984811798371, + "grad_norm": 2.2383044967302643, + "learning_rate": 8.102816010657121e-06, + "loss": 0.3229, + "step": 14006 + }, + { + "epoch": 0.30832049306625575, + "grad_norm": 4.120753985943498, + "learning_rate": 8.102536485209304e-06, + "loss": 0.2652, + "step": 14007 + }, + { + "epoch": 0.30834250495267446, + "grad_norm": 1.4934636263011205, + "learning_rate": 8.102256943993014e-06, + "loss": 0.2861, + "step": 14008 + }, + { + "epoch": 0.3083645168390931, + "grad_norm": 2.161282097962183, + "learning_rate": 8.101977387009672e-06, + "loss": 0.3963, + "step": 14009 + }, + { + "epoch": 0.30838652872551175, + "grad_norm": 1.6255071230091056, + "learning_rate": 8.101697814260694e-06, + "loss": 0.3897, + "step": 14010 + }, + { + "epoch": 0.30840854061193046, + "grad_norm": 1.8994889661307806, + "learning_rate": 8.101418225747509e-06, + "loss": 0.3487, + "step": 14011 + }, + { + "epoch": 0.3084305524983491, + "grad_norm": 0.7048719136548015, + "learning_rate": 8.10113862147153e-06, + "loss": 0.5273, + "step": 14012 + }, + { + "epoch": 0.30845256438476776, + "grad_norm": 1.9229050161033725, + "learning_rate": 8.100859001434181e-06, + "loss": 0.421, + "step": 14013 + }, + { + "epoch": 0.30847457627118646, + "grad_norm": 1.2871324948279699, + "learning_rate": 8.100579365636885e-06, + "loss": 0.3857, + "step": 14014 + }, + { + "epoch": 0.3084965881576051, + "grad_norm": 0.6629878750136465, + "learning_rate": 8.100299714081062e-06, + "loss": 0.5005, + "step": 14015 + }, + { + "epoch": 0.30851860004402376, + "grad_norm": 1.7986260320031064, + "learning_rate": 8.100020046768131e-06, + "loss": 0.3268, + "step": 14016 + }, + { + "epoch": 0.30854061193044247, + "grad_norm": 1.142262453444607, + "learning_rate": 8.099740363699518e-06, + "loss": 0.3117, + "step": 14017 + }, + { + "epoch": 0.3085626238168611, + "grad_norm": 1.4565892778200997, + "learning_rate": 8.09946066487664e-06, + "loss": 0.2692, + "step": 14018 + }, + { + "epoch": 0.30858463570327976, + "grad_norm": 0.6196972417742153, + "learning_rate": 8.099180950300923e-06, + "loss": 0.5163, + "step": 14019 + }, + { + "epoch": 0.3086066475896984, + "grad_norm": 2.116430690381873, + "learning_rate": 8.098901219973785e-06, + "loss": 0.297, + "step": 14020 + }, + { + "epoch": 0.3086286594761171, + "grad_norm": 1.4511490352756367, + "learning_rate": 8.098621473896646e-06, + "loss": 0.4285, + "step": 14021 + }, + { + "epoch": 0.30865067136253577, + "grad_norm": 1.7344771713493143, + "learning_rate": 8.098341712070933e-06, + "loss": 0.3142, + "step": 14022 + }, + { + "epoch": 0.3086726832489544, + "grad_norm": 1.3971820985799632, + "learning_rate": 8.098061934498065e-06, + "loss": 0.3514, + "step": 14023 + }, + { + "epoch": 0.3086946951353731, + "grad_norm": 1.493195238493708, + "learning_rate": 8.097782141179465e-06, + "loss": 0.291, + "step": 14024 + }, + { + "epoch": 0.30871670702179177, + "grad_norm": 1.3941207952397603, + "learning_rate": 8.097502332116556e-06, + "loss": 0.4136, + "step": 14025 + }, + { + "epoch": 0.3087387189082104, + "grad_norm": 1.1912924152703872, + "learning_rate": 8.097222507310756e-06, + "loss": 0.3507, + "step": 14026 + }, + { + "epoch": 0.3087607307946291, + "grad_norm": 1.413993743523234, + "learning_rate": 8.096942666763493e-06, + "loss": 0.3156, + "step": 14027 + }, + { + "epoch": 0.3087827426810478, + "grad_norm": 0.696917623482296, + "learning_rate": 8.096662810476182e-06, + "loss": 0.4883, + "step": 14028 + }, + { + "epoch": 0.3088047545674664, + "grad_norm": 0.6692204450542268, + "learning_rate": 8.096382938450252e-06, + "loss": 0.4894, + "step": 14029 + }, + { + "epoch": 0.30882676645388507, + "grad_norm": 0.6170801745876201, + "learning_rate": 8.096103050687123e-06, + "loss": 0.4947, + "step": 14030 + }, + { + "epoch": 0.3088487783403038, + "grad_norm": 1.494615979987122, + "learning_rate": 8.095823147188215e-06, + "loss": 0.2878, + "step": 14031 + }, + { + "epoch": 0.3088707902267224, + "grad_norm": 1.3420729727083838, + "learning_rate": 8.095543227954955e-06, + "loss": 0.255, + "step": 14032 + }, + { + "epoch": 0.3088928021131411, + "grad_norm": 2.5003415068763943, + "learning_rate": 8.095263292988763e-06, + "loss": 0.3268, + "step": 14033 + }, + { + "epoch": 0.3089148139995598, + "grad_norm": 0.7536697141916284, + "learning_rate": 8.094983342291064e-06, + "loss": 0.5289, + "step": 14034 + }, + { + "epoch": 0.3089368258859784, + "grad_norm": 2.033564959886513, + "learning_rate": 8.094703375863278e-06, + "loss": 0.2606, + "step": 14035 + }, + { + "epoch": 0.3089588377723971, + "grad_norm": 1.835012641100944, + "learning_rate": 8.094423393706829e-06, + "loss": 0.387, + "step": 14036 + }, + { + "epoch": 0.3089808496588158, + "grad_norm": 0.6892202536196822, + "learning_rate": 8.09414339582314e-06, + "loss": 0.5254, + "step": 14037 + }, + { + "epoch": 0.30900286154523443, + "grad_norm": 1.6545216299032686, + "learning_rate": 8.093863382213635e-06, + "loss": 0.2727, + "step": 14038 + }, + { + "epoch": 0.3090248734316531, + "grad_norm": 1.4599708731466154, + "learning_rate": 8.093583352879737e-06, + "loss": 0.3193, + "step": 14039 + }, + { + "epoch": 0.3090468853180718, + "grad_norm": 1.6410530648530761, + "learning_rate": 8.093303307822867e-06, + "loss": 0.3001, + "step": 14040 + }, + { + "epoch": 0.30906889720449043, + "grad_norm": 0.613580384138168, + "learning_rate": 8.093023247044448e-06, + "loss": 0.5087, + "step": 14041 + }, + { + "epoch": 0.3090909090909091, + "grad_norm": 1.7656075543711345, + "learning_rate": 8.092743170545909e-06, + "loss": 0.3641, + "step": 14042 + }, + { + "epoch": 0.30911292097732773, + "grad_norm": 1.9782482057841584, + "learning_rate": 8.09246307832867e-06, + "loss": 0.3093, + "step": 14043 + }, + { + "epoch": 0.30913493286374644, + "grad_norm": 1.367288300392573, + "learning_rate": 8.092182970394152e-06, + "loss": 0.3765, + "step": 14044 + }, + { + "epoch": 0.3091569447501651, + "grad_norm": 2.2477284410949214, + "learning_rate": 8.091902846743782e-06, + "loss": 0.2858, + "step": 14045 + }, + { + "epoch": 0.30917895663658373, + "grad_norm": 1.401387939549117, + "learning_rate": 8.091622707378984e-06, + "loss": 0.3277, + "step": 14046 + }, + { + "epoch": 0.30920096852300244, + "grad_norm": 1.8619615342847926, + "learning_rate": 8.091342552301179e-06, + "loss": 0.3511, + "step": 14047 + }, + { + "epoch": 0.3092229804094211, + "grad_norm": 2.1443006462054814, + "learning_rate": 8.091062381511792e-06, + "loss": 0.4058, + "step": 14048 + }, + { + "epoch": 0.30924499229583974, + "grad_norm": 1.347174408995968, + "learning_rate": 8.090782195012248e-06, + "loss": 0.3693, + "step": 14049 + }, + { + "epoch": 0.30926700418225844, + "grad_norm": 1.8829416732783542, + "learning_rate": 8.09050199280397e-06, + "loss": 0.2773, + "step": 14050 + }, + { + "epoch": 0.3092890160686771, + "grad_norm": 1.351828028383004, + "learning_rate": 8.090221774888383e-06, + "loss": 0.259, + "step": 14051 + }, + { + "epoch": 0.30931102795509574, + "grad_norm": 1.1813310901617413, + "learning_rate": 8.08994154126691e-06, + "loss": 0.2636, + "step": 14052 + }, + { + "epoch": 0.3093330398415144, + "grad_norm": 2.028751629836391, + "learning_rate": 8.089661291940977e-06, + "loss": 0.3835, + "step": 14053 + }, + { + "epoch": 0.3093550517279331, + "grad_norm": 1.623682205946114, + "learning_rate": 8.089381026912006e-06, + "loss": 0.304, + "step": 14054 + }, + { + "epoch": 0.30937706361435174, + "grad_norm": 0.7032743022513406, + "learning_rate": 8.089100746181424e-06, + "loss": 0.5292, + "step": 14055 + }, + { + "epoch": 0.3093990755007704, + "grad_norm": 1.3632359435084087, + "learning_rate": 8.088820449750654e-06, + "loss": 0.3607, + "step": 14056 + }, + { + "epoch": 0.3094210873871891, + "grad_norm": 0.6284483716936692, + "learning_rate": 8.08854013762112e-06, + "loss": 0.5021, + "step": 14057 + }, + { + "epoch": 0.30944309927360775, + "grad_norm": 2.180067063776699, + "learning_rate": 8.088259809794248e-06, + "loss": 0.3807, + "step": 14058 + }, + { + "epoch": 0.3094651111600264, + "grad_norm": 1.7142190723479613, + "learning_rate": 8.087979466271463e-06, + "loss": 0.3576, + "step": 14059 + }, + { + "epoch": 0.3094871230464451, + "grad_norm": 0.6316512000916603, + "learning_rate": 8.087699107054188e-06, + "loss": 0.5134, + "step": 14060 + }, + { + "epoch": 0.30950913493286375, + "grad_norm": 2.284341808166811, + "learning_rate": 8.087418732143849e-06, + "loss": 0.3418, + "step": 14061 + }, + { + "epoch": 0.3095311468192824, + "grad_norm": 1.9429284356259175, + "learning_rate": 8.087138341541872e-06, + "loss": 0.2329, + "step": 14062 + }, + { + "epoch": 0.3095531587057011, + "grad_norm": 1.6596829016714305, + "learning_rate": 8.086857935249678e-06, + "loss": 0.2367, + "step": 14063 + }, + { + "epoch": 0.30957517059211975, + "grad_norm": 1.3106641436059274, + "learning_rate": 8.086577513268698e-06, + "loss": 0.2997, + "step": 14064 + }, + { + "epoch": 0.3095971824785384, + "grad_norm": 1.7529031849597847, + "learning_rate": 8.086297075600356e-06, + "loss": 0.3736, + "step": 14065 + }, + { + "epoch": 0.30961919436495705, + "grad_norm": 1.5568197611023773, + "learning_rate": 8.086016622246072e-06, + "loss": 0.3221, + "step": 14066 + }, + { + "epoch": 0.30964120625137576, + "grad_norm": 1.998679020550544, + "learning_rate": 8.085736153207277e-06, + "loss": 0.3562, + "step": 14067 + }, + { + "epoch": 0.3096632181377944, + "grad_norm": 1.446500683900497, + "learning_rate": 8.085455668485393e-06, + "loss": 0.3356, + "step": 14068 + }, + { + "epoch": 0.30968523002421305, + "grad_norm": 0.6659039252915528, + "learning_rate": 8.085175168081849e-06, + "loss": 0.4846, + "step": 14069 + }, + { + "epoch": 0.30970724191063176, + "grad_norm": 1.3048689785193786, + "learning_rate": 8.08489465199807e-06, + "loss": 0.3632, + "step": 14070 + }, + { + "epoch": 0.3097292537970504, + "grad_norm": 1.3698796460102451, + "learning_rate": 8.084614120235477e-06, + "loss": 0.4521, + "step": 14071 + }, + { + "epoch": 0.30975126568346906, + "grad_norm": 1.5563951585144753, + "learning_rate": 8.084333572795502e-06, + "loss": 0.3389, + "step": 14072 + }, + { + "epoch": 0.30977327756988776, + "grad_norm": 1.6173312745605386, + "learning_rate": 8.084053009679566e-06, + "loss": 0.393, + "step": 14073 + }, + { + "epoch": 0.3097952894563064, + "grad_norm": 1.6941327077366841, + "learning_rate": 8.083772430889098e-06, + "loss": 0.3696, + "step": 14074 + }, + { + "epoch": 0.30981730134272506, + "grad_norm": 1.7052810363747797, + "learning_rate": 8.083491836425525e-06, + "loss": 0.4118, + "step": 14075 + }, + { + "epoch": 0.30983931322914376, + "grad_norm": 1.8372233080883933, + "learning_rate": 8.083211226290269e-06, + "loss": 0.3025, + "step": 14076 + }, + { + "epoch": 0.3098613251155624, + "grad_norm": 1.3089095788902405, + "learning_rate": 8.082930600484759e-06, + "loss": 0.3547, + "step": 14077 + }, + { + "epoch": 0.30988333700198106, + "grad_norm": 1.5989061201131303, + "learning_rate": 8.08264995901042e-06, + "loss": 0.3603, + "step": 14078 + }, + { + "epoch": 0.3099053488883997, + "grad_norm": 1.527513361772938, + "learning_rate": 8.082369301868679e-06, + "loss": 0.3603, + "step": 14079 + }, + { + "epoch": 0.3099273607748184, + "grad_norm": 1.9966386232683697, + "learning_rate": 8.082088629060964e-06, + "loss": 0.2585, + "step": 14080 + }, + { + "epoch": 0.30994937266123707, + "grad_norm": 1.573365247828195, + "learning_rate": 8.081807940588699e-06, + "loss": 0.2624, + "step": 14081 + }, + { + "epoch": 0.3099713845476557, + "grad_norm": 1.2263695220247635, + "learning_rate": 8.08152723645331e-06, + "loss": 0.3786, + "step": 14082 + }, + { + "epoch": 0.3099933964340744, + "grad_norm": 1.3305436817992788, + "learning_rate": 8.081246516656227e-06, + "loss": 0.3052, + "step": 14083 + }, + { + "epoch": 0.31001540832049307, + "grad_norm": 1.1830278156626812, + "learning_rate": 8.080965781198875e-06, + "loss": 0.3429, + "step": 14084 + }, + { + "epoch": 0.3100374202069117, + "grad_norm": 4.905750344798983, + "learning_rate": 8.08068503008268e-06, + "loss": 0.3762, + "step": 14085 + }, + { + "epoch": 0.3100594320933304, + "grad_norm": 1.2400475142897034, + "learning_rate": 8.080404263309071e-06, + "loss": 0.3102, + "step": 14086 + }, + { + "epoch": 0.31008144397974907, + "grad_norm": 1.3427274004494403, + "learning_rate": 8.080123480879471e-06, + "loss": 0.3198, + "step": 14087 + }, + { + "epoch": 0.3101034558661677, + "grad_norm": 1.2975319898491862, + "learning_rate": 8.079842682795312e-06, + "loss": 0.3309, + "step": 14088 + }, + { + "epoch": 0.31012546775258637, + "grad_norm": 1.4718935099494208, + "learning_rate": 8.07956186905802e-06, + "loss": 0.3756, + "step": 14089 + }, + { + "epoch": 0.3101474796390051, + "grad_norm": 1.756690164087483, + "learning_rate": 8.079281039669018e-06, + "loss": 0.4428, + "step": 14090 + }, + { + "epoch": 0.3101694915254237, + "grad_norm": 1.7278284010101, + "learning_rate": 8.079000194629737e-06, + "loss": 0.4431, + "step": 14091 + }, + { + "epoch": 0.3101915034118424, + "grad_norm": 1.8343808423927734, + "learning_rate": 8.078719333941604e-06, + "loss": 0.3093, + "step": 14092 + }, + { + "epoch": 0.3102135152982611, + "grad_norm": 2.3290215738926534, + "learning_rate": 8.078438457606047e-06, + "loss": 0.2963, + "step": 14093 + }, + { + "epoch": 0.3102355271846797, + "grad_norm": 1.421401577399348, + "learning_rate": 8.078157565624493e-06, + "loss": 0.2739, + "step": 14094 + }, + { + "epoch": 0.3102575390710984, + "grad_norm": 1.3215432705549044, + "learning_rate": 8.077876657998369e-06, + "loss": 0.3718, + "step": 14095 + }, + { + "epoch": 0.3102795509575171, + "grad_norm": 1.6729192902361092, + "learning_rate": 8.077595734729101e-06, + "loss": 0.3951, + "step": 14096 + }, + { + "epoch": 0.31030156284393573, + "grad_norm": 1.5037504634773102, + "learning_rate": 8.077314795818121e-06, + "loss": 0.3915, + "step": 14097 + }, + { + "epoch": 0.3103235747303544, + "grad_norm": 1.332424138297827, + "learning_rate": 8.077033841266855e-06, + "loss": 0.3559, + "step": 14098 + }, + { + "epoch": 0.3103455866167731, + "grad_norm": 1.6456984274705506, + "learning_rate": 8.07675287107673e-06, + "loss": 0.2534, + "step": 14099 + }, + { + "epoch": 0.31036759850319173, + "grad_norm": 0.6832255321368335, + "learning_rate": 8.076471885249173e-06, + "loss": 0.5158, + "step": 14100 + }, + { + "epoch": 0.3103896103896104, + "grad_norm": 1.996168748949592, + "learning_rate": 8.076190883785615e-06, + "loss": 0.3768, + "step": 14101 + }, + { + "epoch": 0.31041162227602903, + "grad_norm": 1.7388068247352617, + "learning_rate": 8.075909866687484e-06, + "loss": 0.2885, + "step": 14102 + }, + { + "epoch": 0.31043363416244774, + "grad_norm": 0.621786025036995, + "learning_rate": 8.075628833956206e-06, + "loss": 0.5359, + "step": 14103 + }, + { + "epoch": 0.3104556460488664, + "grad_norm": 2.0197647877445615, + "learning_rate": 8.07534778559321e-06, + "loss": 0.327, + "step": 14104 + }, + { + "epoch": 0.31047765793528503, + "grad_norm": 1.6189875889755752, + "learning_rate": 8.075066721599926e-06, + "loss": 0.3388, + "step": 14105 + }, + { + "epoch": 0.31049966982170374, + "grad_norm": 1.7838305533740393, + "learning_rate": 8.074785641977782e-06, + "loss": 0.3051, + "step": 14106 + }, + { + "epoch": 0.3105216817081224, + "grad_norm": 0.667487179012389, + "learning_rate": 8.074504546728205e-06, + "loss": 0.5336, + "step": 14107 + }, + { + "epoch": 0.31054369359454104, + "grad_norm": 1.2849804315532123, + "learning_rate": 8.074223435852623e-06, + "loss": 0.3482, + "step": 14108 + }, + { + "epoch": 0.31056570548095974, + "grad_norm": 1.3649579808278594, + "learning_rate": 8.073942309352469e-06, + "loss": 0.2747, + "step": 14109 + }, + { + "epoch": 0.3105877173673784, + "grad_norm": 0.6461257611459403, + "learning_rate": 8.073661167229168e-06, + "loss": 0.5396, + "step": 14110 + }, + { + "epoch": 0.31060972925379704, + "grad_norm": 1.3574521487703473, + "learning_rate": 8.073380009484152e-06, + "loss": 0.3581, + "step": 14111 + }, + { + "epoch": 0.3106317411402157, + "grad_norm": 1.4137855121850718, + "learning_rate": 8.073098836118845e-06, + "loss": 0.3731, + "step": 14112 + }, + { + "epoch": 0.3106537530266344, + "grad_norm": 1.4421939704293, + "learning_rate": 8.07281764713468e-06, + "loss": 0.3841, + "step": 14113 + }, + { + "epoch": 0.31067576491305304, + "grad_norm": 2.3233761419687027, + "learning_rate": 8.072536442533086e-06, + "loss": 0.2697, + "step": 14114 + }, + { + "epoch": 0.3106977767994717, + "grad_norm": 1.3737740295273357, + "learning_rate": 8.07225522231549e-06, + "loss": 0.3919, + "step": 14115 + }, + { + "epoch": 0.3107197886858904, + "grad_norm": 1.5160527350689341, + "learning_rate": 8.071973986483323e-06, + "loss": 0.4152, + "step": 14116 + }, + { + "epoch": 0.31074180057230905, + "grad_norm": 1.549884323609316, + "learning_rate": 8.071692735038016e-06, + "loss": 0.3475, + "step": 14117 + }, + { + "epoch": 0.3107638124587277, + "grad_norm": 1.6519636771342399, + "learning_rate": 8.071411467980994e-06, + "loss": 0.341, + "step": 14118 + }, + { + "epoch": 0.3107858243451464, + "grad_norm": 6.098654753035222, + "learning_rate": 8.071130185313692e-06, + "loss": 0.3742, + "step": 14119 + }, + { + "epoch": 0.31080783623156505, + "grad_norm": 1.2106967412556486, + "learning_rate": 8.070848887037534e-06, + "loss": 0.2833, + "step": 14120 + }, + { + "epoch": 0.3108298481179837, + "grad_norm": 1.409971537034557, + "learning_rate": 8.070567573153953e-06, + "loss": 0.3004, + "step": 14121 + }, + { + "epoch": 0.3108518600044024, + "grad_norm": 2.012618223474938, + "learning_rate": 8.070286243664379e-06, + "loss": 0.3643, + "step": 14122 + }, + { + "epoch": 0.31087387189082105, + "grad_norm": 1.5108833145831762, + "learning_rate": 8.07000489857024e-06, + "loss": 0.297, + "step": 14123 + }, + { + "epoch": 0.3108958837772397, + "grad_norm": 1.625370840853454, + "learning_rate": 8.069723537872967e-06, + "loss": 0.2849, + "step": 14124 + }, + { + "epoch": 0.31091789566365835, + "grad_norm": 2.3142194667744347, + "learning_rate": 8.06944216157399e-06, + "loss": 0.3435, + "step": 14125 + }, + { + "epoch": 0.31093990755007705, + "grad_norm": 1.4625526284949817, + "learning_rate": 8.069160769674739e-06, + "loss": 0.3629, + "step": 14126 + }, + { + "epoch": 0.3109619194364957, + "grad_norm": 1.58920057102173, + "learning_rate": 8.068879362176642e-06, + "loss": 0.2552, + "step": 14127 + }, + { + "epoch": 0.31098393132291435, + "grad_norm": 2.4547228960259564, + "learning_rate": 8.068597939081133e-06, + "loss": 0.2856, + "step": 14128 + }, + { + "epoch": 0.31100594320933306, + "grad_norm": 1.3823111424885963, + "learning_rate": 8.068316500389642e-06, + "loss": 0.3207, + "step": 14129 + }, + { + "epoch": 0.3110279550957517, + "grad_norm": 0.7605980863073794, + "learning_rate": 8.068035046103597e-06, + "loss": 0.5409, + "step": 14130 + }, + { + "epoch": 0.31104996698217036, + "grad_norm": 1.4484936165974924, + "learning_rate": 8.06775357622443e-06, + "loss": 0.3699, + "step": 14131 + }, + { + "epoch": 0.31107197886858906, + "grad_norm": 1.5878545242717446, + "learning_rate": 8.06747209075357e-06, + "loss": 0.3444, + "step": 14132 + }, + { + "epoch": 0.3110939907550077, + "grad_norm": 1.3178720176221768, + "learning_rate": 8.067190589692448e-06, + "loss": 0.3617, + "step": 14133 + }, + { + "epoch": 0.31111600264142636, + "grad_norm": 3.6495844479087327, + "learning_rate": 8.066909073042497e-06, + "loss": 0.2395, + "step": 14134 + }, + { + "epoch": 0.31113801452784506, + "grad_norm": 1.3451394143497255, + "learning_rate": 8.066627540805148e-06, + "loss": 0.3906, + "step": 14135 + }, + { + "epoch": 0.3111600264142637, + "grad_norm": 1.3855758313898856, + "learning_rate": 8.066345992981827e-06, + "loss": 0.3371, + "step": 14136 + }, + { + "epoch": 0.31118203830068236, + "grad_norm": 1.3699434306080378, + "learning_rate": 8.066064429573968e-06, + "loss": 0.2879, + "step": 14137 + }, + { + "epoch": 0.311204050187101, + "grad_norm": 1.3715693562821218, + "learning_rate": 8.065782850583004e-06, + "loss": 0.3142, + "step": 14138 + }, + { + "epoch": 0.3112260620735197, + "grad_norm": 1.3927699971195346, + "learning_rate": 8.065501256010362e-06, + "loss": 0.3853, + "step": 14139 + }, + { + "epoch": 0.31124807395993837, + "grad_norm": 10.418757991503883, + "learning_rate": 8.065219645857478e-06, + "loss": 0.3839, + "step": 14140 + }, + { + "epoch": 0.311270085846357, + "grad_norm": 1.4821598117663164, + "learning_rate": 8.06493802012578e-06, + "loss": 0.3275, + "step": 14141 + }, + { + "epoch": 0.3112920977327757, + "grad_norm": 0.7222588352596262, + "learning_rate": 8.064656378816697e-06, + "loss": 0.5171, + "step": 14142 + }, + { + "epoch": 0.31131410961919437, + "grad_norm": 0.6431679211808168, + "learning_rate": 8.064374721931667e-06, + "loss": 0.4894, + "step": 14143 + }, + { + "epoch": 0.311336121505613, + "grad_norm": 1.1207206719246146, + "learning_rate": 8.064093049472115e-06, + "loss": 0.3482, + "step": 14144 + }, + { + "epoch": 0.3113581333920317, + "grad_norm": 1.4818000886628704, + "learning_rate": 8.063811361439478e-06, + "loss": 0.3697, + "step": 14145 + }, + { + "epoch": 0.31138014527845037, + "grad_norm": 1.0547929088535482, + "learning_rate": 8.063529657835183e-06, + "loss": 0.3219, + "step": 14146 + }, + { + "epoch": 0.311402157164869, + "grad_norm": 2.498947751682051, + "learning_rate": 8.063247938660665e-06, + "loss": 0.3332, + "step": 14147 + }, + { + "epoch": 0.31142416905128767, + "grad_norm": 1.6863575425850226, + "learning_rate": 8.062966203917355e-06, + "loss": 0.3232, + "step": 14148 + }, + { + "epoch": 0.3114461809377064, + "grad_norm": 1.2748222907907198, + "learning_rate": 8.062684453606685e-06, + "loss": 0.2886, + "step": 14149 + }, + { + "epoch": 0.311468192824125, + "grad_norm": 1.4159134278322751, + "learning_rate": 8.062402687730087e-06, + "loss": 0.3075, + "step": 14150 + }, + { + "epoch": 0.31149020471054367, + "grad_norm": 0.746304377640295, + "learning_rate": 8.06212090628899e-06, + "loss": 0.5122, + "step": 14151 + }, + { + "epoch": 0.3115122165969624, + "grad_norm": 1.648866911404441, + "learning_rate": 8.061839109284831e-06, + "loss": 0.308, + "step": 14152 + }, + { + "epoch": 0.311534228483381, + "grad_norm": 0.6911271493848843, + "learning_rate": 8.06155729671904e-06, + "loss": 0.5212, + "step": 14153 + }, + { + "epoch": 0.3115562403697997, + "grad_norm": 18.210085093175092, + "learning_rate": 8.061275468593048e-06, + "loss": 0.3765, + "step": 14154 + }, + { + "epoch": 0.3115782522562184, + "grad_norm": 1.1953452376557823, + "learning_rate": 8.06099362490829e-06, + "loss": 0.3599, + "step": 14155 + }, + { + "epoch": 0.31160026414263703, + "grad_norm": 1.47756080613251, + "learning_rate": 8.060711765666197e-06, + "loss": 0.4018, + "step": 14156 + }, + { + "epoch": 0.3116222760290557, + "grad_norm": 2.0354683757267846, + "learning_rate": 8.060429890868203e-06, + "loss": 0.3761, + "step": 14157 + }, + { + "epoch": 0.3116442879154744, + "grad_norm": 1.432704471205729, + "learning_rate": 8.060148000515737e-06, + "loss": 0.2716, + "step": 14158 + }, + { + "epoch": 0.31166629980189303, + "grad_norm": 1.207342167080137, + "learning_rate": 8.059866094610234e-06, + "loss": 0.2976, + "step": 14159 + }, + { + "epoch": 0.3116883116883117, + "grad_norm": 1.3690016493570096, + "learning_rate": 8.059584173153127e-06, + "loss": 0.4539, + "step": 14160 + }, + { + "epoch": 0.31171032357473033, + "grad_norm": 1.661574685986532, + "learning_rate": 8.05930223614585e-06, + "loss": 0.3567, + "step": 14161 + }, + { + "epoch": 0.31173233546114903, + "grad_norm": 1.5851670438693835, + "learning_rate": 8.059020283589834e-06, + "loss": 0.3338, + "step": 14162 + }, + { + "epoch": 0.3117543473475677, + "grad_norm": 1.2616930370069692, + "learning_rate": 8.058738315486512e-06, + "loss": 0.3483, + "step": 14163 + }, + { + "epoch": 0.31177635923398633, + "grad_norm": 0.854145193793541, + "learning_rate": 8.058456331837316e-06, + "loss": 0.4835, + "step": 14164 + }, + { + "epoch": 0.31179837112040504, + "grad_norm": 1.1229848983621615, + "learning_rate": 8.058174332643685e-06, + "loss": 0.3848, + "step": 14165 + }, + { + "epoch": 0.3118203830068237, + "grad_norm": 1.6216044646395673, + "learning_rate": 8.057892317907045e-06, + "loss": 0.4192, + "step": 14166 + }, + { + "epoch": 0.31184239489324234, + "grad_norm": 1.9505071056917147, + "learning_rate": 8.057610287628832e-06, + "loss": 0.4137, + "step": 14167 + }, + { + "epoch": 0.31186440677966104, + "grad_norm": 1.4511887889005368, + "learning_rate": 8.057328241810481e-06, + "loss": 0.342, + "step": 14168 + }, + { + "epoch": 0.3118864186660797, + "grad_norm": 1.3595135105193488, + "learning_rate": 8.057046180453426e-06, + "loss": 0.289, + "step": 14169 + }, + { + "epoch": 0.31190843055249834, + "grad_norm": 1.3185198938865044, + "learning_rate": 8.056764103559096e-06, + "loss": 0.3257, + "step": 14170 + }, + { + "epoch": 0.311930442438917, + "grad_norm": 3.044560835633917, + "learning_rate": 8.056482011128928e-06, + "loss": 0.2733, + "step": 14171 + }, + { + "epoch": 0.3119524543253357, + "grad_norm": 1.3735832850612886, + "learning_rate": 8.056199903164355e-06, + "loss": 0.3725, + "step": 14172 + }, + { + "epoch": 0.31197446621175434, + "grad_norm": 1.0933609985014823, + "learning_rate": 8.055917779666811e-06, + "loss": 0.2899, + "step": 14173 + }, + { + "epoch": 0.311996478098173, + "grad_norm": 1.653376670428998, + "learning_rate": 8.05563564063773e-06, + "loss": 0.3504, + "step": 14174 + }, + { + "epoch": 0.3120184899845917, + "grad_norm": 1.910585972538306, + "learning_rate": 8.055353486078548e-06, + "loss": 0.3002, + "step": 14175 + }, + { + "epoch": 0.31204050187101035, + "grad_norm": 1.0664153630092659, + "learning_rate": 8.055071315990695e-06, + "loss": 0.278, + "step": 14176 + }, + { + "epoch": 0.312062513757429, + "grad_norm": 3.8198847266425373, + "learning_rate": 8.054789130375607e-06, + "loss": 0.4154, + "step": 14177 + }, + { + "epoch": 0.3120845256438477, + "grad_norm": 1.3351190417378074, + "learning_rate": 8.054506929234719e-06, + "loss": 0.3882, + "step": 14178 + }, + { + "epoch": 0.31210653753026635, + "grad_norm": 3.147965306211863, + "learning_rate": 8.054224712569463e-06, + "loss": 0.3803, + "step": 14179 + }, + { + "epoch": 0.312128549416685, + "grad_norm": 0.7013716465456745, + "learning_rate": 8.053942480381276e-06, + "loss": 0.5102, + "step": 14180 + }, + { + "epoch": 0.3121505613031037, + "grad_norm": 1.365281980061174, + "learning_rate": 8.053660232671591e-06, + "loss": 0.3162, + "step": 14181 + }, + { + "epoch": 0.31217257318952235, + "grad_norm": 1.1418151141648114, + "learning_rate": 8.053377969441843e-06, + "loss": 0.2721, + "step": 14182 + }, + { + "epoch": 0.312194585075941, + "grad_norm": 1.7077350000728722, + "learning_rate": 8.053095690693464e-06, + "loss": 0.3472, + "step": 14183 + }, + { + "epoch": 0.31221659696235965, + "grad_norm": 1.3547310817246718, + "learning_rate": 8.052813396427895e-06, + "loss": 0.2707, + "step": 14184 + }, + { + "epoch": 0.31223860884877835, + "grad_norm": 0.6048565633171361, + "learning_rate": 8.052531086646564e-06, + "loss": 0.4831, + "step": 14185 + }, + { + "epoch": 0.312260620735197, + "grad_norm": 1.1749842358562956, + "learning_rate": 8.052248761350908e-06, + "loss": 0.2989, + "step": 14186 + }, + { + "epoch": 0.31228263262161565, + "grad_norm": 1.2649897995946513, + "learning_rate": 8.051966420542364e-06, + "loss": 0.2947, + "step": 14187 + }, + { + "epoch": 0.31230464450803436, + "grad_norm": 1.3602982407771054, + "learning_rate": 8.051684064222364e-06, + "loss": 0.3166, + "step": 14188 + }, + { + "epoch": 0.312326656394453, + "grad_norm": 1.5635600311593925, + "learning_rate": 8.051401692392346e-06, + "loss": 0.463, + "step": 14189 + }, + { + "epoch": 0.31234866828087166, + "grad_norm": 1.193762862141683, + "learning_rate": 8.051119305053742e-06, + "loss": 0.3453, + "step": 14190 + }, + { + "epoch": 0.31237068016729036, + "grad_norm": 1.6896201943611444, + "learning_rate": 8.05083690220799e-06, + "loss": 0.3464, + "step": 14191 + }, + { + "epoch": 0.312392692053709, + "grad_norm": 3.8461066183979518, + "learning_rate": 8.050554483856525e-06, + "loss": 0.2403, + "step": 14192 + }, + { + "epoch": 0.31241470394012766, + "grad_norm": 1.7701075445679872, + "learning_rate": 8.050272050000782e-06, + "loss": 0.3005, + "step": 14193 + }, + { + "epoch": 0.3124367158265463, + "grad_norm": 1.4005843936405529, + "learning_rate": 8.049989600642194e-06, + "loss": 0.326, + "step": 14194 + }, + { + "epoch": 0.312458727712965, + "grad_norm": 1.8055454160300326, + "learning_rate": 8.049707135782198e-06, + "loss": 0.3128, + "step": 14195 + }, + { + "epoch": 0.31248073959938366, + "grad_norm": 1.3992354062089387, + "learning_rate": 8.049424655422231e-06, + "loss": 0.3213, + "step": 14196 + }, + { + "epoch": 0.3125027514858023, + "grad_norm": 0.7507837346729677, + "learning_rate": 8.049142159563728e-06, + "loss": 0.5105, + "step": 14197 + }, + { + "epoch": 0.312524763372221, + "grad_norm": 1.1760884639314935, + "learning_rate": 8.048859648208126e-06, + "loss": 0.2854, + "step": 14198 + }, + { + "epoch": 0.31254677525863966, + "grad_norm": 1.443492800838699, + "learning_rate": 8.048577121356858e-06, + "loss": 0.2895, + "step": 14199 + }, + { + "epoch": 0.3125687871450583, + "grad_norm": 2.466552432729313, + "learning_rate": 8.04829457901136e-06, + "loss": 0.2851, + "step": 14200 + }, + { + "epoch": 0.312590799031477, + "grad_norm": 1.3332704242133362, + "learning_rate": 8.048012021173072e-06, + "loss": 0.3688, + "step": 14201 + }, + { + "epoch": 0.31261281091789567, + "grad_norm": 2.1188332198614046, + "learning_rate": 8.047729447843426e-06, + "loss": 0.3569, + "step": 14202 + }, + { + "epoch": 0.3126348228043143, + "grad_norm": 1.7782748461018394, + "learning_rate": 8.04744685902386e-06, + "loss": 0.4127, + "step": 14203 + }, + { + "epoch": 0.312656834690733, + "grad_norm": 1.8561078572668235, + "learning_rate": 8.047164254715809e-06, + "loss": 0.3845, + "step": 14204 + }, + { + "epoch": 0.31267884657715167, + "grad_norm": 1.2005423575262886, + "learning_rate": 8.04688163492071e-06, + "loss": 0.304, + "step": 14205 + }, + { + "epoch": 0.3127008584635703, + "grad_norm": 1.2191116989190247, + "learning_rate": 8.046598999640003e-06, + "loss": 0.365, + "step": 14206 + }, + { + "epoch": 0.31272287034998897, + "grad_norm": 1.2763998978398172, + "learning_rate": 8.046316348875119e-06, + "loss": 0.2693, + "step": 14207 + }, + { + "epoch": 0.3127448822364077, + "grad_norm": 1.4761910545897634, + "learning_rate": 8.046033682627494e-06, + "loss": 0.3995, + "step": 14208 + }, + { + "epoch": 0.3127668941228263, + "grad_norm": 1.39969039577748, + "learning_rate": 8.04575100089857e-06, + "loss": 0.3936, + "step": 14209 + }, + { + "epoch": 0.31278890600924497, + "grad_norm": 1.139842107251039, + "learning_rate": 8.04546830368978e-06, + "loss": 0.3323, + "step": 14210 + }, + { + "epoch": 0.3128109178956637, + "grad_norm": 1.2942149427401752, + "learning_rate": 8.045185591002563e-06, + "loss": 0.3665, + "step": 14211 + }, + { + "epoch": 0.3128329297820823, + "grad_norm": 1.3478493572641963, + "learning_rate": 8.044902862838354e-06, + "loss": 0.3014, + "step": 14212 + }, + { + "epoch": 0.312854941668501, + "grad_norm": 1.5258456468976278, + "learning_rate": 8.04462011919859e-06, + "loss": 0.3554, + "step": 14213 + }, + { + "epoch": 0.3128769535549197, + "grad_norm": 1.315683147090964, + "learning_rate": 8.04433736008471e-06, + "loss": 0.3507, + "step": 14214 + }, + { + "epoch": 0.31289896544133833, + "grad_norm": 1.6795533443314146, + "learning_rate": 8.044054585498149e-06, + "loss": 0.3751, + "step": 14215 + }, + { + "epoch": 0.312920977327757, + "grad_norm": 1.555828176298367, + "learning_rate": 8.043771795440345e-06, + "loss": 0.3595, + "step": 14216 + }, + { + "epoch": 0.3129429892141757, + "grad_norm": 1.2078371694748622, + "learning_rate": 8.043488989912734e-06, + "loss": 0.3124, + "step": 14217 + }, + { + "epoch": 0.31296500110059433, + "grad_norm": 1.136295795259417, + "learning_rate": 8.043206168916755e-06, + "loss": 0.2643, + "step": 14218 + }, + { + "epoch": 0.312987012987013, + "grad_norm": 1.2677562719735311, + "learning_rate": 8.042923332453845e-06, + "loss": 0.3768, + "step": 14219 + }, + { + "epoch": 0.31300902487343163, + "grad_norm": 1.4656563245977912, + "learning_rate": 8.042640480525443e-06, + "loss": 0.3614, + "step": 14220 + }, + { + "epoch": 0.31303103675985033, + "grad_norm": 1.8109052820801488, + "learning_rate": 8.042357613132983e-06, + "loss": 0.3525, + "step": 14221 + }, + { + "epoch": 0.313053048646269, + "grad_norm": 1.4636976065864309, + "learning_rate": 8.042074730277905e-06, + "loss": 0.32, + "step": 14222 + }, + { + "epoch": 0.31307506053268763, + "grad_norm": 1.3363670557033114, + "learning_rate": 8.041791831961648e-06, + "loss": 0.2965, + "step": 14223 + }, + { + "epoch": 0.31309707241910634, + "grad_norm": 1.1769350015152393, + "learning_rate": 8.041508918185647e-06, + "loss": 0.3555, + "step": 14224 + }, + { + "epoch": 0.313119084305525, + "grad_norm": 1.8003075314789836, + "learning_rate": 8.04122598895134e-06, + "loss": 0.3359, + "step": 14225 + }, + { + "epoch": 0.31314109619194364, + "grad_norm": 1.4719747010769517, + "learning_rate": 8.040943044260167e-06, + "loss": 0.3431, + "step": 14226 + }, + { + "epoch": 0.31316310807836234, + "grad_norm": 1.355107533665633, + "learning_rate": 8.040660084113566e-06, + "loss": 0.3656, + "step": 14227 + }, + { + "epoch": 0.313185119964781, + "grad_norm": 0.683235997732661, + "learning_rate": 8.040377108512975e-06, + "loss": 0.53, + "step": 14228 + }, + { + "epoch": 0.31320713185119964, + "grad_norm": 1.5543623840987875, + "learning_rate": 8.04009411745983e-06, + "loss": 0.3498, + "step": 14229 + }, + { + "epoch": 0.3132291437376183, + "grad_norm": 1.848416951603303, + "learning_rate": 8.03981111095557e-06, + "loss": 0.3427, + "step": 14230 + }, + { + "epoch": 0.313251155624037, + "grad_norm": 1.1433744703476358, + "learning_rate": 8.039528089001636e-06, + "loss": 0.2357, + "step": 14231 + }, + { + "epoch": 0.31327316751045564, + "grad_norm": 1.7859037518689476, + "learning_rate": 8.039245051599464e-06, + "loss": 0.3859, + "step": 14232 + }, + { + "epoch": 0.3132951793968743, + "grad_norm": 1.3673545792710187, + "learning_rate": 8.038961998750492e-06, + "loss": 0.2549, + "step": 14233 + }, + { + "epoch": 0.313317191283293, + "grad_norm": 1.7304338148413245, + "learning_rate": 8.038678930456161e-06, + "loss": 0.3776, + "step": 14234 + }, + { + "epoch": 0.31333920316971164, + "grad_norm": 1.2442313670918428, + "learning_rate": 8.038395846717907e-06, + "loss": 0.4072, + "step": 14235 + }, + { + "epoch": 0.3133612150561303, + "grad_norm": 1.7526473698428415, + "learning_rate": 8.038112747537172e-06, + "loss": 0.3726, + "step": 14236 + }, + { + "epoch": 0.313383226942549, + "grad_norm": 1.4710912759762447, + "learning_rate": 8.037829632915392e-06, + "loss": 0.3137, + "step": 14237 + }, + { + "epoch": 0.31340523882896765, + "grad_norm": 1.2544114493753464, + "learning_rate": 8.037546502854007e-06, + "loss": 0.3069, + "step": 14238 + }, + { + "epoch": 0.3134272507153863, + "grad_norm": 0.6216783795909981, + "learning_rate": 8.037263357354456e-06, + "loss": 0.5233, + "step": 14239 + }, + { + "epoch": 0.313449262601805, + "grad_norm": 1.759485608787577, + "learning_rate": 8.036980196418179e-06, + "loss": 0.372, + "step": 14240 + }, + { + "epoch": 0.31347127448822365, + "grad_norm": 1.4690230317993946, + "learning_rate": 8.03669702004661e-06, + "loss": 0.2902, + "step": 14241 + }, + { + "epoch": 0.3134932863746423, + "grad_norm": 0.6532479924377156, + "learning_rate": 8.036413828241196e-06, + "loss": 0.5251, + "step": 14242 + }, + { + "epoch": 0.31351529826106095, + "grad_norm": 1.8258812662768948, + "learning_rate": 8.036130621003374e-06, + "loss": 0.3769, + "step": 14243 + }, + { + "epoch": 0.31353731014747965, + "grad_norm": 1.4906986091549832, + "learning_rate": 8.03584739833458e-06, + "loss": 0.4402, + "step": 14244 + }, + { + "epoch": 0.3135593220338983, + "grad_norm": 0.6111247112511918, + "learning_rate": 8.035564160236255e-06, + "loss": 0.4948, + "step": 14245 + }, + { + "epoch": 0.31358133392031695, + "grad_norm": 0.6337758650267671, + "learning_rate": 8.03528090670984e-06, + "loss": 0.4962, + "step": 14246 + }, + { + "epoch": 0.31360334580673566, + "grad_norm": 1.4175643836169205, + "learning_rate": 8.034997637756771e-06, + "loss": 0.3483, + "step": 14247 + }, + { + "epoch": 0.3136253576931543, + "grad_norm": 1.9612574507370004, + "learning_rate": 8.034714353378494e-06, + "loss": 0.3507, + "step": 14248 + }, + { + "epoch": 0.31364736957957295, + "grad_norm": 1.64715409765182, + "learning_rate": 8.034431053576444e-06, + "loss": 0.3547, + "step": 14249 + }, + { + "epoch": 0.31366938146599166, + "grad_norm": 2.528930378236149, + "learning_rate": 8.03414773835206e-06, + "loss": 0.3309, + "step": 14250 + }, + { + "epoch": 0.3136913933524103, + "grad_norm": 2.2538249735051212, + "learning_rate": 8.033864407706786e-06, + "loss": 0.2871, + "step": 14251 + }, + { + "epoch": 0.31371340523882896, + "grad_norm": 1.4259676359653204, + "learning_rate": 8.033581061642059e-06, + "loss": 0.3749, + "step": 14252 + }, + { + "epoch": 0.3137354171252476, + "grad_norm": 1.173766672688996, + "learning_rate": 8.033297700159321e-06, + "loss": 0.265, + "step": 14253 + }, + { + "epoch": 0.3137574290116663, + "grad_norm": 1.3031212181140317, + "learning_rate": 8.033014323260012e-06, + "loss": 0.3696, + "step": 14254 + }, + { + "epoch": 0.31377944089808496, + "grad_norm": 1.2377509469632713, + "learning_rate": 8.03273093094557e-06, + "loss": 0.3099, + "step": 14255 + }, + { + "epoch": 0.3138014527845036, + "grad_norm": 2.0976106481630445, + "learning_rate": 8.032447523217437e-06, + "loss": 0.3527, + "step": 14256 + }, + { + "epoch": 0.3138234646709223, + "grad_norm": 1.922206407462672, + "learning_rate": 8.032164100077053e-06, + "loss": 0.366, + "step": 14257 + }, + { + "epoch": 0.31384547655734096, + "grad_norm": 1.4328250735420376, + "learning_rate": 8.031880661525858e-06, + "loss": 0.3523, + "step": 14258 + }, + { + "epoch": 0.3138674884437596, + "grad_norm": 1.3045681397894229, + "learning_rate": 8.031597207565293e-06, + "loss": 0.3618, + "step": 14259 + }, + { + "epoch": 0.3138895003301783, + "grad_norm": 1.3223393630607088, + "learning_rate": 8.0313137381968e-06, + "loss": 0.2701, + "step": 14260 + }, + { + "epoch": 0.31391151221659697, + "grad_norm": 1.3883448504931317, + "learning_rate": 8.03103025342182e-06, + "loss": 0.3369, + "step": 14261 + }, + { + "epoch": 0.3139335241030156, + "grad_norm": 1.2726328350072402, + "learning_rate": 8.03074675324179e-06, + "loss": 0.3593, + "step": 14262 + }, + { + "epoch": 0.3139555359894343, + "grad_norm": 0.6721088405955857, + "learning_rate": 8.030463237658156e-06, + "loss": 0.5135, + "step": 14263 + }, + { + "epoch": 0.31397754787585297, + "grad_norm": 1.4100650581597458, + "learning_rate": 8.030179706672352e-06, + "loss": 0.3106, + "step": 14264 + }, + { + "epoch": 0.3139995597622716, + "grad_norm": 3.03316339378151, + "learning_rate": 8.029896160285827e-06, + "loss": 0.3184, + "step": 14265 + }, + { + "epoch": 0.31402157164869027, + "grad_norm": 2.0121018325420077, + "learning_rate": 8.029612598500016e-06, + "loss": 0.379, + "step": 14266 + }, + { + "epoch": 0.31404358353510897, + "grad_norm": 1.5048270047716399, + "learning_rate": 8.029329021316364e-06, + "loss": 0.3558, + "step": 14267 + }, + { + "epoch": 0.3140655954215276, + "grad_norm": 1.1202907389959698, + "learning_rate": 8.02904542873631e-06, + "loss": 0.3214, + "step": 14268 + }, + { + "epoch": 0.31408760730794627, + "grad_norm": 1.288939172068346, + "learning_rate": 8.028761820761296e-06, + "loss": 0.3939, + "step": 14269 + }, + { + "epoch": 0.314109619194365, + "grad_norm": 2.8990134153754092, + "learning_rate": 8.028478197392764e-06, + "loss": 0.2743, + "step": 14270 + }, + { + "epoch": 0.3141316310807836, + "grad_norm": 1.3897908792931368, + "learning_rate": 8.028194558632155e-06, + "loss": 0.3476, + "step": 14271 + }, + { + "epoch": 0.3141536429672023, + "grad_norm": 1.4299871510569961, + "learning_rate": 8.02791090448091e-06, + "loss": 0.3379, + "step": 14272 + }, + { + "epoch": 0.314175654853621, + "grad_norm": 1.4161729792680982, + "learning_rate": 8.027627234940471e-06, + "loss": 0.3925, + "step": 14273 + }, + { + "epoch": 0.3141976667400396, + "grad_norm": 1.4147071027047144, + "learning_rate": 8.027343550012282e-06, + "loss": 0.3907, + "step": 14274 + }, + { + "epoch": 0.3142196786264583, + "grad_norm": 0.6570611165148647, + "learning_rate": 8.027059849697783e-06, + "loss": 0.5081, + "step": 14275 + }, + { + "epoch": 0.314241690512877, + "grad_norm": 1.5710250789742697, + "learning_rate": 8.026776133998414e-06, + "loss": 0.3548, + "step": 14276 + }, + { + "epoch": 0.31426370239929563, + "grad_norm": 1.0829769175664476, + "learning_rate": 8.02649240291562e-06, + "loss": 0.2461, + "step": 14277 + }, + { + "epoch": 0.3142857142857143, + "grad_norm": 1.7272374948192344, + "learning_rate": 8.02620865645084e-06, + "loss": 0.4085, + "step": 14278 + }, + { + "epoch": 0.31430772617213293, + "grad_norm": 0.624683134407591, + "learning_rate": 8.025924894605519e-06, + "loss": 0.5041, + "step": 14279 + }, + { + "epoch": 0.31432973805855163, + "grad_norm": 1.4331036356822313, + "learning_rate": 8.025641117381098e-06, + "loss": 0.3415, + "step": 14280 + }, + { + "epoch": 0.3143517499449703, + "grad_norm": 0.6527059624932375, + "learning_rate": 8.025357324779019e-06, + "loss": 0.5181, + "step": 14281 + }, + { + "epoch": 0.31437376183138893, + "grad_norm": 1.3674446038980357, + "learning_rate": 8.025073516800724e-06, + "loss": 0.3734, + "step": 14282 + }, + { + "epoch": 0.31439577371780764, + "grad_norm": 1.6483249441089656, + "learning_rate": 8.024789693447657e-06, + "loss": 0.444, + "step": 14283 + }, + { + "epoch": 0.3144177856042263, + "grad_norm": 1.670367401436822, + "learning_rate": 8.02450585472126e-06, + "loss": 0.3052, + "step": 14284 + }, + { + "epoch": 0.31443979749064493, + "grad_norm": 1.5356892435837337, + "learning_rate": 8.024222000622976e-06, + "loss": 0.2667, + "step": 14285 + }, + { + "epoch": 0.31446180937706364, + "grad_norm": 2.5005884470351134, + "learning_rate": 8.023938131154245e-06, + "loss": 0.3188, + "step": 14286 + }, + { + "epoch": 0.3144838212634823, + "grad_norm": 0.6054902707559675, + "learning_rate": 8.023654246316514e-06, + "loss": 0.4879, + "step": 14287 + }, + { + "epoch": 0.31450583314990094, + "grad_norm": 1.1913166583447465, + "learning_rate": 8.023370346111223e-06, + "loss": 0.3303, + "step": 14288 + }, + { + "epoch": 0.3145278450363196, + "grad_norm": 1.436430379458168, + "learning_rate": 8.023086430539815e-06, + "loss": 0.3752, + "step": 14289 + }, + { + "epoch": 0.3145498569227383, + "grad_norm": 1.7772007132619874, + "learning_rate": 8.022802499603732e-06, + "loss": 0.3396, + "step": 14290 + }, + { + "epoch": 0.31457186880915694, + "grad_norm": 2.002677586641801, + "learning_rate": 8.022518553304419e-06, + "loss": 0.3123, + "step": 14291 + }, + { + "epoch": 0.3145938806955756, + "grad_norm": 1.2237692988232443, + "learning_rate": 8.02223459164332e-06, + "loss": 0.3203, + "step": 14292 + }, + { + "epoch": 0.3146158925819943, + "grad_norm": 2.002694175938301, + "learning_rate": 8.021950614621876e-06, + "loss": 0.3269, + "step": 14293 + }, + { + "epoch": 0.31463790446841294, + "grad_norm": 1.598029091555748, + "learning_rate": 8.021666622241531e-06, + "loss": 0.2833, + "step": 14294 + }, + { + "epoch": 0.3146599163548316, + "grad_norm": 1.9274172666595637, + "learning_rate": 8.02138261450373e-06, + "loss": 0.3214, + "step": 14295 + }, + { + "epoch": 0.3146819282412503, + "grad_norm": 1.625895820463225, + "learning_rate": 8.021098591409913e-06, + "loss": 0.3165, + "step": 14296 + }, + { + "epoch": 0.31470394012766895, + "grad_norm": 1.5087357157946895, + "learning_rate": 8.020814552961527e-06, + "loss": 0.2716, + "step": 14297 + }, + { + "epoch": 0.3147259520140876, + "grad_norm": 1.3327101373142598, + "learning_rate": 8.020530499160013e-06, + "loss": 0.295, + "step": 14298 + }, + { + "epoch": 0.3147479639005063, + "grad_norm": 1.242795042597378, + "learning_rate": 8.020246430006817e-06, + "loss": 0.2979, + "step": 14299 + }, + { + "epoch": 0.31476997578692495, + "grad_norm": 1.6798921421564141, + "learning_rate": 8.019962345503381e-06, + "loss": 0.3157, + "step": 14300 + }, + { + "epoch": 0.3147919876733436, + "grad_norm": 2.136397650611086, + "learning_rate": 8.019678245651147e-06, + "loss": 0.2631, + "step": 14301 + }, + { + "epoch": 0.31481399955976225, + "grad_norm": 1.4196490150805539, + "learning_rate": 8.019394130451565e-06, + "loss": 0.3431, + "step": 14302 + }, + { + "epoch": 0.31483601144618095, + "grad_norm": 1.402829762304737, + "learning_rate": 8.019109999906074e-06, + "loss": 0.3292, + "step": 14303 + }, + { + "epoch": 0.3148580233325996, + "grad_norm": 1.3852570039908536, + "learning_rate": 8.01882585401612e-06, + "loss": 0.2913, + "step": 14304 + }, + { + "epoch": 0.31488003521901825, + "grad_norm": 1.7591836847428368, + "learning_rate": 8.018541692783145e-06, + "loss": 0.3886, + "step": 14305 + }, + { + "epoch": 0.31490204710543696, + "grad_norm": 1.359559611463225, + "learning_rate": 8.018257516208597e-06, + "loss": 0.3205, + "step": 14306 + }, + { + "epoch": 0.3149240589918556, + "grad_norm": 1.4665161665257123, + "learning_rate": 8.017973324293916e-06, + "loss": 0.3764, + "step": 14307 + }, + { + "epoch": 0.31494607087827425, + "grad_norm": 1.4000349122471383, + "learning_rate": 8.017689117040549e-06, + "loss": 0.3808, + "step": 14308 + }, + { + "epoch": 0.31496808276469296, + "grad_norm": 1.7140099669630862, + "learning_rate": 8.017404894449941e-06, + "loss": 0.3559, + "step": 14309 + }, + { + "epoch": 0.3149900946511116, + "grad_norm": 1.4694656061643483, + "learning_rate": 8.017120656523534e-06, + "loss": 0.3097, + "step": 14310 + }, + { + "epoch": 0.31501210653753026, + "grad_norm": 1.542565959207443, + "learning_rate": 8.016836403262775e-06, + "loss": 0.3953, + "step": 14311 + }, + { + "epoch": 0.3150341184239489, + "grad_norm": 1.4611586507890635, + "learning_rate": 8.016552134669108e-06, + "loss": 0.3913, + "step": 14312 + }, + { + "epoch": 0.3150561303103676, + "grad_norm": 1.2744667530601734, + "learning_rate": 8.016267850743978e-06, + "loss": 0.3374, + "step": 14313 + }, + { + "epoch": 0.31507814219678626, + "grad_norm": 1.2631835481635012, + "learning_rate": 8.015983551488828e-06, + "loss": 0.3016, + "step": 14314 + }, + { + "epoch": 0.3151001540832049, + "grad_norm": 1.2105209122736877, + "learning_rate": 8.015699236905108e-06, + "loss": 0.3417, + "step": 14315 + }, + { + "epoch": 0.3151221659696236, + "grad_norm": 2.0915580738645905, + "learning_rate": 8.015414906994255e-06, + "loss": 0.3251, + "step": 14316 + }, + { + "epoch": 0.31514417785604226, + "grad_norm": 0.6784382307166543, + "learning_rate": 8.01513056175772e-06, + "loss": 0.5221, + "step": 14317 + }, + { + "epoch": 0.3151661897424609, + "grad_norm": 2.2703771027038573, + "learning_rate": 8.014846201196948e-06, + "loss": 0.3844, + "step": 14318 + }, + { + "epoch": 0.3151882016288796, + "grad_norm": 1.9673524335628654, + "learning_rate": 8.014561825313382e-06, + "loss": 0.3225, + "step": 14319 + }, + { + "epoch": 0.31521021351529827, + "grad_norm": 1.3262745610034894, + "learning_rate": 8.014277434108469e-06, + "loss": 0.2633, + "step": 14320 + }, + { + "epoch": 0.3152322254017169, + "grad_norm": 0.6223835360174518, + "learning_rate": 8.013993027583652e-06, + "loss": 0.5043, + "step": 14321 + }, + { + "epoch": 0.3152542372881356, + "grad_norm": 1.35347306255454, + "learning_rate": 8.01370860574038e-06, + "loss": 0.3136, + "step": 14322 + }, + { + "epoch": 0.31527624917455427, + "grad_norm": 1.4311658790139778, + "learning_rate": 8.013424168580095e-06, + "loss": 0.3183, + "step": 14323 + }, + { + "epoch": 0.3152982610609729, + "grad_norm": 1.4934056595499074, + "learning_rate": 8.013139716104245e-06, + "loss": 0.3912, + "step": 14324 + }, + { + "epoch": 0.31532027294739157, + "grad_norm": 1.264327232959634, + "learning_rate": 8.012855248314276e-06, + "loss": 0.384, + "step": 14325 + }, + { + "epoch": 0.31534228483381027, + "grad_norm": 0.6252307460091642, + "learning_rate": 8.012570765211631e-06, + "loss": 0.506, + "step": 14326 + }, + { + "epoch": 0.3153642967202289, + "grad_norm": 9.525335611105058, + "learning_rate": 8.01228626679776e-06, + "loss": 0.3267, + "step": 14327 + }, + { + "epoch": 0.31538630860664757, + "grad_norm": 1.5835974250658336, + "learning_rate": 8.012001753074107e-06, + "loss": 0.4032, + "step": 14328 + }, + { + "epoch": 0.3154083204930663, + "grad_norm": 1.341519246743626, + "learning_rate": 8.011717224042115e-06, + "loss": 0.2321, + "step": 14329 + }, + { + "epoch": 0.3154303323794849, + "grad_norm": 0.6104965095871535, + "learning_rate": 8.011432679703234e-06, + "loss": 0.5206, + "step": 14330 + }, + { + "epoch": 0.3154523442659036, + "grad_norm": 1.3558339684592304, + "learning_rate": 8.011148120058909e-06, + "loss": 0.354, + "step": 14331 + }, + { + "epoch": 0.3154743561523223, + "grad_norm": 1.620484532931889, + "learning_rate": 8.010863545110586e-06, + "loss": 0.3645, + "step": 14332 + }, + { + "epoch": 0.3154963680387409, + "grad_norm": 1.42719554196191, + "learning_rate": 8.010578954859712e-06, + "loss": 0.3473, + "step": 14333 + }, + { + "epoch": 0.3155183799251596, + "grad_norm": 1.5367311437649662, + "learning_rate": 8.010294349307733e-06, + "loss": 0.3773, + "step": 14334 + }, + { + "epoch": 0.3155403918115782, + "grad_norm": 1.616381750523241, + "learning_rate": 8.010009728456094e-06, + "loss": 0.3253, + "step": 14335 + }, + { + "epoch": 0.31556240369799693, + "grad_norm": 1.2391997429662571, + "learning_rate": 8.009725092306244e-06, + "loss": 0.2995, + "step": 14336 + }, + { + "epoch": 0.3155844155844156, + "grad_norm": 1.2298030129934527, + "learning_rate": 8.00944044085963e-06, + "loss": 0.2662, + "step": 14337 + }, + { + "epoch": 0.3156064274708342, + "grad_norm": 0.6328617175125334, + "learning_rate": 8.009155774117698e-06, + "loss": 0.5046, + "step": 14338 + }, + { + "epoch": 0.31562843935725293, + "grad_norm": 1.1493067243102013, + "learning_rate": 8.008871092081891e-06, + "loss": 0.2595, + "step": 14339 + }, + { + "epoch": 0.3156504512436716, + "grad_norm": 1.4990944801661577, + "learning_rate": 8.00858639475366e-06, + "loss": 0.3499, + "step": 14340 + }, + { + "epoch": 0.31567246313009023, + "grad_norm": 1.875668149416027, + "learning_rate": 8.008301682134454e-06, + "loss": 0.367, + "step": 14341 + }, + { + "epoch": 0.31569447501650894, + "grad_norm": 1.419920598369987, + "learning_rate": 8.008016954225714e-06, + "loss": 0.33, + "step": 14342 + }, + { + "epoch": 0.3157164869029276, + "grad_norm": 1.2678286978083766, + "learning_rate": 8.007732211028893e-06, + "loss": 0.3087, + "step": 14343 + }, + { + "epoch": 0.31573849878934623, + "grad_norm": 1.8796316749888837, + "learning_rate": 8.007447452545434e-06, + "loss": 0.3038, + "step": 14344 + }, + { + "epoch": 0.31576051067576494, + "grad_norm": 1.3125975386869573, + "learning_rate": 8.007162678776784e-06, + "loss": 0.3057, + "step": 14345 + }, + { + "epoch": 0.3157825225621836, + "grad_norm": 1.8413084664357926, + "learning_rate": 8.006877889724394e-06, + "loss": 0.4048, + "step": 14346 + }, + { + "epoch": 0.31580453444860224, + "grad_norm": 0.6167371789462188, + "learning_rate": 8.00659308538971e-06, + "loss": 0.4933, + "step": 14347 + }, + { + "epoch": 0.3158265463350209, + "grad_norm": 1.7030445496841249, + "learning_rate": 8.006308265774177e-06, + "loss": 0.2807, + "step": 14348 + }, + { + "epoch": 0.3158485582214396, + "grad_norm": 1.4513688840107892, + "learning_rate": 8.006023430879245e-06, + "loss": 0.3715, + "step": 14349 + }, + { + "epoch": 0.31587057010785824, + "grad_norm": 0.61703532321994, + "learning_rate": 8.005738580706362e-06, + "loss": 0.5219, + "step": 14350 + }, + { + "epoch": 0.3158925819942769, + "grad_norm": 1.7906693170381294, + "learning_rate": 8.005453715256973e-06, + "loss": 0.324, + "step": 14351 + }, + { + "epoch": 0.3159145938806956, + "grad_norm": 1.749489796320375, + "learning_rate": 8.00516883453253e-06, + "loss": 0.3357, + "step": 14352 + }, + { + "epoch": 0.31593660576711424, + "grad_norm": 1.710817240491679, + "learning_rate": 8.004883938534478e-06, + "loss": 0.3701, + "step": 14353 + }, + { + "epoch": 0.3159586176535329, + "grad_norm": 1.203588963111041, + "learning_rate": 8.004599027264266e-06, + "loss": 0.2888, + "step": 14354 + }, + { + "epoch": 0.3159806295399516, + "grad_norm": 1.5043206741434962, + "learning_rate": 8.004314100723341e-06, + "loss": 0.3446, + "step": 14355 + }, + { + "epoch": 0.31600264142637025, + "grad_norm": 1.9670296784544368, + "learning_rate": 8.004029158913152e-06, + "loss": 0.3879, + "step": 14356 + }, + { + "epoch": 0.3160246533127889, + "grad_norm": 1.100776830644722, + "learning_rate": 8.003744201835146e-06, + "loss": 0.2705, + "step": 14357 + }, + { + "epoch": 0.3160466651992076, + "grad_norm": 2.5713026239873567, + "learning_rate": 8.003459229490775e-06, + "loss": 0.4493, + "step": 14358 + }, + { + "epoch": 0.31606867708562625, + "grad_norm": 1.6185510069334472, + "learning_rate": 8.003174241881482e-06, + "loss": 0.2775, + "step": 14359 + }, + { + "epoch": 0.3160906889720449, + "grad_norm": 4.43489883635027, + "learning_rate": 8.002889239008717e-06, + "loss": 0.2975, + "step": 14360 + }, + { + "epoch": 0.31611270085846355, + "grad_norm": 1.3000598028625614, + "learning_rate": 8.002604220873934e-06, + "loss": 0.2898, + "step": 14361 + }, + { + "epoch": 0.31613471274488225, + "grad_norm": 0.6687561688869474, + "learning_rate": 8.002319187478573e-06, + "loss": 0.528, + "step": 14362 + }, + { + "epoch": 0.3161567246313009, + "grad_norm": 1.6822413114805432, + "learning_rate": 8.002034138824089e-06, + "loss": 0.2954, + "step": 14363 + }, + { + "epoch": 0.31617873651771955, + "grad_norm": 1.3616750338217694, + "learning_rate": 8.001749074911928e-06, + "loss": 0.3029, + "step": 14364 + }, + { + "epoch": 0.31620074840413825, + "grad_norm": 0.6422621852432716, + "learning_rate": 8.001463995743539e-06, + "loss": 0.5265, + "step": 14365 + }, + { + "epoch": 0.3162227602905569, + "grad_norm": 1.9227591251365093, + "learning_rate": 8.001178901320372e-06, + "loss": 0.4062, + "step": 14366 + }, + { + "epoch": 0.31624477217697555, + "grad_norm": 1.2509738444581828, + "learning_rate": 8.000893791643876e-06, + "loss": 0.2808, + "step": 14367 + }, + { + "epoch": 0.31626678406339426, + "grad_norm": 1.6024157326219197, + "learning_rate": 8.000608666715497e-06, + "loss": 0.2901, + "step": 14368 + }, + { + "epoch": 0.3162887959498129, + "grad_norm": 1.3163787063385142, + "learning_rate": 8.000323526536688e-06, + "loss": 0.3001, + "step": 14369 + }, + { + "epoch": 0.31631080783623156, + "grad_norm": 1.5643506798642173, + "learning_rate": 8.000038371108895e-06, + "loss": 0.3637, + "step": 14370 + }, + { + "epoch": 0.3163328197226502, + "grad_norm": 1.8691462907514238, + "learning_rate": 7.999753200433571e-06, + "loss": 0.3006, + "step": 14371 + }, + { + "epoch": 0.3163548316090689, + "grad_norm": 1.755961131067297, + "learning_rate": 7.999468014512162e-06, + "loss": 0.3428, + "step": 14372 + }, + { + "epoch": 0.31637684349548756, + "grad_norm": 2.603238028735634, + "learning_rate": 7.99918281334612e-06, + "loss": 0.2421, + "step": 14373 + }, + { + "epoch": 0.3163988553819062, + "grad_norm": 8.558315884539141, + "learning_rate": 7.998897596936892e-06, + "loss": 0.3548, + "step": 14374 + }, + { + "epoch": 0.3164208672683249, + "grad_norm": 1.418902107955041, + "learning_rate": 7.998612365285931e-06, + "loss": 0.3805, + "step": 14375 + }, + { + "epoch": 0.31644287915474356, + "grad_norm": 2.7745299018110194, + "learning_rate": 7.998327118394682e-06, + "loss": 0.4209, + "step": 14376 + }, + { + "epoch": 0.3164648910411622, + "grad_norm": 1.4239123525221071, + "learning_rate": 7.998041856264598e-06, + "loss": 0.3409, + "step": 14377 + }, + { + "epoch": 0.3164869029275809, + "grad_norm": 1.277644913872946, + "learning_rate": 7.997756578897129e-06, + "loss": 0.3329, + "step": 14378 + }, + { + "epoch": 0.31650891481399956, + "grad_norm": 1.6062150389182894, + "learning_rate": 7.997471286293723e-06, + "loss": 0.3225, + "step": 14379 + }, + { + "epoch": 0.3165309267004182, + "grad_norm": 2.06723265452873, + "learning_rate": 7.997185978455832e-06, + "loss": 0.2716, + "step": 14380 + }, + { + "epoch": 0.3165529385868369, + "grad_norm": 0.6616798713480514, + "learning_rate": 7.996900655384905e-06, + "loss": 0.5129, + "step": 14381 + }, + { + "epoch": 0.31657495047325557, + "grad_norm": 1.694862808226522, + "learning_rate": 7.99661531708239e-06, + "loss": 0.3926, + "step": 14382 + }, + { + "epoch": 0.3165969623596742, + "grad_norm": 3.092674558027368, + "learning_rate": 7.996329963549743e-06, + "loss": 0.378, + "step": 14383 + }, + { + "epoch": 0.31661897424609287, + "grad_norm": 3.3574079154169816, + "learning_rate": 7.996044594788407e-06, + "loss": 0.4131, + "step": 14384 + }, + { + "epoch": 0.31664098613251157, + "grad_norm": 0.6264539888892577, + "learning_rate": 7.99575921079984e-06, + "loss": 0.512, + "step": 14385 + }, + { + "epoch": 0.3166629980189302, + "grad_norm": 1.6267182142016905, + "learning_rate": 7.995473811585485e-06, + "loss": 0.3866, + "step": 14386 + }, + { + "epoch": 0.31668500990534887, + "grad_norm": 0.6467808840709209, + "learning_rate": 7.995188397146798e-06, + "loss": 0.529, + "step": 14387 + }, + { + "epoch": 0.3167070217917676, + "grad_norm": 2.2545939224048435, + "learning_rate": 7.994902967485228e-06, + "loss": 0.3959, + "step": 14388 + }, + { + "epoch": 0.3167290336781862, + "grad_norm": 1.6298440152592815, + "learning_rate": 7.994617522602225e-06, + "loss": 0.3302, + "step": 14389 + }, + { + "epoch": 0.31675104556460487, + "grad_norm": 1.5377995276475065, + "learning_rate": 7.994332062499239e-06, + "loss": 0.3156, + "step": 14390 + }, + { + "epoch": 0.3167730574510236, + "grad_norm": 1.2740136245938167, + "learning_rate": 7.994046587177722e-06, + "loss": 0.3247, + "step": 14391 + }, + { + "epoch": 0.3167950693374422, + "grad_norm": 1.7819446351944708, + "learning_rate": 7.993761096639125e-06, + "loss": 0.386, + "step": 14392 + }, + { + "epoch": 0.3168170812238609, + "grad_norm": 1.4492236709217992, + "learning_rate": 7.9934755908849e-06, + "loss": 0.3772, + "step": 14393 + }, + { + "epoch": 0.3168390931102795, + "grad_norm": 1.7611660587653712, + "learning_rate": 7.993190069916496e-06, + "loss": 0.2913, + "step": 14394 + }, + { + "epoch": 0.31686110499669823, + "grad_norm": 1.2708446807610856, + "learning_rate": 7.992904533735366e-06, + "loss": 0.3432, + "step": 14395 + }, + { + "epoch": 0.3168831168831169, + "grad_norm": 1.755575717884137, + "learning_rate": 7.99261898234296e-06, + "loss": 0.3338, + "step": 14396 + }, + { + "epoch": 0.3169051287695355, + "grad_norm": 0.6964212821430452, + "learning_rate": 7.99233341574073e-06, + "loss": 0.5082, + "step": 14397 + }, + { + "epoch": 0.31692714065595423, + "grad_norm": 1.5748475722161472, + "learning_rate": 7.992047833930126e-06, + "loss": 0.2973, + "step": 14398 + }, + { + "epoch": 0.3169491525423729, + "grad_norm": 1.4516876248102137, + "learning_rate": 7.991762236912601e-06, + "loss": 0.3238, + "step": 14399 + }, + { + "epoch": 0.31697116442879153, + "grad_norm": 1.4526181784501455, + "learning_rate": 7.991476624689605e-06, + "loss": 0.3116, + "step": 14400 + }, + { + "epoch": 0.31699317631521023, + "grad_norm": 1.1745080678920814, + "learning_rate": 7.99119099726259e-06, + "loss": 0.3504, + "step": 14401 + }, + { + "epoch": 0.3170151882016289, + "grad_norm": 1.2344459116224795, + "learning_rate": 7.99090535463301e-06, + "loss": 0.3169, + "step": 14402 + }, + { + "epoch": 0.31703720008804753, + "grad_norm": 1.2990856407285547, + "learning_rate": 7.990619696802316e-06, + "loss": 0.3297, + "step": 14403 + }, + { + "epoch": 0.31705921197446624, + "grad_norm": 2.1670774854112596, + "learning_rate": 7.990334023771956e-06, + "loss": 0.4117, + "step": 14404 + }, + { + "epoch": 0.3170812238608849, + "grad_norm": 0.6299023863247891, + "learning_rate": 7.990048335543385e-06, + "loss": 0.4918, + "step": 14405 + }, + { + "epoch": 0.31710323574730354, + "grad_norm": 1.3287457672413827, + "learning_rate": 7.989762632118056e-06, + "loss": 0.338, + "step": 14406 + }, + { + "epoch": 0.3171252476337222, + "grad_norm": 1.3446791665183033, + "learning_rate": 7.98947691349742e-06, + "loss": 0.391, + "step": 14407 + }, + { + "epoch": 0.3171472595201409, + "grad_norm": 1.35101276716535, + "learning_rate": 7.989191179682929e-06, + "loss": 0.3364, + "step": 14408 + }, + { + "epoch": 0.31716927140655954, + "grad_norm": 1.1810463356388232, + "learning_rate": 7.988905430676034e-06, + "loss": 0.2777, + "step": 14409 + }, + { + "epoch": 0.3171912832929782, + "grad_norm": 1.687471478283732, + "learning_rate": 7.98861966647819e-06, + "loss": 0.3437, + "step": 14410 + }, + { + "epoch": 0.3172132951793969, + "grad_norm": 1.700076346757328, + "learning_rate": 7.988333887090848e-06, + "loss": 0.3466, + "step": 14411 + }, + { + "epoch": 0.31723530706581554, + "grad_norm": 1.2092958841816799, + "learning_rate": 7.98804809251546e-06, + "loss": 0.2802, + "step": 14412 + }, + { + "epoch": 0.3172573189522342, + "grad_norm": 1.3874320250426784, + "learning_rate": 7.987762282753477e-06, + "loss": 0.2861, + "step": 14413 + }, + { + "epoch": 0.3172793308386529, + "grad_norm": 0.6338312011722107, + "learning_rate": 7.987476457806357e-06, + "loss": 0.4967, + "step": 14414 + }, + { + "epoch": 0.31730134272507154, + "grad_norm": 1.7427313161438742, + "learning_rate": 7.987190617675546e-06, + "loss": 0.3231, + "step": 14415 + }, + { + "epoch": 0.3173233546114902, + "grad_norm": 1.2746228647925715, + "learning_rate": 7.986904762362501e-06, + "loss": 0.3738, + "step": 14416 + }, + { + "epoch": 0.3173453664979089, + "grad_norm": 0.6414283955275022, + "learning_rate": 7.986618891868675e-06, + "loss": 0.5074, + "step": 14417 + }, + { + "epoch": 0.31736737838432755, + "grad_norm": 1.3950724872228988, + "learning_rate": 7.986333006195516e-06, + "loss": 0.2997, + "step": 14418 + }, + { + "epoch": 0.3173893902707462, + "grad_norm": 0.6331552259967823, + "learning_rate": 7.986047105344484e-06, + "loss": 0.4906, + "step": 14419 + }, + { + "epoch": 0.31741140215716485, + "grad_norm": 1.387300097096766, + "learning_rate": 7.98576118931703e-06, + "loss": 0.3576, + "step": 14420 + }, + { + "epoch": 0.31743341404358355, + "grad_norm": 0.6109628840746348, + "learning_rate": 7.985475258114603e-06, + "loss": 0.4933, + "step": 14421 + }, + { + "epoch": 0.3174554259300022, + "grad_norm": 2.5808774385673483, + "learning_rate": 7.98518931173866e-06, + "loss": 0.4009, + "step": 14422 + }, + { + "epoch": 0.31747743781642085, + "grad_norm": 1.509270076730244, + "learning_rate": 7.984903350190654e-06, + "loss": 0.3938, + "step": 14423 + }, + { + "epoch": 0.31749944970283955, + "grad_norm": 1.2108153929092331, + "learning_rate": 7.984617373472037e-06, + "loss": 0.2847, + "step": 14424 + }, + { + "epoch": 0.3175214615892582, + "grad_norm": 1.5163790221980504, + "learning_rate": 7.984331381584266e-06, + "loss": 0.3565, + "step": 14425 + }, + { + "epoch": 0.31754347347567685, + "grad_norm": 1.492821903131911, + "learning_rate": 7.984045374528789e-06, + "loss": 0.3564, + "step": 14426 + }, + { + "epoch": 0.31756548536209556, + "grad_norm": 1.2831726913146373, + "learning_rate": 7.983759352307062e-06, + "loss": 0.3194, + "step": 14427 + }, + { + "epoch": 0.3175874972485142, + "grad_norm": 1.6260533057778392, + "learning_rate": 7.98347331492054e-06, + "loss": 0.406, + "step": 14428 + }, + { + "epoch": 0.31760950913493285, + "grad_norm": 1.212656111777055, + "learning_rate": 7.983187262370678e-06, + "loss": 0.3191, + "step": 14429 + }, + { + "epoch": 0.3176315210213515, + "grad_norm": 1.4149477714002872, + "learning_rate": 7.982901194658926e-06, + "loss": 0.3085, + "step": 14430 + }, + { + "epoch": 0.3176535329077702, + "grad_norm": 2.6496280194514825, + "learning_rate": 7.98261511178674e-06, + "loss": 0.2876, + "step": 14431 + }, + { + "epoch": 0.31767554479418886, + "grad_norm": 1.1814608578421346, + "learning_rate": 7.982329013755574e-06, + "loss": 0.3105, + "step": 14432 + }, + { + "epoch": 0.3176975566806075, + "grad_norm": 2.36919372179392, + "learning_rate": 7.98204290056688e-06, + "loss": 0.2988, + "step": 14433 + }, + { + "epoch": 0.3177195685670262, + "grad_norm": 1.1497646384111084, + "learning_rate": 7.981756772222116e-06, + "loss": 0.266, + "step": 14434 + }, + { + "epoch": 0.31774158045344486, + "grad_norm": 0.6748135263586678, + "learning_rate": 7.981470628722734e-06, + "loss": 0.499, + "step": 14435 + }, + { + "epoch": 0.3177635923398635, + "grad_norm": 1.5512646565982802, + "learning_rate": 7.981184470070187e-06, + "loss": 0.2919, + "step": 14436 + }, + { + "epoch": 0.3177856042262822, + "grad_norm": 2.188984613209788, + "learning_rate": 7.980898296265933e-06, + "loss": 0.3012, + "step": 14437 + }, + { + "epoch": 0.31780761611270086, + "grad_norm": 1.3422715279878705, + "learning_rate": 7.980612107311423e-06, + "loss": 0.3987, + "step": 14438 + }, + { + "epoch": 0.3178296279991195, + "grad_norm": 1.7098461474390274, + "learning_rate": 7.980325903208114e-06, + "loss": 0.3889, + "step": 14439 + }, + { + "epoch": 0.3178516398855382, + "grad_norm": 0.6106599872231391, + "learning_rate": 7.980039683957458e-06, + "loss": 0.4911, + "step": 14440 + }, + { + "epoch": 0.31787365177195687, + "grad_norm": 2.0129028615315767, + "learning_rate": 7.979753449560911e-06, + "loss": 0.3543, + "step": 14441 + }, + { + "epoch": 0.3178956636583755, + "grad_norm": 1.4983771012206248, + "learning_rate": 7.97946720001993e-06, + "loss": 0.3272, + "step": 14442 + }, + { + "epoch": 0.31791767554479416, + "grad_norm": 1.1513582043341297, + "learning_rate": 7.979180935335966e-06, + "loss": 0.3836, + "step": 14443 + }, + { + "epoch": 0.31793968743121287, + "grad_norm": 1.6412127221819093, + "learning_rate": 7.978894655510476e-06, + "loss": 0.2954, + "step": 14444 + }, + { + "epoch": 0.3179616993176315, + "grad_norm": 1.1479884802257114, + "learning_rate": 7.978608360544916e-06, + "loss": 0.3278, + "step": 14445 + }, + { + "epoch": 0.31798371120405017, + "grad_norm": 0.6157814118571181, + "learning_rate": 7.978322050440739e-06, + "loss": 0.4978, + "step": 14446 + }, + { + "epoch": 0.3180057230904689, + "grad_norm": 1.4295421109353175, + "learning_rate": 7.9780357251994e-06, + "loss": 0.3617, + "step": 14447 + }, + { + "epoch": 0.3180277349768875, + "grad_norm": 1.275931014369335, + "learning_rate": 7.977749384822357e-06, + "loss": 0.3132, + "step": 14448 + }, + { + "epoch": 0.31804974686330617, + "grad_norm": 1.8193087195494244, + "learning_rate": 7.977463029311062e-06, + "loss": 0.4155, + "step": 14449 + }, + { + "epoch": 0.3180717587497249, + "grad_norm": 0.6114543649939645, + "learning_rate": 7.977176658666972e-06, + "loss": 0.5135, + "step": 14450 + }, + { + "epoch": 0.3180937706361435, + "grad_norm": 1.577821709675106, + "learning_rate": 7.976890272891545e-06, + "loss": 0.3515, + "step": 14451 + }, + { + "epoch": 0.3181157825225622, + "grad_norm": 1.5223697106897691, + "learning_rate": 7.97660387198623e-06, + "loss": 0.3058, + "step": 14452 + }, + { + "epoch": 0.3181377944089808, + "grad_norm": 1.1363630320275424, + "learning_rate": 7.976317455952489e-06, + "loss": 0.2868, + "step": 14453 + }, + { + "epoch": 0.3181598062953995, + "grad_norm": 1.529368229652009, + "learning_rate": 7.976031024791774e-06, + "loss": 0.4326, + "step": 14454 + }, + { + "epoch": 0.3181818181818182, + "grad_norm": 0.6304776382355387, + "learning_rate": 7.975744578505542e-06, + "loss": 0.5201, + "step": 14455 + }, + { + "epoch": 0.3182038300682368, + "grad_norm": 1.0836829100282208, + "learning_rate": 7.975458117095251e-06, + "loss": 0.3032, + "step": 14456 + }, + { + "epoch": 0.31822584195465553, + "grad_norm": 2.0770949914411743, + "learning_rate": 7.975171640562353e-06, + "loss": 0.4524, + "step": 14457 + }, + { + "epoch": 0.3182478538410742, + "grad_norm": 1.2170801520647698, + "learning_rate": 7.974885148908305e-06, + "loss": 0.2868, + "step": 14458 + }, + { + "epoch": 0.31826986572749283, + "grad_norm": 1.456233206915508, + "learning_rate": 7.974598642134564e-06, + "loss": 0.2954, + "step": 14459 + }, + { + "epoch": 0.31829187761391153, + "grad_norm": 1.4657493959043444, + "learning_rate": 7.974312120242587e-06, + "loss": 0.3262, + "step": 14460 + }, + { + "epoch": 0.3183138895003302, + "grad_norm": 1.3092560173497794, + "learning_rate": 7.97402558323383e-06, + "loss": 0.305, + "step": 14461 + }, + { + "epoch": 0.31833590138674883, + "grad_norm": 2.2232516961139734, + "learning_rate": 7.973739031109746e-06, + "loss": 0.3454, + "step": 14462 + }, + { + "epoch": 0.31835791327316754, + "grad_norm": 1.517575825662338, + "learning_rate": 7.973452463871796e-06, + "loss": 0.2543, + "step": 14463 + }, + { + "epoch": 0.3183799251595862, + "grad_norm": 1.6834742339629987, + "learning_rate": 7.973165881521435e-06, + "loss": 0.3132, + "step": 14464 + }, + { + "epoch": 0.31840193704600483, + "grad_norm": 1.5695366999558238, + "learning_rate": 7.972879284060116e-06, + "loss": 0.3477, + "step": 14465 + }, + { + "epoch": 0.3184239489324235, + "grad_norm": 1.4698725725405724, + "learning_rate": 7.972592671489301e-06, + "loss": 0.3407, + "step": 14466 + }, + { + "epoch": 0.3184459608188422, + "grad_norm": 1.1505474911689162, + "learning_rate": 7.972306043810443e-06, + "loss": 0.2376, + "step": 14467 + }, + { + "epoch": 0.31846797270526084, + "grad_norm": 1.5777753212731016, + "learning_rate": 7.972019401024999e-06, + "loss": 0.2949, + "step": 14468 + }, + { + "epoch": 0.3184899845916795, + "grad_norm": 1.924611757524245, + "learning_rate": 7.97173274313443e-06, + "loss": 0.376, + "step": 14469 + }, + { + "epoch": 0.3185119964780982, + "grad_norm": 1.8748350229065682, + "learning_rate": 7.971446070140188e-06, + "loss": 0.37, + "step": 14470 + }, + { + "epoch": 0.31853400836451684, + "grad_norm": 1.7967719080654576, + "learning_rate": 7.971159382043732e-06, + "loss": 0.276, + "step": 14471 + }, + { + "epoch": 0.3185560202509355, + "grad_norm": 1.308265457845612, + "learning_rate": 7.97087267884652e-06, + "loss": 0.2842, + "step": 14472 + }, + { + "epoch": 0.3185780321373542, + "grad_norm": 0.6308708937661518, + "learning_rate": 7.970585960550006e-06, + "loss": 0.5154, + "step": 14473 + }, + { + "epoch": 0.31860004402377284, + "grad_norm": 1.4743313021947744, + "learning_rate": 7.970299227155649e-06, + "loss": 0.289, + "step": 14474 + }, + { + "epoch": 0.3186220559101915, + "grad_norm": 1.9930993985759264, + "learning_rate": 7.970012478664908e-06, + "loss": 0.3084, + "step": 14475 + }, + { + "epoch": 0.31864406779661014, + "grad_norm": 1.6971638212282099, + "learning_rate": 7.969725715079237e-06, + "loss": 0.3688, + "step": 14476 + }, + { + "epoch": 0.31866607968302885, + "grad_norm": 4.002139402631908, + "learning_rate": 7.969438936400098e-06, + "loss": 0.3095, + "step": 14477 + }, + { + "epoch": 0.3186880915694475, + "grad_norm": 5.966329182853122, + "learning_rate": 7.969152142628944e-06, + "loss": 0.3913, + "step": 14478 + }, + { + "epoch": 0.31871010345586614, + "grad_norm": 1.493968384254642, + "learning_rate": 7.968865333767234e-06, + "loss": 0.2706, + "step": 14479 + }, + { + "epoch": 0.31873211534228485, + "grad_norm": 1.4390108496986151, + "learning_rate": 7.968578509816429e-06, + "loss": 0.3534, + "step": 14480 + }, + { + "epoch": 0.3187541272287035, + "grad_norm": 1.3183881584406603, + "learning_rate": 7.968291670777981e-06, + "loss": 0.3597, + "step": 14481 + }, + { + "epoch": 0.31877613911512215, + "grad_norm": 1.8239600108610337, + "learning_rate": 7.968004816653355e-06, + "loss": 0.382, + "step": 14482 + }, + { + "epoch": 0.31879815100154085, + "grad_norm": 2.2569376867227517, + "learning_rate": 7.967717947444e-06, + "loss": 0.2948, + "step": 14483 + }, + { + "epoch": 0.3188201628879595, + "grad_norm": 1.6015838135960803, + "learning_rate": 7.96743106315138e-06, + "loss": 0.2573, + "step": 14484 + }, + { + "epoch": 0.31884217477437815, + "grad_norm": 0.6270330599592563, + "learning_rate": 7.967144163776953e-06, + "loss": 0.4915, + "step": 14485 + }, + { + "epoch": 0.31886418666079686, + "grad_norm": 1.4879763202535878, + "learning_rate": 7.966857249322175e-06, + "loss": 0.4091, + "step": 14486 + }, + { + "epoch": 0.3188861985472155, + "grad_norm": 1.4812874519299426, + "learning_rate": 7.966570319788508e-06, + "loss": 0.3562, + "step": 14487 + }, + { + "epoch": 0.31890821043363415, + "grad_norm": 1.6624731149154848, + "learning_rate": 7.966283375177403e-06, + "loss": 0.3843, + "step": 14488 + }, + { + "epoch": 0.3189302223200528, + "grad_norm": 0.6266789679995269, + "learning_rate": 7.965996415490327e-06, + "loss": 0.5154, + "step": 14489 + }, + { + "epoch": 0.3189522342064715, + "grad_norm": 2.2621313645485324, + "learning_rate": 7.965709440728732e-06, + "loss": 0.2892, + "step": 14490 + }, + { + "epoch": 0.31897424609289016, + "grad_norm": 1.5261364921561296, + "learning_rate": 7.96542245089408e-06, + "loss": 0.2692, + "step": 14491 + }, + { + "epoch": 0.3189962579793088, + "grad_norm": 1.8165659191209835, + "learning_rate": 7.965135445987826e-06, + "loss": 0.3251, + "step": 14492 + }, + { + "epoch": 0.3190182698657275, + "grad_norm": 1.5245828558993386, + "learning_rate": 7.964848426011435e-06, + "loss": 0.2427, + "step": 14493 + }, + { + "epoch": 0.31904028175214616, + "grad_norm": 1.315664387101689, + "learning_rate": 7.96456139096636e-06, + "loss": 0.3055, + "step": 14494 + }, + { + "epoch": 0.3190622936385648, + "grad_norm": 1.6734029475757803, + "learning_rate": 7.96427434085406e-06, + "loss": 0.3697, + "step": 14495 + }, + { + "epoch": 0.3190843055249835, + "grad_norm": 1.282805177942854, + "learning_rate": 7.963987275675997e-06, + "loss": 0.4055, + "step": 14496 + }, + { + "epoch": 0.31910631741140216, + "grad_norm": 1.6395862975221223, + "learning_rate": 7.96370019543363e-06, + "loss": 0.3554, + "step": 14497 + }, + { + "epoch": 0.3191283292978208, + "grad_norm": 1.3689870094322532, + "learning_rate": 7.963413100128416e-06, + "loss": 0.3945, + "step": 14498 + }, + { + "epoch": 0.3191503411842395, + "grad_norm": 1.4836316837643833, + "learning_rate": 7.963125989761816e-06, + "loss": 0.3479, + "step": 14499 + }, + { + "epoch": 0.31917235307065817, + "grad_norm": 1.5510719647127928, + "learning_rate": 7.962838864335286e-06, + "loss": 0.3727, + "step": 14500 + }, + { + "epoch": 0.3191943649570768, + "grad_norm": 1.4920377175289778, + "learning_rate": 7.962551723850287e-06, + "loss": 0.4279, + "step": 14501 + }, + { + "epoch": 0.31921637684349546, + "grad_norm": 1.4134705281482074, + "learning_rate": 7.962264568308282e-06, + "loss": 0.2627, + "step": 14502 + }, + { + "epoch": 0.31923838872991417, + "grad_norm": 1.3706004561497491, + "learning_rate": 7.961977397710725e-06, + "loss": 0.3066, + "step": 14503 + }, + { + "epoch": 0.3192604006163328, + "grad_norm": 1.5001013520532018, + "learning_rate": 7.961690212059077e-06, + "loss": 0.3323, + "step": 14504 + }, + { + "epoch": 0.31928241250275147, + "grad_norm": 1.7171973346320029, + "learning_rate": 7.9614030113548e-06, + "loss": 0.47, + "step": 14505 + }, + { + "epoch": 0.31930442438917017, + "grad_norm": 1.6710152128344777, + "learning_rate": 7.961115795599352e-06, + "loss": 0.2877, + "step": 14506 + }, + { + "epoch": 0.3193264362755888, + "grad_norm": 0.6924947480560555, + "learning_rate": 7.960828564794193e-06, + "loss": 0.4914, + "step": 14507 + }, + { + "epoch": 0.31934844816200747, + "grad_norm": 1.8714623783176305, + "learning_rate": 7.960541318940781e-06, + "loss": 0.38, + "step": 14508 + }, + { + "epoch": 0.3193704600484262, + "grad_norm": 1.9694722770825526, + "learning_rate": 7.96025405804058e-06, + "loss": 0.3964, + "step": 14509 + }, + { + "epoch": 0.3193924719348448, + "grad_norm": 0.650886766644552, + "learning_rate": 7.959966782095045e-06, + "loss": 0.5073, + "step": 14510 + }, + { + "epoch": 0.3194144838212635, + "grad_norm": 2.556466574706449, + "learning_rate": 7.95967949110564e-06, + "loss": 0.3278, + "step": 14511 + }, + { + "epoch": 0.3194364957076821, + "grad_norm": 1.3427444194523737, + "learning_rate": 7.959392185073824e-06, + "loss": 0.3603, + "step": 14512 + }, + { + "epoch": 0.3194585075941008, + "grad_norm": 0.6228651177617964, + "learning_rate": 7.959104864001058e-06, + "loss": 0.5205, + "step": 14513 + }, + { + "epoch": 0.3194805194805195, + "grad_norm": 2.0671986066900123, + "learning_rate": 7.958817527888799e-06, + "loss": 0.3653, + "step": 14514 + }, + { + "epoch": 0.3195025313669381, + "grad_norm": 1.4953654221624813, + "learning_rate": 7.958530176738512e-06, + "loss": 0.3357, + "step": 14515 + }, + { + "epoch": 0.31952454325335683, + "grad_norm": 2.8362581213901707, + "learning_rate": 7.958242810551653e-06, + "loss": 0.4157, + "step": 14516 + }, + { + "epoch": 0.3195465551397755, + "grad_norm": 1.8222408843036846, + "learning_rate": 7.957955429329684e-06, + "loss": 0.3535, + "step": 14517 + }, + { + "epoch": 0.31956856702619413, + "grad_norm": 1.6371700247489314, + "learning_rate": 7.95766803307407e-06, + "loss": 0.4232, + "step": 14518 + }, + { + "epoch": 0.31959057891261283, + "grad_norm": 0.6438135317196495, + "learning_rate": 7.957380621786264e-06, + "loss": 0.4999, + "step": 14519 + }, + { + "epoch": 0.3196125907990315, + "grad_norm": 1.5282467431466418, + "learning_rate": 7.957093195467733e-06, + "loss": 0.3406, + "step": 14520 + }, + { + "epoch": 0.31963460268545013, + "grad_norm": 2.610757876547702, + "learning_rate": 7.956805754119933e-06, + "loss": 0.2523, + "step": 14521 + }, + { + "epoch": 0.31965661457186884, + "grad_norm": 1.8038465220667086, + "learning_rate": 7.956518297744329e-06, + "loss": 0.2928, + "step": 14522 + }, + { + "epoch": 0.3196786264582875, + "grad_norm": 1.4545612440955829, + "learning_rate": 7.95623082634238e-06, + "loss": 0.2984, + "step": 14523 + }, + { + "epoch": 0.31970063834470613, + "grad_norm": 1.302780333346452, + "learning_rate": 7.955943339915549e-06, + "loss": 0.3466, + "step": 14524 + }, + { + "epoch": 0.3197226502311248, + "grad_norm": 2.8027001946093146, + "learning_rate": 7.955655838465295e-06, + "loss": 0.3705, + "step": 14525 + }, + { + "epoch": 0.3197446621175435, + "grad_norm": 1.2584266365286663, + "learning_rate": 7.955368321993077e-06, + "loss": 0.2869, + "step": 14526 + }, + { + "epoch": 0.31976667400396214, + "grad_norm": 1.955717898187128, + "learning_rate": 7.955080790500362e-06, + "loss": 0.4331, + "step": 14527 + }, + { + "epoch": 0.3197886858903808, + "grad_norm": 7.004404574775472, + "learning_rate": 7.954793243988608e-06, + "loss": 0.319, + "step": 14528 + }, + { + "epoch": 0.3198106977767995, + "grad_norm": 0.6631530127055145, + "learning_rate": 7.954505682459277e-06, + "loss": 0.5096, + "step": 14529 + }, + { + "epoch": 0.31983270966321814, + "grad_norm": 1.2254296370060125, + "learning_rate": 7.954218105913828e-06, + "loss": 0.2831, + "step": 14530 + }, + { + "epoch": 0.3198547215496368, + "grad_norm": 1.2840798087579135, + "learning_rate": 7.953930514353727e-06, + "loss": 0.3351, + "step": 14531 + }, + { + "epoch": 0.3198767334360555, + "grad_norm": 0.6402320141416842, + "learning_rate": 7.953642907780435e-06, + "loss": 0.5178, + "step": 14532 + }, + { + "epoch": 0.31989874532247414, + "grad_norm": 1.7070024620650248, + "learning_rate": 7.95335528619541e-06, + "loss": 0.3839, + "step": 14533 + }, + { + "epoch": 0.3199207572088928, + "grad_norm": 1.6882718984673364, + "learning_rate": 7.953067649600118e-06, + "loss": 0.3673, + "step": 14534 + }, + { + "epoch": 0.31994276909531144, + "grad_norm": 1.6403718918608488, + "learning_rate": 7.952779997996018e-06, + "loss": 0.3894, + "step": 14535 + }, + { + "epoch": 0.31996478098173015, + "grad_norm": 1.136750815174122, + "learning_rate": 7.952492331384573e-06, + "loss": 0.3279, + "step": 14536 + }, + { + "epoch": 0.3199867928681488, + "grad_norm": 0.6302574770446179, + "learning_rate": 7.952204649767245e-06, + "loss": 0.4928, + "step": 14537 + }, + { + "epoch": 0.32000880475456744, + "grad_norm": 2.225803040798033, + "learning_rate": 7.951916953145497e-06, + "loss": 0.3107, + "step": 14538 + }, + { + "epoch": 0.32003081664098615, + "grad_norm": 1.335187726002736, + "learning_rate": 7.95162924152079e-06, + "loss": 0.3452, + "step": 14539 + }, + { + "epoch": 0.3200528285274048, + "grad_norm": 1.4308893546351888, + "learning_rate": 7.951341514894586e-06, + "loss": 0.3666, + "step": 14540 + }, + { + "epoch": 0.32007484041382345, + "grad_norm": 1.5031368699673855, + "learning_rate": 7.951053773268348e-06, + "loss": 0.3461, + "step": 14541 + }, + { + "epoch": 0.32009685230024215, + "grad_norm": 0.6163718832030968, + "learning_rate": 7.95076601664354e-06, + "loss": 0.4882, + "step": 14542 + }, + { + "epoch": 0.3201188641866608, + "grad_norm": 0.6099183817335759, + "learning_rate": 7.950478245021623e-06, + "loss": 0.4797, + "step": 14543 + }, + { + "epoch": 0.32014087607307945, + "grad_norm": 1.2669726474828948, + "learning_rate": 7.95019045840406e-06, + "loss": 0.3046, + "step": 14544 + }, + { + "epoch": 0.32016288795949815, + "grad_norm": 1.4763256677120435, + "learning_rate": 7.949902656792312e-06, + "loss": 0.3662, + "step": 14545 + }, + { + "epoch": 0.3201848998459168, + "grad_norm": 1.3371944258726205, + "learning_rate": 7.949614840187843e-06, + "loss": 0.4025, + "step": 14546 + }, + { + "epoch": 0.32020691173233545, + "grad_norm": 1.2278995012211291, + "learning_rate": 7.949327008592117e-06, + "loss": 0.2631, + "step": 14547 + }, + { + "epoch": 0.3202289236187541, + "grad_norm": 1.6366257751823519, + "learning_rate": 7.949039162006592e-06, + "loss": 0.3398, + "step": 14548 + }, + { + "epoch": 0.3202509355051728, + "grad_norm": 1.8033405574908445, + "learning_rate": 7.948751300432739e-06, + "loss": 0.2839, + "step": 14549 + }, + { + "epoch": 0.32027294739159146, + "grad_norm": 1.5392613251755665, + "learning_rate": 7.948463423872015e-06, + "loss": 0.4036, + "step": 14550 + }, + { + "epoch": 0.3202949592780101, + "grad_norm": 1.676863699393217, + "learning_rate": 7.948175532325885e-06, + "loss": 0.3465, + "step": 14551 + }, + { + "epoch": 0.3203169711644288, + "grad_norm": 1.5176606004691886, + "learning_rate": 7.947887625795812e-06, + "loss": 0.2669, + "step": 14552 + }, + { + "epoch": 0.32033898305084746, + "grad_norm": 1.7677494035455243, + "learning_rate": 7.94759970428326e-06, + "loss": 0.4016, + "step": 14553 + }, + { + "epoch": 0.3203609949372661, + "grad_norm": 1.82942461298375, + "learning_rate": 7.94731176778969e-06, + "loss": 0.3099, + "step": 14554 + }, + { + "epoch": 0.3203830068236848, + "grad_norm": 1.7293186244579593, + "learning_rate": 7.947023816316568e-06, + "loss": 0.3726, + "step": 14555 + }, + { + "epoch": 0.32040501871010346, + "grad_norm": 1.3183018968169817, + "learning_rate": 7.946735849865356e-06, + "loss": 0.3045, + "step": 14556 + }, + { + "epoch": 0.3204270305965221, + "grad_norm": 0.6827158774430896, + "learning_rate": 7.94644786843752e-06, + "loss": 0.5149, + "step": 14557 + }, + { + "epoch": 0.3204490424829408, + "grad_norm": 1.1325553600595628, + "learning_rate": 7.94615987203452e-06, + "loss": 0.2956, + "step": 14558 + }, + { + "epoch": 0.32047105436935946, + "grad_norm": 1.8002279675800072, + "learning_rate": 7.945871860657821e-06, + "loss": 0.3513, + "step": 14559 + }, + { + "epoch": 0.3204930662557781, + "grad_norm": 1.6415553337289668, + "learning_rate": 7.94558383430889e-06, + "loss": 0.3, + "step": 14560 + }, + { + "epoch": 0.32051507814219676, + "grad_norm": 0.6161115658816737, + "learning_rate": 7.945295792989185e-06, + "loss": 0.5035, + "step": 14561 + }, + { + "epoch": 0.32053709002861547, + "grad_norm": 0.6073836519376272, + "learning_rate": 7.945007736700175e-06, + "loss": 0.4998, + "step": 14562 + }, + { + "epoch": 0.3205591019150341, + "grad_norm": 1.7369105072958737, + "learning_rate": 7.944719665443323e-06, + "loss": 0.3182, + "step": 14563 + }, + { + "epoch": 0.32058111380145277, + "grad_norm": 1.443341025777491, + "learning_rate": 7.94443157922009e-06, + "loss": 0.2778, + "step": 14564 + }, + { + "epoch": 0.32060312568787147, + "grad_norm": 1.3750127774651082, + "learning_rate": 7.944143478031944e-06, + "loss": 0.3368, + "step": 14565 + }, + { + "epoch": 0.3206251375742901, + "grad_norm": 0.624379727211597, + "learning_rate": 7.943855361880346e-06, + "loss": 0.5216, + "step": 14566 + }, + { + "epoch": 0.32064714946070877, + "grad_norm": 1.6352778811384785, + "learning_rate": 7.943567230766765e-06, + "loss": 0.3331, + "step": 14567 + }, + { + "epoch": 0.3206691613471275, + "grad_norm": 1.8600169474273969, + "learning_rate": 7.943279084692661e-06, + "loss": 0.3986, + "step": 14568 + }, + { + "epoch": 0.3206911732335461, + "grad_norm": 1.6242602864753641, + "learning_rate": 7.9429909236595e-06, + "loss": 0.4523, + "step": 14569 + }, + { + "epoch": 0.32071318511996477, + "grad_norm": 1.478854876835805, + "learning_rate": 7.942702747668745e-06, + "loss": 0.3768, + "step": 14570 + }, + { + "epoch": 0.3207351970063834, + "grad_norm": 0.6319544669343218, + "learning_rate": 7.942414556721863e-06, + "loss": 0.4884, + "step": 14571 + }, + { + "epoch": 0.3207572088928021, + "grad_norm": 1.5168114484921869, + "learning_rate": 7.942126350820319e-06, + "loss": 0.2612, + "step": 14572 + }, + { + "epoch": 0.3207792207792208, + "grad_norm": 1.8579122843819602, + "learning_rate": 7.941838129965576e-06, + "loss": 0.3682, + "step": 14573 + }, + { + "epoch": 0.3208012326656394, + "grad_norm": 2.2464984552707525, + "learning_rate": 7.9415498941591e-06, + "loss": 0.2408, + "step": 14574 + }, + { + "epoch": 0.32082324455205813, + "grad_norm": 1.6115375155940506, + "learning_rate": 7.941261643402354e-06, + "loss": 0.3096, + "step": 14575 + }, + { + "epoch": 0.3208452564384768, + "grad_norm": 1.3548156324175091, + "learning_rate": 7.940973377696806e-06, + "loss": 0.3171, + "step": 14576 + }, + { + "epoch": 0.3208672683248954, + "grad_norm": 1.4360737248028892, + "learning_rate": 7.940685097043918e-06, + "loss": 0.2431, + "step": 14577 + }, + { + "epoch": 0.32088928021131413, + "grad_norm": 1.37291266215719, + "learning_rate": 7.940396801445158e-06, + "loss": 0.3622, + "step": 14578 + }, + { + "epoch": 0.3209112920977328, + "grad_norm": 1.7978470030608018, + "learning_rate": 7.94010849090199e-06, + "loss": 0.4394, + "step": 14579 + }, + { + "epoch": 0.32093330398415143, + "grad_norm": 1.5226003547047793, + "learning_rate": 7.93982016541588e-06, + "loss": 0.343, + "step": 14580 + }, + { + "epoch": 0.32095531587057013, + "grad_norm": 1.5729298550613433, + "learning_rate": 7.939531824988292e-06, + "loss": 0.3964, + "step": 14581 + }, + { + "epoch": 0.3209773277569888, + "grad_norm": 1.6087243436725522, + "learning_rate": 7.939243469620692e-06, + "loss": 0.3637, + "step": 14582 + }, + { + "epoch": 0.32099933964340743, + "grad_norm": 1.3696280222069654, + "learning_rate": 7.938955099314545e-06, + "loss": 0.388, + "step": 14583 + }, + { + "epoch": 0.3210213515298261, + "grad_norm": 1.988739574836569, + "learning_rate": 7.938666714071319e-06, + "loss": 0.3736, + "step": 14584 + }, + { + "epoch": 0.3210433634162448, + "grad_norm": 8.30889664363312, + "learning_rate": 7.938378313892477e-06, + "loss": 0.2798, + "step": 14585 + }, + { + "epoch": 0.32106537530266344, + "grad_norm": 1.4066060037518286, + "learning_rate": 7.938089898779485e-06, + "loss": 0.3473, + "step": 14586 + }, + { + "epoch": 0.3210873871890821, + "grad_norm": 1.6446644526532523, + "learning_rate": 7.93780146873381e-06, + "loss": 0.3222, + "step": 14587 + }, + { + "epoch": 0.3211093990755008, + "grad_norm": 1.3415328214568654, + "learning_rate": 7.93751302375692e-06, + "loss": 0.3062, + "step": 14588 + }, + { + "epoch": 0.32113141096191944, + "grad_norm": 1.7324862931202325, + "learning_rate": 7.937224563850277e-06, + "loss": 0.3233, + "step": 14589 + }, + { + "epoch": 0.3211534228483381, + "grad_norm": 2.2689006841670754, + "learning_rate": 7.936936089015348e-06, + "loss": 0.3343, + "step": 14590 + }, + { + "epoch": 0.3211754347347568, + "grad_norm": 0.6302469713565212, + "learning_rate": 7.9366475992536e-06, + "loss": 0.4927, + "step": 14591 + }, + { + "epoch": 0.32119744662117544, + "grad_norm": 1.6090880336632345, + "learning_rate": 7.936359094566499e-06, + "loss": 0.3219, + "step": 14592 + }, + { + "epoch": 0.3212194585075941, + "grad_norm": 1.3472353001062727, + "learning_rate": 7.936070574955514e-06, + "loss": 0.3041, + "step": 14593 + }, + { + "epoch": 0.32124147039401274, + "grad_norm": 4.2176098665324275, + "learning_rate": 7.935782040422107e-06, + "loss": 0.2989, + "step": 14594 + }, + { + "epoch": 0.32126348228043144, + "grad_norm": 1.6213107438160441, + "learning_rate": 7.935493490967747e-06, + "loss": 0.3596, + "step": 14595 + }, + { + "epoch": 0.3212854941668501, + "grad_norm": 1.46287386339821, + "learning_rate": 7.935204926593897e-06, + "loss": 0.326, + "step": 14596 + }, + { + "epoch": 0.32130750605326874, + "grad_norm": 3.0064138638566784, + "learning_rate": 7.93491634730203e-06, + "loss": 0.3281, + "step": 14597 + }, + { + "epoch": 0.32132951793968745, + "grad_norm": 1.657958258191608, + "learning_rate": 7.934627753093606e-06, + "loss": 0.3384, + "step": 14598 + }, + { + "epoch": 0.3213515298261061, + "grad_norm": 1.9366449668549117, + "learning_rate": 7.934339143970098e-06, + "loss": 0.3405, + "step": 14599 + }, + { + "epoch": 0.32137354171252475, + "grad_norm": 0.6440221755884344, + "learning_rate": 7.934050519932968e-06, + "loss": 0.524, + "step": 14600 + }, + { + "epoch": 0.32139555359894345, + "grad_norm": 0.621575986533067, + "learning_rate": 7.933761880983683e-06, + "loss": 0.5244, + "step": 14601 + }, + { + "epoch": 0.3214175654853621, + "grad_norm": 1.4776391064301655, + "learning_rate": 7.933473227123715e-06, + "loss": 0.4503, + "step": 14602 + }, + { + "epoch": 0.32143957737178075, + "grad_norm": 1.5489248465259964, + "learning_rate": 7.933184558354527e-06, + "loss": 0.3463, + "step": 14603 + }, + { + "epoch": 0.32146158925819945, + "grad_norm": 1.6045256770907736, + "learning_rate": 7.932895874677587e-06, + "loss": 0.411, + "step": 14604 + }, + { + "epoch": 0.3214836011446181, + "grad_norm": 2.5326739822432747, + "learning_rate": 7.93260717609436e-06, + "loss": 0.3852, + "step": 14605 + }, + { + "epoch": 0.32150561303103675, + "grad_norm": 1.2738207460188133, + "learning_rate": 7.932318462606317e-06, + "loss": 0.3996, + "step": 14606 + }, + { + "epoch": 0.3215276249174554, + "grad_norm": 5.423381145089028, + "learning_rate": 7.932029734214922e-06, + "loss": 0.3477, + "step": 14607 + }, + { + "epoch": 0.3215496368038741, + "grad_norm": 1.247837320890362, + "learning_rate": 7.931740990921646e-06, + "loss": 0.3015, + "step": 14608 + }, + { + "epoch": 0.32157164869029276, + "grad_norm": 2.135832262441782, + "learning_rate": 7.931452232727954e-06, + "loss": 0.2869, + "step": 14609 + }, + { + "epoch": 0.3215936605767114, + "grad_norm": 1.3178840188588459, + "learning_rate": 7.931163459635314e-06, + "loss": 0.3283, + "step": 14610 + }, + { + "epoch": 0.3216156724631301, + "grad_norm": 1.7138054528218634, + "learning_rate": 7.930874671645195e-06, + "loss": 0.2718, + "step": 14611 + }, + { + "epoch": 0.32163768434954876, + "grad_norm": 1.8597680577861762, + "learning_rate": 7.930585868759062e-06, + "loss": 0.4161, + "step": 14612 + }, + { + "epoch": 0.3216596962359674, + "grad_norm": 1.7311770447517978, + "learning_rate": 7.930297050978386e-06, + "loss": 0.2309, + "step": 14613 + }, + { + "epoch": 0.3216817081223861, + "grad_norm": 0.6755083410671294, + "learning_rate": 7.930008218304633e-06, + "loss": 0.4993, + "step": 14614 + }, + { + "epoch": 0.32170372000880476, + "grad_norm": 1.473198672343142, + "learning_rate": 7.92971937073927e-06, + "loss": 0.3184, + "step": 14615 + }, + { + "epoch": 0.3217257318952234, + "grad_norm": 1.5814713629768544, + "learning_rate": 7.929430508283768e-06, + "loss": 0.2855, + "step": 14616 + }, + { + "epoch": 0.32174774378164206, + "grad_norm": 2.13478258782548, + "learning_rate": 7.929141630939595e-06, + "loss": 0.3294, + "step": 14617 + }, + { + "epoch": 0.32176975566806076, + "grad_norm": 1.232789318772858, + "learning_rate": 7.928852738708214e-06, + "loss": 0.2815, + "step": 14618 + }, + { + "epoch": 0.3217917675544794, + "grad_norm": 1.6336713031562784, + "learning_rate": 7.9285638315911e-06, + "loss": 0.3445, + "step": 14619 + }, + { + "epoch": 0.32181377944089806, + "grad_norm": 2.4642125422409027, + "learning_rate": 7.928274909589719e-06, + "loss": 0.3003, + "step": 14620 + }, + { + "epoch": 0.32183579132731677, + "grad_norm": 3.27423705418632, + "learning_rate": 7.927985972705538e-06, + "loss": 0.4222, + "step": 14621 + }, + { + "epoch": 0.3218578032137354, + "grad_norm": 1.2127081208436876, + "learning_rate": 7.927697020940025e-06, + "loss": 0.1933, + "step": 14622 + }, + { + "epoch": 0.32187981510015407, + "grad_norm": 1.543090966029172, + "learning_rate": 7.927408054294651e-06, + "loss": 0.3441, + "step": 14623 + }, + { + "epoch": 0.32190182698657277, + "grad_norm": 2.115112159964621, + "learning_rate": 7.927119072770883e-06, + "loss": 0.3333, + "step": 14624 + }, + { + "epoch": 0.3219238388729914, + "grad_norm": 0.6234196790401598, + "learning_rate": 7.926830076370192e-06, + "loss": 0.5244, + "step": 14625 + }, + { + "epoch": 0.32194585075941007, + "grad_norm": 1.374935198582762, + "learning_rate": 7.926541065094045e-06, + "loss": 0.2983, + "step": 14626 + }, + { + "epoch": 0.3219678626458288, + "grad_norm": 1.5141175444633053, + "learning_rate": 7.92625203894391e-06, + "loss": 0.3126, + "step": 14627 + }, + { + "epoch": 0.3219898745322474, + "grad_norm": 1.2398637642638082, + "learning_rate": 7.925962997921255e-06, + "loss": 0.2626, + "step": 14628 + }, + { + "epoch": 0.32201188641866607, + "grad_norm": 1.5347178183247447, + "learning_rate": 7.925673942027554e-06, + "loss": 0.3641, + "step": 14629 + }, + { + "epoch": 0.3220338983050847, + "grad_norm": 1.7051992783110126, + "learning_rate": 7.925384871264273e-06, + "loss": 0.3176, + "step": 14630 + }, + { + "epoch": 0.3220559101915034, + "grad_norm": 1.9727968856928138, + "learning_rate": 7.925095785632878e-06, + "loss": 0.4044, + "step": 14631 + }, + { + "epoch": 0.3220779220779221, + "grad_norm": 1.577113106811467, + "learning_rate": 7.924806685134845e-06, + "loss": 0.3255, + "step": 14632 + }, + { + "epoch": 0.3220999339643407, + "grad_norm": 1.9906917254431766, + "learning_rate": 7.92451756977164e-06, + "loss": 0.2863, + "step": 14633 + }, + { + "epoch": 0.32212194585075943, + "grad_norm": 1.6234591607422775, + "learning_rate": 7.924228439544732e-06, + "loss": 0.3614, + "step": 14634 + }, + { + "epoch": 0.3221439577371781, + "grad_norm": 1.8521059767870867, + "learning_rate": 7.923939294455589e-06, + "loss": 0.4247, + "step": 14635 + }, + { + "epoch": 0.3221659696235967, + "grad_norm": 1.7596105737473609, + "learning_rate": 7.923650134505684e-06, + "loss": 0.2829, + "step": 14636 + }, + { + "epoch": 0.32218798151001543, + "grad_norm": 1.4211572512568182, + "learning_rate": 7.923360959696484e-06, + "loss": 0.2342, + "step": 14637 + }, + { + "epoch": 0.3222099933964341, + "grad_norm": 1.285703148970678, + "learning_rate": 7.923071770029461e-06, + "loss": 0.3433, + "step": 14638 + }, + { + "epoch": 0.32223200528285273, + "grad_norm": 4.59395256376375, + "learning_rate": 7.922782565506083e-06, + "loss": 0.3758, + "step": 14639 + }, + { + "epoch": 0.32225401716927143, + "grad_norm": 1.6989267557647263, + "learning_rate": 7.92249334612782e-06, + "loss": 0.3863, + "step": 14640 + }, + { + "epoch": 0.3222760290556901, + "grad_norm": 0.6489794364213329, + "learning_rate": 7.922204111896143e-06, + "loss": 0.5015, + "step": 14641 + }, + { + "epoch": 0.32229804094210873, + "grad_norm": 1.6354609451430302, + "learning_rate": 7.921914862812522e-06, + "loss": 0.2766, + "step": 14642 + }, + { + "epoch": 0.3223200528285274, + "grad_norm": 1.570596811044168, + "learning_rate": 7.921625598878424e-06, + "loss": 0.3151, + "step": 14643 + }, + { + "epoch": 0.3223420647149461, + "grad_norm": 1.8045673963120061, + "learning_rate": 7.921336320095323e-06, + "loss": 0.4101, + "step": 14644 + }, + { + "epoch": 0.32236407660136474, + "grad_norm": 1.1498059395747862, + "learning_rate": 7.921047026464688e-06, + "loss": 0.2977, + "step": 14645 + }, + { + "epoch": 0.3223860884877834, + "grad_norm": 1.5430864363862735, + "learning_rate": 7.92075771798799e-06, + "loss": 0.3214, + "step": 14646 + }, + { + "epoch": 0.3224081003742021, + "grad_norm": 1.969055833358798, + "learning_rate": 7.920468394666698e-06, + "loss": 0.2875, + "step": 14647 + }, + { + "epoch": 0.32243011226062074, + "grad_norm": 1.3596061258910068, + "learning_rate": 7.920179056502283e-06, + "loss": 0.2847, + "step": 14648 + }, + { + "epoch": 0.3224521241470394, + "grad_norm": 0.6616632119431123, + "learning_rate": 7.919889703496215e-06, + "loss": 0.5002, + "step": 14649 + }, + { + "epoch": 0.3224741360334581, + "grad_norm": 1.146809994270193, + "learning_rate": 7.919600335649964e-06, + "loss": 0.3032, + "step": 14650 + }, + { + "epoch": 0.32249614791987674, + "grad_norm": 2.624511843355476, + "learning_rate": 7.919310952965005e-06, + "loss": 0.379, + "step": 14651 + }, + { + "epoch": 0.3225181598062954, + "grad_norm": 1.58845469426925, + "learning_rate": 7.919021555442805e-06, + "loss": 0.2735, + "step": 14652 + }, + { + "epoch": 0.32254017169271404, + "grad_norm": 1.2847491459714468, + "learning_rate": 7.918732143084834e-06, + "loss": 0.2609, + "step": 14653 + }, + { + "epoch": 0.32256218357913274, + "grad_norm": 1.471136612817291, + "learning_rate": 7.918442715892566e-06, + "loss": 0.3441, + "step": 14654 + }, + { + "epoch": 0.3225841954655514, + "grad_norm": 0.6114072452650797, + "learning_rate": 7.918153273867468e-06, + "loss": 0.5098, + "step": 14655 + }, + { + "epoch": 0.32260620735197004, + "grad_norm": 1.5920470992365563, + "learning_rate": 7.917863817011016e-06, + "loss": 0.3571, + "step": 14656 + }, + { + "epoch": 0.32262821923838875, + "grad_norm": 2.7199026403406132, + "learning_rate": 7.91757434532468e-06, + "loss": 0.4011, + "step": 14657 + }, + { + "epoch": 0.3226502311248074, + "grad_norm": 1.7356741952418184, + "learning_rate": 7.917284858809926e-06, + "loss": 0.3167, + "step": 14658 + }, + { + "epoch": 0.32267224301122605, + "grad_norm": 1.3837306930697613, + "learning_rate": 7.916995357468231e-06, + "loss": 0.3077, + "step": 14659 + }, + { + "epoch": 0.32269425489764475, + "grad_norm": 0.6254204723612284, + "learning_rate": 7.916705841301064e-06, + "loss": 0.522, + "step": 14660 + }, + { + "epoch": 0.3227162667840634, + "grad_norm": 2.87144438168846, + "learning_rate": 7.916416310309898e-06, + "loss": 0.3416, + "step": 14661 + }, + { + "epoch": 0.32273827867048205, + "grad_norm": 0.6170027512197034, + "learning_rate": 7.916126764496203e-06, + "loss": 0.5128, + "step": 14662 + }, + { + "epoch": 0.32276029055690075, + "grad_norm": 1.1573994257511824, + "learning_rate": 7.915837203861451e-06, + "loss": 0.2924, + "step": 14663 + }, + { + "epoch": 0.3227823024433194, + "grad_norm": 1.2852606066168375, + "learning_rate": 7.915547628407114e-06, + "loss": 0.2799, + "step": 14664 + }, + { + "epoch": 0.32280431432973805, + "grad_norm": 1.8456082483625251, + "learning_rate": 7.915258038134665e-06, + "loss": 0.3166, + "step": 14665 + }, + { + "epoch": 0.3228263262161567, + "grad_norm": 0.6136482800399492, + "learning_rate": 7.914968433045573e-06, + "loss": 0.5, + "step": 14666 + }, + { + "epoch": 0.3228483381025754, + "grad_norm": 0.6206572278453657, + "learning_rate": 7.914678813141311e-06, + "loss": 0.4919, + "step": 14667 + }, + { + "epoch": 0.32287034998899405, + "grad_norm": 4.0994888792970805, + "learning_rate": 7.914389178423352e-06, + "loss": 0.3508, + "step": 14668 + }, + { + "epoch": 0.3228923618754127, + "grad_norm": 0.6084398704612874, + "learning_rate": 7.914099528893166e-06, + "loss": 0.4971, + "step": 14669 + }, + { + "epoch": 0.3229143737618314, + "grad_norm": 1.9424320302448526, + "learning_rate": 7.913809864552228e-06, + "loss": 0.3086, + "step": 14670 + }, + { + "epoch": 0.32293638564825006, + "grad_norm": 1.4232184716626368, + "learning_rate": 7.913520185402008e-06, + "loss": 0.2995, + "step": 14671 + }, + { + "epoch": 0.3229583975346687, + "grad_norm": 1.9395603153803866, + "learning_rate": 7.913230491443979e-06, + "loss": 0.387, + "step": 14672 + }, + { + "epoch": 0.3229804094210874, + "grad_norm": 1.3438118920966868, + "learning_rate": 7.912940782679615e-06, + "loss": 0.3216, + "step": 14673 + }, + { + "epoch": 0.32300242130750606, + "grad_norm": 2.0272262913660573, + "learning_rate": 7.912651059110385e-06, + "loss": 0.2633, + "step": 14674 + }, + { + "epoch": 0.3230244331939247, + "grad_norm": 1.697151408864604, + "learning_rate": 7.912361320737764e-06, + "loss": 0.4096, + "step": 14675 + }, + { + "epoch": 0.32304644508034336, + "grad_norm": 1.9721165668012102, + "learning_rate": 7.912071567563223e-06, + "loss": 0.2962, + "step": 14676 + }, + { + "epoch": 0.32306845696676206, + "grad_norm": 2.5099527486176934, + "learning_rate": 7.911781799588234e-06, + "loss": 0.269, + "step": 14677 + }, + { + "epoch": 0.3230904688531807, + "grad_norm": 1.6371565876358565, + "learning_rate": 7.911492016814273e-06, + "loss": 0.3448, + "step": 14678 + }, + { + "epoch": 0.32311248073959936, + "grad_norm": 13.020645766587908, + "learning_rate": 7.911202219242813e-06, + "loss": 0.3075, + "step": 14679 + }, + { + "epoch": 0.32313449262601807, + "grad_norm": 1.8420797904654063, + "learning_rate": 7.910912406875322e-06, + "loss": 0.3447, + "step": 14680 + }, + { + "epoch": 0.3231565045124367, + "grad_norm": 4.420275363218661, + "learning_rate": 7.910622579713276e-06, + "loss": 0.378, + "step": 14681 + }, + { + "epoch": 0.32317851639885536, + "grad_norm": 15.726076920874274, + "learning_rate": 7.910332737758149e-06, + "loss": 0.2769, + "step": 14682 + }, + { + "epoch": 0.32320052828527407, + "grad_norm": 0.6519970555921077, + "learning_rate": 7.910042881011413e-06, + "loss": 0.5118, + "step": 14683 + }, + { + "epoch": 0.3232225401716927, + "grad_norm": 2.843609642404291, + "learning_rate": 7.909753009474541e-06, + "loss": 0.2641, + "step": 14684 + }, + { + "epoch": 0.32324455205811137, + "grad_norm": 2.0716402781804835, + "learning_rate": 7.909463123149005e-06, + "loss": 0.3591, + "step": 14685 + }, + { + "epoch": 0.32326656394453007, + "grad_norm": 4.093879033878665, + "learning_rate": 7.909173222036282e-06, + "loss": 0.3579, + "step": 14686 + }, + { + "epoch": 0.3232885758309487, + "grad_norm": 2.277346238383315, + "learning_rate": 7.908883306137841e-06, + "loss": 0.4205, + "step": 14687 + }, + { + "epoch": 0.32331058771736737, + "grad_norm": 1.7154522862065864, + "learning_rate": 7.908593375455158e-06, + "loss": 0.3634, + "step": 14688 + }, + { + "epoch": 0.323332599603786, + "grad_norm": 1.3845186271866787, + "learning_rate": 7.908303429989708e-06, + "loss": 0.3617, + "step": 14689 + }, + { + "epoch": 0.3233546114902047, + "grad_norm": 6.716954994250811, + "learning_rate": 7.908013469742961e-06, + "loss": 0.4071, + "step": 14690 + }, + { + "epoch": 0.3233766233766234, + "grad_norm": 1.590357672066645, + "learning_rate": 7.907723494716393e-06, + "loss": 0.3563, + "step": 14691 + }, + { + "epoch": 0.323398635263042, + "grad_norm": 1.9991716804869968, + "learning_rate": 7.907433504911478e-06, + "loss": 0.3931, + "step": 14692 + }, + { + "epoch": 0.3234206471494607, + "grad_norm": 1.6306512449905102, + "learning_rate": 7.90714350032969e-06, + "loss": 0.2807, + "step": 14693 + }, + { + "epoch": 0.3234426590358794, + "grad_norm": 0.6434556486352329, + "learning_rate": 7.9068534809725e-06, + "loss": 0.5121, + "step": 14694 + }, + { + "epoch": 0.323464670922298, + "grad_norm": 1.5374414935312986, + "learning_rate": 7.906563446841385e-06, + "loss": 0.3438, + "step": 14695 + }, + { + "epoch": 0.32348668280871673, + "grad_norm": 1.8214067693017553, + "learning_rate": 7.906273397937819e-06, + "loss": 0.3597, + "step": 14696 + }, + { + "epoch": 0.3235086946951354, + "grad_norm": 1.2894929035959941, + "learning_rate": 7.905983334263274e-06, + "loss": 0.339, + "step": 14697 + }, + { + "epoch": 0.32353070658155403, + "grad_norm": 2.0595184504453594, + "learning_rate": 7.905693255819228e-06, + "loss": 0.4216, + "step": 14698 + }, + { + "epoch": 0.3235527184679727, + "grad_norm": 1.2913151294355192, + "learning_rate": 7.905403162607148e-06, + "loss": 0.2234, + "step": 14699 + }, + { + "epoch": 0.3235747303543914, + "grad_norm": 1.4050144284980504, + "learning_rate": 7.905113054628516e-06, + "loss": 0.3455, + "step": 14700 + }, + { + "epoch": 0.32359674224081003, + "grad_norm": 1.4549742330710547, + "learning_rate": 7.904822931884806e-06, + "loss": 0.3994, + "step": 14701 + }, + { + "epoch": 0.3236187541272287, + "grad_norm": 0.631339971914105, + "learning_rate": 7.90453279437749e-06, + "loss": 0.517, + "step": 14702 + }, + { + "epoch": 0.3236407660136474, + "grad_norm": 1.6384972227156065, + "learning_rate": 7.904242642108038e-06, + "loss": 0.2979, + "step": 14703 + }, + { + "epoch": 0.32366277790006603, + "grad_norm": 0.6275706256438042, + "learning_rate": 7.903952475077934e-06, + "loss": 0.5146, + "step": 14704 + }, + { + "epoch": 0.3236847897864847, + "grad_norm": 1.3383346529881428, + "learning_rate": 7.903662293288646e-06, + "loss": 0.3587, + "step": 14705 + }, + { + "epoch": 0.3237068016729034, + "grad_norm": 0.5955467767409707, + "learning_rate": 7.903372096741652e-06, + "loss": 0.5293, + "step": 14706 + }, + { + "epoch": 0.32372881355932204, + "grad_norm": 1.6088086345394208, + "learning_rate": 7.903081885438428e-06, + "loss": 0.345, + "step": 14707 + }, + { + "epoch": 0.3237508254457407, + "grad_norm": 0.6254012689019638, + "learning_rate": 7.902791659380445e-06, + "loss": 0.4963, + "step": 14708 + }, + { + "epoch": 0.3237728373321594, + "grad_norm": 1.9811698318752675, + "learning_rate": 7.90250141856918e-06, + "loss": 0.3175, + "step": 14709 + }, + { + "epoch": 0.32379484921857804, + "grad_norm": 3.165496076862607, + "learning_rate": 7.902211163006109e-06, + "loss": 0.3023, + "step": 14710 + }, + { + "epoch": 0.3238168611049967, + "grad_norm": 6.685179715031213, + "learning_rate": 7.901920892692707e-06, + "loss": 0.3442, + "step": 14711 + }, + { + "epoch": 0.32383887299141534, + "grad_norm": 0.6323122205641921, + "learning_rate": 7.901630607630446e-06, + "loss": 0.4877, + "step": 14712 + }, + { + "epoch": 0.32386088487783404, + "grad_norm": 0.6333706837163661, + "learning_rate": 7.901340307820807e-06, + "loss": 0.5344, + "step": 14713 + }, + { + "epoch": 0.3238828967642527, + "grad_norm": 1.4610270297776868, + "learning_rate": 7.901049993265261e-06, + "loss": 0.3353, + "step": 14714 + }, + { + "epoch": 0.32390490865067134, + "grad_norm": 0.6129847383069874, + "learning_rate": 7.900759663965285e-06, + "loss": 0.5012, + "step": 14715 + }, + { + "epoch": 0.32392692053709005, + "grad_norm": 1.1825626561475673, + "learning_rate": 7.900469319922356e-06, + "loss": 0.2803, + "step": 14716 + }, + { + "epoch": 0.3239489324235087, + "grad_norm": 1.377714698198717, + "learning_rate": 7.900178961137947e-06, + "loss": 0.3779, + "step": 14717 + }, + { + "epoch": 0.32397094430992734, + "grad_norm": 0.6091431904165144, + "learning_rate": 7.899888587613536e-06, + "loss": 0.5138, + "step": 14718 + }, + { + "epoch": 0.32399295619634605, + "grad_norm": 1.7113358737853925, + "learning_rate": 7.899598199350596e-06, + "loss": 0.274, + "step": 14719 + }, + { + "epoch": 0.3240149680827647, + "grad_norm": 1.8350011203000316, + "learning_rate": 7.899307796350607e-06, + "loss": 0.3576, + "step": 14720 + }, + { + "epoch": 0.32403697996918335, + "grad_norm": 1.6757827767477895, + "learning_rate": 7.899017378615042e-06, + "loss": 0.3805, + "step": 14721 + }, + { + "epoch": 0.32405899185560205, + "grad_norm": 2.8411114561404447, + "learning_rate": 7.898726946145376e-06, + "loss": 0.3279, + "step": 14722 + }, + { + "epoch": 0.3240810037420207, + "grad_norm": 5.78009021355711, + "learning_rate": 7.898436498943088e-06, + "loss": 0.3557, + "step": 14723 + }, + { + "epoch": 0.32410301562843935, + "grad_norm": 1.1263796679197917, + "learning_rate": 7.898146037009653e-06, + "loss": 0.3041, + "step": 14724 + }, + { + "epoch": 0.324125027514858, + "grad_norm": 1.2406826169209142, + "learning_rate": 7.897855560346548e-06, + "loss": 0.2625, + "step": 14725 + }, + { + "epoch": 0.3241470394012767, + "grad_norm": 1.4672053798884137, + "learning_rate": 7.897565068955248e-06, + "loss": 0.3398, + "step": 14726 + }, + { + "epoch": 0.32416905128769535, + "grad_norm": 1.5048664223695023, + "learning_rate": 7.897274562837229e-06, + "loss": 0.3968, + "step": 14727 + }, + { + "epoch": 0.324191063174114, + "grad_norm": 1.495876578692289, + "learning_rate": 7.89698404199397e-06, + "loss": 0.3224, + "step": 14728 + }, + { + "epoch": 0.3242130750605327, + "grad_norm": 1.5278582982865403, + "learning_rate": 7.896693506426944e-06, + "loss": 0.3133, + "step": 14729 + }, + { + "epoch": 0.32423508694695136, + "grad_norm": 1.912686211026986, + "learning_rate": 7.896402956137631e-06, + "loss": 0.4225, + "step": 14730 + }, + { + "epoch": 0.32425709883337, + "grad_norm": 1.2524008637767385, + "learning_rate": 7.896112391127506e-06, + "loss": 0.3136, + "step": 14731 + }, + { + "epoch": 0.3242791107197887, + "grad_norm": 3.115805850328247, + "learning_rate": 7.895821811398047e-06, + "loss": 0.3671, + "step": 14732 + }, + { + "epoch": 0.32430112260620736, + "grad_norm": 2.481704117865715, + "learning_rate": 7.895531216950729e-06, + "loss": 0.3627, + "step": 14733 + }, + { + "epoch": 0.324323134492626, + "grad_norm": 3.0295252791852905, + "learning_rate": 7.89524060778703e-06, + "loss": 0.361, + "step": 14734 + }, + { + "epoch": 0.32434514637904466, + "grad_norm": 2.8070478619317503, + "learning_rate": 7.894949983908427e-06, + "loss": 0.3514, + "step": 14735 + }, + { + "epoch": 0.32436715826546336, + "grad_norm": 1.5970107855071098, + "learning_rate": 7.894659345316398e-06, + "loss": 0.3034, + "step": 14736 + }, + { + "epoch": 0.324389170151882, + "grad_norm": 1.4877145462407082, + "learning_rate": 7.894368692012417e-06, + "loss": 0.2865, + "step": 14737 + }, + { + "epoch": 0.32441118203830066, + "grad_norm": 1.6696781426132312, + "learning_rate": 7.894078023997966e-06, + "loss": 0.4315, + "step": 14738 + }, + { + "epoch": 0.32443319392471937, + "grad_norm": 0.6753891615320432, + "learning_rate": 7.893787341274518e-06, + "loss": 0.4966, + "step": 14739 + }, + { + "epoch": 0.324455205811138, + "grad_norm": 2.820501418094735, + "learning_rate": 7.893496643843552e-06, + "loss": 0.3578, + "step": 14740 + }, + { + "epoch": 0.32447721769755666, + "grad_norm": 2.097701142444284, + "learning_rate": 7.893205931706546e-06, + "loss": 0.4168, + "step": 14741 + }, + { + "epoch": 0.32449922958397537, + "grad_norm": 1.9591545155151928, + "learning_rate": 7.892915204864977e-06, + "loss": 0.3705, + "step": 14742 + }, + { + "epoch": 0.324521241470394, + "grad_norm": 1.748481432499714, + "learning_rate": 7.892624463320324e-06, + "loss": 0.3712, + "step": 14743 + }, + { + "epoch": 0.32454325335681267, + "grad_norm": 4.0718202845234925, + "learning_rate": 7.89233370707406e-06, + "loss": 0.2775, + "step": 14744 + }, + { + "epoch": 0.32456526524323137, + "grad_norm": 1.4481718499107388, + "learning_rate": 7.892042936127669e-06, + "loss": 0.3117, + "step": 14745 + }, + { + "epoch": 0.32458727712965, + "grad_norm": 1.9938140186531494, + "learning_rate": 7.891752150482625e-06, + "loss": 0.3877, + "step": 14746 + }, + { + "epoch": 0.32460928901606867, + "grad_norm": 2.239640656251709, + "learning_rate": 7.891461350140409e-06, + "loss": 0.2924, + "step": 14747 + }, + { + "epoch": 0.3246313009024873, + "grad_norm": 0.6739872309676318, + "learning_rate": 7.891170535102493e-06, + "loss": 0.5073, + "step": 14748 + }, + { + "epoch": 0.324653312788906, + "grad_norm": 19.49462617273366, + "learning_rate": 7.89087970537036e-06, + "loss": 0.3425, + "step": 14749 + }, + { + "epoch": 0.3246753246753247, + "grad_norm": 1.9408237893116018, + "learning_rate": 7.890588860945488e-06, + "loss": 0.2735, + "step": 14750 + }, + { + "epoch": 0.3246973365617433, + "grad_norm": 1.6969679203736474, + "learning_rate": 7.890298001829353e-06, + "loss": 0.4229, + "step": 14751 + }, + { + "epoch": 0.324719348448162, + "grad_norm": 3.333193994131129, + "learning_rate": 7.890007128023436e-06, + "loss": 0.3693, + "step": 14752 + }, + { + "epoch": 0.3247413603345807, + "grad_norm": 0.6486402588500719, + "learning_rate": 7.889716239529212e-06, + "loss": 0.5239, + "step": 14753 + }, + { + "epoch": 0.3247633722209993, + "grad_norm": 5.622645342983994, + "learning_rate": 7.889425336348162e-06, + "loss": 0.357, + "step": 14754 + }, + { + "epoch": 0.32478538410741803, + "grad_norm": 1.8388829192363818, + "learning_rate": 7.889134418481762e-06, + "loss": 0.3046, + "step": 14755 + }, + { + "epoch": 0.3248073959938367, + "grad_norm": 12.764971377123343, + "learning_rate": 7.888843485931496e-06, + "loss": 0.3536, + "step": 14756 + }, + { + "epoch": 0.3248294078802553, + "grad_norm": 3.9365689788779785, + "learning_rate": 7.888552538698837e-06, + "loss": 0.2586, + "step": 14757 + }, + { + "epoch": 0.324851419766674, + "grad_norm": 0.6376653959766085, + "learning_rate": 7.888261576785266e-06, + "loss": 0.5396, + "step": 14758 + }, + { + "epoch": 0.3248734316530927, + "grad_norm": 32.185851433167564, + "learning_rate": 7.88797060019226e-06, + "loss": 0.3348, + "step": 14759 + }, + { + "epoch": 0.32489544353951133, + "grad_norm": 1.792337649908075, + "learning_rate": 7.8876796089213e-06, + "loss": 0.3899, + "step": 14760 + }, + { + "epoch": 0.32491745542593, + "grad_norm": 3.28532901348182, + "learning_rate": 7.887388602973867e-06, + "loss": 0.3795, + "step": 14761 + }, + { + "epoch": 0.3249394673123487, + "grad_norm": 1.5196320647086479, + "learning_rate": 7.887097582351432e-06, + "loss": 0.3571, + "step": 14762 + }, + { + "epoch": 0.32496147919876733, + "grad_norm": 1.69453076496431, + "learning_rate": 7.886806547055484e-06, + "loss": 0.3363, + "step": 14763 + }, + { + "epoch": 0.324983491085186, + "grad_norm": 3.375923848159126, + "learning_rate": 7.886515497087496e-06, + "loss": 0.4028, + "step": 14764 + }, + { + "epoch": 0.3250055029716047, + "grad_norm": 0.6376483344275898, + "learning_rate": 7.886224432448951e-06, + "loss": 0.4867, + "step": 14765 + }, + { + "epoch": 0.32502751485802334, + "grad_norm": 0.6474659187005888, + "learning_rate": 7.885933353141321e-06, + "loss": 0.5049, + "step": 14766 + }, + { + "epoch": 0.325049526744442, + "grad_norm": 0.6239856887487769, + "learning_rate": 7.885642259166094e-06, + "loss": 0.4988, + "step": 14767 + }, + { + "epoch": 0.3250715386308607, + "grad_norm": 1.3739988478099305, + "learning_rate": 7.885351150524746e-06, + "loss": 0.2332, + "step": 14768 + }, + { + "epoch": 0.32509355051727934, + "grad_norm": 1.9244050216142567, + "learning_rate": 7.885060027218756e-06, + "loss": 0.3716, + "step": 14769 + }, + { + "epoch": 0.325115562403698, + "grad_norm": 1.9183707862574346, + "learning_rate": 7.884768889249605e-06, + "loss": 0.3263, + "step": 14770 + }, + { + "epoch": 0.32513757429011664, + "grad_norm": 1.8770263567660732, + "learning_rate": 7.884477736618771e-06, + "loss": 0.3313, + "step": 14771 + }, + { + "epoch": 0.32515958617653534, + "grad_norm": 3.2977336254989864, + "learning_rate": 7.884186569327734e-06, + "loss": 0.3269, + "step": 14772 + }, + { + "epoch": 0.325181598062954, + "grad_norm": 1.8726568627490268, + "learning_rate": 7.883895387377976e-06, + "loss": 0.3782, + "step": 14773 + }, + { + "epoch": 0.32520360994937264, + "grad_norm": 1.5821560072366834, + "learning_rate": 7.883604190770974e-06, + "loss": 0.281, + "step": 14774 + }, + { + "epoch": 0.32522562183579135, + "grad_norm": 1.6569451370643713, + "learning_rate": 7.883312979508211e-06, + "loss": 0.3483, + "step": 14775 + }, + { + "epoch": 0.32524763372221, + "grad_norm": 2.7732095075421226, + "learning_rate": 7.883021753591164e-06, + "loss": 0.3295, + "step": 14776 + }, + { + "epoch": 0.32526964560862864, + "grad_norm": 1.389220132123309, + "learning_rate": 7.882730513021316e-06, + "loss": 0.3332, + "step": 14777 + }, + { + "epoch": 0.32529165749504735, + "grad_norm": 0.7071902669088717, + "learning_rate": 7.882439257800143e-06, + "loss": 0.4791, + "step": 14778 + }, + { + "epoch": 0.325313669381466, + "grad_norm": 1.8651487653350578, + "learning_rate": 7.88214798792913e-06, + "loss": 0.2843, + "step": 14779 + }, + { + "epoch": 0.32533568126788465, + "grad_norm": 0.6500012603857983, + "learning_rate": 7.881856703409756e-06, + "loss": 0.5053, + "step": 14780 + }, + { + "epoch": 0.32535769315430335, + "grad_norm": 8.396175746376894, + "learning_rate": 7.8815654042435e-06, + "loss": 0.3286, + "step": 14781 + }, + { + "epoch": 0.325379705040722, + "grad_norm": 1.5252089675518035, + "learning_rate": 7.881274090431844e-06, + "loss": 0.3058, + "step": 14782 + }, + { + "epoch": 0.32540171692714065, + "grad_norm": 1.8330612197241791, + "learning_rate": 7.880982761976267e-06, + "loss": 0.3129, + "step": 14783 + }, + { + "epoch": 0.3254237288135593, + "grad_norm": 1.6086230269426132, + "learning_rate": 7.880691418878253e-06, + "loss": 0.2993, + "step": 14784 + }, + { + "epoch": 0.325445740699978, + "grad_norm": 0.6380189369019823, + "learning_rate": 7.880400061139277e-06, + "loss": 0.4899, + "step": 14785 + }, + { + "epoch": 0.32546775258639665, + "grad_norm": 1.626239185353331, + "learning_rate": 7.880108688760826e-06, + "loss": 0.3422, + "step": 14786 + }, + { + "epoch": 0.3254897644728153, + "grad_norm": 1.9324765075338644, + "learning_rate": 7.879817301744376e-06, + "loss": 0.2867, + "step": 14787 + }, + { + "epoch": 0.325511776359234, + "grad_norm": 1.8500567635177974, + "learning_rate": 7.879525900091411e-06, + "loss": 0.3505, + "step": 14788 + }, + { + "epoch": 0.32553378824565266, + "grad_norm": 1.3749939152349449, + "learning_rate": 7.87923448380341e-06, + "loss": 0.2883, + "step": 14789 + }, + { + "epoch": 0.3255558001320713, + "grad_norm": 2.1101416170803837, + "learning_rate": 7.878943052881857e-06, + "loss": 0.3156, + "step": 14790 + }, + { + "epoch": 0.32557781201849, + "grad_norm": 0.6879576818151674, + "learning_rate": 7.878651607328231e-06, + "loss": 0.5109, + "step": 14791 + }, + { + "epoch": 0.32559982390490866, + "grad_norm": 2.4113032917270654, + "learning_rate": 7.878360147144014e-06, + "loss": 0.3003, + "step": 14792 + }, + { + "epoch": 0.3256218357913273, + "grad_norm": 1.4369372089926578, + "learning_rate": 7.878068672330685e-06, + "loss": 0.334, + "step": 14793 + }, + { + "epoch": 0.32564384767774596, + "grad_norm": 1.2913774347038962, + "learning_rate": 7.877777182889729e-06, + "loss": 0.3329, + "step": 14794 + }, + { + "epoch": 0.32566585956416466, + "grad_norm": 1.4645056620514352, + "learning_rate": 7.877485678822625e-06, + "loss": 0.2604, + "step": 14795 + }, + { + "epoch": 0.3256878714505833, + "grad_norm": 1.5123598114966248, + "learning_rate": 7.877194160130855e-06, + "loss": 0.37, + "step": 14796 + }, + { + "epoch": 0.32570988333700196, + "grad_norm": 1.5762029257937629, + "learning_rate": 7.876902626815904e-06, + "loss": 0.3327, + "step": 14797 + }, + { + "epoch": 0.32573189522342066, + "grad_norm": 2.013150752871396, + "learning_rate": 7.876611078879247e-06, + "loss": 0.3, + "step": 14798 + }, + { + "epoch": 0.3257539071098393, + "grad_norm": 1.9269842416183516, + "learning_rate": 7.87631951632237e-06, + "loss": 0.4596, + "step": 14799 + }, + { + "epoch": 0.32577591899625796, + "grad_norm": 1.3002529029955012, + "learning_rate": 7.876027939146756e-06, + "loss": 0.3539, + "step": 14800 + }, + { + "epoch": 0.32579793088267667, + "grad_norm": 1.3700754684707945, + "learning_rate": 7.875736347353883e-06, + "loss": 0.305, + "step": 14801 + }, + { + "epoch": 0.3258199427690953, + "grad_norm": 1.3740074857203777, + "learning_rate": 7.87544474094524e-06, + "loss": 0.4029, + "step": 14802 + }, + { + "epoch": 0.32584195465551397, + "grad_norm": 0.6739561348788714, + "learning_rate": 7.8751531199223e-06, + "loss": 0.4938, + "step": 14803 + }, + { + "epoch": 0.32586396654193267, + "grad_norm": 1.6978184763522164, + "learning_rate": 7.87486148428655e-06, + "loss": 0.3201, + "step": 14804 + }, + { + "epoch": 0.3258859784283513, + "grad_norm": 1.8968355643365478, + "learning_rate": 7.874569834039473e-06, + "loss": 0.3347, + "step": 14805 + }, + { + "epoch": 0.32590799031476997, + "grad_norm": 1.9290962060877501, + "learning_rate": 7.87427816918255e-06, + "loss": 0.3845, + "step": 14806 + }, + { + "epoch": 0.3259300022011886, + "grad_norm": 1.3895058795420612, + "learning_rate": 7.873986489717263e-06, + "loss": 0.2751, + "step": 14807 + }, + { + "epoch": 0.3259520140876073, + "grad_norm": 0.6302553262728886, + "learning_rate": 7.873694795645094e-06, + "loss": 0.5168, + "step": 14808 + }, + { + "epoch": 0.32597402597402597, + "grad_norm": 1.2463538265628007, + "learning_rate": 7.873403086967525e-06, + "loss": 0.2715, + "step": 14809 + }, + { + "epoch": 0.3259960378604446, + "grad_norm": 2.225489710272008, + "learning_rate": 7.873111363686043e-06, + "loss": 0.3506, + "step": 14810 + }, + { + "epoch": 0.3260180497468633, + "grad_norm": 2.0111102531663407, + "learning_rate": 7.872819625802126e-06, + "loss": 0.4416, + "step": 14811 + }, + { + "epoch": 0.326040061633282, + "grad_norm": 0.6216230191608583, + "learning_rate": 7.872527873317259e-06, + "loss": 0.4799, + "step": 14812 + }, + { + "epoch": 0.3260620735197006, + "grad_norm": 1.680972869647057, + "learning_rate": 7.872236106232925e-06, + "loss": 0.3208, + "step": 14813 + }, + { + "epoch": 0.32608408540611933, + "grad_norm": 1.617771058498982, + "learning_rate": 7.871944324550605e-06, + "loss": 0.3388, + "step": 14814 + }, + { + "epoch": 0.326106097292538, + "grad_norm": 1.2888330148634874, + "learning_rate": 7.871652528271784e-06, + "loss": 0.3356, + "step": 14815 + }, + { + "epoch": 0.3261281091789566, + "grad_norm": 0.6108090069001899, + "learning_rate": 7.871360717397942e-06, + "loss": 0.5235, + "step": 14816 + }, + { + "epoch": 0.3261501210653753, + "grad_norm": 0.6474845464777641, + "learning_rate": 7.871068891930566e-06, + "loss": 0.5028, + "step": 14817 + }, + { + "epoch": 0.326172132951794, + "grad_norm": 0.649712473602141, + "learning_rate": 7.870777051871137e-06, + "loss": 0.5096, + "step": 14818 + }, + { + "epoch": 0.32619414483821263, + "grad_norm": 1.659362027576161, + "learning_rate": 7.870485197221136e-06, + "loss": 0.2735, + "step": 14819 + }, + { + "epoch": 0.3262161567246313, + "grad_norm": 2.9687255907594716, + "learning_rate": 7.870193327982052e-06, + "loss": 0.3394, + "step": 14820 + }, + { + "epoch": 0.32623816861105, + "grad_norm": 0.6364888713326352, + "learning_rate": 7.869901444155365e-06, + "loss": 0.5112, + "step": 14821 + }, + { + "epoch": 0.32626018049746863, + "grad_norm": 1.2548075499861733, + "learning_rate": 7.869609545742558e-06, + "loss": 0.3259, + "step": 14822 + }, + { + "epoch": 0.3262821923838873, + "grad_norm": 1.5167595332228958, + "learning_rate": 7.869317632745117e-06, + "loss": 0.3512, + "step": 14823 + }, + { + "epoch": 0.326304204270306, + "grad_norm": 15.54950223953236, + "learning_rate": 7.86902570516452e-06, + "loss": 0.4077, + "step": 14824 + }, + { + "epoch": 0.32632621615672464, + "grad_norm": 0.6426026188601597, + "learning_rate": 7.868733763002261e-06, + "loss": 0.5008, + "step": 14825 + }, + { + "epoch": 0.3263482280431433, + "grad_norm": 1.4599594670940226, + "learning_rate": 7.868441806259813e-06, + "loss": 0.3853, + "step": 14826 + }, + { + "epoch": 0.326370239929562, + "grad_norm": 2.1573753767001187, + "learning_rate": 7.868149834938666e-06, + "loss": 0.3968, + "step": 14827 + }, + { + "epoch": 0.32639225181598064, + "grad_norm": 1.9392336021857042, + "learning_rate": 7.8678578490403e-06, + "loss": 0.328, + "step": 14828 + }, + { + "epoch": 0.3264142637023993, + "grad_norm": 1.6183495137782604, + "learning_rate": 7.867565848566206e-06, + "loss": 0.3757, + "step": 14829 + }, + { + "epoch": 0.32643627558881794, + "grad_norm": 2.5037558691563553, + "learning_rate": 7.867273833517858e-06, + "loss": 0.4143, + "step": 14830 + }, + { + "epoch": 0.32645828747523664, + "grad_norm": 1.7667014694884238, + "learning_rate": 7.86698180389675e-06, + "loss": 0.3876, + "step": 14831 + }, + { + "epoch": 0.3264802993616553, + "grad_norm": 1.4559226579241338, + "learning_rate": 7.86668975970436e-06, + "loss": 0.3564, + "step": 14832 + }, + { + "epoch": 0.32650231124807394, + "grad_norm": 2.185529235378116, + "learning_rate": 7.866397700942173e-06, + "loss": 0.3124, + "step": 14833 + }, + { + "epoch": 0.32652432313449264, + "grad_norm": 0.6296699159236304, + "learning_rate": 7.866105627611676e-06, + "loss": 0.492, + "step": 14834 + }, + { + "epoch": 0.3265463350209113, + "grad_norm": 0.6324988009934713, + "learning_rate": 7.865813539714351e-06, + "loss": 0.5096, + "step": 14835 + }, + { + "epoch": 0.32656834690732994, + "grad_norm": 1.3259722605049544, + "learning_rate": 7.865521437251684e-06, + "loss": 0.2935, + "step": 14836 + }, + { + "epoch": 0.32659035879374865, + "grad_norm": 1.4569206992290595, + "learning_rate": 7.865229320225158e-06, + "loss": 0.3341, + "step": 14837 + }, + { + "epoch": 0.3266123706801673, + "grad_norm": 1.734767603637668, + "learning_rate": 7.86493718863626e-06, + "loss": 0.3084, + "step": 14838 + }, + { + "epoch": 0.32663438256658595, + "grad_norm": 1.7675134425016241, + "learning_rate": 7.864645042486472e-06, + "loss": 0.4129, + "step": 14839 + }, + { + "epoch": 0.3266563944530046, + "grad_norm": 2.5145871531271116, + "learning_rate": 7.864352881777282e-06, + "loss": 0.2947, + "step": 14840 + }, + { + "epoch": 0.3266784063394233, + "grad_norm": 1.5153173979237606, + "learning_rate": 7.86406070651017e-06, + "loss": 0.3269, + "step": 14841 + }, + { + "epoch": 0.32670041822584195, + "grad_norm": 0.6386919859663982, + "learning_rate": 7.863768516686627e-06, + "loss": 0.5256, + "step": 14842 + }, + { + "epoch": 0.3267224301122606, + "grad_norm": 0.6160186111934335, + "learning_rate": 7.863476312308135e-06, + "loss": 0.4925, + "step": 14843 + }, + { + "epoch": 0.3267444419986793, + "grad_norm": 2.4096303072205982, + "learning_rate": 7.863184093376178e-06, + "loss": 0.3438, + "step": 14844 + }, + { + "epoch": 0.32676645388509795, + "grad_norm": 1.5465848552160748, + "learning_rate": 7.862891859892243e-06, + "loss": 0.2821, + "step": 14845 + }, + { + "epoch": 0.3267884657715166, + "grad_norm": 1.8219070024963324, + "learning_rate": 7.862599611857816e-06, + "loss": 0.2808, + "step": 14846 + }, + { + "epoch": 0.3268104776579353, + "grad_norm": 0.6262802607646156, + "learning_rate": 7.862307349274378e-06, + "loss": 0.5192, + "step": 14847 + }, + { + "epoch": 0.32683248954435395, + "grad_norm": 1.597013472290844, + "learning_rate": 7.86201507214342e-06, + "loss": 0.2698, + "step": 14848 + }, + { + "epoch": 0.3268545014307726, + "grad_norm": 2.175246462583156, + "learning_rate": 7.861722780466423e-06, + "loss": 0.3068, + "step": 14849 + }, + { + "epoch": 0.3268765133171913, + "grad_norm": 0.6404861729703937, + "learning_rate": 7.861430474244875e-06, + "loss": 0.5132, + "step": 14850 + }, + { + "epoch": 0.32689852520360996, + "grad_norm": 1.7447941252414927, + "learning_rate": 7.861138153480262e-06, + "loss": 0.3396, + "step": 14851 + }, + { + "epoch": 0.3269205370900286, + "grad_norm": 1.645758038488975, + "learning_rate": 7.860845818174068e-06, + "loss": 0.3532, + "step": 14852 + }, + { + "epoch": 0.32694254897644726, + "grad_norm": 1.1527546718193495, + "learning_rate": 7.860553468327781e-06, + "loss": 0.339, + "step": 14853 + }, + { + "epoch": 0.32696456086286596, + "grad_norm": 0.6219322942277633, + "learning_rate": 7.860261103942884e-06, + "loss": 0.5132, + "step": 14854 + }, + { + "epoch": 0.3269865727492846, + "grad_norm": 7.121300583697464, + "learning_rate": 7.859968725020865e-06, + "loss": 0.4352, + "step": 14855 + }, + { + "epoch": 0.32700858463570326, + "grad_norm": 1.5750552187143856, + "learning_rate": 7.859676331563208e-06, + "loss": 0.2896, + "step": 14856 + }, + { + "epoch": 0.32703059652212196, + "grad_norm": 1.691377881496745, + "learning_rate": 7.859383923571402e-06, + "loss": 0.3304, + "step": 14857 + }, + { + "epoch": 0.3270526084085406, + "grad_norm": 1.620609424708215, + "learning_rate": 7.859091501046932e-06, + "loss": 0.373, + "step": 14858 + }, + { + "epoch": 0.32707462029495926, + "grad_norm": 0.6241872501582192, + "learning_rate": 7.858799063991283e-06, + "loss": 0.493, + "step": 14859 + }, + { + "epoch": 0.32709663218137797, + "grad_norm": 1.4325426286929375, + "learning_rate": 7.85850661240594e-06, + "loss": 0.3891, + "step": 14860 + }, + { + "epoch": 0.3271186440677966, + "grad_norm": 0.635719484013241, + "learning_rate": 7.858214146292394e-06, + "loss": 0.541, + "step": 14861 + }, + { + "epoch": 0.32714065595421526, + "grad_norm": 3.6622050493825573, + "learning_rate": 7.857921665652128e-06, + "loss": 0.346, + "step": 14862 + }, + { + "epoch": 0.32716266784063397, + "grad_norm": 2.1142515102384083, + "learning_rate": 7.857629170486631e-06, + "loss": 0.2881, + "step": 14863 + }, + { + "epoch": 0.3271846797270526, + "grad_norm": 1.6991136391655388, + "learning_rate": 7.857336660797386e-06, + "loss": 0.3114, + "step": 14864 + }, + { + "epoch": 0.32720669161347127, + "grad_norm": 1.7477848552368358, + "learning_rate": 7.857044136585884e-06, + "loss": 0.384, + "step": 14865 + }, + { + "epoch": 0.3272287034998899, + "grad_norm": 0.6304470363504641, + "learning_rate": 7.856751597853606e-06, + "loss": 0.4558, + "step": 14866 + }, + { + "epoch": 0.3272507153863086, + "grad_norm": 1.3989729283749184, + "learning_rate": 7.856459044602046e-06, + "loss": 0.3312, + "step": 14867 + }, + { + "epoch": 0.32727272727272727, + "grad_norm": 0.6332497542364668, + "learning_rate": 7.856166476832684e-06, + "loss": 0.4995, + "step": 14868 + }, + { + "epoch": 0.3272947391591459, + "grad_norm": 1.3966621843507485, + "learning_rate": 7.85587389454701e-06, + "loss": 0.3957, + "step": 14869 + }, + { + "epoch": 0.3273167510455646, + "grad_norm": 0.6455305719763612, + "learning_rate": 7.855581297746514e-06, + "loss": 0.5349, + "step": 14870 + }, + { + "epoch": 0.3273387629319833, + "grad_norm": 1.2712573152814506, + "learning_rate": 7.855288686432678e-06, + "loss": 0.3258, + "step": 14871 + }, + { + "epoch": 0.3273607748184019, + "grad_norm": 1.666322675781705, + "learning_rate": 7.854996060606992e-06, + "loss": 0.3647, + "step": 14872 + }, + { + "epoch": 0.3273827867048206, + "grad_norm": 1.130385586955557, + "learning_rate": 7.854703420270943e-06, + "loss": 0.3241, + "step": 14873 + }, + { + "epoch": 0.3274047985912393, + "grad_norm": 0.63236929912169, + "learning_rate": 7.854410765426017e-06, + "loss": 0.504, + "step": 14874 + }, + { + "epoch": 0.3274268104776579, + "grad_norm": 1.5394072000796888, + "learning_rate": 7.854118096073703e-06, + "loss": 0.3467, + "step": 14875 + }, + { + "epoch": 0.3274488223640766, + "grad_norm": 1.6350798957366977, + "learning_rate": 7.853825412215488e-06, + "loss": 0.3467, + "step": 14876 + }, + { + "epoch": 0.3274708342504953, + "grad_norm": 1.7610887491581917, + "learning_rate": 7.85353271385286e-06, + "loss": 0.3388, + "step": 14877 + }, + { + "epoch": 0.32749284613691393, + "grad_norm": 1.560727641787915, + "learning_rate": 7.853240000987306e-06, + "loss": 0.4304, + "step": 14878 + }, + { + "epoch": 0.3275148580233326, + "grad_norm": 1.5310008786857883, + "learning_rate": 7.852947273620313e-06, + "loss": 0.3246, + "step": 14879 + }, + { + "epoch": 0.3275368699097513, + "grad_norm": 1.2643853893594972, + "learning_rate": 7.85265453175337e-06, + "loss": 0.2508, + "step": 14880 + }, + { + "epoch": 0.32755888179616993, + "grad_norm": 1.7052212601739167, + "learning_rate": 7.852361775387963e-06, + "loss": 0.3716, + "step": 14881 + }, + { + "epoch": 0.3275808936825886, + "grad_norm": 1.3346830181847469, + "learning_rate": 7.852069004525583e-06, + "loss": 0.3127, + "step": 14882 + }, + { + "epoch": 0.3276029055690073, + "grad_norm": 1.4164286529395025, + "learning_rate": 7.851776219167714e-06, + "loss": 0.329, + "step": 14883 + }, + { + "epoch": 0.32762491745542593, + "grad_norm": 1.299152202899248, + "learning_rate": 7.851483419315849e-06, + "loss": 0.2546, + "step": 14884 + }, + { + "epoch": 0.3276469293418446, + "grad_norm": 1.198839070406427, + "learning_rate": 7.851190604971473e-06, + "loss": 0.2438, + "step": 14885 + }, + { + "epoch": 0.3276689412282633, + "grad_norm": 1.6875665499801902, + "learning_rate": 7.850897776136073e-06, + "loss": 0.4046, + "step": 14886 + }, + { + "epoch": 0.32769095311468194, + "grad_norm": 1.405672143488031, + "learning_rate": 7.850604932811139e-06, + "loss": 0.3144, + "step": 14887 + }, + { + "epoch": 0.3277129650011006, + "grad_norm": 1.722985381535546, + "learning_rate": 7.85031207499816e-06, + "loss": 0.327, + "step": 14888 + }, + { + "epoch": 0.32773497688751924, + "grad_norm": 1.4696206704536785, + "learning_rate": 7.850019202698626e-06, + "loss": 0.3274, + "step": 14889 + }, + { + "epoch": 0.32775698877393794, + "grad_norm": 1.4259582232817845, + "learning_rate": 7.84972631591402e-06, + "loss": 0.3281, + "step": 14890 + }, + { + "epoch": 0.3277790006603566, + "grad_norm": 1.8175979287287238, + "learning_rate": 7.849433414645835e-06, + "loss": 0.3832, + "step": 14891 + }, + { + "epoch": 0.32780101254677524, + "grad_norm": 1.9331975171708886, + "learning_rate": 7.849140498895559e-06, + "loss": 0.2905, + "step": 14892 + }, + { + "epoch": 0.32782302443319394, + "grad_norm": 1.5697100199671208, + "learning_rate": 7.84884756866468e-06, + "loss": 0.3878, + "step": 14893 + }, + { + "epoch": 0.3278450363196126, + "grad_norm": 1.774645762362748, + "learning_rate": 7.848554623954687e-06, + "loss": 0.2815, + "step": 14894 + }, + { + "epoch": 0.32786704820603124, + "grad_norm": 2.0199474739293413, + "learning_rate": 7.84826166476707e-06, + "loss": 0.2898, + "step": 14895 + }, + { + "epoch": 0.32788906009244995, + "grad_norm": 1.5560696486749284, + "learning_rate": 7.847968691103314e-06, + "loss": 0.3516, + "step": 14896 + }, + { + "epoch": 0.3279110719788686, + "grad_norm": 1.5177791088514736, + "learning_rate": 7.847675702964912e-06, + "loss": 0.3843, + "step": 14897 + }, + { + "epoch": 0.32793308386528724, + "grad_norm": 1.546425341617578, + "learning_rate": 7.847382700353354e-06, + "loss": 0.3539, + "step": 14898 + }, + { + "epoch": 0.3279550957517059, + "grad_norm": 1.21838075200588, + "learning_rate": 7.847089683270126e-06, + "loss": 0.2582, + "step": 14899 + }, + { + "epoch": 0.3279771076381246, + "grad_norm": 2.3820014627117847, + "learning_rate": 7.846796651716718e-06, + "loss": 0.3166, + "step": 14900 + }, + { + "epoch": 0.32799911952454325, + "grad_norm": 1.612793745069767, + "learning_rate": 7.846503605694619e-06, + "loss": 0.359, + "step": 14901 + }, + { + "epoch": 0.3280211314109619, + "grad_norm": 11.889922856841675, + "learning_rate": 7.84621054520532e-06, + "loss": 0.3954, + "step": 14902 + }, + { + "epoch": 0.3280431432973806, + "grad_norm": 2.195827388281262, + "learning_rate": 7.84591747025031e-06, + "loss": 0.3301, + "step": 14903 + }, + { + "epoch": 0.32806515518379925, + "grad_norm": 1.5806759833469584, + "learning_rate": 7.845624380831077e-06, + "loss": 0.3221, + "step": 14904 + }, + { + "epoch": 0.3280871670702179, + "grad_norm": 1.6504642494477852, + "learning_rate": 7.845331276949112e-06, + "loss": 0.2972, + "step": 14905 + }, + { + "epoch": 0.3281091789566366, + "grad_norm": 2.4085844835945416, + "learning_rate": 7.845038158605906e-06, + "loss": 0.2627, + "step": 14906 + }, + { + "epoch": 0.32813119084305525, + "grad_norm": 1.9237944728753347, + "learning_rate": 7.844745025802944e-06, + "loss": 0.3846, + "step": 14907 + }, + { + "epoch": 0.3281532027294739, + "grad_norm": 1.719299678142934, + "learning_rate": 7.844451878541721e-06, + "loss": 0.3421, + "step": 14908 + }, + { + "epoch": 0.3281752146158926, + "grad_norm": 1.3939511357970777, + "learning_rate": 7.844158716823726e-06, + "loss": 0.3805, + "step": 14909 + }, + { + "epoch": 0.32819722650231126, + "grad_norm": 0.6745226160728249, + "learning_rate": 7.843865540650446e-06, + "loss": 0.5041, + "step": 14910 + }, + { + "epoch": 0.3282192383887299, + "grad_norm": 1.4754543418464832, + "learning_rate": 7.843572350023374e-06, + "loss": 0.2861, + "step": 14911 + }, + { + "epoch": 0.32824125027514856, + "grad_norm": 1.46664977554726, + "learning_rate": 7.843279144943998e-06, + "loss": 0.346, + "step": 14912 + }, + { + "epoch": 0.32826326216156726, + "grad_norm": 1.4205015049272447, + "learning_rate": 7.84298592541381e-06, + "loss": 0.3106, + "step": 14913 + }, + { + "epoch": 0.3282852740479859, + "grad_norm": 0.644914300260618, + "learning_rate": 7.8426926914343e-06, + "loss": 0.5124, + "step": 14914 + }, + { + "epoch": 0.32830728593440456, + "grad_norm": 0.623067019942382, + "learning_rate": 7.842399443006956e-06, + "loss": 0.4891, + "step": 14915 + }, + { + "epoch": 0.32832929782082326, + "grad_norm": 2.3010268958227877, + "learning_rate": 7.842106180133273e-06, + "loss": 0.3893, + "step": 14916 + }, + { + "epoch": 0.3283513097072419, + "grad_norm": 1.5688117491769633, + "learning_rate": 7.841812902814737e-06, + "loss": 0.3087, + "step": 14917 + }, + { + "epoch": 0.32837332159366056, + "grad_norm": 0.6205211133162013, + "learning_rate": 7.841519611052839e-06, + "loss": 0.4956, + "step": 14918 + }, + { + "epoch": 0.32839533348007927, + "grad_norm": 2.1800654895894698, + "learning_rate": 7.841226304849072e-06, + "loss": 0.3359, + "step": 14919 + }, + { + "epoch": 0.3284173453664979, + "grad_norm": 1.6355533734776528, + "learning_rate": 7.840932984204928e-06, + "loss": 0.3974, + "step": 14920 + }, + { + "epoch": 0.32843935725291656, + "grad_norm": 1.2701852554151543, + "learning_rate": 7.840639649121892e-06, + "loss": 0.362, + "step": 14921 + }, + { + "epoch": 0.32846136913933527, + "grad_norm": 1.4384273980760083, + "learning_rate": 7.840346299601462e-06, + "loss": 0.3298, + "step": 14922 + }, + { + "epoch": 0.3284833810257539, + "grad_norm": 1.2805414201712986, + "learning_rate": 7.840052935645121e-06, + "loss": 0.3106, + "step": 14923 + }, + { + "epoch": 0.32850539291217257, + "grad_norm": 1.407218976128601, + "learning_rate": 7.839759557254369e-06, + "loss": 0.3729, + "step": 14924 + }, + { + "epoch": 0.3285274047985912, + "grad_norm": 1.3041214408777946, + "learning_rate": 7.83946616443069e-06, + "loss": 0.3097, + "step": 14925 + }, + { + "epoch": 0.3285494166850099, + "grad_norm": 1.842578559236877, + "learning_rate": 7.839172757175577e-06, + "loss": 0.3698, + "step": 14926 + }, + { + "epoch": 0.32857142857142857, + "grad_norm": 2.187089011132218, + "learning_rate": 7.83887933549052e-06, + "loss": 0.2575, + "step": 14927 + }, + { + "epoch": 0.3285934404578472, + "grad_norm": 1.397490168533022, + "learning_rate": 7.838585899377016e-06, + "loss": 0.291, + "step": 14928 + }, + { + "epoch": 0.3286154523442659, + "grad_norm": 1.3851223818221088, + "learning_rate": 7.83829244883655e-06, + "loss": 0.3868, + "step": 14929 + }, + { + "epoch": 0.3286374642306846, + "grad_norm": 1.4362802654519276, + "learning_rate": 7.837998983870618e-06, + "loss": 0.3126, + "step": 14930 + }, + { + "epoch": 0.3286594761171032, + "grad_norm": 1.305127171533073, + "learning_rate": 7.837705504480709e-06, + "loss": 0.2955, + "step": 14931 + }, + { + "epoch": 0.3286814880035219, + "grad_norm": 1.5228724857111655, + "learning_rate": 7.837412010668312e-06, + "loss": 0.3349, + "step": 14932 + }, + { + "epoch": 0.3287034998899406, + "grad_norm": 0.7966950188256802, + "learning_rate": 7.837118502434924e-06, + "loss": 0.5222, + "step": 14933 + }, + { + "epoch": 0.3287255117763592, + "grad_norm": 0.7337981075847942, + "learning_rate": 7.836824979782034e-06, + "loss": 0.5031, + "step": 14934 + }, + { + "epoch": 0.3287475236627779, + "grad_norm": 1.3944158551107846, + "learning_rate": 7.836531442711135e-06, + "loss": 0.2348, + "step": 14935 + }, + { + "epoch": 0.3287695355491966, + "grad_norm": 2.0238856372356375, + "learning_rate": 7.836237891223717e-06, + "loss": 0.3994, + "step": 14936 + }, + { + "epoch": 0.32879154743561523, + "grad_norm": 1.343230842359054, + "learning_rate": 7.835944325321272e-06, + "loss": 0.3444, + "step": 14937 + }, + { + "epoch": 0.3288135593220339, + "grad_norm": 1.2524796336049562, + "learning_rate": 7.835650745005296e-06, + "loss": 0.29, + "step": 14938 + }, + { + "epoch": 0.3288355712084526, + "grad_norm": 0.7788961492759576, + "learning_rate": 7.835357150277276e-06, + "loss": 0.4899, + "step": 14939 + }, + { + "epoch": 0.32885758309487123, + "grad_norm": 2.082900590386058, + "learning_rate": 7.835063541138707e-06, + "loss": 0.2573, + "step": 14940 + }, + { + "epoch": 0.3288795949812899, + "grad_norm": 1.2291594849509526, + "learning_rate": 7.83476991759108e-06, + "loss": 0.2697, + "step": 14941 + }, + { + "epoch": 0.3289016068677086, + "grad_norm": 0.7098397812773927, + "learning_rate": 7.834476279635888e-06, + "loss": 0.5293, + "step": 14942 + }, + { + "epoch": 0.32892361875412723, + "grad_norm": 2.0600681432179617, + "learning_rate": 7.834182627274624e-06, + "loss": 0.2701, + "step": 14943 + }, + { + "epoch": 0.3289456306405459, + "grad_norm": 1.2030020829364136, + "learning_rate": 7.833888960508779e-06, + "loss": 0.2766, + "step": 14944 + }, + { + "epoch": 0.3289676425269646, + "grad_norm": 1.3960627235642427, + "learning_rate": 7.833595279339847e-06, + "loss": 0.2912, + "step": 14945 + }, + { + "epoch": 0.32898965441338324, + "grad_norm": 2.1133366410941434, + "learning_rate": 7.83330158376932e-06, + "loss": 0.3036, + "step": 14946 + }, + { + "epoch": 0.3290116662998019, + "grad_norm": 0.6353459009918903, + "learning_rate": 7.833007873798692e-06, + "loss": 0.5151, + "step": 14947 + }, + { + "epoch": 0.32903367818622054, + "grad_norm": 1.4095429265003827, + "learning_rate": 7.832714149429452e-06, + "loss": 0.3368, + "step": 14948 + }, + { + "epoch": 0.32905569007263924, + "grad_norm": 1.0996947821969945, + "learning_rate": 7.832420410663097e-06, + "loss": 0.2375, + "step": 14949 + }, + { + "epoch": 0.3290777019590579, + "grad_norm": 1.6502489731219492, + "learning_rate": 7.832126657501116e-06, + "loss": 0.2817, + "step": 14950 + }, + { + "epoch": 0.32909971384547654, + "grad_norm": 0.6280668840909049, + "learning_rate": 7.831832889945006e-06, + "loss": 0.5208, + "step": 14951 + }, + { + "epoch": 0.32912172573189524, + "grad_norm": 3.0290327546399642, + "learning_rate": 7.831539107996258e-06, + "loss": 0.3602, + "step": 14952 + }, + { + "epoch": 0.3291437376183139, + "grad_norm": 0.6265119097587708, + "learning_rate": 7.831245311656366e-06, + "loss": 0.5074, + "step": 14953 + }, + { + "epoch": 0.32916574950473254, + "grad_norm": 0.6148418569349314, + "learning_rate": 7.83095150092682e-06, + "loss": 0.5118, + "step": 14954 + }, + { + "epoch": 0.32918776139115125, + "grad_norm": 0.6047934582374498, + "learning_rate": 7.83065767580912e-06, + "loss": 0.5176, + "step": 14955 + }, + { + "epoch": 0.3292097732775699, + "grad_norm": 1.5760181701195095, + "learning_rate": 7.830363836304752e-06, + "loss": 0.3536, + "step": 14956 + }, + { + "epoch": 0.32923178516398854, + "grad_norm": 1.9248661966171787, + "learning_rate": 7.830069982415214e-06, + "loss": 0.3511, + "step": 14957 + }, + { + "epoch": 0.3292537970504072, + "grad_norm": 0.6219259205830121, + "learning_rate": 7.829776114141996e-06, + "loss": 0.4908, + "step": 14958 + }, + { + "epoch": 0.3292758089368259, + "grad_norm": 1.2891759499378208, + "learning_rate": 7.829482231486596e-06, + "loss": 0.3276, + "step": 14959 + }, + { + "epoch": 0.32929782082324455, + "grad_norm": 1.6625685942436876, + "learning_rate": 7.829188334450504e-06, + "loss": 0.3628, + "step": 14960 + }, + { + "epoch": 0.3293198327096632, + "grad_norm": 1.9456144131986781, + "learning_rate": 7.828894423035216e-06, + "loss": 0.3706, + "step": 14961 + }, + { + "epoch": 0.3293418445960819, + "grad_norm": 1.584819175847438, + "learning_rate": 7.828600497242223e-06, + "loss": 0.3329, + "step": 14962 + }, + { + "epoch": 0.32936385648250055, + "grad_norm": 2.6137777012611654, + "learning_rate": 7.828306557073023e-06, + "loss": 0.3615, + "step": 14963 + }, + { + "epoch": 0.3293858683689192, + "grad_norm": 1.7675163419069817, + "learning_rate": 7.828012602529108e-06, + "loss": 0.3441, + "step": 14964 + }, + { + "epoch": 0.3294078802553379, + "grad_norm": 1.4244194247381976, + "learning_rate": 7.82771863361197e-06, + "loss": 0.3269, + "step": 14965 + }, + { + "epoch": 0.32942989214175655, + "grad_norm": 0.6459574480359016, + "learning_rate": 7.827424650323102e-06, + "loss": 0.4995, + "step": 14966 + }, + { + "epoch": 0.3294519040281752, + "grad_norm": 2.4692622825802872, + "learning_rate": 7.827130652664006e-06, + "loss": 0.3559, + "step": 14967 + }, + { + "epoch": 0.3294739159145939, + "grad_norm": 1.2070351951021265, + "learning_rate": 7.82683664063617e-06, + "loss": 0.2978, + "step": 14968 + }, + { + "epoch": 0.32949592780101256, + "grad_norm": 1.3546362743559832, + "learning_rate": 7.826542614241087e-06, + "loss": 0.3202, + "step": 14969 + }, + { + "epoch": 0.3295179396874312, + "grad_norm": 1.3915493529769607, + "learning_rate": 7.826248573480255e-06, + "loss": 0.3278, + "step": 14970 + }, + { + "epoch": 0.32953995157384985, + "grad_norm": 6.442612794493119, + "learning_rate": 7.825954518355167e-06, + "loss": 0.261, + "step": 14971 + }, + { + "epoch": 0.32956196346026856, + "grad_norm": 0.6276614109735508, + "learning_rate": 7.825660448867317e-06, + "loss": 0.4923, + "step": 14972 + }, + { + "epoch": 0.3295839753466872, + "grad_norm": 1.68217908448689, + "learning_rate": 7.825366365018201e-06, + "loss": 0.3412, + "step": 14973 + }, + { + "epoch": 0.32960598723310586, + "grad_norm": 1.3036301485709738, + "learning_rate": 7.825072266809314e-06, + "loss": 0.3212, + "step": 14974 + }, + { + "epoch": 0.32962799911952456, + "grad_norm": 2.1838058906948437, + "learning_rate": 7.82477815424215e-06, + "loss": 0.2995, + "step": 14975 + }, + { + "epoch": 0.3296500110059432, + "grad_norm": 1.527208881377167, + "learning_rate": 7.824484027318202e-06, + "loss": 0.3881, + "step": 14976 + }, + { + "epoch": 0.32967202289236186, + "grad_norm": 0.6404001138171795, + "learning_rate": 7.824189886038967e-06, + "loss": 0.4872, + "step": 14977 + }, + { + "epoch": 0.32969403477878056, + "grad_norm": 1.4319695812700415, + "learning_rate": 7.823895730405939e-06, + "loss": 0.2723, + "step": 14978 + }, + { + "epoch": 0.3297160466651992, + "grad_norm": 1.675617242988868, + "learning_rate": 7.823601560420613e-06, + "loss": 0.3419, + "step": 14979 + }, + { + "epoch": 0.32973805855161786, + "grad_norm": 2.1117622157403355, + "learning_rate": 7.823307376084486e-06, + "loss": 0.3265, + "step": 14980 + }, + { + "epoch": 0.3297600704380365, + "grad_norm": 2.3647841037899915, + "learning_rate": 7.82301317739905e-06, + "loss": 0.3577, + "step": 14981 + }, + { + "epoch": 0.3297820823244552, + "grad_norm": 2.016980397624821, + "learning_rate": 7.822718964365802e-06, + "loss": 0.376, + "step": 14982 + }, + { + "epoch": 0.32980409421087387, + "grad_norm": 0.6108698147638516, + "learning_rate": 7.82242473698624e-06, + "loss": 0.4934, + "step": 14983 + }, + { + "epoch": 0.3298261060972925, + "grad_norm": 1.1977749333087606, + "learning_rate": 7.822130495261855e-06, + "loss": 0.335, + "step": 14984 + }, + { + "epoch": 0.3298481179837112, + "grad_norm": 1.4012162770464198, + "learning_rate": 7.821836239194143e-06, + "loss": 0.2923, + "step": 14985 + }, + { + "epoch": 0.32987012987012987, + "grad_norm": 1.5067296757315924, + "learning_rate": 7.821541968784601e-06, + "loss": 0.2595, + "step": 14986 + }, + { + "epoch": 0.3298921417565485, + "grad_norm": 1.594790053836337, + "learning_rate": 7.821247684034725e-06, + "loss": 0.3616, + "step": 14987 + }, + { + "epoch": 0.3299141536429672, + "grad_norm": 0.6492964920627835, + "learning_rate": 7.82095338494601e-06, + "loss": 0.5193, + "step": 14988 + }, + { + "epoch": 0.32993616552938587, + "grad_norm": 1.6847984380333723, + "learning_rate": 7.820659071519951e-06, + "loss": 0.2821, + "step": 14989 + }, + { + "epoch": 0.3299581774158045, + "grad_norm": 1.3794514390685946, + "learning_rate": 7.820364743758046e-06, + "loss": 0.3937, + "step": 14990 + }, + { + "epoch": 0.3299801893022232, + "grad_norm": 2.2175345412565197, + "learning_rate": 7.820070401661788e-06, + "loss": 0.3337, + "step": 14991 + }, + { + "epoch": 0.3300022011886419, + "grad_norm": 0.5898967933412192, + "learning_rate": 7.819776045232674e-06, + "loss": 0.472, + "step": 14992 + }, + { + "epoch": 0.3300242130750605, + "grad_norm": 1.2925582349753533, + "learning_rate": 7.819481674472202e-06, + "loss": 0.281, + "step": 14993 + }, + { + "epoch": 0.3300462249614792, + "grad_norm": 1.646828890406282, + "learning_rate": 7.819187289381867e-06, + "loss": 0.3469, + "step": 14994 + }, + { + "epoch": 0.3300682368478979, + "grad_norm": 1.3215815528157233, + "learning_rate": 7.818892889963163e-06, + "loss": 0.3147, + "step": 14995 + }, + { + "epoch": 0.3300902487343165, + "grad_norm": 1.451896390786608, + "learning_rate": 7.81859847621759e-06, + "loss": 0.2847, + "step": 14996 + }, + { + "epoch": 0.3301122606207352, + "grad_norm": 1.794878003815415, + "learning_rate": 7.81830404814664e-06, + "loss": 0.3146, + "step": 14997 + }, + { + "epoch": 0.3301342725071539, + "grad_norm": 1.3942213048443945, + "learning_rate": 7.818009605751812e-06, + "loss": 0.3322, + "step": 14998 + }, + { + "epoch": 0.33015628439357253, + "grad_norm": 1.6988417160172098, + "learning_rate": 7.817715149034604e-06, + "loss": 0.3498, + "step": 14999 + }, + { + "epoch": 0.3301782962799912, + "grad_norm": 1.6997226623490826, + "learning_rate": 7.817420677996512e-06, + "loss": 0.4851, + "step": 15000 + }, + { + "epoch": 0.3302003081664099, + "grad_norm": 1.3265685318565972, + "learning_rate": 7.817126192639029e-06, + "loss": 0.2833, + "step": 15001 + }, + { + "epoch": 0.33022232005282853, + "grad_norm": 1.3970350949399808, + "learning_rate": 7.816831692963655e-06, + "loss": 0.3088, + "step": 15002 + }, + { + "epoch": 0.3302443319392472, + "grad_norm": 1.390512821569755, + "learning_rate": 7.816537178971885e-06, + "loss": 0.2903, + "step": 15003 + }, + { + "epoch": 0.3302663438256659, + "grad_norm": 1.410055296383257, + "learning_rate": 7.816242650665217e-06, + "loss": 0.3092, + "step": 15004 + }, + { + "epoch": 0.33028835571208454, + "grad_norm": 1.8521710327464302, + "learning_rate": 7.81594810804515e-06, + "loss": 0.3726, + "step": 15005 + }, + { + "epoch": 0.3303103675985032, + "grad_norm": 1.6553813109071238, + "learning_rate": 7.815653551113178e-06, + "loss": 0.2685, + "step": 15006 + }, + { + "epoch": 0.33033237948492183, + "grad_norm": 1.0774281740790923, + "learning_rate": 7.815358979870797e-06, + "loss": 0.3524, + "step": 15007 + }, + { + "epoch": 0.33035439137134054, + "grad_norm": 1.22980339440046, + "learning_rate": 7.815064394319508e-06, + "loss": 0.3005, + "step": 15008 + }, + { + "epoch": 0.3303764032577592, + "grad_norm": 1.7528968877919007, + "learning_rate": 7.814769794460806e-06, + "loss": 0.3262, + "step": 15009 + }, + { + "epoch": 0.33039841514417784, + "grad_norm": 1.7234525259059423, + "learning_rate": 7.814475180296187e-06, + "loss": 0.327, + "step": 15010 + }, + { + "epoch": 0.33042042703059654, + "grad_norm": 1.3120422715958728, + "learning_rate": 7.814180551827151e-06, + "loss": 0.3767, + "step": 15011 + }, + { + "epoch": 0.3304424389170152, + "grad_norm": 1.3890270822956463, + "learning_rate": 7.813885909055194e-06, + "loss": 0.3681, + "step": 15012 + }, + { + "epoch": 0.33046445080343384, + "grad_norm": 2.713745846177567, + "learning_rate": 7.813591251981813e-06, + "loss": 0.3855, + "step": 15013 + }, + { + "epoch": 0.33048646268985254, + "grad_norm": 1.7311903390271346, + "learning_rate": 7.813296580608508e-06, + "loss": 0.2553, + "step": 15014 + }, + { + "epoch": 0.3305084745762712, + "grad_norm": 1.6094617806641922, + "learning_rate": 7.813001894936774e-06, + "loss": 0.2643, + "step": 15015 + }, + { + "epoch": 0.33053048646268984, + "grad_norm": 1.4360000947615525, + "learning_rate": 7.81270719496811e-06, + "loss": 0.3397, + "step": 15016 + }, + { + "epoch": 0.3305524983491085, + "grad_norm": 1.4034731295718097, + "learning_rate": 7.812412480704014e-06, + "loss": 0.3999, + "step": 15017 + }, + { + "epoch": 0.3305745102355272, + "grad_norm": 1.775532178887958, + "learning_rate": 7.812117752145983e-06, + "loss": 0.3484, + "step": 15018 + }, + { + "epoch": 0.33059652212194585, + "grad_norm": 1.513919161099913, + "learning_rate": 7.811823009295514e-06, + "loss": 0.3977, + "step": 15019 + }, + { + "epoch": 0.3306185340083645, + "grad_norm": 1.2391785245485198, + "learning_rate": 7.811528252154108e-06, + "loss": 0.4108, + "step": 15020 + }, + { + "epoch": 0.3306405458947832, + "grad_norm": 1.318534255291572, + "learning_rate": 7.811233480723262e-06, + "loss": 0.2666, + "step": 15021 + }, + { + "epoch": 0.33066255778120185, + "grad_norm": 6.719288432482133, + "learning_rate": 7.810938695004473e-06, + "loss": 0.358, + "step": 15022 + }, + { + "epoch": 0.3306845696676205, + "grad_norm": 1.498578699345414, + "learning_rate": 7.810643894999238e-06, + "loss": 0.3056, + "step": 15023 + }, + { + "epoch": 0.3307065815540392, + "grad_norm": 1.4279494553121639, + "learning_rate": 7.81034908070906e-06, + "loss": 0.3086, + "step": 15024 + }, + { + "epoch": 0.33072859344045785, + "grad_norm": 1.3090835910970646, + "learning_rate": 7.810054252135433e-06, + "loss": 0.3163, + "step": 15025 + }, + { + "epoch": 0.3307506053268765, + "grad_norm": 1.863645158875303, + "learning_rate": 7.809759409279856e-06, + "loss": 0.3019, + "step": 15026 + }, + { + "epoch": 0.3307726172132952, + "grad_norm": 3.27229279902055, + "learning_rate": 7.80946455214383e-06, + "loss": 0.3595, + "step": 15027 + }, + { + "epoch": 0.33079462909971385, + "grad_norm": 1.3607260970890913, + "learning_rate": 7.809169680728851e-06, + "loss": 0.3026, + "step": 15028 + }, + { + "epoch": 0.3308166409861325, + "grad_norm": 1.531734398465071, + "learning_rate": 7.80887479503642e-06, + "loss": 0.3129, + "step": 15029 + }, + { + "epoch": 0.33083865287255115, + "grad_norm": 0.6605909196858395, + "learning_rate": 7.808579895068034e-06, + "loss": 0.4926, + "step": 15030 + }, + { + "epoch": 0.33086066475896986, + "grad_norm": 0.6550075864404826, + "learning_rate": 7.808284980825192e-06, + "loss": 0.5245, + "step": 15031 + }, + { + "epoch": 0.3308826766453885, + "grad_norm": 1.6747556638978511, + "learning_rate": 7.807990052309394e-06, + "loss": 0.3418, + "step": 15032 + }, + { + "epoch": 0.33090468853180716, + "grad_norm": 1.6385592954413186, + "learning_rate": 7.807695109522137e-06, + "loss": 0.2794, + "step": 15033 + }, + { + "epoch": 0.33092670041822586, + "grad_norm": 0.6096919553461303, + "learning_rate": 7.807400152464922e-06, + "loss": 0.4974, + "step": 15034 + }, + { + "epoch": 0.3309487123046445, + "grad_norm": 1.6722073187878133, + "learning_rate": 7.807105181139247e-06, + "loss": 0.338, + "step": 15035 + }, + { + "epoch": 0.33097072419106316, + "grad_norm": 1.6847042022135206, + "learning_rate": 7.806810195546611e-06, + "loss": 0.3368, + "step": 15036 + }, + { + "epoch": 0.33099273607748186, + "grad_norm": 1.203107678910685, + "learning_rate": 7.806515195688514e-06, + "loss": 0.2911, + "step": 15037 + }, + { + "epoch": 0.3310147479639005, + "grad_norm": 1.5725967323458268, + "learning_rate": 7.806220181566457e-06, + "loss": 0.2514, + "step": 15038 + }, + { + "epoch": 0.33103675985031916, + "grad_norm": 1.7763549843074817, + "learning_rate": 7.805925153181935e-06, + "loss": 0.3804, + "step": 15039 + }, + { + "epoch": 0.3310587717367378, + "grad_norm": 0.6328204265470017, + "learning_rate": 7.80563011053645e-06, + "loss": 0.5161, + "step": 15040 + }, + { + "epoch": 0.3310807836231565, + "grad_norm": 1.4160226252083665, + "learning_rate": 7.805335053631502e-06, + "loss": 0.3493, + "step": 15041 + }, + { + "epoch": 0.33110279550957517, + "grad_norm": 2.0866711706873966, + "learning_rate": 7.80503998246859e-06, + "loss": 0.3763, + "step": 15042 + }, + { + "epoch": 0.3311248073959938, + "grad_norm": 1.1784683232665778, + "learning_rate": 7.804744897049215e-06, + "loss": 0.2717, + "step": 15043 + }, + { + "epoch": 0.3311468192824125, + "grad_norm": 1.4790817286406954, + "learning_rate": 7.804449797374873e-06, + "loss": 0.3822, + "step": 15044 + }, + { + "epoch": 0.33116883116883117, + "grad_norm": 0.6576356988866933, + "learning_rate": 7.804154683447066e-06, + "loss": 0.5068, + "step": 15045 + }, + { + "epoch": 0.3311908430552498, + "grad_norm": 2.3726295432967994, + "learning_rate": 7.803859555267296e-06, + "loss": 0.3611, + "step": 15046 + }, + { + "epoch": 0.3312128549416685, + "grad_norm": 0.624380021580866, + "learning_rate": 7.803564412837063e-06, + "loss": 0.5083, + "step": 15047 + }, + { + "epoch": 0.33123486682808717, + "grad_norm": 1.6493496751088794, + "learning_rate": 7.803269256157863e-06, + "loss": 0.395, + "step": 15048 + }, + { + "epoch": 0.3312568787145058, + "grad_norm": 0.6283904270438505, + "learning_rate": 7.802974085231198e-06, + "loss": 0.5055, + "step": 15049 + }, + { + "epoch": 0.3312788906009245, + "grad_norm": 1.767332562270113, + "learning_rate": 7.802678900058569e-06, + "loss": 0.3259, + "step": 15050 + }, + { + "epoch": 0.3313009024873432, + "grad_norm": 4.0559949219179705, + "learning_rate": 7.802383700641476e-06, + "loss": 0.2751, + "step": 15051 + }, + { + "epoch": 0.3313229143737618, + "grad_norm": 1.2795600792159127, + "learning_rate": 7.80208848698142e-06, + "loss": 0.3516, + "step": 15052 + }, + { + "epoch": 0.3313449262601805, + "grad_norm": 2.2058327994690865, + "learning_rate": 7.801793259079899e-06, + "loss": 0.3452, + "step": 15053 + }, + { + "epoch": 0.3313669381465992, + "grad_norm": 1.5211439543957408, + "learning_rate": 7.801498016938415e-06, + "loss": 0.3202, + "step": 15054 + }, + { + "epoch": 0.3313889500330178, + "grad_norm": 1.4458409603158133, + "learning_rate": 7.801202760558469e-06, + "loss": 0.2704, + "step": 15055 + }, + { + "epoch": 0.3314109619194365, + "grad_norm": 1.1930958991794198, + "learning_rate": 7.800907489941562e-06, + "loss": 0.2423, + "step": 15056 + }, + { + "epoch": 0.3314329738058552, + "grad_norm": 2.0282991896720435, + "learning_rate": 7.800612205089193e-06, + "loss": 0.3382, + "step": 15057 + }, + { + "epoch": 0.33145498569227383, + "grad_norm": 1.4797489813538032, + "learning_rate": 7.800316906002864e-06, + "loss": 0.4094, + "step": 15058 + }, + { + "epoch": 0.3314769975786925, + "grad_norm": 1.3467852954692685, + "learning_rate": 7.800021592684075e-06, + "loss": 0.3745, + "step": 15059 + }, + { + "epoch": 0.3314990094651112, + "grad_norm": 3.6077719426939208, + "learning_rate": 7.799726265134328e-06, + "loss": 0.3456, + "step": 15060 + }, + { + "epoch": 0.33152102135152983, + "grad_norm": 1.4751539081181668, + "learning_rate": 7.799430923355124e-06, + "loss": 0.3783, + "step": 15061 + }, + { + "epoch": 0.3315430332379485, + "grad_norm": 1.5243381929817628, + "learning_rate": 7.799135567347962e-06, + "loss": 0.3738, + "step": 15062 + }, + { + "epoch": 0.3315650451243672, + "grad_norm": 9.306609249004564, + "learning_rate": 7.798840197114345e-06, + "loss": 0.3732, + "step": 15063 + }, + { + "epoch": 0.33158705701078584, + "grad_norm": 1.153666915206117, + "learning_rate": 7.798544812655775e-06, + "loss": 0.2882, + "step": 15064 + }, + { + "epoch": 0.3316090688972045, + "grad_norm": 1.3748313745413234, + "learning_rate": 7.798249413973751e-06, + "loss": 0.2773, + "step": 15065 + }, + { + "epoch": 0.33163108078362313, + "grad_norm": 1.3313085846934503, + "learning_rate": 7.797954001069777e-06, + "loss": 0.334, + "step": 15066 + }, + { + "epoch": 0.33165309267004184, + "grad_norm": 2.028238236105468, + "learning_rate": 7.797658573945351e-06, + "loss": 0.3077, + "step": 15067 + }, + { + "epoch": 0.3316751045564605, + "grad_norm": 1.8561214503249928, + "learning_rate": 7.797363132601976e-06, + "loss": 0.3317, + "step": 15068 + }, + { + "epoch": 0.33169711644287914, + "grad_norm": 1.6539461047445647, + "learning_rate": 7.797067677041155e-06, + "loss": 0.2361, + "step": 15069 + }, + { + "epoch": 0.33171912832929784, + "grad_norm": 1.3125285892038883, + "learning_rate": 7.796772207264388e-06, + "loss": 0.3834, + "step": 15070 + }, + { + "epoch": 0.3317411402157165, + "grad_norm": 1.9536193957352062, + "learning_rate": 7.796476723273177e-06, + "loss": 0.3675, + "step": 15071 + }, + { + "epoch": 0.33176315210213514, + "grad_norm": 1.8033700328983535, + "learning_rate": 7.796181225069026e-06, + "loss": 0.3329, + "step": 15072 + }, + { + "epoch": 0.33178516398855384, + "grad_norm": 1.4239475694081931, + "learning_rate": 7.795885712653433e-06, + "loss": 0.2595, + "step": 15073 + }, + { + "epoch": 0.3318071758749725, + "grad_norm": 1.5630814013528993, + "learning_rate": 7.795590186027901e-06, + "loss": 0.3277, + "step": 15074 + }, + { + "epoch": 0.33182918776139114, + "grad_norm": 1.324266632813537, + "learning_rate": 7.795294645193934e-06, + "loss": 0.3116, + "step": 15075 + }, + { + "epoch": 0.3318511996478098, + "grad_norm": 1.49292930416256, + "learning_rate": 7.794999090153033e-06, + "loss": 0.2637, + "step": 15076 + }, + { + "epoch": 0.3318732115342285, + "grad_norm": 1.514674293679743, + "learning_rate": 7.7947035209067e-06, + "loss": 0.3328, + "step": 15077 + }, + { + "epoch": 0.33189522342064715, + "grad_norm": 0.7510758738071065, + "learning_rate": 7.794407937456437e-06, + "loss": 0.5285, + "step": 15078 + }, + { + "epoch": 0.3319172353070658, + "grad_norm": 1.1957798027292443, + "learning_rate": 7.794112339803748e-06, + "loss": 0.4138, + "step": 15079 + }, + { + "epoch": 0.3319392471934845, + "grad_norm": 1.5614795491969058, + "learning_rate": 7.793816727950132e-06, + "loss": 0.3766, + "step": 15080 + }, + { + "epoch": 0.33196125907990315, + "grad_norm": 1.834010689359331, + "learning_rate": 7.793521101897093e-06, + "loss": 0.3735, + "step": 15081 + }, + { + "epoch": 0.3319832709663218, + "grad_norm": 1.463468083027403, + "learning_rate": 7.793225461646135e-06, + "loss": 0.4246, + "step": 15082 + }, + { + "epoch": 0.3320052828527405, + "grad_norm": 1.3192407520408145, + "learning_rate": 7.792929807198759e-06, + "loss": 0.3153, + "step": 15083 + }, + { + "epoch": 0.33202729473915915, + "grad_norm": 1.5516530510100799, + "learning_rate": 7.792634138556469e-06, + "loss": 0.3239, + "step": 15084 + }, + { + "epoch": 0.3320493066255778, + "grad_norm": 1.594924717082059, + "learning_rate": 7.792338455720765e-06, + "loss": 0.3393, + "step": 15085 + }, + { + "epoch": 0.3320713185119965, + "grad_norm": 1.4592667391054293, + "learning_rate": 7.792042758693153e-06, + "loss": 0.3415, + "step": 15086 + }, + { + "epoch": 0.33209333039841515, + "grad_norm": 0.6491624451903415, + "learning_rate": 7.791747047475133e-06, + "loss": 0.5079, + "step": 15087 + }, + { + "epoch": 0.3321153422848338, + "grad_norm": 0.6683315747646431, + "learning_rate": 7.79145132206821e-06, + "loss": 0.5097, + "step": 15088 + }, + { + "epoch": 0.33213735417125245, + "grad_norm": 1.9207616169155544, + "learning_rate": 7.791155582473888e-06, + "loss": 0.2982, + "step": 15089 + }, + { + "epoch": 0.33215936605767116, + "grad_norm": 1.2066271119710585, + "learning_rate": 7.790859828693667e-06, + "loss": 0.3219, + "step": 15090 + }, + { + "epoch": 0.3321813779440898, + "grad_norm": 1.6695375531156362, + "learning_rate": 7.790564060729051e-06, + "loss": 0.3609, + "step": 15091 + }, + { + "epoch": 0.33220338983050846, + "grad_norm": 1.7977322772544646, + "learning_rate": 7.790268278581543e-06, + "loss": 0.3567, + "step": 15092 + }, + { + "epoch": 0.33222540171692716, + "grad_norm": 1.4132304458592975, + "learning_rate": 7.78997248225265e-06, + "loss": 0.2979, + "step": 15093 + }, + { + "epoch": 0.3322474136033458, + "grad_norm": 0.6500870758072335, + "learning_rate": 7.78967667174387e-06, + "loss": 0.5067, + "step": 15094 + }, + { + "epoch": 0.33226942548976446, + "grad_norm": 1.7214162304971183, + "learning_rate": 7.78938084705671e-06, + "loss": 0.3861, + "step": 15095 + }, + { + "epoch": 0.33229143737618316, + "grad_norm": 1.367112930290802, + "learning_rate": 7.789085008192672e-06, + "loss": 0.3517, + "step": 15096 + }, + { + "epoch": 0.3323134492626018, + "grad_norm": 1.7495334982921822, + "learning_rate": 7.788789155153262e-06, + "loss": 0.3702, + "step": 15097 + }, + { + "epoch": 0.33233546114902046, + "grad_norm": 1.37017983089824, + "learning_rate": 7.78849328793998e-06, + "loss": 0.3472, + "step": 15098 + }, + { + "epoch": 0.3323574730354391, + "grad_norm": 1.3649222861392654, + "learning_rate": 7.78819740655433e-06, + "loss": 0.3657, + "step": 15099 + }, + { + "epoch": 0.3323794849218578, + "grad_norm": 1.8189274188582578, + "learning_rate": 7.78790151099782e-06, + "loss": 0.2667, + "step": 15100 + }, + { + "epoch": 0.33240149680827646, + "grad_norm": 1.093369178404827, + "learning_rate": 7.787605601271948e-06, + "loss": 0.2556, + "step": 15101 + }, + { + "epoch": 0.3324235086946951, + "grad_norm": 1.4591447524264471, + "learning_rate": 7.787309677378225e-06, + "loss": 0.3666, + "step": 15102 + }, + { + "epoch": 0.3324455205811138, + "grad_norm": 1.3316486913204875, + "learning_rate": 7.787013739318148e-06, + "loss": 0.2315, + "step": 15103 + }, + { + "epoch": 0.33246753246753247, + "grad_norm": 1.610454487604179, + "learning_rate": 7.786717787093226e-06, + "loss": 0.3582, + "step": 15104 + }, + { + "epoch": 0.3324895443539511, + "grad_norm": 1.2932448485568229, + "learning_rate": 7.786421820704959e-06, + "loss": 0.3743, + "step": 15105 + }, + { + "epoch": 0.3325115562403698, + "grad_norm": 1.4702860926621586, + "learning_rate": 7.786125840154855e-06, + "loss": 0.3864, + "step": 15106 + }, + { + "epoch": 0.33253356812678847, + "grad_norm": 1.310896875991663, + "learning_rate": 7.785829845444417e-06, + "loss": 0.2752, + "step": 15107 + }, + { + "epoch": 0.3325555800132071, + "grad_norm": 1.323179185419776, + "learning_rate": 7.785533836575147e-06, + "loss": 0.3115, + "step": 15108 + }, + { + "epoch": 0.3325775918996258, + "grad_norm": 1.5891005791667983, + "learning_rate": 7.785237813548553e-06, + "loss": 0.3223, + "step": 15109 + }, + { + "epoch": 0.3325996037860445, + "grad_norm": 1.1697108096159197, + "learning_rate": 7.78494177636614e-06, + "loss": 0.3275, + "step": 15110 + }, + { + "epoch": 0.3326216156724631, + "grad_norm": 1.2894268326418983, + "learning_rate": 7.78464572502941e-06, + "loss": 0.3116, + "step": 15111 + }, + { + "epoch": 0.33264362755888177, + "grad_norm": 0.6535059407723739, + "learning_rate": 7.784349659539867e-06, + "loss": 0.4869, + "step": 15112 + }, + { + "epoch": 0.3326656394453005, + "grad_norm": 1.54275516165646, + "learning_rate": 7.784053579899019e-06, + "loss": 0.4422, + "step": 15113 + }, + { + "epoch": 0.3326876513317191, + "grad_norm": 1.158200006190516, + "learning_rate": 7.783757486108367e-06, + "loss": 0.2958, + "step": 15114 + }, + { + "epoch": 0.3327096632181378, + "grad_norm": 1.3677382855649185, + "learning_rate": 7.783461378169419e-06, + "loss": 0.3128, + "step": 15115 + }, + { + "epoch": 0.3327316751045565, + "grad_norm": 1.4044834201553065, + "learning_rate": 7.783165256083679e-06, + "loss": 0.3069, + "step": 15116 + }, + { + "epoch": 0.33275368699097513, + "grad_norm": 2.452875319248746, + "learning_rate": 7.782869119852651e-06, + "loss": 0.3945, + "step": 15117 + }, + { + "epoch": 0.3327756988773938, + "grad_norm": 1.5103847912998165, + "learning_rate": 7.782572969477842e-06, + "loss": 0.3485, + "step": 15118 + }, + { + "epoch": 0.3327977107638125, + "grad_norm": 1.71462004013692, + "learning_rate": 7.782276804960756e-06, + "loss": 0.4592, + "step": 15119 + }, + { + "epoch": 0.33281972265023113, + "grad_norm": 1.0409152128714092, + "learning_rate": 7.781980626302898e-06, + "loss": 0.2684, + "step": 15120 + }, + { + "epoch": 0.3328417345366498, + "grad_norm": 1.413489670936227, + "learning_rate": 7.781684433505774e-06, + "loss": 0.3467, + "step": 15121 + }, + { + "epoch": 0.33286374642306843, + "grad_norm": 1.6056576483477258, + "learning_rate": 7.781388226570887e-06, + "loss": 0.3445, + "step": 15122 + }, + { + "epoch": 0.33288575830948713, + "grad_norm": 1.9967335745293153, + "learning_rate": 7.781092005499747e-06, + "loss": 0.4094, + "step": 15123 + }, + { + "epoch": 0.3329077701959058, + "grad_norm": 1.8602312621878796, + "learning_rate": 7.780795770293856e-06, + "loss": 0.403, + "step": 15124 + }, + { + "epoch": 0.33292978208232443, + "grad_norm": 0.6532681339305695, + "learning_rate": 7.78049952095472e-06, + "loss": 0.5109, + "step": 15125 + }, + { + "epoch": 0.33295179396874314, + "grad_norm": 1.381732728814774, + "learning_rate": 7.780203257483847e-06, + "loss": 0.3941, + "step": 15126 + }, + { + "epoch": 0.3329738058551618, + "grad_norm": 0.6306619740639887, + "learning_rate": 7.779906979882741e-06, + "loss": 0.4901, + "step": 15127 + }, + { + "epoch": 0.33299581774158044, + "grad_norm": 1.6841946866457458, + "learning_rate": 7.779610688152905e-06, + "loss": 0.3649, + "step": 15128 + }, + { + "epoch": 0.33301782962799914, + "grad_norm": 1.6562135475742916, + "learning_rate": 7.779314382295851e-06, + "loss": 0.3077, + "step": 15129 + }, + { + "epoch": 0.3330398415144178, + "grad_norm": 2.099741323492388, + "learning_rate": 7.77901806231308e-06, + "loss": 0.2512, + "step": 15130 + }, + { + "epoch": 0.33306185340083644, + "grad_norm": 1.232089713483117, + "learning_rate": 7.7787217282061e-06, + "loss": 0.3902, + "step": 15131 + }, + { + "epoch": 0.33308386528725514, + "grad_norm": 1.4115017159003291, + "learning_rate": 7.778425379976417e-06, + "loss": 0.3366, + "step": 15132 + }, + { + "epoch": 0.3331058771736738, + "grad_norm": 1.232218473460779, + "learning_rate": 7.778129017625538e-06, + "loss": 0.4415, + "step": 15133 + }, + { + "epoch": 0.33312788906009244, + "grad_norm": 3.51249264971443, + "learning_rate": 7.777832641154967e-06, + "loss": 0.34, + "step": 15134 + }, + { + "epoch": 0.3331499009465111, + "grad_norm": 1.4497821563710322, + "learning_rate": 7.777536250566212e-06, + "loss": 0.2918, + "step": 15135 + }, + { + "epoch": 0.3331719128329298, + "grad_norm": 0.6828613581427804, + "learning_rate": 7.777239845860778e-06, + "loss": 0.5243, + "step": 15136 + }, + { + "epoch": 0.33319392471934844, + "grad_norm": 1.2046040338023785, + "learning_rate": 7.776943427040174e-06, + "loss": 0.2562, + "step": 15137 + }, + { + "epoch": 0.3332159366057671, + "grad_norm": 1.2188998285316672, + "learning_rate": 7.776646994105904e-06, + "loss": 0.31, + "step": 15138 + }, + { + "epoch": 0.3332379484921858, + "grad_norm": 1.3306828379653188, + "learning_rate": 7.776350547059475e-06, + "loss": 0.2996, + "step": 15139 + }, + { + "epoch": 0.33325996037860445, + "grad_norm": 1.5693484276393388, + "learning_rate": 7.776054085902396e-06, + "loss": 0.4029, + "step": 15140 + }, + { + "epoch": 0.3332819722650231, + "grad_norm": 1.454237351421891, + "learning_rate": 7.77575761063617e-06, + "loss": 0.3854, + "step": 15141 + }, + { + "epoch": 0.3333039841514418, + "grad_norm": 1.3509713768174023, + "learning_rate": 7.775461121262306e-06, + "loss": 0.3816, + "step": 15142 + }, + { + "epoch": 0.33332599603786045, + "grad_norm": 1.2899234310291008, + "learning_rate": 7.775164617782312e-06, + "loss": 0.3863, + "step": 15143 + }, + { + "epoch": 0.3333480079242791, + "grad_norm": 1.1100686735494794, + "learning_rate": 7.774868100197694e-06, + "loss": 0.3599, + "step": 15144 + }, + { + "epoch": 0.3333700198106978, + "grad_norm": 1.3919938990163934, + "learning_rate": 7.774571568509958e-06, + "loss": 0.2764, + "step": 15145 + }, + { + "epoch": 0.33339203169711645, + "grad_norm": 1.9997274674040155, + "learning_rate": 7.774275022720611e-06, + "loss": 0.3488, + "step": 15146 + }, + { + "epoch": 0.3334140435835351, + "grad_norm": 1.2328362977275311, + "learning_rate": 7.773978462831163e-06, + "loss": 0.3652, + "step": 15147 + }, + { + "epoch": 0.33343605546995375, + "grad_norm": 1.4023503850404502, + "learning_rate": 7.773681888843116e-06, + "loss": 0.3126, + "step": 15148 + }, + { + "epoch": 0.33345806735637246, + "grad_norm": 0.6415032287135002, + "learning_rate": 7.773385300757982e-06, + "loss": 0.5177, + "step": 15149 + }, + { + "epoch": 0.3334800792427911, + "grad_norm": 1.0476152879345086, + "learning_rate": 7.773088698577266e-06, + "loss": 0.3349, + "step": 15150 + }, + { + "epoch": 0.33350209112920975, + "grad_norm": 1.444566662557314, + "learning_rate": 7.772792082302479e-06, + "loss": 0.3655, + "step": 15151 + }, + { + "epoch": 0.33352410301562846, + "grad_norm": 1.9794447269707363, + "learning_rate": 7.772495451935125e-06, + "loss": 0.3688, + "step": 15152 + }, + { + "epoch": 0.3335461149020471, + "grad_norm": 3.2039339003393255, + "learning_rate": 7.772198807476711e-06, + "loss": 0.3282, + "step": 15153 + }, + { + "epoch": 0.33356812678846576, + "grad_norm": 1.086741290611279, + "learning_rate": 7.771902148928748e-06, + "loss": 0.3178, + "step": 15154 + }, + { + "epoch": 0.33359013867488446, + "grad_norm": 1.6264369333064328, + "learning_rate": 7.771605476292742e-06, + "loss": 0.3838, + "step": 15155 + }, + { + "epoch": 0.3336121505613031, + "grad_norm": 1.519757426475682, + "learning_rate": 7.7713087895702e-06, + "loss": 0.3701, + "step": 15156 + }, + { + "epoch": 0.33363416244772176, + "grad_norm": 1.8046108195446575, + "learning_rate": 7.77101208876263e-06, + "loss": 0.3311, + "step": 15157 + }, + { + "epoch": 0.3336561743341404, + "grad_norm": 1.2527116294147607, + "learning_rate": 7.770715373871542e-06, + "loss": 0.2546, + "step": 15158 + }, + { + "epoch": 0.3336781862205591, + "grad_norm": 2.1815369309107617, + "learning_rate": 7.770418644898443e-06, + "loss": 0.2573, + "step": 15159 + }, + { + "epoch": 0.33370019810697776, + "grad_norm": 1.2575683694356652, + "learning_rate": 7.770121901844838e-06, + "loss": 0.4189, + "step": 15160 + }, + { + "epoch": 0.3337222099933964, + "grad_norm": 1.9850871838045707, + "learning_rate": 7.769825144712241e-06, + "loss": 0.3414, + "step": 15161 + }, + { + "epoch": 0.3337442218798151, + "grad_norm": 0.6702056983543329, + "learning_rate": 7.769528373502157e-06, + "loss": 0.4973, + "step": 15162 + }, + { + "epoch": 0.33376623376623377, + "grad_norm": 1.4230772063355583, + "learning_rate": 7.769231588216092e-06, + "loss": 0.3288, + "step": 15163 + }, + { + "epoch": 0.3337882456526524, + "grad_norm": 1.1325114356757409, + "learning_rate": 7.76893478885556e-06, + "loss": 0.2635, + "step": 15164 + }, + { + "epoch": 0.3338102575390711, + "grad_norm": 0.6396673247366153, + "learning_rate": 7.768637975422064e-06, + "loss": 0.5021, + "step": 15165 + }, + { + "epoch": 0.33383226942548977, + "grad_norm": 2.6375055310720503, + "learning_rate": 7.768341147917118e-06, + "loss": 0.3601, + "step": 15166 + }, + { + "epoch": 0.3338542813119084, + "grad_norm": 1.5640175392364328, + "learning_rate": 7.768044306342225e-06, + "loss": 0.3223, + "step": 15167 + }, + { + "epoch": 0.3338762931983271, + "grad_norm": 2.619960630405838, + "learning_rate": 7.767747450698896e-06, + "loss": 0.3654, + "step": 15168 + }, + { + "epoch": 0.3338983050847458, + "grad_norm": 1.2439995540352695, + "learning_rate": 7.767450580988641e-06, + "loss": 0.2889, + "step": 15169 + }, + { + "epoch": 0.3339203169711644, + "grad_norm": 0.6254237941229241, + "learning_rate": 7.767153697212969e-06, + "loss": 0.5186, + "step": 15170 + }, + { + "epoch": 0.33394232885758307, + "grad_norm": 1.3052711513319346, + "learning_rate": 7.766856799373386e-06, + "loss": 0.2999, + "step": 15171 + }, + { + "epoch": 0.3339643407440018, + "grad_norm": 1.267945732898802, + "learning_rate": 7.766559887471404e-06, + "loss": 0.3164, + "step": 15172 + }, + { + "epoch": 0.3339863526304204, + "grad_norm": 1.5676459990809557, + "learning_rate": 7.766262961508528e-06, + "loss": 0.3664, + "step": 15173 + }, + { + "epoch": 0.3340083645168391, + "grad_norm": 1.5007059140328742, + "learning_rate": 7.765966021486273e-06, + "loss": 0.4139, + "step": 15174 + }, + { + "epoch": 0.3340303764032578, + "grad_norm": 1.6827213738229163, + "learning_rate": 7.765669067406144e-06, + "loss": 0.2754, + "step": 15175 + }, + { + "epoch": 0.3340523882896764, + "grad_norm": 1.5493895223887941, + "learning_rate": 7.76537209926965e-06, + "loss": 0.3745, + "step": 15176 + }, + { + "epoch": 0.3340744001760951, + "grad_norm": 1.3290571404466034, + "learning_rate": 7.765075117078303e-06, + "loss": 0.3829, + "step": 15177 + }, + { + "epoch": 0.3340964120625138, + "grad_norm": 1.3617819635508115, + "learning_rate": 7.764778120833611e-06, + "loss": 0.3845, + "step": 15178 + }, + { + "epoch": 0.33411842394893243, + "grad_norm": 1.3201060730585856, + "learning_rate": 7.764481110537083e-06, + "loss": 0.2205, + "step": 15179 + }, + { + "epoch": 0.3341404358353511, + "grad_norm": 0.6568324759488436, + "learning_rate": 7.76418408619023e-06, + "loss": 0.5131, + "step": 15180 + }, + { + "epoch": 0.33416244772176973, + "grad_norm": 1.61173260400131, + "learning_rate": 7.763887047794559e-06, + "loss": 0.3036, + "step": 15181 + }, + { + "epoch": 0.33418445960818843, + "grad_norm": 1.4351404579662475, + "learning_rate": 7.763589995351582e-06, + "loss": 0.3136, + "step": 15182 + }, + { + "epoch": 0.3342064714946071, + "grad_norm": 1.9178294448971294, + "learning_rate": 7.763292928862809e-06, + "loss": 0.2887, + "step": 15183 + }, + { + "epoch": 0.33422848338102573, + "grad_norm": 1.5557841909783523, + "learning_rate": 7.76299584832975e-06, + "loss": 0.3622, + "step": 15184 + }, + { + "epoch": 0.33425049526744444, + "grad_norm": 1.6185138532621488, + "learning_rate": 7.762698753753909e-06, + "loss": 0.3226, + "step": 15185 + }, + { + "epoch": 0.3342725071538631, + "grad_norm": 1.670203518370818, + "learning_rate": 7.762401645136803e-06, + "loss": 0.2606, + "step": 15186 + }, + { + "epoch": 0.33429451904028173, + "grad_norm": 2.110773873360714, + "learning_rate": 7.762104522479941e-06, + "loss": 0.3498, + "step": 15187 + }, + { + "epoch": 0.33431653092670044, + "grad_norm": 1.2137962016238621, + "learning_rate": 7.76180738578483e-06, + "loss": 0.2806, + "step": 15188 + }, + { + "epoch": 0.3343385428131191, + "grad_norm": 1.363991472231592, + "learning_rate": 7.761510235052985e-06, + "loss": 0.3522, + "step": 15189 + }, + { + "epoch": 0.33436055469953774, + "grad_norm": 1.8326785747689986, + "learning_rate": 7.761213070285911e-06, + "loss": 0.372, + "step": 15190 + }, + { + "epoch": 0.33438256658595644, + "grad_norm": 1.406082673992616, + "learning_rate": 7.76091589148512e-06, + "loss": 0.3843, + "step": 15191 + }, + { + "epoch": 0.3344045784723751, + "grad_norm": 1.098177994694796, + "learning_rate": 7.760618698652125e-06, + "loss": 0.2574, + "step": 15192 + }, + { + "epoch": 0.33442659035879374, + "grad_norm": 1.5758362839460494, + "learning_rate": 7.760321491788434e-06, + "loss": 0.3098, + "step": 15193 + }, + { + "epoch": 0.3344486022452124, + "grad_norm": 1.4819170431261204, + "learning_rate": 7.760024270895557e-06, + "loss": 0.3091, + "step": 15194 + }, + { + "epoch": 0.3344706141316311, + "grad_norm": 2.022478282857956, + "learning_rate": 7.759727035975007e-06, + "loss": 0.4282, + "step": 15195 + }, + { + "epoch": 0.33449262601804974, + "grad_norm": 1.2623814834169524, + "learning_rate": 7.759429787028291e-06, + "loss": 0.3456, + "step": 15196 + }, + { + "epoch": 0.3345146379044684, + "grad_norm": 1.4862421620667392, + "learning_rate": 7.759132524056925e-06, + "loss": 0.3294, + "step": 15197 + }, + { + "epoch": 0.3345366497908871, + "grad_norm": 1.4005493488098264, + "learning_rate": 7.758835247062416e-06, + "loss": 0.3393, + "step": 15198 + }, + { + "epoch": 0.33455866167730575, + "grad_norm": 0.686357543761709, + "learning_rate": 7.758537956046275e-06, + "loss": 0.521, + "step": 15199 + }, + { + "epoch": 0.3345806735637244, + "grad_norm": 1.2955404375859725, + "learning_rate": 7.758240651010013e-06, + "loss": 0.3054, + "step": 15200 + }, + { + "epoch": 0.3346026854501431, + "grad_norm": 1.4232039111553851, + "learning_rate": 7.757943331955144e-06, + "loss": 0.2582, + "step": 15201 + }, + { + "epoch": 0.33462469733656175, + "grad_norm": 1.2285722668536108, + "learning_rate": 7.757645998883176e-06, + "loss": 0.3188, + "step": 15202 + }, + { + "epoch": 0.3346467092229804, + "grad_norm": 2.427141184227767, + "learning_rate": 7.757348651795621e-06, + "loss": 0.308, + "step": 15203 + }, + { + "epoch": 0.3346687211093991, + "grad_norm": 2.3018363626333884, + "learning_rate": 7.757051290693991e-06, + "loss": 0.4173, + "step": 15204 + }, + { + "epoch": 0.33469073299581775, + "grad_norm": 1.1870848298135317, + "learning_rate": 7.756753915579795e-06, + "loss": 0.3587, + "step": 15205 + }, + { + "epoch": 0.3347127448822364, + "grad_norm": 1.4645041549515196, + "learning_rate": 7.756456526454547e-06, + "loss": 0.3944, + "step": 15206 + }, + { + "epoch": 0.33473475676865505, + "grad_norm": 1.4582735938338616, + "learning_rate": 7.75615912331976e-06, + "loss": 0.2907, + "step": 15207 + }, + { + "epoch": 0.33475676865507376, + "grad_norm": 2.791912062347661, + "learning_rate": 7.75586170617694e-06, + "loss": 0.3801, + "step": 15208 + }, + { + "epoch": 0.3347787805414924, + "grad_norm": 1.7309671336822423, + "learning_rate": 7.755564275027602e-06, + "loss": 0.2949, + "step": 15209 + }, + { + "epoch": 0.33480079242791105, + "grad_norm": 1.5476997650944564, + "learning_rate": 7.755266829873258e-06, + "loss": 0.4055, + "step": 15210 + }, + { + "epoch": 0.33482280431432976, + "grad_norm": 1.4923362763626635, + "learning_rate": 7.75496937071542e-06, + "loss": 0.3671, + "step": 15211 + }, + { + "epoch": 0.3348448162007484, + "grad_norm": 2.8025489364478404, + "learning_rate": 7.754671897555597e-06, + "loss": 0.2964, + "step": 15212 + }, + { + "epoch": 0.33486682808716706, + "grad_norm": 1.8553371599833708, + "learning_rate": 7.754374410395302e-06, + "loss": 0.3998, + "step": 15213 + }, + { + "epoch": 0.33488883997358576, + "grad_norm": 2.18916586558787, + "learning_rate": 7.754076909236049e-06, + "loss": 0.3611, + "step": 15214 + }, + { + "epoch": 0.3349108518600044, + "grad_norm": 10.999800203543314, + "learning_rate": 7.75377939407935e-06, + "loss": 0.4208, + "step": 15215 + }, + { + "epoch": 0.33493286374642306, + "grad_norm": 1.4064581732734664, + "learning_rate": 7.753481864926716e-06, + "loss": 0.4123, + "step": 15216 + }, + { + "epoch": 0.3349548756328417, + "grad_norm": 1.6340630570357426, + "learning_rate": 7.753184321779656e-06, + "loss": 0.4298, + "step": 15217 + }, + { + "epoch": 0.3349768875192604, + "grad_norm": 1.529859833783657, + "learning_rate": 7.752886764639688e-06, + "loss": 0.3874, + "step": 15218 + }, + { + "epoch": 0.33499889940567906, + "grad_norm": 1.3836328989012803, + "learning_rate": 7.752589193508322e-06, + "loss": 0.402, + "step": 15219 + }, + { + "epoch": 0.3350209112920977, + "grad_norm": 1.3060381192639603, + "learning_rate": 7.752291608387068e-06, + "loss": 0.2369, + "step": 15220 + }, + { + "epoch": 0.3350429231785164, + "grad_norm": 1.8234393559981248, + "learning_rate": 7.751994009277442e-06, + "loss": 0.3306, + "step": 15221 + }, + { + "epoch": 0.33506493506493507, + "grad_norm": 1.2077414447262111, + "learning_rate": 7.751696396180955e-06, + "loss": 0.2206, + "step": 15222 + }, + { + "epoch": 0.3350869469513537, + "grad_norm": 0.6750467887866511, + "learning_rate": 7.751398769099117e-06, + "loss": 0.4903, + "step": 15223 + }, + { + "epoch": 0.3351089588377724, + "grad_norm": 1.3181847258472108, + "learning_rate": 7.751101128033447e-06, + "loss": 0.3681, + "step": 15224 + }, + { + "epoch": 0.33513097072419107, + "grad_norm": 1.6436918643428138, + "learning_rate": 7.750803472985453e-06, + "loss": 0.3115, + "step": 15225 + }, + { + "epoch": 0.3351529826106097, + "grad_norm": 1.2212932238293126, + "learning_rate": 7.750505803956646e-06, + "loss": 0.3089, + "step": 15226 + }, + { + "epoch": 0.3351749944970284, + "grad_norm": 1.6116479927431293, + "learning_rate": 7.750208120948545e-06, + "loss": 0.451, + "step": 15227 + }, + { + "epoch": 0.33519700638344707, + "grad_norm": 1.633238262077996, + "learning_rate": 7.74991042396266e-06, + "loss": 0.2821, + "step": 15228 + }, + { + "epoch": 0.3352190182698657, + "grad_norm": 1.3801193085122487, + "learning_rate": 7.749612713000502e-06, + "loss": 0.3531, + "step": 15229 + }, + { + "epoch": 0.33524103015628437, + "grad_norm": 2.0055486584466298, + "learning_rate": 7.749314988063586e-06, + "loss": 0.347, + "step": 15230 + }, + { + "epoch": 0.3352630420427031, + "grad_norm": 1.7390037202282032, + "learning_rate": 7.749017249153425e-06, + "loss": 0.3274, + "step": 15231 + }, + { + "epoch": 0.3352850539291217, + "grad_norm": 1.4618434104945417, + "learning_rate": 7.748719496271533e-06, + "loss": 0.3695, + "step": 15232 + }, + { + "epoch": 0.3353070658155404, + "grad_norm": 1.5248801588324588, + "learning_rate": 7.748421729419422e-06, + "loss": 0.4845, + "step": 15233 + }, + { + "epoch": 0.3353290777019591, + "grad_norm": 1.226736302404007, + "learning_rate": 7.748123948598607e-06, + "loss": 0.327, + "step": 15234 + }, + { + "epoch": 0.3353510895883777, + "grad_norm": 1.4219220156548307, + "learning_rate": 7.747826153810599e-06, + "loss": 0.2876, + "step": 15235 + }, + { + "epoch": 0.3353731014747964, + "grad_norm": 0.6641367414215876, + "learning_rate": 7.747528345056913e-06, + "loss": 0.5201, + "step": 15236 + }, + { + "epoch": 0.3353951133612151, + "grad_norm": 1.2968658368472825, + "learning_rate": 7.747230522339062e-06, + "loss": 0.3569, + "step": 15237 + }, + { + "epoch": 0.33541712524763373, + "grad_norm": 2.0868520268810533, + "learning_rate": 7.746932685658562e-06, + "loss": 0.3724, + "step": 15238 + }, + { + "epoch": 0.3354391371340524, + "grad_norm": 1.5305972712512725, + "learning_rate": 7.746634835016924e-06, + "loss": 0.2895, + "step": 15239 + }, + { + "epoch": 0.33546114902047103, + "grad_norm": 1.4391365223282926, + "learning_rate": 7.746336970415662e-06, + "loss": 0.2764, + "step": 15240 + }, + { + "epoch": 0.33548316090688973, + "grad_norm": 1.8119466159537383, + "learning_rate": 7.746039091856292e-06, + "loss": 0.2635, + "step": 15241 + }, + { + "epoch": 0.3355051727933084, + "grad_norm": 1.762552410221268, + "learning_rate": 7.745741199340327e-06, + "loss": 0.405, + "step": 15242 + }, + { + "epoch": 0.33552718467972703, + "grad_norm": 0.6142055641728141, + "learning_rate": 7.74544329286928e-06, + "loss": 0.4831, + "step": 15243 + }, + { + "epoch": 0.33554919656614574, + "grad_norm": 1.1863115425833872, + "learning_rate": 7.745145372444664e-06, + "loss": 0.3183, + "step": 15244 + }, + { + "epoch": 0.3355712084525644, + "grad_norm": 2.306508730681289, + "learning_rate": 7.744847438067994e-06, + "loss": 0.28, + "step": 15245 + }, + { + "epoch": 0.33559322033898303, + "grad_norm": 1.2529449490285018, + "learning_rate": 7.744549489740788e-06, + "loss": 0.297, + "step": 15246 + }, + { + "epoch": 0.33561523222540174, + "grad_norm": 1.3792405614651222, + "learning_rate": 7.744251527464557e-06, + "loss": 0.2654, + "step": 15247 + }, + { + "epoch": 0.3356372441118204, + "grad_norm": 1.4001910389323913, + "learning_rate": 7.743953551240816e-06, + "loss": 0.2888, + "step": 15248 + }, + { + "epoch": 0.33565925599823904, + "grad_norm": 0.6359778800592465, + "learning_rate": 7.743655561071079e-06, + "loss": 0.5097, + "step": 15249 + }, + { + "epoch": 0.33568126788465774, + "grad_norm": 1.5565894735408188, + "learning_rate": 7.74335755695686e-06, + "loss": 0.3629, + "step": 15250 + }, + { + "epoch": 0.3357032797710764, + "grad_norm": 0.6349094912921155, + "learning_rate": 7.743059538899673e-06, + "loss": 0.4975, + "step": 15251 + }, + { + "epoch": 0.33572529165749504, + "grad_norm": 1.6563008915778221, + "learning_rate": 7.742761506901038e-06, + "loss": 0.3378, + "step": 15252 + }, + { + "epoch": 0.3357473035439137, + "grad_norm": 1.5847297924962087, + "learning_rate": 7.742463460962462e-06, + "loss": 0.2701, + "step": 15253 + }, + { + "epoch": 0.3357693154303324, + "grad_norm": 1.0496473514496414, + "learning_rate": 7.742165401085463e-06, + "loss": 0.2827, + "step": 15254 + }, + { + "epoch": 0.33579132731675104, + "grad_norm": 2.179367205731364, + "learning_rate": 7.74186732727156e-06, + "loss": 0.4397, + "step": 15255 + }, + { + "epoch": 0.3358133392031697, + "grad_norm": 2.181454028416342, + "learning_rate": 7.741569239522263e-06, + "loss": 0.3345, + "step": 15256 + }, + { + "epoch": 0.3358353510895884, + "grad_norm": 3.572458787474224, + "learning_rate": 7.741271137839088e-06, + "loss": 0.3989, + "step": 15257 + }, + { + "epoch": 0.33585736297600705, + "grad_norm": 1.3116539322386198, + "learning_rate": 7.74097302222355e-06, + "loss": 0.3085, + "step": 15258 + }, + { + "epoch": 0.3358793748624257, + "grad_norm": 0.6670461391701679, + "learning_rate": 7.740674892677165e-06, + "loss": 0.5179, + "step": 15259 + }, + { + "epoch": 0.3359013867488444, + "grad_norm": 0.6641006687009952, + "learning_rate": 7.740376749201447e-06, + "loss": 0.4938, + "step": 15260 + }, + { + "epoch": 0.33592339863526305, + "grad_norm": 2.408103949353037, + "learning_rate": 7.740078591797914e-06, + "loss": 0.3313, + "step": 15261 + }, + { + "epoch": 0.3359454105216817, + "grad_norm": 2.0883782313919634, + "learning_rate": 7.739780420468077e-06, + "loss": 0.2929, + "step": 15262 + }, + { + "epoch": 0.33596742240810035, + "grad_norm": 1.3847383131237114, + "learning_rate": 7.739482235213455e-06, + "loss": 0.388, + "step": 15263 + }, + { + "epoch": 0.33598943429451905, + "grad_norm": 1.6711546477386463, + "learning_rate": 7.739184036035563e-06, + "loss": 0.3766, + "step": 15264 + }, + { + "epoch": 0.3360114461809377, + "grad_norm": 1.512831764912795, + "learning_rate": 7.738885822935917e-06, + "loss": 0.288, + "step": 15265 + }, + { + "epoch": 0.33603345806735635, + "grad_norm": 1.3946678718607983, + "learning_rate": 7.738587595916029e-06, + "loss": 0.3607, + "step": 15266 + }, + { + "epoch": 0.33605546995377505, + "grad_norm": 1.243467436369825, + "learning_rate": 7.73828935497742e-06, + "loss": 0.3214, + "step": 15267 + }, + { + "epoch": 0.3360774818401937, + "grad_norm": 1.467108355083028, + "learning_rate": 7.737991100121602e-06, + "loss": 0.3575, + "step": 15268 + }, + { + "epoch": 0.33609949372661235, + "grad_norm": 1.4817117074138064, + "learning_rate": 7.73769283135009e-06, + "loss": 0.4435, + "step": 15269 + }, + { + "epoch": 0.33612150561303106, + "grad_norm": 1.5062973534405846, + "learning_rate": 7.737394548664406e-06, + "loss": 0.3722, + "step": 15270 + }, + { + "epoch": 0.3361435174994497, + "grad_norm": 1.336892750627254, + "learning_rate": 7.73709625206606e-06, + "loss": 0.3246, + "step": 15271 + }, + { + "epoch": 0.33616552938586836, + "grad_norm": 1.8620370467274627, + "learning_rate": 7.736797941556568e-06, + "loss": 0.324, + "step": 15272 + }, + { + "epoch": 0.33618754127228706, + "grad_norm": 1.6276392300473768, + "learning_rate": 7.736499617137452e-06, + "loss": 0.2731, + "step": 15273 + }, + { + "epoch": 0.3362095531587057, + "grad_norm": 2.024794161839945, + "learning_rate": 7.736201278810222e-06, + "loss": 0.3527, + "step": 15274 + }, + { + "epoch": 0.33623156504512436, + "grad_norm": 1.3282545799081997, + "learning_rate": 7.735902926576397e-06, + "loss": 0.2792, + "step": 15275 + }, + { + "epoch": 0.336253576931543, + "grad_norm": 2.2530423743120482, + "learning_rate": 7.735604560437495e-06, + "loss": 0.2893, + "step": 15276 + }, + { + "epoch": 0.3362755888179617, + "grad_norm": 2.211439679739991, + "learning_rate": 7.735306180395028e-06, + "loss": 0.3195, + "step": 15277 + }, + { + "epoch": 0.33629760070438036, + "grad_norm": 1.08269823100019, + "learning_rate": 7.735007786450517e-06, + "loss": 0.3475, + "step": 15278 + }, + { + "epoch": 0.336319612590799, + "grad_norm": 1.9017431764247101, + "learning_rate": 7.734709378605476e-06, + "loss": 0.2939, + "step": 15279 + }, + { + "epoch": 0.3363416244772177, + "grad_norm": 2.7735598374471135, + "learning_rate": 7.734410956861422e-06, + "loss": 0.3, + "step": 15280 + }, + { + "epoch": 0.33636363636363636, + "grad_norm": 1.402866679022303, + "learning_rate": 7.73411252121987e-06, + "loss": 0.3634, + "step": 15281 + }, + { + "epoch": 0.336385648250055, + "grad_norm": 0.7469042473975144, + "learning_rate": 7.733814071682342e-06, + "loss": 0.5024, + "step": 15282 + }, + { + "epoch": 0.3364076601364737, + "grad_norm": 3.184794671517126, + "learning_rate": 7.733515608250349e-06, + "loss": 0.3038, + "step": 15283 + }, + { + "epoch": 0.33642967202289237, + "grad_norm": 1.3466645321305613, + "learning_rate": 7.733217130925412e-06, + "loss": 0.2745, + "step": 15284 + }, + { + "epoch": 0.336451683909311, + "grad_norm": 2.13622323894502, + "learning_rate": 7.732918639709046e-06, + "loss": 0.285, + "step": 15285 + }, + { + "epoch": 0.3364736957957297, + "grad_norm": 1.3843315219437193, + "learning_rate": 7.732620134602769e-06, + "loss": 0.3146, + "step": 15286 + }, + { + "epoch": 0.33649570768214837, + "grad_norm": 1.1618633387769512, + "learning_rate": 7.732321615608096e-06, + "loss": 0.2891, + "step": 15287 + }, + { + "epoch": 0.336517719568567, + "grad_norm": 1.3132962128100174, + "learning_rate": 7.732023082726548e-06, + "loss": 0.316, + "step": 15288 + }, + { + "epoch": 0.33653973145498567, + "grad_norm": 1.4673717861803548, + "learning_rate": 7.73172453595964e-06, + "loss": 0.3216, + "step": 15289 + }, + { + "epoch": 0.3365617433414044, + "grad_norm": 3.3582477034822644, + "learning_rate": 7.731425975308888e-06, + "loss": 0.3622, + "step": 15290 + }, + { + "epoch": 0.336583755227823, + "grad_norm": 1.3697157979262842, + "learning_rate": 7.731127400775813e-06, + "loss": 0.3159, + "step": 15291 + }, + { + "epoch": 0.33660576711424167, + "grad_norm": 1.2420629280130002, + "learning_rate": 7.730828812361928e-06, + "loss": 0.3, + "step": 15292 + }, + { + "epoch": 0.3366277790006604, + "grad_norm": 1.18707006208735, + "learning_rate": 7.730530210068755e-06, + "loss": 0.3564, + "step": 15293 + }, + { + "epoch": 0.336649790887079, + "grad_norm": 1.4776250198623424, + "learning_rate": 7.730231593897807e-06, + "loss": 0.3299, + "step": 15294 + }, + { + "epoch": 0.3366718027734977, + "grad_norm": 1.6079478881566958, + "learning_rate": 7.729932963850608e-06, + "loss": 0.2917, + "step": 15295 + }, + { + "epoch": 0.3366938146599164, + "grad_norm": 1.2556760847701227, + "learning_rate": 7.72963431992867e-06, + "loss": 0.2899, + "step": 15296 + }, + { + "epoch": 0.33671582654633503, + "grad_norm": 0.6182639175766224, + "learning_rate": 7.729335662133513e-06, + "loss": 0.4742, + "step": 15297 + }, + { + "epoch": 0.3367378384327537, + "grad_norm": 1.403141934648613, + "learning_rate": 7.729036990466656e-06, + "loss": 0.3309, + "step": 15298 + }, + { + "epoch": 0.3367598503191723, + "grad_norm": 1.080113664958713, + "learning_rate": 7.728738304929615e-06, + "loss": 0.3084, + "step": 15299 + }, + { + "epoch": 0.33678186220559103, + "grad_norm": 1.6975403438806245, + "learning_rate": 7.728439605523907e-06, + "loss": 0.3341, + "step": 15300 + }, + { + "epoch": 0.3368038740920097, + "grad_norm": 0.6205282526663431, + "learning_rate": 7.728140892251055e-06, + "loss": 0.5084, + "step": 15301 + }, + { + "epoch": 0.33682588597842833, + "grad_norm": 1.3139139221299438, + "learning_rate": 7.727842165112574e-06, + "loss": 0.3555, + "step": 15302 + }, + { + "epoch": 0.33684789786484703, + "grad_norm": 1.5983484588682164, + "learning_rate": 7.72754342410998e-06, + "loss": 0.274, + "step": 15303 + }, + { + "epoch": 0.3368699097512657, + "grad_norm": 1.5925850562837802, + "learning_rate": 7.727244669244796e-06, + "loss": 0.3653, + "step": 15304 + }, + { + "epoch": 0.33689192163768433, + "grad_norm": 0.6337314276849767, + "learning_rate": 7.726945900518536e-06, + "loss": 0.5057, + "step": 15305 + }, + { + "epoch": 0.33691393352410304, + "grad_norm": 1.6800008605653824, + "learning_rate": 7.726647117932724e-06, + "loss": 0.3797, + "step": 15306 + }, + { + "epoch": 0.3369359454105217, + "grad_norm": 1.3351054509750624, + "learning_rate": 7.726348321488872e-06, + "loss": 0.3487, + "step": 15307 + }, + { + "epoch": 0.33695795729694034, + "grad_norm": 1.0359148987674431, + "learning_rate": 7.726049511188505e-06, + "loss": 0.2256, + "step": 15308 + }, + { + "epoch": 0.33697996918335904, + "grad_norm": 1.2840767461509874, + "learning_rate": 7.725750687033136e-06, + "loss": 0.3579, + "step": 15309 + }, + { + "epoch": 0.3370019810697777, + "grad_norm": 1.873372030717721, + "learning_rate": 7.725451849024286e-06, + "loss": 0.2759, + "step": 15310 + }, + { + "epoch": 0.33702399295619634, + "grad_norm": 1.2711116161092153, + "learning_rate": 7.725152997163476e-06, + "loss": 0.3003, + "step": 15311 + }, + { + "epoch": 0.337046004842615, + "grad_norm": 1.2449468702204316, + "learning_rate": 7.724854131452223e-06, + "loss": 0.3308, + "step": 15312 + }, + { + "epoch": 0.3370680167290337, + "grad_norm": 1.8627333964541406, + "learning_rate": 7.724555251892046e-06, + "loss": 0.3266, + "step": 15313 + }, + { + "epoch": 0.33709002861545234, + "grad_norm": 1.7104826623076972, + "learning_rate": 7.724256358484462e-06, + "loss": 0.3002, + "step": 15314 + }, + { + "epoch": 0.337112040501871, + "grad_norm": 1.895347765869435, + "learning_rate": 7.723957451230994e-06, + "loss": 0.4444, + "step": 15315 + }, + { + "epoch": 0.3371340523882897, + "grad_norm": 0.6817939271634621, + "learning_rate": 7.723658530133158e-06, + "loss": 0.5198, + "step": 15316 + }, + { + "epoch": 0.33715606427470834, + "grad_norm": 0.658575618233043, + "learning_rate": 7.723359595192475e-06, + "loss": 0.4746, + "step": 15317 + }, + { + "epoch": 0.337178076161127, + "grad_norm": 1.3530163140009968, + "learning_rate": 7.723060646410464e-06, + "loss": 0.2616, + "step": 15318 + }, + { + "epoch": 0.3372000880475457, + "grad_norm": 5.07241007889245, + "learning_rate": 7.722761683788644e-06, + "loss": 0.3619, + "step": 15319 + }, + { + "epoch": 0.33722209993396435, + "grad_norm": 1.9155363629503046, + "learning_rate": 7.722462707328536e-06, + "loss": 0.3832, + "step": 15320 + }, + { + "epoch": 0.337244111820383, + "grad_norm": 1.446792576830571, + "learning_rate": 7.722163717031656e-06, + "loss": 0.3957, + "step": 15321 + }, + { + "epoch": 0.33726612370680165, + "grad_norm": 1.5291120816432864, + "learning_rate": 7.721864712899528e-06, + "loss": 0.3691, + "step": 15322 + }, + { + "epoch": 0.33728813559322035, + "grad_norm": 2.231475985135219, + "learning_rate": 7.721565694933669e-06, + "loss": 0.3469, + "step": 15323 + }, + { + "epoch": 0.337310147479639, + "grad_norm": 2.901897689597049, + "learning_rate": 7.7212666631356e-06, + "loss": 0.357, + "step": 15324 + }, + { + "epoch": 0.33733215936605765, + "grad_norm": 1.3983088419101761, + "learning_rate": 7.720967617506838e-06, + "loss": 0.3274, + "step": 15325 + }, + { + "epoch": 0.33735417125247635, + "grad_norm": 1.4384456258557772, + "learning_rate": 7.720668558048907e-06, + "loss": 0.2904, + "step": 15326 + }, + { + "epoch": 0.337376183138895, + "grad_norm": 1.5417405640462036, + "learning_rate": 7.720369484763324e-06, + "loss": 0.3192, + "step": 15327 + }, + { + "epoch": 0.33739819502531365, + "grad_norm": 1.140607091879033, + "learning_rate": 7.720070397651608e-06, + "loss": 0.2586, + "step": 15328 + }, + { + "epoch": 0.33742020691173236, + "grad_norm": 1.7087711105460726, + "learning_rate": 7.719771296715285e-06, + "loss": 0.3358, + "step": 15329 + }, + { + "epoch": 0.337442218798151, + "grad_norm": 1.429506408879782, + "learning_rate": 7.71947218195587e-06, + "loss": 0.2639, + "step": 15330 + }, + { + "epoch": 0.33746423068456965, + "grad_norm": 1.616640565218777, + "learning_rate": 7.719173053374883e-06, + "loss": 0.3401, + "step": 15331 + }, + { + "epoch": 0.33748624257098836, + "grad_norm": 0.6872629280419927, + "learning_rate": 7.718873910973846e-06, + "loss": 0.5095, + "step": 15332 + }, + { + "epoch": 0.337508254457407, + "grad_norm": 0.6884676543124555, + "learning_rate": 7.71857475475428e-06, + "loss": 0.5159, + "step": 15333 + }, + { + "epoch": 0.33753026634382566, + "grad_norm": 1.5324485605624671, + "learning_rate": 7.718275584717704e-06, + "loss": 0.3415, + "step": 15334 + }, + { + "epoch": 0.3375522782302443, + "grad_norm": 1.3273766719689497, + "learning_rate": 7.71797640086564e-06, + "loss": 0.3295, + "step": 15335 + }, + { + "epoch": 0.337574290116663, + "grad_norm": 1.4963839631549087, + "learning_rate": 7.717677203199607e-06, + "loss": 0.3745, + "step": 15336 + }, + { + "epoch": 0.33759630200308166, + "grad_norm": 1.4307048065532808, + "learning_rate": 7.717377991721127e-06, + "loss": 0.3174, + "step": 15337 + }, + { + "epoch": 0.3376183138895003, + "grad_norm": 1.496790679077399, + "learning_rate": 7.717078766431719e-06, + "loss": 0.3422, + "step": 15338 + }, + { + "epoch": 0.337640325775919, + "grad_norm": 2.410671635547054, + "learning_rate": 7.716779527332905e-06, + "loss": 0.2655, + "step": 15339 + }, + { + "epoch": 0.33766233766233766, + "grad_norm": 1.5057695618461735, + "learning_rate": 7.716480274426205e-06, + "loss": 0.384, + "step": 15340 + }, + { + "epoch": 0.3376843495487563, + "grad_norm": 1.4198227036458644, + "learning_rate": 7.716181007713143e-06, + "loss": 0.2927, + "step": 15341 + }, + { + "epoch": 0.337706361435175, + "grad_norm": 1.2531239100479936, + "learning_rate": 7.715881727195236e-06, + "loss": 0.3124, + "step": 15342 + }, + { + "epoch": 0.33772837332159367, + "grad_norm": 0.7325891891348858, + "learning_rate": 7.715582432874008e-06, + "loss": 0.5155, + "step": 15343 + }, + { + "epoch": 0.3377503852080123, + "grad_norm": 2.0652471509990074, + "learning_rate": 7.715283124750976e-06, + "loss": 0.3508, + "step": 15344 + }, + { + "epoch": 0.33777239709443097, + "grad_norm": 0.6777106567562973, + "learning_rate": 7.714983802827667e-06, + "loss": 0.526, + "step": 15345 + }, + { + "epoch": 0.33779440898084967, + "grad_norm": 1.5010882998093746, + "learning_rate": 7.714684467105598e-06, + "loss": 0.3308, + "step": 15346 + }, + { + "epoch": 0.3378164208672683, + "grad_norm": 1.4144638971206347, + "learning_rate": 7.714385117586294e-06, + "loss": 0.2975, + "step": 15347 + }, + { + "epoch": 0.33783843275368697, + "grad_norm": 2.9473842662285983, + "learning_rate": 7.714085754271271e-06, + "loss": 0.3612, + "step": 15348 + }, + { + "epoch": 0.3378604446401057, + "grad_norm": 1.322621723467248, + "learning_rate": 7.713786377162056e-06, + "loss": 0.3043, + "step": 15349 + }, + { + "epoch": 0.3378824565265243, + "grad_norm": 1.4816234919926452, + "learning_rate": 7.713486986260166e-06, + "loss": 0.3023, + "step": 15350 + }, + { + "epoch": 0.33790446841294297, + "grad_norm": 1.3018229773565975, + "learning_rate": 7.713187581567127e-06, + "loss": 0.3048, + "step": 15351 + }, + { + "epoch": 0.3379264802993617, + "grad_norm": 1.3425363253618308, + "learning_rate": 7.712888163084458e-06, + "loss": 0.3292, + "step": 15352 + }, + { + "epoch": 0.3379484921857803, + "grad_norm": 1.4129481306360137, + "learning_rate": 7.712588730813681e-06, + "loss": 0.3681, + "step": 15353 + }, + { + "epoch": 0.337970504072199, + "grad_norm": 1.1940289039762786, + "learning_rate": 7.712289284756318e-06, + "loss": 0.3031, + "step": 15354 + }, + { + "epoch": 0.3379925159586177, + "grad_norm": 2.095625786747911, + "learning_rate": 7.711989824913893e-06, + "loss": 0.39, + "step": 15355 + }, + { + "epoch": 0.33801452784503633, + "grad_norm": 1.5682363397748902, + "learning_rate": 7.711690351287923e-06, + "loss": 0.2796, + "step": 15356 + }, + { + "epoch": 0.338036539731455, + "grad_norm": 1.2270152629299098, + "learning_rate": 7.711390863879936e-06, + "loss": 0.2847, + "step": 15357 + }, + { + "epoch": 0.3380585516178736, + "grad_norm": 2.0042230216713324, + "learning_rate": 7.71109136269145e-06, + "loss": 0.2791, + "step": 15358 + }, + { + "epoch": 0.33808056350429233, + "grad_norm": 1.2844520203260092, + "learning_rate": 7.710791847723988e-06, + "loss": 0.3413, + "step": 15359 + }, + { + "epoch": 0.338102575390711, + "grad_norm": 1.3830308064597825, + "learning_rate": 7.710492318979074e-06, + "loss": 0.369, + "step": 15360 + }, + { + "epoch": 0.33812458727712963, + "grad_norm": 1.6409341410929599, + "learning_rate": 7.710192776458229e-06, + "loss": 0.2803, + "step": 15361 + }, + { + "epoch": 0.33814659916354833, + "grad_norm": 1.3363025593968014, + "learning_rate": 7.709893220162973e-06, + "loss": 0.3123, + "step": 15362 + }, + { + "epoch": 0.338168611049967, + "grad_norm": 1.5622505492201348, + "learning_rate": 7.709593650094835e-06, + "loss": 0.35, + "step": 15363 + }, + { + "epoch": 0.33819062293638563, + "grad_norm": 0.732232228008836, + "learning_rate": 7.70929406625533e-06, + "loss": 0.5167, + "step": 15364 + }, + { + "epoch": 0.33821263482280434, + "grad_norm": 1.277633239946005, + "learning_rate": 7.708994468645988e-06, + "loss": 0.2959, + "step": 15365 + }, + { + "epoch": 0.338234646709223, + "grad_norm": 1.6228229346681993, + "learning_rate": 7.708694857268325e-06, + "loss": 0.2943, + "step": 15366 + }, + { + "epoch": 0.33825665859564163, + "grad_norm": 1.4413925549255886, + "learning_rate": 7.708395232123869e-06, + "loss": 0.2627, + "step": 15367 + }, + { + "epoch": 0.33827867048206034, + "grad_norm": 1.4508031276727065, + "learning_rate": 7.708095593214137e-06, + "loss": 0.2537, + "step": 15368 + }, + { + "epoch": 0.338300682368479, + "grad_norm": 3.008628517124125, + "learning_rate": 7.707795940540657e-06, + "loss": 0.3432, + "step": 15369 + }, + { + "epoch": 0.33832269425489764, + "grad_norm": 1.6214663987003386, + "learning_rate": 7.70749627410495e-06, + "loss": 0.4101, + "step": 15370 + }, + { + "epoch": 0.3383447061413163, + "grad_norm": 1.2080880722293688, + "learning_rate": 7.70719659390854e-06, + "loss": 0.3144, + "step": 15371 + }, + { + "epoch": 0.338366718027735, + "grad_norm": 1.1859695449415548, + "learning_rate": 7.70689689995295e-06, + "loss": 0.3923, + "step": 15372 + }, + { + "epoch": 0.33838872991415364, + "grad_norm": 1.4427020078815487, + "learning_rate": 7.706597192239703e-06, + "loss": 0.264, + "step": 15373 + }, + { + "epoch": 0.3384107418005723, + "grad_norm": 1.259491526886555, + "learning_rate": 7.70629747077032e-06, + "loss": 0.3526, + "step": 15374 + }, + { + "epoch": 0.338432753686991, + "grad_norm": 0.6775026706871384, + "learning_rate": 7.705997735546327e-06, + "loss": 0.5031, + "step": 15375 + }, + { + "epoch": 0.33845476557340964, + "grad_norm": 1.4347540580142712, + "learning_rate": 7.705697986569247e-06, + "loss": 0.3955, + "step": 15376 + }, + { + "epoch": 0.3384767774598283, + "grad_norm": 1.3291421718435863, + "learning_rate": 7.705398223840603e-06, + "loss": 0.3186, + "step": 15377 + }, + { + "epoch": 0.338498789346247, + "grad_norm": 1.6119639068244513, + "learning_rate": 7.705098447361918e-06, + "loss": 0.2926, + "step": 15378 + }, + { + "epoch": 0.33852080123266565, + "grad_norm": 0.6298293564327155, + "learning_rate": 7.704798657134717e-06, + "loss": 0.5215, + "step": 15379 + }, + { + "epoch": 0.3385428131190843, + "grad_norm": 1.31607703946078, + "learning_rate": 7.704498853160523e-06, + "loss": 0.2895, + "step": 15380 + }, + { + "epoch": 0.33856482500550295, + "grad_norm": 1.5035654679494743, + "learning_rate": 7.704199035440858e-06, + "loss": 0.4293, + "step": 15381 + }, + { + "epoch": 0.33858683689192165, + "grad_norm": 1.5305827466756652, + "learning_rate": 7.703899203977248e-06, + "loss": 0.3553, + "step": 15382 + }, + { + "epoch": 0.3386088487783403, + "grad_norm": 1.6072852587160955, + "learning_rate": 7.703599358771217e-06, + "loss": 0.2758, + "step": 15383 + }, + { + "epoch": 0.33863086066475895, + "grad_norm": 2.3798357738726197, + "learning_rate": 7.703299499824288e-06, + "loss": 0.292, + "step": 15384 + }, + { + "epoch": 0.33865287255117765, + "grad_norm": 1.3205746711750705, + "learning_rate": 7.702999627137984e-06, + "loss": 0.343, + "step": 15385 + }, + { + "epoch": 0.3386748844375963, + "grad_norm": 0.6669895667827194, + "learning_rate": 7.702699740713833e-06, + "loss": 0.5178, + "step": 15386 + }, + { + "epoch": 0.33869689632401495, + "grad_norm": 1.502399010027457, + "learning_rate": 7.702399840553353e-06, + "loss": 0.3251, + "step": 15387 + }, + { + "epoch": 0.33871890821043366, + "grad_norm": 0.6222305928887605, + "learning_rate": 7.702099926658075e-06, + "loss": 0.5003, + "step": 15388 + }, + { + "epoch": 0.3387409200968523, + "grad_norm": 0.6362711228204782, + "learning_rate": 7.701799999029518e-06, + "loss": 0.4941, + "step": 15389 + }, + { + "epoch": 0.33876293198327095, + "grad_norm": 0.6317324170940429, + "learning_rate": 7.701500057669209e-06, + "loss": 0.505, + "step": 15390 + }, + { + "epoch": 0.33878494386968966, + "grad_norm": 1.4499343759720904, + "learning_rate": 7.701200102578672e-06, + "loss": 0.3842, + "step": 15391 + }, + { + "epoch": 0.3388069557561083, + "grad_norm": 1.2467561530789841, + "learning_rate": 7.700900133759431e-06, + "loss": 0.3249, + "step": 15392 + }, + { + "epoch": 0.33882896764252696, + "grad_norm": 1.4030364816475975, + "learning_rate": 7.700600151213012e-06, + "loss": 0.3051, + "step": 15393 + }, + { + "epoch": 0.3388509795289456, + "grad_norm": 1.43671040975189, + "learning_rate": 7.700300154940937e-06, + "loss": 0.3112, + "step": 15394 + }, + { + "epoch": 0.3388729914153643, + "grad_norm": 1.3471042867936498, + "learning_rate": 7.700000144944733e-06, + "loss": 0.2677, + "step": 15395 + }, + { + "epoch": 0.33889500330178296, + "grad_norm": 1.3114338559315917, + "learning_rate": 7.699700121225924e-06, + "loss": 0.3342, + "step": 15396 + }, + { + "epoch": 0.3389170151882016, + "grad_norm": 1.4004003161453928, + "learning_rate": 7.699400083786034e-06, + "loss": 0.2754, + "step": 15397 + }, + { + "epoch": 0.3389390270746203, + "grad_norm": 1.4428347285737826, + "learning_rate": 7.69910003262659e-06, + "loss": 0.3466, + "step": 15398 + }, + { + "epoch": 0.33896103896103896, + "grad_norm": 1.4207278746409489, + "learning_rate": 7.698799967749114e-06, + "loss": 0.2943, + "step": 15399 + }, + { + "epoch": 0.3389830508474576, + "grad_norm": 0.7155856427726006, + "learning_rate": 7.698499889155133e-06, + "loss": 0.5338, + "step": 15400 + }, + { + "epoch": 0.3390050627338763, + "grad_norm": 0.6786382501537246, + "learning_rate": 7.698199796846175e-06, + "loss": 0.5103, + "step": 15401 + }, + { + "epoch": 0.33902707462029497, + "grad_norm": 1.4393293855144647, + "learning_rate": 7.69789969082376e-06, + "loss": 0.2924, + "step": 15402 + }, + { + "epoch": 0.3390490865067136, + "grad_norm": 0.6140913800794789, + "learning_rate": 7.697599571089416e-06, + "loss": 0.5257, + "step": 15403 + }, + { + "epoch": 0.33907109839313226, + "grad_norm": 1.6651759988479404, + "learning_rate": 7.697299437644667e-06, + "loss": 0.3352, + "step": 15404 + }, + { + "epoch": 0.33909311027955097, + "grad_norm": 1.2656440312730945, + "learning_rate": 7.69699929049104e-06, + "loss": 0.3068, + "step": 15405 + }, + { + "epoch": 0.3391151221659696, + "grad_norm": 0.6405454811010738, + "learning_rate": 7.69669912963006e-06, + "loss": 0.4819, + "step": 15406 + }, + { + "epoch": 0.33913713405238827, + "grad_norm": 1.4483793216312846, + "learning_rate": 7.69639895506325e-06, + "loss": 0.2999, + "step": 15407 + }, + { + "epoch": 0.33915914593880697, + "grad_norm": 1.8910466115872038, + "learning_rate": 7.69609876679214e-06, + "loss": 0.3524, + "step": 15408 + }, + { + "epoch": 0.3391811578252256, + "grad_norm": 1.3631928830583502, + "learning_rate": 7.695798564818254e-06, + "loss": 0.3238, + "step": 15409 + }, + { + "epoch": 0.33920316971164427, + "grad_norm": 0.6274111659339897, + "learning_rate": 7.695498349143116e-06, + "loss": 0.4977, + "step": 15410 + }, + { + "epoch": 0.339225181598063, + "grad_norm": 1.4919886158472073, + "learning_rate": 7.695198119768255e-06, + "loss": 0.3167, + "step": 15411 + }, + { + "epoch": 0.3392471934844816, + "grad_norm": 1.1427594040508626, + "learning_rate": 7.694897876695194e-06, + "loss": 0.3328, + "step": 15412 + }, + { + "epoch": 0.3392692053709003, + "grad_norm": 0.6444676132677375, + "learning_rate": 7.69459761992546e-06, + "loss": 0.5154, + "step": 15413 + }, + { + "epoch": 0.339291217257319, + "grad_norm": 1.5918432965602882, + "learning_rate": 7.694297349460578e-06, + "loss": 0.2974, + "step": 15414 + }, + { + "epoch": 0.3393132291437376, + "grad_norm": 2.335228484536635, + "learning_rate": 7.693997065302078e-06, + "loss": 0.348, + "step": 15415 + }, + { + "epoch": 0.3393352410301563, + "grad_norm": 0.6213767608083058, + "learning_rate": 7.693696767451479e-06, + "loss": 0.5019, + "step": 15416 + }, + { + "epoch": 0.3393572529165749, + "grad_norm": 1.522931306773969, + "learning_rate": 7.693396455910314e-06, + "loss": 0.3492, + "step": 15417 + }, + { + "epoch": 0.33937926480299363, + "grad_norm": 1.7968352385746615, + "learning_rate": 7.693096130680108e-06, + "loss": 0.3481, + "step": 15418 + }, + { + "epoch": 0.3394012766894123, + "grad_norm": 1.778450092124388, + "learning_rate": 7.692795791762385e-06, + "loss": 0.3372, + "step": 15419 + }, + { + "epoch": 0.33942328857583093, + "grad_norm": 1.2533430113279984, + "learning_rate": 7.692495439158674e-06, + "loss": 0.3608, + "step": 15420 + }, + { + "epoch": 0.33944530046224963, + "grad_norm": 0.5984439037953032, + "learning_rate": 7.692195072870499e-06, + "loss": 0.4949, + "step": 15421 + }, + { + "epoch": 0.3394673123486683, + "grad_norm": 1.4297990629899553, + "learning_rate": 7.691894692899387e-06, + "loss": 0.3169, + "step": 15422 + }, + { + "epoch": 0.33948932423508693, + "grad_norm": 1.874183024262382, + "learning_rate": 7.691594299246867e-06, + "loss": 0.3143, + "step": 15423 + }, + { + "epoch": 0.33951133612150564, + "grad_norm": 2.1714903421426697, + "learning_rate": 7.691293891914465e-06, + "loss": 0.3317, + "step": 15424 + }, + { + "epoch": 0.3395333480079243, + "grad_norm": 2.524999233728957, + "learning_rate": 7.690993470903705e-06, + "loss": 0.3539, + "step": 15425 + }, + { + "epoch": 0.33955535989434293, + "grad_norm": 1.5164267517476944, + "learning_rate": 7.690693036216118e-06, + "loss": 0.3624, + "step": 15426 + }, + { + "epoch": 0.33957737178076164, + "grad_norm": 1.500107434249535, + "learning_rate": 7.690392587853229e-06, + "loss": 0.3445, + "step": 15427 + }, + { + "epoch": 0.3395993836671803, + "grad_norm": 1.447441198590965, + "learning_rate": 7.690092125816562e-06, + "loss": 0.447, + "step": 15428 + }, + { + "epoch": 0.33962139555359894, + "grad_norm": 1.477380412839348, + "learning_rate": 7.689791650107649e-06, + "loss": 0.3153, + "step": 15429 + }, + { + "epoch": 0.3396434074400176, + "grad_norm": 1.712440752190237, + "learning_rate": 7.689491160728015e-06, + "loss": 0.3984, + "step": 15430 + }, + { + "epoch": 0.3396654193264363, + "grad_norm": 1.7503628514184502, + "learning_rate": 7.689190657679188e-06, + "loss": 0.2613, + "step": 15431 + }, + { + "epoch": 0.33968743121285494, + "grad_norm": 1.3074188099766788, + "learning_rate": 7.688890140962694e-06, + "loss": 0.3441, + "step": 15432 + }, + { + "epoch": 0.3397094430992736, + "grad_norm": 0.7116211571112071, + "learning_rate": 7.688589610580062e-06, + "loss": 0.5018, + "step": 15433 + }, + { + "epoch": 0.3397314549856923, + "grad_norm": 1.465072977655341, + "learning_rate": 7.688289066532817e-06, + "loss": 0.3296, + "step": 15434 + }, + { + "epoch": 0.33975346687211094, + "grad_norm": 1.3496633534923417, + "learning_rate": 7.687988508822489e-06, + "loss": 0.3331, + "step": 15435 + }, + { + "epoch": 0.3397754787585296, + "grad_norm": 1.539710717143703, + "learning_rate": 7.687687937450604e-06, + "loss": 0.4371, + "step": 15436 + }, + { + "epoch": 0.3397974906449483, + "grad_norm": 1.2267734213591772, + "learning_rate": 7.687387352418691e-06, + "loss": 0.3145, + "step": 15437 + }, + { + "epoch": 0.33981950253136695, + "grad_norm": 1.6509990160552073, + "learning_rate": 7.687086753728275e-06, + "loss": 0.266, + "step": 15438 + }, + { + "epoch": 0.3398415144177856, + "grad_norm": 1.3353115609177901, + "learning_rate": 7.686786141380887e-06, + "loss": 0.3624, + "step": 15439 + }, + { + "epoch": 0.33986352630420424, + "grad_norm": 1.3861301089934683, + "learning_rate": 7.686485515378053e-06, + "loss": 0.3794, + "step": 15440 + }, + { + "epoch": 0.33988553819062295, + "grad_norm": 1.238508575370066, + "learning_rate": 7.686184875721302e-06, + "loss": 0.2765, + "step": 15441 + }, + { + "epoch": 0.3399075500770416, + "grad_norm": 1.6046225478802456, + "learning_rate": 7.685884222412161e-06, + "loss": 0.4473, + "step": 15442 + }, + { + "epoch": 0.33992956196346025, + "grad_norm": 1.3934383652945808, + "learning_rate": 7.685583555452158e-06, + "loss": 0.2985, + "step": 15443 + }, + { + "epoch": 0.33995157384987895, + "grad_norm": 1.2108241762157725, + "learning_rate": 7.685282874842823e-06, + "loss": 0.2437, + "step": 15444 + }, + { + "epoch": 0.3399735857362976, + "grad_norm": 1.9230299079596436, + "learning_rate": 7.684982180585682e-06, + "loss": 0.3942, + "step": 15445 + }, + { + "epoch": 0.33999559762271625, + "grad_norm": 0.7127174009830285, + "learning_rate": 7.684681472682263e-06, + "loss": 0.5139, + "step": 15446 + }, + { + "epoch": 0.34001760950913495, + "grad_norm": 1.7584577704505508, + "learning_rate": 7.684380751134097e-06, + "loss": 0.376, + "step": 15447 + }, + { + "epoch": 0.3400396213955536, + "grad_norm": 1.3048687083847696, + "learning_rate": 7.68408001594271e-06, + "loss": 0.3003, + "step": 15448 + }, + { + "epoch": 0.34006163328197225, + "grad_norm": 1.3633715620793054, + "learning_rate": 7.683779267109631e-06, + "loss": 0.3692, + "step": 15449 + }, + { + "epoch": 0.34008364516839096, + "grad_norm": 1.4763913959416877, + "learning_rate": 7.683478504636389e-06, + "loss": 0.2679, + "step": 15450 + }, + { + "epoch": 0.3401056570548096, + "grad_norm": 1.3195857563790334, + "learning_rate": 7.683177728524512e-06, + "loss": 0.3697, + "step": 15451 + }, + { + "epoch": 0.34012766894122826, + "grad_norm": 1.3441582067555806, + "learning_rate": 7.68287693877553e-06, + "loss": 0.3954, + "step": 15452 + }, + { + "epoch": 0.3401496808276469, + "grad_norm": 0.6370041785170836, + "learning_rate": 7.682576135390969e-06, + "loss": 0.5188, + "step": 15453 + }, + { + "epoch": 0.3401716927140656, + "grad_norm": 0.614809294571527, + "learning_rate": 7.682275318372362e-06, + "loss": 0.51, + "step": 15454 + }, + { + "epoch": 0.34019370460048426, + "grad_norm": 1.2912753999352418, + "learning_rate": 7.681974487721235e-06, + "loss": 0.3357, + "step": 15455 + }, + { + "epoch": 0.3402157164869029, + "grad_norm": 1.3380769015974656, + "learning_rate": 7.681673643439116e-06, + "loss": 0.339, + "step": 15456 + }, + { + "epoch": 0.3402377283733216, + "grad_norm": 1.1206750090143693, + "learning_rate": 7.681372785527536e-06, + "loss": 0.2668, + "step": 15457 + }, + { + "epoch": 0.34025974025974026, + "grad_norm": 1.248572399704308, + "learning_rate": 7.681071913988023e-06, + "loss": 0.3098, + "step": 15458 + }, + { + "epoch": 0.3402817521461589, + "grad_norm": 1.391740874430646, + "learning_rate": 7.680771028822109e-06, + "loss": 0.3696, + "step": 15459 + }, + { + "epoch": 0.3403037640325776, + "grad_norm": 0.6300472485532904, + "learning_rate": 7.68047013003132e-06, + "loss": 0.4959, + "step": 15460 + }, + { + "epoch": 0.34032577591899627, + "grad_norm": 1.3806692858700322, + "learning_rate": 7.680169217617186e-06, + "loss": 0.3723, + "step": 15461 + }, + { + "epoch": 0.3403477878054149, + "grad_norm": 1.3028995353260118, + "learning_rate": 7.679868291581236e-06, + "loss": 0.3705, + "step": 15462 + }, + { + "epoch": 0.34036979969183356, + "grad_norm": 1.6314747168532968, + "learning_rate": 7.679567351925e-06, + "loss": 0.3368, + "step": 15463 + }, + { + "epoch": 0.34039181157825227, + "grad_norm": 1.814222347262524, + "learning_rate": 7.679266398650007e-06, + "loss": 0.3884, + "step": 15464 + }, + { + "epoch": 0.3404138234646709, + "grad_norm": 2.3072485522919264, + "learning_rate": 7.67896543175779e-06, + "loss": 0.365, + "step": 15465 + }, + { + "epoch": 0.34043583535108957, + "grad_norm": 5.1087440557905195, + "learning_rate": 7.678664451249874e-06, + "loss": 0.2907, + "step": 15466 + }, + { + "epoch": 0.34045784723750827, + "grad_norm": 1.489907222534136, + "learning_rate": 7.678363457127791e-06, + "loss": 0.3446, + "step": 15467 + }, + { + "epoch": 0.3404798591239269, + "grad_norm": 1.4006718958432132, + "learning_rate": 7.678062449393069e-06, + "loss": 0.384, + "step": 15468 + }, + { + "epoch": 0.34050187101034557, + "grad_norm": 1.408144340214556, + "learning_rate": 7.67776142804724e-06, + "loss": 0.3673, + "step": 15469 + }, + { + "epoch": 0.3405238828967643, + "grad_norm": 1.5490238861967542, + "learning_rate": 7.677460393091834e-06, + "loss": 0.32, + "step": 15470 + }, + { + "epoch": 0.3405458947831829, + "grad_norm": 1.8710818923333956, + "learning_rate": 7.677159344528379e-06, + "loss": 0.3395, + "step": 15471 + }, + { + "epoch": 0.34056790666960157, + "grad_norm": 1.2300426550708312, + "learning_rate": 7.676858282358407e-06, + "loss": 0.267, + "step": 15472 + }, + { + "epoch": 0.3405899185560203, + "grad_norm": 1.2654024452994186, + "learning_rate": 7.676557206583448e-06, + "loss": 0.33, + "step": 15473 + }, + { + "epoch": 0.3406119304424389, + "grad_norm": 1.3780985336401395, + "learning_rate": 7.676256117205029e-06, + "loss": 0.3677, + "step": 15474 + }, + { + "epoch": 0.3406339423288576, + "grad_norm": 0.6512324758544259, + "learning_rate": 7.675955014224686e-06, + "loss": 0.5267, + "step": 15475 + }, + { + "epoch": 0.3406559542152762, + "grad_norm": 1.4248077835459574, + "learning_rate": 7.675653897643943e-06, + "loss": 0.3072, + "step": 15476 + }, + { + "epoch": 0.34067796610169493, + "grad_norm": 1.6631729385907732, + "learning_rate": 7.675352767464334e-06, + "loss": 0.2825, + "step": 15477 + }, + { + "epoch": 0.3406999779881136, + "grad_norm": 1.7626218812661498, + "learning_rate": 7.675051623687391e-06, + "loss": 0.311, + "step": 15478 + }, + { + "epoch": 0.3407219898745322, + "grad_norm": 1.401427235263252, + "learning_rate": 7.674750466314639e-06, + "loss": 0.263, + "step": 15479 + }, + { + "epoch": 0.34074400176095093, + "grad_norm": 1.3947119812692588, + "learning_rate": 7.674449295347617e-06, + "loss": 0.3509, + "step": 15480 + }, + { + "epoch": 0.3407660136473696, + "grad_norm": 5.565042377058975, + "learning_rate": 7.674148110787848e-06, + "loss": 0.2844, + "step": 15481 + }, + { + "epoch": 0.34078802553378823, + "grad_norm": 1.3891794166328613, + "learning_rate": 7.673846912636864e-06, + "loss": 0.2723, + "step": 15482 + }, + { + "epoch": 0.34081003742020693, + "grad_norm": 0.6324559228878095, + "learning_rate": 7.673545700896201e-06, + "loss": 0.4889, + "step": 15483 + }, + { + "epoch": 0.3408320493066256, + "grad_norm": 1.7013801217366804, + "learning_rate": 7.673244475567383e-06, + "loss": 0.3689, + "step": 15484 + }, + { + "epoch": 0.34085406119304423, + "grad_norm": 1.6140486133185297, + "learning_rate": 7.672943236651947e-06, + "loss": 0.2258, + "step": 15485 + }, + { + "epoch": 0.3408760730794629, + "grad_norm": 1.6590959409994974, + "learning_rate": 7.67264198415142e-06, + "loss": 0.3611, + "step": 15486 + }, + { + "epoch": 0.3408980849658816, + "grad_norm": 1.6723914193182012, + "learning_rate": 7.672340718067335e-06, + "loss": 0.3707, + "step": 15487 + }, + { + "epoch": 0.34092009685230024, + "grad_norm": 0.6071385354685906, + "learning_rate": 7.672039438401221e-06, + "loss": 0.5255, + "step": 15488 + }, + { + "epoch": 0.3409421087387189, + "grad_norm": 3.35294632316427, + "learning_rate": 7.671738145154612e-06, + "loss": 0.3685, + "step": 15489 + }, + { + "epoch": 0.3409641206251376, + "grad_norm": 1.2528524071885805, + "learning_rate": 7.67143683832904e-06, + "loss": 0.4147, + "step": 15490 + }, + { + "epoch": 0.34098613251155624, + "grad_norm": 2.5853415246165685, + "learning_rate": 7.671135517926032e-06, + "loss": 0.3579, + "step": 15491 + }, + { + "epoch": 0.3410081443979749, + "grad_norm": 1.5872482801147674, + "learning_rate": 7.670834183947122e-06, + "loss": 0.4161, + "step": 15492 + }, + { + "epoch": 0.3410301562843936, + "grad_norm": 2.0251133535211445, + "learning_rate": 7.670532836393842e-06, + "loss": 0.3674, + "step": 15493 + }, + { + "epoch": 0.34105216817081224, + "grad_norm": 0.6159652748239093, + "learning_rate": 7.670231475267725e-06, + "loss": 0.5156, + "step": 15494 + }, + { + "epoch": 0.3410741800572309, + "grad_norm": 1.5207122550656704, + "learning_rate": 7.669930100570298e-06, + "loss": 0.2739, + "step": 15495 + }, + { + "epoch": 0.3410961919436496, + "grad_norm": 1.7966964114683481, + "learning_rate": 7.669628712303096e-06, + "loss": 0.336, + "step": 15496 + }, + { + "epoch": 0.34111820383006825, + "grad_norm": 6.17977439787456, + "learning_rate": 7.669327310467648e-06, + "loss": 0.2392, + "step": 15497 + }, + { + "epoch": 0.3411402157164869, + "grad_norm": 1.5150054210419421, + "learning_rate": 7.669025895065491e-06, + "loss": 0.275, + "step": 15498 + }, + { + "epoch": 0.34116222760290554, + "grad_norm": 0.6181447328015321, + "learning_rate": 7.668724466098155e-06, + "loss": 0.506, + "step": 15499 + }, + { + "epoch": 0.34118423948932425, + "grad_norm": 1.825208174146205, + "learning_rate": 7.66842302356717e-06, + "loss": 0.3748, + "step": 15500 + }, + { + "epoch": 0.3412062513757429, + "grad_norm": 0.5803652628346914, + "learning_rate": 7.668121567474068e-06, + "loss": 0.4806, + "step": 15501 + }, + { + "epoch": 0.34122826326216155, + "grad_norm": 0.9621612876186252, + "learning_rate": 7.667820097820382e-06, + "loss": 0.2292, + "step": 15502 + }, + { + "epoch": 0.34125027514858025, + "grad_norm": 0.6256221496327625, + "learning_rate": 7.667518614607645e-06, + "loss": 0.5335, + "step": 15503 + }, + { + "epoch": 0.3412722870349989, + "grad_norm": 0.6109515262935132, + "learning_rate": 7.667217117837389e-06, + "loss": 0.4877, + "step": 15504 + }, + { + "epoch": 0.34129429892141755, + "grad_norm": 1.1953195681587407, + "learning_rate": 7.666915607511144e-06, + "loss": 0.3248, + "step": 15505 + }, + { + "epoch": 0.34131631080783625, + "grad_norm": 0.6033744900967346, + "learning_rate": 7.666614083630448e-06, + "loss": 0.4877, + "step": 15506 + }, + { + "epoch": 0.3413383226942549, + "grad_norm": 1.4118499635175006, + "learning_rate": 7.666312546196827e-06, + "loss": 0.2969, + "step": 15507 + }, + { + "epoch": 0.34136033458067355, + "grad_norm": 10.291864269010121, + "learning_rate": 7.666010995211817e-06, + "loss": 0.42, + "step": 15508 + }, + { + "epoch": 0.34138234646709226, + "grad_norm": 2.5711718436239703, + "learning_rate": 7.665709430676951e-06, + "loss": 0.2998, + "step": 15509 + }, + { + "epoch": 0.3414043583535109, + "grad_norm": 1.6077199675469536, + "learning_rate": 7.66540785259376e-06, + "loss": 0.3665, + "step": 15510 + }, + { + "epoch": 0.34142637023992956, + "grad_norm": 0.599607932029695, + "learning_rate": 7.665106260963777e-06, + "loss": 0.4995, + "step": 15511 + }, + { + "epoch": 0.3414483821263482, + "grad_norm": 9.081911103739102, + "learning_rate": 7.664804655788536e-06, + "loss": 0.3864, + "step": 15512 + }, + { + "epoch": 0.3414703940127669, + "grad_norm": 1.163590771027976, + "learning_rate": 7.66450303706957e-06, + "loss": 0.3276, + "step": 15513 + }, + { + "epoch": 0.34149240589918556, + "grad_norm": 1.5251775143597193, + "learning_rate": 7.664201404808409e-06, + "loss": 0.4149, + "step": 15514 + }, + { + "epoch": 0.3415144177856042, + "grad_norm": 0.6151611109895905, + "learning_rate": 7.66389975900659e-06, + "loss": 0.4918, + "step": 15515 + }, + { + "epoch": 0.3415364296720229, + "grad_norm": 1.6227152323925564, + "learning_rate": 7.663598099665644e-06, + "loss": 0.3639, + "step": 15516 + }, + { + "epoch": 0.34155844155844156, + "grad_norm": 1.4895308907760056, + "learning_rate": 7.663296426787104e-06, + "loss": 0.359, + "step": 15517 + }, + { + "epoch": 0.3415804534448602, + "grad_norm": 1.3207783979957115, + "learning_rate": 7.662994740372503e-06, + "loss": 0.4033, + "step": 15518 + }, + { + "epoch": 0.3416024653312789, + "grad_norm": 0.6267121439430242, + "learning_rate": 7.662693040423376e-06, + "loss": 0.5032, + "step": 15519 + }, + { + "epoch": 0.34162447721769756, + "grad_norm": 0.6263382393835977, + "learning_rate": 7.662391326941257e-06, + "loss": 0.5126, + "step": 15520 + }, + { + "epoch": 0.3416464891041162, + "grad_norm": 1.2984079079178688, + "learning_rate": 7.662089599927676e-06, + "loss": 0.2863, + "step": 15521 + }, + { + "epoch": 0.34166850099053486, + "grad_norm": 0.6029338094685835, + "learning_rate": 7.66178785938417e-06, + "loss": 0.4926, + "step": 15522 + }, + { + "epoch": 0.34169051287695357, + "grad_norm": 3.863647283583246, + "learning_rate": 7.661486105312266e-06, + "loss": 0.3378, + "step": 15523 + }, + { + "epoch": 0.3417125247633722, + "grad_norm": 0.6030482281055276, + "learning_rate": 7.661184337713508e-06, + "loss": 0.5005, + "step": 15524 + }, + { + "epoch": 0.34173453664979087, + "grad_norm": 1.3407249167526132, + "learning_rate": 7.660882556589421e-06, + "loss": 0.3659, + "step": 15525 + }, + { + "epoch": 0.34175654853620957, + "grad_norm": 1.3789612591577807, + "learning_rate": 7.660580761941543e-06, + "loss": 0.3049, + "step": 15526 + }, + { + "epoch": 0.3417785604226282, + "grad_norm": 1.5075315858201552, + "learning_rate": 7.660278953771408e-06, + "loss": 0.3524, + "step": 15527 + }, + { + "epoch": 0.34180057230904687, + "grad_norm": 1.569365106240783, + "learning_rate": 7.659977132080548e-06, + "loss": 0.3075, + "step": 15528 + }, + { + "epoch": 0.3418225841954656, + "grad_norm": 1.169022972025359, + "learning_rate": 7.659675296870498e-06, + "loss": 0.3647, + "step": 15529 + }, + { + "epoch": 0.3418445960818842, + "grad_norm": 1.5006837703052633, + "learning_rate": 7.659373448142792e-06, + "loss": 0.4075, + "step": 15530 + }, + { + "epoch": 0.34186660796830287, + "grad_norm": 1.257815641885063, + "learning_rate": 7.659071585898963e-06, + "loss": 0.3868, + "step": 15531 + }, + { + "epoch": 0.3418886198547216, + "grad_norm": 1.3465333623207698, + "learning_rate": 7.658769710140548e-06, + "loss": 0.3323, + "step": 15532 + }, + { + "epoch": 0.3419106317411402, + "grad_norm": 1.0941571723777248, + "learning_rate": 7.658467820869076e-06, + "loss": 0.2674, + "step": 15533 + }, + { + "epoch": 0.3419326436275589, + "grad_norm": 0.6622758995827432, + "learning_rate": 7.658165918086089e-06, + "loss": 0.498, + "step": 15534 + }, + { + "epoch": 0.3419546555139775, + "grad_norm": 1.1660336280760388, + "learning_rate": 7.657864001793116e-06, + "loss": 0.3471, + "step": 15535 + }, + { + "epoch": 0.34197666740039623, + "grad_norm": 3.0617337166514487, + "learning_rate": 7.65756207199169e-06, + "loss": 0.361, + "step": 15536 + }, + { + "epoch": 0.3419986792868149, + "grad_norm": 1.6490309138395094, + "learning_rate": 7.657260128683352e-06, + "loss": 0.4264, + "step": 15537 + }, + { + "epoch": 0.3420206911732335, + "grad_norm": 1.3747874341499984, + "learning_rate": 7.65695817186963e-06, + "loss": 0.2472, + "step": 15538 + }, + { + "epoch": 0.34204270305965223, + "grad_norm": 1.2754099096615699, + "learning_rate": 7.656656201552064e-06, + "loss": 0.3092, + "step": 15539 + }, + { + "epoch": 0.3420647149460709, + "grad_norm": 1.6299390393490027, + "learning_rate": 7.656354217732185e-06, + "loss": 0.3479, + "step": 15540 + }, + { + "epoch": 0.34208672683248953, + "grad_norm": 2.7361093570459984, + "learning_rate": 7.65605222041153e-06, + "loss": 0.3766, + "step": 15541 + }, + { + "epoch": 0.34210873871890823, + "grad_norm": 0.6339840704246549, + "learning_rate": 7.655750209591632e-06, + "loss": 0.4933, + "step": 15542 + }, + { + "epoch": 0.3421307506053269, + "grad_norm": 1.4203438439926528, + "learning_rate": 7.655448185274026e-06, + "loss": 0.3055, + "step": 15543 + }, + { + "epoch": 0.34215276249174553, + "grad_norm": 1.309280869250667, + "learning_rate": 7.65514614746025e-06, + "loss": 0.2754, + "step": 15544 + }, + { + "epoch": 0.3421747743781642, + "grad_norm": 1.3700907600147518, + "learning_rate": 7.654844096151836e-06, + "loss": 0.3182, + "step": 15545 + }, + { + "epoch": 0.3421967862645829, + "grad_norm": 1.965081480408397, + "learning_rate": 7.65454203135032e-06, + "loss": 0.4243, + "step": 15546 + }, + { + "epoch": 0.34221879815100154, + "grad_norm": 1.522993637319303, + "learning_rate": 7.654239953057239e-06, + "loss": 0.3293, + "step": 15547 + }, + { + "epoch": 0.3422408100374202, + "grad_norm": 0.6127688949396555, + "learning_rate": 7.653937861274125e-06, + "loss": 0.4749, + "step": 15548 + }, + { + "epoch": 0.3422628219238389, + "grad_norm": 1.7481169225470783, + "learning_rate": 7.653635756002516e-06, + "loss": 0.3225, + "step": 15549 + }, + { + "epoch": 0.34228483381025754, + "grad_norm": 1.3327404800311193, + "learning_rate": 7.653333637243945e-06, + "loss": 0.3308, + "step": 15550 + }, + { + "epoch": 0.3423068456966762, + "grad_norm": 1.3960792416610095, + "learning_rate": 7.65303150499995e-06, + "loss": 0.4032, + "step": 15551 + }, + { + "epoch": 0.3423288575830949, + "grad_norm": 1.2422264547007649, + "learning_rate": 7.652729359272067e-06, + "loss": 0.3628, + "step": 15552 + }, + { + "epoch": 0.34235086946951354, + "grad_norm": 1.79088064062845, + "learning_rate": 7.65242720006183e-06, + "loss": 0.3055, + "step": 15553 + }, + { + "epoch": 0.3423728813559322, + "grad_norm": 1.4046340303220914, + "learning_rate": 7.652125027370773e-06, + "loss": 0.3229, + "step": 15554 + }, + { + "epoch": 0.3423948932423509, + "grad_norm": 1.4381930331799038, + "learning_rate": 7.651822841200433e-06, + "loss": 0.3834, + "step": 15555 + }, + { + "epoch": 0.34241690512876954, + "grad_norm": 1.3476224425409618, + "learning_rate": 7.65152064155235e-06, + "loss": 0.3551, + "step": 15556 + }, + { + "epoch": 0.3424389170151882, + "grad_norm": 1.5761815716914271, + "learning_rate": 7.651218428428055e-06, + "loss": 0.3327, + "step": 15557 + }, + { + "epoch": 0.34246092890160684, + "grad_norm": 1.3501134729007074, + "learning_rate": 7.650916201829085e-06, + "loss": 0.3634, + "step": 15558 + }, + { + "epoch": 0.34248294078802555, + "grad_norm": 1.322347991083888, + "learning_rate": 7.650613961756977e-06, + "loss": 0.3468, + "step": 15559 + }, + { + "epoch": 0.3425049526744442, + "grad_norm": 15.725121977178286, + "learning_rate": 7.650311708213265e-06, + "loss": 0.3401, + "step": 15560 + }, + { + "epoch": 0.34252696456086285, + "grad_norm": 0.6400374747992891, + "learning_rate": 7.650009441199488e-06, + "loss": 0.4946, + "step": 15561 + }, + { + "epoch": 0.34254897644728155, + "grad_norm": 1.3957606432403475, + "learning_rate": 7.649707160717183e-06, + "loss": 0.384, + "step": 15562 + }, + { + "epoch": 0.3425709883337002, + "grad_norm": 1.7453258515984276, + "learning_rate": 7.649404866767882e-06, + "loss": 0.3826, + "step": 15563 + }, + { + "epoch": 0.34259300022011885, + "grad_norm": 1.3769802404669926, + "learning_rate": 7.649102559353124e-06, + "loss": 0.3227, + "step": 15564 + }, + { + "epoch": 0.34261501210653755, + "grad_norm": 1.1619347189492766, + "learning_rate": 7.648800238474447e-06, + "loss": 0.251, + "step": 15565 + }, + { + "epoch": 0.3426370239929562, + "grad_norm": 1.271078734936331, + "learning_rate": 7.648497904133385e-06, + "loss": 0.3122, + "step": 15566 + }, + { + "epoch": 0.34265903587937485, + "grad_norm": 1.6650106460478273, + "learning_rate": 7.648195556331474e-06, + "loss": 0.2297, + "step": 15567 + }, + { + "epoch": 0.34268104776579356, + "grad_norm": 1.392120554645122, + "learning_rate": 7.647893195070252e-06, + "loss": 0.4278, + "step": 15568 + }, + { + "epoch": 0.3427030596522122, + "grad_norm": 1.3292034439013782, + "learning_rate": 7.647590820351259e-06, + "loss": 0.2514, + "step": 15569 + }, + { + "epoch": 0.34272507153863085, + "grad_norm": 1.6139942583404625, + "learning_rate": 7.647288432176026e-06, + "loss": 0.2978, + "step": 15570 + }, + { + "epoch": 0.3427470834250495, + "grad_norm": 0.6220643905538571, + "learning_rate": 7.646986030546093e-06, + "loss": 0.5016, + "step": 15571 + }, + { + "epoch": 0.3427690953114682, + "grad_norm": 7.073823695357616, + "learning_rate": 7.646683615462995e-06, + "loss": 0.3717, + "step": 15572 + }, + { + "epoch": 0.34279110719788686, + "grad_norm": 1.5286604934593035, + "learning_rate": 7.646381186928272e-06, + "loss": 0.3608, + "step": 15573 + }, + { + "epoch": 0.3428131190843055, + "grad_norm": 1.8426297261560287, + "learning_rate": 7.646078744943459e-06, + "loss": 0.3634, + "step": 15574 + }, + { + "epoch": 0.3428351309707242, + "grad_norm": 2.3110305081119367, + "learning_rate": 7.645776289510095e-06, + "loss": 0.3362, + "step": 15575 + }, + { + "epoch": 0.34285714285714286, + "grad_norm": 1.1995395727826614, + "learning_rate": 7.645473820629713e-06, + "loss": 0.2915, + "step": 15576 + }, + { + "epoch": 0.3428791547435615, + "grad_norm": 0.6327476726441612, + "learning_rate": 7.645171338303856e-06, + "loss": 0.5045, + "step": 15577 + }, + { + "epoch": 0.3429011666299802, + "grad_norm": 0.627850023070147, + "learning_rate": 7.644868842534057e-06, + "loss": 0.5274, + "step": 15578 + }, + { + "epoch": 0.34292317851639886, + "grad_norm": 1.1985532407832344, + "learning_rate": 7.644566333321856e-06, + "loss": 0.3554, + "step": 15579 + }, + { + "epoch": 0.3429451904028175, + "grad_norm": 1.620649271742676, + "learning_rate": 7.644263810668789e-06, + "loss": 0.2994, + "step": 15580 + }, + { + "epoch": 0.34296720228923616, + "grad_norm": 1.7091574598441932, + "learning_rate": 7.643961274576392e-06, + "loss": 0.3275, + "step": 15581 + }, + { + "epoch": 0.34298921417565487, + "grad_norm": 1.3143019540183247, + "learning_rate": 7.643658725046206e-06, + "loss": 0.2815, + "step": 15582 + }, + { + "epoch": 0.3430112260620735, + "grad_norm": 1.3433559462608897, + "learning_rate": 7.643356162079766e-06, + "loss": 0.317, + "step": 15583 + }, + { + "epoch": 0.34303323794849216, + "grad_norm": 1.6459002176035282, + "learning_rate": 7.643053585678615e-06, + "loss": 0.3317, + "step": 15584 + }, + { + "epoch": 0.34305524983491087, + "grad_norm": 0.6517049941893804, + "learning_rate": 7.642750995844284e-06, + "loss": 0.5213, + "step": 15585 + }, + { + "epoch": 0.3430772617213295, + "grad_norm": 1.6594168277375305, + "learning_rate": 7.642448392578314e-06, + "loss": 0.3404, + "step": 15586 + }, + { + "epoch": 0.34309927360774817, + "grad_norm": 0.6119849371150163, + "learning_rate": 7.642145775882241e-06, + "loss": 0.4864, + "step": 15587 + }, + { + "epoch": 0.34312128549416687, + "grad_norm": 1.199057828070458, + "learning_rate": 7.641843145757609e-06, + "loss": 0.3587, + "step": 15588 + }, + { + "epoch": 0.3431432973805855, + "grad_norm": 0.6256314951734019, + "learning_rate": 7.641540502205949e-06, + "loss": 0.5188, + "step": 15589 + }, + { + "epoch": 0.34316530926700417, + "grad_norm": 1.5353865035508154, + "learning_rate": 7.641237845228801e-06, + "loss": 0.3639, + "step": 15590 + }, + { + "epoch": 0.3431873211534229, + "grad_norm": 1.809078109515978, + "learning_rate": 7.640935174827704e-06, + "loss": 0.3164, + "step": 15591 + }, + { + "epoch": 0.3432093330398415, + "grad_norm": 0.6256192455526707, + "learning_rate": 7.6406324910042e-06, + "loss": 0.5026, + "step": 15592 + }, + { + "epoch": 0.3432313449262602, + "grad_norm": 1.5870750733492596, + "learning_rate": 7.640329793759823e-06, + "loss": 0.3761, + "step": 15593 + }, + { + "epoch": 0.3432533568126788, + "grad_norm": 1.4241594225162413, + "learning_rate": 7.64002708309611e-06, + "loss": 0.2664, + "step": 15594 + }, + { + "epoch": 0.3432753686990975, + "grad_norm": 0.5968826993273851, + "learning_rate": 7.639724359014603e-06, + "loss": 0.4821, + "step": 15595 + }, + { + "epoch": 0.3432973805855162, + "grad_norm": 1.3813313190912115, + "learning_rate": 7.63942162151684e-06, + "loss": 0.2826, + "step": 15596 + }, + { + "epoch": 0.3433193924719348, + "grad_norm": 1.6655311371290051, + "learning_rate": 7.639118870604358e-06, + "loss": 0.33, + "step": 15597 + }, + { + "epoch": 0.34334140435835353, + "grad_norm": 2.0422325436767244, + "learning_rate": 7.638816106278697e-06, + "loss": 0.3321, + "step": 15598 + }, + { + "epoch": 0.3433634162447722, + "grad_norm": 1.429097211565409, + "learning_rate": 7.638513328541397e-06, + "loss": 0.3363, + "step": 15599 + }, + { + "epoch": 0.34338542813119083, + "grad_norm": 1.1746206231069585, + "learning_rate": 7.638210537393994e-06, + "loss": 0.2481, + "step": 15600 + }, + { + "epoch": 0.34340744001760953, + "grad_norm": 2.055891394386929, + "learning_rate": 7.63790773283803e-06, + "loss": 0.3286, + "step": 15601 + }, + { + "epoch": 0.3434294519040282, + "grad_norm": 1.9274873725392512, + "learning_rate": 7.63760491487504e-06, + "loss": 0.2659, + "step": 15602 + }, + { + "epoch": 0.34345146379044683, + "grad_norm": 1.459576056483758, + "learning_rate": 7.637302083506569e-06, + "loss": 0.372, + "step": 15603 + }, + { + "epoch": 0.3434734756768655, + "grad_norm": 1.5415307518594914, + "learning_rate": 7.636999238734148e-06, + "loss": 0.368, + "step": 15604 + }, + { + "epoch": 0.3434954875632842, + "grad_norm": 1.9044172359874707, + "learning_rate": 7.636696380559325e-06, + "loss": 0.2884, + "step": 15605 + }, + { + "epoch": 0.34351749944970283, + "grad_norm": 1.7268398337557893, + "learning_rate": 7.636393508983632e-06, + "loss": 0.4527, + "step": 15606 + }, + { + "epoch": 0.3435395113361215, + "grad_norm": 1.8536112378653984, + "learning_rate": 7.636090624008614e-06, + "loss": 0.3577, + "step": 15607 + }, + { + "epoch": 0.3435615232225402, + "grad_norm": 1.0977929358431227, + "learning_rate": 7.635787725635804e-06, + "loss": 0.2326, + "step": 15608 + }, + { + "epoch": 0.34358353510895884, + "grad_norm": 1.3491000061742924, + "learning_rate": 7.635484813866747e-06, + "loss": 0.4018, + "step": 15609 + }, + { + "epoch": 0.3436055469953775, + "grad_norm": 1.5771870254375766, + "learning_rate": 7.635181888702981e-06, + "loss": 0.2674, + "step": 15610 + }, + { + "epoch": 0.3436275588817962, + "grad_norm": 1.3489753330695637, + "learning_rate": 7.634878950146046e-06, + "loss": 0.3938, + "step": 15611 + }, + { + "epoch": 0.34364957076821484, + "grad_norm": 1.6536663592543517, + "learning_rate": 7.634575998197479e-06, + "loss": 0.3062, + "step": 15612 + }, + { + "epoch": 0.3436715826546335, + "grad_norm": 1.4710229732713165, + "learning_rate": 7.634273032858824e-06, + "loss": 0.3725, + "step": 15613 + }, + { + "epoch": 0.3436935945410522, + "grad_norm": 0.6651553337717646, + "learning_rate": 7.633970054131617e-06, + "loss": 0.5117, + "step": 15614 + }, + { + "epoch": 0.34371560642747084, + "grad_norm": 1.4971503366133505, + "learning_rate": 7.6336670620174e-06, + "loss": 0.3425, + "step": 15615 + }, + { + "epoch": 0.3437376183138895, + "grad_norm": 0.6115385972580656, + "learning_rate": 7.633364056517713e-06, + "loss": 0.5031, + "step": 15616 + }, + { + "epoch": 0.34375963020030814, + "grad_norm": 1.1940340794900914, + "learning_rate": 7.633061037634095e-06, + "loss": 0.2608, + "step": 15617 + }, + { + "epoch": 0.34378164208672685, + "grad_norm": 1.4215984661640169, + "learning_rate": 7.632758005368086e-06, + "loss": 0.3676, + "step": 15618 + }, + { + "epoch": 0.3438036539731455, + "grad_norm": 1.1271814041666257, + "learning_rate": 7.632454959721226e-06, + "loss": 0.2853, + "step": 15619 + }, + { + "epoch": 0.34382566585956414, + "grad_norm": 0.626530805935541, + "learning_rate": 7.632151900695057e-06, + "loss": 0.5159, + "step": 15620 + }, + { + "epoch": 0.34384767774598285, + "grad_norm": 1.2251957935451767, + "learning_rate": 7.631848828291117e-06, + "loss": 0.3626, + "step": 15621 + }, + { + "epoch": 0.3438696896324015, + "grad_norm": 1.3910353313115504, + "learning_rate": 7.631545742510948e-06, + "loss": 0.315, + "step": 15622 + }, + { + "epoch": 0.34389170151882015, + "grad_norm": 1.3700054357313691, + "learning_rate": 7.63124264335609e-06, + "loss": 0.3758, + "step": 15623 + }, + { + "epoch": 0.34391371340523885, + "grad_norm": 2.317450359834548, + "learning_rate": 7.630939530828082e-06, + "loss": 0.331, + "step": 15624 + }, + { + "epoch": 0.3439357252916575, + "grad_norm": 1.5025322707535356, + "learning_rate": 7.630636404928469e-06, + "loss": 0.3703, + "step": 15625 + }, + { + "epoch": 0.34395773717807615, + "grad_norm": 1.76431857004686, + "learning_rate": 7.630333265658784e-06, + "loss": 0.3034, + "step": 15626 + }, + { + "epoch": 0.3439797490644948, + "grad_norm": 0.6336304655539449, + "learning_rate": 7.630030113020577e-06, + "loss": 0.4882, + "step": 15627 + }, + { + "epoch": 0.3440017609509135, + "grad_norm": 1.31133140836538, + "learning_rate": 7.629726947015382e-06, + "loss": 0.3642, + "step": 15628 + }, + { + "epoch": 0.34402377283733215, + "grad_norm": 1.4381269549376747, + "learning_rate": 7.62942376764474e-06, + "loss": 0.3397, + "step": 15629 + }, + { + "epoch": 0.3440457847237508, + "grad_norm": 1.464510770103208, + "learning_rate": 7.629120574910196e-06, + "loss": 0.4323, + "step": 15630 + }, + { + "epoch": 0.3440677966101695, + "grad_norm": 1.337204028256121, + "learning_rate": 7.628817368813286e-06, + "loss": 0.3541, + "step": 15631 + }, + { + "epoch": 0.34408980849658816, + "grad_norm": 1.3629490933507165, + "learning_rate": 7.628514149355555e-06, + "loss": 0.3287, + "step": 15632 + }, + { + "epoch": 0.3441118203830068, + "grad_norm": 1.4283216810397015, + "learning_rate": 7.628210916538543e-06, + "loss": 0.3029, + "step": 15633 + }, + { + "epoch": 0.3441338322694255, + "grad_norm": 1.4357619547773646, + "learning_rate": 7.6279076703637925e-06, + "loss": 0.3962, + "step": 15634 + }, + { + "epoch": 0.34415584415584416, + "grad_norm": 0.6506660350717164, + "learning_rate": 7.627604410832841e-06, + "loss": 0.4816, + "step": 15635 + }, + { + "epoch": 0.3441778560422628, + "grad_norm": 1.4539136536059156, + "learning_rate": 7.627301137947232e-06, + "loss": 0.3885, + "step": 15636 + }, + { + "epoch": 0.3441998679286815, + "grad_norm": 1.6469613135943506, + "learning_rate": 7.6269978517085066e-06, + "loss": 0.3655, + "step": 15637 + }, + { + "epoch": 0.34422187981510016, + "grad_norm": 1.3123003041553516, + "learning_rate": 7.626694552118208e-06, + "loss": 0.2762, + "step": 15638 + }, + { + "epoch": 0.3442438917015188, + "grad_norm": 1.6302722734738273, + "learning_rate": 7.626391239177875e-06, + "loss": 0.2914, + "step": 15639 + }, + { + "epoch": 0.34426590358793746, + "grad_norm": 1.4076082285322078, + "learning_rate": 7.626087912889051e-06, + "loss": 0.3188, + "step": 15640 + }, + { + "epoch": 0.34428791547435617, + "grad_norm": 0.6482853901125671, + "learning_rate": 7.625784573253276e-06, + "loss": 0.5056, + "step": 15641 + }, + { + "epoch": 0.3443099273607748, + "grad_norm": 1.307833510651176, + "learning_rate": 7.6254812202720925e-06, + "loss": 0.3545, + "step": 15642 + }, + { + "epoch": 0.34433193924719346, + "grad_norm": 1.7606485192307193, + "learning_rate": 7.625177853947046e-06, + "loss": 0.4352, + "step": 15643 + }, + { + "epoch": 0.34435395113361217, + "grad_norm": 1.36879456483059, + "learning_rate": 7.62487447427967e-06, + "loss": 0.3671, + "step": 15644 + }, + { + "epoch": 0.3443759630200308, + "grad_norm": 1.4027153403486674, + "learning_rate": 7.624571081271513e-06, + "loss": 0.385, + "step": 15645 + }, + { + "epoch": 0.34439797490644947, + "grad_norm": 1.1976150196764397, + "learning_rate": 7.624267674924115e-06, + "loss": 0.3621, + "step": 15646 + }, + { + "epoch": 0.34441998679286817, + "grad_norm": 1.3983564467338783, + "learning_rate": 7.623964255239019e-06, + "loss": 0.3188, + "step": 15647 + }, + { + "epoch": 0.3444419986792868, + "grad_norm": 1.4751310721971482, + "learning_rate": 7.623660822217767e-06, + "loss": 0.3529, + "step": 15648 + }, + { + "epoch": 0.34446401056570547, + "grad_norm": 1.266707205266065, + "learning_rate": 7.623357375861897e-06, + "loss": 0.2329, + "step": 15649 + }, + { + "epoch": 0.3444860224521242, + "grad_norm": 0.6353154395885581, + "learning_rate": 7.623053916172957e-06, + "loss": 0.5048, + "step": 15650 + }, + { + "epoch": 0.3445080343385428, + "grad_norm": 1.2275624342134064, + "learning_rate": 7.622750443152489e-06, + "loss": 0.2902, + "step": 15651 + }, + { + "epoch": 0.3445300462249615, + "grad_norm": 1.9971557976675027, + "learning_rate": 7.622446956802032e-06, + "loss": 0.2896, + "step": 15652 + }, + { + "epoch": 0.3445520581113801, + "grad_norm": 1.6474721427556656, + "learning_rate": 7.6221434571231305e-06, + "loss": 0.3369, + "step": 15653 + }, + { + "epoch": 0.3445740699977988, + "grad_norm": 1.747459430036246, + "learning_rate": 7.621839944117324e-06, + "loss": 0.2918, + "step": 15654 + }, + { + "epoch": 0.3445960818842175, + "grad_norm": 0.6375380442624345, + "learning_rate": 7.621536417786159e-06, + "loss": 0.5096, + "step": 15655 + }, + { + "epoch": 0.3446180937706361, + "grad_norm": 1.2022511337986892, + "learning_rate": 7.6212328781311775e-06, + "loss": 0.302, + "step": 15656 + }, + { + "epoch": 0.34464010565705483, + "grad_norm": 1.3835538534766905, + "learning_rate": 7.620929325153921e-06, + "loss": 0.378, + "step": 15657 + }, + { + "epoch": 0.3446621175434735, + "grad_norm": 1.3828008879451865, + "learning_rate": 7.620625758855932e-06, + "loss": 0.3773, + "step": 15658 + }, + { + "epoch": 0.3446841294298921, + "grad_norm": 1.2892365836469786, + "learning_rate": 7.620322179238755e-06, + "loss": 0.378, + "step": 15659 + }, + { + "epoch": 0.34470614131631083, + "grad_norm": 1.499381002212994, + "learning_rate": 7.62001858630393e-06, + "loss": 0.3279, + "step": 15660 + }, + { + "epoch": 0.3447281532027295, + "grad_norm": 1.1431851497192262, + "learning_rate": 7.6197149800530046e-06, + "loss": 0.3118, + "step": 15661 + }, + { + "epoch": 0.34475016508914813, + "grad_norm": 0.6313787298523172, + "learning_rate": 7.6194113604875166e-06, + "loss": 0.5152, + "step": 15662 + }, + { + "epoch": 0.3447721769755668, + "grad_norm": 1.6507204208467614, + "learning_rate": 7.619107727609014e-06, + "loss": 0.325, + "step": 15663 + }, + { + "epoch": 0.3447941888619855, + "grad_norm": 1.082736260041502, + "learning_rate": 7.618804081419038e-06, + "loss": 0.3172, + "step": 15664 + }, + { + "epoch": 0.34481620074840413, + "grad_norm": 1.3993354935474105, + "learning_rate": 7.61850042191913e-06, + "loss": 0.2722, + "step": 15665 + }, + { + "epoch": 0.3448382126348228, + "grad_norm": 1.2933096907050368, + "learning_rate": 7.618196749110835e-06, + "loss": 0.3357, + "step": 15666 + }, + { + "epoch": 0.3448602245212415, + "grad_norm": 0.6248717533483765, + "learning_rate": 7.617893062995697e-06, + "loss": 0.4932, + "step": 15667 + }, + { + "epoch": 0.34488223640766014, + "grad_norm": 0.610963521608389, + "learning_rate": 7.617589363575259e-06, + "loss": 0.4811, + "step": 15668 + }, + { + "epoch": 0.3449042482940788, + "grad_norm": 1.7109379477848412, + "learning_rate": 7.617285650851065e-06, + "loss": 0.3093, + "step": 15669 + }, + { + "epoch": 0.3449262601804975, + "grad_norm": 1.6230062736976176, + "learning_rate": 7.616981924824657e-06, + "loss": 0.4031, + "step": 15670 + }, + { + "epoch": 0.34494827206691614, + "grad_norm": 1.2864379664835996, + "learning_rate": 7.616678185497579e-06, + "loss": 0.3041, + "step": 15671 + }, + { + "epoch": 0.3449702839533348, + "grad_norm": 1.2876646158020568, + "learning_rate": 7.616374432871376e-06, + "loss": 0.3577, + "step": 15672 + }, + { + "epoch": 0.3449922958397535, + "grad_norm": 1.5187534718673186, + "learning_rate": 7.616070666947593e-06, + "loss": 0.3439, + "step": 15673 + }, + { + "epoch": 0.34501430772617214, + "grad_norm": 1.4670175779280688, + "learning_rate": 7.6157668877277705e-06, + "loss": 0.3178, + "step": 15674 + }, + { + "epoch": 0.3450363196125908, + "grad_norm": 1.377135445623549, + "learning_rate": 7.615463095213453e-06, + "loss": 0.3432, + "step": 15675 + }, + { + "epoch": 0.34505833149900944, + "grad_norm": 1.7283713292420293, + "learning_rate": 7.615159289406187e-06, + "loss": 0.2442, + "step": 15676 + }, + { + "epoch": 0.34508034338542815, + "grad_norm": 1.3156365013136608, + "learning_rate": 7.614855470307514e-06, + "loss": 0.2754, + "step": 15677 + }, + { + "epoch": 0.3451023552718468, + "grad_norm": 1.5567075625970597, + "learning_rate": 7.61455163791898e-06, + "loss": 0.3014, + "step": 15678 + }, + { + "epoch": 0.34512436715826544, + "grad_norm": 0.6470380614933124, + "learning_rate": 7.614247792242129e-06, + "loss": 0.5222, + "step": 15679 + }, + { + "epoch": 0.34514637904468415, + "grad_norm": 1.3363554620230078, + "learning_rate": 7.613943933278504e-06, + "loss": 0.4082, + "step": 15680 + }, + { + "epoch": 0.3451683909311028, + "grad_norm": 1.3716065153323882, + "learning_rate": 7.61364006102965e-06, + "loss": 0.2718, + "step": 15681 + }, + { + "epoch": 0.34519040281752145, + "grad_norm": 1.2964882854055684, + "learning_rate": 7.613336175497111e-06, + "loss": 0.2361, + "step": 15682 + }, + { + "epoch": 0.34521241470394015, + "grad_norm": 1.3326638350856808, + "learning_rate": 7.613032276682433e-06, + "loss": 0.312, + "step": 15683 + }, + { + "epoch": 0.3452344265903588, + "grad_norm": 2.770704669970405, + "learning_rate": 7.61272836458716e-06, + "loss": 0.3333, + "step": 15684 + }, + { + "epoch": 0.34525643847677745, + "grad_norm": 1.2778752412144623, + "learning_rate": 7.612424439212834e-06, + "loss": 0.3473, + "step": 15685 + }, + { + "epoch": 0.3452784503631961, + "grad_norm": 1.3105661732077754, + "learning_rate": 7.612120500561003e-06, + "loss": 0.3085, + "step": 15686 + }, + { + "epoch": 0.3453004622496148, + "grad_norm": 1.4015305767026558, + "learning_rate": 7.6118165486332105e-06, + "loss": 0.3705, + "step": 15687 + }, + { + "epoch": 0.34532247413603345, + "grad_norm": 0.6008313118838295, + "learning_rate": 7.611512583431002e-06, + "loss": 0.4903, + "step": 15688 + }, + { + "epoch": 0.3453444860224521, + "grad_norm": 1.270952689194938, + "learning_rate": 7.611208604955922e-06, + "loss": 0.2653, + "step": 15689 + }, + { + "epoch": 0.3453664979088708, + "grad_norm": 1.5648705875104205, + "learning_rate": 7.6109046132095135e-06, + "loss": 0.378, + "step": 15690 + }, + { + "epoch": 0.34538850979528946, + "grad_norm": 1.3824832220778864, + "learning_rate": 7.610600608193325e-06, + "loss": 0.4025, + "step": 15691 + }, + { + "epoch": 0.3454105216817081, + "grad_norm": 1.4194871035894947, + "learning_rate": 7.6102965899088986e-06, + "loss": 0.3182, + "step": 15692 + }, + { + "epoch": 0.3454325335681268, + "grad_norm": 1.3726914595515758, + "learning_rate": 7.609992558357782e-06, + "loss": 0.361, + "step": 15693 + }, + { + "epoch": 0.34545454545454546, + "grad_norm": 1.3600313994575524, + "learning_rate": 7.609688513541519e-06, + "loss": 0.4041, + "step": 15694 + }, + { + "epoch": 0.3454765573409641, + "grad_norm": 1.5432821053361425, + "learning_rate": 7.6093844554616545e-06, + "loss": 0.3233, + "step": 15695 + }, + { + "epoch": 0.3454985692273828, + "grad_norm": 1.2815353520155246, + "learning_rate": 7.609080384119734e-06, + "loss": 0.4074, + "step": 15696 + }, + { + "epoch": 0.34552058111380146, + "grad_norm": 2.1705868699863933, + "learning_rate": 7.608776299517305e-06, + "loss": 0.3589, + "step": 15697 + }, + { + "epoch": 0.3455425930002201, + "grad_norm": 2.021826240439203, + "learning_rate": 7.608472201655908e-06, + "loss": 0.3448, + "step": 15698 + }, + { + "epoch": 0.34556460488663876, + "grad_norm": 0.6278035441829287, + "learning_rate": 7.608168090537094e-06, + "loss": 0.5117, + "step": 15699 + }, + { + "epoch": 0.34558661677305746, + "grad_norm": 1.4181503284707782, + "learning_rate": 7.607863966162408e-06, + "loss": 0.4202, + "step": 15700 + }, + { + "epoch": 0.3456086286594761, + "grad_norm": 1.528357474450818, + "learning_rate": 7.60755982853339e-06, + "loss": 0.3495, + "step": 15701 + }, + { + "epoch": 0.34563064054589476, + "grad_norm": 1.414184531837259, + "learning_rate": 7.6072556776515945e-06, + "loss": 0.2906, + "step": 15702 + }, + { + "epoch": 0.34565265243231347, + "grad_norm": 1.4295368136953739, + "learning_rate": 7.606951513518559e-06, + "loss": 0.3043, + "step": 15703 + }, + { + "epoch": 0.3456746643187321, + "grad_norm": 1.8209003700158373, + "learning_rate": 7.606647336135835e-06, + "loss": 0.4219, + "step": 15704 + }, + { + "epoch": 0.34569667620515077, + "grad_norm": 1.6483870777160738, + "learning_rate": 7.606343145504968e-06, + "loss": 0.3264, + "step": 15705 + }, + { + "epoch": 0.34571868809156947, + "grad_norm": 1.2950282264017687, + "learning_rate": 7.606038941627501e-06, + "loss": 0.3151, + "step": 15706 + }, + { + "epoch": 0.3457406999779881, + "grad_norm": 1.5197276931858072, + "learning_rate": 7.605734724504982e-06, + "loss": 0.41, + "step": 15707 + }, + { + "epoch": 0.34576271186440677, + "grad_norm": 1.6129096458014502, + "learning_rate": 7.605430494138956e-06, + "loss": 0.3725, + "step": 15708 + }, + { + "epoch": 0.3457847237508255, + "grad_norm": 1.5315074849315853, + "learning_rate": 7.605126250530971e-06, + "loss": 0.3406, + "step": 15709 + }, + { + "epoch": 0.3458067356372441, + "grad_norm": 1.3500854347112021, + "learning_rate": 7.604821993682573e-06, + "loss": 0.3444, + "step": 15710 + }, + { + "epoch": 0.34582874752366277, + "grad_norm": 1.669348119670624, + "learning_rate": 7.604517723595307e-06, + "loss": 0.2492, + "step": 15711 + }, + { + "epoch": 0.3458507594100814, + "grad_norm": 2.388062683440495, + "learning_rate": 7.6042134402707215e-06, + "loss": 0.3895, + "step": 15712 + }, + { + "epoch": 0.3458727712965001, + "grad_norm": 1.6440166776774578, + "learning_rate": 7.603909143710361e-06, + "loss": 0.3256, + "step": 15713 + }, + { + "epoch": 0.3458947831829188, + "grad_norm": 1.3155635936659447, + "learning_rate": 7.603604833915772e-06, + "loss": 0.2859, + "step": 15714 + }, + { + "epoch": 0.3459167950693374, + "grad_norm": 1.668550381442606, + "learning_rate": 7.603300510888504e-06, + "loss": 0.3396, + "step": 15715 + }, + { + "epoch": 0.34593880695575613, + "grad_norm": 1.5875987971369807, + "learning_rate": 7.6029961746301e-06, + "loss": 0.3319, + "step": 15716 + }, + { + "epoch": 0.3459608188421748, + "grad_norm": 1.807660553240386, + "learning_rate": 7.602691825142109e-06, + "loss": 0.3083, + "step": 15717 + }, + { + "epoch": 0.3459828307285934, + "grad_norm": 1.3193880107217517, + "learning_rate": 7.602387462426078e-06, + "loss": 0.3193, + "step": 15718 + }, + { + "epoch": 0.34600484261501213, + "grad_norm": 0.6472318570343522, + "learning_rate": 7.602083086483552e-06, + "loss": 0.5266, + "step": 15719 + }, + { + "epoch": 0.3460268545014308, + "grad_norm": 1.0699197228600337, + "learning_rate": 7.60177869731608e-06, + "loss": 0.2952, + "step": 15720 + }, + { + "epoch": 0.34604886638784943, + "grad_norm": 0.6137223473718766, + "learning_rate": 7.601474294925207e-06, + "loss": 0.5103, + "step": 15721 + }, + { + "epoch": 0.3460708782742681, + "grad_norm": 2.3839345978232895, + "learning_rate": 7.601169879312482e-06, + "loss": 0.3272, + "step": 15722 + }, + { + "epoch": 0.3460928901606868, + "grad_norm": 1.4468740758374423, + "learning_rate": 7.600865450479453e-06, + "loss": 0.2822, + "step": 15723 + }, + { + "epoch": 0.34611490204710543, + "grad_norm": 1.4966199253915224, + "learning_rate": 7.600561008427665e-06, + "loss": 0.3503, + "step": 15724 + }, + { + "epoch": 0.3461369139335241, + "grad_norm": 1.3077056000947855, + "learning_rate": 7.600256553158665e-06, + "loss": 0.3027, + "step": 15725 + }, + { + "epoch": 0.3461589258199428, + "grad_norm": 1.8372309158295845, + "learning_rate": 7.5999520846740005e-06, + "loss": 0.3704, + "step": 15726 + }, + { + "epoch": 0.34618093770636144, + "grad_norm": 1.2953713135811358, + "learning_rate": 7.599647602975222e-06, + "loss": 0.3832, + "step": 15727 + }, + { + "epoch": 0.3462029495927801, + "grad_norm": 0.6224232207543672, + "learning_rate": 7.599343108063874e-06, + "loss": 0.4985, + "step": 15728 + }, + { + "epoch": 0.3462249614791988, + "grad_norm": 1.45041056083548, + "learning_rate": 7.599038599941506e-06, + "loss": 0.3595, + "step": 15729 + }, + { + "epoch": 0.34624697336561744, + "grad_norm": 1.9772467741570434, + "learning_rate": 7.598734078609664e-06, + "loss": 0.3946, + "step": 15730 + }, + { + "epoch": 0.3462689852520361, + "grad_norm": 1.7435513342705513, + "learning_rate": 7.5984295440698976e-06, + "loss": 0.4022, + "step": 15731 + }, + { + "epoch": 0.3462909971384548, + "grad_norm": 1.3575699832202432, + "learning_rate": 7.59812499632375e-06, + "loss": 0.372, + "step": 15732 + }, + { + "epoch": 0.34631300902487344, + "grad_norm": 1.4879829584248214, + "learning_rate": 7.597820435372776e-06, + "loss": 0.4443, + "step": 15733 + }, + { + "epoch": 0.3463350209112921, + "grad_norm": 1.5747412476131408, + "learning_rate": 7.597515861218519e-06, + "loss": 0.3363, + "step": 15734 + }, + { + "epoch": 0.34635703279771074, + "grad_norm": 1.7915364096986302, + "learning_rate": 7.597211273862526e-06, + "loss": 0.3016, + "step": 15735 + }, + { + "epoch": 0.34637904468412944, + "grad_norm": 1.4921000627373864, + "learning_rate": 7.596906673306349e-06, + "loss": 0.3148, + "step": 15736 + }, + { + "epoch": 0.3464010565705481, + "grad_norm": 1.652128962245235, + "learning_rate": 7.596602059551532e-06, + "loss": 0.3165, + "step": 15737 + }, + { + "epoch": 0.34642306845696674, + "grad_norm": 1.8781192028113056, + "learning_rate": 7.596297432599628e-06, + "loss": 0.2924, + "step": 15738 + }, + { + "epoch": 0.34644508034338545, + "grad_norm": 0.6911417746472669, + "learning_rate": 7.59599279245218e-06, + "loss": 0.5246, + "step": 15739 + }, + { + "epoch": 0.3464670922298041, + "grad_norm": 2.0182292892148443, + "learning_rate": 7.59568813911074e-06, + "loss": 0.3437, + "step": 15740 + }, + { + "epoch": 0.34648910411622275, + "grad_norm": 1.4507952493841485, + "learning_rate": 7.5953834725768535e-06, + "loss": 0.4425, + "step": 15741 + }, + { + "epoch": 0.34651111600264145, + "grad_norm": 1.3599034285061224, + "learning_rate": 7.595078792852072e-06, + "loss": 0.3031, + "step": 15742 + }, + { + "epoch": 0.3465331278890601, + "grad_norm": 0.6430980996811737, + "learning_rate": 7.594774099937944e-06, + "loss": 0.521, + "step": 15743 + }, + { + "epoch": 0.34655513977547875, + "grad_norm": 1.3642747760600622, + "learning_rate": 7.594469393836015e-06, + "loss": 0.3346, + "step": 15744 + }, + { + "epoch": 0.3465771516618974, + "grad_norm": 0.6115875082703217, + "learning_rate": 7.594164674547835e-06, + "loss": 0.4846, + "step": 15745 + }, + { + "epoch": 0.3465991635483161, + "grad_norm": 1.4475272835927382, + "learning_rate": 7.593859942074954e-06, + "loss": 0.3575, + "step": 15746 + }, + { + "epoch": 0.34662117543473475, + "grad_norm": 1.7694778295590246, + "learning_rate": 7.593555196418921e-06, + "loss": 0.3414, + "step": 15747 + }, + { + "epoch": 0.3466431873211534, + "grad_norm": 1.7095776131579108, + "learning_rate": 7.593250437581281e-06, + "loss": 0.385, + "step": 15748 + }, + { + "epoch": 0.3466651992075721, + "grad_norm": 1.583926761544975, + "learning_rate": 7.592945665563587e-06, + "loss": 0.3604, + "step": 15749 + }, + { + "epoch": 0.34668721109399075, + "grad_norm": 1.2114239141900056, + "learning_rate": 7.592640880367385e-06, + "loss": 0.2991, + "step": 15750 + }, + { + "epoch": 0.3467092229804094, + "grad_norm": 1.2371279901469527, + "learning_rate": 7.5923360819942295e-06, + "loss": 0.2315, + "step": 15751 + }, + { + "epoch": 0.3467312348668281, + "grad_norm": 1.2541868622952235, + "learning_rate": 7.5920312704456635e-06, + "loss": 0.3247, + "step": 15752 + }, + { + "epoch": 0.34675324675324676, + "grad_norm": 2.0775820581235003, + "learning_rate": 7.591726445723239e-06, + "loss": 0.3956, + "step": 15753 + }, + { + "epoch": 0.3467752586396654, + "grad_norm": 1.6128825512222738, + "learning_rate": 7.591421607828504e-06, + "loss": 0.4044, + "step": 15754 + }, + { + "epoch": 0.3467972705260841, + "grad_norm": 1.6226680312847552, + "learning_rate": 7.591116756763008e-06, + "loss": 0.3045, + "step": 15755 + }, + { + "epoch": 0.34681928241250276, + "grad_norm": 1.5791669774750827, + "learning_rate": 7.5908118925283035e-06, + "loss": 0.2516, + "step": 15756 + }, + { + "epoch": 0.3468412942989214, + "grad_norm": 1.457565515140829, + "learning_rate": 7.590507015125935e-06, + "loss": 0.3027, + "step": 15757 + }, + { + "epoch": 0.34686330618534006, + "grad_norm": 0.7726676698110408, + "learning_rate": 7.5902021245574555e-06, + "loss": 0.5161, + "step": 15758 + }, + { + "epoch": 0.34688531807175876, + "grad_norm": 1.9107847566023197, + "learning_rate": 7.589897220824413e-06, + "loss": 0.338, + "step": 15759 + }, + { + "epoch": 0.3469073299581774, + "grad_norm": 0.6709107284829873, + "learning_rate": 7.589592303928358e-06, + "loss": 0.5085, + "step": 15760 + }, + { + "epoch": 0.34692934184459606, + "grad_norm": 1.3370473046918427, + "learning_rate": 7.5892873738708415e-06, + "loss": 0.3303, + "step": 15761 + }, + { + "epoch": 0.34695135373101477, + "grad_norm": 1.9903130634995123, + "learning_rate": 7.58898243065341e-06, + "loss": 0.3881, + "step": 15762 + }, + { + "epoch": 0.3469733656174334, + "grad_norm": 1.3504827373796495, + "learning_rate": 7.588677474277615e-06, + "loss": 0.3885, + "step": 15763 + }, + { + "epoch": 0.34699537750385206, + "grad_norm": 1.4224979534208566, + "learning_rate": 7.588372504745006e-06, + "loss": 0.3511, + "step": 15764 + }, + { + "epoch": 0.34701738939027077, + "grad_norm": 0.629062315273986, + "learning_rate": 7.588067522057134e-06, + "loss": 0.5008, + "step": 15765 + }, + { + "epoch": 0.3470394012766894, + "grad_norm": 1.5553638647201697, + "learning_rate": 7.58776252621555e-06, + "loss": 0.3642, + "step": 15766 + }, + { + "epoch": 0.34706141316310807, + "grad_norm": 1.4521602142905685, + "learning_rate": 7.587457517221801e-06, + "loss": 0.4081, + "step": 15767 + }, + { + "epoch": 0.3470834250495267, + "grad_norm": 2.7914516263359994, + "learning_rate": 7.587152495077439e-06, + "loss": 0.3314, + "step": 15768 + }, + { + "epoch": 0.3471054369359454, + "grad_norm": 1.4178944593504892, + "learning_rate": 7.586847459784015e-06, + "loss": 0.3451, + "step": 15769 + }, + { + "epoch": 0.34712744882236407, + "grad_norm": 0.6662650772613781, + "learning_rate": 7.586542411343079e-06, + "loss": 0.5021, + "step": 15770 + }, + { + "epoch": 0.3471494607087827, + "grad_norm": 1.2640649377410553, + "learning_rate": 7.586237349756178e-06, + "loss": 0.3008, + "step": 15771 + }, + { + "epoch": 0.3471714725952014, + "grad_norm": 1.3365538316795538, + "learning_rate": 7.585932275024868e-06, + "loss": 0.2506, + "step": 15772 + }, + { + "epoch": 0.3471934844816201, + "grad_norm": 0.6093483201056005, + "learning_rate": 7.585627187150695e-06, + "loss": 0.4818, + "step": 15773 + }, + { + "epoch": 0.3472154963680387, + "grad_norm": 1.952783209497921, + "learning_rate": 7.585322086135214e-06, + "loss": 0.3501, + "step": 15774 + }, + { + "epoch": 0.3472375082544574, + "grad_norm": 1.5298004092309947, + "learning_rate": 7.585016971979971e-06, + "loss": 0.2575, + "step": 15775 + }, + { + "epoch": 0.3472595201408761, + "grad_norm": 0.5876575902118554, + "learning_rate": 7.584711844686518e-06, + "loss": 0.4897, + "step": 15776 + }, + { + "epoch": 0.3472815320272947, + "grad_norm": 1.432669014458462, + "learning_rate": 7.584406704256408e-06, + "loss": 0.377, + "step": 15777 + }, + { + "epoch": 0.34730354391371343, + "grad_norm": 0.6028104185860773, + "learning_rate": 7.58410155069119e-06, + "loss": 0.4905, + "step": 15778 + }, + { + "epoch": 0.3473255558001321, + "grad_norm": 1.3000357150845971, + "learning_rate": 7.583796383992417e-06, + "loss": 0.3801, + "step": 15779 + }, + { + "epoch": 0.34734756768655073, + "grad_norm": 1.8195136563118741, + "learning_rate": 7.583491204161637e-06, + "loss": 0.3369, + "step": 15780 + }, + { + "epoch": 0.3473695795729694, + "grad_norm": 1.3287645908816696, + "learning_rate": 7.583186011200401e-06, + "loss": 0.2589, + "step": 15781 + }, + { + "epoch": 0.3473915914593881, + "grad_norm": 0.644230291237614, + "learning_rate": 7.582880805110263e-06, + "loss": 0.5164, + "step": 15782 + }, + { + "epoch": 0.34741360334580673, + "grad_norm": 0.6572236122685997, + "learning_rate": 7.582575585892772e-06, + "loss": 0.4992, + "step": 15783 + }, + { + "epoch": 0.3474356152322254, + "grad_norm": 5.882265597881211, + "learning_rate": 7.582270353549481e-06, + "loss": 0.2875, + "step": 15784 + }, + { + "epoch": 0.3474576271186441, + "grad_norm": 1.2477442189008392, + "learning_rate": 7.581965108081942e-06, + "loss": 0.2955, + "step": 15785 + }, + { + "epoch": 0.34747963900506273, + "grad_norm": 1.5616537641700936, + "learning_rate": 7.5816598494917e-06, + "loss": 0.2999, + "step": 15786 + }, + { + "epoch": 0.3475016508914814, + "grad_norm": 1.6774016385317092, + "learning_rate": 7.581354577780315e-06, + "loss": 0.2775, + "step": 15787 + }, + { + "epoch": 0.3475236627779001, + "grad_norm": 1.3804685879554381, + "learning_rate": 7.581049292949334e-06, + "loss": 0.3092, + "step": 15788 + }, + { + "epoch": 0.34754567466431874, + "grad_norm": 1.7726175135492943, + "learning_rate": 7.580743995000309e-06, + "loss": 0.3677, + "step": 15789 + }, + { + "epoch": 0.3475676865507374, + "grad_norm": 1.5153717723156337, + "learning_rate": 7.5804386839347925e-06, + "loss": 0.3826, + "step": 15790 + }, + { + "epoch": 0.3475896984371561, + "grad_norm": 0.6610187131415288, + "learning_rate": 7.580133359754334e-06, + "loss": 0.5185, + "step": 15791 + }, + { + "epoch": 0.34761171032357474, + "grad_norm": 1.9388543128693572, + "learning_rate": 7.5798280224604895e-06, + "loss": 0.3365, + "step": 15792 + }, + { + "epoch": 0.3476337222099934, + "grad_norm": 1.3451251075108854, + "learning_rate": 7.5795226720548065e-06, + "loss": 0.3473, + "step": 15793 + }, + { + "epoch": 0.34765573409641204, + "grad_norm": 0.6194573495091801, + "learning_rate": 7.57921730853884e-06, + "loss": 0.5172, + "step": 15794 + }, + { + "epoch": 0.34767774598283074, + "grad_norm": 1.1760708071485513, + "learning_rate": 7.578911931914139e-06, + "loss": 0.2506, + "step": 15795 + }, + { + "epoch": 0.3476997578692494, + "grad_norm": 1.127547757643309, + "learning_rate": 7.578606542182259e-06, + "loss": 0.3001, + "step": 15796 + }, + { + "epoch": 0.34772176975566804, + "grad_norm": 1.0757170257455944, + "learning_rate": 7.57830113934475e-06, + "loss": 0.2522, + "step": 15797 + }, + { + "epoch": 0.34774378164208675, + "grad_norm": 1.776083670399273, + "learning_rate": 7.577995723403164e-06, + "loss": 0.3537, + "step": 15798 + }, + { + "epoch": 0.3477657935285054, + "grad_norm": 1.7845044824554832, + "learning_rate": 7.577690294359055e-06, + "loss": 0.3656, + "step": 15799 + }, + { + "epoch": 0.34778780541492404, + "grad_norm": 1.1917348605281461, + "learning_rate": 7.577384852213973e-06, + "loss": 0.3006, + "step": 15800 + }, + { + "epoch": 0.34780981730134275, + "grad_norm": 1.625357764064958, + "learning_rate": 7.577079396969471e-06, + "loss": 0.4103, + "step": 15801 + }, + { + "epoch": 0.3478318291877614, + "grad_norm": 1.061814038917378, + "learning_rate": 7.576773928627105e-06, + "loss": 0.3176, + "step": 15802 + }, + { + "epoch": 0.34785384107418005, + "grad_norm": 3.7433378445019336, + "learning_rate": 7.576468447188423e-06, + "loss": 0.2776, + "step": 15803 + }, + { + "epoch": 0.3478758529605987, + "grad_norm": 1.5595432391331288, + "learning_rate": 7.57616295265498e-06, + "loss": 0.3339, + "step": 15804 + }, + { + "epoch": 0.3478978648470174, + "grad_norm": 1.3607583105864902, + "learning_rate": 7.575857445028326e-06, + "loss": 0.303, + "step": 15805 + }, + { + "epoch": 0.34791987673343605, + "grad_norm": 1.2131450751783042, + "learning_rate": 7.575551924310017e-06, + "loss": 0.2791, + "step": 15806 + }, + { + "epoch": 0.3479418886198547, + "grad_norm": 1.3358703233611486, + "learning_rate": 7.575246390501605e-06, + "loss": 0.3133, + "step": 15807 + }, + { + "epoch": 0.3479639005062734, + "grad_norm": 1.5257087972359973, + "learning_rate": 7.574940843604641e-06, + "loss": 0.3036, + "step": 15808 + }, + { + "epoch": 0.34798591239269205, + "grad_norm": 1.1153046582270711, + "learning_rate": 7.574635283620679e-06, + "loss": 0.2807, + "step": 15809 + }, + { + "epoch": 0.3480079242791107, + "grad_norm": 1.2510747805551157, + "learning_rate": 7.574329710551273e-06, + "loss": 0.3597, + "step": 15810 + }, + { + "epoch": 0.3480299361655294, + "grad_norm": 1.2052367441418463, + "learning_rate": 7.574024124397974e-06, + "loss": 0.3097, + "step": 15811 + }, + { + "epoch": 0.34805194805194806, + "grad_norm": 1.479254997638285, + "learning_rate": 7.573718525162338e-06, + "loss": 0.3979, + "step": 15812 + }, + { + "epoch": 0.3480739599383667, + "grad_norm": 1.1438114767385321, + "learning_rate": 7.573412912845917e-06, + "loss": 0.2766, + "step": 15813 + }, + { + "epoch": 0.3480959718247854, + "grad_norm": 1.3056302623345168, + "learning_rate": 7.573107287450261e-06, + "loss": 0.2793, + "step": 15814 + }, + { + "epoch": 0.34811798371120406, + "grad_norm": 1.229675132383167, + "learning_rate": 7.572801648976929e-06, + "loss": 0.3115, + "step": 15815 + }, + { + "epoch": 0.3481399955976227, + "grad_norm": 1.59240504759757, + "learning_rate": 7.57249599742747e-06, + "loss": 0.336, + "step": 15816 + }, + { + "epoch": 0.34816200748404136, + "grad_norm": 1.4561106202019414, + "learning_rate": 7.572190332803439e-06, + "loss": 0.3804, + "step": 15817 + }, + { + "epoch": 0.34818401937046006, + "grad_norm": 1.096819312201823, + "learning_rate": 7.571884655106389e-06, + "loss": 0.2556, + "step": 15818 + }, + { + "epoch": 0.3482060312568787, + "grad_norm": 0.6790694269192367, + "learning_rate": 7.5715789643378755e-06, + "loss": 0.4942, + "step": 15819 + }, + { + "epoch": 0.34822804314329736, + "grad_norm": 1.3954488460474521, + "learning_rate": 7.57127326049945e-06, + "loss": 0.3475, + "step": 15820 + }, + { + "epoch": 0.34825005502971607, + "grad_norm": 1.544878722938137, + "learning_rate": 7.570967543592668e-06, + "loss": 0.3049, + "step": 15821 + }, + { + "epoch": 0.3482720669161347, + "grad_norm": 1.1851711819588773, + "learning_rate": 7.57066181361908e-06, + "loss": 0.2675, + "step": 15822 + }, + { + "epoch": 0.34829407880255336, + "grad_norm": 0.637002232181654, + "learning_rate": 7.5703560705802435e-06, + "loss": 0.5107, + "step": 15823 + }, + { + "epoch": 0.34831609068897207, + "grad_norm": 2.2755457908366576, + "learning_rate": 7.570050314477712e-06, + "loss": 0.3484, + "step": 15824 + }, + { + "epoch": 0.3483381025753907, + "grad_norm": 0.6257121300863159, + "learning_rate": 7.569744545313037e-06, + "loss": 0.5147, + "step": 15825 + }, + { + "epoch": 0.34836011446180937, + "grad_norm": 1.1094091190806326, + "learning_rate": 7.5694387630877755e-06, + "loss": 0.2997, + "step": 15826 + }, + { + "epoch": 0.348382126348228, + "grad_norm": 1.3486414196136225, + "learning_rate": 7.5691329678034784e-06, + "loss": 0.3354, + "step": 15827 + }, + { + "epoch": 0.3484041382346467, + "grad_norm": 1.6135822772098305, + "learning_rate": 7.568827159461703e-06, + "loss": 0.3426, + "step": 15828 + }, + { + "epoch": 0.34842615012106537, + "grad_norm": 1.3439288384494794, + "learning_rate": 7.568521338064001e-06, + "loss": 0.3698, + "step": 15829 + }, + { + "epoch": 0.348448162007484, + "grad_norm": 1.9423889788556286, + "learning_rate": 7.568215503611929e-06, + "loss": 0.4496, + "step": 15830 + }, + { + "epoch": 0.3484701738939027, + "grad_norm": 10.787086958562531, + "learning_rate": 7.56790965610704e-06, + "loss": 0.2189, + "step": 15831 + }, + { + "epoch": 0.3484921857803214, + "grad_norm": 0.7169334654246649, + "learning_rate": 7.567603795550889e-06, + "loss": 0.5302, + "step": 15832 + }, + { + "epoch": 0.34851419766674, + "grad_norm": 1.3054150523632804, + "learning_rate": 7.56729792194503e-06, + "loss": 0.2494, + "step": 15833 + }, + { + "epoch": 0.3485362095531587, + "grad_norm": 1.325351255526378, + "learning_rate": 7.56699203529102e-06, + "loss": 0.3593, + "step": 15834 + }, + { + "epoch": 0.3485582214395774, + "grad_norm": 1.6134586358446177, + "learning_rate": 7.566686135590408e-06, + "loss": 0.4115, + "step": 15835 + }, + { + "epoch": 0.348580233325996, + "grad_norm": 1.4959286271085834, + "learning_rate": 7.566380222844755e-06, + "loss": 0.3537, + "step": 15836 + }, + { + "epoch": 0.34860224521241473, + "grad_norm": 0.6256586530377339, + "learning_rate": 7.5660742970556125e-06, + "loss": 0.4875, + "step": 15837 + }, + { + "epoch": 0.3486242570988334, + "grad_norm": 1.6530926803625678, + "learning_rate": 7.565768358224534e-06, + "loss": 0.3143, + "step": 15838 + }, + { + "epoch": 0.34864626898525203, + "grad_norm": 1.9053241631492581, + "learning_rate": 7.565462406353079e-06, + "loss": 0.3855, + "step": 15839 + }, + { + "epoch": 0.3486682808716707, + "grad_norm": 1.590533750035983, + "learning_rate": 7.565156441442798e-06, + "loss": 0.3604, + "step": 15840 + }, + { + "epoch": 0.3486902927580894, + "grad_norm": 1.3788073800091256, + "learning_rate": 7.564850463495249e-06, + "loss": 0.3518, + "step": 15841 + }, + { + "epoch": 0.34871230464450803, + "grad_norm": 1.328315484046918, + "learning_rate": 7.564544472511985e-06, + "loss": 0.3621, + "step": 15842 + }, + { + "epoch": 0.3487343165309267, + "grad_norm": 1.2049002779417382, + "learning_rate": 7.564238468494564e-06, + "loss": 0.3248, + "step": 15843 + }, + { + "epoch": 0.3487563284173454, + "grad_norm": 1.6550894563327347, + "learning_rate": 7.563932451444538e-06, + "loss": 0.2509, + "step": 15844 + }, + { + "epoch": 0.34877834030376403, + "grad_norm": 2.5066715164114877, + "learning_rate": 7.5636264213634634e-06, + "loss": 0.3958, + "step": 15845 + }, + { + "epoch": 0.3488003521901827, + "grad_norm": 1.5320246712737542, + "learning_rate": 7.563320378252896e-06, + "loss": 0.3701, + "step": 15846 + }, + { + "epoch": 0.3488223640766014, + "grad_norm": 1.3837044570425554, + "learning_rate": 7.563014322114393e-06, + "loss": 0.2919, + "step": 15847 + }, + { + "epoch": 0.34884437596302004, + "grad_norm": 0.7019153203301428, + "learning_rate": 7.5627082529495065e-06, + "loss": 0.5322, + "step": 15848 + }, + { + "epoch": 0.3488663878494387, + "grad_norm": 1.2054535965080462, + "learning_rate": 7.562402170759793e-06, + "loss": 0.2869, + "step": 15849 + }, + { + "epoch": 0.34888839973585734, + "grad_norm": 9.197662605976364, + "learning_rate": 7.56209607554681e-06, + "loss": 0.3137, + "step": 15850 + }, + { + "epoch": 0.34891041162227604, + "grad_norm": 1.6586154386107748, + "learning_rate": 7.561789967312111e-06, + "loss": 0.3203, + "step": 15851 + }, + { + "epoch": 0.3489324235086947, + "grad_norm": 0.6301734912095137, + "learning_rate": 7.561483846057254e-06, + "loss": 0.4764, + "step": 15852 + }, + { + "epoch": 0.34895443539511334, + "grad_norm": 1.4940835909271817, + "learning_rate": 7.5611777117837925e-06, + "loss": 0.3308, + "step": 15853 + }, + { + "epoch": 0.34897644728153204, + "grad_norm": 0.6016068681215359, + "learning_rate": 7.560871564493284e-06, + "loss": 0.4921, + "step": 15854 + }, + { + "epoch": 0.3489984591679507, + "grad_norm": 1.314205535825813, + "learning_rate": 7.560565404187285e-06, + "loss": 0.3061, + "step": 15855 + }, + { + "epoch": 0.34902047105436934, + "grad_norm": 1.4433664653806353, + "learning_rate": 7.560259230867351e-06, + "loss": 0.3135, + "step": 15856 + }, + { + "epoch": 0.34904248294078805, + "grad_norm": 3.5060505749934046, + "learning_rate": 7.559953044535036e-06, + "loss": 0.3152, + "step": 15857 + }, + { + "epoch": 0.3490644948272067, + "grad_norm": 1.1743223992260183, + "learning_rate": 7.559646845191897e-06, + "loss": 0.2555, + "step": 15858 + }, + { + "epoch": 0.34908650671362534, + "grad_norm": 1.7877614039415248, + "learning_rate": 7.5593406328394924e-06, + "loss": 0.3209, + "step": 15859 + }, + { + "epoch": 0.34910851860004405, + "grad_norm": 1.4572382481234851, + "learning_rate": 7.559034407479377e-06, + "loss": 0.4274, + "step": 15860 + }, + { + "epoch": 0.3491305304864627, + "grad_norm": 1.5265343774359228, + "learning_rate": 7.558728169113108e-06, + "loss": 0.2699, + "step": 15861 + }, + { + "epoch": 0.34915254237288135, + "grad_norm": 1.5033773553712326, + "learning_rate": 7.55842191774224e-06, + "loss": 0.3056, + "step": 15862 + }, + { + "epoch": 0.3491745542593, + "grad_norm": 1.5589728418538324, + "learning_rate": 7.558115653368332e-06, + "loss": 0.2751, + "step": 15863 + }, + { + "epoch": 0.3491965661457187, + "grad_norm": 2.0821741838723393, + "learning_rate": 7.557809375992939e-06, + "loss": 0.3362, + "step": 15864 + }, + { + "epoch": 0.34921857803213735, + "grad_norm": 1.3817146514817455, + "learning_rate": 7.557503085617617e-06, + "loss": 0.3323, + "step": 15865 + }, + { + "epoch": 0.349240589918556, + "grad_norm": 1.5598159176024569, + "learning_rate": 7.557196782243925e-06, + "loss": 0.3396, + "step": 15866 + }, + { + "epoch": 0.3492626018049747, + "grad_norm": 1.704095225450137, + "learning_rate": 7.556890465873419e-06, + "loss": 0.2673, + "step": 15867 + }, + { + "epoch": 0.34928461369139335, + "grad_norm": 1.6104103857141883, + "learning_rate": 7.556584136507652e-06, + "loss": 0.3957, + "step": 15868 + }, + { + "epoch": 0.349306625577812, + "grad_norm": 1.6104514563457064, + "learning_rate": 7.556277794148187e-06, + "loss": 0.3627, + "step": 15869 + }, + { + "epoch": 0.3493286374642307, + "grad_norm": 1.4816513022223248, + "learning_rate": 7.555971438796579e-06, + "loss": 0.4486, + "step": 15870 + }, + { + "epoch": 0.34935064935064936, + "grad_norm": 1.548078655178255, + "learning_rate": 7.555665070454381e-06, + "loss": 0.2923, + "step": 15871 + }, + { + "epoch": 0.349372661237068, + "grad_norm": 1.307043937950947, + "learning_rate": 7.5553586891231544e-06, + "loss": 0.3335, + "step": 15872 + }, + { + "epoch": 0.3493946731234867, + "grad_norm": 1.0880166449614574, + "learning_rate": 7.555052294804456e-06, + "loss": 0.3064, + "step": 15873 + }, + { + "epoch": 0.34941668500990536, + "grad_norm": 2.517503651058209, + "learning_rate": 7.554745887499842e-06, + "loss": 0.4094, + "step": 15874 + }, + { + "epoch": 0.349438696896324, + "grad_norm": 1.7391250641076605, + "learning_rate": 7.55443946721087e-06, + "loss": 0.3148, + "step": 15875 + }, + { + "epoch": 0.34946070878274266, + "grad_norm": 1.2350778886449791, + "learning_rate": 7.554133033939096e-06, + "loss": 0.2882, + "step": 15876 + }, + { + "epoch": 0.34948272066916136, + "grad_norm": 1.9595769733544681, + "learning_rate": 7.553826587686079e-06, + "loss": 0.3777, + "step": 15877 + }, + { + "epoch": 0.34950473255558, + "grad_norm": 0.662456949821849, + "learning_rate": 7.553520128453378e-06, + "loss": 0.5059, + "step": 15878 + }, + { + "epoch": 0.34952674444199866, + "grad_norm": 1.4298878878026717, + "learning_rate": 7.553213656242548e-06, + "loss": 0.3327, + "step": 15879 + }, + { + "epoch": 0.34954875632841736, + "grad_norm": 1.4606573833639136, + "learning_rate": 7.552907171055149e-06, + "loss": 0.2962, + "step": 15880 + }, + { + "epoch": 0.349570768214836, + "grad_norm": 1.3932905600390182, + "learning_rate": 7.552600672892734e-06, + "loss": 0.3822, + "step": 15881 + }, + { + "epoch": 0.34959278010125466, + "grad_norm": 1.2729410012946436, + "learning_rate": 7.552294161756865e-06, + "loss": 0.3445, + "step": 15882 + }, + { + "epoch": 0.34961479198767337, + "grad_norm": 1.872304586453354, + "learning_rate": 7.551987637649099e-06, + "loss": 0.272, + "step": 15883 + }, + { + "epoch": 0.349636803874092, + "grad_norm": 0.6204249298814407, + "learning_rate": 7.551681100570993e-06, + "loss": 0.5113, + "step": 15884 + }, + { + "epoch": 0.34965881576051067, + "grad_norm": 0.6172888681061679, + "learning_rate": 7.551374550524106e-06, + "loss": 0.4979, + "step": 15885 + }, + { + "epoch": 0.3496808276469293, + "grad_norm": 1.4490390995918936, + "learning_rate": 7.551067987509996e-06, + "loss": 0.415, + "step": 15886 + }, + { + "epoch": 0.349702839533348, + "grad_norm": 6.316249503009572, + "learning_rate": 7.550761411530221e-06, + "loss": 0.4098, + "step": 15887 + }, + { + "epoch": 0.34972485141976667, + "grad_norm": 1.3649813397257742, + "learning_rate": 7.55045482258634e-06, + "loss": 0.3116, + "step": 15888 + }, + { + "epoch": 0.3497468633061853, + "grad_norm": 1.253822226118703, + "learning_rate": 7.550148220679907e-06, + "loss": 0.3277, + "step": 15889 + }, + { + "epoch": 0.349768875192604, + "grad_norm": 1.3371739455095848, + "learning_rate": 7.549841605812484e-06, + "loss": 0.4322, + "step": 15890 + }, + { + "epoch": 0.34979088707902267, + "grad_norm": 1.9219480253775632, + "learning_rate": 7.54953497798563e-06, + "loss": 0.3429, + "step": 15891 + }, + { + "epoch": 0.3498128989654413, + "grad_norm": 1.203624085746062, + "learning_rate": 7.5492283372009e-06, + "loss": 0.254, + "step": 15892 + }, + { + "epoch": 0.34983491085186, + "grad_norm": 1.4684968504415692, + "learning_rate": 7.548921683459857e-06, + "loss": 0.3728, + "step": 15893 + }, + { + "epoch": 0.3498569227382787, + "grad_norm": 1.248404101959778, + "learning_rate": 7.548615016764056e-06, + "loss": 0.323, + "step": 15894 + }, + { + "epoch": 0.3498789346246973, + "grad_norm": 0.6642867686468349, + "learning_rate": 7.548308337115058e-06, + "loss": 0.5139, + "step": 15895 + }, + { + "epoch": 0.34990094651111603, + "grad_norm": 0.7405516734495255, + "learning_rate": 7.54800164451442e-06, + "loss": 0.5128, + "step": 15896 + }, + { + "epoch": 0.3499229583975347, + "grad_norm": 1.4591012221168786, + "learning_rate": 7.5476949389637e-06, + "loss": 0.3863, + "step": 15897 + }, + { + "epoch": 0.3499449702839533, + "grad_norm": 0.6391076396019412, + "learning_rate": 7.547388220464459e-06, + "loss": 0.4994, + "step": 15898 + }, + { + "epoch": 0.349966982170372, + "grad_norm": 1.2990160210133623, + "learning_rate": 7.547081489018253e-06, + "loss": 0.4215, + "step": 15899 + }, + { + "epoch": 0.3499889940567907, + "grad_norm": 1.2819701473901441, + "learning_rate": 7.546774744626646e-06, + "loss": 0.3098, + "step": 15900 + }, + { + "epoch": 0.35001100594320933, + "grad_norm": 1.6724022129414162, + "learning_rate": 7.546467987291192e-06, + "loss": 0.329, + "step": 15901 + }, + { + "epoch": 0.350033017829628, + "grad_norm": 0.6495613689227145, + "learning_rate": 7.546161217013453e-06, + "loss": 0.5022, + "step": 15902 + }, + { + "epoch": 0.3500550297160467, + "grad_norm": 1.374711136012863, + "learning_rate": 7.5458544337949855e-06, + "loss": 0.3965, + "step": 15903 + }, + { + "epoch": 0.35007704160246533, + "grad_norm": 1.631384803929602, + "learning_rate": 7.545547637637351e-06, + "loss": 0.2828, + "step": 15904 + }, + { + "epoch": 0.350099053488884, + "grad_norm": 1.2797614338458208, + "learning_rate": 7.5452408285421075e-06, + "loss": 0.3834, + "step": 15905 + }, + { + "epoch": 0.3501210653753027, + "grad_norm": 0.6391498924456445, + "learning_rate": 7.544934006510817e-06, + "loss": 0.4926, + "step": 15906 + }, + { + "epoch": 0.35014307726172134, + "grad_norm": 1.2073912888907057, + "learning_rate": 7.544627171545034e-06, + "loss": 0.2624, + "step": 15907 + }, + { + "epoch": 0.35016508914814, + "grad_norm": 1.3348434369090822, + "learning_rate": 7.544320323646323e-06, + "loss": 0.3723, + "step": 15908 + }, + { + "epoch": 0.35018710103455863, + "grad_norm": 1.0665835230105054, + "learning_rate": 7.54401346281624e-06, + "loss": 0.2577, + "step": 15909 + }, + { + "epoch": 0.35020911292097734, + "grad_norm": 1.5116446288273182, + "learning_rate": 7.543706589056346e-06, + "loss": 0.3459, + "step": 15910 + }, + { + "epoch": 0.350231124807396, + "grad_norm": 1.7448737211657104, + "learning_rate": 7.5433997023682015e-06, + "loss": 0.276, + "step": 15911 + }, + { + "epoch": 0.35025313669381464, + "grad_norm": 1.4700428022604146, + "learning_rate": 7.543092802753364e-06, + "loss": 0.3484, + "step": 15912 + }, + { + "epoch": 0.35027514858023334, + "grad_norm": 1.4138404229816124, + "learning_rate": 7.542785890213396e-06, + "loss": 0.3333, + "step": 15913 + }, + { + "epoch": 0.350297160466652, + "grad_norm": 1.3585920741711397, + "learning_rate": 7.542478964749855e-06, + "loss": 0.3107, + "step": 15914 + }, + { + "epoch": 0.35031917235307064, + "grad_norm": 1.434544014052775, + "learning_rate": 7.5421720263643024e-06, + "loss": 0.309, + "step": 15915 + }, + { + "epoch": 0.35034118423948934, + "grad_norm": 0.6738496662891322, + "learning_rate": 7.541865075058299e-06, + "loss": 0.5014, + "step": 15916 + }, + { + "epoch": 0.350363196125908, + "grad_norm": 1.6350235868301566, + "learning_rate": 7.5415581108334e-06, + "loss": 0.3693, + "step": 15917 + }, + { + "epoch": 0.35038520801232664, + "grad_norm": 1.2378411687226363, + "learning_rate": 7.541251133691172e-06, + "loss": 0.2167, + "step": 15918 + }, + { + "epoch": 0.35040721989874535, + "grad_norm": 0.6222816345711046, + "learning_rate": 7.540944143633172e-06, + "loss": 0.5052, + "step": 15919 + }, + { + "epoch": 0.350429231785164, + "grad_norm": 1.2156207085454074, + "learning_rate": 7.54063714066096e-06, + "loss": 0.3493, + "step": 15920 + }, + { + "epoch": 0.35045124367158265, + "grad_norm": 1.1664330890330206, + "learning_rate": 7.540330124776098e-06, + "loss": 0.3233, + "step": 15921 + }, + { + "epoch": 0.3504732555580013, + "grad_norm": 2.0267332919928656, + "learning_rate": 7.540023095980144e-06, + "loss": 0.3819, + "step": 15922 + }, + { + "epoch": 0.35049526744442, + "grad_norm": 2.1731071993558406, + "learning_rate": 7.53971605427466e-06, + "loss": 0.2998, + "step": 15923 + }, + { + "epoch": 0.35051727933083865, + "grad_norm": 1.595860183068517, + "learning_rate": 7.539408999661208e-06, + "loss": 0.4247, + "step": 15924 + }, + { + "epoch": 0.3505392912172573, + "grad_norm": 1.8135859434118637, + "learning_rate": 7.539101932141344e-06, + "loss": 0.3063, + "step": 15925 + }, + { + "epoch": 0.350561303103676, + "grad_norm": 1.1626584979028542, + "learning_rate": 7.538794851716633e-06, + "loss": 0.2548, + "step": 15926 + }, + { + "epoch": 0.35058331499009465, + "grad_norm": 1.513149549778543, + "learning_rate": 7.538487758388634e-06, + "loss": 0.3061, + "step": 15927 + }, + { + "epoch": 0.3506053268765133, + "grad_norm": 1.1603130247474553, + "learning_rate": 7.538180652158907e-06, + "loss": 0.2428, + "step": 15928 + }, + { + "epoch": 0.350627338762932, + "grad_norm": 1.6533314875071445, + "learning_rate": 7.537873533029017e-06, + "loss": 0.2948, + "step": 15929 + }, + { + "epoch": 0.35064935064935066, + "grad_norm": 1.4365290054230289, + "learning_rate": 7.5375664010005176e-06, + "loss": 0.2933, + "step": 15930 + }, + { + "epoch": 0.3506713625357693, + "grad_norm": 1.2522604858196564, + "learning_rate": 7.537259256074977e-06, + "loss": 0.2539, + "step": 15931 + }, + { + "epoch": 0.350693374422188, + "grad_norm": 0.8054419683253393, + "learning_rate": 7.536952098253952e-06, + "loss": 0.5495, + "step": 15932 + }, + { + "epoch": 0.35071538630860666, + "grad_norm": 3.0196927313101196, + "learning_rate": 7.536644927539005e-06, + "loss": 0.2893, + "step": 15933 + }, + { + "epoch": 0.3507373981950253, + "grad_norm": 1.4986980806290517, + "learning_rate": 7.5363377439316985e-06, + "loss": 0.4223, + "step": 15934 + }, + { + "epoch": 0.35075941008144396, + "grad_norm": 1.2161752001856916, + "learning_rate": 7.5360305474335905e-06, + "loss": 0.441, + "step": 15935 + }, + { + "epoch": 0.35078142196786266, + "grad_norm": 1.6843371686803403, + "learning_rate": 7.535723338046243e-06, + "loss": 0.3331, + "step": 15936 + }, + { + "epoch": 0.3508034338542813, + "grad_norm": 1.7023622015856705, + "learning_rate": 7.535416115771221e-06, + "loss": 0.2589, + "step": 15937 + }, + { + "epoch": 0.35082544574069996, + "grad_norm": 1.4594111524761861, + "learning_rate": 7.535108880610083e-06, + "loss": 0.3743, + "step": 15938 + }, + { + "epoch": 0.35084745762711866, + "grad_norm": 1.6857361260082269, + "learning_rate": 7.534801632564391e-06, + "loss": 0.3811, + "step": 15939 + }, + { + "epoch": 0.3508694695135373, + "grad_norm": 1.2901372018530068, + "learning_rate": 7.5344943716357055e-06, + "loss": 0.3222, + "step": 15940 + }, + { + "epoch": 0.35089148139995596, + "grad_norm": 1.188216469780628, + "learning_rate": 7.534187097825592e-06, + "loss": 0.3217, + "step": 15941 + }, + { + "epoch": 0.35091349328637467, + "grad_norm": 1.5616455637520064, + "learning_rate": 7.533879811135607e-06, + "loss": 0.3321, + "step": 15942 + }, + { + "epoch": 0.3509355051727933, + "grad_norm": 1.3207867260491117, + "learning_rate": 7.533572511567316e-06, + "loss": 0.2648, + "step": 15943 + }, + { + "epoch": 0.35095751705921197, + "grad_norm": 0.6452575900776868, + "learning_rate": 7.533265199122277e-06, + "loss": 0.5193, + "step": 15944 + }, + { + "epoch": 0.3509795289456306, + "grad_norm": 1.3166161656733326, + "learning_rate": 7.532957873802056e-06, + "loss": 0.3824, + "step": 15945 + }, + { + "epoch": 0.3510015408320493, + "grad_norm": 0.6291304278564672, + "learning_rate": 7.5326505356082126e-06, + "loss": 0.5052, + "step": 15946 + }, + { + "epoch": 0.35102355271846797, + "grad_norm": 1.418674843271306, + "learning_rate": 7.532343184542313e-06, + "loss": 0.4171, + "step": 15947 + }, + { + "epoch": 0.3510455646048866, + "grad_norm": 10.274704738353249, + "learning_rate": 7.532035820605911e-06, + "loss": 0.3003, + "step": 15948 + }, + { + "epoch": 0.3510675764913053, + "grad_norm": 1.5140273488881708, + "learning_rate": 7.531728443800577e-06, + "loss": 0.2459, + "step": 15949 + }, + { + "epoch": 0.35108958837772397, + "grad_norm": 1.3776576795600484, + "learning_rate": 7.531421054127868e-06, + "loss": 0.3527, + "step": 15950 + }, + { + "epoch": 0.3511116002641426, + "grad_norm": 1.2742161599881325, + "learning_rate": 7.531113651589348e-06, + "loss": 0.3281, + "step": 15951 + }, + { + "epoch": 0.3511336121505613, + "grad_norm": 1.377442978260972, + "learning_rate": 7.530806236186581e-06, + "loss": 0.3682, + "step": 15952 + }, + { + "epoch": 0.35115562403698, + "grad_norm": 1.3400625013103329, + "learning_rate": 7.530498807921127e-06, + "loss": 0.3037, + "step": 15953 + }, + { + "epoch": 0.3511776359233986, + "grad_norm": 3.4313911159586503, + "learning_rate": 7.5301913667945505e-06, + "loss": 0.4249, + "step": 15954 + }, + { + "epoch": 0.35119964780981733, + "grad_norm": 1.4734597693894298, + "learning_rate": 7.529883912808411e-06, + "loss": 0.4026, + "step": 15955 + }, + { + "epoch": 0.351221659696236, + "grad_norm": 1.355762160691372, + "learning_rate": 7.529576445964275e-06, + "loss": 0.3164, + "step": 15956 + }, + { + "epoch": 0.3512436715826546, + "grad_norm": 1.5924431255576328, + "learning_rate": 7.529268966263703e-06, + "loss": 0.32, + "step": 15957 + }, + { + "epoch": 0.3512656834690733, + "grad_norm": 1.4251921375806917, + "learning_rate": 7.5289614737082565e-06, + "loss": 0.3177, + "step": 15958 + }, + { + "epoch": 0.351287695355492, + "grad_norm": 1.1114329315562144, + "learning_rate": 7.528653968299501e-06, + "loss": 0.2232, + "step": 15959 + }, + { + "epoch": 0.35130970724191063, + "grad_norm": 2.0525598137873167, + "learning_rate": 7.528346450038998e-06, + "loss": 0.3569, + "step": 15960 + }, + { + "epoch": 0.3513317191283293, + "grad_norm": 1.3521154511750157, + "learning_rate": 7.528038918928311e-06, + "loss": 0.2635, + "step": 15961 + }, + { + "epoch": 0.351353731014748, + "grad_norm": 0.7775672039271989, + "learning_rate": 7.527731374969003e-06, + "loss": 0.5112, + "step": 15962 + }, + { + "epoch": 0.35137574290116663, + "grad_norm": 1.7597190934094071, + "learning_rate": 7.527423818162636e-06, + "loss": 0.3074, + "step": 15963 + }, + { + "epoch": 0.3513977547875853, + "grad_norm": 3.648730227159635, + "learning_rate": 7.527116248510773e-06, + "loss": 0.39, + "step": 15964 + }, + { + "epoch": 0.351419766674004, + "grad_norm": 1.0707553047940392, + "learning_rate": 7.526808666014981e-06, + "loss": 0.2964, + "step": 15965 + }, + { + "epoch": 0.35144177856042264, + "grad_norm": 1.2364501365029936, + "learning_rate": 7.526501070676817e-06, + "loss": 0.3343, + "step": 15966 + }, + { + "epoch": 0.3514637904468413, + "grad_norm": 1.0714146589210547, + "learning_rate": 7.526193462497851e-06, + "loss": 0.2159, + "step": 15967 + }, + { + "epoch": 0.35148580233325993, + "grad_norm": 3.341460296113395, + "learning_rate": 7.525885841479641e-06, + "loss": 0.299, + "step": 15968 + }, + { + "epoch": 0.35150781421967864, + "grad_norm": 1.1924186591322088, + "learning_rate": 7.5255782076237515e-06, + "loss": 0.3586, + "step": 15969 + }, + { + "epoch": 0.3515298261060973, + "grad_norm": 1.047380018870169, + "learning_rate": 7.52527056093175e-06, + "loss": 0.2416, + "step": 15970 + }, + { + "epoch": 0.35155183799251594, + "grad_norm": 0.7105302100011707, + "learning_rate": 7.524962901405195e-06, + "loss": 0.5301, + "step": 15971 + }, + { + "epoch": 0.35157384987893464, + "grad_norm": 1.1778652167469175, + "learning_rate": 7.5246552290456545e-06, + "loss": 0.3741, + "step": 15972 + }, + { + "epoch": 0.3515958617653533, + "grad_norm": 1.6142763189621119, + "learning_rate": 7.524347543854689e-06, + "loss": 0.3538, + "step": 15973 + }, + { + "epoch": 0.35161787365177194, + "grad_norm": 0.6353296545523033, + "learning_rate": 7.5240398458338615e-06, + "loss": 0.4813, + "step": 15974 + }, + { + "epoch": 0.35163988553819064, + "grad_norm": 1.4829550592060257, + "learning_rate": 7.5237321349847406e-06, + "loss": 0.3332, + "step": 15975 + }, + { + "epoch": 0.3516618974246093, + "grad_norm": 1.5243217070515798, + "learning_rate": 7.523424411308885e-06, + "loss": 0.3249, + "step": 15976 + }, + { + "epoch": 0.35168390931102794, + "grad_norm": 1.348371589274163, + "learning_rate": 7.523116674807862e-06, + "loss": 0.3712, + "step": 15977 + }, + { + "epoch": 0.35170592119744665, + "grad_norm": 1.6666658076829162, + "learning_rate": 7.522808925483236e-06, + "loss": 0.3116, + "step": 15978 + }, + { + "epoch": 0.3517279330838653, + "grad_norm": 1.2192239758453849, + "learning_rate": 7.522501163336568e-06, + "loss": 0.3279, + "step": 15979 + }, + { + "epoch": 0.35174994497028395, + "grad_norm": 2.9428315818835995, + "learning_rate": 7.522193388369426e-06, + "loss": 0.3821, + "step": 15980 + }, + { + "epoch": 0.3517719568567026, + "grad_norm": 1.6743470856762097, + "learning_rate": 7.5218856005833704e-06, + "loss": 0.3767, + "step": 15981 + }, + { + "epoch": 0.3517939687431213, + "grad_norm": 0.6891035874515261, + "learning_rate": 7.521577799979968e-06, + "loss": 0.5067, + "step": 15982 + }, + { + "epoch": 0.35181598062953995, + "grad_norm": 0.6551259137925091, + "learning_rate": 7.521269986560782e-06, + "loss": 0.4844, + "step": 15983 + }, + { + "epoch": 0.3518379925159586, + "grad_norm": 1.4774087336811221, + "learning_rate": 7.520962160327379e-06, + "loss": 0.2873, + "step": 15984 + }, + { + "epoch": 0.3518600044023773, + "grad_norm": 1.4619270699332982, + "learning_rate": 7.520654321281321e-06, + "loss": 0.3074, + "step": 15985 + }, + { + "epoch": 0.35188201628879595, + "grad_norm": 1.5967174277954785, + "learning_rate": 7.5203464694241735e-06, + "loss": 0.3753, + "step": 15986 + }, + { + "epoch": 0.3519040281752146, + "grad_norm": 1.3711674487995558, + "learning_rate": 7.5200386047575006e-06, + "loss": 0.3376, + "step": 15987 + }, + { + "epoch": 0.3519260400616333, + "grad_norm": 1.5055668720642188, + "learning_rate": 7.519730727282869e-06, + "loss": 0.2642, + "step": 15988 + }, + { + "epoch": 0.35194805194805195, + "grad_norm": 1.081254349464854, + "learning_rate": 7.51942283700184e-06, + "loss": 0.3022, + "step": 15989 + }, + { + "epoch": 0.3519700638344706, + "grad_norm": 0.6788747460442832, + "learning_rate": 7.519114933915982e-06, + "loss": 0.4893, + "step": 15990 + }, + { + "epoch": 0.35199207572088925, + "grad_norm": 0.6535426630688215, + "learning_rate": 7.518807018026858e-06, + "loss": 0.5133, + "step": 15991 + }, + { + "epoch": 0.35201408760730796, + "grad_norm": 1.339124931917093, + "learning_rate": 7.518499089336032e-06, + "loss": 0.317, + "step": 15992 + }, + { + "epoch": 0.3520360994937266, + "grad_norm": 1.5381415904996005, + "learning_rate": 7.518191147845073e-06, + "loss": 0.3416, + "step": 15993 + }, + { + "epoch": 0.35205811138014526, + "grad_norm": 1.4484592552748334, + "learning_rate": 7.517883193555542e-06, + "loss": 0.3821, + "step": 15994 + }, + { + "epoch": 0.35208012326656396, + "grad_norm": 1.3774535648287125, + "learning_rate": 7.517575226469005e-06, + "loss": 0.2754, + "step": 15995 + }, + { + "epoch": 0.3521021351529826, + "grad_norm": 1.9641116875324698, + "learning_rate": 7.517267246587028e-06, + "loss": 0.3308, + "step": 15996 + }, + { + "epoch": 0.35212414703940126, + "grad_norm": 1.462846553452845, + "learning_rate": 7.5169592539111776e-06, + "loss": 0.3338, + "step": 15997 + }, + { + "epoch": 0.35214615892581996, + "grad_norm": 1.4845400187278477, + "learning_rate": 7.516651248443017e-06, + "loss": 0.3551, + "step": 15998 + }, + { + "epoch": 0.3521681708122386, + "grad_norm": 1.560626943351854, + "learning_rate": 7.516343230184111e-06, + "loss": 0.3688, + "step": 15999 + }, + { + "epoch": 0.35219018269865726, + "grad_norm": 1.427325650371291, + "learning_rate": 7.516035199136027e-06, + "loss": 0.4008, + "step": 16000 + }, + { + "epoch": 0.35221219458507597, + "grad_norm": 1.4707825403901016, + "learning_rate": 7.515727155300329e-06, + "loss": 0.3806, + "step": 16001 + }, + { + "epoch": 0.3522342064714946, + "grad_norm": 0.6443672600733418, + "learning_rate": 7.515419098678587e-06, + "loss": 0.5076, + "step": 16002 + }, + { + "epoch": 0.35225621835791326, + "grad_norm": 1.4114795884374576, + "learning_rate": 7.51511102927236e-06, + "loss": 0.3652, + "step": 16003 + }, + { + "epoch": 0.3522782302443319, + "grad_norm": 0.9923743660101206, + "learning_rate": 7.514802947083219e-06, + "loss": 0.2745, + "step": 16004 + }, + { + "epoch": 0.3523002421307506, + "grad_norm": 0.6393780613443751, + "learning_rate": 7.514494852112726e-06, + "loss": 0.496, + "step": 16005 + }, + { + "epoch": 0.35232225401716927, + "grad_norm": 1.3201210217057047, + "learning_rate": 7.514186744362451e-06, + "loss": 0.3072, + "step": 16006 + }, + { + "epoch": 0.3523442659035879, + "grad_norm": 1.1710310610299146, + "learning_rate": 7.513878623833954e-06, + "loss": 0.3318, + "step": 16007 + }, + { + "epoch": 0.3523662777900066, + "grad_norm": 1.8160602294061785, + "learning_rate": 7.513570490528808e-06, + "loss": 0.307, + "step": 16008 + }, + { + "epoch": 0.35238828967642527, + "grad_norm": 1.256731635031091, + "learning_rate": 7.513262344448573e-06, + "loss": 0.3522, + "step": 16009 + }, + { + "epoch": 0.3524103015628439, + "grad_norm": 1.6956671565180133, + "learning_rate": 7.512954185594819e-06, + "loss": 0.3981, + "step": 16010 + }, + { + "epoch": 0.3524323134492626, + "grad_norm": 1.3859195493419658, + "learning_rate": 7.512646013969111e-06, + "loss": 0.2566, + "step": 16011 + }, + { + "epoch": 0.3524543253356813, + "grad_norm": 1.271703094310674, + "learning_rate": 7.512337829573015e-06, + "loss": 0.3555, + "step": 16012 + }, + { + "epoch": 0.3524763372220999, + "grad_norm": 1.415494838485628, + "learning_rate": 7.5120296324081e-06, + "loss": 0.4599, + "step": 16013 + }, + { + "epoch": 0.3524983491085186, + "grad_norm": 1.2286070815850245, + "learning_rate": 7.511721422475927e-06, + "loss": 0.3338, + "step": 16014 + }, + { + "epoch": 0.3525203609949373, + "grad_norm": 0.6797567643317974, + "learning_rate": 7.5114131997780674e-06, + "loss": 0.5074, + "step": 16015 + }, + { + "epoch": 0.3525423728813559, + "grad_norm": 1.6286502188919965, + "learning_rate": 7.511104964316085e-06, + "loss": 0.2704, + "step": 16016 + }, + { + "epoch": 0.3525643847677746, + "grad_norm": 1.594626339606646, + "learning_rate": 7.510796716091547e-06, + "loss": 0.4102, + "step": 16017 + }, + { + "epoch": 0.3525863966541933, + "grad_norm": 1.8445176582663283, + "learning_rate": 7.510488455106019e-06, + "loss": 0.2835, + "step": 16018 + }, + { + "epoch": 0.35260840854061193, + "grad_norm": 1.9860423134672478, + "learning_rate": 7.510180181361071e-06, + "loss": 0.4211, + "step": 16019 + }, + { + "epoch": 0.3526304204270306, + "grad_norm": 1.4507167549152598, + "learning_rate": 7.509871894858267e-06, + "loss": 0.3813, + "step": 16020 + }, + { + "epoch": 0.3526524323134493, + "grad_norm": 1.6226066916575002, + "learning_rate": 7.509563595599175e-06, + "loss": 0.2888, + "step": 16021 + }, + { + "epoch": 0.35267444419986793, + "grad_norm": 1.3960649697014584, + "learning_rate": 7.5092552835853615e-06, + "loss": 0.2936, + "step": 16022 + }, + { + "epoch": 0.3526964560862866, + "grad_norm": 1.5695940962914723, + "learning_rate": 7.508946958818393e-06, + "loss": 0.2686, + "step": 16023 + }, + { + "epoch": 0.3527184679727053, + "grad_norm": 0.6950291584166342, + "learning_rate": 7.508638621299838e-06, + "loss": 0.5188, + "step": 16024 + }, + { + "epoch": 0.35274047985912393, + "grad_norm": 1.3714077870152803, + "learning_rate": 7.508330271031261e-06, + "loss": 0.2654, + "step": 16025 + }, + { + "epoch": 0.3527624917455426, + "grad_norm": 1.8663808260308627, + "learning_rate": 7.508021908014231e-06, + "loss": 0.3215, + "step": 16026 + }, + { + "epoch": 0.35278450363196123, + "grad_norm": 1.6225269786928926, + "learning_rate": 7.5077135322503165e-06, + "loss": 0.3275, + "step": 16027 + }, + { + "epoch": 0.35280651551837994, + "grad_norm": 1.5944063540234998, + "learning_rate": 7.507405143741082e-06, + "loss": 0.331, + "step": 16028 + }, + { + "epoch": 0.3528285274047986, + "grad_norm": 1.9295749105363373, + "learning_rate": 7.507096742488098e-06, + "loss": 0.394, + "step": 16029 + }, + { + "epoch": 0.35285053929121724, + "grad_norm": 1.4115267007488053, + "learning_rate": 7.506788328492929e-06, + "loss": 0.335, + "step": 16030 + }, + { + "epoch": 0.35287255117763594, + "grad_norm": 1.343441847940102, + "learning_rate": 7.506479901757143e-06, + "loss": 0.3832, + "step": 16031 + }, + { + "epoch": 0.3528945630640546, + "grad_norm": 1.6993190525560267, + "learning_rate": 7.506171462282309e-06, + "loss": 0.3448, + "step": 16032 + }, + { + "epoch": 0.35291657495047324, + "grad_norm": 1.3578412385108263, + "learning_rate": 7.505863010069994e-06, + "loss": 0.2384, + "step": 16033 + }, + { + "epoch": 0.35293858683689194, + "grad_norm": 1.6226397206840197, + "learning_rate": 7.505554545121765e-06, + "loss": 0.4365, + "step": 16034 + }, + { + "epoch": 0.3529605987233106, + "grad_norm": 1.4518329489790334, + "learning_rate": 7.505246067439192e-06, + "loss": 0.2819, + "step": 16035 + }, + { + "epoch": 0.35298261060972924, + "grad_norm": 1.4191198465632222, + "learning_rate": 7.504937577023838e-06, + "loss": 0.2798, + "step": 16036 + }, + { + "epoch": 0.35300462249614795, + "grad_norm": 4.103379536272405, + "learning_rate": 7.5046290738772754e-06, + "loss": 0.4829, + "step": 16037 + }, + { + "epoch": 0.3530266343825666, + "grad_norm": 0.6915098806086365, + "learning_rate": 7.504320558001071e-06, + "loss": 0.4867, + "step": 16038 + }, + { + "epoch": 0.35304864626898524, + "grad_norm": 1.8888517935354652, + "learning_rate": 7.504012029396793e-06, + "loss": 0.2311, + "step": 16039 + }, + { + "epoch": 0.3530706581554039, + "grad_norm": 0.630005486238212, + "learning_rate": 7.503703488066009e-06, + "loss": 0.4666, + "step": 16040 + }, + { + "epoch": 0.3530926700418226, + "grad_norm": 1.8992769905419615, + "learning_rate": 7.503394934010286e-06, + "loss": 0.3211, + "step": 16041 + }, + { + "epoch": 0.35311468192824125, + "grad_norm": 1.182502117899558, + "learning_rate": 7.503086367231194e-06, + "loss": 0.2919, + "step": 16042 + }, + { + "epoch": 0.3531366938146599, + "grad_norm": 1.326634127553808, + "learning_rate": 7.502777787730301e-06, + "loss": 0.4016, + "step": 16043 + }, + { + "epoch": 0.3531587057010786, + "grad_norm": 1.4293286340103482, + "learning_rate": 7.502469195509175e-06, + "loss": 0.4076, + "step": 16044 + }, + { + "epoch": 0.35318071758749725, + "grad_norm": 3.503155670513619, + "learning_rate": 7.502160590569385e-06, + "loss": 0.357, + "step": 16045 + }, + { + "epoch": 0.3532027294739159, + "grad_norm": 2.182235176859835, + "learning_rate": 7.5018519729124975e-06, + "loss": 0.3846, + "step": 16046 + }, + { + "epoch": 0.3532247413603346, + "grad_norm": 1.6091235234031669, + "learning_rate": 7.501543342540084e-06, + "loss": 0.3857, + "step": 16047 + }, + { + "epoch": 0.35324675324675325, + "grad_norm": 1.8415410300097803, + "learning_rate": 7.50123469945371e-06, + "loss": 0.2081, + "step": 16048 + }, + { + "epoch": 0.3532687651331719, + "grad_norm": 0.7463664087696226, + "learning_rate": 7.5009260436549455e-06, + "loss": 0.4876, + "step": 16049 + }, + { + "epoch": 0.35329077701959055, + "grad_norm": 1.828536307713549, + "learning_rate": 7.50061737514536e-06, + "loss": 0.3485, + "step": 16050 + }, + { + "epoch": 0.35331278890600926, + "grad_norm": 1.3997689615759465, + "learning_rate": 7.500308693926521e-06, + "loss": 0.3388, + "step": 16051 + }, + { + "epoch": 0.3533348007924279, + "grad_norm": 1.792692568399905, + "learning_rate": 7.500000000000001e-06, + "loss": 0.4553, + "step": 16052 + }, + { + "epoch": 0.35335681267884655, + "grad_norm": 2.3589462793665987, + "learning_rate": 7.499691293367364e-06, + "loss": 0.2951, + "step": 16053 + }, + { + "epoch": 0.35337882456526526, + "grad_norm": 1.271403221016129, + "learning_rate": 7.499382574030181e-06, + "loss": 0.3328, + "step": 16054 + }, + { + "epoch": 0.3534008364516839, + "grad_norm": 1.4735004856751541, + "learning_rate": 7.49907384199002e-06, + "loss": 0.3696, + "step": 16055 + }, + { + "epoch": 0.35342284833810256, + "grad_norm": 1.3948895387752598, + "learning_rate": 7.4987650972484526e-06, + "loss": 0.3109, + "step": 16056 + }, + { + "epoch": 0.35344486022452126, + "grad_norm": 0.6511132554721115, + "learning_rate": 7.498456339807045e-06, + "loss": 0.5212, + "step": 16057 + }, + { + "epoch": 0.3534668721109399, + "grad_norm": 1.5337844897584336, + "learning_rate": 7.49814756966737e-06, + "loss": 0.3559, + "step": 16058 + }, + { + "epoch": 0.35348888399735856, + "grad_norm": 1.3240452608147057, + "learning_rate": 7.497838786830993e-06, + "loss": 0.3931, + "step": 16059 + }, + { + "epoch": 0.35351089588377727, + "grad_norm": 1.8334937656448802, + "learning_rate": 7.497529991299486e-06, + "loss": 0.4386, + "step": 16060 + }, + { + "epoch": 0.3535329077701959, + "grad_norm": 1.2628601910238404, + "learning_rate": 7.497221183074417e-06, + "loss": 0.2993, + "step": 16061 + }, + { + "epoch": 0.35355491965661456, + "grad_norm": 1.560662989525497, + "learning_rate": 7.496912362157357e-06, + "loss": 0.3194, + "step": 16062 + }, + { + "epoch": 0.3535769315430332, + "grad_norm": 0.6329047683196292, + "learning_rate": 7.496603528549875e-06, + "loss": 0.5228, + "step": 16063 + }, + { + "epoch": 0.3535989434294519, + "grad_norm": 2.4999689420606406, + "learning_rate": 7.496294682253539e-06, + "loss": 0.2655, + "step": 16064 + }, + { + "epoch": 0.35362095531587057, + "grad_norm": 0.6190774399729481, + "learning_rate": 7.495985823269922e-06, + "loss": 0.4967, + "step": 16065 + }, + { + "epoch": 0.3536429672022892, + "grad_norm": 0.6077696842027261, + "learning_rate": 7.49567695160059e-06, + "loss": 0.4965, + "step": 16066 + }, + { + "epoch": 0.3536649790887079, + "grad_norm": 0.6235062647348178, + "learning_rate": 7.495368067247115e-06, + "loss": 0.5032, + "step": 16067 + }, + { + "epoch": 0.35368699097512657, + "grad_norm": 1.2396979789042775, + "learning_rate": 7.495059170211067e-06, + "loss": 0.2696, + "step": 16068 + }, + { + "epoch": 0.3537090028615452, + "grad_norm": 0.6253294754034106, + "learning_rate": 7.494750260494014e-06, + "loss": 0.4757, + "step": 16069 + }, + { + "epoch": 0.3537310147479639, + "grad_norm": 1.3410498022823556, + "learning_rate": 7.49444133809753e-06, + "loss": 0.3188, + "step": 16070 + }, + { + "epoch": 0.3537530266343826, + "grad_norm": 1.4877370703102182, + "learning_rate": 7.494132403023182e-06, + "loss": 0.3281, + "step": 16071 + }, + { + "epoch": 0.3537750385208012, + "grad_norm": 0.6529634153526483, + "learning_rate": 7.493823455272539e-06, + "loss": 0.5234, + "step": 16072 + }, + { + "epoch": 0.3537970504072199, + "grad_norm": 1.576685229834573, + "learning_rate": 7.493514494847174e-06, + "loss": 0.3371, + "step": 16073 + }, + { + "epoch": 0.3538190622936386, + "grad_norm": 1.3103957001149535, + "learning_rate": 7.4932055217486565e-06, + "loss": 0.2773, + "step": 16074 + }, + { + "epoch": 0.3538410741800572, + "grad_norm": 1.121879020185731, + "learning_rate": 7.492896535978556e-06, + "loss": 0.341, + "step": 16075 + }, + { + "epoch": 0.3538630860664759, + "grad_norm": 1.2018247680525105, + "learning_rate": 7.492587537538444e-06, + "loss": 0.3306, + "step": 16076 + }, + { + "epoch": 0.3538850979528946, + "grad_norm": 2.4570028128107593, + "learning_rate": 7.492278526429889e-06, + "loss": 0.3097, + "step": 16077 + }, + { + "epoch": 0.3539071098393132, + "grad_norm": 1.7751795904350702, + "learning_rate": 7.4919695026544635e-06, + "loss": 0.3595, + "step": 16078 + }, + { + "epoch": 0.3539291217257319, + "grad_norm": 1.496511971810956, + "learning_rate": 7.491660466213738e-06, + "loss": 0.3691, + "step": 16079 + }, + { + "epoch": 0.3539511336121506, + "grad_norm": 1.3344143571693141, + "learning_rate": 7.491351417109284e-06, + "loss": 0.3457, + "step": 16080 + }, + { + "epoch": 0.35397314549856923, + "grad_norm": 1.5652745567428492, + "learning_rate": 7.491042355342669e-06, + "loss": 0.3496, + "step": 16081 + }, + { + "epoch": 0.3539951573849879, + "grad_norm": 1.2233737523825, + "learning_rate": 7.490733280915465e-06, + "loss": 0.2735, + "step": 16082 + }, + { + "epoch": 0.3540171692714066, + "grad_norm": 1.5433857885300832, + "learning_rate": 7.4904241938292456e-06, + "loss": 0.3607, + "step": 16083 + }, + { + "epoch": 0.35403918115782523, + "grad_norm": 1.6774554636807908, + "learning_rate": 7.49011509408558e-06, + "loss": 0.3694, + "step": 16084 + }, + { + "epoch": 0.3540611930442439, + "grad_norm": 0.6702547233444195, + "learning_rate": 7.489805981686037e-06, + "loss": 0.5095, + "step": 16085 + }, + { + "epoch": 0.35408320493066253, + "grad_norm": 1.280294651318586, + "learning_rate": 7.48949685663219e-06, + "loss": 0.3215, + "step": 16086 + }, + { + "epoch": 0.35410521681708124, + "grad_norm": 1.8519439502459987, + "learning_rate": 7.4891877189256104e-06, + "loss": 0.2831, + "step": 16087 + }, + { + "epoch": 0.3541272287034999, + "grad_norm": 1.2858785571084848, + "learning_rate": 7.488878568567869e-06, + "loss": 0.3475, + "step": 16088 + }, + { + "epoch": 0.35414924058991853, + "grad_norm": 1.2728357260277439, + "learning_rate": 7.488569405560536e-06, + "loss": 0.2692, + "step": 16089 + }, + { + "epoch": 0.35417125247633724, + "grad_norm": 1.8040461060995179, + "learning_rate": 7.488260229905182e-06, + "loss": 0.3781, + "step": 16090 + }, + { + "epoch": 0.3541932643627559, + "grad_norm": 1.149313291846953, + "learning_rate": 7.48795104160338e-06, + "loss": 0.2925, + "step": 16091 + }, + { + "epoch": 0.35421527624917454, + "grad_norm": 3.3830703852164197, + "learning_rate": 7.487641840656703e-06, + "loss": 0.34, + "step": 16092 + }, + { + "epoch": 0.35423728813559324, + "grad_norm": 1.6398772389622367, + "learning_rate": 7.48733262706672e-06, + "loss": 0.3726, + "step": 16093 + }, + { + "epoch": 0.3542593000220119, + "grad_norm": 2.0172016741804764, + "learning_rate": 7.487023400835003e-06, + "loss": 0.3101, + "step": 16094 + }, + { + "epoch": 0.35428131190843054, + "grad_norm": 1.458326230409744, + "learning_rate": 7.486714161963124e-06, + "loss": 0.3634, + "step": 16095 + }, + { + "epoch": 0.35430332379484925, + "grad_norm": 1.5840169362522862, + "learning_rate": 7.486404910452654e-06, + "loss": 0.3742, + "step": 16096 + }, + { + "epoch": 0.3543253356812679, + "grad_norm": 1.4675468708993773, + "learning_rate": 7.486095646305166e-06, + "loss": 0.3612, + "step": 16097 + }, + { + "epoch": 0.35434734756768654, + "grad_norm": 0.6671907580104157, + "learning_rate": 7.48578636952223e-06, + "loss": 0.5119, + "step": 16098 + }, + { + "epoch": 0.3543693594541052, + "grad_norm": 1.3981279996431966, + "learning_rate": 7.48547708010542e-06, + "loss": 0.3832, + "step": 16099 + }, + { + "epoch": 0.3543913713405239, + "grad_norm": 1.7357822590177387, + "learning_rate": 7.485167778056306e-06, + "loss": 0.2825, + "step": 16100 + }, + { + "epoch": 0.35441338322694255, + "grad_norm": 1.4387818482720074, + "learning_rate": 7.484858463376461e-06, + "loss": 0.3004, + "step": 16101 + }, + { + "epoch": 0.3544353951133612, + "grad_norm": 1.558137531598019, + "learning_rate": 7.484549136067459e-06, + "loss": 0.304, + "step": 16102 + }, + { + "epoch": 0.3544574069997799, + "grad_norm": 2.212520143545027, + "learning_rate": 7.484239796130868e-06, + "loss": 0.4652, + "step": 16103 + }, + { + "epoch": 0.35447941888619855, + "grad_norm": 2.0153404545362603, + "learning_rate": 7.4839304435682634e-06, + "loss": 0.3648, + "step": 16104 + }, + { + "epoch": 0.3545014307726172, + "grad_norm": 2.578085353111104, + "learning_rate": 7.483621078381216e-06, + "loss": 0.3181, + "step": 16105 + }, + { + "epoch": 0.3545234426590359, + "grad_norm": 0.6576559079207001, + "learning_rate": 7.483311700571299e-06, + "loss": 0.5056, + "step": 16106 + }, + { + "epoch": 0.35454545454545455, + "grad_norm": 1.0797657030163472, + "learning_rate": 7.483002310140084e-06, + "loss": 0.2804, + "step": 16107 + }, + { + "epoch": 0.3545674664318732, + "grad_norm": 1.2004853129332431, + "learning_rate": 7.482692907089143e-06, + "loss": 0.3188, + "step": 16108 + }, + { + "epoch": 0.35458947831829185, + "grad_norm": 1.121698493486619, + "learning_rate": 7.48238349142005e-06, + "loss": 0.298, + "step": 16109 + }, + { + "epoch": 0.35461149020471056, + "grad_norm": 1.7099451561368233, + "learning_rate": 7.482074063134377e-06, + "loss": 0.2785, + "step": 16110 + }, + { + "epoch": 0.3546335020911292, + "grad_norm": 1.377924248350434, + "learning_rate": 7.481764622233697e-06, + "loss": 0.3985, + "step": 16111 + }, + { + "epoch": 0.35465551397754785, + "grad_norm": 1.2596491733953992, + "learning_rate": 7.481455168719581e-06, + "loss": 0.3201, + "step": 16112 + }, + { + "epoch": 0.35467752586396656, + "grad_norm": 1.3388521723917244, + "learning_rate": 7.481145702593604e-06, + "loss": 0.3042, + "step": 16113 + }, + { + "epoch": 0.3546995377503852, + "grad_norm": 1.5535879817227687, + "learning_rate": 7.480836223857337e-06, + "loss": 0.3235, + "step": 16114 + }, + { + "epoch": 0.35472154963680386, + "grad_norm": 1.1284583793458527, + "learning_rate": 7.480526732512356e-06, + "loss": 0.2398, + "step": 16115 + }, + { + "epoch": 0.35474356152322256, + "grad_norm": 1.3159910648079145, + "learning_rate": 7.48021722856023e-06, + "loss": 0.3851, + "step": 16116 + }, + { + "epoch": 0.3547655734096412, + "grad_norm": 1.2942534886045522, + "learning_rate": 7.4799077120025345e-06, + "loss": 0.3123, + "step": 16117 + }, + { + "epoch": 0.35478758529605986, + "grad_norm": 1.713306196189726, + "learning_rate": 7.479598182840842e-06, + "loss": 0.2619, + "step": 16118 + }, + { + "epoch": 0.35480959718247856, + "grad_norm": 1.0915934335831432, + "learning_rate": 7.479288641076725e-06, + "loss": 0.3452, + "step": 16119 + }, + { + "epoch": 0.3548316090688972, + "grad_norm": 1.2163754295578861, + "learning_rate": 7.478979086711758e-06, + "loss": 0.2863, + "step": 16120 + }, + { + "epoch": 0.35485362095531586, + "grad_norm": 1.3108038162607172, + "learning_rate": 7.4786695197475135e-06, + "loss": 0.2499, + "step": 16121 + }, + { + "epoch": 0.3548756328417345, + "grad_norm": 2.7987844407306293, + "learning_rate": 7.478359940185565e-06, + "loss": 0.3282, + "step": 16122 + }, + { + "epoch": 0.3548976447281532, + "grad_norm": 1.224967814638971, + "learning_rate": 7.478050348027486e-06, + "loss": 0.2647, + "step": 16123 + }, + { + "epoch": 0.35491965661457187, + "grad_norm": 0.6870697660764234, + "learning_rate": 7.477740743274849e-06, + "loss": 0.5343, + "step": 16124 + }, + { + "epoch": 0.3549416685009905, + "grad_norm": 1.2261311714698078, + "learning_rate": 7.477431125929229e-06, + "loss": 0.3388, + "step": 16125 + }, + { + "epoch": 0.3549636803874092, + "grad_norm": 0.6072533328164558, + "learning_rate": 7.477121495992199e-06, + "loss": 0.4797, + "step": 16126 + }, + { + "epoch": 0.35498569227382787, + "grad_norm": 2.5009506958414565, + "learning_rate": 7.476811853465333e-06, + "loss": 0.38, + "step": 16127 + }, + { + "epoch": 0.3550077041602465, + "grad_norm": 2.2459723943221146, + "learning_rate": 7.476502198350204e-06, + "loss": 0.3297, + "step": 16128 + }, + { + "epoch": 0.3550297160466652, + "grad_norm": 1.2733236179095346, + "learning_rate": 7.476192530648386e-06, + "loss": 0.2995, + "step": 16129 + }, + { + "epoch": 0.35505172793308387, + "grad_norm": 1.3066846696349816, + "learning_rate": 7.4758828503614545e-06, + "loss": 0.3095, + "step": 16130 + }, + { + "epoch": 0.3550737398195025, + "grad_norm": 1.5097978273235884, + "learning_rate": 7.475573157490981e-06, + "loss": 0.3443, + "step": 16131 + }, + { + "epoch": 0.35509575170592117, + "grad_norm": 1.3026711504568012, + "learning_rate": 7.47526345203854e-06, + "loss": 0.2938, + "step": 16132 + }, + { + "epoch": 0.3551177635923399, + "grad_norm": 2.6761907136309557, + "learning_rate": 7.474953734005708e-06, + "loss": 0.4059, + "step": 16133 + }, + { + "epoch": 0.3551397754787585, + "grad_norm": 1.5062194276217156, + "learning_rate": 7.474644003394056e-06, + "loss": 0.3395, + "step": 16134 + }, + { + "epoch": 0.3551617873651772, + "grad_norm": 1.3638388633932856, + "learning_rate": 7.474334260205161e-06, + "loss": 0.2567, + "step": 16135 + }, + { + "epoch": 0.3551837992515959, + "grad_norm": 1.4714083506603337, + "learning_rate": 7.4740245044405925e-06, + "loss": 0.3984, + "step": 16136 + }, + { + "epoch": 0.3552058111380145, + "grad_norm": 2.6892267018661347, + "learning_rate": 7.4737147361019305e-06, + "loss": 0.4224, + "step": 16137 + }, + { + "epoch": 0.3552278230244332, + "grad_norm": 1.3691096966518468, + "learning_rate": 7.473404955190747e-06, + "loss": 0.2901, + "step": 16138 + }, + { + "epoch": 0.3552498349108519, + "grad_norm": 1.4438554321212687, + "learning_rate": 7.473095161708614e-06, + "loss": 0.3103, + "step": 16139 + }, + { + "epoch": 0.35527184679727053, + "grad_norm": 1.7594737870381065, + "learning_rate": 7.4727853556571106e-06, + "loss": 0.3977, + "step": 16140 + }, + { + "epoch": 0.3552938586836892, + "grad_norm": 1.7699180183515177, + "learning_rate": 7.472475537037809e-06, + "loss": 0.3235, + "step": 16141 + }, + { + "epoch": 0.3553158705701079, + "grad_norm": 1.4814403022856217, + "learning_rate": 7.4721657058522835e-06, + "loss": 0.3306, + "step": 16142 + }, + { + "epoch": 0.35533788245652653, + "grad_norm": 1.528653407709061, + "learning_rate": 7.4718558621021096e-06, + "loss": 0.4016, + "step": 16143 + }, + { + "epoch": 0.3553598943429452, + "grad_norm": 0.7855728692430638, + "learning_rate": 7.4715460057888616e-06, + "loss": 0.4886, + "step": 16144 + }, + { + "epoch": 0.35538190622936383, + "grad_norm": 15.585670198913855, + "learning_rate": 7.471236136914114e-06, + "loss": 0.308, + "step": 16145 + }, + { + "epoch": 0.35540391811578254, + "grad_norm": 1.772438765848665, + "learning_rate": 7.470926255479443e-06, + "loss": 0.2473, + "step": 16146 + }, + { + "epoch": 0.3554259300022012, + "grad_norm": 2.380148447732256, + "learning_rate": 7.470616361486423e-06, + "loss": 0.3682, + "step": 16147 + }, + { + "epoch": 0.35544794188861983, + "grad_norm": 1.5721693153980056, + "learning_rate": 7.470306454936626e-06, + "loss": 0.2917, + "step": 16148 + }, + { + "epoch": 0.35546995377503854, + "grad_norm": 1.5967728721476409, + "learning_rate": 7.469996535831632e-06, + "loss": 0.3273, + "step": 16149 + }, + { + "epoch": 0.3554919656614572, + "grad_norm": 1.2754152551611464, + "learning_rate": 7.469686604173013e-06, + "loss": 0.3213, + "step": 16150 + }, + { + "epoch": 0.35551397754787584, + "grad_norm": 2.9963142198788097, + "learning_rate": 7.4693766599623465e-06, + "loss": 0.3325, + "step": 16151 + }, + { + "epoch": 0.35553598943429454, + "grad_norm": 0.6400384389896894, + "learning_rate": 7.469066703201206e-06, + "loss": 0.4696, + "step": 16152 + }, + { + "epoch": 0.3555580013207132, + "grad_norm": 0.6566749887268089, + "learning_rate": 7.468756733891167e-06, + "loss": 0.492, + "step": 16153 + }, + { + "epoch": 0.35558001320713184, + "grad_norm": 1.3205309070369005, + "learning_rate": 7.468446752033805e-06, + "loss": 0.3285, + "step": 16154 + }, + { + "epoch": 0.35560202509355054, + "grad_norm": 1.7203776568004951, + "learning_rate": 7.468136757630695e-06, + "loss": 0.3443, + "step": 16155 + }, + { + "epoch": 0.3556240369799692, + "grad_norm": 2.5936410113193245, + "learning_rate": 7.4678267506834156e-06, + "loss": 0.4019, + "step": 16156 + }, + { + "epoch": 0.35564604886638784, + "grad_norm": 1.7320516344840942, + "learning_rate": 7.4675167311935384e-06, + "loss": 0.3265, + "step": 16157 + }, + { + "epoch": 0.3556680607528065, + "grad_norm": 1.2983910476505172, + "learning_rate": 7.467206699162641e-06, + "loss": 0.2679, + "step": 16158 + }, + { + "epoch": 0.3556900726392252, + "grad_norm": 2.118549206950126, + "learning_rate": 7.4668966545922975e-06, + "loss": 0.3792, + "step": 16159 + }, + { + "epoch": 0.35571208452564385, + "grad_norm": 0.6451570068062159, + "learning_rate": 7.466586597484085e-06, + "loss": 0.5033, + "step": 16160 + }, + { + "epoch": 0.3557340964120625, + "grad_norm": 1.4508533763301608, + "learning_rate": 7.4662765278395825e-06, + "loss": 0.3996, + "step": 16161 + }, + { + "epoch": 0.3557561082984812, + "grad_norm": 2.371712201589423, + "learning_rate": 7.465966445660359e-06, + "loss": 0.3405, + "step": 16162 + }, + { + "epoch": 0.35577812018489985, + "grad_norm": 0.6093218924143432, + "learning_rate": 7.465656350947995e-06, + "loss": 0.493, + "step": 16163 + }, + { + "epoch": 0.3558001320713185, + "grad_norm": 1.1320156173043023, + "learning_rate": 7.465346243704066e-06, + "loss": 0.3135, + "step": 16164 + }, + { + "epoch": 0.3558221439577372, + "grad_norm": 1.936202803221111, + "learning_rate": 7.4650361239301484e-06, + "loss": 0.3939, + "step": 16165 + }, + { + "epoch": 0.35584415584415585, + "grad_norm": 1.5666259978947434, + "learning_rate": 7.464725991627818e-06, + "loss": 0.3347, + "step": 16166 + }, + { + "epoch": 0.3558661677305745, + "grad_norm": 1.206741657101796, + "learning_rate": 7.46441584679865e-06, + "loss": 0.3062, + "step": 16167 + }, + { + "epoch": 0.35588817961699315, + "grad_norm": 3.000031173607821, + "learning_rate": 7.464105689444221e-06, + "loss": 0.2624, + "step": 16168 + }, + { + "epoch": 0.35591019150341185, + "grad_norm": 1.2657455411774905, + "learning_rate": 7.4637955195661085e-06, + "loss": 0.3111, + "step": 16169 + }, + { + "epoch": 0.3559322033898305, + "grad_norm": 1.3981479040189364, + "learning_rate": 7.463485337165889e-06, + "loss": 0.3861, + "step": 16170 + }, + { + "epoch": 0.35595421527624915, + "grad_norm": 0.6505566358112024, + "learning_rate": 7.463175142245138e-06, + "loss": 0.5292, + "step": 16171 + }, + { + "epoch": 0.35597622716266786, + "grad_norm": 1.173752300053747, + "learning_rate": 7.46286493480543e-06, + "loss": 0.2941, + "step": 16172 + }, + { + "epoch": 0.3559982390490865, + "grad_norm": 2.211544901108548, + "learning_rate": 7.462554714848347e-06, + "loss": 0.31, + "step": 16173 + }, + { + "epoch": 0.35602025093550516, + "grad_norm": 0.6179833423771165, + "learning_rate": 7.462244482375462e-06, + "loss": 0.5034, + "step": 16174 + }, + { + "epoch": 0.35604226282192386, + "grad_norm": 1.3923671248947822, + "learning_rate": 7.461934237388351e-06, + "loss": 0.3107, + "step": 16175 + }, + { + "epoch": 0.3560642747083425, + "grad_norm": 1.3601423195547537, + "learning_rate": 7.461623979888594e-06, + "loss": 0.3566, + "step": 16176 + }, + { + "epoch": 0.35608628659476116, + "grad_norm": 1.2483436085065518, + "learning_rate": 7.461313709877765e-06, + "loss": 0.2941, + "step": 16177 + }, + { + "epoch": 0.35610829848117986, + "grad_norm": 1.35065104339488, + "learning_rate": 7.46100342735744e-06, + "loss": 0.2976, + "step": 16178 + }, + { + "epoch": 0.3561303103675985, + "grad_norm": 1.453923426503896, + "learning_rate": 7.460693132329202e-06, + "loss": 0.2806, + "step": 16179 + }, + { + "epoch": 0.35615232225401716, + "grad_norm": 1.3464650206787132, + "learning_rate": 7.460382824794621e-06, + "loss": 0.2773, + "step": 16180 + }, + { + "epoch": 0.3561743341404358, + "grad_norm": 1.4829567865098279, + "learning_rate": 7.460072504755278e-06, + "loss": 0.3684, + "step": 16181 + }, + { + "epoch": 0.3561963460268545, + "grad_norm": 1.1585319257649496, + "learning_rate": 7.45976217221275e-06, + "loss": 0.2957, + "step": 16182 + }, + { + "epoch": 0.35621835791327316, + "grad_norm": 0.637468665340524, + "learning_rate": 7.459451827168611e-06, + "loss": 0.5058, + "step": 16183 + }, + { + "epoch": 0.3562403697996918, + "grad_norm": 1.5158802766566923, + "learning_rate": 7.459141469624444e-06, + "loss": 0.3927, + "step": 16184 + }, + { + "epoch": 0.3562623816861105, + "grad_norm": 1.428573025721034, + "learning_rate": 7.458831099581822e-06, + "loss": 0.3638, + "step": 16185 + }, + { + "epoch": 0.35628439357252917, + "grad_norm": 1.4934657995555893, + "learning_rate": 7.458520717042323e-06, + "loss": 0.3484, + "step": 16186 + }, + { + "epoch": 0.3563064054589478, + "grad_norm": 1.5053018489414036, + "learning_rate": 7.458210322007527e-06, + "loss": 0.3438, + "step": 16187 + }, + { + "epoch": 0.3563284173453665, + "grad_norm": 6.319475353717218, + "learning_rate": 7.457899914479009e-06, + "loss": 0.2454, + "step": 16188 + }, + { + "epoch": 0.35635042923178517, + "grad_norm": 1.4426615550861392, + "learning_rate": 7.457589494458346e-06, + "loss": 0.3, + "step": 16189 + }, + { + "epoch": 0.3563724411182038, + "grad_norm": 3.4258257595872035, + "learning_rate": 7.457279061947118e-06, + "loss": 0.2654, + "step": 16190 + }, + { + "epoch": 0.35639445300462247, + "grad_norm": 1.6332053963618585, + "learning_rate": 7.456968616946903e-06, + "loss": 0.2629, + "step": 16191 + }, + { + "epoch": 0.3564164648910412, + "grad_norm": 1.3419927537991714, + "learning_rate": 7.456658159459276e-06, + "loss": 0.3928, + "step": 16192 + }, + { + "epoch": 0.3564384767774598, + "grad_norm": 1.4290045584594588, + "learning_rate": 7.456347689485816e-06, + "loss": 0.3031, + "step": 16193 + }, + { + "epoch": 0.35646048866387847, + "grad_norm": 0.6253898143797355, + "learning_rate": 7.456037207028104e-06, + "loss": 0.4892, + "step": 16194 + }, + { + "epoch": 0.3564825005502972, + "grad_norm": 36.70919952204365, + "learning_rate": 7.4557267120877144e-06, + "loss": 0.3249, + "step": 16195 + }, + { + "epoch": 0.3565045124367158, + "grad_norm": 1.8390302312720095, + "learning_rate": 7.455416204666225e-06, + "loss": 0.2363, + "step": 16196 + }, + { + "epoch": 0.3565265243231345, + "grad_norm": 1.9323938026956597, + "learning_rate": 7.455105684765217e-06, + "loss": 0.2798, + "step": 16197 + }, + { + "epoch": 0.3565485362095532, + "grad_norm": 1.5854890567261852, + "learning_rate": 7.454795152386265e-06, + "loss": 0.4014, + "step": 16198 + }, + { + "epoch": 0.35657054809597183, + "grad_norm": 0.6193573563638112, + "learning_rate": 7.454484607530951e-06, + "loss": 0.4968, + "step": 16199 + }, + { + "epoch": 0.3565925599823905, + "grad_norm": 2.732693488092165, + "learning_rate": 7.454174050200851e-06, + "loss": 0.3159, + "step": 16200 + }, + { + "epoch": 0.3566145718688092, + "grad_norm": 1.8104485270914235, + "learning_rate": 7.453863480397543e-06, + "loss": 0.2801, + "step": 16201 + }, + { + "epoch": 0.35663658375522783, + "grad_norm": 3.3243607548732785, + "learning_rate": 7.453552898122608e-06, + "loss": 0.3481, + "step": 16202 + }, + { + "epoch": 0.3566585956416465, + "grad_norm": 1.7274358932786387, + "learning_rate": 7.4532423033776216e-06, + "loss": 0.3208, + "step": 16203 + }, + { + "epoch": 0.35668060752806513, + "grad_norm": 1.661076493293969, + "learning_rate": 7.452931696164164e-06, + "loss": 0.3716, + "step": 16204 + }, + { + "epoch": 0.35670261941448383, + "grad_norm": 1.5299365432563983, + "learning_rate": 7.452621076483814e-06, + "loss": 0.3597, + "step": 16205 + }, + { + "epoch": 0.3567246313009025, + "grad_norm": 1.4263645308467083, + "learning_rate": 7.452310444338149e-06, + "loss": 0.3123, + "step": 16206 + }, + { + "epoch": 0.35674664318732113, + "grad_norm": 1.1993574921620953, + "learning_rate": 7.451999799728749e-06, + "loss": 0.3255, + "step": 16207 + }, + { + "epoch": 0.35676865507373984, + "grad_norm": 0.6709836618493852, + "learning_rate": 7.451689142657192e-06, + "loss": 0.5127, + "step": 16208 + }, + { + "epoch": 0.3567906669601585, + "grad_norm": 0.6411040284550781, + "learning_rate": 7.451378473125057e-06, + "loss": 0.4938, + "step": 16209 + }, + { + "epoch": 0.35681267884657714, + "grad_norm": 1.948067399890884, + "learning_rate": 7.451067791133924e-06, + "loss": 0.2734, + "step": 16210 + }, + { + "epoch": 0.35683469073299584, + "grad_norm": 1.4061841547217981, + "learning_rate": 7.450757096685372e-06, + "loss": 0.3567, + "step": 16211 + }, + { + "epoch": 0.3568567026194145, + "grad_norm": 1.1313600113986144, + "learning_rate": 7.450446389780977e-06, + "loss": 0.4596, + "step": 16212 + }, + { + "epoch": 0.35687871450583314, + "grad_norm": 1.5726835813582853, + "learning_rate": 7.450135670422323e-06, + "loss": 0.3333, + "step": 16213 + }, + { + "epoch": 0.35690072639225184, + "grad_norm": 0.6122257204622873, + "learning_rate": 7.449824938610986e-06, + "loss": 0.4978, + "step": 16214 + }, + { + "epoch": 0.3569227382786705, + "grad_norm": 1.1198336175347527, + "learning_rate": 7.4495141943485465e-06, + "loss": 0.2378, + "step": 16215 + }, + { + "epoch": 0.35694475016508914, + "grad_norm": 1.427073880433249, + "learning_rate": 7.449203437636581e-06, + "loss": 0.3582, + "step": 16216 + }, + { + "epoch": 0.3569667620515078, + "grad_norm": 0.6336404706194794, + "learning_rate": 7.448892668476673e-06, + "loss": 0.4783, + "step": 16217 + }, + { + "epoch": 0.3569887739379265, + "grad_norm": 2.334473882927871, + "learning_rate": 7.448581886870401e-06, + "loss": 0.336, + "step": 16218 + }, + { + "epoch": 0.35701078582434514, + "grad_norm": 2.9447954833292993, + "learning_rate": 7.448271092819341e-06, + "loss": 0.3935, + "step": 16219 + }, + { + "epoch": 0.3570327977107638, + "grad_norm": 1.5369497986190133, + "learning_rate": 7.4479602863250775e-06, + "loss": 0.3711, + "step": 16220 + }, + { + "epoch": 0.3570548095971825, + "grad_norm": 0.6141209597987524, + "learning_rate": 7.4476494673891875e-06, + "loss": 0.5118, + "step": 16221 + }, + { + "epoch": 0.35707682148360115, + "grad_norm": 1.2856589038388748, + "learning_rate": 7.447338636013252e-06, + "loss": 0.2732, + "step": 16222 + }, + { + "epoch": 0.3570988333700198, + "grad_norm": 1.3923857641478168, + "learning_rate": 7.447027792198849e-06, + "loss": 0.3139, + "step": 16223 + }, + { + "epoch": 0.3571208452564385, + "grad_norm": 1.5448411398037658, + "learning_rate": 7.44671693594756e-06, + "loss": 0.3974, + "step": 16224 + }, + { + "epoch": 0.35714285714285715, + "grad_norm": 5.837261037407273, + "learning_rate": 7.446406067260964e-06, + "loss": 0.3587, + "step": 16225 + }, + { + "epoch": 0.3571648690292758, + "grad_norm": 1.7881766426550372, + "learning_rate": 7.44609518614064e-06, + "loss": 0.3577, + "step": 16226 + }, + { + "epoch": 0.35718688091569445, + "grad_norm": 1.2713458877554906, + "learning_rate": 7.4457842925881705e-06, + "loss": 0.3724, + "step": 16227 + }, + { + "epoch": 0.35720889280211315, + "grad_norm": 1.7903714330371598, + "learning_rate": 7.4454733866051355e-06, + "loss": 0.3314, + "step": 16228 + }, + { + "epoch": 0.3572309046885318, + "grad_norm": 1.12837769039109, + "learning_rate": 7.445162468193113e-06, + "loss": 0.3027, + "step": 16229 + }, + { + "epoch": 0.35725291657495045, + "grad_norm": 1.5715401666630484, + "learning_rate": 7.444851537353683e-06, + "loss": 0.2783, + "step": 16230 + }, + { + "epoch": 0.35727492846136916, + "grad_norm": 1.6717999100825425, + "learning_rate": 7.4445405940884295e-06, + "loss": 0.2909, + "step": 16231 + }, + { + "epoch": 0.3572969403477878, + "grad_norm": 1.4760621814156234, + "learning_rate": 7.444229638398928e-06, + "loss": 0.2792, + "step": 16232 + }, + { + "epoch": 0.35731895223420646, + "grad_norm": 1.5144886317767852, + "learning_rate": 7.443918670286762e-06, + "loss": 0.368, + "step": 16233 + }, + { + "epoch": 0.35734096412062516, + "grad_norm": 1.127333347427768, + "learning_rate": 7.4436076897535125e-06, + "loss": 0.2958, + "step": 16234 + }, + { + "epoch": 0.3573629760070438, + "grad_norm": 1.679956530014878, + "learning_rate": 7.443296696800758e-06, + "loss": 0.306, + "step": 16235 + }, + { + "epoch": 0.35738498789346246, + "grad_norm": 1.3613191639326236, + "learning_rate": 7.4429856914300806e-06, + "loss": 0.3389, + "step": 16236 + }, + { + "epoch": 0.35740699977988116, + "grad_norm": 0.713789639088407, + "learning_rate": 7.442674673643059e-06, + "loss": 0.494, + "step": 16237 + }, + { + "epoch": 0.3574290116662998, + "grad_norm": 1.5909186212897077, + "learning_rate": 7.4423636434412775e-06, + "loss": 0.3107, + "step": 16238 + }, + { + "epoch": 0.35745102355271846, + "grad_norm": 1.4976485677327505, + "learning_rate": 7.4420526008263114e-06, + "loss": 0.2598, + "step": 16239 + }, + { + "epoch": 0.3574730354391371, + "grad_norm": 1.6673787790524763, + "learning_rate": 7.441741545799748e-06, + "loss": 0.3002, + "step": 16240 + }, + { + "epoch": 0.3574950473255558, + "grad_norm": 0.6111895038084908, + "learning_rate": 7.441430478363163e-06, + "loss": 0.4764, + "step": 16241 + }, + { + "epoch": 0.35751705921197446, + "grad_norm": 1.874694384081093, + "learning_rate": 7.4411193985181394e-06, + "loss": 0.3095, + "step": 16242 + }, + { + "epoch": 0.3575390710983931, + "grad_norm": 2.5236411117656528, + "learning_rate": 7.440808306266259e-06, + "loss": 0.3381, + "step": 16243 + }, + { + "epoch": 0.3575610829848118, + "grad_norm": 1.474834110599872, + "learning_rate": 7.440497201609102e-06, + "loss": 0.373, + "step": 16244 + }, + { + "epoch": 0.35758309487123047, + "grad_norm": 2.0402503818211506, + "learning_rate": 7.440186084548249e-06, + "loss": 0.2389, + "step": 16245 + }, + { + "epoch": 0.3576051067576491, + "grad_norm": 2.254287897891157, + "learning_rate": 7.439874955085283e-06, + "loss": 0.3568, + "step": 16246 + }, + { + "epoch": 0.3576271186440678, + "grad_norm": 2.399463027726118, + "learning_rate": 7.439563813221784e-06, + "loss": 0.2942, + "step": 16247 + }, + { + "epoch": 0.35764913053048647, + "grad_norm": 1.7402096777465563, + "learning_rate": 7.4392526589593335e-06, + "loss": 0.3777, + "step": 16248 + }, + { + "epoch": 0.3576711424169051, + "grad_norm": 1.1209129484261982, + "learning_rate": 7.438941492299514e-06, + "loss": 0.2688, + "step": 16249 + }, + { + "epoch": 0.35769315430332377, + "grad_norm": 1.6042038285536642, + "learning_rate": 7.438630313243904e-06, + "loss": 0.3501, + "step": 16250 + }, + { + "epoch": 0.3577151661897425, + "grad_norm": 1.9166533936474854, + "learning_rate": 7.438319121794089e-06, + "loss": 0.361, + "step": 16251 + }, + { + "epoch": 0.3577371780761611, + "grad_norm": 2.2900577363517276, + "learning_rate": 7.438007917951647e-06, + "loss": 0.2962, + "step": 16252 + }, + { + "epoch": 0.35775918996257977, + "grad_norm": 1.3817512711786053, + "learning_rate": 7.4376967017181626e-06, + "loss": 0.3498, + "step": 16253 + }, + { + "epoch": 0.3577812018489985, + "grad_norm": 1.6159235108572392, + "learning_rate": 7.437385473095217e-06, + "loss": 0.3129, + "step": 16254 + }, + { + "epoch": 0.3578032137354171, + "grad_norm": 1.5054810172628832, + "learning_rate": 7.437074232084389e-06, + "loss": 0.3655, + "step": 16255 + }, + { + "epoch": 0.3578252256218358, + "grad_norm": 1.4598053530768755, + "learning_rate": 7.4367629786872654e-06, + "loss": 0.3432, + "step": 16256 + }, + { + "epoch": 0.3578472375082545, + "grad_norm": 1.1456202005553193, + "learning_rate": 7.436451712905424e-06, + "loss": 0.3049, + "step": 16257 + }, + { + "epoch": 0.35786924939467313, + "grad_norm": 1.2797020999836428, + "learning_rate": 7.436140434740447e-06, + "loss": 0.3106, + "step": 16258 + }, + { + "epoch": 0.3578912612810918, + "grad_norm": 1.2972958099375698, + "learning_rate": 7.4358291441939204e-06, + "loss": 0.2874, + "step": 16259 + }, + { + "epoch": 0.3579132731675105, + "grad_norm": 1.6003008237793948, + "learning_rate": 7.435517841267421e-06, + "loss": 0.3703, + "step": 16260 + }, + { + "epoch": 0.35793528505392913, + "grad_norm": 1.5427686746772762, + "learning_rate": 7.435206525962537e-06, + "loss": 0.334, + "step": 16261 + }, + { + "epoch": 0.3579572969403478, + "grad_norm": 1.7228346816343756, + "learning_rate": 7.434895198280845e-06, + "loss": 0.3884, + "step": 16262 + }, + { + "epoch": 0.35797930882676643, + "grad_norm": 2.1011602429037897, + "learning_rate": 7.4345838582239306e-06, + "loss": 0.2437, + "step": 16263 + }, + { + "epoch": 0.35800132071318513, + "grad_norm": 1.829820325448296, + "learning_rate": 7.434272505793374e-06, + "loss": 0.3482, + "step": 16264 + }, + { + "epoch": 0.3580233325996038, + "grad_norm": 1.7701369829036397, + "learning_rate": 7.4339611409907596e-06, + "loss": 0.3419, + "step": 16265 + }, + { + "epoch": 0.35804534448602243, + "grad_norm": 1.2352074125954458, + "learning_rate": 7.433649763817669e-06, + "loss": 0.3846, + "step": 16266 + }, + { + "epoch": 0.35806735637244114, + "grad_norm": 0.748927964368957, + "learning_rate": 7.433338374275686e-06, + "loss": 0.497, + "step": 16267 + }, + { + "epoch": 0.3580893682588598, + "grad_norm": 5.904083270595377, + "learning_rate": 7.433026972366391e-06, + "loss": 0.3924, + "step": 16268 + }, + { + "epoch": 0.35811138014527844, + "grad_norm": 0.6428203022809184, + "learning_rate": 7.432715558091367e-06, + "loss": 0.4981, + "step": 16269 + }, + { + "epoch": 0.35813339203169714, + "grad_norm": 1.304146112979575, + "learning_rate": 7.432404131452198e-06, + "loss": 0.2807, + "step": 16270 + }, + { + "epoch": 0.3581554039181158, + "grad_norm": 1.2486549344597566, + "learning_rate": 7.432092692450466e-06, + "loss": 0.3599, + "step": 16271 + }, + { + "epoch": 0.35817741580453444, + "grad_norm": 1.717666293835596, + "learning_rate": 7.431781241087757e-06, + "loss": 0.3378, + "step": 16272 + }, + { + "epoch": 0.3581994276909531, + "grad_norm": 1.4978750501750262, + "learning_rate": 7.431469777365647e-06, + "loss": 0.3267, + "step": 16273 + }, + { + "epoch": 0.3582214395773718, + "grad_norm": 0.6743199133144555, + "learning_rate": 7.431158301285726e-06, + "loss": 0.4891, + "step": 16274 + }, + { + "epoch": 0.35824345146379044, + "grad_norm": 1.5488405614499619, + "learning_rate": 7.430846812849572e-06, + "loss": 0.4023, + "step": 16275 + }, + { + "epoch": 0.3582654633502091, + "grad_norm": 0.6936256383194472, + "learning_rate": 7.430535312058772e-06, + "loss": 0.5283, + "step": 16276 + }, + { + "epoch": 0.3582874752366278, + "grad_norm": 1.87218157829916, + "learning_rate": 7.430223798914907e-06, + "loss": 0.3331, + "step": 16277 + }, + { + "epoch": 0.35830948712304644, + "grad_norm": 1.3580529974346605, + "learning_rate": 7.4299122734195606e-06, + "loss": 0.3284, + "step": 16278 + }, + { + "epoch": 0.3583314990094651, + "grad_norm": 2.5910584945695176, + "learning_rate": 7.429600735574317e-06, + "loss": 0.3073, + "step": 16279 + }, + { + "epoch": 0.3583535108958838, + "grad_norm": 1.8464481779446322, + "learning_rate": 7.4292891853807585e-06, + "loss": 0.3734, + "step": 16280 + }, + { + "epoch": 0.35837552278230245, + "grad_norm": 1.4062792324165767, + "learning_rate": 7.4289776228404675e-06, + "loss": 0.2949, + "step": 16281 + }, + { + "epoch": 0.3583975346687211, + "grad_norm": 1.5362467538768956, + "learning_rate": 7.428666047955031e-06, + "loss": 0.2717, + "step": 16282 + }, + { + "epoch": 0.3584195465551398, + "grad_norm": 1.113850915374847, + "learning_rate": 7.428354460726029e-06, + "loss": 0.3127, + "step": 16283 + }, + { + "epoch": 0.35844155844155845, + "grad_norm": 1.7595094130160862, + "learning_rate": 7.428042861155047e-06, + "loss": 0.3596, + "step": 16284 + }, + { + "epoch": 0.3584635703279771, + "grad_norm": 0.6813330895125778, + "learning_rate": 7.427731249243668e-06, + "loss": 0.4918, + "step": 16285 + }, + { + "epoch": 0.35848558221439575, + "grad_norm": 0.6620233493815325, + "learning_rate": 7.427419624993476e-06, + "loss": 0.4922, + "step": 16286 + }, + { + "epoch": 0.35850759410081445, + "grad_norm": 1.8968499821126128, + "learning_rate": 7.427107988406055e-06, + "loss": 0.2328, + "step": 16287 + }, + { + "epoch": 0.3585296059872331, + "grad_norm": 1.392571610787435, + "learning_rate": 7.426796339482989e-06, + "loss": 0.3609, + "step": 16288 + }, + { + "epoch": 0.35855161787365175, + "grad_norm": 0.6223675022830972, + "learning_rate": 7.426484678225862e-06, + "loss": 0.4922, + "step": 16289 + }, + { + "epoch": 0.35857362976007046, + "grad_norm": 3.2591284467755504, + "learning_rate": 7.4261730046362565e-06, + "loss": 0.2769, + "step": 16290 + }, + { + "epoch": 0.3585956416464891, + "grad_norm": 3.1551774659511684, + "learning_rate": 7.425861318715757e-06, + "loss": 0.2888, + "step": 16291 + }, + { + "epoch": 0.35861765353290775, + "grad_norm": 1.2622424365666296, + "learning_rate": 7.425549620465951e-06, + "loss": 0.2861, + "step": 16292 + }, + { + "epoch": 0.35863966541932646, + "grad_norm": 1.4942629527617421, + "learning_rate": 7.425237909888417e-06, + "loss": 0.296, + "step": 16293 + }, + { + "epoch": 0.3586616773057451, + "grad_norm": 5.570901175678799, + "learning_rate": 7.424926186984744e-06, + "loss": 0.4192, + "step": 16294 + }, + { + "epoch": 0.35868368919216376, + "grad_norm": 1.574554614779264, + "learning_rate": 7.424614451756516e-06, + "loss": 0.3656, + "step": 16295 + }, + { + "epoch": 0.35870570107858246, + "grad_norm": 2.215372806287489, + "learning_rate": 7.424302704205313e-06, + "loss": 0.2903, + "step": 16296 + }, + { + "epoch": 0.3587277129650011, + "grad_norm": 1.534817850831299, + "learning_rate": 7.4239909443327245e-06, + "loss": 0.3042, + "step": 16297 + }, + { + "epoch": 0.35874972485141976, + "grad_norm": 0.7939956272229334, + "learning_rate": 7.423679172140331e-06, + "loss": 0.5294, + "step": 16298 + }, + { + "epoch": 0.3587717367378384, + "grad_norm": 2.125075061504867, + "learning_rate": 7.423367387629719e-06, + "loss": 0.277, + "step": 16299 + }, + { + "epoch": 0.3587937486242571, + "grad_norm": 2.37096799546527, + "learning_rate": 7.423055590802475e-06, + "loss": 0.3978, + "step": 16300 + }, + { + "epoch": 0.35881576051067576, + "grad_norm": 2.31747692067814, + "learning_rate": 7.42274378166018e-06, + "loss": 0.3865, + "step": 16301 + }, + { + "epoch": 0.3588377723970944, + "grad_norm": 1.396412593281249, + "learning_rate": 7.422431960204422e-06, + "loss": 0.3338, + "step": 16302 + }, + { + "epoch": 0.3588597842835131, + "grad_norm": 1.338062527291137, + "learning_rate": 7.422120126436784e-06, + "loss": 0.2686, + "step": 16303 + }, + { + "epoch": 0.35888179616993177, + "grad_norm": 0.6321419054197764, + "learning_rate": 7.421808280358851e-06, + "loss": 0.5105, + "step": 16304 + }, + { + "epoch": 0.3589038080563504, + "grad_norm": 1.6874327051187183, + "learning_rate": 7.421496421972207e-06, + "loss": 0.313, + "step": 16305 + }, + { + "epoch": 0.3589258199427691, + "grad_norm": 1.4167507297152557, + "learning_rate": 7.42118455127844e-06, + "loss": 0.3216, + "step": 16306 + }, + { + "epoch": 0.35894783182918777, + "grad_norm": 1.6494333899459268, + "learning_rate": 7.420872668279133e-06, + "loss": 0.3749, + "step": 16307 + }, + { + "epoch": 0.3589698437156064, + "grad_norm": 2.2059999049436874, + "learning_rate": 7.420560772975871e-06, + "loss": 0.4484, + "step": 16308 + }, + { + "epoch": 0.35899185560202507, + "grad_norm": 1.258678085173311, + "learning_rate": 7.420248865370238e-06, + "loss": 0.2946, + "step": 16309 + }, + { + "epoch": 0.35901386748844377, + "grad_norm": 1.8359294320533086, + "learning_rate": 7.419936945463823e-06, + "loss": 0.3347, + "step": 16310 + }, + { + "epoch": 0.3590358793748624, + "grad_norm": 0.648561474359502, + "learning_rate": 7.419625013258208e-06, + "loss": 0.5053, + "step": 16311 + }, + { + "epoch": 0.35905789126128107, + "grad_norm": 1.2113534533722417, + "learning_rate": 7.41931306875498e-06, + "loss": 0.3236, + "step": 16312 + }, + { + "epoch": 0.3590799031476998, + "grad_norm": 1.4846127832805625, + "learning_rate": 7.419001111955723e-06, + "loss": 0.302, + "step": 16313 + }, + { + "epoch": 0.3591019150341184, + "grad_norm": 0.6325964413213369, + "learning_rate": 7.418689142862024e-06, + "loss": 0.5052, + "step": 16314 + }, + { + "epoch": 0.3591239269205371, + "grad_norm": 1.686442508092066, + "learning_rate": 7.41837716147547e-06, + "loss": 0.2967, + "step": 16315 + }, + { + "epoch": 0.3591459388069558, + "grad_norm": 1.722380902503849, + "learning_rate": 7.418065167797642e-06, + "loss": 0.3258, + "step": 16316 + }, + { + "epoch": 0.3591679506933744, + "grad_norm": 1.8852721040771814, + "learning_rate": 7.417753161830129e-06, + "loss": 0.2492, + "step": 16317 + }, + { + "epoch": 0.3591899625797931, + "grad_norm": 0.6370695215611315, + "learning_rate": 7.417441143574516e-06, + "loss": 0.5033, + "step": 16318 + }, + { + "epoch": 0.3592119744662118, + "grad_norm": 1.5251279631936348, + "learning_rate": 7.4171291130323886e-06, + "loss": 0.2748, + "step": 16319 + }, + { + "epoch": 0.35923398635263043, + "grad_norm": 1.5769729403790265, + "learning_rate": 7.4168170702053345e-06, + "loss": 0.3704, + "step": 16320 + }, + { + "epoch": 0.3592559982390491, + "grad_norm": 1.9950363303724952, + "learning_rate": 7.416505015094937e-06, + "loss": 0.4316, + "step": 16321 + }, + { + "epoch": 0.35927801012546773, + "grad_norm": 5.212553792149605, + "learning_rate": 7.4161929477027826e-06, + "loss": 0.3821, + "step": 16322 + }, + { + "epoch": 0.35930002201188643, + "grad_norm": 1.5149575041020178, + "learning_rate": 7.415880868030459e-06, + "loss": 0.2612, + "step": 16323 + }, + { + "epoch": 0.3593220338983051, + "grad_norm": 1.7451149876842875, + "learning_rate": 7.415568776079552e-06, + "loss": 0.3543, + "step": 16324 + }, + { + "epoch": 0.35934404578472373, + "grad_norm": 0.6252410506990738, + "learning_rate": 7.415256671851646e-06, + "loss": 0.5081, + "step": 16325 + }, + { + "epoch": 0.35936605767114244, + "grad_norm": 1.3009639174312964, + "learning_rate": 7.414944555348329e-06, + "loss": 0.3741, + "step": 16326 + }, + { + "epoch": 0.3593880695575611, + "grad_norm": 1.2746752990239227, + "learning_rate": 7.414632426571185e-06, + "loss": 0.2781, + "step": 16327 + }, + { + "epoch": 0.35941008144397973, + "grad_norm": 5.72310438774963, + "learning_rate": 7.414320285521803e-06, + "loss": 0.3051, + "step": 16328 + }, + { + "epoch": 0.35943209333039844, + "grad_norm": 1.6553829123274892, + "learning_rate": 7.414008132201769e-06, + "loss": 0.4368, + "step": 16329 + }, + { + "epoch": 0.3594541052168171, + "grad_norm": 1.6539597372300794, + "learning_rate": 7.41369596661267e-06, + "loss": 0.3293, + "step": 16330 + }, + { + "epoch": 0.35947611710323574, + "grad_norm": 1.4389394699054612, + "learning_rate": 7.41338378875609e-06, + "loss": 0.2909, + "step": 16331 + }, + { + "epoch": 0.3594981289896544, + "grad_norm": 1.0822854254386403, + "learning_rate": 7.413071598633617e-06, + "loss": 0.3743, + "step": 16332 + }, + { + "epoch": 0.3595201408760731, + "grad_norm": 1.8145598183120528, + "learning_rate": 7.412759396246838e-06, + "loss": 0.3127, + "step": 16333 + }, + { + "epoch": 0.35954215276249174, + "grad_norm": 2.1790007703541003, + "learning_rate": 7.412447181597341e-06, + "loss": 0.2611, + "step": 16334 + }, + { + "epoch": 0.3595641646489104, + "grad_norm": 1.508727593094242, + "learning_rate": 7.412134954686709e-06, + "loss": 0.4004, + "step": 16335 + }, + { + "epoch": 0.3595861765353291, + "grad_norm": 1.60919152358216, + "learning_rate": 7.411822715516534e-06, + "loss": 0.345, + "step": 16336 + }, + { + "epoch": 0.35960818842174774, + "grad_norm": 20.02109268496599, + "learning_rate": 7.411510464088398e-06, + "loss": 0.362, + "step": 16337 + }, + { + "epoch": 0.3596302003081664, + "grad_norm": 1.2526188358563157, + "learning_rate": 7.411198200403892e-06, + "loss": 0.2977, + "step": 16338 + }, + { + "epoch": 0.3596522121945851, + "grad_norm": 1.4381128598146575, + "learning_rate": 7.410885924464602e-06, + "loss": 0.3323, + "step": 16339 + }, + { + "epoch": 0.35967422408100375, + "grad_norm": 1.2591959466385367, + "learning_rate": 7.410573636272111e-06, + "loss": 0.3031, + "step": 16340 + }, + { + "epoch": 0.3596962359674224, + "grad_norm": 2.0304058329054953, + "learning_rate": 7.410261335828012e-06, + "loss": 0.3451, + "step": 16341 + }, + { + "epoch": 0.3597182478538411, + "grad_norm": 0.6426633747235666, + "learning_rate": 7.40994902313389e-06, + "loss": 0.5045, + "step": 16342 + }, + { + "epoch": 0.35974025974025975, + "grad_norm": 1.5095955332935616, + "learning_rate": 7.409636698191331e-06, + "loss": 0.3608, + "step": 16343 + }, + { + "epoch": 0.3597622716266784, + "grad_norm": 1.431630963335186, + "learning_rate": 7.4093243610019246e-06, + "loss": 0.311, + "step": 16344 + }, + { + "epoch": 0.35978428351309705, + "grad_norm": 1.5839683242491869, + "learning_rate": 7.409012011567256e-06, + "loss": 0.2949, + "step": 16345 + }, + { + "epoch": 0.35980629539951575, + "grad_norm": 1.3202623738237755, + "learning_rate": 7.408699649888915e-06, + "loss": 0.2851, + "step": 16346 + }, + { + "epoch": 0.3598283072859344, + "grad_norm": 4.237201046449071, + "learning_rate": 7.408387275968488e-06, + "loss": 0.3824, + "step": 16347 + }, + { + "epoch": 0.35985031917235305, + "grad_norm": 1.5890068031930717, + "learning_rate": 7.408074889807564e-06, + "loss": 0.2703, + "step": 16348 + }, + { + "epoch": 0.35987233105877175, + "grad_norm": 0.6502488546559186, + "learning_rate": 7.407762491407727e-06, + "loss": 0.505, + "step": 16349 + }, + { + "epoch": 0.3598943429451904, + "grad_norm": 0.6468365616186486, + "learning_rate": 7.407450080770567e-06, + "loss": 0.4998, + "step": 16350 + }, + { + "epoch": 0.35991635483160905, + "grad_norm": 1.6700760085823625, + "learning_rate": 7.407137657897674e-06, + "loss": 0.3223, + "step": 16351 + }, + { + "epoch": 0.35993836671802776, + "grad_norm": 2.1457378772936897, + "learning_rate": 7.406825222790634e-06, + "loss": 0.35, + "step": 16352 + }, + { + "epoch": 0.3599603786044464, + "grad_norm": 0.6315488894508205, + "learning_rate": 7.406512775451033e-06, + "loss": 0.5173, + "step": 16353 + }, + { + "epoch": 0.35998239049086506, + "grad_norm": 1.6265368853304112, + "learning_rate": 7.406200315880461e-06, + "loss": 0.3312, + "step": 16354 + }, + { + "epoch": 0.36000440237728376, + "grad_norm": 1.7016883741598368, + "learning_rate": 7.405887844080506e-06, + "loss": 0.3561, + "step": 16355 + }, + { + "epoch": 0.3600264142637024, + "grad_norm": 1.5127349230046614, + "learning_rate": 7.405575360052757e-06, + "loss": 0.3549, + "step": 16356 + }, + { + "epoch": 0.36004842615012106, + "grad_norm": 1.7492871892699569, + "learning_rate": 7.4052628637988e-06, + "loss": 0.3805, + "step": 16357 + }, + { + "epoch": 0.3600704380365397, + "grad_norm": 0.6081817924013824, + "learning_rate": 7.404950355320224e-06, + "loss": 0.5085, + "step": 16358 + }, + { + "epoch": 0.3600924499229584, + "grad_norm": 1.5221298981359648, + "learning_rate": 7.404637834618618e-06, + "loss": 0.3467, + "step": 16359 + }, + { + "epoch": 0.36011446180937706, + "grad_norm": 1.3234484858635116, + "learning_rate": 7.4043253016955705e-06, + "loss": 0.3113, + "step": 16360 + }, + { + "epoch": 0.3601364736957957, + "grad_norm": 3.5683454115965225, + "learning_rate": 7.40401275655267e-06, + "loss": 0.318, + "step": 16361 + }, + { + "epoch": 0.3601584855822144, + "grad_norm": 2.8398727846563028, + "learning_rate": 7.4037001991915035e-06, + "loss": 0.4356, + "step": 16362 + }, + { + "epoch": 0.36018049746863307, + "grad_norm": 0.6310880836662816, + "learning_rate": 7.4033876296136585e-06, + "loss": 0.5177, + "step": 16363 + }, + { + "epoch": 0.3602025093550517, + "grad_norm": 2.1460534798868234, + "learning_rate": 7.403075047820729e-06, + "loss": 0.2164, + "step": 16364 + }, + { + "epoch": 0.3602245212414704, + "grad_norm": 1.4255692119256798, + "learning_rate": 7.402762453814298e-06, + "loss": 0.2766, + "step": 16365 + }, + { + "epoch": 0.36024653312788907, + "grad_norm": 1.8804005454454642, + "learning_rate": 7.4024498475959585e-06, + "loss": 0.3534, + "step": 16366 + }, + { + "epoch": 0.3602685450143077, + "grad_norm": 1.406505265815135, + "learning_rate": 7.402137229167297e-06, + "loss": 0.3573, + "step": 16367 + }, + { + "epoch": 0.36029055690072637, + "grad_norm": 3.7753126028577553, + "learning_rate": 7.401824598529901e-06, + "loss": 0.2491, + "step": 16368 + }, + { + "epoch": 0.36031256878714507, + "grad_norm": 1.395030014636714, + "learning_rate": 7.401511955685363e-06, + "loss": 0.2233, + "step": 16369 + }, + { + "epoch": 0.3603345806735637, + "grad_norm": 0.6466873787716915, + "learning_rate": 7.4011993006352704e-06, + "loss": 0.4996, + "step": 16370 + }, + { + "epoch": 0.36035659255998237, + "grad_norm": 2.1826720609387245, + "learning_rate": 7.40088663338121e-06, + "loss": 0.3939, + "step": 16371 + }, + { + "epoch": 0.3603786044464011, + "grad_norm": 1.4154283814666815, + "learning_rate": 7.400573953924774e-06, + "loss": 0.284, + "step": 16372 + }, + { + "epoch": 0.3604006163328197, + "grad_norm": 1.307702868861581, + "learning_rate": 7.400261262267549e-06, + "loss": 0.3338, + "step": 16373 + }, + { + "epoch": 0.3604226282192384, + "grad_norm": 1.5761526276792412, + "learning_rate": 7.399948558411129e-06, + "loss": 0.3355, + "step": 16374 + }, + { + "epoch": 0.3604446401056571, + "grad_norm": 1.2757569692079447, + "learning_rate": 7.399635842357098e-06, + "loss": 0.3553, + "step": 16375 + }, + { + "epoch": 0.3604666519920757, + "grad_norm": 1.391626760252696, + "learning_rate": 7.3993231141070466e-06, + "loss": 0.3045, + "step": 16376 + }, + { + "epoch": 0.3604886638784944, + "grad_norm": 1.2588810748361352, + "learning_rate": 7.399010373662566e-06, + "loss": 0.2813, + "step": 16377 + }, + { + "epoch": 0.3605106757649131, + "grad_norm": 1.3734276178765616, + "learning_rate": 7.398697621025245e-06, + "loss": 0.2989, + "step": 16378 + }, + { + "epoch": 0.36053268765133173, + "grad_norm": 1.480402153341525, + "learning_rate": 7.398384856196673e-06, + "loss": 0.3534, + "step": 16379 + }, + { + "epoch": 0.3605546995377504, + "grad_norm": 1.3417694428435571, + "learning_rate": 7.398072079178439e-06, + "loss": 0.3516, + "step": 16380 + }, + { + "epoch": 0.360576711424169, + "grad_norm": 1.2733678457453896, + "learning_rate": 7.397759289972132e-06, + "loss": 0.3943, + "step": 16381 + }, + { + "epoch": 0.36059872331058773, + "grad_norm": 1.3789386867517544, + "learning_rate": 7.397446488579344e-06, + "loss": 0.2644, + "step": 16382 + }, + { + "epoch": 0.3606207351970064, + "grad_norm": 1.5655507695155793, + "learning_rate": 7.397133675001665e-06, + "loss": 0.3347, + "step": 16383 + }, + { + "epoch": 0.36064274708342503, + "grad_norm": 1.6618644743451925, + "learning_rate": 7.396820849240682e-06, + "loss": 0.3331, + "step": 16384 + }, + { + "epoch": 0.36066475896984374, + "grad_norm": 1.184877045574725, + "learning_rate": 7.396508011297986e-06, + "loss": 0.2882, + "step": 16385 + }, + { + "epoch": 0.3606867708562624, + "grad_norm": 1.5900837975576088, + "learning_rate": 7.396195161175167e-06, + "loss": 0.3236, + "step": 16386 + }, + { + "epoch": 0.36070878274268103, + "grad_norm": 3.154013436785222, + "learning_rate": 7.395882298873817e-06, + "loss": 0.3684, + "step": 16387 + }, + { + "epoch": 0.36073079462909974, + "grad_norm": 1.6262011807862324, + "learning_rate": 7.395569424395525e-06, + "loss": 0.4591, + "step": 16388 + }, + { + "epoch": 0.3607528065155184, + "grad_norm": 1.4723209674335, + "learning_rate": 7.395256537741878e-06, + "loss": 0.3248, + "step": 16389 + }, + { + "epoch": 0.36077481840193704, + "grad_norm": 1.9938298896820945, + "learning_rate": 7.394943638914471e-06, + "loss": 0.4593, + "step": 16390 + }, + { + "epoch": 0.3607968302883557, + "grad_norm": 1.9119101414865094, + "learning_rate": 7.394630727914892e-06, + "loss": 0.3425, + "step": 16391 + }, + { + "epoch": 0.3608188421747744, + "grad_norm": 1.4902959727829495, + "learning_rate": 7.39431780474473e-06, + "loss": 0.2468, + "step": 16392 + }, + { + "epoch": 0.36084085406119304, + "grad_norm": 1.4756381259297033, + "learning_rate": 7.39400486940558e-06, + "loss": 0.3102, + "step": 16393 + }, + { + "epoch": 0.3608628659476117, + "grad_norm": 1.9476804954326448, + "learning_rate": 7.393691921899028e-06, + "loss": 0.3897, + "step": 16394 + }, + { + "epoch": 0.3608848778340304, + "grad_norm": 1.3622682370623542, + "learning_rate": 7.393378962226666e-06, + "loss": 0.3381, + "step": 16395 + }, + { + "epoch": 0.36090688972044904, + "grad_norm": 0.6720160699717692, + "learning_rate": 7.393065990390085e-06, + "loss": 0.5284, + "step": 16396 + }, + { + "epoch": 0.3609289016068677, + "grad_norm": 1.9815256825230005, + "learning_rate": 7.392753006390875e-06, + "loss": 0.3496, + "step": 16397 + }, + { + "epoch": 0.3609509134932864, + "grad_norm": 1.407255769787322, + "learning_rate": 7.392440010230627e-06, + "loss": 0.35, + "step": 16398 + }, + { + "epoch": 0.36097292537970505, + "grad_norm": 1.9809564313860117, + "learning_rate": 7.392127001910931e-06, + "loss": 0.3705, + "step": 16399 + }, + { + "epoch": 0.3609949372661237, + "grad_norm": 1.8583016940970338, + "learning_rate": 7.39181398143338e-06, + "loss": 0.3848, + "step": 16400 + }, + { + "epoch": 0.3610169491525424, + "grad_norm": 0.620308666031337, + "learning_rate": 7.391500948799563e-06, + "loss": 0.517, + "step": 16401 + }, + { + "epoch": 0.36103896103896105, + "grad_norm": 1.6832533500779887, + "learning_rate": 7.391187904011072e-06, + "loss": 0.3634, + "step": 16402 + }, + { + "epoch": 0.3610609729253797, + "grad_norm": 1.5059707002147342, + "learning_rate": 7.390874847069497e-06, + "loss": 0.3191, + "step": 16403 + }, + { + "epoch": 0.36108298481179835, + "grad_norm": 1.8350020920491565, + "learning_rate": 7.39056177797643e-06, + "loss": 0.3111, + "step": 16404 + }, + { + "epoch": 0.36110499669821705, + "grad_norm": 1.676859852736946, + "learning_rate": 7.390248696733461e-06, + "loss": 0.3381, + "step": 16405 + }, + { + "epoch": 0.3611270085846357, + "grad_norm": 1.4852048332869243, + "learning_rate": 7.389935603342183e-06, + "loss": 0.3461, + "step": 16406 + }, + { + "epoch": 0.36114902047105435, + "grad_norm": 1.3602361570183303, + "learning_rate": 7.389622497804185e-06, + "loss": 0.3217, + "step": 16407 + }, + { + "epoch": 0.36117103235747305, + "grad_norm": 3.114363630744484, + "learning_rate": 7.389309380121062e-06, + "loss": 0.3244, + "step": 16408 + }, + { + "epoch": 0.3611930442438917, + "grad_norm": 1.4578195399126632, + "learning_rate": 7.388996250294401e-06, + "loss": 0.2776, + "step": 16409 + }, + { + "epoch": 0.36121505613031035, + "grad_norm": 1.797209021154636, + "learning_rate": 7.388683108325796e-06, + "loss": 0.3446, + "step": 16410 + }, + { + "epoch": 0.36123706801672906, + "grad_norm": 0.62111456485102, + "learning_rate": 7.38836995421684e-06, + "loss": 0.4899, + "step": 16411 + }, + { + "epoch": 0.3612590799031477, + "grad_norm": 3.4316419556234647, + "learning_rate": 7.38805678796912e-06, + "loss": 0.2946, + "step": 16412 + }, + { + "epoch": 0.36128109178956636, + "grad_norm": 1.3401164813060298, + "learning_rate": 7.387743609584231e-06, + "loss": 0.3563, + "step": 16413 + }, + { + "epoch": 0.361303103675985, + "grad_norm": 0.6254661031624148, + "learning_rate": 7.387430419063765e-06, + "loss": 0.5111, + "step": 16414 + }, + { + "epoch": 0.3613251155624037, + "grad_norm": 1.7553300677214418, + "learning_rate": 7.387117216409312e-06, + "loss": 0.284, + "step": 16415 + }, + { + "epoch": 0.36134712744882236, + "grad_norm": 1.3610126579671633, + "learning_rate": 7.386804001622465e-06, + "loss": 0.376, + "step": 16416 + }, + { + "epoch": 0.361369139335241, + "grad_norm": 1.754244487677981, + "learning_rate": 7.3864907747048145e-06, + "loss": 0.3986, + "step": 16417 + }, + { + "epoch": 0.3613911512216597, + "grad_norm": 1.5409273426744545, + "learning_rate": 7.386177535657955e-06, + "loss": 0.3517, + "step": 16418 + }, + { + "epoch": 0.36141316310807836, + "grad_norm": 1.2557130890737442, + "learning_rate": 7.385864284483475e-06, + "loss": 0.3003, + "step": 16419 + }, + { + "epoch": 0.361435174994497, + "grad_norm": 0.6356747645759565, + "learning_rate": 7.38555102118297e-06, + "loss": 0.5116, + "step": 16420 + }, + { + "epoch": 0.3614571868809157, + "grad_norm": 1.752587247800203, + "learning_rate": 7.385237745758032e-06, + "loss": 0.416, + "step": 16421 + }, + { + "epoch": 0.36147919876733436, + "grad_norm": 2.4395456498055315, + "learning_rate": 7.384924458210249e-06, + "loss": 0.2887, + "step": 16422 + }, + { + "epoch": 0.361501210653753, + "grad_norm": 1.217619577996348, + "learning_rate": 7.384611158541217e-06, + "loss": 0.198, + "step": 16423 + }, + { + "epoch": 0.3615232225401717, + "grad_norm": 1.6749475836315983, + "learning_rate": 7.384297846752529e-06, + "loss": 0.3769, + "step": 16424 + }, + { + "epoch": 0.36154523442659037, + "grad_norm": 0.6008320548888489, + "learning_rate": 7.3839845228457754e-06, + "loss": 0.4897, + "step": 16425 + }, + { + "epoch": 0.361567246313009, + "grad_norm": 0.6291053461311684, + "learning_rate": 7.383671186822548e-06, + "loss": 0.503, + "step": 16426 + }, + { + "epoch": 0.36158925819942767, + "grad_norm": 1.6279722523675249, + "learning_rate": 7.383357838684442e-06, + "loss": 0.3292, + "step": 16427 + }, + { + "epoch": 0.36161127008584637, + "grad_norm": 1.1689365611650735, + "learning_rate": 7.3830444784330465e-06, + "loss": 0.3301, + "step": 16428 + }, + { + "epoch": 0.361633281972265, + "grad_norm": 1.7252317289721382, + "learning_rate": 7.3827311060699594e-06, + "loss": 0.2473, + "step": 16429 + }, + { + "epoch": 0.36165529385868367, + "grad_norm": 1.443907866320894, + "learning_rate": 7.382417721596766e-06, + "loss": 0.3497, + "step": 16430 + }, + { + "epoch": 0.3616773057451024, + "grad_norm": 1.6574385677225014, + "learning_rate": 7.382104325015066e-06, + "loss": 0.3833, + "step": 16431 + }, + { + "epoch": 0.361699317631521, + "grad_norm": 2.056876731028499, + "learning_rate": 7.381790916326449e-06, + "loss": 0.4537, + "step": 16432 + }, + { + "epoch": 0.36172132951793967, + "grad_norm": 2.7196707818749806, + "learning_rate": 7.381477495532507e-06, + "loss": 0.3918, + "step": 16433 + }, + { + "epoch": 0.3617433414043584, + "grad_norm": 1.7133855313715352, + "learning_rate": 7.381164062634836e-06, + "loss": 0.4202, + "step": 16434 + }, + { + "epoch": 0.361765353290777, + "grad_norm": 1.6441561032424128, + "learning_rate": 7.3808506176350255e-06, + "loss": 0.281, + "step": 16435 + }, + { + "epoch": 0.3617873651771957, + "grad_norm": 1.4470094794433155, + "learning_rate": 7.380537160534671e-06, + "loss": 0.3046, + "step": 16436 + }, + { + "epoch": 0.3618093770636144, + "grad_norm": 1.4398573296390929, + "learning_rate": 7.3802236913353655e-06, + "loss": 0.325, + "step": 16437 + }, + { + "epoch": 0.36183138895003303, + "grad_norm": 1.479039268591454, + "learning_rate": 7.379910210038701e-06, + "loss": 0.3189, + "step": 16438 + }, + { + "epoch": 0.3618534008364517, + "grad_norm": 1.3040251633399775, + "learning_rate": 7.379596716646271e-06, + "loss": 0.3197, + "step": 16439 + }, + { + "epoch": 0.3618754127228703, + "grad_norm": 1.3279504705060958, + "learning_rate": 7.379283211159669e-06, + "loss": 0.3336, + "step": 16440 + }, + { + "epoch": 0.36189742460928903, + "grad_norm": 1.671096884850996, + "learning_rate": 7.378969693580489e-06, + "loss": 0.3769, + "step": 16441 + }, + { + "epoch": 0.3619194364957077, + "grad_norm": 1.6650852162646956, + "learning_rate": 7.378656163910325e-06, + "loss": 0.3581, + "step": 16442 + }, + { + "epoch": 0.36194144838212633, + "grad_norm": 2.040032077201386, + "learning_rate": 7.3783426221507684e-06, + "loss": 0.3713, + "step": 16443 + }, + { + "epoch": 0.36196346026854503, + "grad_norm": 2.1037768117506737, + "learning_rate": 7.378029068303414e-06, + "loss": 0.3838, + "step": 16444 + }, + { + "epoch": 0.3619854721549637, + "grad_norm": 1.694315615412568, + "learning_rate": 7.3777155023698555e-06, + "loss": 0.3108, + "step": 16445 + }, + { + "epoch": 0.36200748404138233, + "grad_norm": 1.4223328615152375, + "learning_rate": 7.377401924351685e-06, + "loss": 0.3422, + "step": 16446 + }, + { + "epoch": 0.36202949592780104, + "grad_norm": 1.6985619234100842, + "learning_rate": 7.377088334250499e-06, + "loss": 0.3628, + "step": 16447 + }, + { + "epoch": 0.3620515078142197, + "grad_norm": 1.4225705751809496, + "learning_rate": 7.376774732067889e-06, + "loss": 0.3141, + "step": 16448 + }, + { + "epoch": 0.36207351970063834, + "grad_norm": 1.0665364519285903, + "learning_rate": 7.37646111780545e-06, + "loss": 0.2743, + "step": 16449 + }, + { + "epoch": 0.362095531587057, + "grad_norm": 0.7149746769298123, + "learning_rate": 7.376147491464777e-06, + "loss": 0.5001, + "step": 16450 + }, + { + "epoch": 0.3621175434734757, + "grad_norm": 2.4885734466888283, + "learning_rate": 7.375833853047461e-06, + "loss": 0.351, + "step": 16451 + }, + { + "epoch": 0.36213955535989434, + "grad_norm": 1.6654964806656298, + "learning_rate": 7.375520202555099e-06, + "loss": 0.3517, + "step": 16452 + }, + { + "epoch": 0.362161567246313, + "grad_norm": 2.3002211604551572, + "learning_rate": 7.375206539989281e-06, + "loss": 0.3094, + "step": 16453 + }, + { + "epoch": 0.3621835791327317, + "grad_norm": 1.2052827025007904, + "learning_rate": 7.374892865351607e-06, + "loss": 0.3048, + "step": 16454 + }, + { + "epoch": 0.36220559101915034, + "grad_norm": 1.3810415421289441, + "learning_rate": 7.374579178643667e-06, + "loss": 0.3175, + "step": 16455 + }, + { + "epoch": 0.362227602905569, + "grad_norm": 1.4594437960646096, + "learning_rate": 7.374265479867057e-06, + "loss": 0.318, + "step": 16456 + }, + { + "epoch": 0.3622496147919877, + "grad_norm": 1.4930733194217203, + "learning_rate": 7.373951769023371e-06, + "loss": 0.3654, + "step": 16457 + }, + { + "epoch": 0.36227162667840634, + "grad_norm": 1.6480572333880803, + "learning_rate": 7.373638046114202e-06, + "loss": 0.3021, + "step": 16458 + }, + { + "epoch": 0.362293638564825, + "grad_norm": 0.6488590833562822, + "learning_rate": 7.373324311141147e-06, + "loss": 0.5058, + "step": 16459 + }, + { + "epoch": 0.3623156504512437, + "grad_norm": 1.3274889953876985, + "learning_rate": 7.373010564105799e-06, + "loss": 0.3378, + "step": 16460 + }, + { + "epoch": 0.36233766233766235, + "grad_norm": 2.3682087495056563, + "learning_rate": 7.3726968050097535e-06, + "loss": 0.4245, + "step": 16461 + }, + { + "epoch": 0.362359674224081, + "grad_norm": 1.162948004336671, + "learning_rate": 7.372383033854604e-06, + "loss": 0.2792, + "step": 16462 + }, + { + "epoch": 0.36238168611049965, + "grad_norm": 1.2243129620449962, + "learning_rate": 7.372069250641945e-06, + "loss": 0.2909, + "step": 16463 + }, + { + "epoch": 0.36240369799691835, + "grad_norm": 1.7942071647300724, + "learning_rate": 7.371755455373373e-06, + "loss": 0.3592, + "step": 16464 + }, + { + "epoch": 0.362425709883337, + "grad_norm": 1.8114268383934238, + "learning_rate": 7.371441648050484e-06, + "loss": 0.3446, + "step": 16465 + }, + { + "epoch": 0.36244772176975565, + "grad_norm": 2.387765677388218, + "learning_rate": 7.371127828674867e-06, + "loss": 0.345, + "step": 16466 + }, + { + "epoch": 0.36246973365617435, + "grad_norm": 2.501794874058928, + "learning_rate": 7.370813997248123e-06, + "loss": 0.3433, + "step": 16467 + }, + { + "epoch": 0.362491745542593, + "grad_norm": 1.2226977550012652, + "learning_rate": 7.370500153771844e-06, + "loss": 0.2529, + "step": 16468 + }, + { + "epoch": 0.36251375742901165, + "grad_norm": 1.7126179114406288, + "learning_rate": 7.370186298247626e-06, + "loss": 0.336, + "step": 16469 + }, + { + "epoch": 0.36253576931543036, + "grad_norm": 8.783266593481185, + "learning_rate": 7.369872430677065e-06, + "loss": 0.3111, + "step": 16470 + }, + { + "epoch": 0.362557781201849, + "grad_norm": 13.384927938712801, + "learning_rate": 7.369558551061754e-06, + "loss": 0.3256, + "step": 16471 + }, + { + "epoch": 0.36257979308826765, + "grad_norm": 0.6029700689769942, + "learning_rate": 7.369244659403291e-06, + "loss": 0.5134, + "step": 16472 + }, + { + "epoch": 0.3626018049746863, + "grad_norm": 1.8899780985233239, + "learning_rate": 7.36893075570327e-06, + "loss": 0.3838, + "step": 16473 + }, + { + "epoch": 0.362623816861105, + "grad_norm": 3.1761092720463333, + "learning_rate": 7.3686168399632855e-06, + "loss": 0.2928, + "step": 16474 + }, + { + "epoch": 0.36264582874752366, + "grad_norm": 1.4515845358845252, + "learning_rate": 7.368302912184933e-06, + "loss": 0.2653, + "step": 16475 + }, + { + "epoch": 0.3626678406339423, + "grad_norm": 2.5650365906880666, + "learning_rate": 7.367988972369809e-06, + "loss": 0.362, + "step": 16476 + }, + { + "epoch": 0.362689852520361, + "grad_norm": 1.650988955830562, + "learning_rate": 7.36767502051951e-06, + "loss": 0.2717, + "step": 16477 + }, + { + "epoch": 0.36271186440677966, + "grad_norm": 1.6035511542820575, + "learning_rate": 7.36736105663563e-06, + "loss": 0.3389, + "step": 16478 + }, + { + "epoch": 0.3627338762931983, + "grad_norm": 1.5760248078912098, + "learning_rate": 7.367047080719766e-06, + "loss": 0.3907, + "step": 16479 + }, + { + "epoch": 0.362755888179617, + "grad_norm": 1.9525804844067713, + "learning_rate": 7.3667330927735125e-06, + "loss": 0.305, + "step": 16480 + }, + { + "epoch": 0.36277790006603566, + "grad_norm": 0.6295160087411135, + "learning_rate": 7.3664190927984655e-06, + "loss": 0.5121, + "step": 16481 + }, + { + "epoch": 0.3627999119524543, + "grad_norm": 1.884983933726689, + "learning_rate": 7.366105080796221e-06, + "loss": 0.3569, + "step": 16482 + }, + { + "epoch": 0.362821923838873, + "grad_norm": 3.1567229656664155, + "learning_rate": 7.365791056768375e-06, + "loss": 0.3124, + "step": 16483 + }, + { + "epoch": 0.36284393572529167, + "grad_norm": 2.052169729477187, + "learning_rate": 7.365477020716525e-06, + "loss": 0.3044, + "step": 16484 + }, + { + "epoch": 0.3628659476117103, + "grad_norm": 3.6105775915784246, + "learning_rate": 7.3651629726422654e-06, + "loss": 0.2461, + "step": 16485 + }, + { + "epoch": 0.36288795949812896, + "grad_norm": 1.5290672386351665, + "learning_rate": 7.364848912547192e-06, + "loss": 0.3473, + "step": 16486 + }, + { + "epoch": 0.36290997138454767, + "grad_norm": 1.7770567137770632, + "learning_rate": 7.364534840432902e-06, + "loss": 0.3685, + "step": 16487 + }, + { + "epoch": 0.3629319832709663, + "grad_norm": 1.4738722269341142, + "learning_rate": 7.364220756300992e-06, + "loss": 0.2772, + "step": 16488 + }, + { + "epoch": 0.36295399515738497, + "grad_norm": 3.2583712224760633, + "learning_rate": 7.363906660153056e-06, + "loss": 0.267, + "step": 16489 + }, + { + "epoch": 0.3629760070438037, + "grad_norm": 0.6448035932158801, + "learning_rate": 7.363592551990694e-06, + "loss": 0.5101, + "step": 16490 + }, + { + "epoch": 0.3629980189302223, + "grad_norm": 0.6279217651966382, + "learning_rate": 7.363278431815499e-06, + "loss": 0.5119, + "step": 16491 + }, + { + "epoch": 0.36302003081664097, + "grad_norm": 1.5098700404658503, + "learning_rate": 7.362964299629068e-06, + "loss": 0.3336, + "step": 16492 + }, + { + "epoch": 0.3630420427030597, + "grad_norm": 0.6102183648520485, + "learning_rate": 7.362650155433002e-06, + "loss": 0.5033, + "step": 16493 + }, + { + "epoch": 0.3630640545894783, + "grad_norm": 2.0180620707599277, + "learning_rate": 7.362335999228891e-06, + "loss": 0.3787, + "step": 16494 + }, + { + "epoch": 0.363086066475897, + "grad_norm": 1.2228905121694387, + "learning_rate": 7.362021831018336e-06, + "loss": 0.2918, + "step": 16495 + }, + { + "epoch": 0.3631080783623156, + "grad_norm": 1.8312110677928723, + "learning_rate": 7.361707650802932e-06, + "loss": 0.3278, + "step": 16496 + }, + { + "epoch": 0.3631300902487343, + "grad_norm": 2.401476076939704, + "learning_rate": 7.3613934585842775e-06, + "loss": 0.3695, + "step": 16497 + }, + { + "epoch": 0.363152102135153, + "grad_norm": 1.609664739924432, + "learning_rate": 7.3610792543639676e-06, + "loss": 0.2469, + "step": 16498 + }, + { + "epoch": 0.3631741140215716, + "grad_norm": 1.5567672244703257, + "learning_rate": 7.3607650381436e-06, + "loss": 0.3065, + "step": 16499 + }, + { + "epoch": 0.36319612590799033, + "grad_norm": 1.1961351142828376, + "learning_rate": 7.360450809924771e-06, + "loss": 0.2666, + "step": 16500 + }, + { + "epoch": 0.363218137794409, + "grad_norm": 2.061615766915055, + "learning_rate": 7.360136569709078e-06, + "loss": 0.3739, + "step": 16501 + }, + { + "epoch": 0.36324014968082763, + "grad_norm": 1.7109818216202326, + "learning_rate": 7.359822317498119e-06, + "loss": 0.3717, + "step": 16502 + }, + { + "epoch": 0.36326216156724633, + "grad_norm": 1.506548224829694, + "learning_rate": 7.35950805329349e-06, + "loss": 0.331, + "step": 16503 + }, + { + "epoch": 0.363284173453665, + "grad_norm": 18.424288018448642, + "learning_rate": 7.359193777096789e-06, + "loss": 0.3119, + "step": 16504 + }, + { + "epoch": 0.36330618534008363, + "grad_norm": 0.6840737745825716, + "learning_rate": 7.358879488909613e-06, + "loss": 0.5215, + "step": 16505 + }, + { + "epoch": 0.36332819722650234, + "grad_norm": 1.6567178032898988, + "learning_rate": 7.35856518873356e-06, + "loss": 0.3557, + "step": 16506 + }, + { + "epoch": 0.363350209112921, + "grad_norm": 1.9588464948652586, + "learning_rate": 7.3582508765702256e-06, + "loss": 0.4241, + "step": 16507 + }, + { + "epoch": 0.36337222099933963, + "grad_norm": 1.4283345091234467, + "learning_rate": 7.357936552421209e-06, + "loss": 0.2833, + "step": 16508 + }, + { + "epoch": 0.3633942328857583, + "grad_norm": 1.802051835080346, + "learning_rate": 7.357622216288107e-06, + "loss": 0.2642, + "step": 16509 + }, + { + "epoch": 0.363416244772177, + "grad_norm": 5.912982789379122, + "learning_rate": 7.357307868172516e-06, + "loss": 0.3753, + "step": 16510 + }, + { + "epoch": 0.36343825665859564, + "grad_norm": 3.347043842336044, + "learning_rate": 7.356993508076038e-06, + "loss": 0.4621, + "step": 16511 + }, + { + "epoch": 0.3634602685450143, + "grad_norm": 1.3731889866169613, + "learning_rate": 7.356679136000267e-06, + "loss": 0.4047, + "step": 16512 + }, + { + "epoch": 0.363482280431433, + "grad_norm": 1.3973926465631452, + "learning_rate": 7.356364751946799e-06, + "loss": 0.3843, + "step": 16513 + }, + { + "epoch": 0.36350429231785164, + "grad_norm": 3.0253030203498184, + "learning_rate": 7.356050355917237e-06, + "loss": 0.3946, + "step": 16514 + }, + { + "epoch": 0.3635263042042703, + "grad_norm": 2.2506557039673436, + "learning_rate": 7.355735947913176e-06, + "loss": 0.3265, + "step": 16515 + }, + { + "epoch": 0.363548316090689, + "grad_norm": 1.2276314109945237, + "learning_rate": 7.3554215279362126e-06, + "loss": 0.2609, + "step": 16516 + }, + { + "epoch": 0.36357032797710764, + "grad_norm": 1.2723777133987635, + "learning_rate": 7.355107095987948e-06, + "loss": 0.3012, + "step": 16517 + }, + { + "epoch": 0.3635923398635263, + "grad_norm": 1.6988061776437462, + "learning_rate": 7.354792652069977e-06, + "loss": 0.3224, + "step": 16518 + }, + { + "epoch": 0.363614351749945, + "grad_norm": 1.5302330990003996, + "learning_rate": 7.354478196183902e-06, + "loss": 0.3607, + "step": 16519 + }, + { + "epoch": 0.36363636363636365, + "grad_norm": 1.7085474719816693, + "learning_rate": 7.354163728331317e-06, + "loss": 0.2985, + "step": 16520 + }, + { + "epoch": 0.3636583755227823, + "grad_norm": 1.192454422015776, + "learning_rate": 7.353849248513823e-06, + "loss": 0.3036, + "step": 16521 + }, + { + "epoch": 0.36368038740920094, + "grad_norm": 1.314380680468848, + "learning_rate": 7.353534756733016e-06, + "loss": 0.3433, + "step": 16522 + }, + { + "epoch": 0.36370239929561965, + "grad_norm": 1.3828040614055794, + "learning_rate": 7.353220252990496e-06, + "loss": 0.2785, + "step": 16523 + }, + { + "epoch": 0.3637244111820383, + "grad_norm": 0.6673488735299975, + "learning_rate": 7.352905737287862e-06, + "loss": 0.5059, + "step": 16524 + }, + { + "epoch": 0.36374642306845695, + "grad_norm": 1.524476427739322, + "learning_rate": 7.352591209626711e-06, + "loss": 0.3112, + "step": 16525 + }, + { + "epoch": 0.36376843495487565, + "grad_norm": 1.810252226562956, + "learning_rate": 7.3522766700086425e-06, + "loss": 0.2779, + "step": 16526 + }, + { + "epoch": 0.3637904468412943, + "grad_norm": 1.591476269570239, + "learning_rate": 7.351962118435254e-06, + "loss": 0.292, + "step": 16527 + }, + { + "epoch": 0.36381245872771295, + "grad_norm": 3.6064364612768154, + "learning_rate": 7.351647554908144e-06, + "loss": 0.3087, + "step": 16528 + }, + { + "epoch": 0.36383447061413166, + "grad_norm": 1.515769669034593, + "learning_rate": 7.351332979428916e-06, + "loss": 0.3345, + "step": 16529 + }, + { + "epoch": 0.3638564825005503, + "grad_norm": 0.607476066767645, + "learning_rate": 7.351018391999161e-06, + "loss": 0.4932, + "step": 16530 + }, + { + "epoch": 0.36387849438696895, + "grad_norm": 1.7758262828578266, + "learning_rate": 7.350703792620484e-06, + "loss": 0.2806, + "step": 16531 + }, + { + "epoch": 0.3639005062733876, + "grad_norm": 1.5187414325088484, + "learning_rate": 7.350389181294482e-06, + "loss": 0.236, + "step": 16532 + }, + { + "epoch": 0.3639225181598063, + "grad_norm": 2.198620876355254, + "learning_rate": 7.3500745580227526e-06, + "loss": 0.4042, + "step": 16533 + }, + { + "epoch": 0.36394453004622496, + "grad_norm": 2.1483459017660236, + "learning_rate": 7.349759922806898e-06, + "loss": 0.3888, + "step": 16534 + }, + { + "epoch": 0.3639665419326436, + "grad_norm": 1.6241989439417155, + "learning_rate": 7.349445275648515e-06, + "loss": 0.3308, + "step": 16535 + }, + { + "epoch": 0.3639885538190623, + "grad_norm": 1.4660831335262448, + "learning_rate": 7.3491306165492015e-06, + "loss": 0.3554, + "step": 16536 + }, + { + "epoch": 0.36401056570548096, + "grad_norm": 0.6547978054778025, + "learning_rate": 7.3488159455105605e-06, + "loss": 0.4962, + "step": 16537 + }, + { + "epoch": 0.3640325775918996, + "grad_norm": 2.8678772839249778, + "learning_rate": 7.348501262534189e-06, + "loss": 0.2995, + "step": 16538 + }, + { + "epoch": 0.3640545894783183, + "grad_norm": 1.3645158937574937, + "learning_rate": 7.348186567621686e-06, + "loss": 0.391, + "step": 16539 + }, + { + "epoch": 0.36407660136473696, + "grad_norm": 0.5900933542502562, + "learning_rate": 7.347871860774651e-06, + "loss": 0.495, + "step": 16540 + }, + { + "epoch": 0.3640986132511556, + "grad_norm": 1.6462763890004886, + "learning_rate": 7.347557141994684e-06, + "loss": 0.2643, + "step": 16541 + }, + { + "epoch": 0.3641206251375743, + "grad_norm": 1.612487012129776, + "learning_rate": 7.347242411283386e-06, + "loss": 0.3366, + "step": 16542 + }, + { + "epoch": 0.36414263702399297, + "grad_norm": 1.7300260272783234, + "learning_rate": 7.346927668642354e-06, + "loss": 0.3493, + "step": 16543 + }, + { + "epoch": 0.3641646489104116, + "grad_norm": 1.5915220903199712, + "learning_rate": 7.34661291407319e-06, + "loss": 0.3756, + "step": 16544 + }, + { + "epoch": 0.36418666079683026, + "grad_norm": 1.5918052903834616, + "learning_rate": 7.3462981475774906e-06, + "loss": 0.3419, + "step": 16545 + }, + { + "epoch": 0.36420867268324897, + "grad_norm": 1.753463993033178, + "learning_rate": 7.345983369156859e-06, + "loss": 0.3046, + "step": 16546 + }, + { + "epoch": 0.3642306845696676, + "grad_norm": 1.4524120655958779, + "learning_rate": 7.345668578812894e-06, + "loss": 0.4218, + "step": 16547 + }, + { + "epoch": 0.36425269645608627, + "grad_norm": 1.9677456161393705, + "learning_rate": 7.3453537765471935e-06, + "loss": 0.388, + "step": 16548 + }, + { + "epoch": 0.36427470834250497, + "grad_norm": 0.6497291852593822, + "learning_rate": 7.345038962361361e-06, + "loss": 0.4666, + "step": 16549 + }, + { + "epoch": 0.3642967202289236, + "grad_norm": 1.3927086629772205, + "learning_rate": 7.344724136256993e-06, + "loss": 0.2954, + "step": 16550 + }, + { + "epoch": 0.36431873211534227, + "grad_norm": 0.6225178901824863, + "learning_rate": 7.34440929823569e-06, + "loss": 0.4882, + "step": 16551 + }, + { + "epoch": 0.364340744001761, + "grad_norm": 2.8400509219704064, + "learning_rate": 7.344094448299058e-06, + "loss": 0.4053, + "step": 16552 + }, + { + "epoch": 0.3643627558881796, + "grad_norm": 1.4678915688961256, + "learning_rate": 7.343779586448688e-06, + "loss": 0.3588, + "step": 16553 + }, + { + "epoch": 0.3643847677745983, + "grad_norm": 1.3857233772595583, + "learning_rate": 7.343464712686186e-06, + "loss": 0.3741, + "step": 16554 + }, + { + "epoch": 0.3644067796610169, + "grad_norm": 1.6452825496321977, + "learning_rate": 7.3431498270131515e-06, + "loss": 0.3194, + "step": 16555 + }, + { + "epoch": 0.3644287915474356, + "grad_norm": 0.6515974446157454, + "learning_rate": 7.342834929431184e-06, + "loss": 0.5219, + "step": 16556 + }, + { + "epoch": 0.3644508034338543, + "grad_norm": 1.3998346267563164, + "learning_rate": 7.342520019941886e-06, + "loss": 0.343, + "step": 16557 + }, + { + "epoch": 0.3644728153202729, + "grad_norm": 2.4883020029610687, + "learning_rate": 7.342205098546855e-06, + "loss": 0.275, + "step": 16558 + }, + { + "epoch": 0.36449482720669163, + "grad_norm": 1.4716839020684693, + "learning_rate": 7.3418901652476925e-06, + "loss": 0.2363, + "step": 16559 + }, + { + "epoch": 0.3645168390931103, + "grad_norm": 1.3277146172377738, + "learning_rate": 7.341575220046001e-06, + "loss": 0.3205, + "step": 16560 + }, + { + "epoch": 0.36453885097952893, + "grad_norm": 1.86120739559593, + "learning_rate": 7.341260262943379e-06, + "loss": 0.2876, + "step": 16561 + }, + { + "epoch": 0.36456086286594763, + "grad_norm": 0.6298862261738231, + "learning_rate": 7.340945293941428e-06, + "loss": 0.514, + "step": 16562 + }, + { + "epoch": 0.3645828747523663, + "grad_norm": 2.4942307954119087, + "learning_rate": 7.34063031304175e-06, + "loss": 0.3172, + "step": 16563 + }, + { + "epoch": 0.36460488663878493, + "grad_norm": 1.4030271280080073, + "learning_rate": 7.340315320245943e-06, + "loss": 0.3424, + "step": 16564 + }, + { + "epoch": 0.36462689852520364, + "grad_norm": 1.6002848185312868, + "learning_rate": 7.340000315555613e-06, + "loss": 0.3419, + "step": 16565 + }, + { + "epoch": 0.3646489104116223, + "grad_norm": 1.9844831110816876, + "learning_rate": 7.339685298972354e-06, + "loss": 0.3106, + "step": 16566 + }, + { + "epoch": 0.36467092229804093, + "grad_norm": 1.4300283606958932, + "learning_rate": 7.339370270497773e-06, + "loss": 0.3363, + "step": 16567 + }, + { + "epoch": 0.3646929341844596, + "grad_norm": 1.927794513488813, + "learning_rate": 7.33905523013347e-06, + "loss": 0.3497, + "step": 16568 + }, + { + "epoch": 0.3647149460708783, + "grad_norm": 1.244374200367629, + "learning_rate": 7.338740177881042e-06, + "loss": 0.3048, + "step": 16569 + }, + { + "epoch": 0.36473695795729694, + "grad_norm": 1.604982353391935, + "learning_rate": 7.3384251137420956e-06, + "loss": 0.2963, + "step": 16570 + }, + { + "epoch": 0.3647589698437156, + "grad_norm": 1.175565707435424, + "learning_rate": 7.338110037718229e-06, + "loss": 0.278, + "step": 16571 + }, + { + "epoch": 0.3647809817301343, + "grad_norm": 0.6586763976062461, + "learning_rate": 7.337794949811044e-06, + "loss": 0.4771, + "step": 16572 + }, + { + "epoch": 0.36480299361655294, + "grad_norm": 1.38806367887622, + "learning_rate": 7.337479850022142e-06, + "loss": 0.2963, + "step": 16573 + }, + { + "epoch": 0.3648250055029716, + "grad_norm": 1.2430241184887127, + "learning_rate": 7.337164738353127e-06, + "loss": 0.3326, + "step": 16574 + }, + { + "epoch": 0.3648470173893903, + "grad_norm": 1.575590053289781, + "learning_rate": 7.336849614805596e-06, + "loss": 0.3707, + "step": 16575 + }, + { + "epoch": 0.36486902927580894, + "grad_norm": 1.6788204720740707, + "learning_rate": 7.336534479381154e-06, + "loss": 0.3179, + "step": 16576 + }, + { + "epoch": 0.3648910411622276, + "grad_norm": 0.6151618933810068, + "learning_rate": 7.3362193320814e-06, + "loss": 0.4894, + "step": 16577 + }, + { + "epoch": 0.3649130530486463, + "grad_norm": 1.4457954844414798, + "learning_rate": 7.3359041729079395e-06, + "loss": 0.27, + "step": 16578 + }, + { + "epoch": 0.36493506493506495, + "grad_norm": 1.767195519859603, + "learning_rate": 7.335589001862372e-06, + "loss": 0.3302, + "step": 16579 + }, + { + "epoch": 0.3649570768214836, + "grad_norm": 1.816828751502054, + "learning_rate": 7.335273818946299e-06, + "loss": 0.3529, + "step": 16580 + }, + { + "epoch": 0.36497908870790224, + "grad_norm": 2.407259046870729, + "learning_rate": 7.334958624161323e-06, + "loss": 0.2978, + "step": 16581 + }, + { + "epoch": 0.36500110059432095, + "grad_norm": 1.963378115914016, + "learning_rate": 7.334643417509045e-06, + "loss": 0.4581, + "step": 16582 + }, + { + "epoch": 0.3650231124807396, + "grad_norm": 1.9722360805451782, + "learning_rate": 7.3343281989910694e-06, + "loss": 0.3399, + "step": 16583 + }, + { + "epoch": 0.36504512436715825, + "grad_norm": 0.6387115877416362, + "learning_rate": 7.334012968608996e-06, + "loss": 0.4927, + "step": 16584 + }, + { + "epoch": 0.36506713625357695, + "grad_norm": 1.2790276817646236, + "learning_rate": 7.3336977263644285e-06, + "loss": 0.249, + "step": 16585 + }, + { + "epoch": 0.3650891481399956, + "grad_norm": 0.5951353663813349, + "learning_rate": 7.3333824722589675e-06, + "loss": 0.4847, + "step": 16586 + }, + { + "epoch": 0.36511116002641425, + "grad_norm": 2.9735867774249773, + "learning_rate": 7.333067206294215e-06, + "loss": 0.2963, + "step": 16587 + }, + { + "epoch": 0.36513317191283295, + "grad_norm": 0.6254571569439434, + "learning_rate": 7.3327519284717775e-06, + "loss": 0.5198, + "step": 16588 + }, + { + "epoch": 0.3651551837992516, + "grad_norm": 1.8882835774504054, + "learning_rate": 7.332436638793252e-06, + "loss": 0.3641, + "step": 16589 + }, + { + "epoch": 0.36517719568567025, + "grad_norm": 1.3570878218354112, + "learning_rate": 7.332121337260243e-06, + "loss": 0.2412, + "step": 16590 + }, + { + "epoch": 0.3651992075720889, + "grad_norm": 0.6201971523976639, + "learning_rate": 7.3318060238743525e-06, + "loss": 0.5117, + "step": 16591 + }, + { + "epoch": 0.3652212194585076, + "grad_norm": 1.4401763547525106, + "learning_rate": 7.3314906986371856e-06, + "loss": 0.3207, + "step": 16592 + }, + { + "epoch": 0.36524323134492626, + "grad_norm": 1.2885398321176997, + "learning_rate": 7.331175361550342e-06, + "loss": 0.4135, + "step": 16593 + }, + { + "epoch": 0.3652652432313449, + "grad_norm": 0.6142685126269801, + "learning_rate": 7.330860012615426e-06, + "loss": 0.4847, + "step": 16594 + }, + { + "epoch": 0.3652872551177636, + "grad_norm": 0.6003415601764355, + "learning_rate": 7.330544651834039e-06, + "loss": 0.4834, + "step": 16595 + }, + { + "epoch": 0.36530926700418226, + "grad_norm": 0.5794205737753494, + "learning_rate": 7.330229279207785e-06, + "loss": 0.4851, + "step": 16596 + }, + { + "epoch": 0.3653312788906009, + "grad_norm": 6.161760709600856, + "learning_rate": 7.3299138947382685e-06, + "loss": 0.2663, + "step": 16597 + }, + { + "epoch": 0.3653532907770196, + "grad_norm": 0.614195136899076, + "learning_rate": 7.329598498427089e-06, + "loss": 0.4933, + "step": 16598 + }, + { + "epoch": 0.36537530266343826, + "grad_norm": 0.6085453031172643, + "learning_rate": 7.32928309027585e-06, + "loss": 0.4785, + "step": 16599 + }, + { + "epoch": 0.3653973145498569, + "grad_norm": 1.2560552193505867, + "learning_rate": 7.328967670286156e-06, + "loss": 0.3066, + "step": 16600 + }, + { + "epoch": 0.3654193264362756, + "grad_norm": 2.129018022894797, + "learning_rate": 7.328652238459609e-06, + "loss": 0.3372, + "step": 16601 + }, + { + "epoch": 0.36544133832269426, + "grad_norm": 1.208162480393562, + "learning_rate": 7.3283367947978145e-06, + "loss": 0.3108, + "step": 16602 + }, + { + "epoch": 0.3654633502091129, + "grad_norm": 1.6916968372097254, + "learning_rate": 7.3280213393023725e-06, + "loss": 0.3359, + "step": 16603 + }, + { + "epoch": 0.36548536209553156, + "grad_norm": 0.6293142896802546, + "learning_rate": 7.327705871974888e-06, + "loss": 0.512, + "step": 16604 + }, + { + "epoch": 0.36550737398195027, + "grad_norm": 1.8243945309155927, + "learning_rate": 7.327390392816964e-06, + "loss": 0.4405, + "step": 16605 + }, + { + "epoch": 0.3655293858683689, + "grad_norm": 2.1663018264903053, + "learning_rate": 7.327074901830204e-06, + "loss": 0.2556, + "step": 16606 + }, + { + "epoch": 0.36555139775478757, + "grad_norm": 2.6233171496048255, + "learning_rate": 7.326759399016212e-06, + "loss": 0.4433, + "step": 16607 + }, + { + "epoch": 0.36557340964120627, + "grad_norm": 1.3271638130723526, + "learning_rate": 7.32644388437659e-06, + "loss": 0.3062, + "step": 16608 + }, + { + "epoch": 0.3655954215276249, + "grad_norm": 1.3911800091260835, + "learning_rate": 7.3261283579129434e-06, + "loss": 0.3759, + "step": 16609 + }, + { + "epoch": 0.36561743341404357, + "grad_norm": 1.4074755601227023, + "learning_rate": 7.325812819626875e-06, + "loss": 0.266, + "step": 16610 + }, + { + "epoch": 0.3656394453004623, + "grad_norm": 0.6164375327179531, + "learning_rate": 7.325497269519987e-06, + "loss": 0.4987, + "step": 16611 + }, + { + "epoch": 0.3656614571868809, + "grad_norm": 2.5449310897789186, + "learning_rate": 7.3251817075938866e-06, + "loss": 0.3515, + "step": 16612 + }, + { + "epoch": 0.36568346907329957, + "grad_norm": 1.705585148624487, + "learning_rate": 7.324866133850174e-06, + "loss": 0.3309, + "step": 16613 + }, + { + "epoch": 0.3657054809597182, + "grad_norm": 2.0692301867507465, + "learning_rate": 7.324550548290457e-06, + "loss": 0.4039, + "step": 16614 + }, + { + "epoch": 0.3657274928461369, + "grad_norm": 5.612702469818249, + "learning_rate": 7.3242349509163355e-06, + "loss": 0.3645, + "step": 16615 + }, + { + "epoch": 0.3657495047325556, + "grad_norm": 2.214062963669387, + "learning_rate": 7.3239193417294165e-06, + "loss": 0.3081, + "step": 16616 + }, + { + "epoch": 0.3657715166189742, + "grad_norm": 1.7180952222822612, + "learning_rate": 7.323603720731302e-06, + "loss": 0.2858, + "step": 16617 + }, + { + "epoch": 0.36579352850539293, + "grad_norm": 1.5629966872853032, + "learning_rate": 7.323288087923597e-06, + "loss": 0.3146, + "step": 16618 + }, + { + "epoch": 0.3658155403918116, + "grad_norm": 2.448157768293156, + "learning_rate": 7.322972443307907e-06, + "loss": 0.3171, + "step": 16619 + }, + { + "epoch": 0.3658375522782302, + "grad_norm": 0.6018771075810254, + "learning_rate": 7.322656786885835e-06, + "loss": 0.4869, + "step": 16620 + }, + { + "epoch": 0.36585956416464893, + "grad_norm": 4.433835600803831, + "learning_rate": 7.322341118658984e-06, + "loss": 0.1691, + "step": 16621 + }, + { + "epoch": 0.3658815760510676, + "grad_norm": 0.61781753523034, + "learning_rate": 7.322025438628959e-06, + "loss": 0.4748, + "step": 16622 + }, + { + "epoch": 0.36590358793748623, + "grad_norm": 5.459106887392812, + "learning_rate": 7.321709746797366e-06, + "loss": 0.3028, + "step": 16623 + }, + { + "epoch": 0.36592559982390493, + "grad_norm": 3.885998581109844, + "learning_rate": 7.32139404316581e-06, + "loss": 0.2474, + "step": 16624 + }, + { + "epoch": 0.3659476117103236, + "grad_norm": 0.6286936685682872, + "learning_rate": 7.321078327735892e-06, + "loss": 0.4955, + "step": 16625 + }, + { + "epoch": 0.36596962359674223, + "grad_norm": 1.7870902080434325, + "learning_rate": 7.320762600509219e-06, + "loss": 0.3101, + "step": 16626 + }, + { + "epoch": 0.3659916354831609, + "grad_norm": 1.9171837349691319, + "learning_rate": 7.320446861487396e-06, + "loss": 0.327, + "step": 16627 + }, + { + "epoch": 0.3660136473695796, + "grad_norm": 2.704186681260993, + "learning_rate": 7.320131110672027e-06, + "loss": 0.3503, + "step": 16628 + }, + { + "epoch": 0.36603565925599824, + "grad_norm": 0.6557662246503978, + "learning_rate": 7.319815348064716e-06, + "loss": 0.5289, + "step": 16629 + }, + { + "epoch": 0.3660576711424169, + "grad_norm": 1.3099673168653516, + "learning_rate": 7.319499573667071e-06, + "loss": 0.304, + "step": 16630 + }, + { + "epoch": 0.3660796830288356, + "grad_norm": 0.6112911390948524, + "learning_rate": 7.319183787480691e-06, + "loss": 0.4728, + "step": 16631 + }, + { + "epoch": 0.36610169491525424, + "grad_norm": 1.338308388091568, + "learning_rate": 7.318867989507188e-06, + "loss": 0.3989, + "step": 16632 + }, + { + "epoch": 0.3661237068016729, + "grad_norm": 2.311585422791702, + "learning_rate": 7.318552179748162e-06, + "loss": 0.3976, + "step": 16633 + }, + { + "epoch": 0.3661457186880916, + "grad_norm": 1.3821920629797222, + "learning_rate": 7.3182363582052205e-06, + "loss": 0.3908, + "step": 16634 + }, + { + "epoch": 0.36616773057451024, + "grad_norm": 1.6737869800910339, + "learning_rate": 7.3179205248799675e-06, + "loss": 0.3552, + "step": 16635 + }, + { + "epoch": 0.3661897424609289, + "grad_norm": 1.5998846683061756, + "learning_rate": 7.3176046797740074e-06, + "loss": 0.3475, + "step": 16636 + }, + { + "epoch": 0.36621175434734754, + "grad_norm": 1.9177633995024075, + "learning_rate": 7.317288822888948e-06, + "loss": 0.3122, + "step": 16637 + }, + { + "epoch": 0.36623376623376624, + "grad_norm": 0.6454848669560241, + "learning_rate": 7.3169729542263945e-06, + "loss": 0.5144, + "step": 16638 + }, + { + "epoch": 0.3662557781201849, + "grad_norm": 1.7454202772169674, + "learning_rate": 7.316657073787949e-06, + "loss": 0.4436, + "step": 16639 + }, + { + "epoch": 0.36627779000660354, + "grad_norm": 1.407194099461648, + "learning_rate": 7.316341181575219e-06, + "loss": 0.3194, + "step": 16640 + }, + { + "epoch": 0.36629980189302225, + "grad_norm": 3.475541234109115, + "learning_rate": 7.316025277589811e-06, + "loss": 0.3405, + "step": 16641 + }, + { + "epoch": 0.3663218137794409, + "grad_norm": 2.2127675605369332, + "learning_rate": 7.3157093618333285e-06, + "loss": 0.3696, + "step": 16642 + }, + { + "epoch": 0.36634382566585955, + "grad_norm": 2.25199481321196, + "learning_rate": 7.315393434307381e-06, + "loss": 0.2652, + "step": 16643 + }, + { + "epoch": 0.36636583755227825, + "grad_norm": 1.324337865420245, + "learning_rate": 7.315077495013568e-06, + "loss": 0.2916, + "step": 16644 + }, + { + "epoch": 0.3663878494386969, + "grad_norm": 1.53284124850489, + "learning_rate": 7.314761543953499e-06, + "loss": 0.2612, + "step": 16645 + }, + { + "epoch": 0.36640986132511555, + "grad_norm": 2.16194799144283, + "learning_rate": 7.314445581128781e-06, + "loss": 0.3121, + "step": 16646 + }, + { + "epoch": 0.36643187321153425, + "grad_norm": 1.4186540991965653, + "learning_rate": 7.314129606541019e-06, + "loss": 0.3322, + "step": 16647 + }, + { + "epoch": 0.3664538850979529, + "grad_norm": 1.6784817165404602, + "learning_rate": 7.313813620191815e-06, + "loss": 0.237, + "step": 16648 + }, + { + "epoch": 0.36647589698437155, + "grad_norm": 0.6238871911341465, + "learning_rate": 7.313497622082779e-06, + "loss": 0.4853, + "step": 16649 + }, + { + "epoch": 0.3664979088707902, + "grad_norm": 1.3483724608136183, + "learning_rate": 7.3131816122155186e-06, + "loss": 0.3149, + "step": 16650 + }, + { + "epoch": 0.3665199207572089, + "grad_norm": 2.0550380624981397, + "learning_rate": 7.312865590591635e-06, + "loss": 0.4849, + "step": 16651 + }, + { + "epoch": 0.36654193264362755, + "grad_norm": 1.293938795149448, + "learning_rate": 7.312549557212739e-06, + "loss": 0.3515, + "step": 16652 + }, + { + "epoch": 0.3665639445300462, + "grad_norm": 3.2163697455171407, + "learning_rate": 7.312233512080433e-06, + "loss": 0.3406, + "step": 16653 + }, + { + "epoch": 0.3665859564164649, + "grad_norm": 1.4834446536136523, + "learning_rate": 7.311917455196326e-06, + "loss": 0.3418, + "step": 16654 + }, + { + "epoch": 0.36660796830288356, + "grad_norm": 1.3448166645316484, + "learning_rate": 7.311601386562023e-06, + "loss": 0.2559, + "step": 16655 + }, + { + "epoch": 0.3666299801893022, + "grad_norm": 0.5996724014648565, + "learning_rate": 7.311285306179132e-06, + "loss": 0.4749, + "step": 16656 + }, + { + "epoch": 0.3666519920757209, + "grad_norm": 1.544676216602698, + "learning_rate": 7.310969214049256e-06, + "loss": 0.3896, + "step": 16657 + }, + { + "epoch": 0.36667400396213956, + "grad_norm": 1.9886028451444313, + "learning_rate": 7.310653110174006e-06, + "loss": 0.2902, + "step": 16658 + }, + { + "epoch": 0.3666960158485582, + "grad_norm": 6.964424881975869, + "learning_rate": 7.310336994554984e-06, + "loss": 0.31, + "step": 16659 + }, + { + "epoch": 0.3667180277349769, + "grad_norm": 3.049438201023727, + "learning_rate": 7.3100208671938e-06, + "loss": 0.3254, + "step": 16660 + }, + { + "epoch": 0.36674003962139556, + "grad_norm": 1.6004619275585588, + "learning_rate": 7.3097047280920605e-06, + "loss": 0.3934, + "step": 16661 + }, + { + "epoch": 0.3667620515078142, + "grad_norm": 1.665642047128072, + "learning_rate": 7.309388577251369e-06, + "loss": 0.3231, + "step": 16662 + }, + { + "epoch": 0.36678406339423286, + "grad_norm": 1.7993804119646017, + "learning_rate": 7.309072414673337e-06, + "loss": 0.3404, + "step": 16663 + }, + { + "epoch": 0.36680607528065157, + "grad_norm": 2.4456659502635865, + "learning_rate": 7.308756240359569e-06, + "loss": 0.2934, + "step": 16664 + }, + { + "epoch": 0.3668280871670702, + "grad_norm": 1.641861464956522, + "learning_rate": 7.308440054311672e-06, + "loss": 0.3138, + "step": 16665 + }, + { + "epoch": 0.36685009905348887, + "grad_norm": 1.4957328033714112, + "learning_rate": 7.308123856531253e-06, + "loss": 0.3443, + "step": 16666 + }, + { + "epoch": 0.36687211093990757, + "grad_norm": 2.117895291117046, + "learning_rate": 7.307807647019918e-06, + "loss": 0.342, + "step": 16667 + }, + { + "epoch": 0.3668941228263262, + "grad_norm": 1.3185922923173938, + "learning_rate": 7.307491425779275e-06, + "loss": 0.3428, + "step": 16668 + }, + { + "epoch": 0.36691613471274487, + "grad_norm": 1.2434249287601296, + "learning_rate": 7.307175192810932e-06, + "loss": 0.2709, + "step": 16669 + }, + { + "epoch": 0.3669381465991636, + "grad_norm": 1.6083478114647327, + "learning_rate": 7.3068589481164976e-06, + "loss": 0.3515, + "step": 16670 + }, + { + "epoch": 0.3669601584855822, + "grad_norm": 1.637956090425358, + "learning_rate": 7.306542691697576e-06, + "loss": 0.3424, + "step": 16671 + }, + { + "epoch": 0.36698217037200087, + "grad_norm": 1.683120777163302, + "learning_rate": 7.306226423555773e-06, + "loss": 0.3563, + "step": 16672 + }, + { + "epoch": 0.3670041822584195, + "grad_norm": 1.5250697894768959, + "learning_rate": 7.305910143692702e-06, + "loss": 0.2956, + "step": 16673 + }, + { + "epoch": 0.3670261941448382, + "grad_norm": 1.368583648854662, + "learning_rate": 7.305593852109967e-06, + "loss": 0.2564, + "step": 16674 + }, + { + "epoch": 0.3670482060312569, + "grad_norm": 1.4388569712796824, + "learning_rate": 7.305277548809174e-06, + "loss": 0.3147, + "step": 16675 + }, + { + "epoch": 0.3670702179176755, + "grad_norm": 2.1119895538747713, + "learning_rate": 7.304961233791933e-06, + "loss": 0.3492, + "step": 16676 + }, + { + "epoch": 0.36709222980409423, + "grad_norm": 1.7863260401882288, + "learning_rate": 7.304644907059851e-06, + "loss": 0.3454, + "step": 16677 + }, + { + "epoch": 0.3671142416905129, + "grad_norm": 1.6068141994441567, + "learning_rate": 7.304328568614535e-06, + "loss": 0.2934, + "step": 16678 + }, + { + "epoch": 0.3671362535769315, + "grad_norm": 0.6546456580542074, + "learning_rate": 7.304012218457596e-06, + "loss": 0.4878, + "step": 16679 + }, + { + "epoch": 0.36715826546335023, + "grad_norm": 1.6849013105629076, + "learning_rate": 7.3036958565906355e-06, + "loss": 0.3575, + "step": 16680 + }, + { + "epoch": 0.3671802773497689, + "grad_norm": 1.9957982204071336, + "learning_rate": 7.303379483015268e-06, + "loss": 0.3348, + "step": 16681 + }, + { + "epoch": 0.36720228923618753, + "grad_norm": 1.6396368045113583, + "learning_rate": 7.303063097733098e-06, + "loss": 0.3446, + "step": 16682 + }, + { + "epoch": 0.36722430112260623, + "grad_norm": 1.5670433150166998, + "learning_rate": 7.302746700745733e-06, + "loss": 0.4362, + "step": 16683 + }, + { + "epoch": 0.3672463130090249, + "grad_norm": 1.7739992295315363, + "learning_rate": 7.302430292054785e-06, + "loss": 0.2596, + "step": 16684 + }, + { + "epoch": 0.36726832489544353, + "grad_norm": 1.6067662740413442, + "learning_rate": 7.302113871661855e-06, + "loss": 0.3276, + "step": 16685 + }, + { + "epoch": 0.3672903367818622, + "grad_norm": 1.8284710800097381, + "learning_rate": 7.30179743956856e-06, + "loss": 0.3717, + "step": 16686 + }, + { + "epoch": 0.3673123486682809, + "grad_norm": 1.5033148777117318, + "learning_rate": 7.301480995776502e-06, + "loss": 0.4208, + "step": 16687 + }, + { + "epoch": 0.36733436055469953, + "grad_norm": 1.4600757496357437, + "learning_rate": 7.301164540287291e-06, + "loss": 0.2948, + "step": 16688 + }, + { + "epoch": 0.3673563724411182, + "grad_norm": 1.6334731551921144, + "learning_rate": 7.300848073102536e-06, + "loss": 0.3303, + "step": 16689 + }, + { + "epoch": 0.3673783843275369, + "grad_norm": 2.9255069711052717, + "learning_rate": 7.300531594223843e-06, + "loss": 0.4713, + "step": 16690 + }, + { + "epoch": 0.36740039621395554, + "grad_norm": 1.3162993582523186, + "learning_rate": 7.300215103652824e-06, + "loss": 0.2323, + "step": 16691 + }, + { + "epoch": 0.3674224081003742, + "grad_norm": 1.5237068965732679, + "learning_rate": 7.299898601391086e-06, + "loss": 0.3946, + "step": 16692 + }, + { + "epoch": 0.3674444199867929, + "grad_norm": 1.0642574632138064, + "learning_rate": 7.299582087440236e-06, + "loss": 0.2444, + "step": 16693 + }, + { + "epoch": 0.36746643187321154, + "grad_norm": 1.3210651274541343, + "learning_rate": 7.299265561801886e-06, + "loss": 0.3129, + "step": 16694 + }, + { + "epoch": 0.3674884437596302, + "grad_norm": 1.5211784105980515, + "learning_rate": 7.2989490244776415e-06, + "loss": 0.3825, + "step": 16695 + }, + { + "epoch": 0.36751045564604884, + "grad_norm": 1.376842813935256, + "learning_rate": 7.298632475469113e-06, + "loss": 0.3628, + "step": 16696 + }, + { + "epoch": 0.36753246753246754, + "grad_norm": 1.754912252954385, + "learning_rate": 7.298315914777912e-06, + "loss": 0.3203, + "step": 16697 + }, + { + "epoch": 0.3675544794188862, + "grad_norm": 1.4043643122142964, + "learning_rate": 7.29799934240564e-06, + "loss": 0.2852, + "step": 16698 + }, + { + "epoch": 0.36757649130530484, + "grad_norm": 1.3103993772110964, + "learning_rate": 7.297682758353912e-06, + "loss": 0.3146, + "step": 16699 + }, + { + "epoch": 0.36759850319172355, + "grad_norm": 1.8690598543010797, + "learning_rate": 7.297366162624337e-06, + "loss": 0.357, + "step": 16700 + }, + { + "epoch": 0.3676205150781422, + "grad_norm": 2.07624379746563, + "learning_rate": 7.297049555218519e-06, + "loss": 0.2868, + "step": 16701 + }, + { + "epoch": 0.36764252696456085, + "grad_norm": 1.235434847396477, + "learning_rate": 7.296732936138075e-06, + "loss": 0.2492, + "step": 16702 + }, + { + "epoch": 0.36766453885097955, + "grad_norm": 2.619335857133041, + "learning_rate": 7.296416305384606e-06, + "loss": 0.302, + "step": 16703 + }, + { + "epoch": 0.3676865507373982, + "grad_norm": 0.7177436309247626, + "learning_rate": 7.296099662959728e-06, + "loss": 0.5116, + "step": 16704 + }, + { + "epoch": 0.36770856262381685, + "grad_norm": 0.6521735827606978, + "learning_rate": 7.295783008865046e-06, + "loss": 0.5007, + "step": 16705 + }, + { + "epoch": 0.36773057451023555, + "grad_norm": 1.616381308945004, + "learning_rate": 7.295466343102171e-06, + "loss": 0.3357, + "step": 16706 + }, + { + "epoch": 0.3677525863966542, + "grad_norm": 0.6268600622789591, + "learning_rate": 7.295149665672713e-06, + "loss": 0.5033, + "step": 16707 + }, + { + "epoch": 0.36777459828307285, + "grad_norm": 1.4998664839720597, + "learning_rate": 7.294832976578279e-06, + "loss": 0.2647, + "step": 16708 + }, + { + "epoch": 0.3677966101694915, + "grad_norm": 1.7486199169625896, + "learning_rate": 7.294516275820481e-06, + "loss": 0.2751, + "step": 16709 + }, + { + "epoch": 0.3678186220559102, + "grad_norm": 1.379980958585232, + "learning_rate": 7.294199563400928e-06, + "loss": 0.3862, + "step": 16710 + }, + { + "epoch": 0.36784063394232885, + "grad_norm": 4.726353786223502, + "learning_rate": 7.29388283932123e-06, + "loss": 0.3097, + "step": 16711 + }, + { + "epoch": 0.3678626458287475, + "grad_norm": 2.4593318792050103, + "learning_rate": 7.293566103582996e-06, + "loss": 0.3315, + "step": 16712 + }, + { + "epoch": 0.3678846577151662, + "grad_norm": 1.6778821737921843, + "learning_rate": 7.293249356187837e-06, + "loss": 0.3579, + "step": 16713 + }, + { + "epoch": 0.36790666960158486, + "grad_norm": 1.744125595713275, + "learning_rate": 7.292932597137361e-06, + "loss": 0.2879, + "step": 16714 + }, + { + "epoch": 0.3679286814880035, + "grad_norm": 5.297348813352164, + "learning_rate": 7.292615826433179e-06, + "loss": 0.3479, + "step": 16715 + }, + { + "epoch": 0.3679506933744222, + "grad_norm": 1.9986634828662784, + "learning_rate": 7.2922990440769e-06, + "loss": 0.343, + "step": 16716 + }, + { + "epoch": 0.36797270526084086, + "grad_norm": 8.611279886994467, + "learning_rate": 7.2919822500701365e-06, + "loss": 0.2924, + "step": 16717 + }, + { + "epoch": 0.3679947171472595, + "grad_norm": 1.6661282515624805, + "learning_rate": 7.291665444414496e-06, + "loss": 0.3378, + "step": 16718 + }, + { + "epoch": 0.3680167290336782, + "grad_norm": 1.2134768215226668, + "learning_rate": 7.291348627111589e-06, + "loss": 0.2487, + "step": 16719 + }, + { + "epoch": 0.36803874092009686, + "grad_norm": 0.8620524204552563, + "learning_rate": 7.291031798163029e-06, + "loss": 0.5056, + "step": 16720 + }, + { + "epoch": 0.3680607528065155, + "grad_norm": 1.3969576803445511, + "learning_rate": 7.29071495757042e-06, + "loss": 0.3057, + "step": 16721 + }, + { + "epoch": 0.36808276469293416, + "grad_norm": 2.0648828756423336, + "learning_rate": 7.2903981053353775e-06, + "loss": 0.3871, + "step": 16722 + }, + { + "epoch": 0.36810477657935287, + "grad_norm": 1.3911164715993387, + "learning_rate": 7.290081241459511e-06, + "loss": 0.272, + "step": 16723 + }, + { + "epoch": 0.3681267884657715, + "grad_norm": 1.640761435229192, + "learning_rate": 7.289764365944429e-06, + "loss": 0.3474, + "step": 16724 + }, + { + "epoch": 0.36814880035219016, + "grad_norm": 1.52739181164404, + "learning_rate": 7.289447478791744e-06, + "loss": 0.3631, + "step": 16725 + }, + { + "epoch": 0.36817081223860887, + "grad_norm": 1.7152323173332877, + "learning_rate": 7.289130580003065e-06, + "loss": 0.339, + "step": 16726 + }, + { + "epoch": 0.3681928241250275, + "grad_norm": 1.3370497173111973, + "learning_rate": 7.288813669580004e-06, + "loss": 0.3269, + "step": 16727 + }, + { + "epoch": 0.36821483601144617, + "grad_norm": 0.6939268815001376, + "learning_rate": 7.288496747524172e-06, + "loss": 0.525, + "step": 16728 + }, + { + "epoch": 0.36823684789786487, + "grad_norm": 1.969782327854833, + "learning_rate": 7.288179813837177e-06, + "loss": 0.4374, + "step": 16729 + }, + { + "epoch": 0.3682588597842835, + "grad_norm": 1.9022647255247302, + "learning_rate": 7.287862868520634e-06, + "loss": 0.306, + "step": 16730 + }, + { + "epoch": 0.36828087167070217, + "grad_norm": 1.8280845117212439, + "learning_rate": 7.28754591157615e-06, + "loss": 0.375, + "step": 16731 + }, + { + "epoch": 0.3683028835571208, + "grad_norm": 1.840972314820879, + "learning_rate": 7.287228943005335e-06, + "loss": 0.2801, + "step": 16732 + }, + { + "epoch": 0.3683248954435395, + "grad_norm": 0.6232540780889013, + "learning_rate": 7.286911962809806e-06, + "loss": 0.5001, + "step": 16733 + }, + { + "epoch": 0.3683469073299582, + "grad_norm": 2.008281576398433, + "learning_rate": 7.28659497099117e-06, + "loss": 0.2637, + "step": 16734 + }, + { + "epoch": 0.3683689192163768, + "grad_norm": 0.6237257682162458, + "learning_rate": 7.2862779675510375e-06, + "loss": 0.5082, + "step": 16735 + }, + { + "epoch": 0.3683909311027955, + "grad_norm": 1.5376601281940045, + "learning_rate": 7.285960952491022e-06, + "loss": 0.3231, + "step": 16736 + }, + { + "epoch": 0.3684129429892142, + "grad_norm": 1.6963345074678158, + "learning_rate": 7.285643925812731e-06, + "loss": 0.3979, + "step": 16737 + }, + { + "epoch": 0.3684349548756328, + "grad_norm": 1.280407167070609, + "learning_rate": 7.2853268875177796e-06, + "loss": 0.2763, + "step": 16738 + }, + { + "epoch": 0.36845696676205153, + "grad_norm": 1.484314717853336, + "learning_rate": 7.285009837607777e-06, + "loss": 0.2829, + "step": 16739 + }, + { + "epoch": 0.3684789786484702, + "grad_norm": 2.9339344574294204, + "learning_rate": 7.284692776084336e-06, + "loss": 0.3165, + "step": 16740 + }, + { + "epoch": 0.36850099053488883, + "grad_norm": 2.249086974267224, + "learning_rate": 7.284375702949067e-06, + "loss": 0.2739, + "step": 16741 + }, + { + "epoch": 0.36852300242130753, + "grad_norm": 1.9849428009748367, + "learning_rate": 7.284058618203581e-06, + "loss": 0.2557, + "step": 16742 + }, + { + "epoch": 0.3685450143077262, + "grad_norm": 1.6074623541663553, + "learning_rate": 7.283741521849492e-06, + "loss": 0.3707, + "step": 16743 + }, + { + "epoch": 0.36856702619414483, + "grad_norm": 1.3299938784413259, + "learning_rate": 7.283424413888409e-06, + "loss": 0.3244, + "step": 16744 + }, + { + "epoch": 0.3685890380805635, + "grad_norm": 1.4061830293331503, + "learning_rate": 7.283107294321945e-06, + "loss": 0.3655, + "step": 16745 + }, + { + "epoch": 0.3686110499669822, + "grad_norm": 1.3659896192164693, + "learning_rate": 7.282790163151712e-06, + "loss": 0.3441, + "step": 16746 + }, + { + "epoch": 0.36863306185340083, + "grad_norm": 0.7541185094404357, + "learning_rate": 7.28247302037932e-06, + "loss": 0.5194, + "step": 16747 + }, + { + "epoch": 0.3686550737398195, + "grad_norm": 0.6952040760114427, + "learning_rate": 7.282155866006384e-06, + "loss": 0.4927, + "step": 16748 + }, + { + "epoch": 0.3686770856262382, + "grad_norm": 0.6328138103400133, + "learning_rate": 7.281838700034513e-06, + "loss": 0.5103, + "step": 16749 + }, + { + "epoch": 0.36869909751265684, + "grad_norm": 1.6013246212511423, + "learning_rate": 7.281521522465318e-06, + "loss": 0.271, + "step": 16750 + }, + { + "epoch": 0.3687211093990755, + "grad_norm": 3.746780221056407, + "learning_rate": 7.281204333300416e-06, + "loss": 0.3032, + "step": 16751 + }, + { + "epoch": 0.3687431212854942, + "grad_norm": 1.6654173680560036, + "learning_rate": 7.2808871325414145e-06, + "loss": 0.2855, + "step": 16752 + }, + { + "epoch": 0.36876513317191284, + "grad_norm": 0.7021676931366804, + "learning_rate": 7.280569920189927e-06, + "loss": 0.4976, + "step": 16753 + }, + { + "epoch": 0.3687871450583315, + "grad_norm": 1.6269123430248704, + "learning_rate": 7.280252696247567e-06, + "loss": 0.4638, + "step": 16754 + }, + { + "epoch": 0.36880915694475014, + "grad_norm": 1.9921422243040088, + "learning_rate": 7.279935460715946e-06, + "loss": 0.2435, + "step": 16755 + }, + { + "epoch": 0.36883116883116884, + "grad_norm": 1.2716432519579564, + "learning_rate": 7.279618213596675e-06, + "loss": 0.3314, + "step": 16756 + }, + { + "epoch": 0.3688531807175875, + "grad_norm": 1.4892013396495487, + "learning_rate": 7.279300954891368e-06, + "loss": 0.2888, + "step": 16757 + }, + { + "epoch": 0.36887519260400614, + "grad_norm": 1.6778700667959894, + "learning_rate": 7.278983684601637e-06, + "loss": 0.3878, + "step": 16758 + }, + { + "epoch": 0.36889720449042485, + "grad_norm": 2.5414178779137275, + "learning_rate": 7.278666402729096e-06, + "loss": 0.4483, + "step": 16759 + }, + { + "epoch": 0.3689192163768435, + "grad_norm": 2.2604857384708423, + "learning_rate": 7.278349109275352e-06, + "loss": 0.3597, + "step": 16760 + }, + { + "epoch": 0.36894122826326214, + "grad_norm": 1.396225135286199, + "learning_rate": 7.278031804242026e-06, + "loss": 0.3333, + "step": 16761 + }, + { + "epoch": 0.36896324014968085, + "grad_norm": 1.8706575401177665, + "learning_rate": 7.277714487630724e-06, + "loss": 0.3044, + "step": 16762 + }, + { + "epoch": 0.3689852520360995, + "grad_norm": 0.7736513926350987, + "learning_rate": 7.277397159443061e-06, + "loss": 0.5049, + "step": 16763 + }, + { + "epoch": 0.36900726392251815, + "grad_norm": 0.6732755603464944, + "learning_rate": 7.277079819680651e-06, + "loss": 0.4821, + "step": 16764 + }, + { + "epoch": 0.36902927580893685, + "grad_norm": 1.942638925150015, + "learning_rate": 7.276762468345105e-06, + "loss": 0.386, + "step": 16765 + }, + { + "epoch": 0.3690512876953555, + "grad_norm": 1.866367869412572, + "learning_rate": 7.276445105438037e-06, + "loss": 0.3254, + "step": 16766 + }, + { + "epoch": 0.36907329958177415, + "grad_norm": 1.5798933824647101, + "learning_rate": 7.27612773096106e-06, + "loss": 0.3358, + "step": 16767 + }, + { + "epoch": 0.3690953114681928, + "grad_norm": 1.6725991476897064, + "learning_rate": 7.275810344915786e-06, + "loss": 0.396, + "step": 16768 + }, + { + "epoch": 0.3691173233546115, + "grad_norm": 1.6636827666590692, + "learning_rate": 7.275492947303828e-06, + "loss": 0.3678, + "step": 16769 + }, + { + "epoch": 0.36913933524103015, + "grad_norm": 0.7609159161523892, + "learning_rate": 7.275175538126802e-06, + "loss": 0.5208, + "step": 16770 + }, + { + "epoch": 0.3691613471274488, + "grad_norm": 1.6917194638467077, + "learning_rate": 7.274858117386319e-06, + "loss": 0.4038, + "step": 16771 + }, + { + "epoch": 0.3691833590138675, + "grad_norm": 1.398700843461752, + "learning_rate": 7.274540685083991e-06, + "loss": 0.353, + "step": 16772 + }, + { + "epoch": 0.36920537090028616, + "grad_norm": 1.3634438031880474, + "learning_rate": 7.274223241221432e-06, + "loss": 0.2664, + "step": 16773 + }, + { + "epoch": 0.3692273827867048, + "grad_norm": 1.8242819068322695, + "learning_rate": 7.273905785800259e-06, + "loss": 0.2865, + "step": 16774 + }, + { + "epoch": 0.3692493946731235, + "grad_norm": 1.4432076707000527, + "learning_rate": 7.2735883188220805e-06, + "loss": 0.2927, + "step": 16775 + }, + { + "epoch": 0.36927140655954216, + "grad_norm": 1.774966830942673, + "learning_rate": 7.273270840288513e-06, + "loss": 0.3757, + "step": 16776 + }, + { + "epoch": 0.3692934184459608, + "grad_norm": 1.555770821098745, + "learning_rate": 7.272953350201169e-06, + "loss": 0.2687, + "step": 16777 + }, + { + "epoch": 0.36931543033237946, + "grad_norm": 1.6088602840915145, + "learning_rate": 7.272635848561661e-06, + "loss": 0.3399, + "step": 16778 + }, + { + "epoch": 0.36933744221879816, + "grad_norm": 0.7057037754300559, + "learning_rate": 7.272318335371607e-06, + "loss": 0.4932, + "step": 16779 + }, + { + "epoch": 0.3693594541052168, + "grad_norm": 1.142601557343421, + "learning_rate": 7.272000810632614e-06, + "loss": 0.2607, + "step": 16780 + }, + { + "epoch": 0.36938146599163546, + "grad_norm": 1.6070253452065477, + "learning_rate": 7.271683274346303e-06, + "loss": 0.3315, + "step": 16781 + }, + { + "epoch": 0.36940347787805417, + "grad_norm": 2.0275707066385102, + "learning_rate": 7.271365726514282e-06, + "loss": 0.3583, + "step": 16782 + }, + { + "epoch": 0.3694254897644728, + "grad_norm": 1.1985248843239562, + "learning_rate": 7.271048167138168e-06, + "loss": 0.3037, + "step": 16783 + }, + { + "epoch": 0.36944750165089146, + "grad_norm": 1.4700804690302174, + "learning_rate": 7.270730596219575e-06, + "loss": 0.3507, + "step": 16784 + }, + { + "epoch": 0.36946951353731017, + "grad_norm": 2.4784922853831777, + "learning_rate": 7.270413013760116e-06, + "loss": 0.3666, + "step": 16785 + }, + { + "epoch": 0.3694915254237288, + "grad_norm": 0.6149065785579674, + "learning_rate": 7.270095419761405e-06, + "loss": 0.4739, + "step": 16786 + }, + { + "epoch": 0.36951353731014747, + "grad_norm": 1.7560070011057598, + "learning_rate": 7.269777814225056e-06, + "loss": 0.3316, + "step": 16787 + }, + { + "epoch": 0.36953554919656617, + "grad_norm": 0.6356852165456891, + "learning_rate": 7.269460197152686e-06, + "loss": 0.5036, + "step": 16788 + }, + { + "epoch": 0.3695575610829848, + "grad_norm": 1.513847430207826, + "learning_rate": 7.269142568545906e-06, + "loss": 0.4047, + "step": 16789 + }, + { + "epoch": 0.36957957296940347, + "grad_norm": 1.9831198954222478, + "learning_rate": 7.268824928406331e-06, + "loss": 0.3193, + "step": 16790 + }, + { + "epoch": 0.3696015848558221, + "grad_norm": 1.9499132621390312, + "learning_rate": 7.268507276735575e-06, + "loss": 0.2955, + "step": 16791 + }, + { + "epoch": 0.3696235967422408, + "grad_norm": 0.6066835683049737, + "learning_rate": 7.268189613535255e-06, + "loss": 0.4907, + "step": 16792 + }, + { + "epoch": 0.36964560862865947, + "grad_norm": 3.421032795924978, + "learning_rate": 7.267871938806983e-06, + "loss": 0.3903, + "step": 16793 + }, + { + "epoch": 0.3696676205150781, + "grad_norm": 0.6207763414412246, + "learning_rate": 7.267554252552373e-06, + "loss": 0.502, + "step": 16794 + }, + { + "epoch": 0.3696896324014968, + "grad_norm": 1.3752453722101619, + "learning_rate": 7.267236554773043e-06, + "loss": 0.3156, + "step": 16795 + }, + { + "epoch": 0.3697116442879155, + "grad_norm": 1.4461192686237037, + "learning_rate": 7.266918845470604e-06, + "loss": 0.2578, + "step": 16796 + }, + { + "epoch": 0.3697336561743341, + "grad_norm": 1.3347634593438935, + "learning_rate": 7.266601124646675e-06, + "loss": 0.3331, + "step": 16797 + }, + { + "epoch": 0.36975566806075283, + "grad_norm": 1.9278253749118124, + "learning_rate": 7.266283392302864e-06, + "loss": 0.4062, + "step": 16798 + }, + { + "epoch": 0.3697776799471715, + "grad_norm": 2.1669997588936436, + "learning_rate": 7.265965648440794e-06, + "loss": 0.4468, + "step": 16799 + }, + { + "epoch": 0.3697996918335901, + "grad_norm": 1.33874017670019, + "learning_rate": 7.2656478930620735e-06, + "loss": 0.3378, + "step": 16800 + }, + { + "epoch": 0.36982170372000883, + "grad_norm": 1.7477854589470454, + "learning_rate": 7.26533012616832e-06, + "loss": 0.3365, + "step": 16801 + }, + { + "epoch": 0.3698437156064275, + "grad_norm": 3.0917700112932853, + "learning_rate": 7.26501234776115e-06, + "loss": 0.3702, + "step": 16802 + }, + { + "epoch": 0.36986572749284613, + "grad_norm": 1.3867042381883747, + "learning_rate": 7.264694557842176e-06, + "loss": 0.3274, + "step": 16803 + }, + { + "epoch": 0.3698877393792648, + "grad_norm": 1.6136599334659663, + "learning_rate": 7.264376756413014e-06, + "loss": 0.4584, + "step": 16804 + }, + { + "epoch": 0.3699097512656835, + "grad_norm": 2.5193040789473, + "learning_rate": 7.26405894347528e-06, + "loss": 0.3149, + "step": 16805 + }, + { + "epoch": 0.36993176315210213, + "grad_norm": 2.573448655520048, + "learning_rate": 7.263741119030587e-06, + "loss": 0.2954, + "step": 16806 + }, + { + "epoch": 0.3699537750385208, + "grad_norm": 4.526309525456897, + "learning_rate": 7.263423283080554e-06, + "loss": 0.3223, + "step": 16807 + }, + { + "epoch": 0.3699757869249395, + "grad_norm": 1.6095565353343677, + "learning_rate": 7.263105435626794e-06, + "loss": 0.4307, + "step": 16808 + }, + { + "epoch": 0.36999779881135814, + "grad_norm": 1.3878744828155665, + "learning_rate": 7.262787576670921e-06, + "loss": 0.2777, + "step": 16809 + }, + { + "epoch": 0.3700198106977768, + "grad_norm": 0.6675884387305226, + "learning_rate": 7.262469706214555e-06, + "loss": 0.4914, + "step": 16810 + }, + { + "epoch": 0.3700418225841955, + "grad_norm": 1.5591142993563116, + "learning_rate": 7.262151824259307e-06, + "loss": 0.3781, + "step": 16811 + }, + { + "epoch": 0.37006383447061414, + "grad_norm": 0.6312584063286232, + "learning_rate": 7.261833930806795e-06, + "loss": 0.5006, + "step": 16812 + }, + { + "epoch": 0.3700858463570328, + "grad_norm": 1.3294523518803125, + "learning_rate": 7.261516025858635e-06, + "loss": 0.2781, + "step": 16813 + }, + { + "epoch": 0.37010785824345144, + "grad_norm": 0.6252269772809383, + "learning_rate": 7.261198109416439e-06, + "loss": 0.4827, + "step": 16814 + }, + { + "epoch": 0.37012987012987014, + "grad_norm": 1.4475492207530356, + "learning_rate": 7.26088018148183e-06, + "loss": 0.2684, + "step": 16815 + }, + { + "epoch": 0.3701518820162888, + "grad_norm": 3.171911870046151, + "learning_rate": 7.260562242056416e-06, + "loss": 0.374, + "step": 16816 + }, + { + "epoch": 0.37017389390270744, + "grad_norm": 0.6305126540261912, + "learning_rate": 7.260244291141818e-06, + "loss": 0.5002, + "step": 16817 + }, + { + "epoch": 0.37019590578912615, + "grad_norm": 0.6124416669176815, + "learning_rate": 7.259926328739651e-06, + "loss": 0.4836, + "step": 16818 + }, + { + "epoch": 0.3702179176755448, + "grad_norm": 1.8510418622757157, + "learning_rate": 7.2596083548515285e-06, + "loss": 0.3273, + "step": 16819 + }, + { + "epoch": 0.37023992956196344, + "grad_norm": 1.3994087682393368, + "learning_rate": 7.259290369479071e-06, + "loss": 0.2922, + "step": 16820 + }, + { + "epoch": 0.37026194144838215, + "grad_norm": 1.6697843795943643, + "learning_rate": 7.25897237262389e-06, + "loss": 0.3848, + "step": 16821 + }, + { + "epoch": 0.3702839533348008, + "grad_norm": 1.480484851307251, + "learning_rate": 7.258654364287605e-06, + "loss": 0.2417, + "step": 16822 + }, + { + "epoch": 0.37030596522121945, + "grad_norm": 2.194509389584817, + "learning_rate": 7.2583363444718305e-06, + "loss": 0.3785, + "step": 16823 + }, + { + "epoch": 0.37032797710763815, + "grad_norm": 9.547235503013152, + "learning_rate": 7.258018313178183e-06, + "loss": 0.2668, + "step": 16824 + }, + { + "epoch": 0.3703499889940568, + "grad_norm": 15.377429864912461, + "learning_rate": 7.257700270408281e-06, + "loss": 0.2897, + "step": 16825 + }, + { + "epoch": 0.37037200088047545, + "grad_norm": 1.4961732185100984, + "learning_rate": 7.257382216163738e-06, + "loss": 0.3114, + "step": 16826 + }, + { + "epoch": 0.3703940127668941, + "grad_norm": 1.8868198341048588, + "learning_rate": 7.257064150446171e-06, + "loss": 0.3529, + "step": 16827 + }, + { + "epoch": 0.3704160246533128, + "grad_norm": 0.6264910856543958, + "learning_rate": 7.256746073257199e-06, + "loss": 0.4929, + "step": 16828 + }, + { + "epoch": 0.37043803653973145, + "grad_norm": 1.8830177654667617, + "learning_rate": 7.256427984598438e-06, + "loss": 0.3133, + "step": 16829 + }, + { + "epoch": 0.3704600484261501, + "grad_norm": 1.6773476613616785, + "learning_rate": 7.256109884471501e-06, + "loss": 0.3744, + "step": 16830 + }, + { + "epoch": 0.3704820603125688, + "grad_norm": 1.6070779950075957, + "learning_rate": 7.255791772878009e-06, + "loss": 0.3706, + "step": 16831 + }, + { + "epoch": 0.37050407219898746, + "grad_norm": 1.6659691135550596, + "learning_rate": 7.255473649819575e-06, + "loss": 0.3117, + "step": 16832 + }, + { + "epoch": 0.3705260840854061, + "grad_norm": 1.1516261852485017, + "learning_rate": 7.2551555152978194e-06, + "loss": 0.2678, + "step": 16833 + }, + { + "epoch": 0.3705480959718248, + "grad_norm": 1.5913741299294428, + "learning_rate": 7.254837369314358e-06, + "loss": 0.454, + "step": 16834 + }, + { + "epoch": 0.37057010785824346, + "grad_norm": 1.7480063296894224, + "learning_rate": 7.2545192118708076e-06, + "loss": 0.3363, + "step": 16835 + }, + { + "epoch": 0.3705921197446621, + "grad_norm": 2.455833891396795, + "learning_rate": 7.254201042968784e-06, + "loss": 0.3764, + "step": 16836 + }, + { + "epoch": 0.37061413163108076, + "grad_norm": 1.5591714516405704, + "learning_rate": 7.253882862609904e-06, + "loss": 0.3075, + "step": 16837 + }, + { + "epoch": 0.37063614351749946, + "grad_norm": 1.8218044816714336, + "learning_rate": 7.253564670795789e-06, + "loss": 0.3994, + "step": 16838 + }, + { + "epoch": 0.3706581554039181, + "grad_norm": 1.5511111273161389, + "learning_rate": 7.253246467528053e-06, + "loss": 0.3056, + "step": 16839 + }, + { + "epoch": 0.37068016729033676, + "grad_norm": 1.7043832664725267, + "learning_rate": 7.252928252808311e-06, + "loss": 0.4589, + "step": 16840 + }, + { + "epoch": 0.37070217917675546, + "grad_norm": 2.314951747542513, + "learning_rate": 7.252610026638185e-06, + "loss": 0.2388, + "step": 16841 + }, + { + "epoch": 0.3707241910631741, + "grad_norm": 1.3865683285674473, + "learning_rate": 7.252291789019289e-06, + "loss": 0.3096, + "step": 16842 + }, + { + "epoch": 0.37074620294959276, + "grad_norm": 1.5137429044870763, + "learning_rate": 7.2519735399532424e-06, + "loss": 0.3251, + "step": 16843 + }, + { + "epoch": 0.37076821483601147, + "grad_norm": 1.3541308703123698, + "learning_rate": 7.251655279441662e-06, + "loss": 0.2392, + "step": 16844 + }, + { + "epoch": 0.3707902267224301, + "grad_norm": 1.3740671984013177, + "learning_rate": 7.2513370074861635e-06, + "loss": 0.3176, + "step": 16845 + }, + { + "epoch": 0.37081223860884877, + "grad_norm": 2.2941381639730047, + "learning_rate": 7.251018724088367e-06, + "loss": 0.3911, + "step": 16846 + }, + { + "epoch": 0.37083425049526747, + "grad_norm": 1.4753071937653106, + "learning_rate": 7.25070042924989e-06, + "loss": 0.3408, + "step": 16847 + }, + { + "epoch": 0.3708562623816861, + "grad_norm": 1.2485038253702352, + "learning_rate": 7.250382122972349e-06, + "loss": 0.265, + "step": 16848 + }, + { + "epoch": 0.37087827426810477, + "grad_norm": 1.9542327624275215, + "learning_rate": 7.2500638052573635e-06, + "loss": 0.3137, + "step": 16849 + }, + { + "epoch": 0.3709002861545234, + "grad_norm": 1.395473330872639, + "learning_rate": 7.249745476106548e-06, + "loss": 0.2538, + "step": 16850 + }, + { + "epoch": 0.3709222980409421, + "grad_norm": 1.6316185247641442, + "learning_rate": 7.249427135521524e-06, + "loss": 0.34, + "step": 16851 + }, + { + "epoch": 0.37094430992736077, + "grad_norm": 1.4658699455015034, + "learning_rate": 7.2491087835039075e-06, + "loss": 0.3177, + "step": 16852 + }, + { + "epoch": 0.3709663218137794, + "grad_norm": 0.6435022192209756, + "learning_rate": 7.248790420055318e-06, + "loss": 0.4839, + "step": 16853 + }, + { + "epoch": 0.3709883337001981, + "grad_norm": 4.7794157984493495, + "learning_rate": 7.248472045177372e-06, + "loss": 0.3574, + "step": 16854 + }, + { + "epoch": 0.3710103455866168, + "grad_norm": 1.593621571830624, + "learning_rate": 7.248153658871687e-06, + "loss": 0.3065, + "step": 16855 + }, + { + "epoch": 0.3710323574730354, + "grad_norm": 1.711384441025852, + "learning_rate": 7.247835261139885e-06, + "loss": 0.292, + "step": 16856 + }, + { + "epoch": 0.37105436935945413, + "grad_norm": 1.8107148699774136, + "learning_rate": 7.2475168519835805e-06, + "loss": 0.3043, + "step": 16857 + }, + { + "epoch": 0.3710763812458728, + "grad_norm": 1.3529788585262177, + "learning_rate": 7.247198431404392e-06, + "loss": 0.3033, + "step": 16858 + }, + { + "epoch": 0.3710983931322914, + "grad_norm": 1.9062840097305542, + "learning_rate": 7.24687999940394e-06, + "loss": 0.353, + "step": 16859 + }, + { + "epoch": 0.37112040501871013, + "grad_norm": 1.9701247492212282, + "learning_rate": 7.246561555983841e-06, + "loss": 0.3052, + "step": 16860 + }, + { + "epoch": 0.3711424169051288, + "grad_norm": 1.7964555142857652, + "learning_rate": 7.246243101145715e-06, + "loss": 0.3497, + "step": 16861 + }, + { + "epoch": 0.37116442879154743, + "grad_norm": 0.6165521254709108, + "learning_rate": 7.245924634891179e-06, + "loss": 0.5166, + "step": 16862 + }, + { + "epoch": 0.3711864406779661, + "grad_norm": 1.9278965855835322, + "learning_rate": 7.245606157221851e-06, + "loss": 0.3716, + "step": 16863 + }, + { + "epoch": 0.3712084525643848, + "grad_norm": 1.7459215461096054, + "learning_rate": 7.245287668139353e-06, + "loss": 0.3835, + "step": 16864 + }, + { + "epoch": 0.37123046445080343, + "grad_norm": 1.5975121410361708, + "learning_rate": 7.244969167645301e-06, + "loss": 0.247, + "step": 16865 + }, + { + "epoch": 0.3712524763372221, + "grad_norm": 1.6028308677020326, + "learning_rate": 7.2446506557413145e-06, + "loss": 0.3107, + "step": 16866 + }, + { + "epoch": 0.3712744882236408, + "grad_norm": 1.776411391710852, + "learning_rate": 7.2443321324290115e-06, + "loss": 0.4074, + "step": 16867 + }, + { + "epoch": 0.37129650011005944, + "grad_norm": 1.5902506390157778, + "learning_rate": 7.244013597710012e-06, + "loss": 0.2774, + "step": 16868 + }, + { + "epoch": 0.3713185119964781, + "grad_norm": 1.7301546946993893, + "learning_rate": 7.243695051585935e-06, + "loss": 0.3924, + "step": 16869 + }, + { + "epoch": 0.3713405238828968, + "grad_norm": 2.8159186143225687, + "learning_rate": 7.243376494058398e-06, + "loss": 0.3619, + "step": 16870 + }, + { + "epoch": 0.37136253576931544, + "grad_norm": 1.4150284199746763, + "learning_rate": 7.2430579251290225e-06, + "loss": 0.2892, + "step": 16871 + }, + { + "epoch": 0.3713845476557341, + "grad_norm": 0.6621424739814761, + "learning_rate": 7.242739344799425e-06, + "loss": 0.5202, + "step": 16872 + }, + { + "epoch": 0.37140655954215274, + "grad_norm": 0.6407818655580607, + "learning_rate": 7.242420753071227e-06, + "loss": 0.4947, + "step": 16873 + }, + { + "epoch": 0.37142857142857144, + "grad_norm": 2.771312625929008, + "learning_rate": 7.242102149946047e-06, + "loss": 0.3218, + "step": 16874 + }, + { + "epoch": 0.3714505833149901, + "grad_norm": 1.4860438901366853, + "learning_rate": 7.241783535425503e-06, + "loss": 0.327, + "step": 16875 + }, + { + "epoch": 0.37147259520140874, + "grad_norm": 3.3937669122260035, + "learning_rate": 7.241464909511213e-06, + "loss": 0.4118, + "step": 16876 + }, + { + "epoch": 0.37149460708782744, + "grad_norm": 1.7318226457877284, + "learning_rate": 7.241146272204802e-06, + "loss": 0.3561, + "step": 16877 + }, + { + "epoch": 0.3715166189742461, + "grad_norm": 1.717730559972475, + "learning_rate": 7.2408276235078845e-06, + "loss": 0.2452, + "step": 16878 + }, + { + "epoch": 0.37153863086066474, + "grad_norm": 1.494139411922597, + "learning_rate": 7.2405089634220816e-06, + "loss": 0.323, + "step": 16879 + }, + { + "epoch": 0.37156064274708345, + "grad_norm": 1.440696837132316, + "learning_rate": 7.240190291949014e-06, + "loss": 0.3437, + "step": 16880 + }, + { + "epoch": 0.3715826546335021, + "grad_norm": 2.124054401515624, + "learning_rate": 7.2398716090902985e-06, + "loss": 0.3054, + "step": 16881 + }, + { + "epoch": 0.37160466651992075, + "grad_norm": 1.962732048669517, + "learning_rate": 7.239552914847557e-06, + "loss": 0.3142, + "step": 16882 + }, + { + "epoch": 0.37162667840633945, + "grad_norm": 2.1145079098099036, + "learning_rate": 7.239234209222409e-06, + "loss": 0.3116, + "step": 16883 + }, + { + "epoch": 0.3716486902927581, + "grad_norm": 2.526280404266999, + "learning_rate": 7.238915492216474e-06, + "loss": 0.3181, + "step": 16884 + }, + { + "epoch": 0.37167070217917675, + "grad_norm": 1.3562456629479194, + "learning_rate": 7.238596763831373e-06, + "loss": 0.3241, + "step": 16885 + }, + { + "epoch": 0.3716927140655954, + "grad_norm": 1.7847249089396575, + "learning_rate": 7.238278024068722e-06, + "loss": 0.2961, + "step": 16886 + }, + { + "epoch": 0.3717147259520141, + "grad_norm": 0.7204938775515585, + "learning_rate": 7.237959272930146e-06, + "loss": 0.511, + "step": 16887 + }, + { + "epoch": 0.37173673783843275, + "grad_norm": 2.0533860195280362, + "learning_rate": 7.237640510417263e-06, + "loss": 0.3458, + "step": 16888 + }, + { + "epoch": 0.3717587497248514, + "grad_norm": 2.9117000692365154, + "learning_rate": 7.237321736531691e-06, + "loss": 0.3505, + "step": 16889 + }, + { + "epoch": 0.3717807616112701, + "grad_norm": 1.3817065221545624, + "learning_rate": 7.237002951275054e-06, + "loss": 0.342, + "step": 16890 + }, + { + "epoch": 0.37180277349768875, + "grad_norm": 1.6357086850089724, + "learning_rate": 7.236684154648968e-06, + "loss": 0.3653, + "step": 16891 + }, + { + "epoch": 0.3718247853841074, + "grad_norm": 1.2915391840890555, + "learning_rate": 7.236365346655057e-06, + "loss": 0.2776, + "step": 16892 + }, + { + "epoch": 0.3718467972705261, + "grad_norm": 2.2980570872904518, + "learning_rate": 7.236046527294942e-06, + "loss": 0.2791, + "step": 16893 + }, + { + "epoch": 0.37186880915694476, + "grad_norm": 1.7183297271478388, + "learning_rate": 7.235727696570238e-06, + "loss": 0.3894, + "step": 16894 + }, + { + "epoch": 0.3718908210433634, + "grad_norm": 1.7672768742254201, + "learning_rate": 7.2354088544825694e-06, + "loss": 0.3092, + "step": 16895 + }, + { + "epoch": 0.37191283292978206, + "grad_norm": 1.5245356784986475, + "learning_rate": 7.235090001033557e-06, + "loss": 0.2542, + "step": 16896 + }, + { + "epoch": 0.37193484481620076, + "grad_norm": 1.5031893317152443, + "learning_rate": 7.2347711362248196e-06, + "loss": 0.3354, + "step": 16897 + }, + { + "epoch": 0.3719568567026194, + "grad_norm": 2.2235729479628814, + "learning_rate": 7.234452260057979e-06, + "loss": 0.3291, + "step": 16898 + }, + { + "epoch": 0.37197886858903806, + "grad_norm": 1.5094123568792341, + "learning_rate": 7.234133372534654e-06, + "loss": 0.2393, + "step": 16899 + }, + { + "epoch": 0.37200088047545676, + "grad_norm": 0.6497162288495634, + "learning_rate": 7.233814473656469e-06, + "loss": 0.5046, + "step": 16900 + }, + { + "epoch": 0.3720228923618754, + "grad_norm": 1.4492496411240081, + "learning_rate": 7.2334955634250415e-06, + "loss": 0.3444, + "step": 16901 + }, + { + "epoch": 0.37204490424829406, + "grad_norm": 1.7927276323663104, + "learning_rate": 7.233176641841992e-06, + "loss": 0.3348, + "step": 16902 + }, + { + "epoch": 0.37206691613471277, + "grad_norm": 1.5742104738141385, + "learning_rate": 7.232857708908944e-06, + "loss": 0.3725, + "step": 16903 + }, + { + "epoch": 0.3720889280211314, + "grad_norm": 1.8433532276770443, + "learning_rate": 7.232538764627517e-06, + "loss": 0.3633, + "step": 16904 + }, + { + "epoch": 0.37211093990755006, + "grad_norm": 1.567269871763693, + "learning_rate": 7.2322198089993335e-06, + "loss": 0.3518, + "step": 16905 + }, + { + "epoch": 0.37213295179396877, + "grad_norm": 2.0918273798145006, + "learning_rate": 7.231900842026014e-06, + "loss": 0.3487, + "step": 16906 + }, + { + "epoch": 0.3721549636803874, + "grad_norm": 1.4979849411029669, + "learning_rate": 7.231581863709176e-06, + "loss": 0.2703, + "step": 16907 + }, + { + "epoch": 0.37217697556680607, + "grad_norm": 1.3960201550971645, + "learning_rate": 7.231262874050445e-06, + "loss": 0.3906, + "step": 16908 + }, + { + "epoch": 0.3721989874532247, + "grad_norm": 0.5971307971722308, + "learning_rate": 7.23094387305144e-06, + "loss": 0.4854, + "step": 16909 + }, + { + "epoch": 0.3722209993396434, + "grad_norm": 0.6187658926521252, + "learning_rate": 7.230624860713784e-06, + "loss": 0.4881, + "step": 16910 + }, + { + "epoch": 0.37224301122606207, + "grad_norm": 1.53562833381537, + "learning_rate": 7.230305837039099e-06, + "loss": 0.2529, + "step": 16911 + }, + { + "epoch": 0.3722650231124807, + "grad_norm": 1.7672401282795378, + "learning_rate": 7.229986802029003e-06, + "loss": 0.3062, + "step": 16912 + }, + { + "epoch": 0.3722870349988994, + "grad_norm": 1.7777674636607474, + "learning_rate": 7.2296677556851205e-06, + "loss": 0.3798, + "step": 16913 + }, + { + "epoch": 0.3723090468853181, + "grad_norm": 2.013277432222312, + "learning_rate": 7.229348698009071e-06, + "loss": 0.355, + "step": 16914 + }, + { + "epoch": 0.3723310587717367, + "grad_norm": 3.6907042802437418, + "learning_rate": 7.229029629002479e-06, + "loss": 0.2736, + "step": 16915 + }, + { + "epoch": 0.3723530706581554, + "grad_norm": 1.2725913549972148, + "learning_rate": 7.228710548666963e-06, + "loss": 0.2857, + "step": 16916 + }, + { + "epoch": 0.3723750825445741, + "grad_norm": 1.5265635993305582, + "learning_rate": 7.2283914570041445e-06, + "loss": 0.397, + "step": 16917 + }, + { + "epoch": 0.3723970944309927, + "grad_norm": 0.6355824145968498, + "learning_rate": 7.228072354015649e-06, + "loss": 0.5229, + "step": 16918 + }, + { + "epoch": 0.3724191063174114, + "grad_norm": 1.262863499552534, + "learning_rate": 7.227753239703095e-06, + "loss": 0.2977, + "step": 16919 + }, + { + "epoch": 0.3724411182038301, + "grad_norm": 1.8876438421144301, + "learning_rate": 7.227434114068106e-06, + "loss": 0.311, + "step": 16920 + }, + { + "epoch": 0.37246313009024873, + "grad_norm": 1.5878266753178678, + "learning_rate": 7.227114977112302e-06, + "loss": 0.3709, + "step": 16921 + }, + { + "epoch": 0.3724851419766674, + "grad_norm": 1.6761775583839489, + "learning_rate": 7.226795828837308e-06, + "loss": 0.3357, + "step": 16922 + }, + { + "epoch": 0.3725071538630861, + "grad_norm": 1.4609323449661895, + "learning_rate": 7.226476669244744e-06, + "loss": 0.333, + "step": 16923 + }, + { + "epoch": 0.37252916574950473, + "grad_norm": 2.3346319752766482, + "learning_rate": 7.226157498336232e-06, + "loss": 0.3254, + "step": 16924 + }, + { + "epoch": 0.3725511776359234, + "grad_norm": 1.3216864216102817, + "learning_rate": 7.225838316113396e-06, + "loss": 0.302, + "step": 16925 + }, + { + "epoch": 0.3725731895223421, + "grad_norm": 1.9560316308140482, + "learning_rate": 7.225519122577855e-06, + "loss": 0.502, + "step": 16926 + }, + { + "epoch": 0.37259520140876073, + "grad_norm": 1.3814511238075606, + "learning_rate": 7.225199917731234e-06, + "loss": 0.2812, + "step": 16927 + }, + { + "epoch": 0.3726172132951794, + "grad_norm": 1.6832265549033003, + "learning_rate": 7.224880701575154e-06, + "loss": 0.3492, + "step": 16928 + }, + { + "epoch": 0.3726392251815981, + "grad_norm": 1.6763474064641462, + "learning_rate": 7.22456147411124e-06, + "loss": 0.3123, + "step": 16929 + }, + { + "epoch": 0.37266123706801674, + "grad_norm": 1.7497483353736254, + "learning_rate": 7.224242235341111e-06, + "loss": 0.309, + "step": 16930 + }, + { + "epoch": 0.3726832489544354, + "grad_norm": 1.5066952845266175, + "learning_rate": 7.223922985266391e-06, + "loss": 0.2672, + "step": 16931 + }, + { + "epoch": 0.37270526084085404, + "grad_norm": 1.5628760757290345, + "learning_rate": 7.2236037238887025e-06, + "loss": 0.4003, + "step": 16932 + }, + { + "epoch": 0.37272727272727274, + "grad_norm": 1.4985349133741936, + "learning_rate": 7.223284451209668e-06, + "loss": 0.2423, + "step": 16933 + }, + { + "epoch": 0.3727492846136914, + "grad_norm": 1.4602429868301336, + "learning_rate": 7.222965167230911e-06, + "loss": 0.244, + "step": 16934 + }, + { + "epoch": 0.37277129650011004, + "grad_norm": 1.562767159505403, + "learning_rate": 7.222645871954052e-06, + "loss": 0.29, + "step": 16935 + }, + { + "epoch": 0.37279330838652874, + "grad_norm": 1.3813533384688654, + "learning_rate": 7.222326565380718e-06, + "loss": 0.3334, + "step": 16936 + }, + { + "epoch": 0.3728153202729474, + "grad_norm": 1.5216102753094929, + "learning_rate": 7.2220072475125254e-06, + "loss": 0.3359, + "step": 16937 + }, + { + "epoch": 0.37283733215936604, + "grad_norm": 1.6111514023886855, + "learning_rate": 7.221687918351105e-06, + "loss": 0.3008, + "step": 16938 + }, + { + "epoch": 0.37285934404578475, + "grad_norm": 3.166155867735243, + "learning_rate": 7.221368577898073e-06, + "loss": 0.3521, + "step": 16939 + }, + { + "epoch": 0.3728813559322034, + "grad_norm": 2.383212359905861, + "learning_rate": 7.221049226155055e-06, + "loss": 0.3623, + "step": 16940 + }, + { + "epoch": 0.37290336781862204, + "grad_norm": 0.6614909594610131, + "learning_rate": 7.220729863123676e-06, + "loss": 0.5321, + "step": 16941 + }, + { + "epoch": 0.37292537970504075, + "grad_norm": 1.3409124899003038, + "learning_rate": 7.220410488805556e-06, + "loss": 0.302, + "step": 16942 + }, + { + "epoch": 0.3729473915914594, + "grad_norm": 1.6684573780164405, + "learning_rate": 7.22009110320232e-06, + "loss": 0.3108, + "step": 16943 + }, + { + "epoch": 0.37296940347787805, + "grad_norm": 1.7350118505479268, + "learning_rate": 7.219771706315591e-06, + "loss": 0.3275, + "step": 16944 + }, + { + "epoch": 0.3729914153642967, + "grad_norm": 2.471956366789797, + "learning_rate": 7.219452298146993e-06, + "loss": 0.4298, + "step": 16945 + }, + { + "epoch": 0.3730134272507154, + "grad_norm": 0.6163871143538799, + "learning_rate": 7.219132878698147e-06, + "loss": 0.5097, + "step": 16946 + }, + { + "epoch": 0.37303543913713405, + "grad_norm": 5.0616371274214025, + "learning_rate": 7.218813447970678e-06, + "loss": 0.1892, + "step": 16947 + }, + { + "epoch": 0.3730574510235527, + "grad_norm": 1.5278583778606525, + "learning_rate": 7.21849400596621e-06, + "loss": 0.2776, + "step": 16948 + }, + { + "epoch": 0.3730794629099714, + "grad_norm": 1.5659944367853669, + "learning_rate": 7.218174552686366e-06, + "loss": 0.3362, + "step": 16949 + }, + { + "epoch": 0.37310147479639005, + "grad_norm": 1.4578656774583587, + "learning_rate": 7.21785508813277e-06, + "loss": 0.3158, + "step": 16950 + }, + { + "epoch": 0.3731234866828087, + "grad_norm": 1.2979534750091515, + "learning_rate": 7.217535612307044e-06, + "loss": 0.2851, + "step": 16951 + }, + { + "epoch": 0.3731454985692274, + "grad_norm": 1.8142143867090006, + "learning_rate": 7.217216125210814e-06, + "loss": 0.2922, + "step": 16952 + }, + { + "epoch": 0.37316751045564606, + "grad_norm": 1.4434878932468886, + "learning_rate": 7.216896626845702e-06, + "loss": 0.2804, + "step": 16953 + }, + { + "epoch": 0.3731895223420647, + "grad_norm": 1.4374070409659165, + "learning_rate": 7.216577117213333e-06, + "loss": 0.3263, + "step": 16954 + }, + { + "epoch": 0.37321153422848335, + "grad_norm": 0.645503931039898, + "learning_rate": 7.216257596315331e-06, + "loss": 0.5118, + "step": 16955 + }, + { + "epoch": 0.37323354611490206, + "grad_norm": 1.4087835111410731, + "learning_rate": 7.2159380641533196e-06, + "loss": 0.333, + "step": 16956 + }, + { + "epoch": 0.3732555580013207, + "grad_norm": 1.6995858844238863, + "learning_rate": 7.215618520728922e-06, + "loss": 0.4334, + "step": 16957 + }, + { + "epoch": 0.37327756988773936, + "grad_norm": 1.255919156024046, + "learning_rate": 7.215298966043762e-06, + "loss": 0.2843, + "step": 16958 + }, + { + "epoch": 0.37329958177415806, + "grad_norm": 1.1751215774612465, + "learning_rate": 7.2149794000994656e-06, + "loss": 0.3635, + "step": 16959 + }, + { + "epoch": 0.3733215936605767, + "grad_norm": 1.3951032600091893, + "learning_rate": 7.214659822897656e-06, + "loss": 0.2812, + "step": 16960 + }, + { + "epoch": 0.37334360554699536, + "grad_norm": 1.9169625419212637, + "learning_rate": 7.214340234439957e-06, + "loss": 0.3367, + "step": 16961 + }, + { + "epoch": 0.37336561743341407, + "grad_norm": 0.6339375584178761, + "learning_rate": 7.214020634727994e-06, + "loss": 0.4799, + "step": 16962 + }, + { + "epoch": 0.3733876293198327, + "grad_norm": 1.5947987600827627, + "learning_rate": 7.213701023763389e-06, + "loss": 0.3092, + "step": 16963 + }, + { + "epoch": 0.37340964120625136, + "grad_norm": 2.1223115524451224, + "learning_rate": 7.21338140154777e-06, + "loss": 0.4416, + "step": 16964 + }, + { + "epoch": 0.37343165309267007, + "grad_norm": 1.532621968929774, + "learning_rate": 7.213061768082758e-06, + "loss": 0.36, + "step": 16965 + }, + { + "epoch": 0.3734536649790887, + "grad_norm": 1.7111072029376186, + "learning_rate": 7.21274212336998e-06, + "loss": 0.3371, + "step": 16966 + }, + { + "epoch": 0.37347567686550737, + "grad_norm": 1.6931932682227433, + "learning_rate": 7.21242246741106e-06, + "loss": 0.2378, + "step": 16967 + }, + { + "epoch": 0.373497688751926, + "grad_norm": 0.6265671775557037, + "learning_rate": 7.212102800207621e-06, + "loss": 0.515, + "step": 16968 + }, + { + "epoch": 0.3735197006383447, + "grad_norm": 3.2055893817757037, + "learning_rate": 7.211783121761289e-06, + "loss": 0.3162, + "step": 16969 + }, + { + "epoch": 0.37354171252476337, + "grad_norm": 1.8918171834717272, + "learning_rate": 7.211463432073691e-06, + "loss": 0.3554, + "step": 16970 + }, + { + "epoch": 0.373563724411182, + "grad_norm": 1.145316875187701, + "learning_rate": 7.2111437311464476e-06, + "loss": 0.2582, + "step": 16971 + }, + { + "epoch": 0.3735857362976007, + "grad_norm": 2.006750206192233, + "learning_rate": 7.210824018981185e-06, + "loss": 0.3447, + "step": 16972 + }, + { + "epoch": 0.3736077481840194, + "grad_norm": 1.7545514645843274, + "learning_rate": 7.2105042955795305e-06, + "loss": 0.3316, + "step": 16973 + }, + { + "epoch": 0.373629760070438, + "grad_norm": 1.3696184181527107, + "learning_rate": 7.210184560943105e-06, + "loss": 0.3777, + "step": 16974 + }, + { + "epoch": 0.3736517719568567, + "grad_norm": 2.7954903092754666, + "learning_rate": 7.209864815073537e-06, + "loss": 0.3484, + "step": 16975 + }, + { + "epoch": 0.3736737838432754, + "grad_norm": 1.0557205221506973, + "learning_rate": 7.20954505797245e-06, + "loss": 0.2744, + "step": 16976 + }, + { + "epoch": 0.373695795729694, + "grad_norm": 1.8509449334176107, + "learning_rate": 7.209225289641471e-06, + "loss": 0.2904, + "step": 16977 + }, + { + "epoch": 0.3737178076161127, + "grad_norm": 0.6543037646053286, + "learning_rate": 7.208905510082223e-06, + "loss": 0.5005, + "step": 16978 + }, + { + "epoch": 0.3737398195025314, + "grad_norm": 0.6488786674864278, + "learning_rate": 7.2085857192963305e-06, + "loss": 0.5013, + "step": 16979 + }, + { + "epoch": 0.37376183138895, + "grad_norm": 1.3150440599737794, + "learning_rate": 7.208265917285422e-06, + "loss": 0.2972, + "step": 16980 + }, + { + "epoch": 0.3737838432753687, + "grad_norm": 1.5329183982060732, + "learning_rate": 7.207946104051121e-06, + "loss": 0.3893, + "step": 16981 + }, + { + "epoch": 0.3738058551617874, + "grad_norm": 1.6028413282125655, + "learning_rate": 7.207626279595052e-06, + "loss": 0.4007, + "step": 16982 + }, + { + "epoch": 0.37382786704820603, + "grad_norm": 2.0549293880810917, + "learning_rate": 7.207306443918842e-06, + "loss": 0.3912, + "step": 16983 + }, + { + "epoch": 0.3738498789346247, + "grad_norm": 1.4214549038758408, + "learning_rate": 7.206986597024117e-06, + "loss": 0.2755, + "step": 16984 + }, + { + "epoch": 0.3738718908210434, + "grad_norm": 1.6398524816653515, + "learning_rate": 7.206666738912502e-06, + "loss": 0.3789, + "step": 16985 + }, + { + "epoch": 0.37389390270746203, + "grad_norm": 1.9575555886178453, + "learning_rate": 7.206346869585622e-06, + "loss": 0.3535, + "step": 16986 + }, + { + "epoch": 0.3739159145938807, + "grad_norm": 2.1397729692538734, + "learning_rate": 7.206026989045102e-06, + "loss": 0.3564, + "step": 16987 + }, + { + "epoch": 0.3739379264802994, + "grad_norm": 1.7452755493032077, + "learning_rate": 7.205707097292571e-06, + "loss": 0.2763, + "step": 16988 + }, + { + "epoch": 0.37395993836671804, + "grad_norm": 0.6432751464058394, + "learning_rate": 7.20538719432965e-06, + "loss": 0.4955, + "step": 16989 + }, + { + "epoch": 0.3739819502531367, + "grad_norm": 1.4323792867681029, + "learning_rate": 7.2050672801579705e-06, + "loss": 0.2379, + "step": 16990 + }, + { + "epoch": 0.37400396213955533, + "grad_norm": 1.4912344039398187, + "learning_rate": 7.204747354779154e-06, + "loss": 0.3128, + "step": 16991 + }, + { + "epoch": 0.37402597402597404, + "grad_norm": 1.3988551724370397, + "learning_rate": 7.204427418194828e-06, + "loss": 0.3636, + "step": 16992 + }, + { + "epoch": 0.3740479859123927, + "grad_norm": 1.46082237572676, + "learning_rate": 7.20410747040662e-06, + "loss": 0.3952, + "step": 16993 + }, + { + "epoch": 0.37406999779881134, + "grad_norm": 0.6206488674567777, + "learning_rate": 7.203787511416153e-06, + "loss": 0.526, + "step": 16994 + }, + { + "epoch": 0.37409200968523004, + "grad_norm": 1.5670656727956194, + "learning_rate": 7.203467541225055e-06, + "loss": 0.2995, + "step": 16995 + }, + { + "epoch": 0.3741140215716487, + "grad_norm": 1.3084761204788522, + "learning_rate": 7.203147559834954e-06, + "loss": 0.3695, + "step": 16996 + }, + { + "epoch": 0.37413603345806734, + "grad_norm": 1.3877847483224657, + "learning_rate": 7.202827567247474e-06, + "loss": 0.2216, + "step": 16997 + }, + { + "epoch": 0.37415804534448605, + "grad_norm": 1.405291993225619, + "learning_rate": 7.2025075634642415e-06, + "loss": 0.3628, + "step": 16998 + }, + { + "epoch": 0.3741800572309047, + "grad_norm": 1.7976786015509632, + "learning_rate": 7.202187548486883e-06, + "loss": 0.3585, + "step": 16999 + }, + { + "epoch": 0.37420206911732334, + "grad_norm": 2.1685171723539085, + "learning_rate": 7.201867522317024e-06, + "loss": 0.3523, + "step": 17000 + }, + { + "epoch": 0.37422408100374205, + "grad_norm": 1.5279128852051775, + "learning_rate": 7.201547484956294e-06, + "loss": 0.3622, + "step": 17001 + }, + { + "epoch": 0.3742460928901607, + "grad_norm": 2.147446288857207, + "learning_rate": 7.201227436406317e-06, + "loss": 0.2101, + "step": 17002 + }, + { + "epoch": 0.37426810477657935, + "grad_norm": 1.289302790080782, + "learning_rate": 7.200907376668721e-06, + "loss": 0.3063, + "step": 17003 + }, + { + "epoch": 0.374290116662998, + "grad_norm": 1.5553292790261461, + "learning_rate": 7.200587305745132e-06, + "loss": 0.4734, + "step": 17004 + }, + { + "epoch": 0.3743121285494167, + "grad_norm": 2.2353033916650125, + "learning_rate": 7.200267223637175e-06, + "loss": 0.3296, + "step": 17005 + }, + { + "epoch": 0.37433414043583535, + "grad_norm": 1.3437764841530073, + "learning_rate": 7.199947130346482e-06, + "loss": 0.3487, + "step": 17006 + }, + { + "epoch": 0.374356152322254, + "grad_norm": 1.900188125709893, + "learning_rate": 7.199627025874674e-06, + "loss": 0.385, + "step": 17007 + }, + { + "epoch": 0.3743781642086727, + "grad_norm": 0.6806453124274807, + "learning_rate": 7.19930691022338e-06, + "loss": 0.5268, + "step": 17008 + }, + { + "epoch": 0.37440017609509135, + "grad_norm": 1.8962415212597266, + "learning_rate": 7.198986783394229e-06, + "loss": 0.4004, + "step": 17009 + }, + { + "epoch": 0.37442218798151, + "grad_norm": 1.5767639855703117, + "learning_rate": 7.198666645388844e-06, + "loss": 0.3569, + "step": 17010 + }, + { + "epoch": 0.3744441998679287, + "grad_norm": 2.736103124863356, + "learning_rate": 7.1983464962088574e-06, + "loss": 0.3476, + "step": 17011 + }, + { + "epoch": 0.37446621175434736, + "grad_norm": 4.233460026787009, + "learning_rate": 7.198026335855892e-06, + "loss": 0.3462, + "step": 17012 + }, + { + "epoch": 0.374488223640766, + "grad_norm": 1.5522359650539996, + "learning_rate": 7.1977061643315775e-06, + "loss": 0.4054, + "step": 17013 + }, + { + "epoch": 0.37451023552718465, + "grad_norm": 2.92092626153159, + "learning_rate": 7.1973859816375394e-06, + "loss": 0.2574, + "step": 17014 + }, + { + "epoch": 0.37453224741360336, + "grad_norm": 1.2801301232648714, + "learning_rate": 7.197065787775405e-06, + "loss": 0.3462, + "step": 17015 + }, + { + "epoch": 0.374554259300022, + "grad_norm": 1.6861934201754998, + "learning_rate": 7.196745582746802e-06, + "loss": 0.3214, + "step": 17016 + }, + { + "epoch": 0.37457627118644066, + "grad_norm": 2.853585133564491, + "learning_rate": 7.19642536655336e-06, + "loss": 0.2853, + "step": 17017 + }, + { + "epoch": 0.37459828307285936, + "grad_norm": 1.415696591540017, + "learning_rate": 7.196105139196703e-06, + "loss": 0.2784, + "step": 17018 + }, + { + "epoch": 0.374620294959278, + "grad_norm": 1.2980809717141795, + "learning_rate": 7.195784900678461e-06, + "loss": 0.4264, + "step": 17019 + }, + { + "epoch": 0.37464230684569666, + "grad_norm": 1.4572360698205231, + "learning_rate": 7.19546465100026e-06, + "loss": 0.3345, + "step": 17020 + }, + { + "epoch": 0.37466431873211536, + "grad_norm": 1.4136736952255349, + "learning_rate": 7.19514439016373e-06, + "loss": 0.3529, + "step": 17021 + }, + { + "epoch": 0.374686330618534, + "grad_norm": 1.7311178417910167, + "learning_rate": 7.194824118170496e-06, + "loss": 0.3083, + "step": 17022 + }, + { + "epoch": 0.37470834250495266, + "grad_norm": 1.6867558632895667, + "learning_rate": 7.194503835022186e-06, + "loss": 0.4029, + "step": 17023 + }, + { + "epoch": 0.37473035439137137, + "grad_norm": 1.4726121862209174, + "learning_rate": 7.194183540720432e-06, + "loss": 0.3308, + "step": 17024 + }, + { + "epoch": 0.37475236627779, + "grad_norm": 0.6457711042788283, + "learning_rate": 7.193863235266855e-06, + "loss": 0.4924, + "step": 17025 + }, + { + "epoch": 0.37477437816420867, + "grad_norm": 5.283157949945035, + "learning_rate": 7.193542918663088e-06, + "loss": 0.3112, + "step": 17026 + }, + { + "epoch": 0.3747963900506273, + "grad_norm": 1.6677083226658225, + "learning_rate": 7.193222590910758e-06, + "loss": 0.4392, + "step": 17027 + }, + { + "epoch": 0.374818401937046, + "grad_norm": 1.4062061749960468, + "learning_rate": 7.192902252011491e-06, + "loss": 0.3891, + "step": 17028 + }, + { + "epoch": 0.37484041382346467, + "grad_norm": 1.6070437248152094, + "learning_rate": 7.192581901966919e-06, + "loss": 0.3944, + "step": 17029 + }, + { + "epoch": 0.3748624257098833, + "grad_norm": 1.300485695552813, + "learning_rate": 7.192261540778666e-06, + "loss": 0.3999, + "step": 17030 + }, + { + "epoch": 0.374884437596302, + "grad_norm": 1.3956345371672452, + "learning_rate": 7.191941168448363e-06, + "loss": 0.3443, + "step": 17031 + }, + { + "epoch": 0.37490644948272067, + "grad_norm": 1.1994426510917748, + "learning_rate": 7.191620784977638e-06, + "loss": 0.3264, + "step": 17032 + }, + { + "epoch": 0.3749284613691393, + "grad_norm": 1.2408328774877482, + "learning_rate": 7.1913003903681154e-06, + "loss": 0.2792, + "step": 17033 + }, + { + "epoch": 0.374950473255558, + "grad_norm": 1.416925997983458, + "learning_rate": 7.190979984621431e-06, + "loss": 0.3559, + "step": 17034 + }, + { + "epoch": 0.3749724851419767, + "grad_norm": 1.2902185913577309, + "learning_rate": 7.190659567739207e-06, + "loss": 0.2542, + "step": 17035 + }, + { + "epoch": 0.3749944970283953, + "grad_norm": 1.885528388841812, + "learning_rate": 7.190339139723074e-06, + "loss": 0.3724, + "step": 17036 + }, + { + "epoch": 0.375016508914814, + "grad_norm": 1.2637095524945374, + "learning_rate": 7.190018700574661e-06, + "loss": 0.3512, + "step": 17037 + }, + { + "epoch": 0.3750385208012327, + "grad_norm": 1.4896492249967934, + "learning_rate": 7.189698250295596e-06, + "loss": 0.2261, + "step": 17038 + }, + { + "epoch": 0.3750605326876513, + "grad_norm": 1.4827979409695227, + "learning_rate": 7.189377788887508e-06, + "loss": 0.3498, + "step": 17039 + }, + { + "epoch": 0.37508254457407, + "grad_norm": 2.0781683876420463, + "learning_rate": 7.189057316352026e-06, + "loss": 0.3858, + "step": 17040 + }, + { + "epoch": 0.3751045564604887, + "grad_norm": 0.6450234976960998, + "learning_rate": 7.188736832690777e-06, + "loss": 0.5006, + "step": 17041 + }, + { + "epoch": 0.37512656834690733, + "grad_norm": 0.646069871445484, + "learning_rate": 7.188416337905393e-06, + "loss": 0.5029, + "step": 17042 + }, + { + "epoch": 0.375148580233326, + "grad_norm": 1.650983025273392, + "learning_rate": 7.1880958319975e-06, + "loss": 0.3068, + "step": 17043 + }, + { + "epoch": 0.3751705921197447, + "grad_norm": 1.6595064862493607, + "learning_rate": 7.187775314968729e-06, + "loss": 0.4402, + "step": 17044 + }, + { + "epoch": 0.37519260400616333, + "grad_norm": 2.0683529231986033, + "learning_rate": 7.187454786820707e-06, + "loss": 0.3666, + "step": 17045 + }, + { + "epoch": 0.375214615892582, + "grad_norm": 1.4579147198905857, + "learning_rate": 7.1871342475550645e-06, + "loss": 0.3574, + "step": 17046 + }, + { + "epoch": 0.3752366277790007, + "grad_norm": 1.3338147305448955, + "learning_rate": 7.186813697173431e-06, + "loss": 0.2609, + "step": 17047 + }, + { + "epoch": 0.37525863966541934, + "grad_norm": 1.4586222438176288, + "learning_rate": 7.186493135677434e-06, + "loss": 0.3285, + "step": 17048 + }, + { + "epoch": 0.375280651551838, + "grad_norm": 1.4774920114566346, + "learning_rate": 7.186172563068703e-06, + "loss": 0.3312, + "step": 17049 + }, + { + "epoch": 0.37530266343825663, + "grad_norm": 2.15908716028853, + "learning_rate": 7.1858519793488705e-06, + "loss": 0.4062, + "step": 17050 + }, + { + "epoch": 0.37532467532467534, + "grad_norm": 1.5258100134220678, + "learning_rate": 7.185531384519559e-06, + "loss": 0.3209, + "step": 17051 + }, + { + "epoch": 0.375346687211094, + "grad_norm": 0.7027157746739355, + "learning_rate": 7.185210778582407e-06, + "loss": 0.4967, + "step": 17052 + }, + { + "epoch": 0.37536869909751264, + "grad_norm": 1.416706614370595, + "learning_rate": 7.184890161539038e-06, + "loss": 0.192, + "step": 17053 + }, + { + "epoch": 0.37539071098393134, + "grad_norm": 1.4795013290990655, + "learning_rate": 7.18456953339108e-06, + "loss": 0.3509, + "step": 17054 + }, + { + "epoch": 0.37541272287035, + "grad_norm": 1.8274849701949034, + "learning_rate": 7.184248894140168e-06, + "loss": 0.3461, + "step": 17055 + }, + { + "epoch": 0.37543473475676864, + "grad_norm": 1.4887609059748903, + "learning_rate": 7.183928243787928e-06, + "loss": 0.3665, + "step": 17056 + }, + { + "epoch": 0.37545674664318734, + "grad_norm": 1.3417481100135253, + "learning_rate": 7.183607582335991e-06, + "loss": 0.2762, + "step": 17057 + }, + { + "epoch": 0.375478758529606, + "grad_norm": 1.1784710603356077, + "learning_rate": 7.183286909785986e-06, + "loss": 0.2545, + "step": 17058 + }, + { + "epoch": 0.37550077041602464, + "grad_norm": 2.3982520410149877, + "learning_rate": 7.1829662261395425e-06, + "loss": 0.3177, + "step": 17059 + }, + { + "epoch": 0.3755227823024433, + "grad_norm": 1.206663111114515, + "learning_rate": 7.182645531398292e-06, + "loss": 0.3041, + "step": 17060 + }, + { + "epoch": 0.375544794188862, + "grad_norm": 1.6082867759517399, + "learning_rate": 7.1823248255638635e-06, + "loss": 0.4745, + "step": 17061 + }, + { + "epoch": 0.37556680607528065, + "grad_norm": 1.791907262372607, + "learning_rate": 7.182004108637887e-06, + "loss": 0.3068, + "step": 17062 + }, + { + "epoch": 0.3755888179616993, + "grad_norm": 1.4249458602378606, + "learning_rate": 7.181683380621991e-06, + "loss": 0.363, + "step": 17063 + }, + { + "epoch": 0.375610829848118, + "grad_norm": 1.5544568534569023, + "learning_rate": 7.181362641517807e-06, + "loss": 0.2954, + "step": 17064 + }, + { + "epoch": 0.37563284173453665, + "grad_norm": 2.0370211810122356, + "learning_rate": 7.181041891326967e-06, + "loss": 0.3213, + "step": 17065 + }, + { + "epoch": 0.3756548536209553, + "grad_norm": 1.4735852485539744, + "learning_rate": 7.180721130051098e-06, + "loss": 0.3709, + "step": 17066 + }, + { + "epoch": 0.375676865507374, + "grad_norm": 1.9898559908981077, + "learning_rate": 7.180400357691832e-06, + "loss": 0.283, + "step": 17067 + }, + { + "epoch": 0.37569887739379265, + "grad_norm": 1.4526058886436963, + "learning_rate": 7.1800795742508e-06, + "loss": 0.2919, + "step": 17068 + }, + { + "epoch": 0.3757208892802113, + "grad_norm": 1.4884296727408999, + "learning_rate": 7.179758779729629e-06, + "loss": 0.2794, + "step": 17069 + }, + { + "epoch": 0.37574290116663, + "grad_norm": 1.2561549414612678, + "learning_rate": 7.179437974129954e-06, + "loss": 0.2887, + "step": 17070 + }, + { + "epoch": 0.37576491305304865, + "grad_norm": 1.4422349241243024, + "learning_rate": 7.179117157453402e-06, + "loss": 0.3058, + "step": 17071 + }, + { + "epoch": 0.3757869249394673, + "grad_norm": 2.0646640374147314, + "learning_rate": 7.178796329701604e-06, + "loss": 0.3031, + "step": 17072 + }, + { + "epoch": 0.37580893682588595, + "grad_norm": 1.3751510372245077, + "learning_rate": 7.178475490876192e-06, + "loss": 0.2996, + "step": 17073 + }, + { + "epoch": 0.37583094871230466, + "grad_norm": 1.4056753014239882, + "learning_rate": 7.1781546409787965e-06, + "loss": 0.313, + "step": 17074 + }, + { + "epoch": 0.3758529605987233, + "grad_norm": 1.8286936320338532, + "learning_rate": 7.177833780011046e-06, + "loss": 0.369, + "step": 17075 + }, + { + "epoch": 0.37587497248514196, + "grad_norm": 0.6687939682878059, + "learning_rate": 7.177512907974576e-06, + "loss": 0.4848, + "step": 17076 + }, + { + "epoch": 0.37589698437156066, + "grad_norm": 1.476144591921415, + "learning_rate": 7.17719202487101e-06, + "loss": 0.2916, + "step": 17077 + }, + { + "epoch": 0.3759189962579793, + "grad_norm": 1.5902155397306943, + "learning_rate": 7.176871130701985e-06, + "loss": 0.3865, + "step": 17078 + }, + { + "epoch": 0.37594100814439796, + "grad_norm": 1.6727246803801115, + "learning_rate": 7.176550225469131e-06, + "loss": 0.3617, + "step": 17079 + }, + { + "epoch": 0.37596302003081666, + "grad_norm": 1.4625679314736204, + "learning_rate": 7.1762293091740775e-06, + "loss": 0.3472, + "step": 17080 + }, + { + "epoch": 0.3759850319172353, + "grad_norm": 1.4024683822749837, + "learning_rate": 7.175908381818456e-06, + "loss": 0.2543, + "step": 17081 + }, + { + "epoch": 0.37600704380365396, + "grad_norm": 0.6345365602706824, + "learning_rate": 7.1755874434038965e-06, + "loss": 0.4936, + "step": 17082 + }, + { + "epoch": 0.37602905569007267, + "grad_norm": 1.5709675815024196, + "learning_rate": 7.175266493932032e-06, + "loss": 0.3618, + "step": 17083 + }, + { + "epoch": 0.3760510675764913, + "grad_norm": 1.5329375905956408, + "learning_rate": 7.174945533404493e-06, + "loss": 0.3923, + "step": 17084 + }, + { + "epoch": 0.37607307946290996, + "grad_norm": 4.883898631222193, + "learning_rate": 7.1746245618229115e-06, + "loss": 0.3902, + "step": 17085 + }, + { + "epoch": 0.3760950913493286, + "grad_norm": 1.6948832338835436, + "learning_rate": 7.174303579188917e-06, + "loss": 0.2577, + "step": 17086 + }, + { + "epoch": 0.3761171032357473, + "grad_norm": 1.9909212922953718, + "learning_rate": 7.173982585504142e-06, + "loss": 0.3174, + "step": 17087 + }, + { + "epoch": 0.37613911512216597, + "grad_norm": 2.7648531829027108, + "learning_rate": 7.173661580770219e-06, + "loss": 0.22, + "step": 17088 + }, + { + "epoch": 0.3761611270085846, + "grad_norm": 1.257375250735078, + "learning_rate": 7.173340564988778e-06, + "loss": 0.3149, + "step": 17089 + }, + { + "epoch": 0.3761831388950033, + "grad_norm": 1.5266492458993168, + "learning_rate": 7.17301953816145e-06, + "loss": 0.2958, + "step": 17090 + }, + { + "epoch": 0.37620515078142197, + "grad_norm": 1.5683443626340048, + "learning_rate": 7.172698500289867e-06, + "loss": 0.339, + "step": 17091 + }, + { + "epoch": 0.3762271626678406, + "grad_norm": 1.1794113099870218, + "learning_rate": 7.172377451375663e-06, + "loss": 0.3201, + "step": 17092 + }, + { + "epoch": 0.3762491745542593, + "grad_norm": 1.6677920183258845, + "learning_rate": 7.172056391420467e-06, + "loss": 0.363, + "step": 17093 + }, + { + "epoch": 0.376271186440678, + "grad_norm": 0.7046132970499264, + "learning_rate": 7.171735320425912e-06, + "loss": 0.5177, + "step": 17094 + }, + { + "epoch": 0.3762931983270966, + "grad_norm": 0.6704852071317587, + "learning_rate": 7.171414238393628e-06, + "loss": 0.5128, + "step": 17095 + }, + { + "epoch": 0.37631521021351527, + "grad_norm": 1.7296625791031797, + "learning_rate": 7.1710931453252495e-06, + "loss": 0.2679, + "step": 17096 + }, + { + "epoch": 0.376337222099934, + "grad_norm": 2.889308856097422, + "learning_rate": 7.170772041222408e-06, + "loss": 0.3077, + "step": 17097 + }, + { + "epoch": 0.3763592339863526, + "grad_norm": 1.279771449900305, + "learning_rate": 7.170450926086735e-06, + "loss": 0.2598, + "step": 17098 + }, + { + "epoch": 0.3763812458727713, + "grad_norm": 1.4899630298790552, + "learning_rate": 7.170129799919861e-06, + "loss": 0.4012, + "step": 17099 + }, + { + "epoch": 0.37640325775919, + "grad_norm": 1.4517356124295122, + "learning_rate": 7.16980866272342e-06, + "loss": 0.3229, + "step": 17100 + }, + { + "epoch": 0.37642526964560863, + "grad_norm": 1.3718757666075176, + "learning_rate": 7.169487514499043e-06, + "loss": 0.314, + "step": 17101 + }, + { + "epoch": 0.3764472815320273, + "grad_norm": 0.7108445797698053, + "learning_rate": 7.1691663552483635e-06, + "loss": 0.5039, + "step": 17102 + }, + { + "epoch": 0.376469293418446, + "grad_norm": 1.2114196084071922, + "learning_rate": 7.168845184973013e-06, + "loss": 0.3003, + "step": 17103 + }, + { + "epoch": 0.37649130530486463, + "grad_norm": 1.307410525393725, + "learning_rate": 7.168524003674624e-06, + "loss": 0.2746, + "step": 17104 + }, + { + "epoch": 0.3765133171912833, + "grad_norm": 0.6512915279787678, + "learning_rate": 7.168202811354828e-06, + "loss": 0.508, + "step": 17105 + }, + { + "epoch": 0.376535329077702, + "grad_norm": 1.1929820247041798, + "learning_rate": 7.167881608015262e-06, + "loss": 0.3728, + "step": 17106 + }, + { + "epoch": 0.37655734096412063, + "grad_norm": 1.5916949438043988, + "learning_rate": 7.167560393657551e-06, + "loss": 0.3111, + "step": 17107 + }, + { + "epoch": 0.3765793528505393, + "grad_norm": 1.7839604550442447, + "learning_rate": 7.167239168283331e-06, + "loss": 0.4502, + "step": 17108 + }, + { + "epoch": 0.37660136473695793, + "grad_norm": 1.2548033960126015, + "learning_rate": 7.166917931894236e-06, + "loss": 0.2664, + "step": 17109 + }, + { + "epoch": 0.37662337662337664, + "grad_norm": 1.2071584465414191, + "learning_rate": 7.166596684491898e-06, + "loss": 0.284, + "step": 17110 + }, + { + "epoch": 0.3766453885097953, + "grad_norm": 1.485714956280974, + "learning_rate": 7.166275426077949e-06, + "loss": 0.2763, + "step": 17111 + }, + { + "epoch": 0.37666740039621394, + "grad_norm": 1.4818768080906177, + "learning_rate": 7.165954156654023e-06, + "loss": 0.3383, + "step": 17112 + }, + { + "epoch": 0.37668941228263264, + "grad_norm": 1.2130577445510984, + "learning_rate": 7.16563287622175e-06, + "loss": 0.3264, + "step": 17113 + }, + { + "epoch": 0.3767114241690513, + "grad_norm": 1.2872728890867031, + "learning_rate": 7.1653115847827655e-06, + "loss": 0.283, + "step": 17114 + }, + { + "epoch": 0.37673343605546994, + "grad_norm": 1.4928988379086778, + "learning_rate": 7.1649902823387015e-06, + "loss": 0.2916, + "step": 17115 + }, + { + "epoch": 0.37675544794188864, + "grad_norm": 1.66007585143624, + "learning_rate": 7.164668968891193e-06, + "loss": 0.3423, + "step": 17116 + }, + { + "epoch": 0.3767774598283073, + "grad_norm": 1.267361185069857, + "learning_rate": 7.164347644441871e-06, + "loss": 0.2755, + "step": 17117 + }, + { + "epoch": 0.37679947171472594, + "grad_norm": 18.65736155419564, + "learning_rate": 7.1640263089923665e-06, + "loss": 0.2822, + "step": 17118 + }, + { + "epoch": 0.3768214836011446, + "grad_norm": 1.3767822535958478, + "learning_rate": 7.1637049625443175e-06, + "loss": 0.2645, + "step": 17119 + }, + { + "epoch": 0.3768434954875633, + "grad_norm": 1.678150834232892, + "learning_rate": 7.163383605099354e-06, + "loss": 0.4138, + "step": 17120 + }, + { + "epoch": 0.37686550737398194, + "grad_norm": 1.60110605628893, + "learning_rate": 7.163062236659111e-06, + "loss": 0.2797, + "step": 17121 + }, + { + "epoch": 0.3768875192604006, + "grad_norm": 2.1491002488501283, + "learning_rate": 7.16274085722522e-06, + "loss": 0.26, + "step": 17122 + }, + { + "epoch": 0.3769095311468193, + "grad_norm": 1.4315530399650518, + "learning_rate": 7.162419466799314e-06, + "loss": 0.3517, + "step": 17123 + }, + { + "epoch": 0.37693154303323795, + "grad_norm": 1.7472926060043061, + "learning_rate": 7.1620980653830305e-06, + "loss": 0.4356, + "step": 17124 + }, + { + "epoch": 0.3769535549196566, + "grad_norm": 1.8368686229547726, + "learning_rate": 7.161776652978e-06, + "loss": 0.3625, + "step": 17125 + }, + { + "epoch": 0.3769755668060753, + "grad_norm": 2.195776609048237, + "learning_rate": 7.161455229585854e-06, + "loss": 0.3137, + "step": 17126 + }, + { + "epoch": 0.37699757869249395, + "grad_norm": 1.6192393604805309, + "learning_rate": 7.16113379520823e-06, + "loss": 0.3871, + "step": 17127 + }, + { + "epoch": 0.3770195905789126, + "grad_norm": 0.7657650474050851, + "learning_rate": 7.160812349846761e-06, + "loss": 0.4831, + "step": 17128 + }, + { + "epoch": 0.3770416024653313, + "grad_norm": 1.456323670304144, + "learning_rate": 7.160490893503079e-06, + "loss": 0.3162, + "step": 17129 + }, + { + "epoch": 0.37706361435174995, + "grad_norm": 1.6490737320693167, + "learning_rate": 7.160169426178819e-06, + "loss": 0.343, + "step": 17130 + }, + { + "epoch": 0.3770856262381686, + "grad_norm": 1.5691729209755698, + "learning_rate": 7.1598479478756135e-06, + "loss": 0.3245, + "step": 17131 + }, + { + "epoch": 0.37710763812458725, + "grad_norm": 1.5140725606538772, + "learning_rate": 7.159526458595099e-06, + "loss": 0.2908, + "step": 17132 + }, + { + "epoch": 0.37712965001100596, + "grad_norm": 2.240012638391975, + "learning_rate": 7.159204958338907e-06, + "loss": 0.4081, + "step": 17133 + }, + { + "epoch": 0.3771516618974246, + "grad_norm": 1.7166494338687799, + "learning_rate": 7.158883447108672e-06, + "loss": 0.3412, + "step": 17134 + }, + { + "epoch": 0.37717367378384326, + "grad_norm": 1.7907056649970998, + "learning_rate": 7.158561924906029e-06, + "loss": 0.3897, + "step": 17135 + }, + { + "epoch": 0.37719568567026196, + "grad_norm": 2.0194779195570907, + "learning_rate": 7.158240391732611e-06, + "loss": 0.4161, + "step": 17136 + }, + { + "epoch": 0.3772176975566806, + "grad_norm": 1.9782103719152004, + "learning_rate": 7.157918847590053e-06, + "loss": 0.4184, + "step": 17137 + }, + { + "epoch": 0.37723970944309926, + "grad_norm": 1.5472326651862611, + "learning_rate": 7.15759729247999e-06, + "loss": 0.3359, + "step": 17138 + }, + { + "epoch": 0.37726172132951796, + "grad_norm": 6.1121142102199615, + "learning_rate": 7.157275726404054e-06, + "loss": 0.4534, + "step": 17139 + }, + { + "epoch": 0.3772837332159366, + "grad_norm": 1.348586670741124, + "learning_rate": 7.156954149363881e-06, + "loss": 0.3225, + "step": 17140 + }, + { + "epoch": 0.37730574510235526, + "grad_norm": 1.6941298567482563, + "learning_rate": 7.156632561361103e-06, + "loss": 0.3677, + "step": 17141 + }, + { + "epoch": 0.3773277569887739, + "grad_norm": 1.5211988900726698, + "learning_rate": 7.156310962397359e-06, + "loss": 0.3494, + "step": 17142 + }, + { + "epoch": 0.3773497688751926, + "grad_norm": 0.7483173805015242, + "learning_rate": 7.155989352474282e-06, + "loss": 0.4743, + "step": 17143 + }, + { + "epoch": 0.37737178076161126, + "grad_norm": 1.4699166355120692, + "learning_rate": 7.155667731593503e-06, + "loss": 0.3057, + "step": 17144 + }, + { + "epoch": 0.3773937926480299, + "grad_norm": 1.5542439327737483, + "learning_rate": 7.155346099756659e-06, + "loss": 0.3456, + "step": 17145 + }, + { + "epoch": 0.3774158045344486, + "grad_norm": 1.6152402072033238, + "learning_rate": 7.155024456965387e-06, + "loss": 0.3458, + "step": 17146 + }, + { + "epoch": 0.37743781642086727, + "grad_norm": 1.360785726731681, + "learning_rate": 7.154702803221318e-06, + "loss": 0.2955, + "step": 17147 + }, + { + "epoch": 0.3774598283072859, + "grad_norm": 1.7721614237015746, + "learning_rate": 7.154381138526088e-06, + "loss": 0.3402, + "step": 17148 + }, + { + "epoch": 0.3774818401937046, + "grad_norm": 1.6744424863933247, + "learning_rate": 7.154059462881333e-06, + "loss": 0.3695, + "step": 17149 + }, + { + "epoch": 0.37750385208012327, + "grad_norm": 1.7251575458212227, + "learning_rate": 7.153737776288687e-06, + "loss": 0.3656, + "step": 17150 + }, + { + "epoch": 0.3775258639665419, + "grad_norm": 1.332509046416029, + "learning_rate": 7.153416078749784e-06, + "loss": 0.242, + "step": 17151 + }, + { + "epoch": 0.3775478758529606, + "grad_norm": 2.2357454713498814, + "learning_rate": 7.153094370266261e-06, + "loss": 0.2608, + "step": 17152 + }, + { + "epoch": 0.3775698877393793, + "grad_norm": 1.5416529019220695, + "learning_rate": 7.152772650839754e-06, + "loss": 0.3098, + "step": 17153 + }, + { + "epoch": 0.3775918996257979, + "grad_norm": 1.5164214337123951, + "learning_rate": 7.152450920471892e-06, + "loss": 0.2689, + "step": 17154 + }, + { + "epoch": 0.37761391151221657, + "grad_norm": 0.6413553473158848, + "learning_rate": 7.152129179164317e-06, + "loss": 0.4838, + "step": 17155 + }, + { + "epoch": 0.3776359233986353, + "grad_norm": 1.6441134803476662, + "learning_rate": 7.151807426918662e-06, + "loss": 0.3885, + "step": 17156 + }, + { + "epoch": 0.3776579352850539, + "grad_norm": 1.9757838371134575, + "learning_rate": 7.1514856637365615e-06, + "loss": 0.3368, + "step": 17157 + }, + { + "epoch": 0.3776799471714726, + "grad_norm": 1.649404815184744, + "learning_rate": 7.15116388961965e-06, + "loss": 0.3361, + "step": 17158 + }, + { + "epoch": 0.3777019590578913, + "grad_norm": 1.4558181243086676, + "learning_rate": 7.150842104569564e-06, + "loss": 0.3642, + "step": 17159 + }, + { + "epoch": 0.37772397094430993, + "grad_norm": 1.5103697639722933, + "learning_rate": 7.150520308587942e-06, + "loss": 0.3211, + "step": 17160 + }, + { + "epoch": 0.3777459828307286, + "grad_norm": 1.4765977937664652, + "learning_rate": 7.150198501676415e-06, + "loss": 0.2612, + "step": 17161 + }, + { + "epoch": 0.3777679947171473, + "grad_norm": 0.6280229131628169, + "learning_rate": 7.14987668383662e-06, + "loss": 0.5072, + "step": 17162 + }, + { + "epoch": 0.37779000660356593, + "grad_norm": 1.4463328520670484, + "learning_rate": 7.149554855070193e-06, + "loss": 0.3299, + "step": 17163 + }, + { + "epoch": 0.3778120184899846, + "grad_norm": 1.3141115081207335, + "learning_rate": 7.1492330153787695e-06, + "loss": 0.2795, + "step": 17164 + }, + { + "epoch": 0.3778340303764033, + "grad_norm": 1.7600049874387604, + "learning_rate": 7.148911164763985e-06, + "loss": 0.3562, + "step": 17165 + }, + { + "epoch": 0.37785604226282193, + "grad_norm": 1.632623763613543, + "learning_rate": 7.148589303227475e-06, + "loss": 0.321, + "step": 17166 + }, + { + "epoch": 0.3778780541492406, + "grad_norm": 3.9500585790722953, + "learning_rate": 7.148267430770877e-06, + "loss": 0.2878, + "step": 17167 + }, + { + "epoch": 0.37790006603565923, + "grad_norm": 1.9148333201513308, + "learning_rate": 7.147945547395825e-06, + "loss": 0.3748, + "step": 17168 + }, + { + "epoch": 0.37792207792207794, + "grad_norm": 1.7140736702111248, + "learning_rate": 7.1476236531039565e-06, + "loss": 0.3187, + "step": 17169 + }, + { + "epoch": 0.3779440898084966, + "grad_norm": 1.3991663372771868, + "learning_rate": 7.147301747896906e-06, + "loss": 0.3144, + "step": 17170 + }, + { + "epoch": 0.37796610169491524, + "grad_norm": 2.2551561540272007, + "learning_rate": 7.1469798317763105e-06, + "loss": 0.3538, + "step": 17171 + }, + { + "epoch": 0.37798811358133394, + "grad_norm": 1.3985976897460095, + "learning_rate": 7.146657904743805e-06, + "loss": 0.2878, + "step": 17172 + }, + { + "epoch": 0.3780101254677526, + "grad_norm": 2.020141081451291, + "learning_rate": 7.146335966801028e-06, + "loss": 0.3594, + "step": 17173 + }, + { + "epoch": 0.37803213735417124, + "grad_norm": 1.4110964791861489, + "learning_rate": 7.146014017949615e-06, + "loss": 0.3258, + "step": 17174 + }, + { + "epoch": 0.37805414924058994, + "grad_norm": 1.2215389796448226, + "learning_rate": 7.145692058191199e-06, + "loss": 0.2335, + "step": 17175 + }, + { + "epoch": 0.3780761611270086, + "grad_norm": 1.9125917848852874, + "learning_rate": 7.145370087527421e-06, + "loss": 0.4078, + "step": 17176 + }, + { + "epoch": 0.37809817301342724, + "grad_norm": 2.48747342707507, + "learning_rate": 7.145048105959914e-06, + "loss": 0.2764, + "step": 17177 + }, + { + "epoch": 0.3781201848998459, + "grad_norm": 2.3136129022587264, + "learning_rate": 7.144726113490317e-06, + "loss": 0.4285, + "step": 17178 + }, + { + "epoch": 0.3781421967862646, + "grad_norm": 2.1794381918896844, + "learning_rate": 7.1444041101202664e-06, + "loss": 0.3158, + "step": 17179 + }, + { + "epoch": 0.37816420867268324, + "grad_norm": 0.6995197122891625, + "learning_rate": 7.144082095851394e-06, + "loss": 0.523, + "step": 17180 + }, + { + "epoch": 0.3781862205591019, + "grad_norm": 0.6934808219824183, + "learning_rate": 7.143760070685344e-06, + "loss": 0.5236, + "step": 17181 + }, + { + "epoch": 0.3782082324455206, + "grad_norm": 1.7744330326611126, + "learning_rate": 7.143438034623746e-06, + "loss": 0.3192, + "step": 17182 + }, + { + "epoch": 0.37823024433193925, + "grad_norm": 1.4807611764255606, + "learning_rate": 7.143115987668241e-06, + "loss": 0.3323, + "step": 17183 + }, + { + "epoch": 0.3782522562183579, + "grad_norm": 1.7854782500504667, + "learning_rate": 7.1427939298204665e-06, + "loss": 0.2993, + "step": 17184 + }, + { + "epoch": 0.3782742681047766, + "grad_norm": 2.1966513302078807, + "learning_rate": 7.142471861082056e-06, + "loss": 0.2747, + "step": 17185 + }, + { + "epoch": 0.37829627999119525, + "grad_norm": 1.90737123612295, + "learning_rate": 7.142149781454648e-06, + "loss": 0.3273, + "step": 17186 + }, + { + "epoch": 0.3783182918776139, + "grad_norm": 1.2890302937236018, + "learning_rate": 7.14182769093988e-06, + "loss": 0.2829, + "step": 17187 + }, + { + "epoch": 0.3783403037640326, + "grad_norm": 0.6470677201273739, + "learning_rate": 7.141505589539387e-06, + "loss": 0.4944, + "step": 17188 + }, + { + "epoch": 0.37836231565045125, + "grad_norm": 1.2636486251485954, + "learning_rate": 7.141183477254808e-06, + "loss": 0.2915, + "step": 17189 + }, + { + "epoch": 0.3783843275368699, + "grad_norm": 1.9159618230888835, + "learning_rate": 7.140861354087779e-06, + "loss": 0.403, + "step": 17190 + }, + { + "epoch": 0.37840633942328855, + "grad_norm": 1.41913295426419, + "learning_rate": 7.140539220039939e-06, + "loss": 0.3337, + "step": 17191 + }, + { + "epoch": 0.37842835130970726, + "grad_norm": 1.2418386636550967, + "learning_rate": 7.140217075112924e-06, + "loss": 0.3088, + "step": 17192 + }, + { + "epoch": 0.3784503631961259, + "grad_norm": 1.3907141953605133, + "learning_rate": 7.139894919308371e-06, + "loss": 0.309, + "step": 17193 + }, + { + "epoch": 0.37847237508254455, + "grad_norm": 0.6290574052833492, + "learning_rate": 7.139572752627918e-06, + "loss": 0.4969, + "step": 17194 + }, + { + "epoch": 0.37849438696896326, + "grad_norm": 1.4574974299172705, + "learning_rate": 7.1392505750732e-06, + "loss": 0.398, + "step": 17195 + }, + { + "epoch": 0.3785163988553819, + "grad_norm": 0.6437316313433019, + "learning_rate": 7.138928386645858e-06, + "loss": 0.4986, + "step": 17196 + }, + { + "epoch": 0.37853841074180056, + "grad_norm": 0.6318642409502675, + "learning_rate": 7.1386061873475285e-06, + "loss": 0.4911, + "step": 17197 + }, + { + "epoch": 0.37856042262821926, + "grad_norm": 1.4921911228282225, + "learning_rate": 7.138283977179848e-06, + "loss": 0.3225, + "step": 17198 + }, + { + "epoch": 0.3785824345146379, + "grad_norm": 1.7109799094745235, + "learning_rate": 7.137961756144454e-06, + "loss": 0.3588, + "step": 17199 + }, + { + "epoch": 0.37860444640105656, + "grad_norm": 0.6421158038840363, + "learning_rate": 7.137639524242985e-06, + "loss": 0.5149, + "step": 17200 + }, + { + "epoch": 0.3786264582874752, + "grad_norm": 2.084787958435974, + "learning_rate": 7.137317281477078e-06, + "loss": 0.3155, + "step": 17201 + }, + { + "epoch": 0.3786484701738939, + "grad_norm": 1.4928649723304073, + "learning_rate": 7.136995027848373e-06, + "loss": 0.3154, + "step": 17202 + }, + { + "epoch": 0.37867048206031256, + "grad_norm": 1.2015179229754207, + "learning_rate": 7.136672763358505e-06, + "loss": 0.3068, + "step": 17203 + }, + { + "epoch": 0.3786924939467312, + "grad_norm": 1.309487439835522, + "learning_rate": 7.136350488009114e-06, + "loss": 0.3018, + "step": 17204 + }, + { + "epoch": 0.3787145058331499, + "grad_norm": 1.4461288689378677, + "learning_rate": 7.136028201801835e-06, + "loss": 0.3863, + "step": 17205 + }, + { + "epoch": 0.37873651771956857, + "grad_norm": 1.3289892910401437, + "learning_rate": 7.135705904738311e-06, + "loss": 0.3464, + "step": 17206 + }, + { + "epoch": 0.3787585296059872, + "grad_norm": 1.1404651135448791, + "learning_rate": 7.135383596820175e-06, + "loss": 0.3546, + "step": 17207 + }, + { + "epoch": 0.3787805414924059, + "grad_norm": 1.7040049490235092, + "learning_rate": 7.135061278049067e-06, + "loss": 0.3599, + "step": 17208 + }, + { + "epoch": 0.37880255337882457, + "grad_norm": 3.0751313911706872, + "learning_rate": 7.134738948426625e-06, + "loss": 0.324, + "step": 17209 + }, + { + "epoch": 0.3788245652652432, + "grad_norm": 1.4204455864058343, + "learning_rate": 7.134416607954489e-06, + "loss": 0.3766, + "step": 17210 + }, + { + "epoch": 0.3788465771516619, + "grad_norm": 1.9062816039218136, + "learning_rate": 7.134094256634295e-06, + "loss": 0.2911, + "step": 17211 + }, + { + "epoch": 0.37886858903808057, + "grad_norm": 1.6136125631712364, + "learning_rate": 7.1337718944676826e-06, + "loss": 0.2782, + "step": 17212 + }, + { + "epoch": 0.3788906009244992, + "grad_norm": 2.030419818195085, + "learning_rate": 7.133449521456288e-06, + "loss": 0.2865, + "step": 17213 + }, + { + "epoch": 0.37891261281091787, + "grad_norm": 1.4415493380895394, + "learning_rate": 7.133127137601753e-06, + "loss": 0.248, + "step": 17214 + }, + { + "epoch": 0.3789346246973366, + "grad_norm": 1.6229239332778147, + "learning_rate": 7.1328047429057156e-06, + "loss": 0.3266, + "step": 17215 + }, + { + "epoch": 0.3789566365837552, + "grad_norm": 0.7354886908612023, + "learning_rate": 7.132482337369812e-06, + "loss": 0.5081, + "step": 17216 + }, + { + "epoch": 0.3789786484701739, + "grad_norm": 1.4349240997490686, + "learning_rate": 7.132159920995682e-06, + "loss": 0.3241, + "step": 17217 + }, + { + "epoch": 0.3790006603565926, + "grad_norm": 0.6594921399702081, + "learning_rate": 7.131837493784964e-06, + "loss": 0.5026, + "step": 17218 + }, + { + "epoch": 0.3790226722430112, + "grad_norm": 0.6366917626192813, + "learning_rate": 7.131515055739297e-06, + "loss": 0.5059, + "step": 17219 + }, + { + "epoch": 0.3790446841294299, + "grad_norm": 1.3637285610862024, + "learning_rate": 7.13119260686032e-06, + "loss": 0.3159, + "step": 17220 + }, + { + "epoch": 0.3790666960158486, + "grad_norm": 1.6528734335595772, + "learning_rate": 7.130870147149671e-06, + "loss": 0.3428, + "step": 17221 + }, + { + "epoch": 0.37908870790226723, + "grad_norm": 1.4676434265128464, + "learning_rate": 7.130547676608991e-06, + "loss": 0.3912, + "step": 17222 + }, + { + "epoch": 0.3791107197886859, + "grad_norm": 1.6783692179934644, + "learning_rate": 7.130225195239917e-06, + "loss": 0.4287, + "step": 17223 + }, + { + "epoch": 0.3791327316751046, + "grad_norm": 1.2225482426947891, + "learning_rate": 7.129902703044088e-06, + "loss": 0.259, + "step": 17224 + }, + { + "epoch": 0.37915474356152323, + "grad_norm": 0.6671994842410558, + "learning_rate": 7.129580200023144e-06, + "loss": 0.5017, + "step": 17225 + }, + { + "epoch": 0.3791767554479419, + "grad_norm": 0.6668261362425698, + "learning_rate": 7.129257686178722e-06, + "loss": 0.4963, + "step": 17226 + }, + { + "epoch": 0.37919876733436053, + "grad_norm": 1.546213277239104, + "learning_rate": 7.128935161512465e-06, + "loss": 0.3624, + "step": 17227 + }, + { + "epoch": 0.37922077922077924, + "grad_norm": 1.6777498824247583, + "learning_rate": 7.128612626026008e-06, + "loss": 0.315, + "step": 17228 + }, + { + "epoch": 0.3792427911071979, + "grad_norm": 1.426725210408156, + "learning_rate": 7.1282900797209946e-06, + "loss": 0.3388, + "step": 17229 + }, + { + "epoch": 0.37926480299361653, + "grad_norm": 4.481172301800063, + "learning_rate": 7.12796752259906e-06, + "loss": 0.3737, + "step": 17230 + }, + { + "epoch": 0.37928681488003524, + "grad_norm": 1.5727310553316016, + "learning_rate": 7.127644954661846e-06, + "loss": 0.3996, + "step": 17231 + }, + { + "epoch": 0.3793088267664539, + "grad_norm": 1.444638225375246, + "learning_rate": 7.12732237591099e-06, + "loss": 0.2975, + "step": 17232 + }, + { + "epoch": 0.37933083865287254, + "grad_norm": 2.551581051942691, + "learning_rate": 7.126999786348134e-06, + "loss": 0.2926, + "step": 17233 + }, + { + "epoch": 0.37935285053929124, + "grad_norm": 2.3007540732813108, + "learning_rate": 7.126677185974918e-06, + "loss": 0.3786, + "step": 17234 + }, + { + "epoch": 0.3793748624257099, + "grad_norm": 1.5460748751201752, + "learning_rate": 7.126354574792978e-06, + "loss": 0.3327, + "step": 17235 + }, + { + "epoch": 0.37939687431212854, + "grad_norm": 1.303031432561941, + "learning_rate": 7.126031952803955e-06, + "loss": 0.3358, + "step": 17236 + }, + { + "epoch": 0.3794188861985472, + "grad_norm": 0.7581125920468557, + "learning_rate": 7.12570932000949e-06, + "loss": 0.5284, + "step": 17237 + }, + { + "epoch": 0.3794408980849659, + "grad_norm": 0.6840362480986906, + "learning_rate": 7.125386676411224e-06, + "loss": 0.5019, + "step": 17238 + }, + { + "epoch": 0.37946290997138454, + "grad_norm": 1.2493534452361659, + "learning_rate": 7.125064022010794e-06, + "loss": 0.2771, + "step": 17239 + }, + { + "epoch": 0.3794849218578032, + "grad_norm": 2.6376812556752656, + "learning_rate": 7.12474135680984e-06, + "loss": 0.3246, + "step": 17240 + }, + { + "epoch": 0.3795069337442219, + "grad_norm": 1.9805497655771187, + "learning_rate": 7.124418680810003e-06, + "loss": 0.2586, + "step": 17241 + }, + { + "epoch": 0.37952894563064055, + "grad_norm": 1.9692538491300793, + "learning_rate": 7.124095994012923e-06, + "loss": 0.3211, + "step": 17242 + }, + { + "epoch": 0.3795509575170592, + "grad_norm": 0.7021453181553733, + "learning_rate": 7.12377329642024e-06, + "loss": 0.5264, + "step": 17243 + }, + { + "epoch": 0.3795729694034779, + "grad_norm": 2.1445861270911157, + "learning_rate": 7.123450588033594e-06, + "loss": 0.2831, + "step": 17244 + }, + { + "epoch": 0.37959498128989655, + "grad_norm": 0.7293204872402148, + "learning_rate": 7.1231278688546235e-06, + "loss": 0.4792, + "step": 17245 + }, + { + "epoch": 0.3796169931763152, + "grad_norm": 1.2407456844731832, + "learning_rate": 7.122805138884972e-06, + "loss": 0.3085, + "step": 17246 + }, + { + "epoch": 0.3796390050627339, + "grad_norm": 1.6753446461380241, + "learning_rate": 7.122482398126278e-06, + "loss": 0.256, + "step": 17247 + }, + { + "epoch": 0.37966101694915255, + "grad_norm": 1.4688308479329748, + "learning_rate": 7.12215964658018e-06, + "loss": 0.3764, + "step": 17248 + }, + { + "epoch": 0.3796830288355712, + "grad_norm": 2.0493078242453193, + "learning_rate": 7.121836884248322e-06, + "loss": 0.3652, + "step": 17249 + }, + { + "epoch": 0.37970504072198985, + "grad_norm": 0.6318092205846909, + "learning_rate": 7.121514111132342e-06, + "loss": 0.4884, + "step": 17250 + }, + { + "epoch": 0.37972705260840856, + "grad_norm": 2.2636852445700457, + "learning_rate": 7.12119132723388e-06, + "loss": 0.3227, + "step": 17251 + }, + { + "epoch": 0.3797490644948272, + "grad_norm": 1.3527789024812418, + "learning_rate": 7.120868532554579e-06, + "loss": 0.2531, + "step": 17252 + }, + { + "epoch": 0.37977107638124585, + "grad_norm": 1.5708620521322407, + "learning_rate": 7.1205457270960775e-06, + "loss": 0.3552, + "step": 17253 + }, + { + "epoch": 0.37979308826766456, + "grad_norm": 0.6464949244336093, + "learning_rate": 7.1202229108600175e-06, + "loss": 0.4859, + "step": 17254 + }, + { + "epoch": 0.3798151001540832, + "grad_norm": 1.3308551785626348, + "learning_rate": 7.119900083848038e-06, + "loss": 0.3636, + "step": 17255 + }, + { + "epoch": 0.37983711204050186, + "grad_norm": 0.629795373930596, + "learning_rate": 7.119577246061783e-06, + "loss": 0.5311, + "step": 17256 + }, + { + "epoch": 0.37985912392692056, + "grad_norm": 1.4773258287875255, + "learning_rate": 7.119254397502888e-06, + "loss": 0.3314, + "step": 17257 + }, + { + "epoch": 0.3798811358133392, + "grad_norm": 1.6756480122682575, + "learning_rate": 7.118931538172999e-06, + "loss": 0.4052, + "step": 17258 + }, + { + "epoch": 0.37990314769975786, + "grad_norm": 1.5551114705102969, + "learning_rate": 7.118608668073754e-06, + "loss": 0.3157, + "step": 17259 + }, + { + "epoch": 0.3799251595861765, + "grad_norm": 1.8271902861378713, + "learning_rate": 7.118285787206795e-06, + "loss": 0.2867, + "step": 17260 + }, + { + "epoch": 0.3799471714725952, + "grad_norm": 1.2420620080660876, + "learning_rate": 7.117962895573763e-06, + "loss": 0.3427, + "step": 17261 + }, + { + "epoch": 0.37996918335901386, + "grad_norm": 1.3146197395926544, + "learning_rate": 7.117639993176299e-06, + "loss": 0.2915, + "step": 17262 + }, + { + "epoch": 0.3799911952454325, + "grad_norm": 1.6154518100042163, + "learning_rate": 7.117317080016044e-06, + "loss": 0.3384, + "step": 17263 + }, + { + "epoch": 0.3800132071318512, + "grad_norm": 1.462175857504191, + "learning_rate": 7.116994156094638e-06, + "loss": 0.2518, + "step": 17264 + }, + { + "epoch": 0.38003521901826987, + "grad_norm": 0.6304823234662481, + "learning_rate": 7.116671221413725e-06, + "loss": 0.4646, + "step": 17265 + }, + { + "epoch": 0.3800572309046885, + "grad_norm": 1.693075000133773, + "learning_rate": 7.116348275974946e-06, + "loss": 0.2845, + "step": 17266 + }, + { + "epoch": 0.3800792427911072, + "grad_norm": 2.4577156869187733, + "learning_rate": 7.116025319779939e-06, + "loss": 0.3819, + "step": 17267 + }, + { + "epoch": 0.38010125467752587, + "grad_norm": 0.6084899449243045, + "learning_rate": 7.115702352830347e-06, + "loss": 0.5027, + "step": 17268 + }, + { + "epoch": 0.3801232665639445, + "grad_norm": 1.0935681174434664, + "learning_rate": 7.115379375127813e-06, + "loss": 0.2419, + "step": 17269 + }, + { + "epoch": 0.3801452784503632, + "grad_norm": 1.4536639051350333, + "learning_rate": 7.115056386673979e-06, + "loss": 0.3587, + "step": 17270 + }, + { + "epoch": 0.38016729033678187, + "grad_norm": 0.6192837791795378, + "learning_rate": 7.1147333874704835e-06, + "loss": 0.4851, + "step": 17271 + }, + { + "epoch": 0.3801893022232005, + "grad_norm": 2.2709654517588174, + "learning_rate": 7.11441037751897e-06, + "loss": 0.3039, + "step": 17272 + }, + { + "epoch": 0.38021131410961917, + "grad_norm": 0.6038588727363832, + "learning_rate": 7.11408735682108e-06, + "loss": 0.4925, + "step": 17273 + }, + { + "epoch": 0.3802333259960379, + "grad_norm": 1.6173135762009752, + "learning_rate": 7.1137643253784565e-06, + "loss": 0.3655, + "step": 17274 + }, + { + "epoch": 0.3802553378824565, + "grad_norm": 0.6202274233194367, + "learning_rate": 7.113441283192737e-06, + "loss": 0.5003, + "step": 17275 + }, + { + "epoch": 0.3802773497688752, + "grad_norm": 1.324107367597156, + "learning_rate": 7.113118230265568e-06, + "loss": 0.2937, + "step": 17276 + }, + { + "epoch": 0.3802993616552939, + "grad_norm": 1.5049755659772708, + "learning_rate": 7.112795166598589e-06, + "loss": 0.3704, + "step": 17277 + }, + { + "epoch": 0.3803213735417125, + "grad_norm": 2.4297748607507086, + "learning_rate": 7.112472092193443e-06, + "loss": 0.3424, + "step": 17278 + }, + { + "epoch": 0.3803433854281312, + "grad_norm": 1.472749044872322, + "learning_rate": 7.112149007051773e-06, + "loss": 0.3058, + "step": 17279 + }, + { + "epoch": 0.3803653973145499, + "grad_norm": 1.2730993002067514, + "learning_rate": 7.111825911175217e-06, + "loss": 0.2247, + "step": 17280 + }, + { + "epoch": 0.38038740920096853, + "grad_norm": 2.226610090897932, + "learning_rate": 7.1115028045654225e-06, + "loss": 0.2647, + "step": 17281 + }, + { + "epoch": 0.3804094210873872, + "grad_norm": 1.2359356556667593, + "learning_rate": 7.111179687224027e-06, + "loss": 0.3288, + "step": 17282 + }, + { + "epoch": 0.3804314329738058, + "grad_norm": 1.6847111764484866, + "learning_rate": 7.110856559152675e-06, + "loss": 0.3007, + "step": 17283 + }, + { + "epoch": 0.38045344486022453, + "grad_norm": 1.5667544762367862, + "learning_rate": 7.11053342035301e-06, + "loss": 0.397, + "step": 17284 + }, + { + "epoch": 0.3804754567466432, + "grad_norm": 1.6979400078146918, + "learning_rate": 7.1102102708266714e-06, + "loss": 0.3134, + "step": 17285 + }, + { + "epoch": 0.38049746863306183, + "grad_norm": 1.6912061429231562, + "learning_rate": 7.1098871105753034e-06, + "loss": 0.2761, + "step": 17286 + }, + { + "epoch": 0.38051948051948054, + "grad_norm": 0.7082029742888023, + "learning_rate": 7.1095639396005475e-06, + "loss": 0.4879, + "step": 17287 + }, + { + "epoch": 0.3805414924058992, + "grad_norm": 2.8552787609442314, + "learning_rate": 7.1092407579040476e-06, + "loss": 0.3725, + "step": 17288 + }, + { + "epoch": 0.38056350429231783, + "grad_norm": 1.7507311296753252, + "learning_rate": 7.108917565487445e-06, + "loss": 0.3548, + "step": 17289 + }, + { + "epoch": 0.38058551617873654, + "grad_norm": 1.39021172362148, + "learning_rate": 7.108594362352383e-06, + "loss": 0.2725, + "step": 17290 + }, + { + "epoch": 0.3806075280651552, + "grad_norm": 1.9632796357467464, + "learning_rate": 7.108271148500503e-06, + "loss": 0.2316, + "step": 17291 + }, + { + "epoch": 0.38062953995157384, + "grad_norm": 1.294200818089542, + "learning_rate": 7.10794792393345e-06, + "loss": 0.348, + "step": 17292 + }, + { + "epoch": 0.38065155183799254, + "grad_norm": 1.6091376836313571, + "learning_rate": 7.107624688652865e-06, + "loss": 0.2291, + "step": 17293 + }, + { + "epoch": 0.3806735637244112, + "grad_norm": 1.8577695852571847, + "learning_rate": 7.107301442660392e-06, + "loss": 0.3759, + "step": 17294 + }, + { + "epoch": 0.38069557561082984, + "grad_norm": 0.6561136365828981, + "learning_rate": 7.106978185957672e-06, + "loss": 0.5127, + "step": 17295 + }, + { + "epoch": 0.3807175874972485, + "grad_norm": 1.5288662431249023, + "learning_rate": 7.106654918546349e-06, + "loss": 0.3635, + "step": 17296 + }, + { + "epoch": 0.3807395993836672, + "grad_norm": 1.2956636586459096, + "learning_rate": 7.106331640428068e-06, + "loss": 0.3544, + "step": 17297 + }, + { + "epoch": 0.38076161127008584, + "grad_norm": 0.6040567470270647, + "learning_rate": 7.106008351604469e-06, + "loss": 0.4941, + "step": 17298 + }, + { + "epoch": 0.3807836231565045, + "grad_norm": 1.741175449432893, + "learning_rate": 7.105685052077196e-06, + "loss": 0.3333, + "step": 17299 + }, + { + "epoch": 0.3808056350429232, + "grad_norm": 0.596393689744981, + "learning_rate": 7.1053617418478925e-06, + "loss": 0.4856, + "step": 17300 + }, + { + "epoch": 0.38082764692934185, + "grad_norm": 1.206704648460182, + "learning_rate": 7.105038420918201e-06, + "loss": 0.3067, + "step": 17301 + }, + { + "epoch": 0.3808496588157605, + "grad_norm": 0.5949914753552942, + "learning_rate": 7.104715089289767e-06, + "loss": 0.4771, + "step": 17302 + }, + { + "epoch": 0.3808716707021792, + "grad_norm": 1.6832109863730056, + "learning_rate": 7.104391746964232e-06, + "loss": 0.3296, + "step": 17303 + }, + { + "epoch": 0.38089368258859785, + "grad_norm": 1.4940231454438953, + "learning_rate": 7.1040683939432376e-06, + "loss": 0.317, + "step": 17304 + }, + { + "epoch": 0.3809156944750165, + "grad_norm": 0.6377775757016282, + "learning_rate": 7.103745030228431e-06, + "loss": 0.5106, + "step": 17305 + }, + { + "epoch": 0.3809377063614352, + "grad_norm": 2.355103231177144, + "learning_rate": 7.103421655821453e-06, + "loss": 0.3407, + "step": 17306 + }, + { + "epoch": 0.38095971824785385, + "grad_norm": 1.5699784930749592, + "learning_rate": 7.1030982707239486e-06, + "loss": 0.3381, + "step": 17307 + }, + { + "epoch": 0.3809817301342725, + "grad_norm": 2.067117605671575, + "learning_rate": 7.1027748749375606e-06, + "loss": 0.2726, + "step": 17308 + }, + { + "epoch": 0.38100374202069115, + "grad_norm": 0.6320356982745908, + "learning_rate": 7.102451468463931e-06, + "loss": 0.4991, + "step": 17309 + }, + { + "epoch": 0.38102575390710985, + "grad_norm": 1.4511320548271212, + "learning_rate": 7.102128051304708e-06, + "loss": 0.2636, + "step": 17310 + }, + { + "epoch": 0.3810477657935285, + "grad_norm": 0.6424267132228274, + "learning_rate": 7.101804623461532e-06, + "loss": 0.4857, + "step": 17311 + }, + { + "epoch": 0.38106977767994715, + "grad_norm": 1.5605299703053481, + "learning_rate": 7.101481184936048e-06, + "loss": 0.3017, + "step": 17312 + }, + { + "epoch": 0.38109178956636586, + "grad_norm": 1.549599882598224, + "learning_rate": 7.101157735729898e-06, + "loss": 0.3125, + "step": 17313 + }, + { + "epoch": 0.3811138014527845, + "grad_norm": 1.4327890343796719, + "learning_rate": 7.100834275844727e-06, + "loss": 0.3782, + "step": 17314 + }, + { + "epoch": 0.38113581333920316, + "grad_norm": 1.6104886242875578, + "learning_rate": 7.10051080528218e-06, + "loss": 0.2584, + "step": 17315 + }, + { + "epoch": 0.38115782522562186, + "grad_norm": 0.6375913354174683, + "learning_rate": 7.1001873240439004e-06, + "loss": 0.4959, + "step": 17316 + }, + { + "epoch": 0.3811798371120405, + "grad_norm": 1.5027898178711978, + "learning_rate": 7.099863832131532e-06, + "loss": 0.3247, + "step": 17317 + }, + { + "epoch": 0.38120184899845916, + "grad_norm": 1.4915984464209544, + "learning_rate": 7.099540329546719e-06, + "loss": 0.3107, + "step": 17318 + }, + { + "epoch": 0.3812238608848778, + "grad_norm": 1.6517764146308578, + "learning_rate": 7.099216816291104e-06, + "loss": 0.2606, + "step": 17319 + }, + { + "epoch": 0.3812458727712965, + "grad_norm": 1.697816610251985, + "learning_rate": 7.0988932923663355e-06, + "loss": 0.3291, + "step": 17320 + }, + { + "epoch": 0.38126788465771516, + "grad_norm": 1.483007824025454, + "learning_rate": 7.098569757774054e-06, + "loss": 0.3309, + "step": 17321 + }, + { + "epoch": 0.3812898965441338, + "grad_norm": 0.620750488992707, + "learning_rate": 7.0982462125159045e-06, + "loss": 0.4934, + "step": 17322 + }, + { + "epoch": 0.3813119084305525, + "grad_norm": 1.7088233204816439, + "learning_rate": 7.097922656593532e-06, + "loss": 0.3059, + "step": 17323 + }, + { + "epoch": 0.38133392031697116, + "grad_norm": 1.2512919737830561, + "learning_rate": 7.0975990900085825e-06, + "loss": 0.3834, + "step": 17324 + }, + { + "epoch": 0.3813559322033898, + "grad_norm": 1.687528792185676, + "learning_rate": 7.097275512762698e-06, + "loss": 0.3413, + "step": 17325 + }, + { + "epoch": 0.3813779440898085, + "grad_norm": 1.6043516138278977, + "learning_rate": 7.096951924857524e-06, + "loss": 0.3158, + "step": 17326 + }, + { + "epoch": 0.38139995597622717, + "grad_norm": 0.6189647565141005, + "learning_rate": 7.0966283262947035e-06, + "loss": 0.5034, + "step": 17327 + }, + { + "epoch": 0.3814219678626458, + "grad_norm": 1.344910391629211, + "learning_rate": 7.096304717075883e-06, + "loss": 0.2361, + "step": 17328 + }, + { + "epoch": 0.3814439797490645, + "grad_norm": 20.797790513867675, + "learning_rate": 7.095981097202708e-06, + "loss": 0.3598, + "step": 17329 + }, + { + "epoch": 0.38146599163548317, + "grad_norm": 2.7431408062865072, + "learning_rate": 7.095657466676823e-06, + "loss": 0.3533, + "step": 17330 + }, + { + "epoch": 0.3814880035219018, + "grad_norm": 1.6082606759720586, + "learning_rate": 7.09533382549987e-06, + "loss": 0.2913, + "step": 17331 + }, + { + "epoch": 0.38151001540832047, + "grad_norm": 1.298438308355301, + "learning_rate": 7.095010173673498e-06, + "loss": 0.3078, + "step": 17332 + }, + { + "epoch": 0.3815320272947392, + "grad_norm": 1.3844079431244332, + "learning_rate": 7.094686511199349e-06, + "loss": 0.3203, + "step": 17333 + }, + { + "epoch": 0.3815540391811578, + "grad_norm": 1.8780496739521688, + "learning_rate": 7.094362838079068e-06, + "loss": 0.2874, + "step": 17334 + }, + { + "epoch": 0.38157605106757647, + "grad_norm": 1.455047785052789, + "learning_rate": 7.094039154314303e-06, + "loss": 0.3701, + "step": 17335 + }, + { + "epoch": 0.3815980629539952, + "grad_norm": 1.4905860732202771, + "learning_rate": 7.093715459906696e-06, + "loss": 0.3744, + "step": 17336 + }, + { + "epoch": 0.3816200748404138, + "grad_norm": 0.653394191746682, + "learning_rate": 7.093391754857892e-06, + "loss": 0.4971, + "step": 17337 + }, + { + "epoch": 0.3816420867268325, + "grad_norm": 2.8372184034966255, + "learning_rate": 7.093068039169539e-06, + "loss": 0.3913, + "step": 17338 + }, + { + "epoch": 0.3816640986132512, + "grad_norm": 1.3843840469657531, + "learning_rate": 7.09274431284328e-06, + "loss": 0.3584, + "step": 17339 + }, + { + "epoch": 0.38168611049966983, + "grad_norm": 1.1592061317700113, + "learning_rate": 7.092420575880761e-06, + "loss": 0.1704, + "step": 17340 + }, + { + "epoch": 0.3817081223860885, + "grad_norm": 1.5537028482486126, + "learning_rate": 7.0920968282836275e-06, + "loss": 0.3322, + "step": 17341 + }, + { + "epoch": 0.3817301342725071, + "grad_norm": 1.4030297568763435, + "learning_rate": 7.091773070053524e-06, + "loss": 0.3032, + "step": 17342 + }, + { + "epoch": 0.38175214615892583, + "grad_norm": 1.6545700918005533, + "learning_rate": 7.091449301192099e-06, + "loss": 0.3172, + "step": 17343 + }, + { + "epoch": 0.3817741580453445, + "grad_norm": 1.2419786901685994, + "learning_rate": 7.091125521700995e-06, + "loss": 0.2308, + "step": 17344 + }, + { + "epoch": 0.38179616993176313, + "grad_norm": 1.523506338093274, + "learning_rate": 7.090801731581857e-06, + "loss": 0.2326, + "step": 17345 + }, + { + "epoch": 0.38181818181818183, + "grad_norm": 1.5924341787570104, + "learning_rate": 7.0904779308363336e-06, + "loss": 0.3208, + "step": 17346 + }, + { + "epoch": 0.3818401937046005, + "grad_norm": 2.0292166325367957, + "learning_rate": 7.0901541194660685e-06, + "loss": 0.3721, + "step": 17347 + }, + { + "epoch": 0.38186220559101913, + "grad_norm": 1.1033446813550614, + "learning_rate": 7.089830297472708e-06, + "loss": 0.3553, + "step": 17348 + }, + { + "epoch": 0.38188421747743784, + "grad_norm": 0.6174390443249609, + "learning_rate": 7.089506464857898e-06, + "loss": 0.5019, + "step": 17349 + }, + { + "epoch": 0.3819062293638565, + "grad_norm": 1.5267704246650111, + "learning_rate": 7.089182621623283e-06, + "loss": 0.3261, + "step": 17350 + }, + { + "epoch": 0.38192824125027514, + "grad_norm": 0.6393409814543436, + "learning_rate": 7.088858767770511e-06, + "loss": 0.537, + "step": 17351 + }, + { + "epoch": 0.38195025313669384, + "grad_norm": 1.5102077377038803, + "learning_rate": 7.088534903301227e-06, + "loss": 0.3065, + "step": 17352 + }, + { + "epoch": 0.3819722650231125, + "grad_norm": 2.326990228519882, + "learning_rate": 7.0882110282170775e-06, + "loss": 0.3756, + "step": 17353 + }, + { + "epoch": 0.38199427690953114, + "grad_norm": 2.34644923554596, + "learning_rate": 7.087887142519709e-06, + "loss": 0.341, + "step": 17354 + }, + { + "epoch": 0.3820162887959498, + "grad_norm": 1.4952899986801163, + "learning_rate": 7.087563246210763e-06, + "loss": 0.3368, + "step": 17355 + }, + { + "epoch": 0.3820383006823685, + "grad_norm": 0.6486318501201016, + "learning_rate": 7.087239339291894e-06, + "loss": 0.5204, + "step": 17356 + }, + { + "epoch": 0.38206031256878714, + "grad_norm": 0.6299227427331714, + "learning_rate": 7.086915421764742e-06, + "loss": 0.5035, + "step": 17357 + }, + { + "epoch": 0.3820823244552058, + "grad_norm": 2.5313197179727394, + "learning_rate": 7.086591493630954e-06, + "loss": 0.3229, + "step": 17358 + }, + { + "epoch": 0.3821043363416245, + "grad_norm": 1.5201234186404875, + "learning_rate": 7.086267554892179e-06, + "loss": 0.3641, + "step": 17359 + }, + { + "epoch": 0.38212634822804314, + "grad_norm": 0.619886654985634, + "learning_rate": 7.0859436055500605e-06, + "loss": 0.5071, + "step": 17360 + }, + { + "epoch": 0.3821483601144618, + "grad_norm": 0.6052311314107623, + "learning_rate": 7.085619645606247e-06, + "loss": 0.4943, + "step": 17361 + }, + { + "epoch": 0.3821703720008805, + "grad_norm": 2.324083911264664, + "learning_rate": 7.085295675062385e-06, + "loss": 0.3339, + "step": 17362 + }, + { + "epoch": 0.38219238388729915, + "grad_norm": 0.6340653687742841, + "learning_rate": 7.084971693920119e-06, + "loss": 0.5079, + "step": 17363 + }, + { + "epoch": 0.3822143957737178, + "grad_norm": 1.915438371975488, + "learning_rate": 7.084647702181098e-06, + "loss": 0.3048, + "step": 17364 + }, + { + "epoch": 0.3822364076601365, + "grad_norm": 1.8476229526421353, + "learning_rate": 7.084323699846967e-06, + "loss": 0.4052, + "step": 17365 + }, + { + "epoch": 0.38225841954655515, + "grad_norm": 1.5918372920053656, + "learning_rate": 7.083999686919373e-06, + "loss": 0.288, + "step": 17366 + }, + { + "epoch": 0.3822804314329738, + "grad_norm": 1.7119588368908918, + "learning_rate": 7.083675663399964e-06, + "loss": 0.3745, + "step": 17367 + }, + { + "epoch": 0.38230244331939245, + "grad_norm": 1.2508320581265755, + "learning_rate": 7.0833516292903845e-06, + "loss": 0.3066, + "step": 17368 + }, + { + "epoch": 0.38232445520581115, + "grad_norm": 2.0980266051458543, + "learning_rate": 7.0830275845922834e-06, + "loss": 0.3641, + "step": 17369 + }, + { + "epoch": 0.3823464670922298, + "grad_norm": 1.4041286331538498, + "learning_rate": 7.082703529307307e-06, + "loss": 0.3136, + "step": 17370 + }, + { + "epoch": 0.38236847897864845, + "grad_norm": 0.6270617031448733, + "learning_rate": 7.082379463437104e-06, + "loss": 0.5061, + "step": 17371 + }, + { + "epoch": 0.38239049086506716, + "grad_norm": 1.5494374326031293, + "learning_rate": 7.082055386983317e-06, + "loss": 0.3757, + "step": 17372 + }, + { + "epoch": 0.3824125027514858, + "grad_norm": 1.3478960205884147, + "learning_rate": 7.081731299947597e-06, + "loss": 0.2809, + "step": 17373 + }, + { + "epoch": 0.38243451463790445, + "grad_norm": 1.4739978837963206, + "learning_rate": 7.0814072023315915e-06, + "loss": 0.3115, + "step": 17374 + }, + { + "epoch": 0.38245652652432316, + "grad_norm": 1.5505650188391105, + "learning_rate": 7.0810830941369455e-06, + "loss": 0.3006, + "step": 17375 + }, + { + "epoch": 0.3824785384107418, + "grad_norm": 0.6341319833832714, + "learning_rate": 7.080758975365306e-06, + "loss": 0.5196, + "step": 17376 + }, + { + "epoch": 0.38250055029716046, + "grad_norm": 1.617952355095707, + "learning_rate": 7.080434846018323e-06, + "loss": 0.3288, + "step": 17377 + }, + { + "epoch": 0.3825225621835791, + "grad_norm": 1.4141635709907532, + "learning_rate": 7.080110706097641e-06, + "loss": 0.405, + "step": 17378 + }, + { + "epoch": 0.3825445740699978, + "grad_norm": 1.6077361077323726, + "learning_rate": 7.0797865556049094e-06, + "loss": 0.3191, + "step": 17379 + }, + { + "epoch": 0.38256658595641646, + "grad_norm": 1.938553586061553, + "learning_rate": 7.079462394541774e-06, + "loss": 0.3035, + "step": 17380 + }, + { + "epoch": 0.3825885978428351, + "grad_norm": 1.6461877353920522, + "learning_rate": 7.0791382229098824e-06, + "loss": 0.3603, + "step": 17381 + }, + { + "epoch": 0.3826106097292538, + "grad_norm": 2.298852951659676, + "learning_rate": 7.078814040710885e-06, + "loss": 0.2832, + "step": 17382 + }, + { + "epoch": 0.38263262161567246, + "grad_norm": 1.9729836176894562, + "learning_rate": 7.078489847946427e-06, + "loss": 0.3535, + "step": 17383 + }, + { + "epoch": 0.3826546335020911, + "grad_norm": 16.96126538900951, + "learning_rate": 7.078165644618157e-06, + "loss": 0.2952, + "step": 17384 + }, + { + "epoch": 0.3826766453885098, + "grad_norm": 1.5707480585505134, + "learning_rate": 7.077841430727721e-06, + "loss": 0.2803, + "step": 17385 + }, + { + "epoch": 0.38269865727492847, + "grad_norm": 2.542067858816969, + "learning_rate": 7.077517206276768e-06, + "loss": 0.31, + "step": 17386 + }, + { + "epoch": 0.3827206691613471, + "grad_norm": 1.4053327623367649, + "learning_rate": 7.077192971266947e-06, + "loss": 0.2665, + "step": 17387 + }, + { + "epoch": 0.3827426810477658, + "grad_norm": 1.7853579033080316, + "learning_rate": 7.076868725699905e-06, + "loss": 0.3501, + "step": 17388 + }, + { + "epoch": 0.38276469293418447, + "grad_norm": 0.6304618416885558, + "learning_rate": 7.076544469577289e-06, + "loss": 0.5128, + "step": 17389 + }, + { + "epoch": 0.3827867048206031, + "grad_norm": 0.6142417611975314, + "learning_rate": 7.076220202900749e-06, + "loss": 0.4996, + "step": 17390 + }, + { + "epoch": 0.38280871670702177, + "grad_norm": 1.7659662687033175, + "learning_rate": 7.07589592567193e-06, + "loss": 0.3372, + "step": 17391 + }, + { + "epoch": 0.3828307285934405, + "grad_norm": 1.3681284737613424, + "learning_rate": 7.075571637892483e-06, + "loss": 0.3231, + "step": 17392 + }, + { + "epoch": 0.3828527404798591, + "grad_norm": 1.7898205367719502, + "learning_rate": 7.075247339564057e-06, + "loss": 0.2865, + "step": 17393 + }, + { + "epoch": 0.38287475236627777, + "grad_norm": 1.6845443566958804, + "learning_rate": 7.074923030688296e-06, + "loss": 0.2855, + "step": 17394 + }, + { + "epoch": 0.3828967642526965, + "grad_norm": 2.125403251036601, + "learning_rate": 7.074598711266851e-06, + "loss": 0.4038, + "step": 17395 + }, + { + "epoch": 0.3829187761391151, + "grad_norm": 1.6417561412352475, + "learning_rate": 7.074274381301371e-06, + "loss": 0.3781, + "step": 17396 + }, + { + "epoch": 0.3829407880255338, + "grad_norm": 2.4021764711626767, + "learning_rate": 7.073950040793503e-06, + "loss": 0.2493, + "step": 17397 + }, + { + "epoch": 0.3829627999119525, + "grad_norm": 1.177876843319399, + "learning_rate": 7.073625689744897e-06, + "loss": 0.31, + "step": 17398 + }, + { + "epoch": 0.3829848117983711, + "grad_norm": 1.3467538503476097, + "learning_rate": 7.0733013281571986e-06, + "loss": 0.244, + "step": 17399 + }, + { + "epoch": 0.3830068236847898, + "grad_norm": 0.6936925471431209, + "learning_rate": 7.072976956032058e-06, + "loss": 0.516, + "step": 17400 + }, + { + "epoch": 0.3830288355712084, + "grad_norm": 1.3000324367073766, + "learning_rate": 7.072652573371126e-06, + "loss": 0.2752, + "step": 17401 + }, + { + "epoch": 0.38305084745762713, + "grad_norm": 2.158876320042566, + "learning_rate": 7.072328180176047e-06, + "loss": 0.321, + "step": 17402 + }, + { + "epoch": 0.3830728593440458, + "grad_norm": 2.121480858007537, + "learning_rate": 7.072003776448473e-06, + "loss": 0.3823, + "step": 17403 + }, + { + "epoch": 0.38309487123046443, + "grad_norm": 5.989126546069515, + "learning_rate": 7.071679362190051e-06, + "loss": 0.3254, + "step": 17404 + }, + { + "epoch": 0.38311688311688313, + "grad_norm": 2.2569090917712558, + "learning_rate": 7.071354937402431e-06, + "loss": 0.3801, + "step": 17405 + }, + { + "epoch": 0.3831388950033018, + "grad_norm": 5.718721007254126, + "learning_rate": 7.071030502087262e-06, + "loss": 0.2777, + "step": 17406 + }, + { + "epoch": 0.38316090688972043, + "grad_norm": 1.3927159621370957, + "learning_rate": 7.070706056246192e-06, + "loss": 0.2749, + "step": 17407 + }, + { + "epoch": 0.38318291877613914, + "grad_norm": 0.6277136011086223, + "learning_rate": 7.07038159988087e-06, + "loss": 0.4849, + "step": 17408 + }, + { + "epoch": 0.3832049306625578, + "grad_norm": 1.2671898111864348, + "learning_rate": 7.070057132992945e-06, + "loss": 0.3195, + "step": 17409 + }, + { + "epoch": 0.38322694254897643, + "grad_norm": 1.556991450105587, + "learning_rate": 7.069732655584066e-06, + "loss": 0.3036, + "step": 17410 + }, + { + "epoch": 0.38324895443539514, + "grad_norm": 0.6335830586220923, + "learning_rate": 7.069408167655885e-06, + "loss": 0.4933, + "step": 17411 + }, + { + "epoch": 0.3832709663218138, + "grad_norm": 1.3429226156988818, + "learning_rate": 7.069083669210045e-06, + "loss": 0.3319, + "step": 17412 + }, + { + "epoch": 0.38329297820823244, + "grad_norm": 0.6084895879856866, + "learning_rate": 7.068759160248201e-06, + "loss": 0.4947, + "step": 17413 + }, + { + "epoch": 0.3833149900946511, + "grad_norm": 2.342031780296288, + "learning_rate": 7.0684346407720005e-06, + "loss": 0.2817, + "step": 17414 + }, + { + "epoch": 0.3833370019810698, + "grad_norm": 1.421118511340764, + "learning_rate": 7.068110110783092e-06, + "loss": 0.3585, + "step": 17415 + }, + { + "epoch": 0.38335901386748844, + "grad_norm": 1.373616600137706, + "learning_rate": 7.067785570283125e-06, + "loss": 0.3266, + "step": 17416 + }, + { + "epoch": 0.3833810257539071, + "grad_norm": 1.549874881092643, + "learning_rate": 7.06746101927375e-06, + "loss": 0.3279, + "step": 17417 + }, + { + "epoch": 0.3834030376403258, + "grad_norm": 0.6433911746852646, + "learning_rate": 7.067136457756616e-06, + "loss": 0.5265, + "step": 17418 + }, + { + "epoch": 0.38342504952674444, + "grad_norm": 1.270000260940432, + "learning_rate": 7.0668118857333735e-06, + "loss": 0.4093, + "step": 17419 + }, + { + "epoch": 0.3834470614131631, + "grad_norm": 0.6116747606363226, + "learning_rate": 7.06648730320567e-06, + "loss": 0.4585, + "step": 17420 + }, + { + "epoch": 0.3834690732995818, + "grad_norm": 1.6901316071459063, + "learning_rate": 7.066162710175157e-06, + "loss": 0.3065, + "step": 17421 + }, + { + "epoch": 0.38349108518600045, + "grad_norm": 0.6138427972915425, + "learning_rate": 7.065838106643483e-06, + "loss": 0.4816, + "step": 17422 + }, + { + "epoch": 0.3835130970724191, + "grad_norm": 1.3649329331701463, + "learning_rate": 7.065513492612298e-06, + "loss": 0.2119, + "step": 17423 + }, + { + "epoch": 0.38353510895883774, + "grad_norm": 1.5603079063271055, + "learning_rate": 7.065188868083253e-06, + "loss": 0.4037, + "step": 17424 + }, + { + "epoch": 0.38355712084525645, + "grad_norm": 1.2698445552588946, + "learning_rate": 7.064864233057997e-06, + "loss": 0.2586, + "step": 17425 + }, + { + "epoch": 0.3835791327316751, + "grad_norm": 0.6459767343997921, + "learning_rate": 7.064539587538181e-06, + "loss": 0.4988, + "step": 17426 + }, + { + "epoch": 0.38360114461809375, + "grad_norm": 1.208652907551593, + "learning_rate": 7.064214931525452e-06, + "loss": 0.3566, + "step": 17427 + }, + { + "epoch": 0.38362315650451245, + "grad_norm": 1.6184030609748594, + "learning_rate": 7.063890265021462e-06, + "loss": 0.3794, + "step": 17428 + }, + { + "epoch": 0.3836451683909311, + "grad_norm": 2.416457071317115, + "learning_rate": 7.063565588027863e-06, + "loss": 0.3557, + "step": 17429 + }, + { + "epoch": 0.38366718027734975, + "grad_norm": 3.0740155807028304, + "learning_rate": 7.0632409005463025e-06, + "loss": 0.3542, + "step": 17430 + }, + { + "epoch": 0.38368919216376846, + "grad_norm": 0.6234384832602634, + "learning_rate": 7.062916202578432e-06, + "loss": 0.4935, + "step": 17431 + }, + { + "epoch": 0.3837112040501871, + "grad_norm": 1.74962977089465, + "learning_rate": 7.0625914941259e-06, + "loss": 0.3199, + "step": 17432 + }, + { + "epoch": 0.38373321593660575, + "grad_norm": 1.4523553417667234, + "learning_rate": 7.062266775190358e-06, + "loss": 0.3133, + "step": 17433 + }, + { + "epoch": 0.38375522782302446, + "grad_norm": 1.3558179243767123, + "learning_rate": 7.061942045773458e-06, + "loss": 0.4376, + "step": 17434 + }, + { + "epoch": 0.3837772397094431, + "grad_norm": 0.6159340855289362, + "learning_rate": 7.0616173058768475e-06, + "loss": 0.4913, + "step": 17435 + }, + { + "epoch": 0.38379925159586176, + "grad_norm": 1.5773921988077806, + "learning_rate": 7.061292555502179e-06, + "loss": 0.3216, + "step": 17436 + }, + { + "epoch": 0.3838212634822804, + "grad_norm": 2.0501941412876517, + "learning_rate": 7.060967794651102e-06, + "loss": 0.4364, + "step": 17437 + }, + { + "epoch": 0.3838432753686991, + "grad_norm": 1.8055192828866842, + "learning_rate": 7.060643023325269e-06, + "loss": 0.2856, + "step": 17438 + }, + { + "epoch": 0.38386528725511776, + "grad_norm": 0.6099148551875962, + "learning_rate": 7.060318241526327e-06, + "loss": 0.5056, + "step": 17439 + }, + { + "epoch": 0.3838872991415364, + "grad_norm": 1.3420300201602933, + "learning_rate": 7.05999344925593e-06, + "loss": 0.346, + "step": 17440 + }, + { + "epoch": 0.3839093110279551, + "grad_norm": 1.4202887991425492, + "learning_rate": 7.059668646515727e-06, + "loss": 0.2628, + "step": 17441 + }, + { + "epoch": 0.38393132291437376, + "grad_norm": 2.1989170468439285, + "learning_rate": 7.059343833307369e-06, + "loss": 0.3477, + "step": 17442 + }, + { + "epoch": 0.3839533348007924, + "grad_norm": 0.6153167802036146, + "learning_rate": 7.059019009632509e-06, + "loss": 0.4764, + "step": 17443 + }, + { + "epoch": 0.3839753466872111, + "grad_norm": 1.6781499679568674, + "learning_rate": 7.058694175492793e-06, + "loss": 0.309, + "step": 17444 + }, + { + "epoch": 0.38399735857362977, + "grad_norm": 1.3506171134959102, + "learning_rate": 7.058369330889876e-06, + "loss": 0.3426, + "step": 17445 + }, + { + "epoch": 0.3840193704600484, + "grad_norm": 1.586564954427803, + "learning_rate": 7.0580444758254096e-06, + "loss": 0.3019, + "step": 17446 + }, + { + "epoch": 0.3840413823464671, + "grad_norm": 3.1862093278750017, + "learning_rate": 7.057719610301042e-06, + "loss": 0.2628, + "step": 17447 + }, + { + "epoch": 0.38406339423288577, + "grad_norm": 0.6216624376906996, + "learning_rate": 7.057394734318426e-06, + "loss": 0.4821, + "step": 17448 + }, + { + "epoch": 0.3840854061193044, + "grad_norm": 1.3713294546244483, + "learning_rate": 7.057069847879212e-06, + "loss": 0.3047, + "step": 17449 + }, + { + "epoch": 0.38410741800572307, + "grad_norm": 1.7469857614307895, + "learning_rate": 7.056744950985053e-06, + "loss": 0.3397, + "step": 17450 + }, + { + "epoch": 0.38412942989214177, + "grad_norm": 1.7543202209155309, + "learning_rate": 7.0564200436375954e-06, + "loss": 0.3065, + "step": 17451 + }, + { + "epoch": 0.3841514417785604, + "grad_norm": 0.616315473138736, + "learning_rate": 7.056095125838498e-06, + "loss": 0.5051, + "step": 17452 + }, + { + "epoch": 0.38417345366497907, + "grad_norm": 2.828352415780748, + "learning_rate": 7.055770197589404e-06, + "loss": 0.3244, + "step": 17453 + }, + { + "epoch": 0.3841954655513978, + "grad_norm": 2.097229128185279, + "learning_rate": 7.055445258891973e-06, + "loss": 0.3605, + "step": 17454 + }, + { + "epoch": 0.3842174774378164, + "grad_norm": 1.7379112486846995, + "learning_rate": 7.055120309747851e-06, + "loss": 0.3092, + "step": 17455 + }, + { + "epoch": 0.3842394893242351, + "grad_norm": 1.5895915992796823, + "learning_rate": 7.05479535015869e-06, + "loss": 0.406, + "step": 17456 + }, + { + "epoch": 0.3842615012106538, + "grad_norm": 1.6767996188965266, + "learning_rate": 7.054470380126144e-06, + "loss": 0.3458, + "step": 17457 + }, + { + "epoch": 0.3842835130970724, + "grad_norm": 1.5376726051764762, + "learning_rate": 7.0541453996518604e-06, + "loss": 0.3175, + "step": 17458 + }, + { + "epoch": 0.3843055249834911, + "grad_norm": 1.5288301706811376, + "learning_rate": 7.053820408737496e-06, + "loss": 0.4254, + "step": 17459 + }, + { + "epoch": 0.3843275368699097, + "grad_norm": 2.304774716036901, + "learning_rate": 7.053495407384701e-06, + "loss": 0.3456, + "step": 17460 + }, + { + "epoch": 0.38434954875632843, + "grad_norm": 1.4047552496217028, + "learning_rate": 7.0531703955951244e-06, + "loss": 0.2126, + "step": 17461 + }, + { + "epoch": 0.3843715606427471, + "grad_norm": 1.3776087679827775, + "learning_rate": 7.052845373370422e-06, + "loss": 0.4054, + "step": 17462 + }, + { + "epoch": 0.38439357252916573, + "grad_norm": 2.196280836444891, + "learning_rate": 7.0525203407122415e-06, + "loss": 0.3127, + "step": 17463 + }, + { + "epoch": 0.38441558441558443, + "grad_norm": 1.5312218307130816, + "learning_rate": 7.05219529762224e-06, + "loss": 0.4029, + "step": 17464 + }, + { + "epoch": 0.3844375963020031, + "grad_norm": 1.385719940574989, + "learning_rate": 7.051870244102066e-06, + "loss": 0.2999, + "step": 17465 + }, + { + "epoch": 0.38445960818842173, + "grad_norm": 1.1700232684660337, + "learning_rate": 7.05154518015337e-06, + "loss": 0.1925, + "step": 17466 + }, + { + "epoch": 0.38448162007484044, + "grad_norm": 1.9347577750783629, + "learning_rate": 7.051220105777808e-06, + "loss": 0.287, + "step": 17467 + }, + { + "epoch": 0.3845036319612591, + "grad_norm": 1.5627496290349223, + "learning_rate": 7.050895020977031e-06, + "loss": 0.2972, + "step": 17468 + }, + { + "epoch": 0.38452564384767773, + "grad_norm": 1.8549147678708044, + "learning_rate": 7.0505699257526895e-06, + "loss": 0.3863, + "step": 17469 + }, + { + "epoch": 0.38454765573409644, + "grad_norm": 1.7495095611746592, + "learning_rate": 7.050244820106439e-06, + "loss": 0.3651, + "step": 17470 + }, + { + "epoch": 0.3845696676205151, + "grad_norm": 1.429229626478062, + "learning_rate": 7.049919704039928e-06, + "loss": 0.368, + "step": 17471 + }, + { + "epoch": 0.38459167950693374, + "grad_norm": 1.3594724431736611, + "learning_rate": 7.049594577554811e-06, + "loss": 0.3109, + "step": 17472 + }, + { + "epoch": 0.3846136913933524, + "grad_norm": 1.5814211474079793, + "learning_rate": 7.049269440652742e-06, + "loss": 0.3648, + "step": 17473 + }, + { + "epoch": 0.3846357032797711, + "grad_norm": 1.9358024909354086, + "learning_rate": 7.048944293335369e-06, + "loss": 0.384, + "step": 17474 + }, + { + "epoch": 0.38465771516618974, + "grad_norm": 1.3453100418133206, + "learning_rate": 7.04861913560435e-06, + "loss": 0.3511, + "step": 17475 + }, + { + "epoch": 0.3846797270526084, + "grad_norm": 1.4090611749553927, + "learning_rate": 7.048293967461333e-06, + "loss": 0.2702, + "step": 17476 + }, + { + "epoch": 0.3847017389390271, + "grad_norm": 1.543292318342798, + "learning_rate": 7.047968788907973e-06, + "loss": 0.3363, + "step": 17477 + }, + { + "epoch": 0.38472375082544574, + "grad_norm": 1.521852180454973, + "learning_rate": 7.047643599945921e-06, + "loss": 0.3502, + "step": 17478 + }, + { + "epoch": 0.3847457627118644, + "grad_norm": 1.1427176317451089, + "learning_rate": 7.047318400576832e-06, + "loss": 0.2652, + "step": 17479 + }, + { + "epoch": 0.3847677745982831, + "grad_norm": 1.8505055817004912, + "learning_rate": 7.046993190802359e-06, + "loss": 0.2847, + "step": 17480 + }, + { + "epoch": 0.38478978648470175, + "grad_norm": 1.926578577861763, + "learning_rate": 7.046667970624152e-06, + "loss": 0.3947, + "step": 17481 + }, + { + "epoch": 0.3848117983711204, + "grad_norm": 0.6970774503378948, + "learning_rate": 7.0463427400438655e-06, + "loss": 0.511, + "step": 17482 + }, + { + "epoch": 0.38483381025753904, + "grad_norm": 1.9195687002183819, + "learning_rate": 7.046017499063154e-06, + "loss": 0.3397, + "step": 17483 + }, + { + "epoch": 0.38485582214395775, + "grad_norm": 1.3231329848403142, + "learning_rate": 7.045692247683668e-06, + "loss": 0.2983, + "step": 17484 + }, + { + "epoch": 0.3848778340303764, + "grad_norm": 1.424188512219736, + "learning_rate": 7.045366985907062e-06, + "loss": 0.3378, + "step": 17485 + }, + { + "epoch": 0.38489984591679505, + "grad_norm": 3.01452691418199, + "learning_rate": 7.045041713734988e-06, + "loss": 0.2551, + "step": 17486 + }, + { + "epoch": 0.38492185780321375, + "grad_norm": 1.7935436027728366, + "learning_rate": 7.0447164311691e-06, + "loss": 0.2562, + "step": 17487 + }, + { + "epoch": 0.3849438696896324, + "grad_norm": 1.2837730362049848, + "learning_rate": 7.044391138211052e-06, + "loss": 0.3243, + "step": 17488 + }, + { + "epoch": 0.38496588157605105, + "grad_norm": 0.6241497459098102, + "learning_rate": 7.044065834862494e-06, + "loss": 0.5092, + "step": 17489 + }, + { + "epoch": 0.38498789346246975, + "grad_norm": 0.6212771484899334, + "learning_rate": 7.043740521125084e-06, + "loss": 0.5147, + "step": 17490 + }, + { + "epoch": 0.3850099053488884, + "grad_norm": 1.5349999963933498, + "learning_rate": 7.043415197000474e-06, + "loss": 0.2959, + "step": 17491 + }, + { + "epoch": 0.38503191723530705, + "grad_norm": 1.2775242948155818, + "learning_rate": 7.043089862490315e-06, + "loss": 0.2692, + "step": 17492 + }, + { + "epoch": 0.38505392912172576, + "grad_norm": 1.7766231127025245, + "learning_rate": 7.0427645175962626e-06, + "loss": 0.2996, + "step": 17493 + }, + { + "epoch": 0.3850759410081444, + "grad_norm": 0.6257665651315517, + "learning_rate": 7.0424391623199685e-06, + "loss": 0.4937, + "step": 17494 + }, + { + "epoch": 0.38509795289456306, + "grad_norm": 2.1449887083889303, + "learning_rate": 7.042113796663089e-06, + "loss": 0.3515, + "step": 17495 + }, + { + "epoch": 0.3851199647809817, + "grad_norm": 0.632194082989624, + "learning_rate": 7.041788420627275e-06, + "loss": 0.5041, + "step": 17496 + }, + { + "epoch": 0.3851419766674004, + "grad_norm": 1.56095931466135, + "learning_rate": 7.041463034214184e-06, + "loss": 0.3607, + "step": 17497 + }, + { + "epoch": 0.38516398855381906, + "grad_norm": 2.1914196258554406, + "learning_rate": 7.041137637425466e-06, + "loss": 0.3423, + "step": 17498 + }, + { + "epoch": 0.3851860004402377, + "grad_norm": 9.0619108014591, + "learning_rate": 7.040812230262777e-06, + "loss": 0.2602, + "step": 17499 + }, + { + "epoch": 0.3852080123266564, + "grad_norm": 1.2752384289845728, + "learning_rate": 7.040486812727768e-06, + "loss": 0.2655, + "step": 17500 + }, + { + "epoch": 0.38523002421307506, + "grad_norm": 1.454698058848925, + "learning_rate": 7.040161384822096e-06, + "loss": 0.3191, + "step": 17501 + }, + { + "epoch": 0.3852520360994937, + "grad_norm": 2.2584958030785924, + "learning_rate": 7.039835946547414e-06, + "loss": 0.371, + "step": 17502 + }, + { + "epoch": 0.3852740479859124, + "grad_norm": 1.3267254618949202, + "learning_rate": 7.039510497905376e-06, + "loss": 0.2511, + "step": 17503 + }, + { + "epoch": 0.38529605987233106, + "grad_norm": 2.2275231077756303, + "learning_rate": 7.039185038897636e-06, + "loss": 0.3376, + "step": 17504 + }, + { + "epoch": 0.3853180717587497, + "grad_norm": 1.4293016524239135, + "learning_rate": 7.038859569525848e-06, + "loss": 0.2698, + "step": 17505 + }, + { + "epoch": 0.3853400836451684, + "grad_norm": 0.6618832239075563, + "learning_rate": 7.0385340897916666e-06, + "loss": 0.4863, + "step": 17506 + }, + { + "epoch": 0.38536209553158707, + "grad_norm": 1.298382311001232, + "learning_rate": 7.038208599696745e-06, + "loss": 0.3078, + "step": 17507 + }, + { + "epoch": 0.3853841074180057, + "grad_norm": 1.6886651869087872, + "learning_rate": 7.037883099242739e-06, + "loss": 0.2933, + "step": 17508 + }, + { + "epoch": 0.38540611930442437, + "grad_norm": 0.6432539840754741, + "learning_rate": 7.037557588431302e-06, + "loss": 0.4999, + "step": 17509 + }, + { + "epoch": 0.38542813119084307, + "grad_norm": 1.892030986773143, + "learning_rate": 7.0372320672640885e-06, + "loss": 0.3117, + "step": 17510 + }, + { + "epoch": 0.3854501430772617, + "grad_norm": 0.6079154448211183, + "learning_rate": 7.036906535742754e-06, + "loss": 0.4799, + "step": 17511 + }, + { + "epoch": 0.38547215496368037, + "grad_norm": 1.8821603185776299, + "learning_rate": 7.03658099386895e-06, + "loss": 0.3392, + "step": 17512 + }, + { + "epoch": 0.3854941668500991, + "grad_norm": 1.2319200815405416, + "learning_rate": 7.036255441644334e-06, + "loss": 0.2759, + "step": 17513 + }, + { + "epoch": 0.3855161787365177, + "grad_norm": 1.4492525881768903, + "learning_rate": 7.03592987907056e-06, + "loss": 0.3488, + "step": 17514 + }, + { + "epoch": 0.38553819062293637, + "grad_norm": 0.6327704079276217, + "learning_rate": 7.035604306149282e-06, + "loss": 0.4846, + "step": 17515 + }, + { + "epoch": 0.3855602025093551, + "grad_norm": 1.5333743457408364, + "learning_rate": 7.035278722882155e-06, + "loss": 0.3011, + "step": 17516 + }, + { + "epoch": 0.3855822143957737, + "grad_norm": 1.4475605984749629, + "learning_rate": 7.034953129270834e-06, + "loss": 0.3447, + "step": 17517 + }, + { + "epoch": 0.3856042262821924, + "grad_norm": 0.5958124256117369, + "learning_rate": 7.034627525316973e-06, + "loss": 0.4919, + "step": 17518 + }, + { + "epoch": 0.385626238168611, + "grad_norm": 1.7081012558542392, + "learning_rate": 7.034301911022227e-06, + "loss": 0.3082, + "step": 17519 + }, + { + "epoch": 0.38564825005502973, + "grad_norm": 1.4642847977800664, + "learning_rate": 7.033976286388253e-06, + "loss": 0.302, + "step": 17520 + }, + { + "epoch": 0.3856702619414484, + "grad_norm": 1.9134343804681002, + "learning_rate": 7.033650651416704e-06, + "loss": 0.2337, + "step": 17521 + }, + { + "epoch": 0.385692273827867, + "grad_norm": 4.9392792642534795, + "learning_rate": 7.033325006109234e-06, + "loss": 0.3246, + "step": 17522 + }, + { + "epoch": 0.38571428571428573, + "grad_norm": 1.5397512573002454, + "learning_rate": 7.0329993504675e-06, + "loss": 0.3622, + "step": 17523 + }, + { + "epoch": 0.3857362976007044, + "grad_norm": 1.6173924285944807, + "learning_rate": 7.032673684493157e-06, + "loss": 0.3739, + "step": 17524 + }, + { + "epoch": 0.38575830948712303, + "grad_norm": 1.4182674173101013, + "learning_rate": 7.032348008187859e-06, + "loss": 0.2829, + "step": 17525 + }, + { + "epoch": 0.38578032137354173, + "grad_norm": 7.28648019530958, + "learning_rate": 7.032022321553264e-06, + "loss": 0.4524, + "step": 17526 + }, + { + "epoch": 0.3858023332599604, + "grad_norm": 0.690824481973088, + "learning_rate": 7.0316966245910236e-06, + "loss": 0.4884, + "step": 17527 + }, + { + "epoch": 0.38582434514637903, + "grad_norm": 1.2824459446785705, + "learning_rate": 7.0313709173027935e-06, + "loss": 0.2626, + "step": 17528 + }, + { + "epoch": 0.38584635703279774, + "grad_norm": 1.6287761416940119, + "learning_rate": 7.031045199690233e-06, + "loss": 0.335, + "step": 17529 + }, + { + "epoch": 0.3858683689192164, + "grad_norm": 1.6098339429485604, + "learning_rate": 7.030719471754993e-06, + "loss": 0.3506, + "step": 17530 + }, + { + "epoch": 0.38589038080563504, + "grad_norm": 1.943264300588545, + "learning_rate": 7.030393733498733e-06, + "loss": 0.2836, + "step": 17531 + }, + { + "epoch": 0.3859123926920537, + "grad_norm": 1.7603436898832336, + "learning_rate": 7.030067984923105e-06, + "loss": 0.2311, + "step": 17532 + }, + { + "epoch": 0.3859344045784724, + "grad_norm": 1.3073669994393826, + "learning_rate": 7.029742226029765e-06, + "loss": 0.3333, + "step": 17533 + }, + { + "epoch": 0.38595641646489104, + "grad_norm": 1.4907922850964215, + "learning_rate": 7.029416456820373e-06, + "loss": 0.335, + "step": 17534 + }, + { + "epoch": 0.3859784283513097, + "grad_norm": 1.2318016180567313, + "learning_rate": 7.0290906772965786e-06, + "loss": 0.267, + "step": 17535 + }, + { + "epoch": 0.3860004402377284, + "grad_norm": 0.6391484191893925, + "learning_rate": 7.02876488746004e-06, + "loss": 0.5039, + "step": 17536 + }, + { + "epoch": 0.38602245212414704, + "grad_norm": 1.4556114684298014, + "learning_rate": 7.028439087312415e-06, + "loss": 0.3419, + "step": 17537 + }, + { + "epoch": 0.3860444640105657, + "grad_norm": 1.6465602614732688, + "learning_rate": 7.0281132768553576e-06, + "loss": 0.3836, + "step": 17538 + }, + { + "epoch": 0.3860664758969844, + "grad_norm": 4.406116213207712, + "learning_rate": 7.027787456090523e-06, + "loss": 0.2769, + "step": 17539 + }, + { + "epoch": 0.38608848778340304, + "grad_norm": 1.0968641653997904, + "learning_rate": 7.027461625019569e-06, + "loss": 0.2743, + "step": 17540 + }, + { + "epoch": 0.3861104996698217, + "grad_norm": 0.61527017028203, + "learning_rate": 7.02713578364415e-06, + "loss": 0.5059, + "step": 17541 + }, + { + "epoch": 0.38613251155624034, + "grad_norm": 1.9902617805554892, + "learning_rate": 7.026809931965924e-06, + "loss": 0.3881, + "step": 17542 + }, + { + "epoch": 0.38615452344265905, + "grad_norm": 1.7835102821632571, + "learning_rate": 7.026484069986545e-06, + "loss": 0.3174, + "step": 17543 + }, + { + "epoch": 0.3861765353290777, + "grad_norm": 0.6233444103316674, + "learning_rate": 7.026158197707671e-06, + "loss": 0.5222, + "step": 17544 + }, + { + "epoch": 0.38619854721549635, + "grad_norm": 1.2348267897044916, + "learning_rate": 7.025832315130958e-06, + "loss": 0.2474, + "step": 17545 + }, + { + "epoch": 0.38622055910191505, + "grad_norm": 1.1346129458438952, + "learning_rate": 7.025506422258059e-06, + "loss": 0.3557, + "step": 17546 + }, + { + "epoch": 0.3862425709883337, + "grad_norm": 1.4468608806785594, + "learning_rate": 7.025180519090635e-06, + "loss": 0.3088, + "step": 17547 + }, + { + "epoch": 0.38626458287475235, + "grad_norm": 0.6027077017519307, + "learning_rate": 7.0248546056303406e-06, + "loss": 0.4751, + "step": 17548 + }, + { + "epoch": 0.38628659476117105, + "grad_norm": 1.5813021465484844, + "learning_rate": 7.02452868187883e-06, + "loss": 0.3462, + "step": 17549 + }, + { + "epoch": 0.3863086066475897, + "grad_norm": 1.2248120104420754, + "learning_rate": 7.024202747837764e-06, + "loss": 0.3114, + "step": 17550 + }, + { + "epoch": 0.38633061853400835, + "grad_norm": 0.6118629832049157, + "learning_rate": 7.0238768035087945e-06, + "loss": 0.524, + "step": 17551 + }, + { + "epoch": 0.38635263042042706, + "grad_norm": 1.5012306854133772, + "learning_rate": 7.023550848893583e-06, + "loss": 0.3636, + "step": 17552 + }, + { + "epoch": 0.3863746423068457, + "grad_norm": 1.1942043993274711, + "learning_rate": 7.023224883993782e-06, + "loss": 0.2648, + "step": 17553 + }, + { + "epoch": 0.38639665419326436, + "grad_norm": 1.4087533287004048, + "learning_rate": 7.022898908811049e-06, + "loss": 0.3041, + "step": 17554 + }, + { + "epoch": 0.386418666079683, + "grad_norm": 1.9261156858387527, + "learning_rate": 7.022572923347043e-06, + "loss": 0.3926, + "step": 17555 + }, + { + "epoch": 0.3864406779661017, + "grad_norm": 1.5424334008238658, + "learning_rate": 7.022246927603419e-06, + "loss": 0.2511, + "step": 17556 + }, + { + "epoch": 0.38646268985252036, + "grad_norm": 1.874871597416907, + "learning_rate": 7.021920921581834e-06, + "loss": 0.3511, + "step": 17557 + }, + { + "epoch": 0.386484701738939, + "grad_norm": 0.6323180825369976, + "learning_rate": 7.021594905283945e-06, + "loss": 0.5111, + "step": 17558 + }, + { + "epoch": 0.3865067136253577, + "grad_norm": 1.5754468664683752, + "learning_rate": 7.021268878711408e-06, + "loss": 0.3098, + "step": 17559 + }, + { + "epoch": 0.38652872551177636, + "grad_norm": 1.6725054696978903, + "learning_rate": 7.020942841865882e-06, + "loss": 0.3508, + "step": 17560 + }, + { + "epoch": 0.386550737398195, + "grad_norm": 1.2253987184992297, + "learning_rate": 7.0206167947490235e-06, + "loss": 0.2607, + "step": 17561 + }, + { + "epoch": 0.3865727492846137, + "grad_norm": 3.07221695236158, + "learning_rate": 7.0202907373624896e-06, + "loss": 0.3774, + "step": 17562 + }, + { + "epoch": 0.38659476117103236, + "grad_norm": 1.2424284049341214, + "learning_rate": 7.019964669707936e-06, + "loss": 0.3442, + "step": 17563 + }, + { + "epoch": 0.386616773057451, + "grad_norm": 1.5224111791352557, + "learning_rate": 7.01963859178702e-06, + "loss": 0.3559, + "step": 17564 + }, + { + "epoch": 0.38663878494386966, + "grad_norm": 1.29312197969331, + "learning_rate": 7.0193125036014036e-06, + "loss": 0.362, + "step": 17565 + }, + { + "epoch": 0.38666079683028837, + "grad_norm": 3.3116291367551085, + "learning_rate": 7.018986405152737e-06, + "loss": 0.3014, + "step": 17566 + }, + { + "epoch": 0.386682808716707, + "grad_norm": 1.6746707043391074, + "learning_rate": 7.018660296442682e-06, + "loss": 0.3211, + "step": 17567 + }, + { + "epoch": 0.38670482060312567, + "grad_norm": 1.2599794950213723, + "learning_rate": 7.018334177472895e-06, + "loss": 0.3995, + "step": 17568 + }, + { + "epoch": 0.38672683248954437, + "grad_norm": 1.4315696319901308, + "learning_rate": 7.018008048245033e-06, + "loss": 0.3666, + "step": 17569 + }, + { + "epoch": 0.386748844375963, + "grad_norm": 1.8879687935439853, + "learning_rate": 7.0176819087607565e-06, + "loss": 0.3907, + "step": 17570 + }, + { + "epoch": 0.38677085626238167, + "grad_norm": 3.3201422232533, + "learning_rate": 7.017355759021718e-06, + "loss": 0.3604, + "step": 17571 + }, + { + "epoch": 0.3867928681488004, + "grad_norm": 1.7020173867050572, + "learning_rate": 7.017029599029578e-06, + "loss": 0.3634, + "step": 17572 + }, + { + "epoch": 0.386814880035219, + "grad_norm": 1.6926632491770044, + "learning_rate": 7.016703428785995e-06, + "loss": 0.434, + "step": 17573 + }, + { + "epoch": 0.38683689192163767, + "grad_norm": 1.6321818359038274, + "learning_rate": 7.016377248292625e-06, + "loss": 0.2803, + "step": 17574 + }, + { + "epoch": 0.3868589038080564, + "grad_norm": 1.2904498412073293, + "learning_rate": 7.016051057551127e-06, + "loss": 0.272, + "step": 17575 + }, + { + "epoch": 0.386880915694475, + "grad_norm": 6.480822845187075, + "learning_rate": 7.015724856563158e-06, + "loss": 0.31, + "step": 17576 + }, + { + "epoch": 0.3869029275808937, + "grad_norm": 1.5360087447018493, + "learning_rate": 7.015398645330376e-06, + "loss": 0.2754, + "step": 17577 + }, + { + "epoch": 0.3869249394673123, + "grad_norm": 1.3553231284191452, + "learning_rate": 7.015072423854439e-06, + "loss": 0.2744, + "step": 17578 + }, + { + "epoch": 0.38694695135373103, + "grad_norm": 1.3359438903973364, + "learning_rate": 7.0147461921370065e-06, + "loss": 0.334, + "step": 17579 + }, + { + "epoch": 0.3869689632401497, + "grad_norm": 0.6604282608851223, + "learning_rate": 7.0144199501797335e-06, + "loss": 0.4725, + "step": 17580 + }, + { + "epoch": 0.3869909751265683, + "grad_norm": 1.891256240181059, + "learning_rate": 7.0140936979842826e-06, + "loss": 0.4031, + "step": 17581 + }, + { + "epoch": 0.38701298701298703, + "grad_norm": 1.3342349992262699, + "learning_rate": 7.0137674355523055e-06, + "loss": 0.3456, + "step": 17582 + }, + { + "epoch": 0.3870349988994057, + "grad_norm": 2.7177817928590167, + "learning_rate": 7.013441162885466e-06, + "loss": 0.2993, + "step": 17583 + }, + { + "epoch": 0.38705701078582433, + "grad_norm": 1.5847322184360288, + "learning_rate": 7.013114879985421e-06, + "loss": 0.3371, + "step": 17584 + }, + { + "epoch": 0.38707902267224303, + "grad_norm": 1.8565195212422592, + "learning_rate": 7.012788586853828e-06, + "loss": 0.2558, + "step": 17585 + }, + { + "epoch": 0.3871010345586617, + "grad_norm": 1.3718428650682466, + "learning_rate": 7.012462283492346e-06, + "loss": 0.3752, + "step": 17586 + }, + { + "epoch": 0.38712304644508033, + "grad_norm": 1.6117981929810787, + "learning_rate": 7.012135969902632e-06, + "loss": 0.372, + "step": 17587 + }, + { + "epoch": 0.38714505833149904, + "grad_norm": 2.0662183440162663, + "learning_rate": 7.011809646086347e-06, + "loss": 0.2861, + "step": 17588 + }, + { + "epoch": 0.3871670702179177, + "grad_norm": 1.5950625306863617, + "learning_rate": 7.011483312045148e-06, + "loss": 0.3, + "step": 17589 + }, + { + "epoch": 0.38718908210433634, + "grad_norm": 1.206092640244773, + "learning_rate": 7.011156967780691e-06, + "loss": 0.3209, + "step": 17590 + }, + { + "epoch": 0.387211093990755, + "grad_norm": 0.6252386999620049, + "learning_rate": 7.01083061329464e-06, + "loss": 0.4924, + "step": 17591 + }, + { + "epoch": 0.3872331058771737, + "grad_norm": 1.2771379414912494, + "learning_rate": 7.01050424858865e-06, + "loss": 0.3355, + "step": 17592 + }, + { + "epoch": 0.38725511776359234, + "grad_norm": 0.6261198742165295, + "learning_rate": 7.010177873664381e-06, + "loss": 0.5171, + "step": 17593 + }, + { + "epoch": 0.387277129650011, + "grad_norm": 1.3794958333868883, + "learning_rate": 7.0098514885234916e-06, + "loss": 0.3548, + "step": 17594 + }, + { + "epoch": 0.3872991415364297, + "grad_norm": 1.3638873209520015, + "learning_rate": 7.0095250931676395e-06, + "loss": 0.3302, + "step": 17595 + }, + { + "epoch": 0.38732115342284834, + "grad_norm": 1.2913508686577782, + "learning_rate": 7.009198687598486e-06, + "loss": 0.28, + "step": 17596 + }, + { + "epoch": 0.387343165309267, + "grad_norm": 1.642949611909495, + "learning_rate": 7.008872271817688e-06, + "loss": 0.271, + "step": 17597 + }, + { + "epoch": 0.3873651771956857, + "grad_norm": 1.3872223624358888, + "learning_rate": 7.008545845826905e-06, + "loss": 0.3054, + "step": 17598 + }, + { + "epoch": 0.38738718908210434, + "grad_norm": 1.3147427182136788, + "learning_rate": 7.008219409627797e-06, + "loss": 0.3737, + "step": 17599 + }, + { + "epoch": 0.387409200968523, + "grad_norm": 2.165325134222433, + "learning_rate": 7.00789296322202e-06, + "loss": 0.3551, + "step": 17600 + }, + { + "epoch": 0.38743121285494164, + "grad_norm": 2.040856562377586, + "learning_rate": 7.0075665066112375e-06, + "loss": 0.378, + "step": 17601 + }, + { + "epoch": 0.38745322474136035, + "grad_norm": 1.4049236072696047, + "learning_rate": 7.007240039797106e-06, + "loss": 0.3442, + "step": 17602 + }, + { + "epoch": 0.387475236627779, + "grad_norm": 0.6250462913168913, + "learning_rate": 7.006913562781286e-06, + "loss": 0.4807, + "step": 17603 + }, + { + "epoch": 0.38749724851419765, + "grad_norm": 1.5219860743996658, + "learning_rate": 7.006587075565437e-06, + "loss": 0.2913, + "step": 17604 + }, + { + "epoch": 0.38751926040061635, + "grad_norm": 0.6228785076637732, + "learning_rate": 7.006260578151214e-06, + "loss": 0.486, + "step": 17605 + }, + { + "epoch": 0.387541272287035, + "grad_norm": 1.2961581584213218, + "learning_rate": 7.005934070540284e-06, + "loss": 0.3061, + "step": 17606 + }, + { + "epoch": 0.38756328417345365, + "grad_norm": 1.3333931218244206, + "learning_rate": 7.0056075527343e-06, + "loss": 0.3366, + "step": 17607 + }, + { + "epoch": 0.38758529605987235, + "grad_norm": 1.4577626154877779, + "learning_rate": 7.005281024734923e-06, + "loss": 0.3769, + "step": 17608 + }, + { + "epoch": 0.387607307946291, + "grad_norm": 1.82834000445418, + "learning_rate": 7.004954486543816e-06, + "loss": 0.3718, + "step": 17609 + }, + { + "epoch": 0.38762931983270965, + "grad_norm": 1.360469739349094, + "learning_rate": 7.004627938162635e-06, + "loss": 0.293, + "step": 17610 + }, + { + "epoch": 0.38765133171912836, + "grad_norm": 1.7435745309132813, + "learning_rate": 7.00430137959304e-06, + "loss": 0.404, + "step": 17611 + }, + { + "epoch": 0.387673343605547, + "grad_norm": 1.2366777486119935, + "learning_rate": 7.0039748108366925e-06, + "loss": 0.2274, + "step": 17612 + }, + { + "epoch": 0.38769535549196565, + "grad_norm": 0.6570166024245205, + "learning_rate": 7.003648231895251e-06, + "loss": 0.4992, + "step": 17613 + }, + { + "epoch": 0.3877173673783843, + "grad_norm": 1.8791389855925076, + "learning_rate": 7.003321642770375e-06, + "loss": 0.2716, + "step": 17614 + }, + { + "epoch": 0.387739379264803, + "grad_norm": 1.138512429321388, + "learning_rate": 7.002995043463725e-06, + "loss": 0.2934, + "step": 17615 + }, + { + "epoch": 0.38776139115122166, + "grad_norm": 1.522437916971518, + "learning_rate": 7.002668433976961e-06, + "loss": 0.3325, + "step": 17616 + }, + { + "epoch": 0.3877834030376403, + "grad_norm": 1.854508911340642, + "learning_rate": 7.002341814311743e-06, + "loss": 0.31, + "step": 17617 + }, + { + "epoch": 0.387805414924059, + "grad_norm": 1.2720456324872396, + "learning_rate": 7.002015184469731e-06, + "loss": 0.3402, + "step": 17618 + }, + { + "epoch": 0.38782742681047766, + "grad_norm": 1.4647489843512866, + "learning_rate": 7.001688544452584e-06, + "loss": 0.1971, + "step": 17619 + }, + { + "epoch": 0.3878494386968963, + "grad_norm": 1.171156327938519, + "learning_rate": 7.001361894261963e-06, + "loss": 0.2805, + "step": 17620 + }, + { + "epoch": 0.387871450583315, + "grad_norm": 1.4508318826323057, + "learning_rate": 7.001035233899529e-06, + "loss": 0.2572, + "step": 17621 + }, + { + "epoch": 0.38789346246973366, + "grad_norm": 1.3679835917667782, + "learning_rate": 7.000708563366943e-06, + "loss": 0.3507, + "step": 17622 + }, + { + "epoch": 0.3879154743561523, + "grad_norm": 1.3410181244238961, + "learning_rate": 7.0003818826658594e-06, + "loss": 0.2461, + "step": 17623 + }, + { + "epoch": 0.38793748624257096, + "grad_norm": 1.427965903133146, + "learning_rate": 7.000055191797947e-06, + "loss": 0.3631, + "step": 17624 + }, + { + "epoch": 0.38795949812898967, + "grad_norm": 1.4355391027466702, + "learning_rate": 6.999728490764861e-06, + "loss": 0.2953, + "step": 17625 + }, + { + "epoch": 0.3879815100154083, + "grad_norm": 1.401994055386407, + "learning_rate": 6.999401779568263e-06, + "loss": 0.311, + "step": 17626 + }, + { + "epoch": 0.38800352190182696, + "grad_norm": 1.260171301806907, + "learning_rate": 6.999075058209812e-06, + "loss": 0.2622, + "step": 17627 + }, + { + "epoch": 0.38802553378824567, + "grad_norm": 1.5288055793693982, + "learning_rate": 6.998748326691172e-06, + "loss": 0.3056, + "step": 17628 + }, + { + "epoch": 0.3880475456746643, + "grad_norm": 1.935776302387291, + "learning_rate": 6.998421585013999e-06, + "loss": 0.4094, + "step": 17629 + }, + { + "epoch": 0.38806955756108297, + "grad_norm": 1.4971133187221095, + "learning_rate": 6.998094833179958e-06, + "loss": 0.4035, + "step": 17630 + }, + { + "epoch": 0.38809156944750167, + "grad_norm": 0.6525042380517129, + "learning_rate": 6.997768071190707e-06, + "loss": 0.4973, + "step": 17631 + }, + { + "epoch": 0.3881135813339203, + "grad_norm": 0.6149105648552824, + "learning_rate": 6.997441299047908e-06, + "loss": 0.4842, + "step": 17632 + }, + { + "epoch": 0.38813559322033897, + "grad_norm": 1.4900713086395732, + "learning_rate": 6.9971145167532215e-06, + "loss": 0.3557, + "step": 17633 + }, + { + "epoch": 0.3881576051067577, + "grad_norm": 1.4429827094845327, + "learning_rate": 6.9967877243083086e-06, + "loss": 0.3985, + "step": 17634 + }, + { + "epoch": 0.3881796169931763, + "grad_norm": 1.4438690770390594, + "learning_rate": 6.99646092171483e-06, + "loss": 0.3035, + "step": 17635 + }, + { + "epoch": 0.388201628879595, + "grad_norm": 1.5591702477422134, + "learning_rate": 6.996134108974445e-06, + "loss": 0.3027, + "step": 17636 + }, + { + "epoch": 0.3882236407660136, + "grad_norm": 1.4968782435690131, + "learning_rate": 6.995807286088817e-06, + "loss": 0.4029, + "step": 17637 + }, + { + "epoch": 0.3882456526524323, + "grad_norm": 0.6078556693266146, + "learning_rate": 6.995480453059607e-06, + "loss": 0.4751, + "step": 17638 + }, + { + "epoch": 0.388267664538851, + "grad_norm": 1.5242114528640742, + "learning_rate": 6.995153609888475e-06, + "loss": 0.2812, + "step": 17639 + }, + { + "epoch": 0.3882896764252696, + "grad_norm": 1.518185716112483, + "learning_rate": 6.994826756577082e-06, + "loss": 0.3035, + "step": 17640 + }, + { + "epoch": 0.38831168831168833, + "grad_norm": 2.9818330267765587, + "learning_rate": 6.9944998931270906e-06, + "loss": 0.3146, + "step": 17641 + }, + { + "epoch": 0.388333700198107, + "grad_norm": 1.4628685008647897, + "learning_rate": 6.99417301954016e-06, + "loss": 0.3946, + "step": 17642 + }, + { + "epoch": 0.38835571208452563, + "grad_norm": 1.6909016611818295, + "learning_rate": 6.993846135817954e-06, + "loss": 0.3707, + "step": 17643 + }, + { + "epoch": 0.38837772397094433, + "grad_norm": 1.770429152921536, + "learning_rate": 6.993519241962131e-06, + "loss": 0.3149, + "step": 17644 + }, + { + "epoch": 0.388399735857363, + "grad_norm": 2.180067054994241, + "learning_rate": 6.993192337974356e-06, + "loss": 0.3157, + "step": 17645 + }, + { + "epoch": 0.38842174774378163, + "grad_norm": 1.5876291308399573, + "learning_rate": 6.992865423856288e-06, + "loss": 0.3442, + "step": 17646 + }, + { + "epoch": 0.3884437596302003, + "grad_norm": 3.445542313215087, + "learning_rate": 6.992538499609588e-06, + "loss": 0.3543, + "step": 17647 + }, + { + "epoch": 0.388465771516619, + "grad_norm": 1.5967322043579897, + "learning_rate": 6.99221156523592e-06, + "loss": 0.2863, + "step": 17648 + }, + { + "epoch": 0.38848778340303763, + "grad_norm": 1.4055740121798241, + "learning_rate": 6.991884620736943e-06, + "loss": 0.3366, + "step": 17649 + }, + { + "epoch": 0.3885097952894563, + "grad_norm": 0.5936441896445626, + "learning_rate": 6.99155766611432e-06, + "loss": 0.4559, + "step": 17650 + }, + { + "epoch": 0.388531807175875, + "grad_norm": 1.4501276757006896, + "learning_rate": 6.991230701369713e-06, + "loss": 0.3487, + "step": 17651 + }, + { + "epoch": 0.38855381906229364, + "grad_norm": 1.3444961818911705, + "learning_rate": 6.990903726504784e-06, + "loss": 0.4044, + "step": 17652 + }, + { + "epoch": 0.3885758309487123, + "grad_norm": 1.1171127632189761, + "learning_rate": 6.9905767415211945e-06, + "loss": 0.3197, + "step": 17653 + }, + { + "epoch": 0.388597842835131, + "grad_norm": 1.4373591071586145, + "learning_rate": 6.9902497464206055e-06, + "loss": 0.331, + "step": 17654 + }, + { + "epoch": 0.38861985472154964, + "grad_norm": 2.117123758873809, + "learning_rate": 6.989922741204679e-06, + "loss": 0.3161, + "step": 17655 + }, + { + "epoch": 0.3886418666079683, + "grad_norm": 1.3291429650505444, + "learning_rate": 6.989595725875079e-06, + "loss": 0.3754, + "step": 17656 + }, + { + "epoch": 0.388663878494387, + "grad_norm": 1.492031476049053, + "learning_rate": 6.9892687004334646e-06, + "loss": 0.3544, + "step": 17657 + }, + { + "epoch": 0.38868589038080564, + "grad_norm": 1.2882972522878144, + "learning_rate": 6.988941664881501e-06, + "loss": 0.362, + "step": 17658 + }, + { + "epoch": 0.3887079022672243, + "grad_norm": 1.3776099172989473, + "learning_rate": 6.988614619220847e-06, + "loss": 0.2919, + "step": 17659 + }, + { + "epoch": 0.38872991415364294, + "grad_norm": 1.3704687171722127, + "learning_rate": 6.988287563453169e-06, + "loss": 0.2994, + "step": 17660 + }, + { + "epoch": 0.38875192604006165, + "grad_norm": 1.5128779790751998, + "learning_rate": 6.987960497580126e-06, + "loss": 0.3701, + "step": 17661 + }, + { + "epoch": 0.3887739379264803, + "grad_norm": 1.2645773569411574, + "learning_rate": 6.9876334216033795e-06, + "loss": 0.2738, + "step": 17662 + }, + { + "epoch": 0.38879594981289894, + "grad_norm": 0.6812542174267564, + "learning_rate": 6.987306335524595e-06, + "loss": 0.5135, + "step": 17663 + }, + { + "epoch": 0.38881796169931765, + "grad_norm": 2.651130649033968, + "learning_rate": 6.986979239345433e-06, + "loss": 0.3626, + "step": 17664 + }, + { + "epoch": 0.3888399735857363, + "grad_norm": 1.4016892645575854, + "learning_rate": 6.986652133067556e-06, + "loss": 0.3504, + "step": 17665 + }, + { + "epoch": 0.38886198547215495, + "grad_norm": 0.6285470986292978, + "learning_rate": 6.986325016692627e-06, + "loss": 0.5005, + "step": 17666 + }, + { + "epoch": 0.38888399735857365, + "grad_norm": 1.7084178737723976, + "learning_rate": 6.985997890222307e-06, + "loss": 0.3941, + "step": 17667 + }, + { + "epoch": 0.3889060092449923, + "grad_norm": 0.5993451963254404, + "learning_rate": 6.985670753658262e-06, + "loss": 0.4854, + "step": 17668 + }, + { + "epoch": 0.38892802113141095, + "grad_norm": 2.6164014480453304, + "learning_rate": 6.985343607002152e-06, + "loss": 0.301, + "step": 17669 + }, + { + "epoch": 0.38895003301782965, + "grad_norm": 1.5012591389529824, + "learning_rate": 6.98501645025564e-06, + "loss": 0.3628, + "step": 17670 + }, + { + "epoch": 0.3889720449042483, + "grad_norm": 0.6197677067787136, + "learning_rate": 6.98468928342039e-06, + "loss": 0.4906, + "step": 17671 + }, + { + "epoch": 0.38899405679066695, + "grad_norm": 1.5814815310089894, + "learning_rate": 6.984362106498061e-06, + "loss": 0.4306, + "step": 17672 + }, + { + "epoch": 0.3890160686770856, + "grad_norm": 1.5506308058851495, + "learning_rate": 6.984034919490321e-06, + "loss": 0.2862, + "step": 17673 + }, + { + "epoch": 0.3890380805635043, + "grad_norm": 1.9685286100427544, + "learning_rate": 6.983707722398832e-06, + "loss": 0.2944, + "step": 17674 + }, + { + "epoch": 0.38906009244992296, + "grad_norm": 2.1945272095194426, + "learning_rate": 6.983380515225253e-06, + "loss": 0.355, + "step": 17675 + }, + { + "epoch": 0.3890821043363416, + "grad_norm": 1.7696437021252354, + "learning_rate": 6.983053297971251e-06, + "loss": 0.3552, + "step": 17676 + }, + { + "epoch": 0.3891041162227603, + "grad_norm": 1.1110818100380107, + "learning_rate": 6.982726070638486e-06, + "loss": 0.2648, + "step": 17677 + }, + { + "epoch": 0.38912612810917896, + "grad_norm": 1.742061480777864, + "learning_rate": 6.982398833228624e-06, + "loss": 0.3567, + "step": 17678 + }, + { + "epoch": 0.3891481399955976, + "grad_norm": 1.3721765789475588, + "learning_rate": 6.982071585743328e-06, + "loss": 0.3121, + "step": 17679 + }, + { + "epoch": 0.3891701518820163, + "grad_norm": 2.0334850210610322, + "learning_rate": 6.981744328184259e-06, + "loss": 0.3424, + "step": 17680 + }, + { + "epoch": 0.38919216376843496, + "grad_norm": 1.3062727615206007, + "learning_rate": 6.981417060553081e-06, + "loss": 0.33, + "step": 17681 + }, + { + "epoch": 0.3892141756548536, + "grad_norm": 1.223254881375525, + "learning_rate": 6.981089782851459e-06, + "loss": 0.301, + "step": 17682 + }, + { + "epoch": 0.38923618754127226, + "grad_norm": 1.3731822476532154, + "learning_rate": 6.980762495081054e-06, + "loss": 0.3079, + "step": 17683 + }, + { + "epoch": 0.38925819942769097, + "grad_norm": 1.3979944004479854, + "learning_rate": 6.980435197243531e-06, + "loss": 0.2827, + "step": 17684 + }, + { + "epoch": 0.3892802113141096, + "grad_norm": 1.4015425013964908, + "learning_rate": 6.980107889340553e-06, + "loss": 0.277, + "step": 17685 + }, + { + "epoch": 0.38930222320052826, + "grad_norm": 1.53943906808858, + "learning_rate": 6.979780571373784e-06, + "loss": 0.347, + "step": 17686 + }, + { + "epoch": 0.38932423508694697, + "grad_norm": 1.5895137668321326, + "learning_rate": 6.979453243344887e-06, + "loss": 0.4134, + "step": 17687 + }, + { + "epoch": 0.3893462469733656, + "grad_norm": 1.782039398443584, + "learning_rate": 6.9791259052555245e-06, + "loss": 0.2655, + "step": 17688 + }, + { + "epoch": 0.38936825885978427, + "grad_norm": 1.6522858201264643, + "learning_rate": 6.978798557107364e-06, + "loss": 0.3194, + "step": 17689 + }, + { + "epoch": 0.38939027074620297, + "grad_norm": 2.488006202970452, + "learning_rate": 6.978471198902064e-06, + "loss": 0.344, + "step": 17690 + }, + { + "epoch": 0.3894122826326216, + "grad_norm": 2.597154988474102, + "learning_rate": 6.978143830641292e-06, + "loss": 0.3523, + "step": 17691 + }, + { + "epoch": 0.38943429451904027, + "grad_norm": 1.8164917860883274, + "learning_rate": 6.977816452326711e-06, + "loss": 0.3762, + "step": 17692 + }, + { + "epoch": 0.389456306405459, + "grad_norm": 0.6580378573608325, + "learning_rate": 6.977489063959984e-06, + "loss": 0.5142, + "step": 17693 + }, + { + "epoch": 0.3894783182918776, + "grad_norm": 3.076365266905034, + "learning_rate": 6.977161665542776e-06, + "loss": 0.3582, + "step": 17694 + }, + { + "epoch": 0.3895003301782963, + "grad_norm": 1.8050175436085114, + "learning_rate": 6.9768342570767496e-06, + "loss": 0.376, + "step": 17695 + }, + { + "epoch": 0.3895223420647149, + "grad_norm": 1.1710117964825901, + "learning_rate": 6.976506838563572e-06, + "loss": 0.1982, + "step": 17696 + }, + { + "epoch": 0.3895443539511336, + "grad_norm": 2.1169678052584384, + "learning_rate": 6.976179410004905e-06, + "loss": 0.3308, + "step": 17697 + }, + { + "epoch": 0.3895663658375523, + "grad_norm": 1.6602910797378738, + "learning_rate": 6.975851971402411e-06, + "loss": 0.3625, + "step": 17698 + }, + { + "epoch": 0.3895883777239709, + "grad_norm": 1.3204972685359584, + "learning_rate": 6.9755245227577576e-06, + "loss": 0.254, + "step": 17699 + }, + { + "epoch": 0.38961038961038963, + "grad_norm": 1.8798721454414868, + "learning_rate": 6.975197064072606e-06, + "loss": 0.3529, + "step": 17700 + }, + { + "epoch": 0.3896324014968083, + "grad_norm": 0.6174645159740443, + "learning_rate": 6.9748695953486225e-06, + "loss": 0.4813, + "step": 17701 + }, + { + "epoch": 0.3896544133832269, + "grad_norm": 0.6014718964899024, + "learning_rate": 6.974542116587471e-06, + "loss": 0.4786, + "step": 17702 + }, + { + "epoch": 0.38967642526964563, + "grad_norm": 1.5360653213825914, + "learning_rate": 6.974214627790816e-06, + "loss": 0.3358, + "step": 17703 + }, + { + "epoch": 0.3896984371560643, + "grad_norm": 1.5187797215464665, + "learning_rate": 6.973887128960321e-06, + "loss": 0.3591, + "step": 17704 + }, + { + "epoch": 0.38972044904248293, + "grad_norm": 1.4652296213818519, + "learning_rate": 6.973559620097653e-06, + "loss": 0.3956, + "step": 17705 + }, + { + "epoch": 0.3897424609289016, + "grad_norm": 1.4525161496944925, + "learning_rate": 6.973232101204474e-06, + "loss": 0.2072, + "step": 17706 + }, + { + "epoch": 0.3897644728153203, + "grad_norm": 1.7148604699417485, + "learning_rate": 6.972904572282448e-06, + "loss": 0.2751, + "step": 17707 + }, + { + "epoch": 0.38978648470173893, + "grad_norm": 5.182885634350082, + "learning_rate": 6.972577033333242e-06, + "loss": 0.3339, + "step": 17708 + }, + { + "epoch": 0.3898084965881576, + "grad_norm": 1.7190740844684211, + "learning_rate": 6.97224948435852e-06, + "loss": 0.3078, + "step": 17709 + }, + { + "epoch": 0.3898305084745763, + "grad_norm": 1.3543311170473296, + "learning_rate": 6.971921925359946e-06, + "loss": 0.3191, + "step": 17710 + }, + { + "epoch": 0.38985252036099494, + "grad_norm": 1.2606943510244848, + "learning_rate": 6.971594356339186e-06, + "loss": 0.3211, + "step": 17711 + }, + { + "epoch": 0.3898745322474136, + "grad_norm": 1.3473902766005965, + "learning_rate": 6.971266777297904e-06, + "loss": 0.3429, + "step": 17712 + }, + { + "epoch": 0.3898965441338323, + "grad_norm": 0.6639259405973454, + "learning_rate": 6.970939188237763e-06, + "loss": 0.4795, + "step": 17713 + }, + { + "epoch": 0.38991855602025094, + "grad_norm": 1.512812621611336, + "learning_rate": 6.970611589160432e-06, + "loss": 0.3453, + "step": 17714 + }, + { + "epoch": 0.3899405679066696, + "grad_norm": 1.2233369303283175, + "learning_rate": 6.9702839800675745e-06, + "loss": 0.3184, + "step": 17715 + }, + { + "epoch": 0.3899625797930883, + "grad_norm": 1.5799371071967558, + "learning_rate": 6.969956360960853e-06, + "loss": 0.328, + "step": 17716 + }, + { + "epoch": 0.38998459167950694, + "grad_norm": 1.7696302369088892, + "learning_rate": 6.969628731841937e-06, + "loss": 0.2869, + "step": 17717 + }, + { + "epoch": 0.3900066035659256, + "grad_norm": 4.92206419733657, + "learning_rate": 6.969301092712488e-06, + "loss": 0.2931, + "step": 17718 + }, + { + "epoch": 0.39002861545234424, + "grad_norm": 1.3964980108825094, + "learning_rate": 6.9689734435741715e-06, + "loss": 0.3083, + "step": 17719 + }, + { + "epoch": 0.39005062733876295, + "grad_norm": 1.3465450133615144, + "learning_rate": 6.968645784428656e-06, + "loss": 0.3799, + "step": 17720 + }, + { + "epoch": 0.3900726392251816, + "grad_norm": 1.670838766989084, + "learning_rate": 6.968318115277603e-06, + "loss": 0.3141, + "step": 17721 + }, + { + "epoch": 0.39009465111160024, + "grad_norm": 1.2476052228892727, + "learning_rate": 6.9679904361226805e-06, + "loss": 0.3505, + "step": 17722 + }, + { + "epoch": 0.39011666299801895, + "grad_norm": 2.274222822398082, + "learning_rate": 6.967662746965551e-06, + "loss": 0.3858, + "step": 17723 + }, + { + "epoch": 0.3901386748844376, + "grad_norm": 1.4041299189933885, + "learning_rate": 6.967335047807884e-06, + "loss": 0.3331, + "step": 17724 + }, + { + "epoch": 0.39016068677085625, + "grad_norm": 1.5365151241854664, + "learning_rate": 6.967007338651342e-06, + "loss": 0.3033, + "step": 17725 + }, + { + "epoch": 0.39018269865727495, + "grad_norm": 1.2202713423944465, + "learning_rate": 6.966679619497591e-06, + "loss": 0.3339, + "step": 17726 + }, + { + "epoch": 0.3902047105436936, + "grad_norm": 1.3982989962628636, + "learning_rate": 6.966351890348297e-06, + "loss": 0.315, + "step": 17727 + }, + { + "epoch": 0.39022672243011225, + "grad_norm": 2.127068406958828, + "learning_rate": 6.966024151205126e-06, + "loss": 0.3847, + "step": 17728 + }, + { + "epoch": 0.39024873431653095, + "grad_norm": 1.7830980192523103, + "learning_rate": 6.965696402069743e-06, + "loss": 0.2894, + "step": 17729 + }, + { + "epoch": 0.3902707462029496, + "grad_norm": 1.2459793361785079, + "learning_rate": 6.965368642943815e-06, + "loss": 0.2636, + "step": 17730 + }, + { + "epoch": 0.39029275808936825, + "grad_norm": 1.315666043724786, + "learning_rate": 6.965040873829005e-06, + "loss": 0.3421, + "step": 17731 + }, + { + "epoch": 0.3903147699757869, + "grad_norm": 1.4521706571133217, + "learning_rate": 6.964713094726984e-06, + "loss": 0.3943, + "step": 17732 + }, + { + "epoch": 0.3903367818622056, + "grad_norm": 1.2930682980263966, + "learning_rate": 6.964385305639413e-06, + "loss": 0.3452, + "step": 17733 + }, + { + "epoch": 0.39035879374862426, + "grad_norm": 1.4677231108882767, + "learning_rate": 6.964057506567959e-06, + "loss": 0.4097, + "step": 17734 + }, + { + "epoch": 0.3903808056350429, + "grad_norm": 2.088970220061326, + "learning_rate": 6.96372969751429e-06, + "loss": 0.3235, + "step": 17735 + }, + { + "epoch": 0.3904028175214616, + "grad_norm": 1.8021792627528719, + "learning_rate": 6.9634018784800715e-06, + "loss": 0.361, + "step": 17736 + }, + { + "epoch": 0.39042482940788026, + "grad_norm": 6.400931651813787, + "learning_rate": 6.963074049466967e-06, + "loss": 0.2437, + "step": 17737 + }, + { + "epoch": 0.3904468412942989, + "grad_norm": 1.3410949499546894, + "learning_rate": 6.962746210476647e-06, + "loss": 0.3597, + "step": 17738 + }, + { + "epoch": 0.3904688531807176, + "grad_norm": 1.5621293085870684, + "learning_rate": 6.962418361510772e-06, + "loss": 0.3176, + "step": 17739 + }, + { + "epoch": 0.39049086506713626, + "grad_norm": 1.1215502515235805, + "learning_rate": 6.962090502571014e-06, + "loss": 0.3069, + "step": 17740 + }, + { + "epoch": 0.3905128769535549, + "grad_norm": 0.717594479975354, + "learning_rate": 6.961762633659037e-06, + "loss": 0.5234, + "step": 17741 + }, + { + "epoch": 0.39053488883997356, + "grad_norm": 0.6637874137592356, + "learning_rate": 6.961434754776506e-06, + "loss": 0.5255, + "step": 17742 + }, + { + "epoch": 0.39055690072639226, + "grad_norm": 2.0891756756664517, + "learning_rate": 6.96110686592509e-06, + "loss": 0.4006, + "step": 17743 + }, + { + "epoch": 0.3905789126128109, + "grad_norm": 1.276445548134552, + "learning_rate": 6.960778967106453e-06, + "loss": 0.3005, + "step": 17744 + }, + { + "epoch": 0.39060092449922956, + "grad_norm": 1.7798520724551485, + "learning_rate": 6.960451058322264e-06, + "loss": 0.3388, + "step": 17745 + }, + { + "epoch": 0.39062293638564827, + "grad_norm": 1.3890249219563977, + "learning_rate": 6.960123139574187e-06, + "loss": 0.2826, + "step": 17746 + }, + { + "epoch": 0.3906449482720669, + "grad_norm": 2.5949687469286187, + "learning_rate": 6.9597952108638915e-06, + "loss": 0.2967, + "step": 17747 + }, + { + "epoch": 0.39066696015848557, + "grad_norm": 1.3165348595773714, + "learning_rate": 6.959467272193042e-06, + "loss": 0.3909, + "step": 17748 + }, + { + "epoch": 0.39068897204490427, + "grad_norm": 1.6441511338746582, + "learning_rate": 6.959139323563305e-06, + "loss": 0.3217, + "step": 17749 + }, + { + "epoch": 0.3907109839313229, + "grad_norm": 1.4005798946013517, + "learning_rate": 6.95881136497635e-06, + "loss": 0.3152, + "step": 17750 + }, + { + "epoch": 0.39073299581774157, + "grad_norm": 1.4399886124526216, + "learning_rate": 6.958483396433841e-06, + "loss": 0.2986, + "step": 17751 + }, + { + "epoch": 0.3907550077041603, + "grad_norm": 2.1683817139202826, + "learning_rate": 6.958155417937445e-06, + "loss": 0.3015, + "step": 17752 + }, + { + "epoch": 0.3907770195905789, + "grad_norm": 1.3366648898643965, + "learning_rate": 6.957827429488831e-06, + "loss": 0.3754, + "step": 17753 + }, + { + "epoch": 0.39079903147699757, + "grad_norm": 1.7407904600811215, + "learning_rate": 6.957499431089665e-06, + "loss": 0.3268, + "step": 17754 + }, + { + "epoch": 0.3908210433634162, + "grad_norm": 0.9721895111550097, + "learning_rate": 6.957171422741613e-06, + "loss": 0.5205, + "step": 17755 + }, + { + "epoch": 0.3908430552498349, + "grad_norm": 1.3860217176874599, + "learning_rate": 6.956843404446344e-06, + "loss": 0.3278, + "step": 17756 + }, + { + "epoch": 0.3908650671362536, + "grad_norm": 1.5542798293315245, + "learning_rate": 6.9565153762055225e-06, + "loss": 0.3429, + "step": 17757 + }, + { + "epoch": 0.3908870790226722, + "grad_norm": 1.310933591433322, + "learning_rate": 6.9561873380208175e-06, + "loss": 0.2295, + "step": 17758 + }, + { + "epoch": 0.39090909090909093, + "grad_norm": 2.068487948286671, + "learning_rate": 6.955859289893897e-06, + "loss": 0.2553, + "step": 17759 + }, + { + "epoch": 0.3909311027955096, + "grad_norm": 1.8341413198919034, + "learning_rate": 6.9555312318264255e-06, + "loss": 0.2316, + "step": 17760 + }, + { + "epoch": 0.3909531146819282, + "grad_norm": 0.6525282566284085, + "learning_rate": 6.955203163820075e-06, + "loss": 0.5134, + "step": 17761 + }, + { + "epoch": 0.39097512656834693, + "grad_norm": 127.79294414987963, + "learning_rate": 6.9548750858765066e-06, + "loss": 0.2869, + "step": 17762 + }, + { + "epoch": 0.3909971384547656, + "grad_norm": 0.651669413730817, + "learning_rate": 6.954546997997393e-06, + "loss": 0.4883, + "step": 17763 + }, + { + "epoch": 0.39101915034118423, + "grad_norm": 1.6637791204211345, + "learning_rate": 6.9542189001844e-06, + "loss": 0.3533, + "step": 17764 + }, + { + "epoch": 0.3910411622276029, + "grad_norm": 2.1659945755767205, + "learning_rate": 6.953890792439195e-06, + "loss": 0.3421, + "step": 17765 + }, + { + "epoch": 0.3910631741140216, + "grad_norm": 0.6598356418836129, + "learning_rate": 6.953562674763446e-06, + "loss": 0.5107, + "step": 17766 + }, + { + "epoch": 0.39108518600044023, + "grad_norm": 1.5234499192058206, + "learning_rate": 6.953234547158817e-06, + "loss": 0.3618, + "step": 17767 + }, + { + "epoch": 0.3911071978868589, + "grad_norm": 1.6833311874780021, + "learning_rate": 6.952906409626981e-06, + "loss": 0.3985, + "step": 17768 + }, + { + "epoch": 0.3911292097732776, + "grad_norm": 1.5947360229602463, + "learning_rate": 6.952578262169605e-06, + "loss": 0.2503, + "step": 17769 + }, + { + "epoch": 0.39115122165969624, + "grad_norm": 1.5801745640481974, + "learning_rate": 6.952250104788354e-06, + "loss": 0.3611, + "step": 17770 + }, + { + "epoch": 0.3911732335461149, + "grad_norm": 2.652287582128711, + "learning_rate": 6.9519219374848964e-06, + "loss": 0.3088, + "step": 17771 + }, + { + "epoch": 0.3911952454325336, + "grad_norm": 1.3327153113946963, + "learning_rate": 6.951593760260903e-06, + "loss": 0.2743, + "step": 17772 + }, + { + "epoch": 0.39121725731895224, + "grad_norm": 1.6422669999709032, + "learning_rate": 6.951265573118038e-06, + "loss": 0.2647, + "step": 17773 + }, + { + "epoch": 0.3912392692053709, + "grad_norm": 1.9158626510632217, + "learning_rate": 6.950937376057972e-06, + "loss": 0.345, + "step": 17774 + }, + { + "epoch": 0.3912612810917896, + "grad_norm": 1.1443071397677254, + "learning_rate": 6.950609169082371e-06, + "loss": 0.2773, + "step": 17775 + }, + { + "epoch": 0.39128329297820824, + "grad_norm": 1.3654418620026416, + "learning_rate": 6.950280952192905e-06, + "loss": 0.3174, + "step": 17776 + }, + { + "epoch": 0.3913053048646269, + "grad_norm": 1.4323332706870533, + "learning_rate": 6.949952725391241e-06, + "loss": 0.3302, + "step": 17777 + }, + { + "epoch": 0.39132731675104554, + "grad_norm": 1.4048618091234804, + "learning_rate": 6.949624488679048e-06, + "loss": 0.3576, + "step": 17778 + }, + { + "epoch": 0.39134932863746424, + "grad_norm": 1.3997507392577855, + "learning_rate": 6.949296242057995e-06, + "loss": 0.3233, + "step": 17779 + }, + { + "epoch": 0.3913713405238829, + "grad_norm": 1.513835378277979, + "learning_rate": 6.9489679855297455e-06, + "loss": 0.3817, + "step": 17780 + }, + { + "epoch": 0.39139335241030154, + "grad_norm": 1.4693791387249828, + "learning_rate": 6.948639719095974e-06, + "loss": 0.3453, + "step": 17781 + }, + { + "epoch": 0.39141536429672025, + "grad_norm": 1.5851531728597297, + "learning_rate": 6.948311442758347e-06, + "loss": 0.2569, + "step": 17782 + }, + { + "epoch": 0.3914373761831389, + "grad_norm": 1.490554428446487, + "learning_rate": 6.9479831565185295e-06, + "loss": 0.3157, + "step": 17783 + }, + { + "epoch": 0.39145938806955755, + "grad_norm": 0.7523558401660837, + "learning_rate": 6.947654860378196e-06, + "loss": 0.5215, + "step": 17784 + }, + { + "epoch": 0.39148139995597625, + "grad_norm": 1.4627841352074122, + "learning_rate": 6.94732655433901e-06, + "loss": 0.4329, + "step": 17785 + }, + { + "epoch": 0.3915034118423949, + "grad_norm": 4.63763538995131, + "learning_rate": 6.946998238402642e-06, + "loss": 0.3046, + "step": 17786 + }, + { + "epoch": 0.39152542372881355, + "grad_norm": 1.4240365929961765, + "learning_rate": 6.9466699125707605e-06, + "loss": 0.2584, + "step": 17787 + }, + { + "epoch": 0.3915474356152322, + "grad_norm": 1.5336544426118768, + "learning_rate": 6.946341576845035e-06, + "loss": 0.3543, + "step": 17788 + }, + { + "epoch": 0.3915694475016509, + "grad_norm": 0.6322900967308498, + "learning_rate": 6.946013231227133e-06, + "loss": 0.5276, + "step": 17789 + }, + { + "epoch": 0.39159145938806955, + "grad_norm": 0.9736890996234605, + "learning_rate": 6.945684875718723e-06, + "loss": 0.253, + "step": 17790 + }, + { + "epoch": 0.3916134712744882, + "grad_norm": 3.0763797007872546, + "learning_rate": 6.945356510321476e-06, + "loss": 0.292, + "step": 17791 + }, + { + "epoch": 0.3916354831609069, + "grad_norm": 1.350817958319504, + "learning_rate": 6.945028135037058e-06, + "loss": 0.298, + "step": 17792 + }, + { + "epoch": 0.39165749504732555, + "grad_norm": 1.1367680010924655, + "learning_rate": 6.944699749867142e-06, + "loss": 0.2779, + "step": 17793 + }, + { + "epoch": 0.3916795069337442, + "grad_norm": 2.57429067245924, + "learning_rate": 6.944371354813393e-06, + "loss": 0.3836, + "step": 17794 + }, + { + "epoch": 0.3917015188201629, + "grad_norm": 0.6431235474776483, + "learning_rate": 6.944042949877482e-06, + "loss": 0.4955, + "step": 17795 + }, + { + "epoch": 0.39172353070658156, + "grad_norm": 1.8332521109336861, + "learning_rate": 6.943714535061077e-06, + "loss": 0.3856, + "step": 17796 + }, + { + "epoch": 0.3917455425930002, + "grad_norm": 1.7331961451654463, + "learning_rate": 6.943386110365848e-06, + "loss": 0.3769, + "step": 17797 + }, + { + "epoch": 0.3917675544794189, + "grad_norm": 1.558903055925297, + "learning_rate": 6.943057675793464e-06, + "loss": 0.264, + "step": 17798 + }, + { + "epoch": 0.39178956636583756, + "grad_norm": 1.3109343750405695, + "learning_rate": 6.942729231345595e-06, + "loss": 0.2936, + "step": 17799 + }, + { + "epoch": 0.3918115782522562, + "grad_norm": 1.9223130990300734, + "learning_rate": 6.942400777023909e-06, + "loss": 0.4055, + "step": 17800 + }, + { + "epoch": 0.39183359013867486, + "grad_norm": 0.6224182665430753, + "learning_rate": 6.942072312830075e-06, + "loss": 0.4882, + "step": 17801 + }, + { + "epoch": 0.39185560202509356, + "grad_norm": 1.7991400803293005, + "learning_rate": 6.941743838765765e-06, + "loss": 0.3417, + "step": 17802 + }, + { + "epoch": 0.3918776139115122, + "grad_norm": 1.996578137048256, + "learning_rate": 6.941415354832647e-06, + "loss": 0.2999, + "step": 17803 + }, + { + "epoch": 0.39189962579793086, + "grad_norm": 1.530419115335413, + "learning_rate": 6.941086861032388e-06, + "loss": 0.3205, + "step": 17804 + }, + { + "epoch": 0.39192163768434957, + "grad_norm": 1.3165258455270847, + "learning_rate": 6.940758357366661e-06, + "loss": 0.2686, + "step": 17805 + }, + { + "epoch": 0.3919436495707682, + "grad_norm": 1.7221468235248838, + "learning_rate": 6.940429843837135e-06, + "loss": 0.4395, + "step": 17806 + }, + { + "epoch": 0.39196566145718686, + "grad_norm": 1.5605869255706928, + "learning_rate": 6.940101320445478e-06, + "loss": 0.2808, + "step": 17807 + }, + { + "epoch": 0.39198767334360557, + "grad_norm": 0.6355414587175638, + "learning_rate": 6.939772787193362e-06, + "loss": 0.4925, + "step": 17808 + }, + { + "epoch": 0.3920096852300242, + "grad_norm": 0.6163906817462198, + "learning_rate": 6.9394442440824536e-06, + "loss": 0.471, + "step": 17809 + }, + { + "epoch": 0.39203169711644287, + "grad_norm": 1.3724707111997791, + "learning_rate": 6.939115691114426e-06, + "loss": 0.2612, + "step": 17810 + }, + { + "epoch": 0.3920537090028616, + "grad_norm": 1.579779366728631, + "learning_rate": 6.938787128290948e-06, + "loss": 0.283, + "step": 17811 + }, + { + "epoch": 0.3920757208892802, + "grad_norm": 1.5261241154416578, + "learning_rate": 6.93845855561369e-06, + "loss": 0.3133, + "step": 17812 + }, + { + "epoch": 0.39209773277569887, + "grad_norm": 2.920802531776467, + "learning_rate": 6.938129973084319e-06, + "loss": 0.3241, + "step": 17813 + }, + { + "epoch": 0.3921197446621175, + "grad_norm": 1.3057243563246292, + "learning_rate": 6.937801380704506e-06, + "loss": 0.3306, + "step": 17814 + }, + { + "epoch": 0.3921417565485362, + "grad_norm": 2.801445833107322, + "learning_rate": 6.937472778475925e-06, + "loss": 0.3609, + "step": 17815 + }, + { + "epoch": 0.3921637684349549, + "grad_norm": 0.6278924302758895, + "learning_rate": 6.937144166400241e-06, + "loss": 0.4827, + "step": 17816 + }, + { + "epoch": 0.3921857803213735, + "grad_norm": 1.3779334394639147, + "learning_rate": 6.9368155444791274e-06, + "loss": 0.3043, + "step": 17817 + }, + { + "epoch": 0.3922077922077922, + "grad_norm": 2.4455193352925875, + "learning_rate": 6.936486912714254e-06, + "loss": 0.3387, + "step": 17818 + }, + { + "epoch": 0.3922298040942109, + "grad_norm": 1.555482005199081, + "learning_rate": 6.936158271107288e-06, + "loss": 0.2734, + "step": 17819 + }, + { + "epoch": 0.3922518159806295, + "grad_norm": 1.3920201960485856, + "learning_rate": 6.935829619659906e-06, + "loss": 0.2807, + "step": 17820 + }, + { + "epoch": 0.39227382786704823, + "grad_norm": 0.6002661661562624, + "learning_rate": 6.935500958373772e-06, + "loss": 0.5068, + "step": 17821 + }, + { + "epoch": 0.3922958397534669, + "grad_norm": 1.3665477546903448, + "learning_rate": 6.9351722872505575e-06, + "loss": 0.3613, + "step": 17822 + }, + { + "epoch": 0.39231785163988553, + "grad_norm": 0.6088398614034065, + "learning_rate": 6.934843606291937e-06, + "loss": 0.4873, + "step": 17823 + }, + { + "epoch": 0.3923398635263042, + "grad_norm": 1.574859757421875, + "learning_rate": 6.934514915499577e-06, + "loss": 0.4375, + "step": 17824 + }, + { + "epoch": 0.3923618754127229, + "grad_norm": 3.480988713380254, + "learning_rate": 6.934186214875151e-06, + "loss": 0.3284, + "step": 17825 + }, + { + "epoch": 0.39238388729914153, + "grad_norm": 1.2059744828567371, + "learning_rate": 6.933857504420326e-06, + "loss": 0.3055, + "step": 17826 + }, + { + "epoch": 0.3924058991855602, + "grad_norm": 2.242877037402224, + "learning_rate": 6.933528784136774e-06, + "loss": 0.267, + "step": 17827 + }, + { + "epoch": 0.3924279110719789, + "grad_norm": 0.6219185637433486, + "learning_rate": 6.933200054026168e-06, + "loss": 0.5112, + "step": 17828 + }, + { + "epoch": 0.39244992295839753, + "grad_norm": 1.5670729851110476, + "learning_rate": 6.932871314090177e-06, + "loss": 0.2912, + "step": 17829 + }, + { + "epoch": 0.3924719348448162, + "grad_norm": 1.609826288294827, + "learning_rate": 6.9325425643304715e-06, + "loss": 0.4262, + "step": 17830 + }, + { + "epoch": 0.3924939467312349, + "grad_norm": 0.590002831683937, + "learning_rate": 6.932213804748723e-06, + "loss": 0.4939, + "step": 17831 + }, + { + "epoch": 0.39251595861765354, + "grad_norm": 1.3544832318388904, + "learning_rate": 6.931885035346601e-06, + "loss": 0.4025, + "step": 17832 + }, + { + "epoch": 0.3925379705040722, + "grad_norm": 2.675261837141294, + "learning_rate": 6.93155625612578e-06, + "loss": 0.2854, + "step": 17833 + }, + { + "epoch": 0.3925599823904909, + "grad_norm": 1.6104411657708986, + "learning_rate": 6.931227467087927e-06, + "loss": 0.3662, + "step": 17834 + }, + { + "epoch": 0.39258199427690954, + "grad_norm": 1.9175855208594712, + "learning_rate": 6.930898668234714e-06, + "loss": 0.25, + "step": 17835 + }, + { + "epoch": 0.3926040061633282, + "grad_norm": 1.5437036214819257, + "learning_rate": 6.930569859567815e-06, + "loss": 0.2855, + "step": 17836 + }, + { + "epoch": 0.39262601804974684, + "grad_norm": 1.5502849409336104, + "learning_rate": 6.930241041088896e-06, + "loss": 0.3872, + "step": 17837 + }, + { + "epoch": 0.39264802993616554, + "grad_norm": 0.5960516643115416, + "learning_rate": 6.929912212799634e-06, + "loss": 0.5008, + "step": 17838 + }, + { + "epoch": 0.3926700418225842, + "grad_norm": 1.4619158212878627, + "learning_rate": 6.929583374701696e-06, + "loss": 0.2709, + "step": 17839 + }, + { + "epoch": 0.39269205370900284, + "grad_norm": 1.352062262468448, + "learning_rate": 6.929254526796753e-06, + "loss": 0.2515, + "step": 17840 + }, + { + "epoch": 0.39271406559542155, + "grad_norm": 1.232216000361511, + "learning_rate": 6.928925669086482e-06, + "loss": 0.32, + "step": 17841 + }, + { + "epoch": 0.3927360774818402, + "grad_norm": 1.4440462978192035, + "learning_rate": 6.928596801572547e-06, + "loss": 0.3222, + "step": 17842 + }, + { + "epoch": 0.39275808936825884, + "grad_norm": 1.8338721138111278, + "learning_rate": 6.928267924256624e-06, + "loss": 0.3604, + "step": 17843 + }, + { + "epoch": 0.39278010125467755, + "grad_norm": 1.347523624478149, + "learning_rate": 6.927939037140384e-06, + "loss": 0.2431, + "step": 17844 + }, + { + "epoch": 0.3928021131410962, + "grad_norm": 1.2821315863866682, + "learning_rate": 6.927610140225497e-06, + "loss": 0.338, + "step": 17845 + }, + { + "epoch": 0.39282412502751485, + "grad_norm": 1.494155174248245, + "learning_rate": 6.9272812335136354e-06, + "loss": 0.3215, + "step": 17846 + }, + { + "epoch": 0.3928461369139335, + "grad_norm": 2.005838522531643, + "learning_rate": 6.926952317006473e-06, + "loss": 0.2922, + "step": 17847 + }, + { + "epoch": 0.3928681488003522, + "grad_norm": 2.0219269266286752, + "learning_rate": 6.926623390705677e-06, + "loss": 0.4355, + "step": 17848 + }, + { + "epoch": 0.39289016068677085, + "grad_norm": 1.775719513428057, + "learning_rate": 6.926294454612923e-06, + "loss": 0.3672, + "step": 17849 + }, + { + "epoch": 0.3929121725731895, + "grad_norm": 1.787449960944091, + "learning_rate": 6.925965508729879e-06, + "loss": 0.272, + "step": 17850 + }, + { + "epoch": 0.3929341844596082, + "grad_norm": 1.420115031145131, + "learning_rate": 6.925636553058223e-06, + "loss": 0.3764, + "step": 17851 + }, + { + "epoch": 0.39295619634602685, + "grad_norm": 2.4730651872593437, + "learning_rate": 6.9253075875996215e-06, + "loss": 0.3778, + "step": 17852 + }, + { + "epoch": 0.3929782082324455, + "grad_norm": 1.544531121425028, + "learning_rate": 6.9249786123557484e-06, + "loss": 0.445, + "step": 17853 + }, + { + "epoch": 0.3930002201188642, + "grad_norm": 1.2862346969400813, + "learning_rate": 6.924649627328275e-06, + "loss": 0.2795, + "step": 17854 + }, + { + "epoch": 0.39302223200528286, + "grad_norm": 0.6488348683455375, + "learning_rate": 6.924320632518873e-06, + "loss": 0.518, + "step": 17855 + }, + { + "epoch": 0.3930442438917015, + "grad_norm": 1.3842810282555538, + "learning_rate": 6.923991627929217e-06, + "loss": 0.2869, + "step": 17856 + }, + { + "epoch": 0.3930662557781202, + "grad_norm": 0.614358971774325, + "learning_rate": 6.9236626135609765e-06, + "loss": 0.5014, + "step": 17857 + }, + { + "epoch": 0.39308826766453886, + "grad_norm": 1.9046430535998893, + "learning_rate": 6.923333589415824e-06, + "loss": 0.3798, + "step": 17858 + }, + { + "epoch": 0.3931102795509575, + "grad_norm": 0.600818661683987, + "learning_rate": 6.923004555495433e-06, + "loss": 0.5019, + "step": 17859 + }, + { + "epoch": 0.39313229143737616, + "grad_norm": 1.2887800780928411, + "learning_rate": 6.922675511801476e-06, + "loss": 0.3498, + "step": 17860 + }, + { + "epoch": 0.39315430332379486, + "grad_norm": 0.622971594443828, + "learning_rate": 6.922346458335624e-06, + "loss": 0.502, + "step": 17861 + }, + { + "epoch": 0.3931763152102135, + "grad_norm": 1.3288055693275533, + "learning_rate": 6.92201739509955e-06, + "loss": 0.3867, + "step": 17862 + }, + { + "epoch": 0.39319832709663216, + "grad_norm": 1.6171667788846, + "learning_rate": 6.921688322094925e-06, + "loss": 0.2383, + "step": 17863 + }, + { + "epoch": 0.39322033898305087, + "grad_norm": 2.052544934920393, + "learning_rate": 6.921359239323424e-06, + "loss": 0.4307, + "step": 17864 + }, + { + "epoch": 0.3932423508694695, + "grad_norm": 1.9945218017780004, + "learning_rate": 6.921030146786718e-06, + "loss": 0.3341, + "step": 17865 + }, + { + "epoch": 0.39326436275588816, + "grad_norm": 1.6099397398625308, + "learning_rate": 6.9207010444864795e-06, + "loss": 0.3758, + "step": 17866 + }, + { + "epoch": 0.39328637464230687, + "grad_norm": 1.743585022472109, + "learning_rate": 6.920371932424383e-06, + "loss": 0.3912, + "step": 17867 + }, + { + "epoch": 0.3933083865287255, + "grad_norm": 1.329280883860334, + "learning_rate": 6.920042810602099e-06, + "loss": 0.3647, + "step": 17868 + }, + { + "epoch": 0.39333039841514417, + "grad_norm": 1.416529734688274, + "learning_rate": 6.919713679021301e-06, + "loss": 0.3035, + "step": 17869 + }, + { + "epoch": 0.39335241030156287, + "grad_norm": 2.21452558184692, + "learning_rate": 6.919384537683662e-06, + "loss": 0.3695, + "step": 17870 + }, + { + "epoch": 0.3933744221879815, + "grad_norm": 1.8136302384065657, + "learning_rate": 6.919055386590856e-06, + "loss": 0.3618, + "step": 17871 + }, + { + "epoch": 0.39339643407440017, + "grad_norm": 1.3162892751553419, + "learning_rate": 6.918726225744553e-06, + "loss": 0.3035, + "step": 17872 + }, + { + "epoch": 0.3934184459608188, + "grad_norm": 1.465660009471826, + "learning_rate": 6.918397055146427e-06, + "loss": 0.303, + "step": 17873 + }, + { + "epoch": 0.3934404578472375, + "grad_norm": 1.869846871932321, + "learning_rate": 6.918067874798154e-06, + "loss": 0.3534, + "step": 17874 + }, + { + "epoch": 0.3934624697336562, + "grad_norm": 1.455652588213274, + "learning_rate": 6.9177386847014026e-06, + "loss": 0.3768, + "step": 17875 + }, + { + "epoch": 0.3934844816200748, + "grad_norm": 1.2576403283466686, + "learning_rate": 6.917409484857849e-06, + "loss": 0.2523, + "step": 17876 + }, + { + "epoch": 0.3935064935064935, + "grad_norm": 1.5108049103570869, + "learning_rate": 6.917080275269163e-06, + "loss": 0.2594, + "step": 17877 + }, + { + "epoch": 0.3935285053929122, + "grad_norm": 1.602752227576599, + "learning_rate": 6.916751055937023e-06, + "loss": 0.3333, + "step": 17878 + }, + { + "epoch": 0.3935505172793308, + "grad_norm": 1.2727362748899087, + "learning_rate": 6.9164218268630975e-06, + "loss": 0.2793, + "step": 17879 + }, + { + "epoch": 0.39357252916574953, + "grad_norm": 1.7937990982918683, + "learning_rate": 6.916092588049062e-06, + "loss": 0.3309, + "step": 17880 + }, + { + "epoch": 0.3935945410521682, + "grad_norm": 1.3053783592253914, + "learning_rate": 6.915763339496588e-06, + "loss": 0.3474, + "step": 17881 + }, + { + "epoch": 0.39361655293858683, + "grad_norm": 1.1993791953360526, + "learning_rate": 6.915434081207353e-06, + "loss": 0.3669, + "step": 17882 + }, + { + "epoch": 0.3936385648250055, + "grad_norm": 1.7855085221523468, + "learning_rate": 6.915104813183026e-06, + "loss": 0.3442, + "step": 17883 + }, + { + "epoch": 0.3936605767114242, + "grad_norm": 1.5457792116765103, + "learning_rate": 6.914775535425282e-06, + "loss": 0.3752, + "step": 17884 + }, + { + "epoch": 0.39368258859784283, + "grad_norm": 1.7339905184710118, + "learning_rate": 6.914446247935795e-06, + "loss": 0.3897, + "step": 17885 + }, + { + "epoch": 0.3937046004842615, + "grad_norm": 1.3485077221292532, + "learning_rate": 6.914116950716239e-06, + "loss": 0.2905, + "step": 17886 + }, + { + "epoch": 0.3937266123706802, + "grad_norm": 1.5435172616836137, + "learning_rate": 6.913787643768286e-06, + "loss": 0.3606, + "step": 17887 + }, + { + "epoch": 0.39374862425709883, + "grad_norm": 2.1301499805832886, + "learning_rate": 6.913458327093611e-06, + "loss": 0.2657, + "step": 17888 + }, + { + "epoch": 0.3937706361435175, + "grad_norm": 0.6967060391571682, + "learning_rate": 6.913129000693888e-06, + "loss": 0.5124, + "step": 17889 + }, + { + "epoch": 0.3937926480299362, + "grad_norm": 1.399125867734061, + "learning_rate": 6.912799664570788e-06, + "loss": 0.3224, + "step": 17890 + }, + { + "epoch": 0.39381465991635484, + "grad_norm": 1.0069699896951008, + "learning_rate": 6.912470318725988e-06, + "loss": 0.244, + "step": 17891 + }, + { + "epoch": 0.3938366718027735, + "grad_norm": 1.1081748284799726, + "learning_rate": 6.9121409631611615e-06, + "loss": 0.2056, + "step": 17892 + }, + { + "epoch": 0.3938586836891922, + "grad_norm": 2.56630681573719, + "learning_rate": 6.911811597877983e-06, + "loss": 0.3758, + "step": 17893 + }, + { + "epoch": 0.39388069557561084, + "grad_norm": 1.586674445197361, + "learning_rate": 6.911482222878123e-06, + "loss": 0.3338, + "step": 17894 + }, + { + "epoch": 0.3939027074620295, + "grad_norm": 1.3166319746287125, + "learning_rate": 6.911152838163259e-06, + "loss": 0.2916, + "step": 17895 + }, + { + "epoch": 0.39392471934844814, + "grad_norm": 1.6104006123039356, + "learning_rate": 6.910823443735063e-06, + "loss": 0.4028, + "step": 17896 + }, + { + "epoch": 0.39394673123486684, + "grad_norm": 1.4609986824478853, + "learning_rate": 6.910494039595211e-06, + "loss": 0.3534, + "step": 17897 + }, + { + "epoch": 0.3939687431212855, + "grad_norm": 1.183226038284274, + "learning_rate": 6.910164625745376e-06, + "loss": 0.2913, + "step": 17898 + }, + { + "epoch": 0.39399075500770414, + "grad_norm": 2.2466142036788472, + "learning_rate": 6.9098352021872315e-06, + "loss": 0.2987, + "step": 17899 + }, + { + "epoch": 0.39401276689412285, + "grad_norm": 1.4070797954235181, + "learning_rate": 6.909505768922453e-06, + "loss": 0.2667, + "step": 17900 + }, + { + "epoch": 0.3940347787805415, + "grad_norm": 1.2514049752043057, + "learning_rate": 6.909176325952716e-06, + "loss": 0.3337, + "step": 17901 + }, + { + "epoch": 0.39405679066696014, + "grad_norm": 1.3169381792332224, + "learning_rate": 6.908846873279691e-06, + "loss": 0.1944, + "step": 17902 + }, + { + "epoch": 0.39407880255337885, + "grad_norm": 1.3422890706166415, + "learning_rate": 6.908517410905057e-06, + "loss": 0.3575, + "step": 17903 + }, + { + "epoch": 0.3941008144397975, + "grad_norm": 1.343795859938083, + "learning_rate": 6.908187938830484e-06, + "loss": 0.3661, + "step": 17904 + }, + { + "epoch": 0.39412282632621615, + "grad_norm": 1.6746367045988009, + "learning_rate": 6.90785845705765e-06, + "loss": 0.3936, + "step": 17905 + }, + { + "epoch": 0.3941448382126348, + "grad_norm": 1.2250420994707725, + "learning_rate": 6.907528965588229e-06, + "loss": 0.2961, + "step": 17906 + }, + { + "epoch": 0.3941668500990535, + "grad_norm": 1.238424180827098, + "learning_rate": 6.9071994644238945e-06, + "loss": 0.3375, + "step": 17907 + }, + { + "epoch": 0.39418886198547215, + "grad_norm": 1.3302291198031375, + "learning_rate": 6.906869953566322e-06, + "loss": 0.3444, + "step": 17908 + }, + { + "epoch": 0.3942108738718908, + "grad_norm": 1.4875188126841934, + "learning_rate": 6.906540433017185e-06, + "loss": 0.2371, + "step": 17909 + }, + { + "epoch": 0.3942328857583095, + "grad_norm": 0.6845312091699245, + "learning_rate": 6.90621090277816e-06, + "loss": 0.5135, + "step": 17910 + }, + { + "epoch": 0.39425489764472815, + "grad_norm": 1.609391729288759, + "learning_rate": 6.905881362850921e-06, + "loss": 0.3582, + "step": 17911 + }, + { + "epoch": 0.3942769095311468, + "grad_norm": 1.3064777375870205, + "learning_rate": 6.905551813237142e-06, + "loss": 0.3849, + "step": 17912 + }, + { + "epoch": 0.3942989214175655, + "grad_norm": 1.3821687741726676, + "learning_rate": 6.9052222539385e-06, + "loss": 0.2515, + "step": 17913 + }, + { + "epoch": 0.39432093330398416, + "grad_norm": 1.285581931015166, + "learning_rate": 6.9048926849566685e-06, + "loss": 0.2774, + "step": 17914 + }, + { + "epoch": 0.3943429451904028, + "grad_norm": 1.3843926889230005, + "learning_rate": 6.904563106293323e-06, + "loss": 0.2982, + "step": 17915 + }, + { + "epoch": 0.3943649570768215, + "grad_norm": 0.6575946487590405, + "learning_rate": 6.904233517950138e-06, + "loss": 0.5129, + "step": 17916 + }, + { + "epoch": 0.39438696896324016, + "grad_norm": 2.008769165460793, + "learning_rate": 6.903903919928787e-06, + "loss": 0.4121, + "step": 17917 + }, + { + "epoch": 0.3944089808496588, + "grad_norm": 1.1703784811111617, + "learning_rate": 6.90357431223095e-06, + "loss": 0.2775, + "step": 17918 + }, + { + "epoch": 0.39443099273607746, + "grad_norm": 0.6306851880425884, + "learning_rate": 6.903244694858298e-06, + "loss": 0.5124, + "step": 17919 + }, + { + "epoch": 0.39445300462249616, + "grad_norm": 1.3108095653230394, + "learning_rate": 6.902915067812508e-06, + "loss": 0.2786, + "step": 17920 + }, + { + "epoch": 0.3944750165089148, + "grad_norm": 3.0828327968733684, + "learning_rate": 6.902585431095255e-06, + "loss": 0.2953, + "step": 17921 + }, + { + "epoch": 0.39449702839533346, + "grad_norm": 1.8060303303630916, + "learning_rate": 6.902255784708214e-06, + "loss": 0.3648, + "step": 17922 + }, + { + "epoch": 0.39451904028175216, + "grad_norm": 1.7572877034108076, + "learning_rate": 6.9019261286530606e-06, + "loss": 0.2987, + "step": 17923 + }, + { + "epoch": 0.3945410521681708, + "grad_norm": 1.6988057520575854, + "learning_rate": 6.9015964629314705e-06, + "loss": 0.2612, + "step": 17924 + }, + { + "epoch": 0.39456306405458946, + "grad_norm": 1.3932059818732196, + "learning_rate": 6.901266787545119e-06, + "loss": 0.3234, + "step": 17925 + }, + { + "epoch": 0.39458507594100817, + "grad_norm": 1.483209331669373, + "learning_rate": 6.9009371024956815e-06, + "loss": 0.2987, + "step": 17926 + }, + { + "epoch": 0.3946070878274268, + "grad_norm": 1.2879376236003313, + "learning_rate": 6.900607407784833e-06, + "loss": 0.3562, + "step": 17927 + }, + { + "epoch": 0.39462909971384547, + "grad_norm": 1.4935696368747924, + "learning_rate": 6.900277703414251e-06, + "loss": 0.4191, + "step": 17928 + }, + { + "epoch": 0.3946511116002641, + "grad_norm": 1.4719629603784607, + "learning_rate": 6.899947989385611e-06, + "loss": 0.4059, + "step": 17929 + }, + { + "epoch": 0.3946731234866828, + "grad_norm": 1.6102720142864078, + "learning_rate": 6.8996182657005865e-06, + "loss": 0.355, + "step": 17930 + }, + { + "epoch": 0.39469513537310147, + "grad_norm": 3.159384412310306, + "learning_rate": 6.899288532360855e-06, + "loss": 0.2672, + "step": 17931 + }, + { + "epoch": 0.3947171472595201, + "grad_norm": 1.2436599370481407, + "learning_rate": 6.898958789368093e-06, + "loss": 0.3441, + "step": 17932 + }, + { + "epoch": 0.3947391591459388, + "grad_norm": 1.3330592949390825, + "learning_rate": 6.898629036723974e-06, + "loss": 0.3183, + "step": 17933 + }, + { + "epoch": 0.39476117103235747, + "grad_norm": 0.6828038599627824, + "learning_rate": 6.898299274430177e-06, + "loss": 0.4908, + "step": 17934 + }, + { + "epoch": 0.3947831829187761, + "grad_norm": 1.3265423815885742, + "learning_rate": 6.897969502488374e-06, + "loss": 0.3167, + "step": 17935 + }, + { + "epoch": 0.3948051948051948, + "grad_norm": 1.469451286352311, + "learning_rate": 6.897639720900245e-06, + "loss": 0.283, + "step": 17936 + }, + { + "epoch": 0.3948272066916135, + "grad_norm": 1.40464595863451, + "learning_rate": 6.897309929667466e-06, + "loss": 0.2713, + "step": 17937 + }, + { + "epoch": 0.3948492185780321, + "grad_norm": 1.187197281409653, + "learning_rate": 6.896980128791711e-06, + "loss": 0.2669, + "step": 17938 + }, + { + "epoch": 0.39487123046445083, + "grad_norm": 1.2953290311931127, + "learning_rate": 6.8966503182746555e-06, + "loss": 0.3527, + "step": 17939 + }, + { + "epoch": 0.3948932423508695, + "grad_norm": 1.1865726473799074, + "learning_rate": 6.896320498117976e-06, + "loss": 0.304, + "step": 17940 + }, + { + "epoch": 0.3949152542372881, + "grad_norm": 1.5975266687957332, + "learning_rate": 6.895990668323353e-06, + "loss": 0.2843, + "step": 17941 + }, + { + "epoch": 0.3949372661237068, + "grad_norm": 0.6066357886434118, + "learning_rate": 6.895660828892459e-06, + "loss": 0.5137, + "step": 17942 + }, + { + "epoch": 0.3949592780101255, + "grad_norm": 1.6803783529218224, + "learning_rate": 6.895330979826971e-06, + "loss": 0.3014, + "step": 17943 + }, + { + "epoch": 0.39498128989654413, + "grad_norm": 0.6139036251714203, + "learning_rate": 6.895001121128565e-06, + "loss": 0.4743, + "step": 17944 + }, + { + "epoch": 0.3950033017829628, + "grad_norm": 0.6202179043646826, + "learning_rate": 6.894671252798918e-06, + "loss": 0.5011, + "step": 17945 + }, + { + "epoch": 0.3950253136693815, + "grad_norm": 2.9184036996851663, + "learning_rate": 6.894341374839707e-06, + "loss": 0.4001, + "step": 17946 + }, + { + "epoch": 0.39504732555580013, + "grad_norm": 2.0275839868252015, + "learning_rate": 6.89401148725261e-06, + "loss": 0.2764, + "step": 17947 + }, + { + "epoch": 0.3950693374422188, + "grad_norm": 1.378532019553706, + "learning_rate": 6.8936815900393e-06, + "loss": 0.3458, + "step": 17948 + }, + { + "epoch": 0.3950913493286375, + "grad_norm": 1.387132140055337, + "learning_rate": 6.893351683201455e-06, + "loss": 0.3713, + "step": 17949 + }, + { + "epoch": 0.39511336121505614, + "grad_norm": 1.438054228411537, + "learning_rate": 6.893021766740753e-06, + "loss": 0.3093, + "step": 17950 + }, + { + "epoch": 0.3951353731014748, + "grad_norm": 1.244387516378042, + "learning_rate": 6.89269184065887e-06, + "loss": 0.373, + "step": 17951 + }, + { + "epoch": 0.3951573849878935, + "grad_norm": 1.1707766201130012, + "learning_rate": 6.8923619049574845e-06, + "loss": 0.3381, + "step": 17952 + }, + { + "epoch": 0.39517939687431214, + "grad_norm": 2.0818513456137913, + "learning_rate": 6.892031959638269e-06, + "loss": 0.4455, + "step": 17953 + }, + { + "epoch": 0.3952014087607308, + "grad_norm": 1.8909306454012733, + "learning_rate": 6.891702004702905e-06, + "loss": 0.3235, + "step": 17954 + }, + { + "epoch": 0.39522342064714944, + "grad_norm": 1.1706984969898386, + "learning_rate": 6.891372040153067e-06, + "loss": 0.2462, + "step": 17955 + }, + { + "epoch": 0.39524543253356814, + "grad_norm": 2.7586870826238874, + "learning_rate": 6.891042065990433e-06, + "loss": 0.3556, + "step": 17956 + }, + { + "epoch": 0.3952674444199868, + "grad_norm": 1.5188982550714354, + "learning_rate": 6.890712082216681e-06, + "loss": 0.3646, + "step": 17957 + }, + { + "epoch": 0.39528945630640544, + "grad_norm": 0.5996765902978607, + "learning_rate": 6.890382088833484e-06, + "loss": 0.472, + "step": 17958 + }, + { + "epoch": 0.39531146819282414, + "grad_norm": 1.512413091064238, + "learning_rate": 6.890052085842524e-06, + "loss": 0.3414, + "step": 17959 + }, + { + "epoch": 0.3953334800792428, + "grad_norm": 3.132332014932226, + "learning_rate": 6.889722073245476e-06, + "loss": 0.3577, + "step": 17960 + }, + { + "epoch": 0.39535549196566144, + "grad_norm": 1.4613198689506717, + "learning_rate": 6.889392051044018e-06, + "loss": 0.3832, + "step": 17961 + }, + { + "epoch": 0.39537750385208015, + "grad_norm": 1.5225752524666216, + "learning_rate": 6.8890620192398274e-06, + "loss": 0.4143, + "step": 17962 + }, + { + "epoch": 0.3953995157384988, + "grad_norm": 1.3437538341097834, + "learning_rate": 6.88873197783458e-06, + "loss": 0.3437, + "step": 17963 + }, + { + "epoch": 0.39542152762491745, + "grad_norm": 0.6184692692733404, + "learning_rate": 6.888401926829954e-06, + "loss": 0.4869, + "step": 17964 + }, + { + "epoch": 0.3954435395113361, + "grad_norm": 1.5870343038318093, + "learning_rate": 6.888071866227628e-06, + "loss": 0.3611, + "step": 17965 + }, + { + "epoch": 0.3954655513977548, + "grad_norm": 0.6297370846974005, + "learning_rate": 6.887741796029278e-06, + "loss": 0.5149, + "step": 17966 + }, + { + "epoch": 0.39548756328417345, + "grad_norm": 1.5980415619795296, + "learning_rate": 6.887411716236583e-06, + "loss": 0.3417, + "step": 17967 + }, + { + "epoch": 0.3955095751705921, + "grad_norm": 0.6058913868150226, + "learning_rate": 6.8870816268512196e-06, + "loss": 0.502, + "step": 17968 + }, + { + "epoch": 0.3955315870570108, + "grad_norm": 1.6658473087551735, + "learning_rate": 6.886751527874865e-06, + "loss": 0.2283, + "step": 17969 + }, + { + "epoch": 0.39555359894342945, + "grad_norm": 1.3867163300401397, + "learning_rate": 6.8864214193092e-06, + "loss": 0.4131, + "step": 17970 + }, + { + "epoch": 0.3955756108298481, + "grad_norm": 1.2195043086541042, + "learning_rate": 6.886091301155896e-06, + "loss": 0.2423, + "step": 17971 + }, + { + "epoch": 0.3955976227162668, + "grad_norm": 1.6721304368912548, + "learning_rate": 6.885761173416638e-06, + "loss": 0.3305, + "step": 17972 + }, + { + "epoch": 0.39561963460268545, + "grad_norm": 1.2431947603607232, + "learning_rate": 6.8854310360931e-06, + "loss": 0.3418, + "step": 17973 + }, + { + "epoch": 0.3956416464891041, + "grad_norm": 1.0325275728426633, + "learning_rate": 6.885100889186959e-06, + "loss": 0.2722, + "step": 17974 + }, + { + "epoch": 0.3956636583755228, + "grad_norm": 3.8760123790515513, + "learning_rate": 6.884770732699897e-06, + "loss": 0.2526, + "step": 17975 + }, + { + "epoch": 0.39568567026194146, + "grad_norm": 1.4029773395271066, + "learning_rate": 6.884440566633587e-06, + "loss": 0.3221, + "step": 17976 + }, + { + "epoch": 0.3957076821483601, + "grad_norm": 1.4740214543820673, + "learning_rate": 6.884110390989711e-06, + "loss": 0.3868, + "step": 17977 + }, + { + "epoch": 0.39572969403477876, + "grad_norm": 2.416004888482245, + "learning_rate": 6.8837802057699455e-06, + "loss": 0.3107, + "step": 17978 + }, + { + "epoch": 0.39575170592119746, + "grad_norm": 1.3292386003476115, + "learning_rate": 6.883450010975968e-06, + "loss": 0.2861, + "step": 17979 + }, + { + "epoch": 0.3957737178076161, + "grad_norm": 2.139445900825737, + "learning_rate": 6.883119806609458e-06, + "loss": 0.2657, + "step": 17980 + }, + { + "epoch": 0.39579572969403476, + "grad_norm": 1.3966456316046039, + "learning_rate": 6.8827895926720935e-06, + "loss": 0.2805, + "step": 17981 + }, + { + "epoch": 0.39581774158045346, + "grad_norm": 0.6641558858379394, + "learning_rate": 6.882459369165553e-06, + "loss": 0.5371, + "step": 17982 + }, + { + "epoch": 0.3958397534668721, + "grad_norm": 1.58422755165239, + "learning_rate": 6.882129136091513e-06, + "loss": 0.3009, + "step": 17983 + }, + { + "epoch": 0.39586176535329076, + "grad_norm": 1.9711695492940091, + "learning_rate": 6.881798893451654e-06, + "loss": 0.3473, + "step": 17984 + }, + { + "epoch": 0.39588377723970947, + "grad_norm": 1.2748953575306872, + "learning_rate": 6.881468641247655e-06, + "loss": 0.2681, + "step": 17985 + }, + { + "epoch": 0.3959057891261281, + "grad_norm": 1.9634803986017344, + "learning_rate": 6.8811383794811914e-06, + "loss": 0.3252, + "step": 17986 + }, + { + "epoch": 0.39592780101254677, + "grad_norm": 1.4345308120554674, + "learning_rate": 6.8808081081539444e-06, + "loss": 0.3736, + "step": 17987 + }, + { + "epoch": 0.3959498128989654, + "grad_norm": 1.2584433656839733, + "learning_rate": 6.880477827267592e-06, + "loss": 0.2255, + "step": 17988 + }, + { + "epoch": 0.3959718247853841, + "grad_norm": 1.4407816777740934, + "learning_rate": 6.880147536823812e-06, + "loss": 0.389, + "step": 17989 + }, + { + "epoch": 0.39599383667180277, + "grad_norm": 1.5704228407621126, + "learning_rate": 6.879817236824283e-06, + "loss": 0.2887, + "step": 17990 + }, + { + "epoch": 0.3960158485582214, + "grad_norm": 1.5875401600747003, + "learning_rate": 6.879486927270685e-06, + "loss": 0.2666, + "step": 17991 + }, + { + "epoch": 0.3960378604446401, + "grad_norm": 1.396695548829438, + "learning_rate": 6.8791566081646965e-06, + "loss": 0.4299, + "step": 17992 + }, + { + "epoch": 0.39605987233105877, + "grad_norm": 1.3909531642501278, + "learning_rate": 6.878826279507996e-06, + "loss": 0.3051, + "step": 17993 + }, + { + "epoch": 0.3960818842174774, + "grad_norm": 6.597675027883233, + "learning_rate": 6.878495941302262e-06, + "loss": 0.375, + "step": 17994 + }, + { + "epoch": 0.3961038961038961, + "grad_norm": 1.5006444256932003, + "learning_rate": 6.878165593549174e-06, + "loss": 0.2437, + "step": 17995 + }, + { + "epoch": 0.3961259079903148, + "grad_norm": 1.1607011838633352, + "learning_rate": 6.877835236250411e-06, + "loss": 0.3497, + "step": 17996 + }, + { + "epoch": 0.3961479198767334, + "grad_norm": 1.4102104107122464, + "learning_rate": 6.877504869407652e-06, + "loss": 0.3525, + "step": 17997 + }, + { + "epoch": 0.39616993176315213, + "grad_norm": 1.78887053229995, + "learning_rate": 6.877174493022575e-06, + "loss": 0.3611, + "step": 17998 + }, + { + "epoch": 0.3961919436495708, + "grad_norm": 1.2245956018230275, + "learning_rate": 6.876844107096859e-06, + "loss": 0.3027, + "step": 17999 + }, + { + "epoch": 0.3962139555359894, + "grad_norm": 1.5054650691592208, + "learning_rate": 6.876513711632187e-06, + "loss": 0.3025, + "step": 18000 + }, + { + "epoch": 0.3962359674224081, + "grad_norm": 1.4459527876262404, + "learning_rate": 6.876183306630233e-06, + "loss": 0.3304, + "step": 18001 + }, + { + "epoch": 0.3962579793088268, + "grad_norm": 1.3432304772902852, + "learning_rate": 6.87585289209268e-06, + "loss": 0.2688, + "step": 18002 + }, + { + "epoch": 0.39627999119524543, + "grad_norm": 1.2570817122709916, + "learning_rate": 6.875522468021205e-06, + "loss": 0.3682, + "step": 18003 + }, + { + "epoch": 0.3963020030816641, + "grad_norm": 1.1621866784341246, + "learning_rate": 6.875192034417489e-06, + "loss": 0.2365, + "step": 18004 + }, + { + "epoch": 0.3963240149680828, + "grad_norm": 1.3456517589568413, + "learning_rate": 6.874861591283211e-06, + "loss": 0.3785, + "step": 18005 + }, + { + "epoch": 0.39634602685450143, + "grad_norm": 1.864728894943572, + "learning_rate": 6.874531138620049e-06, + "loss": 0.3812, + "step": 18006 + }, + { + "epoch": 0.3963680387409201, + "grad_norm": 1.6885970540941508, + "learning_rate": 6.874200676429684e-06, + "loss": 0.3153, + "step": 18007 + }, + { + "epoch": 0.3963900506273388, + "grad_norm": 0.597485047922467, + "learning_rate": 6.873870204713796e-06, + "loss": 0.4754, + "step": 18008 + }, + { + "epoch": 0.39641206251375743, + "grad_norm": 1.3686555609234863, + "learning_rate": 6.8735397234740646e-06, + "loss": 0.3402, + "step": 18009 + }, + { + "epoch": 0.3964340744001761, + "grad_norm": 1.5086430654046579, + "learning_rate": 6.873209232712166e-06, + "loss": 0.2702, + "step": 18010 + }, + { + "epoch": 0.3964560862865948, + "grad_norm": 1.3878178248953583, + "learning_rate": 6.872878732429786e-06, + "loss": 0.3009, + "step": 18011 + }, + { + "epoch": 0.39647809817301344, + "grad_norm": 1.7237116812482072, + "learning_rate": 6.872548222628598e-06, + "loss": 0.3659, + "step": 18012 + }, + { + "epoch": 0.3965001100594321, + "grad_norm": 1.248634596228457, + "learning_rate": 6.8722177033102855e-06, + "loss": 0.3411, + "step": 18013 + }, + { + "epoch": 0.39652212194585074, + "grad_norm": 1.796300527155228, + "learning_rate": 6.871887174476528e-06, + "loss": 0.3343, + "step": 18014 + }, + { + "epoch": 0.39654413383226944, + "grad_norm": 1.3754684840309495, + "learning_rate": 6.871556636129005e-06, + "loss": 0.3651, + "step": 18015 + }, + { + "epoch": 0.3965661457186881, + "grad_norm": 1.3031793205968376, + "learning_rate": 6.871226088269397e-06, + "loss": 0.3064, + "step": 18016 + }, + { + "epoch": 0.39658815760510674, + "grad_norm": 1.309298673754934, + "learning_rate": 6.8708955308993815e-06, + "loss": 0.2992, + "step": 18017 + }, + { + "epoch": 0.39661016949152544, + "grad_norm": 1.3279765262420495, + "learning_rate": 6.8705649640206425e-06, + "loss": 0.3315, + "step": 18018 + }, + { + "epoch": 0.3966321813779441, + "grad_norm": 1.4185082186723903, + "learning_rate": 6.870234387634857e-06, + "loss": 0.2923, + "step": 18019 + }, + { + "epoch": 0.39665419326436274, + "grad_norm": 1.6892311496630654, + "learning_rate": 6.8699038017437066e-06, + "loss": 0.3921, + "step": 18020 + }, + { + "epoch": 0.39667620515078145, + "grad_norm": 1.7954733528684212, + "learning_rate": 6.8695732063488715e-06, + "loss": 0.2942, + "step": 18021 + }, + { + "epoch": 0.3966982170372001, + "grad_norm": 1.3288482883441062, + "learning_rate": 6.869242601452031e-06, + "loss": 0.3465, + "step": 18022 + }, + { + "epoch": 0.39672022892361875, + "grad_norm": 1.4656238953724934, + "learning_rate": 6.868911987054865e-06, + "loss": 0.357, + "step": 18023 + }, + { + "epoch": 0.3967422408100374, + "grad_norm": 1.6893300987556688, + "learning_rate": 6.868581363159057e-06, + "loss": 0.3544, + "step": 18024 + }, + { + "epoch": 0.3967642526964561, + "grad_norm": 1.5563087212466036, + "learning_rate": 6.868250729766284e-06, + "loss": 0.4276, + "step": 18025 + }, + { + "epoch": 0.39678626458287475, + "grad_norm": 1.2124074949248707, + "learning_rate": 6.867920086878227e-06, + "loss": 0.2925, + "step": 18026 + }, + { + "epoch": 0.3968082764692934, + "grad_norm": 1.5654979373626503, + "learning_rate": 6.867589434496567e-06, + "loss": 0.3565, + "step": 18027 + }, + { + "epoch": 0.3968302883557121, + "grad_norm": 1.7775788489565347, + "learning_rate": 6.8672587726229845e-06, + "loss": 0.3649, + "step": 18028 + }, + { + "epoch": 0.39685230024213075, + "grad_norm": 1.7266294392992085, + "learning_rate": 6.866928101259161e-06, + "loss": 0.2768, + "step": 18029 + }, + { + "epoch": 0.3968743121285494, + "grad_norm": 2.6744297996284234, + "learning_rate": 6.866597420406776e-06, + "loss": 0.3446, + "step": 18030 + }, + { + "epoch": 0.3968963240149681, + "grad_norm": 0.6299560472506613, + "learning_rate": 6.866266730067509e-06, + "loss": 0.4855, + "step": 18031 + }, + { + "epoch": 0.39691833590138675, + "grad_norm": 1.2724965975526574, + "learning_rate": 6.865936030243044e-06, + "loss": 0.2757, + "step": 18032 + }, + { + "epoch": 0.3969403477878054, + "grad_norm": 1.2952292764289348, + "learning_rate": 6.865605320935058e-06, + "loss": 0.2706, + "step": 18033 + }, + { + "epoch": 0.3969623596742241, + "grad_norm": 1.650966150303574, + "learning_rate": 6.865274602145236e-06, + "loss": 0.3625, + "step": 18034 + }, + { + "epoch": 0.39698437156064276, + "grad_norm": 0.6326185902744698, + "learning_rate": 6.864943873875254e-06, + "loss": 0.4932, + "step": 18035 + }, + { + "epoch": 0.3970063834470614, + "grad_norm": 1.7496014314876553, + "learning_rate": 6.864613136126796e-06, + "loss": 0.3179, + "step": 18036 + }, + { + "epoch": 0.39702839533348006, + "grad_norm": 1.3174157491509564, + "learning_rate": 6.864282388901544e-06, + "loss": 0.3891, + "step": 18037 + }, + { + "epoch": 0.39705040721989876, + "grad_norm": 0.6200399660448728, + "learning_rate": 6.863951632201176e-06, + "loss": 0.5115, + "step": 18038 + }, + { + "epoch": 0.3970724191063174, + "grad_norm": 1.2756492565035478, + "learning_rate": 6.863620866027375e-06, + "loss": 0.271, + "step": 18039 + }, + { + "epoch": 0.39709443099273606, + "grad_norm": 1.3254947017972318, + "learning_rate": 6.863290090381822e-06, + "loss": 0.3022, + "step": 18040 + }, + { + "epoch": 0.39711644287915476, + "grad_norm": 1.527818402659253, + "learning_rate": 6.862959305266197e-06, + "loss": 0.3026, + "step": 18041 + }, + { + "epoch": 0.3971384547655734, + "grad_norm": 0.9530097518651025, + "learning_rate": 6.862628510682181e-06, + "loss": 0.2528, + "step": 18042 + }, + { + "epoch": 0.39716046665199206, + "grad_norm": 1.8435303707673432, + "learning_rate": 6.862297706631457e-06, + "loss": 0.3142, + "step": 18043 + }, + { + "epoch": 0.39718247853841077, + "grad_norm": 1.7360223026478951, + "learning_rate": 6.861966893115707e-06, + "loss": 0.2572, + "step": 18044 + }, + { + "epoch": 0.3972044904248294, + "grad_norm": 1.332737276014049, + "learning_rate": 6.861636070136609e-06, + "loss": 0.306, + "step": 18045 + }, + { + "epoch": 0.39722650231124806, + "grad_norm": 1.5862889998814889, + "learning_rate": 6.861305237695845e-06, + "loss": 0.3125, + "step": 18046 + }, + { + "epoch": 0.3972485141976667, + "grad_norm": 2.4145903874246213, + "learning_rate": 6.860974395795102e-06, + "loss": 0.3055, + "step": 18047 + }, + { + "epoch": 0.3972705260840854, + "grad_norm": 1.1960092476871458, + "learning_rate": 6.860643544436053e-06, + "loss": 0.268, + "step": 18048 + }, + { + "epoch": 0.39729253797050407, + "grad_norm": 1.8432639550821288, + "learning_rate": 6.8603126836203844e-06, + "loss": 0.3257, + "step": 18049 + }, + { + "epoch": 0.3973145498569227, + "grad_norm": 0.6747380702857947, + "learning_rate": 6.859981813349778e-06, + "loss": 0.4947, + "step": 18050 + }, + { + "epoch": 0.3973365617433414, + "grad_norm": 1.5552246908263412, + "learning_rate": 6.8596509336259145e-06, + "loss": 0.3521, + "step": 18051 + }, + { + "epoch": 0.39735857362976007, + "grad_norm": 1.7714511408986395, + "learning_rate": 6.859320044450476e-06, + "loss": 0.4354, + "step": 18052 + }, + { + "epoch": 0.3973805855161787, + "grad_norm": 1.392078788447922, + "learning_rate": 6.858989145825141e-06, + "loss": 0.3266, + "step": 18053 + }, + { + "epoch": 0.3974025974025974, + "grad_norm": 1.7496659673385722, + "learning_rate": 6.858658237751596e-06, + "loss": 0.4069, + "step": 18054 + }, + { + "epoch": 0.3974246092890161, + "grad_norm": 1.2430668022238816, + "learning_rate": 6.858327320231521e-06, + "loss": 0.3234, + "step": 18055 + }, + { + "epoch": 0.3974466211754347, + "grad_norm": 0.6130099518035993, + "learning_rate": 6.857996393266598e-06, + "loss": 0.47, + "step": 18056 + }, + { + "epoch": 0.3974686330618534, + "grad_norm": 1.632314305344205, + "learning_rate": 6.857665456858509e-06, + "loss": 0.2924, + "step": 18057 + }, + { + "epoch": 0.3974906449482721, + "grad_norm": 0.6102273362203736, + "learning_rate": 6.857334511008935e-06, + "loss": 0.4938, + "step": 18058 + }, + { + "epoch": 0.3975126568346907, + "grad_norm": 1.6334068198978806, + "learning_rate": 6.857003555719557e-06, + "loss": 0.3236, + "step": 18059 + }, + { + "epoch": 0.3975346687211094, + "grad_norm": 2.8791298043611926, + "learning_rate": 6.856672590992059e-06, + "loss": 0.3247, + "step": 18060 + }, + { + "epoch": 0.3975566806075281, + "grad_norm": 1.4859743762585125, + "learning_rate": 6.856341616828125e-06, + "loss": 0.2802, + "step": 18061 + }, + { + "epoch": 0.39757869249394673, + "grad_norm": 1.778156074157537, + "learning_rate": 6.856010633229434e-06, + "loss": 0.2677, + "step": 18062 + }, + { + "epoch": 0.3976007043803654, + "grad_norm": 1.627062548057024, + "learning_rate": 6.855679640197669e-06, + "loss": 0.2935, + "step": 18063 + }, + { + "epoch": 0.3976227162667841, + "grad_norm": 20.68591158760833, + "learning_rate": 6.855348637734511e-06, + "loss": 0.3427, + "step": 18064 + }, + { + "epoch": 0.39764472815320273, + "grad_norm": 4.169811460454246, + "learning_rate": 6.855017625841647e-06, + "loss": 0.2887, + "step": 18065 + }, + { + "epoch": 0.3976667400396214, + "grad_norm": 1.3568567574968742, + "learning_rate": 6.854686604520752e-06, + "loss": 0.2544, + "step": 18066 + }, + { + "epoch": 0.3976887519260401, + "grad_norm": 1.8716174562812853, + "learning_rate": 6.854355573773516e-06, + "loss": 0.3191, + "step": 18067 + }, + { + "epoch": 0.39771076381245873, + "grad_norm": 1.3348556348894258, + "learning_rate": 6.854024533601616e-06, + "loss": 0.3539, + "step": 18068 + }, + { + "epoch": 0.3977327756988774, + "grad_norm": 2.6353970891780554, + "learning_rate": 6.853693484006736e-06, + "loss": 0.385, + "step": 18069 + }, + { + "epoch": 0.39775478758529603, + "grad_norm": 1.470766942531004, + "learning_rate": 6.853362424990561e-06, + "loss": 0.3614, + "step": 18070 + }, + { + "epoch": 0.39777679947171474, + "grad_norm": 1.5247605295957563, + "learning_rate": 6.853031356554771e-06, + "loss": 0.2952, + "step": 18071 + }, + { + "epoch": 0.3977988113581334, + "grad_norm": 0.6626207268158981, + "learning_rate": 6.852700278701048e-06, + "loss": 0.5013, + "step": 18072 + }, + { + "epoch": 0.39782082324455204, + "grad_norm": 2.855584042445974, + "learning_rate": 6.852369191431077e-06, + "loss": 0.326, + "step": 18073 + }, + { + "epoch": 0.39784283513097074, + "grad_norm": 1.866386883462099, + "learning_rate": 6.8520380947465385e-06, + "loss": 0.3097, + "step": 18074 + }, + { + "epoch": 0.3978648470173894, + "grad_norm": 2.491513521703607, + "learning_rate": 6.851706988649116e-06, + "loss": 0.2961, + "step": 18075 + }, + { + "epoch": 0.39788685890380804, + "grad_norm": 1.7200189530735865, + "learning_rate": 6.851375873140494e-06, + "loss": 0.3216, + "step": 18076 + }, + { + "epoch": 0.39790887079022674, + "grad_norm": 1.519324826884597, + "learning_rate": 6.851044748222353e-06, + "loss": 0.191, + "step": 18077 + }, + { + "epoch": 0.3979308826766454, + "grad_norm": 1.4197002456803312, + "learning_rate": 6.850713613896378e-06, + "loss": 0.3037, + "step": 18078 + }, + { + "epoch": 0.39795289456306404, + "grad_norm": 2.283250676523903, + "learning_rate": 6.8503824701642505e-06, + "loss": 0.3953, + "step": 18079 + }, + { + "epoch": 0.39797490644948275, + "grad_norm": 1.3414828709630833, + "learning_rate": 6.850051317027655e-06, + "loss": 0.2274, + "step": 18080 + }, + { + "epoch": 0.3979969183359014, + "grad_norm": 1.261665118935245, + "learning_rate": 6.849720154488272e-06, + "loss": 0.2897, + "step": 18081 + }, + { + "epoch": 0.39801893022232004, + "grad_norm": 1.545567322442988, + "learning_rate": 6.8493889825477864e-06, + "loss": 0.3199, + "step": 18082 + }, + { + "epoch": 0.3980409421087387, + "grad_norm": 0.6369737358115533, + "learning_rate": 6.849057801207881e-06, + "loss": 0.4776, + "step": 18083 + }, + { + "epoch": 0.3980629539951574, + "grad_norm": 1.8462246977249883, + "learning_rate": 6.84872661047024e-06, + "loss": 0.2885, + "step": 18084 + }, + { + "epoch": 0.39808496588157605, + "grad_norm": 1.9072464187861768, + "learning_rate": 6.848395410336546e-06, + "loss": 0.3774, + "step": 18085 + }, + { + "epoch": 0.3981069777679947, + "grad_norm": 2.973529672444521, + "learning_rate": 6.848064200808482e-06, + "loss": 0.2822, + "step": 18086 + }, + { + "epoch": 0.3981289896544134, + "grad_norm": 2.1713710439318636, + "learning_rate": 6.84773298188773e-06, + "loss": 0.3816, + "step": 18087 + }, + { + "epoch": 0.39815100154083205, + "grad_norm": 1.51612370482631, + "learning_rate": 6.847401753575977e-06, + "loss": 0.4677, + "step": 18088 + }, + { + "epoch": 0.3981730134272507, + "grad_norm": 1.3412313241446425, + "learning_rate": 6.847070515874904e-06, + "loss": 0.4529, + "step": 18089 + }, + { + "epoch": 0.3981950253136694, + "grad_norm": 2.3360758252003047, + "learning_rate": 6.846739268786194e-06, + "loss": 0.3854, + "step": 18090 + }, + { + "epoch": 0.39821703720008805, + "grad_norm": 1.337477711334321, + "learning_rate": 6.84640801231153e-06, + "loss": 0.3867, + "step": 18091 + }, + { + "epoch": 0.3982390490865067, + "grad_norm": 2.933482406126175, + "learning_rate": 6.846076746452598e-06, + "loss": 0.3233, + "step": 18092 + }, + { + "epoch": 0.3982610609729254, + "grad_norm": 0.6650969010728364, + "learning_rate": 6.845745471211083e-06, + "loss": 0.4964, + "step": 18093 + }, + { + "epoch": 0.39828307285934406, + "grad_norm": 1.6263503275407354, + "learning_rate": 6.8454141865886635e-06, + "loss": 0.2862, + "step": 18094 + }, + { + "epoch": 0.3983050847457627, + "grad_norm": 1.2921205938202611, + "learning_rate": 6.845082892587027e-06, + "loss": 0.3422, + "step": 18095 + }, + { + "epoch": 0.39832709663218135, + "grad_norm": 1.5039148138033325, + "learning_rate": 6.8447515892078565e-06, + "loss": 0.3116, + "step": 18096 + }, + { + "epoch": 0.39834910851860006, + "grad_norm": 1.4823135252473787, + "learning_rate": 6.844420276452834e-06, + "loss": 0.3135, + "step": 18097 + }, + { + "epoch": 0.3983711204050187, + "grad_norm": 0.618760607096405, + "learning_rate": 6.844088954323648e-06, + "loss": 0.4837, + "step": 18098 + }, + { + "epoch": 0.39839313229143736, + "grad_norm": 22.924520871184253, + "learning_rate": 6.843757622821977e-06, + "loss": 0.3044, + "step": 18099 + }, + { + "epoch": 0.39841514417785606, + "grad_norm": 0.6194359393153699, + "learning_rate": 6.843426281949508e-06, + "loss": 0.5192, + "step": 18100 + }, + { + "epoch": 0.3984371560642747, + "grad_norm": 1.4008592993154383, + "learning_rate": 6.843094931707924e-06, + "loss": 0.284, + "step": 18101 + }, + { + "epoch": 0.39845916795069336, + "grad_norm": 1.4116910529341067, + "learning_rate": 6.84276357209891e-06, + "loss": 0.3429, + "step": 18102 + }, + { + "epoch": 0.39848117983711207, + "grad_norm": 1.4433679819664318, + "learning_rate": 6.84243220312415e-06, + "loss": 0.3396, + "step": 18103 + }, + { + "epoch": 0.3985031917235307, + "grad_norm": 0.5908027075591954, + "learning_rate": 6.842100824785328e-06, + "loss": 0.4549, + "step": 18104 + }, + { + "epoch": 0.39852520360994936, + "grad_norm": 1.7468763829915526, + "learning_rate": 6.841769437084127e-06, + "loss": 0.425, + "step": 18105 + }, + { + "epoch": 0.398547215496368, + "grad_norm": 2.1198587222682503, + "learning_rate": 6.841438040022235e-06, + "loss": 0.3414, + "step": 18106 + }, + { + "epoch": 0.3985692273827867, + "grad_norm": 1.0100208655134062, + "learning_rate": 6.841106633601331e-06, + "loss": 0.2162, + "step": 18107 + }, + { + "epoch": 0.39859123926920537, + "grad_norm": 1.6236853328335843, + "learning_rate": 6.840775217823101e-06, + "loss": 0.3538, + "step": 18108 + }, + { + "epoch": 0.398613251155624, + "grad_norm": 1.1512643457698541, + "learning_rate": 6.840443792689232e-06, + "loss": 0.2391, + "step": 18109 + }, + { + "epoch": 0.3986352630420427, + "grad_norm": 0.6535683578779032, + "learning_rate": 6.840112358201406e-06, + "loss": 0.4974, + "step": 18110 + }, + { + "epoch": 0.39865727492846137, + "grad_norm": 0.6241349945726548, + "learning_rate": 6.8397809143613105e-06, + "loss": 0.489, + "step": 18111 + }, + { + "epoch": 0.39867928681488, + "grad_norm": 0.6335786464850884, + "learning_rate": 6.839449461170626e-06, + "loss": 0.4988, + "step": 18112 + }, + { + "epoch": 0.3987012987012987, + "grad_norm": 1.9143529323314694, + "learning_rate": 6.839117998631038e-06, + "loss": 0.3981, + "step": 18113 + }, + { + "epoch": 0.39872331058771737, + "grad_norm": 3.3438724727016536, + "learning_rate": 6.838786526744233e-06, + "loss": 0.3042, + "step": 18114 + }, + { + "epoch": 0.398745322474136, + "grad_norm": 1.4130444982238801, + "learning_rate": 6.838455045511896e-06, + "loss": 0.2534, + "step": 18115 + }, + { + "epoch": 0.3987673343605547, + "grad_norm": 1.4708354172490636, + "learning_rate": 6.838123554935708e-06, + "loss": 0.2, + "step": 18116 + }, + { + "epoch": 0.3987893462469734, + "grad_norm": 1.758923450344646, + "learning_rate": 6.837792055017358e-06, + "loss": 0.3985, + "step": 18117 + }, + { + "epoch": 0.398811358133392, + "grad_norm": 1.8130571413370884, + "learning_rate": 6.837460545758529e-06, + "loss": 0.3421, + "step": 18118 + }, + { + "epoch": 0.3988333700198107, + "grad_norm": 1.6674498875749546, + "learning_rate": 6.837129027160905e-06, + "loss": 0.3743, + "step": 18119 + }, + { + "epoch": 0.3988553819062294, + "grad_norm": 1.326296618340492, + "learning_rate": 6.836797499226173e-06, + "loss": 0.2812, + "step": 18120 + }, + { + "epoch": 0.398877393792648, + "grad_norm": 1.4332069519428743, + "learning_rate": 6.836465961956016e-06, + "loss": 0.3487, + "step": 18121 + }, + { + "epoch": 0.3988994056790667, + "grad_norm": 1.6072820293649208, + "learning_rate": 6.836134415352121e-06, + "loss": 0.3004, + "step": 18122 + }, + { + "epoch": 0.3989214175654854, + "grad_norm": 1.6522954702111867, + "learning_rate": 6.83580285941617e-06, + "loss": 0.3891, + "step": 18123 + }, + { + "epoch": 0.39894342945190403, + "grad_norm": 1.3701892106635236, + "learning_rate": 6.835471294149854e-06, + "loss": 0.2763, + "step": 18124 + }, + { + "epoch": 0.3989654413383227, + "grad_norm": 1.4157696615149864, + "learning_rate": 6.83513971955485e-06, + "loss": 0.3424, + "step": 18125 + }, + { + "epoch": 0.3989874532247414, + "grad_norm": 1.671804098129028, + "learning_rate": 6.8348081356328485e-06, + "loss": 0.2978, + "step": 18126 + }, + { + "epoch": 0.39900946511116003, + "grad_norm": 3.6348277109344878, + "learning_rate": 6.834476542385535e-06, + "loss": 0.3545, + "step": 18127 + }, + { + "epoch": 0.3990314769975787, + "grad_norm": 0.8549149391270425, + "learning_rate": 6.834144939814593e-06, + "loss": 0.5215, + "step": 18128 + }, + { + "epoch": 0.39905348888399733, + "grad_norm": 1.280032417069085, + "learning_rate": 6.833813327921708e-06, + "loss": 0.31, + "step": 18129 + }, + { + "epoch": 0.39907550077041604, + "grad_norm": 0.7300952706489175, + "learning_rate": 6.833481706708567e-06, + "loss": 0.5113, + "step": 18130 + }, + { + "epoch": 0.3990975126568347, + "grad_norm": 1.4264398705851784, + "learning_rate": 6.833150076176851e-06, + "loss": 0.3588, + "step": 18131 + }, + { + "epoch": 0.39911952454325333, + "grad_norm": 1.5298963122070919, + "learning_rate": 6.832818436328251e-06, + "loss": 0.3729, + "step": 18132 + }, + { + "epoch": 0.39914153642967204, + "grad_norm": 1.3515395267737549, + "learning_rate": 6.8324867871644504e-06, + "loss": 0.2623, + "step": 18133 + }, + { + "epoch": 0.3991635483160907, + "grad_norm": 1.6527771228268124, + "learning_rate": 6.832155128687135e-06, + "loss": 0.2594, + "step": 18134 + }, + { + "epoch": 0.39918556020250934, + "grad_norm": 1.335007315429478, + "learning_rate": 6.831823460897989e-06, + "loss": 0.2944, + "step": 18135 + }, + { + "epoch": 0.39920757208892804, + "grad_norm": 1.3039649840536132, + "learning_rate": 6.8314917837987e-06, + "loss": 0.2364, + "step": 18136 + }, + { + "epoch": 0.3992295839753467, + "grad_norm": 0.7775625032435867, + "learning_rate": 6.831160097390952e-06, + "loss": 0.5366, + "step": 18137 + }, + { + "epoch": 0.39925159586176534, + "grad_norm": 1.6832210967601968, + "learning_rate": 6.830828401676433e-06, + "loss": 0.2747, + "step": 18138 + }, + { + "epoch": 0.39927360774818405, + "grad_norm": 0.7229486964441801, + "learning_rate": 6.830496696656826e-06, + "loss": 0.4917, + "step": 18139 + }, + { + "epoch": 0.3992956196346027, + "grad_norm": 3.1969317896310696, + "learning_rate": 6.830164982333819e-06, + "loss": 0.3897, + "step": 18140 + }, + { + "epoch": 0.39931763152102134, + "grad_norm": 0.6462870594589374, + "learning_rate": 6.8298332587090966e-06, + "loss": 0.5062, + "step": 18141 + }, + { + "epoch": 0.39933964340744, + "grad_norm": 1.7780967715273757, + "learning_rate": 6.829501525784346e-06, + "loss": 0.3832, + "step": 18142 + }, + { + "epoch": 0.3993616552938587, + "grad_norm": 1.2442692608481425, + "learning_rate": 6.829169783561253e-06, + "loss": 0.2842, + "step": 18143 + }, + { + "epoch": 0.39938366718027735, + "grad_norm": 1.9210559134646985, + "learning_rate": 6.8288380320415026e-06, + "loss": 0.3511, + "step": 18144 + }, + { + "epoch": 0.399405679066696, + "grad_norm": 1.430640748646688, + "learning_rate": 6.828506271226782e-06, + "loss": 0.3237, + "step": 18145 + }, + { + "epoch": 0.3994276909531147, + "grad_norm": 1.480433304738662, + "learning_rate": 6.828174501118778e-06, + "loss": 0.3371, + "step": 18146 + }, + { + "epoch": 0.39944970283953335, + "grad_norm": 1.4342369066100762, + "learning_rate": 6.827842721719175e-06, + "loss": 0.3708, + "step": 18147 + }, + { + "epoch": 0.399471714725952, + "grad_norm": 1.2780510971397472, + "learning_rate": 6.82751093302966e-06, + "loss": 0.2961, + "step": 18148 + }, + { + "epoch": 0.3994937266123707, + "grad_norm": 2.6508999675210094, + "learning_rate": 6.827179135051919e-06, + "loss": 0.322, + "step": 18149 + }, + { + "epoch": 0.39951573849878935, + "grad_norm": 1.9392182423606068, + "learning_rate": 6.826847327787639e-06, + "loss": 0.2514, + "step": 18150 + }, + { + "epoch": 0.399537750385208, + "grad_norm": 1.361461348604318, + "learning_rate": 6.8265155112385065e-06, + "loss": 0.3587, + "step": 18151 + }, + { + "epoch": 0.3995597622716267, + "grad_norm": 1.4919564307702444, + "learning_rate": 6.826183685406207e-06, + "loss": 0.3227, + "step": 18152 + }, + { + "epoch": 0.39958177415804536, + "grad_norm": 1.5715961942298629, + "learning_rate": 6.825851850292428e-06, + "loss": 0.2957, + "step": 18153 + }, + { + "epoch": 0.399603786044464, + "grad_norm": 6.335181802936324, + "learning_rate": 6.825520005898853e-06, + "loss": 0.3601, + "step": 18154 + }, + { + "epoch": 0.39962579793088265, + "grad_norm": 0.9160771968222656, + "learning_rate": 6.825188152227175e-06, + "loss": 0.5069, + "step": 18155 + }, + { + "epoch": 0.39964780981730136, + "grad_norm": 0.7748476286894246, + "learning_rate": 6.824856289279075e-06, + "loss": 0.5019, + "step": 18156 + }, + { + "epoch": 0.39966982170372, + "grad_norm": 2.8939767768910496, + "learning_rate": 6.824524417056241e-06, + "loss": 0.3537, + "step": 18157 + }, + { + "epoch": 0.39969183359013866, + "grad_norm": 1.3144796094650169, + "learning_rate": 6.824192535560361e-06, + "loss": 0.3095, + "step": 18158 + }, + { + "epoch": 0.39971384547655736, + "grad_norm": 2.2133451999827494, + "learning_rate": 6.82386064479312e-06, + "loss": 0.3474, + "step": 18159 + }, + { + "epoch": 0.399735857362976, + "grad_norm": 1.3458564296163733, + "learning_rate": 6.823528744756206e-06, + "loss": 0.3681, + "step": 18160 + }, + { + "epoch": 0.39975786924939466, + "grad_norm": 0.7529029451090429, + "learning_rate": 6.823196835451307e-06, + "loss": 0.5185, + "step": 18161 + }, + { + "epoch": 0.39977988113581336, + "grad_norm": 1.3094831212649132, + "learning_rate": 6.822864916880107e-06, + "loss": 0.3395, + "step": 18162 + }, + { + "epoch": 0.399801893022232, + "grad_norm": 2.0375741186354364, + "learning_rate": 6.822532989044295e-06, + "loss": 0.3023, + "step": 18163 + }, + { + "epoch": 0.39982390490865066, + "grad_norm": 1.583434414758628, + "learning_rate": 6.8222010519455575e-06, + "loss": 0.3566, + "step": 18164 + }, + { + "epoch": 0.3998459167950693, + "grad_norm": 0.8137885143903631, + "learning_rate": 6.821869105585581e-06, + "loss": 0.4996, + "step": 18165 + }, + { + "epoch": 0.399867928681488, + "grad_norm": 1.7047953811821601, + "learning_rate": 6.821537149966054e-06, + "loss": 0.3565, + "step": 18166 + }, + { + "epoch": 0.39988994056790667, + "grad_norm": 1.3093514737676843, + "learning_rate": 6.821205185088662e-06, + "loss": 0.3613, + "step": 18167 + }, + { + "epoch": 0.3999119524543253, + "grad_norm": 0.6884682413938907, + "learning_rate": 6.8208732109550925e-06, + "loss": 0.5149, + "step": 18168 + }, + { + "epoch": 0.399933964340744, + "grad_norm": 1.5520074241085804, + "learning_rate": 6.820541227567034e-06, + "loss": 0.3499, + "step": 18169 + }, + { + "epoch": 0.39995597622716267, + "grad_norm": 0.6118246343193753, + "learning_rate": 6.8202092349261735e-06, + "loss": 0.4903, + "step": 18170 + }, + { + "epoch": 0.3999779881135813, + "grad_norm": 1.3293070491391659, + "learning_rate": 6.819877233034198e-06, + "loss": 0.3093, + "step": 18171 + }, + { + "epoch": 0.4, + "grad_norm": 0.6432793439334487, + "learning_rate": 6.819545221892792e-06, + "loss": 0.4848, + "step": 18172 + }, + { + "epoch": 0.40002201188641867, + "grad_norm": 3.23448936869306, + "learning_rate": 6.819213201503648e-06, + "loss": 0.3264, + "step": 18173 + }, + { + "epoch": 0.4000440237728373, + "grad_norm": 1.4061404969414304, + "learning_rate": 6.8188811718684515e-06, + "loss": 0.3057, + "step": 18174 + }, + { + "epoch": 0.400066035659256, + "grad_norm": 2.1867549122395316, + "learning_rate": 6.818549132988889e-06, + "loss": 0.4135, + "step": 18175 + }, + { + "epoch": 0.4000880475456747, + "grad_norm": 1.221685114534512, + "learning_rate": 6.8182170848666485e-06, + "loss": 0.2061, + "step": 18176 + }, + { + "epoch": 0.4001100594320933, + "grad_norm": 1.2968824653407263, + "learning_rate": 6.817885027503417e-06, + "loss": 0.2844, + "step": 18177 + }, + { + "epoch": 0.400132071318512, + "grad_norm": 1.4576910255666229, + "learning_rate": 6.817552960900884e-06, + "loss": 0.2685, + "step": 18178 + }, + { + "epoch": 0.4001540832049307, + "grad_norm": 1.4184873595562684, + "learning_rate": 6.817220885060738e-06, + "loss": 0.3365, + "step": 18179 + }, + { + "epoch": 0.4001760950913493, + "grad_norm": 2.1069001846065887, + "learning_rate": 6.816888799984662e-06, + "loss": 0.3349, + "step": 18180 + }, + { + "epoch": 0.400198106977768, + "grad_norm": 1.6378789992001954, + "learning_rate": 6.8165567056743476e-06, + "loss": 0.2842, + "step": 18181 + }, + { + "epoch": 0.4002201188641867, + "grad_norm": 1.479681062218621, + "learning_rate": 6.8162246021314825e-06, + "loss": 0.3181, + "step": 18182 + }, + { + "epoch": 0.40024213075060533, + "grad_norm": 1.3903560573314877, + "learning_rate": 6.815892489357752e-06, + "loss": 0.293, + "step": 18183 + }, + { + "epoch": 0.400264142637024, + "grad_norm": 1.35657198161357, + "learning_rate": 6.815560367354849e-06, + "loss": 0.2595, + "step": 18184 + }, + { + "epoch": 0.4002861545234427, + "grad_norm": 1.6014131566694816, + "learning_rate": 6.815228236124456e-06, + "loss": 0.3331, + "step": 18185 + }, + { + "epoch": 0.40030816640986133, + "grad_norm": 2.203752904128692, + "learning_rate": 6.814896095668264e-06, + "loss": 0.3081, + "step": 18186 + }, + { + "epoch": 0.40033017829628, + "grad_norm": 1.164051829730383, + "learning_rate": 6.814563945987962e-06, + "loss": 0.2497, + "step": 18187 + }, + { + "epoch": 0.40035219018269863, + "grad_norm": 1.3847313479571888, + "learning_rate": 6.814231787085236e-06, + "loss": 0.292, + "step": 18188 + }, + { + "epoch": 0.40037420206911734, + "grad_norm": 2.495281029696839, + "learning_rate": 6.813899618961775e-06, + "loss": 0.3917, + "step": 18189 + }, + { + "epoch": 0.400396213955536, + "grad_norm": 0.780595643815459, + "learning_rate": 6.813567441619265e-06, + "loss": 0.4998, + "step": 18190 + }, + { + "epoch": 0.40041822584195463, + "grad_norm": 0.7333867398916278, + "learning_rate": 6.813235255059398e-06, + "loss": 0.506, + "step": 18191 + }, + { + "epoch": 0.40044023772837334, + "grad_norm": 0.6696914983226939, + "learning_rate": 6.812903059283862e-06, + "loss": 0.5221, + "step": 18192 + }, + { + "epoch": 0.400462249614792, + "grad_norm": 1.9220254238210168, + "learning_rate": 6.812570854294342e-06, + "loss": 0.3648, + "step": 18193 + }, + { + "epoch": 0.40048426150121064, + "grad_norm": 1.1343446684948468, + "learning_rate": 6.81223864009253e-06, + "loss": 0.248, + "step": 18194 + }, + { + "epoch": 0.40050627338762934, + "grad_norm": 1.8491229494688841, + "learning_rate": 6.811906416680111e-06, + "loss": 0.3228, + "step": 18195 + }, + { + "epoch": 0.400528285274048, + "grad_norm": 1.5181655456219225, + "learning_rate": 6.811574184058777e-06, + "loss": 0.2861, + "step": 18196 + }, + { + "epoch": 0.40055029716046664, + "grad_norm": 1.5445265575758709, + "learning_rate": 6.811241942230216e-06, + "loss": 0.3138, + "step": 18197 + }, + { + "epoch": 0.40057230904688534, + "grad_norm": 1.927475997004713, + "learning_rate": 6.8109096911961135e-06, + "loss": 0.3509, + "step": 18198 + }, + { + "epoch": 0.400594320933304, + "grad_norm": 1.5309319278311921, + "learning_rate": 6.8105774309581604e-06, + "loss": 0.2787, + "step": 18199 + }, + { + "epoch": 0.40061633281972264, + "grad_norm": 1.3365320067033994, + "learning_rate": 6.810245161518047e-06, + "loss": 0.4411, + "step": 18200 + }, + { + "epoch": 0.4006383447061413, + "grad_norm": 1.5671155196057758, + "learning_rate": 6.809912882877459e-06, + "loss": 0.3907, + "step": 18201 + }, + { + "epoch": 0.40066035659256, + "grad_norm": 1.7164490757681252, + "learning_rate": 6.809580595038088e-06, + "loss": 0.3409, + "step": 18202 + }, + { + "epoch": 0.40068236847897865, + "grad_norm": 1.943053610453913, + "learning_rate": 6.809248298001618e-06, + "loss": 0.309, + "step": 18203 + }, + { + "epoch": 0.4007043803653973, + "grad_norm": 1.5830273541105842, + "learning_rate": 6.808915991769745e-06, + "loss": 0.2617, + "step": 18204 + }, + { + "epoch": 0.400726392251816, + "grad_norm": 1.7400017799503333, + "learning_rate": 6.808583676344151e-06, + "loss": 0.3361, + "step": 18205 + }, + { + "epoch": 0.40074840413823465, + "grad_norm": 1.4820233630822628, + "learning_rate": 6.8082513517265306e-06, + "loss": 0.3092, + "step": 18206 + }, + { + "epoch": 0.4007704160246533, + "grad_norm": 0.9800761678957842, + "learning_rate": 6.8079190179185695e-06, + "loss": 0.4992, + "step": 18207 + }, + { + "epoch": 0.400792427911072, + "grad_norm": 1.6916623838414933, + "learning_rate": 6.807586674921955e-06, + "loss": 0.3194, + "step": 18208 + }, + { + "epoch": 0.40081443979749065, + "grad_norm": 1.1395884222124948, + "learning_rate": 6.8072543227383825e-06, + "loss": 0.2391, + "step": 18209 + }, + { + "epoch": 0.4008364516839093, + "grad_norm": 1.2866153043702409, + "learning_rate": 6.806921961369536e-06, + "loss": 0.304, + "step": 18210 + }, + { + "epoch": 0.40085846357032795, + "grad_norm": 1.3610889035768523, + "learning_rate": 6.806589590817106e-06, + "loss": 0.3026, + "step": 18211 + }, + { + "epoch": 0.40088047545674665, + "grad_norm": 0.6684589752681316, + "learning_rate": 6.806257211082783e-06, + "loss": 0.4938, + "step": 18212 + }, + { + "epoch": 0.4009024873431653, + "grad_norm": 0.6355563829710901, + "learning_rate": 6.805924822168254e-06, + "loss": 0.4903, + "step": 18213 + }, + { + "epoch": 0.40092449922958395, + "grad_norm": 1.3074773212135726, + "learning_rate": 6.805592424075209e-06, + "loss": 0.3225, + "step": 18214 + }, + { + "epoch": 0.40094651111600266, + "grad_norm": 1.3299310241704585, + "learning_rate": 6.805260016805342e-06, + "loss": 0.2463, + "step": 18215 + }, + { + "epoch": 0.4009685230024213, + "grad_norm": 1.7506993841312828, + "learning_rate": 6.804927600360334e-06, + "loss": 0.3711, + "step": 18216 + }, + { + "epoch": 0.40099053488883996, + "grad_norm": 1.4446341485319765, + "learning_rate": 6.80459517474188e-06, + "loss": 0.2443, + "step": 18217 + }, + { + "epoch": 0.40101254677525866, + "grad_norm": 1.322227926119691, + "learning_rate": 6.804262739951669e-06, + "loss": 0.3295, + "step": 18218 + }, + { + "epoch": 0.4010345586616773, + "grad_norm": 1.464987928031249, + "learning_rate": 6.80393029599139e-06, + "loss": 0.2995, + "step": 18219 + }, + { + "epoch": 0.40105657054809596, + "grad_norm": 1.6727975588541093, + "learning_rate": 6.803597842862733e-06, + "loss": 0.26, + "step": 18220 + }, + { + "epoch": 0.40107858243451466, + "grad_norm": 1.706361086005788, + "learning_rate": 6.803265380567386e-06, + "loss": 0.2755, + "step": 18221 + }, + { + "epoch": 0.4011005943209333, + "grad_norm": 1.4909022967988057, + "learning_rate": 6.802932909107041e-06, + "loss": 0.3805, + "step": 18222 + }, + { + "epoch": 0.40112260620735196, + "grad_norm": 1.835759429264371, + "learning_rate": 6.802600428483387e-06, + "loss": 0.3898, + "step": 18223 + }, + { + "epoch": 0.4011446180937706, + "grad_norm": 0.902654858950652, + "learning_rate": 6.802267938698114e-06, + "loss": 0.5104, + "step": 18224 + }, + { + "epoch": 0.4011666299801893, + "grad_norm": 1.5329449998287001, + "learning_rate": 6.801935439752911e-06, + "loss": 0.3024, + "step": 18225 + }, + { + "epoch": 0.40118864186660796, + "grad_norm": 1.5117079216617242, + "learning_rate": 6.801602931649467e-06, + "loss": 0.2783, + "step": 18226 + }, + { + "epoch": 0.4012106537530266, + "grad_norm": 1.7183768218790052, + "learning_rate": 6.801270414389475e-06, + "loss": 0.3176, + "step": 18227 + }, + { + "epoch": 0.4012326656394453, + "grad_norm": 0.7084548528668904, + "learning_rate": 6.8009378879746235e-06, + "loss": 0.475, + "step": 18228 + }, + { + "epoch": 0.40125467752586397, + "grad_norm": 1.407266923993952, + "learning_rate": 6.800605352406602e-06, + "loss": 0.2866, + "step": 18229 + }, + { + "epoch": 0.4012766894122826, + "grad_norm": 1.5386604281581542, + "learning_rate": 6.8002728076871005e-06, + "loss": 0.3146, + "step": 18230 + }, + { + "epoch": 0.4012987012987013, + "grad_norm": 1.8602062416012388, + "learning_rate": 6.79994025381781e-06, + "loss": 0.3394, + "step": 18231 + }, + { + "epoch": 0.40132071318511997, + "grad_norm": 0.6465781710516197, + "learning_rate": 6.7996076908004205e-06, + "loss": 0.5164, + "step": 18232 + }, + { + "epoch": 0.4013427250715386, + "grad_norm": 1.3273532863118265, + "learning_rate": 6.7992751186366215e-06, + "loss": 0.2611, + "step": 18233 + }, + { + "epoch": 0.4013647369579573, + "grad_norm": 1.5517167652539194, + "learning_rate": 6.798942537328105e-06, + "loss": 0.3423, + "step": 18234 + }, + { + "epoch": 0.401386748844376, + "grad_norm": 1.9042518903878007, + "learning_rate": 6.7986099468765595e-06, + "loss": 0.3784, + "step": 18235 + }, + { + "epoch": 0.4014087607307946, + "grad_norm": 0.6484492570465338, + "learning_rate": 6.798277347283676e-06, + "loss": 0.4698, + "step": 18236 + }, + { + "epoch": 0.40143077261721327, + "grad_norm": 1.5804212037831409, + "learning_rate": 6.797944738551144e-06, + "loss": 0.392, + "step": 18237 + }, + { + "epoch": 0.401452784503632, + "grad_norm": 1.634753238014445, + "learning_rate": 6.797612120680658e-06, + "loss": 0.2985, + "step": 18238 + }, + { + "epoch": 0.4014747963900506, + "grad_norm": 1.7245977121397602, + "learning_rate": 6.797279493673903e-06, + "loss": 0.4284, + "step": 18239 + }, + { + "epoch": 0.4014968082764693, + "grad_norm": 1.3678372547137267, + "learning_rate": 6.796946857532571e-06, + "loss": 0.2628, + "step": 18240 + }, + { + "epoch": 0.401518820162888, + "grad_norm": 2.0277507478370804, + "learning_rate": 6.796614212258357e-06, + "loss": 0.3662, + "step": 18241 + }, + { + "epoch": 0.40154083204930663, + "grad_norm": 1.2670339834668969, + "learning_rate": 6.796281557852944e-06, + "loss": 0.4136, + "step": 18242 + }, + { + "epoch": 0.4015628439357253, + "grad_norm": 1.731416527101924, + "learning_rate": 6.7959488943180305e-06, + "loss": 0.3931, + "step": 18243 + }, + { + "epoch": 0.401584855822144, + "grad_norm": 1.5838298590206012, + "learning_rate": 6.795616221655302e-06, + "loss": 0.3119, + "step": 18244 + }, + { + "epoch": 0.40160686770856263, + "grad_norm": 1.8521960580011174, + "learning_rate": 6.795283539866452e-06, + "loss": 0.3184, + "step": 18245 + }, + { + "epoch": 0.4016288795949813, + "grad_norm": 0.6520807976644803, + "learning_rate": 6.79495084895317e-06, + "loss": 0.5003, + "step": 18246 + }, + { + "epoch": 0.40165089148139993, + "grad_norm": 1.6430326924445133, + "learning_rate": 6.794618148917147e-06, + "loss": 0.346, + "step": 18247 + }, + { + "epoch": 0.40167290336781863, + "grad_norm": 1.7255037221818676, + "learning_rate": 6.7942854397600746e-06, + "loss": 0.3445, + "step": 18248 + }, + { + "epoch": 0.4016949152542373, + "grad_norm": 1.4656594726936605, + "learning_rate": 6.793952721483641e-06, + "loss": 0.4009, + "step": 18249 + }, + { + "epoch": 0.40171692714065593, + "grad_norm": 0.6205356370958514, + "learning_rate": 6.793619994089542e-06, + "loss": 0.4914, + "step": 18250 + }, + { + "epoch": 0.40173893902707464, + "grad_norm": 1.3779098522205129, + "learning_rate": 6.793287257579466e-06, + "loss": 0.2517, + "step": 18251 + }, + { + "epoch": 0.4017609509134933, + "grad_norm": 1.3274998655102104, + "learning_rate": 6.792954511955104e-06, + "loss": 0.3132, + "step": 18252 + }, + { + "epoch": 0.40178296279991194, + "grad_norm": 2.221315930193603, + "learning_rate": 6.7926217572181474e-06, + "loss": 0.3375, + "step": 18253 + }, + { + "epoch": 0.40180497468633064, + "grad_norm": 0.6083445723276911, + "learning_rate": 6.7922889933702884e-06, + "loss": 0.4694, + "step": 18254 + }, + { + "epoch": 0.4018269865727493, + "grad_norm": 1.4404518487089362, + "learning_rate": 6.791956220413215e-06, + "loss": 0.3303, + "step": 18255 + }, + { + "epoch": 0.40184899845916794, + "grad_norm": 1.5106182294910433, + "learning_rate": 6.7916234383486235e-06, + "loss": 0.3689, + "step": 18256 + }, + { + "epoch": 0.40187101034558664, + "grad_norm": 1.6257491828975625, + "learning_rate": 6.7912906471782015e-06, + "loss": 0.3395, + "step": 18257 + }, + { + "epoch": 0.4018930222320053, + "grad_norm": 1.2052955586849345, + "learning_rate": 6.790957846903641e-06, + "loss": 0.3287, + "step": 18258 + }, + { + "epoch": 0.40191503411842394, + "grad_norm": 0.6307572107594243, + "learning_rate": 6.790625037526635e-06, + "loss": 0.4819, + "step": 18259 + }, + { + "epoch": 0.4019370460048426, + "grad_norm": 1.586643913986849, + "learning_rate": 6.790292219048872e-06, + "loss": 0.2434, + "step": 18260 + }, + { + "epoch": 0.4019590578912613, + "grad_norm": 1.4997243899340447, + "learning_rate": 6.789959391472047e-06, + "loss": 0.2738, + "step": 18261 + }, + { + "epoch": 0.40198106977767994, + "grad_norm": 1.7190099744296037, + "learning_rate": 6.7896265547978506e-06, + "loss": 0.3485, + "step": 18262 + }, + { + "epoch": 0.4020030816640986, + "grad_norm": 3.4710024297466004, + "learning_rate": 6.789293709027972e-06, + "loss": 0.388, + "step": 18263 + }, + { + "epoch": 0.4020250935505173, + "grad_norm": 1.3063093531066858, + "learning_rate": 6.788960854164106e-06, + "loss": 0.3607, + "step": 18264 + }, + { + "epoch": 0.40204710543693595, + "grad_norm": 0.6297108563143957, + "learning_rate": 6.788627990207943e-06, + "loss": 0.5024, + "step": 18265 + }, + { + "epoch": 0.4020691173233546, + "grad_norm": 1.157259786771357, + "learning_rate": 6.788295117161174e-06, + "loss": 0.3017, + "step": 18266 + }, + { + "epoch": 0.4020911292097733, + "grad_norm": 1.2391322732836485, + "learning_rate": 6.787962235025493e-06, + "loss": 0.3748, + "step": 18267 + }, + { + "epoch": 0.40211314109619195, + "grad_norm": 1.5506284483248436, + "learning_rate": 6.7876293438025886e-06, + "loss": 0.388, + "step": 18268 + }, + { + "epoch": 0.4021351529826106, + "grad_norm": 0.6257124134651749, + "learning_rate": 6.787296443494156e-06, + "loss": 0.4894, + "step": 18269 + }, + { + "epoch": 0.40215716486902925, + "grad_norm": 1.4968414916770105, + "learning_rate": 6.786963534101886e-06, + "loss": 0.3394, + "step": 18270 + }, + { + "epoch": 0.40217917675544795, + "grad_norm": 1.4867305926384864, + "learning_rate": 6.7866306156274695e-06, + "loss": 0.3128, + "step": 18271 + }, + { + "epoch": 0.4022011886418666, + "grad_norm": 1.3553925189520222, + "learning_rate": 6.7862976880726e-06, + "loss": 0.3614, + "step": 18272 + }, + { + "epoch": 0.40222320052828525, + "grad_norm": 1.6539083423250722, + "learning_rate": 6.785964751438967e-06, + "loss": 0.288, + "step": 18273 + }, + { + "epoch": 0.40224521241470396, + "grad_norm": 1.1840561388931885, + "learning_rate": 6.785631805728268e-06, + "loss": 0.2251, + "step": 18274 + }, + { + "epoch": 0.4022672243011226, + "grad_norm": 1.2952068743414245, + "learning_rate": 6.785298850942188e-06, + "loss": 0.3433, + "step": 18275 + }, + { + "epoch": 0.40228923618754125, + "grad_norm": 1.2316622923723861, + "learning_rate": 6.7849658870824255e-06, + "loss": 0.2645, + "step": 18276 + }, + { + "epoch": 0.40231124807395996, + "grad_norm": 0.5997856095961489, + "learning_rate": 6.784632914150669e-06, + "loss": 0.5062, + "step": 18277 + }, + { + "epoch": 0.4023332599603786, + "grad_norm": 1.386397162292039, + "learning_rate": 6.7842999321486125e-06, + "loss": 0.3757, + "step": 18278 + }, + { + "epoch": 0.40235527184679726, + "grad_norm": 0.6095206987441837, + "learning_rate": 6.783966941077948e-06, + "loss": 0.5014, + "step": 18279 + }, + { + "epoch": 0.40237728373321596, + "grad_norm": 1.3731809950079548, + "learning_rate": 6.783633940940369e-06, + "loss": 0.3114, + "step": 18280 + }, + { + "epoch": 0.4023992956196346, + "grad_norm": 1.724886078750301, + "learning_rate": 6.783300931737564e-06, + "loss": 0.3719, + "step": 18281 + }, + { + "epoch": 0.40242130750605326, + "grad_norm": 1.1589829371716989, + "learning_rate": 6.782967913471231e-06, + "loss": 0.3279, + "step": 18282 + }, + { + "epoch": 0.4024433193924719, + "grad_norm": 1.2687747331014727, + "learning_rate": 6.782634886143058e-06, + "loss": 0.2949, + "step": 18283 + }, + { + "epoch": 0.4024653312788906, + "grad_norm": 0.6082077135439015, + "learning_rate": 6.782301849754741e-06, + "loss": 0.4807, + "step": 18284 + }, + { + "epoch": 0.40248734316530926, + "grad_norm": 1.4824739716039004, + "learning_rate": 6.78196880430797e-06, + "loss": 0.3261, + "step": 18285 + }, + { + "epoch": 0.4025093550517279, + "grad_norm": 0.6170837598416196, + "learning_rate": 6.781635749804439e-06, + "loss": 0.5059, + "step": 18286 + }, + { + "epoch": 0.4025313669381466, + "grad_norm": 1.8515683110887018, + "learning_rate": 6.781302686245841e-06, + "loss": 0.5134, + "step": 18287 + }, + { + "epoch": 0.40255337882456527, + "grad_norm": 1.1374792561764124, + "learning_rate": 6.7809696136338675e-06, + "loss": 0.3062, + "step": 18288 + }, + { + "epoch": 0.4025753907109839, + "grad_norm": 1.3739855344552132, + "learning_rate": 6.780636531970213e-06, + "loss": 0.3711, + "step": 18289 + }, + { + "epoch": 0.4025974025974026, + "grad_norm": 1.4338744861745454, + "learning_rate": 6.780303441256568e-06, + "loss": 0.3345, + "step": 18290 + }, + { + "epoch": 0.40261941448382127, + "grad_norm": 0.5980140594714559, + "learning_rate": 6.779970341494627e-06, + "loss": 0.4843, + "step": 18291 + }, + { + "epoch": 0.4026414263702399, + "grad_norm": 1.4643136680152804, + "learning_rate": 6.779637232686084e-06, + "loss": 0.2925, + "step": 18292 + }, + { + "epoch": 0.40266343825665857, + "grad_norm": 1.2578753771225148, + "learning_rate": 6.77930411483263e-06, + "loss": 0.28, + "step": 18293 + }, + { + "epoch": 0.4026854501430773, + "grad_norm": 1.3992746540852041, + "learning_rate": 6.77897098793596e-06, + "loss": 0.331, + "step": 18294 + }, + { + "epoch": 0.4027074620294959, + "grad_norm": 1.7253121041925026, + "learning_rate": 6.778637851997765e-06, + "loss": 0.3385, + "step": 18295 + }, + { + "epoch": 0.40272947391591457, + "grad_norm": 1.4820368526660885, + "learning_rate": 6.778304707019739e-06, + "loss": 0.4013, + "step": 18296 + }, + { + "epoch": 0.4027514858023333, + "grad_norm": 1.646927012333159, + "learning_rate": 6.777971553003576e-06, + "loss": 0.3817, + "step": 18297 + }, + { + "epoch": 0.4027734976887519, + "grad_norm": 1.3488158972265705, + "learning_rate": 6.777638389950968e-06, + "loss": 0.3382, + "step": 18298 + }, + { + "epoch": 0.4027955095751706, + "grad_norm": 1.7700466716645924, + "learning_rate": 6.777305217863609e-06, + "loss": 0.2645, + "step": 18299 + }, + { + "epoch": 0.4028175214615893, + "grad_norm": 0.6162747198735498, + "learning_rate": 6.776972036743192e-06, + "loss": 0.5199, + "step": 18300 + }, + { + "epoch": 0.4028395333480079, + "grad_norm": 1.4026471841434685, + "learning_rate": 6.77663884659141e-06, + "loss": 0.3035, + "step": 18301 + }, + { + "epoch": 0.4028615452344266, + "grad_norm": 1.0522643516882633, + "learning_rate": 6.776305647409959e-06, + "loss": 0.3175, + "step": 18302 + }, + { + "epoch": 0.4028835571208453, + "grad_norm": 2.133983430535429, + "learning_rate": 6.775972439200529e-06, + "loss": 0.2806, + "step": 18303 + }, + { + "epoch": 0.40290556900726393, + "grad_norm": 1.28040378311849, + "learning_rate": 6.775639221964816e-06, + "loss": 0.2985, + "step": 18304 + }, + { + "epoch": 0.4029275808936826, + "grad_norm": 0.6299468184196726, + "learning_rate": 6.775305995704511e-06, + "loss": 0.4951, + "step": 18305 + }, + { + "epoch": 0.40294959278010123, + "grad_norm": 1.6718795745229762, + "learning_rate": 6.77497276042131e-06, + "loss": 0.3662, + "step": 18306 + }, + { + "epoch": 0.40297160466651993, + "grad_norm": 1.7256529225763635, + "learning_rate": 6.774639516116907e-06, + "loss": 0.3187, + "step": 18307 + }, + { + "epoch": 0.4029936165529386, + "grad_norm": 1.2106463895191304, + "learning_rate": 6.774306262792992e-06, + "loss": 0.2606, + "step": 18308 + }, + { + "epoch": 0.40301562843935723, + "grad_norm": 2.3561814003802977, + "learning_rate": 6.773973000451261e-06, + "loss": 0.2865, + "step": 18309 + }, + { + "epoch": 0.40303764032577594, + "grad_norm": 1.548825244474207, + "learning_rate": 6.773639729093411e-06, + "loss": 0.2771, + "step": 18310 + }, + { + "epoch": 0.4030596522121946, + "grad_norm": 0.6306131493671178, + "learning_rate": 6.7733064487211305e-06, + "loss": 0.5167, + "step": 18311 + }, + { + "epoch": 0.40308166409861323, + "grad_norm": 2.348553315660305, + "learning_rate": 6.772973159336117e-06, + "loss": 0.4304, + "step": 18312 + }, + { + "epoch": 0.40310367598503194, + "grad_norm": 1.5598049699742431, + "learning_rate": 6.772639860940062e-06, + "loss": 0.4084, + "step": 18313 + }, + { + "epoch": 0.4031256878714506, + "grad_norm": 1.302925056152374, + "learning_rate": 6.772306553534661e-06, + "loss": 0.2866, + "step": 18314 + }, + { + "epoch": 0.40314769975786924, + "grad_norm": 1.316882640776479, + "learning_rate": 6.771973237121609e-06, + "loss": 0.2948, + "step": 18315 + }, + { + "epoch": 0.40316971164428794, + "grad_norm": 1.908828141873711, + "learning_rate": 6.771639911702596e-06, + "loss": 0.3183, + "step": 18316 + }, + { + "epoch": 0.4031917235307066, + "grad_norm": 1.2353181490082765, + "learning_rate": 6.7713065772793204e-06, + "loss": 0.2944, + "step": 18317 + }, + { + "epoch": 0.40321373541712524, + "grad_norm": 1.1230031856013223, + "learning_rate": 6.770973233853475e-06, + "loss": 0.2318, + "step": 18318 + }, + { + "epoch": 0.4032357473035439, + "grad_norm": 1.4187087512259284, + "learning_rate": 6.770639881426752e-06, + "loss": 0.3494, + "step": 18319 + }, + { + "epoch": 0.4032577591899626, + "grad_norm": 1.6524098858621443, + "learning_rate": 6.770306520000851e-06, + "loss": 0.3776, + "step": 18320 + }, + { + "epoch": 0.40327977107638124, + "grad_norm": 1.8696607633952609, + "learning_rate": 6.769973149577459e-06, + "loss": 0.3413, + "step": 18321 + }, + { + "epoch": 0.4033017829627999, + "grad_norm": 1.3231613085054048, + "learning_rate": 6.769639770158276e-06, + "loss": 0.2569, + "step": 18322 + }, + { + "epoch": 0.4033237948492186, + "grad_norm": 1.6206108464818443, + "learning_rate": 6.769306381744993e-06, + "loss": 0.3463, + "step": 18323 + }, + { + "epoch": 0.40334580673563725, + "grad_norm": 0.6274107648495644, + "learning_rate": 6.768972984339308e-06, + "loss": 0.4749, + "step": 18324 + }, + { + "epoch": 0.4033678186220559, + "grad_norm": 2.1290827340683904, + "learning_rate": 6.768639577942912e-06, + "loss": 0.1913, + "step": 18325 + }, + { + "epoch": 0.4033898305084746, + "grad_norm": 2.0010589049991303, + "learning_rate": 6.768306162557501e-06, + "loss": 0.2484, + "step": 18326 + }, + { + "epoch": 0.40341184239489325, + "grad_norm": 4.884359126647389, + "learning_rate": 6.767972738184768e-06, + "loss": 0.3072, + "step": 18327 + }, + { + "epoch": 0.4034338542813119, + "grad_norm": 2.1605783248471737, + "learning_rate": 6.76763930482641e-06, + "loss": 0.2699, + "step": 18328 + }, + { + "epoch": 0.40345586616773055, + "grad_norm": 1.731749794097708, + "learning_rate": 6.7673058624841215e-06, + "loss": 0.3101, + "step": 18329 + }, + { + "epoch": 0.40347787805414925, + "grad_norm": 0.5914660096672159, + "learning_rate": 6.766972411159596e-06, + "loss": 0.48, + "step": 18330 + }, + { + "epoch": 0.4034998899405679, + "grad_norm": 0.605863357249557, + "learning_rate": 6.766638950854528e-06, + "loss": 0.4736, + "step": 18331 + }, + { + "epoch": 0.40352190182698655, + "grad_norm": 1.3101417184194792, + "learning_rate": 6.766305481570613e-06, + "loss": 0.3396, + "step": 18332 + }, + { + "epoch": 0.40354391371340526, + "grad_norm": 1.6229645960356442, + "learning_rate": 6.765972003309547e-06, + "loss": 0.3587, + "step": 18333 + }, + { + "epoch": 0.4035659255998239, + "grad_norm": 1.4606933319896218, + "learning_rate": 6.765638516073022e-06, + "loss": 0.3393, + "step": 18334 + }, + { + "epoch": 0.40358793748624255, + "grad_norm": 1.5858402723892844, + "learning_rate": 6.765305019862736e-06, + "loss": 0.3311, + "step": 18335 + }, + { + "epoch": 0.40360994937266126, + "grad_norm": 1.3103753593304879, + "learning_rate": 6.76497151468038e-06, + "loss": 0.3289, + "step": 18336 + }, + { + "epoch": 0.4036319612590799, + "grad_norm": 1.2950772496153042, + "learning_rate": 6.7646380005276534e-06, + "loss": 0.3371, + "step": 18337 + }, + { + "epoch": 0.40365397314549856, + "grad_norm": 1.8093604498789648, + "learning_rate": 6.764304477406251e-06, + "loss": 0.3136, + "step": 18338 + }, + { + "epoch": 0.40367598503191726, + "grad_norm": 1.3192901826814794, + "learning_rate": 6.763970945317865e-06, + "loss": 0.2851, + "step": 18339 + }, + { + "epoch": 0.4036979969183359, + "grad_norm": 3.1838496408252754, + "learning_rate": 6.763637404264191e-06, + "loss": 0.2945, + "step": 18340 + }, + { + "epoch": 0.40372000880475456, + "grad_norm": 0.6429921565164188, + "learning_rate": 6.763303854246925e-06, + "loss": 0.4776, + "step": 18341 + }, + { + "epoch": 0.4037420206911732, + "grad_norm": 0.6099229036453274, + "learning_rate": 6.762970295267762e-06, + "loss": 0.479, + "step": 18342 + }, + { + "epoch": 0.4037640325775919, + "grad_norm": 0.6080801888909235, + "learning_rate": 6.7626367273284e-06, + "loss": 0.4804, + "step": 18343 + }, + { + "epoch": 0.40378604446401056, + "grad_norm": 2.7770268263511175, + "learning_rate": 6.762303150430532e-06, + "loss": 0.3079, + "step": 18344 + }, + { + "epoch": 0.4038080563504292, + "grad_norm": 0.6110687337929367, + "learning_rate": 6.761969564575851e-06, + "loss": 0.4878, + "step": 18345 + }, + { + "epoch": 0.4038300682368479, + "grad_norm": 1.2417578517205679, + "learning_rate": 6.761635969766056e-06, + "loss": 0.3307, + "step": 18346 + }, + { + "epoch": 0.40385208012326657, + "grad_norm": 1.6180174011838266, + "learning_rate": 6.761302366002841e-06, + "loss": 0.3192, + "step": 18347 + }, + { + "epoch": 0.4038740920096852, + "grad_norm": 1.7404260227160504, + "learning_rate": 6.760968753287902e-06, + "loss": 0.372, + "step": 18348 + }, + { + "epoch": 0.4038961038961039, + "grad_norm": 0.6385446896643711, + "learning_rate": 6.760635131622935e-06, + "loss": 0.4897, + "step": 18349 + }, + { + "epoch": 0.40391811578252257, + "grad_norm": 2.051658482417941, + "learning_rate": 6.760301501009634e-06, + "loss": 0.3184, + "step": 18350 + }, + { + "epoch": 0.4039401276689412, + "grad_norm": 1.929066267140672, + "learning_rate": 6.7599678614496956e-06, + "loss": 0.2892, + "step": 18351 + }, + { + "epoch": 0.40396213955535987, + "grad_norm": 1.583066863809051, + "learning_rate": 6.759634212944817e-06, + "loss": 0.3567, + "step": 18352 + }, + { + "epoch": 0.40398415144177857, + "grad_norm": 1.1686304359265403, + "learning_rate": 6.759300555496691e-06, + "loss": 0.3053, + "step": 18353 + }, + { + "epoch": 0.4040061633281972, + "grad_norm": 0.633595253799297, + "learning_rate": 6.758966889107015e-06, + "loss": 0.5004, + "step": 18354 + }, + { + "epoch": 0.40402817521461587, + "grad_norm": 1.7083538374177842, + "learning_rate": 6.758633213777485e-06, + "loss": 0.3794, + "step": 18355 + }, + { + "epoch": 0.4040501871010346, + "grad_norm": 3.0256590735143423, + "learning_rate": 6.758299529509797e-06, + "loss": 0.2572, + "step": 18356 + }, + { + "epoch": 0.4040721989874532, + "grad_norm": 1.4740653131317838, + "learning_rate": 6.757965836305646e-06, + "loss": 0.2649, + "step": 18357 + }, + { + "epoch": 0.4040942108738719, + "grad_norm": 1.7926826345457436, + "learning_rate": 6.757632134166729e-06, + "loss": 0.3886, + "step": 18358 + }, + { + "epoch": 0.4041162227602906, + "grad_norm": 2.3528684452866258, + "learning_rate": 6.757298423094743e-06, + "loss": 0.4055, + "step": 18359 + }, + { + "epoch": 0.4041382346467092, + "grad_norm": 1.5354119221785765, + "learning_rate": 6.75696470309138e-06, + "loss": 0.3116, + "step": 18360 + }, + { + "epoch": 0.4041602465331279, + "grad_norm": 2.3868639656307256, + "learning_rate": 6.75663097415834e-06, + "loss": 0.3494, + "step": 18361 + }, + { + "epoch": 0.4041822584195466, + "grad_norm": 0.6302346078573892, + "learning_rate": 6.756297236297318e-06, + "loss": 0.5101, + "step": 18362 + }, + { + "epoch": 0.40420427030596523, + "grad_norm": 2.1695926742268448, + "learning_rate": 6.755963489510009e-06, + "loss": 0.3791, + "step": 18363 + }, + { + "epoch": 0.4042262821923839, + "grad_norm": 2.3518484537324422, + "learning_rate": 6.755629733798111e-06, + "loss": 0.2861, + "step": 18364 + }, + { + "epoch": 0.40424829407880253, + "grad_norm": 1.413230982018351, + "learning_rate": 6.75529596916332e-06, + "loss": 0.345, + "step": 18365 + }, + { + "epoch": 0.40427030596522123, + "grad_norm": 1.2302468104862574, + "learning_rate": 6.754962195607333e-06, + "loss": 0.3025, + "step": 18366 + }, + { + "epoch": 0.4042923178516399, + "grad_norm": 0.6214159298038971, + "learning_rate": 6.754628413131844e-06, + "loss": 0.4858, + "step": 18367 + }, + { + "epoch": 0.40431432973805853, + "grad_norm": 1.5905086398354666, + "learning_rate": 6.754294621738551e-06, + "loss": 0.3165, + "step": 18368 + }, + { + "epoch": 0.40433634162447724, + "grad_norm": 3.335007859692354, + "learning_rate": 6.753960821429151e-06, + "loss": 0.3987, + "step": 18369 + }, + { + "epoch": 0.4043583535108959, + "grad_norm": 1.338930027296818, + "learning_rate": 6.75362701220534e-06, + "loss": 0.3394, + "step": 18370 + }, + { + "epoch": 0.40438036539731453, + "grad_norm": 1.4509153463465885, + "learning_rate": 6.753293194068813e-06, + "loss": 0.3233, + "step": 18371 + }, + { + "epoch": 0.40440237728373324, + "grad_norm": 1.134389381819471, + "learning_rate": 6.752959367021268e-06, + "loss": 0.3518, + "step": 18372 + }, + { + "epoch": 0.4044243891701519, + "grad_norm": 1.2100235734181666, + "learning_rate": 6.7526255310644015e-06, + "loss": 0.3397, + "step": 18373 + }, + { + "epoch": 0.40444640105657054, + "grad_norm": 1.4729566302031907, + "learning_rate": 6.752291686199912e-06, + "loss": 0.3745, + "step": 18374 + }, + { + "epoch": 0.40446841294298924, + "grad_norm": 2.0199066128607592, + "learning_rate": 6.751957832429494e-06, + "loss": 0.3241, + "step": 18375 + }, + { + "epoch": 0.4044904248294079, + "grad_norm": 1.1712826108515386, + "learning_rate": 6.7516239697548435e-06, + "loss": 0.3439, + "step": 18376 + }, + { + "epoch": 0.40451243671582654, + "grad_norm": 1.3513620397579111, + "learning_rate": 6.7512900981776604e-06, + "loss": 0.3958, + "step": 18377 + }, + { + "epoch": 0.4045344486022452, + "grad_norm": 1.4121383554628344, + "learning_rate": 6.750956217699639e-06, + "loss": 0.3232, + "step": 18378 + }, + { + "epoch": 0.4045564604886639, + "grad_norm": 1.331233630628075, + "learning_rate": 6.750622328322478e-06, + "loss": 0.3799, + "step": 18379 + }, + { + "epoch": 0.40457847237508254, + "grad_norm": 1.273706662652324, + "learning_rate": 6.7502884300478745e-06, + "loss": 0.3327, + "step": 18380 + }, + { + "epoch": 0.4046004842615012, + "grad_norm": 1.4682454415781454, + "learning_rate": 6.7499545228775225e-06, + "loss": 0.3465, + "step": 18381 + }, + { + "epoch": 0.4046224961479199, + "grad_norm": 0.6318600969370107, + "learning_rate": 6.749620606813121e-06, + "loss": 0.5194, + "step": 18382 + }, + { + "epoch": 0.40464450803433855, + "grad_norm": 1.6009690600070403, + "learning_rate": 6.749286681856368e-06, + "loss": 0.3453, + "step": 18383 + }, + { + "epoch": 0.4046665199207572, + "grad_norm": 1.6661573410413562, + "learning_rate": 6.74895274800896e-06, + "loss": 0.3282, + "step": 18384 + }, + { + "epoch": 0.4046885318071759, + "grad_norm": 0.6387660932996413, + "learning_rate": 6.748618805272594e-06, + "loss": 0.4988, + "step": 18385 + }, + { + "epoch": 0.40471054369359455, + "grad_norm": 1.6568813506589315, + "learning_rate": 6.748284853648966e-06, + "loss": 0.3896, + "step": 18386 + }, + { + "epoch": 0.4047325555800132, + "grad_norm": 1.0958663752357252, + "learning_rate": 6.747950893139777e-06, + "loss": 0.3446, + "step": 18387 + }, + { + "epoch": 0.40475456746643185, + "grad_norm": 1.099206584095303, + "learning_rate": 6.747616923746721e-06, + "loss": 0.2827, + "step": 18388 + }, + { + "epoch": 0.40477657935285055, + "grad_norm": 14.77188975214311, + "learning_rate": 6.747282945471496e-06, + "loss": 0.2171, + "step": 18389 + }, + { + "epoch": 0.4047985912392692, + "grad_norm": 1.4911347318344088, + "learning_rate": 6.7469489583158e-06, + "loss": 0.3406, + "step": 18390 + }, + { + "epoch": 0.40482060312568785, + "grad_norm": 1.3647342228559405, + "learning_rate": 6.746614962281329e-06, + "loss": 0.2908, + "step": 18391 + }, + { + "epoch": 0.40484261501210655, + "grad_norm": 1.6954535267266662, + "learning_rate": 6.7462809573697845e-06, + "loss": 0.3355, + "step": 18392 + }, + { + "epoch": 0.4048646268985252, + "grad_norm": 1.8104938013806824, + "learning_rate": 6.745946943582859e-06, + "loss": 0.2806, + "step": 18393 + }, + { + "epoch": 0.40488663878494385, + "grad_norm": 0.6242819370450118, + "learning_rate": 6.745612920922253e-06, + "loss": 0.4967, + "step": 18394 + }, + { + "epoch": 0.40490865067136256, + "grad_norm": 1.4098146790864121, + "learning_rate": 6.7452788893896635e-06, + "loss": 0.3124, + "step": 18395 + }, + { + "epoch": 0.4049306625577812, + "grad_norm": 1.8525614575897877, + "learning_rate": 6.744944848986789e-06, + "loss": 0.3904, + "step": 18396 + }, + { + "epoch": 0.40495267444419986, + "grad_norm": 1.4876472074332336, + "learning_rate": 6.744610799715327e-06, + "loss": 0.3786, + "step": 18397 + }, + { + "epoch": 0.40497468633061856, + "grad_norm": 1.6844865870784265, + "learning_rate": 6.744276741576974e-06, + "loss": 0.3973, + "step": 18398 + }, + { + "epoch": 0.4049966982170372, + "grad_norm": 1.512301937396879, + "learning_rate": 6.743942674573427e-06, + "loss": 0.3619, + "step": 18399 + }, + { + "epoch": 0.40501871010345586, + "grad_norm": 1.4110309082241892, + "learning_rate": 6.743608598706387e-06, + "loss": 0.3052, + "step": 18400 + }, + { + "epoch": 0.4050407219898745, + "grad_norm": 1.684554007566257, + "learning_rate": 6.743274513977551e-06, + "loss": 0.3502, + "step": 18401 + }, + { + "epoch": 0.4050627338762932, + "grad_norm": 0.617170345462909, + "learning_rate": 6.742940420388615e-06, + "loss": 0.4815, + "step": 18402 + }, + { + "epoch": 0.40508474576271186, + "grad_norm": 1.2758253304584315, + "learning_rate": 6.74260631794128e-06, + "loss": 0.2884, + "step": 18403 + }, + { + "epoch": 0.4051067576491305, + "grad_norm": 0.6182629637356701, + "learning_rate": 6.742272206637241e-06, + "loss": 0.5055, + "step": 18404 + }, + { + "epoch": 0.4051287695355492, + "grad_norm": 1.4720312040495196, + "learning_rate": 6.741938086478198e-06, + "loss": 0.2729, + "step": 18405 + }, + { + "epoch": 0.40515078142196786, + "grad_norm": 1.2984354809516705, + "learning_rate": 6.7416039574658485e-06, + "loss": 0.3663, + "step": 18406 + }, + { + "epoch": 0.4051727933083865, + "grad_norm": 1.5191598989649961, + "learning_rate": 6.741269819601892e-06, + "loss": 0.2735, + "step": 18407 + }, + { + "epoch": 0.4051948051948052, + "grad_norm": 1.3960434227296568, + "learning_rate": 6.740935672888025e-06, + "loss": 0.3427, + "step": 18408 + }, + { + "epoch": 0.40521681708122387, + "grad_norm": 1.115010537465641, + "learning_rate": 6.7406015173259465e-06, + "loss": 0.2141, + "step": 18409 + }, + { + "epoch": 0.4052388289676425, + "grad_norm": 1.2435952576008287, + "learning_rate": 6.7402673529173534e-06, + "loss": 0.3039, + "step": 18410 + }, + { + "epoch": 0.40526084085406117, + "grad_norm": 1.3651693278254866, + "learning_rate": 6.739933179663949e-06, + "loss": 0.38, + "step": 18411 + }, + { + "epoch": 0.40528285274047987, + "grad_norm": 1.5467689732151573, + "learning_rate": 6.739598997567425e-06, + "loss": 0.3594, + "step": 18412 + }, + { + "epoch": 0.4053048646268985, + "grad_norm": 1.5595358484856274, + "learning_rate": 6.739264806629484e-06, + "loss": 0.4531, + "step": 18413 + }, + { + "epoch": 0.40532687651331717, + "grad_norm": 1.526137249667607, + "learning_rate": 6.738930606851823e-06, + "loss": 0.2631, + "step": 18414 + }, + { + "epoch": 0.4053488883997359, + "grad_norm": 1.740181447739791, + "learning_rate": 6.738596398236141e-06, + "loss": 0.3305, + "step": 18415 + }, + { + "epoch": 0.4053709002861545, + "grad_norm": 2.0976026705965385, + "learning_rate": 6.738262180784139e-06, + "loss": 0.3974, + "step": 18416 + }, + { + "epoch": 0.40539291217257317, + "grad_norm": 1.3700017039498773, + "learning_rate": 6.73792795449751e-06, + "loss": 0.3563, + "step": 18417 + }, + { + "epoch": 0.4054149240589919, + "grad_norm": 0.6532860573169581, + "learning_rate": 6.737593719377957e-06, + "loss": 0.4923, + "step": 18418 + }, + { + "epoch": 0.4054369359454105, + "grad_norm": 1.1424137917086876, + "learning_rate": 6.737259475427178e-06, + "loss": 0.3219, + "step": 18419 + }, + { + "epoch": 0.4054589478318292, + "grad_norm": 1.7828303224364415, + "learning_rate": 6.736925222646873e-06, + "loss": 0.3277, + "step": 18420 + }, + { + "epoch": 0.4054809597182479, + "grad_norm": 0.6383548106946256, + "learning_rate": 6.736590961038738e-06, + "loss": 0.5182, + "step": 18421 + }, + { + "epoch": 0.40550297160466653, + "grad_norm": 0.5911435093753017, + "learning_rate": 6.736256690604473e-06, + "loss": 0.4691, + "step": 18422 + }, + { + "epoch": 0.4055249834910852, + "grad_norm": 1.2855357984471985, + "learning_rate": 6.735922411345777e-06, + "loss": 0.3572, + "step": 18423 + }, + { + "epoch": 0.4055469953775038, + "grad_norm": 1.687218868177462, + "learning_rate": 6.735588123264349e-06, + "loss": 0.3308, + "step": 18424 + }, + { + "epoch": 0.40556900726392253, + "grad_norm": 3.9914150600895906, + "learning_rate": 6.735253826361889e-06, + "loss": 0.3062, + "step": 18425 + }, + { + "epoch": 0.4055910191503412, + "grad_norm": 1.2214304432028544, + "learning_rate": 6.7349195206400945e-06, + "loss": 0.3365, + "step": 18426 + }, + { + "epoch": 0.40561303103675983, + "grad_norm": 1.3489045062717895, + "learning_rate": 6.7345852061006646e-06, + "loss": 0.2731, + "step": 18427 + }, + { + "epoch": 0.40563504292317853, + "grad_norm": 1.036613833845642, + "learning_rate": 6.734250882745299e-06, + "loss": 0.263, + "step": 18428 + }, + { + "epoch": 0.4056570548095972, + "grad_norm": 1.3739346752055852, + "learning_rate": 6.733916550575699e-06, + "loss": 0.3789, + "step": 18429 + }, + { + "epoch": 0.40567906669601583, + "grad_norm": 2.263838891830678, + "learning_rate": 6.7335822095935595e-06, + "loss": 0.2489, + "step": 18430 + }, + { + "epoch": 0.40570107858243454, + "grad_norm": 0.6315876240906823, + "learning_rate": 6.733247859800583e-06, + "loss": 0.479, + "step": 18431 + }, + { + "epoch": 0.4057230904688532, + "grad_norm": 0.6130584032431117, + "learning_rate": 6.732913501198468e-06, + "loss": 0.4863, + "step": 18432 + }, + { + "epoch": 0.40574510235527184, + "grad_norm": 5.104215564963546, + "learning_rate": 6.732579133788913e-06, + "loss": 0.3189, + "step": 18433 + }, + { + "epoch": 0.4057671142416905, + "grad_norm": 0.6225156156264687, + "learning_rate": 6.732244757573619e-06, + "loss": 0.491, + "step": 18434 + }, + { + "epoch": 0.4057891261281092, + "grad_norm": 3.4144896532445324, + "learning_rate": 6.731910372554284e-06, + "loss": 0.3342, + "step": 18435 + }, + { + "epoch": 0.40581113801452784, + "grad_norm": 0.6142368943208734, + "learning_rate": 6.731575978732607e-06, + "loss": 0.509, + "step": 18436 + }, + { + "epoch": 0.4058331499009465, + "grad_norm": 0.6320503054656142, + "learning_rate": 6.73124157611029e-06, + "loss": 0.5135, + "step": 18437 + }, + { + "epoch": 0.4058551617873652, + "grad_norm": 1.2970746380158273, + "learning_rate": 6.730907164689031e-06, + "loss": 0.34, + "step": 18438 + }, + { + "epoch": 0.40587717367378384, + "grad_norm": 1.7023659960066384, + "learning_rate": 6.730572744470529e-06, + "loss": 0.2929, + "step": 18439 + }, + { + "epoch": 0.4058991855602025, + "grad_norm": 1.2463455282477756, + "learning_rate": 6.730238315456485e-06, + "loss": 0.2395, + "step": 18440 + }, + { + "epoch": 0.4059211974466212, + "grad_norm": 1.5776332341446035, + "learning_rate": 6.729903877648598e-06, + "loss": 0.2457, + "step": 18441 + }, + { + "epoch": 0.40594320933303984, + "grad_norm": 1.4713725515033396, + "learning_rate": 6.7295694310485685e-06, + "loss": 0.2689, + "step": 18442 + }, + { + "epoch": 0.4059652212194585, + "grad_norm": 1.408357479190559, + "learning_rate": 6.729234975658096e-06, + "loss": 0.2587, + "step": 18443 + }, + { + "epoch": 0.4059872331058772, + "grad_norm": 1.30780063545872, + "learning_rate": 6.7289005114788785e-06, + "loss": 0.3254, + "step": 18444 + }, + { + "epoch": 0.40600924499229585, + "grad_norm": 1.5695177102875362, + "learning_rate": 6.7285660385126185e-06, + "loss": 0.3866, + "step": 18445 + }, + { + "epoch": 0.4060312568787145, + "grad_norm": 1.5011400625006133, + "learning_rate": 6.728231556761015e-06, + "loss": 0.2607, + "step": 18446 + }, + { + "epoch": 0.40605326876513315, + "grad_norm": 0.6831611358849609, + "learning_rate": 6.727897066225768e-06, + "loss": 0.5126, + "step": 18447 + }, + { + "epoch": 0.40607528065155185, + "grad_norm": 1.3587473662279836, + "learning_rate": 6.727562566908576e-06, + "loss": 0.3036, + "step": 18448 + }, + { + "epoch": 0.4060972925379705, + "grad_norm": 1.9458392981521975, + "learning_rate": 6.727228058811143e-06, + "loss": 0.3817, + "step": 18449 + }, + { + "epoch": 0.40611930442438915, + "grad_norm": 0.6338243175559104, + "learning_rate": 6.726893541935165e-06, + "loss": 0.5006, + "step": 18450 + }, + { + "epoch": 0.40614131631080785, + "grad_norm": 0.6189773447511087, + "learning_rate": 6.726559016282344e-06, + "loss": 0.5047, + "step": 18451 + }, + { + "epoch": 0.4061633281972265, + "grad_norm": 0.628900339231597, + "learning_rate": 6.726224481854381e-06, + "loss": 0.4898, + "step": 18452 + }, + { + "epoch": 0.40618534008364515, + "grad_norm": 2.0649351564413445, + "learning_rate": 6.725889938652973e-06, + "loss": 0.3673, + "step": 18453 + }, + { + "epoch": 0.40620735197006386, + "grad_norm": 1.4185812795163157, + "learning_rate": 6.7255553866798254e-06, + "loss": 0.2839, + "step": 18454 + }, + { + "epoch": 0.4062293638564825, + "grad_norm": 1.8802033671372163, + "learning_rate": 6.725220825936634e-06, + "loss": 0.3237, + "step": 18455 + }, + { + "epoch": 0.40625137574290116, + "grad_norm": 1.9012432885787294, + "learning_rate": 6.7248862564251e-06, + "loss": 0.3212, + "step": 18456 + }, + { + "epoch": 0.40627338762931986, + "grad_norm": 1.5447295749419834, + "learning_rate": 6.7245516781469266e-06, + "loss": 0.3834, + "step": 18457 + }, + { + "epoch": 0.4062953995157385, + "grad_norm": 0.6055411504071091, + "learning_rate": 6.724217091103811e-06, + "loss": 0.4688, + "step": 18458 + }, + { + "epoch": 0.40631741140215716, + "grad_norm": 2.012347082560418, + "learning_rate": 6.723882495297456e-06, + "loss": 0.3262, + "step": 18459 + }, + { + "epoch": 0.4063394232885758, + "grad_norm": 1.8360777391691523, + "learning_rate": 6.723547890729561e-06, + "loss": 0.3377, + "step": 18460 + }, + { + "epoch": 0.4063614351749945, + "grad_norm": 10.58250303646567, + "learning_rate": 6.723213277401827e-06, + "loss": 0.3333, + "step": 18461 + }, + { + "epoch": 0.40638344706141316, + "grad_norm": 1.6418286910862883, + "learning_rate": 6.722878655315955e-06, + "loss": 0.2719, + "step": 18462 + }, + { + "epoch": 0.4064054589478318, + "grad_norm": 1.9597005799272138, + "learning_rate": 6.722544024473645e-06, + "loss": 0.3095, + "step": 18463 + }, + { + "epoch": 0.4064274708342505, + "grad_norm": 1.3637991757251262, + "learning_rate": 6.722209384876598e-06, + "loss": 0.3369, + "step": 18464 + }, + { + "epoch": 0.40644948272066916, + "grad_norm": 1.2025078835817873, + "learning_rate": 6.7218747365265155e-06, + "loss": 0.3299, + "step": 18465 + }, + { + "epoch": 0.4064714946070878, + "grad_norm": 1.6844511138344995, + "learning_rate": 6.7215400794250955e-06, + "loss": 0.3995, + "step": 18466 + }, + { + "epoch": 0.4064935064935065, + "grad_norm": 2.877473782100401, + "learning_rate": 6.721205413574042e-06, + "loss": 0.3463, + "step": 18467 + }, + { + "epoch": 0.40651551837992517, + "grad_norm": 2.0668100957662086, + "learning_rate": 6.720870738975056e-06, + "loss": 0.3155, + "step": 18468 + }, + { + "epoch": 0.4065375302663438, + "grad_norm": 1.3190919932816632, + "learning_rate": 6.720536055629836e-06, + "loss": 0.259, + "step": 18469 + }, + { + "epoch": 0.40655954215276247, + "grad_norm": 1.366646520257851, + "learning_rate": 6.720201363540086e-06, + "loss": 0.3654, + "step": 18470 + }, + { + "epoch": 0.40658155403918117, + "grad_norm": 1.175613003270705, + "learning_rate": 6.719866662707502e-06, + "loss": 0.2778, + "step": 18471 + }, + { + "epoch": 0.4066035659255998, + "grad_norm": 1.7547008249248748, + "learning_rate": 6.719531953133792e-06, + "loss": 0.3738, + "step": 18472 + }, + { + "epoch": 0.40662557781201847, + "grad_norm": 1.3212968870917818, + "learning_rate": 6.719197234820653e-06, + "loss": 0.344, + "step": 18473 + }, + { + "epoch": 0.4066475896984372, + "grad_norm": 1.4008944080955128, + "learning_rate": 6.718862507769787e-06, + "loss": 0.3569, + "step": 18474 + }, + { + "epoch": 0.4066696015848558, + "grad_norm": 0.6697469950738818, + "learning_rate": 6.718527771982895e-06, + "loss": 0.4959, + "step": 18475 + }, + { + "epoch": 0.40669161347127447, + "grad_norm": 1.4598610467501152, + "learning_rate": 6.718193027461676e-06, + "loss": 0.4157, + "step": 18476 + }, + { + "epoch": 0.4067136253576932, + "grad_norm": 1.8536313751168523, + "learning_rate": 6.717858274207835e-06, + "loss": 0.3731, + "step": 18477 + }, + { + "epoch": 0.4067356372441118, + "grad_norm": 0.6454283137722547, + "learning_rate": 6.717523512223074e-06, + "loss": 0.5163, + "step": 18478 + }, + { + "epoch": 0.4067576491305305, + "grad_norm": 1.3174465819187877, + "learning_rate": 6.717188741509091e-06, + "loss": 0.3323, + "step": 18479 + }, + { + "epoch": 0.4067796610169492, + "grad_norm": 1.5872952544048526, + "learning_rate": 6.716853962067589e-06, + "loss": 0.3508, + "step": 18480 + }, + { + "epoch": 0.40680167290336783, + "grad_norm": 0.6158808886849104, + "learning_rate": 6.716519173900269e-06, + "loss": 0.4901, + "step": 18481 + }, + { + "epoch": 0.4068236847897865, + "grad_norm": 1.4806568406854692, + "learning_rate": 6.7161843770088334e-06, + "loss": 0.3568, + "step": 18482 + }, + { + "epoch": 0.4068456966762051, + "grad_norm": 1.7166641058691878, + "learning_rate": 6.715849571394985e-06, + "loss": 0.3206, + "step": 18483 + }, + { + "epoch": 0.40686770856262383, + "grad_norm": 1.6070287134533385, + "learning_rate": 6.715514757060421e-06, + "loss": 0.4166, + "step": 18484 + }, + { + "epoch": 0.4068897204490425, + "grad_norm": 1.5940964992637845, + "learning_rate": 6.7151799340068465e-06, + "loss": 0.3506, + "step": 18485 + }, + { + "epoch": 0.40691173233546113, + "grad_norm": 2.0836674466213116, + "learning_rate": 6.714845102235964e-06, + "loss": 0.4195, + "step": 18486 + }, + { + "epoch": 0.40693374422187983, + "grad_norm": 1.3970230889003323, + "learning_rate": 6.714510261749471e-06, + "loss": 0.3252, + "step": 18487 + }, + { + "epoch": 0.4069557561082985, + "grad_norm": 2.7636553389693783, + "learning_rate": 6.714175412549077e-06, + "loss": 0.3323, + "step": 18488 + }, + { + "epoch": 0.40697776799471713, + "grad_norm": 1.780211610939703, + "learning_rate": 6.7138405546364745e-06, + "loss": 0.2962, + "step": 18489 + }, + { + "epoch": 0.40699977988113584, + "grad_norm": 1.6876497878555223, + "learning_rate": 6.713505688013372e-06, + "loss": 0.3974, + "step": 18490 + }, + { + "epoch": 0.4070217917675545, + "grad_norm": 1.41776309130784, + "learning_rate": 6.7131708126814684e-06, + "loss": 0.3064, + "step": 18491 + }, + { + "epoch": 0.40704380365397314, + "grad_norm": 1.4216420749492678, + "learning_rate": 6.712835928642467e-06, + "loss": 0.3284, + "step": 18492 + }, + { + "epoch": 0.4070658155403918, + "grad_norm": 1.188189928996896, + "learning_rate": 6.71250103589807e-06, + "loss": 0.3327, + "step": 18493 + }, + { + "epoch": 0.4070878274268105, + "grad_norm": 1.6368966078594496, + "learning_rate": 6.712166134449978e-06, + "loss": 0.3643, + "step": 18494 + }, + { + "epoch": 0.40710983931322914, + "grad_norm": 2.289812829894527, + "learning_rate": 6.711831224299895e-06, + "loss": 0.3369, + "step": 18495 + }, + { + "epoch": 0.4071318511996478, + "grad_norm": 1.4966623595406319, + "learning_rate": 6.711496305449522e-06, + "loss": 0.3322, + "step": 18496 + }, + { + "epoch": 0.4071538630860665, + "grad_norm": 1.2564586297746856, + "learning_rate": 6.711161377900561e-06, + "loss": 0.3107, + "step": 18497 + }, + { + "epoch": 0.40717587497248514, + "grad_norm": 1.201317183288262, + "learning_rate": 6.710826441654716e-06, + "loss": 0.3012, + "step": 18498 + }, + { + "epoch": 0.4071978868589038, + "grad_norm": 1.7252115159266908, + "learning_rate": 6.7104914967136846e-06, + "loss": 0.3216, + "step": 18499 + }, + { + "epoch": 0.4072198987453225, + "grad_norm": 1.190333453254983, + "learning_rate": 6.710156543079176e-06, + "loss": 0.2945, + "step": 18500 + }, + { + "epoch": 0.40724191063174114, + "grad_norm": 0.7822264923009395, + "learning_rate": 6.709821580752887e-06, + "loss": 0.5107, + "step": 18501 + }, + { + "epoch": 0.4072639225181598, + "grad_norm": 1.9383070713847348, + "learning_rate": 6.709486609736524e-06, + "loss": 0.3223, + "step": 18502 + }, + { + "epoch": 0.4072859344045785, + "grad_norm": 1.894625931464534, + "learning_rate": 6.709151630031787e-06, + "loss": 0.2379, + "step": 18503 + }, + { + "epoch": 0.40730794629099715, + "grad_norm": 1.8875707723010486, + "learning_rate": 6.708816641640379e-06, + "loss": 0.349, + "step": 18504 + }, + { + "epoch": 0.4073299581774158, + "grad_norm": 0.643203896801335, + "learning_rate": 6.708481644564001e-06, + "loss": 0.4835, + "step": 18505 + }, + { + "epoch": 0.40735197006383445, + "grad_norm": 1.8845029766479107, + "learning_rate": 6.7081466388043605e-06, + "loss": 0.3554, + "step": 18506 + }, + { + "epoch": 0.40737398195025315, + "grad_norm": 1.3235887152679888, + "learning_rate": 6.707811624363154e-06, + "loss": 0.2931, + "step": 18507 + }, + { + "epoch": 0.4073959938366718, + "grad_norm": 1.290850464392361, + "learning_rate": 6.707476601242089e-06, + "loss": 0.3239, + "step": 18508 + }, + { + "epoch": 0.40741800572309045, + "grad_norm": 1.3413637751692589, + "learning_rate": 6.707141569442865e-06, + "loss": 0.3042, + "step": 18509 + }, + { + "epoch": 0.40744001760950915, + "grad_norm": 1.6444024410501488, + "learning_rate": 6.706806528967186e-06, + "loss": 0.3259, + "step": 18510 + }, + { + "epoch": 0.4074620294959278, + "grad_norm": 1.369568622283027, + "learning_rate": 6.706471479816756e-06, + "loss": 0.3102, + "step": 18511 + }, + { + "epoch": 0.40748404138234645, + "grad_norm": 1.483647846021489, + "learning_rate": 6.7061364219932765e-06, + "loss": 0.3044, + "step": 18512 + }, + { + "epoch": 0.40750605326876516, + "grad_norm": 1.372510954635099, + "learning_rate": 6.705801355498451e-06, + "loss": 0.3176, + "step": 18513 + }, + { + "epoch": 0.4075280651551838, + "grad_norm": 1.6105152925478468, + "learning_rate": 6.705466280333982e-06, + "loss": 0.359, + "step": 18514 + }, + { + "epoch": 0.40755007704160245, + "grad_norm": 2.0613792438821847, + "learning_rate": 6.705131196501573e-06, + "loss": 0.3412, + "step": 18515 + }, + { + "epoch": 0.40757208892802116, + "grad_norm": 2.003217843382949, + "learning_rate": 6.7047961040029265e-06, + "loss": 0.2425, + "step": 18516 + }, + { + "epoch": 0.4075941008144398, + "grad_norm": 1.5500889905907167, + "learning_rate": 6.704461002839745e-06, + "loss": 0.3421, + "step": 18517 + }, + { + "epoch": 0.40761611270085846, + "grad_norm": 1.53861553217104, + "learning_rate": 6.7041258930137335e-06, + "loss": 0.3451, + "step": 18518 + }, + { + "epoch": 0.4076381245872771, + "grad_norm": 1.8788231249718743, + "learning_rate": 6.703790774526594e-06, + "loss": 0.2591, + "step": 18519 + }, + { + "epoch": 0.4076601364736958, + "grad_norm": 1.72723730176226, + "learning_rate": 6.70345564738003e-06, + "loss": 0.3467, + "step": 18520 + }, + { + "epoch": 0.40768214836011446, + "grad_norm": 1.7242177217852694, + "learning_rate": 6.703120511575745e-06, + "loss": 0.2884, + "step": 18521 + }, + { + "epoch": 0.4077041602465331, + "grad_norm": 1.7708182415074647, + "learning_rate": 6.702785367115441e-06, + "loss": 0.2775, + "step": 18522 + }, + { + "epoch": 0.4077261721329518, + "grad_norm": 1.7313800852339374, + "learning_rate": 6.7024502140008214e-06, + "loss": 0.3023, + "step": 18523 + }, + { + "epoch": 0.40774818401937046, + "grad_norm": 0.6942926787692921, + "learning_rate": 6.702115052233593e-06, + "loss": 0.4984, + "step": 18524 + }, + { + "epoch": 0.4077701959057891, + "grad_norm": 1.6828669438541313, + "learning_rate": 6.701779881815455e-06, + "loss": 0.3008, + "step": 18525 + }, + { + "epoch": 0.4077922077922078, + "grad_norm": 2.7069127210885697, + "learning_rate": 6.7014447027481135e-06, + "loss": 0.2921, + "step": 18526 + }, + { + "epoch": 0.40781421967862647, + "grad_norm": 0.597906631967788, + "learning_rate": 6.701109515033271e-06, + "loss": 0.4716, + "step": 18527 + }, + { + "epoch": 0.4078362315650451, + "grad_norm": 3.1571608376982594, + "learning_rate": 6.700774318672629e-06, + "loss": 0.3099, + "step": 18528 + }, + { + "epoch": 0.40785824345146376, + "grad_norm": 0.6100590071168106, + "learning_rate": 6.700439113667897e-06, + "loss": 0.5032, + "step": 18529 + }, + { + "epoch": 0.40788025533788247, + "grad_norm": 1.5025403842248672, + "learning_rate": 6.700103900020772e-06, + "loss": 0.2462, + "step": 18530 + }, + { + "epoch": 0.4079022672243011, + "grad_norm": 1.4360645804771406, + "learning_rate": 6.699768677732963e-06, + "loss": 0.312, + "step": 18531 + }, + { + "epoch": 0.40792427911071977, + "grad_norm": 1.7709129011669673, + "learning_rate": 6.6994334468061706e-06, + "loss": 0.3256, + "step": 18532 + }, + { + "epoch": 0.40794629099713847, + "grad_norm": 0.6461722407814999, + "learning_rate": 6.6990982072421e-06, + "loss": 0.5152, + "step": 18533 + }, + { + "epoch": 0.4079683028835571, + "grad_norm": 1.3836085723073916, + "learning_rate": 6.6987629590424526e-06, + "loss": 0.3139, + "step": 18534 + }, + { + "epoch": 0.40799031476997577, + "grad_norm": 1.2267514572439333, + "learning_rate": 6.698427702208936e-06, + "loss": 0.3409, + "step": 18535 + }, + { + "epoch": 0.4080123266563945, + "grad_norm": 1.310109416284817, + "learning_rate": 6.698092436743251e-06, + "loss": 0.3465, + "step": 18536 + }, + { + "epoch": 0.4080343385428131, + "grad_norm": 1.857457494611012, + "learning_rate": 6.697757162647104e-06, + "loss": 0.3545, + "step": 18537 + }, + { + "epoch": 0.4080563504292318, + "grad_norm": 0.6581011114531772, + "learning_rate": 6.697421879922198e-06, + "loss": 0.496, + "step": 18538 + }, + { + "epoch": 0.4080783623156505, + "grad_norm": 1.496008866199345, + "learning_rate": 6.697086588570236e-06, + "loss": 0.2728, + "step": 18539 + }, + { + "epoch": 0.4081003742020691, + "grad_norm": 1.3460300941466439, + "learning_rate": 6.696751288592924e-06, + "loss": 0.3835, + "step": 18540 + }, + { + "epoch": 0.4081223860884878, + "grad_norm": 1.3590121979364607, + "learning_rate": 6.696415979991964e-06, + "loss": 0.3208, + "step": 18541 + }, + { + "epoch": 0.4081443979749064, + "grad_norm": 1.8326902300462213, + "learning_rate": 6.696080662769062e-06, + "loss": 0.3731, + "step": 18542 + }, + { + "epoch": 0.40816640986132513, + "grad_norm": 5.011262804197252, + "learning_rate": 6.695745336925922e-06, + "loss": 0.4382, + "step": 18543 + }, + { + "epoch": 0.4081884217477438, + "grad_norm": 2.824561683811269, + "learning_rate": 6.695410002464248e-06, + "loss": 0.2879, + "step": 18544 + }, + { + "epoch": 0.40821043363416243, + "grad_norm": 1.8070384605120606, + "learning_rate": 6.695074659385743e-06, + "loss": 0.3444, + "step": 18545 + }, + { + "epoch": 0.40823244552058113, + "grad_norm": 1.3068507052666607, + "learning_rate": 6.694739307692113e-06, + "loss": 0.3186, + "step": 18546 + }, + { + "epoch": 0.4082544574069998, + "grad_norm": 1.8522551170330508, + "learning_rate": 6.694403947385063e-06, + "loss": 0.3342, + "step": 18547 + }, + { + "epoch": 0.40827646929341843, + "grad_norm": 1.9252284201240892, + "learning_rate": 6.694068578466295e-06, + "loss": 0.4161, + "step": 18548 + }, + { + "epoch": 0.40829848117983714, + "grad_norm": 1.7975519392499593, + "learning_rate": 6.693733200937516e-06, + "loss": 0.3243, + "step": 18549 + }, + { + "epoch": 0.4083204930662558, + "grad_norm": 1.1174454647101824, + "learning_rate": 6.693397814800428e-06, + "loss": 0.2948, + "step": 18550 + }, + { + "epoch": 0.40834250495267443, + "grad_norm": 1.80084958978768, + "learning_rate": 6.693062420056738e-06, + "loss": 0.3422, + "step": 18551 + }, + { + "epoch": 0.4083645168390931, + "grad_norm": 1.9213888655656755, + "learning_rate": 6.69272701670815e-06, + "loss": 0.2714, + "step": 18552 + }, + { + "epoch": 0.4083865287255118, + "grad_norm": 1.7387556464290308, + "learning_rate": 6.6923916047563675e-06, + "loss": 0.3303, + "step": 18553 + }, + { + "epoch": 0.40840854061193044, + "grad_norm": 1.348258225737876, + "learning_rate": 6.6920561842030975e-06, + "loss": 0.3212, + "step": 18554 + }, + { + "epoch": 0.4084305524983491, + "grad_norm": 1.4584128626319701, + "learning_rate": 6.691720755050042e-06, + "loss": 0.3039, + "step": 18555 + }, + { + "epoch": 0.4084525643847678, + "grad_norm": 2.270253643756174, + "learning_rate": 6.691385317298908e-06, + "loss": 0.3171, + "step": 18556 + }, + { + "epoch": 0.40847457627118644, + "grad_norm": 1.6326829413305377, + "learning_rate": 6.691049870951398e-06, + "loss": 0.3037, + "step": 18557 + }, + { + "epoch": 0.4084965881576051, + "grad_norm": 0.6724263907125362, + "learning_rate": 6.69071441600922e-06, + "loss": 0.5273, + "step": 18558 + }, + { + "epoch": 0.4085186000440238, + "grad_norm": 2.230919157574597, + "learning_rate": 6.690378952474074e-06, + "loss": 0.4072, + "step": 18559 + }, + { + "epoch": 0.40854061193044244, + "grad_norm": 0.6105266535334012, + "learning_rate": 6.690043480347672e-06, + "loss": 0.4956, + "step": 18560 + }, + { + "epoch": 0.4085626238168611, + "grad_norm": 0.5936750902217954, + "learning_rate": 6.689707999631714e-06, + "loss": 0.4725, + "step": 18561 + }, + { + "epoch": 0.4085846357032798, + "grad_norm": 1.7387367833670653, + "learning_rate": 6.689372510327906e-06, + "loss": 0.3211, + "step": 18562 + }, + { + "epoch": 0.40860664758969845, + "grad_norm": 1.3111439432415055, + "learning_rate": 6.689037012437955e-06, + "loss": 0.2128, + "step": 18563 + }, + { + "epoch": 0.4086286594761171, + "grad_norm": 1.2251070013857663, + "learning_rate": 6.688701505963562e-06, + "loss": 0.3185, + "step": 18564 + }, + { + "epoch": 0.40865067136253574, + "grad_norm": 0.6326265813455534, + "learning_rate": 6.688365990906437e-06, + "loss": 0.4909, + "step": 18565 + }, + { + "epoch": 0.40867268324895445, + "grad_norm": 0.6388519274484098, + "learning_rate": 6.688030467268281e-06, + "loss": 0.4781, + "step": 18566 + }, + { + "epoch": 0.4086946951353731, + "grad_norm": 1.6149889258835102, + "learning_rate": 6.687694935050803e-06, + "loss": 0.3659, + "step": 18567 + }, + { + "epoch": 0.40871670702179175, + "grad_norm": 2.2537589530668924, + "learning_rate": 6.687359394255706e-06, + "loss": 0.3865, + "step": 18568 + }, + { + "epoch": 0.40873871890821045, + "grad_norm": 1.6437114135399726, + "learning_rate": 6.687023844884695e-06, + "loss": 0.3147, + "step": 18569 + }, + { + "epoch": 0.4087607307946291, + "grad_norm": 2.8423713193005593, + "learning_rate": 6.686688286939478e-06, + "loss": 0.2696, + "step": 18570 + }, + { + "epoch": 0.40878274268104775, + "grad_norm": 1.9409941641245871, + "learning_rate": 6.686352720421757e-06, + "loss": 0.356, + "step": 18571 + }, + { + "epoch": 0.40880475456746646, + "grad_norm": 1.8310054241574694, + "learning_rate": 6.68601714533324e-06, + "loss": 0.3297, + "step": 18572 + }, + { + "epoch": 0.4088267664538851, + "grad_norm": 1.5996949214751626, + "learning_rate": 6.685681561675633e-06, + "loss": 0.3329, + "step": 18573 + }, + { + "epoch": 0.40884877834030375, + "grad_norm": 1.4460347874335344, + "learning_rate": 6.68534596945064e-06, + "loss": 0.2458, + "step": 18574 + }, + { + "epoch": 0.4088707902267224, + "grad_norm": 1.6579582864337004, + "learning_rate": 6.6850103686599664e-06, + "loss": 0.3044, + "step": 18575 + }, + { + "epoch": 0.4088928021131411, + "grad_norm": 0.6581285680812151, + "learning_rate": 6.68467475930532e-06, + "loss": 0.49, + "step": 18576 + }, + { + "epoch": 0.40891481399955976, + "grad_norm": 1.7363564104234492, + "learning_rate": 6.684339141388403e-06, + "loss": 0.2606, + "step": 18577 + }, + { + "epoch": 0.4089368258859784, + "grad_norm": 0.6414942163812153, + "learning_rate": 6.6840035149109235e-06, + "loss": 0.5081, + "step": 18578 + }, + { + "epoch": 0.4089588377723971, + "grad_norm": 1.5052584274277279, + "learning_rate": 6.683667879874589e-06, + "loss": 0.2851, + "step": 18579 + }, + { + "epoch": 0.40898084965881576, + "grad_norm": 1.626474038301375, + "learning_rate": 6.683332236281103e-06, + "loss": 0.3136, + "step": 18580 + }, + { + "epoch": 0.4090028615452344, + "grad_norm": 1.6738263694790358, + "learning_rate": 6.682996584132171e-06, + "loss": 0.325, + "step": 18581 + }, + { + "epoch": 0.4090248734316531, + "grad_norm": 1.3759567491789204, + "learning_rate": 6.6826609234294984e-06, + "loss": 0.4039, + "step": 18582 + }, + { + "epoch": 0.40904688531807176, + "grad_norm": 1.4921345689860541, + "learning_rate": 6.682325254174795e-06, + "loss": 0.3603, + "step": 18583 + }, + { + "epoch": 0.4090688972044904, + "grad_norm": 1.2031700760601909, + "learning_rate": 6.681989576369763e-06, + "loss": 0.2727, + "step": 18584 + }, + { + "epoch": 0.4090909090909091, + "grad_norm": 2.078587333984654, + "learning_rate": 6.68165389001611e-06, + "loss": 0.2622, + "step": 18585 + }, + { + "epoch": 0.40911292097732777, + "grad_norm": 1.2153374528047836, + "learning_rate": 6.681318195115543e-06, + "loss": 0.3067, + "step": 18586 + }, + { + "epoch": 0.4091349328637464, + "grad_norm": 1.3182257014956027, + "learning_rate": 6.680982491669764e-06, + "loss": 0.327, + "step": 18587 + }, + { + "epoch": 0.40915694475016506, + "grad_norm": 0.6463794991748467, + "learning_rate": 6.6806467796804855e-06, + "loss": 0.4924, + "step": 18588 + }, + { + "epoch": 0.40917895663658377, + "grad_norm": 1.0629630495738338, + "learning_rate": 6.680311059149408e-06, + "loss": 0.2727, + "step": 18589 + }, + { + "epoch": 0.4092009685230024, + "grad_norm": 1.552621392216875, + "learning_rate": 6.679975330078242e-06, + "loss": 0.3875, + "step": 18590 + }, + { + "epoch": 0.40922298040942107, + "grad_norm": 1.4186428056473877, + "learning_rate": 6.679639592468692e-06, + "loss": 0.3515, + "step": 18591 + }, + { + "epoch": 0.40924499229583977, + "grad_norm": 1.3395940818173546, + "learning_rate": 6.679303846322462e-06, + "loss": 0.3841, + "step": 18592 + }, + { + "epoch": 0.4092670041822584, + "grad_norm": 2.002117746549156, + "learning_rate": 6.678968091641264e-06, + "loss": 0.3851, + "step": 18593 + }, + { + "epoch": 0.40928901606867707, + "grad_norm": 0.6383434076423452, + "learning_rate": 6.6786323284268015e-06, + "loss": 0.5116, + "step": 18594 + }, + { + "epoch": 0.4093110279550958, + "grad_norm": 1.8067526641369231, + "learning_rate": 6.678296556680778e-06, + "loss": 0.2851, + "step": 18595 + }, + { + "epoch": 0.4093330398415144, + "grad_norm": 1.3526432032633056, + "learning_rate": 6.677960776404905e-06, + "loss": 0.3482, + "step": 18596 + }, + { + "epoch": 0.4093550517279331, + "grad_norm": 1.5129255129023291, + "learning_rate": 6.6776249876008855e-06, + "loss": 0.3728, + "step": 18597 + }, + { + "epoch": 0.4093770636143518, + "grad_norm": 1.1158713467024084, + "learning_rate": 6.677289190270428e-06, + "loss": 0.2617, + "step": 18598 + }, + { + "epoch": 0.4093990755007704, + "grad_norm": 2.89154541993696, + "learning_rate": 6.67695338441524e-06, + "loss": 0.3216, + "step": 18599 + }, + { + "epoch": 0.4094210873871891, + "grad_norm": 3.3294142489703695, + "learning_rate": 6.676617570037025e-06, + "loss": 0.3371, + "step": 18600 + }, + { + "epoch": 0.4094430992736077, + "grad_norm": 2.0418967499847436, + "learning_rate": 6.676281747137493e-06, + "loss": 0.4261, + "step": 18601 + }, + { + "epoch": 0.40946511116002643, + "grad_norm": 1.3536663600552574, + "learning_rate": 6.675945915718349e-06, + "loss": 0.4097, + "step": 18602 + }, + { + "epoch": 0.4094871230464451, + "grad_norm": 1.3854673071768178, + "learning_rate": 6.675610075781301e-06, + "loss": 0.3776, + "step": 18603 + }, + { + "epoch": 0.4095091349328637, + "grad_norm": 2.1349134265766136, + "learning_rate": 6.675274227328056e-06, + "loss": 0.3221, + "step": 18604 + }, + { + "epoch": 0.40953114681928243, + "grad_norm": 1.884362710310961, + "learning_rate": 6.674938370360318e-06, + "loss": 0.28, + "step": 18605 + }, + { + "epoch": 0.4095531587057011, + "grad_norm": 1.41393561870847, + "learning_rate": 6.674602504879798e-06, + "loss": 0.3502, + "step": 18606 + }, + { + "epoch": 0.40957517059211973, + "grad_norm": 1.330132998042442, + "learning_rate": 6.6742666308882e-06, + "loss": 0.2648, + "step": 18607 + }, + { + "epoch": 0.40959718247853844, + "grad_norm": 1.6144929623647488, + "learning_rate": 6.673930748387233e-06, + "loss": 0.3869, + "step": 18608 + }, + { + "epoch": 0.4096191943649571, + "grad_norm": 2.757462617831344, + "learning_rate": 6.673594857378603e-06, + "loss": 0.3994, + "step": 18609 + }, + { + "epoch": 0.40964120625137573, + "grad_norm": 1.5409681597958216, + "learning_rate": 6.673258957864016e-06, + "loss": 0.3228, + "step": 18610 + }, + { + "epoch": 0.4096632181377944, + "grad_norm": 1.3553584656978606, + "learning_rate": 6.672923049845183e-06, + "loss": 0.3905, + "step": 18611 + }, + { + "epoch": 0.4096852300242131, + "grad_norm": 1.7189872948880904, + "learning_rate": 6.672587133323809e-06, + "loss": 0.4123, + "step": 18612 + }, + { + "epoch": 0.40970724191063174, + "grad_norm": 1.8699474981948303, + "learning_rate": 6.672251208301599e-06, + "loss": 0.4112, + "step": 18613 + }, + { + "epoch": 0.4097292537970504, + "grad_norm": 1.4285572060202547, + "learning_rate": 6.671915274780264e-06, + "loss": 0.3636, + "step": 18614 + }, + { + "epoch": 0.4097512656834691, + "grad_norm": 1.612038588204738, + "learning_rate": 6.671579332761511e-06, + "loss": 0.3673, + "step": 18615 + }, + { + "epoch": 0.40977327756988774, + "grad_norm": 1.3179666788679052, + "learning_rate": 6.6712433822470446e-06, + "loss": 0.2724, + "step": 18616 + }, + { + "epoch": 0.4097952894563064, + "grad_norm": 1.0592150714926931, + "learning_rate": 6.670907423238574e-06, + "loss": 0.1801, + "step": 18617 + }, + { + "epoch": 0.4098173013427251, + "grad_norm": 1.5984403516867893, + "learning_rate": 6.670571455737807e-06, + "loss": 0.301, + "step": 18618 + }, + { + "epoch": 0.40983931322914374, + "grad_norm": 1.2207061662333283, + "learning_rate": 6.67023547974645e-06, + "loss": 0.3051, + "step": 18619 + }, + { + "epoch": 0.4098613251155624, + "grad_norm": 1.543461886168096, + "learning_rate": 6.669899495266212e-06, + "loss": 0.3315, + "step": 18620 + }, + { + "epoch": 0.4098833370019811, + "grad_norm": 9.612998895358485, + "learning_rate": 6.669563502298801e-06, + "loss": 0.2747, + "step": 18621 + }, + { + "epoch": 0.40990534888839975, + "grad_norm": 2.318207611498643, + "learning_rate": 6.669227500845923e-06, + "loss": 0.2526, + "step": 18622 + }, + { + "epoch": 0.4099273607748184, + "grad_norm": 1.5207643257094257, + "learning_rate": 6.668891490909284e-06, + "loss": 0.3082, + "step": 18623 + }, + { + "epoch": 0.40994937266123704, + "grad_norm": 0.6449041252636833, + "learning_rate": 6.6685554724905975e-06, + "loss": 0.4856, + "step": 18624 + }, + { + "epoch": 0.40997138454765575, + "grad_norm": 2.419225448800439, + "learning_rate": 6.668219445591566e-06, + "loss": 0.3286, + "step": 18625 + }, + { + "epoch": 0.4099933964340744, + "grad_norm": 2.245401813571251, + "learning_rate": 6.6678834102139e-06, + "loss": 0.3241, + "step": 18626 + }, + { + "epoch": 0.41001540832049305, + "grad_norm": 1.1574426704979295, + "learning_rate": 6.667547366359306e-06, + "loss": 0.2135, + "step": 18627 + }, + { + "epoch": 0.41003742020691175, + "grad_norm": 0.6159192823656885, + "learning_rate": 6.667211314029492e-06, + "loss": 0.4659, + "step": 18628 + }, + { + "epoch": 0.4100594320933304, + "grad_norm": 0.6337074525892543, + "learning_rate": 6.666875253226168e-06, + "loss": 0.4959, + "step": 18629 + }, + { + "epoch": 0.41008144397974905, + "grad_norm": 1.6731252621946953, + "learning_rate": 6.666539183951041e-06, + "loss": 0.2873, + "step": 18630 + }, + { + "epoch": 0.41010345586616775, + "grad_norm": 1.5265420802198006, + "learning_rate": 6.666203106205816e-06, + "loss": 0.37, + "step": 18631 + }, + { + "epoch": 0.4101254677525864, + "grad_norm": 1.352303512589682, + "learning_rate": 6.665867019992205e-06, + "loss": 0.3136, + "step": 18632 + }, + { + "epoch": 0.41014747963900505, + "grad_norm": 1.132022842437004, + "learning_rate": 6.665530925311916e-06, + "loss": 0.2507, + "step": 18633 + }, + { + "epoch": 0.4101694915254237, + "grad_norm": 1.2493295036379901, + "learning_rate": 6.665194822166654e-06, + "loss": 0.2894, + "step": 18634 + }, + { + "epoch": 0.4101915034118424, + "grad_norm": 0.6434132495421663, + "learning_rate": 6.664858710558131e-06, + "loss": 0.5025, + "step": 18635 + }, + { + "epoch": 0.41021351529826106, + "grad_norm": 2.179898398073388, + "learning_rate": 6.664522590488051e-06, + "loss": 0.3958, + "step": 18636 + }, + { + "epoch": 0.4102355271846797, + "grad_norm": 0.6206831891568526, + "learning_rate": 6.664186461958127e-06, + "loss": 0.492, + "step": 18637 + }, + { + "epoch": 0.4102575390710984, + "grad_norm": 1.3287967464282073, + "learning_rate": 6.663850324970065e-06, + "loss": 0.3089, + "step": 18638 + }, + { + "epoch": 0.41027955095751706, + "grad_norm": 1.6970128796993207, + "learning_rate": 6.6635141795255735e-06, + "loss": 0.3448, + "step": 18639 + }, + { + "epoch": 0.4103015628439357, + "grad_norm": 1.0716203686676526, + "learning_rate": 6.66317802562636e-06, + "loss": 0.1879, + "step": 18640 + }, + { + "epoch": 0.4103235747303544, + "grad_norm": 0.6332299355519423, + "learning_rate": 6.6628418632741345e-06, + "loss": 0.4901, + "step": 18641 + }, + { + "epoch": 0.41034558661677306, + "grad_norm": 2.7157848143116508, + "learning_rate": 6.662505692470605e-06, + "loss": 0.3405, + "step": 18642 + }, + { + "epoch": 0.4103675985031917, + "grad_norm": 1.158947262661863, + "learning_rate": 6.662169513217481e-06, + "loss": 0.3505, + "step": 18643 + }, + { + "epoch": 0.4103896103896104, + "grad_norm": 2.0987860421006843, + "learning_rate": 6.661833325516469e-06, + "loss": 0.2369, + "step": 18644 + }, + { + "epoch": 0.41041162227602906, + "grad_norm": 1.8763449106994512, + "learning_rate": 6.661497129369279e-06, + "loss": 0.319, + "step": 18645 + }, + { + "epoch": 0.4104336341624477, + "grad_norm": 0.6549109803201599, + "learning_rate": 6.661160924777619e-06, + "loss": 0.5032, + "step": 18646 + }, + { + "epoch": 0.41045564604886636, + "grad_norm": 1.4194525940668223, + "learning_rate": 6.6608247117432005e-06, + "loss": 0.3907, + "step": 18647 + }, + { + "epoch": 0.41047765793528507, + "grad_norm": 1.3738274219390219, + "learning_rate": 6.660488490267728e-06, + "loss": 0.3487, + "step": 18648 + }, + { + "epoch": 0.4104996698217037, + "grad_norm": 0.6218104653119033, + "learning_rate": 6.660152260352912e-06, + "loss": 0.4986, + "step": 18649 + }, + { + "epoch": 0.41052168170812237, + "grad_norm": 0.5982755351870752, + "learning_rate": 6.6598160220004625e-06, + "loss": 0.4834, + "step": 18650 + }, + { + "epoch": 0.41054369359454107, + "grad_norm": 1.5690314195309434, + "learning_rate": 6.659479775212089e-06, + "loss": 0.3185, + "step": 18651 + }, + { + "epoch": 0.4105657054809597, + "grad_norm": 1.4999127082216794, + "learning_rate": 6.659143519989498e-06, + "loss": 0.3808, + "step": 18652 + }, + { + "epoch": 0.41058771736737837, + "grad_norm": 1.4348003319131917, + "learning_rate": 6.6588072563344e-06, + "loss": 0.3538, + "step": 18653 + }, + { + "epoch": 0.4106097292537971, + "grad_norm": 1.9955296961290303, + "learning_rate": 6.658470984248502e-06, + "loss": 0.392, + "step": 18654 + }, + { + "epoch": 0.4106317411402157, + "grad_norm": 1.3088651796203545, + "learning_rate": 6.658134703733516e-06, + "loss": 0.3159, + "step": 18655 + }, + { + "epoch": 0.41065375302663437, + "grad_norm": 1.8738634649152086, + "learning_rate": 6.65779841479115e-06, + "loss": 0.2515, + "step": 18656 + }, + { + "epoch": 0.4106757649130531, + "grad_norm": 0.6569669428451765, + "learning_rate": 6.657462117423113e-06, + "loss": 0.4776, + "step": 18657 + }, + { + "epoch": 0.4106977767994717, + "grad_norm": 1.3991839634577923, + "learning_rate": 6.657125811631116e-06, + "loss": 0.3412, + "step": 18658 + }, + { + "epoch": 0.4107197886858904, + "grad_norm": 0.6144890647687572, + "learning_rate": 6.656789497416864e-06, + "loss": 0.5027, + "step": 18659 + }, + { + "epoch": 0.410741800572309, + "grad_norm": 1.4843210394710624, + "learning_rate": 6.656453174782069e-06, + "loss": 0.408, + "step": 18660 + }, + { + "epoch": 0.41076381245872773, + "grad_norm": 1.5147900893260662, + "learning_rate": 6.6561168437284415e-06, + "loss": 0.3217, + "step": 18661 + }, + { + "epoch": 0.4107858243451464, + "grad_norm": 0.6335161480615196, + "learning_rate": 6.655780504257688e-06, + "loss": 0.4934, + "step": 18662 + }, + { + "epoch": 0.410807836231565, + "grad_norm": 1.268089680318383, + "learning_rate": 6.655444156371521e-06, + "loss": 0.2943, + "step": 18663 + }, + { + "epoch": 0.41082984811798373, + "grad_norm": 0.6368688736644549, + "learning_rate": 6.6551078000716475e-06, + "loss": 0.505, + "step": 18664 + }, + { + "epoch": 0.4108518600044024, + "grad_norm": 1.8355877540841876, + "learning_rate": 6.654771435359779e-06, + "loss": 0.3051, + "step": 18665 + }, + { + "epoch": 0.41087387189082103, + "grad_norm": 1.4849404261206685, + "learning_rate": 6.654435062237623e-06, + "loss": 0.339, + "step": 18666 + }, + { + "epoch": 0.41089588377723973, + "grad_norm": 1.2795917935995313, + "learning_rate": 6.6540986807068895e-06, + "loss": 0.3184, + "step": 18667 + }, + { + "epoch": 0.4109178956636584, + "grad_norm": 0.6451533859651958, + "learning_rate": 6.65376229076929e-06, + "loss": 0.5115, + "step": 18668 + }, + { + "epoch": 0.41093990755007703, + "grad_norm": 1.5440959438019126, + "learning_rate": 6.653425892426533e-06, + "loss": 0.3326, + "step": 18669 + }, + { + "epoch": 0.4109619194364957, + "grad_norm": 0.6092156804800521, + "learning_rate": 6.653089485680328e-06, + "loss": 0.4769, + "step": 18670 + }, + { + "epoch": 0.4109839313229144, + "grad_norm": 0.6143396129626187, + "learning_rate": 6.652753070532385e-06, + "loss": 0.4909, + "step": 18671 + }, + { + "epoch": 0.41100594320933304, + "grad_norm": 1.6397282574380116, + "learning_rate": 6.652416646984412e-06, + "loss": 0.3647, + "step": 18672 + }, + { + "epoch": 0.4110279550957517, + "grad_norm": 1.4495894572421617, + "learning_rate": 6.6520802150381225e-06, + "loss": 0.2247, + "step": 18673 + }, + { + "epoch": 0.4110499669821704, + "grad_norm": 1.8794817704370324, + "learning_rate": 6.651743774695224e-06, + "loss": 0.3454, + "step": 18674 + }, + { + "epoch": 0.41107197886858904, + "grad_norm": 10.142734937889074, + "learning_rate": 6.651407325957427e-06, + "loss": 0.3206, + "step": 18675 + }, + { + "epoch": 0.4110939907550077, + "grad_norm": 1.489681987716777, + "learning_rate": 6.651070868826443e-06, + "loss": 0.2924, + "step": 18676 + }, + { + "epoch": 0.4111160026414264, + "grad_norm": 1.3749982754374248, + "learning_rate": 6.6507344033039765e-06, + "loss": 0.3932, + "step": 18677 + }, + { + "epoch": 0.41113801452784504, + "grad_norm": 1.2635223293677518, + "learning_rate": 6.650397929391745e-06, + "loss": 0.2689, + "step": 18678 + }, + { + "epoch": 0.4111600264142637, + "grad_norm": 0.6188686458383919, + "learning_rate": 6.650061447091455e-06, + "loss": 0.4969, + "step": 18679 + }, + { + "epoch": 0.4111820383006824, + "grad_norm": 1.5692819435759962, + "learning_rate": 6.649724956404815e-06, + "loss": 0.3236, + "step": 18680 + }, + { + "epoch": 0.41120405018710104, + "grad_norm": 1.8053340351597416, + "learning_rate": 6.649388457333538e-06, + "loss": 0.3001, + "step": 18681 + }, + { + "epoch": 0.4112260620735197, + "grad_norm": 1.1345385116888997, + "learning_rate": 6.6490519498793325e-06, + "loss": 0.2512, + "step": 18682 + }, + { + "epoch": 0.41124807395993834, + "grad_norm": 1.1154592457082888, + "learning_rate": 6.6487154340439105e-06, + "loss": 0.396, + "step": 18683 + }, + { + "epoch": 0.41127008584635705, + "grad_norm": 0.6149638077252948, + "learning_rate": 6.648378909828981e-06, + "loss": 0.4822, + "step": 18684 + }, + { + "epoch": 0.4112920977327757, + "grad_norm": 0.640433111591014, + "learning_rate": 6.648042377236253e-06, + "loss": 0.4928, + "step": 18685 + }, + { + "epoch": 0.41131410961919435, + "grad_norm": 1.4157411756891116, + "learning_rate": 6.647705836267441e-06, + "loss": 0.249, + "step": 18686 + }, + { + "epoch": 0.41133612150561305, + "grad_norm": 1.3065777941575794, + "learning_rate": 6.6473692869242526e-06, + "loss": 0.3063, + "step": 18687 + }, + { + "epoch": 0.4113581333920317, + "grad_norm": 1.568497998966328, + "learning_rate": 6.647032729208398e-06, + "loss": 0.3081, + "step": 18688 + }, + { + "epoch": 0.41138014527845035, + "grad_norm": 1.1035372746272054, + "learning_rate": 6.646696163121589e-06, + "loss": 0.2971, + "step": 18689 + }, + { + "epoch": 0.41140215716486905, + "grad_norm": 1.3025106630249876, + "learning_rate": 6.646359588665534e-06, + "loss": 0.3093, + "step": 18690 + }, + { + "epoch": 0.4114241690512877, + "grad_norm": 1.5771434177323296, + "learning_rate": 6.646023005841946e-06, + "loss": 0.3501, + "step": 18691 + }, + { + "epoch": 0.41144618093770635, + "grad_norm": 0.6465228924331451, + "learning_rate": 6.645686414652537e-06, + "loss": 0.5033, + "step": 18692 + }, + { + "epoch": 0.411468192824125, + "grad_norm": 1.3177053109699992, + "learning_rate": 6.645349815099014e-06, + "loss": 0.3312, + "step": 18693 + }, + { + "epoch": 0.4114902047105437, + "grad_norm": 2.282969594313594, + "learning_rate": 6.64501320718309e-06, + "loss": 0.3396, + "step": 18694 + }, + { + "epoch": 0.41151221659696235, + "grad_norm": 1.5407445697536064, + "learning_rate": 6.6446765909064736e-06, + "loss": 0.3242, + "step": 18695 + }, + { + "epoch": 0.411534228483381, + "grad_norm": 2.589248660323248, + "learning_rate": 6.644339966270878e-06, + "loss": 0.3612, + "step": 18696 + }, + { + "epoch": 0.4115562403697997, + "grad_norm": 1.9862630479141772, + "learning_rate": 6.644003333278015e-06, + "loss": 0.4186, + "step": 18697 + }, + { + "epoch": 0.41157825225621836, + "grad_norm": 1.7126423414686185, + "learning_rate": 6.643666691929592e-06, + "loss": 0.2958, + "step": 18698 + }, + { + "epoch": 0.411600264142637, + "grad_norm": 1.0915727517542952, + "learning_rate": 6.643330042227323e-06, + "loss": 0.3162, + "step": 18699 + }, + { + "epoch": 0.4116222760290557, + "grad_norm": 1.4985114649921054, + "learning_rate": 6.642993384172917e-06, + "loss": 0.358, + "step": 18700 + }, + { + "epoch": 0.41164428791547436, + "grad_norm": 1.9968587045544712, + "learning_rate": 6.642656717768085e-06, + "loss": 0.3501, + "step": 18701 + }, + { + "epoch": 0.411666299801893, + "grad_norm": 0.6533892489151407, + "learning_rate": 6.642320043014541e-06, + "loss": 0.5153, + "step": 18702 + }, + { + "epoch": 0.4116883116883117, + "grad_norm": 7.9591196330873375, + "learning_rate": 6.641983359913991e-06, + "loss": 0.4122, + "step": 18703 + }, + { + "epoch": 0.41171032357473036, + "grad_norm": 1.7066674086720148, + "learning_rate": 6.641646668468151e-06, + "loss": 0.3336, + "step": 18704 + }, + { + "epoch": 0.411732335461149, + "grad_norm": 2.1222872567137063, + "learning_rate": 6.641309968678731e-06, + "loss": 0.3633, + "step": 18705 + }, + { + "epoch": 0.41175434734756766, + "grad_norm": 1.1113006713630549, + "learning_rate": 6.640973260547442e-06, + "loss": 0.2931, + "step": 18706 + }, + { + "epoch": 0.41177635923398637, + "grad_norm": 1.5886475457115004, + "learning_rate": 6.640636544075995e-06, + "loss": 0.3396, + "step": 18707 + }, + { + "epoch": 0.411798371120405, + "grad_norm": 1.3440651428539352, + "learning_rate": 6.640299819266098e-06, + "loss": 0.3484, + "step": 18708 + }, + { + "epoch": 0.41182038300682366, + "grad_norm": 1.5674202178828502, + "learning_rate": 6.63996308611947e-06, + "loss": 0.3235, + "step": 18709 + }, + { + "epoch": 0.41184239489324237, + "grad_norm": 1.1867082853533728, + "learning_rate": 6.6396263446378164e-06, + "loss": 0.2744, + "step": 18710 + }, + { + "epoch": 0.411864406779661, + "grad_norm": 1.8237537177075787, + "learning_rate": 6.63928959482285e-06, + "loss": 0.3427, + "step": 18711 + }, + { + "epoch": 0.41188641866607967, + "grad_norm": 1.2387048929465394, + "learning_rate": 6.638952836676284e-06, + "loss": 0.4213, + "step": 18712 + }, + { + "epoch": 0.4119084305524984, + "grad_norm": 1.3780850142133103, + "learning_rate": 6.6386160701998256e-06, + "loss": 0.2428, + "step": 18713 + }, + { + "epoch": 0.411930442438917, + "grad_norm": 1.3966200762799141, + "learning_rate": 6.638279295395192e-06, + "loss": 0.2644, + "step": 18714 + }, + { + "epoch": 0.41195245432533567, + "grad_norm": 1.5232524840401973, + "learning_rate": 6.6379425122640935e-06, + "loss": 0.3109, + "step": 18715 + }, + { + "epoch": 0.4119744662117543, + "grad_norm": 0.6425103150049665, + "learning_rate": 6.637605720808239e-06, + "loss": 0.5028, + "step": 18716 + }, + { + "epoch": 0.411996478098173, + "grad_norm": 1.5707556829064468, + "learning_rate": 6.637268921029341e-06, + "loss": 0.3826, + "step": 18717 + }, + { + "epoch": 0.4120184899845917, + "grad_norm": 1.1454275783953558, + "learning_rate": 6.636932112929114e-06, + "loss": 0.305, + "step": 18718 + }, + { + "epoch": 0.4120405018710103, + "grad_norm": 1.5894919453172798, + "learning_rate": 6.636595296509266e-06, + "loss": 0.3053, + "step": 18719 + }, + { + "epoch": 0.412062513757429, + "grad_norm": 1.511407614888257, + "learning_rate": 6.6362584717715125e-06, + "loss": 0.3318, + "step": 18720 + }, + { + "epoch": 0.4120845256438477, + "grad_norm": 1.8133368154869978, + "learning_rate": 6.6359216387175625e-06, + "loss": 0.2486, + "step": 18721 + }, + { + "epoch": 0.4121065375302663, + "grad_norm": 0.6127422708305168, + "learning_rate": 6.63558479734913e-06, + "loss": 0.5029, + "step": 18722 + }, + { + "epoch": 0.41212854941668503, + "grad_norm": 2.172421923209309, + "learning_rate": 6.635247947667925e-06, + "loss": 0.3706, + "step": 18723 + }, + { + "epoch": 0.4121505613031037, + "grad_norm": 1.3123315307592256, + "learning_rate": 6.634911089675662e-06, + "loss": 0.344, + "step": 18724 + }, + { + "epoch": 0.41217257318952233, + "grad_norm": 0.6232162333797171, + "learning_rate": 6.634574223374051e-06, + "loss": 0.4952, + "step": 18725 + }, + { + "epoch": 0.41219458507594103, + "grad_norm": 1.7547690508522547, + "learning_rate": 6.634237348764804e-06, + "loss": 0.3086, + "step": 18726 + }, + { + "epoch": 0.4122165969623597, + "grad_norm": 1.0345920388362368, + "learning_rate": 6.633900465849635e-06, + "loss": 0.3191, + "step": 18727 + }, + { + "epoch": 0.41223860884877833, + "grad_norm": 1.4181688111211839, + "learning_rate": 6.633563574630254e-06, + "loss": 0.3588, + "step": 18728 + }, + { + "epoch": 0.412260620735197, + "grad_norm": 2.0915947054990993, + "learning_rate": 6.633226675108375e-06, + "loss": 0.3208, + "step": 18729 + }, + { + "epoch": 0.4122826326216157, + "grad_norm": 1.5332487634732306, + "learning_rate": 6.6328897672857105e-06, + "loss": 0.3396, + "step": 18730 + }, + { + "epoch": 0.41230464450803433, + "grad_norm": 1.3689468432300886, + "learning_rate": 6.6325528511639695e-06, + "loss": 0.3675, + "step": 18731 + }, + { + "epoch": 0.412326656394453, + "grad_norm": 1.4348349030986587, + "learning_rate": 6.632215926744868e-06, + "loss": 0.3666, + "step": 18732 + }, + { + "epoch": 0.4123486682808717, + "grad_norm": 1.204300544762728, + "learning_rate": 6.631878994030119e-06, + "loss": 0.2734, + "step": 18733 + }, + { + "epoch": 0.41237068016729034, + "grad_norm": 0.6508046017309208, + "learning_rate": 6.6315420530214316e-06, + "loss": 0.4743, + "step": 18734 + }, + { + "epoch": 0.412392692053709, + "grad_norm": 0.6292083111984071, + "learning_rate": 6.63120510372052e-06, + "loss": 0.5101, + "step": 18735 + }, + { + "epoch": 0.4124147039401277, + "grad_norm": 1.6342531514181695, + "learning_rate": 6.630868146129096e-06, + "loss": 0.3333, + "step": 18736 + }, + { + "epoch": 0.41243671582654634, + "grad_norm": 1.4106204456400822, + "learning_rate": 6.630531180248871e-06, + "loss": 0.2844, + "step": 18737 + }, + { + "epoch": 0.412458727712965, + "grad_norm": 1.447175856791743, + "learning_rate": 6.6301942060815625e-06, + "loss": 0.4012, + "step": 18738 + }, + { + "epoch": 0.4124807395993837, + "grad_norm": 1.405909727271869, + "learning_rate": 6.629857223628878e-06, + "loss": 0.3395, + "step": 18739 + }, + { + "epoch": 0.41250275148580234, + "grad_norm": 0.633738085124273, + "learning_rate": 6.629520232892533e-06, + "loss": 0.4855, + "step": 18740 + }, + { + "epoch": 0.412524763372221, + "grad_norm": 1.5678618559392743, + "learning_rate": 6.629183233874239e-06, + "loss": 0.3321, + "step": 18741 + }, + { + "epoch": 0.41254677525863964, + "grad_norm": 2.5273901319892413, + "learning_rate": 6.628846226575709e-06, + "loss": 0.2897, + "step": 18742 + }, + { + "epoch": 0.41256878714505835, + "grad_norm": 1.5413805720035516, + "learning_rate": 6.628509210998657e-06, + "loss": 0.3111, + "step": 18743 + }, + { + "epoch": 0.412590799031477, + "grad_norm": 0.6160098072771636, + "learning_rate": 6.628172187144795e-06, + "loss": 0.4773, + "step": 18744 + }, + { + "epoch": 0.41261281091789564, + "grad_norm": 1.1862426913078874, + "learning_rate": 6.627835155015834e-06, + "loss": 0.2501, + "step": 18745 + }, + { + "epoch": 0.41263482280431435, + "grad_norm": 1.3332366908777669, + "learning_rate": 6.62749811461349e-06, + "loss": 0.3085, + "step": 18746 + }, + { + "epoch": 0.412656834690733, + "grad_norm": 1.6972258696405393, + "learning_rate": 6.627161065939475e-06, + "loss": 0.3588, + "step": 18747 + }, + { + "epoch": 0.41267884657715165, + "grad_norm": 1.461737448702887, + "learning_rate": 6.6268240089955005e-06, + "loss": 0.3479, + "step": 18748 + }, + { + "epoch": 0.41270085846357035, + "grad_norm": 1.4962965088231837, + "learning_rate": 6.626486943783282e-06, + "loss": 0.3583, + "step": 18749 + }, + { + "epoch": 0.412722870349989, + "grad_norm": 0.6271658534914765, + "learning_rate": 6.62614987030453e-06, + "loss": 0.4957, + "step": 18750 + }, + { + "epoch": 0.41274488223640765, + "grad_norm": 1.8681716360625902, + "learning_rate": 6.62581278856096e-06, + "loss": 0.2987, + "step": 18751 + }, + { + "epoch": 0.4127668941228263, + "grad_norm": 1.236012377727643, + "learning_rate": 6.625475698554285e-06, + "loss": 0.2686, + "step": 18752 + }, + { + "epoch": 0.412788906009245, + "grad_norm": 1.5995500645421827, + "learning_rate": 6.625138600286216e-06, + "loss": 0.41, + "step": 18753 + }, + { + "epoch": 0.41281091789566365, + "grad_norm": 1.638237081873195, + "learning_rate": 6.624801493758468e-06, + "loss": 0.3281, + "step": 18754 + }, + { + "epoch": 0.4128329297820823, + "grad_norm": 1.5984348718645212, + "learning_rate": 6.624464378972755e-06, + "loss": 0.3171, + "step": 18755 + }, + { + "epoch": 0.412854941668501, + "grad_norm": 2.3461015390134, + "learning_rate": 6.62412725593079e-06, + "loss": 0.2561, + "step": 18756 + }, + { + "epoch": 0.41287695355491966, + "grad_norm": 20.915728792623195, + "learning_rate": 6.623790124634284e-06, + "loss": 0.2739, + "step": 18757 + }, + { + "epoch": 0.4128989654413383, + "grad_norm": 1.1881162663748182, + "learning_rate": 6.623452985084954e-06, + "loss": 0.3041, + "step": 18758 + }, + { + "epoch": 0.412920977327757, + "grad_norm": 1.4864068307965117, + "learning_rate": 6.623115837284511e-06, + "loss": 0.3501, + "step": 18759 + }, + { + "epoch": 0.41294298921417566, + "grad_norm": 1.5650758819109873, + "learning_rate": 6.622778681234669e-06, + "loss": 0.2537, + "step": 18760 + }, + { + "epoch": 0.4129650011005943, + "grad_norm": 1.453669755771795, + "learning_rate": 6.6224415169371434e-06, + "loss": 0.3088, + "step": 18761 + }, + { + "epoch": 0.412987012987013, + "grad_norm": 1.2898310773629698, + "learning_rate": 6.622104344393645e-06, + "loss": 0.271, + "step": 18762 + }, + { + "epoch": 0.41300902487343166, + "grad_norm": 1.7348719350128505, + "learning_rate": 6.62176716360589e-06, + "loss": 0.2807, + "step": 18763 + }, + { + "epoch": 0.4130310367598503, + "grad_norm": 1.2544655760000454, + "learning_rate": 6.621429974575591e-06, + "loss": 0.2663, + "step": 18764 + }, + { + "epoch": 0.41305304864626896, + "grad_norm": 1.6903618814539207, + "learning_rate": 6.6210927773044605e-06, + "loss": 0.3514, + "step": 18765 + }, + { + "epoch": 0.41307506053268767, + "grad_norm": 1.3022323006764767, + "learning_rate": 6.620755571794214e-06, + "loss": 0.3969, + "step": 18766 + }, + { + "epoch": 0.4130970724191063, + "grad_norm": 1.2863715472198063, + "learning_rate": 6.620418358046564e-06, + "loss": 0.2952, + "step": 18767 + }, + { + "epoch": 0.41311908430552496, + "grad_norm": 1.2880273959546154, + "learning_rate": 6.620081136063226e-06, + "loss": 0.3555, + "step": 18768 + }, + { + "epoch": 0.41314109619194367, + "grad_norm": 1.6120600449091673, + "learning_rate": 6.619743905845913e-06, + "loss": 0.2747, + "step": 18769 + }, + { + "epoch": 0.4131631080783623, + "grad_norm": 1.4278142370809204, + "learning_rate": 6.61940666739634e-06, + "loss": 0.33, + "step": 18770 + }, + { + "epoch": 0.41318511996478097, + "grad_norm": 1.6505065028298658, + "learning_rate": 6.619069420716218e-06, + "loss": 0.2157, + "step": 18771 + }, + { + "epoch": 0.41320713185119967, + "grad_norm": 1.600505335955668, + "learning_rate": 6.618732165807266e-06, + "loss": 0.3191, + "step": 18772 + }, + { + "epoch": 0.4132291437376183, + "grad_norm": 1.0438164700092945, + "learning_rate": 6.618394902671193e-06, + "loss": 0.2334, + "step": 18773 + }, + { + "epoch": 0.41325115562403697, + "grad_norm": 1.758226418252415, + "learning_rate": 6.618057631309715e-06, + "loss": 0.3773, + "step": 18774 + }, + { + "epoch": 0.4132731675104556, + "grad_norm": 1.4894919718531217, + "learning_rate": 6.617720351724547e-06, + "loss": 0.3483, + "step": 18775 + }, + { + "epoch": 0.4132951793968743, + "grad_norm": 1.5352076542923694, + "learning_rate": 6.617383063917403e-06, + "loss": 0.363, + "step": 18776 + }, + { + "epoch": 0.413317191283293, + "grad_norm": 1.5705209300574337, + "learning_rate": 6.617045767889997e-06, + "loss": 0.2767, + "step": 18777 + }, + { + "epoch": 0.4133392031697116, + "grad_norm": 1.3256622832947014, + "learning_rate": 6.616708463644042e-06, + "loss": 0.3155, + "step": 18778 + }, + { + "epoch": 0.4133612150561303, + "grad_norm": 1.338383776411618, + "learning_rate": 6.616371151181256e-06, + "loss": 0.3136, + "step": 18779 + }, + { + "epoch": 0.413383226942549, + "grad_norm": 1.3510036776490872, + "learning_rate": 6.616033830503348e-06, + "loss": 0.2794, + "step": 18780 + }, + { + "epoch": 0.4134052388289676, + "grad_norm": 1.8362381931820733, + "learning_rate": 6.615696501612036e-06, + "loss": 0.4184, + "step": 18781 + }, + { + "epoch": 0.41342725071538633, + "grad_norm": 1.6929474977368084, + "learning_rate": 6.615359164509035e-06, + "loss": 0.3118, + "step": 18782 + }, + { + "epoch": 0.413449262601805, + "grad_norm": 1.2927772664513053, + "learning_rate": 6.615021819196058e-06, + "loss": 0.3121, + "step": 18783 + }, + { + "epoch": 0.41347127448822363, + "grad_norm": 0.664329541752334, + "learning_rate": 6.6146844656748195e-06, + "loss": 0.5074, + "step": 18784 + }, + { + "epoch": 0.41349328637464233, + "grad_norm": 1.5222163617986386, + "learning_rate": 6.614347103947034e-06, + "loss": 0.375, + "step": 18785 + }, + { + "epoch": 0.413515298261061, + "grad_norm": 1.3084882259524446, + "learning_rate": 6.614009734014416e-06, + "loss": 0.272, + "step": 18786 + }, + { + "epoch": 0.41353731014747963, + "grad_norm": 1.429621979125079, + "learning_rate": 6.613672355878682e-06, + "loss": 0.4026, + "step": 18787 + }, + { + "epoch": 0.4135593220338983, + "grad_norm": 0.618208315116707, + "learning_rate": 6.6133349695415446e-06, + "loss": 0.5058, + "step": 18788 + }, + { + "epoch": 0.413581333920317, + "grad_norm": 1.2745951266960627, + "learning_rate": 6.612997575004721e-06, + "loss": 0.3065, + "step": 18789 + }, + { + "epoch": 0.41360334580673563, + "grad_norm": 1.1540515652645928, + "learning_rate": 6.6126601722699225e-06, + "loss": 0.3094, + "step": 18790 + }, + { + "epoch": 0.4136253576931543, + "grad_norm": 1.700465252450133, + "learning_rate": 6.612322761338865e-06, + "loss": 0.365, + "step": 18791 + }, + { + "epoch": 0.413647369579573, + "grad_norm": 1.391943352570697, + "learning_rate": 6.611985342213265e-06, + "loss": 0.2822, + "step": 18792 + }, + { + "epoch": 0.41366938146599164, + "grad_norm": 1.524209111242416, + "learning_rate": 6.611647914894836e-06, + "loss": 0.3219, + "step": 18793 + }, + { + "epoch": 0.4136913933524103, + "grad_norm": 1.6966493009477075, + "learning_rate": 6.611310479385296e-06, + "loss": 0.4157, + "step": 18794 + }, + { + "epoch": 0.413713405238829, + "grad_norm": 1.5875878419128429, + "learning_rate": 6.6109730356863555e-06, + "loss": 0.371, + "step": 18795 + }, + { + "epoch": 0.41373541712524764, + "grad_norm": 1.4262857588550615, + "learning_rate": 6.61063558379973e-06, + "loss": 0.3255, + "step": 18796 + }, + { + "epoch": 0.4137574290116663, + "grad_norm": 1.7596625159611086, + "learning_rate": 6.610298123727139e-06, + "loss": 0.3904, + "step": 18797 + }, + { + "epoch": 0.413779440898085, + "grad_norm": 1.2425044481832097, + "learning_rate": 6.609960655470293e-06, + "loss": 0.3485, + "step": 18798 + }, + { + "epoch": 0.41380145278450364, + "grad_norm": 1.4435346900987545, + "learning_rate": 6.60962317903091e-06, + "loss": 0.3359, + "step": 18799 + }, + { + "epoch": 0.4138234646709223, + "grad_norm": 1.4876175117532182, + "learning_rate": 6.609285694410703e-06, + "loss": 0.3754, + "step": 18800 + }, + { + "epoch": 0.41384547655734094, + "grad_norm": 31.312197136548175, + "learning_rate": 6.608948201611389e-06, + "loss": 0.309, + "step": 18801 + }, + { + "epoch": 0.41386748844375965, + "grad_norm": 1.4176789083212846, + "learning_rate": 6.6086107006346835e-06, + "loss": 0.3033, + "step": 18802 + }, + { + "epoch": 0.4138895003301783, + "grad_norm": 3.3246439610510814, + "learning_rate": 6.608273191482299e-06, + "loss": 0.362, + "step": 18803 + }, + { + "epoch": 0.41391151221659694, + "grad_norm": 1.5250961418590931, + "learning_rate": 6.607935674155953e-06, + "loss": 0.4332, + "step": 18804 + }, + { + "epoch": 0.41393352410301565, + "grad_norm": 1.3678256201490753, + "learning_rate": 6.607598148657362e-06, + "loss": 0.2977, + "step": 18805 + }, + { + "epoch": 0.4139555359894343, + "grad_norm": 1.6312937224832196, + "learning_rate": 6.60726061498824e-06, + "loss": 0.4198, + "step": 18806 + }, + { + "epoch": 0.41397754787585295, + "grad_norm": 0.7152179827169133, + "learning_rate": 6.606923073150302e-06, + "loss": 0.4895, + "step": 18807 + }, + { + "epoch": 0.41399955976227165, + "grad_norm": 1.1569300460564647, + "learning_rate": 6.606585523145266e-06, + "loss": 0.3309, + "step": 18808 + }, + { + "epoch": 0.4140215716486903, + "grad_norm": 2.788212415076201, + "learning_rate": 6.606247964974842e-06, + "loss": 0.2519, + "step": 18809 + }, + { + "epoch": 0.41404358353510895, + "grad_norm": 0.6464300096703582, + "learning_rate": 6.605910398640753e-06, + "loss": 0.4966, + "step": 18810 + }, + { + "epoch": 0.4140655954215276, + "grad_norm": 1.2660155782911515, + "learning_rate": 6.60557282414471e-06, + "loss": 0.3276, + "step": 18811 + }, + { + "epoch": 0.4140876073079463, + "grad_norm": 1.2123092665706343, + "learning_rate": 6.60523524148843e-06, + "loss": 0.2928, + "step": 18812 + }, + { + "epoch": 0.41410961919436495, + "grad_norm": 1.427096813953944, + "learning_rate": 6.604897650673628e-06, + "loss": 0.2914, + "step": 18813 + }, + { + "epoch": 0.4141316310807836, + "grad_norm": 2.8113228876933642, + "learning_rate": 6.60456005170202e-06, + "loss": 0.3451, + "step": 18814 + }, + { + "epoch": 0.4141536429672023, + "grad_norm": 1.4052275465287785, + "learning_rate": 6.604222444575324e-06, + "loss": 0.3465, + "step": 18815 + }, + { + "epoch": 0.41417565485362096, + "grad_norm": 1.1020442656751097, + "learning_rate": 6.6038848292952515e-06, + "loss": 0.3242, + "step": 18816 + }, + { + "epoch": 0.4141976667400396, + "grad_norm": 1.2686620649040679, + "learning_rate": 6.603547205863523e-06, + "loss": 0.3215, + "step": 18817 + }, + { + "epoch": 0.4142196786264583, + "grad_norm": 1.3310617575315429, + "learning_rate": 6.603209574281853e-06, + "loss": 0.3111, + "step": 18818 + }, + { + "epoch": 0.41424169051287696, + "grad_norm": 1.094087934849545, + "learning_rate": 6.602871934551953e-06, + "loss": 0.2326, + "step": 18819 + }, + { + "epoch": 0.4142637023992956, + "grad_norm": 1.3030260270770722, + "learning_rate": 6.602534286675547e-06, + "loss": 0.3313, + "step": 18820 + }, + { + "epoch": 0.4142857142857143, + "grad_norm": 1.696523920193426, + "learning_rate": 6.602196630654345e-06, + "loss": 0.2398, + "step": 18821 + }, + { + "epoch": 0.41430772617213296, + "grad_norm": 1.4862584839075872, + "learning_rate": 6.601858966490066e-06, + "loss": 0.3976, + "step": 18822 + }, + { + "epoch": 0.4143297380585516, + "grad_norm": 0.7045023024660916, + "learning_rate": 6.601521294184425e-06, + "loss": 0.4834, + "step": 18823 + }, + { + "epoch": 0.41435174994497026, + "grad_norm": 1.3862096566521613, + "learning_rate": 6.601183613739138e-06, + "loss": 0.3203, + "step": 18824 + }, + { + "epoch": 0.41437376183138896, + "grad_norm": 12.888427736416665, + "learning_rate": 6.6008459251559235e-06, + "loss": 0.2502, + "step": 18825 + }, + { + "epoch": 0.4143957737178076, + "grad_norm": 0.6597733170084664, + "learning_rate": 6.600508228436494e-06, + "loss": 0.4927, + "step": 18826 + }, + { + "epoch": 0.41441778560422626, + "grad_norm": 0.6253050328718937, + "learning_rate": 6.600170523582567e-06, + "loss": 0.496, + "step": 18827 + }, + { + "epoch": 0.41443979749064497, + "grad_norm": 1.311675449899406, + "learning_rate": 6.599832810595862e-06, + "loss": 0.319, + "step": 18828 + }, + { + "epoch": 0.4144618093770636, + "grad_norm": 1.776724936744446, + "learning_rate": 6.599495089478092e-06, + "loss": 0.354, + "step": 18829 + }, + { + "epoch": 0.41448382126348227, + "grad_norm": 1.188427913722133, + "learning_rate": 6.5991573602309755e-06, + "loss": 0.3189, + "step": 18830 + }, + { + "epoch": 0.41450583314990097, + "grad_norm": 1.2500446728108945, + "learning_rate": 6.598819622856227e-06, + "loss": 0.3352, + "step": 18831 + }, + { + "epoch": 0.4145278450363196, + "grad_norm": 1.5763285225593964, + "learning_rate": 6.598481877355564e-06, + "loss": 0.3702, + "step": 18832 + }, + { + "epoch": 0.41454985692273827, + "grad_norm": 0.7679649206178205, + "learning_rate": 6.598144123730705e-06, + "loss": 0.5388, + "step": 18833 + }, + { + "epoch": 0.4145718688091569, + "grad_norm": 1.4208586168691693, + "learning_rate": 6.597806361983362e-06, + "loss": 0.3248, + "step": 18834 + }, + { + "epoch": 0.4145938806955756, + "grad_norm": 3.5103696958839667, + "learning_rate": 6.597468592115257e-06, + "loss": 0.2488, + "step": 18835 + }, + { + "epoch": 0.41461589258199427, + "grad_norm": 2.2825645956196494, + "learning_rate": 6.597130814128103e-06, + "loss": 0.2542, + "step": 18836 + }, + { + "epoch": 0.4146379044684129, + "grad_norm": 1.3541085639435992, + "learning_rate": 6.596793028023616e-06, + "loss": 0.3843, + "step": 18837 + }, + { + "epoch": 0.4146599163548316, + "grad_norm": 1.563460989186489, + "learning_rate": 6.596455233803518e-06, + "loss": 0.3619, + "step": 18838 + }, + { + "epoch": 0.4146819282412503, + "grad_norm": 1.4373931286654817, + "learning_rate": 6.59611743146952e-06, + "loss": 0.4381, + "step": 18839 + }, + { + "epoch": 0.4147039401276689, + "grad_norm": 1.3777978816344743, + "learning_rate": 6.595779621023343e-06, + "loss": 0.2696, + "step": 18840 + }, + { + "epoch": 0.41472595201408763, + "grad_norm": 1.6134251161529463, + "learning_rate": 6.595441802466701e-06, + "loss": 0.3717, + "step": 18841 + }, + { + "epoch": 0.4147479639005063, + "grad_norm": 1.1293546065399986, + "learning_rate": 6.595103975801312e-06, + "loss": 0.2942, + "step": 18842 + }, + { + "epoch": 0.4147699757869249, + "grad_norm": 1.4700231180003136, + "learning_rate": 6.5947661410288955e-06, + "loss": 0.3422, + "step": 18843 + }, + { + "epoch": 0.41479198767334363, + "grad_norm": 0.6316190433641052, + "learning_rate": 6.594428298151164e-06, + "loss": 0.4935, + "step": 18844 + }, + { + "epoch": 0.4148139995597623, + "grad_norm": 1.305456191538176, + "learning_rate": 6.594090447169837e-06, + "loss": 0.2856, + "step": 18845 + }, + { + "epoch": 0.41483601144618093, + "grad_norm": 0.6257627945547324, + "learning_rate": 6.593752588086632e-06, + "loss": 0.5074, + "step": 18846 + }, + { + "epoch": 0.4148580233325996, + "grad_norm": 1.9043731027020725, + "learning_rate": 6.593414720903265e-06, + "loss": 0.3531, + "step": 18847 + }, + { + "epoch": 0.4148800352190183, + "grad_norm": 1.700665865286005, + "learning_rate": 6.593076845621453e-06, + "loss": 0.2798, + "step": 18848 + }, + { + "epoch": 0.41490204710543693, + "grad_norm": 1.4973079692026832, + "learning_rate": 6.5927389622429164e-06, + "loss": 0.2704, + "step": 18849 + }, + { + "epoch": 0.4149240589918556, + "grad_norm": 1.2912885633856248, + "learning_rate": 6.592401070769367e-06, + "loss": 0.2392, + "step": 18850 + }, + { + "epoch": 0.4149460708782743, + "grad_norm": 1.5765051972833402, + "learning_rate": 6.592063171202526e-06, + "loss": 0.3097, + "step": 18851 + }, + { + "epoch": 0.41496808276469294, + "grad_norm": 1.277745880339381, + "learning_rate": 6.591725263544111e-06, + "loss": 0.4346, + "step": 18852 + }, + { + "epoch": 0.4149900946511116, + "grad_norm": 1.2374798930258446, + "learning_rate": 6.5913873477958375e-06, + "loss": 0.3953, + "step": 18853 + }, + { + "epoch": 0.4150121065375303, + "grad_norm": 1.4246951938516241, + "learning_rate": 6.591049423959424e-06, + "loss": 0.2799, + "step": 18854 + }, + { + "epoch": 0.41503411842394894, + "grad_norm": 1.3900800116638539, + "learning_rate": 6.590711492036588e-06, + "loss": 0.3314, + "step": 18855 + }, + { + "epoch": 0.4150561303103676, + "grad_norm": 0.6296489404850119, + "learning_rate": 6.590373552029047e-06, + "loss": 0.4884, + "step": 18856 + }, + { + "epoch": 0.41507814219678624, + "grad_norm": 1.5820301758085629, + "learning_rate": 6.590035603938516e-06, + "loss": 0.3182, + "step": 18857 + }, + { + "epoch": 0.41510015408320494, + "grad_norm": 1.1470387802608661, + "learning_rate": 6.5896976477667175e-06, + "loss": 0.2721, + "step": 18858 + }, + { + "epoch": 0.4151221659696236, + "grad_norm": 1.368324641166345, + "learning_rate": 6.589359683515365e-06, + "loss": 0.3656, + "step": 18859 + }, + { + "epoch": 0.41514417785604224, + "grad_norm": 1.495228993553325, + "learning_rate": 6.5890217111861774e-06, + "loss": 0.3564, + "step": 18860 + }, + { + "epoch": 0.41516618974246094, + "grad_norm": 1.2698447674415163, + "learning_rate": 6.588683730780875e-06, + "loss": 0.2642, + "step": 18861 + }, + { + "epoch": 0.4151882016288796, + "grad_norm": 1.1135731160574296, + "learning_rate": 6.588345742301172e-06, + "loss": 0.299, + "step": 18862 + }, + { + "epoch": 0.41521021351529824, + "grad_norm": 1.2952094082691732, + "learning_rate": 6.588007745748787e-06, + "loss": 0.3772, + "step": 18863 + }, + { + "epoch": 0.41523222540171695, + "grad_norm": 0.6219732713898122, + "learning_rate": 6.587669741125439e-06, + "loss": 0.4963, + "step": 18864 + }, + { + "epoch": 0.4152542372881356, + "grad_norm": 1.426935076749408, + "learning_rate": 6.587331728432844e-06, + "loss": 0.337, + "step": 18865 + }, + { + "epoch": 0.41527624917455425, + "grad_norm": 0.6261157803537154, + "learning_rate": 6.586993707672723e-06, + "loss": 0.4853, + "step": 18866 + }, + { + "epoch": 0.41529826106097295, + "grad_norm": 1.4830126876852092, + "learning_rate": 6.586655678846792e-06, + "loss": 0.3919, + "step": 18867 + }, + { + "epoch": 0.4153202729473916, + "grad_norm": 0.5998850505829451, + "learning_rate": 6.586317641956768e-06, + "loss": 0.4754, + "step": 18868 + }, + { + "epoch": 0.41534228483381025, + "grad_norm": 1.3326645323935167, + "learning_rate": 6.58597959700437e-06, + "loss": 0.3333, + "step": 18869 + }, + { + "epoch": 0.4153642967202289, + "grad_norm": 1.665072570359928, + "learning_rate": 6.585641543991317e-06, + "loss": 0.4102, + "step": 18870 + }, + { + "epoch": 0.4153863086066476, + "grad_norm": 1.448366851201305, + "learning_rate": 6.585303482919327e-06, + "loss": 0.3196, + "step": 18871 + }, + { + "epoch": 0.41540832049306625, + "grad_norm": 0.608584174498156, + "learning_rate": 6.584965413790118e-06, + "loss": 0.497, + "step": 18872 + }, + { + "epoch": 0.4154303323794849, + "grad_norm": 1.4798989396852233, + "learning_rate": 6.584627336605406e-06, + "loss": 0.3089, + "step": 18873 + }, + { + "epoch": 0.4154523442659036, + "grad_norm": 0.6057746748625357, + "learning_rate": 6.584289251366913e-06, + "loss": 0.4803, + "step": 18874 + }, + { + "epoch": 0.41547435615232226, + "grad_norm": 1.7760913525303192, + "learning_rate": 6.583951158076354e-06, + "loss": 0.2674, + "step": 18875 + }, + { + "epoch": 0.4154963680387409, + "grad_norm": 1.6668952730809887, + "learning_rate": 6.58361305673545e-06, + "loss": 0.3797, + "step": 18876 + }, + { + "epoch": 0.4155183799251596, + "grad_norm": 1.7968568712412791, + "learning_rate": 6.583274947345916e-06, + "loss": 0.3454, + "step": 18877 + }, + { + "epoch": 0.41554039181157826, + "grad_norm": 1.9220060280301907, + "learning_rate": 6.582936829909474e-06, + "loss": 0.2964, + "step": 18878 + }, + { + "epoch": 0.4155624036979969, + "grad_norm": 1.8048310419314302, + "learning_rate": 6.582598704427843e-06, + "loss": 0.3123, + "step": 18879 + }, + { + "epoch": 0.4155844155844156, + "grad_norm": 0.6264123979925112, + "learning_rate": 6.582260570902737e-06, + "loss": 0.4888, + "step": 18880 + }, + { + "epoch": 0.41560642747083426, + "grad_norm": 0.6381890374276151, + "learning_rate": 6.581922429335876e-06, + "loss": 0.5025, + "step": 18881 + }, + { + "epoch": 0.4156284393572529, + "grad_norm": 1.4315253613478982, + "learning_rate": 6.581584279728982e-06, + "loss": 0.264, + "step": 18882 + }, + { + "epoch": 0.41565045124367156, + "grad_norm": 0.6118544575063123, + "learning_rate": 6.581246122083771e-06, + "loss": 0.5045, + "step": 18883 + }, + { + "epoch": 0.41567246313009026, + "grad_norm": 1.2691017445954917, + "learning_rate": 6.580907956401962e-06, + "loss": 0.2382, + "step": 18884 + }, + { + "epoch": 0.4156944750165089, + "grad_norm": 1.1692627638743363, + "learning_rate": 6.580569782685273e-06, + "loss": 0.3017, + "step": 18885 + }, + { + "epoch": 0.41571648690292756, + "grad_norm": 1.3178489933069653, + "learning_rate": 6.580231600935423e-06, + "loss": 0.2968, + "step": 18886 + }, + { + "epoch": 0.41573849878934627, + "grad_norm": 1.6637185906757905, + "learning_rate": 6.5798934111541314e-06, + "loss": 0.343, + "step": 18887 + }, + { + "epoch": 0.4157605106757649, + "grad_norm": 1.6663104281176762, + "learning_rate": 6.5795552133431175e-06, + "loss": 0.3444, + "step": 18888 + }, + { + "epoch": 0.41578252256218357, + "grad_norm": 1.529761347445037, + "learning_rate": 6.5792170075041e-06, + "loss": 0.3942, + "step": 18889 + }, + { + "epoch": 0.41580453444860227, + "grad_norm": 0.6461469178057057, + "learning_rate": 6.578878793638797e-06, + "loss": 0.5044, + "step": 18890 + }, + { + "epoch": 0.4158265463350209, + "grad_norm": 0.6379264638874569, + "learning_rate": 6.578540571748926e-06, + "loss": 0.4899, + "step": 18891 + }, + { + "epoch": 0.41584855822143957, + "grad_norm": 0.6165031341425726, + "learning_rate": 6.578202341836209e-06, + "loss": 0.4829, + "step": 18892 + }, + { + "epoch": 0.4158705701078582, + "grad_norm": 1.6308838939090247, + "learning_rate": 6.577864103902364e-06, + "loss": 0.3784, + "step": 18893 + }, + { + "epoch": 0.4158925819942769, + "grad_norm": 5.084768764019693, + "learning_rate": 6.57752585794911e-06, + "loss": 0.3167, + "step": 18894 + }, + { + "epoch": 0.41591459388069557, + "grad_norm": 0.6387826004985827, + "learning_rate": 6.577187603978166e-06, + "loss": 0.4977, + "step": 18895 + }, + { + "epoch": 0.4159366057671142, + "grad_norm": 1.9102903684561576, + "learning_rate": 6.576849341991249e-06, + "loss": 0.2984, + "step": 18896 + }, + { + "epoch": 0.4159586176535329, + "grad_norm": 0.6411612667991156, + "learning_rate": 6.576511071990084e-06, + "loss": 0.4925, + "step": 18897 + }, + { + "epoch": 0.4159806295399516, + "grad_norm": 1.4412599976747424, + "learning_rate": 6.576172793976386e-06, + "loss": 0.2833, + "step": 18898 + }, + { + "epoch": 0.4160026414263702, + "grad_norm": 1.5245018970874782, + "learning_rate": 6.575834507951872e-06, + "loss": 0.2488, + "step": 18899 + }, + { + "epoch": 0.41602465331278893, + "grad_norm": 0.6262113260330924, + "learning_rate": 6.5754962139182656e-06, + "loss": 0.4931, + "step": 18900 + }, + { + "epoch": 0.4160466651992076, + "grad_norm": 1.2085222630998291, + "learning_rate": 6.575157911877286e-06, + "loss": 0.3345, + "step": 18901 + }, + { + "epoch": 0.4160686770856262, + "grad_norm": 1.393228391522262, + "learning_rate": 6.57481960183065e-06, + "loss": 0.2694, + "step": 18902 + }, + { + "epoch": 0.41609068897204493, + "grad_norm": 1.442934127428915, + "learning_rate": 6.574481283780079e-06, + "loss": 0.3648, + "step": 18903 + }, + { + "epoch": 0.4161127008584636, + "grad_norm": 1.1618684314285406, + "learning_rate": 6.574142957727291e-06, + "loss": 0.3259, + "step": 18904 + }, + { + "epoch": 0.41613471274488223, + "grad_norm": 1.349613262981667, + "learning_rate": 6.573804623674007e-06, + "loss": 0.291, + "step": 18905 + }, + { + "epoch": 0.4161567246313009, + "grad_norm": 1.6432911099384822, + "learning_rate": 6.5734662816219464e-06, + "loss": 0.3615, + "step": 18906 + }, + { + "epoch": 0.4161787365177196, + "grad_norm": 1.5278529117162931, + "learning_rate": 6.573127931572828e-06, + "loss": 0.2859, + "step": 18907 + }, + { + "epoch": 0.41620074840413823, + "grad_norm": 1.4396999665689532, + "learning_rate": 6.572789573528372e-06, + "loss": 0.3204, + "step": 18908 + }, + { + "epoch": 0.4162227602905569, + "grad_norm": 3.2300330900544845, + "learning_rate": 6.5724512074902964e-06, + "loss": 0.2998, + "step": 18909 + }, + { + "epoch": 0.4162447721769756, + "grad_norm": 0.6495911616940268, + "learning_rate": 6.572112833460325e-06, + "loss": 0.4704, + "step": 18910 + }, + { + "epoch": 0.41626678406339424, + "grad_norm": 0.6266313865317383, + "learning_rate": 6.5717744514401735e-06, + "loss": 0.4932, + "step": 18911 + }, + { + "epoch": 0.4162887959498129, + "grad_norm": 1.1691594526752542, + "learning_rate": 6.571436061431563e-06, + "loss": 0.3403, + "step": 18912 + }, + { + "epoch": 0.4163108078362316, + "grad_norm": 1.505035077462779, + "learning_rate": 6.5710976634362145e-06, + "loss": 0.3284, + "step": 18913 + }, + { + "epoch": 0.41633281972265024, + "grad_norm": 1.5776606794705936, + "learning_rate": 6.570759257455846e-06, + "loss": 0.3048, + "step": 18914 + }, + { + "epoch": 0.4163548316090689, + "grad_norm": 1.4038397360205024, + "learning_rate": 6.5704208434921805e-06, + "loss": 0.3857, + "step": 18915 + }, + { + "epoch": 0.41637684349548754, + "grad_norm": 0.6219665673134204, + "learning_rate": 6.570082421546933e-06, + "loss": 0.4832, + "step": 18916 + }, + { + "epoch": 0.41639885538190624, + "grad_norm": 1.3748816559444117, + "learning_rate": 6.569743991621826e-06, + "loss": 0.3688, + "step": 18917 + }, + { + "epoch": 0.4164208672683249, + "grad_norm": 1.2711295346261617, + "learning_rate": 6.569405553718583e-06, + "loss": 0.2862, + "step": 18918 + }, + { + "epoch": 0.41644287915474354, + "grad_norm": 1.2755284602537897, + "learning_rate": 6.569067107838919e-06, + "loss": 0.3068, + "step": 18919 + }, + { + "epoch": 0.41646489104116224, + "grad_norm": 0.6824779420885652, + "learning_rate": 6.568728653984557e-06, + "loss": 0.5094, + "step": 18920 + }, + { + "epoch": 0.4164869029275809, + "grad_norm": 2.3435692654439535, + "learning_rate": 6.568390192157216e-06, + "loss": 0.3938, + "step": 18921 + }, + { + "epoch": 0.41650891481399954, + "grad_norm": 1.8416597151011014, + "learning_rate": 6.568051722358616e-06, + "loss": 0.3134, + "step": 18922 + }, + { + "epoch": 0.41653092670041825, + "grad_norm": 1.399092284532338, + "learning_rate": 6.567713244590478e-06, + "loss": 0.3944, + "step": 18923 + }, + { + "epoch": 0.4165529385868369, + "grad_norm": 1.229089968020003, + "learning_rate": 6.5673747588545225e-06, + "loss": 0.3121, + "step": 18924 + }, + { + "epoch": 0.41657495047325555, + "grad_norm": 1.5796264445631625, + "learning_rate": 6.567036265152469e-06, + "loss": 0.3428, + "step": 18925 + }, + { + "epoch": 0.41659696235967425, + "grad_norm": 0.615733455510559, + "learning_rate": 6.566697763486038e-06, + "loss": 0.477, + "step": 18926 + }, + { + "epoch": 0.4166189742460929, + "grad_norm": 1.4927333632446556, + "learning_rate": 6.566359253856951e-06, + "loss": 0.2264, + "step": 18927 + }, + { + "epoch": 0.41664098613251155, + "grad_norm": 1.2307139283936708, + "learning_rate": 6.566020736266927e-06, + "loss": 0.298, + "step": 18928 + }, + { + "epoch": 0.4166629980189302, + "grad_norm": 1.6148250577234669, + "learning_rate": 6.565682210717689e-06, + "loss": 0.3003, + "step": 18929 + }, + { + "epoch": 0.4166850099053489, + "grad_norm": 1.1280458148449253, + "learning_rate": 6.565343677210954e-06, + "loss": 0.2753, + "step": 18930 + }, + { + "epoch": 0.41670702179176755, + "grad_norm": 1.5192984095986857, + "learning_rate": 6.565005135748444e-06, + "loss": 0.3056, + "step": 18931 + }, + { + "epoch": 0.4167290336781862, + "grad_norm": 2.1251775238756174, + "learning_rate": 6.564666586331879e-06, + "loss": 0.3609, + "step": 18932 + }, + { + "epoch": 0.4167510455646049, + "grad_norm": 1.3071024192766434, + "learning_rate": 6.564328028962983e-06, + "loss": 0.3422, + "step": 18933 + }, + { + "epoch": 0.41677305745102355, + "grad_norm": 1.495472263905854, + "learning_rate": 6.563989463643473e-06, + "loss": 0.2897, + "step": 18934 + }, + { + "epoch": 0.4167950693374422, + "grad_norm": 8.58197076339144, + "learning_rate": 6.56365089037507e-06, + "loss": 0.3167, + "step": 18935 + }, + { + "epoch": 0.4168170812238609, + "grad_norm": 1.1426042343962053, + "learning_rate": 6.563312309159497e-06, + "loss": 0.2017, + "step": 18936 + }, + { + "epoch": 0.41683909311027956, + "grad_norm": 1.4385733546642274, + "learning_rate": 6.5629737199984746e-06, + "loss": 0.3252, + "step": 18937 + }, + { + "epoch": 0.4168611049966982, + "grad_norm": 1.3917874956546605, + "learning_rate": 6.562635122893721e-06, + "loss": 0.3031, + "step": 18938 + }, + { + "epoch": 0.41688311688311686, + "grad_norm": 2.0304299173641804, + "learning_rate": 6.56229651784696e-06, + "loss": 0.2917, + "step": 18939 + }, + { + "epoch": 0.41690512876953556, + "grad_norm": 1.4342067272834145, + "learning_rate": 6.561957904859909e-06, + "loss": 0.3126, + "step": 18940 + }, + { + "epoch": 0.4169271406559542, + "grad_norm": 1.4320498164429176, + "learning_rate": 6.561619283934293e-06, + "loss": 0.4155, + "step": 18941 + }, + { + "epoch": 0.41694915254237286, + "grad_norm": 3.2523459898409808, + "learning_rate": 6.561280655071831e-06, + "loss": 0.2674, + "step": 18942 + }, + { + "epoch": 0.41697116442879156, + "grad_norm": 1.940857092248009, + "learning_rate": 6.5609420182742454e-06, + "loss": 0.36, + "step": 18943 + }, + { + "epoch": 0.4169931763152102, + "grad_norm": 0.6624993580142776, + "learning_rate": 6.560603373543255e-06, + "loss": 0.5456, + "step": 18944 + }, + { + "epoch": 0.41701518820162886, + "grad_norm": 2.5383828566306894, + "learning_rate": 6.560264720880583e-06, + "loss": 0.2518, + "step": 18945 + }, + { + "epoch": 0.41703720008804757, + "grad_norm": 1.6655627578306413, + "learning_rate": 6.559926060287949e-06, + "loss": 0.3776, + "step": 18946 + }, + { + "epoch": 0.4170592119744662, + "grad_norm": 1.4691641757476426, + "learning_rate": 6.559587391767076e-06, + "loss": 0.372, + "step": 18947 + }, + { + "epoch": 0.41708122386088486, + "grad_norm": 1.7735228022856688, + "learning_rate": 6.5592487153196825e-06, + "loss": 0.3794, + "step": 18948 + }, + { + "epoch": 0.41710323574730357, + "grad_norm": 1.4353391687481574, + "learning_rate": 6.558910030947493e-06, + "loss": 0.3096, + "step": 18949 + }, + { + "epoch": 0.4171252476337222, + "grad_norm": 1.3010177751973688, + "learning_rate": 6.5585713386522266e-06, + "loss": 0.3057, + "step": 18950 + }, + { + "epoch": 0.41714725952014087, + "grad_norm": 0.6244107199081775, + "learning_rate": 6.558232638435606e-06, + "loss": 0.4902, + "step": 18951 + }, + { + "epoch": 0.4171692714065595, + "grad_norm": 1.6990040163423494, + "learning_rate": 6.557893930299353e-06, + "loss": 0.3117, + "step": 18952 + }, + { + "epoch": 0.4171912832929782, + "grad_norm": 0.601577782702859, + "learning_rate": 6.557555214245187e-06, + "loss": 0.4797, + "step": 18953 + }, + { + "epoch": 0.41721329517939687, + "grad_norm": 1.8922410464178063, + "learning_rate": 6.557216490274832e-06, + "loss": 0.3607, + "step": 18954 + }, + { + "epoch": 0.4172353070658155, + "grad_norm": 1.7471120992499938, + "learning_rate": 6.556877758390007e-06, + "loss": 0.3161, + "step": 18955 + }, + { + "epoch": 0.4172573189522342, + "grad_norm": 0.6119046962718488, + "learning_rate": 6.556539018592435e-06, + "loss": 0.5036, + "step": 18956 + }, + { + "epoch": 0.4172793308386529, + "grad_norm": 2.7435869413109795, + "learning_rate": 6.556200270883838e-06, + "loss": 0.2838, + "step": 18957 + }, + { + "epoch": 0.4173013427250715, + "grad_norm": 1.4567252658541536, + "learning_rate": 6.555861515265936e-06, + "loss": 0.3851, + "step": 18958 + }, + { + "epoch": 0.4173233546114902, + "grad_norm": 4.93189612372247, + "learning_rate": 6.555522751740453e-06, + "loss": 0.3547, + "step": 18959 + }, + { + "epoch": 0.4173453664979089, + "grad_norm": 0.6274810815310146, + "learning_rate": 6.555183980309109e-06, + "loss": 0.4975, + "step": 18960 + }, + { + "epoch": 0.4173673783843275, + "grad_norm": 2.608432227845255, + "learning_rate": 6.554845200973627e-06, + "loss": 0.3319, + "step": 18961 + }, + { + "epoch": 0.41738939027074623, + "grad_norm": 1.8844983151197248, + "learning_rate": 6.5545064137357285e-06, + "loss": 0.3287, + "step": 18962 + }, + { + "epoch": 0.4174114021571649, + "grad_norm": 0.6401043781300132, + "learning_rate": 6.554167618597134e-06, + "loss": 0.5044, + "step": 18963 + }, + { + "epoch": 0.41743341404358353, + "grad_norm": 0.6255947966075394, + "learning_rate": 6.553828815559567e-06, + "loss": 0.4999, + "step": 18964 + }, + { + "epoch": 0.4174554259300022, + "grad_norm": 0.6229174103267822, + "learning_rate": 6.553490004624749e-06, + "loss": 0.5084, + "step": 18965 + }, + { + "epoch": 0.4174774378164209, + "grad_norm": 2.0565594313828295, + "learning_rate": 6.5531511857944005e-06, + "loss": 0.3133, + "step": 18966 + }, + { + "epoch": 0.41749944970283953, + "grad_norm": 1.6938140760847464, + "learning_rate": 6.552812359070247e-06, + "loss": 0.412, + "step": 18967 + }, + { + "epoch": 0.4175214615892582, + "grad_norm": 1.4550722541181915, + "learning_rate": 6.5524735244540084e-06, + "loss": 0.2973, + "step": 18968 + }, + { + "epoch": 0.4175434734756769, + "grad_norm": 1.574730860751235, + "learning_rate": 6.552134681947405e-06, + "loss": 0.3932, + "step": 18969 + }, + { + "epoch": 0.41756548536209553, + "grad_norm": 1.662206165376426, + "learning_rate": 6.551795831552163e-06, + "loss": 0.333, + "step": 18970 + }, + { + "epoch": 0.4175874972485142, + "grad_norm": 1.4369891420082794, + "learning_rate": 6.551456973270001e-06, + "loss": 0.3835, + "step": 18971 + }, + { + "epoch": 0.4176095091349329, + "grad_norm": 1.2461702216305357, + "learning_rate": 6.551118107102643e-06, + "loss": 0.3576, + "step": 18972 + }, + { + "epoch": 0.41763152102135154, + "grad_norm": 1.8341962450313882, + "learning_rate": 6.550779233051812e-06, + "loss": 0.3731, + "step": 18973 + }, + { + "epoch": 0.4176535329077702, + "grad_norm": 3.58258281094641, + "learning_rate": 6.550440351119228e-06, + "loss": 0.366, + "step": 18974 + }, + { + "epoch": 0.41767554479418884, + "grad_norm": 1.563290662636825, + "learning_rate": 6.550101461306616e-06, + "loss": 0.3194, + "step": 18975 + }, + { + "epoch": 0.41769755668060754, + "grad_norm": 2.037436549464129, + "learning_rate": 6.549762563615695e-06, + "loss": 0.2541, + "step": 18976 + }, + { + "epoch": 0.4177195685670262, + "grad_norm": 1.4174046391233237, + "learning_rate": 6.549423658048191e-06, + "loss": 0.3371, + "step": 18977 + }, + { + "epoch": 0.41774158045344484, + "grad_norm": 1.2240987584957657, + "learning_rate": 6.5490847446058245e-06, + "loss": 0.3495, + "step": 18978 + }, + { + "epoch": 0.41776359233986354, + "grad_norm": 1.3883316276396536, + "learning_rate": 6.548745823290319e-06, + "loss": 0.3095, + "step": 18979 + }, + { + "epoch": 0.4177856042262822, + "grad_norm": 1.9608662339468408, + "learning_rate": 6.548406894103396e-06, + "loss": 0.2792, + "step": 18980 + }, + { + "epoch": 0.41780761611270084, + "grad_norm": 0.6902886154934998, + "learning_rate": 6.548067957046777e-06, + "loss": 0.4894, + "step": 18981 + }, + { + "epoch": 0.41782962799911955, + "grad_norm": 2.5953041228019944, + "learning_rate": 6.5477290121221885e-06, + "loss": 0.3085, + "step": 18982 + }, + { + "epoch": 0.4178516398855382, + "grad_norm": 1.2775284530388185, + "learning_rate": 6.5473900593313515e-06, + "loss": 0.3709, + "step": 18983 + }, + { + "epoch": 0.41787365177195684, + "grad_norm": 1.406281397617474, + "learning_rate": 6.547051098675985e-06, + "loss": 0.3764, + "step": 18984 + }, + { + "epoch": 0.41789566365837555, + "grad_norm": 1.2274363988352852, + "learning_rate": 6.546712130157816e-06, + "loss": 0.3086, + "step": 18985 + }, + { + "epoch": 0.4179176755447942, + "grad_norm": 1.5919581905180464, + "learning_rate": 6.5463731537785665e-06, + "loss": 0.3692, + "step": 18986 + }, + { + "epoch": 0.41793968743121285, + "grad_norm": 1.3298504717243391, + "learning_rate": 6.546034169539958e-06, + "loss": 0.3109, + "step": 18987 + }, + { + "epoch": 0.4179616993176315, + "grad_norm": 0.6301517522359927, + "learning_rate": 6.545695177443717e-06, + "loss": 0.4991, + "step": 18988 + }, + { + "epoch": 0.4179837112040502, + "grad_norm": 1.5353232459035382, + "learning_rate": 6.545356177491561e-06, + "loss": 0.3265, + "step": 18989 + }, + { + "epoch": 0.41800572309046885, + "grad_norm": 1.3235212980247948, + "learning_rate": 6.545017169685217e-06, + "loss": 0.2426, + "step": 18990 + }, + { + "epoch": 0.4180277349768875, + "grad_norm": 1.7897864437059177, + "learning_rate": 6.544678154026406e-06, + "loss": 0.3002, + "step": 18991 + }, + { + "epoch": 0.4180497468633062, + "grad_norm": 1.8443886670438732, + "learning_rate": 6.5443391305168515e-06, + "loss": 0.2929, + "step": 18992 + }, + { + "epoch": 0.41807175874972485, + "grad_norm": 2.2506473609678683, + "learning_rate": 6.544000099158277e-06, + "loss": 0.357, + "step": 18993 + }, + { + "epoch": 0.4180937706361435, + "grad_norm": 1.5010754437831015, + "learning_rate": 6.5436610599524055e-06, + "loss": 0.2747, + "step": 18994 + }, + { + "epoch": 0.4181157825225622, + "grad_norm": 1.1046715981277138, + "learning_rate": 6.54332201290096e-06, + "loss": 0.2793, + "step": 18995 + }, + { + "epoch": 0.41813779440898086, + "grad_norm": 1.5777791497913134, + "learning_rate": 6.542982958005663e-06, + "loss": 0.3897, + "step": 18996 + }, + { + "epoch": 0.4181598062953995, + "grad_norm": 0.6208720404098902, + "learning_rate": 6.5426438952682404e-06, + "loss": 0.4752, + "step": 18997 + }, + { + "epoch": 0.41818181818181815, + "grad_norm": 1.2457360580641088, + "learning_rate": 6.542304824690412e-06, + "loss": 0.3352, + "step": 18998 + }, + { + "epoch": 0.41820383006823686, + "grad_norm": 1.380617364624145, + "learning_rate": 6.541965746273902e-06, + "loss": 0.4234, + "step": 18999 + }, + { + "epoch": 0.4182258419546555, + "grad_norm": 0.6323863467712598, + "learning_rate": 6.541626660020435e-06, + "loss": 0.4917, + "step": 19000 + }, + { + "epoch": 0.41824785384107416, + "grad_norm": 0.6259719113900276, + "learning_rate": 6.541287565931735e-06, + "loss": 0.508, + "step": 19001 + }, + { + "epoch": 0.41826986572749286, + "grad_norm": 1.9549614519294494, + "learning_rate": 6.540948464009522e-06, + "loss": 0.3049, + "step": 19002 + }, + { + "epoch": 0.4182918776139115, + "grad_norm": 3.3117333240021876, + "learning_rate": 6.540609354255522e-06, + "loss": 0.321, + "step": 19003 + }, + { + "epoch": 0.41831388950033016, + "grad_norm": 1.41448259138112, + "learning_rate": 6.54027023667146e-06, + "loss": 0.3111, + "step": 19004 + }, + { + "epoch": 0.41833590138674887, + "grad_norm": 1.7757200909225672, + "learning_rate": 6.5399311112590554e-06, + "loss": 0.2589, + "step": 19005 + }, + { + "epoch": 0.4183579132731675, + "grad_norm": 1.2575468047387879, + "learning_rate": 6.539591978020037e-06, + "loss": 0.257, + "step": 19006 + }, + { + "epoch": 0.41837992515958616, + "grad_norm": 1.7841297231037268, + "learning_rate": 6.539252836956123e-06, + "loss": 0.3195, + "step": 19007 + }, + { + "epoch": 0.41840193704600487, + "grad_norm": 1.4309625291189871, + "learning_rate": 6.53891368806904e-06, + "loss": 0.3596, + "step": 19008 + }, + { + "epoch": 0.4184239489324235, + "grad_norm": 1.4126485965330944, + "learning_rate": 6.538574531360511e-06, + "loss": 0.3319, + "step": 19009 + }, + { + "epoch": 0.41844596081884217, + "grad_norm": 1.1132045876731038, + "learning_rate": 6.53823536683226e-06, + "loss": 0.2052, + "step": 19010 + }, + { + "epoch": 0.4184679727052608, + "grad_norm": 1.2290909686849758, + "learning_rate": 6.537896194486012e-06, + "loss": 0.2844, + "step": 19011 + }, + { + "epoch": 0.4184899845916795, + "grad_norm": 0.6601700978423201, + "learning_rate": 6.537557014323488e-06, + "loss": 0.4887, + "step": 19012 + }, + { + "epoch": 0.41851199647809817, + "grad_norm": 1.2979732651404297, + "learning_rate": 6.5372178263464135e-06, + "loss": 0.4301, + "step": 19013 + }, + { + "epoch": 0.4185340083645168, + "grad_norm": 1.3058007001573744, + "learning_rate": 6.536878630556513e-06, + "loss": 0.3722, + "step": 19014 + }, + { + "epoch": 0.4185560202509355, + "grad_norm": 1.2405517524113874, + "learning_rate": 6.536539426955511e-06, + "loss": 0.2928, + "step": 19015 + }, + { + "epoch": 0.4185780321373542, + "grad_norm": 0.6382944371970347, + "learning_rate": 6.53620021554513e-06, + "loss": 0.4808, + "step": 19016 + }, + { + "epoch": 0.4186000440237728, + "grad_norm": 2.27613798911909, + "learning_rate": 6.535860996327093e-06, + "loss": 0.3195, + "step": 19017 + }, + { + "epoch": 0.4186220559101915, + "grad_norm": 2.775230168497618, + "learning_rate": 6.535521769303126e-06, + "loss": 0.2817, + "step": 19018 + }, + { + "epoch": 0.4186440677966102, + "grad_norm": 4.337895986337156, + "learning_rate": 6.5351825344749535e-06, + "loss": 0.3223, + "step": 19019 + }, + { + "epoch": 0.4186660796830288, + "grad_norm": 4.763416462520158, + "learning_rate": 6.534843291844297e-06, + "loss": 0.2547, + "step": 19020 + }, + { + "epoch": 0.41868809156944753, + "grad_norm": 1.6477002574446484, + "learning_rate": 6.534504041412883e-06, + "loss": 0.3096, + "step": 19021 + }, + { + "epoch": 0.4187101034558662, + "grad_norm": 1.575549326563204, + "learning_rate": 6.534164783182435e-06, + "loss": 0.343, + "step": 19022 + }, + { + "epoch": 0.4187321153422848, + "grad_norm": 1.6516335497252275, + "learning_rate": 6.533825517154677e-06, + "loss": 0.3374, + "step": 19023 + }, + { + "epoch": 0.4187541272287035, + "grad_norm": 0.6279101403447904, + "learning_rate": 6.533486243331335e-06, + "loss": 0.4812, + "step": 19024 + }, + { + "epoch": 0.4187761391151222, + "grad_norm": 0.6490810360765029, + "learning_rate": 6.533146961714131e-06, + "loss": 0.5066, + "step": 19025 + }, + { + "epoch": 0.41879815100154083, + "grad_norm": 2.327652185611296, + "learning_rate": 6.53280767230479e-06, + "loss": 0.2832, + "step": 19026 + }, + { + "epoch": 0.4188201628879595, + "grad_norm": 1.5493523179024338, + "learning_rate": 6.532468375105037e-06, + "loss": 0.3347, + "step": 19027 + }, + { + "epoch": 0.4188421747743782, + "grad_norm": 1.2567048999044035, + "learning_rate": 6.532129070116595e-06, + "loss": 0.3361, + "step": 19028 + }, + { + "epoch": 0.41886418666079683, + "grad_norm": 1.130508163825911, + "learning_rate": 6.531789757341192e-06, + "loss": 0.3355, + "step": 19029 + }, + { + "epoch": 0.4188861985472155, + "grad_norm": 1.409335559570706, + "learning_rate": 6.531450436780549e-06, + "loss": 0.3, + "step": 19030 + }, + { + "epoch": 0.4189082104336342, + "grad_norm": 3.9087931178984987, + "learning_rate": 6.531111108436392e-06, + "loss": 0.2252, + "step": 19031 + }, + { + "epoch": 0.41893022232005284, + "grad_norm": 1.3504204444645036, + "learning_rate": 6.530771772310447e-06, + "loss": 0.3539, + "step": 19032 + }, + { + "epoch": 0.4189522342064715, + "grad_norm": 1.34414757655374, + "learning_rate": 6.530432428404435e-06, + "loss": 0.3797, + "step": 19033 + }, + { + "epoch": 0.41897424609289013, + "grad_norm": 1.1265364801270559, + "learning_rate": 6.530093076720083e-06, + "loss": 0.2899, + "step": 19034 + }, + { + "epoch": 0.41899625797930884, + "grad_norm": 3.884795184792224, + "learning_rate": 6.529753717259116e-06, + "loss": 0.3583, + "step": 19035 + }, + { + "epoch": 0.4190182698657275, + "grad_norm": 1.566826321371347, + "learning_rate": 6.529414350023259e-06, + "loss": 0.2711, + "step": 19036 + }, + { + "epoch": 0.41904028175214614, + "grad_norm": 1.1581036974571206, + "learning_rate": 6.5290749750142355e-06, + "loss": 0.3051, + "step": 19037 + }, + { + "epoch": 0.41906229363856484, + "grad_norm": 1.2003063348739718, + "learning_rate": 6.52873559223377e-06, + "loss": 0.2536, + "step": 19038 + }, + { + "epoch": 0.4190843055249835, + "grad_norm": 0.6720720293333275, + "learning_rate": 6.52839620168359e-06, + "loss": 0.4806, + "step": 19039 + }, + { + "epoch": 0.41910631741140214, + "grad_norm": 1.4908813084854267, + "learning_rate": 6.528056803365419e-06, + "loss": 0.3348, + "step": 19040 + }, + { + "epoch": 0.41912832929782085, + "grad_norm": 1.3499622428438278, + "learning_rate": 6.527717397280979e-06, + "loss": 0.2817, + "step": 19041 + }, + { + "epoch": 0.4191503411842395, + "grad_norm": 1.3425059084050655, + "learning_rate": 6.527377983432e-06, + "loss": 0.3237, + "step": 19042 + }, + { + "epoch": 0.41917235307065814, + "grad_norm": 2.2015855272672735, + "learning_rate": 6.527038561820206e-06, + "loss": 0.3511, + "step": 19043 + }, + { + "epoch": 0.41919436495707685, + "grad_norm": 0.6363571412620322, + "learning_rate": 6.526699132447319e-06, + "loss": 0.5025, + "step": 19044 + }, + { + "epoch": 0.4192163768434955, + "grad_norm": 1.2049744648114789, + "learning_rate": 6.526359695315067e-06, + "loss": 0.299, + "step": 19045 + }, + { + "epoch": 0.41923838872991415, + "grad_norm": 1.2041311699570805, + "learning_rate": 6.5260202504251735e-06, + "loss": 0.277, + "step": 19046 + }, + { + "epoch": 0.4192604006163328, + "grad_norm": 1.2129663613076551, + "learning_rate": 6.525680797779366e-06, + "loss": 0.2889, + "step": 19047 + }, + { + "epoch": 0.4192824125027515, + "grad_norm": 2.234007339385015, + "learning_rate": 6.525341337379366e-06, + "loss": 0.3372, + "step": 19048 + }, + { + "epoch": 0.41930442438917015, + "grad_norm": 0.6312978963708337, + "learning_rate": 6.5250018692269024e-06, + "loss": 0.4976, + "step": 19049 + }, + { + "epoch": 0.4193264362755888, + "grad_norm": 1.4672071731800316, + "learning_rate": 6.524662393323699e-06, + "loss": 0.3075, + "step": 19050 + }, + { + "epoch": 0.4193484481620075, + "grad_norm": 1.4716803099156306, + "learning_rate": 6.52432290967148e-06, + "loss": 0.2716, + "step": 19051 + }, + { + "epoch": 0.41937046004842615, + "grad_norm": 0.6003096756202745, + "learning_rate": 6.523983418271975e-06, + "loss": 0.4787, + "step": 19052 + }, + { + "epoch": 0.4193924719348448, + "grad_norm": 1.6653039129564708, + "learning_rate": 6.523643919126903e-06, + "loss": 0.3717, + "step": 19053 + }, + { + "epoch": 0.4194144838212635, + "grad_norm": 1.2721251728515401, + "learning_rate": 6.523304412237996e-06, + "loss": 0.4136, + "step": 19054 + }, + { + "epoch": 0.41943649570768216, + "grad_norm": 3.611367311048468, + "learning_rate": 6.522964897606975e-06, + "loss": 0.3623, + "step": 19055 + }, + { + "epoch": 0.4194585075941008, + "grad_norm": 0.6192918147980785, + "learning_rate": 6.522625375235567e-06, + "loss": 0.4799, + "step": 19056 + }, + { + "epoch": 0.41948051948051945, + "grad_norm": 1.896878614098917, + "learning_rate": 6.5222858451255e-06, + "loss": 0.3235, + "step": 19057 + }, + { + "epoch": 0.41950253136693816, + "grad_norm": 1.3548399497140993, + "learning_rate": 6.521946307278495e-06, + "loss": 0.3025, + "step": 19058 + }, + { + "epoch": 0.4195245432533568, + "grad_norm": 1.6855300205246029, + "learning_rate": 6.5216067616962795e-06, + "loss": 0.3739, + "step": 19059 + }, + { + "epoch": 0.41954655513977546, + "grad_norm": 1.2723296759054437, + "learning_rate": 6.521267208380581e-06, + "loss": 0.2935, + "step": 19060 + }, + { + "epoch": 0.41956856702619416, + "grad_norm": 1.6575251490583938, + "learning_rate": 6.520927647333125e-06, + "loss": 0.3309, + "step": 19061 + }, + { + "epoch": 0.4195905789126128, + "grad_norm": 1.751622936017623, + "learning_rate": 6.520588078555635e-06, + "loss": 0.383, + "step": 19062 + }, + { + "epoch": 0.41961259079903146, + "grad_norm": 1.635823698427795, + "learning_rate": 6.520248502049839e-06, + "loss": 0.4408, + "step": 19063 + }, + { + "epoch": 0.41963460268545016, + "grad_norm": 1.2403407868166403, + "learning_rate": 6.519908917817461e-06, + "loss": 0.2944, + "step": 19064 + }, + { + "epoch": 0.4196566145718688, + "grad_norm": 0.6287825666290023, + "learning_rate": 6.5195693258602295e-06, + "loss": 0.4544, + "step": 19065 + }, + { + "epoch": 0.41967862645828746, + "grad_norm": 1.5659893501837918, + "learning_rate": 6.519229726179867e-06, + "loss": 0.3708, + "step": 19066 + }, + { + "epoch": 0.41970063834470617, + "grad_norm": 1.551000785145412, + "learning_rate": 6.518890118778103e-06, + "loss": 0.3118, + "step": 19067 + }, + { + "epoch": 0.4197226502311248, + "grad_norm": 4.563573212998824, + "learning_rate": 6.518550503656662e-06, + "loss": 0.3209, + "step": 19068 + }, + { + "epoch": 0.41974466211754347, + "grad_norm": 0.5911386194920348, + "learning_rate": 6.518210880817268e-06, + "loss": 0.4533, + "step": 19069 + }, + { + "epoch": 0.4197666740039621, + "grad_norm": 0.6337976328983803, + "learning_rate": 6.5178712502616514e-06, + "loss": 0.5011, + "step": 19070 + }, + { + "epoch": 0.4197886858903808, + "grad_norm": 0.6103207741611759, + "learning_rate": 6.5175316119915345e-06, + "loss": 0.4913, + "step": 19071 + }, + { + "epoch": 0.41981069777679947, + "grad_norm": 1.7188087937859247, + "learning_rate": 6.517191966008647e-06, + "loss": 0.3513, + "step": 19072 + }, + { + "epoch": 0.4198327096632181, + "grad_norm": 0.6355950181222, + "learning_rate": 6.516852312314712e-06, + "loss": 0.505, + "step": 19073 + }, + { + "epoch": 0.4198547215496368, + "grad_norm": 1.3774090385199365, + "learning_rate": 6.5165126509114576e-06, + "loss": 0.3623, + "step": 19074 + }, + { + "epoch": 0.41987673343605547, + "grad_norm": 1.2740342377694744, + "learning_rate": 6.516172981800609e-06, + "loss": 0.3369, + "step": 19075 + }, + { + "epoch": 0.4198987453224741, + "grad_norm": 2.1029008010945804, + "learning_rate": 6.515833304983894e-06, + "loss": 0.2496, + "step": 19076 + }, + { + "epoch": 0.4199207572088928, + "grad_norm": 1.6623930419354573, + "learning_rate": 6.5154936204630355e-06, + "loss": 0.2614, + "step": 19077 + }, + { + "epoch": 0.4199427690953115, + "grad_norm": 5.34441715549797, + "learning_rate": 6.515153928239765e-06, + "loss": 0.3613, + "step": 19078 + }, + { + "epoch": 0.4199647809817301, + "grad_norm": 1.6631861807334234, + "learning_rate": 6.5148142283158055e-06, + "loss": 0.2512, + "step": 19079 + }, + { + "epoch": 0.4199867928681488, + "grad_norm": 1.7701488441211617, + "learning_rate": 6.514474520692885e-06, + "loss": 0.3514, + "step": 19080 + }, + { + "epoch": 0.4200088047545675, + "grad_norm": 4.081705692114711, + "learning_rate": 6.514134805372729e-06, + "loss": 0.3313, + "step": 19081 + }, + { + "epoch": 0.4200308166409861, + "grad_norm": 1.473157430857103, + "learning_rate": 6.513795082357065e-06, + "loss": 0.2566, + "step": 19082 + }, + { + "epoch": 0.4200528285274048, + "grad_norm": 1.4990413029291747, + "learning_rate": 6.513455351647619e-06, + "loss": 0.4079, + "step": 19083 + }, + { + "epoch": 0.4200748404138235, + "grad_norm": 1.3671106722643205, + "learning_rate": 6.513115613246118e-06, + "loss": 0.2687, + "step": 19084 + }, + { + "epoch": 0.42009685230024213, + "grad_norm": 1.3494839428608116, + "learning_rate": 6.512775867154288e-06, + "loss": 0.3857, + "step": 19085 + }, + { + "epoch": 0.4201188641866608, + "grad_norm": 1.220266101598627, + "learning_rate": 6.512436113373858e-06, + "loss": 0.2019, + "step": 19086 + }, + { + "epoch": 0.4201408760730795, + "grad_norm": 1.3442743696859787, + "learning_rate": 6.5120963519065504e-06, + "loss": 0.3738, + "step": 19087 + }, + { + "epoch": 0.42016288795949813, + "grad_norm": 1.134437081100506, + "learning_rate": 6.5117565827540985e-06, + "loss": 0.2933, + "step": 19088 + }, + { + "epoch": 0.4201848998459168, + "grad_norm": 1.2727896113452752, + "learning_rate": 6.511416805918222e-06, + "loss": 0.3122, + "step": 19089 + }, + { + "epoch": 0.4202069117323355, + "grad_norm": 1.337821900564778, + "learning_rate": 6.511077021400652e-06, + "loss": 0.2188, + "step": 19090 + }, + { + "epoch": 0.42022892361875414, + "grad_norm": 1.7678291096809875, + "learning_rate": 6.510737229203115e-06, + "loss": 0.3129, + "step": 19091 + }, + { + "epoch": 0.4202509355051728, + "grad_norm": 1.4484173503978317, + "learning_rate": 6.510397429327338e-06, + "loss": 0.3147, + "step": 19092 + }, + { + "epoch": 0.42027294739159143, + "grad_norm": 0.6359571588250287, + "learning_rate": 6.510057621775048e-06, + "loss": 0.5004, + "step": 19093 + }, + { + "epoch": 0.42029495927801014, + "grad_norm": 1.5728367849597356, + "learning_rate": 6.50971780654797e-06, + "loss": 0.3644, + "step": 19094 + }, + { + "epoch": 0.4203169711644288, + "grad_norm": 1.9547065822157639, + "learning_rate": 6.509377983647833e-06, + "loss": 0.3125, + "step": 19095 + }, + { + "epoch": 0.42033898305084744, + "grad_norm": 1.5305648955733036, + "learning_rate": 6.509038153076365e-06, + "loss": 0.2805, + "step": 19096 + }, + { + "epoch": 0.42036099493726614, + "grad_norm": 2.050181863375347, + "learning_rate": 6.508698314835291e-06, + "loss": 0.33, + "step": 19097 + }, + { + "epoch": 0.4203830068236848, + "grad_norm": 1.2514455840976857, + "learning_rate": 6.5083584689263406e-06, + "loss": 0.2727, + "step": 19098 + }, + { + "epoch": 0.42040501871010344, + "grad_norm": 1.9534998295513264, + "learning_rate": 6.50801861535124e-06, + "loss": 0.3339, + "step": 19099 + }, + { + "epoch": 0.42042703059652214, + "grad_norm": 1.3486863339188204, + "learning_rate": 6.507678754111713e-06, + "loss": 0.3197, + "step": 19100 + }, + { + "epoch": 0.4204490424829408, + "grad_norm": 1.5320660175371532, + "learning_rate": 6.507338885209492e-06, + "loss": 0.2689, + "step": 19101 + }, + { + "epoch": 0.42047105436935944, + "grad_norm": 1.2216638476417432, + "learning_rate": 6.506999008646303e-06, + "loss": 0.2817, + "step": 19102 + }, + { + "epoch": 0.42049306625577815, + "grad_norm": 1.624859216757084, + "learning_rate": 6.506659124423873e-06, + "loss": 0.3511, + "step": 19103 + }, + { + "epoch": 0.4205150781421968, + "grad_norm": 1.5157526247218704, + "learning_rate": 6.506319232543929e-06, + "loss": 0.3099, + "step": 19104 + }, + { + "epoch": 0.42053709002861545, + "grad_norm": 1.6781422278610123, + "learning_rate": 6.505979333008198e-06, + "loss": 0.3078, + "step": 19105 + }, + { + "epoch": 0.4205591019150341, + "grad_norm": 1.4876837295581073, + "learning_rate": 6.505639425818411e-06, + "loss": 0.3814, + "step": 19106 + }, + { + "epoch": 0.4205811138014528, + "grad_norm": 1.4185570828986627, + "learning_rate": 6.50529951097629e-06, + "loss": 0.2945, + "step": 19107 + }, + { + "epoch": 0.42060312568787145, + "grad_norm": 1.3641795248075559, + "learning_rate": 6.504959588483566e-06, + "loss": 0.2871, + "step": 19108 + }, + { + "epoch": 0.4206251375742901, + "grad_norm": 1.6751169918532376, + "learning_rate": 6.504619658341966e-06, + "loss": 0.3218, + "step": 19109 + }, + { + "epoch": 0.4206471494607088, + "grad_norm": 1.4719623485063398, + "learning_rate": 6.504279720553218e-06, + "loss": 0.3772, + "step": 19110 + }, + { + "epoch": 0.42066916134712745, + "grad_norm": 1.5397637002043527, + "learning_rate": 6.503939775119051e-06, + "loss": 0.3744, + "step": 19111 + }, + { + "epoch": 0.4206911732335461, + "grad_norm": 1.4108070504665557, + "learning_rate": 6.5035998220411915e-06, + "loss": 0.3035, + "step": 19112 + }, + { + "epoch": 0.4207131851199648, + "grad_norm": 0.649417356365693, + "learning_rate": 6.503259861321365e-06, + "loss": 0.5218, + "step": 19113 + }, + { + "epoch": 0.42073519700638345, + "grad_norm": 0.6313994094928616, + "learning_rate": 6.5029198929613025e-06, + "loss": 0.4955, + "step": 19114 + }, + { + "epoch": 0.4207572088928021, + "grad_norm": 2.793681423266433, + "learning_rate": 6.50257991696273e-06, + "loss": 0.2504, + "step": 19115 + }, + { + "epoch": 0.42077922077922075, + "grad_norm": 1.2593182441712516, + "learning_rate": 6.5022399333273775e-06, + "loss": 0.2985, + "step": 19116 + }, + { + "epoch": 0.42080123266563946, + "grad_norm": 0.6160313952576195, + "learning_rate": 6.501899942056971e-06, + "loss": 0.4811, + "step": 19117 + }, + { + "epoch": 0.4208232445520581, + "grad_norm": 1.3733100926922834, + "learning_rate": 6.501559943153238e-06, + "loss": 0.3214, + "step": 19118 + }, + { + "epoch": 0.42084525643847676, + "grad_norm": 1.3084358100800852, + "learning_rate": 6.501219936617909e-06, + "loss": 0.434, + "step": 19119 + }, + { + "epoch": 0.42086726832489546, + "grad_norm": 0.6369510674783893, + "learning_rate": 6.500879922452711e-06, + "loss": 0.5236, + "step": 19120 + }, + { + "epoch": 0.4208892802113141, + "grad_norm": 1.3252906636174855, + "learning_rate": 6.5005399006593704e-06, + "loss": 0.2464, + "step": 19121 + }, + { + "epoch": 0.42091129209773276, + "grad_norm": 1.1468558669029065, + "learning_rate": 6.500199871239619e-06, + "loss": 0.3026, + "step": 19122 + }, + { + "epoch": 0.42093330398415146, + "grad_norm": 0.6134925498211105, + "learning_rate": 6.49985983419518e-06, + "loss": 0.4985, + "step": 19123 + }, + { + "epoch": 0.4209553158705701, + "grad_norm": 1.4305255418453309, + "learning_rate": 6.499519789527786e-06, + "loss": 0.3431, + "step": 19124 + }, + { + "epoch": 0.42097732775698876, + "grad_norm": 1.4084444650334491, + "learning_rate": 6.499179737239163e-06, + "loss": 0.2351, + "step": 19125 + }, + { + "epoch": 0.42099933964340747, + "grad_norm": 1.321933490000465, + "learning_rate": 6.49883967733104e-06, + "loss": 0.3252, + "step": 19126 + }, + { + "epoch": 0.4210213515298261, + "grad_norm": 0.6315683392286222, + "learning_rate": 6.498499609805147e-06, + "loss": 0.5263, + "step": 19127 + }, + { + "epoch": 0.42104336341624476, + "grad_norm": 1.350817456926369, + "learning_rate": 6.498159534663208e-06, + "loss": 0.3856, + "step": 19128 + }, + { + "epoch": 0.4210653753026634, + "grad_norm": 1.6558555555130066, + "learning_rate": 6.497819451906956e-06, + "loss": 0.3387, + "step": 19129 + }, + { + "epoch": 0.4210873871890821, + "grad_norm": 1.201743423783858, + "learning_rate": 6.497479361538117e-06, + "loss": 0.3711, + "step": 19130 + }, + { + "epoch": 0.42110939907550077, + "grad_norm": 0.6240322419966305, + "learning_rate": 6.497139263558418e-06, + "loss": 0.4988, + "step": 19131 + }, + { + "epoch": 0.4211314109619194, + "grad_norm": 0.6411133845677196, + "learning_rate": 6.496799157969592e-06, + "loss": 0.4813, + "step": 19132 + }, + { + "epoch": 0.4211534228483381, + "grad_norm": 1.3254277154114755, + "learning_rate": 6.4964590447733644e-06, + "loss": 0.2623, + "step": 19133 + }, + { + "epoch": 0.42117543473475677, + "grad_norm": 1.4483251951688727, + "learning_rate": 6.496118923971465e-06, + "loss": 0.3269, + "step": 19134 + }, + { + "epoch": 0.4211974466211754, + "grad_norm": 1.3399600254352773, + "learning_rate": 6.4957787955656215e-06, + "loss": 0.288, + "step": 19135 + }, + { + "epoch": 0.4212194585075941, + "grad_norm": 1.7922001337449411, + "learning_rate": 6.495438659557561e-06, + "loss": 0.2738, + "step": 19136 + }, + { + "epoch": 0.4212414703940128, + "grad_norm": 1.4011785001845598, + "learning_rate": 6.495098515949016e-06, + "loss": 0.4026, + "step": 19137 + }, + { + "epoch": 0.4212634822804314, + "grad_norm": 1.7365625066283967, + "learning_rate": 6.494758364741715e-06, + "loss": 0.41, + "step": 19138 + }, + { + "epoch": 0.42128549416685007, + "grad_norm": 1.8776167825357877, + "learning_rate": 6.494418205937384e-06, + "loss": 0.4361, + "step": 19139 + }, + { + "epoch": 0.4213075060532688, + "grad_norm": 1.4258570600763576, + "learning_rate": 6.494078039537753e-06, + "loss": 0.3099, + "step": 19140 + }, + { + "epoch": 0.4213295179396874, + "grad_norm": 1.1394760885423794, + "learning_rate": 6.4937378655445505e-06, + "loss": 0.2689, + "step": 19141 + }, + { + "epoch": 0.4213515298261061, + "grad_norm": 5.9838896080273205, + "learning_rate": 6.493397683959507e-06, + "loss": 0.3007, + "step": 19142 + }, + { + "epoch": 0.4213735417125248, + "grad_norm": 1.9215406192122344, + "learning_rate": 6.493057494784349e-06, + "loss": 0.336, + "step": 19143 + }, + { + "epoch": 0.42139555359894343, + "grad_norm": 2.0345994967659573, + "learning_rate": 6.492717298020809e-06, + "loss": 0.3734, + "step": 19144 + }, + { + "epoch": 0.4214175654853621, + "grad_norm": 0.6521181305081625, + "learning_rate": 6.492377093670612e-06, + "loss": 0.4774, + "step": 19145 + }, + { + "epoch": 0.4214395773717808, + "grad_norm": 1.3940500874877264, + "learning_rate": 6.49203688173549e-06, + "loss": 0.3209, + "step": 19146 + }, + { + "epoch": 0.42146158925819943, + "grad_norm": 0.6817080329283128, + "learning_rate": 6.491696662217172e-06, + "loss": 0.5085, + "step": 19147 + }, + { + "epoch": 0.4214836011446181, + "grad_norm": 1.3115579085254592, + "learning_rate": 6.491356435117385e-06, + "loss": 0.3263, + "step": 19148 + }, + { + "epoch": 0.4215056130310368, + "grad_norm": 1.0757402440062878, + "learning_rate": 6.491016200437858e-06, + "loss": 0.2431, + "step": 19149 + }, + { + "epoch": 0.42152762491745543, + "grad_norm": 1.756778367909711, + "learning_rate": 6.490675958180324e-06, + "loss": 0.3616, + "step": 19150 + }, + { + "epoch": 0.4215496368038741, + "grad_norm": 2.0180645208910044, + "learning_rate": 6.490335708346509e-06, + "loss": 0.4036, + "step": 19151 + }, + { + "epoch": 0.42157164869029273, + "grad_norm": 1.5373890425589625, + "learning_rate": 6.489995450938143e-06, + "loss": 0.2756, + "step": 19152 + }, + { + "epoch": 0.42159366057671144, + "grad_norm": 1.4205197799054434, + "learning_rate": 6.489655185956957e-06, + "loss": 0.3735, + "step": 19153 + }, + { + "epoch": 0.4216156724631301, + "grad_norm": 1.529907007713187, + "learning_rate": 6.489314913404676e-06, + "loss": 0.3294, + "step": 19154 + }, + { + "epoch": 0.42163768434954874, + "grad_norm": 1.29266342984432, + "learning_rate": 6.488974633283035e-06, + "loss": 0.3168, + "step": 19155 + }, + { + "epoch": 0.42165969623596744, + "grad_norm": 8.141116909754155, + "learning_rate": 6.48863434559376e-06, + "loss": 0.274, + "step": 19156 + }, + { + "epoch": 0.4216817081223861, + "grad_norm": 0.6751749663434896, + "learning_rate": 6.4882940503385815e-06, + "loss": 0.5191, + "step": 19157 + }, + { + "epoch": 0.42170372000880474, + "grad_norm": 1.32331385024617, + "learning_rate": 6.487953747519229e-06, + "loss": 0.3216, + "step": 19158 + }, + { + "epoch": 0.42172573189522344, + "grad_norm": 1.0281294504062453, + "learning_rate": 6.48761343713743e-06, + "loss": 0.4958, + "step": 19159 + }, + { + "epoch": 0.4217477437816421, + "grad_norm": 0.638309889398829, + "learning_rate": 6.487273119194918e-06, + "loss": 0.4956, + "step": 19160 + }, + { + "epoch": 0.42176975566806074, + "grad_norm": 1.2069411201115026, + "learning_rate": 6.486932793693421e-06, + "loss": 0.2716, + "step": 19161 + }, + { + "epoch": 0.42179176755447945, + "grad_norm": 1.6633524818166487, + "learning_rate": 6.4865924606346674e-06, + "loss": 0.2858, + "step": 19162 + }, + { + "epoch": 0.4218137794408981, + "grad_norm": 1.3646792220551676, + "learning_rate": 6.486252120020389e-06, + "loss": 0.2692, + "step": 19163 + }, + { + "epoch": 0.42183579132731674, + "grad_norm": 1.5885523503284764, + "learning_rate": 6.485911771852311e-06, + "loss": 0.424, + "step": 19164 + }, + { + "epoch": 0.4218578032137354, + "grad_norm": 1.3098014671697717, + "learning_rate": 6.485571416132172e-06, + "loss": 0.2669, + "step": 19165 + }, + { + "epoch": 0.4218798151001541, + "grad_norm": 1.4717855732397394, + "learning_rate": 6.485231052861693e-06, + "loss": 0.2797, + "step": 19166 + }, + { + "epoch": 0.42190182698657275, + "grad_norm": 0.7167650058125621, + "learning_rate": 6.484890682042607e-06, + "loss": 0.4831, + "step": 19167 + }, + { + "epoch": 0.4219238388729914, + "grad_norm": 1.7589444898654454, + "learning_rate": 6.484550303676646e-06, + "loss": 0.2219, + "step": 19168 + }, + { + "epoch": 0.4219458507594101, + "grad_norm": 0.6449435972590495, + "learning_rate": 6.484209917765537e-06, + "loss": 0.4766, + "step": 19169 + }, + { + "epoch": 0.42196786264582875, + "grad_norm": 2.7936838109731195, + "learning_rate": 6.483869524311012e-06, + "loss": 0.3476, + "step": 19170 + }, + { + "epoch": 0.4219898745322474, + "grad_norm": 1.2292004327705384, + "learning_rate": 6.4835291233148e-06, + "loss": 0.326, + "step": 19171 + }, + { + "epoch": 0.4220118864186661, + "grad_norm": 1.2698009155425594, + "learning_rate": 6.4831887147786295e-06, + "loss": 0.3519, + "step": 19172 + }, + { + "epoch": 0.42203389830508475, + "grad_norm": 1.568735319342088, + "learning_rate": 6.4828482987042356e-06, + "loss": 0.3539, + "step": 19173 + }, + { + "epoch": 0.4220559101915034, + "grad_norm": 1.5939200814112162, + "learning_rate": 6.482507875093345e-06, + "loss": 0.2646, + "step": 19174 + }, + { + "epoch": 0.42207792207792205, + "grad_norm": 1.2801151759421499, + "learning_rate": 6.482167443947687e-06, + "loss": 0.2831, + "step": 19175 + }, + { + "epoch": 0.42209993396434076, + "grad_norm": 1.7382348404722627, + "learning_rate": 6.481827005268992e-06, + "loss": 0.4349, + "step": 19176 + }, + { + "epoch": 0.4221219458507594, + "grad_norm": 1.3560646397115392, + "learning_rate": 6.481486559058991e-06, + "loss": 0.2932, + "step": 19177 + }, + { + "epoch": 0.42214395773717805, + "grad_norm": 0.7006530268749822, + "learning_rate": 6.4811461053194165e-06, + "loss": 0.4889, + "step": 19178 + }, + { + "epoch": 0.42216596962359676, + "grad_norm": 0.6573857103860127, + "learning_rate": 6.480805644051997e-06, + "loss": 0.486, + "step": 19179 + }, + { + "epoch": 0.4221879815100154, + "grad_norm": 0.6114730668035514, + "learning_rate": 6.4804651752584615e-06, + "loss": 0.4907, + "step": 19180 + }, + { + "epoch": 0.42220999339643406, + "grad_norm": 1.3858960914209197, + "learning_rate": 6.480124698940542e-06, + "loss": 0.3176, + "step": 19181 + }, + { + "epoch": 0.42223200528285276, + "grad_norm": 1.3440043508596415, + "learning_rate": 6.479784215099969e-06, + "loss": 0.3864, + "step": 19182 + }, + { + "epoch": 0.4222540171692714, + "grad_norm": 1.7813397946874006, + "learning_rate": 6.4794437237384735e-06, + "loss": 0.4068, + "step": 19183 + }, + { + "epoch": 0.42227602905569006, + "grad_norm": 1.9505363976108692, + "learning_rate": 6.479103224857784e-06, + "loss": 0.3564, + "step": 19184 + }, + { + "epoch": 0.42229804094210877, + "grad_norm": 1.3518664110689929, + "learning_rate": 6.478762718459632e-06, + "loss": 0.3251, + "step": 19185 + }, + { + "epoch": 0.4223200528285274, + "grad_norm": 1.2076260648498351, + "learning_rate": 6.478422204545748e-06, + "loss": 0.3382, + "step": 19186 + }, + { + "epoch": 0.42234206471494606, + "grad_norm": 1.4636358903546545, + "learning_rate": 6.478081683117864e-06, + "loss": 0.3089, + "step": 19187 + }, + { + "epoch": 0.4223640766013647, + "grad_norm": 1.4675261899209264, + "learning_rate": 6.477741154177709e-06, + "loss": 0.3449, + "step": 19188 + }, + { + "epoch": 0.4223860884877834, + "grad_norm": 1.2001139610225822, + "learning_rate": 6.477400617727014e-06, + "loss": 0.2526, + "step": 19189 + }, + { + "epoch": 0.42240810037420207, + "grad_norm": 1.2002499614803248, + "learning_rate": 6.477060073767511e-06, + "loss": 0.2795, + "step": 19190 + }, + { + "epoch": 0.4224301122606207, + "grad_norm": 1.5772439754156513, + "learning_rate": 6.476719522300929e-06, + "loss": 0.272, + "step": 19191 + }, + { + "epoch": 0.4224521241470394, + "grad_norm": 1.2462725221907272, + "learning_rate": 6.476378963329001e-06, + "loss": 0.3277, + "step": 19192 + }, + { + "epoch": 0.42247413603345807, + "grad_norm": 1.4719769759620658, + "learning_rate": 6.4760383968534554e-06, + "loss": 0.2991, + "step": 19193 + }, + { + "epoch": 0.4224961479198767, + "grad_norm": 1.865210447271744, + "learning_rate": 6.475697822876026e-06, + "loss": 0.3709, + "step": 19194 + }, + { + "epoch": 0.4225181598062954, + "grad_norm": 1.2509167168550133, + "learning_rate": 6.475357241398439e-06, + "loss": 0.3839, + "step": 19195 + }, + { + "epoch": 0.4225401716927141, + "grad_norm": 1.6879222933431717, + "learning_rate": 6.475016652422431e-06, + "loss": 0.3768, + "step": 19196 + }, + { + "epoch": 0.4225621835791327, + "grad_norm": 0.9622934351508557, + "learning_rate": 6.4746760559497316e-06, + "loss": 0.5198, + "step": 19197 + }, + { + "epoch": 0.42258419546555137, + "grad_norm": 1.3708422139973535, + "learning_rate": 6.474335451982068e-06, + "loss": 0.3508, + "step": 19198 + }, + { + "epoch": 0.4226062073519701, + "grad_norm": 11.034087058894338, + "learning_rate": 6.473994840521175e-06, + "loss": 0.3631, + "step": 19199 + }, + { + "epoch": 0.4226282192383887, + "grad_norm": 1.2367083564318886, + "learning_rate": 6.473654221568783e-06, + "loss": 0.2846, + "step": 19200 + }, + { + "epoch": 0.4226502311248074, + "grad_norm": 1.2626112614452996, + "learning_rate": 6.473313595126623e-06, + "loss": 0.3381, + "step": 19201 + }, + { + "epoch": 0.4226722430112261, + "grad_norm": 0.6236969933343041, + "learning_rate": 6.472972961196426e-06, + "loss": 0.483, + "step": 19202 + }, + { + "epoch": 0.42269425489764473, + "grad_norm": 1.2720911396464738, + "learning_rate": 6.472632319779923e-06, + "loss": 0.3394, + "step": 19203 + }, + { + "epoch": 0.4227162667840634, + "grad_norm": 1.4551275909039414, + "learning_rate": 6.472291670878846e-06, + "loss": 0.4045, + "step": 19204 + }, + { + "epoch": 0.4227382786704821, + "grad_norm": 1.414624097486344, + "learning_rate": 6.471951014494927e-06, + "loss": 0.372, + "step": 19205 + }, + { + "epoch": 0.42276029055690073, + "grad_norm": 1.7733312934462777, + "learning_rate": 6.4716103506298965e-06, + "loss": 0.2019, + "step": 19206 + }, + { + "epoch": 0.4227823024433194, + "grad_norm": 1.3025986101222315, + "learning_rate": 6.471269679285486e-06, + "loss": 0.3664, + "step": 19207 + }, + { + "epoch": 0.4228043143297381, + "grad_norm": 1.267549982389706, + "learning_rate": 6.470929000463424e-06, + "loss": 0.2838, + "step": 19208 + }, + { + "epoch": 0.42282632621615673, + "grad_norm": 1.3180282701025052, + "learning_rate": 6.470588314165447e-06, + "loss": 0.2704, + "step": 19209 + }, + { + "epoch": 0.4228483381025754, + "grad_norm": 1.3262935468894026, + "learning_rate": 6.470247620393284e-06, + "loss": 0.3037, + "step": 19210 + }, + { + "epoch": 0.42287034998899403, + "grad_norm": 1.5997988887287844, + "learning_rate": 6.469906919148668e-06, + "loss": 0.2704, + "step": 19211 + }, + { + "epoch": 0.42289236187541274, + "grad_norm": 1.454972016735877, + "learning_rate": 6.4695662104333275e-06, + "loss": 0.3627, + "step": 19212 + }, + { + "epoch": 0.4229143737618314, + "grad_norm": 1.1459027905052364, + "learning_rate": 6.4692254942489965e-06, + "loss": 0.275, + "step": 19213 + }, + { + "epoch": 0.42293638564825004, + "grad_norm": 1.848106158111423, + "learning_rate": 6.468884770597407e-06, + "loss": 0.337, + "step": 19214 + }, + { + "epoch": 0.42295839753466874, + "grad_norm": 1.182025726851483, + "learning_rate": 6.46854403948029e-06, + "loss": 0.3142, + "step": 19215 + }, + { + "epoch": 0.4229804094210874, + "grad_norm": 1.2449097748892102, + "learning_rate": 6.468203300899376e-06, + "loss": 0.2694, + "step": 19216 + }, + { + "epoch": 0.42300242130750604, + "grad_norm": 1.3984235161230139, + "learning_rate": 6.467862554856399e-06, + "loss": 0.4059, + "step": 19217 + }, + { + "epoch": 0.42302443319392474, + "grad_norm": 0.8520452674457841, + "learning_rate": 6.467521801353088e-06, + "loss": 0.4801, + "step": 19218 + }, + { + "epoch": 0.4230464450803434, + "grad_norm": 1.4678466474872118, + "learning_rate": 6.46718104039118e-06, + "loss": 0.3646, + "step": 19219 + }, + { + "epoch": 0.42306845696676204, + "grad_norm": 1.7963809867826361, + "learning_rate": 6.466840271972402e-06, + "loss": 0.2978, + "step": 19220 + }, + { + "epoch": 0.4230904688531807, + "grad_norm": 2.235175414616336, + "learning_rate": 6.466499496098486e-06, + "loss": 0.4033, + "step": 19221 + }, + { + "epoch": 0.4231124807395994, + "grad_norm": 2.531485094727673, + "learning_rate": 6.466158712771168e-06, + "loss": 0.3527, + "step": 19222 + }, + { + "epoch": 0.42313449262601804, + "grad_norm": 1.4611291741624377, + "learning_rate": 6.4658179219921755e-06, + "loss": 0.3172, + "step": 19223 + }, + { + "epoch": 0.4231565045124367, + "grad_norm": 1.6232619462203135, + "learning_rate": 6.465477123763244e-06, + "loss": 0.3117, + "step": 19224 + }, + { + "epoch": 0.4231785163988554, + "grad_norm": 1.7743411000611033, + "learning_rate": 6.465136318086103e-06, + "loss": 0.2746, + "step": 19225 + }, + { + "epoch": 0.42320052828527405, + "grad_norm": 3.0623615893313114, + "learning_rate": 6.464795504962485e-06, + "loss": 0.4011, + "step": 19226 + }, + { + "epoch": 0.4232225401716927, + "grad_norm": 1.3431000935092485, + "learning_rate": 6.464454684394126e-06, + "loss": 0.2878, + "step": 19227 + }, + { + "epoch": 0.4232445520581114, + "grad_norm": 1.5153834431754523, + "learning_rate": 6.464113856382752e-06, + "loss": 0.2902, + "step": 19228 + }, + { + "epoch": 0.42326656394453005, + "grad_norm": 2.797175628830449, + "learning_rate": 6.4637730209301004e-06, + "loss": 0.2872, + "step": 19229 + }, + { + "epoch": 0.4232885758309487, + "grad_norm": 1.264569188040105, + "learning_rate": 6.463432178037901e-06, + "loss": 0.3669, + "step": 19230 + }, + { + "epoch": 0.4233105877173674, + "grad_norm": 1.2919979102282706, + "learning_rate": 6.463091327707885e-06, + "loss": 0.3048, + "step": 19231 + }, + { + "epoch": 0.42333259960378605, + "grad_norm": 0.6885897265113112, + "learning_rate": 6.462750469941788e-06, + "loss": 0.5023, + "step": 19232 + }, + { + "epoch": 0.4233546114902047, + "grad_norm": 0.6597699537165946, + "learning_rate": 6.462409604741342e-06, + "loss": 0.4885, + "step": 19233 + }, + { + "epoch": 0.42337662337662335, + "grad_norm": 1.576688221277499, + "learning_rate": 6.462068732108275e-06, + "loss": 0.3406, + "step": 19234 + }, + { + "epoch": 0.42339863526304206, + "grad_norm": 1.5091535229261202, + "learning_rate": 6.461727852044324e-06, + "loss": 0.3071, + "step": 19235 + }, + { + "epoch": 0.4234206471494607, + "grad_norm": 0.6436005019095287, + "learning_rate": 6.46138696455122e-06, + "loss": 0.5071, + "step": 19236 + }, + { + "epoch": 0.42344265903587935, + "grad_norm": 1.4071574101929047, + "learning_rate": 6.4610460696306954e-06, + "loss": 0.3851, + "step": 19237 + }, + { + "epoch": 0.42346467092229806, + "grad_norm": 1.1557453190568037, + "learning_rate": 6.460705167284484e-06, + "loss": 0.376, + "step": 19238 + }, + { + "epoch": 0.4234866828087167, + "grad_norm": 1.2619325886986594, + "learning_rate": 6.460364257514317e-06, + "loss": 0.3064, + "step": 19239 + }, + { + "epoch": 0.42350869469513536, + "grad_norm": 1.543965073529795, + "learning_rate": 6.460023340321927e-06, + "loss": 0.3354, + "step": 19240 + }, + { + "epoch": 0.42353070658155406, + "grad_norm": 1.2089125004381975, + "learning_rate": 6.4596824157090475e-06, + "loss": 0.2667, + "step": 19241 + }, + { + "epoch": 0.4235527184679727, + "grad_norm": 2.3156906557262524, + "learning_rate": 6.45934148367741e-06, + "loss": 0.3269, + "step": 19242 + }, + { + "epoch": 0.42357473035439136, + "grad_norm": 1.8995428419422231, + "learning_rate": 6.45900054422875e-06, + "loss": 0.357, + "step": 19243 + }, + { + "epoch": 0.42359674224081006, + "grad_norm": 0.7124029072236568, + "learning_rate": 6.458659597364798e-06, + "loss": 0.503, + "step": 19244 + }, + { + "epoch": 0.4236187541272287, + "grad_norm": 1.176634394736162, + "learning_rate": 6.458318643087287e-06, + "loss": 0.3519, + "step": 19245 + }, + { + "epoch": 0.42364076601364736, + "grad_norm": 2.290603620915602, + "learning_rate": 6.457977681397951e-06, + "loss": 0.3056, + "step": 19246 + }, + { + "epoch": 0.423662777900066, + "grad_norm": 1.8557897167720978, + "learning_rate": 6.457636712298521e-06, + "loss": 0.3026, + "step": 19247 + }, + { + "epoch": 0.4236847897864847, + "grad_norm": 1.4178357752521324, + "learning_rate": 6.457295735790732e-06, + "loss": 0.331, + "step": 19248 + }, + { + "epoch": 0.42370680167290337, + "grad_norm": 0.6346861456251567, + "learning_rate": 6.456954751876314e-06, + "loss": 0.4774, + "step": 19249 + }, + { + "epoch": 0.423728813559322, + "grad_norm": 1.0928659724608791, + "learning_rate": 6.456613760557005e-06, + "loss": 0.2471, + "step": 19250 + }, + { + "epoch": 0.4237508254457407, + "grad_norm": 1.5305605549282648, + "learning_rate": 6.456272761834533e-06, + "loss": 0.4447, + "step": 19251 + }, + { + "epoch": 0.42377283733215937, + "grad_norm": 1.6766943528724583, + "learning_rate": 6.455931755710635e-06, + "loss": 0.4023, + "step": 19252 + }, + { + "epoch": 0.423794849218578, + "grad_norm": 3.00640946304148, + "learning_rate": 6.455590742187041e-06, + "loss": 0.2473, + "step": 19253 + }, + { + "epoch": 0.4238168611049967, + "grad_norm": 1.3170113248009307, + "learning_rate": 6.455249721265487e-06, + "loss": 0.3171, + "step": 19254 + }, + { + "epoch": 0.42383887299141537, + "grad_norm": 1.6865012566943394, + "learning_rate": 6.454908692947703e-06, + "loss": 0.3912, + "step": 19255 + }, + { + "epoch": 0.423860884877834, + "grad_norm": 1.6545985216418435, + "learning_rate": 6.454567657235427e-06, + "loss": 0.289, + "step": 19256 + }, + { + "epoch": 0.42388289676425267, + "grad_norm": 2.8281419591687342, + "learning_rate": 6.4542266141303865e-06, + "loss": 0.2732, + "step": 19257 + }, + { + "epoch": 0.4239049086506714, + "grad_norm": 1.4312147863354656, + "learning_rate": 6.453885563634319e-06, + "loss": 0.3519, + "step": 19258 + }, + { + "epoch": 0.42392692053709, + "grad_norm": 0.6563089762796682, + "learning_rate": 6.453544505748958e-06, + "loss": 0.484, + "step": 19259 + }, + { + "epoch": 0.4239489324235087, + "grad_norm": 1.4532306351173496, + "learning_rate": 6.453203440476033e-06, + "loss": 0.2794, + "step": 19260 + }, + { + "epoch": 0.4239709443099274, + "grad_norm": 1.637859940747654, + "learning_rate": 6.452862367817282e-06, + "loss": 0.3222, + "step": 19261 + }, + { + "epoch": 0.423992956196346, + "grad_norm": 0.6261461229858527, + "learning_rate": 6.452521287774435e-06, + "loss": 0.4811, + "step": 19262 + }, + { + "epoch": 0.4240149680827647, + "grad_norm": 1.5483046902187376, + "learning_rate": 6.4521802003492274e-06, + "loss": 0.3219, + "step": 19263 + }, + { + "epoch": 0.4240369799691834, + "grad_norm": 1.5000317135460437, + "learning_rate": 6.4518391055433925e-06, + "loss": 0.2915, + "step": 19264 + }, + { + "epoch": 0.42405899185560203, + "grad_norm": 1.4000325580103474, + "learning_rate": 6.451498003358664e-06, + "loss": 0.3216, + "step": 19265 + }, + { + "epoch": 0.4240810037420207, + "grad_norm": 0.6088432591270052, + "learning_rate": 6.451156893796775e-06, + "loss": 0.4577, + "step": 19266 + }, + { + "epoch": 0.4241030156284394, + "grad_norm": 1.056482114202588, + "learning_rate": 6.450815776859458e-06, + "loss": 0.2316, + "step": 19267 + }, + { + "epoch": 0.42412502751485803, + "grad_norm": 1.3916913544330844, + "learning_rate": 6.45047465254845e-06, + "loss": 0.3487, + "step": 19268 + }, + { + "epoch": 0.4241470394012767, + "grad_norm": 0.6064268910763508, + "learning_rate": 6.450133520865482e-06, + "loss": 0.4894, + "step": 19269 + }, + { + "epoch": 0.42416905128769533, + "grad_norm": 1.4328046785557143, + "learning_rate": 6.44979238181229e-06, + "loss": 0.3609, + "step": 19270 + }, + { + "epoch": 0.42419106317411404, + "grad_norm": 1.8886468082495038, + "learning_rate": 6.449451235390605e-06, + "loss": 0.2871, + "step": 19271 + }, + { + "epoch": 0.4242130750605327, + "grad_norm": 2.127786253823263, + "learning_rate": 6.449110081602164e-06, + "loss": 0.2374, + "step": 19272 + }, + { + "epoch": 0.42423508694695133, + "grad_norm": 3.2992299854704568, + "learning_rate": 6.448768920448697e-06, + "loss": 0.3415, + "step": 19273 + }, + { + "epoch": 0.42425709883337004, + "grad_norm": 1.6665682873386316, + "learning_rate": 6.4484277519319424e-06, + "loss": 0.3939, + "step": 19274 + }, + { + "epoch": 0.4242791107197887, + "grad_norm": 0.6075887568355027, + "learning_rate": 6.44808657605363e-06, + "loss": 0.5008, + "step": 19275 + }, + { + "epoch": 0.42430112260620734, + "grad_norm": 1.2458669273533545, + "learning_rate": 6.447745392815497e-06, + "loss": 0.2704, + "step": 19276 + }, + { + "epoch": 0.42432313449262604, + "grad_norm": 1.4747941326053688, + "learning_rate": 6.447404202219275e-06, + "loss": 0.4448, + "step": 19277 + }, + { + "epoch": 0.4243451463790447, + "grad_norm": 0.6254459603598854, + "learning_rate": 6.447063004266699e-06, + "loss": 0.4787, + "step": 19278 + }, + { + "epoch": 0.42436715826546334, + "grad_norm": 1.861293718177767, + "learning_rate": 6.446721798959506e-06, + "loss": 0.2644, + "step": 19279 + }, + { + "epoch": 0.424389170151882, + "grad_norm": 1.5537396141944109, + "learning_rate": 6.446380586299426e-06, + "loss": 0.3004, + "step": 19280 + }, + { + "epoch": 0.4244111820383007, + "grad_norm": 1.442428095950945, + "learning_rate": 6.446039366288195e-06, + "loss": 0.2978, + "step": 19281 + }, + { + "epoch": 0.42443319392471934, + "grad_norm": 2.141690475114178, + "learning_rate": 6.445698138927547e-06, + "loss": 0.2779, + "step": 19282 + }, + { + "epoch": 0.424455205811138, + "grad_norm": 1.53492557614832, + "learning_rate": 6.445356904219215e-06, + "loss": 0.2899, + "step": 19283 + }, + { + "epoch": 0.4244772176975567, + "grad_norm": 3.16607034863169, + "learning_rate": 6.445015662164936e-06, + "loss": 0.3306, + "step": 19284 + }, + { + "epoch": 0.42449922958397535, + "grad_norm": 0.6263351178460635, + "learning_rate": 6.444674412766443e-06, + "loss": 0.4995, + "step": 19285 + }, + { + "epoch": 0.424521241470394, + "grad_norm": 1.523992902388864, + "learning_rate": 6.444333156025469e-06, + "loss": 0.3251, + "step": 19286 + }, + { + "epoch": 0.4245432533568127, + "grad_norm": 1.5387775969882367, + "learning_rate": 6.4439918919437505e-06, + "loss": 0.3216, + "step": 19287 + }, + { + "epoch": 0.42456526524323135, + "grad_norm": 0.6006324404034277, + "learning_rate": 6.443650620523022e-06, + "loss": 0.4752, + "step": 19288 + }, + { + "epoch": 0.42458727712965, + "grad_norm": 1.3360987005344735, + "learning_rate": 6.443309341765016e-06, + "loss": 0.3226, + "step": 19289 + }, + { + "epoch": 0.4246092890160687, + "grad_norm": 1.734889317605349, + "learning_rate": 6.442968055671469e-06, + "loss": 0.2728, + "step": 19290 + }, + { + "epoch": 0.42463130090248735, + "grad_norm": 3.6714934227551366, + "learning_rate": 6.442626762244113e-06, + "loss": 0.2838, + "step": 19291 + }, + { + "epoch": 0.424653312788906, + "grad_norm": 1.5995889595959272, + "learning_rate": 6.4422854614846865e-06, + "loss": 0.412, + "step": 19292 + }, + { + "epoch": 0.42467532467532465, + "grad_norm": 1.3657104408378526, + "learning_rate": 6.441944153394921e-06, + "loss": 0.2873, + "step": 19293 + }, + { + "epoch": 0.42469733656174335, + "grad_norm": 1.413774479158579, + "learning_rate": 6.441602837976553e-06, + "loss": 0.3902, + "step": 19294 + }, + { + "epoch": 0.424719348448162, + "grad_norm": 1.3757493293638328, + "learning_rate": 6.441261515231316e-06, + "loss": 0.4084, + "step": 19295 + }, + { + "epoch": 0.42474136033458065, + "grad_norm": 0.6352810351669179, + "learning_rate": 6.440920185160943e-06, + "loss": 0.4782, + "step": 19296 + }, + { + "epoch": 0.42476337222099936, + "grad_norm": 1.782341007176886, + "learning_rate": 6.4405788477671745e-06, + "loss": 0.3569, + "step": 19297 + }, + { + "epoch": 0.424785384107418, + "grad_norm": 1.6096995381057724, + "learning_rate": 6.440237503051739e-06, + "loss": 0.3588, + "step": 19298 + }, + { + "epoch": 0.42480739599383666, + "grad_norm": 1.6224466907424346, + "learning_rate": 6.439896151016376e-06, + "loss": 0.2732, + "step": 19299 + }, + { + "epoch": 0.42482940788025536, + "grad_norm": 1.5378095565439536, + "learning_rate": 6.439554791662819e-06, + "loss": 0.3254, + "step": 19300 + }, + { + "epoch": 0.424851419766674, + "grad_norm": 1.8642955038648636, + "learning_rate": 6.439213424992799e-06, + "loss": 0.3191, + "step": 19301 + }, + { + "epoch": 0.42487343165309266, + "grad_norm": 1.2627493900645992, + "learning_rate": 6.4388720510080585e-06, + "loss": 0.3349, + "step": 19302 + }, + { + "epoch": 0.42489544353951136, + "grad_norm": 0.6248634870223141, + "learning_rate": 6.438530669710326e-06, + "loss": 0.5267, + "step": 19303 + }, + { + "epoch": 0.42491745542593, + "grad_norm": 2.5056204240292606, + "learning_rate": 6.43818928110134e-06, + "loss": 0.3706, + "step": 19304 + }, + { + "epoch": 0.42493946731234866, + "grad_norm": 1.6972022933173434, + "learning_rate": 6.437847885182835e-06, + "loss": 0.344, + "step": 19305 + }, + { + "epoch": 0.4249614791987673, + "grad_norm": 1.716499933580868, + "learning_rate": 6.437506481956546e-06, + "loss": 0.4093, + "step": 19306 + }, + { + "epoch": 0.424983491085186, + "grad_norm": 1.5874710710415039, + "learning_rate": 6.437165071424206e-06, + "loss": 0.3243, + "step": 19307 + }, + { + "epoch": 0.42500550297160467, + "grad_norm": 1.437675842236055, + "learning_rate": 6.436823653587554e-06, + "loss": 0.3134, + "step": 19308 + }, + { + "epoch": 0.4250275148580233, + "grad_norm": 2.5722489154222465, + "learning_rate": 6.436482228448322e-06, + "loss": 0.3732, + "step": 19309 + }, + { + "epoch": 0.425049526744442, + "grad_norm": 1.475561981698985, + "learning_rate": 6.436140796008247e-06, + "loss": 0.3303, + "step": 19310 + }, + { + "epoch": 0.42507153863086067, + "grad_norm": 1.5888832890818971, + "learning_rate": 6.435799356269064e-06, + "loss": 0.3081, + "step": 19311 + }, + { + "epoch": 0.4250935505172793, + "grad_norm": 2.470082557520197, + "learning_rate": 6.435457909232509e-06, + "loss": 0.3262, + "step": 19312 + }, + { + "epoch": 0.425115562403698, + "grad_norm": 1.826041446699546, + "learning_rate": 6.435116454900317e-06, + "loss": 0.3657, + "step": 19313 + }, + { + "epoch": 0.42513757429011667, + "grad_norm": 1.3832570396299946, + "learning_rate": 6.43477499327422e-06, + "loss": 0.297, + "step": 19314 + }, + { + "epoch": 0.4251595861765353, + "grad_norm": 0.6400903193340509, + "learning_rate": 6.434433524355961e-06, + "loss": 0.4936, + "step": 19315 + }, + { + "epoch": 0.42518159806295397, + "grad_norm": 0.6328645663462091, + "learning_rate": 6.434092048147267e-06, + "loss": 0.4879, + "step": 19316 + }, + { + "epoch": 0.4252036099493727, + "grad_norm": 0.6212064413664297, + "learning_rate": 6.43375056464988e-06, + "loss": 0.4709, + "step": 19317 + }, + { + "epoch": 0.4252256218357913, + "grad_norm": 2.2182480857501425, + "learning_rate": 6.433409073865533e-06, + "loss": 0.2922, + "step": 19318 + }, + { + "epoch": 0.42524763372221, + "grad_norm": 1.5518335340460634, + "learning_rate": 6.43306757579596e-06, + "loss": 0.3851, + "step": 19319 + }, + { + "epoch": 0.4252696456086287, + "grad_norm": 1.7084247148817573, + "learning_rate": 6.432726070442901e-06, + "loss": 0.4206, + "step": 19320 + }, + { + "epoch": 0.4252916574950473, + "grad_norm": 2.2430001379698514, + "learning_rate": 6.432384557808087e-06, + "loss": 0.3159, + "step": 19321 + }, + { + "epoch": 0.425313669381466, + "grad_norm": 1.2015157249691124, + "learning_rate": 6.432043037893256e-06, + "loss": 0.2203, + "step": 19322 + }, + { + "epoch": 0.4253356812678847, + "grad_norm": 2.38990729383829, + "learning_rate": 6.431701510700146e-06, + "loss": 0.2662, + "step": 19323 + }, + { + "epoch": 0.42535769315430333, + "grad_norm": 1.5761757646510441, + "learning_rate": 6.431359976230488e-06, + "loss": 0.3014, + "step": 19324 + }, + { + "epoch": 0.425379705040722, + "grad_norm": 0.678849886243363, + "learning_rate": 6.43101843448602e-06, + "loss": 0.4759, + "step": 19325 + }, + { + "epoch": 0.4254017169271407, + "grad_norm": 0.6477466279744288, + "learning_rate": 6.4306768854684795e-06, + "loss": 0.4961, + "step": 19326 + }, + { + "epoch": 0.42542372881355933, + "grad_norm": 0.6357778518681617, + "learning_rate": 6.4303353291796004e-06, + "loss": 0.4902, + "step": 19327 + }, + { + "epoch": 0.425445740699978, + "grad_norm": 1.5005101793344784, + "learning_rate": 6.429993765621119e-06, + "loss": 0.2549, + "step": 19328 + }, + { + "epoch": 0.42546775258639663, + "grad_norm": 1.3862949335729382, + "learning_rate": 6.429652194794772e-06, + "loss": 0.3374, + "step": 19329 + }, + { + "epoch": 0.42548976447281533, + "grad_norm": 1.552832408933061, + "learning_rate": 6.4293106167022945e-06, + "loss": 0.3361, + "step": 19330 + }, + { + "epoch": 0.425511776359234, + "grad_norm": 2.0716632527507275, + "learning_rate": 6.428969031345423e-06, + "loss": 0.3717, + "step": 19331 + }, + { + "epoch": 0.42553378824565263, + "grad_norm": 1.5524067990225634, + "learning_rate": 6.428627438725893e-06, + "loss": 0.3988, + "step": 19332 + }, + { + "epoch": 0.42555580013207134, + "grad_norm": 0.6618649138696654, + "learning_rate": 6.428285838845443e-06, + "loss": 0.5059, + "step": 19333 + }, + { + "epoch": 0.42557781201849, + "grad_norm": 1.570063155688164, + "learning_rate": 6.4279442317058045e-06, + "loss": 0.3361, + "step": 19334 + }, + { + "epoch": 0.42559982390490864, + "grad_norm": 1.689749964186298, + "learning_rate": 6.4276026173087195e-06, + "loss": 0.3612, + "step": 19335 + }, + { + "epoch": 0.42562183579132734, + "grad_norm": 0.6472866538436377, + "learning_rate": 6.42726099565592e-06, + "loss": 0.502, + "step": 19336 + }, + { + "epoch": 0.425643847677746, + "grad_norm": 1.597953277116423, + "learning_rate": 6.426919366749142e-06, + "loss": 0.4288, + "step": 19337 + }, + { + "epoch": 0.42566585956416464, + "grad_norm": 0.635061272324028, + "learning_rate": 6.426577730590125e-06, + "loss": 0.4974, + "step": 19338 + }, + { + "epoch": 0.4256878714505833, + "grad_norm": 3.4568625710836587, + "learning_rate": 6.426236087180603e-06, + "loss": 0.3806, + "step": 19339 + }, + { + "epoch": 0.425709883337002, + "grad_norm": 2.7127181456603298, + "learning_rate": 6.425894436522313e-06, + "loss": 0.2746, + "step": 19340 + }, + { + "epoch": 0.42573189522342064, + "grad_norm": 1.6356663317309268, + "learning_rate": 6.425552778616993e-06, + "loss": 0.2934, + "step": 19341 + }, + { + "epoch": 0.4257539071098393, + "grad_norm": 1.8863446879413563, + "learning_rate": 6.425211113466376e-06, + "loss": 0.3665, + "step": 19342 + }, + { + "epoch": 0.425775918996258, + "grad_norm": 1.322232936818312, + "learning_rate": 6.4248694410722025e-06, + "loss": 0.2969, + "step": 19343 + }, + { + "epoch": 0.42579793088267665, + "grad_norm": 1.4093115308448079, + "learning_rate": 6.424527761436207e-06, + "loss": 0.2804, + "step": 19344 + }, + { + "epoch": 0.4258199427690953, + "grad_norm": 1.503716204295908, + "learning_rate": 6.424186074560124e-06, + "loss": 0.4118, + "step": 19345 + }, + { + "epoch": 0.425841954655514, + "grad_norm": 1.7056360272802185, + "learning_rate": 6.4238443804456925e-06, + "loss": 0.3241, + "step": 19346 + }, + { + "epoch": 0.42586396654193265, + "grad_norm": 1.2982503473078475, + "learning_rate": 6.42350267909465e-06, + "loss": 0.3322, + "step": 19347 + }, + { + "epoch": 0.4258859784283513, + "grad_norm": 2.1005853917794046, + "learning_rate": 6.423160970508732e-06, + "loss": 0.3366, + "step": 19348 + }, + { + "epoch": 0.42590799031477, + "grad_norm": 1.3632577618813826, + "learning_rate": 6.422819254689674e-06, + "loss": 0.3154, + "step": 19349 + }, + { + "epoch": 0.42593000220118865, + "grad_norm": 1.341648843775997, + "learning_rate": 6.422477531639214e-06, + "loss": 0.326, + "step": 19350 + }, + { + "epoch": 0.4259520140876073, + "grad_norm": 1.6057265135962087, + "learning_rate": 6.42213580135909e-06, + "loss": 0.3571, + "step": 19351 + }, + { + "epoch": 0.42597402597402595, + "grad_norm": 1.9193240510803862, + "learning_rate": 6.421794063851037e-06, + "loss": 0.2923, + "step": 19352 + }, + { + "epoch": 0.42599603786044465, + "grad_norm": 1.2575753771690827, + "learning_rate": 6.421452319116792e-06, + "loss": 0.3343, + "step": 19353 + }, + { + "epoch": 0.4260180497468633, + "grad_norm": 1.5218486880275415, + "learning_rate": 6.421110567158093e-06, + "loss": 0.3971, + "step": 19354 + }, + { + "epoch": 0.42604006163328195, + "grad_norm": 1.3084078651522906, + "learning_rate": 6.420768807976675e-06, + "loss": 0.2924, + "step": 19355 + }, + { + "epoch": 0.42606207351970066, + "grad_norm": 0.7919010235699105, + "learning_rate": 6.420427041574279e-06, + "loss": 0.44, + "step": 19356 + }, + { + "epoch": 0.4260840854061193, + "grad_norm": 0.7780838024468364, + "learning_rate": 6.420085267952636e-06, + "loss": 0.4734, + "step": 19357 + }, + { + "epoch": 0.42610609729253796, + "grad_norm": 2.288131065526581, + "learning_rate": 6.419743487113488e-06, + "loss": 0.302, + "step": 19358 + }, + { + "epoch": 0.42612810917895666, + "grad_norm": 0.6176290195216717, + "learning_rate": 6.419401699058571e-06, + "loss": 0.4984, + "step": 19359 + }, + { + "epoch": 0.4261501210653753, + "grad_norm": 1.2819713367574261, + "learning_rate": 6.419059903789619e-06, + "loss": 0.3014, + "step": 19360 + }, + { + "epoch": 0.42617213295179396, + "grad_norm": 1.2019083956425813, + "learning_rate": 6.418718101308375e-06, + "loss": 0.3284, + "step": 19361 + }, + { + "epoch": 0.4261941448382126, + "grad_norm": 1.313937825350286, + "learning_rate": 6.4183762916165704e-06, + "loss": 0.3187, + "step": 19362 + }, + { + "epoch": 0.4262161567246313, + "grad_norm": 0.718532390906812, + "learning_rate": 6.418034474715945e-06, + "loss": 0.4585, + "step": 19363 + }, + { + "epoch": 0.42623816861104996, + "grad_norm": 1.553881041444106, + "learning_rate": 6.417692650608236e-06, + "loss": 0.3819, + "step": 19364 + }, + { + "epoch": 0.4262601804974686, + "grad_norm": 1.2443551640224766, + "learning_rate": 6.417350819295182e-06, + "loss": 0.3197, + "step": 19365 + }, + { + "epoch": 0.4262821923838873, + "grad_norm": 1.6991059906283794, + "learning_rate": 6.417008980778519e-06, + "loss": 0.3811, + "step": 19366 + }, + { + "epoch": 0.42630420427030596, + "grad_norm": 1.5057150810492235, + "learning_rate": 6.416667135059983e-06, + "loss": 0.468, + "step": 19367 + }, + { + "epoch": 0.4263262161567246, + "grad_norm": 1.392840484551293, + "learning_rate": 6.416325282141312e-06, + "loss": 0.311, + "step": 19368 + }, + { + "epoch": 0.4263482280431433, + "grad_norm": 1.457988893437442, + "learning_rate": 6.415983422024246e-06, + "loss": 0.242, + "step": 19369 + }, + { + "epoch": 0.42637023992956197, + "grad_norm": 2.212685737680946, + "learning_rate": 6.41564155471052e-06, + "loss": 0.2733, + "step": 19370 + }, + { + "epoch": 0.4263922518159806, + "grad_norm": 1.5398366689680272, + "learning_rate": 6.415299680201873e-06, + "loss": 0.3647, + "step": 19371 + }, + { + "epoch": 0.4264142637023993, + "grad_norm": 1.6685603271467098, + "learning_rate": 6.414957798500041e-06, + "loss": 0.3377, + "step": 19372 + }, + { + "epoch": 0.42643627558881797, + "grad_norm": 1.4980416106157968, + "learning_rate": 6.4146159096067614e-06, + "loss": 0.3558, + "step": 19373 + }, + { + "epoch": 0.4264582874752366, + "grad_norm": 1.4054078117284476, + "learning_rate": 6.414274013523776e-06, + "loss": 0.3466, + "step": 19374 + }, + { + "epoch": 0.42648029936165527, + "grad_norm": 1.5017507218284807, + "learning_rate": 6.413932110252816e-06, + "loss": 0.3873, + "step": 19375 + }, + { + "epoch": 0.426502311248074, + "grad_norm": 1.7640817650053278, + "learning_rate": 6.413590199795624e-06, + "loss": 0.3041, + "step": 19376 + }, + { + "epoch": 0.4265243231344926, + "grad_norm": 1.7710573304999748, + "learning_rate": 6.413248282153936e-06, + "loss": 0.2584, + "step": 19377 + }, + { + "epoch": 0.42654633502091127, + "grad_norm": 1.198527812096084, + "learning_rate": 6.412906357329488e-06, + "loss": 0.2537, + "step": 19378 + }, + { + "epoch": 0.42656834690733, + "grad_norm": 1.7047214526077101, + "learning_rate": 6.4125644253240225e-06, + "loss": 0.3102, + "step": 19379 + }, + { + "epoch": 0.4265903587937486, + "grad_norm": 0.6981831659143511, + "learning_rate": 6.412222486139275e-06, + "loss": 0.505, + "step": 19380 + }, + { + "epoch": 0.4266123706801673, + "grad_norm": 1.7466733461652446, + "learning_rate": 6.41188053977698e-06, + "loss": 0.3913, + "step": 19381 + }, + { + "epoch": 0.426634382566586, + "grad_norm": 1.4738404874106872, + "learning_rate": 6.411538586238879e-06, + "loss": 0.349, + "step": 19382 + }, + { + "epoch": 0.42665639445300463, + "grad_norm": 1.5252632208148005, + "learning_rate": 6.411196625526711e-06, + "loss": 0.2885, + "step": 19383 + }, + { + "epoch": 0.4266784063394233, + "grad_norm": 1.5346497281722242, + "learning_rate": 6.410854657642211e-06, + "loss": 0.3091, + "step": 19384 + }, + { + "epoch": 0.426700418225842, + "grad_norm": 0.6046577939081994, + "learning_rate": 6.41051268258712e-06, + "loss": 0.4795, + "step": 19385 + }, + { + "epoch": 0.42672243011226063, + "grad_norm": 0.6032105879152359, + "learning_rate": 6.410170700363172e-06, + "loss": 0.4666, + "step": 19386 + }, + { + "epoch": 0.4267444419986793, + "grad_norm": 0.59927983608745, + "learning_rate": 6.409828710972109e-06, + "loss": 0.484, + "step": 19387 + }, + { + "epoch": 0.42676645388509793, + "grad_norm": 1.541520421756784, + "learning_rate": 6.409486714415667e-06, + "loss": 0.4249, + "step": 19388 + }, + { + "epoch": 0.42678846577151663, + "grad_norm": 1.612455546276226, + "learning_rate": 6.409144710695586e-06, + "loss": 0.3705, + "step": 19389 + }, + { + "epoch": 0.4268104776579353, + "grad_norm": 1.439018860102607, + "learning_rate": 6.408802699813603e-06, + "loss": 0.326, + "step": 19390 + }, + { + "epoch": 0.42683248954435393, + "grad_norm": 0.6185383033309334, + "learning_rate": 6.4084606817714544e-06, + "loss": 0.4676, + "step": 19391 + }, + { + "epoch": 0.42685450143077264, + "grad_norm": 0.6583278843273507, + "learning_rate": 6.408118656570882e-06, + "loss": 0.4481, + "step": 19392 + }, + { + "epoch": 0.4268765133171913, + "grad_norm": 1.4489865350060585, + "learning_rate": 6.407776624213623e-06, + "loss": 0.3665, + "step": 19393 + }, + { + "epoch": 0.42689852520360994, + "grad_norm": 1.2790333402510174, + "learning_rate": 6.407434584701415e-06, + "loss": 0.3372, + "step": 19394 + }, + { + "epoch": 0.42692053709002864, + "grad_norm": 2.9138338739485232, + "learning_rate": 6.407092538035996e-06, + "loss": 0.3742, + "step": 19395 + }, + { + "epoch": 0.4269425489764473, + "grad_norm": 1.7246068654904556, + "learning_rate": 6.406750484219104e-06, + "loss": 0.2833, + "step": 19396 + }, + { + "epoch": 0.42696456086286594, + "grad_norm": 0.6195548857141552, + "learning_rate": 6.406408423252482e-06, + "loss": 0.4925, + "step": 19397 + }, + { + "epoch": 0.4269865727492846, + "grad_norm": 1.74628246255048, + "learning_rate": 6.4060663551378635e-06, + "loss": 0.2395, + "step": 19398 + }, + { + "epoch": 0.4270085846357033, + "grad_norm": 0.61522176449492, + "learning_rate": 6.405724279876988e-06, + "loss": 0.4739, + "step": 19399 + }, + { + "epoch": 0.42703059652212194, + "grad_norm": 2.096869147160015, + "learning_rate": 6.405382197471594e-06, + "loss": 0.2916, + "step": 19400 + }, + { + "epoch": 0.4270526084085406, + "grad_norm": 1.4350439826123331, + "learning_rate": 6.405040107923424e-06, + "loss": 0.3296, + "step": 19401 + }, + { + "epoch": 0.4270746202949593, + "grad_norm": 2.042837932334403, + "learning_rate": 6.404698011234211e-06, + "loss": 0.3824, + "step": 19402 + }, + { + "epoch": 0.42709663218137794, + "grad_norm": 1.5195430575204747, + "learning_rate": 6.404355907405697e-06, + "loss": 0.3203, + "step": 19403 + }, + { + "epoch": 0.4271186440677966, + "grad_norm": 1.3456207522792643, + "learning_rate": 6.404013796439619e-06, + "loss": 0.2617, + "step": 19404 + }, + { + "epoch": 0.4271406559542153, + "grad_norm": 1.6449911361256981, + "learning_rate": 6.403671678337717e-06, + "loss": 0.3431, + "step": 19405 + }, + { + "epoch": 0.42716266784063395, + "grad_norm": 1.3839128472354394, + "learning_rate": 6.40332955310173e-06, + "loss": 0.3809, + "step": 19406 + }, + { + "epoch": 0.4271846797270526, + "grad_norm": 1.2911224135521988, + "learning_rate": 6.402987420733397e-06, + "loss": 0.3346, + "step": 19407 + }, + { + "epoch": 0.4272066916134713, + "grad_norm": 0.6478040859845425, + "learning_rate": 6.402645281234454e-06, + "loss": 0.4933, + "step": 19408 + }, + { + "epoch": 0.42722870349988995, + "grad_norm": 1.4080279040217634, + "learning_rate": 6.4023031346066435e-06, + "loss": 0.3115, + "step": 19409 + }, + { + "epoch": 0.4272507153863086, + "grad_norm": 1.4721783685505612, + "learning_rate": 6.401960980851703e-06, + "loss": 0.3213, + "step": 19410 + }, + { + "epoch": 0.42727272727272725, + "grad_norm": 1.8468196652198858, + "learning_rate": 6.401618819971371e-06, + "loss": 0.335, + "step": 19411 + }, + { + "epoch": 0.42729473915914595, + "grad_norm": 1.4041067893834325, + "learning_rate": 6.401276651967387e-06, + "loss": 0.367, + "step": 19412 + }, + { + "epoch": 0.4273167510455646, + "grad_norm": 1.740654920281043, + "learning_rate": 6.400934476841492e-06, + "loss": 0.3431, + "step": 19413 + }, + { + "epoch": 0.42733876293198325, + "grad_norm": 2.851354007014178, + "learning_rate": 6.4005922945954205e-06, + "loss": 0.2718, + "step": 19414 + }, + { + "epoch": 0.42736077481840196, + "grad_norm": 1.4641133325236704, + "learning_rate": 6.4002501052309165e-06, + "loss": 0.3229, + "step": 19415 + }, + { + "epoch": 0.4273827867048206, + "grad_norm": 1.4863280650847834, + "learning_rate": 6.399907908749716e-06, + "loss": 0.3958, + "step": 19416 + }, + { + "epoch": 0.42740479859123925, + "grad_norm": 1.1902289861404218, + "learning_rate": 6.399565705153558e-06, + "loss": 0.2939, + "step": 19417 + }, + { + "epoch": 0.42742681047765796, + "grad_norm": 1.4457239885327522, + "learning_rate": 6.399223494444184e-06, + "loss": 0.3064, + "step": 19418 + }, + { + "epoch": 0.4274488223640766, + "grad_norm": 1.701888708244715, + "learning_rate": 6.3988812766233325e-06, + "loss": 0.3495, + "step": 19419 + }, + { + "epoch": 0.42747083425049526, + "grad_norm": 1.3807192991780401, + "learning_rate": 6.398539051692742e-06, + "loss": 0.3635, + "step": 19420 + }, + { + "epoch": 0.4274928461369139, + "grad_norm": 0.6286587644058638, + "learning_rate": 6.398196819654152e-06, + "loss": 0.5007, + "step": 19421 + }, + { + "epoch": 0.4275148580233326, + "grad_norm": 2.0502451935227883, + "learning_rate": 6.397854580509302e-06, + "loss": 0.3794, + "step": 19422 + }, + { + "epoch": 0.42753686990975126, + "grad_norm": 1.0053385592102366, + "learning_rate": 6.397512334259931e-06, + "loss": 0.2834, + "step": 19423 + }, + { + "epoch": 0.4275588817961699, + "grad_norm": 1.5595697317468613, + "learning_rate": 6.39717008090778e-06, + "loss": 0.3082, + "step": 19424 + }, + { + "epoch": 0.4275808936825886, + "grad_norm": 1.527178999941504, + "learning_rate": 6.396827820454587e-06, + "loss": 0.3227, + "step": 19425 + }, + { + "epoch": 0.42760290556900726, + "grad_norm": 0.6506288034835869, + "learning_rate": 6.396485552902093e-06, + "loss": 0.4969, + "step": 19426 + }, + { + "epoch": 0.4276249174554259, + "grad_norm": 1.82087550641441, + "learning_rate": 6.3961432782520365e-06, + "loss": 0.2863, + "step": 19427 + }, + { + "epoch": 0.4276469293418446, + "grad_norm": 1.1963171334739757, + "learning_rate": 6.395800996506156e-06, + "loss": 0.2973, + "step": 19428 + }, + { + "epoch": 0.42766894122826327, + "grad_norm": 1.6863417317642118, + "learning_rate": 6.395458707666193e-06, + "loss": 0.384, + "step": 19429 + }, + { + "epoch": 0.4276909531146819, + "grad_norm": 1.2406764872142877, + "learning_rate": 6.395116411733886e-06, + "loss": 0.2851, + "step": 19430 + }, + { + "epoch": 0.4277129650011006, + "grad_norm": 1.3057746953090108, + "learning_rate": 6.3947741087109764e-06, + "loss": 0.332, + "step": 19431 + }, + { + "epoch": 0.42773497688751927, + "grad_norm": 0.6240066343261049, + "learning_rate": 6.394431798599201e-06, + "loss": 0.479, + "step": 19432 + }, + { + "epoch": 0.4277569887739379, + "grad_norm": 1.6012683661288631, + "learning_rate": 6.394089481400303e-06, + "loss": 0.2763, + "step": 19433 + }, + { + "epoch": 0.42777900066035657, + "grad_norm": 1.8923384007916455, + "learning_rate": 6.393747157116019e-06, + "loss": 0.3099, + "step": 19434 + }, + { + "epoch": 0.42780101254677527, + "grad_norm": 2.776731100243902, + "learning_rate": 6.393404825748091e-06, + "loss": 0.3172, + "step": 19435 + }, + { + "epoch": 0.4278230244331939, + "grad_norm": 1.1966115959036752, + "learning_rate": 6.393062487298258e-06, + "loss": 0.2579, + "step": 19436 + }, + { + "epoch": 0.42784503631961257, + "grad_norm": 1.5396153260906835, + "learning_rate": 6.3927201417682605e-06, + "loss": 0.3138, + "step": 19437 + }, + { + "epoch": 0.4278670482060313, + "grad_norm": 0.6024072472522161, + "learning_rate": 6.392377789159838e-06, + "loss": 0.4796, + "step": 19438 + }, + { + "epoch": 0.4278890600924499, + "grad_norm": 1.4429849913125947, + "learning_rate": 6.39203542947473e-06, + "loss": 0.3512, + "step": 19439 + }, + { + "epoch": 0.4279110719788686, + "grad_norm": 1.524209718471798, + "learning_rate": 6.391693062714677e-06, + "loss": 0.2622, + "step": 19440 + }, + { + "epoch": 0.4279330838652873, + "grad_norm": 1.4583844281956964, + "learning_rate": 6.391350688881419e-06, + "loss": 0.3011, + "step": 19441 + }, + { + "epoch": 0.4279550957517059, + "grad_norm": 0.6268372398102646, + "learning_rate": 6.391008307976697e-06, + "loss": 0.4878, + "step": 19442 + }, + { + "epoch": 0.4279771076381246, + "grad_norm": 1.6702675674900382, + "learning_rate": 6.39066592000225e-06, + "loss": 0.3057, + "step": 19443 + }, + { + "epoch": 0.4279991195245432, + "grad_norm": 1.8250402593550292, + "learning_rate": 6.3903235249598194e-06, + "loss": 0.3659, + "step": 19444 + }, + { + "epoch": 0.42802113141096193, + "grad_norm": 0.6094321800364428, + "learning_rate": 6.389981122851143e-06, + "loss": 0.4935, + "step": 19445 + }, + { + "epoch": 0.4280431432973806, + "grad_norm": 0.6250080894168588, + "learning_rate": 6.389638713677963e-06, + "loss": 0.4832, + "step": 19446 + }, + { + "epoch": 0.42806515518379923, + "grad_norm": 2.7407261962502107, + "learning_rate": 6.38929629744202e-06, + "loss": 0.4182, + "step": 19447 + }, + { + "epoch": 0.42808716707021793, + "grad_norm": 1.6421234286326531, + "learning_rate": 6.388953874145054e-06, + "loss": 0.3539, + "step": 19448 + }, + { + "epoch": 0.4281091789566366, + "grad_norm": 0.655832891872347, + "learning_rate": 6.388611443788805e-06, + "loss": 0.524, + "step": 19449 + }, + { + "epoch": 0.42813119084305523, + "grad_norm": 1.8228733158290136, + "learning_rate": 6.388269006375011e-06, + "loss": 0.3535, + "step": 19450 + }, + { + "epoch": 0.42815320272947394, + "grad_norm": 1.3032482649957515, + "learning_rate": 6.387926561905417e-06, + "loss": 0.306, + "step": 19451 + }, + { + "epoch": 0.4281752146158926, + "grad_norm": 1.5938583069406158, + "learning_rate": 6.387584110381762e-06, + "loss": 0.3126, + "step": 19452 + }, + { + "epoch": 0.42819722650231123, + "grad_norm": 1.4614576876125107, + "learning_rate": 6.3872416518057825e-06, + "loss": 0.2414, + "step": 19453 + }, + { + "epoch": 0.42821923838872994, + "grad_norm": 0.6190811416903891, + "learning_rate": 6.386899186179224e-06, + "loss": 0.4679, + "step": 19454 + }, + { + "epoch": 0.4282412502751486, + "grad_norm": 1.1886214758318623, + "learning_rate": 6.3865567135038245e-06, + "loss": 0.2939, + "step": 19455 + }, + { + "epoch": 0.42826326216156724, + "grad_norm": 1.6244564312699454, + "learning_rate": 6.386214233781326e-06, + "loss": 0.3179, + "step": 19456 + }, + { + "epoch": 0.4282852740479859, + "grad_norm": 1.5311716846869246, + "learning_rate": 6.385871747013469e-06, + "loss": 0.3555, + "step": 19457 + }, + { + "epoch": 0.4283072859344046, + "grad_norm": 1.0802820768072987, + "learning_rate": 6.385529253201992e-06, + "loss": 0.2447, + "step": 19458 + }, + { + "epoch": 0.42832929782082324, + "grad_norm": 1.306871276013138, + "learning_rate": 6.3851867523486384e-06, + "loss": 0.3475, + "step": 19459 + }, + { + "epoch": 0.4283513097072419, + "grad_norm": 0.6265130569717691, + "learning_rate": 6.384844244455148e-06, + "loss": 0.4971, + "step": 19460 + }, + { + "epoch": 0.4283733215936606, + "grad_norm": 0.6371806510484302, + "learning_rate": 6.3845017295232625e-06, + "loss": 0.5008, + "step": 19461 + }, + { + "epoch": 0.42839533348007924, + "grad_norm": 0.6199891422343915, + "learning_rate": 6.38415920755472e-06, + "loss": 0.469, + "step": 19462 + }, + { + "epoch": 0.4284173453664979, + "grad_norm": 1.8076528294719127, + "learning_rate": 6.383816678551264e-06, + "loss": 0.3454, + "step": 19463 + }, + { + "epoch": 0.4284393572529166, + "grad_norm": 1.2483541911809786, + "learning_rate": 6.383474142514633e-06, + "loss": 0.3456, + "step": 19464 + }, + { + "epoch": 0.42846136913933525, + "grad_norm": 1.1615465351741217, + "learning_rate": 6.383131599446573e-06, + "loss": 0.2772, + "step": 19465 + }, + { + "epoch": 0.4284833810257539, + "grad_norm": 1.4716336161933756, + "learning_rate": 6.382789049348816e-06, + "loss": 0.3205, + "step": 19466 + }, + { + "epoch": 0.4285053929121726, + "grad_norm": 1.3602198271542518, + "learning_rate": 6.3824464922231125e-06, + "loss": 0.3746, + "step": 19467 + }, + { + "epoch": 0.42852740479859125, + "grad_norm": 2.332063864722657, + "learning_rate": 6.3821039280711975e-06, + "loss": 0.3339, + "step": 19468 + }, + { + "epoch": 0.4285494166850099, + "grad_norm": 1.4755622778564814, + "learning_rate": 6.381761356894814e-06, + "loss": 0.3713, + "step": 19469 + }, + { + "epoch": 0.42857142857142855, + "grad_norm": 1.5699589326810532, + "learning_rate": 6.381418778695705e-06, + "loss": 0.3906, + "step": 19470 + }, + { + "epoch": 0.42859344045784725, + "grad_norm": 1.7674946571141708, + "learning_rate": 6.381076193475606e-06, + "loss": 0.4478, + "step": 19471 + }, + { + "epoch": 0.4286154523442659, + "grad_norm": 0.6458519828477994, + "learning_rate": 6.3807336012362646e-06, + "loss": 0.5, + "step": 19472 + }, + { + "epoch": 0.42863746423068455, + "grad_norm": 1.509968936327042, + "learning_rate": 6.380391001979418e-06, + "loss": 0.3823, + "step": 19473 + }, + { + "epoch": 0.42865947611710326, + "grad_norm": 1.376659134389442, + "learning_rate": 6.380048395706809e-06, + "loss": 0.2948, + "step": 19474 + }, + { + "epoch": 0.4286814880035219, + "grad_norm": 1.235612585632573, + "learning_rate": 6.379705782420179e-06, + "loss": 0.3067, + "step": 19475 + }, + { + "epoch": 0.42870349988994055, + "grad_norm": 1.900834278252256, + "learning_rate": 6.379363162121268e-06, + "loss": 0.3592, + "step": 19476 + }, + { + "epoch": 0.42872551177635926, + "grad_norm": 1.4550844470329227, + "learning_rate": 6.37902053481182e-06, + "loss": 0.3163, + "step": 19477 + }, + { + "epoch": 0.4287475236627779, + "grad_norm": 1.7398899728417494, + "learning_rate": 6.378677900493573e-06, + "loss": 0.3467, + "step": 19478 + }, + { + "epoch": 0.42876953554919656, + "grad_norm": 0.6269980373487238, + "learning_rate": 6.378335259168269e-06, + "loss": 0.485, + "step": 19479 + }, + { + "epoch": 0.4287915474356152, + "grad_norm": 1.2834594941908408, + "learning_rate": 6.377992610837652e-06, + "loss": 0.3127, + "step": 19480 + }, + { + "epoch": 0.4288135593220339, + "grad_norm": 1.2570833688202816, + "learning_rate": 6.377649955503461e-06, + "loss": 0.3359, + "step": 19481 + }, + { + "epoch": 0.42883557120845256, + "grad_norm": 0.6101802530982321, + "learning_rate": 6.37730729316744e-06, + "loss": 0.4733, + "step": 19482 + }, + { + "epoch": 0.4288575830948712, + "grad_norm": 1.2664638759954339, + "learning_rate": 6.376964623831329e-06, + "loss": 0.2997, + "step": 19483 + }, + { + "epoch": 0.4288795949812899, + "grad_norm": 1.3566913862505963, + "learning_rate": 6.376621947496868e-06, + "loss": 0.2525, + "step": 19484 + }, + { + "epoch": 0.42890160686770856, + "grad_norm": 7.5036267623414385, + "learning_rate": 6.376279264165801e-06, + "loss": 0.3094, + "step": 19485 + }, + { + "epoch": 0.4289236187541272, + "grad_norm": 1.3855853721429527, + "learning_rate": 6.3759365738398685e-06, + "loss": 0.3478, + "step": 19486 + }, + { + "epoch": 0.4289456306405459, + "grad_norm": 1.390760639412376, + "learning_rate": 6.375593876520813e-06, + "loss": 0.3975, + "step": 19487 + }, + { + "epoch": 0.42896764252696457, + "grad_norm": 0.5996459267035835, + "learning_rate": 6.375251172210377e-06, + "loss": 0.4682, + "step": 19488 + }, + { + "epoch": 0.4289896544133832, + "grad_norm": 2.4626366910413338, + "learning_rate": 6.3749084609103e-06, + "loss": 0.4157, + "step": 19489 + }, + { + "epoch": 0.4290116662998019, + "grad_norm": 0.6266585319058323, + "learning_rate": 6.374565742622325e-06, + "loss": 0.5193, + "step": 19490 + }, + { + "epoch": 0.42903367818622057, + "grad_norm": 1.7921216903469002, + "learning_rate": 6.374223017348195e-06, + "loss": 0.3278, + "step": 19491 + }, + { + "epoch": 0.4290556900726392, + "grad_norm": 1.5344383261331662, + "learning_rate": 6.373880285089648e-06, + "loss": 0.3023, + "step": 19492 + }, + { + "epoch": 0.42907770195905787, + "grad_norm": 1.3884859406679482, + "learning_rate": 6.373537545848432e-06, + "loss": 0.3679, + "step": 19493 + }, + { + "epoch": 0.42909971384547657, + "grad_norm": 1.3991456908230169, + "learning_rate": 6.373194799626282e-06, + "loss": 0.287, + "step": 19494 + }, + { + "epoch": 0.4291217257318952, + "grad_norm": 1.3104812833921502, + "learning_rate": 6.372852046424947e-06, + "loss": 0.3123, + "step": 19495 + }, + { + "epoch": 0.42914373761831387, + "grad_norm": 1.348658162942093, + "learning_rate": 6.3725092862461634e-06, + "loss": 0.2149, + "step": 19496 + }, + { + "epoch": 0.4291657495047326, + "grad_norm": 1.276045542136213, + "learning_rate": 6.372166519091676e-06, + "loss": 0.3244, + "step": 19497 + }, + { + "epoch": 0.4291877613911512, + "grad_norm": 1.2089717168095242, + "learning_rate": 6.371823744963227e-06, + "loss": 0.3216, + "step": 19498 + }, + { + "epoch": 0.4292097732775699, + "grad_norm": 1.5867827825098564, + "learning_rate": 6.371480963862556e-06, + "loss": 0.3999, + "step": 19499 + }, + { + "epoch": 0.4292317851639886, + "grad_norm": 1.5818452968406715, + "learning_rate": 6.371138175791409e-06, + "loss": 0.3443, + "step": 19500 + }, + { + "epoch": 0.4292537970504072, + "grad_norm": 0.6362436153757217, + "learning_rate": 6.3707953807515256e-06, + "loss": 0.4866, + "step": 19501 + }, + { + "epoch": 0.4292758089368259, + "grad_norm": 2.457636949980245, + "learning_rate": 6.3704525787446495e-06, + "loss": 0.3022, + "step": 19502 + }, + { + "epoch": 0.4292978208232445, + "grad_norm": 0.6246904899342495, + "learning_rate": 6.3701097697725215e-06, + "loss": 0.4577, + "step": 19503 + }, + { + "epoch": 0.42931983270966323, + "grad_norm": 1.1913069148808144, + "learning_rate": 6.369766953836883e-06, + "loss": 0.322, + "step": 19504 + }, + { + "epoch": 0.4293418445960819, + "grad_norm": 2.317047014974573, + "learning_rate": 6.36942413093948e-06, + "loss": 0.3382, + "step": 19505 + }, + { + "epoch": 0.4293638564825005, + "grad_norm": 4.50485007506083, + "learning_rate": 6.369081301082053e-06, + "loss": 0.384, + "step": 19506 + }, + { + "epoch": 0.42938586836891923, + "grad_norm": 1.3486086456109854, + "learning_rate": 6.368738464266342e-06, + "loss": 0.3154, + "step": 19507 + }, + { + "epoch": 0.4294078802553379, + "grad_norm": 5.430424909181947, + "learning_rate": 6.3683956204940935e-06, + "loss": 0.4412, + "step": 19508 + }, + { + "epoch": 0.42942989214175653, + "grad_norm": 1.8827960559751615, + "learning_rate": 6.368052769767048e-06, + "loss": 0.2661, + "step": 19509 + }, + { + "epoch": 0.42945190402817524, + "grad_norm": 0.6124521096556665, + "learning_rate": 6.367709912086947e-06, + "loss": 0.4805, + "step": 19510 + }, + { + "epoch": 0.4294739159145939, + "grad_norm": 1.6014419967154279, + "learning_rate": 6.367367047455536e-06, + "loss": 0.3482, + "step": 19511 + }, + { + "epoch": 0.42949592780101253, + "grad_norm": 1.4901835290870589, + "learning_rate": 6.367024175874554e-06, + "loss": 0.3739, + "step": 19512 + }, + { + "epoch": 0.42951793968743124, + "grad_norm": 1.7092706270298967, + "learning_rate": 6.366681297345746e-06, + "loss": 0.3297, + "step": 19513 + }, + { + "epoch": 0.4295399515738499, + "grad_norm": 1.500891077125655, + "learning_rate": 6.366338411870855e-06, + "loss": 0.2876, + "step": 19514 + }, + { + "epoch": 0.42956196346026854, + "grad_norm": 2.1529864743565543, + "learning_rate": 6.365995519451622e-06, + "loss": 0.3582, + "step": 19515 + }, + { + "epoch": 0.4295839753466872, + "grad_norm": 2.039368090323751, + "learning_rate": 6.365652620089791e-06, + "loss": 0.4268, + "step": 19516 + }, + { + "epoch": 0.4296059872331059, + "grad_norm": 0.6179042708642135, + "learning_rate": 6.365309713787103e-06, + "loss": 0.4892, + "step": 19517 + }, + { + "epoch": 0.42962799911952454, + "grad_norm": 1.6331245765327949, + "learning_rate": 6.364966800545303e-06, + "loss": 0.3469, + "step": 19518 + }, + { + "epoch": 0.4296500110059432, + "grad_norm": 1.2610469563899822, + "learning_rate": 6.364623880366133e-06, + "loss": 0.2874, + "step": 19519 + }, + { + "epoch": 0.4296720228923619, + "grad_norm": 1.4260172312662325, + "learning_rate": 6.364280953251336e-06, + "loss": 0.2889, + "step": 19520 + }, + { + "epoch": 0.42969403477878054, + "grad_norm": 0.61612626401132, + "learning_rate": 6.3639380192026545e-06, + "loss": 0.5037, + "step": 19521 + }, + { + "epoch": 0.4297160466651992, + "grad_norm": 1.4055652921000628, + "learning_rate": 6.363595078221832e-06, + "loss": 0.326, + "step": 19522 + }, + { + "epoch": 0.4297380585516179, + "grad_norm": 1.510236903500654, + "learning_rate": 6.36325213031061e-06, + "loss": 0.2364, + "step": 19523 + }, + { + "epoch": 0.42976007043803655, + "grad_norm": 1.166314270478565, + "learning_rate": 6.362909175470734e-06, + "loss": 0.2671, + "step": 19524 + }, + { + "epoch": 0.4297820823244552, + "grad_norm": 1.6984197102828542, + "learning_rate": 6.3625662137039445e-06, + "loss": 0.2743, + "step": 19525 + }, + { + "epoch": 0.4298040942108739, + "grad_norm": 1.726797313387947, + "learning_rate": 6.362223245011986e-06, + "loss": 0.3721, + "step": 19526 + }, + { + "epoch": 0.42982610609729255, + "grad_norm": 1.2978349033904568, + "learning_rate": 6.361880269396602e-06, + "loss": 0.276, + "step": 19527 + }, + { + "epoch": 0.4298481179837112, + "grad_norm": 1.6572112031154986, + "learning_rate": 6.361537286859533e-06, + "loss": 0.2966, + "step": 19528 + }, + { + "epoch": 0.42987012987012985, + "grad_norm": 1.2626684740774778, + "learning_rate": 6.361194297402529e-06, + "loss": 0.2619, + "step": 19529 + }, + { + "epoch": 0.42989214175654855, + "grad_norm": 1.3206678713003879, + "learning_rate": 6.360851301027324e-06, + "loss": 0.3049, + "step": 19530 + }, + { + "epoch": 0.4299141536429672, + "grad_norm": 0.6443143903046126, + "learning_rate": 6.360508297735667e-06, + "loss": 0.5007, + "step": 19531 + }, + { + "epoch": 0.42993616552938585, + "grad_norm": 1.4479082763081526, + "learning_rate": 6.3601652875293e-06, + "loss": 0.3623, + "step": 19532 + }, + { + "epoch": 0.42995817741580455, + "grad_norm": 1.757969419416614, + "learning_rate": 6.3598222704099675e-06, + "loss": 0.3614, + "step": 19533 + }, + { + "epoch": 0.4299801893022232, + "grad_norm": 2.7884381820715736, + "learning_rate": 6.359479246379411e-06, + "loss": 0.3746, + "step": 19534 + }, + { + "epoch": 0.43000220118864185, + "grad_norm": 2.4781137033855396, + "learning_rate": 6.359136215439374e-06, + "loss": 0.3068, + "step": 19535 + }, + { + "epoch": 0.43002421307506056, + "grad_norm": 1.7059752201355172, + "learning_rate": 6.358793177591601e-06, + "loss": 0.2913, + "step": 19536 + }, + { + "epoch": 0.4300462249614792, + "grad_norm": 1.2751600802232277, + "learning_rate": 6.358450132837834e-06, + "loss": 0.2515, + "step": 19537 + }, + { + "epoch": 0.43006823684789786, + "grad_norm": 2.4767101505714426, + "learning_rate": 6.35810708117982e-06, + "loss": 0.3846, + "step": 19538 + }, + { + "epoch": 0.4300902487343165, + "grad_norm": 1.5068344570433234, + "learning_rate": 6.357764022619299e-06, + "loss": 0.3419, + "step": 19539 + }, + { + "epoch": 0.4301122606207352, + "grad_norm": 2.022710417116615, + "learning_rate": 6.357420957158016e-06, + "loss": 0.3309, + "step": 19540 + }, + { + "epoch": 0.43013427250715386, + "grad_norm": 1.4726191055332616, + "learning_rate": 6.357077884797712e-06, + "loss": 0.3909, + "step": 19541 + }, + { + "epoch": 0.4301562843935725, + "grad_norm": 1.133778800661828, + "learning_rate": 6.356734805540135e-06, + "loss": 0.3177, + "step": 19542 + }, + { + "epoch": 0.4301782962799912, + "grad_norm": 0.6523516221709916, + "learning_rate": 6.356391719387026e-06, + "loss": 0.5009, + "step": 19543 + }, + { + "epoch": 0.43020030816640986, + "grad_norm": 0.6489698102694009, + "learning_rate": 6.35604862634013e-06, + "loss": 0.4919, + "step": 19544 + }, + { + "epoch": 0.4302223200528285, + "grad_norm": 1.346559779216782, + "learning_rate": 6.3557055264011894e-06, + "loss": 0.3554, + "step": 19545 + }, + { + "epoch": 0.4302443319392472, + "grad_norm": 1.9839819299852532, + "learning_rate": 6.3553624195719475e-06, + "loss": 0.3797, + "step": 19546 + }, + { + "epoch": 0.43026634382566586, + "grad_norm": 1.921618649063251, + "learning_rate": 6.3550193058541524e-06, + "loss": 0.3026, + "step": 19547 + }, + { + "epoch": 0.4302883557120845, + "grad_norm": 0.6070994494607429, + "learning_rate": 6.354676185249542e-06, + "loss": 0.493, + "step": 19548 + }, + { + "epoch": 0.4303103675985032, + "grad_norm": 1.2053877807268845, + "learning_rate": 6.354333057759864e-06, + "loss": 0.3344, + "step": 19549 + }, + { + "epoch": 0.43033237948492187, + "grad_norm": 1.711397651654046, + "learning_rate": 6.353989923386861e-06, + "loss": 0.2911, + "step": 19550 + }, + { + "epoch": 0.4303543913713405, + "grad_norm": 2.1646250508155807, + "learning_rate": 6.353646782132277e-06, + "loss": 0.2992, + "step": 19551 + }, + { + "epoch": 0.43037640325775917, + "grad_norm": 1.6562268887876541, + "learning_rate": 6.353303633997858e-06, + "loss": 0.3802, + "step": 19552 + }, + { + "epoch": 0.43039841514417787, + "grad_norm": 1.5162877116853395, + "learning_rate": 6.352960478985345e-06, + "loss": 0.3512, + "step": 19553 + }, + { + "epoch": 0.4304204270305965, + "grad_norm": 1.3481428086306033, + "learning_rate": 6.352617317096483e-06, + "loss": 0.3047, + "step": 19554 + }, + { + "epoch": 0.43044243891701517, + "grad_norm": 1.6756734777877071, + "learning_rate": 6.352274148333018e-06, + "loss": 0.3801, + "step": 19555 + }, + { + "epoch": 0.4304644508034339, + "grad_norm": 1.3854195155320235, + "learning_rate": 6.351930972696692e-06, + "loss": 0.2781, + "step": 19556 + }, + { + "epoch": 0.4304864626898525, + "grad_norm": 2.0093599047246458, + "learning_rate": 6.3515877901892495e-06, + "loss": 0.3254, + "step": 19557 + }, + { + "epoch": 0.43050847457627117, + "grad_norm": 1.038846608392861, + "learning_rate": 6.351244600812436e-06, + "loss": 0.342, + "step": 19558 + }, + { + "epoch": 0.4305304864626899, + "grad_norm": 1.048550778478743, + "learning_rate": 6.350901404567993e-06, + "loss": 0.2508, + "step": 19559 + }, + { + "epoch": 0.4305524983491085, + "grad_norm": 1.8730116246162158, + "learning_rate": 6.350558201457667e-06, + "loss": 0.3729, + "step": 19560 + }, + { + "epoch": 0.4305745102355272, + "grad_norm": 1.6618100077660587, + "learning_rate": 6.350214991483203e-06, + "loss": 0.3957, + "step": 19561 + }, + { + "epoch": 0.4305965221219458, + "grad_norm": 0.680770390296805, + "learning_rate": 6.349871774646343e-06, + "loss": 0.5116, + "step": 19562 + }, + { + "epoch": 0.43061853400836453, + "grad_norm": 0.625895341563295, + "learning_rate": 6.349528550948834e-06, + "loss": 0.4913, + "step": 19563 + }, + { + "epoch": 0.4306405458947832, + "grad_norm": 1.5749996627487197, + "learning_rate": 6.349185320392417e-06, + "loss": 0.3321, + "step": 19564 + }, + { + "epoch": 0.4306625577812018, + "grad_norm": 1.8885999099798663, + "learning_rate": 6.348842082978841e-06, + "loss": 0.2912, + "step": 19565 + }, + { + "epoch": 0.43068456966762053, + "grad_norm": 1.4690960074309045, + "learning_rate": 6.3484988387098455e-06, + "loss": 0.3292, + "step": 19566 + }, + { + "epoch": 0.4307065815540392, + "grad_norm": 1.5666451522112197, + "learning_rate": 6.348155587587179e-06, + "loss": 0.3734, + "step": 19567 + }, + { + "epoch": 0.43072859344045783, + "grad_norm": 0.6293441147033222, + "learning_rate": 6.347812329612584e-06, + "loss": 0.4955, + "step": 19568 + }, + { + "epoch": 0.43075060532687653, + "grad_norm": 2.23706747955579, + "learning_rate": 6.347469064787804e-06, + "loss": 0.3081, + "step": 19569 + }, + { + "epoch": 0.4307726172132952, + "grad_norm": 1.682106183768024, + "learning_rate": 6.347125793114589e-06, + "loss": 0.3366, + "step": 19570 + }, + { + "epoch": 0.43079462909971383, + "grad_norm": 1.2327488188804978, + "learning_rate": 6.346782514594677e-06, + "loss": 0.2849, + "step": 19571 + }, + { + "epoch": 0.43081664098613254, + "grad_norm": 1.514456216058769, + "learning_rate": 6.346439229229816e-06, + "loss": 0.3265, + "step": 19572 + }, + { + "epoch": 0.4308386528725512, + "grad_norm": 1.9732362390525668, + "learning_rate": 6.346095937021751e-06, + "loss": 0.3179, + "step": 19573 + }, + { + "epoch": 0.43086066475896984, + "grad_norm": 0.6311094480133139, + "learning_rate": 6.345752637972226e-06, + "loss": 0.477, + "step": 19574 + }, + { + "epoch": 0.4308826766453885, + "grad_norm": 1.3313613715509112, + "learning_rate": 6.345409332082985e-06, + "loss": 0.3248, + "step": 19575 + }, + { + "epoch": 0.4309046885318072, + "grad_norm": 1.320024509534738, + "learning_rate": 6.345066019355774e-06, + "loss": 0.2377, + "step": 19576 + }, + { + "epoch": 0.43092670041822584, + "grad_norm": 1.8187980267447366, + "learning_rate": 6.344722699792338e-06, + "loss": 0.381, + "step": 19577 + }, + { + "epoch": 0.4309487123046445, + "grad_norm": 1.296905597572285, + "learning_rate": 6.344379373394421e-06, + "loss": 0.4804, + "step": 19578 + }, + { + "epoch": 0.4309707241910632, + "grad_norm": 0.6220157693296317, + "learning_rate": 6.344036040163769e-06, + "loss": 0.4911, + "step": 19579 + }, + { + "epoch": 0.43099273607748184, + "grad_norm": 1.837563137910567, + "learning_rate": 6.343692700102126e-06, + "loss": 0.3578, + "step": 19580 + }, + { + "epoch": 0.4310147479639005, + "grad_norm": 1.217814397082874, + "learning_rate": 6.343349353211239e-06, + "loss": 0.3618, + "step": 19581 + }, + { + "epoch": 0.4310367598503192, + "grad_norm": 1.4645609659456567, + "learning_rate": 6.343005999492848e-06, + "loss": 0.2895, + "step": 19582 + }, + { + "epoch": 0.43105877173673784, + "grad_norm": 0.6139151682517208, + "learning_rate": 6.3426626389487045e-06, + "loss": 0.4898, + "step": 19583 + }, + { + "epoch": 0.4310807836231565, + "grad_norm": 1.4596369961179725, + "learning_rate": 6.342319271580549e-06, + "loss": 0.3074, + "step": 19584 + }, + { + "epoch": 0.43110279550957514, + "grad_norm": 3.337271778202112, + "learning_rate": 6.341975897390129e-06, + "loss": 0.3548, + "step": 19585 + }, + { + "epoch": 0.43112480739599385, + "grad_norm": 2.2293473053987953, + "learning_rate": 6.341632516379189e-06, + "loss": 0.3315, + "step": 19586 + }, + { + "epoch": 0.4311468192824125, + "grad_norm": 0.6265192759530225, + "learning_rate": 6.341289128549473e-06, + "loss": 0.4842, + "step": 19587 + }, + { + "epoch": 0.43116883116883115, + "grad_norm": 1.2595221600303719, + "learning_rate": 6.34094573390273e-06, + "loss": 0.3299, + "step": 19588 + }, + { + "epoch": 0.43119084305524985, + "grad_norm": 0.6004353910538226, + "learning_rate": 6.3406023324407e-06, + "loss": 0.4854, + "step": 19589 + }, + { + "epoch": 0.4312128549416685, + "grad_norm": 1.9502731684140215, + "learning_rate": 6.340258924165132e-06, + "loss": 0.3393, + "step": 19590 + }, + { + "epoch": 0.43123486682808715, + "grad_norm": 0.6052583915249279, + "learning_rate": 6.339915509077771e-06, + "loss": 0.4873, + "step": 19591 + }, + { + "epoch": 0.43125687871450585, + "grad_norm": 1.1480550298058272, + "learning_rate": 6.33957208718036e-06, + "loss": 0.3402, + "step": 19592 + }, + { + "epoch": 0.4312788906009245, + "grad_norm": 0.6108613480440439, + "learning_rate": 6.339228658474649e-06, + "loss": 0.4894, + "step": 19593 + }, + { + "epoch": 0.43130090248734315, + "grad_norm": 1.8755243671448039, + "learning_rate": 6.338885222962379e-06, + "loss": 0.3862, + "step": 19594 + }, + { + "epoch": 0.43132291437376186, + "grad_norm": 1.0656586980059595, + "learning_rate": 6.338541780645297e-06, + "loss": 0.2889, + "step": 19595 + }, + { + "epoch": 0.4313449262601805, + "grad_norm": 0.6072697363865192, + "learning_rate": 6.338198331525148e-06, + "loss": 0.4919, + "step": 19596 + }, + { + "epoch": 0.43136693814659915, + "grad_norm": 1.5684731586271463, + "learning_rate": 6.337854875603679e-06, + "loss": 0.27, + "step": 19597 + }, + { + "epoch": 0.4313889500330178, + "grad_norm": 1.5158746271126695, + "learning_rate": 6.337511412882636e-06, + "loss": 0.3932, + "step": 19598 + }, + { + "epoch": 0.4314109619194365, + "grad_norm": 0.6464455138408207, + "learning_rate": 6.337167943363762e-06, + "loss": 0.5164, + "step": 19599 + }, + { + "epoch": 0.43143297380585516, + "grad_norm": 1.3826729949867662, + "learning_rate": 6.3368244670488035e-06, + "loss": 0.2811, + "step": 19600 + }, + { + "epoch": 0.4314549856922738, + "grad_norm": 1.485533478898363, + "learning_rate": 6.336480983939508e-06, + "loss": 0.3191, + "step": 19601 + }, + { + "epoch": 0.4314769975786925, + "grad_norm": 3.1697370521647477, + "learning_rate": 6.336137494037621e-06, + "loss": 0.4093, + "step": 19602 + }, + { + "epoch": 0.43149900946511116, + "grad_norm": 1.9442335230130867, + "learning_rate": 6.335793997344885e-06, + "loss": 0.4581, + "step": 19603 + }, + { + "epoch": 0.4315210213515298, + "grad_norm": 0.667459006185524, + "learning_rate": 6.33545049386305e-06, + "loss": 0.501, + "step": 19604 + }, + { + "epoch": 0.4315430332379485, + "grad_norm": 1.7550636287946475, + "learning_rate": 6.335106983593858e-06, + "loss": 0.335, + "step": 19605 + }, + { + "epoch": 0.43156504512436716, + "grad_norm": 1.3517440664523273, + "learning_rate": 6.33476346653906e-06, + "loss": 0.3236, + "step": 19606 + }, + { + "epoch": 0.4315870570107858, + "grad_norm": 0.6443090717249584, + "learning_rate": 6.334419942700396e-06, + "loss": 0.4715, + "step": 19607 + }, + { + "epoch": 0.4316090688972045, + "grad_norm": 1.3304755786121782, + "learning_rate": 6.3340764120796165e-06, + "loss": 0.3368, + "step": 19608 + }, + { + "epoch": 0.43163108078362317, + "grad_norm": 2.487248731164498, + "learning_rate": 6.3337328746784646e-06, + "loss": 0.4052, + "step": 19609 + }, + { + "epoch": 0.4316530926700418, + "grad_norm": 1.1935884400202705, + "learning_rate": 6.333389330498686e-06, + "loss": 0.3268, + "step": 19610 + }, + { + "epoch": 0.43167510455646046, + "grad_norm": 1.0651774833302161, + "learning_rate": 6.333045779542031e-06, + "loss": 0.2641, + "step": 19611 + }, + { + "epoch": 0.43169711644287917, + "grad_norm": 2.085071100573112, + "learning_rate": 6.332702221810241e-06, + "loss": 0.2873, + "step": 19612 + }, + { + "epoch": 0.4317191283292978, + "grad_norm": 1.4622545763134525, + "learning_rate": 6.332358657305064e-06, + "loss": 0.3454, + "step": 19613 + }, + { + "epoch": 0.43174114021571647, + "grad_norm": 1.476993989171071, + "learning_rate": 6.3320150860282456e-06, + "loss": 0.3499, + "step": 19614 + }, + { + "epoch": 0.4317631521021352, + "grad_norm": 1.6748960358023208, + "learning_rate": 6.331671507981534e-06, + "loss": 0.3052, + "step": 19615 + }, + { + "epoch": 0.4317851639885538, + "grad_norm": 1.5014886113591768, + "learning_rate": 6.331327923166674e-06, + "loss": 0.286, + "step": 19616 + }, + { + "epoch": 0.43180717587497247, + "grad_norm": 1.4366023221242679, + "learning_rate": 6.33098433158541e-06, + "loss": 0.3118, + "step": 19617 + }, + { + "epoch": 0.4318291877613912, + "grad_norm": 0.6475795640009386, + "learning_rate": 6.33064073323949e-06, + "loss": 0.4947, + "step": 19618 + }, + { + "epoch": 0.4318511996478098, + "grad_norm": 2.190553988431022, + "learning_rate": 6.3302971281306615e-06, + "loss": 0.3496, + "step": 19619 + }, + { + "epoch": 0.4318732115342285, + "grad_norm": 1.2391668932810163, + "learning_rate": 6.329953516260669e-06, + "loss": 0.3033, + "step": 19620 + }, + { + "epoch": 0.4318952234206471, + "grad_norm": 1.5419396429750394, + "learning_rate": 6.329609897631259e-06, + "loss": 0.3105, + "step": 19621 + }, + { + "epoch": 0.4319172353070658, + "grad_norm": 1.557530328469659, + "learning_rate": 6.32926627224418e-06, + "loss": 0.3893, + "step": 19622 + }, + { + "epoch": 0.4319392471934845, + "grad_norm": 1.8685910899878513, + "learning_rate": 6.328922640101176e-06, + "loss": 0.3377, + "step": 19623 + }, + { + "epoch": 0.4319612590799031, + "grad_norm": 1.748299820167822, + "learning_rate": 6.328579001203996e-06, + "loss": 0.2902, + "step": 19624 + }, + { + "epoch": 0.43198327096632183, + "grad_norm": 1.383783821517011, + "learning_rate": 6.328235355554382e-06, + "loss": 0.3377, + "step": 19625 + }, + { + "epoch": 0.4320052828527405, + "grad_norm": 1.5929452123910661, + "learning_rate": 6.3278917031540855e-06, + "loss": 0.2903, + "step": 19626 + }, + { + "epoch": 0.43202729473915913, + "grad_norm": 1.1456547960081351, + "learning_rate": 6.327548044004851e-06, + "loss": 0.3329, + "step": 19627 + }, + { + "epoch": 0.43204930662557783, + "grad_norm": 1.8171272393199542, + "learning_rate": 6.327204378108423e-06, + "loss": 0.376, + "step": 19628 + }, + { + "epoch": 0.4320713185119965, + "grad_norm": 2.744049652553883, + "learning_rate": 6.326860705466555e-06, + "loss": 0.2183, + "step": 19629 + }, + { + "epoch": 0.43209333039841513, + "grad_norm": 0.611588042220014, + "learning_rate": 6.326517026080986e-06, + "loss": 0.479, + "step": 19630 + }, + { + "epoch": 0.43211534228483384, + "grad_norm": 0.6484386218236661, + "learning_rate": 6.326173339953466e-06, + "loss": 0.5086, + "step": 19631 + }, + { + "epoch": 0.4321373541712525, + "grad_norm": 1.2123580524680855, + "learning_rate": 6.325829647085742e-06, + "loss": 0.3072, + "step": 19632 + }, + { + "epoch": 0.43215936605767113, + "grad_norm": 1.826384671488653, + "learning_rate": 6.325485947479559e-06, + "loss": 0.261, + "step": 19633 + }, + { + "epoch": 0.4321813779440898, + "grad_norm": 1.3510928917575955, + "learning_rate": 6.325142241136667e-06, + "loss": 0.3158, + "step": 19634 + }, + { + "epoch": 0.4322033898305085, + "grad_norm": 0.6042679751244507, + "learning_rate": 6.324798528058811e-06, + "loss": 0.4807, + "step": 19635 + }, + { + "epoch": 0.43222540171692714, + "grad_norm": 1.2346984740044709, + "learning_rate": 6.3244548082477365e-06, + "loss": 0.234, + "step": 19636 + }, + { + "epoch": 0.4322474136033458, + "grad_norm": 1.4505857633631782, + "learning_rate": 6.324111081705193e-06, + "loss": 0.3593, + "step": 19637 + }, + { + "epoch": 0.4322694254897645, + "grad_norm": 1.506519531182593, + "learning_rate": 6.323767348432926e-06, + "loss": 0.2716, + "step": 19638 + }, + { + "epoch": 0.43229143737618314, + "grad_norm": 1.8867964808778082, + "learning_rate": 6.3234236084326835e-06, + "loss": 0.3921, + "step": 19639 + }, + { + "epoch": 0.4323134492626018, + "grad_norm": 1.4562230211602152, + "learning_rate": 6.323079861706213e-06, + "loss": 0.2946, + "step": 19640 + }, + { + "epoch": 0.4323354611490205, + "grad_norm": 0.6144214773789507, + "learning_rate": 6.322736108255258e-06, + "loss": 0.4648, + "step": 19641 + }, + { + "epoch": 0.43235747303543914, + "grad_norm": 1.3875275010317087, + "learning_rate": 6.32239234808157e-06, + "loss": 0.2803, + "step": 19642 + }, + { + "epoch": 0.4323794849218578, + "grad_norm": 1.566911485238714, + "learning_rate": 6.322048581186893e-06, + "loss": 0.3685, + "step": 19643 + }, + { + "epoch": 0.43240149680827644, + "grad_norm": 1.5622181363544863, + "learning_rate": 6.321704807572977e-06, + "loss": 0.2447, + "step": 19644 + }, + { + "epoch": 0.43242350869469515, + "grad_norm": 1.450100645340427, + "learning_rate": 6.321361027241567e-06, + "loss": 0.3418, + "step": 19645 + }, + { + "epoch": 0.4324455205811138, + "grad_norm": 0.6365246162791448, + "learning_rate": 6.32101724019441e-06, + "loss": 0.5156, + "step": 19646 + }, + { + "epoch": 0.43246753246753245, + "grad_norm": 1.6531579416474014, + "learning_rate": 6.3206734464332564e-06, + "loss": 0.2446, + "step": 19647 + }, + { + "epoch": 0.43248954435395115, + "grad_norm": 1.0392993854457528, + "learning_rate": 6.320329645959849e-06, + "loss": 0.2047, + "step": 19648 + }, + { + "epoch": 0.4325115562403698, + "grad_norm": 1.5845071658003729, + "learning_rate": 6.319985838775938e-06, + "loss": 0.3526, + "step": 19649 + }, + { + "epoch": 0.43253356812678845, + "grad_norm": 1.5442858490876332, + "learning_rate": 6.319642024883271e-06, + "loss": 0.2866, + "step": 19650 + }, + { + "epoch": 0.43255558001320715, + "grad_norm": 1.7419975630516715, + "learning_rate": 6.319298204283595e-06, + "loss": 0.3239, + "step": 19651 + }, + { + "epoch": 0.4325775918996258, + "grad_norm": 1.247864514996167, + "learning_rate": 6.3189543769786564e-06, + "loss": 0.294, + "step": 19652 + }, + { + "epoch": 0.43259960378604445, + "grad_norm": 1.6449946933981114, + "learning_rate": 6.318610542970203e-06, + "loss": 0.3875, + "step": 19653 + }, + { + "epoch": 0.43262161567246316, + "grad_norm": 1.3728950727986595, + "learning_rate": 6.318266702259982e-06, + "loss": 0.3252, + "step": 19654 + }, + { + "epoch": 0.4326436275588818, + "grad_norm": 1.469039212458143, + "learning_rate": 6.317922854849743e-06, + "loss": 0.2959, + "step": 19655 + }, + { + "epoch": 0.43266563944530045, + "grad_norm": 1.4230430089225976, + "learning_rate": 6.317579000741232e-06, + "loss": 0.3434, + "step": 19656 + }, + { + "epoch": 0.4326876513317191, + "grad_norm": 1.3440876996951303, + "learning_rate": 6.3172351399361975e-06, + "loss": 0.267, + "step": 19657 + }, + { + "epoch": 0.4327096632181378, + "grad_norm": 2.00184570514055, + "learning_rate": 6.316891272436385e-06, + "loss": 0.2997, + "step": 19658 + }, + { + "epoch": 0.43273167510455646, + "grad_norm": 1.6607317197284677, + "learning_rate": 6.316547398243545e-06, + "loss": 0.3627, + "step": 19659 + }, + { + "epoch": 0.4327536869909751, + "grad_norm": 1.5098132840558915, + "learning_rate": 6.316203517359423e-06, + "loss": 0.3451, + "step": 19660 + }, + { + "epoch": 0.4327756988773938, + "grad_norm": 1.9335381712244033, + "learning_rate": 6.315859629785768e-06, + "loss": 0.3269, + "step": 19661 + }, + { + "epoch": 0.43279771076381246, + "grad_norm": 1.2854518902619703, + "learning_rate": 6.315515735524328e-06, + "loss": 0.3148, + "step": 19662 + }, + { + "epoch": 0.4328197226502311, + "grad_norm": 1.8839780898151028, + "learning_rate": 6.31517183457685e-06, + "loss": 0.3535, + "step": 19663 + }, + { + "epoch": 0.4328417345366498, + "grad_norm": 0.6342773525422268, + "learning_rate": 6.314827926945082e-06, + "loss": 0.4792, + "step": 19664 + }, + { + "epoch": 0.43286374642306846, + "grad_norm": 1.460436823111947, + "learning_rate": 6.314484012630773e-06, + "loss": 0.3372, + "step": 19665 + }, + { + "epoch": 0.4328857583094871, + "grad_norm": 4.584169010330182, + "learning_rate": 6.314140091635669e-06, + "loss": 0.2779, + "step": 19666 + }, + { + "epoch": 0.4329077701959058, + "grad_norm": 0.6179295967943669, + "learning_rate": 6.313796163961517e-06, + "loss": 0.5033, + "step": 19667 + }, + { + "epoch": 0.43292978208232447, + "grad_norm": 1.2687074937567855, + "learning_rate": 6.3134522296100696e-06, + "loss": 0.4409, + "step": 19668 + }, + { + "epoch": 0.4329517939687431, + "grad_norm": 0.630379543059007, + "learning_rate": 6.313108288583072e-06, + "loss": 0.4764, + "step": 19669 + }, + { + "epoch": 0.43297380585516176, + "grad_norm": 0.6158650386240891, + "learning_rate": 6.312764340882272e-06, + "loss": 0.4747, + "step": 19670 + }, + { + "epoch": 0.43299581774158047, + "grad_norm": 1.3002610441455216, + "learning_rate": 6.312420386509418e-06, + "loss": 0.3091, + "step": 19671 + }, + { + "epoch": 0.4330178296279991, + "grad_norm": 1.4937462724882755, + "learning_rate": 6.312076425466258e-06, + "loss": 0.2951, + "step": 19672 + }, + { + "epoch": 0.43303984151441777, + "grad_norm": 1.6383010200313644, + "learning_rate": 6.31173245775454e-06, + "loss": 0.3566, + "step": 19673 + }, + { + "epoch": 0.43306185340083647, + "grad_norm": 1.9962301554360222, + "learning_rate": 6.311388483376014e-06, + "loss": 0.3422, + "step": 19674 + }, + { + "epoch": 0.4330838652872551, + "grad_norm": 1.3970188807354054, + "learning_rate": 6.311044502332427e-06, + "loss": 0.3461, + "step": 19675 + }, + { + "epoch": 0.43310587717367377, + "grad_norm": 1.2536431864106925, + "learning_rate": 6.310700514625525e-06, + "loss": 0.3387, + "step": 19676 + }, + { + "epoch": 0.4331278890600925, + "grad_norm": 1.162335201120264, + "learning_rate": 6.31035652025706e-06, + "loss": 0.2744, + "step": 19677 + }, + { + "epoch": 0.4331499009465111, + "grad_norm": 0.6309270075481941, + "learning_rate": 6.310012519228778e-06, + "loss": 0.4785, + "step": 19678 + }, + { + "epoch": 0.4331719128329298, + "grad_norm": 2.223777774857546, + "learning_rate": 6.309668511542429e-06, + "loss": 0.2859, + "step": 19679 + }, + { + "epoch": 0.4331939247193484, + "grad_norm": 1.7086764405636885, + "learning_rate": 6.309324497199759e-06, + "loss": 0.3625, + "step": 19680 + }, + { + "epoch": 0.4332159366057671, + "grad_norm": 0.6284944587363331, + "learning_rate": 6.30898047620252e-06, + "loss": 0.5066, + "step": 19681 + }, + { + "epoch": 0.4332379484921858, + "grad_norm": 1.3076220050994163, + "learning_rate": 6.308636448552455e-06, + "loss": 0.3601, + "step": 19682 + }, + { + "epoch": 0.4332599603786044, + "grad_norm": 1.2161235662828833, + "learning_rate": 6.308292414251321e-06, + "loss": 0.3614, + "step": 19683 + }, + { + "epoch": 0.43328197226502313, + "grad_norm": 1.0571471158227328, + "learning_rate": 6.307948373300857e-06, + "loss": 0.2188, + "step": 19684 + }, + { + "epoch": 0.4333039841514418, + "grad_norm": 1.5010980118543433, + "learning_rate": 6.307604325702818e-06, + "loss": 0.2497, + "step": 19685 + }, + { + "epoch": 0.43332599603786043, + "grad_norm": 0.6240202746619312, + "learning_rate": 6.307260271458949e-06, + "loss": 0.4899, + "step": 19686 + }, + { + "epoch": 0.43334800792427913, + "grad_norm": 1.3515916990913472, + "learning_rate": 6.3069162105710025e-06, + "loss": 0.3235, + "step": 19687 + }, + { + "epoch": 0.4333700198106978, + "grad_norm": 0.5997043042102905, + "learning_rate": 6.306572143040724e-06, + "loss": 0.4611, + "step": 19688 + }, + { + "epoch": 0.43339203169711643, + "grad_norm": 1.7498016767233628, + "learning_rate": 6.306228068869863e-06, + "loss": 0.3515, + "step": 19689 + }, + { + "epoch": 0.43341404358353514, + "grad_norm": 1.3974427714847906, + "learning_rate": 6.305883988060167e-06, + "loss": 0.2472, + "step": 19690 + }, + { + "epoch": 0.4334360554699538, + "grad_norm": 1.7838102395549054, + "learning_rate": 6.305539900613388e-06, + "loss": 0.3732, + "step": 19691 + }, + { + "epoch": 0.43345806735637243, + "grad_norm": 0.620739479056993, + "learning_rate": 6.305195806531272e-06, + "loss": 0.4914, + "step": 19692 + }, + { + "epoch": 0.4334800792427911, + "grad_norm": 0.6129212118620719, + "learning_rate": 6.304851705815569e-06, + "loss": 0.4908, + "step": 19693 + }, + { + "epoch": 0.4335020911292098, + "grad_norm": 1.6957007760912521, + "learning_rate": 6.304507598468028e-06, + "loss": 0.3869, + "step": 19694 + }, + { + "epoch": 0.43352410301562844, + "grad_norm": 2.511711725454772, + "learning_rate": 6.304163484490398e-06, + "loss": 0.3036, + "step": 19695 + }, + { + "epoch": 0.4335461149020471, + "grad_norm": 1.1250015765561376, + "learning_rate": 6.303819363884426e-06, + "loss": 0.2876, + "step": 19696 + }, + { + "epoch": 0.4335681267884658, + "grad_norm": 1.4749559891425217, + "learning_rate": 6.303475236651863e-06, + "loss": 0.3027, + "step": 19697 + }, + { + "epoch": 0.43359013867488444, + "grad_norm": 1.797672341407148, + "learning_rate": 6.303131102794459e-06, + "loss": 0.3527, + "step": 19698 + }, + { + "epoch": 0.4336121505613031, + "grad_norm": 1.1883589903053895, + "learning_rate": 6.30278696231396e-06, + "loss": 0.369, + "step": 19699 + }, + { + "epoch": 0.4336341624477218, + "grad_norm": 1.7691578376566757, + "learning_rate": 6.3024428152121165e-06, + "loss": 0.355, + "step": 19700 + }, + { + "epoch": 0.43365617433414044, + "grad_norm": 1.4061093783669563, + "learning_rate": 6.302098661490678e-06, + "loss": 0.2615, + "step": 19701 + }, + { + "epoch": 0.4336781862205591, + "grad_norm": 1.5818083331895796, + "learning_rate": 6.301754501151395e-06, + "loss": 0.2943, + "step": 19702 + }, + { + "epoch": 0.43370019810697774, + "grad_norm": 1.4766142552645405, + "learning_rate": 6.301410334196012e-06, + "loss": 0.2727, + "step": 19703 + }, + { + "epoch": 0.43372220999339645, + "grad_norm": 0.61762255648415, + "learning_rate": 6.301066160626283e-06, + "loss": 0.4841, + "step": 19704 + }, + { + "epoch": 0.4337442218798151, + "grad_norm": 2.07338962923563, + "learning_rate": 6.3007219804439556e-06, + "loss": 0.3209, + "step": 19705 + }, + { + "epoch": 0.43376623376623374, + "grad_norm": 1.417230646176898, + "learning_rate": 6.300377793650778e-06, + "loss": 0.2856, + "step": 19706 + }, + { + "epoch": 0.43378824565265245, + "grad_norm": 1.5225487106691393, + "learning_rate": 6.300033600248502e-06, + "loss": 0.382, + "step": 19707 + }, + { + "epoch": 0.4338102575390711, + "grad_norm": 1.448054380720207, + "learning_rate": 6.299689400238873e-06, + "loss": 0.2659, + "step": 19708 + }, + { + "epoch": 0.43383226942548975, + "grad_norm": 1.6497041870982627, + "learning_rate": 6.299345193623645e-06, + "loss": 0.2825, + "step": 19709 + }, + { + "epoch": 0.43385428131190845, + "grad_norm": 1.4387077849501066, + "learning_rate": 6.299000980404565e-06, + "loss": 0.298, + "step": 19710 + }, + { + "epoch": 0.4338762931983271, + "grad_norm": 1.531531444243253, + "learning_rate": 6.2986567605833814e-06, + "loss": 0.3405, + "step": 19711 + }, + { + "epoch": 0.43389830508474575, + "grad_norm": 1.6802441084566517, + "learning_rate": 6.298312534161847e-06, + "loss": 0.3824, + "step": 19712 + }, + { + "epoch": 0.43392031697116445, + "grad_norm": 1.516528482585636, + "learning_rate": 6.297968301141706e-06, + "loss": 0.3807, + "step": 19713 + }, + { + "epoch": 0.4339423288575831, + "grad_norm": 1.5551455971354655, + "learning_rate": 6.2976240615247145e-06, + "loss": 0.3046, + "step": 19714 + }, + { + "epoch": 0.43396434074400175, + "grad_norm": 1.4797833777423708, + "learning_rate": 6.297279815312617e-06, + "loss": 0.3271, + "step": 19715 + }, + { + "epoch": 0.4339863526304204, + "grad_norm": 2.9426423315154198, + "learning_rate": 6.296935562507165e-06, + "loss": 0.3907, + "step": 19716 + }, + { + "epoch": 0.4340083645168391, + "grad_norm": 0.609014092422997, + "learning_rate": 6.296591303110109e-06, + "loss": 0.4907, + "step": 19717 + }, + { + "epoch": 0.43403037640325776, + "grad_norm": 1.6689505268062934, + "learning_rate": 6.296247037123195e-06, + "loss": 0.3437, + "step": 19718 + }, + { + "epoch": 0.4340523882896764, + "grad_norm": 1.3082954317734838, + "learning_rate": 6.295902764548178e-06, + "loss": 0.378, + "step": 19719 + }, + { + "epoch": 0.4340744001760951, + "grad_norm": 1.3992775278567544, + "learning_rate": 6.295558485386806e-06, + "loss": 0.3845, + "step": 19720 + }, + { + "epoch": 0.43409641206251376, + "grad_norm": 1.7157409947739632, + "learning_rate": 6.295214199640826e-06, + "loss": 0.3277, + "step": 19721 + }, + { + "epoch": 0.4341184239489324, + "grad_norm": 1.62766030338567, + "learning_rate": 6.294869907311989e-06, + "loss": 0.3211, + "step": 19722 + }, + { + "epoch": 0.4341404358353511, + "grad_norm": 1.539086569615367, + "learning_rate": 6.294525608402048e-06, + "loss": 0.3373, + "step": 19723 + }, + { + "epoch": 0.43416244772176976, + "grad_norm": 8.329329779933065, + "learning_rate": 6.29418130291275e-06, + "loss": 0.3801, + "step": 19724 + }, + { + "epoch": 0.4341844596081884, + "grad_norm": 1.4669761795955778, + "learning_rate": 6.293836990845844e-06, + "loss": 0.2861, + "step": 19725 + }, + { + "epoch": 0.43420647149460706, + "grad_norm": 2.4047112460199975, + "learning_rate": 6.293492672203081e-06, + "loss": 0.2979, + "step": 19726 + }, + { + "epoch": 0.43422848338102576, + "grad_norm": 1.1278222546282362, + "learning_rate": 6.293148346986213e-06, + "loss": 0.2468, + "step": 19727 + }, + { + "epoch": 0.4342504952674444, + "grad_norm": 2.037377400516322, + "learning_rate": 6.292804015196987e-06, + "loss": 0.3347, + "step": 19728 + }, + { + "epoch": 0.43427250715386306, + "grad_norm": 1.4789191750668025, + "learning_rate": 6.292459676837156e-06, + "loss": 0.4017, + "step": 19729 + }, + { + "epoch": 0.43429451904028177, + "grad_norm": 1.5907395603317334, + "learning_rate": 6.2921153319084675e-06, + "loss": 0.3396, + "step": 19730 + }, + { + "epoch": 0.4343165309267004, + "grad_norm": 1.6406885005121523, + "learning_rate": 6.291770980412672e-06, + "loss": 0.3388, + "step": 19731 + }, + { + "epoch": 0.43433854281311907, + "grad_norm": 0.6583446188018051, + "learning_rate": 6.291426622351521e-06, + "loss": 0.5065, + "step": 19732 + }, + { + "epoch": 0.43436055469953777, + "grad_norm": 1.3983494692459335, + "learning_rate": 6.291082257726764e-06, + "loss": 0.2836, + "step": 19733 + }, + { + "epoch": 0.4343825665859564, + "grad_norm": 1.4337743476007927, + "learning_rate": 6.2907378865401506e-06, + "loss": 0.2905, + "step": 19734 + }, + { + "epoch": 0.43440457847237507, + "grad_norm": 1.3570676336855796, + "learning_rate": 6.290393508793432e-06, + "loss": 0.279, + "step": 19735 + }, + { + "epoch": 0.4344265903587938, + "grad_norm": 1.4922995698481487, + "learning_rate": 6.2900491244883575e-06, + "loss": 0.2582, + "step": 19736 + }, + { + "epoch": 0.4344486022452124, + "grad_norm": 1.2514297707332966, + "learning_rate": 6.289704733626678e-06, + "loss": 0.3456, + "step": 19737 + }, + { + "epoch": 0.43447061413163107, + "grad_norm": 1.308293830714036, + "learning_rate": 6.289360336210146e-06, + "loss": 0.2793, + "step": 19738 + }, + { + "epoch": 0.4344926260180497, + "grad_norm": 2.1780862523871822, + "learning_rate": 6.2890159322405066e-06, + "loss": 0.2755, + "step": 19739 + }, + { + "epoch": 0.4345146379044684, + "grad_norm": 1.3450440113629047, + "learning_rate": 6.288671521719515e-06, + "loss": 0.2797, + "step": 19740 + }, + { + "epoch": 0.4345366497908871, + "grad_norm": 2.7756353889046137, + "learning_rate": 6.288327104648919e-06, + "loss": 0.2925, + "step": 19741 + }, + { + "epoch": 0.4345586616773057, + "grad_norm": 1.7878617296467025, + "learning_rate": 6.287982681030471e-06, + "loss": 0.3631, + "step": 19742 + }, + { + "epoch": 0.43458067356372443, + "grad_norm": 0.6199707564048628, + "learning_rate": 6.28763825086592e-06, + "loss": 0.4967, + "step": 19743 + }, + { + "epoch": 0.4346026854501431, + "grad_norm": 1.5337230079625246, + "learning_rate": 6.287293814157018e-06, + "loss": 0.2475, + "step": 19744 + }, + { + "epoch": 0.4346246973365617, + "grad_norm": 1.5317822961666174, + "learning_rate": 6.286949370905513e-06, + "loss": 0.3421, + "step": 19745 + }, + { + "epoch": 0.43464670922298043, + "grad_norm": 1.4952190491086828, + "learning_rate": 6.28660492111316e-06, + "loss": 0.3313, + "step": 19746 + }, + { + "epoch": 0.4346687211093991, + "grad_norm": 0.6090960262242177, + "learning_rate": 6.286260464781705e-06, + "loss": 0.463, + "step": 19747 + }, + { + "epoch": 0.43469073299581773, + "grad_norm": 1.6736834738689987, + "learning_rate": 6.285916001912901e-06, + "loss": 0.2813, + "step": 19748 + }, + { + "epoch": 0.43471274488223643, + "grad_norm": 1.5396772559678555, + "learning_rate": 6.285571532508497e-06, + "loss": 0.361, + "step": 19749 + }, + { + "epoch": 0.4347347567686551, + "grad_norm": 1.2458410560093205, + "learning_rate": 6.285227056570247e-06, + "loss": 0.2722, + "step": 19750 + }, + { + "epoch": 0.43475676865507373, + "grad_norm": 0.6175445240369879, + "learning_rate": 6.2848825740998985e-06, + "loss": 0.5042, + "step": 19751 + }, + { + "epoch": 0.4347787805414924, + "grad_norm": 0.6032700741478981, + "learning_rate": 6.284538085099206e-06, + "loss": 0.4747, + "step": 19752 + }, + { + "epoch": 0.4348007924279111, + "grad_norm": 1.3903793862481837, + "learning_rate": 6.284193589569915e-06, + "loss": 0.3188, + "step": 19753 + }, + { + "epoch": 0.43482280431432974, + "grad_norm": 1.981194269125437, + "learning_rate": 6.28384908751378e-06, + "loss": 0.3455, + "step": 19754 + }, + { + "epoch": 0.4348448162007484, + "grad_norm": 1.2273719896950093, + "learning_rate": 6.283504578932553e-06, + "loss": 0.2782, + "step": 19755 + }, + { + "epoch": 0.4348668280871671, + "grad_norm": 1.7284264934063343, + "learning_rate": 6.283160063827983e-06, + "loss": 0.329, + "step": 19756 + }, + { + "epoch": 0.43488883997358574, + "grad_norm": 1.4963800164832424, + "learning_rate": 6.28281554220182e-06, + "loss": 0.2873, + "step": 19757 + }, + { + "epoch": 0.4349108518600044, + "grad_norm": 1.3939354902467, + "learning_rate": 6.2824710140558174e-06, + "loss": 0.3897, + "step": 19758 + }, + { + "epoch": 0.4349328637464231, + "grad_norm": 1.6712898210980902, + "learning_rate": 6.282126479391724e-06, + "loss": 0.3663, + "step": 19759 + }, + { + "epoch": 0.43495487563284174, + "grad_norm": 1.3443474251483432, + "learning_rate": 6.2817819382112924e-06, + "loss": 0.3303, + "step": 19760 + }, + { + "epoch": 0.4349768875192604, + "grad_norm": 0.618390176923909, + "learning_rate": 6.281437390516274e-06, + "loss": 0.4886, + "step": 19761 + }, + { + "epoch": 0.43499889940567904, + "grad_norm": 1.4784682764469899, + "learning_rate": 6.281092836308419e-06, + "loss": 0.3665, + "step": 19762 + }, + { + "epoch": 0.43502091129209775, + "grad_norm": 1.7164852204544163, + "learning_rate": 6.280748275589479e-06, + "loss": 0.3169, + "step": 19763 + }, + { + "epoch": 0.4350429231785164, + "grad_norm": 2.1186540634969235, + "learning_rate": 6.280403708361205e-06, + "loss": 0.3365, + "step": 19764 + }, + { + "epoch": 0.43506493506493504, + "grad_norm": 1.476392940657707, + "learning_rate": 6.280059134625347e-06, + "loss": 0.3794, + "step": 19765 + }, + { + "epoch": 0.43508694695135375, + "grad_norm": 1.320771931908502, + "learning_rate": 6.279714554383659e-06, + "loss": 0.3806, + "step": 19766 + }, + { + "epoch": 0.4351089588377724, + "grad_norm": 1.2908183641924484, + "learning_rate": 6.279369967637889e-06, + "loss": 0.3155, + "step": 19767 + }, + { + "epoch": 0.43513097072419105, + "grad_norm": 1.151584168741497, + "learning_rate": 6.279025374389792e-06, + "loss": 0.2854, + "step": 19768 + }, + { + "epoch": 0.43515298261060975, + "grad_norm": 2.6646785439046448, + "learning_rate": 6.278680774641117e-06, + "loss": 0.3543, + "step": 19769 + }, + { + "epoch": 0.4351749944970284, + "grad_norm": 0.6641502283712151, + "learning_rate": 6.278336168393616e-06, + "loss": 0.5267, + "step": 19770 + }, + { + "epoch": 0.43519700638344705, + "grad_norm": 0.6452659307976002, + "learning_rate": 6.277991555649041e-06, + "loss": 0.5166, + "step": 19771 + }, + { + "epoch": 0.43521901826986575, + "grad_norm": 2.3796841590746123, + "learning_rate": 6.277646936409142e-06, + "loss": 0.4018, + "step": 19772 + }, + { + "epoch": 0.4352410301562844, + "grad_norm": 1.3320875149995073, + "learning_rate": 6.277302310675672e-06, + "loss": 0.3372, + "step": 19773 + }, + { + "epoch": 0.43526304204270305, + "grad_norm": 1.2780952483486074, + "learning_rate": 6.276957678450383e-06, + "loss": 0.2875, + "step": 19774 + }, + { + "epoch": 0.4352850539291217, + "grad_norm": 0.6208483470868754, + "learning_rate": 6.276613039735024e-06, + "loss": 0.4624, + "step": 19775 + }, + { + "epoch": 0.4353070658155404, + "grad_norm": 1.5497407115970316, + "learning_rate": 6.276268394531349e-06, + "loss": 0.3745, + "step": 19776 + }, + { + "epoch": 0.43532907770195906, + "grad_norm": 3.1329592114248483, + "learning_rate": 6.275923742841108e-06, + "loss": 0.2811, + "step": 19777 + }, + { + "epoch": 0.4353510895883777, + "grad_norm": 0.6406942360773925, + "learning_rate": 6.275579084666052e-06, + "loss": 0.5071, + "step": 19778 + }, + { + "epoch": 0.4353731014747964, + "grad_norm": 1.3936993257334516, + "learning_rate": 6.275234420007939e-06, + "loss": 0.3744, + "step": 19779 + }, + { + "epoch": 0.43539511336121506, + "grad_norm": 1.4070588313423016, + "learning_rate": 6.2748897488685115e-06, + "loss": 0.2888, + "step": 19780 + }, + { + "epoch": 0.4354171252476337, + "grad_norm": 0.6043409427790305, + "learning_rate": 6.274545071249527e-06, + "loss": 0.4811, + "step": 19781 + }, + { + "epoch": 0.4354391371340524, + "grad_norm": 0.6146365590130867, + "learning_rate": 6.274200387152736e-06, + "loss": 0.4788, + "step": 19782 + }, + { + "epoch": 0.43546114902047106, + "grad_norm": 2.890024482219217, + "learning_rate": 6.27385569657989e-06, + "loss": 0.2476, + "step": 19783 + }, + { + "epoch": 0.4354831609068897, + "grad_norm": 1.5687668123068785, + "learning_rate": 6.273510999532741e-06, + "loss": 0.3989, + "step": 19784 + }, + { + "epoch": 0.43550517279330836, + "grad_norm": 0.6108847840225149, + "learning_rate": 6.273166296013041e-06, + "loss": 0.4861, + "step": 19785 + }, + { + "epoch": 0.43552718467972706, + "grad_norm": 2.021624540420299, + "learning_rate": 6.272821586022544e-06, + "loss": 0.3003, + "step": 19786 + }, + { + "epoch": 0.4355491965661457, + "grad_norm": 1.9805114536967758, + "learning_rate": 6.272476869562998e-06, + "loss": 0.3699, + "step": 19787 + }, + { + "epoch": 0.43557120845256436, + "grad_norm": 1.7457496462663111, + "learning_rate": 6.272132146636158e-06, + "loss": 0.3848, + "step": 19788 + }, + { + "epoch": 0.43559322033898307, + "grad_norm": 1.3093746386758067, + "learning_rate": 6.2717874172437755e-06, + "loss": 0.3103, + "step": 19789 + }, + { + "epoch": 0.4356152322254017, + "grad_norm": 1.335342574369695, + "learning_rate": 6.2714426813876005e-06, + "loss": 0.2863, + "step": 19790 + }, + { + "epoch": 0.43563724411182037, + "grad_norm": 1.7945510980943478, + "learning_rate": 6.2710979390693875e-06, + "loss": 0.2696, + "step": 19791 + }, + { + "epoch": 0.43565925599823907, + "grad_norm": 1.4801294537694991, + "learning_rate": 6.270753190290888e-06, + "loss": 0.3691, + "step": 19792 + }, + { + "epoch": 0.4356812678846577, + "grad_norm": 2.4385466503149447, + "learning_rate": 6.270408435053854e-06, + "loss": 0.2475, + "step": 19793 + }, + { + "epoch": 0.43570327977107637, + "grad_norm": 1.3811285456315807, + "learning_rate": 6.270063673360038e-06, + "loss": 0.3153, + "step": 19794 + }, + { + "epoch": 0.4357252916574951, + "grad_norm": 0.6224616734261874, + "learning_rate": 6.269718905211192e-06, + "loss": 0.4987, + "step": 19795 + }, + { + "epoch": 0.4357473035439137, + "grad_norm": 4.5996604267614565, + "learning_rate": 6.269374130609067e-06, + "loss": 0.3233, + "step": 19796 + }, + { + "epoch": 0.43576931543033237, + "grad_norm": 1.2152119494103508, + "learning_rate": 6.269029349555417e-06, + "loss": 0.2912, + "step": 19797 + }, + { + "epoch": 0.435791327316751, + "grad_norm": 1.4676192114918687, + "learning_rate": 6.268684562051995e-06, + "loss": 0.37, + "step": 19798 + }, + { + "epoch": 0.4358133392031697, + "grad_norm": 1.4359365404645479, + "learning_rate": 6.268339768100551e-06, + "loss": 0.2868, + "step": 19799 + }, + { + "epoch": 0.4358353510895884, + "grad_norm": 1.3869533272485666, + "learning_rate": 6.267994967702839e-06, + "loss": 0.4028, + "step": 19800 + }, + { + "epoch": 0.435857362976007, + "grad_norm": 1.2209348726607674, + "learning_rate": 6.26765016086061e-06, + "loss": 0.341, + "step": 19801 + }, + { + "epoch": 0.43587937486242573, + "grad_norm": 1.7363703121193927, + "learning_rate": 6.26730534757562e-06, + "loss": 0.4112, + "step": 19802 + }, + { + "epoch": 0.4359013867488444, + "grad_norm": 1.7429793691426811, + "learning_rate": 6.266960527849617e-06, + "loss": 0.2858, + "step": 19803 + }, + { + "epoch": 0.435923398635263, + "grad_norm": 1.4925582777101143, + "learning_rate": 6.2666157016843556e-06, + "loss": 0.3291, + "step": 19804 + }, + { + "epoch": 0.43594541052168173, + "grad_norm": 1.4330793955643613, + "learning_rate": 6.2662708690815885e-06, + "loss": 0.3556, + "step": 19805 + }, + { + "epoch": 0.4359674224081004, + "grad_norm": 0.6442001988515499, + "learning_rate": 6.265926030043069e-06, + "loss": 0.508, + "step": 19806 + }, + { + "epoch": 0.43598943429451903, + "grad_norm": 1.3782481179096222, + "learning_rate": 6.265581184570548e-06, + "loss": 0.3389, + "step": 19807 + }, + { + "epoch": 0.43601144618093773, + "grad_norm": 1.4827732419650237, + "learning_rate": 6.26523633266578e-06, + "loss": 0.306, + "step": 19808 + }, + { + "epoch": 0.4360334580673564, + "grad_norm": 2.0098832921776326, + "learning_rate": 6.264891474330514e-06, + "loss": 0.419, + "step": 19809 + }, + { + "epoch": 0.43605546995377503, + "grad_norm": 2.4097819753321623, + "learning_rate": 6.264546609566507e-06, + "loss": 0.3739, + "step": 19810 + }, + { + "epoch": 0.4360774818401937, + "grad_norm": 1.858649562269404, + "learning_rate": 6.26420173837551e-06, + "loss": 0.3157, + "step": 19811 + }, + { + "epoch": 0.4360994937266124, + "grad_norm": 1.780351448154736, + "learning_rate": 6.263856860759276e-06, + "loss": 0.3924, + "step": 19812 + }, + { + "epoch": 0.43612150561303104, + "grad_norm": 1.9789895568044051, + "learning_rate": 6.263511976719558e-06, + "loss": 0.2792, + "step": 19813 + }, + { + "epoch": 0.4361435174994497, + "grad_norm": 1.6530841100253444, + "learning_rate": 6.2631670862581086e-06, + "loss": 0.2942, + "step": 19814 + }, + { + "epoch": 0.4361655293858684, + "grad_norm": 1.0637528945858998, + "learning_rate": 6.262822189376682e-06, + "loss": 0.3246, + "step": 19815 + }, + { + "epoch": 0.43618754127228704, + "grad_norm": 0.6422814038399668, + "learning_rate": 6.262477286077026e-06, + "loss": 0.4826, + "step": 19816 + }, + { + "epoch": 0.4362095531587057, + "grad_norm": 1.7607109919194108, + "learning_rate": 6.262132376360901e-06, + "loss": 0.318, + "step": 19817 + }, + { + "epoch": 0.4362315650451244, + "grad_norm": 1.2323781421819577, + "learning_rate": 6.261787460230055e-06, + "loss": 0.3724, + "step": 19818 + }, + { + "epoch": 0.43625357693154304, + "grad_norm": 1.648639761047014, + "learning_rate": 6.261442537686241e-06, + "loss": 0.2799, + "step": 19819 + }, + { + "epoch": 0.4362755888179617, + "grad_norm": 1.5438464059863595, + "learning_rate": 6.2610976087312155e-06, + "loss": 0.3401, + "step": 19820 + }, + { + "epoch": 0.43629760070438034, + "grad_norm": 1.4443723419133043, + "learning_rate": 6.260752673366728e-06, + "loss": 0.3365, + "step": 19821 + }, + { + "epoch": 0.43631961259079904, + "grad_norm": 5.3512650876493595, + "learning_rate": 6.2604077315945346e-06, + "loss": 0.2635, + "step": 19822 + }, + { + "epoch": 0.4363416244772177, + "grad_norm": 1.9289500183358832, + "learning_rate": 6.260062783416386e-06, + "loss": 0.385, + "step": 19823 + }, + { + "epoch": 0.43636363636363634, + "grad_norm": 1.3378915190941996, + "learning_rate": 6.259717828834036e-06, + "loss": 0.3084, + "step": 19824 + }, + { + "epoch": 0.43638564825005505, + "grad_norm": 1.2958257999337262, + "learning_rate": 6.259372867849239e-06, + "loss": 0.386, + "step": 19825 + }, + { + "epoch": 0.4364076601364737, + "grad_norm": 1.2126148761384934, + "learning_rate": 6.259027900463748e-06, + "loss": 0.3148, + "step": 19826 + }, + { + "epoch": 0.43642967202289235, + "grad_norm": 1.481324336219987, + "learning_rate": 6.258682926679314e-06, + "loss": 0.3237, + "step": 19827 + }, + { + "epoch": 0.43645168390931105, + "grad_norm": 0.6228367523230525, + "learning_rate": 6.258337946497692e-06, + "loss": 0.4908, + "step": 19828 + }, + { + "epoch": 0.4364736957957297, + "grad_norm": 1.4312517468152754, + "learning_rate": 6.2579929599206355e-06, + "loss": 0.3621, + "step": 19829 + }, + { + "epoch": 0.43649570768214835, + "grad_norm": 1.5668440519997748, + "learning_rate": 6.257647966949899e-06, + "loss": 0.3173, + "step": 19830 + }, + { + "epoch": 0.43651771956856705, + "grad_norm": 0.6278761823584009, + "learning_rate": 6.257302967587234e-06, + "loss": 0.486, + "step": 19831 + }, + { + "epoch": 0.4365397314549857, + "grad_norm": 0.6138249698939049, + "learning_rate": 6.256957961834393e-06, + "loss": 0.4694, + "step": 19832 + }, + { + "epoch": 0.43656174334140435, + "grad_norm": 1.4298207628069992, + "learning_rate": 6.256612949693135e-06, + "loss": 0.354, + "step": 19833 + }, + { + "epoch": 0.436583755227823, + "grad_norm": 1.5315172956674867, + "learning_rate": 6.256267931165205e-06, + "loss": 0.3254, + "step": 19834 + }, + { + "epoch": 0.4366057671142417, + "grad_norm": 1.6402289700687234, + "learning_rate": 6.255922906252363e-06, + "loss": 0.3179, + "step": 19835 + }, + { + "epoch": 0.43662777900066035, + "grad_norm": 1.4956377605557123, + "learning_rate": 6.255577874956361e-06, + "loss": 0.3409, + "step": 19836 + }, + { + "epoch": 0.436649790887079, + "grad_norm": 2.739674940833938, + "learning_rate": 6.25523283727895e-06, + "loss": 0.3569, + "step": 19837 + }, + { + "epoch": 0.4366718027734977, + "grad_norm": 1.6236668897312907, + "learning_rate": 6.25488779322189e-06, + "loss": 0.3671, + "step": 19838 + }, + { + "epoch": 0.43669381465991636, + "grad_norm": 3.697148444957728, + "learning_rate": 6.254542742786928e-06, + "loss": 0.2933, + "step": 19839 + }, + { + "epoch": 0.436715826546335, + "grad_norm": 4.880208055856797, + "learning_rate": 6.25419768597582e-06, + "loss": 0.4209, + "step": 19840 + }, + { + "epoch": 0.4367378384327537, + "grad_norm": 1.4586819034214078, + "learning_rate": 6.2538526227903205e-06, + "loss": 0.3177, + "step": 19841 + }, + { + "epoch": 0.43675985031917236, + "grad_norm": 1.97570670468327, + "learning_rate": 6.253507553232182e-06, + "loss": 0.4403, + "step": 19842 + }, + { + "epoch": 0.436781862205591, + "grad_norm": 1.3871012553201942, + "learning_rate": 6.2531624773031606e-06, + "loss": 0.3465, + "step": 19843 + }, + { + "epoch": 0.43680387409200966, + "grad_norm": 0.6530753095209401, + "learning_rate": 6.252817395005009e-06, + "loss": 0.4706, + "step": 19844 + }, + { + "epoch": 0.43682588597842836, + "grad_norm": 2.309212552835217, + "learning_rate": 6.252472306339478e-06, + "loss": 0.3298, + "step": 19845 + }, + { + "epoch": 0.436847897864847, + "grad_norm": 0.6453702978677744, + "learning_rate": 6.2521272113083255e-06, + "loss": 0.4879, + "step": 19846 + }, + { + "epoch": 0.43686990975126566, + "grad_norm": 1.5545157633715483, + "learning_rate": 6.251782109913303e-06, + "loss": 0.3388, + "step": 19847 + }, + { + "epoch": 0.43689192163768437, + "grad_norm": 0.6151425901526214, + "learning_rate": 6.251437002156168e-06, + "loss": 0.4939, + "step": 19848 + }, + { + "epoch": 0.436913933524103, + "grad_norm": 1.213536994225798, + "learning_rate": 6.251091888038669e-06, + "loss": 0.2795, + "step": 19849 + }, + { + "epoch": 0.43693594541052166, + "grad_norm": 1.0974249092967647, + "learning_rate": 6.250746767562565e-06, + "loss": 0.244, + "step": 19850 + }, + { + "epoch": 0.43695795729694037, + "grad_norm": 0.6139403013760278, + "learning_rate": 6.2504016407296074e-06, + "loss": 0.4966, + "step": 19851 + }, + { + "epoch": 0.436979969183359, + "grad_norm": 1.8889819085522375, + "learning_rate": 6.250056507541551e-06, + "loss": 0.3607, + "step": 19852 + }, + { + "epoch": 0.43700198106977767, + "grad_norm": 0.6445051902013551, + "learning_rate": 6.249711368000151e-06, + "loss": 0.5019, + "step": 19853 + }, + { + "epoch": 0.43702399295619637, + "grad_norm": 1.4913867672537684, + "learning_rate": 6.249366222107159e-06, + "loss": 0.2191, + "step": 19854 + }, + { + "epoch": 0.437046004842615, + "grad_norm": 1.0251972108578953, + "learning_rate": 6.249021069864331e-06, + "loss": 0.242, + "step": 19855 + }, + { + "epoch": 0.43706801672903367, + "grad_norm": 1.7334448898252737, + "learning_rate": 6.248675911273421e-06, + "loss": 0.3193, + "step": 19856 + }, + { + "epoch": 0.4370900286154523, + "grad_norm": 1.4179534717395492, + "learning_rate": 6.248330746336182e-06, + "loss": 0.3095, + "step": 19857 + }, + { + "epoch": 0.437112040501871, + "grad_norm": 1.713755300392688, + "learning_rate": 6.247985575054372e-06, + "loss": 0.3401, + "step": 19858 + }, + { + "epoch": 0.4371340523882897, + "grad_norm": 2.968267454526008, + "learning_rate": 6.2476403974297405e-06, + "loss": 0.2939, + "step": 19859 + }, + { + "epoch": 0.4371560642747083, + "grad_norm": 1.2574267480807693, + "learning_rate": 6.247295213464044e-06, + "loss": 0.358, + "step": 19860 + }, + { + "epoch": 0.437178076161127, + "grad_norm": 1.270052535706322, + "learning_rate": 6.246950023159038e-06, + "loss": 0.4071, + "step": 19861 + }, + { + "epoch": 0.4372000880475457, + "grad_norm": 1.852393505402452, + "learning_rate": 6.246604826516476e-06, + "loss": 0.3146, + "step": 19862 + }, + { + "epoch": 0.4372220999339643, + "grad_norm": 1.6876153097272588, + "learning_rate": 6.246259623538111e-06, + "loss": 0.2629, + "step": 19863 + }, + { + "epoch": 0.43724411182038303, + "grad_norm": 1.6522344285277684, + "learning_rate": 6.2459144142256985e-06, + "loss": 0.2488, + "step": 19864 + }, + { + "epoch": 0.4372661237068017, + "grad_norm": 2.698167781845724, + "learning_rate": 6.245569198580995e-06, + "loss": 0.3177, + "step": 19865 + }, + { + "epoch": 0.43728813559322033, + "grad_norm": 2.502543988228236, + "learning_rate": 6.245223976605752e-06, + "loss": 0.2901, + "step": 19866 + }, + { + "epoch": 0.437310147479639, + "grad_norm": 1.696722185423382, + "learning_rate": 6.244878748301727e-06, + "loss": 0.3619, + "step": 19867 + }, + { + "epoch": 0.4373321593660577, + "grad_norm": 1.4907028121271593, + "learning_rate": 6.2445335136706705e-06, + "loss": 0.4031, + "step": 19868 + }, + { + "epoch": 0.43735417125247633, + "grad_norm": 1.5688240843958718, + "learning_rate": 6.244188272714341e-06, + "loss": 0.3728, + "step": 19869 + }, + { + "epoch": 0.437376183138895, + "grad_norm": 1.690689731339492, + "learning_rate": 6.2438430254344914e-06, + "loss": 0.329, + "step": 19870 + }, + { + "epoch": 0.4373981950253137, + "grad_norm": 4.143699236842803, + "learning_rate": 6.243497771832878e-06, + "loss": 0.2834, + "step": 19871 + }, + { + "epoch": 0.43742020691173233, + "grad_norm": 1.63506431454528, + "learning_rate": 6.243152511911253e-06, + "loss": 0.1896, + "step": 19872 + }, + { + "epoch": 0.437442218798151, + "grad_norm": 0.6673048322170599, + "learning_rate": 6.2428072456713715e-06, + "loss": 0.4908, + "step": 19873 + }, + { + "epoch": 0.4374642306845697, + "grad_norm": 0.6705113492768449, + "learning_rate": 6.242461973114992e-06, + "loss": 0.5055, + "step": 19874 + }, + { + "epoch": 0.43748624257098834, + "grad_norm": 0.6271611813043934, + "learning_rate": 6.242116694243865e-06, + "loss": 0.5042, + "step": 19875 + }, + { + "epoch": 0.437508254457407, + "grad_norm": 1.372903508940757, + "learning_rate": 6.241771409059747e-06, + "loss": 0.4169, + "step": 19876 + }, + { + "epoch": 0.4375302663438257, + "grad_norm": 2.7870421136367063, + "learning_rate": 6.241426117564393e-06, + "loss": 0.2791, + "step": 19877 + }, + { + "epoch": 0.43755227823024434, + "grad_norm": 30.91883851640215, + "learning_rate": 6.2410808197595576e-06, + "loss": 0.2383, + "step": 19878 + }, + { + "epoch": 0.437574290116663, + "grad_norm": 1.6858726792746679, + "learning_rate": 6.240735515646997e-06, + "loss": 0.3787, + "step": 19879 + }, + { + "epoch": 0.43759630200308164, + "grad_norm": 1.4485833438326106, + "learning_rate": 6.240390205228464e-06, + "loss": 0.3443, + "step": 19880 + }, + { + "epoch": 0.43761831388950034, + "grad_norm": 1.7142046079174766, + "learning_rate": 6.240044888505714e-06, + "loss": 0.3058, + "step": 19881 + }, + { + "epoch": 0.437640325775919, + "grad_norm": 1.3490087999785978, + "learning_rate": 6.239699565480505e-06, + "loss": 0.2854, + "step": 19882 + }, + { + "epoch": 0.43766233766233764, + "grad_norm": 3.239737484831348, + "learning_rate": 6.239354236154588e-06, + "loss": 0.4254, + "step": 19883 + }, + { + "epoch": 0.43768434954875635, + "grad_norm": 1.7369792275705116, + "learning_rate": 6.239008900529721e-06, + "loss": 0.3766, + "step": 19884 + }, + { + "epoch": 0.437706361435175, + "grad_norm": 1.8263981602948252, + "learning_rate": 6.238663558607658e-06, + "loss": 0.3466, + "step": 19885 + }, + { + "epoch": 0.43772837332159364, + "grad_norm": 2.0250144375745935, + "learning_rate": 6.238318210390153e-06, + "loss": 0.3278, + "step": 19886 + }, + { + "epoch": 0.43775038520801235, + "grad_norm": 0.6623923076745339, + "learning_rate": 6.237972855878964e-06, + "loss": 0.5073, + "step": 19887 + }, + { + "epoch": 0.437772397094431, + "grad_norm": 1.6191741364495165, + "learning_rate": 6.2376274950758456e-06, + "loss": 0.3425, + "step": 19888 + }, + { + "epoch": 0.43779440898084965, + "grad_norm": 1.2869132826208574, + "learning_rate": 6.237282127982551e-06, + "loss": 0.3543, + "step": 19889 + }, + { + "epoch": 0.43781642086726835, + "grad_norm": 1.7270020778989439, + "learning_rate": 6.2369367546008365e-06, + "loss": 0.3975, + "step": 19890 + }, + { + "epoch": 0.437838432753687, + "grad_norm": 1.3774329807763563, + "learning_rate": 6.236591374932458e-06, + "loss": 0.3739, + "step": 19891 + }, + { + "epoch": 0.43786044464010565, + "grad_norm": 1.1626947712197853, + "learning_rate": 6.23624598897917e-06, + "loss": 0.2701, + "step": 19892 + }, + { + "epoch": 0.4378824565265243, + "grad_norm": 1.4033340632621676, + "learning_rate": 6.23590059674273e-06, + "loss": 0.3679, + "step": 19893 + }, + { + "epoch": 0.437904468412943, + "grad_norm": 0.630901125407864, + "learning_rate": 6.235555198224892e-06, + "loss": 0.4892, + "step": 19894 + }, + { + "epoch": 0.43792648029936165, + "grad_norm": 1.6472298066331672, + "learning_rate": 6.235209793427411e-06, + "loss": 0.3647, + "step": 19895 + }, + { + "epoch": 0.4379484921857803, + "grad_norm": 0.6201823942385273, + "learning_rate": 6.2348643823520404e-06, + "loss": 0.4892, + "step": 19896 + }, + { + "epoch": 0.437970504072199, + "grad_norm": 1.3074848238667192, + "learning_rate": 6.234518965000542e-06, + "loss": 0.33, + "step": 19897 + }, + { + "epoch": 0.43799251595861766, + "grad_norm": 1.4907514092327308, + "learning_rate": 6.234173541374666e-06, + "loss": 0.3556, + "step": 19898 + }, + { + "epoch": 0.4380145278450363, + "grad_norm": 0.588961239400469, + "learning_rate": 6.233828111476169e-06, + "loss": 0.4856, + "step": 19899 + }, + { + "epoch": 0.438036539731455, + "grad_norm": 0.5911341585037521, + "learning_rate": 6.233482675306808e-06, + "loss": 0.4926, + "step": 19900 + }, + { + "epoch": 0.43805855161787366, + "grad_norm": 1.1294363204946407, + "learning_rate": 6.233137232868338e-06, + "loss": 0.2368, + "step": 19901 + }, + { + "epoch": 0.4380805635042923, + "grad_norm": 0.6288041688139827, + "learning_rate": 6.232791784162514e-06, + "loss": 0.4931, + "step": 19902 + }, + { + "epoch": 0.43810257539071096, + "grad_norm": 1.4189278003837227, + "learning_rate": 6.232446329191093e-06, + "loss": 0.3653, + "step": 19903 + }, + { + "epoch": 0.43812458727712966, + "grad_norm": 1.4444245646688672, + "learning_rate": 6.232100867955829e-06, + "loss": 0.3377, + "step": 19904 + }, + { + "epoch": 0.4381465991635483, + "grad_norm": 1.9917801618661364, + "learning_rate": 6.2317554004584805e-06, + "loss": 0.4197, + "step": 19905 + }, + { + "epoch": 0.43816861104996696, + "grad_norm": 1.1788856740040201, + "learning_rate": 6.2314099267008e-06, + "loss": 0.2273, + "step": 19906 + }, + { + "epoch": 0.43819062293638567, + "grad_norm": 1.5178291993078896, + "learning_rate": 6.231064446684546e-06, + "loss": 0.3238, + "step": 19907 + }, + { + "epoch": 0.4382126348228043, + "grad_norm": 1.3706705158768748, + "learning_rate": 6.230718960411474e-06, + "loss": 0.2719, + "step": 19908 + }, + { + "epoch": 0.43823464670922296, + "grad_norm": 1.8538048712849047, + "learning_rate": 6.230373467883337e-06, + "loss": 0.38, + "step": 19909 + }, + { + "epoch": 0.43825665859564167, + "grad_norm": 1.580360005579734, + "learning_rate": 6.230027969101895e-06, + "loss": 0.3691, + "step": 19910 + }, + { + "epoch": 0.4382786704820603, + "grad_norm": 1.9224730711237419, + "learning_rate": 6.2296824640689025e-06, + "loss": 0.3463, + "step": 19911 + }, + { + "epoch": 0.43830068236847897, + "grad_norm": 1.6216129864097664, + "learning_rate": 6.229336952786114e-06, + "loss": 0.3529, + "step": 19912 + }, + { + "epoch": 0.43832269425489767, + "grad_norm": 1.7913310436984695, + "learning_rate": 6.2289914352552885e-06, + "loss": 0.3166, + "step": 19913 + }, + { + "epoch": 0.4383447061413163, + "grad_norm": 1.9748309941365156, + "learning_rate": 6.228645911478178e-06, + "loss": 0.2692, + "step": 19914 + }, + { + "epoch": 0.43836671802773497, + "grad_norm": 0.6599775518417406, + "learning_rate": 6.2283003814565446e-06, + "loss": 0.4689, + "step": 19915 + }, + { + "epoch": 0.4383887299141536, + "grad_norm": 1.4554853854188627, + "learning_rate": 6.227954845192139e-06, + "loss": 0.3723, + "step": 19916 + }, + { + "epoch": 0.4384107418005723, + "grad_norm": 1.5371558251766542, + "learning_rate": 6.227609302686718e-06, + "loss": 0.2502, + "step": 19917 + }, + { + "epoch": 0.438432753686991, + "grad_norm": 1.7451024576982663, + "learning_rate": 6.22726375394204e-06, + "loss": 0.372, + "step": 19918 + }, + { + "epoch": 0.4384547655734096, + "grad_norm": 3.2079144698926014, + "learning_rate": 6.22691819895986e-06, + "loss": 0.3727, + "step": 19919 + }, + { + "epoch": 0.4384767774598283, + "grad_norm": 0.6325271232268207, + "learning_rate": 6.226572637741934e-06, + "loss": 0.4821, + "step": 19920 + }, + { + "epoch": 0.438498789346247, + "grad_norm": 1.3499930656778258, + "learning_rate": 6.226227070290021e-06, + "loss": 0.3436, + "step": 19921 + }, + { + "epoch": 0.4385208012326656, + "grad_norm": 0.634879963646129, + "learning_rate": 6.225881496605872e-06, + "loss": 0.4858, + "step": 19922 + }, + { + "epoch": 0.43854281311908433, + "grad_norm": 1.493302318949625, + "learning_rate": 6.225535916691248e-06, + "loss": 0.2941, + "step": 19923 + }, + { + "epoch": 0.438564825005503, + "grad_norm": 1.493514523430016, + "learning_rate": 6.2251903305479045e-06, + "loss": 0.2998, + "step": 19924 + }, + { + "epoch": 0.4385868368919216, + "grad_norm": 1.7966489859873456, + "learning_rate": 6.224844738177597e-06, + "loss": 0.3094, + "step": 19925 + }, + { + "epoch": 0.4386088487783403, + "grad_norm": 1.612672112629226, + "learning_rate": 6.224499139582082e-06, + "loss": 0.2625, + "step": 19926 + }, + { + "epoch": 0.438630860664759, + "grad_norm": 6.870456548309375, + "learning_rate": 6.2241535347631155e-06, + "loss": 0.4094, + "step": 19927 + }, + { + "epoch": 0.43865287255117763, + "grad_norm": 1.3540134436486113, + "learning_rate": 6.223807923722454e-06, + "loss": 0.2621, + "step": 19928 + }, + { + "epoch": 0.4386748844375963, + "grad_norm": 1.9730543313929194, + "learning_rate": 6.2234623064618575e-06, + "loss": 0.3736, + "step": 19929 + }, + { + "epoch": 0.438696896324015, + "grad_norm": 1.5408459320398937, + "learning_rate": 6.223116682983078e-06, + "loss": 0.4314, + "step": 19930 + }, + { + "epoch": 0.43871890821043363, + "grad_norm": 1.73303554643164, + "learning_rate": 6.222771053287875e-06, + "loss": 0.2924, + "step": 19931 + }, + { + "epoch": 0.4387409200968523, + "grad_norm": 1.134449512677606, + "learning_rate": 6.222425417378003e-06, + "loss": 0.2735, + "step": 19932 + }, + { + "epoch": 0.438762931983271, + "grad_norm": 0.6082119980747024, + "learning_rate": 6.222079775255222e-06, + "loss": 0.4615, + "step": 19933 + }, + { + "epoch": 0.43878494386968964, + "grad_norm": 2.3250948575421724, + "learning_rate": 6.221734126921285e-06, + "loss": 0.3375, + "step": 19934 + }, + { + "epoch": 0.4388069557561083, + "grad_norm": 1.8433374954993793, + "learning_rate": 6.2213884723779494e-06, + "loss": 0.38, + "step": 19935 + }, + { + "epoch": 0.438828967642527, + "grad_norm": 4.807559341842866, + "learning_rate": 6.2210428116269736e-06, + "loss": 0.2925, + "step": 19936 + }, + { + "epoch": 0.43885097952894564, + "grad_norm": 0.6388769910466342, + "learning_rate": 6.220697144670113e-06, + "loss": 0.486, + "step": 19937 + }, + { + "epoch": 0.4388729914153643, + "grad_norm": 1.4258696995964126, + "learning_rate": 6.220351471509127e-06, + "loss": 0.3189, + "step": 19938 + }, + { + "epoch": 0.43889500330178294, + "grad_norm": 0.6208709200313876, + "learning_rate": 6.220005792145769e-06, + "loss": 0.4744, + "step": 19939 + }, + { + "epoch": 0.43891701518820164, + "grad_norm": 2.1590542984085443, + "learning_rate": 6.219660106581797e-06, + "loss": 0.3019, + "step": 19940 + }, + { + "epoch": 0.4389390270746203, + "grad_norm": 1.2751190918650748, + "learning_rate": 6.21931441481897e-06, + "loss": 0.3198, + "step": 19941 + }, + { + "epoch": 0.43896103896103894, + "grad_norm": 1.8872991438384925, + "learning_rate": 6.218968716859042e-06, + "loss": 0.2473, + "step": 19942 + }, + { + "epoch": 0.43898305084745765, + "grad_norm": 1.552848523500631, + "learning_rate": 6.218623012703772e-06, + "loss": 0.3324, + "step": 19943 + }, + { + "epoch": 0.4390050627338763, + "grad_norm": 0.6203996604749034, + "learning_rate": 6.2182773023549156e-06, + "loss": 0.4741, + "step": 19944 + }, + { + "epoch": 0.43902707462029494, + "grad_norm": 1.6732394284805352, + "learning_rate": 6.217931585814231e-06, + "loss": 0.264, + "step": 19945 + }, + { + "epoch": 0.43904908650671365, + "grad_norm": 1.1870086291308257, + "learning_rate": 6.217585863083475e-06, + "loss": 0.3635, + "step": 19946 + }, + { + "epoch": 0.4390710983931323, + "grad_norm": 1.285948643766468, + "learning_rate": 6.217240134164404e-06, + "loss": 0.26, + "step": 19947 + }, + { + "epoch": 0.43909311027955095, + "grad_norm": 1.7077246153239156, + "learning_rate": 6.216894399058777e-06, + "loss": 0.3359, + "step": 19948 + }, + { + "epoch": 0.43911512216596965, + "grad_norm": 1.5234986486409636, + "learning_rate": 6.216548657768349e-06, + "loss": 0.3642, + "step": 19949 + }, + { + "epoch": 0.4391371340523883, + "grad_norm": 1.3243319404694869, + "learning_rate": 6.216202910294877e-06, + "loss": 0.2822, + "step": 19950 + }, + { + "epoch": 0.43915914593880695, + "grad_norm": 1.5801032692809003, + "learning_rate": 6.2158571566401215e-06, + "loss": 0.4004, + "step": 19951 + }, + { + "epoch": 0.4391811578252256, + "grad_norm": 1.918960788565516, + "learning_rate": 6.215511396805838e-06, + "loss": 0.3627, + "step": 19952 + }, + { + "epoch": 0.4392031697116443, + "grad_norm": 0.6796548093955243, + "learning_rate": 6.215165630793781e-06, + "loss": 0.4585, + "step": 19953 + }, + { + "epoch": 0.43922518159806295, + "grad_norm": 1.4653550418798857, + "learning_rate": 6.214819858605712e-06, + "loss": 0.326, + "step": 19954 + }, + { + "epoch": 0.4392471934844816, + "grad_norm": 1.515443667310614, + "learning_rate": 6.214474080243387e-06, + "loss": 0.2829, + "step": 19955 + }, + { + "epoch": 0.4392692053709003, + "grad_norm": 1.2481324549948871, + "learning_rate": 6.214128295708561e-06, + "loss": 0.3168, + "step": 19956 + }, + { + "epoch": 0.43929121725731896, + "grad_norm": 1.6120514187316397, + "learning_rate": 6.213782505002995e-06, + "loss": 0.2394, + "step": 19957 + }, + { + "epoch": 0.4393132291437376, + "grad_norm": 1.2779235051181335, + "learning_rate": 6.213436708128444e-06, + "loss": 0.3168, + "step": 19958 + }, + { + "epoch": 0.4393352410301563, + "grad_norm": 1.9629005247238847, + "learning_rate": 6.213090905086667e-06, + "loss": 0.2683, + "step": 19959 + }, + { + "epoch": 0.43935725291657496, + "grad_norm": 1.8855843434570083, + "learning_rate": 6.212745095879422e-06, + "loss": 0.2778, + "step": 19960 + }, + { + "epoch": 0.4393792648029936, + "grad_norm": 2.382606605433226, + "learning_rate": 6.2123992805084645e-06, + "loss": 0.3339, + "step": 19961 + }, + { + "epoch": 0.43940127668941226, + "grad_norm": 1.7717747113626716, + "learning_rate": 6.212053458975553e-06, + "loss": 0.3306, + "step": 19962 + }, + { + "epoch": 0.43942328857583096, + "grad_norm": 1.4822037238039403, + "learning_rate": 6.211707631282444e-06, + "loss": 0.3341, + "step": 19963 + }, + { + "epoch": 0.4394453004622496, + "grad_norm": 1.5248825756902327, + "learning_rate": 6.211361797430899e-06, + "loss": 0.3053, + "step": 19964 + }, + { + "epoch": 0.43946731234866826, + "grad_norm": 1.2420685791452206, + "learning_rate": 6.211015957422672e-06, + "loss": 0.3398, + "step": 19965 + }, + { + "epoch": 0.43948932423508696, + "grad_norm": 2.0100035307415114, + "learning_rate": 6.210670111259521e-06, + "loss": 0.3491, + "step": 19966 + }, + { + "epoch": 0.4395113361215056, + "grad_norm": 1.4061016287928934, + "learning_rate": 6.210324258943205e-06, + "loss": 0.2756, + "step": 19967 + }, + { + "epoch": 0.43953334800792426, + "grad_norm": 1.708083007201764, + "learning_rate": 6.209978400475481e-06, + "loss": 0.3637, + "step": 19968 + }, + { + "epoch": 0.43955535989434297, + "grad_norm": 1.1841010282034208, + "learning_rate": 6.209632535858108e-06, + "loss": 0.2611, + "step": 19969 + }, + { + "epoch": 0.4395773717807616, + "grad_norm": 4.294097336891283, + "learning_rate": 6.209286665092844e-06, + "loss": 0.3702, + "step": 19970 + }, + { + "epoch": 0.43959938366718027, + "grad_norm": 0.6428017082885596, + "learning_rate": 6.208940788181443e-06, + "loss": 0.5244, + "step": 19971 + }, + { + "epoch": 0.43962139555359897, + "grad_norm": 2.2097664629402143, + "learning_rate": 6.208594905125668e-06, + "loss": 0.3058, + "step": 19972 + }, + { + "epoch": 0.4396434074400176, + "grad_norm": 0.6420610759390204, + "learning_rate": 6.208249015927274e-06, + "loss": 0.5079, + "step": 19973 + }, + { + "epoch": 0.43966541932643627, + "grad_norm": 1.738154587013818, + "learning_rate": 6.207903120588019e-06, + "loss": 0.3501, + "step": 19974 + }, + { + "epoch": 0.4396874312128549, + "grad_norm": 1.3241369129582843, + "learning_rate": 6.207557219109662e-06, + "loss": 0.2657, + "step": 19975 + }, + { + "epoch": 0.4397094430992736, + "grad_norm": 1.6854609238117364, + "learning_rate": 6.207211311493959e-06, + "loss": 0.4184, + "step": 19976 + }, + { + "epoch": 0.43973145498569227, + "grad_norm": 1.5734994116266199, + "learning_rate": 6.206865397742672e-06, + "loss": 0.3154, + "step": 19977 + }, + { + "epoch": 0.4397534668721109, + "grad_norm": 1.0639736615137587, + "learning_rate": 6.206519477857556e-06, + "loss": 0.3148, + "step": 19978 + }, + { + "epoch": 0.4397754787585296, + "grad_norm": 2.472293356972957, + "learning_rate": 6.20617355184037e-06, + "loss": 0.324, + "step": 19979 + }, + { + "epoch": 0.4397974906449483, + "grad_norm": 1.9680842791666597, + "learning_rate": 6.205827619692872e-06, + "loss": 0.3147, + "step": 19980 + }, + { + "epoch": 0.4398195025313669, + "grad_norm": 1.542831772371925, + "learning_rate": 6.20548168141682e-06, + "loss": 0.3371, + "step": 19981 + }, + { + "epoch": 0.43984151441778563, + "grad_norm": 1.4993344286732297, + "learning_rate": 6.205135737013972e-06, + "loss": 0.2883, + "step": 19982 + }, + { + "epoch": 0.4398635263042043, + "grad_norm": 1.5413480236753179, + "learning_rate": 6.204789786486087e-06, + "loss": 0.2725, + "step": 19983 + }, + { + "epoch": 0.4398855381906229, + "grad_norm": 1.4199102973399709, + "learning_rate": 6.2044438298349245e-06, + "loss": 0.2663, + "step": 19984 + }, + { + "epoch": 0.4399075500770416, + "grad_norm": 2.606566331645207, + "learning_rate": 6.20409786706224e-06, + "loss": 0.344, + "step": 19985 + }, + { + "epoch": 0.4399295619634603, + "grad_norm": 1.3834034338150898, + "learning_rate": 6.2037518981697916e-06, + "loss": 0.2256, + "step": 19986 + }, + { + "epoch": 0.43995157384987893, + "grad_norm": 2.8454479041799723, + "learning_rate": 6.20340592315934e-06, + "loss": 0.3857, + "step": 19987 + }, + { + "epoch": 0.4399735857362976, + "grad_norm": 1.6619486395348133, + "learning_rate": 6.203059942032645e-06, + "loss": 0.3432, + "step": 19988 + }, + { + "epoch": 0.4399955976227163, + "grad_norm": 0.6643713045388249, + "learning_rate": 6.20271395479146e-06, + "loss": 0.4865, + "step": 19989 + }, + { + "epoch": 0.44001760950913493, + "grad_norm": 3.7974424855655884, + "learning_rate": 6.202367961437548e-06, + "loss": 0.4003, + "step": 19990 + }, + { + "epoch": 0.4400396213955536, + "grad_norm": 1.3345548349499807, + "learning_rate": 6.202021961972665e-06, + "loss": 0.3121, + "step": 19991 + }, + { + "epoch": 0.4400616332819723, + "grad_norm": 1.5524648168452602, + "learning_rate": 6.2016759563985695e-06, + "loss": 0.4319, + "step": 19992 + }, + { + "epoch": 0.44008364516839094, + "grad_norm": 1.5411895109930478, + "learning_rate": 6.201329944717024e-06, + "loss": 0.3698, + "step": 19993 + }, + { + "epoch": 0.4401056570548096, + "grad_norm": 2.132655291135161, + "learning_rate": 6.20098392692978e-06, + "loss": 0.3535, + "step": 19994 + }, + { + "epoch": 0.4401276689412283, + "grad_norm": 0.6342287120925291, + "learning_rate": 6.200637903038602e-06, + "loss": 0.4881, + "step": 19995 + }, + { + "epoch": 0.44014968082764694, + "grad_norm": 1.6092713120671311, + "learning_rate": 6.200291873045246e-06, + "loss": 0.3323, + "step": 19996 + }, + { + "epoch": 0.4401716927140656, + "grad_norm": 0.631641272672932, + "learning_rate": 6.199945836951473e-06, + "loss": 0.5229, + "step": 19997 + }, + { + "epoch": 0.44019370460048424, + "grad_norm": 0.6279621592489706, + "learning_rate": 6.1995997947590396e-06, + "loss": 0.5024, + "step": 19998 + }, + { + "epoch": 0.44021571648690294, + "grad_norm": 0.6026587486876592, + "learning_rate": 6.1992537464697024e-06, + "loss": 0.478, + "step": 19999 + }, + { + "epoch": 0.4402377283733216, + "grad_norm": 2.4852439359599625, + "learning_rate": 6.1989076920852255e-06, + "loss": 0.2649, + "step": 20000 + }, + { + "epoch": 0.44025974025974024, + "grad_norm": 1.6865558173739803, + "learning_rate": 6.198561631607364e-06, + "loss": 0.3437, + "step": 20001 + }, + { + "epoch": 0.44028175214615894, + "grad_norm": 1.4346148060205068, + "learning_rate": 6.1982155650378774e-06, + "loss": 0.3048, + "step": 20002 + }, + { + "epoch": 0.4403037640325776, + "grad_norm": 1.6085146450454875, + "learning_rate": 6.197869492378526e-06, + "loss": 0.3185, + "step": 20003 + }, + { + "epoch": 0.44032577591899624, + "grad_norm": 2.343155722772675, + "learning_rate": 6.1975234136310655e-06, + "loss": 0.3429, + "step": 20004 + }, + { + "epoch": 0.44034778780541495, + "grad_norm": 0.6571119183147431, + "learning_rate": 6.197177328797258e-06, + "loss": 0.4749, + "step": 20005 + }, + { + "epoch": 0.4403697996918336, + "grad_norm": 20.062103593542176, + "learning_rate": 6.196831237878863e-06, + "loss": 0.3761, + "step": 20006 + }, + { + "epoch": 0.44039181157825225, + "grad_norm": 1.3367273474928068, + "learning_rate": 6.196485140877636e-06, + "loss": 0.2964, + "step": 20007 + }, + { + "epoch": 0.4404138234646709, + "grad_norm": 0.6680509159613401, + "learning_rate": 6.1961390377953376e-06, + "loss": 0.5049, + "step": 20008 + }, + { + "epoch": 0.4404358353510896, + "grad_norm": 1.7556619717126714, + "learning_rate": 6.195792928633728e-06, + "loss": 0.3182, + "step": 20009 + }, + { + "epoch": 0.44045784723750825, + "grad_norm": 2.564271744717032, + "learning_rate": 6.195446813394565e-06, + "loss": 0.3524, + "step": 20010 + }, + { + "epoch": 0.4404798591239269, + "grad_norm": 1.1993218734200926, + "learning_rate": 6.19510069207961e-06, + "loss": 0.2711, + "step": 20011 + }, + { + "epoch": 0.4405018710103456, + "grad_norm": 1.3502495654409157, + "learning_rate": 6.194754564690617e-06, + "loss": 0.3417, + "step": 20012 + }, + { + "epoch": 0.44052388289676425, + "grad_norm": 1.5286650090792797, + "learning_rate": 6.19440843122935e-06, + "loss": 0.3605, + "step": 20013 + }, + { + "epoch": 0.4405458947831829, + "grad_norm": 1.855070141351342, + "learning_rate": 6.194062291697567e-06, + "loss": 0.4096, + "step": 20014 + }, + { + "epoch": 0.4405679066696016, + "grad_norm": 1.4188137918462655, + "learning_rate": 6.193716146097027e-06, + "loss": 0.3012, + "step": 20015 + }, + { + "epoch": 0.44058991855602025, + "grad_norm": 1.4099506908763726, + "learning_rate": 6.193369994429489e-06, + "loss": 0.3459, + "step": 20016 + }, + { + "epoch": 0.4406119304424389, + "grad_norm": 2.4342205194863236, + "learning_rate": 6.193023836696711e-06, + "loss": 0.3943, + "step": 20017 + }, + { + "epoch": 0.4406339423288576, + "grad_norm": 1.5539231397422801, + "learning_rate": 6.192677672900455e-06, + "loss": 0.3032, + "step": 20018 + }, + { + "epoch": 0.44065595421527626, + "grad_norm": 0.6491549218674776, + "learning_rate": 6.192331503042479e-06, + "loss": 0.5014, + "step": 20019 + }, + { + "epoch": 0.4406779661016949, + "grad_norm": 0.6162681340616941, + "learning_rate": 6.1919853271245435e-06, + "loss": 0.4458, + "step": 20020 + }, + { + "epoch": 0.44069997798811356, + "grad_norm": 1.5062719473217834, + "learning_rate": 6.1916391451484046e-06, + "loss": 0.3157, + "step": 20021 + }, + { + "epoch": 0.44072198987453226, + "grad_norm": 1.2025974777191013, + "learning_rate": 6.191292957115825e-06, + "loss": 0.2462, + "step": 20022 + }, + { + "epoch": 0.4407440017609509, + "grad_norm": 1.77943131790662, + "learning_rate": 6.1909467630285634e-06, + "loss": 0.3541, + "step": 20023 + }, + { + "epoch": 0.44076601364736956, + "grad_norm": 0.6194762319812325, + "learning_rate": 6.1906005628883815e-06, + "loss": 0.5023, + "step": 20024 + }, + { + "epoch": 0.44078802553378826, + "grad_norm": 1.6030102101020838, + "learning_rate": 6.190254356697033e-06, + "loss": 0.3582, + "step": 20025 + }, + { + "epoch": 0.4408100374202069, + "grad_norm": 0.6089584102877273, + "learning_rate": 6.189908144456283e-06, + "loss": 0.5111, + "step": 20026 + }, + { + "epoch": 0.44083204930662556, + "grad_norm": 0.6080737363229926, + "learning_rate": 6.189561926167889e-06, + "loss": 0.4637, + "step": 20027 + }, + { + "epoch": 0.44085406119304427, + "grad_norm": 1.7146868536257704, + "learning_rate": 6.189215701833608e-06, + "loss": 0.3006, + "step": 20028 + }, + { + "epoch": 0.4408760730794629, + "grad_norm": 0.6005603259076516, + "learning_rate": 6.1888694714552066e-06, + "loss": 0.4801, + "step": 20029 + }, + { + "epoch": 0.44089808496588156, + "grad_norm": 1.902549573991875, + "learning_rate": 6.1885232350344385e-06, + "loss": 0.2, + "step": 20030 + }, + { + "epoch": 0.44092009685230027, + "grad_norm": 1.7798745393685214, + "learning_rate": 6.188176992573064e-06, + "loss": 0.4664, + "step": 20031 + }, + { + "epoch": 0.4409421087387189, + "grad_norm": 1.2825342409582399, + "learning_rate": 6.187830744072846e-06, + "loss": 0.3648, + "step": 20032 + }, + { + "epoch": 0.44096412062513757, + "grad_norm": 1.2726518407748384, + "learning_rate": 6.187484489535541e-06, + "loss": 0.2537, + "step": 20033 + }, + { + "epoch": 0.4409861325115562, + "grad_norm": 1.1464737657446067, + "learning_rate": 6.187138228962911e-06, + "loss": 0.3016, + "step": 20034 + }, + { + "epoch": 0.4410081443979749, + "grad_norm": 1.64236194104382, + "learning_rate": 6.1867919623567144e-06, + "loss": 0.2829, + "step": 20035 + }, + { + "epoch": 0.44103015628439357, + "grad_norm": 2.9733779948317354, + "learning_rate": 6.186445689718712e-06, + "loss": 0.3013, + "step": 20036 + }, + { + "epoch": 0.4410521681708122, + "grad_norm": 2.1125872958305107, + "learning_rate": 6.186099411050663e-06, + "loss": 0.3024, + "step": 20037 + }, + { + "epoch": 0.4410741800572309, + "grad_norm": 1.8481881397612072, + "learning_rate": 6.185753126354329e-06, + "loss": 0.2096, + "step": 20038 + }, + { + "epoch": 0.4410961919436496, + "grad_norm": 0.6605133257813223, + "learning_rate": 6.185406835631468e-06, + "loss": 0.4902, + "step": 20039 + }, + { + "epoch": 0.4411182038300682, + "grad_norm": 1.2696357356456385, + "learning_rate": 6.18506053888384e-06, + "loss": 0.3098, + "step": 20040 + }, + { + "epoch": 0.4411402157164869, + "grad_norm": 1.8814685310761643, + "learning_rate": 6.1847142361132075e-06, + "loss": 0.3454, + "step": 20041 + }, + { + "epoch": 0.4411622276029056, + "grad_norm": 1.7288534482765563, + "learning_rate": 6.184367927321328e-06, + "loss": 0.3762, + "step": 20042 + }, + { + "epoch": 0.4411842394893242, + "grad_norm": 1.5428298861502399, + "learning_rate": 6.184021612509962e-06, + "loss": 0.2696, + "step": 20043 + }, + { + "epoch": 0.4412062513757429, + "grad_norm": 1.53389931736021, + "learning_rate": 6.1836752916808705e-06, + "loss": 0.3595, + "step": 20044 + }, + { + "epoch": 0.4412282632621616, + "grad_norm": 1.3038390823892205, + "learning_rate": 6.183328964835813e-06, + "loss": 0.271, + "step": 20045 + }, + { + "epoch": 0.44125027514858023, + "grad_norm": 1.4352748608837733, + "learning_rate": 6.1829826319765495e-06, + "loss": 0.2746, + "step": 20046 + }, + { + "epoch": 0.4412722870349989, + "grad_norm": 1.7299227354075002, + "learning_rate": 6.1826362931048425e-06, + "loss": 0.3571, + "step": 20047 + }, + { + "epoch": 0.4412942989214176, + "grad_norm": 0.6146559307849272, + "learning_rate": 6.182289948222448e-06, + "loss": 0.4615, + "step": 20048 + }, + { + "epoch": 0.44131631080783623, + "grad_norm": 2.5252486149845534, + "learning_rate": 6.18194359733113e-06, + "loss": 0.2679, + "step": 20049 + }, + { + "epoch": 0.4413383226942549, + "grad_norm": 0.6527854098471405, + "learning_rate": 6.181597240432649e-06, + "loss": 0.4948, + "step": 20050 + }, + { + "epoch": 0.4413603345806736, + "grad_norm": 3.7605026628087788, + "learning_rate": 6.18125087752876e-06, + "loss": 0.3038, + "step": 20051 + }, + { + "epoch": 0.44138234646709223, + "grad_norm": 2.9487658270077617, + "learning_rate": 6.18090450862123e-06, + "loss": 0.3061, + "step": 20052 + }, + { + "epoch": 0.4414043583535109, + "grad_norm": 1.4498610677754349, + "learning_rate": 6.180558133711816e-06, + "loss": 0.3414, + "step": 20053 + }, + { + "epoch": 0.4414263702399296, + "grad_norm": 1.6516811384140893, + "learning_rate": 6.180211752802279e-06, + "loss": 0.3055, + "step": 20054 + }, + { + "epoch": 0.44144838212634824, + "grad_norm": 1.5990650033160676, + "learning_rate": 6.17986536589438e-06, + "loss": 0.3125, + "step": 20055 + }, + { + "epoch": 0.4414703940127669, + "grad_norm": 1.2932824661135496, + "learning_rate": 6.179518972989877e-06, + "loss": 0.364, + "step": 20056 + }, + { + "epoch": 0.44149240589918554, + "grad_norm": 1.2248184064424907, + "learning_rate": 6.179172574090535e-06, + "loss": 0.2962, + "step": 20057 + }, + { + "epoch": 0.44151441778560424, + "grad_norm": 1.6424719910396033, + "learning_rate": 6.1788261691981096e-06, + "loss": 0.2661, + "step": 20058 + }, + { + "epoch": 0.4415364296720229, + "grad_norm": 0.6335888336005606, + "learning_rate": 6.178479758314365e-06, + "loss": 0.4748, + "step": 20059 + }, + { + "epoch": 0.44155844155844154, + "grad_norm": 1.4252474179915033, + "learning_rate": 6.178133341441062e-06, + "loss": 0.3888, + "step": 20060 + }, + { + "epoch": 0.44158045344486024, + "grad_norm": 0.6159427862116854, + "learning_rate": 6.177786918579959e-06, + "loss": 0.4844, + "step": 20061 + }, + { + "epoch": 0.4416024653312789, + "grad_norm": 0.6046082834616269, + "learning_rate": 6.177440489732817e-06, + "loss": 0.4606, + "step": 20062 + }, + { + "epoch": 0.44162447721769754, + "grad_norm": 1.2895514545570648, + "learning_rate": 6.177094054901398e-06, + "loss": 0.3321, + "step": 20063 + }, + { + "epoch": 0.44164648910411625, + "grad_norm": 1.5986347087379351, + "learning_rate": 6.17674761408746e-06, + "loss": 0.3142, + "step": 20064 + }, + { + "epoch": 0.4416685009905349, + "grad_norm": 1.1690517074984381, + "learning_rate": 6.17640116729277e-06, + "loss": 0.3217, + "step": 20065 + }, + { + "epoch": 0.44169051287695354, + "grad_norm": 4.022552758565086, + "learning_rate": 6.1760547145190795e-06, + "loss": 0.3188, + "step": 20066 + }, + { + "epoch": 0.4417125247633722, + "grad_norm": 2.404824818076216, + "learning_rate": 6.175708255768157e-06, + "loss": 0.251, + "step": 20067 + }, + { + "epoch": 0.4417345366497909, + "grad_norm": 1.1229847153708679, + "learning_rate": 6.175361791041761e-06, + "loss": 0.2189, + "step": 20068 + }, + { + "epoch": 0.44175654853620955, + "grad_norm": 6.701990899972851, + "learning_rate": 6.175015320341651e-06, + "loss": 0.3158, + "step": 20069 + }, + { + "epoch": 0.4417785604226282, + "grad_norm": 1.0357425677165601, + "learning_rate": 6.1746688436695915e-06, + "loss": 0.2169, + "step": 20070 + }, + { + "epoch": 0.4418005723090469, + "grad_norm": 2.2009933556387486, + "learning_rate": 6.1743223610273384e-06, + "loss": 0.3237, + "step": 20071 + }, + { + "epoch": 0.44182258419546555, + "grad_norm": 2.4402501230065132, + "learning_rate": 6.173975872416656e-06, + "loss": 0.4191, + "step": 20072 + }, + { + "epoch": 0.4418445960818842, + "grad_norm": 1.9317648236831988, + "learning_rate": 6.173629377839305e-06, + "loss": 0.3577, + "step": 20073 + }, + { + "epoch": 0.4418666079683029, + "grad_norm": 1.3854728296134124, + "learning_rate": 6.173282877297046e-06, + "loss": 0.298, + "step": 20074 + }, + { + "epoch": 0.44188861985472155, + "grad_norm": 1.3878612058994653, + "learning_rate": 6.17293637079164e-06, + "loss": 0.3491, + "step": 20075 + }, + { + "epoch": 0.4419106317411402, + "grad_norm": 2.135555468472072, + "learning_rate": 6.1725898583248476e-06, + "loss": 0.3491, + "step": 20076 + }, + { + "epoch": 0.4419326436275589, + "grad_norm": 1.2408567462615752, + "learning_rate": 6.1722433398984315e-06, + "loss": 0.2789, + "step": 20077 + }, + { + "epoch": 0.44195465551397756, + "grad_norm": 1.6856181063240123, + "learning_rate": 6.171896815514151e-06, + "loss": 0.2914, + "step": 20078 + }, + { + "epoch": 0.4419766674003962, + "grad_norm": 1.2987733620336894, + "learning_rate": 6.17155028517377e-06, + "loss": 0.2842, + "step": 20079 + }, + { + "epoch": 0.44199867928681486, + "grad_norm": 1.5561297015930315, + "learning_rate": 6.1712037488790465e-06, + "loss": 0.3897, + "step": 20080 + }, + { + "epoch": 0.44202069117323356, + "grad_norm": 1.3755162194167088, + "learning_rate": 6.170857206631743e-06, + "loss": 0.332, + "step": 20081 + }, + { + "epoch": 0.4420427030596522, + "grad_norm": 2.2633989964956065, + "learning_rate": 6.170510658433621e-06, + "loss": 0.3015, + "step": 20082 + }, + { + "epoch": 0.44206471494607086, + "grad_norm": 1.8888984800755544, + "learning_rate": 6.170164104286444e-06, + "loss": 0.3882, + "step": 20083 + }, + { + "epoch": 0.44208672683248956, + "grad_norm": 1.6097662913999022, + "learning_rate": 6.169817544191967e-06, + "loss": 0.3477, + "step": 20084 + }, + { + "epoch": 0.4421087387189082, + "grad_norm": 0.6822899413267238, + "learning_rate": 6.169470978151959e-06, + "loss": 0.5071, + "step": 20085 + }, + { + "epoch": 0.44213075060532686, + "grad_norm": 1.269227629062923, + "learning_rate": 6.1691244061681764e-06, + "loss": 0.2997, + "step": 20086 + }, + { + "epoch": 0.44215276249174557, + "grad_norm": 1.4819771486575068, + "learning_rate": 6.168777828242382e-06, + "loss": 0.3096, + "step": 20087 + }, + { + "epoch": 0.4421747743781642, + "grad_norm": 1.357972832434578, + "learning_rate": 6.168431244376338e-06, + "loss": 0.279, + "step": 20088 + }, + { + "epoch": 0.44219678626458286, + "grad_norm": 1.2948453578227597, + "learning_rate": 6.168084654571804e-06, + "loss": 0.3012, + "step": 20089 + }, + { + "epoch": 0.4422187981510015, + "grad_norm": 1.8668901774056146, + "learning_rate": 6.167738058830545e-06, + "loss": 0.3892, + "step": 20090 + }, + { + "epoch": 0.4422408100374202, + "grad_norm": 0.6058932909315758, + "learning_rate": 6.1673914571543195e-06, + "loss": 0.4802, + "step": 20091 + }, + { + "epoch": 0.44226282192383887, + "grad_norm": 1.812591253481365, + "learning_rate": 6.167044849544888e-06, + "loss": 0.3523, + "step": 20092 + }, + { + "epoch": 0.4422848338102575, + "grad_norm": 0.6252003786455061, + "learning_rate": 6.166698236004017e-06, + "loss": 0.4823, + "step": 20093 + }, + { + "epoch": 0.4423068456966762, + "grad_norm": 1.4315165782234553, + "learning_rate": 6.166351616533464e-06, + "loss": 0.3596, + "step": 20094 + }, + { + "epoch": 0.44232885758309487, + "grad_norm": 1.4478150908108416, + "learning_rate": 6.166004991134991e-06, + "loss": 0.3146, + "step": 20095 + }, + { + "epoch": 0.4423508694695135, + "grad_norm": 1.6763809323127692, + "learning_rate": 6.165658359810361e-06, + "loss": 0.2948, + "step": 20096 + }, + { + "epoch": 0.4423728813559322, + "grad_norm": 1.620215580968843, + "learning_rate": 6.1653117225613355e-06, + "loss": 0.3469, + "step": 20097 + }, + { + "epoch": 0.4423948932423509, + "grad_norm": 1.8454198151085763, + "learning_rate": 6.164965079389676e-06, + "loss": 0.3944, + "step": 20098 + }, + { + "epoch": 0.4424169051287695, + "grad_norm": 2.658585703710601, + "learning_rate": 6.1646184302971455e-06, + "loss": 0.3217, + "step": 20099 + }, + { + "epoch": 0.4424389170151882, + "grad_norm": 0.6292546266679944, + "learning_rate": 6.164271775285503e-06, + "loss": 0.4822, + "step": 20100 + }, + { + "epoch": 0.4424609289016069, + "grad_norm": 1.2597665198344181, + "learning_rate": 6.163925114356513e-06, + "loss": 0.2819, + "step": 20101 + }, + { + "epoch": 0.4424829407880255, + "grad_norm": 0.5972518920053667, + "learning_rate": 6.1635784475119366e-06, + "loss": 0.4979, + "step": 20102 + }, + { + "epoch": 0.4425049526744442, + "grad_norm": 2.0571425138849087, + "learning_rate": 6.163231774753535e-06, + "loss": 0.2544, + "step": 20103 + }, + { + "epoch": 0.4425269645608629, + "grad_norm": 1.287795198861094, + "learning_rate": 6.162885096083072e-06, + "loss": 0.3078, + "step": 20104 + }, + { + "epoch": 0.44254897644728153, + "grad_norm": 1.3548715609658848, + "learning_rate": 6.162538411502305e-06, + "loss": 0.3323, + "step": 20105 + }, + { + "epoch": 0.4425709883337002, + "grad_norm": 1.2951297039106664, + "learning_rate": 6.1621917210130045e-06, + "loss": 0.3155, + "step": 20106 + }, + { + "epoch": 0.4425930002201189, + "grad_norm": 1.4397897267348363, + "learning_rate": 6.161845024616923e-06, + "loss": 0.3613, + "step": 20107 + }, + { + "epoch": 0.44261501210653753, + "grad_norm": 1.9827209841878488, + "learning_rate": 6.161498322315829e-06, + "loss": 0.3026, + "step": 20108 + }, + { + "epoch": 0.4426370239929562, + "grad_norm": 1.6334106002366524, + "learning_rate": 6.161151614111481e-06, + "loss": 0.3029, + "step": 20109 + }, + { + "epoch": 0.4426590358793749, + "grad_norm": 8.002374410937453, + "learning_rate": 6.160804900005643e-06, + "loss": 0.434, + "step": 20110 + }, + { + "epoch": 0.44268104776579353, + "grad_norm": 0.6366960728259, + "learning_rate": 6.160458180000079e-06, + "loss": 0.4972, + "step": 20111 + }, + { + "epoch": 0.4427030596522122, + "grad_norm": 0.6358968246024566, + "learning_rate": 6.1601114540965465e-06, + "loss": 0.4892, + "step": 20112 + }, + { + "epoch": 0.4427250715386309, + "grad_norm": 1.4431697068554914, + "learning_rate": 6.159764722296809e-06, + "loss": 0.3793, + "step": 20113 + }, + { + "epoch": 0.44274708342504954, + "grad_norm": 1.2787197626122546, + "learning_rate": 6.1594179846026325e-06, + "loss": 0.2662, + "step": 20114 + }, + { + "epoch": 0.4427690953114682, + "grad_norm": 1.2199643525968713, + "learning_rate": 6.159071241015776e-06, + "loss": 0.2949, + "step": 20115 + }, + { + "epoch": 0.44279110719788684, + "grad_norm": 1.4133997058504124, + "learning_rate": 6.1587244915380015e-06, + "loss": 0.3269, + "step": 20116 + }, + { + "epoch": 0.44281311908430554, + "grad_norm": 1.526040607281399, + "learning_rate": 6.158377736171074e-06, + "loss": 0.3479, + "step": 20117 + }, + { + "epoch": 0.4428351309707242, + "grad_norm": 1.5539612524513924, + "learning_rate": 6.158030974916752e-06, + "loss": 0.3299, + "step": 20118 + }, + { + "epoch": 0.44285714285714284, + "grad_norm": 1.3174795414328533, + "learning_rate": 6.1576842077768016e-06, + "loss": 0.2855, + "step": 20119 + }, + { + "epoch": 0.44287915474356154, + "grad_norm": 1.4788121276284427, + "learning_rate": 6.157337434752985e-06, + "loss": 0.2519, + "step": 20120 + }, + { + "epoch": 0.4429011666299802, + "grad_norm": 1.3857501316613654, + "learning_rate": 6.156990655847062e-06, + "loss": 0.3467, + "step": 20121 + }, + { + "epoch": 0.44292317851639884, + "grad_norm": 1.4841152250656502, + "learning_rate": 6.156643871060796e-06, + "loss": 0.3661, + "step": 20122 + }, + { + "epoch": 0.44294519040281755, + "grad_norm": 1.3976417278517448, + "learning_rate": 6.15629708039595e-06, + "loss": 0.3347, + "step": 20123 + }, + { + "epoch": 0.4429672022892362, + "grad_norm": 1.399000940083721, + "learning_rate": 6.155950283854287e-06, + "loss": 0.3343, + "step": 20124 + }, + { + "epoch": 0.44298921417565484, + "grad_norm": 1.2924124216774717, + "learning_rate": 6.155603481437569e-06, + "loss": 0.3112, + "step": 20125 + }, + { + "epoch": 0.4430112260620735, + "grad_norm": 1.780373754281966, + "learning_rate": 6.155256673147558e-06, + "loss": 0.2814, + "step": 20126 + }, + { + "epoch": 0.4430332379484922, + "grad_norm": 1.4306046758134956, + "learning_rate": 6.154909858986019e-06, + "loss": 0.3204, + "step": 20127 + }, + { + "epoch": 0.44305524983491085, + "grad_norm": 1.2739250853156026, + "learning_rate": 6.154563038954711e-06, + "loss": 0.3334, + "step": 20128 + }, + { + "epoch": 0.4430772617213295, + "grad_norm": 24.83513576633779, + "learning_rate": 6.154216213055401e-06, + "loss": 0.2931, + "step": 20129 + }, + { + "epoch": 0.4430992736077482, + "grad_norm": 2.2039118354569416, + "learning_rate": 6.153869381289847e-06, + "loss": 0.2917, + "step": 20130 + }, + { + "epoch": 0.44312128549416685, + "grad_norm": 1.0560541234939973, + "learning_rate": 6.1535225436598156e-06, + "loss": 0.3554, + "step": 20131 + }, + { + "epoch": 0.4431432973805855, + "grad_norm": 1.4491038479448206, + "learning_rate": 6.153175700167068e-06, + "loss": 0.3237, + "step": 20132 + }, + { + "epoch": 0.4431653092670042, + "grad_norm": 1.693263603202029, + "learning_rate": 6.152828850813366e-06, + "loss": 0.3236, + "step": 20133 + }, + { + "epoch": 0.44318732115342285, + "grad_norm": 1.8352704720687538, + "learning_rate": 6.152481995600475e-06, + "loss": 0.3552, + "step": 20134 + }, + { + "epoch": 0.4432093330398415, + "grad_norm": 2.2268101939582574, + "learning_rate": 6.152135134530156e-06, + "loss": 0.2311, + "step": 20135 + }, + { + "epoch": 0.4432313449262602, + "grad_norm": 1.3312593960235521, + "learning_rate": 6.151788267604172e-06, + "loss": 0.2872, + "step": 20136 + }, + { + "epoch": 0.44325335681267886, + "grad_norm": 1.2754039954218428, + "learning_rate": 6.151441394824287e-06, + "loss": 0.3155, + "step": 20137 + }, + { + "epoch": 0.4432753686990975, + "grad_norm": 0.7322906089671136, + "learning_rate": 6.151094516192263e-06, + "loss": 0.4803, + "step": 20138 + }, + { + "epoch": 0.44329738058551615, + "grad_norm": 1.5386743811110377, + "learning_rate": 6.1507476317098635e-06, + "loss": 0.3074, + "step": 20139 + }, + { + "epoch": 0.44331939247193486, + "grad_norm": 1.2430240534703156, + "learning_rate": 6.150400741378851e-06, + "loss": 0.3428, + "step": 20140 + }, + { + "epoch": 0.4433414043583535, + "grad_norm": 2.3973886481620648, + "learning_rate": 6.150053845200988e-06, + "loss": 0.3345, + "step": 20141 + }, + { + "epoch": 0.44336341624477216, + "grad_norm": 1.8600493442203558, + "learning_rate": 6.14970694317804e-06, + "loss": 0.3425, + "step": 20142 + }, + { + "epoch": 0.44338542813119086, + "grad_norm": 1.6883473972451362, + "learning_rate": 6.149360035311767e-06, + "loss": 0.3435, + "step": 20143 + }, + { + "epoch": 0.4434074400176095, + "grad_norm": 1.8226694796835679, + "learning_rate": 6.149013121603935e-06, + "loss": 0.3479, + "step": 20144 + }, + { + "epoch": 0.44342945190402816, + "grad_norm": 1.926167953446319, + "learning_rate": 6.148666202056306e-06, + "loss": 0.382, + "step": 20145 + }, + { + "epoch": 0.44345146379044686, + "grad_norm": 1.4361463609702783, + "learning_rate": 6.148319276670641e-06, + "loss": 0.3446, + "step": 20146 + }, + { + "epoch": 0.4434734756768655, + "grad_norm": 1.7535388214274061, + "learning_rate": 6.147972345448708e-06, + "loss": 0.3688, + "step": 20147 + }, + { + "epoch": 0.44349548756328416, + "grad_norm": 1.2706232705053073, + "learning_rate": 6.147625408392266e-06, + "loss": 0.2976, + "step": 20148 + }, + { + "epoch": 0.4435174994497028, + "grad_norm": 1.0843457415704667, + "learning_rate": 6.14727846550308e-06, + "loss": 0.304, + "step": 20149 + }, + { + "epoch": 0.4435395113361215, + "grad_norm": 1.390854735984333, + "learning_rate": 6.1469315167829125e-06, + "loss": 0.4185, + "step": 20150 + }, + { + "epoch": 0.44356152322254017, + "grad_norm": 1.7621694129348828, + "learning_rate": 6.146584562233528e-06, + "loss": 0.2606, + "step": 20151 + }, + { + "epoch": 0.4435835351089588, + "grad_norm": 0.6589363520647813, + "learning_rate": 6.146237601856691e-06, + "loss": 0.473, + "step": 20152 + }, + { + "epoch": 0.4436055469953775, + "grad_norm": 1.0195258303335046, + "learning_rate": 6.145890635654162e-06, + "loss": 0.248, + "step": 20153 + }, + { + "epoch": 0.44362755888179617, + "grad_norm": 1.3943344042263741, + "learning_rate": 6.1455436636277035e-06, + "loss": 0.3639, + "step": 20154 + }, + { + "epoch": 0.4436495707682148, + "grad_norm": 1.6464691845658785, + "learning_rate": 6.145196685779084e-06, + "loss": 0.3762, + "step": 20155 + }, + { + "epoch": 0.4436715826546335, + "grad_norm": 0.6402019450819755, + "learning_rate": 6.144849702110063e-06, + "loss": 0.4698, + "step": 20156 + }, + { + "epoch": 0.44369359454105217, + "grad_norm": 0.6120956432876775, + "learning_rate": 6.144502712622406e-06, + "loss": 0.4734, + "step": 20157 + }, + { + "epoch": 0.4437156064274708, + "grad_norm": 1.6154369321727062, + "learning_rate": 6.144155717317874e-06, + "loss": 0.3749, + "step": 20158 + }, + { + "epoch": 0.4437376183138895, + "grad_norm": 1.5261612721579245, + "learning_rate": 6.143808716198234e-06, + "loss": 0.2998, + "step": 20159 + }, + { + "epoch": 0.4437596302003082, + "grad_norm": 0.6198033841629448, + "learning_rate": 6.143461709265247e-06, + "loss": 0.4865, + "step": 20160 + }, + { + "epoch": 0.4437816420867268, + "grad_norm": 1.558299106432202, + "learning_rate": 6.143114696520679e-06, + "loss": 0.2926, + "step": 20161 + }, + { + "epoch": 0.4438036539731455, + "grad_norm": 1.354180167333091, + "learning_rate": 6.142767677966291e-06, + "loss": 0.3575, + "step": 20162 + }, + { + "epoch": 0.4438256658595642, + "grad_norm": 1.520573058885616, + "learning_rate": 6.142420653603849e-06, + "loss": 0.2685, + "step": 20163 + }, + { + "epoch": 0.4438476777459828, + "grad_norm": 1.5846834090160329, + "learning_rate": 6.142073623435114e-06, + "loss": 0.3257, + "step": 20164 + }, + { + "epoch": 0.4438696896324015, + "grad_norm": 1.56642745062107, + "learning_rate": 6.141726587461854e-06, + "loss": 0.2303, + "step": 20165 + }, + { + "epoch": 0.4438917015188202, + "grad_norm": 2.1824417460392547, + "learning_rate": 6.141379545685829e-06, + "loss": 0.3277, + "step": 20166 + }, + { + "epoch": 0.44391371340523883, + "grad_norm": 3.138631296238824, + "learning_rate": 6.141032498108803e-06, + "loss": 0.3175, + "step": 20167 + }, + { + "epoch": 0.4439357252916575, + "grad_norm": 1.9245649175576836, + "learning_rate": 6.140685444732542e-06, + "loss": 0.3399, + "step": 20168 + }, + { + "epoch": 0.4439577371780762, + "grad_norm": 1.3410985229713872, + "learning_rate": 6.140338385558808e-06, + "loss": 0.3534, + "step": 20169 + }, + { + "epoch": 0.44397974906449483, + "grad_norm": 1.5329908900035303, + "learning_rate": 6.139991320589367e-06, + "loss": 0.334, + "step": 20170 + }, + { + "epoch": 0.4440017609509135, + "grad_norm": 1.3944609385986373, + "learning_rate": 6.139644249825981e-06, + "loss": 0.3373, + "step": 20171 + }, + { + "epoch": 0.4440237728373322, + "grad_norm": 0.7245279704082973, + "learning_rate": 6.139297173270414e-06, + "loss": 0.4869, + "step": 20172 + }, + { + "epoch": 0.44404578472375084, + "grad_norm": 1.779201705807992, + "learning_rate": 6.1389500909244314e-06, + "loss": 0.2773, + "step": 20173 + }, + { + "epoch": 0.4440677966101695, + "grad_norm": 1.2847070782824537, + "learning_rate": 6.138603002789797e-06, + "loss": 0.2922, + "step": 20174 + }, + { + "epoch": 0.44408980849658813, + "grad_norm": 1.4668638608008902, + "learning_rate": 6.138255908868274e-06, + "loss": 0.3578, + "step": 20175 + }, + { + "epoch": 0.44411182038300684, + "grad_norm": 1.4932478257154465, + "learning_rate": 6.137908809161626e-06, + "loss": 0.318, + "step": 20176 + }, + { + "epoch": 0.4441338322694255, + "grad_norm": 1.1458688123335568, + "learning_rate": 6.1375617036716175e-06, + "loss": 0.2843, + "step": 20177 + }, + { + "epoch": 0.44415584415584414, + "grad_norm": 1.9963068471057377, + "learning_rate": 6.137214592400014e-06, + "loss": 0.2892, + "step": 20178 + }, + { + "epoch": 0.44417785604226284, + "grad_norm": 1.5934196353609484, + "learning_rate": 6.136867475348579e-06, + "loss": 0.3033, + "step": 20179 + }, + { + "epoch": 0.4441998679286815, + "grad_norm": 1.5874419275834997, + "learning_rate": 6.136520352519077e-06, + "loss": 0.2696, + "step": 20180 + }, + { + "epoch": 0.44422187981510014, + "grad_norm": 1.7899085612762964, + "learning_rate": 6.13617322391327e-06, + "loss": 0.357, + "step": 20181 + }, + { + "epoch": 0.44424389170151884, + "grad_norm": 1.6451596132049198, + "learning_rate": 6.135826089532923e-06, + "loss": 0.3121, + "step": 20182 + }, + { + "epoch": 0.4442659035879375, + "grad_norm": 1.213432246752528, + "learning_rate": 6.135478949379805e-06, + "loss": 0.308, + "step": 20183 + }, + { + "epoch": 0.44428791547435614, + "grad_norm": 0.6515667999847322, + "learning_rate": 6.135131803455674e-06, + "loss": 0.4976, + "step": 20184 + }, + { + "epoch": 0.4443099273607748, + "grad_norm": 1.6327698038313028, + "learning_rate": 6.134784651762297e-06, + "loss": 0.3992, + "step": 20185 + }, + { + "epoch": 0.4443319392471935, + "grad_norm": 1.3781822679940985, + "learning_rate": 6.134437494301439e-06, + "loss": 0.3029, + "step": 20186 + }, + { + "epoch": 0.44435395113361215, + "grad_norm": 1.7057841372509444, + "learning_rate": 6.134090331074863e-06, + "loss": 0.3526, + "step": 20187 + }, + { + "epoch": 0.4443759630200308, + "grad_norm": 1.3810264173793079, + "learning_rate": 6.133743162084335e-06, + "loss": 0.3489, + "step": 20188 + }, + { + "epoch": 0.4443979749064495, + "grad_norm": 0.6094139762559277, + "learning_rate": 6.133395987331617e-06, + "loss": 0.4849, + "step": 20189 + }, + { + "epoch": 0.44441998679286815, + "grad_norm": 1.3185706277458138, + "learning_rate": 6.133048806818476e-06, + "loss": 0.2244, + "step": 20190 + }, + { + "epoch": 0.4444419986792868, + "grad_norm": 0.6132569799245657, + "learning_rate": 6.132701620546676e-06, + "loss": 0.4878, + "step": 20191 + }, + { + "epoch": 0.4444640105657055, + "grad_norm": 1.2285410383017445, + "learning_rate": 6.13235442851798e-06, + "loss": 0.2346, + "step": 20192 + }, + { + "epoch": 0.44448602245212415, + "grad_norm": 1.4171418346206885, + "learning_rate": 6.132007230734155e-06, + "loss": 0.2794, + "step": 20193 + }, + { + "epoch": 0.4445080343385428, + "grad_norm": 1.3912638126030887, + "learning_rate": 6.131660027196964e-06, + "loss": 0.3627, + "step": 20194 + }, + { + "epoch": 0.4445300462249615, + "grad_norm": 1.529885015137091, + "learning_rate": 6.131312817908171e-06, + "loss": 0.3268, + "step": 20195 + }, + { + "epoch": 0.44455205811138016, + "grad_norm": 1.466370106932077, + "learning_rate": 6.130965602869543e-06, + "loss": 0.3403, + "step": 20196 + }, + { + "epoch": 0.4445740699977988, + "grad_norm": 1.4910330653388897, + "learning_rate": 6.130618382082843e-06, + "loss": 0.3377, + "step": 20197 + }, + { + "epoch": 0.44459608188421745, + "grad_norm": 1.8690765150289335, + "learning_rate": 6.130271155549836e-06, + "loss": 0.3218, + "step": 20198 + }, + { + "epoch": 0.44461809377063616, + "grad_norm": 1.4980039896536497, + "learning_rate": 6.129923923272287e-06, + "loss": 0.3144, + "step": 20199 + }, + { + "epoch": 0.4446401056570548, + "grad_norm": 1.4290246813352088, + "learning_rate": 6.1295766852519616e-06, + "loss": 0.3135, + "step": 20200 + }, + { + "epoch": 0.44466211754347346, + "grad_norm": 1.534477119484208, + "learning_rate": 6.129229441490622e-06, + "loss": 0.3124, + "step": 20201 + }, + { + "epoch": 0.44468412942989216, + "grad_norm": 2.5723169391231533, + "learning_rate": 6.128882191990037e-06, + "loss": 0.306, + "step": 20202 + }, + { + "epoch": 0.4447061413163108, + "grad_norm": 0.6814833693285407, + "learning_rate": 6.128534936751967e-06, + "loss": 0.4932, + "step": 20203 + }, + { + "epoch": 0.44472815320272946, + "grad_norm": 1.4235166457116775, + "learning_rate": 6.128187675778179e-06, + "loss": 0.3303, + "step": 20204 + }, + { + "epoch": 0.44475016508914816, + "grad_norm": 0.6455120189601506, + "learning_rate": 6.127840409070439e-06, + "loss": 0.4956, + "step": 20205 + }, + { + "epoch": 0.4447721769755668, + "grad_norm": 1.22375620247849, + "learning_rate": 6.127493136630511e-06, + "loss": 0.2414, + "step": 20206 + }, + { + "epoch": 0.44479418886198546, + "grad_norm": 0.6118033455698689, + "learning_rate": 6.12714585846016e-06, + "loss": 0.504, + "step": 20207 + }, + { + "epoch": 0.4448162007484041, + "grad_norm": 2.9334778228475047, + "learning_rate": 6.1267985745611505e-06, + "loss": 0.2562, + "step": 20208 + }, + { + "epoch": 0.4448382126348228, + "grad_norm": 0.6035784238658626, + "learning_rate": 6.1264512849352485e-06, + "loss": 0.5035, + "step": 20209 + }, + { + "epoch": 0.44486022452124147, + "grad_norm": 1.268085143811631, + "learning_rate": 6.126103989584219e-06, + "loss": 0.2698, + "step": 20210 + }, + { + "epoch": 0.4448822364076601, + "grad_norm": 1.5765079086571065, + "learning_rate": 6.125756688509826e-06, + "loss": 0.3352, + "step": 20211 + }, + { + "epoch": 0.4449042482940788, + "grad_norm": 1.3637684702695416, + "learning_rate": 6.125409381713836e-06, + "loss": 0.3357, + "step": 20212 + }, + { + "epoch": 0.44492626018049747, + "grad_norm": 1.5763340271746848, + "learning_rate": 6.125062069198012e-06, + "loss": 0.3549, + "step": 20213 + }, + { + "epoch": 0.4449482720669161, + "grad_norm": 1.335403220479436, + "learning_rate": 6.124714750964123e-06, + "loss": 0.2857, + "step": 20214 + }, + { + "epoch": 0.4449702839533348, + "grad_norm": 0.6660234298684938, + "learning_rate": 6.124367427013931e-06, + "loss": 0.4922, + "step": 20215 + }, + { + "epoch": 0.44499229583975347, + "grad_norm": 1.9822423924086756, + "learning_rate": 6.1240200973492026e-06, + "loss": 0.3305, + "step": 20216 + }, + { + "epoch": 0.4450143077261721, + "grad_norm": 1.1790155759975773, + "learning_rate": 6.123672761971703e-06, + "loss": 0.248, + "step": 20217 + }, + { + "epoch": 0.4450363196125908, + "grad_norm": 1.5498790037988026, + "learning_rate": 6.123325420883195e-06, + "loss": 0.4021, + "step": 20218 + }, + { + "epoch": 0.4450583314990095, + "grad_norm": 1.3160623449096907, + "learning_rate": 6.122978074085448e-06, + "loss": 0.3915, + "step": 20219 + }, + { + "epoch": 0.4450803433854281, + "grad_norm": 1.2007880099228452, + "learning_rate": 6.122630721580227e-06, + "loss": 0.2502, + "step": 20220 + }, + { + "epoch": 0.4451023552718468, + "grad_norm": 1.634479079279915, + "learning_rate": 6.122283363369293e-06, + "loss": 0.2354, + "step": 20221 + }, + { + "epoch": 0.4451243671582655, + "grad_norm": 1.318023656537462, + "learning_rate": 6.121935999454417e-06, + "loss": 0.3376, + "step": 20222 + }, + { + "epoch": 0.4451463790446841, + "grad_norm": 0.6080629629629509, + "learning_rate": 6.121588629837361e-06, + "loss": 0.4858, + "step": 20223 + }, + { + "epoch": 0.4451683909311028, + "grad_norm": 1.447180374218204, + "learning_rate": 6.121241254519891e-06, + "loss": 0.2911, + "step": 20224 + }, + { + "epoch": 0.4451904028175215, + "grad_norm": 1.765261883289921, + "learning_rate": 6.120893873503774e-06, + "loss": 0.2792, + "step": 20225 + }, + { + "epoch": 0.44521241470394013, + "grad_norm": 2.3848112255763154, + "learning_rate": 6.120546486790772e-06, + "loss": 0.3737, + "step": 20226 + }, + { + "epoch": 0.4452344265903588, + "grad_norm": 3.7752940589471886, + "learning_rate": 6.120199094382654e-06, + "loss": 0.4153, + "step": 20227 + }, + { + "epoch": 0.4452564384767775, + "grad_norm": 1.6830522455579304, + "learning_rate": 6.119851696281185e-06, + "loss": 0.3668, + "step": 20228 + }, + { + "epoch": 0.44527845036319613, + "grad_norm": 1.659096698979193, + "learning_rate": 6.119504292488131e-06, + "loss": 0.3523, + "step": 20229 + }, + { + "epoch": 0.4453004622496148, + "grad_norm": 1.5836880779462867, + "learning_rate": 6.119156883005255e-06, + "loss": 0.3407, + "step": 20230 + }, + { + "epoch": 0.44532247413603343, + "grad_norm": 1.4635825202644765, + "learning_rate": 6.118809467834326e-06, + "loss": 0.288, + "step": 20231 + }, + { + "epoch": 0.44534448602245214, + "grad_norm": 1.3806141399025795, + "learning_rate": 6.118462046977107e-06, + "loss": 0.3183, + "step": 20232 + }, + { + "epoch": 0.4453664979088708, + "grad_norm": 1.9208908794067245, + "learning_rate": 6.118114620435366e-06, + "loss": 0.3104, + "step": 20233 + }, + { + "epoch": 0.44538850979528943, + "grad_norm": 1.7021225564366769, + "learning_rate": 6.117767188210868e-06, + "loss": 0.4418, + "step": 20234 + }, + { + "epoch": 0.44541052168170814, + "grad_norm": 1.5936569513113163, + "learning_rate": 6.117419750305379e-06, + "loss": 0.2209, + "step": 20235 + }, + { + "epoch": 0.4454325335681268, + "grad_norm": 1.8601420365375618, + "learning_rate": 6.117072306720663e-06, + "loss": 0.3336, + "step": 20236 + }, + { + "epoch": 0.44545454545454544, + "grad_norm": 3.069908376833328, + "learning_rate": 6.116724857458487e-06, + "loss": 0.3966, + "step": 20237 + }, + { + "epoch": 0.44547655734096414, + "grad_norm": 1.726342535219393, + "learning_rate": 6.11637740252062e-06, + "loss": 0.3644, + "step": 20238 + }, + { + "epoch": 0.4454985692273828, + "grad_norm": 1.7984767183799126, + "learning_rate": 6.116029941908822e-06, + "loss": 0.2788, + "step": 20239 + }, + { + "epoch": 0.44552058111380144, + "grad_norm": 1.4304274789224558, + "learning_rate": 6.1156824756248645e-06, + "loss": 0.239, + "step": 20240 + }, + { + "epoch": 0.44554259300022014, + "grad_norm": 1.7537779305130048, + "learning_rate": 6.115335003670509e-06, + "loss": 0.3026, + "step": 20241 + }, + { + "epoch": 0.4455646048866388, + "grad_norm": 1.8142114144599988, + "learning_rate": 6.114987526047524e-06, + "loss": 0.4668, + "step": 20242 + }, + { + "epoch": 0.44558661677305744, + "grad_norm": 1.4535201242654137, + "learning_rate": 6.114640042757677e-06, + "loss": 0.2793, + "step": 20243 + }, + { + "epoch": 0.4456086286594761, + "grad_norm": 0.6410844011688568, + "learning_rate": 6.11429255380273e-06, + "loss": 0.4888, + "step": 20244 + }, + { + "epoch": 0.4456306405458948, + "grad_norm": 1.0092094627495551, + "learning_rate": 6.113945059184452e-06, + "loss": 0.2925, + "step": 20245 + }, + { + "epoch": 0.44565265243231345, + "grad_norm": 2.21760732977654, + "learning_rate": 6.113597558904609e-06, + "loss": 0.3768, + "step": 20246 + }, + { + "epoch": 0.4456746643187321, + "grad_norm": 2.012098790027866, + "learning_rate": 6.113250052964966e-06, + "loss": 0.3729, + "step": 20247 + }, + { + "epoch": 0.4456966762051508, + "grad_norm": 1.920808382267861, + "learning_rate": 6.11290254136729e-06, + "loss": 0.2465, + "step": 20248 + }, + { + "epoch": 0.44571868809156945, + "grad_norm": 1.5549310558311684, + "learning_rate": 6.112555024113345e-06, + "loss": 0.3251, + "step": 20249 + }, + { + "epoch": 0.4457406999779881, + "grad_norm": 1.443461163943899, + "learning_rate": 6.112207501204901e-06, + "loss": 0.3157, + "step": 20250 + }, + { + "epoch": 0.4457627118644068, + "grad_norm": 0.6015100032361169, + "learning_rate": 6.111859972643722e-06, + "loss": 0.4716, + "step": 20251 + }, + { + "epoch": 0.44578472375082545, + "grad_norm": 1.2136810347947529, + "learning_rate": 6.111512438431575e-06, + "loss": 0.3366, + "step": 20252 + }, + { + "epoch": 0.4458067356372441, + "grad_norm": 1.5058476336554145, + "learning_rate": 6.111164898570226e-06, + "loss": 0.3384, + "step": 20253 + }, + { + "epoch": 0.4458287475236628, + "grad_norm": 1.706332755669334, + "learning_rate": 6.110817353061441e-06, + "loss": 0.2851, + "step": 20254 + }, + { + "epoch": 0.44585075941008145, + "grad_norm": 1.4793340370484855, + "learning_rate": 6.110469801906988e-06, + "loss": 0.2694, + "step": 20255 + }, + { + "epoch": 0.4458727712965001, + "grad_norm": 1.5948167376781475, + "learning_rate": 6.110122245108631e-06, + "loss": 0.3422, + "step": 20256 + }, + { + "epoch": 0.44589478318291875, + "grad_norm": 1.3895945188943108, + "learning_rate": 6.109774682668137e-06, + "loss": 0.3765, + "step": 20257 + }, + { + "epoch": 0.44591679506933746, + "grad_norm": 1.8973893858337563, + "learning_rate": 6.109427114587274e-06, + "loss": 0.3747, + "step": 20258 + }, + { + "epoch": 0.4459388069557561, + "grad_norm": 1.79502676387892, + "learning_rate": 6.1090795408678085e-06, + "loss": 0.4186, + "step": 20259 + }, + { + "epoch": 0.44596081884217476, + "grad_norm": 0.6015517094269512, + "learning_rate": 6.1087319615115036e-06, + "loss": 0.4835, + "step": 20260 + }, + { + "epoch": 0.44598283072859346, + "grad_norm": 1.9785830966133575, + "learning_rate": 6.1083843765201315e-06, + "loss": 0.3427, + "step": 20261 + }, + { + "epoch": 0.4460048426150121, + "grad_norm": 1.8052863934340608, + "learning_rate": 6.108036785895453e-06, + "loss": 0.3264, + "step": 20262 + }, + { + "epoch": 0.44602685450143076, + "grad_norm": 0.6231034969353113, + "learning_rate": 6.1076891896392385e-06, + "loss": 0.4918, + "step": 20263 + }, + { + "epoch": 0.44604886638784946, + "grad_norm": 1.3971392176297712, + "learning_rate": 6.107341587753254e-06, + "loss": 0.3185, + "step": 20264 + }, + { + "epoch": 0.4460708782742681, + "grad_norm": 1.3006085647494088, + "learning_rate": 6.1069939802392655e-06, + "loss": 0.2649, + "step": 20265 + }, + { + "epoch": 0.44609289016068676, + "grad_norm": 1.5657954330477004, + "learning_rate": 6.10664636709904e-06, + "loss": 0.2295, + "step": 20266 + }, + { + "epoch": 0.4461149020471054, + "grad_norm": 2.1360367736234225, + "learning_rate": 6.106298748334341e-06, + "loss": 0.3395, + "step": 20267 + }, + { + "epoch": 0.4461369139335241, + "grad_norm": 1.3570520407656024, + "learning_rate": 6.105951123946942e-06, + "loss": 0.3608, + "step": 20268 + }, + { + "epoch": 0.44615892581994276, + "grad_norm": 1.8790750826028975, + "learning_rate": 6.105603493938605e-06, + "loss": 0.3977, + "step": 20269 + }, + { + "epoch": 0.4461809377063614, + "grad_norm": 1.5676523661000656, + "learning_rate": 6.105255858311098e-06, + "loss": 0.3134, + "step": 20270 + }, + { + "epoch": 0.4462029495927801, + "grad_norm": 1.3069505795119165, + "learning_rate": 6.104908217066187e-06, + "loss": 0.3344, + "step": 20271 + }, + { + "epoch": 0.44622496147919877, + "grad_norm": 0.6217200098620076, + "learning_rate": 6.10456057020564e-06, + "loss": 0.457, + "step": 20272 + }, + { + "epoch": 0.4462469733656174, + "grad_norm": 1.2256715184927567, + "learning_rate": 6.104212917731225e-06, + "loss": 0.3575, + "step": 20273 + }, + { + "epoch": 0.4462689852520361, + "grad_norm": 2.102627452662569, + "learning_rate": 6.103865259644707e-06, + "loss": 0.2862, + "step": 20274 + }, + { + "epoch": 0.44629099713845477, + "grad_norm": 1.6462521089491506, + "learning_rate": 6.103517595947852e-06, + "loss": 0.3132, + "step": 20275 + }, + { + "epoch": 0.4463130090248734, + "grad_norm": 1.906228034521299, + "learning_rate": 6.103169926642429e-06, + "loss": 0.3374, + "step": 20276 + }, + { + "epoch": 0.4463350209112921, + "grad_norm": 1.9005145644749983, + "learning_rate": 6.102822251730205e-06, + "loss": 0.3131, + "step": 20277 + }, + { + "epoch": 0.4463570327977108, + "grad_norm": 1.5320909489194021, + "learning_rate": 6.102474571212944e-06, + "loss": 0.3742, + "step": 20278 + }, + { + "epoch": 0.4463790446841294, + "grad_norm": 0.625814904571274, + "learning_rate": 6.10212688509242e-06, + "loss": 0.494, + "step": 20279 + }, + { + "epoch": 0.44640105657054807, + "grad_norm": 0.6070037931635287, + "learning_rate": 6.101779193370392e-06, + "loss": 0.4807, + "step": 20280 + }, + { + "epoch": 0.4464230684569668, + "grad_norm": 1.266672565430352, + "learning_rate": 6.101431496048631e-06, + "loss": 0.328, + "step": 20281 + }, + { + "epoch": 0.4464450803433854, + "grad_norm": 1.360727135368357, + "learning_rate": 6.101083793128906e-06, + "loss": 0.2702, + "step": 20282 + }, + { + "epoch": 0.4464670922298041, + "grad_norm": 2.380837795518149, + "learning_rate": 6.100736084612981e-06, + "loss": 0.3211, + "step": 20283 + }, + { + "epoch": 0.4464891041162228, + "grad_norm": 1.2540054522869606, + "learning_rate": 6.1003883705026234e-06, + "loss": 0.3708, + "step": 20284 + }, + { + "epoch": 0.44651111600264143, + "grad_norm": 2.049737756892733, + "learning_rate": 6.1000406507996015e-06, + "loss": 0.3333, + "step": 20285 + }, + { + "epoch": 0.4465331278890601, + "grad_norm": 1.851623254138009, + "learning_rate": 6.099692925505683e-06, + "loss": 0.2386, + "step": 20286 + }, + { + "epoch": 0.4465551397754788, + "grad_norm": 1.8662258568680443, + "learning_rate": 6.099345194622635e-06, + "loss": 0.3763, + "step": 20287 + }, + { + "epoch": 0.44657715166189743, + "grad_norm": 1.5624462893623063, + "learning_rate": 6.098997458152222e-06, + "loss": 0.2695, + "step": 20288 + }, + { + "epoch": 0.4465991635483161, + "grad_norm": 0.6263916164950637, + "learning_rate": 6.098649716096217e-06, + "loss": 0.4858, + "step": 20289 + }, + { + "epoch": 0.44662117543473473, + "grad_norm": 1.412014791240765, + "learning_rate": 6.098301968456381e-06, + "loss": 0.3509, + "step": 20290 + }, + { + "epoch": 0.44664318732115343, + "grad_norm": 0.6867514291679957, + "learning_rate": 6.097954215234487e-06, + "loss": 0.5143, + "step": 20291 + }, + { + "epoch": 0.4466651992075721, + "grad_norm": 1.4868654173293743, + "learning_rate": 6.097606456432299e-06, + "loss": 0.3192, + "step": 20292 + }, + { + "epoch": 0.44668721109399073, + "grad_norm": 1.4892032336297685, + "learning_rate": 6.0972586920515866e-06, + "loss": 0.2993, + "step": 20293 + }, + { + "epoch": 0.44670922298040944, + "grad_norm": 1.594721390971017, + "learning_rate": 6.096910922094115e-06, + "loss": 0.3259, + "step": 20294 + }, + { + "epoch": 0.4467312348668281, + "grad_norm": 1.9682222904394486, + "learning_rate": 6.096563146561653e-06, + "loss": 0.3066, + "step": 20295 + }, + { + "epoch": 0.44675324675324674, + "grad_norm": 2.9080012231673265, + "learning_rate": 6.096215365455967e-06, + "loss": 0.292, + "step": 20296 + }, + { + "epoch": 0.44677525863966544, + "grad_norm": 2.2223492042150492, + "learning_rate": 6.095867578778828e-06, + "loss": 0.2798, + "step": 20297 + }, + { + "epoch": 0.4467972705260841, + "grad_norm": 1.3487870609150885, + "learning_rate": 6.095519786531999e-06, + "loss": 0.311, + "step": 20298 + }, + { + "epoch": 0.44681928241250274, + "grad_norm": 1.7285134054350668, + "learning_rate": 6.0951719887172515e-06, + "loss": 0.3388, + "step": 20299 + }, + { + "epoch": 0.44684129429892144, + "grad_norm": 1.1269089342792904, + "learning_rate": 6.09482418533635e-06, + "loss": 0.2727, + "step": 20300 + }, + { + "epoch": 0.4468633061853401, + "grad_norm": 1.7084694524944117, + "learning_rate": 6.094476376391064e-06, + "loss": 0.364, + "step": 20301 + }, + { + "epoch": 0.44688531807175874, + "grad_norm": 2.648275832861396, + "learning_rate": 6.094128561883163e-06, + "loss": 0.3608, + "step": 20302 + }, + { + "epoch": 0.4469073299581774, + "grad_norm": 1.483623561848633, + "learning_rate": 6.093780741814409e-06, + "loss": 0.2908, + "step": 20303 + }, + { + "epoch": 0.4469293418445961, + "grad_norm": 2.0687925414223898, + "learning_rate": 6.0934329161865755e-06, + "loss": 0.2558, + "step": 20304 + }, + { + "epoch": 0.44695135373101474, + "grad_norm": 1.3450663886961296, + "learning_rate": 6.093085085001429e-06, + "loss": 0.2921, + "step": 20305 + }, + { + "epoch": 0.4469733656174334, + "grad_norm": 1.2791877696818146, + "learning_rate": 6.092737248260735e-06, + "loss": 0.2879, + "step": 20306 + }, + { + "epoch": 0.4469953775038521, + "grad_norm": 1.4528687193610412, + "learning_rate": 6.092389405966264e-06, + "loss": 0.3398, + "step": 20307 + }, + { + "epoch": 0.44701738939027075, + "grad_norm": 1.770416813108519, + "learning_rate": 6.092041558119781e-06, + "loss": 0.361, + "step": 20308 + }, + { + "epoch": 0.4470394012766894, + "grad_norm": 1.2167126249667752, + "learning_rate": 6.0916937047230575e-06, + "loss": 0.3817, + "step": 20309 + }, + { + "epoch": 0.4470614131631081, + "grad_norm": 1.626862515491968, + "learning_rate": 6.09134584577786e-06, + "loss": 0.3364, + "step": 20310 + }, + { + "epoch": 0.44708342504952675, + "grad_norm": 1.7624358059363, + "learning_rate": 6.090997981285956e-06, + "loss": 0.3344, + "step": 20311 + }, + { + "epoch": 0.4471054369359454, + "grad_norm": 2.15107576281484, + "learning_rate": 6.090650111249112e-06, + "loss": 0.2624, + "step": 20312 + }, + { + "epoch": 0.4471274488223641, + "grad_norm": 1.1222592385454278, + "learning_rate": 6.0903022356691e-06, + "loss": 0.2787, + "step": 20313 + }, + { + "epoch": 0.44714946070878275, + "grad_norm": 1.20229664779309, + "learning_rate": 6.089954354547683e-06, + "loss": 0.3279, + "step": 20314 + }, + { + "epoch": 0.4471714725952014, + "grad_norm": 1.4569818509139556, + "learning_rate": 6.089606467886635e-06, + "loss": 0.3459, + "step": 20315 + }, + { + "epoch": 0.44719348448162005, + "grad_norm": 1.4304076418995972, + "learning_rate": 6.089258575687718e-06, + "loss": 0.3227, + "step": 20316 + }, + { + "epoch": 0.44721549636803876, + "grad_norm": 1.4944544100102921, + "learning_rate": 6.088910677952705e-06, + "loss": 0.3938, + "step": 20317 + }, + { + "epoch": 0.4472375082544574, + "grad_norm": 1.2095103709524446, + "learning_rate": 6.088562774683362e-06, + "loss": 0.3187, + "step": 20318 + }, + { + "epoch": 0.44725952014087605, + "grad_norm": 1.1734743672334407, + "learning_rate": 6.088214865881457e-06, + "loss": 0.2642, + "step": 20319 + }, + { + "epoch": 0.44728153202729476, + "grad_norm": 1.1836557892716029, + "learning_rate": 6.0878669515487595e-06, + "loss": 0.2517, + "step": 20320 + }, + { + "epoch": 0.4473035439137134, + "grad_norm": 1.5557659589683912, + "learning_rate": 6.087519031687035e-06, + "loss": 0.3642, + "step": 20321 + }, + { + "epoch": 0.44732555580013206, + "grad_norm": 1.4582704460083296, + "learning_rate": 6.087171106298056e-06, + "loss": 0.3625, + "step": 20322 + }, + { + "epoch": 0.44734756768655076, + "grad_norm": 1.58957554184219, + "learning_rate": 6.086823175383587e-06, + "loss": 0.3361, + "step": 20323 + }, + { + "epoch": 0.4473695795729694, + "grad_norm": 1.4683474111802428, + "learning_rate": 6.086475238945398e-06, + "loss": 0.2718, + "step": 20324 + }, + { + "epoch": 0.44739159145938806, + "grad_norm": 1.3294383160477665, + "learning_rate": 6.086127296985259e-06, + "loss": 0.2872, + "step": 20325 + }, + { + "epoch": 0.4474136033458067, + "grad_norm": 1.3506774466400242, + "learning_rate": 6.085779349504935e-06, + "loss": 0.3412, + "step": 20326 + }, + { + "epoch": 0.4474356152322254, + "grad_norm": 1.2194904904871937, + "learning_rate": 6.085431396506196e-06, + "loss": 0.3349, + "step": 20327 + }, + { + "epoch": 0.44745762711864406, + "grad_norm": 1.27637028344898, + "learning_rate": 6.085083437990811e-06, + "loss": 0.2668, + "step": 20328 + }, + { + "epoch": 0.4474796390050627, + "grad_norm": 1.6056083884088188, + "learning_rate": 6.084735473960548e-06, + "loss": 0.3263, + "step": 20329 + }, + { + "epoch": 0.4475016508914814, + "grad_norm": 1.2783490654368748, + "learning_rate": 6.0843875044171766e-06, + "loss": 0.4075, + "step": 20330 + }, + { + "epoch": 0.44752366277790007, + "grad_norm": 0.6539797305762804, + "learning_rate": 6.084039529362462e-06, + "loss": 0.4847, + "step": 20331 + }, + { + "epoch": 0.4475456746643187, + "grad_norm": 1.6260450506330921, + "learning_rate": 6.083691548798176e-06, + "loss": 0.2954, + "step": 20332 + }, + { + "epoch": 0.4475676865507374, + "grad_norm": 1.308499246176222, + "learning_rate": 6.083343562726087e-06, + "loss": 0.2929, + "step": 20333 + }, + { + "epoch": 0.44758969843715607, + "grad_norm": 1.65784296695462, + "learning_rate": 6.082995571147962e-06, + "loss": 0.2593, + "step": 20334 + }, + { + "epoch": 0.4476117103235747, + "grad_norm": 3.2006279634866925, + "learning_rate": 6.082647574065571e-06, + "loss": 0.3336, + "step": 20335 + }, + { + "epoch": 0.4476337222099934, + "grad_norm": 1.3514221622608615, + "learning_rate": 6.082299571480682e-06, + "loss": 0.3556, + "step": 20336 + }, + { + "epoch": 0.4476557340964121, + "grad_norm": 2.146550592115287, + "learning_rate": 6.081951563395062e-06, + "loss": 0.3452, + "step": 20337 + }, + { + "epoch": 0.4476777459828307, + "grad_norm": 1.4396300139750355, + "learning_rate": 6.081603549810486e-06, + "loss": 0.3778, + "step": 20338 + }, + { + "epoch": 0.44769975786924937, + "grad_norm": 0.6121638351073371, + "learning_rate": 6.081255530728713e-06, + "loss": 0.4827, + "step": 20339 + }, + { + "epoch": 0.4477217697556681, + "grad_norm": 1.585617965529424, + "learning_rate": 6.080907506151521e-06, + "loss": 0.3684, + "step": 20340 + }, + { + "epoch": 0.4477437816420867, + "grad_norm": 1.9912705867258202, + "learning_rate": 6.080559476080674e-06, + "loss": 0.3049, + "step": 20341 + }, + { + "epoch": 0.4477657935285054, + "grad_norm": 1.7518552900464357, + "learning_rate": 6.08021144051794e-06, + "loss": 0.2354, + "step": 20342 + }, + { + "epoch": 0.4477878054149241, + "grad_norm": 0.6144790451826422, + "learning_rate": 6.079863399465092e-06, + "loss": 0.4756, + "step": 20343 + }, + { + "epoch": 0.4478098173013427, + "grad_norm": 0.6283508652957244, + "learning_rate": 6.079515352923896e-06, + "loss": 0.48, + "step": 20344 + }, + { + "epoch": 0.4478318291877614, + "grad_norm": 1.1746300346159546, + "learning_rate": 6.079167300896122e-06, + "loss": 0.2683, + "step": 20345 + }, + { + "epoch": 0.4478538410741801, + "grad_norm": 1.2036252426858856, + "learning_rate": 6.078819243383537e-06, + "loss": 0.2943, + "step": 20346 + }, + { + "epoch": 0.44787585296059873, + "grad_norm": 0.615708810945396, + "learning_rate": 6.078471180387913e-06, + "loss": 0.4923, + "step": 20347 + }, + { + "epoch": 0.4478978648470174, + "grad_norm": 3.285686504730427, + "learning_rate": 6.078123111911016e-06, + "loss": 0.2782, + "step": 20348 + }, + { + "epoch": 0.44791987673343603, + "grad_norm": 0.6343444173680576, + "learning_rate": 6.077775037954618e-06, + "loss": 0.4954, + "step": 20349 + }, + { + "epoch": 0.44794188861985473, + "grad_norm": 1.2469500000654685, + "learning_rate": 6.077426958520485e-06, + "loss": 0.2668, + "step": 20350 + }, + { + "epoch": 0.4479639005062734, + "grad_norm": 1.4921700242950593, + "learning_rate": 6.077078873610388e-06, + "loss": 0.2652, + "step": 20351 + }, + { + "epoch": 0.44798591239269203, + "grad_norm": 1.4561031890301068, + "learning_rate": 6.076730783226096e-06, + "loss": 0.3328, + "step": 20352 + }, + { + "epoch": 0.44800792427911074, + "grad_norm": 1.7509152007910906, + "learning_rate": 6.076382687369379e-06, + "loss": 0.3482, + "step": 20353 + }, + { + "epoch": 0.4480299361655294, + "grad_norm": 1.269239675576307, + "learning_rate": 6.076034586042003e-06, + "loss": 0.2991, + "step": 20354 + }, + { + "epoch": 0.44805194805194803, + "grad_norm": 1.231389192848917, + "learning_rate": 6.07568647924574e-06, + "loss": 0.2075, + "step": 20355 + }, + { + "epoch": 0.44807395993836674, + "grad_norm": 0.6439274811408396, + "learning_rate": 6.0753383669823606e-06, + "loss": 0.494, + "step": 20356 + }, + { + "epoch": 0.4480959718247854, + "grad_norm": 1.495910713826393, + "learning_rate": 6.0749902492536296e-06, + "loss": 0.3548, + "step": 20357 + }, + { + "epoch": 0.44811798371120404, + "grad_norm": 0.6381239180747962, + "learning_rate": 6.07464212606132e-06, + "loss": 0.4923, + "step": 20358 + }, + { + "epoch": 0.44813999559762274, + "grad_norm": 1.3447427285675726, + "learning_rate": 6.074293997407199e-06, + "loss": 0.3179, + "step": 20359 + }, + { + "epoch": 0.4481620074840414, + "grad_norm": 2.050171330307794, + "learning_rate": 6.073945863293036e-06, + "loss": 0.3469, + "step": 20360 + }, + { + "epoch": 0.44818401937046004, + "grad_norm": 1.3085468653304468, + "learning_rate": 6.073597723720603e-06, + "loss": 0.326, + "step": 20361 + }, + { + "epoch": 0.4482060312568787, + "grad_norm": 1.9314303336009384, + "learning_rate": 6.0732495786916666e-06, + "loss": 0.2652, + "step": 20362 + }, + { + "epoch": 0.4482280431432974, + "grad_norm": 1.075907157827635, + "learning_rate": 6.072901428207997e-06, + "loss": 0.2765, + "step": 20363 + }, + { + "epoch": 0.44825005502971604, + "grad_norm": 1.5741515288847514, + "learning_rate": 6.072553272271364e-06, + "loss": 0.278, + "step": 20364 + }, + { + "epoch": 0.4482720669161347, + "grad_norm": 9.340560591308426, + "learning_rate": 6.072205110883538e-06, + "loss": 0.2086, + "step": 20365 + }, + { + "epoch": 0.4482940788025534, + "grad_norm": 1.6576763801397176, + "learning_rate": 6.071856944046286e-06, + "loss": 0.2973, + "step": 20366 + }, + { + "epoch": 0.44831609068897205, + "grad_norm": 1.7389750972639797, + "learning_rate": 6.071508771761379e-06, + "loss": 0.315, + "step": 20367 + }, + { + "epoch": 0.4483381025753907, + "grad_norm": 1.666486430071725, + "learning_rate": 6.071160594030586e-06, + "loss": 0.2772, + "step": 20368 + }, + { + "epoch": 0.4483601144618094, + "grad_norm": 1.1817617621011485, + "learning_rate": 6.070812410855679e-06, + "loss": 0.1924, + "step": 20369 + }, + { + "epoch": 0.44838212634822805, + "grad_norm": 0.6903609956133652, + "learning_rate": 6.070464222238424e-06, + "loss": 0.4785, + "step": 20370 + }, + { + "epoch": 0.4484041382346467, + "grad_norm": 1.7672746382228097, + "learning_rate": 6.070116028180594e-06, + "loss": 0.3428, + "step": 20371 + }, + { + "epoch": 0.44842615012106535, + "grad_norm": 2.0248029587646634, + "learning_rate": 6.069767828683956e-06, + "loss": 0.2763, + "step": 20372 + }, + { + "epoch": 0.44844816200748405, + "grad_norm": 0.6203031781151263, + "learning_rate": 6.06941962375028e-06, + "loss": 0.4858, + "step": 20373 + }, + { + "epoch": 0.4484701738939027, + "grad_norm": 2.1808179026805092, + "learning_rate": 6.069071413381339e-06, + "loss": 0.4093, + "step": 20374 + }, + { + "epoch": 0.44849218578032135, + "grad_norm": 1.3818121170233364, + "learning_rate": 6.068723197578897e-06, + "loss": 0.361, + "step": 20375 + }, + { + "epoch": 0.44851419766674006, + "grad_norm": 1.400188490725645, + "learning_rate": 6.068374976344729e-06, + "loss": 0.4063, + "step": 20376 + }, + { + "epoch": 0.4485362095531587, + "grad_norm": 1.4927083236587713, + "learning_rate": 6.068026749680603e-06, + "loss": 0.3287, + "step": 20377 + }, + { + "epoch": 0.44855822143957735, + "grad_norm": 1.8017209820010176, + "learning_rate": 6.067678517588287e-06, + "loss": 0.3801, + "step": 20378 + }, + { + "epoch": 0.44858023332599606, + "grad_norm": 1.4044852631924778, + "learning_rate": 6.067330280069556e-06, + "loss": 0.4303, + "step": 20379 + }, + { + "epoch": 0.4486022452124147, + "grad_norm": 1.6140525481900738, + "learning_rate": 6.0669820371261745e-06, + "loss": 0.3293, + "step": 20380 + }, + { + "epoch": 0.44862425709883336, + "grad_norm": 0.7109499416185671, + "learning_rate": 6.066633788759913e-06, + "loss": 0.4994, + "step": 20381 + }, + { + "epoch": 0.44864626898525206, + "grad_norm": 1.575753572742922, + "learning_rate": 6.066285534972544e-06, + "loss": 0.3351, + "step": 20382 + }, + { + "epoch": 0.4486682808716707, + "grad_norm": 1.1260083498689526, + "learning_rate": 6.065937275765836e-06, + "loss": 0.2626, + "step": 20383 + }, + { + "epoch": 0.44869029275808936, + "grad_norm": 1.224377470426343, + "learning_rate": 6.0655890111415604e-06, + "loss": 0.3653, + "step": 20384 + }, + { + "epoch": 0.448712304644508, + "grad_norm": 1.3829828627166114, + "learning_rate": 6.065240741101486e-06, + "loss": 0.3546, + "step": 20385 + }, + { + "epoch": 0.4487343165309267, + "grad_norm": 1.9813799209481722, + "learning_rate": 6.064892465647382e-06, + "loss": 0.3707, + "step": 20386 + }, + { + "epoch": 0.44875632841734536, + "grad_norm": 1.2509643499561882, + "learning_rate": 6.06454418478102e-06, + "loss": 0.3241, + "step": 20387 + }, + { + "epoch": 0.448778340303764, + "grad_norm": 2.4876185520466674, + "learning_rate": 6.06419589850417e-06, + "loss": 0.3664, + "step": 20388 + }, + { + "epoch": 0.4488003521901827, + "grad_norm": 1.3292356386503814, + "learning_rate": 6.063847606818601e-06, + "loss": 0.3215, + "step": 20389 + }, + { + "epoch": 0.44882236407660137, + "grad_norm": 1.5505444165464204, + "learning_rate": 6.0634993097260855e-06, + "loss": 0.3145, + "step": 20390 + }, + { + "epoch": 0.44884437596302, + "grad_norm": 1.841434817197548, + "learning_rate": 6.063151007228389e-06, + "loss": 0.3112, + "step": 20391 + }, + { + "epoch": 0.4488663878494387, + "grad_norm": 1.4107557422397552, + "learning_rate": 6.062802699327288e-06, + "loss": 0.3126, + "step": 20392 + }, + { + "epoch": 0.44888839973585737, + "grad_norm": 1.3792835019429008, + "learning_rate": 6.062454386024549e-06, + "loss": 0.3262, + "step": 20393 + }, + { + "epoch": 0.448910411622276, + "grad_norm": 1.8452898735164591, + "learning_rate": 6.062106067321944e-06, + "loss": 0.3042, + "step": 20394 + }, + { + "epoch": 0.4489324235086947, + "grad_norm": 1.707037629307067, + "learning_rate": 6.061757743221241e-06, + "loss": 0.2656, + "step": 20395 + }, + { + "epoch": 0.44895443539511337, + "grad_norm": 0.6350222366857735, + "learning_rate": 6.06140941372421e-06, + "loss": 0.4896, + "step": 20396 + }, + { + "epoch": 0.448976447281532, + "grad_norm": 0.6622339942942875, + "learning_rate": 6.061061078832625e-06, + "loss": 0.4877, + "step": 20397 + }, + { + "epoch": 0.44899845916795067, + "grad_norm": 1.7362685899998442, + "learning_rate": 6.0607127385482545e-06, + "loss": 0.3492, + "step": 20398 + }, + { + "epoch": 0.4490204710543694, + "grad_norm": 0.6167176857274254, + "learning_rate": 6.0603643928728675e-06, + "loss": 0.4928, + "step": 20399 + }, + { + "epoch": 0.449042482940788, + "grad_norm": 1.5719851223492216, + "learning_rate": 6.060016041808237e-06, + "loss": 0.3561, + "step": 20400 + }, + { + "epoch": 0.4490644948272067, + "grad_norm": 0.6239443429616514, + "learning_rate": 6.05966768535613e-06, + "loss": 0.4721, + "step": 20401 + }, + { + "epoch": 0.4490865067136254, + "grad_norm": 7.62245176807361, + "learning_rate": 6.059319323518322e-06, + "loss": 0.3712, + "step": 20402 + }, + { + "epoch": 0.449108518600044, + "grad_norm": 1.413943542250418, + "learning_rate": 6.058970956296579e-06, + "loss": 0.2931, + "step": 20403 + }, + { + "epoch": 0.4491305304864627, + "grad_norm": 2.182924246238492, + "learning_rate": 6.058622583692674e-06, + "loss": 0.329, + "step": 20404 + }, + { + "epoch": 0.4491525423728814, + "grad_norm": 1.9595715553279656, + "learning_rate": 6.058274205708377e-06, + "loss": 0.369, + "step": 20405 + }, + { + "epoch": 0.44917455425930003, + "grad_norm": 1.4629718033711399, + "learning_rate": 6.057925822345457e-06, + "loss": 0.3129, + "step": 20406 + }, + { + "epoch": 0.4491965661457187, + "grad_norm": 1.8617516537465943, + "learning_rate": 6.057577433605688e-06, + "loss": 0.3831, + "step": 20407 + }, + { + "epoch": 0.44921857803213733, + "grad_norm": 1.284139782970847, + "learning_rate": 6.057229039490838e-06, + "loss": 0.3148, + "step": 20408 + }, + { + "epoch": 0.44924058991855603, + "grad_norm": 0.6406933563593106, + "learning_rate": 6.056880640002678e-06, + "loss": 0.4818, + "step": 20409 + }, + { + "epoch": 0.4492626018049747, + "grad_norm": 1.3184000491943602, + "learning_rate": 6.05653223514298e-06, + "loss": 0.3763, + "step": 20410 + }, + { + "epoch": 0.44928461369139333, + "grad_norm": 1.3405392979679658, + "learning_rate": 6.056183824913514e-06, + "loss": 0.3378, + "step": 20411 + }, + { + "epoch": 0.44930662557781204, + "grad_norm": 1.9518604289275108, + "learning_rate": 6.05583540931605e-06, + "loss": 0.3239, + "step": 20412 + }, + { + "epoch": 0.4493286374642307, + "grad_norm": 2.19222735911214, + "learning_rate": 6.05548698835236e-06, + "loss": 0.3561, + "step": 20413 + }, + { + "epoch": 0.44935064935064933, + "grad_norm": 1.9654344755803745, + "learning_rate": 6.055138562024214e-06, + "loss": 0.3452, + "step": 20414 + }, + { + "epoch": 0.44937266123706804, + "grad_norm": 1.9938867282462822, + "learning_rate": 6.054790130333384e-06, + "loss": 0.3434, + "step": 20415 + }, + { + "epoch": 0.4493946731234867, + "grad_norm": 1.5591133609203573, + "learning_rate": 6.0544416932816405e-06, + "loss": 0.3548, + "step": 20416 + }, + { + "epoch": 0.44941668500990534, + "grad_norm": 1.4855288669353988, + "learning_rate": 6.054093250870752e-06, + "loss": 0.3736, + "step": 20417 + }, + { + "epoch": 0.44943869689632404, + "grad_norm": 1.4827110275332032, + "learning_rate": 6.053744803102493e-06, + "loss": 0.3305, + "step": 20418 + }, + { + "epoch": 0.4494607087827427, + "grad_norm": 0.6203750037184279, + "learning_rate": 6.053396349978632e-06, + "loss": 0.4728, + "step": 20419 + }, + { + "epoch": 0.44948272066916134, + "grad_norm": 1.444929273725009, + "learning_rate": 6.0530478915009426e-06, + "loss": 0.3431, + "step": 20420 + }, + { + "epoch": 0.44950473255558, + "grad_norm": 1.2927255264898054, + "learning_rate": 6.0526994276711925e-06, + "loss": 0.296, + "step": 20421 + }, + { + "epoch": 0.4495267444419987, + "grad_norm": 0.6302329236519564, + "learning_rate": 6.052350958491155e-06, + "loss": 0.4769, + "step": 20422 + }, + { + "epoch": 0.44954875632841734, + "grad_norm": 0.6086961261510636, + "learning_rate": 6.0520024839626e-06, + "loss": 0.4899, + "step": 20423 + }, + { + "epoch": 0.449570768214836, + "grad_norm": 1.7224596258283287, + "learning_rate": 6.0516540040873e-06, + "loss": 0.3493, + "step": 20424 + }, + { + "epoch": 0.4495927801012547, + "grad_norm": 1.6602340458782487, + "learning_rate": 6.0513055188670245e-06, + "loss": 0.2797, + "step": 20425 + }, + { + "epoch": 0.44961479198767335, + "grad_norm": 1.8389743413037432, + "learning_rate": 6.050957028303547e-06, + "loss": 0.3392, + "step": 20426 + }, + { + "epoch": 0.449636803874092, + "grad_norm": 1.6843939907966028, + "learning_rate": 6.050608532398634e-06, + "loss": 0.3105, + "step": 20427 + }, + { + "epoch": 0.4496588157605107, + "grad_norm": 1.3989806331597086, + "learning_rate": 6.0502600311540625e-06, + "loss": 0.3371, + "step": 20428 + }, + { + "epoch": 0.44968082764692935, + "grad_norm": 1.7558823730004685, + "learning_rate": 6.049911524571601e-06, + "loss": 0.3144, + "step": 20429 + }, + { + "epoch": 0.449702839533348, + "grad_norm": 1.29696455456361, + "learning_rate": 6.049563012653021e-06, + "loss": 0.4099, + "step": 20430 + }, + { + "epoch": 0.44972485141976665, + "grad_norm": 1.7350826235972197, + "learning_rate": 6.049214495400092e-06, + "loss": 0.2758, + "step": 20431 + }, + { + "epoch": 0.44974686330618535, + "grad_norm": 1.4518063801145784, + "learning_rate": 6.048865972814587e-06, + "loss": 0.306, + "step": 20432 + }, + { + "epoch": 0.449768875192604, + "grad_norm": 0.6165954014993587, + "learning_rate": 6.04851744489828e-06, + "loss": 0.4816, + "step": 20433 + }, + { + "epoch": 0.44979088707902265, + "grad_norm": 1.6816858843906464, + "learning_rate": 6.048168911652938e-06, + "loss": 0.2883, + "step": 20434 + }, + { + "epoch": 0.44981289896544135, + "grad_norm": 0.6280103595961806, + "learning_rate": 6.0478203730803335e-06, + "loss": 0.4803, + "step": 20435 + }, + { + "epoch": 0.44983491085186, + "grad_norm": 1.415279590200175, + "learning_rate": 6.047471829182239e-06, + "loss": 0.3245, + "step": 20436 + }, + { + "epoch": 0.44985692273827865, + "grad_norm": 1.473653750103871, + "learning_rate": 6.0471232799604265e-06, + "loss": 0.2531, + "step": 20437 + }, + { + "epoch": 0.44987893462469736, + "grad_norm": 3.1707320170483198, + "learning_rate": 6.046774725416666e-06, + "loss": 0.2542, + "step": 20438 + }, + { + "epoch": 0.449900946511116, + "grad_norm": 0.6077014783217392, + "learning_rate": 6.046426165552728e-06, + "loss": 0.4936, + "step": 20439 + }, + { + "epoch": 0.44992295839753466, + "grad_norm": 0.6204489194535969, + "learning_rate": 6.046077600370386e-06, + "loss": 0.5046, + "step": 20440 + }, + { + "epoch": 0.44994497028395336, + "grad_norm": 2.242976578252008, + "learning_rate": 6.045729029871412e-06, + "loss": 0.3487, + "step": 20441 + }, + { + "epoch": 0.449966982170372, + "grad_norm": 0.6079509160588682, + "learning_rate": 6.045380454057576e-06, + "loss": 0.504, + "step": 20442 + }, + { + "epoch": 0.44998899405679066, + "grad_norm": 0.6030167435672233, + "learning_rate": 6.045031872930651e-06, + "loss": 0.4863, + "step": 20443 + }, + { + "epoch": 0.4500110059432093, + "grad_norm": 1.728588195127237, + "learning_rate": 6.044683286492408e-06, + "loss": 0.4101, + "step": 20444 + }, + { + "epoch": 0.450033017829628, + "grad_norm": 1.379857307447896, + "learning_rate": 6.044334694744617e-06, + "loss": 0.294, + "step": 20445 + }, + { + "epoch": 0.45005502971604666, + "grad_norm": 1.5199650758178025, + "learning_rate": 6.043986097689052e-06, + "loss": 0.374, + "step": 20446 + }, + { + "epoch": 0.4500770416024653, + "grad_norm": 1.3320445348791288, + "learning_rate": 6.043637495327484e-06, + "loss": 0.3427, + "step": 20447 + }, + { + "epoch": 0.450099053488884, + "grad_norm": 3.3326391604344376, + "learning_rate": 6.0432888876616855e-06, + "loss": 0.4437, + "step": 20448 + }, + { + "epoch": 0.45012106537530266, + "grad_norm": 0.6323197456123207, + "learning_rate": 6.042940274693427e-06, + "loss": 0.4642, + "step": 20449 + }, + { + "epoch": 0.4501430772617213, + "grad_norm": 1.7295125292808544, + "learning_rate": 6.042591656424481e-06, + "loss": 0.3103, + "step": 20450 + }, + { + "epoch": 0.45016508914814, + "grad_norm": 4.686175907396241, + "learning_rate": 6.042243032856619e-06, + "loss": 0.3928, + "step": 20451 + }, + { + "epoch": 0.45018710103455867, + "grad_norm": 1.4460723789108911, + "learning_rate": 6.041894403991615e-06, + "loss": 0.2719, + "step": 20452 + }, + { + "epoch": 0.4502091129209773, + "grad_norm": 1.1690077811067212, + "learning_rate": 6.041545769831237e-06, + "loss": 0.2607, + "step": 20453 + }, + { + "epoch": 0.450231124807396, + "grad_norm": 0.6336840005799278, + "learning_rate": 6.0411971303772586e-06, + "loss": 0.4832, + "step": 20454 + }, + { + "epoch": 0.45025313669381467, + "grad_norm": 2.2743354645980816, + "learning_rate": 6.040848485631452e-06, + "loss": 0.363, + "step": 20455 + }, + { + "epoch": 0.4502751485802333, + "grad_norm": 1.547504338722041, + "learning_rate": 6.04049983559559e-06, + "loss": 0.3362, + "step": 20456 + }, + { + "epoch": 0.45029716046665197, + "grad_norm": 1.3171321425474105, + "learning_rate": 6.040151180271444e-06, + "loss": 0.2905, + "step": 20457 + }, + { + "epoch": 0.4503191723530707, + "grad_norm": 1.3583934220914822, + "learning_rate": 6.0398025196607845e-06, + "loss": 0.2967, + "step": 20458 + }, + { + "epoch": 0.4503411842394893, + "grad_norm": 1.5544777233804425, + "learning_rate": 6.0394538537653855e-06, + "loss": 0.2654, + "step": 20459 + }, + { + "epoch": 0.45036319612590797, + "grad_norm": 1.313513815420778, + "learning_rate": 6.039105182587018e-06, + "loss": 0.3495, + "step": 20460 + }, + { + "epoch": 0.4503852080123267, + "grad_norm": 2.0817218230029106, + "learning_rate": 6.038756506127456e-06, + "loss": 0.338, + "step": 20461 + }, + { + "epoch": 0.4504072198987453, + "grad_norm": 1.382592228969089, + "learning_rate": 6.038407824388469e-06, + "loss": 0.2984, + "step": 20462 + }, + { + "epoch": 0.450429231785164, + "grad_norm": 1.497957491616577, + "learning_rate": 6.038059137371829e-06, + "loss": 0.351, + "step": 20463 + }, + { + "epoch": 0.4504512436715827, + "grad_norm": 2.2493344309803986, + "learning_rate": 6.037710445079311e-06, + "loss": 0.2873, + "step": 20464 + }, + { + "epoch": 0.45047325555800133, + "grad_norm": 0.6427986632213121, + "learning_rate": 6.037361747512686e-06, + "loss": 0.4884, + "step": 20465 + }, + { + "epoch": 0.45049526744442, + "grad_norm": 1.8113966112225317, + "learning_rate": 6.037013044673725e-06, + "loss": 0.3421, + "step": 20466 + }, + { + "epoch": 0.4505172793308386, + "grad_norm": 1.5013581569882177, + "learning_rate": 6.036664336564202e-06, + "loss": 0.2674, + "step": 20467 + }, + { + "epoch": 0.45053929121725733, + "grad_norm": 1.1176762373515399, + "learning_rate": 6.036315623185888e-06, + "loss": 0.2625, + "step": 20468 + }, + { + "epoch": 0.450561303103676, + "grad_norm": 1.503206685168303, + "learning_rate": 6.035966904540556e-06, + "loss": 0.348, + "step": 20469 + }, + { + "epoch": 0.45058331499009463, + "grad_norm": 1.6599871003262674, + "learning_rate": 6.035618180629979e-06, + "loss": 0.2972, + "step": 20470 + }, + { + "epoch": 0.45060532687651333, + "grad_norm": 1.6119308460465067, + "learning_rate": 6.035269451455926e-06, + "loss": 0.2697, + "step": 20471 + }, + { + "epoch": 0.450627338762932, + "grad_norm": 0.6278637250639975, + "learning_rate": 6.034920717020174e-06, + "loss": 0.505, + "step": 20472 + }, + { + "epoch": 0.45064935064935063, + "grad_norm": 1.316669526974313, + "learning_rate": 6.034571977324493e-06, + "loss": 0.2902, + "step": 20473 + }, + { + "epoch": 0.45067136253576934, + "grad_norm": 2.4534731351500207, + "learning_rate": 6.034223232370656e-06, + "loss": 0.3977, + "step": 20474 + }, + { + "epoch": 0.450693374422188, + "grad_norm": 0.591930272708362, + "learning_rate": 6.033874482160436e-06, + "loss": 0.465, + "step": 20475 + }, + { + "epoch": 0.45071538630860664, + "grad_norm": 2.589223214984489, + "learning_rate": 6.0335257266956015e-06, + "loss": 0.2383, + "step": 20476 + }, + { + "epoch": 0.45073739819502534, + "grad_norm": 1.228201964455313, + "learning_rate": 6.03317696597793e-06, + "loss": 0.2858, + "step": 20477 + }, + { + "epoch": 0.450759410081444, + "grad_norm": 1.7129655913517983, + "learning_rate": 6.032828200009193e-06, + "loss": 0.3095, + "step": 20478 + }, + { + "epoch": 0.45078142196786264, + "grad_norm": 3.7638499332762736, + "learning_rate": 6.032479428791162e-06, + "loss": 0.4271, + "step": 20479 + }, + { + "epoch": 0.4508034338542813, + "grad_norm": 0.6441457705979198, + "learning_rate": 6.032130652325612e-06, + "loss": 0.4968, + "step": 20480 + }, + { + "epoch": 0.4508254457407, + "grad_norm": 0.6180416418493871, + "learning_rate": 6.031781870614311e-06, + "loss": 0.4691, + "step": 20481 + }, + { + "epoch": 0.45084745762711864, + "grad_norm": 1.2515800844070415, + "learning_rate": 6.031433083659035e-06, + "loss": 0.3155, + "step": 20482 + }, + { + "epoch": 0.4508694695135373, + "grad_norm": 1.474516510065507, + "learning_rate": 6.031084291461556e-06, + "loss": 0.2721, + "step": 20483 + }, + { + "epoch": 0.450891481399956, + "grad_norm": 1.2399082971977364, + "learning_rate": 6.030735494023647e-06, + "loss": 0.3587, + "step": 20484 + }, + { + "epoch": 0.45091349328637464, + "grad_norm": 2.860039464074338, + "learning_rate": 6.030386691347081e-06, + "loss": 0.2782, + "step": 20485 + }, + { + "epoch": 0.4509355051727933, + "grad_norm": 1.142205591387136, + "learning_rate": 6.030037883433629e-06, + "loss": 0.2735, + "step": 20486 + }, + { + "epoch": 0.450957517059212, + "grad_norm": 1.3442166434153053, + "learning_rate": 6.029689070285066e-06, + "loss": 0.2558, + "step": 20487 + }, + { + "epoch": 0.45097952894563065, + "grad_norm": 1.4041974626519333, + "learning_rate": 6.0293402519031654e-06, + "loss": 0.2883, + "step": 20488 + }, + { + "epoch": 0.4510015408320493, + "grad_norm": 1.1849865525315366, + "learning_rate": 6.0289914282896955e-06, + "loss": 0.3099, + "step": 20489 + }, + { + "epoch": 0.45102355271846795, + "grad_norm": 1.7562529924164791, + "learning_rate": 6.028642599446434e-06, + "loss": 0.3351, + "step": 20490 + }, + { + "epoch": 0.45104556460488665, + "grad_norm": 7.772739481082122, + "learning_rate": 6.028293765375151e-06, + "loss": 0.4364, + "step": 20491 + }, + { + "epoch": 0.4510675764913053, + "grad_norm": 0.6276605841012467, + "learning_rate": 6.027944926077621e-06, + "loss": 0.5057, + "step": 20492 + }, + { + "epoch": 0.45108958837772395, + "grad_norm": 1.3591343995966931, + "learning_rate": 6.027596081555618e-06, + "loss": 0.3242, + "step": 20493 + }, + { + "epoch": 0.45111160026414265, + "grad_norm": 1.7245966398210375, + "learning_rate": 6.027247231810911e-06, + "loss": 0.4158, + "step": 20494 + }, + { + "epoch": 0.4511336121505613, + "grad_norm": 1.5901163728491963, + "learning_rate": 6.026898376845276e-06, + "loss": 0.4159, + "step": 20495 + }, + { + "epoch": 0.45115562403697995, + "grad_norm": 1.4371727448085716, + "learning_rate": 6.026549516660486e-06, + "loss": 0.4302, + "step": 20496 + }, + { + "epoch": 0.45117763592339866, + "grad_norm": 1.6074656343219864, + "learning_rate": 6.026200651258314e-06, + "loss": 0.2295, + "step": 20497 + }, + { + "epoch": 0.4511996478098173, + "grad_norm": 1.1134136779380368, + "learning_rate": 6.025851780640532e-06, + "loss": 0.2184, + "step": 20498 + }, + { + "epoch": 0.45122165969623595, + "grad_norm": 4.171314924105105, + "learning_rate": 6.025502904808912e-06, + "loss": 0.3024, + "step": 20499 + }, + { + "epoch": 0.45124367158265466, + "grad_norm": 1.6495327264356021, + "learning_rate": 6.025154023765231e-06, + "loss": 0.3475, + "step": 20500 + }, + { + "epoch": 0.4512656834690733, + "grad_norm": 0.6144456242673513, + "learning_rate": 6.0248051375112596e-06, + "loss": 0.4749, + "step": 20501 + }, + { + "epoch": 0.45128769535549196, + "grad_norm": 1.5394417897991453, + "learning_rate": 6.024456246048771e-06, + "loss": 0.3001, + "step": 20502 + }, + { + "epoch": 0.4513097072419106, + "grad_norm": 3.9530409173658834, + "learning_rate": 6.0241073493795385e-06, + "loss": 0.2889, + "step": 20503 + }, + { + "epoch": 0.4513317191283293, + "grad_norm": 1.3636231480680456, + "learning_rate": 6.023758447505336e-06, + "loss": 0.3558, + "step": 20504 + }, + { + "epoch": 0.45135373101474796, + "grad_norm": 1.3277444828238967, + "learning_rate": 6.023409540427937e-06, + "loss": 0.3136, + "step": 20505 + }, + { + "epoch": 0.4513757429011666, + "grad_norm": 1.5161509806759552, + "learning_rate": 6.023060628149116e-06, + "loss": 0.3798, + "step": 20506 + }, + { + "epoch": 0.4513977547875853, + "grad_norm": 1.382725225776993, + "learning_rate": 6.022711710670641e-06, + "loss": 0.373, + "step": 20507 + }, + { + "epoch": 0.45141976667400396, + "grad_norm": 0.6053359377527588, + "learning_rate": 6.02236278799429e-06, + "loss": 0.4856, + "step": 20508 + }, + { + "epoch": 0.4514417785604226, + "grad_norm": 1.6611970040318969, + "learning_rate": 6.0220138601218355e-06, + "loss": 0.4664, + "step": 20509 + }, + { + "epoch": 0.4514637904468413, + "grad_norm": 1.5689039985968374, + "learning_rate": 6.021664927055051e-06, + "loss": 0.3018, + "step": 20510 + }, + { + "epoch": 0.45148580233325997, + "grad_norm": 1.5073354525890539, + "learning_rate": 6.021315988795711e-06, + "loss": 0.2621, + "step": 20511 + }, + { + "epoch": 0.4515078142196786, + "grad_norm": 2.344486032089574, + "learning_rate": 6.020967045345585e-06, + "loss": 0.3135, + "step": 20512 + }, + { + "epoch": 0.45152982610609727, + "grad_norm": 1.467622336478022, + "learning_rate": 6.020618096706451e-06, + "loss": 0.2416, + "step": 20513 + }, + { + "epoch": 0.45155183799251597, + "grad_norm": 1.2783047179451086, + "learning_rate": 6.02026914288008e-06, + "loss": 0.3617, + "step": 20514 + }, + { + "epoch": 0.4515738498789346, + "grad_norm": 1.9059589494505105, + "learning_rate": 6.019920183868247e-06, + "loss": 0.2704, + "step": 20515 + }, + { + "epoch": 0.45159586176535327, + "grad_norm": 1.3723441284991054, + "learning_rate": 6.019571219672724e-06, + "loss": 0.3106, + "step": 20516 + }, + { + "epoch": 0.451617873651772, + "grad_norm": 1.5529577316054477, + "learning_rate": 6.019222250295285e-06, + "loss": 0.3171, + "step": 20517 + }, + { + "epoch": 0.4516398855381906, + "grad_norm": 1.8496640929406953, + "learning_rate": 6.018873275737704e-06, + "loss": 0.2874, + "step": 20518 + }, + { + "epoch": 0.45166189742460927, + "grad_norm": 1.3424703167184708, + "learning_rate": 6.018524296001754e-06, + "loss": 0.3429, + "step": 20519 + }, + { + "epoch": 0.451683909311028, + "grad_norm": 1.3018914315323937, + "learning_rate": 6.0181753110892105e-06, + "loss": 0.357, + "step": 20520 + }, + { + "epoch": 0.4517059211974466, + "grad_norm": 1.4334008070343296, + "learning_rate": 6.017826321001846e-06, + "loss": 0.4286, + "step": 20521 + }, + { + "epoch": 0.4517279330838653, + "grad_norm": 1.4066239025035001, + "learning_rate": 6.0174773257414336e-06, + "loss": 0.3434, + "step": 20522 + }, + { + "epoch": 0.451749944970284, + "grad_norm": 1.4429725220148797, + "learning_rate": 6.017128325309747e-06, + "loss": 0.2835, + "step": 20523 + }, + { + "epoch": 0.45177195685670263, + "grad_norm": 1.3688407379321967, + "learning_rate": 6.016779319708562e-06, + "loss": 0.2863, + "step": 20524 + }, + { + "epoch": 0.4517939687431213, + "grad_norm": 1.3851180990458243, + "learning_rate": 6.01643030893965e-06, + "loss": 0.3723, + "step": 20525 + }, + { + "epoch": 0.4518159806295399, + "grad_norm": 1.647777890131045, + "learning_rate": 6.016081293004786e-06, + "loss": 0.4025, + "step": 20526 + }, + { + "epoch": 0.45183799251595863, + "grad_norm": 1.1594700268077822, + "learning_rate": 6.015732271905744e-06, + "loss": 0.2773, + "step": 20527 + }, + { + "epoch": 0.4518600044023773, + "grad_norm": 1.5798363426997188, + "learning_rate": 6.0153832456442974e-06, + "loss": 0.3013, + "step": 20528 + }, + { + "epoch": 0.45188201628879593, + "grad_norm": 1.1339323423756844, + "learning_rate": 6.015034214222221e-06, + "loss": 0.2877, + "step": 20529 + }, + { + "epoch": 0.45190402817521463, + "grad_norm": 1.2481077822901718, + "learning_rate": 6.014685177641286e-06, + "loss": 0.2346, + "step": 20530 + }, + { + "epoch": 0.4519260400616333, + "grad_norm": 1.238407568150613, + "learning_rate": 6.0143361359032705e-06, + "loss": 0.3869, + "step": 20531 + }, + { + "epoch": 0.45194805194805193, + "grad_norm": 1.6333408306046633, + "learning_rate": 6.013987089009945e-06, + "loss": 0.2631, + "step": 20532 + }, + { + "epoch": 0.45197006383447064, + "grad_norm": 1.4080312617106014, + "learning_rate": 6.013638036963086e-06, + "loss": 0.2633, + "step": 20533 + }, + { + "epoch": 0.4519920757208893, + "grad_norm": 1.3751944293076628, + "learning_rate": 6.0132889797644655e-06, + "loss": 0.2625, + "step": 20534 + }, + { + "epoch": 0.45201408760730794, + "grad_norm": 0.6618215161838894, + "learning_rate": 6.012939917415858e-06, + "loss": 0.4701, + "step": 20535 + }, + { + "epoch": 0.45203609949372664, + "grad_norm": 1.566426308465524, + "learning_rate": 6.012590849919039e-06, + "loss": 0.3105, + "step": 20536 + }, + { + "epoch": 0.4520581113801453, + "grad_norm": 1.449319777171302, + "learning_rate": 6.012241777275781e-06, + "loss": 0.3273, + "step": 20537 + }, + { + "epoch": 0.45208012326656394, + "grad_norm": 1.4224100944610005, + "learning_rate": 6.01189269948786e-06, + "loss": 0.3078, + "step": 20538 + }, + { + "epoch": 0.4521021351529826, + "grad_norm": 2.052126734664008, + "learning_rate": 6.0115436165570485e-06, + "loss": 0.385, + "step": 20539 + }, + { + "epoch": 0.4521241470394013, + "grad_norm": 1.1932279175881146, + "learning_rate": 6.01119452848512e-06, + "loss": 0.2912, + "step": 20540 + }, + { + "epoch": 0.45214615892581994, + "grad_norm": 1.4079435257697652, + "learning_rate": 6.010845435273851e-06, + "loss": 0.3347, + "step": 20541 + }, + { + "epoch": 0.4521681708122386, + "grad_norm": 2.8751681995220264, + "learning_rate": 6.010496336925015e-06, + "loss": 0.3608, + "step": 20542 + }, + { + "epoch": 0.4521901826986573, + "grad_norm": 1.560880619842431, + "learning_rate": 6.010147233440385e-06, + "loss": 0.3547, + "step": 20543 + }, + { + "epoch": 0.45221219458507594, + "grad_norm": 1.346210527769296, + "learning_rate": 6.009798124821737e-06, + "loss": 0.3294, + "step": 20544 + }, + { + "epoch": 0.4522342064714946, + "grad_norm": 1.4286363899649994, + "learning_rate": 6.009449011070845e-06, + "loss": 0.3685, + "step": 20545 + }, + { + "epoch": 0.4522562183579133, + "grad_norm": 1.2356573874481962, + "learning_rate": 6.00909989218948e-06, + "loss": 0.3413, + "step": 20546 + }, + { + "epoch": 0.45227823024433195, + "grad_norm": 0.6102861712561694, + "learning_rate": 6.008750768179423e-06, + "loss": 0.4651, + "step": 20547 + }, + { + "epoch": 0.4523002421307506, + "grad_norm": 2.5182979854115772, + "learning_rate": 6.008401639042442e-06, + "loss": 0.2971, + "step": 20548 + }, + { + "epoch": 0.45232225401716925, + "grad_norm": 1.8176644904085555, + "learning_rate": 6.008052504780315e-06, + "loss": 0.3755, + "step": 20549 + }, + { + "epoch": 0.45234426590358795, + "grad_norm": 1.2594443931635957, + "learning_rate": 6.007703365394816e-06, + "loss": 0.3257, + "step": 20550 + }, + { + "epoch": 0.4523662777900066, + "grad_norm": 1.2089315523746784, + "learning_rate": 6.0073542208877174e-06, + "loss": 0.2375, + "step": 20551 + }, + { + "epoch": 0.45238828967642525, + "grad_norm": 1.2445330213402175, + "learning_rate": 6.007005071260799e-06, + "loss": 0.2198, + "step": 20552 + }, + { + "epoch": 0.45241030156284395, + "grad_norm": 1.570716795207798, + "learning_rate": 6.006655916515828e-06, + "loss": 0.3781, + "step": 20553 + }, + { + "epoch": 0.4524323134492626, + "grad_norm": 1.2782923776656867, + "learning_rate": 6.0063067566545845e-06, + "loss": 0.308, + "step": 20554 + }, + { + "epoch": 0.45245432533568125, + "grad_norm": 1.5966503355855384, + "learning_rate": 6.005957591678841e-06, + "loss": 0.3398, + "step": 20555 + }, + { + "epoch": 0.45247633722209996, + "grad_norm": 2.583900068528565, + "learning_rate": 6.005608421590372e-06, + "loss": 0.3324, + "step": 20556 + }, + { + "epoch": 0.4524983491085186, + "grad_norm": 1.3417279346503195, + "learning_rate": 6.005259246390953e-06, + "loss": 0.2799, + "step": 20557 + }, + { + "epoch": 0.45252036099493725, + "grad_norm": 1.8495640216399964, + "learning_rate": 6.004910066082357e-06, + "loss": 0.3076, + "step": 20558 + }, + { + "epoch": 0.45254237288135596, + "grad_norm": 1.557642955041598, + "learning_rate": 6.0045608806663604e-06, + "loss": 0.2982, + "step": 20559 + }, + { + "epoch": 0.4525643847677746, + "grad_norm": 1.9907697428677293, + "learning_rate": 6.0042116901447375e-06, + "loss": 0.314, + "step": 20560 + }, + { + "epoch": 0.45258639665419326, + "grad_norm": 1.5036775941384815, + "learning_rate": 6.003862494519263e-06, + "loss": 0.408, + "step": 20561 + }, + { + "epoch": 0.4526084085406119, + "grad_norm": 1.3355096442348415, + "learning_rate": 6.0035132937917116e-06, + "loss": 0.3456, + "step": 20562 + }, + { + "epoch": 0.4526304204270306, + "grad_norm": 2.129339041643367, + "learning_rate": 6.0031640879638585e-06, + "loss": 0.355, + "step": 20563 + }, + { + "epoch": 0.45265243231344926, + "grad_norm": 1.9569165674590212, + "learning_rate": 6.002814877037477e-06, + "loss": 0.293, + "step": 20564 + }, + { + "epoch": 0.4526744441998679, + "grad_norm": 1.7433742995322803, + "learning_rate": 6.002465661014345e-06, + "loss": 0.2966, + "step": 20565 + }, + { + "epoch": 0.4526964560862866, + "grad_norm": 1.8776684620498354, + "learning_rate": 6.0021164398962326e-06, + "loss": 0.2766, + "step": 20566 + }, + { + "epoch": 0.45271846797270526, + "grad_norm": 1.563054959495006, + "learning_rate": 6.0017672136849196e-06, + "loss": 0.2408, + "step": 20567 + }, + { + "epoch": 0.4527404798591239, + "grad_norm": 1.3428803825515432, + "learning_rate": 6.001417982382178e-06, + "loss": 0.3993, + "step": 20568 + }, + { + "epoch": 0.4527624917455426, + "grad_norm": 1.4307615739919703, + "learning_rate": 6.001068745989783e-06, + "loss": 0.385, + "step": 20569 + }, + { + "epoch": 0.45278450363196127, + "grad_norm": 0.6681215021655238, + "learning_rate": 6.000719504509512e-06, + "loss": 0.513, + "step": 20570 + }, + { + "epoch": 0.4528065155183799, + "grad_norm": 1.7658150296142974, + "learning_rate": 6.000370257943136e-06, + "loss": 0.3389, + "step": 20571 + }, + { + "epoch": 0.45282852740479856, + "grad_norm": 1.715375783943655, + "learning_rate": 6.000021006292433e-06, + "loss": 0.3727, + "step": 20572 + }, + { + "epoch": 0.45285053929121727, + "grad_norm": 0.6133952938921625, + "learning_rate": 5.9996717495591775e-06, + "loss": 0.4441, + "step": 20573 + }, + { + "epoch": 0.4528725511776359, + "grad_norm": 1.7624104295126743, + "learning_rate": 5.999322487745145e-06, + "loss": 0.4177, + "step": 20574 + }, + { + "epoch": 0.45289456306405457, + "grad_norm": 1.4122040091663977, + "learning_rate": 5.998973220852109e-06, + "loss": 0.3356, + "step": 20575 + }, + { + "epoch": 0.45291657495047327, + "grad_norm": 4.531873224609208, + "learning_rate": 5.998623948881844e-06, + "loss": 0.2865, + "step": 20576 + }, + { + "epoch": 0.4529385868368919, + "grad_norm": 3.1539993130108894, + "learning_rate": 5.99827467183613e-06, + "loss": 0.4153, + "step": 20577 + }, + { + "epoch": 0.45296059872331057, + "grad_norm": 1.2897772937242802, + "learning_rate": 5.997925389716737e-06, + "loss": 0.285, + "step": 20578 + }, + { + "epoch": 0.4529826106097293, + "grad_norm": 1.206951612317637, + "learning_rate": 5.997576102525442e-06, + "loss": 0.2755, + "step": 20579 + }, + { + "epoch": 0.4530046224961479, + "grad_norm": 1.5831617371127598, + "learning_rate": 5.997226810264021e-06, + "loss": 0.3156, + "step": 20580 + }, + { + "epoch": 0.4530266343825666, + "grad_norm": 1.2485138518349441, + "learning_rate": 5.996877512934248e-06, + "loss": 0.3463, + "step": 20581 + }, + { + "epoch": 0.4530486462689853, + "grad_norm": 1.949086433171905, + "learning_rate": 5.996528210537898e-06, + "loss": 0.3189, + "step": 20582 + }, + { + "epoch": 0.4530706581554039, + "grad_norm": 1.8745086685152246, + "learning_rate": 5.996178903076749e-06, + "loss": 0.3169, + "step": 20583 + }, + { + "epoch": 0.4530926700418226, + "grad_norm": 0.6528168042872943, + "learning_rate": 5.995829590552573e-06, + "loss": 0.4897, + "step": 20584 + }, + { + "epoch": 0.4531146819282412, + "grad_norm": 2.044916396602956, + "learning_rate": 5.995480272967148e-06, + "loss": 0.3065, + "step": 20585 + }, + { + "epoch": 0.45313669381465993, + "grad_norm": 1.238143232574108, + "learning_rate": 5.995130950322248e-06, + "loss": 0.23, + "step": 20586 + }, + { + "epoch": 0.4531587057010786, + "grad_norm": 2.081000732876949, + "learning_rate": 5.994781622619647e-06, + "loss": 0.3012, + "step": 20587 + }, + { + "epoch": 0.45318071758749723, + "grad_norm": 1.4803202305640024, + "learning_rate": 5.994432289861125e-06, + "loss": 0.427, + "step": 20588 + }, + { + "epoch": 0.45320272947391593, + "grad_norm": 1.526824241991177, + "learning_rate": 5.994082952048452e-06, + "loss": 0.3201, + "step": 20589 + }, + { + "epoch": 0.4532247413603346, + "grad_norm": 1.7151116044048405, + "learning_rate": 5.993733609183407e-06, + "loss": 0.3388, + "step": 20590 + }, + { + "epoch": 0.45324675324675323, + "grad_norm": 1.2816316066669524, + "learning_rate": 5.993384261267766e-06, + "loss": 0.2542, + "step": 20591 + }, + { + "epoch": 0.45326876513317194, + "grad_norm": 1.6959143828650975, + "learning_rate": 5.993034908303301e-06, + "loss": 0.3487, + "step": 20592 + }, + { + "epoch": 0.4532907770195906, + "grad_norm": 1.6442911827859952, + "learning_rate": 5.992685550291792e-06, + "loss": 0.307, + "step": 20593 + }, + { + "epoch": 0.45331278890600923, + "grad_norm": 1.8028279833495098, + "learning_rate": 5.9923361872350104e-06, + "loss": 0.3925, + "step": 20594 + }, + { + "epoch": 0.45333480079242794, + "grad_norm": 1.570511659866759, + "learning_rate": 5.991986819134734e-06, + "loss": 0.406, + "step": 20595 + }, + { + "epoch": 0.4533568126788466, + "grad_norm": 1.5264352424190835, + "learning_rate": 5.9916374459927384e-06, + "loss": 0.3727, + "step": 20596 + }, + { + "epoch": 0.45337882456526524, + "grad_norm": 0.5992838111847112, + "learning_rate": 5.991288067810799e-06, + "loss": 0.4622, + "step": 20597 + }, + { + "epoch": 0.4534008364516839, + "grad_norm": 1.506269282674118, + "learning_rate": 5.990938684590693e-06, + "loss": 0.2611, + "step": 20598 + }, + { + "epoch": 0.4534228483381026, + "grad_norm": 2.1424483667087184, + "learning_rate": 5.990589296334193e-06, + "loss": 0.3203, + "step": 20599 + }, + { + "epoch": 0.45344486022452124, + "grad_norm": 1.5896664389663282, + "learning_rate": 5.9902399030430744e-06, + "loss": 0.3242, + "step": 20600 + }, + { + "epoch": 0.4534668721109399, + "grad_norm": 1.2868133806403144, + "learning_rate": 5.989890504719118e-06, + "loss": 0.2969, + "step": 20601 + }, + { + "epoch": 0.4534888839973586, + "grad_norm": 1.2267518629468803, + "learning_rate": 5.989541101364096e-06, + "loss": 0.3001, + "step": 20602 + }, + { + "epoch": 0.45351089588377724, + "grad_norm": 1.4482317939352078, + "learning_rate": 5.989191692979784e-06, + "loss": 0.3135, + "step": 20603 + }, + { + "epoch": 0.4535329077701959, + "grad_norm": 1.909918506591231, + "learning_rate": 5.988842279567959e-06, + "loss": 0.4045, + "step": 20604 + }, + { + "epoch": 0.4535549196566146, + "grad_norm": 2.098719765282688, + "learning_rate": 5.988492861130396e-06, + "loss": 0.3196, + "step": 20605 + }, + { + "epoch": 0.45357693154303325, + "grad_norm": 1.2962437432975311, + "learning_rate": 5.988143437668873e-06, + "loss": 0.3637, + "step": 20606 + }, + { + "epoch": 0.4535989434294519, + "grad_norm": 1.543997890467662, + "learning_rate": 5.9877940091851625e-06, + "loss": 0.2587, + "step": 20607 + }, + { + "epoch": 0.45362095531587054, + "grad_norm": 1.6116314735142863, + "learning_rate": 5.9874445756810425e-06, + "loss": 0.3109, + "step": 20608 + }, + { + "epoch": 0.45364296720228925, + "grad_norm": 0.9657864396930691, + "learning_rate": 5.9870951371582896e-06, + "loss": 0.1945, + "step": 20609 + }, + { + "epoch": 0.4536649790887079, + "grad_norm": 1.2893203745183004, + "learning_rate": 5.986745693618678e-06, + "loss": 0.3196, + "step": 20610 + }, + { + "epoch": 0.45368699097512655, + "grad_norm": 1.4648971354098543, + "learning_rate": 5.986396245063986e-06, + "loss": 0.3842, + "step": 20611 + }, + { + "epoch": 0.45370900286154525, + "grad_norm": 0.6090408558980799, + "learning_rate": 5.986046791495988e-06, + "loss": 0.4745, + "step": 20612 + }, + { + "epoch": 0.4537310147479639, + "grad_norm": 1.4125296600839887, + "learning_rate": 5.985697332916459e-06, + "loss": 0.3247, + "step": 20613 + }, + { + "epoch": 0.45375302663438255, + "grad_norm": 1.6786598927417633, + "learning_rate": 5.985347869327178e-06, + "loss": 0.3106, + "step": 20614 + }, + { + "epoch": 0.45377503852080125, + "grad_norm": 1.8041086187266624, + "learning_rate": 5.98499840072992e-06, + "loss": 0.3926, + "step": 20615 + }, + { + "epoch": 0.4537970504072199, + "grad_norm": 0.6163239329408485, + "learning_rate": 5.98464892712646e-06, + "loss": 0.4763, + "step": 20616 + }, + { + "epoch": 0.45381906229363855, + "grad_norm": 1.2438398857717716, + "learning_rate": 5.984299448518576e-06, + "loss": 0.3081, + "step": 20617 + }, + { + "epoch": 0.45384107418005726, + "grad_norm": 0.6059099007044807, + "learning_rate": 5.9839499649080405e-06, + "loss": 0.4834, + "step": 20618 + }, + { + "epoch": 0.4538630860664759, + "grad_norm": 2.055202624620249, + "learning_rate": 5.983600476296634e-06, + "loss": 0.2838, + "step": 20619 + }, + { + "epoch": 0.45388509795289456, + "grad_norm": 1.2601113406924127, + "learning_rate": 5.983250982686132e-06, + "loss": 0.3148, + "step": 20620 + }, + { + "epoch": 0.4539071098393132, + "grad_norm": 1.4179499031283154, + "learning_rate": 5.98290148407831e-06, + "loss": 0.2877, + "step": 20621 + }, + { + "epoch": 0.4539291217257319, + "grad_norm": 1.469912225559996, + "learning_rate": 5.982551980474943e-06, + "loss": 0.3069, + "step": 20622 + }, + { + "epoch": 0.45395113361215056, + "grad_norm": 1.0856888141387133, + "learning_rate": 5.9822024718778085e-06, + "loss": 0.3034, + "step": 20623 + }, + { + "epoch": 0.4539731454985692, + "grad_norm": 1.589136904604446, + "learning_rate": 5.981852958288684e-06, + "loss": 0.3711, + "step": 20624 + }, + { + "epoch": 0.4539951573849879, + "grad_norm": 1.202783159494141, + "learning_rate": 5.981503439709344e-06, + "loss": 0.2671, + "step": 20625 + }, + { + "epoch": 0.45401716927140656, + "grad_norm": 1.643005540044523, + "learning_rate": 5.981153916141567e-06, + "loss": 0.3833, + "step": 20626 + }, + { + "epoch": 0.4540391811578252, + "grad_norm": 2.00174977106142, + "learning_rate": 5.980804387587127e-06, + "loss": 0.3595, + "step": 20627 + }, + { + "epoch": 0.4540611930442439, + "grad_norm": 0.6241684732434045, + "learning_rate": 5.980454854047801e-06, + "loss": 0.4731, + "step": 20628 + }, + { + "epoch": 0.45408320493066257, + "grad_norm": 1.6397793074430846, + "learning_rate": 5.980105315525368e-06, + "loss": 0.3259, + "step": 20629 + }, + { + "epoch": 0.4541052168170812, + "grad_norm": 1.3870191733861623, + "learning_rate": 5.979755772021602e-06, + "loss": 0.4248, + "step": 20630 + }, + { + "epoch": 0.45412722870349986, + "grad_norm": 0.5922146607783264, + "learning_rate": 5.97940622353828e-06, + "loss": 0.4627, + "step": 20631 + }, + { + "epoch": 0.45414924058991857, + "grad_norm": 1.6075091565766264, + "learning_rate": 5.979056670077179e-06, + "loss": 0.3, + "step": 20632 + }, + { + "epoch": 0.4541712524763372, + "grad_norm": 1.453682568499131, + "learning_rate": 5.978707111640075e-06, + "loss": 0.2827, + "step": 20633 + }, + { + "epoch": 0.45419326436275587, + "grad_norm": 0.6205051587347411, + "learning_rate": 5.978357548228746e-06, + "loss": 0.5099, + "step": 20634 + }, + { + "epoch": 0.45421527624917457, + "grad_norm": 1.3762704745792333, + "learning_rate": 5.978007979844967e-06, + "loss": 0.3315, + "step": 20635 + }, + { + "epoch": 0.4542372881355932, + "grad_norm": 1.6588340249017628, + "learning_rate": 5.977658406490513e-06, + "loss": 0.309, + "step": 20636 + }, + { + "epoch": 0.45425930002201187, + "grad_norm": 1.3163349472078365, + "learning_rate": 5.977308828167165e-06, + "loss": 0.3317, + "step": 20637 + }, + { + "epoch": 0.4542813119084306, + "grad_norm": 1.7142621469316328, + "learning_rate": 5.976959244876697e-06, + "loss": 0.248, + "step": 20638 + }, + { + "epoch": 0.4543033237948492, + "grad_norm": 1.4306981119781956, + "learning_rate": 5.976609656620888e-06, + "loss": 0.2622, + "step": 20639 + }, + { + "epoch": 0.4543253356812679, + "grad_norm": 0.624324541930729, + "learning_rate": 5.976260063401513e-06, + "loss": 0.5029, + "step": 20640 + }, + { + "epoch": 0.4543473475676866, + "grad_norm": 1.4022230937285907, + "learning_rate": 5.975910465220346e-06, + "loss": 0.3156, + "step": 20641 + }, + { + "epoch": 0.4543693594541052, + "grad_norm": 1.2762077076312706, + "learning_rate": 5.9755608620791685e-06, + "loss": 0.278, + "step": 20642 + }, + { + "epoch": 0.4543913713405239, + "grad_norm": 1.6151493520353142, + "learning_rate": 5.9752112539797554e-06, + "loss": 0.3766, + "step": 20643 + }, + { + "epoch": 0.4544133832269425, + "grad_norm": 2.1003783104690155, + "learning_rate": 5.974861640923885e-06, + "loss": 0.3604, + "step": 20644 + }, + { + "epoch": 0.45443539511336123, + "grad_norm": 1.368846354112464, + "learning_rate": 5.974512022913333e-06, + "loss": 0.2418, + "step": 20645 + }, + { + "epoch": 0.4544574069997799, + "grad_norm": 1.3017112954218044, + "learning_rate": 5.974162399949873e-06, + "loss": 0.2089, + "step": 20646 + }, + { + "epoch": 0.4544794188861985, + "grad_norm": 1.6038105394315147, + "learning_rate": 5.973812772035289e-06, + "loss": 0.3603, + "step": 20647 + }, + { + "epoch": 0.45450143077261723, + "grad_norm": 1.4365163429888506, + "learning_rate": 5.973463139171354e-06, + "loss": 0.3282, + "step": 20648 + }, + { + "epoch": 0.4545234426590359, + "grad_norm": 1.655509475735543, + "learning_rate": 5.973113501359842e-06, + "loss": 0.3177, + "step": 20649 + }, + { + "epoch": 0.45454545454545453, + "grad_norm": 1.362028299550445, + "learning_rate": 5.972763858602535e-06, + "loss": 0.3717, + "step": 20650 + }, + { + "epoch": 0.45456746643187323, + "grad_norm": 2.041165211941965, + "learning_rate": 5.972414210901207e-06, + "loss": 0.3622, + "step": 20651 + }, + { + "epoch": 0.4545894783182919, + "grad_norm": 1.4342808773446356, + "learning_rate": 5.97206455825764e-06, + "loss": 0.334, + "step": 20652 + }, + { + "epoch": 0.45461149020471053, + "grad_norm": 2.7158978469253165, + "learning_rate": 5.971714900673604e-06, + "loss": 0.2816, + "step": 20653 + }, + { + "epoch": 0.4546335020911292, + "grad_norm": 0.6527295150848649, + "learning_rate": 5.971365238150879e-06, + "loss": 0.5123, + "step": 20654 + }, + { + "epoch": 0.4546555139775479, + "grad_norm": 0.652828678611904, + "learning_rate": 5.9710155706912446e-06, + "loss": 0.513, + "step": 20655 + }, + { + "epoch": 0.45467752586396654, + "grad_norm": 1.3359583858244803, + "learning_rate": 5.970665898296475e-06, + "loss": 0.3623, + "step": 20656 + }, + { + "epoch": 0.4546995377503852, + "grad_norm": 1.6360416050071622, + "learning_rate": 5.9703162209683495e-06, + "loss": 0.3545, + "step": 20657 + }, + { + "epoch": 0.4547215496368039, + "grad_norm": 1.503222062681089, + "learning_rate": 5.969966538708644e-06, + "loss": 0.2799, + "step": 20658 + }, + { + "epoch": 0.45474356152322254, + "grad_norm": 1.2630056518388182, + "learning_rate": 5.969616851519134e-06, + "loss": 0.3894, + "step": 20659 + }, + { + "epoch": 0.4547655734096412, + "grad_norm": 2.1825227770894045, + "learning_rate": 5.969267159401599e-06, + "loss": 0.3579, + "step": 20660 + }, + { + "epoch": 0.4547875852960599, + "grad_norm": 1.822557399823794, + "learning_rate": 5.968917462357819e-06, + "loss": 0.325, + "step": 20661 + }, + { + "epoch": 0.45480959718247854, + "grad_norm": 0.6407352727259615, + "learning_rate": 5.9685677603895655e-06, + "loss": 0.4826, + "step": 20662 + }, + { + "epoch": 0.4548316090688972, + "grad_norm": 1.9602995935387506, + "learning_rate": 5.96821805349862e-06, + "loss": 0.2843, + "step": 20663 + }, + { + "epoch": 0.4548536209553159, + "grad_norm": 1.2455779860092477, + "learning_rate": 5.967868341686757e-06, + "loss": 0.3133, + "step": 20664 + }, + { + "epoch": 0.45487563284173455, + "grad_norm": 1.1997427853195912, + "learning_rate": 5.967518624955757e-06, + "loss": 0.2612, + "step": 20665 + }, + { + "epoch": 0.4548976447281532, + "grad_norm": 1.643328438943119, + "learning_rate": 5.967168903307396e-06, + "loss": 0.3685, + "step": 20666 + }, + { + "epoch": 0.45491965661457184, + "grad_norm": 2.455517404738464, + "learning_rate": 5.966819176743449e-06, + "loss": 0.3779, + "step": 20667 + }, + { + "epoch": 0.45494166850099055, + "grad_norm": 1.3402819837724143, + "learning_rate": 5.966469445265697e-06, + "loss": 0.3196, + "step": 20668 + }, + { + "epoch": 0.4549636803874092, + "grad_norm": 0.6019595954852546, + "learning_rate": 5.966119708875916e-06, + "loss": 0.4534, + "step": 20669 + }, + { + "epoch": 0.45498569227382785, + "grad_norm": 0.6116192426946543, + "learning_rate": 5.965769967575885e-06, + "loss": 0.5037, + "step": 20670 + }, + { + "epoch": 0.45500770416024655, + "grad_norm": 1.3118269874689532, + "learning_rate": 5.9654202213673794e-06, + "loss": 0.2874, + "step": 20671 + }, + { + "epoch": 0.4550297160466652, + "grad_norm": 0.6035509499890821, + "learning_rate": 5.965070470252177e-06, + "loss": 0.494, + "step": 20672 + }, + { + "epoch": 0.45505172793308385, + "grad_norm": 1.6501505180196332, + "learning_rate": 5.964720714232058e-06, + "loss": 0.317, + "step": 20673 + }, + { + "epoch": 0.45507373981950255, + "grad_norm": 1.7928157327612322, + "learning_rate": 5.964370953308796e-06, + "loss": 0.2603, + "step": 20674 + }, + { + "epoch": 0.4550957517059212, + "grad_norm": 5.15908478837917, + "learning_rate": 5.9640211874841715e-06, + "loss": 0.3493, + "step": 20675 + }, + { + "epoch": 0.45511776359233985, + "grad_norm": 2.8743917687528553, + "learning_rate": 5.963671416759962e-06, + "loss": 0.3632, + "step": 20676 + }, + { + "epoch": 0.45513977547875856, + "grad_norm": 1.562150995848832, + "learning_rate": 5.963321641137942e-06, + "loss": 0.2626, + "step": 20677 + }, + { + "epoch": 0.4551617873651772, + "grad_norm": 1.0925676092569716, + "learning_rate": 5.962971860619894e-06, + "loss": 0.3084, + "step": 20678 + }, + { + "epoch": 0.45518379925159586, + "grad_norm": 1.4394948735442283, + "learning_rate": 5.962622075207593e-06, + "loss": 0.3787, + "step": 20679 + }, + { + "epoch": 0.4552058111380145, + "grad_norm": 3.2730259039210137, + "learning_rate": 5.962272284902818e-06, + "loss": 0.3724, + "step": 20680 + }, + { + "epoch": 0.4552278230244332, + "grad_norm": 1.5123762759933903, + "learning_rate": 5.9619224897073456e-06, + "loss": 0.2642, + "step": 20681 + }, + { + "epoch": 0.45524983491085186, + "grad_norm": 1.3949641532896848, + "learning_rate": 5.961572689622953e-06, + "loss": 0.3139, + "step": 20682 + }, + { + "epoch": 0.4552718467972705, + "grad_norm": 2.442185585637363, + "learning_rate": 5.961222884651422e-06, + "loss": 0.3609, + "step": 20683 + }, + { + "epoch": 0.4552938586836892, + "grad_norm": 1.4671017626202532, + "learning_rate": 5.960873074794526e-06, + "loss": 0.2635, + "step": 20684 + }, + { + "epoch": 0.45531587057010786, + "grad_norm": 1.3274185420060143, + "learning_rate": 5.960523260054043e-06, + "loss": 0.3308, + "step": 20685 + }, + { + "epoch": 0.4553378824565265, + "grad_norm": 2.094036797999601, + "learning_rate": 5.960173440431753e-06, + "loss": 0.3356, + "step": 20686 + }, + { + "epoch": 0.4553598943429452, + "grad_norm": 1.4506632097972305, + "learning_rate": 5.959823615929433e-06, + "loss": 0.4218, + "step": 20687 + }, + { + "epoch": 0.45538190622936386, + "grad_norm": 1.6638661658797826, + "learning_rate": 5.959473786548864e-06, + "loss": 0.3429, + "step": 20688 + }, + { + "epoch": 0.4554039181157825, + "grad_norm": 1.418416892672428, + "learning_rate": 5.959123952291819e-06, + "loss": 0.3807, + "step": 20689 + }, + { + "epoch": 0.45542593000220116, + "grad_norm": 0.6712371765419632, + "learning_rate": 5.958774113160077e-06, + "loss": 0.4961, + "step": 20690 + }, + { + "epoch": 0.45544794188861987, + "grad_norm": 1.757117365370137, + "learning_rate": 5.958424269155421e-06, + "loss": 0.3395, + "step": 20691 + }, + { + "epoch": 0.4554699537750385, + "grad_norm": 1.4407874662138875, + "learning_rate": 5.958074420279622e-06, + "loss": 0.312, + "step": 20692 + }, + { + "epoch": 0.45549196566145717, + "grad_norm": 1.3242222583413632, + "learning_rate": 5.957724566534464e-06, + "loss": 0.4082, + "step": 20693 + }, + { + "epoch": 0.45551397754787587, + "grad_norm": 1.6478999173423146, + "learning_rate": 5.9573747079217215e-06, + "loss": 0.3593, + "step": 20694 + }, + { + "epoch": 0.4555359894342945, + "grad_norm": 1.6122482618354905, + "learning_rate": 5.9570248444431725e-06, + "loss": 0.4097, + "step": 20695 + }, + { + "epoch": 0.45555800132071317, + "grad_norm": 1.8780479317222962, + "learning_rate": 5.956674976100598e-06, + "loss": 0.3091, + "step": 20696 + }, + { + "epoch": 0.4555800132071319, + "grad_norm": 1.4467703091012598, + "learning_rate": 5.9563251028957744e-06, + "loss": 0.3572, + "step": 20697 + }, + { + "epoch": 0.4556020250935505, + "grad_norm": 1.438039401197217, + "learning_rate": 5.95597522483048e-06, + "loss": 0.2684, + "step": 20698 + }, + { + "epoch": 0.45562403697996917, + "grad_norm": 0.6043137278098663, + "learning_rate": 5.9556253419064935e-06, + "loss": 0.5188, + "step": 20699 + }, + { + "epoch": 0.4556460488663879, + "grad_norm": 1.591593182255556, + "learning_rate": 5.955275454125591e-06, + "loss": 0.3435, + "step": 20700 + }, + { + "epoch": 0.4556680607528065, + "grad_norm": 2.4491322294019353, + "learning_rate": 5.954925561489554e-06, + "loss": 0.3972, + "step": 20701 + }, + { + "epoch": 0.4556900726392252, + "grad_norm": 0.5834463272372751, + "learning_rate": 5.9545756640001605e-06, + "loss": 0.4544, + "step": 20702 + }, + { + "epoch": 0.4557120845256438, + "grad_norm": 1.448625000642085, + "learning_rate": 5.954225761659186e-06, + "loss": 0.3641, + "step": 20703 + }, + { + "epoch": 0.45573409641206253, + "grad_norm": 1.283139121848818, + "learning_rate": 5.9538758544684115e-06, + "loss": 0.2843, + "step": 20704 + }, + { + "epoch": 0.4557561082984812, + "grad_norm": 0.595690379350157, + "learning_rate": 5.953525942429614e-06, + "loss": 0.4794, + "step": 20705 + }, + { + "epoch": 0.4557781201848998, + "grad_norm": 1.350039478658091, + "learning_rate": 5.9531760255445725e-06, + "loss": 0.3425, + "step": 20706 + }, + { + "epoch": 0.45580013207131853, + "grad_norm": 1.9025201427484637, + "learning_rate": 5.952826103815066e-06, + "loss": 0.4299, + "step": 20707 + }, + { + "epoch": 0.4558221439577372, + "grad_norm": 1.2781993881843643, + "learning_rate": 5.952476177242869e-06, + "loss": 0.3741, + "step": 20708 + }, + { + "epoch": 0.45584415584415583, + "grad_norm": 2.028507166728423, + "learning_rate": 5.9521262458297666e-06, + "loss": 0.3274, + "step": 20709 + }, + { + "epoch": 0.45586616773057453, + "grad_norm": 1.2544656561414336, + "learning_rate": 5.951776309577534e-06, + "loss": 0.3367, + "step": 20710 + }, + { + "epoch": 0.4558881796169932, + "grad_norm": 0.6361456876588008, + "learning_rate": 5.951426368487948e-06, + "loss": 0.4856, + "step": 20711 + }, + { + "epoch": 0.45591019150341183, + "grad_norm": 2.041366851444173, + "learning_rate": 5.951076422562791e-06, + "loss": 0.2874, + "step": 20712 + }, + { + "epoch": 0.4559322033898305, + "grad_norm": 1.3376730056775787, + "learning_rate": 5.9507264718038364e-06, + "loss": 0.2217, + "step": 20713 + }, + { + "epoch": 0.4559542152762492, + "grad_norm": 2.0857885240455705, + "learning_rate": 5.950376516212868e-06, + "loss": 0.3797, + "step": 20714 + }, + { + "epoch": 0.45597622716266784, + "grad_norm": 1.6506715779810401, + "learning_rate": 5.950026555791661e-06, + "loss": 0.2906, + "step": 20715 + }, + { + "epoch": 0.4559982390490865, + "grad_norm": 1.1319180014787624, + "learning_rate": 5.949676590541998e-06, + "loss": 0.2882, + "step": 20716 + }, + { + "epoch": 0.4560202509355052, + "grad_norm": 1.4538544165357905, + "learning_rate": 5.949326620465652e-06, + "loss": 0.3418, + "step": 20717 + }, + { + "epoch": 0.45604226282192384, + "grad_norm": 2.0940370865730293, + "learning_rate": 5.948976645564405e-06, + "loss": 0.3546, + "step": 20718 + }, + { + "epoch": 0.4560642747083425, + "grad_norm": 1.350362061227478, + "learning_rate": 5.948626665840036e-06, + "loss": 0.3212, + "step": 20719 + }, + { + "epoch": 0.4560862865947612, + "grad_norm": 2.0724162396832524, + "learning_rate": 5.948276681294325e-06, + "loss": 0.4775, + "step": 20720 + }, + { + "epoch": 0.45610829848117984, + "grad_norm": 1.1231884214896228, + "learning_rate": 5.947926691929047e-06, + "loss": 0.3094, + "step": 20721 + }, + { + "epoch": 0.4561303103675985, + "grad_norm": 1.5759469181936623, + "learning_rate": 5.947576697745983e-06, + "loss": 0.2706, + "step": 20722 + }, + { + "epoch": 0.4561523222540172, + "grad_norm": 1.9726765163991453, + "learning_rate": 5.947226698746912e-06, + "loss": 0.3175, + "step": 20723 + }, + { + "epoch": 0.45617433414043584, + "grad_norm": 1.4165044249016157, + "learning_rate": 5.946876694933612e-06, + "loss": 0.2335, + "step": 20724 + }, + { + "epoch": 0.4561963460268545, + "grad_norm": 2.2301953765959754, + "learning_rate": 5.946526686307863e-06, + "loss": 0.2443, + "step": 20725 + }, + { + "epoch": 0.45621835791327314, + "grad_norm": 1.4018256187978222, + "learning_rate": 5.946176672871442e-06, + "loss": 0.3018, + "step": 20726 + }, + { + "epoch": 0.45624036979969185, + "grad_norm": 1.251451852872578, + "learning_rate": 5.94582665462613e-06, + "loss": 0.3355, + "step": 20727 + }, + { + "epoch": 0.4562623816861105, + "grad_norm": 0.6261982103260268, + "learning_rate": 5.945476631573705e-06, + "loss": 0.4643, + "step": 20728 + }, + { + "epoch": 0.45628439357252915, + "grad_norm": 1.4333822965654974, + "learning_rate": 5.945126603715946e-06, + "loss": 0.3744, + "step": 20729 + }, + { + "epoch": 0.45630640545894785, + "grad_norm": 1.3445449480715874, + "learning_rate": 5.944776571054632e-06, + "loss": 0.3813, + "step": 20730 + }, + { + "epoch": 0.4563284173453665, + "grad_norm": 2.504274415637146, + "learning_rate": 5.944426533591542e-06, + "loss": 0.389, + "step": 20731 + }, + { + "epoch": 0.45635042923178515, + "grad_norm": 4.785856451791534, + "learning_rate": 5.944076491328456e-06, + "loss": 0.3081, + "step": 20732 + }, + { + "epoch": 0.45637244111820385, + "grad_norm": 1.3351110144593044, + "learning_rate": 5.94372644426715e-06, + "loss": 0.3323, + "step": 20733 + }, + { + "epoch": 0.4563944530046225, + "grad_norm": 1.8146532439607537, + "learning_rate": 5.943376392409408e-06, + "loss": 0.2958, + "step": 20734 + }, + { + "epoch": 0.45641646489104115, + "grad_norm": 1.1925520590641054, + "learning_rate": 5.943026335757005e-06, + "loss": 0.3288, + "step": 20735 + }, + { + "epoch": 0.4564384767774598, + "grad_norm": 1.476270404427071, + "learning_rate": 5.942676274311721e-06, + "loss": 0.2801, + "step": 20736 + }, + { + "epoch": 0.4564604886638785, + "grad_norm": 1.61142290989003, + "learning_rate": 5.942326208075336e-06, + "loss": 0.3088, + "step": 20737 + }, + { + "epoch": 0.45648250055029715, + "grad_norm": 0.6580008508837701, + "learning_rate": 5.941976137049631e-06, + "loss": 0.485, + "step": 20738 + }, + { + "epoch": 0.4565045124367158, + "grad_norm": 1.172483778798489, + "learning_rate": 5.941626061236381e-06, + "loss": 0.2253, + "step": 20739 + }, + { + "epoch": 0.4565265243231345, + "grad_norm": 1.7923582114603689, + "learning_rate": 5.941275980637368e-06, + "loss": 0.317, + "step": 20740 + }, + { + "epoch": 0.45654853620955316, + "grad_norm": 1.6135421248924615, + "learning_rate": 5.940925895254371e-06, + "loss": 0.2907, + "step": 20741 + }, + { + "epoch": 0.4565705480959718, + "grad_norm": 1.2654211309379577, + "learning_rate": 5.940575805089167e-06, + "loss": 0.2318, + "step": 20742 + }, + { + "epoch": 0.4565925599823905, + "grad_norm": 1.7612676109026348, + "learning_rate": 5.9402257101435405e-06, + "loss": 0.3246, + "step": 20743 + }, + { + "epoch": 0.45661457186880916, + "grad_norm": 1.3405905347993232, + "learning_rate": 5.939875610419264e-06, + "loss": 0.3265, + "step": 20744 + }, + { + "epoch": 0.4566365837552278, + "grad_norm": 1.7795022852707638, + "learning_rate": 5.939525505918124e-06, + "loss": 0.3425, + "step": 20745 + }, + { + "epoch": 0.4566585956416465, + "grad_norm": 1.4826696330477596, + "learning_rate": 5.939175396641894e-06, + "loss": 0.2868, + "step": 20746 + }, + { + "epoch": 0.45668060752806516, + "grad_norm": 1.4734669834551404, + "learning_rate": 5.938825282592356e-06, + "loss": 0.3618, + "step": 20747 + }, + { + "epoch": 0.4567026194144838, + "grad_norm": 1.3378907003948675, + "learning_rate": 5.93847516377129e-06, + "loss": 0.2588, + "step": 20748 + }, + { + "epoch": 0.45672463130090246, + "grad_norm": 0.6347926538348272, + "learning_rate": 5.938125040180473e-06, + "loss": 0.4737, + "step": 20749 + }, + { + "epoch": 0.45674664318732117, + "grad_norm": 1.454066294706678, + "learning_rate": 5.937774911821689e-06, + "loss": 0.3335, + "step": 20750 + }, + { + "epoch": 0.4567686550737398, + "grad_norm": 0.6052062128003586, + "learning_rate": 5.937424778696712e-06, + "loss": 0.4633, + "step": 20751 + }, + { + "epoch": 0.45679066696015846, + "grad_norm": 2.1067035751716494, + "learning_rate": 5.9370746408073255e-06, + "loss": 0.2673, + "step": 20752 + }, + { + "epoch": 0.45681267884657717, + "grad_norm": 1.240035364900612, + "learning_rate": 5.9367244981553086e-06, + "loss": 0.3193, + "step": 20753 + }, + { + "epoch": 0.4568346907329958, + "grad_norm": 1.7525011138057802, + "learning_rate": 5.936374350742438e-06, + "loss": 0.2941, + "step": 20754 + }, + { + "epoch": 0.45685670261941447, + "grad_norm": 2.360829563728583, + "learning_rate": 5.936024198570496e-06, + "loss": 0.3518, + "step": 20755 + }, + { + "epoch": 0.45687871450583317, + "grad_norm": 1.4646876499890056, + "learning_rate": 5.935674041641263e-06, + "loss": 0.3537, + "step": 20756 + }, + { + "epoch": 0.4569007263922518, + "grad_norm": 1.4374884278869293, + "learning_rate": 5.935323879956514e-06, + "loss": 0.3229, + "step": 20757 + }, + { + "epoch": 0.45692273827867047, + "grad_norm": 1.2432791342934872, + "learning_rate": 5.934973713518034e-06, + "loss": 0.316, + "step": 20758 + }, + { + "epoch": 0.4569447501650892, + "grad_norm": 1.85010647045555, + "learning_rate": 5.934623542327601e-06, + "loss": 0.3471, + "step": 20759 + }, + { + "epoch": 0.4569667620515078, + "grad_norm": 1.4645948745235593, + "learning_rate": 5.934273366386993e-06, + "loss": 0.4298, + "step": 20760 + }, + { + "epoch": 0.4569887739379265, + "grad_norm": 3.197603809104609, + "learning_rate": 5.933923185697993e-06, + "loss": 0.3499, + "step": 20761 + }, + { + "epoch": 0.4570107858243451, + "grad_norm": 1.9137465812713146, + "learning_rate": 5.933573000262376e-06, + "loss": 0.3941, + "step": 20762 + }, + { + "epoch": 0.4570327977107638, + "grad_norm": 1.2745812901054137, + "learning_rate": 5.9332228100819265e-06, + "loss": 0.2778, + "step": 20763 + }, + { + "epoch": 0.4570548095971825, + "grad_norm": 1.519742124047819, + "learning_rate": 5.932872615158423e-06, + "loss": 0.3754, + "step": 20764 + }, + { + "epoch": 0.4570768214836011, + "grad_norm": 1.4416427976810682, + "learning_rate": 5.932522415493643e-06, + "loss": 0.2749, + "step": 20765 + }, + { + "epoch": 0.45709883337001983, + "grad_norm": 1.4715018075741986, + "learning_rate": 5.932172211089369e-06, + "loss": 0.3156, + "step": 20766 + }, + { + "epoch": 0.4571208452564385, + "grad_norm": 0.6518232351318595, + "learning_rate": 5.9318220019473794e-06, + "loss": 0.4815, + "step": 20767 + }, + { + "epoch": 0.45714285714285713, + "grad_norm": 1.5097639253513322, + "learning_rate": 5.931471788069456e-06, + "loss": 0.3651, + "step": 20768 + }, + { + "epoch": 0.45716486902927583, + "grad_norm": 1.2976953636871014, + "learning_rate": 5.931121569457376e-06, + "loss": 0.2813, + "step": 20769 + }, + { + "epoch": 0.4571868809156945, + "grad_norm": 1.5026452046367857, + "learning_rate": 5.930771346112923e-06, + "loss": 0.406, + "step": 20770 + }, + { + "epoch": 0.45720889280211313, + "grad_norm": 1.1600483658956653, + "learning_rate": 5.930421118037875e-06, + "loss": 0.2562, + "step": 20771 + }, + { + "epoch": 0.4572309046885318, + "grad_norm": 1.9340098838631097, + "learning_rate": 5.930070885234009e-06, + "loss": 0.3626, + "step": 20772 + }, + { + "epoch": 0.4572529165749505, + "grad_norm": 2.9829861768207393, + "learning_rate": 5.9297206477031095e-06, + "loss": 0.2986, + "step": 20773 + }, + { + "epoch": 0.45727492846136913, + "grad_norm": 1.4452584347001742, + "learning_rate": 5.929370405446957e-06, + "loss": 0.3182, + "step": 20774 + }, + { + "epoch": 0.4572969403477878, + "grad_norm": 1.1438825989289185, + "learning_rate": 5.929020158467327e-06, + "loss": 0.351, + "step": 20775 + }, + { + "epoch": 0.4573189522342065, + "grad_norm": 1.1785512717027316, + "learning_rate": 5.928669906766003e-06, + "loss": 0.299, + "step": 20776 + }, + { + "epoch": 0.45734096412062514, + "grad_norm": 1.5816657891582397, + "learning_rate": 5.928319650344764e-06, + "loss": 0.291, + "step": 20777 + }, + { + "epoch": 0.4573629760070438, + "grad_norm": 1.1657081897790686, + "learning_rate": 5.92796938920539e-06, + "loss": 0.2534, + "step": 20778 + }, + { + "epoch": 0.4573849878934625, + "grad_norm": 1.3050775048275682, + "learning_rate": 5.9276191233496635e-06, + "loss": 0.3542, + "step": 20779 + }, + { + "epoch": 0.45740699977988114, + "grad_norm": 2.745419340453565, + "learning_rate": 5.927268852779361e-06, + "loss": 0.3015, + "step": 20780 + }, + { + "epoch": 0.4574290116662998, + "grad_norm": 1.747397474205174, + "learning_rate": 5.9269185774962655e-06, + "loss": 0.2817, + "step": 20781 + }, + { + "epoch": 0.4574510235527185, + "grad_norm": 1.5701257136016462, + "learning_rate": 5.9265682975021565e-06, + "loss": 0.1766, + "step": 20782 + }, + { + "epoch": 0.45747303543913714, + "grad_norm": 1.5196133751344933, + "learning_rate": 5.9262180127988146e-06, + "loss": 0.2863, + "step": 20783 + }, + { + "epoch": 0.4574950473255558, + "grad_norm": 1.512475658144088, + "learning_rate": 5.925867723388019e-06, + "loss": 0.3328, + "step": 20784 + }, + { + "epoch": 0.45751705921197444, + "grad_norm": 1.4781407340500663, + "learning_rate": 5.925517429271549e-06, + "loss": 0.2792, + "step": 20785 + }, + { + "epoch": 0.45753907109839315, + "grad_norm": 7.791057454553366, + "learning_rate": 5.9251671304511895e-06, + "loss": 0.3828, + "step": 20786 + }, + { + "epoch": 0.4575610829848118, + "grad_norm": 1.2693659647729616, + "learning_rate": 5.924816826928717e-06, + "loss": 0.3656, + "step": 20787 + }, + { + "epoch": 0.45758309487123044, + "grad_norm": 0.6527728782145805, + "learning_rate": 5.924466518705913e-06, + "loss": 0.5059, + "step": 20788 + }, + { + "epoch": 0.45760510675764915, + "grad_norm": 0.6417800905369605, + "learning_rate": 5.924116205784557e-06, + "loss": 0.5062, + "step": 20789 + }, + { + "epoch": 0.4576271186440678, + "grad_norm": 1.5781890080883874, + "learning_rate": 5.923765888166429e-06, + "loss": 0.2688, + "step": 20790 + }, + { + "epoch": 0.45764913053048645, + "grad_norm": 1.3735129929536145, + "learning_rate": 5.923415565853313e-06, + "loss": 0.2823, + "step": 20791 + }, + { + "epoch": 0.45767114241690515, + "grad_norm": 1.1315142087686294, + "learning_rate": 5.923065238846988e-06, + "loss": 0.2594, + "step": 20792 + }, + { + "epoch": 0.4576931543033238, + "grad_norm": 1.5872780985436263, + "learning_rate": 5.922714907149233e-06, + "loss": 0.3204, + "step": 20793 + }, + { + "epoch": 0.45771516618974245, + "grad_norm": 1.6356777989358462, + "learning_rate": 5.9223645707618285e-06, + "loss": 0.3325, + "step": 20794 + }, + { + "epoch": 0.4577371780761611, + "grad_norm": 1.6836106970964053, + "learning_rate": 5.922014229686557e-06, + "loss": 0.3644, + "step": 20795 + }, + { + "epoch": 0.4577591899625798, + "grad_norm": 1.688555155024731, + "learning_rate": 5.921663883925196e-06, + "loss": 0.3857, + "step": 20796 + }, + { + "epoch": 0.45778120184899845, + "grad_norm": 0.6739060850944688, + "learning_rate": 5.92131353347953e-06, + "loss": 0.4767, + "step": 20797 + }, + { + "epoch": 0.4578032137354171, + "grad_norm": 0.650573435747209, + "learning_rate": 5.920963178351336e-06, + "loss": 0.4823, + "step": 20798 + }, + { + "epoch": 0.4578252256218358, + "grad_norm": 1.2843417745284629, + "learning_rate": 5.920612818542397e-06, + "loss": 0.3672, + "step": 20799 + }, + { + "epoch": 0.45784723750825446, + "grad_norm": 1.2558136075477575, + "learning_rate": 5.920262454054494e-06, + "loss": 0.2795, + "step": 20800 + }, + { + "epoch": 0.4578692493946731, + "grad_norm": 1.3781603410061838, + "learning_rate": 5.919912084889405e-06, + "loss": 0.2722, + "step": 20801 + }, + { + "epoch": 0.4578912612810918, + "grad_norm": 2.5366370830392264, + "learning_rate": 5.919561711048916e-06, + "loss": 0.3318, + "step": 20802 + }, + { + "epoch": 0.45791327316751046, + "grad_norm": 0.6221963683157707, + "learning_rate": 5.9192113325348e-06, + "loss": 0.494, + "step": 20803 + }, + { + "epoch": 0.4579352850539291, + "grad_norm": 1.8159404813001971, + "learning_rate": 5.9188609493488445e-06, + "loss": 0.3178, + "step": 20804 + }, + { + "epoch": 0.4579572969403478, + "grad_norm": 2.516959802452487, + "learning_rate": 5.918510561492827e-06, + "loss": 0.3379, + "step": 20805 + }, + { + "epoch": 0.45797930882676646, + "grad_norm": 1.1627950291832545, + "learning_rate": 5.9181601689685295e-06, + "loss": 0.2915, + "step": 20806 + }, + { + "epoch": 0.4580013207131851, + "grad_norm": 2.1983488068517207, + "learning_rate": 5.917809771777732e-06, + "loss": 0.3218, + "step": 20807 + }, + { + "epoch": 0.45802333259960376, + "grad_norm": 4.078717281931187, + "learning_rate": 5.917459369922216e-06, + "loss": 0.2443, + "step": 20808 + }, + { + "epoch": 0.45804534448602247, + "grad_norm": 0.679613744916689, + "learning_rate": 5.917108963403764e-06, + "loss": 0.5053, + "step": 20809 + }, + { + "epoch": 0.4580673563724411, + "grad_norm": 1.3668612705497132, + "learning_rate": 5.916758552224153e-06, + "loss": 0.3207, + "step": 20810 + }, + { + "epoch": 0.45808936825885976, + "grad_norm": 2.123353346718285, + "learning_rate": 5.9164081363851676e-06, + "loss": 0.3526, + "step": 20811 + }, + { + "epoch": 0.45811138014527847, + "grad_norm": 1.2106524862455166, + "learning_rate": 5.916057715888587e-06, + "loss": 0.2455, + "step": 20812 + }, + { + "epoch": 0.4581333920316971, + "grad_norm": 0.643849775710331, + "learning_rate": 5.915707290736193e-06, + "loss": 0.5302, + "step": 20813 + }, + { + "epoch": 0.45815540391811577, + "grad_norm": 1.5266807362318062, + "learning_rate": 5.915356860929764e-06, + "loss": 0.409, + "step": 20814 + }, + { + "epoch": 0.45817741580453447, + "grad_norm": 1.4247139689318733, + "learning_rate": 5.9150064264710845e-06, + "loss": 0.3747, + "step": 20815 + }, + { + "epoch": 0.4581994276909531, + "grad_norm": 0.6201760161289812, + "learning_rate": 5.914655987361934e-06, + "loss": 0.5005, + "step": 20816 + }, + { + "epoch": 0.45822143957737177, + "grad_norm": 1.2378158640192358, + "learning_rate": 5.914305543604094e-06, + "loss": 0.2741, + "step": 20817 + }, + { + "epoch": 0.4582434514637905, + "grad_norm": 1.462523795586683, + "learning_rate": 5.9139550951993466e-06, + "loss": 0.3671, + "step": 20818 + }, + { + "epoch": 0.4582654633502091, + "grad_norm": 1.6097855507265406, + "learning_rate": 5.913604642149468e-06, + "loss": 0.2573, + "step": 20819 + }, + { + "epoch": 0.4582874752366278, + "grad_norm": 1.6310614725975667, + "learning_rate": 5.913254184456247e-06, + "loss": 0.2981, + "step": 20820 + }, + { + "epoch": 0.4583094871230464, + "grad_norm": 1.5336085197982605, + "learning_rate": 5.912903722121461e-06, + "loss": 0.3449, + "step": 20821 + }, + { + "epoch": 0.4583314990094651, + "grad_norm": 1.3054497279580775, + "learning_rate": 5.9125532551468885e-06, + "loss": 0.3196, + "step": 20822 + }, + { + "epoch": 0.4583535108958838, + "grad_norm": 1.7890999648552894, + "learning_rate": 5.9122027835343155e-06, + "loss": 0.2505, + "step": 20823 + }, + { + "epoch": 0.4583755227823024, + "grad_norm": 1.4190153190103827, + "learning_rate": 5.91185230728552e-06, + "loss": 0.2544, + "step": 20824 + }, + { + "epoch": 0.45839753466872113, + "grad_norm": 0.6850614604614784, + "learning_rate": 5.911501826402285e-06, + "loss": 0.4862, + "step": 20825 + }, + { + "epoch": 0.4584195465551398, + "grad_norm": 1.3488233277425103, + "learning_rate": 5.91115134088639e-06, + "loss": 0.3249, + "step": 20826 + }, + { + "epoch": 0.4584415584415584, + "grad_norm": 0.666361733868587, + "learning_rate": 5.910800850739619e-06, + "loss": 0.5068, + "step": 20827 + }, + { + "epoch": 0.45846357032797713, + "grad_norm": 1.957463649229665, + "learning_rate": 5.910450355963751e-06, + "loss": 0.3242, + "step": 20828 + }, + { + "epoch": 0.4584855822143958, + "grad_norm": 1.6809670940035006, + "learning_rate": 5.910099856560568e-06, + "loss": 0.347, + "step": 20829 + }, + { + "epoch": 0.45850759410081443, + "grad_norm": 1.6498294084410574, + "learning_rate": 5.909749352531853e-06, + "loss": 0.3897, + "step": 20830 + }, + { + "epoch": 0.4585296059872331, + "grad_norm": 1.8021314438321077, + "learning_rate": 5.909398843879384e-06, + "loss": 0.2732, + "step": 20831 + }, + { + "epoch": 0.4585516178736518, + "grad_norm": 1.0756667945966714, + "learning_rate": 5.909048330604945e-06, + "loss": 0.2189, + "step": 20832 + }, + { + "epoch": 0.45857362976007043, + "grad_norm": 0.6527922146527422, + "learning_rate": 5.908697812710319e-06, + "loss": 0.4797, + "step": 20833 + }, + { + "epoch": 0.4585956416464891, + "grad_norm": 0.6418880205806812, + "learning_rate": 5.9083472901972825e-06, + "loss": 0.4706, + "step": 20834 + }, + { + "epoch": 0.4586176535329078, + "grad_norm": 2.225039342911597, + "learning_rate": 5.907996763067621e-06, + "loss": 0.3738, + "step": 20835 + }, + { + "epoch": 0.45863966541932644, + "grad_norm": 1.2960059055171205, + "learning_rate": 5.9076462313231165e-06, + "loss": 0.3756, + "step": 20836 + }, + { + "epoch": 0.4586616773057451, + "grad_norm": 1.7986175484372853, + "learning_rate": 5.907295694965547e-06, + "loss": 0.2853, + "step": 20837 + }, + { + "epoch": 0.4586836891921638, + "grad_norm": 1.6480759169382908, + "learning_rate": 5.9069451539966985e-06, + "loss": 0.339, + "step": 20838 + }, + { + "epoch": 0.45870570107858244, + "grad_norm": 0.6342194959524279, + "learning_rate": 5.9065946084183464e-06, + "loss": 0.4992, + "step": 20839 + }, + { + "epoch": 0.4587277129650011, + "grad_norm": 2.436365366640796, + "learning_rate": 5.906244058232279e-06, + "loss": 0.3522, + "step": 20840 + }, + { + "epoch": 0.4587497248514198, + "grad_norm": 2.1436689815026857, + "learning_rate": 5.9058935034402745e-06, + "loss": 0.2922, + "step": 20841 + }, + { + "epoch": 0.45877173673783844, + "grad_norm": 1.3864279538520856, + "learning_rate": 5.905542944044115e-06, + "loss": 0.2909, + "step": 20842 + }, + { + "epoch": 0.4587937486242571, + "grad_norm": 1.3678374790106504, + "learning_rate": 5.905192380045584e-06, + "loss": 0.3254, + "step": 20843 + }, + { + "epoch": 0.45881576051067574, + "grad_norm": 1.7169824301604864, + "learning_rate": 5.904841811446459e-06, + "loss": 0.311, + "step": 20844 + }, + { + "epoch": 0.45883777239709445, + "grad_norm": 0.6189306371164861, + "learning_rate": 5.904491238248525e-06, + "loss": 0.4713, + "step": 20845 + }, + { + "epoch": 0.4588597842835131, + "grad_norm": 1.5045584456185266, + "learning_rate": 5.904140660453564e-06, + "loss": 0.3716, + "step": 20846 + }, + { + "epoch": 0.45888179616993174, + "grad_norm": 1.303316598142898, + "learning_rate": 5.903790078063356e-06, + "loss": 0.2818, + "step": 20847 + }, + { + "epoch": 0.45890380805635045, + "grad_norm": 1.0224021734684305, + "learning_rate": 5.903439491079686e-06, + "loss": 0.2451, + "step": 20848 + }, + { + "epoch": 0.4589258199427691, + "grad_norm": 1.5611837810947282, + "learning_rate": 5.903088899504331e-06, + "loss": 0.3095, + "step": 20849 + }, + { + "epoch": 0.45894783182918775, + "grad_norm": 1.334980886725765, + "learning_rate": 5.902738303339075e-06, + "loss": 0.1983, + "step": 20850 + }, + { + "epoch": 0.45896984371560645, + "grad_norm": 1.5345639757822287, + "learning_rate": 5.902387702585701e-06, + "loss": 0.2852, + "step": 20851 + }, + { + "epoch": 0.4589918556020251, + "grad_norm": 3.9139778904919433, + "learning_rate": 5.902037097245992e-06, + "loss": 0.2848, + "step": 20852 + }, + { + "epoch": 0.45901386748844375, + "grad_norm": 1.611530632271289, + "learning_rate": 5.901686487321727e-06, + "loss": 0.294, + "step": 20853 + }, + { + "epoch": 0.4590358793748624, + "grad_norm": 0.6205376966702271, + "learning_rate": 5.901335872814688e-06, + "loss": 0.4937, + "step": 20854 + }, + { + "epoch": 0.4590578912612811, + "grad_norm": 1.317932116956278, + "learning_rate": 5.900985253726659e-06, + "loss": 0.252, + "step": 20855 + }, + { + "epoch": 0.45907990314769975, + "grad_norm": 1.9628385833391808, + "learning_rate": 5.900634630059423e-06, + "loss": 0.2511, + "step": 20856 + }, + { + "epoch": 0.4591019150341184, + "grad_norm": 0.6036135644162052, + "learning_rate": 5.900284001814757e-06, + "loss": 0.4777, + "step": 20857 + }, + { + "epoch": 0.4591239269205371, + "grad_norm": 1.9720981331420442, + "learning_rate": 5.899933368994447e-06, + "loss": 0.3333, + "step": 20858 + }, + { + "epoch": 0.45914593880695576, + "grad_norm": 1.329222418830116, + "learning_rate": 5.899582731600275e-06, + "loss": 0.372, + "step": 20859 + }, + { + "epoch": 0.4591679506933744, + "grad_norm": 1.9562841186189237, + "learning_rate": 5.899232089634022e-06, + "loss": 0.2927, + "step": 20860 + }, + { + "epoch": 0.4591899625797931, + "grad_norm": 1.211465057997507, + "learning_rate": 5.8988814430974716e-06, + "loss": 0.2432, + "step": 20861 + }, + { + "epoch": 0.45921197446621176, + "grad_norm": 1.4099393513782206, + "learning_rate": 5.898530791992403e-06, + "loss": 0.3162, + "step": 20862 + }, + { + "epoch": 0.4592339863526304, + "grad_norm": 1.376314428366996, + "learning_rate": 5.898180136320601e-06, + "loss": 0.3052, + "step": 20863 + }, + { + "epoch": 0.4592559982390491, + "grad_norm": 0.62482362617361, + "learning_rate": 5.897829476083846e-06, + "loss": 0.4938, + "step": 20864 + }, + { + "epoch": 0.45927801012546776, + "grad_norm": 1.6430141372082285, + "learning_rate": 5.897478811283922e-06, + "loss": 0.2943, + "step": 20865 + }, + { + "epoch": 0.4593000220118864, + "grad_norm": 1.4581902314863955, + "learning_rate": 5.897128141922611e-06, + "loss": 0.3909, + "step": 20866 + }, + { + "epoch": 0.45932203389830506, + "grad_norm": 1.4322894685729086, + "learning_rate": 5.896777468001694e-06, + "loss": 0.389, + "step": 20867 + }, + { + "epoch": 0.45934404578472376, + "grad_norm": 1.2825570719318038, + "learning_rate": 5.8964267895229536e-06, + "loss": 0.2867, + "step": 20868 + }, + { + "epoch": 0.4593660576711424, + "grad_norm": 1.2990072772516197, + "learning_rate": 5.896076106488172e-06, + "loss": 0.2945, + "step": 20869 + }, + { + "epoch": 0.45938806955756106, + "grad_norm": 0.6091383252418308, + "learning_rate": 5.895725418899133e-06, + "loss": 0.4622, + "step": 20870 + }, + { + "epoch": 0.45941008144397977, + "grad_norm": 2.7188547481310823, + "learning_rate": 5.895374726757618e-06, + "loss": 0.3569, + "step": 20871 + }, + { + "epoch": 0.4594320933303984, + "grad_norm": 1.3468560420171727, + "learning_rate": 5.895024030065409e-06, + "loss": 0.3378, + "step": 20872 + }, + { + "epoch": 0.45945410521681707, + "grad_norm": 1.3352436907243352, + "learning_rate": 5.894673328824287e-06, + "loss": 0.2626, + "step": 20873 + }, + { + "epoch": 0.45947611710323577, + "grad_norm": 0.6197202233018079, + "learning_rate": 5.89432262303604e-06, + "loss": 0.4786, + "step": 20874 + }, + { + "epoch": 0.4594981289896544, + "grad_norm": 1.4789867150742897, + "learning_rate": 5.893971912702443e-06, + "loss": 0.2773, + "step": 20875 + }, + { + "epoch": 0.45952014087607307, + "grad_norm": 2.7529624803006025, + "learning_rate": 5.893621197825283e-06, + "loss": 0.3303, + "step": 20876 + }, + { + "epoch": 0.4595421527624917, + "grad_norm": 1.418125702652837, + "learning_rate": 5.893270478406342e-06, + "loss": 0.3136, + "step": 20877 + }, + { + "epoch": 0.4595641646489104, + "grad_norm": 1.2078722137348934, + "learning_rate": 5.892919754447401e-06, + "loss": 0.3474, + "step": 20878 + }, + { + "epoch": 0.45958617653532907, + "grad_norm": 2.9569201202842117, + "learning_rate": 5.8925690259502455e-06, + "loss": 0.4642, + "step": 20879 + }, + { + "epoch": 0.4596081884217477, + "grad_norm": 1.9291903886368722, + "learning_rate": 5.892218292916655e-06, + "loss": 0.3803, + "step": 20880 + }, + { + "epoch": 0.4596302003081664, + "grad_norm": 0.610589186227768, + "learning_rate": 5.891867555348413e-06, + "loss": 0.4751, + "step": 20881 + }, + { + "epoch": 0.4596522121945851, + "grad_norm": 2.1800637872583946, + "learning_rate": 5.8915168132473024e-06, + "loss": 0.328, + "step": 20882 + }, + { + "epoch": 0.4596742240810037, + "grad_norm": 1.4079061359499798, + "learning_rate": 5.891166066615106e-06, + "loss": 0.3532, + "step": 20883 + }, + { + "epoch": 0.45969623596742243, + "grad_norm": 1.1798821857293667, + "learning_rate": 5.890815315453607e-06, + "loss": 0.3692, + "step": 20884 + }, + { + "epoch": 0.4597182478538411, + "grad_norm": 63.015532161767965, + "learning_rate": 5.890464559764586e-06, + "loss": 0.394, + "step": 20885 + }, + { + "epoch": 0.4597402597402597, + "grad_norm": 1.2919312553617865, + "learning_rate": 5.8901137995498255e-06, + "loss": 0.3431, + "step": 20886 + }, + { + "epoch": 0.45976227162667843, + "grad_norm": 1.2044303698593468, + "learning_rate": 5.889763034811112e-06, + "loss": 0.2377, + "step": 20887 + }, + { + "epoch": 0.4597842835130971, + "grad_norm": 1.8149799209246922, + "learning_rate": 5.889412265550225e-06, + "loss": 0.3666, + "step": 20888 + }, + { + "epoch": 0.45980629539951573, + "grad_norm": 1.1093483784377756, + "learning_rate": 5.8890614917689495e-06, + "loss": 0.2412, + "step": 20889 + }, + { + "epoch": 0.4598283072859344, + "grad_norm": 0.6322095469832477, + "learning_rate": 5.888710713469066e-06, + "loss": 0.5017, + "step": 20890 + }, + { + "epoch": 0.4598503191723531, + "grad_norm": 1.358486486994077, + "learning_rate": 5.888359930652356e-06, + "loss": 0.3399, + "step": 20891 + }, + { + "epoch": 0.45987233105877173, + "grad_norm": 1.3210885822233804, + "learning_rate": 5.888009143320608e-06, + "loss": 0.224, + "step": 20892 + }, + { + "epoch": 0.4598943429451904, + "grad_norm": 1.692187511222834, + "learning_rate": 5.8876583514756e-06, + "loss": 0.2952, + "step": 20893 + }, + { + "epoch": 0.4599163548316091, + "grad_norm": 1.50289154368091, + "learning_rate": 5.887307555119117e-06, + "loss": 0.3569, + "step": 20894 + }, + { + "epoch": 0.45993836671802774, + "grad_norm": 1.1681418240024066, + "learning_rate": 5.886956754252941e-06, + "loss": 0.2182, + "step": 20895 + }, + { + "epoch": 0.4599603786044464, + "grad_norm": 1.5213782980725303, + "learning_rate": 5.886605948878854e-06, + "loss": 0.3516, + "step": 20896 + }, + { + "epoch": 0.4599823904908651, + "grad_norm": 1.6047887910843233, + "learning_rate": 5.886255138998642e-06, + "loss": 0.3165, + "step": 20897 + }, + { + "epoch": 0.46000440237728374, + "grad_norm": 0.6199886261894043, + "learning_rate": 5.885904324614084e-06, + "loss": 0.4988, + "step": 20898 + }, + { + "epoch": 0.4600264142637024, + "grad_norm": 1.8477251665496943, + "learning_rate": 5.885553505726967e-06, + "loss": 0.2647, + "step": 20899 + }, + { + "epoch": 0.4600484261501211, + "grad_norm": 1.543841204009622, + "learning_rate": 5.8852026823390726e-06, + "loss": 0.3261, + "step": 20900 + }, + { + "epoch": 0.46007043803653974, + "grad_norm": 2.466632321399542, + "learning_rate": 5.88485185445218e-06, + "loss": 0.3812, + "step": 20901 + }, + { + "epoch": 0.4600924499229584, + "grad_norm": 1.574262774335933, + "learning_rate": 5.88450102206808e-06, + "loss": 0.3292, + "step": 20902 + }, + { + "epoch": 0.46011446180937704, + "grad_norm": 1.4157513712279175, + "learning_rate": 5.88415018518855e-06, + "loss": 0.2655, + "step": 20903 + }, + { + "epoch": 0.46013647369579574, + "grad_norm": 1.4349723965585637, + "learning_rate": 5.883799343815372e-06, + "loss": 0.2968, + "step": 20904 + }, + { + "epoch": 0.4601584855822144, + "grad_norm": 1.2617362499869673, + "learning_rate": 5.883448497950334e-06, + "loss": 0.3423, + "step": 20905 + }, + { + "epoch": 0.46018049746863304, + "grad_norm": 0.5928482461957714, + "learning_rate": 5.883097647595215e-06, + "loss": 0.4687, + "step": 20906 + }, + { + "epoch": 0.46020250935505175, + "grad_norm": 1.4488741241057819, + "learning_rate": 5.882746792751801e-06, + "loss": 0.3363, + "step": 20907 + }, + { + "epoch": 0.4602245212414704, + "grad_norm": 1.5472317091947154, + "learning_rate": 5.882395933421874e-06, + "loss": 0.3091, + "step": 20908 + }, + { + "epoch": 0.46024653312788905, + "grad_norm": 0.641065749581014, + "learning_rate": 5.882045069607216e-06, + "loss": 0.518, + "step": 20909 + }, + { + "epoch": 0.46026854501430775, + "grad_norm": 0.6516753633002671, + "learning_rate": 5.881694201309614e-06, + "loss": 0.4942, + "step": 20910 + }, + { + "epoch": 0.4602905569007264, + "grad_norm": 1.5209487463909717, + "learning_rate": 5.881343328530848e-06, + "loss": 0.2981, + "step": 20911 + }, + { + "epoch": 0.46031256878714505, + "grad_norm": 1.3847833920650114, + "learning_rate": 5.880992451272701e-06, + "loss": 0.3493, + "step": 20912 + }, + { + "epoch": 0.4603345806735637, + "grad_norm": 1.4253306587094399, + "learning_rate": 5.8806415695369575e-06, + "loss": 0.3362, + "step": 20913 + }, + { + "epoch": 0.4603565925599824, + "grad_norm": 1.6057693756744544, + "learning_rate": 5.880290683325402e-06, + "loss": 0.3928, + "step": 20914 + }, + { + "epoch": 0.46037860444640105, + "grad_norm": 1.6861861872854675, + "learning_rate": 5.879939792639816e-06, + "loss": 0.3194, + "step": 20915 + }, + { + "epoch": 0.4604006163328197, + "grad_norm": 0.621349426396647, + "learning_rate": 5.879588897481982e-06, + "loss": 0.4672, + "step": 20916 + }, + { + "epoch": 0.4604226282192384, + "grad_norm": 1.3221277531431086, + "learning_rate": 5.879237997853687e-06, + "loss": 0.3194, + "step": 20917 + }, + { + "epoch": 0.46044464010565705, + "grad_norm": 1.6622084665717605, + "learning_rate": 5.878887093756712e-06, + "loss": 0.3721, + "step": 20918 + }, + { + "epoch": 0.4604666519920757, + "grad_norm": 1.341034234925228, + "learning_rate": 5.878536185192841e-06, + "loss": 0.3283, + "step": 20919 + }, + { + "epoch": 0.4604886638784944, + "grad_norm": 1.247956950463767, + "learning_rate": 5.878185272163857e-06, + "loss": 0.2774, + "step": 20920 + }, + { + "epoch": 0.46051067576491306, + "grad_norm": 1.5959095450975327, + "learning_rate": 5.8778343546715436e-06, + "loss": 0.282, + "step": 20921 + }, + { + "epoch": 0.4605326876513317, + "grad_norm": 3.296695454655425, + "learning_rate": 5.8774834327176835e-06, + "loss": 0.3919, + "step": 20922 + }, + { + "epoch": 0.4605546995377504, + "grad_norm": 1.3899229629842489, + "learning_rate": 5.877132506304062e-06, + "loss": 0.2447, + "step": 20923 + }, + { + "epoch": 0.46057671142416906, + "grad_norm": 1.9192982064572799, + "learning_rate": 5.876781575432462e-06, + "loss": 0.3421, + "step": 20924 + }, + { + "epoch": 0.4605987233105877, + "grad_norm": 1.3269466286546094, + "learning_rate": 5.876430640104667e-06, + "loss": 0.298, + "step": 20925 + }, + { + "epoch": 0.46062073519700636, + "grad_norm": 2.509119457162833, + "learning_rate": 5.876079700322461e-06, + "loss": 0.3193, + "step": 20926 + }, + { + "epoch": 0.46064274708342506, + "grad_norm": 1.2058544861271796, + "learning_rate": 5.8757287560876265e-06, + "loss": 0.3115, + "step": 20927 + }, + { + "epoch": 0.4606647589698437, + "grad_norm": 1.2982017276971412, + "learning_rate": 5.875377807401949e-06, + "loss": 0.3338, + "step": 20928 + }, + { + "epoch": 0.46068677085626236, + "grad_norm": 1.444704593285612, + "learning_rate": 5.875026854267209e-06, + "loss": 0.3, + "step": 20929 + }, + { + "epoch": 0.46070878274268107, + "grad_norm": 2.290229548821067, + "learning_rate": 5.874675896685194e-06, + "loss": 0.3479, + "step": 20930 + }, + { + "epoch": 0.4607307946290997, + "grad_norm": 0.5998930229113725, + "learning_rate": 5.874324934657686e-06, + "loss": 0.4746, + "step": 20931 + }, + { + "epoch": 0.46075280651551836, + "grad_norm": 1.3625746504513154, + "learning_rate": 5.873973968186467e-06, + "loss": 0.298, + "step": 20932 + }, + { + "epoch": 0.46077481840193707, + "grad_norm": 1.433468628335486, + "learning_rate": 5.873622997273325e-06, + "loss": 0.3441, + "step": 20933 + }, + { + "epoch": 0.4607968302883557, + "grad_norm": 1.1641117752904049, + "learning_rate": 5.87327202192004e-06, + "loss": 0.2472, + "step": 20934 + }, + { + "epoch": 0.46081884217477437, + "grad_norm": 4.571548343776017, + "learning_rate": 5.872921042128397e-06, + "loss": 0.4562, + "step": 20935 + }, + { + "epoch": 0.460840854061193, + "grad_norm": 0.6367742255494289, + "learning_rate": 5.87257005790018e-06, + "loss": 0.4978, + "step": 20936 + }, + { + "epoch": 0.4608628659476117, + "grad_norm": 1.7404524309987737, + "learning_rate": 5.872219069237172e-06, + "loss": 0.3491, + "step": 20937 + }, + { + "epoch": 0.46088487783403037, + "grad_norm": 1.395299694045473, + "learning_rate": 5.871868076141159e-06, + "loss": 0.2316, + "step": 20938 + }, + { + "epoch": 0.460906889720449, + "grad_norm": 1.2435050741955787, + "learning_rate": 5.8715170786139235e-06, + "loss": 0.2332, + "step": 20939 + }, + { + "epoch": 0.4609289016068677, + "grad_norm": 2.2343479929358336, + "learning_rate": 5.8711660766572476e-06, + "loss": 0.2666, + "step": 20940 + }, + { + "epoch": 0.4609509134932864, + "grad_norm": 0.6444086549551559, + "learning_rate": 5.8708150702729195e-06, + "loss": 0.5224, + "step": 20941 + }, + { + "epoch": 0.460972925379705, + "grad_norm": 1.2380744927591096, + "learning_rate": 5.870464059462719e-06, + "loss": 0.379, + "step": 20942 + }, + { + "epoch": 0.4609949372661237, + "grad_norm": 1.4415265616990462, + "learning_rate": 5.870113044228434e-06, + "loss": 0.2938, + "step": 20943 + }, + { + "epoch": 0.4610169491525424, + "grad_norm": 1.348250279308345, + "learning_rate": 5.869762024571846e-06, + "loss": 0.255, + "step": 20944 + }, + { + "epoch": 0.461038961038961, + "grad_norm": 1.6464963989396255, + "learning_rate": 5.869411000494737e-06, + "loss": 0.3199, + "step": 20945 + }, + { + "epoch": 0.46106097292537973, + "grad_norm": 1.7480220498550698, + "learning_rate": 5.869059971998895e-06, + "loss": 0.3258, + "step": 20946 + }, + { + "epoch": 0.4610829848117984, + "grad_norm": 1.369627665965918, + "learning_rate": 5.868708939086103e-06, + "loss": 0.397, + "step": 20947 + }, + { + "epoch": 0.46110499669821703, + "grad_norm": 1.3073328106844266, + "learning_rate": 5.8683579017581445e-06, + "loss": 0.3953, + "step": 20948 + }, + { + "epoch": 0.4611270085846357, + "grad_norm": 1.4370810632848623, + "learning_rate": 5.868006860016805e-06, + "loss": 0.252, + "step": 20949 + }, + { + "epoch": 0.4611490204710544, + "grad_norm": 1.7623241529492015, + "learning_rate": 5.8676558138638655e-06, + "loss": 0.3454, + "step": 20950 + }, + { + "epoch": 0.46117103235747303, + "grad_norm": 4.269651016842863, + "learning_rate": 5.867304763301113e-06, + "loss": 0.2954, + "step": 20951 + }, + { + "epoch": 0.4611930442438917, + "grad_norm": 1.3178180191759288, + "learning_rate": 5.866953708330332e-06, + "loss": 0.2973, + "step": 20952 + }, + { + "epoch": 0.4612150561303104, + "grad_norm": 2.87062970196766, + "learning_rate": 5.8666026489533036e-06, + "loss": 0.3656, + "step": 20953 + }, + { + "epoch": 0.46123706801672903, + "grad_norm": 1.428932674495159, + "learning_rate": 5.866251585171815e-06, + "loss": 0.3132, + "step": 20954 + }, + { + "epoch": 0.4612590799031477, + "grad_norm": 1.4843559875989283, + "learning_rate": 5.865900516987648e-06, + "loss": 0.3502, + "step": 20955 + }, + { + "epoch": 0.4612810917895664, + "grad_norm": 1.4725320316643602, + "learning_rate": 5.865549444402592e-06, + "loss": 0.3834, + "step": 20956 + }, + { + "epoch": 0.46130310367598504, + "grad_norm": 1.8998119930289126, + "learning_rate": 5.865198367418426e-06, + "loss": 0.338, + "step": 20957 + }, + { + "epoch": 0.4613251155624037, + "grad_norm": 1.0735638532669431, + "learning_rate": 5.864847286036934e-06, + "loss": 0.2571, + "step": 20958 + }, + { + "epoch": 0.4613471274488224, + "grad_norm": 1.4909070599947143, + "learning_rate": 5.864496200259904e-06, + "loss": 0.2867, + "step": 20959 + }, + { + "epoch": 0.46136913933524104, + "grad_norm": 0.6475596583786974, + "learning_rate": 5.864145110089119e-06, + "loss": 0.5146, + "step": 20960 + }, + { + "epoch": 0.4613911512216597, + "grad_norm": 1.3330035549258448, + "learning_rate": 5.863794015526363e-06, + "loss": 0.3164, + "step": 20961 + }, + { + "epoch": 0.46141316310807834, + "grad_norm": 1.6639319244249604, + "learning_rate": 5.86344291657342e-06, + "loss": 0.2648, + "step": 20962 + }, + { + "epoch": 0.46143517499449704, + "grad_norm": 0.6291728505703756, + "learning_rate": 5.863091813232075e-06, + "loss": 0.5041, + "step": 20963 + }, + { + "epoch": 0.4614571868809157, + "grad_norm": 1.4141116520823054, + "learning_rate": 5.862740705504112e-06, + "loss": 0.2548, + "step": 20964 + }, + { + "epoch": 0.46147919876733434, + "grad_norm": 1.1231745959848645, + "learning_rate": 5.862389593391316e-06, + "loss": 0.2262, + "step": 20965 + }, + { + "epoch": 0.46150121065375305, + "grad_norm": 2.1890499599317406, + "learning_rate": 5.862038476895473e-06, + "loss": 0.3979, + "step": 20966 + }, + { + "epoch": 0.4615232225401717, + "grad_norm": 1.0775055979327024, + "learning_rate": 5.861687356018365e-06, + "loss": 0.2489, + "step": 20967 + }, + { + "epoch": 0.46154523442659035, + "grad_norm": 0.9773244407082078, + "learning_rate": 5.861336230761777e-06, + "loss": 0.2451, + "step": 20968 + }, + { + "epoch": 0.46156724631300905, + "grad_norm": 1.0482638973464171, + "learning_rate": 5.860985101127496e-06, + "loss": 0.2509, + "step": 20969 + }, + { + "epoch": 0.4615892581994277, + "grad_norm": 1.2230285630016138, + "learning_rate": 5.860633967117304e-06, + "loss": 0.2751, + "step": 20970 + }, + { + "epoch": 0.46161127008584635, + "grad_norm": 1.4669569408291168, + "learning_rate": 5.860282828732985e-06, + "loss": 0.294, + "step": 20971 + }, + { + "epoch": 0.461633281972265, + "grad_norm": 1.9229970159690173, + "learning_rate": 5.859931685976326e-06, + "loss": 0.3662, + "step": 20972 + }, + { + "epoch": 0.4616552938586837, + "grad_norm": 1.6326168184168706, + "learning_rate": 5.859580538849109e-06, + "loss": 0.2931, + "step": 20973 + }, + { + "epoch": 0.46167730574510235, + "grad_norm": 1.540394293139832, + "learning_rate": 5.859229387353123e-06, + "loss": 0.3511, + "step": 20974 + }, + { + "epoch": 0.461699317631521, + "grad_norm": 1.2825585150106187, + "learning_rate": 5.858878231490149e-06, + "loss": 0.2913, + "step": 20975 + }, + { + "epoch": 0.4617213295179397, + "grad_norm": 0.6618237429210136, + "learning_rate": 5.8585270712619716e-06, + "loss": 0.4895, + "step": 20976 + }, + { + "epoch": 0.46174334140435835, + "grad_norm": 1.7719982351802848, + "learning_rate": 5.858175906670377e-06, + "loss": 0.305, + "step": 20977 + }, + { + "epoch": 0.461765353290777, + "grad_norm": 1.8742725795780912, + "learning_rate": 5.85782473771715e-06, + "loss": 0.3165, + "step": 20978 + }, + { + "epoch": 0.4617873651771957, + "grad_norm": 1.5556882968812453, + "learning_rate": 5.857473564404076e-06, + "loss": 0.2648, + "step": 20979 + }, + { + "epoch": 0.46180937706361436, + "grad_norm": 1.402865613280923, + "learning_rate": 5.857122386732938e-06, + "loss": 0.305, + "step": 20980 + }, + { + "epoch": 0.461831388950033, + "grad_norm": 0.6151064627357519, + "learning_rate": 5.856771204705521e-06, + "loss": 0.4702, + "step": 20981 + }, + { + "epoch": 0.4618534008364517, + "grad_norm": 0.6197430327712373, + "learning_rate": 5.856420018323612e-06, + "loss": 0.4907, + "step": 20982 + }, + { + "epoch": 0.46187541272287036, + "grad_norm": 1.2457149339839915, + "learning_rate": 5.856068827588994e-06, + "loss": 0.2927, + "step": 20983 + }, + { + "epoch": 0.461897424609289, + "grad_norm": 1.8156283524187817, + "learning_rate": 5.855717632503454e-06, + "loss": 0.3798, + "step": 20984 + }, + { + "epoch": 0.46191943649570766, + "grad_norm": 1.1457264822747115, + "learning_rate": 5.8553664330687735e-06, + "loss": 0.236, + "step": 20985 + }, + { + "epoch": 0.46194144838212636, + "grad_norm": 1.5948610460750727, + "learning_rate": 5.8550152292867375e-06, + "loss": 0.3307, + "step": 20986 + }, + { + "epoch": 0.461963460268545, + "grad_norm": 1.3968441942160994, + "learning_rate": 5.854664021159135e-06, + "loss": 0.2989, + "step": 20987 + }, + { + "epoch": 0.46198547215496366, + "grad_norm": 1.0219656834922115, + "learning_rate": 5.854312808687751e-06, + "loss": 0.3349, + "step": 20988 + }, + { + "epoch": 0.46200748404138237, + "grad_norm": 1.3420155291422597, + "learning_rate": 5.8539615918743645e-06, + "loss": 0.2871, + "step": 20989 + }, + { + "epoch": 0.462029495927801, + "grad_norm": 1.8961476213343191, + "learning_rate": 5.853610370720766e-06, + "loss": 0.4197, + "step": 20990 + }, + { + "epoch": 0.46205150781421966, + "grad_norm": 0.6340728765207924, + "learning_rate": 5.853259145228739e-06, + "loss": 0.4921, + "step": 20991 + }, + { + "epoch": 0.46207351970063837, + "grad_norm": 1.4563263760698972, + "learning_rate": 5.852907915400067e-06, + "loss": 0.2811, + "step": 20992 + }, + { + "epoch": 0.462095531587057, + "grad_norm": 1.3071085100883095, + "learning_rate": 5.852556681236539e-06, + "loss": 0.4103, + "step": 20993 + }, + { + "epoch": 0.46211754347347567, + "grad_norm": 1.5367877510431847, + "learning_rate": 5.8522054427399355e-06, + "loss": 0.3132, + "step": 20994 + }, + { + "epoch": 0.4621395553598943, + "grad_norm": 1.5887500848124427, + "learning_rate": 5.851854199912045e-06, + "loss": 0.4197, + "step": 20995 + }, + { + "epoch": 0.462161567246313, + "grad_norm": 1.3435531361629356, + "learning_rate": 5.851502952754651e-06, + "loss": 0.3163, + "step": 20996 + }, + { + "epoch": 0.46218357913273167, + "grad_norm": 1.135681531371535, + "learning_rate": 5.8511517012695394e-06, + "loss": 0.2876, + "step": 20997 + }, + { + "epoch": 0.4622055910191503, + "grad_norm": 1.211571278173032, + "learning_rate": 5.850800445458495e-06, + "loss": 0.3066, + "step": 20998 + }, + { + "epoch": 0.462227602905569, + "grad_norm": 1.613930580310826, + "learning_rate": 5.850449185323302e-06, + "loss": 0.3841, + "step": 20999 + }, + { + "epoch": 0.4622496147919877, + "grad_norm": 0.602189069774939, + "learning_rate": 5.850097920865749e-06, + "loss": 0.4538, + "step": 21000 + }, + { + "epoch": 0.4622716266784063, + "grad_norm": 0.6315568270902874, + "learning_rate": 5.849746652087619e-06, + "loss": 0.5015, + "step": 21001 + }, + { + "epoch": 0.462293638564825, + "grad_norm": 1.34615041329923, + "learning_rate": 5.849395378990698e-06, + "loss": 0.3399, + "step": 21002 + }, + { + "epoch": 0.4623156504512437, + "grad_norm": 1.4940413238672041, + "learning_rate": 5.849044101576769e-06, + "loss": 0.2853, + "step": 21003 + }, + { + "epoch": 0.4623376623376623, + "grad_norm": 1.3703638395829603, + "learning_rate": 5.84869281984762e-06, + "loss": 0.2681, + "step": 21004 + }, + { + "epoch": 0.46235967422408103, + "grad_norm": 1.7238474612962678, + "learning_rate": 5.8483415338050355e-06, + "loss": 0.326, + "step": 21005 + }, + { + "epoch": 0.4623816861104997, + "grad_norm": 1.731017488095277, + "learning_rate": 5.847990243450803e-06, + "loss": 0.3612, + "step": 21006 + }, + { + "epoch": 0.46240369799691833, + "grad_norm": 1.2980164299236643, + "learning_rate": 5.847638948786702e-06, + "loss": 0.3478, + "step": 21007 + }, + { + "epoch": 0.462425709883337, + "grad_norm": 1.3814257157381749, + "learning_rate": 5.847287649814525e-06, + "loss": 0.3309, + "step": 21008 + }, + { + "epoch": 0.4624477217697557, + "grad_norm": 0.6175780918646582, + "learning_rate": 5.846936346536053e-06, + "loss": 0.4657, + "step": 21009 + }, + { + "epoch": 0.46246973365617433, + "grad_norm": 0.6172095757061112, + "learning_rate": 5.846585038953072e-06, + "loss": 0.4534, + "step": 21010 + }, + { + "epoch": 0.462491745542593, + "grad_norm": 1.539302508678517, + "learning_rate": 5.84623372706737e-06, + "loss": 0.3781, + "step": 21011 + }, + { + "epoch": 0.4625137574290117, + "grad_norm": 1.5125537896825652, + "learning_rate": 5.8458824108807285e-06, + "loss": 0.2561, + "step": 21012 + }, + { + "epoch": 0.46253576931543033, + "grad_norm": 0.6143347958278546, + "learning_rate": 5.845531090394937e-06, + "loss": 0.4913, + "step": 21013 + }, + { + "epoch": 0.462557781201849, + "grad_norm": 1.6493769072426083, + "learning_rate": 5.8451797656117794e-06, + "loss": 0.3057, + "step": 21014 + }, + { + "epoch": 0.4625797930882677, + "grad_norm": 1.5050891306697516, + "learning_rate": 5.844828436533041e-06, + "loss": 0.3671, + "step": 21015 + }, + { + "epoch": 0.46260180497468634, + "grad_norm": 1.4104368781606285, + "learning_rate": 5.844477103160509e-06, + "loss": 0.3494, + "step": 21016 + }, + { + "epoch": 0.462623816861105, + "grad_norm": 1.2220502715230586, + "learning_rate": 5.844125765495965e-06, + "loss": 0.3257, + "step": 21017 + }, + { + "epoch": 0.46264582874752364, + "grad_norm": 1.353510679668588, + "learning_rate": 5.843774423541199e-06, + "loss": 0.3218, + "step": 21018 + }, + { + "epoch": 0.46266784063394234, + "grad_norm": 1.8553379889352213, + "learning_rate": 5.843423077297994e-06, + "loss": 0.3316, + "step": 21019 + }, + { + "epoch": 0.462689852520361, + "grad_norm": 3.443020459031213, + "learning_rate": 5.843071726768138e-06, + "loss": 0.2735, + "step": 21020 + }, + { + "epoch": 0.46271186440677964, + "grad_norm": 1.2331249109673363, + "learning_rate": 5.8427203719534155e-06, + "loss": 0.268, + "step": 21021 + }, + { + "epoch": 0.46273387629319834, + "grad_norm": 1.174503547278974, + "learning_rate": 5.84236901285561e-06, + "loss": 0.3505, + "step": 21022 + }, + { + "epoch": 0.462755888179617, + "grad_norm": 0.6708551375401293, + "learning_rate": 5.842017649476511e-06, + "loss": 0.4683, + "step": 21023 + }, + { + "epoch": 0.46277790006603564, + "grad_norm": 0.6117777193326357, + "learning_rate": 5.841666281817905e-06, + "loss": 0.4766, + "step": 21024 + }, + { + "epoch": 0.46279991195245435, + "grad_norm": 1.3039429596309238, + "learning_rate": 5.8413149098815725e-06, + "loss": 0.2203, + "step": 21025 + }, + { + "epoch": 0.462821923838873, + "grad_norm": 1.9163898849798828, + "learning_rate": 5.840963533669303e-06, + "loss": 0.2983, + "step": 21026 + }, + { + "epoch": 0.46284393572529164, + "grad_norm": 1.3408060537124176, + "learning_rate": 5.840612153182882e-06, + "loss": 0.2642, + "step": 21027 + }, + { + "epoch": 0.46286594761171035, + "grad_norm": 1.5647618642268937, + "learning_rate": 5.840260768424094e-06, + "loss": 0.3532, + "step": 21028 + }, + { + "epoch": 0.462887959498129, + "grad_norm": 1.623909525437491, + "learning_rate": 5.839909379394729e-06, + "loss": 0.3566, + "step": 21029 + }, + { + "epoch": 0.46290997138454765, + "grad_norm": 1.4491713899498688, + "learning_rate": 5.839557986096567e-06, + "loss": 0.3613, + "step": 21030 + }, + { + "epoch": 0.4629319832709663, + "grad_norm": 1.0837368853260538, + "learning_rate": 5.8392065885313985e-06, + "loss": 0.2979, + "step": 21031 + }, + { + "epoch": 0.462953995157385, + "grad_norm": 1.501817887782812, + "learning_rate": 5.838855186701007e-06, + "loss": 0.3603, + "step": 21032 + }, + { + "epoch": 0.46297600704380365, + "grad_norm": 1.1203299733582062, + "learning_rate": 5.83850378060718e-06, + "loss": 0.2884, + "step": 21033 + }, + { + "epoch": 0.4629980189302223, + "grad_norm": 2.1613010054904436, + "learning_rate": 5.838152370251703e-06, + "loss": 0.3018, + "step": 21034 + }, + { + "epoch": 0.463020030816641, + "grad_norm": 1.1975990739953448, + "learning_rate": 5.83780095563636e-06, + "loss": 0.3118, + "step": 21035 + }, + { + "epoch": 0.46304204270305965, + "grad_norm": 1.1004660612098536, + "learning_rate": 5.83744953676294e-06, + "loss": 0.2694, + "step": 21036 + }, + { + "epoch": 0.4630640545894783, + "grad_norm": 1.6140814955251765, + "learning_rate": 5.837098113633228e-06, + "loss": 0.2953, + "step": 21037 + }, + { + "epoch": 0.463086066475897, + "grad_norm": 2.343832606474759, + "learning_rate": 5.8367466862490104e-06, + "loss": 0.2775, + "step": 21038 + }, + { + "epoch": 0.46310807836231566, + "grad_norm": 1.5815517553982412, + "learning_rate": 5.8363952546120724e-06, + "loss": 0.2661, + "step": 21039 + }, + { + "epoch": 0.4631300902487343, + "grad_norm": 2.6789497177789903, + "learning_rate": 5.8360438187242005e-06, + "loss": 0.4042, + "step": 21040 + }, + { + "epoch": 0.463152102135153, + "grad_norm": 1.555327278417466, + "learning_rate": 5.835692378587181e-06, + "loss": 0.3277, + "step": 21041 + }, + { + "epoch": 0.46317411402157166, + "grad_norm": 0.6669603668618695, + "learning_rate": 5.835340934202801e-06, + "loss": 0.4914, + "step": 21042 + }, + { + "epoch": 0.4631961259079903, + "grad_norm": 0.6379184729774038, + "learning_rate": 5.834989485572846e-06, + "loss": 0.4869, + "step": 21043 + }, + { + "epoch": 0.46321813779440896, + "grad_norm": 0.6352874716255317, + "learning_rate": 5.834638032699101e-06, + "loss": 0.4924, + "step": 21044 + }, + { + "epoch": 0.46324014968082766, + "grad_norm": 1.604491387861129, + "learning_rate": 5.8342865755833544e-06, + "loss": 0.2922, + "step": 21045 + }, + { + "epoch": 0.4632621615672463, + "grad_norm": 1.5541632592774326, + "learning_rate": 5.833935114227389e-06, + "loss": 0.3219, + "step": 21046 + }, + { + "epoch": 0.46328417345366496, + "grad_norm": 3.062122531269796, + "learning_rate": 5.8335836486329965e-06, + "loss": 0.3417, + "step": 21047 + }, + { + "epoch": 0.46330618534008366, + "grad_norm": 1.2408952046134376, + "learning_rate": 5.833232178801957e-06, + "loss": 0.3579, + "step": 21048 + }, + { + "epoch": 0.4633281972265023, + "grad_norm": 1.2428474921792159, + "learning_rate": 5.832880704736062e-06, + "loss": 0.2805, + "step": 21049 + }, + { + "epoch": 0.46335020911292096, + "grad_norm": 1.7603817255908505, + "learning_rate": 5.832529226437096e-06, + "loss": 0.3939, + "step": 21050 + }, + { + "epoch": 0.46337222099933967, + "grad_norm": 1.898100605909164, + "learning_rate": 5.832177743906843e-06, + "loss": 0.3018, + "step": 21051 + }, + { + "epoch": 0.4633942328857583, + "grad_norm": 1.5971132256999017, + "learning_rate": 5.831826257147094e-06, + "loss": 0.4369, + "step": 21052 + }, + { + "epoch": 0.46341624477217697, + "grad_norm": 1.575984283673373, + "learning_rate": 5.831474766159632e-06, + "loss": 0.3569, + "step": 21053 + }, + { + "epoch": 0.4634382566585956, + "grad_norm": 1.4171553030429718, + "learning_rate": 5.831123270946244e-06, + "loss": 0.2282, + "step": 21054 + }, + { + "epoch": 0.4634602685450143, + "grad_norm": 1.297262550018615, + "learning_rate": 5.8307717715087165e-06, + "loss": 0.2844, + "step": 21055 + }, + { + "epoch": 0.46348228043143297, + "grad_norm": 1.3919342458929964, + "learning_rate": 5.830420267848837e-06, + "loss": 0.2688, + "step": 21056 + }, + { + "epoch": 0.4635042923178516, + "grad_norm": 1.2115598071563982, + "learning_rate": 5.830068759968393e-06, + "loss": 0.2852, + "step": 21057 + }, + { + "epoch": 0.4635263042042703, + "grad_norm": 2.1133783664785506, + "learning_rate": 5.829717247869166e-06, + "loss": 0.3847, + "step": 21058 + }, + { + "epoch": 0.46354831609068897, + "grad_norm": 1.4129917091230404, + "learning_rate": 5.829365731552947e-06, + "loss": 0.3584, + "step": 21059 + }, + { + "epoch": 0.4635703279771076, + "grad_norm": 0.6879649763197486, + "learning_rate": 5.829014211021522e-06, + "loss": 0.4983, + "step": 21060 + }, + { + "epoch": 0.4635923398635263, + "grad_norm": 1.6866774169347412, + "learning_rate": 5.828662686276678e-06, + "loss": 0.3775, + "step": 21061 + }, + { + "epoch": 0.463614351749945, + "grad_norm": 1.5253078377189166, + "learning_rate": 5.828311157320199e-06, + "loss": 0.3014, + "step": 21062 + }, + { + "epoch": 0.4636363636363636, + "grad_norm": 1.7599695822521004, + "learning_rate": 5.8279596241538746e-06, + "loss": 0.3363, + "step": 21063 + }, + { + "epoch": 0.46365837552278233, + "grad_norm": 4.475067332950463, + "learning_rate": 5.827608086779488e-06, + "loss": 0.3429, + "step": 21064 + }, + { + "epoch": 0.463680387409201, + "grad_norm": 1.501932457138226, + "learning_rate": 5.82725654519883e-06, + "loss": 0.4636, + "step": 21065 + }, + { + "epoch": 0.4637023992956196, + "grad_norm": 1.265069529403674, + "learning_rate": 5.826904999413685e-06, + "loss": 0.3393, + "step": 21066 + }, + { + "epoch": 0.4637244111820383, + "grad_norm": 1.3890526741085563, + "learning_rate": 5.826553449425839e-06, + "loss": 0.3639, + "step": 21067 + }, + { + "epoch": 0.463746423068457, + "grad_norm": 1.3375426436445605, + "learning_rate": 5.826201895237081e-06, + "loss": 0.2816, + "step": 21068 + }, + { + "epoch": 0.46376843495487563, + "grad_norm": 4.911234446532059, + "learning_rate": 5.825850336849194e-06, + "loss": 0.3979, + "step": 21069 + }, + { + "epoch": 0.4637904468412943, + "grad_norm": 1.1668876781100521, + "learning_rate": 5.8254987742639714e-06, + "loss": 0.2277, + "step": 21070 + }, + { + "epoch": 0.463812458727713, + "grad_norm": 1.6012026380148299, + "learning_rate": 5.825147207483193e-06, + "loss": 0.3924, + "step": 21071 + }, + { + "epoch": 0.46383447061413163, + "grad_norm": 1.780479005393261, + "learning_rate": 5.824795636508649e-06, + "loss": 0.3899, + "step": 21072 + }, + { + "epoch": 0.4638564825005503, + "grad_norm": 1.6721172303429102, + "learning_rate": 5.824444061342125e-06, + "loss": 0.2957, + "step": 21073 + }, + { + "epoch": 0.463878494386969, + "grad_norm": 1.510018547967141, + "learning_rate": 5.824092481985409e-06, + "loss": 0.3162, + "step": 21074 + }, + { + "epoch": 0.46390050627338764, + "grad_norm": 1.365138972422364, + "learning_rate": 5.823740898440289e-06, + "loss": 0.2526, + "step": 21075 + }, + { + "epoch": 0.4639225181598063, + "grad_norm": 1.4458815514664949, + "learning_rate": 5.823389310708547e-06, + "loss": 0.3093, + "step": 21076 + }, + { + "epoch": 0.46394453004622493, + "grad_norm": 1.3048883897181176, + "learning_rate": 5.8230377187919765e-06, + "loss": 0.2729, + "step": 21077 + }, + { + "epoch": 0.46396654193264364, + "grad_norm": 0.6273372366399814, + "learning_rate": 5.82268612269236e-06, + "loss": 0.4824, + "step": 21078 + }, + { + "epoch": 0.4639885538190623, + "grad_norm": 1.3326406001931053, + "learning_rate": 5.822334522411486e-06, + "loss": 0.2831, + "step": 21079 + }, + { + "epoch": 0.46401056570548094, + "grad_norm": 1.315563259683057, + "learning_rate": 5.821982917951141e-06, + "loss": 0.326, + "step": 21080 + }, + { + "epoch": 0.46403257759189964, + "grad_norm": 1.4689318558446531, + "learning_rate": 5.821631309313112e-06, + "loss": 0.3426, + "step": 21081 + }, + { + "epoch": 0.4640545894783183, + "grad_norm": 1.3475144845210292, + "learning_rate": 5.821279696499186e-06, + "loss": 0.3658, + "step": 21082 + }, + { + "epoch": 0.46407660136473694, + "grad_norm": 1.3441151812933307, + "learning_rate": 5.820928079511151e-06, + "loss": 0.2295, + "step": 21083 + }, + { + "epoch": 0.46409861325115565, + "grad_norm": 1.4027465745459167, + "learning_rate": 5.8205764583507925e-06, + "loss": 0.3136, + "step": 21084 + }, + { + "epoch": 0.4641206251375743, + "grad_norm": 1.3284367598327904, + "learning_rate": 5.8202248330199e-06, + "loss": 0.2611, + "step": 21085 + }, + { + "epoch": 0.46414263702399294, + "grad_norm": 1.8494700124890633, + "learning_rate": 5.819873203520257e-06, + "loss": 0.3006, + "step": 21086 + }, + { + "epoch": 0.46416464891041165, + "grad_norm": 1.2007541408919165, + "learning_rate": 5.819521569853653e-06, + "loss": 0.3094, + "step": 21087 + }, + { + "epoch": 0.4641866607968303, + "grad_norm": 1.987774906007892, + "learning_rate": 5.819169932021878e-06, + "loss": 0.3332, + "step": 21088 + }, + { + "epoch": 0.46420867268324895, + "grad_norm": 1.3551636180950433, + "learning_rate": 5.818818290026712e-06, + "loss": 0.3673, + "step": 21089 + }, + { + "epoch": 0.4642306845696676, + "grad_norm": 1.931154220290153, + "learning_rate": 5.818466643869947e-06, + "loss": 0.2918, + "step": 21090 + }, + { + "epoch": 0.4642526964560863, + "grad_norm": 1.3306193979946348, + "learning_rate": 5.818114993553371e-06, + "loss": 0.3171, + "step": 21091 + }, + { + "epoch": 0.46427470834250495, + "grad_norm": 1.39805870332487, + "learning_rate": 5.817763339078768e-06, + "loss": 0.342, + "step": 21092 + }, + { + "epoch": 0.4642967202289236, + "grad_norm": 1.5777101752765628, + "learning_rate": 5.81741168044793e-06, + "loss": 0.3803, + "step": 21093 + }, + { + "epoch": 0.4643187321153423, + "grad_norm": 1.6348274822496487, + "learning_rate": 5.817060017662638e-06, + "loss": 0.306, + "step": 21094 + }, + { + "epoch": 0.46434074400176095, + "grad_norm": 0.6107886219809101, + "learning_rate": 5.816708350724685e-06, + "loss": 0.4539, + "step": 21095 + }, + { + "epoch": 0.4643627558881796, + "grad_norm": 1.4990714230007305, + "learning_rate": 5.816356679635854e-06, + "loss": 0.4106, + "step": 21096 + }, + { + "epoch": 0.4643847677745983, + "grad_norm": 1.13917109300975, + "learning_rate": 5.816005004397935e-06, + "loss": 0.3001, + "step": 21097 + }, + { + "epoch": 0.46440677966101696, + "grad_norm": 1.5288518806918632, + "learning_rate": 5.815653325012716e-06, + "loss": 0.3515, + "step": 21098 + }, + { + "epoch": 0.4644287915474356, + "grad_norm": 1.4464307023912437, + "learning_rate": 5.815301641481981e-06, + "loss": 0.3288, + "step": 21099 + }, + { + "epoch": 0.4644508034338543, + "grad_norm": 1.6117994647638325, + "learning_rate": 5.814949953807519e-06, + "loss": 0.2965, + "step": 21100 + }, + { + "epoch": 0.46447281532027296, + "grad_norm": 1.6866605021851424, + "learning_rate": 5.814598261991119e-06, + "loss": 0.3281, + "step": 21101 + }, + { + "epoch": 0.4644948272066916, + "grad_norm": 1.8598379144096588, + "learning_rate": 5.814246566034567e-06, + "loss": 0.3784, + "step": 21102 + }, + { + "epoch": 0.46451683909311026, + "grad_norm": 1.6586332663543837, + "learning_rate": 5.8138948659396524e-06, + "loss": 0.3392, + "step": 21103 + }, + { + "epoch": 0.46453885097952896, + "grad_norm": 1.9974976528554171, + "learning_rate": 5.81354316170816e-06, + "loss": 0.3677, + "step": 21104 + }, + { + "epoch": 0.4645608628659476, + "grad_norm": 1.2951608512315484, + "learning_rate": 5.813191453341877e-06, + "loss": 0.2313, + "step": 21105 + }, + { + "epoch": 0.46458287475236626, + "grad_norm": 1.9446983211460642, + "learning_rate": 5.812839740842594e-06, + "loss": 0.3225, + "step": 21106 + }, + { + "epoch": 0.46460488663878496, + "grad_norm": 1.7924098271068818, + "learning_rate": 5.812488024212095e-06, + "loss": 0.2825, + "step": 21107 + }, + { + "epoch": 0.4646268985252036, + "grad_norm": 1.319098127697932, + "learning_rate": 5.8121363034521715e-06, + "loss": 0.3118, + "step": 21108 + }, + { + "epoch": 0.46464891041162226, + "grad_norm": 1.290565342405357, + "learning_rate": 5.811784578564609e-06, + "loss": 0.2377, + "step": 21109 + }, + { + "epoch": 0.46467092229804097, + "grad_norm": 0.664440200982261, + "learning_rate": 5.811432849551193e-06, + "loss": 0.5073, + "step": 21110 + }, + { + "epoch": 0.4646929341844596, + "grad_norm": 1.3406702115035396, + "learning_rate": 5.811081116413717e-06, + "loss": 0.2505, + "step": 21111 + }, + { + "epoch": 0.46471494607087827, + "grad_norm": 2.099047361091606, + "learning_rate": 5.810729379153962e-06, + "loss": 0.3022, + "step": 21112 + }, + { + "epoch": 0.4647369579572969, + "grad_norm": 1.7853495013173033, + "learning_rate": 5.81037763777372e-06, + "loss": 0.3103, + "step": 21113 + }, + { + "epoch": 0.4647589698437156, + "grad_norm": 3.164679270239226, + "learning_rate": 5.810025892274776e-06, + "loss": 0.2928, + "step": 21114 + }, + { + "epoch": 0.46478098173013427, + "grad_norm": 1.2356046324827996, + "learning_rate": 5.8096741426589205e-06, + "loss": 0.281, + "step": 21115 + }, + { + "epoch": 0.4648029936165529, + "grad_norm": 0.6153408098163474, + "learning_rate": 5.809322388927941e-06, + "loss": 0.4739, + "step": 21116 + }, + { + "epoch": 0.4648250055029716, + "grad_norm": 1.1455385865498409, + "learning_rate": 5.808970631083622e-06, + "loss": 0.2392, + "step": 21117 + }, + { + "epoch": 0.46484701738939027, + "grad_norm": 1.2975789631479016, + "learning_rate": 5.808618869127753e-06, + "loss": 0.2207, + "step": 21118 + }, + { + "epoch": 0.4648690292758089, + "grad_norm": 1.3237459292053064, + "learning_rate": 5.808267103062124e-06, + "loss": 0.395, + "step": 21119 + }, + { + "epoch": 0.4648910411622276, + "grad_norm": 2.529816764518958, + "learning_rate": 5.807915332888521e-06, + "loss": 0.3421, + "step": 21120 + }, + { + "epoch": 0.4649130530486463, + "grad_norm": 1.76457466906688, + "learning_rate": 5.807563558608732e-06, + "loss": 0.3871, + "step": 21121 + }, + { + "epoch": 0.4649350649350649, + "grad_norm": 0.6120670584201611, + "learning_rate": 5.807211780224544e-06, + "loss": 0.4794, + "step": 21122 + }, + { + "epoch": 0.46495707682148363, + "grad_norm": 1.4826509038470175, + "learning_rate": 5.806859997737745e-06, + "loss": 0.337, + "step": 21123 + }, + { + "epoch": 0.4649790887079023, + "grad_norm": 1.472020985706968, + "learning_rate": 5.806508211150126e-06, + "loss": 0.2833, + "step": 21124 + }, + { + "epoch": 0.4650011005943209, + "grad_norm": 1.4625511252255434, + "learning_rate": 5.8061564204634705e-06, + "loss": 0.3371, + "step": 21125 + }, + { + "epoch": 0.4650231124807396, + "grad_norm": 1.6284463431383158, + "learning_rate": 5.80580462567957e-06, + "loss": 0.3167, + "step": 21126 + }, + { + "epoch": 0.4650451243671583, + "grad_norm": 1.8747736760987865, + "learning_rate": 5.805452826800209e-06, + "loss": 0.3139, + "step": 21127 + }, + { + "epoch": 0.46506713625357693, + "grad_norm": 1.3160091564446728, + "learning_rate": 5.805101023827179e-06, + "loss": 0.2425, + "step": 21128 + }, + { + "epoch": 0.4650891481399956, + "grad_norm": 1.4121739814617482, + "learning_rate": 5.8047492167622664e-06, + "loss": 0.2992, + "step": 21129 + }, + { + "epoch": 0.4651111600264143, + "grad_norm": 1.1791260658647649, + "learning_rate": 5.804397405607258e-06, + "loss": 0.2236, + "step": 21130 + }, + { + "epoch": 0.46513317191283293, + "grad_norm": 0.6653311445456269, + "learning_rate": 5.804045590363945e-06, + "loss": 0.5075, + "step": 21131 + }, + { + "epoch": 0.4651551837992516, + "grad_norm": 3.1719406367235243, + "learning_rate": 5.803693771034114e-06, + "loss": 0.3961, + "step": 21132 + }, + { + "epoch": 0.4651771956856703, + "grad_norm": 1.653887256370879, + "learning_rate": 5.803341947619552e-06, + "loss": 0.2795, + "step": 21133 + }, + { + "epoch": 0.46519920757208894, + "grad_norm": 2.691353356348352, + "learning_rate": 5.802990120122047e-06, + "loss": 0.3949, + "step": 21134 + }, + { + "epoch": 0.4652212194585076, + "grad_norm": 2.583381271948715, + "learning_rate": 5.8026382885433895e-06, + "loss": 0.3528, + "step": 21135 + }, + { + "epoch": 0.46524323134492623, + "grad_norm": 1.6517690737654769, + "learning_rate": 5.802286452885364e-06, + "loss": 0.3764, + "step": 21136 + }, + { + "epoch": 0.46526524323134494, + "grad_norm": 3.936335843798277, + "learning_rate": 5.8019346131497634e-06, + "loss": 0.3618, + "step": 21137 + }, + { + "epoch": 0.4652872551177636, + "grad_norm": 1.4074938215441868, + "learning_rate": 5.801582769338372e-06, + "loss": 0.3025, + "step": 21138 + }, + { + "epoch": 0.46530926700418224, + "grad_norm": 0.6357991870082582, + "learning_rate": 5.801230921452981e-06, + "loss": 0.4869, + "step": 21139 + }, + { + "epoch": 0.46533127889060094, + "grad_norm": 1.5020838744298983, + "learning_rate": 5.800879069495375e-06, + "loss": 0.3178, + "step": 21140 + }, + { + "epoch": 0.4653532907770196, + "grad_norm": 3.4228137482342373, + "learning_rate": 5.8005272134673455e-06, + "loss": 0.431, + "step": 21141 + }, + { + "epoch": 0.46537530266343824, + "grad_norm": 1.3688019871549826, + "learning_rate": 5.800175353370679e-06, + "loss": 0.3214, + "step": 21142 + }, + { + "epoch": 0.46539731454985694, + "grad_norm": 0.6300788280367299, + "learning_rate": 5.7998234892071655e-06, + "loss": 0.4999, + "step": 21143 + }, + { + "epoch": 0.4654193264362756, + "grad_norm": 1.417622852759916, + "learning_rate": 5.7994716209785925e-06, + "loss": 0.339, + "step": 21144 + }, + { + "epoch": 0.46544133832269424, + "grad_norm": 1.6433909503408206, + "learning_rate": 5.799119748686748e-06, + "loss": 0.2987, + "step": 21145 + }, + { + "epoch": 0.46546335020911295, + "grad_norm": 1.5690499261422792, + "learning_rate": 5.798767872333418e-06, + "loss": 0.3482, + "step": 21146 + }, + { + "epoch": 0.4654853620955316, + "grad_norm": 1.607833010713794, + "learning_rate": 5.798415991920397e-06, + "loss": 0.2957, + "step": 21147 + }, + { + "epoch": 0.46550737398195025, + "grad_norm": 1.4476613025610545, + "learning_rate": 5.798064107449468e-06, + "loss": 0.2719, + "step": 21148 + }, + { + "epoch": 0.4655293858683689, + "grad_norm": 1.6081127006596883, + "learning_rate": 5.797712218922421e-06, + "loss": 0.245, + "step": 21149 + }, + { + "epoch": 0.4655513977547876, + "grad_norm": 1.3218547456587273, + "learning_rate": 5.797360326341046e-06, + "loss": 0.3394, + "step": 21150 + }, + { + "epoch": 0.46557340964120625, + "grad_norm": 1.4012100810313486, + "learning_rate": 5.797008429707128e-06, + "loss": 0.3481, + "step": 21151 + }, + { + "epoch": 0.4655954215276249, + "grad_norm": 1.261577681602222, + "learning_rate": 5.796656529022461e-06, + "loss": 0.3165, + "step": 21152 + }, + { + "epoch": 0.4656174334140436, + "grad_norm": 1.928686064113259, + "learning_rate": 5.796304624288828e-06, + "loss": 0.3714, + "step": 21153 + }, + { + "epoch": 0.46563944530046225, + "grad_norm": 0.6310806062774852, + "learning_rate": 5.79595271550802e-06, + "loss": 0.4592, + "step": 21154 + }, + { + "epoch": 0.4656614571868809, + "grad_norm": 1.4003384063903932, + "learning_rate": 5.7956008026818245e-06, + "loss": 0.3066, + "step": 21155 + }, + { + "epoch": 0.4656834690732996, + "grad_norm": 0.6945003659722262, + "learning_rate": 5.795248885812033e-06, + "loss": 0.4942, + "step": 21156 + }, + { + "epoch": 0.46570548095971825, + "grad_norm": 1.298427675710992, + "learning_rate": 5.79489696490043e-06, + "loss": 0.3686, + "step": 21157 + }, + { + "epoch": 0.4657274928461369, + "grad_norm": 1.2841705680328803, + "learning_rate": 5.794545039948807e-06, + "loss": 0.4021, + "step": 21158 + }, + { + "epoch": 0.46574950473255555, + "grad_norm": 1.2792042999985307, + "learning_rate": 5.79419311095895e-06, + "loss": 0.3271, + "step": 21159 + }, + { + "epoch": 0.46577151661897426, + "grad_norm": 1.5306146257882045, + "learning_rate": 5.793841177932651e-06, + "loss": 0.3322, + "step": 21160 + }, + { + "epoch": 0.4657935285053929, + "grad_norm": 1.4716403393982416, + "learning_rate": 5.793489240871697e-06, + "loss": 0.3422, + "step": 21161 + }, + { + "epoch": 0.46581554039181156, + "grad_norm": 1.1917614668100613, + "learning_rate": 5.793137299777878e-06, + "loss": 0.2767, + "step": 21162 + }, + { + "epoch": 0.46583755227823026, + "grad_norm": 1.5274608943271077, + "learning_rate": 5.79278535465298e-06, + "loss": 0.3062, + "step": 21163 + }, + { + "epoch": 0.4658595641646489, + "grad_norm": 1.7040409848499605, + "learning_rate": 5.792433405498792e-06, + "loss": 0.313, + "step": 21164 + }, + { + "epoch": 0.46588157605106756, + "grad_norm": 1.9568657225731743, + "learning_rate": 5.792081452317106e-06, + "loss": 0.3321, + "step": 21165 + }, + { + "epoch": 0.46590358793748626, + "grad_norm": 1.0095285947550587, + "learning_rate": 5.7917294951097085e-06, + "loss": 0.213, + "step": 21166 + }, + { + "epoch": 0.4659255998239049, + "grad_norm": 1.142745981832498, + "learning_rate": 5.791377533878388e-06, + "loss": 0.2805, + "step": 21167 + }, + { + "epoch": 0.46594761171032356, + "grad_norm": 1.3936401391840125, + "learning_rate": 5.791025568624935e-06, + "loss": 0.3858, + "step": 21168 + }, + { + "epoch": 0.46596962359674227, + "grad_norm": 1.2628929410257732, + "learning_rate": 5.790673599351136e-06, + "loss": 0.1831, + "step": 21169 + }, + { + "epoch": 0.4659916354831609, + "grad_norm": 1.5140966494065327, + "learning_rate": 5.790321626058782e-06, + "loss": 0.3061, + "step": 21170 + }, + { + "epoch": 0.46601364736957956, + "grad_norm": 0.6637735754617146, + "learning_rate": 5.78996964874966e-06, + "loss": 0.4887, + "step": 21171 + }, + { + "epoch": 0.4660356592559982, + "grad_norm": 1.5012948908695076, + "learning_rate": 5.78961766742556e-06, + "loss": 0.3282, + "step": 21172 + }, + { + "epoch": 0.4660576711424169, + "grad_norm": 2.0520286850102876, + "learning_rate": 5.789265682088272e-06, + "loss": 0.3452, + "step": 21173 + }, + { + "epoch": 0.46607968302883557, + "grad_norm": 0.5960436847331725, + "learning_rate": 5.788913692739584e-06, + "loss": 0.4877, + "step": 21174 + }, + { + "epoch": 0.4661016949152542, + "grad_norm": 1.2137729237228527, + "learning_rate": 5.7885616993812834e-06, + "loss": 0.2992, + "step": 21175 + }, + { + "epoch": 0.4661237068016729, + "grad_norm": 1.7461050357208567, + "learning_rate": 5.788209702015162e-06, + "loss": 0.3271, + "step": 21176 + }, + { + "epoch": 0.46614571868809157, + "grad_norm": 1.3272359132887652, + "learning_rate": 5.787857700643005e-06, + "loss": 0.327, + "step": 21177 + }, + { + "epoch": 0.4661677305745102, + "grad_norm": 1.1264868887378265, + "learning_rate": 5.787505695266605e-06, + "loss": 0.254, + "step": 21178 + }, + { + "epoch": 0.4661897424609289, + "grad_norm": 1.2860701156491396, + "learning_rate": 5.78715368588775e-06, + "loss": 0.3391, + "step": 21179 + }, + { + "epoch": 0.4662117543473476, + "grad_norm": 1.2113822788740045, + "learning_rate": 5.786801672508229e-06, + "loss": 0.3932, + "step": 21180 + }, + { + "epoch": 0.4662337662337662, + "grad_norm": 1.141736307699184, + "learning_rate": 5.786449655129831e-06, + "loss": 0.2824, + "step": 21181 + }, + { + "epoch": 0.4662557781201849, + "grad_norm": 1.8533123858726206, + "learning_rate": 5.7860976337543436e-06, + "loss": 0.3055, + "step": 21182 + }, + { + "epoch": 0.4662777900066036, + "grad_norm": 1.9369811262915448, + "learning_rate": 5.785745608383561e-06, + "loss": 0.3239, + "step": 21183 + }, + { + "epoch": 0.4662998018930222, + "grad_norm": 1.2471622687054276, + "learning_rate": 5.785393579019265e-06, + "loss": 0.2994, + "step": 21184 + }, + { + "epoch": 0.4663218137794409, + "grad_norm": 1.3662032063201637, + "learning_rate": 5.7850415456632504e-06, + "loss": 0.2958, + "step": 21185 + }, + { + "epoch": 0.4663438256658596, + "grad_norm": 2.213901324875055, + "learning_rate": 5.784689508317304e-06, + "loss": 0.2781, + "step": 21186 + }, + { + "epoch": 0.46636583755227823, + "grad_norm": 0.6642797024533155, + "learning_rate": 5.7843374669832155e-06, + "loss": 0.4814, + "step": 21187 + }, + { + "epoch": 0.4663878494386969, + "grad_norm": 1.4207122444396503, + "learning_rate": 5.783985421662774e-06, + "loss": 0.2516, + "step": 21188 + }, + { + "epoch": 0.4664098613251156, + "grad_norm": 2.411729338430214, + "learning_rate": 5.783633372357771e-06, + "loss": 0.3442, + "step": 21189 + }, + { + "epoch": 0.46643187321153423, + "grad_norm": 1.6033479927039689, + "learning_rate": 5.783281319069991e-06, + "loss": 0.3075, + "step": 21190 + }, + { + "epoch": 0.4664538850979529, + "grad_norm": 1.6562868922453766, + "learning_rate": 5.7829292618012266e-06, + "loss": 0.2864, + "step": 21191 + }, + { + "epoch": 0.4664758969843716, + "grad_norm": 1.180630430502139, + "learning_rate": 5.7825772005532665e-06, + "loss": 0.2551, + "step": 21192 + }, + { + "epoch": 0.46649790887079023, + "grad_norm": 1.5679708592115864, + "learning_rate": 5.782225135327901e-06, + "loss": 0.3151, + "step": 21193 + }, + { + "epoch": 0.4665199207572089, + "grad_norm": 2.2520936167145513, + "learning_rate": 5.7818730661269186e-06, + "loss": 0.3348, + "step": 21194 + }, + { + "epoch": 0.46654193264362753, + "grad_norm": 1.5738599646200533, + "learning_rate": 5.781520992952107e-06, + "loss": 0.2253, + "step": 21195 + }, + { + "epoch": 0.46656394453004624, + "grad_norm": 1.3458159437656332, + "learning_rate": 5.781168915805258e-06, + "loss": 0.2256, + "step": 21196 + }, + { + "epoch": 0.4665859564164649, + "grad_norm": 1.7462996168280744, + "learning_rate": 5.78081683468816e-06, + "loss": 0.3323, + "step": 21197 + }, + { + "epoch": 0.46660796830288354, + "grad_norm": 1.4316888819870375, + "learning_rate": 5.780464749602602e-06, + "loss": 0.2994, + "step": 21198 + }, + { + "epoch": 0.46662998018930224, + "grad_norm": 1.1778083053690844, + "learning_rate": 5.7801126605503754e-06, + "loss": 0.2355, + "step": 21199 + }, + { + "epoch": 0.4666519920757209, + "grad_norm": 1.1592077884845764, + "learning_rate": 5.7797605675332664e-06, + "loss": 0.2844, + "step": 21200 + }, + { + "epoch": 0.46667400396213954, + "grad_norm": 1.7234317568695157, + "learning_rate": 5.779408470553068e-06, + "loss": 0.3352, + "step": 21201 + }, + { + "epoch": 0.46669601584855824, + "grad_norm": 2.03267094287746, + "learning_rate": 5.779056369611569e-06, + "loss": 0.3946, + "step": 21202 + }, + { + "epoch": 0.4667180277349769, + "grad_norm": 1.4345579905733896, + "learning_rate": 5.778704264710556e-06, + "loss": 0.2889, + "step": 21203 + }, + { + "epoch": 0.46674003962139554, + "grad_norm": 1.4706266795424652, + "learning_rate": 5.778352155851822e-06, + "loss": 0.337, + "step": 21204 + }, + { + "epoch": 0.46676205150781425, + "grad_norm": 1.1784910067848717, + "learning_rate": 5.778000043037153e-06, + "loss": 0.2689, + "step": 21205 + }, + { + "epoch": 0.4667840633942329, + "grad_norm": 1.1460460331300457, + "learning_rate": 5.777647926268344e-06, + "loss": 0.2253, + "step": 21206 + }, + { + "epoch": 0.46680607528065154, + "grad_norm": 1.0323616628886891, + "learning_rate": 5.77729580554718e-06, + "loss": 0.2991, + "step": 21207 + }, + { + "epoch": 0.4668280871670702, + "grad_norm": 2.7548471067910554, + "learning_rate": 5.776943680875451e-06, + "loss": 0.3747, + "step": 21208 + }, + { + "epoch": 0.4668500990534889, + "grad_norm": 2.893036699158458, + "learning_rate": 5.776591552254949e-06, + "loss": 0.3959, + "step": 21209 + }, + { + "epoch": 0.46687211093990755, + "grad_norm": 1.5610223657459217, + "learning_rate": 5.776239419687462e-06, + "loss": 0.2787, + "step": 21210 + }, + { + "epoch": 0.4668941228263262, + "grad_norm": 1.2237049775058697, + "learning_rate": 5.77588728317478e-06, + "loss": 0.2259, + "step": 21211 + }, + { + "epoch": 0.4669161347127449, + "grad_norm": 1.305572698478968, + "learning_rate": 5.775535142718694e-06, + "loss": 0.2866, + "step": 21212 + }, + { + "epoch": 0.46693814659916355, + "grad_norm": 1.2719142544165243, + "learning_rate": 5.77518299832099e-06, + "loss": 0.3082, + "step": 21213 + }, + { + "epoch": 0.4669601584855822, + "grad_norm": 3.181504186555973, + "learning_rate": 5.774830849983462e-06, + "loss": 0.2577, + "step": 21214 + }, + { + "epoch": 0.4669821703720009, + "grad_norm": 1.4813817882606428, + "learning_rate": 5.774478697707897e-06, + "loss": 0.2963, + "step": 21215 + }, + { + "epoch": 0.46700418225841955, + "grad_norm": 1.4329641347757025, + "learning_rate": 5.774126541496087e-06, + "loss": 0.3751, + "step": 21216 + }, + { + "epoch": 0.4670261941448382, + "grad_norm": 1.2703049995010138, + "learning_rate": 5.773774381349819e-06, + "loss": 0.2352, + "step": 21217 + }, + { + "epoch": 0.46704820603125685, + "grad_norm": 1.5684769412829953, + "learning_rate": 5.7734222172708855e-06, + "loss": 0.3904, + "step": 21218 + }, + { + "epoch": 0.46707021791767556, + "grad_norm": 0.682351442321392, + "learning_rate": 5.773070049261076e-06, + "loss": 0.4999, + "step": 21219 + }, + { + "epoch": 0.4670922298040942, + "grad_norm": 0.6589616249156903, + "learning_rate": 5.772717877322179e-06, + "loss": 0.4917, + "step": 21220 + }, + { + "epoch": 0.46711424169051285, + "grad_norm": 1.2593760727172967, + "learning_rate": 5.772365701455985e-06, + "loss": 0.3285, + "step": 21221 + }, + { + "epoch": 0.46713625357693156, + "grad_norm": 1.5250851308705222, + "learning_rate": 5.772013521664283e-06, + "loss": 0.3712, + "step": 21222 + }, + { + "epoch": 0.4671582654633502, + "grad_norm": 1.2966569090600462, + "learning_rate": 5.771661337948865e-06, + "loss": 0.2966, + "step": 21223 + }, + { + "epoch": 0.46718027734976886, + "grad_norm": 1.4152720319805951, + "learning_rate": 5.771309150311521e-06, + "loss": 0.3459, + "step": 21224 + }, + { + "epoch": 0.46720228923618756, + "grad_norm": 1.690666714244617, + "learning_rate": 5.770956958754039e-06, + "loss": 0.3559, + "step": 21225 + }, + { + "epoch": 0.4672243011226062, + "grad_norm": 1.2096791001169684, + "learning_rate": 5.7706047632782095e-06, + "loss": 0.2143, + "step": 21226 + }, + { + "epoch": 0.46724631300902486, + "grad_norm": 1.2435314198408551, + "learning_rate": 5.770252563885823e-06, + "loss": 0.3468, + "step": 21227 + }, + { + "epoch": 0.46726832489544357, + "grad_norm": 1.414179114715392, + "learning_rate": 5.769900360578671e-06, + "loss": 0.2625, + "step": 21228 + }, + { + "epoch": 0.4672903367818622, + "grad_norm": 1.0916853414229581, + "learning_rate": 5.769548153358541e-06, + "loss": 0.3189, + "step": 21229 + }, + { + "epoch": 0.46731234866828086, + "grad_norm": 0.6843831315566774, + "learning_rate": 5.769195942227224e-06, + "loss": 0.5088, + "step": 21230 + }, + { + "epoch": 0.4673343605546995, + "grad_norm": 1.5940204978641777, + "learning_rate": 5.768843727186509e-06, + "loss": 0.3837, + "step": 21231 + }, + { + "epoch": 0.4673563724411182, + "grad_norm": 1.0940123712570051, + "learning_rate": 5.768491508238187e-06, + "loss": 0.2104, + "step": 21232 + }, + { + "epoch": 0.46737838432753687, + "grad_norm": 0.6569276031692193, + "learning_rate": 5.76813928538405e-06, + "loss": 0.5194, + "step": 21233 + }, + { + "epoch": 0.4674003962139555, + "grad_norm": 1.7776559132541094, + "learning_rate": 5.7677870586258865e-06, + "loss": 0.3356, + "step": 21234 + }, + { + "epoch": 0.4674224081003742, + "grad_norm": 1.648643855067385, + "learning_rate": 5.767434827965486e-06, + "loss": 0.3455, + "step": 21235 + }, + { + "epoch": 0.46744441998679287, + "grad_norm": 1.7101133074651982, + "learning_rate": 5.76708259340464e-06, + "loss": 0.3178, + "step": 21236 + }, + { + "epoch": 0.4674664318732115, + "grad_norm": 1.3541643043907698, + "learning_rate": 5.766730354945137e-06, + "loss": 0.2706, + "step": 21237 + }, + { + "epoch": 0.4674884437596302, + "grad_norm": 1.2503284485726498, + "learning_rate": 5.766378112588771e-06, + "loss": 0.34, + "step": 21238 + }, + { + "epoch": 0.4675104556460489, + "grad_norm": 1.5647145601585664, + "learning_rate": 5.766025866337326e-06, + "loss": 0.2688, + "step": 21239 + }, + { + "epoch": 0.4675324675324675, + "grad_norm": 1.3884502283078004, + "learning_rate": 5.765673616192597e-06, + "loss": 0.2959, + "step": 21240 + }, + { + "epoch": 0.46755447941888617, + "grad_norm": 0.6412062395295228, + "learning_rate": 5.765321362156374e-06, + "loss": 0.4953, + "step": 21241 + }, + { + "epoch": 0.4675764913053049, + "grad_norm": 8.840896756205487, + "learning_rate": 5.764969104230447e-06, + "loss": 0.3646, + "step": 21242 + }, + { + "epoch": 0.4675985031917235, + "grad_norm": 1.9899129154559296, + "learning_rate": 5.764616842416606e-06, + "loss": 0.2568, + "step": 21243 + }, + { + "epoch": 0.4676205150781422, + "grad_norm": 1.261192022310354, + "learning_rate": 5.764264576716639e-06, + "loss": 0.3543, + "step": 21244 + }, + { + "epoch": 0.4676425269645609, + "grad_norm": 1.3805743665270336, + "learning_rate": 5.76391230713234e-06, + "loss": 0.3481, + "step": 21245 + }, + { + "epoch": 0.4676645388509795, + "grad_norm": 1.3170043198244965, + "learning_rate": 5.763560033665498e-06, + "loss": 0.2685, + "step": 21246 + }, + { + "epoch": 0.4676865507373982, + "grad_norm": 1.2331875130388008, + "learning_rate": 5.763207756317903e-06, + "loss": 0.355, + "step": 21247 + }, + { + "epoch": 0.4677085626238169, + "grad_norm": 1.583994382035423, + "learning_rate": 5.7628554750913464e-06, + "loss": 0.3986, + "step": 21248 + }, + { + "epoch": 0.46773057451023553, + "grad_norm": 1.7267995224451596, + "learning_rate": 5.7625031899876175e-06, + "loss": 0.3685, + "step": 21249 + }, + { + "epoch": 0.4677525863966542, + "grad_norm": 1.3236710874727382, + "learning_rate": 5.762150901008508e-06, + "loss": 0.273, + "step": 21250 + }, + { + "epoch": 0.4677745982830729, + "grad_norm": 1.8067542131347605, + "learning_rate": 5.761798608155806e-06, + "loss": 0.4094, + "step": 21251 + }, + { + "epoch": 0.46779661016949153, + "grad_norm": 0.5923561964281383, + "learning_rate": 5.761446311431305e-06, + "loss": 0.4531, + "step": 21252 + }, + { + "epoch": 0.4678186220559102, + "grad_norm": 1.2287066292483158, + "learning_rate": 5.7610940108367956e-06, + "loss": 0.2708, + "step": 21253 + }, + { + "epoch": 0.46784063394232883, + "grad_norm": 0.6143239983272358, + "learning_rate": 5.760741706374064e-06, + "loss": 0.5012, + "step": 21254 + }, + { + "epoch": 0.46786264582874754, + "grad_norm": 1.8379385283782066, + "learning_rate": 5.760389398044905e-06, + "loss": 0.2935, + "step": 21255 + }, + { + "epoch": 0.4678846577151662, + "grad_norm": 0.6256076159656775, + "learning_rate": 5.76003708585111e-06, + "loss": 0.4915, + "step": 21256 + }, + { + "epoch": 0.46790666960158483, + "grad_norm": 1.679610101985901, + "learning_rate": 5.759684769794465e-06, + "loss": 0.392, + "step": 21257 + }, + { + "epoch": 0.46792868148800354, + "grad_norm": 0.5988646965359714, + "learning_rate": 5.759332449876763e-06, + "loss": 0.4649, + "step": 21258 + }, + { + "epoch": 0.4679506933744222, + "grad_norm": 2.895012725734332, + "learning_rate": 5.758980126099796e-06, + "loss": 0.3434, + "step": 21259 + }, + { + "epoch": 0.46797270526084084, + "grad_norm": 1.9686103349330464, + "learning_rate": 5.758627798465353e-06, + "loss": 0.2739, + "step": 21260 + }, + { + "epoch": 0.46799471714725954, + "grad_norm": 1.8321834110425341, + "learning_rate": 5.7582754669752264e-06, + "loss": 0.3295, + "step": 21261 + }, + { + "epoch": 0.4680167290336782, + "grad_norm": 0.6434807928446785, + "learning_rate": 5.7579231316312036e-06, + "loss": 0.4976, + "step": 21262 + }, + { + "epoch": 0.46803874092009684, + "grad_norm": 0.6436708860533404, + "learning_rate": 5.757570792435079e-06, + "loss": 0.4792, + "step": 21263 + }, + { + "epoch": 0.46806075280651555, + "grad_norm": 1.1866898180775918, + "learning_rate": 5.757218449388641e-06, + "loss": 0.3183, + "step": 21264 + }, + { + "epoch": 0.4680827646929342, + "grad_norm": 1.3976425137217419, + "learning_rate": 5.756866102493681e-06, + "loss": 0.3833, + "step": 21265 + }, + { + "epoch": 0.46810477657935284, + "grad_norm": 1.3446446351117318, + "learning_rate": 5.756513751751991e-06, + "loss": 0.3509, + "step": 21266 + }, + { + "epoch": 0.4681267884657715, + "grad_norm": 1.3115098147911834, + "learning_rate": 5.756161397165358e-06, + "loss": 0.3219, + "step": 21267 + }, + { + "epoch": 0.4681488003521902, + "grad_norm": 1.4101752486035666, + "learning_rate": 5.755809038735578e-06, + "loss": 0.3324, + "step": 21268 + }, + { + "epoch": 0.46817081223860885, + "grad_norm": 1.239080088898218, + "learning_rate": 5.755456676464439e-06, + "loss": 0.3373, + "step": 21269 + }, + { + "epoch": 0.4681928241250275, + "grad_norm": 0.6464264322882538, + "learning_rate": 5.755104310353732e-06, + "loss": 0.5101, + "step": 21270 + }, + { + "epoch": 0.4682148360114462, + "grad_norm": 1.4909030453291074, + "learning_rate": 5.754751940405248e-06, + "loss": 0.3521, + "step": 21271 + }, + { + "epoch": 0.46823684789786485, + "grad_norm": 1.615362289172357, + "learning_rate": 5.7543995666207776e-06, + "loss": 0.3696, + "step": 21272 + }, + { + "epoch": 0.4682588597842835, + "grad_norm": 1.2669220989384948, + "learning_rate": 5.754047189002112e-06, + "loss": 0.3297, + "step": 21273 + }, + { + "epoch": 0.4682808716707022, + "grad_norm": 1.366491335654113, + "learning_rate": 5.753694807551045e-06, + "loss": 0.3096, + "step": 21274 + }, + { + "epoch": 0.46830288355712085, + "grad_norm": 1.4362131964128355, + "learning_rate": 5.753342422269361e-06, + "loss": 0.2949, + "step": 21275 + }, + { + "epoch": 0.4683248954435395, + "grad_norm": 1.3157787233269875, + "learning_rate": 5.752990033158857e-06, + "loss": 0.3003, + "step": 21276 + }, + { + "epoch": 0.46834690732995815, + "grad_norm": 1.38306623346688, + "learning_rate": 5.752637640221321e-06, + "loss": 0.3863, + "step": 21277 + }, + { + "epoch": 0.46836891921637686, + "grad_norm": 1.4189605218172976, + "learning_rate": 5.7522852434585454e-06, + "loss": 0.3264, + "step": 21278 + }, + { + "epoch": 0.4683909311027955, + "grad_norm": 2.2134057279194717, + "learning_rate": 5.7519328428723205e-06, + "loss": 0.3752, + "step": 21279 + }, + { + "epoch": 0.46841294298921415, + "grad_norm": 1.6629738406522168, + "learning_rate": 5.751580438464437e-06, + "loss": 0.3197, + "step": 21280 + }, + { + "epoch": 0.46843495487563286, + "grad_norm": 1.515205064723485, + "learning_rate": 5.751228030236688e-06, + "loss": 0.322, + "step": 21281 + }, + { + "epoch": 0.4684569667620515, + "grad_norm": 1.196382526621193, + "learning_rate": 5.750875618190861e-06, + "loss": 0.2098, + "step": 21282 + }, + { + "epoch": 0.46847897864847016, + "grad_norm": 1.0594202812248779, + "learning_rate": 5.750523202328751e-06, + "loss": 0.2287, + "step": 21283 + }, + { + "epoch": 0.46850099053488886, + "grad_norm": 1.329998438030238, + "learning_rate": 5.750170782652147e-06, + "loss": 0.2518, + "step": 21284 + }, + { + "epoch": 0.4685230024213075, + "grad_norm": 2.0465626568437263, + "learning_rate": 5.7498183591628385e-06, + "loss": 0.2948, + "step": 21285 + }, + { + "epoch": 0.46854501430772616, + "grad_norm": 1.5079016036356434, + "learning_rate": 5.749465931862621e-06, + "loss": 0.2278, + "step": 21286 + }, + { + "epoch": 0.46856702619414486, + "grad_norm": 1.7100307756024877, + "learning_rate": 5.7491135007532825e-06, + "loss": 0.2876, + "step": 21287 + }, + { + "epoch": 0.4685890380805635, + "grad_norm": 1.6455118189782998, + "learning_rate": 5.748761065836616e-06, + "loss": 0.319, + "step": 21288 + }, + { + "epoch": 0.46861104996698216, + "grad_norm": 1.307993620423668, + "learning_rate": 5.74840862711441e-06, + "loss": 0.3542, + "step": 21289 + }, + { + "epoch": 0.4686330618534008, + "grad_norm": 1.4160403076463293, + "learning_rate": 5.748056184588458e-06, + "loss": 0.3421, + "step": 21290 + }, + { + "epoch": 0.4686550737398195, + "grad_norm": 2.6219592048138822, + "learning_rate": 5.747703738260551e-06, + "loss": 0.2805, + "step": 21291 + }, + { + "epoch": 0.46867708562623817, + "grad_norm": 1.918960038277846, + "learning_rate": 5.747351288132481e-06, + "loss": 0.4025, + "step": 21292 + }, + { + "epoch": 0.4686990975126568, + "grad_norm": 1.268057934506506, + "learning_rate": 5.746998834206038e-06, + "loss": 0.3164, + "step": 21293 + }, + { + "epoch": 0.4687211093990755, + "grad_norm": 1.560156413344658, + "learning_rate": 5.746646376483014e-06, + "loss": 0.3989, + "step": 21294 + }, + { + "epoch": 0.46874312128549417, + "grad_norm": 1.450473505845693, + "learning_rate": 5.7462939149652e-06, + "loss": 0.3085, + "step": 21295 + }, + { + "epoch": 0.4687651331719128, + "grad_norm": 1.9974761707999744, + "learning_rate": 5.7459414496543865e-06, + "loss": 0.3438, + "step": 21296 + }, + { + "epoch": 0.4687871450583315, + "grad_norm": 1.8148968589055574, + "learning_rate": 5.745588980552368e-06, + "loss": 0.3418, + "step": 21297 + }, + { + "epoch": 0.46880915694475017, + "grad_norm": 1.3109659682401262, + "learning_rate": 5.745236507660931e-06, + "loss": 0.3047, + "step": 21298 + }, + { + "epoch": 0.4688311688311688, + "grad_norm": 1.4902532846450574, + "learning_rate": 5.744884030981871e-06, + "loss": 0.3298, + "step": 21299 + }, + { + "epoch": 0.46885318071758747, + "grad_norm": 1.1849361419177014, + "learning_rate": 5.744531550516978e-06, + "loss": 0.3516, + "step": 21300 + }, + { + "epoch": 0.4688751926040062, + "grad_norm": 1.3420974950497269, + "learning_rate": 5.744179066268043e-06, + "loss": 0.2831, + "step": 21301 + }, + { + "epoch": 0.4688972044904248, + "grad_norm": 1.2609605275439015, + "learning_rate": 5.7438265782368605e-06, + "loss": 0.2409, + "step": 21302 + }, + { + "epoch": 0.4689192163768435, + "grad_norm": 1.476530064391135, + "learning_rate": 5.743474086425217e-06, + "loss": 0.3658, + "step": 21303 + }, + { + "epoch": 0.4689412282632622, + "grad_norm": 0.6420921719336181, + "learning_rate": 5.743121590834908e-06, + "loss": 0.4921, + "step": 21304 + }, + { + "epoch": 0.4689632401496808, + "grad_norm": 1.4860417054478499, + "learning_rate": 5.742769091467724e-06, + "loss": 0.309, + "step": 21305 + }, + { + "epoch": 0.4689852520360995, + "grad_norm": 1.568133229667351, + "learning_rate": 5.7424165883254545e-06, + "loss": 0.2906, + "step": 21306 + }, + { + "epoch": 0.4690072639225182, + "grad_norm": 1.52122573081452, + "learning_rate": 5.742064081409894e-06, + "loss": 0.3512, + "step": 21307 + }, + { + "epoch": 0.46902927580893683, + "grad_norm": 1.2459407372627187, + "learning_rate": 5.741711570722831e-06, + "loss": 0.2885, + "step": 21308 + }, + { + "epoch": 0.4690512876953555, + "grad_norm": 1.206319279017109, + "learning_rate": 5.741359056266062e-06, + "loss": 0.2865, + "step": 21309 + }, + { + "epoch": 0.4690732995817742, + "grad_norm": 1.3778891348138658, + "learning_rate": 5.741006538041374e-06, + "loss": 0.34, + "step": 21310 + }, + { + "epoch": 0.46909531146819283, + "grad_norm": 1.9231289845055046, + "learning_rate": 5.740654016050559e-06, + "loss": 0.2718, + "step": 21311 + }, + { + "epoch": 0.4691173233546115, + "grad_norm": 1.2688341646392, + "learning_rate": 5.740301490295412e-06, + "loss": 0.3115, + "step": 21312 + }, + { + "epoch": 0.46913933524103013, + "grad_norm": 1.0753728587535314, + "learning_rate": 5.739948960777722e-06, + "loss": 0.2742, + "step": 21313 + }, + { + "epoch": 0.46916134712744884, + "grad_norm": 0.6301886935321889, + "learning_rate": 5.739596427499281e-06, + "loss": 0.4925, + "step": 21314 + }, + { + "epoch": 0.4691833590138675, + "grad_norm": 1.4692159213106595, + "learning_rate": 5.739243890461882e-06, + "loss": 0.3422, + "step": 21315 + }, + { + "epoch": 0.46920537090028613, + "grad_norm": 1.4912545477303045, + "learning_rate": 5.738891349667315e-06, + "loss": 0.3515, + "step": 21316 + }, + { + "epoch": 0.46922738278670484, + "grad_norm": 1.3274200151796585, + "learning_rate": 5.738538805117373e-06, + "loss": 0.4247, + "step": 21317 + }, + { + "epoch": 0.4692493946731235, + "grad_norm": 1.5360854772774704, + "learning_rate": 5.738186256813847e-06, + "loss": 0.3406, + "step": 21318 + }, + { + "epoch": 0.46927140655954214, + "grad_norm": 1.0960547322212937, + "learning_rate": 5.737833704758529e-06, + "loss": 0.3177, + "step": 21319 + }, + { + "epoch": 0.46929341844596084, + "grad_norm": 1.4344959031011368, + "learning_rate": 5.737481148953212e-06, + "loss": 0.3993, + "step": 21320 + }, + { + "epoch": 0.4693154303323795, + "grad_norm": 1.145782472825396, + "learning_rate": 5.737128589399685e-06, + "loss": 0.2799, + "step": 21321 + }, + { + "epoch": 0.46933744221879814, + "grad_norm": 1.3238348931472697, + "learning_rate": 5.736776026099744e-06, + "loss": 0.3556, + "step": 21322 + }, + { + "epoch": 0.46935945410521684, + "grad_norm": 1.1703567423596928, + "learning_rate": 5.736423459055178e-06, + "loss": 0.3563, + "step": 21323 + }, + { + "epoch": 0.4693814659916355, + "grad_norm": 1.420874125375945, + "learning_rate": 5.73607088826778e-06, + "loss": 0.2805, + "step": 21324 + }, + { + "epoch": 0.46940347787805414, + "grad_norm": 1.5024595401241063, + "learning_rate": 5.73571831373934e-06, + "loss": 0.3328, + "step": 21325 + }, + { + "epoch": 0.4694254897644728, + "grad_norm": 1.711554586998718, + "learning_rate": 5.735365735471651e-06, + "loss": 0.3881, + "step": 21326 + }, + { + "epoch": 0.4694475016508915, + "grad_norm": 0.6478566398085067, + "learning_rate": 5.735013153466507e-06, + "loss": 0.4953, + "step": 21327 + }, + { + "epoch": 0.46946951353731015, + "grad_norm": 1.4668599865304022, + "learning_rate": 5.734660567725697e-06, + "loss": 0.3183, + "step": 21328 + }, + { + "epoch": 0.4694915254237288, + "grad_norm": 5.089861760895246, + "learning_rate": 5.734307978251016e-06, + "loss": 0.2993, + "step": 21329 + }, + { + "epoch": 0.4695135373101475, + "grad_norm": 0.6309137665550012, + "learning_rate": 5.733955385044253e-06, + "loss": 0.5053, + "step": 21330 + }, + { + "epoch": 0.46953554919656615, + "grad_norm": 1.57337908151699, + "learning_rate": 5.733602788107202e-06, + "loss": 0.2895, + "step": 21331 + }, + { + "epoch": 0.4695575610829848, + "grad_norm": 1.2378262922640368, + "learning_rate": 5.733250187441654e-06, + "loss": 0.298, + "step": 21332 + }, + { + "epoch": 0.4695795729694035, + "grad_norm": 1.3498292932786418, + "learning_rate": 5.732897583049402e-06, + "loss": 0.2636, + "step": 21333 + }, + { + "epoch": 0.46960158485582215, + "grad_norm": 1.9151811087323451, + "learning_rate": 5.732544974932237e-06, + "loss": 0.2705, + "step": 21334 + }, + { + "epoch": 0.4696235967422408, + "grad_norm": 1.6106120981412877, + "learning_rate": 5.732192363091952e-06, + "loss": 0.2905, + "step": 21335 + }, + { + "epoch": 0.46964560862865945, + "grad_norm": 1.2567538976648323, + "learning_rate": 5.7318397475303385e-06, + "loss": 0.2936, + "step": 21336 + }, + { + "epoch": 0.46966762051507815, + "grad_norm": 0.6310157884270613, + "learning_rate": 5.7314871282491895e-06, + "loss": 0.5064, + "step": 21337 + }, + { + "epoch": 0.4696896324014968, + "grad_norm": 1.7383930214498642, + "learning_rate": 5.731134505250298e-06, + "loss": 0.3394, + "step": 21338 + }, + { + "epoch": 0.46971164428791545, + "grad_norm": 1.646191055708359, + "learning_rate": 5.730781878535452e-06, + "loss": 0.3618, + "step": 21339 + }, + { + "epoch": 0.46973365617433416, + "grad_norm": 1.9338570185258823, + "learning_rate": 5.730429248106448e-06, + "loss": 0.3237, + "step": 21340 + }, + { + "epoch": 0.4697556680607528, + "grad_norm": 1.382817057580582, + "learning_rate": 5.7300766139650765e-06, + "loss": 0.3082, + "step": 21341 + }, + { + "epoch": 0.46977767994717146, + "grad_norm": 1.2749193667921797, + "learning_rate": 5.72972397611313e-06, + "loss": 0.371, + "step": 21342 + }, + { + "epoch": 0.46979969183359016, + "grad_norm": 1.3323572686081477, + "learning_rate": 5.729371334552403e-06, + "loss": 0.324, + "step": 21343 + }, + { + "epoch": 0.4698217037200088, + "grad_norm": 1.3579834445325867, + "learning_rate": 5.729018689284682e-06, + "loss": 0.3008, + "step": 21344 + }, + { + "epoch": 0.46984371560642746, + "grad_norm": 0.6219316514345199, + "learning_rate": 5.728666040311764e-06, + "loss": 0.4817, + "step": 21345 + }, + { + "epoch": 0.46986572749284616, + "grad_norm": 0.614997909332389, + "learning_rate": 5.728313387635441e-06, + "loss": 0.4822, + "step": 21346 + }, + { + "epoch": 0.4698877393792648, + "grad_norm": 1.151320608969408, + "learning_rate": 5.727960731257505e-06, + "loss": 0.2978, + "step": 21347 + }, + { + "epoch": 0.46990975126568346, + "grad_norm": 1.4456372400022306, + "learning_rate": 5.7276080711797465e-06, + "loss": 0.2183, + "step": 21348 + }, + { + "epoch": 0.4699317631521021, + "grad_norm": 1.482240312036316, + "learning_rate": 5.72725540740396e-06, + "loss": 0.3666, + "step": 21349 + }, + { + "epoch": 0.4699537750385208, + "grad_norm": 0.6190099329596136, + "learning_rate": 5.726902739931936e-06, + "loss": 0.461, + "step": 21350 + }, + { + "epoch": 0.46997578692493946, + "grad_norm": 1.3112082122622144, + "learning_rate": 5.72655006876547e-06, + "loss": 0.3854, + "step": 21351 + }, + { + "epoch": 0.4699977988113581, + "grad_norm": 1.209844741100063, + "learning_rate": 5.726197393906352e-06, + "loss": 0.3052, + "step": 21352 + }, + { + "epoch": 0.4700198106977768, + "grad_norm": 1.3880007883441583, + "learning_rate": 5.725844715356375e-06, + "loss": 0.2713, + "step": 21353 + }, + { + "epoch": 0.47004182258419547, + "grad_norm": 1.4560229931518718, + "learning_rate": 5.72549203311733e-06, + "loss": 0.4223, + "step": 21354 + }, + { + "epoch": 0.4700638344706141, + "grad_norm": 0.6176935356019663, + "learning_rate": 5.725139347191011e-06, + "loss": 0.467, + "step": 21355 + }, + { + "epoch": 0.4700858463570328, + "grad_norm": 1.389360690603818, + "learning_rate": 5.724786657579212e-06, + "loss": 0.348, + "step": 21356 + }, + { + "epoch": 0.47010785824345147, + "grad_norm": 1.682105697122939, + "learning_rate": 5.724433964283724e-06, + "loss": 0.4149, + "step": 21357 + }, + { + "epoch": 0.4701298701298701, + "grad_norm": 1.3568368750476543, + "learning_rate": 5.724081267306337e-06, + "loss": 0.3249, + "step": 21358 + }, + { + "epoch": 0.47015188201628877, + "grad_norm": 11.287549691122305, + "learning_rate": 5.723728566648849e-06, + "loss": 0.2639, + "step": 21359 + }, + { + "epoch": 0.4701738939027075, + "grad_norm": 0.6243797711007275, + "learning_rate": 5.723375862313047e-06, + "loss": 0.4749, + "step": 21360 + }, + { + "epoch": 0.4701959057891261, + "grad_norm": 1.2814191477487211, + "learning_rate": 5.723023154300729e-06, + "loss": 0.3082, + "step": 21361 + }, + { + "epoch": 0.47021791767554477, + "grad_norm": 1.7864697730152455, + "learning_rate": 5.7226704426136826e-06, + "loss": 0.4, + "step": 21362 + }, + { + "epoch": 0.4702399295619635, + "grad_norm": 1.4652796040212288, + "learning_rate": 5.722317727253703e-06, + "loss": 0.3122, + "step": 21363 + }, + { + "epoch": 0.4702619414483821, + "grad_norm": 0.6252399836961864, + "learning_rate": 5.721965008222584e-06, + "loss": 0.5006, + "step": 21364 + }, + { + "epoch": 0.4702839533348008, + "grad_norm": 1.6461990292441437, + "learning_rate": 5.721612285522115e-06, + "loss": 0.3371, + "step": 21365 + }, + { + "epoch": 0.4703059652212195, + "grad_norm": 1.3159789035357337, + "learning_rate": 5.721259559154093e-06, + "loss": 0.3481, + "step": 21366 + }, + { + "epoch": 0.47032797710763813, + "grad_norm": 1.222708953642083, + "learning_rate": 5.7209068291203065e-06, + "loss": 0.2254, + "step": 21367 + }, + { + "epoch": 0.4703499889940568, + "grad_norm": 1.4066265808938159, + "learning_rate": 5.720554095422549e-06, + "loss": 0.3321, + "step": 21368 + }, + { + "epoch": 0.4703720008804755, + "grad_norm": 2.797127308898661, + "learning_rate": 5.720201358062616e-06, + "loss": 0.3332, + "step": 21369 + }, + { + "epoch": 0.47039401276689413, + "grad_norm": 1.2137697212266574, + "learning_rate": 5.719848617042297e-06, + "loss": 0.2558, + "step": 21370 + }, + { + "epoch": 0.4704160246533128, + "grad_norm": 1.1901341607441713, + "learning_rate": 5.719495872363388e-06, + "loss": 0.2329, + "step": 21371 + }, + { + "epoch": 0.47043803653973143, + "grad_norm": 1.2773147218633703, + "learning_rate": 5.7191431240276795e-06, + "loss": 0.3625, + "step": 21372 + }, + { + "epoch": 0.47046004842615013, + "grad_norm": 1.498044494428587, + "learning_rate": 5.718790372036963e-06, + "loss": 0.3522, + "step": 21373 + }, + { + "epoch": 0.4704820603125688, + "grad_norm": 2.006166229209105, + "learning_rate": 5.718437616393036e-06, + "loss": 0.32, + "step": 21374 + }, + { + "epoch": 0.47050407219898743, + "grad_norm": 1.1251738821968316, + "learning_rate": 5.7180848570976866e-06, + "loss": 0.2945, + "step": 21375 + }, + { + "epoch": 0.47052608408540614, + "grad_norm": 1.0960051279801817, + "learning_rate": 5.71773209415271e-06, + "loss": 0.2551, + "step": 21376 + }, + { + "epoch": 0.4705480959718248, + "grad_norm": 0.6428238093218208, + "learning_rate": 5.7173793275598995e-06, + "loss": 0.4858, + "step": 21377 + }, + { + "epoch": 0.47057010785824344, + "grad_norm": 1.580081671858157, + "learning_rate": 5.717026557321047e-06, + "loss": 0.2238, + "step": 21378 + }, + { + "epoch": 0.47059211974466214, + "grad_norm": 0.6219679497428025, + "learning_rate": 5.7166737834379475e-06, + "loss": 0.4799, + "step": 21379 + }, + { + "epoch": 0.4706141316310808, + "grad_norm": 1.4772861346555697, + "learning_rate": 5.716321005912389e-06, + "loss": 0.3575, + "step": 21380 + }, + { + "epoch": 0.47063614351749944, + "grad_norm": 1.3019095386578583, + "learning_rate": 5.715968224746169e-06, + "loss": 0.3326, + "step": 21381 + }, + { + "epoch": 0.4706581554039181, + "grad_norm": 1.9163488670348483, + "learning_rate": 5.7156154399410805e-06, + "loss": 0.4328, + "step": 21382 + }, + { + "epoch": 0.4706801672903368, + "grad_norm": 1.430184387678814, + "learning_rate": 5.715262651498913e-06, + "loss": 0.2813, + "step": 21383 + }, + { + "epoch": 0.47070217917675544, + "grad_norm": 1.73627583266827, + "learning_rate": 5.714909859421463e-06, + "loss": 0.4243, + "step": 21384 + }, + { + "epoch": 0.4707241910631741, + "grad_norm": 1.381691088288411, + "learning_rate": 5.714557063710522e-06, + "loss": 0.3226, + "step": 21385 + }, + { + "epoch": 0.4707462029495928, + "grad_norm": 0.6349377617976085, + "learning_rate": 5.714204264367882e-06, + "loss": 0.4791, + "step": 21386 + }, + { + "epoch": 0.47076821483601144, + "grad_norm": 1.8815724248410415, + "learning_rate": 5.713851461395339e-06, + "loss": 0.3606, + "step": 21387 + }, + { + "epoch": 0.4707902267224301, + "grad_norm": 6.372344868726662, + "learning_rate": 5.713498654794685e-06, + "loss": 0.324, + "step": 21388 + }, + { + "epoch": 0.4708122386088488, + "grad_norm": 8.369359138127171, + "learning_rate": 5.713145844567711e-06, + "loss": 0.3, + "step": 21389 + }, + { + "epoch": 0.47083425049526745, + "grad_norm": 1.6310012653795982, + "learning_rate": 5.712793030716213e-06, + "loss": 0.3417, + "step": 21390 + }, + { + "epoch": 0.4708562623816861, + "grad_norm": 1.2163538511522858, + "learning_rate": 5.712440213241981e-06, + "loss": 0.2839, + "step": 21391 + }, + { + "epoch": 0.4708782742681048, + "grad_norm": 0.6076024788836295, + "learning_rate": 5.712087392146811e-06, + "loss": 0.5021, + "step": 21392 + }, + { + "epoch": 0.47090028615452345, + "grad_norm": 1.6312949590671928, + "learning_rate": 5.711734567432496e-06, + "loss": 0.337, + "step": 21393 + }, + { + "epoch": 0.4709222980409421, + "grad_norm": 1.383631417787871, + "learning_rate": 5.711381739100828e-06, + "loss": 0.3299, + "step": 21394 + }, + { + "epoch": 0.47094430992736075, + "grad_norm": 1.5964332301241004, + "learning_rate": 5.711028907153602e-06, + "loss": 0.3467, + "step": 21395 + }, + { + "epoch": 0.47096632181377945, + "grad_norm": 1.5263385414005417, + "learning_rate": 5.710676071592608e-06, + "loss": 0.3539, + "step": 21396 + }, + { + "epoch": 0.4709883337001981, + "grad_norm": 1.8806610014180352, + "learning_rate": 5.710323232419644e-06, + "loss": 0.2999, + "step": 21397 + }, + { + "epoch": 0.47101034558661675, + "grad_norm": 1.1727698755956517, + "learning_rate": 5.7099703896364966e-06, + "loss": 0.3041, + "step": 21398 + }, + { + "epoch": 0.47103235747303546, + "grad_norm": 3.129774450799256, + "learning_rate": 5.7096175432449655e-06, + "loss": 0.3643, + "step": 21399 + }, + { + "epoch": 0.4710543693594541, + "grad_norm": 1.1932797641741846, + "learning_rate": 5.709264693246841e-06, + "loss": 0.2635, + "step": 21400 + }, + { + "epoch": 0.47107638124587276, + "grad_norm": 0.6219582873807581, + "learning_rate": 5.708911839643916e-06, + "loss": 0.4796, + "step": 21401 + }, + { + "epoch": 0.47109839313229146, + "grad_norm": 0.6275698754534256, + "learning_rate": 5.7085589824379875e-06, + "loss": 0.4916, + "step": 21402 + }, + { + "epoch": 0.4711204050187101, + "grad_norm": 1.608275605364261, + "learning_rate": 5.708206121630843e-06, + "loss": 0.3543, + "step": 21403 + }, + { + "epoch": 0.47114241690512876, + "grad_norm": 1.5547723670393112, + "learning_rate": 5.707853257224281e-06, + "loss": 0.3077, + "step": 21404 + }, + { + "epoch": 0.47116442879154746, + "grad_norm": 1.2532054100543146, + "learning_rate": 5.7075003892200935e-06, + "loss": 0.3664, + "step": 21405 + }, + { + "epoch": 0.4711864406779661, + "grad_norm": 0.5943091186329421, + "learning_rate": 5.707147517620073e-06, + "loss": 0.4852, + "step": 21406 + }, + { + "epoch": 0.47120845256438476, + "grad_norm": 1.7319357266433548, + "learning_rate": 5.706794642426013e-06, + "loss": 0.361, + "step": 21407 + }, + { + "epoch": 0.4712304644508034, + "grad_norm": 0.6313118167119285, + "learning_rate": 5.706441763639707e-06, + "loss": 0.4645, + "step": 21408 + }, + { + "epoch": 0.4712524763372221, + "grad_norm": 1.620211884750234, + "learning_rate": 5.706088881262949e-06, + "loss": 0.2772, + "step": 21409 + }, + { + "epoch": 0.47127448822364076, + "grad_norm": 1.5633252293035622, + "learning_rate": 5.7057359952975335e-06, + "loss": 0.3513, + "step": 21410 + }, + { + "epoch": 0.4712965001100594, + "grad_norm": 1.4575746925778854, + "learning_rate": 5.705383105745251e-06, + "loss": 0.2661, + "step": 21411 + }, + { + "epoch": 0.4713185119964781, + "grad_norm": 1.3179574549224702, + "learning_rate": 5.705030212607899e-06, + "loss": 0.3154, + "step": 21412 + }, + { + "epoch": 0.47134052388289677, + "grad_norm": 1.378914131040648, + "learning_rate": 5.7046773158872685e-06, + "loss": 0.3645, + "step": 21413 + }, + { + "epoch": 0.4713625357693154, + "grad_norm": 2.0138343000063785, + "learning_rate": 5.7043244155851514e-06, + "loss": 0.2856, + "step": 21414 + }, + { + "epoch": 0.4713845476557341, + "grad_norm": 1.33794986439929, + "learning_rate": 5.703971511703348e-06, + "loss": 0.3798, + "step": 21415 + }, + { + "epoch": 0.47140655954215277, + "grad_norm": 2.3543320301864057, + "learning_rate": 5.703618604243644e-06, + "loss": 0.3072, + "step": 21416 + }, + { + "epoch": 0.4714285714285714, + "grad_norm": 1.3667216306029557, + "learning_rate": 5.7032656932078365e-06, + "loss": 0.2835, + "step": 21417 + }, + { + "epoch": 0.47145058331499007, + "grad_norm": 1.4601254199639115, + "learning_rate": 5.70291277859772e-06, + "loss": 0.3083, + "step": 21418 + }, + { + "epoch": 0.4714725952014088, + "grad_norm": 1.80757015287415, + "learning_rate": 5.702559860415086e-06, + "loss": 0.4072, + "step": 21419 + }, + { + "epoch": 0.4714946070878274, + "grad_norm": 1.37015392806848, + "learning_rate": 5.7022069386617316e-06, + "loss": 0.2871, + "step": 21420 + }, + { + "epoch": 0.47151661897424607, + "grad_norm": 0.6156850446972582, + "learning_rate": 5.701854013339447e-06, + "loss": 0.4791, + "step": 21421 + }, + { + "epoch": 0.4715386308606648, + "grad_norm": 1.6074126821509476, + "learning_rate": 5.701501084450027e-06, + "loss": 0.2939, + "step": 21422 + }, + { + "epoch": 0.4715606427470834, + "grad_norm": 1.6704164954059346, + "learning_rate": 5.7011481519952665e-06, + "loss": 0.3408, + "step": 21423 + }, + { + "epoch": 0.4715826546335021, + "grad_norm": 1.1731057751181866, + "learning_rate": 5.7007952159769574e-06, + "loss": 0.3483, + "step": 21424 + }, + { + "epoch": 0.4716046665199208, + "grad_norm": 0.6471135987266523, + "learning_rate": 5.700442276396896e-06, + "loss": 0.5376, + "step": 21425 + }, + { + "epoch": 0.47162667840633943, + "grad_norm": 1.299487704185978, + "learning_rate": 5.700089333256873e-06, + "loss": 0.3224, + "step": 21426 + }, + { + "epoch": 0.4716486902927581, + "grad_norm": 1.6875943528677302, + "learning_rate": 5.699736386558683e-06, + "loss": 0.3192, + "step": 21427 + }, + { + "epoch": 0.4716707021791768, + "grad_norm": 1.5515719870641858, + "learning_rate": 5.699383436304123e-06, + "loss": 0.405, + "step": 21428 + }, + { + "epoch": 0.47169271406559543, + "grad_norm": 1.4765814988894557, + "learning_rate": 5.6990304824949826e-06, + "loss": 0.388, + "step": 21429 + }, + { + "epoch": 0.4717147259520141, + "grad_norm": 1.3017105725836509, + "learning_rate": 5.6986775251330586e-06, + "loss": 0.3615, + "step": 21430 + }, + { + "epoch": 0.47173673783843273, + "grad_norm": 0.6203476666139325, + "learning_rate": 5.698324564220144e-06, + "loss": 0.4948, + "step": 21431 + }, + { + "epoch": 0.47175874972485143, + "grad_norm": 1.4539967239174478, + "learning_rate": 5.697971599758031e-06, + "loss": 0.3162, + "step": 21432 + }, + { + "epoch": 0.4717807616112701, + "grad_norm": 1.1746888214131788, + "learning_rate": 5.697618631748517e-06, + "loss": 0.3182, + "step": 21433 + }, + { + "epoch": 0.47180277349768873, + "grad_norm": 1.6114370067518473, + "learning_rate": 5.697265660193392e-06, + "loss": 0.3445, + "step": 21434 + }, + { + "epoch": 0.47182478538410744, + "grad_norm": 2.153770789289796, + "learning_rate": 5.696912685094453e-06, + "loss": 0.3137, + "step": 21435 + }, + { + "epoch": 0.4718467972705261, + "grad_norm": 1.3365040112521722, + "learning_rate": 5.696559706453493e-06, + "loss": 0.3953, + "step": 21436 + }, + { + "epoch": 0.47186880915694474, + "grad_norm": 2.3393193395970746, + "learning_rate": 5.696206724272304e-06, + "loss": 0.3268, + "step": 21437 + }, + { + "epoch": 0.47189082104336344, + "grad_norm": 1.5303587626087138, + "learning_rate": 5.695853738552686e-06, + "loss": 0.2495, + "step": 21438 + }, + { + "epoch": 0.4719128329297821, + "grad_norm": 1.6234841552232926, + "learning_rate": 5.695500749296425e-06, + "loss": 0.405, + "step": 21439 + }, + { + "epoch": 0.47193484481620074, + "grad_norm": 1.1853270990667133, + "learning_rate": 5.69514775650532e-06, + "loss": 0.3144, + "step": 21440 + }, + { + "epoch": 0.4719568567026194, + "grad_norm": 1.5820092422634275, + "learning_rate": 5.694794760181165e-06, + "loss": 0.3224, + "step": 21441 + }, + { + "epoch": 0.4719788685890381, + "grad_norm": 1.3207941622477493, + "learning_rate": 5.694441760325753e-06, + "loss": 0.2663, + "step": 21442 + }, + { + "epoch": 0.47200088047545674, + "grad_norm": 2.027792349747659, + "learning_rate": 5.694088756940878e-06, + "loss": 0.2956, + "step": 21443 + }, + { + "epoch": 0.4720228923618754, + "grad_norm": 1.1978315098857815, + "learning_rate": 5.693735750028334e-06, + "loss": 0.2796, + "step": 21444 + }, + { + "epoch": 0.4720449042482941, + "grad_norm": 2.739727890000983, + "learning_rate": 5.693382739589916e-06, + "loss": 0.3874, + "step": 21445 + }, + { + "epoch": 0.47206691613471274, + "grad_norm": 1.3707571653488604, + "learning_rate": 5.693029725627417e-06, + "loss": 0.3222, + "step": 21446 + }, + { + "epoch": 0.4720889280211314, + "grad_norm": 1.7776814175187168, + "learning_rate": 5.692676708142632e-06, + "loss": 0.3819, + "step": 21447 + }, + { + "epoch": 0.4721109399075501, + "grad_norm": 6.128763741808417, + "learning_rate": 5.6923236871373555e-06, + "loss": 0.3041, + "step": 21448 + }, + { + "epoch": 0.47213295179396875, + "grad_norm": 1.6445575087606195, + "learning_rate": 5.691970662613382e-06, + "loss": 0.3707, + "step": 21449 + }, + { + "epoch": 0.4721549636803874, + "grad_norm": 0.587406571567538, + "learning_rate": 5.691617634572503e-06, + "loss": 0.4514, + "step": 21450 + }, + { + "epoch": 0.4721769755668061, + "grad_norm": 1.394934013382292, + "learning_rate": 5.691264603016515e-06, + "loss": 0.3561, + "step": 21451 + }, + { + "epoch": 0.47219898745322475, + "grad_norm": 0.6334979772187603, + "learning_rate": 5.690911567947213e-06, + "loss": 0.4849, + "step": 21452 + }, + { + "epoch": 0.4722209993396434, + "grad_norm": 1.3333419751272422, + "learning_rate": 5.69055852936639e-06, + "loss": 0.3288, + "step": 21453 + }, + { + "epoch": 0.47224301122606205, + "grad_norm": 1.8989074250221494, + "learning_rate": 5.690205487275842e-06, + "loss": 0.2434, + "step": 21454 + }, + { + "epoch": 0.47226502311248075, + "grad_norm": 1.3576596082976005, + "learning_rate": 5.68985244167736e-06, + "loss": 0.3025, + "step": 21455 + }, + { + "epoch": 0.4722870349988994, + "grad_norm": 1.1612023933666817, + "learning_rate": 5.689499392572741e-06, + "loss": 0.2351, + "step": 21456 + }, + { + "epoch": 0.47230904688531805, + "grad_norm": 1.69307648191496, + "learning_rate": 5.689146339963778e-06, + "loss": 0.4151, + "step": 21457 + }, + { + "epoch": 0.47233105877173676, + "grad_norm": 1.55587902845996, + "learning_rate": 5.688793283852265e-06, + "loss": 0.3102, + "step": 21458 + }, + { + "epoch": 0.4723530706581554, + "grad_norm": 0.6354869722854308, + "learning_rate": 5.688440224239998e-06, + "loss": 0.4827, + "step": 21459 + }, + { + "epoch": 0.47237508254457405, + "grad_norm": 1.2780098919150287, + "learning_rate": 5.688087161128773e-06, + "loss": 0.3455, + "step": 21460 + }, + { + "epoch": 0.47239709443099276, + "grad_norm": 1.1256454205572382, + "learning_rate": 5.68773409452038e-06, + "loss": 0.2586, + "step": 21461 + }, + { + "epoch": 0.4724191063174114, + "grad_norm": 1.2420007725293192, + "learning_rate": 5.687381024416615e-06, + "loss": 0.3369, + "step": 21462 + }, + { + "epoch": 0.47244111820383006, + "grad_norm": 1.276347676602508, + "learning_rate": 5.687027950819274e-06, + "loss": 0.3114, + "step": 21463 + }, + { + "epoch": 0.47246313009024876, + "grad_norm": 0.6338721470777662, + "learning_rate": 5.68667487373015e-06, + "loss": 0.4811, + "step": 21464 + }, + { + "epoch": 0.4724851419766674, + "grad_norm": 1.3188964181715133, + "learning_rate": 5.6863217931510385e-06, + "loss": 0.2862, + "step": 21465 + }, + { + "epoch": 0.47250715386308606, + "grad_norm": 1.4979390888445654, + "learning_rate": 5.685968709083734e-06, + "loss": 0.3762, + "step": 21466 + }, + { + "epoch": 0.4725291657495047, + "grad_norm": 2.0189534654884804, + "learning_rate": 5.685615621530029e-06, + "loss": 0.2816, + "step": 21467 + }, + { + "epoch": 0.4725511776359234, + "grad_norm": 1.5163921419661046, + "learning_rate": 5.685262530491721e-06, + "loss": 0.3365, + "step": 21468 + }, + { + "epoch": 0.47257318952234206, + "grad_norm": 1.685610397489484, + "learning_rate": 5.684909435970602e-06, + "loss": 0.3254, + "step": 21469 + }, + { + "epoch": 0.4725952014087607, + "grad_norm": 0.6207854051428868, + "learning_rate": 5.68455633796847e-06, + "loss": 0.4912, + "step": 21470 + }, + { + "epoch": 0.4726172132951794, + "grad_norm": 0.6165184723354272, + "learning_rate": 5.684203236487116e-06, + "loss": 0.4993, + "step": 21471 + }, + { + "epoch": 0.47263922518159807, + "grad_norm": 0.6182071135300817, + "learning_rate": 5.683850131528336e-06, + "loss": 0.5054, + "step": 21472 + }, + { + "epoch": 0.4726612370680167, + "grad_norm": 1.2337836687270027, + "learning_rate": 5.683497023093925e-06, + "loss": 0.2809, + "step": 21473 + }, + { + "epoch": 0.4726832489544354, + "grad_norm": 1.5602781865377893, + "learning_rate": 5.683143911185677e-06, + "loss": 0.2907, + "step": 21474 + }, + { + "epoch": 0.47270526084085407, + "grad_norm": 1.9496027067517951, + "learning_rate": 5.6827907958053875e-06, + "loss": 0.3976, + "step": 21475 + }, + { + "epoch": 0.4727272727272727, + "grad_norm": 0.599375236532148, + "learning_rate": 5.682437676954848e-06, + "loss": 0.4872, + "step": 21476 + }, + { + "epoch": 0.47274928461369137, + "grad_norm": 1.5870320959185698, + "learning_rate": 5.682084554635858e-06, + "loss": 0.2694, + "step": 21477 + }, + { + "epoch": 0.47277129650011007, + "grad_norm": 0.6016713899798187, + "learning_rate": 5.681731428850211e-06, + "loss": 0.4646, + "step": 21478 + }, + { + "epoch": 0.4727933083865287, + "grad_norm": 1.4337311206358874, + "learning_rate": 5.6813782995997e-06, + "loss": 0.3283, + "step": 21479 + }, + { + "epoch": 0.47281532027294737, + "grad_norm": 2.258495251471584, + "learning_rate": 5.6810251668861205e-06, + "loss": 0.3163, + "step": 21480 + }, + { + "epoch": 0.4728373321593661, + "grad_norm": 0.6071283340681927, + "learning_rate": 5.680672030711266e-06, + "loss": 0.4817, + "step": 21481 + }, + { + "epoch": 0.4728593440457847, + "grad_norm": 1.320836371739927, + "learning_rate": 5.680318891076935e-06, + "loss": 0.3323, + "step": 21482 + }, + { + "epoch": 0.4728813559322034, + "grad_norm": 1.5257797618625828, + "learning_rate": 5.679965747984918e-06, + "loss": 0.3429, + "step": 21483 + }, + { + "epoch": 0.4729033678186221, + "grad_norm": 1.275510770929627, + "learning_rate": 5.679612601437013e-06, + "loss": 0.3308, + "step": 21484 + }, + { + "epoch": 0.4729253797050407, + "grad_norm": 0.6137955662543392, + "learning_rate": 5.679259451435013e-06, + "loss": 0.4813, + "step": 21485 + }, + { + "epoch": 0.4729473915914594, + "grad_norm": 0.6138857418235678, + "learning_rate": 5.678906297980713e-06, + "loss": 0.4691, + "step": 21486 + }, + { + "epoch": 0.4729694034778781, + "grad_norm": 3.6224294821898213, + "learning_rate": 5.67855314107591e-06, + "loss": 0.3789, + "step": 21487 + }, + { + "epoch": 0.47299141536429673, + "grad_norm": 1.332725641609664, + "learning_rate": 5.678199980722398e-06, + "loss": 0.2859, + "step": 21488 + }, + { + "epoch": 0.4730134272507154, + "grad_norm": 0.6342006670462661, + "learning_rate": 5.677846816921968e-06, + "loss": 0.4842, + "step": 21489 + }, + { + "epoch": 0.47303543913713403, + "grad_norm": 1.4373200250711966, + "learning_rate": 5.67749364967642e-06, + "loss": 0.2882, + "step": 21490 + }, + { + "epoch": 0.47305745102355273, + "grad_norm": 1.5787983154087522, + "learning_rate": 5.677140478987546e-06, + "loss": 0.319, + "step": 21491 + }, + { + "epoch": 0.4730794629099714, + "grad_norm": 0.6086886946583803, + "learning_rate": 5.676787304857143e-06, + "loss": 0.4755, + "step": 21492 + }, + { + "epoch": 0.47310147479639003, + "grad_norm": 2.148857650189706, + "learning_rate": 5.676434127287007e-06, + "loss": 0.3577, + "step": 21493 + }, + { + "epoch": 0.47312348668280874, + "grad_norm": 1.3695013434942487, + "learning_rate": 5.676080946278928e-06, + "loss": 0.3492, + "step": 21494 + }, + { + "epoch": 0.4731454985692274, + "grad_norm": 1.248272980292909, + "learning_rate": 5.675727761834705e-06, + "loss": 0.3292, + "step": 21495 + }, + { + "epoch": 0.47316751045564603, + "grad_norm": 1.5167172210561781, + "learning_rate": 5.675374573956132e-06, + "loss": 0.3272, + "step": 21496 + }, + { + "epoch": 0.47318952234206474, + "grad_norm": 0.6431459714205413, + "learning_rate": 5.675021382645005e-06, + "loss": 0.492, + "step": 21497 + }, + { + "epoch": 0.4732115342284834, + "grad_norm": 1.573000263908603, + "learning_rate": 5.674668187903118e-06, + "loss": 0.3307, + "step": 21498 + }, + { + "epoch": 0.47323354611490204, + "grad_norm": 1.5407148064597171, + "learning_rate": 5.674314989732266e-06, + "loss": 0.2143, + "step": 21499 + }, + { + "epoch": 0.4732555580013207, + "grad_norm": 1.35605669998281, + "learning_rate": 5.673961788134245e-06, + "loss": 0.2701, + "step": 21500 + }, + { + "epoch": 0.4732775698877394, + "grad_norm": 1.4149395767903405, + "learning_rate": 5.673608583110849e-06, + "loss": 0.3218, + "step": 21501 + }, + { + "epoch": 0.47329958177415804, + "grad_norm": 1.6327333460622049, + "learning_rate": 5.6732553746638744e-06, + "loss": 0.3851, + "step": 21502 + }, + { + "epoch": 0.4733215936605767, + "grad_norm": 1.993122407936871, + "learning_rate": 5.672902162795115e-06, + "loss": 0.3934, + "step": 21503 + }, + { + "epoch": 0.4733436055469954, + "grad_norm": 0.6320000951080802, + "learning_rate": 5.672548947506366e-06, + "loss": 0.4817, + "step": 21504 + }, + { + "epoch": 0.47336561743341404, + "grad_norm": 0.6020617615959601, + "learning_rate": 5.672195728799425e-06, + "loss": 0.4645, + "step": 21505 + }, + { + "epoch": 0.4733876293198327, + "grad_norm": 1.884165737437312, + "learning_rate": 5.671842506676085e-06, + "loss": 0.311, + "step": 21506 + }, + { + "epoch": 0.4734096412062514, + "grad_norm": 1.205690948956202, + "learning_rate": 5.67148928113814e-06, + "loss": 0.2094, + "step": 21507 + }, + { + "epoch": 0.47343165309267005, + "grad_norm": 1.6555415475545905, + "learning_rate": 5.67113605218739e-06, + "loss": 0.3829, + "step": 21508 + }, + { + "epoch": 0.4734536649790887, + "grad_norm": 0.5993143640278307, + "learning_rate": 5.670782819825625e-06, + "loss": 0.4638, + "step": 21509 + }, + { + "epoch": 0.4734756768655074, + "grad_norm": 1.2718693681126827, + "learning_rate": 5.670429584054643e-06, + "loss": 0.2366, + "step": 21510 + }, + { + "epoch": 0.47349768875192605, + "grad_norm": 1.5493580609534152, + "learning_rate": 5.670076344876239e-06, + "loss": 0.3774, + "step": 21511 + }, + { + "epoch": 0.4735197006383447, + "grad_norm": 4.062440206639655, + "learning_rate": 5.669723102292208e-06, + "loss": 0.2423, + "step": 21512 + }, + { + "epoch": 0.47354171252476335, + "grad_norm": 1.278809932291928, + "learning_rate": 5.669369856304347e-06, + "loss": 0.2398, + "step": 21513 + }, + { + "epoch": 0.47356372441118205, + "grad_norm": 1.548961447675634, + "learning_rate": 5.669016606914448e-06, + "loss": 0.4654, + "step": 21514 + }, + { + "epoch": 0.4735857362976007, + "grad_norm": 1.3822257375306208, + "learning_rate": 5.668663354124308e-06, + "loss": 0.339, + "step": 21515 + }, + { + "epoch": 0.47360774818401935, + "grad_norm": 0.6084206737074168, + "learning_rate": 5.668310097935724e-06, + "loss": 0.4656, + "step": 21516 + }, + { + "epoch": 0.47362976007043806, + "grad_norm": 1.5601468316485054, + "learning_rate": 5.667956838350489e-06, + "loss": 0.3314, + "step": 21517 + }, + { + "epoch": 0.4736517719568567, + "grad_norm": 1.7307662196576155, + "learning_rate": 5.6676035753703996e-06, + "loss": 0.341, + "step": 21518 + }, + { + "epoch": 0.47367378384327535, + "grad_norm": 1.4356939916755689, + "learning_rate": 5.667250308997252e-06, + "loss": 0.276, + "step": 21519 + }, + { + "epoch": 0.47369579572969406, + "grad_norm": 1.416010914551755, + "learning_rate": 5.666897039232839e-06, + "loss": 0.3329, + "step": 21520 + }, + { + "epoch": 0.4737178076161127, + "grad_norm": 1.8472368928100658, + "learning_rate": 5.66654376607896e-06, + "loss": 0.2963, + "step": 21521 + }, + { + "epoch": 0.47373981950253136, + "grad_norm": 0.6510856901123343, + "learning_rate": 5.666190489537407e-06, + "loss": 0.4923, + "step": 21522 + }, + { + "epoch": 0.47376183138895, + "grad_norm": 1.2426999118928497, + "learning_rate": 5.665837209609975e-06, + "loss": 0.2683, + "step": 21523 + }, + { + "epoch": 0.4737838432753687, + "grad_norm": 0.6200924035734628, + "learning_rate": 5.665483926298465e-06, + "loss": 0.4794, + "step": 21524 + }, + { + "epoch": 0.47380585516178736, + "grad_norm": 2.3188461335567037, + "learning_rate": 5.665130639604666e-06, + "loss": 0.3399, + "step": 21525 + }, + { + "epoch": 0.473827867048206, + "grad_norm": 2.686713256868064, + "learning_rate": 5.664777349530378e-06, + "loss": 0.3528, + "step": 21526 + }, + { + "epoch": 0.4738498789346247, + "grad_norm": 1.8441876401850388, + "learning_rate": 5.664424056077395e-06, + "loss": 0.2317, + "step": 21527 + }, + { + "epoch": 0.47387189082104336, + "grad_norm": 1.3899299126790998, + "learning_rate": 5.664070759247511e-06, + "loss": 0.3118, + "step": 21528 + }, + { + "epoch": 0.473893902707462, + "grad_norm": 1.6846463237217537, + "learning_rate": 5.663717459042526e-06, + "loss": 0.3857, + "step": 21529 + }, + { + "epoch": 0.4739159145938807, + "grad_norm": 0.6489565799228609, + "learning_rate": 5.66336415546423e-06, + "loss": 0.4729, + "step": 21530 + }, + { + "epoch": 0.47393792648029937, + "grad_norm": 0.6356649674272892, + "learning_rate": 5.663010848514423e-06, + "loss": 0.4972, + "step": 21531 + }, + { + "epoch": 0.473959938366718, + "grad_norm": 0.5979989306815685, + "learning_rate": 5.662657538194899e-06, + "loss": 0.4443, + "step": 21532 + }, + { + "epoch": 0.4739819502531367, + "grad_norm": 1.1505044069365242, + "learning_rate": 5.662304224507454e-06, + "loss": 0.2407, + "step": 21533 + }, + { + "epoch": 0.47400396213955537, + "grad_norm": 1.5335964323929743, + "learning_rate": 5.6619509074538835e-06, + "loss": 0.2868, + "step": 21534 + }, + { + "epoch": 0.474025974025974, + "grad_norm": 1.6738985770268497, + "learning_rate": 5.661597587035982e-06, + "loss": 0.3035, + "step": 21535 + }, + { + "epoch": 0.47404798591239267, + "grad_norm": 1.3888863366965052, + "learning_rate": 5.6612442632555476e-06, + "loss": 0.3371, + "step": 21536 + }, + { + "epoch": 0.47406999779881137, + "grad_norm": 2.144733250632617, + "learning_rate": 5.660890936114375e-06, + "loss": 0.3586, + "step": 21537 + }, + { + "epoch": 0.47409200968523, + "grad_norm": 1.3486072485173317, + "learning_rate": 5.66053760561426e-06, + "loss": 0.2711, + "step": 21538 + }, + { + "epoch": 0.47411402157164867, + "grad_norm": 2.5315132134924405, + "learning_rate": 5.660184271756997e-06, + "loss": 0.2878, + "step": 21539 + }, + { + "epoch": 0.4741360334580674, + "grad_norm": 1.6064324182599803, + "learning_rate": 5.659830934544385e-06, + "loss": 0.3749, + "step": 21540 + }, + { + "epoch": 0.474158045344486, + "grad_norm": 1.7550827829590963, + "learning_rate": 5.6594775939782164e-06, + "loss": 0.3615, + "step": 21541 + }, + { + "epoch": 0.4741800572309047, + "grad_norm": 1.5513192540367664, + "learning_rate": 5.6591242500602895e-06, + "loss": 0.286, + "step": 21542 + }, + { + "epoch": 0.4742020691173234, + "grad_norm": 1.5381032883364618, + "learning_rate": 5.658770902792399e-06, + "loss": 0.252, + "step": 21543 + }, + { + "epoch": 0.474224081003742, + "grad_norm": 0.6373461177103985, + "learning_rate": 5.6584175521763404e-06, + "loss": 0.4942, + "step": 21544 + }, + { + "epoch": 0.4742460928901607, + "grad_norm": 2.2567644015760266, + "learning_rate": 5.6580641982139106e-06, + "loss": 0.324, + "step": 21545 + }, + { + "epoch": 0.4742681047765794, + "grad_norm": 1.4059715847697416, + "learning_rate": 5.657710840906903e-06, + "loss": 0.38, + "step": 21546 + }, + { + "epoch": 0.47429011666299803, + "grad_norm": 1.3986132543855143, + "learning_rate": 5.6573574802571196e-06, + "loss": 0.2401, + "step": 21547 + }, + { + "epoch": 0.4743121285494167, + "grad_norm": 1.4413206741517464, + "learning_rate": 5.657004116266349e-06, + "loss": 0.2834, + "step": 21548 + }, + { + "epoch": 0.4743341404358353, + "grad_norm": 3.6826702689136295, + "learning_rate": 5.656650748936392e-06, + "loss": 0.265, + "step": 21549 + }, + { + "epoch": 0.47435615232225403, + "grad_norm": 1.5528077171453676, + "learning_rate": 5.656297378269043e-06, + "loss": 0.3011, + "step": 21550 + }, + { + "epoch": 0.4743781642086727, + "grad_norm": 1.129676597283282, + "learning_rate": 5.6559440042660966e-06, + "loss": 0.2813, + "step": 21551 + }, + { + "epoch": 0.47440017609509133, + "grad_norm": 2.0484496233298386, + "learning_rate": 5.655590626929353e-06, + "loss": 0.2784, + "step": 21552 + }, + { + "epoch": 0.47442218798151004, + "grad_norm": 1.3943748507451255, + "learning_rate": 5.655237246260602e-06, + "loss": 0.3976, + "step": 21553 + }, + { + "epoch": 0.4744441998679287, + "grad_norm": 1.3138096216492787, + "learning_rate": 5.654883862261645e-06, + "loss": 0.3066, + "step": 21554 + }, + { + "epoch": 0.47446621175434733, + "grad_norm": 1.2945249801956853, + "learning_rate": 5.6545304749342765e-06, + "loss": 0.2843, + "step": 21555 + }, + { + "epoch": 0.47448822364076604, + "grad_norm": 1.524897450672485, + "learning_rate": 5.654177084280291e-06, + "loss": 0.2959, + "step": 21556 + }, + { + "epoch": 0.4745102355271847, + "grad_norm": 1.2044058860388898, + "learning_rate": 5.653823690301488e-06, + "loss": 0.314, + "step": 21557 + }, + { + "epoch": 0.47453224741360334, + "grad_norm": 0.6400521565942855, + "learning_rate": 5.653470292999658e-06, + "loss": 0.4725, + "step": 21558 + }, + { + "epoch": 0.474554259300022, + "grad_norm": 1.2736293799701295, + "learning_rate": 5.653116892376602e-06, + "loss": 0.2601, + "step": 21559 + }, + { + "epoch": 0.4745762711864407, + "grad_norm": 1.6461934063865795, + "learning_rate": 5.652763488434115e-06, + "loss": 0.3041, + "step": 21560 + }, + { + "epoch": 0.47459828307285934, + "grad_norm": 1.394864136709364, + "learning_rate": 5.652410081173993e-06, + "loss": 0.3325, + "step": 21561 + }, + { + "epoch": 0.474620294959278, + "grad_norm": 2.71190218881159, + "learning_rate": 5.652056670598032e-06, + "loss": 0.3171, + "step": 21562 + }, + { + "epoch": 0.4746423068456967, + "grad_norm": 0.5932266418783623, + "learning_rate": 5.6517032567080285e-06, + "loss": 0.478, + "step": 21563 + }, + { + "epoch": 0.47466431873211534, + "grad_norm": 1.5003039447787236, + "learning_rate": 5.6513498395057766e-06, + "loss": 0.2973, + "step": 21564 + }, + { + "epoch": 0.474686330618534, + "grad_norm": 1.484476100584934, + "learning_rate": 5.650996418993077e-06, + "loss": 0.2977, + "step": 21565 + }, + { + "epoch": 0.4747083425049527, + "grad_norm": 1.703695481912317, + "learning_rate": 5.65064299517172e-06, + "loss": 0.3559, + "step": 21566 + }, + { + "epoch": 0.47473035439137135, + "grad_norm": 1.5043004948689984, + "learning_rate": 5.650289568043508e-06, + "loss": 0.3445, + "step": 21567 + }, + { + "epoch": 0.47475236627779, + "grad_norm": 0.6293700686712748, + "learning_rate": 5.649936137610234e-06, + "loss": 0.4755, + "step": 21568 + }, + { + "epoch": 0.4747743781642087, + "grad_norm": 1.7337373671691712, + "learning_rate": 5.649582703873691e-06, + "loss": 0.3183, + "step": 21569 + }, + { + "epoch": 0.47479639005062735, + "grad_norm": 1.5503236620384988, + "learning_rate": 5.649229266835684e-06, + "loss": 0.3156, + "step": 21570 + }, + { + "epoch": 0.474818401937046, + "grad_norm": 1.4792852591071577, + "learning_rate": 5.648875826498002e-06, + "loss": 0.3038, + "step": 21571 + }, + { + "epoch": 0.47484041382346465, + "grad_norm": 1.3038602346380908, + "learning_rate": 5.648522382862443e-06, + "loss": 0.2895, + "step": 21572 + }, + { + "epoch": 0.47486242570988335, + "grad_norm": 1.416900920345485, + "learning_rate": 5.648168935930805e-06, + "loss": 0.272, + "step": 21573 + }, + { + "epoch": 0.474884437596302, + "grad_norm": 1.9533010195740563, + "learning_rate": 5.6478154857048836e-06, + "loss": 0.3112, + "step": 21574 + }, + { + "epoch": 0.47490644948272065, + "grad_norm": 2.638345100100277, + "learning_rate": 5.647462032186475e-06, + "loss": 0.4384, + "step": 21575 + }, + { + "epoch": 0.47492846136913935, + "grad_norm": 1.2759063845542393, + "learning_rate": 5.6471085753773735e-06, + "loss": 0.3109, + "step": 21576 + }, + { + "epoch": 0.474950473255558, + "grad_norm": 1.4894780810477777, + "learning_rate": 5.64675511527938e-06, + "loss": 0.2715, + "step": 21577 + }, + { + "epoch": 0.47497248514197665, + "grad_norm": 1.4248462716562116, + "learning_rate": 5.646401651894288e-06, + "loss": 0.2784, + "step": 21578 + }, + { + "epoch": 0.47499449702839536, + "grad_norm": 0.6222578081686078, + "learning_rate": 5.646048185223893e-06, + "loss": 0.4949, + "step": 21579 + }, + { + "epoch": 0.475016508914814, + "grad_norm": 1.5588452508078134, + "learning_rate": 5.6456947152699956e-06, + "loss": 0.3454, + "step": 21580 + }, + { + "epoch": 0.47503852080123266, + "grad_norm": 1.4320253743721363, + "learning_rate": 5.645341242034388e-06, + "loss": 0.3329, + "step": 21581 + }, + { + "epoch": 0.4750605326876513, + "grad_norm": 1.7887394587936467, + "learning_rate": 5.6449877655188675e-06, + "loss": 0.3671, + "step": 21582 + }, + { + "epoch": 0.47508254457407, + "grad_norm": 0.605480052932262, + "learning_rate": 5.6446342857252346e-06, + "loss": 0.4733, + "step": 21583 + }, + { + "epoch": 0.47510455646048866, + "grad_norm": 1.5818992310090374, + "learning_rate": 5.64428080265528e-06, + "loss": 0.3434, + "step": 21584 + }, + { + "epoch": 0.4751265683469073, + "grad_norm": 1.5654513944123936, + "learning_rate": 5.643927316310804e-06, + "loss": 0.3373, + "step": 21585 + }, + { + "epoch": 0.475148580233326, + "grad_norm": 0.6232269043570087, + "learning_rate": 5.6435738266936024e-06, + "loss": 0.453, + "step": 21586 + }, + { + "epoch": 0.47517059211974466, + "grad_norm": 1.3319840056013261, + "learning_rate": 5.64322033380547e-06, + "loss": 0.397, + "step": 21587 + }, + { + "epoch": 0.4751926040061633, + "grad_norm": 1.1859779379260929, + "learning_rate": 5.642866837648208e-06, + "loss": 0.2681, + "step": 21588 + }, + { + "epoch": 0.475214615892582, + "grad_norm": 1.4656385487582604, + "learning_rate": 5.642513338223606e-06, + "loss": 0.2929, + "step": 21589 + }, + { + "epoch": 0.47523662777900066, + "grad_norm": 1.977965204418965, + "learning_rate": 5.642159835533467e-06, + "loss": 0.2084, + "step": 21590 + }, + { + "epoch": 0.4752586396654193, + "grad_norm": 1.3345973900708203, + "learning_rate": 5.641806329579585e-06, + "loss": 0.2922, + "step": 21591 + }, + { + "epoch": 0.475280651551838, + "grad_norm": 0.6059817665174363, + "learning_rate": 5.641452820363758e-06, + "loss": 0.4873, + "step": 21592 + }, + { + "epoch": 0.47530266343825667, + "grad_norm": 1.4358092124206923, + "learning_rate": 5.6410993078877805e-06, + "loss": 0.3365, + "step": 21593 + }, + { + "epoch": 0.4753246753246753, + "grad_norm": 1.2702128593195048, + "learning_rate": 5.64074579215345e-06, + "loss": 0.3041, + "step": 21594 + }, + { + "epoch": 0.47534668721109397, + "grad_norm": 0.6255004756746433, + "learning_rate": 5.640392273162565e-06, + "loss": 0.4755, + "step": 21595 + }, + { + "epoch": 0.47536869909751267, + "grad_norm": 1.3807372577324413, + "learning_rate": 5.6400387509169195e-06, + "loss": 0.3665, + "step": 21596 + }, + { + "epoch": 0.4753907109839313, + "grad_norm": 1.6793962686598671, + "learning_rate": 5.639685225418313e-06, + "loss": 0.2779, + "step": 21597 + }, + { + "epoch": 0.47541272287034997, + "grad_norm": 1.6775538020642369, + "learning_rate": 5.6393316966685395e-06, + "loss": 0.2542, + "step": 21598 + }, + { + "epoch": 0.4754347347567687, + "grad_norm": 1.2004730568842037, + "learning_rate": 5.638978164669399e-06, + "loss": 0.3234, + "step": 21599 + }, + { + "epoch": 0.4754567466431873, + "grad_norm": 1.7563664640556926, + "learning_rate": 5.638624629422685e-06, + "loss": 0.3527, + "step": 21600 + }, + { + "epoch": 0.47547875852960597, + "grad_norm": 1.4344978255250473, + "learning_rate": 5.638271090930195e-06, + "loss": 0.302, + "step": 21601 + }, + { + "epoch": 0.4755007704160247, + "grad_norm": 0.6254411721714922, + "learning_rate": 5.637917549193727e-06, + "loss": 0.4991, + "step": 21602 + }, + { + "epoch": 0.4755227823024433, + "grad_norm": 1.2017791911860882, + "learning_rate": 5.637564004215079e-06, + "loss": 0.324, + "step": 21603 + }, + { + "epoch": 0.475544794188862, + "grad_norm": 1.8398656149635577, + "learning_rate": 5.637210455996046e-06, + "loss": 0.2965, + "step": 21604 + }, + { + "epoch": 0.4755668060752807, + "grad_norm": 0.6245398393042287, + "learning_rate": 5.636856904538424e-06, + "loss": 0.482, + "step": 21605 + }, + { + "epoch": 0.47558881796169933, + "grad_norm": 1.3553948312303514, + "learning_rate": 5.636503349844014e-06, + "loss": 0.3345, + "step": 21606 + }, + { + "epoch": 0.475610829848118, + "grad_norm": 1.4041835142664312, + "learning_rate": 5.636149791914606e-06, + "loss": 0.3413, + "step": 21607 + }, + { + "epoch": 0.4756328417345366, + "grad_norm": 0.6245605063116991, + "learning_rate": 5.635796230752003e-06, + "loss": 0.4832, + "step": 21608 + }, + { + "epoch": 0.47565485362095533, + "grad_norm": 1.319748005200155, + "learning_rate": 5.635442666358001e-06, + "loss": 0.2734, + "step": 21609 + }, + { + "epoch": 0.475676865507374, + "grad_norm": 1.5330795202471685, + "learning_rate": 5.635089098734394e-06, + "loss": 0.2983, + "step": 21610 + }, + { + "epoch": 0.47569887739379263, + "grad_norm": 0.6192032963427814, + "learning_rate": 5.634735527882983e-06, + "loss": 0.4705, + "step": 21611 + }, + { + "epoch": 0.47572088928021133, + "grad_norm": 0.6266270515688761, + "learning_rate": 5.634381953805561e-06, + "loss": 0.4943, + "step": 21612 + }, + { + "epoch": 0.47574290116663, + "grad_norm": 1.2050379210926963, + "learning_rate": 5.634028376503928e-06, + "loss": 0.21, + "step": 21613 + }, + { + "epoch": 0.47576491305304863, + "grad_norm": 1.4969772587973504, + "learning_rate": 5.63367479597988e-06, + "loss": 0.307, + "step": 21614 + }, + { + "epoch": 0.47578692493946734, + "grad_norm": 1.4389010149295747, + "learning_rate": 5.6333212122352145e-06, + "loss": 0.352, + "step": 21615 + }, + { + "epoch": 0.475808936825886, + "grad_norm": 1.4250720533323389, + "learning_rate": 5.632967625271727e-06, + "loss": 0.283, + "step": 21616 + }, + { + "epoch": 0.47583094871230464, + "grad_norm": 0.6156799126988457, + "learning_rate": 5.632614035091217e-06, + "loss": 0.4555, + "step": 21617 + }, + { + "epoch": 0.4758529605987233, + "grad_norm": 0.592813648261131, + "learning_rate": 5.632260441695479e-06, + "loss": 0.4611, + "step": 21618 + }, + { + "epoch": 0.475874972485142, + "grad_norm": 1.2154372126318793, + "learning_rate": 5.631906845086311e-06, + "loss": 0.2865, + "step": 21619 + }, + { + "epoch": 0.47589698437156064, + "grad_norm": 1.1779772695934763, + "learning_rate": 5.631553245265512e-06, + "loss": 0.2777, + "step": 21620 + }, + { + "epoch": 0.4759189962579793, + "grad_norm": 0.6231964743194458, + "learning_rate": 5.631199642234878e-06, + "loss": 0.4659, + "step": 21621 + }, + { + "epoch": 0.475941008144398, + "grad_norm": 1.2925351476003737, + "learning_rate": 5.6308460359962045e-06, + "loss": 0.2265, + "step": 21622 + }, + { + "epoch": 0.47596302003081664, + "grad_norm": 2.403356011420762, + "learning_rate": 5.6304924265512905e-06, + "loss": 0.449, + "step": 21623 + }, + { + "epoch": 0.4759850319172353, + "grad_norm": 1.6947637870630543, + "learning_rate": 5.630138813901934e-06, + "loss": 0.3296, + "step": 21624 + }, + { + "epoch": 0.476007043803654, + "grad_norm": 1.4201812340442175, + "learning_rate": 5.62978519804993e-06, + "loss": 0.2591, + "step": 21625 + }, + { + "epoch": 0.47602905569007264, + "grad_norm": 1.7630791315448358, + "learning_rate": 5.629431578997076e-06, + "loss": 0.2838, + "step": 21626 + }, + { + "epoch": 0.4760510675764913, + "grad_norm": 1.41179484389838, + "learning_rate": 5.629077956745171e-06, + "loss": 0.3331, + "step": 21627 + }, + { + "epoch": 0.47607307946291, + "grad_norm": 0.6311687573951701, + "learning_rate": 5.62872433129601e-06, + "loss": 0.4753, + "step": 21628 + }, + { + "epoch": 0.47609509134932865, + "grad_norm": 1.7074255522296715, + "learning_rate": 5.628370702651395e-06, + "loss": 0.3123, + "step": 21629 + }, + { + "epoch": 0.4761171032357473, + "grad_norm": 1.2544451936966354, + "learning_rate": 5.6280170708131165e-06, + "loss": 0.2231, + "step": 21630 + }, + { + "epoch": 0.47613911512216595, + "grad_norm": 0.6210960692434377, + "learning_rate": 5.627663435782975e-06, + "loss": 0.4738, + "step": 21631 + }, + { + "epoch": 0.47616112700858465, + "grad_norm": 1.624829123625041, + "learning_rate": 5.62730979756277e-06, + "loss": 0.2734, + "step": 21632 + }, + { + "epoch": 0.4761831388950033, + "grad_norm": 1.2519196021115233, + "learning_rate": 5.626956156154297e-06, + "loss": 0.2832, + "step": 21633 + }, + { + "epoch": 0.47620515078142195, + "grad_norm": 1.3602947875630633, + "learning_rate": 5.626602511559352e-06, + "loss": 0.3169, + "step": 21634 + }, + { + "epoch": 0.47622716266784065, + "grad_norm": 1.1194233711255221, + "learning_rate": 5.626248863779733e-06, + "loss": 0.2635, + "step": 21635 + }, + { + "epoch": 0.4762491745542593, + "grad_norm": 1.2762892233501466, + "learning_rate": 5.6258952128172375e-06, + "loss": 0.2993, + "step": 21636 + }, + { + "epoch": 0.47627118644067795, + "grad_norm": 1.459566678129946, + "learning_rate": 5.625541558673665e-06, + "loss": 0.3214, + "step": 21637 + }, + { + "epoch": 0.47629319832709666, + "grad_norm": 1.2781933230306204, + "learning_rate": 5.6251879013508106e-06, + "loss": 0.2351, + "step": 21638 + }, + { + "epoch": 0.4763152102135153, + "grad_norm": 1.3276693810481595, + "learning_rate": 5.624834240850473e-06, + "loss": 0.2301, + "step": 21639 + }, + { + "epoch": 0.47633722209993395, + "grad_norm": 0.6346407473131016, + "learning_rate": 5.624480577174448e-06, + "loss": 0.4798, + "step": 21640 + }, + { + "epoch": 0.4763592339863526, + "grad_norm": 1.3166815365046431, + "learning_rate": 5.624126910324534e-06, + "loss": 0.2859, + "step": 21641 + }, + { + "epoch": 0.4763812458727713, + "grad_norm": 1.862266505624262, + "learning_rate": 5.62377324030253e-06, + "loss": 0.3026, + "step": 21642 + }, + { + "epoch": 0.47640325775918996, + "grad_norm": 0.6206918772220312, + "learning_rate": 5.623419567110232e-06, + "loss": 0.4728, + "step": 21643 + }, + { + "epoch": 0.4764252696456086, + "grad_norm": 1.505499146585196, + "learning_rate": 5.623065890749437e-06, + "loss": 0.2741, + "step": 21644 + }, + { + "epoch": 0.4764472815320273, + "grad_norm": 2.71290505902497, + "learning_rate": 5.6227122112219446e-06, + "loss": 0.4514, + "step": 21645 + }, + { + "epoch": 0.47646929341844596, + "grad_norm": 1.5263586666432274, + "learning_rate": 5.622358528529549e-06, + "loss": 0.3389, + "step": 21646 + }, + { + "epoch": 0.4764913053048646, + "grad_norm": 1.2708751184561786, + "learning_rate": 5.622004842674051e-06, + "loss": 0.2475, + "step": 21647 + }, + { + "epoch": 0.4765133171912833, + "grad_norm": 1.3662576040777281, + "learning_rate": 5.621651153657247e-06, + "loss": 0.3204, + "step": 21648 + }, + { + "epoch": 0.47653532907770196, + "grad_norm": 1.5681820835025866, + "learning_rate": 5.621297461480934e-06, + "loss": 0.2566, + "step": 21649 + }, + { + "epoch": 0.4765573409641206, + "grad_norm": 1.081303443004666, + "learning_rate": 5.6209437661469115e-06, + "loss": 0.3021, + "step": 21650 + }, + { + "epoch": 0.4765793528505393, + "grad_norm": 1.8896098646365806, + "learning_rate": 5.620590067656974e-06, + "loss": 0.3319, + "step": 21651 + }, + { + "epoch": 0.47660136473695797, + "grad_norm": 1.1953785651924742, + "learning_rate": 5.620236366012923e-06, + "loss": 0.3206, + "step": 21652 + }, + { + "epoch": 0.4766233766233766, + "grad_norm": 1.2949805363632039, + "learning_rate": 5.619882661216553e-06, + "loss": 0.2596, + "step": 21653 + }, + { + "epoch": 0.47664538850979526, + "grad_norm": 0.6575948290730446, + "learning_rate": 5.619528953269662e-06, + "loss": 0.4927, + "step": 21654 + }, + { + "epoch": 0.47666740039621397, + "grad_norm": 5.527738167191839, + "learning_rate": 5.61917524217405e-06, + "loss": 0.3097, + "step": 21655 + }, + { + "epoch": 0.4766894122826326, + "grad_norm": 1.142613575541282, + "learning_rate": 5.618821527931513e-06, + "loss": 0.2547, + "step": 21656 + }, + { + "epoch": 0.47671142416905127, + "grad_norm": 2.502151865086824, + "learning_rate": 5.618467810543848e-06, + "loss": 0.2291, + "step": 21657 + }, + { + "epoch": 0.47673343605547, + "grad_norm": 1.1549133954393316, + "learning_rate": 5.618114090012855e-06, + "loss": 0.261, + "step": 21658 + }, + { + "epoch": 0.4767554479418886, + "grad_norm": 0.6399403605633943, + "learning_rate": 5.617760366340329e-06, + "loss": 0.4811, + "step": 21659 + }, + { + "epoch": 0.47677745982830727, + "grad_norm": 1.3090427639216693, + "learning_rate": 5.61740663952807e-06, + "loss": 0.2724, + "step": 21660 + }, + { + "epoch": 0.476799471714726, + "grad_norm": 1.2857550531565478, + "learning_rate": 5.617052909577876e-06, + "loss": 0.3234, + "step": 21661 + }, + { + "epoch": 0.4768214836011446, + "grad_norm": 3.8040593406791974, + "learning_rate": 5.616699176491544e-06, + "loss": 0.2992, + "step": 21662 + }, + { + "epoch": 0.4768434954875633, + "grad_norm": 0.6259067655679649, + "learning_rate": 5.6163454402708705e-06, + "loss": 0.4996, + "step": 21663 + }, + { + "epoch": 0.4768655073739819, + "grad_norm": 1.2314057253259096, + "learning_rate": 5.6159917009176535e-06, + "loss": 0.3832, + "step": 21664 + }, + { + "epoch": 0.4768875192604006, + "grad_norm": 1.7732997571972886, + "learning_rate": 5.615637958433695e-06, + "loss": 0.3797, + "step": 21665 + }, + { + "epoch": 0.4769095311468193, + "grad_norm": 1.971356111368387, + "learning_rate": 5.615284212820788e-06, + "loss": 0.271, + "step": 21666 + }, + { + "epoch": 0.4769315430332379, + "grad_norm": 1.8091568601031818, + "learning_rate": 5.614930464080732e-06, + "loss": 0.3099, + "step": 21667 + }, + { + "epoch": 0.47695355491965663, + "grad_norm": 1.5160209036870136, + "learning_rate": 5.614576712215326e-06, + "loss": 0.3216, + "step": 21668 + }, + { + "epoch": 0.4769755668060753, + "grad_norm": 1.8384231782576546, + "learning_rate": 5.614222957226365e-06, + "loss": 0.3436, + "step": 21669 + }, + { + "epoch": 0.47699757869249393, + "grad_norm": 1.4562827259199091, + "learning_rate": 5.6138691991156525e-06, + "loss": 0.2746, + "step": 21670 + }, + { + "epoch": 0.47701959057891263, + "grad_norm": 2.3396743770242825, + "learning_rate": 5.613515437884981e-06, + "loss": 0.3187, + "step": 21671 + }, + { + "epoch": 0.4770416024653313, + "grad_norm": 3.1966023372440997, + "learning_rate": 5.613161673536149e-06, + "loss": 0.3031, + "step": 21672 + }, + { + "epoch": 0.47706361435174993, + "grad_norm": 1.5794848084655917, + "learning_rate": 5.612807906070958e-06, + "loss": 0.3321, + "step": 21673 + }, + { + "epoch": 0.47708562623816864, + "grad_norm": 1.2537954047140045, + "learning_rate": 5.612454135491203e-06, + "loss": 0.3118, + "step": 21674 + }, + { + "epoch": 0.4771076381245873, + "grad_norm": 1.0963977353826184, + "learning_rate": 5.612100361798683e-06, + "loss": 0.2208, + "step": 21675 + }, + { + "epoch": 0.47712965001100593, + "grad_norm": 0.6328762445328417, + "learning_rate": 5.611746584995197e-06, + "loss": 0.485, + "step": 21676 + }, + { + "epoch": 0.4771516618974246, + "grad_norm": 2.7368173800506406, + "learning_rate": 5.61139280508254e-06, + "loss": 0.281, + "step": 21677 + }, + { + "epoch": 0.4771736737838433, + "grad_norm": 5.357293461322258, + "learning_rate": 5.611039022062514e-06, + "loss": 0.2772, + "step": 21678 + }, + { + "epoch": 0.47719568567026194, + "grad_norm": 2.6407565121048746, + "learning_rate": 5.610685235936914e-06, + "loss": 0.3821, + "step": 21679 + }, + { + "epoch": 0.4772176975566806, + "grad_norm": 1.4563438031831932, + "learning_rate": 5.6103314467075396e-06, + "loss": 0.365, + "step": 21680 + }, + { + "epoch": 0.4772397094430993, + "grad_norm": 2.2391513056632384, + "learning_rate": 5.6099776543761895e-06, + "loss": 0.2483, + "step": 21681 + }, + { + "epoch": 0.47726172132951794, + "grad_norm": 1.448924512013447, + "learning_rate": 5.60962385894466e-06, + "loss": 0.3182, + "step": 21682 + }, + { + "epoch": 0.4772837332159366, + "grad_norm": 1.296912750725357, + "learning_rate": 5.60927006041475e-06, + "loss": 0.263, + "step": 21683 + }, + { + "epoch": 0.4773057451023553, + "grad_norm": 0.6214863303066286, + "learning_rate": 5.6089162587882574e-06, + "loss": 0.4753, + "step": 21684 + }, + { + "epoch": 0.47732775698877394, + "grad_norm": 1.857046231948373, + "learning_rate": 5.608562454066981e-06, + "loss": 0.2657, + "step": 21685 + }, + { + "epoch": 0.4773497688751926, + "grad_norm": 0.6279105888554557, + "learning_rate": 5.60820864625272e-06, + "loss": 0.5059, + "step": 21686 + }, + { + "epoch": 0.4773717807616113, + "grad_norm": 1.1529622158415753, + "learning_rate": 5.60785483534727e-06, + "loss": 0.3207, + "step": 21687 + }, + { + "epoch": 0.47739379264802995, + "grad_norm": 1.9321020571886203, + "learning_rate": 5.607501021352432e-06, + "loss": 0.3004, + "step": 21688 + }, + { + "epoch": 0.4774158045344486, + "grad_norm": 1.5608343851163313, + "learning_rate": 5.607147204270002e-06, + "loss": 0.3456, + "step": 21689 + }, + { + "epoch": 0.47743781642086724, + "grad_norm": 1.5672855317029066, + "learning_rate": 5.606793384101779e-06, + "loss": 0.3232, + "step": 21690 + }, + { + "epoch": 0.47745982830728595, + "grad_norm": 1.5344816952455556, + "learning_rate": 5.60643956084956e-06, + "loss": 0.3924, + "step": 21691 + }, + { + "epoch": 0.4774818401937046, + "grad_norm": 0.629102077713213, + "learning_rate": 5.606085734515147e-06, + "loss": 0.4898, + "step": 21692 + }, + { + "epoch": 0.47750385208012325, + "grad_norm": 1.5420110478186018, + "learning_rate": 5.605731905100336e-06, + "loss": 0.4313, + "step": 21693 + }, + { + "epoch": 0.47752586396654195, + "grad_norm": 1.6460090551837658, + "learning_rate": 5.605378072606924e-06, + "loss": 0.3252, + "step": 21694 + }, + { + "epoch": 0.4775478758529606, + "grad_norm": 1.6142538618554074, + "learning_rate": 5.605024237036711e-06, + "loss": 0.3072, + "step": 21695 + }, + { + "epoch": 0.47756988773937925, + "grad_norm": 1.5499473045974514, + "learning_rate": 5.604670398391494e-06, + "loss": 0.299, + "step": 21696 + }, + { + "epoch": 0.47759189962579796, + "grad_norm": 2.3385880805131904, + "learning_rate": 5.604316556673074e-06, + "loss": 0.3657, + "step": 21697 + }, + { + "epoch": 0.4776139115122166, + "grad_norm": 1.8312620287450176, + "learning_rate": 5.603962711883247e-06, + "loss": 0.3437, + "step": 21698 + }, + { + "epoch": 0.47763592339863525, + "grad_norm": 0.5908218024418728, + "learning_rate": 5.603608864023813e-06, + "loss": 0.4667, + "step": 21699 + }, + { + "epoch": 0.4776579352850539, + "grad_norm": 2.118464140423785, + "learning_rate": 5.603255013096568e-06, + "loss": 0.2868, + "step": 21700 + }, + { + "epoch": 0.4776799471714726, + "grad_norm": 1.568473299044405, + "learning_rate": 5.6029011591033124e-06, + "loss": 0.3664, + "step": 21701 + }, + { + "epoch": 0.47770195905789126, + "grad_norm": 1.7436278992114256, + "learning_rate": 5.602547302045844e-06, + "loss": 0.295, + "step": 21702 + }, + { + "epoch": 0.4777239709443099, + "grad_norm": 1.7504501467957412, + "learning_rate": 5.602193441925963e-06, + "loss": 0.2249, + "step": 21703 + }, + { + "epoch": 0.4777459828307286, + "grad_norm": 0.5909839571214158, + "learning_rate": 5.601839578745465e-06, + "loss": 0.4591, + "step": 21704 + }, + { + "epoch": 0.47776799471714726, + "grad_norm": 1.688622843741272, + "learning_rate": 5.601485712506149e-06, + "loss": 0.2573, + "step": 21705 + }, + { + "epoch": 0.4777900066035659, + "grad_norm": 1.5666773799802642, + "learning_rate": 5.601131843209817e-06, + "loss": 0.3442, + "step": 21706 + }, + { + "epoch": 0.4778120184899846, + "grad_norm": 1.3153065414898948, + "learning_rate": 5.600777970858265e-06, + "loss": 0.2966, + "step": 21707 + }, + { + "epoch": 0.47783403037640326, + "grad_norm": 1.2921378299956745, + "learning_rate": 5.6004240954532885e-06, + "loss": 0.3047, + "step": 21708 + }, + { + "epoch": 0.4778560422628219, + "grad_norm": 1.3582124137973144, + "learning_rate": 5.6000702169966905e-06, + "loss": 0.2939, + "step": 21709 + }, + { + "epoch": 0.4778780541492406, + "grad_norm": 0.5812497178433603, + "learning_rate": 5.599716335490267e-06, + "loss": 0.4676, + "step": 21710 + }, + { + "epoch": 0.47790006603565927, + "grad_norm": 1.4902251939876912, + "learning_rate": 5.599362450935819e-06, + "loss": 0.3654, + "step": 21711 + }, + { + "epoch": 0.4779220779220779, + "grad_norm": 2.5400435814903317, + "learning_rate": 5.599008563335143e-06, + "loss": 0.2868, + "step": 21712 + }, + { + "epoch": 0.47794408980849656, + "grad_norm": 2.5467006250739255, + "learning_rate": 5.598654672690037e-06, + "loss": 0.3967, + "step": 21713 + }, + { + "epoch": 0.47796610169491527, + "grad_norm": 0.6156767091821588, + "learning_rate": 5.598300779002303e-06, + "loss": 0.4861, + "step": 21714 + }, + { + "epoch": 0.4779881135813339, + "grad_norm": 1.3061128664890158, + "learning_rate": 5.597946882273737e-06, + "loss": 0.2512, + "step": 21715 + }, + { + "epoch": 0.47801012546775257, + "grad_norm": 1.4544621179919213, + "learning_rate": 5.597592982506138e-06, + "loss": 0.2563, + "step": 21716 + }, + { + "epoch": 0.47803213735417127, + "grad_norm": 2.796236429824334, + "learning_rate": 5.5972390797013055e-06, + "loss": 0.365, + "step": 21717 + }, + { + "epoch": 0.4780541492405899, + "grad_norm": 1.3049011304336093, + "learning_rate": 5.596885173861036e-06, + "loss": 0.2663, + "step": 21718 + }, + { + "epoch": 0.47807616112700857, + "grad_norm": 1.3267250296425532, + "learning_rate": 5.596531264987132e-06, + "loss": 0.3079, + "step": 21719 + }, + { + "epoch": 0.4780981730134273, + "grad_norm": 1.82412556986074, + "learning_rate": 5.596177353081389e-06, + "loss": 0.3906, + "step": 21720 + }, + { + "epoch": 0.4781201848998459, + "grad_norm": 2.256718207334771, + "learning_rate": 5.595823438145607e-06, + "loss": 0.3214, + "step": 21721 + }, + { + "epoch": 0.4781421967862646, + "grad_norm": 1.4043570908631902, + "learning_rate": 5.595469520181585e-06, + "loss": 0.2867, + "step": 21722 + }, + { + "epoch": 0.4781642086726832, + "grad_norm": 1.7304131857009852, + "learning_rate": 5.595115599191119e-06, + "loss": 0.2437, + "step": 21723 + }, + { + "epoch": 0.4781862205591019, + "grad_norm": 1.6369917155345124, + "learning_rate": 5.594761675176014e-06, + "loss": 0.3032, + "step": 21724 + }, + { + "epoch": 0.4782082324455206, + "grad_norm": 1.292375798395104, + "learning_rate": 5.594407748138062e-06, + "loss": 0.3379, + "step": 21725 + }, + { + "epoch": 0.4782302443319392, + "grad_norm": 1.3028319865073044, + "learning_rate": 5.594053818079065e-06, + "loss": 0.2646, + "step": 21726 + }, + { + "epoch": 0.47825225621835793, + "grad_norm": 1.4302087532838887, + "learning_rate": 5.5936998850008214e-06, + "loss": 0.2706, + "step": 21727 + }, + { + "epoch": 0.4782742681047766, + "grad_norm": 0.6217477776956519, + "learning_rate": 5.5933459489051315e-06, + "loss": 0.4719, + "step": 21728 + }, + { + "epoch": 0.47829627999119523, + "grad_norm": 1.3655937575258272, + "learning_rate": 5.5929920097937925e-06, + "loss": 0.2874, + "step": 21729 + }, + { + "epoch": 0.47831829187761393, + "grad_norm": 0.617874987505202, + "learning_rate": 5.5926380676686035e-06, + "loss": 0.4654, + "step": 21730 + }, + { + "epoch": 0.4783403037640326, + "grad_norm": 0.6063825389488782, + "learning_rate": 5.592284122531362e-06, + "loss": 0.4401, + "step": 21731 + }, + { + "epoch": 0.47836231565045123, + "grad_norm": 1.2300415610654085, + "learning_rate": 5.59193017438387e-06, + "loss": 0.2086, + "step": 21732 + }, + { + "epoch": 0.47838432753686994, + "grad_norm": 1.329707347321106, + "learning_rate": 5.5915762232279245e-06, + "loss": 0.2775, + "step": 21733 + }, + { + "epoch": 0.4784063394232886, + "grad_norm": 1.8572858676163222, + "learning_rate": 5.591222269065325e-06, + "loss": 0.3236, + "step": 21734 + }, + { + "epoch": 0.47842835130970723, + "grad_norm": 1.391404872553924, + "learning_rate": 5.59086831189787e-06, + "loss": 0.3599, + "step": 21735 + }, + { + "epoch": 0.4784503631961259, + "grad_norm": 2.760149425087843, + "learning_rate": 5.590514351727358e-06, + "loss": 0.2818, + "step": 21736 + }, + { + "epoch": 0.4784723750825446, + "grad_norm": 1.557396543418795, + "learning_rate": 5.59016038855559e-06, + "loss": 0.2982, + "step": 21737 + }, + { + "epoch": 0.47849438696896324, + "grad_norm": 2.436681631539601, + "learning_rate": 5.589806422384363e-06, + "loss": 0.2887, + "step": 21738 + }, + { + "epoch": 0.4785163988553819, + "grad_norm": 1.3098274085476178, + "learning_rate": 5.589452453215477e-06, + "loss": 0.3018, + "step": 21739 + }, + { + "epoch": 0.4785384107418006, + "grad_norm": 0.654332189595303, + "learning_rate": 5.589098481050731e-06, + "loss": 0.476, + "step": 21740 + }, + { + "epoch": 0.47856042262821924, + "grad_norm": 1.336756290818647, + "learning_rate": 5.588744505891922e-06, + "loss": 0.2946, + "step": 21741 + }, + { + "epoch": 0.4785824345146379, + "grad_norm": 1.9856341046715522, + "learning_rate": 5.588390527740854e-06, + "loss": 0.3344, + "step": 21742 + }, + { + "epoch": 0.4786044464010566, + "grad_norm": 1.8809840832482432, + "learning_rate": 5.588036546599321e-06, + "loss": 0.2747, + "step": 21743 + }, + { + "epoch": 0.47862645828747524, + "grad_norm": 1.161241360062317, + "learning_rate": 5.587682562469123e-06, + "loss": 0.2519, + "step": 21744 + }, + { + "epoch": 0.4786484701738939, + "grad_norm": 2.559363069888653, + "learning_rate": 5.5873285753520615e-06, + "loss": 0.3148, + "step": 21745 + }, + { + "epoch": 0.4786704820603126, + "grad_norm": 1.2808090993577546, + "learning_rate": 5.586974585249934e-06, + "loss": 0.3378, + "step": 21746 + }, + { + "epoch": 0.47869249394673125, + "grad_norm": 2.3801231237488243, + "learning_rate": 5.58662059216454e-06, + "loss": 0.4394, + "step": 21747 + }, + { + "epoch": 0.4787145058331499, + "grad_norm": 0.6424439558697265, + "learning_rate": 5.586266596097678e-06, + "loss": 0.4917, + "step": 21748 + }, + { + "epoch": 0.47873651771956854, + "grad_norm": 1.537973534854353, + "learning_rate": 5.585912597051147e-06, + "loss": 0.4291, + "step": 21749 + }, + { + "epoch": 0.47875852960598725, + "grad_norm": 1.1698300274943134, + "learning_rate": 5.585558595026748e-06, + "loss": 0.2731, + "step": 21750 + }, + { + "epoch": 0.4787805414924059, + "grad_norm": 1.5507936525661514, + "learning_rate": 5.5852045900262794e-06, + "loss": 0.3529, + "step": 21751 + }, + { + "epoch": 0.47880255337882455, + "grad_norm": 1.4094424018029712, + "learning_rate": 5.58485058205154e-06, + "loss": 0.2365, + "step": 21752 + }, + { + "epoch": 0.47882456526524325, + "grad_norm": 5.9478541134189395, + "learning_rate": 5.584496571104328e-06, + "loss": 0.3377, + "step": 21753 + }, + { + "epoch": 0.4788465771516619, + "grad_norm": 4.107745260564755, + "learning_rate": 5.5841425571864435e-06, + "loss": 0.383, + "step": 21754 + }, + { + "epoch": 0.47886858903808055, + "grad_norm": 1.186205630173608, + "learning_rate": 5.583788540299687e-06, + "loss": 0.3094, + "step": 21755 + }, + { + "epoch": 0.47889060092449925, + "grad_norm": 1.8352306716377378, + "learning_rate": 5.5834345204458574e-06, + "loss": 0.2519, + "step": 21756 + }, + { + "epoch": 0.4789126128109179, + "grad_norm": 1.410942513756032, + "learning_rate": 5.5830804976267525e-06, + "loss": 0.3508, + "step": 21757 + }, + { + "epoch": 0.47893462469733655, + "grad_norm": 1.4050053800573699, + "learning_rate": 5.5827264718441724e-06, + "loss": 0.2941, + "step": 21758 + }, + { + "epoch": 0.4789566365837552, + "grad_norm": 1.32531539608953, + "learning_rate": 5.582372443099916e-06, + "loss": 0.3316, + "step": 21759 + }, + { + "epoch": 0.4789786484701739, + "grad_norm": 1.3058485141660143, + "learning_rate": 5.582018411395783e-06, + "loss": 0.2541, + "step": 21760 + }, + { + "epoch": 0.47900066035659256, + "grad_norm": 1.8246967284596531, + "learning_rate": 5.581664376733575e-06, + "loss": 0.3058, + "step": 21761 + }, + { + "epoch": 0.4790226722430112, + "grad_norm": 1.2970037051386851, + "learning_rate": 5.581310339115087e-06, + "loss": 0.2995, + "step": 21762 + }, + { + "epoch": 0.4790446841294299, + "grad_norm": 1.2531648169360528, + "learning_rate": 5.5809562985421215e-06, + "loss": 0.3122, + "step": 21763 + }, + { + "epoch": 0.47906669601584856, + "grad_norm": 1.7328708561564754, + "learning_rate": 5.580602255016477e-06, + "loss": 0.277, + "step": 21764 + }, + { + "epoch": 0.4790887079022672, + "grad_norm": 0.640153831529318, + "learning_rate": 5.580248208539954e-06, + "loss": 0.4867, + "step": 21765 + }, + { + "epoch": 0.4791107197886859, + "grad_norm": 0.6445662126248977, + "learning_rate": 5.579894159114349e-06, + "loss": 0.4599, + "step": 21766 + }, + { + "epoch": 0.47913273167510456, + "grad_norm": 1.3622205327311452, + "learning_rate": 5.579540106741462e-06, + "loss": 0.2824, + "step": 21767 + }, + { + "epoch": 0.4791547435615232, + "grad_norm": 1.4161698140313568, + "learning_rate": 5.579186051423096e-06, + "loss": 0.3385, + "step": 21768 + }, + { + "epoch": 0.4791767554479419, + "grad_norm": 1.6027832094881105, + "learning_rate": 5.578831993161048e-06, + "loss": 0.4278, + "step": 21769 + }, + { + "epoch": 0.47919876733436056, + "grad_norm": 0.6143281229705372, + "learning_rate": 5.578477931957117e-06, + "loss": 0.4901, + "step": 21770 + }, + { + "epoch": 0.4792207792207792, + "grad_norm": 2.2675118373327137, + "learning_rate": 5.578123867813104e-06, + "loss": 0.404, + "step": 21771 + }, + { + "epoch": 0.47924279110719786, + "grad_norm": 2.58773361276895, + "learning_rate": 5.5777698007308055e-06, + "loss": 0.3309, + "step": 21772 + }, + { + "epoch": 0.47926480299361657, + "grad_norm": 1.327387382624431, + "learning_rate": 5.577415730712025e-06, + "loss": 0.3804, + "step": 21773 + }, + { + "epoch": 0.4792868148800352, + "grad_norm": 6.76508558479979, + "learning_rate": 5.577061657758561e-06, + "loss": 0.3734, + "step": 21774 + }, + { + "epoch": 0.47930882676645387, + "grad_norm": 0.6648160588455897, + "learning_rate": 5.576707581872211e-06, + "loss": 0.4816, + "step": 21775 + }, + { + "epoch": 0.47933083865287257, + "grad_norm": 1.646381838921842, + "learning_rate": 5.576353503054775e-06, + "loss": 0.3234, + "step": 21776 + }, + { + "epoch": 0.4793528505392912, + "grad_norm": 1.5314941613512674, + "learning_rate": 5.575999421308054e-06, + "loss": 0.2904, + "step": 21777 + }, + { + "epoch": 0.47937486242570987, + "grad_norm": 2.6611612183427904, + "learning_rate": 5.575645336633848e-06, + "loss": 0.3839, + "step": 21778 + }, + { + "epoch": 0.4793968743121286, + "grad_norm": 0.6573115096271396, + "learning_rate": 5.575291249033956e-06, + "loss": 0.4723, + "step": 21779 + }, + { + "epoch": 0.4794188861985472, + "grad_norm": 0.6201401957524111, + "learning_rate": 5.574937158510176e-06, + "loss": 0.4881, + "step": 21780 + }, + { + "epoch": 0.47944089808496587, + "grad_norm": 1.3461352292147275, + "learning_rate": 5.57458306506431e-06, + "loss": 0.3287, + "step": 21781 + }, + { + "epoch": 0.4794629099713845, + "grad_norm": 1.2999616520688795, + "learning_rate": 5.574228968698155e-06, + "loss": 0.3005, + "step": 21782 + }, + { + "epoch": 0.4794849218578032, + "grad_norm": 1.806393701584187, + "learning_rate": 5.573874869413513e-06, + "loss": 0.3363, + "step": 21783 + }, + { + "epoch": 0.4795069337442219, + "grad_norm": 1.459175645665949, + "learning_rate": 5.573520767212184e-06, + "loss": 0.3348, + "step": 21784 + }, + { + "epoch": 0.4795289456306405, + "grad_norm": 1.7231774451835318, + "learning_rate": 5.5731666620959644e-06, + "loss": 0.338, + "step": 21785 + }, + { + "epoch": 0.47955095751705923, + "grad_norm": 1.3596812735141532, + "learning_rate": 5.572812554066659e-06, + "loss": 0.2365, + "step": 21786 + }, + { + "epoch": 0.4795729694034779, + "grad_norm": 1.5197547931960331, + "learning_rate": 5.572458443126063e-06, + "loss": 0.3277, + "step": 21787 + }, + { + "epoch": 0.4795949812898965, + "grad_norm": 1.2228833000282806, + "learning_rate": 5.572104329275978e-06, + "loss": 0.3015, + "step": 21788 + }, + { + "epoch": 0.47961699317631523, + "grad_norm": 1.5737568478251425, + "learning_rate": 5.5717502125182046e-06, + "loss": 0.2899, + "step": 21789 + }, + { + "epoch": 0.4796390050627339, + "grad_norm": 1.4994712693251127, + "learning_rate": 5.57139609285454e-06, + "loss": 0.334, + "step": 21790 + }, + { + "epoch": 0.47966101694915253, + "grad_norm": 1.3971761878464166, + "learning_rate": 5.571041970286787e-06, + "loss": 0.3665, + "step": 21791 + }, + { + "epoch": 0.47968302883557123, + "grad_norm": 1.1742669392491707, + "learning_rate": 5.570687844816744e-06, + "loss": 0.3371, + "step": 21792 + }, + { + "epoch": 0.4797050407219899, + "grad_norm": 0.7131577664403382, + "learning_rate": 5.57033371644621e-06, + "loss": 0.4886, + "step": 21793 + }, + { + "epoch": 0.47972705260840853, + "grad_norm": 2.0317883008874755, + "learning_rate": 5.569979585176986e-06, + "loss": 0.3016, + "step": 21794 + }, + { + "epoch": 0.4797490644948272, + "grad_norm": 0.6167068586516448, + "learning_rate": 5.56962545101087e-06, + "loss": 0.4659, + "step": 21795 + }, + { + "epoch": 0.4797710763812459, + "grad_norm": 1.6681056786704453, + "learning_rate": 5.569271313949666e-06, + "loss": 0.2408, + "step": 21796 + }, + { + "epoch": 0.47979308826766454, + "grad_norm": 1.0885894888568937, + "learning_rate": 5.568917173995171e-06, + "loss": 0.2727, + "step": 21797 + }, + { + "epoch": 0.4798151001540832, + "grad_norm": 1.4362423515833758, + "learning_rate": 5.5685630311491834e-06, + "loss": 0.3302, + "step": 21798 + }, + { + "epoch": 0.4798371120405019, + "grad_norm": 1.4802690313971907, + "learning_rate": 5.5682088854135065e-06, + "loss": 0.323, + "step": 21799 + }, + { + "epoch": 0.47985912392692054, + "grad_norm": 1.6072248186283293, + "learning_rate": 5.5678547367899385e-06, + "loss": 0.3375, + "step": 21800 + }, + { + "epoch": 0.4798811358133392, + "grad_norm": 1.8376108831139284, + "learning_rate": 5.567500585280278e-06, + "loss": 0.2745, + "step": 21801 + }, + { + "epoch": 0.4799031476997579, + "grad_norm": 58.45139731034675, + "learning_rate": 5.567146430886327e-06, + "loss": 0.3014, + "step": 21802 + }, + { + "epoch": 0.47992515958617654, + "grad_norm": 1.454048350534772, + "learning_rate": 5.566792273609886e-06, + "loss": 0.2675, + "step": 21803 + }, + { + "epoch": 0.4799471714725952, + "grad_norm": 1.6356978470819001, + "learning_rate": 5.566438113452752e-06, + "loss": 0.2566, + "step": 21804 + }, + { + "epoch": 0.47996918335901384, + "grad_norm": 0.7392566226417292, + "learning_rate": 5.566083950416729e-06, + "loss": 0.4905, + "step": 21805 + }, + { + "epoch": 0.47999119524543254, + "grad_norm": 1.5171618772430262, + "learning_rate": 5.565729784503614e-06, + "loss": 0.4232, + "step": 21806 + }, + { + "epoch": 0.4800132071318512, + "grad_norm": 1.4143287043877038, + "learning_rate": 5.565375615715209e-06, + "loss": 0.2708, + "step": 21807 + }, + { + "epoch": 0.48003521901826984, + "grad_norm": 1.2421831924371207, + "learning_rate": 5.56502144405331e-06, + "loss": 0.3149, + "step": 21808 + }, + { + "epoch": 0.48005723090468855, + "grad_norm": 1.558833704513359, + "learning_rate": 5.564667269519722e-06, + "loss": 0.279, + "step": 21809 + }, + { + "epoch": 0.4800792427911072, + "grad_norm": 1.5728731784827834, + "learning_rate": 5.5643130921162424e-06, + "loss": 0.28, + "step": 21810 + }, + { + "epoch": 0.48010125467752585, + "grad_norm": 1.3890502014688646, + "learning_rate": 5.563958911844672e-06, + "loss": 0.3285, + "step": 21811 + }, + { + "epoch": 0.48012326656394455, + "grad_norm": 1.4424074780953566, + "learning_rate": 5.563604728706813e-06, + "loss": 0.3017, + "step": 21812 + }, + { + "epoch": 0.4801452784503632, + "grad_norm": 1.5508365861879998, + "learning_rate": 5.5632505427044605e-06, + "loss": 0.4427, + "step": 21813 + }, + { + "epoch": 0.48016729033678185, + "grad_norm": 1.372454114236917, + "learning_rate": 5.5628963538394185e-06, + "loss": 0.2596, + "step": 21814 + }, + { + "epoch": 0.48018930222320055, + "grad_norm": 1.203334420737365, + "learning_rate": 5.562542162113488e-06, + "loss": 0.2446, + "step": 21815 + }, + { + "epoch": 0.4802113141096192, + "grad_norm": 1.2672240344935446, + "learning_rate": 5.562187967528464e-06, + "loss": 0.2878, + "step": 21816 + }, + { + "epoch": 0.48023332599603785, + "grad_norm": 1.5774754283614503, + "learning_rate": 5.561833770086153e-06, + "loss": 0.306, + "step": 21817 + }, + { + "epoch": 0.4802553378824565, + "grad_norm": 1.4492521423566744, + "learning_rate": 5.5614795697883504e-06, + "loss": 0.2909, + "step": 21818 + }, + { + "epoch": 0.4802773497688752, + "grad_norm": 0.6637226684750819, + "learning_rate": 5.561125366636858e-06, + "loss": 0.4732, + "step": 21819 + }, + { + "epoch": 0.48029936165529385, + "grad_norm": 1.3232114828349169, + "learning_rate": 5.560771160633478e-06, + "loss": 0.352, + "step": 21820 + }, + { + "epoch": 0.4803213735417125, + "grad_norm": 1.3363239032583611, + "learning_rate": 5.560416951780007e-06, + "loss": 0.256, + "step": 21821 + }, + { + "epoch": 0.4803433854281312, + "grad_norm": 0.6631119379824364, + "learning_rate": 5.560062740078248e-06, + "loss": 0.4949, + "step": 21822 + }, + { + "epoch": 0.48036539731454986, + "grad_norm": 1.889027063836861, + "learning_rate": 5.559708525530001e-06, + "loss": 0.3786, + "step": 21823 + }, + { + "epoch": 0.4803874092009685, + "grad_norm": 1.159172132603085, + "learning_rate": 5.559354308137065e-06, + "loss": 0.2445, + "step": 21824 + }, + { + "epoch": 0.4804094210873872, + "grad_norm": 1.3342553294729784, + "learning_rate": 5.559000087901241e-06, + "loss": 0.3129, + "step": 21825 + }, + { + "epoch": 0.48043143297380586, + "grad_norm": 0.6255029472250612, + "learning_rate": 5.558645864824328e-06, + "loss": 0.5288, + "step": 21826 + }, + { + "epoch": 0.4804534448602245, + "grad_norm": 1.38620785555661, + "learning_rate": 5.558291638908129e-06, + "loss": 0.3689, + "step": 21827 + }, + { + "epoch": 0.4804754567466432, + "grad_norm": 1.4037721420616336, + "learning_rate": 5.557937410154443e-06, + "loss": 0.3706, + "step": 21828 + }, + { + "epoch": 0.48049746863306186, + "grad_norm": 1.6217141019122097, + "learning_rate": 5.557583178565069e-06, + "loss": 0.34, + "step": 21829 + }, + { + "epoch": 0.4805194805194805, + "grad_norm": 1.5482812836524456, + "learning_rate": 5.557228944141809e-06, + "loss": 0.273, + "step": 21830 + }, + { + "epoch": 0.48054149240589916, + "grad_norm": 1.575546234017865, + "learning_rate": 5.556874706886464e-06, + "loss": 0.3157, + "step": 21831 + }, + { + "epoch": 0.48056350429231787, + "grad_norm": 1.4000627431839878, + "learning_rate": 5.556520466800832e-06, + "loss": 0.3605, + "step": 21832 + }, + { + "epoch": 0.4805855161787365, + "grad_norm": 1.2775888097964816, + "learning_rate": 5.556166223886717e-06, + "loss": 0.2865, + "step": 21833 + }, + { + "epoch": 0.48060752806515517, + "grad_norm": 0.6186567613747262, + "learning_rate": 5.555811978145914e-06, + "loss": 0.484, + "step": 21834 + }, + { + "epoch": 0.48062953995157387, + "grad_norm": 0.6274506168055699, + "learning_rate": 5.5554577295802285e-06, + "loss": 0.4531, + "step": 21835 + }, + { + "epoch": 0.4806515518379925, + "grad_norm": 1.706260701640101, + "learning_rate": 5.55510347819146e-06, + "loss": 0.4144, + "step": 21836 + }, + { + "epoch": 0.48067356372441117, + "grad_norm": 1.322398048442124, + "learning_rate": 5.554749223981405e-06, + "loss": 0.3317, + "step": 21837 + }, + { + "epoch": 0.4806955756108299, + "grad_norm": 1.7241611676024804, + "learning_rate": 5.5543949669518705e-06, + "loss": 0.3929, + "step": 21838 + }, + { + "epoch": 0.4807175874972485, + "grad_norm": 1.3686377876327145, + "learning_rate": 5.554040707104651e-06, + "loss": 0.3734, + "step": 21839 + }, + { + "epoch": 0.48073959938366717, + "grad_norm": 1.9341475879847916, + "learning_rate": 5.553686444441551e-06, + "loss": 0.4125, + "step": 21840 + }, + { + "epoch": 0.4807616112700858, + "grad_norm": 1.4315006043781888, + "learning_rate": 5.553332178964368e-06, + "loss": 0.2393, + "step": 21841 + }, + { + "epoch": 0.4807836231565045, + "grad_norm": 1.3527881272300617, + "learning_rate": 5.552977910674906e-06, + "loss": 0.371, + "step": 21842 + }, + { + "epoch": 0.4808056350429232, + "grad_norm": 1.6101603670079143, + "learning_rate": 5.5526236395749625e-06, + "loss": 0.3512, + "step": 21843 + }, + { + "epoch": 0.4808276469293418, + "grad_norm": 1.3376998142040755, + "learning_rate": 5.552269365666339e-06, + "loss": 0.303, + "step": 21844 + }, + { + "epoch": 0.48084965881576053, + "grad_norm": 1.6241606377195226, + "learning_rate": 5.551915088950836e-06, + "loss": 0.4174, + "step": 21845 + }, + { + "epoch": 0.4808716707021792, + "grad_norm": 1.5578630416811297, + "learning_rate": 5.551560809430255e-06, + "loss": 0.3269, + "step": 21846 + }, + { + "epoch": 0.4808936825885978, + "grad_norm": 1.307566985439162, + "learning_rate": 5.551206527106397e-06, + "loss": 0.2758, + "step": 21847 + }, + { + "epoch": 0.48091569447501653, + "grad_norm": 2.151267771915992, + "learning_rate": 5.550852241981059e-06, + "loss": 0.2907, + "step": 21848 + }, + { + "epoch": 0.4809377063614352, + "grad_norm": 0.6227998177539996, + "learning_rate": 5.550497954056047e-06, + "loss": 0.4719, + "step": 21849 + }, + { + "epoch": 0.48095971824785383, + "grad_norm": 1.6378060229586038, + "learning_rate": 5.550143663333156e-06, + "loss": 0.3426, + "step": 21850 + }, + { + "epoch": 0.48098173013427253, + "grad_norm": 1.4427919665057383, + "learning_rate": 5.5497893698141914e-06, + "loss": 0.3363, + "step": 21851 + }, + { + "epoch": 0.4810037420206912, + "grad_norm": 1.2532064286747528, + "learning_rate": 5.5494350735009525e-06, + "loss": 0.2367, + "step": 21852 + }, + { + "epoch": 0.48102575390710983, + "grad_norm": 1.7068198834937658, + "learning_rate": 5.549080774395239e-06, + "loss": 0.3501, + "step": 21853 + }, + { + "epoch": 0.4810477657935285, + "grad_norm": 1.3161213124480002, + "learning_rate": 5.548726472498852e-06, + "loss": 0.2496, + "step": 21854 + }, + { + "epoch": 0.4810697776799472, + "grad_norm": 1.247644329885376, + "learning_rate": 5.548372167813592e-06, + "loss": 0.234, + "step": 21855 + }, + { + "epoch": 0.48109178956636584, + "grad_norm": 1.4252041060492842, + "learning_rate": 5.54801786034126e-06, + "loss": 0.3146, + "step": 21856 + }, + { + "epoch": 0.4811138014527845, + "grad_norm": 1.2286703558067122, + "learning_rate": 5.547663550083656e-06, + "loss": 0.3019, + "step": 21857 + }, + { + "epoch": 0.4811358133392032, + "grad_norm": 3.434536630710558, + "learning_rate": 5.547309237042583e-06, + "loss": 0.3901, + "step": 21858 + }, + { + "epoch": 0.48115782522562184, + "grad_norm": 1.384026310438916, + "learning_rate": 5.5469549212198405e-06, + "loss": 0.2194, + "step": 21859 + }, + { + "epoch": 0.4811798371120405, + "grad_norm": 1.3971001007239359, + "learning_rate": 5.5466006026172285e-06, + "loss": 0.3314, + "step": 21860 + }, + { + "epoch": 0.4812018489984592, + "grad_norm": 1.3853826294485756, + "learning_rate": 5.546246281236549e-06, + "loss": 0.3295, + "step": 21861 + }, + { + "epoch": 0.48122386088487784, + "grad_norm": 2.241091068432383, + "learning_rate": 5.545891957079601e-06, + "loss": 0.3066, + "step": 21862 + }, + { + "epoch": 0.4812458727712965, + "grad_norm": 0.6671392528136992, + "learning_rate": 5.545537630148188e-06, + "loss": 0.4879, + "step": 21863 + }, + { + "epoch": 0.48126788465771514, + "grad_norm": 1.4985847812396211, + "learning_rate": 5.545183300444109e-06, + "loss": 0.3108, + "step": 21864 + }, + { + "epoch": 0.48128989654413384, + "grad_norm": 1.8098304343330562, + "learning_rate": 5.5448289679691645e-06, + "loss": 0.4181, + "step": 21865 + }, + { + "epoch": 0.4813119084305525, + "grad_norm": 1.4183161670400315, + "learning_rate": 5.544474632725157e-06, + "loss": 0.3115, + "step": 21866 + }, + { + "epoch": 0.48133392031697114, + "grad_norm": 1.207059872356812, + "learning_rate": 5.544120294713887e-06, + "loss": 0.3013, + "step": 21867 + }, + { + "epoch": 0.48135593220338985, + "grad_norm": 0.6356746261970653, + "learning_rate": 5.543765953937154e-06, + "loss": 0.4915, + "step": 21868 + }, + { + "epoch": 0.4813779440898085, + "grad_norm": 1.4408393588718533, + "learning_rate": 5.543411610396761e-06, + "loss": 0.3424, + "step": 21869 + }, + { + "epoch": 0.48139995597622715, + "grad_norm": 1.5982296258581454, + "learning_rate": 5.543057264094507e-06, + "loss": 0.2826, + "step": 21870 + }, + { + "epoch": 0.48142196786264585, + "grad_norm": 1.379148210126191, + "learning_rate": 5.542702915032193e-06, + "loss": 0.2327, + "step": 21871 + }, + { + "epoch": 0.4814439797490645, + "grad_norm": 1.1902047951409036, + "learning_rate": 5.542348563211622e-06, + "loss": 0.2836, + "step": 21872 + }, + { + "epoch": 0.48146599163548315, + "grad_norm": 1.4259384135591329, + "learning_rate": 5.541994208634593e-06, + "loss": 0.387, + "step": 21873 + }, + { + "epoch": 0.48148800352190185, + "grad_norm": 0.6482019612267921, + "learning_rate": 5.541639851302909e-06, + "loss": 0.4968, + "step": 21874 + }, + { + "epoch": 0.4815100154083205, + "grad_norm": 1.60521977261272, + "learning_rate": 5.541285491218367e-06, + "loss": 0.3041, + "step": 21875 + }, + { + "epoch": 0.48153202729473915, + "grad_norm": 0.6150852314346394, + "learning_rate": 5.540931128382774e-06, + "loss": 0.4731, + "step": 21876 + }, + { + "epoch": 0.4815540391811578, + "grad_norm": 2.064190084289621, + "learning_rate": 5.5405767627979255e-06, + "loss": 0.2691, + "step": 21877 + }, + { + "epoch": 0.4815760510675765, + "grad_norm": 0.6106294988209446, + "learning_rate": 5.5402223944656255e-06, + "loss": 0.4678, + "step": 21878 + }, + { + "epoch": 0.48159806295399515, + "grad_norm": 1.2505852054331905, + "learning_rate": 5.539868023387676e-06, + "loss": 0.2961, + "step": 21879 + }, + { + "epoch": 0.4816200748404138, + "grad_norm": 2.3784945347557036, + "learning_rate": 5.539513649565873e-06, + "loss": 0.2567, + "step": 21880 + }, + { + "epoch": 0.4816420867268325, + "grad_norm": 4.396289607849467, + "learning_rate": 5.539159273002023e-06, + "loss": 0.2782, + "step": 21881 + }, + { + "epoch": 0.48166409861325116, + "grad_norm": 0.6034669594355375, + "learning_rate": 5.538804893697926e-06, + "loss": 0.4867, + "step": 21882 + }, + { + "epoch": 0.4816861104996698, + "grad_norm": 1.2726017275592663, + "learning_rate": 5.53845051165538e-06, + "loss": 0.3749, + "step": 21883 + }, + { + "epoch": 0.4817081223860885, + "grad_norm": 1.4790884092693015, + "learning_rate": 5.5380961268761905e-06, + "loss": 0.3421, + "step": 21884 + }, + { + "epoch": 0.48173013427250716, + "grad_norm": 1.1377573376341013, + "learning_rate": 5.537741739362154e-06, + "loss": 0.2395, + "step": 21885 + }, + { + "epoch": 0.4817521461589258, + "grad_norm": 7.138644567325517, + "learning_rate": 5.537387349115076e-06, + "loss": 0.2927, + "step": 21886 + }, + { + "epoch": 0.48177415804534446, + "grad_norm": 1.3697301530551997, + "learning_rate": 5.537032956136755e-06, + "loss": 0.3633, + "step": 21887 + }, + { + "epoch": 0.48179616993176316, + "grad_norm": 1.649632602206807, + "learning_rate": 5.536678560428994e-06, + "loss": 0.3178, + "step": 21888 + }, + { + "epoch": 0.4818181818181818, + "grad_norm": 1.1541186017382175, + "learning_rate": 5.536324161993592e-06, + "loss": 0.2956, + "step": 21889 + }, + { + "epoch": 0.48184019370460046, + "grad_norm": 1.2935777117316718, + "learning_rate": 5.535969760832353e-06, + "loss": 0.3807, + "step": 21890 + }, + { + "epoch": 0.48186220559101917, + "grad_norm": 1.3134773639413662, + "learning_rate": 5.535615356947074e-06, + "loss": 0.3988, + "step": 21891 + }, + { + "epoch": 0.4818842174774378, + "grad_norm": 1.3428529623545944, + "learning_rate": 5.535260950339561e-06, + "loss": 0.3832, + "step": 21892 + }, + { + "epoch": 0.48190622936385646, + "grad_norm": 1.3636191912731292, + "learning_rate": 5.534906541011612e-06, + "loss": 0.3627, + "step": 21893 + }, + { + "epoch": 0.48192824125027517, + "grad_norm": 1.7066982515888292, + "learning_rate": 5.534552128965031e-06, + "loss": 0.3637, + "step": 21894 + }, + { + "epoch": 0.4819502531366938, + "grad_norm": 1.1300807546053024, + "learning_rate": 5.534197714201617e-06, + "loss": 0.2542, + "step": 21895 + }, + { + "epoch": 0.48197226502311247, + "grad_norm": 1.2076916562389732, + "learning_rate": 5.53384329672317e-06, + "loss": 0.2905, + "step": 21896 + }, + { + "epoch": 0.48199427690953117, + "grad_norm": 0.7051512109944017, + "learning_rate": 5.533488876531497e-06, + "loss": 0.4797, + "step": 21897 + }, + { + "epoch": 0.4820162887959498, + "grad_norm": 0.6605217138796705, + "learning_rate": 5.533134453628392e-06, + "loss": 0.5037, + "step": 21898 + }, + { + "epoch": 0.48203830068236847, + "grad_norm": 2.248324552436445, + "learning_rate": 5.5327800280156605e-06, + "loss": 0.3546, + "step": 21899 + }, + { + "epoch": 0.4820603125687871, + "grad_norm": 0.6233409333367923, + "learning_rate": 5.532425599695105e-06, + "loss": 0.4832, + "step": 21900 + }, + { + "epoch": 0.4820823244552058, + "grad_norm": 1.314909561078424, + "learning_rate": 5.532071168668524e-06, + "loss": 0.2465, + "step": 21901 + }, + { + "epoch": 0.4821043363416245, + "grad_norm": 2.102130142872311, + "learning_rate": 5.531716734937721e-06, + "loss": 0.349, + "step": 21902 + }, + { + "epoch": 0.4821263482280431, + "grad_norm": 1.2661743374977126, + "learning_rate": 5.531362298504496e-06, + "loss": 0.4002, + "step": 21903 + }, + { + "epoch": 0.4821483601144618, + "grad_norm": 0.663049352081386, + "learning_rate": 5.53100785937065e-06, + "loss": 0.4909, + "step": 21904 + }, + { + "epoch": 0.4821703720008805, + "grad_norm": 3.100378044840032, + "learning_rate": 5.530653417537985e-06, + "loss": 0.3265, + "step": 21905 + }, + { + "epoch": 0.4821923838872991, + "grad_norm": 1.498209015161953, + "learning_rate": 5.530298973008303e-06, + "loss": 0.314, + "step": 21906 + }, + { + "epoch": 0.48221439577371783, + "grad_norm": 3.795296089118485, + "learning_rate": 5.529944525783406e-06, + "loss": 0.3553, + "step": 21907 + }, + { + "epoch": 0.4822364076601365, + "grad_norm": 1.499862389621409, + "learning_rate": 5.529590075865094e-06, + "loss": 0.2502, + "step": 21908 + }, + { + "epoch": 0.48225841954655513, + "grad_norm": 1.6608505391124724, + "learning_rate": 5.529235623255168e-06, + "loss": 0.3287, + "step": 21909 + }, + { + "epoch": 0.48228043143297383, + "grad_norm": 1.2952620539313628, + "learning_rate": 5.5288811679554325e-06, + "loss": 0.3184, + "step": 21910 + }, + { + "epoch": 0.4823024433193925, + "grad_norm": 1.2570378275028735, + "learning_rate": 5.528526709967685e-06, + "loss": 0.2484, + "step": 21911 + }, + { + "epoch": 0.48232445520581113, + "grad_norm": 1.786861607357095, + "learning_rate": 5.528172249293731e-06, + "loss": 0.4077, + "step": 21912 + }, + { + "epoch": 0.4823464670922298, + "grad_norm": 0.6448587098272891, + "learning_rate": 5.527817785935369e-06, + "loss": 0.4556, + "step": 21913 + }, + { + "epoch": 0.4823684789786485, + "grad_norm": 1.2385489296828294, + "learning_rate": 5.527463319894402e-06, + "loss": 0.3079, + "step": 21914 + }, + { + "epoch": 0.48239049086506713, + "grad_norm": 1.4914141770222205, + "learning_rate": 5.527108851172631e-06, + "loss": 0.2765, + "step": 21915 + }, + { + "epoch": 0.4824125027514858, + "grad_norm": 1.208438740713221, + "learning_rate": 5.526754379771857e-06, + "loss": 0.336, + "step": 21916 + }, + { + "epoch": 0.4824345146379045, + "grad_norm": 2.634367909559265, + "learning_rate": 5.526399905693884e-06, + "loss": 0.3326, + "step": 21917 + }, + { + "epoch": 0.48245652652432314, + "grad_norm": 1.3669238647498683, + "learning_rate": 5.52604542894051e-06, + "loss": 0.3698, + "step": 21918 + }, + { + "epoch": 0.4824785384107418, + "grad_norm": 0.6135114274260681, + "learning_rate": 5.525690949513538e-06, + "loss": 0.4595, + "step": 21919 + }, + { + "epoch": 0.4825005502971605, + "grad_norm": 0.6374027448793835, + "learning_rate": 5.525336467414774e-06, + "loss": 0.4764, + "step": 21920 + }, + { + "epoch": 0.48252256218357914, + "grad_norm": 1.425051212419593, + "learning_rate": 5.524981982646013e-06, + "loss": 0.3395, + "step": 21921 + }, + { + "epoch": 0.4825445740699978, + "grad_norm": 1.2599767143881573, + "learning_rate": 5.524627495209059e-06, + "loss": 0.319, + "step": 21922 + }, + { + "epoch": 0.48256658595641644, + "grad_norm": 1.4439141330171754, + "learning_rate": 5.524273005105714e-06, + "loss": 0.3703, + "step": 21923 + }, + { + "epoch": 0.48258859784283514, + "grad_norm": 1.742195900079235, + "learning_rate": 5.523918512337781e-06, + "loss": 0.3205, + "step": 21924 + }, + { + "epoch": 0.4826106097292538, + "grad_norm": 1.367396871221195, + "learning_rate": 5.52356401690706e-06, + "loss": 0.342, + "step": 21925 + }, + { + "epoch": 0.48263262161567244, + "grad_norm": 1.918182021793423, + "learning_rate": 5.523209518815353e-06, + "loss": 0.3324, + "step": 21926 + }, + { + "epoch": 0.48265463350209115, + "grad_norm": 1.5009241092226198, + "learning_rate": 5.522855018064461e-06, + "loss": 0.2929, + "step": 21927 + }, + { + "epoch": 0.4826766453885098, + "grad_norm": 1.8813311556222299, + "learning_rate": 5.522500514656188e-06, + "loss": 0.2955, + "step": 21928 + }, + { + "epoch": 0.48269865727492844, + "grad_norm": 0.6452661933208969, + "learning_rate": 5.522146008592334e-06, + "loss": 0.4772, + "step": 21929 + }, + { + "epoch": 0.48272066916134715, + "grad_norm": 1.9160301012803145, + "learning_rate": 5.521791499874701e-06, + "loss": 0.3347, + "step": 21930 + }, + { + "epoch": 0.4827426810477658, + "grad_norm": 1.7804479124681625, + "learning_rate": 5.521436988505091e-06, + "loss": 0.2921, + "step": 21931 + }, + { + "epoch": 0.48276469293418445, + "grad_norm": 1.4499631704158662, + "learning_rate": 5.521082474485304e-06, + "loss": 0.4029, + "step": 21932 + }, + { + "epoch": 0.48278670482060315, + "grad_norm": 1.4809344009092968, + "learning_rate": 5.520727957817147e-06, + "loss": 0.2433, + "step": 21933 + }, + { + "epoch": 0.4828087167070218, + "grad_norm": 1.7416559004304777, + "learning_rate": 5.520373438502415e-06, + "loss": 0.3229, + "step": 21934 + }, + { + "epoch": 0.48283072859344045, + "grad_norm": 0.6533383675979227, + "learning_rate": 5.520018916542915e-06, + "loss": 0.4916, + "step": 21935 + }, + { + "epoch": 0.4828527404798591, + "grad_norm": 1.567773319245159, + "learning_rate": 5.519664391940445e-06, + "loss": 0.3262, + "step": 21936 + }, + { + "epoch": 0.4828747523662778, + "grad_norm": 1.5540537108726848, + "learning_rate": 5.51930986469681e-06, + "loss": 0.3271, + "step": 21937 + }, + { + "epoch": 0.48289676425269645, + "grad_norm": 1.3177599544813805, + "learning_rate": 5.518955334813812e-06, + "loss": 0.2675, + "step": 21938 + }, + { + "epoch": 0.4829187761391151, + "grad_norm": 1.5131453485631225, + "learning_rate": 5.5186008022932506e-06, + "loss": 0.2262, + "step": 21939 + }, + { + "epoch": 0.4829407880255338, + "grad_norm": 1.157812283521351, + "learning_rate": 5.518246267136927e-06, + "loss": 0.2524, + "step": 21940 + }, + { + "epoch": 0.48296279991195246, + "grad_norm": 1.7830557799814932, + "learning_rate": 5.517891729346646e-06, + "loss": 0.3504, + "step": 21941 + }, + { + "epoch": 0.4829848117983711, + "grad_norm": 1.695259733724354, + "learning_rate": 5.517537188924209e-06, + "loss": 0.38, + "step": 21942 + }, + { + "epoch": 0.4830068236847898, + "grad_norm": 0.6329532138846976, + "learning_rate": 5.517182645871416e-06, + "loss": 0.5038, + "step": 21943 + }, + { + "epoch": 0.48302883557120846, + "grad_norm": 1.3730047362869906, + "learning_rate": 5.516828100190072e-06, + "loss": 0.2724, + "step": 21944 + }, + { + "epoch": 0.4830508474576271, + "grad_norm": 0.6078329102692828, + "learning_rate": 5.516473551881974e-06, + "loss": 0.5033, + "step": 21945 + }, + { + "epoch": 0.48307285934404576, + "grad_norm": 1.5639690753013302, + "learning_rate": 5.516119000948928e-06, + "loss": 0.2409, + "step": 21946 + }, + { + "epoch": 0.48309487123046446, + "grad_norm": 1.8180178098169746, + "learning_rate": 5.515764447392736e-06, + "loss": 0.3359, + "step": 21947 + }, + { + "epoch": 0.4831168831168831, + "grad_norm": 1.2529694315242943, + "learning_rate": 5.515409891215199e-06, + "loss": 0.3424, + "step": 21948 + }, + { + "epoch": 0.48313889500330176, + "grad_norm": 1.1524028190322582, + "learning_rate": 5.5150553324181184e-06, + "loss": 0.27, + "step": 21949 + }, + { + "epoch": 0.48316090688972047, + "grad_norm": 1.5693430075221535, + "learning_rate": 5.514700771003298e-06, + "loss": 0.4102, + "step": 21950 + }, + { + "epoch": 0.4831829187761391, + "grad_norm": 3.654137926762567, + "learning_rate": 5.514346206972537e-06, + "loss": 0.3831, + "step": 21951 + }, + { + "epoch": 0.48320493066255776, + "grad_norm": 2.22908223560169, + "learning_rate": 5.51399164032764e-06, + "loss": 0.3073, + "step": 21952 + }, + { + "epoch": 0.48322694254897647, + "grad_norm": 1.4808289392283194, + "learning_rate": 5.513637071070409e-06, + "loss": 0.3138, + "step": 21953 + }, + { + "epoch": 0.4832489544353951, + "grad_norm": 1.2211498653074233, + "learning_rate": 5.5132824992026445e-06, + "loss": 0.2205, + "step": 21954 + }, + { + "epoch": 0.48327096632181377, + "grad_norm": 1.4100058708160101, + "learning_rate": 5.512927924726149e-06, + "loss": 0.2515, + "step": 21955 + }, + { + "epoch": 0.48329297820823247, + "grad_norm": 1.4515523590192103, + "learning_rate": 5.512573347642726e-06, + "loss": 0.3025, + "step": 21956 + }, + { + "epoch": 0.4833149900946511, + "grad_norm": 0.6117415592205961, + "learning_rate": 5.512218767954177e-06, + "loss": 0.4774, + "step": 21957 + }, + { + "epoch": 0.48333700198106977, + "grad_norm": 1.3287706103526986, + "learning_rate": 5.511864185662302e-06, + "loss": 0.2345, + "step": 21958 + }, + { + "epoch": 0.4833590138674884, + "grad_norm": 1.8248543873323508, + "learning_rate": 5.5115096007689075e-06, + "loss": 0.31, + "step": 21959 + }, + { + "epoch": 0.4833810257539071, + "grad_norm": 1.1573260780087176, + "learning_rate": 5.5111550132757906e-06, + "loss": 0.2207, + "step": 21960 + }, + { + "epoch": 0.4834030376403258, + "grad_norm": 0.6116102365481624, + "learning_rate": 5.5108004231847576e-06, + "loss": 0.4927, + "step": 21961 + }, + { + "epoch": 0.4834250495267444, + "grad_norm": 2.8072263055778226, + "learning_rate": 5.510445830497608e-06, + "loss": 0.3452, + "step": 21962 + }, + { + "epoch": 0.4834470614131631, + "grad_norm": 0.6271806252513353, + "learning_rate": 5.510091235216144e-06, + "loss": 0.5022, + "step": 21963 + }, + { + "epoch": 0.4834690732995818, + "grad_norm": 1.2322049229425398, + "learning_rate": 5.5097366373421714e-06, + "loss": 0.2557, + "step": 21964 + }, + { + "epoch": 0.4834910851860004, + "grad_norm": 1.1162666749301302, + "learning_rate": 5.509382036877489e-06, + "loss": 0.281, + "step": 21965 + }, + { + "epoch": 0.48351309707241913, + "grad_norm": 1.450005328105403, + "learning_rate": 5.509027433823899e-06, + "loss": 0.3145, + "step": 21966 + }, + { + "epoch": 0.4835351089588378, + "grad_norm": 1.7619119926519529, + "learning_rate": 5.508672828183204e-06, + "loss": 0.2997, + "step": 21967 + }, + { + "epoch": 0.4835571208452564, + "grad_norm": 1.4071227437214557, + "learning_rate": 5.508318219957209e-06, + "loss": 0.3238, + "step": 21968 + }, + { + "epoch": 0.48357913273167513, + "grad_norm": 1.3539232452764538, + "learning_rate": 5.507963609147712e-06, + "loss": 0.2779, + "step": 21969 + }, + { + "epoch": 0.4836011446180938, + "grad_norm": 1.655635223099948, + "learning_rate": 5.507608995756518e-06, + "loss": 0.3015, + "step": 21970 + }, + { + "epoch": 0.48362315650451243, + "grad_norm": 0.6089542479713913, + "learning_rate": 5.5072543797854295e-06, + "loss": 0.4875, + "step": 21971 + }, + { + "epoch": 0.4836451683909311, + "grad_norm": 1.5144635410845886, + "learning_rate": 5.5068997612362485e-06, + "loss": 0.2672, + "step": 21972 + }, + { + "epoch": 0.4836671802773498, + "grad_norm": 1.3980505372418832, + "learning_rate": 5.506545140110774e-06, + "loss": 0.3405, + "step": 21973 + }, + { + "epoch": 0.48368919216376843, + "grad_norm": 1.324516158667347, + "learning_rate": 5.506190516410814e-06, + "loss": 0.3656, + "step": 21974 + }, + { + "epoch": 0.4837112040501871, + "grad_norm": 0.6313581782813511, + "learning_rate": 5.505835890138168e-06, + "loss": 0.4814, + "step": 21975 + }, + { + "epoch": 0.4837332159366058, + "grad_norm": 1.366103187628816, + "learning_rate": 5.505481261294636e-06, + "loss": 0.3222, + "step": 21976 + }, + { + "epoch": 0.48375522782302444, + "grad_norm": 1.7336870319773905, + "learning_rate": 5.505126629882025e-06, + "loss": 0.3199, + "step": 21977 + }, + { + "epoch": 0.4837772397094431, + "grad_norm": 1.696706418318098, + "learning_rate": 5.504771995902134e-06, + "loss": 0.3166, + "step": 21978 + }, + { + "epoch": 0.4837992515958618, + "grad_norm": 0.6351433305399753, + "learning_rate": 5.504417359356767e-06, + "loss": 0.4536, + "step": 21979 + }, + { + "epoch": 0.48382126348228044, + "grad_norm": 2.0106641289897973, + "learning_rate": 5.504062720247727e-06, + "loss": 0.2267, + "step": 21980 + }, + { + "epoch": 0.4838432753686991, + "grad_norm": 0.6193740257487091, + "learning_rate": 5.503708078576813e-06, + "loss": 0.4951, + "step": 21981 + }, + { + "epoch": 0.48386528725511774, + "grad_norm": 1.3582889754367284, + "learning_rate": 5.503353434345832e-06, + "loss": 0.3372, + "step": 21982 + }, + { + "epoch": 0.48388729914153644, + "grad_norm": 2.0027890789480898, + "learning_rate": 5.502998787556584e-06, + "loss": 0.3776, + "step": 21983 + }, + { + "epoch": 0.4839093110279551, + "grad_norm": 1.6210377295757816, + "learning_rate": 5.502644138210871e-06, + "loss": 0.3588, + "step": 21984 + }, + { + "epoch": 0.48393132291437374, + "grad_norm": 1.5499878893309254, + "learning_rate": 5.502289486310497e-06, + "loss": 0.2887, + "step": 21985 + }, + { + "epoch": 0.48395333480079245, + "grad_norm": 2.0834240364364094, + "learning_rate": 5.501934831857263e-06, + "loss": 0.3624, + "step": 21986 + }, + { + "epoch": 0.4839753466872111, + "grad_norm": 1.7952098930428046, + "learning_rate": 5.501580174852973e-06, + "loss": 0.3544, + "step": 21987 + }, + { + "epoch": 0.48399735857362974, + "grad_norm": 1.588722724658186, + "learning_rate": 5.501225515299429e-06, + "loss": 0.3738, + "step": 21988 + }, + { + "epoch": 0.48401937046004845, + "grad_norm": 1.2196092372238767, + "learning_rate": 5.500870853198432e-06, + "loss": 0.2471, + "step": 21989 + }, + { + "epoch": 0.4840413823464671, + "grad_norm": 2.8326271950753195, + "learning_rate": 5.500516188551788e-06, + "loss": 0.3489, + "step": 21990 + }, + { + "epoch": 0.48406339423288575, + "grad_norm": 2.1316959948674525, + "learning_rate": 5.500161521361296e-06, + "loss": 0.2879, + "step": 21991 + }, + { + "epoch": 0.48408540611930445, + "grad_norm": 0.6644511968514103, + "learning_rate": 5.499806851628762e-06, + "loss": 0.5164, + "step": 21992 + }, + { + "epoch": 0.4841074180057231, + "grad_norm": 0.5998755813330958, + "learning_rate": 5.499452179355985e-06, + "loss": 0.4783, + "step": 21993 + }, + { + "epoch": 0.48412942989214175, + "grad_norm": 1.3591980521341134, + "learning_rate": 5.499097504544768e-06, + "loss": 0.2966, + "step": 21994 + }, + { + "epoch": 0.4841514417785604, + "grad_norm": 1.5126764112432514, + "learning_rate": 5.4987428271969165e-06, + "loss": 0.3089, + "step": 21995 + }, + { + "epoch": 0.4841734536649791, + "grad_norm": 0.5931045042683167, + "learning_rate": 5.498388147314232e-06, + "loss": 0.4733, + "step": 21996 + }, + { + "epoch": 0.48419546555139775, + "grad_norm": 2.249121358842545, + "learning_rate": 5.4980334648985164e-06, + "loss": 0.3234, + "step": 21997 + }, + { + "epoch": 0.4842174774378164, + "grad_norm": 1.5846251994051428, + "learning_rate": 5.497678779951572e-06, + "loss": 0.419, + "step": 21998 + }, + { + "epoch": 0.4842394893242351, + "grad_norm": 1.1527468524531659, + "learning_rate": 5.497324092475201e-06, + "loss": 0.2741, + "step": 21999 + }, + { + "epoch": 0.48426150121065376, + "grad_norm": 1.3844119692519257, + "learning_rate": 5.496969402471209e-06, + "loss": 0.3024, + "step": 22000 + }, + { + "epoch": 0.4842835130970724, + "grad_norm": 1.2804498474501476, + "learning_rate": 5.4966147099413966e-06, + "loss": 0.327, + "step": 22001 + }, + { + "epoch": 0.4843055249834911, + "grad_norm": 1.515132220734515, + "learning_rate": 5.496260014887567e-06, + "loss": 0.2798, + "step": 22002 + }, + { + "epoch": 0.48432753686990976, + "grad_norm": 1.3131687428753411, + "learning_rate": 5.495905317311522e-06, + "loss": 0.3505, + "step": 22003 + }, + { + "epoch": 0.4843495487563284, + "grad_norm": 0.6684694722367561, + "learning_rate": 5.4955506172150645e-06, + "loss": 0.4861, + "step": 22004 + }, + { + "epoch": 0.48437156064274706, + "grad_norm": 1.3977124662172735, + "learning_rate": 5.495195914599998e-06, + "loss": 0.3638, + "step": 22005 + }, + { + "epoch": 0.48439357252916576, + "grad_norm": 1.5801449323104317, + "learning_rate": 5.494841209468126e-06, + "loss": 0.2889, + "step": 22006 + }, + { + "epoch": 0.4844155844155844, + "grad_norm": 0.63160266588428, + "learning_rate": 5.49448650182125e-06, + "loss": 0.4735, + "step": 22007 + }, + { + "epoch": 0.48443759630200306, + "grad_norm": 1.6099915690162407, + "learning_rate": 5.494131791661172e-06, + "loss": 0.3649, + "step": 22008 + }, + { + "epoch": 0.48445960818842176, + "grad_norm": 0.6192824838166318, + "learning_rate": 5.493777078989696e-06, + "loss": 0.4795, + "step": 22009 + }, + { + "epoch": 0.4844816200748404, + "grad_norm": 0.6270977938689201, + "learning_rate": 5.493422363808626e-06, + "loss": 0.4832, + "step": 22010 + }, + { + "epoch": 0.48450363196125906, + "grad_norm": 1.4746799935697228, + "learning_rate": 5.4930676461197625e-06, + "loss": 0.384, + "step": 22011 + }, + { + "epoch": 0.48452564384767777, + "grad_norm": 1.2489166835942143, + "learning_rate": 5.492712925924909e-06, + "loss": 0.3222, + "step": 22012 + }, + { + "epoch": 0.4845476557340964, + "grad_norm": 0.6329642894075777, + "learning_rate": 5.492358203225869e-06, + "loss": 0.4802, + "step": 22013 + }, + { + "epoch": 0.48456966762051507, + "grad_norm": 1.0513434359852667, + "learning_rate": 5.492003478024446e-06, + "loss": 0.2883, + "step": 22014 + }, + { + "epoch": 0.48459167950693377, + "grad_norm": 0.6280825545411558, + "learning_rate": 5.491648750322441e-06, + "loss": 0.5086, + "step": 22015 + }, + { + "epoch": 0.4846136913933524, + "grad_norm": 1.0722995969956575, + "learning_rate": 5.491294020121658e-06, + "loss": 0.2997, + "step": 22016 + }, + { + "epoch": 0.48463570327977107, + "grad_norm": 2.5245475341647423, + "learning_rate": 5.490939287423899e-06, + "loss": 0.3286, + "step": 22017 + }, + { + "epoch": 0.4846577151661897, + "grad_norm": 6.476030176668981, + "learning_rate": 5.4905845522309675e-06, + "loss": 0.333, + "step": 22018 + }, + { + "epoch": 0.4846797270526084, + "grad_norm": 1.984238461909765, + "learning_rate": 5.490229814544668e-06, + "loss": 0.3354, + "step": 22019 + }, + { + "epoch": 0.48470173893902707, + "grad_norm": 1.5255899035395224, + "learning_rate": 5.489875074366801e-06, + "loss": 0.2542, + "step": 22020 + }, + { + "epoch": 0.4847237508254457, + "grad_norm": 1.4767154554841373, + "learning_rate": 5.4895203316991705e-06, + "loss": 0.3586, + "step": 22021 + }, + { + "epoch": 0.4847457627118644, + "grad_norm": 0.6501289443559489, + "learning_rate": 5.489165586543578e-06, + "loss": 0.4588, + "step": 22022 + }, + { + "epoch": 0.4847677745982831, + "grad_norm": 1.7983961619952047, + "learning_rate": 5.488810838901829e-06, + "loss": 0.259, + "step": 22023 + }, + { + "epoch": 0.4847897864847017, + "grad_norm": 0.6571437189666385, + "learning_rate": 5.4884560887757255e-06, + "loss": 0.4691, + "step": 22024 + }, + { + "epoch": 0.48481179837112043, + "grad_norm": 1.8019459718913398, + "learning_rate": 5.488101336167071e-06, + "loss": 0.2602, + "step": 22025 + }, + { + "epoch": 0.4848338102575391, + "grad_norm": 1.5955753036923912, + "learning_rate": 5.4877465810776664e-06, + "loss": 0.3253, + "step": 22026 + }, + { + "epoch": 0.4848558221439577, + "grad_norm": 3.2574847215892944, + "learning_rate": 5.487391823509316e-06, + "loss": 0.2627, + "step": 22027 + }, + { + "epoch": 0.4848778340303764, + "grad_norm": 1.86638474129709, + "learning_rate": 5.487037063463824e-06, + "loss": 0.2487, + "step": 22028 + }, + { + "epoch": 0.4848998459167951, + "grad_norm": 1.3409572809220827, + "learning_rate": 5.4866823009429925e-06, + "loss": 0.2451, + "step": 22029 + }, + { + "epoch": 0.48492185780321373, + "grad_norm": 2.787169245336162, + "learning_rate": 5.4863275359486235e-06, + "loss": 0.3082, + "step": 22030 + }, + { + "epoch": 0.4849438696896324, + "grad_norm": 1.7308532081471095, + "learning_rate": 5.485972768482522e-06, + "loss": 0.2609, + "step": 22031 + }, + { + "epoch": 0.4849658815760511, + "grad_norm": 1.0848672300590518, + "learning_rate": 5.4856179985464895e-06, + "loss": 0.3135, + "step": 22032 + }, + { + "epoch": 0.48498789346246973, + "grad_norm": 1.8197278158538637, + "learning_rate": 5.48526322614233e-06, + "loss": 0.3462, + "step": 22033 + }, + { + "epoch": 0.4850099053488884, + "grad_norm": 0.6988993097763104, + "learning_rate": 5.484908451271847e-06, + "loss": 0.4729, + "step": 22034 + }, + { + "epoch": 0.4850319172353071, + "grad_norm": 1.948953529164374, + "learning_rate": 5.484553673936841e-06, + "loss": 0.3397, + "step": 22035 + }, + { + "epoch": 0.48505392912172574, + "grad_norm": 1.6019761913694166, + "learning_rate": 5.484198894139119e-06, + "loss": 0.2929, + "step": 22036 + }, + { + "epoch": 0.4850759410081444, + "grad_norm": 1.5430532395970373, + "learning_rate": 5.483844111880482e-06, + "loss": 0.3075, + "step": 22037 + }, + { + "epoch": 0.4850979528945631, + "grad_norm": 1.8652659888606318, + "learning_rate": 5.483489327162733e-06, + "loss": 0.3352, + "step": 22038 + }, + { + "epoch": 0.48511996478098174, + "grad_norm": 1.8997893548573541, + "learning_rate": 5.483134539987675e-06, + "loss": 0.3876, + "step": 22039 + }, + { + "epoch": 0.4851419766674004, + "grad_norm": 2.7710028786647665, + "learning_rate": 5.482779750357112e-06, + "loss": 0.3363, + "step": 22040 + }, + { + "epoch": 0.48516398855381904, + "grad_norm": 1.2799100585910685, + "learning_rate": 5.4824249582728474e-06, + "loss": 0.3366, + "step": 22041 + }, + { + "epoch": 0.48518600044023774, + "grad_norm": 1.6579402194018333, + "learning_rate": 5.482070163736686e-06, + "loss": 0.3069, + "step": 22042 + }, + { + "epoch": 0.4852080123266564, + "grad_norm": 0.6329988634604874, + "learning_rate": 5.481715366750425e-06, + "loss": 0.4504, + "step": 22043 + }, + { + "epoch": 0.48523002421307504, + "grad_norm": 1.2746673543201998, + "learning_rate": 5.481360567315874e-06, + "loss": 0.2758, + "step": 22044 + }, + { + "epoch": 0.48525203609949374, + "grad_norm": 1.2717009962100339, + "learning_rate": 5.481005765434834e-06, + "loss": 0.328, + "step": 22045 + }, + { + "epoch": 0.4852740479859124, + "grad_norm": 2.4938053705154735, + "learning_rate": 5.480650961109107e-06, + "loss": 0.3114, + "step": 22046 + }, + { + "epoch": 0.48529605987233104, + "grad_norm": 1.4445025734303039, + "learning_rate": 5.4802961543405e-06, + "loss": 0.237, + "step": 22047 + }, + { + "epoch": 0.48531807175874975, + "grad_norm": 1.2607504986613443, + "learning_rate": 5.47994134513081e-06, + "loss": 0.2911, + "step": 22048 + }, + { + "epoch": 0.4853400836451684, + "grad_norm": 1.5228349931221403, + "learning_rate": 5.479586533481847e-06, + "loss": 0.3365, + "step": 22049 + }, + { + "epoch": 0.48536209553158705, + "grad_norm": 0.635507581585502, + "learning_rate": 5.479231719395411e-06, + "loss": 0.4709, + "step": 22050 + }, + { + "epoch": 0.48538410741800575, + "grad_norm": 2.3201858410461824, + "learning_rate": 5.478876902873303e-06, + "loss": 0.3861, + "step": 22051 + }, + { + "epoch": 0.4854061193044244, + "grad_norm": 1.3321482206489579, + "learning_rate": 5.478522083917332e-06, + "loss": 0.3536, + "step": 22052 + }, + { + "epoch": 0.48542813119084305, + "grad_norm": 2.4491226975912275, + "learning_rate": 5.478167262529297e-06, + "loss": 0.3659, + "step": 22053 + }, + { + "epoch": 0.4854501430772617, + "grad_norm": 2.178174773123319, + "learning_rate": 5.477812438711004e-06, + "loss": 0.3286, + "step": 22054 + }, + { + "epoch": 0.4854721549636804, + "grad_norm": 1.673870359110393, + "learning_rate": 5.477457612464253e-06, + "loss": 0.3637, + "step": 22055 + }, + { + "epoch": 0.48549416685009905, + "grad_norm": 1.4946218304952177, + "learning_rate": 5.477102783790851e-06, + "loss": 0.3952, + "step": 22056 + }, + { + "epoch": 0.4855161787365177, + "grad_norm": 1.6888135398732833, + "learning_rate": 5.476747952692599e-06, + "loss": 0.3989, + "step": 22057 + }, + { + "epoch": 0.4855381906229364, + "grad_norm": 3.7514860237574017, + "learning_rate": 5.476393119171301e-06, + "loss": 0.2883, + "step": 22058 + }, + { + "epoch": 0.48556020250935505, + "grad_norm": 1.4693907112275355, + "learning_rate": 5.476038283228761e-06, + "loss": 0.319, + "step": 22059 + }, + { + "epoch": 0.4855822143957737, + "grad_norm": 1.2333219034625047, + "learning_rate": 5.475683444866784e-06, + "loss": 0.3236, + "step": 22060 + }, + { + "epoch": 0.4856042262821924, + "grad_norm": 0.6441187394449628, + "learning_rate": 5.47532860408717e-06, + "loss": 0.4922, + "step": 22061 + }, + { + "epoch": 0.48562623816861106, + "grad_norm": 1.5585283097787146, + "learning_rate": 5.474973760891724e-06, + "loss": 0.3748, + "step": 22062 + }, + { + "epoch": 0.4856482500550297, + "grad_norm": 1.626697525913141, + "learning_rate": 5.4746189152822485e-06, + "loss": 0.266, + "step": 22063 + }, + { + "epoch": 0.48567026194144836, + "grad_norm": 0.6576970575356073, + "learning_rate": 5.4742640672605495e-06, + "loss": 0.4809, + "step": 22064 + }, + { + "epoch": 0.48569227382786706, + "grad_norm": 2.421619562230478, + "learning_rate": 5.47390921682843e-06, + "loss": 0.3517, + "step": 22065 + }, + { + "epoch": 0.4857142857142857, + "grad_norm": 1.605539474974134, + "learning_rate": 5.47355436398769e-06, + "loss": 0.3527, + "step": 22066 + }, + { + "epoch": 0.48573629760070436, + "grad_norm": 1.240096862713061, + "learning_rate": 5.473199508740136e-06, + "loss": 0.262, + "step": 22067 + }, + { + "epoch": 0.48575830948712306, + "grad_norm": 1.5483792680726065, + "learning_rate": 5.472844651087572e-06, + "loss": 0.2766, + "step": 22068 + }, + { + "epoch": 0.4857803213735417, + "grad_norm": 1.2853710071653472, + "learning_rate": 5.4724897910318e-06, + "loss": 0.2973, + "step": 22069 + }, + { + "epoch": 0.48580233325996036, + "grad_norm": 0.6252818358984897, + "learning_rate": 5.472134928574625e-06, + "loss": 0.5048, + "step": 22070 + }, + { + "epoch": 0.48582434514637907, + "grad_norm": 0.6592052522591225, + "learning_rate": 5.471780063717849e-06, + "loss": 0.4968, + "step": 22071 + }, + { + "epoch": 0.4858463570327977, + "grad_norm": 1.235135117008762, + "learning_rate": 5.4714251964632755e-06, + "loss": 0.3642, + "step": 22072 + }, + { + "epoch": 0.48586836891921636, + "grad_norm": 1.645202025462141, + "learning_rate": 5.4710703268127105e-06, + "loss": 0.3129, + "step": 22073 + }, + { + "epoch": 0.48589038080563507, + "grad_norm": 1.4530419664294787, + "learning_rate": 5.470715454767955e-06, + "loss": 0.3513, + "step": 22074 + }, + { + "epoch": 0.4859123926920537, + "grad_norm": 3.116103508559128, + "learning_rate": 5.4703605803308146e-06, + "loss": 0.3694, + "step": 22075 + }, + { + "epoch": 0.48593440457847237, + "grad_norm": 1.6928771998517298, + "learning_rate": 5.4700057035030895e-06, + "loss": 0.3394, + "step": 22076 + }, + { + "epoch": 0.485956416464891, + "grad_norm": 1.962438734555462, + "learning_rate": 5.4696508242865875e-06, + "loss": 0.4409, + "step": 22077 + }, + { + "epoch": 0.4859784283513097, + "grad_norm": 1.596757765739077, + "learning_rate": 5.469295942683112e-06, + "loss": 0.3311, + "step": 22078 + }, + { + "epoch": 0.48600044023772837, + "grad_norm": 1.516631138525969, + "learning_rate": 5.468941058694464e-06, + "loss": 0.3356, + "step": 22079 + }, + { + "epoch": 0.486022452124147, + "grad_norm": 1.260613860943731, + "learning_rate": 5.4685861723224475e-06, + "loss": 0.3681, + "step": 22080 + }, + { + "epoch": 0.4860444640105657, + "grad_norm": 1.9463944776241644, + "learning_rate": 5.468231283568867e-06, + "loss": 0.3072, + "step": 22081 + }, + { + "epoch": 0.4860664758969844, + "grad_norm": 1.7286646888350303, + "learning_rate": 5.467876392435527e-06, + "loss": 0.3405, + "step": 22082 + }, + { + "epoch": 0.486088487783403, + "grad_norm": 0.6619769758653663, + "learning_rate": 5.467521498924231e-06, + "loss": 0.4785, + "step": 22083 + }, + { + "epoch": 0.4861104996698217, + "grad_norm": 1.2952643623659756, + "learning_rate": 5.467166603036781e-06, + "loss": 0.2736, + "step": 22084 + }, + { + "epoch": 0.4861325115562404, + "grad_norm": 1.2158198417963098, + "learning_rate": 5.4668117047749835e-06, + "loss": 0.3496, + "step": 22085 + }, + { + "epoch": 0.486154523442659, + "grad_norm": 1.7895290140322884, + "learning_rate": 5.466456804140639e-06, + "loss": 0.4128, + "step": 22086 + }, + { + "epoch": 0.4861765353290777, + "grad_norm": 0.6357758499777169, + "learning_rate": 5.466101901135553e-06, + "loss": 0.4788, + "step": 22087 + }, + { + "epoch": 0.4861985472154964, + "grad_norm": 1.5812000272777038, + "learning_rate": 5.465746995761531e-06, + "loss": 0.3732, + "step": 22088 + }, + { + "epoch": 0.48622055910191503, + "grad_norm": 1.2156800413891786, + "learning_rate": 5.465392088020374e-06, + "loss": 0.1873, + "step": 22089 + }, + { + "epoch": 0.4862425709883337, + "grad_norm": 1.9521098703212094, + "learning_rate": 5.4650371779138866e-06, + "loss": 0.4231, + "step": 22090 + }, + { + "epoch": 0.4862645828747524, + "grad_norm": 1.584725162463234, + "learning_rate": 5.464682265443874e-06, + "loss": 0.243, + "step": 22091 + }, + { + "epoch": 0.48628659476117103, + "grad_norm": 0.6485844015682426, + "learning_rate": 5.464327350612139e-06, + "loss": 0.4837, + "step": 22092 + }, + { + "epoch": 0.4863086066475897, + "grad_norm": 0.6279618317507957, + "learning_rate": 5.463972433420483e-06, + "loss": 0.4944, + "step": 22093 + }, + { + "epoch": 0.4863306185340084, + "grad_norm": 1.5253824326010659, + "learning_rate": 5.4636175138707134e-06, + "loss": 0.2928, + "step": 22094 + }, + { + "epoch": 0.48635263042042703, + "grad_norm": 1.5720977305773032, + "learning_rate": 5.4632625919646335e-06, + "loss": 0.2829, + "step": 22095 + }, + { + "epoch": 0.4863746423068457, + "grad_norm": 1.4674783441425008, + "learning_rate": 5.462907667704046e-06, + "loss": 0.3638, + "step": 22096 + }, + { + "epoch": 0.4863966541932644, + "grad_norm": 2.5268212200814353, + "learning_rate": 5.462552741090755e-06, + "loss": 0.3572, + "step": 22097 + }, + { + "epoch": 0.48641866607968304, + "grad_norm": 2.986815657552705, + "learning_rate": 5.4621978121265654e-06, + "loss": 0.3422, + "step": 22098 + }, + { + "epoch": 0.4864406779661017, + "grad_norm": 2.195425312950169, + "learning_rate": 5.461842880813278e-06, + "loss": 0.226, + "step": 22099 + }, + { + "epoch": 0.48646268985252034, + "grad_norm": 1.7100069195062237, + "learning_rate": 5.461487947152702e-06, + "loss": 0.2928, + "step": 22100 + }, + { + "epoch": 0.48648470173893904, + "grad_norm": 2.478084672082011, + "learning_rate": 5.461133011146638e-06, + "loss": 0.3875, + "step": 22101 + }, + { + "epoch": 0.4865067136253577, + "grad_norm": 1.4563205029420379, + "learning_rate": 5.46077807279689e-06, + "loss": 0.3615, + "step": 22102 + }, + { + "epoch": 0.48652872551177634, + "grad_norm": 1.26912421270915, + "learning_rate": 5.4604231321052615e-06, + "loss": 0.2451, + "step": 22103 + }, + { + "epoch": 0.48655073739819504, + "grad_norm": 1.4867780135763677, + "learning_rate": 5.460068189073558e-06, + "loss": 0.3218, + "step": 22104 + }, + { + "epoch": 0.4865727492846137, + "grad_norm": 1.9662484262778146, + "learning_rate": 5.4597132437035814e-06, + "loss": 0.309, + "step": 22105 + }, + { + "epoch": 0.48659476117103234, + "grad_norm": 1.1894101847271508, + "learning_rate": 5.4593582959971405e-06, + "loss": 0.2324, + "step": 22106 + }, + { + "epoch": 0.48661677305745105, + "grad_norm": 1.3154439176768573, + "learning_rate": 5.459003345956032e-06, + "loss": 0.3533, + "step": 22107 + }, + { + "epoch": 0.4866387849438697, + "grad_norm": 1.484115564881786, + "learning_rate": 5.458648393582065e-06, + "loss": 0.3586, + "step": 22108 + }, + { + "epoch": 0.48666079683028834, + "grad_norm": 0.6376173760195543, + "learning_rate": 5.458293438877043e-06, + "loss": 0.4866, + "step": 22109 + }, + { + "epoch": 0.48668280871670705, + "grad_norm": 1.3749162092203482, + "learning_rate": 5.457938481842768e-06, + "loss": 0.3407, + "step": 22110 + }, + { + "epoch": 0.4867048206031257, + "grad_norm": 0.6239926458839469, + "learning_rate": 5.457583522481048e-06, + "loss": 0.4646, + "step": 22111 + }, + { + "epoch": 0.48672683248954435, + "grad_norm": 1.5537963263862626, + "learning_rate": 5.457228560793683e-06, + "loss": 0.3653, + "step": 22112 + }, + { + "epoch": 0.486748844375963, + "grad_norm": 1.265363135567195, + "learning_rate": 5.456873596782479e-06, + "loss": 0.2993, + "step": 22113 + }, + { + "epoch": 0.4867708562623817, + "grad_norm": 1.436900748205748, + "learning_rate": 5.456518630449238e-06, + "loss": 0.2367, + "step": 22114 + }, + { + "epoch": 0.48679286814880035, + "grad_norm": 1.4559132859587578, + "learning_rate": 5.456163661795767e-06, + "loss": 0.3167, + "step": 22115 + }, + { + "epoch": 0.486814880035219, + "grad_norm": 1.6126321044587337, + "learning_rate": 5.455808690823869e-06, + "loss": 0.3434, + "step": 22116 + }, + { + "epoch": 0.4868368919216377, + "grad_norm": 1.814908489841785, + "learning_rate": 5.455453717535347e-06, + "loss": 0.2204, + "step": 22117 + }, + { + "epoch": 0.48685890380805635, + "grad_norm": 1.426269012703983, + "learning_rate": 5.455098741932006e-06, + "loss": 0.2873, + "step": 22118 + }, + { + "epoch": 0.486880915694475, + "grad_norm": 1.5019155322314053, + "learning_rate": 5.454743764015651e-06, + "loss": 0.2626, + "step": 22119 + }, + { + "epoch": 0.4869029275808937, + "grad_norm": 1.6111149862747158, + "learning_rate": 5.454388783788085e-06, + "loss": 0.2962, + "step": 22120 + }, + { + "epoch": 0.48692493946731236, + "grad_norm": 2.2415962637575064, + "learning_rate": 5.454033801251113e-06, + "loss": 0.3306, + "step": 22121 + }, + { + "epoch": 0.486946951353731, + "grad_norm": 1.5325778720657555, + "learning_rate": 5.453678816406539e-06, + "loss": 0.398, + "step": 22122 + }, + { + "epoch": 0.48696896324014965, + "grad_norm": 2.927388521018485, + "learning_rate": 5.4533238292561655e-06, + "loss": 0.2971, + "step": 22123 + }, + { + "epoch": 0.48699097512656836, + "grad_norm": 9.687559657421334, + "learning_rate": 5.4529688398018e-06, + "loss": 0.2749, + "step": 22124 + }, + { + "epoch": 0.487012987012987, + "grad_norm": 1.8197304467278002, + "learning_rate": 5.452613848045243e-06, + "loss": 0.2736, + "step": 22125 + }, + { + "epoch": 0.48703499889940566, + "grad_norm": 0.6299296316703689, + "learning_rate": 5.452258853988301e-06, + "loss": 0.4393, + "step": 22126 + }, + { + "epoch": 0.48705701078582436, + "grad_norm": 1.336938155414302, + "learning_rate": 5.451903857632779e-06, + "loss": 0.3864, + "step": 22127 + }, + { + "epoch": 0.487079022672243, + "grad_norm": 1.4330640478616041, + "learning_rate": 5.451548858980477e-06, + "loss": 0.2648, + "step": 22128 + }, + { + "epoch": 0.48710103455866166, + "grad_norm": 4.055364415197879, + "learning_rate": 5.451193858033206e-06, + "loss": 0.3671, + "step": 22129 + }, + { + "epoch": 0.48712304644508037, + "grad_norm": 1.3046375733611735, + "learning_rate": 5.450838854792764e-06, + "loss": 0.2473, + "step": 22130 + }, + { + "epoch": 0.487145058331499, + "grad_norm": 1.6277452693441188, + "learning_rate": 5.450483849260959e-06, + "loss": 0.344, + "step": 22131 + }, + { + "epoch": 0.48716707021791766, + "grad_norm": 13.94356059577083, + "learning_rate": 5.4501288414395945e-06, + "loss": 0.2495, + "step": 22132 + }, + { + "epoch": 0.48718908210433637, + "grad_norm": 1.691014815813356, + "learning_rate": 5.449773831330474e-06, + "loss": 0.2493, + "step": 22133 + }, + { + "epoch": 0.487211093990755, + "grad_norm": 1.1998530649393746, + "learning_rate": 5.449418818935402e-06, + "loss": 0.354, + "step": 22134 + }, + { + "epoch": 0.48723310587717367, + "grad_norm": 1.54029464639089, + "learning_rate": 5.449063804256184e-06, + "loss": 0.3634, + "step": 22135 + }, + { + "epoch": 0.4872551177635923, + "grad_norm": 2.245618362101049, + "learning_rate": 5.4487087872946224e-06, + "loss": 0.3355, + "step": 22136 + }, + { + "epoch": 0.487277129650011, + "grad_norm": 1.7937282007401945, + "learning_rate": 5.448353768052522e-06, + "loss": 0.2782, + "step": 22137 + }, + { + "epoch": 0.48729914153642967, + "grad_norm": 2.2848212653311073, + "learning_rate": 5.44799874653169e-06, + "loss": 0.35, + "step": 22138 + }, + { + "epoch": 0.4873211534228483, + "grad_norm": 1.4669519495625774, + "learning_rate": 5.447643722733927e-06, + "loss": 0.2604, + "step": 22139 + }, + { + "epoch": 0.487343165309267, + "grad_norm": 2.232826345282364, + "learning_rate": 5.44728869666104e-06, + "loss": 0.3882, + "step": 22140 + }, + { + "epoch": 0.4873651771956857, + "grad_norm": 0.6113630530158326, + "learning_rate": 5.446933668314831e-06, + "loss": 0.4704, + "step": 22141 + }, + { + "epoch": 0.4873871890821043, + "grad_norm": 0.6309293996124222, + "learning_rate": 5.446578637697107e-06, + "loss": 0.4801, + "step": 22142 + }, + { + "epoch": 0.487409200968523, + "grad_norm": 1.720663326058057, + "learning_rate": 5.44622360480967e-06, + "loss": 0.2966, + "step": 22143 + }, + { + "epoch": 0.4874312128549417, + "grad_norm": 1.1519661411866324, + "learning_rate": 5.445868569654327e-06, + "loss": 0.2732, + "step": 22144 + }, + { + "epoch": 0.4874532247413603, + "grad_norm": 2.0195896357430345, + "learning_rate": 5.44551353223288e-06, + "loss": 0.4438, + "step": 22145 + }, + { + "epoch": 0.487475236627779, + "grad_norm": 1.745120564427927, + "learning_rate": 5.445158492547133e-06, + "loss": 0.3711, + "step": 22146 + }, + { + "epoch": 0.4874972485141977, + "grad_norm": 1.3749178931851784, + "learning_rate": 5.444803450598896e-06, + "loss": 0.3234, + "step": 22147 + }, + { + "epoch": 0.4875192604006163, + "grad_norm": 1.7310110352773496, + "learning_rate": 5.444448406389966e-06, + "loss": 0.2243, + "step": 22148 + }, + { + "epoch": 0.487541272287035, + "grad_norm": 0.6256656248135916, + "learning_rate": 5.444093359922152e-06, + "loss": 0.4698, + "step": 22149 + }, + { + "epoch": 0.4875632841734537, + "grad_norm": 5.813857540774177, + "learning_rate": 5.4437383111972585e-06, + "loss": 0.3382, + "step": 22150 + }, + { + "epoch": 0.48758529605987233, + "grad_norm": 1.3626545769357439, + "learning_rate": 5.443383260217088e-06, + "loss": 0.2304, + "step": 22151 + }, + { + "epoch": 0.487607307946291, + "grad_norm": 1.327759356781787, + "learning_rate": 5.443028206983447e-06, + "loss": 0.2512, + "step": 22152 + }, + { + "epoch": 0.4876293198327097, + "grad_norm": 0.6198305904450635, + "learning_rate": 5.442673151498139e-06, + "loss": 0.4666, + "step": 22153 + }, + { + "epoch": 0.48765133171912833, + "grad_norm": 0.6268305484415341, + "learning_rate": 5.442318093762966e-06, + "loss": 0.4663, + "step": 22154 + }, + { + "epoch": 0.487673343605547, + "grad_norm": 1.4203034254478064, + "learning_rate": 5.441963033779738e-06, + "loss": 0.3629, + "step": 22155 + }, + { + "epoch": 0.4876953554919657, + "grad_norm": 1.5227583158305993, + "learning_rate": 5.441607971550255e-06, + "loss": 0.3791, + "step": 22156 + }, + { + "epoch": 0.48771736737838434, + "grad_norm": 1.3470170478056749, + "learning_rate": 5.4412529070763256e-06, + "loss": 0.2943, + "step": 22157 + }, + { + "epoch": 0.487739379264803, + "grad_norm": 1.8365719652437238, + "learning_rate": 5.44089784035975e-06, + "loss": 0.2907, + "step": 22158 + }, + { + "epoch": 0.48776139115122163, + "grad_norm": 1.1136965298679486, + "learning_rate": 5.440542771402335e-06, + "loss": 0.2619, + "step": 22159 + }, + { + "epoch": 0.48778340303764034, + "grad_norm": 1.4933770829559805, + "learning_rate": 5.440187700205886e-06, + "loss": 0.3793, + "step": 22160 + }, + { + "epoch": 0.487805414924059, + "grad_norm": 1.379529165289593, + "learning_rate": 5.4398326267722065e-06, + "loss": 0.3198, + "step": 22161 + }, + { + "epoch": 0.48782742681047764, + "grad_norm": 1.4229872271355137, + "learning_rate": 5.4394775511031014e-06, + "loss": 0.3154, + "step": 22162 + }, + { + "epoch": 0.48784943869689634, + "grad_norm": 1.8184346020448683, + "learning_rate": 5.439122473200376e-06, + "loss": 0.3343, + "step": 22163 + }, + { + "epoch": 0.487871450583315, + "grad_norm": 2.0048475963392023, + "learning_rate": 5.438767393065832e-06, + "loss": 0.3581, + "step": 22164 + }, + { + "epoch": 0.48789346246973364, + "grad_norm": 0.6550323934503525, + "learning_rate": 5.43841231070128e-06, + "loss": 0.5067, + "step": 22165 + }, + { + "epoch": 0.48791547435615235, + "grad_norm": 0.6935806540496704, + "learning_rate": 5.438057226108519e-06, + "loss": 0.511, + "step": 22166 + }, + { + "epoch": 0.487937486242571, + "grad_norm": 1.3685636190464472, + "learning_rate": 5.437702139289356e-06, + "loss": 0.3373, + "step": 22167 + }, + { + "epoch": 0.48795949812898964, + "grad_norm": 1.35454605952711, + "learning_rate": 5.437347050245596e-06, + "loss": 0.3482, + "step": 22168 + }, + { + "epoch": 0.4879815100154083, + "grad_norm": 1.8957946809783222, + "learning_rate": 5.436991958979041e-06, + "loss": 0.3244, + "step": 22169 + }, + { + "epoch": 0.488003521901827, + "grad_norm": 1.4383168499206544, + "learning_rate": 5.436636865491501e-06, + "loss": 0.2572, + "step": 22170 + }, + { + "epoch": 0.48802553378824565, + "grad_norm": 1.425703562579318, + "learning_rate": 5.436281769784777e-06, + "loss": 0.3275, + "step": 22171 + }, + { + "epoch": 0.4880475456746643, + "grad_norm": 1.4154458734521336, + "learning_rate": 5.4359266718606725e-06, + "loss": 0.2667, + "step": 22172 + }, + { + "epoch": 0.488069557561083, + "grad_norm": 3.555471030394711, + "learning_rate": 5.4355715717209966e-06, + "loss": 0.2677, + "step": 22173 + }, + { + "epoch": 0.48809156944750165, + "grad_norm": 2.0686963418476227, + "learning_rate": 5.435216469367552e-06, + "loss": 0.2986, + "step": 22174 + }, + { + "epoch": 0.4881135813339203, + "grad_norm": 1.2862264978622477, + "learning_rate": 5.4348613648021405e-06, + "loss": 0.3312, + "step": 22175 + }, + { + "epoch": 0.488135593220339, + "grad_norm": 2.0368507530401097, + "learning_rate": 5.434506258026572e-06, + "loss": 0.2844, + "step": 22176 + }, + { + "epoch": 0.48815760510675765, + "grad_norm": 1.8445892737481435, + "learning_rate": 5.4341511490426466e-06, + "loss": 0.3434, + "step": 22177 + }, + { + "epoch": 0.4881796169931763, + "grad_norm": 0.7227671637376194, + "learning_rate": 5.433796037852172e-06, + "loss": 0.4783, + "step": 22178 + }, + { + "epoch": 0.488201628879595, + "grad_norm": 1.657193518609371, + "learning_rate": 5.433440924456954e-06, + "loss": 0.3592, + "step": 22179 + }, + { + "epoch": 0.48822364076601366, + "grad_norm": 0.6846696097361761, + "learning_rate": 5.433085808858796e-06, + "loss": 0.5053, + "step": 22180 + }, + { + "epoch": 0.4882456526524323, + "grad_norm": 1.2985864994886087, + "learning_rate": 5.432730691059501e-06, + "loss": 0.3146, + "step": 22181 + }, + { + "epoch": 0.48826766453885095, + "grad_norm": 1.5200623730824954, + "learning_rate": 5.432375571060876e-06, + "loss": 0.3215, + "step": 22182 + }, + { + "epoch": 0.48828967642526966, + "grad_norm": 1.3286529595741685, + "learning_rate": 5.432020448864728e-06, + "loss": 0.3108, + "step": 22183 + }, + { + "epoch": 0.4883116883116883, + "grad_norm": 1.3609720118660726, + "learning_rate": 5.431665324472855e-06, + "loss": 0.3011, + "step": 22184 + }, + { + "epoch": 0.48833370019810696, + "grad_norm": 0.639334190526405, + "learning_rate": 5.43131019788707e-06, + "loss": 0.4452, + "step": 22185 + }, + { + "epoch": 0.48835571208452566, + "grad_norm": 0.64499683015481, + "learning_rate": 5.430955069109173e-06, + "loss": 0.464, + "step": 22186 + }, + { + "epoch": 0.4883777239709443, + "grad_norm": 1.2821523251466609, + "learning_rate": 5.430599938140969e-06, + "loss": 0.2575, + "step": 22187 + }, + { + "epoch": 0.48839973585736296, + "grad_norm": 1.32711378505681, + "learning_rate": 5.430244804984266e-06, + "loss": 0.28, + "step": 22188 + }, + { + "epoch": 0.48842174774378166, + "grad_norm": 1.4592982412436029, + "learning_rate": 5.429889669640867e-06, + "loss": 0.3077, + "step": 22189 + }, + { + "epoch": 0.4884437596302003, + "grad_norm": 1.4994201497025157, + "learning_rate": 5.429534532112575e-06, + "loss": 0.3162, + "step": 22190 + }, + { + "epoch": 0.48846577151661896, + "grad_norm": 1.6741269924079634, + "learning_rate": 5.429179392401198e-06, + "loss": 0.2631, + "step": 22191 + }, + { + "epoch": 0.48848778340303767, + "grad_norm": 1.9919025733347977, + "learning_rate": 5.42882425050854e-06, + "loss": 0.3151, + "step": 22192 + }, + { + "epoch": 0.4885097952894563, + "grad_norm": 1.8266758586710239, + "learning_rate": 5.428469106436406e-06, + "loss": 0.3885, + "step": 22193 + }, + { + "epoch": 0.48853180717587497, + "grad_norm": 1.5087573857436856, + "learning_rate": 5.428113960186601e-06, + "loss": 0.3814, + "step": 22194 + }, + { + "epoch": 0.4885538190622936, + "grad_norm": 1.8337708854237136, + "learning_rate": 5.427758811760928e-06, + "loss": 0.3806, + "step": 22195 + }, + { + "epoch": 0.4885758309487123, + "grad_norm": 2.0704858714394403, + "learning_rate": 5.427403661161197e-06, + "loss": 0.2324, + "step": 22196 + }, + { + "epoch": 0.48859784283513097, + "grad_norm": 0.6099863753809133, + "learning_rate": 5.4270485083892075e-06, + "loss": 0.4548, + "step": 22197 + }, + { + "epoch": 0.4886198547215496, + "grad_norm": 1.6760155170708357, + "learning_rate": 5.426693353446769e-06, + "loss": 0.3051, + "step": 22198 + }, + { + "epoch": 0.4886418666079683, + "grad_norm": 0.6336223678613413, + "learning_rate": 5.4263381963356835e-06, + "loss": 0.4943, + "step": 22199 + }, + { + "epoch": 0.48866387849438697, + "grad_norm": 1.648438473280906, + "learning_rate": 5.425983037057756e-06, + "loss": 0.3136, + "step": 22200 + }, + { + "epoch": 0.4886858903808056, + "grad_norm": 1.384459694945932, + "learning_rate": 5.425627875614794e-06, + "loss": 0.3447, + "step": 22201 + }, + { + "epoch": 0.4887079022672243, + "grad_norm": 1.3868769439178192, + "learning_rate": 5.425272712008601e-06, + "loss": 0.4252, + "step": 22202 + }, + { + "epoch": 0.488729914153643, + "grad_norm": 1.947990697139927, + "learning_rate": 5.424917546240983e-06, + "loss": 0.3396, + "step": 22203 + }, + { + "epoch": 0.4887519260400616, + "grad_norm": 1.3641903163027678, + "learning_rate": 5.424562378313745e-06, + "loss": 0.2905, + "step": 22204 + }, + { + "epoch": 0.4887739379264803, + "grad_norm": 1.845735477384447, + "learning_rate": 5.42420720822869e-06, + "loss": 0.2657, + "step": 22205 + }, + { + "epoch": 0.488795949812899, + "grad_norm": 1.5428865823447078, + "learning_rate": 5.4238520359876265e-06, + "loss": 0.3576, + "step": 22206 + }, + { + "epoch": 0.4888179616993176, + "grad_norm": 2.030267750943702, + "learning_rate": 5.423496861592357e-06, + "loss": 0.2898, + "step": 22207 + }, + { + "epoch": 0.4888399735857363, + "grad_norm": 0.6174517165482261, + "learning_rate": 5.423141685044687e-06, + "loss": 0.4611, + "step": 22208 + }, + { + "epoch": 0.488861985472155, + "grad_norm": 1.3888567450470362, + "learning_rate": 5.422786506346422e-06, + "loss": 0.2461, + "step": 22209 + }, + { + "epoch": 0.48888399735857363, + "grad_norm": 1.376315706388292, + "learning_rate": 5.422431325499369e-06, + "loss": 0.3435, + "step": 22210 + }, + { + "epoch": 0.4889060092449923, + "grad_norm": 3.163713890598166, + "learning_rate": 5.42207614250533e-06, + "loss": 0.341, + "step": 22211 + }, + { + "epoch": 0.488928021131411, + "grad_norm": 1.5257017095922514, + "learning_rate": 5.421720957366114e-06, + "loss": 0.3351, + "step": 22212 + }, + { + "epoch": 0.48895003301782963, + "grad_norm": 1.8549803602783896, + "learning_rate": 5.421365770083521e-06, + "loss": 0.3193, + "step": 22213 + }, + { + "epoch": 0.4889720449042483, + "grad_norm": 1.375507783532882, + "learning_rate": 5.421010580659362e-06, + "loss": 0.3775, + "step": 22214 + }, + { + "epoch": 0.488994056790667, + "grad_norm": 1.5045868837448029, + "learning_rate": 5.420655389095437e-06, + "loss": 0.3126, + "step": 22215 + }, + { + "epoch": 0.48901606867708564, + "grad_norm": 1.5371299820411104, + "learning_rate": 5.420300195393556e-06, + "loss": 0.4565, + "step": 22216 + }, + { + "epoch": 0.4890380805635043, + "grad_norm": 1.4854431267043402, + "learning_rate": 5.4199449995555205e-06, + "loss": 0.2729, + "step": 22217 + }, + { + "epoch": 0.48906009244992293, + "grad_norm": 1.4393556768789628, + "learning_rate": 5.419589801583137e-06, + "loss": 0.3046, + "step": 22218 + }, + { + "epoch": 0.48908210433634164, + "grad_norm": 1.1934408319535637, + "learning_rate": 5.419234601478211e-06, + "loss": 0.2601, + "step": 22219 + }, + { + "epoch": 0.4891041162227603, + "grad_norm": 2.054019568292817, + "learning_rate": 5.418879399242549e-06, + "loss": 0.3276, + "step": 22220 + }, + { + "epoch": 0.48912612810917894, + "grad_norm": 3.3305393943334924, + "learning_rate": 5.418524194877954e-06, + "loss": 0.4372, + "step": 22221 + }, + { + "epoch": 0.48914813999559764, + "grad_norm": 1.6759591219687628, + "learning_rate": 5.418168988386233e-06, + "loss": 0.3393, + "step": 22222 + }, + { + "epoch": 0.4891701518820163, + "grad_norm": 1.1825376901936384, + "learning_rate": 5.41781377976919e-06, + "loss": 0.2875, + "step": 22223 + }, + { + "epoch": 0.48919216376843494, + "grad_norm": 1.5727352909868488, + "learning_rate": 5.417458569028633e-06, + "loss": 0.3428, + "step": 22224 + }, + { + "epoch": 0.48921417565485364, + "grad_norm": 1.5572066977796895, + "learning_rate": 5.417103356166363e-06, + "loss": 0.3073, + "step": 22225 + }, + { + "epoch": 0.4892361875412723, + "grad_norm": 1.7850691862408572, + "learning_rate": 5.416748141184188e-06, + "loss": 0.3189, + "step": 22226 + }, + { + "epoch": 0.48925819942769094, + "grad_norm": 0.6892915568177135, + "learning_rate": 5.416392924083914e-06, + "loss": 0.4977, + "step": 22227 + }, + { + "epoch": 0.4892802113141096, + "grad_norm": 1.9834417760780554, + "learning_rate": 5.4160377048673455e-06, + "loss": 0.2852, + "step": 22228 + }, + { + "epoch": 0.4893022232005283, + "grad_norm": 1.826810104388874, + "learning_rate": 5.415682483536288e-06, + "loss": 0.3071, + "step": 22229 + }, + { + "epoch": 0.48932423508694695, + "grad_norm": 1.9682180649843406, + "learning_rate": 5.415327260092546e-06, + "loss": 0.3252, + "step": 22230 + }, + { + "epoch": 0.4893462469733656, + "grad_norm": 1.6993680800233641, + "learning_rate": 5.414972034537924e-06, + "loss": 0.3053, + "step": 22231 + }, + { + "epoch": 0.4893682588597843, + "grad_norm": 1.0990711014755208, + "learning_rate": 5.414616806874231e-06, + "loss": 0.2859, + "step": 22232 + }, + { + "epoch": 0.48939027074620295, + "grad_norm": 1.3315111220107356, + "learning_rate": 5.414261577103272e-06, + "loss": 0.3513, + "step": 22233 + }, + { + "epoch": 0.4894122826326216, + "grad_norm": 0.5966884876449128, + "learning_rate": 5.413906345226848e-06, + "loss": 0.4673, + "step": 22234 + }, + { + "epoch": 0.4894342945190403, + "grad_norm": 1.7242145609538158, + "learning_rate": 5.413551111246769e-06, + "loss": 0.3449, + "step": 22235 + }, + { + "epoch": 0.48945630640545895, + "grad_norm": 1.5010030929569051, + "learning_rate": 5.413195875164837e-06, + "loss": 0.3349, + "step": 22236 + }, + { + "epoch": 0.4894783182918776, + "grad_norm": 0.6254483516672702, + "learning_rate": 5.412840636982861e-06, + "loss": 0.4834, + "step": 22237 + }, + { + "epoch": 0.4895003301782963, + "grad_norm": 1.1395330581640375, + "learning_rate": 5.4124853967026446e-06, + "loss": 0.2229, + "step": 22238 + }, + { + "epoch": 0.48952234206471495, + "grad_norm": 1.5023933410000465, + "learning_rate": 5.412130154325993e-06, + "loss": 0.359, + "step": 22239 + }, + { + "epoch": 0.4895443539511336, + "grad_norm": 1.5972956555994178, + "learning_rate": 5.411774909854712e-06, + "loss": 0.3323, + "step": 22240 + }, + { + "epoch": 0.48956636583755225, + "grad_norm": 1.7868136001661583, + "learning_rate": 5.411419663290607e-06, + "loss": 0.3263, + "step": 22241 + }, + { + "epoch": 0.48958837772397096, + "grad_norm": 1.3755066167722387, + "learning_rate": 5.411064414635486e-06, + "loss": 0.2741, + "step": 22242 + }, + { + "epoch": 0.4896103896103896, + "grad_norm": 1.9130588959377084, + "learning_rate": 5.41070916389115e-06, + "loss": 0.4145, + "step": 22243 + }, + { + "epoch": 0.48963240149680826, + "grad_norm": 8.783813266367554, + "learning_rate": 5.410353911059408e-06, + "loss": 0.2624, + "step": 22244 + }, + { + "epoch": 0.48965441338322696, + "grad_norm": 1.794424517283639, + "learning_rate": 5.409998656142064e-06, + "loss": 0.2964, + "step": 22245 + }, + { + "epoch": 0.4896764252696456, + "grad_norm": 1.3994576830134169, + "learning_rate": 5.409643399140924e-06, + "loss": 0.3499, + "step": 22246 + }, + { + "epoch": 0.48969843715606426, + "grad_norm": 1.4239542013177922, + "learning_rate": 5.409288140057792e-06, + "loss": 0.2873, + "step": 22247 + }, + { + "epoch": 0.48972044904248296, + "grad_norm": 1.2549563268409607, + "learning_rate": 5.408932878894477e-06, + "loss": 0.3275, + "step": 22248 + }, + { + "epoch": 0.4897424609289016, + "grad_norm": 4.083440794752122, + "learning_rate": 5.408577615652781e-06, + "loss": 0.2829, + "step": 22249 + }, + { + "epoch": 0.48976447281532026, + "grad_norm": 1.7451350854146013, + "learning_rate": 5.408222350334513e-06, + "loss": 0.3901, + "step": 22250 + }, + { + "epoch": 0.48978648470173897, + "grad_norm": 0.6319638656926407, + "learning_rate": 5.407867082941477e-06, + "loss": 0.4578, + "step": 22251 + }, + { + "epoch": 0.4898084965881576, + "grad_norm": 1.5578779925538087, + "learning_rate": 5.407511813475479e-06, + "loss": 0.367, + "step": 22252 + }, + { + "epoch": 0.48983050847457626, + "grad_norm": 0.6315493224370207, + "learning_rate": 5.407156541938323e-06, + "loss": 0.4788, + "step": 22253 + }, + { + "epoch": 0.4898525203609949, + "grad_norm": 1.6213549984955236, + "learning_rate": 5.4068012683318146e-06, + "loss": 0.2692, + "step": 22254 + }, + { + "epoch": 0.4898745322474136, + "grad_norm": 0.6187396124385116, + "learning_rate": 5.406445992657763e-06, + "loss": 0.4648, + "step": 22255 + }, + { + "epoch": 0.48989654413383227, + "grad_norm": 0.6109685745261912, + "learning_rate": 5.406090714917971e-06, + "loss": 0.4705, + "step": 22256 + }, + { + "epoch": 0.4899185560202509, + "grad_norm": 1.3125669334799488, + "learning_rate": 5.405735435114244e-06, + "loss": 0.3398, + "step": 22257 + }, + { + "epoch": 0.4899405679066696, + "grad_norm": 2.7029268665142356, + "learning_rate": 5.405380153248389e-06, + "loss": 0.3942, + "step": 22258 + }, + { + "epoch": 0.48996257979308827, + "grad_norm": 0.6505019775372995, + "learning_rate": 5.4050248693222116e-06, + "loss": 0.4844, + "step": 22259 + }, + { + "epoch": 0.4899845916795069, + "grad_norm": 1.6547916687761615, + "learning_rate": 5.404669583337518e-06, + "loss": 0.2893, + "step": 22260 + }, + { + "epoch": 0.4900066035659256, + "grad_norm": 1.670784822388627, + "learning_rate": 5.4043142952961125e-06, + "loss": 0.3528, + "step": 22261 + }, + { + "epoch": 0.4900286154523443, + "grad_norm": 1.5766178366361066, + "learning_rate": 5.4039590051998e-06, + "loss": 0.3189, + "step": 22262 + }, + { + "epoch": 0.4900506273387629, + "grad_norm": 1.3566366516442308, + "learning_rate": 5.40360371305039e-06, + "loss": 0.3514, + "step": 22263 + }, + { + "epoch": 0.49007263922518157, + "grad_norm": 1.1547654685256674, + "learning_rate": 5.403248418849684e-06, + "loss": 0.3691, + "step": 22264 + }, + { + "epoch": 0.4900946511116003, + "grad_norm": 1.7352546094272252, + "learning_rate": 5.402893122599491e-06, + "loss": 0.3343, + "step": 22265 + }, + { + "epoch": 0.4901166629980189, + "grad_norm": 2.3544186090040675, + "learning_rate": 5.402537824301614e-06, + "loss": 0.3043, + "step": 22266 + }, + { + "epoch": 0.4901386748844376, + "grad_norm": 1.6204955625050805, + "learning_rate": 5.40218252395786e-06, + "loss": 0.3144, + "step": 22267 + }, + { + "epoch": 0.4901606867708563, + "grad_norm": 2.53825398543979, + "learning_rate": 5.401827221570037e-06, + "loss": 0.3019, + "step": 22268 + }, + { + "epoch": 0.49018269865727493, + "grad_norm": 1.258908255769937, + "learning_rate": 5.401471917139948e-06, + "loss": 0.3484, + "step": 22269 + }, + { + "epoch": 0.4902047105436936, + "grad_norm": 2.56427018417972, + "learning_rate": 5.4011166106694006e-06, + "loss": 0.2877, + "step": 22270 + }, + { + "epoch": 0.4902267224301123, + "grad_norm": 0.6403582918865985, + "learning_rate": 5.400761302160198e-06, + "loss": 0.4856, + "step": 22271 + }, + { + "epoch": 0.49024873431653093, + "grad_norm": 1.3582547998792471, + "learning_rate": 5.400405991614147e-06, + "loss": 0.283, + "step": 22272 + }, + { + "epoch": 0.4902707462029496, + "grad_norm": 1.427735134690509, + "learning_rate": 5.400050679033057e-06, + "loss": 0.293, + "step": 22273 + }, + { + "epoch": 0.4902927580893683, + "grad_norm": 2.4867276772736933, + "learning_rate": 5.399695364418729e-06, + "loss": 0.3162, + "step": 22274 + }, + { + "epoch": 0.49031476997578693, + "grad_norm": 1.828671671260169, + "learning_rate": 5.399340047772972e-06, + "loss": 0.2756, + "step": 22275 + }, + { + "epoch": 0.4903367818622056, + "grad_norm": 1.0719224777933634, + "learning_rate": 5.3989847290975895e-06, + "loss": 0.3023, + "step": 22276 + }, + { + "epoch": 0.49035879374862423, + "grad_norm": 1.4761762749678622, + "learning_rate": 5.398629408394389e-06, + "loss": 0.2987, + "step": 22277 + }, + { + "epoch": 0.49038080563504294, + "grad_norm": 4.031448421388828, + "learning_rate": 5.398274085665176e-06, + "loss": 0.3114, + "step": 22278 + }, + { + "epoch": 0.4904028175214616, + "grad_norm": 1.3344574666968394, + "learning_rate": 5.397918760911759e-06, + "loss": 0.2538, + "step": 22279 + }, + { + "epoch": 0.49042482940788024, + "grad_norm": 1.6506044180634178, + "learning_rate": 5.3975634341359385e-06, + "loss": 0.3991, + "step": 22280 + }, + { + "epoch": 0.49044684129429894, + "grad_norm": 5.640973119361925, + "learning_rate": 5.3972081053395245e-06, + "loss": 0.2955, + "step": 22281 + }, + { + "epoch": 0.4904688531807176, + "grad_norm": 1.570713240261025, + "learning_rate": 5.39685277452432e-06, + "loss": 0.3151, + "step": 22282 + }, + { + "epoch": 0.49049086506713624, + "grad_norm": 1.6165607423608344, + "learning_rate": 5.396497441692135e-06, + "loss": 0.3801, + "step": 22283 + }, + { + "epoch": 0.49051287695355494, + "grad_norm": 1.6069264956556282, + "learning_rate": 5.396142106844773e-06, + "loss": 0.2568, + "step": 22284 + }, + { + "epoch": 0.4905348888399736, + "grad_norm": 0.6193043602856111, + "learning_rate": 5.395786769984037e-06, + "loss": 0.4835, + "step": 22285 + }, + { + "epoch": 0.49055690072639224, + "grad_norm": 1.1858420646915604, + "learning_rate": 5.395431431111739e-06, + "loss": 0.2679, + "step": 22286 + }, + { + "epoch": 0.4905789126128109, + "grad_norm": 1.1434550532872878, + "learning_rate": 5.395076090229682e-06, + "loss": 0.2973, + "step": 22287 + }, + { + "epoch": 0.4906009244992296, + "grad_norm": 1.9536676425786623, + "learning_rate": 5.394720747339671e-06, + "loss": 0.3127, + "step": 22288 + }, + { + "epoch": 0.49062293638564825, + "grad_norm": 1.5244626953064033, + "learning_rate": 5.394365402443514e-06, + "loss": 0.352, + "step": 22289 + }, + { + "epoch": 0.4906449482720669, + "grad_norm": 1.5615231861126102, + "learning_rate": 5.394010055543015e-06, + "loss": 0.3786, + "step": 22290 + }, + { + "epoch": 0.4906669601584856, + "grad_norm": 1.5960834841496718, + "learning_rate": 5.393654706639983e-06, + "loss": 0.4024, + "step": 22291 + }, + { + "epoch": 0.49068897204490425, + "grad_norm": 0.626721496552089, + "learning_rate": 5.3932993557362215e-06, + "loss": 0.4794, + "step": 22292 + }, + { + "epoch": 0.4907109839313229, + "grad_norm": 1.2970528845225293, + "learning_rate": 5.392944002833537e-06, + "loss": 0.2446, + "step": 22293 + }, + { + "epoch": 0.4907329958177416, + "grad_norm": 3.189794335636496, + "learning_rate": 5.392588647933737e-06, + "loss": 0.3352, + "step": 22294 + }, + { + "epoch": 0.49075500770416025, + "grad_norm": 1.515617524007322, + "learning_rate": 5.392233291038623e-06, + "loss": 0.3812, + "step": 22295 + }, + { + "epoch": 0.4907770195905789, + "grad_norm": 1.519502330921022, + "learning_rate": 5.391877932150007e-06, + "loss": 0.3109, + "step": 22296 + }, + { + "epoch": 0.4907990314769976, + "grad_norm": 1.7636526378594506, + "learning_rate": 5.391522571269694e-06, + "loss": 0.3846, + "step": 22297 + }, + { + "epoch": 0.49082104336341625, + "grad_norm": 1.4090954717360735, + "learning_rate": 5.391167208399486e-06, + "loss": 0.2725, + "step": 22298 + }, + { + "epoch": 0.4908430552498349, + "grad_norm": 1.7292721432436309, + "learning_rate": 5.390811843541193e-06, + "loss": 0.26, + "step": 22299 + }, + { + "epoch": 0.49086506713625355, + "grad_norm": 1.5005933577752708, + "learning_rate": 5.39045647669662e-06, + "loss": 0.3042, + "step": 22300 + }, + { + "epoch": 0.49088707902267226, + "grad_norm": 1.1606168533144343, + "learning_rate": 5.39010110786757e-06, + "loss": 0.2879, + "step": 22301 + }, + { + "epoch": 0.4909090909090909, + "grad_norm": 0.6198879711679026, + "learning_rate": 5.389745737055857e-06, + "loss": 0.4922, + "step": 22302 + }, + { + "epoch": 0.49093110279550956, + "grad_norm": 1.4028298469329654, + "learning_rate": 5.38939036426328e-06, + "loss": 0.3204, + "step": 22303 + }, + { + "epoch": 0.49095311468192826, + "grad_norm": 1.820218665383878, + "learning_rate": 5.389034989491647e-06, + "loss": 0.3697, + "step": 22304 + }, + { + "epoch": 0.4909751265683469, + "grad_norm": 1.2854957022634224, + "learning_rate": 5.388679612742766e-06, + "loss": 0.2553, + "step": 22305 + }, + { + "epoch": 0.49099713845476556, + "grad_norm": 0.6413334071066269, + "learning_rate": 5.388324234018441e-06, + "loss": 0.4984, + "step": 22306 + }, + { + "epoch": 0.49101915034118426, + "grad_norm": 1.923076759224958, + "learning_rate": 5.387968853320479e-06, + "loss": 0.3557, + "step": 22307 + }, + { + "epoch": 0.4910411622276029, + "grad_norm": 3.050195129711785, + "learning_rate": 5.3876134706506855e-06, + "loss": 0.2596, + "step": 22308 + }, + { + "epoch": 0.49106317411402156, + "grad_norm": 1.3660667444468988, + "learning_rate": 5.387258086010869e-06, + "loss": 0.3331, + "step": 22309 + }, + { + "epoch": 0.4910851860004402, + "grad_norm": 1.876537188996713, + "learning_rate": 5.386902699402834e-06, + "loss": 0.3057, + "step": 22310 + }, + { + "epoch": 0.4911071978868589, + "grad_norm": 1.230047970052013, + "learning_rate": 5.386547310828386e-06, + "loss": 0.3511, + "step": 22311 + }, + { + "epoch": 0.49112920977327756, + "grad_norm": 0.6171595363568733, + "learning_rate": 5.3861919202893334e-06, + "loss": 0.4406, + "step": 22312 + }, + { + "epoch": 0.4911512216596962, + "grad_norm": 1.6477513784545186, + "learning_rate": 5.385836527787479e-06, + "loss": 0.3499, + "step": 22313 + }, + { + "epoch": 0.4911732335461149, + "grad_norm": 1.9502897861450628, + "learning_rate": 5.385481133324633e-06, + "loss": 0.3536, + "step": 22314 + }, + { + "epoch": 0.49119524543253357, + "grad_norm": 2.348350188395834, + "learning_rate": 5.385125736902601e-06, + "loss": 0.3088, + "step": 22315 + }, + { + "epoch": 0.4912172573189522, + "grad_norm": 1.2685509822788272, + "learning_rate": 5.384770338523186e-06, + "loss": 0.301, + "step": 22316 + }, + { + "epoch": 0.4912392692053709, + "grad_norm": 1.4797409451854084, + "learning_rate": 5.384414938188198e-06, + "loss": 0.3485, + "step": 22317 + }, + { + "epoch": 0.49126128109178957, + "grad_norm": 2.182794111894329, + "learning_rate": 5.3840595358994415e-06, + "loss": 0.3768, + "step": 22318 + }, + { + "epoch": 0.4912832929782082, + "grad_norm": 1.8311803960314808, + "learning_rate": 5.383704131658721e-06, + "loss": 0.4041, + "step": 22319 + }, + { + "epoch": 0.4913053048646269, + "grad_norm": 1.2725007213302066, + "learning_rate": 5.3833487254678496e-06, + "loss": 0.2242, + "step": 22320 + }, + { + "epoch": 0.4913273167510456, + "grad_norm": 1.8929568315266063, + "learning_rate": 5.382993317328625e-06, + "loss": 0.4115, + "step": 22321 + }, + { + "epoch": 0.4913493286374642, + "grad_norm": 1.5662636943413504, + "learning_rate": 5.382637907242859e-06, + "loss": 0.3663, + "step": 22322 + }, + { + "epoch": 0.49137134052388287, + "grad_norm": 0.6126228285792957, + "learning_rate": 5.382282495212356e-06, + "loss": 0.4744, + "step": 22323 + }, + { + "epoch": 0.4913933524103016, + "grad_norm": 2.002766021027288, + "learning_rate": 5.381927081238924e-06, + "loss": 0.2331, + "step": 22324 + }, + { + "epoch": 0.4914153642967202, + "grad_norm": 1.433049546090839, + "learning_rate": 5.381571665324367e-06, + "loss": 0.3805, + "step": 22325 + }, + { + "epoch": 0.4914373761831389, + "grad_norm": 1.045719649785182, + "learning_rate": 5.3812162474704925e-06, + "loss": 0.2494, + "step": 22326 + }, + { + "epoch": 0.4914593880695576, + "grad_norm": 1.23070546559302, + "learning_rate": 5.380860827679108e-06, + "loss": 0.2376, + "step": 22327 + }, + { + "epoch": 0.49148139995597623, + "grad_norm": 1.3215812519588581, + "learning_rate": 5.380505405952019e-06, + "loss": 0.3077, + "step": 22328 + }, + { + "epoch": 0.4915034118423949, + "grad_norm": 0.6451666557879632, + "learning_rate": 5.380149982291031e-06, + "loss": 0.4972, + "step": 22329 + }, + { + "epoch": 0.4915254237288136, + "grad_norm": 0.6445624045599906, + "learning_rate": 5.379794556697952e-06, + "loss": 0.4753, + "step": 22330 + }, + { + "epoch": 0.49154743561523223, + "grad_norm": 1.4193608502260602, + "learning_rate": 5.379439129174586e-06, + "loss": 0.312, + "step": 22331 + }, + { + "epoch": 0.4915694475016509, + "grad_norm": 1.9003444275149433, + "learning_rate": 5.379083699722743e-06, + "loss": 0.3192, + "step": 22332 + }, + { + "epoch": 0.4915914593880696, + "grad_norm": 0.6117258818412709, + "learning_rate": 5.378728268344228e-06, + "loss": 0.4846, + "step": 22333 + }, + { + "epoch": 0.49161347127448823, + "grad_norm": 0.6118596599665999, + "learning_rate": 5.378372835040845e-06, + "loss": 0.4909, + "step": 22334 + }, + { + "epoch": 0.4916354831609069, + "grad_norm": 2.4005920620123837, + "learning_rate": 5.378017399814403e-06, + "loss": 0.2021, + "step": 22335 + }, + { + "epoch": 0.49165749504732553, + "grad_norm": 0.6112100822551692, + "learning_rate": 5.377661962666708e-06, + "loss": 0.4797, + "step": 22336 + }, + { + "epoch": 0.49167950693374424, + "grad_norm": 1.319555097887024, + "learning_rate": 5.377306523599566e-06, + "loss": 0.2883, + "step": 22337 + }, + { + "epoch": 0.4917015188201629, + "grad_norm": 0.604868317354969, + "learning_rate": 5.376951082614786e-06, + "loss": 0.4431, + "step": 22338 + }, + { + "epoch": 0.49172353070658154, + "grad_norm": 1.2282965714309502, + "learning_rate": 5.37659563971417e-06, + "loss": 0.2335, + "step": 22339 + }, + { + "epoch": 0.49174554259300024, + "grad_norm": 1.492916406247757, + "learning_rate": 5.376240194899529e-06, + "loss": 0.3301, + "step": 22340 + }, + { + "epoch": 0.4917675544794189, + "grad_norm": 1.3088990709344042, + "learning_rate": 5.3758847481726675e-06, + "loss": 0.3215, + "step": 22341 + }, + { + "epoch": 0.49178956636583754, + "grad_norm": 11.091646953550022, + "learning_rate": 5.375529299535391e-06, + "loss": 0.2356, + "step": 22342 + }, + { + "epoch": 0.49181157825225624, + "grad_norm": 1.7991021641551659, + "learning_rate": 5.375173848989508e-06, + "loss": 0.3941, + "step": 22343 + }, + { + "epoch": 0.4918335901386749, + "grad_norm": 1.2083390705033175, + "learning_rate": 5.374818396536822e-06, + "loss": 0.31, + "step": 22344 + }, + { + "epoch": 0.49185560202509354, + "grad_norm": 4.583590288338069, + "learning_rate": 5.374462942179144e-06, + "loss": 0.3175, + "step": 22345 + }, + { + "epoch": 0.4918776139115122, + "grad_norm": 1.3402278634585325, + "learning_rate": 5.374107485918278e-06, + "loss": 0.3262, + "step": 22346 + }, + { + "epoch": 0.4918996257979309, + "grad_norm": 1.601224555518088, + "learning_rate": 5.373752027756031e-06, + "loss": 0.2608, + "step": 22347 + }, + { + "epoch": 0.49192163768434954, + "grad_norm": 1.4347482054509355, + "learning_rate": 5.373396567694209e-06, + "loss": 0.3399, + "step": 22348 + }, + { + "epoch": 0.4919436495707682, + "grad_norm": 0.6201076502424384, + "learning_rate": 5.373041105734618e-06, + "loss": 0.4579, + "step": 22349 + }, + { + "epoch": 0.4919656614571869, + "grad_norm": 1.46207784997565, + "learning_rate": 5.372685641879067e-06, + "loss": 0.4071, + "step": 22350 + }, + { + "epoch": 0.49198767334360555, + "grad_norm": 0.6389076236927653, + "learning_rate": 5.372330176129361e-06, + "loss": 0.4845, + "step": 22351 + }, + { + "epoch": 0.4920096852300242, + "grad_norm": 1.3405648463132127, + "learning_rate": 5.371974708487308e-06, + "loss": 0.242, + "step": 22352 + }, + { + "epoch": 0.4920316971164429, + "grad_norm": 2.206180419489303, + "learning_rate": 5.371619238954713e-06, + "loss": 0.3789, + "step": 22353 + }, + { + "epoch": 0.49205370900286155, + "grad_norm": 2.037489187039296, + "learning_rate": 5.3712637675333835e-06, + "loss": 0.2725, + "step": 22354 + }, + { + "epoch": 0.4920757208892802, + "grad_norm": 0.5995776934264386, + "learning_rate": 5.370908294225125e-06, + "loss": 0.4739, + "step": 22355 + }, + { + "epoch": 0.4920977327756989, + "grad_norm": 1.780720784673834, + "learning_rate": 5.370552819031746e-06, + "loss": 0.3027, + "step": 22356 + }, + { + "epoch": 0.49211974466211755, + "grad_norm": 1.2077939353488092, + "learning_rate": 5.370197341955051e-06, + "loss": 0.3116, + "step": 22357 + }, + { + "epoch": 0.4921417565485362, + "grad_norm": 2.5266307093494507, + "learning_rate": 5.36984186299685e-06, + "loss": 0.309, + "step": 22358 + }, + { + "epoch": 0.49216376843495485, + "grad_norm": 1.4426511009938172, + "learning_rate": 5.369486382158946e-06, + "loss": 0.3672, + "step": 22359 + }, + { + "epoch": 0.49218578032137356, + "grad_norm": 1.4056954824062562, + "learning_rate": 5.369130899443147e-06, + "loss": 0.3045, + "step": 22360 + }, + { + "epoch": 0.4922077922077922, + "grad_norm": 1.792507477289056, + "learning_rate": 5.368775414851263e-06, + "loss": 0.284, + "step": 22361 + }, + { + "epoch": 0.49222980409421085, + "grad_norm": 1.8067438019880078, + "learning_rate": 5.368419928385095e-06, + "loss": 0.3655, + "step": 22362 + }, + { + "epoch": 0.49225181598062956, + "grad_norm": 1.3302395878316737, + "learning_rate": 5.368064440046453e-06, + "loss": 0.289, + "step": 22363 + }, + { + "epoch": 0.4922738278670482, + "grad_norm": 1.9631662063582693, + "learning_rate": 5.367708949837145e-06, + "loss": 0.2758, + "step": 22364 + }, + { + "epoch": 0.49229583975346686, + "grad_norm": 1.1090862617159085, + "learning_rate": 5.367353457758975e-06, + "loss": 0.4873, + "step": 22365 + }, + { + "epoch": 0.49231785163988556, + "grad_norm": 1.6904232758586968, + "learning_rate": 5.366997963813751e-06, + "loss": 0.308, + "step": 22366 + }, + { + "epoch": 0.4923398635263042, + "grad_norm": 1.6866016646734143, + "learning_rate": 5.366642468003279e-06, + "loss": 0.3331, + "step": 22367 + }, + { + "epoch": 0.49236187541272286, + "grad_norm": 0.5947745729370453, + "learning_rate": 5.366286970329368e-06, + "loss": 0.4453, + "step": 22368 + }, + { + "epoch": 0.4923838872991415, + "grad_norm": 1.786983321186212, + "learning_rate": 5.365931470793821e-06, + "loss": 0.26, + "step": 22369 + }, + { + "epoch": 0.4924058991855602, + "grad_norm": 0.6160295832403037, + "learning_rate": 5.36557596939845e-06, + "loss": 0.4594, + "step": 22370 + }, + { + "epoch": 0.49242791107197886, + "grad_norm": 1.5777762135054847, + "learning_rate": 5.365220466145057e-06, + "loss": 0.2164, + "step": 22371 + }, + { + "epoch": 0.4924499229583975, + "grad_norm": 1.2004625985480843, + "learning_rate": 5.364864961035451e-06, + "loss": 0.2914, + "step": 22372 + }, + { + "epoch": 0.4924719348448162, + "grad_norm": 1.405498366652889, + "learning_rate": 5.364509454071439e-06, + "loss": 0.3745, + "step": 22373 + }, + { + "epoch": 0.49249394673123487, + "grad_norm": 1.7389482238801932, + "learning_rate": 5.364153945254828e-06, + "loss": 0.2421, + "step": 22374 + }, + { + "epoch": 0.4925159586176535, + "grad_norm": 0.6411243488584261, + "learning_rate": 5.363798434587422e-06, + "loss": 0.4609, + "step": 22375 + }, + { + "epoch": 0.4925379705040722, + "grad_norm": 1.4131321284870988, + "learning_rate": 5.36344292207103e-06, + "loss": 0.3021, + "step": 22376 + }, + { + "epoch": 0.49255998239049087, + "grad_norm": 1.9972393287795362, + "learning_rate": 5.3630874077074615e-06, + "loss": 0.3604, + "step": 22377 + }, + { + "epoch": 0.4925819942769095, + "grad_norm": 1.3267448334445688, + "learning_rate": 5.362731891498518e-06, + "loss": 0.3884, + "step": 22378 + }, + { + "epoch": 0.4926040061633282, + "grad_norm": 1.1511457139873873, + "learning_rate": 5.362376373446012e-06, + "loss": 0.2657, + "step": 22379 + }, + { + "epoch": 0.49262601804974687, + "grad_norm": 0.6388935986265779, + "learning_rate": 5.362020853551746e-06, + "loss": 0.4761, + "step": 22380 + }, + { + "epoch": 0.4926480299361655, + "grad_norm": 2.445027058691113, + "learning_rate": 5.3616653318175295e-06, + "loss": 0.3094, + "step": 22381 + }, + { + "epoch": 0.49267004182258417, + "grad_norm": 1.763962384337407, + "learning_rate": 5.3613098082451676e-06, + "loss": 0.3515, + "step": 22382 + }, + { + "epoch": 0.4926920537090029, + "grad_norm": 1.3079115164734891, + "learning_rate": 5.360954282836469e-06, + "loss": 0.286, + "step": 22383 + }, + { + "epoch": 0.4927140655954215, + "grad_norm": 1.819007521228426, + "learning_rate": 5.36059875559324e-06, + "loss": 0.3506, + "step": 22384 + }, + { + "epoch": 0.4927360774818402, + "grad_norm": 1.252404398541566, + "learning_rate": 5.360243226517286e-06, + "loss": 0.2763, + "step": 22385 + }, + { + "epoch": 0.4927580893682589, + "grad_norm": 1.2501670042738158, + "learning_rate": 5.359887695610415e-06, + "loss": 0.3726, + "step": 22386 + }, + { + "epoch": 0.4927801012546775, + "grad_norm": 1.3751691068407577, + "learning_rate": 5.359532162874436e-06, + "loss": 0.2779, + "step": 22387 + }, + { + "epoch": 0.4928021131410962, + "grad_norm": 2.015858656543773, + "learning_rate": 5.359176628311152e-06, + "loss": 0.2847, + "step": 22388 + }, + { + "epoch": 0.4928241250275149, + "grad_norm": 2.147661315285253, + "learning_rate": 5.358821091922374e-06, + "loss": 0.2313, + "step": 22389 + }, + { + "epoch": 0.49284613691393353, + "grad_norm": 1.3860853208764918, + "learning_rate": 5.358465553709907e-06, + "loss": 0.3137, + "step": 22390 + }, + { + "epoch": 0.4928681488003522, + "grad_norm": 1.7326659014821069, + "learning_rate": 5.358110013675557e-06, + "loss": 0.3477, + "step": 22391 + }, + { + "epoch": 0.4928901606867709, + "grad_norm": 0.6163431674272042, + "learning_rate": 5.3577544718211335e-06, + "loss": 0.4679, + "step": 22392 + }, + { + "epoch": 0.49291217257318953, + "grad_norm": 1.5455171913772132, + "learning_rate": 5.357398928148441e-06, + "loss": 0.2388, + "step": 22393 + }, + { + "epoch": 0.4929341844596082, + "grad_norm": 1.0998289518697912, + "learning_rate": 5.357043382659288e-06, + "loss": 0.2576, + "step": 22394 + }, + { + "epoch": 0.49295619634602683, + "grad_norm": 1.7177114584072557, + "learning_rate": 5.356687835355483e-06, + "loss": 0.2596, + "step": 22395 + }, + { + "epoch": 0.49297820823244554, + "grad_norm": 0.6146549033072162, + "learning_rate": 5.356332286238829e-06, + "loss": 0.5002, + "step": 22396 + }, + { + "epoch": 0.4930002201188642, + "grad_norm": 2.1508723151562035, + "learning_rate": 5.355976735311137e-06, + "loss": 0.3479, + "step": 22397 + }, + { + "epoch": 0.49302223200528283, + "grad_norm": 0.6235570768159724, + "learning_rate": 5.355621182574211e-06, + "loss": 0.4624, + "step": 22398 + }, + { + "epoch": 0.49304424389170154, + "grad_norm": 0.6052208235498975, + "learning_rate": 5.355265628029861e-06, + "loss": 0.4673, + "step": 22399 + }, + { + "epoch": 0.4930662557781202, + "grad_norm": 2.1144575566119115, + "learning_rate": 5.354910071679892e-06, + "loss": 0.297, + "step": 22400 + }, + { + "epoch": 0.49308826766453884, + "grad_norm": 1.201073948645048, + "learning_rate": 5.35455451352611e-06, + "loss": 0.304, + "step": 22401 + }, + { + "epoch": 0.49311027955095754, + "grad_norm": 1.5454632195939155, + "learning_rate": 5.354198953570328e-06, + "loss": 0.2836, + "step": 22402 + }, + { + "epoch": 0.4931322914373762, + "grad_norm": 1.294532718296039, + "learning_rate": 5.353843391814346e-06, + "loss": 0.3328, + "step": 22403 + }, + { + "epoch": 0.49315430332379484, + "grad_norm": 1.5183612791468426, + "learning_rate": 5.353487828259973e-06, + "loss": 0.2807, + "step": 22404 + }, + { + "epoch": 0.4931763152102135, + "grad_norm": 1.4248638658128157, + "learning_rate": 5.35313226290902e-06, + "loss": 0.3493, + "step": 22405 + }, + { + "epoch": 0.4931983270966322, + "grad_norm": 1.9458611298489954, + "learning_rate": 5.35277669576329e-06, + "loss": 0.3344, + "step": 22406 + }, + { + "epoch": 0.49322033898305084, + "grad_norm": 1.456737125501485, + "learning_rate": 5.352421126824592e-06, + "loss": 0.4252, + "step": 22407 + }, + { + "epoch": 0.4932423508694695, + "grad_norm": 1.4651991094616879, + "learning_rate": 5.352065556094732e-06, + "loss": 0.3529, + "step": 22408 + }, + { + "epoch": 0.4932643627558882, + "grad_norm": 1.63705548790028, + "learning_rate": 5.351709983575517e-06, + "loss": 0.3379, + "step": 22409 + }, + { + "epoch": 0.49328637464230685, + "grad_norm": 0.6421177050919479, + "learning_rate": 5.351354409268756e-06, + "loss": 0.4603, + "step": 22410 + }, + { + "epoch": 0.4933083865287255, + "grad_norm": 1.842161947430242, + "learning_rate": 5.350998833176256e-06, + "loss": 0.385, + "step": 22411 + }, + { + "epoch": 0.4933303984151442, + "grad_norm": 1.0762136527541057, + "learning_rate": 5.3506432552998225e-06, + "loss": 0.3227, + "step": 22412 + }, + { + "epoch": 0.49335241030156285, + "grad_norm": 1.7572342067364974, + "learning_rate": 5.350287675641264e-06, + "loss": 0.3042, + "step": 22413 + }, + { + "epoch": 0.4933744221879815, + "grad_norm": 1.6871447409223752, + "learning_rate": 5.349932094202386e-06, + "loss": 0.2914, + "step": 22414 + }, + { + "epoch": 0.4933964340744002, + "grad_norm": 1.2941585729373615, + "learning_rate": 5.3495765109849995e-06, + "loss": 0.2534, + "step": 22415 + }, + { + "epoch": 0.49341844596081885, + "grad_norm": 1.5722570929473616, + "learning_rate": 5.3492209259909065e-06, + "loss": 0.2467, + "step": 22416 + }, + { + "epoch": 0.4934404578472375, + "grad_norm": 1.7132817899721793, + "learning_rate": 5.3488653392219196e-06, + "loss": 0.2767, + "step": 22417 + }, + { + "epoch": 0.49346246973365615, + "grad_norm": 1.8555899361586776, + "learning_rate": 5.348509750679841e-06, + "loss": 0.2969, + "step": 22418 + }, + { + "epoch": 0.49348448162007486, + "grad_norm": 1.529653295626097, + "learning_rate": 5.348154160366481e-06, + "loss": 0.331, + "step": 22419 + }, + { + "epoch": 0.4935064935064935, + "grad_norm": 1.3299008167681665, + "learning_rate": 5.347798568283648e-06, + "loss": 0.3516, + "step": 22420 + }, + { + "epoch": 0.49352850539291215, + "grad_norm": 1.6858109914783794, + "learning_rate": 5.3474429744331465e-06, + "loss": 0.4731, + "step": 22421 + }, + { + "epoch": 0.49355051727933086, + "grad_norm": 1.6198078692465823, + "learning_rate": 5.347087378816784e-06, + "loss": 0.3121, + "step": 22422 + }, + { + "epoch": 0.4935725291657495, + "grad_norm": 2.2813120877912283, + "learning_rate": 5.346731781436369e-06, + "loss": 0.3824, + "step": 22423 + }, + { + "epoch": 0.49359454105216816, + "grad_norm": 1.4262044461482002, + "learning_rate": 5.346376182293709e-06, + "loss": 0.2602, + "step": 22424 + }, + { + "epoch": 0.49361655293858686, + "grad_norm": 1.434031808734102, + "learning_rate": 5.346020581390611e-06, + "loss": 0.3025, + "step": 22425 + }, + { + "epoch": 0.4936385648250055, + "grad_norm": 1.7954458287552104, + "learning_rate": 5.345664978728881e-06, + "loss": 0.3607, + "step": 22426 + }, + { + "epoch": 0.49366057671142416, + "grad_norm": 1.5065362441235253, + "learning_rate": 5.345309374310328e-06, + "loss": 0.2706, + "step": 22427 + }, + { + "epoch": 0.4936825885978428, + "grad_norm": 1.5016862968245432, + "learning_rate": 5.344953768136759e-06, + "loss": 0.5292, + "step": 22428 + }, + { + "epoch": 0.4937046004842615, + "grad_norm": 1.4248360451388933, + "learning_rate": 5.344598160209981e-06, + "loss": 0.2814, + "step": 22429 + }, + { + "epoch": 0.49372661237068016, + "grad_norm": 1.30958847947349, + "learning_rate": 5.344242550531802e-06, + "loss": 0.3363, + "step": 22430 + }, + { + "epoch": 0.4937486242570988, + "grad_norm": 0.6340012847896637, + "learning_rate": 5.3438869391040274e-06, + "loss": 0.4928, + "step": 22431 + }, + { + "epoch": 0.4937706361435175, + "grad_norm": 0.6378990673676058, + "learning_rate": 5.343531325928466e-06, + "loss": 0.4578, + "step": 22432 + }, + { + "epoch": 0.49379264802993617, + "grad_norm": 1.2121944930921806, + "learning_rate": 5.3431757110069275e-06, + "loss": 0.3506, + "step": 22433 + }, + { + "epoch": 0.4938146599163548, + "grad_norm": 1.0766655332875434, + "learning_rate": 5.342820094341215e-06, + "loss": 0.2678, + "step": 22434 + }, + { + "epoch": 0.4938366718027735, + "grad_norm": 1.5136256741563883, + "learning_rate": 5.342464475933138e-06, + "loss": 0.4223, + "step": 22435 + }, + { + "epoch": 0.49385868368919217, + "grad_norm": 1.6623859738639224, + "learning_rate": 5.342108855784504e-06, + "loss": 0.3214, + "step": 22436 + }, + { + "epoch": 0.4938806955756108, + "grad_norm": 1.4015043093419777, + "learning_rate": 5.341753233897119e-06, + "loss": 0.3471, + "step": 22437 + }, + { + "epoch": 0.4939027074620295, + "grad_norm": 1.5501590189635137, + "learning_rate": 5.3413976102727935e-06, + "loss": 0.334, + "step": 22438 + }, + { + "epoch": 0.49392471934844817, + "grad_norm": 1.4652258870302928, + "learning_rate": 5.341041984913334e-06, + "loss": 0.3458, + "step": 22439 + }, + { + "epoch": 0.4939467312348668, + "grad_norm": 0.6091432897132029, + "learning_rate": 5.340686357820544e-06, + "loss": 0.4608, + "step": 22440 + }, + { + "epoch": 0.49396874312128547, + "grad_norm": 1.4117948653291439, + "learning_rate": 5.340330728996236e-06, + "loss": 0.2691, + "step": 22441 + }, + { + "epoch": 0.4939907550077042, + "grad_norm": 1.7486411756309783, + "learning_rate": 5.339975098442215e-06, + "loss": 0.2746, + "step": 22442 + }, + { + "epoch": 0.4940127668941228, + "grad_norm": 5.357194468502759, + "learning_rate": 5.339619466160289e-06, + "loss": 0.4449, + "step": 22443 + }, + { + "epoch": 0.4940347787805415, + "grad_norm": 1.6633544502487316, + "learning_rate": 5.339263832152265e-06, + "loss": 0.2522, + "step": 22444 + }, + { + "epoch": 0.4940567906669602, + "grad_norm": 1.108159093912474, + "learning_rate": 5.338908196419951e-06, + "loss": 0.2265, + "step": 22445 + }, + { + "epoch": 0.4940788025533788, + "grad_norm": 1.4764864480802495, + "learning_rate": 5.338552558965154e-06, + "loss": 0.3155, + "step": 22446 + }, + { + "epoch": 0.4941008144397975, + "grad_norm": 1.4673830628840725, + "learning_rate": 5.338196919789683e-06, + "loss": 0.1943, + "step": 22447 + }, + { + "epoch": 0.4941228263262162, + "grad_norm": 0.6301053883281102, + "learning_rate": 5.3378412788953434e-06, + "loss": 0.5013, + "step": 22448 + }, + { + "epoch": 0.49414483821263483, + "grad_norm": 1.394999880116209, + "learning_rate": 5.337485636283944e-06, + "loss": 0.2798, + "step": 22449 + }, + { + "epoch": 0.4941668500990535, + "grad_norm": 1.4867453781901248, + "learning_rate": 5.337129991957291e-06, + "loss": 0.2665, + "step": 22450 + }, + { + "epoch": 0.4941888619854721, + "grad_norm": 1.5798938162330782, + "learning_rate": 5.336774345917195e-06, + "loss": 0.2438, + "step": 22451 + }, + { + "epoch": 0.49421087387189083, + "grad_norm": 0.633853096002065, + "learning_rate": 5.336418698165462e-06, + "loss": 0.4865, + "step": 22452 + }, + { + "epoch": 0.4942328857583095, + "grad_norm": 1.2770648862690324, + "learning_rate": 5.336063048703898e-06, + "loss": 0.2378, + "step": 22453 + }, + { + "epoch": 0.49425489764472813, + "grad_norm": 1.3518785017493267, + "learning_rate": 5.335707397534311e-06, + "loss": 0.3098, + "step": 22454 + }, + { + "epoch": 0.49427690953114684, + "grad_norm": 1.252365461072134, + "learning_rate": 5.335351744658509e-06, + "loss": 0.2983, + "step": 22455 + }, + { + "epoch": 0.4942989214175655, + "grad_norm": 1.5103059864443928, + "learning_rate": 5.334996090078303e-06, + "loss": 0.3456, + "step": 22456 + }, + { + "epoch": 0.49432093330398413, + "grad_norm": 1.2508749586638916, + "learning_rate": 5.3346404337954945e-06, + "loss": 0.3323, + "step": 22457 + }, + { + "epoch": 0.49434294519040284, + "grad_norm": 1.2702550732078963, + "learning_rate": 5.334284775811894e-06, + "loss": 0.2025, + "step": 22458 + }, + { + "epoch": 0.4943649570768215, + "grad_norm": 1.2415868393350822, + "learning_rate": 5.333929116129311e-06, + "loss": 0.3026, + "step": 22459 + }, + { + "epoch": 0.49438696896324014, + "grad_norm": 3.088080694978613, + "learning_rate": 5.33357345474955e-06, + "loss": 0.342, + "step": 22460 + }, + { + "epoch": 0.49440898084965884, + "grad_norm": 3.79466638519633, + "learning_rate": 5.33321779167442e-06, + "loss": 0.3332, + "step": 22461 + }, + { + "epoch": 0.4944309927360775, + "grad_norm": 0.6070585142636689, + "learning_rate": 5.332862126905729e-06, + "loss": 0.4609, + "step": 22462 + }, + { + "epoch": 0.49445300462249614, + "grad_norm": 1.277832283895102, + "learning_rate": 5.332506460445284e-06, + "loss": 0.2284, + "step": 22463 + }, + { + "epoch": 0.4944750165089148, + "grad_norm": 1.2140674454781657, + "learning_rate": 5.332150792294893e-06, + "loss": 0.309, + "step": 22464 + }, + { + "epoch": 0.4944970283953335, + "grad_norm": 1.1543332257086418, + "learning_rate": 5.331795122456362e-06, + "loss": 0.3265, + "step": 22465 + }, + { + "epoch": 0.49451904028175214, + "grad_norm": 0.6250140127248464, + "learning_rate": 5.3314394509315025e-06, + "loss": 0.496, + "step": 22466 + }, + { + "epoch": 0.4945410521681708, + "grad_norm": 1.5246203609195508, + "learning_rate": 5.331083777722119e-06, + "loss": 0.4235, + "step": 22467 + }, + { + "epoch": 0.4945630640545895, + "grad_norm": 1.5014454804799395, + "learning_rate": 5.33072810283002e-06, + "loss": 0.2983, + "step": 22468 + }, + { + "epoch": 0.49458507594100815, + "grad_norm": 1.559075271639131, + "learning_rate": 5.330372426257013e-06, + "loss": 0.3083, + "step": 22469 + }, + { + "epoch": 0.4946070878274268, + "grad_norm": 2.6197597661690315, + "learning_rate": 5.330016748004906e-06, + "loss": 0.3163, + "step": 22470 + }, + { + "epoch": 0.4946290997138455, + "grad_norm": 1.2379280303315892, + "learning_rate": 5.329661068075508e-06, + "loss": 0.2913, + "step": 22471 + }, + { + "epoch": 0.49465111160026415, + "grad_norm": 1.8802550910201936, + "learning_rate": 5.329305386470625e-06, + "loss": 0.3776, + "step": 22472 + }, + { + "epoch": 0.4946731234866828, + "grad_norm": 1.5902429700825726, + "learning_rate": 5.3289497031920625e-06, + "loss": 0.3424, + "step": 22473 + }, + { + "epoch": 0.4946951353731015, + "grad_norm": 1.5045549742964428, + "learning_rate": 5.328594018241634e-06, + "loss": 0.2932, + "step": 22474 + }, + { + "epoch": 0.49471714725952015, + "grad_norm": 1.4732796294826844, + "learning_rate": 5.328238331621144e-06, + "loss": 0.3035, + "step": 22475 + }, + { + "epoch": 0.4947391591459388, + "grad_norm": 1.3237013538868259, + "learning_rate": 5.327882643332398e-06, + "loss": 0.2822, + "step": 22476 + }, + { + "epoch": 0.49476117103235745, + "grad_norm": 0.6442028845635472, + "learning_rate": 5.3275269533772076e-06, + "loss": 0.4934, + "step": 22477 + }, + { + "epoch": 0.49478318291877615, + "grad_norm": 1.1927228934082281, + "learning_rate": 5.32717126175738e-06, + "loss": 0.2989, + "step": 22478 + }, + { + "epoch": 0.4948051948051948, + "grad_norm": 1.3805285801383445, + "learning_rate": 5.3268155684747214e-06, + "loss": 0.3247, + "step": 22479 + }, + { + "epoch": 0.49482720669161345, + "grad_norm": 1.5345384542077802, + "learning_rate": 5.32645987353104e-06, + "loss": 0.3292, + "step": 22480 + }, + { + "epoch": 0.49484921857803216, + "grad_norm": 1.4241262414024198, + "learning_rate": 5.3261041769281444e-06, + "loss": 0.2601, + "step": 22481 + }, + { + "epoch": 0.4948712304644508, + "grad_norm": 0.6298643975548173, + "learning_rate": 5.325748478667842e-06, + "loss": 0.473, + "step": 22482 + }, + { + "epoch": 0.49489324235086946, + "grad_norm": 1.8654239271548358, + "learning_rate": 5.32539277875194e-06, + "loss": 0.2386, + "step": 22483 + }, + { + "epoch": 0.49491525423728816, + "grad_norm": 2.1726432758535332, + "learning_rate": 5.325037077182247e-06, + "loss": 0.3155, + "step": 22484 + }, + { + "epoch": 0.4949372661237068, + "grad_norm": 1.2983270227411132, + "learning_rate": 5.324681373960571e-06, + "loss": 0.3084, + "step": 22485 + }, + { + "epoch": 0.49495927801012546, + "grad_norm": 1.4648215451810755, + "learning_rate": 5.324325669088718e-06, + "loss": 0.3235, + "step": 22486 + }, + { + "epoch": 0.4949812898965441, + "grad_norm": 1.4861750891776118, + "learning_rate": 5.323969962568498e-06, + "loss": 0.28, + "step": 22487 + }, + { + "epoch": 0.4950033017829628, + "grad_norm": 1.7045748125211178, + "learning_rate": 5.3236142544017185e-06, + "loss": 0.3485, + "step": 22488 + }, + { + "epoch": 0.49502531366938146, + "grad_norm": 1.6362405232437351, + "learning_rate": 5.323258544590187e-06, + "loss": 0.4248, + "step": 22489 + }, + { + "epoch": 0.4950473255558001, + "grad_norm": 1.3757181471840065, + "learning_rate": 5.322902833135712e-06, + "loss": 0.3611, + "step": 22490 + }, + { + "epoch": 0.4950693374422188, + "grad_norm": 1.07035649812248, + "learning_rate": 5.3225471200400995e-06, + "loss": 0.2512, + "step": 22491 + }, + { + "epoch": 0.49509134932863746, + "grad_norm": 1.1159115928616752, + "learning_rate": 5.32219140530516e-06, + "loss": 0.3075, + "step": 22492 + }, + { + "epoch": 0.4951133612150561, + "grad_norm": 1.3301431699458492, + "learning_rate": 5.3218356889327005e-06, + "loss": 0.324, + "step": 22493 + }, + { + "epoch": 0.4951353731014748, + "grad_norm": 1.1718133180411623, + "learning_rate": 5.321479970924526e-06, + "loss": 0.2704, + "step": 22494 + }, + { + "epoch": 0.49515738498789347, + "grad_norm": 2.7344538442911017, + "learning_rate": 5.321124251282449e-06, + "loss": 0.2896, + "step": 22495 + }, + { + "epoch": 0.4951793968743121, + "grad_norm": 1.3912497863172217, + "learning_rate": 5.320768530008275e-06, + "loss": 0.4958, + "step": 22496 + }, + { + "epoch": 0.4952014087607308, + "grad_norm": 0.6308569605301679, + "learning_rate": 5.320412807103812e-06, + "loss": 0.4766, + "step": 22497 + }, + { + "epoch": 0.49522342064714947, + "grad_norm": 2.1096377036218494, + "learning_rate": 5.320057082570869e-06, + "loss": 0.3355, + "step": 22498 + }, + { + "epoch": 0.4952454325335681, + "grad_norm": 1.216834286887466, + "learning_rate": 5.319701356411251e-06, + "loss": 0.3179, + "step": 22499 + }, + { + "epoch": 0.49526744441998677, + "grad_norm": 1.276784496853571, + "learning_rate": 5.319345628626769e-06, + "loss": 0.2836, + "step": 22500 + }, + { + "epoch": 0.4952894563064055, + "grad_norm": 1.265312288336017, + "learning_rate": 5.318989899219231e-06, + "loss": 0.3225, + "step": 22501 + }, + { + "epoch": 0.4953114681928241, + "grad_norm": 1.5206543074807573, + "learning_rate": 5.318634168190443e-06, + "loss": 0.3276, + "step": 22502 + }, + { + "epoch": 0.49533348007924277, + "grad_norm": 2.3151697200475607, + "learning_rate": 5.318278435542216e-06, + "loss": 0.3563, + "step": 22503 + }, + { + "epoch": 0.4953554919656615, + "grad_norm": 2.0436370274002917, + "learning_rate": 5.3179227012763526e-06, + "loss": 0.3542, + "step": 22504 + }, + { + "epoch": 0.4953775038520801, + "grad_norm": 1.2624264890303083, + "learning_rate": 5.317566965394667e-06, + "loss": 0.3188, + "step": 22505 + }, + { + "epoch": 0.4953995157384988, + "grad_norm": 4.382007673675832, + "learning_rate": 5.317211227898964e-06, + "loss": 0.4034, + "step": 22506 + }, + { + "epoch": 0.4954215276249175, + "grad_norm": 0.6629614967170563, + "learning_rate": 5.316855488791052e-06, + "loss": 0.4722, + "step": 22507 + }, + { + "epoch": 0.49544353951133613, + "grad_norm": 1.6283318256744512, + "learning_rate": 5.316499748072738e-06, + "loss": 0.3119, + "step": 22508 + }, + { + "epoch": 0.4954655513977548, + "grad_norm": 1.5874919885800076, + "learning_rate": 5.316144005745831e-06, + "loss": 0.323, + "step": 22509 + }, + { + "epoch": 0.4954875632841734, + "grad_norm": 1.2502687302009023, + "learning_rate": 5.315788261812142e-06, + "loss": 0.308, + "step": 22510 + }, + { + "epoch": 0.49550957517059213, + "grad_norm": 0.610199919122195, + "learning_rate": 5.3154325162734745e-06, + "loss": 0.4669, + "step": 22511 + }, + { + "epoch": 0.4955315870570108, + "grad_norm": 0.6195528374087704, + "learning_rate": 5.3150767691316374e-06, + "loss": 0.4741, + "step": 22512 + }, + { + "epoch": 0.49555359894342943, + "grad_norm": 0.6106430910153234, + "learning_rate": 5.3147210203884406e-06, + "loss": 0.4931, + "step": 22513 + }, + { + "epoch": 0.49557561082984813, + "grad_norm": 1.4832683183973545, + "learning_rate": 5.3143652700456906e-06, + "loss": 0.3267, + "step": 22514 + }, + { + "epoch": 0.4955976227162668, + "grad_norm": 2.0435026163273045, + "learning_rate": 5.314009518105197e-06, + "loss": 0.3615, + "step": 22515 + }, + { + "epoch": 0.49561963460268543, + "grad_norm": 3.524918847599501, + "learning_rate": 5.3136537645687666e-06, + "loss": 0.2899, + "step": 22516 + }, + { + "epoch": 0.49564164648910414, + "grad_norm": 1.2574575606136706, + "learning_rate": 5.313298009438207e-06, + "loss": 0.2562, + "step": 22517 + }, + { + "epoch": 0.4956636583755228, + "grad_norm": 0.6412931523331266, + "learning_rate": 5.312942252715328e-06, + "loss": 0.5048, + "step": 22518 + }, + { + "epoch": 0.49568567026194144, + "grad_norm": 1.3423565034782816, + "learning_rate": 5.312586494401937e-06, + "loss": 0.3278, + "step": 22519 + }, + { + "epoch": 0.49570768214836014, + "grad_norm": 0.647398620931615, + "learning_rate": 5.312230734499841e-06, + "loss": 0.511, + "step": 22520 + }, + { + "epoch": 0.4957296940347788, + "grad_norm": 1.4400361461860023, + "learning_rate": 5.311874973010851e-06, + "loss": 0.225, + "step": 22521 + }, + { + "epoch": 0.49575170592119744, + "grad_norm": 2.5696337577940263, + "learning_rate": 5.311519209936771e-06, + "loss": 0.2804, + "step": 22522 + }, + { + "epoch": 0.4957737178076161, + "grad_norm": 1.460206762705158, + "learning_rate": 5.311163445279413e-06, + "loss": 0.3396, + "step": 22523 + }, + { + "epoch": 0.4957957296940348, + "grad_norm": 1.4755778277309268, + "learning_rate": 5.3108076790405835e-06, + "loss": 0.2776, + "step": 22524 + }, + { + "epoch": 0.49581774158045344, + "grad_norm": 1.064013290561198, + "learning_rate": 5.310451911222091e-06, + "loss": 0.2577, + "step": 22525 + }, + { + "epoch": 0.4958397534668721, + "grad_norm": 1.3191799729243032, + "learning_rate": 5.3100961418257424e-06, + "loss": 0.3606, + "step": 22526 + }, + { + "epoch": 0.4958617653532908, + "grad_norm": 1.6710761395658424, + "learning_rate": 5.309740370853347e-06, + "loss": 0.3387, + "step": 22527 + }, + { + "epoch": 0.49588377723970944, + "grad_norm": 0.619196427276924, + "learning_rate": 5.309384598306713e-06, + "loss": 0.4909, + "step": 22528 + }, + { + "epoch": 0.4959057891261281, + "grad_norm": 0.6334243760501757, + "learning_rate": 5.30902882418765e-06, + "loss": 0.4795, + "step": 22529 + }, + { + "epoch": 0.4959278010125468, + "grad_norm": 0.6304209326896478, + "learning_rate": 5.308673048497963e-06, + "loss": 0.477, + "step": 22530 + }, + { + "epoch": 0.49594981289896545, + "grad_norm": 0.609129488256815, + "learning_rate": 5.308317271239462e-06, + "loss": 0.4436, + "step": 22531 + }, + { + "epoch": 0.4959718247853841, + "grad_norm": 1.6731501478364057, + "learning_rate": 5.3079614924139555e-06, + "loss": 0.4118, + "step": 22532 + }, + { + "epoch": 0.49599383667180275, + "grad_norm": 1.3439308684481837, + "learning_rate": 5.307605712023251e-06, + "loss": 0.3242, + "step": 22533 + }, + { + "epoch": 0.49601584855822145, + "grad_norm": 0.6499507573479385, + "learning_rate": 5.307249930069158e-06, + "loss": 0.4795, + "step": 22534 + }, + { + "epoch": 0.4960378604446401, + "grad_norm": 0.6332631191078528, + "learning_rate": 5.306894146553483e-06, + "loss": 0.4721, + "step": 22535 + }, + { + "epoch": 0.49605987233105875, + "grad_norm": 1.6728097774811361, + "learning_rate": 5.306538361478035e-06, + "loss": 0.286, + "step": 22536 + }, + { + "epoch": 0.49608188421747745, + "grad_norm": 1.582058220122049, + "learning_rate": 5.306182574844623e-06, + "loss": 0.2972, + "step": 22537 + }, + { + "epoch": 0.4961038961038961, + "grad_norm": 1.4694927618458677, + "learning_rate": 5.305826786655054e-06, + "loss": 0.3362, + "step": 22538 + }, + { + "epoch": 0.49612590799031475, + "grad_norm": 1.5304827930044713, + "learning_rate": 5.305470996911137e-06, + "loss": 0.308, + "step": 22539 + }, + { + "epoch": 0.49614791987673346, + "grad_norm": 1.5163553275493509, + "learning_rate": 5.305115205614679e-06, + "loss": 0.3689, + "step": 22540 + }, + { + "epoch": 0.4961699317631521, + "grad_norm": 0.6527865126733821, + "learning_rate": 5.304759412767492e-06, + "loss": 0.4708, + "step": 22541 + }, + { + "epoch": 0.49619194364957075, + "grad_norm": 2.012218637912472, + "learning_rate": 5.30440361837138e-06, + "loss": 0.2933, + "step": 22542 + }, + { + "epoch": 0.49621395553598946, + "grad_norm": 3.1733815761743998, + "learning_rate": 5.3040478224281536e-06, + "loss": 0.2093, + "step": 22543 + }, + { + "epoch": 0.4962359674224081, + "grad_norm": 1.3789309615989045, + "learning_rate": 5.303692024939621e-06, + "loss": 0.2791, + "step": 22544 + }, + { + "epoch": 0.49625797930882676, + "grad_norm": 0.6028714398512071, + "learning_rate": 5.303336225907588e-06, + "loss": 0.4957, + "step": 22545 + }, + { + "epoch": 0.4962799911952454, + "grad_norm": 1.52061630567336, + "learning_rate": 5.302980425333867e-06, + "loss": 0.3506, + "step": 22546 + }, + { + "epoch": 0.4963020030816641, + "grad_norm": 1.6934349630981511, + "learning_rate": 5.302624623220265e-06, + "loss": 0.1914, + "step": 22547 + }, + { + "epoch": 0.49632401496808276, + "grad_norm": 0.6256220065147053, + "learning_rate": 5.3022688195685875e-06, + "loss": 0.4656, + "step": 22548 + }, + { + "epoch": 0.4963460268545014, + "grad_norm": 0.6200235607979013, + "learning_rate": 5.3019130143806464e-06, + "loss": 0.4925, + "step": 22549 + }, + { + "epoch": 0.4963680387409201, + "grad_norm": 0.5874128164159275, + "learning_rate": 5.301557207658249e-06, + "loss": 0.4887, + "step": 22550 + }, + { + "epoch": 0.49639005062733876, + "grad_norm": 0.6061552155130101, + "learning_rate": 5.301201399403202e-06, + "loss": 0.4848, + "step": 22551 + }, + { + "epoch": 0.4964120625137574, + "grad_norm": 1.3915669496098002, + "learning_rate": 5.300845589617316e-06, + "loss": 0.3956, + "step": 22552 + }, + { + "epoch": 0.4964340744001761, + "grad_norm": 1.22007869633474, + "learning_rate": 5.300489778302398e-06, + "loss": 0.3117, + "step": 22553 + }, + { + "epoch": 0.49645608628659477, + "grad_norm": 1.961869352498593, + "learning_rate": 5.300133965460258e-06, + "loss": 0.2916, + "step": 22554 + }, + { + "epoch": 0.4964780981730134, + "grad_norm": 3.8768610648469406, + "learning_rate": 5.299778151092702e-06, + "loss": 0.2661, + "step": 22555 + }, + { + "epoch": 0.4965001100594321, + "grad_norm": 1.214138908567677, + "learning_rate": 5.29942233520154e-06, + "loss": 0.247, + "step": 22556 + }, + { + "epoch": 0.49652212194585077, + "grad_norm": 1.574477921781867, + "learning_rate": 5.299066517788581e-06, + "loss": 0.3112, + "step": 22557 + }, + { + "epoch": 0.4965441338322694, + "grad_norm": 1.4470768742290026, + "learning_rate": 5.2987106988556314e-06, + "loss": 0.3608, + "step": 22558 + }, + { + "epoch": 0.49656614571868807, + "grad_norm": 1.9542705889996972, + "learning_rate": 5.298354878404502e-06, + "loss": 0.3817, + "step": 22559 + }, + { + "epoch": 0.4965881576051068, + "grad_norm": 1.5531223164284713, + "learning_rate": 5.2979990564369994e-06, + "loss": 0.2936, + "step": 22560 + }, + { + "epoch": 0.4966101694915254, + "grad_norm": 1.3502197087709185, + "learning_rate": 5.297643232954933e-06, + "loss": 0.2983, + "step": 22561 + }, + { + "epoch": 0.49663218137794407, + "grad_norm": 1.360164600527659, + "learning_rate": 5.297287407960111e-06, + "loss": 0.3383, + "step": 22562 + }, + { + "epoch": 0.4966541932643628, + "grad_norm": 1.278517586313448, + "learning_rate": 5.296931581454341e-06, + "loss": 0.3109, + "step": 22563 + }, + { + "epoch": 0.4966762051507814, + "grad_norm": 1.1057031703654274, + "learning_rate": 5.296575753439434e-06, + "loss": 0.2872, + "step": 22564 + }, + { + "epoch": 0.4966982170372001, + "grad_norm": 0.6444855817726342, + "learning_rate": 5.296219923917197e-06, + "loss": 0.4788, + "step": 22565 + }, + { + "epoch": 0.4967202289236188, + "grad_norm": 1.4570500396356523, + "learning_rate": 5.295864092889435e-06, + "loss": 0.4748, + "step": 22566 + }, + { + "epoch": 0.4967422408100374, + "grad_norm": 1.3932372958311312, + "learning_rate": 5.295508260357962e-06, + "loss": 0.3157, + "step": 22567 + }, + { + "epoch": 0.4967642526964561, + "grad_norm": 1.5287848684624967, + "learning_rate": 5.295152426324584e-06, + "loss": 0.3392, + "step": 22568 + }, + { + "epoch": 0.4967862645828747, + "grad_norm": 1.9033509596658125, + "learning_rate": 5.294796590791108e-06, + "loss": 0.3746, + "step": 22569 + }, + { + "epoch": 0.49680827646929343, + "grad_norm": 1.5278222132509849, + "learning_rate": 5.294440753759347e-06, + "loss": 0.3222, + "step": 22570 + }, + { + "epoch": 0.4968302883557121, + "grad_norm": 1.5294091634129565, + "learning_rate": 5.294084915231105e-06, + "loss": 0.3011, + "step": 22571 + }, + { + "epoch": 0.49685230024213073, + "grad_norm": 1.4637904906011392, + "learning_rate": 5.293729075208193e-06, + "loss": 0.4112, + "step": 22572 + }, + { + "epoch": 0.49687431212854943, + "grad_norm": 1.3173192718147138, + "learning_rate": 5.293373233692418e-06, + "loss": 0.3586, + "step": 22573 + }, + { + "epoch": 0.4968963240149681, + "grad_norm": 1.7210848654563076, + "learning_rate": 5.293017390685591e-06, + "loss": 0.3521, + "step": 22574 + }, + { + "epoch": 0.49691833590138673, + "grad_norm": 1.6519039924180245, + "learning_rate": 5.292661546189518e-06, + "loss": 0.3276, + "step": 22575 + }, + { + "epoch": 0.49694034778780544, + "grad_norm": 1.148560323684656, + "learning_rate": 5.292305700206008e-06, + "loss": 0.2843, + "step": 22576 + }, + { + "epoch": 0.4969623596742241, + "grad_norm": 1.1390860070424287, + "learning_rate": 5.29194985273687e-06, + "loss": 0.3204, + "step": 22577 + }, + { + "epoch": 0.49698437156064273, + "grad_norm": 1.4255385975311694, + "learning_rate": 5.291594003783913e-06, + "loss": 0.3093, + "step": 22578 + }, + { + "epoch": 0.49700638344706144, + "grad_norm": 1.9918626787357567, + "learning_rate": 5.2912381533489455e-06, + "loss": 0.2656, + "step": 22579 + }, + { + "epoch": 0.4970283953334801, + "grad_norm": 3.9388069525188425, + "learning_rate": 5.290882301433776e-06, + "loss": 0.3854, + "step": 22580 + }, + { + "epoch": 0.49705040721989874, + "grad_norm": 1.516143556849338, + "learning_rate": 5.290526448040212e-06, + "loss": 0.3219, + "step": 22581 + }, + { + "epoch": 0.4970724191063174, + "grad_norm": 1.386423874683922, + "learning_rate": 5.290170593170063e-06, + "loss": 0.3469, + "step": 22582 + }, + { + "epoch": 0.4970944309927361, + "grad_norm": 1.5388405104698413, + "learning_rate": 5.2898147368251395e-06, + "loss": 0.275, + "step": 22583 + }, + { + "epoch": 0.49711644287915474, + "grad_norm": 1.5163603421040033, + "learning_rate": 5.289458879007247e-06, + "loss": 0.4125, + "step": 22584 + }, + { + "epoch": 0.4971384547655734, + "grad_norm": 1.370614473344888, + "learning_rate": 5.289103019718195e-06, + "loss": 0.2439, + "step": 22585 + }, + { + "epoch": 0.4971604666519921, + "grad_norm": 1.6215065597944271, + "learning_rate": 5.2887471589597926e-06, + "loss": 0.2933, + "step": 22586 + }, + { + "epoch": 0.49718247853841074, + "grad_norm": 2.0690437974359366, + "learning_rate": 5.288391296733848e-06, + "loss": 0.314, + "step": 22587 + }, + { + "epoch": 0.4972044904248294, + "grad_norm": 1.3754935908757784, + "learning_rate": 5.288035433042172e-06, + "loss": 0.3014, + "step": 22588 + }, + { + "epoch": 0.4972265023112481, + "grad_norm": 2.0665905804928992, + "learning_rate": 5.2876795678865685e-06, + "loss": 0.3131, + "step": 22589 + }, + { + "epoch": 0.49724851419766675, + "grad_norm": 0.61415317932046, + "learning_rate": 5.287323701268851e-06, + "loss": 0.4587, + "step": 22590 + }, + { + "epoch": 0.4972705260840854, + "grad_norm": 1.2576801441563645, + "learning_rate": 5.286967833190827e-06, + "loss": 0.369, + "step": 22591 + }, + { + "epoch": 0.49729253797050404, + "grad_norm": 1.456619464379993, + "learning_rate": 5.286611963654303e-06, + "loss": 0.2707, + "step": 22592 + }, + { + "epoch": 0.49731454985692275, + "grad_norm": 1.559225126146758, + "learning_rate": 5.286256092661091e-06, + "loss": 0.2645, + "step": 22593 + }, + { + "epoch": 0.4973365617433414, + "grad_norm": 1.438317534026851, + "learning_rate": 5.285900220212996e-06, + "loss": 0.3519, + "step": 22594 + }, + { + "epoch": 0.49735857362976005, + "grad_norm": 2.734307197684149, + "learning_rate": 5.2855443463118294e-06, + "loss": 0.3365, + "step": 22595 + }, + { + "epoch": 0.49738058551617875, + "grad_norm": 1.706405267084092, + "learning_rate": 5.285188470959399e-06, + "loss": 0.3974, + "step": 22596 + }, + { + "epoch": 0.4974025974025974, + "grad_norm": 1.2888858345103116, + "learning_rate": 5.284832594157513e-06, + "loss": 0.2642, + "step": 22597 + }, + { + "epoch": 0.49742460928901605, + "grad_norm": 1.537918522289905, + "learning_rate": 5.2844767159079815e-06, + "loss": 0.3081, + "step": 22598 + }, + { + "epoch": 0.49744662117543476, + "grad_norm": 1.256051281488394, + "learning_rate": 5.284120836212612e-06, + "loss": 0.2963, + "step": 22599 + }, + { + "epoch": 0.4974686330618534, + "grad_norm": 1.477214810756236, + "learning_rate": 5.283764955073214e-06, + "loss": 0.2992, + "step": 22600 + }, + { + "epoch": 0.49749064494827205, + "grad_norm": 0.6341821942924409, + "learning_rate": 5.283409072491595e-06, + "loss": 0.5203, + "step": 22601 + }, + { + "epoch": 0.49751265683469076, + "grad_norm": 1.216505750854587, + "learning_rate": 5.283053188469566e-06, + "loss": 0.2956, + "step": 22602 + }, + { + "epoch": 0.4975346687211094, + "grad_norm": 1.8284431864413029, + "learning_rate": 5.282697303008935e-06, + "loss": 0.3311, + "step": 22603 + }, + { + "epoch": 0.49755668060752806, + "grad_norm": 1.2874980520781596, + "learning_rate": 5.282341416111508e-06, + "loss": 0.2694, + "step": 22604 + }, + { + "epoch": 0.4975786924939467, + "grad_norm": 1.1692846101918568, + "learning_rate": 5.281985527779097e-06, + "loss": 0.1802, + "step": 22605 + }, + { + "epoch": 0.4976007043803654, + "grad_norm": 0.615959798248093, + "learning_rate": 5.2816296380135116e-06, + "loss": 0.4631, + "step": 22606 + }, + { + "epoch": 0.49762271626678406, + "grad_norm": 1.2066051473293886, + "learning_rate": 5.281273746816555e-06, + "loss": 0.2809, + "step": 22607 + }, + { + "epoch": 0.4976447281532027, + "grad_norm": 1.3872756146548193, + "learning_rate": 5.280917854190043e-06, + "loss": 0.3537, + "step": 22608 + }, + { + "epoch": 0.4976667400396214, + "grad_norm": 1.4014630837074433, + "learning_rate": 5.28056196013578e-06, + "loss": 0.3056, + "step": 22609 + }, + { + "epoch": 0.49768875192604006, + "grad_norm": 1.121154357293123, + "learning_rate": 5.280206064655575e-06, + "loss": 0.255, + "step": 22610 + }, + { + "epoch": 0.4977107638124587, + "grad_norm": 0.6119466143531761, + "learning_rate": 5.2798501677512405e-06, + "loss": 0.5003, + "step": 22611 + }, + { + "epoch": 0.4977327756988774, + "grad_norm": 0.610694751795891, + "learning_rate": 5.27949426942458e-06, + "loss": 0.4806, + "step": 22612 + }, + { + "epoch": 0.49775478758529607, + "grad_norm": 0.6176625972250214, + "learning_rate": 5.279138369677406e-06, + "loss": 0.4772, + "step": 22613 + }, + { + "epoch": 0.4977767994717147, + "grad_norm": 1.481704060212767, + "learning_rate": 5.278782468511526e-06, + "loss": 0.2563, + "step": 22614 + }, + { + "epoch": 0.4977988113581334, + "grad_norm": 3.663831329764161, + "learning_rate": 5.27842656592875e-06, + "loss": 0.3105, + "step": 22615 + }, + { + "epoch": 0.49782082324455207, + "grad_norm": 1.5100637074569627, + "learning_rate": 5.2780706619308855e-06, + "loss": 0.3577, + "step": 22616 + }, + { + "epoch": 0.4978428351309707, + "grad_norm": 2.2172770440624596, + "learning_rate": 5.27771475651974e-06, + "loss": 0.2277, + "step": 22617 + }, + { + "epoch": 0.49786484701738937, + "grad_norm": 0.655460732971807, + "learning_rate": 5.277358849697127e-06, + "loss": 0.4712, + "step": 22618 + }, + { + "epoch": 0.49788685890380807, + "grad_norm": 2.041043999413996, + "learning_rate": 5.277002941464852e-06, + "loss": 0.2932, + "step": 22619 + }, + { + "epoch": 0.4979088707902267, + "grad_norm": 0.6432548694602361, + "learning_rate": 5.276647031824724e-06, + "loss": 0.4738, + "step": 22620 + }, + { + "epoch": 0.49793088267664537, + "grad_norm": 1.4952721137834986, + "learning_rate": 5.276291120778553e-06, + "loss": 0.3362, + "step": 22621 + }, + { + "epoch": 0.4979528945630641, + "grad_norm": 1.2167043752520355, + "learning_rate": 5.275935208328147e-06, + "loss": 0.2811, + "step": 22622 + }, + { + "epoch": 0.4979749064494827, + "grad_norm": 1.6015518007949971, + "learning_rate": 5.275579294475314e-06, + "loss": 0.2751, + "step": 22623 + }, + { + "epoch": 0.4979969183359014, + "grad_norm": 1.1569061147822135, + "learning_rate": 5.275223379221867e-06, + "loss": 0.3067, + "step": 22624 + }, + { + "epoch": 0.4980189302223201, + "grad_norm": 1.413542772666291, + "learning_rate": 5.2748674625696085e-06, + "loss": 0.2733, + "step": 22625 + }, + { + "epoch": 0.4980409421087387, + "grad_norm": 1.4641897870850569, + "learning_rate": 5.274511544520353e-06, + "loss": 0.47, + "step": 22626 + }, + { + "epoch": 0.4980629539951574, + "grad_norm": 1.5511665429171355, + "learning_rate": 5.274155625075907e-06, + "loss": 0.3106, + "step": 22627 + }, + { + "epoch": 0.498084965881576, + "grad_norm": 1.4187143229378658, + "learning_rate": 5.27379970423808e-06, + "loss": 0.2852, + "step": 22628 + }, + { + "epoch": 0.49810697776799473, + "grad_norm": 0.6261780611816351, + "learning_rate": 5.273443782008681e-06, + "loss": 0.4535, + "step": 22629 + }, + { + "epoch": 0.4981289896544134, + "grad_norm": 3.424438772665812, + "learning_rate": 5.273087858389518e-06, + "loss": 0.4016, + "step": 22630 + }, + { + "epoch": 0.49815100154083203, + "grad_norm": 2.1499375886063783, + "learning_rate": 5.2727319333824015e-06, + "loss": 0.3856, + "step": 22631 + }, + { + "epoch": 0.49817301342725073, + "grad_norm": 1.549286286982, + "learning_rate": 5.272376006989138e-06, + "loss": 0.3861, + "step": 22632 + }, + { + "epoch": 0.4981950253136694, + "grad_norm": 1.3249835048163754, + "learning_rate": 5.272020079211541e-06, + "loss": 0.299, + "step": 22633 + }, + { + "epoch": 0.49821703720008803, + "grad_norm": 1.2572715716132346, + "learning_rate": 5.2716641500514145e-06, + "loss": 0.2956, + "step": 22634 + }, + { + "epoch": 0.49823904908650674, + "grad_norm": 1.846742230043107, + "learning_rate": 5.2713082195105695e-06, + "loss": 0.3123, + "step": 22635 + }, + { + "epoch": 0.4982610609729254, + "grad_norm": 2.3107684244654325, + "learning_rate": 5.270952287590816e-06, + "loss": 0.3331, + "step": 22636 + }, + { + "epoch": 0.49828307285934403, + "grad_norm": 1.6073517427825563, + "learning_rate": 5.270596354293962e-06, + "loss": 0.3269, + "step": 22637 + }, + { + "epoch": 0.49830508474576274, + "grad_norm": 1.1335986570528225, + "learning_rate": 5.270240419621817e-06, + "loss": 0.2469, + "step": 22638 + }, + { + "epoch": 0.4983270966321814, + "grad_norm": 1.6167028261584597, + "learning_rate": 5.269884483576189e-06, + "loss": 0.3557, + "step": 22639 + }, + { + "epoch": 0.49834910851860004, + "grad_norm": 1.671334539255958, + "learning_rate": 5.269528546158888e-06, + "loss": 0.2755, + "step": 22640 + }, + { + "epoch": 0.4983711204050187, + "grad_norm": 1.3208426534926059, + "learning_rate": 5.269172607371722e-06, + "loss": 0.3455, + "step": 22641 + }, + { + "epoch": 0.4983931322914374, + "grad_norm": 1.7052935744851438, + "learning_rate": 5.268816667216502e-06, + "loss": 0.3361, + "step": 22642 + }, + { + "epoch": 0.49841514417785604, + "grad_norm": 1.4150588762899663, + "learning_rate": 5.268460725695035e-06, + "loss": 0.332, + "step": 22643 + }, + { + "epoch": 0.4984371560642747, + "grad_norm": 0.6475548996756332, + "learning_rate": 5.268104782809132e-06, + "loss": 0.4959, + "step": 22644 + }, + { + "epoch": 0.4984591679506934, + "grad_norm": 1.1794136687005357, + "learning_rate": 5.267748838560599e-06, + "loss": 0.2464, + "step": 22645 + }, + { + "epoch": 0.49848117983711204, + "grad_norm": 0.6252026982849349, + "learning_rate": 5.2673928929512474e-06, + "loss": 0.4734, + "step": 22646 + }, + { + "epoch": 0.4985031917235307, + "grad_norm": 1.4348506575941973, + "learning_rate": 5.267036945982887e-06, + "loss": 0.3559, + "step": 22647 + }, + { + "epoch": 0.4985252036099494, + "grad_norm": 1.165323497267912, + "learning_rate": 5.266680997657324e-06, + "loss": 0.2389, + "step": 22648 + }, + { + "epoch": 0.49854721549636805, + "grad_norm": 1.6753292364921677, + "learning_rate": 5.2663250479763704e-06, + "loss": 0.2477, + "step": 22649 + }, + { + "epoch": 0.4985692273827867, + "grad_norm": 1.356558461207417, + "learning_rate": 5.265969096941835e-06, + "loss": 0.2787, + "step": 22650 + }, + { + "epoch": 0.49859123926920534, + "grad_norm": 1.7082451239710315, + "learning_rate": 5.2656131445555235e-06, + "loss": 0.2608, + "step": 22651 + }, + { + "epoch": 0.49861325115562405, + "grad_norm": 0.6291683265554527, + "learning_rate": 5.26525719081925e-06, + "loss": 0.481, + "step": 22652 + }, + { + "epoch": 0.4986352630420427, + "grad_norm": 2.3175978459062376, + "learning_rate": 5.26490123573482e-06, + "loss": 0.3125, + "step": 22653 + }, + { + "epoch": 0.49865727492846135, + "grad_norm": 1.8072823624728735, + "learning_rate": 5.264545279304044e-06, + "loss": 0.3092, + "step": 22654 + }, + { + "epoch": 0.49867928681488005, + "grad_norm": 1.8499901098866158, + "learning_rate": 5.264189321528732e-06, + "loss": 0.3306, + "step": 22655 + }, + { + "epoch": 0.4987012987012987, + "grad_norm": 1.242107467014773, + "learning_rate": 5.263833362410691e-06, + "loss": 0.2641, + "step": 22656 + }, + { + "epoch": 0.49872331058771735, + "grad_norm": 0.6236830705439463, + "learning_rate": 5.263477401951731e-06, + "loss": 0.4538, + "step": 22657 + }, + { + "epoch": 0.49874532247413605, + "grad_norm": 1.3059734278417519, + "learning_rate": 5.263121440153662e-06, + "loss": 0.29, + "step": 22658 + }, + { + "epoch": 0.4987673343605547, + "grad_norm": 1.8139881642238194, + "learning_rate": 5.2627654770182915e-06, + "loss": 0.2952, + "step": 22659 + }, + { + "epoch": 0.49878934624697335, + "grad_norm": 2.0617453195871396, + "learning_rate": 5.262409512547429e-06, + "loss": 0.3906, + "step": 22660 + }, + { + "epoch": 0.49881135813339206, + "grad_norm": 2.572148990454646, + "learning_rate": 5.262053546742887e-06, + "loss": 0.3408, + "step": 22661 + }, + { + "epoch": 0.4988333700198107, + "grad_norm": 1.265948107986547, + "learning_rate": 5.261697579606469e-06, + "loss": 0.3276, + "step": 22662 + }, + { + "epoch": 0.49885538190622936, + "grad_norm": 1.1845534207680852, + "learning_rate": 5.261341611139989e-06, + "loss": 0.2668, + "step": 22663 + }, + { + "epoch": 0.498877393792648, + "grad_norm": 1.2754407979376718, + "learning_rate": 5.260985641345254e-06, + "loss": 0.2555, + "step": 22664 + }, + { + "epoch": 0.4988994056790667, + "grad_norm": 0.6083893848792585, + "learning_rate": 5.2606296702240744e-06, + "loss": 0.4775, + "step": 22665 + }, + { + "epoch": 0.49892141756548536, + "grad_norm": 1.6370875784144732, + "learning_rate": 5.260273697778257e-06, + "loss": 0.3501, + "step": 22666 + }, + { + "epoch": 0.498943429451904, + "grad_norm": 1.7312169744621164, + "learning_rate": 5.2599177240096136e-06, + "loss": 0.3074, + "step": 22667 + }, + { + "epoch": 0.4989654413383227, + "grad_norm": 1.3153570065178448, + "learning_rate": 5.2595617489199525e-06, + "loss": 0.2956, + "step": 22668 + }, + { + "epoch": 0.49898745322474136, + "grad_norm": 1.254992541994862, + "learning_rate": 5.259205772511082e-06, + "loss": 0.2359, + "step": 22669 + }, + { + "epoch": 0.49900946511116, + "grad_norm": 1.9198166114848318, + "learning_rate": 5.258849794784813e-06, + "loss": 0.3792, + "step": 22670 + }, + { + "epoch": 0.4990314769975787, + "grad_norm": 2.006932705570492, + "learning_rate": 5.2584938157429544e-06, + "loss": 0.3391, + "step": 22671 + }, + { + "epoch": 0.49905348888399736, + "grad_norm": 1.4289413260863868, + "learning_rate": 5.258137835387312e-06, + "loss": 0.3079, + "step": 22672 + }, + { + "epoch": 0.499075500770416, + "grad_norm": 0.6151415097875259, + "learning_rate": 5.257781853719701e-06, + "loss": 0.4627, + "step": 22673 + }, + { + "epoch": 0.49909751265683466, + "grad_norm": 1.6266105362234509, + "learning_rate": 5.257425870741927e-06, + "loss": 0.3316, + "step": 22674 + }, + { + "epoch": 0.49911952454325337, + "grad_norm": 1.6176291849768842, + "learning_rate": 5.257069886455799e-06, + "loss": 0.3016, + "step": 22675 + }, + { + "epoch": 0.499141536429672, + "grad_norm": 2.1602678964657867, + "learning_rate": 5.256713900863128e-06, + "loss": 0.3745, + "step": 22676 + }, + { + "epoch": 0.49916354831609067, + "grad_norm": 1.4865137004238667, + "learning_rate": 5.256357913965722e-06, + "loss": 0.3764, + "step": 22677 + }, + { + "epoch": 0.49918556020250937, + "grad_norm": 2.122076458185763, + "learning_rate": 5.25600192576539e-06, + "loss": 0.3094, + "step": 22678 + }, + { + "epoch": 0.499207572088928, + "grad_norm": 1.2704446942257759, + "learning_rate": 5.255645936263943e-06, + "loss": 0.294, + "step": 22679 + }, + { + "epoch": 0.49922958397534667, + "grad_norm": 1.5753831647250665, + "learning_rate": 5.25528994546319e-06, + "loss": 0.309, + "step": 22680 + }, + { + "epoch": 0.4992515958617654, + "grad_norm": 1.525650082976183, + "learning_rate": 5.254933953364939e-06, + "loss": 0.4196, + "step": 22681 + }, + { + "epoch": 0.499273607748184, + "grad_norm": 1.4758337699650115, + "learning_rate": 5.254577959970999e-06, + "loss": 0.3119, + "step": 22682 + }, + { + "epoch": 0.49929561963460267, + "grad_norm": 0.6225395268058629, + "learning_rate": 5.254221965283183e-06, + "loss": 0.4679, + "step": 22683 + }, + { + "epoch": 0.4993176315210214, + "grad_norm": 1.35017134214284, + "learning_rate": 5.253865969303295e-06, + "loss": 0.2688, + "step": 22684 + }, + { + "epoch": 0.49933964340744, + "grad_norm": 1.6880841449082955, + "learning_rate": 5.253509972033147e-06, + "loss": 0.2845, + "step": 22685 + }, + { + "epoch": 0.4993616552938587, + "grad_norm": 1.7179649686252574, + "learning_rate": 5.25315397347455e-06, + "loss": 0.3143, + "step": 22686 + }, + { + "epoch": 0.4993836671802773, + "grad_norm": 1.2378127024917758, + "learning_rate": 5.252797973629309e-06, + "loss": 0.2846, + "step": 22687 + }, + { + "epoch": 0.49940567906669603, + "grad_norm": 1.3986383634856663, + "learning_rate": 5.25244197249924e-06, + "loss": 0.4006, + "step": 22688 + }, + { + "epoch": 0.4994276909531147, + "grad_norm": 1.3108103609797304, + "learning_rate": 5.252085970086145e-06, + "loss": 0.2743, + "step": 22689 + }, + { + "epoch": 0.4994497028395333, + "grad_norm": 1.454762175804389, + "learning_rate": 5.251729966391836e-06, + "loss": 0.3147, + "step": 22690 + }, + { + "epoch": 0.49947171472595203, + "grad_norm": 1.3469712881339133, + "learning_rate": 5.251373961418124e-06, + "loss": 0.3018, + "step": 22691 + }, + { + "epoch": 0.4994937266123707, + "grad_norm": 5.344405688913151, + "learning_rate": 5.251017955166819e-06, + "loss": 0.3171, + "step": 22692 + }, + { + "epoch": 0.49951573849878933, + "grad_norm": 1.2679569869239686, + "learning_rate": 5.250661947639728e-06, + "loss": 0.3091, + "step": 22693 + }, + { + "epoch": 0.49953775038520803, + "grad_norm": 0.6253002942149246, + "learning_rate": 5.250305938838661e-06, + "loss": 0.4738, + "step": 22694 + }, + { + "epoch": 0.4995597622716267, + "grad_norm": 1.258854179863138, + "learning_rate": 5.249949928765427e-06, + "loss": 0.2603, + "step": 22695 + }, + { + "epoch": 0.49958177415804533, + "grad_norm": 1.5630844146791218, + "learning_rate": 5.2495939174218355e-06, + "loss": 0.299, + "step": 22696 + }, + { + "epoch": 0.49960378604446404, + "grad_norm": 1.3116159046457379, + "learning_rate": 5.249237904809699e-06, + "loss": 0.274, + "step": 22697 + }, + { + "epoch": 0.4996257979308827, + "grad_norm": 1.1749004545718533, + "learning_rate": 5.2488818909308214e-06, + "loss": 0.2954, + "step": 22698 + }, + { + "epoch": 0.49964780981730134, + "grad_norm": 1.44913674157347, + "learning_rate": 5.248525875787017e-06, + "loss": 0.3736, + "step": 22699 + }, + { + "epoch": 0.49966982170372, + "grad_norm": 0.6187236414379937, + "learning_rate": 5.248169859380091e-06, + "loss": 0.4962, + "step": 22700 + }, + { + "epoch": 0.4996918335901387, + "grad_norm": 0.6278596008179617, + "learning_rate": 5.247813841711857e-06, + "loss": 0.4884, + "step": 22701 + }, + { + "epoch": 0.49971384547655734, + "grad_norm": 1.81791998722487, + "learning_rate": 5.247457822784122e-06, + "loss": 0.3021, + "step": 22702 + }, + { + "epoch": 0.499735857362976, + "grad_norm": 1.626799866278856, + "learning_rate": 5.247101802598696e-06, + "loss": 0.2604, + "step": 22703 + }, + { + "epoch": 0.4997578692493947, + "grad_norm": 1.6185773013018638, + "learning_rate": 5.24674578115739e-06, + "loss": 0.3109, + "step": 22704 + }, + { + "epoch": 0.49977988113581334, + "grad_norm": 0.6265437943577488, + "learning_rate": 5.24638975846201e-06, + "loss": 0.4927, + "step": 22705 + }, + { + "epoch": 0.499801893022232, + "grad_norm": 1.4020824163011676, + "learning_rate": 5.2460337345143676e-06, + "loss": 0.3294, + "step": 22706 + }, + { + "epoch": 0.4998239049086507, + "grad_norm": 18.266247547016643, + "learning_rate": 5.245677709316273e-06, + "loss": 0.3084, + "step": 22707 + }, + { + "epoch": 0.49984591679506934, + "grad_norm": 1.9646979733447139, + "learning_rate": 5.245321682869533e-06, + "loss": 0.3446, + "step": 22708 + }, + { + "epoch": 0.499867928681488, + "grad_norm": 0.6214872580233548, + "learning_rate": 5.2449656551759595e-06, + "loss": 0.4644, + "step": 22709 + }, + { + "epoch": 0.49988994056790664, + "grad_norm": 1.4370045700045324, + "learning_rate": 5.24460962623736e-06, + "loss": 0.3296, + "step": 22710 + }, + { + "epoch": 0.49991195245432535, + "grad_norm": 1.46027894928508, + "learning_rate": 5.244253596055548e-06, + "loss": 0.2463, + "step": 22711 + }, + { + "epoch": 0.499933964340744, + "grad_norm": 1.5690747898132074, + "learning_rate": 5.2438975646323295e-06, + "loss": 0.2794, + "step": 22712 + }, + { + "epoch": 0.49995597622716265, + "grad_norm": 2.1928725736337094, + "learning_rate": 5.243541531969514e-06, + "loss": 0.3721, + "step": 22713 + }, + { + "epoch": 0.49997798811358135, + "grad_norm": 0.6145477922129822, + "learning_rate": 5.243185498068913e-06, + "loss": 0.463, + "step": 22714 + }, + { + "epoch": 0.5, + "grad_norm": 1.1401027110048758, + "learning_rate": 5.242829462932333e-06, + "loss": 0.2728, + "step": 22715 + }, + { + "epoch": 0.5000220118864187, + "grad_norm": 5.791849747531467, + "learning_rate": 5.242473426561586e-06, + "loss": 0.3431, + "step": 22716 + }, + { + "epoch": 0.5000440237728373, + "grad_norm": 1.7425383619441344, + "learning_rate": 5.242117388958482e-06, + "loss": 0.2486, + "step": 22717 + }, + { + "epoch": 0.500066035659256, + "grad_norm": 0.6299688007132214, + "learning_rate": 5.241761350124827e-06, + "loss": 0.4945, + "step": 22718 + }, + { + "epoch": 0.5000880475456747, + "grad_norm": 0.6260125120568524, + "learning_rate": 5.241405310062435e-06, + "loss": 0.4726, + "step": 22719 + }, + { + "epoch": 0.5001100594320933, + "grad_norm": 1.385576383522629, + "learning_rate": 5.241049268773114e-06, + "loss": 0.3045, + "step": 22720 + }, + { + "epoch": 0.500132071318512, + "grad_norm": 1.3586279483267216, + "learning_rate": 5.240693226258673e-06, + "loss": 0.368, + "step": 22721 + }, + { + "epoch": 0.5001540832049307, + "grad_norm": 3.761742913671549, + "learning_rate": 5.24033718252092e-06, + "loss": 0.2852, + "step": 22722 + }, + { + "epoch": 0.5001760950913493, + "grad_norm": 3.7011446918644806, + "learning_rate": 5.239981137561667e-06, + "loss": 0.3591, + "step": 22723 + }, + { + "epoch": 0.500198106977768, + "grad_norm": 1.31186132333607, + "learning_rate": 5.239625091382724e-06, + "loss": 0.3913, + "step": 22724 + }, + { + "epoch": 0.5002201188641867, + "grad_norm": 2.761544433809958, + "learning_rate": 5.239269043985899e-06, + "loss": 0.3096, + "step": 22725 + }, + { + "epoch": 0.5002421307506053, + "grad_norm": 1.8937772287127077, + "learning_rate": 5.238912995373e-06, + "loss": 0.3351, + "step": 22726 + }, + { + "epoch": 0.500264142637024, + "grad_norm": 1.919516409829173, + "learning_rate": 5.238556945545841e-06, + "loss": 0.2508, + "step": 22727 + }, + { + "epoch": 0.5002861545234426, + "grad_norm": 1.4458253664900347, + "learning_rate": 5.238200894506227e-06, + "loss": 0.2984, + "step": 22728 + }, + { + "epoch": 0.5003081664098613, + "grad_norm": 1.122038063996468, + "learning_rate": 5.237844842255972e-06, + "loss": 0.2102, + "step": 22729 + }, + { + "epoch": 0.50033017829628, + "grad_norm": 1.5895886856013364, + "learning_rate": 5.2374887887968825e-06, + "loss": 0.4233, + "step": 22730 + }, + { + "epoch": 0.5003521901826986, + "grad_norm": 0.6627346682872554, + "learning_rate": 5.237132734130769e-06, + "loss": 0.4875, + "step": 22731 + }, + { + "epoch": 0.5003742020691173, + "grad_norm": 0.6093229945483407, + "learning_rate": 5.236776678259441e-06, + "loss": 0.4627, + "step": 22732 + }, + { + "epoch": 0.500396213955536, + "grad_norm": 0.6343747942753802, + "learning_rate": 5.236420621184709e-06, + "loss": 0.5085, + "step": 22733 + }, + { + "epoch": 0.5004182258419546, + "grad_norm": 1.5071100964299138, + "learning_rate": 5.2360645629083816e-06, + "loss": 0.2748, + "step": 22734 + }, + { + "epoch": 0.5004402377283733, + "grad_norm": 0.6053484597004831, + "learning_rate": 5.235708503432269e-06, + "loss": 0.4749, + "step": 22735 + }, + { + "epoch": 0.500462249614792, + "grad_norm": 1.6782822264594008, + "learning_rate": 5.235352442758179e-06, + "loss": 0.3904, + "step": 22736 + }, + { + "epoch": 0.5004842615012106, + "grad_norm": 1.4279139496079027, + "learning_rate": 5.234996380887924e-06, + "loss": 0.3508, + "step": 22737 + }, + { + "epoch": 0.5005062733876293, + "grad_norm": 0.6333968824825029, + "learning_rate": 5.234640317823314e-06, + "loss": 0.4691, + "step": 22738 + }, + { + "epoch": 0.500528285274048, + "grad_norm": 1.5615106135861976, + "learning_rate": 5.234284253566156e-06, + "loss": 0.3248, + "step": 22739 + }, + { + "epoch": 0.5005502971604666, + "grad_norm": 1.6935095592636908, + "learning_rate": 5.233928188118261e-06, + "loss": 0.2325, + "step": 22740 + }, + { + "epoch": 0.5005723090468853, + "grad_norm": 1.7527837954817436, + "learning_rate": 5.2335721214814384e-06, + "loss": 0.312, + "step": 22741 + }, + { + "epoch": 0.500594320933304, + "grad_norm": 0.619514859000717, + "learning_rate": 5.233216053657499e-06, + "loss": 0.4858, + "step": 22742 + }, + { + "epoch": 0.5006163328197226, + "grad_norm": 1.7933097022639344, + "learning_rate": 5.232859984648252e-06, + "loss": 0.349, + "step": 22743 + }, + { + "epoch": 0.5006383447061413, + "grad_norm": 1.45215593071372, + "learning_rate": 5.2325039144555046e-06, + "loss": 0.311, + "step": 22744 + }, + { + "epoch": 0.50066035659256, + "grad_norm": 1.4403333954301925, + "learning_rate": 5.23214784308107e-06, + "loss": 0.3552, + "step": 22745 + }, + { + "epoch": 0.5006823684789786, + "grad_norm": 1.00988674056824, + "learning_rate": 5.231791770526756e-06, + "loss": 0.3277, + "step": 22746 + }, + { + "epoch": 0.5007043803653973, + "grad_norm": 0.6368256041711414, + "learning_rate": 5.231435696794373e-06, + "loss": 0.5041, + "step": 22747 + }, + { + "epoch": 0.500726392251816, + "grad_norm": 1.6118550488129548, + "learning_rate": 5.231079621885731e-06, + "loss": 0.3266, + "step": 22748 + }, + { + "epoch": 0.5007484041382346, + "grad_norm": 2.238882187115398, + "learning_rate": 5.230723545802638e-06, + "loss": 0.3061, + "step": 22749 + }, + { + "epoch": 0.5007704160246533, + "grad_norm": 1.4665787574384392, + "learning_rate": 5.230367468546906e-06, + "loss": 0.377, + "step": 22750 + }, + { + "epoch": 0.5007924279110719, + "grad_norm": 1.749149610790496, + "learning_rate": 5.230011390120344e-06, + "loss": 0.3323, + "step": 22751 + }, + { + "epoch": 0.5008144397974906, + "grad_norm": 1.4997398320351463, + "learning_rate": 5.2296553105247614e-06, + "loss": 0.3406, + "step": 22752 + }, + { + "epoch": 0.5008364516839093, + "grad_norm": 1.803582752063348, + "learning_rate": 5.2292992297619685e-06, + "loss": 0.3616, + "step": 22753 + }, + { + "epoch": 0.5008584635703279, + "grad_norm": 1.2229966843993618, + "learning_rate": 5.228943147833773e-06, + "loss": 0.2807, + "step": 22754 + }, + { + "epoch": 0.5008804754567466, + "grad_norm": 1.4471439300923825, + "learning_rate": 5.228587064741988e-06, + "loss": 0.3375, + "step": 22755 + }, + { + "epoch": 0.5009024873431653, + "grad_norm": 0.592293091609271, + "learning_rate": 5.22823098048842e-06, + "loss": 0.4569, + "step": 22756 + }, + { + "epoch": 0.5009244992295839, + "grad_norm": 2.0017674990562706, + "learning_rate": 5.227874895074882e-06, + "loss": 0.3654, + "step": 22757 + }, + { + "epoch": 0.5009465111160026, + "grad_norm": 1.695760416210777, + "learning_rate": 5.227518808503182e-06, + "loss": 0.3778, + "step": 22758 + }, + { + "epoch": 0.5009685230024213, + "grad_norm": 0.6150973491926988, + "learning_rate": 5.227162720775128e-06, + "loss": 0.4948, + "step": 22759 + }, + { + "epoch": 0.5009905348888399, + "grad_norm": 0.6069573565053406, + "learning_rate": 5.2268066318925335e-06, + "loss": 0.4607, + "step": 22760 + }, + { + "epoch": 0.5010125467752586, + "grad_norm": 2.0639692608398534, + "learning_rate": 5.226450541857207e-06, + "loss": 0.334, + "step": 22761 + }, + { + "epoch": 0.5010345586616773, + "grad_norm": 1.2469502975021776, + "learning_rate": 5.226094450670957e-06, + "loss": 0.2222, + "step": 22762 + }, + { + "epoch": 0.5010565705480959, + "grad_norm": 0.6233043336929204, + "learning_rate": 5.225738358335594e-06, + "loss": 0.4887, + "step": 22763 + }, + { + "epoch": 0.5010785824345146, + "grad_norm": 1.4121562954684224, + "learning_rate": 5.225382264852928e-06, + "loss": 0.2729, + "step": 22764 + }, + { + "epoch": 0.5011005943209333, + "grad_norm": 3.9080284414255266, + "learning_rate": 5.225026170224769e-06, + "loss": 0.4044, + "step": 22765 + }, + { + "epoch": 0.5011226062073519, + "grad_norm": 1.4312691126752373, + "learning_rate": 5.224670074452928e-06, + "loss": 0.3499, + "step": 22766 + }, + { + "epoch": 0.5011446180937706, + "grad_norm": 4.1313139974822635, + "learning_rate": 5.2243139775392105e-06, + "loss": 0.336, + "step": 22767 + }, + { + "epoch": 0.5011666299801893, + "grad_norm": 1.194233557173055, + "learning_rate": 5.223957879485431e-06, + "loss": 0.3164, + "step": 22768 + }, + { + "epoch": 0.5011886418666079, + "grad_norm": 2.931631246603779, + "learning_rate": 5.223601780293398e-06, + "loss": 0.2058, + "step": 22769 + }, + { + "epoch": 0.5012106537530266, + "grad_norm": 3.3646957237044246, + "learning_rate": 5.22324567996492e-06, + "loss": 0.341, + "step": 22770 + }, + { + "epoch": 0.5012326656394454, + "grad_norm": 1.5603969182371773, + "learning_rate": 5.22288957850181e-06, + "loss": 0.3736, + "step": 22771 + }, + { + "epoch": 0.501254677525864, + "grad_norm": 6.990350042853028, + "learning_rate": 5.222533475905872e-06, + "loss": 0.2856, + "step": 22772 + }, + { + "epoch": 0.5012766894122826, + "grad_norm": 1.5719253533266295, + "learning_rate": 5.222177372178922e-06, + "loss": 0.3408, + "step": 22773 + }, + { + "epoch": 0.5012987012987012, + "grad_norm": 0.6144086358066699, + "learning_rate": 5.2218212673227686e-06, + "loss": 0.4844, + "step": 22774 + }, + { + "epoch": 0.50132071318512, + "grad_norm": 1.415416097353517, + "learning_rate": 5.221465161339219e-06, + "loss": 0.2972, + "step": 22775 + }, + { + "epoch": 0.5013427250715387, + "grad_norm": 1.9779658035151568, + "learning_rate": 5.221109054230085e-06, + "loss": 0.3464, + "step": 22776 + }, + { + "epoch": 0.5013647369579572, + "grad_norm": 1.188230023040174, + "learning_rate": 5.220752945997175e-06, + "loss": 0.2251, + "step": 22777 + }, + { + "epoch": 0.501386748844376, + "grad_norm": 1.8167304194986706, + "learning_rate": 5.220396836642302e-06, + "loss": 0.3116, + "step": 22778 + }, + { + "epoch": 0.5014087607307947, + "grad_norm": 2.236747091267747, + "learning_rate": 5.220040726167275e-06, + "loss": 0.3628, + "step": 22779 + }, + { + "epoch": 0.5014307726172132, + "grad_norm": 1.3312670223498804, + "learning_rate": 5.2196846145739e-06, + "loss": 0.3556, + "step": 22780 + }, + { + "epoch": 0.501452784503632, + "grad_norm": 1.9807622145436383, + "learning_rate": 5.219328501863991e-06, + "loss": 0.3192, + "step": 22781 + }, + { + "epoch": 0.5014747963900507, + "grad_norm": 1.377302875929335, + "learning_rate": 5.218972388039356e-06, + "loss": 0.2962, + "step": 22782 + }, + { + "epoch": 0.5014968082764693, + "grad_norm": 1.670313780112276, + "learning_rate": 5.218616273101808e-06, + "loss": 0.2508, + "step": 22783 + }, + { + "epoch": 0.501518820162888, + "grad_norm": 1.4257469737754775, + "learning_rate": 5.218260157053152e-06, + "loss": 0.2518, + "step": 22784 + }, + { + "epoch": 0.5015408320493067, + "grad_norm": 1.2641910281814464, + "learning_rate": 5.217904039895201e-06, + "loss": 0.3428, + "step": 22785 + }, + { + "epoch": 0.5015628439357253, + "grad_norm": 0.617381068972005, + "learning_rate": 5.217547921629766e-06, + "loss": 0.4645, + "step": 22786 + }, + { + "epoch": 0.501584855822144, + "grad_norm": 1.6711563724499818, + "learning_rate": 5.217191802258653e-06, + "loss": 0.2525, + "step": 22787 + }, + { + "epoch": 0.5016068677085627, + "grad_norm": 1.9024899378549935, + "learning_rate": 5.2168356817836765e-06, + "loss": 0.3369, + "step": 22788 + }, + { + "epoch": 0.5016288795949813, + "grad_norm": 0.6294869269240643, + "learning_rate": 5.216479560206643e-06, + "loss": 0.4861, + "step": 22789 + }, + { + "epoch": 0.5016508914814, + "grad_norm": 0.6119864834032245, + "learning_rate": 5.216123437529364e-06, + "loss": 0.4724, + "step": 22790 + }, + { + "epoch": 0.5016729033678187, + "grad_norm": 1.3355271417173071, + "learning_rate": 5.2157673137536505e-06, + "loss": 0.302, + "step": 22791 + }, + { + "epoch": 0.5016949152542373, + "grad_norm": 0.6104544338905403, + "learning_rate": 5.215411188881309e-06, + "loss": 0.4786, + "step": 22792 + }, + { + "epoch": 0.501716927140656, + "grad_norm": 1.6060274724127697, + "learning_rate": 5.215055062914155e-06, + "loss": 0.3906, + "step": 22793 + }, + { + "epoch": 0.5017389390270747, + "grad_norm": 1.927164244259235, + "learning_rate": 5.214698935853993e-06, + "loss": 0.3124, + "step": 22794 + }, + { + "epoch": 0.5017609509134933, + "grad_norm": 2.4715775219425127, + "learning_rate": 5.214342807702635e-06, + "loss": 0.272, + "step": 22795 + }, + { + "epoch": 0.501782962799912, + "grad_norm": 1.1899837943205307, + "learning_rate": 5.213986678461892e-06, + "loss": 0.291, + "step": 22796 + }, + { + "epoch": 0.5018049746863307, + "grad_norm": 1.4004650433500596, + "learning_rate": 5.213630548133574e-06, + "loss": 0.3776, + "step": 22797 + }, + { + "epoch": 0.5018269865727493, + "grad_norm": 0.611599425108497, + "learning_rate": 5.213274416719489e-06, + "loss": 0.472, + "step": 22798 + }, + { + "epoch": 0.501848998459168, + "grad_norm": 0.6506489219499879, + "learning_rate": 5.2129182842214495e-06, + "loss": 0.5047, + "step": 22799 + }, + { + "epoch": 0.5018710103455866, + "grad_norm": 2.045602046509256, + "learning_rate": 5.212562150641264e-06, + "loss": 0.2904, + "step": 22800 + }, + { + "epoch": 0.5018930222320053, + "grad_norm": 1.496745331621589, + "learning_rate": 5.212206015980742e-06, + "loss": 0.32, + "step": 22801 + }, + { + "epoch": 0.501915034118424, + "grad_norm": 1.447458239797136, + "learning_rate": 5.211849880241696e-06, + "loss": 0.3209, + "step": 22802 + }, + { + "epoch": 0.5019370460048426, + "grad_norm": 1.4308217857231258, + "learning_rate": 5.211493743425933e-06, + "loss": 0.2902, + "step": 22803 + }, + { + "epoch": 0.5019590578912613, + "grad_norm": 1.3717651348507658, + "learning_rate": 5.211137605535264e-06, + "loss": 0.287, + "step": 22804 + }, + { + "epoch": 0.50198106977768, + "grad_norm": 1.4664239220944548, + "learning_rate": 5.2107814665715e-06, + "loss": 0.3543, + "step": 22805 + }, + { + "epoch": 0.5020030816640986, + "grad_norm": 3.1454339644684226, + "learning_rate": 5.210425326536451e-06, + "loss": 0.2854, + "step": 22806 + }, + { + "epoch": 0.5020250935505173, + "grad_norm": 1.625895922545402, + "learning_rate": 5.210069185431927e-06, + "loss": 0.3397, + "step": 22807 + }, + { + "epoch": 0.502047105436936, + "grad_norm": 0.6356425609377812, + "learning_rate": 5.209713043259736e-06, + "loss": 0.468, + "step": 22808 + }, + { + "epoch": 0.5020691173233546, + "grad_norm": 1.1581656744279785, + "learning_rate": 5.209356900021691e-06, + "loss": 0.2716, + "step": 22809 + }, + { + "epoch": 0.5020911292097733, + "grad_norm": 0.599540119647808, + "learning_rate": 5.209000755719602e-06, + "loss": 0.4652, + "step": 22810 + }, + { + "epoch": 0.502113141096192, + "grad_norm": 1.6063194638288119, + "learning_rate": 5.208644610355277e-06, + "loss": 0.3146, + "step": 22811 + }, + { + "epoch": 0.5021351529826106, + "grad_norm": 0.6606503730786248, + "learning_rate": 5.208288463930527e-06, + "loss": 0.4918, + "step": 22812 + }, + { + "epoch": 0.5021571648690293, + "grad_norm": 1.3416834354138234, + "learning_rate": 5.207932316447161e-06, + "loss": 0.3449, + "step": 22813 + }, + { + "epoch": 0.502179176755448, + "grad_norm": 2.387116259145049, + "learning_rate": 5.207576167906991e-06, + "loss": 0.28, + "step": 22814 + }, + { + "epoch": 0.5022011886418666, + "grad_norm": 1.5726869648794048, + "learning_rate": 5.207220018311828e-06, + "loss": 0.3063, + "step": 22815 + }, + { + "epoch": 0.5022232005282853, + "grad_norm": 2.099015672182192, + "learning_rate": 5.206863867663479e-06, + "loss": 0.3469, + "step": 22816 + }, + { + "epoch": 0.502245212414704, + "grad_norm": 1.590609863622636, + "learning_rate": 5.206507715963754e-06, + "loss": 0.3582, + "step": 22817 + }, + { + "epoch": 0.5022672243011226, + "grad_norm": 1.893250466106406, + "learning_rate": 5.206151563214468e-06, + "loss": 0.3644, + "step": 22818 + }, + { + "epoch": 0.5022892361875413, + "grad_norm": 1.3651427334896054, + "learning_rate": 5.2057954094174246e-06, + "loss": 0.3552, + "step": 22819 + }, + { + "epoch": 0.50231124807396, + "grad_norm": 1.0967586631808213, + "learning_rate": 5.20543925457444e-06, + "loss": 0.2361, + "step": 22820 + }, + { + "epoch": 0.5023332599603786, + "grad_norm": 1.322198800971141, + "learning_rate": 5.205083098687319e-06, + "loss": 0.2047, + "step": 22821 + }, + { + "epoch": 0.5023552718467973, + "grad_norm": 1.2501489618282804, + "learning_rate": 5.204726941757875e-06, + "loss": 0.2503, + "step": 22822 + }, + { + "epoch": 0.5023772837332159, + "grad_norm": 1.692498957836004, + "learning_rate": 5.204370783787918e-06, + "loss": 0.4236, + "step": 22823 + }, + { + "epoch": 0.5023992956196346, + "grad_norm": 1.8282381628205397, + "learning_rate": 5.204014624779257e-06, + "loss": 0.3811, + "step": 22824 + }, + { + "epoch": 0.5024213075060533, + "grad_norm": 1.2807107048076167, + "learning_rate": 5.203658464733703e-06, + "loss": 0.3267, + "step": 22825 + }, + { + "epoch": 0.5024433193924719, + "grad_norm": 1.4508368315688434, + "learning_rate": 5.203302303653065e-06, + "loss": 0.3382, + "step": 22826 + }, + { + "epoch": 0.5024653312788906, + "grad_norm": 1.517224709669656, + "learning_rate": 5.2029461415391544e-06, + "loss": 0.3275, + "step": 22827 + }, + { + "epoch": 0.5024873431653093, + "grad_norm": 1.5308989206931543, + "learning_rate": 5.202589978393782e-06, + "loss": 0.32, + "step": 22828 + }, + { + "epoch": 0.5025093550517279, + "grad_norm": 1.3562250438407697, + "learning_rate": 5.202233814218756e-06, + "loss": 0.3206, + "step": 22829 + }, + { + "epoch": 0.5025313669381466, + "grad_norm": 1.50420491004272, + "learning_rate": 5.2018776490158885e-06, + "loss": 0.3504, + "step": 22830 + }, + { + "epoch": 0.5025533788245653, + "grad_norm": 1.5758262131657301, + "learning_rate": 5.201521482786987e-06, + "loss": 0.2712, + "step": 22831 + }, + { + "epoch": 0.5025753907109839, + "grad_norm": 1.8020532724158214, + "learning_rate": 5.2011653155338645e-06, + "loss": 0.2716, + "step": 22832 + }, + { + "epoch": 0.5025974025974026, + "grad_norm": 0.7134284400329868, + "learning_rate": 5.200809147258331e-06, + "loss": 0.4982, + "step": 22833 + }, + { + "epoch": 0.5026194144838213, + "grad_norm": 1.6805234872835235, + "learning_rate": 5.200452977962195e-06, + "loss": 0.3733, + "step": 22834 + }, + { + "epoch": 0.5026414263702399, + "grad_norm": 1.6872570218061376, + "learning_rate": 5.200096807647268e-06, + "loss": 0.4112, + "step": 22835 + }, + { + "epoch": 0.5026634382566586, + "grad_norm": 1.451208660793019, + "learning_rate": 5.199740636315359e-06, + "loss": 0.2977, + "step": 22836 + }, + { + "epoch": 0.5026854501430773, + "grad_norm": 1.3743615102182334, + "learning_rate": 5.199384463968279e-06, + "loss": 0.2388, + "step": 22837 + }, + { + "epoch": 0.5027074620294959, + "grad_norm": 1.559760475061277, + "learning_rate": 5.199028290607839e-06, + "loss": 0.2511, + "step": 22838 + }, + { + "epoch": 0.5027294739159146, + "grad_norm": 19.904420945066953, + "learning_rate": 5.198672116235848e-06, + "loss": 0.3301, + "step": 22839 + }, + { + "epoch": 0.5027514858023333, + "grad_norm": 0.6250345809074047, + "learning_rate": 5.198315940854116e-06, + "loss": 0.4706, + "step": 22840 + }, + { + "epoch": 0.5027734976887519, + "grad_norm": 1.959187455021105, + "learning_rate": 5.1979597644644555e-06, + "loss": 0.3546, + "step": 22841 + }, + { + "epoch": 0.5027955095751706, + "grad_norm": 1.3954986113752696, + "learning_rate": 5.197603587068674e-06, + "loss": 0.307, + "step": 22842 + }, + { + "epoch": 0.5028175214615893, + "grad_norm": 1.398756114418229, + "learning_rate": 5.1972474086685834e-06, + "loss": 0.2914, + "step": 22843 + }, + { + "epoch": 0.5028395333480079, + "grad_norm": 1.5908854158401535, + "learning_rate": 5.1968912292659925e-06, + "loss": 0.2929, + "step": 22844 + }, + { + "epoch": 0.5028615452344266, + "grad_norm": 0.6301403656311292, + "learning_rate": 5.196535048862714e-06, + "loss": 0.4814, + "step": 22845 + }, + { + "epoch": 0.5028835571208452, + "grad_norm": 1.687939451399152, + "learning_rate": 5.196178867460556e-06, + "loss": 0.3716, + "step": 22846 + }, + { + "epoch": 0.5029055690072639, + "grad_norm": 1.3521196038569492, + "learning_rate": 5.195822685061331e-06, + "loss": 0.2968, + "step": 22847 + }, + { + "epoch": 0.5029275808936826, + "grad_norm": 1.5379233444436937, + "learning_rate": 5.195466501666846e-06, + "loss": 0.2946, + "step": 22848 + }, + { + "epoch": 0.5029495927801012, + "grad_norm": 1.6100859560067915, + "learning_rate": 5.1951103172789115e-06, + "loss": 0.372, + "step": 22849 + }, + { + "epoch": 0.5029716046665199, + "grad_norm": 1.367492530629339, + "learning_rate": 5.194754131899342e-06, + "loss": 0.3593, + "step": 22850 + }, + { + "epoch": 0.5029936165529386, + "grad_norm": 1.4626525863472921, + "learning_rate": 5.1943979455299456e-06, + "loss": 0.3492, + "step": 22851 + }, + { + "epoch": 0.5030156284393572, + "grad_norm": 0.6351645081154758, + "learning_rate": 5.19404175817253e-06, + "loss": 0.4836, + "step": 22852 + }, + { + "epoch": 0.5030376403257759, + "grad_norm": 1.7387057727623618, + "learning_rate": 5.193685569828909e-06, + "loss": 0.4131, + "step": 22853 + }, + { + "epoch": 0.5030596522121946, + "grad_norm": 1.5477004474456746, + "learning_rate": 5.193329380500891e-06, + "loss": 0.3582, + "step": 22854 + }, + { + "epoch": 0.5030816640986132, + "grad_norm": 1.4357870337113117, + "learning_rate": 5.192973190190287e-06, + "loss": 0.2555, + "step": 22855 + }, + { + "epoch": 0.5031036759850319, + "grad_norm": 1.7426101692130176, + "learning_rate": 5.192616998898909e-06, + "loss": 0.3604, + "step": 22856 + }, + { + "epoch": 0.5031256878714506, + "grad_norm": 1.375145672801936, + "learning_rate": 5.192260806628562e-06, + "loss": 0.3352, + "step": 22857 + }, + { + "epoch": 0.5031476997578692, + "grad_norm": 1.142412685285029, + "learning_rate": 5.191904613381061e-06, + "loss": 0.2724, + "step": 22858 + }, + { + "epoch": 0.5031697116442879, + "grad_norm": 2.1494888539564774, + "learning_rate": 5.191548419158217e-06, + "loss": 0.3665, + "step": 22859 + }, + { + "epoch": 0.5031917235307066, + "grad_norm": 1.6383222961483201, + "learning_rate": 5.191192223961836e-06, + "loss": 0.2633, + "step": 22860 + }, + { + "epoch": 0.5032137354171252, + "grad_norm": 1.456354563640145, + "learning_rate": 5.190836027793733e-06, + "loss": 0.3228, + "step": 22861 + }, + { + "epoch": 0.5032357473035439, + "grad_norm": 0.6167492133059072, + "learning_rate": 5.190479830655715e-06, + "loss": 0.4565, + "step": 22862 + }, + { + "epoch": 0.5032577591899626, + "grad_norm": 1.5298221890418622, + "learning_rate": 5.190123632549594e-06, + "loss": 0.2759, + "step": 22863 + }, + { + "epoch": 0.5032797710763812, + "grad_norm": 1.757759299204799, + "learning_rate": 5.18976743347718e-06, + "loss": 0.3253, + "step": 22864 + }, + { + "epoch": 0.5033017829627999, + "grad_norm": 2.231826126335819, + "learning_rate": 5.189411233440282e-06, + "loss": 0.3721, + "step": 22865 + }, + { + "epoch": 0.5033237948492186, + "grad_norm": 1.341496660383982, + "learning_rate": 5.189055032440714e-06, + "loss": 0.2962, + "step": 22866 + }, + { + "epoch": 0.5033458067356372, + "grad_norm": 1.2946888575108646, + "learning_rate": 5.188698830480281e-06, + "loss": 0.2872, + "step": 22867 + }, + { + "epoch": 0.5033678186220559, + "grad_norm": 0.613665024891153, + "learning_rate": 5.188342627560798e-06, + "loss": 0.4772, + "step": 22868 + }, + { + "epoch": 0.5033898305084745, + "grad_norm": 1.3874461795013417, + "learning_rate": 5.1879864236840745e-06, + "loss": 0.3112, + "step": 22869 + }, + { + "epoch": 0.5034118423948932, + "grad_norm": 1.3769809417219236, + "learning_rate": 5.18763021885192e-06, + "loss": 0.2921, + "step": 22870 + }, + { + "epoch": 0.5034338542813119, + "grad_norm": 1.8926548485950532, + "learning_rate": 5.187274013066144e-06, + "loss": 0.302, + "step": 22871 + }, + { + "epoch": 0.5034558661677305, + "grad_norm": 1.3358394473248343, + "learning_rate": 5.186917806328559e-06, + "loss": 0.3637, + "step": 22872 + }, + { + "epoch": 0.5034778780541492, + "grad_norm": 1.6701050395885972, + "learning_rate": 5.186561598640972e-06, + "loss": 0.3483, + "step": 22873 + }, + { + "epoch": 0.5034998899405679, + "grad_norm": 1.1564340594082683, + "learning_rate": 5.186205390005199e-06, + "loss": 0.3528, + "step": 22874 + }, + { + "epoch": 0.5035219018269865, + "grad_norm": 0.6089407591543439, + "learning_rate": 5.185849180423045e-06, + "loss": 0.4535, + "step": 22875 + }, + { + "epoch": 0.5035439137134052, + "grad_norm": 1.5784910700533616, + "learning_rate": 5.185492969896323e-06, + "loss": 0.3602, + "step": 22876 + }, + { + "epoch": 0.5035659255998239, + "grad_norm": 1.3485249142858182, + "learning_rate": 5.185136758426844e-06, + "loss": 0.2901, + "step": 22877 + }, + { + "epoch": 0.5035879374862425, + "grad_norm": 1.3897031341522184, + "learning_rate": 5.184780546016416e-06, + "loss": 0.4612, + "step": 22878 + }, + { + "epoch": 0.5036099493726612, + "grad_norm": 1.1084661111949543, + "learning_rate": 5.184424332666852e-06, + "loss": 0.2825, + "step": 22879 + }, + { + "epoch": 0.5036319612590799, + "grad_norm": 2.0110968948990724, + "learning_rate": 5.18406811837996e-06, + "loss": 0.2695, + "step": 22880 + }, + { + "epoch": 0.5036539731454985, + "grad_norm": 0.6027343041892517, + "learning_rate": 5.183711903157554e-06, + "loss": 0.4638, + "step": 22881 + }, + { + "epoch": 0.5036759850319172, + "grad_norm": 1.343393598956052, + "learning_rate": 5.18335568700144e-06, + "loss": 0.3254, + "step": 22882 + }, + { + "epoch": 0.5036979969183359, + "grad_norm": 1.933787479503174, + "learning_rate": 5.182999469913431e-06, + "loss": 0.3549, + "step": 22883 + }, + { + "epoch": 0.5037200088047545, + "grad_norm": 1.9776361613410023, + "learning_rate": 5.182643251895337e-06, + "loss": 0.2313, + "step": 22884 + }, + { + "epoch": 0.5037420206911732, + "grad_norm": 1.516344481097314, + "learning_rate": 5.182287032948968e-06, + "loss": 0.2942, + "step": 22885 + }, + { + "epoch": 0.503764032577592, + "grad_norm": 1.5870876883817608, + "learning_rate": 5.1819308130761366e-06, + "loss": 0.3339, + "step": 22886 + }, + { + "epoch": 0.5037860444640105, + "grad_norm": 1.386151909513119, + "learning_rate": 5.181574592278651e-06, + "loss": 0.2956, + "step": 22887 + }, + { + "epoch": 0.5038080563504292, + "grad_norm": 1.984066728764093, + "learning_rate": 5.181218370558323e-06, + "loss": 0.2474, + "step": 22888 + }, + { + "epoch": 0.503830068236848, + "grad_norm": 0.6063287834246447, + "learning_rate": 5.180862147916962e-06, + "loss": 0.4316, + "step": 22889 + }, + { + "epoch": 0.5038520801232665, + "grad_norm": 2.101855127785787, + "learning_rate": 5.180505924356379e-06, + "loss": 0.2367, + "step": 22890 + }, + { + "epoch": 0.5038740920096852, + "grad_norm": 1.9127679899038483, + "learning_rate": 5.180149699878383e-06, + "loss": 0.3702, + "step": 22891 + }, + { + "epoch": 0.5038961038961038, + "grad_norm": 2.400347813050351, + "learning_rate": 5.179793474484788e-06, + "loss": 0.2558, + "step": 22892 + }, + { + "epoch": 0.5039181157825225, + "grad_norm": 3.605579410619498, + "learning_rate": 5.1794372481774e-06, + "loss": 0.4378, + "step": 22893 + }, + { + "epoch": 0.5039401276689413, + "grad_norm": 1.3119070372327346, + "learning_rate": 5.1790810209580335e-06, + "loss": 0.2256, + "step": 22894 + }, + { + "epoch": 0.5039621395553598, + "grad_norm": 1.7929321828793592, + "learning_rate": 5.178724792828497e-06, + "loss": 0.3264, + "step": 22895 + }, + { + "epoch": 0.5039841514417785, + "grad_norm": 1.7724424902430018, + "learning_rate": 5.1783685637906e-06, + "loss": 0.3632, + "step": 22896 + }, + { + "epoch": 0.5040061633281973, + "grad_norm": 1.3712718433571272, + "learning_rate": 5.178012333846157e-06, + "loss": 0.2992, + "step": 22897 + }, + { + "epoch": 0.5040281752146158, + "grad_norm": 1.4695986046130611, + "learning_rate": 5.177656102996974e-06, + "loss": 0.2841, + "step": 22898 + }, + { + "epoch": 0.5040501871010346, + "grad_norm": 1.321776101232956, + "learning_rate": 5.177299871244864e-06, + "loss": 0.3239, + "step": 22899 + }, + { + "epoch": 0.5040721989874533, + "grad_norm": 1.2328198175876117, + "learning_rate": 5.176943638591639e-06, + "loss": 0.2314, + "step": 22900 + }, + { + "epoch": 0.5040942108738719, + "grad_norm": 1.7205547192263009, + "learning_rate": 5.176587405039104e-06, + "loss": 0.2685, + "step": 22901 + }, + { + "epoch": 0.5041162227602906, + "grad_norm": 1.0387367532433829, + "learning_rate": 5.176231170589076e-06, + "loss": 0.2176, + "step": 22902 + }, + { + "epoch": 0.5041382346467093, + "grad_norm": 0.6145038292129889, + "learning_rate": 5.17587493524336e-06, + "loss": 0.4703, + "step": 22903 + }, + { + "epoch": 0.5041602465331279, + "grad_norm": 1.3734951772094024, + "learning_rate": 5.175518699003771e-06, + "loss": 0.2592, + "step": 22904 + }, + { + "epoch": 0.5041822584195466, + "grad_norm": 1.6175780506509103, + "learning_rate": 5.175162461872116e-06, + "loss": 0.3152, + "step": 22905 + }, + { + "epoch": 0.5042042703059653, + "grad_norm": 1.293905300947427, + "learning_rate": 5.1748062238502095e-06, + "loss": 0.267, + "step": 22906 + }, + { + "epoch": 0.5042262821923839, + "grad_norm": 1.7514233572359865, + "learning_rate": 5.1744499849398585e-06, + "loss": 0.326, + "step": 22907 + }, + { + "epoch": 0.5042482940788026, + "grad_norm": 1.7606034862845585, + "learning_rate": 5.174093745142876e-06, + "loss": 0.2959, + "step": 22908 + }, + { + "epoch": 0.5042703059652213, + "grad_norm": 0.6126204868778966, + "learning_rate": 5.173737504461069e-06, + "loss": 0.4617, + "step": 22909 + }, + { + "epoch": 0.5042923178516399, + "grad_norm": 1.3436534452405449, + "learning_rate": 5.173381262896252e-06, + "loss": 0.304, + "step": 22910 + }, + { + "epoch": 0.5043143297380586, + "grad_norm": 1.4128624275811132, + "learning_rate": 5.173025020450234e-06, + "loss": 0.3118, + "step": 22911 + }, + { + "epoch": 0.5043363416244773, + "grad_norm": 0.6094089718274831, + "learning_rate": 5.172668777124827e-06, + "loss": 0.497, + "step": 22912 + }, + { + "epoch": 0.5043583535108959, + "grad_norm": 1.349735288128378, + "learning_rate": 5.172312532921838e-06, + "loss": 0.3181, + "step": 22913 + }, + { + "epoch": 0.5043803653973146, + "grad_norm": 3.636325985715598, + "learning_rate": 5.171956287843081e-06, + "loss": 0.3445, + "step": 22914 + }, + { + "epoch": 0.5044023772837332, + "grad_norm": 2.1844690275987872, + "learning_rate": 5.171600041890366e-06, + "loss": 0.2793, + "step": 22915 + }, + { + "epoch": 0.5044243891701519, + "grad_norm": 1.3673621273545276, + "learning_rate": 5.171243795065501e-06, + "loss": 0.2785, + "step": 22916 + }, + { + "epoch": 0.5044464010565706, + "grad_norm": 1.166918457285404, + "learning_rate": 5.1708875473703005e-06, + "loss": 0.2105, + "step": 22917 + }, + { + "epoch": 0.5044684129429892, + "grad_norm": 1.6436187776144615, + "learning_rate": 5.170531298806572e-06, + "loss": 0.3424, + "step": 22918 + }, + { + "epoch": 0.5044904248294079, + "grad_norm": 1.3676060276233493, + "learning_rate": 5.170175049376128e-06, + "loss": 0.3405, + "step": 22919 + }, + { + "epoch": 0.5045124367158266, + "grad_norm": 1.8020908342883468, + "learning_rate": 5.16981879908078e-06, + "loss": 0.3066, + "step": 22920 + }, + { + "epoch": 0.5045344486022452, + "grad_norm": 1.5745957390282181, + "learning_rate": 5.169462547922335e-06, + "loss": 0.3406, + "step": 22921 + }, + { + "epoch": 0.5045564604886639, + "grad_norm": 1.721004376609238, + "learning_rate": 5.169106295902606e-06, + "loss": 0.3007, + "step": 22922 + }, + { + "epoch": 0.5045784723750826, + "grad_norm": 1.7153493065766952, + "learning_rate": 5.168750043023405e-06, + "loss": 0.3389, + "step": 22923 + }, + { + "epoch": 0.5046004842615012, + "grad_norm": 0.6291685116092188, + "learning_rate": 5.16839378928654e-06, + "loss": 0.483, + "step": 22924 + }, + { + "epoch": 0.5046224961479199, + "grad_norm": 1.4837338481130617, + "learning_rate": 5.168037534693823e-06, + "loss": 0.3304, + "step": 22925 + }, + { + "epoch": 0.5046445080343386, + "grad_norm": 1.490364414269159, + "learning_rate": 5.1676812792470635e-06, + "loss": 0.3368, + "step": 22926 + }, + { + "epoch": 0.5046665199207572, + "grad_norm": 1.3952948151213367, + "learning_rate": 5.167325022948072e-06, + "loss": 0.403, + "step": 22927 + }, + { + "epoch": 0.5046885318071759, + "grad_norm": 1.2815016654352571, + "learning_rate": 5.166968765798662e-06, + "loss": 0.2871, + "step": 22928 + }, + { + "epoch": 0.5047105436935946, + "grad_norm": 1.4682762276801522, + "learning_rate": 5.166612507800642e-06, + "loss": 0.2525, + "step": 22929 + }, + { + "epoch": 0.5047325555800132, + "grad_norm": 1.4420182921468856, + "learning_rate": 5.166256248955823e-06, + "loss": 0.3883, + "step": 22930 + }, + { + "epoch": 0.5047545674664319, + "grad_norm": 0.5814740406299443, + "learning_rate": 5.165899989266016e-06, + "loss": 0.473, + "step": 22931 + }, + { + "epoch": 0.5047765793528506, + "grad_norm": 1.3858549073435729, + "learning_rate": 5.165543728733029e-06, + "loss": 0.341, + "step": 22932 + }, + { + "epoch": 0.5047985912392692, + "grad_norm": 1.7673394383338448, + "learning_rate": 5.1651874673586774e-06, + "loss": 0.2762, + "step": 22933 + }, + { + "epoch": 0.5048206031256879, + "grad_norm": 7.665032769307584, + "learning_rate": 5.164831205144768e-06, + "loss": 0.2761, + "step": 22934 + }, + { + "epoch": 0.5048426150121066, + "grad_norm": 1.2609939585938068, + "learning_rate": 5.164474942093113e-06, + "loss": 0.2881, + "step": 22935 + }, + { + "epoch": 0.5048646268985252, + "grad_norm": 0.614672229810092, + "learning_rate": 5.164118678205523e-06, + "loss": 0.4825, + "step": 22936 + }, + { + "epoch": 0.5048866387849439, + "grad_norm": 1.7694118213476855, + "learning_rate": 5.163762413483809e-06, + "loss": 0.3298, + "step": 22937 + }, + { + "epoch": 0.5049086506713626, + "grad_norm": 1.45591747878399, + "learning_rate": 5.163406147929783e-06, + "loss": 0.3198, + "step": 22938 + }, + { + "epoch": 0.5049306625577812, + "grad_norm": 1.4927652914471103, + "learning_rate": 5.1630498815452516e-06, + "loss": 0.3379, + "step": 22939 + }, + { + "epoch": 0.5049526744441999, + "grad_norm": 1.4704345187318668, + "learning_rate": 5.162693614332028e-06, + "loss": 0.3986, + "step": 22940 + }, + { + "epoch": 0.5049746863306185, + "grad_norm": 1.2542240723406664, + "learning_rate": 5.162337346291924e-06, + "loss": 0.351, + "step": 22941 + }, + { + "epoch": 0.5049966982170372, + "grad_norm": 1.47108755640607, + "learning_rate": 5.1619810774267485e-06, + "loss": 0.3291, + "step": 22942 + }, + { + "epoch": 0.5050187101034559, + "grad_norm": 1.60328764974866, + "learning_rate": 5.161624807738314e-06, + "loss": 0.2453, + "step": 22943 + }, + { + "epoch": 0.5050407219898745, + "grad_norm": 1.6715248486502783, + "learning_rate": 5.16126853722843e-06, + "loss": 0.2993, + "step": 22944 + }, + { + "epoch": 0.5050627338762932, + "grad_norm": 1.6451525251366832, + "learning_rate": 5.160912265898905e-06, + "loss": 0.3142, + "step": 22945 + }, + { + "epoch": 0.5050847457627119, + "grad_norm": 3.0545949797502328, + "learning_rate": 5.160555993751555e-06, + "loss": 0.3209, + "step": 22946 + }, + { + "epoch": 0.5051067576491305, + "grad_norm": 1.2679707385791819, + "learning_rate": 5.160199720788187e-06, + "loss": 0.3265, + "step": 22947 + }, + { + "epoch": 0.5051287695355492, + "grad_norm": 1.4035827410154353, + "learning_rate": 5.159843447010613e-06, + "loss": 0.2747, + "step": 22948 + }, + { + "epoch": 0.5051507814219679, + "grad_norm": 2.2664140510682076, + "learning_rate": 5.159487172420642e-06, + "loss": 0.3601, + "step": 22949 + }, + { + "epoch": 0.5051727933083865, + "grad_norm": 2.0231356337691344, + "learning_rate": 5.159130897020086e-06, + "loss": 0.2369, + "step": 22950 + }, + { + "epoch": 0.5051948051948052, + "grad_norm": 1.6792902776999645, + "learning_rate": 5.158774620810757e-06, + "loss": 0.3854, + "step": 22951 + }, + { + "epoch": 0.5052168170812239, + "grad_norm": 1.755213848730644, + "learning_rate": 5.158418343794463e-06, + "loss": 0.3931, + "step": 22952 + }, + { + "epoch": 0.5052388289676425, + "grad_norm": 1.451779401135863, + "learning_rate": 5.158062065973019e-06, + "loss": 0.4591, + "step": 22953 + }, + { + "epoch": 0.5052608408540612, + "grad_norm": 2.057994827012591, + "learning_rate": 5.15770578734823e-06, + "loss": 0.3496, + "step": 22954 + }, + { + "epoch": 0.5052828527404799, + "grad_norm": 2.6798596442106137, + "learning_rate": 5.157349507921911e-06, + "loss": 0.3679, + "step": 22955 + }, + { + "epoch": 0.5053048646268985, + "grad_norm": 1.280125046266829, + "learning_rate": 5.156993227695873e-06, + "loss": 0.3335, + "step": 22956 + }, + { + "epoch": 0.5053268765133172, + "grad_norm": 1.1752312792604265, + "learning_rate": 5.156636946671926e-06, + "loss": 0.2936, + "step": 22957 + }, + { + "epoch": 0.5053488883997359, + "grad_norm": 1.7042710220628376, + "learning_rate": 5.156280664851877e-06, + "loss": 0.2896, + "step": 22958 + }, + { + "epoch": 0.5053709002861545, + "grad_norm": 4.027832520361686, + "learning_rate": 5.155924382237542e-06, + "loss": 0.3364, + "step": 22959 + }, + { + "epoch": 0.5053929121725732, + "grad_norm": 1.7015544332699672, + "learning_rate": 5.155568098830728e-06, + "loss": 0.2366, + "step": 22960 + }, + { + "epoch": 0.5054149240589919, + "grad_norm": 1.4723082501470264, + "learning_rate": 5.1552118146332495e-06, + "loss": 0.3173, + "step": 22961 + }, + { + "epoch": 0.5054369359454105, + "grad_norm": 1.4977726374501863, + "learning_rate": 5.154855529646915e-06, + "loss": 0.3364, + "step": 22962 + }, + { + "epoch": 0.5054589478318292, + "grad_norm": 1.930228247953816, + "learning_rate": 5.154499243873534e-06, + "loss": 0.3889, + "step": 22963 + }, + { + "epoch": 0.5054809597182478, + "grad_norm": 1.706162892228345, + "learning_rate": 5.15414295731492e-06, + "loss": 0.3476, + "step": 22964 + }, + { + "epoch": 0.5055029716046665, + "grad_norm": 1.3505824318749893, + "learning_rate": 5.153786669972884e-06, + "loss": 0.2899, + "step": 22965 + }, + { + "epoch": 0.5055249834910852, + "grad_norm": 1.5153032966695024, + "learning_rate": 5.153430381849235e-06, + "loss": 0.2787, + "step": 22966 + }, + { + "epoch": 0.5055469953775038, + "grad_norm": 1.6402984900713702, + "learning_rate": 5.153074092945784e-06, + "loss": 0.3011, + "step": 22967 + }, + { + "epoch": 0.5055690072639225, + "grad_norm": 1.515511010198138, + "learning_rate": 5.15271780326434e-06, + "loss": 0.3417, + "step": 22968 + }, + { + "epoch": 0.5055910191503412, + "grad_norm": 0.6972569126750553, + "learning_rate": 5.1523615128067186e-06, + "loss": 0.4732, + "step": 22969 + }, + { + "epoch": 0.5056130310367598, + "grad_norm": 0.6794826170568815, + "learning_rate": 5.152005221574727e-06, + "loss": 0.472, + "step": 22970 + }, + { + "epoch": 0.5056350429231785, + "grad_norm": 0.6256217998855862, + "learning_rate": 5.151648929570178e-06, + "loss": 0.4745, + "step": 22971 + }, + { + "epoch": 0.5056570548095972, + "grad_norm": 1.422535258183841, + "learning_rate": 5.151292636794881e-06, + "loss": 0.2857, + "step": 22972 + }, + { + "epoch": 0.5056790666960158, + "grad_norm": 1.5336950077300928, + "learning_rate": 5.150936343250647e-06, + "loss": 0.2386, + "step": 22973 + }, + { + "epoch": 0.5057010785824345, + "grad_norm": 1.662996068560206, + "learning_rate": 5.150580048939289e-06, + "loss": 0.3607, + "step": 22974 + }, + { + "epoch": 0.5057230904688532, + "grad_norm": 1.2670510137323103, + "learning_rate": 5.1502237538626145e-06, + "loss": 0.3263, + "step": 22975 + }, + { + "epoch": 0.5057451023552718, + "grad_norm": 2.034031898184641, + "learning_rate": 5.149867458022435e-06, + "loss": 0.3584, + "step": 22976 + }, + { + "epoch": 0.5057671142416905, + "grad_norm": 1.59498079320767, + "learning_rate": 5.149511161420564e-06, + "loss": 0.3163, + "step": 22977 + }, + { + "epoch": 0.5057891261281092, + "grad_norm": 1.438469269086457, + "learning_rate": 5.149154864058808e-06, + "loss": 0.3064, + "step": 22978 + }, + { + "epoch": 0.5058111380145278, + "grad_norm": 2.1300453794146144, + "learning_rate": 5.148798565938984e-06, + "loss": 0.4031, + "step": 22979 + }, + { + "epoch": 0.5058331499009465, + "grad_norm": 1.0834263665461186, + "learning_rate": 5.148442267062897e-06, + "loss": 0.2809, + "step": 22980 + }, + { + "epoch": 0.5058551617873652, + "grad_norm": 1.7426277730144222, + "learning_rate": 5.14808596743236e-06, + "loss": 0.3771, + "step": 22981 + }, + { + "epoch": 0.5058771736737838, + "grad_norm": 1.8106326730069258, + "learning_rate": 5.147729667049185e-06, + "loss": 0.3525, + "step": 22982 + }, + { + "epoch": 0.5058991855602025, + "grad_norm": 0.7922224423939869, + "learning_rate": 5.1473733659151826e-06, + "loss": 0.4812, + "step": 22983 + }, + { + "epoch": 0.5059211974466212, + "grad_norm": 1.3698985565751562, + "learning_rate": 5.147017064032162e-06, + "loss": 0.4099, + "step": 22984 + }, + { + "epoch": 0.5059432093330398, + "grad_norm": 1.4454402500237402, + "learning_rate": 5.146660761401936e-06, + "loss": 0.3511, + "step": 22985 + }, + { + "epoch": 0.5059652212194585, + "grad_norm": 1.327314925537339, + "learning_rate": 5.146304458026312e-06, + "loss": 0.3436, + "step": 22986 + }, + { + "epoch": 0.5059872331058771, + "grad_norm": 1.7610858735308461, + "learning_rate": 5.145948153907107e-06, + "loss": 0.4051, + "step": 22987 + }, + { + "epoch": 0.5060092449922958, + "grad_norm": 1.3414486988778382, + "learning_rate": 5.145591849046127e-06, + "loss": 0.2611, + "step": 22988 + }, + { + "epoch": 0.5060312568787145, + "grad_norm": 1.3944162404991498, + "learning_rate": 5.145235543445183e-06, + "loss": 0.3192, + "step": 22989 + }, + { + "epoch": 0.5060532687651331, + "grad_norm": 1.5940400362591909, + "learning_rate": 5.144879237106089e-06, + "loss": 0.2833, + "step": 22990 + }, + { + "epoch": 0.5060752806515518, + "grad_norm": 1.665171295580072, + "learning_rate": 5.144522930030652e-06, + "loss": 0.3552, + "step": 22991 + }, + { + "epoch": 0.5060972925379705, + "grad_norm": 1.6638239448231107, + "learning_rate": 5.144166622220688e-06, + "loss": 0.3357, + "step": 22992 + }, + { + "epoch": 0.5061193044243891, + "grad_norm": 0.6213997953524951, + "learning_rate": 5.143810313678003e-06, + "loss": 0.4819, + "step": 22993 + }, + { + "epoch": 0.5061413163108078, + "grad_norm": 1.3367110990386117, + "learning_rate": 5.143454004404409e-06, + "loss": 0.2885, + "step": 22994 + }, + { + "epoch": 0.5061633281972265, + "grad_norm": 1.5373205407529906, + "learning_rate": 5.14309769440172e-06, + "loss": 0.4011, + "step": 22995 + }, + { + "epoch": 0.5061853400836451, + "grad_norm": 1.3452435656815003, + "learning_rate": 5.142741383671742e-06, + "loss": 0.2833, + "step": 22996 + }, + { + "epoch": 0.5062073519700638, + "grad_norm": 2.1003730943990013, + "learning_rate": 5.142385072216292e-06, + "loss": 0.3827, + "step": 22997 + }, + { + "epoch": 0.5062293638564825, + "grad_norm": 3.4166370693293584, + "learning_rate": 5.142028760037176e-06, + "loss": 0.3578, + "step": 22998 + }, + { + "epoch": 0.5062513757429011, + "grad_norm": 1.2630140433764836, + "learning_rate": 5.1416724471362045e-06, + "loss": 0.2879, + "step": 22999 + }, + { + "epoch": 0.5062733876293198, + "grad_norm": 0.6225796718966572, + "learning_rate": 5.141316133515193e-06, + "loss": 0.4804, + "step": 23000 + }, + { + "epoch": 0.5062953995157385, + "grad_norm": 1.556163053959755, + "learning_rate": 5.1409598191759495e-06, + "loss": 0.3069, + "step": 23001 + }, + { + "epoch": 0.5063174114021571, + "grad_norm": 1.1374906204582005, + "learning_rate": 5.140603504120285e-06, + "loss": 0.3, + "step": 23002 + }, + { + "epoch": 0.5063394232885758, + "grad_norm": 1.8965224712104427, + "learning_rate": 5.14024718835001e-06, + "loss": 0.3182, + "step": 23003 + }, + { + "epoch": 0.5063614351749945, + "grad_norm": 1.2048001126129269, + "learning_rate": 5.139890871866936e-06, + "loss": 0.2886, + "step": 23004 + }, + { + "epoch": 0.5063834470614131, + "grad_norm": 1.3449979737634785, + "learning_rate": 5.139534554672876e-06, + "loss": 0.3548, + "step": 23005 + }, + { + "epoch": 0.5064054589478318, + "grad_norm": 1.3562115068075826, + "learning_rate": 5.139178236769639e-06, + "loss": 0.2777, + "step": 23006 + }, + { + "epoch": 0.5064274708342505, + "grad_norm": 1.7190087424834746, + "learning_rate": 5.138821918159035e-06, + "loss": 0.2954, + "step": 23007 + }, + { + "epoch": 0.5064494827206691, + "grad_norm": 1.7947340803176481, + "learning_rate": 5.1384655988428765e-06, + "loss": 0.3338, + "step": 23008 + }, + { + "epoch": 0.5064714946070878, + "grad_norm": 1.9328174010534358, + "learning_rate": 5.138109278822974e-06, + "loss": 0.3262, + "step": 23009 + }, + { + "epoch": 0.5064935064935064, + "grad_norm": 2.214928233936534, + "learning_rate": 5.137752958101137e-06, + "loss": 0.4061, + "step": 23010 + }, + { + "epoch": 0.5065155183799251, + "grad_norm": 2.2114120637931207, + "learning_rate": 5.137396636679182e-06, + "loss": 0.2812, + "step": 23011 + }, + { + "epoch": 0.5065375302663438, + "grad_norm": 0.6389205607449777, + "learning_rate": 5.137040314558912e-06, + "loss": 0.4507, + "step": 23012 + }, + { + "epoch": 0.5065595421527624, + "grad_norm": 1.2941322778133009, + "learning_rate": 5.136683991742144e-06, + "loss": 0.3012, + "step": 23013 + }, + { + "epoch": 0.5065815540391811, + "grad_norm": 0.6011177621116188, + "learning_rate": 5.136327668230687e-06, + "loss": 0.452, + "step": 23014 + }, + { + "epoch": 0.5066035659255999, + "grad_norm": 0.6095757575845712, + "learning_rate": 5.135971344026353e-06, + "loss": 0.4706, + "step": 23015 + }, + { + "epoch": 0.5066255778120184, + "grad_norm": 1.3759252308311276, + "learning_rate": 5.1356150191309496e-06, + "loss": 0.2622, + "step": 23016 + }, + { + "epoch": 0.5066475896984372, + "grad_norm": 1.3698331688668488, + "learning_rate": 5.13525869354629e-06, + "loss": 0.3352, + "step": 23017 + }, + { + "epoch": 0.5066696015848559, + "grad_norm": 1.5462040016160796, + "learning_rate": 5.1349023672741875e-06, + "loss": 0.3625, + "step": 23018 + }, + { + "epoch": 0.5066916134712744, + "grad_norm": 1.3089810572868008, + "learning_rate": 5.1345460403164505e-06, + "loss": 0.3195, + "step": 23019 + }, + { + "epoch": 0.5067136253576932, + "grad_norm": 0.6146732164640311, + "learning_rate": 5.1341897126748894e-06, + "loss": 0.4873, + "step": 23020 + }, + { + "epoch": 0.5067356372441119, + "grad_norm": 1.2546232458178568, + "learning_rate": 5.133833384351318e-06, + "loss": 0.3071, + "step": 23021 + }, + { + "epoch": 0.5067576491305305, + "grad_norm": 1.9614802585250268, + "learning_rate": 5.133477055347543e-06, + "loss": 0.3836, + "step": 23022 + }, + { + "epoch": 0.5067796610169492, + "grad_norm": 1.4063251636990142, + "learning_rate": 5.1331207256653805e-06, + "loss": 0.2881, + "step": 23023 + }, + { + "epoch": 0.5068016729033679, + "grad_norm": 1.2391331290185508, + "learning_rate": 5.1327643953066386e-06, + "loss": 0.3207, + "step": 23024 + }, + { + "epoch": 0.5068236847897865, + "grad_norm": 0.6429735238375636, + "learning_rate": 5.132408064273129e-06, + "loss": 0.4843, + "step": 23025 + }, + { + "epoch": 0.5068456966762052, + "grad_norm": 1.5313471985893023, + "learning_rate": 5.132051732566662e-06, + "loss": 0.2621, + "step": 23026 + }, + { + "epoch": 0.5068677085626239, + "grad_norm": 1.3877319409503326, + "learning_rate": 5.131695400189048e-06, + "loss": 0.2908, + "step": 23027 + }, + { + "epoch": 0.5068897204490425, + "grad_norm": 1.3317829242843198, + "learning_rate": 5.1313390671421015e-06, + "loss": 0.2934, + "step": 23028 + }, + { + "epoch": 0.5069117323354612, + "grad_norm": 1.4566615182234297, + "learning_rate": 5.130982733427632e-06, + "loss": 0.3065, + "step": 23029 + }, + { + "epoch": 0.5069337442218799, + "grad_norm": 1.658501527557603, + "learning_rate": 5.130626399047447e-06, + "loss": 0.3714, + "step": 23030 + }, + { + "epoch": 0.5069557561082985, + "grad_norm": 1.2202757162870586, + "learning_rate": 5.130270064003361e-06, + "loss": 0.3144, + "step": 23031 + }, + { + "epoch": 0.5069777679947172, + "grad_norm": 1.2426229384118885, + "learning_rate": 5.129913728297185e-06, + "loss": 0.38, + "step": 23032 + }, + { + "epoch": 0.5069997798811358, + "grad_norm": 1.2985338328251632, + "learning_rate": 5.129557391930731e-06, + "loss": 0.3013, + "step": 23033 + }, + { + "epoch": 0.5070217917675545, + "grad_norm": 0.641784833263402, + "learning_rate": 5.129201054905807e-06, + "loss": 0.4872, + "step": 23034 + }, + { + "epoch": 0.5070438036539732, + "grad_norm": 2.2679282027230454, + "learning_rate": 5.128844717224224e-06, + "loss": 0.3924, + "step": 23035 + }, + { + "epoch": 0.5070658155403918, + "grad_norm": 1.5743198035759818, + "learning_rate": 5.128488378887797e-06, + "loss": 0.2762, + "step": 23036 + }, + { + "epoch": 0.5070878274268105, + "grad_norm": 1.4500806689177268, + "learning_rate": 5.1281320398983334e-06, + "loss": 0.3234, + "step": 23037 + }, + { + "epoch": 0.5071098393132292, + "grad_norm": 1.2763802985105115, + "learning_rate": 5.127775700257647e-06, + "loss": 0.2415, + "step": 23038 + }, + { + "epoch": 0.5071318511996478, + "grad_norm": 1.7328699052931826, + "learning_rate": 5.127419359967547e-06, + "loss": 0.3327, + "step": 23039 + }, + { + "epoch": 0.5071538630860665, + "grad_norm": 1.2603843771630592, + "learning_rate": 5.1270630190298446e-06, + "loss": 0.2964, + "step": 23040 + }, + { + "epoch": 0.5071758749724852, + "grad_norm": 2.7073681306014064, + "learning_rate": 5.126706677446352e-06, + "loss": 0.3293, + "step": 23041 + }, + { + "epoch": 0.5071978868589038, + "grad_norm": 1.301552557734695, + "learning_rate": 5.126350335218878e-06, + "loss": 0.3426, + "step": 23042 + }, + { + "epoch": 0.5072198987453225, + "grad_norm": 1.8993117236901924, + "learning_rate": 5.125993992349237e-06, + "loss": 0.2201, + "step": 23043 + }, + { + "epoch": 0.5072419106317412, + "grad_norm": 0.6474295405396201, + "learning_rate": 5.125637648839238e-06, + "loss": 0.4847, + "step": 23044 + }, + { + "epoch": 0.5072639225181598, + "grad_norm": 0.6382199815440299, + "learning_rate": 5.12528130469069e-06, + "loss": 0.4656, + "step": 23045 + }, + { + "epoch": 0.5072859344045785, + "grad_norm": 0.6094550551807235, + "learning_rate": 5.12492495990541e-06, + "loss": 0.4563, + "step": 23046 + }, + { + "epoch": 0.5073079462909972, + "grad_norm": 0.6254705532051521, + "learning_rate": 5.124568614485205e-06, + "loss": 0.5008, + "step": 23047 + }, + { + "epoch": 0.5073299581774158, + "grad_norm": 1.4393850661177394, + "learning_rate": 5.124212268431885e-06, + "loss": 0.3542, + "step": 23048 + }, + { + "epoch": 0.5073519700638345, + "grad_norm": 1.7110735515611388, + "learning_rate": 5.123855921747264e-06, + "loss": 0.2997, + "step": 23049 + }, + { + "epoch": 0.5073739819502532, + "grad_norm": 1.299707300739303, + "learning_rate": 5.123499574433152e-06, + "loss": 0.3824, + "step": 23050 + }, + { + "epoch": 0.5073959938366718, + "grad_norm": 1.6157608078688144, + "learning_rate": 5.123143226491358e-06, + "loss": 0.3231, + "step": 23051 + }, + { + "epoch": 0.5074180057230905, + "grad_norm": 2.87737284596524, + "learning_rate": 5.122786877923699e-06, + "loss": 0.3121, + "step": 23052 + }, + { + "epoch": 0.5074400176095092, + "grad_norm": 1.4780431918946848, + "learning_rate": 5.12243052873198e-06, + "loss": 0.3322, + "step": 23053 + }, + { + "epoch": 0.5074620294959278, + "grad_norm": 1.5451935183512786, + "learning_rate": 5.122074178918014e-06, + "loss": 0.3657, + "step": 23054 + }, + { + "epoch": 0.5074840413823465, + "grad_norm": 1.4392940923691149, + "learning_rate": 5.121717828483615e-06, + "loss": 0.3212, + "step": 23055 + }, + { + "epoch": 0.5075060532687651, + "grad_norm": 0.6460433765211158, + "learning_rate": 5.12136147743059e-06, + "loss": 0.4964, + "step": 23056 + }, + { + "epoch": 0.5075280651551838, + "grad_norm": 1.1270264967873782, + "learning_rate": 5.121005125760752e-06, + "loss": 0.2308, + "step": 23057 + }, + { + "epoch": 0.5075500770416025, + "grad_norm": 1.443709093403872, + "learning_rate": 5.12064877347591e-06, + "loss": 0.2954, + "step": 23058 + }, + { + "epoch": 0.5075720889280211, + "grad_norm": 1.3654348430025836, + "learning_rate": 5.120292420577879e-06, + "loss": 0.401, + "step": 23059 + }, + { + "epoch": 0.5075941008144398, + "grad_norm": 0.6225506718510034, + "learning_rate": 5.1199360670684694e-06, + "loss": 0.4848, + "step": 23060 + }, + { + "epoch": 0.5076161127008585, + "grad_norm": 0.6465712254076352, + "learning_rate": 5.119579712949489e-06, + "loss": 0.4899, + "step": 23061 + }, + { + "epoch": 0.5076381245872771, + "grad_norm": 1.6102559258719624, + "learning_rate": 5.119223358222754e-06, + "loss": 0.371, + "step": 23062 + }, + { + "epoch": 0.5076601364736958, + "grad_norm": 1.7179038446767867, + "learning_rate": 5.11886700289007e-06, + "loss": 0.3264, + "step": 23063 + }, + { + "epoch": 0.5076821483601145, + "grad_norm": 1.6672624007006775, + "learning_rate": 5.118510646953252e-06, + "loss": 0.3232, + "step": 23064 + }, + { + "epoch": 0.5077041602465331, + "grad_norm": 1.212493557046863, + "learning_rate": 5.118154290414111e-06, + "loss": 0.2264, + "step": 23065 + }, + { + "epoch": 0.5077261721329518, + "grad_norm": 1.5976746413521623, + "learning_rate": 5.117797933274455e-06, + "loss": 0.3351, + "step": 23066 + }, + { + "epoch": 0.5077481840193705, + "grad_norm": 1.7579911780334712, + "learning_rate": 5.117441575536098e-06, + "loss": 0.3677, + "step": 23067 + }, + { + "epoch": 0.5077701959057891, + "grad_norm": 1.3433219763834379, + "learning_rate": 5.117085217200852e-06, + "loss": 0.3081, + "step": 23068 + }, + { + "epoch": 0.5077922077922078, + "grad_norm": 1.4399986442749135, + "learning_rate": 5.1167288582705245e-06, + "loss": 0.3, + "step": 23069 + }, + { + "epoch": 0.5078142196786265, + "grad_norm": 0.6395341159510688, + "learning_rate": 5.116372498746932e-06, + "loss": 0.4902, + "step": 23070 + }, + { + "epoch": 0.5078362315650451, + "grad_norm": 2.341454243166033, + "learning_rate": 5.1160161386318795e-06, + "loss": 0.3029, + "step": 23071 + }, + { + "epoch": 0.5078582434514638, + "grad_norm": 0.6286709556912725, + "learning_rate": 5.115659777927182e-06, + "loss": 0.4932, + "step": 23072 + }, + { + "epoch": 0.5078802553378825, + "grad_norm": 1.017583807861174, + "learning_rate": 5.11530341663465e-06, + "loss": 0.2359, + "step": 23073 + }, + { + "epoch": 0.5079022672243011, + "grad_norm": 1.349081552330727, + "learning_rate": 5.114947054756096e-06, + "loss": 0.2697, + "step": 23074 + }, + { + "epoch": 0.5079242791107198, + "grad_norm": 1.2004680814555806, + "learning_rate": 5.114590692293329e-06, + "loss": 0.2617, + "step": 23075 + }, + { + "epoch": 0.5079462909971385, + "grad_norm": 0.635901929683602, + "learning_rate": 5.11423432924816e-06, + "loss": 0.5016, + "step": 23076 + }, + { + "epoch": 0.5079683028835571, + "grad_norm": 1.9107928645309002, + "learning_rate": 5.113877965622401e-06, + "loss": 0.3043, + "step": 23077 + }, + { + "epoch": 0.5079903147699758, + "grad_norm": 1.1968101301111147, + "learning_rate": 5.113521601417866e-06, + "loss": 0.2526, + "step": 23078 + }, + { + "epoch": 0.5080123266563945, + "grad_norm": 0.5944800383656416, + "learning_rate": 5.1131652366363616e-06, + "loss": 0.4391, + "step": 23079 + }, + { + "epoch": 0.5080343385428131, + "grad_norm": 1.2749058229627857, + "learning_rate": 5.1128088712797005e-06, + "loss": 0.2797, + "step": 23080 + }, + { + "epoch": 0.5080563504292318, + "grad_norm": 1.5646407314318371, + "learning_rate": 5.112452505349696e-06, + "loss": 0.2675, + "step": 23081 + }, + { + "epoch": 0.5080783623156504, + "grad_norm": 0.616725922511337, + "learning_rate": 5.112096138848156e-06, + "loss": 0.513, + "step": 23082 + }, + { + "epoch": 0.5081003742020691, + "grad_norm": 1.8625677103838265, + "learning_rate": 5.111739771776896e-06, + "loss": 0.2947, + "step": 23083 + }, + { + "epoch": 0.5081223860884878, + "grad_norm": 1.1441286227431102, + "learning_rate": 5.111383404137722e-06, + "loss": 0.3125, + "step": 23084 + }, + { + "epoch": 0.5081443979749064, + "grad_norm": 1.692199177966602, + "learning_rate": 5.111027035932448e-06, + "loss": 0.3229, + "step": 23085 + }, + { + "epoch": 0.5081664098613251, + "grad_norm": 1.5489733565397115, + "learning_rate": 5.110670667162887e-06, + "loss": 0.342, + "step": 23086 + }, + { + "epoch": 0.5081884217477438, + "grad_norm": 1.4273814076309026, + "learning_rate": 5.110314297830846e-06, + "loss": 0.2932, + "step": 23087 + }, + { + "epoch": 0.5082104336341624, + "grad_norm": 1.6137178570773898, + "learning_rate": 5.10995792793814e-06, + "loss": 0.4056, + "step": 23088 + }, + { + "epoch": 0.5082324455205811, + "grad_norm": 1.6042067541208413, + "learning_rate": 5.109601557486579e-06, + "loss": 0.4045, + "step": 23089 + }, + { + "epoch": 0.5082544574069998, + "grad_norm": 0.6218124861441114, + "learning_rate": 5.1092451864779734e-06, + "loss": 0.4667, + "step": 23090 + }, + { + "epoch": 0.5082764692934184, + "grad_norm": 1.4559149832853895, + "learning_rate": 5.1088888149141355e-06, + "loss": 0.3343, + "step": 23091 + }, + { + "epoch": 0.5082984811798371, + "grad_norm": 2.135283005064158, + "learning_rate": 5.108532442796875e-06, + "loss": 0.3848, + "step": 23092 + }, + { + "epoch": 0.5083204930662558, + "grad_norm": 2.0004647082393605, + "learning_rate": 5.108176070128006e-06, + "loss": 0.2553, + "step": 23093 + }, + { + "epoch": 0.5083425049526744, + "grad_norm": 1.3366889778210826, + "learning_rate": 5.107819696909336e-06, + "loss": 0.2534, + "step": 23094 + }, + { + "epoch": 0.5083645168390931, + "grad_norm": 1.5276163328571792, + "learning_rate": 5.107463323142679e-06, + "loss": 0.2927, + "step": 23095 + }, + { + "epoch": 0.5083865287255118, + "grad_norm": 1.4701101496297362, + "learning_rate": 5.107106948829847e-06, + "loss": 0.2865, + "step": 23096 + }, + { + "epoch": 0.5084085406119304, + "grad_norm": 1.3145242312823018, + "learning_rate": 5.106750573972647e-06, + "loss": 0.3619, + "step": 23097 + }, + { + "epoch": 0.5084305524983491, + "grad_norm": 1.8707016388768234, + "learning_rate": 5.106394198572895e-06, + "loss": 0.2991, + "step": 23098 + }, + { + "epoch": 0.5084525643847678, + "grad_norm": 1.4473895103160104, + "learning_rate": 5.106037822632398e-06, + "loss": 0.33, + "step": 23099 + }, + { + "epoch": 0.5084745762711864, + "grad_norm": 1.4386167955592473, + "learning_rate": 5.1056814461529715e-06, + "loss": 0.3416, + "step": 23100 + }, + { + "epoch": 0.5084965881576051, + "grad_norm": 1.4652530383058975, + "learning_rate": 5.105325069136424e-06, + "loss": 0.2892, + "step": 23101 + }, + { + "epoch": 0.5085186000440238, + "grad_norm": 1.2633573039930384, + "learning_rate": 5.104968691584569e-06, + "loss": 0.3006, + "step": 23102 + }, + { + "epoch": 0.5085406119304424, + "grad_norm": 1.315143571059554, + "learning_rate": 5.1046123134992144e-06, + "loss": 0.3098, + "step": 23103 + }, + { + "epoch": 0.5085626238168611, + "grad_norm": 2.3900301547775125, + "learning_rate": 5.104255934882174e-06, + "loss": 0.3229, + "step": 23104 + }, + { + "epoch": 0.5085846357032797, + "grad_norm": 1.4314468002612968, + "learning_rate": 5.103899555735257e-06, + "loss": 0.4214, + "step": 23105 + }, + { + "epoch": 0.5086066475896984, + "grad_norm": 0.6223733481035334, + "learning_rate": 5.10354317606028e-06, + "loss": 0.4717, + "step": 23106 + }, + { + "epoch": 0.5086286594761171, + "grad_norm": 1.2665118643064814, + "learning_rate": 5.1031867958590474e-06, + "loss": 0.2674, + "step": 23107 + }, + { + "epoch": 0.5086506713625357, + "grad_norm": 2.316849268111147, + "learning_rate": 5.102830415133374e-06, + "loss": 0.3226, + "step": 23108 + }, + { + "epoch": 0.5086726832489544, + "grad_norm": 1.3691207960778498, + "learning_rate": 5.102474033885071e-06, + "loss": 0.2881, + "step": 23109 + }, + { + "epoch": 0.5086946951353731, + "grad_norm": 1.3538999924300938, + "learning_rate": 5.102117652115948e-06, + "loss": 0.3256, + "step": 23110 + }, + { + "epoch": 0.5087167070217917, + "grad_norm": 2.0950958343789114, + "learning_rate": 5.10176126982782e-06, + "loss": 0.4098, + "step": 23111 + }, + { + "epoch": 0.5087387189082104, + "grad_norm": 2.0216806978348, + "learning_rate": 5.101404887022493e-06, + "loss": 0.3607, + "step": 23112 + }, + { + "epoch": 0.5087607307946291, + "grad_norm": 1.4492934310778955, + "learning_rate": 5.101048503701782e-06, + "loss": 0.3102, + "step": 23113 + }, + { + "epoch": 0.5087827426810477, + "grad_norm": 1.6550360838596316, + "learning_rate": 5.100692119867499e-06, + "loss": 0.3265, + "step": 23114 + }, + { + "epoch": 0.5088047545674664, + "grad_norm": 1.3776762925841965, + "learning_rate": 5.100335735521452e-06, + "loss": 0.3716, + "step": 23115 + }, + { + "epoch": 0.5088267664538851, + "grad_norm": 1.2829096266697815, + "learning_rate": 5.099979350665455e-06, + "loss": 0.2657, + "step": 23116 + }, + { + "epoch": 0.5088487783403037, + "grad_norm": 1.3784388574928093, + "learning_rate": 5.099622965301317e-06, + "loss": 0.3605, + "step": 23117 + }, + { + "epoch": 0.5088707902267224, + "grad_norm": 1.6294799079738502, + "learning_rate": 5.099266579430851e-06, + "loss": 0.3254, + "step": 23118 + }, + { + "epoch": 0.5088928021131411, + "grad_norm": 0.632766318390272, + "learning_rate": 5.098910193055869e-06, + "loss": 0.4667, + "step": 23119 + }, + { + "epoch": 0.5089148139995597, + "grad_norm": 1.3959820550753348, + "learning_rate": 5.098553806178181e-06, + "loss": 0.2701, + "step": 23120 + }, + { + "epoch": 0.5089368258859784, + "grad_norm": 1.4564671907029898, + "learning_rate": 5.0981974187995994e-06, + "loss": 0.3554, + "step": 23121 + }, + { + "epoch": 0.5089588377723971, + "grad_norm": 1.427305146494502, + "learning_rate": 5.097841030921934e-06, + "loss": 0.3023, + "step": 23122 + }, + { + "epoch": 0.5089808496588157, + "grad_norm": 0.6035068629480068, + "learning_rate": 5.097484642546996e-06, + "loss": 0.4652, + "step": 23123 + }, + { + "epoch": 0.5090028615452344, + "grad_norm": 3.003924535973509, + "learning_rate": 5.097128253676599e-06, + "loss": 0.3511, + "step": 23124 + }, + { + "epoch": 0.5090248734316531, + "grad_norm": 1.80625069288457, + "learning_rate": 5.0967718643125506e-06, + "loss": 0.3718, + "step": 23125 + }, + { + "epoch": 0.5090468853180717, + "grad_norm": 4.20941367722239, + "learning_rate": 5.096415474456666e-06, + "loss": 0.4021, + "step": 23126 + }, + { + "epoch": 0.5090688972044904, + "grad_norm": 1.444995590136031, + "learning_rate": 5.096059084110756e-06, + "loss": 0.3407, + "step": 23127 + }, + { + "epoch": 0.509090909090909, + "grad_norm": 1.5458904724810605, + "learning_rate": 5.095702693276629e-06, + "loss": 0.3361, + "step": 23128 + }, + { + "epoch": 0.5091129209773277, + "grad_norm": 1.466745730912858, + "learning_rate": 5.0953463019561e-06, + "loss": 0.2533, + "step": 23129 + }, + { + "epoch": 0.5091349328637464, + "grad_norm": 0.6118048537835236, + "learning_rate": 5.094989910150977e-06, + "loss": 0.4627, + "step": 23130 + }, + { + "epoch": 0.509156944750165, + "grad_norm": 1.318086408470246, + "learning_rate": 5.094633517863074e-06, + "loss": 0.392, + "step": 23131 + }, + { + "epoch": 0.5091789566365837, + "grad_norm": 1.68281152647555, + "learning_rate": 5.094277125094201e-06, + "loss": 0.3712, + "step": 23132 + }, + { + "epoch": 0.5092009685230025, + "grad_norm": 0.5940788695612849, + "learning_rate": 5.093920731846169e-06, + "loss": 0.4733, + "step": 23133 + }, + { + "epoch": 0.509222980409421, + "grad_norm": 1.532999835583918, + "learning_rate": 5.09356433812079e-06, + "loss": 0.3296, + "step": 23134 + }, + { + "epoch": 0.5092449922958397, + "grad_norm": 1.4395405024439178, + "learning_rate": 5.093207943919875e-06, + "loss": 0.3415, + "step": 23135 + }, + { + "epoch": 0.5092670041822585, + "grad_norm": 1.6453963631662238, + "learning_rate": 5.092851549245237e-06, + "loss": 0.3464, + "step": 23136 + }, + { + "epoch": 0.509289016068677, + "grad_norm": 1.3595375597662145, + "learning_rate": 5.092495154098684e-06, + "loss": 0.2663, + "step": 23137 + }, + { + "epoch": 0.5093110279550958, + "grad_norm": 1.3757157424504876, + "learning_rate": 5.092138758482031e-06, + "loss": 0.2571, + "step": 23138 + }, + { + "epoch": 0.5093330398415145, + "grad_norm": 1.3811703060686769, + "learning_rate": 5.0917823623970885e-06, + "loss": 0.3068, + "step": 23139 + }, + { + "epoch": 0.509355051727933, + "grad_norm": 1.2667475017777352, + "learning_rate": 5.091425965845665e-06, + "loss": 0.2722, + "step": 23140 + }, + { + "epoch": 0.5093770636143518, + "grad_norm": 1.5260497629247114, + "learning_rate": 5.091069568829576e-06, + "loss": 0.313, + "step": 23141 + }, + { + "epoch": 0.5093990755007705, + "grad_norm": 2.3006776093836363, + "learning_rate": 5.09071317135063e-06, + "loss": 0.3266, + "step": 23142 + }, + { + "epoch": 0.509421087387189, + "grad_norm": 1.5590224293051123, + "learning_rate": 5.0903567734106376e-06, + "loss": 0.2799, + "step": 23143 + }, + { + "epoch": 0.5094430992736078, + "grad_norm": 1.9637845802903755, + "learning_rate": 5.0900003750114135e-06, + "loss": 0.2979, + "step": 23144 + }, + { + "epoch": 0.5094651111600265, + "grad_norm": 1.3650019960073458, + "learning_rate": 5.089643976154768e-06, + "loss": 0.2633, + "step": 23145 + }, + { + "epoch": 0.5094871230464451, + "grad_norm": 1.4858718996897298, + "learning_rate": 5.0892875768425085e-06, + "loss": 0.3142, + "step": 23146 + }, + { + "epoch": 0.5095091349328638, + "grad_norm": 1.315010829512601, + "learning_rate": 5.088931177076454e-06, + "loss": 0.2861, + "step": 23147 + }, + { + "epoch": 0.5095311468192825, + "grad_norm": 1.4274039041105595, + "learning_rate": 5.088574776858409e-06, + "loss": 0.3537, + "step": 23148 + }, + { + "epoch": 0.5095531587057011, + "grad_norm": 1.7368798831401682, + "learning_rate": 5.088218376190188e-06, + "loss": 0.3434, + "step": 23149 + }, + { + "epoch": 0.5095751705921198, + "grad_norm": 1.237396228178786, + "learning_rate": 5.087861975073603e-06, + "loss": 0.2664, + "step": 23150 + }, + { + "epoch": 0.5095971824785384, + "grad_norm": 2.618404160445216, + "learning_rate": 5.087505573510462e-06, + "loss": 0.3235, + "step": 23151 + }, + { + "epoch": 0.5096191943649571, + "grad_norm": 1.8260587002662017, + "learning_rate": 5.087149171502581e-06, + "loss": 0.312, + "step": 23152 + }, + { + "epoch": 0.5096412062513758, + "grad_norm": 1.6439278511419924, + "learning_rate": 5.086792769051767e-06, + "loss": 0.3132, + "step": 23153 + }, + { + "epoch": 0.5096632181377944, + "grad_norm": 1.6149634477111663, + "learning_rate": 5.0864363661598335e-06, + "loss": 0.2312, + "step": 23154 + }, + { + "epoch": 0.5096852300242131, + "grad_norm": 1.347000480982436, + "learning_rate": 5.086079962828594e-06, + "loss": 0.2937, + "step": 23155 + }, + { + "epoch": 0.5097072419106318, + "grad_norm": 1.5834815573555767, + "learning_rate": 5.085723559059856e-06, + "loss": 0.29, + "step": 23156 + }, + { + "epoch": 0.5097292537970504, + "grad_norm": 1.7813925099072152, + "learning_rate": 5.085367154855434e-06, + "loss": 0.4522, + "step": 23157 + }, + { + "epoch": 0.5097512656834691, + "grad_norm": 1.6413934598771742, + "learning_rate": 5.085010750217137e-06, + "loss": 0.264, + "step": 23158 + }, + { + "epoch": 0.5097732775698878, + "grad_norm": 0.6342404822368056, + "learning_rate": 5.084654345146777e-06, + "loss": 0.4807, + "step": 23159 + }, + { + "epoch": 0.5097952894563064, + "grad_norm": 1.524228435569064, + "learning_rate": 5.084297939646166e-06, + "loss": 0.3495, + "step": 23160 + }, + { + "epoch": 0.5098173013427251, + "grad_norm": 0.6445947612471443, + "learning_rate": 5.083941533717116e-06, + "loss": 0.4692, + "step": 23161 + }, + { + "epoch": 0.5098393132291438, + "grad_norm": 1.649733250473463, + "learning_rate": 5.083585127361438e-06, + "loss": 0.3165, + "step": 23162 + }, + { + "epoch": 0.5098613251155624, + "grad_norm": 0.6315829922240541, + "learning_rate": 5.083228720580942e-06, + "loss": 0.474, + "step": 23163 + }, + { + "epoch": 0.5098833370019811, + "grad_norm": 1.740096192774199, + "learning_rate": 5.08287231337744e-06, + "loss": 0.2342, + "step": 23164 + }, + { + "epoch": 0.5099053488883998, + "grad_norm": 2.102494989650764, + "learning_rate": 5.082515905752746e-06, + "loss": 0.383, + "step": 23165 + }, + { + "epoch": 0.5099273607748184, + "grad_norm": 1.6996639983524502, + "learning_rate": 5.082159497708668e-06, + "loss": 0.312, + "step": 23166 + }, + { + "epoch": 0.5099493726612371, + "grad_norm": 0.6237453700865059, + "learning_rate": 5.081803089247019e-06, + "loss": 0.5011, + "step": 23167 + }, + { + "epoch": 0.5099713845476558, + "grad_norm": 1.7430984927974345, + "learning_rate": 5.0814466803696105e-06, + "loss": 0.2138, + "step": 23168 + }, + { + "epoch": 0.5099933964340744, + "grad_norm": 1.345059028733982, + "learning_rate": 5.081090271078253e-06, + "loss": 0.2917, + "step": 23169 + }, + { + "epoch": 0.5100154083204931, + "grad_norm": 1.4752212912303697, + "learning_rate": 5.080733861374761e-06, + "loss": 0.2814, + "step": 23170 + }, + { + "epoch": 0.5100374202069118, + "grad_norm": 1.595678563914866, + "learning_rate": 5.08037745126094e-06, + "loss": 0.2647, + "step": 23171 + }, + { + "epoch": 0.5100594320933304, + "grad_norm": 0.6479379282220729, + "learning_rate": 5.0800210407386065e-06, + "loss": 0.4872, + "step": 23172 + }, + { + "epoch": 0.5100814439797491, + "grad_norm": 1.8471649579798801, + "learning_rate": 5.0796646298095705e-06, + "loss": 0.2851, + "step": 23173 + }, + { + "epoch": 0.5101034558661677, + "grad_norm": 1.4509590144825446, + "learning_rate": 5.0793082184756435e-06, + "loss": 0.2716, + "step": 23174 + }, + { + "epoch": 0.5101254677525864, + "grad_norm": 1.273079879012474, + "learning_rate": 5.078951806738637e-06, + "loss": 0.2348, + "step": 23175 + }, + { + "epoch": 0.5101474796390051, + "grad_norm": 0.6331791839127838, + "learning_rate": 5.078595394600361e-06, + "loss": 0.4976, + "step": 23176 + }, + { + "epoch": 0.5101694915254237, + "grad_norm": 1.6874187778938468, + "learning_rate": 5.078238982062628e-06, + "loss": 0.2849, + "step": 23177 + }, + { + "epoch": 0.5101915034118424, + "grad_norm": 1.888344569151552, + "learning_rate": 5.0778825691272505e-06, + "loss": 0.281, + "step": 23178 + }, + { + "epoch": 0.5102135152982611, + "grad_norm": 0.6221284195973035, + "learning_rate": 5.07752615579604e-06, + "loss": 0.4644, + "step": 23179 + }, + { + "epoch": 0.5102355271846797, + "grad_norm": 1.5399715286282958, + "learning_rate": 5.077169742070805e-06, + "loss": 0.333, + "step": 23180 + }, + { + "epoch": 0.5102575390710984, + "grad_norm": 1.6916983006275665, + "learning_rate": 5.07681332795336e-06, + "loss": 0.3146, + "step": 23181 + }, + { + "epoch": 0.5102795509575171, + "grad_norm": 2.7811692371804235, + "learning_rate": 5.076456913445515e-06, + "loss": 0.364, + "step": 23182 + }, + { + "epoch": 0.5103015628439357, + "grad_norm": 1.3101364975772762, + "learning_rate": 5.0761004985490835e-06, + "loss": 0.2297, + "step": 23183 + }, + { + "epoch": 0.5103235747303544, + "grad_norm": 2.016353106935193, + "learning_rate": 5.075744083265873e-06, + "loss": 0.3267, + "step": 23184 + }, + { + "epoch": 0.5103455866167731, + "grad_norm": 1.4413518286707343, + "learning_rate": 5.075387667597699e-06, + "loss": 0.299, + "step": 23185 + }, + { + "epoch": 0.5103675985031917, + "grad_norm": 1.4793753248653103, + "learning_rate": 5.075031251546371e-06, + "loss": 0.2372, + "step": 23186 + }, + { + "epoch": 0.5103896103896104, + "grad_norm": 1.6982642078439574, + "learning_rate": 5.074674835113699e-06, + "loss": 0.4074, + "step": 23187 + }, + { + "epoch": 0.5104116222760291, + "grad_norm": 1.43995640359744, + "learning_rate": 5.074318418301498e-06, + "loss": 0.3805, + "step": 23188 + }, + { + "epoch": 0.5104336341624477, + "grad_norm": 0.608627216062382, + "learning_rate": 5.073962001111577e-06, + "loss": 0.4485, + "step": 23189 + }, + { + "epoch": 0.5104556460488664, + "grad_norm": 1.5519565176103054, + "learning_rate": 5.073605583545748e-06, + "loss": 0.3684, + "step": 23190 + }, + { + "epoch": 0.5104776579352851, + "grad_norm": 2.73141647405689, + "learning_rate": 5.073249165605823e-06, + "loss": 0.3311, + "step": 23191 + }, + { + "epoch": 0.5104996698217037, + "grad_norm": 1.5858515268747508, + "learning_rate": 5.072892747293613e-06, + "loss": 0.2619, + "step": 23192 + }, + { + "epoch": 0.5105216817081224, + "grad_norm": 1.7460434284857511, + "learning_rate": 5.07253632861093e-06, + "loss": 0.3003, + "step": 23193 + }, + { + "epoch": 0.5105436935945411, + "grad_norm": 1.439624918557623, + "learning_rate": 5.072179909559584e-06, + "loss": 0.3434, + "step": 23194 + }, + { + "epoch": 0.5105657054809597, + "grad_norm": 0.5995252519975665, + "learning_rate": 5.071823490141387e-06, + "loss": 0.4468, + "step": 23195 + }, + { + "epoch": 0.5105877173673784, + "grad_norm": 1.4225439811343368, + "learning_rate": 5.0714670703581525e-06, + "loss": 0.2727, + "step": 23196 + }, + { + "epoch": 0.510609729253797, + "grad_norm": 2.1335520521199203, + "learning_rate": 5.07111065021169e-06, + "loss": 0.3313, + "step": 23197 + }, + { + "epoch": 0.5106317411402157, + "grad_norm": 1.348296688309, + "learning_rate": 5.070754229703811e-06, + "loss": 0.3162, + "step": 23198 + }, + { + "epoch": 0.5106537530266344, + "grad_norm": 1.3431729883877457, + "learning_rate": 5.0703978088363284e-06, + "loss": 0.3197, + "step": 23199 + }, + { + "epoch": 0.510675764913053, + "grad_norm": 1.870957737519371, + "learning_rate": 5.070041387611051e-06, + "loss": 0.3415, + "step": 23200 + }, + { + "epoch": 0.5106977767994717, + "grad_norm": 1.6851772619035943, + "learning_rate": 5.069684966029793e-06, + "loss": 0.3721, + "step": 23201 + }, + { + "epoch": 0.5107197886858904, + "grad_norm": 1.521293241937979, + "learning_rate": 5.069328544094366e-06, + "loss": 0.3047, + "step": 23202 + }, + { + "epoch": 0.510741800572309, + "grad_norm": 0.6512962436640216, + "learning_rate": 5.068972121806579e-06, + "loss": 0.5004, + "step": 23203 + }, + { + "epoch": 0.5107638124587277, + "grad_norm": 1.741588443040777, + "learning_rate": 5.0686156991682465e-06, + "loss": 0.2975, + "step": 23204 + }, + { + "epoch": 0.5107858243451464, + "grad_norm": 3.4362433488162023, + "learning_rate": 5.0682592761811764e-06, + "loss": 0.3358, + "step": 23205 + }, + { + "epoch": 0.510807836231565, + "grad_norm": 1.6626388967744334, + "learning_rate": 5.067902852847184e-06, + "loss": 0.2945, + "step": 23206 + }, + { + "epoch": 0.5108298481179837, + "grad_norm": 1.673187647730831, + "learning_rate": 5.0675464291680775e-06, + "loss": 0.3045, + "step": 23207 + }, + { + "epoch": 0.5108518600044024, + "grad_norm": 2.1735083763330887, + "learning_rate": 5.0671900051456715e-06, + "loss": 0.2953, + "step": 23208 + }, + { + "epoch": 0.510873871890821, + "grad_norm": 3.1759647720720166, + "learning_rate": 5.066833580781775e-06, + "loss": 0.3099, + "step": 23209 + }, + { + "epoch": 0.5108958837772397, + "grad_norm": 3.0424785068951117, + "learning_rate": 5.066477156078199e-06, + "loss": 0.2993, + "step": 23210 + }, + { + "epoch": 0.5109178956636584, + "grad_norm": 0.6252568028658156, + "learning_rate": 5.066120731036759e-06, + "loss": 0.5007, + "step": 23211 + }, + { + "epoch": 0.510939907550077, + "grad_norm": 1.5554012544772038, + "learning_rate": 5.065764305659263e-06, + "loss": 0.3386, + "step": 23212 + }, + { + "epoch": 0.5109619194364957, + "grad_norm": 3.683094811976212, + "learning_rate": 5.0654078799475225e-06, + "loss": 0.2992, + "step": 23213 + }, + { + "epoch": 0.5109839313229144, + "grad_norm": 2.5022873764935096, + "learning_rate": 5.065051453903351e-06, + "loss": 0.3162, + "step": 23214 + }, + { + "epoch": 0.511005943209333, + "grad_norm": 3.410724522882019, + "learning_rate": 5.0646950275285586e-06, + "loss": 0.3093, + "step": 23215 + }, + { + "epoch": 0.5110279550957517, + "grad_norm": 1.1736397948541937, + "learning_rate": 5.064338600824959e-06, + "loss": 0.1996, + "step": 23216 + }, + { + "epoch": 0.5110499669821704, + "grad_norm": 1.7345412842191525, + "learning_rate": 5.06398217379436e-06, + "loss": 0.3625, + "step": 23217 + }, + { + "epoch": 0.511071978868589, + "grad_norm": 1.5710697871466126, + "learning_rate": 5.063625746438574e-06, + "loss": 0.2701, + "step": 23218 + }, + { + "epoch": 0.5110939907550077, + "grad_norm": 1.4992209866612085, + "learning_rate": 5.063269318759415e-06, + "loss": 0.2303, + "step": 23219 + }, + { + "epoch": 0.5111160026414264, + "grad_norm": 3.5297615904433717, + "learning_rate": 5.0629128907586936e-06, + "loss": 0.3064, + "step": 23220 + }, + { + "epoch": 0.511138014527845, + "grad_norm": 1.3089747824047986, + "learning_rate": 5.06255646243822e-06, + "loss": 0.2873, + "step": 23221 + }, + { + "epoch": 0.5111600264142637, + "grad_norm": 1.0706284281247076, + "learning_rate": 5.0622000337998075e-06, + "loss": 0.2527, + "step": 23222 + }, + { + "epoch": 0.5111820383006823, + "grad_norm": 1.3897208439773483, + "learning_rate": 5.061843604845266e-06, + "loss": 0.3813, + "step": 23223 + }, + { + "epoch": 0.511204050187101, + "grad_norm": 1.5225630366164922, + "learning_rate": 5.061487175576408e-06, + "loss": 0.2754, + "step": 23224 + }, + { + "epoch": 0.5112260620735197, + "grad_norm": 1.8242839769148733, + "learning_rate": 5.061130745995046e-06, + "loss": 0.2874, + "step": 23225 + }, + { + "epoch": 0.5112480739599383, + "grad_norm": 1.3145253840850015, + "learning_rate": 5.060774316102987e-06, + "loss": 0.2845, + "step": 23226 + }, + { + "epoch": 0.511270085846357, + "grad_norm": 1.713194713515254, + "learning_rate": 5.060417885902048e-06, + "loss": 0.342, + "step": 23227 + }, + { + "epoch": 0.5112920977327757, + "grad_norm": 0.6446330600257528, + "learning_rate": 5.060061455394036e-06, + "loss": 0.5003, + "step": 23228 + }, + { + "epoch": 0.5113141096191943, + "grad_norm": 1.86701681837185, + "learning_rate": 5.059705024580769e-06, + "loss": 0.3654, + "step": 23229 + }, + { + "epoch": 0.511336121505613, + "grad_norm": 1.369746360999691, + "learning_rate": 5.059348593464053e-06, + "loss": 0.2902, + "step": 23230 + }, + { + "epoch": 0.5113581333920317, + "grad_norm": 1.369311067756662, + "learning_rate": 5.0589921620456995e-06, + "loss": 0.2887, + "step": 23231 + }, + { + "epoch": 0.5113801452784503, + "grad_norm": 1.516242555379846, + "learning_rate": 5.058635730327522e-06, + "loss": 0.2399, + "step": 23232 + }, + { + "epoch": 0.511402157164869, + "grad_norm": 1.498877348413027, + "learning_rate": 5.058279298311332e-06, + "loss": 0.2375, + "step": 23233 + }, + { + "epoch": 0.5114241690512877, + "grad_norm": 1.1170004722084572, + "learning_rate": 5.057922865998941e-06, + "loss": 0.2384, + "step": 23234 + }, + { + "epoch": 0.5114461809377063, + "grad_norm": 2.1089644163664762, + "learning_rate": 5.0575664333921595e-06, + "loss": 0.3791, + "step": 23235 + }, + { + "epoch": 0.511468192824125, + "grad_norm": 1.2805562735129188, + "learning_rate": 5.057210000492799e-06, + "loss": 0.2802, + "step": 23236 + }, + { + "epoch": 0.5114902047105437, + "grad_norm": 1.6019873759362404, + "learning_rate": 5.056853567302673e-06, + "loss": 0.2781, + "step": 23237 + }, + { + "epoch": 0.5115122165969623, + "grad_norm": 2.070349877604465, + "learning_rate": 5.056497133823592e-06, + "loss": 0.2439, + "step": 23238 + }, + { + "epoch": 0.511534228483381, + "grad_norm": 2.079188728796833, + "learning_rate": 5.056140700057366e-06, + "loss": 0.3108, + "step": 23239 + }, + { + "epoch": 0.5115562403697997, + "grad_norm": 1.5447322488121222, + "learning_rate": 5.05578426600581e-06, + "loss": 0.3086, + "step": 23240 + }, + { + "epoch": 0.5115782522562183, + "grad_norm": 1.740050782060246, + "learning_rate": 5.05542783167073e-06, + "loss": 0.3398, + "step": 23241 + }, + { + "epoch": 0.511600264142637, + "grad_norm": 0.672923926676718, + "learning_rate": 5.055071397053945e-06, + "loss": 0.4792, + "step": 23242 + }, + { + "epoch": 0.5116222760290557, + "grad_norm": 1.674111158183214, + "learning_rate": 5.054714962157261e-06, + "loss": 0.3639, + "step": 23243 + }, + { + "epoch": 0.5116442879154743, + "grad_norm": 1.4779611941888082, + "learning_rate": 5.05435852698249e-06, + "loss": 0.377, + "step": 23244 + }, + { + "epoch": 0.511666299801893, + "grad_norm": 1.3198321738097805, + "learning_rate": 5.054002091531446e-06, + "loss": 0.2728, + "step": 23245 + }, + { + "epoch": 0.5116883116883116, + "grad_norm": 1.2654887294870374, + "learning_rate": 5.053645655805939e-06, + "loss": 0.3053, + "step": 23246 + }, + { + "epoch": 0.5117103235747303, + "grad_norm": 0.6026648564315633, + "learning_rate": 5.053289219807781e-06, + "loss": 0.4686, + "step": 23247 + }, + { + "epoch": 0.511732335461149, + "grad_norm": 0.6275598297306363, + "learning_rate": 5.0529327835387845e-06, + "loss": 0.4854, + "step": 23248 + }, + { + "epoch": 0.5117543473475676, + "grad_norm": 8.402550886924178, + "learning_rate": 5.052576347000757e-06, + "loss": 0.3104, + "step": 23249 + }, + { + "epoch": 0.5117763592339863, + "grad_norm": 1.3123893639801176, + "learning_rate": 5.052219910195515e-06, + "loss": 0.2683, + "step": 23250 + }, + { + "epoch": 0.511798371120405, + "grad_norm": 2.1166012783342194, + "learning_rate": 5.0518634731248686e-06, + "loss": 0.325, + "step": 23251 + }, + { + "epoch": 0.5118203830068236, + "grad_norm": 1.3798299220426231, + "learning_rate": 5.051507035790628e-06, + "loss": 0.3065, + "step": 23252 + }, + { + "epoch": 0.5118423948932423, + "grad_norm": 1.9650548233055136, + "learning_rate": 5.051150598194606e-06, + "loss": 0.3444, + "step": 23253 + }, + { + "epoch": 0.511864406779661, + "grad_norm": 1.5345212987230172, + "learning_rate": 5.050794160338613e-06, + "loss": 0.2976, + "step": 23254 + }, + { + "epoch": 0.5118864186660796, + "grad_norm": 1.8774461827869855, + "learning_rate": 5.050437722224463e-06, + "loss": 0.4458, + "step": 23255 + }, + { + "epoch": 0.5119084305524984, + "grad_norm": 1.631638128294467, + "learning_rate": 5.050081283853965e-06, + "loss": 0.3087, + "step": 23256 + }, + { + "epoch": 0.511930442438917, + "grad_norm": 1.9939746575887252, + "learning_rate": 5.049724845228931e-06, + "loss": 0.305, + "step": 23257 + }, + { + "epoch": 0.5119524543253356, + "grad_norm": 2.1852966483195946, + "learning_rate": 5.049368406351175e-06, + "loss": 0.2056, + "step": 23258 + }, + { + "epoch": 0.5119744662117544, + "grad_norm": 0.99815542928468, + "learning_rate": 5.049011967222505e-06, + "loss": 0.2699, + "step": 23259 + }, + { + "epoch": 0.5119964780981731, + "grad_norm": 1.1391060510491569, + "learning_rate": 5.048655527844735e-06, + "loss": 0.3523, + "step": 23260 + }, + { + "epoch": 0.5120184899845917, + "grad_norm": 4.473434097899968, + "learning_rate": 5.048299088219677e-06, + "loss": 0.3251, + "step": 23261 + }, + { + "epoch": 0.5120405018710104, + "grad_norm": 3.4869555445810914, + "learning_rate": 5.04794264834914e-06, + "loss": 0.3422, + "step": 23262 + }, + { + "epoch": 0.5120625137574291, + "grad_norm": 1.4915211644434088, + "learning_rate": 5.047586208234937e-06, + "loss": 0.345, + "step": 23263 + }, + { + "epoch": 0.5120845256438477, + "grad_norm": 1.1727210324027804, + "learning_rate": 5.0472297678788795e-06, + "loss": 0.2488, + "step": 23264 + }, + { + "epoch": 0.5121065375302664, + "grad_norm": 0.6789601912073121, + "learning_rate": 5.0468733272827805e-06, + "loss": 0.4977, + "step": 23265 + }, + { + "epoch": 0.5121285494166851, + "grad_norm": 1.3673433675673377, + "learning_rate": 5.04651688644845e-06, + "loss": 0.3325, + "step": 23266 + }, + { + "epoch": 0.5121505613031037, + "grad_norm": 3.32461913563483, + "learning_rate": 5.046160445377699e-06, + "loss": 0.3709, + "step": 23267 + }, + { + "epoch": 0.5121725731895224, + "grad_norm": 1.2362701556259634, + "learning_rate": 5.0458040040723414e-06, + "loss": 0.2395, + "step": 23268 + }, + { + "epoch": 0.512194585075941, + "grad_norm": 1.6452814610857687, + "learning_rate": 5.045447562534187e-06, + "loss": 0.3306, + "step": 23269 + }, + { + "epoch": 0.5122165969623597, + "grad_norm": 1.3005701935609366, + "learning_rate": 5.045091120765048e-06, + "loss": 0.2535, + "step": 23270 + }, + { + "epoch": 0.5122386088487784, + "grad_norm": 1.6589429691556978, + "learning_rate": 5.044734678766735e-06, + "loss": 0.3173, + "step": 23271 + }, + { + "epoch": 0.512260620735197, + "grad_norm": 1.6824852246592794, + "learning_rate": 5.04437823654106e-06, + "loss": 0.2553, + "step": 23272 + }, + { + "epoch": 0.5122826326216157, + "grad_norm": 0.6173592703885382, + "learning_rate": 5.044021794089836e-06, + "loss": 0.4727, + "step": 23273 + }, + { + "epoch": 0.5123046445080344, + "grad_norm": 1.3747902670257581, + "learning_rate": 5.043665351414874e-06, + "loss": 0.2288, + "step": 23274 + }, + { + "epoch": 0.512326656394453, + "grad_norm": 1.4385175669164234, + "learning_rate": 5.0433089085179846e-06, + "loss": 0.3004, + "step": 23275 + }, + { + "epoch": 0.5123486682808717, + "grad_norm": 1.4012976533262373, + "learning_rate": 5.04295246540098e-06, + "loss": 0.261, + "step": 23276 + }, + { + "epoch": 0.5123706801672904, + "grad_norm": 1.6370793813382558, + "learning_rate": 5.042596022065671e-06, + "loss": 0.3281, + "step": 23277 + }, + { + "epoch": 0.512392692053709, + "grad_norm": 1.455653692061104, + "learning_rate": 5.042239578513871e-06, + "loss": 0.3419, + "step": 23278 + }, + { + "epoch": 0.5124147039401277, + "grad_norm": 2.291942717245963, + "learning_rate": 5.041883134747391e-06, + "loss": 0.2619, + "step": 23279 + }, + { + "epoch": 0.5124367158265464, + "grad_norm": 1.553839023053685, + "learning_rate": 5.041526690768041e-06, + "loss": 0.2806, + "step": 23280 + }, + { + "epoch": 0.512458727712965, + "grad_norm": 4.277037527593942, + "learning_rate": 5.041170246577635e-06, + "loss": 0.2961, + "step": 23281 + }, + { + "epoch": 0.5124807395993837, + "grad_norm": 0.6181097702159292, + "learning_rate": 5.040813802177982e-06, + "loss": 0.47, + "step": 23282 + }, + { + "epoch": 0.5125027514858024, + "grad_norm": 1.865456086409836, + "learning_rate": 5.040457357570897e-06, + "loss": 0.3092, + "step": 23283 + }, + { + "epoch": 0.512524763372221, + "grad_norm": 2.6322163769912437, + "learning_rate": 5.040100912758189e-06, + "loss": 0.377, + "step": 23284 + }, + { + "epoch": 0.5125467752586397, + "grad_norm": 1.4419177831364907, + "learning_rate": 5.0397444677416695e-06, + "loss": 0.3384, + "step": 23285 + }, + { + "epoch": 0.5125687871450584, + "grad_norm": 1.7764255824207857, + "learning_rate": 5.039388022523151e-06, + "loss": 0.3961, + "step": 23286 + }, + { + "epoch": 0.512590799031477, + "grad_norm": 1.3011586805854367, + "learning_rate": 5.039031577104445e-06, + "loss": 0.3373, + "step": 23287 + }, + { + "epoch": 0.5126128109178957, + "grad_norm": 2.0491166620588497, + "learning_rate": 5.0386751314873635e-06, + "loss": 0.2389, + "step": 23288 + }, + { + "epoch": 0.5126348228043144, + "grad_norm": 1.6214205413141995, + "learning_rate": 5.038318685673719e-06, + "loss": 0.3475, + "step": 23289 + }, + { + "epoch": 0.512656834690733, + "grad_norm": 1.4510990654353864, + "learning_rate": 5.03796223966532e-06, + "loss": 0.3121, + "step": 23290 + }, + { + "epoch": 0.5126788465771517, + "grad_norm": 1.4151042374899252, + "learning_rate": 5.03760579346398e-06, + "loss": 0.3583, + "step": 23291 + }, + { + "epoch": 0.5127008584635703, + "grad_norm": 1.50919952712181, + "learning_rate": 5.037249347071511e-06, + "loss": 0.2772, + "step": 23292 + }, + { + "epoch": 0.512722870349989, + "grad_norm": 1.683774599747363, + "learning_rate": 5.036892900489725e-06, + "loss": 0.2916, + "step": 23293 + }, + { + "epoch": 0.5127448822364077, + "grad_norm": 0.6166225709442377, + "learning_rate": 5.036536453720432e-06, + "loss": 0.4844, + "step": 23294 + }, + { + "epoch": 0.5127668941228263, + "grad_norm": 1.8429249939559187, + "learning_rate": 5.036180006765443e-06, + "loss": 0.3233, + "step": 23295 + }, + { + "epoch": 0.512788906009245, + "grad_norm": 1.8297336006252156, + "learning_rate": 5.035823559626574e-06, + "loss": 0.4107, + "step": 23296 + }, + { + "epoch": 0.5128109178956637, + "grad_norm": 1.6689262523967499, + "learning_rate": 5.035467112305633e-06, + "loss": 0.3305, + "step": 23297 + }, + { + "epoch": 0.5128329297820823, + "grad_norm": 1.4837644513632977, + "learning_rate": 5.0351106648044314e-06, + "loss": 0.3183, + "step": 23298 + }, + { + "epoch": 0.512854941668501, + "grad_norm": 1.3938111947111278, + "learning_rate": 5.034754217124782e-06, + "loss": 0.2683, + "step": 23299 + }, + { + "epoch": 0.5128769535549197, + "grad_norm": 1.458269485462961, + "learning_rate": 5.034397769268496e-06, + "loss": 0.3233, + "step": 23300 + }, + { + "epoch": 0.5128989654413383, + "grad_norm": 1.571919074691828, + "learning_rate": 5.034041321237384e-06, + "loss": 0.3191, + "step": 23301 + }, + { + "epoch": 0.512920977327757, + "grad_norm": 1.2639100328177102, + "learning_rate": 5.033684873033262e-06, + "loss": 0.2643, + "step": 23302 + }, + { + "epoch": 0.5129429892141757, + "grad_norm": 0.6331941465759506, + "learning_rate": 5.033328424657935e-06, + "loss": 0.4801, + "step": 23303 + }, + { + "epoch": 0.5129650011005943, + "grad_norm": 1.6805749711895206, + "learning_rate": 5.03297197611322e-06, + "loss": 0.2238, + "step": 23304 + }, + { + "epoch": 0.512987012987013, + "grad_norm": 1.478949935292354, + "learning_rate": 5.032615527400927e-06, + "loss": 0.2965, + "step": 23305 + }, + { + "epoch": 0.5130090248734317, + "grad_norm": 1.4198548714231818, + "learning_rate": 5.032259078522867e-06, + "loss": 0.2677, + "step": 23306 + }, + { + "epoch": 0.5130310367598503, + "grad_norm": 1.7578067384227933, + "learning_rate": 5.0319026294808515e-06, + "loss": 0.3279, + "step": 23307 + }, + { + "epoch": 0.513053048646269, + "grad_norm": 1.1666493328982908, + "learning_rate": 5.031546180276692e-06, + "loss": 0.2799, + "step": 23308 + }, + { + "epoch": 0.5130750605326877, + "grad_norm": 1.2787991886399817, + "learning_rate": 5.031189730912201e-06, + "loss": 0.2851, + "step": 23309 + }, + { + "epoch": 0.5130970724191063, + "grad_norm": 0.6144876671600609, + "learning_rate": 5.030833281389191e-06, + "loss": 0.4719, + "step": 23310 + }, + { + "epoch": 0.513119084305525, + "grad_norm": 1.555688719007785, + "learning_rate": 5.030476831709473e-06, + "loss": 0.4286, + "step": 23311 + }, + { + "epoch": 0.5131410961919437, + "grad_norm": 1.1784922062798484, + "learning_rate": 5.030120381874857e-06, + "loss": 0.3936, + "step": 23312 + }, + { + "epoch": 0.5131631080783623, + "grad_norm": 1.4449968473146368, + "learning_rate": 5.029763931887155e-06, + "loss": 0.3478, + "step": 23313 + }, + { + "epoch": 0.513185119964781, + "grad_norm": 1.479304160435425, + "learning_rate": 5.029407481748181e-06, + "loss": 0.3281, + "step": 23314 + }, + { + "epoch": 0.5132071318511996, + "grad_norm": 1.9276833332768646, + "learning_rate": 5.0290510314597455e-06, + "loss": 0.2941, + "step": 23315 + }, + { + "epoch": 0.5132291437376183, + "grad_norm": 1.3822226162876299, + "learning_rate": 5.028694581023658e-06, + "loss": 0.2744, + "step": 23316 + }, + { + "epoch": 0.513251155624037, + "grad_norm": 1.4237413339044012, + "learning_rate": 5.028338130441733e-06, + "loss": 0.318, + "step": 23317 + }, + { + "epoch": 0.5132731675104556, + "grad_norm": 1.3395091985552383, + "learning_rate": 5.02798167971578e-06, + "loss": 0.2761, + "step": 23318 + }, + { + "epoch": 0.5132951793968743, + "grad_norm": 1.481725908873524, + "learning_rate": 5.027625228847611e-06, + "loss": 0.3242, + "step": 23319 + }, + { + "epoch": 0.513317191283293, + "grad_norm": 2.133049374793269, + "learning_rate": 5.02726877783904e-06, + "loss": 0.3454, + "step": 23320 + }, + { + "epoch": 0.5133392031697116, + "grad_norm": 0.6066678280233165, + "learning_rate": 5.026912326691876e-06, + "loss": 0.4748, + "step": 23321 + }, + { + "epoch": 0.5133612150561303, + "grad_norm": 1.3711714805293527, + "learning_rate": 5.0265558754079315e-06, + "loss": 0.2997, + "step": 23322 + }, + { + "epoch": 0.513383226942549, + "grad_norm": 1.694427355446167, + "learning_rate": 5.026199423989019e-06, + "loss": 0.3345, + "step": 23323 + }, + { + "epoch": 0.5134052388289676, + "grad_norm": 3.437688017075241, + "learning_rate": 5.0258429724369486e-06, + "loss": 0.3353, + "step": 23324 + }, + { + "epoch": 0.5134272507153863, + "grad_norm": 1.4472567573598025, + "learning_rate": 5.025486520753533e-06, + "loss": 0.3602, + "step": 23325 + }, + { + "epoch": 0.513449262601805, + "grad_norm": 1.7409819250626626, + "learning_rate": 5.0251300689405824e-06, + "loss": 0.3467, + "step": 23326 + }, + { + "epoch": 0.5134712744882236, + "grad_norm": 0.6222062129784903, + "learning_rate": 5.02477361699991e-06, + "loss": 0.4803, + "step": 23327 + }, + { + "epoch": 0.5134932863746423, + "grad_norm": 1.6460715256145735, + "learning_rate": 5.024417164933328e-06, + "loss": 0.372, + "step": 23328 + }, + { + "epoch": 0.513515298261061, + "grad_norm": 2.308887928153292, + "learning_rate": 5.024060712742648e-06, + "loss": 0.3205, + "step": 23329 + }, + { + "epoch": 0.5135373101474796, + "grad_norm": 1.4453155150839652, + "learning_rate": 5.023704260429679e-06, + "loss": 0.2947, + "step": 23330 + }, + { + "epoch": 0.5135593220338983, + "grad_norm": 1.8240514805397527, + "learning_rate": 5.023347807996234e-06, + "loss": 0.298, + "step": 23331 + }, + { + "epoch": 0.513581333920317, + "grad_norm": 1.6643001597244635, + "learning_rate": 5.022991355444127e-06, + "loss": 0.3212, + "step": 23332 + }, + { + "epoch": 0.5136033458067356, + "grad_norm": 1.4664247437449316, + "learning_rate": 5.022634902775168e-06, + "loss": 0.3704, + "step": 23333 + }, + { + "epoch": 0.5136253576931543, + "grad_norm": 1.3403853575580686, + "learning_rate": 5.022278449991165e-06, + "loss": 0.3278, + "step": 23334 + }, + { + "epoch": 0.513647369579573, + "grad_norm": 0.6164024973575066, + "learning_rate": 5.0219219970939345e-06, + "loss": 0.4469, + "step": 23335 + }, + { + "epoch": 0.5136693814659916, + "grad_norm": 1.8011269963990515, + "learning_rate": 5.021565544085288e-06, + "loss": 0.3801, + "step": 23336 + }, + { + "epoch": 0.5136913933524103, + "grad_norm": 1.6235406613654653, + "learning_rate": 5.021209090967034e-06, + "loss": 0.2521, + "step": 23337 + }, + { + "epoch": 0.5137134052388289, + "grad_norm": 0.604142140191883, + "learning_rate": 5.0208526377409875e-06, + "loss": 0.4749, + "step": 23338 + }, + { + "epoch": 0.5137354171252476, + "grad_norm": 1.4459171233786172, + "learning_rate": 5.020496184408957e-06, + "loss": 0.3107, + "step": 23339 + }, + { + "epoch": 0.5137574290116663, + "grad_norm": 0.6368242245480636, + "learning_rate": 5.020139730972756e-06, + "loss": 0.4764, + "step": 23340 + }, + { + "epoch": 0.5137794408980849, + "grad_norm": 2.014356437922738, + "learning_rate": 5.019783277434197e-06, + "loss": 0.271, + "step": 23341 + }, + { + "epoch": 0.5138014527845036, + "grad_norm": 0.5860172006127995, + "learning_rate": 5.0194268237950905e-06, + "loss": 0.4528, + "step": 23342 + }, + { + "epoch": 0.5138234646709223, + "grad_norm": 0.5841685694589466, + "learning_rate": 5.019070370057247e-06, + "loss": 0.4626, + "step": 23343 + }, + { + "epoch": 0.5138454765573409, + "grad_norm": 1.6973670583432716, + "learning_rate": 5.01871391622248e-06, + "loss": 0.2822, + "step": 23344 + }, + { + "epoch": 0.5138674884437596, + "grad_norm": 1.3716822916195817, + "learning_rate": 5.018357462292601e-06, + "loss": 0.337, + "step": 23345 + }, + { + "epoch": 0.5138895003301783, + "grad_norm": 1.4856710410363672, + "learning_rate": 5.018001008269421e-06, + "loss": 0.415, + "step": 23346 + }, + { + "epoch": 0.5139115122165969, + "grad_norm": 1.2314245745753192, + "learning_rate": 5.017644554154752e-06, + "loss": 0.2793, + "step": 23347 + }, + { + "epoch": 0.5139335241030156, + "grad_norm": 1.5759025676993839, + "learning_rate": 5.017288099950404e-06, + "loss": 0.2826, + "step": 23348 + }, + { + "epoch": 0.5139555359894343, + "grad_norm": 1.275384821903302, + "learning_rate": 5.016931645658192e-06, + "loss": 0.3074, + "step": 23349 + }, + { + "epoch": 0.5139775478758529, + "grad_norm": 1.462281652087095, + "learning_rate": 5.016575191279924e-06, + "loss": 0.3136, + "step": 23350 + }, + { + "epoch": 0.5139995597622716, + "grad_norm": 1.3536426273540685, + "learning_rate": 5.016218736817415e-06, + "loss": 0.3434, + "step": 23351 + }, + { + "epoch": 0.5140215716486903, + "grad_norm": 2.1129005186372285, + "learning_rate": 5.015862282272476e-06, + "loss": 0.3394, + "step": 23352 + }, + { + "epoch": 0.5140435835351089, + "grad_norm": 3.6220593646769914, + "learning_rate": 5.015505827646916e-06, + "loss": 0.2845, + "step": 23353 + }, + { + "epoch": 0.5140655954215276, + "grad_norm": 1.47591296918854, + "learning_rate": 5.015149372942549e-06, + "loss": 0.2835, + "step": 23354 + }, + { + "epoch": 0.5140876073079463, + "grad_norm": 5.4450716297862245, + "learning_rate": 5.014792918161185e-06, + "loss": 0.4139, + "step": 23355 + }, + { + "epoch": 0.5141096191943649, + "grad_norm": 2.8184089851829666, + "learning_rate": 5.01443646330464e-06, + "loss": 0.3443, + "step": 23356 + }, + { + "epoch": 0.5141316310807836, + "grad_norm": 1.366129841748849, + "learning_rate": 5.01408000837472e-06, + "loss": 0.3444, + "step": 23357 + }, + { + "epoch": 0.5141536429672023, + "grad_norm": 1.3869039151329816, + "learning_rate": 5.01372355337324e-06, + "loss": 0.3016, + "step": 23358 + }, + { + "epoch": 0.5141756548536209, + "grad_norm": 1.888073058279532, + "learning_rate": 5.01336709830201e-06, + "loss": 0.3823, + "step": 23359 + }, + { + "epoch": 0.5141976667400396, + "grad_norm": 1.5327740368320677, + "learning_rate": 5.013010643162842e-06, + "loss": 0.3018, + "step": 23360 + }, + { + "epoch": 0.5142196786264582, + "grad_norm": 0.6526281791523716, + "learning_rate": 5.01265418795755e-06, + "loss": 0.4939, + "step": 23361 + }, + { + "epoch": 0.5142416905128769, + "grad_norm": 2.325891663029763, + "learning_rate": 5.012297732687942e-06, + "loss": 0.3109, + "step": 23362 + }, + { + "epoch": 0.5142637023992956, + "grad_norm": 1.5595093305741896, + "learning_rate": 5.011941277355832e-06, + "loss": 0.3791, + "step": 23363 + }, + { + "epoch": 0.5142857142857142, + "grad_norm": 2.2365835228441435, + "learning_rate": 5.011584821963031e-06, + "loss": 0.3759, + "step": 23364 + }, + { + "epoch": 0.5143077261721329, + "grad_norm": 1.0978717321055451, + "learning_rate": 5.0112283665113514e-06, + "loss": 0.2432, + "step": 23365 + }, + { + "epoch": 0.5143297380585516, + "grad_norm": 0.6049658769546323, + "learning_rate": 5.010871911002604e-06, + "loss": 0.4763, + "step": 23366 + }, + { + "epoch": 0.5143517499449702, + "grad_norm": 0.6108316179388028, + "learning_rate": 5.010515455438599e-06, + "loss": 0.4565, + "step": 23367 + }, + { + "epoch": 0.5143737618313889, + "grad_norm": 1.7163746736702452, + "learning_rate": 5.010158999821152e-06, + "loss": 0.321, + "step": 23368 + }, + { + "epoch": 0.5143957737178076, + "grad_norm": 1.2824871124743973, + "learning_rate": 5.009802544152072e-06, + "loss": 0.2973, + "step": 23369 + }, + { + "epoch": 0.5144177856042262, + "grad_norm": 1.5018760207527702, + "learning_rate": 5.009446088433171e-06, + "loss": 0.3059, + "step": 23370 + }, + { + "epoch": 0.514439797490645, + "grad_norm": 2.066721739970153, + "learning_rate": 5.0090896326662606e-06, + "loss": 0.2789, + "step": 23371 + }, + { + "epoch": 0.5144618093770637, + "grad_norm": 1.4075489723425165, + "learning_rate": 5.008733176853154e-06, + "loss": 0.2938, + "step": 23372 + }, + { + "epoch": 0.5144838212634822, + "grad_norm": 0.6271064197626993, + "learning_rate": 5.008376720995658e-06, + "loss": 0.4847, + "step": 23373 + }, + { + "epoch": 0.514505833149901, + "grad_norm": 0.6056280813984695, + "learning_rate": 5.008020265095591e-06, + "loss": 0.4838, + "step": 23374 + }, + { + "epoch": 0.5145278450363197, + "grad_norm": 0.6273587993038264, + "learning_rate": 5.0076638091547595e-06, + "loss": 0.4781, + "step": 23375 + }, + { + "epoch": 0.5145498569227382, + "grad_norm": 1.3778838199378456, + "learning_rate": 5.007307353174979e-06, + "loss": 0.2493, + "step": 23376 + }, + { + "epoch": 0.514571868809157, + "grad_norm": 1.8199897364834698, + "learning_rate": 5.006950897158057e-06, + "loss": 0.2774, + "step": 23377 + }, + { + "epoch": 0.5145938806955757, + "grad_norm": 0.6046030420875126, + "learning_rate": 5.006594441105809e-06, + "loss": 0.4752, + "step": 23378 + }, + { + "epoch": 0.5146158925819942, + "grad_norm": 1.1833988322297853, + "learning_rate": 5.006237985020045e-06, + "loss": 0.3546, + "step": 23379 + }, + { + "epoch": 0.514637904468413, + "grad_norm": 2.3744732036892797, + "learning_rate": 5.005881528902575e-06, + "loss": 0.2702, + "step": 23380 + }, + { + "epoch": 0.5146599163548317, + "grad_norm": 2.0619250643234244, + "learning_rate": 5.0055250727552155e-06, + "loss": 0.3523, + "step": 23381 + }, + { + "epoch": 0.5146819282412503, + "grad_norm": 1.5361994734198865, + "learning_rate": 5.005168616579773e-06, + "loss": 0.3058, + "step": 23382 + }, + { + "epoch": 0.514703940127669, + "grad_norm": 2.196028148149519, + "learning_rate": 5.004812160378062e-06, + "loss": 0.3061, + "step": 23383 + }, + { + "epoch": 0.5147259520140877, + "grad_norm": 1.333644448150838, + "learning_rate": 5.004455704151894e-06, + "loss": 0.3237, + "step": 23384 + }, + { + "epoch": 0.5147479639005063, + "grad_norm": 2.911263088703022, + "learning_rate": 5.004099247903078e-06, + "loss": 0.3054, + "step": 23385 + }, + { + "epoch": 0.514769975786925, + "grad_norm": 1.7023414278536249, + "learning_rate": 5.003742791633429e-06, + "loss": 0.2707, + "step": 23386 + }, + { + "epoch": 0.5147919876733436, + "grad_norm": 1.3392155874516867, + "learning_rate": 5.003386335344759e-06, + "loss": 0.3718, + "step": 23387 + }, + { + "epoch": 0.5148139995597623, + "grad_norm": 1.535782228835972, + "learning_rate": 5.003029879038876e-06, + "loss": 0.3479, + "step": 23388 + }, + { + "epoch": 0.514836011446181, + "grad_norm": 1.501823892960074, + "learning_rate": 5.002673422717595e-06, + "loss": 0.2859, + "step": 23389 + }, + { + "epoch": 0.5148580233325996, + "grad_norm": 1.5488546441931705, + "learning_rate": 5.002316966382726e-06, + "loss": 0.3546, + "step": 23390 + }, + { + "epoch": 0.5148800352190183, + "grad_norm": 3.1199020146105028, + "learning_rate": 5.001960510036081e-06, + "loss": 0.3094, + "step": 23391 + }, + { + "epoch": 0.514902047105437, + "grad_norm": 2.0159056082907734, + "learning_rate": 5.001604053679473e-06, + "loss": 0.3807, + "step": 23392 + }, + { + "epoch": 0.5149240589918556, + "grad_norm": 1.350784425326172, + "learning_rate": 5.001247597314711e-06, + "loss": 0.2414, + "step": 23393 + }, + { + "epoch": 0.5149460708782743, + "grad_norm": 1.6133261122915414, + "learning_rate": 5.0008911409436086e-06, + "loss": 0.3149, + "step": 23394 + }, + { + "epoch": 0.514968082764693, + "grad_norm": 0.6947867907847979, + "learning_rate": 5.000534684567977e-06, + "loss": 0.4693, + "step": 23395 + }, + { + "epoch": 0.5149900946511116, + "grad_norm": 0.6684331321232388, + "learning_rate": 5.000178228189627e-06, + "loss": 0.4783, + "step": 23396 + }, + { + "epoch": 0.5150121065375303, + "grad_norm": 1.4002650440279214, + "learning_rate": 4.9998217718103745e-06, + "loss": 0.3533, + "step": 23397 + }, + { + "epoch": 0.515034118423949, + "grad_norm": 1.3014890984473875, + "learning_rate": 4.999465315432024e-06, + "loss": 0.2508, + "step": 23398 + }, + { + "epoch": 0.5150561303103676, + "grad_norm": 1.5089774408830354, + "learning_rate": 4.999108859056393e-06, + "loss": 0.3052, + "step": 23399 + }, + { + "epoch": 0.5150781421967863, + "grad_norm": 1.8903608472789994, + "learning_rate": 4.9987524026852906e-06, + "loss": 0.3786, + "step": 23400 + }, + { + "epoch": 0.515100154083205, + "grad_norm": 1.243770337272223, + "learning_rate": 4.998395946320528e-06, + "loss": 0.2919, + "step": 23401 + }, + { + "epoch": 0.5151221659696236, + "grad_norm": 1.5431832699543895, + "learning_rate": 4.998039489963921e-06, + "loss": 0.2482, + "step": 23402 + }, + { + "epoch": 0.5151441778560423, + "grad_norm": 1.573104710845855, + "learning_rate": 4.997683033617275e-06, + "loss": 0.2404, + "step": 23403 + }, + { + "epoch": 0.515166189742461, + "grad_norm": 1.1493754068495434, + "learning_rate": 4.997326577282406e-06, + "loss": 0.3597, + "step": 23404 + }, + { + "epoch": 0.5151882016288796, + "grad_norm": 1.70510393414731, + "learning_rate": 4.996970120961126e-06, + "loss": 0.3752, + "step": 23405 + }, + { + "epoch": 0.5152102135152983, + "grad_norm": 1.6926126895355251, + "learning_rate": 4.996613664655242e-06, + "loss": 0.3255, + "step": 23406 + }, + { + "epoch": 0.515232225401717, + "grad_norm": 1.2738560217275474, + "learning_rate": 4.996257208366572e-06, + "loss": 0.2805, + "step": 23407 + }, + { + "epoch": 0.5152542372881356, + "grad_norm": 1.1857523553314093, + "learning_rate": 4.9959007520969235e-06, + "loss": 0.2795, + "step": 23408 + }, + { + "epoch": 0.5152762491745543, + "grad_norm": 1.3980609863849431, + "learning_rate": 4.995544295848108e-06, + "loss": 0.3392, + "step": 23409 + }, + { + "epoch": 0.5152982610609729, + "grad_norm": 1.5715760617400982, + "learning_rate": 4.99518783962194e-06, + "loss": 0.34, + "step": 23410 + }, + { + "epoch": 0.5153202729473916, + "grad_norm": 0.7071103817895736, + "learning_rate": 4.994831383420228e-06, + "loss": 0.4904, + "step": 23411 + }, + { + "epoch": 0.5153422848338103, + "grad_norm": 1.2502929001219243, + "learning_rate": 4.994474927244787e-06, + "loss": 0.3055, + "step": 23412 + }, + { + "epoch": 0.5153642967202289, + "grad_norm": 0.6359209138654563, + "learning_rate": 4.994118471097426e-06, + "loss": 0.4514, + "step": 23413 + }, + { + "epoch": 0.5153863086066476, + "grad_norm": 1.564366365304865, + "learning_rate": 4.993762014979956e-06, + "loss": 0.2043, + "step": 23414 + }, + { + "epoch": 0.5154083204930663, + "grad_norm": 1.2712242976438108, + "learning_rate": 4.993405558894194e-06, + "loss": 0.2628, + "step": 23415 + }, + { + "epoch": 0.5154303323794849, + "grad_norm": 0.6158446791302842, + "learning_rate": 4.993049102841944e-06, + "loss": 0.479, + "step": 23416 + }, + { + "epoch": 0.5154523442659036, + "grad_norm": 1.2353870879008217, + "learning_rate": 4.992692646825023e-06, + "loss": 0.2922, + "step": 23417 + }, + { + "epoch": 0.5154743561523223, + "grad_norm": 2.889401758030084, + "learning_rate": 4.992336190845241e-06, + "loss": 0.3075, + "step": 23418 + }, + { + "epoch": 0.5154963680387409, + "grad_norm": 1.4169511157217745, + "learning_rate": 4.99197973490441e-06, + "loss": 0.3403, + "step": 23419 + }, + { + "epoch": 0.5155183799251596, + "grad_norm": 1.3061614172076643, + "learning_rate": 4.991623279004344e-06, + "loss": 0.2656, + "step": 23420 + }, + { + "epoch": 0.5155403918115783, + "grad_norm": 1.3554888545569261, + "learning_rate": 4.991266823146849e-06, + "loss": 0.3195, + "step": 23421 + }, + { + "epoch": 0.5155624036979969, + "grad_norm": 1.5243155971527422, + "learning_rate": 4.99091036733374e-06, + "loss": 0.327, + "step": 23422 + }, + { + "epoch": 0.5155844155844156, + "grad_norm": 1.2092981286646975, + "learning_rate": 4.990553911566831e-06, + "loss": 0.3325, + "step": 23423 + }, + { + "epoch": 0.5156064274708343, + "grad_norm": 1.8443629149328962, + "learning_rate": 4.990197455847928e-06, + "loss": 0.4126, + "step": 23424 + }, + { + "epoch": 0.5156284393572529, + "grad_norm": 0.7847177655732394, + "learning_rate": 4.989841000178849e-06, + "loss": 0.452, + "step": 23425 + }, + { + "epoch": 0.5156504512436716, + "grad_norm": 1.6192601747325086, + "learning_rate": 4.989484544561401e-06, + "loss": 0.3272, + "step": 23426 + }, + { + "epoch": 0.5156724631300903, + "grad_norm": 2.2337152980337134, + "learning_rate": 4.989128088997398e-06, + "loss": 0.3084, + "step": 23427 + }, + { + "epoch": 0.5156944750165089, + "grad_norm": 1.3933306400517185, + "learning_rate": 4.988771633488651e-06, + "loss": 0.3158, + "step": 23428 + }, + { + "epoch": 0.5157164869029276, + "grad_norm": 1.8478339302214684, + "learning_rate": 4.98841517803697e-06, + "loss": 0.338, + "step": 23429 + }, + { + "epoch": 0.5157384987893463, + "grad_norm": 1.2666387752993669, + "learning_rate": 4.98805872264417e-06, + "loss": 0.3084, + "step": 23430 + }, + { + "epoch": 0.5157605106757649, + "grad_norm": 1.3763549192853908, + "learning_rate": 4.987702267312059e-06, + "loss": 0.294, + "step": 23431 + }, + { + "epoch": 0.5157825225621836, + "grad_norm": 2.7521592748259414, + "learning_rate": 4.987345812042452e-06, + "loss": 0.2653, + "step": 23432 + }, + { + "epoch": 0.5158045344486022, + "grad_norm": 1.5984609610140414, + "learning_rate": 4.98698935683716e-06, + "loss": 0.2821, + "step": 23433 + }, + { + "epoch": 0.5158265463350209, + "grad_norm": 1.2204248791614214, + "learning_rate": 4.986632901697991e-06, + "loss": 0.2593, + "step": 23434 + }, + { + "epoch": 0.5158485582214396, + "grad_norm": 1.5163104983588496, + "learning_rate": 4.986276446626762e-06, + "loss": 0.3191, + "step": 23435 + }, + { + "epoch": 0.5158705701078582, + "grad_norm": 1.6123437725246876, + "learning_rate": 4.9859199916252824e-06, + "loss": 0.3599, + "step": 23436 + }, + { + "epoch": 0.5158925819942769, + "grad_norm": 1.48097480368047, + "learning_rate": 4.9855635366953616e-06, + "loss": 0.3508, + "step": 23437 + }, + { + "epoch": 0.5159145938806956, + "grad_norm": 0.6469198752014382, + "learning_rate": 4.985207081838816e-06, + "loss": 0.4517, + "step": 23438 + }, + { + "epoch": 0.5159366057671142, + "grad_norm": 1.8346846783344575, + "learning_rate": 4.984850627057452e-06, + "loss": 0.3321, + "step": 23439 + }, + { + "epoch": 0.5159586176535329, + "grad_norm": 1.6604054147183276, + "learning_rate": 4.984494172353085e-06, + "loss": 0.3263, + "step": 23440 + }, + { + "epoch": 0.5159806295399516, + "grad_norm": 3.6901679172032793, + "learning_rate": 4.9841377177275265e-06, + "loss": 0.3386, + "step": 23441 + }, + { + "epoch": 0.5160026414263702, + "grad_norm": 1.5341635594644027, + "learning_rate": 4.983781263182585e-06, + "loss": 0.3192, + "step": 23442 + }, + { + "epoch": 0.5160246533127889, + "grad_norm": 1.3291255689497434, + "learning_rate": 4.983424808720077e-06, + "loss": 0.367, + "step": 23443 + }, + { + "epoch": 0.5160466651992076, + "grad_norm": 0.6295081362758124, + "learning_rate": 4.983068354341809e-06, + "loss": 0.4612, + "step": 23444 + }, + { + "epoch": 0.5160686770856262, + "grad_norm": 1.6060888195217777, + "learning_rate": 4.9827119000495965e-06, + "loss": 0.3224, + "step": 23445 + }, + { + "epoch": 0.5160906889720449, + "grad_norm": 1.7606384598653475, + "learning_rate": 4.982355445845251e-06, + "loss": 0.3531, + "step": 23446 + }, + { + "epoch": 0.5161127008584636, + "grad_norm": 1.1717008467119716, + "learning_rate": 4.9819989917305805e-06, + "loss": 0.2676, + "step": 23447 + }, + { + "epoch": 0.5161347127448822, + "grad_norm": 2.8599384339666303, + "learning_rate": 4.981642537707401e-06, + "loss": 0.3202, + "step": 23448 + }, + { + "epoch": 0.5161567246313009, + "grad_norm": 1.5178119202917129, + "learning_rate": 4.981286083777521e-06, + "loss": 0.355, + "step": 23449 + }, + { + "epoch": 0.5161787365177196, + "grad_norm": 0.6015727893934038, + "learning_rate": 4.9809296299427534e-06, + "loss": 0.4928, + "step": 23450 + }, + { + "epoch": 0.5162007484041382, + "grad_norm": 3.9124135674797786, + "learning_rate": 4.980573176204912e-06, + "loss": 0.2429, + "step": 23451 + }, + { + "epoch": 0.5162227602905569, + "grad_norm": 0.6141598496820759, + "learning_rate": 4.980216722565804e-06, + "loss": 0.4647, + "step": 23452 + }, + { + "epoch": 0.5162447721769756, + "grad_norm": 0.6221366741117267, + "learning_rate": 4.979860269027245e-06, + "loss": 0.5003, + "step": 23453 + }, + { + "epoch": 0.5162667840633942, + "grad_norm": 2.0584190944642655, + "learning_rate": 4.9795038155910455e-06, + "loss": 0.2672, + "step": 23454 + }, + { + "epoch": 0.5162887959498129, + "grad_norm": 4.120746858700356, + "learning_rate": 4.979147362259013e-06, + "loss": 0.2883, + "step": 23455 + }, + { + "epoch": 0.5163108078362315, + "grad_norm": 1.752097559659737, + "learning_rate": 4.978790909032968e-06, + "loss": 0.3737, + "step": 23456 + }, + { + "epoch": 0.5163328197226502, + "grad_norm": 1.301130527090414, + "learning_rate": 4.978434455914713e-06, + "loss": 0.335, + "step": 23457 + }, + { + "epoch": 0.5163548316090689, + "grad_norm": 1.8654758908513212, + "learning_rate": 4.978078002906066e-06, + "loss": 0.3317, + "step": 23458 + }, + { + "epoch": 0.5163768434954875, + "grad_norm": 0.6312838101132436, + "learning_rate": 4.977721550008837e-06, + "loss": 0.4906, + "step": 23459 + }, + { + "epoch": 0.5163988553819062, + "grad_norm": 0.6077557446072761, + "learning_rate": 4.977365097224835e-06, + "loss": 0.4538, + "step": 23460 + }, + { + "epoch": 0.5164208672683249, + "grad_norm": 1.6715737198171345, + "learning_rate": 4.977008644555875e-06, + "loss": 0.3365, + "step": 23461 + }, + { + "epoch": 0.5164428791547435, + "grad_norm": 1.3439971350092947, + "learning_rate": 4.976652192003765e-06, + "loss": 0.3387, + "step": 23462 + }, + { + "epoch": 0.5164648910411622, + "grad_norm": 1.1241347966816475, + "learning_rate": 4.976295739570322e-06, + "loss": 0.273, + "step": 23463 + }, + { + "epoch": 0.5164869029275809, + "grad_norm": 1.787497588334728, + "learning_rate": 4.975939287257355e-06, + "loss": 0.3343, + "step": 23464 + }, + { + "epoch": 0.5165089148139995, + "grad_norm": 1.1744425648557013, + "learning_rate": 4.9755828350666724e-06, + "loss": 0.2355, + "step": 23465 + }, + { + "epoch": 0.5165309267004182, + "grad_norm": 1.5818386543478191, + "learning_rate": 4.9752263830000905e-06, + "loss": 0.2922, + "step": 23466 + }, + { + "epoch": 0.5165529385868369, + "grad_norm": 1.6569667841660571, + "learning_rate": 4.974869931059419e-06, + "loss": 0.2638, + "step": 23467 + }, + { + "epoch": 0.5165749504732555, + "grad_norm": 1.7077685640025848, + "learning_rate": 4.974513479246469e-06, + "loss": 0.3124, + "step": 23468 + }, + { + "epoch": 0.5165969623596742, + "grad_norm": 3.9139302557532027, + "learning_rate": 4.974157027563053e-06, + "loss": 0.3282, + "step": 23469 + }, + { + "epoch": 0.5166189742460929, + "grad_norm": 1.3011137540063613, + "learning_rate": 4.973800576010983e-06, + "loss": 0.226, + "step": 23470 + }, + { + "epoch": 0.5166409861325115, + "grad_norm": 1.379795290955909, + "learning_rate": 4.97344412459207e-06, + "loss": 0.3615, + "step": 23471 + }, + { + "epoch": 0.5166629980189302, + "grad_norm": 1.3767123144858309, + "learning_rate": 4.973087673308126e-06, + "loss": 0.3235, + "step": 23472 + }, + { + "epoch": 0.5166850099053489, + "grad_norm": 2.361213450324461, + "learning_rate": 4.972731222160961e-06, + "loss": 0.3655, + "step": 23473 + }, + { + "epoch": 0.5167070217917675, + "grad_norm": 1.8635421796151561, + "learning_rate": 4.972374771152391e-06, + "loss": 0.3351, + "step": 23474 + }, + { + "epoch": 0.5167290336781862, + "grad_norm": 1.3643262454548275, + "learning_rate": 4.9720183202842214e-06, + "loss": 0.271, + "step": 23475 + }, + { + "epoch": 0.5167510455646049, + "grad_norm": 1.9693811486541124, + "learning_rate": 4.971661869558269e-06, + "loss": 0.4123, + "step": 23476 + }, + { + "epoch": 0.5167730574510235, + "grad_norm": 2.2682439228081663, + "learning_rate": 4.9713054189763445e-06, + "loss": 0.3145, + "step": 23477 + }, + { + "epoch": 0.5167950693374422, + "grad_norm": 1.328967972026273, + "learning_rate": 4.970948968540256e-06, + "loss": 0.3081, + "step": 23478 + }, + { + "epoch": 0.5168170812238608, + "grad_norm": 1.5462948926006004, + "learning_rate": 4.970592518251821e-06, + "loss": 0.3063, + "step": 23479 + }, + { + "epoch": 0.5168390931102795, + "grad_norm": 0.6889292993465164, + "learning_rate": 4.970236068112845e-06, + "loss": 0.4727, + "step": 23480 + }, + { + "epoch": 0.5168611049966982, + "grad_norm": 0.6354254770306227, + "learning_rate": 4.969879618125144e-06, + "loss": 0.4663, + "step": 23481 + }, + { + "epoch": 0.5168831168831168, + "grad_norm": 1.4143689838024491, + "learning_rate": 4.9695231682905295e-06, + "loss": 0.3105, + "step": 23482 + }, + { + "epoch": 0.5169051287695355, + "grad_norm": 0.6108402885427081, + "learning_rate": 4.96916671861081e-06, + "loss": 0.4568, + "step": 23483 + }, + { + "epoch": 0.5169271406559542, + "grad_norm": 1.5743477309518872, + "learning_rate": 4.9688102690878e-06, + "loss": 0.3552, + "step": 23484 + }, + { + "epoch": 0.5169491525423728, + "grad_norm": 2.3558279398716935, + "learning_rate": 4.968453819723308e-06, + "loss": 0.3117, + "step": 23485 + }, + { + "epoch": 0.5169711644287915, + "grad_norm": 1.6121565923969763, + "learning_rate": 4.96809737051915e-06, + "loss": 0.3444, + "step": 23486 + }, + { + "epoch": 0.5169931763152102, + "grad_norm": 1.7043485103676463, + "learning_rate": 4.967740921477136e-06, + "loss": 0.2864, + "step": 23487 + }, + { + "epoch": 0.5170151882016288, + "grad_norm": 1.3357998533908537, + "learning_rate": 4.967384472599075e-06, + "loss": 0.3085, + "step": 23488 + }, + { + "epoch": 0.5170372000880475, + "grad_norm": 2.4308924109598924, + "learning_rate": 4.967028023886781e-06, + "loss": 0.3781, + "step": 23489 + }, + { + "epoch": 0.5170592119744662, + "grad_norm": 1.4697647160157168, + "learning_rate": 4.966671575342067e-06, + "loss": 0.3752, + "step": 23490 + }, + { + "epoch": 0.5170812238608848, + "grad_norm": 1.212273343963193, + "learning_rate": 4.96631512696674e-06, + "loss": 0.2687, + "step": 23491 + }, + { + "epoch": 0.5171032357473035, + "grad_norm": 1.4614494150900983, + "learning_rate": 4.965958678762617e-06, + "loss": 0.3151, + "step": 23492 + }, + { + "epoch": 0.5171252476337223, + "grad_norm": 0.6671981239534851, + "learning_rate": 4.9656022307315055e-06, + "loss": 0.4667, + "step": 23493 + }, + { + "epoch": 0.5171472595201408, + "grad_norm": 5.359938912521037, + "learning_rate": 4.96524578287522e-06, + "loss": 0.3862, + "step": 23494 + }, + { + "epoch": 0.5171692714065595, + "grad_norm": 0.6322525089286554, + "learning_rate": 4.964889335195572e-06, + "loss": 0.4772, + "step": 23495 + }, + { + "epoch": 0.5171912832929783, + "grad_norm": 1.3055904543470027, + "learning_rate": 4.964532887694368e-06, + "loss": 0.322, + "step": 23496 + }, + { + "epoch": 0.5172132951793968, + "grad_norm": 1.2139100845569182, + "learning_rate": 4.964176440373428e-06, + "loss": 0.3061, + "step": 23497 + }, + { + "epoch": 0.5172353070658156, + "grad_norm": 1.4178439074881226, + "learning_rate": 4.963819993234557e-06, + "loss": 0.2909, + "step": 23498 + }, + { + "epoch": 0.5172573189522343, + "grad_norm": 1.4130490775711042, + "learning_rate": 4.96346354627957e-06, + "loss": 0.3081, + "step": 23499 + }, + { + "epoch": 0.5172793308386529, + "grad_norm": 1.6145410565247513, + "learning_rate": 4.9631070995102785e-06, + "loss": 0.3017, + "step": 23500 + }, + { + "epoch": 0.5173013427250716, + "grad_norm": 0.6393928603989827, + "learning_rate": 4.96275065292849e-06, + "loss": 0.4697, + "step": 23501 + }, + { + "epoch": 0.5173233546114901, + "grad_norm": 2.675463945802597, + "learning_rate": 4.962394206536022e-06, + "loss": 0.3396, + "step": 23502 + }, + { + "epoch": 0.5173453664979089, + "grad_norm": 1.4165434843893743, + "learning_rate": 4.962037760334681e-06, + "loss": 0.3092, + "step": 23503 + }, + { + "epoch": 0.5173673783843276, + "grad_norm": 1.649816900615651, + "learning_rate": 4.961681314326282e-06, + "loss": 0.3008, + "step": 23504 + }, + { + "epoch": 0.5173893902707462, + "grad_norm": 0.6393524526807449, + "learning_rate": 4.961324868512638e-06, + "loss": 0.4641, + "step": 23505 + }, + { + "epoch": 0.5174114021571649, + "grad_norm": 1.8397446635677968, + "learning_rate": 4.960968422895556e-06, + "loss": 0.2938, + "step": 23506 + }, + { + "epoch": 0.5174334140435836, + "grad_norm": 2.68639271542658, + "learning_rate": 4.96061197747685e-06, + "loss": 0.2993, + "step": 23507 + }, + { + "epoch": 0.5174554259300022, + "grad_norm": 1.7665595155277758, + "learning_rate": 4.960255532258334e-06, + "loss": 0.2644, + "step": 23508 + }, + { + "epoch": 0.5174774378164209, + "grad_norm": 1.7955317471725405, + "learning_rate": 4.959899087241812e-06, + "loss": 0.3188, + "step": 23509 + }, + { + "epoch": 0.5174994497028396, + "grad_norm": 2.3332201547370337, + "learning_rate": 4.959542642429105e-06, + "loss": 0.2942, + "step": 23510 + }, + { + "epoch": 0.5175214615892582, + "grad_norm": 1.5356131371760005, + "learning_rate": 4.959186197822019e-06, + "loss": 0.2971, + "step": 23511 + }, + { + "epoch": 0.5175434734756769, + "grad_norm": 1.5575281671114598, + "learning_rate": 4.958829753422367e-06, + "loss": 0.309, + "step": 23512 + }, + { + "epoch": 0.5175654853620956, + "grad_norm": 2.3158136475368853, + "learning_rate": 4.958473309231961e-06, + "loss": 0.3938, + "step": 23513 + }, + { + "epoch": 0.5175874972485142, + "grad_norm": 1.5458830043068938, + "learning_rate": 4.95811686525261e-06, + "loss": 0.2341, + "step": 23514 + }, + { + "epoch": 0.5176095091349329, + "grad_norm": 1.947834799584538, + "learning_rate": 4.9577604214861305e-06, + "loss": 0.299, + "step": 23515 + }, + { + "epoch": 0.5176315210213516, + "grad_norm": 1.8691484439672033, + "learning_rate": 4.9574039779343295e-06, + "loss": 0.2855, + "step": 23516 + }, + { + "epoch": 0.5176535329077702, + "grad_norm": 0.6595331443989139, + "learning_rate": 4.957047534599022e-06, + "loss": 0.4857, + "step": 23517 + }, + { + "epoch": 0.5176755447941889, + "grad_norm": 1.5161332246764787, + "learning_rate": 4.956691091482018e-06, + "loss": 0.2534, + "step": 23518 + }, + { + "epoch": 0.5176975566806076, + "grad_norm": 1.5172584080840974, + "learning_rate": 4.9563346485851284e-06, + "loss": 0.4048, + "step": 23519 + }, + { + "epoch": 0.5177195685670262, + "grad_norm": 1.5903991149272467, + "learning_rate": 4.955978205910166e-06, + "loss": 0.3683, + "step": 23520 + }, + { + "epoch": 0.5177415804534449, + "grad_norm": 1.5590851140808724, + "learning_rate": 4.95562176345894e-06, + "loss": 0.3107, + "step": 23521 + }, + { + "epoch": 0.5177635923398636, + "grad_norm": 3.1740104344844777, + "learning_rate": 4.955265321233267e-06, + "loss": 0.3658, + "step": 23522 + }, + { + "epoch": 0.5177856042262822, + "grad_norm": 1.3077373854789103, + "learning_rate": 4.954908879234955e-06, + "loss": 0.297, + "step": 23523 + }, + { + "epoch": 0.5178076161127009, + "grad_norm": 1.555407561993446, + "learning_rate": 4.954552437465815e-06, + "loss": 0.2928, + "step": 23524 + }, + { + "epoch": 0.5178296279991196, + "grad_norm": 1.8074991447161861, + "learning_rate": 4.95419599592766e-06, + "loss": 0.2759, + "step": 23525 + }, + { + "epoch": 0.5178516398855382, + "grad_norm": 1.544529060325719, + "learning_rate": 4.953839554622301e-06, + "loss": 0.2378, + "step": 23526 + }, + { + "epoch": 0.5178736517719569, + "grad_norm": 1.2680513207190973, + "learning_rate": 4.953483113551551e-06, + "loss": 0.2517, + "step": 23527 + }, + { + "epoch": 0.5178956636583755, + "grad_norm": 2.097339884844544, + "learning_rate": 4.953126672717221e-06, + "loss": 0.3211, + "step": 23528 + }, + { + "epoch": 0.5179176755447942, + "grad_norm": 1.6106306639403345, + "learning_rate": 4.9527702321211205e-06, + "loss": 0.256, + "step": 23529 + }, + { + "epoch": 0.5179396874312129, + "grad_norm": 0.6614061519398263, + "learning_rate": 4.952413791765064e-06, + "loss": 0.4764, + "step": 23530 + }, + { + "epoch": 0.5179616993176315, + "grad_norm": 2.0950777427930176, + "learning_rate": 4.952057351650863e-06, + "loss": 0.2764, + "step": 23531 + }, + { + "epoch": 0.5179837112040502, + "grad_norm": 1.3785727651025552, + "learning_rate": 4.951700911780324e-06, + "loss": 0.2777, + "step": 23532 + }, + { + "epoch": 0.5180057230904689, + "grad_norm": 0.626496760314821, + "learning_rate": 4.951344472155266e-06, + "loss": 0.474, + "step": 23533 + }, + { + "epoch": 0.5180277349768875, + "grad_norm": 1.3286313509043974, + "learning_rate": 4.950988032777495e-06, + "loss": 0.2602, + "step": 23534 + }, + { + "epoch": 0.5180497468633062, + "grad_norm": 1.4425326135678622, + "learning_rate": 4.950631593648827e-06, + "loss": 0.293, + "step": 23535 + }, + { + "epoch": 0.5180717587497249, + "grad_norm": 2.236459231155583, + "learning_rate": 4.95027515477107e-06, + "loss": 0.3036, + "step": 23536 + }, + { + "epoch": 0.5180937706361435, + "grad_norm": 0.6044703290612562, + "learning_rate": 4.9499187161460365e-06, + "loss": 0.4985, + "step": 23537 + }, + { + "epoch": 0.5181157825225622, + "grad_norm": 1.2905201387457554, + "learning_rate": 4.9495622777755386e-06, + "loss": 0.2534, + "step": 23538 + }, + { + "epoch": 0.5181377944089809, + "grad_norm": 1.3960703085197672, + "learning_rate": 4.949205839661387e-06, + "loss": 0.3269, + "step": 23539 + }, + { + "epoch": 0.5181598062953995, + "grad_norm": 1.7217055580125162, + "learning_rate": 4.948849401805395e-06, + "loss": 0.3485, + "step": 23540 + }, + { + "epoch": 0.5181818181818182, + "grad_norm": 1.4269976549902519, + "learning_rate": 4.948492964209374e-06, + "loss": 0.2721, + "step": 23541 + }, + { + "epoch": 0.5182038300682369, + "grad_norm": 1.681197003482402, + "learning_rate": 4.948136526875133e-06, + "loss": 0.3394, + "step": 23542 + }, + { + "epoch": 0.5182258419546555, + "grad_norm": 1.7004943824194179, + "learning_rate": 4.947780089804486e-06, + "loss": 0.3092, + "step": 23543 + }, + { + "epoch": 0.5182478538410742, + "grad_norm": 1.1876412094752664, + "learning_rate": 4.947423652999243e-06, + "loss": 0.2626, + "step": 23544 + }, + { + "epoch": 0.5182698657274929, + "grad_norm": 2.5570338730374735, + "learning_rate": 4.947067216461218e-06, + "loss": 0.3454, + "step": 23545 + }, + { + "epoch": 0.5182918776139115, + "grad_norm": 0.6613628196871576, + "learning_rate": 4.946710780192221e-06, + "loss": 0.4664, + "step": 23546 + }, + { + "epoch": 0.5183138895003302, + "grad_norm": 1.3567061449266917, + "learning_rate": 4.946354344194062e-06, + "loss": 0.3917, + "step": 23547 + }, + { + "epoch": 0.5183359013867489, + "grad_norm": 1.7632885795458693, + "learning_rate": 4.945997908468555e-06, + "loss": 0.251, + "step": 23548 + }, + { + "epoch": 0.5183579132731675, + "grad_norm": 3.8435249680021144, + "learning_rate": 4.945641473017512e-06, + "loss": 0.3628, + "step": 23549 + }, + { + "epoch": 0.5183799251595862, + "grad_norm": 1.3917038107018587, + "learning_rate": 4.94528503784274e-06, + "loss": 0.2983, + "step": 23550 + }, + { + "epoch": 0.5184019370460048, + "grad_norm": 1.4455155361854939, + "learning_rate": 4.944928602946057e-06, + "loss": 0.3455, + "step": 23551 + }, + { + "epoch": 0.5184239489324235, + "grad_norm": 1.5271277673318815, + "learning_rate": 4.9445721683292696e-06, + "loss": 0.3136, + "step": 23552 + }, + { + "epoch": 0.5184459608188422, + "grad_norm": 1.4805128071236158, + "learning_rate": 4.9442157339941925e-06, + "loss": 0.2347, + "step": 23553 + }, + { + "epoch": 0.5184679727052608, + "grad_norm": 1.6244570201123532, + "learning_rate": 4.943859299942636e-06, + "loss": 0.2069, + "step": 23554 + }, + { + "epoch": 0.5184899845916795, + "grad_norm": 1.2544034149191903, + "learning_rate": 4.94350286617641e-06, + "loss": 0.255, + "step": 23555 + }, + { + "epoch": 0.5185119964780982, + "grad_norm": 1.32460169104547, + "learning_rate": 4.943146432697329e-06, + "loss": 0.2116, + "step": 23556 + }, + { + "epoch": 0.5185340083645168, + "grad_norm": 0.6436710495333718, + "learning_rate": 4.942789999507202e-06, + "loss": 0.4823, + "step": 23557 + }, + { + "epoch": 0.5185560202509355, + "grad_norm": 1.8355137574552194, + "learning_rate": 4.942433566607842e-06, + "loss": 0.3076, + "step": 23558 + }, + { + "epoch": 0.5185780321373542, + "grad_norm": 0.6275975265062163, + "learning_rate": 4.942077134001062e-06, + "loss": 0.4642, + "step": 23559 + }, + { + "epoch": 0.5186000440237728, + "grad_norm": 1.3882963402973936, + "learning_rate": 4.941720701688669e-06, + "loss": 0.3491, + "step": 23560 + }, + { + "epoch": 0.5186220559101915, + "grad_norm": 1.3973804847016544, + "learning_rate": 4.9413642696724795e-06, + "loss": 0.3884, + "step": 23561 + }, + { + "epoch": 0.5186440677966102, + "grad_norm": 5.255488944699806, + "learning_rate": 4.941007837954301e-06, + "loss": 0.3041, + "step": 23562 + }, + { + "epoch": 0.5186660796830288, + "grad_norm": 1.4419165863550196, + "learning_rate": 4.940651406535949e-06, + "loss": 0.3249, + "step": 23563 + }, + { + "epoch": 0.5186880915694475, + "grad_norm": 1.2135897800394417, + "learning_rate": 4.940294975419234e-06, + "loss": 0.2335, + "step": 23564 + }, + { + "epoch": 0.5187101034558662, + "grad_norm": 1.5037212497713393, + "learning_rate": 4.939938544605963e-06, + "loss": 0.2978, + "step": 23565 + }, + { + "epoch": 0.5187321153422848, + "grad_norm": 1.404568129772169, + "learning_rate": 4.939582114097954e-06, + "loss": 0.3324, + "step": 23566 + }, + { + "epoch": 0.5187541272287035, + "grad_norm": 2.009205293014541, + "learning_rate": 4.939225683897016e-06, + "loss": 0.3176, + "step": 23567 + }, + { + "epoch": 0.5187761391151222, + "grad_norm": 1.352423431264203, + "learning_rate": 4.938869254004957e-06, + "loss": 0.2722, + "step": 23568 + }, + { + "epoch": 0.5187981510015408, + "grad_norm": 1.8978815379124965, + "learning_rate": 4.938512824423594e-06, + "loss": 0.2584, + "step": 23569 + }, + { + "epoch": 0.5188201628879595, + "grad_norm": 1.704281888264156, + "learning_rate": 4.938156395154735e-06, + "loss": 0.3533, + "step": 23570 + }, + { + "epoch": 0.5188421747743782, + "grad_norm": 1.412741193809605, + "learning_rate": 4.937799966200194e-06, + "loss": 0.2788, + "step": 23571 + }, + { + "epoch": 0.5188641866607968, + "grad_norm": 0.6583648387453604, + "learning_rate": 4.937443537561782e-06, + "loss": 0.4796, + "step": 23572 + }, + { + "epoch": 0.5188861985472155, + "grad_norm": 1.586603136785446, + "learning_rate": 4.937087109241307e-06, + "loss": 0.3832, + "step": 23573 + }, + { + "epoch": 0.5189082104336341, + "grad_norm": 0.6290556037844172, + "learning_rate": 4.9367306812405865e-06, + "loss": 0.4641, + "step": 23574 + }, + { + "epoch": 0.5189302223200528, + "grad_norm": 1.335676015395556, + "learning_rate": 4.9363742535614265e-06, + "loss": 0.2653, + "step": 23575 + }, + { + "epoch": 0.5189522342064715, + "grad_norm": 1.290894997190314, + "learning_rate": 4.936017826205642e-06, + "loss": 0.2648, + "step": 23576 + }, + { + "epoch": 0.5189742460928901, + "grad_norm": 1.7793755913455016, + "learning_rate": 4.935661399175045e-06, + "loss": 0.3783, + "step": 23577 + }, + { + "epoch": 0.5189962579793088, + "grad_norm": 1.3098677203515536, + "learning_rate": 4.935304972471442e-06, + "loss": 0.2444, + "step": 23578 + }, + { + "epoch": 0.5190182698657275, + "grad_norm": 1.469914130393657, + "learning_rate": 4.93494854609665e-06, + "loss": 0.2708, + "step": 23579 + }, + { + "epoch": 0.5190402817521461, + "grad_norm": 1.5331735873889736, + "learning_rate": 4.9345921200524774e-06, + "loss": 0.4373, + "step": 23580 + }, + { + "epoch": 0.5190622936385648, + "grad_norm": 1.5566124017849199, + "learning_rate": 4.934235694340738e-06, + "loss": 0.291, + "step": 23581 + }, + { + "epoch": 0.5190843055249835, + "grad_norm": 2.288324383604757, + "learning_rate": 4.933879268963243e-06, + "loss": 0.3355, + "step": 23582 + }, + { + "epoch": 0.5191063174114021, + "grad_norm": 1.3376337331654724, + "learning_rate": 4.933522843921802e-06, + "loss": 0.257, + "step": 23583 + }, + { + "epoch": 0.5191283292978208, + "grad_norm": 1.773428619779248, + "learning_rate": 4.933166419218227e-06, + "loss": 0.2715, + "step": 23584 + }, + { + "epoch": 0.5191503411842395, + "grad_norm": 1.2503547783970537, + "learning_rate": 4.932809994854329e-06, + "loss": 0.3284, + "step": 23585 + }, + { + "epoch": 0.5191723530706581, + "grad_norm": 1.3188087908283166, + "learning_rate": 4.932453570831924e-06, + "loss": 0.2638, + "step": 23586 + }, + { + "epoch": 0.5191943649570768, + "grad_norm": 1.51850943845285, + "learning_rate": 4.932097147152819e-06, + "loss": 0.3318, + "step": 23587 + }, + { + "epoch": 0.5192163768434955, + "grad_norm": 1.6043234891817342, + "learning_rate": 4.931740723818824e-06, + "loss": 0.297, + "step": 23588 + }, + { + "epoch": 0.5192383887299141, + "grad_norm": 1.5134221655221158, + "learning_rate": 4.931384300831755e-06, + "loss": 0.271, + "step": 23589 + }, + { + "epoch": 0.5192604006163328, + "grad_norm": 1.689638938021708, + "learning_rate": 4.931027878193423e-06, + "loss": 0.4226, + "step": 23590 + }, + { + "epoch": 0.5192824125027515, + "grad_norm": 2.0099848602237858, + "learning_rate": 4.930671455905635e-06, + "loss": 0.3092, + "step": 23591 + }, + { + "epoch": 0.5193044243891701, + "grad_norm": 1.4671683008998642, + "learning_rate": 4.9303150339702075e-06, + "loss": 0.2471, + "step": 23592 + }, + { + "epoch": 0.5193264362755888, + "grad_norm": 1.9029189027047626, + "learning_rate": 4.9299586123889495e-06, + "loss": 0.3605, + "step": 23593 + }, + { + "epoch": 0.5193484481620075, + "grad_norm": 2.4905221301756404, + "learning_rate": 4.929602191163673e-06, + "loss": 0.3582, + "step": 23594 + }, + { + "epoch": 0.5193704600484261, + "grad_norm": 3.3430016082500975, + "learning_rate": 4.929245770296191e-06, + "loss": 0.3811, + "step": 23595 + }, + { + "epoch": 0.5193924719348448, + "grad_norm": 1.3441228388211723, + "learning_rate": 4.928889349788311e-06, + "loss": 0.3367, + "step": 23596 + }, + { + "epoch": 0.5194144838212634, + "grad_norm": 1.7670441017940481, + "learning_rate": 4.928532929641849e-06, + "loss": 0.3544, + "step": 23597 + }, + { + "epoch": 0.5194364957076821, + "grad_norm": 1.5051623719904095, + "learning_rate": 4.9281765098586135e-06, + "loss": 0.3646, + "step": 23598 + }, + { + "epoch": 0.5194585075941008, + "grad_norm": 1.5583304389709631, + "learning_rate": 4.927820090440417e-06, + "loss": 0.3558, + "step": 23599 + }, + { + "epoch": 0.5194805194805194, + "grad_norm": 0.643096250903571, + "learning_rate": 4.9274636713890734e-06, + "loss": 0.4441, + "step": 23600 + }, + { + "epoch": 0.5195025313669381, + "grad_norm": 1.8119135569844118, + "learning_rate": 4.9271072527063885e-06, + "loss": 0.3578, + "step": 23601 + }, + { + "epoch": 0.5195245432533568, + "grad_norm": 1.310657467555578, + "learning_rate": 4.926750834394179e-06, + "loss": 0.274, + "step": 23602 + }, + { + "epoch": 0.5195465551397754, + "grad_norm": 1.4645965081665513, + "learning_rate": 4.926394416454253e-06, + "loss": 0.3678, + "step": 23603 + }, + { + "epoch": 0.5195685670261941, + "grad_norm": 1.2966463509673698, + "learning_rate": 4.926037998888425e-06, + "loss": 0.2864, + "step": 23604 + }, + { + "epoch": 0.5195905789126128, + "grad_norm": 1.3993565419812688, + "learning_rate": 4.925681581698504e-06, + "loss": 0.3204, + "step": 23605 + }, + { + "epoch": 0.5196125907990314, + "grad_norm": 1.2930161994061773, + "learning_rate": 4.925325164886302e-06, + "loss": 0.2722, + "step": 23606 + }, + { + "epoch": 0.5196346026854501, + "grad_norm": 1.5514822956449374, + "learning_rate": 4.924968748453631e-06, + "loss": 0.3096, + "step": 23607 + }, + { + "epoch": 0.5196566145718688, + "grad_norm": 0.648899224600508, + "learning_rate": 4.924612332402304e-06, + "loss": 0.4951, + "step": 23608 + }, + { + "epoch": 0.5196786264582874, + "grad_norm": 4.477747982314056, + "learning_rate": 4.924255916734129e-06, + "loss": 0.2627, + "step": 23609 + }, + { + "epoch": 0.5197006383447061, + "grad_norm": 2.0295436026129066, + "learning_rate": 4.923899501450919e-06, + "loss": 0.3039, + "step": 23610 + }, + { + "epoch": 0.5197226502311248, + "grad_norm": 1.2146625604016652, + "learning_rate": 4.923543086554486e-06, + "loss": 0.3378, + "step": 23611 + }, + { + "epoch": 0.5197446621175434, + "grad_norm": 2.0366089115971753, + "learning_rate": 4.923186672046641e-06, + "loss": 0.3907, + "step": 23612 + }, + { + "epoch": 0.5197666740039621, + "grad_norm": 1.8160361461755166, + "learning_rate": 4.922830257929197e-06, + "loss": 0.3458, + "step": 23613 + }, + { + "epoch": 0.5197886858903809, + "grad_norm": 1.568440748778605, + "learning_rate": 4.9224738442039624e-06, + "loss": 0.3035, + "step": 23614 + }, + { + "epoch": 0.5198106977767994, + "grad_norm": 8.259738000729174, + "learning_rate": 4.92211743087275e-06, + "loss": 0.3034, + "step": 23615 + }, + { + "epoch": 0.5198327096632182, + "grad_norm": 1.296392245775936, + "learning_rate": 4.921761017937372e-06, + "loss": 0.3501, + "step": 23616 + }, + { + "epoch": 0.5198547215496369, + "grad_norm": 2.0111215757489247, + "learning_rate": 4.92140460539964e-06, + "loss": 0.3295, + "step": 23617 + }, + { + "epoch": 0.5198767334360554, + "grad_norm": 1.7741075603695908, + "learning_rate": 4.921048193261366e-06, + "loss": 0.3193, + "step": 23618 + }, + { + "epoch": 0.5198987453224742, + "grad_norm": 1.8526488925257873, + "learning_rate": 4.920691781524358e-06, + "loss": 0.2596, + "step": 23619 + }, + { + "epoch": 0.5199207572088927, + "grad_norm": 1.5120295646346058, + "learning_rate": 4.920335370190431e-06, + "loss": 0.3047, + "step": 23620 + }, + { + "epoch": 0.5199427690953115, + "grad_norm": 1.5085331863767635, + "learning_rate": 4.919978959261394e-06, + "loss": 0.2872, + "step": 23621 + }, + { + "epoch": 0.5199647809817302, + "grad_norm": 1.4249806432535967, + "learning_rate": 4.919622548739062e-06, + "loss": 0.3281, + "step": 23622 + }, + { + "epoch": 0.5199867928681488, + "grad_norm": 1.5393118934442873, + "learning_rate": 4.919266138625243e-06, + "loss": 0.3025, + "step": 23623 + }, + { + "epoch": 0.5200088047545675, + "grad_norm": 1.3106158908825074, + "learning_rate": 4.918909728921747e-06, + "loss": 0.2961, + "step": 23624 + }, + { + "epoch": 0.5200308166409862, + "grad_norm": 1.7321508073996694, + "learning_rate": 4.918553319630391e-06, + "loss": 0.3129, + "step": 23625 + }, + { + "epoch": 0.5200528285274048, + "grad_norm": 1.393710923018905, + "learning_rate": 4.918196910752981e-06, + "loss": 0.2952, + "step": 23626 + }, + { + "epoch": 0.5200748404138235, + "grad_norm": 1.504213192479104, + "learning_rate": 4.917840502291333e-06, + "loss": 0.306, + "step": 23627 + }, + { + "epoch": 0.5200968523002422, + "grad_norm": 0.637614270321197, + "learning_rate": 4.917484094247257e-06, + "loss": 0.4834, + "step": 23628 + }, + { + "epoch": 0.5201188641866608, + "grad_norm": 21.10193052932501, + "learning_rate": 4.91712768662256e-06, + "loss": 0.3639, + "step": 23629 + }, + { + "epoch": 0.5201408760730795, + "grad_norm": 1.6353117824947174, + "learning_rate": 4.916771279419059e-06, + "loss": 0.295, + "step": 23630 + }, + { + "epoch": 0.5201628879594982, + "grad_norm": 1.4028256812431397, + "learning_rate": 4.916414872638565e-06, + "loss": 0.3192, + "step": 23631 + }, + { + "epoch": 0.5201848998459168, + "grad_norm": 1.1950834173957632, + "learning_rate": 4.9160584662828854e-06, + "loss": 0.271, + "step": 23632 + }, + { + "epoch": 0.5202069117323355, + "grad_norm": 0.6179038019912279, + "learning_rate": 4.915702060353835e-06, + "loss": 0.4739, + "step": 23633 + }, + { + "epoch": 0.5202289236187542, + "grad_norm": 1.1793434665883287, + "learning_rate": 4.915345654853224e-06, + "loss": 0.2894, + "step": 23634 + }, + { + "epoch": 0.5202509355051728, + "grad_norm": 1.560793038565103, + "learning_rate": 4.914989249782864e-06, + "loss": 0.3351, + "step": 23635 + }, + { + "epoch": 0.5202729473915915, + "grad_norm": 1.5419834845138793, + "learning_rate": 4.914632845144569e-06, + "loss": 0.3478, + "step": 23636 + }, + { + "epoch": 0.5202949592780102, + "grad_norm": 0.6010632396579286, + "learning_rate": 4.914276440940145e-06, + "loss": 0.4916, + "step": 23637 + }, + { + "epoch": 0.5203169711644288, + "grad_norm": 1.5205653450865444, + "learning_rate": 4.913920037171408e-06, + "loss": 0.3524, + "step": 23638 + }, + { + "epoch": 0.5203389830508475, + "grad_norm": 1.9113010943449964, + "learning_rate": 4.913563633840166e-06, + "loss": 0.3824, + "step": 23639 + }, + { + "epoch": 0.5203609949372662, + "grad_norm": 0.6161499750627438, + "learning_rate": 4.913207230948235e-06, + "loss": 0.477, + "step": 23640 + }, + { + "epoch": 0.5203830068236848, + "grad_norm": 1.7548454594696952, + "learning_rate": 4.912850828497422e-06, + "loss": 0.2581, + "step": 23641 + }, + { + "epoch": 0.5204050187101035, + "grad_norm": 7.394115061250311, + "learning_rate": 4.912494426489538e-06, + "loss": 0.263, + "step": 23642 + }, + { + "epoch": 0.5204270305965221, + "grad_norm": 1.7317035260649813, + "learning_rate": 4.9121380249264e-06, + "loss": 0.3088, + "step": 23643 + }, + { + "epoch": 0.5204490424829408, + "grad_norm": 1.4275561955819929, + "learning_rate": 4.911781623809812e-06, + "loss": 0.2875, + "step": 23644 + }, + { + "epoch": 0.5204710543693595, + "grad_norm": 2.614197436088159, + "learning_rate": 4.911425223141592e-06, + "loss": 0.256, + "step": 23645 + }, + { + "epoch": 0.5204930662557781, + "grad_norm": 1.4203056024842597, + "learning_rate": 4.911068822923549e-06, + "loss": 0.2868, + "step": 23646 + }, + { + "epoch": 0.5205150781421968, + "grad_norm": 1.8645934835071618, + "learning_rate": 4.910712423157491e-06, + "loss": 0.3103, + "step": 23647 + }, + { + "epoch": 0.5205370900286155, + "grad_norm": 1.5830385248127306, + "learning_rate": 4.9103560238452345e-06, + "loss": 0.2371, + "step": 23648 + }, + { + "epoch": 0.5205591019150341, + "grad_norm": 1.2118115333427983, + "learning_rate": 4.909999624988589e-06, + "loss": 0.3123, + "step": 23649 + }, + { + "epoch": 0.5205811138014528, + "grad_norm": 0.6160723421909449, + "learning_rate": 4.909643226589363e-06, + "loss": 0.4684, + "step": 23650 + }, + { + "epoch": 0.5206031256878715, + "grad_norm": 1.9384135277736247, + "learning_rate": 4.9092868286493725e-06, + "loss": 0.4632, + "step": 23651 + }, + { + "epoch": 0.5206251375742901, + "grad_norm": 2.475731065744186, + "learning_rate": 4.908930431170425e-06, + "loss": 0.3382, + "step": 23652 + }, + { + "epoch": 0.5206471494607088, + "grad_norm": 2.2390048431512497, + "learning_rate": 4.908574034154336e-06, + "loss": 0.2722, + "step": 23653 + }, + { + "epoch": 0.5206691613471275, + "grad_norm": 0.6316869984036996, + "learning_rate": 4.908217637602915e-06, + "loss": 0.4898, + "step": 23654 + }, + { + "epoch": 0.5206911732335461, + "grad_norm": 1.6260004577981533, + "learning_rate": 4.90786124151797e-06, + "loss": 0.3536, + "step": 23655 + }, + { + "epoch": 0.5207131851199648, + "grad_norm": 1.5184270259323795, + "learning_rate": 4.907504845901317e-06, + "loss": 0.3345, + "step": 23656 + }, + { + "epoch": 0.5207351970063835, + "grad_norm": 2.018199972648707, + "learning_rate": 4.907148450754764e-06, + "loss": 0.3183, + "step": 23657 + }, + { + "epoch": 0.5207572088928021, + "grad_norm": 1.545807739546066, + "learning_rate": 4.906792056080126e-06, + "loss": 0.2929, + "step": 23658 + }, + { + "epoch": 0.5207792207792208, + "grad_norm": 1.6278341026731726, + "learning_rate": 4.906435661879212e-06, + "loss": 0.3176, + "step": 23659 + }, + { + "epoch": 0.5208012326656395, + "grad_norm": 1.4346020439852898, + "learning_rate": 4.906079268153833e-06, + "loss": 0.2808, + "step": 23660 + }, + { + "epoch": 0.5208232445520581, + "grad_norm": 1.4558714482489175, + "learning_rate": 4.905722874905801e-06, + "loss": 0.3624, + "step": 23661 + }, + { + "epoch": 0.5208452564384768, + "grad_norm": 1.268967687293123, + "learning_rate": 4.905366482136927e-06, + "loss": 0.3764, + "step": 23662 + }, + { + "epoch": 0.5208672683248955, + "grad_norm": 1.5477795755898913, + "learning_rate": 4.905010089849025e-06, + "loss": 0.3225, + "step": 23663 + }, + { + "epoch": 0.5208892802113141, + "grad_norm": 2.193572078671465, + "learning_rate": 4.904653698043903e-06, + "loss": 0.338, + "step": 23664 + }, + { + "epoch": 0.5209112920977328, + "grad_norm": 1.391504873021934, + "learning_rate": 4.904297306723372e-06, + "loss": 0.2871, + "step": 23665 + }, + { + "epoch": 0.5209333039841515, + "grad_norm": 1.334154620881478, + "learning_rate": 4.903940915889246e-06, + "loss": 0.2983, + "step": 23666 + }, + { + "epoch": 0.5209553158705701, + "grad_norm": 1.216964252672544, + "learning_rate": 4.903584525543336e-06, + "loss": 0.3371, + "step": 23667 + }, + { + "epoch": 0.5209773277569888, + "grad_norm": 1.4572145003763854, + "learning_rate": 4.90322813568745e-06, + "loss": 0.3007, + "step": 23668 + }, + { + "epoch": 0.5209993396434074, + "grad_norm": 1.6342486697402288, + "learning_rate": 4.902871746323403e-06, + "loss": 0.2642, + "step": 23669 + }, + { + "epoch": 0.5210213515298261, + "grad_norm": 1.4789306935664968, + "learning_rate": 4.902515357453005e-06, + "loss": 0.3598, + "step": 23670 + }, + { + "epoch": 0.5210433634162448, + "grad_norm": 1.8442315968504521, + "learning_rate": 4.902158969078068e-06, + "loss": 0.2669, + "step": 23671 + }, + { + "epoch": 0.5210653753026634, + "grad_norm": 1.06671519464018, + "learning_rate": 4.901802581200403e-06, + "loss": 0.192, + "step": 23672 + }, + { + "epoch": 0.5210873871890821, + "grad_norm": 1.2569422046452479, + "learning_rate": 4.90144619382182e-06, + "loss": 0.2981, + "step": 23673 + }, + { + "epoch": 0.5211093990755008, + "grad_norm": 1.665256896799242, + "learning_rate": 4.901089806944132e-06, + "loss": 0.3071, + "step": 23674 + }, + { + "epoch": 0.5211314109619194, + "grad_norm": 2.433160196807086, + "learning_rate": 4.900733420569149e-06, + "loss": 0.298, + "step": 23675 + }, + { + "epoch": 0.5211534228483381, + "grad_norm": 1.8533514275243161, + "learning_rate": 4.9003770346986845e-06, + "loss": 0.2908, + "step": 23676 + }, + { + "epoch": 0.5211754347347568, + "grad_norm": 1.2532032394070476, + "learning_rate": 4.900020649334548e-06, + "loss": 0.2746, + "step": 23677 + }, + { + "epoch": 0.5211974466211754, + "grad_norm": 1.5968029753587971, + "learning_rate": 4.899664264478549e-06, + "loss": 0.2608, + "step": 23678 + }, + { + "epoch": 0.5212194585075941, + "grad_norm": 1.974271978647757, + "learning_rate": 4.8993078801325035e-06, + "loss": 0.3163, + "step": 23679 + }, + { + "epoch": 0.5212414703940128, + "grad_norm": 1.8069968955816014, + "learning_rate": 4.8989514962982185e-06, + "loss": 0.3316, + "step": 23680 + }, + { + "epoch": 0.5212634822804314, + "grad_norm": 1.3387040368156489, + "learning_rate": 4.898595112977509e-06, + "loss": 0.2296, + "step": 23681 + }, + { + "epoch": 0.5212854941668501, + "grad_norm": 1.9414423212688838, + "learning_rate": 4.898238730172184e-06, + "loss": 0.3065, + "step": 23682 + }, + { + "epoch": 0.5213075060532688, + "grad_norm": 1.695285882276269, + "learning_rate": 4.897882347884052e-06, + "loss": 0.3446, + "step": 23683 + }, + { + "epoch": 0.5213295179396874, + "grad_norm": 0.6406794211248689, + "learning_rate": 4.897525966114931e-06, + "loss": 0.458, + "step": 23684 + }, + { + "epoch": 0.5213515298261061, + "grad_norm": 3.857382252919925, + "learning_rate": 4.897169584866627e-06, + "loss": 0.2832, + "step": 23685 + }, + { + "epoch": 0.5213735417125248, + "grad_norm": 2.685199401070445, + "learning_rate": 4.896813204140954e-06, + "loss": 0.2854, + "step": 23686 + }, + { + "epoch": 0.5213955535989434, + "grad_norm": 1.5849872580687783, + "learning_rate": 4.896456823939722e-06, + "loss": 0.3178, + "step": 23687 + }, + { + "epoch": 0.5214175654853621, + "grad_norm": 1.3697051074437299, + "learning_rate": 4.896100444264742e-06, + "loss": 0.2869, + "step": 23688 + }, + { + "epoch": 0.5214395773717808, + "grad_norm": 1.4175342862021754, + "learning_rate": 4.895744065117827e-06, + "loss": 0.3167, + "step": 23689 + }, + { + "epoch": 0.5214615892581994, + "grad_norm": 1.6345525365870333, + "learning_rate": 4.895387686500787e-06, + "loss": 0.255, + "step": 23690 + }, + { + "epoch": 0.5214836011446181, + "grad_norm": 0.6351757471700823, + "learning_rate": 4.895031308415433e-06, + "loss": 0.4781, + "step": 23691 + }, + { + "epoch": 0.5215056130310367, + "grad_norm": 1.729931633030047, + "learning_rate": 4.8946749308635774e-06, + "loss": 0.2981, + "step": 23692 + }, + { + "epoch": 0.5215276249174554, + "grad_norm": 9.084772320210737, + "learning_rate": 4.894318553847029e-06, + "loss": 0.2565, + "step": 23693 + }, + { + "epoch": 0.5215496368038741, + "grad_norm": 1.7384980534658219, + "learning_rate": 4.893962177367603e-06, + "loss": 0.3562, + "step": 23694 + }, + { + "epoch": 0.5215716486902927, + "grad_norm": 1.5888502304384453, + "learning_rate": 4.893605801427107e-06, + "loss": 0.2995, + "step": 23695 + }, + { + "epoch": 0.5215936605767114, + "grad_norm": 1.3983399039021907, + "learning_rate": 4.893249426027354e-06, + "loss": 0.2926, + "step": 23696 + }, + { + "epoch": 0.5216156724631301, + "grad_norm": 0.6182350079038997, + "learning_rate": 4.892893051170156e-06, + "loss": 0.4841, + "step": 23697 + }, + { + "epoch": 0.5216376843495487, + "grad_norm": 0.6207949512158221, + "learning_rate": 4.892536676857322e-06, + "loss": 0.4789, + "step": 23698 + }, + { + "epoch": 0.5216596962359674, + "grad_norm": 2.3673672185486527, + "learning_rate": 4.892180303090665e-06, + "loss": 0.2619, + "step": 23699 + }, + { + "epoch": 0.5216817081223861, + "grad_norm": 1.4364586139319333, + "learning_rate": 4.891823929871997e-06, + "loss": 0.3068, + "step": 23700 + }, + { + "epoch": 0.5217037200088047, + "grad_norm": 2.0403162286969336, + "learning_rate": 4.891467557203126e-06, + "loss": 0.3418, + "step": 23701 + }, + { + "epoch": 0.5217257318952234, + "grad_norm": 1.7803749627345467, + "learning_rate": 4.891111185085866e-06, + "loss": 0.3077, + "step": 23702 + }, + { + "epoch": 0.5217477437816421, + "grad_norm": 1.5593936479690564, + "learning_rate": 4.8907548135220265e-06, + "loss": 0.3318, + "step": 23703 + }, + { + "epoch": 0.5217697556680607, + "grad_norm": 2.212027307895078, + "learning_rate": 4.890398442513423e-06, + "loss": 0.3571, + "step": 23704 + }, + { + "epoch": 0.5217917675544794, + "grad_norm": 1.9050266558511857, + "learning_rate": 4.890042072061861e-06, + "loss": 0.4176, + "step": 23705 + }, + { + "epoch": 0.5218137794408981, + "grad_norm": 1.2986668297473776, + "learning_rate": 4.889685702169155e-06, + "loss": 0.3444, + "step": 23706 + }, + { + "epoch": 0.5218357913273167, + "grad_norm": 1.6241472628992648, + "learning_rate": 4.889329332837115e-06, + "loss": 0.3064, + "step": 23707 + }, + { + "epoch": 0.5218578032137354, + "grad_norm": 2.2530747847042254, + "learning_rate": 4.888972964067554e-06, + "loss": 0.4024, + "step": 23708 + }, + { + "epoch": 0.5218798151001541, + "grad_norm": 1.4809140476850484, + "learning_rate": 4.88861659586228e-06, + "loss": 0.2412, + "step": 23709 + }, + { + "epoch": 0.5219018269865727, + "grad_norm": 2.249554606775327, + "learning_rate": 4.888260228223106e-06, + "loss": 0.2686, + "step": 23710 + }, + { + "epoch": 0.5219238388729914, + "grad_norm": 1.7345121961436891, + "learning_rate": 4.887903861151844e-06, + "loss": 0.3338, + "step": 23711 + }, + { + "epoch": 0.5219458507594101, + "grad_norm": 1.5554458891433962, + "learning_rate": 4.887547494650307e-06, + "loss": 0.3389, + "step": 23712 + }, + { + "epoch": 0.5219678626458287, + "grad_norm": 0.6345982252198944, + "learning_rate": 4.887191128720302e-06, + "loss": 0.4768, + "step": 23713 + }, + { + "epoch": 0.5219898745322474, + "grad_norm": 2.445234458828091, + "learning_rate": 4.88683476336364e-06, + "loss": 0.3049, + "step": 23714 + }, + { + "epoch": 0.522011886418666, + "grad_norm": 3.5565465929808298, + "learning_rate": 4.8864783985821364e-06, + "loss": 0.3713, + "step": 23715 + }, + { + "epoch": 0.5220338983050847, + "grad_norm": 2.052563391049091, + "learning_rate": 4.886122034377599e-06, + "loss": 0.3595, + "step": 23716 + }, + { + "epoch": 0.5220559101915034, + "grad_norm": 1.7918723307179423, + "learning_rate": 4.885765670751842e-06, + "loss": 0.3516, + "step": 23717 + }, + { + "epoch": 0.522077922077922, + "grad_norm": 1.6819397206400397, + "learning_rate": 4.8854093077066735e-06, + "loss": 0.2844, + "step": 23718 + }, + { + "epoch": 0.5220999339643407, + "grad_norm": 1.875980110923973, + "learning_rate": 4.885052945243906e-06, + "loss": 0.3729, + "step": 23719 + }, + { + "epoch": 0.5221219458507594, + "grad_norm": 1.5995023248013962, + "learning_rate": 4.884696583365351e-06, + "loss": 0.3005, + "step": 23720 + }, + { + "epoch": 0.522143957737178, + "grad_norm": 2.203836618659774, + "learning_rate": 4.884340222072818e-06, + "loss": 0.2742, + "step": 23721 + }, + { + "epoch": 0.5221659696235967, + "grad_norm": 1.2385797840277535, + "learning_rate": 4.883983861368122e-06, + "loss": 0.2801, + "step": 23722 + }, + { + "epoch": 0.5221879815100154, + "grad_norm": 2.5891815694592677, + "learning_rate": 4.883627501253071e-06, + "loss": 0.378, + "step": 23723 + }, + { + "epoch": 0.522209993396434, + "grad_norm": 1.2043184151778685, + "learning_rate": 4.8832711417294755e-06, + "loss": 0.2375, + "step": 23724 + }, + { + "epoch": 0.5222320052828527, + "grad_norm": 1.6321071884242628, + "learning_rate": 4.882914782799149e-06, + "loss": 0.3285, + "step": 23725 + }, + { + "epoch": 0.5222540171692714, + "grad_norm": 1.5472883303920009, + "learning_rate": 4.882558424463902e-06, + "loss": 0.4, + "step": 23726 + }, + { + "epoch": 0.52227602905569, + "grad_norm": 2.2245010479825975, + "learning_rate": 4.8822020667255465e-06, + "loss": 0.2704, + "step": 23727 + }, + { + "epoch": 0.5222980409421087, + "grad_norm": 1.269978964746333, + "learning_rate": 4.881845709585892e-06, + "loss": 0.3106, + "step": 23728 + }, + { + "epoch": 0.5223200528285274, + "grad_norm": 1.4075294189891199, + "learning_rate": 4.881489353046749e-06, + "loss": 0.3036, + "step": 23729 + }, + { + "epoch": 0.522342064714946, + "grad_norm": 1.4754901443279735, + "learning_rate": 4.881132997109932e-06, + "loss": 0.3286, + "step": 23730 + }, + { + "epoch": 0.5223640766013647, + "grad_norm": 1.627177089698097, + "learning_rate": 4.880776641777249e-06, + "loss": 0.2793, + "step": 23731 + }, + { + "epoch": 0.5223860884877835, + "grad_norm": 0.5967843232329211, + "learning_rate": 4.8804202870505114e-06, + "loss": 0.4741, + "step": 23732 + }, + { + "epoch": 0.522408100374202, + "grad_norm": 1.505972013522304, + "learning_rate": 4.880063932931532e-06, + "loss": 0.344, + "step": 23733 + }, + { + "epoch": 0.5224301122606207, + "grad_norm": 2.98004893791876, + "learning_rate": 4.87970757942212e-06, + "loss": 0.3719, + "step": 23734 + }, + { + "epoch": 0.5224521241470395, + "grad_norm": 1.5119551945970158, + "learning_rate": 4.879351226524091e-06, + "loss": 0.2588, + "step": 23735 + }, + { + "epoch": 0.522474136033458, + "grad_norm": 1.3409525169631449, + "learning_rate": 4.878994874239251e-06, + "loss": 0.2761, + "step": 23736 + }, + { + "epoch": 0.5224961479198768, + "grad_norm": 1.385737104682687, + "learning_rate": 4.878638522569412e-06, + "loss": 0.3247, + "step": 23737 + }, + { + "epoch": 0.5225181598062953, + "grad_norm": 1.3373354090477643, + "learning_rate": 4.878282171516388e-06, + "loss": 0.3312, + "step": 23738 + }, + { + "epoch": 0.522540171692714, + "grad_norm": 1.4558892706331292, + "learning_rate": 4.877925821081986e-06, + "loss": 0.27, + "step": 23739 + }, + { + "epoch": 0.5225621835791328, + "grad_norm": 1.4902379815685762, + "learning_rate": 4.877569471268022e-06, + "loss": 0.2972, + "step": 23740 + }, + { + "epoch": 0.5225841954655513, + "grad_norm": 1.6909041171190173, + "learning_rate": 4.877213122076304e-06, + "loss": 0.3204, + "step": 23741 + }, + { + "epoch": 0.52260620735197, + "grad_norm": 1.315191262909157, + "learning_rate": 4.876856773508641e-06, + "loss": 0.2475, + "step": 23742 + }, + { + "epoch": 0.5226282192383888, + "grad_norm": 0.6169212916827346, + "learning_rate": 4.87650042556685e-06, + "loss": 0.4964, + "step": 23743 + }, + { + "epoch": 0.5226502311248074, + "grad_norm": 2.4567907058916036, + "learning_rate": 4.876144078252737e-06, + "loss": 0.322, + "step": 23744 + }, + { + "epoch": 0.5226722430112261, + "grad_norm": 0.6433961850294961, + "learning_rate": 4.8757877315681165e-06, + "loss": 0.4635, + "step": 23745 + }, + { + "epoch": 0.5226942548976448, + "grad_norm": 1.4367402769028552, + "learning_rate": 4.875431385514798e-06, + "loss": 0.3471, + "step": 23746 + }, + { + "epoch": 0.5227162667840634, + "grad_norm": 1.326697990950431, + "learning_rate": 4.875075040094591e-06, + "loss": 0.2141, + "step": 23747 + }, + { + "epoch": 0.5227382786704821, + "grad_norm": 0.5979894130778698, + "learning_rate": 4.874718695309311e-06, + "loss": 0.4731, + "step": 23748 + }, + { + "epoch": 0.5227602905569008, + "grad_norm": 1.7586294938559357, + "learning_rate": 4.874362351160765e-06, + "loss": 0.3566, + "step": 23749 + }, + { + "epoch": 0.5227823024433194, + "grad_norm": 0.5896344018523939, + "learning_rate": 4.874006007650764e-06, + "loss": 0.4684, + "step": 23750 + }, + { + "epoch": 0.5228043143297381, + "grad_norm": 1.2585132147991442, + "learning_rate": 4.873649664781124e-06, + "loss": 0.284, + "step": 23751 + }, + { + "epoch": 0.5228263262161568, + "grad_norm": 1.2737229807262664, + "learning_rate": 4.873293322553649e-06, + "loss": 0.3118, + "step": 23752 + }, + { + "epoch": 0.5228483381025754, + "grad_norm": 1.8394500466065407, + "learning_rate": 4.872936980970158e-06, + "loss": 0.3423, + "step": 23753 + }, + { + "epoch": 0.5228703499889941, + "grad_norm": 1.56084909402115, + "learning_rate": 4.8725806400324556e-06, + "loss": 0.3099, + "step": 23754 + }, + { + "epoch": 0.5228923618754128, + "grad_norm": 1.2901191570292179, + "learning_rate": 4.872224299742354e-06, + "loss": 0.3459, + "step": 23755 + }, + { + "epoch": 0.5229143737618314, + "grad_norm": 1.5241334732313991, + "learning_rate": 4.871867960101667e-06, + "loss": 0.3118, + "step": 23756 + }, + { + "epoch": 0.5229363856482501, + "grad_norm": 0.6000404096420637, + "learning_rate": 4.8715116211122035e-06, + "loss": 0.4677, + "step": 23757 + }, + { + "epoch": 0.5229583975346688, + "grad_norm": 1.6186182564235598, + "learning_rate": 4.871155282775777e-06, + "loss": 0.2365, + "step": 23758 + }, + { + "epoch": 0.5229804094210874, + "grad_norm": 0.6107517651220778, + "learning_rate": 4.870798945094196e-06, + "loss": 0.4855, + "step": 23759 + }, + { + "epoch": 0.5230024213075061, + "grad_norm": 1.3651655323883778, + "learning_rate": 4.870442608069272e-06, + "loss": 0.3336, + "step": 23760 + }, + { + "epoch": 0.5230244331939247, + "grad_norm": 1.7863207857427856, + "learning_rate": 4.8700862717028155e-06, + "loss": 0.3527, + "step": 23761 + }, + { + "epoch": 0.5230464450803434, + "grad_norm": 0.6191226284885416, + "learning_rate": 4.8697299359966384e-06, + "loss": 0.4795, + "step": 23762 + }, + { + "epoch": 0.5230684569667621, + "grad_norm": 0.5857879271380336, + "learning_rate": 4.869373600952555e-06, + "loss": 0.442, + "step": 23763 + }, + { + "epoch": 0.5230904688531807, + "grad_norm": 7.182104517714296, + "learning_rate": 4.8690172665723716e-06, + "loss": 0.2721, + "step": 23764 + }, + { + "epoch": 0.5231124807395994, + "grad_norm": 1.4596949599819589, + "learning_rate": 4.868660932857899e-06, + "loss": 0.2926, + "step": 23765 + }, + { + "epoch": 0.5231344926260181, + "grad_norm": 1.5925387853101671, + "learning_rate": 4.868304599810953e-06, + "loss": 0.3722, + "step": 23766 + }, + { + "epoch": 0.5231565045124367, + "grad_norm": 1.4004061423782113, + "learning_rate": 4.867948267433341e-06, + "loss": 0.2676, + "step": 23767 + }, + { + "epoch": 0.5231785163988554, + "grad_norm": 1.1476923276177162, + "learning_rate": 4.867591935726873e-06, + "loss": 0.1962, + "step": 23768 + }, + { + "epoch": 0.5232005282852741, + "grad_norm": 1.459074311851151, + "learning_rate": 4.867235604693363e-06, + "loss": 0.2703, + "step": 23769 + }, + { + "epoch": 0.5232225401716927, + "grad_norm": 1.4652365463253298, + "learning_rate": 4.86687927433462e-06, + "loss": 0.2752, + "step": 23770 + }, + { + "epoch": 0.5232445520581114, + "grad_norm": 1.6555057968900793, + "learning_rate": 4.866522944652458e-06, + "loss": 0.351, + "step": 23771 + }, + { + "epoch": 0.5232665639445301, + "grad_norm": 2.081000391916713, + "learning_rate": 4.866166615648685e-06, + "loss": 0.3037, + "step": 23772 + }, + { + "epoch": 0.5232885758309487, + "grad_norm": 1.4498158878861904, + "learning_rate": 4.865810287325111e-06, + "loss": 0.3024, + "step": 23773 + }, + { + "epoch": 0.5233105877173674, + "grad_norm": 1.8468646873459946, + "learning_rate": 4.865453959683551e-06, + "loss": 0.4157, + "step": 23774 + }, + { + "epoch": 0.5233325996037861, + "grad_norm": 1.2169329925045396, + "learning_rate": 4.865097632725813e-06, + "loss": 0.2491, + "step": 23775 + }, + { + "epoch": 0.5233546114902047, + "grad_norm": 2.0237338812746213, + "learning_rate": 4.8647413064537106e-06, + "loss": 0.2224, + "step": 23776 + }, + { + "epoch": 0.5233766233766234, + "grad_norm": 1.985765196264481, + "learning_rate": 4.864384980869052e-06, + "loss": 0.3055, + "step": 23777 + }, + { + "epoch": 0.5233986352630421, + "grad_norm": 1.6848274355806798, + "learning_rate": 4.86402865597365e-06, + "loss": 0.3294, + "step": 23778 + }, + { + "epoch": 0.5234206471494607, + "grad_norm": 0.6773863016607186, + "learning_rate": 4.863672331769314e-06, + "loss": 0.4986, + "step": 23779 + }, + { + "epoch": 0.5234426590358794, + "grad_norm": 1.4097107383264549, + "learning_rate": 4.863316008257856e-06, + "loss": 0.2354, + "step": 23780 + }, + { + "epoch": 0.5234646709222981, + "grad_norm": 2.2623014166477073, + "learning_rate": 4.862959685441089e-06, + "loss": 0.3459, + "step": 23781 + }, + { + "epoch": 0.5234866828087167, + "grad_norm": 1.5040625758019146, + "learning_rate": 4.862603363320821e-06, + "loss": 0.2999, + "step": 23782 + }, + { + "epoch": 0.5235086946951354, + "grad_norm": 3.8197494644372543, + "learning_rate": 4.862247041898862e-06, + "loss": 0.2414, + "step": 23783 + }, + { + "epoch": 0.523530706581554, + "grad_norm": 1.5524008125737028, + "learning_rate": 4.861890721177029e-06, + "loss": 0.2618, + "step": 23784 + }, + { + "epoch": 0.5235527184679727, + "grad_norm": 1.6372773984147322, + "learning_rate": 4.8615344011571235e-06, + "loss": 0.2947, + "step": 23785 + }, + { + "epoch": 0.5235747303543914, + "grad_norm": 1.397382799963277, + "learning_rate": 4.861178081840966e-06, + "loss": 0.3707, + "step": 23786 + }, + { + "epoch": 0.52359674224081, + "grad_norm": 3.1978205930896593, + "learning_rate": 4.860821763230364e-06, + "loss": 0.3497, + "step": 23787 + }, + { + "epoch": 0.5236187541272287, + "grad_norm": 6.7402524416181295, + "learning_rate": 4.860465445327124e-06, + "loss": 0.3433, + "step": 23788 + }, + { + "epoch": 0.5236407660136474, + "grad_norm": 1.6521307088837593, + "learning_rate": 4.860109128133065e-06, + "loss": 0.3609, + "step": 23789 + }, + { + "epoch": 0.523662777900066, + "grad_norm": 2.2644624114454057, + "learning_rate": 4.859752811649992e-06, + "loss": 0.2945, + "step": 23790 + }, + { + "epoch": 0.5236847897864847, + "grad_norm": 1.3698511198665535, + "learning_rate": 4.859396495879717e-06, + "loss": 0.3083, + "step": 23791 + }, + { + "epoch": 0.5237068016729034, + "grad_norm": 2.365433150536548, + "learning_rate": 4.859040180824052e-06, + "loss": 0.3497, + "step": 23792 + }, + { + "epoch": 0.523728813559322, + "grad_norm": 1.4260212147619105, + "learning_rate": 4.858683866484807e-06, + "loss": 0.3257, + "step": 23793 + }, + { + "epoch": 0.5237508254457407, + "grad_norm": 1.4593857310091471, + "learning_rate": 4.858327552863796e-06, + "loss": 0.388, + "step": 23794 + }, + { + "epoch": 0.5237728373321594, + "grad_norm": 1.4270757459315888, + "learning_rate": 4.857971239962827e-06, + "loss": 0.2813, + "step": 23795 + }, + { + "epoch": 0.523794849218578, + "grad_norm": 1.4691304397161113, + "learning_rate": 4.85761492778371e-06, + "loss": 0.3285, + "step": 23796 + }, + { + "epoch": 0.5238168611049967, + "grad_norm": 1.5545082758420974, + "learning_rate": 4.857258616328259e-06, + "loss": 0.3394, + "step": 23797 + }, + { + "epoch": 0.5238388729914154, + "grad_norm": 1.3907672246902134, + "learning_rate": 4.856902305598281e-06, + "loss": 0.3415, + "step": 23798 + }, + { + "epoch": 0.523860884877834, + "grad_norm": 1.3550192751459735, + "learning_rate": 4.856545995595592e-06, + "loss": 0.3426, + "step": 23799 + }, + { + "epoch": 0.5238828967642527, + "grad_norm": 1.3324057043714592, + "learning_rate": 4.856189686321999e-06, + "loss": 0.2966, + "step": 23800 + }, + { + "epoch": 0.5239049086506714, + "grad_norm": 1.7750195606390626, + "learning_rate": 4.855833377779314e-06, + "loss": 0.3835, + "step": 23801 + }, + { + "epoch": 0.52392692053709, + "grad_norm": 2.0885581022909205, + "learning_rate": 4.85547706996935e-06, + "loss": 0.309, + "step": 23802 + }, + { + "epoch": 0.5239489324235087, + "grad_norm": 1.8003946199177825, + "learning_rate": 4.855120762893912e-06, + "loss": 0.3204, + "step": 23803 + }, + { + "epoch": 0.5239709443099274, + "grad_norm": 1.5453808417384853, + "learning_rate": 4.854764456554818e-06, + "loss": 0.265, + "step": 23804 + }, + { + "epoch": 0.523992956196346, + "grad_norm": 1.4293096504932956, + "learning_rate": 4.854408150953876e-06, + "loss": 0.3448, + "step": 23805 + }, + { + "epoch": 0.5240149680827647, + "grad_norm": 2.234718622599445, + "learning_rate": 4.854051846092894e-06, + "loss": 0.2917, + "step": 23806 + }, + { + "epoch": 0.5240369799691834, + "grad_norm": 1.365288071192899, + "learning_rate": 4.853695541973689e-06, + "loss": 0.2865, + "step": 23807 + }, + { + "epoch": 0.524058991855602, + "grad_norm": 0.6308076573606796, + "learning_rate": 4.853339238598067e-06, + "loss": 0.4792, + "step": 23808 + }, + { + "epoch": 0.5240810037420207, + "grad_norm": 1.9582181750476382, + "learning_rate": 4.852982935967839e-06, + "loss": 0.238, + "step": 23809 + }, + { + "epoch": 0.5241030156284393, + "grad_norm": 1.4840900802188008, + "learning_rate": 4.852626634084819e-06, + "loss": 0.3635, + "step": 23810 + }, + { + "epoch": 0.524125027514858, + "grad_norm": 1.3232781824165687, + "learning_rate": 4.8522703329508145e-06, + "loss": 0.2827, + "step": 23811 + }, + { + "epoch": 0.5241470394012767, + "grad_norm": 1.0752747406071146, + "learning_rate": 4.851914032567641e-06, + "loss": 0.2302, + "step": 23812 + }, + { + "epoch": 0.5241690512876953, + "grad_norm": 1.3214045328347237, + "learning_rate": 4.8515577329371046e-06, + "loss": 0.3472, + "step": 23813 + }, + { + "epoch": 0.524191063174114, + "grad_norm": 1.39523102608845, + "learning_rate": 4.851201434061018e-06, + "loss": 0.3047, + "step": 23814 + }, + { + "epoch": 0.5242130750605327, + "grad_norm": 2.8366209187593876, + "learning_rate": 4.8508451359411935e-06, + "loss": 0.3771, + "step": 23815 + }, + { + "epoch": 0.5242350869469513, + "grad_norm": 1.5127340465405303, + "learning_rate": 4.8504888385794365e-06, + "loss": 0.3463, + "step": 23816 + }, + { + "epoch": 0.52425709883337, + "grad_norm": 1.3232636627417045, + "learning_rate": 4.850132541977566e-06, + "loss": 0.3515, + "step": 23817 + }, + { + "epoch": 0.5242791107197887, + "grad_norm": 1.7458984356945508, + "learning_rate": 4.849776246137388e-06, + "loss": 0.4055, + "step": 23818 + }, + { + "epoch": 0.5243011226062073, + "grad_norm": 1.2275566969279903, + "learning_rate": 4.8494199510607125e-06, + "loss": 0.2884, + "step": 23819 + }, + { + "epoch": 0.524323134492626, + "grad_norm": 1.2668001362929937, + "learning_rate": 4.849063656749354e-06, + "loss": 0.2685, + "step": 23820 + }, + { + "epoch": 0.5243451463790447, + "grad_norm": 2.092282852737576, + "learning_rate": 4.848707363205119e-06, + "loss": 0.2815, + "step": 23821 + }, + { + "epoch": 0.5243671582654633, + "grad_norm": 1.5189543848784868, + "learning_rate": 4.848351070429823e-06, + "loss": 0.2807, + "step": 23822 + }, + { + "epoch": 0.524389170151882, + "grad_norm": 1.2621501984965884, + "learning_rate": 4.847994778425274e-06, + "loss": 0.3035, + "step": 23823 + }, + { + "epoch": 0.5244111820383007, + "grad_norm": 1.2689361830648362, + "learning_rate": 4.847638487193282e-06, + "loss": 0.2166, + "step": 23824 + }, + { + "epoch": 0.5244331939247193, + "grad_norm": 0.6233633950199627, + "learning_rate": 4.847282196735661e-06, + "loss": 0.4963, + "step": 23825 + }, + { + "epoch": 0.524455205811138, + "grad_norm": 0.6304220913854228, + "learning_rate": 4.846925907054218e-06, + "loss": 0.5078, + "step": 23826 + }, + { + "epoch": 0.5244772176975567, + "grad_norm": 1.4271832483900833, + "learning_rate": 4.846569618150767e-06, + "loss": 0.2601, + "step": 23827 + }, + { + "epoch": 0.5244992295839753, + "grad_norm": 1.9311718109023948, + "learning_rate": 4.846213330027117e-06, + "loss": 0.3306, + "step": 23828 + }, + { + "epoch": 0.524521241470394, + "grad_norm": 0.5821730294159553, + "learning_rate": 4.84585704268508e-06, + "loss": 0.476, + "step": 23829 + }, + { + "epoch": 0.5245432533568127, + "grad_norm": 2.5262521871366137, + "learning_rate": 4.845500756126467e-06, + "loss": 0.2748, + "step": 23830 + }, + { + "epoch": 0.5245652652432313, + "grad_norm": 1.444044504865074, + "learning_rate": 4.8451444703530875e-06, + "loss": 0.274, + "step": 23831 + }, + { + "epoch": 0.52458727712965, + "grad_norm": 1.6434717502593592, + "learning_rate": 4.844788185366751e-06, + "loss": 0.3568, + "step": 23832 + }, + { + "epoch": 0.5246092890160686, + "grad_norm": 1.4291332176628886, + "learning_rate": 4.844431901169274e-06, + "loss": 0.3528, + "step": 23833 + }, + { + "epoch": 0.5246313009024873, + "grad_norm": 2.78577860841011, + "learning_rate": 4.844075617762459e-06, + "loss": 0.2774, + "step": 23834 + }, + { + "epoch": 0.524653312788906, + "grad_norm": 0.640542772041113, + "learning_rate": 4.843719335148125e-06, + "loss": 0.4515, + "step": 23835 + }, + { + "epoch": 0.5246753246753246, + "grad_norm": 0.6307800256696929, + "learning_rate": 4.843363053328078e-06, + "loss": 0.4564, + "step": 23836 + }, + { + "epoch": 0.5246973365617433, + "grad_norm": 3.5918965120475645, + "learning_rate": 4.843006772304128e-06, + "loss": 0.3756, + "step": 23837 + }, + { + "epoch": 0.524719348448162, + "grad_norm": 1.628770660447093, + "learning_rate": 4.8426504920780895e-06, + "loss": 0.3234, + "step": 23838 + }, + { + "epoch": 0.5247413603345806, + "grad_norm": 1.2367083726393602, + "learning_rate": 4.842294212651769e-06, + "loss": 0.3459, + "step": 23839 + }, + { + "epoch": 0.5247633722209993, + "grad_norm": 1.8331962296517936, + "learning_rate": 4.841937934026983e-06, + "loss": 0.2383, + "step": 23840 + }, + { + "epoch": 0.524785384107418, + "grad_norm": 1.9419416259270073, + "learning_rate": 4.8415816562055375e-06, + "loss": 0.2914, + "step": 23841 + }, + { + "epoch": 0.5248073959938366, + "grad_norm": 1.2253896416924117, + "learning_rate": 4.8412253791892434e-06, + "loss": 0.2449, + "step": 23842 + }, + { + "epoch": 0.5248294078802553, + "grad_norm": 0.6474486454407601, + "learning_rate": 4.840869102979915e-06, + "loss": 0.464, + "step": 23843 + }, + { + "epoch": 0.524851419766674, + "grad_norm": 0.6607044594413319, + "learning_rate": 4.840512827579359e-06, + "loss": 0.4904, + "step": 23844 + }, + { + "epoch": 0.5248734316530926, + "grad_norm": 1.7711604475693319, + "learning_rate": 4.840156552989389e-06, + "loss": 0.3139, + "step": 23845 + }, + { + "epoch": 0.5248954435395113, + "grad_norm": 0.6337075000703553, + "learning_rate": 4.839800279211815e-06, + "loss": 0.5086, + "step": 23846 + }, + { + "epoch": 0.52491745542593, + "grad_norm": 1.6479948613844997, + "learning_rate": 4.839444006248445e-06, + "loss": 0.3953, + "step": 23847 + }, + { + "epoch": 0.5249394673123486, + "grad_norm": 2.346543402575773, + "learning_rate": 4.839087734101096e-06, + "loss": 0.2894, + "step": 23848 + }, + { + "epoch": 0.5249614791987673, + "grad_norm": 1.5486571768989488, + "learning_rate": 4.838731462771572e-06, + "loss": 0.3305, + "step": 23849 + }, + { + "epoch": 0.524983491085186, + "grad_norm": 0.6068407145520532, + "learning_rate": 4.838375192261687e-06, + "loss": 0.478, + "step": 23850 + }, + { + "epoch": 0.5250055029716046, + "grad_norm": 1.87743143942802, + "learning_rate": 4.838018922573253e-06, + "loss": 0.3018, + "step": 23851 + }, + { + "epoch": 0.5250275148580233, + "grad_norm": 1.7379893900527166, + "learning_rate": 4.8376626537080766e-06, + "loss": 0.3168, + "step": 23852 + }, + { + "epoch": 0.525049526744442, + "grad_norm": 1.2788653229251303, + "learning_rate": 4.837306385667973e-06, + "loss": 0.3128, + "step": 23853 + }, + { + "epoch": 0.5250715386308606, + "grad_norm": 1.6865904980285715, + "learning_rate": 4.836950118454751e-06, + "loss": 0.3415, + "step": 23854 + }, + { + "epoch": 0.5250935505172794, + "grad_norm": 0.6355042749128451, + "learning_rate": 4.83659385207022e-06, + "loss": 0.4858, + "step": 23855 + }, + { + "epoch": 0.5251155624036979, + "grad_norm": 1.5157298263060903, + "learning_rate": 4.836237586516192e-06, + "loss": 0.3254, + "step": 23856 + }, + { + "epoch": 0.5251375742901166, + "grad_norm": 0.6380544875445201, + "learning_rate": 4.835881321794476e-06, + "loss": 0.504, + "step": 23857 + }, + { + "epoch": 0.5251595861765354, + "grad_norm": 1.4389226492843727, + "learning_rate": 4.835525057906889e-06, + "loss": 0.3017, + "step": 23858 + }, + { + "epoch": 0.525181598062954, + "grad_norm": 0.6156990255502411, + "learning_rate": 4.835168794855234e-06, + "loss": 0.4704, + "step": 23859 + }, + { + "epoch": 0.5252036099493727, + "grad_norm": 1.2064217613605428, + "learning_rate": 4.834812532641323e-06, + "loss": 0.2453, + "step": 23860 + }, + { + "epoch": 0.5252256218357914, + "grad_norm": 1.6047410426663533, + "learning_rate": 4.834456271266973e-06, + "loss": 0.2702, + "step": 23861 + }, + { + "epoch": 0.52524763372221, + "grad_norm": 0.6292772828025543, + "learning_rate": 4.834100010733986e-06, + "loss": 0.442, + "step": 23862 + }, + { + "epoch": 0.5252696456086287, + "grad_norm": 0.6157875885243814, + "learning_rate": 4.833743751044179e-06, + "loss": 0.4763, + "step": 23863 + }, + { + "epoch": 0.5252916574950474, + "grad_norm": 2.2096324027911103, + "learning_rate": 4.8333874921993595e-06, + "loss": 0.2971, + "step": 23864 + }, + { + "epoch": 0.525313669381466, + "grad_norm": 1.184202071778023, + "learning_rate": 4.833031234201338e-06, + "loss": 0.2569, + "step": 23865 + }, + { + "epoch": 0.5253356812678847, + "grad_norm": 1.7592195772271673, + "learning_rate": 4.832674977051929e-06, + "loss": 0.3114, + "step": 23866 + }, + { + "epoch": 0.5253576931543034, + "grad_norm": 1.3288688114971015, + "learning_rate": 4.832318720752939e-06, + "loss": 0.2958, + "step": 23867 + }, + { + "epoch": 0.525379705040722, + "grad_norm": 1.1251035221133803, + "learning_rate": 4.831962465306179e-06, + "loss": 0.3023, + "step": 23868 + }, + { + "epoch": 0.5254017169271407, + "grad_norm": 1.4089498395386386, + "learning_rate": 4.8316062107134625e-06, + "loss": 0.2535, + "step": 23869 + }, + { + "epoch": 0.5254237288135594, + "grad_norm": 1.8168990596847683, + "learning_rate": 4.831249956976596e-06, + "loss": 0.2784, + "step": 23870 + }, + { + "epoch": 0.525445740699978, + "grad_norm": 1.3966459428733544, + "learning_rate": 4.830893704097395e-06, + "loss": 0.3843, + "step": 23871 + }, + { + "epoch": 0.5254677525863967, + "grad_norm": 1.1929974824321619, + "learning_rate": 4.830537452077667e-06, + "loss": 0.3279, + "step": 23872 + }, + { + "epoch": 0.5254897644728154, + "grad_norm": 1.477899818893515, + "learning_rate": 4.830181200919221e-06, + "loss": 0.3036, + "step": 23873 + }, + { + "epoch": 0.525511776359234, + "grad_norm": 1.49985367613149, + "learning_rate": 4.829824950623873e-06, + "loss": 0.3254, + "step": 23874 + }, + { + "epoch": 0.5255337882456527, + "grad_norm": 1.3992987466009073, + "learning_rate": 4.8294687011934275e-06, + "loss": 0.3223, + "step": 23875 + }, + { + "epoch": 0.5255558001320714, + "grad_norm": 1.2256263969439043, + "learning_rate": 4.829112452629701e-06, + "loss": 0.3129, + "step": 23876 + }, + { + "epoch": 0.52557781201849, + "grad_norm": 1.2429619847257372, + "learning_rate": 4.8287562049345e-06, + "loss": 0.289, + "step": 23877 + }, + { + "epoch": 0.5255998239049087, + "grad_norm": 1.7183579269974967, + "learning_rate": 4.828399958109636e-06, + "loss": 0.3155, + "step": 23878 + }, + { + "epoch": 0.5256218357913273, + "grad_norm": 1.4783452171649645, + "learning_rate": 4.828043712156921e-06, + "loss": 0.312, + "step": 23879 + }, + { + "epoch": 0.525643847677746, + "grad_norm": 1.7069791737314661, + "learning_rate": 4.827687467078163e-06, + "loss": 0.3035, + "step": 23880 + }, + { + "epoch": 0.5256658595641647, + "grad_norm": 1.826041448189267, + "learning_rate": 4.827331222875175e-06, + "loss": 0.4111, + "step": 23881 + }, + { + "epoch": 0.5256878714505833, + "grad_norm": 1.3758936831932302, + "learning_rate": 4.826974979549767e-06, + "loss": 0.3533, + "step": 23882 + }, + { + "epoch": 0.525709883337002, + "grad_norm": 1.5477330420544038, + "learning_rate": 4.826618737103748e-06, + "loss": 0.3295, + "step": 23883 + }, + { + "epoch": 0.5257318952234207, + "grad_norm": 2.093209540861194, + "learning_rate": 4.826262495538932e-06, + "loss": 0.3638, + "step": 23884 + }, + { + "epoch": 0.5257539071098393, + "grad_norm": 1.5698871692978185, + "learning_rate": 4.825906254857126e-06, + "loss": 0.271, + "step": 23885 + }, + { + "epoch": 0.525775918996258, + "grad_norm": 0.6656939126790468, + "learning_rate": 4.825550015060143e-06, + "loss": 0.5026, + "step": 23886 + }, + { + "epoch": 0.5257979308826767, + "grad_norm": 0.6500131941227658, + "learning_rate": 4.825193776149794e-06, + "loss": 0.4801, + "step": 23887 + }, + { + "epoch": 0.5258199427690953, + "grad_norm": 1.5715337216224967, + "learning_rate": 4.824837538127884e-06, + "loss": 0.324, + "step": 23888 + }, + { + "epoch": 0.525841954655514, + "grad_norm": 2.3940550031409678, + "learning_rate": 4.824481300996231e-06, + "loss": 0.3346, + "step": 23889 + }, + { + "epoch": 0.5258639665419327, + "grad_norm": 1.2038513461373144, + "learning_rate": 4.824125064756641e-06, + "loss": 0.2416, + "step": 23890 + }, + { + "epoch": 0.5258859784283513, + "grad_norm": 1.4835364140094103, + "learning_rate": 4.823768829410926e-06, + "loss": 0.2947, + "step": 23891 + }, + { + "epoch": 0.52590799031477, + "grad_norm": 0.614559358003832, + "learning_rate": 4.823412594960898e-06, + "loss": 0.4802, + "step": 23892 + }, + { + "epoch": 0.5259300022011887, + "grad_norm": 0.5917371111037811, + "learning_rate": 4.823056361408362e-06, + "loss": 0.4612, + "step": 23893 + }, + { + "epoch": 0.5259520140876073, + "grad_norm": 0.5986610128711651, + "learning_rate": 4.822700128755137e-06, + "loss": 0.4815, + "step": 23894 + }, + { + "epoch": 0.525974025974026, + "grad_norm": 1.2085994784085061, + "learning_rate": 4.822343897003027e-06, + "loss": 0.3179, + "step": 23895 + }, + { + "epoch": 0.5259960378604447, + "grad_norm": 1.2387877592228547, + "learning_rate": 4.821987666153844e-06, + "loss": 0.3007, + "step": 23896 + }, + { + "epoch": 0.5260180497468633, + "grad_norm": 0.616278047546431, + "learning_rate": 4.821631436209402e-06, + "loss": 0.489, + "step": 23897 + }, + { + "epoch": 0.526040061633282, + "grad_norm": 1.5181416983521552, + "learning_rate": 4.821275207171505e-06, + "loss": 0.3301, + "step": 23898 + }, + { + "epoch": 0.5260620735197007, + "grad_norm": 0.6047231693620931, + "learning_rate": 4.820918979041968e-06, + "loss": 0.4374, + "step": 23899 + }, + { + "epoch": 0.5260840854061193, + "grad_norm": 1.6889953038004655, + "learning_rate": 4.820562751822601e-06, + "loss": 0.4218, + "step": 23900 + }, + { + "epoch": 0.526106097292538, + "grad_norm": 2.197457276387712, + "learning_rate": 4.820206525515214e-06, + "loss": 0.4354, + "step": 23901 + }, + { + "epoch": 0.5261281091789566, + "grad_norm": 2.0430656347679648, + "learning_rate": 4.819850300121619e-06, + "loss": 0.4206, + "step": 23902 + }, + { + "epoch": 0.5261501210653753, + "grad_norm": 0.6224162411585503, + "learning_rate": 4.819494075643623e-06, + "loss": 0.476, + "step": 23903 + }, + { + "epoch": 0.526172132951794, + "grad_norm": 2.772446227823127, + "learning_rate": 4.81913785208304e-06, + "loss": 0.2407, + "step": 23904 + }, + { + "epoch": 0.5261941448382126, + "grad_norm": 1.4142308948646871, + "learning_rate": 4.81878162944168e-06, + "loss": 0.3564, + "step": 23905 + }, + { + "epoch": 0.5262161567246313, + "grad_norm": 1.9645378275824728, + "learning_rate": 4.818425407721349e-06, + "loss": 0.3094, + "step": 23906 + }, + { + "epoch": 0.52623816861105, + "grad_norm": 2.140216881151334, + "learning_rate": 4.818069186923865e-06, + "loss": 0.3096, + "step": 23907 + }, + { + "epoch": 0.5262601804974686, + "grad_norm": 3.9489265561580824, + "learning_rate": 4.817712967051033e-06, + "loss": 0.4138, + "step": 23908 + }, + { + "epoch": 0.5262821923838873, + "grad_norm": 1.4524862728674164, + "learning_rate": 4.817356748104664e-06, + "loss": 0.286, + "step": 23909 + }, + { + "epoch": 0.526304204270306, + "grad_norm": 1.5146830985695645, + "learning_rate": 4.817000530086572e-06, + "loss": 0.2786, + "step": 23910 + }, + { + "epoch": 0.5263262161567246, + "grad_norm": 1.3544827871489493, + "learning_rate": 4.816644312998561e-06, + "loss": 0.344, + "step": 23911 + }, + { + "epoch": 0.5263482280431433, + "grad_norm": 1.5295545044236967, + "learning_rate": 4.816288096842449e-06, + "loss": 0.3825, + "step": 23912 + }, + { + "epoch": 0.526370239929562, + "grad_norm": 4.48965273360613, + "learning_rate": 4.8159318816200415e-06, + "loss": 0.2506, + "step": 23913 + }, + { + "epoch": 0.5263922518159806, + "grad_norm": 1.8779221293536155, + "learning_rate": 4.815575667333149e-06, + "loss": 0.352, + "step": 23914 + }, + { + "epoch": 0.5264142637023993, + "grad_norm": 9.746079244354611, + "learning_rate": 4.815219453983586e-06, + "loss": 0.2966, + "step": 23915 + }, + { + "epoch": 0.526436275588818, + "grad_norm": 1.5817272104916391, + "learning_rate": 4.814863241573158e-06, + "loss": 0.3359, + "step": 23916 + }, + { + "epoch": 0.5264582874752366, + "grad_norm": 1.7664687844871365, + "learning_rate": 4.814507030103678e-06, + "loss": 0.3552, + "step": 23917 + }, + { + "epoch": 0.5264802993616553, + "grad_norm": 1.3132542829894667, + "learning_rate": 4.8141508195769565e-06, + "loss": 0.2884, + "step": 23918 + }, + { + "epoch": 0.526502311248074, + "grad_norm": 0.6535242433067143, + "learning_rate": 4.813794609994802e-06, + "loss": 0.4857, + "step": 23919 + }, + { + "epoch": 0.5265243231344926, + "grad_norm": 0.6213743948696061, + "learning_rate": 4.813438401359029e-06, + "loss": 0.4919, + "step": 23920 + }, + { + "epoch": 0.5265463350209113, + "grad_norm": 1.0851419914482612, + "learning_rate": 4.813082193671443e-06, + "loss": 0.2792, + "step": 23921 + }, + { + "epoch": 0.52656834690733, + "grad_norm": 1.3533348857154046, + "learning_rate": 4.812725986933857e-06, + "loss": 0.3527, + "step": 23922 + }, + { + "epoch": 0.5265903587937486, + "grad_norm": 1.5512786935067069, + "learning_rate": 4.812369781148083e-06, + "loss": 0.2762, + "step": 23923 + }, + { + "epoch": 0.5266123706801673, + "grad_norm": 1.238081000621937, + "learning_rate": 4.812013576315926e-06, + "loss": 0.3484, + "step": 23924 + }, + { + "epoch": 0.5266343825665859, + "grad_norm": 1.4725748733301813, + "learning_rate": 4.811657372439203e-06, + "loss": 0.2639, + "step": 23925 + }, + { + "epoch": 0.5266563944530046, + "grad_norm": 1.4160791039846612, + "learning_rate": 4.8113011695197186e-06, + "loss": 0.3408, + "step": 23926 + }, + { + "epoch": 0.5266784063394233, + "grad_norm": 1.955728046178194, + "learning_rate": 4.8109449675592875e-06, + "loss": 0.3056, + "step": 23927 + }, + { + "epoch": 0.5267004182258419, + "grad_norm": 1.5073273249608596, + "learning_rate": 4.81058876655972e-06, + "loss": 0.4215, + "step": 23928 + }, + { + "epoch": 0.5267224301122606, + "grad_norm": 1.630692488606046, + "learning_rate": 4.810232566522822e-06, + "loss": 0.3275, + "step": 23929 + }, + { + "epoch": 0.5267444419986793, + "grad_norm": 1.2885839961583097, + "learning_rate": 4.809876367450408e-06, + "loss": 0.266, + "step": 23930 + }, + { + "epoch": 0.5267664538850979, + "grad_norm": 1.5079450917691204, + "learning_rate": 4.809520169344287e-06, + "loss": 0.3789, + "step": 23931 + }, + { + "epoch": 0.5267884657715166, + "grad_norm": 0.6776965099225932, + "learning_rate": 4.8091639722062676e-06, + "loss": 0.4775, + "step": 23932 + }, + { + "epoch": 0.5268104776579353, + "grad_norm": 1.5354137841990056, + "learning_rate": 4.8088077760381655e-06, + "loss": 0.3761, + "step": 23933 + }, + { + "epoch": 0.5268324895443539, + "grad_norm": 1.7224342237829464, + "learning_rate": 4.808451580841785e-06, + "loss": 0.3779, + "step": 23934 + }, + { + "epoch": 0.5268545014307726, + "grad_norm": 1.4539789460067478, + "learning_rate": 4.8080953866189394e-06, + "loss": 0.2834, + "step": 23935 + }, + { + "epoch": 0.5268765133171913, + "grad_norm": 0.6363806167290673, + "learning_rate": 4.80773919337144e-06, + "loss": 0.4887, + "step": 23936 + }, + { + "epoch": 0.5268985252036099, + "grad_norm": 3.705689735270932, + "learning_rate": 4.807383001101093e-06, + "loss": 0.3258, + "step": 23937 + }, + { + "epoch": 0.5269205370900286, + "grad_norm": 1.1628261564781148, + "learning_rate": 4.807026809809714e-06, + "loss": 0.2674, + "step": 23938 + }, + { + "epoch": 0.5269425489764473, + "grad_norm": 0.6248774744260324, + "learning_rate": 4.806670619499109e-06, + "loss": 0.486, + "step": 23939 + }, + { + "epoch": 0.5269645608628659, + "grad_norm": 1.9070419495226443, + "learning_rate": 4.806314430171092e-06, + "loss": 0.3124, + "step": 23940 + }, + { + "epoch": 0.5269865727492846, + "grad_norm": 1.2032795384339536, + "learning_rate": 4.805958241827472e-06, + "loss": 0.338, + "step": 23941 + }, + { + "epoch": 0.5270085846357033, + "grad_norm": 1.5293471582824345, + "learning_rate": 4.805602054470055e-06, + "loss": 0.3437, + "step": 23942 + }, + { + "epoch": 0.5270305965221219, + "grad_norm": 2.360893091707528, + "learning_rate": 4.805245868100659e-06, + "loss": 0.3638, + "step": 23943 + }, + { + "epoch": 0.5270526084085406, + "grad_norm": 1.406605590676683, + "learning_rate": 4.8048896827210885e-06, + "loss": 0.2028, + "step": 23944 + }, + { + "epoch": 0.5270746202949593, + "grad_norm": 1.8270707148322194, + "learning_rate": 4.804533498333156e-06, + "loss": 0.3521, + "step": 23945 + }, + { + "epoch": 0.5270966321813779, + "grad_norm": 1.3273495941526257, + "learning_rate": 4.804177314938673e-06, + "loss": 0.3955, + "step": 23946 + }, + { + "epoch": 0.5271186440677966, + "grad_norm": 1.4328858245780982, + "learning_rate": 4.803821132539445e-06, + "loss": 0.3105, + "step": 23947 + }, + { + "epoch": 0.5271406559542153, + "grad_norm": 1.2063179977316087, + "learning_rate": 4.803464951137287e-06, + "loss": 0.2981, + "step": 23948 + }, + { + "epoch": 0.5271626678406339, + "grad_norm": 1.887991870681517, + "learning_rate": 4.803108770734008e-06, + "loss": 0.4476, + "step": 23949 + }, + { + "epoch": 0.5271846797270526, + "grad_norm": 1.5289857165868053, + "learning_rate": 4.802752591331417e-06, + "loss": 0.2457, + "step": 23950 + }, + { + "epoch": 0.5272066916134712, + "grad_norm": 1.4081025576814887, + "learning_rate": 4.802396412931328e-06, + "loss": 0.336, + "step": 23951 + }, + { + "epoch": 0.5272287034998899, + "grad_norm": 1.1137831520553805, + "learning_rate": 4.802040235535546e-06, + "loss": 0.2813, + "step": 23952 + }, + { + "epoch": 0.5272507153863086, + "grad_norm": 1.656454418752272, + "learning_rate": 4.801684059145885e-06, + "loss": 0.3207, + "step": 23953 + }, + { + "epoch": 0.5272727272727272, + "grad_norm": 2.2248745860759533, + "learning_rate": 4.801327883764154e-06, + "loss": 0.2866, + "step": 23954 + }, + { + "epoch": 0.5272947391591459, + "grad_norm": 0.6225028225663273, + "learning_rate": 4.800971709392162e-06, + "loss": 0.4529, + "step": 23955 + }, + { + "epoch": 0.5273167510455646, + "grad_norm": 1.7694417475041424, + "learning_rate": 4.800615536031723e-06, + "loss": 0.2995, + "step": 23956 + }, + { + "epoch": 0.5273387629319832, + "grad_norm": 0.6372540547785067, + "learning_rate": 4.800259363684642e-06, + "loss": 0.4733, + "step": 23957 + }, + { + "epoch": 0.5273607748184019, + "grad_norm": 1.2138545360909492, + "learning_rate": 4.799903192352734e-06, + "loss": 0.2682, + "step": 23958 + }, + { + "epoch": 0.5273827867048206, + "grad_norm": 0.6079639154080521, + "learning_rate": 4.799547022037807e-06, + "loss": 0.4499, + "step": 23959 + }, + { + "epoch": 0.5274047985912392, + "grad_norm": 1.2809780997102729, + "learning_rate": 4.79919085274167e-06, + "loss": 0.3171, + "step": 23960 + }, + { + "epoch": 0.5274268104776579, + "grad_norm": 1.6575312121388066, + "learning_rate": 4.798834684466136e-06, + "loss": 0.3891, + "step": 23961 + }, + { + "epoch": 0.5274488223640766, + "grad_norm": 2.140012946925574, + "learning_rate": 4.798478517213013e-06, + "loss": 0.3152, + "step": 23962 + }, + { + "epoch": 0.5274708342504952, + "grad_norm": 1.3810614134038117, + "learning_rate": 4.798122350984113e-06, + "loss": 0.2934, + "step": 23963 + }, + { + "epoch": 0.5274928461369139, + "grad_norm": 4.451149764814168, + "learning_rate": 4.797766185781246e-06, + "loss": 0.3164, + "step": 23964 + }, + { + "epoch": 0.5275148580233326, + "grad_norm": 3.6778416145134156, + "learning_rate": 4.79741002160622e-06, + "loss": 0.3254, + "step": 23965 + }, + { + "epoch": 0.5275368699097512, + "grad_norm": 1.3953348078124594, + "learning_rate": 4.797053858460846e-06, + "loss": 0.3121, + "step": 23966 + }, + { + "epoch": 0.5275588817961699, + "grad_norm": 1.342722158970799, + "learning_rate": 4.7966976963469365e-06, + "loss": 0.3208, + "step": 23967 + }, + { + "epoch": 0.5275808936825886, + "grad_norm": 2.1945216569886328, + "learning_rate": 4.796341535266298e-06, + "loss": 0.2881, + "step": 23968 + }, + { + "epoch": 0.5276029055690072, + "grad_norm": 1.6733863113923808, + "learning_rate": 4.795985375220746e-06, + "loss": 0.3444, + "step": 23969 + }, + { + "epoch": 0.527624917455426, + "grad_norm": 1.6082010839161738, + "learning_rate": 4.795629216212084e-06, + "loss": 0.3111, + "step": 23970 + }, + { + "epoch": 0.5276469293418447, + "grad_norm": 1.5784828402971776, + "learning_rate": 4.795273058242126e-06, + "loss": 0.3367, + "step": 23971 + }, + { + "epoch": 0.5276689412282632, + "grad_norm": 0.633425967253714, + "learning_rate": 4.794916901312683e-06, + "loss": 0.4964, + "step": 23972 + }, + { + "epoch": 0.527690953114682, + "grad_norm": 1.0931623227200096, + "learning_rate": 4.794560745425562e-06, + "loss": 0.1977, + "step": 23973 + }, + { + "epoch": 0.5277129650011005, + "grad_norm": 0.5957561539296043, + "learning_rate": 4.794204590582577e-06, + "loss": 0.4632, + "step": 23974 + }, + { + "epoch": 0.5277349768875192, + "grad_norm": 1.324642989347583, + "learning_rate": 4.793848436785535e-06, + "loss": 0.3309, + "step": 23975 + }, + { + "epoch": 0.527756988773938, + "grad_norm": 1.607714586987672, + "learning_rate": 4.7934922840362465e-06, + "loss": 0.2959, + "step": 23976 + }, + { + "epoch": 0.5277790006603565, + "grad_norm": 0.6096489587925263, + "learning_rate": 4.793136132336525e-06, + "loss": 0.5037, + "step": 23977 + }, + { + "epoch": 0.5278010125467753, + "grad_norm": 1.3510241169824737, + "learning_rate": 4.792779981688173e-06, + "loss": 0.3571, + "step": 23978 + }, + { + "epoch": 0.527823024433194, + "grad_norm": 1.4955921408621602, + "learning_rate": 4.79242383209301e-06, + "loss": 0.2903, + "step": 23979 + }, + { + "epoch": 0.5278450363196125, + "grad_norm": 1.6686570151697593, + "learning_rate": 4.79206768355284e-06, + "loss": 0.4098, + "step": 23980 + }, + { + "epoch": 0.5278670482060313, + "grad_norm": 0.6004937713092486, + "learning_rate": 4.791711536069474e-06, + "loss": 0.4841, + "step": 23981 + }, + { + "epoch": 0.52788906009245, + "grad_norm": 1.3615359398583924, + "learning_rate": 4.791355389644725e-06, + "loss": 0.2972, + "step": 23982 + }, + { + "epoch": 0.5279110719788686, + "grad_norm": 6.006849962536829, + "learning_rate": 4.7909992442804e-06, + "loss": 0.3282, + "step": 23983 + }, + { + "epoch": 0.5279330838652873, + "grad_norm": 1.6192952626122283, + "learning_rate": 4.79064309997831e-06, + "loss": 0.3342, + "step": 23984 + }, + { + "epoch": 0.527955095751706, + "grad_norm": 1.733446118880402, + "learning_rate": 4.790286956740264e-06, + "loss": 0.3367, + "step": 23985 + }, + { + "epoch": 0.5279771076381246, + "grad_norm": 1.920286306491675, + "learning_rate": 4.789930814568074e-06, + "loss": 0.3287, + "step": 23986 + }, + { + "epoch": 0.5279991195245433, + "grad_norm": 0.5988152008829544, + "learning_rate": 4.78957467346355e-06, + "loss": 0.4457, + "step": 23987 + }, + { + "epoch": 0.528021131410962, + "grad_norm": 1.815001195505931, + "learning_rate": 4.7892185334285004e-06, + "loss": 0.3882, + "step": 23988 + }, + { + "epoch": 0.5280431432973806, + "grad_norm": 2.0017480118808755, + "learning_rate": 4.788862394464737e-06, + "loss": 0.344, + "step": 23989 + }, + { + "epoch": 0.5280651551837993, + "grad_norm": 3.914230676949172, + "learning_rate": 4.78850625657407e-06, + "loss": 0.3384, + "step": 23990 + }, + { + "epoch": 0.528087167070218, + "grad_norm": 1.5565374384848722, + "learning_rate": 4.7881501197583056e-06, + "loss": 0.3054, + "step": 23991 + }, + { + "epoch": 0.5281091789566366, + "grad_norm": 1.7330031401206512, + "learning_rate": 4.78779398401926e-06, + "loss": 0.3507, + "step": 23992 + }, + { + "epoch": 0.5281311908430553, + "grad_norm": 1.1948340477595367, + "learning_rate": 4.787437849358738e-06, + "loss": 0.2363, + "step": 23993 + }, + { + "epoch": 0.528153202729474, + "grad_norm": 1.3047930319167291, + "learning_rate": 4.787081715778551e-06, + "loss": 0.3336, + "step": 23994 + }, + { + "epoch": 0.5281752146158926, + "grad_norm": 1.26640124233929, + "learning_rate": 4.786725583280512e-06, + "loss": 0.4006, + "step": 23995 + }, + { + "epoch": 0.5281972265023113, + "grad_norm": 1.0890435885069303, + "learning_rate": 4.7863694518664265e-06, + "loss": 0.2482, + "step": 23996 + }, + { + "epoch": 0.5282192383887299, + "grad_norm": 1.6721333474907998, + "learning_rate": 4.786013321538109e-06, + "loss": 0.2523, + "step": 23997 + }, + { + "epoch": 0.5282412502751486, + "grad_norm": 1.6862658542546205, + "learning_rate": 4.785657192297365e-06, + "loss": 0.26, + "step": 23998 + }, + { + "epoch": 0.5282632621615673, + "grad_norm": 1.9017603078781298, + "learning_rate": 4.7853010641460075e-06, + "loss": 0.2982, + "step": 23999 + }, + { + "epoch": 0.5282852740479859, + "grad_norm": 1.8413614903222275, + "learning_rate": 4.784944937085848e-06, + "loss": 0.3189, + "step": 24000 + }, + { + "epoch": 0.5283072859344046, + "grad_norm": 1.2307284906303475, + "learning_rate": 4.784588811118691e-06, + "loss": 0.27, + "step": 24001 + }, + { + "epoch": 0.5283292978208233, + "grad_norm": 1.6763126438567488, + "learning_rate": 4.784232686246351e-06, + "loss": 0.2737, + "step": 24002 + }, + { + "epoch": 0.5283513097072419, + "grad_norm": 1.1688745318956415, + "learning_rate": 4.7838765624706364e-06, + "loss": 0.2225, + "step": 24003 + }, + { + "epoch": 0.5283733215936606, + "grad_norm": 0.6608732500215799, + "learning_rate": 4.7835204397933575e-06, + "loss": 0.4882, + "step": 24004 + }, + { + "epoch": 0.5283953334800793, + "grad_norm": 1.1244252344666867, + "learning_rate": 4.783164318216326e-06, + "loss": 0.2303, + "step": 24005 + }, + { + "epoch": 0.5284173453664979, + "grad_norm": 0.6411499011835967, + "learning_rate": 4.782808197741348e-06, + "loss": 0.49, + "step": 24006 + }, + { + "epoch": 0.5284393572529166, + "grad_norm": 1.3861148074977274, + "learning_rate": 4.782452078370237e-06, + "loss": 0.3071, + "step": 24007 + }, + { + "epoch": 0.5284613691393353, + "grad_norm": 0.6222174525694578, + "learning_rate": 4.782095960104802e-06, + "loss": 0.4798, + "step": 24008 + }, + { + "epoch": 0.5284833810257539, + "grad_norm": 1.3895831065433153, + "learning_rate": 4.78173984294685e-06, + "loss": 0.2447, + "step": 24009 + }, + { + "epoch": 0.5285053929121726, + "grad_norm": 1.6725906041825458, + "learning_rate": 4.7813837268981956e-06, + "loss": 0.313, + "step": 24010 + }, + { + "epoch": 0.5285274047985913, + "grad_norm": 1.355864764559823, + "learning_rate": 4.7810276119606446e-06, + "loss": 0.3142, + "step": 24011 + }, + { + "epoch": 0.5285494166850099, + "grad_norm": 1.6319253464991628, + "learning_rate": 4.780671498136011e-06, + "loss": 0.3562, + "step": 24012 + }, + { + "epoch": 0.5285714285714286, + "grad_norm": 0.6395185419074113, + "learning_rate": 4.780315385426103e-06, + "loss": 0.4945, + "step": 24013 + }, + { + "epoch": 0.5285934404578473, + "grad_norm": 1.6447529866439645, + "learning_rate": 4.779959273832727e-06, + "loss": 0.4979, + "step": 24014 + }, + { + "epoch": 0.5286154523442659, + "grad_norm": 1.585010067141756, + "learning_rate": 4.779603163357699e-06, + "loss": 0.3515, + "step": 24015 + }, + { + "epoch": 0.5286374642306846, + "grad_norm": 1.30874175675784, + "learning_rate": 4.7792470540028246e-06, + "loss": 0.2955, + "step": 24016 + }, + { + "epoch": 0.5286594761171033, + "grad_norm": 1.6811482515231, + "learning_rate": 4.778890945769916e-06, + "loss": 0.3292, + "step": 24017 + }, + { + "epoch": 0.5286814880035219, + "grad_norm": 1.4373809247909162, + "learning_rate": 4.7785348386607836e-06, + "loss": 0.3108, + "step": 24018 + }, + { + "epoch": 0.5287034998899406, + "grad_norm": 0.62459485438555, + "learning_rate": 4.778178732677233e-06, + "loss": 0.4495, + "step": 24019 + }, + { + "epoch": 0.5287255117763592, + "grad_norm": 3.125434899229882, + "learning_rate": 4.777822627821078e-06, + "loss": 0.3766, + "step": 24020 + }, + { + "epoch": 0.5287475236627779, + "grad_norm": 0.6177573297671768, + "learning_rate": 4.777466524094128e-06, + "loss": 0.496, + "step": 24021 + }, + { + "epoch": 0.5287695355491966, + "grad_norm": 1.3624280651376823, + "learning_rate": 4.7771104214981926e-06, + "loss": 0.2562, + "step": 24022 + }, + { + "epoch": 0.5287915474356152, + "grad_norm": 0.6038195453467903, + "learning_rate": 4.776754320035082e-06, + "loss": 0.4828, + "step": 24023 + }, + { + "epoch": 0.5288135593220339, + "grad_norm": 1.1393586636935087, + "learning_rate": 4.7763982197066035e-06, + "loss": 0.3022, + "step": 24024 + }, + { + "epoch": 0.5288355712084526, + "grad_norm": 1.4887560404907778, + "learning_rate": 4.77604212051457e-06, + "loss": 0.2605, + "step": 24025 + }, + { + "epoch": 0.5288575830948712, + "grad_norm": 1.443162619650667, + "learning_rate": 4.7756860224607895e-06, + "loss": 0.294, + "step": 24026 + }, + { + "epoch": 0.5288795949812899, + "grad_norm": 0.6265483882593571, + "learning_rate": 4.775329925547074e-06, + "loss": 0.4675, + "step": 24027 + }, + { + "epoch": 0.5289016068677086, + "grad_norm": 1.3054174879162408, + "learning_rate": 4.774973829775233e-06, + "loss": 0.2422, + "step": 24028 + }, + { + "epoch": 0.5289236187541272, + "grad_norm": 1.9316251636533095, + "learning_rate": 4.774617735147073e-06, + "loss": 0.3162, + "step": 24029 + }, + { + "epoch": 0.5289456306405459, + "grad_norm": 1.555042011274913, + "learning_rate": 4.774261641664408e-06, + "loss": 0.3108, + "step": 24030 + }, + { + "epoch": 0.5289676425269646, + "grad_norm": 1.3464259772634315, + "learning_rate": 4.773905549329046e-06, + "loss": 0.3333, + "step": 24031 + }, + { + "epoch": 0.5289896544133832, + "grad_norm": 1.635530434373326, + "learning_rate": 4.773549458142793e-06, + "loss": 0.375, + "step": 24032 + }, + { + "epoch": 0.5290116662998019, + "grad_norm": 2.05057033380982, + "learning_rate": 4.773193368107468e-06, + "loss": 0.3816, + "step": 24033 + }, + { + "epoch": 0.5290336781862206, + "grad_norm": 1.8306433555696318, + "learning_rate": 4.7728372792248725e-06, + "loss": 0.377, + "step": 24034 + }, + { + "epoch": 0.5290556900726392, + "grad_norm": 1.4535110846758992, + "learning_rate": 4.77248119149682e-06, + "loss": 0.3532, + "step": 24035 + }, + { + "epoch": 0.5290777019590579, + "grad_norm": 0.6208133922157822, + "learning_rate": 4.77212510492512e-06, + "loss": 0.4754, + "step": 24036 + }, + { + "epoch": 0.5290997138454766, + "grad_norm": 1.3067059367232061, + "learning_rate": 4.771769019511581e-06, + "loss": 0.3468, + "step": 24037 + }, + { + "epoch": 0.5291217257318952, + "grad_norm": 1.476203491924623, + "learning_rate": 4.7714129352580145e-06, + "loss": 0.2766, + "step": 24038 + }, + { + "epoch": 0.5291437376183139, + "grad_norm": 1.379903271795214, + "learning_rate": 4.771056852166227e-06, + "loss": 0.3381, + "step": 24039 + }, + { + "epoch": 0.5291657495047326, + "grad_norm": 1.932701252205027, + "learning_rate": 4.770700770238033e-06, + "loss": 0.3402, + "step": 24040 + }, + { + "epoch": 0.5291877613911512, + "grad_norm": 1.8197038023716294, + "learning_rate": 4.770344689475241e-06, + "loss": 0.3271, + "step": 24041 + }, + { + "epoch": 0.5292097732775699, + "grad_norm": 1.4278487505834283, + "learning_rate": 4.769988609879657e-06, + "loss": 0.3357, + "step": 24042 + }, + { + "epoch": 0.5292317851639885, + "grad_norm": 1.3387837288207032, + "learning_rate": 4.769632531453095e-06, + "loss": 0.3158, + "step": 24043 + }, + { + "epoch": 0.5292537970504072, + "grad_norm": 0.6177421615265901, + "learning_rate": 4.769276454197363e-06, + "loss": 0.461, + "step": 24044 + }, + { + "epoch": 0.5292758089368259, + "grad_norm": 1.9378906436679573, + "learning_rate": 4.76892037811427e-06, + "loss": 0.3613, + "step": 24045 + }, + { + "epoch": 0.5292978208232445, + "grad_norm": 1.9413839361300398, + "learning_rate": 4.768564303205629e-06, + "loss": 0.3313, + "step": 24046 + }, + { + "epoch": 0.5293198327096632, + "grad_norm": 0.6437222356243366, + "learning_rate": 4.768208229473245e-06, + "loss": 0.4768, + "step": 24047 + }, + { + "epoch": 0.5293418445960819, + "grad_norm": 1.341102785086406, + "learning_rate": 4.767852156918932e-06, + "loss": 0.3021, + "step": 24048 + }, + { + "epoch": 0.5293638564825005, + "grad_norm": 0.6643605168980263, + "learning_rate": 4.767496085544498e-06, + "loss": 0.4875, + "step": 24049 + }, + { + "epoch": 0.5293858683689192, + "grad_norm": 1.4694540244980656, + "learning_rate": 4.76714001535175e-06, + "loss": 0.2179, + "step": 24050 + }, + { + "epoch": 0.5294078802553379, + "grad_norm": 1.654906466270114, + "learning_rate": 4.766783946342503e-06, + "loss": 0.3117, + "step": 24051 + }, + { + "epoch": 0.5294298921417565, + "grad_norm": 1.324707517284943, + "learning_rate": 4.7664278785185615e-06, + "loss": 0.2687, + "step": 24052 + }, + { + "epoch": 0.5294519040281752, + "grad_norm": 1.6857466724286307, + "learning_rate": 4.7660718118817405e-06, + "loss": 0.2448, + "step": 24053 + }, + { + "epoch": 0.5294739159145939, + "grad_norm": 1.3203614064449383, + "learning_rate": 4.765715746433846e-06, + "loss": 0.2731, + "step": 24054 + }, + { + "epoch": 0.5294959278010125, + "grad_norm": 1.2728630610344267, + "learning_rate": 4.765359682176687e-06, + "loss": 0.3904, + "step": 24055 + }, + { + "epoch": 0.5295179396874312, + "grad_norm": 1.6973500474027199, + "learning_rate": 4.765003619112077e-06, + "loss": 0.3158, + "step": 24056 + }, + { + "epoch": 0.5295399515738499, + "grad_norm": 1.3239137799240877, + "learning_rate": 4.764647557241821e-06, + "loss": 0.3119, + "step": 24057 + }, + { + "epoch": 0.5295619634602685, + "grad_norm": 1.3935395977233274, + "learning_rate": 4.764291496567732e-06, + "loss": 0.3674, + "step": 24058 + }, + { + "epoch": 0.5295839753466872, + "grad_norm": 1.343658735769162, + "learning_rate": 4.763935437091621e-06, + "loss": 0.3721, + "step": 24059 + }, + { + "epoch": 0.5296059872331059, + "grad_norm": 0.6097309778365415, + "learning_rate": 4.763579378815292e-06, + "loss": 0.4768, + "step": 24060 + }, + { + "epoch": 0.5296279991195245, + "grad_norm": 2.0108177156472107, + "learning_rate": 4.76322332174056e-06, + "loss": 0.3047, + "step": 24061 + }, + { + "epoch": 0.5296500110059432, + "grad_norm": 3.3579944782614355, + "learning_rate": 4.762867265869232e-06, + "loss": 0.3021, + "step": 24062 + }, + { + "epoch": 0.5296720228923619, + "grad_norm": 0.6416861993018728, + "learning_rate": 4.762511211203118e-06, + "loss": 0.4878, + "step": 24063 + }, + { + "epoch": 0.5296940347787805, + "grad_norm": 3.646533141846734, + "learning_rate": 4.76215515774403e-06, + "loss": 0.3179, + "step": 24064 + }, + { + "epoch": 0.5297160466651992, + "grad_norm": 1.7471556750704038, + "learning_rate": 4.761799105493773e-06, + "loss": 0.2971, + "step": 24065 + }, + { + "epoch": 0.5297380585516178, + "grad_norm": 0.625226623040381, + "learning_rate": 4.761443054454161e-06, + "loss": 0.4777, + "step": 24066 + }, + { + "epoch": 0.5297600704380365, + "grad_norm": 0.6286990043142849, + "learning_rate": 4.761087004627002e-06, + "loss": 0.4721, + "step": 24067 + }, + { + "epoch": 0.5297820823244552, + "grad_norm": 2.4159734118556453, + "learning_rate": 4.760730956014102e-06, + "loss": 0.4309, + "step": 24068 + }, + { + "epoch": 0.5298040942108738, + "grad_norm": 0.6417564463256654, + "learning_rate": 4.760374908617278e-06, + "loss": 0.5052, + "step": 24069 + }, + { + "epoch": 0.5298261060972925, + "grad_norm": 1.5752499010750918, + "learning_rate": 4.760018862438333e-06, + "loss": 0.3197, + "step": 24070 + }, + { + "epoch": 0.5298481179837112, + "grad_norm": 1.6799443813068249, + "learning_rate": 4.759662817479081e-06, + "loss": 0.2765, + "step": 24071 + }, + { + "epoch": 0.5298701298701298, + "grad_norm": 1.7452382762070784, + "learning_rate": 4.75930677374133e-06, + "loss": 0.3444, + "step": 24072 + }, + { + "epoch": 0.5298921417565485, + "grad_norm": 1.39035746870618, + "learning_rate": 4.758950731226888e-06, + "loss": 0.2645, + "step": 24073 + }, + { + "epoch": 0.5299141536429672, + "grad_norm": 1.4461408922574537, + "learning_rate": 4.758594689937566e-06, + "loss": 0.3539, + "step": 24074 + }, + { + "epoch": 0.5299361655293858, + "grad_norm": 1.8065440431724884, + "learning_rate": 4.758238649875173e-06, + "loss": 0.3244, + "step": 24075 + }, + { + "epoch": 0.5299581774158045, + "grad_norm": 1.3338162635024822, + "learning_rate": 4.757882611041519e-06, + "loss": 0.2852, + "step": 24076 + }, + { + "epoch": 0.5299801893022232, + "grad_norm": 2.778742512612965, + "learning_rate": 4.757526573438416e-06, + "loss": 0.3935, + "step": 24077 + }, + { + "epoch": 0.5300022011886418, + "grad_norm": 1.3598402371223135, + "learning_rate": 4.757170537067668e-06, + "loss": 0.2438, + "step": 24078 + }, + { + "epoch": 0.5300242130750605, + "grad_norm": 2.247219316988461, + "learning_rate": 4.75681450193109e-06, + "loss": 0.3182, + "step": 24079 + }, + { + "epoch": 0.5300462249614792, + "grad_norm": 1.1241975051916349, + "learning_rate": 4.756458468030486e-06, + "loss": 0.2429, + "step": 24080 + }, + { + "epoch": 0.5300682368478978, + "grad_norm": 1.4304961836208299, + "learning_rate": 4.756102435367671e-06, + "loss": 0.2827, + "step": 24081 + }, + { + "epoch": 0.5300902487343165, + "grad_norm": 0.7118787471717769, + "learning_rate": 4.755746403944454e-06, + "loss": 0.5002, + "step": 24082 + }, + { + "epoch": 0.5301122606207352, + "grad_norm": 0.6526700124256279, + "learning_rate": 4.75539037376264e-06, + "loss": 0.5058, + "step": 24083 + }, + { + "epoch": 0.5301342725071538, + "grad_norm": 1.4521462397769567, + "learning_rate": 4.755034344824042e-06, + "loss": 0.2987, + "step": 24084 + }, + { + "epoch": 0.5301562843935725, + "grad_norm": 1.440319046652869, + "learning_rate": 4.754678317130468e-06, + "loss": 0.2976, + "step": 24085 + }, + { + "epoch": 0.5301782962799912, + "grad_norm": 1.6986329876257737, + "learning_rate": 4.754322290683729e-06, + "loss": 0.2828, + "step": 24086 + }, + { + "epoch": 0.5302003081664098, + "grad_norm": 1.2844008458082148, + "learning_rate": 4.753966265485634e-06, + "loss": 0.2087, + "step": 24087 + }, + { + "epoch": 0.5302223200528285, + "grad_norm": 1.7143907619312508, + "learning_rate": 4.753610241537991e-06, + "loss": 0.3162, + "step": 24088 + }, + { + "epoch": 0.5302443319392472, + "grad_norm": 1.7026630800548512, + "learning_rate": 4.753254218842612e-06, + "loss": 0.3511, + "step": 24089 + }, + { + "epoch": 0.5302663438256658, + "grad_norm": 1.2367848073015741, + "learning_rate": 4.7528981974013055e-06, + "loss": 0.3079, + "step": 24090 + }, + { + "epoch": 0.5302883557120845, + "grad_norm": 3.2983976523812957, + "learning_rate": 4.752542177215879e-06, + "loss": 0.3109, + "step": 24091 + }, + { + "epoch": 0.5303103675985031, + "grad_norm": 1.3232563486017066, + "learning_rate": 4.752186158288144e-06, + "loss": 0.2916, + "step": 24092 + }, + { + "epoch": 0.5303323794849218, + "grad_norm": 1.4679929327114365, + "learning_rate": 4.751830140619908e-06, + "loss": 0.2571, + "step": 24093 + }, + { + "epoch": 0.5303543913713405, + "grad_norm": 0.6408863193505123, + "learning_rate": 4.751474124212985e-06, + "loss": 0.4648, + "step": 24094 + }, + { + "epoch": 0.5303764032577591, + "grad_norm": 1.472322483542273, + "learning_rate": 4.75111810906918e-06, + "loss": 0.3518, + "step": 24095 + }, + { + "epoch": 0.5303984151441778, + "grad_norm": 1.5291839840632897, + "learning_rate": 4.7507620951903035e-06, + "loss": 0.2969, + "step": 24096 + }, + { + "epoch": 0.5304204270305966, + "grad_norm": 0.6536312354662106, + "learning_rate": 4.750406082578165e-06, + "loss": 0.492, + "step": 24097 + }, + { + "epoch": 0.5304424389170151, + "grad_norm": 2.3780940202561065, + "learning_rate": 4.750050071234573e-06, + "loss": 0.3379, + "step": 24098 + }, + { + "epoch": 0.5304644508034339, + "grad_norm": 1.7803254939067017, + "learning_rate": 4.74969406116134e-06, + "loss": 0.3798, + "step": 24099 + }, + { + "epoch": 0.5304864626898526, + "grad_norm": 1.3498956368298838, + "learning_rate": 4.749338052360274e-06, + "loss": 0.2751, + "step": 24100 + }, + { + "epoch": 0.5305084745762711, + "grad_norm": 1.682885249746686, + "learning_rate": 4.7489820448331825e-06, + "loss": 0.4096, + "step": 24101 + }, + { + "epoch": 0.5305304864626899, + "grad_norm": 1.5085794150276657, + "learning_rate": 4.7486260385818764e-06, + "loss": 0.2715, + "step": 24102 + }, + { + "epoch": 0.5305524983491086, + "grad_norm": 1.4307899352788906, + "learning_rate": 4.748270033608164e-06, + "loss": 0.2236, + "step": 24103 + }, + { + "epoch": 0.5305745102355272, + "grad_norm": 1.2449334477638152, + "learning_rate": 4.747914029913857e-06, + "loss": 0.21, + "step": 24104 + }, + { + "epoch": 0.5305965221219459, + "grad_norm": 1.4149358480043999, + "learning_rate": 4.747558027500764e-06, + "loss": 0.3079, + "step": 24105 + }, + { + "epoch": 0.5306185340083646, + "grad_norm": 0.6305024933141318, + "learning_rate": 4.747202026370691e-06, + "loss": 0.4675, + "step": 24106 + }, + { + "epoch": 0.5306405458947832, + "grad_norm": 1.457982315851646, + "learning_rate": 4.746846026525452e-06, + "loss": 0.2809, + "step": 24107 + }, + { + "epoch": 0.5306625577812019, + "grad_norm": 1.925960006026872, + "learning_rate": 4.746490027966855e-06, + "loss": 0.3298, + "step": 24108 + }, + { + "epoch": 0.5306845696676206, + "grad_norm": 1.4818681439776262, + "learning_rate": 4.746134030696707e-06, + "loss": 0.3553, + "step": 24109 + }, + { + "epoch": 0.5307065815540392, + "grad_norm": 6.759307959934557, + "learning_rate": 4.7457780347168195e-06, + "loss": 0.3238, + "step": 24110 + }, + { + "epoch": 0.5307285934404579, + "grad_norm": 1.7500950040755663, + "learning_rate": 4.7454220400290005e-06, + "loss": 0.2941, + "step": 24111 + }, + { + "epoch": 0.5307506053268766, + "grad_norm": 1.3155577222815722, + "learning_rate": 4.745066046635062e-06, + "loss": 0.2543, + "step": 24112 + }, + { + "epoch": 0.5307726172132952, + "grad_norm": 0.6350599483930495, + "learning_rate": 4.744710054536812e-06, + "loss": 0.4772, + "step": 24113 + }, + { + "epoch": 0.5307946290997139, + "grad_norm": 1.0378091530962505, + "learning_rate": 4.744354063736058e-06, + "loss": 0.2587, + "step": 24114 + }, + { + "epoch": 0.5308166409861325, + "grad_norm": 1.1762985134198534, + "learning_rate": 4.7439980742346106e-06, + "loss": 0.2562, + "step": 24115 + }, + { + "epoch": 0.5308386528725512, + "grad_norm": 1.5524344353978967, + "learning_rate": 4.743642086034279e-06, + "loss": 0.2884, + "step": 24116 + }, + { + "epoch": 0.5308606647589699, + "grad_norm": 1.3572260640042848, + "learning_rate": 4.7432860991368725e-06, + "loss": 0.2501, + "step": 24117 + }, + { + "epoch": 0.5308826766453885, + "grad_norm": 1.3894431411734591, + "learning_rate": 4.742930113544203e-06, + "loss": 0.2616, + "step": 24118 + }, + { + "epoch": 0.5309046885318072, + "grad_norm": 1.8742136220870251, + "learning_rate": 4.742574129258075e-06, + "loss": 0.3122, + "step": 24119 + }, + { + "epoch": 0.5309267004182259, + "grad_norm": 1.33375821713349, + "learning_rate": 4.742218146280301e-06, + "loss": 0.252, + "step": 24120 + }, + { + "epoch": 0.5309487123046445, + "grad_norm": 1.7834230351262093, + "learning_rate": 4.7418621646126876e-06, + "loss": 0.28, + "step": 24121 + }, + { + "epoch": 0.5309707241910632, + "grad_norm": 2.7151370321964987, + "learning_rate": 4.741506184257047e-06, + "loss": 0.3324, + "step": 24122 + }, + { + "epoch": 0.5309927360774819, + "grad_norm": 1.767493752250418, + "learning_rate": 4.741150205215189e-06, + "loss": 0.345, + "step": 24123 + }, + { + "epoch": 0.5310147479639005, + "grad_norm": 1.4935396222371407, + "learning_rate": 4.740794227488919e-06, + "loss": 0.31, + "step": 24124 + }, + { + "epoch": 0.5310367598503192, + "grad_norm": 6.645552441921993, + "learning_rate": 4.740438251080049e-06, + "loss": 0.3171, + "step": 24125 + }, + { + "epoch": 0.5310587717367379, + "grad_norm": 1.3217700362947593, + "learning_rate": 4.740082275990386e-06, + "loss": 0.332, + "step": 24126 + }, + { + "epoch": 0.5310807836231565, + "grad_norm": 1.1654723867825973, + "learning_rate": 4.739726302221744e-06, + "loss": 0.3214, + "step": 24127 + }, + { + "epoch": 0.5311027955095752, + "grad_norm": 1.4739546480375094, + "learning_rate": 4.739370329775928e-06, + "loss": 0.2914, + "step": 24128 + }, + { + "epoch": 0.5311248073959939, + "grad_norm": 1.2911602109705118, + "learning_rate": 4.739014358654747e-06, + "loss": 0.297, + "step": 24129 + }, + { + "epoch": 0.5311468192824125, + "grad_norm": 1.3554293418391519, + "learning_rate": 4.7386583888600115e-06, + "loss": 0.4019, + "step": 24130 + }, + { + "epoch": 0.5311688311688312, + "grad_norm": 1.6721269785235708, + "learning_rate": 4.738302420393532e-06, + "loss": 0.3143, + "step": 24131 + }, + { + "epoch": 0.5311908430552499, + "grad_norm": 0.6188586947173454, + "learning_rate": 4.737946453257115e-06, + "loss": 0.4918, + "step": 24132 + }, + { + "epoch": 0.5312128549416685, + "grad_norm": 1.7601012481849674, + "learning_rate": 4.7375904874525715e-06, + "loss": 0.2675, + "step": 24133 + }, + { + "epoch": 0.5312348668280872, + "grad_norm": 2.6816646388923724, + "learning_rate": 4.737234522981709e-06, + "loss": 0.301, + "step": 24134 + }, + { + "epoch": 0.5312568787145059, + "grad_norm": 1.547582838883036, + "learning_rate": 4.736878559846339e-06, + "loss": 0.2933, + "step": 24135 + }, + { + "epoch": 0.5312788906009245, + "grad_norm": 1.6267172996543544, + "learning_rate": 4.736522598048272e-06, + "loss": 0.3087, + "step": 24136 + }, + { + "epoch": 0.5313009024873432, + "grad_norm": 1.6847132015719855, + "learning_rate": 4.736166637589311e-06, + "loss": 0.3038, + "step": 24137 + }, + { + "epoch": 0.5313229143737618, + "grad_norm": 1.8161229151205096, + "learning_rate": 4.73581067847127e-06, + "loss": 0.414, + "step": 24138 + }, + { + "epoch": 0.5313449262601805, + "grad_norm": 1.8038739264848476, + "learning_rate": 4.735454720695955e-06, + "loss": 0.3876, + "step": 24139 + }, + { + "epoch": 0.5313669381465992, + "grad_norm": 1.3290168416859587, + "learning_rate": 4.735098764265181e-06, + "loss": 0.3059, + "step": 24140 + }, + { + "epoch": 0.5313889500330178, + "grad_norm": 1.609693803562621, + "learning_rate": 4.734742809180751e-06, + "loss": 0.3134, + "step": 24141 + }, + { + "epoch": 0.5314109619194365, + "grad_norm": 1.5100629268881656, + "learning_rate": 4.734386855444476e-06, + "loss": 0.3372, + "step": 24142 + }, + { + "epoch": 0.5314329738058552, + "grad_norm": 2.4171221939481944, + "learning_rate": 4.7340309030581665e-06, + "loss": 0.3129, + "step": 24143 + }, + { + "epoch": 0.5314549856922738, + "grad_norm": 1.3826717042464376, + "learning_rate": 4.7336749520236295e-06, + "loss": 0.3137, + "step": 24144 + }, + { + "epoch": 0.5314769975786925, + "grad_norm": 1.4911713477144568, + "learning_rate": 4.733319002342677e-06, + "loss": 0.3678, + "step": 24145 + }, + { + "epoch": 0.5314990094651112, + "grad_norm": 1.2013905921054098, + "learning_rate": 4.7329630540171154e-06, + "loss": 0.3433, + "step": 24146 + }, + { + "epoch": 0.5315210213515298, + "grad_norm": 1.6593192103998633, + "learning_rate": 4.732607107048753e-06, + "loss": 0.2668, + "step": 24147 + }, + { + "epoch": 0.5315430332379485, + "grad_norm": 1.3105262854322342, + "learning_rate": 4.732251161439402e-06, + "loss": 0.2943, + "step": 24148 + }, + { + "epoch": 0.5315650451243672, + "grad_norm": 1.9775842492458933, + "learning_rate": 4.7318952171908715e-06, + "loss": 0.3317, + "step": 24149 + }, + { + "epoch": 0.5315870570107858, + "grad_norm": 1.6457752983398979, + "learning_rate": 4.731539274304967e-06, + "loss": 0.3685, + "step": 24150 + }, + { + "epoch": 0.5316090688972045, + "grad_norm": 1.0609615096952867, + "learning_rate": 4.7311833327835e-06, + "loss": 0.2923, + "step": 24151 + }, + { + "epoch": 0.5316310807836232, + "grad_norm": 7.43530845585968, + "learning_rate": 4.730827392628278e-06, + "loss": 0.3466, + "step": 24152 + }, + { + "epoch": 0.5316530926700418, + "grad_norm": 1.7631436331425925, + "learning_rate": 4.730471453841113e-06, + "loss": 0.2995, + "step": 24153 + }, + { + "epoch": 0.5316751045564605, + "grad_norm": 2.120161402391776, + "learning_rate": 4.7301155164238136e-06, + "loss": 0.3537, + "step": 24154 + }, + { + "epoch": 0.5316971164428792, + "grad_norm": 1.5958155540813024, + "learning_rate": 4.7297595803781845e-06, + "loss": 0.267, + "step": 24155 + }, + { + "epoch": 0.5317191283292978, + "grad_norm": 1.2318449818099808, + "learning_rate": 4.72940364570604e-06, + "loss": 0.2843, + "step": 24156 + }, + { + "epoch": 0.5317411402157165, + "grad_norm": 1.445106548685443, + "learning_rate": 4.729047712409184e-06, + "loss": 0.319, + "step": 24157 + }, + { + "epoch": 0.5317631521021352, + "grad_norm": 1.741324290093707, + "learning_rate": 4.728691780489432e-06, + "loss": 0.2667, + "step": 24158 + }, + { + "epoch": 0.5317851639885538, + "grad_norm": 0.63098582525642, + "learning_rate": 4.728335849948588e-06, + "loss": 0.474, + "step": 24159 + }, + { + "epoch": 0.5318071758749725, + "grad_norm": 0.6337819912095741, + "learning_rate": 4.727979920788462e-06, + "loss": 0.4953, + "step": 24160 + }, + { + "epoch": 0.5318291877613911, + "grad_norm": 1.6575259583914668, + "learning_rate": 4.727623993010863e-06, + "loss": 0.3301, + "step": 24161 + }, + { + "epoch": 0.5318511996478098, + "grad_norm": 2.8355087217134702, + "learning_rate": 4.727268066617599e-06, + "loss": 0.3457, + "step": 24162 + }, + { + "epoch": 0.5318732115342285, + "grad_norm": 2.003072881721789, + "learning_rate": 4.726912141610484e-06, + "loss": 0.2961, + "step": 24163 + }, + { + "epoch": 0.5318952234206471, + "grad_norm": 0.6152132164805145, + "learning_rate": 4.726556217991321e-06, + "loss": 0.4616, + "step": 24164 + }, + { + "epoch": 0.5319172353070658, + "grad_norm": 1.354670240195274, + "learning_rate": 4.726200295761921e-06, + "loss": 0.2823, + "step": 24165 + }, + { + "epoch": 0.5319392471934845, + "grad_norm": 0.6015918938939703, + "learning_rate": 4.725844374924094e-06, + "loss": 0.4764, + "step": 24166 + }, + { + "epoch": 0.5319612590799031, + "grad_norm": 0.6209124654188323, + "learning_rate": 4.72548845547965e-06, + "loss": 0.4813, + "step": 24167 + }, + { + "epoch": 0.5319832709663218, + "grad_norm": 2.398461072119521, + "learning_rate": 4.725132537430393e-06, + "loss": 0.3941, + "step": 24168 + }, + { + "epoch": 0.5320052828527405, + "grad_norm": 1.3756326517607136, + "learning_rate": 4.7247766207781365e-06, + "loss": 0.3809, + "step": 24169 + }, + { + "epoch": 0.5320272947391591, + "grad_norm": 1.1935282343721927, + "learning_rate": 4.724420705524686e-06, + "loss": 0.3079, + "step": 24170 + }, + { + "epoch": 0.5320493066255778, + "grad_norm": 0.6304502661696273, + "learning_rate": 4.724064791671855e-06, + "loss": 0.4881, + "step": 24171 + }, + { + "epoch": 0.5320713185119965, + "grad_norm": 0.6315038237032397, + "learning_rate": 4.72370887922145e-06, + "loss": 0.4934, + "step": 24172 + }, + { + "epoch": 0.5320933303984151, + "grad_norm": 2.3166889938106445, + "learning_rate": 4.723352968175277e-06, + "loss": 0.3613, + "step": 24173 + }, + { + "epoch": 0.5321153422848338, + "grad_norm": 1.8989205113782512, + "learning_rate": 4.72299705853515e-06, + "loss": 0.2996, + "step": 24174 + }, + { + "epoch": 0.5321373541712525, + "grad_norm": 1.7086232704693005, + "learning_rate": 4.7226411503028734e-06, + "loss": 0.2877, + "step": 24175 + }, + { + "epoch": 0.5321593660576711, + "grad_norm": 2.1474855097958683, + "learning_rate": 4.7222852434802605e-06, + "loss": 0.4006, + "step": 24176 + }, + { + "epoch": 0.5321813779440898, + "grad_norm": 1.7110834680982616, + "learning_rate": 4.721929338069117e-06, + "loss": 0.4083, + "step": 24177 + }, + { + "epoch": 0.5322033898305085, + "grad_norm": 1.4865122971959368, + "learning_rate": 4.721573434071251e-06, + "loss": 0.3313, + "step": 24178 + }, + { + "epoch": 0.5322254017169271, + "grad_norm": 1.5368061493079381, + "learning_rate": 4.721217531488475e-06, + "loss": 0.3036, + "step": 24179 + }, + { + "epoch": 0.5322474136033458, + "grad_norm": 1.2710651443138088, + "learning_rate": 4.720861630322595e-06, + "loss": 0.2959, + "step": 24180 + }, + { + "epoch": 0.5322694254897645, + "grad_norm": 1.7696082775035742, + "learning_rate": 4.720505730575422e-06, + "loss": 0.3632, + "step": 24181 + }, + { + "epoch": 0.5322914373761831, + "grad_norm": 1.886402579093017, + "learning_rate": 4.720149832248762e-06, + "loss": 0.2534, + "step": 24182 + }, + { + "epoch": 0.5323134492626018, + "grad_norm": 1.5721131042596133, + "learning_rate": 4.719793935344425e-06, + "loss": 0.2316, + "step": 24183 + }, + { + "epoch": 0.5323354611490204, + "grad_norm": 1.4994862401001363, + "learning_rate": 4.7194380398642215e-06, + "loss": 0.2966, + "step": 24184 + }, + { + "epoch": 0.5323574730354391, + "grad_norm": 1.493296438178116, + "learning_rate": 4.719082145809958e-06, + "loss": 0.318, + "step": 24185 + }, + { + "epoch": 0.5323794849218578, + "grad_norm": 2.2289931272335233, + "learning_rate": 4.718726253183446e-06, + "loss": 0.3459, + "step": 24186 + }, + { + "epoch": 0.5324014968082764, + "grad_norm": 1.2923925979631168, + "learning_rate": 4.718370361986491e-06, + "loss": 0.2905, + "step": 24187 + }, + { + "epoch": 0.5324235086946951, + "grad_norm": 1.6716337765661193, + "learning_rate": 4.718014472220903e-06, + "loss": 0.3292, + "step": 24188 + }, + { + "epoch": 0.5324455205811138, + "grad_norm": 1.646148350447867, + "learning_rate": 4.717658583888493e-06, + "loss": 0.5484, + "step": 24189 + }, + { + "epoch": 0.5324675324675324, + "grad_norm": 2.1068304790547656, + "learning_rate": 4.717302696991068e-06, + "loss": 0.4114, + "step": 24190 + }, + { + "epoch": 0.5324895443539511, + "grad_norm": 0.5962038300116488, + "learning_rate": 4.716946811530435e-06, + "loss": 0.4709, + "step": 24191 + }, + { + "epoch": 0.5325115562403698, + "grad_norm": 1.3856208213401713, + "learning_rate": 4.716590927508406e-06, + "loss": 0.3016, + "step": 24192 + }, + { + "epoch": 0.5325335681267884, + "grad_norm": 1.4897113867688132, + "learning_rate": 4.7162350449267865e-06, + "loss": 0.2641, + "step": 24193 + }, + { + "epoch": 0.5325555800132071, + "grad_norm": 1.263170670508797, + "learning_rate": 4.71587916378739e-06, + "loss": 0.3142, + "step": 24194 + }, + { + "epoch": 0.5325775918996258, + "grad_norm": 1.9021186270492068, + "learning_rate": 4.715523284092021e-06, + "loss": 0.3369, + "step": 24195 + }, + { + "epoch": 0.5325996037860444, + "grad_norm": 1.5320984778544964, + "learning_rate": 4.715167405842488e-06, + "loss": 0.3108, + "step": 24196 + }, + { + "epoch": 0.5326216156724631, + "grad_norm": 0.590049932832201, + "learning_rate": 4.7148115290406025e-06, + "loss": 0.4577, + "step": 24197 + }, + { + "epoch": 0.5326436275588818, + "grad_norm": 1.5114341819647827, + "learning_rate": 4.714455653688171e-06, + "loss": 0.3302, + "step": 24198 + }, + { + "epoch": 0.5326656394453004, + "grad_norm": 5.345881921173674, + "learning_rate": 4.714099779787006e-06, + "loss": 0.238, + "step": 24199 + }, + { + "epoch": 0.5326876513317191, + "grad_norm": 2.953252386558568, + "learning_rate": 4.713743907338912e-06, + "loss": 0.495, + "step": 24200 + }, + { + "epoch": 0.5327096632181378, + "grad_norm": 1.7974366262719943, + "learning_rate": 4.713388036345698e-06, + "loss": 0.3083, + "step": 24201 + }, + { + "epoch": 0.5327316751045564, + "grad_norm": 1.254425286763116, + "learning_rate": 4.713032166809174e-06, + "loss": 0.3645, + "step": 24202 + }, + { + "epoch": 0.5327536869909751, + "grad_norm": 2.401181758165216, + "learning_rate": 4.712676298731149e-06, + "loss": 0.2697, + "step": 24203 + }, + { + "epoch": 0.5327756988773938, + "grad_norm": 1.336825788584444, + "learning_rate": 4.712320432113432e-06, + "loss": 0.2447, + "step": 24204 + }, + { + "epoch": 0.5327977107638124, + "grad_norm": 1.6299471360731894, + "learning_rate": 4.71196456695783e-06, + "loss": 0.3383, + "step": 24205 + }, + { + "epoch": 0.5328197226502311, + "grad_norm": 1.3862802004742232, + "learning_rate": 4.711608703266152e-06, + "loss": 0.368, + "step": 24206 + }, + { + "epoch": 0.5328417345366497, + "grad_norm": 1.2881310600509224, + "learning_rate": 4.711252841040208e-06, + "loss": 0.3295, + "step": 24207 + }, + { + "epoch": 0.5328637464230684, + "grad_norm": 1.640223822287802, + "learning_rate": 4.7108969802818076e-06, + "loss": 0.3525, + "step": 24208 + }, + { + "epoch": 0.5328857583094871, + "grad_norm": 3.3313659865006757, + "learning_rate": 4.710541120992756e-06, + "loss": 0.3781, + "step": 24209 + }, + { + "epoch": 0.5329077701959057, + "grad_norm": 1.8879888563239497, + "learning_rate": 4.710185263174863e-06, + "loss": 0.3663, + "step": 24210 + }, + { + "epoch": 0.5329297820823244, + "grad_norm": 1.4056095274647462, + "learning_rate": 4.709829406829937e-06, + "loss": 0.2285, + "step": 24211 + }, + { + "epoch": 0.5329517939687431, + "grad_norm": 1.8646077665378107, + "learning_rate": 4.70947355195979e-06, + "loss": 0.3462, + "step": 24212 + }, + { + "epoch": 0.5329738058551617, + "grad_norm": 1.4211597257291395, + "learning_rate": 4.709117698566227e-06, + "loss": 0.2911, + "step": 24213 + }, + { + "epoch": 0.5329958177415804, + "grad_norm": 1.6659231808462633, + "learning_rate": 4.708761846651056e-06, + "loss": 0.262, + "step": 24214 + }, + { + "epoch": 0.5330178296279992, + "grad_norm": 1.7649817479467393, + "learning_rate": 4.708405996216088e-06, + "loss": 0.3383, + "step": 24215 + }, + { + "epoch": 0.5330398415144177, + "grad_norm": 1.8822465365971264, + "learning_rate": 4.70805014726313e-06, + "loss": 0.3742, + "step": 24216 + }, + { + "epoch": 0.5330618534008364, + "grad_norm": 1.6616923067250375, + "learning_rate": 4.707694299793995e-06, + "loss": 0.2569, + "step": 24217 + }, + { + "epoch": 0.5330838652872552, + "grad_norm": 0.6161461019095408, + "learning_rate": 4.707338453810485e-06, + "loss": 0.4874, + "step": 24218 + }, + { + "epoch": 0.5331058771736737, + "grad_norm": 3.1255052741576375, + "learning_rate": 4.706982609314411e-06, + "loss": 0.2618, + "step": 24219 + }, + { + "epoch": 0.5331278890600925, + "grad_norm": 1.4886265009743165, + "learning_rate": 4.706626766307583e-06, + "loss": 0.3283, + "step": 24220 + }, + { + "epoch": 0.5331499009465112, + "grad_norm": 1.335380028306351, + "learning_rate": 4.706270924791807e-06, + "loss": 0.3362, + "step": 24221 + }, + { + "epoch": 0.5331719128329298, + "grad_norm": 1.8419031824707228, + "learning_rate": 4.705915084768897e-06, + "loss": 0.3823, + "step": 24222 + }, + { + "epoch": 0.5331939247193485, + "grad_norm": 1.9749077600375138, + "learning_rate": 4.705559246240655e-06, + "loss": 0.3711, + "step": 24223 + }, + { + "epoch": 0.5332159366057672, + "grad_norm": 1.4996686900220595, + "learning_rate": 4.705203409208892e-06, + "loss": 0.2888, + "step": 24224 + }, + { + "epoch": 0.5332379484921858, + "grad_norm": 1.321443264938793, + "learning_rate": 4.704847573675418e-06, + "loss": 0.3153, + "step": 24225 + }, + { + "epoch": 0.5332599603786045, + "grad_norm": 1.2806004651926515, + "learning_rate": 4.7044917396420385e-06, + "loss": 0.2373, + "step": 24226 + }, + { + "epoch": 0.5332819722650232, + "grad_norm": 5.3979895419684425, + "learning_rate": 4.704135907110566e-06, + "loss": 0.3205, + "step": 24227 + }, + { + "epoch": 0.5333039841514418, + "grad_norm": 0.6288453529566341, + "learning_rate": 4.703780076082807e-06, + "loss": 0.4621, + "step": 24228 + }, + { + "epoch": 0.5333259960378605, + "grad_norm": 0.6160071014547627, + "learning_rate": 4.703424246560567e-06, + "loss": 0.4796, + "step": 24229 + }, + { + "epoch": 0.5333480079242792, + "grad_norm": 1.1991730234278615, + "learning_rate": 4.703068418545661e-06, + "loss": 0.267, + "step": 24230 + }, + { + "epoch": 0.5333700198106978, + "grad_norm": 1.4712089116551643, + "learning_rate": 4.702712592039891e-06, + "loss": 0.2356, + "step": 24231 + }, + { + "epoch": 0.5333920316971165, + "grad_norm": 1.3845426475563143, + "learning_rate": 4.702356767045068e-06, + "loss": 0.2629, + "step": 24232 + }, + { + "epoch": 0.5334140435835351, + "grad_norm": 1.9205734099424927, + "learning_rate": 4.702000943563001e-06, + "loss": 0.3493, + "step": 24233 + }, + { + "epoch": 0.5334360554699538, + "grad_norm": 2.3148378677292962, + "learning_rate": 4.7016451215954986e-06, + "loss": 0.2865, + "step": 24234 + }, + { + "epoch": 0.5334580673563725, + "grad_norm": 1.464010839890124, + "learning_rate": 4.701289301144369e-06, + "loss": 0.3281, + "step": 24235 + }, + { + "epoch": 0.5334800792427911, + "grad_norm": 1.2810534330414696, + "learning_rate": 4.7009334822114206e-06, + "loss": 0.3848, + "step": 24236 + }, + { + "epoch": 0.5335020911292098, + "grad_norm": 2.08041213910262, + "learning_rate": 4.7005776647984605e-06, + "loss": 0.3506, + "step": 24237 + }, + { + "epoch": 0.5335241030156285, + "grad_norm": 1.284727214967137, + "learning_rate": 4.700221848907299e-06, + "loss": 0.2724, + "step": 24238 + }, + { + "epoch": 0.5335461149020471, + "grad_norm": 1.3815249856615905, + "learning_rate": 4.699866034539743e-06, + "loss": 0.2596, + "step": 24239 + }, + { + "epoch": 0.5335681267884658, + "grad_norm": 1.791951587050005, + "learning_rate": 4.699510221697604e-06, + "loss": 0.3968, + "step": 24240 + }, + { + "epoch": 0.5335901386748845, + "grad_norm": 0.6333212999501949, + "learning_rate": 4.6991544103826855e-06, + "loss": 0.4533, + "step": 24241 + }, + { + "epoch": 0.5336121505613031, + "grad_norm": 0.6276873901667035, + "learning_rate": 4.698798600596799e-06, + "loss": 0.4883, + "step": 24242 + }, + { + "epoch": 0.5336341624477218, + "grad_norm": 1.7837030921066017, + "learning_rate": 4.698442792341753e-06, + "loss": 0.3161, + "step": 24243 + }, + { + "epoch": 0.5336561743341405, + "grad_norm": 1.4255559091482184, + "learning_rate": 4.6980869856193535e-06, + "loss": 0.269, + "step": 24244 + }, + { + "epoch": 0.5336781862205591, + "grad_norm": 0.5985291692354366, + "learning_rate": 4.697731180431413e-06, + "loss": 0.4323, + "step": 24245 + }, + { + "epoch": 0.5337001981069778, + "grad_norm": 1.568681819960109, + "learning_rate": 4.697375376779738e-06, + "loss": 0.3071, + "step": 24246 + }, + { + "epoch": 0.5337222099933965, + "grad_norm": 1.4949891929618693, + "learning_rate": 4.697019574666133e-06, + "loss": 0.2765, + "step": 24247 + }, + { + "epoch": 0.5337442218798151, + "grad_norm": 1.8104232977997183, + "learning_rate": 4.696663774092413e-06, + "loss": 0.2608, + "step": 24248 + }, + { + "epoch": 0.5337662337662338, + "grad_norm": 1.2084213142710323, + "learning_rate": 4.696307975060382e-06, + "loss": 0.2522, + "step": 24249 + }, + { + "epoch": 0.5337882456526525, + "grad_norm": 1.2539757707408434, + "learning_rate": 4.695952177571847e-06, + "loss": 0.2845, + "step": 24250 + }, + { + "epoch": 0.5338102575390711, + "grad_norm": 1.7227914947578882, + "learning_rate": 4.695596381628621e-06, + "loss": 0.2928, + "step": 24251 + }, + { + "epoch": 0.5338322694254898, + "grad_norm": 1.659716569698073, + "learning_rate": 4.695240587232508e-06, + "loss": 0.2948, + "step": 24252 + }, + { + "epoch": 0.5338542813119085, + "grad_norm": 1.6518612297859825, + "learning_rate": 4.6948847943853215e-06, + "loss": 0.3568, + "step": 24253 + }, + { + "epoch": 0.5338762931983271, + "grad_norm": 1.659945137521183, + "learning_rate": 4.694529003088866e-06, + "loss": 0.2697, + "step": 24254 + }, + { + "epoch": 0.5338983050847458, + "grad_norm": 1.3788886242385556, + "learning_rate": 4.6941732133449466e-06, + "loss": 0.3134, + "step": 24255 + }, + { + "epoch": 0.5339203169711644, + "grad_norm": 1.3493085552509896, + "learning_rate": 4.693817425155379e-06, + "loss": 0.2602, + "step": 24256 + }, + { + "epoch": 0.5339423288575831, + "grad_norm": 1.5030990119481875, + "learning_rate": 4.6934616385219654e-06, + "loss": 0.2561, + "step": 24257 + }, + { + "epoch": 0.5339643407440018, + "grad_norm": 1.236868250279708, + "learning_rate": 4.6931058534465185e-06, + "loss": 0.3589, + "step": 24258 + }, + { + "epoch": 0.5339863526304204, + "grad_norm": 1.4237376928274952, + "learning_rate": 4.6927500699308446e-06, + "loss": 0.3045, + "step": 24259 + }, + { + "epoch": 0.5340083645168391, + "grad_norm": 1.5244623760625662, + "learning_rate": 4.6923942879767496e-06, + "loss": 0.3012, + "step": 24260 + }, + { + "epoch": 0.5340303764032578, + "grad_norm": 1.6162402526484325, + "learning_rate": 4.692038507586045e-06, + "loss": 0.2935, + "step": 24261 + }, + { + "epoch": 0.5340523882896764, + "grad_norm": 1.0822173314511663, + "learning_rate": 4.6916827287605385e-06, + "loss": 0.2258, + "step": 24262 + }, + { + "epoch": 0.5340744001760951, + "grad_norm": 1.5151037180877025, + "learning_rate": 4.6913269515020385e-06, + "loss": 0.2803, + "step": 24263 + }, + { + "epoch": 0.5340964120625138, + "grad_norm": 0.6487900728483493, + "learning_rate": 4.690971175812352e-06, + "loss": 0.4797, + "step": 24264 + }, + { + "epoch": 0.5341184239489324, + "grad_norm": 1.2211109406990956, + "learning_rate": 4.690615401693287e-06, + "loss": 0.3246, + "step": 24265 + }, + { + "epoch": 0.5341404358353511, + "grad_norm": 2.0259176492618782, + "learning_rate": 4.690259629146654e-06, + "loss": 0.307, + "step": 24266 + }, + { + "epoch": 0.5341624477217698, + "grad_norm": 1.5213459814758954, + "learning_rate": 4.68990385817426e-06, + "loss": 0.3635, + "step": 24267 + }, + { + "epoch": 0.5341844596081884, + "grad_norm": 1.3812791258487134, + "learning_rate": 4.689548088777911e-06, + "loss": 0.4058, + "step": 24268 + }, + { + "epoch": 0.5342064714946071, + "grad_norm": 1.7110361761376902, + "learning_rate": 4.689192320959418e-06, + "loss": 0.3703, + "step": 24269 + }, + { + "epoch": 0.5342284833810258, + "grad_norm": 1.3120066505899401, + "learning_rate": 4.688836554720587e-06, + "loss": 0.3523, + "step": 24270 + }, + { + "epoch": 0.5342504952674444, + "grad_norm": 1.2785200199664373, + "learning_rate": 4.68848079006323e-06, + "loss": 0.3533, + "step": 24271 + }, + { + "epoch": 0.5342725071538631, + "grad_norm": 1.2214210090083018, + "learning_rate": 4.688125026989152e-06, + "loss": 0.2485, + "step": 24272 + }, + { + "epoch": 0.5342945190402818, + "grad_norm": 0.6243229411057775, + "learning_rate": 4.68776926550016e-06, + "loss": 0.4723, + "step": 24273 + }, + { + "epoch": 0.5343165309267004, + "grad_norm": 1.3223247329377088, + "learning_rate": 4.687413505598065e-06, + "loss": 0.2942, + "step": 24274 + }, + { + "epoch": 0.5343385428131191, + "grad_norm": 0.6168294625382905, + "learning_rate": 4.687057747284672e-06, + "loss": 0.4446, + "step": 24275 + }, + { + "epoch": 0.5343605546995378, + "grad_norm": 1.6402810758714037, + "learning_rate": 4.686701990561795e-06, + "loss": 0.4126, + "step": 24276 + }, + { + "epoch": 0.5343825665859564, + "grad_norm": 1.392650797244201, + "learning_rate": 4.686346235431236e-06, + "loss": 0.3194, + "step": 24277 + }, + { + "epoch": 0.5344045784723751, + "grad_norm": 1.3935902680379577, + "learning_rate": 4.685990481894804e-06, + "loss": 0.3056, + "step": 24278 + }, + { + "epoch": 0.5344265903587937, + "grad_norm": 1.5696335959032592, + "learning_rate": 4.68563472995431e-06, + "loss": 0.3223, + "step": 24279 + }, + { + "epoch": 0.5344486022452124, + "grad_norm": 1.3247835066429279, + "learning_rate": 4.68527897961156e-06, + "loss": 0.2715, + "step": 24280 + }, + { + "epoch": 0.5344706141316311, + "grad_norm": 1.2804976732504247, + "learning_rate": 4.684923230868364e-06, + "loss": 0.3602, + "step": 24281 + }, + { + "epoch": 0.5344926260180497, + "grad_norm": 1.6315123042664088, + "learning_rate": 4.684567483726528e-06, + "loss": 0.3745, + "step": 24282 + }, + { + "epoch": 0.5345146379044684, + "grad_norm": 1.6976050931127824, + "learning_rate": 4.684211738187859e-06, + "loss": 0.3191, + "step": 24283 + }, + { + "epoch": 0.5345366497908871, + "grad_norm": 1.9542096691378683, + "learning_rate": 4.68385599425417e-06, + "loss": 0.3322, + "step": 24284 + }, + { + "epoch": 0.5345586616773057, + "grad_norm": 1.3407349618875104, + "learning_rate": 4.683500251927262e-06, + "loss": 0.3063, + "step": 24285 + }, + { + "epoch": 0.5345806735637244, + "grad_norm": 1.1652664076008412, + "learning_rate": 4.68314451120895e-06, + "loss": 0.2905, + "step": 24286 + }, + { + "epoch": 0.5346026854501431, + "grad_norm": 2.2687061106819377, + "learning_rate": 4.682788772101038e-06, + "loss": 0.35, + "step": 24287 + }, + { + "epoch": 0.5346246973365617, + "grad_norm": 1.3018160597134723, + "learning_rate": 4.682433034605333e-06, + "loss": 0.3463, + "step": 24288 + }, + { + "epoch": 0.5346467092229804, + "grad_norm": 0.6633432751514678, + "learning_rate": 4.682077298723648e-06, + "loss": 0.5035, + "step": 24289 + }, + { + "epoch": 0.5346687211093991, + "grad_norm": 1.336462866671133, + "learning_rate": 4.681721564457788e-06, + "loss": 0.259, + "step": 24290 + }, + { + "epoch": 0.5346907329958177, + "grad_norm": 2.0123527530486314, + "learning_rate": 4.681365831809558e-06, + "loss": 0.4298, + "step": 24291 + }, + { + "epoch": 0.5347127448822364, + "grad_norm": 1.2123894493282092, + "learning_rate": 4.681010100780771e-06, + "loss": 0.2788, + "step": 24292 + }, + { + "epoch": 0.5347347567686551, + "grad_norm": 1.2734696659648566, + "learning_rate": 4.680654371373231e-06, + "loss": 0.3307, + "step": 24293 + }, + { + "epoch": 0.5347567686550737, + "grad_norm": 1.3858282695096529, + "learning_rate": 4.680298643588751e-06, + "loss": 0.3175, + "step": 24294 + }, + { + "epoch": 0.5347787805414924, + "grad_norm": 1.1781574660631868, + "learning_rate": 4.6799429174291344e-06, + "loss": 0.2732, + "step": 24295 + }, + { + "epoch": 0.5348007924279111, + "grad_norm": 1.3897356786982522, + "learning_rate": 4.67958719289619e-06, + "loss": 0.3348, + "step": 24296 + }, + { + "epoch": 0.5348228043143297, + "grad_norm": 0.617030772579953, + "learning_rate": 4.679231469991727e-06, + "loss": 0.4693, + "step": 24297 + }, + { + "epoch": 0.5348448162007484, + "grad_norm": 1.425859362492497, + "learning_rate": 4.678875748717552e-06, + "loss": 0.3126, + "step": 24298 + }, + { + "epoch": 0.5348668280871671, + "grad_norm": 1.0723246490488427, + "learning_rate": 4.678520029075475e-06, + "loss": 0.2462, + "step": 24299 + }, + { + "epoch": 0.5348888399735857, + "grad_norm": 1.7637245538830537, + "learning_rate": 4.678164311067303e-06, + "loss": 0.2831, + "step": 24300 + }, + { + "epoch": 0.5349108518600044, + "grad_norm": 1.353588907210936, + "learning_rate": 4.6778085946948406e-06, + "loss": 0.2947, + "step": 24301 + }, + { + "epoch": 0.534932863746423, + "grad_norm": 1.2273586390378386, + "learning_rate": 4.677452879959902e-06, + "loss": 0.289, + "step": 24302 + }, + { + "epoch": 0.5349548756328417, + "grad_norm": 1.3361615635994448, + "learning_rate": 4.6770971668642885e-06, + "loss": 0.27, + "step": 24303 + }, + { + "epoch": 0.5349768875192604, + "grad_norm": 1.5274800629503853, + "learning_rate": 4.676741455409814e-06, + "loss": 0.2898, + "step": 24304 + }, + { + "epoch": 0.534998899405679, + "grad_norm": 1.3269008564105476, + "learning_rate": 4.676385745598282e-06, + "loss": 0.3074, + "step": 24305 + }, + { + "epoch": 0.5350209112920977, + "grad_norm": 0.6346672902546437, + "learning_rate": 4.676030037431502e-06, + "loss": 0.4848, + "step": 24306 + }, + { + "epoch": 0.5350429231785164, + "grad_norm": 1.4115411708731778, + "learning_rate": 4.675674330911284e-06, + "loss": 0.3433, + "step": 24307 + }, + { + "epoch": 0.535064935064935, + "grad_norm": 0.642632628485464, + "learning_rate": 4.675318626039432e-06, + "loss": 0.5249, + "step": 24308 + }, + { + "epoch": 0.5350869469513537, + "grad_norm": 1.8353099141129712, + "learning_rate": 4.6749629228177545e-06, + "loss": 0.3031, + "step": 24309 + }, + { + "epoch": 0.5351089588377724, + "grad_norm": 1.6398563459165072, + "learning_rate": 4.674607221248062e-06, + "loss": 0.3544, + "step": 24310 + }, + { + "epoch": 0.535130970724191, + "grad_norm": 1.3502386326443403, + "learning_rate": 4.674251521332159e-06, + "loss": 0.3033, + "step": 24311 + }, + { + "epoch": 0.5351529826106097, + "grad_norm": 1.378440509070159, + "learning_rate": 4.673895823071858e-06, + "loss": 0.2955, + "step": 24312 + }, + { + "epoch": 0.5351749944970284, + "grad_norm": 0.6070548822609002, + "learning_rate": 4.673540126468963e-06, + "loss": 0.4774, + "step": 24313 + }, + { + "epoch": 0.535197006383447, + "grad_norm": 1.3490864950183223, + "learning_rate": 4.67318443152528e-06, + "loss": 0.2496, + "step": 24314 + }, + { + "epoch": 0.5352190182698657, + "grad_norm": 1.28951470348213, + "learning_rate": 4.672828738242621e-06, + "loss": 0.3317, + "step": 24315 + }, + { + "epoch": 0.5352410301562844, + "grad_norm": 1.4451086330865577, + "learning_rate": 4.6724730466227916e-06, + "loss": 0.3707, + "step": 24316 + }, + { + "epoch": 0.535263042042703, + "grad_norm": 1.4036976407828379, + "learning_rate": 4.672117356667603e-06, + "loss": 0.3001, + "step": 24317 + }, + { + "epoch": 0.5352850539291217, + "grad_norm": 2.1803490681641184, + "learning_rate": 4.6717616683788585e-06, + "loss": 0.2895, + "step": 24318 + }, + { + "epoch": 0.5353070658155404, + "grad_norm": 1.441273758910359, + "learning_rate": 4.671405981758367e-06, + "loss": 0.3921, + "step": 24319 + }, + { + "epoch": 0.535329077701959, + "grad_norm": 1.7221680476427736, + "learning_rate": 4.671050296807938e-06, + "loss": 0.4147, + "step": 24320 + }, + { + "epoch": 0.5353510895883777, + "grad_norm": 1.796532317159881, + "learning_rate": 4.670694613529376e-06, + "loss": 0.3055, + "step": 24321 + }, + { + "epoch": 0.5353731014747964, + "grad_norm": 1.883371177922237, + "learning_rate": 4.670338931924494e-06, + "loss": 0.351, + "step": 24322 + }, + { + "epoch": 0.535395113361215, + "grad_norm": 1.5018968427139405, + "learning_rate": 4.669983251995095e-06, + "loss": 0.3463, + "step": 24323 + }, + { + "epoch": 0.5354171252476337, + "grad_norm": 1.5714447227582296, + "learning_rate": 4.669627573742988e-06, + "loss": 0.3703, + "step": 24324 + }, + { + "epoch": 0.5354391371340523, + "grad_norm": 1.3611656216076369, + "learning_rate": 4.669271897169982e-06, + "loss": 0.2958, + "step": 24325 + }, + { + "epoch": 0.535461149020471, + "grad_norm": 1.5981917244780337, + "learning_rate": 4.668916222277882e-06, + "loss": 0.2897, + "step": 24326 + }, + { + "epoch": 0.5354831609068897, + "grad_norm": 0.659868204089871, + "learning_rate": 4.668560549068499e-06, + "loss": 0.4816, + "step": 24327 + }, + { + "epoch": 0.5355051727933083, + "grad_norm": 1.3891199398675789, + "learning_rate": 4.668204877543639e-06, + "loss": 0.2868, + "step": 24328 + }, + { + "epoch": 0.535527184679727, + "grad_norm": 0.6462170584393399, + "learning_rate": 4.667849207705108e-06, + "loss": 0.4756, + "step": 24329 + }, + { + "epoch": 0.5355491965661457, + "grad_norm": 1.4392802216760154, + "learning_rate": 4.667493539554719e-06, + "loss": 0.3219, + "step": 24330 + }, + { + "epoch": 0.5355712084525643, + "grad_norm": 1.305857920227348, + "learning_rate": 4.667137873094273e-06, + "loss": 0.313, + "step": 24331 + }, + { + "epoch": 0.535593220338983, + "grad_norm": 1.1884586746846686, + "learning_rate": 4.666782208325581e-06, + "loss": 0.2914, + "step": 24332 + }, + { + "epoch": 0.5356152322254017, + "grad_norm": 1.3993639746753268, + "learning_rate": 4.6664265452504515e-06, + "loss": 0.2952, + "step": 24333 + }, + { + "epoch": 0.5356372441118203, + "grad_norm": 1.5026127899769537, + "learning_rate": 4.666070883870689e-06, + "loss": 0.3, + "step": 24334 + }, + { + "epoch": 0.535659255998239, + "grad_norm": 1.2307912858517989, + "learning_rate": 4.665715224188107e-06, + "loss": 0.2979, + "step": 24335 + }, + { + "epoch": 0.5356812678846578, + "grad_norm": 1.4395934637476917, + "learning_rate": 4.665359566204508e-06, + "loss": 0.37, + "step": 24336 + }, + { + "epoch": 0.5357032797710763, + "grad_norm": 0.639842057390563, + "learning_rate": 4.665003909921699e-06, + "loss": 0.463, + "step": 24337 + }, + { + "epoch": 0.535725291657495, + "grad_norm": 1.8020784823013276, + "learning_rate": 4.664648255341492e-06, + "loss": 0.4184, + "step": 24338 + }, + { + "epoch": 0.5357473035439138, + "grad_norm": 1.382412346412425, + "learning_rate": 4.664292602465689e-06, + "loss": 0.2671, + "step": 24339 + }, + { + "epoch": 0.5357693154303323, + "grad_norm": 0.6178283687726983, + "learning_rate": 4.6639369512961035e-06, + "loss": 0.4668, + "step": 24340 + }, + { + "epoch": 0.535791327316751, + "grad_norm": 1.3460132073717408, + "learning_rate": 4.66358130183454e-06, + "loss": 0.3376, + "step": 24341 + }, + { + "epoch": 0.5358133392031698, + "grad_norm": 0.6212798682450983, + "learning_rate": 4.663225654082805e-06, + "loss": 0.4713, + "step": 24342 + }, + { + "epoch": 0.5358353510895884, + "grad_norm": 1.1415309107821385, + "learning_rate": 4.6628700080427094e-06, + "loss": 0.2917, + "step": 24343 + }, + { + "epoch": 0.5358573629760071, + "grad_norm": 1.7469434413533662, + "learning_rate": 4.662514363716057e-06, + "loss": 0.2993, + "step": 24344 + }, + { + "epoch": 0.5358793748624258, + "grad_norm": 1.6469960515560214, + "learning_rate": 4.662158721104657e-06, + "loss": 0.3612, + "step": 24345 + }, + { + "epoch": 0.5359013867488444, + "grad_norm": 1.900953707783868, + "learning_rate": 4.6618030802103184e-06, + "loss": 0.3919, + "step": 24346 + }, + { + "epoch": 0.5359233986352631, + "grad_norm": 2.3453961257138274, + "learning_rate": 4.661447441034846e-06, + "loss": 0.3243, + "step": 24347 + }, + { + "epoch": 0.5359454105216817, + "grad_norm": 0.648203235845132, + "learning_rate": 4.661091803580051e-06, + "loss": 0.4594, + "step": 24348 + }, + { + "epoch": 0.5359674224081004, + "grad_norm": 1.3987049510319172, + "learning_rate": 4.660736167847737e-06, + "loss": 0.3049, + "step": 24349 + }, + { + "epoch": 0.5359894342945191, + "grad_norm": 1.164889074888893, + "learning_rate": 4.660380533839712e-06, + "loss": 0.217, + "step": 24350 + }, + { + "epoch": 0.5360114461809377, + "grad_norm": 1.91609982027395, + "learning_rate": 4.660024901557787e-06, + "loss": 0.2358, + "step": 24351 + }, + { + "epoch": 0.5360334580673564, + "grad_norm": 0.6505980178655199, + "learning_rate": 4.659669271003765e-06, + "loss": 0.4934, + "step": 24352 + }, + { + "epoch": 0.5360554699537751, + "grad_norm": 1.7898746092862938, + "learning_rate": 4.659313642179457e-06, + "loss": 0.3349, + "step": 24353 + }, + { + "epoch": 0.5360774818401937, + "grad_norm": 1.5830838921022978, + "learning_rate": 4.658958015086669e-06, + "loss": 0.3926, + "step": 24354 + }, + { + "epoch": 0.5360994937266124, + "grad_norm": 1.8401427137378399, + "learning_rate": 4.658602389727207e-06, + "loss": 0.404, + "step": 24355 + }, + { + "epoch": 0.5361215056130311, + "grad_norm": 1.3637956718522446, + "learning_rate": 4.6582467661028825e-06, + "loss": 0.2596, + "step": 24356 + }, + { + "epoch": 0.5361435174994497, + "grad_norm": 1.7270977456151575, + "learning_rate": 4.657891144215497e-06, + "loss": 0.2907, + "step": 24357 + }, + { + "epoch": 0.5361655293858684, + "grad_norm": 6.367002002528483, + "learning_rate": 4.657535524066864e-06, + "loss": 0.3153, + "step": 24358 + }, + { + "epoch": 0.5361875412722871, + "grad_norm": 1.243889331757116, + "learning_rate": 4.657179905658788e-06, + "loss": 0.2809, + "step": 24359 + }, + { + "epoch": 0.5362095531587057, + "grad_norm": 1.2721672582969954, + "learning_rate": 4.656824288993074e-06, + "loss": 0.2639, + "step": 24360 + }, + { + "epoch": 0.5362315650451244, + "grad_norm": 1.2159014245413307, + "learning_rate": 4.656468674071535e-06, + "loss": 0.3182, + "step": 24361 + }, + { + "epoch": 0.5362535769315431, + "grad_norm": 1.4715820326915567, + "learning_rate": 4.656113060895973e-06, + "loss": 0.3148, + "step": 24362 + }, + { + "epoch": 0.5362755888179617, + "grad_norm": 1.873793505855733, + "learning_rate": 4.6557574494682e-06, + "loss": 0.3097, + "step": 24363 + }, + { + "epoch": 0.5362976007043804, + "grad_norm": 1.4073094023624022, + "learning_rate": 4.655401839790021e-06, + "loss": 0.3185, + "step": 24364 + }, + { + "epoch": 0.5363196125907991, + "grad_norm": 1.2988021330129704, + "learning_rate": 4.655046231863242e-06, + "loss": 0.2949, + "step": 24365 + }, + { + "epoch": 0.5363416244772177, + "grad_norm": 0.6572236934226343, + "learning_rate": 4.6546906256896736e-06, + "loss": 0.4768, + "step": 24366 + }, + { + "epoch": 0.5363636363636364, + "grad_norm": 1.2921165943817037, + "learning_rate": 4.654335021271121e-06, + "loss": 0.2441, + "step": 24367 + }, + { + "epoch": 0.5363856482500551, + "grad_norm": 1.5531062545066785, + "learning_rate": 4.6539794186093906e-06, + "loss": 0.2974, + "step": 24368 + }, + { + "epoch": 0.5364076601364737, + "grad_norm": 1.3266297109296226, + "learning_rate": 4.6536238177062935e-06, + "loss": 0.3545, + "step": 24369 + }, + { + "epoch": 0.5364296720228924, + "grad_norm": 1.9054749837132274, + "learning_rate": 4.653268218563631e-06, + "loss": 0.3387, + "step": 24370 + }, + { + "epoch": 0.5364516839093111, + "grad_norm": 0.5988014222122323, + "learning_rate": 4.652912621183217e-06, + "loss": 0.447, + "step": 24371 + }, + { + "epoch": 0.5364736957957297, + "grad_norm": 1.2901220968448994, + "learning_rate": 4.652557025566856e-06, + "loss": 0.2993, + "step": 24372 + }, + { + "epoch": 0.5364957076821484, + "grad_norm": 0.6423299726072488, + "learning_rate": 4.6522014317163535e-06, + "loss": 0.4895, + "step": 24373 + }, + { + "epoch": 0.536517719568567, + "grad_norm": 1.75705968446275, + "learning_rate": 4.6518458396335205e-06, + "loss": 0.358, + "step": 24374 + }, + { + "epoch": 0.5365397314549857, + "grad_norm": 1.3602630901500905, + "learning_rate": 4.651490249320159e-06, + "loss": 0.4924, + "step": 24375 + }, + { + "epoch": 0.5365617433414044, + "grad_norm": 2.292692914139928, + "learning_rate": 4.651134660778083e-06, + "loss": 0.3557, + "step": 24376 + }, + { + "epoch": 0.536583755227823, + "grad_norm": 1.4808510843373117, + "learning_rate": 4.650779074009094e-06, + "loss": 0.3171, + "step": 24377 + }, + { + "epoch": 0.5366057671142417, + "grad_norm": 1.077939637965602, + "learning_rate": 4.650423489015002e-06, + "loss": 0.2671, + "step": 24378 + }, + { + "epoch": 0.5366277790006604, + "grad_norm": 1.4050151968482618, + "learning_rate": 4.650067905797616e-06, + "loss": 0.2981, + "step": 24379 + }, + { + "epoch": 0.536649790887079, + "grad_norm": 1.6695581501625223, + "learning_rate": 4.649712324358737e-06, + "loss": 0.3084, + "step": 24380 + }, + { + "epoch": 0.5366718027734977, + "grad_norm": 1.4724065317775359, + "learning_rate": 4.649356744700178e-06, + "loss": 0.4027, + "step": 24381 + }, + { + "epoch": 0.5366938146599164, + "grad_norm": 1.4627344613985565, + "learning_rate": 4.6490011668237455e-06, + "loss": 0.3354, + "step": 24382 + }, + { + "epoch": 0.536715826546335, + "grad_norm": 1.270696691357768, + "learning_rate": 4.648645590731244e-06, + "loss": 0.2623, + "step": 24383 + }, + { + "epoch": 0.5367378384327537, + "grad_norm": 1.5108152735939873, + "learning_rate": 4.648290016424484e-06, + "loss": 0.3391, + "step": 24384 + }, + { + "epoch": 0.5367598503191724, + "grad_norm": 0.6145630413846673, + "learning_rate": 4.6479344439052695e-06, + "loss": 0.4508, + "step": 24385 + }, + { + "epoch": 0.536781862205591, + "grad_norm": 1.854923825083925, + "learning_rate": 4.64757887317541e-06, + "loss": 0.291, + "step": 24386 + }, + { + "epoch": 0.5368038740920097, + "grad_norm": 1.5084646704960774, + "learning_rate": 4.647223304236713e-06, + "loss": 0.2914, + "step": 24387 + }, + { + "epoch": 0.5368258859784284, + "grad_norm": 1.1209711624997245, + "learning_rate": 4.646867737090981e-06, + "loss": 0.2669, + "step": 24388 + }, + { + "epoch": 0.536847897864847, + "grad_norm": 0.6297171495778677, + "learning_rate": 4.646512171740028e-06, + "loss": 0.4841, + "step": 24389 + }, + { + "epoch": 0.5368699097512657, + "grad_norm": 1.2835200547590813, + "learning_rate": 4.646156608185656e-06, + "loss": 0.3123, + "step": 24390 + }, + { + "epoch": 0.5368919216376844, + "grad_norm": 1.3727610645744515, + "learning_rate": 4.645801046429674e-06, + "loss": 0.2402, + "step": 24391 + }, + { + "epoch": 0.536913933524103, + "grad_norm": 1.5766405681998372, + "learning_rate": 4.645445486473891e-06, + "loss": 0.3863, + "step": 24392 + }, + { + "epoch": 0.5369359454105217, + "grad_norm": 1.5486150891674362, + "learning_rate": 4.645089928320109e-06, + "loss": 0.4032, + "step": 24393 + }, + { + "epoch": 0.5369579572969404, + "grad_norm": 0.6175861945989285, + "learning_rate": 4.644734371970141e-06, + "loss": 0.4725, + "step": 24394 + }, + { + "epoch": 0.536979969183359, + "grad_norm": 1.728140678987246, + "learning_rate": 4.64437881742579e-06, + "loss": 0.2932, + "step": 24395 + }, + { + "epoch": 0.5370019810697777, + "grad_norm": 1.4310005390923644, + "learning_rate": 4.644023264688864e-06, + "loss": 0.4248, + "step": 24396 + }, + { + "epoch": 0.5370239929561963, + "grad_norm": 1.6123771371668187, + "learning_rate": 4.643667713761173e-06, + "loss": 0.324, + "step": 24397 + }, + { + "epoch": 0.537046004842615, + "grad_norm": 0.6317440720706465, + "learning_rate": 4.643312164644519e-06, + "loss": 0.4719, + "step": 24398 + }, + { + "epoch": 0.5370680167290337, + "grad_norm": 1.176559642757679, + "learning_rate": 4.642956617340713e-06, + "loss": 0.2871, + "step": 24399 + }, + { + "epoch": 0.5370900286154523, + "grad_norm": 1.1627619198537673, + "learning_rate": 4.64260107185156e-06, + "loss": 0.2644, + "step": 24400 + }, + { + "epoch": 0.537112040501871, + "grad_norm": 1.3732534684092839, + "learning_rate": 4.642245528178867e-06, + "loss": 0.3075, + "step": 24401 + }, + { + "epoch": 0.5371340523882897, + "grad_norm": 0.6104403385440286, + "learning_rate": 4.641889986324445e-06, + "loss": 0.449, + "step": 24402 + }, + { + "epoch": 0.5371560642747083, + "grad_norm": 1.3567840340598731, + "learning_rate": 4.6415344462900946e-06, + "loss": 0.3167, + "step": 24403 + }, + { + "epoch": 0.537178076161127, + "grad_norm": 1.7844157553064255, + "learning_rate": 4.641178908077627e-06, + "loss": 0.2281, + "step": 24404 + }, + { + "epoch": 0.5372000880475457, + "grad_norm": 1.547017042567915, + "learning_rate": 4.64082337168885e-06, + "loss": 0.2909, + "step": 24405 + }, + { + "epoch": 0.5372220999339643, + "grad_norm": 4.641282199102921, + "learning_rate": 4.640467837125565e-06, + "loss": 0.2779, + "step": 24406 + }, + { + "epoch": 0.537244111820383, + "grad_norm": 1.7035339325341103, + "learning_rate": 4.6401123043895865e-06, + "loss": 0.2988, + "step": 24407 + }, + { + "epoch": 0.5372661237068017, + "grad_norm": 1.532466774819921, + "learning_rate": 4.639756773482716e-06, + "loss": 0.2955, + "step": 24408 + }, + { + "epoch": 0.5372881355932203, + "grad_norm": 0.5924112979165436, + "learning_rate": 4.639401244406762e-06, + "loss": 0.4527, + "step": 24409 + }, + { + "epoch": 0.537310147479639, + "grad_norm": 1.4563026292324899, + "learning_rate": 4.639045717163533e-06, + "loss": 0.2843, + "step": 24410 + }, + { + "epoch": 0.5373321593660577, + "grad_norm": 1.5409436447483809, + "learning_rate": 4.638690191754832e-06, + "loss": 0.3262, + "step": 24411 + }, + { + "epoch": 0.5373541712524763, + "grad_norm": 1.1666299064942058, + "learning_rate": 4.638334668182472e-06, + "loss": 0.3287, + "step": 24412 + }, + { + "epoch": 0.537376183138895, + "grad_norm": 0.6090251831837407, + "learning_rate": 4.637979146448256e-06, + "loss": 0.4706, + "step": 24413 + }, + { + "epoch": 0.5373981950253137, + "grad_norm": 1.412514081806763, + "learning_rate": 4.6376236265539886e-06, + "loss": 0.368, + "step": 24414 + }, + { + "epoch": 0.5374202069117323, + "grad_norm": 1.360325055574604, + "learning_rate": 4.637268108501483e-06, + "loss": 0.2536, + "step": 24415 + }, + { + "epoch": 0.537442218798151, + "grad_norm": 1.2933373562343657, + "learning_rate": 4.63691259229254e-06, + "loss": 0.2944, + "step": 24416 + }, + { + "epoch": 0.5374642306845697, + "grad_norm": 1.2386162048462661, + "learning_rate": 4.6365570779289705e-06, + "loss": 0.2497, + "step": 24417 + }, + { + "epoch": 0.5374862425709883, + "grad_norm": 1.5306231157920795, + "learning_rate": 4.636201565412581e-06, + "loss": 0.2643, + "step": 24418 + }, + { + "epoch": 0.537508254457407, + "grad_norm": 1.034469048817585, + "learning_rate": 4.635846054745175e-06, + "loss": 0.1909, + "step": 24419 + }, + { + "epoch": 0.5375302663438256, + "grad_norm": 2.0685038880591518, + "learning_rate": 4.635490545928564e-06, + "loss": 0.3826, + "step": 24420 + }, + { + "epoch": 0.5375522782302443, + "grad_norm": 1.6383660384493597, + "learning_rate": 4.63513503896455e-06, + "loss": 0.3042, + "step": 24421 + }, + { + "epoch": 0.537574290116663, + "grad_norm": 1.1896118277986742, + "learning_rate": 4.634779533854944e-06, + "loss": 0.2591, + "step": 24422 + }, + { + "epoch": 0.5375963020030816, + "grad_norm": 1.2940064005036869, + "learning_rate": 4.634424030601553e-06, + "loss": 0.2842, + "step": 24423 + }, + { + "epoch": 0.5376183138895003, + "grad_norm": 1.2847013282086521, + "learning_rate": 4.634068529206178e-06, + "loss": 0.2893, + "step": 24424 + }, + { + "epoch": 0.537640325775919, + "grad_norm": 1.2249414395801816, + "learning_rate": 4.633713029670634e-06, + "loss": 0.2846, + "step": 24425 + }, + { + "epoch": 0.5376623376623376, + "grad_norm": 1.959535449058916, + "learning_rate": 4.633357531996721e-06, + "loss": 0.3411, + "step": 24426 + }, + { + "epoch": 0.5376843495487563, + "grad_norm": 1.340490360351156, + "learning_rate": 4.6330020361862495e-06, + "loss": 0.3836, + "step": 24427 + }, + { + "epoch": 0.537706361435175, + "grad_norm": 0.6592272657533821, + "learning_rate": 4.6326465422410274e-06, + "loss": 0.4867, + "step": 24428 + }, + { + "epoch": 0.5377283733215936, + "grad_norm": 1.3092613952407606, + "learning_rate": 4.632291050162856e-06, + "loss": 0.3009, + "step": 24429 + }, + { + "epoch": 0.5377503852080123, + "grad_norm": 1.3139542280850982, + "learning_rate": 4.631935559953548e-06, + "loss": 0.3212, + "step": 24430 + }, + { + "epoch": 0.537772397094431, + "grad_norm": 1.3959339324998066, + "learning_rate": 4.631580071614907e-06, + "loss": 0.3071, + "step": 24431 + }, + { + "epoch": 0.5377944089808496, + "grad_norm": 0.6090111908935781, + "learning_rate": 4.6312245851487385e-06, + "loss": 0.4472, + "step": 24432 + }, + { + "epoch": 0.5378164208672683, + "grad_norm": 1.2662635753088811, + "learning_rate": 4.630869100556854e-06, + "loss": 0.3808, + "step": 24433 + }, + { + "epoch": 0.537838432753687, + "grad_norm": 1.8651316876150903, + "learning_rate": 4.630513617841055e-06, + "loss": 0.4231, + "step": 24434 + }, + { + "epoch": 0.5378604446401056, + "grad_norm": 1.8777477870180794, + "learning_rate": 4.630158137003152e-06, + "loss": 0.2958, + "step": 24435 + }, + { + "epoch": 0.5378824565265243, + "grad_norm": 1.910119678811877, + "learning_rate": 4.62980265804495e-06, + "loss": 0.4247, + "step": 24436 + }, + { + "epoch": 0.537904468412943, + "grad_norm": 1.564722791740377, + "learning_rate": 4.629447180968255e-06, + "loss": 0.2985, + "step": 24437 + }, + { + "epoch": 0.5379264802993616, + "grad_norm": 0.63428266212687, + "learning_rate": 4.629091705774878e-06, + "loss": 0.5138, + "step": 24438 + }, + { + "epoch": 0.5379484921857803, + "grad_norm": 0.6123243381222121, + "learning_rate": 4.628736232466618e-06, + "loss": 0.4825, + "step": 24439 + }, + { + "epoch": 0.537970504072199, + "grad_norm": 1.5818954613030365, + "learning_rate": 4.628380761045289e-06, + "loss": 0.3331, + "step": 24440 + }, + { + "epoch": 0.5379925159586176, + "grad_norm": 2.752642484017383, + "learning_rate": 4.628025291512695e-06, + "loss": 0.4391, + "step": 24441 + }, + { + "epoch": 0.5380145278450363, + "grad_norm": 1.4724628360107643, + "learning_rate": 4.627669823870639e-06, + "loss": 0.3662, + "step": 24442 + }, + { + "epoch": 0.5380365397314549, + "grad_norm": 0.583484190835391, + "learning_rate": 4.627314358120935e-06, + "loss": 0.4835, + "step": 24443 + }, + { + "epoch": 0.5380585516178736, + "grad_norm": 1.3198490700591885, + "learning_rate": 4.626958894265382e-06, + "loss": 0.2768, + "step": 24444 + }, + { + "epoch": 0.5380805635042923, + "grad_norm": 1.7768643575538066, + "learning_rate": 4.626603432305793e-06, + "loss": 0.2928, + "step": 24445 + }, + { + "epoch": 0.5381025753907109, + "grad_norm": 1.3396057261160408, + "learning_rate": 4.626247972243972e-06, + "loss": 0.3323, + "step": 24446 + }, + { + "epoch": 0.5381245872771296, + "grad_norm": 0.6400041091167121, + "learning_rate": 4.625892514081722e-06, + "loss": 0.4853, + "step": 24447 + }, + { + "epoch": 0.5381465991635483, + "grad_norm": 4.10225571368569, + "learning_rate": 4.625537057820857e-06, + "loss": 0.4214, + "step": 24448 + }, + { + "epoch": 0.5381686110499669, + "grad_norm": 1.1928481400626596, + "learning_rate": 4.625181603463179e-06, + "loss": 0.2295, + "step": 24449 + }, + { + "epoch": 0.5381906229363856, + "grad_norm": 0.6156689270851907, + "learning_rate": 4.624826151010494e-06, + "loss": 0.4712, + "step": 24450 + }, + { + "epoch": 0.5382126348228043, + "grad_norm": 1.44944096727645, + "learning_rate": 4.6244707004646115e-06, + "loss": 0.3305, + "step": 24451 + }, + { + "epoch": 0.5382346467092229, + "grad_norm": 1.7299195754966223, + "learning_rate": 4.624115251827334e-06, + "loss": 0.3492, + "step": 24452 + }, + { + "epoch": 0.5382566585956416, + "grad_norm": 1.7439390285640894, + "learning_rate": 4.623759805100472e-06, + "loss": 0.3522, + "step": 24453 + }, + { + "epoch": 0.5382786704820604, + "grad_norm": 1.2395237554630025, + "learning_rate": 4.623404360285831e-06, + "loss": 0.294, + "step": 24454 + }, + { + "epoch": 0.5383006823684789, + "grad_norm": 2.3645502779057446, + "learning_rate": 4.623048917385215e-06, + "loss": 0.3101, + "step": 24455 + }, + { + "epoch": 0.5383226942548976, + "grad_norm": 1.3710570328333287, + "learning_rate": 4.622693476400435e-06, + "loss": 0.3686, + "step": 24456 + }, + { + "epoch": 0.5383447061413164, + "grad_norm": 2.4140365335766747, + "learning_rate": 4.622338037333293e-06, + "loss": 0.22, + "step": 24457 + }, + { + "epoch": 0.538366718027735, + "grad_norm": 1.4263516949694706, + "learning_rate": 4.621982600185598e-06, + "loss": 0.3525, + "step": 24458 + }, + { + "epoch": 0.5383887299141537, + "grad_norm": 1.287772944372648, + "learning_rate": 4.621627164959158e-06, + "loss": 0.313, + "step": 24459 + }, + { + "epoch": 0.5384107418005724, + "grad_norm": 1.6190898855739129, + "learning_rate": 4.621271731655774e-06, + "loss": 0.372, + "step": 24460 + }, + { + "epoch": 0.538432753686991, + "grad_norm": 2.247095618877364, + "learning_rate": 4.620916300277259e-06, + "loss": 0.3121, + "step": 24461 + }, + { + "epoch": 0.5384547655734097, + "grad_norm": 1.2950449701590105, + "learning_rate": 4.620560870825415e-06, + "loss": 0.3563, + "step": 24462 + }, + { + "epoch": 0.5384767774598284, + "grad_norm": 1.3321906800577223, + "learning_rate": 4.62020544330205e-06, + "loss": 0.3351, + "step": 24463 + }, + { + "epoch": 0.538498789346247, + "grad_norm": 0.5836935078335719, + "learning_rate": 4.6198500177089715e-06, + "loss": 0.4331, + "step": 24464 + }, + { + "epoch": 0.5385208012326657, + "grad_norm": 1.208276088590884, + "learning_rate": 4.619494594047982e-06, + "loss": 0.2714, + "step": 24465 + }, + { + "epoch": 0.5385428131190843, + "grad_norm": 1.272683213747142, + "learning_rate": 4.619139172320893e-06, + "loss": 0.2897, + "step": 24466 + }, + { + "epoch": 0.538564825005503, + "grad_norm": 1.837425201155863, + "learning_rate": 4.618783752529508e-06, + "loss": 0.2822, + "step": 24467 + }, + { + "epoch": 0.5385868368919217, + "grad_norm": 1.748967564004906, + "learning_rate": 4.618428334675634e-06, + "loss": 0.3116, + "step": 24468 + }, + { + "epoch": 0.5386088487783403, + "grad_norm": 1.463175346744923, + "learning_rate": 4.618072918761079e-06, + "loss": 0.3634, + "step": 24469 + }, + { + "epoch": 0.538630860664759, + "grad_norm": 1.41416961647466, + "learning_rate": 4.617717504787645e-06, + "loss": 0.4046, + "step": 24470 + }, + { + "epoch": 0.5386528725511777, + "grad_norm": 1.8367807591808645, + "learning_rate": 4.617362092757143e-06, + "loss": 0.3442, + "step": 24471 + }, + { + "epoch": 0.5386748844375963, + "grad_norm": 1.3169765910029279, + "learning_rate": 4.617006682671376e-06, + "loss": 0.2573, + "step": 24472 + }, + { + "epoch": 0.538696896324015, + "grad_norm": 1.4458154230205387, + "learning_rate": 4.616651274532152e-06, + "loss": 0.3127, + "step": 24473 + }, + { + "epoch": 0.5387189082104337, + "grad_norm": 1.3484227953826275, + "learning_rate": 4.61629586834128e-06, + "loss": 0.2112, + "step": 24474 + }, + { + "epoch": 0.5387409200968523, + "grad_norm": 1.3807013973948736, + "learning_rate": 4.61594046410056e-06, + "loss": 0.3275, + "step": 24475 + }, + { + "epoch": 0.538762931983271, + "grad_norm": 1.7223177977290367, + "learning_rate": 4.615585061811803e-06, + "loss": 0.3208, + "step": 24476 + }, + { + "epoch": 0.5387849438696897, + "grad_norm": 1.5283921218921948, + "learning_rate": 4.615229661476816e-06, + "loss": 0.3707, + "step": 24477 + }, + { + "epoch": 0.5388069557561083, + "grad_norm": 1.4298446036563532, + "learning_rate": 4.6148742630974e-06, + "loss": 0.3786, + "step": 24478 + }, + { + "epoch": 0.538828967642527, + "grad_norm": 0.6638729062398071, + "learning_rate": 4.614518866675368e-06, + "loss": 0.4695, + "step": 24479 + }, + { + "epoch": 0.5388509795289457, + "grad_norm": 1.4899317412477817, + "learning_rate": 4.614163472212521e-06, + "loss": 0.2866, + "step": 24480 + }, + { + "epoch": 0.5388729914153643, + "grad_norm": 1.186854602061531, + "learning_rate": 4.613808079710668e-06, + "loss": 0.2334, + "step": 24481 + }, + { + "epoch": 0.538895003301783, + "grad_norm": 0.6064326605056698, + "learning_rate": 4.613452689171616e-06, + "loss": 0.4528, + "step": 24482 + }, + { + "epoch": 0.5389170151882017, + "grad_norm": 2.006185392310625, + "learning_rate": 4.613097300597168e-06, + "loss": 0.3363, + "step": 24483 + }, + { + "epoch": 0.5389390270746203, + "grad_norm": 0.6090659786944423, + "learning_rate": 4.612741913989133e-06, + "loss": 0.4617, + "step": 24484 + }, + { + "epoch": 0.538961038961039, + "grad_norm": 0.614997906359348, + "learning_rate": 4.6123865293493145e-06, + "loss": 0.488, + "step": 24485 + }, + { + "epoch": 0.5389830508474577, + "grad_norm": 1.5811113381548472, + "learning_rate": 4.612031146679522e-06, + "loss": 0.2762, + "step": 24486 + }, + { + "epoch": 0.5390050627338763, + "grad_norm": 1.3123866611236865, + "learning_rate": 4.611675765981562e-06, + "loss": 0.3169, + "step": 24487 + }, + { + "epoch": 0.539027074620295, + "grad_norm": 0.6453157521153666, + "learning_rate": 4.611320387257236e-06, + "loss": 0.5089, + "step": 24488 + }, + { + "epoch": 0.5390490865067136, + "grad_norm": 1.4322404676253493, + "learning_rate": 4.610965010508355e-06, + "loss": 0.3182, + "step": 24489 + }, + { + "epoch": 0.5390710983931323, + "grad_norm": 1.1205307714769852, + "learning_rate": 4.610609635736723e-06, + "loss": 0.1998, + "step": 24490 + }, + { + "epoch": 0.539093110279551, + "grad_norm": 1.1777483116688343, + "learning_rate": 4.610254262944145e-06, + "loss": 0.1784, + "step": 24491 + }, + { + "epoch": 0.5391151221659696, + "grad_norm": 0.6326946160959934, + "learning_rate": 4.6098988921324305e-06, + "loss": 0.4673, + "step": 24492 + }, + { + "epoch": 0.5391371340523883, + "grad_norm": 1.561768464582091, + "learning_rate": 4.609543523303383e-06, + "loss": 0.2697, + "step": 24493 + }, + { + "epoch": 0.539159145938807, + "grad_norm": 1.6761439560711417, + "learning_rate": 4.6091881564588095e-06, + "loss": 0.2955, + "step": 24494 + }, + { + "epoch": 0.5391811578252256, + "grad_norm": 1.2407233363380856, + "learning_rate": 4.6088327916005174e-06, + "loss": 0.2911, + "step": 24495 + }, + { + "epoch": 0.5392031697116443, + "grad_norm": 0.604682259442425, + "learning_rate": 4.608477428730308e-06, + "loss": 0.4924, + "step": 24496 + }, + { + "epoch": 0.539225181598063, + "grad_norm": 1.3927318545514324, + "learning_rate": 4.608122067849995e-06, + "loss": 0.3648, + "step": 24497 + }, + { + "epoch": 0.5392471934844816, + "grad_norm": 13.74416595465086, + "learning_rate": 4.607766708961377e-06, + "loss": 0.3093, + "step": 24498 + }, + { + "epoch": 0.5392692053709003, + "grad_norm": 1.9283325378530711, + "learning_rate": 4.607411352066266e-06, + "loss": 0.3431, + "step": 24499 + }, + { + "epoch": 0.539291217257319, + "grad_norm": 1.5758664688976827, + "learning_rate": 4.6070559971664655e-06, + "loss": 0.3598, + "step": 24500 + }, + { + "epoch": 0.5393132291437376, + "grad_norm": 1.3987583134511221, + "learning_rate": 4.60670064426378e-06, + "loss": 0.3288, + "step": 24501 + }, + { + "epoch": 0.5393352410301563, + "grad_norm": 1.0069982451067812, + "learning_rate": 4.606345293360019e-06, + "loss": 0.226, + "step": 24502 + }, + { + "epoch": 0.539357252916575, + "grad_norm": 1.8057395995047187, + "learning_rate": 4.605989944456985e-06, + "loss": 0.364, + "step": 24503 + }, + { + "epoch": 0.5393792648029936, + "grad_norm": 0.6211727719580054, + "learning_rate": 4.605634597556487e-06, + "loss": 0.467, + "step": 24504 + }, + { + "epoch": 0.5394012766894123, + "grad_norm": 2.1703847783351393, + "learning_rate": 4.605279252660331e-06, + "loss": 0.2688, + "step": 24505 + }, + { + "epoch": 0.539423288575831, + "grad_norm": 1.2989155210042023, + "learning_rate": 4.60492390977032e-06, + "loss": 0.3229, + "step": 24506 + }, + { + "epoch": 0.5394453004622496, + "grad_norm": 1.2622216261027952, + "learning_rate": 4.6045685688882625e-06, + "loss": 0.3408, + "step": 24507 + }, + { + "epoch": 0.5394673123486683, + "grad_norm": 1.5977347766328955, + "learning_rate": 4.604213230015964e-06, + "loss": 0.336, + "step": 24508 + }, + { + "epoch": 0.539489324235087, + "grad_norm": 1.2020320543984457, + "learning_rate": 4.6038578931552295e-06, + "loss": 0.2115, + "step": 24509 + }, + { + "epoch": 0.5395113361215056, + "grad_norm": 1.3561553284693684, + "learning_rate": 4.603502558307867e-06, + "loss": 0.278, + "step": 24510 + }, + { + "epoch": 0.5395333480079243, + "grad_norm": 1.703808579787304, + "learning_rate": 4.603147225475681e-06, + "loss": 0.2853, + "step": 24511 + }, + { + "epoch": 0.5395553598943429, + "grad_norm": 1.6082192559620478, + "learning_rate": 4.602791894660478e-06, + "loss": 0.3477, + "step": 24512 + }, + { + "epoch": 0.5395773717807616, + "grad_norm": 1.6861945680246575, + "learning_rate": 4.602436565864065e-06, + "loss": 0.3599, + "step": 24513 + }, + { + "epoch": 0.5395993836671803, + "grad_norm": 1.8942237331184506, + "learning_rate": 4.602081239088243e-06, + "loss": 0.3607, + "step": 24514 + }, + { + "epoch": 0.5396213955535989, + "grad_norm": 1.5612753209101242, + "learning_rate": 4.601725914334825e-06, + "loss": 0.2815, + "step": 24515 + }, + { + "epoch": 0.5396434074400176, + "grad_norm": 1.4995775493437697, + "learning_rate": 4.601370591605612e-06, + "loss": 0.3258, + "step": 24516 + }, + { + "epoch": 0.5396654193264363, + "grad_norm": 1.5643385256581208, + "learning_rate": 4.601015270902411e-06, + "loss": 0.211, + "step": 24517 + }, + { + "epoch": 0.5396874312128549, + "grad_norm": 1.4626494435260895, + "learning_rate": 4.600659952227031e-06, + "loss": 0.3914, + "step": 24518 + }, + { + "epoch": 0.5397094430992736, + "grad_norm": 1.4879616803725912, + "learning_rate": 4.600304635581273e-06, + "loss": 0.3681, + "step": 24519 + }, + { + "epoch": 0.5397314549856923, + "grad_norm": 1.2635707226144341, + "learning_rate": 4.599949320966946e-06, + "loss": 0.2752, + "step": 24520 + }, + { + "epoch": 0.5397534668721109, + "grad_norm": 1.646147210154037, + "learning_rate": 4.5995940083858526e-06, + "loss": 0.253, + "step": 24521 + }, + { + "epoch": 0.5397754787585296, + "grad_norm": 1.2893427483601023, + "learning_rate": 4.5992386978398035e-06, + "loss": 0.3522, + "step": 24522 + }, + { + "epoch": 0.5397974906449483, + "grad_norm": 1.444217444465969, + "learning_rate": 4.598883389330603e-06, + "loss": 0.3578, + "step": 24523 + }, + { + "epoch": 0.5398195025313669, + "grad_norm": 0.6518753362131002, + "learning_rate": 4.598528082860053e-06, + "loss": 0.4859, + "step": 24524 + }, + { + "epoch": 0.5398415144177856, + "grad_norm": 1.271471376799299, + "learning_rate": 4.5981727784299645e-06, + "loss": 0.3074, + "step": 24525 + }, + { + "epoch": 0.5398635263042043, + "grad_norm": 1.1889240103718857, + "learning_rate": 4.59781747604214e-06, + "loss": 0.33, + "step": 24526 + }, + { + "epoch": 0.5398855381906229, + "grad_norm": 1.3245744685826566, + "learning_rate": 4.5974621756983864e-06, + "loss": 0.2905, + "step": 24527 + }, + { + "epoch": 0.5399075500770416, + "grad_norm": 1.7954457940336876, + "learning_rate": 4.597106877400512e-06, + "loss": 0.3818, + "step": 24528 + }, + { + "epoch": 0.5399295619634603, + "grad_norm": 1.4645686538604785, + "learning_rate": 4.596751581150317e-06, + "loss": 0.2326, + "step": 24529 + }, + { + "epoch": 0.5399515738498789, + "grad_norm": 1.105456728341392, + "learning_rate": 4.596396286949612e-06, + "loss": 0.2072, + "step": 24530 + }, + { + "epoch": 0.5399735857362976, + "grad_norm": 0.5957894819565102, + "learning_rate": 4.596040994800202e-06, + "loss": 0.4634, + "step": 24531 + }, + { + "epoch": 0.5399955976227163, + "grad_norm": 0.6077361309082618, + "learning_rate": 4.595685704703889e-06, + "loss": 0.4402, + "step": 24532 + }, + { + "epoch": 0.5400176095091349, + "grad_norm": 2.305816082652097, + "learning_rate": 4.595330416662484e-06, + "loss": 0.405, + "step": 24533 + }, + { + "epoch": 0.5400396213955536, + "grad_norm": 1.5578077475231766, + "learning_rate": 4.594975130677788e-06, + "loss": 0.3034, + "step": 24534 + }, + { + "epoch": 0.5400616332819723, + "grad_norm": 1.5385183132991285, + "learning_rate": 4.5946198467516115e-06, + "loss": 0.3157, + "step": 24535 + }, + { + "epoch": 0.5400836451683909, + "grad_norm": 1.4567994827733546, + "learning_rate": 4.594264564885758e-06, + "loss": 0.3195, + "step": 24536 + }, + { + "epoch": 0.5401056570548096, + "grad_norm": 1.4768622196080738, + "learning_rate": 4.593909285082031e-06, + "loss": 0.2793, + "step": 24537 + }, + { + "epoch": 0.5401276689412282, + "grad_norm": 1.146957140164239, + "learning_rate": 4.593554007342239e-06, + "loss": 0.3261, + "step": 24538 + }, + { + "epoch": 0.5401496808276469, + "grad_norm": 2.1127890524025466, + "learning_rate": 4.593198731668185e-06, + "loss": 0.326, + "step": 24539 + }, + { + "epoch": 0.5401716927140656, + "grad_norm": 1.5494758749556845, + "learning_rate": 4.592843458061679e-06, + "loss": 0.2443, + "step": 24540 + }, + { + "epoch": 0.5401937046004842, + "grad_norm": 1.1999581139000803, + "learning_rate": 4.592488186524524e-06, + "loss": 0.2771, + "step": 24541 + }, + { + "epoch": 0.5402157164869029, + "grad_norm": 1.836248129588349, + "learning_rate": 4.592132917058524e-06, + "loss": 0.3099, + "step": 24542 + }, + { + "epoch": 0.5402377283733216, + "grad_norm": 1.3950165346803425, + "learning_rate": 4.591777649665488e-06, + "loss": 0.3449, + "step": 24543 + }, + { + "epoch": 0.5402597402597402, + "grad_norm": 1.3893097017375626, + "learning_rate": 4.591422384347218e-06, + "loss": 0.3154, + "step": 24544 + }, + { + "epoch": 0.5402817521461589, + "grad_norm": 1.4569483568406334, + "learning_rate": 4.591067121105524e-06, + "loss": 0.3231, + "step": 24545 + }, + { + "epoch": 0.5403037640325776, + "grad_norm": 2.049096904972248, + "learning_rate": 4.59071185994221e-06, + "loss": 0.3155, + "step": 24546 + }, + { + "epoch": 0.5403257759189962, + "grad_norm": 0.6381752628517923, + "learning_rate": 4.590356600859078e-06, + "loss": 0.4878, + "step": 24547 + }, + { + "epoch": 0.5403477878054149, + "grad_norm": 1.4221473911596578, + "learning_rate": 4.590001343857938e-06, + "loss": 0.3443, + "step": 24548 + }, + { + "epoch": 0.5403697996918336, + "grad_norm": 1.4222835032645962, + "learning_rate": 4.589646088940595e-06, + "loss": 0.3991, + "step": 24549 + }, + { + "epoch": 0.5403918115782522, + "grad_norm": 1.1754599315260306, + "learning_rate": 4.589290836108851e-06, + "loss": 0.2589, + "step": 24550 + }, + { + "epoch": 0.5404138234646709, + "grad_norm": 1.3618835701300616, + "learning_rate": 4.5889355853645164e-06, + "loss": 0.3754, + "step": 24551 + }, + { + "epoch": 0.5404358353510896, + "grad_norm": 0.6409918777479392, + "learning_rate": 4.588580336709393e-06, + "loss": 0.4667, + "step": 24552 + }, + { + "epoch": 0.5404578472375082, + "grad_norm": 1.3053302048779556, + "learning_rate": 4.588225090145289e-06, + "loss": 0.3189, + "step": 24553 + }, + { + "epoch": 0.5404798591239269, + "grad_norm": 1.3106274242932545, + "learning_rate": 4.58786984567401e-06, + "loss": 0.318, + "step": 24554 + }, + { + "epoch": 0.5405018710103456, + "grad_norm": 1.6042163375745286, + "learning_rate": 4.587514603297357e-06, + "loss": 0.2699, + "step": 24555 + }, + { + "epoch": 0.5405238828967642, + "grad_norm": 1.6112391810490554, + "learning_rate": 4.5871593630171405e-06, + "loss": 0.3313, + "step": 24556 + }, + { + "epoch": 0.5405458947831829, + "grad_norm": 1.1413953841127782, + "learning_rate": 4.586804124835163e-06, + "loss": 0.2651, + "step": 24557 + }, + { + "epoch": 0.5405679066696016, + "grad_norm": 2.1207975513788675, + "learning_rate": 4.586448888753233e-06, + "loss": 0.3375, + "step": 24558 + }, + { + "epoch": 0.5405899185560202, + "grad_norm": 2.6985010362693567, + "learning_rate": 4.586093654773154e-06, + "loss": 0.3197, + "step": 24559 + }, + { + "epoch": 0.5406119304424389, + "grad_norm": 1.2912914146327978, + "learning_rate": 4.585738422896731e-06, + "loss": 0.321, + "step": 24560 + }, + { + "epoch": 0.5406339423288575, + "grad_norm": 2.359105119959502, + "learning_rate": 4.58538319312577e-06, + "loss": 0.318, + "step": 24561 + }, + { + "epoch": 0.5406559542152762, + "grad_norm": 2.105002769340467, + "learning_rate": 4.585027965462076e-06, + "loss": 0.3267, + "step": 24562 + }, + { + "epoch": 0.5406779661016949, + "grad_norm": 1.2809140883258379, + "learning_rate": 4.584672739907456e-06, + "loss": 0.2891, + "step": 24563 + }, + { + "epoch": 0.5406999779881135, + "grad_norm": 1.2482333719493317, + "learning_rate": 4.584317516463715e-06, + "loss": 0.3207, + "step": 24564 + }, + { + "epoch": 0.5407219898745322, + "grad_norm": 1.2371479495820406, + "learning_rate": 4.583962295132656e-06, + "loss": 0.2368, + "step": 24565 + }, + { + "epoch": 0.5407440017609509, + "grad_norm": 2.3639139684423935, + "learning_rate": 4.583607075916087e-06, + "loss": 0.2179, + "step": 24566 + }, + { + "epoch": 0.5407660136473695, + "grad_norm": 1.1236970612455248, + "learning_rate": 4.583251858815814e-06, + "loss": 0.2408, + "step": 24567 + }, + { + "epoch": 0.5407880255337882, + "grad_norm": 0.6348333576095403, + "learning_rate": 4.5828966438336375e-06, + "loss": 0.477, + "step": 24568 + }, + { + "epoch": 0.540810037420207, + "grad_norm": 1.6240135507973017, + "learning_rate": 4.58254143097137e-06, + "loss": 0.3499, + "step": 24569 + }, + { + "epoch": 0.5408320493066255, + "grad_norm": 1.3128147663797507, + "learning_rate": 4.58218622023081e-06, + "loss": 0.3073, + "step": 24570 + }, + { + "epoch": 0.5408540611930442, + "grad_norm": 1.5255395654370867, + "learning_rate": 4.5818310116137675e-06, + "loss": 0.2975, + "step": 24571 + }, + { + "epoch": 0.540876073079463, + "grad_norm": 1.3851274671656617, + "learning_rate": 4.5814758051220484e-06, + "loss": 0.3521, + "step": 24572 + }, + { + "epoch": 0.5408980849658815, + "grad_norm": 3.8133983425412423, + "learning_rate": 4.581120600757453e-06, + "loss": 0.3151, + "step": 24573 + }, + { + "epoch": 0.5409200968523002, + "grad_norm": 1.250970853251242, + "learning_rate": 4.5807653985217895e-06, + "loss": 0.284, + "step": 24574 + }, + { + "epoch": 0.540942108738719, + "grad_norm": 1.2950577973552617, + "learning_rate": 4.580410198416864e-06, + "loss": 0.2861, + "step": 24575 + }, + { + "epoch": 0.5409641206251375, + "grad_norm": 1.3674815553037618, + "learning_rate": 4.580055000444481e-06, + "loss": 0.287, + "step": 24576 + }, + { + "epoch": 0.5409861325115563, + "grad_norm": 1.3298076420719442, + "learning_rate": 4.579699804606447e-06, + "loss": 0.3743, + "step": 24577 + }, + { + "epoch": 0.541008144397975, + "grad_norm": 6.763774708109274, + "learning_rate": 4.5793446109045645e-06, + "loss": 0.2528, + "step": 24578 + }, + { + "epoch": 0.5410301562843935, + "grad_norm": 1.6531874415608867, + "learning_rate": 4.578989419340641e-06, + "loss": 0.375, + "step": 24579 + }, + { + "epoch": 0.5410521681708123, + "grad_norm": 0.6255712437663126, + "learning_rate": 4.5786342299164795e-06, + "loss": 0.4494, + "step": 24580 + }, + { + "epoch": 0.541074180057231, + "grad_norm": 1.796555571366653, + "learning_rate": 4.578279042633888e-06, + "loss": 0.2793, + "step": 24581 + }, + { + "epoch": 0.5410961919436496, + "grad_norm": 1.9371374535929538, + "learning_rate": 4.577923857494672e-06, + "loss": 0.307, + "step": 24582 + }, + { + "epoch": 0.5411182038300683, + "grad_norm": 2.000283508508323, + "learning_rate": 4.5775686745006326e-06, + "loss": 0.314, + "step": 24583 + }, + { + "epoch": 0.5411402157164868, + "grad_norm": 1.3821770773470727, + "learning_rate": 4.577213493653579e-06, + "loss": 0.2861, + "step": 24584 + }, + { + "epoch": 0.5411622276029056, + "grad_norm": 0.6181973400338965, + "learning_rate": 4.576858314955314e-06, + "loss": 0.4725, + "step": 24585 + }, + { + "epoch": 0.5411842394893243, + "grad_norm": 1.5310352865387036, + "learning_rate": 4.576503138407644e-06, + "loss": 0.2592, + "step": 24586 + }, + { + "epoch": 0.5412062513757429, + "grad_norm": 0.6208860471391897, + "learning_rate": 4.576147964012376e-06, + "loss": 0.4886, + "step": 24587 + }, + { + "epoch": 0.5412282632621616, + "grad_norm": 1.7657986125352712, + "learning_rate": 4.575792791771311e-06, + "loss": 0.2695, + "step": 24588 + }, + { + "epoch": 0.5412502751485803, + "grad_norm": 1.8377953957009732, + "learning_rate": 4.575437621686257e-06, + "loss": 0.3532, + "step": 24589 + }, + { + "epoch": 0.5412722870349989, + "grad_norm": 0.6096354310667679, + "learning_rate": 4.575082453759019e-06, + "loss": 0.4599, + "step": 24590 + }, + { + "epoch": 0.5412942989214176, + "grad_norm": 0.6177705354211454, + "learning_rate": 4.5747272879913995e-06, + "loss": 0.4902, + "step": 24591 + }, + { + "epoch": 0.5413163108078363, + "grad_norm": 1.3592102505584909, + "learning_rate": 4.574372124385207e-06, + "loss": 0.2689, + "step": 24592 + }, + { + "epoch": 0.5413383226942549, + "grad_norm": 1.5055365213812921, + "learning_rate": 4.574016962942245e-06, + "loss": 0.2603, + "step": 24593 + }, + { + "epoch": 0.5413603345806736, + "grad_norm": 1.2560012280336208, + "learning_rate": 4.573661803664318e-06, + "loss": 0.2428, + "step": 24594 + }, + { + "epoch": 0.5413823464670923, + "grad_norm": 0.6042172745774219, + "learning_rate": 4.573306646553234e-06, + "loss": 0.4363, + "step": 24595 + }, + { + "epoch": 0.5414043583535109, + "grad_norm": 1.612453288837779, + "learning_rate": 4.572951491610793e-06, + "loss": 0.3455, + "step": 24596 + }, + { + "epoch": 0.5414263702399296, + "grad_norm": 1.4790073038307572, + "learning_rate": 4.572596338838805e-06, + "loss": 0.3168, + "step": 24597 + }, + { + "epoch": 0.5414483821263483, + "grad_norm": 3.16507928251064, + "learning_rate": 4.572241188239072e-06, + "loss": 0.2906, + "step": 24598 + }, + { + "epoch": 0.5414703940127669, + "grad_norm": 1.5122156387409282, + "learning_rate": 4.571886039813401e-06, + "loss": 0.2744, + "step": 24599 + }, + { + "epoch": 0.5414924058991856, + "grad_norm": 0.5941305119493899, + "learning_rate": 4.571530893563597e-06, + "loss": 0.4698, + "step": 24600 + }, + { + "epoch": 0.5415144177856043, + "grad_norm": 1.7291499958625265, + "learning_rate": 4.571175749491461e-06, + "loss": 0.2377, + "step": 24601 + }, + { + "epoch": 0.5415364296720229, + "grad_norm": 1.5149826052244273, + "learning_rate": 4.570820607598803e-06, + "loss": 0.2787, + "step": 24602 + }, + { + "epoch": 0.5415584415584416, + "grad_norm": 0.611522075838695, + "learning_rate": 4.570465467887425e-06, + "loss": 0.4501, + "step": 24603 + }, + { + "epoch": 0.5415804534448603, + "grad_norm": 1.8154971265678732, + "learning_rate": 4.570110330359135e-06, + "loss": 0.3278, + "step": 24604 + }, + { + "epoch": 0.5416024653312789, + "grad_norm": 2.3247031451434976, + "learning_rate": 4.569755195015736e-06, + "loss": 0.2599, + "step": 24605 + }, + { + "epoch": 0.5416244772176976, + "grad_norm": 1.3845897005543948, + "learning_rate": 4.569400061859031e-06, + "loss": 0.2041, + "step": 24606 + }, + { + "epoch": 0.5416464891041162, + "grad_norm": 1.2205165837554828, + "learning_rate": 4.569044930890829e-06, + "loss": 0.278, + "step": 24607 + }, + { + "epoch": 0.5416685009905349, + "grad_norm": 1.5279191183914933, + "learning_rate": 4.568689802112932e-06, + "loss": 0.318, + "step": 24608 + }, + { + "epoch": 0.5416905128769536, + "grad_norm": 1.5296228496162338, + "learning_rate": 4.5683346755271455e-06, + "loss": 0.3293, + "step": 24609 + }, + { + "epoch": 0.5417125247633722, + "grad_norm": 1.2580626860359376, + "learning_rate": 4.567979551135275e-06, + "loss": 0.3437, + "step": 24610 + }, + { + "epoch": 0.5417345366497909, + "grad_norm": 1.526993665596192, + "learning_rate": 4.567624428939124e-06, + "loss": 0.3321, + "step": 24611 + }, + { + "epoch": 0.5417565485362096, + "grad_norm": 1.4839960889091361, + "learning_rate": 4.5672693089405e-06, + "loss": 0.2051, + "step": 24612 + }, + { + "epoch": 0.5417785604226282, + "grad_norm": 1.3224787663182087, + "learning_rate": 4.566914191141207e-06, + "loss": 0.2505, + "step": 24613 + }, + { + "epoch": 0.5418005723090469, + "grad_norm": 1.650211594633263, + "learning_rate": 4.5665590755430476e-06, + "loss": 0.4099, + "step": 24614 + }, + { + "epoch": 0.5418225841954656, + "grad_norm": 1.3002134616693335, + "learning_rate": 4.566203962147829e-06, + "loss": 0.3541, + "step": 24615 + }, + { + "epoch": 0.5418445960818842, + "grad_norm": 1.3565826761075575, + "learning_rate": 4.565848850957354e-06, + "loss": 0.2731, + "step": 24616 + }, + { + "epoch": 0.5418666079683029, + "grad_norm": 0.6451930115915472, + "learning_rate": 4.5654937419734295e-06, + "loss": 0.4636, + "step": 24617 + }, + { + "epoch": 0.5418886198547216, + "grad_norm": 1.8857844894989293, + "learning_rate": 4.565138635197862e-06, + "loss": 0.3194, + "step": 24618 + }, + { + "epoch": 0.5419106317411402, + "grad_norm": 1.77699639761898, + "learning_rate": 4.564783530632451e-06, + "loss": 0.3515, + "step": 24619 + }, + { + "epoch": 0.5419326436275589, + "grad_norm": 1.7870567324756153, + "learning_rate": 4.564428428279005e-06, + "loss": 0.2938, + "step": 24620 + }, + { + "epoch": 0.5419546555139776, + "grad_norm": 1.938224476350985, + "learning_rate": 4.5640733281393275e-06, + "loss": 0.3811, + "step": 24621 + }, + { + "epoch": 0.5419766674003962, + "grad_norm": 1.266282331808433, + "learning_rate": 4.563718230215225e-06, + "loss": 0.2542, + "step": 24622 + }, + { + "epoch": 0.5419986792868149, + "grad_norm": 1.9861629124198028, + "learning_rate": 4.5633631345085015e-06, + "loss": 0.3609, + "step": 24623 + }, + { + "epoch": 0.5420206911732336, + "grad_norm": 1.7922554500885628, + "learning_rate": 4.563008041020959e-06, + "loss": 0.2962, + "step": 24624 + }, + { + "epoch": 0.5420427030596522, + "grad_norm": 1.1480747294389484, + "learning_rate": 4.5626529497544055e-06, + "loss": 0.2606, + "step": 24625 + }, + { + "epoch": 0.5420647149460709, + "grad_norm": 1.4852949277820782, + "learning_rate": 4.562297860710644e-06, + "loss": 0.2726, + "step": 24626 + }, + { + "epoch": 0.5420867268324896, + "grad_norm": 0.6445633851453327, + "learning_rate": 4.561942773891483e-06, + "loss": 0.4611, + "step": 24627 + }, + { + "epoch": 0.5421087387189082, + "grad_norm": 0.637314604541937, + "learning_rate": 4.5615876892987225e-06, + "loss": 0.5052, + "step": 24628 + }, + { + "epoch": 0.5421307506053269, + "grad_norm": 1.284378230595582, + "learning_rate": 4.561232606934168e-06, + "loss": 0.3575, + "step": 24629 + }, + { + "epoch": 0.5421527624917455, + "grad_norm": 1.418025833487079, + "learning_rate": 4.560877526799625e-06, + "loss": 0.2933, + "step": 24630 + }, + { + "epoch": 0.5421747743781642, + "grad_norm": 1.2379579913477008, + "learning_rate": 4.560522448896901e-06, + "loss": 0.2839, + "step": 24631 + }, + { + "epoch": 0.5421967862645829, + "grad_norm": 1.1050216025103907, + "learning_rate": 4.560167373227794e-06, + "loss": 0.2856, + "step": 24632 + }, + { + "epoch": 0.5422187981510015, + "grad_norm": 1.179675639609551, + "learning_rate": 4.559812299794115e-06, + "loss": 0.2974, + "step": 24633 + }, + { + "epoch": 0.5422408100374202, + "grad_norm": 1.3239164489142852, + "learning_rate": 4.559457228597665e-06, + "loss": 0.3344, + "step": 24634 + }, + { + "epoch": 0.5422628219238389, + "grad_norm": 1.6308120412402582, + "learning_rate": 4.559102159640251e-06, + "loss": 0.3363, + "step": 24635 + }, + { + "epoch": 0.5422848338102575, + "grad_norm": 1.5181705557967444, + "learning_rate": 4.558747092923678e-06, + "loss": 0.3347, + "step": 24636 + }, + { + "epoch": 0.5423068456966762, + "grad_norm": 1.785671199105883, + "learning_rate": 4.558392028449745e-06, + "loss": 0.2439, + "step": 24637 + }, + { + "epoch": 0.5423288575830949, + "grad_norm": 1.5765885786240754, + "learning_rate": 4.558036966220264e-06, + "loss": 0.3649, + "step": 24638 + }, + { + "epoch": 0.5423508694695135, + "grad_norm": 4.172048676888377, + "learning_rate": 4.5576819062370335e-06, + "loss": 0.3899, + "step": 24639 + }, + { + "epoch": 0.5423728813559322, + "grad_norm": 1.3750949757290645, + "learning_rate": 4.557326848501863e-06, + "loss": 0.3673, + "step": 24640 + }, + { + "epoch": 0.5423948932423509, + "grad_norm": 1.7452709240084778, + "learning_rate": 4.5569717930165555e-06, + "loss": 0.3847, + "step": 24641 + }, + { + "epoch": 0.5424169051287695, + "grad_norm": 1.7219232774952662, + "learning_rate": 4.556616739782913e-06, + "loss": 0.3353, + "step": 24642 + }, + { + "epoch": 0.5424389170151882, + "grad_norm": 1.7109289050806555, + "learning_rate": 4.556261688802743e-06, + "loss": 0.3015, + "step": 24643 + }, + { + "epoch": 0.5424609289016069, + "grad_norm": 1.3261921950385305, + "learning_rate": 4.555906640077848e-06, + "loss": 0.3474, + "step": 24644 + }, + { + "epoch": 0.5424829407880255, + "grad_norm": 1.5572142480698183, + "learning_rate": 4.555551593610035e-06, + "loss": 0.2561, + "step": 24645 + }, + { + "epoch": 0.5425049526744442, + "grad_norm": 1.462097446981825, + "learning_rate": 4.555196549401107e-06, + "loss": 0.3305, + "step": 24646 + }, + { + "epoch": 0.5425269645608629, + "grad_norm": 1.2985436951366909, + "learning_rate": 4.554841507452866e-06, + "loss": 0.2992, + "step": 24647 + }, + { + "epoch": 0.5425489764472815, + "grad_norm": 1.624788064384212, + "learning_rate": 4.554486467767122e-06, + "loss": 0.3501, + "step": 24648 + }, + { + "epoch": 0.5425709883337002, + "grad_norm": 1.3995323613443384, + "learning_rate": 4.554131430345676e-06, + "loss": 0.3452, + "step": 24649 + }, + { + "epoch": 0.5425930002201189, + "grad_norm": 0.634434395514612, + "learning_rate": 4.553776395190331e-06, + "loss": 0.4529, + "step": 24650 + }, + { + "epoch": 0.5426150121065375, + "grad_norm": 0.6702328494993084, + "learning_rate": 4.553421362302895e-06, + "loss": 0.5111, + "step": 24651 + }, + { + "epoch": 0.5426370239929562, + "grad_norm": 3.5932789399583784, + "learning_rate": 4.55306633168517e-06, + "loss": 0.2498, + "step": 24652 + }, + { + "epoch": 0.5426590358793748, + "grad_norm": 2.8726625794991394, + "learning_rate": 4.552711303338961e-06, + "loss": 0.3705, + "step": 24653 + }, + { + "epoch": 0.5426810477657935, + "grad_norm": 1.3693926688783133, + "learning_rate": 4.552356277266075e-06, + "loss": 0.319, + "step": 24654 + }, + { + "epoch": 0.5427030596522122, + "grad_norm": 1.237725620939127, + "learning_rate": 4.552001253468312e-06, + "loss": 0.2866, + "step": 24655 + }, + { + "epoch": 0.5427250715386308, + "grad_norm": 1.3501290259972838, + "learning_rate": 4.551646231947479e-06, + "loss": 0.3101, + "step": 24656 + }, + { + "epoch": 0.5427470834250495, + "grad_norm": 0.6059126057255838, + "learning_rate": 4.551291212705378e-06, + "loss": 0.4783, + "step": 24657 + }, + { + "epoch": 0.5427690953114682, + "grad_norm": 1.4561747466774217, + "learning_rate": 4.550936195743818e-06, + "loss": 0.3575, + "step": 24658 + }, + { + "epoch": 0.5427911071978868, + "grad_norm": 1.4028762592293327, + "learning_rate": 4.5505811810646e-06, + "loss": 0.2536, + "step": 24659 + }, + { + "epoch": 0.5428131190843055, + "grad_norm": 0.626786613616595, + "learning_rate": 4.550226168669528e-06, + "loss": 0.4911, + "step": 24660 + }, + { + "epoch": 0.5428351309707242, + "grad_norm": 1.1700050507160493, + "learning_rate": 4.549871158560407e-06, + "loss": 0.309, + "step": 24661 + }, + { + "epoch": 0.5428571428571428, + "grad_norm": 0.5923949121671753, + "learning_rate": 4.549516150739041e-06, + "loss": 0.4658, + "step": 24662 + }, + { + "epoch": 0.5428791547435615, + "grad_norm": 1.7606136895586728, + "learning_rate": 4.549161145207237e-06, + "loss": 0.3482, + "step": 24663 + }, + { + "epoch": 0.5429011666299802, + "grad_norm": 1.2869222665904276, + "learning_rate": 4.548806141966796e-06, + "loss": 0.2799, + "step": 24664 + }, + { + "epoch": 0.5429231785163988, + "grad_norm": 0.6208706344189794, + "learning_rate": 4.5484511410195235e-06, + "loss": 0.5001, + "step": 24665 + }, + { + "epoch": 0.5429451904028175, + "grad_norm": 1.358080653186838, + "learning_rate": 4.548096142367223e-06, + "loss": 0.2945, + "step": 24666 + }, + { + "epoch": 0.5429672022892362, + "grad_norm": 1.9727544774730175, + "learning_rate": 4.547741146011701e-06, + "loss": 0.4121, + "step": 24667 + }, + { + "epoch": 0.5429892141756548, + "grad_norm": 1.4063416917497884, + "learning_rate": 4.547386151954759e-06, + "loss": 0.2658, + "step": 24668 + }, + { + "epoch": 0.5430112260620735, + "grad_norm": 0.6235222426510505, + "learning_rate": 4.547031160198203e-06, + "loss": 0.4828, + "step": 24669 + }, + { + "epoch": 0.5430332379484922, + "grad_norm": 2.294569056404869, + "learning_rate": 4.546676170743835e-06, + "loss": 0.318, + "step": 24670 + }, + { + "epoch": 0.5430552498349108, + "grad_norm": 1.4590835971819378, + "learning_rate": 4.546321183593463e-06, + "loss": 0.2743, + "step": 24671 + }, + { + "epoch": 0.5430772617213295, + "grad_norm": 1.4841645418340648, + "learning_rate": 4.5459661987488895e-06, + "loss": 0.3141, + "step": 24672 + }, + { + "epoch": 0.5430992736077482, + "grad_norm": 1.4248776429103047, + "learning_rate": 4.545611216211916e-06, + "loss": 0.3693, + "step": 24673 + }, + { + "epoch": 0.5431212854941668, + "grad_norm": 1.696980377993285, + "learning_rate": 4.545256235984351e-06, + "loss": 0.2524, + "step": 24674 + }, + { + "epoch": 0.5431432973805855, + "grad_norm": 1.2583723614097877, + "learning_rate": 4.544901258067994e-06, + "loss": 0.3324, + "step": 24675 + }, + { + "epoch": 0.5431653092670042, + "grad_norm": 1.361767230913855, + "learning_rate": 4.544546282464654e-06, + "loss": 0.2366, + "step": 24676 + }, + { + "epoch": 0.5431873211534228, + "grad_norm": 0.6297229270404361, + "learning_rate": 4.544191309176134e-06, + "loss": 0.4793, + "step": 24677 + }, + { + "epoch": 0.5432093330398415, + "grad_norm": 1.3005432689726453, + "learning_rate": 4.543836338204235e-06, + "loss": 0.2635, + "step": 24678 + }, + { + "epoch": 0.5432313449262601, + "grad_norm": 3.1978149124665474, + "learning_rate": 4.543481369550763e-06, + "loss": 0.3499, + "step": 24679 + }, + { + "epoch": 0.5432533568126788, + "grad_norm": 5.779994667949255, + "learning_rate": 4.543126403217522e-06, + "loss": 0.2483, + "step": 24680 + }, + { + "epoch": 0.5432753686990975, + "grad_norm": 1.5854804091383012, + "learning_rate": 4.54277143920632e-06, + "loss": 0.3044, + "step": 24681 + }, + { + "epoch": 0.5432973805855161, + "grad_norm": 1.8783154259866397, + "learning_rate": 4.5424164775189545e-06, + "loss": 0.302, + "step": 24682 + }, + { + "epoch": 0.5433193924719348, + "grad_norm": 1.7417832310596713, + "learning_rate": 4.542061518157232e-06, + "loss": 0.403, + "step": 24683 + }, + { + "epoch": 0.5433414043583535, + "grad_norm": 1.5159730488173657, + "learning_rate": 4.5417065611229575e-06, + "loss": 0.2542, + "step": 24684 + }, + { + "epoch": 0.5433634162447721, + "grad_norm": 0.5931594921838285, + "learning_rate": 4.541351606417935e-06, + "loss": 0.4561, + "step": 24685 + }, + { + "epoch": 0.5433854281311908, + "grad_norm": 1.62045371622196, + "learning_rate": 4.54099665404397e-06, + "loss": 0.352, + "step": 24686 + }, + { + "epoch": 0.5434074400176095, + "grad_norm": 1.3662369621452906, + "learning_rate": 4.540641704002863e-06, + "loss": 0.3043, + "step": 24687 + }, + { + "epoch": 0.5434294519040281, + "grad_norm": 1.6450797534144177, + "learning_rate": 4.5402867562964185e-06, + "loss": 0.2911, + "step": 24688 + }, + { + "epoch": 0.5434514637904468, + "grad_norm": 2.4197247273881244, + "learning_rate": 4.5399318109264435e-06, + "loss": 0.2741, + "step": 24689 + }, + { + "epoch": 0.5434734756768655, + "grad_norm": 0.6259593535540078, + "learning_rate": 4.539576867894741e-06, + "loss": 0.5019, + "step": 24690 + }, + { + "epoch": 0.5434954875632841, + "grad_norm": 1.2447037237663843, + "learning_rate": 4.539221927203113e-06, + "loss": 0.3024, + "step": 24691 + }, + { + "epoch": 0.5435174994497028, + "grad_norm": 1.6484566894291246, + "learning_rate": 4.5388669888533645e-06, + "loss": 0.3612, + "step": 24692 + }, + { + "epoch": 0.5435395113361216, + "grad_norm": 1.276138578561189, + "learning_rate": 4.538512052847299e-06, + "loss": 0.2835, + "step": 24693 + }, + { + "epoch": 0.5435615232225401, + "grad_norm": 1.2873080180436525, + "learning_rate": 4.538157119186723e-06, + "loss": 0.2783, + "step": 24694 + }, + { + "epoch": 0.5435835351089588, + "grad_norm": 1.4566959968250903, + "learning_rate": 4.537802187873437e-06, + "loss": 0.2557, + "step": 24695 + }, + { + "epoch": 0.5436055469953776, + "grad_norm": 1.266638081399015, + "learning_rate": 4.537447258909247e-06, + "loss": 0.2575, + "step": 24696 + }, + { + "epoch": 0.5436275588817961, + "grad_norm": 1.3315069663373034, + "learning_rate": 4.537092332295956e-06, + "loss": 0.2353, + "step": 24697 + }, + { + "epoch": 0.5436495707682149, + "grad_norm": 1.335225560521879, + "learning_rate": 4.536737408035367e-06, + "loss": 0.1913, + "step": 24698 + }, + { + "epoch": 0.5436715826546336, + "grad_norm": 1.681231531341532, + "learning_rate": 4.536382486129288e-06, + "loss": 0.2829, + "step": 24699 + }, + { + "epoch": 0.5436935945410521, + "grad_norm": 0.615433536768541, + "learning_rate": 4.536027566579519e-06, + "loss": 0.5067, + "step": 24700 + }, + { + "epoch": 0.5437156064274709, + "grad_norm": 1.4113958089113372, + "learning_rate": 4.535672649387863e-06, + "loss": 0.3163, + "step": 24701 + }, + { + "epoch": 0.5437376183138894, + "grad_norm": 1.4493822377417178, + "learning_rate": 4.535317734556127e-06, + "loss": 0.3354, + "step": 24702 + }, + { + "epoch": 0.5437596302003082, + "grad_norm": 1.4146968035045036, + "learning_rate": 4.534962822086113e-06, + "loss": 0.2868, + "step": 24703 + }, + { + "epoch": 0.5437816420867269, + "grad_norm": 1.5279651483147423, + "learning_rate": 4.534607911979628e-06, + "loss": 0.3404, + "step": 24704 + }, + { + "epoch": 0.5438036539731455, + "grad_norm": 0.6169213633079463, + "learning_rate": 4.53425300423847e-06, + "loss": 0.4603, + "step": 24705 + }, + { + "epoch": 0.5438256658595642, + "grad_norm": 1.9673634772232262, + "learning_rate": 4.5338980988644475e-06, + "loss": 0.2804, + "step": 24706 + }, + { + "epoch": 0.5438476777459829, + "grad_norm": 0.6092080317659003, + "learning_rate": 4.5335431958593615e-06, + "loss": 0.4648, + "step": 24707 + }, + { + "epoch": 0.5438696896324015, + "grad_norm": 0.6068494137586814, + "learning_rate": 4.53318829522502e-06, + "loss": 0.4576, + "step": 24708 + }, + { + "epoch": 0.5438917015188202, + "grad_norm": 1.4270935142885641, + "learning_rate": 4.53283339696322e-06, + "loss": 0.3686, + "step": 24709 + }, + { + "epoch": 0.5439137134052389, + "grad_norm": 3.095736585216453, + "learning_rate": 4.532478501075771e-06, + "loss": 0.3026, + "step": 24710 + }, + { + "epoch": 0.5439357252916575, + "grad_norm": 1.410004261230203, + "learning_rate": 4.532123607564474e-06, + "loss": 0.3805, + "step": 24711 + }, + { + "epoch": 0.5439577371780762, + "grad_norm": 1.9765448606469103, + "learning_rate": 4.531768716431135e-06, + "loss": 0.323, + "step": 24712 + }, + { + "epoch": 0.5439797490644949, + "grad_norm": 1.6854318242941981, + "learning_rate": 4.531413827677556e-06, + "loss": 0.3664, + "step": 24713 + }, + { + "epoch": 0.5440017609509135, + "grad_norm": 1.5686685596802077, + "learning_rate": 4.531058941305538e-06, + "loss": 0.3995, + "step": 24714 + }, + { + "epoch": 0.5440237728373322, + "grad_norm": 1.3931184497250422, + "learning_rate": 4.5307040573168906e-06, + "loss": 0.3423, + "step": 24715 + }, + { + "epoch": 0.5440457847237509, + "grad_norm": 1.489911245841897, + "learning_rate": 4.5303491757134125e-06, + "loss": 0.3351, + "step": 24716 + }, + { + "epoch": 0.5440677966101695, + "grad_norm": 1.3478093615717144, + "learning_rate": 4.529994296496912e-06, + "loss": 0.3059, + "step": 24717 + }, + { + "epoch": 0.5440898084965882, + "grad_norm": 1.3555997985349066, + "learning_rate": 4.529639419669189e-06, + "loss": 0.2947, + "step": 24718 + }, + { + "epoch": 0.5441118203830069, + "grad_norm": 1.3680256021460364, + "learning_rate": 4.529284545232046e-06, + "loss": 0.2845, + "step": 24719 + }, + { + "epoch": 0.5441338322694255, + "grad_norm": 1.4955262236444207, + "learning_rate": 4.528929673187291e-06, + "loss": 0.2335, + "step": 24720 + }, + { + "epoch": 0.5441558441558442, + "grad_norm": 1.257840665223826, + "learning_rate": 4.5285748035367244e-06, + "loss": 0.2746, + "step": 24721 + }, + { + "epoch": 0.5441778560422629, + "grad_norm": 1.357066312187208, + "learning_rate": 4.528219936282153e-06, + "loss": 0.2774, + "step": 24722 + }, + { + "epoch": 0.5441998679286815, + "grad_norm": 0.6375475362215127, + "learning_rate": 4.527865071425378e-06, + "loss": 0.4678, + "step": 24723 + }, + { + "epoch": 0.5442218798151002, + "grad_norm": 1.4672744031278444, + "learning_rate": 4.527510208968201e-06, + "loss": 0.2958, + "step": 24724 + }, + { + "epoch": 0.5442438917015188, + "grad_norm": 1.8775132840623838, + "learning_rate": 4.52715534891243e-06, + "loss": 0.3025, + "step": 24725 + }, + { + "epoch": 0.5442659035879375, + "grad_norm": 4.264413281556776, + "learning_rate": 4.526800491259864e-06, + "loss": 0.3452, + "step": 24726 + }, + { + "epoch": 0.5442879154743562, + "grad_norm": 1.725790233896183, + "learning_rate": 4.526445636012312e-06, + "loss": 0.4086, + "step": 24727 + }, + { + "epoch": 0.5443099273607748, + "grad_norm": 0.5985546956074933, + "learning_rate": 4.526090783171573e-06, + "loss": 0.4778, + "step": 24728 + }, + { + "epoch": 0.5443319392471935, + "grad_norm": 1.233850787068581, + "learning_rate": 4.525735932739451e-06, + "loss": 0.2988, + "step": 24729 + }, + { + "epoch": 0.5443539511336122, + "grad_norm": 1.7195032834210575, + "learning_rate": 4.525381084717752e-06, + "loss": 0.2758, + "step": 24730 + }, + { + "epoch": 0.5443759630200308, + "grad_norm": 1.5625635080906666, + "learning_rate": 4.525026239108279e-06, + "loss": 0.2984, + "step": 24731 + }, + { + "epoch": 0.5443979749064495, + "grad_norm": 1.8642663070553886, + "learning_rate": 4.524671395912832e-06, + "loss": 0.3854, + "step": 24732 + }, + { + "epoch": 0.5444199867928682, + "grad_norm": 1.7044511311835728, + "learning_rate": 4.524316555133218e-06, + "loss": 0.3033, + "step": 24733 + }, + { + "epoch": 0.5444419986792868, + "grad_norm": 8.214306330832809, + "learning_rate": 4.523961716771239e-06, + "loss": 0.3401, + "step": 24734 + }, + { + "epoch": 0.5444640105657055, + "grad_norm": 2.412276412875096, + "learning_rate": 4.5236068808287e-06, + "loss": 0.2556, + "step": 24735 + }, + { + "epoch": 0.5444860224521242, + "grad_norm": 0.6508978790726712, + "learning_rate": 4.523252047307403e-06, + "loss": 0.5043, + "step": 24736 + }, + { + "epoch": 0.5445080343385428, + "grad_norm": 1.9365600025416256, + "learning_rate": 4.522897216209151e-06, + "loss": 0.2561, + "step": 24737 + }, + { + "epoch": 0.5445300462249615, + "grad_norm": 1.9275681664434663, + "learning_rate": 4.522542387535748e-06, + "loss": 0.3202, + "step": 24738 + }, + { + "epoch": 0.5445520581113802, + "grad_norm": 1.1323552688499374, + "learning_rate": 4.522187561288997e-06, + "loss": 0.2019, + "step": 24739 + }, + { + "epoch": 0.5445740699977988, + "grad_norm": 0.6198952713741137, + "learning_rate": 4.521832737470705e-06, + "loss": 0.4435, + "step": 24740 + }, + { + "epoch": 0.5445960818842175, + "grad_norm": 1.621242102324368, + "learning_rate": 4.52147791608267e-06, + "loss": 0.3345, + "step": 24741 + }, + { + "epoch": 0.5446180937706362, + "grad_norm": 0.6049832674170147, + "learning_rate": 4.5211230971266975e-06, + "loss": 0.4635, + "step": 24742 + }, + { + "epoch": 0.5446401056570548, + "grad_norm": 1.2536815014034088, + "learning_rate": 4.520768280604591e-06, + "loss": 0.3827, + "step": 24743 + }, + { + "epoch": 0.5446621175434735, + "grad_norm": 0.5988519053881917, + "learning_rate": 4.520413466518154e-06, + "loss": 0.4851, + "step": 24744 + }, + { + "epoch": 0.5446841294298922, + "grad_norm": 1.4283003899068227, + "learning_rate": 4.520058654869191e-06, + "loss": 0.2688, + "step": 24745 + }, + { + "epoch": 0.5447061413163108, + "grad_norm": 1.379466718901602, + "learning_rate": 4.519703845659503e-06, + "loss": 0.354, + "step": 24746 + }, + { + "epoch": 0.5447281532027295, + "grad_norm": 1.3874800561095615, + "learning_rate": 4.5193490388908935e-06, + "loss": 0.3109, + "step": 24747 + }, + { + "epoch": 0.5447501650891481, + "grad_norm": 0.6166342657509439, + "learning_rate": 4.518994234565169e-06, + "loss": 0.4664, + "step": 24748 + }, + { + "epoch": 0.5447721769755668, + "grad_norm": 1.6538759065739872, + "learning_rate": 4.518639432684128e-06, + "loss": 0.3687, + "step": 24749 + }, + { + "epoch": 0.5447941888619855, + "grad_norm": 0.6354191800704548, + "learning_rate": 4.518284633249576e-06, + "loss": 0.4769, + "step": 24750 + }, + { + "epoch": 0.5448162007484041, + "grad_norm": 0.6001346463466046, + "learning_rate": 4.517929836263317e-06, + "loss": 0.4711, + "step": 24751 + }, + { + "epoch": 0.5448382126348228, + "grad_norm": 2.443746705380671, + "learning_rate": 4.5175750417271525e-06, + "loss": 0.3257, + "step": 24752 + }, + { + "epoch": 0.5448602245212415, + "grad_norm": 2.0173229506584183, + "learning_rate": 4.517220249642889e-06, + "loss": 0.3386, + "step": 24753 + }, + { + "epoch": 0.5448822364076601, + "grad_norm": 1.3703698419257582, + "learning_rate": 4.516865460012327e-06, + "loss": 0.3038, + "step": 24754 + }, + { + "epoch": 0.5449042482940788, + "grad_norm": 1.4352751207083516, + "learning_rate": 4.516510672837269e-06, + "loss": 0.2967, + "step": 24755 + }, + { + "epoch": 0.5449262601804975, + "grad_norm": 0.6095287324573323, + "learning_rate": 4.51615588811952e-06, + "loss": 0.463, + "step": 24756 + }, + { + "epoch": 0.5449482720669161, + "grad_norm": 0.6184276187887552, + "learning_rate": 4.515801105860882e-06, + "loss": 0.4493, + "step": 24757 + }, + { + "epoch": 0.5449702839533348, + "grad_norm": 1.599512022410702, + "learning_rate": 4.51544632606316e-06, + "loss": 0.3236, + "step": 24758 + }, + { + "epoch": 0.5449922958397535, + "grad_norm": 0.6196972128509424, + "learning_rate": 4.515091548728156e-06, + "loss": 0.4641, + "step": 24759 + }, + { + "epoch": 0.5450143077261721, + "grad_norm": 1.5197176683822287, + "learning_rate": 4.514736773857671e-06, + "loss": 0.2057, + "step": 24760 + }, + { + "epoch": 0.5450363196125908, + "grad_norm": 1.7925110481112925, + "learning_rate": 4.514382001453511e-06, + "loss": 0.3957, + "step": 24761 + }, + { + "epoch": 0.5450583314990095, + "grad_norm": 1.0500947374834264, + "learning_rate": 4.514027231517479e-06, + "loss": 0.2485, + "step": 24762 + }, + { + "epoch": 0.5450803433854281, + "grad_norm": 1.625227721556649, + "learning_rate": 4.513672464051378e-06, + "loss": 0.2677, + "step": 24763 + }, + { + "epoch": 0.5451023552718468, + "grad_norm": 1.3039116519623748, + "learning_rate": 4.513317699057009e-06, + "loss": 0.3018, + "step": 24764 + }, + { + "epoch": 0.5451243671582655, + "grad_norm": 1.1133415321936229, + "learning_rate": 4.512962936536177e-06, + "loss": 0.2898, + "step": 24765 + }, + { + "epoch": 0.5451463790446841, + "grad_norm": 1.637550862980303, + "learning_rate": 4.512608176490685e-06, + "loss": 0.356, + "step": 24766 + }, + { + "epoch": 0.5451683909311028, + "grad_norm": 2.026161181124228, + "learning_rate": 4.512253418922336e-06, + "loss": 0.2898, + "step": 24767 + }, + { + "epoch": 0.5451904028175215, + "grad_norm": 2.739831584303911, + "learning_rate": 4.5118986638329306e-06, + "loss": 0.2416, + "step": 24768 + }, + { + "epoch": 0.5452124147039401, + "grad_norm": 3.48813992697358, + "learning_rate": 4.511543911224276e-06, + "loss": 0.2452, + "step": 24769 + }, + { + "epoch": 0.5452344265903588, + "grad_norm": 1.435667431742801, + "learning_rate": 4.511189161098171e-06, + "loss": 0.2558, + "step": 24770 + }, + { + "epoch": 0.5452564384767774, + "grad_norm": 1.4991428796161816, + "learning_rate": 4.510834413456424e-06, + "loss": 0.29, + "step": 24771 + }, + { + "epoch": 0.5452784503631961, + "grad_norm": 0.6280516066619665, + "learning_rate": 4.510479668300832e-06, + "loss": 0.4655, + "step": 24772 + }, + { + "epoch": 0.5453004622496148, + "grad_norm": 1.8022863566886975, + "learning_rate": 4.510124925633201e-06, + "loss": 0.2804, + "step": 24773 + }, + { + "epoch": 0.5453224741360334, + "grad_norm": 1.4446940035037545, + "learning_rate": 4.509770185455334e-06, + "loss": 0.3014, + "step": 24774 + }, + { + "epoch": 0.5453444860224521, + "grad_norm": 1.4670919616481788, + "learning_rate": 4.5094154477690324e-06, + "loss": 0.329, + "step": 24775 + }, + { + "epoch": 0.5453664979088708, + "grad_norm": 1.6634650312021795, + "learning_rate": 4.509060712576103e-06, + "loss": 0.3378, + "step": 24776 + }, + { + "epoch": 0.5453885097952894, + "grad_norm": 1.4744743706273797, + "learning_rate": 4.508705979878345e-06, + "loss": 0.354, + "step": 24777 + }, + { + "epoch": 0.5454105216817081, + "grad_norm": 1.335185593062059, + "learning_rate": 4.50835124967756e-06, + "loss": 0.1919, + "step": 24778 + }, + { + "epoch": 0.5454325335681268, + "grad_norm": 1.1535076985050625, + "learning_rate": 4.507996521975555e-06, + "loss": 0.2801, + "step": 24779 + }, + { + "epoch": 0.5454545454545454, + "grad_norm": 1.5280062832994257, + "learning_rate": 4.507641796774131e-06, + "loss": 0.2933, + "step": 24780 + }, + { + "epoch": 0.5454765573409641, + "grad_norm": 1.8004054367017404, + "learning_rate": 4.5072870740750915e-06, + "loss": 0.3111, + "step": 24781 + }, + { + "epoch": 0.5454985692273828, + "grad_norm": 1.7172670634083131, + "learning_rate": 4.506932353880239e-06, + "loss": 0.3043, + "step": 24782 + }, + { + "epoch": 0.5455205811138014, + "grad_norm": 1.3543362004135637, + "learning_rate": 4.506577636191375e-06, + "loss": 0.3887, + "step": 24783 + }, + { + "epoch": 0.5455425930002201, + "grad_norm": 1.2208773755293392, + "learning_rate": 4.506222921010306e-06, + "loss": 0.2468, + "step": 24784 + }, + { + "epoch": 0.5455646048866388, + "grad_norm": 1.3936302631857083, + "learning_rate": 4.505868208338828e-06, + "loss": 0.4026, + "step": 24785 + }, + { + "epoch": 0.5455866167730574, + "grad_norm": 1.4821912883107908, + "learning_rate": 4.505513498178752e-06, + "loss": 0.4241, + "step": 24786 + }, + { + "epoch": 0.5456086286594761, + "grad_norm": 3.001857949180946, + "learning_rate": 4.505158790531875e-06, + "loss": 0.3037, + "step": 24787 + }, + { + "epoch": 0.5456306405458948, + "grad_norm": 1.7676063372661137, + "learning_rate": 4.504804085400002e-06, + "loss": 0.3271, + "step": 24788 + }, + { + "epoch": 0.5456526524323134, + "grad_norm": 1.577895104336252, + "learning_rate": 4.504449382784937e-06, + "loss": 0.2763, + "step": 24789 + }, + { + "epoch": 0.5456746643187321, + "grad_norm": 2.1369470675064157, + "learning_rate": 4.50409468268848e-06, + "loss": 0.2659, + "step": 24790 + }, + { + "epoch": 0.5456966762051508, + "grad_norm": 1.2282780737853243, + "learning_rate": 4.503739985112435e-06, + "loss": 0.1957, + "step": 24791 + }, + { + "epoch": 0.5457186880915694, + "grad_norm": 1.1873886801763662, + "learning_rate": 4.503385290058604e-06, + "loss": 0.252, + "step": 24792 + }, + { + "epoch": 0.5457406999779881, + "grad_norm": 1.42309870302666, + "learning_rate": 4.503030597528791e-06, + "loss": 0.3068, + "step": 24793 + }, + { + "epoch": 0.5457627118644067, + "grad_norm": 1.637581232330845, + "learning_rate": 4.5026759075248e-06, + "loss": 0.3727, + "step": 24794 + }, + { + "epoch": 0.5457847237508254, + "grad_norm": 1.314406574830222, + "learning_rate": 4.50232122004843e-06, + "loss": 0.2452, + "step": 24795 + }, + { + "epoch": 0.5458067356372441, + "grad_norm": 1.2401752742471372, + "learning_rate": 4.501966535101485e-06, + "loss": 0.2385, + "step": 24796 + }, + { + "epoch": 0.5458287475236627, + "grad_norm": 1.5671424160613163, + "learning_rate": 4.501611852685769e-06, + "loss": 0.2252, + "step": 24797 + }, + { + "epoch": 0.5458507594100814, + "grad_norm": 1.3117449301595139, + "learning_rate": 4.5012571728030834e-06, + "loss": 0.2722, + "step": 24798 + }, + { + "epoch": 0.5458727712965001, + "grad_norm": 1.3254455265389422, + "learning_rate": 4.500902495455233e-06, + "loss": 0.2448, + "step": 24799 + }, + { + "epoch": 0.5458947831829187, + "grad_norm": 1.747360534713582, + "learning_rate": 4.5005478206440175e-06, + "loss": 0.3374, + "step": 24800 + }, + { + "epoch": 0.5459167950693374, + "grad_norm": 1.253727435546676, + "learning_rate": 4.50019314837124e-06, + "loss": 0.2693, + "step": 24801 + }, + { + "epoch": 0.5459388069557561, + "grad_norm": 1.4468108418168237, + "learning_rate": 4.499838478638706e-06, + "loss": 0.2991, + "step": 24802 + }, + { + "epoch": 0.5459608188421747, + "grad_norm": 1.4060699644693102, + "learning_rate": 4.499483811448212e-06, + "loss": 0.255, + "step": 24803 + }, + { + "epoch": 0.5459828307285934, + "grad_norm": 1.3613293777560096, + "learning_rate": 4.499129146801569e-06, + "loss": 0.4396, + "step": 24804 + }, + { + "epoch": 0.5460048426150121, + "grad_norm": 0.657358862333676, + "learning_rate": 4.4987744847005735e-06, + "loss": 0.4805, + "step": 24805 + }, + { + "epoch": 0.5460268545014307, + "grad_norm": 1.5283483814015006, + "learning_rate": 4.498419825147028e-06, + "loss": 0.3545, + "step": 24806 + }, + { + "epoch": 0.5460488663878494, + "grad_norm": 1.5408206514098886, + "learning_rate": 4.498065168142739e-06, + "loss": 0.3428, + "step": 24807 + }, + { + "epoch": 0.5460708782742681, + "grad_norm": 1.2599486114241232, + "learning_rate": 4.497710513689506e-06, + "loss": 0.3395, + "step": 24808 + }, + { + "epoch": 0.5460928901606867, + "grad_norm": 1.5051568669516286, + "learning_rate": 4.497355861789131e-06, + "loss": 0.3338, + "step": 24809 + }, + { + "epoch": 0.5461149020471054, + "grad_norm": 1.417343477404083, + "learning_rate": 4.497001212443418e-06, + "loss": 0.2524, + "step": 24810 + }, + { + "epoch": 0.5461369139335241, + "grad_norm": 1.536489338980521, + "learning_rate": 4.496646565654168e-06, + "loss": 0.2644, + "step": 24811 + }, + { + "epoch": 0.5461589258199427, + "grad_norm": 1.3353281996120752, + "learning_rate": 4.496291921423188e-06, + "loss": 0.2827, + "step": 24812 + }, + { + "epoch": 0.5461809377063614, + "grad_norm": 1.5005942033847273, + "learning_rate": 4.495937279752276e-06, + "loss": 0.3473, + "step": 24813 + }, + { + "epoch": 0.5462029495927802, + "grad_norm": 0.6282802629416462, + "learning_rate": 4.495582640643234e-06, + "loss": 0.4655, + "step": 24814 + }, + { + "epoch": 0.5462249614791987, + "grad_norm": 1.5160924754710088, + "learning_rate": 4.495228004097867e-06, + "loss": 0.2916, + "step": 24815 + }, + { + "epoch": 0.5462469733656174, + "grad_norm": 1.8389703798799797, + "learning_rate": 4.494873370117976e-06, + "loss": 0.4427, + "step": 24816 + }, + { + "epoch": 0.5462689852520362, + "grad_norm": 1.419483518302988, + "learning_rate": 4.494518738705365e-06, + "loss": 0.3303, + "step": 24817 + }, + { + "epoch": 0.5462909971384547, + "grad_norm": 2.5059467975520073, + "learning_rate": 4.494164109861835e-06, + "loss": 0.3624, + "step": 24818 + }, + { + "epoch": 0.5463130090248735, + "grad_norm": 1.6051835053754278, + "learning_rate": 4.493809483589187e-06, + "loss": 0.2836, + "step": 24819 + }, + { + "epoch": 0.546335020911292, + "grad_norm": 1.7939748162898044, + "learning_rate": 4.493454859889227e-06, + "loss": 0.3076, + "step": 24820 + }, + { + "epoch": 0.5463570327977108, + "grad_norm": 1.4303297869882112, + "learning_rate": 4.493100238763752e-06, + "loss": 0.2812, + "step": 24821 + }, + { + "epoch": 0.5463790446841295, + "grad_norm": 1.2918318150387464, + "learning_rate": 4.492745620214571e-06, + "loss": 0.3843, + "step": 24822 + }, + { + "epoch": 0.546401056570548, + "grad_norm": 1.8952302173304387, + "learning_rate": 4.4923910042434825e-06, + "loss": 0.3471, + "step": 24823 + }, + { + "epoch": 0.5464230684569668, + "grad_norm": 1.7003057150673533, + "learning_rate": 4.4920363908522875e-06, + "loss": 0.3109, + "step": 24824 + }, + { + "epoch": 0.5464450803433855, + "grad_norm": 1.7644402781364104, + "learning_rate": 4.491681780042794e-06, + "loss": 0.3279, + "step": 24825 + }, + { + "epoch": 0.546467092229804, + "grad_norm": 1.062488307586852, + "learning_rate": 4.4913271718167965e-06, + "loss": 0.2965, + "step": 24826 + }, + { + "epoch": 0.5464891041162228, + "grad_norm": 1.1691205759081327, + "learning_rate": 4.490972566176103e-06, + "loss": 0.2186, + "step": 24827 + }, + { + "epoch": 0.5465111160026415, + "grad_norm": 1.5231040003115501, + "learning_rate": 4.490617963122513e-06, + "loss": 0.2497, + "step": 24828 + }, + { + "epoch": 0.54653312788906, + "grad_norm": 1.3057351736273892, + "learning_rate": 4.490263362657829e-06, + "loss": 0.2397, + "step": 24829 + }, + { + "epoch": 0.5465551397754788, + "grad_norm": 1.3799780327809454, + "learning_rate": 4.489908764783857e-06, + "loss": 0.2973, + "step": 24830 + }, + { + "epoch": 0.5465771516618975, + "grad_norm": 1.2898975149375462, + "learning_rate": 4.489554169502394e-06, + "loss": 0.2795, + "step": 24831 + }, + { + "epoch": 0.5465991635483161, + "grad_norm": 0.6364729335884878, + "learning_rate": 4.489199576815244e-06, + "loss": 0.4697, + "step": 24832 + }, + { + "epoch": 0.5466211754347348, + "grad_norm": 1.4266294700515583, + "learning_rate": 4.48884498672421e-06, + "loss": 0.337, + "step": 24833 + }, + { + "epoch": 0.5466431873211535, + "grad_norm": 1.4907786848773321, + "learning_rate": 4.488490399231093e-06, + "loss": 0.2389, + "step": 24834 + }, + { + "epoch": 0.5466651992075721, + "grad_norm": 1.4501255794079229, + "learning_rate": 4.4881358143376984e-06, + "loss": 0.3537, + "step": 24835 + }, + { + "epoch": 0.5466872110939908, + "grad_norm": 0.6084285198962379, + "learning_rate": 4.487781232045825e-06, + "loss": 0.4774, + "step": 24836 + }, + { + "epoch": 0.5467092229804095, + "grad_norm": 1.6223075609528297, + "learning_rate": 4.487426652357274e-06, + "loss": 0.3369, + "step": 24837 + }, + { + "epoch": 0.5467312348668281, + "grad_norm": 4.492263977663555, + "learning_rate": 4.487072075273852e-06, + "loss": 0.3131, + "step": 24838 + }, + { + "epoch": 0.5467532467532468, + "grad_norm": 0.6138959224919898, + "learning_rate": 4.4867175007973554e-06, + "loss": 0.4581, + "step": 24839 + }, + { + "epoch": 0.5467752586396655, + "grad_norm": 0.6164123949419613, + "learning_rate": 4.4863629289295925e-06, + "loss": 0.4575, + "step": 24840 + }, + { + "epoch": 0.5467972705260841, + "grad_norm": 0.6308057581100253, + "learning_rate": 4.486008359672361e-06, + "loss": 0.4573, + "step": 24841 + }, + { + "epoch": 0.5468192824125028, + "grad_norm": 1.1952011693547522, + "learning_rate": 4.485653793027463e-06, + "loss": 0.2911, + "step": 24842 + }, + { + "epoch": 0.5468412942989214, + "grad_norm": 1.59375798617763, + "learning_rate": 4.485299228996705e-06, + "loss": 0.3801, + "step": 24843 + }, + { + "epoch": 0.5468633061853401, + "grad_norm": 0.6128826467732496, + "learning_rate": 4.484944667581882e-06, + "loss": 0.4753, + "step": 24844 + }, + { + "epoch": 0.5468853180717588, + "grad_norm": 1.3872479334484349, + "learning_rate": 4.484590108784802e-06, + "loss": 0.2962, + "step": 24845 + }, + { + "epoch": 0.5469073299581774, + "grad_norm": 0.5864622924913558, + "learning_rate": 4.4842355526072654e-06, + "loss": 0.4449, + "step": 24846 + }, + { + "epoch": 0.5469293418445961, + "grad_norm": 1.1381378019667001, + "learning_rate": 4.483880999051072e-06, + "loss": 0.1941, + "step": 24847 + }, + { + "epoch": 0.5469513537310148, + "grad_norm": 1.3123741909771605, + "learning_rate": 4.483526448118028e-06, + "loss": 0.2382, + "step": 24848 + }, + { + "epoch": 0.5469733656174334, + "grad_norm": 1.6052594940239429, + "learning_rate": 4.4831718998099316e-06, + "loss": 0.2511, + "step": 24849 + }, + { + "epoch": 0.5469953775038521, + "grad_norm": 4.202204530114004, + "learning_rate": 4.482817354128585e-06, + "loss": 0.376, + "step": 24850 + }, + { + "epoch": 0.5470173893902708, + "grad_norm": 1.695720616312214, + "learning_rate": 4.482462811075793e-06, + "loss": 0.288, + "step": 24851 + }, + { + "epoch": 0.5470394012766894, + "grad_norm": 0.6402645378965374, + "learning_rate": 4.482108270653354e-06, + "loss": 0.4852, + "step": 24852 + }, + { + "epoch": 0.5470614131631081, + "grad_norm": 1.9293403493102357, + "learning_rate": 4.481753732863074e-06, + "loss": 0.3894, + "step": 24853 + }, + { + "epoch": 0.5470834250495268, + "grad_norm": 1.4677438601900474, + "learning_rate": 4.481399197706752e-06, + "loss": 0.3236, + "step": 24854 + }, + { + "epoch": 0.5471054369359454, + "grad_norm": 1.4842887735052832, + "learning_rate": 4.481044665186189e-06, + "loss": 0.2867, + "step": 24855 + }, + { + "epoch": 0.5471274488223641, + "grad_norm": 2.164224429145753, + "learning_rate": 4.480690135303191e-06, + "loss": 0.3456, + "step": 24856 + }, + { + "epoch": 0.5471494607087828, + "grad_norm": 1.5349392472204677, + "learning_rate": 4.4803356080595546e-06, + "loss": 0.284, + "step": 24857 + }, + { + "epoch": 0.5471714725952014, + "grad_norm": 1.2104809182430158, + "learning_rate": 4.4799810834570865e-06, + "loss": 0.3242, + "step": 24858 + }, + { + "epoch": 0.5471934844816201, + "grad_norm": 1.2932676593022445, + "learning_rate": 4.4796265614975865e-06, + "loss": 0.2584, + "step": 24859 + }, + { + "epoch": 0.5472154963680388, + "grad_norm": 4.679639428853932, + "learning_rate": 4.4792720421828546e-06, + "loss": 0.3708, + "step": 24860 + }, + { + "epoch": 0.5472375082544574, + "grad_norm": 1.1641946427357954, + "learning_rate": 4.4789175255146966e-06, + "loss": 0.2537, + "step": 24861 + }, + { + "epoch": 0.5472595201408761, + "grad_norm": 0.6517792375270564, + "learning_rate": 4.47856301149491e-06, + "loss": 0.4619, + "step": 24862 + }, + { + "epoch": 0.5472815320272948, + "grad_norm": 1.404442151770974, + "learning_rate": 4.4782085001253e-06, + "loss": 0.248, + "step": 24863 + }, + { + "epoch": 0.5473035439137134, + "grad_norm": 1.9203406557620397, + "learning_rate": 4.477853991407667e-06, + "loss": 0.3332, + "step": 24864 + }, + { + "epoch": 0.5473255558001321, + "grad_norm": 0.6383372379043475, + "learning_rate": 4.477499485343813e-06, + "loss": 0.4554, + "step": 24865 + }, + { + "epoch": 0.5473475676865507, + "grad_norm": 1.572207610573071, + "learning_rate": 4.4771449819355406e-06, + "loss": 0.2811, + "step": 24866 + }, + { + "epoch": 0.5473695795729694, + "grad_norm": 1.4857239994745943, + "learning_rate": 4.4767904811846495e-06, + "loss": 0.2957, + "step": 24867 + }, + { + "epoch": 0.5473915914593881, + "grad_norm": 1.1529924733675565, + "learning_rate": 4.476435983092941e-06, + "loss": 0.2166, + "step": 24868 + }, + { + "epoch": 0.5474136033458067, + "grad_norm": 1.271007279446498, + "learning_rate": 4.47608148766222e-06, + "loss": 0.3555, + "step": 24869 + }, + { + "epoch": 0.5474356152322254, + "grad_norm": 2.642943868079972, + "learning_rate": 4.475726994894286e-06, + "loss": 0.2939, + "step": 24870 + }, + { + "epoch": 0.5474576271186441, + "grad_norm": 0.6168737738869287, + "learning_rate": 4.475372504790943e-06, + "loss": 0.4783, + "step": 24871 + }, + { + "epoch": 0.5474796390050627, + "grad_norm": 1.8203785548986309, + "learning_rate": 4.4750180173539894e-06, + "loss": 0.3324, + "step": 24872 + }, + { + "epoch": 0.5475016508914814, + "grad_norm": 1.308289577010371, + "learning_rate": 4.474663532585229e-06, + "loss": 0.2953, + "step": 24873 + }, + { + "epoch": 0.5475236627779001, + "grad_norm": 1.535916885119075, + "learning_rate": 4.474309050486463e-06, + "loss": 0.297, + "step": 24874 + }, + { + "epoch": 0.5475456746643187, + "grad_norm": 0.976252771799783, + "learning_rate": 4.47395457105949e-06, + "loss": 0.2356, + "step": 24875 + }, + { + "epoch": 0.5475676865507374, + "grad_norm": 0.6082894547306943, + "learning_rate": 4.473600094306119e-06, + "loss": 0.452, + "step": 24876 + }, + { + "epoch": 0.5475896984371561, + "grad_norm": 1.4757575803660128, + "learning_rate": 4.473245620228145e-06, + "loss": 0.3657, + "step": 24877 + }, + { + "epoch": 0.5476117103235747, + "grad_norm": 1.771115964247799, + "learning_rate": 4.47289114882737e-06, + "loss": 0.3771, + "step": 24878 + }, + { + "epoch": 0.5476337222099934, + "grad_norm": 1.67351778536623, + "learning_rate": 4.4725366801056005e-06, + "loss": 0.3618, + "step": 24879 + }, + { + "epoch": 0.5476557340964121, + "grad_norm": 0.6175132037703297, + "learning_rate": 4.472182214064632e-06, + "loss": 0.4539, + "step": 24880 + }, + { + "epoch": 0.5476777459828307, + "grad_norm": 2.0128501632454867, + "learning_rate": 4.47182775070627e-06, + "loss": 0.4035, + "step": 24881 + }, + { + "epoch": 0.5476997578692494, + "grad_norm": 1.1286468394246705, + "learning_rate": 4.471473290032316e-06, + "loss": 0.2541, + "step": 24882 + }, + { + "epoch": 0.5477217697556681, + "grad_norm": 0.610450923331588, + "learning_rate": 4.471118832044568e-06, + "loss": 0.4569, + "step": 24883 + }, + { + "epoch": 0.5477437816420867, + "grad_norm": 1.6312203630469881, + "learning_rate": 4.470764376744833e-06, + "loss": 0.3373, + "step": 24884 + }, + { + "epoch": 0.5477657935285054, + "grad_norm": 1.1457606578263857, + "learning_rate": 4.470409924134907e-06, + "loss": 0.2824, + "step": 24885 + }, + { + "epoch": 0.5477878054149241, + "grad_norm": 0.6197963256444011, + "learning_rate": 4.4700554742165955e-06, + "loss": 0.4694, + "step": 24886 + }, + { + "epoch": 0.5478098173013427, + "grad_norm": 0.6474106506873608, + "learning_rate": 4.4697010269916976e-06, + "loss": 0.4617, + "step": 24887 + }, + { + "epoch": 0.5478318291877614, + "grad_norm": 1.4326492133498143, + "learning_rate": 4.469346582462015e-06, + "loss": 0.2927, + "step": 24888 + }, + { + "epoch": 0.54785384107418, + "grad_norm": 1.5582243206540058, + "learning_rate": 4.468992140629352e-06, + "loss": 0.3341, + "step": 24889 + }, + { + "epoch": 0.5478758529605987, + "grad_norm": 1.4886990993964413, + "learning_rate": 4.468637701495507e-06, + "loss": 0.2935, + "step": 24890 + }, + { + "epoch": 0.5478978648470174, + "grad_norm": 2.118918089494663, + "learning_rate": 4.468283265062281e-06, + "loss": 0.3235, + "step": 24891 + }, + { + "epoch": 0.547919876733436, + "grad_norm": 0.6297245577393525, + "learning_rate": 4.467928831331478e-06, + "loss": 0.4745, + "step": 24892 + }, + { + "epoch": 0.5479418886198547, + "grad_norm": 12.46874267881038, + "learning_rate": 4.467574400304895e-06, + "loss": 0.3711, + "step": 24893 + }, + { + "epoch": 0.5479639005062734, + "grad_norm": 3.589752390227497, + "learning_rate": 4.46721997198434e-06, + "loss": 0.2809, + "step": 24894 + }, + { + "epoch": 0.547985912392692, + "grad_norm": 1.7283266903436854, + "learning_rate": 4.46686554637161e-06, + "loss": 0.3744, + "step": 24895 + }, + { + "epoch": 0.5480079242791107, + "grad_norm": 1.7410447412022891, + "learning_rate": 4.466511123468505e-06, + "loss": 0.2538, + "step": 24896 + }, + { + "epoch": 0.5480299361655294, + "grad_norm": 1.5279380294073477, + "learning_rate": 4.466156703276831e-06, + "loss": 0.3066, + "step": 24897 + }, + { + "epoch": 0.548051948051948, + "grad_norm": 0.6099510274496618, + "learning_rate": 4.4658022857983855e-06, + "loss": 0.4986, + "step": 24898 + }, + { + "epoch": 0.5480739599383667, + "grad_norm": 1.732814328505293, + "learning_rate": 4.465447871034971e-06, + "loss": 0.3553, + "step": 24899 + }, + { + "epoch": 0.5480959718247854, + "grad_norm": 0.6075078317166169, + "learning_rate": 4.465093458988389e-06, + "loss": 0.4724, + "step": 24900 + }, + { + "epoch": 0.548117983711204, + "grad_norm": 3.3200036210350294, + "learning_rate": 4.4647390496604395e-06, + "loss": 0.2817, + "step": 24901 + }, + { + "epoch": 0.5481399955976227, + "grad_norm": 15.652615191021084, + "learning_rate": 4.464384643052927e-06, + "loss": 0.3209, + "step": 24902 + }, + { + "epoch": 0.5481620074840414, + "grad_norm": 1.565090178662261, + "learning_rate": 4.464030239167649e-06, + "loss": 0.342, + "step": 24903 + }, + { + "epoch": 0.54818401937046, + "grad_norm": 1.399604255354126, + "learning_rate": 4.4636758380064094e-06, + "loss": 0.282, + "step": 24904 + }, + { + "epoch": 0.5482060312568787, + "grad_norm": 4.633785600110308, + "learning_rate": 4.463321439571009e-06, + "loss": 0.2865, + "step": 24905 + }, + { + "epoch": 0.5482280431432974, + "grad_norm": 1.5115344075620674, + "learning_rate": 4.4629670438632454e-06, + "loss": 0.3541, + "step": 24906 + }, + { + "epoch": 0.548250055029716, + "grad_norm": 1.9922854763282287, + "learning_rate": 4.462612650884926e-06, + "loss": 0.352, + "step": 24907 + }, + { + "epoch": 0.5482720669161347, + "grad_norm": 2.4882145499240176, + "learning_rate": 4.462258260637847e-06, + "loss": 0.293, + "step": 24908 + }, + { + "epoch": 0.5482940788025534, + "grad_norm": 1.9038951263975659, + "learning_rate": 4.461903873123812e-06, + "loss": 0.2856, + "step": 24909 + }, + { + "epoch": 0.548316090688972, + "grad_norm": 0.6412817234092321, + "learning_rate": 4.461549488344622e-06, + "loss": 0.4938, + "step": 24910 + }, + { + "epoch": 0.5483381025753907, + "grad_norm": 1.37161511183514, + "learning_rate": 4.461195106302075e-06, + "loss": 0.2444, + "step": 24911 + }, + { + "epoch": 0.5483601144618093, + "grad_norm": 0.622289478129377, + "learning_rate": 4.460840726997979e-06, + "loss": 0.4876, + "step": 24912 + }, + { + "epoch": 0.548382126348228, + "grad_norm": 0.6003554108564305, + "learning_rate": 4.460486350434129e-06, + "loss": 0.4504, + "step": 24913 + }, + { + "epoch": 0.5484041382346467, + "grad_norm": 1.463833365124515, + "learning_rate": 4.460131976612327e-06, + "loss": 0.2676, + "step": 24914 + }, + { + "epoch": 0.5484261501210653, + "grad_norm": 0.6326876926801632, + "learning_rate": 4.459777605534376e-06, + "loss": 0.4776, + "step": 24915 + }, + { + "epoch": 0.548448162007484, + "grad_norm": 3.9369117645915734, + "learning_rate": 4.459423237202075e-06, + "loss": 0.3172, + "step": 24916 + }, + { + "epoch": 0.5484701738939027, + "grad_norm": 0.585262519987575, + "learning_rate": 4.4590688716172275e-06, + "loss": 0.4231, + "step": 24917 + }, + { + "epoch": 0.5484921857803213, + "grad_norm": 3.0556329764949, + "learning_rate": 4.458714508781633e-06, + "loss": 0.3041, + "step": 24918 + }, + { + "epoch": 0.54851419766674, + "grad_norm": 1.6740110940521127, + "learning_rate": 4.458360148697092e-06, + "loss": 0.2772, + "step": 24919 + }, + { + "epoch": 0.5485362095531587, + "grad_norm": 1.3806374297776012, + "learning_rate": 4.458005791365408e-06, + "loss": 0.2704, + "step": 24920 + }, + { + "epoch": 0.5485582214395773, + "grad_norm": 1.5328577008727313, + "learning_rate": 4.457651436788379e-06, + "loss": 0.3846, + "step": 24921 + }, + { + "epoch": 0.548580233325996, + "grad_norm": 0.6321697532968591, + "learning_rate": 4.457297084967808e-06, + "loss": 0.4583, + "step": 24922 + }, + { + "epoch": 0.5486022452124147, + "grad_norm": 1.3489303866088158, + "learning_rate": 4.456942735905496e-06, + "loss": 0.3709, + "step": 24923 + }, + { + "epoch": 0.5486242570988333, + "grad_norm": 1.45825718010633, + "learning_rate": 4.45658838960324e-06, + "loss": 0.2712, + "step": 24924 + }, + { + "epoch": 0.548646268985252, + "grad_norm": 1.5133085242910294, + "learning_rate": 4.456234046062848e-06, + "loss": 0.3474, + "step": 24925 + }, + { + "epoch": 0.5486682808716707, + "grad_norm": 1.4751827328863036, + "learning_rate": 4.4558797052861145e-06, + "loss": 0.3571, + "step": 24926 + }, + { + "epoch": 0.5486902927580893, + "grad_norm": 0.6207347804751638, + "learning_rate": 4.455525367274844e-06, + "loss": 0.4468, + "step": 24927 + }, + { + "epoch": 0.548712304644508, + "grad_norm": 0.6099162530547755, + "learning_rate": 4.455171032030837e-06, + "loss": 0.4683, + "step": 24928 + }, + { + "epoch": 0.5487343165309267, + "grad_norm": 1.6055634311856193, + "learning_rate": 4.4548166995558914e-06, + "loss": 0.2595, + "step": 24929 + }, + { + "epoch": 0.5487563284173453, + "grad_norm": 1.2557549801992454, + "learning_rate": 4.454462369851813e-06, + "loss": 0.2999, + "step": 24930 + }, + { + "epoch": 0.548778340303764, + "grad_norm": 5.073776147442694, + "learning_rate": 4.4541080429204e-06, + "loss": 0.2919, + "step": 24931 + }, + { + "epoch": 0.5488003521901827, + "grad_norm": 1.4075467637006351, + "learning_rate": 4.4537537187634525e-06, + "loss": 0.2756, + "step": 24932 + }, + { + "epoch": 0.5488223640766013, + "grad_norm": 0.6220778888747226, + "learning_rate": 4.453399397382774e-06, + "loss": 0.4823, + "step": 24933 + }, + { + "epoch": 0.54884437596302, + "grad_norm": 0.6354339769542845, + "learning_rate": 4.453045078780161e-06, + "loss": 0.4696, + "step": 24934 + }, + { + "epoch": 0.5488663878494386, + "grad_norm": 1.7301509225443885, + "learning_rate": 4.452690762957418e-06, + "loss": 0.4211, + "step": 24935 + }, + { + "epoch": 0.5488883997358573, + "grad_norm": 2.2788546723310255, + "learning_rate": 4.452336449916345e-06, + "loss": 0.3582, + "step": 24936 + }, + { + "epoch": 0.548910411622276, + "grad_norm": 2.201573294817111, + "learning_rate": 4.451982139658741e-06, + "loss": 0.3309, + "step": 24937 + }, + { + "epoch": 0.5489324235086946, + "grad_norm": 0.6259986874723628, + "learning_rate": 4.45162783218641e-06, + "loss": 0.4818, + "step": 24938 + }, + { + "epoch": 0.5489544353951133, + "grad_norm": 1.4484908879138154, + "learning_rate": 4.45127352750115e-06, + "loss": 0.3446, + "step": 24939 + }, + { + "epoch": 0.548976447281532, + "grad_norm": 1.368735251636041, + "learning_rate": 4.450919225604763e-06, + "loss": 0.2521, + "step": 24940 + }, + { + "epoch": 0.5489984591679506, + "grad_norm": 0.6014695361132565, + "learning_rate": 4.45056492649905e-06, + "loss": 0.4882, + "step": 24941 + }, + { + "epoch": 0.5490204710543694, + "grad_norm": 1.4714887410099242, + "learning_rate": 4.4502106301858085e-06, + "loss": 0.2932, + "step": 24942 + }, + { + "epoch": 0.5490424829407881, + "grad_norm": 0.6171345631795505, + "learning_rate": 4.449856336666845e-06, + "loss": 0.4698, + "step": 24943 + }, + { + "epoch": 0.5490644948272067, + "grad_norm": 0.6288851678207477, + "learning_rate": 4.449502045943955e-06, + "loss": 0.4748, + "step": 24944 + }, + { + "epoch": 0.5490865067136254, + "grad_norm": 0.6212819243194299, + "learning_rate": 4.4491477580189415e-06, + "loss": 0.4741, + "step": 24945 + }, + { + "epoch": 0.5491085186000441, + "grad_norm": 1.2290409882838955, + "learning_rate": 4.448793472893606e-06, + "loss": 0.25, + "step": 24946 + }, + { + "epoch": 0.5491305304864627, + "grad_norm": 0.6112471999917282, + "learning_rate": 4.4484391905697446e-06, + "loss": 0.4841, + "step": 24947 + }, + { + "epoch": 0.5491525423728814, + "grad_norm": 2.2485646661828516, + "learning_rate": 4.448084911049165e-06, + "loss": 0.306, + "step": 24948 + }, + { + "epoch": 0.5491745542593001, + "grad_norm": 1.500552911662868, + "learning_rate": 4.447730634333664e-06, + "loss": 0.4231, + "step": 24949 + }, + { + "epoch": 0.5491965661457187, + "grad_norm": 1.5971165526068212, + "learning_rate": 4.447376360425039e-06, + "loss": 0.3923, + "step": 24950 + }, + { + "epoch": 0.5492185780321374, + "grad_norm": 1.5136023438906825, + "learning_rate": 4.447022089325097e-06, + "loss": 0.2897, + "step": 24951 + }, + { + "epoch": 0.5492405899185561, + "grad_norm": 0.6038997731023269, + "learning_rate": 4.446667821035633e-06, + "loss": 0.4686, + "step": 24952 + }, + { + "epoch": 0.5492626018049747, + "grad_norm": 1.3003555167095502, + "learning_rate": 4.446313555558452e-06, + "loss": 0.3247, + "step": 24953 + }, + { + "epoch": 0.5492846136913934, + "grad_norm": 1.4717714377862499, + "learning_rate": 4.445959292895351e-06, + "loss": 0.3608, + "step": 24954 + }, + { + "epoch": 0.5493066255778121, + "grad_norm": 1.8676912703410737, + "learning_rate": 4.445605033048131e-06, + "loss": 0.358, + "step": 24955 + }, + { + "epoch": 0.5493286374642307, + "grad_norm": 1.5451618177189304, + "learning_rate": 4.445250776018596e-06, + "loss": 0.296, + "step": 24956 + }, + { + "epoch": 0.5493506493506494, + "grad_norm": 1.9614616776813174, + "learning_rate": 4.444896521808543e-06, + "loss": 0.3152, + "step": 24957 + }, + { + "epoch": 0.5493726612370681, + "grad_norm": 1.7816953047688056, + "learning_rate": 4.444542270419772e-06, + "loss": 0.3283, + "step": 24958 + }, + { + "epoch": 0.5493946731234867, + "grad_norm": 1.9010100209688083, + "learning_rate": 4.444188021854088e-06, + "loss": 0.3285, + "step": 24959 + }, + { + "epoch": 0.5494166850099054, + "grad_norm": 1.299529102041574, + "learning_rate": 4.443833776113285e-06, + "loss": 0.2912, + "step": 24960 + }, + { + "epoch": 0.549438696896324, + "grad_norm": 0.6253635027302483, + "learning_rate": 4.44347953319917e-06, + "loss": 0.4638, + "step": 24961 + }, + { + "epoch": 0.5494607087827427, + "grad_norm": 1.542087708386708, + "learning_rate": 4.443125293113538e-06, + "loss": 0.3659, + "step": 24962 + }, + { + "epoch": 0.5494827206691614, + "grad_norm": 0.6091019654246641, + "learning_rate": 4.442771055858192e-06, + "loss": 0.4704, + "step": 24963 + }, + { + "epoch": 0.54950473255558, + "grad_norm": 1.959706965799246, + "learning_rate": 4.442416821434934e-06, + "loss": 0.3432, + "step": 24964 + }, + { + "epoch": 0.5495267444419987, + "grad_norm": 3.622265812094465, + "learning_rate": 4.442062589845558e-06, + "loss": 0.306, + "step": 24965 + }, + { + "epoch": 0.5495487563284174, + "grad_norm": 1.3162632847585871, + "learning_rate": 4.441708361091873e-06, + "loss": 0.2443, + "step": 24966 + }, + { + "epoch": 0.549570768214836, + "grad_norm": 2.109033229243954, + "learning_rate": 4.441354135175674e-06, + "loss": 0.341, + "step": 24967 + }, + { + "epoch": 0.5495927801012547, + "grad_norm": 1.7724255534232398, + "learning_rate": 4.440999912098761e-06, + "loss": 0.4354, + "step": 24968 + }, + { + "epoch": 0.5496147919876734, + "grad_norm": 1.4727375827162654, + "learning_rate": 4.440645691862938e-06, + "loss": 0.343, + "step": 24969 + }, + { + "epoch": 0.549636803874092, + "grad_norm": 1.5073205926319502, + "learning_rate": 4.440291474470001e-06, + "loss": 0.2675, + "step": 24970 + }, + { + "epoch": 0.5496588157605107, + "grad_norm": 1.5557905104186975, + "learning_rate": 4.439937259921753e-06, + "loss": 0.3244, + "step": 24971 + }, + { + "epoch": 0.5496808276469294, + "grad_norm": 1.9881195558457812, + "learning_rate": 4.439583048219994e-06, + "loss": 0.3744, + "step": 24972 + }, + { + "epoch": 0.549702839533348, + "grad_norm": 0.6499413086772708, + "learning_rate": 4.439228839366524e-06, + "loss": 0.4714, + "step": 24973 + }, + { + "epoch": 0.5497248514197667, + "grad_norm": 1.7848001234532016, + "learning_rate": 4.438874633363144e-06, + "loss": 0.2751, + "step": 24974 + }, + { + "epoch": 0.5497468633061854, + "grad_norm": 8.576970077134352, + "learning_rate": 4.43852043021165e-06, + "loss": 0.3109, + "step": 24975 + }, + { + "epoch": 0.549768875192604, + "grad_norm": 1.4533598061151938, + "learning_rate": 4.438166229913849e-06, + "loss": 0.2347, + "step": 24976 + }, + { + "epoch": 0.5497908870790227, + "grad_norm": 1.827777436411631, + "learning_rate": 4.437812032471538e-06, + "loss": 0.3698, + "step": 24977 + }, + { + "epoch": 0.5498128989654414, + "grad_norm": 1.8711374051402347, + "learning_rate": 4.437457837886514e-06, + "loss": 0.3833, + "step": 24978 + }, + { + "epoch": 0.54983491085186, + "grad_norm": 0.5988557249363998, + "learning_rate": 4.437103646160583e-06, + "loss": 0.4663, + "step": 24979 + }, + { + "epoch": 0.5498569227382787, + "grad_norm": 1.5167977922284883, + "learning_rate": 4.43674945729554e-06, + "loss": 0.3532, + "step": 24980 + }, + { + "epoch": 0.5498789346246974, + "grad_norm": 1.3674515449590599, + "learning_rate": 4.436395271293189e-06, + "loss": 0.3586, + "step": 24981 + }, + { + "epoch": 0.549900946511116, + "grad_norm": 1.5289884453426559, + "learning_rate": 4.436041088155329e-06, + "loss": 0.3502, + "step": 24982 + }, + { + "epoch": 0.5499229583975347, + "grad_norm": 1.585788494026465, + "learning_rate": 4.4356869078837575e-06, + "loss": 0.3503, + "step": 24983 + }, + { + "epoch": 0.5499449702839533, + "grad_norm": 1.2771762790028431, + "learning_rate": 4.435332730480279e-06, + "loss": 0.3661, + "step": 24984 + }, + { + "epoch": 0.549966982170372, + "grad_norm": 2.2877366848712946, + "learning_rate": 4.434978555946691e-06, + "loss": 0.358, + "step": 24985 + }, + { + "epoch": 0.5499889940567907, + "grad_norm": 1.4605046445516305, + "learning_rate": 4.434624384284794e-06, + "loss": 0.3291, + "step": 24986 + }, + { + "epoch": 0.5500110059432093, + "grad_norm": 1.1153435821481101, + "learning_rate": 4.4342702154963885e-06, + "loss": 0.2549, + "step": 24987 + }, + { + "epoch": 0.550033017829628, + "grad_norm": 0.612003948339067, + "learning_rate": 4.433916049583272e-06, + "loss": 0.4692, + "step": 24988 + }, + { + "epoch": 0.5500550297160467, + "grad_norm": 2.2399036692872323, + "learning_rate": 4.433561886547249e-06, + "loss": 0.3153, + "step": 24989 + }, + { + "epoch": 0.5500770416024653, + "grad_norm": 1.9837228787176084, + "learning_rate": 4.433207726390117e-06, + "loss": 0.292, + "step": 24990 + }, + { + "epoch": 0.550099053488884, + "grad_norm": 1.5655485873687225, + "learning_rate": 4.4328535691136735e-06, + "loss": 0.3605, + "step": 24991 + }, + { + "epoch": 0.5501210653753027, + "grad_norm": 1.8288768410008074, + "learning_rate": 4.432499414719724e-06, + "loss": 0.3167, + "step": 24992 + }, + { + "epoch": 0.5501430772617213, + "grad_norm": 1.3096134454449182, + "learning_rate": 4.432145263210064e-06, + "loss": 0.3292, + "step": 24993 + }, + { + "epoch": 0.55016508914814, + "grad_norm": 1.1861134204508155, + "learning_rate": 4.431791114586495e-06, + "loss": 0.2081, + "step": 24994 + }, + { + "epoch": 0.5501871010345587, + "grad_norm": 1.4119108353569962, + "learning_rate": 4.431436968850819e-06, + "loss": 0.2581, + "step": 24995 + }, + { + "epoch": 0.5502091129209773, + "grad_norm": 1.3219984178953847, + "learning_rate": 4.431082826004831e-06, + "loss": 0.2572, + "step": 24996 + }, + { + "epoch": 0.550231124807396, + "grad_norm": 0.6044765725064065, + "learning_rate": 4.430728686050336e-06, + "loss": 0.4917, + "step": 24997 + }, + { + "epoch": 0.5502531366938147, + "grad_norm": 1.603024619709826, + "learning_rate": 4.43037454898913e-06, + "loss": 0.2763, + "step": 24998 + }, + { + "epoch": 0.5502751485802333, + "grad_norm": 1.7463923099948377, + "learning_rate": 4.430020414823016e-06, + "loss": 0.3918, + "step": 24999 + }, + { + "epoch": 0.550297160466652, + "grad_norm": 1.72414157890529, + "learning_rate": 4.429666283553793e-06, + "loss": 0.3202, + "step": 25000 + }, + { + "epoch": 0.5503191723530707, + "grad_norm": 1.9446658952729516, + "learning_rate": 4.4293121551832575e-06, + "loss": 0.3546, + "step": 25001 + }, + { + "epoch": 0.5503411842394893, + "grad_norm": 1.264494248002177, + "learning_rate": 4.428958029713215e-06, + "loss": 0.3473, + "step": 25002 + }, + { + "epoch": 0.550363196125908, + "grad_norm": 15.997817752501804, + "learning_rate": 4.42860390714546e-06, + "loss": 0.366, + "step": 25003 + }, + { + "epoch": 0.5503852080123267, + "grad_norm": 1.4374412962517562, + "learning_rate": 4.428249787481797e-06, + "loss": 0.406, + "step": 25004 + }, + { + "epoch": 0.5504072198987453, + "grad_norm": 1.5291369407680035, + "learning_rate": 4.427895670724024e-06, + "loss": 0.2968, + "step": 25005 + }, + { + "epoch": 0.550429231785164, + "grad_norm": 1.4949099581120904, + "learning_rate": 4.427541556873938e-06, + "loss": 0.2336, + "step": 25006 + }, + { + "epoch": 0.5504512436715826, + "grad_norm": 1.333658331929255, + "learning_rate": 4.427187445933343e-06, + "loss": 0.3129, + "step": 25007 + }, + { + "epoch": 0.5504732555580013, + "grad_norm": 1.5152553264762865, + "learning_rate": 4.426833337904036e-06, + "loss": 0.2982, + "step": 25008 + }, + { + "epoch": 0.55049526744442, + "grad_norm": 0.6333478139286794, + "learning_rate": 4.426479232787817e-06, + "loss": 0.4498, + "step": 25009 + }, + { + "epoch": 0.5505172793308386, + "grad_norm": 2.0100202813715993, + "learning_rate": 4.426125130586489e-06, + "loss": 0.3275, + "step": 25010 + }, + { + "epoch": 0.5505392912172573, + "grad_norm": 1.8943605066596378, + "learning_rate": 4.425771031301846e-06, + "loss": 0.3204, + "step": 25011 + }, + { + "epoch": 0.550561303103676, + "grad_norm": 1.7874636147362493, + "learning_rate": 4.4254169349356926e-06, + "loss": 0.3015, + "step": 25012 + }, + { + "epoch": 0.5505833149900946, + "grad_norm": 1.4359023252952612, + "learning_rate": 4.425062841489827e-06, + "loss": 0.3031, + "step": 25013 + }, + { + "epoch": 0.5506053268765133, + "grad_norm": 0.6234887018925943, + "learning_rate": 4.424708750966045e-06, + "loss": 0.488, + "step": 25014 + }, + { + "epoch": 0.550627338762932, + "grad_norm": 1.7194119442715925, + "learning_rate": 4.424354663366154e-06, + "loss": 0.3286, + "step": 25015 + }, + { + "epoch": 0.5506493506493506, + "grad_norm": 1.1714575791493849, + "learning_rate": 4.424000578691946e-06, + "loss": 0.2489, + "step": 25016 + }, + { + "epoch": 0.5506713625357693, + "grad_norm": 0.6022290763841284, + "learning_rate": 4.423646496945226e-06, + "loss": 0.4784, + "step": 25017 + }, + { + "epoch": 0.550693374422188, + "grad_norm": 1.189233148735049, + "learning_rate": 4.423292418127792e-06, + "loss": 0.1964, + "step": 25018 + }, + { + "epoch": 0.5507153863086066, + "grad_norm": 1.050625892889675, + "learning_rate": 4.42293834224144e-06, + "loss": 0.2138, + "step": 25019 + }, + { + "epoch": 0.5507373981950253, + "grad_norm": 1.6489871932773574, + "learning_rate": 4.422584269287977e-06, + "loss": 0.2868, + "step": 25020 + }, + { + "epoch": 0.550759410081444, + "grad_norm": 2.006640839979731, + "learning_rate": 4.4222301992691944e-06, + "loss": 0.3553, + "step": 25021 + }, + { + "epoch": 0.5507814219678626, + "grad_norm": 2.0481839157460975, + "learning_rate": 4.421876132186898e-06, + "loss": 0.326, + "step": 25022 + }, + { + "epoch": 0.5508034338542813, + "grad_norm": 1.4036356728766661, + "learning_rate": 4.421522068042885e-06, + "loss": 0.359, + "step": 25023 + }, + { + "epoch": 0.5508254457407, + "grad_norm": 1.7981746582880889, + "learning_rate": 4.421168006838954e-06, + "loss": 0.3225, + "step": 25024 + }, + { + "epoch": 0.5508474576271186, + "grad_norm": 1.185709106503438, + "learning_rate": 4.420813948576906e-06, + "loss": 0.298, + "step": 25025 + }, + { + "epoch": 0.5508694695135373, + "grad_norm": 0.6101779475036992, + "learning_rate": 4.420459893258538e-06, + "loss": 0.4779, + "step": 25026 + }, + { + "epoch": 0.550891481399956, + "grad_norm": 1.6464030141677763, + "learning_rate": 4.420105840885653e-06, + "loss": 0.3262, + "step": 25027 + }, + { + "epoch": 0.5509134932863746, + "grad_norm": 0.6045251472505236, + "learning_rate": 4.41975179146005e-06, + "loss": 0.4755, + "step": 25028 + }, + { + "epoch": 0.5509355051727933, + "grad_norm": 2.5516590032049336, + "learning_rate": 4.4193977449835245e-06, + "loss": 0.2506, + "step": 25029 + }, + { + "epoch": 0.5509575170592119, + "grad_norm": 1.5371082208374867, + "learning_rate": 4.419043701457879e-06, + "loss": 0.2484, + "step": 25030 + }, + { + "epoch": 0.5509795289456306, + "grad_norm": 1.3215537135041728, + "learning_rate": 4.4186896608849156e-06, + "loss": 0.2724, + "step": 25031 + }, + { + "epoch": 0.5510015408320493, + "grad_norm": 1.6228532576204644, + "learning_rate": 4.418335623266426e-06, + "loss": 0.3377, + "step": 25032 + }, + { + "epoch": 0.5510235527184679, + "grad_norm": 1.475956770950819, + "learning_rate": 4.417981588604218e-06, + "loss": 0.3229, + "step": 25033 + }, + { + "epoch": 0.5510455646048866, + "grad_norm": 1.8996351262670437, + "learning_rate": 4.417627556900085e-06, + "loss": 0.3381, + "step": 25034 + }, + { + "epoch": 0.5510675764913053, + "grad_norm": 1.9596099573264112, + "learning_rate": 4.417273528155828e-06, + "loss": 0.2765, + "step": 25035 + }, + { + "epoch": 0.5510895883777239, + "grad_norm": 0.628462072698734, + "learning_rate": 4.41691950237325e-06, + "loss": 0.4472, + "step": 25036 + }, + { + "epoch": 0.5511116002641426, + "grad_norm": 2.06511119609555, + "learning_rate": 4.416565479554144e-06, + "loss": 0.3241, + "step": 25037 + }, + { + "epoch": 0.5511336121505613, + "grad_norm": 1.8105127915698707, + "learning_rate": 4.4162114597003145e-06, + "loss": 0.306, + "step": 25038 + }, + { + "epoch": 0.5511556240369799, + "grad_norm": 1.9853437096751763, + "learning_rate": 4.4158574428135564e-06, + "loss": 0.3459, + "step": 25039 + }, + { + "epoch": 0.5511776359233986, + "grad_norm": 1.4156852126147816, + "learning_rate": 4.4155034288956735e-06, + "loss": 0.3783, + "step": 25040 + }, + { + "epoch": 0.5511996478098173, + "grad_norm": 0.6180682549186515, + "learning_rate": 4.415149417948463e-06, + "loss": 0.4687, + "step": 25041 + }, + { + "epoch": 0.5512216596962359, + "grad_norm": 1.5773160541739182, + "learning_rate": 4.414795409973722e-06, + "loss": 0.3289, + "step": 25042 + }, + { + "epoch": 0.5512436715826546, + "grad_norm": 2.080617836409204, + "learning_rate": 4.414441404973253e-06, + "loss": 0.3034, + "step": 25043 + }, + { + "epoch": 0.5512656834690733, + "grad_norm": 1.6972620964500134, + "learning_rate": 4.414087402948853e-06, + "loss": 0.326, + "step": 25044 + }, + { + "epoch": 0.5512876953554919, + "grad_norm": 1.2635079197080135, + "learning_rate": 4.413733403902323e-06, + "loss": 0.2537, + "step": 25045 + }, + { + "epoch": 0.5513097072419106, + "grad_norm": 2.0001628189053378, + "learning_rate": 4.413379407835463e-06, + "loss": 0.2809, + "step": 25046 + }, + { + "epoch": 0.5513317191283293, + "grad_norm": 3.548738223927849, + "learning_rate": 4.413025414750067e-06, + "loss": 0.3276, + "step": 25047 + }, + { + "epoch": 0.5513537310147479, + "grad_norm": 1.4618461404384488, + "learning_rate": 4.41267142464794e-06, + "loss": 0.3466, + "step": 25048 + }, + { + "epoch": 0.5513757429011666, + "grad_norm": 2.041927127770071, + "learning_rate": 4.412317437530879e-06, + "loss": 0.3383, + "step": 25049 + }, + { + "epoch": 0.5513977547875853, + "grad_norm": 1.4341394653491601, + "learning_rate": 4.41196345340068e-06, + "loss": 0.2958, + "step": 25050 + }, + { + "epoch": 0.5514197666740039, + "grad_norm": 1.9600341372957384, + "learning_rate": 4.411609472259149e-06, + "loss": 0.2809, + "step": 25051 + }, + { + "epoch": 0.5514417785604226, + "grad_norm": 2.36925765353905, + "learning_rate": 4.411255494108078e-06, + "loss": 0.1528, + "step": 25052 + }, + { + "epoch": 0.5514637904468412, + "grad_norm": 2.0881187515301094, + "learning_rate": 4.41090151894927e-06, + "loss": 0.3139, + "step": 25053 + }, + { + "epoch": 0.55148580233326, + "grad_norm": 0.6496912196575738, + "learning_rate": 4.410547546784525e-06, + "loss": 0.4903, + "step": 25054 + }, + { + "epoch": 0.5515078142196786, + "grad_norm": 1.225336266002154, + "learning_rate": 4.410193577615638e-06, + "loss": 0.2931, + "step": 25055 + }, + { + "epoch": 0.5515298261060972, + "grad_norm": 3.0947199199546622, + "learning_rate": 4.409839611444411e-06, + "loss": 0.3305, + "step": 25056 + }, + { + "epoch": 0.551551837992516, + "grad_norm": 1.3663707962513194, + "learning_rate": 4.4094856482726425e-06, + "loss": 0.3372, + "step": 25057 + }, + { + "epoch": 0.5515738498789347, + "grad_norm": 1.3296175872448048, + "learning_rate": 4.4091316881021314e-06, + "loss": 0.2783, + "step": 25058 + }, + { + "epoch": 0.5515958617653532, + "grad_norm": 1.7345181896763986, + "learning_rate": 4.4087777309346775e-06, + "loss": 0.4041, + "step": 25059 + }, + { + "epoch": 0.551617873651772, + "grad_norm": 1.41238407044097, + "learning_rate": 4.408423776772076e-06, + "loss": 0.3115, + "step": 25060 + }, + { + "epoch": 0.5516398855381907, + "grad_norm": 1.3080279625855527, + "learning_rate": 4.408069825616131e-06, + "loss": 0.2084, + "step": 25061 + }, + { + "epoch": 0.5516618974246092, + "grad_norm": 1.241935647167239, + "learning_rate": 4.407715877468638e-06, + "loss": 0.2653, + "step": 25062 + }, + { + "epoch": 0.551683909311028, + "grad_norm": 1.5165731680398904, + "learning_rate": 4.407361932331398e-06, + "loss": 0.2794, + "step": 25063 + }, + { + "epoch": 0.5517059211974467, + "grad_norm": 1.674393127994249, + "learning_rate": 4.40700799020621e-06, + "loss": 0.3198, + "step": 25064 + }, + { + "epoch": 0.5517279330838653, + "grad_norm": 0.6175188500402998, + "learning_rate": 4.406654051094869e-06, + "loss": 0.4839, + "step": 25065 + }, + { + "epoch": 0.551749944970284, + "grad_norm": 2.9292918188492365, + "learning_rate": 4.406300114999179e-06, + "loss": 0.3272, + "step": 25066 + }, + { + "epoch": 0.5517719568567027, + "grad_norm": 0.609863754246864, + "learning_rate": 4.405946181920938e-06, + "loss": 0.4718, + "step": 25067 + }, + { + "epoch": 0.5517939687431213, + "grad_norm": 1.8082287470929248, + "learning_rate": 4.405592251861939e-06, + "loss": 0.2872, + "step": 25068 + }, + { + "epoch": 0.55181598062954, + "grad_norm": 1.3282115108730306, + "learning_rate": 4.405238324823989e-06, + "loss": 0.3172, + "step": 25069 + }, + { + "epoch": 0.5518379925159587, + "grad_norm": 1.3965057975957094, + "learning_rate": 4.404884400808881e-06, + "loss": 0.3025, + "step": 25070 + }, + { + "epoch": 0.5518600044023773, + "grad_norm": 2.44552824075696, + "learning_rate": 4.404530479818417e-06, + "loss": 0.297, + "step": 25071 + }, + { + "epoch": 0.551882016288796, + "grad_norm": 1.3639142799808661, + "learning_rate": 4.4041765618543955e-06, + "loss": 0.3056, + "step": 25072 + }, + { + "epoch": 0.5519040281752147, + "grad_norm": 1.168672933536819, + "learning_rate": 4.403822646918613e-06, + "loss": 0.2802, + "step": 25073 + }, + { + "epoch": 0.5519260400616333, + "grad_norm": 0.6257721376477774, + "learning_rate": 4.40346873501287e-06, + "loss": 0.4882, + "step": 25074 + }, + { + "epoch": 0.551948051948052, + "grad_norm": 1.7037196465715838, + "learning_rate": 4.403114826138964e-06, + "loss": 0.327, + "step": 25075 + }, + { + "epoch": 0.5519700638344706, + "grad_norm": 1.5531322640404768, + "learning_rate": 4.402760920298696e-06, + "loss": 0.2736, + "step": 25076 + }, + { + "epoch": 0.5519920757208893, + "grad_norm": 1.5446204443520868, + "learning_rate": 4.402407017493864e-06, + "loss": 0.2886, + "step": 25077 + }, + { + "epoch": 0.552014087607308, + "grad_norm": 1.2907277691626584, + "learning_rate": 4.4020531177262645e-06, + "loss": 0.2662, + "step": 25078 + }, + { + "epoch": 0.5520360994937266, + "grad_norm": 1.530580519827412, + "learning_rate": 4.401699220997698e-06, + "loss": 0.4037, + "step": 25079 + }, + { + "epoch": 0.5520581113801453, + "grad_norm": 1.1219012741435064, + "learning_rate": 4.401345327309963e-06, + "loss": 0.2083, + "step": 25080 + }, + { + "epoch": 0.552080123266564, + "grad_norm": 1.2133895390225935, + "learning_rate": 4.400991436664858e-06, + "loss": 0.3534, + "step": 25081 + }, + { + "epoch": 0.5521021351529826, + "grad_norm": 1.1985955312373477, + "learning_rate": 4.400637549064184e-06, + "loss": 0.2782, + "step": 25082 + }, + { + "epoch": 0.5521241470394013, + "grad_norm": 1.5174185708176604, + "learning_rate": 4.400283664509734e-06, + "loss": 0.3757, + "step": 25083 + }, + { + "epoch": 0.55214615892582, + "grad_norm": 1.3852595027770718, + "learning_rate": 4.399929783003312e-06, + "loss": 0.2763, + "step": 25084 + }, + { + "epoch": 0.5521681708122386, + "grad_norm": 1.292121609578989, + "learning_rate": 4.399575904546712e-06, + "loss": 0.2798, + "step": 25085 + }, + { + "epoch": 0.5521901826986573, + "grad_norm": 1.2404282154970359, + "learning_rate": 4.399222029141738e-06, + "loss": 0.2297, + "step": 25086 + }, + { + "epoch": 0.552212194585076, + "grad_norm": 1.6668187850204776, + "learning_rate": 4.398868156790185e-06, + "loss": 0.48, + "step": 25087 + }, + { + "epoch": 0.5522342064714946, + "grad_norm": 1.4569012095598748, + "learning_rate": 4.398514287493851e-06, + "loss": 0.3231, + "step": 25088 + }, + { + "epoch": 0.5522562183579133, + "grad_norm": 0.6461501577250658, + "learning_rate": 4.398160421254536e-06, + "loss": 0.4775, + "step": 25089 + }, + { + "epoch": 0.552278230244332, + "grad_norm": 1.8573045594561388, + "learning_rate": 4.39780655807404e-06, + "loss": 0.3137, + "step": 25090 + }, + { + "epoch": 0.5523002421307506, + "grad_norm": 1.8860375352873249, + "learning_rate": 4.397452697954157e-06, + "loss": 0.321, + "step": 25091 + }, + { + "epoch": 0.5523222540171693, + "grad_norm": 0.6294675833616515, + "learning_rate": 4.397098840896689e-06, + "loss": 0.4791, + "step": 25092 + }, + { + "epoch": 0.552344265903588, + "grad_norm": 1.6479681203999845, + "learning_rate": 4.396744986903433e-06, + "loss": 0.3128, + "step": 25093 + }, + { + "epoch": 0.5523662777900066, + "grad_norm": 1.4792830349094301, + "learning_rate": 4.396391135976188e-06, + "loss": 0.4083, + "step": 25094 + }, + { + "epoch": 0.5523882896764253, + "grad_norm": 0.5973022634456083, + "learning_rate": 4.3960372881167555e-06, + "loss": 0.4777, + "step": 25095 + }, + { + "epoch": 0.552410301562844, + "grad_norm": 1.5384109476004793, + "learning_rate": 4.395683443326927e-06, + "loss": 0.2712, + "step": 25096 + }, + { + "epoch": 0.5524323134492626, + "grad_norm": 1.831112016044107, + "learning_rate": 4.395329601608507e-06, + "loss": 0.3363, + "step": 25097 + }, + { + "epoch": 0.5524543253356813, + "grad_norm": 1.4298999961067838, + "learning_rate": 4.39497576296329e-06, + "loss": 0.3892, + "step": 25098 + }, + { + "epoch": 0.5524763372221, + "grad_norm": 2.8383193471795796, + "learning_rate": 4.394621927393077e-06, + "loss": 0.3502, + "step": 25099 + }, + { + "epoch": 0.5524983491085186, + "grad_norm": 1.9001872469787977, + "learning_rate": 4.394268094899667e-06, + "loss": 0.276, + "step": 25100 + }, + { + "epoch": 0.5525203609949373, + "grad_norm": 1.6244535419295643, + "learning_rate": 4.393914265484854e-06, + "loss": 0.3128, + "step": 25101 + }, + { + "epoch": 0.5525423728813559, + "grad_norm": 2.0503207374862056, + "learning_rate": 4.393560439150441e-06, + "loss": 0.2551, + "step": 25102 + }, + { + "epoch": 0.5525643847677746, + "grad_norm": 1.7339354027550675, + "learning_rate": 4.393206615898222e-06, + "loss": 0.2612, + "step": 25103 + }, + { + "epoch": 0.5525863966541933, + "grad_norm": 1.685052190665235, + "learning_rate": 4.392852795729999e-06, + "loss": 0.2784, + "step": 25104 + }, + { + "epoch": 0.5526084085406119, + "grad_norm": 1.2989282105869737, + "learning_rate": 4.3924989786475704e-06, + "loss": 0.2739, + "step": 25105 + }, + { + "epoch": 0.5526304204270306, + "grad_norm": 1.6980455193790778, + "learning_rate": 4.392145164652731e-06, + "loss": 0.3469, + "step": 25106 + }, + { + "epoch": 0.5526524323134493, + "grad_norm": 0.6117774274120513, + "learning_rate": 4.391791353747282e-06, + "loss": 0.4788, + "step": 25107 + }, + { + "epoch": 0.5526744441998679, + "grad_norm": 1.4916806542817203, + "learning_rate": 4.391437545933021e-06, + "loss": 0.2782, + "step": 25108 + }, + { + "epoch": 0.5526964560862866, + "grad_norm": 1.5451413505445917, + "learning_rate": 4.391083741211744e-06, + "loss": 0.343, + "step": 25109 + }, + { + "epoch": 0.5527184679727053, + "grad_norm": 0.6279997275985977, + "learning_rate": 4.390729939585252e-06, + "loss": 0.4794, + "step": 25110 + }, + { + "epoch": 0.5527404798591239, + "grad_norm": 1.5073078666200495, + "learning_rate": 4.390376141055341e-06, + "loss": 0.3492, + "step": 25111 + }, + { + "epoch": 0.5527624917455426, + "grad_norm": 1.3704126620597736, + "learning_rate": 4.390022345623812e-06, + "loss": 0.4346, + "step": 25112 + }, + { + "epoch": 0.5527845036319613, + "grad_norm": 1.5030019674031156, + "learning_rate": 4.389668553292462e-06, + "loss": 0.2336, + "step": 25113 + }, + { + "epoch": 0.5528065155183799, + "grad_norm": 1.3305736042626861, + "learning_rate": 4.389314764063087e-06, + "loss": 0.3092, + "step": 25114 + }, + { + "epoch": 0.5528285274047986, + "grad_norm": 4.42767668699522, + "learning_rate": 4.388960977937488e-06, + "loss": 0.2612, + "step": 25115 + }, + { + "epoch": 0.5528505392912173, + "grad_norm": 1.603146955946323, + "learning_rate": 4.38860719491746e-06, + "loss": 0.3071, + "step": 25116 + }, + { + "epoch": 0.5528725511776359, + "grad_norm": 2.3934543044001275, + "learning_rate": 4.388253415004804e-06, + "loss": 0.3343, + "step": 25117 + }, + { + "epoch": 0.5528945630640546, + "grad_norm": 1.3039771347387723, + "learning_rate": 4.387899638201318e-06, + "loss": 0.2852, + "step": 25118 + }, + { + "epoch": 0.5529165749504733, + "grad_norm": 1.4760798039853038, + "learning_rate": 4.387545864508798e-06, + "loss": 0.3629, + "step": 25119 + }, + { + "epoch": 0.5529385868368919, + "grad_norm": 1.37739503250752, + "learning_rate": 4.387192093929044e-06, + "loss": 0.3483, + "step": 25120 + }, + { + "epoch": 0.5529605987233106, + "grad_norm": 1.3338782056142207, + "learning_rate": 4.386838326463851e-06, + "loss": 0.3616, + "step": 25121 + }, + { + "epoch": 0.5529826106097293, + "grad_norm": 1.3910848925847825, + "learning_rate": 4.38648456211502e-06, + "loss": 0.3203, + "step": 25122 + }, + { + "epoch": 0.5530046224961479, + "grad_norm": 1.4438763286523113, + "learning_rate": 4.38613080088435e-06, + "loss": 0.191, + "step": 25123 + }, + { + "epoch": 0.5530266343825666, + "grad_norm": 1.645972214134221, + "learning_rate": 4.385777042773635e-06, + "loss": 0.3443, + "step": 25124 + }, + { + "epoch": 0.5530486462689852, + "grad_norm": 1.1712134017448157, + "learning_rate": 4.385423287784676e-06, + "loss": 0.2591, + "step": 25125 + }, + { + "epoch": 0.5530706581554039, + "grad_norm": 1.3381408203367597, + "learning_rate": 4.38506953591927e-06, + "loss": 0.2263, + "step": 25126 + }, + { + "epoch": 0.5530926700418226, + "grad_norm": 1.815950011910284, + "learning_rate": 4.384715787179214e-06, + "loss": 0.2986, + "step": 25127 + }, + { + "epoch": 0.5531146819282412, + "grad_norm": 2.2122871762237972, + "learning_rate": 4.384362041566307e-06, + "loss": 0.3413, + "step": 25128 + }, + { + "epoch": 0.5531366938146599, + "grad_norm": 1.2762916222538006, + "learning_rate": 4.3840082990823464e-06, + "loss": 0.3378, + "step": 25129 + }, + { + "epoch": 0.5531587057010786, + "grad_norm": 1.526518970901334, + "learning_rate": 4.383654559729131e-06, + "loss": 0.3857, + "step": 25130 + }, + { + "epoch": 0.5531807175874972, + "grad_norm": 1.9090189766595678, + "learning_rate": 4.38330082350846e-06, + "loss": 0.309, + "step": 25131 + }, + { + "epoch": 0.5532027294739159, + "grad_norm": 2.037544695479238, + "learning_rate": 4.382947090422125e-06, + "loss": 0.3787, + "step": 25132 + }, + { + "epoch": 0.5532247413603346, + "grad_norm": 1.474856854126979, + "learning_rate": 4.382593360471931e-06, + "loss": 0.3481, + "step": 25133 + }, + { + "epoch": 0.5532467532467532, + "grad_norm": 1.5450971089054975, + "learning_rate": 4.382239633659671e-06, + "loss": 0.3682, + "step": 25134 + }, + { + "epoch": 0.5532687651331719, + "grad_norm": 1.4956193670853812, + "learning_rate": 4.3818859099871465e-06, + "loss": 0.2823, + "step": 25135 + }, + { + "epoch": 0.5532907770195906, + "grad_norm": 2.216141720962471, + "learning_rate": 4.3815321894561544e-06, + "loss": 0.3372, + "step": 25136 + }, + { + "epoch": 0.5533127889060092, + "grad_norm": 0.632230022664533, + "learning_rate": 4.381178472068489e-06, + "loss": 0.4584, + "step": 25137 + }, + { + "epoch": 0.5533348007924279, + "grad_norm": 2.115416041187312, + "learning_rate": 4.3808247578259515e-06, + "loss": 0.3224, + "step": 25138 + }, + { + "epoch": 0.5533568126788466, + "grad_norm": 1.994482713962467, + "learning_rate": 4.380471046730338e-06, + "loss": 0.3194, + "step": 25139 + }, + { + "epoch": 0.5533788245652652, + "grad_norm": 1.541746846379001, + "learning_rate": 4.3801173387834485e-06, + "loss": 0.3233, + "step": 25140 + }, + { + "epoch": 0.5534008364516839, + "grad_norm": 2.300015681415883, + "learning_rate": 4.379763633987079e-06, + "loss": 0.2747, + "step": 25141 + }, + { + "epoch": 0.5534228483381026, + "grad_norm": 1.4876019762151549, + "learning_rate": 4.379409932343026e-06, + "loss": 0.2934, + "step": 25142 + }, + { + "epoch": 0.5534448602245212, + "grad_norm": 1.4751411280317046, + "learning_rate": 4.37905623385309e-06, + "loss": 0.3749, + "step": 25143 + }, + { + "epoch": 0.5534668721109399, + "grad_norm": 1.3164820972066669, + "learning_rate": 4.378702538519066e-06, + "loss": 0.3424, + "step": 25144 + }, + { + "epoch": 0.5534888839973586, + "grad_norm": 1.4441700995672122, + "learning_rate": 4.378348846342755e-06, + "loss": 0.3645, + "step": 25145 + }, + { + "epoch": 0.5535108958837772, + "grad_norm": 1.6411791181704787, + "learning_rate": 4.3779951573259505e-06, + "loss": 0.2842, + "step": 25146 + }, + { + "epoch": 0.5535329077701959, + "grad_norm": 1.3593766750907297, + "learning_rate": 4.377641471470451e-06, + "loss": 0.4181, + "step": 25147 + }, + { + "epoch": 0.5535549196566145, + "grad_norm": 6.134149942498325, + "learning_rate": 4.377287788778057e-06, + "loss": 0.274, + "step": 25148 + }, + { + "epoch": 0.5535769315430332, + "grad_norm": 1.4662919894272481, + "learning_rate": 4.376934109250565e-06, + "loss": 0.3609, + "step": 25149 + }, + { + "epoch": 0.5535989434294519, + "grad_norm": 0.6221981926125436, + "learning_rate": 4.376580432889769e-06, + "loss": 0.4731, + "step": 25150 + }, + { + "epoch": 0.5536209553158705, + "grad_norm": 0.6129611418825345, + "learning_rate": 4.376226759697471e-06, + "loss": 0.4662, + "step": 25151 + }, + { + "epoch": 0.5536429672022892, + "grad_norm": 0.6034173295901567, + "learning_rate": 4.375873089675466e-06, + "loss": 0.4511, + "step": 25152 + }, + { + "epoch": 0.5536649790887079, + "grad_norm": 1.4556800387414797, + "learning_rate": 4.375519422825552e-06, + "loss": 0.3045, + "step": 25153 + }, + { + "epoch": 0.5536869909751265, + "grad_norm": 1.1293232025831563, + "learning_rate": 4.37516575914953e-06, + "loss": 0.2965, + "step": 25154 + }, + { + "epoch": 0.5537090028615452, + "grad_norm": 0.608920979930995, + "learning_rate": 4.374812098649191e-06, + "loss": 0.4611, + "step": 25155 + }, + { + "epoch": 0.5537310147479639, + "grad_norm": 1.7648058430236373, + "learning_rate": 4.374458441326337e-06, + "loss": 0.2916, + "step": 25156 + }, + { + "epoch": 0.5537530266343825, + "grad_norm": 1.3822501628651367, + "learning_rate": 4.374104787182763e-06, + "loss": 0.2014, + "step": 25157 + }, + { + "epoch": 0.5537750385208012, + "grad_norm": 1.5092757506631613, + "learning_rate": 4.3737511362202685e-06, + "loss": 0.2513, + "step": 25158 + }, + { + "epoch": 0.5537970504072199, + "grad_norm": 1.668336811864797, + "learning_rate": 4.373397488440652e-06, + "loss": 0.4031, + "step": 25159 + }, + { + "epoch": 0.5538190622936385, + "grad_norm": 1.5210456139880817, + "learning_rate": 4.373043843845706e-06, + "loss": 0.2929, + "step": 25160 + }, + { + "epoch": 0.5538410741800572, + "grad_norm": 1.8689417749475317, + "learning_rate": 4.372690202437231e-06, + "loss": 0.2669, + "step": 25161 + }, + { + "epoch": 0.5538630860664759, + "grad_norm": 0.6114675439712295, + "learning_rate": 4.372336564217025e-06, + "loss": 0.4609, + "step": 25162 + }, + { + "epoch": 0.5538850979528945, + "grad_norm": 1.3293001905117638, + "learning_rate": 4.371982929186885e-06, + "loss": 0.2918, + "step": 25163 + }, + { + "epoch": 0.5539071098393132, + "grad_norm": 3.514104596289953, + "learning_rate": 4.371629297348608e-06, + "loss": 0.3377, + "step": 25164 + }, + { + "epoch": 0.5539291217257319, + "grad_norm": 7.973759399627385, + "learning_rate": 4.37127566870399e-06, + "loss": 0.3005, + "step": 25165 + }, + { + "epoch": 0.5539511336121505, + "grad_norm": 1.5899192691306763, + "learning_rate": 4.37092204325483e-06, + "loss": 0.3037, + "step": 25166 + }, + { + "epoch": 0.5539731454985692, + "grad_norm": 1.2642470337310876, + "learning_rate": 4.3705684210029255e-06, + "loss": 0.2585, + "step": 25167 + }, + { + "epoch": 0.553995157384988, + "grad_norm": 1.5395765975541167, + "learning_rate": 4.370214801950072e-06, + "loss": 0.3081, + "step": 25168 + }, + { + "epoch": 0.5540171692714065, + "grad_norm": 1.1941076144090126, + "learning_rate": 4.369861186098068e-06, + "loss": 0.2806, + "step": 25169 + }, + { + "epoch": 0.5540391811578252, + "grad_norm": 1.3114905042912006, + "learning_rate": 4.3695075734487094e-06, + "loss": 0.2409, + "step": 25170 + }, + { + "epoch": 0.5540611930442438, + "grad_norm": 1.3685467514098397, + "learning_rate": 4.369153964003796e-06, + "loss": 0.3723, + "step": 25171 + }, + { + "epoch": 0.5540832049306625, + "grad_norm": 1.2534312407782968, + "learning_rate": 4.368800357765125e-06, + "loss": 0.2845, + "step": 25172 + }, + { + "epoch": 0.5541052168170812, + "grad_norm": 1.4318865063738029, + "learning_rate": 4.368446754734489e-06, + "loss": 0.342, + "step": 25173 + }, + { + "epoch": 0.5541272287034998, + "grad_norm": 1.7577849483923718, + "learning_rate": 4.3680931549136895e-06, + "loss": 0.376, + "step": 25174 + }, + { + "epoch": 0.5541492405899185, + "grad_norm": 4.758023745112148, + "learning_rate": 4.367739558304522e-06, + "loss": 0.3124, + "step": 25175 + }, + { + "epoch": 0.5541712524763373, + "grad_norm": 1.4225761294007306, + "learning_rate": 4.367385964908784e-06, + "loss": 0.3604, + "step": 25176 + }, + { + "epoch": 0.5541932643627558, + "grad_norm": 1.706087811489289, + "learning_rate": 4.367032374728275e-06, + "loss": 0.3974, + "step": 25177 + }, + { + "epoch": 0.5542152762491745, + "grad_norm": 1.8206074267427867, + "learning_rate": 4.366678787764787e-06, + "loss": 0.2577, + "step": 25178 + }, + { + "epoch": 0.5542372881355933, + "grad_norm": 1.764376805792436, + "learning_rate": 4.366325204020122e-06, + "loss": 0.3477, + "step": 25179 + }, + { + "epoch": 0.5542593000220118, + "grad_norm": 1.2935351540748123, + "learning_rate": 4.365971623496073e-06, + "loss": 0.2815, + "step": 25180 + }, + { + "epoch": 0.5542813119084306, + "grad_norm": 1.5654927170096025, + "learning_rate": 4.3656180461944404e-06, + "loss": 0.3072, + "step": 25181 + }, + { + "epoch": 0.5543033237948493, + "grad_norm": 1.8085097230395788, + "learning_rate": 4.3652644721170195e-06, + "loss": 0.3112, + "step": 25182 + }, + { + "epoch": 0.5543253356812679, + "grad_norm": 0.6029316287834493, + "learning_rate": 4.364910901265607e-06, + "loss": 0.4702, + "step": 25183 + }, + { + "epoch": 0.5543473475676866, + "grad_norm": 0.6103101949832238, + "learning_rate": 4.364557333642001e-06, + "loss": 0.4748, + "step": 25184 + }, + { + "epoch": 0.5543693594541053, + "grad_norm": 1.6346646491229677, + "learning_rate": 4.364203769247997e-06, + "loss": 0.2332, + "step": 25185 + }, + { + "epoch": 0.5543913713405239, + "grad_norm": 0.6176280054840018, + "learning_rate": 4.363850208085395e-06, + "loss": 0.482, + "step": 25186 + }, + { + "epoch": 0.5544133832269426, + "grad_norm": 0.615598061594499, + "learning_rate": 4.36349665015599e-06, + "loss": 0.4819, + "step": 25187 + }, + { + "epoch": 0.5544353951133613, + "grad_norm": 1.4246645230296997, + "learning_rate": 4.363143095461577e-06, + "loss": 0.3524, + "step": 25188 + }, + { + "epoch": 0.5544574069997799, + "grad_norm": 1.1719076344893238, + "learning_rate": 4.362789544003955e-06, + "loss": 0.3669, + "step": 25189 + }, + { + "epoch": 0.5544794188861986, + "grad_norm": 1.5222812377648, + "learning_rate": 4.362435995784924e-06, + "loss": 0.2672, + "step": 25190 + }, + { + "epoch": 0.5545014307726173, + "grad_norm": 1.9677440982235808, + "learning_rate": 4.3620824508062735e-06, + "loss": 0.303, + "step": 25191 + }, + { + "epoch": 0.5545234426590359, + "grad_norm": 1.5152076113828874, + "learning_rate": 4.361728909069806e-06, + "loss": 0.2404, + "step": 25192 + }, + { + "epoch": 0.5545454545454546, + "grad_norm": 1.6547587153207861, + "learning_rate": 4.361375370577316e-06, + "loss": 0.3506, + "step": 25193 + }, + { + "epoch": 0.5545674664318732, + "grad_norm": 2.0930240233498085, + "learning_rate": 4.361021835330603e-06, + "loss": 0.3633, + "step": 25194 + }, + { + "epoch": 0.5545894783182919, + "grad_norm": 1.0738048953701576, + "learning_rate": 4.360668303331462e-06, + "loss": 0.2709, + "step": 25195 + }, + { + "epoch": 0.5546114902047106, + "grad_norm": 1.26708214241109, + "learning_rate": 4.360314774581689e-06, + "loss": 0.3631, + "step": 25196 + }, + { + "epoch": 0.5546335020911292, + "grad_norm": 1.6696242210155152, + "learning_rate": 4.359961249083081e-06, + "loss": 0.3556, + "step": 25197 + }, + { + "epoch": 0.5546555139775479, + "grad_norm": 1.5527620894946395, + "learning_rate": 4.359607726837435e-06, + "loss": 0.3102, + "step": 25198 + }, + { + "epoch": 0.5546775258639666, + "grad_norm": 1.8586514261869815, + "learning_rate": 4.359254207846551e-06, + "loss": 0.3405, + "step": 25199 + }, + { + "epoch": 0.5546995377503852, + "grad_norm": 1.6871628102867415, + "learning_rate": 4.358900692112221e-06, + "loss": 0.2882, + "step": 25200 + }, + { + "epoch": 0.5547215496368039, + "grad_norm": 1.2463855363661664, + "learning_rate": 4.358547179636243e-06, + "loss": 0.2412, + "step": 25201 + }, + { + "epoch": 0.5547435615232226, + "grad_norm": 1.2658548583064733, + "learning_rate": 4.358193670420416e-06, + "loss": 0.2549, + "step": 25202 + }, + { + "epoch": 0.5547655734096412, + "grad_norm": 1.3819475313426326, + "learning_rate": 4.357840164466533e-06, + "loss": 0.2109, + "step": 25203 + }, + { + "epoch": 0.5547875852960599, + "grad_norm": 1.5175418302979897, + "learning_rate": 4.357486661776395e-06, + "loss": 0.4179, + "step": 25204 + }, + { + "epoch": 0.5548095971824786, + "grad_norm": 2.7662440239025066, + "learning_rate": 4.357133162351796e-06, + "loss": 0.309, + "step": 25205 + }, + { + "epoch": 0.5548316090688972, + "grad_norm": 0.6378260229473506, + "learning_rate": 4.3567796661945305e-06, + "loss": 0.4509, + "step": 25206 + }, + { + "epoch": 0.5548536209553159, + "grad_norm": 1.7935043685517689, + "learning_rate": 4.356426173306399e-06, + "loss": 0.4215, + "step": 25207 + }, + { + "epoch": 0.5548756328417346, + "grad_norm": 1.3688086995207869, + "learning_rate": 4.356072683689198e-06, + "loss": 0.3239, + "step": 25208 + }, + { + "epoch": 0.5548976447281532, + "grad_norm": 0.6521171729688219, + "learning_rate": 4.355719197344722e-06, + "loss": 0.4847, + "step": 25209 + }, + { + "epoch": 0.5549196566145719, + "grad_norm": 1.3923214935986787, + "learning_rate": 4.355365714274768e-06, + "loss": 0.3519, + "step": 25210 + }, + { + "epoch": 0.5549416685009906, + "grad_norm": 1.2324928008730358, + "learning_rate": 4.355012234481132e-06, + "loss": 0.2439, + "step": 25211 + }, + { + "epoch": 0.5549636803874092, + "grad_norm": 2.601486722063899, + "learning_rate": 4.354658757965613e-06, + "loss": 0.3057, + "step": 25212 + }, + { + "epoch": 0.5549856922738279, + "grad_norm": 1.3737610631037858, + "learning_rate": 4.354305284730007e-06, + "loss": 0.279, + "step": 25213 + }, + { + "epoch": 0.5550077041602466, + "grad_norm": 1.7331898645290356, + "learning_rate": 4.3539518147761076e-06, + "loss": 0.3276, + "step": 25214 + }, + { + "epoch": 0.5550297160466652, + "grad_norm": 1.3825878271731213, + "learning_rate": 4.353598348105714e-06, + "loss": 0.3679, + "step": 25215 + }, + { + "epoch": 0.5550517279330839, + "grad_norm": 1.7772478068728734, + "learning_rate": 4.353244884720621e-06, + "loss": 0.3967, + "step": 25216 + }, + { + "epoch": 0.5550737398195025, + "grad_norm": 1.5091897215898749, + "learning_rate": 4.352891424622627e-06, + "loss": 0.2582, + "step": 25217 + }, + { + "epoch": 0.5550957517059212, + "grad_norm": 1.2331547502839, + "learning_rate": 4.352537967813527e-06, + "loss": 0.2372, + "step": 25218 + }, + { + "epoch": 0.5551177635923399, + "grad_norm": 1.3321756045731588, + "learning_rate": 4.352184514295118e-06, + "loss": 0.2651, + "step": 25219 + }, + { + "epoch": 0.5551397754787585, + "grad_norm": 0.6874279703777821, + "learning_rate": 4.351831064069196e-06, + "loss": 0.4726, + "step": 25220 + }, + { + "epoch": 0.5551617873651772, + "grad_norm": 1.3525273918182559, + "learning_rate": 4.3514776171375575e-06, + "loss": 0.3341, + "step": 25221 + }, + { + "epoch": 0.5551837992515959, + "grad_norm": 4.097163095244118, + "learning_rate": 4.351124173502e-06, + "loss": 0.3115, + "step": 25222 + }, + { + "epoch": 0.5552058111380145, + "grad_norm": 1.314380218383354, + "learning_rate": 4.350770733164319e-06, + "loss": 0.3088, + "step": 25223 + }, + { + "epoch": 0.5552278230244332, + "grad_norm": 4.032343989674859, + "learning_rate": 4.350417296126308e-06, + "loss": 0.2692, + "step": 25224 + }, + { + "epoch": 0.5552498349108519, + "grad_norm": 1.883704636922687, + "learning_rate": 4.350063862389769e-06, + "loss": 0.3906, + "step": 25225 + }, + { + "epoch": 0.5552718467972705, + "grad_norm": 3.4423434033579365, + "learning_rate": 4.3497104319564945e-06, + "loss": 0.267, + "step": 25226 + }, + { + "epoch": 0.5552938586836892, + "grad_norm": 1.652889413540482, + "learning_rate": 4.349357004828281e-06, + "loss": 0.2625, + "step": 25227 + }, + { + "epoch": 0.5553158705701079, + "grad_norm": 1.570172696587558, + "learning_rate": 4.349003581006926e-06, + "loss": 0.3225, + "step": 25228 + }, + { + "epoch": 0.5553378824565265, + "grad_norm": 1.9156177404347523, + "learning_rate": 4.348650160494223e-06, + "loss": 0.2819, + "step": 25229 + }, + { + "epoch": 0.5553598943429452, + "grad_norm": 1.6643209332932674, + "learning_rate": 4.348296743291973e-06, + "loss": 0.3191, + "step": 25230 + }, + { + "epoch": 0.5553819062293639, + "grad_norm": 1.5436814585191214, + "learning_rate": 4.34794332940197e-06, + "loss": 0.3337, + "step": 25231 + }, + { + "epoch": 0.5554039181157825, + "grad_norm": 0.6133321073667343, + "learning_rate": 4.347589918826008e-06, + "loss": 0.4832, + "step": 25232 + }, + { + "epoch": 0.5554259300022012, + "grad_norm": 1.3518007156778324, + "learning_rate": 4.347236511565886e-06, + "loss": 0.3323, + "step": 25233 + }, + { + "epoch": 0.5554479418886199, + "grad_norm": 1.931084024380171, + "learning_rate": 4.346883107623398e-06, + "loss": 0.3319, + "step": 25234 + }, + { + "epoch": 0.5554699537750385, + "grad_norm": 2.0680821700490113, + "learning_rate": 4.346529707000343e-06, + "loss": 0.373, + "step": 25235 + }, + { + "epoch": 0.5554919656614572, + "grad_norm": 1.7623483830914364, + "learning_rate": 4.346176309698515e-06, + "loss": 0.3328, + "step": 25236 + }, + { + "epoch": 0.5555139775478759, + "grad_norm": 2.995158154734642, + "learning_rate": 4.34582291571971e-06, + "loss": 0.2645, + "step": 25237 + }, + { + "epoch": 0.5555359894342945, + "grad_norm": 1.1626379997406218, + "learning_rate": 4.345469525065725e-06, + "loss": 0.2318, + "step": 25238 + }, + { + "epoch": 0.5555580013207132, + "grad_norm": 1.6467734625725619, + "learning_rate": 4.345116137738355e-06, + "loss": 0.262, + "step": 25239 + }, + { + "epoch": 0.5555800132071319, + "grad_norm": 1.547916351229816, + "learning_rate": 4.344762753739399e-06, + "loss": 0.3048, + "step": 25240 + }, + { + "epoch": 0.5556020250935505, + "grad_norm": 1.5230706886427614, + "learning_rate": 4.34440937307065e-06, + "loss": 0.302, + "step": 25241 + }, + { + "epoch": 0.5556240369799692, + "grad_norm": 1.2370542868314542, + "learning_rate": 4.344055995733903e-06, + "loss": 0.3134, + "step": 25242 + }, + { + "epoch": 0.5556460488663878, + "grad_norm": 1.52462796120137, + "learning_rate": 4.343702621730959e-06, + "loss": 0.352, + "step": 25243 + }, + { + "epoch": 0.5556680607528065, + "grad_norm": 1.447407696996518, + "learning_rate": 4.3433492510636075e-06, + "loss": 0.2806, + "step": 25244 + }, + { + "epoch": 0.5556900726392252, + "grad_norm": 1.4262932207979062, + "learning_rate": 4.342995883733652e-06, + "loss": 0.3385, + "step": 25245 + }, + { + "epoch": 0.5557120845256438, + "grad_norm": 1.231023630617537, + "learning_rate": 4.342642519742882e-06, + "loss": 0.2677, + "step": 25246 + }, + { + "epoch": 0.5557340964120625, + "grad_norm": 0.6077745583958444, + "learning_rate": 4.342289159093096e-06, + "loss": 0.4583, + "step": 25247 + }, + { + "epoch": 0.5557561082984812, + "grad_norm": 1.7258122832381497, + "learning_rate": 4.341935801786091e-06, + "loss": 0.3856, + "step": 25248 + }, + { + "epoch": 0.5557781201848998, + "grad_norm": 1.5402286885823446, + "learning_rate": 4.341582447823662e-06, + "loss": 0.3004, + "step": 25249 + }, + { + "epoch": 0.5558001320713185, + "grad_norm": 1.1861242066539124, + "learning_rate": 4.341229097207603e-06, + "loss": 0.301, + "step": 25250 + }, + { + "epoch": 0.5558221439577372, + "grad_norm": 1.5618143433106269, + "learning_rate": 4.340875749939712e-06, + "loss": 0.3127, + "step": 25251 + }, + { + "epoch": 0.5558441558441558, + "grad_norm": 1.8330856018105948, + "learning_rate": 4.3405224060217835e-06, + "loss": 0.2373, + "step": 25252 + }, + { + "epoch": 0.5558661677305745, + "grad_norm": 1.442796091007278, + "learning_rate": 4.340169065455618e-06, + "loss": 0.3316, + "step": 25253 + }, + { + "epoch": 0.5558881796169932, + "grad_norm": 0.6017528574224, + "learning_rate": 4.339815728243004e-06, + "loss": 0.483, + "step": 25254 + }, + { + "epoch": 0.5559101915034118, + "grad_norm": 1.4735133287072169, + "learning_rate": 4.339462394385742e-06, + "loss": 0.3381, + "step": 25255 + }, + { + "epoch": 0.5559322033898305, + "grad_norm": 0.605954101992466, + "learning_rate": 4.339109063885626e-06, + "loss": 0.4731, + "step": 25256 + }, + { + "epoch": 0.5559542152762492, + "grad_norm": 1.4097729250152775, + "learning_rate": 4.338755736744452e-06, + "loss": 0.2798, + "step": 25257 + }, + { + "epoch": 0.5559762271626678, + "grad_norm": 1.4576984447377357, + "learning_rate": 4.33840241296402e-06, + "loss": 0.3396, + "step": 25258 + }, + { + "epoch": 0.5559982390490865, + "grad_norm": 0.5844029376090119, + "learning_rate": 4.338049092546119e-06, + "loss": 0.456, + "step": 25259 + }, + { + "epoch": 0.5560202509355052, + "grad_norm": 1.329771616339673, + "learning_rate": 4.3376957754925476e-06, + "loss": 0.3104, + "step": 25260 + }, + { + "epoch": 0.5560422628219238, + "grad_norm": 1.7819502636675655, + "learning_rate": 4.337342461805102e-06, + "loss": 0.3443, + "step": 25261 + }, + { + "epoch": 0.5560642747083425, + "grad_norm": 1.38528292646815, + "learning_rate": 4.336989151485577e-06, + "loss": 0.2994, + "step": 25262 + }, + { + "epoch": 0.5560862865947612, + "grad_norm": 1.3847087172437638, + "learning_rate": 4.336635844535771e-06, + "loss": 0.3412, + "step": 25263 + }, + { + "epoch": 0.5561082984811798, + "grad_norm": 0.6166379135237008, + "learning_rate": 4.336282540957477e-06, + "loss": 0.4638, + "step": 25264 + }, + { + "epoch": 0.5561303103675985, + "grad_norm": 0.6270717073958805, + "learning_rate": 4.335929240752489e-06, + "loss": 0.4919, + "step": 25265 + }, + { + "epoch": 0.5561523222540171, + "grad_norm": 1.872709009255098, + "learning_rate": 4.335575943922608e-06, + "loss": 0.3589, + "step": 25266 + }, + { + "epoch": 0.5561743341404358, + "grad_norm": 1.319612942976721, + "learning_rate": 4.335222650469624e-06, + "loss": 0.3121, + "step": 25267 + }, + { + "epoch": 0.5561963460268545, + "grad_norm": 3.0882707151861606, + "learning_rate": 4.3348693603953344e-06, + "loss": 0.3414, + "step": 25268 + }, + { + "epoch": 0.5562183579132731, + "grad_norm": 1.687344246240633, + "learning_rate": 4.334516073701537e-06, + "loss": 0.35, + "step": 25269 + }, + { + "epoch": 0.5562403697996918, + "grad_norm": 1.6547994162435746, + "learning_rate": 4.334162790390024e-06, + "loss": 0.3496, + "step": 25270 + }, + { + "epoch": 0.5562623816861105, + "grad_norm": 1.7768112770772206, + "learning_rate": 4.333809510462595e-06, + "loss": 0.2682, + "step": 25271 + }, + { + "epoch": 0.5562843935725291, + "grad_norm": 1.4588923885119032, + "learning_rate": 4.3334562339210425e-06, + "loss": 0.4135, + "step": 25272 + }, + { + "epoch": 0.5563064054589478, + "grad_norm": 0.6248370077559209, + "learning_rate": 4.333102960767162e-06, + "loss": 0.4654, + "step": 25273 + }, + { + "epoch": 0.5563284173453665, + "grad_norm": 1.264930319206553, + "learning_rate": 4.33274969100275e-06, + "loss": 0.2438, + "step": 25274 + }, + { + "epoch": 0.5563504292317851, + "grad_norm": 0.5928195759113271, + "learning_rate": 4.3323964246296e-06, + "loss": 0.4479, + "step": 25275 + }, + { + "epoch": 0.5563724411182038, + "grad_norm": 1.862658097100159, + "learning_rate": 4.332043161649513e-06, + "loss": 0.4522, + "step": 25276 + }, + { + "epoch": 0.5563944530046225, + "grad_norm": 0.6317474140307405, + "learning_rate": 4.331689902064278e-06, + "loss": 0.473, + "step": 25277 + }, + { + "epoch": 0.5564164648910411, + "grad_norm": 1.1852920128242332, + "learning_rate": 4.3313366458756925e-06, + "loss": 0.2997, + "step": 25278 + }, + { + "epoch": 0.5564384767774598, + "grad_norm": 0.6266878410879959, + "learning_rate": 4.330983393085554e-06, + "loss": 0.4724, + "step": 25279 + }, + { + "epoch": 0.5564604886638785, + "grad_norm": 1.2610512823355455, + "learning_rate": 4.330630143695654e-06, + "loss": 0.2399, + "step": 25280 + }, + { + "epoch": 0.5564825005502971, + "grad_norm": 1.9637490007110427, + "learning_rate": 4.330276897707793e-06, + "loss": 0.3259, + "step": 25281 + }, + { + "epoch": 0.5565045124367158, + "grad_norm": 1.4213724841712914, + "learning_rate": 4.329923655123762e-06, + "loss": 0.2812, + "step": 25282 + }, + { + "epoch": 0.5565265243231345, + "grad_norm": 1.7267480407956575, + "learning_rate": 4.329570415945358e-06, + "loss": 0.2617, + "step": 25283 + }, + { + "epoch": 0.5565485362095531, + "grad_norm": 1.640171721936158, + "learning_rate": 4.329217180174377e-06, + "loss": 0.2487, + "step": 25284 + }, + { + "epoch": 0.5565705480959718, + "grad_norm": 1.4329644911818484, + "learning_rate": 4.32886394781261e-06, + "loss": 0.2904, + "step": 25285 + }, + { + "epoch": 0.5565925599823905, + "grad_norm": 2.287069616741023, + "learning_rate": 4.3285107188618605e-06, + "loss": 0.3687, + "step": 25286 + }, + { + "epoch": 0.5566145718688091, + "grad_norm": 1.8638489099809903, + "learning_rate": 4.3281574933239175e-06, + "loss": 0.3459, + "step": 25287 + }, + { + "epoch": 0.5566365837552278, + "grad_norm": 1.4913034932556073, + "learning_rate": 4.327804271200576e-06, + "loss": 0.2692, + "step": 25288 + }, + { + "epoch": 0.5566585956416464, + "grad_norm": 4.589358921994532, + "learning_rate": 4.327451052493635e-06, + "loss": 0.3263, + "step": 25289 + }, + { + "epoch": 0.5566806075280651, + "grad_norm": 1.2192261499299522, + "learning_rate": 4.327097837204888e-06, + "loss": 0.2599, + "step": 25290 + }, + { + "epoch": 0.5567026194144838, + "grad_norm": 1.5674406066924564, + "learning_rate": 4.326744625336128e-06, + "loss": 0.3034, + "step": 25291 + }, + { + "epoch": 0.5567246313009024, + "grad_norm": 1.2644633605406805, + "learning_rate": 4.326391416889153e-06, + "loss": 0.2921, + "step": 25292 + }, + { + "epoch": 0.5567466431873211, + "grad_norm": 1.2122841254457073, + "learning_rate": 4.326038211865756e-06, + "loss": 0.2232, + "step": 25293 + }, + { + "epoch": 0.5567686550737398, + "grad_norm": 1.345976906274907, + "learning_rate": 4.3256850102677365e-06, + "loss": 0.2896, + "step": 25294 + }, + { + "epoch": 0.5567906669601584, + "grad_norm": 1.145822204595256, + "learning_rate": 4.325331812096884e-06, + "loss": 0.2683, + "step": 25295 + }, + { + "epoch": 0.5568126788465771, + "grad_norm": 1.7020780731787593, + "learning_rate": 4.3249786173549965e-06, + "loss": 0.3975, + "step": 25296 + }, + { + "epoch": 0.5568346907329959, + "grad_norm": 1.8781102915627297, + "learning_rate": 4.3246254260438684e-06, + "loss": 0.3227, + "step": 25297 + }, + { + "epoch": 0.5568567026194144, + "grad_norm": 1.4069742247242139, + "learning_rate": 4.324272238165295e-06, + "loss": 0.3094, + "step": 25298 + }, + { + "epoch": 0.5568787145058332, + "grad_norm": 1.820584262529505, + "learning_rate": 4.323919053721074e-06, + "loss": 0.3021, + "step": 25299 + }, + { + "epoch": 0.5569007263922519, + "grad_norm": 2.1754119826569402, + "learning_rate": 4.323565872712996e-06, + "loss": 0.2922, + "step": 25300 + }, + { + "epoch": 0.5569227382786704, + "grad_norm": 1.743925808989691, + "learning_rate": 4.323212695142857e-06, + "loss": 0.4539, + "step": 25301 + }, + { + "epoch": 0.5569447501650892, + "grad_norm": 0.6599815740950085, + "learning_rate": 4.322859521012455e-06, + "loss": 0.4813, + "step": 25302 + }, + { + "epoch": 0.5569667620515079, + "grad_norm": 1.755511241769375, + "learning_rate": 4.32250635032358e-06, + "loss": 0.3056, + "step": 25303 + }, + { + "epoch": 0.5569887739379265, + "grad_norm": 1.9064048481017613, + "learning_rate": 4.322153183078034e-06, + "loss": 0.3657, + "step": 25304 + }, + { + "epoch": 0.5570107858243452, + "grad_norm": 1.280951844333757, + "learning_rate": 4.321800019277605e-06, + "loss": 0.2172, + "step": 25305 + }, + { + "epoch": 0.5570327977107639, + "grad_norm": 0.6155815640795178, + "learning_rate": 4.321446858924091e-06, + "loss": 0.4715, + "step": 25306 + }, + { + "epoch": 0.5570548095971825, + "grad_norm": 2.291797013989265, + "learning_rate": 4.321093702019288e-06, + "loss": 0.2879, + "step": 25307 + }, + { + "epoch": 0.5570768214836012, + "grad_norm": 1.2489213751815946, + "learning_rate": 4.320740548564989e-06, + "loss": 0.2648, + "step": 25308 + }, + { + "epoch": 0.5570988333700199, + "grad_norm": 1.148635563709267, + "learning_rate": 4.3203873985629885e-06, + "loss": 0.2203, + "step": 25309 + }, + { + "epoch": 0.5571208452564385, + "grad_norm": 1.270985173167821, + "learning_rate": 4.3200342520150826e-06, + "loss": 0.2298, + "step": 25310 + }, + { + "epoch": 0.5571428571428572, + "grad_norm": 1.3469893440933234, + "learning_rate": 4.319681108923066e-06, + "loss": 0.348, + "step": 25311 + }, + { + "epoch": 0.5571648690292758, + "grad_norm": 2.5918615485922207, + "learning_rate": 4.319327969288735e-06, + "loss": 0.385, + "step": 25312 + }, + { + "epoch": 0.5571868809156945, + "grad_norm": 1.7339536975917398, + "learning_rate": 4.318974833113882e-06, + "loss": 0.2896, + "step": 25313 + }, + { + "epoch": 0.5572088928021132, + "grad_norm": 1.2778071197022591, + "learning_rate": 4.318621700400302e-06, + "loss": 0.199, + "step": 25314 + }, + { + "epoch": 0.5572309046885318, + "grad_norm": 2.2470727896951836, + "learning_rate": 4.31826857114979e-06, + "loss": 0.2952, + "step": 25315 + }, + { + "epoch": 0.5572529165749505, + "grad_norm": 1.5909581008373743, + "learning_rate": 4.317915445364142e-06, + "loss": 0.3015, + "step": 25316 + }, + { + "epoch": 0.5572749284613692, + "grad_norm": 1.4952793120975751, + "learning_rate": 4.3175623230451526e-06, + "loss": 0.2806, + "step": 25317 + }, + { + "epoch": 0.5572969403477878, + "grad_norm": 1.743240194019872, + "learning_rate": 4.317209204194615e-06, + "loss": 0.3008, + "step": 25318 + }, + { + "epoch": 0.5573189522342065, + "grad_norm": 1.3111096384198528, + "learning_rate": 4.316856088814324e-06, + "loss": 0.2643, + "step": 25319 + }, + { + "epoch": 0.5573409641206252, + "grad_norm": 1.8946229099322576, + "learning_rate": 4.316502976906077e-06, + "loss": 0.3049, + "step": 25320 + }, + { + "epoch": 0.5573629760070438, + "grad_norm": 1.4566387589902452, + "learning_rate": 4.3161498684716645e-06, + "loss": 0.2954, + "step": 25321 + }, + { + "epoch": 0.5573849878934625, + "grad_norm": 0.625514206822162, + "learning_rate": 4.315796763512885e-06, + "loss": 0.4571, + "step": 25322 + }, + { + "epoch": 0.5574069997798812, + "grad_norm": 1.3050843916921542, + "learning_rate": 4.315443662031531e-06, + "loss": 0.3445, + "step": 25323 + }, + { + "epoch": 0.5574290116662998, + "grad_norm": 1.46237028341723, + "learning_rate": 4.315090564029398e-06, + "loss": 0.3738, + "step": 25324 + }, + { + "epoch": 0.5574510235527185, + "grad_norm": 1.05617830628007, + "learning_rate": 4.3147374695082806e-06, + "loss": 0.2076, + "step": 25325 + }, + { + "epoch": 0.5574730354391372, + "grad_norm": 1.3910924448396285, + "learning_rate": 4.3143843784699724e-06, + "loss": 0.343, + "step": 25326 + }, + { + "epoch": 0.5574950473255558, + "grad_norm": 1.3686723397568525, + "learning_rate": 4.314031290916267e-06, + "loss": 0.2616, + "step": 25327 + }, + { + "epoch": 0.5575170592119745, + "grad_norm": 1.321197023369826, + "learning_rate": 4.313678206848962e-06, + "loss": 0.2909, + "step": 25328 + }, + { + "epoch": 0.5575390710983932, + "grad_norm": 1.6962193219035542, + "learning_rate": 4.313325126269851e-06, + "loss": 0.3327, + "step": 25329 + }, + { + "epoch": 0.5575610829848118, + "grad_norm": 1.4142531873928907, + "learning_rate": 4.312972049180728e-06, + "loss": 0.2308, + "step": 25330 + }, + { + "epoch": 0.5575830948712305, + "grad_norm": 2.185812298346207, + "learning_rate": 4.312618975583387e-06, + "loss": 0.3478, + "step": 25331 + }, + { + "epoch": 0.5576051067576492, + "grad_norm": 1.6925789681453436, + "learning_rate": 4.312265905479622e-06, + "loss": 0.3587, + "step": 25332 + }, + { + "epoch": 0.5576271186440678, + "grad_norm": 2.1229692041302735, + "learning_rate": 4.31191283887123e-06, + "loss": 0.343, + "step": 25333 + }, + { + "epoch": 0.5576491305304865, + "grad_norm": 2.588336194666438, + "learning_rate": 4.311559775760002e-06, + "loss": 0.3587, + "step": 25334 + }, + { + "epoch": 0.5576711424169051, + "grad_norm": 2.1453830248300525, + "learning_rate": 4.311206716147737e-06, + "loss": 0.2589, + "step": 25335 + }, + { + "epoch": 0.5576931543033238, + "grad_norm": 1.5218963634874394, + "learning_rate": 4.310853660036225e-06, + "loss": 0.2808, + "step": 25336 + }, + { + "epoch": 0.5577151661897425, + "grad_norm": 1.626815949236285, + "learning_rate": 4.310500607427261e-06, + "loss": 0.3657, + "step": 25337 + }, + { + "epoch": 0.5577371780761611, + "grad_norm": 1.6101595707706515, + "learning_rate": 4.310147558322643e-06, + "loss": 0.3079, + "step": 25338 + }, + { + "epoch": 0.5577591899625798, + "grad_norm": 1.6750153930059133, + "learning_rate": 4.309794512724159e-06, + "loss": 0.3413, + "step": 25339 + }, + { + "epoch": 0.5577812018489985, + "grad_norm": 1.8361163105087757, + "learning_rate": 4.309441470633611e-06, + "loss": 0.3164, + "step": 25340 + }, + { + "epoch": 0.5578032137354171, + "grad_norm": 1.4977810783610859, + "learning_rate": 4.309088432052788e-06, + "loss": 0.3768, + "step": 25341 + }, + { + "epoch": 0.5578252256218358, + "grad_norm": 2.147003253199924, + "learning_rate": 4.308735396983484e-06, + "loss": 0.4054, + "step": 25342 + }, + { + "epoch": 0.5578472375082545, + "grad_norm": 1.545857163328501, + "learning_rate": 4.308382365427499e-06, + "loss": 0.371, + "step": 25343 + }, + { + "epoch": 0.5578692493946731, + "grad_norm": 1.3079274137925292, + "learning_rate": 4.308029337386619e-06, + "loss": 0.2943, + "step": 25344 + }, + { + "epoch": 0.5578912612810918, + "grad_norm": 1.2802711818276387, + "learning_rate": 4.307676312862645e-06, + "loss": 0.3798, + "step": 25345 + }, + { + "epoch": 0.5579132731675105, + "grad_norm": 1.6176563576018361, + "learning_rate": 4.3073232918573694e-06, + "loss": 0.2972, + "step": 25346 + }, + { + "epoch": 0.5579352850539291, + "grad_norm": 1.5299464419997895, + "learning_rate": 4.3069702743725835e-06, + "loss": 0.3211, + "step": 25347 + }, + { + "epoch": 0.5579572969403478, + "grad_norm": 1.643027562071548, + "learning_rate": 4.306617260410086e-06, + "loss": 0.3425, + "step": 25348 + }, + { + "epoch": 0.5579793088267665, + "grad_norm": 0.6460591281872892, + "learning_rate": 4.306264249971668e-06, + "loss": 0.4782, + "step": 25349 + }, + { + "epoch": 0.5580013207131851, + "grad_norm": 1.2036897923174799, + "learning_rate": 4.305911243059123e-06, + "loss": 0.2614, + "step": 25350 + }, + { + "epoch": 0.5580233325996038, + "grad_norm": 1.815346893988722, + "learning_rate": 4.305558239674248e-06, + "loss": 0.2469, + "step": 25351 + }, + { + "epoch": 0.5580453444860225, + "grad_norm": 2.0506249536975107, + "learning_rate": 4.305205239818835e-06, + "loss": 0.2708, + "step": 25352 + }, + { + "epoch": 0.5580673563724411, + "grad_norm": 1.454832397146721, + "learning_rate": 4.304852243494681e-06, + "loss": 0.3753, + "step": 25353 + }, + { + "epoch": 0.5580893682588598, + "grad_norm": 1.4367981572813004, + "learning_rate": 4.3044992507035766e-06, + "loss": 0.3397, + "step": 25354 + }, + { + "epoch": 0.5581113801452785, + "grad_norm": 2.067183781719178, + "learning_rate": 4.304146261447316e-06, + "loss": 0.268, + "step": 25355 + }, + { + "epoch": 0.5581333920316971, + "grad_norm": 1.9663358327143654, + "learning_rate": 4.303793275727697e-06, + "loss": 0.3951, + "step": 25356 + }, + { + "epoch": 0.5581554039181158, + "grad_norm": 1.320587370366301, + "learning_rate": 4.303440293546508e-06, + "loss": 0.2821, + "step": 25357 + }, + { + "epoch": 0.5581774158045344, + "grad_norm": 1.4587139817676742, + "learning_rate": 4.303087314905548e-06, + "loss": 0.3461, + "step": 25358 + }, + { + "epoch": 0.5581994276909531, + "grad_norm": 1.3662030853840625, + "learning_rate": 4.30273433980661e-06, + "loss": 0.3576, + "step": 25359 + }, + { + "epoch": 0.5582214395773718, + "grad_norm": 1.3393005677075018, + "learning_rate": 4.302381368251485e-06, + "loss": 0.3137, + "step": 25360 + }, + { + "epoch": 0.5582434514637904, + "grad_norm": 1.45293989222092, + "learning_rate": 4.302028400241971e-06, + "loss": 0.248, + "step": 25361 + }, + { + "epoch": 0.5582654633502091, + "grad_norm": 1.3903819752690036, + "learning_rate": 4.301675435779857e-06, + "loss": 0.2533, + "step": 25362 + }, + { + "epoch": 0.5582874752366278, + "grad_norm": 0.6251724128545249, + "learning_rate": 4.301322474866942e-06, + "loss": 0.4649, + "step": 25363 + }, + { + "epoch": 0.5583094871230464, + "grad_norm": 0.6275374647179497, + "learning_rate": 4.300969517505019e-06, + "loss": 0.4822, + "step": 25364 + }, + { + "epoch": 0.5583314990094651, + "grad_norm": 1.5408505106183752, + "learning_rate": 4.300616563695878e-06, + "loss": 0.3159, + "step": 25365 + }, + { + "epoch": 0.5583535108958838, + "grad_norm": 1.4056689989498823, + "learning_rate": 4.300263613441318e-06, + "loss": 0.2884, + "step": 25366 + }, + { + "epoch": 0.5583755227823024, + "grad_norm": 1.6622080318307493, + "learning_rate": 4.29991066674313e-06, + "loss": 0.2957, + "step": 25367 + }, + { + "epoch": 0.5583975346687211, + "grad_norm": 2.243413297563104, + "learning_rate": 4.2995577236031065e-06, + "loss": 0.3367, + "step": 25368 + }, + { + "epoch": 0.5584195465551398, + "grad_norm": 1.4088728824607442, + "learning_rate": 4.299204784023044e-06, + "loss": 0.3293, + "step": 25369 + }, + { + "epoch": 0.5584415584415584, + "grad_norm": 1.4994571984445253, + "learning_rate": 4.298851848004734e-06, + "loss": 0.3326, + "step": 25370 + }, + { + "epoch": 0.5584635703279771, + "grad_norm": 2.092199840944173, + "learning_rate": 4.298498915549974e-06, + "loss": 0.2548, + "step": 25371 + }, + { + "epoch": 0.5584855822143958, + "grad_norm": 1.473528985423412, + "learning_rate": 4.298145986660555e-06, + "loss": 0.2142, + "step": 25372 + }, + { + "epoch": 0.5585075941008144, + "grad_norm": 1.7038757311874873, + "learning_rate": 4.29779306133827e-06, + "loss": 0.3208, + "step": 25373 + }, + { + "epoch": 0.5585296059872331, + "grad_norm": 0.5883339607916775, + "learning_rate": 4.297440139584915e-06, + "loss": 0.467, + "step": 25374 + }, + { + "epoch": 0.5585516178736518, + "grad_norm": 1.473941416958276, + "learning_rate": 4.2970872214022805e-06, + "loss": 0.3046, + "step": 25375 + }, + { + "epoch": 0.5585736297600704, + "grad_norm": 1.2546690240067209, + "learning_rate": 4.296734306792165e-06, + "loss": 0.3077, + "step": 25376 + }, + { + "epoch": 0.5585956416464891, + "grad_norm": 1.876697720057782, + "learning_rate": 4.2963813957563585e-06, + "loss": 0.2471, + "step": 25377 + }, + { + "epoch": 0.5586176535329078, + "grad_norm": 1.2150190144777566, + "learning_rate": 4.296028488296654e-06, + "loss": 0.2739, + "step": 25378 + }, + { + "epoch": 0.5586396654193264, + "grad_norm": 1.2524672705381963, + "learning_rate": 4.295675584414849e-06, + "loss": 0.2098, + "step": 25379 + }, + { + "epoch": 0.5586616773057451, + "grad_norm": 1.491685851173262, + "learning_rate": 4.295322684112732e-06, + "loss": 0.2905, + "step": 25380 + }, + { + "epoch": 0.5586836891921638, + "grad_norm": 1.433658635718352, + "learning_rate": 4.294969787392103e-06, + "loss": 0.3275, + "step": 25381 + }, + { + "epoch": 0.5587057010785824, + "grad_norm": 1.516360670461447, + "learning_rate": 4.294616894254749e-06, + "loss": 0.2158, + "step": 25382 + }, + { + "epoch": 0.5587277129650011, + "grad_norm": 1.7144779021554444, + "learning_rate": 4.294264004702467e-06, + "loss": 0.2749, + "step": 25383 + }, + { + "epoch": 0.5587497248514197, + "grad_norm": 1.6374835741122116, + "learning_rate": 4.293911118737052e-06, + "loss": 0.3264, + "step": 25384 + }, + { + "epoch": 0.5587717367378384, + "grad_norm": 1.373808804222431, + "learning_rate": 4.293558236360294e-06, + "loss": 0.3527, + "step": 25385 + }, + { + "epoch": 0.5587937486242571, + "grad_norm": 1.3962209960821645, + "learning_rate": 4.293205357573988e-06, + "loss": 0.2683, + "step": 25386 + }, + { + "epoch": 0.5588157605106757, + "grad_norm": 1.9293912933751225, + "learning_rate": 4.292852482379929e-06, + "loss": 0.3644, + "step": 25387 + }, + { + "epoch": 0.5588377723970944, + "grad_norm": 1.4377099701615303, + "learning_rate": 4.292499610779907e-06, + "loss": 0.2453, + "step": 25388 + }, + { + "epoch": 0.5588597842835131, + "grad_norm": 1.330752187486077, + "learning_rate": 4.29214674277572e-06, + "loss": 0.3122, + "step": 25389 + }, + { + "epoch": 0.5588817961699317, + "grad_norm": 1.3705019916336678, + "learning_rate": 4.291793878369158e-06, + "loss": 0.2271, + "step": 25390 + }, + { + "epoch": 0.5589038080563504, + "grad_norm": 0.6377670036908164, + "learning_rate": 4.291441017562014e-06, + "loss": 0.4654, + "step": 25391 + }, + { + "epoch": 0.5589258199427691, + "grad_norm": 1.475757293241099, + "learning_rate": 4.291088160356086e-06, + "loss": 0.3634, + "step": 25392 + }, + { + "epoch": 0.5589478318291877, + "grad_norm": 1.3011350722575619, + "learning_rate": 4.290735306753159e-06, + "loss": 0.3267, + "step": 25393 + }, + { + "epoch": 0.5589698437156064, + "grad_norm": 2.3277517096110745, + "learning_rate": 4.290382456755036e-06, + "loss": 0.3149, + "step": 25394 + }, + { + "epoch": 0.5589918556020251, + "grad_norm": 1.7245427141694027, + "learning_rate": 4.290029610363505e-06, + "loss": 0.3135, + "step": 25395 + }, + { + "epoch": 0.5590138674884437, + "grad_norm": 1.4782737542385864, + "learning_rate": 4.289676767580359e-06, + "loss": 0.284, + "step": 25396 + }, + { + "epoch": 0.5590358793748624, + "grad_norm": 1.7589412749897784, + "learning_rate": 4.289323928407394e-06, + "loss": 0.2868, + "step": 25397 + }, + { + "epoch": 0.5590578912612811, + "grad_norm": 1.4250558272459544, + "learning_rate": 4.2889710928464e-06, + "loss": 0.3031, + "step": 25398 + }, + { + "epoch": 0.5590799031476997, + "grad_norm": 1.420752514008301, + "learning_rate": 4.2886182608991726e-06, + "loss": 0.3058, + "step": 25399 + }, + { + "epoch": 0.5591019150341184, + "grad_norm": 1.4432345664234332, + "learning_rate": 4.288265432567505e-06, + "loss": 0.2964, + "step": 25400 + }, + { + "epoch": 0.5591239269205371, + "grad_norm": 1.6947301621114177, + "learning_rate": 4.287912607853189e-06, + "loss": 0.3169, + "step": 25401 + }, + { + "epoch": 0.5591459388069557, + "grad_norm": 1.3518598947131808, + "learning_rate": 4.287559786758021e-06, + "loss": 0.295, + "step": 25402 + }, + { + "epoch": 0.5591679506933744, + "grad_norm": 1.77279025994476, + "learning_rate": 4.287206969283789e-06, + "loss": 0.3995, + "step": 25403 + }, + { + "epoch": 0.5591899625797931, + "grad_norm": 0.6215545888895081, + "learning_rate": 4.28685415543229e-06, + "loss": 0.4767, + "step": 25404 + }, + { + "epoch": 0.5592119744662117, + "grad_norm": 2.347056250795696, + "learning_rate": 4.286501345205317e-06, + "loss": 0.3502, + "step": 25405 + }, + { + "epoch": 0.5592339863526304, + "grad_norm": 2.1830764260380695, + "learning_rate": 4.286148538604661e-06, + "loss": 0.3348, + "step": 25406 + }, + { + "epoch": 0.559255998239049, + "grad_norm": 1.5644714259521164, + "learning_rate": 4.285795735632119e-06, + "loss": 0.325, + "step": 25407 + }, + { + "epoch": 0.5592780101254677, + "grad_norm": 2.230963575908203, + "learning_rate": 4.28544293628948e-06, + "loss": 0.3033, + "step": 25408 + }, + { + "epoch": 0.5593000220118864, + "grad_norm": 1.4536948780500587, + "learning_rate": 4.285090140578538e-06, + "loss": 0.2605, + "step": 25409 + }, + { + "epoch": 0.559322033898305, + "grad_norm": 1.3457122925586738, + "learning_rate": 4.284737348501089e-06, + "loss": 0.2367, + "step": 25410 + }, + { + "epoch": 0.5593440457847237, + "grad_norm": 1.8311612198811709, + "learning_rate": 4.28438456005892e-06, + "loss": 0.2942, + "step": 25411 + }, + { + "epoch": 0.5593660576711424, + "grad_norm": 7.251056278129254, + "learning_rate": 4.284031775253832e-06, + "loss": 0.2692, + "step": 25412 + }, + { + "epoch": 0.559388069557561, + "grad_norm": 1.3368306656862652, + "learning_rate": 4.283678994087612e-06, + "loss": 0.2699, + "step": 25413 + }, + { + "epoch": 0.5594100814439797, + "grad_norm": 1.6335373176367591, + "learning_rate": 4.283326216562055e-06, + "loss": 0.3733, + "step": 25414 + }, + { + "epoch": 0.5594320933303984, + "grad_norm": 1.7357623131656197, + "learning_rate": 4.282973442678955e-06, + "loss": 0.3163, + "step": 25415 + }, + { + "epoch": 0.559454105216817, + "grad_norm": 1.4541614701918377, + "learning_rate": 4.282620672440101e-06, + "loss": 0.2877, + "step": 25416 + }, + { + "epoch": 0.5594761171032357, + "grad_norm": 1.3252864745344533, + "learning_rate": 4.282267905847291e-06, + "loss": 0.2298, + "step": 25417 + }, + { + "epoch": 0.5594981289896545, + "grad_norm": 0.6096006421308335, + "learning_rate": 4.281915142902315e-06, + "loss": 0.4753, + "step": 25418 + }, + { + "epoch": 0.559520140876073, + "grad_norm": 1.5259029667639212, + "learning_rate": 4.281562383606965e-06, + "loss": 0.3192, + "step": 25419 + }, + { + "epoch": 0.5595421527624918, + "grad_norm": 1.3717674757279896, + "learning_rate": 4.281209627963038e-06, + "loss": 0.2784, + "step": 25420 + }, + { + "epoch": 0.5595641646489105, + "grad_norm": 1.4506621074160153, + "learning_rate": 4.280856875972322e-06, + "loss": 0.2993, + "step": 25421 + }, + { + "epoch": 0.559586176535329, + "grad_norm": 1.3638029914665948, + "learning_rate": 4.280504127636614e-06, + "loss": 0.3003, + "step": 25422 + }, + { + "epoch": 0.5596081884217478, + "grad_norm": 1.5332786317486016, + "learning_rate": 4.280151382957704e-06, + "loss": 0.3662, + "step": 25423 + }, + { + "epoch": 0.5596302003081665, + "grad_norm": 1.8629716822685347, + "learning_rate": 4.279798641937385e-06, + "loss": 0.3456, + "step": 25424 + }, + { + "epoch": 0.559652212194585, + "grad_norm": 1.9460627469406753, + "learning_rate": 4.2794459045774526e-06, + "loss": 0.2268, + "step": 25425 + }, + { + "epoch": 0.5596742240810038, + "grad_norm": 1.4864555424992347, + "learning_rate": 4.279093170879696e-06, + "loss": 0.3382, + "step": 25426 + }, + { + "epoch": 0.5596962359674225, + "grad_norm": 2.6396584304275925, + "learning_rate": 4.2787404408459085e-06, + "loss": 0.3468, + "step": 25427 + }, + { + "epoch": 0.559718247853841, + "grad_norm": 2.2565472717223956, + "learning_rate": 4.278387714477886e-06, + "loss": 0.367, + "step": 25428 + }, + { + "epoch": 0.5597402597402598, + "grad_norm": 4.494422610888667, + "learning_rate": 4.278034991777417e-06, + "loss": 0.3415, + "step": 25429 + }, + { + "epoch": 0.5597622716266784, + "grad_norm": 1.636528548067086, + "learning_rate": 4.277682272746298e-06, + "loss": 0.2871, + "step": 25430 + }, + { + "epoch": 0.5597842835130971, + "grad_norm": 1.4741311914082362, + "learning_rate": 4.277329557386319e-06, + "loss": 0.2974, + "step": 25431 + }, + { + "epoch": 0.5598062953995158, + "grad_norm": 1.4439535960796595, + "learning_rate": 4.276976845699272e-06, + "loss": 0.3079, + "step": 25432 + }, + { + "epoch": 0.5598283072859344, + "grad_norm": 1.1969437957543139, + "learning_rate": 4.276624137686954e-06, + "loss": 0.2639, + "step": 25433 + }, + { + "epoch": 0.5598503191723531, + "grad_norm": 1.99859192292819, + "learning_rate": 4.276271433351153e-06, + "loss": 0.3743, + "step": 25434 + }, + { + "epoch": 0.5598723310587718, + "grad_norm": 1.4941439681853645, + "learning_rate": 4.275918732693664e-06, + "loss": 0.2706, + "step": 25435 + }, + { + "epoch": 0.5598943429451904, + "grad_norm": 1.9681365012523415, + "learning_rate": 4.275566035716279e-06, + "loss": 0.2898, + "step": 25436 + }, + { + "epoch": 0.5599163548316091, + "grad_norm": 1.5859172204102108, + "learning_rate": 4.275213342420789e-06, + "loss": 0.4106, + "step": 25437 + }, + { + "epoch": 0.5599383667180278, + "grad_norm": 1.6038590330929354, + "learning_rate": 4.2748606528089905e-06, + "loss": 0.2858, + "step": 25438 + }, + { + "epoch": 0.5599603786044464, + "grad_norm": 1.2496176471629106, + "learning_rate": 4.274507966882671e-06, + "loss": 0.3243, + "step": 25439 + }, + { + "epoch": 0.5599823904908651, + "grad_norm": 1.3491496104761014, + "learning_rate": 4.274155284643627e-06, + "loss": 0.3544, + "step": 25440 + }, + { + "epoch": 0.5600044023772838, + "grad_norm": 1.8457645671244003, + "learning_rate": 4.27380260609365e-06, + "loss": 0.2955, + "step": 25441 + }, + { + "epoch": 0.5600264142637024, + "grad_norm": 0.6430018147139914, + "learning_rate": 4.27344993123453e-06, + "loss": 0.444, + "step": 25442 + }, + { + "epoch": 0.5600484261501211, + "grad_norm": 1.8033249813452783, + "learning_rate": 4.273097260068065e-06, + "loss": 0.3223, + "step": 25443 + }, + { + "epoch": 0.5600704380365398, + "grad_norm": 0.6332564917172253, + "learning_rate": 4.272744592596041e-06, + "loss": 0.4948, + "step": 25444 + }, + { + "epoch": 0.5600924499229584, + "grad_norm": 1.7602953315085728, + "learning_rate": 4.272391928820254e-06, + "loss": 0.3997, + "step": 25445 + }, + { + "epoch": 0.5601144618093771, + "grad_norm": 1.4437738290075914, + "learning_rate": 4.272039268742498e-06, + "loss": 0.3143, + "step": 25446 + }, + { + "epoch": 0.5601364736957958, + "grad_norm": 1.5343965387922878, + "learning_rate": 4.271686612364559e-06, + "loss": 0.2607, + "step": 25447 + }, + { + "epoch": 0.5601584855822144, + "grad_norm": 3.4949946052027414, + "learning_rate": 4.271333959688237e-06, + "loss": 0.3787, + "step": 25448 + }, + { + "epoch": 0.5601804974686331, + "grad_norm": 2.868204949877726, + "learning_rate": 4.27098131071532e-06, + "loss": 0.3303, + "step": 25449 + }, + { + "epoch": 0.5602025093550518, + "grad_norm": 2.9327532948865693, + "learning_rate": 4.270628665447599e-06, + "loss": 0.269, + "step": 25450 + }, + { + "epoch": 0.5602245212414704, + "grad_norm": 1.4162758797388972, + "learning_rate": 4.270276023886872e-06, + "loss": 0.3061, + "step": 25451 + }, + { + "epoch": 0.5602465331278891, + "grad_norm": 1.226973967248595, + "learning_rate": 4.269923386034924e-06, + "loss": 0.209, + "step": 25452 + }, + { + "epoch": 0.5602685450143077, + "grad_norm": 1.861966849926157, + "learning_rate": 4.269570751893554e-06, + "loss": 0.3463, + "step": 25453 + }, + { + "epoch": 0.5602905569007264, + "grad_norm": 1.5161425986933803, + "learning_rate": 4.26921812146455e-06, + "loss": 0.2511, + "step": 25454 + }, + { + "epoch": 0.5603125687871451, + "grad_norm": 4.263272081801264, + "learning_rate": 4.268865494749704e-06, + "loss": 0.2947, + "step": 25455 + }, + { + "epoch": 0.5603345806735637, + "grad_norm": 4.626094982658681, + "learning_rate": 4.268512871750812e-06, + "loss": 0.3615, + "step": 25456 + }, + { + "epoch": 0.5603565925599824, + "grad_norm": 0.666697252119046, + "learning_rate": 4.268160252469662e-06, + "loss": 0.5033, + "step": 25457 + }, + { + "epoch": 0.5603786044464011, + "grad_norm": 1.7363383766012024, + "learning_rate": 4.267807636908049e-06, + "loss": 0.3054, + "step": 25458 + }, + { + "epoch": 0.5604006163328197, + "grad_norm": 1.2199414809922597, + "learning_rate": 4.267455025067764e-06, + "loss": 0.2721, + "step": 25459 + }, + { + "epoch": 0.5604226282192384, + "grad_norm": 1.5187841693555806, + "learning_rate": 4.2671024169505985e-06, + "loss": 0.3136, + "step": 25460 + }, + { + "epoch": 0.5604446401056571, + "grad_norm": 0.6074866317343246, + "learning_rate": 4.266749812558348e-06, + "loss": 0.4438, + "step": 25461 + }, + { + "epoch": 0.5604666519920757, + "grad_norm": 1.3659360111395666, + "learning_rate": 4.2663972118927995e-06, + "loss": 0.3512, + "step": 25462 + }, + { + "epoch": 0.5604886638784944, + "grad_norm": 1.944956923047252, + "learning_rate": 4.2660446149557485e-06, + "loss": 0.3008, + "step": 25463 + }, + { + "epoch": 0.5605106757649131, + "grad_norm": 1.46767071115766, + "learning_rate": 4.265692021748987e-06, + "loss": 0.3072, + "step": 25464 + }, + { + "epoch": 0.5605326876513317, + "grad_norm": 1.4340668678827093, + "learning_rate": 4.265339432274303e-06, + "loss": 0.3426, + "step": 25465 + }, + { + "epoch": 0.5605546995377504, + "grad_norm": 1.7308811802666837, + "learning_rate": 4.264986846533495e-06, + "loss": 0.2627, + "step": 25466 + }, + { + "epoch": 0.5605767114241691, + "grad_norm": 6.88032878102385, + "learning_rate": 4.26463426452835e-06, + "loss": 0.2634, + "step": 25467 + }, + { + "epoch": 0.5605987233105877, + "grad_norm": 1.5732074543639132, + "learning_rate": 4.2642816862606615e-06, + "loss": 0.306, + "step": 25468 + }, + { + "epoch": 0.5606207351970064, + "grad_norm": 1.4396508185537276, + "learning_rate": 4.263929111732224e-06, + "loss": 0.2638, + "step": 25469 + }, + { + "epoch": 0.5606427470834251, + "grad_norm": 1.6779011158553454, + "learning_rate": 4.263576540944824e-06, + "loss": 0.2396, + "step": 25470 + }, + { + "epoch": 0.5606647589698437, + "grad_norm": 2.3049362749824143, + "learning_rate": 4.263223973900258e-06, + "loss": 0.3578, + "step": 25471 + }, + { + "epoch": 0.5606867708562624, + "grad_norm": 1.1381799090613811, + "learning_rate": 4.262871410600316e-06, + "loss": 0.2312, + "step": 25472 + }, + { + "epoch": 0.5607087827426811, + "grad_norm": 1.2982726620947442, + "learning_rate": 4.262518851046789e-06, + "loss": 0.3247, + "step": 25473 + }, + { + "epoch": 0.5607307946290997, + "grad_norm": 1.4811600786754384, + "learning_rate": 4.262166295241473e-06, + "loss": 0.3055, + "step": 25474 + }, + { + "epoch": 0.5607528065155184, + "grad_norm": 1.4600035248252703, + "learning_rate": 4.261813743186154e-06, + "loss": 0.3184, + "step": 25475 + }, + { + "epoch": 0.560774818401937, + "grad_norm": 2.794938754769172, + "learning_rate": 4.261461194882629e-06, + "loss": 0.2669, + "step": 25476 + }, + { + "epoch": 0.5607968302883557, + "grad_norm": 1.6027111778222232, + "learning_rate": 4.261108650332688e-06, + "loss": 0.3102, + "step": 25477 + }, + { + "epoch": 0.5608188421747744, + "grad_norm": 1.8677908317008167, + "learning_rate": 4.260756109538119e-06, + "loss": 0.3859, + "step": 25478 + }, + { + "epoch": 0.560840854061193, + "grad_norm": 1.7227407642491168, + "learning_rate": 4.260403572500721e-06, + "loss": 0.4535, + "step": 25479 + }, + { + "epoch": 0.5608628659476117, + "grad_norm": 1.2851653779041845, + "learning_rate": 4.260051039222279e-06, + "loss": 0.2633, + "step": 25480 + }, + { + "epoch": 0.5608848778340304, + "grad_norm": 1.3273753515761868, + "learning_rate": 4.259698509704589e-06, + "loss": 0.2382, + "step": 25481 + }, + { + "epoch": 0.560906889720449, + "grad_norm": 1.115838199140346, + "learning_rate": 4.259345983949443e-06, + "loss": 0.2716, + "step": 25482 + }, + { + "epoch": 0.5609289016068677, + "grad_norm": 0.6357304110731946, + "learning_rate": 4.258993461958627e-06, + "loss": 0.493, + "step": 25483 + }, + { + "epoch": 0.5609509134932864, + "grad_norm": 1.8272628332786776, + "learning_rate": 4.258640943733941e-06, + "loss": 0.3651, + "step": 25484 + }, + { + "epoch": 0.560972925379705, + "grad_norm": 1.4765758152419848, + "learning_rate": 4.2582884292771685e-06, + "loss": 0.2532, + "step": 25485 + }, + { + "epoch": 0.5609949372661237, + "grad_norm": 0.5876003041840956, + "learning_rate": 4.257935918590108e-06, + "loss": 0.4512, + "step": 25486 + }, + { + "epoch": 0.5610169491525424, + "grad_norm": 1.1705910599474256, + "learning_rate": 4.257583411674548e-06, + "loss": 0.2737, + "step": 25487 + }, + { + "epoch": 0.561038961038961, + "grad_norm": 1.469209515779545, + "learning_rate": 4.2572309085322784e-06, + "loss": 0.347, + "step": 25488 + }, + { + "epoch": 0.5610609729253797, + "grad_norm": 0.6335889362098946, + "learning_rate": 4.256878409165094e-06, + "loss": 0.4893, + "step": 25489 + }, + { + "epoch": 0.5610829848117984, + "grad_norm": 0.6138399331282571, + "learning_rate": 4.256525913574785e-06, + "loss": 0.4642, + "step": 25490 + }, + { + "epoch": 0.561104996698217, + "grad_norm": 1.9078356784594765, + "learning_rate": 4.25617342176314e-06, + "loss": 0.3325, + "step": 25491 + }, + { + "epoch": 0.5611270085846357, + "grad_norm": 1.32749201519141, + "learning_rate": 4.255820933731958e-06, + "loss": 0.3462, + "step": 25492 + }, + { + "epoch": 0.5611490204710544, + "grad_norm": 1.432813483746848, + "learning_rate": 4.255468449483023e-06, + "loss": 0.3016, + "step": 25493 + }, + { + "epoch": 0.561171032357473, + "grad_norm": 1.526213794040723, + "learning_rate": 4.25511596901813e-06, + "loss": 0.296, + "step": 25494 + }, + { + "epoch": 0.5611930442438917, + "grad_norm": 2.077095326834451, + "learning_rate": 4.254763492339071e-06, + "loss": 0.2439, + "step": 25495 + }, + { + "epoch": 0.5612150561303104, + "grad_norm": 1.651984465873784, + "learning_rate": 4.254411019447634e-06, + "loss": 0.2772, + "step": 25496 + }, + { + "epoch": 0.561237068016729, + "grad_norm": 0.6348590152964128, + "learning_rate": 4.254058550345615e-06, + "loss": 0.4928, + "step": 25497 + }, + { + "epoch": 0.5612590799031477, + "grad_norm": 2.425799050424701, + "learning_rate": 4.253706085034801e-06, + "loss": 0.2682, + "step": 25498 + }, + { + "epoch": 0.5612810917895663, + "grad_norm": 1.4873585092168486, + "learning_rate": 4.253353623516987e-06, + "loss": 0.2416, + "step": 25499 + }, + { + "epoch": 0.561303103675985, + "grad_norm": 0.6180565481108861, + "learning_rate": 4.253001165793964e-06, + "loss": 0.4516, + "step": 25500 + }, + { + "epoch": 0.5613251155624037, + "grad_norm": 1.2371147091189436, + "learning_rate": 4.252648711867519e-06, + "loss": 0.2478, + "step": 25501 + }, + { + "epoch": 0.5613471274488223, + "grad_norm": 1.2531468875118421, + "learning_rate": 4.252296261739449e-06, + "loss": 0.2825, + "step": 25502 + }, + { + "epoch": 0.561369139335241, + "grad_norm": 2.190384377599222, + "learning_rate": 4.251943815411542e-06, + "loss": 0.3632, + "step": 25503 + }, + { + "epoch": 0.5613911512216597, + "grad_norm": 2.2519345038507974, + "learning_rate": 4.251591372885591e-06, + "loss": 0.2779, + "step": 25504 + }, + { + "epoch": 0.5614131631080783, + "grad_norm": 1.4016794162595851, + "learning_rate": 4.251238934163388e-06, + "loss": 0.327, + "step": 25505 + }, + { + "epoch": 0.561435174994497, + "grad_norm": 1.3489813605110452, + "learning_rate": 4.250886499246719e-06, + "loss": 0.2994, + "step": 25506 + }, + { + "epoch": 0.5614571868809157, + "grad_norm": 1.4511350164369892, + "learning_rate": 4.250534068137381e-06, + "loss": 0.2448, + "step": 25507 + }, + { + "epoch": 0.5614791987673343, + "grad_norm": 0.6307277274583963, + "learning_rate": 4.250181640837162e-06, + "loss": 0.457, + "step": 25508 + }, + { + "epoch": 0.561501210653753, + "grad_norm": 1.5162254405699915, + "learning_rate": 4.249829217347855e-06, + "loss": 0.3692, + "step": 25509 + }, + { + "epoch": 0.5615232225401717, + "grad_norm": 1.367170669848866, + "learning_rate": 4.249476797671252e-06, + "loss": 0.3158, + "step": 25510 + }, + { + "epoch": 0.5615452344265903, + "grad_norm": 1.5762979224320026, + "learning_rate": 4.24912438180914e-06, + "loss": 0.2086, + "step": 25511 + }, + { + "epoch": 0.561567246313009, + "grad_norm": 1.395693083333549, + "learning_rate": 4.248771969763315e-06, + "loss": 0.2823, + "step": 25512 + }, + { + "epoch": 0.5615892581994277, + "grad_norm": 1.3396904548882829, + "learning_rate": 4.248419561535565e-06, + "loss": 0.296, + "step": 25513 + }, + { + "epoch": 0.5616112700858463, + "grad_norm": 1.7106062974690641, + "learning_rate": 4.24806715712768e-06, + "loss": 0.3517, + "step": 25514 + }, + { + "epoch": 0.561633281972265, + "grad_norm": 1.8909866945240696, + "learning_rate": 4.247714756541457e-06, + "loss": 0.2966, + "step": 25515 + }, + { + "epoch": 0.5616552938586837, + "grad_norm": 2.265589390538561, + "learning_rate": 4.24736235977868e-06, + "loss": 0.3675, + "step": 25516 + }, + { + "epoch": 0.5616773057451023, + "grad_norm": 1.5653080710916847, + "learning_rate": 4.2470099668411445e-06, + "loss": 0.3214, + "step": 25517 + }, + { + "epoch": 0.561699317631521, + "grad_norm": 1.3617521152290921, + "learning_rate": 4.246657577730641e-06, + "loss": 0.2858, + "step": 25518 + }, + { + "epoch": 0.5617213295179397, + "grad_norm": 1.615328850218886, + "learning_rate": 4.246305192448957e-06, + "loss": 0.324, + "step": 25519 + }, + { + "epoch": 0.5617433414043583, + "grad_norm": 1.4081399696966603, + "learning_rate": 4.245952810997889e-06, + "loss": 0.3067, + "step": 25520 + }, + { + "epoch": 0.561765353290777, + "grad_norm": 1.5849668571567086, + "learning_rate": 4.245600433379223e-06, + "loss": 0.3678, + "step": 25521 + }, + { + "epoch": 0.5617873651771957, + "grad_norm": 1.8693442105528226, + "learning_rate": 4.245248059594753e-06, + "loss": 0.325, + "step": 25522 + }, + { + "epoch": 0.5618093770636143, + "grad_norm": 2.112807192717795, + "learning_rate": 4.244895689646271e-06, + "loss": 0.3214, + "step": 25523 + }, + { + "epoch": 0.561831388950033, + "grad_norm": 1.6233127793107665, + "learning_rate": 4.244543323535562e-06, + "loss": 0.3116, + "step": 25524 + }, + { + "epoch": 0.5618534008364516, + "grad_norm": 0.6066946294198093, + "learning_rate": 4.244190961264424e-06, + "loss": 0.4566, + "step": 25525 + }, + { + "epoch": 0.5618754127228703, + "grad_norm": 2.012849836533378, + "learning_rate": 4.243838602834643e-06, + "loss": 0.3695, + "step": 25526 + }, + { + "epoch": 0.561897424609289, + "grad_norm": 1.3468342755350626, + "learning_rate": 4.2434862482480115e-06, + "loss": 0.2304, + "step": 25527 + }, + { + "epoch": 0.5619194364957076, + "grad_norm": 1.3567198979142467, + "learning_rate": 4.243133897506321e-06, + "loss": 0.2488, + "step": 25528 + }, + { + "epoch": 0.5619414483821263, + "grad_norm": 0.5906037784677137, + "learning_rate": 4.242781550611361e-06, + "loss": 0.4539, + "step": 25529 + }, + { + "epoch": 0.561963460268545, + "grad_norm": 1.6240679238110638, + "learning_rate": 4.2424292075649235e-06, + "loss": 0.2252, + "step": 25530 + }, + { + "epoch": 0.5619854721549636, + "grad_norm": 3.5710887819313015, + "learning_rate": 4.242076868368799e-06, + "loss": 0.3497, + "step": 25531 + }, + { + "epoch": 0.5620074840413823, + "grad_norm": 1.4696969894476926, + "learning_rate": 4.241724533024775e-06, + "loss": 0.2627, + "step": 25532 + }, + { + "epoch": 0.562029495927801, + "grad_norm": 1.6256177088664152, + "learning_rate": 4.241372201534649e-06, + "loss": 0.3532, + "step": 25533 + }, + { + "epoch": 0.5620515078142196, + "grad_norm": 1.4483772724401298, + "learning_rate": 4.241019873900205e-06, + "loss": 0.4147, + "step": 25534 + }, + { + "epoch": 0.5620735197006383, + "grad_norm": 1.4795410792785024, + "learning_rate": 4.240667550123238e-06, + "loss": 0.3175, + "step": 25535 + }, + { + "epoch": 0.562095531587057, + "grad_norm": 0.6148797173775785, + "learning_rate": 4.240315230205539e-06, + "loss": 0.4664, + "step": 25536 + }, + { + "epoch": 0.5621175434734756, + "grad_norm": 1.5881840094668038, + "learning_rate": 4.2399629141488915e-06, + "loss": 0.2938, + "step": 25537 + }, + { + "epoch": 0.5621395553598943, + "grad_norm": 1.5956850882485212, + "learning_rate": 4.2396106019550956e-06, + "loss": 0.2727, + "step": 25538 + }, + { + "epoch": 0.562161567246313, + "grad_norm": 1.3814811187850937, + "learning_rate": 4.239258293625937e-06, + "loss": 0.3195, + "step": 25539 + }, + { + "epoch": 0.5621835791327316, + "grad_norm": 0.6389902121294476, + "learning_rate": 4.238905989163206e-06, + "loss": 0.4541, + "step": 25540 + }, + { + "epoch": 0.5622055910191504, + "grad_norm": 1.7290252820197105, + "learning_rate": 4.238553688568697e-06, + "loss": 0.3233, + "step": 25541 + }, + { + "epoch": 0.5622276029055691, + "grad_norm": 2.2632751059926934, + "learning_rate": 4.2382013918441945e-06, + "loss": 0.3247, + "step": 25542 + }, + { + "epoch": 0.5622496147919877, + "grad_norm": 1.7202743789599957, + "learning_rate": 4.237849098991494e-06, + "loss": 0.3518, + "step": 25543 + }, + { + "epoch": 0.5622716266784064, + "grad_norm": 1.6619272346686964, + "learning_rate": 4.2374968100123825e-06, + "loss": 0.4257, + "step": 25544 + }, + { + "epoch": 0.5622936385648251, + "grad_norm": 8.921116293751053, + "learning_rate": 4.237144524908654e-06, + "loss": 0.2236, + "step": 25545 + }, + { + "epoch": 0.5623156504512437, + "grad_norm": 2.0643576569320685, + "learning_rate": 4.2367922436820984e-06, + "loss": 0.3037, + "step": 25546 + }, + { + "epoch": 0.5623376623376624, + "grad_norm": 1.675224614429267, + "learning_rate": 4.236439966334503e-06, + "loss": 0.3066, + "step": 25547 + }, + { + "epoch": 0.562359674224081, + "grad_norm": 1.628397649752135, + "learning_rate": 4.2360876928676605e-06, + "loss": 0.4383, + "step": 25548 + }, + { + "epoch": 0.5623816861104997, + "grad_norm": 1.5354060358364956, + "learning_rate": 4.235735423283363e-06, + "loss": 0.3716, + "step": 25549 + }, + { + "epoch": 0.5624036979969184, + "grad_norm": 1.8318168779472805, + "learning_rate": 4.2353831575833956e-06, + "loss": 0.3154, + "step": 25550 + }, + { + "epoch": 0.562425709883337, + "grad_norm": 1.11508713391572, + "learning_rate": 4.235030895769555e-06, + "loss": 0.2976, + "step": 25551 + }, + { + "epoch": 0.5624477217697557, + "grad_norm": 1.5793748914410657, + "learning_rate": 4.234678637843627e-06, + "loss": 0.2522, + "step": 25552 + }, + { + "epoch": 0.5624697336561744, + "grad_norm": 1.233795165706342, + "learning_rate": 4.234326383807404e-06, + "loss": 0.2807, + "step": 25553 + }, + { + "epoch": 0.562491745542593, + "grad_norm": 3.2254677021304303, + "learning_rate": 4.233974133662676e-06, + "loss": 0.3157, + "step": 25554 + }, + { + "epoch": 0.5625137574290117, + "grad_norm": 1.4333575494348676, + "learning_rate": 4.233621887411231e-06, + "loss": 0.3296, + "step": 25555 + }, + { + "epoch": 0.5625357693154304, + "grad_norm": 1.3901366188892035, + "learning_rate": 4.2332696450548645e-06, + "loss": 0.248, + "step": 25556 + }, + { + "epoch": 0.562557781201849, + "grad_norm": 1.8816990590059361, + "learning_rate": 4.232917406595361e-06, + "loss": 0.3607, + "step": 25557 + }, + { + "epoch": 0.5625797930882677, + "grad_norm": 1.6381801918710424, + "learning_rate": 4.232565172034515e-06, + "loss": 0.3241, + "step": 25558 + }, + { + "epoch": 0.5626018049746864, + "grad_norm": 1.6386149568005206, + "learning_rate": 4.232212941374117e-06, + "loss": 0.221, + "step": 25559 + }, + { + "epoch": 0.562623816861105, + "grad_norm": 1.3401215328573919, + "learning_rate": 4.231860714615952e-06, + "loss": 0.3462, + "step": 25560 + }, + { + "epoch": 0.5626458287475237, + "grad_norm": 1.525514043505792, + "learning_rate": 4.231508491761814e-06, + "loss": 0.3583, + "step": 25561 + }, + { + "epoch": 0.5626678406339424, + "grad_norm": 2.4129229839644046, + "learning_rate": 4.231156272813492e-06, + "loss": 0.3732, + "step": 25562 + }, + { + "epoch": 0.562689852520361, + "grad_norm": 0.6369515190337222, + "learning_rate": 4.230804057772779e-06, + "loss": 0.4564, + "step": 25563 + }, + { + "epoch": 0.5627118644067797, + "grad_norm": 2.105705424334278, + "learning_rate": 4.230451846641463e-06, + "loss": 0.4005, + "step": 25564 + }, + { + "epoch": 0.5627338762931984, + "grad_norm": 2.18274315965033, + "learning_rate": 4.2300996394213316e-06, + "loss": 0.3564, + "step": 25565 + }, + { + "epoch": 0.562755888179617, + "grad_norm": 0.622662098904963, + "learning_rate": 4.2297474361141775e-06, + "loss": 0.4648, + "step": 25566 + }, + { + "epoch": 0.5627779000660357, + "grad_norm": 7.137738024988368, + "learning_rate": 4.229395236721793e-06, + "loss": 0.3344, + "step": 25567 + }, + { + "epoch": 0.5627999119524544, + "grad_norm": 0.613634020757034, + "learning_rate": 4.229043041245962e-06, + "loss": 0.4725, + "step": 25568 + }, + { + "epoch": 0.562821923838873, + "grad_norm": 6.9282504238973335, + "learning_rate": 4.22869084968848e-06, + "loss": 0.3186, + "step": 25569 + }, + { + "epoch": 0.5628439357252917, + "grad_norm": 0.6299230904718599, + "learning_rate": 4.228338662051135e-06, + "loss": 0.4597, + "step": 25570 + }, + { + "epoch": 0.5628659476117103, + "grad_norm": 1.3523929452844123, + "learning_rate": 4.2279864783357176e-06, + "loss": 0.3232, + "step": 25571 + }, + { + "epoch": 0.562887959498129, + "grad_norm": 1.3742391759026116, + "learning_rate": 4.227634298544018e-06, + "loss": 0.3152, + "step": 25572 + }, + { + "epoch": 0.5629099713845477, + "grad_norm": 2.061643657580466, + "learning_rate": 4.2272821226778215e-06, + "loss": 0.2868, + "step": 25573 + }, + { + "epoch": 0.5629319832709663, + "grad_norm": 1.180243878472679, + "learning_rate": 4.226929950738926e-06, + "loss": 0.2947, + "step": 25574 + }, + { + "epoch": 0.562953995157385, + "grad_norm": 1.4292888335458596, + "learning_rate": 4.2265777827291145e-06, + "loss": 0.2619, + "step": 25575 + }, + { + "epoch": 0.5629760070438037, + "grad_norm": 1.3209594726159686, + "learning_rate": 4.2262256186501815e-06, + "loss": 0.2973, + "step": 25576 + }, + { + "epoch": 0.5629980189302223, + "grad_norm": 2.1724900905085494, + "learning_rate": 4.2258734585039165e-06, + "loss": 0.2599, + "step": 25577 + }, + { + "epoch": 0.563020030816641, + "grad_norm": 1.8527618209811727, + "learning_rate": 4.225521302292104e-06, + "loss": 0.2895, + "step": 25578 + }, + { + "epoch": 0.5630420427030597, + "grad_norm": 1.7640532420235933, + "learning_rate": 4.225169150016541e-06, + "loss": 0.2238, + "step": 25579 + }, + { + "epoch": 0.5630640545894783, + "grad_norm": 1.644560914990655, + "learning_rate": 4.224817001679011e-06, + "loss": 0.3817, + "step": 25580 + }, + { + "epoch": 0.563086066475897, + "grad_norm": 2.5804004019537765, + "learning_rate": 4.224464857281309e-06, + "loss": 0.255, + "step": 25581 + }, + { + "epoch": 0.5631080783623157, + "grad_norm": 1.9277632724952487, + "learning_rate": 4.224112716825223e-06, + "loss": 0.2207, + "step": 25582 + }, + { + "epoch": 0.5631300902487343, + "grad_norm": 2.1406992173833372, + "learning_rate": 4.223760580312539e-06, + "loss": 0.3234, + "step": 25583 + }, + { + "epoch": 0.563152102135153, + "grad_norm": 1.3210651018070483, + "learning_rate": 4.223408447745053e-06, + "loss": 0.3022, + "step": 25584 + }, + { + "epoch": 0.5631741140215717, + "grad_norm": 0.6388382190443317, + "learning_rate": 4.2230563191245485e-06, + "loss": 0.485, + "step": 25585 + }, + { + "epoch": 0.5631961259079903, + "grad_norm": 0.6335021618596625, + "learning_rate": 4.222704194452821e-06, + "loss": 0.4965, + "step": 25586 + }, + { + "epoch": 0.563218137794409, + "grad_norm": 0.6482197341017825, + "learning_rate": 4.222352073731658e-06, + "loss": 0.4915, + "step": 25587 + }, + { + "epoch": 0.5632401496808277, + "grad_norm": 1.5427320419852342, + "learning_rate": 4.221999956962847e-06, + "loss": 0.3493, + "step": 25588 + }, + { + "epoch": 0.5632621615672463, + "grad_norm": 1.3881942838133023, + "learning_rate": 4.22164784414818e-06, + "loss": 0.3037, + "step": 25589 + }, + { + "epoch": 0.563284173453665, + "grad_norm": 0.6147707792165418, + "learning_rate": 4.2212957352894464e-06, + "loss": 0.4541, + "step": 25590 + }, + { + "epoch": 0.5633061853400837, + "grad_norm": 1.5287811361731882, + "learning_rate": 4.220943630388432e-06, + "loss": 0.3026, + "step": 25591 + }, + { + "epoch": 0.5633281972265023, + "grad_norm": 1.6264802604964637, + "learning_rate": 4.2205915294469335e-06, + "loss": 0.283, + "step": 25592 + }, + { + "epoch": 0.563350209112921, + "grad_norm": 1.4516228854552442, + "learning_rate": 4.2202394324667335e-06, + "loss": 0.3585, + "step": 25593 + }, + { + "epoch": 0.5633722209993396, + "grad_norm": 1.5839819847128063, + "learning_rate": 4.219887339449626e-06, + "loss": 0.2713, + "step": 25594 + }, + { + "epoch": 0.5633942328857583, + "grad_norm": 1.3706818264641423, + "learning_rate": 4.2195352503974005e-06, + "loss": 0.2468, + "step": 25595 + }, + { + "epoch": 0.563416244772177, + "grad_norm": 1.446196932158685, + "learning_rate": 4.219183165311842e-06, + "loss": 0.3078, + "step": 25596 + }, + { + "epoch": 0.5634382566585956, + "grad_norm": 1.2989088907708501, + "learning_rate": 4.2188310841947446e-06, + "loss": 0.2828, + "step": 25597 + }, + { + "epoch": 0.5634602685450143, + "grad_norm": 1.3769039449998295, + "learning_rate": 4.218479007047894e-06, + "loss": 0.2591, + "step": 25598 + }, + { + "epoch": 0.563482280431433, + "grad_norm": 1.278949049500536, + "learning_rate": 4.218126933873083e-06, + "loss": 0.3051, + "step": 25599 + }, + { + "epoch": 0.5635042923178516, + "grad_norm": 1.4036981289141481, + "learning_rate": 4.217774864672102e-06, + "loss": 0.3389, + "step": 25600 + }, + { + "epoch": 0.5635263042042703, + "grad_norm": 2.021616885515198, + "learning_rate": 4.217422799446734e-06, + "loss": 0.3935, + "step": 25601 + }, + { + "epoch": 0.563548316090689, + "grad_norm": 4.001830203823014, + "learning_rate": 4.217070738198775e-06, + "loss": 0.2333, + "step": 25602 + }, + { + "epoch": 0.5635703279771076, + "grad_norm": 1.6095302723227036, + "learning_rate": 4.21671868093001e-06, + "loss": 0.3676, + "step": 25603 + }, + { + "epoch": 0.5635923398635263, + "grad_norm": 1.8746690414256661, + "learning_rate": 4.216366627642231e-06, + "loss": 0.2611, + "step": 25604 + }, + { + "epoch": 0.563614351749945, + "grad_norm": 0.6473737895947206, + "learning_rate": 4.216014578337228e-06, + "loss": 0.444, + "step": 25605 + }, + { + "epoch": 0.5636363636363636, + "grad_norm": 1.5283021940538042, + "learning_rate": 4.215662533016785e-06, + "loss": 0.3745, + "step": 25606 + }, + { + "epoch": 0.5636583755227823, + "grad_norm": 2.0751836284462053, + "learning_rate": 4.215310491682697e-06, + "loss": 0.2259, + "step": 25607 + }, + { + "epoch": 0.563680387409201, + "grad_norm": 2.8465512821246524, + "learning_rate": 4.214958454336752e-06, + "loss": 0.2639, + "step": 25608 + }, + { + "epoch": 0.5637023992956196, + "grad_norm": 1.3749217968674177, + "learning_rate": 4.2146064209807365e-06, + "loss": 0.3289, + "step": 25609 + }, + { + "epoch": 0.5637244111820383, + "grad_norm": 1.378537813180798, + "learning_rate": 4.214254391616442e-06, + "loss": 0.2253, + "step": 25610 + }, + { + "epoch": 0.563746423068457, + "grad_norm": 1.4814752498571406, + "learning_rate": 4.213902366245656e-06, + "loss": 0.3187, + "step": 25611 + }, + { + "epoch": 0.5637684349548756, + "grad_norm": 1.473606995335963, + "learning_rate": 4.2135503448701706e-06, + "loss": 0.291, + "step": 25612 + }, + { + "epoch": 0.5637904468412943, + "grad_norm": 4.064035612054369, + "learning_rate": 4.213198327491773e-06, + "loss": 0.2937, + "step": 25613 + }, + { + "epoch": 0.563812458727713, + "grad_norm": 1.3281383309500925, + "learning_rate": 4.212846314112251e-06, + "loss": 0.2624, + "step": 25614 + }, + { + "epoch": 0.5638344706141316, + "grad_norm": 0.648630304157236, + "learning_rate": 4.212494304733396e-06, + "loss": 0.4738, + "step": 25615 + }, + { + "epoch": 0.5638564825005503, + "grad_norm": 1.8353139967239418, + "learning_rate": 4.212142299356996e-06, + "loss": 0.4166, + "step": 25616 + }, + { + "epoch": 0.5638784943869689, + "grad_norm": 0.6004363536770119, + "learning_rate": 4.21179029798484e-06, + "loss": 0.4714, + "step": 25617 + }, + { + "epoch": 0.5639005062733876, + "grad_norm": 1.629778769679219, + "learning_rate": 4.211438300618719e-06, + "loss": 0.3117, + "step": 25618 + }, + { + "epoch": 0.5639225181598063, + "grad_norm": 1.3803029228068675, + "learning_rate": 4.211086307260418e-06, + "loss": 0.395, + "step": 25619 + }, + { + "epoch": 0.5639445300462249, + "grad_norm": 1.3006682296446048, + "learning_rate": 4.210734317911729e-06, + "loss": 0.2529, + "step": 25620 + }, + { + "epoch": 0.5639665419326436, + "grad_norm": 1.8267034801128994, + "learning_rate": 4.210382332574439e-06, + "loss": 0.2571, + "step": 25621 + }, + { + "epoch": 0.5639885538190623, + "grad_norm": 1.4014943310369445, + "learning_rate": 4.210030351250341e-06, + "loss": 0.3589, + "step": 25622 + }, + { + "epoch": 0.5640105657054809, + "grad_norm": 1.5517963031111555, + "learning_rate": 4.20967837394122e-06, + "loss": 0.3683, + "step": 25623 + }, + { + "epoch": 0.5640325775918996, + "grad_norm": 6.985520699234765, + "learning_rate": 4.209326400648865e-06, + "loss": 0.3026, + "step": 25624 + }, + { + "epoch": 0.5640545894783183, + "grad_norm": 3.082383117249368, + "learning_rate": 4.208974431375067e-06, + "loss": 0.2833, + "step": 25625 + }, + { + "epoch": 0.5640766013647369, + "grad_norm": 1.3866791326715755, + "learning_rate": 4.208622466121614e-06, + "loss": 0.2933, + "step": 25626 + }, + { + "epoch": 0.5640986132511556, + "grad_norm": 1.352841554811307, + "learning_rate": 4.208270504890292e-06, + "loss": 0.3132, + "step": 25627 + }, + { + "epoch": 0.5641206251375743, + "grad_norm": 0.6301633342124625, + "learning_rate": 4.207918547682895e-06, + "loss": 0.4646, + "step": 25628 + }, + { + "epoch": 0.5641426370239929, + "grad_norm": 67.93172994486798, + "learning_rate": 4.207566594501208e-06, + "loss": 0.3461, + "step": 25629 + }, + { + "epoch": 0.5641646489104116, + "grad_norm": 0.6265060919340278, + "learning_rate": 4.2072146453470216e-06, + "loss": 0.512, + "step": 25630 + }, + { + "epoch": 0.5641866607968303, + "grad_norm": 1.610919534458696, + "learning_rate": 4.206862700222125e-06, + "loss": 0.2272, + "step": 25631 + }, + { + "epoch": 0.5642086726832489, + "grad_norm": 2.0979246283275454, + "learning_rate": 4.2065107591283035e-06, + "loss": 0.2508, + "step": 25632 + }, + { + "epoch": 0.5642306845696676, + "grad_norm": 1.2476118069241053, + "learning_rate": 4.20615882206735e-06, + "loss": 0.3204, + "step": 25633 + }, + { + "epoch": 0.5642526964560863, + "grad_norm": 2.173856229869115, + "learning_rate": 4.20580688904105e-06, + "loss": 0.3017, + "step": 25634 + }, + { + "epoch": 0.5642747083425049, + "grad_norm": 3.377480785045392, + "learning_rate": 4.2054549600511946e-06, + "loss": 0.2906, + "step": 25635 + }, + { + "epoch": 0.5642967202289236, + "grad_norm": 1.568218971943548, + "learning_rate": 4.205103035099573e-06, + "loss": 0.2568, + "step": 25636 + }, + { + "epoch": 0.5643187321153423, + "grad_norm": 2.3488950298966014, + "learning_rate": 4.20475111418797e-06, + "loss": 0.3137, + "step": 25637 + }, + { + "epoch": 0.5643407440017609, + "grad_norm": 1.5734075764566975, + "learning_rate": 4.204399197318177e-06, + "loss": 0.2885, + "step": 25638 + }, + { + "epoch": 0.5643627558881796, + "grad_norm": 1.180203473687259, + "learning_rate": 4.204047284491981e-06, + "loss": 0.2488, + "step": 25639 + }, + { + "epoch": 0.5643847677745982, + "grad_norm": 1.6530223319052195, + "learning_rate": 4.203695375711173e-06, + "loss": 0.2966, + "step": 25640 + }, + { + "epoch": 0.5644067796610169, + "grad_norm": 3.616413851904271, + "learning_rate": 4.2033434709775425e-06, + "loss": 0.2963, + "step": 25641 + }, + { + "epoch": 0.5644287915474356, + "grad_norm": 1.6875738227149069, + "learning_rate": 4.202991570292872e-06, + "loss": 0.384, + "step": 25642 + }, + { + "epoch": 0.5644508034338542, + "grad_norm": 1.5232156745124477, + "learning_rate": 4.202639673658956e-06, + "loss": 0.3255, + "step": 25643 + }, + { + "epoch": 0.5644728153202729, + "grad_norm": 1.840055688178346, + "learning_rate": 4.202287781077579e-06, + "loss": 0.2484, + "step": 25644 + }, + { + "epoch": 0.5644948272066916, + "grad_norm": 1.4596396455392378, + "learning_rate": 4.201935892550534e-06, + "loss": 0.2541, + "step": 25645 + }, + { + "epoch": 0.5645168390931102, + "grad_norm": 1.4108721241889168, + "learning_rate": 4.201584008079605e-06, + "loss": 0.3342, + "step": 25646 + }, + { + "epoch": 0.5645388509795289, + "grad_norm": 0.6351981933646317, + "learning_rate": 4.201232127666582e-06, + "loss": 0.4999, + "step": 25647 + }, + { + "epoch": 0.5645608628659476, + "grad_norm": 1.4373593966074654, + "learning_rate": 4.200880251313254e-06, + "loss": 0.313, + "step": 25648 + }, + { + "epoch": 0.5645828747523662, + "grad_norm": 1.608318165866991, + "learning_rate": 4.20052837902141e-06, + "loss": 0.3663, + "step": 25649 + }, + { + "epoch": 0.5646048866387849, + "grad_norm": 1.5660427097580871, + "learning_rate": 4.200176510792835e-06, + "loss": 0.3099, + "step": 25650 + }, + { + "epoch": 0.5646268985252036, + "grad_norm": 1.5479701368403447, + "learning_rate": 4.199824646629322e-06, + "loss": 0.3692, + "step": 25651 + }, + { + "epoch": 0.5646489104116222, + "grad_norm": 1.3630642420890873, + "learning_rate": 4.199472786532655e-06, + "loss": 0.3578, + "step": 25652 + }, + { + "epoch": 0.564670922298041, + "grad_norm": 2.276817052532591, + "learning_rate": 4.1991209305046255e-06, + "loss": 0.2807, + "step": 25653 + }, + { + "epoch": 0.5646929341844596, + "grad_norm": 1.8625704317948242, + "learning_rate": 4.198769078547022e-06, + "loss": 0.3234, + "step": 25654 + }, + { + "epoch": 0.5647149460708782, + "grad_norm": 1.783868502271555, + "learning_rate": 4.198417230661629e-06, + "loss": 0.238, + "step": 25655 + }, + { + "epoch": 0.564736957957297, + "grad_norm": 1.6437559159845625, + "learning_rate": 4.198065386850238e-06, + "loss": 0.3725, + "step": 25656 + }, + { + "epoch": 0.5647589698437157, + "grad_norm": 1.376316114926709, + "learning_rate": 4.1977135471146355e-06, + "loss": 0.3674, + "step": 25657 + }, + { + "epoch": 0.5647809817301342, + "grad_norm": 1.1591653890744218, + "learning_rate": 4.197361711456612e-06, + "loss": 0.3245, + "step": 25658 + }, + { + "epoch": 0.564802993616553, + "grad_norm": 1.0899474801553166, + "learning_rate": 4.197009879877955e-06, + "loss": 0.2026, + "step": 25659 + }, + { + "epoch": 0.5648250055029717, + "grad_norm": 2.239798151885799, + "learning_rate": 4.19665805238045e-06, + "loss": 0.2785, + "step": 25660 + }, + { + "epoch": 0.5648470173893902, + "grad_norm": 1.4965920889637507, + "learning_rate": 4.196306228965888e-06, + "loss": 0.3435, + "step": 25661 + }, + { + "epoch": 0.564869029275809, + "grad_norm": 2.1128276433067503, + "learning_rate": 4.195954409636055e-06, + "loss": 0.4204, + "step": 25662 + }, + { + "epoch": 0.5648910411622277, + "grad_norm": 1.1001700662623368, + "learning_rate": 4.195602594392743e-06, + "loss": 0.2716, + "step": 25663 + }, + { + "epoch": 0.5649130530486463, + "grad_norm": 0.6202517754397728, + "learning_rate": 4.195250783237735e-06, + "loss": 0.4828, + "step": 25664 + }, + { + "epoch": 0.564935064935065, + "grad_norm": 1.472984856871643, + "learning_rate": 4.194898976172822e-06, + "loss": 0.3019, + "step": 25665 + }, + { + "epoch": 0.5649570768214836, + "grad_norm": 1.4799806222127714, + "learning_rate": 4.1945471731997915e-06, + "loss": 0.2903, + "step": 25666 + }, + { + "epoch": 0.5649790887079023, + "grad_norm": 1.3976883763422576, + "learning_rate": 4.194195374320433e-06, + "loss": 0.4023, + "step": 25667 + }, + { + "epoch": 0.565001100594321, + "grad_norm": 0.6095401155829572, + "learning_rate": 4.19384357953653e-06, + "loss": 0.4897, + "step": 25668 + }, + { + "epoch": 0.5650231124807396, + "grad_norm": 1.4451102876682547, + "learning_rate": 4.1934917888498765e-06, + "loss": 0.2777, + "step": 25669 + }, + { + "epoch": 0.5650451243671583, + "grad_norm": 0.6084596450864387, + "learning_rate": 4.193140002262255e-06, + "loss": 0.4685, + "step": 25670 + }, + { + "epoch": 0.565067136253577, + "grad_norm": 1.4011249089245799, + "learning_rate": 4.1927882197754574e-06, + "loss": 0.3251, + "step": 25671 + }, + { + "epoch": 0.5650891481399956, + "grad_norm": 0.9605916006359778, + "learning_rate": 4.192436441391271e-06, + "loss": 0.19, + "step": 25672 + }, + { + "epoch": 0.5651111600264143, + "grad_norm": 1.5336383493338759, + "learning_rate": 4.19208466711148e-06, + "loss": 0.3106, + "step": 25673 + }, + { + "epoch": 0.565133171912833, + "grad_norm": 1.604287938771196, + "learning_rate": 4.191732896937877e-06, + "loss": 0.3089, + "step": 25674 + }, + { + "epoch": 0.5651551837992516, + "grad_norm": 1.2967940075186444, + "learning_rate": 4.191381130872246e-06, + "loss": 0.2617, + "step": 25675 + }, + { + "epoch": 0.5651771956856703, + "grad_norm": 1.4045281066186734, + "learning_rate": 4.191029368916379e-06, + "loss": 0.301, + "step": 25676 + }, + { + "epoch": 0.565199207572089, + "grad_norm": 0.6033512935270188, + "learning_rate": 4.190677611072062e-06, + "loss": 0.4638, + "step": 25677 + }, + { + "epoch": 0.5652212194585076, + "grad_norm": 0.6172743006444253, + "learning_rate": 4.19032585734108e-06, + "loss": 0.4725, + "step": 25678 + }, + { + "epoch": 0.5652432313449263, + "grad_norm": 1.233828890515779, + "learning_rate": 4.189974107725224e-06, + "loss": 0.2273, + "step": 25679 + }, + { + "epoch": 0.565265243231345, + "grad_norm": 5.1878157241572325, + "learning_rate": 4.189622362226281e-06, + "loss": 0.2174, + "step": 25680 + }, + { + "epoch": 0.5652872551177636, + "grad_norm": 1.3298639755591621, + "learning_rate": 4.18927062084604e-06, + "loss": 0.2587, + "step": 25681 + }, + { + "epoch": 0.5653092670041823, + "grad_norm": 0.610571024214511, + "learning_rate": 4.188918883586286e-06, + "loss": 0.4849, + "step": 25682 + }, + { + "epoch": 0.565331278890601, + "grad_norm": 0.6220794372974771, + "learning_rate": 4.188567150448807e-06, + "loss": 0.4667, + "step": 25683 + }, + { + "epoch": 0.5653532907770196, + "grad_norm": 1.7383821012192562, + "learning_rate": 4.188215421435393e-06, + "loss": 0.3213, + "step": 25684 + }, + { + "epoch": 0.5653753026634383, + "grad_norm": 1.7252882154726925, + "learning_rate": 4.1878636965478285e-06, + "loss": 0.3258, + "step": 25685 + }, + { + "epoch": 0.565397314549857, + "grad_norm": 1.8099916933986717, + "learning_rate": 4.187511975787906e-06, + "loss": 0.2726, + "step": 25686 + }, + { + "epoch": 0.5654193264362756, + "grad_norm": 1.1383091046648006, + "learning_rate": 4.187160259157409e-06, + "loss": 0.3087, + "step": 25687 + }, + { + "epoch": 0.5654413383226943, + "grad_norm": 9.496068016671742, + "learning_rate": 4.186808546658124e-06, + "loss": 0.2553, + "step": 25688 + }, + { + "epoch": 0.5654633502091129, + "grad_norm": 1.4453241381185642, + "learning_rate": 4.186456838291843e-06, + "loss": 0.3123, + "step": 25689 + }, + { + "epoch": 0.5654853620955316, + "grad_norm": 1.4863014349006989, + "learning_rate": 4.186105134060351e-06, + "loss": 0.378, + "step": 25690 + }, + { + "epoch": 0.5655073739819503, + "grad_norm": 2.276842294466443, + "learning_rate": 4.1857534339654335e-06, + "loss": 0.4117, + "step": 25691 + }, + { + "epoch": 0.5655293858683689, + "grad_norm": 1.355443333221461, + "learning_rate": 4.185401738008883e-06, + "loss": 0.3398, + "step": 25692 + }, + { + "epoch": 0.5655513977547876, + "grad_norm": 2.4517504256457734, + "learning_rate": 4.185050046192482e-06, + "loss": 0.2483, + "step": 25693 + }, + { + "epoch": 0.5655734096412063, + "grad_norm": 1.255646955993595, + "learning_rate": 4.184698358518021e-06, + "loss": 0.2922, + "step": 25694 + }, + { + "epoch": 0.5655954215276249, + "grad_norm": 1.1532896339636858, + "learning_rate": 4.184346674987288e-06, + "loss": 0.3296, + "step": 25695 + }, + { + "epoch": 0.5656174334140436, + "grad_norm": 1.6675756322106183, + "learning_rate": 4.183994995602067e-06, + "loss": 0.3705, + "step": 25696 + }, + { + "epoch": 0.5656394453004623, + "grad_norm": 1.9631547608588724, + "learning_rate": 4.183643320364148e-06, + "loss": 0.3951, + "step": 25697 + }, + { + "epoch": 0.5656614571868809, + "grad_norm": 1.2946576899072775, + "learning_rate": 4.183291649275316e-06, + "loss": 0.2831, + "step": 25698 + }, + { + "epoch": 0.5656834690732996, + "grad_norm": 1.4562884086242192, + "learning_rate": 4.182939982337364e-06, + "loss": 0.2684, + "step": 25699 + }, + { + "epoch": 0.5657054809597183, + "grad_norm": 2.404724989236636, + "learning_rate": 4.1825883195520724e-06, + "loss": 0.3367, + "step": 25700 + }, + { + "epoch": 0.5657274928461369, + "grad_norm": 0.6573002707799515, + "learning_rate": 4.182236660921231e-06, + "loss": 0.4869, + "step": 25701 + }, + { + "epoch": 0.5657495047325556, + "grad_norm": 0.6489447428186086, + "learning_rate": 4.18188500644663e-06, + "loss": 0.4686, + "step": 25702 + }, + { + "epoch": 0.5657715166189743, + "grad_norm": 1.804484669425019, + "learning_rate": 4.181533356130053e-06, + "loss": 0.2618, + "step": 25703 + }, + { + "epoch": 0.5657935285053929, + "grad_norm": 5.204176330918124, + "learning_rate": 4.1811817099732895e-06, + "loss": 0.3105, + "step": 25704 + }, + { + "epoch": 0.5658155403918116, + "grad_norm": 1.385269988522539, + "learning_rate": 4.1808300679781255e-06, + "loss": 0.3414, + "step": 25705 + }, + { + "epoch": 0.5658375522782303, + "grad_norm": 1.5280432723993502, + "learning_rate": 4.180478430146348e-06, + "loss": 0.2934, + "step": 25706 + }, + { + "epoch": 0.5658595641646489, + "grad_norm": 1.2360153962490732, + "learning_rate": 4.1801267964797444e-06, + "loss": 0.3116, + "step": 25707 + }, + { + "epoch": 0.5658815760510676, + "grad_norm": 1.084767233287189, + "learning_rate": 4.179775166980104e-06, + "loss": 0.2804, + "step": 25708 + }, + { + "epoch": 0.5659035879374863, + "grad_norm": 1.555567932166337, + "learning_rate": 4.179423541649209e-06, + "loss": 0.4004, + "step": 25709 + }, + { + "epoch": 0.5659255998239049, + "grad_norm": 0.6492794400382381, + "learning_rate": 4.179071920488851e-06, + "loss": 0.4696, + "step": 25710 + }, + { + "epoch": 0.5659476117103236, + "grad_norm": 2.2952977703378346, + "learning_rate": 4.178720303500815e-06, + "loss": 0.402, + "step": 25711 + }, + { + "epoch": 0.5659696235967422, + "grad_norm": 1.4369752573412955, + "learning_rate": 4.178368690686889e-06, + "loss": 0.2988, + "step": 25712 + }, + { + "epoch": 0.5659916354831609, + "grad_norm": 4.189030040593421, + "learning_rate": 4.1780170820488615e-06, + "loss": 0.3063, + "step": 25713 + }, + { + "epoch": 0.5660136473695796, + "grad_norm": 1.602685207162789, + "learning_rate": 4.177665477588516e-06, + "loss": 0.2902, + "step": 25714 + }, + { + "epoch": 0.5660356592559982, + "grad_norm": 1.7267426209718644, + "learning_rate": 4.1773138773076416e-06, + "loss": 0.2897, + "step": 25715 + }, + { + "epoch": 0.5660576711424169, + "grad_norm": 2.1082078744836164, + "learning_rate": 4.176962281208024e-06, + "loss": 0.3366, + "step": 25716 + }, + { + "epoch": 0.5660796830288356, + "grad_norm": 0.6032577160388062, + "learning_rate": 4.176610689291454e-06, + "loss": 0.4523, + "step": 25717 + }, + { + "epoch": 0.5661016949152542, + "grad_norm": 1.4580483030804892, + "learning_rate": 4.176259101559714e-06, + "loss": 0.2973, + "step": 25718 + }, + { + "epoch": 0.5661237068016729, + "grad_norm": 4.191073528464875, + "learning_rate": 4.175907518014592e-06, + "loss": 0.3197, + "step": 25719 + }, + { + "epoch": 0.5661457186880916, + "grad_norm": 1.6554773425790126, + "learning_rate": 4.175555938657876e-06, + "loss": 0.2965, + "step": 25720 + }, + { + "epoch": 0.5661677305745102, + "grad_norm": 1.9198284655542284, + "learning_rate": 4.175204363491352e-06, + "loss": 0.2546, + "step": 25721 + }, + { + "epoch": 0.5661897424609289, + "grad_norm": 0.6223865758473611, + "learning_rate": 4.174852792516809e-06, + "loss": 0.4679, + "step": 25722 + }, + { + "epoch": 0.5662117543473476, + "grad_norm": 1.5280792627558222, + "learning_rate": 4.174501225736031e-06, + "loss": 0.25, + "step": 25723 + }, + { + "epoch": 0.5662337662337662, + "grad_norm": 1.4636563168679162, + "learning_rate": 4.1741496631508056e-06, + "loss": 0.2744, + "step": 25724 + }, + { + "epoch": 0.5662557781201849, + "grad_norm": 1.6356350228278471, + "learning_rate": 4.1737981047629205e-06, + "loss": 0.2738, + "step": 25725 + }, + { + "epoch": 0.5662777900066036, + "grad_norm": 1.2180827925731714, + "learning_rate": 4.173446550574163e-06, + "loss": 0.2463, + "step": 25726 + }, + { + "epoch": 0.5662998018930222, + "grad_norm": 0.5998037039735887, + "learning_rate": 4.173095000586317e-06, + "loss": 0.4795, + "step": 25727 + }, + { + "epoch": 0.5663218137794409, + "grad_norm": 2.4077642879718493, + "learning_rate": 4.172743454801171e-06, + "loss": 0.2756, + "step": 25728 + }, + { + "epoch": 0.5663438256658596, + "grad_norm": 2.5595969587499328, + "learning_rate": 4.172391913220512e-06, + "loss": 0.3357, + "step": 25729 + }, + { + "epoch": 0.5663658375522782, + "grad_norm": 1.4913628353921402, + "learning_rate": 4.172040375846127e-06, + "loss": 0.3351, + "step": 25730 + }, + { + "epoch": 0.5663878494386969, + "grad_norm": 0.5874920544756378, + "learning_rate": 4.1716888426798034e-06, + "loss": 0.4848, + "step": 25731 + }, + { + "epoch": 0.5664098613251156, + "grad_norm": 1.5431024630162016, + "learning_rate": 4.171337313723324e-06, + "loss": 0.3018, + "step": 25732 + }, + { + "epoch": 0.5664318732115342, + "grad_norm": 0.6073948951313558, + "learning_rate": 4.170985788978479e-06, + "loss": 0.4871, + "step": 25733 + }, + { + "epoch": 0.5664538850979529, + "grad_norm": 2.3046453158489, + "learning_rate": 4.170634268447053e-06, + "loss": 0.3209, + "step": 25734 + }, + { + "epoch": 0.5664758969843715, + "grad_norm": 1.7661359265976302, + "learning_rate": 4.170282752130836e-06, + "loss": 0.2978, + "step": 25735 + }, + { + "epoch": 0.5664979088707902, + "grad_norm": 1.4629426635832934, + "learning_rate": 4.169931240031611e-06, + "loss": 0.3073, + "step": 25736 + }, + { + "epoch": 0.5665199207572089, + "grad_norm": 0.615575606122431, + "learning_rate": 4.169579732151164e-06, + "loss": 0.4775, + "step": 25737 + }, + { + "epoch": 0.5665419326436275, + "grad_norm": 1.3107816923725024, + "learning_rate": 4.169228228491284e-06, + "loss": 0.2737, + "step": 25738 + }, + { + "epoch": 0.5665639445300462, + "grad_norm": 1.3683037885420433, + "learning_rate": 4.168876729053757e-06, + "loss": 0.3022, + "step": 25739 + }, + { + "epoch": 0.5665859564164649, + "grad_norm": 1.3390002106109757, + "learning_rate": 4.1685252338403705e-06, + "loss": 0.2498, + "step": 25740 + }, + { + "epoch": 0.5666079683028835, + "grad_norm": 0.6202200178109764, + "learning_rate": 4.168173742852909e-06, + "loss": 0.4804, + "step": 25741 + }, + { + "epoch": 0.5666299801893022, + "grad_norm": 1.4970112508928028, + "learning_rate": 4.167822256093157e-06, + "loss": 0.2887, + "step": 25742 + }, + { + "epoch": 0.5666519920757209, + "grad_norm": 1.5936351785138338, + "learning_rate": 4.167470773562906e-06, + "loss": 0.3298, + "step": 25743 + }, + { + "epoch": 0.5666740039621395, + "grad_norm": 2.152952941216778, + "learning_rate": 4.1671192952639385e-06, + "loss": 0.3283, + "step": 25744 + }, + { + "epoch": 0.5666960158485582, + "grad_norm": 2.3217219228052675, + "learning_rate": 4.166767821198044e-06, + "loss": 0.2961, + "step": 25745 + }, + { + "epoch": 0.5667180277349769, + "grad_norm": 1.347380294974181, + "learning_rate": 4.166416351367006e-06, + "loss": 0.3078, + "step": 25746 + }, + { + "epoch": 0.5667400396213955, + "grad_norm": 1.2909720909725841, + "learning_rate": 4.166064885772611e-06, + "loss": 0.3081, + "step": 25747 + }, + { + "epoch": 0.5667620515078142, + "grad_norm": 1.897498835098001, + "learning_rate": 4.165713424416647e-06, + "loss": 0.3351, + "step": 25748 + }, + { + "epoch": 0.5667840633942329, + "grad_norm": 1.3240116835189182, + "learning_rate": 4.1653619673009014e-06, + "loss": 0.2833, + "step": 25749 + }, + { + "epoch": 0.5668060752806515, + "grad_norm": 1.2805006358497069, + "learning_rate": 4.165010514427156e-06, + "loss": 0.2528, + "step": 25750 + }, + { + "epoch": 0.5668280871670702, + "grad_norm": 1.4086338016429258, + "learning_rate": 4.1646590657972005e-06, + "loss": 0.2975, + "step": 25751 + }, + { + "epoch": 0.5668500990534889, + "grad_norm": 0.634591111855536, + "learning_rate": 4.164307621412819e-06, + "loss": 0.4757, + "step": 25752 + }, + { + "epoch": 0.5668721109399075, + "grad_norm": 1.913814263374787, + "learning_rate": 4.163956181275801e-06, + "loss": 0.2897, + "step": 25753 + }, + { + "epoch": 0.5668941228263262, + "grad_norm": 1.4399764233379362, + "learning_rate": 4.16360474538793e-06, + "loss": 0.2993, + "step": 25754 + }, + { + "epoch": 0.5669161347127449, + "grad_norm": 1.2774435317444097, + "learning_rate": 4.16325331375099e-06, + "loss": 0.3554, + "step": 25755 + }, + { + "epoch": 0.5669381465991635, + "grad_norm": 1.5757464716040657, + "learning_rate": 4.162901886366773e-06, + "loss": 0.2715, + "step": 25756 + }, + { + "epoch": 0.5669601584855822, + "grad_norm": 5.938052201674158, + "learning_rate": 4.16255046323706e-06, + "loss": 0.3255, + "step": 25757 + }, + { + "epoch": 0.5669821703720008, + "grad_norm": 1.4846689631751258, + "learning_rate": 4.1621990443636415e-06, + "loss": 0.2943, + "step": 25758 + }, + { + "epoch": 0.5670041822584195, + "grad_norm": 1.3076824755150156, + "learning_rate": 4.1618476297483e-06, + "loss": 0.3167, + "step": 25759 + }, + { + "epoch": 0.5670261941448382, + "grad_norm": 0.6270680148525541, + "learning_rate": 4.1614962193928215e-06, + "loss": 0.481, + "step": 25760 + }, + { + "epoch": 0.5670482060312568, + "grad_norm": 1.413622578588673, + "learning_rate": 4.161144813298994e-06, + "loss": 0.3768, + "step": 25761 + }, + { + "epoch": 0.5670702179176755, + "grad_norm": 1.6350715159377651, + "learning_rate": 4.160793411468602e-06, + "loss": 0.4164, + "step": 25762 + }, + { + "epoch": 0.5670922298040942, + "grad_norm": 1.6040744268590603, + "learning_rate": 4.160442013903434e-06, + "loss": 0.3736, + "step": 25763 + }, + { + "epoch": 0.5671142416905128, + "grad_norm": 1.3413470691432905, + "learning_rate": 4.160090620605274e-06, + "loss": 0.2864, + "step": 25764 + }, + { + "epoch": 0.5671362535769315, + "grad_norm": 1.3898542993962177, + "learning_rate": 4.159739231575905e-06, + "loss": 0.3374, + "step": 25765 + }, + { + "epoch": 0.5671582654633502, + "grad_norm": 1.0088902105618809, + "learning_rate": 4.159387846817119e-06, + "loss": 0.1904, + "step": 25766 + }, + { + "epoch": 0.5671802773497688, + "grad_norm": 0.6122572477738841, + "learning_rate": 4.159036466330699e-06, + "loss": 0.4751, + "step": 25767 + }, + { + "epoch": 0.5672022892361875, + "grad_norm": 1.3801204646855298, + "learning_rate": 4.158685090118429e-06, + "loss": 0.3084, + "step": 25768 + }, + { + "epoch": 0.5672243011226062, + "grad_norm": 1.5238913283018587, + "learning_rate": 4.158333718182098e-06, + "loss": 0.3294, + "step": 25769 + }, + { + "epoch": 0.5672463130090248, + "grad_norm": 0.6225926413553599, + "learning_rate": 4.157982350523489e-06, + "loss": 0.4798, + "step": 25770 + }, + { + "epoch": 0.5672683248954435, + "grad_norm": 0.5904174457914405, + "learning_rate": 4.157630987144391e-06, + "loss": 0.469, + "step": 25771 + }, + { + "epoch": 0.5672903367818622, + "grad_norm": 1.3726029894677476, + "learning_rate": 4.157279628046588e-06, + "loss": 0.3147, + "step": 25772 + }, + { + "epoch": 0.5673123486682808, + "grad_norm": 1.7822242385622975, + "learning_rate": 4.156928273231864e-06, + "loss": 0.3093, + "step": 25773 + }, + { + "epoch": 0.5673343605546995, + "grad_norm": 0.6312197717499705, + "learning_rate": 4.156576922702007e-06, + "loss": 0.4754, + "step": 25774 + }, + { + "epoch": 0.5673563724411183, + "grad_norm": 1.6370803827124614, + "learning_rate": 4.156225576458802e-06, + "loss": 0.3073, + "step": 25775 + }, + { + "epoch": 0.5673783843275368, + "grad_norm": 1.715854736737786, + "learning_rate": 4.155874234504037e-06, + "loss": 0.3267, + "step": 25776 + }, + { + "epoch": 0.5674003962139555, + "grad_norm": 1.844252189659008, + "learning_rate": 4.155522896839495e-06, + "loss": 0.3026, + "step": 25777 + }, + { + "epoch": 0.5674224081003743, + "grad_norm": 1.3576127033353793, + "learning_rate": 4.155171563466961e-06, + "loss": 0.3342, + "step": 25778 + }, + { + "epoch": 0.5674444199867928, + "grad_norm": 0.5859100249103574, + "learning_rate": 4.154820234388221e-06, + "loss": 0.4506, + "step": 25779 + }, + { + "epoch": 0.5674664318732116, + "grad_norm": 0.6296122010547687, + "learning_rate": 4.154468909605063e-06, + "loss": 0.4542, + "step": 25780 + }, + { + "epoch": 0.5674884437596301, + "grad_norm": 1.4912924447582536, + "learning_rate": 4.154117589119272e-06, + "loss": 0.3483, + "step": 25781 + }, + { + "epoch": 0.5675104556460489, + "grad_norm": 1.600720973307553, + "learning_rate": 4.153766272932632e-06, + "loss": 0.2932, + "step": 25782 + }, + { + "epoch": 0.5675324675324676, + "grad_norm": 1.486573556716317, + "learning_rate": 4.1534149610469285e-06, + "loss": 0.2643, + "step": 25783 + }, + { + "epoch": 0.5675544794188861, + "grad_norm": 1.6942858392080398, + "learning_rate": 4.153063653463948e-06, + "loss": 0.338, + "step": 25784 + }, + { + "epoch": 0.5675764913053049, + "grad_norm": 1.3121733874687826, + "learning_rate": 4.152712350185476e-06, + "loss": 0.2955, + "step": 25785 + }, + { + "epoch": 0.5675985031917236, + "grad_norm": 1.3310020173747306, + "learning_rate": 4.152361051213299e-06, + "loss": 0.2912, + "step": 25786 + }, + { + "epoch": 0.5676205150781422, + "grad_norm": 1.3980647117671214, + "learning_rate": 4.1520097565492005e-06, + "loss": 0.2917, + "step": 25787 + }, + { + "epoch": 0.5676425269645609, + "grad_norm": 0.6378991462432343, + "learning_rate": 4.151658466194965e-06, + "loss": 0.4695, + "step": 25788 + }, + { + "epoch": 0.5676645388509796, + "grad_norm": 3.8962939223943227, + "learning_rate": 4.151307180152382e-06, + "loss": 0.2944, + "step": 25789 + }, + { + "epoch": 0.5676865507373982, + "grad_norm": 0.6309082734209536, + "learning_rate": 4.150955898423233e-06, + "loss": 0.4743, + "step": 25790 + }, + { + "epoch": 0.5677085626238169, + "grad_norm": 1.2627485945402348, + "learning_rate": 4.150604621009304e-06, + "loss": 0.2392, + "step": 25791 + }, + { + "epoch": 0.5677305745102356, + "grad_norm": 1.6423362483139463, + "learning_rate": 4.1502533479123826e-06, + "loss": 0.3845, + "step": 25792 + }, + { + "epoch": 0.5677525863966542, + "grad_norm": 1.2372423318243952, + "learning_rate": 4.149902079134251e-06, + "loss": 0.3898, + "step": 25793 + }, + { + "epoch": 0.5677745982830729, + "grad_norm": 2.030855637743229, + "learning_rate": 4.1495508146766986e-06, + "loss": 0.3621, + "step": 25794 + }, + { + "epoch": 0.5677966101694916, + "grad_norm": 1.4948663692385185, + "learning_rate": 4.149199554541507e-06, + "loss": 0.3073, + "step": 25795 + }, + { + "epoch": 0.5678186220559102, + "grad_norm": 1.2814642102769938, + "learning_rate": 4.148848298730462e-06, + "loss": 0.2319, + "step": 25796 + }, + { + "epoch": 0.5678406339423289, + "grad_norm": 0.6199954030779192, + "learning_rate": 4.148497047245351e-06, + "loss": 0.4948, + "step": 25797 + }, + { + "epoch": 0.5678626458287476, + "grad_norm": 1.6120600209934066, + "learning_rate": 4.148145800087956e-06, + "loss": 0.2918, + "step": 25798 + }, + { + "epoch": 0.5678846577151662, + "grad_norm": 1.4033827210734158, + "learning_rate": 4.147794557260066e-06, + "loss": 0.3051, + "step": 25799 + }, + { + "epoch": 0.5679066696015849, + "grad_norm": 1.5652294742344908, + "learning_rate": 4.147443318763464e-06, + "loss": 0.3501, + "step": 25800 + }, + { + "epoch": 0.5679286814880036, + "grad_norm": 0.6217117334325968, + "learning_rate": 4.147092084599933e-06, + "loss": 0.4657, + "step": 25801 + }, + { + "epoch": 0.5679506933744222, + "grad_norm": 1.4612145670351073, + "learning_rate": 4.146740854771263e-06, + "loss": 0.3173, + "step": 25802 + }, + { + "epoch": 0.5679727052608409, + "grad_norm": 2.677656545699003, + "learning_rate": 4.146389629279235e-06, + "loss": 0.4077, + "step": 25803 + }, + { + "epoch": 0.5679947171472595, + "grad_norm": 2.3418243131400396, + "learning_rate": 4.146038408125636e-06, + "loss": 0.3294, + "step": 25804 + }, + { + "epoch": 0.5680167290336782, + "grad_norm": 2.0738024143543057, + "learning_rate": 4.145687191312252e-06, + "loss": 0.3204, + "step": 25805 + }, + { + "epoch": 0.5680387409200969, + "grad_norm": 1.356446711686604, + "learning_rate": 4.145335978840865e-06, + "loss": 0.2803, + "step": 25806 + }, + { + "epoch": 0.5680607528065155, + "grad_norm": 2.1711765655669906, + "learning_rate": 4.144984770713263e-06, + "loss": 0.318, + "step": 25807 + }, + { + "epoch": 0.5680827646929342, + "grad_norm": 2.3027693642602203, + "learning_rate": 4.144633566931229e-06, + "loss": 0.3137, + "step": 25808 + }, + { + "epoch": 0.5681047765793529, + "grad_norm": 1.5428023281352548, + "learning_rate": 4.144282367496548e-06, + "loss": 0.3169, + "step": 25809 + }, + { + "epoch": 0.5681267884657715, + "grad_norm": 1.4232886246291705, + "learning_rate": 4.143931172411007e-06, + "loss": 0.3508, + "step": 25810 + }, + { + "epoch": 0.5681488003521902, + "grad_norm": 1.3301700819348696, + "learning_rate": 4.143579981676388e-06, + "loss": 0.3473, + "step": 25811 + }, + { + "epoch": 0.5681708122386089, + "grad_norm": 1.3345712778004852, + "learning_rate": 4.14322879529448e-06, + "loss": 0.3029, + "step": 25812 + }, + { + "epoch": 0.5681928241250275, + "grad_norm": 1.2600045978288206, + "learning_rate": 4.142877613267064e-06, + "loss": 0.2134, + "step": 25813 + }, + { + "epoch": 0.5682148360114462, + "grad_norm": 1.4509160441513789, + "learning_rate": 4.142526435595925e-06, + "loss": 0.3828, + "step": 25814 + }, + { + "epoch": 0.5682368478978649, + "grad_norm": 0.6425759964122518, + "learning_rate": 4.142175262282851e-06, + "loss": 0.499, + "step": 25815 + }, + { + "epoch": 0.5682588597842835, + "grad_norm": 1.9178738301694727, + "learning_rate": 4.141824093329623e-06, + "loss": 0.3903, + "step": 25816 + }, + { + "epoch": 0.5682808716707022, + "grad_norm": 1.7237067179944674, + "learning_rate": 4.14147292873803e-06, + "loss": 0.3086, + "step": 25817 + }, + { + "epoch": 0.5683028835571209, + "grad_norm": 1.4710103355891615, + "learning_rate": 4.141121768509854e-06, + "loss": 0.316, + "step": 25818 + }, + { + "epoch": 0.5683248954435395, + "grad_norm": 1.660229424495705, + "learning_rate": 4.140770612646879e-06, + "loss": 0.2792, + "step": 25819 + }, + { + "epoch": 0.5683469073299582, + "grad_norm": 2.1302081569554163, + "learning_rate": 4.1404194611508925e-06, + "loss": 0.2777, + "step": 25820 + }, + { + "epoch": 0.5683689192163769, + "grad_norm": 1.3508247826728499, + "learning_rate": 4.140068314023675e-06, + "loss": 0.2752, + "step": 25821 + }, + { + "epoch": 0.5683909311027955, + "grad_norm": 1.5018533238236194, + "learning_rate": 4.139717171267017e-06, + "loss": 0.3256, + "step": 25822 + }, + { + "epoch": 0.5684129429892142, + "grad_norm": 1.5155214321700479, + "learning_rate": 4.139366032882699e-06, + "loss": 0.3035, + "step": 25823 + }, + { + "epoch": 0.5684349548756329, + "grad_norm": 1.3481416241341877, + "learning_rate": 4.139014898872505e-06, + "loss": 0.3244, + "step": 25824 + }, + { + "epoch": 0.5684569667620515, + "grad_norm": 1.6363115705029596, + "learning_rate": 4.138663769238224e-06, + "loss": 0.3194, + "step": 25825 + }, + { + "epoch": 0.5684789786484702, + "grad_norm": 2.6375504325414383, + "learning_rate": 4.138312643981637e-06, + "loss": 0.283, + "step": 25826 + }, + { + "epoch": 0.5685009905348889, + "grad_norm": 1.6671360057339693, + "learning_rate": 4.137961523104528e-06, + "loss": 0.236, + "step": 25827 + }, + { + "epoch": 0.5685230024213075, + "grad_norm": 1.2238742742985134, + "learning_rate": 4.1376104066086845e-06, + "loss": 0.329, + "step": 25828 + }, + { + "epoch": 0.5685450143077262, + "grad_norm": 4.2134086182463335, + "learning_rate": 4.137259294495889e-06, + "loss": 0.3756, + "step": 25829 + }, + { + "epoch": 0.5685670261941448, + "grad_norm": 1.1223156177272389, + "learning_rate": 4.136908186767928e-06, + "loss": 0.2965, + "step": 25830 + }, + { + "epoch": 0.5685890380805635, + "grad_norm": 2.1885316423553407, + "learning_rate": 4.1365570834265825e-06, + "loss": 0.3427, + "step": 25831 + }, + { + "epoch": 0.5686110499669822, + "grad_norm": 2.042841285695545, + "learning_rate": 4.136205984473639e-06, + "loss": 0.2908, + "step": 25832 + }, + { + "epoch": 0.5686330618534008, + "grad_norm": 1.8935485977856488, + "learning_rate": 4.135854889910883e-06, + "loss": 0.3685, + "step": 25833 + }, + { + "epoch": 0.5686550737398195, + "grad_norm": 1.4323576369275857, + "learning_rate": 4.135503799740096e-06, + "loss": 0.36, + "step": 25834 + }, + { + "epoch": 0.5686770856262382, + "grad_norm": 1.6764992096667166, + "learning_rate": 4.135152713963068e-06, + "loss": 0.3865, + "step": 25835 + }, + { + "epoch": 0.5686990975126568, + "grad_norm": 1.867548039859978, + "learning_rate": 4.134801632581577e-06, + "loss": 0.4043, + "step": 25836 + }, + { + "epoch": 0.5687211093990755, + "grad_norm": 1.3467585640995374, + "learning_rate": 4.134450555597409e-06, + "loss": 0.2997, + "step": 25837 + }, + { + "epoch": 0.5687431212854942, + "grad_norm": 1.4716667272295565, + "learning_rate": 4.134099483012353e-06, + "loss": 0.3526, + "step": 25838 + }, + { + "epoch": 0.5687651331719128, + "grad_norm": 0.6449776869491343, + "learning_rate": 4.133748414828185e-06, + "loss": 0.4796, + "step": 25839 + }, + { + "epoch": 0.5687871450583315, + "grad_norm": 1.7038446947385912, + "learning_rate": 4.133397351046698e-06, + "loss": 0.2787, + "step": 25840 + }, + { + "epoch": 0.5688091569447502, + "grad_norm": 1.5387762797437003, + "learning_rate": 4.1330462916696706e-06, + "loss": 0.3113, + "step": 25841 + }, + { + "epoch": 0.5688311688311688, + "grad_norm": 1.148727994707076, + "learning_rate": 4.132695236698887e-06, + "loss": 0.2724, + "step": 25842 + }, + { + "epoch": 0.5688531807175875, + "grad_norm": 1.4541804213101448, + "learning_rate": 4.132344186136136e-06, + "loss": 0.2734, + "step": 25843 + }, + { + "epoch": 0.5688751926040062, + "grad_norm": 1.6062480384471212, + "learning_rate": 4.131993139983196e-06, + "loss": 0.2791, + "step": 25844 + }, + { + "epoch": 0.5688972044904248, + "grad_norm": 1.6324239242462115, + "learning_rate": 4.131642098241856e-06, + "loss": 0.3157, + "step": 25845 + }, + { + "epoch": 0.5689192163768435, + "grad_norm": 1.2322441997980904, + "learning_rate": 4.131291060913898e-06, + "loss": 0.2124, + "step": 25846 + }, + { + "epoch": 0.5689412282632622, + "grad_norm": 1.9181007974566069, + "learning_rate": 4.130940028001105e-06, + "loss": 0.3035, + "step": 25847 + }, + { + "epoch": 0.5689632401496808, + "grad_norm": 1.6327729200943246, + "learning_rate": 4.130588999505265e-06, + "loss": 0.3071, + "step": 25848 + }, + { + "epoch": 0.5689852520360995, + "grad_norm": 1.4813367422722987, + "learning_rate": 4.1302379754281574e-06, + "loss": 0.3146, + "step": 25849 + }, + { + "epoch": 0.5690072639225182, + "grad_norm": 2.8555157910855846, + "learning_rate": 4.129886955771568e-06, + "loss": 0.3148, + "step": 25850 + }, + { + "epoch": 0.5690292758089368, + "grad_norm": 1.7390414507952965, + "learning_rate": 4.129535940537282e-06, + "loss": 0.2309, + "step": 25851 + }, + { + "epoch": 0.5690512876953555, + "grad_norm": 0.6327104280924734, + "learning_rate": 4.129184929727081e-06, + "loss": 0.4802, + "step": 25852 + }, + { + "epoch": 0.5690732995817741, + "grad_norm": 1.2127910449057937, + "learning_rate": 4.128833923342753e-06, + "loss": 0.2731, + "step": 25853 + }, + { + "epoch": 0.5690953114681928, + "grad_norm": 2.352151426766818, + "learning_rate": 4.128482921386079e-06, + "loss": 0.3196, + "step": 25854 + }, + { + "epoch": 0.5691173233546115, + "grad_norm": 0.6471307634155972, + "learning_rate": 4.1281319238588425e-06, + "loss": 0.4683, + "step": 25855 + }, + { + "epoch": 0.5691393352410301, + "grad_norm": 0.6040936717521967, + "learning_rate": 4.12778093076283e-06, + "loss": 0.4897, + "step": 25856 + }, + { + "epoch": 0.5691613471274488, + "grad_norm": 2.3614249905779423, + "learning_rate": 4.127429942099821e-06, + "loss": 0.3249, + "step": 25857 + }, + { + "epoch": 0.5691833590138675, + "grad_norm": 0.613955847081559, + "learning_rate": 4.127078957871605e-06, + "loss": 0.4629, + "step": 25858 + }, + { + "epoch": 0.5692053709002861, + "grad_norm": 1.4917705397236625, + "learning_rate": 4.126727978079963e-06, + "loss": 0.3745, + "step": 25859 + }, + { + "epoch": 0.5692273827867048, + "grad_norm": 1.2392886772056158, + "learning_rate": 4.1263770027266766e-06, + "loss": 0.2606, + "step": 25860 + }, + { + "epoch": 0.5692493946731235, + "grad_norm": 1.3818837879597619, + "learning_rate": 4.1260260318135345e-06, + "loss": 0.3137, + "step": 25861 + }, + { + "epoch": 0.5692714065595421, + "grad_norm": 1.6231502641254072, + "learning_rate": 4.125675065342315e-06, + "loss": 0.4433, + "step": 25862 + }, + { + "epoch": 0.5692934184459608, + "grad_norm": 1.5647190895781071, + "learning_rate": 4.125324103314807e-06, + "loss": 0.3112, + "step": 25863 + }, + { + "epoch": 0.5693154303323795, + "grad_norm": 1.3796967043556732, + "learning_rate": 4.124973145732792e-06, + "loss": 0.3357, + "step": 25864 + }, + { + "epoch": 0.5693374422187981, + "grad_norm": 1.3626504069227674, + "learning_rate": 4.124622192598052e-06, + "loss": 0.3001, + "step": 25865 + }, + { + "epoch": 0.5693594541052168, + "grad_norm": 1.2362515690527893, + "learning_rate": 4.124271243912375e-06, + "loss": 0.273, + "step": 25866 + }, + { + "epoch": 0.5693814659916355, + "grad_norm": 1.418760738862208, + "learning_rate": 4.123920299677541e-06, + "loss": 0.303, + "step": 25867 + }, + { + "epoch": 0.5694034778780541, + "grad_norm": 1.5067954386185893, + "learning_rate": 4.123569359895334e-06, + "loss": 0.3949, + "step": 25868 + }, + { + "epoch": 0.5694254897644728, + "grad_norm": 2.1246529365045093, + "learning_rate": 4.123218424567539e-06, + "loss": 0.2741, + "step": 25869 + }, + { + "epoch": 0.5694475016508915, + "grad_norm": 1.507673182742027, + "learning_rate": 4.122867493695938e-06, + "loss": 0.3559, + "step": 25870 + }, + { + "epoch": 0.5694695135373101, + "grad_norm": 1.3451396015550672, + "learning_rate": 4.122516567282318e-06, + "loss": 0.3012, + "step": 25871 + }, + { + "epoch": 0.5694915254237288, + "grad_norm": 1.9518942260432148, + "learning_rate": 4.122165645328459e-06, + "loss": 0.3703, + "step": 25872 + }, + { + "epoch": 0.5695135373101475, + "grad_norm": 2.816243484028221, + "learning_rate": 4.121814727836145e-06, + "loss": 0.2222, + "step": 25873 + }, + { + "epoch": 0.5695355491965661, + "grad_norm": 1.575091968444406, + "learning_rate": 4.121463814807162e-06, + "loss": 0.31, + "step": 25874 + }, + { + "epoch": 0.5695575610829848, + "grad_norm": 1.4089306131425559, + "learning_rate": 4.1211129062432884e-06, + "loss": 0.2962, + "step": 25875 + }, + { + "epoch": 0.5695795729694034, + "grad_norm": 1.2095985284702018, + "learning_rate": 4.120762002146314e-06, + "loss": 0.2347, + "step": 25876 + }, + { + "epoch": 0.5696015848558221, + "grad_norm": 1.603311664575749, + "learning_rate": 4.1204111025180185e-06, + "loss": 0.2999, + "step": 25877 + }, + { + "epoch": 0.5696235967422408, + "grad_norm": 1.7705741598049893, + "learning_rate": 4.1200602073601854e-06, + "loss": 0.3323, + "step": 25878 + }, + { + "epoch": 0.5696456086286594, + "grad_norm": 1.3611531981946114, + "learning_rate": 4.119709316674601e-06, + "loss": 0.3305, + "step": 25879 + }, + { + "epoch": 0.5696676205150781, + "grad_norm": 1.6707901283785402, + "learning_rate": 4.119358430463042e-06, + "loss": 0.3812, + "step": 25880 + }, + { + "epoch": 0.5696896324014968, + "grad_norm": 1.7153200300607134, + "learning_rate": 4.1190075487273e-06, + "loss": 0.2621, + "step": 25881 + }, + { + "epoch": 0.5697116442879154, + "grad_norm": 1.6215192153857503, + "learning_rate": 4.118656671469155e-06, + "loss": 0.3201, + "step": 25882 + }, + { + "epoch": 0.5697336561743341, + "grad_norm": 1.180403193884069, + "learning_rate": 4.118305798690387e-06, + "loss": 0.3163, + "step": 25883 + }, + { + "epoch": 0.5697556680607528, + "grad_norm": 1.202110131499194, + "learning_rate": 4.117954930392784e-06, + "loss": 0.2216, + "step": 25884 + }, + { + "epoch": 0.5697776799471714, + "grad_norm": 1.2096856287043263, + "learning_rate": 4.1176040665781266e-06, + "loss": 0.2903, + "step": 25885 + }, + { + "epoch": 0.5697996918335901, + "grad_norm": 0.6839852900667671, + "learning_rate": 4.1172532072482e-06, + "loss": 0.4967, + "step": 25886 + }, + { + "epoch": 0.5698217037200088, + "grad_norm": 2.1414190806545865, + "learning_rate": 4.116902352404786e-06, + "loss": 0.323, + "step": 25887 + }, + { + "epoch": 0.5698437156064274, + "grad_norm": 1.3637219887599674, + "learning_rate": 4.116551502049667e-06, + "loss": 0.2999, + "step": 25888 + }, + { + "epoch": 0.5698657274928461, + "grad_norm": 1.00464718519552, + "learning_rate": 4.116200656184629e-06, + "loss": 0.2481, + "step": 25889 + }, + { + "epoch": 0.5698877393792648, + "grad_norm": 3.550664027018793, + "learning_rate": 4.115849814811453e-06, + "loss": 0.2877, + "step": 25890 + }, + { + "epoch": 0.5699097512656834, + "grad_norm": 1.4016201506560342, + "learning_rate": 4.115498977931921e-06, + "loss": 0.284, + "step": 25891 + }, + { + "epoch": 0.5699317631521021, + "grad_norm": 1.4168457781246613, + "learning_rate": 4.1151481455478205e-06, + "loss": 0.2653, + "step": 25892 + }, + { + "epoch": 0.5699537750385208, + "grad_norm": 1.3577754498952586, + "learning_rate": 4.114797317660928e-06, + "loss": 0.2776, + "step": 25893 + }, + { + "epoch": 0.5699757869249394, + "grad_norm": 0.6144720847808158, + "learning_rate": 4.114446494273034e-06, + "loss": 0.4831, + "step": 25894 + }, + { + "epoch": 0.5699977988113581, + "grad_norm": 0.6248874541008332, + "learning_rate": 4.1140956753859165e-06, + "loss": 0.4686, + "step": 25895 + }, + { + "epoch": 0.5700198106977769, + "grad_norm": 1.4488279037454026, + "learning_rate": 4.113744861001359e-06, + "loss": 0.3283, + "step": 25896 + }, + { + "epoch": 0.5700418225841954, + "grad_norm": 1.344495675127111, + "learning_rate": 4.113394051121148e-06, + "loss": 0.3926, + "step": 25897 + }, + { + "epoch": 0.5700638344706142, + "grad_norm": 1.24739863596684, + "learning_rate": 4.11304324574706e-06, + "loss": 0.3073, + "step": 25898 + }, + { + "epoch": 0.5700858463570327, + "grad_norm": 5.335171489935009, + "learning_rate": 4.112692444880884e-06, + "loss": 0.2651, + "step": 25899 + }, + { + "epoch": 0.5701078582434514, + "grad_norm": 1.2627098524388813, + "learning_rate": 4.112341648524401e-06, + "loss": 0.2951, + "step": 25900 + }, + { + "epoch": 0.5701298701298702, + "grad_norm": 1.8781526922804954, + "learning_rate": 4.111990856679393e-06, + "loss": 0.4345, + "step": 25901 + }, + { + "epoch": 0.5701518820162887, + "grad_norm": 0.6228937421586159, + "learning_rate": 4.1116400693476445e-06, + "loss": 0.4914, + "step": 25902 + }, + { + "epoch": 0.5701738939027075, + "grad_norm": 2.09752620759767, + "learning_rate": 4.111289286530936e-06, + "loss": 0.2825, + "step": 25903 + }, + { + "epoch": 0.5701959057891262, + "grad_norm": 1.5724193510417068, + "learning_rate": 4.110938508231053e-06, + "loss": 0.2913, + "step": 25904 + }, + { + "epoch": 0.5702179176755447, + "grad_norm": 1.5378572604779979, + "learning_rate": 4.110587734449776e-06, + "loss": 0.279, + "step": 25905 + }, + { + "epoch": 0.5702399295619635, + "grad_norm": 2.175472985896711, + "learning_rate": 4.1102369651888886e-06, + "loss": 0.3566, + "step": 25906 + }, + { + "epoch": 0.5702619414483822, + "grad_norm": 1.2892133382222764, + "learning_rate": 4.109886200450175e-06, + "loss": 0.3189, + "step": 25907 + }, + { + "epoch": 0.5702839533348008, + "grad_norm": 1.7386573223370465, + "learning_rate": 4.109535440235417e-06, + "loss": 0.2525, + "step": 25908 + }, + { + "epoch": 0.5703059652212195, + "grad_norm": 1.5114144440794344, + "learning_rate": 4.109184684546395e-06, + "loss": 0.251, + "step": 25909 + }, + { + "epoch": 0.5703279771076382, + "grad_norm": 1.2067259449666898, + "learning_rate": 4.1088339333848965e-06, + "loss": 0.2807, + "step": 25910 + }, + { + "epoch": 0.5703499889940568, + "grad_norm": 1.556736838294911, + "learning_rate": 4.1084831867526975e-06, + "loss": 0.2472, + "step": 25911 + }, + { + "epoch": 0.5703720008804755, + "grad_norm": 0.9887596156821644, + "learning_rate": 4.108132444651589e-06, + "loss": 0.2339, + "step": 25912 + }, + { + "epoch": 0.5703940127668942, + "grad_norm": 1.297167161142025, + "learning_rate": 4.107781707083347e-06, + "loss": 0.3445, + "step": 25913 + }, + { + "epoch": 0.5704160246533128, + "grad_norm": 1.4188214420554326, + "learning_rate": 4.107430974049755e-06, + "loss": 0.3052, + "step": 25914 + }, + { + "epoch": 0.5704380365397315, + "grad_norm": 1.545046583355383, + "learning_rate": 4.1070802455525995e-06, + "loss": 0.3126, + "step": 25915 + }, + { + "epoch": 0.5704600484261502, + "grad_norm": 1.1927282664765202, + "learning_rate": 4.106729521593657e-06, + "loss": 0.3015, + "step": 25916 + }, + { + "epoch": 0.5704820603125688, + "grad_norm": 1.7893308971925912, + "learning_rate": 4.106378802174718e-06, + "loss": 0.2962, + "step": 25917 + }, + { + "epoch": 0.5705040721989875, + "grad_norm": 1.8297543450229217, + "learning_rate": 4.1060280872975585e-06, + "loss": 0.328, + "step": 25918 + }, + { + "epoch": 0.5705260840854062, + "grad_norm": 1.4252132754522477, + "learning_rate": 4.105677376963961e-06, + "loss": 0.3135, + "step": 25919 + }, + { + "epoch": 0.5705480959718248, + "grad_norm": 2.4693781722291863, + "learning_rate": 4.105326671175714e-06, + "loss": 0.2898, + "step": 25920 + }, + { + "epoch": 0.5705701078582435, + "grad_norm": 1.3016112927245094, + "learning_rate": 4.1049759699345925e-06, + "loss": 0.285, + "step": 25921 + }, + { + "epoch": 0.5705921197446621, + "grad_norm": 1.6225221800017429, + "learning_rate": 4.104625273242383e-06, + "loss": 0.3007, + "step": 25922 + }, + { + "epoch": 0.5706141316310808, + "grad_norm": 1.3884715866407225, + "learning_rate": 4.104274581100868e-06, + "loss": 0.3168, + "step": 25923 + }, + { + "epoch": 0.5706361435174995, + "grad_norm": 1.2187283067678008, + "learning_rate": 4.103923893511828e-06, + "loss": 0.324, + "step": 25924 + }, + { + "epoch": 0.5706581554039181, + "grad_norm": 1.2974836897090534, + "learning_rate": 4.103573210477049e-06, + "loss": 0.325, + "step": 25925 + }, + { + "epoch": 0.5706801672903368, + "grad_norm": 0.6614603082396067, + "learning_rate": 4.103222531998308e-06, + "loss": 0.475, + "step": 25926 + }, + { + "epoch": 0.5707021791767555, + "grad_norm": 1.8263638196323122, + "learning_rate": 4.10287185807739e-06, + "loss": 0.3305, + "step": 25927 + }, + { + "epoch": 0.5707241910631741, + "grad_norm": 1.6936320853997031, + "learning_rate": 4.10252118871608e-06, + "loss": 0.3495, + "step": 25928 + }, + { + "epoch": 0.5707462029495928, + "grad_norm": 1.7248905854090464, + "learning_rate": 4.1021705239161544e-06, + "loss": 0.2931, + "step": 25929 + }, + { + "epoch": 0.5707682148360115, + "grad_norm": 1.8281068564506857, + "learning_rate": 4.101819863679401e-06, + "loss": 0.4056, + "step": 25930 + }, + { + "epoch": 0.5707902267224301, + "grad_norm": 1.3979843413004436, + "learning_rate": 4.101469208007599e-06, + "loss": 0.296, + "step": 25931 + }, + { + "epoch": 0.5708122386088488, + "grad_norm": 2.2856448940446703, + "learning_rate": 4.10111855690253e-06, + "loss": 0.3633, + "step": 25932 + }, + { + "epoch": 0.5708342504952675, + "grad_norm": 1.3357389325152942, + "learning_rate": 4.10076791036598e-06, + "loss": 0.2861, + "step": 25933 + }, + { + "epoch": 0.5708562623816861, + "grad_norm": 1.2593290230222778, + "learning_rate": 4.100417268399725e-06, + "loss": 0.2922, + "step": 25934 + }, + { + "epoch": 0.5708782742681048, + "grad_norm": 1.7046505972995982, + "learning_rate": 4.100066631005554e-06, + "loss": 0.3302, + "step": 25935 + }, + { + "epoch": 0.5709002861545235, + "grad_norm": 1.5234929222406706, + "learning_rate": 4.099715998185245e-06, + "loss": 0.3956, + "step": 25936 + }, + { + "epoch": 0.5709222980409421, + "grad_norm": 0.5919532787573843, + "learning_rate": 4.0993653699405786e-06, + "loss": 0.4616, + "step": 25937 + }, + { + "epoch": 0.5709443099273608, + "grad_norm": 1.3435679987211995, + "learning_rate": 4.099014746273342e-06, + "loss": 0.3988, + "step": 25938 + }, + { + "epoch": 0.5709663218137795, + "grad_norm": 1.3888632341259624, + "learning_rate": 4.098664127185312e-06, + "loss": 0.2496, + "step": 25939 + }, + { + "epoch": 0.5709883337001981, + "grad_norm": 1.4500962698738855, + "learning_rate": 4.0983135126782744e-06, + "loss": 0.3703, + "step": 25940 + }, + { + "epoch": 0.5710103455866168, + "grad_norm": 1.3961010092211872, + "learning_rate": 4.09796290275401e-06, + "loss": 0.2967, + "step": 25941 + }, + { + "epoch": 0.5710323574730355, + "grad_norm": 1.6443029693065385, + "learning_rate": 4.097612297414299e-06, + "loss": 0.2697, + "step": 25942 + }, + { + "epoch": 0.5710543693594541, + "grad_norm": 1.0704886851697197, + "learning_rate": 4.097261696660926e-06, + "loss": 0.2231, + "step": 25943 + }, + { + "epoch": 0.5710763812458728, + "grad_norm": 1.4194797304473952, + "learning_rate": 4.096911100495671e-06, + "loss": 0.3, + "step": 25944 + }, + { + "epoch": 0.5710983931322914, + "grad_norm": 1.8099971929104812, + "learning_rate": 4.096560508920317e-06, + "loss": 0.3044, + "step": 25945 + }, + { + "epoch": 0.5711204050187101, + "grad_norm": 2.119449672345513, + "learning_rate": 4.096209921936645e-06, + "loss": 0.3266, + "step": 25946 + }, + { + "epoch": 0.5711424169051288, + "grad_norm": 1.2350151845117385, + "learning_rate": 4.095859339546436e-06, + "loss": 0.2777, + "step": 25947 + }, + { + "epoch": 0.5711644287915474, + "grad_norm": 1.7347989906757026, + "learning_rate": 4.0955087617514765e-06, + "loss": 0.3882, + "step": 25948 + }, + { + "epoch": 0.5711864406779661, + "grad_norm": 1.4973354050411336, + "learning_rate": 4.095158188553543e-06, + "loss": 0.2746, + "step": 25949 + }, + { + "epoch": 0.5712084525643848, + "grad_norm": 1.1732125338285926, + "learning_rate": 4.094807619954418e-06, + "loss": 0.2827, + "step": 25950 + }, + { + "epoch": 0.5712304644508034, + "grad_norm": 1.6524748808565044, + "learning_rate": 4.094457055955887e-06, + "loss": 0.376, + "step": 25951 + }, + { + "epoch": 0.5712524763372221, + "grad_norm": 0.6367030736135173, + "learning_rate": 4.094106496559726e-06, + "loss": 0.4629, + "step": 25952 + }, + { + "epoch": 0.5712744882236408, + "grad_norm": 1.4817891098833333, + "learning_rate": 4.0937559417677225e-06, + "loss": 0.3084, + "step": 25953 + }, + { + "epoch": 0.5712965001100594, + "grad_norm": 1.4138459179732608, + "learning_rate": 4.093405391581654e-06, + "loss": 0.3291, + "step": 25954 + }, + { + "epoch": 0.5713185119964781, + "grad_norm": 1.542149591914878, + "learning_rate": 4.093054846003304e-06, + "loss": 0.2504, + "step": 25955 + }, + { + "epoch": 0.5713405238828968, + "grad_norm": 1.710946308733332, + "learning_rate": 4.092704305034455e-06, + "loss": 0.309, + "step": 25956 + }, + { + "epoch": 0.5713625357693154, + "grad_norm": 1.313109246252494, + "learning_rate": 4.092353768676885e-06, + "loss": 0.266, + "step": 25957 + }, + { + "epoch": 0.5713845476557341, + "grad_norm": 1.6232318651793132, + "learning_rate": 4.092003236932379e-06, + "loss": 0.3411, + "step": 25958 + }, + { + "epoch": 0.5714065595421528, + "grad_norm": 0.6343360782427617, + "learning_rate": 4.091652709802718e-06, + "loss": 0.4862, + "step": 25959 + }, + { + "epoch": 0.5714285714285714, + "grad_norm": 1.3575881218815427, + "learning_rate": 4.091302187289682e-06, + "loss": 0.3159, + "step": 25960 + }, + { + "epoch": 0.5714505833149901, + "grad_norm": 1.3469078557316994, + "learning_rate": 4.090951669395056e-06, + "loss": 0.323, + "step": 25961 + }, + { + "epoch": 0.5714725952014088, + "grad_norm": 1.3988041709006704, + "learning_rate": 4.090601156120617e-06, + "loss": 0.3126, + "step": 25962 + }, + { + "epoch": 0.5714946070878274, + "grad_norm": 1.3199465849575434, + "learning_rate": 4.090250647468149e-06, + "loss": 0.2798, + "step": 25963 + }, + { + "epoch": 0.5715166189742461, + "grad_norm": 0.9341813648123419, + "learning_rate": 4.0899001434394345e-06, + "loss": 0.1865, + "step": 25964 + }, + { + "epoch": 0.5715386308606648, + "grad_norm": 0.627607352372689, + "learning_rate": 4.08954964403625e-06, + "loss": 0.455, + "step": 25965 + }, + { + "epoch": 0.5715606427470834, + "grad_norm": 1.6137321241494893, + "learning_rate": 4.089199149260383e-06, + "loss": 0.3526, + "step": 25966 + }, + { + "epoch": 0.5715826546335021, + "grad_norm": 0.631540445972741, + "learning_rate": 4.088848659113612e-06, + "loss": 0.4874, + "step": 25967 + }, + { + "epoch": 0.5716046665199208, + "grad_norm": 5.953341649817451, + "learning_rate": 4.088498173597716e-06, + "loss": 0.3182, + "step": 25968 + }, + { + "epoch": 0.5716266784063394, + "grad_norm": 1.412350898868204, + "learning_rate": 4.088147692714482e-06, + "loss": 0.2837, + "step": 25969 + }, + { + "epoch": 0.5716486902927581, + "grad_norm": 1.1282446679716969, + "learning_rate": 4.087797216465686e-06, + "loss": 0.2852, + "step": 25970 + }, + { + "epoch": 0.5716707021791767, + "grad_norm": 1.5995742100773855, + "learning_rate": 4.087446744853112e-06, + "loss": 0.3211, + "step": 25971 + }, + { + "epoch": 0.5716927140655954, + "grad_norm": 1.061472208076118, + "learning_rate": 4.087096277878542e-06, + "loss": 0.3029, + "step": 25972 + }, + { + "epoch": 0.5717147259520141, + "grad_norm": 1.4796356483765272, + "learning_rate": 4.086745815543754e-06, + "loss": 0.4032, + "step": 25973 + }, + { + "epoch": 0.5717367378384327, + "grad_norm": 1.8531085380478827, + "learning_rate": 4.086395357850532e-06, + "loss": 0.3796, + "step": 25974 + }, + { + "epoch": 0.5717587497248514, + "grad_norm": 0.6253405879517043, + "learning_rate": 4.086044904800656e-06, + "loss": 0.4937, + "step": 25975 + }, + { + "epoch": 0.5717807616112701, + "grad_norm": 1.3064501095805214, + "learning_rate": 4.085694456395907e-06, + "loss": 0.2994, + "step": 25976 + }, + { + "epoch": 0.5718027734976887, + "grad_norm": 1.6137223495537094, + "learning_rate": 4.085344012638067e-06, + "loss": 0.3766, + "step": 25977 + }, + { + "epoch": 0.5718247853841074, + "grad_norm": 1.6497973241443638, + "learning_rate": 4.084993573528916e-06, + "loss": 0.3071, + "step": 25978 + }, + { + "epoch": 0.5718467972705261, + "grad_norm": 1.4840998178890503, + "learning_rate": 4.084643139070238e-06, + "loss": 0.3637, + "step": 25979 + }, + { + "epoch": 0.5718688091569447, + "grad_norm": 1.3192216356184059, + "learning_rate": 4.08429270926381e-06, + "loss": 0.308, + "step": 25980 + }, + { + "epoch": 0.5718908210433634, + "grad_norm": 1.5520004142483552, + "learning_rate": 4.083942284111415e-06, + "loss": 0.3173, + "step": 25981 + }, + { + "epoch": 0.5719128329297821, + "grad_norm": 0.5947985192037354, + "learning_rate": 4.083591863614835e-06, + "loss": 0.4553, + "step": 25982 + }, + { + "epoch": 0.5719348448162007, + "grad_norm": 1.493498079160975, + "learning_rate": 4.083241447775847e-06, + "loss": 0.3066, + "step": 25983 + }, + { + "epoch": 0.5719568567026194, + "grad_norm": 1.36109694211395, + "learning_rate": 4.082891036596238e-06, + "loss": 0.2009, + "step": 25984 + }, + { + "epoch": 0.5719788685890381, + "grad_norm": 1.2689213278570903, + "learning_rate": 4.0825406300777835e-06, + "loss": 0.2492, + "step": 25985 + }, + { + "epoch": 0.5720008804754567, + "grad_norm": 1.590050845223479, + "learning_rate": 4.082190228222268e-06, + "loss": 0.3404, + "step": 25986 + }, + { + "epoch": 0.5720228923618754, + "grad_norm": 1.8253480519882972, + "learning_rate": 4.081839831031472e-06, + "loss": 0.2725, + "step": 25987 + }, + { + "epoch": 0.5720449042482941, + "grad_norm": 1.572206248193786, + "learning_rate": 4.081489438507174e-06, + "loss": 0.2624, + "step": 25988 + }, + { + "epoch": 0.5720669161347127, + "grad_norm": 1.344143171936232, + "learning_rate": 4.081139050651156e-06, + "loss": 0.3052, + "step": 25989 + }, + { + "epoch": 0.5720889280211314, + "grad_norm": 1.198246911328093, + "learning_rate": 4.080788667465201e-06, + "loss": 0.2588, + "step": 25990 + }, + { + "epoch": 0.5721109399075501, + "grad_norm": 1.3845472942675494, + "learning_rate": 4.080438288951086e-06, + "loss": 0.4093, + "step": 25991 + }, + { + "epoch": 0.5721329517939687, + "grad_norm": 1.723938882767543, + "learning_rate": 4.080087915110596e-06, + "loss": 0.3913, + "step": 25992 + }, + { + "epoch": 0.5721549636803874, + "grad_norm": 2.330741304775987, + "learning_rate": 4.079737545945507e-06, + "loss": 0.3519, + "step": 25993 + }, + { + "epoch": 0.572176975566806, + "grad_norm": 1.4729237493217264, + "learning_rate": 4.079387181457604e-06, + "loss": 0.2967, + "step": 25994 + }, + { + "epoch": 0.5721989874532247, + "grad_norm": 0.6166845562967214, + "learning_rate": 4.079036821648665e-06, + "loss": 0.4477, + "step": 25995 + }, + { + "epoch": 0.5722209993396434, + "grad_norm": 1.262110591669821, + "learning_rate": 4.0786864665204715e-06, + "loss": 0.2842, + "step": 25996 + }, + { + "epoch": 0.572243011226062, + "grad_norm": 1.456329092849414, + "learning_rate": 4.078336116074805e-06, + "loss": 0.3525, + "step": 25997 + }, + { + "epoch": 0.5722650231124807, + "grad_norm": 0.6428191330762162, + "learning_rate": 4.077985770313445e-06, + "loss": 0.4878, + "step": 25998 + }, + { + "epoch": 0.5722870349988994, + "grad_norm": 0.635059365412004, + "learning_rate": 4.077635429238173e-06, + "loss": 0.4695, + "step": 25999 + }, + { + "epoch": 0.572309046885318, + "grad_norm": 0.62739915923003, + "learning_rate": 4.0772850928507705e-06, + "loss": 0.48, + "step": 26000 + }, + { + "epoch": 0.5723310587717367, + "grad_norm": 1.5933435181703317, + "learning_rate": 4.076934761153012e-06, + "loss": 0.3693, + "step": 26001 + }, + { + "epoch": 0.5723530706581554, + "grad_norm": 1.449676731161963, + "learning_rate": 4.076584434146688e-06, + "loss": 0.3614, + "step": 26002 + }, + { + "epoch": 0.572375082544574, + "grad_norm": 1.3512091148680254, + "learning_rate": 4.07623411183357e-06, + "loss": 0.2807, + "step": 26003 + }, + { + "epoch": 0.5723970944309927, + "grad_norm": 0.606772877238532, + "learning_rate": 4.075883794215444e-06, + "loss": 0.4736, + "step": 26004 + }, + { + "epoch": 0.5724191063174114, + "grad_norm": 1.3870368604867458, + "learning_rate": 4.075533481294091e-06, + "loss": 0.2588, + "step": 26005 + }, + { + "epoch": 0.57244111820383, + "grad_norm": 3.0167547652148894, + "learning_rate": 4.075183173071285e-06, + "loss": 0.2678, + "step": 26006 + }, + { + "epoch": 0.5724631300902487, + "grad_norm": 1.9387533993001536, + "learning_rate": 4.074832869548813e-06, + "loss": 0.3915, + "step": 26007 + }, + { + "epoch": 0.5724851419766674, + "grad_norm": 1.4466870597618902, + "learning_rate": 4.0744825707284515e-06, + "loss": 0.2356, + "step": 26008 + }, + { + "epoch": 0.572507153863086, + "grad_norm": 0.6423048206128522, + "learning_rate": 4.0741322766119824e-06, + "loss": 0.4708, + "step": 26009 + }, + { + "epoch": 0.5725291657495047, + "grad_norm": 1.735808119560169, + "learning_rate": 4.073781987201189e-06, + "loss": 0.353, + "step": 26010 + }, + { + "epoch": 0.5725511776359234, + "grad_norm": 1.326847459716784, + "learning_rate": 4.073431702497845e-06, + "loss": 0.2984, + "step": 26011 + }, + { + "epoch": 0.572573189522342, + "grad_norm": 0.643317738747117, + "learning_rate": 4.073081422503736e-06, + "loss": 0.5005, + "step": 26012 + }, + { + "epoch": 0.5725952014087607, + "grad_norm": 1.8550185789388915, + "learning_rate": 4.072731147220641e-06, + "loss": 0.3283, + "step": 26013 + }, + { + "epoch": 0.5726172132951795, + "grad_norm": 1.984998373606749, + "learning_rate": 4.072380876650338e-06, + "loss": 0.2691, + "step": 26014 + }, + { + "epoch": 0.572639225181598, + "grad_norm": 1.3056447500174635, + "learning_rate": 4.072030610794612e-06, + "loss": 0.3162, + "step": 26015 + }, + { + "epoch": 0.5726612370680167, + "grad_norm": 1.5670185865501711, + "learning_rate": 4.071680349655238e-06, + "loss": 0.3435, + "step": 26016 + }, + { + "epoch": 0.5726832489544353, + "grad_norm": 1.687610197795782, + "learning_rate": 4.071330093233999e-06, + "loss": 0.4271, + "step": 26017 + }, + { + "epoch": 0.572705260840854, + "grad_norm": 1.944194791196964, + "learning_rate": 4.070979841532676e-06, + "loss": 0.3182, + "step": 26018 + }, + { + "epoch": 0.5727272727272728, + "grad_norm": 0.6244096424305161, + "learning_rate": 4.070629594553045e-06, + "loss": 0.4706, + "step": 26019 + }, + { + "epoch": 0.5727492846136913, + "grad_norm": 1.74451981856742, + "learning_rate": 4.070279352296892e-06, + "loss": 0.2598, + "step": 26020 + }, + { + "epoch": 0.57277129650011, + "grad_norm": 1.297971458198817, + "learning_rate": 4.0699291147659915e-06, + "loss": 0.2193, + "step": 26021 + }, + { + "epoch": 0.5727933083865288, + "grad_norm": 0.6174923451399899, + "learning_rate": 4.069578881962128e-06, + "loss": 0.4752, + "step": 26022 + }, + { + "epoch": 0.5728153202729473, + "grad_norm": 0.6343936032657735, + "learning_rate": 4.069228653887079e-06, + "loss": 0.4843, + "step": 26023 + }, + { + "epoch": 0.572837332159366, + "grad_norm": 1.2125920759351216, + "learning_rate": 4.0688784305426246e-06, + "loss": 0.1763, + "step": 26024 + }, + { + "epoch": 0.5728593440457848, + "grad_norm": 1.3537401817565804, + "learning_rate": 4.068528211930546e-06, + "loss": 0.2801, + "step": 26025 + }, + { + "epoch": 0.5728813559322034, + "grad_norm": 1.3504975201921459, + "learning_rate": 4.068177998052621e-06, + "loss": 0.2758, + "step": 26026 + }, + { + "epoch": 0.5729033678186221, + "grad_norm": 1.3275595668705198, + "learning_rate": 4.067827788910632e-06, + "loss": 0.2971, + "step": 26027 + }, + { + "epoch": 0.5729253797050408, + "grad_norm": 1.5318561154978974, + "learning_rate": 4.067477584506359e-06, + "loss": 0.4096, + "step": 26028 + }, + { + "epoch": 0.5729473915914594, + "grad_norm": 1.7207240424497523, + "learning_rate": 4.0671273848415795e-06, + "loss": 0.3344, + "step": 26029 + }, + { + "epoch": 0.5729694034778781, + "grad_norm": 1.471044842966772, + "learning_rate": 4.066777189918075e-06, + "loss": 0.3301, + "step": 26030 + }, + { + "epoch": 0.5729914153642968, + "grad_norm": 1.3788771798483517, + "learning_rate": 4.066426999737627e-06, + "loss": 0.3378, + "step": 26031 + }, + { + "epoch": 0.5730134272507154, + "grad_norm": 1.4208120973790672, + "learning_rate": 4.0660768143020095e-06, + "loss": 0.2374, + "step": 26032 + }, + { + "epoch": 0.5730354391371341, + "grad_norm": 0.6212837933609293, + "learning_rate": 4.065726633613008e-06, + "loss": 0.4777, + "step": 26033 + }, + { + "epoch": 0.5730574510235528, + "grad_norm": 1.7460018445569074, + "learning_rate": 4.065376457672401e-06, + "loss": 0.3987, + "step": 26034 + }, + { + "epoch": 0.5730794629099714, + "grad_norm": 2.802511316555719, + "learning_rate": 4.065026286481967e-06, + "loss": 0.229, + "step": 26035 + }, + { + "epoch": 0.5731014747963901, + "grad_norm": 0.6227958409655967, + "learning_rate": 4.064676120043488e-06, + "loss": 0.4444, + "step": 26036 + }, + { + "epoch": 0.5731234866828088, + "grad_norm": 1.344442379248822, + "learning_rate": 4.064325958358739e-06, + "loss": 0.2611, + "step": 26037 + }, + { + "epoch": 0.5731454985692274, + "grad_norm": 1.2818318659928545, + "learning_rate": 4.063975801429506e-06, + "loss": 0.2531, + "step": 26038 + }, + { + "epoch": 0.5731675104556461, + "grad_norm": 1.318670924833526, + "learning_rate": 4.063625649257563e-06, + "loss": 0.341, + "step": 26039 + }, + { + "epoch": 0.5731895223420647, + "grad_norm": 4.539213303436551, + "learning_rate": 4.063275501844693e-06, + "loss": 0.3065, + "step": 26040 + }, + { + "epoch": 0.5732115342284834, + "grad_norm": 1.5654005736751546, + "learning_rate": 4.062925359192676e-06, + "loss": 0.349, + "step": 26041 + }, + { + "epoch": 0.5732335461149021, + "grad_norm": 2.2833533096926613, + "learning_rate": 4.062575221303289e-06, + "loss": 0.3484, + "step": 26042 + }, + { + "epoch": 0.5732555580013207, + "grad_norm": 1.7643235903325598, + "learning_rate": 4.062225088178313e-06, + "loss": 0.3224, + "step": 26043 + }, + { + "epoch": 0.5732775698877394, + "grad_norm": 1.7038039697069378, + "learning_rate": 4.061874959819526e-06, + "loss": 0.3772, + "step": 26044 + }, + { + "epoch": 0.5732995817741581, + "grad_norm": 0.625713649823949, + "learning_rate": 4.061524836228711e-06, + "loss": 0.4648, + "step": 26045 + }, + { + "epoch": 0.5733215936605767, + "grad_norm": 1.6581351122511014, + "learning_rate": 4.0611747174076455e-06, + "loss": 0.1784, + "step": 26046 + }, + { + "epoch": 0.5733436055469954, + "grad_norm": 1.658116293571733, + "learning_rate": 4.0608246033581075e-06, + "loss": 0.2886, + "step": 26047 + }, + { + "epoch": 0.5733656174334141, + "grad_norm": 2.475363194191243, + "learning_rate": 4.060474494081878e-06, + "loss": 0.3454, + "step": 26048 + }, + { + "epoch": 0.5733876293198327, + "grad_norm": 1.7345271134225526, + "learning_rate": 4.060124389580737e-06, + "loss": 0.3078, + "step": 26049 + }, + { + "epoch": 0.5734096412062514, + "grad_norm": 1.3979067833753254, + "learning_rate": 4.05977428985646e-06, + "loss": 0.2905, + "step": 26050 + }, + { + "epoch": 0.5734316530926701, + "grad_norm": 1.5790963210678206, + "learning_rate": 4.059424194910834e-06, + "loss": 0.3242, + "step": 26051 + }, + { + "epoch": 0.5734536649790887, + "grad_norm": 1.351768397689848, + "learning_rate": 4.05907410474563e-06, + "loss": 0.2619, + "step": 26052 + }, + { + "epoch": 0.5734756768655074, + "grad_norm": 1.3881448294503118, + "learning_rate": 4.058724019362633e-06, + "loss": 0.3494, + "step": 26053 + }, + { + "epoch": 0.5734976887519261, + "grad_norm": 0.6544458039758176, + "learning_rate": 4.0583739387636215e-06, + "loss": 0.4937, + "step": 26054 + }, + { + "epoch": 0.5735197006383447, + "grad_norm": 0.652406669860696, + "learning_rate": 4.05802386295037e-06, + "loss": 0.4682, + "step": 26055 + }, + { + "epoch": 0.5735417125247634, + "grad_norm": 1.6458102586646939, + "learning_rate": 4.057673791924665e-06, + "loss": 0.3719, + "step": 26056 + }, + { + "epoch": 0.5735637244111821, + "grad_norm": 1.3223920257310806, + "learning_rate": 4.057323725688279e-06, + "loss": 0.3306, + "step": 26057 + }, + { + "epoch": 0.5735857362976007, + "grad_norm": 1.6045453070508042, + "learning_rate": 4.0569736642429964e-06, + "loss": 0.3227, + "step": 26058 + }, + { + "epoch": 0.5736077481840194, + "grad_norm": 1.2964815445747524, + "learning_rate": 4.056623607590595e-06, + "loss": 0.282, + "step": 26059 + }, + { + "epoch": 0.5736297600704381, + "grad_norm": 3.775829041821452, + "learning_rate": 4.056273555732851e-06, + "loss": 0.3725, + "step": 26060 + }, + { + "epoch": 0.5736517719568567, + "grad_norm": 0.6311650882434929, + "learning_rate": 4.055923508671546e-06, + "loss": 0.4928, + "step": 26061 + }, + { + "epoch": 0.5736737838432754, + "grad_norm": 1.1434204716000358, + "learning_rate": 4.055573466408459e-06, + "loss": 0.3142, + "step": 26062 + }, + { + "epoch": 0.573695795729694, + "grad_norm": 1.2657482740179629, + "learning_rate": 4.055223428945369e-06, + "loss": 0.2892, + "step": 26063 + }, + { + "epoch": 0.5737178076161127, + "grad_norm": 2.158829003144811, + "learning_rate": 4.054873396284056e-06, + "loss": 0.251, + "step": 26064 + }, + { + "epoch": 0.5737398195025314, + "grad_norm": 0.6087192643698679, + "learning_rate": 4.054523368426296e-06, + "loss": 0.461, + "step": 26065 + }, + { + "epoch": 0.57376183138895, + "grad_norm": 1.1809244522863376, + "learning_rate": 4.054173345373872e-06, + "loss": 0.294, + "step": 26066 + }, + { + "epoch": 0.5737838432753687, + "grad_norm": 1.3205552029457637, + "learning_rate": 4.053823327128561e-06, + "loss": 0.2626, + "step": 26067 + }, + { + "epoch": 0.5738058551617874, + "grad_norm": 1.4014947507405338, + "learning_rate": 4.053473313692138e-06, + "loss": 0.4167, + "step": 26068 + }, + { + "epoch": 0.573827867048206, + "grad_norm": 1.8127673146113064, + "learning_rate": 4.05312330506639e-06, + "loss": 0.3084, + "step": 26069 + }, + { + "epoch": 0.5738498789346247, + "grad_norm": 1.5535701729059272, + "learning_rate": 4.052773301253089e-06, + "loss": 0.2951, + "step": 26070 + }, + { + "epoch": 0.5738718908210434, + "grad_norm": 2.0226891591140177, + "learning_rate": 4.052423302254018e-06, + "loss": 0.2977, + "step": 26071 + }, + { + "epoch": 0.573893902707462, + "grad_norm": 1.4639774035058644, + "learning_rate": 4.052073308070956e-06, + "loss": 0.3423, + "step": 26072 + }, + { + "epoch": 0.5739159145938807, + "grad_norm": 1.8888026315998572, + "learning_rate": 4.051723318705676e-06, + "loss": 0.3693, + "step": 26073 + }, + { + "epoch": 0.5739379264802994, + "grad_norm": 1.4731028516529046, + "learning_rate": 4.051373334159965e-06, + "loss": 0.3319, + "step": 26074 + }, + { + "epoch": 0.573959938366718, + "grad_norm": 1.5329786637501228, + "learning_rate": 4.051023354435595e-06, + "loss": 0.3355, + "step": 26075 + }, + { + "epoch": 0.5739819502531367, + "grad_norm": 1.8322126674367292, + "learning_rate": 4.050673379534349e-06, + "loss": 0.2506, + "step": 26076 + }, + { + "epoch": 0.5740039621395554, + "grad_norm": 2.015310137443821, + "learning_rate": 4.050323409458006e-06, + "loss": 0.2613, + "step": 26077 + }, + { + "epoch": 0.574025974025974, + "grad_norm": 1.254417854648138, + "learning_rate": 4.04997344420834e-06, + "loss": 0.2546, + "step": 26078 + }, + { + "epoch": 0.5740479859123927, + "grad_norm": 1.1878458692955451, + "learning_rate": 4.049623483787133e-06, + "loss": 0.2288, + "step": 26079 + }, + { + "epoch": 0.5740699977988114, + "grad_norm": 0.6406832966651858, + "learning_rate": 4.0492735281961635e-06, + "loss": 0.486, + "step": 26080 + }, + { + "epoch": 0.57409200968523, + "grad_norm": 1.5733064505804009, + "learning_rate": 4.048923577437211e-06, + "loss": 0.2806, + "step": 26081 + }, + { + "epoch": 0.5741140215716487, + "grad_norm": 1.312926066008043, + "learning_rate": 4.048573631512054e-06, + "loss": 0.2968, + "step": 26082 + }, + { + "epoch": 0.5741360334580674, + "grad_norm": 2.1048322504586032, + "learning_rate": 4.0482236904224685e-06, + "loss": 0.3727, + "step": 26083 + }, + { + "epoch": 0.574158045344486, + "grad_norm": 1.0798566692841813, + "learning_rate": 4.047873754170234e-06, + "loss": 0.3038, + "step": 26084 + }, + { + "epoch": 0.5741800572309047, + "grad_norm": 1.3898511618768719, + "learning_rate": 4.047523822757131e-06, + "loss": 0.2788, + "step": 26085 + }, + { + "epoch": 0.5742020691173233, + "grad_norm": 1.3320169391913763, + "learning_rate": 4.047173896184936e-06, + "loss": 0.2709, + "step": 26086 + }, + { + "epoch": 0.574224081003742, + "grad_norm": 1.3387216703605, + "learning_rate": 4.04682397445543e-06, + "loss": 0.2551, + "step": 26087 + }, + { + "epoch": 0.5742460928901607, + "grad_norm": 1.7599211718756873, + "learning_rate": 4.046474057570387e-06, + "loss": 0.3058, + "step": 26088 + }, + { + "epoch": 0.5742681047765793, + "grad_norm": 0.6127942682981372, + "learning_rate": 4.04612414553159e-06, + "loss": 0.4455, + "step": 26089 + }, + { + "epoch": 0.574290116662998, + "grad_norm": 0.5998239757100021, + "learning_rate": 4.045774238340817e-06, + "loss": 0.4647, + "step": 26090 + }, + { + "epoch": 0.5743121285494167, + "grad_norm": 1.2714570198950939, + "learning_rate": 4.04542433599984e-06, + "loss": 0.3978, + "step": 26091 + }, + { + "epoch": 0.5743341404358353, + "grad_norm": 1.7235232661457633, + "learning_rate": 4.045074438510448e-06, + "loss": 0.3276, + "step": 26092 + }, + { + "epoch": 0.574356152322254, + "grad_norm": 1.679347546475105, + "learning_rate": 4.044724545874409e-06, + "loss": 0.4196, + "step": 26093 + }, + { + "epoch": 0.5743781642086727, + "grad_norm": 1.9760194487632172, + "learning_rate": 4.044374658093508e-06, + "loss": 0.3659, + "step": 26094 + }, + { + "epoch": 0.5744001760950913, + "grad_norm": 1.5839976222052123, + "learning_rate": 4.044024775169523e-06, + "loss": 0.386, + "step": 26095 + }, + { + "epoch": 0.57442218798151, + "grad_norm": 1.5337499942991148, + "learning_rate": 4.043674897104227e-06, + "loss": 0.3728, + "step": 26096 + }, + { + "epoch": 0.5744441998679287, + "grad_norm": 1.4787993944939057, + "learning_rate": 4.043325023899404e-06, + "loss": 0.3388, + "step": 26097 + }, + { + "epoch": 0.5744662117543473, + "grad_norm": 0.6504688251071821, + "learning_rate": 4.0429751555568275e-06, + "loss": 0.4839, + "step": 26098 + }, + { + "epoch": 0.574488223640766, + "grad_norm": 1.556787704738624, + "learning_rate": 4.04262529207828e-06, + "loss": 0.3472, + "step": 26099 + }, + { + "epoch": 0.5745102355271847, + "grad_norm": 3.9744333467323125, + "learning_rate": 4.042275433465539e-06, + "loss": 0.2537, + "step": 26100 + }, + { + "epoch": 0.5745322474136033, + "grad_norm": 2.667297034647091, + "learning_rate": 4.041925579720379e-06, + "loss": 0.2928, + "step": 26101 + }, + { + "epoch": 0.574554259300022, + "grad_norm": 0.6396801211309547, + "learning_rate": 4.041575730844581e-06, + "loss": 0.5, + "step": 26102 + }, + { + "epoch": 0.5745762711864407, + "grad_norm": 0.6261184750337836, + "learning_rate": 4.0412258868399226e-06, + "loss": 0.4867, + "step": 26103 + }, + { + "epoch": 0.5745982830728593, + "grad_norm": 0.619281851480475, + "learning_rate": 4.040876047708182e-06, + "loss": 0.4603, + "step": 26104 + }, + { + "epoch": 0.574620294959278, + "grad_norm": 0.5978386524296001, + "learning_rate": 4.040526213451139e-06, + "loss": 0.4502, + "step": 26105 + }, + { + "epoch": 0.5746423068456967, + "grad_norm": 0.5892598281385751, + "learning_rate": 4.040176384070567e-06, + "loss": 0.4356, + "step": 26106 + }, + { + "epoch": 0.5746643187321153, + "grad_norm": 1.0521931105886384, + "learning_rate": 4.039826559568248e-06, + "loss": 0.252, + "step": 26107 + }, + { + "epoch": 0.574686330618534, + "grad_norm": 2.310445835078973, + "learning_rate": 4.03947673994596e-06, + "loss": 0.32, + "step": 26108 + }, + { + "epoch": 0.5747083425049527, + "grad_norm": 7.402618346807968, + "learning_rate": 4.039126925205476e-06, + "loss": 0.3688, + "step": 26109 + }, + { + "epoch": 0.5747303543913713, + "grad_norm": 0.633120729539556, + "learning_rate": 4.0387771153485815e-06, + "loss": 0.4537, + "step": 26110 + }, + { + "epoch": 0.57475236627779, + "grad_norm": 1.6647912619233007, + "learning_rate": 4.038427310377047e-06, + "loss": 0.3623, + "step": 26111 + }, + { + "epoch": 0.5747743781642086, + "grad_norm": 1.1770950132575797, + "learning_rate": 4.038077510292655e-06, + "loss": 0.3025, + "step": 26112 + }, + { + "epoch": 0.5747963900506273, + "grad_norm": 1.940901228102241, + "learning_rate": 4.0377277150971845e-06, + "loss": 0.4061, + "step": 26113 + }, + { + "epoch": 0.574818401937046, + "grad_norm": 1.3617819521326289, + "learning_rate": 4.037377924792407e-06, + "loss": 0.3205, + "step": 26114 + }, + { + "epoch": 0.5748404138234646, + "grad_norm": 0.6159758364835134, + "learning_rate": 4.037028139380107e-06, + "loss": 0.4523, + "step": 26115 + }, + { + "epoch": 0.5748624257098833, + "grad_norm": 1.8113110562923795, + "learning_rate": 4.036678358862058e-06, + "loss": 0.2945, + "step": 26116 + }, + { + "epoch": 0.574884437596302, + "grad_norm": 1.3369996913289608, + "learning_rate": 4.0363285832400405e-06, + "loss": 0.3099, + "step": 26117 + }, + { + "epoch": 0.5749064494827206, + "grad_norm": 1.7037474997761612, + "learning_rate": 4.035978812515831e-06, + "loss": 0.2838, + "step": 26118 + }, + { + "epoch": 0.5749284613691393, + "grad_norm": 2.1287546249373914, + "learning_rate": 4.035629046691205e-06, + "loss": 0.2302, + "step": 26119 + }, + { + "epoch": 0.574950473255558, + "grad_norm": 1.515048023854152, + "learning_rate": 4.035279285767945e-06, + "loss": 0.3237, + "step": 26120 + }, + { + "epoch": 0.5749724851419766, + "grad_norm": 2.534162801219787, + "learning_rate": 4.034929529747823e-06, + "loss": 0.3586, + "step": 26121 + }, + { + "epoch": 0.5749944970283953, + "grad_norm": 2.171018977061617, + "learning_rate": 4.034579778632621e-06, + "loss": 0.4542, + "step": 26122 + }, + { + "epoch": 0.575016508914814, + "grad_norm": 0.594101176496039, + "learning_rate": 4.034230032424117e-06, + "loss": 0.4281, + "step": 26123 + }, + { + "epoch": 0.5750385208012326, + "grad_norm": 0.6128952200284974, + "learning_rate": 4.033880291124085e-06, + "loss": 0.4476, + "step": 26124 + }, + { + "epoch": 0.5750605326876513, + "grad_norm": 1.5286075127333743, + "learning_rate": 4.033530554734304e-06, + "loss": 0.2658, + "step": 26125 + }, + { + "epoch": 0.57508254457407, + "grad_norm": 2.3206741637362187, + "learning_rate": 4.033180823256554e-06, + "loss": 0.333, + "step": 26126 + }, + { + "epoch": 0.5751045564604886, + "grad_norm": 1.459543498999898, + "learning_rate": 4.032831096692606e-06, + "loss": 0.3607, + "step": 26127 + }, + { + "epoch": 0.5751265683469073, + "grad_norm": 1.4513530844314613, + "learning_rate": 4.0324813750442454e-06, + "loss": 0.2777, + "step": 26128 + }, + { + "epoch": 0.575148580233326, + "grad_norm": 6.3380622430567986, + "learning_rate": 4.032131658313244e-06, + "loss": 0.2924, + "step": 26129 + }, + { + "epoch": 0.5751705921197446, + "grad_norm": 1.9599854173795186, + "learning_rate": 4.031781946501382e-06, + "loss": 0.2426, + "step": 26130 + }, + { + "epoch": 0.5751926040061633, + "grad_norm": 1.4898266694682825, + "learning_rate": 4.031432239610437e-06, + "loss": 0.3211, + "step": 26131 + }, + { + "epoch": 0.575214615892582, + "grad_norm": 4.118940987696487, + "learning_rate": 4.031082537642184e-06, + "loss": 0.3629, + "step": 26132 + }, + { + "epoch": 0.5752366277790006, + "grad_norm": 1.3273036451591613, + "learning_rate": 4.030732840598402e-06, + "loss": 0.2601, + "step": 26133 + }, + { + "epoch": 0.5752586396654193, + "grad_norm": 1.7181302654280444, + "learning_rate": 4.030383148480866e-06, + "loss": 0.3581, + "step": 26134 + }, + { + "epoch": 0.5752806515518379, + "grad_norm": 1.5432087351190917, + "learning_rate": 4.030033461291359e-06, + "loss": 0.3426, + "step": 26135 + }, + { + "epoch": 0.5753026634382566, + "grad_norm": 1.6678386777232779, + "learning_rate": 4.029683779031653e-06, + "loss": 0.2304, + "step": 26136 + }, + { + "epoch": 0.5753246753246753, + "grad_norm": 0.6310078980453134, + "learning_rate": 4.029334101703526e-06, + "loss": 0.4784, + "step": 26137 + }, + { + "epoch": 0.5753466872110939, + "grad_norm": 1.2227063749049714, + "learning_rate": 4.028984429308756e-06, + "loss": 0.2942, + "step": 26138 + }, + { + "epoch": 0.5753686990975126, + "grad_norm": 1.346107385522678, + "learning_rate": 4.028634761849121e-06, + "loss": 0.3044, + "step": 26139 + }, + { + "epoch": 0.5753907109839314, + "grad_norm": 1.8003595966004733, + "learning_rate": 4.028285099326397e-06, + "loss": 0.3173, + "step": 26140 + }, + { + "epoch": 0.57541272287035, + "grad_norm": 1.592683213747821, + "learning_rate": 4.0279354417423635e-06, + "loss": 0.3389, + "step": 26141 + }, + { + "epoch": 0.5754347347567687, + "grad_norm": 1.6452892012463156, + "learning_rate": 4.027585789098793e-06, + "loss": 0.2796, + "step": 26142 + }, + { + "epoch": 0.5754567466431874, + "grad_norm": 0.6110836934188308, + "learning_rate": 4.027236141397466e-06, + "loss": 0.4539, + "step": 26143 + }, + { + "epoch": 0.575478758529606, + "grad_norm": 2.1963999869901762, + "learning_rate": 4.026886498640158e-06, + "loss": 0.2628, + "step": 26144 + }, + { + "epoch": 0.5755007704160247, + "grad_norm": 1.8004461222339845, + "learning_rate": 4.026536860828648e-06, + "loss": 0.3793, + "step": 26145 + }, + { + "epoch": 0.5755227823024434, + "grad_norm": 1.87799077388617, + "learning_rate": 4.026187227964713e-06, + "loss": 0.2415, + "step": 26146 + }, + { + "epoch": 0.575544794188862, + "grad_norm": 1.8595019601705558, + "learning_rate": 4.025837600050126e-06, + "loss": 0.3293, + "step": 26147 + }, + { + "epoch": 0.5755668060752807, + "grad_norm": 1.192533936758505, + "learning_rate": 4.02548797708667e-06, + "loss": 0.282, + "step": 26148 + }, + { + "epoch": 0.5755888179616994, + "grad_norm": 1.3621097444353154, + "learning_rate": 4.0251383590761175e-06, + "loss": 0.2855, + "step": 26149 + }, + { + "epoch": 0.575610829848118, + "grad_norm": 3.689956662516299, + "learning_rate": 4.024788746020245e-06, + "loss": 0.35, + "step": 26150 + }, + { + "epoch": 0.5756328417345367, + "grad_norm": 1.3490397342765903, + "learning_rate": 4.024439137920832e-06, + "loss": 0.318, + "step": 26151 + }, + { + "epoch": 0.5756548536209554, + "grad_norm": 1.2207766628506094, + "learning_rate": 4.0240895347796545e-06, + "loss": 0.3096, + "step": 26152 + }, + { + "epoch": 0.575676865507374, + "grad_norm": 1.2252799070183011, + "learning_rate": 4.02373993659849e-06, + "loss": 0.3659, + "step": 26153 + }, + { + "epoch": 0.5756988773937927, + "grad_norm": 0.6174244568710281, + "learning_rate": 4.023390343379115e-06, + "loss": 0.4616, + "step": 26154 + }, + { + "epoch": 0.5757208892802114, + "grad_norm": 0.6220915558655931, + "learning_rate": 4.023040755123303e-06, + "loss": 0.4825, + "step": 26155 + }, + { + "epoch": 0.57574290116663, + "grad_norm": 0.6331933982705578, + "learning_rate": 4.0226911718328366e-06, + "loss": 0.4761, + "step": 26156 + }, + { + "epoch": 0.5757649130530487, + "grad_norm": 1.999131537926539, + "learning_rate": 4.022341593509487e-06, + "loss": 0.2977, + "step": 26157 + }, + { + "epoch": 0.5757869249394673, + "grad_norm": 1.279313272008948, + "learning_rate": 4.021992020155036e-06, + "loss": 0.308, + "step": 26158 + }, + { + "epoch": 0.575808936825886, + "grad_norm": 0.6353727769452021, + "learning_rate": 4.021642451771257e-06, + "loss": 0.4517, + "step": 26159 + }, + { + "epoch": 0.5758309487123047, + "grad_norm": 0.6191560174564784, + "learning_rate": 4.021292888359926e-06, + "loss": 0.4656, + "step": 26160 + }, + { + "epoch": 0.5758529605987233, + "grad_norm": 1.7512665069561355, + "learning_rate": 4.020943329922823e-06, + "loss": 0.2924, + "step": 26161 + }, + { + "epoch": 0.575874972485142, + "grad_norm": 0.6232445442401268, + "learning_rate": 4.020593776461721e-06, + "loss": 0.4881, + "step": 26162 + }, + { + "epoch": 0.5758969843715607, + "grad_norm": 1.523224231121135, + "learning_rate": 4.0202442279784e-06, + "loss": 0.4035, + "step": 26163 + }, + { + "epoch": 0.5759189962579793, + "grad_norm": 2.2049730736365403, + "learning_rate": 4.019894684474634e-06, + "loss": 0.3253, + "step": 26164 + }, + { + "epoch": 0.575941008144398, + "grad_norm": 1.9817354463068064, + "learning_rate": 4.019545145952199e-06, + "loss": 0.394, + "step": 26165 + }, + { + "epoch": 0.5759630200308167, + "grad_norm": 1.1623989090662141, + "learning_rate": 4.0191956124128745e-06, + "loss": 0.2377, + "step": 26166 + }, + { + "epoch": 0.5759850319172353, + "grad_norm": 1.5417656258513268, + "learning_rate": 4.018846083858436e-06, + "loss": 0.3253, + "step": 26167 + }, + { + "epoch": 0.576007043803654, + "grad_norm": 1.3400322726612335, + "learning_rate": 4.018496560290658e-06, + "loss": 0.2222, + "step": 26168 + }, + { + "epoch": 0.5760290556900727, + "grad_norm": 1.417667005704872, + "learning_rate": 4.0181470417113175e-06, + "loss": 0.3568, + "step": 26169 + }, + { + "epoch": 0.5760510675764913, + "grad_norm": 1.2231357056289416, + "learning_rate": 4.017797528122192e-06, + "loss": 0.2919, + "step": 26170 + }, + { + "epoch": 0.57607307946291, + "grad_norm": 1.7061128449974226, + "learning_rate": 4.0174480195250584e-06, + "loss": 0.3152, + "step": 26171 + }, + { + "epoch": 0.5760950913493287, + "grad_norm": 1.2084521060524418, + "learning_rate": 4.017098515921694e-06, + "loss": 0.339, + "step": 26172 + }, + { + "epoch": 0.5761171032357473, + "grad_norm": 1.8319753021172382, + "learning_rate": 4.016749017313869e-06, + "loss": 0.3544, + "step": 26173 + }, + { + "epoch": 0.576139115122166, + "grad_norm": 1.4036288971480544, + "learning_rate": 4.016399523703367e-06, + "loss": 0.2768, + "step": 26174 + }, + { + "epoch": 0.5761611270085847, + "grad_norm": 0.6269118381804162, + "learning_rate": 4.01605003509196e-06, + "loss": 0.4766, + "step": 26175 + }, + { + "epoch": 0.5761831388950033, + "grad_norm": 1.2555142516211608, + "learning_rate": 4.015700551481426e-06, + "loss": 0.3043, + "step": 26176 + }, + { + "epoch": 0.576205150781422, + "grad_norm": 1.050841938365177, + "learning_rate": 4.015351072873543e-06, + "loss": 0.2559, + "step": 26177 + }, + { + "epoch": 0.5762271626678407, + "grad_norm": 1.3914645151547018, + "learning_rate": 4.015001599270082e-06, + "loss": 0.3153, + "step": 26178 + }, + { + "epoch": 0.5762491745542593, + "grad_norm": 1.6484463593890648, + "learning_rate": 4.014652130672823e-06, + "loss": 0.3511, + "step": 26179 + }, + { + "epoch": 0.576271186440678, + "grad_norm": 2.198686812130103, + "learning_rate": 4.014302667083541e-06, + "loss": 0.3287, + "step": 26180 + }, + { + "epoch": 0.5762931983270966, + "grad_norm": 1.6684193527743523, + "learning_rate": 4.013953208504014e-06, + "loss": 0.343, + "step": 26181 + }, + { + "epoch": 0.5763152102135153, + "grad_norm": 1.4527842603382675, + "learning_rate": 4.013603754936016e-06, + "loss": 0.2454, + "step": 26182 + }, + { + "epoch": 0.576337222099934, + "grad_norm": 1.5136911603626577, + "learning_rate": 4.013254306381323e-06, + "loss": 0.3033, + "step": 26183 + }, + { + "epoch": 0.5763592339863526, + "grad_norm": 1.4674061355398775, + "learning_rate": 4.012904862841711e-06, + "loss": 0.335, + "step": 26184 + }, + { + "epoch": 0.5763812458727713, + "grad_norm": 1.2994061773846537, + "learning_rate": 4.0125554243189575e-06, + "loss": 0.2887, + "step": 26185 + }, + { + "epoch": 0.57640325775919, + "grad_norm": 1.3810069544843404, + "learning_rate": 4.012205990814839e-06, + "loss": 0.3112, + "step": 26186 + }, + { + "epoch": 0.5764252696456086, + "grad_norm": 1.3574677797003876, + "learning_rate": 4.011856562331129e-06, + "loss": 0.2912, + "step": 26187 + }, + { + "epoch": 0.5764472815320273, + "grad_norm": 1.3673898495382126, + "learning_rate": 4.0115071388696046e-06, + "loss": 0.3518, + "step": 26188 + }, + { + "epoch": 0.576469293418446, + "grad_norm": 1.5978211568236924, + "learning_rate": 4.0111577204320425e-06, + "loss": 0.3452, + "step": 26189 + }, + { + "epoch": 0.5764913053048646, + "grad_norm": 1.60088455890838, + "learning_rate": 4.010808307020218e-06, + "loss": 0.3626, + "step": 26190 + }, + { + "epoch": 0.5765133171912833, + "grad_norm": 1.4465815445741341, + "learning_rate": 4.010458898635906e-06, + "loss": 0.3335, + "step": 26191 + }, + { + "epoch": 0.576535329077702, + "grad_norm": 6.160088165767851, + "learning_rate": 4.010109495280884e-06, + "loss": 0.3312, + "step": 26192 + }, + { + "epoch": 0.5765573409641206, + "grad_norm": 1.4629578572872466, + "learning_rate": 4.0097600969569255e-06, + "loss": 0.3595, + "step": 26193 + }, + { + "epoch": 0.5765793528505393, + "grad_norm": 1.4677939592316165, + "learning_rate": 4.009410703665809e-06, + "loss": 0.3896, + "step": 26194 + }, + { + "epoch": 0.576601364736958, + "grad_norm": 1.738718791622497, + "learning_rate": 4.009061315409311e-06, + "loss": 0.3006, + "step": 26195 + }, + { + "epoch": 0.5766233766233766, + "grad_norm": 1.3913438499400228, + "learning_rate": 4.008711932189202e-06, + "loss": 0.3239, + "step": 26196 + }, + { + "epoch": 0.5766453885097953, + "grad_norm": 4.925749134707641, + "learning_rate": 4.008362554007263e-06, + "loss": 0.325, + "step": 26197 + }, + { + "epoch": 0.576667400396214, + "grad_norm": 0.6239389832928233, + "learning_rate": 4.0080131808652666e-06, + "loss": 0.4817, + "step": 26198 + }, + { + "epoch": 0.5766894122826326, + "grad_norm": 1.9618496974397215, + "learning_rate": 4.007663812764991e-06, + "loss": 0.3412, + "step": 26199 + }, + { + "epoch": 0.5767114241690513, + "grad_norm": 2.1214607274001587, + "learning_rate": 4.007314449708211e-06, + "loss": 0.2529, + "step": 26200 + }, + { + "epoch": 0.57673343605547, + "grad_norm": 1.922386439865847, + "learning_rate": 4.006965091696699e-06, + "loss": 0.3373, + "step": 26201 + }, + { + "epoch": 0.5767554479418886, + "grad_norm": 1.5197337041557795, + "learning_rate": 4.0066157387322354e-06, + "loss": 0.2859, + "step": 26202 + }, + { + "epoch": 0.5767774598283073, + "grad_norm": 1.8917703421361927, + "learning_rate": 4.006266390816592e-06, + "loss": 0.3839, + "step": 26203 + }, + { + "epoch": 0.5767994717147259, + "grad_norm": 1.4834335705824435, + "learning_rate": 4.005917047951549e-06, + "loss": 0.3257, + "step": 26204 + }, + { + "epoch": 0.5768214836011446, + "grad_norm": 1.4889356514584273, + "learning_rate": 4.005567710138877e-06, + "loss": 0.2684, + "step": 26205 + }, + { + "epoch": 0.5768434954875633, + "grad_norm": 1.1944815412128162, + "learning_rate": 4.005218377380353e-06, + "loss": 0.3208, + "step": 26206 + }, + { + "epoch": 0.5768655073739819, + "grad_norm": 1.7344396833174296, + "learning_rate": 4.004869049677753e-06, + "loss": 0.3859, + "step": 26207 + }, + { + "epoch": 0.5768875192604006, + "grad_norm": 1.658911500245017, + "learning_rate": 4.004519727032854e-06, + "loss": 0.2871, + "step": 26208 + }, + { + "epoch": 0.5769095311468193, + "grad_norm": 1.4307917074065528, + "learning_rate": 4.0041704094474285e-06, + "loss": 0.2945, + "step": 26209 + }, + { + "epoch": 0.5769315430332379, + "grad_norm": 2.237402284440963, + "learning_rate": 4.003821096923252e-06, + "loss": 0.3309, + "step": 26210 + }, + { + "epoch": 0.5769535549196566, + "grad_norm": 1.5663103675535668, + "learning_rate": 4.003471789462102e-06, + "loss": 0.263, + "step": 26211 + }, + { + "epoch": 0.5769755668060753, + "grad_norm": 0.5968689508251637, + "learning_rate": 4.003122487065754e-06, + "loss": 0.5048, + "step": 26212 + }, + { + "epoch": 0.5769975786924939, + "grad_norm": 1.8010198701000444, + "learning_rate": 4.002773189735982e-06, + "loss": 0.3443, + "step": 26213 + }, + { + "epoch": 0.5770195905789126, + "grad_norm": 1.0224275611539435, + "learning_rate": 4.002423897474559e-06, + "loss": 0.2424, + "step": 26214 + }, + { + "epoch": 0.5770416024653313, + "grad_norm": 1.3834010159027383, + "learning_rate": 4.002074610283265e-06, + "loss": 0.3734, + "step": 26215 + }, + { + "epoch": 0.5770636143517499, + "grad_norm": 1.3404112143693838, + "learning_rate": 4.001725328163871e-06, + "loss": 0.2675, + "step": 26216 + }, + { + "epoch": 0.5770856262381686, + "grad_norm": 0.5964937959801774, + "learning_rate": 4.001376051118157e-06, + "loss": 0.4164, + "step": 26217 + }, + { + "epoch": 0.5771076381245873, + "grad_norm": 1.405819814149312, + "learning_rate": 4.001026779147894e-06, + "loss": 0.2424, + "step": 26218 + }, + { + "epoch": 0.5771296500110059, + "grad_norm": 1.4401985288469488, + "learning_rate": 4.000677512254857e-06, + "loss": 0.293, + "step": 26219 + }, + { + "epoch": 0.5771516618974246, + "grad_norm": 1.711657463865814, + "learning_rate": 4.000328250440823e-06, + "loss": 0.2495, + "step": 26220 + }, + { + "epoch": 0.5771736737838433, + "grad_norm": 1.2788376812332996, + "learning_rate": 3.9999789937075674e-06, + "loss": 0.279, + "step": 26221 + }, + { + "epoch": 0.5771956856702619, + "grad_norm": 1.2105233123762882, + "learning_rate": 3.999629742056865e-06, + "loss": 0.2664, + "step": 26222 + }, + { + "epoch": 0.5772176975566806, + "grad_norm": 1.8699804707676124, + "learning_rate": 3.999280495490491e-06, + "loss": 0.3446, + "step": 26223 + }, + { + "epoch": 0.5772397094430993, + "grad_norm": 1.7213288045861115, + "learning_rate": 3.998931254010218e-06, + "loss": 0.2561, + "step": 26224 + }, + { + "epoch": 0.5772617213295179, + "grad_norm": 0.627677974877664, + "learning_rate": 3.998582017617824e-06, + "loss": 0.4688, + "step": 26225 + }, + { + "epoch": 0.5772837332159366, + "grad_norm": 1.3926854785416392, + "learning_rate": 3.998232786315083e-06, + "loss": 0.3202, + "step": 26226 + }, + { + "epoch": 0.5773057451023552, + "grad_norm": 1.0539950377682268, + "learning_rate": 3.997883560103768e-06, + "loss": 0.2671, + "step": 26227 + }, + { + "epoch": 0.5773277569887739, + "grad_norm": 1.261288914858343, + "learning_rate": 3.997534338985658e-06, + "loss": 0.2698, + "step": 26228 + }, + { + "epoch": 0.5773497688751926, + "grad_norm": 1.6423885670472331, + "learning_rate": 3.997185122962524e-06, + "loss": 0.3539, + "step": 26229 + }, + { + "epoch": 0.5773717807616112, + "grad_norm": 1.6859593671901598, + "learning_rate": 3.996835912036143e-06, + "loss": 0.33, + "step": 26230 + }, + { + "epoch": 0.5773937926480299, + "grad_norm": 7.1953296427594875, + "learning_rate": 3.99648670620829e-06, + "loss": 0.2776, + "step": 26231 + }, + { + "epoch": 0.5774158045344486, + "grad_norm": 1.331611900322754, + "learning_rate": 3.996137505480738e-06, + "loss": 0.2894, + "step": 26232 + }, + { + "epoch": 0.5774378164208672, + "grad_norm": 1.180200294839666, + "learning_rate": 3.995788309855263e-06, + "loss": 0.2396, + "step": 26233 + }, + { + "epoch": 0.5774598283072859, + "grad_norm": 1.5244238072861787, + "learning_rate": 3.99543911933364e-06, + "loss": 0.2778, + "step": 26234 + }, + { + "epoch": 0.5774818401937046, + "grad_norm": 1.4174557043437677, + "learning_rate": 3.995089933917645e-06, + "loss": 0.3997, + "step": 26235 + }, + { + "epoch": 0.5775038520801232, + "grad_norm": 7.119888707399769, + "learning_rate": 3.99474075360905e-06, + "loss": 0.3079, + "step": 26236 + }, + { + "epoch": 0.5775258639665419, + "grad_norm": 0.6085371812797756, + "learning_rate": 3.994391578409629e-06, + "loss": 0.4607, + "step": 26237 + }, + { + "epoch": 0.5775478758529606, + "grad_norm": 0.6152269426093223, + "learning_rate": 3.99404240832116e-06, + "loss": 0.4718, + "step": 26238 + }, + { + "epoch": 0.5775698877393792, + "grad_norm": 1.6436151347002863, + "learning_rate": 3.993693243345416e-06, + "loss": 0.3467, + "step": 26239 + }, + { + "epoch": 0.5775918996257979, + "grad_norm": 1.478525793266143, + "learning_rate": 3.993344083484173e-06, + "loss": 0.3173, + "step": 26240 + }, + { + "epoch": 0.5776139115122166, + "grad_norm": 1.42928195908225, + "learning_rate": 3.992994928739204e-06, + "loss": 0.2779, + "step": 26241 + }, + { + "epoch": 0.5776359233986352, + "grad_norm": 2.714452267532044, + "learning_rate": 3.9926457791122825e-06, + "loss": 0.3137, + "step": 26242 + }, + { + "epoch": 0.5776579352850539, + "grad_norm": 1.8290455370033705, + "learning_rate": 3.992296634605185e-06, + "loss": 0.3315, + "step": 26243 + }, + { + "epoch": 0.5776799471714726, + "grad_norm": 1.3905023123331535, + "learning_rate": 3.9919474952196855e-06, + "loss": 0.2996, + "step": 26244 + }, + { + "epoch": 0.5777019590578912, + "grad_norm": 1.8571347651988095, + "learning_rate": 3.991598360957559e-06, + "loss": 0.3266, + "step": 26245 + }, + { + "epoch": 0.5777239709443099, + "grad_norm": 2.0161242753178987, + "learning_rate": 3.99124923182058e-06, + "loss": 0.3335, + "step": 26246 + }, + { + "epoch": 0.5777459828307286, + "grad_norm": 0.6180828658076495, + "learning_rate": 3.990900107810521e-06, + "loss": 0.4659, + "step": 26247 + }, + { + "epoch": 0.5777679947171472, + "grad_norm": 2.56378297423758, + "learning_rate": 3.990550988929157e-06, + "loss": 0.2981, + "step": 26248 + }, + { + "epoch": 0.5777900066035659, + "grad_norm": 1.3976161806025997, + "learning_rate": 3.990201875178265e-06, + "loss": 0.2738, + "step": 26249 + }, + { + "epoch": 0.5778120184899846, + "grad_norm": 0.608541449716901, + "learning_rate": 3.9898527665596165e-06, + "loss": 0.4694, + "step": 26250 + }, + { + "epoch": 0.5778340303764032, + "grad_norm": 0.6197811407214169, + "learning_rate": 3.989503663074987e-06, + "loss": 0.4741, + "step": 26251 + }, + { + "epoch": 0.577856042262822, + "grad_norm": 2.190218182831692, + "learning_rate": 3.989154564726149e-06, + "loss": 0.3447, + "step": 26252 + }, + { + "epoch": 0.5778780541492405, + "grad_norm": 1.3568170921727771, + "learning_rate": 3.9888054715148816e-06, + "loss": 0.3327, + "step": 26253 + }, + { + "epoch": 0.5779000660356592, + "grad_norm": 1.4211300301557257, + "learning_rate": 3.988456383442954e-06, + "loss": 0.3163, + "step": 26254 + }, + { + "epoch": 0.577922077922078, + "grad_norm": 3.832795589208847, + "learning_rate": 3.988107300512142e-06, + "loss": 0.2917, + "step": 26255 + }, + { + "epoch": 0.5779440898084965, + "grad_norm": 1.120968439125918, + "learning_rate": 3.98775822272422e-06, + "loss": 0.3399, + "step": 26256 + }, + { + "epoch": 0.5779661016949152, + "grad_norm": 0.5994655585714724, + "learning_rate": 3.987409150080962e-06, + "loss": 0.4485, + "step": 26257 + }, + { + "epoch": 0.577988113581334, + "grad_norm": 1.323093404953906, + "learning_rate": 3.987060082584143e-06, + "loss": 0.3046, + "step": 26258 + }, + { + "epoch": 0.5780101254677525, + "grad_norm": 1.845508818084944, + "learning_rate": 3.986711020235537e-06, + "loss": 0.3654, + "step": 26259 + }, + { + "epoch": 0.5780321373541712, + "grad_norm": 12.21824183240933, + "learning_rate": 3.986361963036915e-06, + "loss": 0.2913, + "step": 26260 + }, + { + "epoch": 0.57805414924059, + "grad_norm": 2.0897461991510147, + "learning_rate": 3.986012910990056e-06, + "loss": 0.2828, + "step": 26261 + }, + { + "epoch": 0.5780761611270085, + "grad_norm": 1.3046388429651494, + "learning_rate": 3.98566386409673e-06, + "loss": 0.2489, + "step": 26262 + }, + { + "epoch": 0.5780981730134273, + "grad_norm": 1.3120349320240938, + "learning_rate": 3.9853148223587155e-06, + "loss": 0.2576, + "step": 26263 + }, + { + "epoch": 0.578120184899846, + "grad_norm": 0.6028270184898631, + "learning_rate": 3.984965785777781e-06, + "loss": 0.4527, + "step": 26264 + }, + { + "epoch": 0.5781421967862646, + "grad_norm": 1.8098959217941484, + "learning_rate": 3.984616754355703e-06, + "loss": 0.2632, + "step": 26265 + }, + { + "epoch": 0.5781642086726833, + "grad_norm": 1.466338140519855, + "learning_rate": 3.984267728094257e-06, + "loss": 0.2608, + "step": 26266 + }, + { + "epoch": 0.578186220559102, + "grad_norm": 0.602589689747413, + "learning_rate": 3.983918706995216e-06, + "loss": 0.454, + "step": 26267 + }, + { + "epoch": 0.5782082324455206, + "grad_norm": 1.5886412954452724, + "learning_rate": 3.983569691060352e-06, + "loss": 0.2059, + "step": 26268 + }, + { + "epoch": 0.5782302443319393, + "grad_norm": 0.5969931185467272, + "learning_rate": 3.983220680291439e-06, + "loss": 0.4503, + "step": 26269 + }, + { + "epoch": 0.578252256218358, + "grad_norm": 1.4373478912939948, + "learning_rate": 3.982871674690254e-06, + "loss": 0.3114, + "step": 26270 + }, + { + "epoch": 0.5782742681047766, + "grad_norm": 1.7549189062194863, + "learning_rate": 3.982522674258569e-06, + "loss": 0.3041, + "step": 26271 + }, + { + "epoch": 0.5782962799911953, + "grad_norm": 1.8848531799599901, + "learning_rate": 3.9821736789981565e-06, + "loss": 0.3872, + "step": 26272 + }, + { + "epoch": 0.578318291877614, + "grad_norm": 0.6088710190900593, + "learning_rate": 3.98182468891079e-06, + "loss": 0.4694, + "step": 26273 + }, + { + "epoch": 0.5783403037640326, + "grad_norm": 1.2825807925768535, + "learning_rate": 3.9814757039982465e-06, + "loss": 0.3111, + "step": 26274 + }, + { + "epoch": 0.5783623156504513, + "grad_norm": 1.296817196307066, + "learning_rate": 3.981126724262297e-06, + "loss": 0.2763, + "step": 26275 + }, + { + "epoch": 0.5783843275368699, + "grad_norm": 1.364812111597697, + "learning_rate": 3.980777749704718e-06, + "loss": 0.3292, + "step": 26276 + }, + { + "epoch": 0.5784063394232886, + "grad_norm": 1.3352508064495212, + "learning_rate": 3.980428780327279e-06, + "loss": 0.2719, + "step": 26277 + }, + { + "epoch": 0.5784283513097073, + "grad_norm": 1.4299488365815045, + "learning_rate": 3.980079816131755e-06, + "loss": 0.256, + "step": 26278 + }, + { + "epoch": 0.5784503631961259, + "grad_norm": 1.5876608013889746, + "learning_rate": 3.979730857119921e-06, + "loss": 0.2559, + "step": 26279 + }, + { + "epoch": 0.5784723750825446, + "grad_norm": 2.4067414558380626, + "learning_rate": 3.979381903293549e-06, + "loss": 0.3595, + "step": 26280 + }, + { + "epoch": 0.5784943869689633, + "grad_norm": 1.5578007861331162, + "learning_rate": 3.979032954654416e-06, + "loss": 0.3475, + "step": 26281 + }, + { + "epoch": 0.5785163988553819, + "grad_norm": 1.3628290900947115, + "learning_rate": 3.978684011204291e-06, + "loss": 0.327, + "step": 26282 + }, + { + "epoch": 0.5785384107418006, + "grad_norm": 0.623143192277984, + "learning_rate": 3.978335072944949e-06, + "loss": 0.4584, + "step": 26283 + }, + { + "epoch": 0.5785604226282193, + "grad_norm": 1.4349168455262995, + "learning_rate": 3.977986139878165e-06, + "loss": 0.3558, + "step": 26284 + }, + { + "epoch": 0.5785824345146379, + "grad_norm": 1.2633587948034513, + "learning_rate": 3.9776372120057096e-06, + "loss": 0.312, + "step": 26285 + }, + { + "epoch": 0.5786044464010566, + "grad_norm": 1.2511744252160557, + "learning_rate": 3.9772882893293605e-06, + "loss": 0.359, + "step": 26286 + }, + { + "epoch": 0.5786264582874753, + "grad_norm": 1.7796595194695242, + "learning_rate": 3.9769393718508876e-06, + "loss": 0.3551, + "step": 26287 + }, + { + "epoch": 0.5786484701738939, + "grad_norm": 1.1697880715271494, + "learning_rate": 3.976590459572064e-06, + "loss": 0.2623, + "step": 26288 + }, + { + "epoch": 0.5786704820603126, + "grad_norm": 1.4292581294945261, + "learning_rate": 3.976241552494666e-06, + "loss": 0.3095, + "step": 26289 + }, + { + "epoch": 0.5786924939467313, + "grad_norm": 1.3237226487374216, + "learning_rate": 3.975892650620463e-06, + "loss": 0.4044, + "step": 26290 + }, + { + "epoch": 0.5787145058331499, + "grad_norm": 1.3835774041437097, + "learning_rate": 3.97554375395123e-06, + "loss": 0.209, + "step": 26291 + }, + { + "epoch": 0.5787365177195686, + "grad_norm": 0.6094000974109701, + "learning_rate": 3.975194862488743e-06, + "loss": 0.4783, + "step": 26292 + }, + { + "epoch": 0.5787585296059873, + "grad_norm": 2.348760236522269, + "learning_rate": 3.97484597623477e-06, + "loss": 0.24, + "step": 26293 + }, + { + "epoch": 0.5787805414924059, + "grad_norm": 1.9070367747322596, + "learning_rate": 3.97449709519109e-06, + "loss": 0.3406, + "step": 26294 + }, + { + "epoch": 0.5788025533788246, + "grad_norm": 1.6026099333495698, + "learning_rate": 3.974148219359471e-06, + "loss": 0.3351, + "step": 26295 + }, + { + "epoch": 0.5788245652652433, + "grad_norm": 1.624350896868092, + "learning_rate": 3.973799348741688e-06, + "loss": 0.2712, + "step": 26296 + }, + { + "epoch": 0.5788465771516619, + "grad_norm": 2.5081303799497965, + "learning_rate": 3.973450483339516e-06, + "loss": 0.3355, + "step": 26297 + }, + { + "epoch": 0.5788685890380806, + "grad_norm": 1.4749133612029852, + "learning_rate": 3.973101623154724e-06, + "loss": 0.319, + "step": 26298 + }, + { + "epoch": 0.5788906009244992, + "grad_norm": 1.2542049752614, + "learning_rate": 3.972752768189091e-06, + "loss": 0.2969, + "step": 26299 + }, + { + "epoch": 0.5789126128109179, + "grad_norm": 1.7425200784434716, + "learning_rate": 3.972403918444385e-06, + "loss": 0.2781, + "step": 26300 + }, + { + "epoch": 0.5789346246973366, + "grad_norm": 1.885900899940508, + "learning_rate": 3.9720550739223795e-06, + "loss": 0.359, + "step": 26301 + }, + { + "epoch": 0.5789566365837552, + "grad_norm": 2.7618991664621824, + "learning_rate": 3.971706234624849e-06, + "loss": 0.3163, + "step": 26302 + }, + { + "epoch": 0.5789786484701739, + "grad_norm": 2.6165279259914507, + "learning_rate": 3.971357400553567e-06, + "loss": 0.3898, + "step": 26303 + }, + { + "epoch": 0.5790006603565926, + "grad_norm": 1.7940468150510511, + "learning_rate": 3.971008571710306e-06, + "loss": 0.2526, + "step": 26304 + }, + { + "epoch": 0.5790226722430112, + "grad_norm": 1.5494892136004166, + "learning_rate": 3.970659748096838e-06, + "loss": 0.4152, + "step": 26305 + }, + { + "epoch": 0.5790446841294299, + "grad_norm": 1.2456143082057318, + "learning_rate": 3.970310929714935e-06, + "loss": 0.2562, + "step": 26306 + }, + { + "epoch": 0.5790666960158486, + "grad_norm": 1.3418203709602199, + "learning_rate": 3.969962116566372e-06, + "loss": 0.2646, + "step": 26307 + }, + { + "epoch": 0.5790887079022672, + "grad_norm": 0.6275464994011691, + "learning_rate": 3.9696133086529214e-06, + "loss": 0.4716, + "step": 26308 + }, + { + "epoch": 0.5791107197886859, + "grad_norm": 1.9323288195663624, + "learning_rate": 3.969264505976354e-06, + "loss": 0.3147, + "step": 26309 + }, + { + "epoch": 0.5791327316751046, + "grad_norm": 1.8451201678083287, + "learning_rate": 3.968915708538445e-06, + "loss": 0.3215, + "step": 26310 + }, + { + "epoch": 0.5791547435615232, + "grad_norm": 1.3383370874949199, + "learning_rate": 3.968566916340966e-06, + "loss": 0.3334, + "step": 26311 + }, + { + "epoch": 0.5791767554479419, + "grad_norm": 0.6119578403676509, + "learning_rate": 3.968218129385692e-06, + "loss": 0.452, + "step": 26312 + }, + { + "epoch": 0.5791987673343606, + "grad_norm": 1.3227386403731485, + "learning_rate": 3.967869347674391e-06, + "loss": 0.3118, + "step": 26313 + }, + { + "epoch": 0.5792207792207792, + "grad_norm": 1.9054800998288333, + "learning_rate": 3.967520571208838e-06, + "loss": 0.3827, + "step": 26314 + }, + { + "epoch": 0.5792427911071979, + "grad_norm": 3.0222425781168565, + "learning_rate": 3.967171799990808e-06, + "loss": 0.3526, + "step": 26315 + }, + { + "epoch": 0.5792648029936166, + "grad_norm": 1.5677061499504654, + "learning_rate": 3.96682303402207e-06, + "loss": 0.3017, + "step": 26316 + }, + { + "epoch": 0.5792868148800352, + "grad_norm": 1.254494696426197, + "learning_rate": 3.9664742733044e-06, + "loss": 0.3121, + "step": 26317 + }, + { + "epoch": 0.5793088267664539, + "grad_norm": 1.4255045250043819, + "learning_rate": 3.966125517839567e-06, + "loss": 0.2795, + "step": 26318 + }, + { + "epoch": 0.5793308386528726, + "grad_norm": 0.6109187694214951, + "learning_rate": 3.965776767629346e-06, + "loss": 0.4601, + "step": 26319 + }, + { + "epoch": 0.5793528505392912, + "grad_norm": 1.6347797308674674, + "learning_rate": 3.965428022675508e-06, + "loss": 0.306, + "step": 26320 + }, + { + "epoch": 0.5793748624257099, + "grad_norm": 2.4796747979023674, + "learning_rate": 3.965079282979826e-06, + "loss": 0.3565, + "step": 26321 + }, + { + "epoch": 0.5793968743121285, + "grad_norm": 1.0535762041785095, + "learning_rate": 3.9647305485440744e-06, + "loss": 0.2634, + "step": 26322 + }, + { + "epoch": 0.5794188861985472, + "grad_norm": 2.155308777833857, + "learning_rate": 3.964381819370024e-06, + "loss": 0.3082, + "step": 26323 + }, + { + "epoch": 0.5794408980849659, + "grad_norm": 0.6453863051050892, + "learning_rate": 3.964033095459445e-06, + "loss": 0.4883, + "step": 26324 + }, + { + "epoch": 0.5794629099713845, + "grad_norm": 4.472193557168427, + "learning_rate": 3.963684376814114e-06, + "loss": 0.2051, + "step": 26325 + }, + { + "epoch": 0.5794849218578032, + "grad_norm": 1.271534548083205, + "learning_rate": 3.9633356634358e-06, + "loss": 0.2478, + "step": 26326 + }, + { + "epoch": 0.5795069337442219, + "grad_norm": 1.846782623007854, + "learning_rate": 3.962986955326276e-06, + "loss": 0.3721, + "step": 26327 + }, + { + "epoch": 0.5795289456306405, + "grad_norm": 1.465704526748658, + "learning_rate": 3.962638252487315e-06, + "loss": 0.3575, + "step": 26328 + }, + { + "epoch": 0.5795509575170592, + "grad_norm": 1.9702133655684457, + "learning_rate": 3.962289554920689e-06, + "loss": 0.2768, + "step": 26329 + }, + { + "epoch": 0.5795729694034779, + "grad_norm": 1.3950049276787686, + "learning_rate": 3.961940862628172e-06, + "loss": 0.2372, + "step": 26330 + }, + { + "epoch": 0.5795949812898965, + "grad_norm": 1.4482448361243534, + "learning_rate": 3.961592175611534e-06, + "loss": 0.2471, + "step": 26331 + }, + { + "epoch": 0.5796169931763152, + "grad_norm": 1.464651970563222, + "learning_rate": 3.961243493872546e-06, + "loss": 0.3422, + "step": 26332 + }, + { + "epoch": 0.5796390050627339, + "grad_norm": 2.305211057970228, + "learning_rate": 3.960894817412983e-06, + "loss": 0.2821, + "step": 26333 + }, + { + "epoch": 0.5796610169491525, + "grad_norm": 1.4354232035022758, + "learning_rate": 3.960546146234615e-06, + "loss": 0.2301, + "step": 26334 + }, + { + "epoch": 0.5796830288355712, + "grad_norm": 1.5178587118655793, + "learning_rate": 3.960197480339217e-06, + "loss": 0.2688, + "step": 26335 + }, + { + "epoch": 0.5797050407219899, + "grad_norm": 2.2179125217834716, + "learning_rate": 3.959848819728559e-06, + "loss": 0.294, + "step": 26336 + }, + { + "epoch": 0.5797270526084085, + "grad_norm": 1.5916042762266078, + "learning_rate": 3.959500164404411e-06, + "loss": 0.2649, + "step": 26337 + }, + { + "epoch": 0.5797490644948272, + "grad_norm": 1.2458133323448326, + "learning_rate": 3.9591515143685485e-06, + "loss": 0.2378, + "step": 26338 + }, + { + "epoch": 0.5797710763812459, + "grad_norm": 1.6948541281770062, + "learning_rate": 3.958802869622742e-06, + "loss": 0.3562, + "step": 26339 + }, + { + "epoch": 0.5797930882676645, + "grad_norm": 1.2274992429897535, + "learning_rate": 3.958454230168766e-06, + "loss": 0.2868, + "step": 26340 + }, + { + "epoch": 0.5798151001540832, + "grad_norm": 1.359784038361767, + "learning_rate": 3.958105596008387e-06, + "loss": 0.348, + "step": 26341 + }, + { + "epoch": 0.5798371120405019, + "grad_norm": 0.6176756014445538, + "learning_rate": 3.957756967143381e-06, + "loss": 0.4622, + "step": 26342 + }, + { + "epoch": 0.5798591239269205, + "grad_norm": 1.1188922998679012, + "learning_rate": 3.957408343575521e-06, + "loss": 0.2546, + "step": 26343 + }, + { + "epoch": 0.5798811358133392, + "grad_norm": 1.4105058687604888, + "learning_rate": 3.957059725306573e-06, + "loss": 0.2651, + "step": 26344 + }, + { + "epoch": 0.5799031476997578, + "grad_norm": 1.2774772772257232, + "learning_rate": 3.956711112338315e-06, + "loss": 0.3071, + "step": 26345 + }, + { + "epoch": 0.5799251595861765, + "grad_norm": 0.6403401555776125, + "learning_rate": 3.9563625046725165e-06, + "loss": 0.484, + "step": 26346 + }, + { + "epoch": 0.5799471714725952, + "grad_norm": 1.528545703314764, + "learning_rate": 3.956013902310948e-06, + "loss": 0.2892, + "step": 26347 + }, + { + "epoch": 0.5799691833590138, + "grad_norm": 1.6878300913864082, + "learning_rate": 3.9556653052553854e-06, + "loss": 0.3132, + "step": 26348 + }, + { + "epoch": 0.5799911952454325, + "grad_norm": 2.3543400630608162, + "learning_rate": 3.9553167135075955e-06, + "loss": 0.3054, + "step": 26349 + }, + { + "epoch": 0.5800132071318512, + "grad_norm": 1.5811070677170853, + "learning_rate": 3.95496812706935e-06, + "loss": 0.273, + "step": 26350 + }, + { + "epoch": 0.5800352190182698, + "grad_norm": 1.9285981053730161, + "learning_rate": 3.954619545942425e-06, + "loss": 0.2703, + "step": 26351 + }, + { + "epoch": 0.5800572309046885, + "grad_norm": 1.2581286596811407, + "learning_rate": 3.954270970128589e-06, + "loss": 0.2154, + "step": 26352 + }, + { + "epoch": 0.5800792427911072, + "grad_norm": 1.1640447566136976, + "learning_rate": 3.953922399629616e-06, + "loss": 0.2783, + "step": 26353 + }, + { + "epoch": 0.5801012546775258, + "grad_norm": 0.6099439321910749, + "learning_rate": 3.953573834447274e-06, + "loss": 0.4556, + "step": 26354 + }, + { + "epoch": 0.5801232665639445, + "grad_norm": 1.7492189781137664, + "learning_rate": 3.953225274583337e-06, + "loss": 0.2751, + "step": 26355 + }, + { + "epoch": 0.5801452784503632, + "grad_norm": 1.2699207295048451, + "learning_rate": 3.952876720039576e-06, + "loss": 0.3064, + "step": 26356 + }, + { + "epoch": 0.5801672903367818, + "grad_norm": 2.548731937319234, + "learning_rate": 3.952528170817761e-06, + "loss": 0.3034, + "step": 26357 + }, + { + "epoch": 0.5801893022232005, + "grad_norm": 1.6521338097901088, + "learning_rate": 3.952179626919668e-06, + "loss": 0.2869, + "step": 26358 + }, + { + "epoch": 0.5802113141096192, + "grad_norm": 1.2541185895115845, + "learning_rate": 3.951831088347064e-06, + "loss": 0.3426, + "step": 26359 + }, + { + "epoch": 0.5802333259960378, + "grad_norm": 0.5916634614541746, + "learning_rate": 3.951482555101721e-06, + "loss": 0.4598, + "step": 26360 + }, + { + "epoch": 0.5802553378824565, + "grad_norm": 1.0112582169817024, + "learning_rate": 3.951134027185414e-06, + "loss": 0.1694, + "step": 26361 + }, + { + "epoch": 0.5802773497688752, + "grad_norm": 0.6084960775828114, + "learning_rate": 3.950785504599908e-06, + "loss": 0.4555, + "step": 26362 + }, + { + "epoch": 0.5802993616552938, + "grad_norm": 2.7179069880876923, + "learning_rate": 3.950436987346981e-06, + "loss": 0.3538, + "step": 26363 + }, + { + "epoch": 0.5803213735417125, + "grad_norm": 0.6181172816387553, + "learning_rate": 3.950088475428401e-06, + "loss": 0.4699, + "step": 26364 + }, + { + "epoch": 0.5803433854281312, + "grad_norm": 1.9038033797020482, + "learning_rate": 3.9497399688459375e-06, + "loss": 0.323, + "step": 26365 + }, + { + "epoch": 0.5803653973145498, + "grad_norm": 1.2869318883783802, + "learning_rate": 3.949391467601367e-06, + "loss": 0.2585, + "step": 26366 + }, + { + "epoch": 0.5803874092009685, + "grad_norm": 1.4922001561795444, + "learning_rate": 3.949042971696456e-06, + "loss": 0.2696, + "step": 26367 + }, + { + "epoch": 0.5804094210873871, + "grad_norm": 2.507212929653827, + "learning_rate": 3.948694481132976e-06, + "loss": 0.327, + "step": 26368 + }, + { + "epoch": 0.5804314329738058, + "grad_norm": 2.6392230784846853, + "learning_rate": 3.948345995912702e-06, + "loss": 0.2183, + "step": 26369 + }, + { + "epoch": 0.5804534448602245, + "grad_norm": 2.110830699940401, + "learning_rate": 3.947997516037401e-06, + "loss": 0.3458, + "step": 26370 + }, + { + "epoch": 0.5804754567466431, + "grad_norm": 3.598241267462542, + "learning_rate": 3.947649041508847e-06, + "loss": 0.2904, + "step": 26371 + }, + { + "epoch": 0.5804974686330618, + "grad_norm": 0.620379906446684, + "learning_rate": 3.94730057232881e-06, + "loss": 0.4701, + "step": 26372 + }, + { + "epoch": 0.5805194805194805, + "grad_norm": 1.401176322390554, + "learning_rate": 3.946952108499059e-06, + "loss": 0.2872, + "step": 26373 + }, + { + "epoch": 0.5805414924058991, + "grad_norm": 1.24951829148178, + "learning_rate": 3.94660365002137e-06, + "loss": 0.2196, + "step": 26374 + }, + { + "epoch": 0.5805635042923178, + "grad_norm": 1.1319613673869875, + "learning_rate": 3.946255196897508e-06, + "loss": 0.2592, + "step": 26375 + }, + { + "epoch": 0.5805855161787365, + "grad_norm": 2.4551514119356246, + "learning_rate": 3.945906749129249e-06, + "loss": 0.3067, + "step": 26376 + }, + { + "epoch": 0.5806075280651551, + "grad_norm": 0.6227879271355035, + "learning_rate": 3.945558306718362e-06, + "loss": 0.4893, + "step": 26377 + }, + { + "epoch": 0.5806295399515738, + "grad_norm": 1.6757681183080226, + "learning_rate": 3.945209869666617e-06, + "loss": 0.3721, + "step": 26378 + }, + { + "epoch": 0.5806515518379926, + "grad_norm": 0.617607313316515, + "learning_rate": 3.944861437975788e-06, + "loss": 0.476, + "step": 26379 + }, + { + "epoch": 0.5806735637244111, + "grad_norm": 1.2493517283170936, + "learning_rate": 3.94451301164764e-06, + "loss": 0.2772, + "step": 26380 + }, + { + "epoch": 0.5806955756108299, + "grad_norm": 0.6194072290842307, + "learning_rate": 3.944164590683951e-06, + "loss": 0.4712, + "step": 26381 + }, + { + "epoch": 0.5807175874972486, + "grad_norm": 1.2853703378677408, + "learning_rate": 3.943816175086488e-06, + "loss": 0.3193, + "step": 26382 + }, + { + "epoch": 0.5807395993836671, + "grad_norm": 0.621577564121973, + "learning_rate": 3.943467764857021e-06, + "loss": 0.4424, + "step": 26383 + }, + { + "epoch": 0.5807616112700859, + "grad_norm": 1.2794622785136742, + "learning_rate": 3.943119359997324e-06, + "loss": 0.2084, + "step": 26384 + }, + { + "epoch": 0.5807836231565046, + "grad_norm": 1.2764217715329422, + "learning_rate": 3.942770960509163e-06, + "loss": 0.2922, + "step": 26385 + }, + { + "epoch": 0.5808056350429232, + "grad_norm": 1.3756454317871951, + "learning_rate": 3.942422566394313e-06, + "loss": 0.289, + "step": 26386 + }, + { + "epoch": 0.5808276469293419, + "grad_norm": 8.827355360117108, + "learning_rate": 3.942074177654544e-06, + "loss": 0.2781, + "step": 26387 + }, + { + "epoch": 0.5808496588157606, + "grad_norm": 0.59356042105051, + "learning_rate": 3.941725794291624e-06, + "loss": 0.4375, + "step": 26388 + }, + { + "epoch": 0.5808716707021792, + "grad_norm": 1.6397621439840617, + "learning_rate": 3.9413774163073284e-06, + "loss": 0.2798, + "step": 26389 + }, + { + "epoch": 0.5808936825885979, + "grad_norm": 1.323687483479037, + "learning_rate": 3.941029043703422e-06, + "loss": 0.3311, + "step": 26390 + }, + { + "epoch": 0.5809156944750166, + "grad_norm": 1.542100966118987, + "learning_rate": 3.94068067648168e-06, + "loss": 0.3252, + "step": 26391 + }, + { + "epoch": 0.5809377063614352, + "grad_norm": 1.4844359947089918, + "learning_rate": 3.9403323146438716e-06, + "loss": 0.3279, + "step": 26392 + }, + { + "epoch": 0.5809597182478539, + "grad_norm": 1.6519552353896774, + "learning_rate": 3.939983958191764e-06, + "loss": 0.2842, + "step": 26393 + }, + { + "epoch": 0.5809817301342725, + "grad_norm": 0.5965397588002619, + "learning_rate": 3.939635607127134e-06, + "loss": 0.457, + "step": 26394 + }, + { + "epoch": 0.5810037420206912, + "grad_norm": 1.3015690982911692, + "learning_rate": 3.939287261451748e-06, + "loss": 0.3685, + "step": 26395 + }, + { + "epoch": 0.5810257539071099, + "grad_norm": 0.6180208939667128, + "learning_rate": 3.938938921167376e-06, + "loss": 0.4804, + "step": 26396 + }, + { + "epoch": 0.5810477657935285, + "grad_norm": 1.2947999398003869, + "learning_rate": 3.938590586275792e-06, + "loss": 0.2434, + "step": 26397 + }, + { + "epoch": 0.5810697776799472, + "grad_norm": 1.4329215777462316, + "learning_rate": 3.93824225677876e-06, + "loss": 0.2973, + "step": 26398 + }, + { + "epoch": 0.5810917895663659, + "grad_norm": 1.469868283051376, + "learning_rate": 3.937893932678059e-06, + "loss": 0.285, + "step": 26399 + }, + { + "epoch": 0.5811138014527845, + "grad_norm": 1.7745356072110714, + "learning_rate": 3.937545613975452e-06, + "loss": 0.3611, + "step": 26400 + }, + { + "epoch": 0.5811358133392032, + "grad_norm": 2.3724349833579788, + "learning_rate": 3.937197300672712e-06, + "loss": 0.4365, + "step": 26401 + }, + { + "epoch": 0.5811578252256219, + "grad_norm": 1.9032359935216123, + "learning_rate": 3.9368489927716114e-06, + "loss": 0.3209, + "step": 26402 + }, + { + "epoch": 0.5811798371120405, + "grad_norm": 1.2155719966343816, + "learning_rate": 3.936500690273916e-06, + "loss": 0.2602, + "step": 26403 + }, + { + "epoch": 0.5812018489984592, + "grad_norm": 1.590460195388238, + "learning_rate": 3.9361523931814e-06, + "loss": 0.2413, + "step": 26404 + }, + { + "epoch": 0.5812238608848779, + "grad_norm": 1.7712631887869572, + "learning_rate": 3.935804101495831e-06, + "loss": 0.3237, + "step": 26405 + }, + { + "epoch": 0.5812458727712965, + "grad_norm": 1.3127228760433636, + "learning_rate": 3.93545581521898e-06, + "loss": 0.2805, + "step": 26406 + }, + { + "epoch": 0.5812678846577152, + "grad_norm": 1.81602976390322, + "learning_rate": 3.93510753435262e-06, + "loss": 0.2307, + "step": 26407 + }, + { + "epoch": 0.5812898965441339, + "grad_norm": 1.5264277669340742, + "learning_rate": 3.934759258898517e-06, + "loss": 0.2959, + "step": 26408 + }, + { + "epoch": 0.5813119084305525, + "grad_norm": 3.1359983368239615, + "learning_rate": 3.934410988858441e-06, + "loss": 0.3714, + "step": 26409 + }, + { + "epoch": 0.5813339203169712, + "grad_norm": 1.1787879860930108, + "learning_rate": 3.9340627242341665e-06, + "loss": 0.3143, + "step": 26410 + }, + { + "epoch": 0.5813559322033899, + "grad_norm": 1.3762124301593126, + "learning_rate": 3.933714465027456e-06, + "loss": 0.3732, + "step": 26411 + }, + { + "epoch": 0.5813779440898085, + "grad_norm": 1.7962315526222379, + "learning_rate": 3.933366211240089e-06, + "loss": 0.3117, + "step": 26412 + }, + { + "epoch": 0.5813999559762272, + "grad_norm": 1.1438249196425947, + "learning_rate": 3.933017962873829e-06, + "loss": 0.2511, + "step": 26413 + }, + { + "epoch": 0.5814219678626459, + "grad_norm": 1.9089704276469233, + "learning_rate": 3.932669719930446e-06, + "loss": 0.3568, + "step": 26414 + }, + { + "epoch": 0.5814439797490645, + "grad_norm": 1.6101222631532408, + "learning_rate": 3.932321482411714e-06, + "loss": 0.2849, + "step": 26415 + }, + { + "epoch": 0.5814659916354832, + "grad_norm": 2.225358783782355, + "learning_rate": 3.9319732503193974e-06, + "loss": 0.3796, + "step": 26416 + }, + { + "epoch": 0.5814880035219018, + "grad_norm": 1.7037753220136453, + "learning_rate": 3.931625023655272e-06, + "loss": 0.303, + "step": 26417 + }, + { + "epoch": 0.5815100154083205, + "grad_norm": 0.6272857387737327, + "learning_rate": 3.9312768024211036e-06, + "loss": 0.4832, + "step": 26418 + }, + { + "epoch": 0.5815320272947392, + "grad_norm": 1.3333582744068355, + "learning_rate": 3.930928586618663e-06, + "loss": 0.2532, + "step": 26419 + }, + { + "epoch": 0.5815540391811578, + "grad_norm": 2.087148433571703, + "learning_rate": 3.9305803762497215e-06, + "loss": 0.3104, + "step": 26420 + }, + { + "epoch": 0.5815760510675765, + "grad_norm": 0.6454047065954551, + "learning_rate": 3.930232171316045e-06, + "loss": 0.4748, + "step": 26421 + }, + { + "epoch": 0.5815980629539952, + "grad_norm": 1.3485408659610711, + "learning_rate": 3.929883971819408e-06, + "loss": 0.2984, + "step": 26422 + }, + { + "epoch": 0.5816200748404138, + "grad_norm": 1.616209328419153, + "learning_rate": 3.929535777761577e-06, + "loss": 0.321, + "step": 26423 + }, + { + "epoch": 0.5816420867268325, + "grad_norm": 3.166479710141628, + "learning_rate": 3.929187589144322e-06, + "loss": 0.2953, + "step": 26424 + }, + { + "epoch": 0.5816640986132512, + "grad_norm": 1.2932202321540716, + "learning_rate": 3.928839405969415e-06, + "loss": 0.3033, + "step": 26425 + }, + { + "epoch": 0.5816861104996698, + "grad_norm": 1.1710577889797973, + "learning_rate": 3.928491228238623e-06, + "loss": 0.2419, + "step": 26426 + }, + { + "epoch": 0.5817081223860885, + "grad_norm": 0.6101584963317139, + "learning_rate": 3.928143055953716e-06, + "loss": 0.4638, + "step": 26427 + }, + { + "epoch": 0.5817301342725072, + "grad_norm": 1.2805417555491088, + "learning_rate": 3.927794889116466e-06, + "loss": 0.3244, + "step": 26428 + }, + { + "epoch": 0.5817521461589258, + "grad_norm": 1.3718397835012641, + "learning_rate": 3.927446727728636e-06, + "loss": 0.3326, + "step": 26429 + }, + { + "epoch": 0.5817741580453445, + "grad_norm": 1.660417428545452, + "learning_rate": 3.927098571792005e-06, + "loss": 0.328, + "step": 26430 + }, + { + "epoch": 0.5817961699317632, + "grad_norm": 1.672342616199101, + "learning_rate": 3.926750421308335e-06, + "loss": 0.3131, + "step": 26431 + }, + { + "epoch": 0.5818181818181818, + "grad_norm": 0.6310442075507676, + "learning_rate": 3.926402276279398e-06, + "loss": 0.4468, + "step": 26432 + }, + { + "epoch": 0.5818401937046005, + "grad_norm": 0.6242580462239082, + "learning_rate": 3.926054136706965e-06, + "loss": 0.4757, + "step": 26433 + }, + { + "epoch": 0.5818622055910192, + "grad_norm": 0.6193586294149661, + "learning_rate": 3.925706002592802e-06, + "loss": 0.4658, + "step": 26434 + }, + { + "epoch": 0.5818842174774378, + "grad_norm": 1.358536653228351, + "learning_rate": 3.9253578739386824e-06, + "loss": 0.3375, + "step": 26435 + }, + { + "epoch": 0.5819062293638565, + "grad_norm": 1.303772058790504, + "learning_rate": 3.925009750746372e-06, + "loss": 0.2737, + "step": 26436 + }, + { + "epoch": 0.5819282412502752, + "grad_norm": 1.8019299203673196, + "learning_rate": 3.924661633017641e-06, + "loss": 0.2916, + "step": 26437 + }, + { + "epoch": 0.5819502531366938, + "grad_norm": 1.0973825107597264, + "learning_rate": 3.9243135207542614e-06, + "loss": 0.2591, + "step": 26438 + }, + { + "epoch": 0.5819722650231125, + "grad_norm": 1.2223250220370403, + "learning_rate": 3.923965413957998e-06, + "loss": 0.3019, + "step": 26439 + }, + { + "epoch": 0.5819942769095311, + "grad_norm": 17.887722864760377, + "learning_rate": 3.923617312630623e-06, + "loss": 0.3668, + "step": 26440 + }, + { + "epoch": 0.5820162887959498, + "grad_norm": 1.4881314092430658, + "learning_rate": 3.923269216773905e-06, + "loss": 0.3532, + "step": 26441 + }, + { + "epoch": 0.5820383006823685, + "grad_norm": 2.5890487326482883, + "learning_rate": 3.922921126389613e-06, + "loss": 0.2414, + "step": 26442 + }, + { + "epoch": 0.5820603125687871, + "grad_norm": 0.6181854611117162, + "learning_rate": 3.9225730414795175e-06, + "loss": 0.4693, + "step": 26443 + }, + { + "epoch": 0.5820823244552058, + "grad_norm": 1.1310442464694441, + "learning_rate": 3.922224962045384e-06, + "loss": 0.3214, + "step": 26444 + }, + { + "epoch": 0.5821043363416245, + "grad_norm": 1.723026504124769, + "learning_rate": 3.921876888088985e-06, + "loss": 0.316, + "step": 26445 + }, + { + "epoch": 0.5821263482280431, + "grad_norm": 0.6137084347549729, + "learning_rate": 3.92152881961209e-06, + "loss": 0.455, + "step": 26446 + }, + { + "epoch": 0.5821483601144618, + "grad_norm": 1.2342386364125733, + "learning_rate": 3.921180756616464e-06, + "loss": 0.2475, + "step": 26447 + }, + { + "epoch": 0.5821703720008805, + "grad_norm": 2.4046493850925073, + "learning_rate": 3.92083269910388e-06, + "loss": 0.2677, + "step": 26448 + }, + { + "epoch": 0.5821923838872991, + "grad_norm": 1.1858190556900658, + "learning_rate": 3.920484647076106e-06, + "loss": 0.2792, + "step": 26449 + }, + { + "epoch": 0.5822143957737178, + "grad_norm": 0.6045445800396508, + "learning_rate": 3.920136600534909e-06, + "loss": 0.4423, + "step": 26450 + }, + { + "epoch": 0.5822364076601365, + "grad_norm": 0.607585536130879, + "learning_rate": 3.9197885594820615e-06, + "loss": 0.4435, + "step": 26451 + }, + { + "epoch": 0.5822584195465551, + "grad_norm": 1.3641303586223288, + "learning_rate": 3.919440523919327e-06, + "loss": 0.2851, + "step": 26452 + }, + { + "epoch": 0.5822804314329738, + "grad_norm": 1.227952762621566, + "learning_rate": 3.919092493848481e-06, + "loss": 0.3364, + "step": 26453 + }, + { + "epoch": 0.5823024433193925, + "grad_norm": 1.3918286372314248, + "learning_rate": 3.918744469271287e-06, + "loss": 0.3309, + "step": 26454 + }, + { + "epoch": 0.5823244552058111, + "grad_norm": 1.6475509954287488, + "learning_rate": 3.918396450189517e-06, + "loss": 0.335, + "step": 26455 + }, + { + "epoch": 0.5823464670922298, + "grad_norm": 1.4818579454261345, + "learning_rate": 3.918048436604939e-06, + "loss": 0.2992, + "step": 26456 + }, + { + "epoch": 0.5823684789786485, + "grad_norm": 1.5774659078481084, + "learning_rate": 3.91770042851932e-06, + "loss": 0.3414, + "step": 26457 + }, + { + "epoch": 0.5823904908650671, + "grad_norm": 0.5954196952520274, + "learning_rate": 3.917352425934431e-06, + "loss": 0.4582, + "step": 26458 + }, + { + "epoch": 0.5824125027514858, + "grad_norm": 3.657589686525899, + "learning_rate": 3.91700442885204e-06, + "loss": 0.3528, + "step": 26459 + }, + { + "epoch": 0.5824345146379045, + "grad_norm": 1.3895204175025175, + "learning_rate": 3.916656437273914e-06, + "loss": 0.2538, + "step": 26460 + }, + { + "epoch": 0.5824565265243231, + "grad_norm": 1.7954631729220432, + "learning_rate": 3.9163084512018254e-06, + "loss": 0.2354, + "step": 26461 + }, + { + "epoch": 0.5824785384107418, + "grad_norm": 0.6114344754682055, + "learning_rate": 3.915960470637539e-06, + "loss": 0.434, + "step": 26462 + }, + { + "epoch": 0.5825005502971604, + "grad_norm": 1.506546820439529, + "learning_rate": 3.915612495582826e-06, + "loss": 0.1962, + "step": 26463 + }, + { + "epoch": 0.5825225621835791, + "grad_norm": 1.9082590264682076, + "learning_rate": 3.915264526039455e-06, + "loss": 0.3578, + "step": 26464 + }, + { + "epoch": 0.5825445740699978, + "grad_norm": 1.777099519040802, + "learning_rate": 3.91491656200919e-06, + "loss": 0.298, + "step": 26465 + }, + { + "epoch": 0.5825665859564164, + "grad_norm": 1.2690874117551914, + "learning_rate": 3.914568603493806e-06, + "loss": 0.2367, + "step": 26466 + }, + { + "epoch": 0.5825885978428351, + "grad_norm": 3.942784467801424, + "learning_rate": 3.914220650495068e-06, + "loss": 0.2646, + "step": 26467 + }, + { + "epoch": 0.5826106097292538, + "grad_norm": 0.6102664264428111, + "learning_rate": 3.913872703014743e-06, + "loss": 0.466, + "step": 26468 + }, + { + "epoch": 0.5826326216156724, + "grad_norm": 1.1456744234234608, + "learning_rate": 3.913524761054603e-06, + "loss": 0.3184, + "step": 26469 + }, + { + "epoch": 0.5826546335020911, + "grad_norm": 1.3314284682040773, + "learning_rate": 3.913176824616413e-06, + "loss": 0.2417, + "step": 26470 + }, + { + "epoch": 0.5826766453885098, + "grad_norm": 1.5879850654752599, + "learning_rate": 3.912828893701946e-06, + "loss": 0.3435, + "step": 26471 + }, + { + "epoch": 0.5826986572749284, + "grad_norm": 1.6871581833248768, + "learning_rate": 3.912480968312966e-06, + "loss": 0.2953, + "step": 26472 + }, + { + "epoch": 0.5827206691613471, + "grad_norm": 1.4098395825692231, + "learning_rate": 3.912133048451242e-06, + "loss": 0.2988, + "step": 26473 + }, + { + "epoch": 0.5827426810477658, + "grad_norm": 1.296834062742399, + "learning_rate": 3.911785134118546e-06, + "loss": 0.2288, + "step": 26474 + }, + { + "epoch": 0.5827646929341844, + "grad_norm": 1.7424993637076178, + "learning_rate": 3.91143722531664e-06, + "loss": 0.349, + "step": 26475 + }, + { + "epoch": 0.5827867048206031, + "grad_norm": 1.6146555233717843, + "learning_rate": 3.911089322047296e-06, + "loss": 0.305, + "step": 26476 + }, + { + "epoch": 0.5828087167070218, + "grad_norm": 1.4137926704485944, + "learning_rate": 3.910741424312283e-06, + "loss": 0.2389, + "step": 26477 + }, + { + "epoch": 0.5828307285934404, + "grad_norm": 1.4830079357065007, + "learning_rate": 3.910393532113367e-06, + "loss": 0.3392, + "step": 26478 + }, + { + "epoch": 0.5828527404798591, + "grad_norm": 1.4998910219651078, + "learning_rate": 3.910045645452318e-06, + "loss": 0.3132, + "step": 26479 + }, + { + "epoch": 0.5828747523662778, + "grad_norm": 0.621678630141668, + "learning_rate": 3.909697764330902e-06, + "loss": 0.4677, + "step": 26480 + }, + { + "epoch": 0.5828967642526964, + "grad_norm": 0.6220999671358564, + "learning_rate": 3.9093498887508885e-06, + "loss": 0.4799, + "step": 26481 + }, + { + "epoch": 0.5829187761391151, + "grad_norm": 1.2485567570969867, + "learning_rate": 3.909002018714047e-06, + "loss": 0.2348, + "step": 26482 + }, + { + "epoch": 0.5829407880255338, + "grad_norm": 1.5014761713515845, + "learning_rate": 3.908654154222141e-06, + "loss": 0.3353, + "step": 26483 + }, + { + "epoch": 0.5829627999119524, + "grad_norm": 1.8483320301649624, + "learning_rate": 3.908306295276943e-06, + "loss": 0.3856, + "step": 26484 + }, + { + "epoch": 0.5829848117983711, + "grad_norm": 1.1856852950211185, + "learning_rate": 3.907958441880219e-06, + "loss": 0.3032, + "step": 26485 + }, + { + "epoch": 0.5830068236847897, + "grad_norm": 1.5542884028774862, + "learning_rate": 3.907610594033737e-06, + "loss": 0.3663, + "step": 26486 + }, + { + "epoch": 0.5830288355712084, + "grad_norm": 1.8251309955781745, + "learning_rate": 3.9072627517392666e-06, + "loss": 0.3426, + "step": 26487 + }, + { + "epoch": 0.5830508474576271, + "grad_norm": 0.5923979465021927, + "learning_rate": 3.906914914998572e-06, + "loss": 0.4576, + "step": 26488 + }, + { + "epoch": 0.5830728593440457, + "grad_norm": 1.561718229187053, + "learning_rate": 3.906567083813425e-06, + "loss": 0.3354, + "step": 26489 + }, + { + "epoch": 0.5830948712304644, + "grad_norm": 3.516927363581997, + "learning_rate": 3.906219258185593e-06, + "loss": 0.273, + "step": 26490 + }, + { + "epoch": 0.5831168831168831, + "grad_norm": 1.717823446830588, + "learning_rate": 3.90587143811684e-06, + "loss": 0.3651, + "step": 26491 + }, + { + "epoch": 0.5831388950033017, + "grad_norm": 0.608813227793092, + "learning_rate": 3.905523623608938e-06, + "loss": 0.475, + "step": 26492 + }, + { + "epoch": 0.5831609068897204, + "grad_norm": 1.4481539950888043, + "learning_rate": 3.905175814663651e-06, + "loss": 0.3027, + "step": 26493 + }, + { + "epoch": 0.5831829187761391, + "grad_norm": 1.700116199498295, + "learning_rate": 3.90482801128275e-06, + "loss": 0.2657, + "step": 26494 + }, + { + "epoch": 0.5832049306625577, + "grad_norm": 1.3219729557849176, + "learning_rate": 3.904480213468002e-06, + "loss": 0.2368, + "step": 26495 + }, + { + "epoch": 0.5832269425489764, + "grad_norm": 0.5961697165304177, + "learning_rate": 3.904132421221173e-06, + "loss": 0.4574, + "step": 26496 + }, + { + "epoch": 0.5832489544353952, + "grad_norm": 1.6373916047813872, + "learning_rate": 3.903784634544034e-06, + "loss": 0.2901, + "step": 26497 + }, + { + "epoch": 0.5832709663218137, + "grad_norm": 1.417823900345224, + "learning_rate": 3.903436853438348e-06, + "loss": 0.2743, + "step": 26498 + }, + { + "epoch": 0.5832929782082324, + "grad_norm": 1.472401841516195, + "learning_rate": 3.903089077905886e-06, + "loss": 0.3241, + "step": 26499 + }, + { + "epoch": 0.5833149900946512, + "grad_norm": 1.980758243477076, + "learning_rate": 3.902741307948416e-06, + "loss": 0.2371, + "step": 26500 + }, + { + "epoch": 0.5833370019810697, + "grad_norm": 0.6290082643077308, + "learning_rate": 3.902393543567701e-06, + "loss": 0.4867, + "step": 26501 + }, + { + "epoch": 0.5833590138674885, + "grad_norm": 1.523456967011245, + "learning_rate": 3.902045784765514e-06, + "loss": 0.2545, + "step": 26502 + }, + { + "epoch": 0.5833810257539072, + "grad_norm": 1.684762065880592, + "learning_rate": 3.9016980315436186e-06, + "loss": 0.3044, + "step": 26503 + }, + { + "epoch": 0.5834030376403258, + "grad_norm": 0.5697700660257126, + "learning_rate": 3.901350283903784e-06, + "loss": 0.4362, + "step": 26504 + }, + { + "epoch": 0.5834250495267445, + "grad_norm": 1.7361726293127413, + "learning_rate": 3.901002541847779e-06, + "loss": 0.3604, + "step": 26505 + }, + { + "epoch": 0.5834470614131632, + "grad_norm": 1.3398155540203058, + "learning_rate": 3.900654805377368e-06, + "loss": 0.2535, + "step": 26506 + }, + { + "epoch": 0.5834690732995818, + "grad_norm": 1.4968947204896497, + "learning_rate": 3.900307074494319e-06, + "loss": 0.2915, + "step": 26507 + }, + { + "epoch": 0.5834910851860005, + "grad_norm": 2.918205801829427, + "learning_rate": 3.8999593492004e-06, + "loss": 0.2807, + "step": 26508 + }, + { + "epoch": 0.583513097072419, + "grad_norm": 1.5403803616960525, + "learning_rate": 3.899611629497378e-06, + "loss": 0.2685, + "step": 26509 + }, + { + "epoch": 0.5835351089588378, + "grad_norm": 0.6056216976383658, + "learning_rate": 3.899263915387022e-06, + "loss": 0.46, + "step": 26510 + }, + { + "epoch": 0.5835571208452565, + "grad_norm": 1.2116658624584862, + "learning_rate": 3.898916206871096e-06, + "loss": 0.2676, + "step": 26511 + }, + { + "epoch": 0.583579132731675, + "grad_norm": 2.7590763854643123, + "learning_rate": 3.89856850395137e-06, + "loss": 0.3319, + "step": 26512 + }, + { + "epoch": 0.5836011446180938, + "grad_norm": 1.809116226543922, + "learning_rate": 3.89822080662961e-06, + "loss": 0.3577, + "step": 26513 + }, + { + "epoch": 0.5836231565045125, + "grad_norm": 5.007302212928348, + "learning_rate": 3.897873114907582e-06, + "loss": 0.4145, + "step": 26514 + }, + { + "epoch": 0.5836451683909311, + "grad_norm": 1.5852589291415826, + "learning_rate": 3.897525428787057e-06, + "loss": 0.3172, + "step": 26515 + }, + { + "epoch": 0.5836671802773498, + "grad_norm": 1.4870642428467469, + "learning_rate": 3.897177748269797e-06, + "loss": 0.2543, + "step": 26516 + }, + { + "epoch": 0.5836891921637685, + "grad_norm": 1.636725695364964, + "learning_rate": 3.896830073357572e-06, + "loss": 0.319, + "step": 26517 + }, + { + "epoch": 0.5837112040501871, + "grad_norm": 1.3934433405206386, + "learning_rate": 3.896482404052151e-06, + "loss": 0.305, + "step": 26518 + }, + { + "epoch": 0.5837332159366058, + "grad_norm": 1.1637574740940217, + "learning_rate": 3.896134740355295e-06, + "loss": 0.3524, + "step": 26519 + }, + { + "epoch": 0.5837552278230245, + "grad_norm": 1.1818720146484878, + "learning_rate": 3.8957870822687775e-06, + "loss": 0.2135, + "step": 26520 + }, + { + "epoch": 0.5837772397094431, + "grad_norm": 1.4987917631540753, + "learning_rate": 3.89543942979436e-06, + "loss": 0.2278, + "step": 26521 + }, + { + "epoch": 0.5837992515958618, + "grad_norm": 0.6369690877013315, + "learning_rate": 3.895091782933813e-06, + "loss": 0.4742, + "step": 26522 + }, + { + "epoch": 0.5838212634822805, + "grad_norm": 1.7835274546500048, + "learning_rate": 3.894744141688904e-06, + "loss": 0.328, + "step": 26523 + }, + { + "epoch": 0.5838432753686991, + "grad_norm": 1.7328143986806959, + "learning_rate": 3.8943965060613965e-06, + "loss": 0.3552, + "step": 26524 + }, + { + "epoch": 0.5838652872551178, + "grad_norm": 2.1389021490645717, + "learning_rate": 3.894048876053059e-06, + "loss": 0.3162, + "step": 26525 + }, + { + "epoch": 0.5838872991415365, + "grad_norm": 1.0883984887547578, + "learning_rate": 3.8937012516656596e-06, + "loss": 0.2968, + "step": 26526 + }, + { + "epoch": 0.5839093110279551, + "grad_norm": 3.6725614041861028, + "learning_rate": 3.893353632900962e-06, + "loss": 0.2124, + "step": 26527 + }, + { + "epoch": 0.5839313229143738, + "grad_norm": 1.4288353955053903, + "learning_rate": 3.893006019760737e-06, + "loss": 0.3121, + "step": 26528 + }, + { + "epoch": 0.5839533348007925, + "grad_norm": 1.2482550901404448, + "learning_rate": 3.892658412246748e-06, + "loss": 0.3236, + "step": 26529 + }, + { + "epoch": 0.5839753466872111, + "grad_norm": 1.518755530423902, + "learning_rate": 3.892310810360762e-06, + "loss": 0.3542, + "step": 26530 + }, + { + "epoch": 0.5839973585736298, + "grad_norm": 1.4444998172119168, + "learning_rate": 3.891963214104548e-06, + "loss": 0.3137, + "step": 26531 + }, + { + "epoch": 0.5840193704600485, + "grad_norm": 2.0499642764624744, + "learning_rate": 3.891615623479869e-06, + "loss": 0.3032, + "step": 26532 + }, + { + "epoch": 0.5840413823464671, + "grad_norm": 1.4009066853478584, + "learning_rate": 3.891268038488497e-06, + "loss": 0.2945, + "step": 26533 + }, + { + "epoch": 0.5840633942328858, + "grad_norm": 1.5464399507428255, + "learning_rate": 3.890920459132193e-06, + "loss": 0.2772, + "step": 26534 + }, + { + "epoch": 0.5840854061193044, + "grad_norm": 1.4270248291197551, + "learning_rate": 3.890572885412726e-06, + "loss": 0.3415, + "step": 26535 + }, + { + "epoch": 0.5841074180057231, + "grad_norm": 1.7942912984155917, + "learning_rate": 3.890225317331865e-06, + "loss": 0.3185, + "step": 26536 + }, + { + "epoch": 0.5841294298921418, + "grad_norm": 1.688320197142362, + "learning_rate": 3.889877754891369e-06, + "loss": 0.2702, + "step": 26537 + }, + { + "epoch": 0.5841514417785604, + "grad_norm": 0.6317868489274705, + "learning_rate": 3.889530198093015e-06, + "loss": 0.47, + "step": 26538 + }, + { + "epoch": 0.5841734536649791, + "grad_norm": 1.3728485937378854, + "learning_rate": 3.88918264693856e-06, + "loss": 0.2377, + "step": 26539 + }, + { + "epoch": 0.5841954655513978, + "grad_norm": 0.6030609040208476, + "learning_rate": 3.8888351014297745e-06, + "loss": 0.4359, + "step": 26540 + }, + { + "epoch": 0.5842174774378164, + "grad_norm": 5.33614476380747, + "learning_rate": 3.888487561568427e-06, + "loss": 0.3395, + "step": 26541 + }, + { + "epoch": 0.5842394893242351, + "grad_norm": 1.994098959510004, + "learning_rate": 3.888140027356279e-06, + "loss": 0.3555, + "step": 26542 + }, + { + "epoch": 0.5842615012106538, + "grad_norm": 1.6895482292894053, + "learning_rate": 3.8877924987951e-06, + "loss": 0.2908, + "step": 26543 + }, + { + "epoch": 0.5842835130970724, + "grad_norm": 1.6042723248642021, + "learning_rate": 3.887444975886655e-06, + "loss": 0.2295, + "step": 26544 + }, + { + "epoch": 0.5843055249834911, + "grad_norm": 1.3112475652829052, + "learning_rate": 3.887097458632711e-06, + "loss": 0.3132, + "step": 26545 + }, + { + "epoch": 0.5843275368699098, + "grad_norm": 0.6058155776906243, + "learning_rate": 3.886749947035037e-06, + "loss": 0.4727, + "step": 26546 + }, + { + "epoch": 0.5843495487563284, + "grad_norm": 1.3113604360300755, + "learning_rate": 3.8864024410953925e-06, + "loss": 0.2749, + "step": 26547 + }, + { + "epoch": 0.5843715606427471, + "grad_norm": 0.6017703688476147, + "learning_rate": 3.886054940815549e-06, + "loss": 0.4336, + "step": 26548 + }, + { + "epoch": 0.5843935725291658, + "grad_norm": 1.1537996600075282, + "learning_rate": 3.8857074461972715e-06, + "loss": 0.2903, + "step": 26549 + }, + { + "epoch": 0.5844155844155844, + "grad_norm": 1.6618966754168056, + "learning_rate": 3.885359957242325e-06, + "loss": 0.2795, + "step": 26550 + }, + { + "epoch": 0.5844375963020031, + "grad_norm": 0.6090436709840953, + "learning_rate": 3.8850124739524774e-06, + "loss": 0.4718, + "step": 26551 + }, + { + "epoch": 0.5844596081884218, + "grad_norm": 2.305085691950357, + "learning_rate": 3.8846649963294915e-06, + "loss": 0.3491, + "step": 26552 + }, + { + "epoch": 0.5844816200748404, + "grad_norm": 1.8108038383137406, + "learning_rate": 3.884317524375137e-06, + "loss": 0.3242, + "step": 26553 + }, + { + "epoch": 0.5845036319612591, + "grad_norm": 0.6288798719850619, + "learning_rate": 3.88397005809118e-06, + "loss": 0.471, + "step": 26554 + }, + { + "epoch": 0.5845256438476778, + "grad_norm": 0.6254111588100986, + "learning_rate": 3.883622597479382e-06, + "loss": 0.4624, + "step": 26555 + }, + { + "epoch": 0.5845476557340964, + "grad_norm": 1.9283041404663503, + "learning_rate": 3.883275142541514e-06, + "loss": 0.2676, + "step": 26556 + }, + { + "epoch": 0.5845696676205151, + "grad_norm": 15.448606741883822, + "learning_rate": 3.882927693279338e-06, + "loss": 0.2706, + "step": 26557 + }, + { + "epoch": 0.5845916795069337, + "grad_norm": 1.2997275773938428, + "learning_rate": 3.882580249694623e-06, + "loss": 0.3355, + "step": 26558 + }, + { + "epoch": 0.5846136913933524, + "grad_norm": 1.5258626958860755, + "learning_rate": 3.882232811789134e-06, + "loss": 0.3829, + "step": 26559 + }, + { + "epoch": 0.5846357032797711, + "grad_norm": 2.4653415956167732, + "learning_rate": 3.881885379564635e-06, + "loss": 0.2668, + "step": 26560 + }, + { + "epoch": 0.5846577151661897, + "grad_norm": 0.5960224939598994, + "learning_rate": 3.881537953022894e-06, + "loss": 0.4631, + "step": 26561 + }, + { + "epoch": 0.5846797270526084, + "grad_norm": 1.847262714671769, + "learning_rate": 3.881190532165675e-06, + "loss": 0.2211, + "step": 26562 + }, + { + "epoch": 0.5847017389390271, + "grad_norm": 0.6256603401343075, + "learning_rate": 3.880843116994746e-06, + "loss": 0.4768, + "step": 26563 + }, + { + "epoch": 0.5847237508254457, + "grad_norm": 1.7088911630621075, + "learning_rate": 3.880495707511872e-06, + "loss": 0.4179, + "step": 26564 + }, + { + "epoch": 0.5847457627118644, + "grad_norm": 1.536377637295077, + "learning_rate": 3.880148303718816e-06, + "loss": 0.292, + "step": 26565 + }, + { + "epoch": 0.5847677745982831, + "grad_norm": 1.4995436078044555, + "learning_rate": 3.879800905617347e-06, + "loss": 0.258, + "step": 26566 + }, + { + "epoch": 0.5847897864847017, + "grad_norm": 0.6124930798156303, + "learning_rate": 3.87945351320923e-06, + "loss": 0.455, + "step": 26567 + }, + { + "epoch": 0.5848117983711204, + "grad_norm": 1.8443020028239991, + "learning_rate": 3.8791061264962286e-06, + "loss": 0.37, + "step": 26568 + }, + { + "epoch": 0.5848338102575391, + "grad_norm": 1.1301703320765721, + "learning_rate": 3.878758745480112e-06, + "loss": 0.2527, + "step": 26569 + }, + { + "epoch": 0.5848558221439577, + "grad_norm": 0.6282226068616794, + "learning_rate": 3.87841137016264e-06, + "loss": 0.4777, + "step": 26570 + }, + { + "epoch": 0.5848778340303764, + "grad_norm": 1.2583089629974265, + "learning_rate": 3.878064000545584e-06, + "loss": 0.3647, + "step": 26571 + }, + { + "epoch": 0.5848998459167951, + "grad_norm": 2.0493217888289035, + "learning_rate": 3.877716636630708e-06, + "loss": 0.3576, + "step": 26572 + }, + { + "epoch": 0.5849218578032137, + "grad_norm": 0.6198018051663806, + "learning_rate": 3.877369278419774e-06, + "loss": 0.4644, + "step": 26573 + }, + { + "epoch": 0.5849438696896324, + "grad_norm": 1.6729644545958307, + "learning_rate": 3.877021925914553e-06, + "loss": 0.4097, + "step": 26574 + }, + { + "epoch": 0.5849658815760511, + "grad_norm": 1.718306408141729, + "learning_rate": 3.876674579116805e-06, + "loss": 0.289, + "step": 26575 + }, + { + "epoch": 0.5849878934624697, + "grad_norm": 0.6249403058072942, + "learning_rate": 3.876327238028299e-06, + "loss": 0.4456, + "step": 26576 + }, + { + "epoch": 0.5850099053488884, + "grad_norm": 0.6138100715555646, + "learning_rate": 3.8759799026508e-06, + "loss": 0.4859, + "step": 26577 + }, + { + "epoch": 0.5850319172353071, + "grad_norm": 1.3797351238269573, + "learning_rate": 3.875632572986071e-06, + "loss": 0.2671, + "step": 26578 + }, + { + "epoch": 0.5850539291217257, + "grad_norm": 1.03828305591825, + "learning_rate": 3.875285249035878e-06, + "loss": 0.2385, + "step": 26579 + }, + { + "epoch": 0.5850759410081444, + "grad_norm": 1.5584386447684877, + "learning_rate": 3.874937930801988e-06, + "loss": 0.3503, + "step": 26580 + }, + { + "epoch": 0.585097952894563, + "grad_norm": 1.2299797111337398, + "learning_rate": 3.874590618286166e-06, + "loss": 0.3016, + "step": 26581 + }, + { + "epoch": 0.5851199647809817, + "grad_norm": 1.8204838494137368, + "learning_rate": 3.8742433114901765e-06, + "loss": 0.3179, + "step": 26582 + }, + { + "epoch": 0.5851419766674004, + "grad_norm": 1.750044321848186, + "learning_rate": 3.873896010415783e-06, + "loss": 0.2705, + "step": 26583 + }, + { + "epoch": 0.585163988553819, + "grad_norm": 1.7610675341688868, + "learning_rate": 3.873548715064753e-06, + "loss": 0.3365, + "step": 26584 + }, + { + "epoch": 0.5851860004402377, + "grad_norm": 1.4770961137018144, + "learning_rate": 3.87320142543885e-06, + "loss": 0.2521, + "step": 26585 + }, + { + "epoch": 0.5852080123266564, + "grad_norm": 1.925344306749888, + "learning_rate": 3.872854141539841e-06, + "loss": 0.237, + "step": 26586 + }, + { + "epoch": 0.585230024213075, + "grad_norm": 1.3988992182407995, + "learning_rate": 3.872506863369491e-06, + "loss": 0.3505, + "step": 26587 + }, + { + "epoch": 0.5852520360994937, + "grad_norm": 1.9846769387628418, + "learning_rate": 3.872159590929562e-06, + "loss": 0.3649, + "step": 26588 + }, + { + "epoch": 0.5852740479859124, + "grad_norm": 1.567122027987744, + "learning_rate": 3.871812324221822e-06, + "loss": 0.3453, + "step": 26589 + }, + { + "epoch": 0.585296059872331, + "grad_norm": 1.5629528487455526, + "learning_rate": 3.871465063248036e-06, + "loss": 0.3306, + "step": 26590 + }, + { + "epoch": 0.5853180717587497, + "grad_norm": 0.625823324755554, + "learning_rate": 3.871117808009965e-06, + "loss": 0.4569, + "step": 26591 + }, + { + "epoch": 0.5853400836451684, + "grad_norm": 1.2065627240909929, + "learning_rate": 3.87077055850938e-06, + "loss": 0.2242, + "step": 26592 + }, + { + "epoch": 0.585362095531587, + "grad_norm": 1.5723615470701586, + "learning_rate": 3.870423314748039e-06, + "loss": 0.2745, + "step": 26593 + }, + { + "epoch": 0.5853841074180057, + "grad_norm": 1.707175337617464, + "learning_rate": 3.870076076727714e-06, + "loss": 0.4812, + "step": 26594 + }, + { + "epoch": 0.5854061193044244, + "grad_norm": 2.0666231709790446, + "learning_rate": 3.869728844450165e-06, + "loss": 0.3038, + "step": 26595 + }, + { + "epoch": 0.585428131190843, + "grad_norm": 2.1788057721891425, + "learning_rate": 3.869381617917157e-06, + "loss": 0.3058, + "step": 26596 + }, + { + "epoch": 0.5854501430772617, + "grad_norm": 2.334494385998404, + "learning_rate": 3.869034397130458e-06, + "loss": 0.3429, + "step": 26597 + }, + { + "epoch": 0.5854721549636804, + "grad_norm": 0.6239546977057681, + "learning_rate": 3.868687182091829e-06, + "loss": 0.4674, + "step": 26598 + }, + { + "epoch": 0.585494166850099, + "grad_norm": 2.1588926200056315, + "learning_rate": 3.868339972803037e-06, + "loss": 0.2429, + "step": 26599 + }, + { + "epoch": 0.5855161787365177, + "grad_norm": 1.7375283515521862, + "learning_rate": 3.867992769265847e-06, + "loss": 0.2476, + "step": 26600 + }, + { + "epoch": 0.5855381906229364, + "grad_norm": 1.6416756728778739, + "learning_rate": 3.867645571482021e-06, + "loss": 0.355, + "step": 26601 + }, + { + "epoch": 0.585560202509355, + "grad_norm": 1.3742328810926847, + "learning_rate": 3.867298379453326e-06, + "loss": 0.2309, + "step": 26602 + }, + { + "epoch": 0.5855822143957737, + "grad_norm": 1.3350317560215896, + "learning_rate": 3.866951193181525e-06, + "loss": 0.2557, + "step": 26603 + }, + { + "epoch": 0.5856042262821923, + "grad_norm": 1.479468604963421, + "learning_rate": 3.866604012668384e-06, + "loss": 0.3131, + "step": 26604 + }, + { + "epoch": 0.585626238168611, + "grad_norm": 0.6307267103524197, + "learning_rate": 3.866256837915668e-06, + "loss": 0.4706, + "step": 26605 + }, + { + "epoch": 0.5856482500550297, + "grad_norm": 1.2562639031388183, + "learning_rate": 3.865909668925138e-06, + "loss": 0.3015, + "step": 26606 + }, + { + "epoch": 0.5856702619414483, + "grad_norm": 1.6680284277278177, + "learning_rate": 3.865562505698563e-06, + "loss": 0.2452, + "step": 26607 + }, + { + "epoch": 0.585692273827867, + "grad_norm": 5.863514058074611, + "learning_rate": 3.8652153482377054e-06, + "loss": 0.3379, + "step": 26608 + }, + { + "epoch": 0.5857142857142857, + "grad_norm": 2.1703660677638483, + "learning_rate": 3.8648681965443265e-06, + "loss": 0.2333, + "step": 26609 + }, + { + "epoch": 0.5857362976007043, + "grad_norm": 1.6939364036630236, + "learning_rate": 3.8645210506201975e-06, + "loss": 0.3679, + "step": 26610 + }, + { + "epoch": 0.585758309487123, + "grad_norm": 1.4863362234780477, + "learning_rate": 3.864173910467077e-06, + "loss": 0.3072, + "step": 26611 + }, + { + "epoch": 0.5857803213735417, + "grad_norm": 0.6084621111283631, + "learning_rate": 3.863826776086732e-06, + "loss": 0.4616, + "step": 26612 + }, + { + "epoch": 0.5858023332599603, + "grad_norm": 1.3968119831807795, + "learning_rate": 3.863479647480927e-06, + "loss": 0.2877, + "step": 26613 + }, + { + "epoch": 0.585824345146379, + "grad_norm": 1.5256061979765843, + "learning_rate": 3.863132524651423e-06, + "loss": 0.3141, + "step": 26614 + }, + { + "epoch": 0.5858463570327977, + "grad_norm": 1.4798732013863483, + "learning_rate": 3.862785407599987e-06, + "loss": 0.3379, + "step": 26615 + }, + { + "epoch": 0.5858683689192163, + "grad_norm": 0.5895922028593193, + "learning_rate": 3.862438296328383e-06, + "loss": 0.4529, + "step": 26616 + }, + { + "epoch": 0.585890380805635, + "grad_norm": 1.3442893501474191, + "learning_rate": 3.862091190838375e-06, + "loss": 0.288, + "step": 26617 + }, + { + "epoch": 0.5859123926920538, + "grad_norm": 1.7161685146579702, + "learning_rate": 3.861744091131729e-06, + "loss": 0.4127, + "step": 26618 + }, + { + "epoch": 0.5859344045784723, + "grad_norm": 1.5397015277442883, + "learning_rate": 3.861396997210205e-06, + "loss": 0.2324, + "step": 26619 + }, + { + "epoch": 0.585956416464891, + "grad_norm": 1.7170242382439733, + "learning_rate": 3.86104990907557e-06, + "loss": 0.3785, + "step": 26620 + }, + { + "epoch": 0.5859784283513098, + "grad_norm": 1.465986784065969, + "learning_rate": 3.860702826729586e-06, + "loss": 0.2669, + "step": 26621 + }, + { + "epoch": 0.5860004402377283, + "grad_norm": 1.3925132777533824, + "learning_rate": 3.86035575017402e-06, + "loss": 0.3312, + "step": 26622 + }, + { + "epoch": 0.586022452124147, + "grad_norm": 1.3654583201679589, + "learning_rate": 3.860008679410635e-06, + "loss": 0.2666, + "step": 26623 + }, + { + "epoch": 0.5860444640105658, + "grad_norm": 0.6482681551095071, + "learning_rate": 3.859661614441193e-06, + "loss": 0.4641, + "step": 26624 + }, + { + "epoch": 0.5860664758969844, + "grad_norm": 1.5527485185270185, + "learning_rate": 3.859314555267459e-06, + "loss": 0.3169, + "step": 26625 + }, + { + "epoch": 0.5860884877834031, + "grad_norm": 1.3936742732196779, + "learning_rate": 3.8589675018912e-06, + "loss": 0.2835, + "step": 26626 + }, + { + "epoch": 0.5861104996698216, + "grad_norm": 1.790510534680278, + "learning_rate": 3.858620454314173e-06, + "loss": 0.2894, + "step": 26627 + }, + { + "epoch": 0.5861325115562404, + "grad_norm": 1.2333111258935576, + "learning_rate": 3.8582734125381485e-06, + "loss": 0.2496, + "step": 26628 + }, + { + "epoch": 0.5861545234426591, + "grad_norm": 0.61755651094491, + "learning_rate": 3.857926376564886e-06, + "loss": 0.4442, + "step": 26629 + }, + { + "epoch": 0.5861765353290777, + "grad_norm": 1.9924519322678629, + "learning_rate": 3.857579346396153e-06, + "loss": 0.2914, + "step": 26630 + }, + { + "epoch": 0.5861985472154964, + "grad_norm": 0.6165678078327208, + "learning_rate": 3.857232322033711e-06, + "loss": 0.4736, + "step": 26631 + }, + { + "epoch": 0.5862205591019151, + "grad_norm": 0.6183217665249133, + "learning_rate": 3.856885303479322e-06, + "loss": 0.4572, + "step": 26632 + }, + { + "epoch": 0.5862425709883337, + "grad_norm": 1.244045384915079, + "learning_rate": 3.856538290734754e-06, + "loss": 0.3379, + "step": 26633 + }, + { + "epoch": 0.5862645828747524, + "grad_norm": 1.315848366111673, + "learning_rate": 3.856191283801767e-06, + "loss": 0.2843, + "step": 26634 + }, + { + "epoch": 0.5862865947611711, + "grad_norm": 1.5666019965135363, + "learning_rate": 3.8558442826821265e-06, + "loss": 0.34, + "step": 26635 + }, + { + "epoch": 0.5863086066475897, + "grad_norm": 1.4163172959063886, + "learning_rate": 3.855497287377597e-06, + "loss": 0.3209, + "step": 26636 + }, + { + "epoch": 0.5863306185340084, + "grad_norm": 1.4743914390738053, + "learning_rate": 3.855150297889939e-06, + "loss": 0.3991, + "step": 26637 + }, + { + "epoch": 0.5863526304204271, + "grad_norm": 1.6632005983133986, + "learning_rate": 3.854803314220918e-06, + "loss": 0.2884, + "step": 26638 + }, + { + "epoch": 0.5863746423068457, + "grad_norm": 1.5499398851477035, + "learning_rate": 3.8544563363722965e-06, + "loss": 0.2683, + "step": 26639 + }, + { + "epoch": 0.5863966541932644, + "grad_norm": 1.6388329383378641, + "learning_rate": 3.854109364345841e-06, + "loss": 0.3486, + "step": 26640 + }, + { + "epoch": 0.5864186660796831, + "grad_norm": 0.6184412909247939, + "learning_rate": 3.8537623981433125e-06, + "loss": 0.4648, + "step": 26641 + }, + { + "epoch": 0.5864406779661017, + "grad_norm": 0.6336039682065522, + "learning_rate": 3.853415437766473e-06, + "loss": 0.4646, + "step": 26642 + }, + { + "epoch": 0.5864626898525204, + "grad_norm": 1.6281617357642189, + "learning_rate": 3.853068483217088e-06, + "loss": 0.2989, + "step": 26643 + }, + { + "epoch": 0.5864847017389391, + "grad_norm": 2.3258061308205478, + "learning_rate": 3.852721534496921e-06, + "loss": 0.2482, + "step": 26644 + }, + { + "epoch": 0.5865067136253577, + "grad_norm": 1.3404625036669802, + "learning_rate": 3.852374591607735e-06, + "loss": 0.239, + "step": 26645 + }, + { + "epoch": 0.5865287255117764, + "grad_norm": 0.6171017492033553, + "learning_rate": 3.852027654551295e-06, + "loss": 0.4808, + "step": 26646 + }, + { + "epoch": 0.5865507373981951, + "grad_norm": 8.667155115840336, + "learning_rate": 3.851680723329359e-06, + "loss": 0.2899, + "step": 26647 + }, + { + "epoch": 0.5865727492846137, + "grad_norm": 1.102399112323243, + "learning_rate": 3.851333797943696e-06, + "loss": 0.2395, + "step": 26648 + }, + { + "epoch": 0.5865947611710324, + "grad_norm": 1.3294492625159027, + "learning_rate": 3.850986878396068e-06, + "loss": 0.3588, + "step": 26649 + }, + { + "epoch": 0.586616773057451, + "grad_norm": 1.6789499828393608, + "learning_rate": 3.850639964688234e-06, + "loss": 0.3871, + "step": 26650 + }, + { + "epoch": 0.5866387849438697, + "grad_norm": 1.2976305591260076, + "learning_rate": 3.850293056821962e-06, + "loss": 0.2833, + "step": 26651 + }, + { + "epoch": 0.5866607968302884, + "grad_norm": 1.4690428174056551, + "learning_rate": 3.849946154799013e-06, + "loss": 0.3329, + "step": 26652 + }, + { + "epoch": 0.586682808716707, + "grad_norm": 1.4085686822269585, + "learning_rate": 3.8495992586211505e-06, + "loss": 0.3053, + "step": 26653 + }, + { + "epoch": 0.5867048206031257, + "grad_norm": 1.7834010926079584, + "learning_rate": 3.84925236829014e-06, + "loss": 0.3282, + "step": 26654 + }, + { + "epoch": 0.5867268324895444, + "grad_norm": 1.794547797214375, + "learning_rate": 3.8489054838077384e-06, + "loss": 0.2914, + "step": 26655 + }, + { + "epoch": 0.586748844375963, + "grad_norm": 1.5517339702140323, + "learning_rate": 3.848558605175715e-06, + "loss": 0.2374, + "step": 26656 + }, + { + "epoch": 0.5867708562623817, + "grad_norm": 2.4268242271205893, + "learning_rate": 3.848211732395828e-06, + "loss": 0.3215, + "step": 26657 + }, + { + "epoch": 0.5867928681488004, + "grad_norm": 1.222501544213515, + "learning_rate": 3.847864865469845e-06, + "loss": 0.29, + "step": 26658 + }, + { + "epoch": 0.586814880035219, + "grad_norm": 1.4513588719998254, + "learning_rate": 3.847518004399527e-06, + "loss": 0.3252, + "step": 26659 + }, + { + "epoch": 0.5868368919216377, + "grad_norm": 1.31712865269105, + "learning_rate": 3.847171149186635e-06, + "loss": 0.245, + "step": 26660 + }, + { + "epoch": 0.5868589038080564, + "grad_norm": 2.4909069809581004, + "learning_rate": 3.846824299832934e-06, + "loss": 0.3838, + "step": 26661 + }, + { + "epoch": 0.586880915694475, + "grad_norm": 1.7466009386575747, + "learning_rate": 3.846477456340185e-06, + "loss": 0.2973, + "step": 26662 + }, + { + "epoch": 0.5869029275808937, + "grad_norm": 1.4409806064723094, + "learning_rate": 3.8461306187101535e-06, + "loss": 0.3147, + "step": 26663 + }, + { + "epoch": 0.5869249394673124, + "grad_norm": 1.319822699401265, + "learning_rate": 3.845783786944602e-06, + "loss": 0.2342, + "step": 26664 + }, + { + "epoch": 0.586946951353731, + "grad_norm": 1.1902091252450657, + "learning_rate": 3.845436961045289e-06, + "loss": 0.3686, + "step": 26665 + }, + { + "epoch": 0.5869689632401497, + "grad_norm": 1.746725973006264, + "learning_rate": 3.845090141013983e-06, + "loss": 0.3255, + "step": 26666 + }, + { + "epoch": 0.5869909751265684, + "grad_norm": 1.2662118635446642, + "learning_rate": 3.844743326852444e-06, + "loss": 0.3215, + "step": 26667 + }, + { + "epoch": 0.587012987012987, + "grad_norm": 1.393283377894565, + "learning_rate": 3.844396518562433e-06, + "loss": 0.3457, + "step": 26668 + }, + { + "epoch": 0.5870349988994057, + "grad_norm": 1.3025920784794216, + "learning_rate": 3.8440497161457145e-06, + "loss": 0.2469, + "step": 26669 + }, + { + "epoch": 0.5870570107858244, + "grad_norm": 1.3075000998456978, + "learning_rate": 3.843702919604051e-06, + "loss": 0.255, + "step": 26670 + }, + { + "epoch": 0.587079022672243, + "grad_norm": 1.336442350914779, + "learning_rate": 3.843356128939205e-06, + "loss": 0.3634, + "step": 26671 + }, + { + "epoch": 0.5871010345586617, + "grad_norm": 1.3654694971839043, + "learning_rate": 3.843009344152941e-06, + "loss": 0.2714, + "step": 26672 + }, + { + "epoch": 0.5871230464450804, + "grad_norm": 0.6699032827864374, + "learning_rate": 3.842662565247017e-06, + "loss": 0.4769, + "step": 26673 + }, + { + "epoch": 0.587145058331499, + "grad_norm": 1.2858175000565863, + "learning_rate": 3.842315792223199e-06, + "loss": 0.3035, + "step": 26674 + }, + { + "epoch": 0.5871670702179177, + "grad_norm": 1.3056517953954032, + "learning_rate": 3.841969025083248e-06, + "loss": 0.2113, + "step": 26675 + }, + { + "epoch": 0.5871890821043363, + "grad_norm": 1.2173249979447935, + "learning_rate": 3.841622263828927e-06, + "loss": 0.2173, + "step": 26676 + }, + { + "epoch": 0.587211093990755, + "grad_norm": 1.3804819640747963, + "learning_rate": 3.841275508462e-06, + "loss": 0.3215, + "step": 26677 + }, + { + "epoch": 0.5872331058771737, + "grad_norm": 0.6489199237448677, + "learning_rate": 3.840928758984226e-06, + "loss": 0.4855, + "step": 26678 + }, + { + "epoch": 0.5872551177635923, + "grad_norm": 1.6335016610025317, + "learning_rate": 3.840582015397369e-06, + "loss": 0.2968, + "step": 26679 + }, + { + "epoch": 0.587277129650011, + "grad_norm": 1.5533734862050008, + "learning_rate": 3.840235277703191e-06, + "loss": 0.3001, + "step": 26680 + }, + { + "epoch": 0.5872991415364297, + "grad_norm": 3.9249538792569534, + "learning_rate": 3.839888545903455e-06, + "loss": 0.2577, + "step": 26681 + }, + { + "epoch": 0.5873211534228483, + "grad_norm": 0.6287221835441891, + "learning_rate": 3.8395418199999245e-06, + "loss": 0.4767, + "step": 26682 + }, + { + "epoch": 0.587343165309267, + "grad_norm": 1.6323111569140463, + "learning_rate": 3.839195099994358e-06, + "loss": 0.3765, + "step": 26683 + }, + { + "epoch": 0.5873651771956857, + "grad_norm": 1.2916304824732099, + "learning_rate": 3.838848385888521e-06, + "loss": 0.3135, + "step": 26684 + }, + { + "epoch": 0.5873871890821043, + "grad_norm": 1.4630568405170612, + "learning_rate": 3.838501677684172e-06, + "loss": 0.304, + "step": 26685 + }, + { + "epoch": 0.587409200968523, + "grad_norm": 0.6325156443765295, + "learning_rate": 3.838154975383079e-06, + "loss": 0.4773, + "step": 26686 + }, + { + "epoch": 0.5874312128549417, + "grad_norm": 1.614126409345197, + "learning_rate": 3.837808278986999e-06, + "loss": 0.3125, + "step": 26687 + }, + { + "epoch": 0.5874532247413603, + "grad_norm": 1.341803349355462, + "learning_rate": 3.837461588497694e-06, + "loss": 0.3349, + "step": 26688 + }, + { + "epoch": 0.587475236627779, + "grad_norm": 1.4202160321845316, + "learning_rate": 3.83711490391693e-06, + "loss": 0.2832, + "step": 26689 + }, + { + "epoch": 0.5874972485141977, + "grad_norm": 1.413049780105372, + "learning_rate": 3.836768225246467e-06, + "loss": 0.3312, + "step": 26690 + }, + { + "epoch": 0.5875192604006163, + "grad_norm": 1.288700760742034, + "learning_rate": 3.836421552488065e-06, + "loss": 0.2926, + "step": 26691 + }, + { + "epoch": 0.587541272287035, + "grad_norm": 1.5116863253330979, + "learning_rate": 3.836074885643488e-06, + "loss": 0.2796, + "step": 26692 + }, + { + "epoch": 0.5875632841734537, + "grad_norm": 0.6251801429493068, + "learning_rate": 3.835728224714497e-06, + "loss": 0.4686, + "step": 26693 + }, + { + "epoch": 0.5875852960598723, + "grad_norm": 1.7033642971305218, + "learning_rate": 3.835381569702855e-06, + "loss": 0.3188, + "step": 26694 + }, + { + "epoch": 0.587607307946291, + "grad_norm": 1.2048186285984457, + "learning_rate": 3.8350349206103255e-06, + "loss": 0.2559, + "step": 26695 + }, + { + "epoch": 0.5876293198327097, + "grad_norm": 1.6326420114026703, + "learning_rate": 3.834688277438665e-06, + "loss": 0.3232, + "step": 26696 + }, + { + "epoch": 0.5876513317191283, + "grad_norm": 1.261402409283222, + "learning_rate": 3.834341640189641e-06, + "loss": 0.2595, + "step": 26697 + }, + { + "epoch": 0.587673343605547, + "grad_norm": 0.6352748949361818, + "learning_rate": 3.833995008865009e-06, + "loss": 0.4943, + "step": 26698 + }, + { + "epoch": 0.5876953554919656, + "grad_norm": 0.6312418761007612, + "learning_rate": 3.833648383466537e-06, + "loss": 0.4938, + "step": 26699 + }, + { + "epoch": 0.5877173673783843, + "grad_norm": 0.5948036515705593, + "learning_rate": 3.833301763995986e-06, + "loss": 0.4614, + "step": 26700 + }, + { + "epoch": 0.587739379264803, + "grad_norm": 0.598284362638334, + "learning_rate": 3.832955150455112e-06, + "loss": 0.4623, + "step": 26701 + }, + { + "epoch": 0.5877613911512216, + "grad_norm": 1.35195841627164, + "learning_rate": 3.832608542845683e-06, + "loss": 0.279, + "step": 26702 + }, + { + "epoch": 0.5877834030376403, + "grad_norm": 1.2245221200303025, + "learning_rate": 3.832261941169456e-06, + "loss": 0.2058, + "step": 26703 + }, + { + "epoch": 0.587805414924059, + "grad_norm": 0.6009295605628359, + "learning_rate": 3.831915345428197e-06, + "loss": 0.4666, + "step": 26704 + }, + { + "epoch": 0.5878274268104776, + "grad_norm": 1.175815920543718, + "learning_rate": 3.8315687556236646e-06, + "loss": 0.2663, + "step": 26705 + }, + { + "epoch": 0.5878494386968963, + "grad_norm": 1.4229678416512612, + "learning_rate": 3.831222171757619e-06, + "loss": 0.2511, + "step": 26706 + }, + { + "epoch": 0.587871450583315, + "grad_norm": 1.5159970699133036, + "learning_rate": 3.830875593831825e-06, + "loss": 0.3508, + "step": 26707 + }, + { + "epoch": 0.5878934624697336, + "grad_norm": 11.558462076160058, + "learning_rate": 3.830529021848044e-06, + "loss": 0.3435, + "step": 26708 + }, + { + "epoch": 0.5879154743561523, + "grad_norm": 1.6885079417419704, + "learning_rate": 3.830182455808034e-06, + "loss": 0.4066, + "step": 26709 + }, + { + "epoch": 0.587937486242571, + "grad_norm": 1.5033349935895306, + "learning_rate": 3.82983589571356e-06, + "loss": 0.3144, + "step": 26710 + }, + { + "epoch": 0.5879594981289896, + "grad_norm": 1.095104970345893, + "learning_rate": 3.829489341566379e-06, + "loss": 0.2848, + "step": 26711 + }, + { + "epoch": 0.5879815100154083, + "grad_norm": 1.2100835072998513, + "learning_rate": 3.829142793368258e-06, + "loss": 0.2276, + "step": 26712 + }, + { + "epoch": 0.588003521901827, + "grad_norm": 1.6354199333054147, + "learning_rate": 3.828796251120956e-06, + "loss": 0.1984, + "step": 26713 + }, + { + "epoch": 0.5880255337882456, + "grad_norm": 1.5123137646629385, + "learning_rate": 3.8284497148262325e-06, + "loss": 0.3371, + "step": 26714 + }, + { + "epoch": 0.5880475456746643, + "grad_norm": 2.2330190551267304, + "learning_rate": 3.82810318448585e-06, + "loss": 0.3804, + "step": 26715 + }, + { + "epoch": 0.588069557561083, + "grad_norm": 3.121164410870142, + "learning_rate": 3.827756660101569e-06, + "loss": 0.2786, + "step": 26716 + }, + { + "epoch": 0.5880915694475016, + "grad_norm": 0.6491605350842253, + "learning_rate": 3.827410141675154e-06, + "loss": 0.4889, + "step": 26717 + }, + { + "epoch": 0.5881135813339203, + "grad_norm": 1.6658865257691748, + "learning_rate": 3.827063629208363e-06, + "loss": 0.2773, + "step": 26718 + }, + { + "epoch": 0.588135593220339, + "grad_norm": 2.07317275106469, + "learning_rate": 3.826717122702955e-06, + "loss": 0.3335, + "step": 26719 + }, + { + "epoch": 0.5881576051067576, + "grad_norm": 1.8186110341753947, + "learning_rate": 3.826370622160696e-06, + "loss": 0.3592, + "step": 26720 + }, + { + "epoch": 0.5881796169931763, + "grad_norm": 1.4380105050176613, + "learning_rate": 3.826024127583344e-06, + "loss": 0.3068, + "step": 26721 + }, + { + "epoch": 0.5882016288795949, + "grad_norm": 2.05705248911502, + "learning_rate": 3.825677638972663e-06, + "loss": 0.2551, + "step": 26722 + }, + { + "epoch": 0.5882236407660136, + "grad_norm": 1.4222399474667136, + "learning_rate": 3.825331156330412e-06, + "loss": 0.3215, + "step": 26723 + }, + { + "epoch": 0.5882456526524323, + "grad_norm": 1.1000284303296028, + "learning_rate": 3.824984679658349e-06, + "loss": 0.3208, + "step": 26724 + }, + { + "epoch": 0.5882676645388509, + "grad_norm": 1.3760049767705136, + "learning_rate": 3.82463820895824e-06, + "loss": 0.3548, + "step": 26725 + }, + { + "epoch": 0.5882896764252696, + "grad_norm": 0.6097072155990204, + "learning_rate": 3.824291744231845e-06, + "loss": 0.4596, + "step": 26726 + }, + { + "epoch": 0.5883116883116883, + "grad_norm": 1.6116506128943704, + "learning_rate": 3.823945285480921e-06, + "loss": 0.2895, + "step": 26727 + }, + { + "epoch": 0.5883337001981069, + "grad_norm": 1.8154867317866377, + "learning_rate": 3.823598832707234e-06, + "loss": 0.3548, + "step": 26728 + }, + { + "epoch": 0.5883557120845256, + "grad_norm": 1.580884016941639, + "learning_rate": 3.823252385912541e-06, + "loss": 0.2695, + "step": 26729 + }, + { + "epoch": 0.5883777239709443, + "grad_norm": 1.3341146298624078, + "learning_rate": 3.822905945098604e-06, + "loss": 0.3008, + "step": 26730 + }, + { + "epoch": 0.5883997358573629, + "grad_norm": 1.0845681287926197, + "learning_rate": 3.822559510267186e-06, + "loss": 0.1986, + "step": 26731 + }, + { + "epoch": 0.5884217477437816, + "grad_norm": 1.2729362593187883, + "learning_rate": 3.822213081420043e-06, + "loss": 0.2155, + "step": 26732 + }, + { + "epoch": 0.5884437596302003, + "grad_norm": 1.618227783058984, + "learning_rate": 3.82186665855894e-06, + "loss": 0.2284, + "step": 26733 + }, + { + "epoch": 0.5884657715166189, + "grad_norm": 0.6285157082687294, + "learning_rate": 3.821520241685635e-06, + "loss": 0.4824, + "step": 26734 + }, + { + "epoch": 0.5884877834030376, + "grad_norm": 1.3092698111734618, + "learning_rate": 3.821173830801891e-06, + "loss": 0.3434, + "step": 26735 + }, + { + "epoch": 0.5885097952894563, + "grad_norm": 1.4493991007506766, + "learning_rate": 3.820827425909468e-06, + "loss": 0.3318, + "step": 26736 + }, + { + "epoch": 0.5885318071758749, + "grad_norm": 2.289704542046814, + "learning_rate": 3.820481027010124e-06, + "loss": 0.2811, + "step": 26737 + }, + { + "epoch": 0.5885538190622936, + "grad_norm": 1.2026120987422044, + "learning_rate": 3.820134634105623e-06, + "loss": 0.3236, + "step": 26738 + }, + { + "epoch": 0.5885758309487124, + "grad_norm": 1.9417837264836457, + "learning_rate": 3.819788247197722e-06, + "loss": 0.3992, + "step": 26739 + }, + { + "epoch": 0.588597842835131, + "grad_norm": 1.4130550837057063, + "learning_rate": 3.819441866288186e-06, + "loss": 0.2822, + "step": 26740 + }, + { + "epoch": 0.5886198547215497, + "grad_norm": 1.212462608308072, + "learning_rate": 3.819095491378772e-06, + "loss": 0.3262, + "step": 26741 + }, + { + "epoch": 0.5886418666079684, + "grad_norm": 1.4864039807441436, + "learning_rate": 3.8187491224712405e-06, + "loss": 0.3196, + "step": 26742 + }, + { + "epoch": 0.588663878494387, + "grad_norm": 1.3071280308797466, + "learning_rate": 3.818402759567354e-06, + "loss": 0.3156, + "step": 26743 + }, + { + "epoch": 0.5886858903808057, + "grad_norm": 1.6457065259237222, + "learning_rate": 3.81805640266887e-06, + "loss": 0.3501, + "step": 26744 + }, + { + "epoch": 0.5887079022672242, + "grad_norm": 1.5953634149043223, + "learning_rate": 3.8177100517775526e-06, + "loss": 0.3269, + "step": 26745 + }, + { + "epoch": 0.588729914153643, + "grad_norm": 2.124971307401284, + "learning_rate": 3.81736370689516e-06, + "loss": 0.2674, + "step": 26746 + }, + { + "epoch": 0.5887519260400617, + "grad_norm": 3.1592740355852382, + "learning_rate": 3.8170173680234505e-06, + "loss": 0.4024, + "step": 26747 + }, + { + "epoch": 0.5887739379264803, + "grad_norm": 1.7338610279847355, + "learning_rate": 3.8166710351641875e-06, + "loss": 0.2406, + "step": 26748 + }, + { + "epoch": 0.588795949812899, + "grad_norm": 1.4349927007798768, + "learning_rate": 3.816324708319132e-06, + "loss": 0.2926, + "step": 26749 + }, + { + "epoch": 0.5888179616993177, + "grad_norm": 1.719554381136948, + "learning_rate": 3.815978387490039e-06, + "loss": 0.277, + "step": 26750 + }, + { + "epoch": 0.5888399735857363, + "grad_norm": 1.3734502204138708, + "learning_rate": 3.815632072678674e-06, + "loss": 0.2907, + "step": 26751 + }, + { + "epoch": 0.588861985472155, + "grad_norm": 2.0004216633474, + "learning_rate": 3.815285763886793e-06, + "loss": 0.3762, + "step": 26752 + }, + { + "epoch": 0.5888839973585737, + "grad_norm": 1.2930613902973491, + "learning_rate": 3.814939461116162e-06, + "loss": 0.2741, + "step": 26753 + }, + { + "epoch": 0.5889060092449923, + "grad_norm": 1.4018282277517704, + "learning_rate": 3.814593164368534e-06, + "loss": 0.3987, + "step": 26754 + }, + { + "epoch": 0.588928021131411, + "grad_norm": 0.6523497431267592, + "learning_rate": 3.8142468736456724e-06, + "loss": 0.4654, + "step": 26755 + }, + { + "epoch": 0.5889500330178297, + "grad_norm": 1.7610133614777987, + "learning_rate": 3.813900588949338e-06, + "loss": 0.2458, + "step": 26756 + }, + { + "epoch": 0.5889720449042483, + "grad_norm": 0.6337346001893877, + "learning_rate": 3.8135543102812885e-06, + "loss": 0.4638, + "step": 26757 + }, + { + "epoch": 0.588994056790667, + "grad_norm": 2.686759885361052, + "learning_rate": 3.8132080376432872e-06, + "loss": 0.3392, + "step": 26758 + }, + { + "epoch": 0.5890160686770857, + "grad_norm": 4.3551760571403975, + "learning_rate": 3.8128617710370914e-06, + "loss": 0.2758, + "step": 26759 + }, + { + "epoch": 0.5890380805635043, + "grad_norm": 1.5019947476257052, + "learning_rate": 3.81251551046446e-06, + "loss": 0.3249, + "step": 26760 + }, + { + "epoch": 0.589060092449923, + "grad_norm": 1.6810731878568963, + "learning_rate": 3.8121692559271556e-06, + "loss": 0.2597, + "step": 26761 + }, + { + "epoch": 0.5890821043363417, + "grad_norm": 0.6298540193804565, + "learning_rate": 3.8118230074269357e-06, + "loss": 0.4804, + "step": 26762 + }, + { + "epoch": 0.5891041162227603, + "grad_norm": 1.3304719631313806, + "learning_rate": 3.8114767649655636e-06, + "loss": 0.2642, + "step": 26763 + }, + { + "epoch": 0.589126128109179, + "grad_norm": 0.6147884579753533, + "learning_rate": 3.811130528544796e-06, + "loss": 0.4535, + "step": 26764 + }, + { + "epoch": 0.5891481399955977, + "grad_norm": 0.6294007224754133, + "learning_rate": 3.810784298166391e-06, + "loss": 0.4713, + "step": 26765 + }, + { + "epoch": 0.5891701518820163, + "grad_norm": 1.7551528919993504, + "learning_rate": 3.810438073832113e-06, + "loss": 0.3381, + "step": 26766 + }, + { + "epoch": 0.589192163768435, + "grad_norm": 1.3195163683913205, + "learning_rate": 3.8100918555437193e-06, + "loss": 0.3363, + "step": 26767 + }, + { + "epoch": 0.5892141756548536, + "grad_norm": 6.389294860105685, + "learning_rate": 3.809745643302968e-06, + "loss": 0.2923, + "step": 26768 + }, + { + "epoch": 0.5892361875412723, + "grad_norm": 0.588132423589378, + "learning_rate": 3.809399437111621e-06, + "loss": 0.4573, + "step": 26769 + }, + { + "epoch": 0.589258199427691, + "grad_norm": 1.773493179116749, + "learning_rate": 3.809053236971436e-06, + "loss": 0.2934, + "step": 26770 + }, + { + "epoch": 0.5892802113141096, + "grad_norm": 0.631756916333928, + "learning_rate": 3.808707042884176e-06, + "loss": 0.5092, + "step": 26771 + }, + { + "epoch": 0.5893022232005283, + "grad_norm": 1.5648487176354944, + "learning_rate": 3.808360854851597e-06, + "loss": 0.2405, + "step": 26772 + }, + { + "epoch": 0.589324235086947, + "grad_norm": 1.442401339171542, + "learning_rate": 3.8080146728754586e-06, + "loss": 0.3529, + "step": 26773 + }, + { + "epoch": 0.5893462469733656, + "grad_norm": 0.5995070897880185, + "learning_rate": 3.807668496957522e-06, + "loss": 0.4504, + "step": 26774 + }, + { + "epoch": 0.5893682588597843, + "grad_norm": 4.327347562893048, + "learning_rate": 3.807322327099545e-06, + "loss": 0.3575, + "step": 26775 + }, + { + "epoch": 0.589390270746203, + "grad_norm": 0.5974289553644643, + "learning_rate": 3.8069761633032905e-06, + "loss": 0.4638, + "step": 26776 + }, + { + "epoch": 0.5894122826326216, + "grad_norm": 1.7140024069109823, + "learning_rate": 3.8066300055705135e-06, + "loss": 0.1948, + "step": 26777 + }, + { + "epoch": 0.5894342945190403, + "grad_norm": 1.58008906985131, + "learning_rate": 3.8062838539029745e-06, + "loss": 0.3105, + "step": 26778 + }, + { + "epoch": 0.589456306405459, + "grad_norm": 0.6186088753395985, + "learning_rate": 3.8059377083024342e-06, + "loss": 0.4607, + "step": 26779 + }, + { + "epoch": 0.5894783182918776, + "grad_norm": 1.4207416704857416, + "learning_rate": 3.80559156877065e-06, + "loss": 0.2766, + "step": 26780 + }, + { + "epoch": 0.5895003301782963, + "grad_norm": 1.4372695942623035, + "learning_rate": 3.8052454353093845e-06, + "loss": 0.2992, + "step": 26781 + }, + { + "epoch": 0.589522342064715, + "grad_norm": 1.5399340987718302, + "learning_rate": 3.8048993079203927e-06, + "loss": 0.3254, + "step": 26782 + }, + { + "epoch": 0.5895443539511336, + "grad_norm": 0.6077713357188765, + "learning_rate": 3.804553186605435e-06, + "loss": 0.472, + "step": 26783 + }, + { + "epoch": 0.5895663658375523, + "grad_norm": 0.6127740209724178, + "learning_rate": 3.8042070713662726e-06, + "loss": 0.4664, + "step": 26784 + }, + { + "epoch": 0.589588377723971, + "grad_norm": 0.6285168548675416, + "learning_rate": 3.803860962204662e-06, + "loss": 0.4808, + "step": 26785 + }, + { + "epoch": 0.5896103896103896, + "grad_norm": 0.6253082264085755, + "learning_rate": 3.8035148591223657e-06, + "loss": 0.4895, + "step": 26786 + }, + { + "epoch": 0.5896324014968083, + "grad_norm": 1.34188850427878, + "learning_rate": 3.8031687621211395e-06, + "loss": 0.2662, + "step": 26787 + }, + { + "epoch": 0.589654413383227, + "grad_norm": 4.747531322417167, + "learning_rate": 3.802822671202742e-06, + "loss": 0.314, + "step": 26788 + }, + { + "epoch": 0.5896764252696456, + "grad_norm": 1.781455911736579, + "learning_rate": 3.802476586368936e-06, + "loss": 0.3506, + "step": 26789 + }, + { + "epoch": 0.5896984371560643, + "grad_norm": 1.6360375902432303, + "learning_rate": 3.802130507621477e-06, + "loss": 0.3188, + "step": 26790 + }, + { + "epoch": 0.5897204490424829, + "grad_norm": 1.5463120455686652, + "learning_rate": 3.8017844349621242e-06, + "loss": 0.2667, + "step": 26791 + }, + { + "epoch": 0.5897424609289016, + "grad_norm": 1.2178931989551063, + "learning_rate": 3.8014383683926383e-06, + "loss": 0.2857, + "step": 26792 + }, + { + "epoch": 0.5897644728153203, + "grad_norm": 2.9794366110834956, + "learning_rate": 3.8010923079147753e-06, + "loss": 0.2934, + "step": 26793 + }, + { + "epoch": 0.5897864847017389, + "grad_norm": 0.6540165470057657, + "learning_rate": 3.8007462535302992e-06, + "loss": 0.4704, + "step": 26794 + }, + { + "epoch": 0.5898084965881576, + "grad_norm": 1.2597451289267727, + "learning_rate": 3.800400205240964e-06, + "loss": 0.3216, + "step": 26795 + }, + { + "epoch": 0.5898305084745763, + "grad_norm": 1.5836977609851095, + "learning_rate": 3.8000541630485287e-06, + "loss": 0.3112, + "step": 26796 + }, + { + "epoch": 0.5898525203609949, + "grad_norm": 0.5954889603959932, + "learning_rate": 3.7997081269547546e-06, + "loss": 0.4592, + "step": 26797 + }, + { + "epoch": 0.5898745322474136, + "grad_norm": 0.6451687929258445, + "learning_rate": 3.799362096961398e-06, + "loss": 0.4711, + "step": 26798 + }, + { + "epoch": 0.5898965441338323, + "grad_norm": 1.315022519301917, + "learning_rate": 3.799016073070221e-06, + "loss": 0.3148, + "step": 26799 + }, + { + "epoch": 0.5899185560202509, + "grad_norm": 0.620492056276728, + "learning_rate": 3.7986700552829794e-06, + "loss": 0.4621, + "step": 26800 + }, + { + "epoch": 0.5899405679066696, + "grad_norm": 1.93654905952299, + "learning_rate": 3.79832404360143e-06, + "loss": 0.259, + "step": 26801 + }, + { + "epoch": 0.5899625797930883, + "grad_norm": 1.61126332378214, + "learning_rate": 3.7979780380273355e-06, + "loss": 0.377, + "step": 26802 + }, + { + "epoch": 0.5899845916795069, + "grad_norm": 1.8137084912546306, + "learning_rate": 3.797632038562452e-06, + "loss": 0.2402, + "step": 26803 + }, + { + "epoch": 0.5900066035659256, + "grad_norm": 1.5797241385272376, + "learning_rate": 3.797286045208541e-06, + "loss": 0.3296, + "step": 26804 + }, + { + "epoch": 0.5900286154523443, + "grad_norm": 1.2443730521151577, + "learning_rate": 3.796940057967357e-06, + "loss": 0.2559, + "step": 26805 + }, + { + "epoch": 0.5900506273387629, + "grad_norm": 1.381458221904044, + "learning_rate": 3.7965940768406593e-06, + "loss": 0.3382, + "step": 26806 + }, + { + "epoch": 0.5900726392251816, + "grad_norm": 0.6694204116785336, + "learning_rate": 3.79624810183021e-06, + "loss": 0.4903, + "step": 26807 + }, + { + "epoch": 0.5900946511116003, + "grad_norm": 0.6175263494196112, + "learning_rate": 3.7959021329377634e-06, + "loss": 0.4446, + "step": 26808 + }, + { + "epoch": 0.5901166629980189, + "grad_norm": 1.4770900074908004, + "learning_rate": 3.795556170165078e-06, + "loss": 0.2874, + "step": 26809 + }, + { + "epoch": 0.5901386748844376, + "grad_norm": 1.2359655112483499, + "learning_rate": 3.795210213513914e-06, + "loss": 0.3504, + "step": 26810 + }, + { + "epoch": 0.5901606867708563, + "grad_norm": 1.4942447446697738, + "learning_rate": 3.794864262986028e-06, + "loss": 0.2558, + "step": 26811 + }, + { + "epoch": 0.5901826986572749, + "grad_norm": 1.3362719611337126, + "learning_rate": 3.7945183185831826e-06, + "loss": 0.283, + "step": 26812 + }, + { + "epoch": 0.5902047105436936, + "grad_norm": 1.25442036621131, + "learning_rate": 3.7941723803071305e-06, + "loss": 0.2888, + "step": 26813 + }, + { + "epoch": 0.5902267224301123, + "grad_norm": 1.3668534989772305, + "learning_rate": 3.7938264481596316e-06, + "loss": 0.2578, + "step": 26814 + }, + { + "epoch": 0.5902487343165309, + "grad_norm": 1.5701173596066882, + "learning_rate": 3.793480522142445e-06, + "loss": 0.3616, + "step": 26815 + }, + { + "epoch": 0.5902707462029496, + "grad_norm": 1.0354512685270316, + "learning_rate": 3.7931346022573283e-06, + "loss": 0.2589, + "step": 26816 + }, + { + "epoch": 0.5902927580893682, + "grad_norm": 1.463043324715491, + "learning_rate": 3.7927886885060416e-06, + "loss": 0.3318, + "step": 26817 + }, + { + "epoch": 0.5903147699757869, + "grad_norm": 1.309395373015781, + "learning_rate": 3.792442780890341e-06, + "loss": 0.1926, + "step": 26818 + }, + { + "epoch": 0.5903367818622056, + "grad_norm": 1.8981075461107653, + "learning_rate": 3.7920968794119827e-06, + "loss": 0.367, + "step": 26819 + }, + { + "epoch": 0.5903587937486242, + "grad_norm": 0.6037160981983732, + "learning_rate": 3.791750984072728e-06, + "loss": 0.4722, + "step": 26820 + }, + { + "epoch": 0.5903808056350429, + "grad_norm": 1.991633010372287, + "learning_rate": 3.791405094874333e-06, + "loss": 0.3084, + "step": 26821 + }, + { + "epoch": 0.5904028175214616, + "grad_norm": 1.146914395142644, + "learning_rate": 3.7910592118185578e-06, + "loss": 0.3121, + "step": 26822 + }, + { + "epoch": 0.5904248294078802, + "grad_norm": 1.3631177728752382, + "learning_rate": 3.7907133349071586e-06, + "loss": 0.2513, + "step": 26823 + }, + { + "epoch": 0.5904468412942989, + "grad_norm": 1.5642903738344407, + "learning_rate": 3.790367464141892e-06, + "loss": 0.2934, + "step": 26824 + }, + { + "epoch": 0.5904688531807176, + "grad_norm": 1.204394251612968, + "learning_rate": 3.79002159952452e-06, + "loss": 0.3133, + "step": 26825 + }, + { + "epoch": 0.5904908650671362, + "grad_norm": 2.2653695765649084, + "learning_rate": 3.7896757410567964e-06, + "loss": 0.2929, + "step": 26826 + }, + { + "epoch": 0.5905128769535549, + "grad_norm": 1.3675501363826494, + "learning_rate": 3.7893298887404795e-06, + "loss": 0.3026, + "step": 26827 + }, + { + "epoch": 0.5905348888399736, + "grad_norm": 1.684610100796868, + "learning_rate": 3.7889840425773296e-06, + "loss": 0.3236, + "step": 26828 + }, + { + "epoch": 0.5905569007263922, + "grad_norm": 1.436062293321296, + "learning_rate": 3.7886382025691014e-06, + "loss": 0.2476, + "step": 26829 + }, + { + "epoch": 0.5905789126128109, + "grad_norm": 1.8389685370455144, + "learning_rate": 3.7882923687175564e-06, + "loss": 0.3316, + "step": 26830 + }, + { + "epoch": 0.5906009244992296, + "grad_norm": 1.4412279701975963, + "learning_rate": 3.787946541024449e-06, + "loss": 0.3311, + "step": 26831 + }, + { + "epoch": 0.5906229363856482, + "grad_norm": 0.6258029463379797, + "learning_rate": 3.787600719491537e-06, + "loss": 0.4846, + "step": 26832 + }, + { + "epoch": 0.5906449482720669, + "grad_norm": 1.4216728459906973, + "learning_rate": 3.78725490412058e-06, + "loss": 0.3113, + "step": 26833 + }, + { + "epoch": 0.5906669601584856, + "grad_norm": 2.825510681602293, + "learning_rate": 3.786909094913333e-06, + "loss": 0.3002, + "step": 26834 + }, + { + "epoch": 0.5906889720449042, + "grad_norm": 2.3678913082937916, + "learning_rate": 3.7865632918715573e-06, + "loss": 0.3432, + "step": 26835 + }, + { + "epoch": 0.5907109839313229, + "grad_norm": 0.6292650817290163, + "learning_rate": 3.786217494997007e-06, + "loss": 0.4659, + "step": 26836 + }, + { + "epoch": 0.5907329958177416, + "grad_norm": 1.3135779627578472, + "learning_rate": 3.7858717042914394e-06, + "loss": 0.336, + "step": 26837 + }, + { + "epoch": 0.5907550077041602, + "grad_norm": 2.7476665558628355, + "learning_rate": 3.7855259197566153e-06, + "loss": 0.3104, + "step": 26838 + }, + { + "epoch": 0.5907770195905789, + "grad_norm": 1.4821325554695364, + "learning_rate": 3.7851801413942884e-06, + "loss": 0.2812, + "step": 26839 + }, + { + "epoch": 0.5907990314769975, + "grad_norm": 1.3509953631494624, + "learning_rate": 3.78483436920622e-06, + "loss": 0.2347, + "step": 26840 + }, + { + "epoch": 0.5908210433634162, + "grad_norm": 1.9082794957470202, + "learning_rate": 3.7844886031941647e-06, + "loss": 0.3476, + "step": 26841 + }, + { + "epoch": 0.5908430552498349, + "grad_norm": 1.4223575682330487, + "learning_rate": 3.7841428433598793e-06, + "loss": 0.3632, + "step": 26842 + }, + { + "epoch": 0.5908650671362535, + "grad_norm": 1.7265253887110763, + "learning_rate": 3.7837970897051236e-06, + "loss": 0.3392, + "step": 26843 + }, + { + "epoch": 0.5908870790226722, + "grad_norm": 10.864923401396439, + "learning_rate": 3.783451342231651e-06, + "loss": 0.2784, + "step": 26844 + }, + { + "epoch": 0.5909090909090909, + "grad_norm": 0.6118375448722968, + "learning_rate": 3.783105600941225e-06, + "loss": 0.4517, + "step": 26845 + }, + { + "epoch": 0.5909311027955095, + "grad_norm": 1.6058700188092165, + "learning_rate": 3.7827598658355967e-06, + "loss": 0.3382, + "step": 26846 + }, + { + "epoch": 0.5909531146819282, + "grad_norm": 1.3705703461770309, + "learning_rate": 3.7824141369165257e-06, + "loss": 0.291, + "step": 26847 + }, + { + "epoch": 0.5909751265683469, + "grad_norm": 1.7659005573064022, + "learning_rate": 3.7820684141857706e-06, + "loss": 0.3043, + "step": 26848 + }, + { + "epoch": 0.5909971384547655, + "grad_norm": 1.6557895485282068, + "learning_rate": 3.781722697645086e-06, + "loss": 0.2714, + "step": 26849 + }, + { + "epoch": 0.5910191503411842, + "grad_norm": 1.40746502069421, + "learning_rate": 3.781376987296229e-06, + "loss": 0.2917, + "step": 26850 + }, + { + "epoch": 0.591041162227603, + "grad_norm": 0.6286086322864313, + "learning_rate": 3.781031283140959e-06, + "loss": 0.4785, + "step": 26851 + }, + { + "epoch": 0.5910631741140215, + "grad_norm": 1.9238511276541073, + "learning_rate": 3.780685585181031e-06, + "loss": 0.3231, + "step": 26852 + }, + { + "epoch": 0.5910851860004402, + "grad_norm": 1.4841095587159998, + "learning_rate": 3.780339893418204e-06, + "loss": 0.2462, + "step": 26853 + }, + { + "epoch": 0.591107197886859, + "grad_norm": 1.3097037013243467, + "learning_rate": 3.779994207854233e-06, + "loss": 0.2789, + "step": 26854 + }, + { + "epoch": 0.5911292097732775, + "grad_norm": 1.2482498630078238, + "learning_rate": 3.7796485284908747e-06, + "loss": 0.3148, + "step": 26855 + }, + { + "epoch": 0.5911512216596962, + "grad_norm": 1.347381930468577, + "learning_rate": 3.779302855329887e-06, + "loss": 0.2987, + "step": 26856 + }, + { + "epoch": 0.591173233546115, + "grad_norm": 1.2790919908606893, + "learning_rate": 3.778957188373027e-06, + "loss": 0.2538, + "step": 26857 + }, + { + "epoch": 0.5911952454325335, + "grad_norm": 1.5158181554492969, + "learning_rate": 3.7786115276220526e-06, + "loss": 0.3444, + "step": 26858 + }, + { + "epoch": 0.5912172573189522, + "grad_norm": 1.3360686836977802, + "learning_rate": 3.7782658730787174e-06, + "loss": 0.3432, + "step": 26859 + }, + { + "epoch": 0.591239269205371, + "grad_norm": 1.3906853397365035, + "learning_rate": 3.77792022474478e-06, + "loss": 0.3694, + "step": 26860 + }, + { + "epoch": 0.5912612810917895, + "grad_norm": 1.987809237611576, + "learning_rate": 3.7775745826219983e-06, + "loss": 0.2973, + "step": 26861 + }, + { + "epoch": 0.5912832929782083, + "grad_norm": 1.5926617822299654, + "learning_rate": 3.777228946712125e-06, + "loss": 0.3031, + "step": 26862 + }, + { + "epoch": 0.5913053048646268, + "grad_norm": 1.423757470143943, + "learning_rate": 3.7768833170169226e-06, + "loss": 0.2498, + "step": 26863 + }, + { + "epoch": 0.5913273167510456, + "grad_norm": 1.4682372036662126, + "learning_rate": 3.776537693538144e-06, + "loss": 0.352, + "step": 26864 + }, + { + "epoch": 0.5913493286374643, + "grad_norm": 1.6243164592175927, + "learning_rate": 3.7761920762775456e-06, + "loss": 0.3162, + "step": 26865 + }, + { + "epoch": 0.5913713405238828, + "grad_norm": 1.909134642340648, + "learning_rate": 3.7758464652368865e-06, + "loss": 0.2978, + "step": 26866 + }, + { + "epoch": 0.5913933524103016, + "grad_norm": 1.1343761015972362, + "learning_rate": 3.7755008604179212e-06, + "loss": 0.1928, + "step": 26867 + }, + { + "epoch": 0.5914153642967203, + "grad_norm": 1.3729256010311737, + "learning_rate": 3.7751552618224047e-06, + "loss": 0.3015, + "step": 26868 + }, + { + "epoch": 0.5914373761831389, + "grad_norm": 1.9803393513399679, + "learning_rate": 3.7748096694520976e-06, + "loss": 0.2783, + "step": 26869 + }, + { + "epoch": 0.5914593880695576, + "grad_norm": 0.6318777853896588, + "learning_rate": 3.774464083308752e-06, + "loss": 0.4828, + "step": 26870 + }, + { + "epoch": 0.5914813999559763, + "grad_norm": 0.6415464615500799, + "learning_rate": 3.77411850339413e-06, + "loss": 0.486, + "step": 26871 + }, + { + "epoch": 0.5915034118423949, + "grad_norm": 2.6459719227945486, + "learning_rate": 3.7737729297099823e-06, + "loss": 0.2912, + "step": 26872 + }, + { + "epoch": 0.5915254237288136, + "grad_norm": 1.495497535700887, + "learning_rate": 3.7734273622580665e-06, + "loss": 0.3383, + "step": 26873 + }, + { + "epoch": 0.5915474356152323, + "grad_norm": 6.765939498624773, + "learning_rate": 3.7730818010401416e-06, + "loss": 0.3981, + "step": 26874 + }, + { + "epoch": 0.5915694475016509, + "grad_norm": 1.5911542601631021, + "learning_rate": 3.7727362460579606e-06, + "loss": 0.2727, + "step": 26875 + }, + { + "epoch": 0.5915914593880696, + "grad_norm": 1.7767489398962577, + "learning_rate": 3.772390697313284e-06, + "loss": 0.2483, + "step": 26876 + }, + { + "epoch": 0.5916134712744883, + "grad_norm": 1.6564557728209797, + "learning_rate": 3.772045154807864e-06, + "loss": 0.2503, + "step": 26877 + }, + { + "epoch": 0.5916354831609069, + "grad_norm": 1.8028015641372983, + "learning_rate": 3.7716996185434567e-06, + "loss": 0.3019, + "step": 26878 + }, + { + "epoch": 0.5916574950473256, + "grad_norm": 1.715890696312368, + "learning_rate": 3.7713540885218226e-06, + "loss": 0.2641, + "step": 26879 + }, + { + "epoch": 0.5916795069337443, + "grad_norm": 0.6125197095750489, + "learning_rate": 3.7710085647447115e-06, + "loss": 0.459, + "step": 26880 + }, + { + "epoch": 0.5917015188201629, + "grad_norm": 1.4925892389730087, + "learning_rate": 3.7706630472138865e-06, + "loss": 0.2988, + "step": 26881 + }, + { + "epoch": 0.5917235307065816, + "grad_norm": 1.4205285011051427, + "learning_rate": 3.7703175359310988e-06, + "loss": 0.2722, + "step": 26882 + }, + { + "epoch": 0.5917455425930003, + "grad_norm": 3.076745106180931, + "learning_rate": 3.769972030898105e-06, + "loss": 0.3024, + "step": 26883 + }, + { + "epoch": 0.5917675544794189, + "grad_norm": 1.2938800730196602, + "learning_rate": 3.7696265321166644e-06, + "loss": 0.2938, + "step": 26884 + }, + { + "epoch": 0.5917895663658376, + "grad_norm": 0.6046413677076268, + "learning_rate": 3.769281039588528e-06, + "loss": 0.4495, + "step": 26885 + }, + { + "epoch": 0.5918115782522562, + "grad_norm": 2.1280182490277575, + "learning_rate": 3.7689355533154555e-06, + "loss": 0.2602, + "step": 26886 + }, + { + "epoch": 0.5918335901386749, + "grad_norm": 1.6888834598841127, + "learning_rate": 3.768590073299201e-06, + "loss": 0.3915, + "step": 26887 + }, + { + "epoch": 0.5918556020250936, + "grad_norm": 1.4321874581081817, + "learning_rate": 3.7682445995415207e-06, + "loss": 0.2558, + "step": 26888 + }, + { + "epoch": 0.5918776139115122, + "grad_norm": 1.422937074535162, + "learning_rate": 3.767899132044172e-06, + "loss": 0.2413, + "step": 26889 + }, + { + "epoch": 0.5918996257979309, + "grad_norm": 1.7002109835016852, + "learning_rate": 3.767553670808909e-06, + "loss": 0.2859, + "step": 26890 + }, + { + "epoch": 0.5919216376843496, + "grad_norm": 1.7062412933256133, + "learning_rate": 3.767208215837487e-06, + "loss": 0.2955, + "step": 26891 + }, + { + "epoch": 0.5919436495707682, + "grad_norm": 2.632599622873781, + "learning_rate": 3.7668627671316633e-06, + "loss": 0.3144, + "step": 26892 + }, + { + "epoch": 0.5919656614571869, + "grad_norm": 0.6094644255840754, + "learning_rate": 3.7665173246931923e-06, + "loss": 0.4576, + "step": 26893 + }, + { + "epoch": 0.5919876733436056, + "grad_norm": 1.9760164424946662, + "learning_rate": 3.7661718885238324e-06, + "loss": 0.3695, + "step": 26894 + }, + { + "epoch": 0.5920096852300242, + "grad_norm": 0.6015907288609165, + "learning_rate": 3.765826458625336e-06, + "loss": 0.4452, + "step": 26895 + }, + { + "epoch": 0.5920316971164429, + "grad_norm": 0.612933320636711, + "learning_rate": 3.7654810349994593e-06, + "loss": 0.4622, + "step": 26896 + }, + { + "epoch": 0.5920537090028616, + "grad_norm": 1.6556850941676169, + "learning_rate": 3.7651356176479604e-06, + "loss": 0.2077, + "step": 26897 + }, + { + "epoch": 0.5920757208892802, + "grad_norm": 2.941602283945432, + "learning_rate": 3.76479020657259e-06, + "loss": 0.334, + "step": 26898 + }, + { + "epoch": 0.5920977327756989, + "grad_norm": 4.672389236086681, + "learning_rate": 3.7644448017751102e-06, + "loss": 0.4403, + "step": 26899 + }, + { + "epoch": 0.5921197446621176, + "grad_norm": 1.3541679872486632, + "learning_rate": 3.7640994032572716e-06, + "loss": 0.2706, + "step": 26900 + }, + { + "epoch": 0.5921417565485362, + "grad_norm": 3.0060311475412638, + "learning_rate": 3.76375401102083e-06, + "loss": 0.2825, + "step": 26901 + }, + { + "epoch": 0.5921637684349549, + "grad_norm": 1.283084405926077, + "learning_rate": 3.7634086250675438e-06, + "loss": 0.2683, + "step": 26902 + }, + { + "epoch": 0.5921857803213736, + "grad_norm": 1.3053368731779833, + "learning_rate": 3.7630632453991643e-06, + "loss": 0.2405, + "step": 26903 + }, + { + "epoch": 0.5922077922077922, + "grad_norm": 1.8170006532622063, + "learning_rate": 3.7627178720174506e-06, + "loss": 0.3718, + "step": 26904 + }, + { + "epoch": 0.5922298040942109, + "grad_norm": 1.6597423922035246, + "learning_rate": 3.7623725049241565e-06, + "loss": 0.3, + "step": 26905 + }, + { + "epoch": 0.5922518159806296, + "grad_norm": 4.3478070899923, + "learning_rate": 3.762027144121036e-06, + "loss": 0.3202, + "step": 26906 + }, + { + "epoch": 0.5922738278670482, + "grad_norm": 1.6027137235469684, + "learning_rate": 3.761681789609848e-06, + "loss": 0.3719, + "step": 26907 + }, + { + "epoch": 0.5922958397534669, + "grad_norm": 1.9601974463019245, + "learning_rate": 3.7613364413923442e-06, + "loss": 0.2359, + "step": 26908 + }, + { + "epoch": 0.5923178516398855, + "grad_norm": 1.969986050378107, + "learning_rate": 3.7609910994702804e-06, + "loss": 0.2839, + "step": 26909 + }, + { + "epoch": 0.5923398635263042, + "grad_norm": 1.20236287959788, + "learning_rate": 3.760645763845413e-06, + "loss": 0.2057, + "step": 26910 + }, + { + "epoch": 0.5923618754127229, + "grad_norm": 1.2514072297684051, + "learning_rate": 3.760300434519496e-06, + "loss": 0.2677, + "step": 26911 + }, + { + "epoch": 0.5923838872991415, + "grad_norm": 1.3789086344654389, + "learning_rate": 3.759955111494287e-06, + "loss": 0.295, + "step": 26912 + }, + { + "epoch": 0.5924058991855602, + "grad_norm": 1.5476640989219044, + "learning_rate": 3.7596097947715383e-06, + "loss": 0.3301, + "step": 26913 + }, + { + "epoch": 0.5924279110719789, + "grad_norm": 1.4646984545841781, + "learning_rate": 3.7592644843530044e-06, + "loss": 0.2386, + "step": 26914 + }, + { + "epoch": 0.5924499229583975, + "grad_norm": 0.630814702888096, + "learning_rate": 3.7589191802404445e-06, + "loss": 0.4674, + "step": 26915 + }, + { + "epoch": 0.5924719348448162, + "grad_norm": 1.4866035335173529, + "learning_rate": 3.7585738824356073e-06, + "loss": 0.3398, + "step": 26916 + }, + { + "epoch": 0.5924939467312349, + "grad_norm": 1.6053253923414181, + "learning_rate": 3.7582285909402545e-06, + "loss": 0.3275, + "step": 26917 + }, + { + "epoch": 0.5925159586176535, + "grad_norm": 1.715121527873407, + "learning_rate": 3.7578833057561365e-06, + "loss": 0.358, + "step": 26918 + }, + { + "epoch": 0.5925379705040722, + "grad_norm": 2.2305617445401, + "learning_rate": 3.7575380268850094e-06, + "loss": 0.3418, + "step": 26919 + }, + { + "epoch": 0.5925599823904909, + "grad_norm": 1.8255305187828883, + "learning_rate": 3.7571927543286293e-06, + "loss": 0.368, + "step": 26920 + }, + { + "epoch": 0.5925819942769095, + "grad_norm": 1.516928946918685, + "learning_rate": 3.7568474880887484e-06, + "loss": 0.2581, + "step": 26921 + }, + { + "epoch": 0.5926040061633282, + "grad_norm": 1.6139611173366881, + "learning_rate": 3.756502228167124e-06, + "loss": 0.3118, + "step": 26922 + }, + { + "epoch": 0.5926260180497469, + "grad_norm": 1.3120770498382186, + "learning_rate": 3.75615697456551e-06, + "loss": 0.2556, + "step": 26923 + }, + { + "epoch": 0.5926480299361655, + "grad_norm": 1.3841733810339194, + "learning_rate": 3.7558117272856593e-06, + "loss": 0.2941, + "step": 26924 + }, + { + "epoch": 0.5926700418225842, + "grad_norm": 1.8384973919956535, + "learning_rate": 3.755466486329331e-06, + "loss": 0.2943, + "step": 26925 + }, + { + "epoch": 0.5926920537090029, + "grad_norm": 2.0875782980679576, + "learning_rate": 3.755121251698276e-06, + "loss": 0.3224, + "step": 26926 + }, + { + "epoch": 0.5927140655954215, + "grad_norm": 0.5912994986921049, + "learning_rate": 3.754776023394249e-06, + "loss": 0.447, + "step": 26927 + }, + { + "epoch": 0.5927360774818402, + "grad_norm": 1.4231629296203414, + "learning_rate": 3.7544308014190063e-06, + "loss": 0.2436, + "step": 26928 + }, + { + "epoch": 0.5927580893682589, + "grad_norm": 1.186971717397006, + "learning_rate": 3.7540855857743007e-06, + "loss": 0.2941, + "step": 26929 + }, + { + "epoch": 0.5927801012546775, + "grad_norm": 1.9514078180112, + "learning_rate": 3.7537403764618905e-06, + "loss": 0.2762, + "step": 26930 + }, + { + "epoch": 0.5928021131410962, + "grad_norm": 1.2862639202616868, + "learning_rate": 3.7533951734835267e-06, + "loss": 0.2555, + "step": 26931 + }, + { + "epoch": 0.5928241250275148, + "grad_norm": 2.073431801775938, + "learning_rate": 3.753049976840963e-06, + "loss": 0.2214, + "step": 26932 + }, + { + "epoch": 0.5928461369139335, + "grad_norm": 1.3506454713369112, + "learning_rate": 3.752704786535958e-06, + "loss": 0.3015, + "step": 26933 + }, + { + "epoch": 0.5928681488003522, + "grad_norm": 2.773029043072685, + "learning_rate": 3.7523596025702603e-06, + "loss": 0.3179, + "step": 26934 + }, + { + "epoch": 0.5928901606867708, + "grad_norm": 1.4935358562583143, + "learning_rate": 3.75201442494563e-06, + "loss": 0.3021, + "step": 26935 + }, + { + "epoch": 0.5929121725731895, + "grad_norm": 1.4819598762049015, + "learning_rate": 3.751669253663819e-06, + "loss": 0.2816, + "step": 26936 + }, + { + "epoch": 0.5929341844596082, + "grad_norm": 1.4692481150971823, + "learning_rate": 3.75132408872658e-06, + "loss": 0.2318, + "step": 26937 + }, + { + "epoch": 0.5929561963460268, + "grad_norm": 1.3999644339015753, + "learning_rate": 3.7509789301356716e-06, + "loss": 0.3869, + "step": 26938 + }, + { + "epoch": 0.5929782082324455, + "grad_norm": 0.6099280544206357, + "learning_rate": 3.7506337778928423e-06, + "loss": 0.4475, + "step": 26939 + }, + { + "epoch": 0.5930002201188642, + "grad_norm": 1.2386800191463896, + "learning_rate": 3.750288631999851e-06, + "loss": 0.2728, + "step": 26940 + }, + { + "epoch": 0.5930222320052828, + "grad_norm": 1.917392279451373, + "learning_rate": 3.74994349245845e-06, + "loss": 0.2958, + "step": 26941 + }, + { + "epoch": 0.5930442438917015, + "grad_norm": 1.4215217263543107, + "learning_rate": 3.7495983592703934e-06, + "loss": 0.3553, + "step": 26942 + }, + { + "epoch": 0.5930662557781202, + "grad_norm": 1.5584653275387488, + "learning_rate": 3.749253232437437e-06, + "loss": 0.2979, + "step": 26943 + }, + { + "epoch": 0.5930882676645388, + "grad_norm": 1.5887056654588982, + "learning_rate": 3.7489081119613313e-06, + "loss": 0.2535, + "step": 26944 + }, + { + "epoch": 0.5931102795509575, + "grad_norm": 1.3685163597204735, + "learning_rate": 3.7485629978438345e-06, + "loss": 0.2275, + "step": 26945 + }, + { + "epoch": 0.5931322914373762, + "grad_norm": 1.2365152411998221, + "learning_rate": 3.7482178900866988e-06, + "loss": 0.3017, + "step": 26946 + }, + { + "epoch": 0.5931543033237948, + "grad_norm": 1.479248630886251, + "learning_rate": 3.7478727886916754e-06, + "loss": 0.2732, + "step": 26947 + }, + { + "epoch": 0.5931763152102135, + "grad_norm": 1.4024535454188438, + "learning_rate": 3.7475276936605244e-06, + "loss": 0.2565, + "step": 26948 + }, + { + "epoch": 0.5931983270966322, + "grad_norm": 3.3886160136555876, + "learning_rate": 3.7471826049949942e-06, + "loss": 0.2669, + "step": 26949 + }, + { + "epoch": 0.5932203389830508, + "grad_norm": 1.5854149924843604, + "learning_rate": 3.7468375226968407e-06, + "loss": 0.3417, + "step": 26950 + }, + { + "epoch": 0.5932423508694695, + "grad_norm": 0.6304005017249821, + "learning_rate": 3.74649244676782e-06, + "loss": 0.468, + "step": 26951 + }, + { + "epoch": 0.5932643627558882, + "grad_norm": 1.784575280878601, + "learning_rate": 3.7461473772096803e-06, + "loss": 0.2584, + "step": 26952 + }, + { + "epoch": 0.5932863746423068, + "grad_norm": 1.739606569552087, + "learning_rate": 3.745802314024182e-06, + "loss": 0.308, + "step": 26953 + }, + { + "epoch": 0.5933083865287255, + "grad_norm": 2.141589175611499, + "learning_rate": 3.7454572572130743e-06, + "loss": 0.3467, + "step": 26954 + }, + { + "epoch": 0.5933303984151441, + "grad_norm": 0.5964642738428348, + "learning_rate": 3.7451122067781115e-06, + "loss": 0.4652, + "step": 26955 + }, + { + "epoch": 0.5933524103015628, + "grad_norm": 1.747356306826039, + "learning_rate": 3.74476716272105e-06, + "loss": 0.3379, + "step": 26956 + }, + { + "epoch": 0.5933744221879815, + "grad_norm": 1.5226405488287107, + "learning_rate": 3.74442212504364e-06, + "loss": 0.3411, + "step": 26957 + }, + { + "epoch": 0.5933964340744001, + "grad_norm": 4.756269495741938, + "learning_rate": 3.744077093747638e-06, + "loss": 0.3523, + "step": 26958 + }, + { + "epoch": 0.5934184459608188, + "grad_norm": 1.2582196244520272, + "learning_rate": 3.743732068834796e-06, + "loss": 0.306, + "step": 26959 + }, + { + "epoch": 0.5934404578472375, + "grad_norm": 5.5928737464909295, + "learning_rate": 3.743387050306867e-06, + "loss": 0.3095, + "step": 26960 + }, + { + "epoch": 0.5934624697336561, + "grad_norm": 1.2487797620288257, + "learning_rate": 3.7430420381656075e-06, + "loss": 0.2812, + "step": 26961 + }, + { + "epoch": 0.5934844816200748, + "grad_norm": 1.6628941642986246, + "learning_rate": 3.7426970324127676e-06, + "loss": 0.4354, + "step": 26962 + }, + { + "epoch": 0.5935064935064935, + "grad_norm": 0.6272771850110419, + "learning_rate": 3.742352033050102e-06, + "loss": 0.4739, + "step": 26963 + }, + { + "epoch": 0.5935285053929121, + "grad_norm": 0.6112204913199965, + "learning_rate": 3.742007040079366e-06, + "loss": 0.4659, + "step": 26964 + }, + { + "epoch": 0.5935505172793308, + "grad_norm": 2.6124547803845606, + "learning_rate": 3.7416620535023083e-06, + "loss": 0.2331, + "step": 26965 + }, + { + "epoch": 0.5935725291657495, + "grad_norm": 1.9549478283567574, + "learning_rate": 3.7413170733206884e-06, + "loss": 0.3461, + "step": 26966 + }, + { + "epoch": 0.5935945410521681, + "grad_norm": 0.6255732942478988, + "learning_rate": 3.7409720995362552e-06, + "loss": 0.4704, + "step": 26967 + }, + { + "epoch": 0.5936165529385868, + "grad_norm": 0.6352279342861806, + "learning_rate": 3.740627132150762e-06, + "loss": 0.5109, + "step": 26968 + }, + { + "epoch": 0.5936385648250055, + "grad_norm": 1.6628034360983621, + "learning_rate": 3.7402821711659657e-06, + "loss": 0.4589, + "step": 26969 + }, + { + "epoch": 0.5936605767114241, + "grad_norm": 1.4601943044819206, + "learning_rate": 3.7399372165836145e-06, + "loss": 0.2836, + "step": 26970 + }, + { + "epoch": 0.5936825885978428, + "grad_norm": 1.3422901452500737, + "learning_rate": 3.7395922684054675e-06, + "loss": 0.2803, + "step": 26971 + }, + { + "epoch": 0.5937046004842615, + "grad_norm": 1.470819126691016, + "learning_rate": 3.739247326633273e-06, + "loss": 0.2978, + "step": 26972 + }, + { + "epoch": 0.5937266123706801, + "grad_norm": 0.6398742215142235, + "learning_rate": 3.7389023912687854e-06, + "loss": 0.4831, + "step": 26973 + }, + { + "epoch": 0.5937486242570988, + "grad_norm": 1.292486474490224, + "learning_rate": 3.7385574623137606e-06, + "loss": 0.4066, + "step": 26974 + }, + { + "epoch": 0.5937706361435175, + "grad_norm": 1.569868772234196, + "learning_rate": 3.7382125397699465e-06, + "loss": 0.2652, + "step": 26975 + }, + { + "epoch": 0.5937926480299361, + "grad_norm": 5.120488255294742, + "learning_rate": 3.737867623639101e-06, + "loss": 0.3956, + "step": 26976 + }, + { + "epoch": 0.5938146599163548, + "grad_norm": 1.4846777251374708, + "learning_rate": 3.737522713922975e-06, + "loss": 0.3298, + "step": 26977 + }, + { + "epoch": 0.5938366718027736, + "grad_norm": 1.1648716827807075, + "learning_rate": 3.7371778106233203e-06, + "loss": 0.3455, + "step": 26978 + }, + { + "epoch": 0.5938586836891921, + "grad_norm": 4.42705845432434, + "learning_rate": 3.7368329137418935e-06, + "loss": 0.2254, + "step": 26979 + }, + { + "epoch": 0.5938806955756109, + "grad_norm": 2.236540613968058, + "learning_rate": 3.7364880232804424e-06, + "loss": 0.4181, + "step": 26980 + }, + { + "epoch": 0.5939027074620294, + "grad_norm": 0.6401355437837165, + "learning_rate": 3.7361431392407242e-06, + "loss": 0.4508, + "step": 26981 + }, + { + "epoch": 0.5939247193484481, + "grad_norm": 2.0209319196906677, + "learning_rate": 3.735798261624492e-06, + "loss": 0.37, + "step": 26982 + }, + { + "epoch": 0.5939467312348669, + "grad_norm": 1.5193230012280259, + "learning_rate": 3.7354533904334934e-06, + "loss": 0.2888, + "step": 26983 + }, + { + "epoch": 0.5939687431212854, + "grad_norm": 1.4712749960874623, + "learning_rate": 3.7351085256694876e-06, + "loss": 0.3368, + "step": 26984 + }, + { + "epoch": 0.5939907550077042, + "grad_norm": 1.0762700480403, + "learning_rate": 3.7347636673342225e-06, + "loss": 0.2517, + "step": 26985 + }, + { + "epoch": 0.5940127668941229, + "grad_norm": 1.553002023774677, + "learning_rate": 3.7344188154294537e-06, + "loss": 0.3386, + "step": 26986 + }, + { + "epoch": 0.5940347787805415, + "grad_norm": 0.599983759830728, + "learning_rate": 3.734073969956934e-06, + "loss": 0.4383, + "step": 26987 + }, + { + "epoch": 0.5940567906669602, + "grad_norm": 1.7774525889951462, + "learning_rate": 3.7337291309184115e-06, + "loss": 0.291, + "step": 26988 + }, + { + "epoch": 0.5940788025533789, + "grad_norm": 1.4071950446796546, + "learning_rate": 3.7333842983156453e-06, + "loss": 0.2721, + "step": 26989 + }, + { + "epoch": 0.5941008144397975, + "grad_norm": 1.746897443320718, + "learning_rate": 3.733039472150385e-06, + "loss": 0.2777, + "step": 26990 + }, + { + "epoch": 0.5941228263262162, + "grad_norm": 1.3113525006080122, + "learning_rate": 3.7326946524243812e-06, + "loss": 0.3502, + "step": 26991 + }, + { + "epoch": 0.5941448382126349, + "grad_norm": 0.6151846404834315, + "learning_rate": 3.7323498391393914e-06, + "loss": 0.4586, + "step": 26992 + }, + { + "epoch": 0.5941668500990535, + "grad_norm": 1.3365095626205539, + "learning_rate": 3.7320050322971625e-06, + "loss": 0.2554, + "step": 26993 + }, + { + "epoch": 0.5941888619854722, + "grad_norm": 1.852142857544096, + "learning_rate": 3.7316602318994504e-06, + "loss": 0.3544, + "step": 26994 + }, + { + "epoch": 0.5942108738718909, + "grad_norm": 1.449737989087417, + "learning_rate": 3.731315437948007e-06, + "loss": 0.2873, + "step": 26995 + }, + { + "epoch": 0.5942328857583095, + "grad_norm": 4.702461516546614, + "learning_rate": 3.730970650444583e-06, + "loss": 0.2404, + "step": 26996 + }, + { + "epoch": 0.5942548976447282, + "grad_norm": 1.2018272340340386, + "learning_rate": 3.7306258693909345e-06, + "loss": 0.2631, + "step": 26997 + }, + { + "epoch": 0.5942769095311469, + "grad_norm": 2.8296692822843106, + "learning_rate": 3.7302810947888104e-06, + "loss": 0.2701, + "step": 26998 + }, + { + "epoch": 0.5942989214175655, + "grad_norm": 1.2078936766323811, + "learning_rate": 3.7299363266399634e-06, + "loss": 0.3645, + "step": 26999 + }, + { + "epoch": 0.5943209333039842, + "grad_norm": 1.536675814982565, + "learning_rate": 3.7295915649461482e-06, + "loss": 0.3349, + "step": 27000 + }, + { + "epoch": 0.5943429451904029, + "grad_norm": 1.477076450098991, + "learning_rate": 3.7292468097091128e-06, + "loss": 0.2213, + "step": 27001 + }, + { + "epoch": 0.5943649570768215, + "grad_norm": 0.6353962458589033, + "learning_rate": 3.7289020609306138e-06, + "loss": 0.4892, + "step": 27002 + }, + { + "epoch": 0.5943869689632402, + "grad_norm": 1.7456289587631073, + "learning_rate": 3.7285573186124003e-06, + "loss": 0.4148, + "step": 27003 + }, + { + "epoch": 0.5944089808496588, + "grad_norm": 0.6030372716918622, + "learning_rate": 3.7282125827562266e-06, + "loss": 0.457, + "step": 27004 + }, + { + "epoch": 0.5944309927360775, + "grad_norm": 1.6115453372950557, + "learning_rate": 3.7278678533638448e-06, + "loss": 0.2811, + "step": 27005 + }, + { + "epoch": 0.5944530046224962, + "grad_norm": 0.6567651037863759, + "learning_rate": 3.727523130437002e-06, + "loss": 0.5074, + "step": 27006 + }, + { + "epoch": 0.5944750165089148, + "grad_norm": 0.6298278787090414, + "learning_rate": 3.7271784139774582e-06, + "loss": 0.4565, + "step": 27007 + }, + { + "epoch": 0.5944970283953335, + "grad_norm": 3.899543847324677, + "learning_rate": 3.7268337039869594e-06, + "loss": 0.251, + "step": 27008 + }, + { + "epoch": 0.5945190402817522, + "grad_norm": 1.2225322547746038, + "learning_rate": 3.7264890004672595e-06, + "loss": 0.224, + "step": 27009 + }, + { + "epoch": 0.5945410521681708, + "grad_norm": 1.2864950917514977, + "learning_rate": 3.7261443034201123e-06, + "loss": 0.221, + "step": 27010 + }, + { + "epoch": 0.5945630640545895, + "grad_norm": 1.251890821990628, + "learning_rate": 3.7257996128472657e-06, + "loss": 0.204, + "step": 27011 + }, + { + "epoch": 0.5945850759410082, + "grad_norm": 1.4725681121703478, + "learning_rate": 3.7254549287504747e-06, + "loss": 0.3261, + "step": 27012 + }, + { + "epoch": 0.5946070878274268, + "grad_norm": 1.2893310768815276, + "learning_rate": 3.7251102511314906e-06, + "loss": 0.2934, + "step": 27013 + }, + { + "epoch": 0.5946290997138455, + "grad_norm": 1.5815946142112989, + "learning_rate": 3.7247655799920633e-06, + "loss": 0.3537, + "step": 27014 + }, + { + "epoch": 0.5946511116002642, + "grad_norm": 2.085386301106503, + "learning_rate": 3.7244209153339487e-06, + "loss": 0.3163, + "step": 27015 + }, + { + "epoch": 0.5946731234866828, + "grad_norm": 1.2232311947368568, + "learning_rate": 3.724076257158894e-06, + "loss": 0.2619, + "step": 27016 + }, + { + "epoch": 0.5946951353731015, + "grad_norm": 1.5142465862786838, + "learning_rate": 3.7237316054686534e-06, + "loss": 0.2975, + "step": 27017 + }, + { + "epoch": 0.5947171472595202, + "grad_norm": 1.2538977254645538, + "learning_rate": 3.7233869602649785e-06, + "loss": 0.2895, + "step": 27018 + }, + { + "epoch": 0.5947391591459388, + "grad_norm": 0.6542150153731799, + "learning_rate": 3.723042321549618e-06, + "loss": 0.4476, + "step": 27019 + }, + { + "epoch": 0.5947611710323575, + "grad_norm": 1.652018650332328, + "learning_rate": 3.7226976893243293e-06, + "loss": 0.3249, + "step": 27020 + }, + { + "epoch": 0.5947831829187762, + "grad_norm": 1.6924449706253053, + "learning_rate": 3.722353063590859e-06, + "loss": 0.2517, + "step": 27021 + }, + { + "epoch": 0.5948051948051948, + "grad_norm": 1.2788372226576588, + "learning_rate": 3.7220084443509603e-06, + "loss": 0.2843, + "step": 27022 + }, + { + "epoch": 0.5948272066916135, + "grad_norm": 4.0434362003196505, + "learning_rate": 3.721663831606386e-06, + "loss": 0.3565, + "step": 27023 + }, + { + "epoch": 0.5948492185780322, + "grad_norm": 0.6105753590268067, + "learning_rate": 3.7213192253588832e-06, + "loss": 0.4451, + "step": 27024 + }, + { + "epoch": 0.5948712304644508, + "grad_norm": 0.6198788625312466, + "learning_rate": 3.7209746256102095e-06, + "loss": 0.4618, + "step": 27025 + }, + { + "epoch": 0.5948932423508695, + "grad_norm": 1.5221948075450966, + "learning_rate": 3.7206300323621126e-06, + "loss": 0.2508, + "step": 27026 + }, + { + "epoch": 0.5949152542372881, + "grad_norm": 0.6185340314911971, + "learning_rate": 3.720285445616343e-06, + "loss": 0.4748, + "step": 27027 + }, + { + "epoch": 0.5949372661237068, + "grad_norm": 2.1695324849262017, + "learning_rate": 3.719940865374655e-06, + "loss": 0.2855, + "step": 27028 + }, + { + "epoch": 0.5949592780101255, + "grad_norm": 0.6160635958662669, + "learning_rate": 3.7195962916387972e-06, + "loss": 0.4523, + "step": 27029 + }, + { + "epoch": 0.5949812898965441, + "grad_norm": 1.4748147164365608, + "learning_rate": 3.719251724410523e-06, + "loss": 0.3886, + "step": 27030 + }, + { + "epoch": 0.5950033017829628, + "grad_norm": 1.5763278631997444, + "learning_rate": 3.718907163691583e-06, + "loss": 0.3658, + "step": 27031 + }, + { + "epoch": 0.5950253136693815, + "grad_norm": 1.9674160343947604, + "learning_rate": 3.718562609483727e-06, + "loss": 0.2438, + "step": 27032 + }, + { + "epoch": 0.5950473255558001, + "grad_norm": 1.7307449457297799, + "learning_rate": 3.718218061788709e-06, + "loss": 0.2443, + "step": 27033 + }, + { + "epoch": 0.5950693374422188, + "grad_norm": 2.173972796646707, + "learning_rate": 3.7178735206082773e-06, + "loss": 0.3629, + "step": 27034 + }, + { + "epoch": 0.5950913493286375, + "grad_norm": 1.5438237562908406, + "learning_rate": 3.7175289859441842e-06, + "loss": 0.3353, + "step": 27035 + }, + { + "epoch": 0.5951133612150561, + "grad_norm": 0.6052345747582907, + "learning_rate": 3.7171844577981825e-06, + "loss": 0.4541, + "step": 27036 + }, + { + "epoch": 0.5951353731014748, + "grad_norm": 1.3139925251743063, + "learning_rate": 3.716839936172018e-06, + "loss": 0.3458, + "step": 27037 + }, + { + "epoch": 0.5951573849878935, + "grad_norm": 1.4498745384260254, + "learning_rate": 3.7164954210674485e-06, + "loss": 0.2698, + "step": 27038 + }, + { + "epoch": 0.5951793968743121, + "grad_norm": 2.142921111655187, + "learning_rate": 3.71615091248622e-06, + "loss": 0.3802, + "step": 27039 + }, + { + "epoch": 0.5952014087607308, + "grad_norm": 1.4232397287974827, + "learning_rate": 3.7158064104300863e-06, + "loss": 0.2966, + "step": 27040 + }, + { + "epoch": 0.5952234206471495, + "grad_norm": 2.1954050035443458, + "learning_rate": 3.7154619149007975e-06, + "loss": 0.3371, + "step": 27041 + }, + { + "epoch": 0.5952454325335681, + "grad_norm": 1.3748459885181739, + "learning_rate": 3.7151174259001015e-06, + "loss": 0.3146, + "step": 27042 + }, + { + "epoch": 0.5952674444199868, + "grad_norm": 1.3202941837103883, + "learning_rate": 3.714772943429755e-06, + "loss": 0.3433, + "step": 27043 + }, + { + "epoch": 0.5952894563064055, + "grad_norm": 1.7066589287992235, + "learning_rate": 3.714428467491503e-06, + "loss": 0.2718, + "step": 27044 + }, + { + "epoch": 0.5953114681928241, + "grad_norm": 1.1951434270753782, + "learning_rate": 3.7140839980871007e-06, + "loss": 0.288, + "step": 27045 + }, + { + "epoch": 0.5953334800792428, + "grad_norm": 2.108488033756907, + "learning_rate": 3.7137395352182982e-06, + "loss": 0.32, + "step": 27046 + }, + { + "epoch": 0.5953554919656615, + "grad_norm": 1.5335236533895, + "learning_rate": 3.7133950788868423e-06, + "loss": 0.2802, + "step": 27047 + }, + { + "epoch": 0.5953775038520801, + "grad_norm": 1.56195526990621, + "learning_rate": 3.713050629094488e-06, + "loss": 0.3256, + "step": 27048 + }, + { + "epoch": 0.5953995157384988, + "grad_norm": 1.4857285458891407, + "learning_rate": 3.7127061858429836e-06, + "loss": 0.2628, + "step": 27049 + }, + { + "epoch": 0.5954215276249174, + "grad_norm": 1.2379828891430988, + "learning_rate": 3.71236174913408e-06, + "loss": 0.2974, + "step": 27050 + }, + { + "epoch": 0.5954435395113361, + "grad_norm": 1.707056234655498, + "learning_rate": 3.7120173189695306e-06, + "loss": 0.3977, + "step": 27051 + }, + { + "epoch": 0.5954655513977548, + "grad_norm": 2.224516838256702, + "learning_rate": 3.7116728953510816e-06, + "loss": 0.2474, + "step": 27052 + }, + { + "epoch": 0.5954875632841734, + "grad_norm": 2.370848655463083, + "learning_rate": 3.7113284782804866e-06, + "loss": 0.3022, + "step": 27053 + }, + { + "epoch": 0.5955095751705921, + "grad_norm": 0.939386056175795, + "learning_rate": 3.7109840677594955e-06, + "loss": 0.2449, + "step": 27054 + }, + { + "epoch": 0.5955315870570108, + "grad_norm": 1.5596698620471894, + "learning_rate": 3.7106396637898558e-06, + "loss": 0.3506, + "step": 27055 + }, + { + "epoch": 0.5955535989434294, + "grad_norm": 1.4516539972952094, + "learning_rate": 3.710295266373323e-06, + "loss": 0.3221, + "step": 27056 + }, + { + "epoch": 0.5955756108298481, + "grad_norm": 0.6082517210980187, + "learning_rate": 3.7099508755116433e-06, + "loss": 0.4733, + "step": 27057 + }, + { + "epoch": 0.5955976227162668, + "grad_norm": 1.4328402357786234, + "learning_rate": 3.7096064912065693e-06, + "loss": 0.2418, + "step": 27058 + }, + { + "epoch": 0.5956196346026854, + "grad_norm": 1.6857051429756882, + "learning_rate": 3.7092621134598515e-06, + "loss": 0.2823, + "step": 27059 + }, + { + "epoch": 0.5956416464891041, + "grad_norm": 1.6357154960072489, + "learning_rate": 3.708917742273238e-06, + "loss": 0.2386, + "step": 27060 + }, + { + "epoch": 0.5956636583755228, + "grad_norm": 0.597440164947549, + "learning_rate": 3.708573377648481e-06, + "loss": 0.4518, + "step": 27061 + }, + { + "epoch": 0.5956856702619414, + "grad_norm": 1.4143226040832313, + "learning_rate": 3.7082290195873282e-06, + "loss": 0.2748, + "step": 27062 + }, + { + "epoch": 0.5957076821483601, + "grad_norm": 1.4066313016916931, + "learning_rate": 3.7078846680915337e-06, + "loss": 0.2905, + "step": 27063 + }, + { + "epoch": 0.5957296940347788, + "grad_norm": 1.4430002915208218, + "learning_rate": 3.707540323162847e-06, + "loss": 0.3123, + "step": 27064 + }, + { + "epoch": 0.5957517059211974, + "grad_norm": 5.243906010984723, + "learning_rate": 3.7071959848030136e-06, + "loss": 0.2683, + "step": 27065 + }, + { + "epoch": 0.5957737178076161, + "grad_norm": 3.7002869736512487, + "learning_rate": 3.7068516530137887e-06, + "loss": 0.2807, + "step": 27066 + }, + { + "epoch": 0.5957957296940348, + "grad_norm": 1.6785161809915292, + "learning_rate": 3.7065073277969204e-06, + "loss": 0.2251, + "step": 27067 + }, + { + "epoch": 0.5958177415804534, + "grad_norm": 1.4600590550782686, + "learning_rate": 3.706163009154158e-06, + "loss": 0.3945, + "step": 27068 + }, + { + "epoch": 0.5958397534668721, + "grad_norm": 0.6029849184715134, + "learning_rate": 3.7058186970872534e-06, + "loss": 0.4505, + "step": 27069 + }, + { + "epoch": 0.5958617653532908, + "grad_norm": 2.28334673396555, + "learning_rate": 3.705474391597954e-06, + "loss": 0.3263, + "step": 27070 + }, + { + "epoch": 0.5958837772397094, + "grad_norm": 1.3547629353326716, + "learning_rate": 3.7051300926880116e-06, + "loss": 0.3192, + "step": 27071 + }, + { + "epoch": 0.5959057891261281, + "grad_norm": 0.602094999157009, + "learning_rate": 3.7047858003591774e-06, + "loss": 0.4568, + "step": 27072 + }, + { + "epoch": 0.5959278010125467, + "grad_norm": 3.0266988544696125, + "learning_rate": 3.7044415146131956e-06, + "loss": 0.3564, + "step": 27073 + }, + { + "epoch": 0.5959498128989654, + "grad_norm": 1.5632220738525024, + "learning_rate": 3.7040972354518235e-06, + "loss": 0.3025, + "step": 27074 + }, + { + "epoch": 0.5959718247853841, + "grad_norm": 1.4755785812264133, + "learning_rate": 3.7037529628768047e-06, + "loss": 0.2866, + "step": 27075 + }, + { + "epoch": 0.5959938366718027, + "grad_norm": 1.5914977229828133, + "learning_rate": 3.703408696889893e-06, + "loss": 0.3815, + "step": 27076 + }, + { + "epoch": 0.5960158485582214, + "grad_norm": 1.900016929771785, + "learning_rate": 3.703064437492837e-06, + "loss": 0.3258, + "step": 27077 + }, + { + "epoch": 0.5960378604446401, + "grad_norm": 5.4402635187582575, + "learning_rate": 3.702720184687385e-06, + "loss": 0.3384, + "step": 27078 + }, + { + "epoch": 0.5960598723310587, + "grad_norm": 1.3803204547551071, + "learning_rate": 3.7023759384752876e-06, + "loss": 0.2644, + "step": 27079 + }, + { + "epoch": 0.5960818842174774, + "grad_norm": 1.4316910272252978, + "learning_rate": 3.7020316988582937e-06, + "loss": 0.374, + "step": 27080 + }, + { + "epoch": 0.5961038961038961, + "grad_norm": 0.6256837610146039, + "learning_rate": 3.701687465838155e-06, + "loss": 0.4635, + "step": 27081 + }, + { + "epoch": 0.5961259079903147, + "grad_norm": 2.855192483006584, + "learning_rate": 3.7013432394166206e-06, + "loss": 0.4143, + "step": 27082 + }, + { + "epoch": 0.5961479198767334, + "grad_norm": 1.3666390200870409, + "learning_rate": 3.7009990195954372e-06, + "loss": 0.3492, + "step": 27083 + }, + { + "epoch": 0.5961699317631521, + "grad_norm": 0.6122118568072998, + "learning_rate": 3.7006548063763564e-06, + "loss": 0.4569, + "step": 27084 + }, + { + "epoch": 0.5961919436495707, + "grad_norm": 0.6208542380513931, + "learning_rate": 3.7003105997611264e-06, + "loss": 0.4659, + "step": 27085 + }, + { + "epoch": 0.5962139555359894, + "grad_norm": 1.989239009560249, + "learning_rate": 3.699966399751499e-06, + "loss": 0.3805, + "step": 27086 + }, + { + "epoch": 0.5962359674224081, + "grad_norm": 1.2141722614617472, + "learning_rate": 3.6996222063492236e-06, + "loss": 0.2959, + "step": 27087 + }, + { + "epoch": 0.5962579793088267, + "grad_norm": 1.2207751664330533, + "learning_rate": 3.6992780195560457e-06, + "loss": 0.3193, + "step": 27088 + }, + { + "epoch": 0.5962799911952454, + "grad_norm": 1.421321736477025, + "learning_rate": 3.6989338393737185e-06, + "loss": 0.3803, + "step": 27089 + }, + { + "epoch": 0.5963020030816641, + "grad_norm": 1.4153084701123042, + "learning_rate": 3.69858966580399e-06, + "loss": 0.2682, + "step": 27090 + }, + { + "epoch": 0.5963240149680827, + "grad_norm": 1.4893360261620272, + "learning_rate": 3.6982454988486064e-06, + "loss": 0.2969, + "step": 27091 + }, + { + "epoch": 0.5963460268545014, + "grad_norm": 2.2737559479105176, + "learning_rate": 3.6979013385093232e-06, + "loss": 0.2269, + "step": 27092 + }, + { + "epoch": 0.5963680387409201, + "grad_norm": 1.7220891450727536, + "learning_rate": 3.6975571847878843e-06, + "loss": 0.362, + "step": 27093 + }, + { + "epoch": 0.5963900506273387, + "grad_norm": 1.4551181922150123, + "learning_rate": 3.6972130376860417e-06, + "loss": 0.2392, + "step": 27094 + }, + { + "epoch": 0.5964120625137574, + "grad_norm": 1.4435127361055529, + "learning_rate": 3.696868897205544e-06, + "loss": 0.2449, + "step": 27095 + }, + { + "epoch": 0.596434074400176, + "grad_norm": 0.6181373665429618, + "learning_rate": 3.6965247633481376e-06, + "loss": 0.4683, + "step": 27096 + }, + { + "epoch": 0.5964560862865947, + "grad_norm": 0.6040767239264864, + "learning_rate": 3.696180636115575e-06, + "loss": 0.4667, + "step": 27097 + }, + { + "epoch": 0.5964780981730134, + "grad_norm": 1.5236608197721637, + "learning_rate": 3.6958365155096032e-06, + "loss": 0.3931, + "step": 27098 + }, + { + "epoch": 0.596500110059432, + "grad_norm": 1.3814908140317386, + "learning_rate": 3.6954924015319726e-06, + "loss": 0.2883, + "step": 27099 + }, + { + "epoch": 0.5965221219458507, + "grad_norm": 0.6100254146474839, + "learning_rate": 3.695148294184433e-06, + "loss": 0.4551, + "step": 27100 + }, + { + "epoch": 0.5965441338322695, + "grad_norm": 1.7722576180763454, + "learning_rate": 3.6948041934687294e-06, + "loss": 0.3441, + "step": 27101 + }, + { + "epoch": 0.596566145718688, + "grad_norm": 1.3477769497749468, + "learning_rate": 3.6944600993866135e-06, + "loss": 0.3163, + "step": 27102 + }, + { + "epoch": 0.5965881576051068, + "grad_norm": 1.4742553070843918, + "learning_rate": 3.6941160119398335e-06, + "loss": 0.3299, + "step": 27103 + }, + { + "epoch": 0.5966101694915255, + "grad_norm": 1.4723261331677404, + "learning_rate": 3.693771931130139e-06, + "loss": 0.298, + "step": 27104 + }, + { + "epoch": 0.596632181377944, + "grad_norm": 1.463062853049852, + "learning_rate": 3.6934278569592787e-06, + "loss": 0.3072, + "step": 27105 + }, + { + "epoch": 0.5966541932643628, + "grad_norm": 0.6216470484021722, + "learning_rate": 3.6930837894289995e-06, + "loss": 0.4428, + "step": 27106 + }, + { + "epoch": 0.5966762051507815, + "grad_norm": 1.4283813453318828, + "learning_rate": 3.6927397285410515e-06, + "loss": 0.308, + "step": 27107 + }, + { + "epoch": 0.5966982170372, + "grad_norm": 1.153367658313007, + "learning_rate": 3.6923956742971855e-06, + "loss": 0.2813, + "step": 27108 + }, + { + "epoch": 0.5967202289236188, + "grad_norm": 6.340980454340134, + "learning_rate": 3.6920516266991434e-06, + "loss": 0.3594, + "step": 27109 + }, + { + "epoch": 0.5967422408100375, + "grad_norm": 1.487582511636565, + "learning_rate": 3.6917075857486827e-06, + "loss": 0.2682, + "step": 27110 + }, + { + "epoch": 0.596764252696456, + "grad_norm": 0.6110626073447564, + "learning_rate": 3.691363551447545e-06, + "loss": 0.46, + "step": 27111 + }, + { + "epoch": 0.5967862645828748, + "grad_norm": 0.6148917614862297, + "learning_rate": 3.691019523797482e-06, + "loss": 0.4639, + "step": 27112 + }, + { + "epoch": 0.5968082764692935, + "grad_norm": 1.10580508975196, + "learning_rate": 3.6906755028002432e-06, + "loss": 0.2475, + "step": 27113 + }, + { + "epoch": 0.5968302883557121, + "grad_norm": 1.3480597514290769, + "learning_rate": 3.6903314884575735e-06, + "loss": 0.3221, + "step": 27114 + }, + { + "epoch": 0.5968523002421308, + "grad_norm": 2.9931984953314785, + "learning_rate": 3.6899874807712236e-06, + "loss": 0.2738, + "step": 27115 + }, + { + "epoch": 0.5968743121285495, + "grad_norm": 0.6162996617723688, + "learning_rate": 3.6896434797429416e-06, + "loss": 0.464, + "step": 27116 + }, + { + "epoch": 0.5968963240149681, + "grad_norm": 1.5734627599618425, + "learning_rate": 3.6892994853744755e-06, + "loss": 0.2841, + "step": 27117 + }, + { + "epoch": 0.5969183359013868, + "grad_norm": 0.6111601881464744, + "learning_rate": 3.6889554976675767e-06, + "loss": 0.4551, + "step": 27118 + }, + { + "epoch": 0.5969403477878055, + "grad_norm": 1.1932809644447646, + "learning_rate": 3.6886115166239876e-06, + "loss": 0.2874, + "step": 27119 + }, + { + "epoch": 0.5969623596742241, + "grad_norm": 1.5512424270447698, + "learning_rate": 3.688267542245461e-06, + "loss": 0.301, + "step": 27120 + }, + { + "epoch": 0.5969843715606428, + "grad_norm": 0.6235551645204654, + "learning_rate": 3.6879235745337426e-06, + "loss": 0.4663, + "step": 27121 + }, + { + "epoch": 0.5970063834470614, + "grad_norm": 1.2277573134097237, + "learning_rate": 3.687579613490583e-06, + "loss": 0.2958, + "step": 27122 + }, + { + "epoch": 0.5970283953334801, + "grad_norm": 1.6161131057155118, + "learning_rate": 3.6872356591177307e-06, + "loss": 0.3234, + "step": 27123 + }, + { + "epoch": 0.5970504072198988, + "grad_norm": 0.6282130278746021, + "learning_rate": 3.6868917114169294e-06, + "loss": 0.461, + "step": 27124 + }, + { + "epoch": 0.5970724191063174, + "grad_norm": 2.0061357942000577, + "learning_rate": 3.6865477703899312e-06, + "loss": 0.2948, + "step": 27125 + }, + { + "epoch": 0.5970944309927361, + "grad_norm": 2.658508157658505, + "learning_rate": 3.6862038360384844e-06, + "loss": 0.321, + "step": 27126 + }, + { + "epoch": 0.5971164428791548, + "grad_norm": 1.2728502372853507, + "learning_rate": 3.6858599083643326e-06, + "loss": 0.2575, + "step": 27127 + }, + { + "epoch": 0.5971384547655734, + "grad_norm": 1.4747660279383545, + "learning_rate": 3.6855159873692303e-06, + "loss": 0.2973, + "step": 27128 + }, + { + "epoch": 0.5971604666519921, + "grad_norm": 3.935354445246351, + "learning_rate": 3.6851720730549188e-06, + "loss": 0.3038, + "step": 27129 + }, + { + "epoch": 0.5971824785384108, + "grad_norm": 1.2673540290988488, + "learning_rate": 3.684828165423151e-06, + "loss": 0.2882, + "step": 27130 + }, + { + "epoch": 0.5972044904248294, + "grad_norm": 1.1972048362470562, + "learning_rate": 3.6844842644756744e-06, + "loss": 0.2425, + "step": 27131 + }, + { + "epoch": 0.5972265023112481, + "grad_norm": 0.6196126448381692, + "learning_rate": 3.684140370214233e-06, + "loss": 0.46, + "step": 27132 + }, + { + "epoch": 0.5972485141976668, + "grad_norm": 2.020996181007916, + "learning_rate": 3.683796482640578e-06, + "loss": 0.2283, + "step": 27133 + }, + { + "epoch": 0.5972705260840854, + "grad_norm": 1.401326277034996, + "learning_rate": 3.6834526017564558e-06, + "loss": 0.2841, + "step": 27134 + }, + { + "epoch": 0.5972925379705041, + "grad_norm": 4.574384782712003, + "learning_rate": 3.6831087275636157e-06, + "loss": 0.2559, + "step": 27135 + }, + { + "epoch": 0.5973145498569228, + "grad_norm": 0.6226703249939503, + "learning_rate": 3.6827648600638054e-06, + "loss": 0.462, + "step": 27136 + }, + { + "epoch": 0.5973365617433414, + "grad_norm": 1.8917696035042728, + "learning_rate": 3.682420999258769e-06, + "loss": 0.308, + "step": 27137 + }, + { + "epoch": 0.5973585736297601, + "grad_norm": 2.240456557773861, + "learning_rate": 3.682077145150258e-06, + "loss": 0.2357, + "step": 27138 + }, + { + "epoch": 0.5973805855161788, + "grad_norm": 1.5829851786006763, + "learning_rate": 3.6817332977400178e-06, + "loss": 0.319, + "step": 27139 + }, + { + "epoch": 0.5974025974025974, + "grad_norm": 1.5865425363132668, + "learning_rate": 3.681389457029798e-06, + "loss": 0.3686, + "step": 27140 + }, + { + "epoch": 0.5974246092890161, + "grad_norm": 1.481614225810903, + "learning_rate": 3.6810456230213465e-06, + "loss": 0.3518, + "step": 27141 + }, + { + "epoch": 0.5974466211754348, + "grad_norm": 1.197515336104071, + "learning_rate": 3.680701795716407e-06, + "loss": 0.2789, + "step": 27142 + }, + { + "epoch": 0.5974686330618534, + "grad_norm": 0.6202616901095271, + "learning_rate": 3.6803579751167305e-06, + "loss": 0.4763, + "step": 27143 + }, + { + "epoch": 0.5974906449482721, + "grad_norm": 0.6306370587735431, + "learning_rate": 3.680014161224062e-06, + "loss": 0.4736, + "step": 27144 + }, + { + "epoch": 0.5975126568346907, + "grad_norm": 1.5608117237008174, + "learning_rate": 3.6796703540401514e-06, + "loss": 0.3016, + "step": 27145 + }, + { + "epoch": 0.5975346687211094, + "grad_norm": 1.3227555693225932, + "learning_rate": 3.6793265535667465e-06, + "loss": 0.2336, + "step": 27146 + }, + { + "epoch": 0.5975566806075281, + "grad_norm": 1.5052582381463608, + "learning_rate": 3.6789827598055907e-06, + "loss": 0.2844, + "step": 27147 + }, + { + "epoch": 0.5975786924939467, + "grad_norm": 0.6239508122893201, + "learning_rate": 3.6786389727584343e-06, + "loss": 0.4804, + "step": 27148 + }, + { + "epoch": 0.5976007043803654, + "grad_norm": 1.6720072538274011, + "learning_rate": 3.6782951924270257e-06, + "loss": 0.3147, + "step": 27149 + }, + { + "epoch": 0.5976227162667841, + "grad_norm": 0.6023808963640972, + "learning_rate": 3.6779514188131075e-06, + "loss": 0.4808, + "step": 27150 + }, + { + "epoch": 0.5976447281532027, + "grad_norm": 1.3814188621685577, + "learning_rate": 3.6776076519184315e-06, + "loss": 0.2593, + "step": 27151 + }, + { + "epoch": 0.5976667400396214, + "grad_norm": 1.9282592054268817, + "learning_rate": 3.677263891744742e-06, + "loss": 0.2566, + "step": 27152 + }, + { + "epoch": 0.5976887519260401, + "grad_norm": 1.2954213244209778, + "learning_rate": 3.676920138293789e-06, + "loss": 0.2687, + "step": 27153 + }, + { + "epoch": 0.5977107638124587, + "grad_norm": 1.7692648717899022, + "learning_rate": 3.676576391567318e-06, + "loss": 0.333, + "step": 27154 + }, + { + "epoch": 0.5977327756988774, + "grad_norm": 1.2450981363625557, + "learning_rate": 3.676232651567074e-06, + "loss": 0.2878, + "step": 27155 + }, + { + "epoch": 0.5977547875852961, + "grad_norm": 1.5890390500127418, + "learning_rate": 3.675888918294808e-06, + "loss": 0.2926, + "step": 27156 + }, + { + "epoch": 0.5977767994717147, + "grad_norm": 1.520896885918861, + "learning_rate": 3.675545191752264e-06, + "loss": 0.3394, + "step": 27157 + }, + { + "epoch": 0.5977988113581334, + "grad_norm": 1.5077519148938412, + "learning_rate": 3.675201471941191e-06, + "loss": 0.3136, + "step": 27158 + }, + { + "epoch": 0.5978208232445521, + "grad_norm": 1.2449432497781971, + "learning_rate": 3.6748577588633354e-06, + "loss": 0.2459, + "step": 27159 + }, + { + "epoch": 0.5978428351309707, + "grad_norm": 0.6272634703077861, + "learning_rate": 3.674514052520442e-06, + "loss": 0.4656, + "step": 27160 + }, + { + "epoch": 0.5978648470173894, + "grad_norm": 1.4270571355461508, + "learning_rate": 3.67417035291426e-06, + "loss": 0.2672, + "step": 27161 + }, + { + "epoch": 0.5978868589038081, + "grad_norm": 1.1187646757453416, + "learning_rate": 3.6738266600465356e-06, + "loss": 0.2916, + "step": 27162 + }, + { + "epoch": 0.5979088707902267, + "grad_norm": 1.421584438872885, + "learning_rate": 3.673482973919016e-06, + "loss": 0.2996, + "step": 27163 + }, + { + "epoch": 0.5979308826766454, + "grad_norm": 1.3593783678758127, + "learning_rate": 3.6731392945334487e-06, + "loss": 0.3748, + "step": 27164 + }, + { + "epoch": 0.5979528945630641, + "grad_norm": 1.3585299651249927, + "learning_rate": 3.6727956218915766e-06, + "loss": 0.2801, + "step": 27165 + }, + { + "epoch": 0.5979749064494827, + "grad_norm": 1.3933819728329684, + "learning_rate": 3.672451955995151e-06, + "loss": 0.2302, + "step": 27166 + }, + { + "epoch": 0.5979969183359014, + "grad_norm": 1.75799646785836, + "learning_rate": 3.672108296845918e-06, + "loss": 0.2569, + "step": 27167 + }, + { + "epoch": 0.59801893022232, + "grad_norm": 0.6330625614155004, + "learning_rate": 3.6717646444456196e-06, + "loss": 0.4784, + "step": 27168 + }, + { + "epoch": 0.5980409421087387, + "grad_norm": 1.5060408882051317, + "learning_rate": 3.671420998796007e-06, + "loss": 0.2939, + "step": 27169 + }, + { + "epoch": 0.5980629539951574, + "grad_norm": 1.4080954140967978, + "learning_rate": 3.671077359898825e-06, + "loss": 0.2847, + "step": 27170 + }, + { + "epoch": 0.598084965881576, + "grad_norm": 1.5224319630496523, + "learning_rate": 3.670733727755821e-06, + "loss": 0.3099, + "step": 27171 + }, + { + "epoch": 0.5981069777679947, + "grad_norm": 1.2017601491835193, + "learning_rate": 3.6703901023687426e-06, + "loss": 0.2917, + "step": 27172 + }, + { + "epoch": 0.5981289896544134, + "grad_norm": 1.5946420474989704, + "learning_rate": 3.6700464837393325e-06, + "loss": 0.3165, + "step": 27173 + }, + { + "epoch": 0.598151001540832, + "grad_norm": 1.2875123681419605, + "learning_rate": 3.6697028718693405e-06, + "loss": 0.2437, + "step": 27174 + }, + { + "epoch": 0.5981730134272507, + "grad_norm": 2.216371228975343, + "learning_rate": 3.6693592667605105e-06, + "loss": 0.2905, + "step": 27175 + }, + { + "epoch": 0.5981950253136694, + "grad_norm": 1.3124338198381762, + "learning_rate": 3.6690156684145917e-06, + "loss": 0.2658, + "step": 27176 + }, + { + "epoch": 0.598217037200088, + "grad_norm": 1.9109262236664941, + "learning_rate": 3.6686720768333293e-06, + "loss": 0.3341, + "step": 27177 + }, + { + "epoch": 0.5982390490865067, + "grad_norm": 1.454872306755045, + "learning_rate": 3.6683284920184676e-06, + "loss": 0.3405, + "step": 27178 + }, + { + "epoch": 0.5982610609729254, + "grad_norm": 2.0150879947471987, + "learning_rate": 3.667984913971755e-06, + "loss": 0.3845, + "step": 27179 + }, + { + "epoch": 0.598283072859344, + "grad_norm": 1.582518783160152, + "learning_rate": 3.6676413426949364e-06, + "loss": 0.3302, + "step": 27180 + }, + { + "epoch": 0.5983050847457627, + "grad_norm": 1.5023813118045206, + "learning_rate": 3.66729777818976e-06, + "loss": 0.3964, + "step": 27181 + }, + { + "epoch": 0.5983270966321814, + "grad_norm": 1.0438745571608847, + "learning_rate": 3.666954220457972e-06, + "loss": 0.2666, + "step": 27182 + }, + { + "epoch": 0.5983491085186, + "grad_norm": 1.4517241918751789, + "learning_rate": 3.666610669501314e-06, + "loss": 0.2461, + "step": 27183 + }, + { + "epoch": 0.5983711204050187, + "grad_norm": 1.2620954357896048, + "learning_rate": 3.6662671253215375e-06, + "loss": 0.258, + "step": 27184 + }, + { + "epoch": 0.5983931322914374, + "grad_norm": 1.3523004670665781, + "learning_rate": 3.6659235879203848e-06, + "loss": 0.2884, + "step": 27185 + }, + { + "epoch": 0.598415144177856, + "grad_norm": 0.636194821879488, + "learning_rate": 3.6655800572996054e-06, + "loss": 0.4626, + "step": 27186 + }, + { + "epoch": 0.5984371560642747, + "grad_norm": 1.5117976644701758, + "learning_rate": 3.6652365334609426e-06, + "loss": 0.3132, + "step": 27187 + }, + { + "epoch": 0.5984591679506934, + "grad_norm": 1.6748619668306777, + "learning_rate": 3.664893016406141e-06, + "loss": 0.3241, + "step": 27188 + }, + { + "epoch": 0.598481179837112, + "grad_norm": 3.603875044011017, + "learning_rate": 3.664549506136951e-06, + "loss": 0.3217, + "step": 27189 + }, + { + "epoch": 0.5985031917235307, + "grad_norm": 1.460985325438793, + "learning_rate": 3.6642060026551164e-06, + "loss": 0.3231, + "step": 27190 + }, + { + "epoch": 0.5985252036099493, + "grad_norm": 1.5082957293952375, + "learning_rate": 3.663862505962381e-06, + "loss": 0.3497, + "step": 27191 + }, + { + "epoch": 0.598547215496368, + "grad_norm": 1.3649983155106813, + "learning_rate": 3.6635190160604927e-06, + "loss": 0.3279, + "step": 27192 + }, + { + "epoch": 0.5985692273827867, + "grad_norm": 2.0379800143866444, + "learning_rate": 3.6631755329511965e-06, + "loss": 0.3064, + "step": 27193 + }, + { + "epoch": 0.5985912392692053, + "grad_norm": 7.558877242899866, + "learning_rate": 3.662832056636239e-06, + "loss": 0.337, + "step": 27194 + }, + { + "epoch": 0.598613251155624, + "grad_norm": 1.7509002801730797, + "learning_rate": 3.6624885871173665e-06, + "loss": 0.394, + "step": 27195 + }, + { + "epoch": 0.5986352630420427, + "grad_norm": 1.8015207963690292, + "learning_rate": 3.6621451243963215e-06, + "loss": 0.3463, + "step": 27196 + }, + { + "epoch": 0.5986572749284613, + "grad_norm": 1.2421290550478579, + "learning_rate": 3.661801668474853e-06, + "loss": 0.2026, + "step": 27197 + }, + { + "epoch": 0.59867928681488, + "grad_norm": 1.7576284268998237, + "learning_rate": 3.6614582193547037e-06, + "loss": 0.2847, + "step": 27198 + }, + { + "epoch": 0.5987012987012987, + "grad_norm": 1.9153713071424687, + "learning_rate": 3.661114777037622e-06, + "loss": 0.3453, + "step": 27199 + }, + { + "epoch": 0.5987233105877173, + "grad_norm": 2.25402378484813, + "learning_rate": 3.660771341525354e-06, + "loss": 0.3608, + "step": 27200 + }, + { + "epoch": 0.598745322474136, + "grad_norm": 1.2654269191073624, + "learning_rate": 3.66042791281964e-06, + "loss": 0.2412, + "step": 27201 + }, + { + "epoch": 0.5987673343605547, + "grad_norm": 0.5924610966735675, + "learning_rate": 3.6600844909222306e-06, + "loss": 0.464, + "step": 27202 + }, + { + "epoch": 0.5987893462469733, + "grad_norm": 1.6883067358763144, + "learning_rate": 3.659741075834868e-06, + "loss": 0.3362, + "step": 27203 + }, + { + "epoch": 0.598811358133392, + "grad_norm": 1.4153217238767857, + "learning_rate": 3.659397667559301e-06, + "loss": 0.3376, + "step": 27204 + }, + { + "epoch": 0.5988333700198107, + "grad_norm": 0.5947022415512514, + "learning_rate": 3.6590542660972726e-06, + "loss": 0.4682, + "step": 27205 + }, + { + "epoch": 0.5988553819062293, + "grad_norm": 1.3468748014417375, + "learning_rate": 3.658710871450527e-06, + "loss": 0.2929, + "step": 27206 + }, + { + "epoch": 0.598877393792648, + "grad_norm": 1.3892317637353224, + "learning_rate": 3.6583674836208128e-06, + "loss": 0.3547, + "step": 27207 + }, + { + "epoch": 0.5988994056790667, + "grad_norm": 1.3397705792430106, + "learning_rate": 3.6580241026098734e-06, + "loss": 0.3218, + "step": 27208 + }, + { + "epoch": 0.5989214175654853, + "grad_norm": 1.7137812804462624, + "learning_rate": 3.6576807284194525e-06, + "loss": 0.2874, + "step": 27209 + }, + { + "epoch": 0.598943429451904, + "grad_norm": 2.0953550367885176, + "learning_rate": 3.6573373610512976e-06, + "loss": 0.3578, + "step": 27210 + }, + { + "epoch": 0.5989654413383227, + "grad_norm": 1.9387754762534701, + "learning_rate": 3.6569940005071526e-06, + "loss": 0.3197, + "step": 27211 + }, + { + "epoch": 0.5989874532247413, + "grad_norm": 0.6116742348351817, + "learning_rate": 3.656650646788763e-06, + "loss": 0.4621, + "step": 27212 + }, + { + "epoch": 0.59900946511116, + "grad_norm": 1.4391975784458968, + "learning_rate": 3.656307299897876e-06, + "loss": 0.2945, + "step": 27213 + }, + { + "epoch": 0.5990314769975786, + "grad_norm": 2.229207183843334, + "learning_rate": 3.6559639598362316e-06, + "loss": 0.3323, + "step": 27214 + }, + { + "epoch": 0.5990534888839973, + "grad_norm": 1.4013834780141607, + "learning_rate": 3.65562062660558e-06, + "loss": 0.3085, + "step": 27215 + }, + { + "epoch": 0.599075500770416, + "grad_norm": 1.3682063845287422, + "learning_rate": 3.6552773002076626e-06, + "loss": 0.2943, + "step": 27216 + }, + { + "epoch": 0.5990975126568346, + "grad_norm": 1.7877247703266306, + "learning_rate": 3.6549339806442265e-06, + "loss": 0.3095, + "step": 27217 + }, + { + "epoch": 0.5991195245432533, + "grad_norm": 1.116694991056974, + "learning_rate": 3.6545906679170168e-06, + "loss": 0.2897, + "step": 27218 + }, + { + "epoch": 0.599141536429672, + "grad_norm": 1.3311540820557197, + "learning_rate": 3.6542473620277752e-06, + "loss": 0.3084, + "step": 27219 + }, + { + "epoch": 0.5991635483160906, + "grad_norm": 2.197031062570595, + "learning_rate": 3.6539040629782506e-06, + "loss": 0.3386, + "step": 27220 + }, + { + "epoch": 0.5991855602025093, + "grad_norm": 1.4853576103217685, + "learning_rate": 3.653560770770184e-06, + "loss": 0.3455, + "step": 27221 + }, + { + "epoch": 0.599207572088928, + "grad_norm": 1.3366131503986538, + "learning_rate": 3.653217485405325e-06, + "loss": 0.2977, + "step": 27222 + }, + { + "epoch": 0.5992295839753466, + "grad_norm": 3.254725784236732, + "learning_rate": 3.6528742068854133e-06, + "loss": 0.2951, + "step": 27223 + }, + { + "epoch": 0.5992515958617654, + "grad_norm": 1.5425709904401204, + "learning_rate": 3.652530935212195e-06, + "loss": 0.2536, + "step": 27224 + }, + { + "epoch": 0.5992736077481841, + "grad_norm": 1.5839822132300858, + "learning_rate": 3.652187670387417e-06, + "loss": 0.2592, + "step": 27225 + }, + { + "epoch": 0.5992956196346026, + "grad_norm": 1.2352654535046932, + "learning_rate": 3.651844412412824e-06, + "loss": 0.3086, + "step": 27226 + }, + { + "epoch": 0.5993176315210214, + "grad_norm": 1.598674252924881, + "learning_rate": 3.6515011612901553e-06, + "loss": 0.2996, + "step": 27227 + }, + { + "epoch": 0.5993396434074401, + "grad_norm": 1.6405407270008672, + "learning_rate": 3.6511579170211614e-06, + "loss": 0.2747, + "step": 27228 + }, + { + "epoch": 0.5993616552938587, + "grad_norm": 1.3435248207657216, + "learning_rate": 3.650814679607583e-06, + "loss": 0.2155, + "step": 27229 + }, + { + "epoch": 0.5993836671802774, + "grad_norm": 11.999025389150686, + "learning_rate": 3.6504714490511673e-06, + "loss": 0.3282, + "step": 27230 + }, + { + "epoch": 0.5994056790666961, + "grad_norm": 1.8385038513190017, + "learning_rate": 3.6501282253536585e-06, + "loss": 0.35, + "step": 27231 + }, + { + "epoch": 0.5994276909531147, + "grad_norm": 1.6255232503468158, + "learning_rate": 3.6497850085167985e-06, + "loss": 0.2705, + "step": 27232 + }, + { + "epoch": 0.5994497028395334, + "grad_norm": 1.5292441941428379, + "learning_rate": 3.6494417985423335e-06, + "loss": 0.2436, + "step": 27233 + }, + { + "epoch": 0.5994717147259521, + "grad_norm": 1.3666968309895056, + "learning_rate": 3.649098595432008e-06, + "loss": 0.3629, + "step": 27234 + }, + { + "epoch": 0.5994937266123707, + "grad_norm": 1.1672689989134701, + "learning_rate": 3.648755399187566e-06, + "loss": 0.2229, + "step": 27235 + }, + { + "epoch": 0.5995157384987894, + "grad_norm": 2.112047644492077, + "learning_rate": 3.6484122098107526e-06, + "loss": 0.325, + "step": 27236 + }, + { + "epoch": 0.599537750385208, + "grad_norm": 1.308728488341473, + "learning_rate": 3.648069027303309e-06, + "loss": 0.2505, + "step": 27237 + }, + { + "epoch": 0.5995597622716267, + "grad_norm": 1.626160724772805, + "learning_rate": 3.6477258516669835e-06, + "loss": 0.296, + "step": 27238 + }, + { + "epoch": 0.5995817741580454, + "grad_norm": 1.18923164559503, + "learning_rate": 3.647382682903517e-06, + "loss": 0.2949, + "step": 27239 + }, + { + "epoch": 0.599603786044464, + "grad_norm": 0.6526011647506311, + "learning_rate": 3.647039521014657e-06, + "loss": 0.4811, + "step": 27240 + }, + { + "epoch": 0.5996257979308827, + "grad_norm": 0.6365343576737006, + "learning_rate": 3.646696366002144e-06, + "loss": 0.4547, + "step": 27241 + }, + { + "epoch": 0.5996478098173014, + "grad_norm": 1.529948728702643, + "learning_rate": 3.646353217867723e-06, + "loss": 0.3226, + "step": 27242 + }, + { + "epoch": 0.59966982170372, + "grad_norm": 1.3073309522053775, + "learning_rate": 3.64601007661314e-06, + "loss": 0.318, + "step": 27243 + }, + { + "epoch": 0.5996918335901387, + "grad_norm": 3.800444825182288, + "learning_rate": 3.645666942240137e-06, + "loss": 0.3423, + "step": 27244 + }, + { + "epoch": 0.5997138454765574, + "grad_norm": 1.0868963007863703, + "learning_rate": 3.64532381475046e-06, + "loss": 0.2396, + "step": 27245 + }, + { + "epoch": 0.599735857362976, + "grad_norm": 7.728693027522423, + "learning_rate": 3.6449806941458505e-06, + "loss": 0.2466, + "step": 27246 + }, + { + "epoch": 0.5997578692493947, + "grad_norm": 1.6280363268172682, + "learning_rate": 3.6446375804280525e-06, + "loss": 0.3333, + "step": 27247 + }, + { + "epoch": 0.5997798811358134, + "grad_norm": 1.6904184023998738, + "learning_rate": 3.6442944735988122e-06, + "loss": 0.3225, + "step": 27248 + }, + { + "epoch": 0.599801893022232, + "grad_norm": 1.2457948444516256, + "learning_rate": 3.6439513736598733e-06, + "loss": 0.2496, + "step": 27249 + }, + { + "epoch": 0.5998239049086507, + "grad_norm": 1.2142763891451045, + "learning_rate": 3.643608280612976e-06, + "loss": 0.1614, + "step": 27250 + }, + { + "epoch": 0.5998459167950694, + "grad_norm": 1.3205669069632917, + "learning_rate": 3.643265194459867e-06, + "loss": 0.3439, + "step": 27251 + }, + { + "epoch": 0.599867928681488, + "grad_norm": 1.3061400656044504, + "learning_rate": 3.6429221152022886e-06, + "loss": 0.2577, + "step": 27252 + }, + { + "epoch": 0.5998899405679067, + "grad_norm": 2.2744398390853124, + "learning_rate": 3.642579042841986e-06, + "loss": 0.3299, + "step": 27253 + }, + { + "epoch": 0.5999119524543254, + "grad_norm": 1.529185629825749, + "learning_rate": 3.6422359773807036e-06, + "loss": 0.2347, + "step": 27254 + }, + { + "epoch": 0.599933964340744, + "grad_norm": 0.6758640847882593, + "learning_rate": 3.6418929188201813e-06, + "loss": 0.4767, + "step": 27255 + }, + { + "epoch": 0.5999559762271627, + "grad_norm": 2.9129723640853986, + "learning_rate": 3.641549867162166e-06, + "loss": 0.2905, + "step": 27256 + }, + { + "epoch": 0.5999779881135814, + "grad_norm": 0.6612958961110669, + "learning_rate": 3.6412068224083994e-06, + "loss": 0.4704, + "step": 27257 + }, + { + "epoch": 0.6, + "grad_norm": 1.5631871433148623, + "learning_rate": 3.6408637845606275e-06, + "loss": 0.2945, + "step": 27258 + }, + { + "epoch": 0.6000220118864187, + "grad_norm": 2.357081189773261, + "learning_rate": 3.640520753620591e-06, + "loss": 0.3049, + "step": 27259 + }, + { + "epoch": 0.6000440237728374, + "grad_norm": 1.5757601482230883, + "learning_rate": 3.640177729590034e-06, + "loss": 0.2733, + "step": 27260 + }, + { + "epoch": 0.600066035659256, + "grad_norm": 1.5452956707541599, + "learning_rate": 3.6398347124707004e-06, + "loss": 0.27, + "step": 27261 + }, + { + "epoch": 0.6000880475456747, + "grad_norm": 1.4124479331888935, + "learning_rate": 3.639491702264333e-06, + "loss": 0.358, + "step": 27262 + }, + { + "epoch": 0.6001100594320933, + "grad_norm": 1.5445054378131209, + "learning_rate": 3.6391486989726774e-06, + "loss": 0.1996, + "step": 27263 + }, + { + "epoch": 0.600132071318512, + "grad_norm": 2.042383809880389, + "learning_rate": 3.6388057025974745e-06, + "loss": 0.2584, + "step": 27264 + }, + { + "epoch": 0.6001540832049307, + "grad_norm": 1.5074179244762302, + "learning_rate": 3.638462713140466e-06, + "loss": 0.349, + "step": 27265 + }, + { + "epoch": 0.6001760950913493, + "grad_norm": 4.693476315678988, + "learning_rate": 3.638119730603399e-06, + "loss": 0.3294, + "step": 27266 + }, + { + "epoch": 0.600198106977768, + "grad_norm": 1.5823326593442524, + "learning_rate": 3.637776754988016e-06, + "loss": 0.2839, + "step": 27267 + }, + { + "epoch": 0.6002201188641867, + "grad_norm": 1.8066906400312328, + "learning_rate": 3.6374337862960576e-06, + "loss": 0.2511, + "step": 27268 + }, + { + "epoch": 0.6002421307506053, + "grad_norm": 2.3225716995982, + "learning_rate": 3.637090824529268e-06, + "loss": 0.3691, + "step": 27269 + }, + { + "epoch": 0.600264142637024, + "grad_norm": 0.621850304876116, + "learning_rate": 3.636747869689391e-06, + "loss": 0.4551, + "step": 27270 + }, + { + "epoch": 0.6002861545234427, + "grad_norm": 0.6414083617635845, + "learning_rate": 3.6364049217781696e-06, + "loss": 0.4389, + "step": 27271 + }, + { + "epoch": 0.6003081664098613, + "grad_norm": 3.4980819152226643, + "learning_rate": 3.636061980797348e-06, + "loss": 0.3125, + "step": 27272 + }, + { + "epoch": 0.60033017829628, + "grad_norm": 1.2989425125690688, + "learning_rate": 3.6357190467486658e-06, + "loss": 0.3152, + "step": 27273 + }, + { + "epoch": 0.6003521901826987, + "grad_norm": 1.183466472174965, + "learning_rate": 3.635376119633869e-06, + "loss": 0.3104, + "step": 27274 + }, + { + "epoch": 0.6003742020691173, + "grad_norm": 1.2950927818128173, + "learning_rate": 3.635033199454697e-06, + "loss": 0.3241, + "step": 27275 + }, + { + "epoch": 0.600396213955536, + "grad_norm": 3.9226530850506953, + "learning_rate": 3.6346902862128986e-06, + "loss": 0.2883, + "step": 27276 + }, + { + "epoch": 0.6004182258419547, + "grad_norm": 1.38854850410555, + "learning_rate": 3.6343473799102116e-06, + "loss": 0.3516, + "step": 27277 + }, + { + "epoch": 0.6004402377283733, + "grad_norm": 1.4293863143993542, + "learning_rate": 3.6340044805483797e-06, + "loss": 0.3051, + "step": 27278 + }, + { + "epoch": 0.600462249614792, + "grad_norm": 1.5079855824641644, + "learning_rate": 3.633661588129147e-06, + "loss": 0.2602, + "step": 27279 + }, + { + "epoch": 0.6004842615012107, + "grad_norm": 1.3760083187010963, + "learning_rate": 3.6333187026542536e-06, + "loss": 0.2936, + "step": 27280 + }, + { + "epoch": 0.6005062733876293, + "grad_norm": 1.2132654878268954, + "learning_rate": 3.6329758241254474e-06, + "loss": 0.339, + "step": 27281 + }, + { + "epoch": 0.600528285274048, + "grad_norm": 0.6120833679391625, + "learning_rate": 3.6326329525444663e-06, + "loss": 0.4392, + "step": 27282 + }, + { + "epoch": 0.6005502971604667, + "grad_norm": 1.7663553324734516, + "learning_rate": 3.6322900879130533e-06, + "loss": 0.3415, + "step": 27283 + }, + { + "epoch": 0.6005723090468853, + "grad_norm": 1.76838940774375, + "learning_rate": 3.631947230232954e-06, + "loss": 0.267, + "step": 27284 + }, + { + "epoch": 0.600594320933304, + "grad_norm": 3.0688237356969092, + "learning_rate": 3.6316043795059065e-06, + "loss": 0.3553, + "step": 27285 + }, + { + "epoch": 0.6006163328197226, + "grad_norm": 1.4055417822382819, + "learning_rate": 3.631261535733659e-06, + "loss": 0.266, + "step": 27286 + }, + { + "epoch": 0.6006383447061413, + "grad_norm": 10.35332962729144, + "learning_rate": 3.6309186989179496e-06, + "loss": 0.3059, + "step": 27287 + }, + { + "epoch": 0.60066035659256, + "grad_norm": 0.5849521560895179, + "learning_rate": 3.630575869060521e-06, + "loss": 0.4364, + "step": 27288 + }, + { + "epoch": 0.6006823684789786, + "grad_norm": 1.2945262989011306, + "learning_rate": 3.6302330461631184e-06, + "loss": 0.2582, + "step": 27289 + }, + { + "epoch": 0.6007043803653973, + "grad_norm": 1.3826955767027844, + "learning_rate": 3.6298902302274814e-06, + "loss": 0.263, + "step": 27290 + }, + { + "epoch": 0.600726392251816, + "grad_norm": 1.4556958359554208, + "learning_rate": 3.6295474212553526e-06, + "loss": 0.2504, + "step": 27291 + }, + { + "epoch": 0.6007484041382346, + "grad_norm": 1.714496268853817, + "learning_rate": 3.6292046192484753e-06, + "loss": 0.2924, + "step": 27292 + }, + { + "epoch": 0.6007704160246533, + "grad_norm": 1.3526002793917322, + "learning_rate": 3.6288618242085917e-06, + "loss": 0.3375, + "step": 27293 + }, + { + "epoch": 0.600792427911072, + "grad_norm": 0.5926629327269928, + "learning_rate": 3.628519036137445e-06, + "loss": 0.4511, + "step": 27294 + }, + { + "epoch": 0.6008144397974906, + "grad_norm": 1.3962393323086595, + "learning_rate": 3.6281762550367754e-06, + "loss": 0.2423, + "step": 27295 + }, + { + "epoch": 0.6008364516839093, + "grad_norm": 2.2962412676932122, + "learning_rate": 3.627833480908325e-06, + "loss": 0.2611, + "step": 27296 + }, + { + "epoch": 0.600858463570328, + "grad_norm": 1.4517891734514359, + "learning_rate": 3.6274907137538382e-06, + "loss": 0.3064, + "step": 27297 + }, + { + "epoch": 0.6008804754567466, + "grad_norm": 1.3652533465386063, + "learning_rate": 3.6271479535750542e-06, + "loss": 0.3355, + "step": 27298 + }, + { + "epoch": 0.6009024873431653, + "grad_norm": 1.5297228300249783, + "learning_rate": 3.6268052003737185e-06, + "loss": 0.2925, + "step": 27299 + }, + { + "epoch": 0.600924499229584, + "grad_norm": 1.574283470428025, + "learning_rate": 3.6264624541515713e-06, + "loss": 0.4449, + "step": 27300 + }, + { + "epoch": 0.6009465111160026, + "grad_norm": 0.6278976885410088, + "learning_rate": 3.626119714910352e-06, + "loss": 0.4812, + "step": 27301 + }, + { + "epoch": 0.6009685230024213, + "grad_norm": 1.4660967800328755, + "learning_rate": 3.6257769826518073e-06, + "loss": 0.2761, + "step": 27302 + }, + { + "epoch": 0.60099053488884, + "grad_norm": 0.6083622278250302, + "learning_rate": 3.6254342573776756e-06, + "loss": 0.4705, + "step": 27303 + }, + { + "epoch": 0.6010125467752586, + "grad_norm": 1.5737600201891495, + "learning_rate": 3.6250915390897018e-06, + "loss": 0.3691, + "step": 27304 + }, + { + "epoch": 0.6010345586616773, + "grad_norm": 1.246256274012755, + "learning_rate": 3.6247488277896253e-06, + "loss": 0.3174, + "step": 27305 + }, + { + "epoch": 0.601056570548096, + "grad_norm": 1.8188483688013801, + "learning_rate": 3.624406123479187e-06, + "loss": 0.3454, + "step": 27306 + }, + { + "epoch": 0.6010785824345146, + "grad_norm": 0.6375281200527343, + "learning_rate": 3.624063426160133e-06, + "loss": 0.4796, + "step": 27307 + }, + { + "epoch": 0.6011005943209333, + "grad_norm": 3.200465969676052, + "learning_rate": 3.6237207358342013e-06, + "loss": 0.3115, + "step": 27308 + }, + { + "epoch": 0.6011226062073519, + "grad_norm": 1.4215703913935953, + "learning_rate": 3.6233780525031337e-06, + "loss": 0.2438, + "step": 27309 + }, + { + "epoch": 0.6011446180937706, + "grad_norm": 2.591120636007139, + "learning_rate": 3.6230353761686734e-06, + "loss": 0.3336, + "step": 27310 + }, + { + "epoch": 0.6011666299801893, + "grad_norm": 1.5575223411831371, + "learning_rate": 3.6226927068325606e-06, + "loss": 0.4039, + "step": 27311 + }, + { + "epoch": 0.6011886418666079, + "grad_norm": 1.862325679515003, + "learning_rate": 3.62235004449654e-06, + "loss": 0.3157, + "step": 27312 + }, + { + "epoch": 0.6012106537530266, + "grad_norm": 1.293140154994949, + "learning_rate": 3.6220073891623498e-06, + "loss": 0.3233, + "step": 27313 + }, + { + "epoch": 0.6012326656394453, + "grad_norm": 1.3442522678936326, + "learning_rate": 3.6216647408317317e-06, + "loss": 0.3254, + "step": 27314 + }, + { + "epoch": 0.6012546775258639, + "grad_norm": 1.9285895139752374, + "learning_rate": 3.621322099506429e-06, + "loss": 0.2668, + "step": 27315 + }, + { + "epoch": 0.6012766894122826, + "grad_norm": 1.531518092306185, + "learning_rate": 3.620979465188181e-06, + "loss": 0.2792, + "step": 27316 + }, + { + "epoch": 0.6012987012987013, + "grad_norm": 1.6806193558040554, + "learning_rate": 3.620636837878734e-06, + "loss": 0.3578, + "step": 27317 + }, + { + "epoch": 0.6013207131851199, + "grad_norm": 1.9579117111781672, + "learning_rate": 3.620294217579823e-06, + "loss": 0.293, + "step": 27318 + }, + { + "epoch": 0.6013427250715386, + "grad_norm": 1.300381541616661, + "learning_rate": 3.6199516042931914e-06, + "loss": 0.2779, + "step": 27319 + }, + { + "epoch": 0.6013647369579573, + "grad_norm": 1.6106349890895753, + "learning_rate": 3.6196089980205828e-06, + "loss": 0.2649, + "step": 27320 + }, + { + "epoch": 0.6013867488443759, + "grad_norm": 1.421648070769852, + "learning_rate": 3.6192663987637354e-06, + "loss": 0.2715, + "step": 27321 + }, + { + "epoch": 0.6014087607307946, + "grad_norm": 1.5779326141156815, + "learning_rate": 3.6189238065243947e-06, + "loss": 0.2752, + "step": 27322 + }, + { + "epoch": 0.6014307726172133, + "grad_norm": 1.3183932782684604, + "learning_rate": 3.618581221304298e-06, + "loss": 0.2606, + "step": 27323 + }, + { + "epoch": 0.6014527845036319, + "grad_norm": 1.1657255546498362, + "learning_rate": 3.6182386431051863e-06, + "loss": 0.3306, + "step": 27324 + }, + { + "epoch": 0.6014747963900506, + "grad_norm": 1.2667792605188193, + "learning_rate": 3.6178960719288046e-06, + "loss": 0.3281, + "step": 27325 + }, + { + "epoch": 0.6014968082764693, + "grad_norm": 1.4067722355537917, + "learning_rate": 3.6175535077768905e-06, + "loss": 0.3134, + "step": 27326 + }, + { + "epoch": 0.6015188201628879, + "grad_norm": 2.1434066097163416, + "learning_rate": 3.6172109506511845e-06, + "loss": 0.2618, + "step": 27327 + }, + { + "epoch": 0.6015408320493066, + "grad_norm": 1.4071240105705796, + "learning_rate": 3.6168684005534303e-06, + "loss": 0.2618, + "step": 27328 + }, + { + "epoch": 0.6015628439357253, + "grad_norm": 1.3253767458593788, + "learning_rate": 3.616525857485367e-06, + "loss": 0.2601, + "step": 27329 + }, + { + "epoch": 0.6015848558221439, + "grad_norm": 1.7409075012753568, + "learning_rate": 3.6161833214487383e-06, + "loss": 0.2617, + "step": 27330 + }, + { + "epoch": 0.6016068677085626, + "grad_norm": 1.4854825934912723, + "learning_rate": 3.615840792445282e-06, + "loss": 0.2393, + "step": 27331 + }, + { + "epoch": 0.6016288795949812, + "grad_norm": 1.646264444124446, + "learning_rate": 3.6154982704767396e-06, + "loss": 0.3102, + "step": 27332 + }, + { + "epoch": 0.6016508914813999, + "grad_norm": 1.309234504768083, + "learning_rate": 3.615155755544853e-06, + "loss": 0.3242, + "step": 27333 + }, + { + "epoch": 0.6016729033678186, + "grad_norm": 0.6533517534735978, + "learning_rate": 3.614813247651362e-06, + "loss": 0.4797, + "step": 27334 + }, + { + "epoch": 0.6016949152542372, + "grad_norm": 1.6169658091145094, + "learning_rate": 3.6144707467980095e-06, + "loss": 0.2444, + "step": 27335 + }, + { + "epoch": 0.6017169271406559, + "grad_norm": 1.7221601488461542, + "learning_rate": 3.6141282529865336e-06, + "loss": 0.293, + "step": 27336 + }, + { + "epoch": 0.6017389390270746, + "grad_norm": 1.9485145211645545, + "learning_rate": 3.613785766218675e-06, + "loss": 0.2934, + "step": 27337 + }, + { + "epoch": 0.6017609509134932, + "grad_norm": 4.469339961424473, + "learning_rate": 3.6134432864961768e-06, + "loss": 0.3813, + "step": 27338 + }, + { + "epoch": 0.601782962799912, + "grad_norm": 1.3517003258072124, + "learning_rate": 3.6131008138207767e-06, + "loss": 0.2774, + "step": 27339 + }, + { + "epoch": 0.6018049746863307, + "grad_norm": 2.106587191084952, + "learning_rate": 3.612758348194219e-06, + "loss": 0.2849, + "step": 27340 + }, + { + "epoch": 0.6018269865727492, + "grad_norm": 1.331140057213615, + "learning_rate": 3.6124158896182416e-06, + "loss": 0.3198, + "step": 27341 + }, + { + "epoch": 0.601848998459168, + "grad_norm": 1.7042936864701537, + "learning_rate": 3.6120734380945845e-06, + "loss": 0.4146, + "step": 27342 + }, + { + "epoch": 0.6018710103455867, + "grad_norm": 1.3835288704890156, + "learning_rate": 3.61173099362499e-06, + "loss": 0.2888, + "step": 27343 + }, + { + "epoch": 0.6018930222320052, + "grad_norm": 1.5342183852247482, + "learning_rate": 3.611388556211196e-06, + "loss": 0.1972, + "step": 27344 + }, + { + "epoch": 0.601915034118424, + "grad_norm": 21.59303168038557, + "learning_rate": 3.6110461258549474e-06, + "loss": 0.2179, + "step": 27345 + }, + { + "epoch": 0.6019370460048427, + "grad_norm": 1.727702948644789, + "learning_rate": 3.6107037025579805e-06, + "loss": 0.2619, + "step": 27346 + }, + { + "epoch": 0.6019590578912613, + "grad_norm": 0.6104278776192765, + "learning_rate": 3.6103612863220366e-06, + "loss": 0.4577, + "step": 27347 + }, + { + "epoch": 0.60198106977768, + "grad_norm": 2.01714311350488, + "learning_rate": 3.6100188771488588e-06, + "loss": 0.3157, + "step": 27348 + }, + { + "epoch": 0.6020030816640987, + "grad_norm": 1.5894231944540262, + "learning_rate": 3.6096764750401826e-06, + "loss": 0.2881, + "step": 27349 + }, + { + "epoch": 0.6020250935505173, + "grad_norm": 1.6547964134615796, + "learning_rate": 3.6093340799977506e-06, + "loss": 0.3341, + "step": 27350 + }, + { + "epoch": 0.602047105436936, + "grad_norm": 1.1630049162580611, + "learning_rate": 3.6089916920233036e-06, + "loss": 0.2896, + "step": 27351 + }, + { + "epoch": 0.6020691173233547, + "grad_norm": 0.6089069838123784, + "learning_rate": 3.6086493111185807e-06, + "loss": 0.4602, + "step": 27352 + }, + { + "epoch": 0.6020911292097733, + "grad_norm": 1.77652853794316, + "learning_rate": 3.6083069372853246e-06, + "loss": 0.2501, + "step": 27353 + }, + { + "epoch": 0.602113141096192, + "grad_norm": 1.496353794655725, + "learning_rate": 3.607964570525272e-06, + "loss": 0.2663, + "step": 27354 + }, + { + "epoch": 0.6021351529826106, + "grad_norm": 3.3708034107367904, + "learning_rate": 3.607622210840164e-06, + "loss": 0.3311, + "step": 27355 + }, + { + "epoch": 0.6021571648690293, + "grad_norm": 1.3887271542132915, + "learning_rate": 3.6072798582317416e-06, + "loss": 0.2659, + "step": 27356 + }, + { + "epoch": 0.602179176755448, + "grad_norm": 0.5988066588761416, + "learning_rate": 3.6069375127017425e-06, + "loss": 0.4748, + "step": 27357 + }, + { + "epoch": 0.6022011886418666, + "grad_norm": 1.2927624237661994, + "learning_rate": 3.6065951742519105e-06, + "loss": 0.3036, + "step": 27358 + }, + { + "epoch": 0.6022232005282853, + "grad_norm": 0.6083275648870723, + "learning_rate": 3.6062528428839826e-06, + "loss": 0.4502, + "step": 27359 + }, + { + "epoch": 0.602245212414704, + "grad_norm": 1.4124503210407315, + "learning_rate": 3.6059105185996984e-06, + "loss": 0.3141, + "step": 27360 + }, + { + "epoch": 0.6022672243011226, + "grad_norm": 1.858882269277559, + "learning_rate": 3.6055682014008007e-06, + "loss": 0.2194, + "step": 27361 + }, + { + "epoch": 0.6022892361875413, + "grad_norm": 1.5754547487484811, + "learning_rate": 3.6052258912890244e-06, + "loss": 0.3, + "step": 27362 + }, + { + "epoch": 0.60231124807396, + "grad_norm": 0.6088435872808189, + "learning_rate": 3.6048835882661146e-06, + "loss": 0.4823, + "step": 27363 + }, + { + "epoch": 0.6023332599603786, + "grad_norm": 0.5962825728162514, + "learning_rate": 3.6045412923338087e-06, + "loss": 0.4655, + "step": 27364 + }, + { + "epoch": 0.6023552718467973, + "grad_norm": 1.3164093250450888, + "learning_rate": 3.6041990034938444e-06, + "loss": 0.274, + "step": 27365 + }, + { + "epoch": 0.602377283733216, + "grad_norm": 1.597236687028877, + "learning_rate": 3.603856721747966e-06, + "loss": 0.2829, + "step": 27366 + }, + { + "epoch": 0.6023992956196346, + "grad_norm": 0.6073422007695202, + "learning_rate": 3.603514447097909e-06, + "loss": 0.4496, + "step": 27367 + }, + { + "epoch": 0.6024213075060533, + "grad_norm": 1.5365067602129274, + "learning_rate": 3.6031721795454135e-06, + "loss": 0.2952, + "step": 27368 + }, + { + "epoch": 0.602443319392472, + "grad_norm": 1.6880742243236688, + "learning_rate": 3.602829919092221e-06, + "loss": 0.3284, + "step": 27369 + }, + { + "epoch": 0.6024653312788906, + "grad_norm": 1.6376836099528764, + "learning_rate": 3.602487665740069e-06, + "loss": 0.2976, + "step": 27370 + }, + { + "epoch": 0.6024873431653093, + "grad_norm": 0.6542565069400857, + "learning_rate": 3.6021454194907e-06, + "loss": 0.489, + "step": 27371 + }, + { + "epoch": 0.602509355051728, + "grad_norm": 1.5798242460032585, + "learning_rate": 3.601803180345851e-06, + "loss": 0.3703, + "step": 27372 + }, + { + "epoch": 0.6025313669381466, + "grad_norm": 0.6211969906144656, + "learning_rate": 3.60146094830726e-06, + "loss": 0.4649, + "step": 27373 + }, + { + "epoch": 0.6025533788245653, + "grad_norm": 1.5477201218375387, + "learning_rate": 3.6011187233766696e-06, + "loss": 0.2371, + "step": 27374 + }, + { + "epoch": 0.602575390710984, + "grad_norm": 0.6012994883159479, + "learning_rate": 3.6007765055558164e-06, + "loss": 0.4619, + "step": 27375 + }, + { + "epoch": 0.6025974025974026, + "grad_norm": 1.8384682857772303, + "learning_rate": 3.600434294846444e-06, + "loss": 0.2861, + "step": 27376 + }, + { + "epoch": 0.6026194144838213, + "grad_norm": 1.9815172855306487, + "learning_rate": 3.6000920912502866e-06, + "loss": 0.3708, + "step": 27377 + }, + { + "epoch": 0.6026414263702399, + "grad_norm": 1.775440916410291, + "learning_rate": 3.5997498947690856e-06, + "loss": 0.3104, + "step": 27378 + }, + { + "epoch": 0.6026634382566586, + "grad_norm": 1.8279538813097784, + "learning_rate": 3.5994077054045816e-06, + "loss": 0.2683, + "step": 27379 + }, + { + "epoch": 0.6026854501430773, + "grad_norm": 1.318399977874613, + "learning_rate": 3.599065523158509e-06, + "loss": 0.2368, + "step": 27380 + }, + { + "epoch": 0.6027074620294959, + "grad_norm": 1.5722021773621617, + "learning_rate": 3.5987233480326133e-06, + "loss": 0.2734, + "step": 27381 + }, + { + "epoch": 0.6027294739159146, + "grad_norm": 1.439745186059821, + "learning_rate": 3.5983811800286306e-06, + "loss": 0.3605, + "step": 27382 + }, + { + "epoch": 0.6027514858023333, + "grad_norm": 8.74983253800096, + "learning_rate": 3.598039019148298e-06, + "loss": 0.23, + "step": 27383 + }, + { + "epoch": 0.6027734976887519, + "grad_norm": 1.9919502734491168, + "learning_rate": 3.5976968653933585e-06, + "loss": 0.3805, + "step": 27384 + }, + { + "epoch": 0.6027955095751706, + "grad_norm": 1.548539817567662, + "learning_rate": 3.5973547187655453e-06, + "loss": 0.3243, + "step": 27385 + }, + { + "epoch": 0.6028175214615893, + "grad_norm": 0.6433474383042534, + "learning_rate": 3.5970125792666053e-06, + "loss": 0.486, + "step": 27386 + }, + { + "epoch": 0.6028395333480079, + "grad_norm": 1.3724481715484647, + "learning_rate": 3.5966704468982716e-06, + "loss": 0.2776, + "step": 27387 + }, + { + "epoch": 0.6028615452344266, + "grad_norm": 1.643032646791461, + "learning_rate": 3.5963283216622835e-06, + "loss": 0.3348, + "step": 27388 + }, + { + "epoch": 0.6028835571208453, + "grad_norm": 4.2985148009389675, + "learning_rate": 3.595986203560383e-06, + "loss": 0.2936, + "step": 27389 + }, + { + "epoch": 0.6029055690072639, + "grad_norm": 1.8095559992901746, + "learning_rate": 3.595644092594306e-06, + "loss": 0.373, + "step": 27390 + }, + { + "epoch": 0.6029275808936826, + "grad_norm": 1.6114031308531471, + "learning_rate": 3.5953019887657903e-06, + "loss": 0.3879, + "step": 27391 + }, + { + "epoch": 0.6029495927801013, + "grad_norm": 1.2987895564294027, + "learning_rate": 3.5949598920765782e-06, + "loss": 0.2177, + "step": 27392 + }, + { + "epoch": 0.6029716046665199, + "grad_norm": 1.5541677708408002, + "learning_rate": 3.5946178025284053e-06, + "loss": 0.3592, + "step": 27393 + }, + { + "epoch": 0.6029936165529386, + "grad_norm": 1.5726085260422196, + "learning_rate": 3.594275720123014e-06, + "loss": 0.2953, + "step": 27394 + }, + { + "epoch": 0.6030156284393573, + "grad_norm": 1.5070632025724775, + "learning_rate": 3.5939336448621386e-06, + "loss": 0.3082, + "step": 27395 + }, + { + "epoch": 0.6030376403257759, + "grad_norm": 1.6565453677389284, + "learning_rate": 3.5935915767475194e-06, + "loss": 0.4004, + "step": 27396 + }, + { + "epoch": 0.6030596522121946, + "grad_norm": 1.6026916291387423, + "learning_rate": 3.5932495157808965e-06, + "loss": 0.323, + "step": 27397 + }, + { + "epoch": 0.6030816640986133, + "grad_norm": 0.6267117715038992, + "learning_rate": 3.592907461964004e-06, + "loss": 0.4977, + "step": 27398 + }, + { + "epoch": 0.6031036759850319, + "grad_norm": 0.6054477179523738, + "learning_rate": 3.592565415298587e-06, + "loss": 0.436, + "step": 27399 + }, + { + "epoch": 0.6031256878714506, + "grad_norm": 3.4755167498212858, + "learning_rate": 3.592223375786379e-06, + "loss": 0.2898, + "step": 27400 + }, + { + "epoch": 0.6031476997578693, + "grad_norm": 2.8377444519051784, + "learning_rate": 3.591881343429118e-06, + "loss": 0.3566, + "step": 27401 + }, + { + "epoch": 0.6031697116442879, + "grad_norm": 1.6422879025700208, + "learning_rate": 3.591539318228547e-06, + "loss": 0.3247, + "step": 27402 + }, + { + "epoch": 0.6031917235307066, + "grad_norm": 1.7086831345674975, + "learning_rate": 3.5911973001863985e-06, + "loss": 0.4044, + "step": 27403 + }, + { + "epoch": 0.6032137354171252, + "grad_norm": 1.491566410152203, + "learning_rate": 3.5908552893044157e-06, + "loss": 0.3565, + "step": 27404 + }, + { + "epoch": 0.6032357473035439, + "grad_norm": 1.3066137354068041, + "learning_rate": 3.5905132855843337e-06, + "loss": 0.3418, + "step": 27405 + }, + { + "epoch": 0.6032577591899626, + "grad_norm": 1.3355081369888644, + "learning_rate": 3.590171289027892e-06, + "loss": 0.3115, + "step": 27406 + }, + { + "epoch": 0.6032797710763812, + "grad_norm": 1.2363537535376046, + "learning_rate": 3.5898292996368294e-06, + "loss": 0.2742, + "step": 27407 + }, + { + "epoch": 0.6033017829627999, + "grad_norm": 1.702857659368673, + "learning_rate": 3.589487317412883e-06, + "loss": 0.3159, + "step": 27408 + }, + { + "epoch": 0.6033237948492186, + "grad_norm": 3.9581845203659163, + "learning_rate": 3.5891453423577897e-06, + "loss": 0.3612, + "step": 27409 + }, + { + "epoch": 0.6033458067356372, + "grad_norm": 1.3684556435439732, + "learning_rate": 3.5888033744732907e-06, + "loss": 0.3198, + "step": 27410 + }, + { + "epoch": 0.6033678186220559, + "grad_norm": 1.667856227318595, + "learning_rate": 3.5884614137611206e-06, + "loss": 0.3173, + "step": 27411 + }, + { + "epoch": 0.6033898305084746, + "grad_norm": 3.8307734483198823, + "learning_rate": 3.588119460223022e-06, + "loss": 0.3396, + "step": 27412 + }, + { + "epoch": 0.6034118423948932, + "grad_norm": 1.2027456976701523, + "learning_rate": 3.5877775138607285e-06, + "loss": 0.3522, + "step": 27413 + }, + { + "epoch": 0.6034338542813119, + "grad_norm": 1.5239911392014456, + "learning_rate": 3.5874355746759783e-06, + "loss": 0.3414, + "step": 27414 + }, + { + "epoch": 0.6034558661677306, + "grad_norm": 3.9668457598139235, + "learning_rate": 3.587093642670513e-06, + "loss": 0.4175, + "step": 27415 + }, + { + "epoch": 0.6034778780541492, + "grad_norm": 1.9125103821515208, + "learning_rate": 3.586751717846065e-06, + "loss": 0.3708, + "step": 27416 + }, + { + "epoch": 0.6034998899405679, + "grad_norm": 0.6322919400636056, + "learning_rate": 3.586409800204378e-06, + "loss": 0.4688, + "step": 27417 + }, + { + "epoch": 0.6035219018269866, + "grad_norm": 1.37205541059546, + "learning_rate": 3.5860678897471857e-06, + "loss": 0.3244, + "step": 27418 + }, + { + "epoch": 0.6035439137134052, + "grad_norm": 1.21244040573317, + "learning_rate": 3.5857259864762263e-06, + "loss": 0.3497, + "step": 27419 + }, + { + "epoch": 0.6035659255998239, + "grad_norm": 2.2251215421792354, + "learning_rate": 3.58538409039324e-06, + "loss": 0.381, + "step": 27420 + }, + { + "epoch": 0.6035879374862426, + "grad_norm": 1.2987702902997884, + "learning_rate": 3.5850422014999604e-06, + "loss": 0.2471, + "step": 27421 + }, + { + "epoch": 0.6036099493726612, + "grad_norm": 1.8837619545692157, + "learning_rate": 3.584700319798129e-06, + "loss": 0.3132, + "step": 27422 + }, + { + "epoch": 0.6036319612590799, + "grad_norm": 1.7517158769327648, + "learning_rate": 3.584358445289481e-06, + "loss": 0.3721, + "step": 27423 + }, + { + "epoch": 0.6036539731454986, + "grad_norm": 1.25836359813558, + "learning_rate": 3.5840165779757542e-06, + "loss": 0.3038, + "step": 27424 + }, + { + "epoch": 0.6036759850319172, + "grad_norm": 2.0317760444651665, + "learning_rate": 3.5836747178586896e-06, + "loss": 0.2688, + "step": 27425 + }, + { + "epoch": 0.6036979969183359, + "grad_norm": 1.1978866004563529, + "learning_rate": 3.5833328649400196e-06, + "loss": 0.3036, + "step": 27426 + }, + { + "epoch": 0.6037200088047545, + "grad_norm": 1.8349443971284196, + "learning_rate": 3.5829910192214834e-06, + "loss": 0.3036, + "step": 27427 + }, + { + "epoch": 0.6037420206911732, + "grad_norm": 1.4393879137312142, + "learning_rate": 3.582649180704819e-06, + "loss": 0.3112, + "step": 27428 + }, + { + "epoch": 0.6037640325775919, + "grad_norm": 0.6416464219045704, + "learning_rate": 3.5823073493917636e-06, + "loss": 0.4879, + "step": 27429 + }, + { + "epoch": 0.6037860444640105, + "grad_norm": 0.6077186636321753, + "learning_rate": 3.5819655252840567e-06, + "loss": 0.4623, + "step": 27430 + }, + { + "epoch": 0.6038080563504292, + "grad_norm": 2.0475810826993954, + "learning_rate": 3.5816237083834317e-06, + "loss": 0.3109, + "step": 27431 + }, + { + "epoch": 0.6038300682368479, + "grad_norm": 1.7058101617883212, + "learning_rate": 3.581281898691627e-06, + "loss": 0.337, + "step": 27432 + }, + { + "epoch": 0.6038520801232665, + "grad_norm": 1.572450682239065, + "learning_rate": 3.5809400962103823e-06, + "loss": 0.313, + "step": 27433 + }, + { + "epoch": 0.6038740920096852, + "grad_norm": 0.6061491960239382, + "learning_rate": 3.5805983009414304e-06, + "loss": 0.4471, + "step": 27434 + }, + { + "epoch": 0.6038961038961039, + "grad_norm": 1.418983131424428, + "learning_rate": 3.5802565128865128e-06, + "loss": 0.2572, + "step": 27435 + }, + { + "epoch": 0.6039181157825225, + "grad_norm": 1.48464393114018, + "learning_rate": 3.5799147320473655e-06, + "loss": 0.2348, + "step": 27436 + }, + { + "epoch": 0.6039401276689412, + "grad_norm": 1.7417860511578915, + "learning_rate": 3.5795729584257225e-06, + "loss": 0.4028, + "step": 27437 + }, + { + "epoch": 0.6039621395553599, + "grad_norm": 3.0336116978097034, + "learning_rate": 3.5792311920233265e-06, + "loss": 0.2828, + "step": 27438 + }, + { + "epoch": 0.6039841514417785, + "grad_norm": 2.832400404340169, + "learning_rate": 3.5788894328419083e-06, + "loss": 0.4137, + "step": 27439 + }, + { + "epoch": 0.6040061633281972, + "grad_norm": 1.4405364112455858, + "learning_rate": 3.578547680883209e-06, + "loss": 0.2429, + "step": 27440 + }, + { + "epoch": 0.6040281752146159, + "grad_norm": 1.7712671519784846, + "learning_rate": 3.5782059361489645e-06, + "loss": 0.2759, + "step": 27441 + }, + { + "epoch": 0.6040501871010345, + "grad_norm": 3.2786136759484177, + "learning_rate": 3.577864198640911e-06, + "loss": 0.3992, + "step": 27442 + }, + { + "epoch": 0.6040721989874532, + "grad_norm": 2.1347608414002415, + "learning_rate": 3.577522468360787e-06, + "loss": 0.313, + "step": 27443 + }, + { + "epoch": 0.6040942108738719, + "grad_norm": 2.142424130886169, + "learning_rate": 3.5771807453103268e-06, + "loss": 0.3643, + "step": 27444 + }, + { + "epoch": 0.6041162227602905, + "grad_norm": 1.9767121338281173, + "learning_rate": 3.5768390294912696e-06, + "loss": 0.4076, + "step": 27445 + }, + { + "epoch": 0.6041382346467092, + "grad_norm": 1.339973159452155, + "learning_rate": 3.576497320905351e-06, + "loss": 0.2741, + "step": 27446 + }, + { + "epoch": 0.6041602465331279, + "grad_norm": 2.7813482558793634, + "learning_rate": 3.576155619554308e-06, + "loss": 0.3423, + "step": 27447 + }, + { + "epoch": 0.6041822584195465, + "grad_norm": 3.9691843675710876, + "learning_rate": 3.5758139254398783e-06, + "loss": 0.3273, + "step": 27448 + }, + { + "epoch": 0.6042042703059652, + "grad_norm": 1.4510428214146314, + "learning_rate": 3.5754722385637963e-06, + "loss": 0.3074, + "step": 27449 + }, + { + "epoch": 0.6042262821923838, + "grad_norm": 1.5259099015314608, + "learning_rate": 3.5751305589277984e-06, + "loss": 0.3479, + "step": 27450 + }, + { + "epoch": 0.6042482940788025, + "grad_norm": 1.690622563130771, + "learning_rate": 3.5747888865336254e-06, + "loss": 0.321, + "step": 27451 + }, + { + "epoch": 0.6042703059652212, + "grad_norm": 1.400828202082664, + "learning_rate": 3.5744472213830072e-06, + "loss": 0.307, + "step": 27452 + }, + { + "epoch": 0.6042923178516398, + "grad_norm": 0.6241019923207829, + "learning_rate": 3.574105563477688e-06, + "loss": 0.4741, + "step": 27453 + }, + { + "epoch": 0.6043143297380585, + "grad_norm": 0.6284258766092936, + "learning_rate": 3.5737639128193984e-06, + "loss": 0.4771, + "step": 27454 + }, + { + "epoch": 0.6043363416244772, + "grad_norm": 1.8199935748017442, + "learning_rate": 3.5734222694098756e-06, + "loss": 0.3105, + "step": 27455 + }, + { + "epoch": 0.6043583535108958, + "grad_norm": 1.1672820837968991, + "learning_rate": 3.57308063325086e-06, + "loss": 0.2265, + "step": 27456 + }, + { + "epoch": 0.6043803653973145, + "grad_norm": 1.7306568698605465, + "learning_rate": 3.5727390043440823e-06, + "loss": 0.3919, + "step": 27457 + }, + { + "epoch": 0.6044023772837332, + "grad_norm": 3.0286455862743273, + "learning_rate": 3.572397382691283e-06, + "loss": 0.362, + "step": 27458 + }, + { + "epoch": 0.6044243891701518, + "grad_norm": 2.2729897216456916, + "learning_rate": 3.572055768294197e-06, + "loss": 0.344, + "step": 27459 + }, + { + "epoch": 0.6044464010565705, + "grad_norm": 1.367127325685908, + "learning_rate": 3.571714161154559e-06, + "loss": 0.2567, + "step": 27460 + }, + { + "epoch": 0.6044684129429893, + "grad_norm": 3.609607327210182, + "learning_rate": 3.571372561274109e-06, + "loss": 0.357, + "step": 27461 + }, + { + "epoch": 0.6044904248294078, + "grad_norm": 1.3832659036676123, + "learning_rate": 3.5710309686545785e-06, + "loss": 0.2595, + "step": 27462 + }, + { + "epoch": 0.6045124367158266, + "grad_norm": 2.333448583502327, + "learning_rate": 3.5706893832977068e-06, + "loss": 0.3004, + "step": 27463 + }, + { + "epoch": 0.6045344486022453, + "grad_norm": 1.5571928657548877, + "learning_rate": 3.5703478052052298e-06, + "loss": 0.3764, + "step": 27464 + }, + { + "epoch": 0.6045564604886638, + "grad_norm": 0.6233015839154902, + "learning_rate": 3.570006234378881e-06, + "loss": 0.4657, + "step": 27465 + }, + { + "epoch": 0.6045784723750826, + "grad_norm": 3.559421235159853, + "learning_rate": 3.5696646708204012e-06, + "loss": 0.2639, + "step": 27466 + }, + { + "epoch": 0.6046004842615013, + "grad_norm": 0.6278415340861284, + "learning_rate": 3.569323114531522e-06, + "loss": 0.4834, + "step": 27467 + }, + { + "epoch": 0.6046224961479199, + "grad_norm": 0.6156320864986279, + "learning_rate": 3.5689815655139804e-06, + "loss": 0.4349, + "step": 27468 + }, + { + "epoch": 0.6046445080343386, + "grad_norm": 1.1109168457466294, + "learning_rate": 3.5686400237695142e-06, + "loss": 0.2398, + "step": 27469 + }, + { + "epoch": 0.6046665199207573, + "grad_norm": 1.7349003393684417, + "learning_rate": 3.5682984892998553e-06, + "loss": 0.3865, + "step": 27470 + }, + { + "epoch": 0.6046885318071759, + "grad_norm": 1.768435374378835, + "learning_rate": 3.5679569621067443e-06, + "loss": 0.3229, + "step": 27471 + }, + { + "epoch": 0.6047105436935946, + "grad_norm": 2.0394298705850242, + "learning_rate": 3.5676154421919142e-06, + "loss": 0.3167, + "step": 27472 + }, + { + "epoch": 0.6047325555800132, + "grad_norm": 2.2046583851319976, + "learning_rate": 3.5672739295571002e-06, + "loss": 0.339, + "step": 27473 + }, + { + "epoch": 0.6047545674664319, + "grad_norm": 1.1653870257974062, + "learning_rate": 3.5669324242040413e-06, + "loss": 0.2845, + "step": 27474 + }, + { + "epoch": 0.6047765793528506, + "grad_norm": 0.639931334379668, + "learning_rate": 3.566590926134469e-06, + "loss": 0.4729, + "step": 27475 + }, + { + "epoch": 0.6047985912392692, + "grad_norm": 0.6438107444250746, + "learning_rate": 3.5662494353501216e-06, + "loss": 0.4716, + "step": 27476 + }, + { + "epoch": 0.6048206031256879, + "grad_norm": 0.991700858451092, + "learning_rate": 3.565907951852734e-06, + "loss": 0.2784, + "step": 27477 + }, + { + "epoch": 0.6048426150121066, + "grad_norm": 2.0567996806901077, + "learning_rate": 3.565566475644041e-06, + "loss": 0.3138, + "step": 27478 + }, + { + "epoch": 0.6048646268985252, + "grad_norm": 2.483926935908759, + "learning_rate": 3.565225006725781e-06, + "loss": 0.3238, + "step": 27479 + }, + { + "epoch": 0.6048866387849439, + "grad_norm": 0.6149242453320313, + "learning_rate": 3.5648835450996855e-06, + "loss": 0.4462, + "step": 27480 + }, + { + "epoch": 0.6049086506713626, + "grad_norm": 1.9383718691987686, + "learning_rate": 3.5645420907674927e-06, + "loss": 0.3465, + "step": 27481 + }, + { + "epoch": 0.6049306625577812, + "grad_norm": 2.025889360823935, + "learning_rate": 3.5642006437309374e-06, + "loss": 0.2704, + "step": 27482 + }, + { + "epoch": 0.6049526744441999, + "grad_norm": 1.6742078193007055, + "learning_rate": 3.5638592039917534e-06, + "loss": 0.3609, + "step": 27483 + }, + { + "epoch": 0.6049746863306186, + "grad_norm": 1.8498793227063621, + "learning_rate": 3.56351777155168e-06, + "loss": 0.3575, + "step": 27484 + }, + { + "epoch": 0.6049966982170372, + "grad_norm": 1.19899749040087, + "learning_rate": 3.5631763464124474e-06, + "loss": 0.2459, + "step": 27485 + }, + { + "epoch": 0.6050187101034559, + "grad_norm": 1.6465801876093773, + "learning_rate": 3.5628349285757944e-06, + "loss": 0.3546, + "step": 27486 + }, + { + "epoch": 0.6050407219898746, + "grad_norm": 1.3633102922720293, + "learning_rate": 3.5624935180434572e-06, + "loss": 0.2928, + "step": 27487 + }, + { + "epoch": 0.6050627338762932, + "grad_norm": 1.3146664939818817, + "learning_rate": 3.5621521148171654e-06, + "loss": 0.2681, + "step": 27488 + }, + { + "epoch": 0.6050847457627119, + "grad_norm": 1.431091489576816, + "learning_rate": 3.561810718898661e-06, + "loss": 0.2432, + "step": 27489 + }, + { + "epoch": 0.6051067576491306, + "grad_norm": 1.4865197797870375, + "learning_rate": 3.561469330289675e-06, + "loss": 0.3168, + "step": 27490 + }, + { + "epoch": 0.6051287695355492, + "grad_norm": 1.3481827065465093, + "learning_rate": 3.5611279489919427e-06, + "loss": 0.3492, + "step": 27491 + }, + { + "epoch": 0.6051507814219679, + "grad_norm": 1.470631194315275, + "learning_rate": 3.560786575007202e-06, + "loss": 0.3636, + "step": 27492 + }, + { + "epoch": 0.6051727933083866, + "grad_norm": 1.5302201230881305, + "learning_rate": 3.5604452083371834e-06, + "loss": 0.2544, + "step": 27493 + }, + { + "epoch": 0.6051948051948052, + "grad_norm": 1.422335772250942, + "learning_rate": 3.5601038489836255e-06, + "loss": 0.2671, + "step": 27494 + }, + { + "epoch": 0.6052168170812239, + "grad_norm": 1.4069769612225158, + "learning_rate": 3.5597624969482615e-06, + "loss": 0.217, + "step": 27495 + }, + { + "epoch": 0.6052388289676425, + "grad_norm": 1.6016597436984923, + "learning_rate": 3.5594211522328263e-06, + "loss": 0.2775, + "step": 27496 + }, + { + "epoch": 0.6052608408540612, + "grad_norm": 2.0307194811949545, + "learning_rate": 3.5590798148390575e-06, + "loss": 0.311, + "step": 27497 + }, + { + "epoch": 0.6052828527404799, + "grad_norm": 1.4019039238657756, + "learning_rate": 3.5587384847686855e-06, + "loss": 0.2418, + "step": 27498 + }, + { + "epoch": 0.6053048646268985, + "grad_norm": 1.7989520325638737, + "learning_rate": 3.5583971620234486e-06, + "loss": 0.2843, + "step": 27499 + }, + { + "epoch": 0.6053268765133172, + "grad_norm": 2.1825398234439315, + "learning_rate": 3.5580558466050814e-06, + "loss": 0.3685, + "step": 27500 + }, + { + "epoch": 0.6053488883997359, + "grad_norm": 1.5404889375569741, + "learning_rate": 3.557714538515314e-06, + "loss": 0.2701, + "step": 27501 + }, + { + "epoch": 0.6053709002861545, + "grad_norm": 1.8092363661330642, + "learning_rate": 3.557373237755888e-06, + "loss": 0.289, + "step": 27502 + }, + { + "epoch": 0.6053929121725732, + "grad_norm": 1.2675294166980067, + "learning_rate": 3.557031944328532e-06, + "loss": 0.2941, + "step": 27503 + }, + { + "epoch": 0.6054149240589919, + "grad_norm": 1.4790155443985427, + "learning_rate": 3.5566906582349847e-06, + "loss": 0.3474, + "step": 27504 + }, + { + "epoch": 0.6054369359454105, + "grad_norm": 1.6752266049713775, + "learning_rate": 3.556349379476981e-06, + "loss": 0.2669, + "step": 27505 + }, + { + "epoch": 0.6054589478318292, + "grad_norm": 1.4286366426871253, + "learning_rate": 3.556008108056249e-06, + "loss": 0.3568, + "step": 27506 + }, + { + "epoch": 0.6054809597182479, + "grad_norm": 1.316693981187867, + "learning_rate": 3.5556668439745324e-06, + "loss": 0.2818, + "step": 27507 + }, + { + "epoch": 0.6055029716046665, + "grad_norm": 1.5969178258754508, + "learning_rate": 3.555325587233559e-06, + "loss": 0.2903, + "step": 27508 + }, + { + "epoch": 0.6055249834910852, + "grad_norm": 2.222845723538594, + "learning_rate": 3.5549843378350645e-06, + "loss": 0.2824, + "step": 27509 + }, + { + "epoch": 0.6055469953775039, + "grad_norm": 1.9301353375056747, + "learning_rate": 3.5546430957807865e-06, + "loss": 0.3668, + "step": 27510 + }, + { + "epoch": 0.6055690072639225, + "grad_norm": 2.3078791870644046, + "learning_rate": 3.5543018610724546e-06, + "loss": 0.268, + "step": 27511 + }, + { + "epoch": 0.6055910191503412, + "grad_norm": 2.1628571511748054, + "learning_rate": 3.553960633711807e-06, + "loss": 0.3254, + "step": 27512 + }, + { + "epoch": 0.6056130310367599, + "grad_norm": 1.4649054299891953, + "learning_rate": 3.553619413700576e-06, + "loss": 0.285, + "step": 27513 + }, + { + "epoch": 0.6056350429231785, + "grad_norm": 1.580540902832113, + "learning_rate": 3.5532782010404953e-06, + "loss": 0.2995, + "step": 27514 + }, + { + "epoch": 0.6056570548095972, + "grad_norm": 1.6262120102722148, + "learning_rate": 3.5529369957333014e-06, + "loss": 0.3722, + "step": 27515 + }, + { + "epoch": 0.6056790666960159, + "grad_norm": 0.6693397102747848, + "learning_rate": 3.552595797780726e-06, + "loss": 0.477, + "step": 27516 + }, + { + "epoch": 0.6057010785824345, + "grad_norm": 2.046020872760219, + "learning_rate": 3.552254607184505e-06, + "loss": 0.3655, + "step": 27517 + }, + { + "epoch": 0.6057230904688532, + "grad_norm": 1.740144061246107, + "learning_rate": 3.5519134239463725e-06, + "loss": 0.3343, + "step": 27518 + }, + { + "epoch": 0.6057451023552718, + "grad_norm": 2.0246210135072187, + "learning_rate": 3.5515722480680592e-06, + "loss": 0.333, + "step": 27519 + }, + { + "epoch": 0.6057671142416905, + "grad_norm": 1.780700067638473, + "learning_rate": 3.551231079551305e-06, + "loss": 0.2895, + "step": 27520 + }, + { + "epoch": 0.6057891261281092, + "grad_norm": 1.4960200095225742, + "learning_rate": 3.550889918397838e-06, + "loss": 0.3006, + "step": 27521 + }, + { + "epoch": 0.6058111380145278, + "grad_norm": 1.3896406850076903, + "learning_rate": 3.550548764609396e-06, + "loss": 0.2762, + "step": 27522 + }, + { + "epoch": 0.6058331499009465, + "grad_norm": 1.3871845347756717, + "learning_rate": 3.5502076181877133e-06, + "loss": 0.2903, + "step": 27523 + }, + { + "epoch": 0.6058551617873652, + "grad_norm": 1.9023248693166548, + "learning_rate": 3.549866479134518e-06, + "loss": 0.2368, + "step": 27524 + }, + { + "epoch": 0.6058771736737838, + "grad_norm": 2.827505745429575, + "learning_rate": 3.5495253474515514e-06, + "loss": 0.2722, + "step": 27525 + }, + { + "epoch": 0.6058991855602025, + "grad_norm": 0.6224597683969536, + "learning_rate": 3.5491842231405434e-06, + "loss": 0.462, + "step": 27526 + }, + { + "epoch": 0.6059211974466212, + "grad_norm": 2.0662898266030796, + "learning_rate": 3.548843106203227e-06, + "loss": 0.281, + "step": 27527 + }, + { + "epoch": 0.6059432093330398, + "grad_norm": 1.8614535139713206, + "learning_rate": 3.5485019966413387e-06, + "loss": 0.3208, + "step": 27528 + }, + { + "epoch": 0.6059652212194585, + "grad_norm": 1.3661628396699452, + "learning_rate": 3.548160894456609e-06, + "loss": 0.276, + "step": 27529 + }, + { + "epoch": 0.6059872331058772, + "grad_norm": 1.5158875315210867, + "learning_rate": 3.5478197996507746e-06, + "loss": 0.3089, + "step": 27530 + }, + { + "epoch": 0.6060092449922958, + "grad_norm": 1.2159050781399185, + "learning_rate": 3.547478712225567e-06, + "loss": 0.2609, + "step": 27531 + }, + { + "epoch": 0.6060312568787145, + "grad_norm": 1.2196845504073754, + "learning_rate": 3.5471376321827194e-06, + "loss": 0.3195, + "step": 27532 + }, + { + "epoch": 0.6060532687651332, + "grad_norm": 1.9173951241241283, + "learning_rate": 3.546796559523969e-06, + "loss": 0.3407, + "step": 27533 + }, + { + "epoch": 0.6060752806515518, + "grad_norm": 1.5018885961498663, + "learning_rate": 3.5464554942510444e-06, + "loss": 0.3679, + "step": 27534 + }, + { + "epoch": 0.6060972925379705, + "grad_norm": 1.579645379775934, + "learning_rate": 3.546114436365682e-06, + "loss": 0.3039, + "step": 27535 + }, + { + "epoch": 0.6061193044243892, + "grad_norm": 0.61617248572361, + "learning_rate": 3.545773385869615e-06, + "loss": 0.4724, + "step": 27536 + }, + { + "epoch": 0.6061413163108078, + "grad_norm": 1.460225737590389, + "learning_rate": 3.545432342764574e-06, + "loss": 0.3056, + "step": 27537 + }, + { + "epoch": 0.6061633281972265, + "grad_norm": 1.3359864604323068, + "learning_rate": 3.5450913070522976e-06, + "loss": 0.2207, + "step": 27538 + }, + { + "epoch": 0.6061853400836452, + "grad_norm": 3.3673811801593727, + "learning_rate": 3.5447502787345145e-06, + "loss": 0.2997, + "step": 27539 + }, + { + "epoch": 0.6062073519700638, + "grad_norm": 1.4390858175412358, + "learning_rate": 3.5444092578129597e-06, + "loss": 0.3067, + "step": 27540 + }, + { + "epoch": 0.6062293638564825, + "grad_norm": 1.751633453269747, + "learning_rate": 3.5440682442893677e-06, + "loss": 0.3047, + "step": 27541 + }, + { + "epoch": 0.6062513757429012, + "grad_norm": 1.558871453248603, + "learning_rate": 3.543727238165467e-06, + "loss": 0.2909, + "step": 27542 + }, + { + "epoch": 0.6062733876293198, + "grad_norm": 2.0484113669264703, + "learning_rate": 3.5433862394429974e-06, + "loss": 0.3523, + "step": 27543 + }, + { + "epoch": 0.6062953995157385, + "grad_norm": 1.650720101103587, + "learning_rate": 3.5430452481236855e-06, + "loss": 0.2823, + "step": 27544 + }, + { + "epoch": 0.6063174114021571, + "grad_norm": 2.207157766226925, + "learning_rate": 3.5427042642092694e-06, + "loss": 0.292, + "step": 27545 + }, + { + "epoch": 0.6063394232885758, + "grad_norm": 1.5854680586550707, + "learning_rate": 3.5423632877014814e-06, + "loss": 0.2689, + "step": 27546 + }, + { + "epoch": 0.6063614351749945, + "grad_norm": 2.0626315658340406, + "learning_rate": 3.5420223186020507e-06, + "loss": 0.203, + "step": 27547 + }, + { + "epoch": 0.6063834470614131, + "grad_norm": 5.4863493292460195, + "learning_rate": 3.541681356912714e-06, + "loss": 0.279, + "step": 27548 + }, + { + "epoch": 0.6064054589478318, + "grad_norm": 1.6086412154860514, + "learning_rate": 3.5413404026352037e-06, + "loss": 0.2448, + "step": 27549 + }, + { + "epoch": 0.6064274708342505, + "grad_norm": 0.6094578394281462, + "learning_rate": 3.54099945577125e-06, + "loss": 0.4543, + "step": 27550 + }, + { + "epoch": 0.6064494827206691, + "grad_norm": 1.273781797009379, + "learning_rate": 3.5406585163225905e-06, + "loss": 0.2775, + "step": 27551 + }, + { + "epoch": 0.6064714946070878, + "grad_norm": 1.2079108271880228, + "learning_rate": 3.5403175842909533e-06, + "loss": 0.2858, + "step": 27552 + }, + { + "epoch": 0.6064935064935065, + "grad_norm": 1.5476257512421092, + "learning_rate": 3.5399766596780737e-06, + "loss": 0.3404, + "step": 27553 + }, + { + "epoch": 0.6065155183799251, + "grad_norm": 3.0230091009308273, + "learning_rate": 3.539635742485685e-06, + "loss": 0.2714, + "step": 27554 + }, + { + "epoch": 0.6065375302663438, + "grad_norm": 1.409222625969083, + "learning_rate": 3.5392948327155163e-06, + "loss": 0.3179, + "step": 27555 + }, + { + "epoch": 0.6065595421527625, + "grad_norm": 1.3717527432514034, + "learning_rate": 3.5389539303693054e-06, + "loss": 0.2847, + "step": 27556 + }, + { + "epoch": 0.6065815540391811, + "grad_norm": 2.5213005488154407, + "learning_rate": 3.5386130354487808e-06, + "loss": 0.3032, + "step": 27557 + }, + { + "epoch": 0.6066035659255998, + "grad_norm": 1.446070483659366, + "learning_rate": 3.538272147955677e-06, + "loss": 0.3739, + "step": 27558 + }, + { + "epoch": 0.6066255778120185, + "grad_norm": 1.3854643497338508, + "learning_rate": 3.537931267891727e-06, + "loss": 0.2464, + "step": 27559 + }, + { + "epoch": 0.6066475896984371, + "grad_norm": 1.405515944167953, + "learning_rate": 3.53759039525866e-06, + "loss": 0.2203, + "step": 27560 + }, + { + "epoch": 0.6066696015848558, + "grad_norm": 5.230240926058216, + "learning_rate": 3.5372495300582133e-06, + "loss": 0.3317, + "step": 27561 + }, + { + "epoch": 0.6066916134712745, + "grad_norm": 1.511056205350286, + "learning_rate": 3.5369086722921155e-06, + "loss": 0.2936, + "step": 27562 + }, + { + "epoch": 0.6067136253576931, + "grad_norm": 1.542944196548142, + "learning_rate": 3.5365678219621e-06, + "loss": 0.3297, + "step": 27563 + }, + { + "epoch": 0.6067356372441118, + "grad_norm": 1.8601686295863908, + "learning_rate": 3.536226979069902e-06, + "loss": 0.3306, + "step": 27564 + }, + { + "epoch": 0.6067576491305305, + "grad_norm": 1.3105448717014299, + "learning_rate": 3.5358861436172487e-06, + "loss": 0.336, + "step": 27565 + }, + { + "epoch": 0.6067796610169491, + "grad_norm": 2.093822269536983, + "learning_rate": 3.535545315605876e-06, + "loss": 0.3508, + "step": 27566 + }, + { + "epoch": 0.6068016729033678, + "grad_norm": 1.1977342864112928, + "learning_rate": 3.5352044950375155e-06, + "loss": 0.28, + "step": 27567 + }, + { + "epoch": 0.6068236847897864, + "grad_norm": 1.8468669992904552, + "learning_rate": 3.5348636819138983e-06, + "loss": 0.27, + "step": 27568 + }, + { + "epoch": 0.6068456966762051, + "grad_norm": 1.6296119481971856, + "learning_rate": 3.534522876236759e-06, + "loss": 0.3164, + "step": 27569 + }, + { + "epoch": 0.6068677085626238, + "grad_norm": 1.9637185904155692, + "learning_rate": 3.5341820780078253e-06, + "loss": 0.2519, + "step": 27570 + }, + { + "epoch": 0.6068897204490424, + "grad_norm": 1.3680218983740762, + "learning_rate": 3.5338412872288343e-06, + "loss": 0.2443, + "step": 27571 + }, + { + "epoch": 0.6069117323354611, + "grad_norm": 1.775034108535107, + "learning_rate": 3.5335005039015157e-06, + "loss": 0.3674, + "step": 27572 + }, + { + "epoch": 0.6069337442218798, + "grad_norm": 1.605296807908662, + "learning_rate": 3.5331597280275992e-06, + "loss": 0.3714, + "step": 27573 + }, + { + "epoch": 0.6069557561082984, + "grad_norm": 0.6418569233454373, + "learning_rate": 3.5328189596088224e-06, + "loss": 0.5016, + "step": 27574 + }, + { + "epoch": 0.6069777679947171, + "grad_norm": 1.682837819217988, + "learning_rate": 3.5324781986469116e-06, + "loss": 0.4095, + "step": 27575 + }, + { + "epoch": 0.6069997798811358, + "grad_norm": 0.639274768820352, + "learning_rate": 3.5321374451436024e-06, + "loss": 0.4627, + "step": 27576 + }, + { + "epoch": 0.6070217917675544, + "grad_norm": 1.5557835162998284, + "learning_rate": 3.5317966991006265e-06, + "loss": 0.3398, + "step": 27577 + }, + { + "epoch": 0.6070438036539731, + "grad_norm": 2.583554571285174, + "learning_rate": 3.5314559605197114e-06, + "loss": 0.3741, + "step": 27578 + }, + { + "epoch": 0.6070658155403919, + "grad_norm": 1.5716966658906417, + "learning_rate": 3.5311152294025952e-06, + "loss": 0.2881, + "step": 27579 + }, + { + "epoch": 0.6070878274268104, + "grad_norm": 1.1611857704027513, + "learning_rate": 3.5307745057510047e-06, + "loss": 0.239, + "step": 27580 + }, + { + "epoch": 0.6071098393132291, + "grad_norm": 1.1752669635722572, + "learning_rate": 3.530433789566674e-06, + "loss": 0.2818, + "step": 27581 + }, + { + "epoch": 0.6071318511996479, + "grad_norm": 1.9045968523903825, + "learning_rate": 3.5300930808513356e-06, + "loss": 0.3546, + "step": 27582 + }, + { + "epoch": 0.6071538630860664, + "grad_norm": 1.2045707706484876, + "learning_rate": 3.529752379606717e-06, + "loss": 0.304, + "step": 27583 + }, + { + "epoch": 0.6071758749724852, + "grad_norm": 1.35399273127175, + "learning_rate": 3.529411685834554e-06, + "loss": 0.27, + "step": 27584 + }, + { + "epoch": 0.6071978868589039, + "grad_norm": 0.6501424336910191, + "learning_rate": 3.5290709995365765e-06, + "loss": 0.4804, + "step": 27585 + }, + { + "epoch": 0.6072198987453225, + "grad_norm": 1.7471816869597083, + "learning_rate": 3.5287303207145164e-06, + "loss": 0.2816, + "step": 27586 + }, + { + "epoch": 0.6072419106317412, + "grad_norm": 0.6193026151574599, + "learning_rate": 3.5283896493701064e-06, + "loss": 0.4563, + "step": 27587 + }, + { + "epoch": 0.6072639225181599, + "grad_norm": 1.3409082716321237, + "learning_rate": 3.5280489855050744e-06, + "loss": 0.281, + "step": 27588 + }, + { + "epoch": 0.6072859344045785, + "grad_norm": 1.539719360091218, + "learning_rate": 3.5277083291211547e-06, + "loss": 0.279, + "step": 27589 + }, + { + "epoch": 0.6073079462909972, + "grad_norm": 1.2494107635239269, + "learning_rate": 3.5273676802200795e-06, + "loss": 0.2784, + "step": 27590 + }, + { + "epoch": 0.6073299581774158, + "grad_norm": 2.728733553320896, + "learning_rate": 3.527027038803575e-06, + "loss": 0.3964, + "step": 27591 + }, + { + "epoch": 0.6073519700638345, + "grad_norm": 1.2979414083894354, + "learning_rate": 3.5266864048733793e-06, + "loss": 0.3123, + "step": 27592 + }, + { + "epoch": 0.6073739819502532, + "grad_norm": 1.2504078962799756, + "learning_rate": 3.5263457784312184e-06, + "loss": 0.281, + "step": 27593 + }, + { + "epoch": 0.6073959938366718, + "grad_norm": 1.3508106353750424, + "learning_rate": 3.5260051594788264e-06, + "loss": 0.3349, + "step": 27594 + }, + { + "epoch": 0.6074180057230905, + "grad_norm": 1.6374087138585298, + "learning_rate": 3.5256645480179348e-06, + "loss": 0.3016, + "step": 27595 + }, + { + "epoch": 0.6074400176095092, + "grad_norm": 1.3054011567121868, + "learning_rate": 3.5253239440502705e-06, + "loss": 0.2688, + "step": 27596 + }, + { + "epoch": 0.6074620294959278, + "grad_norm": 1.2759133815468306, + "learning_rate": 3.5249833475775706e-06, + "loss": 0.2833, + "step": 27597 + }, + { + "epoch": 0.6074840413823465, + "grad_norm": 1.9926469363429602, + "learning_rate": 3.5246427586015607e-06, + "loss": 0.3177, + "step": 27598 + }, + { + "epoch": 0.6075060532687652, + "grad_norm": 1.910061804138708, + "learning_rate": 3.524302177123976e-06, + "loss": 0.2032, + "step": 27599 + }, + { + "epoch": 0.6075280651551838, + "grad_norm": 1.825469438467821, + "learning_rate": 3.5239616031465467e-06, + "loss": 0.2804, + "step": 27600 + }, + { + "epoch": 0.6075500770416025, + "grad_norm": 1.2164131157826643, + "learning_rate": 3.523621036671001e-06, + "loss": 0.241, + "step": 27601 + }, + { + "epoch": 0.6075720889280212, + "grad_norm": 1.5399584046929522, + "learning_rate": 3.5232804776990724e-06, + "loss": 0.3063, + "step": 27602 + }, + { + "epoch": 0.6075941008144398, + "grad_norm": 1.4099900974336625, + "learning_rate": 3.52293992623249e-06, + "loss": 0.3032, + "step": 27603 + }, + { + "epoch": 0.6076161127008585, + "grad_norm": 1.5734638686748612, + "learning_rate": 3.5225993822729865e-06, + "loss": 0.4125, + "step": 27604 + }, + { + "epoch": 0.6076381245872772, + "grad_norm": 6.462439428401146, + "learning_rate": 3.522258845822294e-06, + "loss": 0.2816, + "step": 27605 + }, + { + "epoch": 0.6076601364736958, + "grad_norm": 1.2477269744432673, + "learning_rate": 3.521918316882138e-06, + "loss": 0.2865, + "step": 27606 + }, + { + "epoch": 0.6076821483601145, + "grad_norm": 3.336473169853591, + "learning_rate": 3.5215777954542535e-06, + "loss": 0.296, + "step": 27607 + }, + { + "epoch": 0.6077041602465332, + "grad_norm": 0.6535413725146421, + "learning_rate": 3.5212372815403717e-06, + "loss": 0.4697, + "step": 27608 + }, + { + "epoch": 0.6077261721329518, + "grad_norm": 1.607485596907742, + "learning_rate": 3.5208967751422176e-06, + "loss": 0.3203, + "step": 27609 + }, + { + "epoch": 0.6077481840193705, + "grad_norm": 1.614827943137988, + "learning_rate": 3.52055627626153e-06, + "loss": 0.3166, + "step": 27610 + }, + { + "epoch": 0.6077701959057892, + "grad_norm": 1.3623332180676684, + "learning_rate": 3.5202157849000317e-06, + "loss": 0.2355, + "step": 27611 + }, + { + "epoch": 0.6077922077922078, + "grad_norm": 3.023151634923806, + "learning_rate": 3.519875301059459e-06, + "loss": 0.2629, + "step": 27612 + }, + { + "epoch": 0.6078142196786265, + "grad_norm": 1.6410262235746345, + "learning_rate": 3.5195348247415405e-06, + "loss": 0.2971, + "step": 27613 + }, + { + "epoch": 0.6078362315650451, + "grad_norm": 0.6520381018361566, + "learning_rate": 3.5191943559480036e-06, + "loss": 0.4872, + "step": 27614 + }, + { + "epoch": 0.6078582434514638, + "grad_norm": 1.9293428615200978, + "learning_rate": 3.518853894680585e-06, + "loss": 0.3024, + "step": 27615 + }, + { + "epoch": 0.6078802553378825, + "grad_norm": 1.2728813095113032, + "learning_rate": 3.518513440941008e-06, + "loss": 0.2649, + "step": 27616 + }, + { + "epoch": 0.6079022672243011, + "grad_norm": 1.3721471077954894, + "learning_rate": 3.5181729947310095e-06, + "loss": 0.3135, + "step": 27617 + }, + { + "epoch": 0.6079242791107198, + "grad_norm": 2.6404634099493176, + "learning_rate": 3.5178325560523165e-06, + "loss": 0.282, + "step": 27618 + }, + { + "epoch": 0.6079462909971385, + "grad_norm": 1.9230209396844995, + "learning_rate": 3.5174921249066573e-06, + "loss": 0.2529, + "step": 27619 + }, + { + "epoch": 0.6079683028835571, + "grad_norm": 0.6413180829703722, + "learning_rate": 3.5171517012957657e-06, + "loss": 0.4801, + "step": 27620 + }, + { + "epoch": 0.6079903147699758, + "grad_norm": 2.0253380251616537, + "learning_rate": 3.5168112852213696e-06, + "loss": 0.3691, + "step": 27621 + }, + { + "epoch": 0.6080123266563945, + "grad_norm": 2.139505638925131, + "learning_rate": 3.5164708766852013e-06, + "loss": 0.2679, + "step": 27622 + }, + { + "epoch": 0.6080343385428131, + "grad_norm": 1.4674703858058118, + "learning_rate": 3.5161304756889904e-06, + "loss": 0.317, + "step": 27623 + }, + { + "epoch": 0.6080563504292318, + "grad_norm": 1.7095705049905203, + "learning_rate": 3.5157900822344643e-06, + "loss": 0.3719, + "step": 27624 + }, + { + "epoch": 0.6080783623156505, + "grad_norm": 1.27221623096752, + "learning_rate": 3.5154496963233558e-06, + "loss": 0.2336, + "step": 27625 + }, + { + "epoch": 0.6081003742020691, + "grad_norm": 1.4151362750196175, + "learning_rate": 3.5151093179573952e-06, + "loss": 0.3102, + "step": 27626 + }, + { + "epoch": 0.6081223860884878, + "grad_norm": 0.6408002186666508, + "learning_rate": 3.5147689471383085e-06, + "loss": 0.4677, + "step": 27627 + }, + { + "epoch": 0.6081443979749065, + "grad_norm": 1.195321305068391, + "learning_rate": 3.5144285838678316e-06, + "loss": 0.2852, + "step": 27628 + }, + { + "epoch": 0.6081664098613251, + "grad_norm": 1.4013794880088941, + "learning_rate": 3.5140882281476883e-06, + "loss": 0.2828, + "step": 27629 + }, + { + "epoch": 0.6081884217477438, + "grad_norm": 2.4669531214523004, + "learning_rate": 3.513747879979613e-06, + "loss": 0.313, + "step": 27630 + }, + { + "epoch": 0.6082104336341625, + "grad_norm": 0.6277142750060415, + "learning_rate": 3.513407539365335e-06, + "loss": 0.4611, + "step": 27631 + }, + { + "epoch": 0.6082324455205811, + "grad_norm": 1.9560433835356679, + "learning_rate": 3.5130672063065807e-06, + "loss": 0.3597, + "step": 27632 + }, + { + "epoch": 0.6082544574069998, + "grad_norm": 0.6324351444945787, + "learning_rate": 3.512726880805083e-06, + "loss": 0.4661, + "step": 27633 + }, + { + "epoch": 0.6082764692934185, + "grad_norm": 0.5943033649258633, + "learning_rate": 3.51238656286257e-06, + "loss": 0.4698, + "step": 27634 + }, + { + "epoch": 0.6082984811798371, + "grad_norm": 0.6278635581577993, + "learning_rate": 3.5120462524807726e-06, + "loss": 0.4819, + "step": 27635 + }, + { + "epoch": 0.6083204930662558, + "grad_norm": 1.6163283654234981, + "learning_rate": 3.511705949661421e-06, + "loss": 0.3506, + "step": 27636 + }, + { + "epoch": 0.6083425049526744, + "grad_norm": 1.4804138236731272, + "learning_rate": 3.5113656544062413e-06, + "loss": 0.264, + "step": 27637 + }, + { + "epoch": 0.6083645168390931, + "grad_norm": 2.2129322374874296, + "learning_rate": 3.511025366716966e-06, + "loss": 0.2965, + "step": 27638 + }, + { + "epoch": 0.6083865287255118, + "grad_norm": 1.3887821145817534, + "learning_rate": 3.5106850865953236e-06, + "loss": 0.2621, + "step": 27639 + }, + { + "epoch": 0.6084085406119304, + "grad_norm": 1.4318029183084287, + "learning_rate": 3.510344814043045e-06, + "loss": 0.3045, + "step": 27640 + }, + { + "epoch": 0.6084305524983491, + "grad_norm": 1.3551473792391002, + "learning_rate": 3.510004549061859e-06, + "loss": 0.3014, + "step": 27641 + }, + { + "epoch": 0.6084525643847678, + "grad_norm": 1.6251567772615625, + "learning_rate": 3.5096642916534925e-06, + "loss": 0.3337, + "step": 27642 + }, + { + "epoch": 0.6084745762711864, + "grad_norm": 1.3334211113962704, + "learning_rate": 3.509324041819677e-06, + "loss": 0.2977, + "step": 27643 + }, + { + "epoch": 0.6084965881576051, + "grad_norm": 1.5131982795360885, + "learning_rate": 3.508983799562142e-06, + "loss": 0.3423, + "step": 27644 + }, + { + "epoch": 0.6085186000440238, + "grad_norm": 1.2140675505659042, + "learning_rate": 3.5086435648826166e-06, + "loss": 0.1889, + "step": 27645 + }, + { + "epoch": 0.6085406119304424, + "grad_norm": 1.9799759661345877, + "learning_rate": 3.5083033377828313e-06, + "loss": 0.2623, + "step": 27646 + }, + { + "epoch": 0.6085626238168611, + "grad_norm": 1.3102966206349997, + "learning_rate": 3.5079631182645106e-06, + "loss": 0.2517, + "step": 27647 + }, + { + "epoch": 0.6085846357032798, + "grad_norm": 0.6102631897096056, + "learning_rate": 3.5076229063293886e-06, + "loss": 0.4764, + "step": 27648 + }, + { + "epoch": 0.6086066475896984, + "grad_norm": 1.4029892182102572, + "learning_rate": 3.507282701979194e-06, + "loss": 0.2478, + "step": 27649 + }, + { + "epoch": 0.6086286594761171, + "grad_norm": 0.6368299479332464, + "learning_rate": 3.5069425052156514e-06, + "loss": 0.4843, + "step": 27650 + }, + { + "epoch": 0.6086506713625358, + "grad_norm": 0.6301596173178406, + "learning_rate": 3.506602316040495e-06, + "loss": 0.4939, + "step": 27651 + }, + { + "epoch": 0.6086726832489544, + "grad_norm": 1.7878763683833063, + "learning_rate": 3.50626213445545e-06, + "loss": 0.2392, + "step": 27652 + }, + { + "epoch": 0.6086946951353731, + "grad_norm": 0.6030714076710545, + "learning_rate": 3.5059219604622485e-06, + "loss": 0.4775, + "step": 27653 + }, + { + "epoch": 0.6087167070217918, + "grad_norm": 1.3660164904070062, + "learning_rate": 3.5055817940626184e-06, + "loss": 0.2701, + "step": 27654 + }, + { + "epoch": 0.6087387189082104, + "grad_norm": 1.3453599099250737, + "learning_rate": 3.5052416352582865e-06, + "loss": 0.3831, + "step": 27655 + }, + { + "epoch": 0.6087607307946291, + "grad_norm": 2.8022291425625876, + "learning_rate": 3.504901484050984e-06, + "loss": 0.312, + "step": 27656 + }, + { + "epoch": 0.6087827426810478, + "grad_norm": 1.339992702929868, + "learning_rate": 3.504561340442438e-06, + "loss": 0.2949, + "step": 27657 + }, + { + "epoch": 0.6088047545674664, + "grad_norm": 1.5278763581516392, + "learning_rate": 3.5042212044343806e-06, + "loss": 0.3011, + "step": 27658 + }, + { + "epoch": 0.6088267664538851, + "grad_norm": 1.5369492544921015, + "learning_rate": 3.503881076028538e-06, + "loss": 0.3113, + "step": 27659 + }, + { + "epoch": 0.6088487783403037, + "grad_norm": 1.3223858718770287, + "learning_rate": 3.5035409552266364e-06, + "loss": 0.3031, + "step": 27660 + }, + { + "epoch": 0.6088707902267224, + "grad_norm": 1.9484293722897563, + "learning_rate": 3.5032008420304094e-06, + "loss": 0.2874, + "step": 27661 + }, + { + "epoch": 0.6088928021131411, + "grad_norm": 1.5305642229480956, + "learning_rate": 3.5028607364415814e-06, + "loss": 0.2451, + "step": 27662 + }, + { + "epoch": 0.6089148139995597, + "grad_norm": 1.7669083998939261, + "learning_rate": 3.502520638461885e-06, + "loss": 0.317, + "step": 27663 + }, + { + "epoch": 0.6089368258859784, + "grad_norm": 2.334108836281458, + "learning_rate": 3.502180548093046e-06, + "loss": 0.3102, + "step": 27664 + }, + { + "epoch": 0.6089588377723971, + "grad_norm": 1.4915378708267775, + "learning_rate": 3.501840465336793e-06, + "loss": 0.2629, + "step": 27665 + }, + { + "epoch": 0.6089808496588157, + "grad_norm": 1.6663662733974323, + "learning_rate": 3.501500390194855e-06, + "loss": 0.3012, + "step": 27666 + }, + { + "epoch": 0.6090028615452344, + "grad_norm": 2.2075986107366568, + "learning_rate": 3.5011603226689615e-06, + "loss": 0.3344, + "step": 27667 + }, + { + "epoch": 0.6090248734316531, + "grad_norm": 1.2091123458364674, + "learning_rate": 3.500820262760838e-06, + "loss": 0.3217, + "step": 27668 + }, + { + "epoch": 0.6090468853180717, + "grad_norm": 0.6337497390433354, + "learning_rate": 3.5004802104722156e-06, + "loss": 0.4687, + "step": 27669 + }, + { + "epoch": 0.6090688972044904, + "grad_norm": 1.4661113820354128, + "learning_rate": 3.5001401658048207e-06, + "loss": 0.3768, + "step": 27670 + }, + { + "epoch": 0.6090909090909091, + "grad_norm": 0.6234746876178393, + "learning_rate": 3.499800128760383e-06, + "loss": 0.4601, + "step": 27671 + }, + { + "epoch": 0.6091129209773277, + "grad_norm": 1.4718898904003699, + "learning_rate": 3.4994600993406312e-06, + "loss": 0.36, + "step": 27672 + }, + { + "epoch": 0.6091349328637464, + "grad_norm": 0.602314754285867, + "learning_rate": 3.4991200775472913e-06, + "loss": 0.4622, + "step": 27673 + }, + { + "epoch": 0.6091569447501651, + "grad_norm": 2.326000669446318, + "learning_rate": 3.4987800633820924e-06, + "loss": 0.303, + "step": 27674 + }, + { + "epoch": 0.6091789566365837, + "grad_norm": 1.6493745007283793, + "learning_rate": 3.4984400568467626e-06, + "loss": 0.2904, + "step": 27675 + }, + { + "epoch": 0.6092009685230024, + "grad_norm": 1.4614887713902012, + "learning_rate": 3.4981000579430303e-06, + "loss": 0.3489, + "step": 27676 + }, + { + "epoch": 0.6092229804094211, + "grad_norm": 1.6319982918652756, + "learning_rate": 3.4977600666726255e-06, + "loss": 0.3395, + "step": 27677 + }, + { + "epoch": 0.6092449922958397, + "grad_norm": 2.175705081783766, + "learning_rate": 3.4974200830372706e-06, + "loss": 0.3701, + "step": 27678 + }, + { + "epoch": 0.6092670041822584, + "grad_norm": 1.3198935290741842, + "learning_rate": 3.4970801070386995e-06, + "loss": 0.2835, + "step": 27679 + }, + { + "epoch": 0.6092890160686771, + "grad_norm": 1.3140932376941972, + "learning_rate": 3.496740138678636e-06, + "loss": 0.3191, + "step": 27680 + }, + { + "epoch": 0.6093110279550957, + "grad_norm": 1.5678086117680279, + "learning_rate": 3.49640017795881e-06, + "loss": 0.2711, + "step": 27681 + }, + { + "epoch": 0.6093330398415144, + "grad_norm": 0.6287916712691161, + "learning_rate": 3.4960602248809507e-06, + "loss": 0.4496, + "step": 27682 + }, + { + "epoch": 0.6093550517279331, + "grad_norm": 1.3197797873739563, + "learning_rate": 3.4957202794467813e-06, + "loss": 0.3129, + "step": 27683 + }, + { + "epoch": 0.6093770636143517, + "grad_norm": 0.6148665715395973, + "learning_rate": 3.4953803416580347e-06, + "loss": 0.4334, + "step": 27684 + }, + { + "epoch": 0.6093990755007704, + "grad_norm": 0.601193423591085, + "learning_rate": 3.4950404115164344e-06, + "loss": 0.4543, + "step": 27685 + }, + { + "epoch": 0.609421087387189, + "grad_norm": 1.6573028474355824, + "learning_rate": 3.4947004890237124e-06, + "loss": 0.2708, + "step": 27686 + }, + { + "epoch": 0.6094430992736077, + "grad_norm": 0.6174822507563804, + "learning_rate": 3.4943605741815923e-06, + "loss": 0.4605, + "step": 27687 + }, + { + "epoch": 0.6094651111600264, + "grad_norm": 0.6191644091474666, + "learning_rate": 3.4940206669918024e-06, + "loss": 0.4383, + "step": 27688 + }, + { + "epoch": 0.609487123046445, + "grad_norm": 1.234704523886852, + "learning_rate": 3.4936807674560725e-06, + "loss": 0.2627, + "step": 27689 + }, + { + "epoch": 0.6095091349328637, + "grad_norm": 1.9817600413410519, + "learning_rate": 3.4933408755761296e-06, + "loss": 0.2953, + "step": 27690 + }, + { + "epoch": 0.6095311468192824, + "grad_norm": 1.3860318294152503, + "learning_rate": 3.4930009913536982e-06, + "loss": 0.295, + "step": 27691 + }, + { + "epoch": 0.609553158705701, + "grad_norm": 0.623715218564321, + "learning_rate": 3.4926611147905094e-06, + "loss": 0.46, + "step": 27692 + }, + { + "epoch": 0.6095751705921197, + "grad_norm": 1.6079981824702554, + "learning_rate": 3.492321245888287e-06, + "loss": 0.3343, + "step": 27693 + }, + { + "epoch": 0.6095971824785384, + "grad_norm": 1.1896600507101724, + "learning_rate": 3.491981384648763e-06, + "loss": 0.168, + "step": 27694 + }, + { + "epoch": 0.609619194364957, + "grad_norm": 1.806510448886957, + "learning_rate": 3.4916415310736624e-06, + "loss": 0.2995, + "step": 27695 + }, + { + "epoch": 0.6096412062513757, + "grad_norm": 1.4063656181271966, + "learning_rate": 3.4913016851647102e-06, + "loss": 0.3365, + "step": 27696 + }, + { + "epoch": 0.6096632181377944, + "grad_norm": 1.5107973625426965, + "learning_rate": 3.4909618469236373e-06, + "loss": 0.3083, + "step": 27697 + }, + { + "epoch": 0.609685230024213, + "grad_norm": 1.7065865240713214, + "learning_rate": 3.4906220163521674e-06, + "loss": 0.3187, + "step": 27698 + }, + { + "epoch": 0.6097072419106317, + "grad_norm": 1.8942612813666602, + "learning_rate": 3.4902821934520314e-06, + "loss": 0.2987, + "step": 27699 + }, + { + "epoch": 0.6097292537970505, + "grad_norm": 1.3601134913697706, + "learning_rate": 3.4899423782249555e-06, + "loss": 0.391, + "step": 27700 + }, + { + "epoch": 0.609751265683469, + "grad_norm": 1.5489247798396264, + "learning_rate": 3.4896025706726634e-06, + "loss": 0.2725, + "step": 27701 + }, + { + "epoch": 0.6097732775698878, + "grad_norm": 2.1874857832230927, + "learning_rate": 3.489262770796886e-06, + "loss": 0.2404, + "step": 27702 + }, + { + "epoch": 0.6097952894563065, + "grad_norm": 1.4284441613034262, + "learning_rate": 3.4889229785993485e-06, + "loss": 0.3277, + "step": 27703 + }, + { + "epoch": 0.609817301342725, + "grad_norm": 1.571938060472137, + "learning_rate": 3.48858319408178e-06, + "loss": 0.3069, + "step": 27704 + }, + { + "epoch": 0.6098393132291438, + "grad_norm": 1.433415458986578, + "learning_rate": 3.4882434172459044e-06, + "loss": 0.3582, + "step": 27705 + }, + { + "epoch": 0.6098613251155625, + "grad_norm": 1.5531569589833567, + "learning_rate": 3.4879036480934495e-06, + "loss": 0.3728, + "step": 27706 + }, + { + "epoch": 0.609883337001981, + "grad_norm": 1.7564071422588081, + "learning_rate": 3.487563886626144e-06, + "loss": 0.3606, + "step": 27707 + }, + { + "epoch": 0.6099053488883998, + "grad_norm": 0.6251724239807076, + "learning_rate": 3.4872241328457134e-06, + "loss": 0.4902, + "step": 27708 + }, + { + "epoch": 0.6099273607748184, + "grad_norm": 1.8859976402973595, + "learning_rate": 3.486884386753884e-06, + "loss": 0.2567, + "step": 27709 + }, + { + "epoch": 0.609949372661237, + "grad_norm": 1.4867382080162208, + "learning_rate": 3.4865446483523824e-06, + "loss": 0.3562, + "step": 27710 + }, + { + "epoch": 0.6099713845476558, + "grad_norm": 2.4871302593463334, + "learning_rate": 3.486204917642936e-06, + "loss": 0.2461, + "step": 27711 + }, + { + "epoch": 0.6099933964340744, + "grad_norm": 0.6209204407530207, + "learning_rate": 3.4858651946272716e-06, + "loss": 0.4688, + "step": 27712 + }, + { + "epoch": 0.6100154083204931, + "grad_norm": 1.3682460076404184, + "learning_rate": 3.485525479307117e-06, + "loss": 0.2649, + "step": 27713 + }, + { + "epoch": 0.6100374202069118, + "grad_norm": 1.1874217230540816, + "learning_rate": 3.4851857716841954e-06, + "loss": 0.2495, + "step": 27714 + }, + { + "epoch": 0.6100594320933304, + "grad_norm": 1.6993773579555738, + "learning_rate": 3.4848460717602362e-06, + "loss": 0.3195, + "step": 27715 + }, + { + "epoch": 0.6100814439797491, + "grad_norm": 1.731993187052846, + "learning_rate": 3.4845063795369645e-06, + "loss": 0.3083, + "step": 27716 + }, + { + "epoch": 0.6101034558661678, + "grad_norm": 2.316110412539757, + "learning_rate": 3.484166695016108e-06, + "loss": 0.3211, + "step": 27717 + }, + { + "epoch": 0.6101254677525864, + "grad_norm": 1.725072328486716, + "learning_rate": 3.4838270181993932e-06, + "loss": 0.3088, + "step": 27718 + }, + { + "epoch": 0.6101474796390051, + "grad_norm": 1.6098657904858797, + "learning_rate": 3.4834873490885437e-06, + "loss": 0.3179, + "step": 27719 + }, + { + "epoch": 0.6101694915254238, + "grad_norm": 0.6116605681332749, + "learning_rate": 3.4831476876852895e-06, + "loss": 0.4508, + "step": 27720 + }, + { + "epoch": 0.6101915034118424, + "grad_norm": 2.696702739775281, + "learning_rate": 3.482808033991354e-06, + "loss": 0.291, + "step": 27721 + }, + { + "epoch": 0.6102135152982611, + "grad_norm": 1.710793446370003, + "learning_rate": 3.4824683880084663e-06, + "loss": 0.387, + "step": 27722 + }, + { + "epoch": 0.6102355271846798, + "grad_norm": 1.5227652380424235, + "learning_rate": 3.4821287497383507e-06, + "loss": 0.3174, + "step": 27723 + }, + { + "epoch": 0.6102575390710984, + "grad_norm": 1.7430462823597404, + "learning_rate": 3.4817891191827324e-06, + "loss": 0.328, + "step": 27724 + }, + { + "epoch": 0.6102795509575171, + "grad_norm": 1.440727645615159, + "learning_rate": 3.48144949634334e-06, + "loss": 0.2625, + "step": 27725 + }, + { + "epoch": 0.6103015628439358, + "grad_norm": 1.2408043274835896, + "learning_rate": 3.4811098812218995e-06, + "loss": 0.2906, + "step": 27726 + }, + { + "epoch": 0.6103235747303544, + "grad_norm": 1.3524834053104806, + "learning_rate": 3.4807702738201344e-06, + "loss": 0.3157, + "step": 27727 + }, + { + "epoch": 0.6103455866167731, + "grad_norm": 2.8345972719764188, + "learning_rate": 3.480430674139773e-06, + "loss": 0.2579, + "step": 27728 + }, + { + "epoch": 0.6103675985031918, + "grad_norm": 2.0140895026711316, + "learning_rate": 3.4800910821825394e-06, + "loss": 0.2468, + "step": 27729 + }, + { + "epoch": 0.6103896103896104, + "grad_norm": 1.9617967204326614, + "learning_rate": 3.4797514979501624e-06, + "loss": 0.3152, + "step": 27730 + }, + { + "epoch": 0.6104116222760291, + "grad_norm": 1.1579326802738092, + "learning_rate": 3.479411921444367e-06, + "loss": 0.2545, + "step": 27731 + }, + { + "epoch": 0.6104336341624477, + "grad_norm": 0.6059889666273212, + "learning_rate": 3.479072352666877e-06, + "loss": 0.4448, + "step": 27732 + }, + { + "epoch": 0.6104556460488664, + "grad_norm": 0.6025205942501709, + "learning_rate": 3.47873279161942e-06, + "loss": 0.4368, + "step": 27733 + }, + { + "epoch": 0.6104776579352851, + "grad_norm": 0.6065561391339951, + "learning_rate": 3.47839323830372e-06, + "loss": 0.4314, + "step": 27734 + }, + { + "epoch": 0.6104996698217037, + "grad_norm": 1.698903264778256, + "learning_rate": 3.478053692721506e-06, + "loss": 0.3787, + "step": 27735 + }, + { + "epoch": 0.6105216817081224, + "grad_norm": 1.2913016633251553, + "learning_rate": 3.477714154874503e-06, + "loss": 0.2832, + "step": 27736 + }, + { + "epoch": 0.6105436935945411, + "grad_norm": 0.6335034910625136, + "learning_rate": 3.4773746247644335e-06, + "loss": 0.4637, + "step": 27737 + }, + { + "epoch": 0.6105657054809597, + "grad_norm": 1.3968690967048938, + "learning_rate": 3.477035102393026e-06, + "loss": 0.3477, + "step": 27738 + }, + { + "epoch": 0.6105877173673784, + "grad_norm": 2.0388088966576228, + "learning_rate": 3.4766955877620047e-06, + "loss": 0.286, + "step": 27739 + }, + { + "epoch": 0.6106097292537971, + "grad_norm": 1.3775605226806769, + "learning_rate": 3.4763560808730976e-06, + "loss": 0.2347, + "step": 27740 + }, + { + "epoch": 0.6106317411402157, + "grad_norm": 1.4707132429149115, + "learning_rate": 3.476016581728028e-06, + "loss": 0.3062, + "step": 27741 + }, + { + "epoch": 0.6106537530266344, + "grad_norm": 2.0152726981349973, + "learning_rate": 3.4756770903285198e-06, + "loss": 0.3356, + "step": 27742 + }, + { + "epoch": 0.6106757649130531, + "grad_norm": 1.713016975024751, + "learning_rate": 3.4753376066763023e-06, + "loss": 0.3324, + "step": 27743 + }, + { + "epoch": 0.6106977767994717, + "grad_norm": 1.4798510573373467, + "learning_rate": 3.474998130773098e-06, + "loss": 0.2571, + "step": 27744 + }, + { + "epoch": 0.6107197886858904, + "grad_norm": 1.6370314523124454, + "learning_rate": 3.474658662620635e-06, + "loss": 0.3212, + "step": 27745 + }, + { + "epoch": 0.6107418005723091, + "grad_norm": 1.1538826582223658, + "learning_rate": 3.4743192022206363e-06, + "loss": 0.3283, + "step": 27746 + }, + { + "epoch": 0.6107638124587277, + "grad_norm": 1.6984161621186176, + "learning_rate": 3.4739797495748273e-06, + "loss": 0.2868, + "step": 27747 + }, + { + "epoch": 0.6107858243451464, + "grad_norm": 1.3452772677644547, + "learning_rate": 3.4736403046849343e-06, + "loss": 0.2909, + "step": 27748 + }, + { + "epoch": 0.6108078362315651, + "grad_norm": 1.7395406872637138, + "learning_rate": 3.473300867552683e-06, + "loss": 0.3371, + "step": 27749 + }, + { + "epoch": 0.6108298481179837, + "grad_norm": 1.4416584272841513, + "learning_rate": 3.472961438179796e-06, + "loss": 0.2339, + "step": 27750 + }, + { + "epoch": 0.6108518600044024, + "grad_norm": 1.4788923182609497, + "learning_rate": 3.4726220165680002e-06, + "loss": 0.2679, + "step": 27751 + }, + { + "epoch": 0.6108738718908211, + "grad_norm": 1.6032882117884737, + "learning_rate": 3.4722826027190204e-06, + "loss": 0.3271, + "step": 27752 + }, + { + "epoch": 0.6108958837772397, + "grad_norm": 1.3350516852595973, + "learning_rate": 3.471943196634583e-06, + "loss": 0.3073, + "step": 27753 + }, + { + "epoch": 0.6109178956636584, + "grad_norm": 1.8694623712122613, + "learning_rate": 3.471603798316412e-06, + "loss": 0.3009, + "step": 27754 + }, + { + "epoch": 0.610939907550077, + "grad_norm": 1.515488469912304, + "learning_rate": 3.4712644077662305e-06, + "loss": 0.3662, + "step": 27755 + }, + { + "epoch": 0.6109619194364957, + "grad_norm": 0.6424400079594476, + "learning_rate": 3.470925024985766e-06, + "loss": 0.4715, + "step": 27756 + }, + { + "epoch": 0.6109839313229144, + "grad_norm": 1.469960010019197, + "learning_rate": 3.4705856499767423e-06, + "loss": 0.4196, + "step": 27757 + }, + { + "epoch": 0.611005943209333, + "grad_norm": 1.5609769695946671, + "learning_rate": 3.4702462827408855e-06, + "loss": 0.3101, + "step": 27758 + }, + { + "epoch": 0.6110279550957517, + "grad_norm": 1.3764232352008674, + "learning_rate": 3.469906923279919e-06, + "loss": 0.237, + "step": 27759 + }, + { + "epoch": 0.6110499669821704, + "grad_norm": 2.399267960701784, + "learning_rate": 3.4695675715955662e-06, + "loss": 0.4328, + "step": 27760 + }, + { + "epoch": 0.611071978868589, + "grad_norm": 1.4156082454412757, + "learning_rate": 3.469228227689555e-06, + "loss": 0.2712, + "step": 27761 + }, + { + "epoch": 0.6110939907550077, + "grad_norm": 2.52500194770405, + "learning_rate": 3.4688888915636077e-06, + "loss": 0.3187, + "step": 27762 + }, + { + "epoch": 0.6111160026414264, + "grad_norm": 3.1091408250540042, + "learning_rate": 3.468549563219452e-06, + "loss": 0.2373, + "step": 27763 + }, + { + "epoch": 0.611138014527845, + "grad_norm": 1.6570983157183146, + "learning_rate": 3.4682102426588093e-06, + "loss": 0.3613, + "step": 27764 + }, + { + "epoch": 0.6111600264142637, + "grad_norm": 1.608980100150409, + "learning_rate": 3.4678709298834047e-06, + "loss": 0.2388, + "step": 27765 + }, + { + "epoch": 0.6111820383006824, + "grad_norm": 1.2559992557930622, + "learning_rate": 3.467531624894964e-06, + "loss": 0.2285, + "step": 27766 + }, + { + "epoch": 0.611204050187101, + "grad_norm": 0.6174714796089021, + "learning_rate": 3.4671923276952123e-06, + "loss": 0.4514, + "step": 27767 + }, + { + "epoch": 0.6112260620735197, + "grad_norm": 0.6048845541421219, + "learning_rate": 3.466853038285871e-06, + "loss": 0.4472, + "step": 27768 + }, + { + "epoch": 0.6112480739599384, + "grad_norm": 1.4523177334499597, + "learning_rate": 3.466513756668667e-06, + "loss": 0.336, + "step": 27769 + }, + { + "epoch": 0.611270085846357, + "grad_norm": 0.6050786168521908, + "learning_rate": 3.4661744828453236e-06, + "loss": 0.4632, + "step": 27770 + }, + { + "epoch": 0.6112920977327757, + "grad_norm": 1.9877821276988703, + "learning_rate": 3.4658352168175657e-06, + "loss": 0.3001, + "step": 27771 + }, + { + "epoch": 0.6113141096191944, + "grad_norm": 1.6436556601460819, + "learning_rate": 3.4654959585871197e-06, + "loss": 0.2958, + "step": 27772 + }, + { + "epoch": 0.611336121505613, + "grad_norm": 1.8924707188111158, + "learning_rate": 3.4651567081557045e-06, + "loss": 0.2686, + "step": 27773 + }, + { + "epoch": 0.6113581333920317, + "grad_norm": 1.3017772456256103, + "learning_rate": 3.464817465525049e-06, + "loss": 0.3169, + "step": 27774 + }, + { + "epoch": 0.6113801452784504, + "grad_norm": 0.6180576334639127, + "learning_rate": 3.4644782306968744e-06, + "loss": 0.4893, + "step": 27775 + }, + { + "epoch": 0.611402157164869, + "grad_norm": 1.3791433384786955, + "learning_rate": 3.464139003672909e-06, + "loss": 0.2938, + "step": 27776 + }, + { + "epoch": 0.6114241690512877, + "grad_norm": 1.7337272681155491, + "learning_rate": 3.4637997844548726e-06, + "loss": 0.3172, + "step": 27777 + }, + { + "epoch": 0.6114461809377063, + "grad_norm": 1.332431588891369, + "learning_rate": 3.4634605730444904e-06, + "loss": 0.2538, + "step": 27778 + }, + { + "epoch": 0.611468192824125, + "grad_norm": 0.6196192709537258, + "learning_rate": 3.4631213694434872e-06, + "loss": 0.4681, + "step": 27779 + }, + { + "epoch": 0.6114902047105437, + "grad_norm": 0.629092455815768, + "learning_rate": 3.462782173653586e-06, + "loss": 0.4546, + "step": 27780 + }, + { + "epoch": 0.6115122165969623, + "grad_norm": 1.606280825632693, + "learning_rate": 3.4624429856765134e-06, + "loss": 0.3671, + "step": 27781 + }, + { + "epoch": 0.611534228483381, + "grad_norm": 1.894452252625675, + "learning_rate": 3.4621038055139906e-06, + "loss": 0.3151, + "step": 27782 + }, + { + "epoch": 0.6115562403697997, + "grad_norm": 1.3081755427771917, + "learning_rate": 3.4617646331677413e-06, + "loss": 0.3032, + "step": 27783 + }, + { + "epoch": 0.6115782522562183, + "grad_norm": 1.4354187022688154, + "learning_rate": 3.4614254686394905e-06, + "loss": 0.3135, + "step": 27784 + }, + { + "epoch": 0.611600264142637, + "grad_norm": 1.2733857141036224, + "learning_rate": 3.4610863119309614e-06, + "loss": 0.2532, + "step": 27785 + }, + { + "epoch": 0.6116222760290557, + "grad_norm": 1.4339610898613986, + "learning_rate": 3.4607471630438795e-06, + "loss": 0.3122, + "step": 27786 + }, + { + "epoch": 0.6116442879154743, + "grad_norm": 1.760137317478087, + "learning_rate": 3.4604080219799664e-06, + "loss": 0.2594, + "step": 27787 + }, + { + "epoch": 0.611666299801893, + "grad_norm": 1.6537224496797496, + "learning_rate": 3.460068888740945e-06, + "loss": 0.2836, + "step": 27788 + }, + { + "epoch": 0.6116883116883117, + "grad_norm": 0.5901022161399258, + "learning_rate": 3.4597297633285414e-06, + "loss": 0.466, + "step": 27789 + }, + { + "epoch": 0.6117103235747303, + "grad_norm": 1.5638477465510963, + "learning_rate": 3.4593906457444794e-06, + "loss": 0.3008, + "step": 27790 + }, + { + "epoch": 0.611732335461149, + "grad_norm": 2.016241135747434, + "learning_rate": 3.4590515359904794e-06, + "loss": 0.3754, + "step": 27791 + }, + { + "epoch": 0.6117543473475677, + "grad_norm": 1.3303239578546064, + "learning_rate": 3.4587124340682675e-06, + "loss": 0.3559, + "step": 27792 + }, + { + "epoch": 0.6117763592339863, + "grad_norm": 1.5511898362783734, + "learning_rate": 3.458373339979565e-06, + "loss": 0.3057, + "step": 27793 + }, + { + "epoch": 0.611798371120405, + "grad_norm": 0.6127557347797179, + "learning_rate": 3.4580342537260993e-06, + "loss": 0.4673, + "step": 27794 + }, + { + "epoch": 0.6118203830068237, + "grad_norm": 0.6130244024366582, + "learning_rate": 3.4576951753095905e-06, + "loss": 0.4662, + "step": 27795 + }, + { + "epoch": 0.6118423948932423, + "grad_norm": 1.2160922844133086, + "learning_rate": 3.4573561047317616e-06, + "loss": 0.2965, + "step": 27796 + }, + { + "epoch": 0.611864406779661, + "grad_norm": 1.3289736334575353, + "learning_rate": 3.457017041994337e-06, + "loss": 0.2489, + "step": 27797 + }, + { + "epoch": 0.6118864186660797, + "grad_norm": 1.2770089558305122, + "learning_rate": 3.4566779870990403e-06, + "loss": 0.2961, + "step": 27798 + }, + { + "epoch": 0.6119084305524983, + "grad_norm": 0.6293320859180758, + "learning_rate": 3.456338940047596e-06, + "loss": 0.4764, + "step": 27799 + }, + { + "epoch": 0.611930442438917, + "grad_norm": 1.4417470547102123, + "learning_rate": 3.4559999008417243e-06, + "loss": 0.2929, + "step": 27800 + }, + { + "epoch": 0.6119524543253356, + "grad_norm": 1.3904159931249696, + "learning_rate": 3.4556608694831493e-06, + "loss": 0.3213, + "step": 27801 + }, + { + "epoch": 0.6119744662117543, + "grad_norm": 1.6525172821202305, + "learning_rate": 3.4553218459735955e-06, + "loss": 0.279, + "step": 27802 + }, + { + "epoch": 0.611996478098173, + "grad_norm": 1.6976850252618643, + "learning_rate": 3.454982830314784e-06, + "loss": 0.3265, + "step": 27803 + }, + { + "epoch": 0.6120184899845916, + "grad_norm": 1.1420762614307867, + "learning_rate": 3.4546438225084407e-06, + "loss": 0.2756, + "step": 27804 + }, + { + "epoch": 0.6120405018710103, + "grad_norm": 0.6422114298814829, + "learning_rate": 3.454304822556286e-06, + "loss": 0.4768, + "step": 27805 + }, + { + "epoch": 0.612062513757429, + "grad_norm": 1.555461258326939, + "learning_rate": 3.453965830460042e-06, + "loss": 0.2265, + "step": 27806 + }, + { + "epoch": 0.6120845256438476, + "grad_norm": 3.1367384310067745, + "learning_rate": 3.4536268462214344e-06, + "loss": 0.3379, + "step": 27807 + }, + { + "epoch": 0.6121065375302663, + "grad_norm": 1.5935769900379444, + "learning_rate": 3.453287869842186e-06, + "loss": 0.2647, + "step": 27808 + }, + { + "epoch": 0.612128549416685, + "grad_norm": 1.520971932270461, + "learning_rate": 3.452948901324017e-06, + "loss": 0.3415, + "step": 27809 + }, + { + "epoch": 0.6121505613031036, + "grad_norm": 1.744519260501281, + "learning_rate": 3.452609940668652e-06, + "loss": 0.4698, + "step": 27810 + }, + { + "epoch": 0.6121725731895223, + "grad_norm": 1.7748222231248674, + "learning_rate": 3.4522709878778127e-06, + "loss": 0.2495, + "step": 27811 + }, + { + "epoch": 0.612194585075941, + "grad_norm": 1.5224052683909857, + "learning_rate": 3.4519320429532245e-06, + "loss": 0.2999, + "step": 27812 + }, + { + "epoch": 0.6122165969623596, + "grad_norm": 1.372236361219215, + "learning_rate": 3.451593105896607e-06, + "loss": 0.2745, + "step": 27813 + }, + { + "epoch": 0.6122386088487783, + "grad_norm": 1.4438854565625523, + "learning_rate": 3.451254176709683e-06, + "loss": 0.4403, + "step": 27814 + }, + { + "epoch": 0.612260620735197, + "grad_norm": 1.3245471911202986, + "learning_rate": 3.450915255394177e-06, + "loss": 0.2846, + "step": 27815 + }, + { + "epoch": 0.6122826326216156, + "grad_norm": 3.011249641604012, + "learning_rate": 3.4505763419518097e-06, + "loss": 0.4268, + "step": 27816 + }, + { + "epoch": 0.6123046445080343, + "grad_norm": 0.6633426889512171, + "learning_rate": 3.450237436384306e-06, + "loss": 0.4758, + "step": 27817 + }, + { + "epoch": 0.612326656394453, + "grad_norm": 2.357750506739243, + "learning_rate": 3.449898538693387e-06, + "loss": 0.3304, + "step": 27818 + }, + { + "epoch": 0.6123486682808716, + "grad_norm": 1.6354385146678545, + "learning_rate": 3.4495596488807725e-06, + "loss": 0.2945, + "step": 27819 + }, + { + "epoch": 0.6123706801672903, + "grad_norm": 2.0465564935528935, + "learning_rate": 3.449220766948189e-06, + "loss": 0.2853, + "step": 27820 + }, + { + "epoch": 0.612392692053709, + "grad_norm": 1.1342369817757592, + "learning_rate": 3.4488818928973565e-06, + "loss": 0.2253, + "step": 27821 + }, + { + "epoch": 0.6124147039401276, + "grad_norm": 1.8945124359170886, + "learning_rate": 3.44854302673e-06, + "loss": 0.3337, + "step": 27822 + }, + { + "epoch": 0.6124367158265464, + "grad_norm": 1.4273201192169134, + "learning_rate": 3.4482041684478396e-06, + "loss": 0.1962, + "step": 27823 + }, + { + "epoch": 0.6124587277129651, + "grad_norm": 1.605980993457377, + "learning_rate": 3.4478653180525955e-06, + "loss": 0.3056, + "step": 27824 + }, + { + "epoch": 0.6124807395993837, + "grad_norm": 1.3998395036394242, + "learning_rate": 3.4475264755459937e-06, + "loss": 0.2988, + "step": 27825 + }, + { + "epoch": 0.6125027514858024, + "grad_norm": 0.6336941332019127, + "learning_rate": 3.4471876409297557e-06, + "loss": 0.4628, + "step": 27826 + }, + { + "epoch": 0.612524763372221, + "grad_norm": 1.4531281659138786, + "learning_rate": 3.4468488142056e-06, + "loss": 0.291, + "step": 27827 + }, + { + "epoch": 0.6125467752586397, + "grad_norm": 1.3016924420320137, + "learning_rate": 3.4465099953752533e-06, + "loss": 0.3396, + "step": 27828 + }, + { + "epoch": 0.6125687871450584, + "grad_norm": 0.6181821992218767, + "learning_rate": 3.4461711844404345e-06, + "loss": 0.4516, + "step": 27829 + }, + { + "epoch": 0.612590799031477, + "grad_norm": 1.8566193681416665, + "learning_rate": 3.445832381402868e-06, + "loss": 0.3204, + "step": 27830 + }, + { + "epoch": 0.6126128109178957, + "grad_norm": 1.4527606540313303, + "learning_rate": 3.445493586264275e-06, + "loss": 0.3114, + "step": 27831 + }, + { + "epoch": 0.6126348228043144, + "grad_norm": 1.9259790962495862, + "learning_rate": 3.4451547990263746e-06, + "loss": 0.3363, + "step": 27832 + }, + { + "epoch": 0.612656834690733, + "grad_norm": 0.6381017032929622, + "learning_rate": 3.444816019690892e-06, + "loss": 0.4691, + "step": 27833 + }, + { + "epoch": 0.6126788465771517, + "grad_norm": 0.6436246302002909, + "learning_rate": 3.4444772482595477e-06, + "loss": 0.4888, + "step": 27834 + }, + { + "epoch": 0.6127008584635704, + "grad_norm": 0.6107354930500701, + "learning_rate": 3.4441384847340655e-06, + "loss": 0.4752, + "step": 27835 + }, + { + "epoch": 0.612722870349989, + "grad_norm": 0.6218900735758385, + "learning_rate": 3.4437997291161644e-06, + "loss": 0.4827, + "step": 27836 + }, + { + "epoch": 0.6127448822364077, + "grad_norm": 2.51154265259672, + "learning_rate": 3.443460981407566e-06, + "loss": 0.2524, + "step": 27837 + }, + { + "epoch": 0.6127668941228264, + "grad_norm": 3.1142094691014677, + "learning_rate": 3.443122241609995e-06, + "loss": 0.4095, + "step": 27838 + }, + { + "epoch": 0.612788906009245, + "grad_norm": 1.5646175234145263, + "learning_rate": 3.4427835097251696e-06, + "loss": 0.3108, + "step": 27839 + }, + { + "epoch": 0.6128109178956637, + "grad_norm": 0.6419614775260478, + "learning_rate": 3.4424447857548144e-06, + "loss": 0.4698, + "step": 27840 + }, + { + "epoch": 0.6128329297820824, + "grad_norm": 1.6461334334297812, + "learning_rate": 3.442106069700649e-06, + "loss": 0.3439, + "step": 27841 + }, + { + "epoch": 0.612854941668501, + "grad_norm": 1.6277004743531185, + "learning_rate": 3.4417673615643945e-06, + "loss": 0.3651, + "step": 27842 + }, + { + "epoch": 0.6128769535549197, + "grad_norm": 1.7971714708957707, + "learning_rate": 3.4414286613477755e-06, + "loss": 0.3583, + "step": 27843 + }, + { + "epoch": 0.6128989654413384, + "grad_norm": 1.7918754078566903, + "learning_rate": 3.4410899690525066e-06, + "loss": 0.2631, + "step": 27844 + }, + { + "epoch": 0.612920977327757, + "grad_norm": 3.1232950394377843, + "learning_rate": 3.4407512846803187e-06, + "loss": 0.356, + "step": 27845 + }, + { + "epoch": 0.6129429892141757, + "grad_norm": 0.601898456668453, + "learning_rate": 3.4404126082329266e-06, + "loss": 0.4357, + "step": 27846 + }, + { + "epoch": 0.6129650011005944, + "grad_norm": 6.119230403013716, + "learning_rate": 3.440073939712052e-06, + "loss": 0.366, + "step": 27847 + }, + { + "epoch": 0.612987012987013, + "grad_norm": 1.6192200080461516, + "learning_rate": 3.4397352791194194e-06, + "loss": 0.261, + "step": 27848 + }, + { + "epoch": 0.6130090248734317, + "grad_norm": 0.6121845389044678, + "learning_rate": 3.4393966264567467e-06, + "loss": 0.4704, + "step": 27849 + }, + { + "epoch": 0.6130310367598503, + "grad_norm": 0.6001691469605291, + "learning_rate": 3.4390579817257562e-06, + "loss": 0.4666, + "step": 27850 + }, + { + "epoch": 0.613053048646269, + "grad_norm": 1.6929314906342314, + "learning_rate": 3.43871934492817e-06, + "loss": 0.3198, + "step": 27851 + }, + { + "epoch": 0.6130750605326877, + "grad_norm": 0.6141769887304624, + "learning_rate": 3.438380716065707e-06, + "loss": 0.434, + "step": 27852 + }, + { + "epoch": 0.6130970724191063, + "grad_norm": 1.4857077508648497, + "learning_rate": 3.4380420951400915e-06, + "loss": 0.3787, + "step": 27853 + }, + { + "epoch": 0.613119084305525, + "grad_norm": 1.872599488196975, + "learning_rate": 3.437703482153043e-06, + "loss": 0.3186, + "step": 27854 + }, + { + "epoch": 0.6131410961919437, + "grad_norm": 1.5786449812494194, + "learning_rate": 3.43736487710628e-06, + "loss": 0.3177, + "step": 27855 + }, + { + "epoch": 0.6131631080783623, + "grad_norm": 0.6242152251921341, + "learning_rate": 3.4370262800015275e-06, + "loss": 0.4545, + "step": 27856 + }, + { + "epoch": 0.613185119964781, + "grad_norm": 1.2769915915400054, + "learning_rate": 3.436687690840503e-06, + "loss": 0.3616, + "step": 27857 + }, + { + "epoch": 0.6132071318511997, + "grad_norm": 1.5417134314700427, + "learning_rate": 3.436349109624931e-06, + "loss": 0.3086, + "step": 27858 + }, + { + "epoch": 0.6132291437376183, + "grad_norm": 0.6183704116544118, + "learning_rate": 3.4360105363565288e-06, + "loss": 0.4766, + "step": 27859 + }, + { + "epoch": 0.613251155624037, + "grad_norm": 1.7272827037213199, + "learning_rate": 3.4356719710370184e-06, + "loss": 0.2786, + "step": 27860 + }, + { + "epoch": 0.6132731675104557, + "grad_norm": 0.6155707711302976, + "learning_rate": 3.4353334136681226e-06, + "loss": 0.4751, + "step": 27861 + }, + { + "epoch": 0.6132951793968743, + "grad_norm": 1.7932368605218325, + "learning_rate": 3.4349948642515567e-06, + "loss": 0.3123, + "step": 27862 + }, + { + "epoch": 0.613317191283293, + "grad_norm": 1.3660968186570324, + "learning_rate": 3.4346563227890485e-06, + "loss": 0.2625, + "step": 27863 + }, + { + "epoch": 0.6133392031697117, + "grad_norm": 1.6684048582913062, + "learning_rate": 3.434317789282314e-06, + "loss": 0.367, + "step": 27864 + }, + { + "epoch": 0.6133612150561303, + "grad_norm": 1.8363980506403974, + "learning_rate": 3.433979263733073e-06, + "loss": 0.252, + "step": 27865 + }, + { + "epoch": 0.613383226942549, + "grad_norm": 1.4639503538752285, + "learning_rate": 3.433640746143051e-06, + "loss": 0.4075, + "step": 27866 + }, + { + "epoch": 0.6134052388289677, + "grad_norm": 0.6239065913362837, + "learning_rate": 3.4333022365139634e-06, + "loss": 0.4616, + "step": 27867 + }, + { + "epoch": 0.6134272507153863, + "grad_norm": 1.4728211669193791, + "learning_rate": 3.432963734847532e-06, + "loss": 0.4403, + "step": 27868 + }, + { + "epoch": 0.613449262601805, + "grad_norm": 4.2983006075336325, + "learning_rate": 3.432625241145479e-06, + "loss": 0.3019, + "step": 27869 + }, + { + "epoch": 0.6134712744882237, + "grad_norm": 0.6260117096054897, + "learning_rate": 3.4322867554095223e-06, + "loss": 0.4534, + "step": 27870 + }, + { + "epoch": 0.6134932863746423, + "grad_norm": 1.5702047077267725, + "learning_rate": 3.431948277641386e-06, + "loss": 0.261, + "step": 27871 + }, + { + "epoch": 0.613515298261061, + "grad_norm": 1.6813023802921425, + "learning_rate": 3.4316098078427863e-06, + "loss": 0.3854, + "step": 27872 + }, + { + "epoch": 0.6135373101474796, + "grad_norm": 1.399211571936977, + "learning_rate": 3.4312713460154445e-06, + "loss": 0.2709, + "step": 27873 + }, + { + "epoch": 0.6135593220338983, + "grad_norm": 1.506132444765815, + "learning_rate": 3.430932892161082e-06, + "loss": 0.2875, + "step": 27874 + }, + { + "epoch": 0.613581333920317, + "grad_norm": 1.5307939839154698, + "learning_rate": 3.430594446281418e-06, + "loss": 0.2896, + "step": 27875 + }, + { + "epoch": 0.6136033458067356, + "grad_norm": 1.9168131033767226, + "learning_rate": 3.4302560083781745e-06, + "loss": 0.3203, + "step": 27876 + }, + { + "epoch": 0.6136253576931543, + "grad_norm": 1.211119474661684, + "learning_rate": 3.429917578453069e-06, + "loss": 0.3149, + "step": 27877 + }, + { + "epoch": 0.613647369579573, + "grad_norm": 0.6133194898812562, + "learning_rate": 3.4295791565078216e-06, + "loss": 0.4698, + "step": 27878 + }, + { + "epoch": 0.6136693814659916, + "grad_norm": 1.3300951777585204, + "learning_rate": 3.429240742544156e-06, + "loss": 0.2886, + "step": 27879 + }, + { + "epoch": 0.6136913933524103, + "grad_norm": 1.4255128009734521, + "learning_rate": 3.428902336563786e-06, + "loss": 0.307, + "step": 27880 + }, + { + "epoch": 0.613713405238829, + "grad_norm": 1.727736575001657, + "learning_rate": 3.4285639385684382e-06, + "loss": 0.3169, + "step": 27881 + }, + { + "epoch": 0.6137354171252476, + "grad_norm": 1.2184552121101848, + "learning_rate": 3.428225548559828e-06, + "loss": 0.2695, + "step": 27882 + }, + { + "epoch": 0.6137574290116663, + "grad_norm": 1.5571284759119486, + "learning_rate": 3.427887166539676e-06, + "loss": 0.3125, + "step": 27883 + }, + { + "epoch": 0.613779440898085, + "grad_norm": 1.4222565894524717, + "learning_rate": 3.4275487925097052e-06, + "loss": 0.3494, + "step": 27884 + }, + { + "epoch": 0.6138014527845036, + "grad_norm": 0.6198608189330052, + "learning_rate": 3.4272104264716287e-06, + "loss": 0.4608, + "step": 27885 + }, + { + "epoch": 0.6138234646709223, + "grad_norm": 1.5155934866861969, + "learning_rate": 3.426872068427174e-06, + "loss": 0.2955, + "step": 27886 + }, + { + "epoch": 0.613845476557341, + "grad_norm": 1.5509453891443148, + "learning_rate": 3.4265337183780556e-06, + "loss": 0.2881, + "step": 27887 + }, + { + "epoch": 0.6138674884437596, + "grad_norm": 1.672894607270994, + "learning_rate": 3.426195376325994e-06, + "loss": 0.3039, + "step": 27888 + }, + { + "epoch": 0.6138895003301783, + "grad_norm": 1.5901740383841716, + "learning_rate": 3.425857042272711e-06, + "loss": 0.2746, + "step": 27889 + }, + { + "epoch": 0.613911512216597, + "grad_norm": 1.7748017807099983, + "learning_rate": 3.4255187162199234e-06, + "loss": 0.2409, + "step": 27890 + }, + { + "epoch": 0.6139335241030156, + "grad_norm": 0.592261884856836, + "learning_rate": 3.4251803981693516e-06, + "loss": 0.436, + "step": 27891 + }, + { + "epoch": 0.6139555359894343, + "grad_norm": 1.1478642394160439, + "learning_rate": 3.424842088122716e-06, + "loss": 0.2816, + "step": 27892 + }, + { + "epoch": 0.613977547875853, + "grad_norm": 1.9344143717211906, + "learning_rate": 3.4245037860817344e-06, + "loss": 0.3008, + "step": 27893 + }, + { + "epoch": 0.6139995597622716, + "grad_norm": 2.4810685184500545, + "learning_rate": 3.424165492048129e-06, + "loss": 0.3177, + "step": 27894 + }, + { + "epoch": 0.6140215716486903, + "grad_norm": 1.5754474520334958, + "learning_rate": 3.4238272060236178e-06, + "loss": 0.2211, + "step": 27895 + }, + { + "epoch": 0.6140435835351089, + "grad_norm": 1.7338058708848587, + "learning_rate": 3.423488928009917e-06, + "loss": 0.2141, + "step": 27896 + }, + { + "epoch": 0.6140655954215276, + "grad_norm": 1.0975111620055624, + "learning_rate": 3.423150658008752e-06, + "loss": 0.2321, + "step": 27897 + }, + { + "epoch": 0.6140876073079463, + "grad_norm": 1.4666463905443252, + "learning_rate": 3.422812396021835e-06, + "loss": 0.3359, + "step": 27898 + }, + { + "epoch": 0.6141096191943649, + "grad_norm": 1.3452117360783735, + "learning_rate": 3.422474142050892e-06, + "loss": 0.2826, + "step": 27899 + }, + { + "epoch": 0.6141316310807836, + "grad_norm": 1.5865512099197543, + "learning_rate": 3.422135896097638e-06, + "loss": 0.2188, + "step": 27900 + }, + { + "epoch": 0.6141536429672023, + "grad_norm": 1.5587899867289032, + "learning_rate": 3.421797658163792e-06, + "loss": 0.265, + "step": 27901 + }, + { + "epoch": 0.6141756548536209, + "grad_norm": 1.279642391462603, + "learning_rate": 3.4214594282510762e-06, + "loss": 0.337, + "step": 27902 + }, + { + "epoch": 0.6141976667400396, + "grad_norm": 2.0496411861733024, + "learning_rate": 3.4211212063612043e-06, + "loss": 0.2997, + "step": 27903 + }, + { + "epoch": 0.6142196786264583, + "grad_norm": 1.4212088492801966, + "learning_rate": 3.4207829924959023e-06, + "loss": 0.3177, + "step": 27904 + }, + { + "epoch": 0.6142416905128769, + "grad_norm": 2.42562015323615, + "learning_rate": 3.420444786656884e-06, + "loss": 0.3694, + "step": 27905 + }, + { + "epoch": 0.6142637023992956, + "grad_norm": 1.6156334522926363, + "learning_rate": 3.420106588845869e-06, + "loss": 0.3248, + "step": 27906 + }, + { + "epoch": 0.6142857142857143, + "grad_norm": 1.5371627504809398, + "learning_rate": 3.419768399064579e-06, + "loss": 0.2946, + "step": 27907 + }, + { + "epoch": 0.6143077261721329, + "grad_norm": 1.3051253963327993, + "learning_rate": 3.41943021731473e-06, + "loss": 0.3191, + "step": 27908 + }, + { + "epoch": 0.6143297380585516, + "grad_norm": 1.672311230953107, + "learning_rate": 3.41909204359804e-06, + "loss": 0.2601, + "step": 27909 + }, + { + "epoch": 0.6143517499449703, + "grad_norm": 2.8160184500921512, + "learning_rate": 3.418753877916231e-06, + "loss": 0.3601, + "step": 27910 + }, + { + "epoch": 0.6143737618313889, + "grad_norm": 1.4424140692207081, + "learning_rate": 3.4184157202710184e-06, + "loss": 0.3306, + "step": 27911 + }, + { + "epoch": 0.6143957737178076, + "grad_norm": 0.6449783729318905, + "learning_rate": 3.4180775706641243e-06, + "loss": 0.4783, + "step": 27912 + }, + { + "epoch": 0.6144177856042263, + "grad_norm": 2.6295934476499605, + "learning_rate": 3.4177394290972653e-06, + "loss": 0.3019, + "step": 27913 + }, + { + "epoch": 0.6144397974906449, + "grad_norm": 1.5447623107866972, + "learning_rate": 3.417401295572159e-06, + "loss": 0.323, + "step": 27914 + }, + { + "epoch": 0.6144618093770636, + "grad_norm": 1.414473664464587, + "learning_rate": 3.417063170090528e-06, + "loss": 0.2917, + "step": 27915 + }, + { + "epoch": 0.6144838212634823, + "grad_norm": 1.4289531722153488, + "learning_rate": 3.4167250526540836e-06, + "loss": 0.3858, + "step": 27916 + }, + { + "epoch": 0.6145058331499009, + "grad_norm": 0.5935675584274361, + "learning_rate": 3.4163869432645523e-06, + "loss": 0.4799, + "step": 27917 + }, + { + "epoch": 0.6145278450363196, + "grad_norm": 1.5963609474730192, + "learning_rate": 3.4160488419236475e-06, + "loss": 0.3885, + "step": 27918 + }, + { + "epoch": 0.6145498569227382, + "grad_norm": 1.058849133653278, + "learning_rate": 3.4157107486330887e-06, + "loss": 0.2538, + "step": 27919 + }, + { + "epoch": 0.6145718688091569, + "grad_norm": 1.4059211634968916, + "learning_rate": 3.4153726633945962e-06, + "loss": 0.3151, + "step": 27920 + }, + { + "epoch": 0.6145938806955756, + "grad_norm": 1.4379375151628595, + "learning_rate": 3.4150345862098827e-06, + "loss": 0.3356, + "step": 27921 + }, + { + "epoch": 0.6146158925819942, + "grad_norm": 0.6105579054219914, + "learning_rate": 3.4146965170806745e-06, + "loss": 0.4393, + "step": 27922 + }, + { + "epoch": 0.6146379044684129, + "grad_norm": 2.0631068054566324, + "learning_rate": 3.414358456008684e-06, + "loss": 0.3491, + "step": 27923 + }, + { + "epoch": 0.6146599163548316, + "grad_norm": 1.384582483161311, + "learning_rate": 3.4140204029956304e-06, + "loss": 0.2666, + "step": 27924 + }, + { + "epoch": 0.6146819282412502, + "grad_norm": 1.765622624169704, + "learning_rate": 3.4136823580432345e-06, + "loss": 0.2401, + "step": 27925 + }, + { + "epoch": 0.6147039401276689, + "grad_norm": 0.6089610991949126, + "learning_rate": 3.4133443211532106e-06, + "loss": 0.4159, + "step": 27926 + }, + { + "epoch": 0.6147259520140876, + "grad_norm": 1.4340492295370995, + "learning_rate": 3.4130062923272784e-06, + "loss": 0.2747, + "step": 27927 + }, + { + "epoch": 0.6147479639005062, + "grad_norm": 1.4148443801396353, + "learning_rate": 3.4126682715671565e-06, + "loss": 0.3086, + "step": 27928 + }, + { + "epoch": 0.6147699757869249, + "grad_norm": 1.971061819368829, + "learning_rate": 3.412330258874562e-06, + "loss": 0.2445, + "step": 27929 + }, + { + "epoch": 0.6147919876733436, + "grad_norm": 1.452903330897249, + "learning_rate": 3.4119922542512145e-06, + "loss": 0.3049, + "step": 27930 + }, + { + "epoch": 0.6148139995597622, + "grad_norm": 1.6745759544099799, + "learning_rate": 3.4116542576988304e-06, + "loss": 0.339, + "step": 27931 + }, + { + "epoch": 0.6148360114461809, + "grad_norm": 2.0765370315605187, + "learning_rate": 3.411316269219126e-06, + "loss": 0.2624, + "step": 27932 + }, + { + "epoch": 0.6148580233325996, + "grad_norm": 1.4630318774363298, + "learning_rate": 3.410978288813824e-06, + "loss": 0.2739, + "step": 27933 + }, + { + "epoch": 0.6148800352190182, + "grad_norm": 1.262699540976746, + "learning_rate": 3.4106403164846347e-06, + "loss": 0.3407, + "step": 27934 + }, + { + "epoch": 0.6149020471054369, + "grad_norm": 1.3753353011317175, + "learning_rate": 3.4103023522332846e-06, + "loss": 0.2958, + "step": 27935 + }, + { + "epoch": 0.6149240589918556, + "grad_norm": 1.5586307413966694, + "learning_rate": 3.409964396061485e-06, + "loss": 0.3327, + "step": 27936 + }, + { + "epoch": 0.6149460708782742, + "grad_norm": 1.792024687730408, + "learning_rate": 3.4096264479709546e-06, + "loss": 0.3274, + "step": 27937 + }, + { + "epoch": 0.614968082764693, + "grad_norm": 1.170570068058244, + "learning_rate": 3.409288507963414e-06, + "loss": 0.2434, + "step": 27938 + }, + { + "epoch": 0.6149900946511117, + "grad_norm": 1.55449900314735, + "learning_rate": 3.408950576040576e-06, + "loss": 0.2517, + "step": 27939 + }, + { + "epoch": 0.6150121065375302, + "grad_norm": 1.21299436250475, + "learning_rate": 3.4086126522041633e-06, + "loss": 0.2955, + "step": 27940 + }, + { + "epoch": 0.615034118423949, + "grad_norm": 1.4304383968515613, + "learning_rate": 3.40827473645589e-06, + "loss": 0.244, + "step": 27941 + }, + { + "epoch": 0.6150561303103675, + "grad_norm": 0.6151409914883691, + "learning_rate": 3.4079368287974733e-06, + "loss": 0.4614, + "step": 27942 + }, + { + "epoch": 0.6150781421967862, + "grad_norm": 1.5954204584180467, + "learning_rate": 3.4075989292306344e-06, + "loss": 0.3236, + "step": 27943 + }, + { + "epoch": 0.615100154083205, + "grad_norm": 1.5916152882218433, + "learning_rate": 3.4072610377570857e-06, + "loss": 0.3498, + "step": 27944 + }, + { + "epoch": 0.6151221659696235, + "grad_norm": 1.5340373877200848, + "learning_rate": 3.4069231543785474e-06, + "loss": 0.2482, + "step": 27945 + }, + { + "epoch": 0.6151441778560423, + "grad_norm": 1.5581698316343051, + "learning_rate": 3.406585279096737e-06, + "loss": 0.3383, + "step": 27946 + }, + { + "epoch": 0.615166189742461, + "grad_norm": 1.9953690823932502, + "learning_rate": 3.4062474119133686e-06, + "loss": 0.3101, + "step": 27947 + }, + { + "epoch": 0.6151882016288795, + "grad_norm": 1.2805315191329385, + "learning_rate": 3.4059095528301645e-06, + "loss": 0.3275, + "step": 27948 + }, + { + "epoch": 0.6152102135152983, + "grad_norm": 1.4800599701927826, + "learning_rate": 3.405571701848838e-06, + "loss": 0.3308, + "step": 27949 + }, + { + "epoch": 0.615232225401717, + "grad_norm": 1.3984215962131534, + "learning_rate": 3.4052338589711066e-06, + "loss": 0.291, + "step": 27950 + }, + { + "epoch": 0.6152542372881356, + "grad_norm": 0.6231948512497976, + "learning_rate": 3.4048960241986896e-06, + "loss": 0.4679, + "step": 27951 + }, + { + "epoch": 0.6152762491745543, + "grad_norm": 0.6335570693308936, + "learning_rate": 3.4045581975332986e-06, + "loss": 0.4794, + "step": 27952 + }, + { + "epoch": 0.615298261060973, + "grad_norm": 1.6071317554507518, + "learning_rate": 3.4042203789766586e-06, + "loss": 0.403, + "step": 27953 + }, + { + "epoch": 0.6153202729473916, + "grad_norm": 1.418726910729419, + "learning_rate": 3.403882568530481e-06, + "loss": 0.3405, + "step": 27954 + }, + { + "epoch": 0.6153422848338103, + "grad_norm": 1.453613747773336, + "learning_rate": 3.4035447661964833e-06, + "loss": 0.3382, + "step": 27955 + }, + { + "epoch": 0.615364296720229, + "grad_norm": 1.4058214090520487, + "learning_rate": 3.403206971976385e-06, + "loss": 0.3347, + "step": 27956 + }, + { + "epoch": 0.6153863086066476, + "grad_norm": 1.3541930031629745, + "learning_rate": 3.4028691858718976e-06, + "loss": 0.2608, + "step": 27957 + }, + { + "epoch": 0.6154083204930663, + "grad_norm": 1.6111096740108888, + "learning_rate": 3.4025314078847447e-06, + "loss": 0.3017, + "step": 27958 + }, + { + "epoch": 0.615430332379485, + "grad_norm": 0.6319438566302726, + "learning_rate": 3.402193638016639e-06, + "loss": 0.4972, + "step": 27959 + }, + { + "epoch": 0.6154523442659036, + "grad_norm": 1.3367596913994353, + "learning_rate": 3.4018558762692968e-06, + "loss": 0.2243, + "step": 27960 + }, + { + "epoch": 0.6154743561523223, + "grad_norm": 0.6211195708274789, + "learning_rate": 3.4015181226444373e-06, + "loss": 0.4523, + "step": 27961 + }, + { + "epoch": 0.615496368038741, + "grad_norm": 1.488023620971957, + "learning_rate": 3.401180377143774e-06, + "loss": 0.2806, + "step": 27962 + }, + { + "epoch": 0.6155183799251596, + "grad_norm": 1.5233669704261001, + "learning_rate": 3.4008426397690266e-06, + "loss": 0.2808, + "step": 27963 + }, + { + "epoch": 0.6155403918115783, + "grad_norm": 1.377242658664503, + "learning_rate": 3.400504910521909e-06, + "loss": 0.32, + "step": 27964 + }, + { + "epoch": 0.615562403697997, + "grad_norm": 1.4748408629082082, + "learning_rate": 3.400167189404139e-06, + "loss": 0.3057, + "step": 27965 + }, + { + "epoch": 0.6155844155844156, + "grad_norm": 1.419435804953465, + "learning_rate": 3.3998294764174337e-06, + "loss": 0.3942, + "step": 27966 + }, + { + "epoch": 0.6156064274708343, + "grad_norm": 1.6017482698463537, + "learning_rate": 3.3994917715635083e-06, + "loss": 0.3361, + "step": 27967 + }, + { + "epoch": 0.6156284393572529, + "grad_norm": 0.620811890682592, + "learning_rate": 3.399154074844079e-06, + "loss": 0.4527, + "step": 27968 + }, + { + "epoch": 0.6156504512436716, + "grad_norm": 1.4869520238575444, + "learning_rate": 3.398816386260864e-06, + "loss": 0.3071, + "step": 27969 + }, + { + "epoch": 0.6156724631300903, + "grad_norm": 1.410888025636092, + "learning_rate": 3.3984787058155753e-06, + "loss": 0.2596, + "step": 27970 + }, + { + "epoch": 0.6156944750165089, + "grad_norm": 0.614574519741756, + "learning_rate": 3.3981410335099356e-06, + "loss": 0.466, + "step": 27971 + }, + { + "epoch": 0.6157164869029276, + "grad_norm": 1.7875978799578478, + "learning_rate": 3.3978033693456565e-06, + "loss": 0.3637, + "step": 27972 + }, + { + "epoch": 0.6157384987893463, + "grad_norm": 1.3051462962010314, + "learning_rate": 3.3974657133244535e-06, + "loss": 0.324, + "step": 27973 + }, + { + "epoch": 0.6157605106757649, + "grad_norm": 1.3208928908224304, + "learning_rate": 3.3971280654480475e-06, + "loss": 0.3328, + "step": 27974 + }, + { + "epoch": 0.6157825225621836, + "grad_norm": 1.4002602227422822, + "learning_rate": 3.3967904257181494e-06, + "loss": 0.2443, + "step": 27975 + }, + { + "epoch": 0.6158045344486023, + "grad_norm": 1.3380277894140857, + "learning_rate": 3.3964527941364783e-06, + "loss": 0.3209, + "step": 27976 + }, + { + "epoch": 0.6158265463350209, + "grad_norm": 1.3881232155552083, + "learning_rate": 3.396115170704749e-06, + "loss": 0.3124, + "step": 27977 + }, + { + "epoch": 0.6158485582214396, + "grad_norm": 1.6812305924759174, + "learning_rate": 3.3957775554246775e-06, + "loss": 0.3001, + "step": 27978 + }, + { + "epoch": 0.6158705701078583, + "grad_norm": 1.4786592438455102, + "learning_rate": 3.3954399482979817e-06, + "loss": 0.2907, + "step": 27979 + }, + { + "epoch": 0.6158925819942769, + "grad_norm": 1.3455416971293874, + "learning_rate": 3.3951023493263735e-06, + "loss": 0.3152, + "step": 27980 + }, + { + "epoch": 0.6159145938806956, + "grad_norm": 1.3743838269400483, + "learning_rate": 3.394764758511572e-06, + "loss": 0.2709, + "step": 27981 + }, + { + "epoch": 0.6159366057671143, + "grad_norm": 1.9774989595063184, + "learning_rate": 3.3944271758552916e-06, + "loss": 0.312, + "step": 27982 + }, + { + "epoch": 0.6159586176535329, + "grad_norm": 1.486070620267651, + "learning_rate": 3.394089601359247e-06, + "loss": 0.2908, + "step": 27983 + }, + { + "epoch": 0.6159806295399516, + "grad_norm": 1.3913387916548152, + "learning_rate": 3.3937520350251583e-06, + "loss": 0.3511, + "step": 27984 + }, + { + "epoch": 0.6160026414263703, + "grad_norm": 1.1679353636750798, + "learning_rate": 3.3934144768547363e-06, + "loss": 0.2454, + "step": 27985 + }, + { + "epoch": 0.6160246533127889, + "grad_norm": 1.42107012375649, + "learning_rate": 3.3930769268496987e-06, + "loss": 0.3871, + "step": 27986 + }, + { + "epoch": 0.6160466651992076, + "grad_norm": 0.6665454889481455, + "learning_rate": 3.392739385011763e-06, + "loss": 0.4541, + "step": 27987 + }, + { + "epoch": 0.6160686770856263, + "grad_norm": 1.3679265510904377, + "learning_rate": 3.3924018513426383e-06, + "loss": 0.3126, + "step": 27988 + }, + { + "epoch": 0.6160906889720449, + "grad_norm": 1.6876937766210027, + "learning_rate": 3.3920643258440478e-06, + "loss": 0.3234, + "step": 27989 + }, + { + "epoch": 0.6161127008584636, + "grad_norm": 1.396172354737234, + "learning_rate": 3.391726808517703e-06, + "loss": 0.2798, + "step": 27990 + }, + { + "epoch": 0.6161347127448822, + "grad_norm": 0.6059168639108597, + "learning_rate": 3.3913892993653186e-06, + "loss": 0.4388, + "step": 27991 + }, + { + "epoch": 0.6161567246313009, + "grad_norm": 1.3080521373418796, + "learning_rate": 3.391051798388613e-06, + "loss": 0.3062, + "step": 27992 + }, + { + "epoch": 0.6161787365177196, + "grad_norm": 2.176009340804374, + "learning_rate": 3.390714305589298e-06, + "loss": 0.3984, + "step": 27993 + }, + { + "epoch": 0.6162007484041382, + "grad_norm": 1.4638218458856576, + "learning_rate": 3.3903768209690916e-06, + "loss": 0.2631, + "step": 27994 + }, + { + "epoch": 0.6162227602905569, + "grad_norm": 0.6336619836110862, + "learning_rate": 3.3900393445297087e-06, + "loss": 0.4621, + "step": 27995 + }, + { + "epoch": 0.6162447721769756, + "grad_norm": 0.6098409369640659, + "learning_rate": 3.389701876272862e-06, + "loss": 0.4455, + "step": 27996 + }, + { + "epoch": 0.6162667840633942, + "grad_norm": 1.4403375298974428, + "learning_rate": 3.389364416200271e-06, + "loss": 0.2724, + "step": 27997 + }, + { + "epoch": 0.6162887959498129, + "grad_norm": 1.6480518301535927, + "learning_rate": 3.3890269643136466e-06, + "loss": 0.2874, + "step": 27998 + }, + { + "epoch": 0.6163108078362316, + "grad_norm": 1.5602715727999525, + "learning_rate": 3.388689520614706e-06, + "loss": 0.4032, + "step": 27999 + }, + { + "epoch": 0.6163328197226502, + "grad_norm": 1.3614384106924617, + "learning_rate": 3.3883520851051644e-06, + "loss": 0.2883, + "step": 28000 + }, + { + "epoch": 0.6163548316090689, + "grad_norm": 1.6734327381453067, + "learning_rate": 3.388014657786735e-06, + "loss": 0.2941, + "step": 28001 + }, + { + "epoch": 0.6163768434954876, + "grad_norm": 1.3654937511391732, + "learning_rate": 3.3876772386611367e-06, + "loss": 0.3368, + "step": 28002 + }, + { + "epoch": 0.6163988553819062, + "grad_norm": 1.372508412075109, + "learning_rate": 3.3873398277300796e-06, + "loss": 0.3966, + "step": 28003 + }, + { + "epoch": 0.6164208672683249, + "grad_norm": 2.5553667027946125, + "learning_rate": 3.3870024249952813e-06, + "loss": 0.2282, + "step": 28004 + }, + { + "epoch": 0.6164428791547436, + "grad_norm": 1.4198272404745167, + "learning_rate": 3.3866650304584575e-06, + "loss": 0.3108, + "step": 28005 + }, + { + "epoch": 0.6164648910411622, + "grad_norm": 1.2825566777440647, + "learning_rate": 3.3863276441213184e-06, + "loss": 0.2973, + "step": 28006 + }, + { + "epoch": 0.6164869029275809, + "grad_norm": 2.7961353141552725, + "learning_rate": 3.3859902659855847e-06, + "loss": 0.3604, + "step": 28007 + }, + { + "epoch": 0.6165089148139996, + "grad_norm": 1.473086638687766, + "learning_rate": 3.3856528960529678e-06, + "loss": 0.2447, + "step": 28008 + }, + { + "epoch": 0.6165309267004182, + "grad_norm": 1.6193768275150364, + "learning_rate": 3.385315534325182e-06, + "loss": 0.3413, + "step": 28009 + }, + { + "epoch": 0.6165529385868369, + "grad_norm": 1.4582149011365875, + "learning_rate": 3.3849781808039446e-06, + "loss": 0.268, + "step": 28010 + }, + { + "epoch": 0.6165749504732556, + "grad_norm": 0.6317369034807627, + "learning_rate": 3.384640835490966e-06, + "loss": 0.4648, + "step": 28011 + }, + { + "epoch": 0.6165969623596742, + "grad_norm": 1.6094399264695476, + "learning_rate": 3.3843034983879645e-06, + "loss": 0.3923, + "step": 28012 + }, + { + "epoch": 0.6166189742460929, + "grad_norm": 1.2725209236415211, + "learning_rate": 3.3839661694966532e-06, + "loss": 0.2838, + "step": 28013 + }, + { + "epoch": 0.6166409861325115, + "grad_norm": 1.4360547785482225, + "learning_rate": 3.3836288488187456e-06, + "loss": 0.2702, + "step": 28014 + }, + { + "epoch": 0.6166629980189302, + "grad_norm": 1.747091725655778, + "learning_rate": 3.3832915363559593e-06, + "loss": 0.2981, + "step": 28015 + }, + { + "epoch": 0.6166850099053489, + "grad_norm": 1.508113039031465, + "learning_rate": 3.3829542321100046e-06, + "loss": 0.2623, + "step": 28016 + }, + { + "epoch": 0.6167070217917675, + "grad_norm": 1.9723522349499603, + "learning_rate": 3.3826169360825985e-06, + "loss": 0.314, + "step": 28017 + }, + { + "epoch": 0.6167290336781862, + "grad_norm": 2.262978243530971, + "learning_rate": 3.382279648275454e-06, + "loss": 0.273, + "step": 28018 + }, + { + "epoch": 0.6167510455646049, + "grad_norm": 3.6336004049154242, + "learning_rate": 3.3819423686902854e-06, + "loss": 0.2918, + "step": 28019 + }, + { + "epoch": 0.6167730574510235, + "grad_norm": 1.4827109986754374, + "learning_rate": 3.3816050973288094e-06, + "loss": 0.2645, + "step": 28020 + }, + { + "epoch": 0.6167950693374422, + "grad_norm": 1.479092636299859, + "learning_rate": 3.381267834192736e-06, + "loss": 0.3368, + "step": 28021 + }, + { + "epoch": 0.6168170812238609, + "grad_norm": 0.6228107586173247, + "learning_rate": 3.380930579283782e-06, + "loss": 0.4469, + "step": 28022 + }, + { + "epoch": 0.6168390931102795, + "grad_norm": 1.65924210546918, + "learning_rate": 3.380593332603663e-06, + "loss": 0.3277, + "step": 28023 + }, + { + "epoch": 0.6168611049966982, + "grad_norm": 0.5977299470416977, + "learning_rate": 3.3802560941540874e-06, + "loss": 0.4526, + "step": 28024 + }, + { + "epoch": 0.6168831168831169, + "grad_norm": 1.5726133099936166, + "learning_rate": 3.379918863936775e-06, + "loss": 0.3367, + "step": 28025 + }, + { + "epoch": 0.6169051287695355, + "grad_norm": 1.2106639679755038, + "learning_rate": 3.3795816419534377e-06, + "loss": 0.323, + "step": 28026 + }, + { + "epoch": 0.6169271406559542, + "grad_norm": 1.5818886827126735, + "learning_rate": 3.3792444282057875e-06, + "loss": 0.2728, + "step": 28027 + }, + { + "epoch": 0.6169491525423729, + "grad_norm": 1.3966515512538122, + "learning_rate": 3.378907222695542e-06, + "loss": 0.2855, + "step": 28028 + }, + { + "epoch": 0.6169711644287915, + "grad_norm": 1.353248476123373, + "learning_rate": 3.3785700254244115e-06, + "loss": 0.2896, + "step": 28029 + }, + { + "epoch": 0.6169931763152102, + "grad_norm": 1.8309561300089423, + "learning_rate": 3.378232836394112e-06, + "loss": 0.3106, + "step": 28030 + }, + { + "epoch": 0.6170151882016289, + "grad_norm": 0.6152221851412519, + "learning_rate": 3.377895655606357e-06, + "loss": 0.4574, + "step": 28031 + }, + { + "epoch": 0.6170372000880475, + "grad_norm": 1.5086756610859984, + "learning_rate": 3.377558483062858e-06, + "loss": 0.3427, + "step": 28032 + }, + { + "epoch": 0.6170592119744662, + "grad_norm": 1.4476466661757175, + "learning_rate": 3.3772213187653325e-06, + "loss": 0.2393, + "step": 28033 + }, + { + "epoch": 0.6170812238608849, + "grad_norm": 1.3856763940135028, + "learning_rate": 3.37688416271549e-06, + "loss": 0.271, + "step": 28034 + }, + { + "epoch": 0.6171032357473035, + "grad_norm": 2.3100689797567933, + "learning_rate": 3.3765470149150483e-06, + "loss": 0.3446, + "step": 28035 + }, + { + "epoch": 0.6171252476337222, + "grad_norm": 1.7460472649629983, + "learning_rate": 3.3762098753657174e-06, + "loss": 0.2933, + "step": 28036 + }, + { + "epoch": 0.6171472595201408, + "grad_norm": 1.117073616032393, + "learning_rate": 3.375872744069212e-06, + "loss": 0.1856, + "step": 28037 + }, + { + "epoch": 0.6171692714065595, + "grad_norm": 1.5184787110613855, + "learning_rate": 3.375535621027247e-06, + "loss": 0.3369, + "step": 28038 + }, + { + "epoch": 0.6171912832929782, + "grad_norm": 2.287833687956236, + "learning_rate": 3.3751985062415325e-06, + "loss": 0.2315, + "step": 28039 + }, + { + "epoch": 0.6172132951793968, + "grad_norm": 2.5213604189282783, + "learning_rate": 3.3748613997137853e-06, + "loss": 0.2692, + "step": 28040 + }, + { + "epoch": 0.6172353070658155, + "grad_norm": 1.7178211687851186, + "learning_rate": 3.3745243014457185e-06, + "loss": 0.2778, + "step": 28041 + }, + { + "epoch": 0.6172573189522342, + "grad_norm": 1.291552477475062, + "learning_rate": 3.3741872114390406e-06, + "loss": 0.2992, + "step": 28042 + }, + { + "epoch": 0.6172793308386528, + "grad_norm": 1.705086834509184, + "learning_rate": 3.373850129695472e-06, + "loss": 0.2983, + "step": 28043 + }, + { + "epoch": 0.6173013427250715, + "grad_norm": 3.470728091437521, + "learning_rate": 3.3735130562167196e-06, + "loss": 0.2544, + "step": 28044 + }, + { + "epoch": 0.6173233546114902, + "grad_norm": 1.53869471386758, + "learning_rate": 3.3731759910045003e-06, + "loss": 0.3533, + "step": 28045 + }, + { + "epoch": 0.6173453664979088, + "grad_norm": 1.535588867188959, + "learning_rate": 3.3728389340605282e-06, + "loss": 0.2929, + "step": 28046 + }, + { + "epoch": 0.6173673783843275, + "grad_norm": 1.3973134228587023, + "learning_rate": 3.3725018853865116e-06, + "loss": 0.3137, + "step": 28047 + }, + { + "epoch": 0.6173893902707462, + "grad_norm": 0.6317820684635436, + "learning_rate": 3.372164844984167e-06, + "loss": 0.4559, + "step": 28048 + }, + { + "epoch": 0.6174114021571648, + "grad_norm": 1.7908040955621964, + "learning_rate": 3.3718278128552074e-06, + "loss": 0.307, + "step": 28049 + }, + { + "epoch": 0.6174334140435835, + "grad_norm": 0.6137797287033682, + "learning_rate": 3.3714907890013436e-06, + "loss": 0.4618, + "step": 28050 + }, + { + "epoch": 0.6174554259300022, + "grad_norm": 1.4484216803729786, + "learning_rate": 3.3711537734242927e-06, + "loss": 0.2533, + "step": 28051 + }, + { + "epoch": 0.6174774378164208, + "grad_norm": 0.6330062831499431, + "learning_rate": 3.370816766125762e-06, + "loss": 0.4719, + "step": 28052 + }, + { + "epoch": 0.6174994497028395, + "grad_norm": 1.9100546681917967, + "learning_rate": 3.3704797671074683e-06, + "loss": 0.3802, + "step": 28053 + }, + { + "epoch": 0.6175214615892582, + "grad_norm": 1.0585402302854368, + "learning_rate": 3.3701427763711244e-06, + "loss": 0.2448, + "step": 28054 + }, + { + "epoch": 0.6175434734756768, + "grad_norm": 2.094381285994046, + "learning_rate": 3.3698057939184387e-06, + "loss": 0.3777, + "step": 28055 + }, + { + "epoch": 0.6175654853620955, + "grad_norm": 2.0816676300295787, + "learning_rate": 3.36946881975113e-06, + "loss": 0.2425, + "step": 28056 + }, + { + "epoch": 0.6175874972485143, + "grad_norm": 1.4979713450206833, + "learning_rate": 3.369131853870906e-06, + "loss": 0.3359, + "step": 28057 + }, + { + "epoch": 0.6176095091349328, + "grad_norm": 1.5463269745813322, + "learning_rate": 3.3687948962794827e-06, + "loss": 0.3877, + "step": 28058 + }, + { + "epoch": 0.6176315210213515, + "grad_norm": 1.6165033254626968, + "learning_rate": 3.368457946978572e-06, + "loss": 0.3311, + "step": 28059 + }, + { + "epoch": 0.6176535329077701, + "grad_norm": 1.5550719993764397, + "learning_rate": 3.3681210059698822e-06, + "loss": 0.389, + "step": 28060 + }, + { + "epoch": 0.6176755447941888, + "grad_norm": 2.548724322203085, + "learning_rate": 3.3677840732551326e-06, + "loss": 0.2868, + "step": 28061 + }, + { + "epoch": 0.6176975566806076, + "grad_norm": 1.8912125894473493, + "learning_rate": 3.3674471488360305e-06, + "loss": 0.3189, + "step": 28062 + }, + { + "epoch": 0.6177195685670261, + "grad_norm": 1.2991211922526165, + "learning_rate": 3.3671102327142912e-06, + "loss": 0.2459, + "step": 28063 + }, + { + "epoch": 0.6177415804534448, + "grad_norm": 2.3246334758474134, + "learning_rate": 3.3667733248916266e-06, + "loss": 0.3249, + "step": 28064 + }, + { + "epoch": 0.6177635923398636, + "grad_norm": 1.8843463117200223, + "learning_rate": 3.3664364253697468e-06, + "loss": 0.3186, + "step": 28065 + }, + { + "epoch": 0.6177856042262821, + "grad_norm": 1.226165097718042, + "learning_rate": 3.3660995341503666e-06, + "loss": 0.2805, + "step": 28066 + }, + { + "epoch": 0.6178076161127009, + "grad_norm": 1.308867300528199, + "learning_rate": 3.365762651235197e-06, + "loss": 0.3186, + "step": 28067 + }, + { + "epoch": 0.6178296279991196, + "grad_norm": 1.9175391240761575, + "learning_rate": 3.36542577662595e-06, + "loss": 0.361, + "step": 28068 + }, + { + "epoch": 0.6178516398855382, + "grad_norm": 1.3960603003440226, + "learning_rate": 3.36508891032434e-06, + "loss": 0.2283, + "step": 28069 + }, + { + "epoch": 0.6178736517719569, + "grad_norm": 1.6846091464213129, + "learning_rate": 3.3647520523320755e-06, + "loss": 0.2178, + "step": 28070 + }, + { + "epoch": 0.6178956636583756, + "grad_norm": 1.5602585133769338, + "learning_rate": 3.3644152026508714e-06, + "loss": 0.3548, + "step": 28071 + }, + { + "epoch": 0.6179176755447942, + "grad_norm": 2.066004205663135, + "learning_rate": 3.3640783612824396e-06, + "loss": 0.3, + "step": 28072 + }, + { + "epoch": 0.6179396874312129, + "grad_norm": 1.8223152720455658, + "learning_rate": 3.363741528228488e-06, + "loss": 0.2436, + "step": 28073 + }, + { + "epoch": 0.6179616993176316, + "grad_norm": 1.5283404849173305, + "learning_rate": 3.3634047034907345e-06, + "loss": 0.3121, + "step": 28074 + }, + { + "epoch": 0.6179837112040502, + "grad_norm": 0.6638261636348364, + "learning_rate": 3.363067887070887e-06, + "loss": 0.485, + "step": 28075 + }, + { + "epoch": 0.6180057230904689, + "grad_norm": 0.6631707351977937, + "learning_rate": 3.362731078970659e-06, + "loss": 0.4594, + "step": 28076 + }, + { + "epoch": 0.6180277349768876, + "grad_norm": 1.5569625366414692, + "learning_rate": 3.362394279191764e-06, + "loss": 0.3489, + "step": 28077 + }, + { + "epoch": 0.6180497468633062, + "grad_norm": 1.109921416496686, + "learning_rate": 3.3620574877359073e-06, + "loss": 0.2273, + "step": 28078 + }, + { + "epoch": 0.6180717587497249, + "grad_norm": 1.3812994635459117, + "learning_rate": 3.3617207046048083e-06, + "loss": 0.2821, + "step": 28079 + }, + { + "epoch": 0.6180937706361436, + "grad_norm": 1.7413337287303507, + "learning_rate": 3.361383929800174e-06, + "loss": 0.3676, + "step": 28080 + }, + { + "epoch": 0.6181157825225622, + "grad_norm": 7.6546517694290515, + "learning_rate": 3.361047163323718e-06, + "loss": 0.2785, + "step": 28081 + }, + { + "epoch": 0.6181377944089809, + "grad_norm": 1.453088059094367, + "learning_rate": 3.360710405177152e-06, + "loss": 0.2781, + "step": 28082 + }, + { + "epoch": 0.6181598062953995, + "grad_norm": 0.6188430527624188, + "learning_rate": 3.3603736553621857e-06, + "loss": 0.4541, + "step": 28083 + }, + { + "epoch": 0.6181818181818182, + "grad_norm": 1.8869913239154295, + "learning_rate": 3.3600369138805322e-06, + "loss": 0.3186, + "step": 28084 + }, + { + "epoch": 0.6182038300682369, + "grad_norm": 1.400372699286516, + "learning_rate": 3.3597001807339015e-06, + "loss": 0.2362, + "step": 28085 + }, + { + "epoch": 0.6182258419546555, + "grad_norm": 0.6382316166768882, + "learning_rate": 3.3593634559240076e-06, + "loss": 0.4826, + "step": 28086 + }, + { + "epoch": 0.6182478538410742, + "grad_norm": 1.4736692164572192, + "learning_rate": 3.3590267394525604e-06, + "loss": 0.2922, + "step": 28087 + }, + { + "epoch": 0.6182698657274929, + "grad_norm": 2.259246471229782, + "learning_rate": 3.35869003132127e-06, + "loss": 0.3118, + "step": 28088 + }, + { + "epoch": 0.6182918776139115, + "grad_norm": 1.2950728835516052, + "learning_rate": 3.3583533315318496e-06, + "loss": 0.3106, + "step": 28089 + }, + { + "epoch": 0.6183138895003302, + "grad_norm": 0.6263439190395115, + "learning_rate": 3.3580166400860108e-06, + "loss": 0.4632, + "step": 28090 + }, + { + "epoch": 0.6183359013867489, + "grad_norm": 1.6118768414499132, + "learning_rate": 3.357679956985461e-06, + "loss": 0.2501, + "step": 28091 + }, + { + "epoch": 0.6183579132731675, + "grad_norm": 1.3828806429517353, + "learning_rate": 3.357343282231917e-06, + "loss": 0.2698, + "step": 28092 + }, + { + "epoch": 0.6183799251595862, + "grad_norm": 0.611134747471379, + "learning_rate": 3.3570066158270847e-06, + "loss": 0.4556, + "step": 28093 + }, + { + "epoch": 0.6184019370460049, + "grad_norm": 1.40898608634453, + "learning_rate": 3.3566699577726793e-06, + "loss": 0.3475, + "step": 28094 + }, + { + "epoch": 0.6184239489324235, + "grad_norm": 1.6629065604708435, + "learning_rate": 3.356333308070411e-06, + "loss": 0.2983, + "step": 28095 + }, + { + "epoch": 0.6184459608188422, + "grad_norm": 7.51409931417434, + "learning_rate": 3.355996666721986e-06, + "loss": 0.2358, + "step": 28096 + }, + { + "epoch": 0.6184679727052609, + "grad_norm": 1.3987987025260893, + "learning_rate": 3.3556600337291235e-06, + "loss": 0.2932, + "step": 28097 + }, + { + "epoch": 0.6184899845916795, + "grad_norm": 1.153944680830019, + "learning_rate": 3.3553234090935272e-06, + "loss": 0.3054, + "step": 28098 + }, + { + "epoch": 0.6185119964780982, + "grad_norm": 2.2137521771522333, + "learning_rate": 3.354986792816912e-06, + "loss": 0.3253, + "step": 28099 + }, + { + "epoch": 0.6185340083645169, + "grad_norm": 1.2725350226731884, + "learning_rate": 3.354650184900989e-06, + "loss": 0.2108, + "step": 28100 + }, + { + "epoch": 0.6185560202509355, + "grad_norm": 0.6307502014092682, + "learning_rate": 3.354313585347465e-06, + "loss": 0.4703, + "step": 28101 + }, + { + "epoch": 0.6185780321373542, + "grad_norm": 1.5447956245019943, + "learning_rate": 3.353976994158055e-06, + "loss": 0.3066, + "step": 28102 + }, + { + "epoch": 0.6186000440237729, + "grad_norm": 1.7219175773644122, + "learning_rate": 3.3536404113344667e-06, + "loss": 0.2716, + "step": 28103 + }, + { + "epoch": 0.6186220559101915, + "grad_norm": 0.6440916266628973, + "learning_rate": 3.353303836878413e-06, + "loss": 0.4818, + "step": 28104 + }, + { + "epoch": 0.6186440677966102, + "grad_norm": 1.4711146796949375, + "learning_rate": 3.352967270791605e-06, + "loss": 0.2427, + "step": 28105 + }, + { + "epoch": 0.6186660796830288, + "grad_norm": 1.5461792918599264, + "learning_rate": 3.3526307130757495e-06, + "loss": 0.3837, + "step": 28106 + }, + { + "epoch": 0.6186880915694475, + "grad_norm": 1.540774674555778, + "learning_rate": 3.3522941637325613e-06, + "loss": 0.3045, + "step": 28107 + }, + { + "epoch": 0.6187101034558662, + "grad_norm": 1.5611774468308979, + "learning_rate": 3.351957622763749e-06, + "loss": 0.3316, + "step": 28108 + }, + { + "epoch": 0.6187321153422848, + "grad_norm": 1.6156429349322774, + "learning_rate": 3.3516210901710205e-06, + "loss": 0.4441, + "step": 28109 + }, + { + "epoch": 0.6187541272287035, + "grad_norm": 2.2817280324735045, + "learning_rate": 3.351284565956092e-06, + "loss": 0.428, + "step": 28110 + }, + { + "epoch": 0.6187761391151222, + "grad_norm": 1.335852458646258, + "learning_rate": 3.3509480501206687e-06, + "loss": 0.2604, + "step": 28111 + }, + { + "epoch": 0.6187981510015408, + "grad_norm": 1.7738285976259551, + "learning_rate": 3.3506115426664636e-06, + "loss": 0.2718, + "step": 28112 + }, + { + "epoch": 0.6188201628879595, + "grad_norm": 1.3955235969526563, + "learning_rate": 3.350275043595188e-06, + "loss": 0.2577, + "step": 28113 + }, + { + "epoch": 0.6188421747743782, + "grad_norm": 2.366383753783427, + "learning_rate": 3.3499385529085466e-06, + "loss": 0.347, + "step": 28114 + }, + { + "epoch": 0.6188641866607968, + "grad_norm": 1.195990284623547, + "learning_rate": 3.3496020706082567e-06, + "loss": 0.3548, + "step": 28115 + }, + { + "epoch": 0.6188861985472155, + "grad_norm": 0.6020920158556743, + "learning_rate": 3.349265596696023e-06, + "loss": 0.4627, + "step": 28116 + }, + { + "epoch": 0.6189082104336342, + "grad_norm": 1.707114738973809, + "learning_rate": 3.3489291311735593e-06, + "loss": 0.3003, + "step": 28117 + }, + { + "epoch": 0.6189302223200528, + "grad_norm": 1.4654080731738794, + "learning_rate": 3.348592674042575e-06, + "loss": 0.2862, + "step": 28118 + }, + { + "epoch": 0.6189522342064715, + "grad_norm": 1.6543095899790576, + "learning_rate": 3.348256225304777e-06, + "loss": 0.3428, + "step": 28119 + }, + { + "epoch": 0.6189742460928902, + "grad_norm": 1.7708770235290494, + "learning_rate": 3.3479197849618787e-06, + "loss": 0.3524, + "step": 28120 + }, + { + "epoch": 0.6189962579793088, + "grad_norm": 1.3905925467405196, + "learning_rate": 3.3475833530155875e-06, + "loss": 0.3055, + "step": 28121 + }, + { + "epoch": 0.6190182698657275, + "grad_norm": 1.6987199008009612, + "learning_rate": 3.347246929467617e-06, + "loss": 0.383, + "step": 28122 + }, + { + "epoch": 0.6190402817521462, + "grad_norm": 0.5919839163150727, + "learning_rate": 3.3469105143196744e-06, + "loss": 0.4686, + "step": 28123 + }, + { + "epoch": 0.6190622936385648, + "grad_norm": 1.3869233726032357, + "learning_rate": 3.346574107573468e-06, + "loss": 0.3276, + "step": 28124 + }, + { + "epoch": 0.6190843055249835, + "grad_norm": 1.5369516457587549, + "learning_rate": 3.3462377092307107e-06, + "loss": 0.2923, + "step": 28125 + }, + { + "epoch": 0.6191063174114022, + "grad_norm": 1.335598857967821, + "learning_rate": 3.345901319293112e-06, + "loss": 0.2388, + "step": 28126 + }, + { + "epoch": 0.6191283292978208, + "grad_norm": 1.434237486711586, + "learning_rate": 3.345564937762378e-06, + "loss": 0.2826, + "step": 28127 + }, + { + "epoch": 0.6191503411842395, + "grad_norm": 18.977514338845452, + "learning_rate": 3.345228564640224e-06, + "loss": 0.2996, + "step": 28128 + }, + { + "epoch": 0.6191723530706582, + "grad_norm": 1.2164959590740774, + "learning_rate": 3.3448921999283533e-06, + "loss": 0.249, + "step": 28129 + }, + { + "epoch": 0.6191943649570768, + "grad_norm": 1.3538658017124139, + "learning_rate": 3.3445558436284804e-06, + "loss": 0.3342, + "step": 28130 + }, + { + "epoch": 0.6192163768434955, + "grad_norm": 1.6464499677628563, + "learning_rate": 3.3442194957423136e-06, + "loss": 0.3687, + "step": 28131 + }, + { + "epoch": 0.6192383887299141, + "grad_norm": 1.688158603446115, + "learning_rate": 3.343883156271559e-06, + "loss": 0.3902, + "step": 28132 + }, + { + "epoch": 0.6192604006163328, + "grad_norm": 1.367516396267549, + "learning_rate": 3.3435468252179324e-06, + "loss": 0.2867, + "step": 28133 + }, + { + "epoch": 0.6192824125027515, + "grad_norm": 1.442566993867311, + "learning_rate": 3.3432105025831375e-06, + "loss": 0.3317, + "step": 28134 + }, + { + "epoch": 0.6193044243891701, + "grad_norm": 1.411815910559684, + "learning_rate": 3.3428741883688864e-06, + "loss": 0.2732, + "step": 28135 + }, + { + "epoch": 0.6193264362755888, + "grad_norm": 1.4356261123198004, + "learning_rate": 3.3425378825768885e-06, + "loss": 0.2522, + "step": 28136 + }, + { + "epoch": 0.6193484481620075, + "grad_norm": 1.6817630427216625, + "learning_rate": 3.342201585208851e-06, + "loss": 0.35, + "step": 28137 + }, + { + "epoch": 0.6193704600484261, + "grad_norm": 0.6357631206801924, + "learning_rate": 3.341865296266485e-06, + "loss": 0.4712, + "step": 28138 + }, + { + "epoch": 0.6193924719348448, + "grad_norm": 1.4884224759960099, + "learning_rate": 3.341529015751498e-06, + "loss": 0.23, + "step": 28139 + }, + { + "epoch": 0.6194144838212635, + "grad_norm": 1.360615461615646, + "learning_rate": 3.341192743665602e-06, + "loss": 0.3225, + "step": 28140 + }, + { + "epoch": 0.6194364957076821, + "grad_norm": 1.2329602737403222, + "learning_rate": 3.3408564800105046e-06, + "loss": 0.2771, + "step": 28141 + }, + { + "epoch": 0.6194585075941008, + "grad_norm": 1.615184512577652, + "learning_rate": 3.3405202247879127e-06, + "loss": 0.3869, + "step": 28142 + }, + { + "epoch": 0.6194805194805195, + "grad_norm": 0.5889192897906549, + "learning_rate": 3.3401839779995383e-06, + "loss": 0.4525, + "step": 28143 + }, + { + "epoch": 0.6195025313669381, + "grad_norm": 1.547239905099911, + "learning_rate": 3.339847739647088e-06, + "loss": 0.4064, + "step": 28144 + }, + { + "epoch": 0.6195245432533568, + "grad_norm": 1.555035899615449, + "learning_rate": 3.339511509732273e-06, + "loss": 0.3766, + "step": 28145 + }, + { + "epoch": 0.6195465551397755, + "grad_norm": 1.342177737158851, + "learning_rate": 3.3391752882568025e-06, + "loss": 0.3005, + "step": 28146 + }, + { + "epoch": 0.6195685670261941, + "grad_norm": 8.483117073393812, + "learning_rate": 3.3388390752223814e-06, + "loss": 0.2371, + "step": 28147 + }, + { + "epoch": 0.6195905789126128, + "grad_norm": 2.088275627899347, + "learning_rate": 3.3385028706307223e-06, + "loss": 0.3821, + "step": 28148 + }, + { + "epoch": 0.6196125907990315, + "grad_norm": 2.4352328842273896, + "learning_rate": 3.338166674483534e-06, + "loss": 0.3201, + "step": 28149 + }, + { + "epoch": 0.6196346026854501, + "grad_norm": 1.4871351738330236, + "learning_rate": 3.3378304867825207e-06, + "loss": 0.2911, + "step": 28150 + }, + { + "epoch": 0.6196566145718688, + "grad_norm": 1.2880341523897538, + "learning_rate": 3.3374943075293965e-06, + "loss": 0.3445, + "step": 28151 + }, + { + "epoch": 0.6196786264582875, + "grad_norm": 1.801292266410048, + "learning_rate": 3.3371581367258663e-06, + "loss": 0.3349, + "step": 28152 + }, + { + "epoch": 0.6197006383447061, + "grad_norm": 2.180347578898203, + "learning_rate": 3.336821974373641e-06, + "loss": 0.2903, + "step": 28153 + }, + { + "epoch": 0.6197226502311248, + "grad_norm": 1.2600067788312674, + "learning_rate": 3.3364858204744295e-06, + "loss": 0.2802, + "step": 28154 + }, + { + "epoch": 0.6197446621175434, + "grad_norm": 1.9203348944904761, + "learning_rate": 3.3361496750299367e-06, + "loss": 0.3618, + "step": 28155 + }, + { + "epoch": 0.6197666740039621, + "grad_norm": 1.2594534783009619, + "learning_rate": 3.3358135380418744e-06, + "loss": 0.235, + "step": 28156 + }, + { + "epoch": 0.6197886858903808, + "grad_norm": 0.6339161083825755, + "learning_rate": 3.335477409511949e-06, + "loss": 0.4662, + "step": 28157 + }, + { + "epoch": 0.6198106977767994, + "grad_norm": 1.375751530691272, + "learning_rate": 3.3351412894418706e-06, + "loss": 0.295, + "step": 28158 + }, + { + "epoch": 0.6198327096632181, + "grad_norm": 1.4487191448168684, + "learning_rate": 3.3348051778333477e-06, + "loss": 0.3178, + "step": 28159 + }, + { + "epoch": 0.6198547215496368, + "grad_norm": 1.398927101772595, + "learning_rate": 3.3344690746880858e-06, + "loss": 0.2875, + "step": 28160 + }, + { + "epoch": 0.6198767334360554, + "grad_norm": 1.5976790319187157, + "learning_rate": 3.334132980007796e-06, + "loss": 0.2175, + "step": 28161 + }, + { + "epoch": 0.6198987453224741, + "grad_norm": 1.150347441384648, + "learning_rate": 3.3337968937941847e-06, + "loss": 0.3355, + "step": 28162 + }, + { + "epoch": 0.6199207572088928, + "grad_norm": 1.3747204093957217, + "learning_rate": 3.333460816048961e-06, + "loss": 0.3154, + "step": 28163 + }, + { + "epoch": 0.6199427690953114, + "grad_norm": 1.123317760229821, + "learning_rate": 3.3331247467738337e-06, + "loss": 0.2323, + "step": 28164 + }, + { + "epoch": 0.6199647809817301, + "grad_norm": 1.4838266419130688, + "learning_rate": 3.3327886859705083e-06, + "loss": 0.3424, + "step": 28165 + }, + { + "epoch": 0.6199867928681488, + "grad_norm": 1.7525243227961103, + "learning_rate": 3.332452633640695e-06, + "loss": 0.3147, + "step": 28166 + }, + { + "epoch": 0.6200088047545674, + "grad_norm": 2.2312363888689877, + "learning_rate": 3.332116589786103e-06, + "loss": 0.3097, + "step": 28167 + }, + { + "epoch": 0.6200308166409861, + "grad_norm": 0.644719756437563, + "learning_rate": 3.3317805544084346e-06, + "loss": 0.4919, + "step": 28168 + }, + { + "epoch": 0.6200528285274048, + "grad_norm": 3.5533588879999014, + "learning_rate": 3.331444527509405e-06, + "loss": 0.4075, + "step": 28169 + }, + { + "epoch": 0.6200748404138234, + "grad_norm": 1.3272606710820265, + "learning_rate": 3.331108509090716e-06, + "loss": 0.2784, + "step": 28170 + }, + { + "epoch": 0.6200968523002421, + "grad_norm": 1.6103470013493366, + "learning_rate": 3.330772499154079e-06, + "loss": 0.2545, + "step": 28171 + }, + { + "epoch": 0.6201188641866608, + "grad_norm": 1.48582086548755, + "learning_rate": 3.3304364977012023e-06, + "loss": 0.2495, + "step": 28172 + }, + { + "epoch": 0.6201408760730794, + "grad_norm": 1.5071217606871703, + "learning_rate": 3.330100504733789e-06, + "loss": 0.3349, + "step": 28173 + }, + { + "epoch": 0.6201628879594981, + "grad_norm": 0.629604920928223, + "learning_rate": 3.329764520253551e-06, + "loss": 0.4926, + "step": 28174 + }, + { + "epoch": 0.6201848998459168, + "grad_norm": 0.5997931788792895, + "learning_rate": 3.3294285442621944e-06, + "loss": 0.4644, + "step": 28175 + }, + { + "epoch": 0.6202069117323354, + "grad_norm": 2.062005507850295, + "learning_rate": 3.329092576761427e-06, + "loss": 0.3454, + "step": 28176 + }, + { + "epoch": 0.6202289236187541, + "grad_norm": 0.6107625795401338, + "learning_rate": 3.3287566177529584e-06, + "loss": 0.4994, + "step": 28177 + }, + { + "epoch": 0.6202509355051727, + "grad_norm": 1.2053505574957777, + "learning_rate": 3.3284206672384913e-06, + "loss": 0.2051, + "step": 28178 + }, + { + "epoch": 0.6202729473915914, + "grad_norm": 1.7932867289208465, + "learning_rate": 3.3280847252197368e-06, + "loss": 0.3037, + "step": 28179 + }, + { + "epoch": 0.6202949592780101, + "grad_norm": 1.190768894777785, + "learning_rate": 3.3277487916984007e-06, + "loss": 0.2664, + "step": 28180 + }, + { + "epoch": 0.6203169711644287, + "grad_norm": 1.0741398610169515, + "learning_rate": 3.3274128666761923e-06, + "loss": 0.2087, + "step": 28181 + }, + { + "epoch": 0.6203389830508474, + "grad_norm": 1.2077142913673165, + "learning_rate": 3.3270769501548185e-06, + "loss": 0.3075, + "step": 28182 + }, + { + "epoch": 0.6203609949372662, + "grad_norm": 1.8865142610524042, + "learning_rate": 3.326741042135984e-06, + "loss": 0.3289, + "step": 28183 + }, + { + "epoch": 0.6203830068236847, + "grad_norm": 1.5580349979434494, + "learning_rate": 3.3264051426213985e-06, + "loss": 0.2959, + "step": 28184 + }, + { + "epoch": 0.6204050187101035, + "grad_norm": 1.9233238825176493, + "learning_rate": 3.326069251612768e-06, + "loss": 0.2944, + "step": 28185 + }, + { + "epoch": 0.6204270305965222, + "grad_norm": 1.9083580437089138, + "learning_rate": 3.3257333691118016e-06, + "loss": 0.2974, + "step": 28186 + }, + { + "epoch": 0.6204490424829407, + "grad_norm": 1.5433304913567982, + "learning_rate": 3.325397495120204e-06, + "loss": 0.3392, + "step": 28187 + }, + { + "epoch": 0.6204710543693595, + "grad_norm": 1.511643333356252, + "learning_rate": 3.3250616296396824e-06, + "loss": 0.3341, + "step": 28188 + }, + { + "epoch": 0.6204930662557782, + "grad_norm": 2.63677201561886, + "learning_rate": 3.3247257726719457e-06, + "loss": 0.2747, + "step": 28189 + }, + { + "epoch": 0.6205150781421968, + "grad_norm": 1.686493673277952, + "learning_rate": 3.3243899242187007e-06, + "loss": 0.2702, + "step": 28190 + }, + { + "epoch": 0.6205370900286155, + "grad_norm": 1.7388994569145344, + "learning_rate": 3.3240540842816515e-06, + "loss": 0.3193, + "step": 28191 + }, + { + "epoch": 0.6205591019150342, + "grad_norm": 1.8237851208599813, + "learning_rate": 3.323718252862508e-06, + "loss": 0.3501, + "step": 28192 + }, + { + "epoch": 0.6205811138014528, + "grad_norm": 1.5955924442100142, + "learning_rate": 3.3233824299629747e-06, + "loss": 0.3349, + "step": 28193 + }, + { + "epoch": 0.6206031256878715, + "grad_norm": 1.514464265148588, + "learning_rate": 3.3230466155847606e-06, + "loss": 0.262, + "step": 28194 + }, + { + "epoch": 0.6206251375742902, + "grad_norm": 2.25322017171446, + "learning_rate": 3.3227108097295733e-06, + "loss": 0.2336, + "step": 28195 + }, + { + "epoch": 0.6206471494607088, + "grad_norm": 1.85184108168786, + "learning_rate": 3.322375012399115e-06, + "loss": 0.2562, + "step": 28196 + }, + { + "epoch": 0.6206691613471275, + "grad_norm": 1.3291783451449006, + "learning_rate": 3.322039223595097e-06, + "loss": 0.3182, + "step": 28197 + }, + { + "epoch": 0.6206911732335462, + "grad_norm": 1.4285992351991719, + "learning_rate": 3.321703443319222e-06, + "loss": 0.2557, + "step": 28198 + }, + { + "epoch": 0.6207131851199648, + "grad_norm": 1.484258182086969, + "learning_rate": 3.3213676715732e-06, + "loss": 0.2969, + "step": 28199 + }, + { + "epoch": 0.6207351970063835, + "grad_norm": 2.890709223702371, + "learning_rate": 3.3210319083587377e-06, + "loss": 0.3692, + "step": 28200 + }, + { + "epoch": 0.6207572088928021, + "grad_norm": 0.6128905770459205, + "learning_rate": 3.320696153677537e-06, + "loss": 0.4407, + "step": 28201 + }, + { + "epoch": 0.6207792207792208, + "grad_norm": 1.4753663119725782, + "learning_rate": 3.32036040753131e-06, + "loss": 0.3271, + "step": 28202 + }, + { + "epoch": 0.6208012326656395, + "grad_norm": 1.600653718519065, + "learning_rate": 3.3200246699217586e-06, + "loss": 0.2664, + "step": 28203 + }, + { + "epoch": 0.6208232445520581, + "grad_norm": 1.6939934881334455, + "learning_rate": 3.319688940850593e-06, + "loss": 0.3718, + "step": 28204 + }, + { + "epoch": 0.6208452564384768, + "grad_norm": 1.1637251848402128, + "learning_rate": 3.3193532203195166e-06, + "loss": 0.2984, + "step": 28205 + }, + { + "epoch": 0.6208672683248955, + "grad_norm": 1.2960979183171435, + "learning_rate": 3.319017508330236e-06, + "loss": 0.3162, + "step": 28206 + }, + { + "epoch": 0.6208892802113141, + "grad_norm": 0.6106732596163432, + "learning_rate": 3.3186818048844593e-06, + "loss": 0.4472, + "step": 28207 + }, + { + "epoch": 0.6209112920977328, + "grad_norm": 2.3594516559510255, + "learning_rate": 3.3183461099838922e-06, + "loss": 0.3245, + "step": 28208 + }, + { + "epoch": 0.6209333039841515, + "grad_norm": 1.6576092686504043, + "learning_rate": 3.3180104236302383e-06, + "loss": 0.3067, + "step": 28209 + }, + { + "epoch": 0.6209553158705701, + "grad_norm": 1.9491121224580301, + "learning_rate": 3.317674745825207e-06, + "loss": 0.3038, + "step": 28210 + }, + { + "epoch": 0.6209773277569888, + "grad_norm": 1.4613690260776242, + "learning_rate": 3.3173390765705015e-06, + "loss": 0.2597, + "step": 28211 + }, + { + "epoch": 0.6209993396434075, + "grad_norm": 0.6187638062179749, + "learning_rate": 3.317003415867831e-06, + "loss": 0.4705, + "step": 28212 + }, + { + "epoch": 0.6210213515298261, + "grad_norm": 1.6876150012502165, + "learning_rate": 3.3166677637188997e-06, + "loss": 0.3077, + "step": 28213 + }, + { + "epoch": 0.6210433634162448, + "grad_norm": 3.523037949249839, + "learning_rate": 3.316332120125412e-06, + "loss": 0.388, + "step": 28214 + }, + { + "epoch": 0.6210653753026635, + "grad_norm": 0.6193472233119935, + "learning_rate": 3.315996485089077e-06, + "loss": 0.4669, + "step": 28215 + }, + { + "epoch": 0.6210873871890821, + "grad_norm": 1.5434743601304313, + "learning_rate": 3.3156608586115976e-06, + "loss": 0.2906, + "step": 28216 + }, + { + "epoch": 0.6211093990755008, + "grad_norm": 1.2630484902347285, + "learning_rate": 3.3153252406946817e-06, + "loss": 0.2778, + "step": 28217 + }, + { + "epoch": 0.6211314109619195, + "grad_norm": 0.6196157818906884, + "learning_rate": 3.3149896313400352e-06, + "loss": 0.4582, + "step": 28218 + }, + { + "epoch": 0.6211534228483381, + "grad_norm": 1.6861287404845895, + "learning_rate": 3.314654030549361e-06, + "loss": 0.3184, + "step": 28219 + }, + { + "epoch": 0.6211754347347568, + "grad_norm": 1.73676722623263, + "learning_rate": 3.3143184383243687e-06, + "loss": 0.2481, + "step": 28220 + }, + { + "epoch": 0.6211974466211755, + "grad_norm": 0.6156723867156401, + "learning_rate": 3.3139828546667596e-06, + "loss": 0.4737, + "step": 28221 + }, + { + "epoch": 0.6212194585075941, + "grad_norm": 1.4219153803258118, + "learning_rate": 3.313647279578245e-06, + "loss": 0.3342, + "step": 28222 + }, + { + "epoch": 0.6212414703940128, + "grad_norm": 1.8264044547858949, + "learning_rate": 3.313311713060525e-06, + "loss": 0.384, + "step": 28223 + }, + { + "epoch": 0.6212634822804314, + "grad_norm": 2.2001666953287593, + "learning_rate": 3.312976155115306e-06, + "loss": 0.2389, + "step": 28224 + }, + { + "epoch": 0.6212854941668501, + "grad_norm": 5.715291394457731, + "learning_rate": 3.312640605744296e-06, + "loss": 0.2789, + "step": 28225 + }, + { + "epoch": 0.6213075060532688, + "grad_norm": 1.5027789836070342, + "learning_rate": 3.3123050649492006e-06, + "loss": 0.2621, + "step": 28226 + }, + { + "epoch": 0.6213295179396874, + "grad_norm": 1.3876893842590954, + "learning_rate": 3.3119695327317202e-06, + "loss": 0.303, + "step": 28227 + }, + { + "epoch": 0.6213515298261061, + "grad_norm": 1.3620170244469059, + "learning_rate": 3.3116340090935656e-06, + "loss": 0.3303, + "step": 28228 + }, + { + "epoch": 0.6213735417125248, + "grad_norm": 1.9149976305041294, + "learning_rate": 3.311298494036439e-06, + "loss": 0.3615, + "step": 28229 + }, + { + "epoch": 0.6213955535989434, + "grad_norm": 3.947693786050761, + "learning_rate": 3.3109629875620477e-06, + "loss": 0.3093, + "step": 28230 + }, + { + "epoch": 0.6214175654853621, + "grad_norm": 1.4535347699443988, + "learning_rate": 3.310627489672096e-06, + "loss": 0.2605, + "step": 28231 + }, + { + "epoch": 0.6214395773717808, + "grad_norm": 0.6091518136369566, + "learning_rate": 3.310292000368287e-06, + "loss": 0.4491, + "step": 28232 + }, + { + "epoch": 0.6214615892581994, + "grad_norm": 1.657101456057209, + "learning_rate": 3.3099565196523297e-06, + "loss": 0.3018, + "step": 28233 + }, + { + "epoch": 0.6214836011446181, + "grad_norm": 1.7298951255591906, + "learning_rate": 3.309621047525925e-06, + "loss": 0.3443, + "step": 28234 + }, + { + "epoch": 0.6215056130310368, + "grad_norm": 0.6198070993721435, + "learning_rate": 3.3092855839907818e-06, + "loss": 0.4519, + "step": 28235 + }, + { + "epoch": 0.6215276249174554, + "grad_norm": 1.2394931043836919, + "learning_rate": 3.3089501290486043e-06, + "loss": 0.3309, + "step": 28236 + }, + { + "epoch": 0.6215496368038741, + "grad_norm": 1.3905864395638534, + "learning_rate": 3.308614682701094e-06, + "loss": 0.3018, + "step": 28237 + }, + { + "epoch": 0.6215716486902928, + "grad_norm": 1.4746483978262148, + "learning_rate": 3.3082792449499596e-06, + "loss": 0.2874, + "step": 28238 + }, + { + "epoch": 0.6215936605767114, + "grad_norm": 1.4187138358123594, + "learning_rate": 3.3079438157969033e-06, + "loss": 0.3727, + "step": 28239 + }, + { + "epoch": 0.6216156724631301, + "grad_norm": 1.709482336936097, + "learning_rate": 3.3076083952436337e-06, + "loss": 0.3935, + "step": 28240 + }, + { + "epoch": 0.6216376843495488, + "grad_norm": 1.8597417869659787, + "learning_rate": 3.3072729832918516e-06, + "loss": 0.2461, + "step": 28241 + }, + { + "epoch": 0.6216596962359674, + "grad_norm": 1.386199239499145, + "learning_rate": 3.3069375799432623e-06, + "loss": 0.2596, + "step": 28242 + }, + { + "epoch": 0.6216817081223861, + "grad_norm": 1.874076776695559, + "learning_rate": 3.306602185199573e-06, + "loss": 0.3302, + "step": 28243 + }, + { + "epoch": 0.6217037200088048, + "grad_norm": 1.3197509442963884, + "learning_rate": 3.306266799062485e-06, + "loss": 0.3397, + "step": 28244 + }, + { + "epoch": 0.6217257318952234, + "grad_norm": 1.4705349942201908, + "learning_rate": 3.305931421533707e-06, + "loss": 0.2384, + "step": 28245 + }, + { + "epoch": 0.6217477437816421, + "grad_norm": 1.4439661895031506, + "learning_rate": 3.3055960526149394e-06, + "loss": 0.2934, + "step": 28246 + }, + { + "epoch": 0.6217697556680607, + "grad_norm": 1.790958210497329, + "learning_rate": 3.3052606923078876e-06, + "loss": 0.3904, + "step": 28247 + }, + { + "epoch": 0.6217917675544794, + "grad_norm": 2.630375975170042, + "learning_rate": 3.304925340614258e-06, + "loss": 0.3515, + "step": 28248 + }, + { + "epoch": 0.6218137794408981, + "grad_norm": 6.1494375753900465, + "learning_rate": 3.3045899975357552e-06, + "loss": 0.3745, + "step": 28249 + }, + { + "epoch": 0.6218357913273167, + "grad_norm": 0.6181033339778588, + "learning_rate": 3.3042546630740796e-06, + "loss": 0.4429, + "step": 28250 + }, + { + "epoch": 0.6218578032137354, + "grad_norm": 0.6077743370175177, + "learning_rate": 3.3039193372309397e-06, + "loss": 0.4706, + "step": 28251 + }, + { + "epoch": 0.6218798151001541, + "grad_norm": 1.8555055195781212, + "learning_rate": 3.3035840200080367e-06, + "loss": 0.3714, + "step": 28252 + }, + { + "epoch": 0.6219018269865727, + "grad_norm": 1.7093882912470626, + "learning_rate": 3.303248711407078e-06, + "loss": 0.2933, + "step": 28253 + }, + { + "epoch": 0.6219238388729914, + "grad_norm": 1.7477238483804687, + "learning_rate": 3.3029134114297657e-06, + "loss": 0.2636, + "step": 28254 + }, + { + "epoch": 0.6219458507594101, + "grad_norm": 2.03487505041363, + "learning_rate": 3.3025781200778036e-06, + "loss": 0.3086, + "step": 28255 + }, + { + "epoch": 0.6219678626458287, + "grad_norm": 1.3286639621408052, + "learning_rate": 3.302242837352897e-06, + "loss": 0.2742, + "step": 28256 + }, + { + "epoch": 0.6219898745322474, + "grad_norm": 1.5097337175504848, + "learning_rate": 3.301907563256749e-06, + "loss": 0.3739, + "step": 28257 + }, + { + "epoch": 0.6220118864186661, + "grad_norm": 1.40993891817152, + "learning_rate": 3.3015722977910657e-06, + "loss": 0.3344, + "step": 28258 + }, + { + "epoch": 0.6220338983050847, + "grad_norm": 1.5120894505002145, + "learning_rate": 3.301237040957549e-06, + "loss": 0.3899, + "step": 28259 + }, + { + "epoch": 0.6220559101915034, + "grad_norm": 0.6005108282970039, + "learning_rate": 3.300901792757902e-06, + "loss": 0.458, + "step": 28260 + }, + { + "epoch": 0.6220779220779221, + "grad_norm": 1.2915171939762273, + "learning_rate": 3.3005665531938307e-06, + "loss": 0.2252, + "step": 28261 + }, + { + "epoch": 0.6220999339643407, + "grad_norm": 1.7970337264899268, + "learning_rate": 3.3002313222670374e-06, + "loss": 0.2504, + "step": 28262 + }, + { + "epoch": 0.6221219458507594, + "grad_norm": 1.4331603114034754, + "learning_rate": 3.2998960999792284e-06, + "loss": 0.288, + "step": 28263 + }, + { + "epoch": 0.6221439577371781, + "grad_norm": 1.0617885092359225, + "learning_rate": 3.299560886332105e-06, + "loss": 0.1989, + "step": 28264 + }, + { + "epoch": 0.6221659696235967, + "grad_norm": 1.6665795719632484, + "learning_rate": 3.2992256813273706e-06, + "loss": 0.2476, + "step": 28265 + }, + { + "epoch": 0.6221879815100154, + "grad_norm": 1.5641358334221729, + "learning_rate": 3.2988904849667306e-06, + "loss": 0.2788, + "step": 28266 + }, + { + "epoch": 0.6222099933964341, + "grad_norm": 1.2478035372298872, + "learning_rate": 3.29855529725189e-06, + "loss": 0.2845, + "step": 28267 + }, + { + "epoch": 0.6222320052828527, + "grad_norm": 1.7458054677549817, + "learning_rate": 3.2982201181845464e-06, + "loss": 0.3377, + "step": 28268 + }, + { + "epoch": 0.6222540171692714, + "grad_norm": 1.412307776285493, + "learning_rate": 3.2978849477664093e-06, + "loss": 0.3035, + "step": 28269 + }, + { + "epoch": 0.6222760290556901, + "grad_norm": 1.4899196901063156, + "learning_rate": 3.297549785999179e-06, + "loss": 0.323, + "step": 28270 + }, + { + "epoch": 0.6222980409421087, + "grad_norm": 1.6008424467084543, + "learning_rate": 3.2972146328845604e-06, + "loss": 0.3819, + "step": 28271 + }, + { + "epoch": 0.6223200528285274, + "grad_norm": 0.6090181402586508, + "learning_rate": 3.2968794884242583e-06, + "loss": 0.4606, + "step": 28272 + }, + { + "epoch": 0.622342064714946, + "grad_norm": 0.6356811702767182, + "learning_rate": 3.2965443526199714e-06, + "loss": 0.4612, + "step": 28273 + }, + { + "epoch": 0.6223640766013647, + "grad_norm": 0.6211205227947997, + "learning_rate": 3.2962092254734078e-06, + "loss": 0.4575, + "step": 28274 + }, + { + "epoch": 0.6223860884877834, + "grad_norm": 1.5444148490843705, + "learning_rate": 3.295874106986267e-06, + "loss": 0.2543, + "step": 28275 + }, + { + "epoch": 0.622408100374202, + "grad_norm": 1.6809666989140652, + "learning_rate": 3.2955389971602565e-06, + "loss": 0.3014, + "step": 28276 + }, + { + "epoch": 0.6224301122606207, + "grad_norm": 1.325420739097804, + "learning_rate": 3.2952038959970756e-06, + "loss": 0.3421, + "step": 28277 + }, + { + "epoch": 0.6224521241470394, + "grad_norm": 1.348993993332581, + "learning_rate": 3.294868803498429e-06, + "loss": 0.2692, + "step": 28278 + }, + { + "epoch": 0.622474136033458, + "grad_norm": 1.481748412164379, + "learning_rate": 3.2945337196660195e-06, + "loss": 0.295, + "step": 28279 + }, + { + "epoch": 0.6224961479198767, + "grad_norm": 1.7737259800922662, + "learning_rate": 3.2941986445015493e-06, + "loss": 0.3259, + "step": 28280 + }, + { + "epoch": 0.6225181598062954, + "grad_norm": 1.3453683068556452, + "learning_rate": 3.293863578006725e-06, + "loss": 0.3165, + "step": 28281 + }, + { + "epoch": 0.622540171692714, + "grad_norm": 1.6525011389850488, + "learning_rate": 3.293528520183246e-06, + "loss": 0.3231, + "step": 28282 + }, + { + "epoch": 0.6225621835791327, + "grad_norm": 1.6649159514482332, + "learning_rate": 3.2931934710328147e-06, + "loss": 0.3107, + "step": 28283 + }, + { + "epoch": 0.6225841954655514, + "grad_norm": 1.32895444089628, + "learning_rate": 3.2928584305571364e-06, + "loss": 0.2645, + "step": 28284 + }, + { + "epoch": 0.62260620735197, + "grad_norm": 1.430894087420569, + "learning_rate": 3.2925233987579117e-06, + "loss": 0.3546, + "step": 28285 + }, + { + "epoch": 0.6226282192383887, + "grad_norm": 1.799046674826251, + "learning_rate": 3.292188375636848e-06, + "loss": 0.2763, + "step": 28286 + }, + { + "epoch": 0.6226502311248074, + "grad_norm": 0.6547659078193865, + "learning_rate": 3.291853361195643e-06, + "loss": 0.4847, + "step": 28287 + }, + { + "epoch": 0.622672243011226, + "grad_norm": 0.6260707383916603, + "learning_rate": 3.291518355435999e-06, + "loss": 0.4528, + "step": 28288 + }, + { + "epoch": 0.6226942548976447, + "grad_norm": 1.5961389950992928, + "learning_rate": 3.291183358359623e-06, + "loss": 0.2339, + "step": 28289 + }, + { + "epoch": 0.6227162667840634, + "grad_norm": 0.6165730134385852, + "learning_rate": 3.2908483699682158e-06, + "loss": 0.474, + "step": 28290 + }, + { + "epoch": 0.622738278670482, + "grad_norm": 1.7339731869109098, + "learning_rate": 3.2905133902634774e-06, + "loss": 0.2573, + "step": 28291 + }, + { + "epoch": 0.6227602905569007, + "grad_norm": 1.5460248848173734, + "learning_rate": 3.290178419247114e-06, + "loss": 0.3237, + "step": 28292 + }, + { + "epoch": 0.6227823024433194, + "grad_norm": 1.688645140790792, + "learning_rate": 3.2898434569208247e-06, + "loss": 0.3813, + "step": 28293 + }, + { + "epoch": 0.622804314329738, + "grad_norm": 2.7422471030987725, + "learning_rate": 3.2895085032863162e-06, + "loss": 0.2318, + "step": 28294 + }, + { + "epoch": 0.6228263262161567, + "grad_norm": 1.6686596612785403, + "learning_rate": 3.2891735583452873e-06, + "loss": 0.3283, + "step": 28295 + }, + { + "epoch": 0.6228483381025753, + "grad_norm": 0.6593778860322447, + "learning_rate": 3.28883862209944e-06, + "loss": 0.4687, + "step": 28296 + }, + { + "epoch": 0.622870349988994, + "grad_norm": 1.4159176011008594, + "learning_rate": 3.2885036945504795e-06, + "loss": 0.3389, + "step": 28297 + }, + { + "epoch": 0.6228923618754127, + "grad_norm": 2.097707572999872, + "learning_rate": 3.288168775700106e-06, + "loss": 0.2986, + "step": 28298 + }, + { + "epoch": 0.6229143737618313, + "grad_norm": 1.193724838725477, + "learning_rate": 3.287833865550023e-06, + "loss": 0.226, + "step": 28299 + }, + { + "epoch": 0.62293638564825, + "grad_norm": 1.6269479907698858, + "learning_rate": 3.287498964101932e-06, + "loss": 0.3313, + "step": 28300 + }, + { + "epoch": 0.6229583975346688, + "grad_norm": 1.4757037370675088, + "learning_rate": 3.2871640713575337e-06, + "loss": 0.2887, + "step": 28301 + }, + { + "epoch": 0.6229804094210873, + "grad_norm": 2.45951938272907, + "learning_rate": 3.286829187318533e-06, + "loss": 0.3388, + "step": 28302 + }, + { + "epoch": 0.623002421307506, + "grad_norm": 1.42724687586035, + "learning_rate": 3.2864943119866286e-06, + "loss": 0.3154, + "step": 28303 + }, + { + "epoch": 0.6230244331939248, + "grad_norm": 2.2287992662171363, + "learning_rate": 3.286159445363527e-06, + "loss": 0.3034, + "step": 28304 + }, + { + "epoch": 0.6230464450803433, + "grad_norm": 2.8155901777320906, + "learning_rate": 3.2858245874509263e-06, + "loss": 0.3517, + "step": 28305 + }, + { + "epoch": 0.623068456966762, + "grad_norm": 0.6222833669594482, + "learning_rate": 3.2854897382505286e-06, + "loss": 0.4509, + "step": 28306 + }, + { + "epoch": 0.6230904688531808, + "grad_norm": 1.977865557312117, + "learning_rate": 3.2851548977640378e-06, + "loss": 0.2611, + "step": 28307 + }, + { + "epoch": 0.6231124807395994, + "grad_norm": 0.6068504953760793, + "learning_rate": 3.2848200659931556e-06, + "loss": 0.4517, + "step": 28308 + }, + { + "epoch": 0.623134492626018, + "grad_norm": 0.6126072114597446, + "learning_rate": 3.2844852429395807e-06, + "loss": 0.4718, + "step": 28309 + }, + { + "epoch": 0.6231565045124368, + "grad_norm": 0.6166804817754723, + "learning_rate": 3.284150428605018e-06, + "loss": 0.4447, + "step": 28310 + }, + { + "epoch": 0.6231785163988554, + "grad_norm": 0.5984924714708243, + "learning_rate": 3.2838156229911674e-06, + "loss": 0.438, + "step": 28311 + }, + { + "epoch": 0.6232005282852741, + "grad_norm": 6.230346798289886, + "learning_rate": 3.283480826099733e-06, + "loss": 0.2924, + "step": 28312 + }, + { + "epoch": 0.6232225401716928, + "grad_norm": 2.9353852897112174, + "learning_rate": 3.283146037932413e-06, + "loss": 0.2932, + "step": 28313 + }, + { + "epoch": 0.6232445520581114, + "grad_norm": 1.4853861812864126, + "learning_rate": 3.2828112584909103e-06, + "loss": 0.3251, + "step": 28314 + }, + { + "epoch": 0.6232665639445301, + "grad_norm": 1.939445529148305, + "learning_rate": 3.2824764877769277e-06, + "loss": 0.3026, + "step": 28315 + }, + { + "epoch": 0.6232885758309488, + "grad_norm": 0.6188119422721582, + "learning_rate": 3.2821417257921646e-06, + "loss": 0.4736, + "step": 28316 + }, + { + "epoch": 0.6233105877173674, + "grad_norm": 1.5349328994330351, + "learning_rate": 3.281806972538325e-06, + "loss": 0.4054, + "step": 28317 + }, + { + "epoch": 0.6233325996037861, + "grad_norm": 3.6468810656442874, + "learning_rate": 3.2814722280171087e-06, + "loss": 0.3291, + "step": 28318 + }, + { + "epoch": 0.6233546114902047, + "grad_norm": 1.5836733604958195, + "learning_rate": 3.2811374922302154e-06, + "loss": 0.3494, + "step": 28319 + }, + { + "epoch": 0.6233766233766234, + "grad_norm": 1.6543944074125791, + "learning_rate": 3.2808027651793484e-06, + "loss": 0.2432, + "step": 28320 + }, + { + "epoch": 0.6233986352630421, + "grad_norm": 3.222086556333803, + "learning_rate": 3.2804680468662086e-06, + "loss": 0.2396, + "step": 28321 + }, + { + "epoch": 0.6234206471494607, + "grad_norm": 1.770220530430956, + "learning_rate": 3.280133337292498e-06, + "loss": 0.44, + "step": 28322 + }, + { + "epoch": 0.6234426590358794, + "grad_norm": 1.4218733715895324, + "learning_rate": 3.2797986364599165e-06, + "loss": 0.2914, + "step": 28323 + }, + { + "epoch": 0.6234646709222981, + "grad_norm": 1.4946311884421366, + "learning_rate": 3.2794639443701646e-06, + "loss": 0.2784, + "step": 28324 + }, + { + "epoch": 0.6234866828087167, + "grad_norm": 1.47792354666452, + "learning_rate": 3.2791292610249458e-06, + "loss": 0.3286, + "step": 28325 + }, + { + "epoch": 0.6235086946951354, + "grad_norm": 2.0460664518882488, + "learning_rate": 3.2787945864259596e-06, + "loss": 0.2518, + "step": 28326 + }, + { + "epoch": 0.6235307065815541, + "grad_norm": 1.6894408253557416, + "learning_rate": 3.2784599205749057e-06, + "loss": 0.2525, + "step": 28327 + }, + { + "epoch": 0.6235527184679727, + "grad_norm": 1.9951695516840122, + "learning_rate": 3.2781252634734875e-06, + "loss": 0.4204, + "step": 28328 + }, + { + "epoch": 0.6235747303543914, + "grad_norm": 2.1319329087032957, + "learning_rate": 3.277790615123403e-06, + "loss": 0.3697, + "step": 28329 + }, + { + "epoch": 0.6235967422408101, + "grad_norm": 1.3423816741197496, + "learning_rate": 3.277455975526357e-06, + "loss": 0.2506, + "step": 28330 + }, + { + "epoch": 0.6236187541272287, + "grad_norm": 1.655854163622248, + "learning_rate": 3.2771213446840467e-06, + "loss": 0.2328, + "step": 28331 + }, + { + "epoch": 0.6236407660136474, + "grad_norm": 6.762792701470714, + "learning_rate": 3.276786722598173e-06, + "loss": 0.3342, + "step": 28332 + }, + { + "epoch": 0.6236627779000661, + "grad_norm": 1.8349595312972284, + "learning_rate": 3.2764521092704395e-06, + "loss": 0.3726, + "step": 28333 + }, + { + "epoch": 0.6236847897864847, + "grad_norm": 1.4873599153034678, + "learning_rate": 3.2761175047025438e-06, + "loss": 0.2661, + "step": 28334 + }, + { + "epoch": 0.6237068016729034, + "grad_norm": 1.3284352059386095, + "learning_rate": 3.27578290889619e-06, + "loss": 0.2855, + "step": 28335 + }, + { + "epoch": 0.6237288135593221, + "grad_norm": 1.6745552966170885, + "learning_rate": 3.275448321853075e-06, + "loss": 0.3461, + "step": 28336 + }, + { + "epoch": 0.6237508254457407, + "grad_norm": 2.1391021303706905, + "learning_rate": 3.2751137435749005e-06, + "loss": 0.3354, + "step": 28337 + }, + { + "epoch": 0.6237728373321594, + "grad_norm": 2.6960996422953607, + "learning_rate": 3.274779174063368e-06, + "loss": 0.2809, + "step": 28338 + }, + { + "epoch": 0.6237948492185781, + "grad_norm": 3.8060064822361053, + "learning_rate": 3.274444613320176e-06, + "loss": 0.2934, + "step": 28339 + }, + { + "epoch": 0.6238168611049967, + "grad_norm": 2.0061333724226347, + "learning_rate": 3.274110061347028e-06, + "loss": 0.3075, + "step": 28340 + }, + { + "epoch": 0.6238388729914154, + "grad_norm": 9.127481575640198, + "learning_rate": 3.2737755181456217e-06, + "loss": 0.2676, + "step": 28341 + }, + { + "epoch": 0.623860884877834, + "grad_norm": 1.4928920578698022, + "learning_rate": 3.273440983717657e-06, + "loss": 0.2892, + "step": 28342 + }, + { + "epoch": 0.6238828967642527, + "grad_norm": 2.01008113526952, + "learning_rate": 3.2731064580648365e-06, + "loss": 0.2695, + "step": 28343 + }, + { + "epoch": 0.6239049086506714, + "grad_norm": 3.4434313438793964, + "learning_rate": 3.2727719411888583e-06, + "loss": 0.2563, + "step": 28344 + }, + { + "epoch": 0.62392692053709, + "grad_norm": 1.6565248982484941, + "learning_rate": 3.2724374330914244e-06, + "loss": 0.315, + "step": 28345 + }, + { + "epoch": 0.6239489324235087, + "grad_norm": 1.4995933146794513, + "learning_rate": 3.2721029337742338e-06, + "loss": 0.3076, + "step": 28346 + }, + { + "epoch": 0.6239709443099274, + "grad_norm": 1.9438108144353237, + "learning_rate": 3.2717684432389863e-06, + "loss": 0.3175, + "step": 28347 + }, + { + "epoch": 0.623992956196346, + "grad_norm": 0.6367678284909368, + "learning_rate": 3.2714339614873836e-06, + "loss": 0.4742, + "step": 28348 + }, + { + "epoch": 0.6240149680827647, + "grad_norm": 13.326008982382584, + "learning_rate": 3.271099488521123e-06, + "loss": 0.3658, + "step": 28349 + }, + { + "epoch": 0.6240369799691834, + "grad_norm": 1.9972012946226383, + "learning_rate": 3.2707650243419064e-06, + "loss": 0.3342, + "step": 28350 + }, + { + "epoch": 0.624058991855602, + "grad_norm": 0.610340556503931, + "learning_rate": 3.270430568951433e-06, + "loss": 0.4767, + "step": 28351 + }, + { + "epoch": 0.6240810037420207, + "grad_norm": 3.22293693365992, + "learning_rate": 3.270096122351402e-06, + "loss": 0.3134, + "step": 28352 + }, + { + "epoch": 0.6241030156284394, + "grad_norm": 1.970282206212901, + "learning_rate": 3.2697616845435164e-06, + "loss": 0.3545, + "step": 28353 + }, + { + "epoch": 0.624125027514858, + "grad_norm": 1.5816919893077719, + "learning_rate": 3.2694272555294725e-06, + "loss": 0.2937, + "step": 28354 + }, + { + "epoch": 0.6241470394012767, + "grad_norm": 1.5566894364329087, + "learning_rate": 3.269092835310971e-06, + "loss": 0.2635, + "step": 28355 + }, + { + "epoch": 0.6241690512876954, + "grad_norm": 2.1331642597159957, + "learning_rate": 3.268758423889711e-06, + "loss": 0.3714, + "step": 28356 + }, + { + "epoch": 0.624191063174114, + "grad_norm": 4.656104847295066, + "learning_rate": 3.2684240212673936e-06, + "loss": 0.2208, + "step": 28357 + }, + { + "epoch": 0.6242130750605327, + "grad_norm": 2.787032828880573, + "learning_rate": 3.2680896274457187e-06, + "loss": 0.2799, + "step": 28358 + }, + { + "epoch": 0.6242350869469514, + "grad_norm": 1.515960945977822, + "learning_rate": 3.2677552424263836e-06, + "loss": 0.3088, + "step": 28359 + }, + { + "epoch": 0.62425709883337, + "grad_norm": 6.622086171720072, + "learning_rate": 3.2674208662110877e-06, + "loss": 0.3189, + "step": 28360 + }, + { + "epoch": 0.6242791107197887, + "grad_norm": 1.602665686549416, + "learning_rate": 3.2670864988015337e-06, + "loss": 0.2699, + "step": 28361 + }, + { + "epoch": 0.6243011226062074, + "grad_norm": 0.606281045810556, + "learning_rate": 3.2667521401994172e-06, + "loss": 0.4516, + "step": 28362 + }, + { + "epoch": 0.624323134492626, + "grad_norm": 2.0213912608393216, + "learning_rate": 3.266417790406442e-06, + "loss": 0.2835, + "step": 28363 + }, + { + "epoch": 0.6243451463790447, + "grad_norm": 1.404442472905602, + "learning_rate": 3.2660834494243033e-06, + "loss": 0.2629, + "step": 28364 + }, + { + "epoch": 0.6243671582654633, + "grad_norm": 0.6108092665610463, + "learning_rate": 3.265749117254701e-06, + "loss": 0.4773, + "step": 28365 + }, + { + "epoch": 0.624389170151882, + "grad_norm": 5.5363930564947745, + "learning_rate": 3.2654147938993375e-06, + "loss": 0.2777, + "step": 28366 + }, + { + "epoch": 0.6244111820383007, + "grad_norm": 1.2496640353446098, + "learning_rate": 3.2650804793599084e-06, + "loss": 0.2234, + "step": 28367 + }, + { + "epoch": 0.6244331939247193, + "grad_norm": 1.5224623429618536, + "learning_rate": 3.2647461736381126e-06, + "loss": 0.2794, + "step": 28368 + }, + { + "epoch": 0.624455205811138, + "grad_norm": 2.1956209595521514, + "learning_rate": 3.2644118767356527e-06, + "loss": 0.2826, + "step": 28369 + }, + { + "epoch": 0.6244772176975567, + "grad_norm": 3.3038877734718732, + "learning_rate": 3.2640775886542243e-06, + "loss": 0.2875, + "step": 28370 + }, + { + "epoch": 0.6244992295839753, + "grad_norm": 0.6069416466690007, + "learning_rate": 3.2637433093955297e-06, + "loss": 0.4559, + "step": 28371 + }, + { + "epoch": 0.624521241470394, + "grad_norm": 1.3871613377302923, + "learning_rate": 3.263409038961265e-06, + "loss": 0.3315, + "step": 28372 + }, + { + "epoch": 0.6245432533568127, + "grad_norm": 1.5347138628056347, + "learning_rate": 3.263074777353129e-06, + "loss": 0.2535, + "step": 28373 + }, + { + "epoch": 0.6245652652432313, + "grad_norm": 1.5149152038702143, + "learning_rate": 3.2627405245728227e-06, + "loss": 0.3284, + "step": 28374 + }, + { + "epoch": 0.62458727712965, + "grad_norm": 2.8265666577127764, + "learning_rate": 3.262406280622043e-06, + "loss": 0.3484, + "step": 28375 + }, + { + "epoch": 0.6246092890160687, + "grad_norm": 0.6138558689449088, + "learning_rate": 3.262072045502492e-06, + "loss": 0.4802, + "step": 28376 + }, + { + "epoch": 0.6246313009024873, + "grad_norm": 1.3568605235343343, + "learning_rate": 3.2617378192158643e-06, + "loss": 0.2747, + "step": 28377 + }, + { + "epoch": 0.624653312788906, + "grad_norm": 0.6233881121596351, + "learning_rate": 3.2614036017638595e-06, + "loss": 0.4837, + "step": 28378 + }, + { + "epoch": 0.6246753246753247, + "grad_norm": 1.6740709981834732, + "learning_rate": 3.2610693931481785e-06, + "loss": 0.2762, + "step": 28379 + }, + { + "epoch": 0.6246973365617433, + "grad_norm": 1.6108989133050655, + "learning_rate": 3.260735193370517e-06, + "loss": 0.3301, + "step": 28380 + }, + { + "epoch": 0.624719348448162, + "grad_norm": 2.1695728845774687, + "learning_rate": 3.2604010024325765e-06, + "loss": 0.3139, + "step": 28381 + }, + { + "epoch": 0.6247413603345807, + "grad_norm": 3.870776329460915, + "learning_rate": 3.2600668203360542e-06, + "loss": 0.2689, + "step": 28382 + }, + { + "epoch": 0.6247633722209993, + "grad_norm": 0.6210861762567383, + "learning_rate": 3.259732647082646e-06, + "loss": 0.4637, + "step": 28383 + }, + { + "epoch": 0.624785384107418, + "grad_norm": 2.1433887772331772, + "learning_rate": 3.2593984826740556e-06, + "loss": 0.3887, + "step": 28384 + }, + { + "epoch": 0.6248073959938367, + "grad_norm": 1.3185061225375203, + "learning_rate": 3.259064327111975e-06, + "loss": 0.246, + "step": 28385 + }, + { + "epoch": 0.6248294078802553, + "grad_norm": 1.348767947936802, + "learning_rate": 3.2587301803981093e-06, + "loss": 0.2845, + "step": 28386 + }, + { + "epoch": 0.624851419766674, + "grad_norm": 1.8651553330911914, + "learning_rate": 3.258396042534152e-06, + "loss": 0.3281, + "step": 28387 + }, + { + "epoch": 0.6248734316530926, + "grad_norm": 1.9558757420726125, + "learning_rate": 3.2580619135218024e-06, + "loss": 0.3131, + "step": 28388 + }, + { + "epoch": 0.6248954435395113, + "grad_norm": 2.7750712950919767, + "learning_rate": 3.2577277933627606e-06, + "loss": 0.3223, + "step": 28389 + }, + { + "epoch": 0.62491745542593, + "grad_norm": 0.6137113369571893, + "learning_rate": 3.2573936820587228e-06, + "loss": 0.4613, + "step": 28390 + }, + { + "epoch": 0.6249394673123486, + "grad_norm": 1.3188127490849493, + "learning_rate": 3.2570595796113856e-06, + "loss": 0.2893, + "step": 28391 + }, + { + "epoch": 0.6249614791987673, + "grad_norm": 0.6168860593157264, + "learning_rate": 3.2567254860224506e-06, + "loss": 0.4705, + "step": 28392 + }, + { + "epoch": 0.624983491085186, + "grad_norm": 1.3712523962687255, + "learning_rate": 3.2563914012936133e-06, + "loss": 0.2698, + "step": 28393 + }, + { + "epoch": 0.6250055029716046, + "grad_norm": 2.7786196191475865, + "learning_rate": 3.2560573254265744e-06, + "loss": 0.3578, + "step": 28394 + }, + { + "epoch": 0.6250275148580233, + "grad_norm": 1.486982065074536, + "learning_rate": 3.2557232584230293e-06, + "loss": 0.3925, + "step": 28395 + }, + { + "epoch": 0.625049526744442, + "grad_norm": 1.4652265994720284, + "learning_rate": 3.255389200284675e-06, + "loss": 0.298, + "step": 28396 + }, + { + "epoch": 0.6250715386308606, + "grad_norm": 1.4251658457584555, + "learning_rate": 3.2550551510132133e-06, + "loss": 0.2453, + "step": 28397 + }, + { + "epoch": 0.6250935505172793, + "grad_norm": 1.8000051735760207, + "learning_rate": 3.254721110610336e-06, + "loss": 0.3044, + "step": 28398 + }, + { + "epoch": 0.625115562403698, + "grad_norm": 1.6882267275574565, + "learning_rate": 3.254387079077748e-06, + "loss": 0.2824, + "step": 28399 + }, + { + "epoch": 0.6251375742901166, + "grad_norm": 2.074642772914539, + "learning_rate": 3.2540530564171424e-06, + "loss": 0.2321, + "step": 28400 + }, + { + "epoch": 0.6251595861765353, + "grad_norm": 1.454443096881189, + "learning_rate": 3.2537190426302168e-06, + "loss": 0.3082, + "step": 28401 + }, + { + "epoch": 0.625181598062954, + "grad_norm": 1.5028966795538965, + "learning_rate": 3.2533850377186713e-06, + "loss": 0.3492, + "step": 28402 + }, + { + "epoch": 0.6252036099493726, + "grad_norm": 1.5770786469727536, + "learning_rate": 3.2530510416841998e-06, + "loss": 0.3412, + "step": 28403 + }, + { + "epoch": 0.6252256218357913, + "grad_norm": 1.8033835177945596, + "learning_rate": 3.2527170545285054e-06, + "loss": 0.2984, + "step": 28404 + }, + { + "epoch": 0.62524763372221, + "grad_norm": 0.6169214111706195, + "learning_rate": 3.2523830762532802e-06, + "loss": 0.4744, + "step": 28405 + }, + { + "epoch": 0.6252696456086286, + "grad_norm": 1.5248722824207475, + "learning_rate": 3.2520491068602234e-06, + "loss": 0.339, + "step": 28406 + }, + { + "epoch": 0.6252916574950473, + "grad_norm": 5.2486322888470225, + "learning_rate": 3.2517151463510344e-06, + "loss": 0.2806, + "step": 28407 + }, + { + "epoch": 0.625313669381466, + "grad_norm": 1.585511649548852, + "learning_rate": 3.251381194727408e-06, + "loss": 0.3028, + "step": 28408 + }, + { + "epoch": 0.6253356812678846, + "grad_norm": 1.576975151768081, + "learning_rate": 3.2510472519910407e-06, + "loss": 0.2573, + "step": 28409 + }, + { + "epoch": 0.6253576931543033, + "grad_norm": 1.2987023411044638, + "learning_rate": 3.2507133181436334e-06, + "loss": 0.2221, + "step": 28410 + }, + { + "epoch": 0.625379705040722, + "grad_norm": 1.75991096257375, + "learning_rate": 3.2503793931868787e-06, + "loss": 0.3104, + "step": 28411 + }, + { + "epoch": 0.6254017169271406, + "grad_norm": 1.8866536028152792, + "learning_rate": 3.25004547712248e-06, + "loss": 0.3678, + "step": 28412 + }, + { + "epoch": 0.6254237288135593, + "grad_norm": 2.900931625027508, + "learning_rate": 3.2497115699521285e-06, + "loss": 0.371, + "step": 28413 + }, + { + "epoch": 0.6254457406999779, + "grad_norm": 5.285488342227792, + "learning_rate": 3.2493776716775228e-06, + "loss": 0.3234, + "step": 28414 + }, + { + "epoch": 0.6254677525863966, + "grad_norm": 0.6038522979717144, + "learning_rate": 3.2490437823003625e-06, + "loss": 0.4252, + "step": 28415 + }, + { + "epoch": 0.6254897644728153, + "grad_norm": 2.5732192609151827, + "learning_rate": 3.2487099018223395e-06, + "loss": 0.2554, + "step": 28416 + }, + { + "epoch": 0.6255117763592339, + "grad_norm": 1.7094951846111677, + "learning_rate": 3.2483760302451574e-06, + "loss": 0.3433, + "step": 28417 + }, + { + "epoch": 0.6255337882456526, + "grad_norm": 1.8198656992918254, + "learning_rate": 3.2480421675705074e-06, + "loss": 0.3332, + "step": 28418 + }, + { + "epoch": 0.6255558001320713, + "grad_norm": 2.1157355995505394, + "learning_rate": 3.2477083138000885e-06, + "loss": 0.3257, + "step": 28419 + }, + { + "epoch": 0.6255778120184899, + "grad_norm": 4.12130682856557, + "learning_rate": 3.2473744689355997e-06, + "loss": 0.2407, + "step": 28420 + }, + { + "epoch": 0.6255998239049086, + "grad_norm": 0.6483741041094256, + "learning_rate": 3.2470406329787318e-06, + "loss": 0.4887, + "step": 28421 + }, + { + "epoch": 0.6256218357913274, + "grad_norm": 0.6056598117385094, + "learning_rate": 3.246706805931189e-06, + "loss": 0.447, + "step": 28422 + }, + { + "epoch": 0.625643847677746, + "grad_norm": 1.742492297169474, + "learning_rate": 3.2463729877946625e-06, + "loss": 0.2359, + "step": 28423 + }, + { + "epoch": 0.6256658595641647, + "grad_norm": 0.6301950751649052, + "learning_rate": 3.24603917857085e-06, + "loss": 0.4694, + "step": 28424 + }, + { + "epoch": 0.6256878714505834, + "grad_norm": 1.6586973986732527, + "learning_rate": 3.2457053782614506e-06, + "loss": 0.3945, + "step": 28425 + }, + { + "epoch": 0.625709883337002, + "grad_norm": 1.768976861027981, + "learning_rate": 3.245371586868158e-06, + "loss": 0.3169, + "step": 28426 + }, + { + "epoch": 0.6257318952234207, + "grad_norm": 1.7321904863306725, + "learning_rate": 3.2450378043926686e-06, + "loss": 0.3496, + "step": 28427 + }, + { + "epoch": 0.6257539071098394, + "grad_norm": 1.9255228375444333, + "learning_rate": 3.2447040308366806e-06, + "loss": 0.2606, + "step": 28428 + }, + { + "epoch": 0.625775918996258, + "grad_norm": 1.4539230724132701, + "learning_rate": 3.2443702662018883e-06, + "loss": 0.2897, + "step": 28429 + }, + { + "epoch": 0.6257979308826767, + "grad_norm": 1.9071458931857546, + "learning_rate": 3.244036510489992e-06, + "loss": 0.2849, + "step": 28430 + }, + { + "epoch": 0.6258199427690954, + "grad_norm": 1.7740709911925197, + "learning_rate": 3.243702763702684e-06, + "loss": 0.3082, + "step": 28431 + }, + { + "epoch": 0.625841954655514, + "grad_norm": 0.6077108206482493, + "learning_rate": 3.243369025841661e-06, + "loss": 0.4457, + "step": 28432 + }, + { + "epoch": 0.6258639665419327, + "grad_norm": 1.3679440001510315, + "learning_rate": 3.2430352969086222e-06, + "loss": 0.2907, + "step": 28433 + }, + { + "epoch": 0.6258859784283514, + "grad_norm": 1.9110896626808302, + "learning_rate": 3.2427015769052585e-06, + "loss": 0.2531, + "step": 28434 + }, + { + "epoch": 0.62590799031477, + "grad_norm": 0.6583293872542058, + "learning_rate": 3.2423678658332723e-06, + "loss": 0.4697, + "step": 28435 + }, + { + "epoch": 0.6259300022011887, + "grad_norm": 2.2275325473975265, + "learning_rate": 3.2420341636943552e-06, + "loss": 0.3691, + "step": 28436 + }, + { + "epoch": 0.6259520140876073, + "grad_norm": 1.4726444686950966, + "learning_rate": 3.2417004704902034e-06, + "loss": 0.3407, + "step": 28437 + }, + { + "epoch": 0.625974025974026, + "grad_norm": 6.715813767595426, + "learning_rate": 3.241366786222517e-06, + "loss": 0.3282, + "step": 28438 + }, + { + "epoch": 0.6259960378604447, + "grad_norm": 1.594957899501223, + "learning_rate": 3.241033110892985e-06, + "loss": 0.3049, + "step": 28439 + }, + { + "epoch": 0.6260180497468633, + "grad_norm": 1.6151432051483456, + "learning_rate": 3.24069944450331e-06, + "loss": 0.2334, + "step": 28440 + }, + { + "epoch": 0.626040061633282, + "grad_norm": 2.0978036564532347, + "learning_rate": 3.240365787055185e-06, + "loss": 0.2524, + "step": 28441 + }, + { + "epoch": 0.6260620735197007, + "grad_norm": 0.624544038852257, + "learning_rate": 3.240032138550305e-06, + "loss": 0.4643, + "step": 28442 + }, + { + "epoch": 0.6260840854061193, + "grad_norm": 0.6068608660895317, + "learning_rate": 3.239698498990368e-06, + "loss": 0.4611, + "step": 28443 + }, + { + "epoch": 0.626106097292538, + "grad_norm": 1.8951948029605987, + "learning_rate": 3.2393648683770663e-06, + "loss": 0.3353, + "step": 28444 + }, + { + "epoch": 0.6261281091789567, + "grad_norm": 2.3897863597699174, + "learning_rate": 3.2390312467120988e-06, + "loss": 0.3269, + "step": 28445 + }, + { + "epoch": 0.6261501210653753, + "grad_norm": 1.5724261482541992, + "learning_rate": 3.2386976339971604e-06, + "loss": 0.3085, + "step": 28446 + }, + { + "epoch": 0.626172132951794, + "grad_norm": 1.7277202048623632, + "learning_rate": 3.238364030233945e-06, + "loss": 0.4167, + "step": 28447 + }, + { + "epoch": 0.6261941448382127, + "grad_norm": 2.1370417381050273, + "learning_rate": 3.238030435424151e-06, + "loss": 0.3396, + "step": 28448 + }, + { + "epoch": 0.6262161567246313, + "grad_norm": 4.054928015061801, + "learning_rate": 3.237696849569471e-06, + "loss": 0.3207, + "step": 28449 + }, + { + "epoch": 0.62623816861105, + "grad_norm": 2.0898177331374046, + "learning_rate": 3.237363272671601e-06, + "loss": 0.3319, + "step": 28450 + }, + { + "epoch": 0.6262601804974687, + "grad_norm": 1.7382610663897937, + "learning_rate": 3.2370297047322387e-06, + "loss": 0.3496, + "step": 28451 + }, + { + "epoch": 0.6262821923838873, + "grad_norm": 1.920494918876322, + "learning_rate": 3.2366961457530747e-06, + "loss": 0.3403, + "step": 28452 + }, + { + "epoch": 0.626304204270306, + "grad_norm": 1.3782226688385821, + "learning_rate": 3.2363625957358107e-06, + "loss": 0.2975, + "step": 28453 + }, + { + "epoch": 0.6263262161567247, + "grad_norm": 0.6470691471546267, + "learning_rate": 3.236029054682138e-06, + "loss": 0.4602, + "step": 28454 + }, + { + "epoch": 0.6263482280431433, + "grad_norm": 1.8209715063031913, + "learning_rate": 3.2356955225937505e-06, + "loss": 0.3234, + "step": 28455 + }, + { + "epoch": 0.626370239929562, + "grad_norm": 1.101863924350601, + "learning_rate": 3.2353619994723474e-06, + "loss": 0.2224, + "step": 28456 + }, + { + "epoch": 0.6263922518159807, + "grad_norm": 1.6542358393584968, + "learning_rate": 3.235028485319619e-06, + "loss": 0.315, + "step": 28457 + }, + { + "epoch": 0.6264142637023993, + "grad_norm": 1.5099976303422575, + "learning_rate": 3.234694980137266e-06, + "loss": 0.3027, + "step": 28458 + }, + { + "epoch": 0.626436275588818, + "grad_norm": 2.43310485794587, + "learning_rate": 3.2343614839269793e-06, + "loss": 0.2996, + "step": 28459 + }, + { + "epoch": 0.6264582874752366, + "grad_norm": 0.6184702476037174, + "learning_rate": 3.234027996690454e-06, + "loss": 0.4656, + "step": 28460 + }, + { + "epoch": 0.6264802993616553, + "grad_norm": 2.6552454674538493, + "learning_rate": 3.2336945184293884e-06, + "loss": 0.3555, + "step": 28461 + }, + { + "epoch": 0.626502311248074, + "grad_norm": 1.6596684745770367, + "learning_rate": 3.2333610491454727e-06, + "loss": 0.2728, + "step": 28462 + }, + { + "epoch": 0.6265243231344926, + "grad_norm": 2.9793130339334497, + "learning_rate": 3.233027588840405e-06, + "loss": 0.3241, + "step": 28463 + }, + { + "epoch": 0.6265463350209113, + "grad_norm": 0.6245510344564801, + "learning_rate": 3.2326941375158797e-06, + "loss": 0.4512, + "step": 28464 + }, + { + "epoch": 0.62656834690733, + "grad_norm": 1.4681210144112864, + "learning_rate": 3.2323606951735897e-06, + "loss": 0.2351, + "step": 28465 + }, + { + "epoch": 0.6265903587937486, + "grad_norm": 2.779270061957728, + "learning_rate": 3.232027261815233e-06, + "loss": 0.3091, + "step": 28466 + }, + { + "epoch": 0.6266123706801673, + "grad_norm": 0.619848315171748, + "learning_rate": 3.2316938374425018e-06, + "loss": 0.4785, + "step": 28467 + }, + { + "epoch": 0.626634382566586, + "grad_norm": 1.8327929153723446, + "learning_rate": 3.2313604220570895e-06, + "loss": 0.2546, + "step": 28468 + }, + { + "epoch": 0.6266563944530046, + "grad_norm": 0.6143461643733895, + "learning_rate": 3.231027015660695e-06, + "loss": 0.4888, + "step": 28469 + }, + { + "epoch": 0.6266784063394233, + "grad_norm": 1.9470685372434045, + "learning_rate": 3.2306936182550062e-06, + "loss": 0.268, + "step": 28470 + }, + { + "epoch": 0.626700418225842, + "grad_norm": 1.3253673967289115, + "learning_rate": 3.230360229841726e-06, + "loss": 0.2553, + "step": 28471 + }, + { + "epoch": 0.6267224301122606, + "grad_norm": 2.4254415700269796, + "learning_rate": 3.2300268504225417e-06, + "loss": 0.2793, + "step": 28472 + }, + { + "epoch": 0.6267444419986793, + "grad_norm": 1.6162628783976836, + "learning_rate": 3.2296934799991507e-06, + "loss": 0.3939, + "step": 28473 + }, + { + "epoch": 0.626766453885098, + "grad_norm": 2.934697975284582, + "learning_rate": 3.2293601185732485e-06, + "loss": 0.3325, + "step": 28474 + }, + { + "epoch": 0.6267884657715166, + "grad_norm": 4.203863545520323, + "learning_rate": 3.229026766146525e-06, + "loss": 0.2553, + "step": 28475 + }, + { + "epoch": 0.6268104776579353, + "grad_norm": 1.7698494956699673, + "learning_rate": 3.22869342272068e-06, + "loss": 0.2912, + "step": 28476 + }, + { + "epoch": 0.626832489544354, + "grad_norm": 1.9469836941447125, + "learning_rate": 3.2283600882974046e-06, + "loss": 0.2893, + "step": 28477 + }, + { + "epoch": 0.6268545014307726, + "grad_norm": 2.247739193316068, + "learning_rate": 3.2280267628783924e-06, + "loss": 0.3523, + "step": 28478 + }, + { + "epoch": 0.6268765133171913, + "grad_norm": 1.9238174033911886, + "learning_rate": 3.22769344646534e-06, + "loss": 0.2719, + "step": 28479 + }, + { + "epoch": 0.62689852520361, + "grad_norm": 1.5154630227617911, + "learning_rate": 3.2273601390599383e-06, + "loss": 0.2809, + "step": 28480 + }, + { + "epoch": 0.6269205370900286, + "grad_norm": 3.014918137941291, + "learning_rate": 3.2270268406638842e-06, + "loss": 0.4131, + "step": 28481 + }, + { + "epoch": 0.6269425489764473, + "grad_norm": 3.836959020664945, + "learning_rate": 3.2266935512788703e-06, + "loss": 0.3018, + "step": 28482 + }, + { + "epoch": 0.6269645608628659, + "grad_norm": 2.0778533264149543, + "learning_rate": 3.22636027090659e-06, + "loss": 0.3874, + "step": 28483 + }, + { + "epoch": 0.6269865727492846, + "grad_norm": 1.41797732072489, + "learning_rate": 3.226026999548739e-06, + "loss": 0.305, + "step": 28484 + }, + { + "epoch": 0.6270085846357033, + "grad_norm": 0.6272137929002146, + "learning_rate": 3.2256937372070097e-06, + "loss": 0.466, + "step": 28485 + }, + { + "epoch": 0.6270305965221219, + "grad_norm": 1.70212518698198, + "learning_rate": 3.2253604838830954e-06, + "loss": 0.3357, + "step": 28486 + }, + { + "epoch": 0.6270526084085406, + "grad_norm": 1.4848030691503613, + "learning_rate": 3.225027239578692e-06, + "loss": 0.3279, + "step": 28487 + }, + { + "epoch": 0.6270746202949593, + "grad_norm": 1.5987961475842263, + "learning_rate": 3.224694004295489e-06, + "loss": 0.2661, + "step": 28488 + }, + { + "epoch": 0.6270966321813779, + "grad_norm": 1.7961310317298338, + "learning_rate": 3.224360778035186e-06, + "loss": 0.3035, + "step": 28489 + }, + { + "epoch": 0.6271186440677966, + "grad_norm": 2.1482031559760504, + "learning_rate": 3.224027560799472e-06, + "loss": 0.2902, + "step": 28490 + }, + { + "epoch": 0.6271406559542153, + "grad_norm": 2.0119375536221624, + "learning_rate": 3.2236943525900423e-06, + "loss": 0.3303, + "step": 28491 + }, + { + "epoch": 0.6271626678406339, + "grad_norm": 2.4288834573789613, + "learning_rate": 3.223361153408591e-06, + "loss": 0.3204, + "step": 28492 + }, + { + "epoch": 0.6271846797270526, + "grad_norm": 1.9509937298272537, + "learning_rate": 3.223027963256808e-06, + "loss": 0.2916, + "step": 28493 + }, + { + "epoch": 0.6272066916134713, + "grad_norm": 1.6461163843465516, + "learning_rate": 3.2226947821363926e-06, + "loss": 0.2801, + "step": 28494 + }, + { + "epoch": 0.6272287034998899, + "grad_norm": 2.397908422724208, + "learning_rate": 3.222361610049034e-06, + "loss": 0.2998, + "step": 28495 + }, + { + "epoch": 0.6272507153863086, + "grad_norm": 0.5954620745421048, + "learning_rate": 3.222028446996425e-06, + "loss": 0.4746, + "step": 28496 + }, + { + "epoch": 0.6272727272727273, + "grad_norm": 1.7593745361479955, + "learning_rate": 3.221695292980263e-06, + "loss": 0.3538, + "step": 28497 + }, + { + "epoch": 0.6272947391591459, + "grad_norm": 0.6070890716983203, + "learning_rate": 3.2213621480022363e-06, + "loss": 0.4689, + "step": 28498 + }, + { + "epoch": 0.6273167510455646, + "grad_norm": 1.1942652059398513, + "learning_rate": 3.221029012064042e-06, + "loss": 0.222, + "step": 28499 + }, + { + "epoch": 0.6273387629319833, + "grad_norm": 3.564879968675019, + "learning_rate": 3.2206958851673713e-06, + "loss": 0.3399, + "step": 28500 + }, + { + "epoch": 0.6273607748184019, + "grad_norm": 2.22356583849697, + "learning_rate": 3.220362767313917e-06, + "loss": 0.373, + "step": 28501 + }, + { + "epoch": 0.6273827867048206, + "grad_norm": 2.6900308415775056, + "learning_rate": 3.220029658505375e-06, + "loss": 0.3196, + "step": 28502 + }, + { + "epoch": 0.6274047985912393, + "grad_norm": 1.660838456878751, + "learning_rate": 3.2196965587434335e-06, + "loss": 0.314, + "step": 28503 + }, + { + "epoch": 0.6274268104776579, + "grad_norm": 1.1538833155719148, + "learning_rate": 3.219363468029789e-06, + "loss": 0.2971, + "step": 28504 + }, + { + "epoch": 0.6274488223640766, + "grad_norm": 1.8325029123144663, + "learning_rate": 3.2190303863661354e-06, + "loss": 0.2767, + "step": 28505 + }, + { + "epoch": 0.6274708342504952, + "grad_norm": 1.6224014090641656, + "learning_rate": 3.21869731375416e-06, + "loss": 0.3352, + "step": 28506 + }, + { + "epoch": 0.6274928461369139, + "grad_norm": 1.4506969096507416, + "learning_rate": 3.2183642501955626e-06, + "loss": 0.2063, + "step": 28507 + }, + { + "epoch": 0.6275148580233326, + "grad_norm": 4.0290634601028295, + "learning_rate": 3.218031195692032e-06, + "loss": 0.3053, + "step": 28508 + }, + { + "epoch": 0.6275368699097512, + "grad_norm": 2.1360823321633093, + "learning_rate": 3.2176981502452605e-06, + "loss": 0.2591, + "step": 28509 + }, + { + "epoch": 0.6275588817961699, + "grad_norm": 2.008847647835037, + "learning_rate": 3.217365113856944e-06, + "loss": 0.3102, + "step": 28510 + }, + { + "epoch": 0.6275808936825886, + "grad_norm": 2.3901855502634732, + "learning_rate": 3.2170320865287696e-06, + "loss": 0.3813, + "step": 28511 + }, + { + "epoch": 0.6276029055690072, + "grad_norm": 1.9076291305717803, + "learning_rate": 3.2166990682624367e-06, + "loss": 0.299, + "step": 28512 + }, + { + "epoch": 0.6276249174554259, + "grad_norm": 1.689346111957165, + "learning_rate": 3.2163660590596337e-06, + "loss": 0.2769, + "step": 28513 + }, + { + "epoch": 0.6276469293418446, + "grad_norm": 1.4374772355850112, + "learning_rate": 3.216033058922052e-06, + "loss": 0.2403, + "step": 28514 + }, + { + "epoch": 0.6276689412282632, + "grad_norm": 0.621897881072273, + "learning_rate": 3.2157000678513896e-06, + "loss": 0.4571, + "step": 28515 + }, + { + "epoch": 0.6276909531146819, + "grad_norm": 1.7083413880786242, + "learning_rate": 3.2153670858493315e-06, + "loss": 0.2996, + "step": 28516 + }, + { + "epoch": 0.6277129650011006, + "grad_norm": 0.6224835736423875, + "learning_rate": 3.2150341129175753e-06, + "loss": 0.4651, + "step": 28517 + }, + { + "epoch": 0.6277349768875192, + "grad_norm": 1.3846963152377285, + "learning_rate": 3.2147011490578127e-06, + "loss": 0.3637, + "step": 28518 + }, + { + "epoch": 0.6277569887739379, + "grad_norm": 1.2388059355282033, + "learning_rate": 3.214368194271734e-06, + "loss": 0.2769, + "step": 28519 + }, + { + "epoch": 0.6277790006603566, + "grad_norm": 0.6364243236570322, + "learning_rate": 3.2140352485610336e-06, + "loss": 0.4668, + "step": 28520 + }, + { + "epoch": 0.6278010125467752, + "grad_norm": 1.6150982476661118, + "learning_rate": 3.2137023119274014e-06, + "loss": 0.3045, + "step": 28521 + }, + { + "epoch": 0.6278230244331939, + "grad_norm": 1.870395758960454, + "learning_rate": 3.2133693843725318e-06, + "loss": 0.3687, + "step": 28522 + }, + { + "epoch": 0.6278450363196126, + "grad_norm": 2.5075969399225544, + "learning_rate": 3.2130364658981167e-06, + "loss": 0.3225, + "step": 28523 + }, + { + "epoch": 0.6278670482060312, + "grad_norm": 1.6909661161421723, + "learning_rate": 3.2127035565058438e-06, + "loss": 0.3207, + "step": 28524 + }, + { + "epoch": 0.6278890600924499, + "grad_norm": 1.9547275974260632, + "learning_rate": 3.2123706561974123e-06, + "loss": 0.3038, + "step": 28525 + }, + { + "epoch": 0.6279110719788686, + "grad_norm": 1.8098790365255044, + "learning_rate": 3.2120377649745094e-06, + "loss": 0.3087, + "step": 28526 + }, + { + "epoch": 0.6279330838652872, + "grad_norm": 5.972097933895802, + "learning_rate": 3.2117048828388263e-06, + "loss": 0.2874, + "step": 28527 + }, + { + "epoch": 0.6279550957517059, + "grad_norm": 0.6034188896289834, + "learning_rate": 3.211372009792059e-06, + "loss": 0.4403, + "step": 28528 + }, + { + "epoch": 0.6279771076381245, + "grad_norm": 1.834368387781794, + "learning_rate": 3.211039145835895e-06, + "loss": 0.3089, + "step": 28529 + }, + { + "epoch": 0.6279991195245432, + "grad_norm": 0.6187437983731094, + "learning_rate": 3.210706290972029e-06, + "loss": 0.4585, + "step": 28530 + }, + { + "epoch": 0.6280211314109619, + "grad_norm": 4.5009246059907655, + "learning_rate": 3.210373445202152e-06, + "loss": 0.212, + "step": 28531 + }, + { + "epoch": 0.6280431432973805, + "grad_norm": 1.5785815405613597, + "learning_rate": 3.210040608527953e-06, + "loss": 0.2649, + "step": 28532 + }, + { + "epoch": 0.6280651551837992, + "grad_norm": 3.130735996799952, + "learning_rate": 3.2097077809511297e-06, + "loss": 0.2884, + "step": 28533 + }, + { + "epoch": 0.6280871670702179, + "grad_norm": 0.6072606711423244, + "learning_rate": 3.209374962473367e-06, + "loss": 0.4542, + "step": 28534 + }, + { + "epoch": 0.6281091789566365, + "grad_norm": 23.323091142119463, + "learning_rate": 3.2090421530963606e-06, + "loss": 0.3019, + "step": 28535 + }, + { + "epoch": 0.6281311908430552, + "grad_norm": 2.084360335654959, + "learning_rate": 3.2087093528218006e-06, + "loss": 0.301, + "step": 28536 + }, + { + "epoch": 0.628153202729474, + "grad_norm": 0.608293030041909, + "learning_rate": 3.2083765616513773e-06, + "loss": 0.4761, + "step": 28537 + }, + { + "epoch": 0.6281752146158925, + "grad_norm": 1.4902595791563726, + "learning_rate": 3.208043779586786e-06, + "loss": 0.2243, + "step": 28538 + }, + { + "epoch": 0.6281972265023112, + "grad_norm": 2.1649552548725746, + "learning_rate": 3.2077110066297136e-06, + "loss": 0.3276, + "step": 28539 + }, + { + "epoch": 0.62821923838873, + "grad_norm": 1.437739528430974, + "learning_rate": 3.2073782427818534e-06, + "loss": 0.2876, + "step": 28540 + }, + { + "epoch": 0.6282412502751485, + "grad_norm": 0.6149263301703389, + "learning_rate": 3.2070454880448985e-06, + "loss": 0.4451, + "step": 28541 + }, + { + "epoch": 0.6282632621615672, + "grad_norm": 0.5938942857594444, + "learning_rate": 3.2067127424205345e-06, + "loss": 0.4371, + "step": 28542 + }, + { + "epoch": 0.628285274047986, + "grad_norm": 1.983155001794, + "learning_rate": 3.2063800059104593e-06, + "loss": 0.3206, + "step": 28543 + }, + { + "epoch": 0.6283072859344045, + "grad_norm": 2.0778504634498405, + "learning_rate": 3.206047278516359e-06, + "loss": 0.3519, + "step": 28544 + }, + { + "epoch": 0.6283292978208233, + "grad_norm": 2.691879747275707, + "learning_rate": 3.2057145602399275e-06, + "loss": 0.3234, + "step": 28545 + }, + { + "epoch": 0.628351309707242, + "grad_norm": 3.836315902914743, + "learning_rate": 3.2053818510828557e-06, + "loss": 0.2801, + "step": 28546 + }, + { + "epoch": 0.6283733215936605, + "grad_norm": 1.7845996020689234, + "learning_rate": 3.205049151046832e-06, + "loss": 0.3162, + "step": 28547 + }, + { + "epoch": 0.6283953334800793, + "grad_norm": 1.5912453405425877, + "learning_rate": 3.20471646013355e-06, + "loss": 0.2902, + "step": 28548 + }, + { + "epoch": 0.628417345366498, + "grad_norm": 1.8610928679208831, + "learning_rate": 3.2043837783446996e-06, + "loss": 0.3018, + "step": 28549 + }, + { + "epoch": 0.6284393572529166, + "grad_norm": 1.6583280940145524, + "learning_rate": 3.20405110568197e-06, + "loss": 0.2135, + "step": 28550 + }, + { + "epoch": 0.6284613691393353, + "grad_norm": 2.3279273334240975, + "learning_rate": 3.203718442147057e-06, + "loss": 0.3558, + "step": 28551 + }, + { + "epoch": 0.628483381025754, + "grad_norm": 1.6444345547823738, + "learning_rate": 3.203385787741645e-06, + "loss": 0.3458, + "step": 28552 + }, + { + "epoch": 0.6285053929121726, + "grad_norm": 1.4214004056402436, + "learning_rate": 3.2030531424674296e-06, + "loss": 0.2868, + "step": 28553 + }, + { + "epoch": 0.6285274047985913, + "grad_norm": 2.3856301724711466, + "learning_rate": 3.202720506326099e-06, + "loss": 0.3991, + "step": 28554 + }, + { + "epoch": 0.6285494166850099, + "grad_norm": 1.724909450433388, + "learning_rate": 3.2023878793193435e-06, + "loss": 0.296, + "step": 28555 + }, + { + "epoch": 0.6285714285714286, + "grad_norm": 1.681758962467825, + "learning_rate": 3.2020552614488563e-06, + "loss": 0.307, + "step": 28556 + }, + { + "epoch": 0.6285934404578473, + "grad_norm": 2.461715749073692, + "learning_rate": 3.201722652716325e-06, + "loss": 0.3163, + "step": 28557 + }, + { + "epoch": 0.6286154523442659, + "grad_norm": 2.320696683756697, + "learning_rate": 3.201390053123442e-06, + "loss": 0.2999, + "step": 28558 + }, + { + "epoch": 0.6286374642306846, + "grad_norm": 0.60901469011239, + "learning_rate": 3.2010574626718972e-06, + "loss": 0.4552, + "step": 28559 + }, + { + "epoch": 0.6286594761171033, + "grad_norm": 2.135047001194181, + "learning_rate": 3.2007248813633785e-06, + "loss": 0.3165, + "step": 28560 + }, + { + "epoch": 0.6286814880035219, + "grad_norm": 1.865412171673514, + "learning_rate": 3.2003923091995815e-06, + "loss": 0.323, + "step": 28561 + }, + { + "epoch": 0.6287034998899406, + "grad_norm": 1.6532505067418688, + "learning_rate": 3.2000597461821904e-06, + "loss": 0.3503, + "step": 28562 + }, + { + "epoch": 0.6287255117763593, + "grad_norm": 1.5976073835185705, + "learning_rate": 3.1997271923129003e-06, + "loss": 0.2963, + "step": 28563 + }, + { + "epoch": 0.6287475236627779, + "grad_norm": 2.008176615797257, + "learning_rate": 3.1993946475934008e-06, + "loss": 0.3416, + "step": 28564 + }, + { + "epoch": 0.6287695355491966, + "grad_norm": 1.2743014037905744, + "learning_rate": 3.1990621120253786e-06, + "loss": 0.3, + "step": 28565 + }, + { + "epoch": 0.6287915474356153, + "grad_norm": 4.167398374784832, + "learning_rate": 3.198729585610526e-06, + "loss": 0.2553, + "step": 28566 + }, + { + "epoch": 0.6288135593220339, + "grad_norm": 1.421947564885693, + "learning_rate": 3.1983970683505337e-06, + "loss": 0.2337, + "step": 28567 + }, + { + "epoch": 0.6288355712084526, + "grad_norm": 1.8909492489119444, + "learning_rate": 3.1980645602470906e-06, + "loss": 0.2945, + "step": 28568 + }, + { + "epoch": 0.6288575830948713, + "grad_norm": 1.593625735441577, + "learning_rate": 3.1977320613018894e-06, + "loss": 0.2593, + "step": 28569 + }, + { + "epoch": 0.6288795949812899, + "grad_norm": 1.5165639895011847, + "learning_rate": 3.1973995715166138e-06, + "loss": 0.39, + "step": 28570 + }, + { + "epoch": 0.6289016068677086, + "grad_norm": 2.0044756870992235, + "learning_rate": 3.19706709089296e-06, + "loss": 0.3089, + "step": 28571 + }, + { + "epoch": 0.6289236187541273, + "grad_norm": 2.1569863606688573, + "learning_rate": 3.1967346194326154e-06, + "loss": 0.2961, + "step": 28572 + }, + { + "epoch": 0.6289456306405459, + "grad_norm": 1.5260755581169898, + "learning_rate": 3.196402157137268e-06, + "loss": 0.3381, + "step": 28573 + }, + { + "epoch": 0.6289676425269646, + "grad_norm": 2.2064904865972585, + "learning_rate": 3.1960697040086118e-06, + "loss": 0.3164, + "step": 28574 + }, + { + "epoch": 0.6289896544133833, + "grad_norm": 0.6366671972893859, + "learning_rate": 3.1957372600483323e-06, + "loss": 0.4807, + "step": 28575 + }, + { + "epoch": 0.6290116662998019, + "grad_norm": 1.821079418726967, + "learning_rate": 3.195404825258121e-06, + "loss": 0.2753, + "step": 28576 + }, + { + "epoch": 0.6290336781862206, + "grad_norm": 1.5429533928921746, + "learning_rate": 3.1950723996396687e-06, + "loss": 0.2808, + "step": 28577 + }, + { + "epoch": 0.6290556900726392, + "grad_norm": 2.333361126056619, + "learning_rate": 3.19473998319466e-06, + "loss": 0.2061, + "step": 28578 + }, + { + "epoch": 0.6290777019590579, + "grad_norm": 1.8045219616764359, + "learning_rate": 3.1944075759247918e-06, + "loss": 0.3227, + "step": 28579 + }, + { + "epoch": 0.6290997138454766, + "grad_norm": 1.9665142768411554, + "learning_rate": 3.194075177831747e-06, + "loss": 0.3829, + "step": 28580 + }, + { + "epoch": 0.6291217257318952, + "grad_norm": 1.7490655141820803, + "learning_rate": 3.1937427889172183e-06, + "loss": 0.3394, + "step": 28581 + }, + { + "epoch": 0.6291437376183139, + "grad_norm": 0.6076686945431178, + "learning_rate": 3.1934104091828953e-06, + "loss": 0.4561, + "step": 28582 + }, + { + "epoch": 0.6291657495047326, + "grad_norm": 3.278245661112386, + "learning_rate": 3.193078038630465e-06, + "loss": 0.2152, + "step": 28583 + }, + { + "epoch": 0.6291877613911512, + "grad_norm": 1.655566772520746, + "learning_rate": 3.1927456772616196e-06, + "loss": 0.3399, + "step": 28584 + }, + { + "epoch": 0.6292097732775699, + "grad_norm": 2.730738382586957, + "learning_rate": 3.1924133250780453e-06, + "loss": 0.3933, + "step": 28585 + }, + { + "epoch": 0.6292317851639886, + "grad_norm": 3.156565360636199, + "learning_rate": 3.1920809820814326e-06, + "loss": 0.3641, + "step": 28586 + }, + { + "epoch": 0.6292537970504072, + "grad_norm": 2.508373069121847, + "learning_rate": 3.1917486482734724e-06, + "loss": 0.311, + "step": 28587 + }, + { + "epoch": 0.6292758089368259, + "grad_norm": 1.3334576600719852, + "learning_rate": 3.1914163236558496e-06, + "loss": 0.2799, + "step": 28588 + }, + { + "epoch": 0.6292978208232446, + "grad_norm": 0.649272203853332, + "learning_rate": 3.1910840082302573e-06, + "loss": 0.4644, + "step": 28589 + }, + { + "epoch": 0.6293198327096632, + "grad_norm": 1.5610468148567456, + "learning_rate": 3.1907517019983825e-06, + "loss": 0.3192, + "step": 28590 + }, + { + "epoch": 0.6293418445960819, + "grad_norm": 2.338458167497602, + "learning_rate": 3.190419404961914e-06, + "loss": 0.3576, + "step": 28591 + }, + { + "epoch": 0.6293638564825006, + "grad_norm": 1.4763026341870429, + "learning_rate": 3.1900871171225436e-06, + "loss": 0.3028, + "step": 28592 + }, + { + "epoch": 0.6293858683689192, + "grad_norm": 0.6047531543434195, + "learning_rate": 3.1897548384819543e-06, + "loss": 0.4506, + "step": 28593 + }, + { + "epoch": 0.6294078802553379, + "grad_norm": 1.3527179400705662, + "learning_rate": 3.18942256904184e-06, + "loss": 0.3116, + "step": 28594 + }, + { + "epoch": 0.6294298921417566, + "grad_norm": 1.50736891104569, + "learning_rate": 3.1890903088038894e-06, + "loss": 0.365, + "step": 28595 + }, + { + "epoch": 0.6294519040281752, + "grad_norm": 2.4928681442951315, + "learning_rate": 3.1887580577697853e-06, + "loss": 0.268, + "step": 28596 + }, + { + "epoch": 0.6294739159145939, + "grad_norm": 1.8133976232589868, + "learning_rate": 3.188425815941224e-06, + "loss": 0.2817, + "step": 28597 + }, + { + "epoch": 0.6294959278010126, + "grad_norm": 1.8437777409845044, + "learning_rate": 3.188093583319889e-06, + "loss": 0.2621, + "step": 28598 + }, + { + "epoch": 0.6295179396874312, + "grad_norm": 2.237334727195444, + "learning_rate": 3.187761359907472e-06, + "loss": 0.3051, + "step": 28599 + }, + { + "epoch": 0.6295399515738499, + "grad_norm": 3.3303095993337455, + "learning_rate": 3.1874291457056605e-06, + "loss": 0.2978, + "step": 28600 + }, + { + "epoch": 0.6295619634602685, + "grad_norm": 0.6138491491693211, + "learning_rate": 3.18709694071614e-06, + "loss": 0.4697, + "step": 28601 + }, + { + "epoch": 0.6295839753466872, + "grad_norm": 2.2256957447672883, + "learning_rate": 3.186764744940603e-06, + "loss": 0.3047, + "step": 28602 + }, + { + "epoch": 0.6296059872331059, + "grad_norm": 1.4636401657329698, + "learning_rate": 3.1864325583807352e-06, + "loss": 0.2353, + "step": 28603 + }, + { + "epoch": 0.6296279991195245, + "grad_norm": 1.8686546264341184, + "learning_rate": 3.1861003810382273e-06, + "loss": 0.3661, + "step": 28604 + }, + { + "epoch": 0.6296500110059432, + "grad_norm": 2.503109084891187, + "learning_rate": 3.185768212914767e-06, + "loss": 0.3533, + "step": 28605 + }, + { + "epoch": 0.6296720228923619, + "grad_norm": 3.8668311861437483, + "learning_rate": 3.18543605401204e-06, + "loss": 0.2708, + "step": 28606 + }, + { + "epoch": 0.6296940347787805, + "grad_norm": 2.6082319656820427, + "learning_rate": 3.1851039043317366e-06, + "loss": 0.3616, + "step": 28607 + }, + { + "epoch": 0.6297160466651992, + "grad_norm": 0.6145452609790573, + "learning_rate": 3.184771763875545e-06, + "loss": 0.457, + "step": 28608 + }, + { + "epoch": 0.6297380585516179, + "grad_norm": 1.4718572940626535, + "learning_rate": 3.184439632645152e-06, + "loss": 0.3182, + "step": 28609 + }, + { + "epoch": 0.6297600704380365, + "grad_norm": 7.94956596112036, + "learning_rate": 3.1841075106422492e-06, + "loss": 0.2497, + "step": 28610 + }, + { + "epoch": 0.6297820823244552, + "grad_norm": 1.7847807867137873, + "learning_rate": 3.183775397868519e-06, + "loss": 0.2186, + "step": 28611 + }, + { + "epoch": 0.6298040942108739, + "grad_norm": 1.5232332182203985, + "learning_rate": 3.1834432943256537e-06, + "loss": 0.2994, + "step": 28612 + }, + { + "epoch": 0.6298261060972925, + "grad_norm": 1.8611632695821554, + "learning_rate": 3.1831112000153407e-06, + "loss": 0.2841, + "step": 28613 + }, + { + "epoch": 0.6298481179837112, + "grad_norm": 3.210322215779207, + "learning_rate": 3.182779114939264e-06, + "loss": 0.2454, + "step": 28614 + }, + { + "epoch": 0.6298701298701299, + "grad_norm": 1.8150683953160671, + "learning_rate": 3.1824470390991167e-06, + "loss": 0.2753, + "step": 28615 + }, + { + "epoch": 0.6298921417565485, + "grad_norm": 2.105107302298326, + "learning_rate": 3.1821149724965833e-06, + "loss": 0.3365, + "step": 28616 + }, + { + "epoch": 0.6299141536429672, + "grad_norm": 1.715390342108122, + "learning_rate": 3.1817829151333528e-06, + "loss": 0.3131, + "step": 28617 + }, + { + "epoch": 0.6299361655293859, + "grad_norm": 3.9480084220622613, + "learning_rate": 3.181450867011113e-06, + "loss": 0.2797, + "step": 28618 + }, + { + "epoch": 0.6299581774158045, + "grad_norm": 2.099946176773421, + "learning_rate": 3.18111882813155e-06, + "loss": 0.2906, + "step": 28619 + }, + { + "epoch": 0.6299801893022232, + "grad_norm": 1.350771113964039, + "learning_rate": 3.180786798496353e-06, + "loss": 0.2501, + "step": 28620 + }, + { + "epoch": 0.6300022011886419, + "grad_norm": 0.6266266640399392, + "learning_rate": 3.180454778107208e-06, + "loss": 0.4854, + "step": 28621 + }, + { + "epoch": 0.6300242130750605, + "grad_norm": 0.6024706001118465, + "learning_rate": 3.180122766965804e-06, + "loss": 0.4346, + "step": 28622 + }, + { + "epoch": 0.6300462249614792, + "grad_norm": 1.7685575318497302, + "learning_rate": 3.179790765073829e-06, + "loss": 0.248, + "step": 28623 + }, + { + "epoch": 0.6300682368478978, + "grad_norm": 0.6124497573550801, + "learning_rate": 3.1794587724329674e-06, + "loss": 0.4705, + "step": 28624 + }, + { + "epoch": 0.6300902487343165, + "grad_norm": 3.2211801331056584, + "learning_rate": 3.1791267890449083e-06, + "loss": 0.299, + "step": 28625 + }, + { + "epoch": 0.6301122606207352, + "grad_norm": 1.516949113559457, + "learning_rate": 3.1787948149113414e-06, + "loss": 0.3169, + "step": 28626 + }, + { + "epoch": 0.6301342725071538, + "grad_norm": 2.3758019900186587, + "learning_rate": 3.1784628500339477e-06, + "loss": 0.3649, + "step": 28627 + }, + { + "epoch": 0.6301562843935725, + "grad_norm": 2.6808716410401874, + "learning_rate": 3.178130894414421e-06, + "loss": 0.3283, + "step": 28628 + }, + { + "epoch": 0.6301782962799912, + "grad_norm": 0.5987700651579937, + "learning_rate": 3.177798948054444e-06, + "loss": 0.4462, + "step": 28629 + }, + { + "epoch": 0.6302003081664098, + "grad_norm": 2.2982436323687963, + "learning_rate": 3.177467010955707e-06, + "loss": 0.3192, + "step": 28630 + }, + { + "epoch": 0.6302223200528285, + "grad_norm": 1.6025321702561859, + "learning_rate": 3.1771350831198953e-06, + "loss": 0.2341, + "step": 28631 + }, + { + "epoch": 0.6302443319392472, + "grad_norm": 1.73650628960953, + "learning_rate": 3.1768031645486942e-06, + "loss": 0.3297, + "step": 28632 + }, + { + "epoch": 0.6302663438256658, + "grad_norm": 0.6060043760666091, + "learning_rate": 3.1764712552437956e-06, + "loss": 0.4499, + "step": 28633 + }, + { + "epoch": 0.6302883557120845, + "grad_norm": 6.495648158990218, + "learning_rate": 3.1761393552068807e-06, + "loss": 0.3728, + "step": 28634 + }, + { + "epoch": 0.6303103675985032, + "grad_norm": 1.7988806140557019, + "learning_rate": 3.1758074644396407e-06, + "loss": 0.2971, + "step": 28635 + }, + { + "epoch": 0.6303323794849218, + "grad_norm": 0.5961052683267588, + "learning_rate": 3.1754755829437612e-06, + "loss": 0.4579, + "step": 28636 + }, + { + "epoch": 0.6303543913713405, + "grad_norm": 2.3554130624879877, + "learning_rate": 3.175143710720927e-06, + "loss": 0.4051, + "step": 28637 + }, + { + "epoch": 0.6303764032577592, + "grad_norm": 16.470135936812454, + "learning_rate": 3.1748118477728273e-06, + "loss": 0.3904, + "step": 28638 + }, + { + "epoch": 0.6303984151441778, + "grad_norm": 1.382030934655383, + "learning_rate": 3.1744799941011462e-06, + "loss": 0.3086, + "step": 28639 + }, + { + "epoch": 0.6304204270305965, + "grad_norm": 0.6129847574662428, + "learning_rate": 3.174148149707574e-06, + "loss": 0.4474, + "step": 28640 + }, + { + "epoch": 0.6304424389170152, + "grad_norm": 1.6400248391180274, + "learning_rate": 3.1738163145937956e-06, + "loss": 0.3406, + "step": 28641 + }, + { + "epoch": 0.6304644508034338, + "grad_norm": 1.6191803639368552, + "learning_rate": 3.173484488761495e-06, + "loss": 0.2695, + "step": 28642 + }, + { + "epoch": 0.6304864626898525, + "grad_norm": 0.6051301001030704, + "learning_rate": 3.173152672212363e-06, + "loss": 0.4473, + "step": 28643 + }, + { + "epoch": 0.6305084745762712, + "grad_norm": 1.7160364354812516, + "learning_rate": 3.1728208649480817e-06, + "loss": 0.2926, + "step": 28644 + }, + { + "epoch": 0.6305304864626898, + "grad_norm": 1.3236530133547257, + "learning_rate": 3.172489066970341e-06, + "loss": 0.3128, + "step": 28645 + }, + { + "epoch": 0.6305524983491085, + "grad_norm": 1.8004692285650183, + "learning_rate": 3.1721572782808276e-06, + "loss": 0.3168, + "step": 28646 + }, + { + "epoch": 0.6305745102355271, + "grad_norm": 9.91704473257633, + "learning_rate": 3.1718254988812237e-06, + "loss": 0.3281, + "step": 28647 + }, + { + "epoch": 0.6305965221219458, + "grad_norm": 0.584190623061732, + "learning_rate": 3.1714937287732184e-06, + "loss": 0.4613, + "step": 28648 + }, + { + "epoch": 0.6306185340083645, + "grad_norm": 0.6164471950475622, + "learning_rate": 3.171161967958499e-06, + "loss": 0.4353, + "step": 28649 + }, + { + "epoch": 0.6306405458947831, + "grad_norm": 0.6292354538070132, + "learning_rate": 3.170830216438747e-06, + "loss": 0.485, + "step": 28650 + }, + { + "epoch": 0.6306625577812018, + "grad_norm": 2.502610947641081, + "learning_rate": 3.170498474215655e-06, + "loss": 0.2826, + "step": 28651 + }, + { + "epoch": 0.6306845696676205, + "grad_norm": 0.6184712727927371, + "learning_rate": 3.1701667412909043e-06, + "loss": 0.4642, + "step": 28652 + }, + { + "epoch": 0.6307065815540391, + "grad_norm": 1.3737344794032038, + "learning_rate": 3.1698350176661825e-06, + "loss": 0.33, + "step": 28653 + }, + { + "epoch": 0.6307285934404578, + "grad_norm": 2.403281637666712, + "learning_rate": 3.1695033033431764e-06, + "loss": 0.3016, + "step": 28654 + }, + { + "epoch": 0.6307506053268765, + "grad_norm": 1.5136753742794264, + "learning_rate": 3.1691715983235693e-06, + "loss": 0.2798, + "step": 28655 + }, + { + "epoch": 0.6307726172132951, + "grad_norm": 3.169459902684219, + "learning_rate": 3.16883990260905e-06, + "loss": 0.3237, + "step": 28656 + }, + { + "epoch": 0.6307946290997138, + "grad_norm": 1.6059742486828115, + "learning_rate": 3.1685082162013016e-06, + "loss": 0.2876, + "step": 28657 + }, + { + "epoch": 0.6308166409861325, + "grad_norm": 0.6249717106828047, + "learning_rate": 3.168176539102012e-06, + "loss": 0.4748, + "step": 28658 + }, + { + "epoch": 0.6308386528725511, + "grad_norm": 2.3699057917615027, + "learning_rate": 3.1678448713128675e-06, + "loss": 0.2702, + "step": 28659 + }, + { + "epoch": 0.6308606647589698, + "grad_norm": 8.405595865518894, + "learning_rate": 3.16751321283555e-06, + "loss": 0.2665, + "step": 28660 + }, + { + "epoch": 0.6308826766453886, + "grad_norm": 1.9621235971302828, + "learning_rate": 3.1671815636717497e-06, + "loss": 0.395, + "step": 28661 + }, + { + "epoch": 0.6309046885318071, + "grad_norm": 6.127699874527144, + "learning_rate": 3.1668499238231486e-06, + "loss": 0.3712, + "step": 28662 + }, + { + "epoch": 0.6309267004182258, + "grad_norm": 1.491222076944537, + "learning_rate": 3.166518293291435e-06, + "loss": 0.2708, + "step": 28663 + }, + { + "epoch": 0.6309487123046446, + "grad_norm": 3.3032850845239388, + "learning_rate": 3.166186672078294e-06, + "loss": 0.3638, + "step": 28664 + }, + { + "epoch": 0.6309707241910631, + "grad_norm": 0.6091432248183433, + "learning_rate": 3.165855060185409e-06, + "loss": 0.4533, + "step": 28665 + }, + { + "epoch": 0.6309927360774819, + "grad_norm": 2.2258597891056078, + "learning_rate": 3.165523457614467e-06, + "loss": 0.2835, + "step": 28666 + }, + { + "epoch": 0.6310147479639006, + "grad_norm": 1.6978715135957958, + "learning_rate": 3.165191864367153e-06, + "loss": 0.3492, + "step": 28667 + }, + { + "epoch": 0.6310367598503192, + "grad_norm": 1.5221200495652463, + "learning_rate": 3.1648602804451504e-06, + "loss": 0.268, + "step": 28668 + }, + { + "epoch": 0.6310587717367379, + "grad_norm": 3.2847339404459475, + "learning_rate": 3.1645287058501494e-06, + "loss": 0.2704, + "step": 28669 + }, + { + "epoch": 0.6310807836231564, + "grad_norm": 1.3807045524756238, + "learning_rate": 3.16419714058383e-06, + "loss": 0.2678, + "step": 28670 + }, + { + "epoch": 0.6311027955095752, + "grad_norm": 1.9992830458541209, + "learning_rate": 3.1638655846478807e-06, + "loss": 0.3612, + "step": 28671 + }, + { + "epoch": 0.6311248073959939, + "grad_norm": 2.2519541332310204, + "learning_rate": 3.163534038043986e-06, + "loss": 0.2753, + "step": 28672 + }, + { + "epoch": 0.6311468192824125, + "grad_norm": 5.1351135037380935, + "learning_rate": 3.1632025007738288e-06, + "loss": 0.3012, + "step": 28673 + }, + { + "epoch": 0.6311688311688312, + "grad_norm": 1.8338038781984833, + "learning_rate": 3.1628709728390962e-06, + "loss": 0.2125, + "step": 28674 + }, + { + "epoch": 0.6311908430552499, + "grad_norm": 1.6533685722122304, + "learning_rate": 3.162539454241472e-06, + "loss": 0.3144, + "step": 28675 + }, + { + "epoch": 0.6312128549416685, + "grad_norm": 0.5985501135987443, + "learning_rate": 3.1622079449826425e-06, + "loss": 0.4571, + "step": 28676 + }, + { + "epoch": 0.6312348668280872, + "grad_norm": 1.8525853933926062, + "learning_rate": 3.1618764450642938e-06, + "loss": 0.301, + "step": 28677 + }, + { + "epoch": 0.6312568787145059, + "grad_norm": 1.403624362163196, + "learning_rate": 3.161544954488106e-06, + "loss": 0.2554, + "step": 28678 + }, + { + "epoch": 0.6312788906009245, + "grad_norm": 1.5555337204454285, + "learning_rate": 3.161213473255768e-06, + "loss": 0.3202, + "step": 28679 + }, + { + "epoch": 0.6313009024873432, + "grad_norm": 2.378973237708859, + "learning_rate": 3.1608820013689624e-06, + "loss": 0.2419, + "step": 28680 + }, + { + "epoch": 0.6313229143737619, + "grad_norm": 2.0487104073389144, + "learning_rate": 3.1605505388293755e-06, + "loss": 0.3902, + "step": 28681 + }, + { + "epoch": 0.6313449262601805, + "grad_norm": 1.5670859703536244, + "learning_rate": 3.160219085638693e-06, + "loss": 0.2435, + "step": 28682 + }, + { + "epoch": 0.6313669381465992, + "grad_norm": 1.6643721689258224, + "learning_rate": 3.1598876417985945e-06, + "loss": 0.373, + "step": 28683 + }, + { + "epoch": 0.6313889500330179, + "grad_norm": 1.594743263818657, + "learning_rate": 3.159556207310769e-06, + "loss": 0.3333, + "step": 28684 + }, + { + "epoch": 0.6314109619194365, + "grad_norm": 1.5828693146339945, + "learning_rate": 3.159224782176901e-06, + "loss": 0.3399, + "step": 28685 + }, + { + "epoch": 0.6314329738058552, + "grad_norm": 4.247355217326408, + "learning_rate": 3.1588933663986703e-06, + "loss": 0.2235, + "step": 28686 + }, + { + "epoch": 0.6314549856922739, + "grad_norm": 1.6485586613702472, + "learning_rate": 3.1585619599777686e-06, + "loss": 0.3178, + "step": 28687 + }, + { + "epoch": 0.6314769975786925, + "grad_norm": 1.424960468184171, + "learning_rate": 3.158230562915873e-06, + "loss": 0.2732, + "step": 28688 + }, + { + "epoch": 0.6314990094651112, + "grad_norm": 2.226262898795056, + "learning_rate": 3.1578991752146724e-06, + "loss": 0.2981, + "step": 28689 + }, + { + "epoch": 0.6315210213515299, + "grad_norm": 1.9075696550568957, + "learning_rate": 3.1575677968758517e-06, + "loss": 0.3452, + "step": 28690 + }, + { + "epoch": 0.6315430332379485, + "grad_norm": 2.9174095645741387, + "learning_rate": 3.1572364279010903e-06, + "loss": 0.331, + "step": 28691 + }, + { + "epoch": 0.6315650451243672, + "grad_norm": 1.6562819608311683, + "learning_rate": 3.1569050682920764e-06, + "loss": 0.257, + "step": 28692 + }, + { + "epoch": 0.6315870570107859, + "grad_norm": 1.1521499416514507, + "learning_rate": 3.156573718050493e-06, + "loss": 0.239, + "step": 28693 + }, + { + "epoch": 0.6316090688972045, + "grad_norm": 1.3691325459979193, + "learning_rate": 3.1562423771780237e-06, + "loss": 0.235, + "step": 28694 + }, + { + "epoch": 0.6316310807836232, + "grad_norm": 0.620448365364088, + "learning_rate": 3.1559110456763553e-06, + "loss": 0.4568, + "step": 28695 + }, + { + "epoch": 0.6316530926700418, + "grad_norm": 0.6244402937457909, + "learning_rate": 3.1555797235471665e-06, + "loss": 0.4593, + "step": 28696 + }, + { + "epoch": 0.6316751045564605, + "grad_norm": 1.6155656211866989, + "learning_rate": 3.1552484107921456e-06, + "loss": 0.3115, + "step": 28697 + }, + { + "epoch": 0.6316971164428792, + "grad_norm": 1.5385333572405957, + "learning_rate": 3.1549171074129734e-06, + "loss": 0.3015, + "step": 28698 + }, + { + "epoch": 0.6317191283292978, + "grad_norm": 2.592607748727272, + "learning_rate": 3.1545858134113373e-06, + "loss": 0.3249, + "step": 28699 + }, + { + "epoch": 0.6317411402157165, + "grad_norm": 3.4028992094144748, + "learning_rate": 3.15425452878892e-06, + "loss": 0.2941, + "step": 28700 + }, + { + "epoch": 0.6317631521021352, + "grad_norm": 1.8176458925121906, + "learning_rate": 3.1539232535474018e-06, + "loss": 0.2616, + "step": 28701 + }, + { + "epoch": 0.6317851639885538, + "grad_norm": 1.780002991934346, + "learning_rate": 3.153591987688471e-06, + "loss": 0.3255, + "step": 28702 + }, + { + "epoch": 0.6318071758749725, + "grad_norm": 2.2691727773286057, + "learning_rate": 3.1532607312138075e-06, + "loss": 0.3387, + "step": 28703 + }, + { + "epoch": 0.6318291877613912, + "grad_norm": 1.5726430383009897, + "learning_rate": 3.152929484125098e-06, + "loss": 0.2569, + "step": 28704 + }, + { + "epoch": 0.6318511996478098, + "grad_norm": 1.9482955574995273, + "learning_rate": 3.152598246424025e-06, + "loss": 0.2704, + "step": 28705 + }, + { + "epoch": 0.6318732115342285, + "grad_norm": 1.9294958160276658, + "learning_rate": 3.1522670181122707e-06, + "loss": 0.2591, + "step": 28706 + }, + { + "epoch": 0.6318952234206472, + "grad_norm": 1.6838140244740658, + "learning_rate": 3.1519357991915194e-06, + "loss": 0.3206, + "step": 28707 + }, + { + "epoch": 0.6319172353070658, + "grad_norm": 0.6303586443924897, + "learning_rate": 3.151604589663456e-06, + "loss": 0.482, + "step": 28708 + }, + { + "epoch": 0.6319392471934845, + "grad_norm": 0.6153935917579966, + "learning_rate": 3.151273389529761e-06, + "loss": 0.4492, + "step": 28709 + }, + { + "epoch": 0.6319612590799032, + "grad_norm": 1.637635787396052, + "learning_rate": 3.1509421987921195e-06, + "loss": 0.2949, + "step": 28710 + }, + { + "epoch": 0.6319832709663218, + "grad_norm": 1.8538966639283665, + "learning_rate": 3.1506110174522144e-06, + "loss": 0.2441, + "step": 28711 + }, + { + "epoch": 0.6320052828527405, + "grad_norm": 2.4067575784776003, + "learning_rate": 3.150279845511729e-06, + "loss": 0.2121, + "step": 28712 + }, + { + "epoch": 0.6320272947391592, + "grad_norm": 2.132048080044747, + "learning_rate": 3.149948682972348e-06, + "loss": 0.3649, + "step": 28713 + }, + { + "epoch": 0.6320493066255778, + "grad_norm": 1.8408483670808908, + "learning_rate": 3.1496175298357507e-06, + "loss": 0.2825, + "step": 28714 + }, + { + "epoch": 0.6320713185119965, + "grad_norm": 2.743039806066501, + "learning_rate": 3.149286386103624e-06, + "loss": 0.2178, + "step": 28715 + }, + { + "epoch": 0.6320933303984152, + "grad_norm": 2.59837379642162, + "learning_rate": 3.1489552517776472e-06, + "loss": 0.3035, + "step": 28716 + }, + { + "epoch": 0.6321153422848338, + "grad_norm": 6.0998338567841115, + "learning_rate": 3.148624126859507e-06, + "loss": 0.3353, + "step": 28717 + }, + { + "epoch": 0.6321373541712525, + "grad_norm": 1.5524395842918757, + "learning_rate": 3.1482930113508857e-06, + "loss": 0.3858, + "step": 28718 + }, + { + "epoch": 0.6321593660576711, + "grad_norm": 1.8415501336900226, + "learning_rate": 3.147961905253463e-06, + "loss": 0.314, + "step": 28719 + }, + { + "epoch": 0.6321813779440898, + "grad_norm": 1.6221766974626564, + "learning_rate": 3.1476308085689255e-06, + "loss": 0.3073, + "step": 28720 + }, + { + "epoch": 0.6322033898305085, + "grad_norm": 0.6010731063550669, + "learning_rate": 3.147299721298953e-06, + "loss": 0.4667, + "step": 28721 + }, + { + "epoch": 0.6322254017169271, + "grad_norm": 2.4463125432798147, + "learning_rate": 3.1469686434452306e-06, + "loss": 0.2889, + "step": 28722 + }, + { + "epoch": 0.6322474136033458, + "grad_norm": 1.5079175485853649, + "learning_rate": 3.146637575009441e-06, + "loss": 0.1935, + "step": 28723 + }, + { + "epoch": 0.6322694254897645, + "grad_norm": 1.8541220566570862, + "learning_rate": 3.1463065159932637e-06, + "loss": 0.3587, + "step": 28724 + }, + { + "epoch": 0.6322914373761831, + "grad_norm": 1.6549150958562475, + "learning_rate": 3.1459754663983856e-06, + "loss": 0.2279, + "step": 28725 + }, + { + "epoch": 0.6323134492626018, + "grad_norm": 1.8561645711217964, + "learning_rate": 3.145644426226487e-06, + "loss": 0.3168, + "step": 28726 + }, + { + "epoch": 0.6323354611490205, + "grad_norm": 1.5921706800860222, + "learning_rate": 3.145313395479248e-06, + "loss": 0.273, + "step": 28727 + }, + { + "epoch": 0.6323574730354391, + "grad_norm": 1.7230614716596044, + "learning_rate": 3.1449823741583555e-06, + "loss": 0.2811, + "step": 28728 + }, + { + "epoch": 0.6323794849218578, + "grad_norm": 1.695308858486247, + "learning_rate": 3.144651362265489e-06, + "loss": 0.2962, + "step": 28729 + }, + { + "epoch": 0.6324014968082765, + "grad_norm": 2.0235489232925077, + "learning_rate": 3.144320359802332e-06, + "loss": 0.2625, + "step": 28730 + }, + { + "epoch": 0.6324235086946951, + "grad_norm": 1.6560551640764445, + "learning_rate": 3.1439893667705684e-06, + "loss": 0.311, + "step": 28731 + }, + { + "epoch": 0.6324455205811138, + "grad_norm": 2.2450003237559866, + "learning_rate": 3.1436583831718765e-06, + "loss": 0.2784, + "step": 28732 + }, + { + "epoch": 0.6324675324675325, + "grad_norm": 1.5791358226697476, + "learning_rate": 3.143327409007941e-06, + "loss": 0.3291, + "step": 28733 + }, + { + "epoch": 0.6324895443539511, + "grad_norm": 1.5476357069586084, + "learning_rate": 3.1429964442804433e-06, + "loss": 0.2533, + "step": 28734 + }, + { + "epoch": 0.6325115562403698, + "grad_norm": 1.4291475541505434, + "learning_rate": 3.142665488991067e-06, + "loss": 0.2296, + "step": 28735 + }, + { + "epoch": 0.6325335681267885, + "grad_norm": 1.6615562084452444, + "learning_rate": 3.1423345431414943e-06, + "loss": 0.248, + "step": 28736 + }, + { + "epoch": 0.6325555800132071, + "grad_norm": 1.5604248784881989, + "learning_rate": 3.142003606733403e-06, + "loss": 0.2487, + "step": 28737 + }, + { + "epoch": 0.6325775918996258, + "grad_norm": 1.7664340216194327, + "learning_rate": 3.141672679768479e-06, + "loss": 0.3279, + "step": 28738 + }, + { + "epoch": 0.6325996037860445, + "grad_norm": 2.5724950588668203, + "learning_rate": 3.1413417622484034e-06, + "loss": 0.2975, + "step": 28739 + }, + { + "epoch": 0.6326216156724631, + "grad_norm": 1.7689838885160023, + "learning_rate": 3.1410108541748596e-06, + "loss": 0.2751, + "step": 28740 + }, + { + "epoch": 0.6326436275588818, + "grad_norm": 1.9271484771117453, + "learning_rate": 3.1406799555495267e-06, + "loss": 0.2895, + "step": 28741 + }, + { + "epoch": 0.6326656394453004, + "grad_norm": 2.1651711885611196, + "learning_rate": 3.1403490663740867e-06, + "loss": 0.3234, + "step": 28742 + }, + { + "epoch": 0.6326876513317191, + "grad_norm": 1.441094809983395, + "learning_rate": 3.140018186650223e-06, + "loss": 0.2621, + "step": 28743 + }, + { + "epoch": 0.6327096632181378, + "grad_norm": 26.12465834361472, + "learning_rate": 3.1396873163796147e-06, + "loss": 0.3073, + "step": 28744 + }, + { + "epoch": 0.6327316751045564, + "grad_norm": 2.104724575164976, + "learning_rate": 3.1393564555639483e-06, + "loss": 0.3011, + "step": 28745 + }, + { + "epoch": 0.6327536869909751, + "grad_norm": 1.5256285433524208, + "learning_rate": 3.1390256042049007e-06, + "loss": 0.3632, + "step": 28746 + }, + { + "epoch": 0.6327756988773938, + "grad_norm": 1.6098783514140154, + "learning_rate": 3.1386947623041543e-06, + "loss": 0.3524, + "step": 28747 + }, + { + "epoch": 0.6327977107638124, + "grad_norm": 1.7996571030702526, + "learning_rate": 3.1383639298633915e-06, + "loss": 0.2356, + "step": 28748 + }, + { + "epoch": 0.6328197226502311, + "grad_norm": 1.316565604072412, + "learning_rate": 3.1380331068842955e-06, + "loss": 0.2445, + "step": 28749 + }, + { + "epoch": 0.6328417345366498, + "grad_norm": 1.6489405970375177, + "learning_rate": 3.137702293368543e-06, + "loss": 0.3053, + "step": 28750 + }, + { + "epoch": 0.6328637464230684, + "grad_norm": 1.194754373315953, + "learning_rate": 3.13737148931782e-06, + "loss": 0.287, + "step": 28751 + }, + { + "epoch": 0.6328857583094871, + "grad_norm": 0.6452559177848335, + "learning_rate": 3.137040694733804e-06, + "loss": 0.4728, + "step": 28752 + }, + { + "epoch": 0.6329077701959058, + "grad_norm": 1.8131989564470676, + "learning_rate": 3.1367099096181797e-06, + "loss": 0.2565, + "step": 28753 + }, + { + "epoch": 0.6329297820823244, + "grad_norm": 2.672500654569871, + "learning_rate": 3.136379133972627e-06, + "loss": 0.2777, + "step": 28754 + }, + { + "epoch": 0.6329517939687431, + "grad_norm": 1.6012259940934297, + "learning_rate": 3.1360483677988253e-06, + "loss": 0.2494, + "step": 28755 + }, + { + "epoch": 0.6329738058551618, + "grad_norm": 1.4268932386633058, + "learning_rate": 3.1357176110984578e-06, + "loss": 0.4391, + "step": 28756 + }, + { + "epoch": 0.6329958177415804, + "grad_norm": 1.8101476433057582, + "learning_rate": 3.1353868638732037e-06, + "loss": 0.3156, + "step": 28757 + }, + { + "epoch": 0.6330178296279991, + "grad_norm": 1.5636837145389562, + "learning_rate": 3.135056126124748e-06, + "loss": 0.352, + "step": 28758 + }, + { + "epoch": 0.6330398415144178, + "grad_norm": 1.796490307362515, + "learning_rate": 3.134725397854767e-06, + "loss": 0.2791, + "step": 28759 + }, + { + "epoch": 0.6330618534008364, + "grad_norm": 3.7127780534695267, + "learning_rate": 3.1343946790649427e-06, + "loss": 0.2881, + "step": 28760 + }, + { + "epoch": 0.6330838652872551, + "grad_norm": 1.5458894047752076, + "learning_rate": 3.134063969756958e-06, + "loss": 0.2554, + "step": 28761 + }, + { + "epoch": 0.6331058771736738, + "grad_norm": 2.814253998646812, + "learning_rate": 3.1337332699324914e-06, + "loss": 0.3212, + "step": 28762 + }, + { + "epoch": 0.6331278890600924, + "grad_norm": 4.399161785832213, + "learning_rate": 3.1334025795932264e-06, + "loss": 0.2948, + "step": 28763 + }, + { + "epoch": 0.6331499009465111, + "grad_norm": 2.9906090930892044, + "learning_rate": 3.1330718987408413e-06, + "loss": 0.3348, + "step": 28764 + }, + { + "epoch": 0.6331719128329297, + "grad_norm": 2.326916559977971, + "learning_rate": 3.1327412273770163e-06, + "loss": 0.3412, + "step": 28765 + }, + { + "epoch": 0.6331939247193484, + "grad_norm": 1.7498226652624884, + "learning_rate": 3.1324105655034344e-06, + "loss": 0.3451, + "step": 28766 + }, + { + "epoch": 0.6332159366057671, + "grad_norm": 1.548132831495811, + "learning_rate": 3.132079913121776e-06, + "loss": 0.2604, + "step": 28767 + }, + { + "epoch": 0.6332379484921857, + "grad_norm": 1.2226433981513822, + "learning_rate": 3.1317492702337182e-06, + "loss": 0.2208, + "step": 28768 + }, + { + "epoch": 0.6332599603786044, + "grad_norm": 2.0309755992637686, + "learning_rate": 3.131418636840945e-06, + "loss": 0.2741, + "step": 28769 + }, + { + "epoch": 0.6332819722650231, + "grad_norm": 3.339317811311067, + "learning_rate": 3.1310880129451347e-06, + "loss": 0.3246, + "step": 28770 + }, + { + "epoch": 0.6333039841514417, + "grad_norm": 1.6415305057750376, + "learning_rate": 3.13075739854797e-06, + "loss": 0.2814, + "step": 28771 + }, + { + "epoch": 0.6333259960378604, + "grad_norm": 1.3703900109836824, + "learning_rate": 3.1304267936511306e-06, + "loss": 0.2116, + "step": 28772 + }, + { + "epoch": 0.6333480079242791, + "grad_norm": 0.6348584313925776, + "learning_rate": 3.130096198256294e-06, + "loss": 0.4711, + "step": 28773 + }, + { + "epoch": 0.6333700198106977, + "grad_norm": 1.290363971923867, + "learning_rate": 3.1297656123651443e-06, + "loss": 0.2704, + "step": 28774 + }, + { + "epoch": 0.6333920316971164, + "grad_norm": 0.6535440256837344, + "learning_rate": 3.129435035979358e-06, + "loss": 0.4729, + "step": 28775 + }, + { + "epoch": 0.6334140435835351, + "grad_norm": 1.3368910241313405, + "learning_rate": 3.1291044691006193e-06, + "loss": 0.2412, + "step": 28776 + }, + { + "epoch": 0.6334360554699537, + "grad_norm": 1.8921468533992984, + "learning_rate": 3.1287739117306052e-06, + "loss": 0.2921, + "step": 28777 + }, + { + "epoch": 0.6334580673563724, + "grad_norm": 2.6602510883579824, + "learning_rate": 3.128443363870996e-06, + "loss": 0.3647, + "step": 28778 + }, + { + "epoch": 0.6334800792427911, + "grad_norm": 1.6530459597035227, + "learning_rate": 3.1281128255234727e-06, + "loss": 0.2751, + "step": 28779 + }, + { + "epoch": 0.6335020911292097, + "grad_norm": 0.620071020983096, + "learning_rate": 3.127782296689715e-06, + "loss": 0.4788, + "step": 28780 + }, + { + "epoch": 0.6335241030156284, + "grad_norm": 2.167137557785421, + "learning_rate": 3.127451777371404e-06, + "loss": 0.3237, + "step": 28781 + }, + { + "epoch": 0.6335461149020472, + "grad_norm": 21.985703519972656, + "learning_rate": 3.127121267570217e-06, + "loss": 0.29, + "step": 28782 + }, + { + "epoch": 0.6335681267884657, + "grad_norm": 1.9638661515348474, + "learning_rate": 3.1267907672878344e-06, + "loss": 0.2943, + "step": 28783 + }, + { + "epoch": 0.6335901386748845, + "grad_norm": 2.2158896129304657, + "learning_rate": 3.126460276525938e-06, + "loss": 0.2831, + "step": 28784 + }, + { + "epoch": 0.6336121505613032, + "grad_norm": 0.5778095493741554, + "learning_rate": 3.126129795286206e-06, + "loss": 0.424, + "step": 28785 + }, + { + "epoch": 0.6336341624477217, + "grad_norm": 1.5537813019050253, + "learning_rate": 3.1257993235703166e-06, + "loss": 0.2396, + "step": 28786 + }, + { + "epoch": 0.6336561743341405, + "grad_norm": 2.398596197813682, + "learning_rate": 3.1254688613799523e-06, + "loss": 0.3947, + "step": 28787 + }, + { + "epoch": 0.633678186220559, + "grad_norm": 1.6380701363773287, + "learning_rate": 3.12513840871679e-06, + "loss": 0.2422, + "step": 28788 + }, + { + "epoch": 0.6337001981069778, + "grad_norm": 2.081207674374005, + "learning_rate": 3.1248079655825115e-06, + "loss": 0.312, + "step": 28789 + }, + { + "epoch": 0.6337222099933965, + "grad_norm": 0.6490243966663145, + "learning_rate": 3.1244775319787968e-06, + "loss": 0.5166, + "step": 28790 + }, + { + "epoch": 0.633744221879815, + "grad_norm": 0.6177050051388281, + "learning_rate": 3.1241471079073215e-06, + "loss": 0.4584, + "step": 28791 + }, + { + "epoch": 0.6337662337662338, + "grad_norm": 4.0082973631040435, + "learning_rate": 3.123816693369768e-06, + "loss": 0.2758, + "step": 28792 + }, + { + "epoch": 0.6337882456526525, + "grad_norm": 4.685206811342448, + "learning_rate": 3.123486288367814e-06, + "loss": 0.4339, + "step": 28793 + }, + { + "epoch": 0.633810257539071, + "grad_norm": 0.6098343056204919, + "learning_rate": 3.1231558929031414e-06, + "loss": 0.4618, + "step": 28794 + }, + { + "epoch": 0.6338322694254898, + "grad_norm": 1.5587446303426218, + "learning_rate": 3.1228255069774273e-06, + "loss": 0.3133, + "step": 28795 + }, + { + "epoch": 0.6338542813119085, + "grad_norm": 1.71879625864639, + "learning_rate": 3.12249513059235e-06, + "loss": 0.3382, + "step": 28796 + }, + { + "epoch": 0.6338762931983271, + "grad_norm": 2.089676789202699, + "learning_rate": 3.1221647637495904e-06, + "loss": 0.3025, + "step": 28797 + }, + { + "epoch": 0.6338983050847458, + "grad_norm": 2.7683130632266377, + "learning_rate": 3.1218344064508265e-06, + "loss": 0.2447, + "step": 28798 + }, + { + "epoch": 0.6339203169711645, + "grad_norm": 1.8284081804198347, + "learning_rate": 3.1215040586977398e-06, + "loss": 0.3058, + "step": 28799 + }, + { + "epoch": 0.6339423288575831, + "grad_norm": 1.9329021559390491, + "learning_rate": 3.1211737204920057e-06, + "loss": 0.3795, + "step": 28800 + }, + { + "epoch": 0.6339643407440018, + "grad_norm": 1.65406854336378, + "learning_rate": 3.1208433918353043e-06, + "loss": 0.2813, + "step": 28801 + }, + { + "epoch": 0.6339863526304205, + "grad_norm": 2.2481489025812684, + "learning_rate": 3.1205130727293153e-06, + "loss": 0.3348, + "step": 28802 + }, + { + "epoch": 0.6340083645168391, + "grad_norm": 1.9226355489851996, + "learning_rate": 3.120182763175717e-06, + "loss": 0.3383, + "step": 28803 + }, + { + "epoch": 0.6340303764032578, + "grad_norm": 2.126603699450155, + "learning_rate": 3.1198524631761896e-06, + "loss": 0.3366, + "step": 28804 + }, + { + "epoch": 0.6340523882896765, + "grad_norm": 1.5643383053097248, + "learning_rate": 3.11952217273241e-06, + "loss": 0.2851, + "step": 28805 + }, + { + "epoch": 0.6340744001760951, + "grad_norm": 2.0295504150396693, + "learning_rate": 3.1191918918460564e-06, + "loss": 0.3557, + "step": 28806 + }, + { + "epoch": 0.6340964120625138, + "grad_norm": 1.8800126539340403, + "learning_rate": 3.1188616205188094e-06, + "loss": 0.291, + "step": 28807 + }, + { + "epoch": 0.6341184239489325, + "grad_norm": 1.6405626959694406, + "learning_rate": 3.1185313587523473e-06, + "loss": 0.2478, + "step": 28808 + }, + { + "epoch": 0.6341404358353511, + "grad_norm": 1.4250085963484036, + "learning_rate": 3.1182011065483463e-06, + "loss": 0.283, + "step": 28809 + }, + { + "epoch": 0.6341624477217698, + "grad_norm": 1.3768472874139746, + "learning_rate": 3.1178708639084874e-06, + "loss": 0.2264, + "step": 28810 + }, + { + "epoch": 0.6341844596081884, + "grad_norm": 1.3315383212820622, + "learning_rate": 3.117540630834448e-06, + "loss": 0.2515, + "step": 28811 + }, + { + "epoch": 0.6342064714946071, + "grad_norm": 1.6802395798524732, + "learning_rate": 3.117210407327908e-06, + "loss": 0.3068, + "step": 28812 + }, + { + "epoch": 0.6342284833810258, + "grad_norm": 1.8675898090842622, + "learning_rate": 3.1168801933905434e-06, + "loss": 0.3728, + "step": 28813 + }, + { + "epoch": 0.6342504952674444, + "grad_norm": 1.2823672171115228, + "learning_rate": 3.116549989024033e-06, + "loss": 0.3297, + "step": 28814 + }, + { + "epoch": 0.6342725071538631, + "grad_norm": 4.830375200522766, + "learning_rate": 3.1162197942300566e-06, + "loss": 0.2972, + "step": 28815 + }, + { + "epoch": 0.6342945190402818, + "grad_norm": 1.53110483298235, + "learning_rate": 3.1158896090102898e-06, + "loss": 0.3226, + "step": 28816 + }, + { + "epoch": 0.6343165309267004, + "grad_norm": 1.8582980268366125, + "learning_rate": 3.1155594333664146e-06, + "loss": 0.3408, + "step": 28817 + }, + { + "epoch": 0.6343385428131191, + "grad_norm": 2.1205730948069217, + "learning_rate": 3.115229267300106e-06, + "loss": 0.3138, + "step": 28818 + }, + { + "epoch": 0.6343605546995378, + "grad_norm": 1.5501966155872458, + "learning_rate": 3.1148991108130423e-06, + "loss": 0.3383, + "step": 28819 + }, + { + "epoch": 0.6343825665859564, + "grad_norm": 2.498387255349626, + "learning_rate": 3.114568963906902e-06, + "loss": 0.32, + "step": 28820 + }, + { + "epoch": 0.6344045784723751, + "grad_norm": 2.0287989159610205, + "learning_rate": 3.114238826583363e-06, + "loss": 0.3497, + "step": 28821 + }, + { + "epoch": 0.6344265903587938, + "grad_norm": 2.03775128173584, + "learning_rate": 3.1139086988441047e-06, + "loss": 0.2983, + "step": 28822 + }, + { + "epoch": 0.6344486022452124, + "grad_norm": 0.622092960595965, + "learning_rate": 3.1135785806908036e-06, + "loss": 0.4597, + "step": 28823 + }, + { + "epoch": 0.6344706141316311, + "grad_norm": 1.6006513903574913, + "learning_rate": 3.1132484721251353e-06, + "loss": 0.2828, + "step": 28824 + }, + { + "epoch": 0.6344926260180498, + "grad_norm": 1.1746615593176892, + "learning_rate": 3.112918373148782e-06, + "loss": 0.1976, + "step": 28825 + }, + { + "epoch": 0.6345146379044684, + "grad_norm": 1.6979444593165636, + "learning_rate": 3.112588283763419e-06, + "loss": 0.2277, + "step": 28826 + }, + { + "epoch": 0.6345366497908871, + "grad_norm": 2.106726688620612, + "learning_rate": 3.1122582039707234e-06, + "loss": 0.3688, + "step": 28827 + }, + { + "epoch": 0.6345586616773058, + "grad_norm": 2.1895554044702696, + "learning_rate": 3.111928133772374e-06, + "loss": 0.3036, + "step": 28828 + }, + { + "epoch": 0.6345806735637244, + "grad_norm": 2.716918063370808, + "learning_rate": 3.1115980731700466e-06, + "loss": 0.2792, + "step": 28829 + }, + { + "epoch": 0.6346026854501431, + "grad_norm": 1.5114076739815232, + "learning_rate": 3.111268022165422e-06, + "loss": 0.3142, + "step": 28830 + }, + { + "epoch": 0.6346246973365618, + "grad_norm": 1.6437719837859806, + "learning_rate": 3.110937980760176e-06, + "loss": 0.2594, + "step": 28831 + }, + { + "epoch": 0.6346467092229804, + "grad_norm": 1.3663663744310603, + "learning_rate": 3.1106079489559834e-06, + "loss": 0.3685, + "step": 28832 + }, + { + "epoch": 0.6346687211093991, + "grad_norm": 1.5268044166094004, + "learning_rate": 3.1102779267545245e-06, + "loss": 0.323, + "step": 28833 + }, + { + "epoch": 0.6346907329958178, + "grad_norm": 2.8413999382342, + "learning_rate": 3.109947914157476e-06, + "loss": 0.3652, + "step": 28834 + }, + { + "epoch": 0.6347127448822364, + "grad_norm": 7.342110175258476, + "learning_rate": 3.109617911166517e-06, + "loss": 0.3531, + "step": 28835 + }, + { + "epoch": 0.6347347567686551, + "grad_norm": 1.572591196088357, + "learning_rate": 3.1092879177833222e-06, + "loss": 0.3072, + "step": 28836 + }, + { + "epoch": 0.6347567686550737, + "grad_norm": 2.166500416511155, + "learning_rate": 3.1089579340095675e-06, + "loss": 0.3214, + "step": 28837 + }, + { + "epoch": 0.6347787805414924, + "grad_norm": 4.180002651822523, + "learning_rate": 3.108627959846934e-06, + "loss": 0.2977, + "step": 28838 + }, + { + "epoch": 0.6348007924279111, + "grad_norm": 1.6014506190018936, + "learning_rate": 3.1082979952970955e-06, + "loss": 0.3248, + "step": 28839 + }, + { + "epoch": 0.6348228043143297, + "grad_norm": 1.9711926161678432, + "learning_rate": 3.1079680403617324e-06, + "loss": 0.2825, + "step": 28840 + }, + { + "epoch": 0.6348448162007484, + "grad_norm": 11.801509002054086, + "learning_rate": 3.1076380950425184e-06, + "loss": 0.3244, + "step": 28841 + }, + { + "epoch": 0.6348668280871671, + "grad_norm": 1.2682877689085206, + "learning_rate": 3.10730815934113e-06, + "loss": 0.2838, + "step": 28842 + }, + { + "epoch": 0.6348888399735857, + "grad_norm": 1.2888211762055375, + "learning_rate": 3.1069782332592478e-06, + "loss": 0.2637, + "step": 28843 + }, + { + "epoch": 0.6349108518600044, + "grad_norm": 1.4767166710252229, + "learning_rate": 3.106648316798545e-06, + "loss": 0.2727, + "step": 28844 + }, + { + "epoch": 0.6349328637464231, + "grad_norm": 2.0771026067408536, + "learning_rate": 3.1063184099607026e-06, + "loss": 0.3272, + "step": 28845 + }, + { + "epoch": 0.6349548756328417, + "grad_norm": 1.989842879456785, + "learning_rate": 3.1059885127473925e-06, + "loss": 0.3515, + "step": 28846 + }, + { + "epoch": 0.6349768875192604, + "grad_norm": 2.391952216880981, + "learning_rate": 3.1056586251602933e-06, + "loss": 0.2678, + "step": 28847 + }, + { + "epoch": 0.6349988994056791, + "grad_norm": 4.316023115550709, + "learning_rate": 3.1053287472010834e-06, + "loss": 0.2584, + "step": 28848 + }, + { + "epoch": 0.6350209112920977, + "grad_norm": 1.8981756900903275, + "learning_rate": 3.104998878871437e-06, + "loss": 0.2282, + "step": 28849 + }, + { + "epoch": 0.6350429231785164, + "grad_norm": 1.6955245028689538, + "learning_rate": 3.104669020173031e-06, + "loss": 0.2904, + "step": 28850 + }, + { + "epoch": 0.6350649350649351, + "grad_norm": 1.6160725129348548, + "learning_rate": 3.1043391711075427e-06, + "loss": 0.292, + "step": 28851 + }, + { + "epoch": 0.6350869469513537, + "grad_norm": 1.9769065769028136, + "learning_rate": 3.104009331676647e-06, + "loss": 0.2483, + "step": 28852 + }, + { + "epoch": 0.6351089588377724, + "grad_norm": 1.8214899170228327, + "learning_rate": 3.1036795018820244e-06, + "loss": 0.3271, + "step": 28853 + }, + { + "epoch": 0.6351309707241911, + "grad_norm": 1.9826557629019659, + "learning_rate": 3.103349681725347e-06, + "loss": 0.3326, + "step": 28854 + }, + { + "epoch": 0.6351529826106097, + "grad_norm": 1.7111461796471992, + "learning_rate": 3.1030198712082917e-06, + "loss": 0.2682, + "step": 28855 + }, + { + "epoch": 0.6351749944970284, + "grad_norm": 0.620181758035699, + "learning_rate": 3.102690070332536e-06, + "loss": 0.4657, + "step": 28856 + }, + { + "epoch": 0.6351970063834471, + "grad_norm": 1.39581907707233, + "learning_rate": 3.102360279099755e-06, + "loss": 0.2859, + "step": 28857 + }, + { + "epoch": 0.6352190182698657, + "grad_norm": 0.6254320603419973, + "learning_rate": 3.1020304975116266e-06, + "loss": 0.4818, + "step": 28858 + }, + { + "epoch": 0.6352410301562844, + "grad_norm": 2.182390448240081, + "learning_rate": 3.1017007255698256e-06, + "loss": 0.308, + "step": 28859 + }, + { + "epoch": 0.635263042042703, + "grad_norm": 1.2602530514244032, + "learning_rate": 3.1013709632760267e-06, + "loss": 0.2213, + "step": 28860 + }, + { + "epoch": 0.6352850539291217, + "grad_norm": 1.7855442207976806, + "learning_rate": 3.101041210631909e-06, + "loss": 0.2966, + "step": 28861 + }, + { + "epoch": 0.6353070658155404, + "grad_norm": 2.00010666556201, + "learning_rate": 3.1007114676391452e-06, + "loss": 0.3076, + "step": 28862 + }, + { + "epoch": 0.635329077701959, + "grad_norm": 0.6085794185473467, + "learning_rate": 3.1003817342994147e-06, + "loss": 0.4396, + "step": 28863 + }, + { + "epoch": 0.6353510895883777, + "grad_norm": 2.081672912313304, + "learning_rate": 3.100052010614391e-06, + "loss": 0.3849, + "step": 28864 + }, + { + "epoch": 0.6353731014747964, + "grad_norm": 2.801799727290644, + "learning_rate": 3.0997222965857493e-06, + "loss": 0.2942, + "step": 28865 + }, + { + "epoch": 0.635395113361215, + "grad_norm": 7.129469668135775, + "learning_rate": 3.0993925922151685e-06, + "loss": 0.3628, + "step": 28866 + }, + { + "epoch": 0.6354171252476337, + "grad_norm": 0.6289728745703598, + "learning_rate": 3.0990628975043214e-06, + "loss": 0.4852, + "step": 28867 + }, + { + "epoch": 0.6354391371340524, + "grad_norm": 2.433015188419646, + "learning_rate": 3.098733212454883e-06, + "loss": 0.2832, + "step": 28868 + }, + { + "epoch": 0.635461149020471, + "grad_norm": 4.002882475216251, + "learning_rate": 3.0984035370685316e-06, + "loss": 0.284, + "step": 28869 + }, + { + "epoch": 0.6354831609068897, + "grad_norm": 0.6226436558221362, + "learning_rate": 3.0980738713469407e-06, + "loss": 0.4479, + "step": 28870 + }, + { + "epoch": 0.6355051727933084, + "grad_norm": 1.6116857171876635, + "learning_rate": 3.097744215291788e-06, + "loss": 0.3296, + "step": 28871 + }, + { + "epoch": 0.635527184679727, + "grad_norm": 1.9409415288918679, + "learning_rate": 3.0974145689047473e-06, + "loss": 0.3073, + "step": 28872 + }, + { + "epoch": 0.6355491965661457, + "grad_norm": 0.6233792218384454, + "learning_rate": 3.097084932187493e-06, + "loss": 0.4578, + "step": 28873 + }, + { + "epoch": 0.6355712084525644, + "grad_norm": 0.5995288569688975, + "learning_rate": 3.0967553051417032e-06, + "loss": 0.4587, + "step": 28874 + }, + { + "epoch": 0.635593220338983, + "grad_norm": 3.646475770976691, + "learning_rate": 3.0964256877690502e-06, + "loss": 0.4338, + "step": 28875 + }, + { + "epoch": 0.6356152322254017, + "grad_norm": 1.7627636711910313, + "learning_rate": 3.0960960800712135e-06, + "loss": 0.3207, + "step": 28876 + }, + { + "epoch": 0.6356372441118204, + "grad_norm": 0.6139328129794287, + "learning_rate": 3.0957664820498647e-06, + "loss": 0.4629, + "step": 28877 + }, + { + "epoch": 0.635659255998239, + "grad_norm": 0.6220031641273063, + "learning_rate": 3.0954368937066783e-06, + "loss": 0.453, + "step": 28878 + }, + { + "epoch": 0.6356812678846577, + "grad_norm": 1.532607199550764, + "learning_rate": 3.0951073150433327e-06, + "loss": 0.2674, + "step": 28879 + }, + { + "epoch": 0.6357032797710764, + "grad_norm": 2.209960462893964, + "learning_rate": 3.0947777460615005e-06, + "loss": 0.2264, + "step": 28880 + }, + { + "epoch": 0.635725291657495, + "grad_norm": 0.6019171519054065, + "learning_rate": 3.0944481867628583e-06, + "loss": 0.4299, + "step": 28881 + }, + { + "epoch": 0.6357473035439137, + "grad_norm": 1.9513696348482912, + "learning_rate": 3.0941186371490807e-06, + "loss": 0.1948, + "step": 28882 + }, + { + "epoch": 0.6357693154303323, + "grad_norm": 4.034415331247063, + "learning_rate": 3.0937890972218405e-06, + "loss": 0.2478, + "step": 28883 + }, + { + "epoch": 0.635791327316751, + "grad_norm": 1.8397066897438408, + "learning_rate": 3.0934595669828168e-06, + "loss": 0.2389, + "step": 28884 + }, + { + "epoch": 0.6358133392031697, + "grad_norm": 1.8549949341794574, + "learning_rate": 3.09313004643368e-06, + "loss": 0.2365, + "step": 28885 + }, + { + "epoch": 0.6358353510895883, + "grad_norm": 1.5937914507413908, + "learning_rate": 3.092800535576107e-06, + "loss": 0.3735, + "step": 28886 + }, + { + "epoch": 0.635857362976007, + "grad_norm": 1.756264625009749, + "learning_rate": 3.0924710344117726e-06, + "loss": 0.2259, + "step": 28887 + }, + { + "epoch": 0.6358793748624257, + "grad_norm": 1.4486185801604505, + "learning_rate": 3.09214154294235e-06, + "loss": 0.312, + "step": 28888 + }, + { + "epoch": 0.6359013867488443, + "grad_norm": 1.7921008775048834, + "learning_rate": 3.0918120611695173e-06, + "loss": 0.2267, + "step": 28889 + }, + { + "epoch": 0.635923398635263, + "grad_norm": 0.6469883341164806, + "learning_rate": 3.0914825890949463e-06, + "loss": 0.4688, + "step": 28890 + }, + { + "epoch": 0.6359454105216817, + "grad_norm": 1.6374043683613952, + "learning_rate": 3.0911531267203098e-06, + "loss": 0.3341, + "step": 28891 + }, + { + "epoch": 0.6359674224081003, + "grad_norm": 1.4118972360479738, + "learning_rate": 3.0908236740472863e-06, + "loss": 0.2195, + "step": 28892 + }, + { + "epoch": 0.635989434294519, + "grad_norm": 9.166237440203801, + "learning_rate": 3.090494231077548e-06, + "loss": 0.2545, + "step": 28893 + }, + { + "epoch": 0.6360114461809377, + "grad_norm": 2.037793932574549, + "learning_rate": 3.0901647978127702e-06, + "loss": 0.3082, + "step": 28894 + }, + { + "epoch": 0.6360334580673563, + "grad_norm": 2.1149021299286495, + "learning_rate": 3.089835374254627e-06, + "loss": 0.34, + "step": 28895 + }, + { + "epoch": 0.636055469953775, + "grad_norm": 1.5200689935594176, + "learning_rate": 3.0895059604047905e-06, + "loss": 0.2683, + "step": 28896 + }, + { + "epoch": 0.6360774818401937, + "grad_norm": 1.8179343372039294, + "learning_rate": 3.089176556264938e-06, + "loss": 0.3622, + "step": 28897 + }, + { + "epoch": 0.6360994937266123, + "grad_norm": 2.6388725485490125, + "learning_rate": 3.0888471618367418e-06, + "loss": 0.2548, + "step": 28898 + }, + { + "epoch": 0.636121505613031, + "grad_norm": 10.548154282091511, + "learning_rate": 3.088517777121879e-06, + "loss": 0.3056, + "step": 28899 + }, + { + "epoch": 0.6361435174994498, + "grad_norm": 0.6262270367194331, + "learning_rate": 3.0881884021220197e-06, + "loss": 0.4755, + "step": 28900 + }, + { + "epoch": 0.6361655293858683, + "grad_norm": 2.245340469543458, + "learning_rate": 3.0878590368388385e-06, + "loss": 0.343, + "step": 28901 + }, + { + "epoch": 0.636187541272287, + "grad_norm": 0.627726219244232, + "learning_rate": 3.087529681274013e-06, + "loss": 0.4689, + "step": 28902 + }, + { + "epoch": 0.6362095531587058, + "grad_norm": 0.6319816902172941, + "learning_rate": 3.0872003354292114e-06, + "loss": 0.4652, + "step": 28903 + }, + { + "epoch": 0.6362315650451243, + "grad_norm": 1.8908256521681275, + "learning_rate": 3.0868709993061143e-06, + "loss": 0.2626, + "step": 28904 + }, + { + "epoch": 0.636253576931543, + "grad_norm": 0.6028224717741785, + "learning_rate": 3.086541672906391e-06, + "loss": 0.4398, + "step": 28905 + }, + { + "epoch": 0.6362755888179616, + "grad_norm": 0.5959384860754834, + "learning_rate": 3.086212356231715e-06, + "loss": 0.4441, + "step": 28906 + }, + { + "epoch": 0.6362976007043804, + "grad_norm": 1.9099170699481784, + "learning_rate": 3.0858830492837634e-06, + "loss": 0.3425, + "step": 28907 + }, + { + "epoch": 0.636319612590799, + "grad_norm": 2.0618894275401725, + "learning_rate": 3.0855537520642065e-06, + "loss": 0.3845, + "step": 28908 + }, + { + "epoch": 0.6363416244772176, + "grad_norm": 2.486431254180207, + "learning_rate": 3.0852244645747186e-06, + "loss": 0.3159, + "step": 28909 + }, + { + "epoch": 0.6363636363636364, + "grad_norm": 2.1885373535805064, + "learning_rate": 3.084895186816975e-06, + "loss": 0.1937, + "step": 28910 + }, + { + "epoch": 0.6363856482500551, + "grad_norm": 6.957328461648497, + "learning_rate": 3.084565918792648e-06, + "loss": 0.2918, + "step": 28911 + }, + { + "epoch": 0.6364076601364737, + "grad_norm": 1.398557787975397, + "learning_rate": 3.0842366605034125e-06, + "loss": 0.2854, + "step": 28912 + }, + { + "epoch": 0.6364296720228924, + "grad_norm": 1.8866178202027106, + "learning_rate": 3.08390741195094e-06, + "loss": 0.3028, + "step": 28913 + }, + { + "epoch": 0.6364516839093111, + "grad_norm": 2.6354914034740435, + "learning_rate": 3.0835781731369037e-06, + "loss": 0.3043, + "step": 28914 + }, + { + "epoch": 0.6364736957957297, + "grad_norm": 1.9065487535625683, + "learning_rate": 3.0832489440629787e-06, + "loss": 0.3145, + "step": 28915 + }, + { + "epoch": 0.6364957076821484, + "grad_norm": 0.6395387042636549, + "learning_rate": 3.0829197247308363e-06, + "loss": 0.4684, + "step": 28916 + }, + { + "epoch": 0.6365177195685671, + "grad_norm": 0.6323741787152665, + "learning_rate": 3.0825905151421532e-06, + "loss": 0.4674, + "step": 28917 + }, + { + "epoch": 0.6365397314549857, + "grad_norm": 1.6192493274152864, + "learning_rate": 3.082261315298599e-06, + "loss": 0.2693, + "step": 28918 + }, + { + "epoch": 0.6365617433414044, + "grad_norm": 1.9523013661257846, + "learning_rate": 3.0819321252018476e-06, + "loss": 0.2397, + "step": 28919 + }, + { + "epoch": 0.6365837552278231, + "grad_norm": 1.576514344338939, + "learning_rate": 3.081602944853574e-06, + "loss": 0.3043, + "step": 28920 + }, + { + "epoch": 0.6366057671142417, + "grad_norm": 2.449656244648203, + "learning_rate": 3.0812737742554473e-06, + "loss": 0.3209, + "step": 28921 + }, + { + "epoch": 0.6366277790006604, + "grad_norm": 1.7052734136217147, + "learning_rate": 3.0809446134091457e-06, + "loss": 0.3757, + "step": 28922 + }, + { + "epoch": 0.6366497908870791, + "grad_norm": 0.6258633023466543, + "learning_rate": 3.0806154623163386e-06, + "loss": 0.4667, + "step": 28923 + }, + { + "epoch": 0.6366718027734977, + "grad_norm": 2.2804249281877937, + "learning_rate": 3.0802863209786995e-06, + "loss": 0.2699, + "step": 28924 + }, + { + "epoch": 0.6366938146599164, + "grad_norm": 1.5562475769674782, + "learning_rate": 3.079957189397903e-06, + "loss": 0.2786, + "step": 28925 + }, + { + "epoch": 0.6367158265463351, + "grad_norm": 0.6113917010553598, + "learning_rate": 3.079628067575619e-06, + "loss": 0.4529, + "step": 28926 + }, + { + "epoch": 0.6367378384327537, + "grad_norm": 1.3065605111201357, + "learning_rate": 3.079298955513521e-06, + "loss": 0.2999, + "step": 28927 + }, + { + "epoch": 0.6367598503191724, + "grad_norm": 1.7172796577455758, + "learning_rate": 3.078969853213283e-06, + "loss": 0.3061, + "step": 28928 + }, + { + "epoch": 0.636781862205591, + "grad_norm": 1.7503443611407619, + "learning_rate": 3.0786407606765766e-06, + "loss": 0.245, + "step": 28929 + }, + { + "epoch": 0.6368038740920097, + "grad_norm": 1.9646352083993335, + "learning_rate": 3.078311677905077e-06, + "loss": 0.3209, + "step": 28930 + }, + { + "epoch": 0.6368258859784284, + "grad_norm": 2.3467898011139168, + "learning_rate": 3.0779826049004528e-06, + "loss": 0.3945, + "step": 28931 + }, + { + "epoch": 0.636847897864847, + "grad_norm": 1.7874140497130178, + "learning_rate": 3.077653541664378e-06, + "loss": 0.334, + "step": 28932 + }, + { + "epoch": 0.6368699097512657, + "grad_norm": 1.698277227941845, + "learning_rate": 3.0773244881985255e-06, + "loss": 0.3097, + "step": 28933 + }, + { + "epoch": 0.6368919216376844, + "grad_norm": 1.727713499775855, + "learning_rate": 3.076995444504567e-06, + "loss": 0.2821, + "step": 28934 + }, + { + "epoch": 0.636913933524103, + "grad_norm": 5.769810386371684, + "learning_rate": 3.076666410584177e-06, + "loss": 0.3466, + "step": 28935 + }, + { + "epoch": 0.6369359454105217, + "grad_norm": 1.5572853761223489, + "learning_rate": 3.0763373864390256e-06, + "loss": 0.3212, + "step": 28936 + }, + { + "epoch": 0.6369579572969404, + "grad_norm": 1.9739602914083896, + "learning_rate": 3.0760083720707846e-06, + "loss": 0.2056, + "step": 28937 + }, + { + "epoch": 0.636979969183359, + "grad_norm": 1.9023681540935615, + "learning_rate": 3.075679367481129e-06, + "loss": 0.2526, + "step": 28938 + }, + { + "epoch": 0.6370019810697777, + "grad_norm": 1.6741754589685973, + "learning_rate": 3.075350372671726e-06, + "loss": 0.2733, + "step": 28939 + }, + { + "epoch": 0.6370239929561964, + "grad_norm": 1.7187459051413194, + "learning_rate": 3.0750213876442536e-06, + "loss": 0.252, + "step": 28940 + }, + { + "epoch": 0.637046004842615, + "grad_norm": 1.7099842057541426, + "learning_rate": 3.0746924124003806e-06, + "loss": 0.3795, + "step": 28941 + }, + { + "epoch": 0.6370680167290337, + "grad_norm": 0.6058811367241643, + "learning_rate": 3.074363446941778e-06, + "loss": 0.4471, + "step": 28942 + }, + { + "epoch": 0.6370900286154524, + "grad_norm": 2.758970398841667, + "learning_rate": 3.0740344912701213e-06, + "loss": 0.3425, + "step": 28943 + }, + { + "epoch": 0.637112040501871, + "grad_norm": 3.68446581625791, + "learning_rate": 3.073705545387079e-06, + "loss": 0.3136, + "step": 28944 + }, + { + "epoch": 0.6371340523882897, + "grad_norm": 1.432179042495415, + "learning_rate": 3.0733766092943247e-06, + "loss": 0.2657, + "step": 28945 + }, + { + "epoch": 0.6371560642747084, + "grad_norm": 5.650220054259968, + "learning_rate": 3.0730476829935294e-06, + "loss": 0.3716, + "step": 28946 + }, + { + "epoch": 0.637178076161127, + "grad_norm": 3.2668794828849275, + "learning_rate": 3.0727187664863645e-06, + "loss": 0.296, + "step": 28947 + }, + { + "epoch": 0.6372000880475457, + "grad_norm": 2.7577200451671455, + "learning_rate": 3.072389859774505e-06, + "loss": 0.2887, + "step": 28948 + }, + { + "epoch": 0.6372220999339644, + "grad_norm": 1.71372133183457, + "learning_rate": 3.072060962859618e-06, + "loss": 0.3, + "step": 28949 + }, + { + "epoch": 0.637244111820383, + "grad_norm": 2.6486224607456705, + "learning_rate": 3.0717320757433766e-06, + "loss": 0.2778, + "step": 28950 + }, + { + "epoch": 0.6372661237068017, + "grad_norm": 1.4558603790333906, + "learning_rate": 3.071403198427454e-06, + "loss": 0.3031, + "step": 28951 + }, + { + "epoch": 0.6372881355932203, + "grad_norm": 1.9844406081791104, + "learning_rate": 3.0710743309135195e-06, + "loss": 0.3265, + "step": 28952 + }, + { + "epoch": 0.637310147479639, + "grad_norm": 1.5049420792696824, + "learning_rate": 3.070745473203247e-06, + "loss": 0.2999, + "step": 28953 + }, + { + "epoch": 0.6373321593660577, + "grad_norm": 1.6591162588728021, + "learning_rate": 3.070416625298306e-06, + "loss": 0.3731, + "step": 28954 + }, + { + "epoch": 0.6373541712524763, + "grad_norm": 2.684354477407238, + "learning_rate": 3.0700877872003677e-06, + "loss": 0.2928, + "step": 28955 + }, + { + "epoch": 0.637376183138895, + "grad_norm": 2.3085001538820724, + "learning_rate": 3.0697589589111053e-06, + "loss": 0.3196, + "step": 28956 + }, + { + "epoch": 0.6373981950253137, + "grad_norm": 2.5362437998498257, + "learning_rate": 3.0694301404321857e-06, + "loss": 0.2788, + "step": 28957 + }, + { + "epoch": 0.6374202069117323, + "grad_norm": 2.0773699430675925, + "learning_rate": 3.069101331765287e-06, + "loss": 0.3157, + "step": 28958 + }, + { + "epoch": 0.637442218798151, + "grad_norm": 0.6145330303950394, + "learning_rate": 3.068772532912075e-06, + "loss": 0.4481, + "step": 28959 + }, + { + "epoch": 0.6374642306845697, + "grad_norm": 0.6333636866268232, + "learning_rate": 3.068443743874221e-06, + "loss": 0.4483, + "step": 28960 + }, + { + "epoch": 0.6374862425709883, + "grad_norm": 3.447540438298685, + "learning_rate": 3.0681149646534003e-06, + "loss": 0.3737, + "step": 28961 + }, + { + "epoch": 0.637508254457407, + "grad_norm": 2.3338723374827133, + "learning_rate": 3.067786195251278e-06, + "loss": 0.3293, + "step": 28962 + }, + { + "epoch": 0.6375302663438257, + "grad_norm": 1.5757124302091159, + "learning_rate": 3.0674574356695297e-06, + "loss": 0.3359, + "step": 28963 + }, + { + "epoch": 0.6375522782302443, + "grad_norm": 2.598173956068253, + "learning_rate": 3.0671286859098244e-06, + "loss": 0.2555, + "step": 28964 + }, + { + "epoch": 0.637574290116663, + "grad_norm": 1.749903739492463, + "learning_rate": 3.0667999459738322e-06, + "loss": 0.2983, + "step": 28965 + }, + { + "epoch": 0.6375963020030817, + "grad_norm": 1.565704391980449, + "learning_rate": 3.066471215863227e-06, + "loss": 0.2558, + "step": 28966 + }, + { + "epoch": 0.6376183138895003, + "grad_norm": 0.6294537961471438, + "learning_rate": 3.0661424955796766e-06, + "loss": 0.4776, + "step": 28967 + }, + { + "epoch": 0.637640325775919, + "grad_norm": 2.93983051487235, + "learning_rate": 3.0658137851248517e-06, + "loss": 0.2883, + "step": 28968 + }, + { + "epoch": 0.6376623376623377, + "grad_norm": 0.6330866923586559, + "learning_rate": 3.0654850845004248e-06, + "loss": 0.4812, + "step": 28969 + }, + { + "epoch": 0.6376843495487563, + "grad_norm": 2.0175304822645663, + "learning_rate": 3.0651563937080632e-06, + "loss": 0.2961, + "step": 28970 + }, + { + "epoch": 0.637706361435175, + "grad_norm": 1.7650583055139883, + "learning_rate": 3.064827712749443e-06, + "loss": 0.3215, + "step": 28971 + }, + { + "epoch": 0.6377283733215937, + "grad_norm": 1.8267053215968012, + "learning_rate": 3.0644990416262304e-06, + "loss": 0.2879, + "step": 28972 + }, + { + "epoch": 0.6377503852080123, + "grad_norm": 0.6296086269128038, + "learning_rate": 3.0641703803400955e-06, + "loss": 0.447, + "step": 28973 + }, + { + "epoch": 0.637772397094431, + "grad_norm": 1.895484553263406, + "learning_rate": 3.0638417288927125e-06, + "loss": 0.3127, + "step": 28974 + }, + { + "epoch": 0.6377944089808497, + "grad_norm": 2.079968830039892, + "learning_rate": 3.0635130872857467e-06, + "loss": 0.3143, + "step": 28975 + }, + { + "epoch": 0.6378164208672683, + "grad_norm": 1.534663628328464, + "learning_rate": 3.0631844555208734e-06, + "loss": 0.2639, + "step": 28976 + }, + { + "epoch": 0.637838432753687, + "grad_norm": 1.9273048315551549, + "learning_rate": 3.06285583359976e-06, + "loss": 0.2862, + "step": 28977 + }, + { + "epoch": 0.6378604446401056, + "grad_norm": 1.108179589687493, + "learning_rate": 3.062527221524076e-06, + "loss": 0.2643, + "step": 28978 + }, + { + "epoch": 0.6378824565265243, + "grad_norm": 1.9143707592096417, + "learning_rate": 3.062198619295495e-06, + "loss": 0.2352, + "step": 28979 + }, + { + "epoch": 0.637904468412943, + "grad_norm": 1.397512814050876, + "learning_rate": 3.0618700269156826e-06, + "loss": 0.2171, + "step": 28980 + }, + { + "epoch": 0.6379264802993616, + "grad_norm": 1.6778585963412627, + "learning_rate": 3.0615414443863125e-06, + "loss": 0.2458, + "step": 28981 + }, + { + "epoch": 0.6379484921857803, + "grad_norm": 1.338851332959788, + "learning_rate": 3.0612128717090528e-06, + "loss": 0.2468, + "step": 28982 + }, + { + "epoch": 0.637970504072199, + "grad_norm": 2.4456163857074293, + "learning_rate": 3.0608843088855734e-06, + "loss": 0.2733, + "step": 28983 + }, + { + "epoch": 0.6379925159586176, + "grad_norm": 1.8543879517854454, + "learning_rate": 3.060555755917547e-06, + "loss": 0.265, + "step": 28984 + }, + { + "epoch": 0.6380145278450363, + "grad_norm": 0.6098216605344007, + "learning_rate": 3.06022721280664e-06, + "loss": 0.4482, + "step": 28985 + }, + { + "epoch": 0.638036539731455, + "grad_norm": 1.6047967041024405, + "learning_rate": 3.0598986795545225e-06, + "loss": 0.2821, + "step": 28986 + }, + { + "epoch": 0.6380585516178736, + "grad_norm": 0.6277274694143946, + "learning_rate": 3.0595701561628676e-06, + "loss": 0.4755, + "step": 28987 + }, + { + "epoch": 0.6380805635042923, + "grad_norm": 1.785360427303744, + "learning_rate": 3.059241642633339e-06, + "loss": 0.2181, + "step": 28988 + }, + { + "epoch": 0.638102575390711, + "grad_norm": 1.8845954030471441, + "learning_rate": 3.0589131389676137e-06, + "loss": 0.3749, + "step": 28989 + }, + { + "epoch": 0.6381245872771296, + "grad_norm": 1.5142314070718015, + "learning_rate": 3.0585846451673563e-06, + "loss": 0.2553, + "step": 28990 + }, + { + "epoch": 0.6381465991635483, + "grad_norm": 1.7050583641638322, + "learning_rate": 3.058256161234236e-06, + "loss": 0.2336, + "step": 28991 + }, + { + "epoch": 0.638168611049967, + "grad_norm": 1.8932184625617807, + "learning_rate": 3.0579276871699264e-06, + "loss": 0.3178, + "step": 28992 + }, + { + "epoch": 0.6381906229363856, + "grad_norm": 1.9307219699874618, + "learning_rate": 3.0575992229760916e-06, + "loss": 0.2949, + "step": 28993 + }, + { + "epoch": 0.6382126348228043, + "grad_norm": 2.4116696334937306, + "learning_rate": 3.057270768654407e-06, + "loss": 0.3921, + "step": 28994 + }, + { + "epoch": 0.638234646709223, + "grad_norm": 1.6795404991181622, + "learning_rate": 3.0569423242065377e-06, + "loss": 0.2402, + "step": 28995 + }, + { + "epoch": 0.6382566585956416, + "grad_norm": 0.614326360442173, + "learning_rate": 3.0566138896341524e-06, + "loss": 0.462, + "step": 28996 + }, + { + "epoch": 0.6382786704820603, + "grad_norm": 2.302597161985636, + "learning_rate": 3.056285464938925e-06, + "loss": 0.3006, + "step": 28997 + }, + { + "epoch": 0.638300682368479, + "grad_norm": 1.5325548261022996, + "learning_rate": 3.0559570501225194e-06, + "loss": 0.319, + "step": 28998 + }, + { + "epoch": 0.6383226942548976, + "grad_norm": 0.6206020921941855, + "learning_rate": 3.0556286451866078e-06, + "loss": 0.4541, + "step": 28999 + }, + { + "epoch": 0.6383447061413163, + "grad_norm": 2.071232560313089, + "learning_rate": 3.0553002501328597e-06, + "loss": 0.3531, + "step": 29000 + }, + { + "epoch": 0.6383667180277349, + "grad_norm": 1.6754943467963226, + "learning_rate": 3.0549718649629413e-06, + "loss": 0.3367, + "step": 29001 + }, + { + "epoch": 0.6383887299141536, + "grad_norm": 1.73104729623289, + "learning_rate": 3.054643489678526e-06, + "loss": 0.2266, + "step": 29002 + }, + { + "epoch": 0.6384107418005723, + "grad_norm": 1.6942938724322585, + "learning_rate": 3.0543151242812776e-06, + "loss": 0.2617, + "step": 29003 + }, + { + "epoch": 0.6384327536869909, + "grad_norm": 2.535033550814877, + "learning_rate": 3.0539867687728686e-06, + "loss": 0.3399, + "step": 29004 + }, + { + "epoch": 0.6384547655734096, + "grad_norm": 1.167969880093586, + "learning_rate": 3.0536584231549678e-06, + "loss": 0.2249, + "step": 29005 + }, + { + "epoch": 0.6384767774598283, + "grad_norm": 1.95062914537893, + "learning_rate": 3.05333008742924e-06, + "loss": 0.3166, + "step": 29006 + }, + { + "epoch": 0.6384987893462469, + "grad_norm": 1.8365074845832463, + "learning_rate": 3.0530017615973596e-06, + "loss": 0.3497, + "step": 29007 + }, + { + "epoch": 0.6385208012326656, + "grad_norm": 1.442446544805627, + "learning_rate": 3.052673445660992e-06, + "loss": 0.2864, + "step": 29008 + }, + { + "epoch": 0.6385428131190843, + "grad_norm": 0.6411872956339478, + "learning_rate": 3.0523451396218055e-06, + "loss": 0.4629, + "step": 29009 + }, + { + "epoch": 0.6385648250055029, + "grad_norm": 0.6156181496328565, + "learning_rate": 3.0520168434814717e-06, + "loss": 0.4616, + "step": 29010 + }, + { + "epoch": 0.6385868368919216, + "grad_norm": 1.9905417395146494, + "learning_rate": 3.0516885572416545e-06, + "loss": 0.4046, + "step": 29011 + }, + { + "epoch": 0.6386088487783403, + "grad_norm": 1.1185957613935484, + "learning_rate": 3.0513602809040277e-06, + "loss": 0.2641, + "step": 29012 + }, + { + "epoch": 0.6386308606647589, + "grad_norm": 1.7360039512947338, + "learning_rate": 3.0510320144702553e-06, + "loss": 0.3117, + "step": 29013 + }, + { + "epoch": 0.6386528725511776, + "grad_norm": 2.9085881044437816, + "learning_rate": 3.050703757942007e-06, + "loss": 0.219, + "step": 29014 + }, + { + "epoch": 0.6386748844375963, + "grad_norm": 2.0652725204161855, + "learning_rate": 3.050375511320954e-06, + "loss": 0.366, + "step": 29015 + }, + { + "epoch": 0.6386968963240149, + "grad_norm": 1.8511402333921108, + "learning_rate": 3.0500472746087602e-06, + "loss": 0.4033, + "step": 29016 + }, + { + "epoch": 0.6387189082104336, + "grad_norm": 1.6104059090464036, + "learning_rate": 3.0497190478070966e-06, + "loss": 0.2792, + "step": 29017 + }, + { + "epoch": 0.6387409200968523, + "grad_norm": 0.6113886627343056, + "learning_rate": 3.0493908309176306e-06, + "loss": 0.4519, + "step": 29018 + }, + { + "epoch": 0.6387629319832709, + "grad_norm": 0.6228101398252525, + "learning_rate": 3.0490626239420296e-06, + "loss": 0.4803, + "step": 29019 + }, + { + "epoch": 0.6387849438696896, + "grad_norm": 3.7719291092487275, + "learning_rate": 3.048734426881964e-06, + "loss": 0.2709, + "step": 29020 + }, + { + "epoch": 0.6388069557561084, + "grad_norm": 1.8438254870074318, + "learning_rate": 3.048406239739099e-06, + "loss": 0.3699, + "step": 29021 + }, + { + "epoch": 0.638828967642527, + "grad_norm": 1.28029326006266, + "learning_rate": 3.048078062515104e-06, + "loss": 0.2738, + "step": 29022 + }, + { + "epoch": 0.6388509795289457, + "grad_norm": 0.6139335088003134, + "learning_rate": 3.0477498952116493e-06, + "loss": 0.4599, + "step": 29023 + }, + { + "epoch": 0.6388729914153642, + "grad_norm": 1.5235288719601234, + "learning_rate": 3.047421737830396e-06, + "loss": 0.397, + "step": 29024 + }, + { + "epoch": 0.638895003301783, + "grad_norm": 1.4544794857437167, + "learning_rate": 3.0470935903730204e-06, + "loss": 0.3224, + "step": 29025 + }, + { + "epoch": 0.6389170151882017, + "grad_norm": 1.4268145936892638, + "learning_rate": 3.046765452841184e-06, + "loss": 0.2945, + "step": 29026 + }, + { + "epoch": 0.6389390270746202, + "grad_norm": 2.2577191735714037, + "learning_rate": 3.0464373252365563e-06, + "loss": 0.2526, + "step": 29027 + }, + { + "epoch": 0.638961038961039, + "grad_norm": 3.883570195181307, + "learning_rate": 3.046109207560808e-06, + "loss": 0.2269, + "step": 29028 + }, + { + "epoch": 0.6389830508474577, + "grad_norm": 1.8859440523347408, + "learning_rate": 3.0457810998156003e-06, + "loss": 0.2993, + "step": 29029 + }, + { + "epoch": 0.6390050627338763, + "grad_norm": 1.3011925225627508, + "learning_rate": 3.045453002002608e-06, + "loss": 0.2596, + "step": 29030 + }, + { + "epoch": 0.639027074620295, + "grad_norm": 2.019381004282107, + "learning_rate": 3.045124914123494e-06, + "loss": 0.2555, + "step": 29031 + }, + { + "epoch": 0.6390490865067137, + "grad_norm": 1.370699238291529, + "learning_rate": 3.0447968361799267e-06, + "loss": 0.3119, + "step": 29032 + }, + { + "epoch": 0.6390710983931323, + "grad_norm": 0.6223118507861918, + "learning_rate": 3.0444687681735757e-06, + "loss": 0.4448, + "step": 29033 + }, + { + "epoch": 0.639093110279551, + "grad_norm": 2.2771827198810097, + "learning_rate": 3.0441407101061043e-06, + "loss": 0.2521, + "step": 29034 + }, + { + "epoch": 0.6391151221659697, + "grad_norm": 1.702122915484603, + "learning_rate": 3.0438126619791837e-06, + "loss": 0.32, + "step": 29035 + }, + { + "epoch": 0.6391371340523883, + "grad_norm": 2.5827914879871345, + "learning_rate": 3.043484623794479e-06, + "loss": 0.1778, + "step": 29036 + }, + { + "epoch": 0.639159145938807, + "grad_norm": 1.7891071753982892, + "learning_rate": 3.0431565955536573e-06, + "loss": 0.3824, + "step": 29037 + }, + { + "epoch": 0.6391811578252257, + "grad_norm": 1.646793164766865, + "learning_rate": 3.042828577258389e-06, + "loss": 0.2638, + "step": 29038 + }, + { + "epoch": 0.6392031697116443, + "grad_norm": 4.047313009488764, + "learning_rate": 3.042500568910336e-06, + "loss": 0.2812, + "step": 29039 + }, + { + "epoch": 0.639225181598063, + "grad_norm": 0.6284564042070796, + "learning_rate": 3.04217257051117e-06, + "loss": 0.4736, + "step": 29040 + }, + { + "epoch": 0.6392471934844817, + "grad_norm": 2.7802195975001838, + "learning_rate": 3.0418445820625563e-06, + "loss": 0.3472, + "step": 29041 + }, + { + "epoch": 0.6392692053709003, + "grad_norm": 1.7056110771727364, + "learning_rate": 3.0415166035661596e-06, + "loss": 0.3653, + "step": 29042 + }, + { + "epoch": 0.639291217257319, + "grad_norm": 1.812079496566676, + "learning_rate": 3.041188635023652e-06, + "loss": 0.3269, + "step": 29043 + }, + { + "epoch": 0.6393132291437377, + "grad_norm": 2.0996180067308097, + "learning_rate": 3.040860676436695e-06, + "loss": 0.3992, + "step": 29044 + }, + { + "epoch": 0.6393352410301563, + "grad_norm": 1.5566769878016538, + "learning_rate": 3.040532727806959e-06, + "loss": 0.2136, + "step": 29045 + }, + { + "epoch": 0.639357252916575, + "grad_norm": 1.399625153431324, + "learning_rate": 3.040204789136111e-06, + "loss": 0.31, + "step": 29046 + }, + { + "epoch": 0.6393792648029936, + "grad_norm": 2.0656494962559053, + "learning_rate": 3.039876860425812e-06, + "loss": 0.2823, + "step": 29047 + }, + { + "epoch": 0.6394012766894123, + "grad_norm": 2.6733998062171342, + "learning_rate": 3.039548941677738e-06, + "loss": 0.2965, + "step": 29048 + }, + { + "epoch": 0.639423288575831, + "grad_norm": 2.0576010443137824, + "learning_rate": 3.0392210328935483e-06, + "loss": 0.3564, + "step": 29049 + }, + { + "epoch": 0.6394453004622496, + "grad_norm": 2.5754748661214903, + "learning_rate": 3.0388931340749113e-06, + "loss": 0.2778, + "step": 29050 + }, + { + "epoch": 0.6394673123486683, + "grad_norm": 1.5377440456695277, + "learning_rate": 3.038565245223496e-06, + "loss": 0.2251, + "step": 29051 + }, + { + "epoch": 0.639489324235087, + "grad_norm": 7.117550427411726, + "learning_rate": 3.038237366340965e-06, + "loss": 0.3602, + "step": 29052 + }, + { + "epoch": 0.6395113361215056, + "grad_norm": 2.4928982358169516, + "learning_rate": 3.0379094974289872e-06, + "loss": 0.3354, + "step": 29053 + }, + { + "epoch": 0.6395333480079243, + "grad_norm": 0.6201870944606127, + "learning_rate": 3.037581638489229e-06, + "loss": 0.4632, + "step": 29054 + }, + { + "epoch": 0.639555359894343, + "grad_norm": 1.9177783316292336, + "learning_rate": 3.037253789523355e-06, + "loss": 0.3564, + "step": 29055 + }, + { + "epoch": 0.6395773717807616, + "grad_norm": 2.2048981756081147, + "learning_rate": 3.036925950533035e-06, + "loss": 0.3556, + "step": 29056 + }, + { + "epoch": 0.6395993836671803, + "grad_norm": 1.378611610707692, + "learning_rate": 3.03659812151993e-06, + "loss": 0.2835, + "step": 29057 + }, + { + "epoch": 0.639621395553599, + "grad_norm": 1.7305692982053686, + "learning_rate": 3.036270302485711e-06, + "loss": 0.4163, + "step": 29058 + }, + { + "epoch": 0.6396434074400176, + "grad_norm": 2.4337343683275448, + "learning_rate": 3.0359424934320426e-06, + "loss": 0.1988, + "step": 29059 + }, + { + "epoch": 0.6396654193264363, + "grad_norm": 1.835157513193283, + "learning_rate": 3.035614694360588e-06, + "loss": 0.2664, + "step": 29060 + }, + { + "epoch": 0.639687431212855, + "grad_norm": 9.370915984687736, + "learning_rate": 3.0352869052730184e-06, + "loss": 0.2673, + "step": 29061 + }, + { + "epoch": 0.6397094430992736, + "grad_norm": 0.6128717822732, + "learning_rate": 3.0349591261709943e-06, + "loss": 0.4566, + "step": 29062 + }, + { + "epoch": 0.6397314549856923, + "grad_norm": 1.5303817899624863, + "learning_rate": 3.0346313570561865e-06, + "loss": 0.2848, + "step": 29063 + }, + { + "epoch": 0.639753466872111, + "grad_norm": 0.6050659738353423, + "learning_rate": 3.034303597930259e-06, + "loss": 0.4526, + "step": 29064 + }, + { + "epoch": 0.6397754787585296, + "grad_norm": 2.182953307318862, + "learning_rate": 3.0339758487948745e-06, + "loss": 0.3154, + "step": 29065 + }, + { + "epoch": 0.6397974906449483, + "grad_norm": 1.6406344902960646, + "learning_rate": 3.0336481096517046e-06, + "loss": 0.2082, + "step": 29066 + }, + { + "epoch": 0.639819502531367, + "grad_norm": 2.289832133438846, + "learning_rate": 3.0333203805024113e-06, + "loss": 0.3479, + "step": 29067 + }, + { + "epoch": 0.6398415144177856, + "grad_norm": 0.6045648772361513, + "learning_rate": 3.0329926613486595e-06, + "loss": 0.4717, + "step": 29068 + }, + { + "epoch": 0.6398635263042043, + "grad_norm": 0.6249959967489601, + "learning_rate": 3.0326649521921186e-06, + "loss": 0.4915, + "step": 29069 + }, + { + "epoch": 0.6398855381906229, + "grad_norm": 3.019409142754754, + "learning_rate": 3.0323372530344498e-06, + "loss": 0.2966, + "step": 29070 + }, + { + "epoch": 0.6399075500770416, + "grad_norm": 1.8991538425744672, + "learning_rate": 3.0320095638773216e-06, + "loss": 0.3266, + "step": 29071 + }, + { + "epoch": 0.6399295619634603, + "grad_norm": 1.729139149478849, + "learning_rate": 3.031681884722399e-06, + "loss": 0.2751, + "step": 29072 + }, + { + "epoch": 0.6399515738498789, + "grad_norm": 0.6353736102650894, + "learning_rate": 3.0313542155713453e-06, + "loss": 0.4684, + "step": 29073 + }, + { + "epoch": 0.6399735857362976, + "grad_norm": 1.4050418410597298, + "learning_rate": 3.03102655642583e-06, + "loss": 0.3349, + "step": 29074 + }, + { + "epoch": 0.6399955976227163, + "grad_norm": 1.5742529319740353, + "learning_rate": 3.030698907287514e-06, + "loss": 0.3031, + "step": 29075 + }, + { + "epoch": 0.6400176095091349, + "grad_norm": 1.5404412212353773, + "learning_rate": 3.030371268158065e-06, + "loss": 0.3236, + "step": 29076 + }, + { + "epoch": 0.6400396213955536, + "grad_norm": 4.401201781827139, + "learning_rate": 3.0300436390391492e-06, + "loss": 0.2839, + "step": 29077 + }, + { + "epoch": 0.6400616332819723, + "grad_norm": 1.4097038739712486, + "learning_rate": 3.029716019932427e-06, + "loss": 0.3485, + "step": 29078 + }, + { + "epoch": 0.6400836451683909, + "grad_norm": 12.18504982789823, + "learning_rate": 3.02938841083957e-06, + "loss": 0.3226, + "step": 29079 + }, + { + "epoch": 0.6401056570548096, + "grad_norm": 4.15474289889367, + "learning_rate": 3.029060811762237e-06, + "loss": 0.3459, + "step": 29080 + }, + { + "epoch": 0.6401276689412283, + "grad_norm": 1.320647146706306, + "learning_rate": 3.0287332227020983e-06, + "loss": 0.2905, + "step": 29081 + }, + { + "epoch": 0.6401496808276469, + "grad_norm": 2.522879460080388, + "learning_rate": 3.0284056436608165e-06, + "loss": 0.3198, + "step": 29082 + }, + { + "epoch": 0.6401716927140656, + "grad_norm": 2.57107413107341, + "learning_rate": 3.028078074640055e-06, + "loss": 0.3276, + "step": 29083 + }, + { + "epoch": 0.6401937046004843, + "grad_norm": 3.5395680198351975, + "learning_rate": 3.027750515641482e-06, + "loss": 0.3425, + "step": 29084 + }, + { + "epoch": 0.6402157164869029, + "grad_norm": 2.0231123816328798, + "learning_rate": 3.02742296666676e-06, + "loss": 0.303, + "step": 29085 + }, + { + "epoch": 0.6402377283733216, + "grad_norm": 0.6061078972943883, + "learning_rate": 3.0270954277175528e-06, + "loss": 0.4874, + "step": 29086 + }, + { + "epoch": 0.6402597402597403, + "grad_norm": 1.4643289453618293, + "learning_rate": 3.026767898795529e-06, + "loss": 0.2672, + "step": 29087 + }, + { + "epoch": 0.6402817521461589, + "grad_norm": 2.1957650497355585, + "learning_rate": 3.026440379902349e-06, + "loss": 0.3143, + "step": 29088 + }, + { + "epoch": 0.6403037640325776, + "grad_norm": 1.2622629479416763, + "learning_rate": 3.0261128710396796e-06, + "loss": 0.2233, + "step": 29089 + }, + { + "epoch": 0.6403257759189963, + "grad_norm": 1.5616181609002713, + "learning_rate": 3.0257853722091856e-06, + "loss": 0.3467, + "step": 29090 + }, + { + "epoch": 0.6403477878054149, + "grad_norm": 1.8398869666900632, + "learning_rate": 3.025457883412529e-06, + "loss": 0.3102, + "step": 29091 + }, + { + "epoch": 0.6403697996918336, + "grad_norm": 0.596411079938935, + "learning_rate": 3.025130404651379e-06, + "loss": 0.453, + "step": 29092 + }, + { + "epoch": 0.6403918115782522, + "grad_norm": 2.705956005056531, + "learning_rate": 3.024802935927395e-06, + "loss": 0.368, + "step": 29093 + }, + { + "epoch": 0.6404138234646709, + "grad_norm": 2.044364647875717, + "learning_rate": 3.0244754772422445e-06, + "loss": 0.3187, + "step": 29094 + }, + { + "epoch": 0.6404358353510896, + "grad_norm": 1.5573833098239676, + "learning_rate": 3.024148028597591e-06, + "loss": 0.3269, + "step": 29095 + }, + { + "epoch": 0.6404578472375082, + "grad_norm": 0.6187122140391351, + "learning_rate": 3.0238205899950962e-06, + "loss": 0.4479, + "step": 29096 + }, + { + "epoch": 0.6404798591239269, + "grad_norm": 1.546440846005272, + "learning_rate": 3.0234931614364295e-06, + "loss": 0.1771, + "step": 29097 + }, + { + "epoch": 0.6405018710103456, + "grad_norm": 1.4177632966418445, + "learning_rate": 3.0231657429232496e-06, + "loss": 0.2847, + "step": 29098 + }, + { + "epoch": 0.6405238828967642, + "grad_norm": 1.7321550278693174, + "learning_rate": 3.022838334457224e-06, + "loss": 0.2847, + "step": 29099 + }, + { + "epoch": 0.6405458947831829, + "grad_norm": 1.534336983086439, + "learning_rate": 3.022510936040018e-06, + "loss": 0.2541, + "step": 29100 + }, + { + "epoch": 0.6405679066696016, + "grad_norm": 0.619131357839906, + "learning_rate": 3.022183547673291e-06, + "loss": 0.4523, + "step": 29101 + }, + { + "epoch": 0.6405899185560202, + "grad_norm": 4.510967997609332, + "learning_rate": 3.0218561693587095e-06, + "loss": 0.3085, + "step": 29102 + }, + { + "epoch": 0.6406119304424389, + "grad_norm": 234.4033272049685, + "learning_rate": 3.0215288010979365e-06, + "loss": 0.309, + "step": 29103 + }, + { + "epoch": 0.6406339423288576, + "grad_norm": 0.6563967366903486, + "learning_rate": 3.021201442892638e-06, + "loss": 0.4906, + "step": 29104 + }, + { + "epoch": 0.6406559542152762, + "grad_norm": 1.8921830074640742, + "learning_rate": 3.020874094744477e-06, + "loss": 0.2311, + "step": 29105 + }, + { + "epoch": 0.6406779661016949, + "grad_norm": 0.6229845998041444, + "learning_rate": 3.020546756655115e-06, + "loss": 0.4744, + "step": 29106 + }, + { + "epoch": 0.6406999779881136, + "grad_norm": 0.6321306972833955, + "learning_rate": 3.0202194286262175e-06, + "loss": 0.4645, + "step": 29107 + }, + { + "epoch": 0.6407219898745322, + "grad_norm": 1.4356764678593397, + "learning_rate": 3.0198921106594482e-06, + "loss": 0.2527, + "step": 29108 + }, + { + "epoch": 0.6407440017609509, + "grad_norm": 1.584842461309313, + "learning_rate": 3.0195648027564693e-06, + "loss": 0.237, + "step": 29109 + }, + { + "epoch": 0.6407660136473696, + "grad_norm": 0.63824455195212, + "learning_rate": 3.019237504918948e-06, + "loss": 0.4711, + "step": 29110 + }, + { + "epoch": 0.6407880255337882, + "grad_norm": 0.5855041366242886, + "learning_rate": 3.018910217148542e-06, + "loss": 0.4551, + "step": 29111 + }, + { + "epoch": 0.6408100374202069, + "grad_norm": 1.639935835955535, + "learning_rate": 3.0185829394469202e-06, + "loss": 0.235, + "step": 29112 + }, + { + "epoch": 0.6408320493066256, + "grad_norm": 0.6220559216525308, + "learning_rate": 3.018255671815744e-06, + "loss": 0.4582, + "step": 29113 + }, + { + "epoch": 0.6408540611930442, + "grad_norm": 2.4689228976917743, + "learning_rate": 3.0179284142566733e-06, + "loss": 0.289, + "step": 29114 + }, + { + "epoch": 0.6408760730794629, + "grad_norm": 2.134950728714514, + "learning_rate": 3.017601166771377e-06, + "loss": 0.3302, + "step": 29115 + }, + { + "epoch": 0.6408980849658816, + "grad_norm": 2.1992822489062887, + "learning_rate": 3.0172739293615146e-06, + "loss": 0.2583, + "step": 29116 + }, + { + "epoch": 0.6409200968523002, + "grad_norm": 1.7168948185536566, + "learning_rate": 3.016946702028751e-06, + "loss": 0.313, + "step": 29117 + }, + { + "epoch": 0.6409421087387189, + "grad_norm": 2.6275254482637584, + "learning_rate": 3.0166194847747497e-06, + "loss": 0.3051, + "step": 29118 + }, + { + "epoch": 0.6409641206251375, + "grad_norm": 2.4902391845658842, + "learning_rate": 3.0162922776011707e-06, + "loss": 0.2986, + "step": 29119 + }, + { + "epoch": 0.6409861325115562, + "grad_norm": 2.012378457489281, + "learning_rate": 3.01596508050968e-06, + "loss": 0.3534, + "step": 29120 + }, + { + "epoch": 0.6410081443979749, + "grad_norm": 0.6112353689849124, + "learning_rate": 3.015637893501939e-06, + "loss": 0.4558, + "step": 29121 + }, + { + "epoch": 0.6410301562843935, + "grad_norm": 1.7540312011797552, + "learning_rate": 3.0153107165796124e-06, + "loss": 0.2104, + "step": 29122 + }, + { + "epoch": 0.6410521681708122, + "grad_norm": 1.9687412598932756, + "learning_rate": 3.0149835497443627e-06, + "loss": 0.3305, + "step": 29123 + }, + { + "epoch": 0.6410741800572309, + "grad_norm": 1.6376650888279165, + "learning_rate": 3.01465639299785e-06, + "loss": 0.2942, + "step": 29124 + }, + { + "epoch": 0.6410961919436495, + "grad_norm": 1.7889477863489507, + "learning_rate": 3.0143292463417393e-06, + "loss": 0.3112, + "step": 29125 + }, + { + "epoch": 0.6411182038300682, + "grad_norm": 1.7055909761472439, + "learning_rate": 3.0140021097776938e-06, + "loss": 0.2935, + "step": 29126 + }, + { + "epoch": 0.6411402157164869, + "grad_norm": 0.6100415006043637, + "learning_rate": 3.0136749833073743e-06, + "loss": 0.4567, + "step": 29127 + }, + { + "epoch": 0.6411622276029055, + "grad_norm": 1.6057157429312294, + "learning_rate": 3.0133478669324467e-06, + "loss": 0.3016, + "step": 29128 + }, + { + "epoch": 0.6411842394893242, + "grad_norm": 1.6710853496538032, + "learning_rate": 3.0130207606545686e-06, + "loss": 0.316, + "step": 29129 + }, + { + "epoch": 0.6412062513757429, + "grad_norm": 2.935671870456815, + "learning_rate": 3.012693664475407e-06, + "loss": 0.2669, + "step": 29130 + }, + { + "epoch": 0.6412282632621615, + "grad_norm": 6.430854952017451, + "learning_rate": 3.0123665783966226e-06, + "loss": 0.2346, + "step": 29131 + }, + { + "epoch": 0.6412502751485802, + "grad_norm": 1.980816211322697, + "learning_rate": 3.0120395024198756e-06, + "loss": 0.3337, + "step": 29132 + }, + { + "epoch": 0.641272287034999, + "grad_norm": 0.6132985143804389, + "learning_rate": 3.011712436546833e-06, + "loss": 0.4465, + "step": 29133 + }, + { + "epoch": 0.6412942989214175, + "grad_norm": 1.6123701002007371, + "learning_rate": 3.0113853807791526e-06, + "loss": 0.3725, + "step": 29134 + }, + { + "epoch": 0.6413163108078362, + "grad_norm": 3.655699945266288, + "learning_rate": 3.0110583351185003e-06, + "loss": 0.3007, + "step": 29135 + }, + { + "epoch": 0.641338322694255, + "grad_norm": 2.595265031075606, + "learning_rate": 3.010731299566537e-06, + "loss": 0.279, + "step": 29136 + }, + { + "epoch": 0.6413603345806735, + "grad_norm": 1.521471822522828, + "learning_rate": 3.010404274124923e-06, + "loss": 0.2749, + "step": 29137 + }, + { + "epoch": 0.6413823464670922, + "grad_norm": 2.097630958322831, + "learning_rate": 3.010077258795322e-06, + "loss": 0.3101, + "step": 29138 + }, + { + "epoch": 0.641404358353511, + "grad_norm": 1.4808433159595493, + "learning_rate": 3.0097502535793953e-06, + "loss": 0.2302, + "step": 29139 + }, + { + "epoch": 0.6414263702399295, + "grad_norm": 1.7360522406965377, + "learning_rate": 3.0094232584788068e-06, + "loss": 0.2829, + "step": 29140 + }, + { + "epoch": 0.6414483821263482, + "grad_norm": 2.07010006787779, + "learning_rate": 3.009096273495218e-06, + "loss": 0.2626, + "step": 29141 + }, + { + "epoch": 0.6414703940127668, + "grad_norm": 2.556702678544933, + "learning_rate": 3.0087692986302875e-06, + "loss": 0.3419, + "step": 29142 + }, + { + "epoch": 0.6414924058991855, + "grad_norm": 1.4999132063733913, + "learning_rate": 3.008442333885681e-06, + "loss": 0.2674, + "step": 29143 + }, + { + "epoch": 0.6415144177856043, + "grad_norm": 1.226589458350333, + "learning_rate": 3.008115379263058e-06, + "loss": 0.245, + "step": 29144 + }, + { + "epoch": 0.6415364296720228, + "grad_norm": 1.8023156278875667, + "learning_rate": 3.0077884347640817e-06, + "loss": 0.2778, + "step": 29145 + }, + { + "epoch": 0.6415584415584416, + "grad_norm": 0.6306064691283736, + "learning_rate": 3.007461500390414e-06, + "loss": 0.4535, + "step": 29146 + }, + { + "epoch": 0.6415804534448603, + "grad_norm": 0.6304875680004062, + "learning_rate": 3.0071345761437144e-06, + "loss": 0.4634, + "step": 29147 + }, + { + "epoch": 0.6416024653312788, + "grad_norm": 1.3874715583215362, + "learning_rate": 3.006807662025646e-06, + "loss": 0.2495, + "step": 29148 + }, + { + "epoch": 0.6416244772176976, + "grad_norm": 3.4705667303247045, + "learning_rate": 3.0064807580378706e-06, + "loss": 0.3788, + "step": 29149 + }, + { + "epoch": 0.6416464891041163, + "grad_norm": 1.5879419320909876, + "learning_rate": 3.006153864182047e-06, + "loss": 0.2971, + "step": 29150 + }, + { + "epoch": 0.6416685009905349, + "grad_norm": 1.7761701467829079, + "learning_rate": 3.0058269804598416e-06, + "loss": 0.2475, + "step": 29151 + }, + { + "epoch": 0.6416905128769536, + "grad_norm": 1.8111048394579339, + "learning_rate": 3.0055001068729107e-06, + "loss": 0.2532, + "step": 29152 + }, + { + "epoch": 0.6417125247633723, + "grad_norm": 1.5524423309107747, + "learning_rate": 3.0051732434229185e-06, + "loss": 0.3292, + "step": 29153 + }, + { + "epoch": 0.6417345366497909, + "grad_norm": 0.6467295125783594, + "learning_rate": 3.004846390111527e-06, + "loss": 0.492, + "step": 29154 + }, + { + "epoch": 0.6417565485362096, + "grad_norm": 3.499237110631492, + "learning_rate": 3.0045195469403944e-06, + "loss": 0.3475, + "step": 29155 + }, + { + "epoch": 0.6417785604226283, + "grad_norm": 3.2223972623314454, + "learning_rate": 3.0041927139111835e-06, + "loss": 0.3197, + "step": 29156 + }, + { + "epoch": 0.6418005723090469, + "grad_norm": 2.4162199357353495, + "learning_rate": 3.003865891025555e-06, + "loss": 0.2896, + "step": 29157 + }, + { + "epoch": 0.6418225841954656, + "grad_norm": 1.7802547272786386, + "learning_rate": 3.0035390782851716e-06, + "loss": 0.3157, + "step": 29158 + }, + { + "epoch": 0.6418445960818843, + "grad_norm": 0.5958723888739313, + "learning_rate": 3.0032122756916935e-06, + "loss": 0.4473, + "step": 29159 + }, + { + "epoch": 0.6418666079683029, + "grad_norm": 1.5506226572967536, + "learning_rate": 3.0028854832467797e-06, + "loss": 0.2627, + "step": 29160 + }, + { + "epoch": 0.6418886198547216, + "grad_norm": 1.9121280172650015, + "learning_rate": 3.0025587009520934e-06, + "loss": 0.2678, + "step": 29161 + }, + { + "epoch": 0.6419106317411403, + "grad_norm": 1.8460137500984732, + "learning_rate": 3.002231928809294e-06, + "loss": 0.3081, + "step": 29162 + }, + { + "epoch": 0.6419326436275589, + "grad_norm": 1.9349855661284383, + "learning_rate": 3.0019051668200434e-06, + "loss": 0.3129, + "step": 29163 + }, + { + "epoch": 0.6419546555139776, + "grad_norm": 1.6815513246083835, + "learning_rate": 3.0015784149860027e-06, + "loss": 0.3553, + "step": 29164 + }, + { + "epoch": 0.6419766674003962, + "grad_norm": 1.7864970174100103, + "learning_rate": 3.0012516733088302e-06, + "loss": 0.3182, + "step": 29165 + }, + { + "epoch": 0.6419986792868149, + "grad_norm": 1.6841489408837795, + "learning_rate": 3.00092494179019e-06, + "loss": 0.2772, + "step": 29166 + }, + { + "epoch": 0.6420206911732336, + "grad_norm": 1.7253981354037795, + "learning_rate": 3.0005982204317408e-06, + "loss": 0.2693, + "step": 29167 + }, + { + "epoch": 0.6420427030596522, + "grad_norm": 0.5899786320105551, + "learning_rate": 3.00027150923514e-06, + "loss": 0.4293, + "step": 29168 + }, + { + "epoch": 0.6420647149460709, + "grad_norm": 1.587591862877704, + "learning_rate": 2.999944808202055e-06, + "loss": 0.3879, + "step": 29169 + }, + { + "epoch": 0.6420867268324896, + "grad_norm": 3.6017250105829253, + "learning_rate": 2.9996181173341397e-06, + "loss": 0.2345, + "step": 29170 + }, + { + "epoch": 0.6421087387189082, + "grad_norm": 2.0600470309766274, + "learning_rate": 2.9992914366330594e-06, + "loss": 0.4054, + "step": 29171 + }, + { + "epoch": 0.6421307506053269, + "grad_norm": 2.3153710561894636, + "learning_rate": 2.998964766100473e-06, + "loss": 0.3422, + "step": 29172 + }, + { + "epoch": 0.6421527624917456, + "grad_norm": 0.5876937925111589, + "learning_rate": 2.9986381057380376e-06, + "loss": 0.4462, + "step": 29173 + }, + { + "epoch": 0.6421747743781642, + "grad_norm": 2.5257965480721056, + "learning_rate": 2.998311455547418e-06, + "loss": 0.2436, + "step": 29174 + }, + { + "epoch": 0.6421967862645829, + "grad_norm": 2.2041786808771024, + "learning_rate": 2.9979848155302706e-06, + "loss": 0.356, + "step": 29175 + }, + { + "epoch": 0.6422187981510016, + "grad_norm": 1.437531652580239, + "learning_rate": 2.9976581856882583e-06, + "loss": 0.3348, + "step": 29176 + }, + { + "epoch": 0.6422408100374202, + "grad_norm": 0.6158920086412834, + "learning_rate": 2.997331566023041e-06, + "loss": 0.4704, + "step": 29177 + }, + { + "epoch": 0.6422628219238389, + "grad_norm": 0.6309944923338604, + "learning_rate": 2.9970049565362765e-06, + "loss": 0.4682, + "step": 29178 + }, + { + "epoch": 0.6422848338102576, + "grad_norm": 1.935006684181213, + "learning_rate": 2.9966783572296266e-06, + "loss": 0.2919, + "step": 29179 + }, + { + "epoch": 0.6423068456966762, + "grad_norm": 1.6009884161834491, + "learning_rate": 2.9963517681047506e-06, + "loss": 0.3118, + "step": 29180 + }, + { + "epoch": 0.6423288575830949, + "grad_norm": 0.6188213652073197, + "learning_rate": 2.9960251891633084e-06, + "loss": 0.4726, + "step": 29181 + }, + { + "epoch": 0.6423508694695136, + "grad_norm": 2.0565315962181723, + "learning_rate": 2.9956986204069617e-06, + "loss": 0.3726, + "step": 29182 + }, + { + "epoch": 0.6423728813559322, + "grad_norm": 1.2788714485386785, + "learning_rate": 2.9953720618373665e-06, + "loss": 0.2993, + "step": 29183 + }, + { + "epoch": 0.6423948932423509, + "grad_norm": 1.5653378348386533, + "learning_rate": 2.9950455134561856e-06, + "loss": 0.2756, + "step": 29184 + }, + { + "epoch": 0.6424169051287696, + "grad_norm": 1.5854186588976193, + "learning_rate": 2.9947189752650783e-06, + "loss": 0.2417, + "step": 29185 + }, + { + "epoch": 0.6424389170151882, + "grad_norm": 1.5606726692815402, + "learning_rate": 2.9943924472657014e-06, + "loss": 0.2541, + "step": 29186 + }, + { + "epoch": 0.6424609289016069, + "grad_norm": 1.5295844625830368, + "learning_rate": 2.994065929459719e-06, + "loss": 0.23, + "step": 29187 + }, + { + "epoch": 0.6424829407880255, + "grad_norm": 1.6143331590860754, + "learning_rate": 2.993739421848786e-06, + "loss": 0.4521, + "step": 29188 + }, + { + "epoch": 0.6425049526744442, + "grad_norm": 1.8225834332514874, + "learning_rate": 2.9934129244345655e-06, + "loss": 0.3775, + "step": 29189 + }, + { + "epoch": 0.6425269645608629, + "grad_norm": 2.100368050697185, + "learning_rate": 2.993086437218716e-06, + "loss": 0.307, + "step": 29190 + }, + { + "epoch": 0.6425489764472815, + "grad_norm": 2.6217459652658173, + "learning_rate": 2.992759960202895e-06, + "loss": 0.2922, + "step": 29191 + }, + { + "epoch": 0.6425709883337002, + "grad_norm": 2.577495935731271, + "learning_rate": 2.9924334933887638e-06, + "loss": 0.2633, + "step": 29192 + }, + { + "epoch": 0.6425930002201189, + "grad_norm": 1.628219737284144, + "learning_rate": 2.9921070367779805e-06, + "loss": 0.3111, + "step": 29193 + }, + { + "epoch": 0.6426150121065375, + "grad_norm": 3.1831545998086384, + "learning_rate": 2.991780590372205e-06, + "loss": 0.3179, + "step": 29194 + }, + { + "epoch": 0.6426370239929562, + "grad_norm": 2.160940425273941, + "learning_rate": 2.9914541541730967e-06, + "loss": 0.2827, + "step": 29195 + }, + { + "epoch": 0.6426590358793749, + "grad_norm": 1.512007170416574, + "learning_rate": 2.9911277281823135e-06, + "loss": 0.2914, + "step": 29196 + }, + { + "epoch": 0.6426810477657935, + "grad_norm": 2.1647521450919305, + "learning_rate": 2.9908013124015155e-06, + "loss": 0.2983, + "step": 29197 + }, + { + "epoch": 0.6427030596522122, + "grad_norm": 1.546015710851682, + "learning_rate": 2.990474906832361e-06, + "loss": 0.3022, + "step": 29198 + }, + { + "epoch": 0.6427250715386309, + "grad_norm": 0.6235887261998979, + "learning_rate": 2.9901485114765093e-06, + "loss": 0.4803, + "step": 29199 + }, + { + "epoch": 0.6427470834250495, + "grad_norm": 1.5844687876596564, + "learning_rate": 2.989822126335621e-06, + "loss": 0.3214, + "step": 29200 + }, + { + "epoch": 0.6427690953114682, + "grad_norm": 1.8389681517114917, + "learning_rate": 2.989495751411351e-06, + "loss": 0.289, + "step": 29201 + }, + { + "epoch": 0.6427911071978869, + "grad_norm": 1.7285590446689705, + "learning_rate": 2.989169386705362e-06, + "loss": 0.2376, + "step": 29202 + }, + { + "epoch": 0.6428131190843055, + "grad_norm": 2.344421347819736, + "learning_rate": 2.9888430322193086e-06, + "loss": 0.3838, + "step": 29203 + }, + { + "epoch": 0.6428351309707242, + "grad_norm": 1.876222532168285, + "learning_rate": 2.988516687954854e-06, + "loss": 0.2936, + "step": 29204 + }, + { + "epoch": 0.6428571428571429, + "grad_norm": 0.6135522947021087, + "learning_rate": 2.988190353913655e-06, + "loss": 0.4596, + "step": 29205 + }, + { + "epoch": 0.6428791547435615, + "grad_norm": 1.7815346484195058, + "learning_rate": 2.987864030097369e-06, + "loss": 0.2844, + "step": 29206 + }, + { + "epoch": 0.6429011666299802, + "grad_norm": 1.7690397938518545, + "learning_rate": 2.987537716507656e-06, + "loss": 0.3565, + "step": 29207 + }, + { + "epoch": 0.6429231785163989, + "grad_norm": 1.5962120267563826, + "learning_rate": 2.987211413146174e-06, + "loss": 0.246, + "step": 29208 + }, + { + "epoch": 0.6429451904028175, + "grad_norm": 1.8352858664477414, + "learning_rate": 2.98688512001458e-06, + "loss": 0.2268, + "step": 29209 + }, + { + "epoch": 0.6429672022892362, + "grad_norm": 0.6190072208458449, + "learning_rate": 2.986558837114535e-06, + "loss": 0.4576, + "step": 29210 + }, + { + "epoch": 0.6429892141756548, + "grad_norm": 1.9534891272854287, + "learning_rate": 2.986232564447694e-06, + "loss": 0.3003, + "step": 29211 + }, + { + "epoch": 0.6430112260620735, + "grad_norm": 10.243092814385726, + "learning_rate": 2.98590630201572e-06, + "loss": 0.2539, + "step": 29212 + }, + { + "epoch": 0.6430332379484922, + "grad_norm": 1.5557451196876422, + "learning_rate": 2.985580049820268e-06, + "loss": 0.2945, + "step": 29213 + }, + { + "epoch": 0.6430552498349108, + "grad_norm": 1.69729730908493, + "learning_rate": 2.985253807862995e-06, + "loss": 0.2571, + "step": 29214 + }, + { + "epoch": 0.6430772617213295, + "grad_norm": 0.6229324146618362, + "learning_rate": 2.984927576145562e-06, + "loss": 0.4652, + "step": 29215 + }, + { + "epoch": 0.6430992736077482, + "grad_norm": 0.6103441337220412, + "learning_rate": 2.9846013546696247e-06, + "loss": 0.4521, + "step": 29216 + }, + { + "epoch": 0.6431212854941668, + "grad_norm": 1.4328726691753602, + "learning_rate": 2.9842751434368434e-06, + "loss": 0.3223, + "step": 29217 + }, + { + "epoch": 0.6431432973805855, + "grad_norm": 0.6105276237554024, + "learning_rate": 2.9839489424488756e-06, + "loss": 0.4876, + "step": 29218 + }, + { + "epoch": 0.6431653092670042, + "grad_norm": 2.1719498483286195, + "learning_rate": 2.9836227517073764e-06, + "loss": 0.316, + "step": 29219 + }, + { + "epoch": 0.6431873211534228, + "grad_norm": 1.6398085877231856, + "learning_rate": 2.9832965712140065e-06, + "loss": 0.2694, + "step": 29220 + }, + { + "epoch": 0.6432093330398415, + "grad_norm": 0.5906126444679772, + "learning_rate": 2.9829704009704226e-06, + "loss": 0.443, + "step": 29221 + }, + { + "epoch": 0.6432313449262602, + "grad_norm": 2.8199219828720397, + "learning_rate": 2.982644240978283e-06, + "loss": 0.2773, + "step": 29222 + }, + { + "epoch": 0.6432533568126788, + "grad_norm": 1.5662476469994608, + "learning_rate": 2.982318091239247e-06, + "loss": 0.3155, + "step": 29223 + }, + { + "epoch": 0.6432753686990975, + "grad_norm": 1.3231213084190045, + "learning_rate": 2.981991951754968e-06, + "loss": 0.2646, + "step": 29224 + }, + { + "epoch": 0.6432973805855162, + "grad_norm": 2.273526416024341, + "learning_rate": 2.981665822527106e-06, + "loss": 0.2738, + "step": 29225 + }, + { + "epoch": 0.6433193924719348, + "grad_norm": 1.9478303250762017, + "learning_rate": 2.9813397035573204e-06, + "loss": 0.3404, + "step": 29226 + }, + { + "epoch": 0.6433414043583535, + "grad_norm": 0.6169175338543008, + "learning_rate": 2.9810135948472653e-06, + "loss": 0.4694, + "step": 29227 + }, + { + "epoch": 0.6433634162447722, + "grad_norm": 0.6113217545114772, + "learning_rate": 2.9806874963985994e-06, + "loss": 0.439, + "step": 29228 + }, + { + "epoch": 0.6433854281311908, + "grad_norm": 5.099035069032668, + "learning_rate": 2.98036140821298e-06, + "loss": 0.3208, + "step": 29229 + }, + { + "epoch": 0.6434074400176095, + "grad_norm": 1.2537364144597443, + "learning_rate": 2.9800353302920653e-06, + "loss": 0.2932, + "step": 29230 + }, + { + "epoch": 0.6434294519040282, + "grad_norm": 0.6140935455311867, + "learning_rate": 2.9797092626375134e-06, + "loss": 0.4441, + "step": 29231 + }, + { + "epoch": 0.6434514637904468, + "grad_norm": 31.18909454672443, + "learning_rate": 2.979383205250978e-06, + "loss": 0.3166, + "step": 29232 + }, + { + "epoch": 0.6434734756768655, + "grad_norm": 3.205624426824685, + "learning_rate": 2.9790571581341194e-06, + "loss": 0.324, + "step": 29233 + }, + { + "epoch": 0.6434954875632841, + "grad_norm": 1.452733085366734, + "learning_rate": 2.9787311212885927e-06, + "loss": 0.2492, + "step": 29234 + }, + { + "epoch": 0.6435174994497028, + "grad_norm": 1.66575730889896, + "learning_rate": 2.978405094716057e-06, + "loss": 0.2834, + "step": 29235 + }, + { + "epoch": 0.6435395113361215, + "grad_norm": 2.9135269047221466, + "learning_rate": 2.978079078418169e-06, + "loss": 0.262, + "step": 29236 + }, + { + "epoch": 0.6435615232225401, + "grad_norm": 1.7426817689337657, + "learning_rate": 2.977753072396583e-06, + "loss": 0.2459, + "step": 29237 + }, + { + "epoch": 0.6435835351089588, + "grad_norm": 2.0598770846891146, + "learning_rate": 2.977427076652959e-06, + "loss": 0.2371, + "step": 29238 + }, + { + "epoch": 0.6436055469953775, + "grad_norm": 2.0304655901206474, + "learning_rate": 2.9771010911889507e-06, + "loss": 0.2688, + "step": 29239 + }, + { + "epoch": 0.6436275588817961, + "grad_norm": 1.5635064466883863, + "learning_rate": 2.9767751160062197e-06, + "loss": 0.3417, + "step": 29240 + }, + { + "epoch": 0.6436495707682148, + "grad_norm": 1.9219088036080074, + "learning_rate": 2.9764491511064196e-06, + "loss": 0.2772, + "step": 29241 + }, + { + "epoch": 0.6436715826546335, + "grad_norm": 0.6296938178403523, + "learning_rate": 2.9761231964912063e-06, + "loss": 0.4658, + "step": 29242 + }, + { + "epoch": 0.6436935945410521, + "grad_norm": 2.1831892816831338, + "learning_rate": 2.975797252162238e-06, + "loss": 0.3152, + "step": 29243 + }, + { + "epoch": 0.6437156064274708, + "grad_norm": 1.6820244473945833, + "learning_rate": 2.97547131812117e-06, + "loss": 0.3324, + "step": 29244 + }, + { + "epoch": 0.6437376183138895, + "grad_norm": 1.8784484101812382, + "learning_rate": 2.975145394369662e-06, + "loss": 0.3206, + "step": 29245 + }, + { + "epoch": 0.6437596302003081, + "grad_norm": 1.8701283728013143, + "learning_rate": 2.9748194809093668e-06, + "loss": 0.289, + "step": 29246 + }, + { + "epoch": 0.6437816420867268, + "grad_norm": 2.6777551372159287, + "learning_rate": 2.9744935777419415e-06, + "loss": 0.3421, + "step": 29247 + }, + { + "epoch": 0.6438036539731455, + "grad_norm": 2.3804864691987304, + "learning_rate": 2.974167684869044e-06, + "loss": 0.2814, + "step": 29248 + }, + { + "epoch": 0.6438256658595641, + "grad_norm": 1.6848177976343954, + "learning_rate": 2.973841802292331e-06, + "loss": 0.2639, + "step": 29249 + }, + { + "epoch": 0.6438476777459828, + "grad_norm": 2.165255530926776, + "learning_rate": 2.973515930013455e-06, + "loss": 0.3401, + "step": 29250 + }, + { + "epoch": 0.6438696896324015, + "grad_norm": 1.7186675746882867, + "learning_rate": 2.973190068034077e-06, + "loss": 0.2744, + "step": 29251 + }, + { + "epoch": 0.6438917015188201, + "grad_norm": 4.823379213452264, + "learning_rate": 2.9728642163558497e-06, + "loss": 0.362, + "step": 29252 + }, + { + "epoch": 0.6439137134052388, + "grad_norm": 2.1819169067128543, + "learning_rate": 2.9725383749804313e-06, + "loss": 0.2733, + "step": 29253 + }, + { + "epoch": 0.6439357252916575, + "grad_norm": 14.67823667638508, + "learning_rate": 2.9722125439094783e-06, + "loss": 0.3067, + "step": 29254 + }, + { + "epoch": 0.6439577371780761, + "grad_norm": 1.7850983284835178, + "learning_rate": 2.9718867231446437e-06, + "loss": 0.3453, + "step": 29255 + }, + { + "epoch": 0.6439797490644948, + "grad_norm": 1.946425640517355, + "learning_rate": 2.9715609126875868e-06, + "loss": 0.3282, + "step": 29256 + }, + { + "epoch": 0.6440017609509135, + "grad_norm": 1.739737311936499, + "learning_rate": 2.97123511253996e-06, + "loss": 0.2541, + "step": 29257 + }, + { + "epoch": 0.6440237728373321, + "grad_norm": 2.0663508263681356, + "learning_rate": 2.9709093227034223e-06, + "loss": 0.3483, + "step": 29258 + }, + { + "epoch": 0.6440457847237508, + "grad_norm": 1.8816779118774158, + "learning_rate": 2.9705835431796304e-06, + "loss": 0.2203, + "step": 29259 + }, + { + "epoch": 0.6440677966101694, + "grad_norm": 2.8505810874538446, + "learning_rate": 2.9702577739702354e-06, + "loss": 0.3139, + "step": 29260 + }, + { + "epoch": 0.6440898084965881, + "grad_norm": 2.0083329701877037, + "learning_rate": 2.969932015076897e-06, + "loss": 0.2824, + "step": 29261 + }, + { + "epoch": 0.6441118203830069, + "grad_norm": 2.0362826099644833, + "learning_rate": 2.969606266501268e-06, + "loss": 0.3172, + "step": 29262 + }, + { + "epoch": 0.6441338322694254, + "grad_norm": 3.257796883724702, + "learning_rate": 2.9692805282450083e-06, + "loss": 0.2373, + "step": 29263 + }, + { + "epoch": 0.6441558441558441, + "grad_norm": 2.7195233740912346, + "learning_rate": 2.968954800309769e-06, + "loss": 0.2819, + "step": 29264 + }, + { + "epoch": 0.6441778560422629, + "grad_norm": 1.4223882431498465, + "learning_rate": 2.9686290826972065e-06, + "loss": 0.2551, + "step": 29265 + }, + { + "epoch": 0.6441998679286814, + "grad_norm": 1.47448842502832, + "learning_rate": 2.9683033754089785e-06, + "loss": 0.2493, + "step": 29266 + }, + { + "epoch": 0.6442218798151002, + "grad_norm": 2.084014424909674, + "learning_rate": 2.967977678446739e-06, + "loss": 0.3616, + "step": 29267 + }, + { + "epoch": 0.6442438917015189, + "grad_norm": 1.7455660075831756, + "learning_rate": 2.9676519918121415e-06, + "loss": 0.3212, + "step": 29268 + }, + { + "epoch": 0.6442659035879374, + "grad_norm": 1.525823549057712, + "learning_rate": 2.9673263155068444e-06, + "loss": 0.2363, + "step": 29269 + }, + { + "epoch": 0.6442879154743562, + "grad_norm": 2.899210946637135, + "learning_rate": 2.9670006495325002e-06, + "loss": 0.3492, + "step": 29270 + }, + { + "epoch": 0.6443099273607749, + "grad_norm": 1.5976278622550448, + "learning_rate": 2.9666749938907667e-06, + "loss": 0.2872, + "step": 29271 + }, + { + "epoch": 0.6443319392471935, + "grad_norm": 2.522920912852565, + "learning_rate": 2.966349348583299e-06, + "loss": 0.3649, + "step": 29272 + }, + { + "epoch": 0.6443539511336122, + "grad_norm": 2.25287040300718, + "learning_rate": 2.966023713611749e-06, + "loss": 0.3027, + "step": 29273 + }, + { + "epoch": 0.6443759630200309, + "grad_norm": 1.333223741159252, + "learning_rate": 2.9656980889777743e-06, + "loss": 0.2598, + "step": 29274 + }, + { + "epoch": 0.6443979749064495, + "grad_norm": 1.9465631915142372, + "learning_rate": 2.9653724746830282e-06, + "loss": 0.2798, + "step": 29275 + }, + { + "epoch": 0.6444199867928682, + "grad_norm": 2.7079491875394583, + "learning_rate": 2.9650468707291674e-06, + "loss": 0.3218, + "step": 29276 + }, + { + "epoch": 0.6444419986792869, + "grad_norm": 1.6944387758913093, + "learning_rate": 2.9647212771178474e-06, + "loss": 0.2646, + "step": 29277 + }, + { + "epoch": 0.6444640105657055, + "grad_norm": 1.8390121656252416, + "learning_rate": 2.9643956938507195e-06, + "loss": 0.2763, + "step": 29278 + }, + { + "epoch": 0.6444860224521242, + "grad_norm": 1.8217051299412679, + "learning_rate": 2.9640701209294422e-06, + "loss": 0.4629, + "step": 29279 + }, + { + "epoch": 0.6445080343385429, + "grad_norm": 1.9661146877299094, + "learning_rate": 2.9637445583556668e-06, + "loss": 0.215, + "step": 29280 + }, + { + "epoch": 0.6445300462249615, + "grad_norm": 2.963688539809894, + "learning_rate": 2.963419006131052e-06, + "loss": 0.2608, + "step": 29281 + }, + { + "epoch": 0.6445520581113802, + "grad_norm": 3.7998889147404196, + "learning_rate": 2.963093464257249e-06, + "loss": 0.2521, + "step": 29282 + }, + { + "epoch": 0.6445740699977988, + "grad_norm": 1.8011146529596016, + "learning_rate": 2.9627679327359128e-06, + "loss": 0.2896, + "step": 29283 + }, + { + "epoch": 0.6445960818842175, + "grad_norm": 1.5512777481215527, + "learning_rate": 2.9624424115686988e-06, + "loss": 0.3431, + "step": 29284 + }, + { + "epoch": 0.6446180937706362, + "grad_norm": 1.8403472873246978, + "learning_rate": 2.962116900757263e-06, + "loss": 0.2512, + "step": 29285 + }, + { + "epoch": 0.6446401056570548, + "grad_norm": 2.226656973397123, + "learning_rate": 2.961791400303256e-06, + "loss": 0.3525, + "step": 29286 + }, + { + "epoch": 0.6446621175434735, + "grad_norm": 2.0110011545212148, + "learning_rate": 2.9614659102083355e-06, + "loss": 0.3627, + "step": 29287 + }, + { + "epoch": 0.6446841294298922, + "grad_norm": 1.8062873937107151, + "learning_rate": 2.961140430474153e-06, + "loss": 0.3361, + "step": 29288 + }, + { + "epoch": 0.6447061413163108, + "grad_norm": 3.4379096577294517, + "learning_rate": 2.9608149611023653e-06, + "loss": 0.3014, + "step": 29289 + }, + { + "epoch": 0.6447281532027295, + "grad_norm": 1.7081114627686864, + "learning_rate": 2.960489502094626e-06, + "loss": 0.2794, + "step": 29290 + }, + { + "epoch": 0.6447501650891482, + "grad_norm": 1.3307099108960796, + "learning_rate": 2.960164053452587e-06, + "loss": 0.3464, + "step": 29291 + }, + { + "epoch": 0.6447721769755668, + "grad_norm": 0.6330812052054565, + "learning_rate": 2.9598386151779057e-06, + "loss": 0.4538, + "step": 29292 + }, + { + "epoch": 0.6447941888619855, + "grad_norm": 1.3947627784553729, + "learning_rate": 2.959513187272233e-06, + "loss": 0.1975, + "step": 29293 + }, + { + "epoch": 0.6448162007484042, + "grad_norm": 6.569536175341359, + "learning_rate": 2.9591877697372253e-06, + "loss": 0.2911, + "step": 29294 + }, + { + "epoch": 0.6448382126348228, + "grad_norm": 2.4529405904780734, + "learning_rate": 2.958862362574536e-06, + "loss": 0.2503, + "step": 29295 + }, + { + "epoch": 0.6448602245212415, + "grad_norm": 0.6199700387224569, + "learning_rate": 2.9585369657858177e-06, + "loss": 0.4433, + "step": 29296 + }, + { + "epoch": 0.6448822364076602, + "grad_norm": 5.389107775088267, + "learning_rate": 2.958211579372725e-06, + "loss": 0.3093, + "step": 29297 + }, + { + "epoch": 0.6449042482940788, + "grad_norm": 1.4682261718097327, + "learning_rate": 2.9578862033369117e-06, + "loss": 0.2525, + "step": 29298 + }, + { + "epoch": 0.6449262601804975, + "grad_norm": 0.6058017377513101, + "learning_rate": 2.9575608376800323e-06, + "loss": 0.4588, + "step": 29299 + }, + { + "epoch": 0.6449482720669162, + "grad_norm": 1.876706401956449, + "learning_rate": 2.9572354824037395e-06, + "loss": 0.3156, + "step": 29300 + }, + { + "epoch": 0.6449702839533348, + "grad_norm": 5.029952719860436, + "learning_rate": 2.9569101375096867e-06, + "loss": 0.2839, + "step": 29301 + }, + { + "epoch": 0.6449922958397535, + "grad_norm": 2.178641628537488, + "learning_rate": 2.956584802999528e-06, + "loss": 0.3601, + "step": 29302 + }, + { + "epoch": 0.6450143077261722, + "grad_norm": 1.7760187583725051, + "learning_rate": 2.9562594788749153e-06, + "loss": 0.3678, + "step": 29303 + }, + { + "epoch": 0.6450363196125908, + "grad_norm": 1.5830192352864458, + "learning_rate": 2.955934165137506e-06, + "loss": 0.2724, + "step": 29304 + }, + { + "epoch": 0.6450583314990095, + "grad_norm": 0.5966738021730361, + "learning_rate": 2.95560886178895e-06, + "loss": 0.427, + "step": 29305 + }, + { + "epoch": 0.6450803433854281, + "grad_norm": 1.7237727839452617, + "learning_rate": 2.955283568830901e-06, + "loss": 0.2612, + "step": 29306 + }, + { + "epoch": 0.6451023552718468, + "grad_norm": 1.822774921771029, + "learning_rate": 2.9549582862650135e-06, + "loss": 0.3058, + "step": 29307 + }, + { + "epoch": 0.6451243671582655, + "grad_norm": 1.6742987593781808, + "learning_rate": 2.954633014092941e-06, + "loss": 0.2626, + "step": 29308 + }, + { + "epoch": 0.6451463790446841, + "grad_norm": 2.169327583344196, + "learning_rate": 2.9543077523163333e-06, + "loss": 0.3492, + "step": 29309 + }, + { + "epoch": 0.6451683909311028, + "grad_norm": 2.0597106817746016, + "learning_rate": 2.9539825009368477e-06, + "loss": 0.2449, + "step": 29310 + }, + { + "epoch": 0.6451904028175215, + "grad_norm": 2.7213543136214566, + "learning_rate": 2.953657259956134e-06, + "loss": 0.3241, + "step": 29311 + }, + { + "epoch": 0.6452124147039401, + "grad_norm": 1.4530739331987945, + "learning_rate": 2.9533320293758495e-06, + "loss": 0.2287, + "step": 29312 + }, + { + "epoch": 0.6452344265903588, + "grad_norm": 0.6353213803196178, + "learning_rate": 2.9530068091976437e-06, + "loss": 0.4745, + "step": 29313 + }, + { + "epoch": 0.6452564384767775, + "grad_norm": 2.742627906120265, + "learning_rate": 2.9526815994231683e-06, + "loss": 0.3534, + "step": 29314 + }, + { + "epoch": 0.6452784503631961, + "grad_norm": 3.183244284803829, + "learning_rate": 2.9523564000540794e-06, + "loss": 0.4172, + "step": 29315 + }, + { + "epoch": 0.6453004622496148, + "grad_norm": 1.7991678111028147, + "learning_rate": 2.952031211092028e-06, + "loss": 0.3095, + "step": 29316 + }, + { + "epoch": 0.6453224741360335, + "grad_norm": 1.414656218899708, + "learning_rate": 2.9517060325386692e-06, + "loss": 0.3125, + "step": 29317 + }, + { + "epoch": 0.6453444860224521, + "grad_norm": 1.496152761193537, + "learning_rate": 2.951380864395653e-06, + "loss": 0.3561, + "step": 29318 + }, + { + "epoch": 0.6453664979088708, + "grad_norm": 1.4443203649423795, + "learning_rate": 2.9510557066646317e-06, + "loss": 0.3026, + "step": 29319 + }, + { + "epoch": 0.6453885097952895, + "grad_norm": 1.7345181591300944, + "learning_rate": 2.95073055934726e-06, + "loss": 0.3206, + "step": 29320 + }, + { + "epoch": 0.6454105216817081, + "grad_norm": 1.6659074801562979, + "learning_rate": 2.950405422445189e-06, + "loss": 0.2329, + "step": 29321 + }, + { + "epoch": 0.6454325335681268, + "grad_norm": 1.2496079074104314, + "learning_rate": 2.950080295960074e-06, + "loss": 0.2174, + "step": 29322 + }, + { + "epoch": 0.6454545454545455, + "grad_norm": 2.122817718838409, + "learning_rate": 2.9497551798935637e-06, + "loss": 0.4033, + "step": 29323 + }, + { + "epoch": 0.6454765573409641, + "grad_norm": 1.9224904312442588, + "learning_rate": 2.9494300742473113e-06, + "loss": 0.322, + "step": 29324 + }, + { + "epoch": 0.6454985692273828, + "grad_norm": 1.9746217870854632, + "learning_rate": 2.9491049790229705e-06, + "loss": 0.2665, + "step": 29325 + }, + { + "epoch": 0.6455205811138015, + "grad_norm": 0.6272589189498032, + "learning_rate": 2.948779894222194e-06, + "loss": 0.4564, + "step": 29326 + }, + { + "epoch": 0.6455425930002201, + "grad_norm": 1.4861578081411218, + "learning_rate": 2.9484548198466314e-06, + "loss": 0.2496, + "step": 29327 + }, + { + "epoch": 0.6455646048866388, + "grad_norm": 2.2636819348167974, + "learning_rate": 2.9481297558979373e-06, + "loss": 0.2464, + "step": 29328 + }, + { + "epoch": 0.6455866167730574, + "grad_norm": 0.6124700190307076, + "learning_rate": 2.9478047023777613e-06, + "loss": 0.4574, + "step": 29329 + }, + { + "epoch": 0.6456086286594761, + "grad_norm": 0.6498821122342494, + "learning_rate": 2.9474796592877597e-06, + "loss": 0.5104, + "step": 29330 + }, + { + "epoch": 0.6456306405458948, + "grad_norm": 3.011113899952654, + "learning_rate": 2.9471546266295804e-06, + "loss": 0.35, + "step": 29331 + }, + { + "epoch": 0.6456526524323134, + "grad_norm": 1.568506228578708, + "learning_rate": 2.946829604404876e-06, + "loss": 0.3679, + "step": 29332 + }, + { + "epoch": 0.6456746643187321, + "grad_norm": 1.5654490437044264, + "learning_rate": 2.9465045926153015e-06, + "loss": 0.2814, + "step": 29333 + }, + { + "epoch": 0.6456966762051508, + "grad_norm": 1.7296674977204198, + "learning_rate": 2.946179591262504e-06, + "loss": 0.2451, + "step": 29334 + }, + { + "epoch": 0.6457186880915694, + "grad_norm": 2.159910552477944, + "learning_rate": 2.9458546003481404e-06, + "loss": 0.2813, + "step": 29335 + }, + { + "epoch": 0.6457406999779881, + "grad_norm": 1.7704950878427497, + "learning_rate": 2.9455296198738593e-06, + "loss": 0.2391, + "step": 29336 + }, + { + "epoch": 0.6457627118644068, + "grad_norm": 4.973264922590916, + "learning_rate": 2.9452046498413113e-06, + "loss": 0.3001, + "step": 29337 + }, + { + "epoch": 0.6457847237508254, + "grad_norm": 1.6855587484032486, + "learning_rate": 2.9448796902521513e-06, + "loss": 0.2375, + "step": 29338 + }, + { + "epoch": 0.6458067356372441, + "grad_norm": 2.507321320121286, + "learning_rate": 2.944554741108028e-06, + "loss": 0.3279, + "step": 29339 + }, + { + "epoch": 0.6458287475236628, + "grad_norm": 2.782571912407239, + "learning_rate": 2.9442298024105963e-06, + "loss": 0.434, + "step": 29340 + }, + { + "epoch": 0.6458507594100814, + "grad_norm": 1.9136447167734165, + "learning_rate": 2.943904874161505e-06, + "loss": 0.2838, + "step": 29341 + }, + { + "epoch": 0.6458727712965001, + "grad_norm": 2.5689681570965726, + "learning_rate": 2.9435799563624045e-06, + "loss": 0.2722, + "step": 29342 + }, + { + "epoch": 0.6458947831829188, + "grad_norm": 1.6208458440998836, + "learning_rate": 2.943255049014949e-06, + "loss": 0.2581, + "step": 29343 + }, + { + "epoch": 0.6459167950693374, + "grad_norm": 0.6195349040084255, + "learning_rate": 2.942930152120788e-06, + "loss": 0.4404, + "step": 29344 + }, + { + "epoch": 0.6459388069557561, + "grad_norm": 1.7253367888323616, + "learning_rate": 2.942605265681575e-06, + "loss": 0.3123, + "step": 29345 + }, + { + "epoch": 0.6459608188421748, + "grad_norm": 1.9869471313465448, + "learning_rate": 2.9422803896989594e-06, + "loss": 0.2946, + "step": 29346 + }, + { + "epoch": 0.6459828307285934, + "grad_norm": 1.5153349282566326, + "learning_rate": 2.9419555241745913e-06, + "loss": 0.2605, + "step": 29347 + }, + { + "epoch": 0.6460048426150121, + "grad_norm": 3.359813168354778, + "learning_rate": 2.941630669110125e-06, + "loss": 0.3248, + "step": 29348 + }, + { + "epoch": 0.6460268545014308, + "grad_norm": 4.583040637032158, + "learning_rate": 2.941305824507209e-06, + "loss": 0.3145, + "step": 29349 + }, + { + "epoch": 0.6460488663878494, + "grad_norm": 0.6250862011986449, + "learning_rate": 2.9409809903674935e-06, + "loss": 0.4892, + "step": 29350 + }, + { + "epoch": 0.6460708782742681, + "grad_norm": 2.497600979478973, + "learning_rate": 2.9406561666926325e-06, + "loss": 0.2897, + "step": 29351 + }, + { + "epoch": 0.6460928901606867, + "grad_norm": 1.2775414963432372, + "learning_rate": 2.940331353484274e-06, + "loss": 0.2137, + "step": 29352 + }, + { + "epoch": 0.6461149020471054, + "grad_norm": 2.12911256763918, + "learning_rate": 2.9400065507440723e-06, + "loss": 0.2903, + "step": 29353 + }, + { + "epoch": 0.6461369139335241, + "grad_norm": 1.881102572094824, + "learning_rate": 2.939681758473675e-06, + "loss": 0.3235, + "step": 29354 + }, + { + "epoch": 0.6461589258199427, + "grad_norm": 2.1509455949269083, + "learning_rate": 2.9393569766747333e-06, + "loss": 0.2913, + "step": 29355 + }, + { + "epoch": 0.6461809377063614, + "grad_norm": 1.4098525646341746, + "learning_rate": 2.9390322053488984e-06, + "loss": 0.3489, + "step": 29356 + }, + { + "epoch": 0.6462029495927801, + "grad_norm": 2.013761160508965, + "learning_rate": 2.9387074444978214e-06, + "loss": 0.2325, + "step": 29357 + }, + { + "epoch": 0.6462249614791987, + "grad_norm": 0.6108285486311533, + "learning_rate": 2.938382694123154e-06, + "loss": 0.4539, + "step": 29358 + }, + { + "epoch": 0.6462469733656174, + "grad_norm": 2.679700647232, + "learning_rate": 2.938057954226544e-06, + "loss": 0.326, + "step": 29359 + }, + { + "epoch": 0.6462689852520361, + "grad_norm": 2.3088841068962886, + "learning_rate": 2.9377332248096425e-06, + "loss": 0.288, + "step": 29360 + }, + { + "epoch": 0.6462909971384547, + "grad_norm": 0.6331859911915642, + "learning_rate": 2.9374085058741014e-06, + "loss": 0.4662, + "step": 29361 + }, + { + "epoch": 0.6463130090248734, + "grad_norm": 0.6149641560810747, + "learning_rate": 2.9370837974215694e-06, + "loss": 0.4553, + "step": 29362 + }, + { + "epoch": 0.6463350209112921, + "grad_norm": 2.272455167815155, + "learning_rate": 2.936759099453699e-06, + "loss": 0.188, + "step": 29363 + }, + { + "epoch": 0.6463570327977107, + "grad_norm": 0.6123317816174404, + "learning_rate": 2.936434411972139e-06, + "loss": 0.466, + "step": 29364 + }, + { + "epoch": 0.6463790446841294, + "grad_norm": 3.169203617983, + "learning_rate": 2.9361097349785383e-06, + "loss": 0.3141, + "step": 29365 + }, + { + "epoch": 0.6464010565705481, + "grad_norm": 1.559609214119138, + "learning_rate": 2.93578506847455e-06, + "loss": 0.3482, + "step": 29366 + }, + { + "epoch": 0.6464230684569667, + "grad_norm": 1.7580879084775876, + "learning_rate": 2.9354604124618223e-06, + "loss": 0.3176, + "step": 29367 + }, + { + "epoch": 0.6464450803433854, + "grad_norm": 1.7561063345618444, + "learning_rate": 2.9351357669420044e-06, + "loss": 0.2941, + "step": 29368 + }, + { + "epoch": 0.6464670922298041, + "grad_norm": 1.6764255120222453, + "learning_rate": 2.934811131916748e-06, + "loss": 0.313, + "step": 29369 + }, + { + "epoch": 0.6464891041162227, + "grad_norm": 1.8208764956278136, + "learning_rate": 2.934486507387702e-06, + "loss": 0.2786, + "step": 29370 + }, + { + "epoch": 0.6465111160026414, + "grad_norm": 2.2267095912750654, + "learning_rate": 2.9341618933565187e-06, + "loss": 0.3314, + "step": 29371 + }, + { + "epoch": 0.6465331278890601, + "grad_norm": 0.6163770367605235, + "learning_rate": 2.9338372898248456e-06, + "loss": 0.4613, + "step": 29372 + }, + { + "epoch": 0.6465551397754787, + "grad_norm": 1.917565077888677, + "learning_rate": 2.9335126967943312e-06, + "loss": 0.2668, + "step": 29373 + }, + { + "epoch": 0.6465771516618974, + "grad_norm": 4.090922502176848, + "learning_rate": 2.9331881142666286e-06, + "loss": 0.3447, + "step": 29374 + }, + { + "epoch": 0.646599163548316, + "grad_norm": 4.001389373521115, + "learning_rate": 2.932863542243384e-06, + "loss": 0.2906, + "step": 29375 + }, + { + "epoch": 0.6466211754347347, + "grad_norm": 4.412899907241722, + "learning_rate": 2.932538980726252e-06, + "loss": 0.2853, + "step": 29376 + }, + { + "epoch": 0.6466431873211534, + "grad_norm": 11.288882962700294, + "learning_rate": 2.9322144297168764e-06, + "loss": 0.3197, + "step": 29377 + }, + { + "epoch": 0.646665199207572, + "grad_norm": 2.5747021338877527, + "learning_rate": 2.931889889216909e-06, + "loss": 0.4091, + "step": 29378 + }, + { + "epoch": 0.6466872110939907, + "grad_norm": 2.529884303290708, + "learning_rate": 2.931565359228001e-06, + "loss": 0.2721, + "step": 29379 + }, + { + "epoch": 0.6467092229804094, + "grad_norm": 1.5802408923403333, + "learning_rate": 2.9312408397517993e-06, + "loss": 0.2593, + "step": 29380 + }, + { + "epoch": 0.646731234866828, + "grad_norm": 2.6006054574690722, + "learning_rate": 2.9309163307899564e-06, + "loss": 0.3171, + "step": 29381 + }, + { + "epoch": 0.6467532467532467, + "grad_norm": 0.6560370478332634, + "learning_rate": 2.9305918323441185e-06, + "loss": 0.4849, + "step": 29382 + }, + { + "epoch": 0.6467752586396655, + "grad_norm": 2.5545409277568707, + "learning_rate": 2.9302673444159345e-06, + "loss": 0.3172, + "step": 29383 + }, + { + "epoch": 0.646797270526084, + "grad_norm": 0.6173854419075625, + "learning_rate": 2.9299428670070573e-06, + "loss": 0.4688, + "step": 29384 + }, + { + "epoch": 0.6468192824125027, + "grad_norm": 3.1638952947952057, + "learning_rate": 2.929618400119133e-06, + "loss": 0.3107, + "step": 29385 + }, + { + "epoch": 0.6468412942989215, + "grad_norm": 1.6278851279201245, + "learning_rate": 2.92929394375381e-06, + "loss": 0.2507, + "step": 29386 + }, + { + "epoch": 0.64686330618534, + "grad_norm": 1.850412552306442, + "learning_rate": 2.928969497912739e-06, + "loss": 0.336, + "step": 29387 + }, + { + "epoch": 0.6468853180717588, + "grad_norm": 2.447136861257638, + "learning_rate": 2.9286450625975695e-06, + "loss": 0.2584, + "step": 29388 + }, + { + "epoch": 0.6469073299581775, + "grad_norm": 2.291899576768397, + "learning_rate": 2.9283206378099505e-06, + "loss": 0.2654, + "step": 29389 + }, + { + "epoch": 0.646929341844596, + "grad_norm": 2.7360812783268185, + "learning_rate": 2.927996223551529e-06, + "loss": 0.2511, + "step": 29390 + }, + { + "epoch": 0.6469513537310148, + "grad_norm": 1.9102645067008852, + "learning_rate": 2.927671819823954e-06, + "loss": 0.1992, + "step": 29391 + }, + { + "epoch": 0.6469733656174335, + "grad_norm": 1.622592056674293, + "learning_rate": 2.927347426628876e-06, + "loss": 0.2791, + "step": 29392 + }, + { + "epoch": 0.646995377503852, + "grad_norm": 1.3746099070453306, + "learning_rate": 2.927023043967942e-06, + "loss": 0.3112, + "step": 29393 + }, + { + "epoch": 0.6470173893902708, + "grad_norm": 2.1758929584305604, + "learning_rate": 2.9266986718428035e-06, + "loss": 0.3335, + "step": 29394 + }, + { + "epoch": 0.6470394012766895, + "grad_norm": 1.4918954297946985, + "learning_rate": 2.926374310255106e-06, + "loss": 0.2373, + "step": 29395 + }, + { + "epoch": 0.6470614131631081, + "grad_norm": 3.2649609265400805, + "learning_rate": 2.926049959206498e-06, + "loss": 0.2951, + "step": 29396 + }, + { + "epoch": 0.6470834250495268, + "grad_norm": 10.24100362175592, + "learning_rate": 2.92572561869863e-06, + "loss": 0.2963, + "step": 29397 + }, + { + "epoch": 0.6471054369359454, + "grad_norm": 1.7441605677930792, + "learning_rate": 2.9254012887331485e-06, + "loss": 0.2036, + "step": 29398 + }, + { + "epoch": 0.6471274488223641, + "grad_norm": 1.4648339232167082, + "learning_rate": 2.9250769693117055e-06, + "loss": 0.2844, + "step": 29399 + }, + { + "epoch": 0.6471494607087828, + "grad_norm": 1.8565176410120436, + "learning_rate": 2.9247526604359456e-06, + "loss": 0.2593, + "step": 29400 + }, + { + "epoch": 0.6471714725952014, + "grad_norm": 2.1646473421026826, + "learning_rate": 2.924428362107517e-06, + "loss": 0.295, + "step": 29401 + }, + { + "epoch": 0.6471934844816201, + "grad_norm": 1.7605608564710187, + "learning_rate": 2.9241040743280714e-06, + "loss": 0.3368, + "step": 29402 + }, + { + "epoch": 0.6472154963680388, + "grad_norm": 1.7032146820604792, + "learning_rate": 2.9237797970992517e-06, + "loss": 0.1983, + "step": 29403 + }, + { + "epoch": 0.6472375082544574, + "grad_norm": 4.250267106561943, + "learning_rate": 2.923455530422712e-06, + "loss": 0.3261, + "step": 29404 + }, + { + "epoch": 0.6472595201408761, + "grad_norm": 1.8182359411965905, + "learning_rate": 2.9231312743000963e-06, + "loss": 0.3334, + "step": 29405 + }, + { + "epoch": 0.6472815320272948, + "grad_norm": 1.8914146779660854, + "learning_rate": 2.9228070287330535e-06, + "loss": 0.2892, + "step": 29406 + }, + { + "epoch": 0.6473035439137134, + "grad_norm": 1.6390017162135457, + "learning_rate": 2.9224827937232333e-06, + "loss": 0.3772, + "step": 29407 + }, + { + "epoch": 0.6473255558001321, + "grad_norm": 2.8458985196170805, + "learning_rate": 2.922158569272281e-06, + "loss": 0.344, + "step": 29408 + }, + { + "epoch": 0.6473475676865508, + "grad_norm": 1.41384302704998, + "learning_rate": 2.921834355381845e-06, + "loss": 0.2786, + "step": 29409 + }, + { + "epoch": 0.6473695795729694, + "grad_norm": 1.7760135343689558, + "learning_rate": 2.9215101520535743e-06, + "loss": 0.2435, + "step": 29410 + }, + { + "epoch": 0.6473915914593881, + "grad_norm": 2.4943785798170373, + "learning_rate": 2.9211859592891157e-06, + "loss": 0.2847, + "step": 29411 + }, + { + "epoch": 0.6474136033458068, + "grad_norm": 0.6312529124026518, + "learning_rate": 2.9208617770901184e-06, + "loss": 0.457, + "step": 29412 + }, + { + "epoch": 0.6474356152322254, + "grad_norm": 2.1288477579101817, + "learning_rate": 2.920537605458228e-06, + "loss": 0.3757, + "step": 29413 + }, + { + "epoch": 0.6474576271186441, + "grad_norm": 1.4520461014093462, + "learning_rate": 2.9202134443950927e-06, + "loss": 0.2844, + "step": 29414 + }, + { + "epoch": 0.6474796390050628, + "grad_norm": 1.6210378088365291, + "learning_rate": 2.9198892939023605e-06, + "loss": 0.3167, + "step": 29415 + }, + { + "epoch": 0.6475016508914814, + "grad_norm": 0.6242478681410053, + "learning_rate": 2.919565153981678e-06, + "loss": 0.4483, + "step": 29416 + }, + { + "epoch": 0.6475236627779001, + "grad_norm": 1.851700640338499, + "learning_rate": 2.9192410246346946e-06, + "loss": 0.3884, + "step": 29417 + }, + { + "epoch": 0.6475456746643188, + "grad_norm": 3.198156101698904, + "learning_rate": 2.9189169058630575e-06, + "loss": 0.3174, + "step": 29418 + }, + { + "epoch": 0.6475676865507374, + "grad_norm": 1.9030344673434343, + "learning_rate": 2.91859279766841e-06, + "loss": 0.2202, + "step": 29419 + }, + { + "epoch": 0.6475896984371561, + "grad_norm": 1.2451377698895145, + "learning_rate": 2.918268700052403e-06, + "loss": 0.2865, + "step": 29420 + }, + { + "epoch": 0.6476117103235748, + "grad_norm": 0.6251352425811476, + "learning_rate": 2.9179446130166823e-06, + "loss": 0.4592, + "step": 29421 + }, + { + "epoch": 0.6476337222099934, + "grad_norm": 1.8735187656639685, + "learning_rate": 2.917620536562898e-06, + "loss": 0.3442, + "step": 29422 + }, + { + "epoch": 0.6476557340964121, + "grad_norm": 1.7443072708676413, + "learning_rate": 2.9172964706926948e-06, + "loss": 0.2303, + "step": 29423 + }, + { + "epoch": 0.6476777459828307, + "grad_norm": 1.6836511660726068, + "learning_rate": 2.9169724154077182e-06, + "loss": 0.2325, + "step": 29424 + }, + { + "epoch": 0.6476997578692494, + "grad_norm": 2.437764051121196, + "learning_rate": 2.916648370709616e-06, + "loss": 0.3203, + "step": 29425 + }, + { + "epoch": 0.6477217697556681, + "grad_norm": 0.6293168622638722, + "learning_rate": 2.9163243366000387e-06, + "loss": 0.4874, + "step": 29426 + }, + { + "epoch": 0.6477437816420867, + "grad_norm": 1.9373135207059073, + "learning_rate": 2.916000313080628e-06, + "loss": 0.2953, + "step": 29427 + }, + { + "epoch": 0.6477657935285054, + "grad_norm": 1.9156054098875084, + "learning_rate": 2.915676300153036e-06, + "loss": 0.3629, + "step": 29428 + }, + { + "epoch": 0.6477878054149241, + "grad_norm": 0.6341381960144437, + "learning_rate": 2.915352297818904e-06, + "loss": 0.4609, + "step": 29429 + }, + { + "epoch": 0.6478098173013427, + "grad_norm": 1.8864220963674405, + "learning_rate": 2.915028306079881e-06, + "loss": 0.3361, + "step": 29430 + }, + { + "epoch": 0.6478318291877614, + "grad_norm": 2.157361069242359, + "learning_rate": 2.9147043249376174e-06, + "loss": 0.3376, + "step": 29431 + }, + { + "epoch": 0.6478538410741801, + "grad_norm": 1.7834771948748362, + "learning_rate": 2.914380354393753e-06, + "loss": 0.2813, + "step": 29432 + }, + { + "epoch": 0.6478758529605987, + "grad_norm": 1.6189407039675798, + "learning_rate": 2.914056394449941e-06, + "loss": 0.2864, + "step": 29433 + }, + { + "epoch": 0.6478978648470174, + "grad_norm": 2.2296415607798927, + "learning_rate": 2.913732445107822e-06, + "loss": 0.3116, + "step": 29434 + }, + { + "epoch": 0.6479198767334361, + "grad_norm": 1.915121701824419, + "learning_rate": 2.913408506369046e-06, + "loss": 0.3272, + "step": 29435 + }, + { + "epoch": 0.6479418886198547, + "grad_norm": 14.041324849586129, + "learning_rate": 2.9130845782352603e-06, + "loss": 0.3188, + "step": 29436 + }, + { + "epoch": 0.6479639005062734, + "grad_norm": 1.602482222252508, + "learning_rate": 2.912760660708107e-06, + "loss": 0.2687, + "step": 29437 + }, + { + "epoch": 0.6479859123926921, + "grad_norm": 0.6093018807853497, + "learning_rate": 2.9124367537892374e-06, + "loss": 0.4651, + "step": 29438 + }, + { + "epoch": 0.6480079242791107, + "grad_norm": 5.446884981891268, + "learning_rate": 2.9121128574802935e-06, + "loss": 0.2753, + "step": 29439 + }, + { + "epoch": 0.6480299361655294, + "grad_norm": 0.6134257626047037, + "learning_rate": 2.911788971782923e-06, + "loss": 0.4731, + "step": 29440 + }, + { + "epoch": 0.6480519480519481, + "grad_norm": 1.776655536768824, + "learning_rate": 2.9114650966987747e-06, + "loss": 0.2536, + "step": 29441 + }, + { + "epoch": 0.6480739599383667, + "grad_norm": 5.8680772920350135, + "learning_rate": 2.9111412322294896e-06, + "loss": 0.3043, + "step": 29442 + }, + { + "epoch": 0.6480959718247854, + "grad_norm": 3.217616786369858, + "learning_rate": 2.910817378376719e-06, + "loss": 0.2299, + "step": 29443 + }, + { + "epoch": 0.6481179837112041, + "grad_norm": 1.7219501005422475, + "learning_rate": 2.9104935351421036e-06, + "loss": 0.3394, + "step": 29444 + }, + { + "epoch": 0.6481399955976227, + "grad_norm": 1.6969777112834832, + "learning_rate": 2.9101697025272925e-06, + "loss": 0.1809, + "step": 29445 + }, + { + "epoch": 0.6481620074840414, + "grad_norm": 2.371931994061474, + "learning_rate": 2.909845880533933e-06, + "loss": 0.3533, + "step": 29446 + }, + { + "epoch": 0.64818401937046, + "grad_norm": 2.5862151081617597, + "learning_rate": 2.9095220691636673e-06, + "loss": 0.3901, + "step": 29447 + }, + { + "epoch": 0.6482060312568787, + "grad_norm": 1.5180764551805277, + "learning_rate": 2.9091982684181443e-06, + "loss": 0.4482, + "step": 29448 + }, + { + "epoch": 0.6482280431432974, + "grad_norm": 1.7317037643730204, + "learning_rate": 2.9088744782990087e-06, + "loss": 0.2699, + "step": 29449 + }, + { + "epoch": 0.648250055029716, + "grad_norm": 1.8621257330576138, + "learning_rate": 2.908550698807903e-06, + "loss": 0.2819, + "step": 29450 + }, + { + "epoch": 0.6482720669161347, + "grad_norm": 5.295915855581679, + "learning_rate": 2.908226929946476e-06, + "loss": 0.2613, + "step": 29451 + }, + { + "epoch": 0.6482940788025534, + "grad_norm": 1.6227279068604736, + "learning_rate": 2.9079031717163725e-06, + "loss": 0.2292, + "step": 29452 + }, + { + "epoch": 0.648316090688972, + "grad_norm": 0.6452359677725935, + "learning_rate": 2.90757942411924e-06, + "loss": 0.4459, + "step": 29453 + }, + { + "epoch": 0.6483381025753907, + "grad_norm": 1.8997645816524085, + "learning_rate": 2.9072556871567227e-06, + "loss": 0.2102, + "step": 29454 + }, + { + "epoch": 0.6483601144618094, + "grad_norm": 1.869752563446749, + "learning_rate": 2.906931960830463e-06, + "loss": 0.3725, + "step": 29455 + }, + { + "epoch": 0.648382126348228, + "grad_norm": 1.7866508765196907, + "learning_rate": 2.9066082451421084e-06, + "loss": 0.2765, + "step": 29456 + }, + { + "epoch": 0.6484041382346467, + "grad_norm": 0.6369824887726723, + "learning_rate": 2.9062845400933048e-06, + "loss": 0.4677, + "step": 29457 + }, + { + "epoch": 0.6484261501210654, + "grad_norm": 2.0432046357550844, + "learning_rate": 2.905960845685699e-06, + "loss": 0.3885, + "step": 29458 + }, + { + "epoch": 0.648448162007484, + "grad_norm": 1.4638250535940625, + "learning_rate": 2.905637161920934e-06, + "loss": 0.3623, + "step": 29459 + }, + { + "epoch": 0.6484701738939027, + "grad_norm": 2.190323335574381, + "learning_rate": 2.905313488800653e-06, + "loss": 0.3993, + "step": 29460 + }, + { + "epoch": 0.6484921857803214, + "grad_norm": 5.16238573330077, + "learning_rate": 2.904989826326503e-06, + "loss": 0.3423, + "step": 29461 + }, + { + "epoch": 0.64851419766674, + "grad_norm": 0.6125908996840168, + "learning_rate": 2.904666174500129e-06, + "loss": 0.4636, + "step": 29462 + }, + { + "epoch": 0.6485362095531587, + "grad_norm": 1.7570597362334146, + "learning_rate": 2.904342533323179e-06, + "loss": 0.3159, + "step": 29463 + }, + { + "epoch": 0.6485582214395774, + "grad_norm": 1.5135346282706552, + "learning_rate": 2.9040189027972937e-06, + "loss": 0.333, + "step": 29464 + }, + { + "epoch": 0.648580233325996, + "grad_norm": 5.575701460336418, + "learning_rate": 2.903695282924118e-06, + "loss": 0.2877, + "step": 29465 + }, + { + "epoch": 0.6486022452124147, + "grad_norm": 2.718554333423406, + "learning_rate": 2.9033716737052973e-06, + "loss": 0.3192, + "step": 29466 + }, + { + "epoch": 0.6486242570988334, + "grad_norm": 3.6761734341221746, + "learning_rate": 2.9030480751424794e-06, + "loss": 0.2886, + "step": 29467 + }, + { + "epoch": 0.648646268985252, + "grad_norm": 0.6085076647921073, + "learning_rate": 2.902724487237304e-06, + "loss": 0.4498, + "step": 29468 + }, + { + "epoch": 0.6486682808716707, + "grad_norm": 1.8693139174025393, + "learning_rate": 2.90240090999142e-06, + "loss": 0.3539, + "step": 29469 + }, + { + "epoch": 0.6486902927580893, + "grad_norm": 2.074530548584994, + "learning_rate": 2.9020773434064685e-06, + "loss": 0.3174, + "step": 29470 + }, + { + "epoch": 0.648712304644508, + "grad_norm": 1.6575219585052947, + "learning_rate": 2.9017537874840955e-06, + "loss": 0.2915, + "step": 29471 + }, + { + "epoch": 0.6487343165309267, + "grad_norm": 1.7407266814424533, + "learning_rate": 2.901430242225948e-06, + "loss": 0.2314, + "step": 29472 + }, + { + "epoch": 0.6487563284173453, + "grad_norm": 0.6395576614808683, + "learning_rate": 2.9011067076336657e-06, + "loss": 0.4588, + "step": 29473 + }, + { + "epoch": 0.648778340303764, + "grad_norm": 1.7014950641621094, + "learning_rate": 2.9007831837088975e-06, + "loss": 0.2842, + "step": 29474 + }, + { + "epoch": 0.6488003521901827, + "grad_norm": 1.397169842205581, + "learning_rate": 2.900459670453283e-06, + "loss": 0.2922, + "step": 29475 + }, + { + "epoch": 0.6488223640766013, + "grad_norm": 0.6192054209702847, + "learning_rate": 2.9001361678684692e-06, + "loss": 0.4579, + "step": 29476 + }, + { + "epoch": 0.64884437596302, + "grad_norm": 0.6149427760577597, + "learning_rate": 2.8998126759561017e-06, + "loss": 0.4607, + "step": 29477 + }, + { + "epoch": 0.6488663878494387, + "grad_norm": 2.2914399240345005, + "learning_rate": 2.8994891947178205e-06, + "loss": 0.3724, + "step": 29478 + }, + { + "epoch": 0.6488883997358573, + "grad_norm": 1.3933635571072145, + "learning_rate": 2.899165724155275e-06, + "loss": 0.3001, + "step": 29479 + }, + { + "epoch": 0.648910411622276, + "grad_norm": 0.6014286589267277, + "learning_rate": 2.898842264270104e-06, + "loss": 0.4553, + "step": 29480 + }, + { + "epoch": 0.6489324235086947, + "grad_norm": 2.0169932054764943, + "learning_rate": 2.898518815063953e-06, + "loss": 0.3253, + "step": 29481 + }, + { + "epoch": 0.6489544353951133, + "grad_norm": 1.662101325360092, + "learning_rate": 2.8981953765384697e-06, + "loss": 0.3407, + "step": 29482 + }, + { + "epoch": 0.648976447281532, + "grad_norm": 2.2005340002635183, + "learning_rate": 2.897871948695292e-06, + "loss": 0.2925, + "step": 29483 + }, + { + "epoch": 0.6489984591679507, + "grad_norm": 1.320324413490739, + "learning_rate": 2.8975485315360695e-06, + "loss": 0.2603, + "step": 29484 + }, + { + "epoch": 0.6490204710543693, + "grad_norm": 3.045154012497872, + "learning_rate": 2.897225125062443e-06, + "loss": 0.3684, + "step": 29485 + }, + { + "epoch": 0.649042482940788, + "grad_norm": 0.5995116542086528, + "learning_rate": 2.8969017292760535e-06, + "loss": 0.4456, + "step": 29486 + }, + { + "epoch": 0.6490644948272067, + "grad_norm": 1.8269859808750997, + "learning_rate": 2.8965783441785483e-06, + "loss": 0.2843, + "step": 29487 + }, + { + "epoch": 0.6490865067136253, + "grad_norm": 1.7705712639651767, + "learning_rate": 2.8962549697715697e-06, + "loss": 0.3228, + "step": 29488 + }, + { + "epoch": 0.649108518600044, + "grad_norm": 2.5118559924546675, + "learning_rate": 2.8959316060567632e-06, + "loss": 0.2637, + "step": 29489 + }, + { + "epoch": 0.6491305304864627, + "grad_norm": 1.8747155037846943, + "learning_rate": 2.8956082530357715e-06, + "loss": 0.3292, + "step": 29490 + }, + { + "epoch": 0.6491525423728813, + "grad_norm": 2.9283202321202553, + "learning_rate": 2.8952849107102356e-06, + "loss": 0.3264, + "step": 29491 + }, + { + "epoch": 0.6491745542593, + "grad_norm": 2.8325788141677726, + "learning_rate": 2.894961579081799e-06, + "loss": 0.3085, + "step": 29492 + }, + { + "epoch": 0.6491965661457186, + "grad_norm": 2.912234443516647, + "learning_rate": 2.8946382581521075e-06, + "loss": 0.3638, + "step": 29493 + }, + { + "epoch": 0.6492185780321373, + "grad_norm": 2.7824985880159883, + "learning_rate": 2.8943149479228056e-06, + "loss": 0.2225, + "step": 29494 + }, + { + "epoch": 0.649240589918556, + "grad_norm": 0.6379892675791362, + "learning_rate": 2.893991648395534e-06, + "loss": 0.4516, + "step": 29495 + }, + { + "epoch": 0.6492626018049746, + "grad_norm": 1.7793265806825331, + "learning_rate": 2.893668359571934e-06, + "loss": 0.2692, + "step": 29496 + }, + { + "epoch": 0.6492846136913933, + "grad_norm": 1.4611824376928393, + "learning_rate": 2.893345081453651e-06, + "loss": 0.4095, + "step": 29497 + }, + { + "epoch": 0.649306625577812, + "grad_norm": 1.2635300018358604, + "learning_rate": 2.8930218140423277e-06, + "loss": 0.2539, + "step": 29498 + }, + { + "epoch": 0.6493286374642306, + "grad_norm": 1.8135376211201655, + "learning_rate": 2.8926985573396094e-06, + "loss": 0.2854, + "step": 29499 + }, + { + "epoch": 0.6493506493506493, + "grad_norm": 2.9979235430161557, + "learning_rate": 2.892375311347137e-06, + "loss": 0.2682, + "step": 29500 + }, + { + "epoch": 0.649372661237068, + "grad_norm": 1.49974350989266, + "learning_rate": 2.8920520760665514e-06, + "loss": 0.2708, + "step": 29501 + }, + { + "epoch": 0.6493946731234866, + "grad_norm": 1.5205553456144014, + "learning_rate": 2.891728851499497e-06, + "loss": 0.3039, + "step": 29502 + }, + { + "epoch": 0.6494166850099053, + "grad_norm": 5.575841855246334, + "learning_rate": 2.8914056376476172e-06, + "loss": 0.3083, + "step": 29503 + }, + { + "epoch": 0.649438696896324, + "grad_norm": 1.5662840610182451, + "learning_rate": 2.891082434512556e-06, + "loss": 0.2787, + "step": 29504 + }, + { + "epoch": 0.6494607087827426, + "grad_norm": 1.4690238595648875, + "learning_rate": 2.890759242095955e-06, + "loss": 0.3398, + "step": 29505 + }, + { + "epoch": 0.6494827206691614, + "grad_norm": 0.6088418457278524, + "learning_rate": 2.8904360603994542e-06, + "loss": 0.4878, + "step": 29506 + }, + { + "epoch": 0.64950473255558, + "grad_norm": 1.655314204810256, + "learning_rate": 2.890112889424698e-06, + "loss": 0.2612, + "step": 29507 + }, + { + "epoch": 0.6495267444419986, + "grad_norm": 1.3889552643543792, + "learning_rate": 2.8897897291733306e-06, + "loss": 0.335, + "step": 29508 + }, + { + "epoch": 0.6495487563284174, + "grad_norm": 1.3324173096129481, + "learning_rate": 2.8894665796469917e-06, + "loss": 0.2614, + "step": 29509 + }, + { + "epoch": 0.6495707682148361, + "grad_norm": 0.6281768336135655, + "learning_rate": 2.889143440847327e-06, + "loss": 0.4794, + "step": 29510 + }, + { + "epoch": 0.6495927801012547, + "grad_norm": 1.7616567485468293, + "learning_rate": 2.8888203127759747e-06, + "loss": 0.2703, + "step": 29511 + }, + { + "epoch": 0.6496147919876734, + "grad_norm": 1.4908225020699835, + "learning_rate": 2.8884971954345788e-06, + "loss": 0.274, + "step": 29512 + }, + { + "epoch": 0.6496368038740921, + "grad_norm": 1.3793618057187553, + "learning_rate": 2.8881740888247845e-06, + "loss": 0.2577, + "step": 29513 + }, + { + "epoch": 0.6496588157605107, + "grad_norm": 8.544949782985954, + "learning_rate": 2.887850992948229e-06, + "loss": 0.3675, + "step": 29514 + }, + { + "epoch": 0.6496808276469294, + "grad_norm": 2.4854155347197735, + "learning_rate": 2.8875279078065587e-06, + "loss": 0.3419, + "step": 29515 + }, + { + "epoch": 0.649702839533348, + "grad_norm": 2.1927419259411556, + "learning_rate": 2.887204833401412e-06, + "loss": 0.2658, + "step": 29516 + }, + { + "epoch": 0.6497248514197667, + "grad_norm": 1.4390454480260881, + "learning_rate": 2.8868817697344327e-06, + "loss": 0.2524, + "step": 29517 + }, + { + "epoch": 0.6497468633061854, + "grad_norm": 1.5537590570109325, + "learning_rate": 2.8865587168072645e-06, + "loss": 0.29, + "step": 29518 + }, + { + "epoch": 0.649768875192604, + "grad_norm": 1.5636363358084633, + "learning_rate": 2.886235674621546e-06, + "loss": 0.2344, + "step": 29519 + }, + { + "epoch": 0.6497908870790227, + "grad_norm": 1.6984940571912532, + "learning_rate": 2.885912643178922e-06, + "loss": 0.301, + "step": 29520 + }, + { + "epoch": 0.6498128989654414, + "grad_norm": 1.98001522509631, + "learning_rate": 2.885589622481031e-06, + "loss": 0.29, + "step": 29521 + }, + { + "epoch": 0.64983491085186, + "grad_norm": 1.5711405547268675, + "learning_rate": 2.8852666125295186e-06, + "loss": 0.2636, + "step": 29522 + }, + { + "epoch": 0.6498569227382787, + "grad_norm": 0.5966488820225367, + "learning_rate": 2.8849436133260224e-06, + "loss": 0.4561, + "step": 29523 + }, + { + "epoch": 0.6498789346246974, + "grad_norm": 0.6224961536860695, + "learning_rate": 2.8846206248721866e-06, + "loss": 0.4677, + "step": 29524 + }, + { + "epoch": 0.649900946511116, + "grad_norm": 2.501626202878977, + "learning_rate": 2.8842976471696536e-06, + "loss": 0.3086, + "step": 29525 + }, + { + "epoch": 0.6499229583975347, + "grad_norm": 2.374796257424131, + "learning_rate": 2.883974680220064e-06, + "loss": 0.2624, + "step": 29526 + }, + { + "epoch": 0.6499449702839534, + "grad_norm": 1.6114509333374933, + "learning_rate": 2.883651724025057e-06, + "loss": 0.2641, + "step": 29527 + }, + { + "epoch": 0.649966982170372, + "grad_norm": 4.6745433979725455, + "learning_rate": 2.883328778586275e-06, + "loss": 0.295, + "step": 29528 + }, + { + "epoch": 0.6499889940567907, + "grad_norm": 1.7819593095906903, + "learning_rate": 2.883005843905361e-06, + "loss": 0.298, + "step": 29529 + }, + { + "epoch": 0.6500110059432094, + "grad_norm": 0.6623749501408376, + "learning_rate": 2.882682919983958e-06, + "loss": 0.4811, + "step": 29530 + }, + { + "epoch": 0.650033017829628, + "grad_norm": 0.6191739611092614, + "learning_rate": 2.8823600068237032e-06, + "loss": 0.4637, + "step": 29531 + }, + { + "epoch": 0.6500550297160467, + "grad_norm": 1.5384399448448511, + "learning_rate": 2.8820371044262384e-06, + "loss": 0.3055, + "step": 29532 + }, + { + "epoch": 0.6500770416024654, + "grad_norm": 2.2352846480462873, + "learning_rate": 2.881714212793206e-06, + "loss": 0.3022, + "step": 29533 + }, + { + "epoch": 0.650099053488884, + "grad_norm": 4.835323683768968, + "learning_rate": 2.881391331926246e-06, + "loss": 0.3626, + "step": 29534 + }, + { + "epoch": 0.6501210653753027, + "grad_norm": 1.342697116691014, + "learning_rate": 2.8810684618270024e-06, + "loss": 0.2959, + "step": 29535 + }, + { + "epoch": 0.6501430772617214, + "grad_norm": 6.92443480262555, + "learning_rate": 2.8807456024971137e-06, + "loss": 0.25, + "step": 29536 + }, + { + "epoch": 0.65016508914814, + "grad_norm": 2.015143130624842, + "learning_rate": 2.8804227539382197e-06, + "loss": 0.2848, + "step": 29537 + }, + { + "epoch": 0.6501871010345587, + "grad_norm": 11.78070076722494, + "learning_rate": 2.8800999161519622e-06, + "loss": 0.29, + "step": 29538 + }, + { + "epoch": 0.6502091129209773, + "grad_norm": 1.5846153183198604, + "learning_rate": 2.8797770891399825e-06, + "loss": 0.2107, + "step": 29539 + }, + { + "epoch": 0.650231124807396, + "grad_norm": 2.2331402583279245, + "learning_rate": 2.8794542729039233e-06, + "loss": 0.335, + "step": 29540 + }, + { + "epoch": 0.6502531366938147, + "grad_norm": 1.382843337336263, + "learning_rate": 2.879131467445423e-06, + "loss": 0.2895, + "step": 29541 + }, + { + "epoch": 0.6502751485802333, + "grad_norm": 0.5956809108355617, + "learning_rate": 2.8788086727661214e-06, + "loss": 0.4528, + "step": 29542 + }, + { + "epoch": 0.650297160466652, + "grad_norm": 2.7692412188940057, + "learning_rate": 2.878485888867659e-06, + "loss": 0.3129, + "step": 29543 + }, + { + "epoch": 0.6503191723530707, + "grad_norm": 1.4198041970954134, + "learning_rate": 2.878163115751678e-06, + "loss": 0.331, + "step": 29544 + }, + { + "epoch": 0.6503411842394893, + "grad_norm": 1.8609788002333105, + "learning_rate": 2.877840353419821e-06, + "loss": 0.2313, + "step": 29545 + }, + { + "epoch": 0.650363196125908, + "grad_norm": 10.843275993324605, + "learning_rate": 2.877517601873725e-06, + "loss": 0.308, + "step": 29546 + }, + { + "epoch": 0.6503852080123267, + "grad_norm": 2.1292750242641985, + "learning_rate": 2.8771948611150292e-06, + "loss": 0.3398, + "step": 29547 + }, + { + "epoch": 0.6504072198987453, + "grad_norm": 0.6215635204959409, + "learning_rate": 2.8768721311453764e-06, + "loss": 0.4518, + "step": 29548 + }, + { + "epoch": 0.650429231785164, + "grad_norm": 2.1404393695487793, + "learning_rate": 2.8765494119664084e-06, + "loss": 0.246, + "step": 29549 + }, + { + "epoch": 0.6504512436715827, + "grad_norm": 1.3833860346770173, + "learning_rate": 2.8762267035797607e-06, + "loss": 0.2629, + "step": 29550 + }, + { + "epoch": 0.6504732555580013, + "grad_norm": 2.025320035920054, + "learning_rate": 2.875904005987079e-06, + "loss": 0.3632, + "step": 29551 + }, + { + "epoch": 0.65049526744442, + "grad_norm": 3.537195189539773, + "learning_rate": 2.8755813191899985e-06, + "loss": 0.251, + "step": 29552 + }, + { + "epoch": 0.6505172793308387, + "grad_norm": 3.0434845838296294, + "learning_rate": 2.875258643190161e-06, + "loss": 0.293, + "step": 29553 + }, + { + "epoch": 0.6505392912172573, + "grad_norm": 0.630246333641282, + "learning_rate": 2.8749359779892084e-06, + "loss": 0.4812, + "step": 29554 + }, + { + "epoch": 0.650561303103676, + "grad_norm": 1.6618218394403397, + "learning_rate": 2.874613323588777e-06, + "loss": 0.31, + "step": 29555 + }, + { + "epoch": 0.6505833149900947, + "grad_norm": 0.6159782541944957, + "learning_rate": 2.8742906799905105e-06, + "loss": 0.4495, + "step": 29556 + }, + { + "epoch": 0.6506053268765133, + "grad_norm": 1.7491099844576985, + "learning_rate": 2.873968047196045e-06, + "loss": 0.3428, + "step": 29557 + }, + { + "epoch": 0.650627338762932, + "grad_norm": 9.872608283788935, + "learning_rate": 2.8736454252070244e-06, + "loss": 0.3442, + "step": 29558 + }, + { + "epoch": 0.6506493506493507, + "grad_norm": 1.59342925329297, + "learning_rate": 2.8733228140250845e-06, + "loss": 0.2621, + "step": 29559 + }, + { + "epoch": 0.6506713625357693, + "grad_norm": 3.4233701405658223, + "learning_rate": 2.873000213651866e-06, + "loss": 0.236, + "step": 29560 + }, + { + "epoch": 0.650693374422188, + "grad_norm": 1.7084681310723082, + "learning_rate": 2.8726776240890106e-06, + "loss": 0.454, + "step": 29561 + }, + { + "epoch": 0.6507153863086067, + "grad_norm": 1.6038004188041242, + "learning_rate": 2.8723550453381556e-06, + "loss": 0.2187, + "step": 29562 + }, + { + "epoch": 0.6507373981950253, + "grad_norm": 2.4233393460388277, + "learning_rate": 2.8720324774009424e-06, + "loss": 0.2373, + "step": 29563 + }, + { + "epoch": 0.650759410081444, + "grad_norm": 2.2166991553094344, + "learning_rate": 2.871709920279008e-06, + "loss": 0.2988, + "step": 29564 + }, + { + "epoch": 0.6507814219678626, + "grad_norm": 3.864260815174639, + "learning_rate": 2.8713873739739917e-06, + "loss": 0.2038, + "step": 29565 + }, + { + "epoch": 0.6508034338542813, + "grad_norm": 2.003255511042477, + "learning_rate": 2.8710648384875374e-06, + "loss": 0.3385, + "step": 29566 + }, + { + "epoch": 0.6508254457407, + "grad_norm": 1.3789214035925919, + "learning_rate": 2.8707423138212797e-06, + "loss": 0.2343, + "step": 29567 + }, + { + "epoch": 0.6508474576271186, + "grad_norm": 2.5553323263041845, + "learning_rate": 2.8704197999768585e-06, + "loss": 0.3194, + "step": 29568 + }, + { + "epoch": 0.6508694695135373, + "grad_norm": 2.079339864851101, + "learning_rate": 2.8700972969559137e-06, + "loss": 0.3544, + "step": 29569 + }, + { + "epoch": 0.650891481399956, + "grad_norm": 2.5587468993701017, + "learning_rate": 2.869774804760084e-06, + "loss": 0.3287, + "step": 29570 + }, + { + "epoch": 0.6509134932863746, + "grad_norm": 2.1257008538140045, + "learning_rate": 2.869452323391011e-06, + "loss": 0.3133, + "step": 29571 + }, + { + "epoch": 0.6509355051727933, + "grad_norm": 2.2919550379111215, + "learning_rate": 2.869129852850331e-06, + "loss": 0.3708, + "step": 29572 + }, + { + "epoch": 0.650957517059212, + "grad_norm": 2.007153690819406, + "learning_rate": 2.868807393139682e-06, + "loss": 0.3151, + "step": 29573 + }, + { + "epoch": 0.6509795289456306, + "grad_norm": 2.7447334322812638, + "learning_rate": 2.868484944260704e-06, + "loss": 0.3234, + "step": 29574 + }, + { + "epoch": 0.6510015408320493, + "grad_norm": 0.6287746363447332, + "learning_rate": 2.8681625062150363e-06, + "loss": 0.4712, + "step": 29575 + }, + { + "epoch": 0.651023552718468, + "grad_norm": 1.8014018999045118, + "learning_rate": 2.8678400790043204e-06, + "loss": 0.318, + "step": 29576 + }, + { + "epoch": 0.6510455646048866, + "grad_norm": 3.8465179203955686, + "learning_rate": 2.8675176626301916e-06, + "loss": 0.226, + "step": 29577 + }, + { + "epoch": 0.6510675764913053, + "grad_norm": 1.6208217781179428, + "learning_rate": 2.867195257094286e-06, + "loss": 0.3494, + "step": 29578 + }, + { + "epoch": 0.651089588377724, + "grad_norm": 7.240656293766373, + "learning_rate": 2.8668728623982466e-06, + "loss": 0.3208, + "step": 29579 + }, + { + "epoch": 0.6511116002641426, + "grad_norm": 3.0142691719929946, + "learning_rate": 2.866550478543711e-06, + "loss": 0.3029, + "step": 29580 + }, + { + "epoch": 0.6511336121505613, + "grad_norm": 2.7167968691244084, + "learning_rate": 2.8662281055323195e-06, + "loss": 0.3503, + "step": 29581 + }, + { + "epoch": 0.65115562403698, + "grad_norm": 2.027985052520985, + "learning_rate": 2.865905743365708e-06, + "loss": 0.3059, + "step": 29582 + }, + { + "epoch": 0.6511776359233986, + "grad_norm": 1.4850148700767878, + "learning_rate": 2.8655833920455124e-06, + "loss": 0.2819, + "step": 29583 + }, + { + "epoch": 0.6511996478098173, + "grad_norm": 2.5478737753711713, + "learning_rate": 2.8652610515733746e-06, + "loss": 0.3018, + "step": 29584 + }, + { + "epoch": 0.651221659696236, + "grad_norm": 0.6424219771292353, + "learning_rate": 2.864938721950935e-06, + "loss": 0.469, + "step": 29585 + }, + { + "epoch": 0.6512436715826546, + "grad_norm": 0.6179954043302625, + "learning_rate": 2.8646164031798263e-06, + "loss": 0.4658, + "step": 29586 + }, + { + "epoch": 0.6512656834690733, + "grad_norm": 1.6839779006778877, + "learning_rate": 2.8642940952616922e-06, + "loss": 0.2793, + "step": 29587 + }, + { + "epoch": 0.6512876953554919, + "grad_norm": 0.6262962732008517, + "learning_rate": 2.863971798198165e-06, + "loss": 0.4409, + "step": 29588 + }, + { + "epoch": 0.6513097072419106, + "grad_norm": 8.050702290503862, + "learning_rate": 2.8636495119908893e-06, + "loss": 0.3793, + "step": 29589 + }, + { + "epoch": 0.6513317191283293, + "grad_norm": 1.695683800942372, + "learning_rate": 2.863327236641496e-06, + "loss": 0.2455, + "step": 29590 + }, + { + "epoch": 0.6513537310147479, + "grad_norm": 0.600146135486102, + "learning_rate": 2.863004972151627e-06, + "loss": 0.4588, + "step": 29591 + }, + { + "epoch": 0.6513757429011666, + "grad_norm": 1.4194729730958144, + "learning_rate": 2.862682718522923e-06, + "loss": 0.2442, + "step": 29592 + }, + { + "epoch": 0.6513977547875853, + "grad_norm": 1.6039617646694344, + "learning_rate": 2.8623604757570156e-06, + "loss": 0.2699, + "step": 29593 + }, + { + "epoch": 0.6514197666740039, + "grad_norm": 2.0661121508502087, + "learning_rate": 2.8620382438555484e-06, + "loss": 0.3147, + "step": 29594 + }, + { + "epoch": 0.6514417785604226, + "grad_norm": 2.2643040468582067, + "learning_rate": 2.8617160228201536e-06, + "loss": 0.3468, + "step": 29595 + }, + { + "epoch": 0.6514637904468413, + "grad_norm": 2.872043692339007, + "learning_rate": 2.8613938126524723e-06, + "loss": 0.2118, + "step": 29596 + }, + { + "epoch": 0.6514858023332599, + "grad_norm": 2.5958436210373597, + "learning_rate": 2.8610716133541435e-06, + "loss": 0.2384, + "step": 29597 + }, + { + "epoch": 0.6515078142196786, + "grad_norm": 7.084869131549848, + "learning_rate": 2.8607494249268003e-06, + "loss": 0.2798, + "step": 29598 + }, + { + "epoch": 0.6515298261060973, + "grad_norm": 1.6236972935558922, + "learning_rate": 2.860427247372085e-06, + "loss": 0.3456, + "step": 29599 + }, + { + "epoch": 0.6515518379925159, + "grad_norm": 1.8622002533742172, + "learning_rate": 2.86010508069163e-06, + "loss": 0.3839, + "step": 29600 + }, + { + "epoch": 0.6515738498789346, + "grad_norm": 17.352493998698517, + "learning_rate": 2.859782924887077e-06, + "loss": 0.3739, + "step": 29601 + }, + { + "epoch": 0.6515958617653533, + "grad_norm": 1.9880997443651645, + "learning_rate": 2.859460779960062e-06, + "loss": 0.2426, + "step": 29602 + }, + { + "epoch": 0.6516178736517719, + "grad_norm": 0.6574595082749634, + "learning_rate": 2.8591386459122204e-06, + "loss": 0.4452, + "step": 29603 + }, + { + "epoch": 0.6516398855381906, + "grad_norm": 1.4059227275239272, + "learning_rate": 2.858816522745194e-06, + "loss": 0.2675, + "step": 29604 + }, + { + "epoch": 0.6516618974246093, + "grad_norm": 0.6620009427000209, + "learning_rate": 2.8584944104606145e-06, + "loss": 0.4563, + "step": 29605 + }, + { + "epoch": 0.6516839093110279, + "grad_norm": 3.0322310258848626, + "learning_rate": 2.858172309060121e-06, + "loss": 0.36, + "step": 29606 + }, + { + "epoch": 0.6517059211974466, + "grad_norm": 2.395227388335162, + "learning_rate": 2.857850218545354e-06, + "loss": 0.2732, + "step": 29607 + }, + { + "epoch": 0.6517279330838653, + "grad_norm": 2.437138651551221, + "learning_rate": 2.857528138917947e-06, + "loss": 0.2758, + "step": 29608 + }, + { + "epoch": 0.6517499449702839, + "grad_norm": 1.7598580172186746, + "learning_rate": 2.8572060701795356e-06, + "loss": 0.2958, + "step": 29609 + }, + { + "epoch": 0.6517719568567026, + "grad_norm": 1.9456371204399272, + "learning_rate": 2.856884012331759e-06, + "loss": 0.2705, + "step": 29610 + }, + { + "epoch": 0.6517939687431212, + "grad_norm": 2.27866515814589, + "learning_rate": 2.856561965376253e-06, + "loss": 0.3709, + "step": 29611 + }, + { + "epoch": 0.6518159806295399, + "grad_norm": 0.6113648499224394, + "learning_rate": 2.856239929314658e-06, + "loss": 0.4517, + "step": 29612 + }, + { + "epoch": 0.6518379925159586, + "grad_norm": 2.7749125837627995, + "learning_rate": 2.8559179041486075e-06, + "loss": 0.3259, + "step": 29613 + }, + { + "epoch": 0.6518600044023772, + "grad_norm": 1.4705375333581436, + "learning_rate": 2.8555958898797365e-06, + "loss": 0.2755, + "step": 29614 + }, + { + "epoch": 0.6518820162887959, + "grad_norm": 1.424406455031017, + "learning_rate": 2.8552738865096834e-06, + "loss": 0.2947, + "step": 29615 + }, + { + "epoch": 0.6519040281752146, + "grad_norm": 1.8718906276853657, + "learning_rate": 2.8549518940400854e-06, + "loss": 0.2682, + "step": 29616 + }, + { + "epoch": 0.6519260400616332, + "grad_norm": 1.9010279425262986, + "learning_rate": 2.8546299124725803e-06, + "loss": 0.286, + "step": 29617 + }, + { + "epoch": 0.6519480519480519, + "grad_norm": 1.3870925403959284, + "learning_rate": 2.854307941808803e-06, + "loss": 0.251, + "step": 29618 + }, + { + "epoch": 0.6519700638344706, + "grad_norm": 0.6353734927310545, + "learning_rate": 2.8539859820503873e-06, + "loss": 0.4698, + "step": 29619 + }, + { + "epoch": 0.6519920757208892, + "grad_norm": 1.6293637057380217, + "learning_rate": 2.8536640331989724e-06, + "loss": 0.2606, + "step": 29620 + }, + { + "epoch": 0.652014087607308, + "grad_norm": 2.6123142694471175, + "learning_rate": 2.853342095256194e-06, + "loss": 0.2787, + "step": 29621 + }, + { + "epoch": 0.6520360994937267, + "grad_norm": 1.4912651160025543, + "learning_rate": 2.8530201682236907e-06, + "loss": 0.2662, + "step": 29622 + }, + { + "epoch": 0.6520581113801452, + "grad_norm": 1.5110127377324152, + "learning_rate": 2.8526982521030964e-06, + "loss": 0.2895, + "step": 29623 + }, + { + "epoch": 0.652080123266564, + "grad_norm": 0.6093689606668133, + "learning_rate": 2.852376346896045e-06, + "loss": 0.4594, + "step": 29624 + }, + { + "epoch": 0.6521021351529827, + "grad_norm": 1.419411716554821, + "learning_rate": 2.8520544526041773e-06, + "loss": 0.2572, + "step": 29625 + }, + { + "epoch": 0.6521241470394012, + "grad_norm": 2.3693819712381456, + "learning_rate": 2.851732569229125e-06, + "loss": 0.3647, + "step": 29626 + }, + { + "epoch": 0.65214615892582, + "grad_norm": 1.415606690770097, + "learning_rate": 2.851410696772525e-06, + "loss": 0.2988, + "step": 29627 + }, + { + "epoch": 0.6521681708122387, + "grad_norm": 0.6190187550333074, + "learning_rate": 2.8510888352360176e-06, + "loss": 0.4778, + "step": 29628 + }, + { + "epoch": 0.6521901826986573, + "grad_norm": 2.1416348614200045, + "learning_rate": 2.850766984621232e-06, + "loss": 0.3196, + "step": 29629 + }, + { + "epoch": 0.652212194585076, + "grad_norm": 1.4610883110906696, + "learning_rate": 2.85044514492981e-06, + "loss": 0.2485, + "step": 29630 + }, + { + "epoch": 0.6522342064714947, + "grad_norm": 11.486340208498627, + "learning_rate": 2.850123316163382e-06, + "loss": 0.3117, + "step": 29631 + }, + { + "epoch": 0.6522562183579133, + "grad_norm": 1.7853016532543813, + "learning_rate": 2.849801498323586e-06, + "loss": 0.2595, + "step": 29632 + }, + { + "epoch": 0.652278230244332, + "grad_norm": 2.008234944379875, + "learning_rate": 2.8494796914120605e-06, + "loss": 0.2375, + "step": 29633 + }, + { + "epoch": 0.6523002421307506, + "grad_norm": 1.6852343562951448, + "learning_rate": 2.8491578954304355e-06, + "loss": 0.3436, + "step": 29634 + }, + { + "epoch": 0.6523222540171693, + "grad_norm": 3.243594114178193, + "learning_rate": 2.8488361103803517e-06, + "loss": 0.2642, + "step": 29635 + }, + { + "epoch": 0.652344265903588, + "grad_norm": 0.6048859391600443, + "learning_rate": 2.8485143362634406e-06, + "loss": 0.4612, + "step": 29636 + }, + { + "epoch": 0.6523662777900066, + "grad_norm": 1.8609299141764035, + "learning_rate": 2.8481925730813386e-06, + "loss": 0.3833, + "step": 29637 + }, + { + "epoch": 0.6523882896764253, + "grad_norm": 2.346211429894925, + "learning_rate": 2.8478708208356843e-06, + "loss": 0.3089, + "step": 29638 + }, + { + "epoch": 0.652410301562844, + "grad_norm": 1.5642813691538375, + "learning_rate": 2.847549079528108e-06, + "loss": 0.2731, + "step": 29639 + }, + { + "epoch": 0.6524323134492626, + "grad_norm": 1.6146482826086523, + "learning_rate": 2.8472273491602498e-06, + "loss": 0.2718, + "step": 29640 + }, + { + "epoch": 0.6524543253356813, + "grad_norm": 1.7654657950586914, + "learning_rate": 2.8469056297337395e-06, + "loss": 0.3912, + "step": 29641 + }, + { + "epoch": 0.6524763372221, + "grad_norm": 2.282152722381504, + "learning_rate": 2.8465839212502156e-06, + "loss": 0.2886, + "step": 29642 + }, + { + "epoch": 0.6524983491085186, + "grad_norm": 2.0829762933528326, + "learning_rate": 2.8462622237113146e-06, + "loss": 0.2514, + "step": 29643 + }, + { + "epoch": 0.6525203609949373, + "grad_norm": 3.048926599158545, + "learning_rate": 2.8459405371186677e-06, + "loss": 0.3216, + "step": 29644 + }, + { + "epoch": 0.652542372881356, + "grad_norm": 3.762380213202389, + "learning_rate": 2.8456188614739133e-06, + "loss": 0.2893, + "step": 29645 + }, + { + "epoch": 0.6525643847677746, + "grad_norm": 2.1034038376084094, + "learning_rate": 2.8452971967786834e-06, + "loss": 0.3015, + "step": 29646 + }, + { + "epoch": 0.6525863966541933, + "grad_norm": 1.8081713239600872, + "learning_rate": 2.8449755430346138e-06, + "loss": 0.312, + "step": 29647 + }, + { + "epoch": 0.652608408540612, + "grad_norm": 1.8380341911796685, + "learning_rate": 2.8446539002433417e-06, + "loss": 0.2791, + "step": 29648 + }, + { + "epoch": 0.6526304204270306, + "grad_norm": 1.7379445095628074, + "learning_rate": 2.8443322684065e-06, + "loss": 0.27, + "step": 29649 + }, + { + "epoch": 0.6526524323134493, + "grad_norm": 2.2632442195451072, + "learning_rate": 2.8440106475257207e-06, + "loss": 0.3295, + "step": 29650 + }, + { + "epoch": 0.652674444199868, + "grad_norm": 0.6442533597717411, + "learning_rate": 2.8436890376026414e-06, + "loss": 0.485, + "step": 29651 + }, + { + "epoch": 0.6526964560862866, + "grad_norm": 2.050431760749633, + "learning_rate": 2.8433674386388954e-06, + "loss": 0.2731, + "step": 29652 + }, + { + "epoch": 0.6527184679727053, + "grad_norm": 1.8654183567500453, + "learning_rate": 2.8430458506361207e-06, + "loss": 0.2691, + "step": 29653 + }, + { + "epoch": 0.652740479859124, + "grad_norm": 0.6108794671724616, + "learning_rate": 2.842724273595949e-06, + "loss": 0.4406, + "step": 29654 + }, + { + "epoch": 0.6527624917455426, + "grad_norm": 1.4169740418335124, + "learning_rate": 2.8424027075200124e-06, + "loss": 0.2995, + "step": 29655 + }, + { + "epoch": 0.6527845036319613, + "grad_norm": 1.7363533139785545, + "learning_rate": 2.842081152409948e-06, + "loss": 0.3234, + "step": 29656 + }, + { + "epoch": 0.6528065155183799, + "grad_norm": 2.9039132996057604, + "learning_rate": 2.841759608267389e-06, + "loss": 0.3382, + "step": 29657 + }, + { + "epoch": 0.6528285274047986, + "grad_norm": 1.400149542283689, + "learning_rate": 2.8414380750939726e-06, + "loss": 0.2818, + "step": 29658 + }, + { + "epoch": 0.6528505392912173, + "grad_norm": 0.6309719260639771, + "learning_rate": 2.8411165528913303e-06, + "loss": 0.4513, + "step": 29659 + }, + { + "epoch": 0.6528725511776359, + "grad_norm": 3.06085914748111, + "learning_rate": 2.8407950416610953e-06, + "loss": 0.3774, + "step": 29660 + }, + { + "epoch": 0.6528945630640546, + "grad_norm": 2.1565905913831727, + "learning_rate": 2.840473541404904e-06, + "loss": 0.3082, + "step": 29661 + }, + { + "epoch": 0.6529165749504733, + "grad_norm": 0.6324179718037285, + "learning_rate": 2.840152052124386e-06, + "loss": 0.4389, + "step": 29662 + }, + { + "epoch": 0.6529385868368919, + "grad_norm": 0.6322204403688946, + "learning_rate": 2.8398305738211825e-06, + "loss": 0.4717, + "step": 29663 + }, + { + "epoch": 0.6529605987233106, + "grad_norm": 0.6081162392751983, + "learning_rate": 2.839509106496923e-06, + "loss": 0.4627, + "step": 29664 + }, + { + "epoch": 0.6529826106097293, + "grad_norm": 1.7669782767746651, + "learning_rate": 2.83918765015324e-06, + "loss": 0.3138, + "step": 29665 + }, + { + "epoch": 0.6530046224961479, + "grad_norm": 2.097320833909693, + "learning_rate": 2.8388662047917715e-06, + "loss": 0.2545, + "step": 29666 + }, + { + "epoch": 0.6530266343825666, + "grad_norm": 1.6540335393743633, + "learning_rate": 2.8385447704141466e-06, + "loss": 0.2092, + "step": 29667 + }, + { + "epoch": 0.6530486462689853, + "grad_norm": 1.8264899758042774, + "learning_rate": 2.8382233470220016e-06, + "loss": 0.2936, + "step": 29668 + }, + { + "epoch": 0.6530706581554039, + "grad_norm": 1.4846369663620722, + "learning_rate": 2.837901934616971e-06, + "loss": 0.2753, + "step": 29669 + }, + { + "epoch": 0.6530926700418226, + "grad_norm": 1.8435383038335673, + "learning_rate": 2.8375805332006856e-06, + "loss": 0.2504, + "step": 29670 + }, + { + "epoch": 0.6531146819282413, + "grad_norm": 6.143219081702842, + "learning_rate": 2.837259142774783e-06, + "loss": 0.2546, + "step": 29671 + }, + { + "epoch": 0.6531366938146599, + "grad_norm": 1.4109709499409873, + "learning_rate": 2.836937763340891e-06, + "loss": 0.2747, + "step": 29672 + }, + { + "epoch": 0.6531587057010786, + "grad_norm": 1.959219683401837, + "learning_rate": 2.8366163949006464e-06, + "loss": 0.2522, + "step": 29673 + }, + { + "epoch": 0.6531807175874973, + "grad_norm": 1.7135531546918759, + "learning_rate": 2.836295037455684e-06, + "loss": 0.2719, + "step": 29674 + }, + { + "epoch": 0.6532027294739159, + "grad_norm": 0.6375780926310992, + "learning_rate": 2.835973691007634e-06, + "loss": 0.4493, + "step": 29675 + }, + { + "epoch": 0.6532247413603346, + "grad_norm": 0.6214126880865912, + "learning_rate": 2.8356523555581315e-06, + "loss": 0.4548, + "step": 29676 + }, + { + "epoch": 0.6532467532467533, + "grad_norm": 1.7180760340337786, + "learning_rate": 2.8353310311088084e-06, + "loss": 0.2709, + "step": 29677 + }, + { + "epoch": 0.6532687651331719, + "grad_norm": 1.2335959538042958, + "learning_rate": 2.835009717661298e-06, + "loss": 0.2231, + "step": 29678 + }, + { + "epoch": 0.6532907770195906, + "grad_norm": 3.0179509529124444, + "learning_rate": 2.834688415217236e-06, + "loss": 0.3888, + "step": 29679 + }, + { + "epoch": 0.6533127889060092, + "grad_norm": 1.5328625964765528, + "learning_rate": 2.834367123778251e-06, + "loss": 0.2607, + "step": 29680 + }, + { + "epoch": 0.6533348007924279, + "grad_norm": 1.7787517560040986, + "learning_rate": 2.83404584334598e-06, + "loss": 0.352, + "step": 29681 + }, + { + "epoch": 0.6533568126788466, + "grad_norm": 1.9293474852711756, + "learning_rate": 2.833724573922052e-06, + "loss": 0.3106, + "step": 29682 + }, + { + "epoch": 0.6533788245652652, + "grad_norm": 1.674355304061984, + "learning_rate": 2.833403315508102e-06, + "loss": 0.3256, + "step": 29683 + }, + { + "epoch": 0.6534008364516839, + "grad_norm": 2.21509461740361, + "learning_rate": 2.8330820681057645e-06, + "loss": 0.277, + "step": 29684 + }, + { + "epoch": 0.6534228483381026, + "grad_norm": 1.6985951118906204, + "learning_rate": 2.832760831716671e-06, + "loss": 0.2402, + "step": 29685 + }, + { + "epoch": 0.6534448602245212, + "grad_norm": 0.608796452402678, + "learning_rate": 2.832439606342452e-06, + "loss": 0.4522, + "step": 29686 + }, + { + "epoch": 0.6534668721109399, + "grad_norm": 1.4711454603043534, + "learning_rate": 2.8321183919847405e-06, + "loss": 0.1976, + "step": 29687 + }, + { + "epoch": 0.6534888839973586, + "grad_norm": 2.720310929669049, + "learning_rate": 2.831797188645171e-06, + "loss": 0.2779, + "step": 29688 + }, + { + "epoch": 0.6535108958837772, + "grad_norm": 2.1353565316464627, + "learning_rate": 2.8314759963253768e-06, + "loss": 0.2774, + "step": 29689 + }, + { + "epoch": 0.6535329077701959, + "grad_norm": 1.8994754933182194, + "learning_rate": 2.8311548150269884e-06, + "loss": 0.2461, + "step": 29690 + }, + { + "epoch": 0.6535549196566146, + "grad_norm": 1.9565960754273606, + "learning_rate": 2.8308336447516377e-06, + "loss": 0.2722, + "step": 29691 + }, + { + "epoch": 0.6535769315430332, + "grad_norm": 0.621317467525779, + "learning_rate": 2.830512485500957e-06, + "loss": 0.4705, + "step": 29692 + }, + { + "epoch": 0.6535989434294519, + "grad_norm": 1.5328245426133489, + "learning_rate": 2.83019133727658e-06, + "loss": 0.2651, + "step": 29693 + }, + { + "epoch": 0.6536209553158706, + "grad_norm": 1.4537512930257774, + "learning_rate": 2.8298702000801404e-06, + "loss": 0.287, + "step": 29694 + }, + { + "epoch": 0.6536429672022892, + "grad_norm": 2.0084346813412353, + "learning_rate": 2.829549073913268e-06, + "loss": 0.3057, + "step": 29695 + }, + { + "epoch": 0.6536649790887079, + "grad_norm": 1.79507919737292, + "learning_rate": 2.8292279587775927e-06, + "loss": 0.3264, + "step": 29696 + }, + { + "epoch": 0.6536869909751266, + "grad_norm": 1.7194003546014507, + "learning_rate": 2.828906854674752e-06, + "loss": 0.3787, + "step": 29697 + }, + { + "epoch": 0.6537090028615452, + "grad_norm": 2.0006586763863257, + "learning_rate": 2.8285857616063704e-06, + "loss": 0.2806, + "step": 29698 + }, + { + "epoch": 0.6537310147479639, + "grad_norm": 2.24361222204795, + "learning_rate": 2.8282646795740893e-06, + "loss": 0.3526, + "step": 29699 + }, + { + "epoch": 0.6537530266343826, + "grad_norm": 0.6096171823397306, + "learning_rate": 2.827943608579535e-06, + "loss": 0.4389, + "step": 29700 + }, + { + "epoch": 0.6537750385208012, + "grad_norm": 0.6361016794948369, + "learning_rate": 2.827622548624338e-06, + "loss": 0.4546, + "step": 29701 + }, + { + "epoch": 0.6537970504072199, + "grad_norm": 2.2887046358825516, + "learning_rate": 2.8273014997101343e-06, + "loss": 0.2867, + "step": 29702 + }, + { + "epoch": 0.6538190622936386, + "grad_norm": 1.8839250007861217, + "learning_rate": 2.82698046183855e-06, + "loss": 0.424, + "step": 29703 + }, + { + "epoch": 0.6538410741800572, + "grad_norm": 2.6008572660523877, + "learning_rate": 2.826659435011223e-06, + "loss": 0.3788, + "step": 29704 + }, + { + "epoch": 0.6538630860664759, + "grad_norm": 2.2473249156004877, + "learning_rate": 2.826338419229783e-06, + "loss": 0.34, + "step": 29705 + }, + { + "epoch": 0.6538850979528945, + "grad_norm": 1.592460966140257, + "learning_rate": 2.8260174144958586e-06, + "loss": 0.3068, + "step": 29706 + }, + { + "epoch": 0.6539071098393132, + "grad_norm": 0.6226539059558163, + "learning_rate": 2.825696420811085e-06, + "loss": 0.4636, + "step": 29707 + }, + { + "epoch": 0.6539291217257319, + "grad_norm": 2.023369537702237, + "learning_rate": 2.82537543817709e-06, + "loss": 0.2587, + "step": 29708 + }, + { + "epoch": 0.6539511336121505, + "grad_norm": 2.7055132750798316, + "learning_rate": 2.825054466595507e-06, + "loss": 0.3548, + "step": 29709 + }, + { + "epoch": 0.6539731454985692, + "grad_norm": 1.901798699888097, + "learning_rate": 2.8247335060679693e-06, + "loss": 0.3512, + "step": 29710 + }, + { + "epoch": 0.6539951573849879, + "grad_norm": 1.6929094583348991, + "learning_rate": 2.8244125565961043e-06, + "loss": 0.3231, + "step": 29711 + }, + { + "epoch": 0.6540171692714065, + "grad_norm": 3.7937217335913798, + "learning_rate": 2.824091618181547e-06, + "loss": 0.2967, + "step": 29712 + }, + { + "epoch": 0.6540391811578252, + "grad_norm": 1.7505541725538707, + "learning_rate": 2.823770690825924e-06, + "loss": 0.2873, + "step": 29713 + }, + { + "epoch": 0.6540611930442439, + "grad_norm": 1.2347935904600817, + "learning_rate": 2.8234497745308697e-06, + "loss": 0.2357, + "step": 29714 + }, + { + "epoch": 0.6540832049306625, + "grad_norm": 9.663916229157362, + "learning_rate": 2.823128869298016e-06, + "loss": 0.2813, + "step": 29715 + }, + { + "epoch": 0.6541052168170812, + "grad_norm": 2.1373478946818563, + "learning_rate": 2.8228079751289905e-06, + "loss": 0.2515, + "step": 29716 + }, + { + "epoch": 0.6541272287034999, + "grad_norm": 2.380115425969157, + "learning_rate": 2.8224870920254278e-06, + "loss": 0.3556, + "step": 29717 + }, + { + "epoch": 0.6541492405899185, + "grad_norm": 1.9083273973720751, + "learning_rate": 2.8221662199889544e-06, + "loss": 0.2547, + "step": 29718 + }, + { + "epoch": 0.6541712524763372, + "grad_norm": 1.8956336601291983, + "learning_rate": 2.8218453590212043e-06, + "loss": 0.22, + "step": 29719 + }, + { + "epoch": 0.6541932643627559, + "grad_norm": 1.9004857593153233, + "learning_rate": 2.821524509123809e-06, + "loss": 0.325, + "step": 29720 + }, + { + "epoch": 0.6542152762491745, + "grad_norm": 6.972252870567395, + "learning_rate": 2.8212036702983963e-06, + "loss": 0.2633, + "step": 29721 + }, + { + "epoch": 0.6542372881355932, + "grad_norm": 4.369699782692345, + "learning_rate": 2.8208828425466007e-06, + "loss": 0.393, + "step": 29722 + }, + { + "epoch": 0.6542593000220119, + "grad_norm": 1.6063207804327322, + "learning_rate": 2.820562025870047e-06, + "loss": 0.2759, + "step": 29723 + }, + { + "epoch": 0.6542813119084305, + "grad_norm": 1.5009256203757164, + "learning_rate": 2.8202412202703705e-06, + "loss": 0.3, + "step": 29724 + }, + { + "epoch": 0.6543033237948492, + "grad_norm": 2.2150678649767723, + "learning_rate": 2.819920425749202e-06, + "loss": 0.2632, + "step": 29725 + }, + { + "epoch": 0.6543253356812679, + "grad_norm": 1.9942791265731317, + "learning_rate": 2.8195996423081695e-06, + "loss": 0.3532, + "step": 29726 + }, + { + "epoch": 0.6543473475676865, + "grad_norm": 1.9253737864202962, + "learning_rate": 2.819278869948903e-06, + "loss": 0.3241, + "step": 29727 + }, + { + "epoch": 0.6543693594541052, + "grad_norm": 3.0024992171608393, + "learning_rate": 2.8189581086730334e-06, + "loss": 0.1874, + "step": 29728 + }, + { + "epoch": 0.6543913713405238, + "grad_norm": 12.896894412904789, + "learning_rate": 2.818637358482192e-06, + "loss": 0.2475, + "step": 29729 + }, + { + "epoch": 0.6544133832269425, + "grad_norm": 2.5584959399238927, + "learning_rate": 2.8183166193780096e-06, + "loss": 0.3105, + "step": 29730 + }, + { + "epoch": 0.6544353951133612, + "grad_norm": 0.629845604152782, + "learning_rate": 2.8179958913621163e-06, + "loss": 0.4752, + "step": 29731 + }, + { + "epoch": 0.6544574069997798, + "grad_norm": 1.3182672835399691, + "learning_rate": 2.817675174436138e-06, + "loss": 0.2519, + "step": 29732 + }, + { + "epoch": 0.6544794188861985, + "grad_norm": 2.8987826066020603, + "learning_rate": 2.8173544686017105e-06, + "loss": 0.3686, + "step": 29733 + }, + { + "epoch": 0.6545014307726172, + "grad_norm": 2.452111140522308, + "learning_rate": 2.817033773860457e-06, + "loss": 0.3579, + "step": 29734 + }, + { + "epoch": 0.6545234426590358, + "grad_norm": 4.6869621671505906, + "learning_rate": 2.816713090214015e-06, + "loss": 0.2722, + "step": 29735 + }, + { + "epoch": 0.6545454545454545, + "grad_norm": 1.7012994800331454, + "learning_rate": 2.8163924176640113e-06, + "loss": 0.2838, + "step": 29736 + }, + { + "epoch": 0.6545674664318732, + "grad_norm": 2.1275465271877465, + "learning_rate": 2.8160717562120734e-06, + "loss": 0.3698, + "step": 29737 + }, + { + "epoch": 0.6545894783182918, + "grad_norm": 2.069330806272639, + "learning_rate": 2.815751105859834e-06, + "loss": 0.3941, + "step": 29738 + }, + { + "epoch": 0.6546114902047105, + "grad_norm": 4.746833297462241, + "learning_rate": 2.8154304666089205e-06, + "loss": 0.2878, + "step": 29739 + }, + { + "epoch": 0.6546335020911292, + "grad_norm": 1.944595814054464, + "learning_rate": 2.8151098384609636e-06, + "loss": 0.2601, + "step": 29740 + }, + { + "epoch": 0.6546555139775478, + "grad_norm": 0.6098812438426542, + "learning_rate": 2.814789221417595e-06, + "loss": 0.4634, + "step": 29741 + }, + { + "epoch": 0.6546775258639665, + "grad_norm": 2.2341184234367963, + "learning_rate": 2.81446861548044e-06, + "loss": 0.3102, + "step": 29742 + }, + { + "epoch": 0.6546995377503853, + "grad_norm": 1.9257989847206611, + "learning_rate": 2.814148020651133e-06, + "loss": 0.3063, + "step": 29743 + }, + { + "epoch": 0.6547215496368038, + "grad_norm": 1.3528228399968625, + "learning_rate": 2.813827436931298e-06, + "loss": 0.2143, + "step": 29744 + }, + { + "epoch": 0.6547435615232226, + "grad_norm": 1.4611292413065204, + "learning_rate": 2.8135068643225673e-06, + "loss": 0.2291, + "step": 29745 + }, + { + "epoch": 0.6547655734096413, + "grad_norm": 1.4252757534468168, + "learning_rate": 2.8131863028265717e-06, + "loss": 0.2498, + "step": 29746 + }, + { + "epoch": 0.6547875852960598, + "grad_norm": 2.837112290673772, + "learning_rate": 2.8128657524449367e-06, + "loss": 0.3203, + "step": 29747 + }, + { + "epoch": 0.6548095971824786, + "grad_norm": 2.233839529635723, + "learning_rate": 2.812545213179295e-06, + "loss": 0.3022, + "step": 29748 + }, + { + "epoch": 0.6548316090688973, + "grad_norm": 8.776186377182757, + "learning_rate": 2.812224685031273e-06, + "loss": 0.2999, + "step": 29749 + }, + { + "epoch": 0.6548536209553159, + "grad_norm": 0.6469060436483112, + "learning_rate": 2.8119041680025e-06, + "loss": 0.4828, + "step": 29750 + }, + { + "epoch": 0.6548756328417346, + "grad_norm": 1.437192066869313, + "learning_rate": 2.8115836620946086e-06, + "loss": 0.3502, + "step": 29751 + }, + { + "epoch": 0.6548976447281531, + "grad_norm": 1.4244971041815475, + "learning_rate": 2.811263167309223e-06, + "loss": 0.2976, + "step": 29752 + }, + { + "epoch": 0.6549196566145719, + "grad_norm": 1.672144619554785, + "learning_rate": 2.810942683647976e-06, + "loss": 0.3371, + "step": 29753 + }, + { + "epoch": 0.6549416685009906, + "grad_norm": 1.8033445186365498, + "learning_rate": 2.8106222111124924e-06, + "loss": 0.3577, + "step": 29754 + }, + { + "epoch": 0.6549636803874092, + "grad_norm": 1.7257306309778015, + "learning_rate": 2.810301749704404e-06, + "loss": 0.2841, + "step": 29755 + }, + { + "epoch": 0.6549856922738279, + "grad_norm": 1.358661121365988, + "learning_rate": 2.809981299425341e-06, + "loss": 0.2582, + "step": 29756 + }, + { + "epoch": 0.6550077041602466, + "grad_norm": 1.4462953125949731, + "learning_rate": 2.809660860276926e-06, + "loss": 0.3479, + "step": 29757 + }, + { + "epoch": 0.6550297160466652, + "grad_norm": 1.7358206070302817, + "learning_rate": 2.8093404322607947e-06, + "loss": 0.3415, + "step": 29758 + }, + { + "epoch": 0.6550517279330839, + "grad_norm": 1.6755436862368454, + "learning_rate": 2.8090200153785708e-06, + "loss": 0.3342, + "step": 29759 + }, + { + "epoch": 0.6550737398195026, + "grad_norm": 1.454535077907201, + "learning_rate": 2.808699609631883e-06, + "loss": 0.2545, + "step": 29760 + }, + { + "epoch": 0.6550957517059212, + "grad_norm": 24.12392956973972, + "learning_rate": 2.8083792150223644e-06, + "loss": 0.3314, + "step": 29761 + }, + { + "epoch": 0.6551177635923399, + "grad_norm": 1.6367982924554678, + "learning_rate": 2.8080588315516375e-06, + "loss": 0.3199, + "step": 29762 + }, + { + "epoch": 0.6551397754787586, + "grad_norm": 1.247739864551862, + "learning_rate": 2.8077384592213353e-06, + "loss": 0.2354, + "step": 29763 + }, + { + "epoch": 0.6551617873651772, + "grad_norm": 3.1985648189137694, + "learning_rate": 2.8074180980330824e-06, + "loss": 0.2783, + "step": 29764 + }, + { + "epoch": 0.6551837992515959, + "grad_norm": 1.4970747262011717, + "learning_rate": 2.807097747988509e-06, + "loss": 0.3163, + "step": 29765 + }, + { + "epoch": 0.6552058111380146, + "grad_norm": 1.8681480327945423, + "learning_rate": 2.8067774090892435e-06, + "loss": 0.3175, + "step": 29766 + }, + { + "epoch": 0.6552278230244332, + "grad_norm": 2.289978290655691, + "learning_rate": 2.806457081336914e-06, + "loss": 0.3213, + "step": 29767 + }, + { + "epoch": 0.6552498349108519, + "grad_norm": 1.9510058274834847, + "learning_rate": 2.806136764733146e-06, + "loss": 0.3093, + "step": 29768 + }, + { + "epoch": 0.6552718467972706, + "grad_norm": 1.2929767302178783, + "learning_rate": 2.805816459279572e-06, + "loss": 0.3489, + "step": 29769 + }, + { + "epoch": 0.6552938586836892, + "grad_norm": 2.1932741432744827, + "learning_rate": 2.8054961649778133e-06, + "loss": 0.3728, + "step": 29770 + }, + { + "epoch": 0.6553158705701079, + "grad_norm": 1.3038628868840154, + "learning_rate": 2.8051758818295056e-06, + "loss": 0.2726, + "step": 29771 + }, + { + "epoch": 0.6553378824565266, + "grad_norm": 0.6152498810434978, + "learning_rate": 2.804855609836273e-06, + "loss": 0.4376, + "step": 29772 + }, + { + "epoch": 0.6553598943429452, + "grad_norm": 2.9561114422967547, + "learning_rate": 2.8045353489997406e-06, + "loss": 0.2315, + "step": 29773 + }, + { + "epoch": 0.6553819062293639, + "grad_norm": 0.6226482381705813, + "learning_rate": 2.8042150993215418e-06, + "loss": 0.4609, + "step": 29774 + }, + { + "epoch": 0.6554039181157825, + "grad_norm": 0.6331284168516921, + "learning_rate": 2.803894860803299e-06, + "loss": 0.4744, + "step": 29775 + }, + { + "epoch": 0.6554259300022012, + "grad_norm": 2.4058084151129804, + "learning_rate": 2.8035746334466417e-06, + "loss": 0.3776, + "step": 29776 + }, + { + "epoch": 0.6554479418886199, + "grad_norm": 1.2711691506744336, + "learning_rate": 2.8032544172531994e-06, + "loss": 0.1668, + "step": 29777 + }, + { + "epoch": 0.6554699537750385, + "grad_norm": 0.6281217658771264, + "learning_rate": 2.8029342122245966e-06, + "loss": 0.4799, + "step": 29778 + }, + { + "epoch": 0.6554919656614572, + "grad_norm": 2.504494701375753, + "learning_rate": 2.802614018362464e-06, + "loss": 0.3797, + "step": 29779 + }, + { + "epoch": 0.6555139775478759, + "grad_norm": 5.251469210184886, + "learning_rate": 2.8022938356684246e-06, + "loss": 0.2785, + "step": 29780 + }, + { + "epoch": 0.6555359894342945, + "grad_norm": 1.9289867130143694, + "learning_rate": 2.8019736641441085e-06, + "loss": 0.3071, + "step": 29781 + }, + { + "epoch": 0.6555580013207132, + "grad_norm": 0.6106776686444763, + "learning_rate": 2.8016535037911442e-06, + "loss": 0.4423, + "step": 29782 + }, + { + "epoch": 0.6555800132071319, + "grad_norm": 0.6418851916883576, + "learning_rate": 2.8013333546111553e-06, + "loss": 0.4535, + "step": 29783 + }, + { + "epoch": 0.6556020250935505, + "grad_norm": 2.006486081099055, + "learning_rate": 2.8010132166057734e-06, + "loss": 0.2585, + "step": 29784 + }, + { + "epoch": 0.6556240369799692, + "grad_norm": 2.491339310223293, + "learning_rate": 2.800693089776621e-06, + "loss": 0.2841, + "step": 29785 + }, + { + "epoch": 0.6556460488663879, + "grad_norm": 2.5234067811502334, + "learning_rate": 2.8003729741253273e-06, + "loss": 0.2717, + "step": 29786 + }, + { + "epoch": 0.6556680607528065, + "grad_norm": 1.7614907526661097, + "learning_rate": 2.800052869653521e-06, + "loss": 0.3092, + "step": 29787 + }, + { + "epoch": 0.6556900726392252, + "grad_norm": 1.6793833181214466, + "learning_rate": 2.7997327763628252e-06, + "loss": 0.2704, + "step": 29788 + }, + { + "epoch": 0.6557120845256439, + "grad_norm": 2.413543098857834, + "learning_rate": 2.7994126942548704e-06, + "loss": 0.2652, + "step": 29789 + }, + { + "epoch": 0.6557340964120625, + "grad_norm": 1.8459050860389112, + "learning_rate": 2.7990926233312806e-06, + "loss": 0.2913, + "step": 29790 + }, + { + "epoch": 0.6557561082984812, + "grad_norm": 1.706948169780479, + "learning_rate": 2.7987725635936836e-06, + "loss": 0.3705, + "step": 29791 + }, + { + "epoch": 0.6557781201848999, + "grad_norm": 0.6250984837364137, + "learning_rate": 2.7984525150437075e-06, + "loss": 0.4424, + "step": 29792 + }, + { + "epoch": 0.6558001320713185, + "grad_norm": 1.6014118573557927, + "learning_rate": 2.798132477682976e-06, + "loss": 0.2576, + "step": 29793 + }, + { + "epoch": 0.6558221439577372, + "grad_norm": 11.367033198868036, + "learning_rate": 2.7978124515131195e-06, + "loss": 0.2472, + "step": 29794 + }, + { + "epoch": 0.6558441558441559, + "grad_norm": 1.6732431275146689, + "learning_rate": 2.79749243653576e-06, + "loss": 0.3946, + "step": 29795 + }, + { + "epoch": 0.6558661677305745, + "grad_norm": 0.6237895191013539, + "learning_rate": 2.797172432752526e-06, + "loss": 0.4701, + "step": 29796 + }, + { + "epoch": 0.6558881796169932, + "grad_norm": 10.26224645539081, + "learning_rate": 2.796852440165047e-06, + "loss": 0.2493, + "step": 29797 + }, + { + "epoch": 0.6559101915034118, + "grad_norm": 2.2471305636775574, + "learning_rate": 2.796532458774944e-06, + "loss": 0.192, + "step": 29798 + }, + { + "epoch": 0.6559322033898305, + "grad_norm": 0.6009660873180735, + "learning_rate": 2.7962124885838484e-06, + "loss": 0.4744, + "step": 29799 + }, + { + "epoch": 0.6559542152762492, + "grad_norm": 0.6189192396392902, + "learning_rate": 2.7958925295933835e-06, + "loss": 0.4652, + "step": 29800 + }, + { + "epoch": 0.6559762271626678, + "grad_norm": 2.292246725158774, + "learning_rate": 2.7955725818051717e-06, + "loss": 0.2708, + "step": 29801 + }, + { + "epoch": 0.6559982390490865, + "grad_norm": 2.7709070210817823, + "learning_rate": 2.795252645220847e-06, + "loss": 0.2342, + "step": 29802 + }, + { + "epoch": 0.6560202509355052, + "grad_norm": 1.6305184850566985, + "learning_rate": 2.7949327198420307e-06, + "loss": 0.3945, + "step": 29803 + }, + { + "epoch": 0.6560422628219238, + "grad_norm": 1.6268757677307957, + "learning_rate": 2.7946128056703504e-06, + "loss": 0.3249, + "step": 29804 + }, + { + "epoch": 0.6560642747083425, + "grad_norm": 3.230375603538713, + "learning_rate": 2.794292902707433e-06, + "loss": 0.304, + "step": 29805 + }, + { + "epoch": 0.6560862865947612, + "grad_norm": 0.6092047912812376, + "learning_rate": 2.793973010954898e-06, + "loss": 0.4667, + "step": 29806 + }, + { + "epoch": 0.6561082984811798, + "grad_norm": 1.986423195896212, + "learning_rate": 2.79365313041438e-06, + "loss": 0.2505, + "step": 29807 + }, + { + "epoch": 0.6561303103675985, + "grad_norm": 1.8472504562638608, + "learning_rate": 2.793333261087501e-06, + "loss": 0.3018, + "step": 29808 + }, + { + "epoch": 0.6561523222540172, + "grad_norm": 1.3514232677973905, + "learning_rate": 2.793013402975884e-06, + "loss": 0.242, + "step": 29809 + }, + { + "epoch": 0.6561743341404358, + "grad_norm": 1.4497627379674611, + "learning_rate": 2.7926935560811596e-06, + "loss": 0.3095, + "step": 29810 + }, + { + "epoch": 0.6561963460268545, + "grad_norm": 1.7567455461302706, + "learning_rate": 2.7923737204049496e-06, + "loss": 0.3092, + "step": 29811 + }, + { + "epoch": 0.6562183579132732, + "grad_norm": 2.4367183496565867, + "learning_rate": 2.7920538959488803e-06, + "loss": 0.2584, + "step": 29812 + }, + { + "epoch": 0.6562403697996918, + "grad_norm": 0.6197076389374436, + "learning_rate": 2.79173408271458e-06, + "loss": 0.4329, + "step": 29813 + }, + { + "epoch": 0.6562623816861105, + "grad_norm": 1.9298031563680866, + "learning_rate": 2.7914142807036703e-06, + "loss": 0.2449, + "step": 29814 + }, + { + "epoch": 0.6562843935725292, + "grad_norm": 1.1818276627818172, + "learning_rate": 2.79109448991778e-06, + "loss": 0.1935, + "step": 29815 + }, + { + "epoch": 0.6563064054589478, + "grad_norm": 1.8214179694820531, + "learning_rate": 2.790774710358531e-06, + "loss": 0.3422, + "step": 29816 + }, + { + "epoch": 0.6563284173453665, + "grad_norm": 2.078200592518385, + "learning_rate": 2.79045494202755e-06, + "loss": 0.2977, + "step": 29817 + }, + { + "epoch": 0.6563504292317852, + "grad_norm": 1.5436294374749397, + "learning_rate": 2.7901351849264646e-06, + "loss": 0.3205, + "step": 29818 + }, + { + "epoch": 0.6563724411182038, + "grad_norm": 1.8247828426546093, + "learning_rate": 2.7898154390568954e-06, + "loss": 0.3466, + "step": 29819 + }, + { + "epoch": 0.6563944530046225, + "grad_norm": 1.9414494095439427, + "learning_rate": 2.789495704420473e-06, + "loss": 0.2993, + "step": 29820 + }, + { + "epoch": 0.6564164648910411, + "grad_norm": 2.324542416142452, + "learning_rate": 2.7891759810188163e-06, + "loss": 0.2686, + "step": 29821 + }, + { + "epoch": 0.6564384767774598, + "grad_norm": 1.629188004832785, + "learning_rate": 2.7888562688535537e-06, + "loss": 0.2927, + "step": 29822 + }, + { + "epoch": 0.6564604886638785, + "grad_norm": 1.919972486944835, + "learning_rate": 2.7885365679263117e-06, + "loss": 0.3002, + "step": 29823 + }, + { + "epoch": 0.6564825005502971, + "grad_norm": 2.0234034082900787, + "learning_rate": 2.788216878238711e-06, + "loss": 0.3705, + "step": 29824 + }, + { + "epoch": 0.6565045124367158, + "grad_norm": 1.4187576894314593, + "learning_rate": 2.787897199792381e-06, + "loss": 0.28, + "step": 29825 + }, + { + "epoch": 0.6565265243231345, + "grad_norm": 1.5097360873143724, + "learning_rate": 2.7875775325889416e-06, + "loss": 0.348, + "step": 29826 + }, + { + "epoch": 0.6565485362095531, + "grad_norm": 1.8517903866849312, + "learning_rate": 2.78725787663002e-06, + "loss": 0.2992, + "step": 29827 + }, + { + "epoch": 0.6565705480959718, + "grad_norm": 0.6217894561873946, + "learning_rate": 2.7869382319172433e-06, + "loss": 0.4609, + "step": 29828 + }, + { + "epoch": 0.6565925599823905, + "grad_norm": 1.6888130734696378, + "learning_rate": 2.786618598452231e-06, + "loss": 0.3132, + "step": 29829 + }, + { + "epoch": 0.6566145718688091, + "grad_norm": 1.515850933567926, + "learning_rate": 2.7862989762366123e-06, + "loss": 0.247, + "step": 29830 + }, + { + "epoch": 0.6566365837552278, + "grad_norm": 2.295365823241386, + "learning_rate": 2.7859793652720078e-06, + "loss": 0.3355, + "step": 29831 + }, + { + "epoch": 0.6566585956416465, + "grad_norm": 9.31579656562604, + "learning_rate": 2.7856597655600437e-06, + "loss": 0.3467, + "step": 29832 + }, + { + "epoch": 0.6566806075280651, + "grad_norm": 0.6138767789087897, + "learning_rate": 2.785340177102346e-06, + "loss": 0.4427, + "step": 29833 + }, + { + "epoch": 0.6567026194144838, + "grad_norm": 0.6013704187247777, + "learning_rate": 2.7850205999005352e-06, + "loss": 0.4659, + "step": 29834 + }, + { + "epoch": 0.6567246313009025, + "grad_norm": 2.1281995502919884, + "learning_rate": 2.7847010339562397e-06, + "loss": 0.3782, + "step": 29835 + }, + { + "epoch": 0.6567466431873211, + "grad_norm": 1.6100511598461762, + "learning_rate": 2.784381479271082e-06, + "loss": 0.3111, + "step": 29836 + }, + { + "epoch": 0.6567686550737398, + "grad_norm": 1.225067095289995, + "learning_rate": 2.7840619358466813e-06, + "loss": 0.2082, + "step": 29837 + }, + { + "epoch": 0.6567906669601585, + "grad_norm": 1.7812652560009417, + "learning_rate": 2.7837424036846706e-06, + "loss": 0.4349, + "step": 29838 + }, + { + "epoch": 0.6568126788465771, + "grad_norm": 1.6308428346801256, + "learning_rate": 2.783422882786667e-06, + "loss": 0.2651, + "step": 29839 + }, + { + "epoch": 0.6568346907329958, + "grad_norm": 1.9113921804238436, + "learning_rate": 2.783103373154299e-06, + "loss": 0.3566, + "step": 29840 + }, + { + "epoch": 0.6568567026194145, + "grad_norm": 2.0822703176872666, + "learning_rate": 2.782783874789189e-06, + "loss": 0.3333, + "step": 29841 + }, + { + "epoch": 0.6568787145058331, + "grad_norm": 0.6278213842246417, + "learning_rate": 2.7824643876929557e-06, + "loss": 0.4514, + "step": 29842 + }, + { + "epoch": 0.6569007263922518, + "grad_norm": 0.6398490528770329, + "learning_rate": 2.7821449118672317e-06, + "loss": 0.4568, + "step": 29843 + }, + { + "epoch": 0.6569227382786705, + "grad_norm": 14.747860925243051, + "learning_rate": 2.781825447313634e-06, + "loss": 0.2751, + "step": 29844 + }, + { + "epoch": 0.6569447501650891, + "grad_norm": 1.3397106646975854, + "learning_rate": 2.7815059940337914e-06, + "loss": 0.2612, + "step": 29845 + }, + { + "epoch": 0.6569667620515078, + "grad_norm": 2.4154130079284557, + "learning_rate": 2.7811865520293248e-06, + "loss": 0.3411, + "step": 29846 + }, + { + "epoch": 0.6569887739379264, + "grad_norm": 2.3001035092819007, + "learning_rate": 2.780867121301855e-06, + "loss": 0.2681, + "step": 29847 + }, + { + "epoch": 0.6570107858243451, + "grad_norm": 1.389728080778025, + "learning_rate": 2.7805477018530092e-06, + "loss": 0.3024, + "step": 29848 + }, + { + "epoch": 0.6570327977107638, + "grad_norm": 2.853600111984573, + "learning_rate": 2.780228293684411e-06, + "loss": 0.3065, + "step": 29849 + }, + { + "epoch": 0.6570548095971824, + "grad_norm": 2.6750627664232454, + "learning_rate": 2.7799088967976805e-06, + "loss": 0.447, + "step": 29850 + }, + { + "epoch": 0.6570768214836011, + "grad_norm": 1.523271643126784, + "learning_rate": 2.779589511194446e-06, + "loss": 0.225, + "step": 29851 + }, + { + "epoch": 0.6570988333700198, + "grad_norm": 1.7832634581284297, + "learning_rate": 2.7792701368763258e-06, + "loss": 0.3196, + "step": 29852 + }, + { + "epoch": 0.6571208452564384, + "grad_norm": 1.6959599861878143, + "learning_rate": 2.778950773844945e-06, + "loss": 0.2574, + "step": 29853 + }, + { + "epoch": 0.6571428571428571, + "grad_norm": 1.7042544483906015, + "learning_rate": 2.7786314221019296e-06, + "loss": 0.2464, + "step": 29854 + }, + { + "epoch": 0.6571648690292758, + "grad_norm": 1.584828657809913, + "learning_rate": 2.7783120816488974e-06, + "loss": 0.2758, + "step": 29855 + }, + { + "epoch": 0.6571868809156944, + "grad_norm": 1.5434220376870615, + "learning_rate": 2.7779927524874762e-06, + "loss": 0.2366, + "step": 29856 + }, + { + "epoch": 0.6572088928021131, + "grad_norm": 4.402733419114195, + "learning_rate": 2.7776734346192845e-06, + "loss": 0.255, + "step": 29857 + }, + { + "epoch": 0.6572309046885318, + "grad_norm": 0.6319393136183781, + "learning_rate": 2.777354128045948e-06, + "loss": 0.4674, + "step": 29858 + }, + { + "epoch": 0.6572529165749504, + "grad_norm": 0.6170728463401212, + "learning_rate": 2.777034832769091e-06, + "loss": 0.4612, + "step": 29859 + }, + { + "epoch": 0.6572749284613691, + "grad_norm": 1.9659315051659656, + "learning_rate": 2.776715548790333e-06, + "loss": 0.4208, + "step": 29860 + }, + { + "epoch": 0.6572969403477879, + "grad_norm": 2.0526442142369614, + "learning_rate": 2.7763962761112995e-06, + "loss": 0.4031, + "step": 29861 + }, + { + "epoch": 0.6573189522342064, + "grad_norm": 2.6449111959545704, + "learning_rate": 2.7760770147336104e-06, + "loss": 0.2695, + "step": 29862 + }, + { + "epoch": 0.6573409641206251, + "grad_norm": 1.32722628216031, + "learning_rate": 2.7757577646588894e-06, + "loss": 0.2763, + "step": 29863 + }, + { + "epoch": 0.6573629760070439, + "grad_norm": 0.6201068718254412, + "learning_rate": 2.7754385258887616e-06, + "loss": 0.4661, + "step": 29864 + }, + { + "epoch": 0.6573849878934624, + "grad_norm": 2.313858906077369, + "learning_rate": 2.7751192984248453e-06, + "loss": 0.2677, + "step": 29865 + }, + { + "epoch": 0.6574069997798812, + "grad_norm": 1.4532433610739632, + "learning_rate": 2.774800082268767e-06, + "loss": 0.2523, + "step": 29866 + }, + { + "epoch": 0.6574290116662999, + "grad_norm": 0.6302019929873198, + "learning_rate": 2.7744808774221454e-06, + "loss": 0.4864, + "step": 29867 + }, + { + "epoch": 0.6574510235527184, + "grad_norm": 1.5620823082549113, + "learning_rate": 2.7741616838866047e-06, + "loss": 0.3432, + "step": 29868 + }, + { + "epoch": 0.6574730354391372, + "grad_norm": 1.8313847592714179, + "learning_rate": 2.7738425016637693e-06, + "loss": 0.248, + "step": 29869 + }, + { + "epoch": 0.6574950473255557, + "grad_norm": 0.6221845230583467, + "learning_rate": 2.773523330755257e-06, + "loss": 0.4703, + "step": 29870 + }, + { + "epoch": 0.6575170592119745, + "grad_norm": 1.780639964120119, + "learning_rate": 2.7732041711626934e-06, + "loss": 0.2646, + "step": 29871 + }, + { + "epoch": 0.6575390710983932, + "grad_norm": 2.3504087370615765, + "learning_rate": 2.7728850228877e-06, + "loss": 0.3655, + "step": 29872 + }, + { + "epoch": 0.6575610829848118, + "grad_norm": 0.624506922696838, + "learning_rate": 2.7725658859318944e-06, + "loss": 0.4786, + "step": 29873 + }, + { + "epoch": 0.6575830948712305, + "grad_norm": 1.8671363504766487, + "learning_rate": 2.772246760296906e-06, + "loss": 0.2926, + "step": 29874 + }, + { + "epoch": 0.6576051067576492, + "grad_norm": 1.690034134644574, + "learning_rate": 2.7719276459843513e-06, + "loss": 0.3275, + "step": 29875 + }, + { + "epoch": 0.6576271186440678, + "grad_norm": 2.1998565466398423, + "learning_rate": 2.771608542995856e-06, + "loss": 0.2068, + "step": 29876 + }, + { + "epoch": 0.6576491305304865, + "grad_norm": 2.2887033372011367, + "learning_rate": 2.7712894513330407e-06, + "loss": 0.3366, + "step": 29877 + }, + { + "epoch": 0.6576711424169052, + "grad_norm": 3.713037930731177, + "learning_rate": 2.770970370997522e-06, + "loss": 0.3435, + "step": 29878 + }, + { + "epoch": 0.6576931543033238, + "grad_norm": 0.6064175295561924, + "learning_rate": 2.7706513019909297e-06, + "loss": 0.435, + "step": 29879 + }, + { + "epoch": 0.6577151661897425, + "grad_norm": 2.007760825682047, + "learning_rate": 2.77033224431488e-06, + "loss": 0.3247, + "step": 29880 + }, + { + "epoch": 0.6577371780761612, + "grad_norm": 3.5923699773017863, + "learning_rate": 2.7700131979709984e-06, + "loss": 0.3109, + "step": 29881 + }, + { + "epoch": 0.6577591899625798, + "grad_norm": 2.07577115406249, + "learning_rate": 2.7696941629609044e-06, + "loss": 0.3157, + "step": 29882 + }, + { + "epoch": 0.6577812018489985, + "grad_norm": 49.624491748965106, + "learning_rate": 2.769375139286217e-06, + "loss": 0.3415, + "step": 29883 + }, + { + "epoch": 0.6578032137354172, + "grad_norm": 1.4974851350013831, + "learning_rate": 2.7690561269485606e-06, + "loss": 0.2908, + "step": 29884 + }, + { + "epoch": 0.6578252256218358, + "grad_norm": 1.2998844029201984, + "learning_rate": 2.7687371259495576e-06, + "loss": 0.3089, + "step": 29885 + }, + { + "epoch": 0.6578472375082545, + "grad_norm": 1.8506560616432375, + "learning_rate": 2.768418136290826e-06, + "loss": 0.3031, + "step": 29886 + }, + { + "epoch": 0.6578692493946732, + "grad_norm": 3.3412510774197512, + "learning_rate": 2.76809915797399e-06, + "loss": 0.2764, + "step": 29887 + }, + { + "epoch": 0.6578912612810918, + "grad_norm": 1.4686369271684776, + "learning_rate": 2.767780191000668e-06, + "loss": 0.28, + "step": 29888 + }, + { + "epoch": 0.6579132731675105, + "grad_norm": 1.8988954674909044, + "learning_rate": 2.7674612353724826e-06, + "loss": 0.2449, + "step": 29889 + }, + { + "epoch": 0.6579352850539292, + "grad_norm": 0.6005904120520748, + "learning_rate": 2.7671422910910572e-06, + "loss": 0.4366, + "step": 29890 + }, + { + "epoch": 0.6579572969403478, + "grad_norm": 1.9796761441009603, + "learning_rate": 2.7668233581580082e-06, + "loss": 0.2598, + "step": 29891 + }, + { + "epoch": 0.6579793088267665, + "grad_norm": 1.5949138534876686, + "learning_rate": 2.7665044365749614e-06, + "loss": 0.2722, + "step": 29892 + }, + { + "epoch": 0.6580013207131851, + "grad_norm": 0.6180837032602864, + "learning_rate": 2.7661855263435334e-06, + "loss": 0.4385, + "step": 29893 + }, + { + "epoch": 0.6580233325996038, + "grad_norm": 1.5407133106762214, + "learning_rate": 2.7658666274653457e-06, + "loss": 0.2417, + "step": 29894 + }, + { + "epoch": 0.6580453444860225, + "grad_norm": 1.4239698115351158, + "learning_rate": 2.765547739942023e-06, + "loss": 0.2898, + "step": 29895 + }, + { + "epoch": 0.6580673563724411, + "grad_norm": 1.7640447588622126, + "learning_rate": 2.7652288637751813e-06, + "loss": 0.2714, + "step": 29896 + }, + { + "epoch": 0.6580893682588598, + "grad_norm": 2.012728264893402, + "learning_rate": 2.764909998966445e-06, + "loss": 0.2461, + "step": 29897 + }, + { + "epoch": 0.6581113801452785, + "grad_norm": 1.6244303431418199, + "learning_rate": 2.7645911455174314e-06, + "loss": 0.2967, + "step": 29898 + }, + { + "epoch": 0.6581333920316971, + "grad_norm": 1.843383436658908, + "learning_rate": 2.7642723034297624e-06, + "loss": 0.3975, + "step": 29899 + }, + { + "epoch": 0.6581554039181158, + "grad_norm": 0.6155732978131806, + "learning_rate": 2.763953472705061e-06, + "loss": 0.4783, + "step": 29900 + }, + { + "epoch": 0.6581774158045345, + "grad_norm": 1.4290295133886637, + "learning_rate": 2.763634653344943e-06, + "loss": 0.2487, + "step": 29901 + }, + { + "epoch": 0.6581994276909531, + "grad_norm": 1.2384302731838093, + "learning_rate": 2.7633158453510334e-06, + "loss": 0.3032, + "step": 29902 + }, + { + "epoch": 0.6582214395773718, + "grad_norm": 0.6154058127230811, + "learning_rate": 2.762997048724948e-06, + "loss": 0.4685, + "step": 29903 + }, + { + "epoch": 0.6582434514637905, + "grad_norm": 3.1813003575593135, + "learning_rate": 2.762678263468309e-06, + "loss": 0.2047, + "step": 29904 + }, + { + "epoch": 0.6582654633502091, + "grad_norm": 1.5258731589955703, + "learning_rate": 2.7623594895827395e-06, + "loss": 0.335, + "step": 29905 + }, + { + "epoch": 0.6582874752366278, + "grad_norm": 1.8059876175420115, + "learning_rate": 2.7620407270698544e-06, + "loss": 0.3716, + "step": 29906 + }, + { + "epoch": 0.6583094871230465, + "grad_norm": 1.4929414525396312, + "learning_rate": 2.7617219759312796e-06, + "loss": 0.2337, + "step": 29907 + }, + { + "epoch": 0.6583314990094651, + "grad_norm": 3.42242181289411, + "learning_rate": 2.761403236168631e-06, + "loss": 0.3129, + "step": 29908 + }, + { + "epoch": 0.6583535108958838, + "grad_norm": 2.320173300923504, + "learning_rate": 2.761084507783526e-06, + "loss": 0.3163, + "step": 29909 + }, + { + "epoch": 0.6583755227823025, + "grad_norm": 1.3080401129337473, + "learning_rate": 2.760765790777592e-06, + "loss": 0.2511, + "step": 29910 + }, + { + "epoch": 0.6583975346687211, + "grad_norm": 1.9821312388214214, + "learning_rate": 2.7604470851524434e-06, + "loss": 0.3542, + "step": 29911 + }, + { + "epoch": 0.6584195465551398, + "grad_norm": 1.9391665017684507, + "learning_rate": 2.760128390909703e-06, + "loss": 0.3085, + "step": 29912 + }, + { + "epoch": 0.6584415584415585, + "grad_norm": 2.004689967267499, + "learning_rate": 2.75980970805099e-06, + "loss": 0.2639, + "step": 29913 + }, + { + "epoch": 0.6584635703279771, + "grad_norm": 1.862911443751022, + "learning_rate": 2.759491036577918e-06, + "loss": 0.2731, + "step": 29914 + }, + { + "epoch": 0.6584855822143958, + "grad_norm": 2.0582214546403215, + "learning_rate": 2.7591723764921168e-06, + "loss": 0.3033, + "step": 29915 + }, + { + "epoch": 0.6585075941008144, + "grad_norm": 2.874710625818024, + "learning_rate": 2.7588537277951987e-06, + "loss": 0.3672, + "step": 29916 + }, + { + "epoch": 0.6585296059872331, + "grad_norm": 1.8428937980872584, + "learning_rate": 2.7585350904887875e-06, + "loss": 0.3047, + "step": 29917 + }, + { + "epoch": 0.6585516178736518, + "grad_norm": 1.91410625691016, + "learning_rate": 2.7582164645745006e-06, + "loss": 0.2536, + "step": 29918 + }, + { + "epoch": 0.6585736297600704, + "grad_norm": 1.7028944088151527, + "learning_rate": 2.7578978500539555e-06, + "loss": 0.4052, + "step": 29919 + }, + { + "epoch": 0.6585956416464891, + "grad_norm": 0.6043389453033637, + "learning_rate": 2.7575792469287737e-06, + "loss": 0.4261, + "step": 29920 + }, + { + "epoch": 0.6586176535329078, + "grad_norm": 1.431532633589564, + "learning_rate": 2.7572606552005744e-06, + "loss": 0.2235, + "step": 29921 + }, + { + "epoch": 0.6586396654193264, + "grad_norm": 0.6587859758268431, + "learning_rate": 2.7569420748709783e-06, + "loss": 0.4568, + "step": 29922 + }, + { + "epoch": 0.6586616773057451, + "grad_norm": 1.7565885052263066, + "learning_rate": 2.7566235059416034e-06, + "loss": 0.2857, + "step": 29923 + }, + { + "epoch": 0.6586836891921638, + "grad_norm": 2.4508428503239856, + "learning_rate": 2.756304948414067e-06, + "loss": 0.2757, + "step": 29924 + }, + { + "epoch": 0.6587057010785824, + "grad_norm": 1.6774885937128834, + "learning_rate": 2.7559864022899885e-06, + "loss": 0.3075, + "step": 29925 + }, + { + "epoch": 0.6587277129650011, + "grad_norm": 1.4789352557759707, + "learning_rate": 2.75566786757099e-06, + "loss": 0.3034, + "step": 29926 + }, + { + "epoch": 0.6587497248514198, + "grad_norm": 3.7592508787310006, + "learning_rate": 2.7553493442586872e-06, + "loss": 0.415, + "step": 29927 + }, + { + "epoch": 0.6587717367378384, + "grad_norm": 2.8114686538340465, + "learning_rate": 2.7550308323547013e-06, + "loss": 0.3349, + "step": 29928 + }, + { + "epoch": 0.6587937486242571, + "grad_norm": 1.8215670499213166, + "learning_rate": 2.754712331860648e-06, + "loss": 0.3595, + "step": 29929 + }, + { + "epoch": 0.6588157605106758, + "grad_norm": 3.7960220235897455, + "learning_rate": 2.754393842778149e-06, + "loss": 0.3786, + "step": 29930 + }, + { + "epoch": 0.6588377723970944, + "grad_norm": 1.7379087036937504, + "learning_rate": 2.7540753651088237e-06, + "loss": 0.3824, + "step": 29931 + }, + { + "epoch": 0.6588597842835131, + "grad_norm": 2.980963837481552, + "learning_rate": 2.753756898854287e-06, + "loss": 0.2701, + "step": 29932 + }, + { + "epoch": 0.6588817961699318, + "grad_norm": 1.8416693063106706, + "learning_rate": 2.753438444016161e-06, + "loss": 0.2996, + "step": 29933 + }, + { + "epoch": 0.6589038080563504, + "grad_norm": 1.4262659646783469, + "learning_rate": 2.753120000596062e-06, + "loss": 0.3526, + "step": 29934 + }, + { + "epoch": 0.6589258199427691, + "grad_norm": 2.7134648953928244, + "learning_rate": 2.7528015685956085e-06, + "loss": 0.2768, + "step": 29935 + }, + { + "epoch": 0.6589478318291878, + "grad_norm": 2.6390801981838177, + "learning_rate": 2.7524831480164216e-06, + "loss": 0.2872, + "step": 29936 + }, + { + "epoch": 0.6589698437156064, + "grad_norm": 2.0467406409064566, + "learning_rate": 2.7521647388601157e-06, + "loss": 0.3261, + "step": 29937 + }, + { + "epoch": 0.6589918556020251, + "grad_norm": 0.6214507716109361, + "learning_rate": 2.7518463411283136e-06, + "loss": 0.4474, + "step": 29938 + }, + { + "epoch": 0.6590138674884437, + "grad_norm": 1.1511456385996612, + "learning_rate": 2.7515279548226293e-06, + "loss": 0.2438, + "step": 29939 + }, + { + "epoch": 0.6590358793748624, + "grad_norm": 0.6279973626513, + "learning_rate": 2.7512095799446824e-06, + "loss": 0.4764, + "step": 29940 + }, + { + "epoch": 0.6590578912612811, + "grad_norm": 1.8448869637261571, + "learning_rate": 2.7508912164960933e-06, + "loss": 0.211, + "step": 29941 + }, + { + "epoch": 0.6590799031476997, + "grad_norm": 1.436483183088861, + "learning_rate": 2.750572864478477e-06, + "loss": 0.3139, + "step": 29942 + }, + { + "epoch": 0.6591019150341184, + "grad_norm": 3.1836360133599455, + "learning_rate": 2.750254523893454e-06, + "loss": 0.2614, + "step": 29943 + }, + { + "epoch": 0.6591239269205371, + "grad_norm": 1.771155775734728, + "learning_rate": 2.7499361947426386e-06, + "loss": 0.2926, + "step": 29944 + }, + { + "epoch": 0.6591459388069557, + "grad_norm": 1.4470997336013316, + "learning_rate": 2.7496178770276517e-06, + "loss": 0.2672, + "step": 29945 + }, + { + "epoch": 0.6591679506933744, + "grad_norm": 1.833394221057357, + "learning_rate": 2.7492995707501112e-06, + "loss": 0.2589, + "step": 29946 + }, + { + "epoch": 0.6591899625797931, + "grad_norm": 0.6181955068951589, + "learning_rate": 2.748981275911633e-06, + "loss": 0.4462, + "step": 29947 + }, + { + "epoch": 0.6592119744662117, + "grad_norm": 1.414417847608155, + "learning_rate": 2.7486629925138386e-06, + "loss": 0.2542, + "step": 29948 + }, + { + "epoch": 0.6592339863526304, + "grad_norm": 1.6903251834245276, + "learning_rate": 2.7483447205583417e-06, + "loss": 0.3487, + "step": 29949 + }, + { + "epoch": 0.6592559982390491, + "grad_norm": 4.807192909308563, + "learning_rate": 2.74802646004676e-06, + "loss": 0.2768, + "step": 29950 + }, + { + "epoch": 0.6592780101254677, + "grad_norm": 1.6409545131268326, + "learning_rate": 2.7477082109807124e-06, + "loss": 0.3345, + "step": 29951 + }, + { + "epoch": 0.6593000220118864, + "grad_norm": 2.0636511263194888, + "learning_rate": 2.7473899733618157e-06, + "loss": 0.2435, + "step": 29952 + }, + { + "epoch": 0.6593220338983051, + "grad_norm": 2.3461620033707353, + "learning_rate": 2.7470717471916895e-06, + "loss": 0.2744, + "step": 29953 + }, + { + "epoch": 0.6593440457847237, + "grad_norm": 1.7004537659213883, + "learning_rate": 2.7467535324719508e-06, + "loss": 0.2339, + "step": 29954 + }, + { + "epoch": 0.6593660576711424, + "grad_norm": 1.7431246912857883, + "learning_rate": 2.7464353292042127e-06, + "loss": 0.2744, + "step": 29955 + }, + { + "epoch": 0.6593880695575611, + "grad_norm": 1.5898610497969743, + "learning_rate": 2.746117137390095e-06, + "loss": 0.3701, + "step": 29956 + }, + { + "epoch": 0.6594100814439797, + "grad_norm": 2.3692360946262467, + "learning_rate": 2.745798957031216e-06, + "loss": 0.291, + "step": 29957 + }, + { + "epoch": 0.6594320933303984, + "grad_norm": 1.5941734824774214, + "learning_rate": 2.7454807881291945e-06, + "loss": 0.3226, + "step": 29958 + }, + { + "epoch": 0.6594541052168171, + "grad_norm": 1.5483219645173378, + "learning_rate": 2.7451626306856445e-06, + "loss": 0.2664, + "step": 29959 + }, + { + "epoch": 0.6594761171032357, + "grad_norm": 1.8284775603806467, + "learning_rate": 2.7448444847021814e-06, + "loss": 0.2326, + "step": 29960 + }, + { + "epoch": 0.6594981289896544, + "grad_norm": 1.6101786740249202, + "learning_rate": 2.7445263501804254e-06, + "loss": 0.2393, + "step": 29961 + }, + { + "epoch": 0.659520140876073, + "grad_norm": 1.2661052620748732, + "learning_rate": 2.7442082271219916e-06, + "loss": 0.2188, + "step": 29962 + }, + { + "epoch": 0.6595421527624917, + "grad_norm": 1.9348425053053044, + "learning_rate": 2.7438901155285e-06, + "loss": 0.3421, + "step": 29963 + }, + { + "epoch": 0.6595641646489104, + "grad_norm": 1.6455263930871697, + "learning_rate": 2.7435720154015653e-06, + "loss": 0.3674, + "step": 29964 + }, + { + "epoch": 0.659586176535329, + "grad_norm": 1.4116689673468832, + "learning_rate": 2.7432539267428017e-06, + "loss": 0.2963, + "step": 29965 + }, + { + "epoch": 0.6596081884217477, + "grad_norm": 3.288761891097639, + "learning_rate": 2.742935849553828e-06, + "loss": 0.3561, + "step": 29966 + }, + { + "epoch": 0.6596302003081664, + "grad_norm": 0.6217477952570785, + "learning_rate": 2.7426177838362634e-06, + "loss": 0.4565, + "step": 29967 + }, + { + "epoch": 0.659652212194585, + "grad_norm": 1.2905789528414982, + "learning_rate": 2.7422997295917204e-06, + "loss": 0.2592, + "step": 29968 + }, + { + "epoch": 0.6596742240810037, + "grad_norm": 1.7809202712195034, + "learning_rate": 2.7419816868218185e-06, + "loss": 0.3123, + "step": 29969 + }, + { + "epoch": 0.6596962359674224, + "grad_norm": 1.919554866330479, + "learning_rate": 2.7416636555281707e-06, + "loss": 0.3003, + "step": 29970 + }, + { + "epoch": 0.659718247853841, + "grad_norm": 15.12893332834264, + "learning_rate": 2.7413456357123958e-06, + "loss": 0.2495, + "step": 29971 + }, + { + "epoch": 0.6597402597402597, + "grad_norm": 1.5289737227579179, + "learning_rate": 2.7410276273761115e-06, + "loss": 0.3125, + "step": 29972 + }, + { + "epoch": 0.6597622716266784, + "grad_norm": 1.4678758456195122, + "learning_rate": 2.7407096305209306e-06, + "loss": 0.3409, + "step": 29973 + }, + { + "epoch": 0.659784283513097, + "grad_norm": 2.9408421489571626, + "learning_rate": 2.740391645148473e-06, + "loss": 0.3478, + "step": 29974 + }, + { + "epoch": 0.6598062953995157, + "grad_norm": 1.437628323031239, + "learning_rate": 2.7400736712603505e-06, + "loss": 0.3061, + "step": 29975 + }, + { + "epoch": 0.6598283072859344, + "grad_norm": 2.021102421160275, + "learning_rate": 2.7397557088581823e-06, + "loss": 0.3537, + "step": 29976 + }, + { + "epoch": 0.659850319172353, + "grad_norm": 1.5990811614677252, + "learning_rate": 2.739437757943585e-06, + "loss": 0.3759, + "step": 29977 + }, + { + "epoch": 0.6598723310587717, + "grad_norm": 1.7560816658481142, + "learning_rate": 2.739119818518172e-06, + "loss": 0.3481, + "step": 29978 + }, + { + "epoch": 0.6598943429451904, + "grad_norm": 4.110693600557093, + "learning_rate": 2.738801890583561e-06, + "loss": 0.2613, + "step": 29979 + }, + { + "epoch": 0.659916354831609, + "grad_norm": 0.6102941869224705, + "learning_rate": 2.738483974141367e-06, + "loss": 0.4655, + "step": 29980 + }, + { + "epoch": 0.6599383667180277, + "grad_norm": 14.064173036506274, + "learning_rate": 2.7381660691932056e-06, + "loss": 0.293, + "step": 29981 + }, + { + "epoch": 0.6599603786044465, + "grad_norm": 1.768318297852186, + "learning_rate": 2.7378481757406945e-06, + "loss": 0.2991, + "step": 29982 + }, + { + "epoch": 0.659982390490865, + "grad_norm": 1.557874646265113, + "learning_rate": 2.7375302937854466e-06, + "loss": 0.3695, + "step": 29983 + }, + { + "epoch": 0.6600044023772837, + "grad_norm": 2.7388158878228874, + "learning_rate": 2.7372124233290796e-06, + "loss": 0.2963, + "step": 29984 + }, + { + "epoch": 0.6600264142637025, + "grad_norm": 1.840129226831585, + "learning_rate": 2.7368945643732093e-06, + "loss": 0.3066, + "step": 29985 + }, + { + "epoch": 0.660048426150121, + "grad_norm": 2.5986597259331403, + "learning_rate": 2.736576716919448e-06, + "loss": 0.3172, + "step": 29986 + }, + { + "epoch": 0.6600704380365398, + "grad_norm": 1.6894361580200012, + "learning_rate": 2.7362588809694134e-06, + "loss": 0.2963, + "step": 29987 + }, + { + "epoch": 0.6600924499229583, + "grad_norm": 1.4038518093971561, + "learning_rate": 2.7359410565247214e-06, + "loss": 0.2691, + "step": 29988 + }, + { + "epoch": 0.660114461809377, + "grad_norm": 2.0492066678636385, + "learning_rate": 2.7356232435869877e-06, + "loss": 0.3197, + "step": 29989 + }, + { + "epoch": 0.6601364736957958, + "grad_norm": 1.363690540243014, + "learning_rate": 2.735305442157827e-06, + "loss": 0.2932, + "step": 29990 + }, + { + "epoch": 0.6601584855822143, + "grad_norm": 1.310614808362179, + "learning_rate": 2.7349876522388525e-06, + "loss": 0.2205, + "step": 29991 + }, + { + "epoch": 0.660180497468633, + "grad_norm": 1.621325702462232, + "learning_rate": 2.7346698738316806e-06, + "loss": 0.296, + "step": 29992 + }, + { + "epoch": 0.6602025093550518, + "grad_norm": 2.046546683827553, + "learning_rate": 2.734352106937927e-06, + "loss": 0.3453, + "step": 29993 + }, + { + "epoch": 0.6602245212414704, + "grad_norm": 1.6437318990177456, + "learning_rate": 2.7340343515592084e-06, + "loss": 0.3534, + "step": 29994 + }, + { + "epoch": 0.6602465331278891, + "grad_norm": 1.8703928124628417, + "learning_rate": 2.7337166076971377e-06, + "loss": 0.3124, + "step": 29995 + }, + { + "epoch": 0.6602685450143078, + "grad_norm": 0.6275103086135715, + "learning_rate": 2.733398875353328e-06, + "loss": 0.4552, + "step": 29996 + }, + { + "epoch": 0.6602905569007264, + "grad_norm": 1.5121423212381706, + "learning_rate": 2.7330811545293966e-06, + "loss": 0.2897, + "step": 29997 + }, + { + "epoch": 0.6603125687871451, + "grad_norm": 1.3409416757161594, + "learning_rate": 2.7327634452269567e-06, + "loss": 0.2928, + "step": 29998 + }, + { + "epoch": 0.6603345806735638, + "grad_norm": 0.6221902015622816, + "learning_rate": 2.732445747447627e-06, + "loss": 0.4574, + "step": 29999 + }, + { + "epoch": 0.6603565925599824, + "grad_norm": 1.7825533563756044, + "learning_rate": 2.7321280611930203e-06, + "loss": 0.3577, + "step": 30000 + }, + { + "epoch": 0.6603786044464011, + "grad_norm": 1.82947316487894, + "learning_rate": 2.731810386464747e-06, + "loss": 0.3026, + "step": 30001 + }, + { + "epoch": 0.6604006163328198, + "grad_norm": 0.5919824899612579, + "learning_rate": 2.731492723264425e-06, + "loss": 0.4352, + "step": 30002 + }, + { + "epoch": 0.6604226282192384, + "grad_norm": 1.4880877469519902, + "learning_rate": 2.731175071593669e-06, + "loss": 0.2539, + "step": 30003 + }, + { + "epoch": 0.6604446401056571, + "grad_norm": 1.612201266802494, + "learning_rate": 2.730857431454096e-06, + "loss": 0.3354, + "step": 30004 + }, + { + "epoch": 0.6604666519920758, + "grad_norm": 1.4405576719281092, + "learning_rate": 2.7305398028473164e-06, + "loss": 0.2131, + "step": 30005 + }, + { + "epoch": 0.6604886638784944, + "grad_norm": 2.2404915098895537, + "learning_rate": 2.7302221857749446e-06, + "loss": 0.3218, + "step": 30006 + }, + { + "epoch": 0.6605106757649131, + "grad_norm": 1.2581319526062993, + "learning_rate": 2.729904580238595e-06, + "loss": 0.2655, + "step": 30007 + }, + { + "epoch": 0.6605326876513318, + "grad_norm": 0.6110891863208319, + "learning_rate": 2.729586986239886e-06, + "loss": 0.4587, + "step": 30008 + }, + { + "epoch": 0.6605546995377504, + "grad_norm": 1.6329256065816016, + "learning_rate": 2.7292694037804256e-06, + "loss": 0.2357, + "step": 30009 + }, + { + "epoch": 0.6605767114241691, + "grad_norm": 1.7274694310022611, + "learning_rate": 2.728951832861834e-06, + "loss": 0.3343, + "step": 30010 + }, + { + "epoch": 0.6605987233105877, + "grad_norm": 2.859009389855359, + "learning_rate": 2.7286342734857185e-06, + "loss": 0.3526, + "step": 30011 + }, + { + "epoch": 0.6606207351970064, + "grad_norm": 0.6203996477507181, + "learning_rate": 2.728316725653698e-06, + "loss": 0.4448, + "step": 30012 + }, + { + "epoch": 0.6606427470834251, + "grad_norm": 2.1338384063456792, + "learning_rate": 2.7279991893673865e-06, + "loss": 0.2886, + "step": 30013 + }, + { + "epoch": 0.6606647589698437, + "grad_norm": 3.3646091455412637, + "learning_rate": 2.7276816646283944e-06, + "loss": 0.3317, + "step": 30014 + }, + { + "epoch": 0.6606867708562624, + "grad_norm": 1.2541663211694887, + "learning_rate": 2.7273641514383396e-06, + "loss": 0.2156, + "step": 30015 + }, + { + "epoch": 0.6607087827426811, + "grad_norm": 1.6803837330770166, + "learning_rate": 2.7270466497988326e-06, + "loss": 0.2607, + "step": 30016 + }, + { + "epoch": 0.6607307946290997, + "grad_norm": 1.5451720177810875, + "learning_rate": 2.7267291597114874e-06, + "loss": 0.2377, + "step": 30017 + }, + { + "epoch": 0.6607528065155184, + "grad_norm": 3.5955938800517697, + "learning_rate": 2.7264116811779207e-06, + "loss": 0.2527, + "step": 30018 + }, + { + "epoch": 0.6607748184019371, + "grad_norm": 1.9237156073674293, + "learning_rate": 2.7260942141997422e-06, + "loss": 0.3035, + "step": 30019 + }, + { + "epoch": 0.6607968302883557, + "grad_norm": 3.351966255146611, + "learning_rate": 2.725776758778569e-06, + "loss": 0.2602, + "step": 30020 + }, + { + "epoch": 0.6608188421747744, + "grad_norm": 1.4165143502248498, + "learning_rate": 2.7254593149160102e-06, + "loss": 0.2171, + "step": 30021 + }, + { + "epoch": 0.6608408540611931, + "grad_norm": 0.6540279131118222, + "learning_rate": 2.7251418826136843e-06, + "loss": 0.4594, + "step": 30022 + }, + { + "epoch": 0.6608628659476117, + "grad_norm": 0.6223187686468589, + "learning_rate": 2.724824461873199e-06, + "loss": 0.4839, + "step": 30023 + }, + { + "epoch": 0.6608848778340304, + "grad_norm": 0.6309138461680495, + "learning_rate": 2.7245070526961714e-06, + "loss": 0.4563, + "step": 30024 + }, + { + "epoch": 0.6609068897204491, + "grad_norm": 4.675209402105075, + "learning_rate": 2.7241896550842158e-06, + "loss": 0.3027, + "step": 30025 + }, + { + "epoch": 0.6609289016068677, + "grad_norm": 1.6930241054200623, + "learning_rate": 2.723872269038943e-06, + "loss": 0.2935, + "step": 30026 + }, + { + "epoch": 0.6609509134932864, + "grad_norm": 2.6395851740082525, + "learning_rate": 2.723554894561965e-06, + "loss": 0.3087, + "step": 30027 + }, + { + "epoch": 0.6609729253797051, + "grad_norm": 1.5781527543419493, + "learning_rate": 2.7232375316548957e-06, + "loss": 0.2187, + "step": 30028 + }, + { + "epoch": 0.6609949372661237, + "grad_norm": 1.0779611753981422, + "learning_rate": 2.7229201803193494e-06, + "loss": 0.2558, + "step": 30029 + }, + { + "epoch": 0.6610169491525424, + "grad_norm": 2.0579888121766814, + "learning_rate": 2.72260284055694e-06, + "loss": 0.3012, + "step": 30030 + }, + { + "epoch": 0.6610389610389611, + "grad_norm": 1.9864222369176328, + "learning_rate": 2.7222855123692786e-06, + "loss": 0.3478, + "step": 30031 + }, + { + "epoch": 0.6610609729253797, + "grad_norm": 0.6068475585195233, + "learning_rate": 2.721968195757976e-06, + "loss": 0.4243, + "step": 30032 + }, + { + "epoch": 0.6610829848117984, + "grad_norm": 1.6141103639814347, + "learning_rate": 2.7216508907246473e-06, + "loss": 0.3327, + "step": 30033 + }, + { + "epoch": 0.661104996698217, + "grad_norm": 0.6244988620067516, + "learning_rate": 2.7213335972709056e-06, + "loss": 0.456, + "step": 30034 + }, + { + "epoch": 0.6611270085846357, + "grad_norm": 2.2167827572102112, + "learning_rate": 2.7210163153983637e-06, + "loss": 0.3267, + "step": 30035 + }, + { + "epoch": 0.6611490204710544, + "grad_norm": 1.4540021522087114, + "learning_rate": 2.7206990451086344e-06, + "loss": 0.2215, + "step": 30036 + }, + { + "epoch": 0.661171032357473, + "grad_norm": 2.738882481231377, + "learning_rate": 2.7203817864033266e-06, + "loss": 0.3039, + "step": 30037 + }, + { + "epoch": 0.6611930442438917, + "grad_norm": 1.9992470525341375, + "learning_rate": 2.7200645392840553e-06, + "loss": 0.1889, + "step": 30038 + }, + { + "epoch": 0.6612150561303104, + "grad_norm": 2.4138322036614173, + "learning_rate": 2.719747303752433e-06, + "loss": 0.2629, + "step": 30039 + }, + { + "epoch": 0.661237068016729, + "grad_norm": 1.1580185613908642, + "learning_rate": 2.7194300798100736e-06, + "loss": 0.2098, + "step": 30040 + }, + { + "epoch": 0.6612590799031477, + "grad_norm": 1.5693014586093579, + "learning_rate": 2.719112867458588e-06, + "loss": 0.3312, + "step": 30041 + }, + { + "epoch": 0.6612810917895664, + "grad_norm": 1.3474620577606862, + "learning_rate": 2.7187956666995864e-06, + "loss": 0.2486, + "step": 30042 + }, + { + "epoch": 0.661303103675985, + "grad_norm": 1.6366492715938123, + "learning_rate": 2.7184784775346824e-06, + "loss": 0.2818, + "step": 30043 + }, + { + "epoch": 0.6613251155624037, + "grad_norm": 1.3472491185292448, + "learning_rate": 2.7181612999654884e-06, + "loss": 0.2467, + "step": 30044 + }, + { + "epoch": 0.6613471274488224, + "grad_norm": 0.5997598701030807, + "learning_rate": 2.717844133993618e-06, + "loss": 0.4102, + "step": 30045 + }, + { + "epoch": 0.661369139335241, + "grad_norm": 0.6473799563029217, + "learning_rate": 2.7175269796206816e-06, + "loss": 0.473, + "step": 30046 + }, + { + "epoch": 0.6613911512216597, + "grad_norm": 1.7572358267428219, + "learning_rate": 2.7172098368482896e-06, + "loss": 0.344, + "step": 30047 + }, + { + "epoch": 0.6614131631080784, + "grad_norm": 2.0658812517901826, + "learning_rate": 2.7168927056780553e-06, + "loss": 0.3015, + "step": 30048 + }, + { + "epoch": 0.661435174994497, + "grad_norm": 2.301738264257325, + "learning_rate": 2.716575586111592e-06, + "loss": 0.3221, + "step": 30049 + }, + { + "epoch": 0.6614571868809157, + "grad_norm": 0.6142650616563288, + "learning_rate": 2.7162584781505085e-06, + "loss": 0.4423, + "step": 30050 + }, + { + "epoch": 0.6614791987673344, + "grad_norm": 3.3914218288767928, + "learning_rate": 2.71594138179642e-06, + "loss": 0.3577, + "step": 30051 + }, + { + "epoch": 0.661501210653753, + "grad_norm": 0.646717446959303, + "learning_rate": 2.715624297050934e-06, + "loss": 0.4969, + "step": 30052 + }, + { + "epoch": 0.6615232225401717, + "grad_norm": 1.4953907832369462, + "learning_rate": 2.7153072239156643e-06, + "loss": 0.3099, + "step": 30053 + }, + { + "epoch": 0.6615452344265904, + "grad_norm": 2.105898116250849, + "learning_rate": 2.714990162392224e-06, + "loss": 0.3573, + "step": 30054 + }, + { + "epoch": 0.661567246313009, + "grad_norm": 1.834937241588076, + "learning_rate": 2.7146731124822213e-06, + "loss": 0.2836, + "step": 30055 + }, + { + "epoch": 0.6615892581994277, + "grad_norm": 1.620040526486717, + "learning_rate": 2.714356074187271e-06, + "loss": 0.2283, + "step": 30056 + }, + { + "epoch": 0.6616112700858463, + "grad_norm": 0.6260231827382812, + "learning_rate": 2.7140390475089805e-06, + "loss": 0.4668, + "step": 30057 + }, + { + "epoch": 0.661633281972265, + "grad_norm": 1.6443015717171672, + "learning_rate": 2.713722032448965e-06, + "loss": 0.2616, + "step": 30058 + }, + { + "epoch": 0.6616552938586837, + "grad_norm": 1.9427565459826779, + "learning_rate": 2.713405029008832e-06, + "loss": 0.2989, + "step": 30059 + }, + { + "epoch": 0.6616773057451023, + "grad_norm": 0.6371059430569412, + "learning_rate": 2.713088037190194e-06, + "loss": 0.4519, + "step": 30060 + }, + { + "epoch": 0.661699317631521, + "grad_norm": 2.5093378796370334, + "learning_rate": 2.7127710569946654e-06, + "loss": 0.2847, + "step": 30061 + }, + { + "epoch": 0.6617213295179397, + "grad_norm": 1.7287884329518206, + "learning_rate": 2.7124540884238517e-06, + "loss": 0.3719, + "step": 30062 + }, + { + "epoch": 0.6617433414043583, + "grad_norm": 1.5924429533642757, + "learning_rate": 2.712137131479369e-06, + "loss": 0.2104, + "step": 30063 + }, + { + "epoch": 0.661765353290777, + "grad_norm": 2.0545890844398795, + "learning_rate": 2.7118201861628236e-06, + "loss": 0.2829, + "step": 30064 + }, + { + "epoch": 0.6617873651771957, + "grad_norm": 1.5838128027218505, + "learning_rate": 2.711503252475829e-06, + "loss": 0.28, + "step": 30065 + }, + { + "epoch": 0.6618093770636143, + "grad_norm": 4.978894971838284, + "learning_rate": 2.711186330419997e-06, + "loss": 0.3071, + "step": 30066 + }, + { + "epoch": 0.661831388950033, + "grad_norm": 2.129582679320449, + "learning_rate": 2.710869419996937e-06, + "loss": 0.3211, + "step": 30067 + }, + { + "epoch": 0.6618534008364517, + "grad_norm": 1.3681125036488806, + "learning_rate": 2.710552521208258e-06, + "loss": 0.2828, + "step": 30068 + }, + { + "epoch": 0.6618754127228703, + "grad_norm": 2.8834954822015484, + "learning_rate": 2.7102356340555715e-06, + "loss": 0.2987, + "step": 30069 + }, + { + "epoch": 0.661897424609289, + "grad_norm": 1.6557171767710115, + "learning_rate": 2.7099187585404898e-06, + "loss": 0.3158, + "step": 30070 + }, + { + "epoch": 0.6619194364957077, + "grad_norm": 23.909550431322412, + "learning_rate": 2.7096018946646237e-06, + "loss": 0.277, + "step": 30071 + }, + { + "epoch": 0.6619414483821263, + "grad_norm": 1.4990169038206314, + "learning_rate": 2.7092850424295823e-06, + "loss": 0.3332, + "step": 30072 + }, + { + "epoch": 0.661963460268545, + "grad_norm": 1.3314624119670797, + "learning_rate": 2.708968201836974e-06, + "loss": 0.2277, + "step": 30073 + }, + { + "epoch": 0.6619854721549637, + "grad_norm": 1.6886310400556772, + "learning_rate": 2.708651372888411e-06, + "loss": 0.3523, + "step": 30074 + }, + { + "epoch": 0.6620074840413823, + "grad_norm": 1.3651736668219758, + "learning_rate": 2.7083345555855044e-06, + "loss": 0.2747, + "step": 30075 + }, + { + "epoch": 0.662029495927801, + "grad_norm": 1.3609374387405542, + "learning_rate": 2.7080177499298648e-06, + "loss": 0.3107, + "step": 30076 + }, + { + "epoch": 0.6620515078142197, + "grad_norm": 2.301821956603492, + "learning_rate": 2.7077009559231017e-06, + "loss": 0.3345, + "step": 30077 + }, + { + "epoch": 0.6620735197006383, + "grad_norm": 2.4868574030212365, + "learning_rate": 2.7073841735668228e-06, + "loss": 0.2816, + "step": 30078 + }, + { + "epoch": 0.662095531587057, + "grad_norm": 2.4543780773869597, + "learning_rate": 2.70706740286264e-06, + "loss": 0.3262, + "step": 30079 + }, + { + "epoch": 0.6621175434734756, + "grad_norm": 1.971614350073571, + "learning_rate": 2.7067506438121634e-06, + "loss": 0.3435, + "step": 30080 + }, + { + "epoch": 0.6621395553598943, + "grad_norm": 1.4257621916184724, + "learning_rate": 2.7064338964170047e-06, + "loss": 0.3002, + "step": 30081 + }, + { + "epoch": 0.662161567246313, + "grad_norm": 1.6605627414316169, + "learning_rate": 2.7061171606787718e-06, + "loss": 0.2896, + "step": 30082 + }, + { + "epoch": 0.6621835791327316, + "grad_norm": 1.2447916203977847, + "learning_rate": 2.7058004365990733e-06, + "loss": 0.2569, + "step": 30083 + }, + { + "epoch": 0.6622055910191503, + "grad_norm": 2.019801195183138, + "learning_rate": 2.7054837241795198e-06, + "loss": 0.2963, + "step": 30084 + }, + { + "epoch": 0.662227602905569, + "grad_norm": 1.629221755765395, + "learning_rate": 2.705167023421723e-06, + "loss": 0.3353, + "step": 30085 + }, + { + "epoch": 0.6622496147919876, + "grad_norm": 0.6307022870968297, + "learning_rate": 2.704850334327289e-06, + "loss": 0.4643, + "step": 30086 + }, + { + "epoch": 0.6622716266784063, + "grad_norm": 1.5156457870371403, + "learning_rate": 2.7045336568978314e-06, + "loss": 0.3097, + "step": 30087 + }, + { + "epoch": 0.662293638564825, + "grad_norm": 0.6425217886476403, + "learning_rate": 2.704216991134956e-06, + "loss": 0.4475, + "step": 30088 + }, + { + "epoch": 0.6623156504512436, + "grad_norm": 1.9689279436666247, + "learning_rate": 2.703900337040273e-06, + "loss": 0.279, + "step": 30089 + }, + { + "epoch": 0.6623376623376623, + "grad_norm": 1.5722619576236343, + "learning_rate": 2.703583694615395e-06, + "loss": 0.3324, + "step": 30090 + }, + { + "epoch": 0.662359674224081, + "grad_norm": 1.5318185381969411, + "learning_rate": 2.703267063861927e-06, + "loss": 0.2436, + "step": 30091 + }, + { + "epoch": 0.6623816861104996, + "grad_norm": 2.4431392908291767, + "learning_rate": 2.702950444781482e-06, + "loss": 0.2577, + "step": 30092 + }, + { + "epoch": 0.6624036979969183, + "grad_norm": 1.6910781435553262, + "learning_rate": 2.7026338373756655e-06, + "loss": 0.3455, + "step": 30093 + }, + { + "epoch": 0.662425709883337, + "grad_norm": 3.2433038595977433, + "learning_rate": 2.7023172416460898e-06, + "loss": 0.2787, + "step": 30094 + }, + { + "epoch": 0.6624477217697556, + "grad_norm": 1.8182563815558346, + "learning_rate": 2.702000657594361e-06, + "loss": 0.3715, + "step": 30095 + }, + { + "epoch": 0.6624697336561743, + "grad_norm": 1.9973406500220128, + "learning_rate": 2.7016840852220904e-06, + "loss": 0.3327, + "step": 30096 + }, + { + "epoch": 0.662491745542593, + "grad_norm": 1.864265902819849, + "learning_rate": 2.701367524530888e-06, + "loss": 0.3232, + "step": 30097 + }, + { + "epoch": 0.6625137574290116, + "grad_norm": 2.265875643758991, + "learning_rate": 2.701050975522359e-06, + "loss": 0.2647, + "step": 30098 + }, + { + "epoch": 0.6625357693154303, + "grad_norm": 1.2485079541655222, + "learning_rate": 2.7007344381981164e-06, + "loss": 0.2421, + "step": 30099 + }, + { + "epoch": 0.662557781201849, + "grad_norm": 1.6749332661905139, + "learning_rate": 2.7004179125597653e-06, + "loss": 0.3004, + "step": 30100 + }, + { + "epoch": 0.6625797930882676, + "grad_norm": 1.592473938293942, + "learning_rate": 2.7001013986089154e-06, + "loss": 0.3044, + "step": 30101 + }, + { + "epoch": 0.6626018049746863, + "grad_norm": 3.459020310965281, + "learning_rate": 2.699784896347178e-06, + "loss": 0.3599, + "step": 30102 + }, + { + "epoch": 0.6626238168611049, + "grad_norm": 1.5063370759457375, + "learning_rate": 2.6994684057761577e-06, + "loss": 0.2945, + "step": 30103 + }, + { + "epoch": 0.6626458287475236, + "grad_norm": 1.973731195222725, + "learning_rate": 2.6991519268974676e-06, + "loss": 0.2293, + "step": 30104 + }, + { + "epoch": 0.6626678406339424, + "grad_norm": 1.6729109750580282, + "learning_rate": 2.6988354597127108e-06, + "loss": 0.3264, + "step": 30105 + }, + { + "epoch": 0.662689852520361, + "grad_norm": 1.7078202721448785, + "learning_rate": 2.698519004223499e-06, + "loss": 0.3683, + "step": 30106 + }, + { + "epoch": 0.6627118644067796, + "grad_norm": 1.2683643280410313, + "learning_rate": 2.698202560431442e-06, + "loss": 0.2639, + "step": 30107 + }, + { + "epoch": 0.6627338762931984, + "grad_norm": 2.2451649575122117, + "learning_rate": 2.6978861283381463e-06, + "loss": 0.3309, + "step": 30108 + }, + { + "epoch": 0.662755888179617, + "grad_norm": 0.6160719001863928, + "learning_rate": 2.6975697079452178e-06, + "loss": 0.4392, + "step": 30109 + }, + { + "epoch": 0.6627779000660357, + "grad_norm": 1.9316132872055314, + "learning_rate": 2.697253299254267e-06, + "loss": 0.2711, + "step": 30110 + }, + { + "epoch": 0.6627999119524544, + "grad_norm": 1.8711035721645395, + "learning_rate": 2.6969369022669024e-06, + "loss": 0.2604, + "step": 30111 + }, + { + "epoch": 0.662821923838873, + "grad_norm": 1.9640424413069884, + "learning_rate": 2.696620516984733e-06, + "loss": 0.3276, + "step": 30112 + }, + { + "epoch": 0.6628439357252917, + "grad_norm": 2.261027893148889, + "learning_rate": 2.696304143409366e-06, + "loss": 0.2345, + "step": 30113 + }, + { + "epoch": 0.6628659476117104, + "grad_norm": 1.8849128275627391, + "learning_rate": 2.695987781542407e-06, + "loss": 0.2487, + "step": 30114 + }, + { + "epoch": 0.662887959498129, + "grad_norm": 1.853480665909603, + "learning_rate": 2.695671431385465e-06, + "loss": 0.3064, + "step": 30115 + }, + { + "epoch": 0.6629099713845477, + "grad_norm": 1.570492590376657, + "learning_rate": 2.6953550929401496e-06, + "loss": 0.2563, + "step": 30116 + }, + { + "epoch": 0.6629319832709664, + "grad_norm": 2.6599414170584876, + "learning_rate": 2.6950387662080684e-06, + "loss": 0.3789, + "step": 30117 + }, + { + "epoch": 0.662953995157385, + "grad_norm": 1.6211664197322673, + "learning_rate": 2.694722451190829e-06, + "loss": 0.2861, + "step": 30118 + }, + { + "epoch": 0.6629760070438037, + "grad_norm": 1.5429739872778956, + "learning_rate": 2.694406147890036e-06, + "loss": 0.2554, + "step": 30119 + }, + { + "epoch": 0.6629980189302224, + "grad_norm": 2.6397032534453007, + "learning_rate": 2.694089856307299e-06, + "loss": 0.4159, + "step": 30120 + }, + { + "epoch": 0.663020030816641, + "grad_norm": 1.6960771624094786, + "learning_rate": 2.6937735764442262e-06, + "loss": 0.3016, + "step": 30121 + }, + { + "epoch": 0.6630420427030597, + "grad_norm": 1.4063249161740268, + "learning_rate": 2.693457308302426e-06, + "loss": 0.3196, + "step": 30122 + }, + { + "epoch": 0.6630640545894784, + "grad_norm": 1.6604017438696639, + "learning_rate": 2.6931410518835054e-06, + "loss": 0.3753, + "step": 30123 + }, + { + "epoch": 0.663086066475897, + "grad_norm": 1.9951677609675926, + "learning_rate": 2.6928248071890685e-06, + "loss": 0.4088, + "step": 30124 + }, + { + "epoch": 0.6631080783623157, + "grad_norm": 2.197375632265431, + "learning_rate": 2.6925085742207245e-06, + "loss": 0.3483, + "step": 30125 + }, + { + "epoch": 0.6631300902487344, + "grad_norm": 2.392406267605677, + "learning_rate": 2.692192352980084e-06, + "loss": 0.3068, + "step": 30126 + }, + { + "epoch": 0.663152102135153, + "grad_norm": 1.4083117030976613, + "learning_rate": 2.691876143468749e-06, + "loss": 0.2338, + "step": 30127 + }, + { + "epoch": 0.6631741140215717, + "grad_norm": 1.5085037875034946, + "learning_rate": 2.691559945688331e-06, + "loss": 0.1941, + "step": 30128 + }, + { + "epoch": 0.6631961259079903, + "grad_norm": 1.9889142653242309, + "learning_rate": 2.691243759640432e-06, + "loss": 0.288, + "step": 30129 + }, + { + "epoch": 0.663218137794409, + "grad_norm": 2.3954261542666813, + "learning_rate": 2.6909275853266646e-06, + "loss": 0.3559, + "step": 30130 + }, + { + "epoch": 0.6632401496808277, + "grad_norm": 1.6352700866270853, + "learning_rate": 2.6906114227486313e-06, + "loss": 0.2139, + "step": 30131 + }, + { + "epoch": 0.6632621615672463, + "grad_norm": 1.8743305542221151, + "learning_rate": 2.690295271907941e-06, + "loss": 0.2629, + "step": 30132 + }, + { + "epoch": 0.663284173453665, + "grad_norm": 0.6297578570381871, + "learning_rate": 2.689979132806202e-06, + "loss": 0.4656, + "step": 30133 + }, + { + "epoch": 0.6633061853400837, + "grad_norm": 2.226243432606182, + "learning_rate": 2.689663005445017e-06, + "loss": 0.3746, + "step": 30134 + }, + { + "epoch": 0.6633281972265023, + "grad_norm": 2.2639883475027083, + "learning_rate": 2.6893468898259973e-06, + "loss": 0.2964, + "step": 30135 + }, + { + "epoch": 0.663350209112921, + "grad_norm": 1.2281812082222638, + "learning_rate": 2.6890307859507447e-06, + "loss": 0.2387, + "step": 30136 + }, + { + "epoch": 0.6633722209993397, + "grad_norm": 1.4980395273346054, + "learning_rate": 2.68871469382087e-06, + "loss": 0.3047, + "step": 30137 + }, + { + "epoch": 0.6633942328857583, + "grad_norm": 2.203555758598424, + "learning_rate": 2.688398613437978e-06, + "loss": 0.248, + "step": 30138 + }, + { + "epoch": 0.663416244772177, + "grad_norm": 2.303011847483414, + "learning_rate": 2.6880825448036745e-06, + "loss": 0.2729, + "step": 30139 + }, + { + "epoch": 0.6634382566585957, + "grad_norm": 0.6323952581978148, + "learning_rate": 2.6877664879195686e-06, + "loss": 0.4588, + "step": 30140 + }, + { + "epoch": 0.6634602685450143, + "grad_norm": 3.456695445105305, + "learning_rate": 2.6874504427872626e-06, + "loss": 0.2881, + "step": 30141 + }, + { + "epoch": 0.663482280431433, + "grad_norm": 2.15503197372403, + "learning_rate": 2.6871344094083646e-06, + "loss": 0.2062, + "step": 30142 + }, + { + "epoch": 0.6635042923178517, + "grad_norm": 1.35384517337186, + "learning_rate": 2.6868183877844835e-06, + "loss": 0.2311, + "step": 30143 + }, + { + "epoch": 0.6635263042042703, + "grad_norm": 1.7832563531403294, + "learning_rate": 2.6865023779172207e-06, + "loss": 0.2429, + "step": 30144 + }, + { + "epoch": 0.663548316090689, + "grad_norm": 0.6056815319597577, + "learning_rate": 2.6861863798081868e-06, + "loss": 0.4517, + "step": 30145 + }, + { + "epoch": 0.6635703279771077, + "grad_norm": 35.86346434890433, + "learning_rate": 2.6858703934589835e-06, + "loss": 0.2689, + "step": 30146 + }, + { + "epoch": 0.6635923398635263, + "grad_norm": 0.5914926040875482, + "learning_rate": 2.685554418871219e-06, + "loss": 0.4523, + "step": 30147 + }, + { + "epoch": 0.663614351749945, + "grad_norm": 2.0938929892600155, + "learning_rate": 2.6852384560465016e-06, + "loss": 0.4003, + "step": 30148 + }, + { + "epoch": 0.6636363636363637, + "grad_norm": 2.356816005879952, + "learning_rate": 2.684922504986435e-06, + "loss": 0.2845, + "step": 30149 + }, + { + "epoch": 0.6636583755227823, + "grad_norm": 1.5509184238685632, + "learning_rate": 2.6846065656926223e-06, + "loss": 0.3169, + "step": 30150 + }, + { + "epoch": 0.663680387409201, + "grad_norm": 0.6355529731994654, + "learning_rate": 2.6842906381666715e-06, + "loss": 0.4761, + "step": 30151 + }, + { + "epoch": 0.6637023992956196, + "grad_norm": 1.4004270370775231, + "learning_rate": 2.6839747224101885e-06, + "loss": 0.2889, + "step": 30152 + }, + { + "epoch": 0.6637244111820383, + "grad_norm": 1.7300441073166581, + "learning_rate": 2.6836588184247812e-06, + "loss": 0.2739, + "step": 30153 + }, + { + "epoch": 0.663746423068457, + "grad_norm": 1.6421622676573921, + "learning_rate": 2.6833429262120537e-06, + "loss": 0.2828, + "step": 30154 + }, + { + "epoch": 0.6637684349548756, + "grad_norm": 2.374217757483383, + "learning_rate": 2.683027045773607e-06, + "loss": 0.3307, + "step": 30155 + }, + { + "epoch": 0.6637904468412943, + "grad_norm": 2.5186539436935527, + "learning_rate": 2.682711177111052e-06, + "loss": 0.3032, + "step": 30156 + }, + { + "epoch": 0.663812458727713, + "grad_norm": 1.477466059086678, + "learning_rate": 2.6823953202259913e-06, + "loss": 0.2637, + "step": 30157 + }, + { + "epoch": 0.6638344706141316, + "grad_norm": 1.9340492374055458, + "learning_rate": 2.682079475120034e-06, + "loss": 0.3527, + "step": 30158 + }, + { + "epoch": 0.6638564825005503, + "grad_norm": 2.458179264352977, + "learning_rate": 2.6817636417947816e-06, + "loss": 0.2899, + "step": 30159 + }, + { + "epoch": 0.663878494386969, + "grad_norm": 0.640265873778427, + "learning_rate": 2.681447820251839e-06, + "loss": 0.4457, + "step": 30160 + }, + { + "epoch": 0.6639005062733876, + "grad_norm": 1.7179786798611416, + "learning_rate": 2.6811320104928147e-06, + "loss": 0.295, + "step": 30161 + }, + { + "epoch": 0.6639225181598063, + "grad_norm": 6.457031193304856, + "learning_rate": 2.6808162125193076e-06, + "loss": 0.3591, + "step": 30162 + }, + { + "epoch": 0.663944530046225, + "grad_norm": 1.3654895008380252, + "learning_rate": 2.680500426332931e-06, + "loss": 0.2334, + "step": 30163 + }, + { + "epoch": 0.6639665419326436, + "grad_norm": 2.2103175150818224, + "learning_rate": 2.680184651935286e-06, + "loss": 0.2996, + "step": 30164 + }, + { + "epoch": 0.6639885538190623, + "grad_norm": 1.3466987246803077, + "learning_rate": 2.6798688893279746e-06, + "loss": 0.3766, + "step": 30165 + }, + { + "epoch": 0.664010565705481, + "grad_norm": 0.630712186737293, + "learning_rate": 2.6795531385126067e-06, + "loss": 0.4658, + "step": 30166 + }, + { + "epoch": 0.6640325775918996, + "grad_norm": 1.477664140485878, + "learning_rate": 2.6792373994907827e-06, + "loss": 0.3054, + "step": 30167 + }, + { + "epoch": 0.6640545894783183, + "grad_norm": 1.4350940624987258, + "learning_rate": 2.6789216722641093e-06, + "loss": 0.2841, + "step": 30168 + }, + { + "epoch": 0.664076601364737, + "grad_norm": 1.69030952236656, + "learning_rate": 2.678605956834193e-06, + "loss": 0.2573, + "step": 30169 + }, + { + "epoch": 0.6640986132511556, + "grad_norm": 1.6138835301912544, + "learning_rate": 2.6782902532026344e-06, + "loss": 0.2663, + "step": 30170 + }, + { + "epoch": 0.6641206251375743, + "grad_norm": 1.6739386780095538, + "learning_rate": 2.6779745613710427e-06, + "loss": 0.3003, + "step": 30171 + }, + { + "epoch": 0.664142637023993, + "grad_norm": 1.3962860222309237, + "learning_rate": 2.677658881341018e-06, + "loss": 0.2797, + "step": 30172 + }, + { + "epoch": 0.6641646489104116, + "grad_norm": 2.266809179975522, + "learning_rate": 2.677343213114166e-06, + "loss": 0.3911, + "step": 30173 + }, + { + "epoch": 0.6641866607968303, + "grad_norm": 0.6294426473484169, + "learning_rate": 2.6770275566920945e-06, + "loss": 0.477, + "step": 30174 + }, + { + "epoch": 0.6642086726832489, + "grad_norm": 1.5347487482110593, + "learning_rate": 2.676711912076403e-06, + "loss": 0.1972, + "step": 30175 + }, + { + "epoch": 0.6642306845696676, + "grad_norm": 1.811739087204536, + "learning_rate": 2.6763962792686992e-06, + "loss": 0.2796, + "step": 30176 + }, + { + "epoch": 0.6642526964560863, + "grad_norm": 1.4127122801635905, + "learning_rate": 2.6760806582705847e-06, + "loss": 0.3039, + "step": 30177 + }, + { + "epoch": 0.6642747083425049, + "grad_norm": 1.7823902164773207, + "learning_rate": 2.675765049083664e-06, + "loss": 0.2995, + "step": 30178 + }, + { + "epoch": 0.6642967202289236, + "grad_norm": 1.6827521134545274, + "learning_rate": 2.675449451709544e-06, + "loss": 0.2899, + "step": 30179 + }, + { + "epoch": 0.6643187321153423, + "grad_norm": 1.5589632261670017, + "learning_rate": 2.6751338661498256e-06, + "loss": 0.3143, + "step": 30180 + }, + { + "epoch": 0.6643407440017609, + "grad_norm": 0.6116420212047974, + "learning_rate": 2.674818292406115e-06, + "loss": 0.4682, + "step": 30181 + }, + { + "epoch": 0.6643627558881796, + "grad_norm": 1.4943571094818675, + "learning_rate": 2.674502730480013e-06, + "loss": 0.4018, + "step": 30182 + }, + { + "epoch": 0.6643847677745983, + "grad_norm": 2.0367444870198317, + "learning_rate": 2.6741871803731255e-06, + "loss": 0.3421, + "step": 30183 + }, + { + "epoch": 0.6644067796610169, + "grad_norm": 1.86883942126979, + "learning_rate": 2.6738716420870582e-06, + "loss": 0.2883, + "step": 30184 + }, + { + "epoch": 0.6644287915474356, + "grad_norm": 2.7744703675910394, + "learning_rate": 2.673556115623412e-06, + "loss": 0.2487, + "step": 30185 + }, + { + "epoch": 0.6644508034338543, + "grad_norm": 6.559655328392547, + "learning_rate": 2.6732406009837898e-06, + "loss": 0.2827, + "step": 30186 + }, + { + "epoch": 0.6644728153202729, + "grad_norm": 0.6240098956459076, + "learning_rate": 2.6729250981697963e-06, + "loss": 0.4722, + "step": 30187 + }, + { + "epoch": 0.6644948272066916, + "grad_norm": 3.6334255736813335, + "learning_rate": 2.6726096071830356e-06, + "loss": 0.2046, + "step": 30188 + }, + { + "epoch": 0.6645168390931103, + "grad_norm": 2.649984206928117, + "learning_rate": 2.672294128025113e-06, + "loss": 0.2989, + "step": 30189 + }, + { + "epoch": 0.6645388509795289, + "grad_norm": 1.7367040593513425, + "learning_rate": 2.6719786606976296e-06, + "loss": 0.3045, + "step": 30190 + }, + { + "epoch": 0.6645608628659476, + "grad_norm": 3.9801427312302464, + "learning_rate": 2.671663205202187e-06, + "loss": 0.2899, + "step": 30191 + }, + { + "epoch": 0.6645828747523663, + "grad_norm": 0.5739071806658731, + "learning_rate": 2.6713477615403906e-06, + "loss": 0.42, + "step": 30192 + }, + { + "epoch": 0.6646048866387849, + "grad_norm": 0.6003080019092002, + "learning_rate": 2.6710323297138436e-06, + "loss": 0.4511, + "step": 30193 + }, + { + "epoch": 0.6646268985252036, + "grad_norm": 6.381440810487485, + "learning_rate": 2.670716909724151e-06, + "loss": 0.3116, + "step": 30194 + }, + { + "epoch": 0.6646489104116223, + "grad_norm": 1.2964031227259019, + "learning_rate": 2.6704015015729133e-06, + "loss": 0.2485, + "step": 30195 + }, + { + "epoch": 0.6646709222980409, + "grad_norm": 1.375465756231398, + "learning_rate": 2.670086105261733e-06, + "loss": 0.2179, + "step": 30196 + }, + { + "epoch": 0.6646929341844596, + "grad_norm": 1.3989904694870208, + "learning_rate": 2.6697707207922163e-06, + "loss": 0.3038, + "step": 30197 + }, + { + "epoch": 0.6647149460708782, + "grad_norm": 2.203051141511565, + "learning_rate": 2.66945534816596e-06, + "loss": 0.3489, + "step": 30198 + }, + { + "epoch": 0.6647369579572969, + "grad_norm": 2.699606273576241, + "learning_rate": 2.669139987384575e-06, + "loss": 0.3929, + "step": 30199 + }, + { + "epoch": 0.6647589698437156, + "grad_norm": 1.4178178203941934, + "learning_rate": 2.6688246384496597e-06, + "loss": 0.2512, + "step": 30200 + }, + { + "epoch": 0.6647809817301342, + "grad_norm": 1.6021220160292977, + "learning_rate": 2.6685093013628157e-06, + "loss": 0.3172, + "step": 30201 + }, + { + "epoch": 0.6648029936165529, + "grad_norm": 1.811178258465693, + "learning_rate": 2.668193976125649e-06, + "loss": 0.3674, + "step": 30202 + }, + { + "epoch": 0.6648250055029716, + "grad_norm": 1.683293693897127, + "learning_rate": 2.667878662739757e-06, + "loss": 0.2993, + "step": 30203 + }, + { + "epoch": 0.6648470173893902, + "grad_norm": 2.493114986753057, + "learning_rate": 2.66756336120675e-06, + "loss": 0.3191, + "step": 30204 + }, + { + "epoch": 0.6648690292758089, + "grad_norm": 1.2493114311275142, + "learning_rate": 2.6672480715282255e-06, + "loss": 0.275, + "step": 30205 + }, + { + "epoch": 0.6648910411622276, + "grad_norm": 0.618346386173183, + "learning_rate": 2.6669327937057852e-06, + "loss": 0.4685, + "step": 30206 + }, + { + "epoch": 0.6649130530486462, + "grad_norm": 1.5990798234121242, + "learning_rate": 2.666617527741035e-06, + "loss": 0.338, + "step": 30207 + }, + { + "epoch": 0.6649350649350649, + "grad_norm": 1.4082358764334502, + "learning_rate": 2.666302273635573e-06, + "loss": 0.1826, + "step": 30208 + }, + { + "epoch": 0.6649570768214836, + "grad_norm": 1.3480765684250702, + "learning_rate": 2.665987031391004e-06, + "loss": 0.2721, + "step": 30209 + }, + { + "epoch": 0.6649790887079022, + "grad_norm": 1.6297711984321346, + "learning_rate": 2.6656718010089322e-06, + "loss": 0.3431, + "step": 30210 + }, + { + "epoch": 0.6650011005943209, + "grad_norm": 1.4515270726843064, + "learning_rate": 2.6653565824909545e-06, + "loss": 0.2749, + "step": 30211 + }, + { + "epoch": 0.6650231124807396, + "grad_norm": 1.5367433229695526, + "learning_rate": 2.6650413758386784e-06, + "loss": 0.3441, + "step": 30212 + }, + { + "epoch": 0.6650451243671582, + "grad_norm": 2.203443426168931, + "learning_rate": 2.6647261810537015e-06, + "loss": 0.249, + "step": 30213 + }, + { + "epoch": 0.6650671362535769, + "grad_norm": 1.422878676652904, + "learning_rate": 2.664410998137628e-06, + "loss": 0.269, + "step": 30214 + }, + { + "epoch": 0.6650891481399956, + "grad_norm": 0.6343211889503403, + "learning_rate": 2.6640958270920613e-06, + "loss": 0.4487, + "step": 30215 + }, + { + "epoch": 0.6651111600264142, + "grad_norm": 2.3422109696362265, + "learning_rate": 2.6637806679185997e-06, + "loss": 0.2447, + "step": 30216 + }, + { + "epoch": 0.6651331719128329, + "grad_norm": 0.6198146247633094, + "learning_rate": 2.6634655206188485e-06, + "loss": 0.4721, + "step": 30217 + }, + { + "epoch": 0.6651551837992516, + "grad_norm": 1.7081905486748419, + "learning_rate": 2.663150385194405e-06, + "loss": 0.2991, + "step": 30218 + }, + { + "epoch": 0.6651771956856702, + "grad_norm": 0.6209849567376584, + "learning_rate": 2.6628352616468745e-06, + "loss": 0.4459, + "step": 30219 + }, + { + "epoch": 0.665199207572089, + "grad_norm": 1.7005030396093472, + "learning_rate": 2.662520149977859e-06, + "loss": 0.2928, + "step": 30220 + }, + { + "epoch": 0.6652212194585075, + "grad_norm": 1.853500013273565, + "learning_rate": 2.6622050501889574e-06, + "loss": 0.3911, + "step": 30221 + }, + { + "epoch": 0.6652432313449262, + "grad_norm": 1.7910237724235112, + "learning_rate": 2.6618899622817732e-06, + "loss": 0.219, + "step": 30222 + }, + { + "epoch": 0.665265243231345, + "grad_norm": 1.50987304372931, + "learning_rate": 2.661574886257906e-06, + "loss": 0.3801, + "step": 30223 + }, + { + "epoch": 0.6652872551177635, + "grad_norm": 3.408649465272855, + "learning_rate": 2.6612598221189577e-06, + "loss": 0.271, + "step": 30224 + }, + { + "epoch": 0.6653092670041822, + "grad_norm": 3.1451591656123505, + "learning_rate": 2.6609447698665324e-06, + "loss": 0.2884, + "step": 30225 + }, + { + "epoch": 0.665331278890601, + "grad_norm": 1.5442891403601202, + "learning_rate": 2.660629729502229e-06, + "loss": 0.3657, + "step": 30226 + }, + { + "epoch": 0.6653532907770195, + "grad_norm": 1.4827674861353348, + "learning_rate": 2.6603147010276463e-06, + "loss": 0.3623, + "step": 30227 + }, + { + "epoch": 0.6653753026634383, + "grad_norm": 1.8126136840462685, + "learning_rate": 2.6599996844443887e-06, + "loss": 0.3271, + "step": 30228 + }, + { + "epoch": 0.665397314549857, + "grad_norm": 1.7568095071521967, + "learning_rate": 2.6596846797540556e-06, + "loss": 0.2663, + "step": 30229 + }, + { + "epoch": 0.6654193264362755, + "grad_norm": 1.665623556446413, + "learning_rate": 2.6593696869582508e-06, + "loss": 0.2745, + "step": 30230 + }, + { + "epoch": 0.6654413383226943, + "grad_norm": 0.6222446387647295, + "learning_rate": 2.6590547060585735e-06, + "loss": 0.4525, + "step": 30231 + }, + { + "epoch": 0.665463350209113, + "grad_norm": 2.707704751910389, + "learning_rate": 2.658739737056622e-06, + "loss": 0.3533, + "step": 30232 + }, + { + "epoch": 0.6654853620955316, + "grad_norm": 1.7188886617434955, + "learning_rate": 2.658424779954002e-06, + "loss": 0.304, + "step": 30233 + }, + { + "epoch": 0.6655073739819503, + "grad_norm": 2.855356409803658, + "learning_rate": 2.658109834752307e-06, + "loss": 0.3312, + "step": 30234 + }, + { + "epoch": 0.665529385868369, + "grad_norm": 1.3935914963452818, + "learning_rate": 2.6577949014531463e-06, + "loss": 0.3141, + "step": 30235 + }, + { + "epoch": 0.6655513977547876, + "grad_norm": 1.8089577053176937, + "learning_rate": 2.657479980058117e-06, + "loss": 0.2966, + "step": 30236 + }, + { + "epoch": 0.6655734096412063, + "grad_norm": 1.468884597472912, + "learning_rate": 2.657165070568817e-06, + "loss": 0.3263, + "step": 30237 + }, + { + "epoch": 0.665595421527625, + "grad_norm": 1.6701432132166738, + "learning_rate": 2.6568501729868514e-06, + "loss": 0.3513, + "step": 30238 + }, + { + "epoch": 0.6656174334140436, + "grad_norm": 1.5495709610173167, + "learning_rate": 2.6565352873138135e-06, + "loss": 0.2923, + "step": 30239 + }, + { + "epoch": 0.6656394453004623, + "grad_norm": 3.591335394900924, + "learning_rate": 2.6562204135513125e-06, + "loss": 0.3495, + "step": 30240 + }, + { + "epoch": 0.665661457186881, + "grad_norm": 2.176857464654049, + "learning_rate": 2.6559055517009458e-06, + "loss": 0.2994, + "step": 30241 + }, + { + "epoch": 0.6656834690732996, + "grad_norm": 1.8330353661746097, + "learning_rate": 2.6555907017643097e-06, + "loss": 0.303, + "step": 30242 + }, + { + "epoch": 0.6657054809597183, + "grad_norm": 1.2168358797567316, + "learning_rate": 2.6552758637430097e-06, + "loss": 0.267, + "step": 30243 + }, + { + "epoch": 0.6657274928461369, + "grad_norm": 1.282117695906022, + "learning_rate": 2.6549610376386412e-06, + "loss": 0.1756, + "step": 30244 + }, + { + "epoch": 0.6657495047325556, + "grad_norm": 1.258486896877322, + "learning_rate": 2.654646223452807e-06, + "loss": 0.3202, + "step": 30245 + }, + { + "epoch": 0.6657715166189743, + "grad_norm": 1.20219552761199, + "learning_rate": 2.6543314211871084e-06, + "loss": 0.2732, + "step": 30246 + }, + { + "epoch": 0.6657935285053929, + "grad_norm": 1.7177190323808056, + "learning_rate": 2.6540166308431416e-06, + "loss": 0.2937, + "step": 30247 + }, + { + "epoch": 0.6658155403918116, + "grad_norm": 1.5527216675067808, + "learning_rate": 2.653701852422511e-06, + "loss": 0.2881, + "step": 30248 + }, + { + "epoch": 0.6658375522782303, + "grad_norm": 1.4256251893872078, + "learning_rate": 2.653387085926812e-06, + "loss": 0.2681, + "step": 30249 + }, + { + "epoch": 0.6658595641646489, + "grad_norm": 1.1827062078307402, + "learning_rate": 2.6530723313576463e-06, + "loss": 0.226, + "step": 30250 + }, + { + "epoch": 0.6658815760510676, + "grad_norm": 1.96701280733867, + "learning_rate": 2.652757588716616e-06, + "loss": 0.354, + "step": 30251 + }, + { + "epoch": 0.6659035879374863, + "grad_norm": 1.5601456219839522, + "learning_rate": 2.6524428580053164e-06, + "loss": 0.276, + "step": 30252 + }, + { + "epoch": 0.6659255998239049, + "grad_norm": 1.5813678068516026, + "learning_rate": 2.6521281392253507e-06, + "loss": 0.2721, + "step": 30253 + }, + { + "epoch": 0.6659476117103236, + "grad_norm": 1.6278325374606333, + "learning_rate": 2.6518134323783156e-06, + "loss": 0.2525, + "step": 30254 + }, + { + "epoch": 0.6659696235967423, + "grad_norm": 1.7723030080967137, + "learning_rate": 2.6514987374658117e-06, + "loss": 0.3332, + "step": 30255 + }, + { + "epoch": 0.6659916354831609, + "grad_norm": 1.5800775625699768, + "learning_rate": 2.6511840544894407e-06, + "loss": 0.3028, + "step": 30256 + }, + { + "epoch": 0.6660136473695796, + "grad_norm": 1.7687935430190307, + "learning_rate": 2.650869383450798e-06, + "loss": 0.3209, + "step": 30257 + }, + { + "epoch": 0.6660356592559983, + "grad_norm": 1.6774295803086483, + "learning_rate": 2.6505547243514875e-06, + "loss": 0.3112, + "step": 30258 + }, + { + "epoch": 0.6660576711424169, + "grad_norm": 1.539656228759875, + "learning_rate": 2.6502400771931032e-06, + "loss": 0.2126, + "step": 30259 + }, + { + "epoch": 0.6660796830288356, + "grad_norm": 1.7590467310274052, + "learning_rate": 2.649925441977247e-06, + "loss": 0.3078, + "step": 30260 + }, + { + "epoch": 0.6661016949152543, + "grad_norm": 1.8861260489203961, + "learning_rate": 2.6496108187055192e-06, + "loss": 0.3184, + "step": 30261 + }, + { + "epoch": 0.6661237068016729, + "grad_norm": 1.4990438022953452, + "learning_rate": 2.6492962073795158e-06, + "loss": 0.2582, + "step": 30262 + }, + { + "epoch": 0.6661457186880916, + "grad_norm": 1.9102448952157387, + "learning_rate": 2.64898160800084e-06, + "loss": 0.3358, + "step": 30263 + }, + { + "epoch": 0.6661677305745103, + "grad_norm": 0.656176084718084, + "learning_rate": 2.648667020571086e-06, + "loss": 0.4747, + "step": 30264 + }, + { + "epoch": 0.6661897424609289, + "grad_norm": 2.158117887107487, + "learning_rate": 2.6483524450918553e-06, + "loss": 0.2482, + "step": 30265 + }, + { + "epoch": 0.6662117543473476, + "grad_norm": 1.7641514510236513, + "learning_rate": 2.6480378815647478e-06, + "loss": 0.3125, + "step": 30266 + }, + { + "epoch": 0.6662337662337663, + "grad_norm": 1.686971790639887, + "learning_rate": 2.647723329991361e-06, + "loss": 0.3518, + "step": 30267 + }, + { + "epoch": 0.6662557781201849, + "grad_norm": 1.7563237120588508, + "learning_rate": 2.647408790373291e-06, + "loss": 0.2353, + "step": 30268 + }, + { + "epoch": 0.6662777900066036, + "grad_norm": 1.3318329131570317, + "learning_rate": 2.647094262712141e-06, + "loss": 0.232, + "step": 30269 + }, + { + "epoch": 0.6662998018930222, + "grad_norm": 1.7214486520929868, + "learning_rate": 2.6467797470095034e-06, + "loss": 0.2795, + "step": 30270 + }, + { + "epoch": 0.6663218137794409, + "grad_norm": 1.2032210267231154, + "learning_rate": 2.6464652432669846e-06, + "loss": 0.225, + "step": 30271 + }, + { + "epoch": 0.6663438256658596, + "grad_norm": 1.5698738018204883, + "learning_rate": 2.6461507514861795e-06, + "loss": 0.3056, + "step": 30272 + }, + { + "epoch": 0.6663658375522782, + "grad_norm": 0.6212281324189944, + "learning_rate": 2.6458362716686843e-06, + "loss": 0.4631, + "step": 30273 + }, + { + "epoch": 0.6663878494386969, + "grad_norm": 1.7540791897515027, + "learning_rate": 2.6455218038161005e-06, + "loss": 0.3259, + "step": 30274 + }, + { + "epoch": 0.6664098613251156, + "grad_norm": 1.2649638459160275, + "learning_rate": 2.6452073479300233e-06, + "loss": 0.2216, + "step": 30275 + }, + { + "epoch": 0.6664318732115342, + "grad_norm": 1.741424949670108, + "learning_rate": 2.644892904012053e-06, + "loss": 0.2855, + "step": 30276 + }, + { + "epoch": 0.6664538850979529, + "grad_norm": 3.8865671069917282, + "learning_rate": 2.644578472063789e-06, + "loss": 0.3417, + "step": 30277 + }, + { + "epoch": 0.6664758969843716, + "grad_norm": 1.7471159430000898, + "learning_rate": 2.644264052086826e-06, + "loss": 0.3143, + "step": 30278 + }, + { + "epoch": 0.6664979088707902, + "grad_norm": 4.968439995478528, + "learning_rate": 2.643949644082765e-06, + "loss": 0.3142, + "step": 30279 + }, + { + "epoch": 0.6665199207572089, + "grad_norm": 1.7022134087155703, + "learning_rate": 2.6436352480532012e-06, + "loss": 0.2845, + "step": 30280 + }, + { + "epoch": 0.6665419326436276, + "grad_norm": 0.6057263460204245, + "learning_rate": 2.6433208639997345e-06, + "loss": 0.4443, + "step": 30281 + }, + { + "epoch": 0.6665639445300462, + "grad_norm": 1.5047739566165335, + "learning_rate": 2.643006491923964e-06, + "loss": 0.3267, + "step": 30282 + }, + { + "epoch": 0.6665859564164649, + "grad_norm": 2.167847193132805, + "learning_rate": 2.6426921318274835e-06, + "loss": 0.2896, + "step": 30283 + }, + { + "epoch": 0.6666079683028836, + "grad_norm": 0.6311384038382056, + "learning_rate": 2.642377783711895e-06, + "loss": 0.4609, + "step": 30284 + }, + { + "epoch": 0.6666299801893022, + "grad_norm": 7.578115019124153, + "learning_rate": 2.642063447578792e-06, + "loss": 0.3005, + "step": 30285 + }, + { + "epoch": 0.6666519920757209, + "grad_norm": 1.5223899113958872, + "learning_rate": 2.641749123429775e-06, + "loss": 0.2843, + "step": 30286 + }, + { + "epoch": 0.6666740039621396, + "grad_norm": 6.308490091639926, + "learning_rate": 2.6414348112664423e-06, + "loss": 0.3063, + "step": 30287 + }, + { + "epoch": 0.6666960158485582, + "grad_norm": 2.41134798103749, + "learning_rate": 2.6411205110903877e-06, + "loss": 0.2457, + "step": 30288 + }, + { + "epoch": 0.6667180277349769, + "grad_norm": 1.3316588169708592, + "learning_rate": 2.640806222903213e-06, + "loss": 0.313, + "step": 30289 + }, + { + "epoch": 0.6667400396213956, + "grad_norm": 1.282465192017502, + "learning_rate": 2.6404919467065108e-06, + "loss": 0.1981, + "step": 30290 + }, + { + "epoch": 0.6667620515078142, + "grad_norm": 1.2530905334294387, + "learning_rate": 2.6401776825018817e-06, + "loss": 0.2914, + "step": 30291 + }, + { + "epoch": 0.6667840633942329, + "grad_norm": 1.4193463503886021, + "learning_rate": 2.6398634302909232e-06, + "loss": 0.304, + "step": 30292 + }, + { + "epoch": 0.6668060752806515, + "grad_norm": 2.85875910247895, + "learning_rate": 2.6395491900752303e-06, + "loss": 0.31, + "step": 30293 + }, + { + "epoch": 0.6668280871670702, + "grad_norm": 2.884198026860784, + "learning_rate": 2.639234961856403e-06, + "loss": 0.2989, + "step": 30294 + }, + { + "epoch": 0.6668500990534889, + "grad_norm": 1.6863291681997064, + "learning_rate": 2.6389207456360337e-06, + "loss": 0.2773, + "step": 30295 + }, + { + "epoch": 0.6668721109399075, + "grad_norm": 2.901084750487009, + "learning_rate": 2.638606541415723e-06, + "loss": 0.2869, + "step": 30296 + }, + { + "epoch": 0.6668941228263262, + "grad_norm": 1.5910960045234406, + "learning_rate": 2.6382923491970693e-06, + "loss": 0.2849, + "step": 30297 + }, + { + "epoch": 0.6669161347127449, + "grad_norm": 1.4561091877514636, + "learning_rate": 2.6379781689816646e-06, + "loss": 0.3984, + "step": 30298 + }, + { + "epoch": 0.6669381465991635, + "grad_norm": 4.677129746701346, + "learning_rate": 2.637664000771111e-06, + "loss": 0.2876, + "step": 30299 + }, + { + "epoch": 0.6669601584855822, + "grad_norm": 1.8512410997616293, + "learning_rate": 2.6373498445669997e-06, + "loss": 0.2915, + "step": 30300 + }, + { + "epoch": 0.6669821703720009, + "grad_norm": 1.2345407450052057, + "learning_rate": 2.6370357003709312e-06, + "loss": 0.308, + "step": 30301 + }, + { + "epoch": 0.6670041822584195, + "grad_norm": 3.224198028604474, + "learning_rate": 2.636721568184503e-06, + "loss": 0.3425, + "step": 30302 + }, + { + "epoch": 0.6670261941448382, + "grad_norm": 1.8224292306068481, + "learning_rate": 2.636407448009307e-06, + "loss": 0.2982, + "step": 30303 + }, + { + "epoch": 0.6670482060312569, + "grad_norm": 1.1564874066592965, + "learning_rate": 2.636093339846945e-06, + "loss": 0.2169, + "step": 30304 + }, + { + "epoch": 0.6670702179176755, + "grad_norm": 1.7892021168303176, + "learning_rate": 2.6357792436990114e-06, + "loss": 0.3611, + "step": 30305 + }, + { + "epoch": 0.6670922298040942, + "grad_norm": 1.383458503549555, + "learning_rate": 2.635465159567098e-06, + "loss": 0.2883, + "step": 30306 + }, + { + "epoch": 0.6671142416905129, + "grad_norm": 1.3190849157834916, + "learning_rate": 2.6351510874528085e-06, + "loss": 0.2695, + "step": 30307 + }, + { + "epoch": 0.6671362535769315, + "grad_norm": 1.4048051192716955, + "learning_rate": 2.634837027357737e-06, + "loss": 0.3517, + "step": 30308 + }, + { + "epoch": 0.6671582654633502, + "grad_norm": 2.366132094627241, + "learning_rate": 2.634522979283476e-06, + "loss": 0.2969, + "step": 30309 + }, + { + "epoch": 0.6671802773497689, + "grad_norm": 1.966687634486448, + "learning_rate": 2.6342089432316263e-06, + "loss": 0.3899, + "step": 30310 + }, + { + "epoch": 0.6672022892361875, + "grad_norm": 0.6295278821784405, + "learning_rate": 2.63389491920378e-06, + "loss": 0.4698, + "step": 30311 + }, + { + "epoch": 0.6672243011226062, + "grad_norm": 1.722152343676899, + "learning_rate": 2.633580907201535e-06, + "loss": 0.3219, + "step": 30312 + }, + { + "epoch": 0.6672463130090249, + "grad_norm": 1.8344608840829173, + "learning_rate": 2.633266907226489e-06, + "loss": 0.3381, + "step": 30313 + }, + { + "epoch": 0.6672683248954435, + "grad_norm": 9.631162940827444, + "learning_rate": 2.632952919280235e-06, + "loss": 0.2591, + "step": 30314 + }, + { + "epoch": 0.6672903367818622, + "grad_norm": 1.809578115044142, + "learning_rate": 2.632638943364372e-06, + "loss": 0.2497, + "step": 30315 + }, + { + "epoch": 0.6673123486682808, + "grad_norm": 1.7011569703654374, + "learning_rate": 2.632324979480491e-06, + "loss": 0.4008, + "step": 30316 + }, + { + "epoch": 0.6673343605546995, + "grad_norm": 0.6237995176851461, + "learning_rate": 2.6320110276301906e-06, + "loss": 0.4797, + "step": 30317 + }, + { + "epoch": 0.6673563724411182, + "grad_norm": 2.277587057556008, + "learning_rate": 2.6316970878150687e-06, + "loss": 0.2466, + "step": 30318 + }, + { + "epoch": 0.6673783843275368, + "grad_norm": 1.530339954758584, + "learning_rate": 2.631383160036716e-06, + "loss": 0.2724, + "step": 30319 + }, + { + "epoch": 0.6674003962139555, + "grad_norm": 1.9015820829871566, + "learning_rate": 2.6310692442967333e-06, + "loss": 0.3535, + "step": 30320 + }, + { + "epoch": 0.6674224081003742, + "grad_norm": 2.0541997689040143, + "learning_rate": 2.6307553405967102e-06, + "loss": 0.422, + "step": 30321 + }, + { + "epoch": 0.6674444199867928, + "grad_norm": 2.0367908884626242, + "learning_rate": 2.6304414489382455e-06, + "loss": 0.3376, + "step": 30322 + }, + { + "epoch": 0.6674664318732115, + "grad_norm": 1.955996386741932, + "learning_rate": 2.630127569322937e-06, + "loss": 0.2865, + "step": 30323 + }, + { + "epoch": 0.6674884437596302, + "grad_norm": 0.5904257764982306, + "learning_rate": 2.6298137017523743e-06, + "loss": 0.4636, + "step": 30324 + }, + { + "epoch": 0.6675104556460488, + "grad_norm": 1.674233210349888, + "learning_rate": 2.6294998462281585e-06, + "loss": 0.3463, + "step": 30325 + }, + { + "epoch": 0.6675324675324675, + "grad_norm": 1.5264365158241813, + "learning_rate": 2.629186002751879e-06, + "loss": 0.2704, + "step": 30326 + }, + { + "epoch": 0.6675544794188862, + "grad_norm": 1.4827155734967399, + "learning_rate": 2.628872171325133e-06, + "loss": 0.2414, + "step": 30327 + }, + { + "epoch": 0.6675764913053048, + "grad_norm": 1.5155157523774132, + "learning_rate": 2.6285583519495196e-06, + "loss": 0.3382, + "step": 30328 + }, + { + "epoch": 0.6675985031917235, + "grad_norm": 0.6248187800560872, + "learning_rate": 2.628244544626627e-06, + "loss": 0.4655, + "step": 30329 + }, + { + "epoch": 0.6676205150781422, + "grad_norm": 1.4307881202783523, + "learning_rate": 2.6279307493580564e-06, + "loss": 0.2496, + "step": 30330 + }, + { + "epoch": 0.6676425269645608, + "grad_norm": 1.7713142620902589, + "learning_rate": 2.627616966145398e-06, + "loss": 0.2618, + "step": 30331 + }, + { + "epoch": 0.6676645388509795, + "grad_norm": 0.5926286000834429, + "learning_rate": 2.6273031949902473e-06, + "loss": 0.443, + "step": 30332 + }, + { + "epoch": 0.6676865507373982, + "grad_norm": 1.3928045444785864, + "learning_rate": 2.626989435894202e-06, + "loss": 0.3142, + "step": 30333 + }, + { + "epoch": 0.6677085626238168, + "grad_norm": 1.6150453015855841, + "learning_rate": 2.6266756888588534e-06, + "loss": 0.2256, + "step": 30334 + }, + { + "epoch": 0.6677305745102355, + "grad_norm": 1.5962769966311836, + "learning_rate": 2.6263619538857994e-06, + "loss": 0.2763, + "step": 30335 + }, + { + "epoch": 0.6677525863966542, + "grad_norm": 2.200471755086744, + "learning_rate": 2.62604823097663e-06, + "loss": 0.3195, + "step": 30336 + }, + { + "epoch": 0.6677745982830728, + "grad_norm": 2.2854707588699865, + "learning_rate": 2.625734520132943e-06, + "loss": 0.2582, + "step": 30337 + }, + { + "epoch": 0.6677966101694915, + "grad_norm": 0.6138784563030759, + "learning_rate": 2.6254208213563336e-06, + "loss": 0.4582, + "step": 30338 + }, + { + "epoch": 0.6678186220559101, + "grad_norm": 2.703578897268161, + "learning_rate": 2.6251071346483935e-06, + "loss": 0.3355, + "step": 30339 + }, + { + "epoch": 0.6678406339423288, + "grad_norm": 1.691331768413392, + "learning_rate": 2.6247934600107195e-06, + "loss": 0.329, + "step": 30340 + }, + { + "epoch": 0.6678626458287475, + "grad_norm": 0.6168853393688077, + "learning_rate": 2.624479797444905e-06, + "loss": 0.4553, + "step": 30341 + }, + { + "epoch": 0.6678846577151661, + "grad_norm": 1.3762827485244618, + "learning_rate": 2.6241661469525394e-06, + "loss": 0.3696, + "step": 30342 + }, + { + "epoch": 0.6679066696015848, + "grad_norm": 0.6221082776936009, + "learning_rate": 2.623852508535225e-06, + "loss": 0.4332, + "step": 30343 + }, + { + "epoch": 0.6679286814880036, + "grad_norm": 0.6179428571336303, + "learning_rate": 2.6235388821945497e-06, + "loss": 0.4446, + "step": 30344 + }, + { + "epoch": 0.6679506933744221, + "grad_norm": 1.4159607107811407, + "learning_rate": 2.6232252679321123e-06, + "loss": 0.314, + "step": 30345 + }, + { + "epoch": 0.6679727052608408, + "grad_norm": 0.6110997693278513, + "learning_rate": 2.6229116657495036e-06, + "loss": 0.4619, + "step": 30346 + }, + { + "epoch": 0.6679947171472596, + "grad_norm": 1.5482700796104303, + "learning_rate": 2.622598075648316e-06, + "loss": 0.2279, + "step": 30347 + }, + { + "epoch": 0.6680167290336781, + "grad_norm": 1.7576551106700293, + "learning_rate": 2.6222844976301458e-06, + "loss": 0.2678, + "step": 30348 + }, + { + "epoch": 0.6680387409200969, + "grad_norm": 2.002215824805247, + "learning_rate": 2.621970931696588e-06, + "loss": 0.2654, + "step": 30349 + }, + { + "epoch": 0.6680607528065156, + "grad_norm": 1.8889286921155657, + "learning_rate": 2.621657377849233e-06, + "loss": 0.291, + "step": 30350 + }, + { + "epoch": 0.6680827646929342, + "grad_norm": 0.6004957280100658, + "learning_rate": 2.6213438360896775e-06, + "loss": 0.4326, + "step": 30351 + }, + { + "epoch": 0.6681047765793529, + "grad_norm": 1.8771669458916107, + "learning_rate": 2.621030306419512e-06, + "loss": 0.3457, + "step": 30352 + }, + { + "epoch": 0.6681267884657716, + "grad_norm": 0.6188674727992595, + "learning_rate": 2.6207167888403313e-06, + "loss": 0.4627, + "step": 30353 + }, + { + "epoch": 0.6681488003521902, + "grad_norm": 1.4766568687548471, + "learning_rate": 2.620403283353731e-06, + "loss": 0.3092, + "step": 30354 + }, + { + "epoch": 0.6681708122386089, + "grad_norm": 1.5411346372729209, + "learning_rate": 2.6200897899613005e-06, + "loss": 0.3726, + "step": 30355 + }, + { + "epoch": 0.6681928241250276, + "grad_norm": 1.3851156165981118, + "learning_rate": 2.6197763086646366e-06, + "loss": 0.2408, + "step": 30356 + }, + { + "epoch": 0.6682148360114462, + "grad_norm": 0.6155375213422815, + "learning_rate": 2.61946283946533e-06, + "loss": 0.4452, + "step": 30357 + }, + { + "epoch": 0.6682368478978649, + "grad_norm": 1.5613711359043754, + "learning_rate": 2.619149382364975e-06, + "loss": 0.2619, + "step": 30358 + }, + { + "epoch": 0.6682588597842836, + "grad_norm": 2.5514396504081507, + "learning_rate": 2.618835937365166e-06, + "loss": 0.2908, + "step": 30359 + }, + { + "epoch": 0.6682808716707022, + "grad_norm": 1.3877158377804133, + "learning_rate": 2.6185225044674935e-06, + "loss": 0.3048, + "step": 30360 + }, + { + "epoch": 0.6683028835571209, + "grad_norm": 3.0828250600154856, + "learning_rate": 2.618209083673553e-06, + "loss": 0.2441, + "step": 30361 + }, + { + "epoch": 0.6683248954435395, + "grad_norm": 1.869118219223107, + "learning_rate": 2.6178956749849347e-06, + "loss": 0.4022, + "step": 30362 + }, + { + "epoch": 0.6683469073299582, + "grad_norm": 1.2111120610015724, + "learning_rate": 2.6175822784032334e-06, + "loss": 0.2626, + "step": 30363 + }, + { + "epoch": 0.6683689192163769, + "grad_norm": 1.4371028330632654, + "learning_rate": 2.6172688939300435e-06, + "loss": 0.2492, + "step": 30364 + }, + { + "epoch": 0.6683909311027955, + "grad_norm": 1.5834624556287, + "learning_rate": 2.6169555215669535e-06, + "loss": 0.2666, + "step": 30365 + }, + { + "epoch": 0.6684129429892142, + "grad_norm": 5.091682710011325, + "learning_rate": 2.61664216131556e-06, + "loss": 0.2912, + "step": 30366 + }, + { + "epoch": 0.6684349548756329, + "grad_norm": 1.5742563881275469, + "learning_rate": 2.6163288131774526e-06, + "loss": 0.3224, + "step": 30367 + }, + { + "epoch": 0.6684569667620515, + "grad_norm": 0.6251476399379661, + "learning_rate": 2.616015477154226e-06, + "loss": 0.4452, + "step": 30368 + }, + { + "epoch": 0.6684789786484702, + "grad_norm": 2.5231645459144096, + "learning_rate": 2.6157021532474724e-06, + "loss": 0.2828, + "step": 30369 + }, + { + "epoch": 0.6685009905348889, + "grad_norm": 0.6078209113119328, + "learning_rate": 2.615388841458783e-06, + "loss": 0.4621, + "step": 30370 + }, + { + "epoch": 0.6685230024213075, + "grad_norm": 1.5125412204527464, + "learning_rate": 2.6150755417897526e-06, + "loss": 0.3706, + "step": 30371 + }, + { + "epoch": 0.6685450143077262, + "grad_norm": 1.7743237208452163, + "learning_rate": 2.6147622542419728e-06, + "loss": 0.4014, + "step": 30372 + }, + { + "epoch": 0.6685670261941449, + "grad_norm": 1.6424364193070207, + "learning_rate": 2.61444897881703e-06, + "loss": 0.3935, + "step": 30373 + }, + { + "epoch": 0.6685890380805635, + "grad_norm": 3.9103352527574313, + "learning_rate": 2.614135715516526e-06, + "loss": 0.2272, + "step": 30374 + }, + { + "epoch": 0.6686110499669822, + "grad_norm": 3.392535800506232, + "learning_rate": 2.6138224643420463e-06, + "loss": 0.3303, + "step": 30375 + }, + { + "epoch": 0.6686330618534009, + "grad_norm": 1.4965352844833482, + "learning_rate": 2.6135092252951875e-06, + "loss": 0.3253, + "step": 30376 + }, + { + "epoch": 0.6686550737398195, + "grad_norm": 1.305078348362841, + "learning_rate": 2.613195998377539e-06, + "loss": 0.2909, + "step": 30377 + }, + { + "epoch": 0.6686770856262382, + "grad_norm": 1.508282339073455, + "learning_rate": 2.612882783590689e-06, + "loss": 0.2802, + "step": 30378 + }, + { + "epoch": 0.6686990975126569, + "grad_norm": 1.9579848580108097, + "learning_rate": 2.6125695809362365e-06, + "loss": 0.2385, + "step": 30379 + }, + { + "epoch": 0.6687211093990755, + "grad_norm": 1.3872259162581082, + "learning_rate": 2.6122563904157695e-06, + "loss": 0.239, + "step": 30380 + }, + { + "epoch": 0.6687431212854942, + "grad_norm": 2.5124993136429903, + "learning_rate": 2.611943212030882e-06, + "loss": 0.3094, + "step": 30381 + }, + { + "epoch": 0.6687651331719129, + "grad_norm": 1.902285563874389, + "learning_rate": 2.611630045783164e-06, + "loss": 0.2811, + "step": 30382 + }, + { + "epoch": 0.6687871450583315, + "grad_norm": 1.4043846878620787, + "learning_rate": 2.6113168916742055e-06, + "loss": 0.2498, + "step": 30383 + }, + { + "epoch": 0.6688091569447502, + "grad_norm": 1.3703658260238234, + "learning_rate": 2.6110037497055996e-06, + "loss": 0.275, + "step": 30384 + }, + { + "epoch": 0.6688311688311688, + "grad_norm": 1.794083328482369, + "learning_rate": 2.6106906198789405e-06, + "loss": 0.2474, + "step": 30385 + }, + { + "epoch": 0.6688531807175875, + "grad_norm": 1.3682875410510345, + "learning_rate": 2.6103775021958153e-06, + "loss": 0.3121, + "step": 30386 + }, + { + "epoch": 0.6688751926040062, + "grad_norm": 1.6538152346936381, + "learning_rate": 2.61006439665782e-06, + "loss": 0.3632, + "step": 30387 + }, + { + "epoch": 0.6688972044904248, + "grad_norm": 1.3956083346095491, + "learning_rate": 2.6097513032665406e-06, + "loss": 0.2564, + "step": 30388 + }, + { + "epoch": 0.6689192163768435, + "grad_norm": 1.617727376660313, + "learning_rate": 2.6094382220235716e-06, + "loss": 0.2765, + "step": 30389 + }, + { + "epoch": 0.6689412282632622, + "grad_norm": 1.5760747546101515, + "learning_rate": 2.609125152930505e-06, + "loss": 0.3252, + "step": 30390 + }, + { + "epoch": 0.6689632401496808, + "grad_norm": 2.344464437803826, + "learning_rate": 2.6088120959889294e-06, + "loss": 0.328, + "step": 30391 + }, + { + "epoch": 0.6689852520360995, + "grad_norm": 2.0515046837540445, + "learning_rate": 2.6084990512004395e-06, + "loss": 0.3877, + "step": 30392 + }, + { + "epoch": 0.6690072639225182, + "grad_norm": 1.8377889412282864, + "learning_rate": 2.608186018566622e-06, + "loss": 0.3427, + "step": 30393 + }, + { + "epoch": 0.6690292758089368, + "grad_norm": 1.3659675168638445, + "learning_rate": 2.6078729980890695e-06, + "loss": 0.2218, + "step": 30394 + }, + { + "epoch": 0.6690512876953555, + "grad_norm": 2.423645011171816, + "learning_rate": 2.6075599897693747e-06, + "loss": 0.2678, + "step": 30395 + }, + { + "epoch": 0.6690732995817742, + "grad_norm": 1.2686455780095358, + "learning_rate": 2.607246993609126e-06, + "loss": 0.2539, + "step": 30396 + }, + { + "epoch": 0.6690953114681928, + "grad_norm": 1.4123007006961268, + "learning_rate": 2.606934009609917e-06, + "loss": 0.3185, + "step": 30397 + }, + { + "epoch": 0.6691173233546115, + "grad_norm": 1.6540159960167902, + "learning_rate": 2.6066210377733353e-06, + "loss": 0.2161, + "step": 30398 + }, + { + "epoch": 0.6691393352410302, + "grad_norm": 1.3013735835848617, + "learning_rate": 2.6063080781009725e-06, + "loss": 0.2412, + "step": 30399 + }, + { + "epoch": 0.6691613471274488, + "grad_norm": 1.5852617413805095, + "learning_rate": 2.605995130594422e-06, + "loss": 0.3614, + "step": 30400 + }, + { + "epoch": 0.6691833590138675, + "grad_norm": 0.6409467745173657, + "learning_rate": 2.6056821952552693e-06, + "loss": 0.4753, + "step": 30401 + }, + { + "epoch": 0.6692053709002862, + "grad_norm": 1.2268760027974883, + "learning_rate": 2.60536927208511e-06, + "loss": 0.3008, + "step": 30402 + }, + { + "epoch": 0.6692273827867048, + "grad_norm": 2.0446472909694733, + "learning_rate": 2.6050563610855303e-06, + "loss": 0.2535, + "step": 30403 + }, + { + "epoch": 0.6692493946731235, + "grad_norm": 1.580634612897566, + "learning_rate": 2.604743462258122e-06, + "loss": 0.3254, + "step": 30404 + }, + { + "epoch": 0.6692714065595422, + "grad_norm": 2.25324376812376, + "learning_rate": 2.6044305756044775e-06, + "loss": 0.3569, + "step": 30405 + }, + { + "epoch": 0.6692934184459608, + "grad_norm": 1.2417260372812382, + "learning_rate": 2.604117701126184e-06, + "loss": 0.2467, + "step": 30406 + }, + { + "epoch": 0.6693154303323795, + "grad_norm": 1.831002321511807, + "learning_rate": 2.6038048388248343e-06, + "loss": 0.3339, + "step": 30407 + }, + { + "epoch": 0.6693374422187982, + "grad_norm": 1.3603528939596676, + "learning_rate": 2.603491988702017e-06, + "loss": 0.1906, + "step": 30408 + }, + { + "epoch": 0.6693594541052168, + "grad_norm": 1.6089289034707868, + "learning_rate": 2.6031791507593196e-06, + "loss": 0.2133, + "step": 30409 + }, + { + "epoch": 0.6693814659916355, + "grad_norm": 1.9365202962592287, + "learning_rate": 2.6028663249983375e-06, + "loss": 0.2802, + "step": 30410 + }, + { + "epoch": 0.6694034778780541, + "grad_norm": 1.4634311298688676, + "learning_rate": 2.6025535114206564e-06, + "loss": 0.3089, + "step": 30411 + }, + { + "epoch": 0.6694254897644728, + "grad_norm": 1.7630330206125533, + "learning_rate": 2.602240710027869e-06, + "loss": 0.3113, + "step": 30412 + }, + { + "epoch": 0.6694475016508915, + "grad_norm": 1.2309052546372719, + "learning_rate": 2.601927920821564e-06, + "loss": 0.3013, + "step": 30413 + }, + { + "epoch": 0.6694695135373101, + "grad_norm": 0.6515860086538642, + "learning_rate": 2.6016151438033275e-06, + "loss": 0.4761, + "step": 30414 + }, + { + "epoch": 0.6694915254237288, + "grad_norm": 1.3667445380150745, + "learning_rate": 2.601302378974756e-06, + "loss": 0.3332, + "step": 30415 + }, + { + "epoch": 0.6695135373101475, + "grad_norm": 1.4349672462720315, + "learning_rate": 2.600989626337434e-06, + "loss": 0.2577, + "step": 30416 + }, + { + "epoch": 0.6695355491965661, + "grad_norm": 1.6073718260284975, + "learning_rate": 2.6006768858929543e-06, + "loss": 0.2925, + "step": 30417 + }, + { + "epoch": 0.6695575610829848, + "grad_norm": 1.432906968475634, + "learning_rate": 2.6003641576429052e-06, + "loss": 0.2944, + "step": 30418 + }, + { + "epoch": 0.6695795729694035, + "grad_norm": 1.5238952794055634, + "learning_rate": 2.6000514415888732e-06, + "loss": 0.2452, + "step": 30419 + }, + { + "epoch": 0.6696015848558221, + "grad_norm": 1.565545098819263, + "learning_rate": 2.5997387377324507e-06, + "loss": 0.3303, + "step": 30420 + }, + { + "epoch": 0.6696235967422408, + "grad_norm": 1.5794916150806106, + "learning_rate": 2.599426046075226e-06, + "loss": 0.2893, + "step": 30421 + }, + { + "epoch": 0.6696456086286595, + "grad_norm": 1.2934985834295252, + "learning_rate": 2.5991133666187917e-06, + "loss": 0.3455, + "step": 30422 + }, + { + "epoch": 0.6696676205150781, + "grad_norm": 1.643171289581546, + "learning_rate": 2.598800699364733e-06, + "loss": 0.2798, + "step": 30423 + }, + { + "epoch": 0.6696896324014968, + "grad_norm": 1.6659640088354244, + "learning_rate": 2.598488044314639e-06, + "loss": 0.277, + "step": 30424 + }, + { + "epoch": 0.6697116442879155, + "grad_norm": 1.4086897560426377, + "learning_rate": 2.5981754014700995e-06, + "loss": 0.3083, + "step": 30425 + }, + { + "epoch": 0.6697336561743341, + "grad_norm": 1.5805800167557111, + "learning_rate": 2.5978627708327054e-06, + "loss": 0.3064, + "step": 30426 + }, + { + "epoch": 0.6697556680607528, + "grad_norm": 0.6235430036056717, + "learning_rate": 2.5975501524040427e-06, + "loss": 0.4644, + "step": 30427 + }, + { + "epoch": 0.6697776799471715, + "grad_norm": 1.6920912348382338, + "learning_rate": 2.5972375461857034e-06, + "loss": 0.2982, + "step": 30428 + }, + { + "epoch": 0.6697996918335901, + "grad_norm": 1.3642358839659046, + "learning_rate": 2.5969249521792728e-06, + "loss": 0.2461, + "step": 30429 + }, + { + "epoch": 0.6698217037200088, + "grad_norm": 1.5260331714851065, + "learning_rate": 2.596612370386341e-06, + "loss": 0.3624, + "step": 30430 + }, + { + "epoch": 0.6698437156064275, + "grad_norm": 1.272881324195715, + "learning_rate": 2.5962998008084995e-06, + "loss": 0.3165, + "step": 30431 + }, + { + "epoch": 0.6698657274928461, + "grad_norm": 1.5397127157795623, + "learning_rate": 2.5959872434473323e-06, + "loss": 0.3307, + "step": 30432 + }, + { + "epoch": 0.6698877393792648, + "grad_norm": 1.5420472790225612, + "learning_rate": 2.5956746983044324e-06, + "loss": 0.3521, + "step": 30433 + }, + { + "epoch": 0.6699097512656834, + "grad_norm": 1.438983127173223, + "learning_rate": 2.5953621653813834e-06, + "loss": 0.3233, + "step": 30434 + }, + { + "epoch": 0.6699317631521021, + "grad_norm": 0.6135542895789737, + "learning_rate": 2.5950496446797768e-06, + "loss": 0.4635, + "step": 30435 + }, + { + "epoch": 0.6699537750385208, + "grad_norm": 1.2947268421776048, + "learning_rate": 2.594737136201202e-06, + "loss": 0.2946, + "step": 30436 + }, + { + "epoch": 0.6699757869249394, + "grad_norm": 1.6245427767068392, + "learning_rate": 2.594424639947245e-06, + "loss": 0.3022, + "step": 30437 + }, + { + "epoch": 0.6699977988113581, + "grad_norm": 1.519890695676787, + "learning_rate": 2.5941121559194958e-06, + "loss": 0.2641, + "step": 30438 + }, + { + "epoch": 0.6700198106977768, + "grad_norm": 2.2933286762870773, + "learning_rate": 2.5937996841195402e-06, + "loss": 0.3044, + "step": 30439 + }, + { + "epoch": 0.6700418225841954, + "grad_norm": 1.6458378041900934, + "learning_rate": 2.593487224548967e-06, + "loss": 0.2877, + "step": 30440 + }, + { + "epoch": 0.6700638344706141, + "grad_norm": 1.2444049808119975, + "learning_rate": 2.5931747772093686e-06, + "loss": 0.2184, + "step": 30441 + }, + { + "epoch": 0.6700858463570328, + "grad_norm": 1.6100733141697738, + "learning_rate": 2.5928623421023264e-06, + "loss": 0.3437, + "step": 30442 + }, + { + "epoch": 0.6701078582434514, + "grad_norm": 9.775821581046348, + "learning_rate": 2.5925499192294334e-06, + "loss": 0.2983, + "step": 30443 + }, + { + "epoch": 0.6701298701298701, + "grad_norm": 1.5962872125829326, + "learning_rate": 2.5922375085922734e-06, + "loss": 0.3303, + "step": 30444 + }, + { + "epoch": 0.6701518820162888, + "grad_norm": 1.6565203813737337, + "learning_rate": 2.591925110192437e-06, + "loss": 0.3422, + "step": 30445 + }, + { + "epoch": 0.6701738939027074, + "grad_norm": 1.6659577349158847, + "learning_rate": 2.5916127240315125e-06, + "loss": 0.3626, + "step": 30446 + }, + { + "epoch": 0.6701959057891261, + "grad_norm": 2.8751197562318485, + "learning_rate": 2.5913003501110844e-06, + "loss": 0.258, + "step": 30447 + }, + { + "epoch": 0.6702179176755448, + "grad_norm": 1.2896618761200584, + "learning_rate": 2.5909879884327448e-06, + "loss": 0.2678, + "step": 30448 + }, + { + "epoch": 0.6702399295619634, + "grad_norm": 6.865957618776875, + "learning_rate": 2.590675638998078e-06, + "loss": 0.3378, + "step": 30449 + }, + { + "epoch": 0.6702619414483821, + "grad_norm": 2.3015883952949547, + "learning_rate": 2.5903633018086687e-06, + "loss": 0.2851, + "step": 30450 + }, + { + "epoch": 0.6702839533348008, + "grad_norm": 1.8973533334824342, + "learning_rate": 2.590050976866112e-06, + "loss": 0.3228, + "step": 30451 + }, + { + "epoch": 0.6703059652212194, + "grad_norm": 1.4797037770836414, + "learning_rate": 2.5897386641719884e-06, + "loss": 0.3209, + "step": 30452 + }, + { + "epoch": 0.6703279771076381, + "grad_norm": 1.4410151405368008, + "learning_rate": 2.5894263637278903e-06, + "loss": 0.3348, + "step": 30453 + }, + { + "epoch": 0.6703499889940568, + "grad_norm": 1.5520107611817517, + "learning_rate": 2.5891140755354026e-06, + "loss": 0.343, + "step": 30454 + }, + { + "epoch": 0.6703720008804754, + "grad_norm": 1.2132687909777744, + "learning_rate": 2.58880179959611e-06, + "loss": 0.2739, + "step": 30455 + }, + { + "epoch": 0.6703940127668941, + "grad_norm": 1.4337239065393268, + "learning_rate": 2.5884895359116024e-06, + "loss": 0.2855, + "step": 30456 + }, + { + "epoch": 0.6704160246533127, + "grad_norm": 1.795521442638723, + "learning_rate": 2.5881772844834667e-06, + "loss": 0.232, + "step": 30457 + }, + { + "epoch": 0.6704380365397314, + "grad_norm": 1.7748631805420725, + "learning_rate": 2.5878650453132916e-06, + "loss": 0.3256, + "step": 30458 + }, + { + "epoch": 0.6704600484261501, + "grad_norm": 0.6166276966153031, + "learning_rate": 2.5875528184026623e-06, + "loss": 0.4475, + "step": 30459 + }, + { + "epoch": 0.6704820603125687, + "grad_norm": 2.1969943758916903, + "learning_rate": 2.587240603753164e-06, + "loss": 0.3448, + "step": 30460 + }, + { + "epoch": 0.6705040721989874, + "grad_norm": 1.262845255832735, + "learning_rate": 2.5869284013663843e-06, + "loss": 0.2057, + "step": 30461 + }, + { + "epoch": 0.6705260840854061, + "grad_norm": 1.1746219296013733, + "learning_rate": 2.5866162112439107e-06, + "loss": 0.2372, + "step": 30462 + }, + { + "epoch": 0.6705480959718247, + "grad_norm": 1.262405734236391, + "learning_rate": 2.5863040333873323e-06, + "loss": 0.2011, + "step": 30463 + }, + { + "epoch": 0.6705701078582434, + "grad_norm": 1.0840515030788316, + "learning_rate": 2.5859918677982334e-06, + "loss": 0.2474, + "step": 30464 + }, + { + "epoch": 0.6705921197446622, + "grad_norm": 6.797813907111857, + "learning_rate": 2.585679714478198e-06, + "loss": 0.2456, + "step": 30465 + }, + { + "epoch": 0.6706141316310807, + "grad_norm": 1.4781162902191938, + "learning_rate": 2.5853675734288154e-06, + "loss": 0.3516, + "step": 30466 + }, + { + "epoch": 0.6706361435174995, + "grad_norm": 1.4907468680155453, + "learning_rate": 2.585055444651674e-06, + "loss": 0.3328, + "step": 30467 + }, + { + "epoch": 0.6706581554039182, + "grad_norm": 1.2869470730121129, + "learning_rate": 2.5847433281483558e-06, + "loss": 0.292, + "step": 30468 + }, + { + "epoch": 0.6706801672903367, + "grad_norm": 2.3193144218871087, + "learning_rate": 2.5844312239204504e-06, + "loss": 0.4257, + "step": 30469 + }, + { + "epoch": 0.6707021791767555, + "grad_norm": 0.6153737441732736, + "learning_rate": 2.5841191319695423e-06, + "loss": 0.4637, + "step": 30470 + }, + { + "epoch": 0.6707241910631742, + "grad_norm": 1.58393785270158, + "learning_rate": 2.5838070522972174e-06, + "loss": 0.3245, + "step": 30471 + }, + { + "epoch": 0.6707462029495928, + "grad_norm": 0.6154900899531044, + "learning_rate": 2.583494984905065e-06, + "loss": 0.4402, + "step": 30472 + }, + { + "epoch": 0.6707682148360115, + "grad_norm": 1.6936002626171074, + "learning_rate": 2.5831829297946663e-06, + "loss": 0.2447, + "step": 30473 + }, + { + "epoch": 0.6707902267224302, + "grad_norm": 0.6311262633601962, + "learning_rate": 2.5828708869676127e-06, + "loss": 0.4677, + "step": 30474 + }, + { + "epoch": 0.6708122386088488, + "grad_norm": 1.7261049620974012, + "learning_rate": 2.582558856425485e-06, + "loss": 0.2635, + "step": 30475 + }, + { + "epoch": 0.6708342504952675, + "grad_norm": 1.2545448060911846, + "learning_rate": 2.582246838169872e-06, + "loss": 0.4005, + "step": 30476 + }, + { + "epoch": 0.6708562623816862, + "grad_norm": 1.4822037684234801, + "learning_rate": 2.5819348322023595e-06, + "loss": 0.3439, + "step": 30477 + }, + { + "epoch": 0.6708782742681048, + "grad_norm": 1.5805994946894533, + "learning_rate": 2.581622838524532e-06, + "loss": 0.3156, + "step": 30478 + }, + { + "epoch": 0.6709002861545235, + "grad_norm": 1.1619633787455956, + "learning_rate": 2.5813108571379775e-06, + "loss": 0.2547, + "step": 30479 + }, + { + "epoch": 0.6709222980409421, + "grad_norm": 1.8963190335288826, + "learning_rate": 2.580998888044277e-06, + "loss": 0.3343, + "step": 30480 + }, + { + "epoch": 0.6709443099273608, + "grad_norm": 0.5900295478396244, + "learning_rate": 2.5806869312450206e-06, + "loss": 0.4357, + "step": 30481 + }, + { + "epoch": 0.6709663218137795, + "grad_norm": 1.7333690124361787, + "learning_rate": 2.5803749867417938e-06, + "loss": 0.3035, + "step": 30482 + }, + { + "epoch": 0.6709883337001981, + "grad_norm": 5.629222366700446, + "learning_rate": 2.5800630545361783e-06, + "loss": 0.29, + "step": 30483 + }, + { + "epoch": 0.6710103455866168, + "grad_norm": 0.6226207023411842, + "learning_rate": 2.5797511346297633e-06, + "loss": 0.4586, + "step": 30484 + }, + { + "epoch": 0.6710323574730355, + "grad_norm": 1.1963787009174855, + "learning_rate": 2.579439227024133e-06, + "loss": 0.311, + "step": 30485 + }, + { + "epoch": 0.6710543693594541, + "grad_norm": 0.6222296922792087, + "learning_rate": 2.5791273317208697e-06, + "loss": 0.4782, + "step": 30486 + }, + { + "epoch": 0.6710763812458728, + "grad_norm": 0.6121612631790596, + "learning_rate": 2.578815448721561e-06, + "loss": 0.4634, + "step": 30487 + }, + { + "epoch": 0.6710983931322915, + "grad_norm": 2.8801566375890153, + "learning_rate": 2.578503578027793e-06, + "loss": 0.2841, + "step": 30488 + }, + { + "epoch": 0.6711204050187101, + "grad_norm": 1.2597919820912604, + "learning_rate": 2.578191719641151e-06, + "loss": 0.3116, + "step": 30489 + }, + { + "epoch": 0.6711424169051288, + "grad_norm": 1.3196511434259055, + "learning_rate": 2.5778798735632194e-06, + "loss": 0.2717, + "step": 30490 + }, + { + "epoch": 0.6711644287915475, + "grad_norm": 5.684455696849424, + "learning_rate": 2.57756803979558e-06, + "loss": 0.4063, + "step": 30491 + }, + { + "epoch": 0.6711864406779661, + "grad_norm": 1.4737892264979056, + "learning_rate": 2.5772562183398204e-06, + "loss": 0.3616, + "step": 30492 + }, + { + "epoch": 0.6712084525643848, + "grad_norm": 1.423849079947998, + "learning_rate": 2.5769444091975254e-06, + "loss": 0.2572, + "step": 30493 + }, + { + "epoch": 0.6712304644508035, + "grad_norm": 2.5742163370772952, + "learning_rate": 2.576632612370281e-06, + "loss": 0.1892, + "step": 30494 + }, + { + "epoch": 0.6712524763372221, + "grad_norm": 1.6868248339364842, + "learning_rate": 2.576320827859672e-06, + "loss": 0.1847, + "step": 30495 + }, + { + "epoch": 0.6712744882236408, + "grad_norm": 1.552350798428342, + "learning_rate": 2.576009055667278e-06, + "loss": 0.3779, + "step": 30496 + }, + { + "epoch": 0.6712965001100595, + "grad_norm": 1.539813247367132, + "learning_rate": 2.575697295794688e-06, + "loss": 0.3339, + "step": 30497 + }, + { + "epoch": 0.6713185119964781, + "grad_norm": 2.334430233338152, + "learning_rate": 2.575385548243485e-06, + "loss": 0.2778, + "step": 30498 + }, + { + "epoch": 0.6713405238828968, + "grad_norm": 1.2580973669431483, + "learning_rate": 2.575073813015257e-06, + "loss": 0.2433, + "step": 30499 + }, + { + "epoch": 0.6713625357693155, + "grad_norm": 0.6248730186260354, + "learning_rate": 2.5747620901115845e-06, + "loss": 0.4508, + "step": 30500 + }, + { + "epoch": 0.6713845476557341, + "grad_norm": 1.4666276442467618, + "learning_rate": 2.5744503795340515e-06, + "loss": 0.2559, + "step": 30501 + }, + { + "epoch": 0.6714065595421528, + "grad_norm": 1.2601061604362134, + "learning_rate": 2.5741386812842427e-06, + "loss": 0.2632, + "step": 30502 + }, + { + "epoch": 0.6714285714285714, + "grad_norm": 3.6146646276679486, + "learning_rate": 2.573826995363744e-06, + "loss": 0.2294, + "step": 30503 + }, + { + "epoch": 0.6714505833149901, + "grad_norm": 1.8631185855672805, + "learning_rate": 2.57351532177414e-06, + "loss": 0.3525, + "step": 30504 + }, + { + "epoch": 0.6714725952014088, + "grad_norm": 2.9407650861789674, + "learning_rate": 2.573203660517013e-06, + "loss": 0.3031, + "step": 30505 + }, + { + "epoch": 0.6714946070878274, + "grad_norm": 0.6273719529071725, + "learning_rate": 2.572892011593946e-06, + "loss": 0.4649, + "step": 30506 + }, + { + "epoch": 0.6715166189742461, + "grad_norm": 2.552070852099691, + "learning_rate": 2.572580375006524e-06, + "loss": 0.3421, + "step": 30507 + }, + { + "epoch": 0.6715386308606648, + "grad_norm": 1.4846422628153748, + "learning_rate": 2.5722687507563338e-06, + "loss": 0.378, + "step": 30508 + }, + { + "epoch": 0.6715606427470834, + "grad_norm": 1.5511739289056488, + "learning_rate": 2.5719571388449538e-06, + "loss": 0.3121, + "step": 30509 + }, + { + "epoch": 0.6715826546335021, + "grad_norm": 1.525389074041327, + "learning_rate": 2.571645539273973e-06, + "loss": 0.3229, + "step": 30510 + }, + { + "epoch": 0.6716046665199208, + "grad_norm": 1.429226811092206, + "learning_rate": 2.571333952044971e-06, + "loss": 0.3581, + "step": 30511 + }, + { + "epoch": 0.6716266784063394, + "grad_norm": 1.6183788805630646, + "learning_rate": 2.571022377159532e-06, + "loss": 0.3019, + "step": 30512 + }, + { + "epoch": 0.6716486902927581, + "grad_norm": 1.3395852435200408, + "learning_rate": 2.570710814619244e-06, + "loss": 0.2969, + "step": 30513 + }, + { + "epoch": 0.6716707021791768, + "grad_norm": 1.6324150807547033, + "learning_rate": 2.5703992644256835e-06, + "loss": 0.2925, + "step": 30514 + }, + { + "epoch": 0.6716927140655954, + "grad_norm": 1.7881902458413355, + "learning_rate": 2.5700877265804415e-06, + "loss": 0.2813, + "step": 30515 + }, + { + "epoch": 0.6717147259520141, + "grad_norm": 3.0571832154900225, + "learning_rate": 2.5697762010850937e-06, + "loss": 0.3788, + "step": 30516 + }, + { + "epoch": 0.6717367378384328, + "grad_norm": 1.5040328140897226, + "learning_rate": 2.5694646879412277e-06, + "loss": 0.3714, + "step": 30517 + }, + { + "epoch": 0.6717587497248514, + "grad_norm": 4.1856037208093975, + "learning_rate": 2.569153187150429e-06, + "loss": 0.2246, + "step": 30518 + }, + { + "epoch": 0.6717807616112701, + "grad_norm": 1.7723638011728013, + "learning_rate": 2.568841698714275e-06, + "loss": 0.3921, + "step": 30519 + }, + { + "epoch": 0.6718027734976888, + "grad_norm": 0.624697323422977, + "learning_rate": 2.5685302226343546e-06, + "loss": 0.456, + "step": 30520 + }, + { + "epoch": 0.6718247853841074, + "grad_norm": 1.6435235040742888, + "learning_rate": 2.5682187589122455e-06, + "loss": 0.311, + "step": 30521 + }, + { + "epoch": 0.6718467972705261, + "grad_norm": 1.7978614061105684, + "learning_rate": 2.5679073075495354e-06, + "loss": 0.327, + "step": 30522 + }, + { + "epoch": 0.6718688091569448, + "grad_norm": 2.047350592820999, + "learning_rate": 2.567595868547803e-06, + "loss": 0.257, + "step": 30523 + }, + { + "epoch": 0.6718908210433634, + "grad_norm": 1.6506625800735315, + "learning_rate": 2.5672844419086336e-06, + "loss": 0.3451, + "step": 30524 + }, + { + "epoch": 0.6719128329297821, + "grad_norm": 5.920992691219836, + "learning_rate": 2.5669730276336114e-06, + "loss": 0.2127, + "step": 30525 + }, + { + "epoch": 0.6719348448162007, + "grad_norm": 0.6309938183233833, + "learning_rate": 2.5666616257243177e-06, + "loss": 0.4634, + "step": 30526 + }, + { + "epoch": 0.6719568567026194, + "grad_norm": 1.2696272339665369, + "learning_rate": 2.5663502361823323e-06, + "loss": 0.2297, + "step": 30527 + }, + { + "epoch": 0.6719788685890381, + "grad_norm": 1.3597616476795171, + "learning_rate": 2.5660388590092412e-06, + "loss": 0.334, + "step": 30528 + }, + { + "epoch": 0.6720008804754567, + "grad_norm": 1.670340918047393, + "learning_rate": 2.565727494206626e-06, + "loss": 0.2779, + "step": 30529 + }, + { + "epoch": 0.6720228923618754, + "grad_norm": 1.3471241332970534, + "learning_rate": 2.5654161417760715e-06, + "loss": 0.2641, + "step": 30530 + }, + { + "epoch": 0.6720449042482941, + "grad_norm": 1.4402415737742462, + "learning_rate": 2.565104801719157e-06, + "loss": 0.309, + "step": 30531 + }, + { + "epoch": 0.6720669161347127, + "grad_norm": 1.6521851117076172, + "learning_rate": 2.5647934740374654e-06, + "loss": 0.3006, + "step": 30532 + }, + { + "epoch": 0.6720889280211314, + "grad_norm": 1.3607857640757726, + "learning_rate": 2.5644821587325785e-06, + "loss": 0.347, + "step": 30533 + }, + { + "epoch": 0.6721109399075501, + "grad_norm": 2.324971437552304, + "learning_rate": 2.5641708558060795e-06, + "loss": 0.2751, + "step": 30534 + }, + { + "epoch": 0.6721329517939687, + "grad_norm": 1.7687220691931687, + "learning_rate": 2.5638595652595532e-06, + "loss": 0.3613, + "step": 30535 + }, + { + "epoch": 0.6721549636803874, + "grad_norm": 1.2810102077043115, + "learning_rate": 2.5635482870945793e-06, + "loss": 0.2548, + "step": 30536 + }, + { + "epoch": 0.6721769755668061, + "grad_norm": 1.6068338447085766, + "learning_rate": 2.563237021312737e-06, + "loss": 0.3069, + "step": 30537 + }, + { + "epoch": 0.6721989874532247, + "grad_norm": 2.127623225398877, + "learning_rate": 2.5629257679156117e-06, + "loss": 0.3058, + "step": 30538 + }, + { + "epoch": 0.6722209993396434, + "grad_norm": 1.617716322158659, + "learning_rate": 2.5626145269047843e-06, + "loss": 0.297, + "step": 30539 + }, + { + "epoch": 0.6722430112260621, + "grad_norm": 0.6107919672450717, + "learning_rate": 2.5623032982818387e-06, + "loss": 0.4565, + "step": 30540 + }, + { + "epoch": 0.6722650231124807, + "grad_norm": 1.5819771863892693, + "learning_rate": 2.5619920820483546e-06, + "loss": 0.3145, + "step": 30541 + }, + { + "epoch": 0.6722870349988994, + "grad_norm": 1.5341999147245484, + "learning_rate": 2.561680878205913e-06, + "loss": 0.282, + "step": 30542 + }, + { + "epoch": 0.6723090468853181, + "grad_norm": 1.5644995471919274, + "learning_rate": 2.561369686756097e-06, + "loss": 0.2734, + "step": 30543 + }, + { + "epoch": 0.6723310587717367, + "grad_norm": 1.967226965509145, + "learning_rate": 2.561058507700487e-06, + "loss": 0.2456, + "step": 30544 + }, + { + "epoch": 0.6723530706581554, + "grad_norm": 2.8010684248582405, + "learning_rate": 2.5607473410406673e-06, + "loss": 0.3384, + "step": 30545 + }, + { + "epoch": 0.6723750825445741, + "grad_norm": 2.299812740123596, + "learning_rate": 2.5604361867782177e-06, + "loss": 0.3054, + "step": 30546 + }, + { + "epoch": 0.6723970944309927, + "grad_norm": 1.356861940046231, + "learning_rate": 2.560125044914718e-06, + "loss": 0.277, + "step": 30547 + }, + { + "epoch": 0.6724191063174114, + "grad_norm": 1.75374259613046, + "learning_rate": 2.559813915451751e-06, + "loss": 0.2481, + "step": 30548 + }, + { + "epoch": 0.67244111820383, + "grad_norm": 1.7222353407264983, + "learning_rate": 2.5595027983908994e-06, + "loss": 0.2945, + "step": 30549 + }, + { + "epoch": 0.6724631300902487, + "grad_norm": 1.9655559583895723, + "learning_rate": 2.5591916937337417e-06, + "loss": 0.2593, + "step": 30550 + }, + { + "epoch": 0.6724851419766674, + "grad_norm": 1.5659718748756428, + "learning_rate": 2.5588806014818622e-06, + "loss": 0.2305, + "step": 30551 + }, + { + "epoch": 0.672507153863086, + "grad_norm": 1.8800644371969173, + "learning_rate": 2.558569521636839e-06, + "loss": 0.3748, + "step": 30552 + }, + { + "epoch": 0.6725291657495047, + "grad_norm": 1.512673160594998, + "learning_rate": 2.558258454200253e-06, + "loss": 0.2439, + "step": 30553 + }, + { + "epoch": 0.6725511776359234, + "grad_norm": 1.7055809239140431, + "learning_rate": 2.557947399173689e-06, + "loss": 0.3203, + "step": 30554 + }, + { + "epoch": 0.672573189522342, + "grad_norm": 1.6847680658231499, + "learning_rate": 2.557636356558724e-06, + "loss": 0.2762, + "step": 30555 + }, + { + "epoch": 0.6725952014087607, + "grad_norm": 1.6178368644045344, + "learning_rate": 2.5573253263569425e-06, + "loss": 0.2131, + "step": 30556 + }, + { + "epoch": 0.6726172132951794, + "grad_norm": 1.542362848467767, + "learning_rate": 2.5570143085699207e-06, + "loss": 0.2909, + "step": 30557 + }, + { + "epoch": 0.672639225181598, + "grad_norm": 1.3184206435301624, + "learning_rate": 2.556703303199244e-06, + "loss": 0.2506, + "step": 30558 + }, + { + "epoch": 0.6726612370680167, + "grad_norm": 2.1525616727603527, + "learning_rate": 2.5563923102464883e-06, + "loss": 0.2506, + "step": 30559 + }, + { + "epoch": 0.6726832489544354, + "grad_norm": 1.6605573984884736, + "learning_rate": 2.5560813297132376e-06, + "loss": 0.3234, + "step": 30560 + }, + { + "epoch": 0.672705260840854, + "grad_norm": 1.309824118039445, + "learning_rate": 2.5557703616010733e-06, + "loss": 0.2536, + "step": 30561 + }, + { + "epoch": 0.6727272727272727, + "grad_norm": 1.5511964445037574, + "learning_rate": 2.555459405911572e-06, + "loss": 0.2701, + "step": 30562 + }, + { + "epoch": 0.6727492846136914, + "grad_norm": 2.5922959578333846, + "learning_rate": 2.5551484626463185e-06, + "loss": 0.3891, + "step": 30563 + }, + { + "epoch": 0.67277129650011, + "grad_norm": 2.3227563185737603, + "learning_rate": 2.5548375318068886e-06, + "loss": 0.2412, + "step": 30564 + }, + { + "epoch": 0.6727933083865287, + "grad_norm": 2.6811871452869735, + "learning_rate": 2.5545266133948653e-06, + "loss": 0.2678, + "step": 30565 + }, + { + "epoch": 0.6728153202729474, + "grad_norm": 6.427263038621792, + "learning_rate": 2.55421570741183e-06, + "loss": 0.227, + "step": 30566 + }, + { + "epoch": 0.672837332159366, + "grad_norm": 1.7952539263067124, + "learning_rate": 2.553904813859362e-06, + "loss": 0.3285, + "step": 30567 + }, + { + "epoch": 0.6728593440457847, + "grad_norm": 2.8021113793103662, + "learning_rate": 2.553593932739038e-06, + "loss": 0.3012, + "step": 30568 + }, + { + "epoch": 0.6728813559322034, + "grad_norm": 1.9562474906815126, + "learning_rate": 2.5532830640524412e-06, + "loss": 0.2928, + "step": 30569 + }, + { + "epoch": 0.672903367818622, + "grad_norm": 2.6055183283591266, + "learning_rate": 2.552972207801151e-06, + "loss": 0.3566, + "step": 30570 + }, + { + "epoch": 0.6729253797050407, + "grad_norm": 2.4173866865628986, + "learning_rate": 2.5526613639867496e-06, + "loss": 0.2309, + "step": 30571 + }, + { + "epoch": 0.6729473915914594, + "grad_norm": 3.2106042685023155, + "learning_rate": 2.5523505326108145e-06, + "loss": 0.3059, + "step": 30572 + }, + { + "epoch": 0.672969403477878, + "grad_norm": 0.6401800247071244, + "learning_rate": 2.552039713674924e-06, + "loss": 0.4569, + "step": 30573 + }, + { + "epoch": 0.6729914153642967, + "grad_norm": 1.7874418397106795, + "learning_rate": 2.5517289071806594e-06, + "loss": 0.3165, + "step": 30574 + }, + { + "epoch": 0.6730134272507153, + "grad_norm": 2.190708222166816, + "learning_rate": 2.5514181131296e-06, + "loss": 0.2818, + "step": 30575 + }, + { + "epoch": 0.673035439137134, + "grad_norm": 0.6131854042942809, + "learning_rate": 2.5511073315233283e-06, + "loss": 0.452, + "step": 30576 + }, + { + "epoch": 0.6730574510235527, + "grad_norm": 1.4753454962283068, + "learning_rate": 2.5507965623634214e-06, + "loss": 0.3081, + "step": 30577 + }, + { + "epoch": 0.6730794629099713, + "grad_norm": 2.028640276893807, + "learning_rate": 2.550485805651457e-06, + "loss": 0.3473, + "step": 30578 + }, + { + "epoch": 0.67310147479639, + "grad_norm": 2.6475657282119784, + "learning_rate": 2.5501750613890154e-06, + "loss": 0.2711, + "step": 30579 + }, + { + "epoch": 0.6731234866828087, + "grad_norm": 0.5988891652367647, + "learning_rate": 2.549864329577677e-06, + "loss": 0.4429, + "step": 30580 + }, + { + "epoch": 0.6731454985692273, + "grad_norm": 1.6644314023469744, + "learning_rate": 2.5495536102190234e-06, + "loss": 0.2921, + "step": 30581 + }, + { + "epoch": 0.673167510455646, + "grad_norm": 1.4287782176229267, + "learning_rate": 2.5492429033146308e-06, + "loss": 0.3437, + "step": 30582 + }, + { + "epoch": 0.6731895223420648, + "grad_norm": 1.5003997961973754, + "learning_rate": 2.5489322088660774e-06, + "loss": 0.2508, + "step": 30583 + }, + { + "epoch": 0.6732115342284833, + "grad_norm": 1.9073551545443717, + "learning_rate": 2.548621526874943e-06, + "loss": 0.3355, + "step": 30584 + }, + { + "epoch": 0.673233546114902, + "grad_norm": 1.4392125674745775, + "learning_rate": 2.5483108573428096e-06, + "loss": 0.3034, + "step": 30585 + }, + { + "epoch": 0.6732555580013208, + "grad_norm": 1.6570074994325703, + "learning_rate": 2.5480002002712524e-06, + "loss": 0.3072, + "step": 30586 + }, + { + "epoch": 0.6732775698877393, + "grad_norm": 1.2554976279697747, + "learning_rate": 2.547689555661853e-06, + "loss": 0.3057, + "step": 30587 + }, + { + "epoch": 0.673299581774158, + "grad_norm": 1.326027523896697, + "learning_rate": 2.5473789235161876e-06, + "loss": 0.231, + "step": 30588 + }, + { + "epoch": 0.6733215936605768, + "grad_norm": 1.3231111223098748, + "learning_rate": 2.547068303835836e-06, + "loss": 0.3076, + "step": 30589 + }, + { + "epoch": 0.6733436055469953, + "grad_norm": 0.6240242377478338, + "learning_rate": 2.54675769662238e-06, + "loss": 0.4811, + "step": 30590 + }, + { + "epoch": 0.673365617433414, + "grad_norm": 1.3759880193452976, + "learning_rate": 2.5464471018773927e-06, + "loss": 0.2887, + "step": 30591 + }, + { + "epoch": 0.6733876293198328, + "grad_norm": 2.059349715478741, + "learning_rate": 2.5461365196024578e-06, + "loss": 0.3452, + "step": 30592 + }, + { + "epoch": 0.6734096412062514, + "grad_norm": 1.579677557832137, + "learning_rate": 2.54582594979915e-06, + "loss": 0.2671, + "step": 30593 + }, + { + "epoch": 0.6734316530926701, + "grad_norm": 2.8564668594488882, + "learning_rate": 2.5455153924690512e-06, + "loss": 0.2586, + "step": 30594 + }, + { + "epoch": 0.6734536649790888, + "grad_norm": 1.5130675820885837, + "learning_rate": 2.545204847613736e-06, + "loss": 0.3248, + "step": 30595 + }, + { + "epoch": 0.6734756768655074, + "grad_norm": 1.6546115887957673, + "learning_rate": 2.544894315234784e-06, + "loss": 0.3228, + "step": 30596 + }, + { + "epoch": 0.6734976887519261, + "grad_norm": 1.7415608261905973, + "learning_rate": 2.544583795333777e-06, + "loss": 0.2729, + "step": 30597 + }, + { + "epoch": 0.6735197006383447, + "grad_norm": 2.077338479000528, + "learning_rate": 2.5442732879122876e-06, + "loss": 0.204, + "step": 30598 + }, + { + "epoch": 0.6735417125247634, + "grad_norm": 1.5229142140591496, + "learning_rate": 2.543962792971899e-06, + "loss": 0.3366, + "step": 30599 + }, + { + "epoch": 0.6735637244111821, + "grad_norm": 3.4043205382510338, + "learning_rate": 2.543652310514184e-06, + "loss": 0.2582, + "step": 30600 + }, + { + "epoch": 0.6735857362976007, + "grad_norm": 1.5926611833007758, + "learning_rate": 2.543341840540725e-06, + "loss": 0.3169, + "step": 30601 + }, + { + "epoch": 0.6736077481840194, + "grad_norm": 1.2561042044820174, + "learning_rate": 2.5430313830530993e-06, + "loss": 0.2121, + "step": 30602 + }, + { + "epoch": 0.6736297600704381, + "grad_norm": 1.4948083711192448, + "learning_rate": 2.542720938052883e-06, + "loss": 0.3811, + "step": 30603 + }, + { + "epoch": 0.6736517719568567, + "grad_norm": 1.5363359035123818, + "learning_rate": 2.542410505541656e-06, + "loss": 0.3066, + "step": 30604 + }, + { + "epoch": 0.6736737838432754, + "grad_norm": 1.7311420000900821, + "learning_rate": 2.542100085520993e-06, + "loss": 0.3299, + "step": 30605 + }, + { + "epoch": 0.6736957957296941, + "grad_norm": 1.6957415900063448, + "learning_rate": 2.5417896779924737e-06, + "loss": 0.2991, + "step": 30606 + }, + { + "epoch": 0.6737178076161127, + "grad_norm": 1.5042850250000808, + "learning_rate": 2.541479282957678e-06, + "loss": 0.2842, + "step": 30607 + }, + { + "epoch": 0.6737398195025314, + "grad_norm": 2.104319705375917, + "learning_rate": 2.5411689004181803e-06, + "loss": 0.2888, + "step": 30608 + }, + { + "epoch": 0.6737618313889501, + "grad_norm": 3.9689742583272913, + "learning_rate": 2.5408585303755577e-06, + "loss": 0.3395, + "step": 30609 + }, + { + "epoch": 0.6737838432753687, + "grad_norm": 4.2133841748488114, + "learning_rate": 2.5405481728313886e-06, + "loss": 0.3561, + "step": 30610 + }, + { + "epoch": 0.6738058551617874, + "grad_norm": 1.2866026310404528, + "learning_rate": 2.540237827787251e-06, + "loss": 0.2811, + "step": 30611 + }, + { + "epoch": 0.6738278670482061, + "grad_norm": 1.9007858838880092, + "learning_rate": 2.5399274952447234e-06, + "loss": 0.3075, + "step": 30612 + }, + { + "epoch": 0.6738498789346247, + "grad_norm": 0.6212814001338055, + "learning_rate": 2.5396171752053813e-06, + "loss": 0.445, + "step": 30613 + }, + { + "epoch": 0.6738718908210434, + "grad_norm": 2.3057494660629136, + "learning_rate": 2.5393068676708006e-06, + "loss": 0.2431, + "step": 30614 + }, + { + "epoch": 0.6738939027074621, + "grad_norm": 0.6287037883289773, + "learning_rate": 2.5389965726425596e-06, + "loss": 0.4648, + "step": 30615 + }, + { + "epoch": 0.6739159145938807, + "grad_norm": 1.072109632163951, + "learning_rate": 2.5386862901222353e-06, + "loss": 0.2396, + "step": 30616 + }, + { + "epoch": 0.6739379264802994, + "grad_norm": 1.3984747360070737, + "learning_rate": 2.538376020111408e-06, + "loss": 0.2351, + "step": 30617 + }, + { + "epoch": 0.6739599383667181, + "grad_norm": 1.7299844436401224, + "learning_rate": 2.538065762611651e-06, + "loss": 0.3178, + "step": 30618 + }, + { + "epoch": 0.6739819502531367, + "grad_norm": 1.2726860892841865, + "learning_rate": 2.53775551762454e-06, + "loss": 0.3383, + "step": 30619 + }, + { + "epoch": 0.6740039621395554, + "grad_norm": 0.6432874734753065, + "learning_rate": 2.537445285151654e-06, + "loss": 0.4591, + "step": 30620 + }, + { + "epoch": 0.674025974025974, + "grad_norm": 0.6052851806787978, + "learning_rate": 2.537135065194569e-06, + "loss": 0.4859, + "step": 30621 + }, + { + "epoch": 0.6740479859123927, + "grad_norm": 1.9489616561358278, + "learning_rate": 2.5368248577548638e-06, + "loss": 0.2455, + "step": 30622 + }, + { + "epoch": 0.6740699977988114, + "grad_norm": 1.4224310277092789, + "learning_rate": 2.536514662834114e-06, + "loss": 0.2574, + "step": 30623 + }, + { + "epoch": 0.67409200968523, + "grad_norm": 1.6004709431580386, + "learning_rate": 2.5362044804338927e-06, + "loss": 0.2845, + "step": 30624 + }, + { + "epoch": 0.6741140215716487, + "grad_norm": 2.205134181531764, + "learning_rate": 2.5358943105557794e-06, + "loss": 0.2836, + "step": 30625 + }, + { + "epoch": 0.6741360334580674, + "grad_norm": 1.214408175348053, + "learning_rate": 2.5355841532013524e-06, + "loss": 0.2012, + "step": 30626 + }, + { + "epoch": 0.674158045344486, + "grad_norm": 1.8494683869574058, + "learning_rate": 2.535274008372184e-06, + "loss": 0.2318, + "step": 30627 + }, + { + "epoch": 0.6741800572309047, + "grad_norm": 1.4726454558547275, + "learning_rate": 2.5349638760698537e-06, + "loss": 0.363, + "step": 30628 + }, + { + "epoch": 0.6742020691173234, + "grad_norm": 1.5490203364516364, + "learning_rate": 2.534653756295935e-06, + "loss": 0.2744, + "step": 30629 + }, + { + "epoch": 0.674224081003742, + "grad_norm": 1.6146097611447456, + "learning_rate": 2.534343649052007e-06, + "loss": 0.299, + "step": 30630 + }, + { + "epoch": 0.6742460928901607, + "grad_norm": 2.0111549864357796, + "learning_rate": 2.5340335543396434e-06, + "loss": 0.4215, + "step": 30631 + }, + { + "epoch": 0.6742681047765794, + "grad_norm": 2.4137373727174127, + "learning_rate": 2.53372347216042e-06, + "loss": 0.2868, + "step": 30632 + }, + { + "epoch": 0.674290116662998, + "grad_norm": 0.6237592837368369, + "learning_rate": 2.533413402515916e-06, + "loss": 0.4548, + "step": 30633 + }, + { + "epoch": 0.6743121285494167, + "grad_norm": 1.4702120372077827, + "learning_rate": 2.5331033454077037e-06, + "loss": 0.3145, + "step": 30634 + }, + { + "epoch": 0.6743341404358354, + "grad_norm": 0.6208154018052914, + "learning_rate": 2.532793300837362e-06, + "loss": 0.4393, + "step": 30635 + }, + { + "epoch": 0.674356152322254, + "grad_norm": 2.962077796650969, + "learning_rate": 2.5324832688064636e-06, + "loss": 0.2269, + "step": 30636 + }, + { + "epoch": 0.6743781642086727, + "grad_norm": 0.6231601494386815, + "learning_rate": 2.5321732493165857e-06, + "loss": 0.4756, + "step": 30637 + }, + { + "epoch": 0.6744001760950914, + "grad_norm": 2.29320737414372, + "learning_rate": 2.5318632423693056e-06, + "loss": 0.291, + "step": 30638 + }, + { + "epoch": 0.67442218798151, + "grad_norm": 1.7272809043294546, + "learning_rate": 2.531553247966196e-06, + "loss": 0.3376, + "step": 30639 + }, + { + "epoch": 0.6744441998679287, + "grad_norm": 1.8066815476926121, + "learning_rate": 2.531243266108835e-06, + "loss": 0.2681, + "step": 30640 + }, + { + "epoch": 0.6744662117543474, + "grad_norm": 1.499420998498729, + "learning_rate": 2.5309332967987955e-06, + "loss": 0.3428, + "step": 30641 + }, + { + "epoch": 0.674488223640766, + "grad_norm": 1.5214798763631652, + "learning_rate": 2.530623340037654e-06, + "loss": 0.2689, + "step": 30642 + }, + { + "epoch": 0.6745102355271847, + "grad_norm": 5.001387868485651, + "learning_rate": 2.5303133958269876e-06, + "loss": 0.2762, + "step": 30643 + }, + { + "epoch": 0.6745322474136033, + "grad_norm": 1.8937400151210746, + "learning_rate": 2.5300034641683683e-06, + "loss": 0.2954, + "step": 30644 + }, + { + "epoch": 0.674554259300022, + "grad_norm": 1.6656373577696562, + "learning_rate": 2.5296935450633753e-06, + "loss": 0.3188, + "step": 30645 + }, + { + "epoch": 0.6745762711864407, + "grad_norm": 1.5386224102463892, + "learning_rate": 2.5293836385135794e-06, + "loss": 0.3101, + "step": 30646 + }, + { + "epoch": 0.6745982830728593, + "grad_norm": 4.454800188629609, + "learning_rate": 2.529073744520558e-06, + "loss": 0.336, + "step": 30647 + }, + { + "epoch": 0.674620294959278, + "grad_norm": 1.6611393675663826, + "learning_rate": 2.5287638630858876e-06, + "loss": 0.3527, + "step": 30648 + }, + { + "epoch": 0.6746423068456967, + "grad_norm": 1.5142264777461916, + "learning_rate": 2.5284539942111414e-06, + "loss": 0.3085, + "step": 30649 + }, + { + "epoch": 0.6746643187321153, + "grad_norm": 1.9802228529882824, + "learning_rate": 2.528144137897892e-06, + "loss": 0.2894, + "step": 30650 + }, + { + "epoch": 0.674686330618534, + "grad_norm": 0.6164398275731101, + "learning_rate": 2.5278342941477173e-06, + "loss": 0.4457, + "step": 30651 + }, + { + "epoch": 0.6747083425049527, + "grad_norm": 2.0622065360322233, + "learning_rate": 2.527524462962191e-06, + "loss": 0.2926, + "step": 30652 + }, + { + "epoch": 0.6747303543913713, + "grad_norm": 0.6152026586047459, + "learning_rate": 2.5272146443428903e-06, + "loss": 0.4592, + "step": 30653 + }, + { + "epoch": 0.67475236627779, + "grad_norm": 0.6009629670923964, + "learning_rate": 2.5269048382913873e-06, + "loss": 0.4361, + "step": 30654 + }, + { + "epoch": 0.6747743781642087, + "grad_norm": 1.3853015937695081, + "learning_rate": 2.526595044809255e-06, + "loss": 0.2941, + "step": 30655 + }, + { + "epoch": 0.6747963900506273, + "grad_norm": 1.4072893112035472, + "learning_rate": 2.5262852638980708e-06, + "loss": 0.3018, + "step": 30656 + }, + { + "epoch": 0.674818401937046, + "grad_norm": 0.5974025231227292, + "learning_rate": 2.525975495559407e-06, + "loss": 0.4498, + "step": 30657 + }, + { + "epoch": 0.6748404138234647, + "grad_norm": 0.6376090982150275, + "learning_rate": 2.5256657397948415e-06, + "loss": 0.4735, + "step": 30658 + }, + { + "epoch": 0.6748624257098833, + "grad_norm": 1.6176823562261986, + "learning_rate": 2.525355996605946e-06, + "loss": 0.2561, + "step": 30659 + }, + { + "epoch": 0.674884437596302, + "grad_norm": 3.0757445869702726, + "learning_rate": 2.525046265994294e-06, + "loss": 0.3662, + "step": 30660 + }, + { + "epoch": 0.6749064494827207, + "grad_norm": 1.5173020470645193, + "learning_rate": 2.5247365479614597e-06, + "loss": 0.2536, + "step": 30661 + }, + { + "epoch": 0.6749284613691393, + "grad_norm": 1.6941475440091351, + "learning_rate": 2.524426842509019e-06, + "loss": 0.3365, + "step": 30662 + }, + { + "epoch": 0.674950473255558, + "grad_norm": 1.5296960625185945, + "learning_rate": 2.5241171496385468e-06, + "loss": 0.3521, + "step": 30663 + }, + { + "epoch": 0.6749724851419767, + "grad_norm": 1.2933292151211595, + "learning_rate": 2.523807469351616e-06, + "loss": 0.2702, + "step": 30664 + }, + { + "epoch": 0.6749944970283953, + "grad_norm": 1.8480780612975711, + "learning_rate": 2.5234978016497975e-06, + "loss": 0.2143, + "step": 30665 + }, + { + "epoch": 0.675016508914814, + "grad_norm": 1.857899975834783, + "learning_rate": 2.52318814653467e-06, + "loss": 0.3035, + "step": 30666 + }, + { + "epoch": 0.6750385208012326, + "grad_norm": 0.6099156520088655, + "learning_rate": 2.522878504007803e-06, + "loss": 0.434, + "step": 30667 + }, + { + "epoch": 0.6750605326876513, + "grad_norm": 1.623827623656193, + "learning_rate": 2.5225688740707723e-06, + "loss": 0.3239, + "step": 30668 + }, + { + "epoch": 0.67508254457407, + "grad_norm": 1.4373393222394448, + "learning_rate": 2.5222592567251538e-06, + "loss": 0.2845, + "step": 30669 + }, + { + "epoch": 0.6751045564604886, + "grad_norm": 0.6234286933288942, + "learning_rate": 2.5219496519725164e-06, + "loss": 0.483, + "step": 30670 + }, + { + "epoch": 0.6751265683469073, + "grad_norm": 1.6275532940384485, + "learning_rate": 2.5216400598144373e-06, + "loss": 0.3011, + "step": 30671 + }, + { + "epoch": 0.675148580233326, + "grad_norm": 1.7005176970216145, + "learning_rate": 2.5213304802524886e-06, + "loss": 0.2951, + "step": 30672 + }, + { + "epoch": 0.6751705921197446, + "grad_norm": 2.7340514709515906, + "learning_rate": 2.521020913288243e-06, + "loss": 0.2409, + "step": 30673 + }, + { + "epoch": 0.6751926040061633, + "grad_norm": 2.4692316567139847, + "learning_rate": 2.5207113589232767e-06, + "loss": 0.3321, + "step": 30674 + }, + { + "epoch": 0.675214615892582, + "grad_norm": 1.3855175973110647, + "learning_rate": 2.5204018171591594e-06, + "loss": 0.2537, + "step": 30675 + }, + { + "epoch": 0.6752366277790006, + "grad_norm": 1.3876944523215398, + "learning_rate": 2.5200922879974676e-06, + "loss": 0.2246, + "step": 30676 + }, + { + "epoch": 0.6752586396654193, + "grad_norm": 0.6391036329314984, + "learning_rate": 2.519782771439771e-06, + "loss": 0.4581, + "step": 30677 + }, + { + "epoch": 0.675280651551838, + "grad_norm": 1.3564292096961679, + "learning_rate": 2.5194732674876444e-06, + "loss": 0.2567, + "step": 30678 + }, + { + "epoch": 0.6753026634382566, + "grad_norm": 0.6300462478828415, + "learning_rate": 2.5191637761426635e-06, + "loss": 0.4666, + "step": 30679 + }, + { + "epoch": 0.6753246753246753, + "grad_norm": 0.6365986500586857, + "learning_rate": 2.518854297406397e-06, + "loss": 0.4624, + "step": 30680 + }, + { + "epoch": 0.675346687211094, + "grad_norm": 1.8299807702008433, + "learning_rate": 2.5185448312804206e-06, + "loss": 0.3111, + "step": 30681 + }, + { + "epoch": 0.6753686990975126, + "grad_norm": 1.331967703073058, + "learning_rate": 2.518235377766305e-06, + "loss": 0.2684, + "step": 30682 + }, + { + "epoch": 0.6753907109839313, + "grad_norm": 1.9336611900942022, + "learning_rate": 2.5179259368656233e-06, + "loss": 0.3591, + "step": 30683 + }, + { + "epoch": 0.67541272287035, + "grad_norm": 1.3800378232554813, + "learning_rate": 2.5176165085799516e-06, + "loss": 0.2909, + "step": 30684 + }, + { + "epoch": 0.6754347347567686, + "grad_norm": 1.623859328589609, + "learning_rate": 2.51730709291086e-06, + "loss": 0.1788, + "step": 30685 + }, + { + "epoch": 0.6754567466431873, + "grad_norm": 1.488603170561668, + "learning_rate": 2.5169976898599186e-06, + "loss": 0.3067, + "step": 30686 + }, + { + "epoch": 0.675478758529606, + "grad_norm": 0.5974619261144855, + "learning_rate": 2.516688299428703e-06, + "loss": 0.4628, + "step": 30687 + }, + { + "epoch": 0.6755007704160246, + "grad_norm": 1.4369788920102313, + "learning_rate": 2.5163789216187845e-06, + "loss": 0.3502, + "step": 30688 + }, + { + "epoch": 0.6755227823024433, + "grad_norm": 1.970917529269905, + "learning_rate": 2.5160695564317382e-06, + "loss": 0.3362, + "step": 30689 + }, + { + "epoch": 0.6755447941888619, + "grad_norm": 1.662222722914649, + "learning_rate": 2.515760203869134e-06, + "loss": 0.3425, + "step": 30690 + }, + { + "epoch": 0.6755668060752806, + "grad_norm": 1.9686361993252566, + "learning_rate": 2.515450863932543e-06, + "loss": 0.2134, + "step": 30691 + }, + { + "epoch": 0.6755888179616993, + "grad_norm": 1.8233363081132399, + "learning_rate": 2.515141536623539e-06, + "loss": 0.3096, + "step": 30692 + }, + { + "epoch": 0.6756108298481179, + "grad_norm": 2.0788372938830917, + "learning_rate": 2.5148322219436937e-06, + "loss": 0.2657, + "step": 30693 + }, + { + "epoch": 0.6756328417345366, + "grad_norm": 1.7220710751543031, + "learning_rate": 2.5145229198945808e-06, + "loss": 0.3574, + "step": 30694 + }, + { + "epoch": 0.6756548536209553, + "grad_norm": 1.4991313594725346, + "learning_rate": 2.514213630477772e-06, + "loss": 0.2872, + "step": 30695 + }, + { + "epoch": 0.6756768655073739, + "grad_norm": 1.4347828164221328, + "learning_rate": 2.513904353694836e-06, + "loss": 0.311, + "step": 30696 + }, + { + "epoch": 0.6756988773937926, + "grad_norm": 0.6149289056533537, + "learning_rate": 2.5135950895473466e-06, + "loss": 0.4427, + "step": 30697 + }, + { + "epoch": 0.6757208892802113, + "grad_norm": 1.7812292348589807, + "learning_rate": 2.5132858380368762e-06, + "loss": 0.3069, + "step": 30698 + }, + { + "epoch": 0.6757429011666299, + "grad_norm": 0.6102442401789454, + "learning_rate": 2.512976599164998e-06, + "loss": 0.4323, + "step": 30699 + }, + { + "epoch": 0.6757649130530486, + "grad_norm": 1.5407290597516696, + "learning_rate": 2.512667372933282e-06, + "loss": 0.355, + "step": 30700 + }, + { + "epoch": 0.6757869249394673, + "grad_norm": 1.230120911147969, + "learning_rate": 2.5123581593432976e-06, + "loss": 0.3012, + "step": 30701 + }, + { + "epoch": 0.6758089368258859, + "grad_norm": 1.8971451044015741, + "learning_rate": 2.5120489583966212e-06, + "loss": 0.3773, + "step": 30702 + }, + { + "epoch": 0.6758309487123046, + "grad_norm": 1.3946776688040186, + "learning_rate": 2.511739770094817e-06, + "loss": 0.2217, + "step": 30703 + }, + { + "epoch": 0.6758529605987234, + "grad_norm": 1.498200739821175, + "learning_rate": 2.5114305944394657e-06, + "loss": 0.2016, + "step": 30704 + }, + { + "epoch": 0.675874972485142, + "grad_norm": 1.6439210088288323, + "learning_rate": 2.5111214314321332e-06, + "loss": 0.3441, + "step": 30705 + }, + { + "epoch": 0.6758969843715606, + "grad_norm": 1.4103972112908356, + "learning_rate": 2.510812281074391e-06, + "loss": 0.3162, + "step": 30706 + }, + { + "epoch": 0.6759189962579794, + "grad_norm": 1.3267087376225921, + "learning_rate": 2.5105031433678118e-06, + "loss": 0.2172, + "step": 30707 + }, + { + "epoch": 0.675941008144398, + "grad_norm": 1.9462347733725949, + "learning_rate": 2.510194018313964e-06, + "loss": 0.3154, + "step": 30708 + }, + { + "epoch": 0.6759630200308167, + "grad_norm": 1.796596368182313, + "learning_rate": 2.5098849059144213e-06, + "loss": 0.2713, + "step": 30709 + }, + { + "epoch": 0.6759850319172354, + "grad_norm": 0.6218765750303428, + "learning_rate": 2.5095758061707553e-06, + "loss": 0.4646, + "step": 30710 + }, + { + "epoch": 0.676007043803654, + "grad_norm": 1.2491944426170407, + "learning_rate": 2.5092667190845342e-06, + "loss": 0.3121, + "step": 30711 + }, + { + "epoch": 0.6760290556900727, + "grad_norm": 1.1832031489874084, + "learning_rate": 2.5089576446573333e-06, + "loss": 0.2585, + "step": 30712 + }, + { + "epoch": 0.6760510675764914, + "grad_norm": 6.266521599413042, + "learning_rate": 2.508648582890718e-06, + "loss": 0.2852, + "step": 30713 + }, + { + "epoch": 0.67607307946291, + "grad_norm": 1.4470336973756739, + "learning_rate": 2.5083395337862616e-06, + "loss": 0.2411, + "step": 30714 + }, + { + "epoch": 0.6760950913493287, + "grad_norm": 1.8438064489312986, + "learning_rate": 2.5080304973455373e-06, + "loss": 0.3538, + "step": 30715 + }, + { + "epoch": 0.6761171032357473, + "grad_norm": 1.6423176788071103, + "learning_rate": 2.5077214735701116e-06, + "loss": 0.33, + "step": 30716 + }, + { + "epoch": 0.676139115122166, + "grad_norm": 0.6078124076858881, + "learning_rate": 2.507412462461559e-06, + "loss": 0.4246, + "step": 30717 + }, + { + "epoch": 0.6761611270085847, + "grad_norm": 1.4149242656532983, + "learning_rate": 2.507103464021445e-06, + "loss": 0.3226, + "step": 30718 + }, + { + "epoch": 0.6761831388950033, + "grad_norm": 1.9704562893422608, + "learning_rate": 2.5067944782513443e-06, + "loss": 0.2725, + "step": 30719 + }, + { + "epoch": 0.676205150781422, + "grad_norm": 1.4505055076536093, + "learning_rate": 2.5064855051528275e-06, + "loss": 0.3637, + "step": 30720 + }, + { + "epoch": 0.6762271626678407, + "grad_norm": 3.2604582051414717, + "learning_rate": 2.506176544727462e-06, + "loss": 0.2448, + "step": 30721 + }, + { + "epoch": 0.6762491745542593, + "grad_norm": 1.5823893436196101, + "learning_rate": 2.505867596976821e-06, + "loss": 0.3103, + "step": 30722 + }, + { + "epoch": 0.676271186440678, + "grad_norm": 2.6098265111020638, + "learning_rate": 2.5055586619024713e-06, + "loss": 0.2727, + "step": 30723 + }, + { + "epoch": 0.6762931983270967, + "grad_norm": 0.6338766638335932, + "learning_rate": 2.5052497395059847e-06, + "loss": 0.4723, + "step": 30724 + }, + { + "epoch": 0.6763152102135153, + "grad_norm": 1.5377569579123245, + "learning_rate": 2.5049408297889342e-06, + "loss": 0.2527, + "step": 30725 + }, + { + "epoch": 0.676337222099934, + "grad_norm": 2.2562523513867108, + "learning_rate": 2.5046319327528873e-06, + "loss": 0.353, + "step": 30726 + }, + { + "epoch": 0.6763592339863527, + "grad_norm": 0.5965276441272617, + "learning_rate": 2.5043230483994117e-06, + "loss": 0.45, + "step": 30727 + }, + { + "epoch": 0.6763812458727713, + "grad_norm": 8.71156844482146, + "learning_rate": 2.5040141767300796e-06, + "loss": 0.3356, + "step": 30728 + }, + { + "epoch": 0.67640325775919, + "grad_norm": 2.0232318854641056, + "learning_rate": 2.503705317746461e-06, + "loss": 0.2763, + "step": 30729 + }, + { + "epoch": 0.6764252696456087, + "grad_norm": 1.1539968349656178, + "learning_rate": 2.503396471450127e-06, + "loss": 0.2281, + "step": 30730 + }, + { + "epoch": 0.6764472815320273, + "grad_norm": 1.9211343317406304, + "learning_rate": 2.503087637842645e-06, + "loss": 0.2944, + "step": 30731 + }, + { + "epoch": 0.676469293418446, + "grad_norm": 6.961841524805174, + "learning_rate": 2.502778816925584e-06, + "loss": 0.2462, + "step": 30732 + }, + { + "epoch": 0.6764913053048647, + "grad_norm": 1.6684000861697648, + "learning_rate": 2.502470008700517e-06, + "loss": 0.3435, + "step": 30733 + }, + { + "epoch": 0.6765133171912833, + "grad_norm": 3.3877303394731135, + "learning_rate": 2.5021612131690065e-06, + "loss": 0.3734, + "step": 30734 + }, + { + "epoch": 0.676535329077702, + "grad_norm": 0.6177662342300204, + "learning_rate": 2.501852430332632e-06, + "loss": 0.433, + "step": 30735 + }, + { + "epoch": 0.6765573409641207, + "grad_norm": 1.321105665964515, + "learning_rate": 2.5015436601929567e-06, + "loss": 0.3129, + "step": 30736 + }, + { + "epoch": 0.6765793528505393, + "grad_norm": 2.646593717604598, + "learning_rate": 2.501234902751549e-06, + "loss": 0.296, + "step": 30737 + }, + { + "epoch": 0.676601364736958, + "grad_norm": 2.8077818891565696, + "learning_rate": 2.500926158009982e-06, + "loss": 0.2504, + "step": 30738 + }, + { + "epoch": 0.6766233766233766, + "grad_norm": 1.2093229812106578, + "learning_rate": 2.500617425969819e-06, + "loss": 0.2408, + "step": 30739 + }, + { + "epoch": 0.6766453885097953, + "grad_norm": 1.7315246501634827, + "learning_rate": 2.500308706632637e-06, + "loss": 0.3205, + "step": 30740 + }, + { + "epoch": 0.676667400396214, + "grad_norm": 1.4391041974965464, + "learning_rate": 2.5000000000000015e-06, + "loss": 0.3433, + "step": 30741 + }, + { + "epoch": 0.6766894122826326, + "grad_norm": 1.3210400624019445, + "learning_rate": 2.4996913060734784e-06, + "loss": 0.2532, + "step": 30742 + }, + { + "epoch": 0.6767114241690513, + "grad_norm": 1.5448204025332057, + "learning_rate": 2.4993826248546417e-06, + "loss": 0.2834, + "step": 30743 + }, + { + "epoch": 0.67673343605547, + "grad_norm": 1.7482854335465137, + "learning_rate": 2.499073956345055e-06, + "loss": 0.2727, + "step": 30744 + }, + { + "epoch": 0.6767554479418886, + "grad_norm": 1.5674199825624715, + "learning_rate": 2.498765300546291e-06, + "loss": 0.2787, + "step": 30745 + }, + { + "epoch": 0.6767774598283073, + "grad_norm": 1.7765216076434514, + "learning_rate": 2.4984566574599182e-06, + "loss": 0.2916, + "step": 30746 + }, + { + "epoch": 0.676799471714726, + "grad_norm": 1.6495174596064, + "learning_rate": 2.498148027087503e-06, + "loss": 0.2677, + "step": 30747 + }, + { + "epoch": 0.6768214836011446, + "grad_norm": 1.6160732151773514, + "learning_rate": 2.4978394094306173e-06, + "loss": 0.2746, + "step": 30748 + }, + { + "epoch": 0.6768434954875633, + "grad_norm": 1.8410644264925073, + "learning_rate": 2.4975308044908257e-06, + "loss": 0.2383, + "step": 30749 + }, + { + "epoch": 0.676865507373982, + "grad_norm": 7.56714622707287, + "learning_rate": 2.497222212269699e-06, + "loss": 0.1776, + "step": 30750 + }, + { + "epoch": 0.6768875192604006, + "grad_norm": 2.06112697693095, + "learning_rate": 2.496913632768807e-06, + "loss": 0.3828, + "step": 30751 + }, + { + "epoch": 0.6769095311468193, + "grad_norm": 2.5490303628308157, + "learning_rate": 2.4966050659897144e-06, + "loss": 0.3045, + "step": 30752 + }, + { + "epoch": 0.676931543033238, + "grad_norm": 2.5876747116390595, + "learning_rate": 2.4962965119339935e-06, + "loss": 0.2983, + "step": 30753 + }, + { + "epoch": 0.6769535549196566, + "grad_norm": 1.7642513714432817, + "learning_rate": 2.4959879706032083e-06, + "loss": 0.2503, + "step": 30754 + }, + { + "epoch": 0.6769755668060753, + "grad_norm": 1.4621610429271434, + "learning_rate": 2.4956794419989288e-06, + "loss": 0.3197, + "step": 30755 + }, + { + "epoch": 0.676997578692494, + "grad_norm": 1.8061662185860794, + "learning_rate": 2.4953709261227254e-06, + "loss": 0.4229, + "step": 30756 + }, + { + "epoch": 0.6770195905789126, + "grad_norm": 1.4269779897755503, + "learning_rate": 2.4950624229761626e-06, + "loss": 0.2481, + "step": 30757 + }, + { + "epoch": 0.6770416024653313, + "grad_norm": 0.6332992336152912, + "learning_rate": 2.494753932560811e-06, + "loss": 0.4705, + "step": 30758 + }, + { + "epoch": 0.67706361435175, + "grad_norm": 1.8757351862776668, + "learning_rate": 2.494445454878236e-06, + "loss": 0.2424, + "step": 30759 + }, + { + "epoch": 0.6770856262381686, + "grad_norm": 1.3006341321510768, + "learning_rate": 2.494136989930007e-06, + "loss": 0.1795, + "step": 30760 + }, + { + "epoch": 0.6771076381245873, + "grad_norm": 1.4227476118838764, + "learning_rate": 2.4938285377176925e-06, + "loss": 0.2902, + "step": 30761 + }, + { + "epoch": 0.6771296500110059, + "grad_norm": 0.6244454592132004, + "learning_rate": 2.4935200982428577e-06, + "loss": 0.4531, + "step": 30762 + }, + { + "epoch": 0.6771516618974246, + "grad_norm": 0.6180522182230499, + "learning_rate": 2.493211671507073e-06, + "loss": 0.4505, + "step": 30763 + }, + { + "epoch": 0.6771736737838433, + "grad_norm": 0.6340610938835257, + "learning_rate": 2.4929032575119032e-06, + "loss": 0.4585, + "step": 30764 + }, + { + "epoch": 0.6771956856702619, + "grad_norm": 0.6220732989293098, + "learning_rate": 2.4925948562589177e-06, + "loss": 0.4744, + "step": 30765 + }, + { + "epoch": 0.6772176975566806, + "grad_norm": 0.6085476766009824, + "learning_rate": 2.4922864677496843e-06, + "loss": 0.4652, + "step": 30766 + }, + { + "epoch": 0.6772397094430993, + "grad_norm": 1.5212094877246611, + "learning_rate": 2.4919780919857704e-06, + "loss": 0.3465, + "step": 30767 + }, + { + "epoch": 0.6772617213295179, + "grad_norm": 1.740763956223403, + "learning_rate": 2.4916697289687402e-06, + "loss": 0.3489, + "step": 30768 + }, + { + "epoch": 0.6772837332159366, + "grad_norm": 2.1620776853678203, + "learning_rate": 2.491361378700165e-06, + "loss": 0.2804, + "step": 30769 + }, + { + "epoch": 0.6773057451023553, + "grad_norm": 2.5661195613970063, + "learning_rate": 2.491053041181607e-06, + "loss": 0.2695, + "step": 30770 + }, + { + "epoch": 0.6773277569887739, + "grad_norm": 1.439009405016603, + "learning_rate": 2.4907447164146397e-06, + "loss": 0.2743, + "step": 30771 + }, + { + "epoch": 0.6773497688751926, + "grad_norm": 1.5838514528917507, + "learning_rate": 2.490436404400827e-06, + "loss": 0.3067, + "step": 30772 + }, + { + "epoch": 0.6773717807616113, + "grad_norm": 1.4735901521234531, + "learning_rate": 2.4901281051417335e-06, + "loss": 0.292, + "step": 30773 + }, + { + "epoch": 0.6773937926480299, + "grad_norm": 1.7122003203808742, + "learning_rate": 2.489819818638931e-06, + "loss": 0.2677, + "step": 30774 + }, + { + "epoch": 0.6774158045344486, + "grad_norm": 1.5676758878332149, + "learning_rate": 2.4895115448939797e-06, + "loss": 0.3176, + "step": 30775 + }, + { + "epoch": 0.6774378164208673, + "grad_norm": 2.1442758169309735, + "learning_rate": 2.489203283908454e-06, + "loss": 0.269, + "step": 30776 + }, + { + "epoch": 0.6774598283072859, + "grad_norm": 0.6694246786699715, + "learning_rate": 2.4888950356839176e-06, + "loss": 0.4685, + "step": 30777 + }, + { + "epoch": 0.6774818401937046, + "grad_norm": 1.7918760059083485, + "learning_rate": 2.4885868002219342e-06, + "loss": 0.3473, + "step": 30778 + }, + { + "epoch": 0.6775038520801233, + "grad_norm": 1.4800747298893795, + "learning_rate": 2.4882785775240746e-06, + "loss": 0.3156, + "step": 30779 + }, + { + "epoch": 0.6775258639665419, + "grad_norm": 1.8807303022247566, + "learning_rate": 2.487970367591902e-06, + "loss": 0.2673, + "step": 30780 + }, + { + "epoch": 0.6775478758529606, + "grad_norm": 0.623336223187167, + "learning_rate": 2.4876621704269843e-06, + "loss": 0.4408, + "step": 30781 + }, + { + "epoch": 0.6775698877393793, + "grad_norm": 1.7501711200261933, + "learning_rate": 2.48735398603089e-06, + "loss": 0.2585, + "step": 30782 + }, + { + "epoch": 0.6775918996257979, + "grad_norm": 1.4313959396287772, + "learning_rate": 2.487045814405181e-06, + "loss": 0.3016, + "step": 30783 + }, + { + "epoch": 0.6776139115122166, + "grad_norm": 1.8164194488021845, + "learning_rate": 2.4867376555514288e-06, + "loss": 0.3207, + "step": 30784 + }, + { + "epoch": 0.6776359233986352, + "grad_norm": 1.8179159358750077, + "learning_rate": 2.4864295094711945e-06, + "loss": 0.28, + "step": 30785 + }, + { + "epoch": 0.6776579352850539, + "grad_norm": 0.614080062504476, + "learning_rate": 2.4861213761660462e-06, + "loss": 0.4752, + "step": 30786 + }, + { + "epoch": 0.6776799471714726, + "grad_norm": 0.6116414184670741, + "learning_rate": 2.4858132556375527e-06, + "loss": 0.4524, + "step": 30787 + }, + { + "epoch": 0.6777019590578912, + "grad_norm": 1.2967939124072956, + "learning_rate": 2.485505147887275e-06, + "loss": 0.2308, + "step": 30788 + }, + { + "epoch": 0.6777239709443099, + "grad_norm": 1.4926348702576815, + "learning_rate": 2.485197052916784e-06, + "loss": 0.3143, + "step": 30789 + }, + { + "epoch": 0.6777459828307286, + "grad_norm": 1.629444206278157, + "learning_rate": 2.484888970727641e-06, + "loss": 0.3775, + "step": 30790 + }, + { + "epoch": 0.6777679947171472, + "grad_norm": 0.622238899718695, + "learning_rate": 2.484580901321414e-06, + "loss": 0.4557, + "step": 30791 + }, + { + "epoch": 0.6777900066035659, + "grad_norm": 1.509325946959715, + "learning_rate": 2.484272844699671e-06, + "loss": 0.2987, + "step": 30792 + }, + { + "epoch": 0.6778120184899846, + "grad_norm": 1.7803021063254538, + "learning_rate": 2.4839648008639732e-06, + "loss": 0.3236, + "step": 30793 + }, + { + "epoch": 0.6778340303764032, + "grad_norm": 1.5721979732075162, + "learning_rate": 2.4836567698158907e-06, + "loss": 0.2502, + "step": 30794 + }, + { + "epoch": 0.6778560422628219, + "grad_norm": 2.364440318687832, + "learning_rate": 2.4833487515569847e-06, + "loss": 0.3804, + "step": 30795 + }, + { + "epoch": 0.6778780541492406, + "grad_norm": 2.23094034810898, + "learning_rate": 2.4830407460888232e-06, + "loss": 0.3169, + "step": 30796 + }, + { + "epoch": 0.6779000660356592, + "grad_norm": 1.391225594653176, + "learning_rate": 2.482732753412973e-06, + "loss": 0.2163, + "step": 30797 + }, + { + "epoch": 0.6779220779220779, + "grad_norm": 1.4098450637840327, + "learning_rate": 2.4824247735309953e-06, + "loss": 0.3122, + "step": 30798 + }, + { + "epoch": 0.6779440898084966, + "grad_norm": 0.6233750170987705, + "learning_rate": 2.48211680644446e-06, + "loss": 0.4698, + "step": 30799 + }, + { + "epoch": 0.6779661016949152, + "grad_norm": 1.5731523741321094, + "learning_rate": 2.4818088521549283e-06, + "loss": 0.2851, + "step": 30800 + }, + { + "epoch": 0.6779881135813339, + "grad_norm": 1.6677746364857935, + "learning_rate": 2.481500910663967e-06, + "loss": 0.2661, + "step": 30801 + }, + { + "epoch": 0.6780101254677526, + "grad_norm": 1.4873890813725, + "learning_rate": 2.4811929819731434e-06, + "loss": 0.4292, + "step": 30802 + }, + { + "epoch": 0.6780321373541712, + "grad_norm": 2.4803066854020086, + "learning_rate": 2.4808850660840185e-06, + "loss": 0.3921, + "step": 30803 + }, + { + "epoch": 0.6780541492405899, + "grad_norm": 1.8118825262991431, + "learning_rate": 2.4805771629981614e-06, + "loss": 0.2318, + "step": 30804 + }, + { + "epoch": 0.6780761611270086, + "grad_norm": 1.734482350512474, + "learning_rate": 2.4802692727171342e-06, + "loss": 0.269, + "step": 30805 + }, + { + "epoch": 0.6780981730134272, + "grad_norm": 1.275067638305317, + "learning_rate": 2.479961395242499e-06, + "loss": 0.2165, + "step": 30806 + }, + { + "epoch": 0.6781201848998459, + "grad_norm": 3.2314384769873667, + "learning_rate": 2.4796535305758278e-06, + "loss": 0.2715, + "step": 30807 + }, + { + "epoch": 0.6781421967862645, + "grad_norm": 1.4695520794231345, + "learning_rate": 2.4793456787186816e-06, + "loss": 0.3331, + "step": 30808 + }, + { + "epoch": 0.6781642086726832, + "grad_norm": 1.6961378123841429, + "learning_rate": 2.479037839672622e-06, + "loss": 0.3464, + "step": 30809 + }, + { + "epoch": 0.6781862205591019, + "grad_norm": 1.5340691659587489, + "learning_rate": 2.4787300134392196e-06, + "loss": 0.2822, + "step": 30810 + }, + { + "epoch": 0.6782082324455205, + "grad_norm": 1.2516678121609626, + "learning_rate": 2.4784222000200314e-06, + "loss": 0.3343, + "step": 30811 + }, + { + "epoch": 0.6782302443319392, + "grad_norm": 1.7221843964393733, + "learning_rate": 2.478114399416631e-06, + "loss": 0.3253, + "step": 30812 + }, + { + "epoch": 0.6782522562183579, + "grad_norm": 1.2822338203005192, + "learning_rate": 2.4778066116305764e-06, + "loss": 0.3329, + "step": 30813 + }, + { + "epoch": 0.6782742681047765, + "grad_norm": 1.4314341865859506, + "learning_rate": 2.4774988366634326e-06, + "loss": 0.3066, + "step": 30814 + }, + { + "epoch": 0.6782962799911952, + "grad_norm": 1.8530004752533071, + "learning_rate": 2.4771910745167665e-06, + "loss": 0.3614, + "step": 30815 + }, + { + "epoch": 0.6783182918776139, + "grad_norm": 2.2367916289166336, + "learning_rate": 2.4768833251921388e-06, + "loss": 0.2779, + "step": 30816 + }, + { + "epoch": 0.6783403037640325, + "grad_norm": 1.7422328487848153, + "learning_rate": 2.4765755886911155e-06, + "loss": 0.3085, + "step": 30817 + }, + { + "epoch": 0.6783623156504512, + "grad_norm": 1.5821306969816047, + "learning_rate": 2.4762678650152624e-06, + "loss": 0.292, + "step": 30818 + }, + { + "epoch": 0.67838432753687, + "grad_norm": 2.662282986696146, + "learning_rate": 2.4759601541661393e-06, + "loss": 0.2499, + "step": 30819 + }, + { + "epoch": 0.6784063394232885, + "grad_norm": 1.6340565495956818, + "learning_rate": 2.475652456145314e-06, + "loss": 0.3495, + "step": 30820 + }, + { + "epoch": 0.6784283513097072, + "grad_norm": 1.5150289279933937, + "learning_rate": 2.4753447709543476e-06, + "loss": 0.3191, + "step": 30821 + }, + { + "epoch": 0.678450363196126, + "grad_norm": 1.918839408108855, + "learning_rate": 2.4750370985948053e-06, + "loss": 0.3786, + "step": 30822 + }, + { + "epoch": 0.6784723750825445, + "grad_norm": 1.406382720123863, + "learning_rate": 2.4747294390682518e-06, + "loss": 0.3338, + "step": 30823 + }, + { + "epoch": 0.6784943869689632, + "grad_norm": 1.4507887616129997, + "learning_rate": 2.474421792376249e-06, + "loss": 0.2179, + "step": 30824 + }, + { + "epoch": 0.678516398855382, + "grad_norm": 0.6376745216156823, + "learning_rate": 2.4741141585203616e-06, + "loss": 0.4325, + "step": 30825 + }, + { + "epoch": 0.6785384107418005, + "grad_norm": 1.6102958608562328, + "learning_rate": 2.4738065375021514e-06, + "loss": 0.3319, + "step": 30826 + }, + { + "epoch": 0.6785604226282193, + "grad_norm": 1.3696520052012313, + "learning_rate": 2.4734989293231824e-06, + "loss": 0.265, + "step": 30827 + }, + { + "epoch": 0.678582434514638, + "grad_norm": 1.4054668734783269, + "learning_rate": 2.4731913339850215e-06, + "loss": 0.345, + "step": 30828 + }, + { + "epoch": 0.6786044464010565, + "grad_norm": 0.6430720630382171, + "learning_rate": 2.472883751489227e-06, + "loss": 0.4772, + "step": 30829 + }, + { + "epoch": 0.6786264582874753, + "grad_norm": 0.6178913648722744, + "learning_rate": 2.4725761818373663e-06, + "loss": 0.4704, + "step": 30830 + }, + { + "epoch": 0.6786484701738938, + "grad_norm": 1.7104347719819015, + "learning_rate": 2.4722686250309985e-06, + "loss": 0.2263, + "step": 30831 + }, + { + "epoch": 0.6786704820603126, + "grad_norm": 1.4035464042503842, + "learning_rate": 2.4719610810716894e-06, + "loss": 0.2661, + "step": 30832 + }, + { + "epoch": 0.6786924939467313, + "grad_norm": 0.6410462139157919, + "learning_rate": 2.4716535499610033e-06, + "loss": 0.4791, + "step": 30833 + }, + { + "epoch": 0.6787145058331499, + "grad_norm": 1.7972972986931255, + "learning_rate": 2.4713460317004994e-06, + "loss": 0.2636, + "step": 30834 + }, + { + "epoch": 0.6787365177195686, + "grad_norm": 1.188333598981781, + "learning_rate": 2.471038526291745e-06, + "loss": 0.2407, + "step": 30835 + }, + { + "epoch": 0.6787585296059873, + "grad_norm": 1.7158221274141454, + "learning_rate": 2.4707310337362988e-06, + "loss": 0.2725, + "step": 30836 + }, + { + "epoch": 0.6787805414924059, + "grad_norm": 1.7849032034485355, + "learning_rate": 2.470423554035726e-06, + "loss": 0.3256, + "step": 30837 + }, + { + "epoch": 0.6788025533788246, + "grad_norm": 3.2571091987356646, + "learning_rate": 2.47011608719159e-06, + "loss": 0.2954, + "step": 30838 + }, + { + "epoch": 0.6788245652652433, + "grad_norm": 1.5970417225441058, + "learning_rate": 2.469808633205451e-06, + "loss": 0.2831, + "step": 30839 + }, + { + "epoch": 0.6788465771516619, + "grad_norm": 1.1615886502455872, + "learning_rate": 2.4695011920788748e-06, + "loss": 0.2954, + "step": 30840 + }, + { + "epoch": 0.6788685890380806, + "grad_norm": 1.675516630305015, + "learning_rate": 2.469193763813422e-06, + "loss": 0.2898, + "step": 30841 + }, + { + "epoch": 0.6788906009244993, + "grad_norm": 1.8573460464028688, + "learning_rate": 2.468886348410652e-06, + "loss": 0.2222, + "step": 30842 + }, + { + "epoch": 0.6789126128109179, + "grad_norm": 2.4336943240158697, + "learning_rate": 2.4685789458721336e-06, + "loss": 0.3048, + "step": 30843 + }, + { + "epoch": 0.6789346246973366, + "grad_norm": 1.614776961488804, + "learning_rate": 2.468271556199424e-06, + "loss": 0.322, + "step": 30844 + }, + { + "epoch": 0.6789566365837553, + "grad_norm": 1.621394951310376, + "learning_rate": 2.4679641793940902e-06, + "loss": 0.2615, + "step": 30845 + }, + { + "epoch": 0.6789786484701739, + "grad_norm": 1.5735697636460928, + "learning_rate": 2.467656815457692e-06, + "loss": 0.2865, + "step": 30846 + }, + { + "epoch": 0.6790006603565926, + "grad_norm": 1.222033048851069, + "learning_rate": 2.4673494643917882e-06, + "loss": 0.2784, + "step": 30847 + }, + { + "epoch": 0.6790226722430113, + "grad_norm": 1.6714697078321106, + "learning_rate": 2.467042126197945e-06, + "loss": 0.3191, + "step": 30848 + }, + { + "epoch": 0.6790446841294299, + "grad_norm": 1.6158606498280794, + "learning_rate": 2.466734800877725e-06, + "loss": 0.2989, + "step": 30849 + }, + { + "epoch": 0.6790666960158486, + "grad_norm": 1.4133653710842122, + "learning_rate": 2.4664274884326864e-06, + "loss": 0.201, + "step": 30850 + }, + { + "epoch": 0.6790887079022673, + "grad_norm": 0.617474027225562, + "learning_rate": 2.466120188864396e-06, + "loss": 0.4621, + "step": 30851 + }, + { + "epoch": 0.6791107197886859, + "grad_norm": 1.6622210654158964, + "learning_rate": 2.46581290217441e-06, + "loss": 0.3217, + "step": 30852 + }, + { + "epoch": 0.6791327316751046, + "grad_norm": 0.6068322041792529, + "learning_rate": 2.465505628364294e-06, + "loss": 0.45, + "step": 30853 + }, + { + "epoch": 0.6791547435615233, + "grad_norm": 1.4787209727877835, + "learning_rate": 2.4651983674356107e-06, + "loss": 0.3126, + "step": 30854 + }, + { + "epoch": 0.6791767554479419, + "grad_norm": 2.927512726284617, + "learning_rate": 2.4648911193899176e-06, + "loss": 0.3814, + "step": 30855 + }, + { + "epoch": 0.6791987673343606, + "grad_norm": 1.7157756608125097, + "learning_rate": 2.4645838842287807e-06, + "loss": 0.3041, + "step": 30856 + }, + { + "epoch": 0.6792207792207792, + "grad_norm": 0.6155098461852692, + "learning_rate": 2.464276661953757e-06, + "loss": 0.4507, + "step": 30857 + }, + { + "epoch": 0.6792427911071979, + "grad_norm": 1.4149784703682087, + "learning_rate": 2.4639694525664103e-06, + "loss": 0.3488, + "step": 30858 + }, + { + "epoch": 0.6792648029936166, + "grad_norm": 1.583984579082334, + "learning_rate": 2.4636622560683044e-06, + "loss": 0.3347, + "step": 30859 + }, + { + "epoch": 0.6792868148800352, + "grad_norm": 1.7848850670797192, + "learning_rate": 2.4633550724609956e-06, + "loss": 0.2743, + "step": 30860 + }, + { + "epoch": 0.6793088267664539, + "grad_norm": 1.313807258559246, + "learning_rate": 2.46304790174605e-06, + "loss": 0.3216, + "step": 30861 + }, + { + "epoch": 0.6793308386528726, + "grad_norm": 0.6303488788872393, + "learning_rate": 2.4627407439250246e-06, + "loss": 0.4745, + "step": 30862 + }, + { + "epoch": 0.6793528505392912, + "grad_norm": 1.5827126688827742, + "learning_rate": 2.4624335989994824e-06, + "loss": 0.2778, + "step": 30863 + }, + { + "epoch": 0.6793748624257099, + "grad_norm": 1.736974698037456, + "learning_rate": 2.462126466970986e-06, + "loss": 0.276, + "step": 30864 + }, + { + "epoch": 0.6793968743121286, + "grad_norm": 1.319397093264422, + "learning_rate": 2.461819347841093e-06, + "loss": 0.207, + "step": 30865 + }, + { + "epoch": 0.6794188861985472, + "grad_norm": 2.756116982108267, + "learning_rate": 2.4615122416113684e-06, + "loss": 0.2711, + "step": 30866 + }, + { + "epoch": 0.6794408980849659, + "grad_norm": 1.667547982108753, + "learning_rate": 2.4612051482833684e-06, + "loss": 0.3093, + "step": 30867 + }, + { + "epoch": 0.6794629099713846, + "grad_norm": 0.6026906512159143, + "learning_rate": 2.4608980678586562e-06, + "loss": 0.433, + "step": 30868 + }, + { + "epoch": 0.6794849218578032, + "grad_norm": 0.615836210240642, + "learning_rate": 2.460591000338795e-06, + "loss": 0.4602, + "step": 30869 + }, + { + "epoch": 0.6795069337442219, + "grad_norm": 1.1122714647941876, + "learning_rate": 2.4602839457253402e-06, + "loss": 0.2154, + "step": 30870 + }, + { + "epoch": 0.6795289456306406, + "grad_norm": 1.6758200910623482, + "learning_rate": 2.459976904019858e-06, + "loss": 0.2766, + "step": 30871 + }, + { + "epoch": 0.6795509575170592, + "grad_norm": 1.7952922792458499, + "learning_rate": 2.4596698752239034e-06, + "loss": 0.2075, + "step": 30872 + }, + { + "epoch": 0.6795729694034779, + "grad_norm": 1.7877597028940748, + "learning_rate": 2.4593628593390395e-06, + "loss": 0.2638, + "step": 30873 + }, + { + "epoch": 0.6795949812898966, + "grad_norm": 1.399941798724802, + "learning_rate": 2.4590558563668295e-06, + "loss": 0.3042, + "step": 30874 + }, + { + "epoch": 0.6796169931763152, + "grad_norm": 0.6152385858526962, + "learning_rate": 2.4587488663088277e-06, + "loss": 0.4573, + "step": 30875 + }, + { + "epoch": 0.6796390050627339, + "grad_norm": 1.349105665471124, + "learning_rate": 2.4584418891666006e-06, + "loss": 0.2964, + "step": 30876 + }, + { + "epoch": 0.6796610169491526, + "grad_norm": 1.7141355758243244, + "learning_rate": 2.458134924941705e-06, + "loss": 0.3414, + "step": 30877 + }, + { + "epoch": 0.6796830288355712, + "grad_norm": 1.3430802334764962, + "learning_rate": 2.4578279736356975e-06, + "loss": 0.2701, + "step": 30878 + }, + { + "epoch": 0.6797050407219899, + "grad_norm": 1.4956056319639552, + "learning_rate": 2.4575210352501457e-06, + "loss": 0.3459, + "step": 30879 + }, + { + "epoch": 0.6797270526084085, + "grad_norm": 1.4794637300215105, + "learning_rate": 2.457214109786604e-06, + "loss": 0.3144, + "step": 30880 + }, + { + "epoch": 0.6797490644948272, + "grad_norm": 2.625330277243976, + "learning_rate": 2.456907197246637e-06, + "loss": 0.2502, + "step": 30881 + }, + { + "epoch": 0.6797710763812459, + "grad_norm": 1.443476180764302, + "learning_rate": 2.4566002976318015e-06, + "loss": 0.2993, + "step": 30882 + }, + { + "epoch": 0.6797930882676645, + "grad_norm": 0.6426347385728328, + "learning_rate": 2.4562934109436554e-06, + "loss": 0.4601, + "step": 30883 + }, + { + "epoch": 0.6798151001540832, + "grad_norm": 6.504528191145506, + "learning_rate": 2.455986537183761e-06, + "loss": 0.3739, + "step": 30884 + }, + { + "epoch": 0.6798371120405019, + "grad_norm": 1.7212541236297632, + "learning_rate": 2.4556796763536795e-06, + "loss": 0.3759, + "step": 30885 + }, + { + "epoch": 0.6798591239269205, + "grad_norm": 1.2553466497181869, + "learning_rate": 2.455372828454967e-06, + "loss": 0.2788, + "step": 30886 + }, + { + "epoch": 0.6798811358133392, + "grad_norm": 2.6272920069324024, + "learning_rate": 2.4550659934891863e-06, + "loss": 0.2574, + "step": 30887 + }, + { + "epoch": 0.6799031476997579, + "grad_norm": 1.652540904047204, + "learning_rate": 2.4547591714578937e-06, + "loss": 0.3719, + "step": 30888 + }, + { + "epoch": 0.6799251595861765, + "grad_norm": 1.857020786036755, + "learning_rate": 2.4544523623626497e-06, + "loss": 0.3484, + "step": 30889 + }, + { + "epoch": 0.6799471714725952, + "grad_norm": 1.5765331925579356, + "learning_rate": 2.454145566205017e-06, + "loss": 0.2766, + "step": 30890 + }, + { + "epoch": 0.6799691833590139, + "grad_norm": 1.4886057482078037, + "learning_rate": 2.453838782986549e-06, + "loss": 0.3834, + "step": 30891 + }, + { + "epoch": 0.6799911952454325, + "grad_norm": 1.4023512687061428, + "learning_rate": 2.4535320127088107e-06, + "loss": 0.2938, + "step": 30892 + }, + { + "epoch": 0.6800132071318512, + "grad_norm": 1.9160926304937795, + "learning_rate": 2.453225255373356e-06, + "loss": 0.2729, + "step": 30893 + }, + { + "epoch": 0.6800352190182699, + "grad_norm": 1.7487882008393176, + "learning_rate": 2.4529185109817466e-06, + "loss": 0.3217, + "step": 30894 + }, + { + "epoch": 0.6800572309046885, + "grad_norm": 1.7797733583403061, + "learning_rate": 2.4526117795355434e-06, + "loss": 0.2712, + "step": 30895 + }, + { + "epoch": 0.6800792427911072, + "grad_norm": 1.7497283686348977, + "learning_rate": 2.4523050610363016e-06, + "loss": 0.3619, + "step": 30896 + }, + { + "epoch": 0.6801012546775259, + "grad_norm": 1.4324966690984344, + "learning_rate": 2.451998355485583e-06, + "loss": 0.2679, + "step": 30897 + }, + { + "epoch": 0.6801232665639445, + "grad_norm": 1.3443649648275469, + "learning_rate": 2.4516916628849442e-06, + "loss": 0.3027, + "step": 30898 + }, + { + "epoch": 0.6801452784503632, + "grad_norm": 1.4100020856646953, + "learning_rate": 2.451384983235944e-06, + "loss": 0.3299, + "step": 30899 + }, + { + "epoch": 0.6801672903367819, + "grad_norm": 1.5106490818648175, + "learning_rate": 2.4510783165401447e-06, + "loss": 0.233, + "step": 30900 + }, + { + "epoch": 0.6801893022232005, + "grad_norm": 1.497843116183158, + "learning_rate": 2.4507716627990996e-06, + "loss": 0.2644, + "step": 30901 + }, + { + "epoch": 0.6802113141096192, + "grad_norm": 1.8738899468468142, + "learning_rate": 2.450465022014372e-06, + "loss": 0.3133, + "step": 30902 + }, + { + "epoch": 0.6802333259960378, + "grad_norm": 0.6209510833418677, + "learning_rate": 2.450158394187517e-06, + "loss": 0.4684, + "step": 30903 + }, + { + "epoch": 0.6802553378824565, + "grad_norm": 2.2047393515930533, + "learning_rate": 2.449851779320094e-06, + "loss": 0.3286, + "step": 30904 + }, + { + "epoch": 0.6802773497688752, + "grad_norm": 1.2126825252682678, + "learning_rate": 2.4495451774136637e-06, + "loss": 0.2795, + "step": 30905 + }, + { + "epoch": 0.6802993616552938, + "grad_norm": 2.030912652386874, + "learning_rate": 2.44923858846978e-06, + "loss": 0.3225, + "step": 30906 + }, + { + "epoch": 0.6803213735417125, + "grad_norm": 1.2024840508401193, + "learning_rate": 2.4489320124900056e-06, + "loss": 0.214, + "step": 30907 + }, + { + "epoch": 0.6803433854281312, + "grad_norm": 1.7813905857357835, + "learning_rate": 2.4486254494758945e-06, + "loss": 0.4425, + "step": 30908 + }, + { + "epoch": 0.6803653973145498, + "grad_norm": 1.3802971933251416, + "learning_rate": 2.448318899429007e-06, + "loss": 0.2912, + "step": 30909 + }, + { + "epoch": 0.6803874092009685, + "grad_norm": 2.2714117823983115, + "learning_rate": 2.4480123623509027e-06, + "loss": 0.3081, + "step": 30910 + }, + { + "epoch": 0.6804094210873872, + "grad_norm": 2.4550669299494836, + "learning_rate": 2.447705838243136e-06, + "loss": 0.2645, + "step": 30911 + }, + { + "epoch": 0.6804314329738058, + "grad_norm": 2.439017668024678, + "learning_rate": 2.447399327107268e-06, + "loss": 0.3393, + "step": 30912 + }, + { + "epoch": 0.6804534448602245, + "grad_norm": 1.6163412902763568, + "learning_rate": 2.447092828944856e-06, + "loss": 0.2874, + "step": 30913 + }, + { + "epoch": 0.6804754567466432, + "grad_norm": 1.4050393429961796, + "learning_rate": 2.4467863437574526e-06, + "loss": 0.298, + "step": 30914 + }, + { + "epoch": 0.6804974686330618, + "grad_norm": 1.7642970640579392, + "learning_rate": 2.446479871546623e-06, + "loss": 0.3868, + "step": 30915 + }, + { + "epoch": 0.6805194805194805, + "grad_norm": 0.6470881976702604, + "learning_rate": 2.4461734123139202e-06, + "loss": 0.4643, + "step": 30916 + }, + { + "epoch": 0.6805414924058992, + "grad_norm": 1.3302615633785384, + "learning_rate": 2.445866966060905e-06, + "loss": 0.2765, + "step": 30917 + }, + { + "epoch": 0.6805635042923178, + "grad_norm": 1.4706956667902855, + "learning_rate": 2.4455605327891335e-06, + "loss": 0.3076, + "step": 30918 + }, + { + "epoch": 0.6805855161787365, + "grad_norm": 1.4689751652336462, + "learning_rate": 2.4452541125001604e-06, + "loss": 0.2889, + "step": 30919 + }, + { + "epoch": 0.6806075280651552, + "grad_norm": 1.4692366481028254, + "learning_rate": 2.4449477051955454e-06, + "loss": 0.219, + "step": 30920 + }, + { + "epoch": 0.6806295399515738, + "grad_norm": 1.5834767791393412, + "learning_rate": 2.444641310876846e-06, + "loss": 0.2559, + "step": 30921 + }, + { + "epoch": 0.6806515518379925, + "grad_norm": 1.836747772054224, + "learning_rate": 2.444334929545621e-06, + "loss": 0.3833, + "step": 30922 + }, + { + "epoch": 0.6806735637244112, + "grad_norm": 1.5790897203552985, + "learning_rate": 2.4440285612034255e-06, + "loss": 0.3496, + "step": 30923 + }, + { + "epoch": 0.6806955756108298, + "grad_norm": 1.3300491872789173, + "learning_rate": 2.4437222058518144e-06, + "loss": 0.3176, + "step": 30924 + }, + { + "epoch": 0.6807175874972485, + "grad_norm": 0.6330946759021862, + "learning_rate": 2.4434158634923483e-06, + "loss": 0.4854, + "step": 30925 + }, + { + "epoch": 0.6807395993836671, + "grad_norm": 1.7524795538665692, + "learning_rate": 2.4431095341265844e-06, + "loss": 0.2492, + "step": 30926 + }, + { + "epoch": 0.6807616112700858, + "grad_norm": 2.4898776396595643, + "learning_rate": 2.4428032177560763e-06, + "loss": 0.3133, + "step": 30927 + }, + { + "epoch": 0.6807836231565045, + "grad_norm": 1.6309368976368739, + "learning_rate": 2.4424969143823845e-06, + "loss": 0.2815, + "step": 30928 + }, + { + "epoch": 0.6808056350429231, + "grad_norm": 1.0979121731390342, + "learning_rate": 2.4421906240070625e-06, + "loss": 0.3008, + "step": 30929 + }, + { + "epoch": 0.6808276469293418, + "grad_norm": 1.7038961382758442, + "learning_rate": 2.441884346631669e-06, + "loss": 0.292, + "step": 30930 + }, + { + "epoch": 0.6808496588157605, + "grad_norm": 5.270854990322882, + "learning_rate": 2.441578082257761e-06, + "loss": 0.2695, + "step": 30931 + }, + { + "epoch": 0.6808716707021791, + "grad_norm": 1.6705687333487116, + "learning_rate": 2.4412718308868934e-06, + "loss": 0.2581, + "step": 30932 + }, + { + "epoch": 0.6808936825885978, + "grad_norm": 1.3749435396549048, + "learning_rate": 2.440965592520625e-06, + "loss": 0.2783, + "step": 30933 + }, + { + "epoch": 0.6809156944750165, + "grad_norm": 1.3119180625387932, + "learning_rate": 2.440659367160509e-06, + "loss": 0.2334, + "step": 30934 + }, + { + "epoch": 0.6809377063614351, + "grad_norm": 1.1199731294241617, + "learning_rate": 2.4403531548081034e-06, + "loss": 0.2869, + "step": 30935 + }, + { + "epoch": 0.6809597182478538, + "grad_norm": 2.4215271943136876, + "learning_rate": 2.440046955464967e-06, + "loss": 0.2768, + "step": 30936 + }, + { + "epoch": 0.6809817301342725, + "grad_norm": 1.315420338748682, + "learning_rate": 2.4397407691326514e-06, + "loss": 0.2924, + "step": 30937 + }, + { + "epoch": 0.6810037420206911, + "grad_norm": 3.350604688128938, + "learning_rate": 2.4394345958127176e-06, + "loss": 0.4498, + "step": 30938 + }, + { + "epoch": 0.6810257539071098, + "grad_norm": 1.676240494474285, + "learning_rate": 2.4391284355067167e-06, + "loss": 0.3002, + "step": 30939 + }, + { + "epoch": 0.6810477657935285, + "grad_norm": 1.4980180409768271, + "learning_rate": 2.4388222882162075e-06, + "loss": 0.2121, + "step": 30940 + }, + { + "epoch": 0.6810697776799471, + "grad_norm": 0.6011540829839885, + "learning_rate": 2.4385161539427477e-06, + "loss": 0.4602, + "step": 30941 + }, + { + "epoch": 0.6810917895663658, + "grad_norm": 1.9057448426474353, + "learning_rate": 2.438210032687889e-06, + "loss": 0.3036, + "step": 30942 + }, + { + "epoch": 0.6811138014527846, + "grad_norm": 0.6253533911108468, + "learning_rate": 2.437903924453192e-06, + "loss": 0.4436, + "step": 30943 + }, + { + "epoch": 0.6811358133392031, + "grad_norm": 3.826552780035336, + "learning_rate": 2.437597829240208e-06, + "loss": 0.3005, + "step": 30944 + }, + { + "epoch": 0.6811578252256218, + "grad_norm": 0.6225448567039469, + "learning_rate": 2.4372917470504948e-06, + "loss": 0.443, + "step": 30945 + }, + { + "epoch": 0.6811798371120406, + "grad_norm": 0.6139513891466986, + "learning_rate": 2.4369856778856095e-06, + "loss": 0.4361, + "step": 30946 + }, + { + "epoch": 0.6812018489984591, + "grad_norm": 1.4939140767624182, + "learning_rate": 2.436679621747104e-06, + "loss": 0.3026, + "step": 30947 + }, + { + "epoch": 0.6812238608848779, + "grad_norm": 31.286422063482412, + "learning_rate": 2.4363735786365382e-06, + "loss": 0.3056, + "step": 30948 + }, + { + "epoch": 0.6812458727712964, + "grad_norm": 0.6167481309513432, + "learning_rate": 2.4360675485554654e-06, + "loss": 0.4582, + "step": 30949 + }, + { + "epoch": 0.6812678846577152, + "grad_norm": 2.2857766538976327, + "learning_rate": 2.4357615315054368e-06, + "loss": 0.3463, + "step": 30950 + }, + { + "epoch": 0.6812898965441339, + "grad_norm": 2.8414603253865423, + "learning_rate": 2.4354555274880154e-06, + "loss": 0.3502, + "step": 30951 + }, + { + "epoch": 0.6813119084305524, + "grad_norm": 0.622089924337865, + "learning_rate": 2.435149536504751e-06, + "loss": 0.4776, + "step": 30952 + }, + { + "epoch": 0.6813339203169712, + "grad_norm": 1.9465662741661949, + "learning_rate": 2.4348435585572035e-06, + "loss": 0.3784, + "step": 30953 + }, + { + "epoch": 0.6813559322033899, + "grad_norm": 1.7475465339076026, + "learning_rate": 2.434537593646924e-06, + "loss": 0.2741, + "step": 30954 + }, + { + "epoch": 0.6813779440898085, + "grad_norm": 1.6195191631772394, + "learning_rate": 2.434231641775467e-06, + "loss": 0.2849, + "step": 30955 + }, + { + "epoch": 0.6813999559762272, + "grad_norm": 2.0147695751485237, + "learning_rate": 2.433925702944389e-06, + "loss": 0.3801, + "step": 30956 + }, + { + "epoch": 0.6814219678626459, + "grad_norm": 1.567416358806643, + "learning_rate": 2.4336197771552454e-06, + "loss": 0.3434, + "step": 30957 + }, + { + "epoch": 0.6814439797490645, + "grad_norm": 2.1120534866105998, + "learning_rate": 2.433313864409593e-06, + "loss": 0.3449, + "step": 30958 + }, + { + "epoch": 0.6814659916354832, + "grad_norm": 1.513454470161476, + "learning_rate": 2.433007964708984e-06, + "loss": 0.279, + "step": 30959 + }, + { + "epoch": 0.6814880035219019, + "grad_norm": 0.6020809951982935, + "learning_rate": 2.432702078054971e-06, + "loss": 0.4415, + "step": 30960 + }, + { + "epoch": 0.6815100154083205, + "grad_norm": 1.8980048836339718, + "learning_rate": 2.4323962044491117e-06, + "loss": 0.3242, + "step": 30961 + }, + { + "epoch": 0.6815320272947392, + "grad_norm": 1.4155576660819793, + "learning_rate": 2.43209034389296e-06, + "loss": 0.3226, + "step": 30962 + }, + { + "epoch": 0.6815540391811579, + "grad_norm": 1.4077502656124656, + "learning_rate": 2.431784496388072e-06, + "loss": 0.319, + "step": 30963 + }, + { + "epoch": 0.6815760510675765, + "grad_norm": 1.7335683256356533, + "learning_rate": 2.431478661936001e-06, + "loss": 0.3526, + "step": 30964 + }, + { + "epoch": 0.6815980629539952, + "grad_norm": 1.7087850430173317, + "learning_rate": 2.431172840538299e-06, + "loss": 0.2751, + "step": 30965 + }, + { + "epoch": 0.6816200748404139, + "grad_norm": 1.4893111097104572, + "learning_rate": 2.4308670321965224e-06, + "loss": 0.2593, + "step": 30966 + }, + { + "epoch": 0.6816420867268325, + "grad_norm": 1.5649242795096527, + "learning_rate": 2.430561236912227e-06, + "loss": 0.2889, + "step": 30967 + }, + { + "epoch": 0.6816640986132512, + "grad_norm": 3.893281118760861, + "learning_rate": 2.4302554546869643e-06, + "loss": 0.4095, + "step": 30968 + }, + { + "epoch": 0.6816861104996699, + "grad_norm": 0.6383659911621183, + "learning_rate": 2.429949685522291e-06, + "loss": 0.4623, + "step": 30969 + }, + { + "epoch": 0.6817081223860885, + "grad_norm": 0.628998663845609, + "learning_rate": 2.4296439294197578e-06, + "loss": 0.4633, + "step": 30970 + }, + { + "epoch": 0.6817301342725072, + "grad_norm": 2.3529611991100157, + "learning_rate": 2.42933818638092e-06, + "loss": 0.3273, + "step": 30971 + }, + { + "epoch": 0.6817521461589258, + "grad_norm": 1.8302171926004611, + "learning_rate": 2.4290324564073347e-06, + "loss": 0.3441, + "step": 30972 + }, + { + "epoch": 0.6817741580453445, + "grad_norm": 1.172160136327841, + "learning_rate": 2.428726739500551e-06, + "loss": 0.2731, + "step": 30973 + }, + { + "epoch": 0.6817961699317632, + "grad_norm": 1.7338223006934084, + "learning_rate": 2.428421035662127e-06, + "loss": 0.3232, + "step": 30974 + }, + { + "epoch": 0.6818181818181818, + "grad_norm": 0.5966583442738468, + "learning_rate": 2.428115344893612e-06, + "loss": 0.4541, + "step": 30975 + }, + { + "epoch": 0.6818401937046005, + "grad_norm": 1.1378930950955797, + "learning_rate": 2.427809667196562e-06, + "loss": 0.2179, + "step": 30976 + }, + { + "epoch": 0.6818622055910192, + "grad_norm": 1.4349050804419266, + "learning_rate": 2.427504002572532e-06, + "loss": 0.2267, + "step": 30977 + }, + { + "epoch": 0.6818842174774378, + "grad_norm": 1.4210484910631689, + "learning_rate": 2.427198351023072e-06, + "loss": 0.2469, + "step": 30978 + }, + { + "epoch": 0.6819062293638565, + "grad_norm": 1.4020888706532664, + "learning_rate": 2.4268927125497403e-06, + "loss": 0.2398, + "step": 30979 + }, + { + "epoch": 0.6819282412502752, + "grad_norm": 1.6700273724923782, + "learning_rate": 2.4265870871540855e-06, + "loss": 0.3007, + "step": 30980 + }, + { + "epoch": 0.6819502531366938, + "grad_norm": 1.807458255254279, + "learning_rate": 2.426281474837662e-06, + "loss": 0.2678, + "step": 30981 + }, + { + "epoch": 0.6819722650231125, + "grad_norm": 1.1998065495650232, + "learning_rate": 2.425975875602027e-06, + "loss": 0.2674, + "step": 30982 + }, + { + "epoch": 0.6819942769095312, + "grad_norm": 1.6246382570083562, + "learning_rate": 2.4256702894487276e-06, + "loss": 0.2935, + "step": 30983 + }, + { + "epoch": 0.6820162887959498, + "grad_norm": 1.5116169072193175, + "learning_rate": 2.4253647163793225e-06, + "loss": 0.2935, + "step": 30984 + }, + { + "epoch": 0.6820383006823685, + "grad_norm": 1.7307945961828268, + "learning_rate": 2.4250591563953623e-06, + "loss": 0.3112, + "step": 30985 + }, + { + "epoch": 0.6820603125687872, + "grad_norm": 1.72797792107584, + "learning_rate": 2.4247536094983957e-06, + "loss": 0.2969, + "step": 30986 + }, + { + "epoch": 0.6820823244552058, + "grad_norm": 1.5263082519298174, + "learning_rate": 2.424448075689984e-06, + "loss": 0.2875, + "step": 30987 + }, + { + "epoch": 0.6821043363416245, + "grad_norm": 1.5479933149274931, + "learning_rate": 2.424142554971674e-06, + "loss": 0.2962, + "step": 30988 + }, + { + "epoch": 0.6821263482280432, + "grad_norm": 1.5392087798540433, + "learning_rate": 2.423837047345022e-06, + "loss": 0.2233, + "step": 30989 + }, + { + "epoch": 0.6821483601144618, + "grad_norm": 1.593419128694304, + "learning_rate": 2.42353155281158e-06, + "loss": 0.3121, + "step": 30990 + }, + { + "epoch": 0.6821703720008805, + "grad_norm": 1.51636251980816, + "learning_rate": 2.423226071372897e-06, + "loss": 0.3435, + "step": 30991 + }, + { + "epoch": 0.6821923838872992, + "grad_norm": 1.5039708918083021, + "learning_rate": 2.4229206030305288e-06, + "loss": 0.2878, + "step": 30992 + }, + { + "epoch": 0.6822143957737178, + "grad_norm": 2.7001057872665113, + "learning_rate": 2.422615147786027e-06, + "loss": 0.3728, + "step": 30993 + }, + { + "epoch": 0.6822364076601365, + "grad_norm": 1.5463613653710537, + "learning_rate": 2.422309705640947e-06, + "loss": 0.2731, + "step": 30994 + }, + { + "epoch": 0.6822584195465552, + "grad_norm": 1.263305944042829, + "learning_rate": 2.422004276596838e-06, + "loss": 0.2712, + "step": 30995 + }, + { + "epoch": 0.6822804314329738, + "grad_norm": 2.2353842977105063, + "learning_rate": 2.421698860655252e-06, + "loss": 0.3223, + "step": 30996 + }, + { + "epoch": 0.6823024433193925, + "grad_norm": 1.3818035929819512, + "learning_rate": 2.4213934578177422e-06, + "loss": 0.2876, + "step": 30997 + }, + { + "epoch": 0.6823244552058111, + "grad_norm": 1.7030415588495988, + "learning_rate": 2.421088068085861e-06, + "loss": 0.251, + "step": 30998 + }, + { + "epoch": 0.6823464670922298, + "grad_norm": 1.4757477710430513, + "learning_rate": 2.420782691461162e-06, + "loss": 0.3304, + "step": 30999 + }, + { + "epoch": 0.6823684789786485, + "grad_norm": 1.2657893687975703, + "learning_rate": 2.420477327945196e-06, + "loss": 0.2565, + "step": 31000 + }, + { + "epoch": 0.6823904908650671, + "grad_norm": 1.6954865273783635, + "learning_rate": 2.4201719775395126e-06, + "loss": 0.3545, + "step": 31001 + }, + { + "epoch": 0.6824125027514858, + "grad_norm": 2.11834273950701, + "learning_rate": 2.4198666402456663e-06, + "loss": 0.3312, + "step": 31002 + }, + { + "epoch": 0.6824345146379045, + "grad_norm": 2.8432149889740836, + "learning_rate": 2.4195613160652084e-06, + "loss": 0.3968, + "step": 31003 + }, + { + "epoch": 0.6824565265243231, + "grad_norm": 3.453403137819337, + "learning_rate": 2.4192560049996924e-06, + "loss": 0.372, + "step": 31004 + }, + { + "epoch": 0.6824785384107418, + "grad_norm": 1.9973623231222462, + "learning_rate": 2.418950707050668e-06, + "loss": 0.2853, + "step": 31005 + }, + { + "epoch": 0.6825005502971605, + "grad_norm": 2.0737662435671753, + "learning_rate": 2.418645422219686e-06, + "loss": 0.3194, + "step": 31006 + }, + { + "epoch": 0.6825225621835791, + "grad_norm": 1.5879760945559194, + "learning_rate": 2.4183401505082993e-06, + "loss": 0.2838, + "step": 31007 + }, + { + "epoch": 0.6825445740699978, + "grad_norm": 0.6147506987083988, + "learning_rate": 2.4180348919180613e-06, + "loss": 0.4477, + "step": 31008 + }, + { + "epoch": 0.6825665859564165, + "grad_norm": 1.5415061798328626, + "learning_rate": 2.4177296464505194e-06, + "loss": 0.2642, + "step": 31009 + }, + { + "epoch": 0.6825885978428351, + "grad_norm": 1.3796773689074142, + "learning_rate": 2.417424414107229e-06, + "loss": 0.2632, + "step": 31010 + }, + { + "epoch": 0.6826106097292538, + "grad_norm": 1.515420593270071, + "learning_rate": 2.4171191948897383e-06, + "loss": 0.2643, + "step": 31011 + }, + { + "epoch": 0.6826326216156725, + "grad_norm": 2.439202514425434, + "learning_rate": 2.4168139887995993e-06, + "loss": 0.3282, + "step": 31012 + }, + { + "epoch": 0.6826546335020911, + "grad_norm": 0.6208100345311871, + "learning_rate": 2.416508795838366e-06, + "loss": 0.4559, + "step": 31013 + }, + { + "epoch": 0.6826766453885098, + "grad_norm": 2.3833669309053898, + "learning_rate": 2.4162036160075846e-06, + "loss": 0.2601, + "step": 31014 + }, + { + "epoch": 0.6826986572749285, + "grad_norm": 1.756646995123752, + "learning_rate": 2.4158984493088115e-06, + "loss": 0.3639, + "step": 31015 + }, + { + "epoch": 0.6827206691613471, + "grad_norm": 1.5774480569186315, + "learning_rate": 2.415593295743593e-06, + "loss": 0.3231, + "step": 31016 + }, + { + "epoch": 0.6827426810477658, + "grad_norm": 1.6444465578532053, + "learning_rate": 2.415288155313482e-06, + "loss": 0.2745, + "step": 31017 + }, + { + "epoch": 0.6827646929341845, + "grad_norm": 1.776311834980742, + "learning_rate": 2.414983028020031e-06, + "loss": 0.2444, + "step": 31018 + }, + { + "epoch": 0.6827867048206031, + "grad_norm": 1.9733552538957981, + "learning_rate": 2.414677913864788e-06, + "loss": 0.3169, + "step": 31019 + }, + { + "epoch": 0.6828087167070218, + "grad_norm": 1.8442226434627453, + "learning_rate": 2.414372812849306e-06, + "loss": 0.3014, + "step": 31020 + }, + { + "epoch": 0.6828307285934404, + "grad_norm": 0.6164907874803417, + "learning_rate": 2.4140677249751333e-06, + "loss": 0.457, + "step": 31021 + }, + { + "epoch": 0.6828527404798591, + "grad_norm": 12.099990207577585, + "learning_rate": 2.4137626502438217e-06, + "loss": 0.2876, + "step": 31022 + }, + { + "epoch": 0.6828747523662778, + "grad_norm": 0.6195089900006878, + "learning_rate": 2.4134575886569233e-06, + "loss": 0.4347, + "step": 31023 + }, + { + "epoch": 0.6828967642526964, + "grad_norm": 2.216668942756138, + "learning_rate": 2.4131525402159856e-06, + "loss": 0.3325, + "step": 31024 + }, + { + "epoch": 0.6829187761391151, + "grad_norm": 1.962770935140645, + "learning_rate": 2.412847504922562e-06, + "loss": 0.3827, + "step": 31025 + }, + { + "epoch": 0.6829407880255338, + "grad_norm": 1.6085884651039872, + "learning_rate": 2.4125424827782017e-06, + "loss": 0.3579, + "step": 31026 + }, + { + "epoch": 0.6829627999119524, + "grad_norm": 0.6080313749469899, + "learning_rate": 2.4122374737844523e-06, + "loss": 0.4698, + "step": 31027 + }, + { + "epoch": 0.6829848117983711, + "grad_norm": 0.6434451039812981, + "learning_rate": 2.4119324779428664e-06, + "loss": 0.4659, + "step": 31028 + }, + { + "epoch": 0.6830068236847898, + "grad_norm": 1.2623502028610285, + "learning_rate": 2.411627495254994e-06, + "loss": 0.2383, + "step": 31029 + }, + { + "epoch": 0.6830288355712084, + "grad_norm": 1.3346008867532775, + "learning_rate": 2.4113225257223865e-06, + "loss": 0.32, + "step": 31030 + }, + { + "epoch": 0.6830508474576271, + "grad_norm": 2.0991548458077127, + "learning_rate": 2.411017569346593e-06, + "loss": 0.2131, + "step": 31031 + }, + { + "epoch": 0.6830728593440458, + "grad_norm": 1.6958319623106184, + "learning_rate": 2.4107126261291614e-06, + "loss": 0.2769, + "step": 31032 + }, + { + "epoch": 0.6830948712304644, + "grad_norm": 1.257474692946979, + "learning_rate": 2.4104076960716423e-06, + "loss": 0.2939, + "step": 31033 + }, + { + "epoch": 0.6831168831168831, + "grad_norm": 1.4186945552229282, + "learning_rate": 2.4101027791755866e-06, + "loss": 0.2778, + "step": 31034 + }, + { + "epoch": 0.6831388950033018, + "grad_norm": 1.473976689599228, + "learning_rate": 2.4097978754425457e-06, + "loss": 0.3508, + "step": 31035 + }, + { + "epoch": 0.6831609068897204, + "grad_norm": 1.6083237559954484, + "learning_rate": 2.409492984874067e-06, + "loss": 0.3244, + "step": 31036 + }, + { + "epoch": 0.6831829187761391, + "grad_norm": 3.8494978117404632, + "learning_rate": 2.409188107471699e-06, + "loss": 0.2316, + "step": 31037 + }, + { + "epoch": 0.6832049306625578, + "grad_norm": 1.8707744117979153, + "learning_rate": 2.408883243236992e-06, + "loss": 0.2945, + "step": 31038 + }, + { + "epoch": 0.6832269425489764, + "grad_norm": 0.6204841701094076, + "learning_rate": 2.4085783921714963e-06, + "loss": 0.427, + "step": 31039 + }, + { + "epoch": 0.6832489544353951, + "grad_norm": 1.8061481473352836, + "learning_rate": 2.408273554276763e-06, + "loss": 0.3474, + "step": 31040 + }, + { + "epoch": 0.6832709663218138, + "grad_norm": 1.1790546837600975, + "learning_rate": 2.407968729554339e-06, + "loss": 0.2906, + "step": 31041 + }, + { + "epoch": 0.6832929782082324, + "grad_norm": 1.1827513112246422, + "learning_rate": 2.4076639180057726e-06, + "loss": 0.1574, + "step": 31042 + }, + { + "epoch": 0.6833149900946511, + "grad_norm": 1.4250735145935838, + "learning_rate": 2.4073591196326142e-06, + "loss": 0.2085, + "step": 31043 + }, + { + "epoch": 0.6833370019810697, + "grad_norm": 2.2453370919931332, + "learning_rate": 2.407054334436413e-06, + "loss": 0.3121, + "step": 31044 + }, + { + "epoch": 0.6833590138674884, + "grad_norm": 1.4203153725090163, + "learning_rate": 2.40674956241872e-06, + "loss": 0.3314, + "step": 31045 + }, + { + "epoch": 0.6833810257539071, + "grad_norm": 1.8841094582381641, + "learning_rate": 2.4064448035810828e-06, + "loss": 0.3531, + "step": 31046 + }, + { + "epoch": 0.6834030376403257, + "grad_norm": 1.4273263652908135, + "learning_rate": 2.4061400579250473e-06, + "loss": 0.3186, + "step": 31047 + }, + { + "epoch": 0.6834250495267444, + "grad_norm": 4.212693181075484, + "learning_rate": 2.4058353254521653e-06, + "loss": 0.31, + "step": 31048 + }, + { + "epoch": 0.6834470614131631, + "grad_norm": 0.6165214178964109, + "learning_rate": 2.4055306061639872e-06, + "loss": 0.4713, + "step": 31049 + }, + { + "epoch": 0.6834690732995817, + "grad_norm": 2.046291349554007, + "learning_rate": 2.4052259000620577e-06, + "loss": 0.3336, + "step": 31050 + }, + { + "epoch": 0.6834910851860004, + "grad_norm": 1.462767982056544, + "learning_rate": 2.4049212071479292e-06, + "loss": 0.2636, + "step": 31051 + }, + { + "epoch": 0.6835130970724191, + "grad_norm": 1.544874488726418, + "learning_rate": 2.404616527423147e-06, + "loss": 0.3688, + "step": 31052 + }, + { + "epoch": 0.6835351089588377, + "grad_norm": 1.945688802266266, + "learning_rate": 2.4043118608892608e-06, + "loss": 0.346, + "step": 31053 + }, + { + "epoch": 0.6835571208452564, + "grad_norm": 1.5087716964450582, + "learning_rate": 2.4040072075478217e-06, + "loss": 0.26, + "step": 31054 + }, + { + "epoch": 0.6835791327316751, + "grad_norm": 1.5281655612713456, + "learning_rate": 2.403702567400374e-06, + "loss": 0.2624, + "step": 31055 + }, + { + "epoch": 0.6836011446180937, + "grad_norm": 1.5493816908573756, + "learning_rate": 2.403397940448469e-06, + "loss": 0.3253, + "step": 31056 + }, + { + "epoch": 0.6836231565045124, + "grad_norm": 0.6262988996366596, + "learning_rate": 2.4030933266936523e-06, + "loss": 0.4793, + "step": 31057 + }, + { + "epoch": 0.6836451683909311, + "grad_norm": 1.5501215946346747, + "learning_rate": 2.402788726137476e-06, + "loss": 0.2927, + "step": 31058 + }, + { + "epoch": 0.6836671802773497, + "grad_norm": 1.7002577506369, + "learning_rate": 2.4024841387814833e-06, + "loss": 0.3189, + "step": 31059 + }, + { + "epoch": 0.6836891921637684, + "grad_norm": 1.7406176316172945, + "learning_rate": 2.402179564627225e-06, + "loss": 0.3374, + "step": 31060 + }, + { + "epoch": 0.6837112040501871, + "grad_norm": 1.27037903601995, + "learning_rate": 2.4018750036762507e-06, + "loss": 0.3267, + "step": 31061 + }, + { + "epoch": 0.6837332159366057, + "grad_norm": 1.6077406731039716, + "learning_rate": 2.401570455930104e-06, + "loss": 0.297, + "step": 31062 + }, + { + "epoch": 0.6837552278230244, + "grad_norm": 1.563168089604862, + "learning_rate": 2.401265921390338e-06, + "loss": 0.3022, + "step": 31063 + }, + { + "epoch": 0.6837772397094432, + "grad_norm": 3.5458811495715787, + "learning_rate": 2.4009614000584945e-06, + "loss": 0.2513, + "step": 31064 + }, + { + "epoch": 0.6837992515958617, + "grad_norm": 1.4952216915902208, + "learning_rate": 2.4006568919361257e-06, + "loss": 0.2954, + "step": 31065 + }, + { + "epoch": 0.6838212634822805, + "grad_norm": 1.4196063702402821, + "learning_rate": 2.400352397024779e-06, + "loss": 0.2904, + "step": 31066 + }, + { + "epoch": 0.683843275368699, + "grad_norm": 1.6318307326366601, + "learning_rate": 2.4000479153260007e-06, + "loss": 0.2949, + "step": 31067 + }, + { + "epoch": 0.6838652872551177, + "grad_norm": 2.0813771151492206, + "learning_rate": 2.399743446841337e-06, + "loss": 0.3592, + "step": 31068 + }, + { + "epoch": 0.6838872991415365, + "grad_norm": 1.5244340898495266, + "learning_rate": 2.399438991572337e-06, + "loss": 0.2857, + "step": 31069 + }, + { + "epoch": 0.683909311027955, + "grad_norm": 1.7520006139996458, + "learning_rate": 2.3991345495205482e-06, + "loss": 0.3167, + "step": 31070 + }, + { + "epoch": 0.6839313229143738, + "grad_norm": 0.6277291392260836, + "learning_rate": 2.3988301206875186e-06, + "loss": 0.4402, + "step": 31071 + }, + { + "epoch": 0.6839533348007925, + "grad_norm": 1.4686327435403594, + "learning_rate": 2.398525705074795e-06, + "loss": 0.2167, + "step": 31072 + }, + { + "epoch": 0.683975346687211, + "grad_norm": 1.45155879458231, + "learning_rate": 2.3982213026839225e-06, + "loss": 0.2291, + "step": 31073 + }, + { + "epoch": 0.6839973585736298, + "grad_norm": 0.6336164352142302, + "learning_rate": 2.3979169135164493e-06, + "loss": 0.4772, + "step": 31074 + }, + { + "epoch": 0.6840193704600485, + "grad_norm": 134.85774746024816, + "learning_rate": 2.3976125375739235e-06, + "loss": 0.2728, + "step": 31075 + }, + { + "epoch": 0.684041382346467, + "grad_norm": 1.5900877496247023, + "learning_rate": 2.3973081748578925e-06, + "loss": 0.2582, + "step": 31076 + }, + { + "epoch": 0.6840633942328858, + "grad_norm": 0.7363909544228954, + "learning_rate": 2.397003825369903e-06, + "loss": 0.4362, + "step": 31077 + }, + { + "epoch": 0.6840854061193045, + "grad_norm": 1.8373441120976246, + "learning_rate": 2.3966994891114988e-06, + "loss": 0.327, + "step": 31078 + }, + { + "epoch": 0.6841074180057231, + "grad_norm": 2.243633677023012, + "learning_rate": 2.396395166084228e-06, + "loss": 0.2899, + "step": 31079 + }, + { + "epoch": 0.6841294298921418, + "grad_norm": 1.641656045140741, + "learning_rate": 2.3960908562896396e-06, + "loss": 0.3658, + "step": 31080 + }, + { + "epoch": 0.6841514417785605, + "grad_norm": 1.2521163632625154, + "learning_rate": 2.39578655972928e-06, + "loss": 0.2418, + "step": 31081 + }, + { + "epoch": 0.6841734536649791, + "grad_norm": 2.2002767492343085, + "learning_rate": 2.395482276404695e-06, + "loss": 0.2226, + "step": 31082 + }, + { + "epoch": 0.6841954655513978, + "grad_norm": 1.8498707324342338, + "learning_rate": 2.3951780063174286e-06, + "loss": 0.2466, + "step": 31083 + }, + { + "epoch": 0.6842174774378165, + "grad_norm": 1.474899877561139, + "learning_rate": 2.3948737494690294e-06, + "loss": 0.2533, + "step": 31084 + }, + { + "epoch": 0.6842394893242351, + "grad_norm": 1.3470143541160946, + "learning_rate": 2.394569505861045e-06, + "loss": 0.3009, + "step": 31085 + }, + { + "epoch": 0.6842615012106538, + "grad_norm": 1.4149893653818182, + "learning_rate": 2.3942652754950196e-06, + "loss": 0.2976, + "step": 31086 + }, + { + "epoch": 0.6842835130970725, + "grad_norm": 1.869901674130157, + "learning_rate": 2.3939610583725016e-06, + "loss": 0.3058, + "step": 31087 + }, + { + "epoch": 0.6843055249834911, + "grad_norm": 0.6274118135699629, + "learning_rate": 2.393656854495034e-06, + "loss": 0.4547, + "step": 31088 + }, + { + "epoch": 0.6843275368699098, + "grad_norm": 1.852911800501983, + "learning_rate": 2.3933526638641648e-06, + "loss": 0.3392, + "step": 31089 + }, + { + "epoch": 0.6843495487563284, + "grad_norm": 1.7492419639727752, + "learning_rate": 2.393048486481442e-06, + "loss": 0.2616, + "step": 31090 + }, + { + "epoch": 0.6843715606427471, + "grad_norm": 1.8141963360674274, + "learning_rate": 2.392744322348407e-06, + "loss": 0.2544, + "step": 31091 + }, + { + "epoch": 0.6843935725291658, + "grad_norm": 1.67933729519776, + "learning_rate": 2.392440171466611e-06, + "loss": 0.2861, + "step": 31092 + }, + { + "epoch": 0.6844155844155844, + "grad_norm": 1.9671128579228414, + "learning_rate": 2.392136033837594e-06, + "loss": 0.3596, + "step": 31093 + }, + { + "epoch": 0.6844375963020031, + "grad_norm": 0.6311942075527145, + "learning_rate": 2.391831909462908e-06, + "loss": 0.4619, + "step": 31094 + }, + { + "epoch": 0.6844596081884218, + "grad_norm": 1.5239345475192045, + "learning_rate": 2.3915277983440933e-06, + "loss": 0.2563, + "step": 31095 + }, + { + "epoch": 0.6844816200748404, + "grad_norm": 2.415446125714228, + "learning_rate": 2.391223700482697e-06, + "loss": 0.3535, + "step": 31096 + }, + { + "epoch": 0.6845036319612591, + "grad_norm": 4.612625646492146, + "learning_rate": 2.390919615880268e-06, + "loss": 0.2872, + "step": 31097 + }, + { + "epoch": 0.6845256438476778, + "grad_norm": 2.5218487868385733, + "learning_rate": 2.3906155445383467e-06, + "loss": 0.2812, + "step": 31098 + }, + { + "epoch": 0.6845476557340964, + "grad_norm": 1.6463277547948583, + "learning_rate": 2.390311486458484e-06, + "loss": 0.297, + "step": 31099 + }, + { + "epoch": 0.6845696676205151, + "grad_norm": 1.4405769662102428, + "learning_rate": 2.3900074416422193e-06, + "loss": 0.33, + "step": 31100 + }, + { + "epoch": 0.6845916795069338, + "grad_norm": 4.400948986068699, + "learning_rate": 2.389703410091101e-06, + "loss": 0.2724, + "step": 31101 + }, + { + "epoch": 0.6846136913933524, + "grad_norm": 2.8600274950400917, + "learning_rate": 2.3893993918066764e-06, + "loss": 0.2849, + "step": 31102 + }, + { + "epoch": 0.6846357032797711, + "grad_norm": 1.7419658079739426, + "learning_rate": 2.3890953867904864e-06, + "loss": 0.3279, + "step": 31103 + }, + { + "epoch": 0.6846577151661898, + "grad_norm": 1.478536110466182, + "learning_rate": 2.38879139504408e-06, + "loss": 0.2695, + "step": 31104 + }, + { + "epoch": 0.6846797270526084, + "grad_norm": 2.5787884352464294, + "learning_rate": 2.388487416568999e-06, + "loss": 0.3247, + "step": 31105 + }, + { + "epoch": 0.6847017389390271, + "grad_norm": 2.8084476430507923, + "learning_rate": 2.388183451366789e-06, + "loss": 0.2893, + "step": 31106 + }, + { + "epoch": 0.6847237508254458, + "grad_norm": 1.6113368089947684, + "learning_rate": 2.387879499438998e-06, + "loss": 0.2814, + "step": 31107 + }, + { + "epoch": 0.6847457627118644, + "grad_norm": 1.7196418045293551, + "learning_rate": 2.387575560787168e-06, + "loss": 0.3155, + "step": 31108 + }, + { + "epoch": 0.6847677745982831, + "grad_norm": 1.4504493520977553, + "learning_rate": 2.3872716354128427e-06, + "loss": 0.2967, + "step": 31109 + }, + { + "epoch": 0.6847897864847018, + "grad_norm": 1.1371557045157579, + "learning_rate": 2.386967723317568e-06, + "loss": 0.2129, + "step": 31110 + }, + { + "epoch": 0.6848117983711204, + "grad_norm": 1.8948325877387118, + "learning_rate": 2.386663824502889e-06, + "loss": 0.2999, + "step": 31111 + }, + { + "epoch": 0.6848338102575391, + "grad_norm": 1.2507452946668811, + "learning_rate": 2.386359938970352e-06, + "loss": 0.2208, + "step": 31112 + }, + { + "epoch": 0.6848558221439577, + "grad_norm": 0.6161666432560573, + "learning_rate": 2.386056066721499e-06, + "loss": 0.4684, + "step": 31113 + }, + { + "epoch": 0.6848778340303764, + "grad_norm": 1.3939166685276543, + "learning_rate": 2.385752207757873e-06, + "loss": 0.2776, + "step": 31114 + }, + { + "epoch": 0.6848998459167951, + "grad_norm": 4.235422597022278, + "learning_rate": 2.3854483620810203e-06, + "loss": 0.3002, + "step": 31115 + }, + { + "epoch": 0.6849218578032137, + "grad_norm": 1.4217136442661833, + "learning_rate": 2.3851445296924858e-06, + "loss": 0.2851, + "step": 31116 + }, + { + "epoch": 0.6849438696896324, + "grad_norm": 1.9241919529152967, + "learning_rate": 2.3848407105938142e-06, + "loss": 0.2222, + "step": 31117 + }, + { + "epoch": 0.6849658815760511, + "grad_norm": 1.845754332619667, + "learning_rate": 2.384536904786549e-06, + "loss": 0.2821, + "step": 31118 + }, + { + "epoch": 0.6849878934624697, + "grad_norm": 1.4242170044049574, + "learning_rate": 2.384233112272232e-06, + "loss": 0.2907, + "step": 31119 + }, + { + "epoch": 0.6850099053488884, + "grad_norm": 2.946230853606498, + "learning_rate": 2.3839293330524086e-06, + "loss": 0.3018, + "step": 31120 + }, + { + "epoch": 0.6850319172353071, + "grad_norm": 0.635343128350976, + "learning_rate": 2.3836255671286237e-06, + "loss": 0.4584, + "step": 31121 + }, + { + "epoch": 0.6850539291217257, + "grad_norm": 0.6239681696601584, + "learning_rate": 2.3833218145024217e-06, + "loss": 0.4657, + "step": 31122 + }, + { + "epoch": 0.6850759410081444, + "grad_norm": 1.979072594665656, + "learning_rate": 2.3830180751753455e-06, + "loss": 0.3613, + "step": 31123 + }, + { + "epoch": 0.6850979528945631, + "grad_norm": 1.6926424100645188, + "learning_rate": 2.3827143491489367e-06, + "loss": 0.3374, + "step": 31124 + }, + { + "epoch": 0.6851199647809817, + "grad_norm": 2.5639066286358023, + "learning_rate": 2.382410636424741e-06, + "loss": 0.3904, + "step": 31125 + }, + { + "epoch": 0.6851419766674004, + "grad_norm": 1.4064867509253303, + "learning_rate": 2.3821069370043043e-06, + "loss": 0.2832, + "step": 31126 + }, + { + "epoch": 0.6851639885538191, + "grad_norm": 1.935174965450214, + "learning_rate": 2.381803250889165e-06, + "loss": 0.2583, + "step": 31127 + }, + { + "epoch": 0.6851860004402377, + "grad_norm": 1.9230051882735986, + "learning_rate": 2.3814995780808725e-06, + "loss": 0.2549, + "step": 31128 + }, + { + "epoch": 0.6852080123266564, + "grad_norm": 1.4325125477145348, + "learning_rate": 2.381195918580964e-06, + "loss": 0.2822, + "step": 31129 + }, + { + "epoch": 0.6852300242130751, + "grad_norm": 1.7780074879387908, + "learning_rate": 2.380892272390988e-06, + "loss": 0.2561, + "step": 31130 + }, + { + "epoch": 0.6852520360994937, + "grad_norm": 2.291537391774921, + "learning_rate": 2.3805886395124843e-06, + "loss": 0.3143, + "step": 31131 + }, + { + "epoch": 0.6852740479859124, + "grad_norm": 0.6092394327121365, + "learning_rate": 2.3802850199469967e-06, + "loss": 0.4467, + "step": 31132 + }, + { + "epoch": 0.6852960598723311, + "grad_norm": 1.6737422970890166, + "learning_rate": 2.3799814136960712e-06, + "loss": 0.2441, + "step": 31133 + }, + { + "epoch": 0.6853180717587497, + "grad_norm": 2.3915048421113596, + "learning_rate": 2.379677820761247e-06, + "loss": 0.2489, + "step": 31134 + }, + { + "epoch": 0.6853400836451684, + "grad_norm": 2.8370826055508074, + "learning_rate": 2.3793742411440707e-06, + "loss": 0.3807, + "step": 31135 + }, + { + "epoch": 0.6853620955315871, + "grad_norm": 2.2738443798336445, + "learning_rate": 2.3790706748460812e-06, + "loss": 0.3203, + "step": 31136 + }, + { + "epoch": 0.6853841074180057, + "grad_norm": 2.238413125942981, + "learning_rate": 2.3787671218688233e-06, + "loss": 0.2935, + "step": 31137 + }, + { + "epoch": 0.6854061193044244, + "grad_norm": 1.512180721823693, + "learning_rate": 2.3784635822138424e-06, + "loss": 0.3472, + "step": 31138 + }, + { + "epoch": 0.685428131190843, + "grad_norm": 1.22981117867727, + "learning_rate": 2.378160055882677e-06, + "loss": 0.2206, + "step": 31139 + }, + { + "epoch": 0.6854501430772617, + "grad_norm": 2.3534330127154064, + "learning_rate": 2.377856542876873e-06, + "loss": 0.3323, + "step": 31140 + }, + { + "epoch": 0.6854721549636804, + "grad_norm": 1.3195106402479595, + "learning_rate": 2.3775530431979697e-06, + "loss": 0.3013, + "step": 31141 + }, + { + "epoch": 0.685494166850099, + "grad_norm": 1.6214960060352868, + "learning_rate": 2.3772495568475116e-06, + "loss": 0.3128, + "step": 31142 + }, + { + "epoch": 0.6855161787365177, + "grad_norm": 1.8088667239557794, + "learning_rate": 2.376946083827043e-06, + "loss": 0.2842, + "step": 31143 + }, + { + "epoch": 0.6855381906229364, + "grad_norm": 0.6193329589925661, + "learning_rate": 2.3766426241381022e-06, + "loss": 0.4927, + "step": 31144 + }, + { + "epoch": 0.685560202509355, + "grad_norm": 1.4757778974322202, + "learning_rate": 2.376339177782236e-06, + "loss": 0.2737, + "step": 31145 + }, + { + "epoch": 0.6855822143957737, + "grad_norm": 0.612079622693102, + "learning_rate": 2.3760357447609823e-06, + "loss": 0.4672, + "step": 31146 + }, + { + "epoch": 0.6856042262821924, + "grad_norm": 1.8572038137230402, + "learning_rate": 2.3757323250758846e-06, + "loss": 0.2843, + "step": 31147 + }, + { + "epoch": 0.685626238168611, + "grad_norm": 1.4667576483531517, + "learning_rate": 2.375428918728488e-06, + "loss": 0.2568, + "step": 31148 + }, + { + "epoch": 0.6856482500550297, + "grad_norm": 1.7261261379996389, + "learning_rate": 2.375125525720332e-06, + "loss": 0.2558, + "step": 31149 + }, + { + "epoch": 0.6856702619414484, + "grad_norm": 2.1266959690974865, + "learning_rate": 2.3748221460529575e-06, + "loss": 0.3684, + "step": 31150 + }, + { + "epoch": 0.685692273827867, + "grad_norm": 0.6090722143292615, + "learning_rate": 2.3745187797279074e-06, + "loss": 0.4404, + "step": 31151 + }, + { + "epoch": 0.6857142857142857, + "grad_norm": 1.4823411645014617, + "learning_rate": 2.374215426746724e-06, + "loss": 0.2444, + "step": 31152 + }, + { + "epoch": 0.6857362976007044, + "grad_norm": 1.8077754306333427, + "learning_rate": 2.3739120871109505e-06, + "loss": 0.3253, + "step": 31153 + }, + { + "epoch": 0.685758309487123, + "grad_norm": 1.4190715305633606, + "learning_rate": 2.373608760822127e-06, + "loss": 0.3202, + "step": 31154 + }, + { + "epoch": 0.6857803213735417, + "grad_norm": 1.5478837286999687, + "learning_rate": 2.3733054478817937e-06, + "loss": 0.2786, + "step": 31155 + }, + { + "epoch": 0.6858023332599604, + "grad_norm": 2.600287728895723, + "learning_rate": 2.3730021482914934e-06, + "loss": 0.2758, + "step": 31156 + }, + { + "epoch": 0.685824345146379, + "grad_norm": 0.615406944523223, + "learning_rate": 2.372698862052768e-06, + "loss": 0.4645, + "step": 31157 + }, + { + "epoch": 0.6858463570327977, + "grad_norm": 1.460654101779242, + "learning_rate": 2.37239558916716e-06, + "loss": 0.2913, + "step": 31158 + }, + { + "epoch": 0.6858683689192164, + "grad_norm": 1.2042767795099902, + "learning_rate": 2.3720923296362104e-06, + "loss": 0.2504, + "step": 31159 + }, + { + "epoch": 0.685890380805635, + "grad_norm": 1.7489839368394975, + "learning_rate": 2.371789083461458e-06, + "loss": 0.2536, + "step": 31160 + }, + { + "epoch": 0.6859123926920537, + "grad_norm": 1.3861893868045398, + "learning_rate": 2.371485850644445e-06, + "loss": 0.2801, + "step": 31161 + }, + { + "epoch": 0.6859344045784723, + "grad_norm": 1.4423078272177996, + "learning_rate": 2.3711826311867132e-06, + "loss": 0.3163, + "step": 31162 + }, + { + "epoch": 0.685956416464891, + "grad_norm": 1.7068078135642453, + "learning_rate": 2.370879425089806e-06, + "loss": 0.3277, + "step": 31163 + }, + { + "epoch": 0.6859784283513097, + "grad_norm": 2.1781747487145595, + "learning_rate": 2.370576232355262e-06, + "loss": 0.4003, + "step": 31164 + }, + { + "epoch": 0.6860004402377283, + "grad_norm": 1.8756983828335654, + "learning_rate": 2.370273052984621e-06, + "loss": 0.2802, + "step": 31165 + }, + { + "epoch": 0.686022452124147, + "grad_norm": 1.1713913070432729, + "learning_rate": 2.3699698869794263e-06, + "loss": 0.2221, + "step": 31166 + }, + { + "epoch": 0.6860444640105657, + "grad_norm": 1.6622217529423629, + "learning_rate": 2.369666734341216e-06, + "loss": 0.2585, + "step": 31167 + }, + { + "epoch": 0.6860664758969843, + "grad_norm": 1.3942498643193226, + "learning_rate": 2.3693635950715327e-06, + "loss": 0.2255, + "step": 31168 + }, + { + "epoch": 0.686088487783403, + "grad_norm": 2.0851111469665993, + "learning_rate": 2.369060469171919e-06, + "loss": 0.2314, + "step": 31169 + }, + { + "epoch": 0.6861104996698217, + "grad_norm": 1.413146168128666, + "learning_rate": 2.368757356643911e-06, + "loss": 0.2758, + "step": 31170 + }, + { + "epoch": 0.6861325115562403, + "grad_norm": 0.6098817061480581, + "learning_rate": 2.3684542574890538e-06, + "loss": 0.4684, + "step": 31171 + }, + { + "epoch": 0.686154523442659, + "grad_norm": 1.7450828142473667, + "learning_rate": 2.368151171708884e-06, + "loss": 0.2561, + "step": 31172 + }, + { + "epoch": 0.6861765353290777, + "grad_norm": 11.089019365083338, + "learning_rate": 2.3678480993049434e-06, + "loss": 0.2762, + "step": 31173 + }, + { + "epoch": 0.6861985472154963, + "grad_norm": 1.740496191186387, + "learning_rate": 2.367545040278775e-06, + "loss": 0.2774, + "step": 31174 + }, + { + "epoch": 0.686220559101915, + "grad_norm": 1.5992004657499785, + "learning_rate": 2.3672419946319154e-06, + "loss": 0.2334, + "step": 31175 + }, + { + "epoch": 0.6862425709883337, + "grad_norm": 1.2912834836791631, + "learning_rate": 2.366938962365907e-06, + "loss": 0.2903, + "step": 31176 + }, + { + "epoch": 0.6862645828747523, + "grad_norm": 1.6747387559902909, + "learning_rate": 2.3666359434822882e-06, + "loss": 0.3583, + "step": 31177 + }, + { + "epoch": 0.686286594761171, + "grad_norm": 2.650994233183441, + "learning_rate": 2.3663329379825994e-06, + "loss": 0.2894, + "step": 31178 + }, + { + "epoch": 0.6863086066475897, + "grad_norm": 2.2900257510518305, + "learning_rate": 2.3660299458683837e-06, + "loss": 0.2937, + "step": 31179 + }, + { + "epoch": 0.6863306185340083, + "grad_norm": 1.4467776478530616, + "learning_rate": 2.3657269671411764e-06, + "loss": 0.3294, + "step": 31180 + }, + { + "epoch": 0.686352630420427, + "grad_norm": 0.6292823090381312, + "learning_rate": 2.3654240018025213e-06, + "loss": 0.4578, + "step": 31181 + }, + { + "epoch": 0.6863746423068458, + "grad_norm": 1.7779047667373773, + "learning_rate": 2.365121049853955e-06, + "loss": 0.419, + "step": 31182 + }, + { + "epoch": 0.6863966541932643, + "grad_norm": 1.3124571156742106, + "learning_rate": 2.364818111297019e-06, + "loss": 0.2342, + "step": 31183 + }, + { + "epoch": 0.686418666079683, + "grad_norm": 0.6536827609740407, + "learning_rate": 2.3645151861332537e-06, + "loss": 0.4888, + "step": 31184 + }, + { + "epoch": 0.6864406779661016, + "grad_norm": 2.618970224309518, + "learning_rate": 2.3642122743641983e-06, + "loss": 0.2598, + "step": 31185 + }, + { + "epoch": 0.6864626898525203, + "grad_norm": 1.824290818919072, + "learning_rate": 2.3639093759913895e-06, + "loss": 0.2874, + "step": 31186 + }, + { + "epoch": 0.686484701738939, + "grad_norm": 1.4556151033313773, + "learning_rate": 2.3636064910163694e-06, + "loss": 0.2264, + "step": 31187 + }, + { + "epoch": 0.6865067136253576, + "grad_norm": 2.0433942477699993, + "learning_rate": 2.363303619440676e-06, + "loss": 0.2925, + "step": 31188 + }, + { + "epoch": 0.6865287255117763, + "grad_norm": 1.2205866317061385, + "learning_rate": 2.3630007612658525e-06, + "loss": 0.2776, + "step": 31189 + }, + { + "epoch": 0.686550737398195, + "grad_norm": 1.8992197655567586, + "learning_rate": 2.362697916493435e-06, + "loss": 0.2772, + "step": 31190 + }, + { + "epoch": 0.6865727492846136, + "grad_norm": 1.350626414931634, + "learning_rate": 2.3623950851249605e-06, + "loss": 0.2718, + "step": 31191 + }, + { + "epoch": 0.6865947611710324, + "grad_norm": 1.6712845688907014, + "learning_rate": 2.362092267161971e-06, + "loss": 0.3469, + "step": 31192 + }, + { + "epoch": 0.6866167730574511, + "grad_norm": 0.6183888039782193, + "learning_rate": 2.3617894626060056e-06, + "loss": 0.4673, + "step": 31193 + }, + { + "epoch": 0.6866387849438697, + "grad_norm": 1.2860741095491006, + "learning_rate": 2.361486671458604e-06, + "loss": 0.2152, + "step": 31194 + }, + { + "epoch": 0.6866607968302884, + "grad_norm": 1.5730823542915169, + "learning_rate": 2.3611838937213043e-06, + "loss": 0.3008, + "step": 31195 + }, + { + "epoch": 0.6866828087167071, + "grad_norm": 1.4553246802437994, + "learning_rate": 2.3608811293956433e-06, + "loss": 0.2781, + "step": 31196 + }, + { + "epoch": 0.6867048206031257, + "grad_norm": 2.148900214806421, + "learning_rate": 2.3605783784831616e-06, + "loss": 0.3035, + "step": 31197 + }, + { + "epoch": 0.6867268324895444, + "grad_norm": 1.4676098422695438, + "learning_rate": 2.3602756409853973e-06, + "loss": 0.3141, + "step": 31198 + }, + { + "epoch": 0.6867488443759631, + "grad_norm": 0.6018618143784579, + "learning_rate": 2.3599729169038916e-06, + "loss": 0.4536, + "step": 31199 + }, + { + "epoch": 0.6867708562623817, + "grad_norm": 1.4697247460753586, + "learning_rate": 2.359670206240181e-06, + "loss": 0.276, + "step": 31200 + }, + { + "epoch": 0.6867928681488004, + "grad_norm": 1.3852850337808897, + "learning_rate": 2.3593675089958017e-06, + "loss": 0.3123, + "step": 31201 + }, + { + "epoch": 0.6868148800352191, + "grad_norm": 1.8363915170077059, + "learning_rate": 2.359064825172297e-06, + "loss": 0.3047, + "step": 31202 + }, + { + "epoch": 0.6868368919216377, + "grad_norm": 2.034441369059629, + "learning_rate": 2.358762154771199e-06, + "loss": 0.1655, + "step": 31203 + }, + { + "epoch": 0.6868589038080564, + "grad_norm": 4.30913923029141, + "learning_rate": 2.3584594977940535e-06, + "loss": 0.3511, + "step": 31204 + }, + { + "epoch": 0.6868809156944751, + "grad_norm": 1.7910310731657493, + "learning_rate": 2.3581568542423943e-06, + "loss": 0.2926, + "step": 31205 + }, + { + "epoch": 0.6869029275808937, + "grad_norm": 1.7448968430119456, + "learning_rate": 2.3578542241177595e-06, + "loss": 0.2242, + "step": 31206 + }, + { + "epoch": 0.6869249394673124, + "grad_norm": 2.358326671200449, + "learning_rate": 2.3575516074216886e-06, + "loss": 0.3675, + "step": 31207 + }, + { + "epoch": 0.686946951353731, + "grad_norm": 1.426526099029137, + "learning_rate": 2.3572490041557176e-06, + "loss": 0.3143, + "step": 31208 + }, + { + "epoch": 0.6869689632401497, + "grad_norm": 1.7523585505162396, + "learning_rate": 2.356946414321386e-06, + "loss": 0.3306, + "step": 31209 + }, + { + "epoch": 0.6869909751265684, + "grad_norm": 0.6293994696803817, + "learning_rate": 2.356643837920234e-06, + "loss": 0.4469, + "step": 31210 + }, + { + "epoch": 0.687012987012987, + "grad_norm": 1.7341974523422832, + "learning_rate": 2.3563412749537945e-06, + "loss": 0.3681, + "step": 31211 + }, + { + "epoch": 0.6870349988994057, + "grad_norm": 1.7248535971342245, + "learning_rate": 2.3560387254236094e-06, + "loss": 0.3771, + "step": 31212 + }, + { + "epoch": 0.6870570107858244, + "grad_norm": 0.6063766674203509, + "learning_rate": 2.3557361893312132e-06, + "loss": 0.464, + "step": 31213 + }, + { + "epoch": 0.687079022672243, + "grad_norm": 2.3721209656431537, + "learning_rate": 2.355433666678145e-06, + "loss": 0.2244, + "step": 31214 + }, + { + "epoch": 0.6871010345586617, + "grad_norm": 1.5656794952244952, + "learning_rate": 2.3551311574659443e-06, + "loss": 0.3362, + "step": 31215 + }, + { + "epoch": 0.6871230464450804, + "grad_norm": 1.9258971102805331, + "learning_rate": 2.3548286616961446e-06, + "loss": 0.3456, + "step": 31216 + }, + { + "epoch": 0.687145058331499, + "grad_norm": 0.5958780630588306, + "learning_rate": 2.354526179370288e-06, + "loss": 0.4476, + "step": 31217 + }, + { + "epoch": 0.6871670702179177, + "grad_norm": 1.7506390776545573, + "learning_rate": 2.354223710489907e-06, + "loss": 0.3325, + "step": 31218 + }, + { + "epoch": 0.6871890821043364, + "grad_norm": 1.882659482734414, + "learning_rate": 2.3539212550565405e-06, + "loss": 0.2588, + "step": 31219 + }, + { + "epoch": 0.687211093990755, + "grad_norm": 1.4296628162022686, + "learning_rate": 2.353618813071729e-06, + "loss": 0.2585, + "step": 31220 + }, + { + "epoch": 0.6872331058771737, + "grad_norm": 2.4398786834035726, + "learning_rate": 2.353316384537005e-06, + "loss": 0.3775, + "step": 31221 + }, + { + "epoch": 0.6872551177635924, + "grad_norm": 1.5011292203185738, + "learning_rate": 2.3530139694539096e-06, + "loss": 0.2477, + "step": 31222 + }, + { + "epoch": 0.687277129650011, + "grad_norm": 1.5496521468816238, + "learning_rate": 2.3527115678239755e-06, + "loss": 0.309, + "step": 31223 + }, + { + "epoch": 0.6872991415364297, + "grad_norm": 1.2618915500824466, + "learning_rate": 2.3524091796487418e-06, + "loss": 0.2953, + "step": 31224 + }, + { + "epoch": 0.6873211534228484, + "grad_norm": 0.6029197579263679, + "learning_rate": 2.352106804929748e-06, + "loss": 0.4637, + "step": 31225 + }, + { + "epoch": 0.687343165309267, + "grad_norm": 2.3230338711436107, + "learning_rate": 2.351804443668528e-06, + "loss": 0.2735, + "step": 31226 + }, + { + "epoch": 0.6873651771956857, + "grad_norm": 0.6123063402312279, + "learning_rate": 2.3515020958666175e-06, + "loss": 0.4294, + "step": 31227 + }, + { + "epoch": 0.6873871890821044, + "grad_norm": 1.4856001969472792, + "learning_rate": 2.3511997615255538e-06, + "loss": 0.2968, + "step": 31228 + }, + { + "epoch": 0.687409200968523, + "grad_norm": 1.7889808397247415, + "learning_rate": 2.3508974406468747e-06, + "loss": 0.2624, + "step": 31229 + }, + { + "epoch": 0.6874312128549417, + "grad_norm": 2.2951238654529433, + "learning_rate": 2.350595133232119e-06, + "loss": 0.3052, + "step": 31230 + }, + { + "epoch": 0.6874532247413603, + "grad_norm": 1.2521291667467682, + "learning_rate": 2.3502928392828194e-06, + "loss": 0.2552, + "step": 31231 + }, + { + "epoch": 0.687475236627779, + "grad_norm": 5.903850699643783, + "learning_rate": 2.3499905588005124e-06, + "loss": 0.307, + "step": 31232 + }, + { + "epoch": 0.6874972485141977, + "grad_norm": 1.5630724437409915, + "learning_rate": 2.349688291786735e-06, + "loss": 0.2808, + "step": 31233 + }, + { + "epoch": 0.6875192604006163, + "grad_norm": 2.8879244650669356, + "learning_rate": 2.3493860382430234e-06, + "loss": 0.3122, + "step": 31234 + }, + { + "epoch": 0.687541272287035, + "grad_norm": 1.6888236186706889, + "learning_rate": 2.3490837981709164e-06, + "loss": 0.2967, + "step": 31235 + }, + { + "epoch": 0.6875632841734537, + "grad_norm": 0.6034192754894878, + "learning_rate": 2.3487815715719476e-06, + "loss": 0.4529, + "step": 31236 + }, + { + "epoch": 0.6875852960598723, + "grad_norm": 1.8953049898873813, + "learning_rate": 2.3484793584476517e-06, + "loss": 0.2467, + "step": 31237 + }, + { + "epoch": 0.687607307946291, + "grad_norm": 2.18257655965928, + "learning_rate": 2.348177158799568e-06, + "loss": 0.2438, + "step": 31238 + }, + { + "epoch": 0.6876293198327097, + "grad_norm": 2.860715461362065, + "learning_rate": 2.347874972629227e-06, + "loss": 0.3293, + "step": 31239 + }, + { + "epoch": 0.6876513317191283, + "grad_norm": 1.3822692247165305, + "learning_rate": 2.347572799938172e-06, + "loss": 0.228, + "step": 31240 + }, + { + "epoch": 0.687673343605547, + "grad_norm": 1.8147791677857301, + "learning_rate": 2.3472706407279356e-06, + "loss": 0.2712, + "step": 31241 + }, + { + "epoch": 0.6876953554919657, + "grad_norm": 1.2897473346368065, + "learning_rate": 2.3469684950000503e-06, + "loss": 0.2303, + "step": 31242 + }, + { + "epoch": 0.6877173673783843, + "grad_norm": 6.146452703085205, + "learning_rate": 2.3466663627560566e-06, + "loss": 0.3909, + "step": 31243 + }, + { + "epoch": 0.687739379264803, + "grad_norm": 0.6375720584941451, + "learning_rate": 2.346364243997486e-06, + "loss": 0.4713, + "step": 31244 + }, + { + "epoch": 0.6877613911512217, + "grad_norm": 1.3438101277736592, + "learning_rate": 2.346062138725876e-06, + "loss": 0.2872, + "step": 31245 + }, + { + "epoch": 0.6877834030376403, + "grad_norm": 1.2291864856697403, + "learning_rate": 2.345760046942764e-06, + "loss": 0.2188, + "step": 31246 + }, + { + "epoch": 0.687805414924059, + "grad_norm": 1.4212484649256951, + "learning_rate": 2.3454579686496804e-06, + "loss": 0.2219, + "step": 31247 + }, + { + "epoch": 0.6878274268104777, + "grad_norm": 1.9906655553058816, + "learning_rate": 2.3451559038481657e-06, + "loss": 0.2816, + "step": 31248 + }, + { + "epoch": 0.6878494386968963, + "grad_norm": 1.6322548908735728, + "learning_rate": 2.344853852539751e-06, + "loss": 0.3306, + "step": 31249 + }, + { + "epoch": 0.687871450583315, + "grad_norm": 3.0624297846211097, + "learning_rate": 2.344551814725974e-06, + "loss": 0.3143, + "step": 31250 + }, + { + "epoch": 0.6878934624697337, + "grad_norm": 2.3311989970211036, + "learning_rate": 2.34424979040837e-06, + "loss": 0.2656, + "step": 31251 + }, + { + "epoch": 0.6879154743561523, + "grad_norm": 0.6301870841228766, + "learning_rate": 2.3439477795884714e-06, + "loss": 0.4624, + "step": 31252 + }, + { + "epoch": 0.687937486242571, + "grad_norm": 1.8772795157823945, + "learning_rate": 2.3436457822678167e-06, + "loss": 0.2876, + "step": 31253 + }, + { + "epoch": 0.6879594981289896, + "grad_norm": 1.385911762817373, + "learning_rate": 2.343343798447937e-06, + "loss": 0.3571, + "step": 31254 + }, + { + "epoch": 0.6879815100154083, + "grad_norm": 1.7317075301402372, + "learning_rate": 2.343041828130369e-06, + "loss": 0.3337, + "step": 31255 + }, + { + "epoch": 0.688003521901827, + "grad_norm": 2.480862672235275, + "learning_rate": 2.3427398713166492e-06, + "loss": 0.3679, + "step": 31256 + }, + { + "epoch": 0.6880255337882456, + "grad_norm": 1.7475019217395735, + "learning_rate": 2.342437928008309e-06, + "loss": 0.3064, + "step": 31257 + }, + { + "epoch": 0.6880475456746643, + "grad_norm": 0.6206348338220873, + "learning_rate": 2.3421359982068864e-06, + "loss": 0.4757, + "step": 31258 + }, + { + "epoch": 0.688069557561083, + "grad_norm": 3.5749566701415394, + "learning_rate": 2.341834081913912e-06, + "loss": 0.3382, + "step": 31259 + }, + { + "epoch": 0.6880915694475016, + "grad_norm": 1.8974363069127071, + "learning_rate": 2.3415321791309224e-06, + "loss": 0.3107, + "step": 31260 + }, + { + "epoch": 0.6881135813339203, + "grad_norm": 0.633741183244607, + "learning_rate": 2.341230289859454e-06, + "loss": 0.4542, + "step": 31261 + }, + { + "epoch": 0.688135593220339, + "grad_norm": 2.1978585656798817, + "learning_rate": 2.3409284141010373e-06, + "loss": 0.2174, + "step": 31262 + }, + { + "epoch": 0.6881576051067576, + "grad_norm": 1.317639610104264, + "learning_rate": 2.3406265518572104e-06, + "loss": 0.3074, + "step": 31263 + }, + { + "epoch": 0.6881796169931763, + "grad_norm": 1.9916795743273867, + "learning_rate": 2.3403247031295035e-06, + "loss": 0.3373, + "step": 31264 + }, + { + "epoch": 0.688201628879595, + "grad_norm": 1.6737114980587016, + "learning_rate": 2.3400228679194527e-06, + "loss": 0.2337, + "step": 31265 + }, + { + "epoch": 0.6882236407660136, + "grad_norm": 1.766747220351927, + "learning_rate": 2.3397210462285936e-06, + "loss": 0.3398, + "step": 31266 + }, + { + "epoch": 0.6882456526524323, + "grad_norm": 1.307063868108474, + "learning_rate": 2.3394192380584585e-06, + "loss": 0.3113, + "step": 31267 + }, + { + "epoch": 0.688267664538851, + "grad_norm": 1.5891610479536917, + "learning_rate": 2.33911744341058e-06, + "loss": 0.2454, + "step": 31268 + }, + { + "epoch": 0.6882896764252696, + "grad_norm": 1.3034944929169234, + "learning_rate": 2.3388156622864956e-06, + "loss": 0.277, + "step": 31269 + }, + { + "epoch": 0.6883116883116883, + "grad_norm": 1.92230011194314, + "learning_rate": 2.3385138946877327e-06, + "loss": 0.4759, + "step": 31270 + }, + { + "epoch": 0.688333700198107, + "grad_norm": 1.7389615118307982, + "learning_rate": 2.338212140615833e-06, + "loss": 0.3504, + "step": 31271 + }, + { + "epoch": 0.6883557120845256, + "grad_norm": 1.8623206896543583, + "learning_rate": 2.337910400072327e-06, + "loss": 0.4119, + "step": 31272 + }, + { + "epoch": 0.6883777239709443, + "grad_norm": 1.5218280725713524, + "learning_rate": 2.337608673058745e-06, + "loss": 0.3309, + "step": 31273 + }, + { + "epoch": 0.688399735857363, + "grad_norm": 1.3435854977576753, + "learning_rate": 2.3373069595766253e-06, + "loss": 0.2288, + "step": 31274 + }, + { + "epoch": 0.6884217477437816, + "grad_norm": 1.564879340222251, + "learning_rate": 2.3370052596274955e-06, + "loss": 0.2009, + "step": 31275 + }, + { + "epoch": 0.6884437596302003, + "grad_norm": 1.9436089553843405, + "learning_rate": 2.336703573212897e-06, + "loss": 0.2757, + "step": 31276 + }, + { + "epoch": 0.688465771516619, + "grad_norm": 1.787551566981769, + "learning_rate": 2.336401900334358e-06, + "loss": 0.3598, + "step": 31277 + }, + { + "epoch": 0.6884877834030376, + "grad_norm": 1.4038603823371831, + "learning_rate": 2.336100240993411e-06, + "loss": 0.255, + "step": 31278 + }, + { + "epoch": 0.6885097952894563, + "grad_norm": 1.8905622188227609, + "learning_rate": 2.3357985951915923e-06, + "loss": 0.2424, + "step": 31279 + }, + { + "epoch": 0.6885318071758749, + "grad_norm": 1.307490293055637, + "learning_rate": 2.335496962930432e-06, + "loss": 0.2762, + "step": 31280 + }, + { + "epoch": 0.6885538190622936, + "grad_norm": 1.7379176906021327, + "learning_rate": 2.3351953442114645e-06, + "loss": 0.2798, + "step": 31281 + }, + { + "epoch": 0.6885758309487123, + "grad_norm": 1.64559221393452, + "learning_rate": 2.3348937390362243e-06, + "loss": 0.2788, + "step": 31282 + }, + { + "epoch": 0.6885978428351309, + "grad_norm": 0.6646444662024804, + "learning_rate": 2.334592147406241e-06, + "loss": 0.4816, + "step": 31283 + }, + { + "epoch": 0.6886198547215496, + "grad_norm": 1.4541647936289832, + "learning_rate": 2.334290569323052e-06, + "loss": 0.3098, + "step": 31284 + }, + { + "epoch": 0.6886418666079683, + "grad_norm": 1.6441992950472382, + "learning_rate": 2.333989004788184e-06, + "loss": 0.2845, + "step": 31285 + }, + { + "epoch": 0.6886638784943869, + "grad_norm": 1.7296621247528785, + "learning_rate": 2.3336874538031735e-06, + "loss": 0.2693, + "step": 31286 + }, + { + "epoch": 0.6886858903808056, + "grad_norm": 1.6852496913385056, + "learning_rate": 2.3333859163695542e-06, + "loss": 0.3698, + "step": 31287 + }, + { + "epoch": 0.6887079022672243, + "grad_norm": 1.272633622238192, + "learning_rate": 2.3330843924888556e-06, + "loss": 0.186, + "step": 31288 + }, + { + "epoch": 0.6887299141536429, + "grad_norm": 1.5093449691589682, + "learning_rate": 2.3327828821626137e-06, + "loss": 0.3369, + "step": 31289 + }, + { + "epoch": 0.6887519260400616, + "grad_norm": 1.9152548205171631, + "learning_rate": 2.332481385392356e-06, + "loss": 0.2891, + "step": 31290 + }, + { + "epoch": 0.6887739379264803, + "grad_norm": 13.377254788686585, + "learning_rate": 2.332179902179619e-06, + "loss": 0.2984, + "step": 31291 + }, + { + "epoch": 0.6887959498128989, + "grad_norm": 1.5828680584499975, + "learning_rate": 2.331878432525934e-06, + "loss": 0.2948, + "step": 31292 + }, + { + "epoch": 0.6888179616993176, + "grad_norm": 1.9123850259891595, + "learning_rate": 2.3315769764328315e-06, + "loss": 0.2364, + "step": 31293 + }, + { + "epoch": 0.6888399735857363, + "grad_norm": 1.4499031319075433, + "learning_rate": 2.331275533901847e-06, + "loss": 0.2919, + "step": 31294 + }, + { + "epoch": 0.6888619854721549, + "grad_norm": 0.6137870277550981, + "learning_rate": 2.3309741049345093e-06, + "loss": 0.4439, + "step": 31295 + }, + { + "epoch": 0.6888839973585736, + "grad_norm": 1.7650906996343656, + "learning_rate": 2.330672689532351e-06, + "loss": 0.3676, + "step": 31296 + }, + { + "epoch": 0.6889060092449923, + "grad_norm": 1.8361849572932978, + "learning_rate": 2.330371287696906e-06, + "loss": 0.3085, + "step": 31297 + }, + { + "epoch": 0.6889280211314109, + "grad_norm": 1.9953803589719432, + "learning_rate": 2.3300698994297033e-06, + "loss": 0.3316, + "step": 31298 + }, + { + "epoch": 0.6889500330178296, + "grad_norm": 1.732254528732738, + "learning_rate": 2.329768524732278e-06, + "loss": 0.3327, + "step": 31299 + }, + { + "epoch": 0.6889720449042483, + "grad_norm": 1.288624013199015, + "learning_rate": 2.3294671636061583e-06, + "loss": 0.255, + "step": 31300 + }, + { + "epoch": 0.6889940567906669, + "grad_norm": 1.5992104025487828, + "learning_rate": 2.3291658160528775e-06, + "loss": 0.2383, + "step": 31301 + }, + { + "epoch": 0.6890160686770856, + "grad_norm": 1.5557288581371225, + "learning_rate": 2.3288644820739692e-06, + "loss": 0.2575, + "step": 31302 + }, + { + "epoch": 0.6890380805635042, + "grad_norm": 1.8590435368861633, + "learning_rate": 2.3285631616709616e-06, + "loss": 0.3529, + "step": 31303 + }, + { + "epoch": 0.689060092449923, + "grad_norm": 0.6351860975407071, + "learning_rate": 2.3282618548453886e-06, + "loss": 0.4677, + "step": 31304 + }, + { + "epoch": 0.6890821043363416, + "grad_norm": 0.6316650827099568, + "learning_rate": 2.3279605615987807e-06, + "loss": 0.4828, + "step": 31305 + }, + { + "epoch": 0.6891041162227602, + "grad_norm": 1.8674776688944144, + "learning_rate": 2.3276592819326648e-06, + "loss": 0.3326, + "step": 31306 + }, + { + "epoch": 0.689126128109179, + "grad_norm": 2.225601120679314, + "learning_rate": 2.3273580158485807e-06, + "loss": 0.3079, + "step": 31307 + }, + { + "epoch": 0.6891481399955977, + "grad_norm": 2.0068968390238977, + "learning_rate": 2.3270567633480553e-06, + "loss": 0.2712, + "step": 31308 + }, + { + "epoch": 0.6891701518820162, + "grad_norm": 1.5945368038963712, + "learning_rate": 2.3267555244326173e-06, + "loss": 0.3196, + "step": 31309 + }, + { + "epoch": 0.689192163768435, + "grad_norm": 1.8186897910035291, + "learning_rate": 2.3264542991038018e-06, + "loss": 0.3583, + "step": 31310 + }, + { + "epoch": 0.6892141756548537, + "grad_norm": 0.5970240062625106, + "learning_rate": 2.326153087363135e-06, + "loss": 0.4558, + "step": 31311 + }, + { + "epoch": 0.6892361875412722, + "grad_norm": 1.6797745012939724, + "learning_rate": 2.3258518892121533e-06, + "loss": 0.3551, + "step": 31312 + }, + { + "epoch": 0.689258199427691, + "grad_norm": 0.6043439781551895, + "learning_rate": 2.325550704652386e-06, + "loss": 0.4315, + "step": 31313 + }, + { + "epoch": 0.6892802113141097, + "grad_norm": 1.9585917467900653, + "learning_rate": 2.325249533685361e-06, + "loss": 0.3909, + "step": 31314 + }, + { + "epoch": 0.6893022232005283, + "grad_norm": 1.9107627956532607, + "learning_rate": 2.3249483763126124e-06, + "loss": 0.2975, + "step": 31315 + }, + { + "epoch": 0.689324235086947, + "grad_norm": 2.441942035411046, + "learning_rate": 2.3246472325356674e-06, + "loss": 0.2892, + "step": 31316 + }, + { + "epoch": 0.6893462469733657, + "grad_norm": 1.5528493216867598, + "learning_rate": 2.324346102356058e-06, + "loss": 0.311, + "step": 31317 + }, + { + "epoch": 0.6893682588597843, + "grad_norm": 1.7038788575603883, + "learning_rate": 2.3240449857753178e-06, + "loss": 0.3354, + "step": 31318 + }, + { + "epoch": 0.689390270746203, + "grad_norm": 2.086884460029749, + "learning_rate": 2.3237438827949722e-06, + "loss": 0.373, + "step": 31319 + }, + { + "epoch": 0.6894122826326217, + "grad_norm": 2.998610516959747, + "learning_rate": 2.3234427934165553e-06, + "loss": 0.2659, + "step": 31320 + }, + { + "epoch": 0.6894342945190403, + "grad_norm": 0.6343692572036558, + "learning_rate": 2.3231417176415943e-06, + "loss": 0.4793, + "step": 31321 + }, + { + "epoch": 0.689456306405459, + "grad_norm": 0.6390870716572787, + "learning_rate": 2.3228406554716214e-06, + "loss": 0.4533, + "step": 31322 + }, + { + "epoch": 0.6894783182918777, + "grad_norm": 1.4365017034588632, + "learning_rate": 2.322539606908168e-06, + "loss": 0.2526, + "step": 31323 + }, + { + "epoch": 0.6895003301782963, + "grad_norm": 1.3126331252889152, + "learning_rate": 2.3222385719527603e-06, + "loss": 0.3023, + "step": 31324 + }, + { + "epoch": 0.689522342064715, + "grad_norm": 1.462607288965697, + "learning_rate": 2.3219375506069324e-06, + "loss": 0.2842, + "step": 31325 + }, + { + "epoch": 0.6895443539511336, + "grad_norm": 2.2621157252339015, + "learning_rate": 2.321636542872211e-06, + "loss": 0.3173, + "step": 31326 + }, + { + "epoch": 0.6895663658375523, + "grad_norm": 0.6075327855734286, + "learning_rate": 2.3213355487501268e-06, + "loss": 0.4627, + "step": 31327 + }, + { + "epoch": 0.689588377723971, + "grad_norm": 2.356476488864151, + "learning_rate": 2.321034568242212e-06, + "loss": 0.3775, + "step": 31328 + }, + { + "epoch": 0.6896103896103896, + "grad_norm": 1.3720146364524295, + "learning_rate": 2.3207336013499926e-06, + "loss": 0.2805, + "step": 31329 + }, + { + "epoch": 0.6896324014968083, + "grad_norm": 1.7517232299319156, + "learning_rate": 2.3204326480750015e-06, + "loss": 0.3177, + "step": 31330 + }, + { + "epoch": 0.689654413383227, + "grad_norm": 1.3471275691273663, + "learning_rate": 2.3201317084187657e-06, + "loss": 0.3395, + "step": 31331 + }, + { + "epoch": 0.6896764252696456, + "grad_norm": 2.0505952784791943, + "learning_rate": 2.319830782382815e-06, + "loss": 0.335, + "step": 31332 + }, + { + "epoch": 0.6896984371560643, + "grad_norm": 3.0033864633101865, + "learning_rate": 2.3195298699686824e-06, + "loss": 0.2994, + "step": 31333 + }, + { + "epoch": 0.689720449042483, + "grad_norm": 3.986730513737973, + "learning_rate": 2.319228971177892e-06, + "loss": 0.265, + "step": 31334 + }, + { + "epoch": 0.6897424609289016, + "grad_norm": 2.0574800219061817, + "learning_rate": 2.318928086011978e-06, + "loss": 0.3204, + "step": 31335 + }, + { + "epoch": 0.6897644728153203, + "grad_norm": 3.106116670321412, + "learning_rate": 2.3186272144724648e-06, + "loss": 0.3193, + "step": 31336 + }, + { + "epoch": 0.689786484701739, + "grad_norm": 1.9937221919977381, + "learning_rate": 2.3183263565608843e-06, + "loss": 0.2242, + "step": 31337 + }, + { + "epoch": 0.6898084965881576, + "grad_norm": 1.7437221881382843, + "learning_rate": 2.318025512278767e-06, + "loss": 0.2475, + "step": 31338 + }, + { + "epoch": 0.6898305084745763, + "grad_norm": 1.662074125766157, + "learning_rate": 2.317724681627639e-06, + "loss": 0.2679, + "step": 31339 + }, + { + "epoch": 0.689852520360995, + "grad_norm": 1.3203586619228784, + "learning_rate": 2.3174238646090313e-06, + "loss": 0.2318, + "step": 31340 + }, + { + "epoch": 0.6898745322474136, + "grad_norm": 1.4519640869002464, + "learning_rate": 2.3171230612244725e-06, + "loss": 0.3881, + "step": 31341 + }, + { + "epoch": 0.6898965441338323, + "grad_norm": 1.6797481167425319, + "learning_rate": 2.316822271475488e-06, + "loss": 0.2831, + "step": 31342 + }, + { + "epoch": 0.689918556020251, + "grad_norm": 1.9806753663068002, + "learning_rate": 2.316521495363612e-06, + "loss": 0.244, + "step": 31343 + }, + { + "epoch": 0.6899405679066696, + "grad_norm": 1.7025727980718155, + "learning_rate": 2.316220732890369e-06, + "loss": 0.2884, + "step": 31344 + }, + { + "epoch": 0.6899625797930883, + "grad_norm": 2.056956388023518, + "learning_rate": 2.3159199840572914e-06, + "loss": 0.2269, + "step": 31345 + }, + { + "epoch": 0.689984591679507, + "grad_norm": 1.806571792615422, + "learning_rate": 2.3156192488659055e-06, + "loss": 0.2433, + "step": 31346 + }, + { + "epoch": 0.6900066035659256, + "grad_norm": 1.4678914595269161, + "learning_rate": 2.3153185273177365e-06, + "loss": 0.2802, + "step": 31347 + }, + { + "epoch": 0.6900286154523443, + "grad_norm": 0.6132467143517331, + "learning_rate": 2.315017819414319e-06, + "loss": 0.4392, + "step": 31348 + }, + { + "epoch": 0.6900506273387629, + "grad_norm": 0.6182188961038336, + "learning_rate": 2.3147171251571795e-06, + "loss": 0.4516, + "step": 31349 + }, + { + "epoch": 0.6900726392251816, + "grad_norm": 1.8781283216511724, + "learning_rate": 2.3144164445478424e-06, + "loss": 0.2424, + "step": 31350 + }, + { + "epoch": 0.6900946511116003, + "grad_norm": 1.3734521711728314, + "learning_rate": 2.314115777587841e-06, + "loss": 0.2817, + "step": 31351 + }, + { + "epoch": 0.6901166629980189, + "grad_norm": 0.5880308231718189, + "learning_rate": 2.3138151242786992e-06, + "loss": 0.4385, + "step": 31352 + }, + { + "epoch": 0.6901386748844376, + "grad_norm": 1.5280509369209465, + "learning_rate": 2.313514484621947e-06, + "loss": 0.2192, + "step": 31353 + }, + { + "epoch": 0.6901606867708563, + "grad_norm": 1.5322975578312157, + "learning_rate": 2.3132138586191144e-06, + "loss": 0.3093, + "step": 31354 + }, + { + "epoch": 0.6901826986572749, + "grad_norm": 3.43214014656806, + "learning_rate": 2.312913246271726e-06, + "loss": 0.2411, + "step": 31355 + }, + { + "epoch": 0.6902047105436936, + "grad_norm": 1.4262351334774008, + "learning_rate": 2.3126126475813114e-06, + "loss": 0.2407, + "step": 31356 + }, + { + "epoch": 0.6902267224301123, + "grad_norm": 1.5896592823584967, + "learning_rate": 2.3123120625493972e-06, + "loss": 0.3077, + "step": 31357 + }, + { + "epoch": 0.6902487343165309, + "grad_norm": 1.7534999198926053, + "learning_rate": 2.3120114911775118e-06, + "loss": 0.3109, + "step": 31358 + }, + { + "epoch": 0.6902707462029496, + "grad_norm": 2.6475030509827935, + "learning_rate": 2.3117109334671844e-06, + "loss": 0.3656, + "step": 31359 + }, + { + "epoch": 0.6902927580893683, + "grad_norm": 1.549810549268368, + "learning_rate": 2.311410389419939e-06, + "loss": 0.305, + "step": 31360 + }, + { + "epoch": 0.6903147699757869, + "grad_norm": 3.376316852426776, + "learning_rate": 2.3111098590373076e-06, + "loss": 0.282, + "step": 31361 + }, + { + "epoch": 0.6903367818622056, + "grad_norm": 0.6414375592024699, + "learning_rate": 2.3108093423208126e-06, + "loss": 0.4664, + "step": 31362 + }, + { + "epoch": 0.6903587937486243, + "grad_norm": 0.6172460936360399, + "learning_rate": 2.3105088392719845e-06, + "loss": 0.464, + "step": 31363 + }, + { + "epoch": 0.6903808056350429, + "grad_norm": 1.499341104356075, + "learning_rate": 2.3102083498923518e-06, + "loss": 0.3162, + "step": 31364 + }, + { + "epoch": 0.6904028175214616, + "grad_norm": 1.3810134242590666, + "learning_rate": 2.3099078741834378e-06, + "loss": 0.2779, + "step": 31365 + }, + { + "epoch": 0.6904248294078803, + "grad_norm": 1.5366159692862318, + "learning_rate": 2.309607412146774e-06, + "loss": 0.3222, + "step": 31366 + }, + { + "epoch": 0.6904468412942989, + "grad_norm": 1.8599736752899751, + "learning_rate": 2.3093069637838833e-06, + "loss": 0.3175, + "step": 31367 + }, + { + "epoch": 0.6904688531807176, + "grad_norm": 2.0126397365796507, + "learning_rate": 2.309006529096295e-06, + "loss": 0.4032, + "step": 31368 + }, + { + "epoch": 0.6904908650671363, + "grad_norm": 0.6270654986541448, + "learning_rate": 2.3087061080855374e-06, + "loss": 0.4482, + "step": 31369 + }, + { + "epoch": 0.6905128769535549, + "grad_norm": 2.0619010473627393, + "learning_rate": 2.308405700753133e-06, + "loss": 0.2791, + "step": 31370 + }, + { + "epoch": 0.6905348888399736, + "grad_norm": 1.4162385194675904, + "learning_rate": 2.3081053071006142e-06, + "loss": 0.3253, + "step": 31371 + }, + { + "epoch": 0.6905569007263922, + "grad_norm": 1.2411040877315307, + "learning_rate": 2.307804927129503e-06, + "loss": 0.2995, + "step": 31372 + }, + { + "epoch": 0.6905789126128109, + "grad_norm": 1.239494459247111, + "learning_rate": 2.3075045608413273e-06, + "loss": 0.2681, + "step": 31373 + }, + { + "epoch": 0.6906009244992296, + "grad_norm": 0.6186950674033735, + "learning_rate": 2.307204208237616e-06, + "loss": 0.4684, + "step": 31374 + }, + { + "epoch": 0.6906229363856482, + "grad_norm": 0.6297717809953781, + "learning_rate": 2.3069038693198933e-06, + "loss": 0.4588, + "step": 31375 + }, + { + "epoch": 0.6906449482720669, + "grad_norm": 1.763549350568479, + "learning_rate": 2.3066035440896866e-06, + "loss": 0.3413, + "step": 31376 + }, + { + "epoch": 0.6906669601584856, + "grad_norm": 1.409948113915875, + "learning_rate": 2.306303232548523e-06, + "loss": 0.3234, + "step": 31377 + }, + { + "epoch": 0.6906889720449042, + "grad_norm": 5.6117579702999745, + "learning_rate": 2.306002934697924e-06, + "loss": 0.3154, + "step": 31378 + }, + { + "epoch": 0.6907109839313229, + "grad_norm": 1.8586378331449875, + "learning_rate": 2.305702650539423e-06, + "loss": 0.299, + "step": 31379 + }, + { + "epoch": 0.6907329958177416, + "grad_norm": 0.6029543269783834, + "learning_rate": 2.305402380074541e-06, + "loss": 0.4667, + "step": 31380 + }, + { + "epoch": 0.6907550077041602, + "grad_norm": 1.7110581386007244, + "learning_rate": 2.305102123304808e-06, + "loss": 0.247, + "step": 31381 + }, + { + "epoch": 0.6907770195905789, + "grad_norm": 0.6151008467424668, + "learning_rate": 2.3048018802317478e-06, + "loss": 0.4801, + "step": 31382 + }, + { + "epoch": 0.6907990314769976, + "grad_norm": 1.7447519848816506, + "learning_rate": 2.3045016508568835e-06, + "loss": 0.3197, + "step": 31383 + }, + { + "epoch": 0.6908210433634162, + "grad_norm": 1.7200448757398141, + "learning_rate": 2.304201435181747e-06, + "loss": 0.3026, + "step": 31384 + }, + { + "epoch": 0.6908430552498349, + "grad_norm": 7.254345590981175, + "learning_rate": 2.3039012332078613e-06, + "loss": 0.2351, + "step": 31385 + }, + { + "epoch": 0.6908650671362536, + "grad_norm": 2.140102433307511, + "learning_rate": 2.30360104493675e-06, + "loss": 0.2719, + "step": 31386 + }, + { + "epoch": 0.6908870790226722, + "grad_norm": 1.5851324826740871, + "learning_rate": 2.3033008703699434e-06, + "loss": 0.4307, + "step": 31387 + }, + { + "epoch": 0.6909090909090909, + "grad_norm": 3.2481758370700273, + "learning_rate": 2.303000709508962e-06, + "loss": 0.2952, + "step": 31388 + }, + { + "epoch": 0.6909311027955096, + "grad_norm": 1.391502696503528, + "learning_rate": 2.302700562355334e-06, + "loss": 0.2193, + "step": 31389 + }, + { + "epoch": 0.6909531146819282, + "grad_norm": 1.706704374558882, + "learning_rate": 2.3024004289105867e-06, + "loss": 0.2731, + "step": 31390 + }, + { + "epoch": 0.6909751265683469, + "grad_norm": 3.7205421175810915, + "learning_rate": 2.302100309176242e-06, + "loss": 0.3153, + "step": 31391 + }, + { + "epoch": 0.6909971384547656, + "grad_norm": 0.59581535652744, + "learning_rate": 2.3018002031538277e-06, + "loss": 0.4326, + "step": 31392 + }, + { + "epoch": 0.6910191503411842, + "grad_norm": 1.7699022724242346, + "learning_rate": 2.3015001108448676e-06, + "loss": 0.2443, + "step": 31393 + }, + { + "epoch": 0.6910411622276029, + "grad_norm": 1.9602648830509832, + "learning_rate": 2.3012000322508867e-06, + "loss": 0.304, + "step": 31394 + }, + { + "epoch": 0.6910631741140215, + "grad_norm": 1.980345270808802, + "learning_rate": 2.3008999673734126e-06, + "loss": 0.3495, + "step": 31395 + }, + { + "epoch": 0.6910851860004402, + "grad_norm": 5.977149470878354, + "learning_rate": 2.3005999162139675e-06, + "loss": 0.2961, + "step": 31396 + }, + { + "epoch": 0.6911071978868589, + "grad_norm": 1.4615079438979808, + "learning_rate": 2.3002998787740793e-06, + "loss": 0.2905, + "step": 31397 + }, + { + "epoch": 0.6911292097732775, + "grad_norm": 1.3341618427395145, + "learning_rate": 2.2999998550552693e-06, + "loss": 0.2772, + "step": 31398 + }, + { + "epoch": 0.6911512216596962, + "grad_norm": 6.081039738314252, + "learning_rate": 2.299699845059064e-06, + "loss": 0.2819, + "step": 31399 + }, + { + "epoch": 0.6911732335461149, + "grad_norm": 2.0379392718473412, + "learning_rate": 2.2993998487869904e-06, + "loss": 0.2954, + "step": 31400 + }, + { + "epoch": 0.6911952454325335, + "grad_norm": 0.6154077140185513, + "learning_rate": 2.29909986624057e-06, + "loss": 0.455, + "step": 31401 + }, + { + "epoch": 0.6912172573189522, + "grad_norm": 0.6429672534596105, + "learning_rate": 2.2987998974213306e-06, + "loss": 0.4864, + "step": 31402 + }, + { + "epoch": 0.6912392692053709, + "grad_norm": 1.4663229053419835, + "learning_rate": 2.2984999423307923e-06, + "loss": 0.289, + "step": 31403 + }, + { + "epoch": 0.6912612810917895, + "grad_norm": 0.6314338648842859, + "learning_rate": 2.2982000009704824e-06, + "loss": 0.4495, + "step": 31404 + }, + { + "epoch": 0.6912832929782082, + "grad_norm": 0.610363800917093, + "learning_rate": 2.297900073341927e-06, + "loss": 0.4584, + "step": 31405 + }, + { + "epoch": 0.6913053048646269, + "grad_norm": 1.8792818243317297, + "learning_rate": 2.297600159446647e-06, + "loss": 0.2408, + "step": 31406 + }, + { + "epoch": 0.6913273167510455, + "grad_norm": 1.3853369109434313, + "learning_rate": 2.29730025928617e-06, + "loss": 0.2834, + "step": 31407 + }, + { + "epoch": 0.6913493286374642, + "grad_norm": 0.6231089723107873, + "learning_rate": 2.2970003728620164e-06, + "loss": 0.4458, + "step": 31408 + }, + { + "epoch": 0.6913713405238829, + "grad_norm": 1.5800598520389935, + "learning_rate": 2.2967005001757124e-06, + "loss": 0.2923, + "step": 31409 + }, + { + "epoch": 0.6913933524103015, + "grad_norm": 2.16290361279587, + "learning_rate": 2.2964006412287847e-06, + "loss": 0.4116, + "step": 31410 + }, + { + "epoch": 0.6914153642967202, + "grad_norm": 0.608635764433671, + "learning_rate": 2.296100796022753e-06, + "loss": 0.4686, + "step": 31411 + }, + { + "epoch": 0.6914373761831389, + "grad_norm": 2.1278264665401725, + "learning_rate": 2.295800964559144e-06, + "loss": 0.2725, + "step": 31412 + }, + { + "epoch": 0.6914593880695575, + "grad_norm": 1.5933988425016663, + "learning_rate": 2.2955011468394812e-06, + "loss": 0.2832, + "step": 31413 + }, + { + "epoch": 0.6914813999559762, + "grad_norm": 1.7961316073540086, + "learning_rate": 2.2952013428652836e-06, + "loss": 0.3887, + "step": 31414 + }, + { + "epoch": 0.6915034118423949, + "grad_norm": 1.3928679123695984, + "learning_rate": 2.294901552638083e-06, + "loss": 0.3236, + "step": 31415 + }, + { + "epoch": 0.6915254237288135, + "grad_norm": 1.969759791900209, + "learning_rate": 2.2946017761593973e-06, + "loss": 0.4332, + "step": 31416 + }, + { + "epoch": 0.6915474356152322, + "grad_norm": 2.4049682080569705, + "learning_rate": 2.2943020134307546e-06, + "loss": 0.231, + "step": 31417 + }, + { + "epoch": 0.691569447501651, + "grad_norm": 1.3765738463529866, + "learning_rate": 2.2940022644536752e-06, + "loss": 0.2823, + "step": 31418 + }, + { + "epoch": 0.6915914593880695, + "grad_norm": 1.5361178215277824, + "learning_rate": 2.293702529229682e-06, + "loss": 0.3103, + "step": 31419 + }, + { + "epoch": 0.6916134712744882, + "grad_norm": 5.908784461689213, + "learning_rate": 2.293402807760299e-06, + "loss": 0.332, + "step": 31420 + }, + { + "epoch": 0.6916354831609068, + "grad_norm": 1.63050603676498, + "learning_rate": 2.29310310004705e-06, + "loss": 0.3397, + "step": 31421 + }, + { + "epoch": 0.6916574950473255, + "grad_norm": 2.0171951090910683, + "learning_rate": 2.2928034060914607e-06, + "loss": 0.2775, + "step": 31422 + }, + { + "epoch": 0.6916795069337442, + "grad_norm": 2.327766270555351, + "learning_rate": 2.292503725895051e-06, + "loss": 0.3531, + "step": 31423 + }, + { + "epoch": 0.6917015188201628, + "grad_norm": 0.5867390251587219, + "learning_rate": 2.292204059459344e-06, + "loss": 0.4452, + "step": 31424 + }, + { + "epoch": 0.6917235307065815, + "grad_norm": 1.7893341521668154, + "learning_rate": 2.2919044067858637e-06, + "loss": 0.2802, + "step": 31425 + }, + { + "epoch": 0.6917455425930003, + "grad_norm": 1.6880094036225917, + "learning_rate": 2.2916047678761343e-06, + "loss": 0.4042, + "step": 31426 + }, + { + "epoch": 0.6917675544794188, + "grad_norm": 1.486220750168008, + "learning_rate": 2.291305142731676e-06, + "loss": 0.2907, + "step": 31427 + }, + { + "epoch": 0.6917895663658375, + "grad_norm": 0.6148100174794562, + "learning_rate": 2.291005531354015e-06, + "loss": 0.4703, + "step": 31428 + }, + { + "epoch": 0.6918115782522563, + "grad_norm": 2.7317061217236893, + "learning_rate": 2.29070593374467e-06, + "loss": 0.3981, + "step": 31429 + }, + { + "epoch": 0.6918335901386748, + "grad_norm": 2.62796628092821, + "learning_rate": 2.290406349905166e-06, + "loss": 0.3131, + "step": 31430 + }, + { + "epoch": 0.6918556020250936, + "grad_norm": 1.8102741151448094, + "learning_rate": 2.290106779837027e-06, + "loss": 0.2797, + "step": 31431 + }, + { + "epoch": 0.6918776139115123, + "grad_norm": 1.8709401160349215, + "learning_rate": 2.289807223541772e-06, + "loss": 0.2649, + "step": 31432 + }, + { + "epoch": 0.6918996257979309, + "grad_norm": 1.5453383211891352, + "learning_rate": 2.289507681020928e-06, + "loss": 0.2642, + "step": 31433 + }, + { + "epoch": 0.6919216376843496, + "grad_norm": 2.4879651988918123, + "learning_rate": 2.289208152276013e-06, + "loss": 0.3031, + "step": 31434 + }, + { + "epoch": 0.6919436495707683, + "grad_norm": 1.8191647137902747, + "learning_rate": 2.2889086373085513e-06, + "loss": 0.2907, + "step": 31435 + }, + { + "epoch": 0.6919656614571869, + "grad_norm": 2.038134377575329, + "learning_rate": 2.288609136120066e-06, + "loss": 0.3273, + "step": 31436 + }, + { + "epoch": 0.6919876733436056, + "grad_norm": 1.5520760489194327, + "learning_rate": 2.288309648712077e-06, + "loss": 0.2477, + "step": 31437 + }, + { + "epoch": 0.6920096852300243, + "grad_norm": 3.2644882317966846, + "learning_rate": 2.28801017508611e-06, + "loss": 0.1909, + "step": 31438 + }, + { + "epoch": 0.6920316971164429, + "grad_norm": 1.4298588727510952, + "learning_rate": 2.2877107152436826e-06, + "loss": 0.2383, + "step": 31439 + }, + { + "epoch": 0.6920537090028616, + "grad_norm": 2.065029050292218, + "learning_rate": 2.2874112691863193e-06, + "loss": 0.3158, + "step": 31440 + }, + { + "epoch": 0.6920757208892803, + "grad_norm": 1.584159780790026, + "learning_rate": 2.2871118369155438e-06, + "loss": 0.2971, + "step": 31441 + }, + { + "epoch": 0.6920977327756989, + "grad_norm": 0.6046425154119338, + "learning_rate": 2.286812418432874e-06, + "loss": 0.4508, + "step": 31442 + }, + { + "epoch": 0.6921197446621176, + "grad_norm": 1.6147174658221561, + "learning_rate": 2.286513013739835e-06, + "loss": 0.3056, + "step": 31443 + }, + { + "epoch": 0.6921417565485362, + "grad_norm": 1.6404731900995173, + "learning_rate": 2.2862136228379454e-06, + "loss": 0.285, + "step": 31444 + }, + { + "epoch": 0.6921637684349549, + "grad_norm": 1.746225967613548, + "learning_rate": 2.285914245728729e-06, + "loss": 0.3651, + "step": 31445 + }, + { + "epoch": 0.6921857803213736, + "grad_norm": 1.5720207506218193, + "learning_rate": 2.285614882413709e-06, + "loss": 0.311, + "step": 31446 + }, + { + "epoch": 0.6922077922077922, + "grad_norm": 1.940492905267259, + "learning_rate": 2.285315532894402e-06, + "loss": 0.2914, + "step": 31447 + }, + { + "epoch": 0.6922298040942109, + "grad_norm": 1.680300768000723, + "learning_rate": 2.2850161971723343e-06, + "loss": 0.3004, + "step": 31448 + }, + { + "epoch": 0.6922518159806296, + "grad_norm": 1.8259059261407353, + "learning_rate": 2.284716875249026e-06, + "loss": 0.2938, + "step": 31449 + }, + { + "epoch": 0.6922738278670482, + "grad_norm": 1.8852368986575523, + "learning_rate": 2.284417567125993e-06, + "loss": 0.3356, + "step": 31450 + }, + { + "epoch": 0.6922958397534669, + "grad_norm": 2.3758848584703407, + "learning_rate": 2.2841182728047657e-06, + "loss": 0.2915, + "step": 31451 + }, + { + "epoch": 0.6923178516398856, + "grad_norm": 0.6238786290554371, + "learning_rate": 2.283818992286858e-06, + "loss": 0.4799, + "step": 31452 + }, + { + "epoch": 0.6923398635263042, + "grad_norm": 2.1676519982822997, + "learning_rate": 2.2835197255737955e-06, + "loss": 0.2715, + "step": 31453 + }, + { + "epoch": 0.6923618754127229, + "grad_norm": 1.637582327180848, + "learning_rate": 2.2832204726670975e-06, + "loss": 0.2416, + "step": 31454 + }, + { + "epoch": 0.6923838872991416, + "grad_norm": 3.303850421306506, + "learning_rate": 2.2829212335682833e-06, + "loss": 0.3532, + "step": 31455 + }, + { + "epoch": 0.6924058991855602, + "grad_norm": 1.8047107473846546, + "learning_rate": 2.282622008278875e-06, + "loss": 0.289, + "step": 31456 + }, + { + "epoch": 0.6924279110719789, + "grad_norm": 1.6015402724567354, + "learning_rate": 2.2823227968003935e-06, + "loss": 0.2182, + "step": 31457 + }, + { + "epoch": 0.6924499229583976, + "grad_norm": 1.841143591951653, + "learning_rate": 2.2820235991343615e-06, + "loss": 0.2901, + "step": 31458 + }, + { + "epoch": 0.6924719348448162, + "grad_norm": 1.9504871316236063, + "learning_rate": 2.281724415282298e-06, + "loss": 0.3287, + "step": 31459 + }, + { + "epoch": 0.6924939467312349, + "grad_norm": 1.6364767690235524, + "learning_rate": 2.2814252452457216e-06, + "loss": 0.2644, + "step": 31460 + }, + { + "epoch": 0.6925159586176536, + "grad_norm": 1.7266523222575672, + "learning_rate": 2.2811260890261544e-06, + "loss": 0.2734, + "step": 31461 + }, + { + "epoch": 0.6925379705040722, + "grad_norm": 0.6118112549733677, + "learning_rate": 2.280826946625117e-06, + "loss": 0.4521, + "step": 31462 + }, + { + "epoch": 0.6925599823904909, + "grad_norm": 1.4821986977785409, + "learning_rate": 2.280527818044132e-06, + "loss": 0.3075, + "step": 31463 + }, + { + "epoch": 0.6925819942769096, + "grad_norm": 1.7832489650723262, + "learning_rate": 2.2802287032847177e-06, + "loss": 0.3809, + "step": 31464 + }, + { + "epoch": 0.6926040061633282, + "grad_norm": 2.253253174816689, + "learning_rate": 2.279929602348392e-06, + "loss": 0.2738, + "step": 31465 + }, + { + "epoch": 0.6926260180497469, + "grad_norm": 0.6384415562969593, + "learning_rate": 2.279630515236677e-06, + "loss": 0.4494, + "step": 31466 + }, + { + "epoch": 0.6926480299361655, + "grad_norm": 1.500288063362925, + "learning_rate": 2.2793314419510956e-06, + "loss": 0.3296, + "step": 31467 + }, + { + "epoch": 0.6926700418225842, + "grad_norm": 1.9215883921869767, + "learning_rate": 2.2790323824931625e-06, + "loss": 0.2865, + "step": 31468 + }, + { + "epoch": 0.6926920537090029, + "grad_norm": 1.9043367567366973, + "learning_rate": 2.278733336864403e-06, + "loss": 0.3272, + "step": 31469 + }, + { + "epoch": 0.6927140655954215, + "grad_norm": 1.4079827881951217, + "learning_rate": 2.2784343050663327e-06, + "loss": 0.2425, + "step": 31470 + }, + { + "epoch": 0.6927360774818402, + "grad_norm": 1.9260424025417773, + "learning_rate": 2.2781352871004726e-06, + "loss": 0.334, + "step": 31471 + }, + { + "epoch": 0.6927580893682589, + "grad_norm": 0.6132419573437309, + "learning_rate": 2.2778362829683447e-06, + "loss": 0.4499, + "step": 31472 + }, + { + "epoch": 0.6927801012546775, + "grad_norm": 6.556163836540827, + "learning_rate": 2.2775372926714654e-06, + "loss": 0.3, + "step": 31473 + }, + { + "epoch": 0.6928021131410962, + "grad_norm": 1.499535679076896, + "learning_rate": 2.2772383162113577e-06, + "loss": 0.2904, + "step": 31474 + }, + { + "epoch": 0.6928241250275149, + "grad_norm": 1.1796918120349853, + "learning_rate": 2.276939353589537e-06, + "loss": 0.3132, + "step": 31475 + }, + { + "epoch": 0.6928461369139335, + "grad_norm": 1.8210164670987705, + "learning_rate": 2.2766404048075256e-06, + "loss": 0.2451, + "step": 31476 + }, + { + "epoch": 0.6928681488003522, + "grad_norm": 2.2367005173119834, + "learning_rate": 2.276341469866844e-06, + "loss": 0.2837, + "step": 31477 + }, + { + "epoch": 0.6928901606867709, + "grad_norm": 1.3172454494014314, + "learning_rate": 2.276042548769008e-06, + "loss": 0.3089, + "step": 31478 + }, + { + "epoch": 0.6929121725731895, + "grad_norm": 1.6302120054084004, + "learning_rate": 2.2757436415155406e-06, + "loss": 0.2702, + "step": 31479 + }, + { + "epoch": 0.6929341844596082, + "grad_norm": 0.6205152104448487, + "learning_rate": 2.275444748107957e-06, + "loss": 0.4512, + "step": 31480 + }, + { + "epoch": 0.6929561963460269, + "grad_norm": 3.7083892606732376, + "learning_rate": 2.275145868547778e-06, + "loss": 0.2305, + "step": 31481 + }, + { + "epoch": 0.6929782082324455, + "grad_norm": 0.6064193362461284, + "learning_rate": 2.2748470028365253e-06, + "loss": 0.4629, + "step": 31482 + }, + { + "epoch": 0.6930002201188642, + "grad_norm": 0.6132160217392922, + "learning_rate": 2.2745481509757137e-06, + "loss": 0.4542, + "step": 31483 + }, + { + "epoch": 0.6930222320052829, + "grad_norm": 0.6236711423211849, + "learning_rate": 2.274249312966866e-06, + "loss": 0.467, + "step": 31484 + }, + { + "epoch": 0.6930442438917015, + "grad_norm": 1.5402889200714935, + "learning_rate": 2.2739504888114988e-06, + "loss": 0.2733, + "step": 31485 + }, + { + "epoch": 0.6930662557781202, + "grad_norm": 2.251538124494182, + "learning_rate": 2.2736516785111274e-06, + "loss": 0.26, + "step": 31486 + }, + { + "epoch": 0.6930882676645389, + "grad_norm": 1.2558431514766457, + "learning_rate": 2.273352882067278e-06, + "loss": 0.2794, + "step": 31487 + }, + { + "epoch": 0.6931102795509575, + "grad_norm": 2.818698632017695, + "learning_rate": 2.273054099481463e-06, + "loss": 0.3824, + "step": 31488 + }, + { + "epoch": 0.6931322914373762, + "grad_norm": 1.7557557318782875, + "learning_rate": 2.2727553307552057e-06, + "loss": 0.3421, + "step": 31489 + }, + { + "epoch": 0.6931543033237948, + "grad_norm": 1.485461440354022, + "learning_rate": 2.2724565758900224e-06, + "loss": 0.2308, + "step": 31490 + }, + { + "epoch": 0.6931763152102135, + "grad_norm": 1.6115501907299141, + "learning_rate": 2.272157834887429e-06, + "loss": 0.2779, + "step": 31491 + }, + { + "epoch": 0.6931983270966322, + "grad_norm": 3.976910473765927, + "learning_rate": 2.271859107748946e-06, + "loss": 0.2879, + "step": 31492 + }, + { + "epoch": 0.6932203389830508, + "grad_norm": 1.3139409089759397, + "learning_rate": 2.2715603944760923e-06, + "loss": 0.3141, + "step": 31493 + }, + { + "epoch": 0.6932423508694695, + "grad_norm": 3.0722768055837477, + "learning_rate": 2.271261695070387e-06, + "loss": 0.2503, + "step": 31494 + }, + { + "epoch": 0.6932643627558882, + "grad_norm": 0.618831466141357, + "learning_rate": 2.2709630095333473e-06, + "loss": 0.4573, + "step": 31495 + }, + { + "epoch": 0.6932863746423068, + "grad_norm": 2.908322094005071, + "learning_rate": 2.2706643378664883e-06, + "loss": 0.2945, + "step": 31496 + }, + { + "epoch": 0.6933083865287255, + "grad_norm": 1.9181874512914394, + "learning_rate": 2.2703656800713307e-06, + "loss": 0.3424, + "step": 31497 + }, + { + "epoch": 0.6933303984151442, + "grad_norm": 2.2914032064479257, + "learning_rate": 2.2700670361493924e-06, + "loss": 0.321, + "step": 31498 + }, + { + "epoch": 0.6933524103015628, + "grad_norm": 0.5962001976609639, + "learning_rate": 2.269768406102193e-06, + "loss": 0.4578, + "step": 31499 + }, + { + "epoch": 0.6933744221879815, + "grad_norm": 1.0998536041163247, + "learning_rate": 2.269469789931248e-06, + "loss": 0.2183, + "step": 31500 + }, + { + "epoch": 0.6933964340744002, + "grad_norm": 1.7399226109774253, + "learning_rate": 2.2691711876380733e-06, + "loss": 0.3737, + "step": 31501 + }, + { + "epoch": 0.6934184459608188, + "grad_norm": 1.568471377738619, + "learning_rate": 2.268872599224189e-06, + "loss": 0.3217, + "step": 31502 + }, + { + "epoch": 0.6934404578472375, + "grad_norm": 1.6223048011350478, + "learning_rate": 2.268574024691112e-06, + "loss": 0.2508, + "step": 31503 + }, + { + "epoch": 0.6934624697336562, + "grad_norm": 1.7006982294670885, + "learning_rate": 2.268275464040362e-06, + "loss": 0.2765, + "step": 31504 + }, + { + "epoch": 0.6934844816200748, + "grad_norm": 1.3624550852271204, + "learning_rate": 2.2679769172734546e-06, + "loss": 0.2514, + "step": 31505 + }, + { + "epoch": 0.6935064935064935, + "grad_norm": 2.6979328005684966, + "learning_rate": 2.2676783843919046e-06, + "loss": 0.2534, + "step": 31506 + }, + { + "epoch": 0.6935285053929122, + "grad_norm": 0.6238001326681261, + "learning_rate": 2.2673798653972324e-06, + "loss": 0.4604, + "step": 31507 + }, + { + "epoch": 0.6935505172793308, + "grad_norm": 1.7616638173319839, + "learning_rate": 2.267081360290956e-06, + "loss": 0.2672, + "step": 31508 + }, + { + "epoch": 0.6935725291657495, + "grad_norm": 1.6653594244378132, + "learning_rate": 2.266782869074589e-06, + "loss": 0.3076, + "step": 31509 + }, + { + "epoch": 0.6935945410521682, + "grad_norm": 2.2012460936112705, + "learning_rate": 2.2664843917496526e-06, + "loss": 0.2644, + "step": 31510 + }, + { + "epoch": 0.6936165529385868, + "grad_norm": 1.4773627899602626, + "learning_rate": 2.26618592831766e-06, + "loss": 0.3025, + "step": 31511 + }, + { + "epoch": 0.6936385648250055, + "grad_norm": 1.3233779889436448, + "learning_rate": 2.26588747878013e-06, + "loss": 0.2637, + "step": 31512 + }, + { + "epoch": 0.6936605767114241, + "grad_norm": 1.5555567376895187, + "learning_rate": 2.2655890431385806e-06, + "loss": 0.2386, + "step": 31513 + }, + { + "epoch": 0.6936825885978428, + "grad_norm": 1.4578861214898011, + "learning_rate": 2.265290621394526e-06, + "loss": 0.3729, + "step": 31514 + }, + { + "epoch": 0.6937046004842615, + "grad_norm": 1.5524675072073864, + "learning_rate": 2.2649922135494855e-06, + "loss": 0.2925, + "step": 31515 + }, + { + "epoch": 0.6937266123706801, + "grad_norm": 1.9515301853034626, + "learning_rate": 2.2646938196049728e-06, + "loss": 0.2637, + "step": 31516 + }, + { + "epoch": 0.6937486242570988, + "grad_norm": 2.995439566553278, + "learning_rate": 2.264395439562506e-06, + "loss": 0.2924, + "step": 31517 + }, + { + "epoch": 0.6937706361435175, + "grad_norm": 1.7428090564505498, + "learning_rate": 2.2640970734236032e-06, + "loss": 0.2824, + "step": 31518 + }, + { + "epoch": 0.6937926480299361, + "grad_norm": 0.5797379946555792, + "learning_rate": 2.2637987211897783e-06, + "loss": 0.4161, + "step": 31519 + }, + { + "epoch": 0.6938146599163548, + "grad_norm": 1.567985838596735, + "learning_rate": 2.26350038286255e-06, + "loss": 0.3334, + "step": 31520 + }, + { + "epoch": 0.6938366718027735, + "grad_norm": 1.6466536974383204, + "learning_rate": 2.2632020584434317e-06, + "loss": 0.2402, + "step": 31521 + }, + { + "epoch": 0.6938586836891921, + "grad_norm": 1.4202743278170435, + "learning_rate": 2.262903747933941e-06, + "loss": 0.3718, + "step": 31522 + }, + { + "epoch": 0.6938806955756108, + "grad_norm": 1.3568627701204754, + "learning_rate": 2.262605451335596e-06, + "loss": 0.2995, + "step": 31523 + }, + { + "epoch": 0.6939027074620295, + "grad_norm": 1.4237704089208483, + "learning_rate": 2.2623071686499094e-06, + "loss": 0.2395, + "step": 31524 + }, + { + "epoch": 0.6939247193484481, + "grad_norm": 1.6376614032033936, + "learning_rate": 2.2620088998784e-06, + "loss": 0.3532, + "step": 31525 + }, + { + "epoch": 0.6939467312348668, + "grad_norm": 1.99777432097611, + "learning_rate": 2.2617106450225834e-06, + "loss": 0.3562, + "step": 31526 + }, + { + "epoch": 0.6939687431212855, + "grad_norm": 2.093325445516558, + "learning_rate": 2.2614124040839725e-06, + "loss": 0.2498, + "step": 31527 + }, + { + "epoch": 0.6939907550077041, + "grad_norm": 1.2868540212635893, + "learning_rate": 2.2611141770640847e-06, + "loss": 0.2838, + "step": 31528 + }, + { + "epoch": 0.6940127668941228, + "grad_norm": 6.218013684705607, + "learning_rate": 2.260815963964437e-06, + "loss": 0.2723, + "step": 31529 + }, + { + "epoch": 0.6940347787805415, + "grad_norm": 0.6023350550393235, + "learning_rate": 2.2605177647865458e-06, + "loss": 0.4499, + "step": 31530 + }, + { + "epoch": 0.6940567906669601, + "grad_norm": 1.4723723072672201, + "learning_rate": 2.2602195795319248e-06, + "loss": 0.3055, + "step": 31531 + }, + { + "epoch": 0.6940788025533788, + "grad_norm": 0.6193797239477686, + "learning_rate": 2.259921408202089e-06, + "loss": 0.4622, + "step": 31532 + }, + { + "epoch": 0.6941008144397975, + "grad_norm": 2.0296755663169614, + "learning_rate": 2.2596232507985536e-06, + "loss": 0.3127, + "step": 31533 + }, + { + "epoch": 0.6941228263262161, + "grad_norm": 0.6307128484590606, + "learning_rate": 2.2593251073228354e-06, + "loss": 0.4768, + "step": 31534 + }, + { + "epoch": 0.6941448382126348, + "grad_norm": 1.4651049270295506, + "learning_rate": 2.2590269777764516e-06, + "loss": 0.3096, + "step": 31535 + }, + { + "epoch": 0.6941668500990534, + "grad_norm": 2.369639283708169, + "learning_rate": 2.258728862160915e-06, + "loss": 0.3152, + "step": 31536 + }, + { + "epoch": 0.6941888619854721, + "grad_norm": 1.5897977617983954, + "learning_rate": 2.258430760477739e-06, + "loss": 0.2305, + "step": 31537 + }, + { + "epoch": 0.6942108738718908, + "grad_norm": 0.6202301052478015, + "learning_rate": 2.2581326727284407e-06, + "loss": 0.4633, + "step": 31538 + }, + { + "epoch": 0.6942328857583094, + "grad_norm": 1.9591983384643628, + "learning_rate": 2.2578345989145357e-06, + "loss": 0.2788, + "step": 31539 + }, + { + "epoch": 0.6942548976447281, + "grad_norm": 1.2666688284113066, + "learning_rate": 2.2575365390375393e-06, + "loss": 0.2305, + "step": 31540 + }, + { + "epoch": 0.6942769095311468, + "grad_norm": 2.543149979673874, + "learning_rate": 2.2572384930989656e-06, + "loss": 0.2826, + "step": 31541 + }, + { + "epoch": 0.6942989214175654, + "grad_norm": 0.6124302467852576, + "learning_rate": 2.2569404611003277e-06, + "loss": 0.4558, + "step": 31542 + }, + { + "epoch": 0.6943209333039841, + "grad_norm": 1.951568286851961, + "learning_rate": 2.256642443043141e-06, + "loss": 0.3216, + "step": 31543 + }, + { + "epoch": 0.6943429451904028, + "grad_norm": 1.3467356224647942, + "learning_rate": 2.256344438928922e-06, + "loss": 0.3006, + "step": 31544 + }, + { + "epoch": 0.6943649570768214, + "grad_norm": 1.9473152176637907, + "learning_rate": 2.2560464487591855e-06, + "loss": 0.406, + "step": 31545 + }, + { + "epoch": 0.6943869689632401, + "grad_norm": 1.5200173470787566, + "learning_rate": 2.2557484725354446e-06, + "loss": 0.2564, + "step": 31546 + }, + { + "epoch": 0.6944089808496589, + "grad_norm": 1.2163914441918282, + "learning_rate": 2.255450510259212e-06, + "loss": 0.2901, + "step": 31547 + }, + { + "epoch": 0.6944309927360774, + "grad_norm": 1.3070471413219602, + "learning_rate": 2.2551525619320045e-06, + "loss": 0.3211, + "step": 31548 + }, + { + "epoch": 0.6944530046224962, + "grad_norm": 1.3808503033685673, + "learning_rate": 2.2548546275553377e-06, + "loss": 0.2864, + "step": 31549 + }, + { + "epoch": 0.6944750165089149, + "grad_norm": 1.6888892057464173, + "learning_rate": 2.2545567071307223e-06, + "loss": 0.2887, + "step": 31550 + }, + { + "epoch": 0.6944970283953334, + "grad_norm": 2.3106670411237697, + "learning_rate": 2.2542588006596763e-06, + "loss": 0.3682, + "step": 31551 + }, + { + "epoch": 0.6945190402817522, + "grad_norm": 1.8830367688835021, + "learning_rate": 2.2539609081437093e-06, + "loss": 0.3132, + "step": 31552 + }, + { + "epoch": 0.6945410521681709, + "grad_norm": 0.6454350372227037, + "learning_rate": 2.2536630295843377e-06, + "loss": 0.4743, + "step": 31553 + }, + { + "epoch": 0.6945630640545895, + "grad_norm": 0.6453896734983923, + "learning_rate": 2.253365164983078e-06, + "loss": 0.4877, + "step": 31554 + }, + { + "epoch": 0.6945850759410082, + "grad_norm": 1.989857956008824, + "learning_rate": 2.253067314341439e-06, + "loss": 0.3514, + "step": 31555 + }, + { + "epoch": 0.6946070878274269, + "grad_norm": 1.8012122109826476, + "learning_rate": 2.2527694776609394e-06, + "loss": 0.3574, + "step": 31556 + }, + { + "epoch": 0.6946290997138455, + "grad_norm": 1.7430124382469108, + "learning_rate": 2.252471654943088e-06, + "loss": 0.2597, + "step": 31557 + }, + { + "epoch": 0.6946511116002642, + "grad_norm": 1.6334649144360003, + "learning_rate": 2.252173846189402e-06, + "loss": 0.2767, + "step": 31558 + }, + { + "epoch": 0.6946731234866829, + "grad_norm": 1.4224126651904787, + "learning_rate": 2.2518760514013954e-06, + "loss": 0.3188, + "step": 31559 + }, + { + "epoch": 0.6946951353731015, + "grad_norm": 0.6185095992216739, + "learning_rate": 2.2515782705805787e-06, + "loss": 0.4694, + "step": 31560 + }, + { + "epoch": 0.6947171472595202, + "grad_norm": 1.219857069073866, + "learning_rate": 2.251280503728469e-06, + "loss": 0.2827, + "step": 31561 + }, + { + "epoch": 0.6947391591459388, + "grad_norm": 1.5435485027681517, + "learning_rate": 2.250982750846576e-06, + "loss": 0.3536, + "step": 31562 + }, + { + "epoch": 0.6947611710323575, + "grad_norm": 1.6118113595581873, + "learning_rate": 2.250685011936416e-06, + "loss": 0.3556, + "step": 31563 + }, + { + "epoch": 0.6947831829187762, + "grad_norm": 1.4511318125285046, + "learning_rate": 2.2503872869994996e-06, + "loss": 0.296, + "step": 31564 + }, + { + "epoch": 0.6948051948051948, + "grad_norm": 2.003143014226349, + "learning_rate": 2.2500895760373416e-06, + "loss": 0.2857, + "step": 31565 + }, + { + "epoch": 0.6948272066916135, + "grad_norm": 1.7415363005023379, + "learning_rate": 2.2497918790514563e-06, + "loss": 0.28, + "step": 31566 + }, + { + "epoch": 0.6948492185780322, + "grad_norm": 2.8349768279624565, + "learning_rate": 2.2494941960433553e-06, + "loss": 0.2611, + "step": 31567 + }, + { + "epoch": 0.6948712304644508, + "grad_norm": 1.6025999673361468, + "learning_rate": 2.24919652701455e-06, + "loss": 0.2922, + "step": 31568 + }, + { + "epoch": 0.6948932423508695, + "grad_norm": 1.9987796381620497, + "learning_rate": 2.248898871966555e-06, + "loss": 0.3321, + "step": 31569 + }, + { + "epoch": 0.6949152542372882, + "grad_norm": 2.624560173449749, + "learning_rate": 2.2486012309008826e-06, + "loss": 0.2978, + "step": 31570 + }, + { + "epoch": 0.6949372661237068, + "grad_norm": 0.6268319538649273, + "learning_rate": 2.2483036038190476e-06, + "loss": 0.4729, + "step": 31571 + }, + { + "epoch": 0.6949592780101255, + "grad_norm": 0.6145120090905938, + "learning_rate": 2.248005990722561e-06, + "loss": 0.4526, + "step": 31572 + }, + { + "epoch": 0.6949812898965442, + "grad_norm": 2.1163002351901286, + "learning_rate": 2.247708391612934e-06, + "loss": 0.3668, + "step": 31573 + }, + { + "epoch": 0.6950033017829628, + "grad_norm": 1.9480254640291639, + "learning_rate": 2.24741080649168e-06, + "loss": 0.3215, + "step": 31574 + }, + { + "epoch": 0.6950253136693815, + "grad_norm": 1.631066177653967, + "learning_rate": 2.2471132353603125e-06, + "loss": 0.3088, + "step": 31575 + }, + { + "epoch": 0.6950473255558002, + "grad_norm": 1.9210047516632167, + "learning_rate": 2.2468156782203448e-06, + "loss": 0.3848, + "step": 31576 + }, + { + "epoch": 0.6950693374422188, + "grad_norm": 1.9236616420188255, + "learning_rate": 2.2465181350732877e-06, + "loss": 0.3247, + "step": 31577 + }, + { + "epoch": 0.6950913493286375, + "grad_norm": 3.517148343942666, + "learning_rate": 2.2462206059206523e-06, + "loss": 0.3379, + "step": 31578 + }, + { + "epoch": 0.6951133612150562, + "grad_norm": 3.273719168026533, + "learning_rate": 2.2459230907639514e-06, + "loss": 0.2504, + "step": 31579 + }, + { + "epoch": 0.6951353731014748, + "grad_norm": 0.5966373485395158, + "learning_rate": 2.245625589604698e-06, + "loss": 0.4488, + "step": 31580 + }, + { + "epoch": 0.6951573849878935, + "grad_norm": 0.5929967344843127, + "learning_rate": 2.245328102444405e-06, + "loss": 0.4563, + "step": 31581 + }, + { + "epoch": 0.6951793968743122, + "grad_norm": 0.6160934026425928, + "learning_rate": 2.245030629284584e-06, + "loss": 0.4585, + "step": 31582 + }, + { + "epoch": 0.6952014087607308, + "grad_norm": 1.640548030325866, + "learning_rate": 2.2447331701267438e-06, + "loss": 0.2479, + "step": 31583 + }, + { + "epoch": 0.6952234206471495, + "grad_norm": 2.070458108181842, + "learning_rate": 2.244435724972399e-06, + "loss": 0.3094, + "step": 31584 + }, + { + "epoch": 0.6952454325335681, + "grad_norm": 1.4961831015018938, + "learning_rate": 2.244138293823062e-06, + "loss": 0.2199, + "step": 31585 + }, + { + "epoch": 0.6952674444199868, + "grad_norm": 1.3573210281371109, + "learning_rate": 2.243840876680242e-06, + "loss": 0.3027, + "step": 31586 + }, + { + "epoch": 0.6952894563064055, + "grad_norm": 0.6130865634202932, + "learning_rate": 2.2435434735454543e-06, + "loss": 0.4559, + "step": 31587 + }, + { + "epoch": 0.6953114681928241, + "grad_norm": 1.76768607930032, + "learning_rate": 2.2432460844202053e-06, + "loss": 0.3182, + "step": 31588 + }, + { + "epoch": 0.6953334800792428, + "grad_norm": 1.4754253836485405, + "learning_rate": 2.24294870930601e-06, + "loss": 0.3074, + "step": 31589 + }, + { + "epoch": 0.6953554919656615, + "grad_norm": 2.8000359031156643, + "learning_rate": 2.24265134820438e-06, + "loss": 0.3114, + "step": 31590 + }, + { + "epoch": 0.6953775038520801, + "grad_norm": 1.5351886837509776, + "learning_rate": 2.2423540011168243e-06, + "loss": 0.3185, + "step": 31591 + }, + { + "epoch": 0.6953995157384988, + "grad_norm": 1.771931659896061, + "learning_rate": 2.2420566680448576e-06, + "loss": 0.2921, + "step": 31592 + }, + { + "epoch": 0.6954215276249175, + "grad_norm": 2.3744693153313867, + "learning_rate": 2.241759348989987e-06, + "loss": 0.3186, + "step": 31593 + }, + { + "epoch": 0.6954435395113361, + "grad_norm": 0.6216780961309075, + "learning_rate": 2.2414620439537253e-06, + "loss": 0.4403, + "step": 31594 + }, + { + "epoch": 0.6954655513977548, + "grad_norm": 1.369033492315519, + "learning_rate": 2.2411647529375857e-06, + "loss": 0.2898, + "step": 31595 + }, + { + "epoch": 0.6954875632841735, + "grad_norm": 1.7364266421952232, + "learning_rate": 2.2408674759430753e-06, + "loss": 0.2242, + "step": 31596 + }, + { + "epoch": 0.6955095751705921, + "grad_norm": 1.9205417042975286, + "learning_rate": 2.2405702129717096e-06, + "loss": 0.376, + "step": 31597 + }, + { + "epoch": 0.6955315870570108, + "grad_norm": 1.6340523121730806, + "learning_rate": 2.2402729640249944e-06, + "loss": 0.3643, + "step": 31598 + }, + { + "epoch": 0.6955535989434295, + "grad_norm": 0.6065990743959611, + "learning_rate": 2.2399757291044448e-06, + "loss": 0.4489, + "step": 31599 + }, + { + "epoch": 0.6955756108298481, + "grad_norm": 0.6282031212141609, + "learning_rate": 2.239678508211568e-06, + "loss": 0.457, + "step": 31600 + }, + { + "epoch": 0.6955976227162668, + "grad_norm": 3.0545974102671143, + "learning_rate": 2.239381301347876e-06, + "loss": 0.3096, + "step": 31601 + }, + { + "epoch": 0.6956196346026855, + "grad_norm": 1.6482319610416658, + "learning_rate": 2.239084108514881e-06, + "loss": 0.3016, + "step": 31602 + }, + { + "epoch": 0.6956416464891041, + "grad_norm": 1.82013327849988, + "learning_rate": 2.2387869297140903e-06, + "loss": 0.3051, + "step": 31603 + }, + { + "epoch": 0.6956636583755228, + "grad_norm": 2.377866875581953, + "learning_rate": 2.238489764947018e-06, + "loss": 0.2882, + "step": 31604 + }, + { + "epoch": 0.6956856702619415, + "grad_norm": 2.105511571115168, + "learning_rate": 2.2381926142151706e-06, + "loss": 0.2454, + "step": 31605 + }, + { + "epoch": 0.6957076821483601, + "grad_norm": 1.5570776105511523, + "learning_rate": 2.2378954775200594e-06, + "loss": 0.3, + "step": 31606 + }, + { + "epoch": 0.6957296940347788, + "grad_norm": 2.2178701868547512, + "learning_rate": 2.237598354863198e-06, + "loss": 0.3231, + "step": 31607 + }, + { + "epoch": 0.6957517059211974, + "grad_norm": 1.9302788223981246, + "learning_rate": 2.2373012462460934e-06, + "loss": 0.2502, + "step": 31608 + }, + { + "epoch": 0.6957737178076161, + "grad_norm": 1.5346802537719537, + "learning_rate": 2.2370041516702545e-06, + "loss": 0.2557, + "step": 31609 + }, + { + "epoch": 0.6957957296940348, + "grad_norm": 1.5648120588702037, + "learning_rate": 2.236707071137193e-06, + "loss": 0.2835, + "step": 31610 + }, + { + "epoch": 0.6958177415804534, + "grad_norm": 1.4863123952875859, + "learning_rate": 2.2364100046484183e-06, + "loss": 0.2977, + "step": 31611 + }, + { + "epoch": 0.6958397534668721, + "grad_norm": 1.4395516705604525, + "learning_rate": 2.236112952205442e-06, + "loss": 0.2787, + "step": 31612 + }, + { + "epoch": 0.6958617653532908, + "grad_norm": 2.1770866779647484, + "learning_rate": 2.235815913809773e-06, + "loss": 0.2538, + "step": 31613 + }, + { + "epoch": 0.6958837772397094, + "grad_norm": 1.2769125550681528, + "learning_rate": 2.235518889462918e-06, + "loss": 0.2951, + "step": 31614 + }, + { + "epoch": 0.6959057891261281, + "grad_norm": 2.2857618995131626, + "learning_rate": 2.23522187916639e-06, + "loss": 0.2626, + "step": 31615 + }, + { + "epoch": 0.6959278010125468, + "grad_norm": 2.1116528037495716, + "learning_rate": 2.234924882921697e-06, + "loss": 0.357, + "step": 31616 + }, + { + "epoch": 0.6959498128989654, + "grad_norm": 1.7638725882616195, + "learning_rate": 2.23462790073035e-06, + "loss": 0.3148, + "step": 31617 + }, + { + "epoch": 0.6959718247853841, + "grad_norm": 0.6349383874731465, + "learning_rate": 2.2343309325938585e-06, + "loss": 0.4573, + "step": 31618 + }, + { + "epoch": 0.6959938366718028, + "grad_norm": 0.6199380110866419, + "learning_rate": 2.2340339785137284e-06, + "loss": 0.4629, + "step": 31619 + }, + { + "epoch": 0.6960158485582214, + "grad_norm": 1.4526499273159081, + "learning_rate": 2.2337370384914716e-06, + "loss": 0.2975, + "step": 31620 + }, + { + "epoch": 0.6960378604446401, + "grad_norm": 1.3259807928347964, + "learning_rate": 2.2334401125285963e-06, + "loss": 0.3753, + "step": 31621 + }, + { + "epoch": 0.6960598723310588, + "grad_norm": 1.8329194269306877, + "learning_rate": 2.2331432006266145e-06, + "loss": 0.3124, + "step": 31622 + }, + { + "epoch": 0.6960818842174774, + "grad_norm": 3.603884523700409, + "learning_rate": 2.2328463027870334e-06, + "loss": 0.3211, + "step": 31623 + }, + { + "epoch": 0.6961038961038961, + "grad_norm": 2.1060588793218513, + "learning_rate": 2.2325494190113596e-06, + "loss": 0.2072, + "step": 31624 + }, + { + "epoch": 0.6961259079903148, + "grad_norm": 2.153576926630478, + "learning_rate": 2.2322525493011036e-06, + "loss": 0.3731, + "step": 31625 + }, + { + "epoch": 0.6961479198767334, + "grad_norm": 1.561975400115179, + "learning_rate": 2.231955693657777e-06, + "loss": 0.2873, + "step": 31626 + }, + { + "epoch": 0.6961699317631521, + "grad_norm": 0.6009959850375843, + "learning_rate": 2.2316588520828834e-06, + "loss": 0.4465, + "step": 31627 + }, + { + "epoch": 0.6961919436495708, + "grad_norm": 1.9546525939042763, + "learning_rate": 2.2313620245779367e-06, + "loss": 0.2923, + "step": 31628 + }, + { + "epoch": 0.6962139555359894, + "grad_norm": 1.4178730056345883, + "learning_rate": 2.231065211144441e-06, + "loss": 0.3008, + "step": 31629 + }, + { + "epoch": 0.6962359674224081, + "grad_norm": 1.3586269365950976, + "learning_rate": 2.230768411783909e-06, + "loss": 0.2404, + "step": 31630 + }, + { + "epoch": 0.6962579793088267, + "grad_norm": 1.5872781186986185, + "learning_rate": 2.2304716264978447e-06, + "loss": 0.333, + "step": 31631 + }, + { + "epoch": 0.6962799911952454, + "grad_norm": 3.6095832939431527, + "learning_rate": 2.2301748552877593e-06, + "loss": 0.3363, + "step": 31632 + }, + { + "epoch": 0.6963020030816641, + "grad_norm": 1.9713886416220197, + "learning_rate": 2.229878098155162e-06, + "loss": 0.2689, + "step": 31633 + }, + { + "epoch": 0.6963240149680827, + "grad_norm": 15.493007347917391, + "learning_rate": 2.229581355101559e-06, + "loss": 0.2858, + "step": 31634 + }, + { + "epoch": 0.6963460268545014, + "grad_norm": 3.462677503103225, + "learning_rate": 2.22928462612846e-06, + "loss": 0.3679, + "step": 31635 + }, + { + "epoch": 0.6963680387409201, + "grad_norm": 1.636874467625807, + "learning_rate": 2.228987911237371e-06, + "loss": 0.265, + "step": 31636 + }, + { + "epoch": 0.6963900506273387, + "grad_norm": 1.718963111427228, + "learning_rate": 2.2286912104298002e-06, + "loss": 0.2578, + "step": 31637 + }, + { + "epoch": 0.6964120625137574, + "grad_norm": 1.966062105077743, + "learning_rate": 2.2283945237072597e-06, + "loss": 0.3173, + "step": 31638 + }, + { + "epoch": 0.6964340744001761, + "grad_norm": 1.9142900468427542, + "learning_rate": 2.2280978510712527e-06, + "loss": 0.3509, + "step": 31639 + }, + { + "epoch": 0.6964560862865947, + "grad_norm": 1.5804728162974393, + "learning_rate": 2.2278011925232895e-06, + "loss": 0.2227, + "step": 31640 + }, + { + "epoch": 0.6964780981730134, + "grad_norm": 1.4290438703577473, + "learning_rate": 2.227504548064876e-06, + "loss": 0.2867, + "step": 31641 + }, + { + "epoch": 0.6965001100594321, + "grad_norm": 0.6166891659425495, + "learning_rate": 2.227207917697521e-06, + "loss": 0.46, + "step": 31642 + }, + { + "epoch": 0.6965221219458507, + "grad_norm": 1.5352381558102473, + "learning_rate": 2.2269113014227334e-06, + "loss": 0.2787, + "step": 31643 + }, + { + "epoch": 0.6965441338322694, + "grad_norm": 4.91916734765409, + "learning_rate": 2.2266146992420177e-06, + "loss": 0.387, + "step": 31644 + }, + { + "epoch": 0.6965661457186881, + "grad_norm": 1.7958567126118998, + "learning_rate": 2.2263181111568852e-06, + "loss": 0.3064, + "step": 31645 + }, + { + "epoch": 0.6965881576051067, + "grad_norm": 0.6230738968519438, + "learning_rate": 2.226021537168839e-06, + "loss": 0.4581, + "step": 31646 + }, + { + "epoch": 0.6966101694915254, + "grad_norm": 1.6698649843411084, + "learning_rate": 2.225724977279389e-06, + "loss": 0.3327, + "step": 31647 + }, + { + "epoch": 0.6966321813779441, + "grad_norm": 1.209787794103491, + "learning_rate": 2.225428431490044e-06, + "loss": 0.2143, + "step": 31648 + }, + { + "epoch": 0.6966541932643627, + "grad_norm": 2.4142506115752265, + "learning_rate": 2.2251318998023086e-06, + "loss": 0.3431, + "step": 31649 + }, + { + "epoch": 0.6966762051507814, + "grad_norm": 1.4029537079452625, + "learning_rate": 2.224835382217689e-06, + "loss": 0.2718, + "step": 31650 + }, + { + "epoch": 0.6966982170372001, + "grad_norm": 1.6095551176645537, + "learning_rate": 2.2245388787376935e-06, + "loss": 0.203, + "step": 31651 + }, + { + "epoch": 0.6967202289236187, + "grad_norm": 1.5473899645457152, + "learning_rate": 2.22424238936383e-06, + "loss": 0.3665, + "step": 31652 + }, + { + "epoch": 0.6967422408100374, + "grad_norm": 1.7702529343227618, + "learning_rate": 2.2239459140976056e-06, + "loss": 0.3513, + "step": 31653 + }, + { + "epoch": 0.696764252696456, + "grad_norm": 2.062104977936053, + "learning_rate": 2.2236494529405268e-06, + "loss": 0.3083, + "step": 31654 + }, + { + "epoch": 0.6967862645828747, + "grad_norm": 1.1979066743400708, + "learning_rate": 2.2233530058940978e-06, + "loss": 0.2333, + "step": 31655 + }, + { + "epoch": 0.6968082764692934, + "grad_norm": 1.6645633207636623, + "learning_rate": 2.2230565729598268e-06, + "loss": 0.3476, + "step": 31656 + }, + { + "epoch": 0.696830288355712, + "grad_norm": 2.591166813263715, + "learning_rate": 2.222760154139221e-06, + "loss": 0.2547, + "step": 31657 + }, + { + "epoch": 0.6968523002421307, + "grad_norm": 1.6679199584041093, + "learning_rate": 2.2224637494337893e-06, + "loss": 0.2777, + "step": 31658 + }, + { + "epoch": 0.6968743121285494, + "grad_norm": 1.471124292583673, + "learning_rate": 2.2221673588450353e-06, + "loss": 0.3033, + "step": 31659 + }, + { + "epoch": 0.696896324014968, + "grad_norm": 0.6267750080314647, + "learning_rate": 2.2218709823744638e-06, + "loss": 0.4576, + "step": 31660 + }, + { + "epoch": 0.6969183359013867, + "grad_norm": 1.9349308743421723, + "learning_rate": 2.2215746200235825e-06, + "loss": 0.2631, + "step": 31661 + }, + { + "epoch": 0.6969403477878054, + "grad_norm": 1.7996588157870224, + "learning_rate": 2.2212782717938993e-06, + "loss": 0.2814, + "step": 31662 + }, + { + "epoch": 0.696962359674224, + "grad_norm": 1.7984428239400476, + "learning_rate": 2.220981937686921e-06, + "loss": 0.2849, + "step": 31663 + }, + { + "epoch": 0.6969843715606427, + "grad_norm": 1.8563670132952643, + "learning_rate": 2.2206856177041508e-06, + "loss": 0.2458, + "step": 31664 + }, + { + "epoch": 0.6970063834470615, + "grad_norm": 2.0232786225288764, + "learning_rate": 2.220389311847095e-06, + "loss": 0.3416, + "step": 31665 + }, + { + "epoch": 0.69702839533348, + "grad_norm": 2.2577818843585065, + "learning_rate": 2.220093020117262e-06, + "loss": 0.3264, + "step": 31666 + }, + { + "epoch": 0.6970504072198987, + "grad_norm": 1.254063179797681, + "learning_rate": 2.219796742516154e-06, + "loss": 0.251, + "step": 31667 + }, + { + "epoch": 0.6970724191063175, + "grad_norm": 2.2735400106249566, + "learning_rate": 2.2195004790452797e-06, + "loss": 0.2638, + "step": 31668 + }, + { + "epoch": 0.697094430992736, + "grad_norm": 1.3794387459605697, + "learning_rate": 2.2192042297061454e-06, + "loss": 0.2212, + "step": 31669 + }, + { + "epoch": 0.6971164428791548, + "grad_norm": 1.6733946384346818, + "learning_rate": 2.218907994500254e-06, + "loss": 0.2399, + "step": 31670 + }, + { + "epoch": 0.6971384547655735, + "grad_norm": 8.44204030310003, + "learning_rate": 2.218611773429114e-06, + "loss": 0.2805, + "step": 31671 + }, + { + "epoch": 0.697160466651992, + "grad_norm": 1.9392021656701284, + "learning_rate": 2.218315566494228e-06, + "loss": 0.2536, + "step": 31672 + }, + { + "epoch": 0.6971824785384108, + "grad_norm": 2.690154519580411, + "learning_rate": 2.2180193736971027e-06, + "loss": 0.2627, + "step": 31673 + }, + { + "epoch": 0.6972044904248295, + "grad_norm": 1.9525824854914557, + "learning_rate": 2.217723195039246e-06, + "loss": 0.3675, + "step": 31674 + }, + { + "epoch": 0.697226502311248, + "grad_norm": 1.8083030864330487, + "learning_rate": 2.217427030522159e-06, + "loss": 0.3366, + "step": 31675 + }, + { + "epoch": 0.6972485141976668, + "grad_norm": 0.6127252358219037, + "learning_rate": 2.21713088014735e-06, + "loss": 0.4548, + "step": 31676 + }, + { + "epoch": 0.6972705260840854, + "grad_norm": 5.852508221488295, + "learning_rate": 2.2168347439163224e-06, + "loss": 0.2579, + "step": 31677 + }, + { + "epoch": 0.6972925379705041, + "grad_norm": 1.538582789595988, + "learning_rate": 2.216538621830581e-06, + "loss": 0.3193, + "step": 31678 + }, + { + "epoch": 0.6973145498569228, + "grad_norm": 2.241998535898161, + "learning_rate": 2.2162425138916343e-06, + "loss": 0.3314, + "step": 31679 + }, + { + "epoch": 0.6973365617433414, + "grad_norm": 0.6148915574650717, + "learning_rate": 2.215946420100982e-06, + "loss": 0.4531, + "step": 31680 + }, + { + "epoch": 0.6973585736297601, + "grad_norm": 1.5295309038431268, + "learning_rate": 2.2156503404601342e-06, + "loss": 0.2727, + "step": 31681 + }, + { + "epoch": 0.6973805855161788, + "grad_norm": 1.2505225623470995, + "learning_rate": 2.2153542749705913e-06, + "loss": 0.2506, + "step": 31682 + }, + { + "epoch": 0.6974025974025974, + "grad_norm": 1.3023421689014893, + "learning_rate": 2.21505822363386e-06, + "loss": 0.2351, + "step": 31683 + }, + { + "epoch": 0.6974246092890161, + "grad_norm": 1.5370793802892855, + "learning_rate": 2.2147621864514468e-06, + "loss": 0.2619, + "step": 31684 + }, + { + "epoch": 0.6974466211754348, + "grad_norm": 0.6181274452426435, + "learning_rate": 2.2144661634248543e-06, + "loss": 0.4476, + "step": 31685 + }, + { + "epoch": 0.6974686330618534, + "grad_norm": 2.3430704378043754, + "learning_rate": 2.2141701545555853e-06, + "loss": 0.2472, + "step": 31686 + }, + { + "epoch": 0.6974906449482721, + "grad_norm": 0.6131251553364874, + "learning_rate": 2.2138741598451462e-06, + "loss": 0.4351, + "step": 31687 + }, + { + "epoch": 0.6975126568346908, + "grad_norm": 1.514551135253641, + "learning_rate": 2.213578179295042e-06, + "loss": 0.2917, + "step": 31688 + }, + { + "epoch": 0.6975346687211094, + "grad_norm": 2.9979981846651143, + "learning_rate": 2.2132822129067766e-06, + "loss": 0.338, + "step": 31689 + }, + { + "epoch": 0.6975566806075281, + "grad_norm": 1.6550299969008828, + "learning_rate": 2.2129862606818547e-06, + "loss": 0.3179, + "step": 31690 + }, + { + "epoch": 0.6975786924939468, + "grad_norm": 2.9613082882696924, + "learning_rate": 2.212690322621778e-06, + "loss": 0.2803, + "step": 31691 + }, + { + "epoch": 0.6976007043803654, + "grad_norm": 1.780677694684729, + "learning_rate": 2.2123943987280515e-06, + "loss": 0.3399, + "step": 31692 + }, + { + "epoch": 0.6976227162667841, + "grad_norm": 1.7165234467121142, + "learning_rate": 2.2120984890021806e-06, + "loss": 0.2797, + "step": 31693 + }, + { + "epoch": 0.6976447281532028, + "grad_norm": 1.4053804373553527, + "learning_rate": 2.2118025934456703e-06, + "loss": 0.2998, + "step": 31694 + }, + { + "epoch": 0.6976667400396214, + "grad_norm": 4.553831794082257, + "learning_rate": 2.211506712060023e-06, + "loss": 0.2931, + "step": 31695 + }, + { + "epoch": 0.6976887519260401, + "grad_norm": 4.604095264351967, + "learning_rate": 2.2112108448467408e-06, + "loss": 0.3026, + "step": 31696 + }, + { + "epoch": 0.6977107638124588, + "grad_norm": 0.6126269579186381, + "learning_rate": 2.210914991807328e-06, + "loss": 0.4483, + "step": 31697 + }, + { + "epoch": 0.6977327756988774, + "grad_norm": 2.1455777260793583, + "learning_rate": 2.21061915294329e-06, + "loss": 0.3003, + "step": 31698 + }, + { + "epoch": 0.6977547875852961, + "grad_norm": 1.2510947959785992, + "learning_rate": 2.2103233282561302e-06, + "loss": 0.2062, + "step": 31699 + }, + { + "epoch": 0.6977767994717147, + "grad_norm": 1.6338286242691837, + "learning_rate": 2.210027517747352e-06, + "loss": 0.2461, + "step": 31700 + }, + { + "epoch": 0.6977988113581334, + "grad_norm": 1.4668540199417128, + "learning_rate": 2.209731721418457e-06, + "loss": 0.3357, + "step": 31701 + }, + { + "epoch": 0.6978208232445521, + "grad_norm": 1.7045089625608316, + "learning_rate": 2.209435939270952e-06, + "loss": 0.3024, + "step": 31702 + }, + { + "epoch": 0.6978428351309707, + "grad_norm": 1.7190538163612765, + "learning_rate": 2.2091401713063336e-06, + "loss": 0.2627, + "step": 31703 + }, + { + "epoch": 0.6978648470173894, + "grad_norm": 1.9370256046031586, + "learning_rate": 2.2088444175261138e-06, + "loss": 0.336, + "step": 31704 + }, + { + "epoch": 0.6978868589038081, + "grad_norm": 4.603511877055595, + "learning_rate": 2.2085486779317917e-06, + "loss": 0.3385, + "step": 31705 + }, + { + "epoch": 0.6979088707902267, + "grad_norm": 2.316997596478765, + "learning_rate": 2.208252952524868e-06, + "loss": 0.2872, + "step": 31706 + }, + { + "epoch": 0.6979308826766454, + "grad_norm": 0.6387060548552637, + "learning_rate": 2.2079572413068496e-06, + "loss": 0.4577, + "step": 31707 + }, + { + "epoch": 0.6979528945630641, + "grad_norm": 1.5617808046328692, + "learning_rate": 2.2076615442792366e-06, + "loss": 0.2722, + "step": 31708 + }, + { + "epoch": 0.6979749064494827, + "grad_norm": 3.3621699034489407, + "learning_rate": 2.2073658614435323e-06, + "loss": 0.3531, + "step": 31709 + }, + { + "epoch": 0.6979969183359014, + "grad_norm": 1.6254619355382676, + "learning_rate": 2.2070701928012426e-06, + "loss": 0.3031, + "step": 31710 + }, + { + "epoch": 0.6980189302223201, + "grad_norm": 2.3233043422219755, + "learning_rate": 2.2067745383538658e-06, + "loss": 0.3865, + "step": 31711 + }, + { + "epoch": 0.6980409421087387, + "grad_norm": 0.6383398234755039, + "learning_rate": 2.206478898102909e-06, + "loss": 0.4655, + "step": 31712 + }, + { + "epoch": 0.6980629539951574, + "grad_norm": 1.5055154791388894, + "learning_rate": 2.20618327204987e-06, + "loss": 0.319, + "step": 31713 + }, + { + "epoch": 0.6980849658815761, + "grad_norm": 2.0028515182919713, + "learning_rate": 2.2058876601962538e-06, + "loss": 0.291, + "step": 31714 + }, + { + "epoch": 0.6981069777679947, + "grad_norm": 1.9706164838413358, + "learning_rate": 2.2055920625435647e-06, + "loss": 0.2854, + "step": 31715 + }, + { + "epoch": 0.6981289896544134, + "grad_norm": 2.0764393169266766, + "learning_rate": 2.2052964790933006e-06, + "loss": 0.275, + "step": 31716 + }, + { + "epoch": 0.6981510015408321, + "grad_norm": 1.800111553993322, + "learning_rate": 2.205000909846969e-06, + "loss": 0.1996, + "step": 31717 + }, + { + "epoch": 0.6981730134272507, + "grad_norm": 1.7086724472847539, + "learning_rate": 2.204705354806067e-06, + "loss": 0.3703, + "step": 31718 + }, + { + "epoch": 0.6981950253136694, + "grad_norm": 1.3661641697912446, + "learning_rate": 2.2044098139720994e-06, + "loss": 0.2102, + "step": 31719 + }, + { + "epoch": 0.6982170372000881, + "grad_norm": 2.3019167259140505, + "learning_rate": 2.204114287346569e-06, + "loss": 0.2448, + "step": 31720 + }, + { + "epoch": 0.6982390490865067, + "grad_norm": 1.8498465257615844, + "learning_rate": 2.2038187749309757e-06, + "loss": 0.331, + "step": 31721 + }, + { + "epoch": 0.6982610609729254, + "grad_norm": 2.325574458021851, + "learning_rate": 2.2035232767268245e-06, + "loss": 0.3033, + "step": 31722 + }, + { + "epoch": 0.6982830728593441, + "grad_norm": 4.2016739703333785, + "learning_rate": 2.2032277927356133e-06, + "loss": 0.3133, + "step": 31723 + }, + { + "epoch": 0.6983050847457627, + "grad_norm": 0.6286815853165403, + "learning_rate": 2.202932322958846e-06, + "loss": 0.465, + "step": 31724 + }, + { + "epoch": 0.6983270966321814, + "grad_norm": 1.4008033668362414, + "learning_rate": 2.2026368673980258e-06, + "loss": 0.2754, + "step": 31725 + }, + { + "epoch": 0.6983491085186, + "grad_norm": 1.9371353112890135, + "learning_rate": 2.202341426054653e-06, + "loss": 0.3749, + "step": 31726 + }, + { + "epoch": 0.6983711204050187, + "grad_norm": 0.600205742461332, + "learning_rate": 2.202045998930226e-06, + "loss": 0.4358, + "step": 31727 + }, + { + "epoch": 0.6983931322914374, + "grad_norm": 1.7813496224003333, + "learning_rate": 2.20175058602625e-06, + "loss": 0.3601, + "step": 31728 + }, + { + "epoch": 0.698415144177856, + "grad_norm": 1.163112536099002, + "learning_rate": 2.201455187344225e-06, + "loss": 0.2697, + "step": 31729 + }, + { + "epoch": 0.6984371560642747, + "grad_norm": 1.929784783439255, + "learning_rate": 2.201159802885656e-06, + "loss": 0.3089, + "step": 31730 + }, + { + "epoch": 0.6984591679506934, + "grad_norm": 0.6350135436904034, + "learning_rate": 2.20086443265204e-06, + "loss": 0.4513, + "step": 31731 + }, + { + "epoch": 0.698481179837112, + "grad_norm": 1.4126608546867048, + "learning_rate": 2.2005690766448777e-06, + "loss": 0.2209, + "step": 31732 + }, + { + "epoch": 0.6985031917235307, + "grad_norm": 1.9051613981951387, + "learning_rate": 2.200273734865673e-06, + "loss": 0.3032, + "step": 31733 + }, + { + "epoch": 0.6985252036099494, + "grad_norm": 1.1236383068889884, + "learning_rate": 2.199978407315925e-06, + "loss": 0.271, + "step": 31734 + }, + { + "epoch": 0.698547215496368, + "grad_norm": 2.0430506729454123, + "learning_rate": 2.1996830939971368e-06, + "loss": 0.3089, + "step": 31735 + }, + { + "epoch": 0.6985692273827867, + "grad_norm": 1.3175448137100285, + "learning_rate": 2.199387794910809e-06, + "loss": 0.2376, + "step": 31736 + }, + { + "epoch": 0.6985912392692054, + "grad_norm": 1.6329312026374618, + "learning_rate": 2.1990925100584397e-06, + "loss": 0.3136, + "step": 31737 + }, + { + "epoch": 0.698613251155624, + "grad_norm": 1.573977373025901, + "learning_rate": 2.198797239441533e-06, + "loss": 0.2637, + "step": 31738 + }, + { + "epoch": 0.6986352630420427, + "grad_norm": 0.6509538466893207, + "learning_rate": 2.1985019830615843e-06, + "loss": 0.4723, + "step": 31739 + }, + { + "epoch": 0.6986572749284614, + "grad_norm": 1.3060870499491442, + "learning_rate": 2.1982067409201025e-06, + "loss": 0.2824, + "step": 31740 + }, + { + "epoch": 0.69867928681488, + "grad_norm": 1.42716147585678, + "learning_rate": 2.197911513018583e-06, + "loss": 0.305, + "step": 31741 + }, + { + "epoch": 0.6987012987012987, + "grad_norm": 0.619637330486564, + "learning_rate": 2.197616299358525e-06, + "loss": 0.4533, + "step": 31742 + }, + { + "epoch": 0.6987233105877174, + "grad_norm": 1.6584167592357946, + "learning_rate": 2.1973210999414328e-06, + "loss": 0.4218, + "step": 31743 + }, + { + "epoch": 0.698745322474136, + "grad_norm": 2.387067577476987, + "learning_rate": 2.197025914768803e-06, + "loss": 0.2299, + "step": 31744 + }, + { + "epoch": 0.6987673343605547, + "grad_norm": 1.6599822828298834, + "learning_rate": 2.1967307438421383e-06, + "loss": 0.19, + "step": 31745 + }, + { + "epoch": 0.6987893462469734, + "grad_norm": 1.3894701558040325, + "learning_rate": 2.1964355871629394e-06, + "loss": 0.3345, + "step": 31746 + }, + { + "epoch": 0.698811358133392, + "grad_norm": 1.512184021576493, + "learning_rate": 2.196140444732704e-06, + "loss": 0.2832, + "step": 31747 + }, + { + "epoch": 0.6988333700198107, + "grad_norm": 1.5781263550150708, + "learning_rate": 2.195845316552935e-06, + "loss": 0.3518, + "step": 31748 + }, + { + "epoch": 0.6988553819062293, + "grad_norm": 2.3183470777566075, + "learning_rate": 2.1955502026251287e-06, + "loss": 0.3083, + "step": 31749 + }, + { + "epoch": 0.698877393792648, + "grad_norm": 1.7122682747430087, + "learning_rate": 2.195255102950787e-06, + "loss": 0.3872, + "step": 31750 + }, + { + "epoch": 0.6988994056790667, + "grad_norm": 1.2296564406346449, + "learning_rate": 2.194960017531412e-06, + "loss": 0.3092, + "step": 31751 + }, + { + "epoch": 0.6989214175654853, + "grad_norm": 1.5876455293522613, + "learning_rate": 2.1946649463684988e-06, + "loss": 0.2602, + "step": 31752 + }, + { + "epoch": 0.698943429451904, + "grad_norm": 0.6286538013021015, + "learning_rate": 2.1943698894635524e-06, + "loss": 0.4581, + "step": 31753 + }, + { + "epoch": 0.6989654413383227, + "grad_norm": 1.4060481990651763, + "learning_rate": 2.1940748468180673e-06, + "loss": 0.2371, + "step": 31754 + }, + { + "epoch": 0.6989874532247413, + "grad_norm": 1.3628372685845824, + "learning_rate": 2.1937798184335453e-06, + "loss": 0.218, + "step": 31755 + }, + { + "epoch": 0.69900946511116, + "grad_norm": 1.529567001938638, + "learning_rate": 2.193484804311487e-06, + "loss": 0.3026, + "step": 31756 + }, + { + "epoch": 0.6990314769975787, + "grad_norm": 1.9063541988081956, + "learning_rate": 2.1931898044533895e-06, + "loss": 0.3805, + "step": 31757 + }, + { + "epoch": 0.6990534888839973, + "grad_norm": 1.2443551517326958, + "learning_rate": 2.192894818860755e-06, + "loss": 0.3429, + "step": 31758 + }, + { + "epoch": 0.699075500770416, + "grad_norm": 4.62343532558691, + "learning_rate": 2.1925998475350796e-06, + "loss": 0.2658, + "step": 31759 + }, + { + "epoch": 0.6990975126568347, + "grad_norm": 1.4110086642033215, + "learning_rate": 2.1923048904778637e-06, + "loss": 0.2246, + "step": 31760 + }, + { + "epoch": 0.6991195245432533, + "grad_norm": 1.4759860748754887, + "learning_rate": 2.192009947690608e-06, + "loss": 0.3332, + "step": 31761 + }, + { + "epoch": 0.699141536429672, + "grad_norm": 1.4675486483051081, + "learning_rate": 2.1917150191748086e-06, + "loss": 0.3231, + "step": 31762 + }, + { + "epoch": 0.6991635483160907, + "grad_norm": 0.6393457813575414, + "learning_rate": 2.1914201049319683e-06, + "loss": 0.4624, + "step": 31763 + }, + { + "epoch": 0.6991855602025093, + "grad_norm": 1.484345027129744, + "learning_rate": 2.1911252049635813e-06, + "loss": 0.2876, + "step": 31764 + }, + { + "epoch": 0.699207572088928, + "grad_norm": 1.5985812519279194, + "learning_rate": 2.190830319271149e-06, + "loss": 0.2681, + "step": 31765 + }, + { + "epoch": 0.6992295839753467, + "grad_norm": 1.5702046724700738, + "learning_rate": 2.190535447856172e-06, + "loss": 0.2476, + "step": 31766 + }, + { + "epoch": 0.6992515958617653, + "grad_norm": 1.6921966332807288, + "learning_rate": 2.1902405907201464e-06, + "loss": 0.3042, + "step": 31767 + }, + { + "epoch": 0.699273607748184, + "grad_norm": 1.550483168190632, + "learning_rate": 2.1899457478645695e-06, + "loss": 0.3126, + "step": 31768 + }, + { + "epoch": 0.6992956196346027, + "grad_norm": 3.930550562146754, + "learning_rate": 2.189650919290942e-06, + "loss": 0.3178, + "step": 31769 + }, + { + "epoch": 0.6993176315210213, + "grad_norm": 2.1651663834857526, + "learning_rate": 2.1893561050007616e-06, + "loss": 0.3017, + "step": 31770 + }, + { + "epoch": 0.69933964340744, + "grad_norm": 3.717537927938726, + "learning_rate": 2.1890613049955294e-06, + "loss": 0.2065, + "step": 31771 + }, + { + "epoch": 0.6993616552938586, + "grad_norm": 1.478894189789156, + "learning_rate": 2.188766519276741e-06, + "loss": 0.3175, + "step": 31772 + }, + { + "epoch": 0.6993836671802773, + "grad_norm": 1.4184943361649869, + "learning_rate": 2.1884717478458933e-06, + "loss": 0.2787, + "step": 31773 + }, + { + "epoch": 0.699405679066696, + "grad_norm": 1.8100786567675222, + "learning_rate": 2.1881769907044877e-06, + "loss": 0.3609, + "step": 31774 + }, + { + "epoch": 0.6994276909531146, + "grad_norm": 1.5487472987035624, + "learning_rate": 2.1878822478540175e-06, + "loss": 0.3406, + "step": 31775 + }, + { + "epoch": 0.6994497028395333, + "grad_norm": 1.8654304448218932, + "learning_rate": 2.1875875192959874e-06, + "loss": 0.2793, + "step": 31776 + }, + { + "epoch": 0.699471714725952, + "grad_norm": 0.6136154434119542, + "learning_rate": 2.187292805031892e-06, + "loss": 0.4535, + "step": 31777 + }, + { + "epoch": 0.6994937266123706, + "grad_norm": 0.6498452222259403, + "learning_rate": 2.1869981050632266e-06, + "loss": 0.4486, + "step": 31778 + }, + { + "epoch": 0.6995157384987893, + "grad_norm": 3.5821400689090876, + "learning_rate": 2.1867034193914944e-06, + "loss": 0.3243, + "step": 31779 + }, + { + "epoch": 0.699537750385208, + "grad_norm": 1.3001047787861513, + "learning_rate": 2.1864087480181874e-06, + "loss": 0.2438, + "step": 31780 + }, + { + "epoch": 0.6995597622716266, + "grad_norm": 1.5258405979996608, + "learning_rate": 2.1861140909448066e-06, + "loss": 0.3156, + "step": 31781 + }, + { + "epoch": 0.6995817741580453, + "grad_norm": 1.4579129538090314, + "learning_rate": 2.1858194481728508e-06, + "loss": 0.3106, + "step": 31782 + }, + { + "epoch": 0.699603786044464, + "grad_norm": 1.3202290589210524, + "learning_rate": 2.1855248197038137e-06, + "loss": 0.1899, + "step": 31783 + }, + { + "epoch": 0.6996257979308826, + "grad_norm": 1.5636677019020644, + "learning_rate": 2.1852302055391967e-06, + "loss": 0.2539, + "step": 31784 + }, + { + "epoch": 0.6996478098173013, + "grad_norm": 2.5309198785586755, + "learning_rate": 2.184935605680493e-06, + "loss": 0.3219, + "step": 31785 + }, + { + "epoch": 0.69966982170372, + "grad_norm": 0.6098977130982112, + "learning_rate": 2.184641020129203e-06, + "loss": 0.4492, + "step": 31786 + }, + { + "epoch": 0.6996918335901386, + "grad_norm": 2.0467108397248914, + "learning_rate": 2.184346448886824e-06, + "loss": 0.3174, + "step": 31787 + }, + { + "epoch": 0.6997138454765574, + "grad_norm": 1.5207205943053412, + "learning_rate": 2.1840518919548506e-06, + "loss": 0.2928, + "step": 31788 + }, + { + "epoch": 0.699735857362976, + "grad_norm": 1.4401489566272536, + "learning_rate": 2.1837573493347838e-06, + "loss": 0.2491, + "step": 31789 + }, + { + "epoch": 0.6997578692493946, + "grad_norm": 1.8594082944501555, + "learning_rate": 2.1834628210281156e-06, + "loss": 0.3003, + "step": 31790 + }, + { + "epoch": 0.6997798811358134, + "grad_norm": 1.458517856971312, + "learning_rate": 2.183168307036346e-06, + "loss": 0.3308, + "step": 31791 + }, + { + "epoch": 0.6998018930222321, + "grad_norm": 1.7225936378671443, + "learning_rate": 2.1828738073609727e-06, + "loss": 0.2758, + "step": 31792 + }, + { + "epoch": 0.6998239049086507, + "grad_norm": 0.6096212938932377, + "learning_rate": 2.1825793220034896e-06, + "loss": 0.4425, + "step": 31793 + }, + { + "epoch": 0.6998459167950694, + "grad_norm": 1.9246693037174138, + "learning_rate": 2.182284850965397e-06, + "loss": 0.2434, + "step": 31794 + }, + { + "epoch": 0.699867928681488, + "grad_norm": 2.045508434522769, + "learning_rate": 2.181990394248188e-06, + "loss": 0.3089, + "step": 31795 + }, + { + "epoch": 0.6998899405679067, + "grad_norm": 1.5327609279592764, + "learning_rate": 2.1816959518533603e-06, + "loss": 0.2775, + "step": 31796 + }, + { + "epoch": 0.6999119524543254, + "grad_norm": 1.4061279331036223, + "learning_rate": 2.1814015237824125e-06, + "loss": 0.2783, + "step": 31797 + }, + { + "epoch": 0.699933964340744, + "grad_norm": 2.90740312487177, + "learning_rate": 2.1811071100368374e-06, + "loss": 0.256, + "step": 31798 + }, + { + "epoch": 0.6999559762271627, + "grad_norm": 2.1401403682338933, + "learning_rate": 2.1808127106181355e-06, + "loss": 0.3209, + "step": 31799 + }, + { + "epoch": 0.6999779881135814, + "grad_norm": 1.2891160505760386, + "learning_rate": 2.180518325527799e-06, + "loss": 0.3174, + "step": 31800 + }, + { + "epoch": 0.7, + "grad_norm": 1.5175161794574177, + "learning_rate": 2.180223954767325e-06, + "loss": 0.2735, + "step": 31801 + }, + { + "epoch": 0.7000220118864187, + "grad_norm": 1.1098935616234376, + "learning_rate": 2.1799295983382136e-06, + "loss": 0.1878, + "step": 31802 + }, + { + "epoch": 0.7000440237728374, + "grad_norm": 0.6021235657108835, + "learning_rate": 2.179635256241955e-06, + "loss": 0.4341, + "step": 31803 + }, + { + "epoch": 0.700066035659256, + "grad_norm": 1.9542895584819728, + "learning_rate": 2.17934092848005e-06, + "loss": 0.2542, + "step": 31804 + }, + { + "epoch": 0.7000880475456747, + "grad_norm": 1.3279915403352303, + "learning_rate": 2.1790466150539908e-06, + "loss": 0.2668, + "step": 31805 + }, + { + "epoch": 0.7001100594320934, + "grad_norm": 1.7052128113479104, + "learning_rate": 2.1787523159652746e-06, + "loss": 0.3475, + "step": 31806 + }, + { + "epoch": 0.700132071318512, + "grad_norm": 1.3967723521130981, + "learning_rate": 2.1784580312153995e-06, + "loss": 0.2865, + "step": 31807 + }, + { + "epoch": 0.7001540832049307, + "grad_norm": 6.339979485583954, + "learning_rate": 2.178163760805859e-06, + "loss": 0.3121, + "step": 31808 + }, + { + "epoch": 0.7001760950913494, + "grad_norm": 0.5954841588362961, + "learning_rate": 2.177869504738147e-06, + "loss": 0.4559, + "step": 31809 + }, + { + "epoch": 0.700198106977768, + "grad_norm": 3.176879500932402, + "learning_rate": 2.1775752630137633e-06, + "loss": 0.3775, + "step": 31810 + }, + { + "epoch": 0.7002201188641867, + "grad_norm": 1.728473794561682, + "learning_rate": 2.1772810356341967e-06, + "loss": 0.287, + "step": 31811 + }, + { + "epoch": 0.7002421307506054, + "grad_norm": 2.6424369651889466, + "learning_rate": 2.1769868226009503e-06, + "loss": 0.3583, + "step": 31812 + }, + { + "epoch": 0.700264142637024, + "grad_norm": 0.6172144371919327, + "learning_rate": 2.1766926239155167e-06, + "loss": 0.4502, + "step": 31813 + }, + { + "epoch": 0.7002861545234427, + "grad_norm": 1.9708728885926026, + "learning_rate": 2.176398439579388e-06, + "loss": 0.3069, + "step": 31814 + }, + { + "epoch": 0.7003081664098614, + "grad_norm": 2.871708873928184, + "learning_rate": 2.1761042695940635e-06, + "loss": 0.3113, + "step": 31815 + }, + { + "epoch": 0.70033017829628, + "grad_norm": 1.5405853868871664, + "learning_rate": 2.1758101139610357e-06, + "loss": 0.2566, + "step": 31816 + }, + { + "epoch": 0.7003521901826987, + "grad_norm": 2.8961618702133465, + "learning_rate": 2.1755159726817994e-06, + "loss": 0.4591, + "step": 31817 + }, + { + "epoch": 0.7003742020691173, + "grad_norm": 0.6188174522046457, + "learning_rate": 2.175221845757853e-06, + "loss": 0.4776, + "step": 31818 + }, + { + "epoch": 0.700396213955536, + "grad_norm": 1.366428411163024, + "learning_rate": 2.174927733190687e-06, + "loss": 0.33, + "step": 31819 + }, + { + "epoch": 0.7004182258419547, + "grad_norm": 1.3606110025024434, + "learning_rate": 2.1746336349818003e-06, + "loss": 0.2183, + "step": 31820 + }, + { + "epoch": 0.7004402377283733, + "grad_norm": 1.8726231567407188, + "learning_rate": 2.1743395511326844e-06, + "loss": 0.2606, + "step": 31821 + }, + { + "epoch": 0.700462249614792, + "grad_norm": 4.100359731903037, + "learning_rate": 2.1740454816448343e-06, + "loss": 0.2805, + "step": 31822 + }, + { + "epoch": 0.7004842615012107, + "grad_norm": 3.3669904589305286, + "learning_rate": 2.1737514265197475e-06, + "loss": 0.3174, + "step": 31823 + }, + { + "epoch": 0.7005062733876293, + "grad_norm": 1.5687420063225306, + "learning_rate": 2.1734573857589148e-06, + "loss": 0.3189, + "step": 31824 + }, + { + "epoch": 0.700528285274048, + "grad_norm": 1.4627540430006798, + "learning_rate": 2.173163359363834e-06, + "loss": 0.2946, + "step": 31825 + }, + { + "epoch": 0.7005502971604667, + "grad_norm": 2.596682028051918, + "learning_rate": 2.1728693473359962e-06, + "loss": 0.2687, + "step": 31826 + }, + { + "epoch": 0.7005723090468853, + "grad_norm": 1.6987605288072898, + "learning_rate": 2.172575349676897e-06, + "loss": 0.2483, + "step": 31827 + }, + { + "epoch": 0.700594320933304, + "grad_norm": 1.828137987127931, + "learning_rate": 2.1722813663880333e-06, + "loss": 0.3006, + "step": 31828 + }, + { + "epoch": 0.7006163328197227, + "grad_norm": 1.710042005675901, + "learning_rate": 2.171987397470895e-06, + "loss": 0.3652, + "step": 31829 + }, + { + "epoch": 0.7006383447061413, + "grad_norm": 0.6399888138018701, + "learning_rate": 2.1716934429269793e-06, + "loss": 0.4647, + "step": 31830 + }, + { + "epoch": 0.70066035659256, + "grad_norm": 1.6541598599673906, + "learning_rate": 2.171399502757778e-06, + "loss": 0.2572, + "step": 31831 + }, + { + "epoch": 0.7006823684789787, + "grad_norm": 0.6098517264427336, + "learning_rate": 2.1711055769647847e-06, + "loss": 0.4306, + "step": 31832 + }, + { + "epoch": 0.7007043803653973, + "grad_norm": 1.5456998800217074, + "learning_rate": 2.1708116655494977e-06, + "loss": 0.2709, + "step": 31833 + }, + { + "epoch": 0.700726392251816, + "grad_norm": 1.5035256130462389, + "learning_rate": 2.1705177685134056e-06, + "loss": 0.2588, + "step": 31834 + }, + { + "epoch": 0.7007484041382347, + "grad_norm": 1.4311622887112656, + "learning_rate": 2.1702238858580055e-06, + "loss": 0.2919, + "step": 31835 + }, + { + "epoch": 0.7007704160246533, + "grad_norm": 1.3978949923159225, + "learning_rate": 2.1699300175847883e-06, + "loss": 0.1937, + "step": 31836 + }, + { + "epoch": 0.700792427911072, + "grad_norm": 0.5984397703327152, + "learning_rate": 2.169636163695249e-06, + "loss": 0.4625, + "step": 31837 + }, + { + "epoch": 0.7008144397974907, + "grad_norm": 1.4457934099348908, + "learning_rate": 2.169342324190882e-06, + "loss": 0.2802, + "step": 31838 + }, + { + "epoch": 0.7008364516839093, + "grad_norm": 1.84421488003407, + "learning_rate": 2.169048499073179e-06, + "loss": 0.306, + "step": 31839 + }, + { + "epoch": 0.700858463570328, + "grad_norm": 2.091982673203608, + "learning_rate": 2.168754688343636e-06, + "loss": 0.3219, + "step": 31840 + }, + { + "epoch": 0.7008804754567466, + "grad_norm": 0.6122661734940363, + "learning_rate": 2.1684608920037444e-06, + "loss": 0.4606, + "step": 31841 + }, + { + "epoch": 0.7009024873431653, + "grad_norm": 1.4175815045604148, + "learning_rate": 2.1681671100549934e-06, + "loss": 0.25, + "step": 31842 + }, + { + "epoch": 0.700924499229584, + "grad_norm": 1.5744366438001751, + "learning_rate": 2.167873342498884e-06, + "loss": 0.2672, + "step": 31843 + }, + { + "epoch": 0.7009465111160026, + "grad_norm": 1.4086287839268532, + "learning_rate": 2.1675795893369037e-06, + "loss": 0.3158, + "step": 31844 + }, + { + "epoch": 0.7009685230024213, + "grad_norm": 1.9054035934541929, + "learning_rate": 2.1672858505705492e-06, + "loss": 0.4088, + "step": 31845 + }, + { + "epoch": 0.70099053488884, + "grad_norm": 1.6420294435567955, + "learning_rate": 2.166992126201311e-06, + "loss": 0.2916, + "step": 31846 + }, + { + "epoch": 0.7010125467752586, + "grad_norm": 2.651317041460069, + "learning_rate": 2.1666984162306794e-06, + "loss": 0.3581, + "step": 31847 + }, + { + "epoch": 0.7010345586616773, + "grad_norm": 1.6085920207853994, + "learning_rate": 2.1664047206601536e-06, + "loss": 0.2838, + "step": 31848 + }, + { + "epoch": 0.701056570548096, + "grad_norm": 3.2286614995729956, + "learning_rate": 2.166111039491222e-06, + "loss": 0.3062, + "step": 31849 + }, + { + "epoch": 0.7010785824345146, + "grad_norm": 1.47134501980204, + "learning_rate": 2.1658173727253774e-06, + "loss": 0.3535, + "step": 31850 + }, + { + "epoch": 0.7011005943209333, + "grad_norm": 1.3956683343803122, + "learning_rate": 2.165523720364114e-06, + "loss": 0.2799, + "step": 31851 + }, + { + "epoch": 0.701122606207352, + "grad_norm": 0.6120524690989224, + "learning_rate": 2.1652300824089218e-06, + "loss": 0.4482, + "step": 31852 + }, + { + "epoch": 0.7011446180937706, + "grad_norm": 1.65949517184351, + "learning_rate": 2.164936458861294e-06, + "loss": 0.2587, + "step": 31853 + }, + { + "epoch": 0.7011666299801893, + "grad_norm": 1.5892459447383671, + "learning_rate": 2.164642849722726e-06, + "loss": 0.2912, + "step": 31854 + }, + { + "epoch": 0.701188641866608, + "grad_norm": 2.060600082877556, + "learning_rate": 2.164349254994706e-06, + "loss": 0.297, + "step": 31855 + }, + { + "epoch": 0.7012106537530266, + "grad_norm": 1.369016374088507, + "learning_rate": 2.1640556746787293e-06, + "loss": 0.231, + "step": 31856 + }, + { + "epoch": 0.7012326656394453, + "grad_norm": 0.6014418376796348, + "learning_rate": 2.163762108776285e-06, + "loss": 0.4492, + "step": 31857 + }, + { + "epoch": 0.701254677525864, + "grad_norm": 3.8851934338484333, + "learning_rate": 2.1634685572888665e-06, + "loss": 0.2862, + "step": 31858 + }, + { + "epoch": 0.7012766894122826, + "grad_norm": 1.4989718661091032, + "learning_rate": 2.1631750202179673e-06, + "loss": 0.2587, + "step": 31859 + }, + { + "epoch": 0.7012987012987013, + "grad_norm": 3.0824893162964613, + "learning_rate": 2.162881497565077e-06, + "loss": 0.4228, + "step": 31860 + }, + { + "epoch": 0.70132071318512, + "grad_norm": 0.6402775726758613, + "learning_rate": 2.1625879893316894e-06, + "loss": 0.4556, + "step": 31861 + }, + { + "epoch": 0.7013427250715386, + "grad_norm": 1.3981614107713496, + "learning_rate": 2.1622944955192935e-06, + "loss": 0.2869, + "step": 31862 + }, + { + "epoch": 0.7013647369579573, + "grad_norm": 5.284262020525956, + "learning_rate": 2.162001016129383e-06, + "loss": 0.2859, + "step": 31863 + }, + { + "epoch": 0.701386748844376, + "grad_norm": 1.3008944271341203, + "learning_rate": 2.1617075511634506e-06, + "loss": 0.3317, + "step": 31864 + }, + { + "epoch": 0.7014087607307946, + "grad_norm": 2.260663531686825, + "learning_rate": 2.1614141006229843e-06, + "loss": 0.3536, + "step": 31865 + }, + { + "epoch": 0.7014307726172133, + "grad_norm": 1.6449793120720821, + "learning_rate": 2.16112066450948e-06, + "loss": 0.2852, + "step": 31866 + }, + { + "epoch": 0.7014527845036319, + "grad_norm": 1.511463939639965, + "learning_rate": 2.1608272428244244e-06, + "loss": 0.286, + "step": 31867 + }, + { + "epoch": 0.7014747963900506, + "grad_norm": 2.117857988049621, + "learning_rate": 2.160533835569311e-06, + "loss": 0.2651, + "step": 31868 + }, + { + "epoch": 0.7014968082764693, + "grad_norm": 1.3791751578818434, + "learning_rate": 2.160240442745633e-06, + "loss": 0.2313, + "step": 31869 + }, + { + "epoch": 0.7015188201628879, + "grad_norm": 1.5679002560129691, + "learning_rate": 2.159947064354878e-06, + "loss": 0.2871, + "step": 31870 + }, + { + "epoch": 0.7015408320493066, + "grad_norm": 0.6174839345326094, + "learning_rate": 2.159653700398541e-06, + "loss": 0.4544, + "step": 31871 + }, + { + "epoch": 0.7015628439357253, + "grad_norm": 1.8231964136787415, + "learning_rate": 2.159360350878108e-06, + "loss": 0.3263, + "step": 31872 + }, + { + "epoch": 0.7015848558221439, + "grad_norm": 0.6343587807455583, + "learning_rate": 2.1590670157950727e-06, + "loss": 0.474, + "step": 31873 + }, + { + "epoch": 0.7016068677085626, + "grad_norm": 1.379772944034642, + "learning_rate": 2.158773695150928e-06, + "loss": 0.2474, + "step": 31874 + }, + { + "epoch": 0.7016288795949813, + "grad_norm": 1.2815293820554838, + "learning_rate": 2.1584803889471613e-06, + "loss": 0.2697, + "step": 31875 + }, + { + "epoch": 0.7016508914813999, + "grad_norm": 0.635749288536804, + "learning_rate": 2.158187097185266e-06, + "loss": 0.497, + "step": 31876 + }, + { + "epoch": 0.7016729033678186, + "grad_norm": 1.7197887127155624, + "learning_rate": 2.157893819866731e-06, + "loss": 0.3049, + "step": 31877 + }, + { + "epoch": 0.7016949152542373, + "grad_norm": 2.085096984970012, + "learning_rate": 2.1576005569930434e-06, + "loss": 0.2601, + "step": 31878 + }, + { + "epoch": 0.7017169271406559, + "grad_norm": 1.5383815997891546, + "learning_rate": 2.157307308565702e-06, + "loss": 0.2479, + "step": 31879 + }, + { + "epoch": 0.7017389390270746, + "grad_norm": 1.3834849625831338, + "learning_rate": 2.1570140745861907e-06, + "loss": 0.2478, + "step": 31880 + }, + { + "epoch": 0.7017609509134933, + "grad_norm": 1.9320935235478434, + "learning_rate": 2.1567208550560037e-06, + "loss": 0.2694, + "step": 31881 + }, + { + "epoch": 0.7017829627999119, + "grad_norm": 1.5656990656946652, + "learning_rate": 2.156427649976629e-06, + "loss": 0.311, + "step": 31882 + }, + { + "epoch": 0.7018049746863306, + "grad_norm": 9.731241224800417, + "learning_rate": 2.156134459349554e-06, + "loss": 0.3756, + "step": 31883 + }, + { + "epoch": 0.7018269865727493, + "grad_norm": 1.7528030258287786, + "learning_rate": 2.1558412831762753e-06, + "loss": 0.3371, + "step": 31884 + }, + { + "epoch": 0.7018489984591679, + "grad_norm": 2.575433651216478, + "learning_rate": 2.1555481214582803e-06, + "loss": 0.2005, + "step": 31885 + }, + { + "epoch": 0.7018710103455866, + "grad_norm": 2.364652501688515, + "learning_rate": 2.155254974197056e-06, + "loss": 0.3352, + "step": 31886 + }, + { + "epoch": 0.7018930222320053, + "grad_norm": 1.8702882713731785, + "learning_rate": 2.1549618413940974e-06, + "loss": 0.3932, + "step": 31887 + }, + { + "epoch": 0.7019150341184239, + "grad_norm": 1.3666993211989085, + "learning_rate": 2.1546687230508894e-06, + "loss": 0.2888, + "step": 31888 + }, + { + "epoch": 0.7019370460048426, + "grad_norm": 1.5001804396372038, + "learning_rate": 2.1543756191689236e-06, + "loss": 0.3618, + "step": 31889 + }, + { + "epoch": 0.7019590578912612, + "grad_norm": 2.8177590622714295, + "learning_rate": 2.1540825297496927e-06, + "loss": 0.3461, + "step": 31890 + }, + { + "epoch": 0.7019810697776799, + "grad_norm": 0.6187733922633932, + "learning_rate": 2.1537894547946813e-06, + "loss": 0.443, + "step": 31891 + }, + { + "epoch": 0.7020030816640986, + "grad_norm": 2.124097400013947, + "learning_rate": 2.1534963943053834e-06, + "loss": 0.3116, + "step": 31892 + }, + { + "epoch": 0.7020250935505172, + "grad_norm": 0.6289992426948756, + "learning_rate": 2.153203348283284e-06, + "loss": 0.4665, + "step": 31893 + }, + { + "epoch": 0.7020471054369359, + "grad_norm": 0.6273518261486183, + "learning_rate": 2.1529103167298755e-06, + "loss": 0.4688, + "step": 31894 + }, + { + "epoch": 0.7020691173233546, + "grad_norm": 0.6234869758300056, + "learning_rate": 2.1526172996466484e-06, + "loss": 0.4798, + "step": 31895 + }, + { + "epoch": 0.7020911292097732, + "grad_norm": 0.5975722737384431, + "learning_rate": 2.1523242970350876e-06, + "loss": 0.4462, + "step": 31896 + }, + { + "epoch": 0.7021131410961919, + "grad_norm": 1.5655876436103446, + "learning_rate": 2.1520313088966877e-06, + "loss": 0.2696, + "step": 31897 + }, + { + "epoch": 0.7021351529826106, + "grad_norm": 2.1486661599703893, + "learning_rate": 2.1517383352329323e-06, + "loss": 0.2375, + "step": 31898 + }, + { + "epoch": 0.7021571648690292, + "grad_norm": 1.3171862125288432, + "learning_rate": 2.1514453760453134e-06, + "loss": 0.2258, + "step": 31899 + }, + { + "epoch": 0.7021791767554479, + "grad_norm": 3.801764065877376, + "learning_rate": 2.1511524313353215e-06, + "loss": 0.3028, + "step": 31900 + }, + { + "epoch": 0.7022011886418666, + "grad_norm": 3.195804107634717, + "learning_rate": 2.150859501104442e-06, + "loss": 0.2803, + "step": 31901 + }, + { + "epoch": 0.7022232005282852, + "grad_norm": 1.6311299943382254, + "learning_rate": 2.150566585354166e-06, + "loss": 0.2533, + "step": 31902 + }, + { + "epoch": 0.702245212414704, + "grad_norm": 1.6768558684171724, + "learning_rate": 2.1502736840859804e-06, + "loss": 0.3122, + "step": 31903 + }, + { + "epoch": 0.7022672243011227, + "grad_norm": 1.525426357230664, + "learning_rate": 2.1499807973013755e-06, + "loss": 0.3293, + "step": 31904 + }, + { + "epoch": 0.7022892361875412, + "grad_norm": 0.652446304532939, + "learning_rate": 2.1496879250018403e-06, + "loss": 0.4777, + "step": 31905 + }, + { + "epoch": 0.70231124807396, + "grad_norm": 1.679210888060395, + "learning_rate": 2.149395067188861e-06, + "loss": 0.2437, + "step": 31906 + }, + { + "epoch": 0.7023332599603787, + "grad_norm": 1.3305153107787124, + "learning_rate": 2.149102223863929e-06, + "loss": 0.2984, + "step": 31907 + }, + { + "epoch": 0.7023552718467972, + "grad_norm": 1.7447850819555832, + "learning_rate": 2.1488093950285293e-06, + "loss": 0.3374, + "step": 31908 + }, + { + "epoch": 0.702377283733216, + "grad_norm": 4.266350207470297, + "learning_rate": 2.1485165806841517e-06, + "loss": 0.3398, + "step": 31909 + }, + { + "epoch": 0.7023992956196347, + "grad_norm": 1.5265284138290274, + "learning_rate": 2.1482237808322864e-06, + "loss": 0.3136, + "step": 31910 + }, + { + "epoch": 0.7024213075060532, + "grad_norm": 2.453514922383005, + "learning_rate": 2.147930995474418e-06, + "loss": 0.3458, + "step": 31911 + }, + { + "epoch": 0.702443319392472, + "grad_norm": 3.205028956455737, + "learning_rate": 2.1476382246120386e-06, + "loss": 0.282, + "step": 31912 + }, + { + "epoch": 0.7024653312788905, + "grad_norm": 1.9761853594834367, + "learning_rate": 2.147345468246634e-06, + "loss": 0.2444, + "step": 31913 + }, + { + "epoch": 0.7024873431653093, + "grad_norm": 1.3354020247752716, + "learning_rate": 2.1470527263796876e-06, + "loss": 0.2106, + "step": 31914 + }, + { + "epoch": 0.702509355051728, + "grad_norm": 1.3056709988864512, + "learning_rate": 2.146759999012696e-06, + "loss": 0.3259, + "step": 31915 + }, + { + "epoch": 0.7025313669381466, + "grad_norm": 1.4277376182300276, + "learning_rate": 2.1464672861471405e-06, + "loss": 0.3175, + "step": 31916 + }, + { + "epoch": 0.7025533788245653, + "grad_norm": 1.771147190503774, + "learning_rate": 2.146174587784513e-06, + "loss": 0.3177, + "step": 31917 + }, + { + "epoch": 0.702575390710984, + "grad_norm": 2.638424273397117, + "learning_rate": 2.145881903926299e-06, + "loss": 0.3437, + "step": 31918 + }, + { + "epoch": 0.7025974025974026, + "grad_norm": 0.6099311329768576, + "learning_rate": 2.1455892345739825e-06, + "loss": 0.4552, + "step": 31919 + }, + { + "epoch": 0.7026194144838213, + "grad_norm": 2.7264964056610297, + "learning_rate": 2.145296579729058e-06, + "loss": 0.2224, + "step": 31920 + }, + { + "epoch": 0.70264142637024, + "grad_norm": 2.761991651505088, + "learning_rate": 2.1450039393930083e-06, + "loss": 0.279, + "step": 31921 + }, + { + "epoch": 0.7026634382566586, + "grad_norm": 1.6218979328882168, + "learning_rate": 2.144711313567323e-06, + "loss": 0.3909, + "step": 31922 + }, + { + "epoch": 0.7026854501430773, + "grad_norm": 1.6563465655060927, + "learning_rate": 2.144418702253489e-06, + "loss": 0.2835, + "step": 31923 + }, + { + "epoch": 0.702707462029496, + "grad_norm": 0.6230704011561337, + "learning_rate": 2.1441261054529903e-06, + "loss": 0.4716, + "step": 31924 + }, + { + "epoch": 0.7027294739159146, + "grad_norm": 1.6300916460688761, + "learning_rate": 2.143833523167317e-06, + "loss": 0.3587, + "step": 31925 + }, + { + "epoch": 0.7027514858023333, + "grad_norm": 0.6080746417465923, + "learning_rate": 2.1435409553979576e-06, + "loss": 0.4731, + "step": 31926 + }, + { + "epoch": 0.702773497688752, + "grad_norm": 1.6118660896490355, + "learning_rate": 2.1432484021463945e-06, + "loss": 0.2632, + "step": 31927 + }, + { + "epoch": 0.7027955095751706, + "grad_norm": 1.6336307990583514, + "learning_rate": 2.1429558634141196e-06, + "loss": 0.2685, + "step": 31928 + }, + { + "epoch": 0.7028175214615893, + "grad_norm": 1.6260848758674598, + "learning_rate": 2.1426633392026153e-06, + "loss": 0.2681, + "step": 31929 + }, + { + "epoch": 0.702839533348008, + "grad_norm": 1.4661239649972855, + "learning_rate": 2.1423708295133706e-06, + "loss": 0.3116, + "step": 31930 + }, + { + "epoch": 0.7028615452344266, + "grad_norm": 0.5811294589764703, + "learning_rate": 2.1420783343478736e-06, + "loss": 0.4236, + "step": 31931 + }, + { + "epoch": 0.7028835571208453, + "grad_norm": 1.6269224680344616, + "learning_rate": 2.141785853707607e-06, + "loss": 0.2763, + "step": 31932 + }, + { + "epoch": 0.702905569007264, + "grad_norm": 2.0043551609525587, + "learning_rate": 2.1414933875940615e-06, + "loss": 0.3125, + "step": 31933 + }, + { + "epoch": 0.7029275808936826, + "grad_norm": 1.554890285042125, + "learning_rate": 2.1412009360087193e-06, + "loss": 0.1928, + "step": 31934 + }, + { + "epoch": 0.7029495927801013, + "grad_norm": 1.8665654806953604, + "learning_rate": 2.14090849895307e-06, + "loss": 0.3518, + "step": 31935 + }, + { + "epoch": 0.7029716046665199, + "grad_norm": 1.4692662290874265, + "learning_rate": 2.1406160764286e-06, + "loss": 0.33, + "step": 31936 + }, + { + "epoch": 0.7029936165529386, + "grad_norm": 1.5748257184764238, + "learning_rate": 2.1403236684367924e-06, + "loss": 0.1947, + "step": 31937 + }, + { + "epoch": 0.7030156284393573, + "grad_norm": 1.2996375792163162, + "learning_rate": 2.1400312749791375e-06, + "loss": 0.2165, + "step": 31938 + }, + { + "epoch": 0.7030376403257759, + "grad_norm": 1.749338431996994, + "learning_rate": 2.1397388960571176e-06, + "loss": 0.2573, + "step": 31939 + }, + { + "epoch": 0.7030596522121946, + "grad_norm": 1.7814527565063139, + "learning_rate": 2.1394465316722203e-06, + "loss": 0.2523, + "step": 31940 + }, + { + "epoch": 0.7030816640986133, + "grad_norm": 0.6071148886074939, + "learning_rate": 2.1391541818259333e-06, + "loss": 0.4494, + "step": 31941 + }, + { + "epoch": 0.7031036759850319, + "grad_norm": 1.6838368033292368, + "learning_rate": 2.1388618465197384e-06, + "loss": 0.3466, + "step": 31942 + }, + { + "epoch": 0.7031256878714506, + "grad_norm": 2.063370329293543, + "learning_rate": 2.1385695257551257e-06, + "loss": 0.3758, + "step": 31943 + }, + { + "epoch": 0.7031476997578693, + "grad_norm": 2.4410858532436155, + "learning_rate": 2.138277219533578e-06, + "loss": 0.3292, + "step": 31944 + }, + { + "epoch": 0.7031697116442879, + "grad_norm": 2.5019191199771256, + "learning_rate": 2.1379849278565813e-06, + "loss": 0.3596, + "step": 31945 + }, + { + "epoch": 0.7031917235307066, + "grad_norm": 0.607598964857968, + "learning_rate": 2.1376926507256227e-06, + "loss": 0.4545, + "step": 31946 + }, + { + "epoch": 0.7032137354171253, + "grad_norm": 2.0057710773545763, + "learning_rate": 2.137400388142186e-06, + "loss": 0.2863, + "step": 31947 + }, + { + "epoch": 0.7032357473035439, + "grad_norm": 1.6486860441449926, + "learning_rate": 2.137108140107759e-06, + "loss": 0.3775, + "step": 31948 + }, + { + "epoch": 0.7032577591899626, + "grad_norm": 1.9226718479286289, + "learning_rate": 2.136815906623825e-06, + "loss": 0.2734, + "step": 31949 + }, + { + "epoch": 0.7032797710763813, + "grad_norm": 1.546503026098251, + "learning_rate": 2.1365236876918657e-06, + "loss": 0.2471, + "step": 31950 + }, + { + "epoch": 0.7033017829627999, + "grad_norm": 3.5022664288548917, + "learning_rate": 2.136231483313374e-06, + "loss": 0.3214, + "step": 31951 + }, + { + "epoch": 0.7033237948492186, + "grad_norm": 4.283643321047306, + "learning_rate": 2.1359392934898293e-06, + "loss": 0.2609, + "step": 31952 + }, + { + "epoch": 0.7033458067356373, + "grad_norm": 1.425130154507839, + "learning_rate": 2.1356471182227205e-06, + "loss": 0.2596, + "step": 31953 + }, + { + "epoch": 0.7033678186220559, + "grad_norm": 1.9357847768066394, + "learning_rate": 2.13535495751353e-06, + "loss": 0.3366, + "step": 31954 + }, + { + "epoch": 0.7033898305084746, + "grad_norm": 0.6161636621218941, + "learning_rate": 2.135062811363742e-06, + "loss": 0.4709, + "step": 31955 + }, + { + "epoch": 0.7034118423948933, + "grad_norm": 1.5809811858413605, + "learning_rate": 2.134770679774843e-06, + "loss": 0.2568, + "step": 31956 + }, + { + "epoch": 0.7034338542813119, + "grad_norm": 0.5889998081602844, + "learning_rate": 2.1344785627483168e-06, + "loss": 0.4223, + "step": 31957 + }, + { + "epoch": 0.7034558661677306, + "grad_norm": 1.727937425576052, + "learning_rate": 2.1341864602856503e-06, + "loss": 0.2369, + "step": 31958 + }, + { + "epoch": 0.7034778780541492, + "grad_norm": 0.6307658075115704, + "learning_rate": 2.1338943723883264e-06, + "loss": 0.4417, + "step": 31959 + }, + { + "epoch": 0.7034998899405679, + "grad_norm": 1.7228709708901317, + "learning_rate": 2.133602299057828e-06, + "loss": 0.3068, + "step": 31960 + }, + { + "epoch": 0.7035219018269866, + "grad_norm": 1.6134362072092339, + "learning_rate": 2.133310240295642e-06, + "loss": 0.254, + "step": 31961 + }, + { + "epoch": 0.7035439137134052, + "grad_norm": 1.4194157945051586, + "learning_rate": 2.133018196103251e-06, + "loss": 0.328, + "step": 31962 + }, + { + "epoch": 0.7035659255998239, + "grad_norm": 3.027694736783173, + "learning_rate": 2.132726166482142e-06, + "loss": 0.2821, + "step": 31963 + }, + { + "epoch": 0.7035879374862426, + "grad_norm": 0.5996402732400834, + "learning_rate": 2.1324341514337978e-06, + "loss": 0.4354, + "step": 31964 + }, + { + "epoch": 0.7036099493726612, + "grad_norm": 2.7371941836674356, + "learning_rate": 2.1321421509597003e-06, + "loss": 0.3264, + "step": 31965 + }, + { + "epoch": 0.7036319612590799, + "grad_norm": 0.611130501077144, + "learning_rate": 2.1318501650613348e-06, + "loss": 0.4395, + "step": 31966 + }, + { + "epoch": 0.7036539731454986, + "grad_norm": 1.525040401144647, + "learning_rate": 2.1315581937401885e-06, + "loss": 0.2569, + "step": 31967 + }, + { + "epoch": 0.7036759850319172, + "grad_norm": 0.6267067593613788, + "learning_rate": 2.131266236997741e-06, + "loss": 0.4571, + "step": 31968 + }, + { + "epoch": 0.7036979969183359, + "grad_norm": 1.8825846322570061, + "learning_rate": 2.1309742948354796e-06, + "loss": 0.1999, + "step": 31969 + }, + { + "epoch": 0.7037200088047546, + "grad_norm": 4.345516486654383, + "learning_rate": 2.130682367254885e-06, + "loss": 0.2491, + "step": 31970 + }, + { + "epoch": 0.7037420206911732, + "grad_norm": 1.8294671487872847, + "learning_rate": 2.130390454257442e-06, + "loss": 0.3581, + "step": 31971 + }, + { + "epoch": 0.7037640325775919, + "grad_norm": 1.5336236703835295, + "learning_rate": 2.1300985558446364e-06, + "loss": 0.2047, + "step": 31972 + }, + { + "epoch": 0.7037860444640106, + "grad_norm": 0.6353234951916521, + "learning_rate": 2.129806672017948e-06, + "loss": 0.4406, + "step": 31973 + }, + { + "epoch": 0.7038080563504292, + "grad_norm": 1.9474590505069467, + "learning_rate": 2.129514802778865e-06, + "loss": 0.3524, + "step": 31974 + }, + { + "epoch": 0.7038300682368479, + "grad_norm": 1.3865473089986156, + "learning_rate": 2.129222948128865e-06, + "loss": 0.2043, + "step": 31975 + }, + { + "epoch": 0.7038520801232666, + "grad_norm": 1.5172130454715655, + "learning_rate": 2.1289311080694354e-06, + "loss": 0.3027, + "step": 31976 + }, + { + "epoch": 0.7038740920096852, + "grad_norm": 2.915463368483446, + "learning_rate": 2.12863928260206e-06, + "loss": 0.2678, + "step": 31977 + }, + { + "epoch": 0.7038961038961039, + "grad_norm": 1.5105050700160383, + "learning_rate": 2.1283474717282175e-06, + "loss": 0.2668, + "step": 31978 + }, + { + "epoch": 0.7039181157825226, + "grad_norm": 2.0795431180541244, + "learning_rate": 2.128055675449397e-06, + "loss": 0.3429, + "step": 31979 + }, + { + "epoch": 0.7039401276689412, + "grad_norm": 1.4839214513911576, + "learning_rate": 2.1277638937670764e-06, + "loss": 0.3644, + "step": 31980 + }, + { + "epoch": 0.7039621395553599, + "grad_norm": 0.6658442969206505, + "learning_rate": 2.12747212668274e-06, + "loss": 0.4769, + "step": 31981 + }, + { + "epoch": 0.7039841514417785, + "grad_norm": 1.5414413402342957, + "learning_rate": 2.127180374197874e-06, + "loss": 0.3128, + "step": 31982 + }, + { + "epoch": 0.7040061633281972, + "grad_norm": 0.6371496412291986, + "learning_rate": 2.126888636313957e-06, + "loss": 0.4522, + "step": 31983 + }, + { + "epoch": 0.7040281752146159, + "grad_norm": 2.912072686163413, + "learning_rate": 2.1265969130324746e-06, + "loss": 0.2086, + "step": 31984 + }, + { + "epoch": 0.7040501871010345, + "grad_norm": 1.7037533787681134, + "learning_rate": 2.1263052043549095e-06, + "loss": 0.2577, + "step": 31985 + }, + { + "epoch": 0.7040721989874532, + "grad_norm": 2.435188219922709, + "learning_rate": 2.1260135102827385e-06, + "loss": 0.4219, + "step": 31986 + }, + { + "epoch": 0.7040942108738719, + "grad_norm": 0.5988388008396476, + "learning_rate": 2.125721830817452e-06, + "loss": 0.4488, + "step": 31987 + }, + { + "epoch": 0.7041162227602905, + "grad_norm": 4.741551114243235, + "learning_rate": 2.1254301659605274e-06, + "loss": 0.3263, + "step": 31988 + }, + { + "epoch": 0.7041382346467092, + "grad_norm": 1.6551951905174547, + "learning_rate": 2.1251385157134506e-06, + "loss": 0.2856, + "step": 31989 + }, + { + "epoch": 0.7041602465331279, + "grad_norm": 0.6505082720703964, + "learning_rate": 2.1248468800777028e-06, + "loss": 0.4998, + "step": 31990 + }, + { + "epoch": 0.7041822584195465, + "grad_norm": 1.7446325531286215, + "learning_rate": 2.124555259054763e-06, + "loss": 0.3039, + "step": 31991 + }, + { + "epoch": 0.7042042703059652, + "grad_norm": 1.3891013289866734, + "learning_rate": 2.1242636526461163e-06, + "loss": 0.2807, + "step": 31992 + }, + { + "epoch": 0.7042262821923839, + "grad_norm": 1.3238967031741442, + "learning_rate": 2.1239720608532444e-06, + "loss": 0.283, + "step": 31993 + }, + { + "epoch": 0.7042482940788025, + "grad_norm": 2.750493430196176, + "learning_rate": 2.1236804836776303e-06, + "loss": 0.2613, + "step": 31994 + }, + { + "epoch": 0.7042703059652212, + "grad_norm": 1.6432460289596431, + "learning_rate": 2.1233889211207554e-06, + "loss": 0.247, + "step": 31995 + }, + { + "epoch": 0.7042923178516399, + "grad_norm": 1.9211784258519868, + "learning_rate": 2.1230973731840994e-06, + "loss": 0.2128, + "step": 31996 + }, + { + "epoch": 0.7043143297380585, + "grad_norm": 1.626019957931438, + "learning_rate": 2.1228058398691455e-06, + "loss": 0.3105, + "step": 31997 + }, + { + "epoch": 0.7043363416244772, + "grad_norm": 1.3005863045485155, + "learning_rate": 2.1225143211773752e-06, + "loss": 0.136, + "step": 31998 + }, + { + "epoch": 0.7043583535108959, + "grad_norm": 1.5562786259058412, + "learning_rate": 2.1222228171102725e-06, + "loss": 0.3032, + "step": 31999 + }, + { + "epoch": 0.7043803653973145, + "grad_norm": 1.3481613710658196, + "learning_rate": 2.121931327669317e-06, + "loss": 0.236, + "step": 32000 + }, + { + "epoch": 0.7044023772837332, + "grad_norm": 0.6031180136188645, + "learning_rate": 2.1216398528559885e-06, + "loss": 0.4552, + "step": 32001 + }, + { + "epoch": 0.7044243891701519, + "grad_norm": 1.3657922751087546, + "learning_rate": 2.1213483926717697e-06, + "loss": 0.29, + "step": 32002 + }, + { + "epoch": 0.7044464010565705, + "grad_norm": 1.5424889772711903, + "learning_rate": 2.121056947118143e-06, + "loss": 0.2381, + "step": 32003 + }, + { + "epoch": 0.7044684129429892, + "grad_norm": 1.5159578683156154, + "learning_rate": 2.12076551619659e-06, + "loss": 0.3409, + "step": 32004 + }, + { + "epoch": 0.7044904248294079, + "grad_norm": 1.776868306279272, + "learning_rate": 2.1204740999085915e-06, + "loss": 0.2528, + "step": 32005 + }, + { + "epoch": 0.7045124367158265, + "grad_norm": 4.0732148415368785, + "learning_rate": 2.120182698255626e-06, + "loss": 0.2223, + "step": 32006 + }, + { + "epoch": 0.7045344486022452, + "grad_norm": 0.5881027383954951, + "learning_rate": 2.1198913112391756e-06, + "loss": 0.4257, + "step": 32007 + }, + { + "epoch": 0.7045564604886638, + "grad_norm": 1.9136189227739644, + "learning_rate": 2.119599938860725e-06, + "loss": 0.34, + "step": 32008 + }, + { + "epoch": 0.7045784723750825, + "grad_norm": 1.6048513537835118, + "learning_rate": 2.1193085811217494e-06, + "loss": 0.2974, + "step": 32009 + }, + { + "epoch": 0.7046004842615012, + "grad_norm": 1.452464709010818, + "learning_rate": 2.119017238023735e-06, + "loss": 0.2028, + "step": 32010 + }, + { + "epoch": 0.7046224961479198, + "grad_norm": 7.600858029259601, + "learning_rate": 2.118725909568158e-06, + "loss": 0.3166, + "step": 32011 + }, + { + "epoch": 0.7046445080343385, + "grad_norm": 1.572215363803999, + "learning_rate": 2.1184345957565005e-06, + "loss": 0.2921, + "step": 32012 + }, + { + "epoch": 0.7046665199207572, + "grad_norm": 2.1718824136330386, + "learning_rate": 2.118143296590246e-06, + "loss": 0.397, + "step": 32013 + }, + { + "epoch": 0.7046885318071758, + "grad_norm": 2.3216008648934054, + "learning_rate": 2.1178520120708705e-06, + "loss": 0.2971, + "step": 32014 + }, + { + "epoch": 0.7047105436935945, + "grad_norm": 1.3334585974922823, + "learning_rate": 2.1175607421998582e-06, + "loss": 0.3178, + "step": 32015 + }, + { + "epoch": 0.7047325555800132, + "grad_norm": 2.0039806876568464, + "learning_rate": 2.117269486978687e-06, + "loss": 0.3097, + "step": 32016 + }, + { + "epoch": 0.7047545674664318, + "grad_norm": 1.9762937380584895, + "learning_rate": 2.116978246408837e-06, + "loss": 0.2663, + "step": 32017 + }, + { + "epoch": 0.7047765793528505, + "grad_norm": 1.9143837717886223, + "learning_rate": 2.1166870204917915e-06, + "loss": 0.3329, + "step": 32018 + }, + { + "epoch": 0.7047985912392692, + "grad_norm": 1.739422468683414, + "learning_rate": 2.1163958092290264e-06, + "loss": 0.3354, + "step": 32019 + }, + { + "epoch": 0.7048206031256878, + "grad_norm": 1.7650004351643345, + "learning_rate": 2.1161046126220264e-06, + "loss": 0.2904, + "step": 32020 + }, + { + "epoch": 0.7048426150121065, + "grad_norm": 1.590214636411738, + "learning_rate": 2.1158134306722667e-06, + "loss": 0.3034, + "step": 32021 + }, + { + "epoch": 0.7048646268985252, + "grad_norm": 0.6042027211239701, + "learning_rate": 2.11552226338123e-06, + "loss": 0.4746, + "step": 32022 + }, + { + "epoch": 0.7048866387849438, + "grad_norm": 1.4616605112204824, + "learning_rate": 2.115231110750397e-06, + "loss": 0.3629, + "step": 32023 + }, + { + "epoch": 0.7049086506713625, + "grad_norm": 2.173650744036596, + "learning_rate": 2.1149399727812448e-06, + "loss": 0.3831, + "step": 32024 + }, + { + "epoch": 0.7049306625577813, + "grad_norm": 2.276395290715465, + "learning_rate": 2.114648849475256e-06, + "loss": 0.3032, + "step": 32025 + }, + { + "epoch": 0.7049526744441998, + "grad_norm": 1.3842229452412533, + "learning_rate": 2.1143577408339082e-06, + "loss": 0.3652, + "step": 32026 + }, + { + "epoch": 0.7049746863306185, + "grad_norm": 2.4962856862502623, + "learning_rate": 2.1140666468586802e-06, + "loss": 0.293, + "step": 32027 + }, + { + "epoch": 0.7049966982170373, + "grad_norm": 1.4292783682291306, + "learning_rate": 2.1137755675510525e-06, + "loss": 0.3793, + "step": 32028 + }, + { + "epoch": 0.7050187101034558, + "grad_norm": 1.7171049378655676, + "learning_rate": 2.1134845029125044e-06, + "loss": 0.2479, + "step": 32029 + }, + { + "epoch": 0.7050407219898746, + "grad_norm": 2.318991537983807, + "learning_rate": 2.1131934529445175e-06, + "loss": 0.3256, + "step": 32030 + }, + { + "epoch": 0.7050627338762931, + "grad_norm": 0.8547007375476804, + "learning_rate": 2.1129024176485692e-06, + "loss": 0.1376, + "step": 32031 + }, + { + "epoch": 0.7050847457627119, + "grad_norm": 1.565531409534334, + "learning_rate": 2.1126113970261365e-06, + "loss": 0.2991, + "step": 32032 + }, + { + "epoch": 0.7051067576491306, + "grad_norm": 0.6121310683531662, + "learning_rate": 2.1123203910787004e-06, + "loss": 0.4477, + "step": 32033 + }, + { + "epoch": 0.7051287695355491, + "grad_norm": 1.4125459079332778, + "learning_rate": 2.1120293998077402e-06, + "loss": 0.2865, + "step": 32034 + }, + { + "epoch": 0.7051507814219679, + "grad_norm": 1.3020436993283566, + "learning_rate": 2.1117384232147366e-06, + "loss": 0.2705, + "step": 32035 + }, + { + "epoch": 0.7051727933083866, + "grad_norm": 1.7556315299102327, + "learning_rate": 2.111447461301166e-06, + "loss": 0.2522, + "step": 32036 + }, + { + "epoch": 0.7051948051948052, + "grad_norm": 1.8575340059216556, + "learning_rate": 2.1111565140685066e-06, + "loss": 0.3664, + "step": 32037 + }, + { + "epoch": 0.7052168170812239, + "grad_norm": 1.9676565497722733, + "learning_rate": 2.1108655815182376e-06, + "loss": 0.3277, + "step": 32038 + }, + { + "epoch": 0.7052388289676426, + "grad_norm": 1.4505602338498875, + "learning_rate": 2.1105746636518387e-06, + "loss": 0.2763, + "step": 32039 + }, + { + "epoch": 0.7052608408540612, + "grad_norm": 3.3894281959286667, + "learning_rate": 2.1102837604707893e-06, + "loss": 0.2728, + "step": 32040 + }, + { + "epoch": 0.7052828527404799, + "grad_norm": 1.9997017945084634, + "learning_rate": 2.1099928719765674e-06, + "loss": 0.2566, + "step": 32041 + }, + { + "epoch": 0.7053048646268986, + "grad_norm": 3.5612331730619533, + "learning_rate": 2.1097019981706485e-06, + "loss": 0.3944, + "step": 32042 + }, + { + "epoch": 0.7053268765133172, + "grad_norm": 1.4471125842964128, + "learning_rate": 2.1094111390545137e-06, + "loss": 0.2985, + "step": 32043 + }, + { + "epoch": 0.7053488883997359, + "grad_norm": 1.603801123570099, + "learning_rate": 2.109120294629642e-06, + "loss": 0.311, + "step": 32044 + }, + { + "epoch": 0.7053709002861546, + "grad_norm": 1.8319613038113285, + "learning_rate": 2.1088294648975087e-06, + "loss": 0.3069, + "step": 32045 + }, + { + "epoch": 0.7053929121725732, + "grad_norm": 3.018777202383269, + "learning_rate": 2.108538649859595e-06, + "loss": 0.4279, + "step": 32046 + }, + { + "epoch": 0.7054149240589919, + "grad_norm": 2.273763413781736, + "learning_rate": 2.1082478495173764e-06, + "loss": 0.3465, + "step": 32047 + }, + { + "epoch": 0.7054369359454106, + "grad_norm": 2.6798244472920802, + "learning_rate": 2.107957063872332e-06, + "loss": 0.2986, + "step": 32048 + }, + { + "epoch": 0.7054589478318292, + "grad_norm": 1.4951161210190818, + "learning_rate": 2.1076662929259407e-06, + "loss": 0.317, + "step": 32049 + }, + { + "epoch": 0.7054809597182479, + "grad_norm": 0.6213985259607888, + "learning_rate": 2.1073755366796776e-06, + "loss": 0.4308, + "step": 32050 + }, + { + "epoch": 0.7055029716046666, + "grad_norm": 4.643179255721983, + "learning_rate": 2.107084795135025e-06, + "loss": 0.3275, + "step": 32051 + }, + { + "epoch": 0.7055249834910852, + "grad_norm": 1.449359763716072, + "learning_rate": 2.106794068293455e-06, + "loss": 0.3012, + "step": 32052 + }, + { + "epoch": 0.7055469953775039, + "grad_norm": 1.8685179768245224, + "learning_rate": 2.106503356156448e-06, + "loss": 0.2902, + "step": 32053 + }, + { + "epoch": 0.7055690072639225, + "grad_norm": 1.3019647616117764, + "learning_rate": 2.106212658725483e-06, + "loss": 0.3283, + "step": 32054 + }, + { + "epoch": 0.7055910191503412, + "grad_norm": 0.6604480609993462, + "learning_rate": 2.105921976002035e-06, + "loss": 0.4821, + "step": 32055 + }, + { + "epoch": 0.7056130310367599, + "grad_norm": 1.2076577671133981, + "learning_rate": 2.105631307987584e-06, + "loss": 0.2551, + "step": 32056 + }, + { + "epoch": 0.7056350429231785, + "grad_norm": 2.7684861599143598, + "learning_rate": 2.1053406546836037e-06, + "loss": 0.2878, + "step": 32057 + }, + { + "epoch": 0.7056570548095972, + "grad_norm": 0.6055265713779036, + "learning_rate": 2.1050500160915727e-06, + "loss": 0.4682, + "step": 32058 + }, + { + "epoch": 0.7056790666960159, + "grad_norm": 1.7426024928496384, + "learning_rate": 2.104759392212971e-06, + "loss": 0.343, + "step": 32059 + }, + { + "epoch": 0.7057010785824345, + "grad_norm": 1.5076516246345604, + "learning_rate": 2.1044687830492718e-06, + "loss": 0.3044, + "step": 32060 + }, + { + "epoch": 0.7057230904688532, + "grad_norm": 2.9604002307065733, + "learning_rate": 2.1041781886019547e-06, + "loss": 0.3264, + "step": 32061 + }, + { + "epoch": 0.7057451023552719, + "grad_norm": 1.3435561333107051, + "learning_rate": 2.1038876088724946e-06, + "loss": 0.2917, + "step": 32062 + }, + { + "epoch": 0.7057671142416905, + "grad_norm": 1.2741734734265258, + "learning_rate": 2.103597043862371e-06, + "loss": 0.2788, + "step": 32063 + }, + { + "epoch": 0.7057891261281092, + "grad_norm": 1.8206086657518465, + "learning_rate": 2.1033064935730567e-06, + "loss": 0.2205, + "step": 32064 + }, + { + "epoch": 0.7058111380145279, + "grad_norm": 1.4655222749525572, + "learning_rate": 2.103015958006031e-06, + "loss": 0.2836, + "step": 32065 + }, + { + "epoch": 0.7058331499009465, + "grad_norm": 0.6286538806208773, + "learning_rate": 2.1027254371627727e-06, + "loss": 0.4518, + "step": 32066 + }, + { + "epoch": 0.7058551617873652, + "grad_norm": 1.4989484910325752, + "learning_rate": 2.1024349310447557e-06, + "loss": 0.2352, + "step": 32067 + }, + { + "epoch": 0.7058771736737839, + "grad_norm": 2.7000463156018553, + "learning_rate": 2.1021444396534547e-06, + "loss": 0.3417, + "step": 32068 + }, + { + "epoch": 0.7058991855602025, + "grad_norm": 1.890419252899695, + "learning_rate": 2.1018539629903478e-06, + "loss": 0.3157, + "step": 32069 + }, + { + "epoch": 0.7059211974466212, + "grad_norm": 1.5582540613518596, + "learning_rate": 2.1015635010569117e-06, + "loss": 0.2908, + "step": 32070 + }, + { + "epoch": 0.7059432093330399, + "grad_norm": 1.658086166109475, + "learning_rate": 2.101273053854625e-06, + "loss": 0.3582, + "step": 32071 + }, + { + "epoch": 0.7059652212194585, + "grad_norm": 0.6053567394782371, + "learning_rate": 2.1009826213849615e-06, + "loss": 0.4723, + "step": 32072 + }, + { + "epoch": 0.7059872331058772, + "grad_norm": 2.3902532376804566, + "learning_rate": 2.1006922036493954e-06, + "loss": 0.2687, + "step": 32073 + }, + { + "epoch": 0.7060092449922959, + "grad_norm": 1.642020495038416, + "learning_rate": 2.100401800649404e-06, + "loss": 0.293, + "step": 32074 + }, + { + "epoch": 0.7060312568787145, + "grad_norm": 1.6015879481700546, + "learning_rate": 2.1001114123864646e-06, + "loss": 0.2266, + "step": 32075 + }, + { + "epoch": 0.7060532687651332, + "grad_norm": 0.607043836901103, + "learning_rate": 2.0998210388620543e-06, + "loss": 0.4647, + "step": 32076 + }, + { + "epoch": 0.7060752806515518, + "grad_norm": 1.5653381858150894, + "learning_rate": 2.0995306800776465e-06, + "loss": 0.2632, + "step": 32077 + }, + { + "epoch": 0.7060972925379705, + "grad_norm": 1.9937877776786255, + "learning_rate": 2.0992403360347164e-06, + "loss": 0.239, + "step": 32078 + }, + { + "epoch": 0.7061193044243892, + "grad_norm": 0.6103948712020786, + "learning_rate": 2.09895000673474e-06, + "loss": 0.4363, + "step": 32079 + }, + { + "epoch": 0.7061413163108078, + "grad_norm": 1.7888134617935099, + "learning_rate": 2.0986596921791934e-06, + "loss": 0.3164, + "step": 32080 + }, + { + "epoch": 0.7061633281972265, + "grad_norm": 0.5864696025148978, + "learning_rate": 2.0983693923695546e-06, + "loss": 0.4504, + "step": 32081 + }, + { + "epoch": 0.7061853400836452, + "grad_norm": 0.5991236790808824, + "learning_rate": 2.0980791073072963e-06, + "loss": 0.4523, + "step": 32082 + }, + { + "epoch": 0.7062073519700638, + "grad_norm": 1.82936379324113, + "learning_rate": 2.097788836993893e-06, + "loss": 0.2895, + "step": 32083 + }, + { + "epoch": 0.7062293638564825, + "grad_norm": 1.471953997437551, + "learning_rate": 2.097498581430821e-06, + "loss": 0.3519, + "step": 32084 + }, + { + "epoch": 0.7062513757429012, + "grad_norm": 2.2719738086954924, + "learning_rate": 2.0972083406195578e-06, + "loss": 0.3285, + "step": 32085 + }, + { + "epoch": 0.7062733876293198, + "grad_norm": 1.589765308923605, + "learning_rate": 2.096918114561574e-06, + "loss": 0.2645, + "step": 32086 + }, + { + "epoch": 0.7062953995157385, + "grad_norm": 1.8293616235429613, + "learning_rate": 2.0966279032583494e-06, + "loss": 0.3821, + "step": 32087 + }, + { + "epoch": 0.7063174114021572, + "grad_norm": 1.5986405711970046, + "learning_rate": 2.096337706711355e-06, + "loss": 0.3367, + "step": 32088 + }, + { + "epoch": 0.7063394232885758, + "grad_norm": 1.4937567985335878, + "learning_rate": 2.0960475249220673e-06, + "loss": 0.4088, + "step": 32089 + }, + { + "epoch": 0.7063614351749945, + "grad_norm": 1.5577188390514873, + "learning_rate": 2.095757357891963e-06, + "loss": 0.279, + "step": 32090 + }, + { + "epoch": 0.7063834470614132, + "grad_norm": 1.6309183844840949, + "learning_rate": 2.0954672056225133e-06, + "loss": 0.2869, + "step": 32091 + }, + { + "epoch": 0.7064054589478318, + "grad_norm": 1.883703399587411, + "learning_rate": 2.0951770681151966e-06, + "loss": 0.2953, + "step": 32092 + }, + { + "epoch": 0.7064274708342505, + "grad_norm": 1.9521215119776245, + "learning_rate": 2.094886945371484e-06, + "loss": 0.3902, + "step": 32093 + }, + { + "epoch": 0.7064494827206692, + "grad_norm": 1.6244828795056463, + "learning_rate": 2.0945968373928514e-06, + "loss": 0.2905, + "step": 32094 + }, + { + "epoch": 0.7064714946070878, + "grad_norm": 1.6723192989984792, + "learning_rate": 2.094306744180776e-06, + "loss": 0.2843, + "step": 32095 + }, + { + "epoch": 0.7064935064935065, + "grad_norm": 1.5593293080717339, + "learning_rate": 2.0940166657367272e-06, + "loss": 0.3385, + "step": 32096 + }, + { + "epoch": 0.7065155183799252, + "grad_norm": 1.5063253010282922, + "learning_rate": 2.0937266020621838e-06, + "loss": 0.2858, + "step": 32097 + }, + { + "epoch": 0.7065375302663438, + "grad_norm": 1.3934932233443849, + "learning_rate": 2.093436553158616e-06, + "loss": 0.2783, + "step": 32098 + }, + { + "epoch": 0.7065595421527625, + "grad_norm": 1.3899555968489372, + "learning_rate": 2.0931465190275017e-06, + "loss": 0.2525, + "step": 32099 + }, + { + "epoch": 0.7065815540391811, + "grad_norm": 2.4732478352474194, + "learning_rate": 2.092856499670312e-06, + "loss": 0.3466, + "step": 32100 + }, + { + "epoch": 0.7066035659255998, + "grad_norm": 1.9870976592540412, + "learning_rate": 2.0925664950885223e-06, + "loss": 0.3401, + "step": 32101 + }, + { + "epoch": 0.7066255778120185, + "grad_norm": 1.5202921314708242, + "learning_rate": 2.0922765052836075e-06, + "loss": 0.2565, + "step": 32102 + }, + { + "epoch": 0.7066475896984371, + "grad_norm": 0.6464022885510893, + "learning_rate": 2.091986530257039e-06, + "loss": 0.4698, + "step": 32103 + }, + { + "epoch": 0.7066696015848558, + "grad_norm": 0.6376794933048563, + "learning_rate": 2.0916965700102943e-06, + "loss": 0.4581, + "step": 32104 + }, + { + "epoch": 0.7066916134712745, + "grad_norm": 1.6865777872003147, + "learning_rate": 2.0914066245448424e-06, + "loss": 0.2921, + "step": 32105 + }, + { + "epoch": 0.7067136253576931, + "grad_norm": 1.751658842821326, + "learning_rate": 2.091116693862159e-06, + "loss": 0.2928, + "step": 32106 + }, + { + "epoch": 0.7067356372441118, + "grad_norm": 1.5669919400704233, + "learning_rate": 2.09082677796372e-06, + "loss": 0.3087, + "step": 32107 + }, + { + "epoch": 0.7067576491305305, + "grad_norm": 1.6840639151708927, + "learning_rate": 2.090536876850997e-06, + "loss": 0.2745, + "step": 32108 + }, + { + "epoch": 0.7067796610169491, + "grad_norm": 0.6052152371409304, + "learning_rate": 2.0902469905254614e-06, + "loss": 0.462, + "step": 32109 + }, + { + "epoch": 0.7068016729033678, + "grad_norm": 1.4199631811444784, + "learning_rate": 2.0899571189885887e-06, + "loss": 0.2724, + "step": 32110 + }, + { + "epoch": 0.7068236847897865, + "grad_norm": 1.4326462013951848, + "learning_rate": 2.0896672622418513e-06, + "loss": 0.2871, + "step": 32111 + }, + { + "epoch": 0.7068456966762051, + "grad_norm": 2.056019277148007, + "learning_rate": 2.089377420286725e-06, + "loss": 0.2756, + "step": 32112 + }, + { + "epoch": 0.7068677085626238, + "grad_norm": 1.7430692727531287, + "learning_rate": 2.0890875931246805e-06, + "loss": 0.355, + "step": 32113 + }, + { + "epoch": 0.7068897204490425, + "grad_norm": 1.7286607772439708, + "learning_rate": 2.0887977807571897e-06, + "loss": 0.3335, + "step": 32114 + }, + { + "epoch": 0.7069117323354611, + "grad_norm": 1.8545921804153782, + "learning_rate": 2.088507983185727e-06, + "loss": 0.309, + "step": 32115 + }, + { + "epoch": 0.7069337442218798, + "grad_norm": 2.9413831673499193, + "learning_rate": 2.0882182004117653e-06, + "loss": 0.292, + "step": 32116 + }, + { + "epoch": 0.7069557561082985, + "grad_norm": 1.7200731037961756, + "learning_rate": 2.0879284324367786e-06, + "loss": 0.3175, + "step": 32117 + }, + { + "epoch": 0.7069777679947171, + "grad_norm": 1.9041048193987378, + "learning_rate": 2.0876386792622392e-06, + "loss": 0.3467, + "step": 32118 + }, + { + "epoch": 0.7069997798811358, + "grad_norm": 2.043255335400061, + "learning_rate": 2.087348940889617e-06, + "loss": 0.2859, + "step": 32119 + }, + { + "epoch": 0.7070217917675545, + "grad_norm": 1.2865977413722032, + "learning_rate": 2.0870592173203865e-06, + "loss": 0.1849, + "step": 32120 + }, + { + "epoch": 0.7070438036539731, + "grad_norm": 1.6386904887924256, + "learning_rate": 2.0867695085560204e-06, + "loss": 0.3365, + "step": 32121 + }, + { + "epoch": 0.7070658155403918, + "grad_norm": 1.3953329389611695, + "learning_rate": 2.0864798145979927e-06, + "loss": 0.317, + "step": 32122 + }, + { + "epoch": 0.7070878274268104, + "grad_norm": 3.6744311642586127, + "learning_rate": 2.0861901354477737e-06, + "loss": 0.2523, + "step": 32123 + }, + { + "epoch": 0.7071098393132291, + "grad_norm": 1.6251585530080142, + "learning_rate": 2.085900471106834e-06, + "loss": 0.2961, + "step": 32124 + }, + { + "epoch": 0.7071318511996478, + "grad_norm": 1.39221844211906, + "learning_rate": 2.085610821576649e-06, + "loss": 0.2834, + "step": 32125 + }, + { + "epoch": 0.7071538630860664, + "grad_norm": 2.532194671921099, + "learning_rate": 2.0853211868586904e-06, + "loss": 0.3126, + "step": 32126 + }, + { + "epoch": 0.7071758749724851, + "grad_norm": 1.6568196097095602, + "learning_rate": 2.0850315669544286e-06, + "loss": 0.3343, + "step": 32127 + }, + { + "epoch": 0.7071978868589038, + "grad_norm": 1.5293329339571935, + "learning_rate": 2.0847419618653375e-06, + "loss": 0.2382, + "step": 32128 + }, + { + "epoch": 0.7072198987453224, + "grad_norm": 2.099912781975351, + "learning_rate": 2.0844523715928865e-06, + "loss": 0.3155, + "step": 32129 + }, + { + "epoch": 0.7072419106317411, + "grad_norm": 2.2173884742708267, + "learning_rate": 2.084162796138549e-06, + "loss": 0.3447, + "step": 32130 + }, + { + "epoch": 0.7072639225181598, + "grad_norm": 0.6224853227978451, + "learning_rate": 2.0838732355037983e-06, + "loss": 0.4571, + "step": 32131 + }, + { + "epoch": 0.7072859344045784, + "grad_norm": 0.6229881633793214, + "learning_rate": 2.083583689690103e-06, + "loss": 0.4615, + "step": 32132 + }, + { + "epoch": 0.7073079462909971, + "grad_norm": 1.737030645704803, + "learning_rate": 2.0832941586989375e-06, + "loss": 0.2248, + "step": 32133 + }, + { + "epoch": 0.7073299581774158, + "grad_norm": 0.6221073443596304, + "learning_rate": 2.08300464253177e-06, + "loss": 0.4706, + "step": 32134 + }, + { + "epoch": 0.7073519700638344, + "grad_norm": 1.7934455480045128, + "learning_rate": 2.0827151411900765e-06, + "loss": 0.2675, + "step": 32135 + }, + { + "epoch": 0.7073739819502531, + "grad_norm": 2.141628499682085, + "learning_rate": 2.0824256546753235e-06, + "loss": 0.338, + "step": 32136 + }, + { + "epoch": 0.7073959938366718, + "grad_norm": 2.0135034727447345, + "learning_rate": 2.0821361829889848e-06, + "loss": 0.2782, + "step": 32137 + }, + { + "epoch": 0.7074180057230904, + "grad_norm": 1.6364031333766764, + "learning_rate": 2.0818467261325327e-06, + "loss": 0.2685, + "step": 32138 + }, + { + "epoch": 0.7074400176095091, + "grad_norm": 1.7107113618914491, + "learning_rate": 2.0815572841074357e-06, + "loss": 0.2712, + "step": 32139 + }, + { + "epoch": 0.7074620294959278, + "grad_norm": 2.255460903643132, + "learning_rate": 2.081267856915168e-06, + "loss": 0.3626, + "step": 32140 + }, + { + "epoch": 0.7074840413823464, + "grad_norm": 1.5016254612352509, + "learning_rate": 2.0809784445571967e-06, + "loss": 0.2517, + "step": 32141 + }, + { + "epoch": 0.7075060532687651, + "grad_norm": 0.6145319757504062, + "learning_rate": 2.0806890470349956e-06, + "loss": 0.458, + "step": 32142 + }, + { + "epoch": 0.7075280651551838, + "grad_norm": 0.6262638838333812, + "learning_rate": 2.0803996643500364e-06, + "loss": 0.4481, + "step": 32143 + }, + { + "epoch": 0.7075500770416024, + "grad_norm": 0.6290927754276735, + "learning_rate": 2.080110296503788e-06, + "loss": 0.4593, + "step": 32144 + }, + { + "epoch": 0.7075720889280211, + "grad_norm": 1.2297100597370612, + "learning_rate": 2.0798209434977198e-06, + "loss": 0.2254, + "step": 32145 + }, + { + "epoch": 0.7075941008144399, + "grad_norm": 1.5506101637782546, + "learning_rate": 2.0795316053333036e-06, + "loss": 0.2634, + "step": 32146 + }, + { + "epoch": 0.7076161127008584, + "grad_norm": 0.6127234300692104, + "learning_rate": 2.0792422820120107e-06, + "loss": 0.4599, + "step": 32147 + }, + { + "epoch": 0.7076381245872772, + "grad_norm": 1.572567820591311, + "learning_rate": 2.0789529735353124e-06, + "loss": 0.335, + "step": 32148 + }, + { + "epoch": 0.7076601364736957, + "grad_norm": 1.327422083922626, + "learning_rate": 2.078663679904679e-06, + "loss": 0.2595, + "step": 32149 + }, + { + "epoch": 0.7076821483601144, + "grad_norm": 2.0520496454377377, + "learning_rate": 2.0783744011215773e-06, + "loss": 0.3504, + "step": 32150 + }, + { + "epoch": 0.7077041602465332, + "grad_norm": 1.8845674871106184, + "learning_rate": 2.07808513718748e-06, + "loss": 0.3481, + "step": 32151 + }, + { + "epoch": 0.7077261721329517, + "grad_norm": 1.7652740972239576, + "learning_rate": 2.0777958881038574e-06, + "loss": 0.284, + "step": 32152 + }, + { + "epoch": 0.7077481840193705, + "grad_norm": 5.670496454892797, + "learning_rate": 2.0775066538721813e-06, + "loss": 0.3646, + "step": 32153 + }, + { + "epoch": 0.7077701959057892, + "grad_norm": 1.8686434225290143, + "learning_rate": 2.0772174344939194e-06, + "loss": 0.2964, + "step": 32154 + }, + { + "epoch": 0.7077922077922078, + "grad_norm": 3.8354420475715827, + "learning_rate": 2.0769282299705407e-06, + "loss": 0.3827, + "step": 32155 + }, + { + "epoch": 0.7078142196786265, + "grad_norm": 9.428437092423534, + "learning_rate": 2.0766390403035164e-06, + "loss": 0.2869, + "step": 32156 + }, + { + "epoch": 0.7078362315650452, + "grad_norm": 2.0350217563696673, + "learning_rate": 2.076349865494316e-06, + "loss": 0.2896, + "step": 32157 + }, + { + "epoch": 0.7078582434514638, + "grad_norm": 1.3697657458213996, + "learning_rate": 2.076060705544412e-06, + "loss": 0.2919, + "step": 32158 + }, + { + "epoch": 0.7078802553378825, + "grad_norm": 0.6308934751077627, + "learning_rate": 2.0757715604552707e-06, + "loss": 0.4493, + "step": 32159 + }, + { + "epoch": 0.7079022672243012, + "grad_norm": 0.6147546650988109, + "learning_rate": 2.0754824302283616e-06, + "loss": 0.4721, + "step": 32160 + }, + { + "epoch": 0.7079242791107198, + "grad_norm": 1.8560457769886645, + "learning_rate": 2.075193314865155e-06, + "loss": 0.3564, + "step": 32161 + }, + { + "epoch": 0.7079462909971385, + "grad_norm": 1.3687698056433706, + "learning_rate": 2.0749042143671207e-06, + "loss": 0.2436, + "step": 32162 + }, + { + "epoch": 0.7079683028835572, + "grad_norm": 1.913191377282696, + "learning_rate": 2.0746151287357293e-06, + "loss": 0.3372, + "step": 32163 + }, + { + "epoch": 0.7079903147699758, + "grad_norm": 2.235071298614769, + "learning_rate": 2.074326057972448e-06, + "loss": 0.3286, + "step": 32164 + }, + { + "epoch": 0.7080123266563945, + "grad_norm": 1.7404096192003486, + "learning_rate": 2.0740370020787454e-06, + "loss": 0.2751, + "step": 32165 + }, + { + "epoch": 0.7080343385428132, + "grad_norm": 1.9802843604659524, + "learning_rate": 2.073747961056094e-06, + "loss": 0.2912, + "step": 32166 + }, + { + "epoch": 0.7080563504292318, + "grad_norm": 1.526722924951872, + "learning_rate": 2.073458934905958e-06, + "loss": 0.3306, + "step": 32167 + }, + { + "epoch": 0.7080783623156505, + "grad_norm": 2.1588669034649453, + "learning_rate": 2.073169923629809e-06, + "loss": 0.3924, + "step": 32168 + }, + { + "epoch": 0.7081003742020692, + "grad_norm": 0.6456611595826995, + "learning_rate": 2.0728809272291182e-06, + "loss": 0.4849, + "step": 32169 + }, + { + "epoch": 0.7081223860884878, + "grad_norm": 2.6834515076913488, + "learning_rate": 2.07259194570535e-06, + "loss": 0.3782, + "step": 32170 + }, + { + "epoch": 0.7081443979749065, + "grad_norm": 1.3868995943393578, + "learning_rate": 2.072302979059977e-06, + "loss": 0.308, + "step": 32171 + }, + { + "epoch": 0.7081664098613251, + "grad_norm": 1.8917507731446659, + "learning_rate": 2.0720140272944646e-06, + "loss": 0.224, + "step": 32172 + }, + { + "epoch": 0.7081884217477438, + "grad_norm": 2.4404238378217933, + "learning_rate": 2.071725090410282e-06, + "loss": 0.2626, + "step": 32173 + }, + { + "epoch": 0.7082104336341625, + "grad_norm": 1.9129903181721644, + "learning_rate": 2.0714361684089006e-06, + "loss": 0.3425, + "step": 32174 + }, + { + "epoch": 0.7082324455205811, + "grad_norm": 1.3660347951511154, + "learning_rate": 2.0711472612917853e-06, + "loss": 0.3017, + "step": 32175 + }, + { + "epoch": 0.7082544574069998, + "grad_norm": 1.4986784633958297, + "learning_rate": 2.070858369060408e-06, + "loss": 0.2411, + "step": 32176 + }, + { + "epoch": 0.7082764692934185, + "grad_norm": 2.307606016052551, + "learning_rate": 2.070569491716232e-06, + "loss": 0.2503, + "step": 32177 + }, + { + "epoch": 0.7082984811798371, + "grad_norm": 1.8487613300035983, + "learning_rate": 2.0702806292607297e-06, + "loss": 0.2947, + "step": 32178 + }, + { + "epoch": 0.7083204930662558, + "grad_norm": 0.6201652697483644, + "learning_rate": 2.069991781695369e-06, + "loss": 0.4491, + "step": 32179 + }, + { + "epoch": 0.7083425049526745, + "grad_norm": 1.8754722075005799, + "learning_rate": 2.0697029490216153e-06, + "loss": 0.363, + "step": 32180 + }, + { + "epoch": 0.7083645168390931, + "grad_norm": 1.5405756799527202, + "learning_rate": 2.06941413124094e-06, + "loss": 0.2529, + "step": 32181 + }, + { + "epoch": 0.7083865287255118, + "grad_norm": 1.5652096319998086, + "learning_rate": 2.0691253283548066e-06, + "loss": 0.3592, + "step": 32182 + }, + { + "epoch": 0.7084085406119305, + "grad_norm": 0.6074444924399007, + "learning_rate": 2.0688365403646866e-06, + "loss": 0.4462, + "step": 32183 + }, + { + "epoch": 0.7084305524983491, + "grad_norm": 1.5791363785508428, + "learning_rate": 2.068547767272048e-06, + "loss": 0.3651, + "step": 32184 + }, + { + "epoch": 0.7084525643847678, + "grad_norm": 5.203928052145521, + "learning_rate": 2.0682590090783573e-06, + "loss": 0.3777, + "step": 32185 + }, + { + "epoch": 0.7084745762711865, + "grad_norm": 1.404743710705193, + "learning_rate": 2.0679702657850797e-06, + "loss": 0.2341, + "step": 32186 + }, + { + "epoch": 0.7084965881576051, + "grad_norm": 1.6283945168377159, + "learning_rate": 2.0676815373936848e-06, + "loss": 0.2803, + "step": 32187 + }, + { + "epoch": 0.7085186000440238, + "grad_norm": 1.49281310120524, + "learning_rate": 2.0673928239056406e-06, + "loss": 0.3963, + "step": 32188 + }, + { + "epoch": 0.7085406119304425, + "grad_norm": 0.615431300121581, + "learning_rate": 2.0671041253224156e-06, + "loss": 0.4767, + "step": 32189 + }, + { + "epoch": 0.7085626238168611, + "grad_norm": 3.1371317217862327, + "learning_rate": 2.0668154416454762e-06, + "loss": 0.3418, + "step": 32190 + }, + { + "epoch": 0.7085846357032798, + "grad_norm": 1.6572802966261468, + "learning_rate": 2.0665267728762862e-06, + "loss": 0.302, + "step": 32191 + }, + { + "epoch": 0.7086066475896985, + "grad_norm": 1.690828101320601, + "learning_rate": 2.066238119016316e-06, + "loss": 0.285, + "step": 32192 + }, + { + "epoch": 0.7086286594761171, + "grad_norm": 2.118170159720011, + "learning_rate": 2.065949480067032e-06, + "loss": 0.2412, + "step": 32193 + }, + { + "epoch": 0.7086506713625358, + "grad_norm": 1.7393525034377866, + "learning_rate": 2.0656608560299035e-06, + "loss": 0.292, + "step": 32194 + }, + { + "epoch": 0.7086726832489544, + "grad_norm": 1.3809149779788825, + "learning_rate": 2.065372246906395e-06, + "loss": 0.4024, + "step": 32195 + }, + { + "epoch": 0.7086946951353731, + "grad_norm": 0.6346094803160874, + "learning_rate": 2.0650836526979716e-06, + "loss": 0.4539, + "step": 32196 + }, + { + "epoch": 0.7087167070217918, + "grad_norm": 0.628795218155144, + "learning_rate": 2.0647950734061027e-06, + "loss": 0.4731, + "step": 32197 + }, + { + "epoch": 0.7087387189082104, + "grad_norm": 1.3204938608397274, + "learning_rate": 2.0645065090322543e-06, + "loss": 0.2855, + "step": 32198 + }, + { + "epoch": 0.7087607307946291, + "grad_norm": 1.5253239956439912, + "learning_rate": 2.0642179595778943e-06, + "loss": 0.3108, + "step": 32199 + }, + { + "epoch": 0.7087827426810478, + "grad_norm": 0.6026076977685744, + "learning_rate": 2.0639294250444885e-06, + "loss": 0.4396, + "step": 32200 + }, + { + "epoch": 0.7088047545674664, + "grad_norm": 1.665374884315295, + "learning_rate": 2.063640905433501e-06, + "loss": 0.3453, + "step": 32201 + }, + { + "epoch": 0.7088267664538851, + "grad_norm": 6.140295999289732, + "learning_rate": 2.0633524007464018e-06, + "loss": 0.2057, + "step": 32202 + }, + { + "epoch": 0.7088487783403038, + "grad_norm": 1.9878991853011754, + "learning_rate": 2.0630639109846516e-06, + "loss": 0.2921, + "step": 32203 + }, + { + "epoch": 0.7088707902267224, + "grad_norm": 1.3589005956293634, + "learning_rate": 2.0627754361497247e-06, + "loss": 0.2926, + "step": 32204 + }, + { + "epoch": 0.7088928021131411, + "grad_norm": 1.166168925423721, + "learning_rate": 2.0624869762430827e-06, + "loss": 0.309, + "step": 32205 + }, + { + "epoch": 0.7089148139995598, + "grad_norm": 1.2512470496710462, + "learning_rate": 2.06219853126619e-06, + "loss": 0.3075, + "step": 32206 + }, + { + "epoch": 0.7089368258859784, + "grad_norm": 1.8645532528695927, + "learning_rate": 2.061910101220517e-06, + "loss": 0.2797, + "step": 32207 + }, + { + "epoch": 0.7089588377723971, + "grad_norm": 1.5461996722760984, + "learning_rate": 2.0616216861075257e-06, + "loss": 0.2719, + "step": 32208 + }, + { + "epoch": 0.7089808496588158, + "grad_norm": 2.3131147691400296, + "learning_rate": 2.0613332859286827e-06, + "loss": 0.3104, + "step": 32209 + }, + { + "epoch": 0.7090028615452344, + "grad_norm": 1.3889885969130695, + "learning_rate": 2.0610449006854573e-06, + "loss": 0.3155, + "step": 32210 + }, + { + "epoch": 0.7090248734316531, + "grad_norm": 1.3712024217355823, + "learning_rate": 2.06075653037931e-06, + "loss": 0.3, + "step": 32211 + }, + { + "epoch": 0.7090468853180718, + "grad_norm": 2.7366459258792775, + "learning_rate": 2.060468175011711e-06, + "loss": 0.3098, + "step": 32212 + }, + { + "epoch": 0.7090688972044904, + "grad_norm": 5.923074963008418, + "learning_rate": 2.0601798345841225e-06, + "loss": 0.3316, + "step": 32213 + }, + { + "epoch": 0.7090909090909091, + "grad_norm": 1.6148772881225628, + "learning_rate": 2.0598915090980106e-06, + "loss": 0.2967, + "step": 32214 + }, + { + "epoch": 0.7091129209773278, + "grad_norm": 0.6150015182157887, + "learning_rate": 2.059603198554843e-06, + "loss": 0.4443, + "step": 32215 + }, + { + "epoch": 0.7091349328637464, + "grad_norm": 1.5473508014470145, + "learning_rate": 2.059314902956082e-06, + "loss": 0.3004, + "step": 32216 + }, + { + "epoch": 0.7091569447501651, + "grad_norm": 2.018301711439828, + "learning_rate": 2.059026622303196e-06, + "loss": 0.339, + "step": 32217 + }, + { + "epoch": 0.7091789566365837, + "grad_norm": 1.672132422737485, + "learning_rate": 2.058738356597647e-06, + "loss": 0.3024, + "step": 32218 + }, + { + "epoch": 0.7092009685230024, + "grad_norm": 1.5313623694483856, + "learning_rate": 2.0584501058409013e-06, + "loss": 0.291, + "step": 32219 + }, + { + "epoch": 0.7092229804094211, + "grad_norm": 3.0910662468057915, + "learning_rate": 2.058161870034426e-06, + "loss": 0.3412, + "step": 32220 + }, + { + "epoch": 0.7092449922958397, + "grad_norm": 0.6197301073850137, + "learning_rate": 2.0578736491796818e-06, + "loss": 0.4335, + "step": 32221 + }, + { + "epoch": 0.7092670041822584, + "grad_norm": 0.6042976864927245, + "learning_rate": 2.0575854432781383e-06, + "loss": 0.4444, + "step": 32222 + }, + { + "epoch": 0.7092890160686771, + "grad_norm": 2.2951731523361825, + "learning_rate": 2.0572972523312555e-06, + "loss": 0.298, + "step": 32223 + }, + { + "epoch": 0.7093110279550957, + "grad_norm": 1.6776946635570724, + "learning_rate": 2.0570090763405016e-06, + "loss": 0.3283, + "step": 32224 + }, + { + "epoch": 0.7093330398415144, + "grad_norm": 1.477964509280304, + "learning_rate": 2.0567209153073407e-06, + "loss": 0.2947, + "step": 32225 + }, + { + "epoch": 0.7093550517279331, + "grad_norm": 2.5098155676555987, + "learning_rate": 2.0564327692332382e-06, + "loss": 0.292, + "step": 32226 + }, + { + "epoch": 0.7093770636143517, + "grad_norm": 1.7833200463521297, + "learning_rate": 2.0561446381196544e-06, + "loss": 0.3113, + "step": 32227 + }, + { + "epoch": 0.7093990755007704, + "grad_norm": 1.2968438211109736, + "learning_rate": 2.0558565219680572e-06, + "loss": 0.2454, + "step": 32228 + }, + { + "epoch": 0.7094210873871891, + "grad_norm": 0.6221615959576396, + "learning_rate": 2.05556842077991e-06, + "loss": 0.4613, + "step": 32229 + }, + { + "epoch": 0.7094430992736077, + "grad_norm": 1.534283243618593, + "learning_rate": 2.0552803345566786e-06, + "loss": 0.2536, + "step": 32230 + }, + { + "epoch": 0.7094651111600264, + "grad_norm": 1.8197971671884687, + "learning_rate": 2.0549922632998266e-06, + "loss": 0.2909, + "step": 32231 + }, + { + "epoch": 0.7094871230464451, + "grad_norm": 3.6739741631056817, + "learning_rate": 2.0547042070108155e-06, + "loss": 0.2778, + "step": 32232 + }, + { + "epoch": 0.7095091349328637, + "grad_norm": 1.631191301127379, + "learning_rate": 2.054416165691111e-06, + "loss": 0.327, + "step": 32233 + }, + { + "epoch": 0.7095311468192824, + "grad_norm": 1.5872380053483508, + "learning_rate": 2.054128139342178e-06, + "loss": 0.3039, + "step": 32234 + }, + { + "epoch": 0.7095531587057011, + "grad_norm": 1.6199844903457508, + "learning_rate": 2.053840127965481e-06, + "loss": 0.2134, + "step": 32235 + }, + { + "epoch": 0.7095751705921197, + "grad_norm": 1.693260321218356, + "learning_rate": 2.053552131562482e-06, + "loss": 0.3239, + "step": 32236 + }, + { + "epoch": 0.7095971824785384, + "grad_norm": 1.7253603085754068, + "learning_rate": 2.053264150134644e-06, + "loss": 0.3794, + "step": 32237 + }, + { + "epoch": 0.7096191943649571, + "grad_norm": 2.332643472103677, + "learning_rate": 2.052976183683434e-06, + "loss": 0.376, + "step": 32238 + }, + { + "epoch": 0.7096412062513757, + "grad_norm": 1.6910216870508141, + "learning_rate": 2.0526882322103093e-06, + "loss": 0.3043, + "step": 32239 + }, + { + "epoch": 0.7096632181377944, + "grad_norm": 5.4269843906468065, + "learning_rate": 2.052400295716741e-06, + "loss": 0.2636, + "step": 32240 + }, + { + "epoch": 0.709685230024213, + "grad_norm": 1.4860956018605909, + "learning_rate": 2.05211237420419e-06, + "loss": 0.27, + "step": 32241 + }, + { + "epoch": 0.7097072419106317, + "grad_norm": 1.6507119019213932, + "learning_rate": 2.051824467674116e-06, + "loss": 0.369, + "step": 32242 + }, + { + "epoch": 0.7097292537970504, + "grad_norm": 2.4071294405588697, + "learning_rate": 2.051536576127987e-06, + "loss": 0.2745, + "step": 32243 + }, + { + "epoch": 0.709751265683469, + "grad_norm": 1.283503547155002, + "learning_rate": 2.0512486995672627e-06, + "loss": 0.2285, + "step": 32244 + }, + { + "epoch": 0.7097732775698877, + "grad_norm": 1.534144789647642, + "learning_rate": 2.050960837993407e-06, + "loss": 0.3334, + "step": 32245 + }, + { + "epoch": 0.7097952894563064, + "grad_norm": 1.5650264103568055, + "learning_rate": 2.050672991407886e-06, + "loss": 0.3326, + "step": 32246 + }, + { + "epoch": 0.709817301342725, + "grad_norm": 3.153147118173596, + "learning_rate": 2.0503851598121584e-06, + "loss": 0.3877, + "step": 32247 + }, + { + "epoch": 0.7098393132291437, + "grad_norm": 2.1155028768242548, + "learning_rate": 2.0500973432076902e-06, + "loss": 0.3208, + "step": 32248 + }, + { + "epoch": 0.7098613251155624, + "grad_norm": 1.6881946359518798, + "learning_rate": 2.0498095415959424e-06, + "loss": 0.269, + "step": 32249 + }, + { + "epoch": 0.709883337001981, + "grad_norm": 1.4727529494918303, + "learning_rate": 2.0495217549783776e-06, + "loss": 0.2935, + "step": 32250 + }, + { + "epoch": 0.7099053488883997, + "grad_norm": 1.8706447469979595, + "learning_rate": 2.0492339833564606e-06, + "loss": 0.2793, + "step": 32251 + }, + { + "epoch": 0.7099273607748184, + "grad_norm": 1.305343201144391, + "learning_rate": 2.048946226731651e-06, + "loss": 0.3278, + "step": 32252 + }, + { + "epoch": 0.709949372661237, + "grad_norm": 2.0122918693319605, + "learning_rate": 2.0486584851054154e-06, + "loss": 0.2644, + "step": 32253 + }, + { + "epoch": 0.7099713845476557, + "grad_norm": 1.485284602975941, + "learning_rate": 2.048370758479211e-06, + "loss": 0.3512, + "step": 32254 + }, + { + "epoch": 0.7099933964340744, + "grad_norm": 1.4365574692842669, + "learning_rate": 2.048083046854504e-06, + "loss": 0.3625, + "step": 32255 + }, + { + "epoch": 0.710015408320493, + "grad_norm": 1.6326558181293211, + "learning_rate": 2.047795350232756e-06, + "loss": 0.3127, + "step": 32256 + }, + { + "epoch": 0.7100374202069117, + "grad_norm": 2.94784087533485, + "learning_rate": 2.0475076686154276e-06, + "loss": 0.2575, + "step": 32257 + }, + { + "epoch": 0.7100594320933304, + "grad_norm": 1.3605230219927344, + "learning_rate": 2.047220002003984e-06, + "loss": 0.3309, + "step": 32258 + }, + { + "epoch": 0.710081443979749, + "grad_norm": 0.6109846193033459, + "learning_rate": 2.0469323503998835e-06, + "loss": 0.476, + "step": 32259 + }, + { + "epoch": 0.7101034558661677, + "grad_norm": 0.6171431725072923, + "learning_rate": 2.0466447138045897e-06, + "loss": 0.4517, + "step": 32260 + }, + { + "epoch": 0.7101254677525864, + "grad_norm": 0.6359226892417262, + "learning_rate": 2.0463570922195664e-06, + "loss": 0.452, + "step": 32261 + }, + { + "epoch": 0.710147479639005, + "grad_norm": 0.6251341183452087, + "learning_rate": 2.046069485646272e-06, + "loss": 0.4471, + "step": 32262 + }, + { + "epoch": 0.7101694915254237, + "grad_norm": 1.8541020997622562, + "learning_rate": 2.045781894086172e-06, + "loss": 0.2926, + "step": 32263 + }, + { + "epoch": 0.7101915034118423, + "grad_norm": 1.7823909650004077, + "learning_rate": 2.0454943175407244e-06, + "loss": 0.342, + "step": 32264 + }, + { + "epoch": 0.710213515298261, + "grad_norm": 1.7115365398483966, + "learning_rate": 2.045206756011392e-06, + "loss": 0.2887, + "step": 32265 + }, + { + "epoch": 0.7102355271846797, + "grad_norm": 1.646085941116464, + "learning_rate": 2.0449192094996383e-06, + "loss": 0.3441, + "step": 32266 + }, + { + "epoch": 0.7102575390710983, + "grad_norm": 1.5734615063009876, + "learning_rate": 2.0446316780069234e-06, + "loss": 0.4279, + "step": 32267 + }, + { + "epoch": 0.710279550957517, + "grad_norm": 1.7649870839571475, + "learning_rate": 2.044344161534707e-06, + "loss": 0.3018, + "step": 32268 + }, + { + "epoch": 0.7103015628439358, + "grad_norm": 1.7341534383299313, + "learning_rate": 2.044056660084452e-06, + "loss": 0.3285, + "step": 32269 + }, + { + "epoch": 0.7103235747303543, + "grad_norm": 3.3435088997536906, + "learning_rate": 2.0437691736576187e-06, + "loss": 0.2681, + "step": 32270 + }, + { + "epoch": 0.710345586616773, + "grad_norm": 1.8061737359656131, + "learning_rate": 2.0434817022556713e-06, + "loss": 0.1798, + "step": 32271 + }, + { + "epoch": 0.7103675985031918, + "grad_norm": 2.248448292479386, + "learning_rate": 2.0431942458800685e-06, + "loss": 0.3338, + "step": 32272 + }, + { + "epoch": 0.7103896103896103, + "grad_norm": 1.4981916658153516, + "learning_rate": 2.0429068045322692e-06, + "loss": 0.3256, + "step": 32273 + }, + { + "epoch": 0.710411622276029, + "grad_norm": 1.5840218273369604, + "learning_rate": 2.042619378213738e-06, + "loss": 0.2827, + "step": 32274 + }, + { + "epoch": 0.7104336341624478, + "grad_norm": 1.5154019980137843, + "learning_rate": 2.042331966925931e-06, + "loss": 0.1879, + "step": 32275 + }, + { + "epoch": 0.7104556460488664, + "grad_norm": 0.6153614539149277, + "learning_rate": 2.042044570670316e-06, + "loss": 0.4295, + "step": 32276 + }, + { + "epoch": 0.7104776579352851, + "grad_norm": 1.967658438308323, + "learning_rate": 2.0417571894483495e-06, + "loss": 0.3171, + "step": 32277 + }, + { + "epoch": 0.7104996698217038, + "grad_norm": 1.860322611993899, + "learning_rate": 2.0414698232614903e-06, + "loss": 0.3286, + "step": 32278 + }, + { + "epoch": 0.7105216817081224, + "grad_norm": 2.1388065178642024, + "learning_rate": 2.0411824721112027e-06, + "loss": 0.2647, + "step": 32279 + }, + { + "epoch": 0.7105436935945411, + "grad_norm": 2.8632948112295193, + "learning_rate": 2.0408951359989424e-06, + "loss": 0.3259, + "step": 32280 + }, + { + "epoch": 0.7105657054809598, + "grad_norm": 0.6152427786841691, + "learning_rate": 2.040607814926176e-06, + "loss": 0.4461, + "step": 32281 + }, + { + "epoch": 0.7105877173673784, + "grad_norm": 0.6174727656148039, + "learning_rate": 2.0403205088943614e-06, + "loss": 0.4678, + "step": 32282 + }, + { + "epoch": 0.7106097292537971, + "grad_norm": 1.5298665920567611, + "learning_rate": 2.0400332179049552e-06, + "loss": 0.3344, + "step": 32283 + }, + { + "epoch": 0.7106317411402158, + "grad_norm": 1.2261324053977614, + "learning_rate": 2.039745941959422e-06, + "loss": 0.1996, + "step": 32284 + }, + { + "epoch": 0.7106537530266344, + "grad_norm": 2.61924850159487, + "learning_rate": 2.039458681059219e-06, + "loss": 0.2363, + "step": 32285 + }, + { + "epoch": 0.7106757649130531, + "grad_norm": 0.6028321670254259, + "learning_rate": 2.039171435205808e-06, + "loss": 0.4474, + "step": 32286 + }, + { + "epoch": 0.7106977767994718, + "grad_norm": 0.6252204067559523, + "learning_rate": 2.0388842044006495e-06, + "loss": 0.4538, + "step": 32287 + }, + { + "epoch": 0.7107197886858904, + "grad_norm": 1.605350623759804, + "learning_rate": 2.0385969886452006e-06, + "loss": 0.2204, + "step": 32288 + }, + { + "epoch": 0.7107418005723091, + "grad_norm": 1.996084098639252, + "learning_rate": 2.038309787940924e-06, + "loss": 0.274, + "step": 32289 + }, + { + "epoch": 0.7107638124587277, + "grad_norm": 1.8174074588081364, + "learning_rate": 2.0380226022892764e-06, + "loss": 0.2563, + "step": 32290 + }, + { + "epoch": 0.7107858243451464, + "grad_norm": 1.3541352534930544, + "learning_rate": 2.037735431691719e-06, + "loss": 0.2571, + "step": 32291 + }, + { + "epoch": 0.7108078362315651, + "grad_norm": 1.3451009917627792, + "learning_rate": 2.0374482761497135e-06, + "loss": 0.2292, + "step": 32292 + }, + { + "epoch": 0.7108298481179837, + "grad_norm": 1.6516463735419866, + "learning_rate": 2.037161135664715e-06, + "loss": 0.3037, + "step": 32293 + }, + { + "epoch": 0.7108518600044024, + "grad_norm": 0.6092270944045741, + "learning_rate": 2.036874010238187e-06, + "loss": 0.4588, + "step": 32294 + }, + { + "epoch": 0.7108738718908211, + "grad_norm": 1.859646095307118, + "learning_rate": 2.0365868998715848e-06, + "loss": 0.3507, + "step": 32295 + }, + { + "epoch": 0.7108958837772397, + "grad_norm": 0.6019283509672833, + "learning_rate": 2.0362998045663703e-06, + "loss": 0.4371, + "step": 32296 + }, + { + "epoch": 0.7109178956636584, + "grad_norm": 1.8928144552007262, + "learning_rate": 2.0360127243240037e-06, + "loss": 0.3651, + "step": 32297 + }, + { + "epoch": 0.7109399075500771, + "grad_norm": 1.4234298273543888, + "learning_rate": 2.0357256591459402e-06, + "loss": 0.2553, + "step": 32298 + }, + { + "epoch": 0.7109619194364957, + "grad_norm": 2.3660941772211275, + "learning_rate": 2.035438609033643e-06, + "loss": 0.4006, + "step": 32299 + }, + { + "epoch": 0.7109839313229144, + "grad_norm": 1.6550851962278685, + "learning_rate": 2.035151573988567e-06, + "loss": 0.303, + "step": 32300 + }, + { + "epoch": 0.7110059432093331, + "grad_norm": 0.6042790982360298, + "learning_rate": 2.0348645540121733e-06, + "loss": 0.4329, + "step": 32301 + }, + { + "epoch": 0.7110279550957517, + "grad_norm": 1.3024235314358819, + "learning_rate": 2.034577549105922e-06, + "loss": 0.2712, + "step": 32302 + }, + { + "epoch": 0.7110499669821704, + "grad_norm": 1.8483772056311751, + "learning_rate": 2.0342905592712686e-06, + "loss": 0.2739, + "step": 32303 + }, + { + "epoch": 0.7110719788685891, + "grad_norm": 1.3295069753200628, + "learning_rate": 2.0340035845096756e-06, + "loss": 0.3024, + "step": 32304 + }, + { + "epoch": 0.7110939907550077, + "grad_norm": 1.6176372306725002, + "learning_rate": 2.033716624822597e-06, + "loss": 0.2884, + "step": 32305 + }, + { + "epoch": 0.7111160026414264, + "grad_norm": 1.235383263846868, + "learning_rate": 2.0334296802114935e-06, + "loss": 0.2455, + "step": 32306 + }, + { + "epoch": 0.7111380145278451, + "grad_norm": 2.1382828921549835, + "learning_rate": 2.033142750677825e-06, + "loss": 0.2859, + "step": 32307 + }, + { + "epoch": 0.7111600264142637, + "grad_norm": 1.8605373017334297, + "learning_rate": 2.0328558362230487e-06, + "loss": 0.2678, + "step": 32308 + }, + { + "epoch": 0.7111820383006824, + "grad_norm": 1.6373760927587386, + "learning_rate": 2.0325689368486205e-06, + "loss": 0.2888, + "step": 32309 + }, + { + "epoch": 0.7112040501871011, + "grad_norm": 1.980116314153587, + "learning_rate": 2.0322820525560023e-06, + "loss": 0.1903, + "step": 32310 + }, + { + "epoch": 0.7112260620735197, + "grad_norm": 0.6216664119172094, + "learning_rate": 2.0319951833466467e-06, + "loss": 0.4553, + "step": 32311 + }, + { + "epoch": 0.7112480739599384, + "grad_norm": 2.052700587467874, + "learning_rate": 2.031708329222019e-06, + "loss": 0.2968, + "step": 32312 + }, + { + "epoch": 0.711270085846357, + "grad_norm": 1.8789277300740939, + "learning_rate": 2.031421490183573e-06, + "loss": 0.2936, + "step": 32313 + }, + { + "epoch": 0.7112920977327757, + "grad_norm": 1.3611959100517406, + "learning_rate": 2.0311346662327656e-06, + "loss": 0.2632, + "step": 32314 + }, + { + "epoch": 0.7113141096191944, + "grad_norm": 1.3676980634874198, + "learning_rate": 2.0308478573710584e-06, + "loss": 0.2088, + "step": 32315 + }, + { + "epoch": 0.711336121505613, + "grad_norm": 0.6080581729009955, + "learning_rate": 2.030561063599904e-06, + "loss": 0.4484, + "step": 32316 + }, + { + "epoch": 0.7113581333920317, + "grad_norm": 1.6637661675370834, + "learning_rate": 2.030274284920763e-06, + "loss": 0.3703, + "step": 32317 + }, + { + "epoch": 0.7113801452784504, + "grad_norm": 1.5366598664342663, + "learning_rate": 2.0299875213350943e-06, + "loss": 0.258, + "step": 32318 + }, + { + "epoch": 0.711402157164869, + "grad_norm": 1.375085901524509, + "learning_rate": 2.0297007728443526e-06, + "loss": 0.2632, + "step": 32319 + }, + { + "epoch": 0.7114241690512877, + "grad_norm": 0.6108708905777703, + "learning_rate": 2.0294140394499975e-06, + "loss": 0.4697, + "step": 32320 + }, + { + "epoch": 0.7114461809377064, + "grad_norm": 2.2895834171985703, + "learning_rate": 2.029127321153483e-06, + "loss": 0.3344, + "step": 32321 + }, + { + "epoch": 0.711468192824125, + "grad_norm": 1.37380124718029, + "learning_rate": 2.028840617956269e-06, + "loss": 0.2269, + "step": 32322 + }, + { + "epoch": 0.7114902047105437, + "grad_norm": 1.5302412187750283, + "learning_rate": 2.028553929859814e-06, + "loss": 0.2817, + "step": 32323 + }, + { + "epoch": 0.7115122165969624, + "grad_norm": 0.6169992623466533, + "learning_rate": 2.028267256865571e-06, + "loss": 0.4694, + "step": 32324 + }, + { + "epoch": 0.711534228483381, + "grad_norm": 1.6081246075081215, + "learning_rate": 2.027980598975001e-06, + "loss": 0.3057, + "step": 32325 + }, + { + "epoch": 0.7115562403697997, + "grad_norm": 1.9268369853253344, + "learning_rate": 2.027693956189558e-06, + "loss": 0.3416, + "step": 32326 + }, + { + "epoch": 0.7115782522562184, + "grad_norm": 1.4151678199216793, + "learning_rate": 2.0274073285107e-06, + "loss": 0.182, + "step": 32327 + }, + { + "epoch": 0.711600264142637, + "grad_norm": 1.9322473453328228, + "learning_rate": 2.0271207159398854e-06, + "loss": 0.2974, + "step": 32328 + }, + { + "epoch": 0.7116222760290557, + "grad_norm": 1.468212970552408, + "learning_rate": 2.0268341184785674e-06, + "loss": 0.2305, + "step": 32329 + }, + { + "epoch": 0.7116442879154744, + "grad_norm": 1.3567904608210029, + "learning_rate": 2.0265475361282057e-06, + "loss": 0.2856, + "step": 32330 + }, + { + "epoch": 0.711666299801893, + "grad_norm": 3.648852873825261, + "learning_rate": 2.0262609688902548e-06, + "loss": 0.3009, + "step": 32331 + }, + { + "epoch": 0.7116883116883117, + "grad_norm": 1.772017325019786, + "learning_rate": 2.025974416766171e-06, + "loss": 0.288, + "step": 32332 + }, + { + "epoch": 0.7117103235747304, + "grad_norm": 2.228255753567288, + "learning_rate": 2.025687879757414e-06, + "loss": 0.3601, + "step": 32333 + }, + { + "epoch": 0.711732335461149, + "grad_norm": 1.4246361909789285, + "learning_rate": 2.025401357865436e-06, + "loss": 0.2786, + "step": 32334 + }, + { + "epoch": 0.7117543473475677, + "grad_norm": 1.8948395873952304, + "learning_rate": 2.025114851091697e-06, + "loss": 0.3082, + "step": 32335 + }, + { + "epoch": 0.7117763592339863, + "grad_norm": 0.6078937273137293, + "learning_rate": 2.024828359437649e-06, + "loss": 0.4598, + "step": 32336 + }, + { + "epoch": 0.711798371120405, + "grad_norm": 1.8894185555395189, + "learning_rate": 2.02454188290475e-06, + "loss": 0.3751, + "step": 32337 + }, + { + "epoch": 0.7118203830068237, + "grad_norm": 4.410143615877621, + "learning_rate": 2.0242554214944587e-06, + "loss": 0.2901, + "step": 32338 + }, + { + "epoch": 0.7118423948932423, + "grad_norm": 2.9920705928478557, + "learning_rate": 2.0239689752082263e-06, + "loss": 0.2815, + "step": 32339 + }, + { + "epoch": 0.711864406779661, + "grad_norm": 1.9433222046132088, + "learning_rate": 2.0236825440475126e-06, + "loss": 0.265, + "step": 32340 + }, + { + "epoch": 0.7118864186660797, + "grad_norm": 1.3894603952148823, + "learning_rate": 2.0233961280137705e-06, + "loss": 0.2588, + "step": 32341 + }, + { + "epoch": 0.7119084305524983, + "grad_norm": 1.4464114238653363, + "learning_rate": 2.0231097271084566e-06, + "loss": 0.2873, + "step": 32342 + }, + { + "epoch": 0.711930442438917, + "grad_norm": 0.6149754257003123, + "learning_rate": 2.0228233413330286e-06, + "loss": 0.4669, + "step": 32343 + }, + { + "epoch": 0.7119524543253357, + "grad_norm": 2.40764951076859, + "learning_rate": 2.02253697068894e-06, + "loss": 0.3398, + "step": 32344 + }, + { + "epoch": 0.7119744662117543, + "grad_norm": 0.6223453237757978, + "learning_rate": 2.022250615177645e-06, + "loss": 0.4758, + "step": 32345 + }, + { + "epoch": 0.711996478098173, + "grad_norm": 2.420998543068796, + "learning_rate": 2.021964274800602e-06, + "loss": 0.2629, + "step": 32346 + }, + { + "epoch": 0.7120184899845917, + "grad_norm": 2.3529179843835526, + "learning_rate": 2.0216779495592616e-06, + "loss": 0.2972, + "step": 32347 + }, + { + "epoch": 0.7120405018710103, + "grad_norm": 1.4769982256807839, + "learning_rate": 2.0213916394550855e-06, + "loss": 0.2967, + "step": 32348 + }, + { + "epoch": 0.712062513757429, + "grad_norm": 0.5940230594469079, + "learning_rate": 2.0211053444895257e-06, + "loss": 0.4505, + "step": 32349 + }, + { + "epoch": 0.7120845256438477, + "grad_norm": 2.7734403580160634, + "learning_rate": 2.020819064664035e-06, + "loss": 0.2379, + "step": 32350 + }, + { + "epoch": 0.7121065375302663, + "grad_norm": 1.3220802559475489, + "learning_rate": 2.0205327999800726e-06, + "loss": 0.3175, + "step": 32351 + }, + { + "epoch": 0.712128549416685, + "grad_norm": 2.2917138373357058, + "learning_rate": 2.02024655043909e-06, + "loss": 0.3365, + "step": 32352 + }, + { + "epoch": 0.7121505613031037, + "grad_norm": 0.6420821300249047, + "learning_rate": 2.019960316042543e-06, + "loss": 0.4566, + "step": 32353 + }, + { + "epoch": 0.7121725731895223, + "grad_norm": 1.2188405239454934, + "learning_rate": 2.019674096791889e-06, + "loss": 0.2883, + "step": 32354 + }, + { + "epoch": 0.712194585075941, + "grad_norm": 1.5622124393408134, + "learning_rate": 2.019387892688578e-06, + "loss": 0.3101, + "step": 32355 + }, + { + "epoch": 0.7122165969623597, + "grad_norm": 2.359451783427872, + "learning_rate": 2.0191017037340695e-06, + "loss": 0.3536, + "step": 32356 + }, + { + "epoch": 0.7122386088487783, + "grad_norm": 0.6297209653065987, + "learning_rate": 2.0188155299298134e-06, + "loss": 0.4851, + "step": 32357 + }, + { + "epoch": 0.712260620735197, + "grad_norm": 3.6483547941890806, + "learning_rate": 2.018529371277267e-06, + "loss": 0.3118, + "step": 32358 + }, + { + "epoch": 0.7122826326216156, + "grad_norm": 1.2049120478385495, + "learning_rate": 2.0182432277778856e-06, + "loss": 0.3186, + "step": 32359 + }, + { + "epoch": 0.7123046445080343, + "grad_norm": 1.711786413700731, + "learning_rate": 2.0179570994331205e-06, + "loss": 0.2172, + "step": 32360 + }, + { + "epoch": 0.712326656394453, + "grad_norm": 1.4825495413302185, + "learning_rate": 2.017670986244429e-06, + "loss": 0.2442, + "step": 32361 + }, + { + "epoch": 0.7123486682808716, + "grad_norm": 1.5864342202574246, + "learning_rate": 2.017384888213262e-06, + "loss": 0.2882, + "step": 32362 + }, + { + "epoch": 0.7123706801672903, + "grad_norm": 2.1585996619631147, + "learning_rate": 2.017098805341075e-06, + "loss": 0.3128, + "step": 32363 + }, + { + "epoch": 0.712392692053709, + "grad_norm": 0.6051128565330287, + "learning_rate": 2.0168127376293246e-06, + "loss": 0.453, + "step": 32364 + }, + { + "epoch": 0.7124147039401276, + "grad_norm": 1.7022092471533645, + "learning_rate": 2.01652668507946e-06, + "loss": 0.3469, + "step": 32365 + }, + { + "epoch": 0.7124367158265463, + "grad_norm": 1.2382660024812813, + "learning_rate": 2.01624064769294e-06, + "loss": 0.2957, + "step": 32366 + }, + { + "epoch": 0.712458727712965, + "grad_norm": 2.144205953236671, + "learning_rate": 2.015954625471213e-06, + "loss": 0.3039, + "step": 32367 + }, + { + "epoch": 0.7124807395993836, + "grad_norm": 1.398911888877772, + "learning_rate": 2.015668618415736e-06, + "loss": 0.231, + "step": 32368 + }, + { + "epoch": 0.7125027514858023, + "grad_norm": 0.6124121507463299, + "learning_rate": 2.015382626527964e-06, + "loss": 0.4556, + "step": 32369 + }, + { + "epoch": 0.712524763372221, + "grad_norm": 3.5915024575525227, + "learning_rate": 2.015096649809347e-06, + "loss": 0.2479, + "step": 32370 + }, + { + "epoch": 0.7125467752586396, + "grad_norm": 1.282854691579183, + "learning_rate": 2.014810688261342e-06, + "loss": 0.2944, + "step": 32371 + }, + { + "epoch": 0.7125687871450583, + "grad_norm": 1.7823151390584604, + "learning_rate": 2.014524741885398e-06, + "loss": 0.3038, + "step": 32372 + }, + { + "epoch": 0.712590799031477, + "grad_norm": 1.1968824858247762, + "learning_rate": 2.014238810682972e-06, + "loss": 0.2752, + "step": 32373 + }, + { + "epoch": 0.7126128109178956, + "grad_norm": 1.8266626882416543, + "learning_rate": 2.0139528946555172e-06, + "loss": 0.2891, + "step": 32374 + }, + { + "epoch": 0.7126348228043143, + "grad_norm": 1.7040009864473467, + "learning_rate": 2.0136669938044834e-06, + "loss": 0.4261, + "step": 32375 + }, + { + "epoch": 0.712656834690733, + "grad_norm": 1.6409363254374596, + "learning_rate": 2.013381108131328e-06, + "loss": 0.2895, + "step": 32376 + }, + { + "epoch": 0.7126788465771516, + "grad_norm": 2.257554688285762, + "learning_rate": 2.0130952376375e-06, + "loss": 0.3103, + "step": 32377 + }, + { + "epoch": 0.7127008584635703, + "grad_norm": 5.771640592849443, + "learning_rate": 2.0128093823244542e-06, + "loss": 0.3155, + "step": 32378 + }, + { + "epoch": 0.712722870349989, + "grad_norm": 1.6464462267291993, + "learning_rate": 2.0125235421936456e-06, + "loss": 0.2828, + "step": 32379 + }, + { + "epoch": 0.7127448822364076, + "grad_norm": 1.3877478548544004, + "learning_rate": 2.012237717246523e-06, + "loss": 0.3676, + "step": 32380 + }, + { + "epoch": 0.7127668941228263, + "grad_norm": 4.845181111163801, + "learning_rate": 2.0119519074845423e-06, + "loss": 0.347, + "step": 32381 + }, + { + "epoch": 0.7127889060092449, + "grad_norm": 1.9257581719176613, + "learning_rate": 2.0116661129091552e-06, + "loss": 0.2782, + "step": 32382 + }, + { + "epoch": 0.7128109178956636, + "grad_norm": 1.7542541530040512, + "learning_rate": 2.0113803335218096e-06, + "loss": 0.333, + "step": 32383 + }, + { + "epoch": 0.7128329297820823, + "grad_norm": 1.9143668802098193, + "learning_rate": 2.0110945693239657e-06, + "loss": 0.3558, + "step": 32384 + }, + { + "epoch": 0.7128549416685009, + "grad_norm": 1.5655560735938447, + "learning_rate": 2.010808820317073e-06, + "loss": 0.256, + "step": 32385 + }, + { + "epoch": 0.7128769535549196, + "grad_norm": 0.6052484665862887, + "learning_rate": 2.010523086502581e-06, + "loss": 0.4597, + "step": 32386 + }, + { + "epoch": 0.7128989654413384, + "grad_norm": 0.6284652212035934, + "learning_rate": 2.0102373678819455e-06, + "loss": 0.4573, + "step": 32387 + }, + { + "epoch": 0.7129209773277569, + "grad_norm": 0.6265597984212579, + "learning_rate": 2.0099516644566154e-06, + "loss": 0.4338, + "step": 32388 + }, + { + "epoch": 0.7129429892141756, + "grad_norm": 1.8894910690737714, + "learning_rate": 2.009665976228045e-06, + "loss": 0.2782, + "step": 32389 + }, + { + "epoch": 0.7129650011005944, + "grad_norm": 2.1561995470825708, + "learning_rate": 2.0093803031976873e-06, + "loss": 0.2678, + "step": 32390 + }, + { + "epoch": 0.712987012987013, + "grad_norm": 2.2941128302498055, + "learning_rate": 2.009094645366991e-06, + "loss": 0.2815, + "step": 32391 + }, + { + "epoch": 0.7130090248734317, + "grad_norm": 1.7156653266996063, + "learning_rate": 2.008809002737411e-06, + "loss": 0.3339, + "step": 32392 + }, + { + "epoch": 0.7130310367598504, + "grad_norm": 0.6032749091795849, + "learning_rate": 2.0085233753103966e-06, + "loss": 0.4442, + "step": 32393 + }, + { + "epoch": 0.713053048646269, + "grad_norm": 1.8420482604267847, + "learning_rate": 2.0082377630874006e-06, + "loss": 0.3521, + "step": 32394 + }, + { + "epoch": 0.7130750605326877, + "grad_norm": 1.7983071657333376, + "learning_rate": 2.007952166069876e-06, + "loss": 0.3566, + "step": 32395 + }, + { + "epoch": 0.7130970724191064, + "grad_norm": 1.6415474846498053, + "learning_rate": 2.007666584259272e-06, + "loss": 0.2855, + "step": 32396 + }, + { + "epoch": 0.713119084305525, + "grad_norm": 0.6017744169655384, + "learning_rate": 2.007381017657042e-06, + "loss": 0.4573, + "step": 32397 + }, + { + "epoch": 0.7131410961919437, + "grad_norm": 2.4095902582107995, + "learning_rate": 2.0070954662646348e-06, + "loss": 0.3644, + "step": 32398 + }, + { + "epoch": 0.7131631080783624, + "grad_norm": 0.6344998026501419, + "learning_rate": 2.0068099300835037e-06, + "loss": 0.4588, + "step": 32399 + }, + { + "epoch": 0.713185119964781, + "grad_norm": 1.3924100216656985, + "learning_rate": 2.006524409115101e-06, + "loss": 0.2562, + "step": 32400 + }, + { + "epoch": 0.7132071318511997, + "grad_norm": 2.7450106283723734, + "learning_rate": 2.0062389033608747e-06, + "loss": 0.2832, + "step": 32401 + }, + { + "epoch": 0.7132291437376184, + "grad_norm": 1.7829111427870001, + "learning_rate": 2.005953412822279e-06, + "loss": 0.3404, + "step": 32402 + }, + { + "epoch": 0.713251155624037, + "grad_norm": 1.5334293284015685, + "learning_rate": 2.0056679375007622e-06, + "loss": 0.3356, + "step": 32403 + }, + { + "epoch": 0.7132731675104557, + "grad_norm": 1.8978938929714828, + "learning_rate": 2.0053824773977757e-06, + "loss": 0.3219, + "step": 32404 + }, + { + "epoch": 0.7132951793968743, + "grad_norm": 0.5949596946943598, + "learning_rate": 2.005097032514774e-06, + "loss": 0.4059, + "step": 32405 + }, + { + "epoch": 0.713317191283293, + "grad_norm": 1.1874369113242453, + "learning_rate": 2.004811602853202e-06, + "loss": 0.2329, + "step": 32406 + }, + { + "epoch": 0.7133392031697117, + "grad_norm": 0.6010438538890871, + "learning_rate": 2.004526188414516e-06, + "loss": 0.4352, + "step": 32407 + }, + { + "epoch": 0.7133612150561303, + "grad_norm": 1.4250591844711407, + "learning_rate": 2.0042407892001618e-06, + "loss": 0.3049, + "step": 32408 + }, + { + "epoch": 0.713383226942549, + "grad_norm": 2.22468583495233, + "learning_rate": 2.0039554052115923e-06, + "loss": 0.3217, + "step": 32409 + }, + { + "epoch": 0.7134052388289677, + "grad_norm": 1.683520873191761, + "learning_rate": 2.003670036450259e-06, + "loss": 0.2435, + "step": 32410 + }, + { + "epoch": 0.7134272507153863, + "grad_norm": 1.3311381131746816, + "learning_rate": 2.0033846829176097e-06, + "loss": 0.2394, + "step": 32411 + }, + { + "epoch": 0.713449262601805, + "grad_norm": 1.5327413273754102, + "learning_rate": 2.003099344615097e-06, + "loss": 0.302, + "step": 32412 + }, + { + "epoch": 0.7134712744882237, + "grad_norm": 1.7466079677704567, + "learning_rate": 2.002814021544171e-06, + "loss": 0.3697, + "step": 32413 + }, + { + "epoch": 0.7134932863746423, + "grad_norm": 1.513446022618694, + "learning_rate": 2.0025287137062766e-06, + "loss": 0.2962, + "step": 32414 + }, + { + "epoch": 0.713515298261061, + "grad_norm": 0.6411774371827806, + "learning_rate": 2.0022434211028725e-06, + "loss": 0.4621, + "step": 32415 + }, + { + "epoch": 0.7135373101474797, + "grad_norm": 1.6443195165587068, + "learning_rate": 2.001958143735402e-06, + "loss": 0.344, + "step": 32416 + }, + { + "epoch": 0.7135593220338983, + "grad_norm": 0.6208351861628529, + "learning_rate": 2.0016728816053194e-06, + "loss": 0.4661, + "step": 32417 + }, + { + "epoch": 0.713581333920317, + "grad_norm": 1.441593807700608, + "learning_rate": 2.001387634714072e-06, + "loss": 0.3452, + "step": 32418 + }, + { + "epoch": 0.7136033458067357, + "grad_norm": 1.8876613450380493, + "learning_rate": 2.001102403063107e-06, + "loss": 0.2409, + "step": 32419 + }, + { + "epoch": 0.7136253576931543, + "grad_norm": 0.6354468875460538, + "learning_rate": 2.0008171866538805e-06, + "loss": 0.4805, + "step": 32420 + }, + { + "epoch": 0.713647369579573, + "grad_norm": 1.1775402845448868, + "learning_rate": 2.0005319854878376e-06, + "loss": 0.2785, + "step": 32421 + }, + { + "epoch": 0.7136693814659917, + "grad_norm": 2.9764144540418687, + "learning_rate": 2.00024679956643e-06, + "loss": 0.3167, + "step": 32422 + }, + { + "epoch": 0.7136913933524103, + "grad_norm": 1.7951056032150339, + "learning_rate": 1.9999616288911062e-06, + "loss": 0.3207, + "step": 32423 + }, + { + "epoch": 0.713713405238829, + "grad_norm": 2.130091600296569, + "learning_rate": 1.999676473463314e-06, + "loss": 0.238, + "step": 32424 + }, + { + "epoch": 0.7137354171252477, + "grad_norm": 1.600081403694588, + "learning_rate": 1.999391333284504e-06, + "loss": 0.2484, + "step": 32425 + }, + { + "epoch": 0.7137574290116663, + "grad_norm": 0.5989873934613437, + "learning_rate": 1.999106208356127e-06, + "loss": 0.4165, + "step": 32426 + }, + { + "epoch": 0.713779440898085, + "grad_norm": 1.9896494887153122, + "learning_rate": 1.998821098679629e-06, + "loss": 0.389, + "step": 32427 + }, + { + "epoch": 0.7138014527845037, + "grad_norm": 4.160058971228294, + "learning_rate": 1.9985360042564627e-06, + "loss": 0.2841, + "step": 32428 + }, + { + "epoch": 0.7138234646709223, + "grad_norm": 0.6496890920558622, + "learning_rate": 1.9982509250880734e-06, + "loss": 0.5, + "step": 32429 + }, + { + "epoch": 0.713845476557341, + "grad_norm": 1.5178491972387644, + "learning_rate": 1.9979658611759117e-06, + "loss": 0.2988, + "step": 32430 + }, + { + "epoch": 0.7138674884437596, + "grad_norm": 2.368183209644214, + "learning_rate": 1.997680812521428e-06, + "loss": 0.2825, + "step": 32431 + }, + { + "epoch": 0.7138895003301783, + "grad_norm": 2.1184386768278807, + "learning_rate": 1.997395779126068e-06, + "loss": 0.2201, + "step": 32432 + }, + { + "epoch": 0.713911512216597, + "grad_norm": 2.2980856034387176, + "learning_rate": 1.997110760991283e-06, + "loss": 0.2935, + "step": 32433 + }, + { + "epoch": 0.7139335241030156, + "grad_norm": 1.7648567050278698, + "learning_rate": 1.9968257581185195e-06, + "loss": 0.2976, + "step": 32434 + }, + { + "epoch": 0.7139555359894343, + "grad_norm": 1.6681027419887122, + "learning_rate": 1.9965407705092266e-06, + "loss": 0.3117, + "step": 32435 + }, + { + "epoch": 0.713977547875853, + "grad_norm": 2.2228973939456522, + "learning_rate": 1.996255798164855e-06, + "loss": 0.3292, + "step": 32436 + }, + { + "epoch": 0.7139995597622716, + "grad_norm": 1.9087489891119358, + "learning_rate": 1.995970841086849e-06, + "loss": 0.3193, + "step": 32437 + }, + { + "epoch": 0.7140215716486903, + "grad_norm": 1.9762309843002284, + "learning_rate": 1.995685899276661e-06, + "loss": 0.3314, + "step": 32438 + }, + { + "epoch": 0.714043583535109, + "grad_norm": 1.5809280478196313, + "learning_rate": 1.9954009727357355e-06, + "loss": 0.2969, + "step": 32439 + }, + { + "epoch": 0.7140655954215276, + "grad_norm": 1.401141333305433, + "learning_rate": 1.9951160614655224e-06, + "loss": 0.318, + "step": 32440 + }, + { + "epoch": 0.7140876073079463, + "grad_norm": 1.5021273822090573, + "learning_rate": 1.9948311654674714e-06, + "loss": 0.2619, + "step": 32441 + }, + { + "epoch": 0.714109619194365, + "grad_norm": 0.6137101363145727, + "learning_rate": 1.994546284743027e-06, + "loss": 0.4406, + "step": 32442 + }, + { + "epoch": 0.7141316310807836, + "grad_norm": 1.623425233626975, + "learning_rate": 1.9942614192936405e-06, + "loss": 0.301, + "step": 32443 + }, + { + "epoch": 0.7141536429672023, + "grad_norm": 1.6028142074564322, + "learning_rate": 1.9939765691207563e-06, + "loss": 0.2424, + "step": 32444 + }, + { + "epoch": 0.714175654853621, + "grad_norm": 1.428083344013412, + "learning_rate": 1.993691734225824e-06, + "loss": 0.2804, + "step": 32445 + }, + { + "epoch": 0.7141976667400396, + "grad_norm": 1.1603452783166481, + "learning_rate": 1.993406914610293e-06, + "loss": 0.299, + "step": 32446 + }, + { + "epoch": 0.7142196786264583, + "grad_norm": 1.3065012272970067, + "learning_rate": 1.993122110275607e-06, + "loss": 0.3575, + "step": 32447 + }, + { + "epoch": 0.714241690512877, + "grad_norm": 0.6239623472912137, + "learning_rate": 1.9928373212232177e-06, + "loss": 0.4756, + "step": 32448 + }, + { + "epoch": 0.7142637023992956, + "grad_norm": 0.5982603299479439, + "learning_rate": 1.9925525474545697e-06, + "loss": 0.4456, + "step": 32449 + }, + { + "epoch": 0.7142857142857143, + "grad_norm": 1.3881922434201697, + "learning_rate": 1.9922677889711085e-06, + "loss": 0.247, + "step": 32450 + }, + { + "epoch": 0.714307726172133, + "grad_norm": 0.6136975578253269, + "learning_rate": 1.9919830457742867e-06, + "loss": 0.4389, + "step": 32451 + }, + { + "epoch": 0.7143297380585516, + "grad_norm": 1.5851089374858844, + "learning_rate": 1.991698317865547e-06, + "loss": 0.3693, + "step": 32452 + }, + { + "epoch": 0.7143517499449703, + "grad_norm": 1.4162967066449315, + "learning_rate": 1.9914136052463403e-06, + "loss": 0.2946, + "step": 32453 + }, + { + "epoch": 0.7143737618313889, + "grad_norm": 1.3290600576096054, + "learning_rate": 1.9911289079181107e-06, + "loss": 0.1929, + "step": 32454 + }, + { + "epoch": 0.7143957737178076, + "grad_norm": 0.6153128870272564, + "learning_rate": 1.990844225882303e-06, + "loss": 0.4815, + "step": 32455 + }, + { + "epoch": 0.7144177856042263, + "grad_norm": 1.581291983991387, + "learning_rate": 1.9905595591403707e-06, + "loss": 0.2979, + "step": 32456 + }, + { + "epoch": 0.7144397974906449, + "grad_norm": 5.737728470278037, + "learning_rate": 1.990274907693755e-06, + "loss": 0.2533, + "step": 32457 + }, + { + "epoch": 0.7144618093770636, + "grad_norm": 1.889965078835017, + "learning_rate": 1.989990271543906e-06, + "loss": 0.2681, + "step": 32458 + }, + { + "epoch": 0.7144838212634823, + "grad_norm": 1.3934409996419523, + "learning_rate": 1.9897056506922696e-06, + "loss": 0.2853, + "step": 32459 + }, + { + "epoch": 0.7145058331499009, + "grad_norm": 1.4591157676258593, + "learning_rate": 1.9894210451402902e-06, + "loss": 0.2485, + "step": 32460 + }, + { + "epoch": 0.7145278450363196, + "grad_norm": 2.274877539379584, + "learning_rate": 1.9891364548894147e-06, + "loss": 0.2565, + "step": 32461 + }, + { + "epoch": 0.7145498569227383, + "grad_norm": 1.3086895342635536, + "learning_rate": 1.9888518799410915e-06, + "loss": 0.2423, + "step": 32462 + }, + { + "epoch": 0.7145718688091569, + "grad_norm": 0.5835548543107766, + "learning_rate": 1.9885673202967677e-06, + "loss": 0.4452, + "step": 32463 + }, + { + "epoch": 0.7145938806955756, + "grad_norm": 1.8096326537833107, + "learning_rate": 1.9882827759578875e-06, + "loss": 0.4094, + "step": 32464 + }, + { + "epoch": 0.7146158925819943, + "grad_norm": 2.2287120134390888, + "learning_rate": 1.9879982469258964e-06, + "loss": 0.3805, + "step": 32465 + }, + { + "epoch": 0.7146379044684129, + "grad_norm": 1.4897726616306468, + "learning_rate": 1.987713733202241e-06, + "loss": 0.2265, + "step": 32466 + }, + { + "epoch": 0.7146599163548316, + "grad_norm": 3.074817688052011, + "learning_rate": 1.98742923478837e-06, + "loss": 0.3128, + "step": 32467 + }, + { + "epoch": 0.7146819282412503, + "grad_norm": 1.1825760538440229, + "learning_rate": 1.9871447516857255e-06, + "loss": 0.2949, + "step": 32468 + }, + { + "epoch": 0.7147039401276689, + "grad_norm": 1.3190863895269147, + "learning_rate": 1.986860283895757e-06, + "loss": 0.2791, + "step": 32469 + }, + { + "epoch": 0.7147259520140876, + "grad_norm": 1.6996919381693893, + "learning_rate": 1.9865758314199063e-06, + "loss": 0.2639, + "step": 32470 + }, + { + "epoch": 0.7147479639005063, + "grad_norm": 1.6369125316549333, + "learning_rate": 1.9862913942596214e-06, + "loss": 0.2555, + "step": 32471 + }, + { + "epoch": 0.7147699757869249, + "grad_norm": 18.39969014482604, + "learning_rate": 1.9860069724163493e-06, + "loss": 0.3332, + "step": 32472 + }, + { + "epoch": 0.7147919876733436, + "grad_norm": 1.819861574365774, + "learning_rate": 1.985722565891533e-06, + "loss": 0.3195, + "step": 32473 + }, + { + "epoch": 0.7148139995597623, + "grad_norm": 1.3220273076563016, + "learning_rate": 1.9854381746866204e-06, + "loss": 0.2546, + "step": 32474 + }, + { + "epoch": 0.7148360114461809, + "grad_norm": 1.9628687921216847, + "learning_rate": 1.9851537988030533e-06, + "loss": 0.2935, + "step": 32475 + }, + { + "epoch": 0.7148580233325996, + "grad_norm": 0.6154359441200241, + "learning_rate": 1.98486943824228e-06, + "loss": 0.4367, + "step": 32476 + }, + { + "epoch": 0.7148800352190182, + "grad_norm": 1.6221586324096047, + "learning_rate": 1.9845850930057463e-06, + "loss": 0.2735, + "step": 32477 + }, + { + "epoch": 0.7149020471054369, + "grad_norm": 1.558725329345767, + "learning_rate": 1.9843007630948945e-06, + "loss": 0.3013, + "step": 32478 + }, + { + "epoch": 0.7149240589918556, + "grad_norm": 2.307881093934736, + "learning_rate": 1.984016448511173e-06, + "loss": 0.2862, + "step": 32479 + }, + { + "epoch": 0.7149460708782742, + "grad_norm": 2.8685720817227587, + "learning_rate": 1.983732149256023e-06, + "loss": 0.3584, + "step": 32480 + }, + { + "epoch": 0.7149680827646929, + "grad_norm": 2.08187183360725, + "learning_rate": 1.983447865330892e-06, + "loss": 0.2868, + "step": 32481 + }, + { + "epoch": 0.7149900946511116, + "grad_norm": 0.6051706506339798, + "learning_rate": 1.9831635967372257e-06, + "loss": 0.4419, + "step": 32482 + }, + { + "epoch": 0.7150121065375302, + "grad_norm": 1.5277917207513056, + "learning_rate": 1.9828793434764664e-06, + "loss": 0.2489, + "step": 32483 + }, + { + "epoch": 0.7150341184239489, + "grad_norm": 1.6928532960104181, + "learning_rate": 1.9825951055500613e-06, + "loss": 0.2898, + "step": 32484 + }, + { + "epoch": 0.7150561303103676, + "grad_norm": 1.5738893407241876, + "learning_rate": 1.9823108829594532e-06, + "loss": 0.2239, + "step": 32485 + }, + { + "epoch": 0.7150781421967862, + "grad_norm": 2.1770598353573263, + "learning_rate": 1.9820266757060847e-06, + "loss": 0.4042, + "step": 32486 + }, + { + "epoch": 0.7151001540832049, + "grad_norm": 1.775678047761952, + "learning_rate": 1.981742483791405e-06, + "loss": 0.2412, + "step": 32487 + }, + { + "epoch": 0.7151221659696236, + "grad_norm": 1.6560653090837547, + "learning_rate": 1.9814583072168553e-06, + "loss": 0.2584, + "step": 32488 + }, + { + "epoch": 0.7151441778560422, + "grad_norm": 1.3578854317999964, + "learning_rate": 1.9811741459838824e-06, + "loss": 0.3722, + "step": 32489 + }, + { + "epoch": 0.7151661897424609, + "grad_norm": 0.6248059172004463, + "learning_rate": 1.9808900000939284e-06, + "loss": 0.4393, + "step": 32490 + }, + { + "epoch": 0.7151882016288796, + "grad_norm": 1.38722950542538, + "learning_rate": 1.9806058695484354e-06, + "loss": 0.3237, + "step": 32491 + }, + { + "epoch": 0.7152102135152982, + "grad_norm": 1.7875147823605246, + "learning_rate": 1.980321754348853e-06, + "loss": 0.3792, + "step": 32492 + }, + { + "epoch": 0.7152322254017169, + "grad_norm": 1.6697072573807596, + "learning_rate": 1.9800376544966206e-06, + "loss": 0.3374, + "step": 32493 + }, + { + "epoch": 0.7152542372881356, + "grad_norm": 1.543150318223968, + "learning_rate": 1.979753569993185e-06, + "loss": 0.2765, + "step": 32494 + }, + { + "epoch": 0.7152762491745542, + "grad_norm": 1.691289839966553, + "learning_rate": 1.9794695008399896e-06, + "loss": 0.3186, + "step": 32495 + }, + { + "epoch": 0.7152982610609729, + "grad_norm": 0.6121202967269155, + "learning_rate": 1.979185447038475e-06, + "loss": 0.4461, + "step": 32496 + }, + { + "epoch": 0.7153202729473916, + "grad_norm": 1.3648733667181572, + "learning_rate": 1.9789014085900875e-06, + "loss": 0.2774, + "step": 32497 + }, + { + "epoch": 0.7153422848338102, + "grad_norm": 0.6105487209165895, + "learning_rate": 1.978617385496271e-06, + "loss": 0.4569, + "step": 32498 + }, + { + "epoch": 0.7153642967202289, + "grad_norm": 0.6011766796086283, + "learning_rate": 1.9783333777584698e-06, + "loss": 0.4458, + "step": 32499 + }, + { + "epoch": 0.7153863086066475, + "grad_norm": 1.4775685730515034, + "learning_rate": 1.9780493853781264e-06, + "loss": 0.3715, + "step": 32500 + }, + { + "epoch": 0.7154083204930662, + "grad_norm": 1.6805722499140605, + "learning_rate": 1.9777654083566816e-06, + "loss": 0.2343, + "step": 32501 + }, + { + "epoch": 0.715430332379485, + "grad_norm": 1.7500735272828536, + "learning_rate": 1.977481446695581e-06, + "loss": 0.2905, + "step": 32502 + }, + { + "epoch": 0.7154523442659035, + "grad_norm": 2.704385212794732, + "learning_rate": 1.977197500396268e-06, + "loss": 0.2095, + "step": 32503 + }, + { + "epoch": 0.7154743561523222, + "grad_norm": 2.752801210094976, + "learning_rate": 1.9769135694601876e-06, + "loss": 0.3574, + "step": 32504 + }, + { + "epoch": 0.715496368038741, + "grad_norm": 1.3903297572739846, + "learning_rate": 1.9766296538887804e-06, + "loss": 0.2866, + "step": 32505 + }, + { + "epoch": 0.7155183799251595, + "grad_norm": 1.520905789130354, + "learning_rate": 1.9763457536834877e-06, + "loss": 0.191, + "step": 32506 + }, + { + "epoch": 0.7155403918115782, + "grad_norm": 1.556779076182136, + "learning_rate": 1.976061868845755e-06, + "loss": 0.2991, + "step": 32507 + }, + { + "epoch": 0.715562403697997, + "grad_norm": 1.8675152804787087, + "learning_rate": 1.975777999377026e-06, + "loss": 0.2556, + "step": 32508 + }, + { + "epoch": 0.7155844155844155, + "grad_norm": 0.600706579300499, + "learning_rate": 1.97549414527874e-06, + "loss": 0.4308, + "step": 32509 + }, + { + "epoch": 0.7156064274708342, + "grad_norm": 1.287383366714964, + "learning_rate": 1.9752103065523443e-06, + "loss": 0.2497, + "step": 32510 + }, + { + "epoch": 0.715628439357253, + "grad_norm": 2.012788103145975, + "learning_rate": 1.9749264831992763e-06, + "loss": 0.1846, + "step": 32511 + }, + { + "epoch": 0.7156504512436715, + "grad_norm": 2.1209998060362083, + "learning_rate": 1.974642675220982e-06, + "loss": 0.3108, + "step": 32512 + }, + { + "epoch": 0.7156724631300903, + "grad_norm": 1.3828081889858248, + "learning_rate": 1.9743588826189043e-06, + "loss": 0.3616, + "step": 32513 + }, + { + "epoch": 0.715694475016509, + "grad_norm": 1.610053712899993, + "learning_rate": 1.9740751053944824e-06, + "loss": 0.2135, + "step": 32514 + }, + { + "epoch": 0.7157164869029276, + "grad_norm": 1.6303659906198638, + "learning_rate": 1.9737913435491625e-06, + "loss": 0.2405, + "step": 32515 + }, + { + "epoch": 0.7157384987893463, + "grad_norm": 1.9416414583815667, + "learning_rate": 1.9735075970843824e-06, + "loss": 0.331, + "step": 32516 + }, + { + "epoch": 0.715760510675765, + "grad_norm": 1.48176052969596, + "learning_rate": 1.9732238660015873e-06, + "loss": 0.2864, + "step": 32517 + }, + { + "epoch": 0.7157825225621836, + "grad_norm": 1.637068660470372, + "learning_rate": 1.9729401503022196e-06, + "loss": 0.3753, + "step": 32518 + }, + { + "epoch": 0.7158045344486023, + "grad_norm": 1.4540217196128677, + "learning_rate": 1.972656449987719e-06, + "loss": 0.2697, + "step": 32519 + }, + { + "epoch": 0.715826546335021, + "grad_norm": 1.9789111983178953, + "learning_rate": 1.9723727650595296e-06, + "loss": 0.3204, + "step": 32520 + }, + { + "epoch": 0.7158485582214396, + "grad_norm": 2.197561452467734, + "learning_rate": 1.9720890955190907e-06, + "loss": 0.3269, + "step": 32521 + }, + { + "epoch": 0.7158705701078583, + "grad_norm": 4.090966513264637, + "learning_rate": 1.9718054413678457e-06, + "loss": 0.2862, + "step": 32522 + }, + { + "epoch": 0.7158925819942769, + "grad_norm": 1.846917787622921, + "learning_rate": 1.9715218026072373e-06, + "loss": 0.2866, + "step": 32523 + }, + { + "epoch": 0.7159145938806956, + "grad_norm": 2.3237505389642528, + "learning_rate": 1.971238179238705e-06, + "loss": 0.2872, + "step": 32524 + }, + { + "epoch": 0.7159366057671143, + "grad_norm": 1.696642444453298, + "learning_rate": 1.970954571263692e-06, + "loss": 0.2967, + "step": 32525 + }, + { + "epoch": 0.7159586176535329, + "grad_norm": 0.6208969419673888, + "learning_rate": 1.970670978683639e-06, + "loss": 0.4635, + "step": 32526 + }, + { + "epoch": 0.7159806295399516, + "grad_norm": 4.177999863396516, + "learning_rate": 1.970387401499986e-06, + "loss": 0.297, + "step": 32527 + }, + { + "epoch": 0.7160026414263703, + "grad_norm": 1.2167800994446862, + "learning_rate": 1.9701038397141746e-06, + "loss": 0.2741, + "step": 32528 + }, + { + "epoch": 0.7160246533127889, + "grad_norm": 0.5825905488705195, + "learning_rate": 1.9698202933276476e-06, + "loss": 0.4502, + "step": 32529 + }, + { + "epoch": 0.7160466651992076, + "grad_norm": 1.341309381745116, + "learning_rate": 1.969536762341847e-06, + "loss": 0.2627, + "step": 32530 + }, + { + "epoch": 0.7160686770856263, + "grad_norm": 1.7638411838439638, + "learning_rate": 1.969253246758212e-06, + "loss": 0.292, + "step": 32531 + }, + { + "epoch": 0.7160906889720449, + "grad_norm": 1.4254030192982214, + "learning_rate": 1.968969746578182e-06, + "loss": 0.2908, + "step": 32532 + }, + { + "epoch": 0.7161127008584636, + "grad_norm": 0.6182853889641746, + "learning_rate": 1.9686862618031998e-06, + "loss": 0.4722, + "step": 32533 + }, + { + "epoch": 0.7161347127448823, + "grad_norm": 5.837524723537334, + "learning_rate": 1.968402792434706e-06, + "loss": 0.2753, + "step": 32534 + }, + { + "epoch": 0.7161567246313009, + "grad_norm": 1.475140944679318, + "learning_rate": 1.968119338474143e-06, + "loss": 0.273, + "step": 32535 + }, + { + "epoch": 0.7161787365177196, + "grad_norm": 1.4784943626986375, + "learning_rate": 1.9678358999229503e-06, + "loss": 0.2668, + "step": 32536 + }, + { + "epoch": 0.7162007484041383, + "grad_norm": 1.6984051348691926, + "learning_rate": 1.9675524767825653e-06, + "loss": 0.2415, + "step": 32537 + }, + { + "epoch": 0.7162227602905569, + "grad_norm": 1.9692916903343034, + "learning_rate": 1.9672690690544317e-06, + "loss": 0.2785, + "step": 32538 + }, + { + "epoch": 0.7162447721769756, + "grad_norm": 1.478787385655611, + "learning_rate": 1.9669856767399893e-06, + "loss": 0.2688, + "step": 32539 + }, + { + "epoch": 0.7162667840633943, + "grad_norm": 1.4866710140678805, + "learning_rate": 1.96670229984068e-06, + "loss": 0.3022, + "step": 32540 + }, + { + "epoch": 0.7162887959498129, + "grad_norm": 8.088555831165278, + "learning_rate": 1.9664189383579424e-06, + "loss": 0.2634, + "step": 32541 + }, + { + "epoch": 0.7163108078362316, + "grad_norm": 1.5808059823281253, + "learning_rate": 1.9661355922932153e-06, + "loss": 0.3494, + "step": 32542 + }, + { + "epoch": 0.7163328197226503, + "grad_norm": 1.5980710718599465, + "learning_rate": 1.9658522616479396e-06, + "loss": 0.2984, + "step": 32543 + }, + { + "epoch": 0.7163548316090689, + "grad_norm": 1.8589134468556703, + "learning_rate": 1.965568946423559e-06, + "loss": 0.3705, + "step": 32544 + }, + { + "epoch": 0.7163768434954876, + "grad_norm": 1.7484595885039043, + "learning_rate": 1.9652856466215078e-06, + "loss": 0.2922, + "step": 32545 + }, + { + "epoch": 0.7163988553819062, + "grad_norm": 1.8578845262285248, + "learning_rate": 1.9650023622432303e-06, + "loss": 0.2573, + "step": 32546 + }, + { + "epoch": 0.7164208672683249, + "grad_norm": 1.67413707694877, + "learning_rate": 1.9647190932901627e-06, + "loss": 0.269, + "step": 32547 + }, + { + "epoch": 0.7164428791547436, + "grad_norm": 1.5798044939204312, + "learning_rate": 1.9644358397637466e-06, + "loss": 0.2888, + "step": 32548 + }, + { + "epoch": 0.7164648910411622, + "grad_norm": 1.7507631922596831, + "learning_rate": 1.964152601665423e-06, + "loss": 0.347, + "step": 32549 + }, + { + "epoch": 0.7164869029275809, + "grad_norm": 1.924712616761782, + "learning_rate": 1.963869378996628e-06, + "loss": 0.2703, + "step": 32550 + }, + { + "epoch": 0.7165089148139996, + "grad_norm": 1.4139051881752547, + "learning_rate": 1.963586171758805e-06, + "loss": 0.1975, + "step": 32551 + }, + { + "epoch": 0.7165309267004182, + "grad_norm": 1.555997391427321, + "learning_rate": 1.963302979953389e-06, + "loss": 0.3177, + "step": 32552 + }, + { + "epoch": 0.7165529385868369, + "grad_norm": 1.6332860245591658, + "learning_rate": 1.9630198035818227e-06, + "loss": 0.3588, + "step": 32553 + }, + { + "epoch": 0.7165749504732556, + "grad_norm": 1.471454161204949, + "learning_rate": 1.962736642645546e-06, + "loss": 0.2668, + "step": 32554 + }, + { + "epoch": 0.7165969623596742, + "grad_norm": 1.7130981174522772, + "learning_rate": 1.9624534971459934e-06, + "loss": 0.3326, + "step": 32555 + }, + { + "epoch": 0.7166189742460929, + "grad_norm": 1.650525096103995, + "learning_rate": 1.96217036708461e-06, + "loss": 0.339, + "step": 32556 + }, + { + "epoch": 0.7166409861325116, + "grad_norm": 3.659289668149816, + "learning_rate": 1.961887252462829e-06, + "loss": 0.2928, + "step": 32557 + }, + { + "epoch": 0.7166629980189302, + "grad_norm": 0.6388988113471176, + "learning_rate": 1.961604153282092e-06, + "loss": 0.4599, + "step": 32558 + }, + { + "epoch": 0.7166850099053489, + "grad_norm": 1.7822889177569585, + "learning_rate": 1.9613210695438407e-06, + "loss": 0.3041, + "step": 32559 + }, + { + "epoch": 0.7167070217917676, + "grad_norm": 2.112847187762771, + "learning_rate": 1.9610380012495084e-06, + "loss": 0.1965, + "step": 32560 + }, + { + "epoch": 0.7167290336781862, + "grad_norm": 2.0022590457599088, + "learning_rate": 1.9607549484005386e-06, + "loss": 0.278, + "step": 32561 + }, + { + "epoch": 0.7167510455646049, + "grad_norm": 1.16650333429221, + "learning_rate": 1.9604719109983648e-06, + "loss": 0.2886, + "step": 32562 + }, + { + "epoch": 0.7167730574510236, + "grad_norm": 2.903842331270836, + "learning_rate": 1.9601888890444317e-06, + "loss": 0.3567, + "step": 32563 + }, + { + "epoch": 0.7167950693374422, + "grad_norm": 1.5186570742903176, + "learning_rate": 1.9599058825401716e-06, + "loss": 0.2617, + "step": 32564 + }, + { + "epoch": 0.7168170812238609, + "grad_norm": 1.74366979122954, + "learning_rate": 1.959622891487026e-06, + "loss": 0.3629, + "step": 32565 + }, + { + "epoch": 0.7168390931102796, + "grad_norm": 1.6699985580331385, + "learning_rate": 1.9593399158864353e-06, + "loss": 0.2577, + "step": 32566 + }, + { + "epoch": 0.7168611049966982, + "grad_norm": 1.5452657310551874, + "learning_rate": 1.959056955739835e-06, + "loss": 0.2859, + "step": 32567 + }, + { + "epoch": 0.7168831168831169, + "grad_norm": 2.1580001066287235, + "learning_rate": 1.9587740110486612e-06, + "loss": 0.2965, + "step": 32568 + }, + { + "epoch": 0.7169051287695356, + "grad_norm": 1.4831517440741555, + "learning_rate": 1.958491081814355e-06, + "loss": 0.2556, + "step": 32569 + }, + { + "epoch": 0.7169271406559542, + "grad_norm": 2.5249114239809733, + "learning_rate": 1.958208168038353e-06, + "loss": 0.3689, + "step": 32570 + }, + { + "epoch": 0.7169491525423729, + "grad_norm": 1.5339486126168411, + "learning_rate": 1.9579252697220956e-06, + "loss": 0.3435, + "step": 32571 + }, + { + "epoch": 0.7169711644287915, + "grad_norm": 1.7590227174026394, + "learning_rate": 1.95764238686702e-06, + "loss": 0.3857, + "step": 32572 + }, + { + "epoch": 0.7169931763152102, + "grad_norm": 1.6550701288510787, + "learning_rate": 1.9573595194745594e-06, + "loss": 0.2947, + "step": 32573 + }, + { + "epoch": 0.7170151882016289, + "grad_norm": 1.3599486726599426, + "learning_rate": 1.9570766675461554e-06, + "loss": 0.277, + "step": 32574 + }, + { + "epoch": 0.7170372000880475, + "grad_norm": 2.0364544088551293, + "learning_rate": 1.9567938310832454e-06, + "loss": 0.4116, + "step": 32575 + }, + { + "epoch": 0.7170592119744662, + "grad_norm": 1.800441906092489, + "learning_rate": 1.9565110100872674e-06, + "loss": 0.195, + "step": 32576 + }, + { + "epoch": 0.7170812238608849, + "grad_norm": 1.4143993315242653, + "learning_rate": 1.956228204559658e-06, + "loss": 0.279, + "step": 32577 + }, + { + "epoch": 0.7171032357473035, + "grad_norm": 1.475397688609443, + "learning_rate": 1.955945414501853e-06, + "loss": 0.2434, + "step": 32578 + }, + { + "epoch": 0.7171252476337222, + "grad_norm": 1.8512034967497488, + "learning_rate": 1.955662639915291e-06, + "loss": 0.3213, + "step": 32579 + }, + { + "epoch": 0.7171472595201409, + "grad_norm": 3.4525842908974687, + "learning_rate": 1.9553798808014094e-06, + "loss": 0.2809, + "step": 32580 + }, + { + "epoch": 0.7171692714065595, + "grad_norm": 1.4641747656026345, + "learning_rate": 1.955097137161647e-06, + "loss": 0.2877, + "step": 32581 + }, + { + "epoch": 0.7171912832929782, + "grad_norm": 1.90189503246057, + "learning_rate": 1.954814408997439e-06, + "loss": 0.2894, + "step": 32582 + }, + { + "epoch": 0.7172132951793969, + "grad_norm": 1.7865553194316912, + "learning_rate": 1.9545316963102206e-06, + "loss": 0.3007, + "step": 32583 + }, + { + "epoch": 0.7172353070658155, + "grad_norm": 0.6222083792429908, + "learning_rate": 1.9542489991014303e-06, + "loss": 0.4626, + "step": 32584 + }, + { + "epoch": 0.7172573189522342, + "grad_norm": 1.6597612814226836, + "learning_rate": 1.9539663173725064e-06, + "loss": 0.3124, + "step": 32585 + }, + { + "epoch": 0.7172793308386529, + "grad_norm": 1.3196420490616196, + "learning_rate": 1.953683651124883e-06, + "loss": 0.3141, + "step": 32586 + }, + { + "epoch": 0.7173013427250715, + "grad_norm": 3.385406617866822, + "learning_rate": 1.9534010003599996e-06, + "loss": 0.3798, + "step": 32587 + }, + { + "epoch": 0.7173233546114902, + "grad_norm": 1.588058010893465, + "learning_rate": 1.9531183650792895e-06, + "loss": 0.3546, + "step": 32588 + }, + { + "epoch": 0.7173453664979089, + "grad_norm": 1.1951949479571853, + "learning_rate": 1.9528357452841907e-06, + "loss": 0.2723, + "step": 32589 + }, + { + "epoch": 0.7173673783843275, + "grad_norm": 1.6938372352522808, + "learning_rate": 1.9525531409761415e-06, + "loss": 0.3198, + "step": 32590 + }, + { + "epoch": 0.7173893902707462, + "grad_norm": 0.6345416366532325, + "learning_rate": 1.9522705521565747e-06, + "loss": 0.4654, + "step": 32591 + }, + { + "epoch": 0.7174114021571649, + "grad_norm": 0.6262505916561746, + "learning_rate": 1.9519879788269302e-06, + "loss": 0.4825, + "step": 32592 + }, + { + "epoch": 0.7174334140435835, + "grad_norm": 1.8648634195248823, + "learning_rate": 1.9517054209886404e-06, + "loss": 0.3115, + "step": 32593 + }, + { + "epoch": 0.7174554259300022, + "grad_norm": 1.5444853015724493, + "learning_rate": 1.9514228786431426e-06, + "loss": 0.296, + "step": 32594 + }, + { + "epoch": 0.7174774378164208, + "grad_norm": 0.6138060534031498, + "learning_rate": 1.951140351791876e-06, + "loss": 0.4599, + "step": 32595 + }, + { + "epoch": 0.7174994497028395, + "grad_norm": 2.039289135705156, + "learning_rate": 1.950857840436272e-06, + "loss": 0.2855, + "step": 32596 + }, + { + "epoch": 0.7175214615892582, + "grad_norm": 0.6216275552458636, + "learning_rate": 1.9505753445777704e-06, + "loss": 0.5014, + "step": 32597 + }, + { + "epoch": 0.7175434734756768, + "grad_norm": 1.8578867764082805, + "learning_rate": 1.950292864217803e-06, + "loss": 0.3261, + "step": 32598 + }, + { + "epoch": 0.7175654853620955, + "grad_norm": 0.6028997169250191, + "learning_rate": 1.9500103993578084e-06, + "loss": 0.4585, + "step": 32599 + }, + { + "epoch": 0.7175874972485142, + "grad_norm": 1.8831219115945237, + "learning_rate": 1.9497279499992204e-06, + "loss": 0.3199, + "step": 32600 + }, + { + "epoch": 0.7176095091349328, + "grad_norm": 1.3743490810395282, + "learning_rate": 1.9494455161434754e-06, + "loss": 0.2386, + "step": 32601 + }, + { + "epoch": 0.7176315210213515, + "grad_norm": 1.713502154766714, + "learning_rate": 1.9491630977920107e-06, + "loss": 0.2665, + "step": 32602 + }, + { + "epoch": 0.7176535329077702, + "grad_norm": 2.613021663034749, + "learning_rate": 1.948880694946258e-06, + "loss": 0.2793, + "step": 32603 + }, + { + "epoch": 0.7176755447941888, + "grad_norm": 3.504326475819058, + "learning_rate": 1.9485983076076557e-06, + "loss": 0.268, + "step": 32604 + }, + { + "epoch": 0.7176975566806075, + "grad_norm": 1.6160184207423676, + "learning_rate": 1.9483159357776367e-06, + "loss": 0.3303, + "step": 32605 + }, + { + "epoch": 0.7177195685670262, + "grad_norm": 0.620591398409444, + "learning_rate": 1.9480335794576364e-06, + "loss": 0.4721, + "step": 32606 + }, + { + "epoch": 0.7177415804534448, + "grad_norm": 2.057667183934727, + "learning_rate": 1.9477512386490934e-06, + "loss": 0.3258, + "step": 32607 + }, + { + "epoch": 0.7177635923398635, + "grad_norm": 1.75023152602027, + "learning_rate": 1.9474689133534395e-06, + "loss": 0.2303, + "step": 32608 + }, + { + "epoch": 0.7177856042262822, + "grad_norm": 0.6198787106370515, + "learning_rate": 1.947186603572108e-06, + "loss": 0.4751, + "step": 32609 + }, + { + "epoch": 0.7178076161127008, + "grad_norm": 0.6139624695415623, + "learning_rate": 1.9469043093065364e-06, + "loss": 0.4347, + "step": 32610 + }, + { + "epoch": 0.7178296279991195, + "grad_norm": 3.8199202537541166, + "learning_rate": 1.9466220305581585e-06, + "loss": 0.3827, + "step": 32611 + }, + { + "epoch": 0.7178516398855382, + "grad_norm": 1.2866174543576927, + "learning_rate": 1.9463397673284106e-06, + "loss": 0.2481, + "step": 32612 + }, + { + "epoch": 0.7178736517719568, + "grad_norm": 1.6256821256840452, + "learning_rate": 1.9460575196187263e-06, + "loss": 0.3008, + "step": 32613 + }, + { + "epoch": 0.7178956636583755, + "grad_norm": 1.6725070521941003, + "learning_rate": 1.9457752874305387e-06, + "loss": 0.3143, + "step": 32614 + }, + { + "epoch": 0.7179176755447942, + "grad_norm": 1.5385592162543098, + "learning_rate": 1.9454930707652823e-06, + "loss": 0.2613, + "step": 32615 + }, + { + "epoch": 0.7179396874312128, + "grad_norm": 0.6416892183332086, + "learning_rate": 1.9452108696243933e-06, + "loss": 0.4754, + "step": 32616 + }, + { + "epoch": 0.7179616993176315, + "grad_norm": 1.2630469731455536, + "learning_rate": 1.9449286840093064e-06, + "loss": 0.2845, + "step": 32617 + }, + { + "epoch": 0.7179837112040501, + "grad_norm": 2.515173738394448, + "learning_rate": 1.944646513921454e-06, + "loss": 0.2745, + "step": 32618 + }, + { + "epoch": 0.7180057230904688, + "grad_norm": 1.3424129022857414, + "learning_rate": 1.94436435936227e-06, + "loss": 0.3026, + "step": 32619 + }, + { + "epoch": 0.7180277349768875, + "grad_norm": 1.3661290465942424, + "learning_rate": 1.944082220333189e-06, + "loss": 0.2331, + "step": 32620 + }, + { + "epoch": 0.7180497468633061, + "grad_norm": 1.481062818876675, + "learning_rate": 1.9438000968356444e-06, + "loss": 0.289, + "step": 32621 + }, + { + "epoch": 0.7180717587497248, + "grad_norm": 1.6681089245423872, + "learning_rate": 1.943517988871073e-06, + "loss": 0.3581, + "step": 32622 + }, + { + "epoch": 0.7180937706361435, + "grad_norm": 1.4696303860224156, + "learning_rate": 1.9432358964409066e-06, + "loss": 0.2746, + "step": 32623 + }, + { + "epoch": 0.7181157825225621, + "grad_norm": 1.5098505511308415, + "learning_rate": 1.942953819546577e-06, + "loss": 0.2572, + "step": 32624 + }, + { + "epoch": 0.7181377944089808, + "grad_norm": 1.3980962970036934, + "learning_rate": 1.9426717581895193e-06, + "loss": 0.3241, + "step": 32625 + }, + { + "epoch": 0.7181598062953995, + "grad_norm": 1.5025595914776995, + "learning_rate": 1.942389712371169e-06, + "loss": 0.3111, + "step": 32626 + }, + { + "epoch": 0.7181818181818181, + "grad_norm": 1.6973128658401515, + "learning_rate": 1.942107682092957e-06, + "loss": 0.2562, + "step": 32627 + }, + { + "epoch": 0.7182038300682368, + "grad_norm": 0.6093120982495307, + "learning_rate": 1.9418256673563186e-06, + "loss": 0.461, + "step": 32628 + }, + { + "epoch": 0.7182258419546556, + "grad_norm": 1.59105241375365, + "learning_rate": 1.941543668162684e-06, + "loss": 0.2582, + "step": 32629 + }, + { + "epoch": 0.7182478538410741, + "grad_norm": 1.854274090892887, + "learning_rate": 1.9412616845134895e-06, + "loss": 0.2991, + "step": 32630 + }, + { + "epoch": 0.7182698657274929, + "grad_norm": 3.8862081768148564, + "learning_rate": 1.9409797164101683e-06, + "loss": 0.2796, + "step": 32631 + }, + { + "epoch": 0.7182918776139116, + "grad_norm": 0.618518365808247, + "learning_rate": 1.9406977638541514e-06, + "loss": 0.451, + "step": 32632 + }, + { + "epoch": 0.7183138895003301, + "grad_norm": 1.457627274628522, + "learning_rate": 1.9404158268468747e-06, + "loss": 0.3516, + "step": 32633 + }, + { + "epoch": 0.7183359013867489, + "grad_norm": 0.6153234768079362, + "learning_rate": 1.9401339053897672e-06, + "loss": 0.4692, + "step": 32634 + }, + { + "epoch": 0.7183579132731676, + "grad_norm": 1.3644123659099727, + "learning_rate": 1.9398519994842664e-06, + "loss": 0.2649, + "step": 32635 + }, + { + "epoch": 0.7183799251595862, + "grad_norm": 1.3960153351018372, + "learning_rate": 1.9395701091318004e-06, + "loss": 0.3254, + "step": 32636 + }, + { + "epoch": 0.7184019370460049, + "grad_norm": 2.747544648969272, + "learning_rate": 1.939288234333804e-06, + "loss": 0.254, + "step": 32637 + }, + { + "epoch": 0.7184239489324236, + "grad_norm": 5.356351856098199, + "learning_rate": 1.9390063750917116e-06, + "loss": 0.2703, + "step": 32638 + }, + { + "epoch": 0.7184459608188422, + "grad_norm": 1.7409918177215944, + "learning_rate": 1.9387245314069526e-06, + "loss": 0.2943, + "step": 32639 + }, + { + "epoch": 0.7184679727052609, + "grad_norm": 1.268376857005547, + "learning_rate": 1.938442703280962e-06, + "loss": 0.283, + "step": 32640 + }, + { + "epoch": 0.7184899845916795, + "grad_norm": 1.620364768087521, + "learning_rate": 1.93816089071517e-06, + "loss": 0.2742, + "step": 32641 + }, + { + "epoch": 0.7185119964780982, + "grad_norm": 1.4442911163648744, + "learning_rate": 1.9378790937110096e-06, + "loss": 0.2069, + "step": 32642 + }, + { + "epoch": 0.7185340083645169, + "grad_norm": 0.622225218803903, + "learning_rate": 1.9375973122699154e-06, + "loss": 0.4296, + "step": 32643 + }, + { + "epoch": 0.7185560202509355, + "grad_norm": 3.7872569556535574, + "learning_rate": 1.9373155463933176e-06, + "loss": 0.3124, + "step": 32644 + }, + { + "epoch": 0.7185780321373542, + "grad_norm": 1.9054834731429504, + "learning_rate": 1.9370337960826463e-06, + "loss": 0.2545, + "step": 32645 + }, + { + "epoch": 0.7186000440237729, + "grad_norm": 2.7749513207803878, + "learning_rate": 1.9367520613393353e-06, + "loss": 0.3116, + "step": 32646 + }, + { + "epoch": 0.7186220559101915, + "grad_norm": 1.4545407210256038, + "learning_rate": 1.9364703421648167e-06, + "loss": 0.2528, + "step": 32647 + }, + { + "epoch": 0.7186440677966102, + "grad_norm": 1.5282287369760135, + "learning_rate": 1.936188638560524e-06, + "loss": 0.2455, + "step": 32648 + }, + { + "epoch": 0.7186660796830289, + "grad_norm": 1.4002914414047727, + "learning_rate": 1.9359069505278866e-06, + "loss": 0.3345, + "step": 32649 + }, + { + "epoch": 0.7186880915694475, + "grad_norm": 1.619933004280677, + "learning_rate": 1.9356252780683354e-06, + "loss": 0.3705, + "step": 32650 + }, + { + "epoch": 0.7187101034558662, + "grad_norm": 2.121554124901369, + "learning_rate": 1.935343621183303e-06, + "loss": 0.2695, + "step": 32651 + }, + { + "epoch": 0.7187321153422849, + "grad_norm": 2.142382114291827, + "learning_rate": 1.9350619798742216e-06, + "loss": 0.3253, + "step": 32652 + }, + { + "epoch": 0.7187541272287035, + "grad_norm": 2.0853842168956658, + "learning_rate": 1.934780354142524e-06, + "loss": 0.3276, + "step": 32653 + }, + { + "epoch": 0.7187761391151222, + "grad_norm": 1.473754185315332, + "learning_rate": 1.9344987439896395e-06, + "loss": 0.3687, + "step": 32654 + }, + { + "epoch": 0.7187981510015409, + "grad_norm": 1.2288255751242478, + "learning_rate": 1.934217149416998e-06, + "loss": 0.2846, + "step": 32655 + }, + { + "epoch": 0.7188201628879595, + "grad_norm": 0.6351770492418426, + "learning_rate": 1.933935570426032e-06, + "loss": 0.4466, + "step": 32656 + }, + { + "epoch": 0.7188421747743782, + "grad_norm": 1.8038551970707803, + "learning_rate": 1.9336540070181734e-06, + "loss": 0.2665, + "step": 32657 + }, + { + "epoch": 0.7188641866607969, + "grad_norm": 1.4742019055299989, + "learning_rate": 1.9333724591948546e-06, + "loss": 0.2453, + "step": 32658 + }, + { + "epoch": 0.7188861985472155, + "grad_norm": 0.6311512191132101, + "learning_rate": 1.9330909269575043e-06, + "loss": 0.4348, + "step": 32659 + }, + { + "epoch": 0.7189082104336342, + "grad_norm": 2.3850354123619377, + "learning_rate": 1.9328094103075517e-06, + "loss": 0.3066, + "step": 32660 + }, + { + "epoch": 0.7189302223200529, + "grad_norm": 1.6211538002249735, + "learning_rate": 1.9325279092464307e-06, + "loss": 0.2689, + "step": 32661 + }, + { + "epoch": 0.7189522342064715, + "grad_norm": 1.4498180960256983, + "learning_rate": 1.93224642377557e-06, + "loss": 0.2958, + "step": 32662 + }, + { + "epoch": 0.7189742460928902, + "grad_norm": 1.8355632057149252, + "learning_rate": 1.9319649538964037e-06, + "loss": 0.3007, + "step": 32663 + }, + { + "epoch": 0.7189962579793088, + "grad_norm": 2.004474804888287, + "learning_rate": 1.9316834996103597e-06, + "loss": 0.2637, + "step": 32664 + }, + { + "epoch": 0.7190182698657275, + "grad_norm": 1.355360738033569, + "learning_rate": 1.9314020609188665e-06, + "loss": 0.281, + "step": 32665 + }, + { + "epoch": 0.7190402817521462, + "grad_norm": 1.8300262239447254, + "learning_rate": 1.931120637823357e-06, + "loss": 0.3006, + "step": 32666 + }, + { + "epoch": 0.7190622936385648, + "grad_norm": 0.6062330053270895, + "learning_rate": 1.930839230325263e-06, + "loss": 0.4682, + "step": 32667 + }, + { + "epoch": 0.7190843055249835, + "grad_norm": 2.354495602112737, + "learning_rate": 1.930557838426011e-06, + "loss": 0.2356, + "step": 32668 + }, + { + "epoch": 0.7191063174114022, + "grad_norm": 2.9658404434889176, + "learning_rate": 1.9302764621270354e-06, + "loss": 0.3211, + "step": 32669 + }, + { + "epoch": 0.7191283292978208, + "grad_norm": 1.4889740049233744, + "learning_rate": 1.9299951014297613e-06, + "loss": 0.2707, + "step": 32670 + }, + { + "epoch": 0.7191503411842395, + "grad_norm": 3.165859556401733, + "learning_rate": 1.929713756335624e-06, + "loss": 0.3049, + "step": 32671 + }, + { + "epoch": 0.7191723530706582, + "grad_norm": 1.4548231730406547, + "learning_rate": 1.929432426846048e-06, + "loss": 0.2435, + "step": 32672 + }, + { + "epoch": 0.7191943649570768, + "grad_norm": 2.057922529225326, + "learning_rate": 1.929151112962467e-06, + "loss": 0.3584, + "step": 32673 + }, + { + "epoch": 0.7192163768434955, + "grad_norm": 3.6455331868673237, + "learning_rate": 1.9288698146863104e-06, + "loss": 0.3005, + "step": 32674 + }, + { + "epoch": 0.7192383887299142, + "grad_norm": 2.1400443631864876, + "learning_rate": 1.928588532019006e-06, + "loss": 0.3713, + "step": 32675 + }, + { + "epoch": 0.7192604006163328, + "grad_norm": 0.6183466654938599, + "learning_rate": 1.9283072649619862e-06, + "loss": 0.4509, + "step": 32676 + }, + { + "epoch": 0.7192824125027515, + "grad_norm": 1.3985029315229973, + "learning_rate": 1.928026013516677e-06, + "loss": 0.2446, + "step": 32677 + }, + { + "epoch": 0.7193044243891702, + "grad_norm": 1.4685280020419698, + "learning_rate": 1.9277447776845097e-06, + "loss": 0.2554, + "step": 32678 + }, + { + "epoch": 0.7193264362755888, + "grad_norm": 1.891018958719857, + "learning_rate": 1.927463557466916e-06, + "loss": 0.3252, + "step": 32679 + }, + { + "epoch": 0.7193484481620075, + "grad_norm": 1.4861283040982203, + "learning_rate": 1.927182352865321e-06, + "loss": 0.2835, + "step": 32680 + }, + { + "epoch": 0.7193704600484262, + "grad_norm": 1.4570123824756307, + "learning_rate": 1.9269011638811576e-06, + "loss": 0.3619, + "step": 32681 + }, + { + "epoch": 0.7193924719348448, + "grad_norm": 1.3680819825474906, + "learning_rate": 1.9266199905158505e-06, + "loss": 0.3266, + "step": 32682 + }, + { + "epoch": 0.7194144838212635, + "grad_norm": 2.0758103450379544, + "learning_rate": 1.926338832770832e-06, + "loss": 0.3113, + "step": 32683 + }, + { + "epoch": 0.7194364957076822, + "grad_norm": 1.6656301780496177, + "learning_rate": 1.9260576906475327e-06, + "loss": 0.3561, + "step": 32684 + }, + { + "epoch": 0.7194585075941008, + "grad_norm": 1.4511542755675269, + "learning_rate": 1.9257765641473786e-06, + "loss": 0.3069, + "step": 32685 + }, + { + "epoch": 0.7194805194805195, + "grad_norm": 1.6927266959049334, + "learning_rate": 1.925495453271797e-06, + "loss": 0.3093, + "step": 32686 + }, + { + "epoch": 0.7195025313669381, + "grad_norm": 1.5603427116607789, + "learning_rate": 1.92521435802222e-06, + "loss": 0.2803, + "step": 32687 + }, + { + "epoch": 0.7195245432533568, + "grad_norm": 1.699142123412683, + "learning_rate": 1.924933278400074e-06, + "loss": 0.2927, + "step": 32688 + }, + { + "epoch": 0.7195465551397755, + "grad_norm": 1.6945936042630985, + "learning_rate": 1.924652214406791e-06, + "loss": 0.2824, + "step": 32689 + }, + { + "epoch": 0.7195685670261941, + "grad_norm": 2.0968726910660878, + "learning_rate": 1.9243711660437962e-06, + "loss": 0.238, + "step": 32690 + }, + { + "epoch": 0.7195905789126128, + "grad_norm": 1.107109384635115, + "learning_rate": 1.924090133312518e-06, + "loss": 0.2518, + "step": 32691 + }, + { + "epoch": 0.7196125907990315, + "grad_norm": 0.5988470880891985, + "learning_rate": 1.923809116214385e-06, + "loss": 0.452, + "step": 32692 + }, + { + "epoch": 0.7196346026854501, + "grad_norm": 1.4176761986603392, + "learning_rate": 1.9235281147508265e-06, + "loss": 0.271, + "step": 32693 + }, + { + "epoch": 0.7196566145718688, + "grad_norm": 1.4839925521281516, + "learning_rate": 1.923247128923272e-06, + "loss": 0.2968, + "step": 32694 + }, + { + "epoch": 0.7196786264582875, + "grad_norm": 1.5918686740192138, + "learning_rate": 1.9229661587331476e-06, + "loss": 0.4435, + "step": 32695 + }, + { + "epoch": 0.7197006383447061, + "grad_norm": 0.6385486822253053, + "learning_rate": 1.9226852041818804e-06, + "loss": 0.459, + "step": 32696 + }, + { + "epoch": 0.7197226502311248, + "grad_norm": 1.643920648158945, + "learning_rate": 1.922404265270899e-06, + "loss": 0.2357, + "step": 32697 + }, + { + "epoch": 0.7197446621175435, + "grad_norm": 2.3893114350934215, + "learning_rate": 1.922123342001632e-06, + "loss": 0.2668, + "step": 32698 + }, + { + "epoch": 0.7197666740039621, + "grad_norm": 0.610806697617451, + "learning_rate": 1.9218424343755087e-06, + "loss": 0.4486, + "step": 32699 + }, + { + "epoch": 0.7197886858903808, + "grad_norm": 1.5017880576850882, + "learning_rate": 1.921561542393955e-06, + "loss": 0.3121, + "step": 32700 + }, + { + "epoch": 0.7198106977767995, + "grad_norm": 1.5731219054958405, + "learning_rate": 1.9212806660583967e-06, + "loss": 0.2599, + "step": 32701 + }, + { + "epoch": 0.7198327096632181, + "grad_norm": 1.5739347226965, + "learning_rate": 1.9209998053702634e-06, + "loss": 0.2639, + "step": 32702 + }, + { + "epoch": 0.7198547215496368, + "grad_norm": 1.5178831787079083, + "learning_rate": 1.920718960330982e-06, + "loss": 0.2725, + "step": 32703 + }, + { + "epoch": 0.7198767334360555, + "grad_norm": 2.7662081702901595, + "learning_rate": 1.9204381309419823e-06, + "loss": 0.3137, + "step": 32704 + }, + { + "epoch": 0.7198987453224741, + "grad_norm": 1.2473381015557867, + "learning_rate": 1.9201573172046897e-06, + "loss": 0.2239, + "step": 32705 + }, + { + "epoch": 0.7199207572088928, + "grad_norm": 1.5956474670785514, + "learning_rate": 1.9198765191205295e-06, + "loss": 0.3113, + "step": 32706 + }, + { + "epoch": 0.7199427690953115, + "grad_norm": 1.2554124337018786, + "learning_rate": 1.919595736690932e-06, + "loss": 0.3307, + "step": 32707 + }, + { + "epoch": 0.7199647809817301, + "grad_norm": 1.3015796093981344, + "learning_rate": 1.919314969917321e-06, + "loss": 0.3338, + "step": 32708 + }, + { + "epoch": 0.7199867928681488, + "grad_norm": 1.166496004175279, + "learning_rate": 1.919034218801126e-06, + "loss": 0.2539, + "step": 32709 + }, + { + "epoch": 0.7200088047545675, + "grad_norm": 2.005079388769738, + "learning_rate": 1.9187534833437743e-06, + "loss": 0.2394, + "step": 32710 + }, + { + "epoch": 0.7200308166409861, + "grad_norm": 4.5189213497477985, + "learning_rate": 1.91847276354669e-06, + "loss": 0.2022, + "step": 32711 + }, + { + "epoch": 0.7200528285274048, + "grad_norm": 1.5356728822581849, + "learning_rate": 1.9181920594113033e-06, + "loss": 0.3348, + "step": 32712 + }, + { + "epoch": 0.7200748404138234, + "grad_norm": 1.664983058728242, + "learning_rate": 1.917911370939038e-06, + "loss": 0.3467, + "step": 32713 + }, + { + "epoch": 0.7200968523002421, + "grad_norm": 1.2934032144938252, + "learning_rate": 1.917630698131321e-06, + "loss": 0.2599, + "step": 32714 + }, + { + "epoch": 0.7201188641866608, + "grad_norm": 0.6143268684274549, + "learning_rate": 1.917350040989582e-06, + "loss": 0.458, + "step": 32715 + }, + { + "epoch": 0.7201408760730794, + "grad_norm": 1.6291359212416945, + "learning_rate": 1.9170693995152424e-06, + "loss": 0.2957, + "step": 32716 + }, + { + "epoch": 0.7201628879594981, + "grad_norm": 1.7480383162855648, + "learning_rate": 1.9167887737097334e-06, + "loss": 0.3128, + "step": 32717 + }, + { + "epoch": 0.7201848998459168, + "grad_norm": 2.040598993545556, + "learning_rate": 1.916508163574477e-06, + "loss": 0.3022, + "step": 32718 + }, + { + "epoch": 0.7202069117323354, + "grad_norm": 1.5452853545207725, + "learning_rate": 1.9162275691109015e-06, + "loss": 0.2187, + "step": 32719 + }, + { + "epoch": 0.7202289236187541, + "grad_norm": 1.3091861379116407, + "learning_rate": 1.915946990320435e-06, + "loss": 0.3222, + "step": 32720 + }, + { + "epoch": 0.7202509355051728, + "grad_norm": 2.473513588245061, + "learning_rate": 1.9156664272044996e-06, + "loss": 0.2518, + "step": 32721 + }, + { + "epoch": 0.7202729473915914, + "grad_norm": 2.146668057323534, + "learning_rate": 1.9153858797645243e-06, + "loss": 0.2219, + "step": 32722 + }, + { + "epoch": 0.7202949592780101, + "grad_norm": 1.7142049723794328, + "learning_rate": 1.915105348001932e-06, + "loss": 0.2337, + "step": 32723 + }, + { + "epoch": 0.7203169711644288, + "grad_norm": 1.6327556193365291, + "learning_rate": 1.9148248319181505e-06, + "loss": 0.3082, + "step": 32724 + }, + { + "epoch": 0.7203389830508474, + "grad_norm": 1.8180841329171782, + "learning_rate": 1.9145443315146073e-06, + "loss": 0.2438, + "step": 32725 + }, + { + "epoch": 0.7203609949372661, + "grad_norm": 2.2036523755059076, + "learning_rate": 1.9142638467927254e-06, + "loss": 0.2958, + "step": 32726 + }, + { + "epoch": 0.7203830068236848, + "grad_norm": 1.5039848240775138, + "learning_rate": 1.9139833777539296e-06, + "loss": 0.2749, + "step": 32727 + }, + { + "epoch": 0.7204050187101034, + "grad_norm": 1.54937714912804, + "learning_rate": 1.9137029243996462e-06, + "loss": 0.3721, + "step": 32728 + }, + { + "epoch": 0.7204270305965221, + "grad_norm": 2.330333105793558, + "learning_rate": 1.9134224867313018e-06, + "loss": 0.3154, + "step": 32729 + }, + { + "epoch": 0.7204490424829408, + "grad_norm": 1.6226329348947017, + "learning_rate": 1.9131420647503218e-06, + "loss": 0.3878, + "step": 32730 + }, + { + "epoch": 0.7204710543693594, + "grad_norm": 6.884930092195116, + "learning_rate": 1.9128616584581313e-06, + "loss": 0.4043, + "step": 32731 + }, + { + "epoch": 0.7204930662557781, + "grad_norm": 1.8064380299974054, + "learning_rate": 1.912581267856153e-06, + "loss": 0.4165, + "step": 32732 + }, + { + "epoch": 0.7205150781421968, + "grad_norm": 3.3597576388069874, + "learning_rate": 1.9123008929458132e-06, + "loss": 0.2919, + "step": 32733 + }, + { + "epoch": 0.7205370900286154, + "grad_norm": 1.722381505895602, + "learning_rate": 1.912020533728538e-06, + "loss": 0.3602, + "step": 32734 + }, + { + "epoch": 0.7205591019150341, + "grad_norm": 1.6047179337964104, + "learning_rate": 1.911740190205753e-06, + "loss": 0.1942, + "step": 32735 + }, + { + "epoch": 0.7205811138014527, + "grad_norm": 2.1838282912545375, + "learning_rate": 1.911459862378882e-06, + "loss": 0.3048, + "step": 32736 + }, + { + "epoch": 0.7206031256878714, + "grad_norm": 1.5789687888913562, + "learning_rate": 1.9111795502493476e-06, + "loss": 0.3663, + "step": 32737 + }, + { + "epoch": 0.7206251375742901, + "grad_norm": 1.7784575223418537, + "learning_rate": 1.9108992538185785e-06, + "loss": 0.3061, + "step": 32738 + }, + { + "epoch": 0.7206471494607087, + "grad_norm": 2.2338151077885096, + "learning_rate": 1.9106189730879932e-06, + "loss": 0.3431, + "step": 32739 + }, + { + "epoch": 0.7206691613471274, + "grad_norm": 0.6566159944488842, + "learning_rate": 1.910338708059024e-06, + "loss": 0.4964, + "step": 32740 + }, + { + "epoch": 0.7206911732335461, + "grad_norm": 1.3975935896014458, + "learning_rate": 1.9100584587330914e-06, + "loss": 0.27, + "step": 32741 + }, + { + "epoch": 0.7207131851199647, + "grad_norm": 1.714808541969215, + "learning_rate": 1.909778225111618e-06, + "loss": 0.265, + "step": 32742 + }, + { + "epoch": 0.7207351970063834, + "grad_norm": 1.722758164628766, + "learning_rate": 1.9094980071960316e-06, + "loss": 0.2499, + "step": 32743 + }, + { + "epoch": 0.7207572088928021, + "grad_norm": 0.6065875854422152, + "learning_rate": 1.9092178049877536e-06, + "loss": 0.4356, + "step": 32744 + }, + { + "epoch": 0.7207792207792207, + "grad_norm": 2.3990208059751743, + "learning_rate": 1.9089376184882085e-06, + "loss": 0.3036, + "step": 32745 + }, + { + "epoch": 0.7208012326656394, + "grad_norm": 1.412801956863856, + "learning_rate": 1.9086574476988233e-06, + "loss": 0.3096, + "step": 32746 + }, + { + "epoch": 0.7208232445520582, + "grad_norm": 1.7311474548804997, + "learning_rate": 1.9083772926210175e-06, + "loss": 0.2972, + "step": 32747 + }, + { + "epoch": 0.7208452564384767, + "grad_norm": 1.4593915188188333, + "learning_rate": 1.908097153256219e-06, + "loss": 0.2791, + "step": 32748 + }, + { + "epoch": 0.7208672683248954, + "grad_norm": 2.1157375314957627, + "learning_rate": 1.9078170296058492e-06, + "loss": 0.3381, + "step": 32749 + }, + { + "epoch": 0.7208892802113142, + "grad_norm": 1.6517137156367823, + "learning_rate": 1.907536921671331e-06, + "loss": 0.344, + "step": 32750 + }, + { + "epoch": 0.7209112920977327, + "grad_norm": 1.5766802891146563, + "learning_rate": 1.9072568294540915e-06, + "loss": 0.2143, + "step": 32751 + }, + { + "epoch": 0.7209333039841515, + "grad_norm": 3.236713839793202, + "learning_rate": 1.906976752955551e-06, + "loss": 0.2467, + "step": 32752 + }, + { + "epoch": 0.7209553158705702, + "grad_norm": 1.8527541193908768, + "learning_rate": 1.9066966921771357e-06, + "loss": 0.4076, + "step": 32753 + }, + { + "epoch": 0.7209773277569888, + "grad_norm": 0.6370915453658921, + "learning_rate": 1.9064166471202656e-06, + "loss": 0.4359, + "step": 32754 + }, + { + "epoch": 0.7209993396434075, + "grad_norm": 0.608182512898054, + "learning_rate": 1.9061366177863655e-06, + "loss": 0.4424, + "step": 32755 + }, + { + "epoch": 0.7210213515298262, + "grad_norm": 0.6076410420060474, + "learning_rate": 1.905856604176861e-06, + "loss": 0.4681, + "step": 32756 + }, + { + "epoch": 0.7210433634162448, + "grad_norm": 2.8708545114133015, + "learning_rate": 1.905576606293172e-06, + "loss": 0.3915, + "step": 32757 + }, + { + "epoch": 0.7210653753026635, + "grad_norm": 4.936974451383228, + "learning_rate": 1.9052966241367243e-06, + "loss": 0.2751, + "step": 32758 + }, + { + "epoch": 0.721087387189082, + "grad_norm": 2.2940329759443716, + "learning_rate": 1.9050166577089375e-06, + "loss": 0.3484, + "step": 32759 + }, + { + "epoch": 0.7211093990755008, + "grad_norm": 1.3195073927927274, + "learning_rate": 1.9047367070112365e-06, + "loss": 0.2392, + "step": 32760 + }, + { + "epoch": 0.7211314109619195, + "grad_norm": 3.85685973984256, + "learning_rate": 1.9044567720450457e-06, + "loss": 0.321, + "step": 32761 + }, + { + "epoch": 0.721153422848338, + "grad_norm": 2.652533518191625, + "learning_rate": 1.904176852811785e-06, + "loss": 0.336, + "step": 32762 + }, + { + "epoch": 0.7211754347347568, + "grad_norm": 1.5665724090102777, + "learning_rate": 1.9038969493128795e-06, + "loss": 0.269, + "step": 32763 + }, + { + "epoch": 0.7211974466211755, + "grad_norm": 2.4009479494894026, + "learning_rate": 1.903617061549749e-06, + "loss": 0.3612, + "step": 32764 + }, + { + "epoch": 0.7212194585075941, + "grad_norm": 4.306743544070224, + "learning_rate": 1.9033371895238178e-06, + "loss": 0.3394, + "step": 32765 + }, + { + "epoch": 0.7212414703940128, + "grad_norm": 1.709524620691805, + "learning_rate": 1.90305733323651e-06, + "loss": 0.3412, + "step": 32766 + }, + { + "epoch": 0.7212634822804315, + "grad_norm": 1.9821786047690386, + "learning_rate": 1.902777492689246e-06, + "loss": 0.2782, + "step": 32767 + }, + { + "epoch": 0.7212854941668501, + "grad_norm": 1.5824551631524624, + "learning_rate": 1.9024976678834462e-06, + "loss": 0.2555, + "step": 32768 + }, + { + "epoch": 0.7213075060532688, + "grad_norm": 1.1547794774339397, + "learning_rate": 1.9022178588205347e-06, + "loss": 0.3279, + "step": 32769 + }, + { + "epoch": 0.7213295179396875, + "grad_norm": 1.5705150395488772, + "learning_rate": 1.9019380655019342e-06, + "loss": 0.336, + "step": 32770 + }, + { + "epoch": 0.7213515298261061, + "grad_norm": 1.4219856044192443, + "learning_rate": 1.901658287929068e-06, + "loss": 0.2913, + "step": 32771 + }, + { + "epoch": 0.7213735417125248, + "grad_norm": 2.460886469578308, + "learning_rate": 1.9013785261033558e-06, + "loss": 0.3346, + "step": 32772 + }, + { + "epoch": 0.7213955535989435, + "grad_norm": 0.6353875600694714, + "learning_rate": 1.9010987800262181e-06, + "loss": 0.4565, + "step": 32773 + }, + { + "epoch": 0.7214175654853621, + "grad_norm": 1.607494129354433, + "learning_rate": 1.900819049699081e-06, + "loss": 0.2737, + "step": 32774 + }, + { + "epoch": 0.7214395773717808, + "grad_norm": 1.8914135691472183, + "learning_rate": 1.9005393351233597e-06, + "loss": 0.3602, + "step": 32775 + }, + { + "epoch": 0.7214615892581995, + "grad_norm": 1.968614143321867, + "learning_rate": 1.9002596363004832e-06, + "loss": 0.2587, + "step": 32776 + }, + { + "epoch": 0.7214836011446181, + "grad_norm": 2.1175507487811642, + "learning_rate": 1.8999799532318702e-06, + "loss": 0.3191, + "step": 32777 + }, + { + "epoch": 0.7215056130310368, + "grad_norm": 3.523078606524268, + "learning_rate": 1.89970028591894e-06, + "loss": 0.3206, + "step": 32778 + }, + { + "epoch": 0.7215276249174555, + "grad_norm": 1.7540412493068098, + "learning_rate": 1.8994206343631172e-06, + "loss": 0.2196, + "step": 32779 + }, + { + "epoch": 0.7215496368038741, + "grad_norm": 0.5985534695443734, + "learning_rate": 1.8991409985658182e-06, + "loss": 0.4619, + "step": 32780 + }, + { + "epoch": 0.7215716486902928, + "grad_norm": 1.385921351334628, + "learning_rate": 1.8988613785284715e-06, + "loss": 0.2586, + "step": 32781 + }, + { + "epoch": 0.7215936605767114, + "grad_norm": 1.621455058807814, + "learning_rate": 1.8985817742524942e-06, + "loss": 0.2655, + "step": 32782 + }, + { + "epoch": 0.7216156724631301, + "grad_norm": 1.4271600707056766, + "learning_rate": 1.898302185739306e-06, + "loss": 0.2819, + "step": 32783 + }, + { + "epoch": 0.7216376843495488, + "grad_norm": 1.5801201449427933, + "learning_rate": 1.8980226129903305e-06, + "loss": 0.2775, + "step": 32784 + }, + { + "epoch": 0.7216596962359674, + "grad_norm": 1.5772925228666537, + "learning_rate": 1.8977430560069865e-06, + "loss": 0.2808, + "step": 32785 + }, + { + "epoch": 0.7216817081223861, + "grad_norm": 1.4599619240255166, + "learning_rate": 1.8974635147906956e-06, + "loss": 0.3044, + "step": 32786 + }, + { + "epoch": 0.7217037200088048, + "grad_norm": 1.447812676686069, + "learning_rate": 1.8971839893428811e-06, + "loss": 0.3586, + "step": 32787 + }, + { + "epoch": 0.7217257318952234, + "grad_norm": 1.205580025950228, + "learning_rate": 1.8969044796649595e-06, + "loss": 0.2536, + "step": 32788 + }, + { + "epoch": 0.7217477437816421, + "grad_norm": 2.1447477579466088, + "learning_rate": 1.8966249857583552e-06, + "loss": 0.2761, + "step": 32789 + }, + { + "epoch": 0.7217697556680608, + "grad_norm": 0.6038483275281845, + "learning_rate": 1.8963455076244847e-06, + "loss": 0.4423, + "step": 32790 + }, + { + "epoch": 0.7217917675544794, + "grad_norm": 1.7735238774488293, + "learning_rate": 1.8960660452647712e-06, + "loss": 0.2496, + "step": 32791 + }, + { + "epoch": 0.7218137794408981, + "grad_norm": 1.2085187471786492, + "learning_rate": 1.8957865986806357e-06, + "loss": 0.2717, + "step": 32792 + }, + { + "epoch": 0.7218357913273168, + "grad_norm": 0.61899154742668, + "learning_rate": 1.8955071678734955e-06, + "loss": 0.473, + "step": 32793 + }, + { + "epoch": 0.7218578032137354, + "grad_norm": 1.3999781908033633, + "learning_rate": 1.8952277528447743e-06, + "loss": 0.2456, + "step": 32794 + }, + { + "epoch": 0.7218798151001541, + "grad_norm": 1.4072787628033994, + "learning_rate": 1.894948353595889e-06, + "loss": 0.2855, + "step": 32795 + }, + { + "epoch": 0.7219018269865728, + "grad_norm": 1.7204186599751863, + "learning_rate": 1.894668970128261e-06, + "loss": 0.3626, + "step": 32796 + }, + { + "epoch": 0.7219238388729914, + "grad_norm": 2.2971748776101775, + "learning_rate": 1.8943896024433116e-06, + "loss": 0.3339, + "step": 32797 + }, + { + "epoch": 0.7219458507594101, + "grad_norm": 1.4131062034503208, + "learning_rate": 1.894110250542458e-06, + "loss": 0.2353, + "step": 32798 + }, + { + "epoch": 0.7219678626458288, + "grad_norm": 2.435250161815264, + "learning_rate": 1.8938309144271232e-06, + "loss": 0.3733, + "step": 32799 + }, + { + "epoch": 0.7219898745322474, + "grad_norm": 1.3567398852260792, + "learning_rate": 1.893551594098723e-06, + "loss": 0.2526, + "step": 32800 + }, + { + "epoch": 0.7220118864186661, + "grad_norm": 1.6579229319638709, + "learning_rate": 1.8932722895586797e-06, + "loss": 0.2545, + "step": 32801 + }, + { + "epoch": 0.7220338983050848, + "grad_norm": 1.3644498740179458, + "learning_rate": 1.8929930008084135e-06, + "loss": 0.2834, + "step": 32802 + }, + { + "epoch": 0.7220559101915034, + "grad_norm": 1.7119141662562347, + "learning_rate": 1.8927137278493414e-06, + "loss": 0.2659, + "step": 32803 + }, + { + "epoch": 0.7220779220779221, + "grad_norm": 0.6127471931523181, + "learning_rate": 1.8924344706828851e-06, + "loss": 0.4357, + "step": 32804 + }, + { + "epoch": 0.7220999339643407, + "grad_norm": 1.3874318166999215, + "learning_rate": 1.892155229310461e-06, + "loss": 0.3232, + "step": 32805 + }, + { + "epoch": 0.7221219458507594, + "grad_norm": 1.954905702110295, + "learning_rate": 1.891876003733491e-06, + "loss": 0.2677, + "step": 32806 + }, + { + "epoch": 0.7221439577371781, + "grad_norm": 0.597937419068766, + "learning_rate": 1.8915967939533946e-06, + "loss": 0.449, + "step": 32807 + }, + { + "epoch": 0.7221659696235967, + "grad_norm": 1.2822463434125282, + "learning_rate": 1.89131759997159e-06, + "loss": 0.2451, + "step": 32808 + }, + { + "epoch": 0.7221879815100154, + "grad_norm": 2.0967711243697447, + "learning_rate": 1.8910384217894933e-06, + "loss": 0.241, + "step": 32809 + }, + { + "epoch": 0.7222099933964341, + "grad_norm": 1.395587285299504, + "learning_rate": 1.890759259408529e-06, + "loss": 0.2632, + "step": 32810 + }, + { + "epoch": 0.7222320052828527, + "grad_norm": 2.1201999063824957, + "learning_rate": 1.8904801128301082e-06, + "loss": 0.285, + "step": 32811 + }, + { + "epoch": 0.7222540171692714, + "grad_norm": 1.4649871100351681, + "learning_rate": 1.8902009820556588e-06, + "loss": 0.2761, + "step": 32812 + }, + { + "epoch": 0.7222760290556901, + "grad_norm": 1.7785985135764661, + "learning_rate": 1.8899218670865942e-06, + "loss": 0.2433, + "step": 32813 + }, + { + "epoch": 0.7222980409421087, + "grad_norm": 1.3570461521586543, + "learning_rate": 1.8896427679243329e-06, + "loss": 0.2198, + "step": 32814 + }, + { + "epoch": 0.7223200528285274, + "grad_norm": 0.6243721928822542, + "learning_rate": 1.889363684570295e-06, + "loss": 0.4684, + "step": 32815 + }, + { + "epoch": 0.7223420647149461, + "grad_norm": 1.6403738859510475, + "learning_rate": 1.8890846170258957e-06, + "loss": 0.3406, + "step": 32816 + }, + { + "epoch": 0.7223640766013647, + "grad_norm": 1.840437259482289, + "learning_rate": 1.888805565292559e-06, + "loss": 0.349, + "step": 32817 + }, + { + "epoch": 0.7223860884877834, + "grad_norm": 1.6416931493025477, + "learning_rate": 1.8885265293717003e-06, + "loss": 0.2268, + "step": 32818 + }, + { + "epoch": 0.7224081003742021, + "grad_norm": 2.030362809170185, + "learning_rate": 1.8882475092647351e-06, + "loss": 0.2856, + "step": 32819 + }, + { + "epoch": 0.7224301122606207, + "grad_norm": 1.725261849458429, + "learning_rate": 1.8879685049730861e-06, + "loss": 0.3299, + "step": 32820 + }, + { + "epoch": 0.7224521241470394, + "grad_norm": 4.046170573989849, + "learning_rate": 1.8876895164981678e-06, + "loss": 0.2369, + "step": 32821 + }, + { + "epoch": 0.7224741360334581, + "grad_norm": 1.6295485186787615, + "learning_rate": 1.8874105438413993e-06, + "loss": 0.3183, + "step": 32822 + }, + { + "epoch": 0.7224961479198767, + "grad_norm": 1.3165072181151545, + "learning_rate": 1.8871315870042e-06, + "loss": 0.2688, + "step": 32823 + }, + { + "epoch": 0.7225181598062954, + "grad_norm": 2.013965672183331, + "learning_rate": 1.8868526459879849e-06, + "loss": 0.2357, + "step": 32824 + }, + { + "epoch": 0.7225401716927141, + "grad_norm": 1.2471544573238538, + "learning_rate": 1.8865737207941747e-06, + "loss": 0.2076, + "step": 32825 + }, + { + "epoch": 0.7225621835791327, + "grad_norm": 0.6215239532701382, + "learning_rate": 1.8862948114241841e-06, + "loss": 0.446, + "step": 32826 + }, + { + "epoch": 0.7225841954655514, + "grad_norm": 2.0435265793141655, + "learning_rate": 1.8860159178794318e-06, + "loss": 0.2639, + "step": 32827 + }, + { + "epoch": 0.72260620735197, + "grad_norm": 2.0134027303232176, + "learning_rate": 1.8857370401613372e-06, + "loss": 0.3026, + "step": 32828 + }, + { + "epoch": 0.7226282192383887, + "grad_norm": 1.2314950942083602, + "learning_rate": 1.8854581782713143e-06, + "loss": 0.2743, + "step": 32829 + }, + { + "epoch": 0.7226502311248074, + "grad_norm": 0.6118375563724833, + "learning_rate": 1.885179332210784e-06, + "loss": 0.4511, + "step": 32830 + }, + { + "epoch": 0.722672243011226, + "grad_norm": 1.473506712039296, + "learning_rate": 1.8849005019811595e-06, + "loss": 0.3448, + "step": 32831 + }, + { + "epoch": 0.7226942548976447, + "grad_norm": 1.266397849456762, + "learning_rate": 1.8846216875838602e-06, + "loss": 0.2933, + "step": 32832 + }, + { + "epoch": 0.7227162667840634, + "grad_norm": 1.307037038761188, + "learning_rate": 1.8843428890203048e-06, + "loss": 0.2142, + "step": 32833 + }, + { + "epoch": 0.722738278670482, + "grad_norm": 2.426851139173158, + "learning_rate": 1.8840641062919067e-06, + "loss": 0.2822, + "step": 32834 + }, + { + "epoch": 0.7227602905569007, + "grad_norm": 2.6136140558914924, + "learning_rate": 1.883785339400086e-06, + "loss": 0.3001, + "step": 32835 + }, + { + "epoch": 0.7227823024433194, + "grad_norm": 1.5548766666454414, + "learning_rate": 1.8835065883462567e-06, + "loss": 0.2956, + "step": 32836 + }, + { + "epoch": 0.722804314329738, + "grad_norm": 0.6275398315078228, + "learning_rate": 1.8832278531318366e-06, + "loss": 0.4532, + "step": 32837 + }, + { + "epoch": 0.7228263262161567, + "grad_norm": 2.0508168397437023, + "learning_rate": 1.8829491337582445e-06, + "loss": 0.2759, + "step": 32838 + }, + { + "epoch": 0.7228483381025754, + "grad_norm": 1.9548918630676029, + "learning_rate": 1.882670430226894e-06, + "loss": 0.2999, + "step": 32839 + }, + { + "epoch": 0.722870349988994, + "grad_norm": 1.8181334784717305, + "learning_rate": 1.8823917425392035e-06, + "loss": 0.2618, + "step": 32840 + }, + { + "epoch": 0.7228923618754127, + "grad_norm": 2.760930435878864, + "learning_rate": 1.8821130706965873e-06, + "loss": 0.3318, + "step": 32841 + }, + { + "epoch": 0.7229143737618314, + "grad_norm": 1.3065269613873403, + "learning_rate": 1.8818344147004637e-06, + "loss": 0.2938, + "step": 32842 + }, + { + "epoch": 0.72293638564825, + "grad_norm": 1.4889725123258304, + "learning_rate": 1.8815557745522494e-06, + "loss": 0.2946, + "step": 32843 + }, + { + "epoch": 0.7229583975346687, + "grad_norm": 2.0189528854621757, + "learning_rate": 1.8812771502533594e-06, + "loss": 0.2674, + "step": 32844 + }, + { + "epoch": 0.7229804094210874, + "grad_norm": 2.0926612854447297, + "learning_rate": 1.8809985418052085e-06, + "loss": 0.2112, + "step": 32845 + }, + { + "epoch": 0.723002421307506, + "grad_norm": 1.981236061139507, + "learning_rate": 1.880719949209216e-06, + "loss": 0.3112, + "step": 32846 + }, + { + "epoch": 0.7230244331939247, + "grad_norm": 1.7654286228184244, + "learning_rate": 1.880441372466792e-06, + "loss": 0.3738, + "step": 32847 + }, + { + "epoch": 0.7230464450803434, + "grad_norm": 1.6134948600451224, + "learning_rate": 1.8801628115793603e-06, + "loss": 0.3298, + "step": 32848 + }, + { + "epoch": 0.723068456966762, + "grad_norm": 2.4390570793335824, + "learning_rate": 1.8798842665483324e-06, + "loss": 0.2758, + "step": 32849 + }, + { + "epoch": 0.7230904688531807, + "grad_norm": 1.204482253240424, + "learning_rate": 1.8796057373751224e-06, + "loss": 0.2483, + "step": 32850 + }, + { + "epoch": 0.7231124807395994, + "grad_norm": 1.6408916068694184, + "learning_rate": 1.8793272240611499e-06, + "loss": 0.2791, + "step": 32851 + }, + { + "epoch": 0.723134492626018, + "grad_norm": 1.479115664753447, + "learning_rate": 1.879048726607826e-06, + "loss": 0.2605, + "step": 32852 + }, + { + "epoch": 0.7231565045124367, + "grad_norm": 1.4393045191084657, + "learning_rate": 1.878770245016569e-06, + "loss": 0.2571, + "step": 32853 + }, + { + "epoch": 0.7231785163988553, + "grad_norm": 0.6309358624348009, + "learning_rate": 1.8784917792887953e-06, + "loss": 0.4572, + "step": 32854 + }, + { + "epoch": 0.723200528285274, + "grad_norm": 1.328223471620264, + "learning_rate": 1.878213329425917e-06, + "loss": 0.292, + "step": 32855 + }, + { + "epoch": 0.7232225401716927, + "grad_norm": 1.7903121006130363, + "learning_rate": 1.8779348954293524e-06, + "loss": 0.3077, + "step": 32856 + }, + { + "epoch": 0.7232445520581113, + "grad_norm": 1.4365878721331464, + "learning_rate": 1.8776564773005136e-06, + "loss": 0.2364, + "step": 32857 + }, + { + "epoch": 0.72326656394453, + "grad_norm": 1.512087244425643, + "learning_rate": 1.8773780750408166e-06, + "loss": 0.3123, + "step": 32858 + }, + { + "epoch": 0.7232885758309487, + "grad_norm": 0.6056141188187825, + "learning_rate": 1.8770996886516795e-06, + "loss": 0.4302, + "step": 32859 + }, + { + "epoch": 0.7233105877173673, + "grad_norm": 1.9860290319327885, + "learning_rate": 1.8768213181345124e-06, + "loss": 0.4246, + "step": 32860 + }, + { + "epoch": 0.723332599603786, + "grad_norm": 1.508385499400513, + "learning_rate": 1.8765429634907345e-06, + "loss": 0.2713, + "step": 32861 + }, + { + "epoch": 0.7233546114902047, + "grad_norm": 0.6132638365061212, + "learning_rate": 1.8762646247217559e-06, + "loss": 0.4687, + "step": 32862 + }, + { + "epoch": 0.7233766233766233, + "grad_norm": 2.4567389285367875, + "learning_rate": 1.8759863018289942e-06, + "loss": 0.3459, + "step": 32863 + }, + { + "epoch": 0.723398635263042, + "grad_norm": 1.2757080024121306, + "learning_rate": 1.8757079948138651e-06, + "loss": 0.2324, + "step": 32864 + }, + { + "epoch": 0.7234206471494607, + "grad_norm": 0.6150136347429417, + "learning_rate": 1.8754297036777802e-06, + "loss": 0.4432, + "step": 32865 + }, + { + "epoch": 0.7234426590358793, + "grad_norm": 2.3265978704973556, + "learning_rate": 1.8751514284221562e-06, + "loss": 0.2685, + "step": 32866 + }, + { + "epoch": 0.723464670922298, + "grad_norm": 1.273477957197216, + "learning_rate": 1.8748731690484052e-06, + "loss": 0.241, + "step": 32867 + }, + { + "epoch": 0.7234866828087168, + "grad_norm": 1.773236433550119, + "learning_rate": 1.874594925557942e-06, + "loss": 0.2957, + "step": 32868 + }, + { + "epoch": 0.7235086946951353, + "grad_norm": 0.626889444135581, + "learning_rate": 1.874316697952183e-06, + "loss": 0.4554, + "step": 32869 + }, + { + "epoch": 0.723530706581554, + "grad_norm": 1.43280539175126, + "learning_rate": 1.8740384862325394e-06, + "loss": 0.2529, + "step": 32870 + }, + { + "epoch": 0.7235527184679728, + "grad_norm": 0.6505613284461699, + "learning_rate": 1.8737602904004282e-06, + "loss": 0.4777, + "step": 32871 + }, + { + "epoch": 0.7235747303543913, + "grad_norm": 2.437524355883037, + "learning_rate": 1.8734821104572598e-06, + "loss": 0.2758, + "step": 32872 + }, + { + "epoch": 0.72359674224081, + "grad_norm": 1.7616550754424478, + "learning_rate": 1.8732039464044494e-06, + "loss": 0.3098, + "step": 32873 + }, + { + "epoch": 0.7236187541272288, + "grad_norm": 2.855341228485423, + "learning_rate": 1.8729257982434134e-06, + "loss": 0.2532, + "step": 32874 + }, + { + "epoch": 0.7236407660136474, + "grad_norm": 1.5088582685750542, + "learning_rate": 1.872647665975561e-06, + "loss": 0.255, + "step": 32875 + }, + { + "epoch": 0.7236627779000661, + "grad_norm": 1.511592067958778, + "learning_rate": 1.8723695496023097e-06, + "loss": 0.2796, + "step": 32876 + }, + { + "epoch": 0.7236847897864847, + "grad_norm": 5.898263911979776, + "learning_rate": 1.8720914491250702e-06, + "loss": 0.2656, + "step": 32877 + }, + { + "epoch": 0.7237068016729034, + "grad_norm": 1.2775613612984833, + "learning_rate": 1.8718133645452563e-06, + "loss": 0.2373, + "step": 32878 + }, + { + "epoch": 0.7237288135593221, + "grad_norm": 1.235827151062608, + "learning_rate": 1.871535295864284e-06, + "loss": 0.2282, + "step": 32879 + }, + { + "epoch": 0.7237508254457407, + "grad_norm": 1.6256238350466587, + "learning_rate": 1.8712572430835634e-06, + "loss": 0.3277, + "step": 32880 + }, + { + "epoch": 0.7237728373321594, + "grad_norm": 2.188242387481919, + "learning_rate": 1.87097920620451e-06, + "loss": 0.2577, + "step": 32881 + }, + { + "epoch": 0.7237948492185781, + "grad_norm": 1.3894567972531657, + "learning_rate": 1.8707011852285368e-06, + "loss": 0.321, + "step": 32882 + }, + { + "epoch": 0.7238168611049967, + "grad_norm": 1.6965761500581031, + "learning_rate": 1.8704231801570517e-06, + "loss": 0.2802, + "step": 32883 + }, + { + "epoch": 0.7238388729914154, + "grad_norm": 2.121835662780147, + "learning_rate": 1.8701451909914753e-06, + "loss": 0.3004, + "step": 32884 + }, + { + "epoch": 0.7238608848778341, + "grad_norm": 1.9664994685526769, + "learning_rate": 1.8698672177332172e-06, + "loss": 0.4213, + "step": 32885 + }, + { + "epoch": 0.7238828967642527, + "grad_norm": 1.3781644300489317, + "learning_rate": 1.869589260383688e-06, + "loss": 0.357, + "step": 32886 + }, + { + "epoch": 0.7239049086506714, + "grad_norm": 3.7726653546866973, + "learning_rate": 1.8693113189443041e-06, + "loss": 0.2932, + "step": 32887 + }, + { + "epoch": 0.7239269205370901, + "grad_norm": 1.694406034738059, + "learning_rate": 1.869033393416475e-06, + "loss": 0.2354, + "step": 32888 + }, + { + "epoch": 0.7239489324235087, + "grad_norm": 1.2292252170705738, + "learning_rate": 1.8687554838016148e-06, + "loss": 0.183, + "step": 32889 + }, + { + "epoch": 0.7239709443099274, + "grad_norm": 1.6234410366224974, + "learning_rate": 1.868477590101137e-06, + "loss": 0.449, + "step": 32890 + }, + { + "epoch": 0.7239929561963461, + "grad_norm": 2.3133111769039814, + "learning_rate": 1.8681997123164513e-06, + "loss": 0.3428, + "step": 32891 + }, + { + "epoch": 0.7240149680827647, + "grad_norm": 1.8250155980498473, + "learning_rate": 1.8679218504489732e-06, + "loss": 0.2451, + "step": 32892 + }, + { + "epoch": 0.7240369799691834, + "grad_norm": 1.7391874437481483, + "learning_rate": 1.8676440045001114e-06, + "loss": 0.3597, + "step": 32893 + }, + { + "epoch": 0.7240589918556021, + "grad_norm": 1.7287828717309859, + "learning_rate": 1.86736617447128e-06, + "loss": 0.2629, + "step": 32894 + }, + { + "epoch": 0.7240810037420207, + "grad_norm": 1.9991284408110621, + "learning_rate": 1.8670883603638923e-06, + "loss": 0.3307, + "step": 32895 + }, + { + "epoch": 0.7241030156284394, + "grad_norm": 1.5555849666739192, + "learning_rate": 1.8668105621793576e-06, + "loss": 0.3301, + "step": 32896 + }, + { + "epoch": 0.7241250275148581, + "grad_norm": 1.3149387349792827, + "learning_rate": 1.8665327799190907e-06, + "loss": 0.3048, + "step": 32897 + }, + { + "epoch": 0.7241470394012767, + "grad_norm": 1.4736520822379464, + "learning_rate": 1.8662550135844997e-06, + "loss": 0.2494, + "step": 32898 + }, + { + "epoch": 0.7241690512876954, + "grad_norm": 2.4014763847793086, + "learning_rate": 1.8659772631769986e-06, + "loss": 0.4028, + "step": 32899 + }, + { + "epoch": 0.724191063174114, + "grad_norm": 1.8111603607815525, + "learning_rate": 1.8656995286980006e-06, + "loss": 0.19, + "step": 32900 + }, + { + "epoch": 0.7242130750605327, + "grad_norm": 1.7393586285972618, + "learning_rate": 1.8654218101489142e-06, + "loss": 0.2453, + "step": 32901 + }, + { + "epoch": 0.7242350869469514, + "grad_norm": 1.600310535452989, + "learning_rate": 1.8651441075311534e-06, + "loss": 0.2443, + "step": 32902 + }, + { + "epoch": 0.72425709883337, + "grad_norm": 2.242535928118377, + "learning_rate": 1.864866420846127e-06, + "loss": 0.3038, + "step": 32903 + }, + { + "epoch": 0.7242791107197887, + "grad_norm": 1.2731054027299165, + "learning_rate": 1.8645887500952475e-06, + "loss": 0.214, + "step": 32904 + }, + { + "epoch": 0.7243011226062074, + "grad_norm": 1.4012830737859894, + "learning_rate": 1.864311095279928e-06, + "loss": 0.2565, + "step": 32905 + }, + { + "epoch": 0.724323134492626, + "grad_norm": 1.725041056497898, + "learning_rate": 1.8640334564015767e-06, + "loss": 0.2944, + "step": 32906 + }, + { + "epoch": 0.7243451463790447, + "grad_norm": 0.6299683329484573, + "learning_rate": 1.863755833461608e-06, + "loss": 0.4649, + "step": 32907 + }, + { + "epoch": 0.7243671582654634, + "grad_norm": 1.483789022097973, + "learning_rate": 1.8634782264614286e-06, + "loss": 0.2629, + "step": 32908 + }, + { + "epoch": 0.724389170151882, + "grad_norm": 1.700028127963101, + "learning_rate": 1.8632006354024523e-06, + "loss": 0.3016, + "step": 32909 + }, + { + "epoch": 0.7244111820383007, + "grad_norm": 1.569559483205032, + "learning_rate": 1.8629230602860909e-06, + "loss": 0.3546, + "step": 32910 + }, + { + "epoch": 0.7244331939247194, + "grad_norm": 1.22685172300432, + "learning_rate": 1.8626455011137518e-06, + "loss": 0.2599, + "step": 32911 + }, + { + "epoch": 0.724455205811138, + "grad_norm": 1.365404132634771, + "learning_rate": 1.8623679578868497e-06, + "loss": 0.1999, + "step": 32912 + }, + { + "epoch": 0.7244772176975567, + "grad_norm": 0.5998237354054187, + "learning_rate": 1.862090430606791e-06, + "loss": 0.4566, + "step": 32913 + }, + { + "epoch": 0.7244992295839754, + "grad_norm": 1.132425398404813, + "learning_rate": 1.8618129192749885e-06, + "loss": 0.2288, + "step": 32914 + }, + { + "epoch": 0.724521241470394, + "grad_norm": 1.9206032265783306, + "learning_rate": 1.861535423892854e-06, + "loss": 0.3695, + "step": 32915 + }, + { + "epoch": 0.7245432533568127, + "grad_norm": 1.3611935594446394, + "learning_rate": 1.8612579444617945e-06, + "loss": 0.2652, + "step": 32916 + }, + { + "epoch": 0.7245652652432314, + "grad_norm": 2.0651472773135695, + "learning_rate": 1.8609804809832237e-06, + "loss": 0.2888, + "step": 32917 + }, + { + "epoch": 0.72458727712965, + "grad_norm": 1.182884357762988, + "learning_rate": 1.8607030334585507e-06, + "loss": 0.2738, + "step": 32918 + }, + { + "epoch": 0.7246092890160687, + "grad_norm": 1.483186839454834, + "learning_rate": 1.860425601889181e-06, + "loss": 0.2954, + "step": 32919 + }, + { + "epoch": 0.7246313009024874, + "grad_norm": 0.6285568778881857, + "learning_rate": 1.8601481862765324e-06, + "loss": 0.4572, + "step": 32920 + }, + { + "epoch": 0.724653312788906, + "grad_norm": 2.4259725794847937, + "learning_rate": 1.8598707866220094e-06, + "loss": 0.3833, + "step": 32921 + }, + { + "epoch": 0.7246753246753247, + "grad_norm": 1.3307339656668729, + "learning_rate": 1.8595934029270252e-06, + "loss": 0.2777, + "step": 32922 + }, + { + "epoch": 0.7246973365617433, + "grad_norm": 2.016567290308082, + "learning_rate": 1.859316035192988e-06, + "loss": 0.3572, + "step": 32923 + }, + { + "epoch": 0.724719348448162, + "grad_norm": 1.5865218995585801, + "learning_rate": 1.8590386834213059e-06, + "loss": 0.3726, + "step": 32924 + }, + { + "epoch": 0.7247413603345807, + "grad_norm": 1.5446754476411286, + "learning_rate": 1.8587613476133897e-06, + "loss": 0.4192, + "step": 32925 + }, + { + "epoch": 0.7247633722209993, + "grad_norm": 0.613507882930518, + "learning_rate": 1.8584840277706518e-06, + "loss": 0.4373, + "step": 32926 + }, + { + "epoch": 0.724785384107418, + "grad_norm": 1.504396047809772, + "learning_rate": 1.8582067238944968e-06, + "loss": 0.3886, + "step": 32927 + }, + { + "epoch": 0.7248073959938367, + "grad_norm": 2.1067276178969077, + "learning_rate": 1.8579294359863387e-06, + "loss": 0.2462, + "step": 32928 + }, + { + "epoch": 0.7248294078802553, + "grad_norm": 2.5147706221405626, + "learning_rate": 1.8576521640475826e-06, + "loss": 0.3312, + "step": 32929 + }, + { + "epoch": 0.724851419766674, + "grad_norm": 4.420897579448096, + "learning_rate": 1.85737490807964e-06, + "loss": 0.3626, + "step": 32930 + }, + { + "epoch": 0.7248734316530927, + "grad_norm": 1.7347153570045517, + "learning_rate": 1.8570976680839215e-06, + "loss": 0.3319, + "step": 32931 + }, + { + "epoch": 0.7248954435395113, + "grad_norm": 1.6738653600143152, + "learning_rate": 1.8568204440618326e-06, + "loss": 0.3825, + "step": 32932 + }, + { + "epoch": 0.72491745542593, + "grad_norm": 2.35203037642986, + "learning_rate": 1.8565432360147857e-06, + "loss": 0.2547, + "step": 32933 + }, + { + "epoch": 0.7249394673123487, + "grad_norm": 2.6145469131141748, + "learning_rate": 1.8562660439441864e-06, + "loss": 0.2889, + "step": 32934 + }, + { + "epoch": 0.7249614791987673, + "grad_norm": 1.6443585989152585, + "learning_rate": 1.8559888678514454e-06, + "loss": 0.2854, + "step": 32935 + }, + { + "epoch": 0.724983491085186, + "grad_norm": 1.4342617546259402, + "learning_rate": 1.8557117077379727e-06, + "loss": 0.2697, + "step": 32936 + }, + { + "epoch": 0.7250055029716047, + "grad_norm": 1.9093240081932623, + "learning_rate": 1.855434563605174e-06, + "loss": 0.2821, + "step": 32937 + }, + { + "epoch": 0.7250275148580233, + "grad_norm": 0.6200984910887675, + "learning_rate": 1.855157435454461e-06, + "loss": 0.4393, + "step": 32938 + }, + { + "epoch": 0.725049526744442, + "grad_norm": 3.233864476543831, + "learning_rate": 1.8548803232872387e-06, + "loss": 0.2941, + "step": 32939 + }, + { + "epoch": 0.7250715386308607, + "grad_norm": 1.9070408150946452, + "learning_rate": 1.8546032271049175e-06, + "loss": 0.3494, + "step": 32940 + }, + { + "epoch": 0.7250935505172793, + "grad_norm": 1.4934179113304422, + "learning_rate": 1.8543261469089074e-06, + "loss": 0.2637, + "step": 32941 + }, + { + "epoch": 0.725115562403698, + "grad_norm": 2.780947871220909, + "learning_rate": 1.854049082700613e-06, + "loss": 0.3467, + "step": 32942 + }, + { + "epoch": 0.7251375742901167, + "grad_norm": 0.6086084798229511, + "learning_rate": 1.8537720344814458e-06, + "loss": 0.4524, + "step": 32943 + }, + { + "epoch": 0.7251595861765353, + "grad_norm": 1.3194244357047078, + "learning_rate": 1.853495002252811e-06, + "loss": 0.2676, + "step": 32944 + }, + { + "epoch": 0.725181598062954, + "grad_norm": 1.6639659809486715, + "learning_rate": 1.8532179860161177e-06, + "loss": 0.316, + "step": 32945 + }, + { + "epoch": 0.7252036099493726, + "grad_norm": 1.5207619868586622, + "learning_rate": 1.8529409857727759e-06, + "loss": 0.2964, + "step": 32946 + }, + { + "epoch": 0.7252256218357913, + "grad_norm": 1.3142713932195862, + "learning_rate": 1.8526640015241904e-06, + "loss": 0.2037, + "step": 32947 + }, + { + "epoch": 0.72524763372221, + "grad_norm": 1.524520604741869, + "learning_rate": 1.8523870332717713e-06, + "loss": 0.3833, + "step": 32948 + }, + { + "epoch": 0.7252696456086286, + "grad_norm": 0.6487979368353294, + "learning_rate": 1.852110081016925e-06, + "loss": 0.462, + "step": 32949 + }, + { + "epoch": 0.7252916574950473, + "grad_norm": 0.6131720475935896, + "learning_rate": 1.8518331447610565e-06, + "loss": 0.4362, + "step": 32950 + }, + { + "epoch": 0.725313669381466, + "grad_norm": 0.5977242630494183, + "learning_rate": 1.851556224505579e-06, + "loss": 0.4409, + "step": 32951 + }, + { + "epoch": 0.7253356812678846, + "grad_norm": 1.3295732902895654, + "learning_rate": 1.8512793202518959e-06, + "loss": 0.2408, + "step": 32952 + }, + { + "epoch": 0.7253576931543033, + "grad_norm": 1.256119606700478, + "learning_rate": 1.851002432001417e-06, + "loss": 0.2437, + "step": 32953 + }, + { + "epoch": 0.725379705040722, + "grad_norm": 1.7943826537374468, + "learning_rate": 1.8507255597555479e-06, + "loss": 0.2806, + "step": 32954 + }, + { + "epoch": 0.7254017169271406, + "grad_norm": 1.8117952811177627, + "learning_rate": 1.8504487035156932e-06, + "loss": 0.1983, + "step": 32955 + }, + { + "epoch": 0.7254237288135593, + "grad_norm": 0.6011648854379128, + "learning_rate": 1.850171863283266e-06, + "loss": 0.4412, + "step": 32956 + }, + { + "epoch": 0.725445740699978, + "grad_norm": 1.7243664317746579, + "learning_rate": 1.8498950390596688e-06, + "loss": 0.2446, + "step": 32957 + }, + { + "epoch": 0.7254677525863966, + "grad_norm": 1.737267394342177, + "learning_rate": 1.8496182308463118e-06, + "loss": 0.3014, + "step": 32958 + }, + { + "epoch": 0.7254897644728153, + "grad_norm": 1.7769623214576062, + "learning_rate": 1.8493414386445997e-06, + "loss": 0.2344, + "step": 32959 + }, + { + "epoch": 0.725511776359234, + "grad_norm": 1.323353487762859, + "learning_rate": 1.8490646624559383e-06, + "loss": 0.2664, + "step": 32960 + }, + { + "epoch": 0.7255337882456526, + "grad_norm": 1.5534001026094777, + "learning_rate": 1.8487879022817352e-06, + "loss": 0.2799, + "step": 32961 + }, + { + "epoch": 0.7255558001320713, + "grad_norm": 1.9657716153810052, + "learning_rate": 1.8485111581233979e-06, + "loss": 0.2374, + "step": 32962 + }, + { + "epoch": 0.72557781201849, + "grad_norm": 1.741022069553524, + "learning_rate": 1.848234429982334e-06, + "loss": 0.2706, + "step": 32963 + }, + { + "epoch": 0.7255998239049086, + "grad_norm": 8.522182678994293, + "learning_rate": 1.8479577178599485e-06, + "loss": 0.2263, + "step": 32964 + }, + { + "epoch": 0.7256218357913273, + "grad_norm": 1.6357315732546691, + "learning_rate": 1.8476810217576458e-06, + "loss": 0.3274, + "step": 32965 + }, + { + "epoch": 0.725643847677746, + "grad_norm": 1.5037541473922302, + "learning_rate": 1.8474043416768344e-06, + "loss": 0.2213, + "step": 32966 + }, + { + "epoch": 0.7256658595641646, + "grad_norm": 20.654505687332556, + "learning_rate": 1.8471276776189217e-06, + "loss": 0.357, + "step": 32967 + }, + { + "epoch": 0.7256878714505833, + "grad_norm": 1.541832516572232, + "learning_rate": 1.846851029585311e-06, + "loss": 0.3192, + "step": 32968 + }, + { + "epoch": 0.7257098833370019, + "grad_norm": 0.6032852394304733, + "learning_rate": 1.846574397577411e-06, + "loss": 0.4412, + "step": 32969 + }, + { + "epoch": 0.7257318952234206, + "grad_norm": 1.871509487004424, + "learning_rate": 1.846297781596625e-06, + "loss": 0.3414, + "step": 32970 + }, + { + "epoch": 0.7257539071098393, + "grad_norm": 1.2423938452083823, + "learning_rate": 1.8460211816443603e-06, + "loss": 0.1991, + "step": 32971 + }, + { + "epoch": 0.7257759189962579, + "grad_norm": 1.868736427574475, + "learning_rate": 1.8457445977220239e-06, + "loss": 0.3889, + "step": 32972 + }, + { + "epoch": 0.7257979308826766, + "grad_norm": 0.6345351551824381, + "learning_rate": 1.845468029831019e-06, + "loss": 0.4941, + "step": 32973 + }, + { + "epoch": 0.7258199427690953, + "grad_norm": 3.685771783123579, + "learning_rate": 1.8451914779727543e-06, + "loss": 0.2774, + "step": 32974 + }, + { + "epoch": 0.7258419546555139, + "grad_norm": 0.5918080155685309, + "learning_rate": 1.8449149421486317e-06, + "loss": 0.4389, + "step": 32975 + }, + { + "epoch": 0.7258639665419326, + "grad_norm": 2.4069376535275597, + "learning_rate": 1.8446384223600588e-06, + "loss": 0.3127, + "step": 32976 + }, + { + "epoch": 0.7258859784283513, + "grad_norm": 1.2994948805411428, + "learning_rate": 1.8443619186084422e-06, + "loss": 0.3147, + "step": 32977 + }, + { + "epoch": 0.7259079903147699, + "grad_norm": 1.7324140627997604, + "learning_rate": 1.844085430895184e-06, + "loss": 0.3377, + "step": 32978 + }, + { + "epoch": 0.7259300022011886, + "grad_norm": 2.714528936646247, + "learning_rate": 1.8438089592216935e-06, + "loss": 0.2419, + "step": 32979 + }, + { + "epoch": 0.7259520140876073, + "grad_norm": 1.6910471884024199, + "learning_rate": 1.8435325035893713e-06, + "loss": 0.228, + "step": 32980 + }, + { + "epoch": 0.7259740259740259, + "grad_norm": 1.2441616613110702, + "learning_rate": 1.843256063999625e-06, + "loss": 0.2572, + "step": 32981 + }, + { + "epoch": 0.7259960378604446, + "grad_norm": 5.645354906979629, + "learning_rate": 1.8429796404538608e-06, + "loss": 0.2615, + "step": 32982 + }, + { + "epoch": 0.7260180497468633, + "grad_norm": 3.344735938095816, + "learning_rate": 1.8427032329534805e-06, + "loss": 0.2501, + "step": 32983 + }, + { + "epoch": 0.7260400616332819, + "grad_norm": 2.2782396131082456, + "learning_rate": 1.8424268414998926e-06, + "loss": 0.2737, + "step": 32984 + }, + { + "epoch": 0.7260620735197006, + "grad_norm": 1.5010444750044563, + "learning_rate": 1.8421504660944995e-06, + "loss": 0.2449, + "step": 32985 + }, + { + "epoch": 0.7260840854061194, + "grad_norm": 0.6120928925425702, + "learning_rate": 1.8418741067387025e-06, + "loss": 0.4335, + "step": 32986 + }, + { + "epoch": 0.7261060972925379, + "grad_norm": 0.6279674480832513, + "learning_rate": 1.841597763433914e-06, + "loss": 0.4445, + "step": 32987 + }, + { + "epoch": 0.7261281091789566, + "grad_norm": 1.9612949300122253, + "learning_rate": 1.8413214361815318e-06, + "loss": 0.2715, + "step": 32988 + }, + { + "epoch": 0.7261501210653754, + "grad_norm": 1.502864851713552, + "learning_rate": 1.8410451249829653e-06, + "loss": 0.3593, + "step": 32989 + }, + { + "epoch": 0.726172132951794, + "grad_norm": 1.5870072297659523, + "learning_rate": 1.8407688298396158e-06, + "loss": 0.2761, + "step": 32990 + }, + { + "epoch": 0.7261941448382127, + "grad_norm": 0.6173665399915181, + "learning_rate": 1.8404925507528853e-06, + "loss": 0.4429, + "step": 32991 + }, + { + "epoch": 0.7262161567246312, + "grad_norm": 2.2698018629932504, + "learning_rate": 1.8402162877241836e-06, + "loss": 0.2742, + "step": 32992 + }, + { + "epoch": 0.72623816861105, + "grad_norm": 1.4941158839504487, + "learning_rate": 1.8399400407549113e-06, + "loss": 0.2103, + "step": 32993 + }, + { + "epoch": 0.7262601804974687, + "grad_norm": 1.4866444699784016, + "learning_rate": 1.8396638098464735e-06, + "loss": 0.3068, + "step": 32994 + }, + { + "epoch": 0.7262821923838872, + "grad_norm": 0.5930843691735574, + "learning_rate": 1.8393875950002748e-06, + "loss": 0.4553, + "step": 32995 + }, + { + "epoch": 0.726304204270306, + "grad_norm": 1.6922440480165188, + "learning_rate": 1.839111396217716e-06, + "loss": 0.3385, + "step": 32996 + }, + { + "epoch": 0.7263262161567247, + "grad_norm": 1.849310431553615, + "learning_rate": 1.838835213500203e-06, + "loss": 0.3489, + "step": 32997 + }, + { + "epoch": 0.7263482280431433, + "grad_norm": 1.6665046877210663, + "learning_rate": 1.8385590468491393e-06, + "loss": 0.3121, + "step": 32998 + }, + { + "epoch": 0.726370239929562, + "grad_norm": 2.028133459549262, + "learning_rate": 1.8382828962659305e-06, + "loss": 0.3096, + "step": 32999 + }, + { + "epoch": 0.7263922518159807, + "grad_norm": 3.984796586199814, + "learning_rate": 1.8380067617519775e-06, + "loss": 0.2565, + "step": 33000 + }, + { + "epoch": 0.7264142637023993, + "grad_norm": 2.141148326496005, + "learning_rate": 1.8377306433086832e-06, + "loss": 0.3572, + "step": 33001 + }, + { + "epoch": 0.726436275588818, + "grad_norm": 1.2289357675677304, + "learning_rate": 1.8374545409374523e-06, + "loss": 0.2033, + "step": 33002 + }, + { + "epoch": 0.7264582874752367, + "grad_norm": 0.6211707984424567, + "learning_rate": 1.8371784546396881e-06, + "loss": 0.4774, + "step": 33003 + }, + { + "epoch": 0.7264802993616553, + "grad_norm": 1.4452103683058801, + "learning_rate": 1.836902384416795e-06, + "loss": 0.276, + "step": 33004 + }, + { + "epoch": 0.726502311248074, + "grad_norm": 17.070509997131747, + "learning_rate": 1.8366263302701748e-06, + "loss": 0.2357, + "step": 33005 + }, + { + "epoch": 0.7265243231344927, + "grad_norm": 1.7337181802300874, + "learning_rate": 1.8363502922012288e-06, + "loss": 0.2557, + "step": 33006 + }, + { + "epoch": 0.7265463350209113, + "grad_norm": 1.9130411658421491, + "learning_rate": 1.8360742702113615e-06, + "loss": 0.2868, + "step": 33007 + }, + { + "epoch": 0.72656834690733, + "grad_norm": 1.6977175963333, + "learning_rate": 1.8357982643019777e-06, + "loss": 0.2575, + "step": 33008 + }, + { + "epoch": 0.7265903587937487, + "grad_norm": 1.7094855267113664, + "learning_rate": 1.835522274474477e-06, + "loss": 0.3091, + "step": 33009 + }, + { + "epoch": 0.7266123706801673, + "grad_norm": 1.8229900532357821, + "learning_rate": 1.835246300730265e-06, + "loss": 0.3817, + "step": 33010 + }, + { + "epoch": 0.726634382566586, + "grad_norm": 1.938517874556577, + "learning_rate": 1.8349703430707411e-06, + "loss": 0.3282, + "step": 33011 + }, + { + "epoch": 0.7266563944530047, + "grad_norm": 1.2992478503601512, + "learning_rate": 1.8346944014973095e-06, + "loss": 0.2723, + "step": 33012 + }, + { + "epoch": 0.7266784063394233, + "grad_norm": 1.593276007426082, + "learning_rate": 1.8344184760113742e-06, + "loss": 0.3044, + "step": 33013 + }, + { + "epoch": 0.726700418225842, + "grad_norm": 1.8952408452163572, + "learning_rate": 1.8341425666143348e-06, + "loss": 0.2735, + "step": 33014 + }, + { + "epoch": 0.7267224301122607, + "grad_norm": 1.432688242593203, + "learning_rate": 1.8338666733075967e-06, + "loss": 0.3115, + "step": 33015 + }, + { + "epoch": 0.7267444419986793, + "grad_norm": 1.9953610458852398, + "learning_rate": 1.833590796092558e-06, + "loss": 0.2925, + "step": 33016 + }, + { + "epoch": 0.726766453885098, + "grad_norm": 1.8935570788172071, + "learning_rate": 1.8333149349706236e-06, + "loss": 0.3742, + "step": 33017 + }, + { + "epoch": 0.7267884657715166, + "grad_norm": 0.6276814695142268, + "learning_rate": 1.8330390899431965e-06, + "loss": 0.469, + "step": 33018 + }, + { + "epoch": 0.7268104776579353, + "grad_norm": 1.6525022507210778, + "learning_rate": 1.8327632610116758e-06, + "loss": 0.2961, + "step": 33019 + }, + { + "epoch": 0.726832489544354, + "grad_norm": 1.398363631028707, + "learning_rate": 1.8324874481774663e-06, + "loss": 0.2372, + "step": 33020 + }, + { + "epoch": 0.7268545014307726, + "grad_norm": 2.2328563535564028, + "learning_rate": 1.8322116514419668e-06, + "loss": 0.3062, + "step": 33021 + }, + { + "epoch": 0.7268765133171913, + "grad_norm": 3.637071005956556, + "learning_rate": 1.8319358708065804e-06, + "loss": 0.2747, + "step": 33022 + }, + { + "epoch": 0.72689852520361, + "grad_norm": 3.359733595832807, + "learning_rate": 1.8316601062727108e-06, + "loss": 0.3058, + "step": 33023 + }, + { + "epoch": 0.7269205370900286, + "grad_norm": 1.9080463455746595, + "learning_rate": 1.8313843578417562e-06, + "loss": 0.2911, + "step": 33024 + }, + { + "epoch": 0.7269425489764473, + "grad_norm": 1.501444386312322, + "learning_rate": 1.8311086255151205e-06, + "loss": 0.3, + "step": 33025 + }, + { + "epoch": 0.726964560862866, + "grad_norm": 0.6184911242648105, + "learning_rate": 1.8308329092942052e-06, + "loss": 0.4815, + "step": 33026 + }, + { + "epoch": 0.7269865727492846, + "grad_norm": 1.6866434743355077, + "learning_rate": 1.8305572091804064e-06, + "loss": 0.302, + "step": 33027 + }, + { + "epoch": 0.7270085846357033, + "grad_norm": 2.3528807307067874, + "learning_rate": 1.8302815251751326e-06, + "loss": 0.2768, + "step": 33028 + }, + { + "epoch": 0.727030596522122, + "grad_norm": 1.997793884245513, + "learning_rate": 1.8300058572797807e-06, + "loss": 0.231, + "step": 33029 + }, + { + "epoch": 0.7270526084085406, + "grad_norm": 1.4820389894932176, + "learning_rate": 1.8297302054957544e-06, + "loss": 0.3286, + "step": 33030 + }, + { + "epoch": 0.7270746202949593, + "grad_norm": 2.4569216210219804, + "learning_rate": 1.829454569824453e-06, + "loss": 0.2865, + "step": 33031 + }, + { + "epoch": 0.727096632181378, + "grad_norm": 1.5975139100802274, + "learning_rate": 1.8291789502672762e-06, + "loss": 0.2699, + "step": 33032 + }, + { + "epoch": 0.7271186440677966, + "grad_norm": 0.6343810659999254, + "learning_rate": 1.8289033468256261e-06, + "loss": 0.4471, + "step": 33033 + }, + { + "epoch": 0.7271406559542153, + "grad_norm": 2.365195528483102, + "learning_rate": 1.8286277595009033e-06, + "loss": 0.3059, + "step": 33034 + }, + { + "epoch": 0.727162667840634, + "grad_norm": 3.0138591030526958, + "learning_rate": 1.8283521882945105e-06, + "loss": 0.2335, + "step": 33035 + }, + { + "epoch": 0.7271846797270526, + "grad_norm": 2.690101347887475, + "learning_rate": 1.828076633207847e-06, + "loss": 0.2424, + "step": 33036 + }, + { + "epoch": 0.7272066916134713, + "grad_norm": 1.6481679404972696, + "learning_rate": 1.8278010942423101e-06, + "loss": 0.3136, + "step": 33037 + }, + { + "epoch": 0.72722870349989, + "grad_norm": 1.8202164334078785, + "learning_rate": 1.8275255713993034e-06, + "loss": 0.2904, + "step": 33038 + }, + { + "epoch": 0.7272507153863086, + "grad_norm": 1.348847084388448, + "learning_rate": 1.8272500646802271e-06, + "loss": 0.2739, + "step": 33039 + }, + { + "epoch": 0.7272727272727273, + "grad_norm": 0.5954158497870573, + "learning_rate": 1.8269745740864824e-06, + "loss": 0.4314, + "step": 33040 + }, + { + "epoch": 0.7272947391591459, + "grad_norm": 1.725744763020237, + "learning_rate": 1.8266990996194678e-06, + "loss": 0.2936, + "step": 33041 + }, + { + "epoch": 0.7273167510455646, + "grad_norm": 1.4163733850587072, + "learning_rate": 1.8264236412805825e-06, + "loss": 0.2932, + "step": 33042 + }, + { + "epoch": 0.7273387629319833, + "grad_norm": 1.6604638477485172, + "learning_rate": 1.8261481990712276e-06, + "loss": 0.2785, + "step": 33043 + }, + { + "epoch": 0.7273607748184019, + "grad_norm": 0.6103607924645759, + "learning_rate": 1.8258727729928044e-06, + "loss": 0.4644, + "step": 33044 + }, + { + "epoch": 0.7273827867048206, + "grad_norm": 1.435807148636642, + "learning_rate": 1.8255973630467105e-06, + "loss": 0.2862, + "step": 33045 + }, + { + "epoch": 0.7274047985912393, + "grad_norm": 1.4501289427780655, + "learning_rate": 1.8253219692343476e-06, + "loss": 0.2295, + "step": 33046 + }, + { + "epoch": 0.7274268104776579, + "grad_norm": 1.678078291076896, + "learning_rate": 1.825046591557113e-06, + "loss": 0.3833, + "step": 33047 + }, + { + "epoch": 0.7274488223640766, + "grad_norm": 1.5602665107977305, + "learning_rate": 1.8247712300164078e-06, + "loss": 0.2412, + "step": 33048 + }, + { + "epoch": 0.7274708342504953, + "grad_norm": 1.5371401823786104, + "learning_rate": 1.824495884613633e-06, + "loss": 0.3027, + "step": 33049 + }, + { + "epoch": 0.7274928461369139, + "grad_norm": 2.939499441511006, + "learning_rate": 1.8242205553501846e-06, + "loss": 0.3392, + "step": 33050 + }, + { + "epoch": 0.7275148580233326, + "grad_norm": 3.2033235624581677, + "learning_rate": 1.823945242227466e-06, + "loss": 0.3773, + "step": 33051 + }, + { + "epoch": 0.7275368699097513, + "grad_norm": 1.700647250413716, + "learning_rate": 1.823669945246872e-06, + "loss": 0.3479, + "step": 33052 + }, + { + "epoch": 0.7275588817961699, + "grad_norm": 1.4640187232480408, + "learning_rate": 1.8233946644098044e-06, + "loss": 0.2642, + "step": 33053 + }, + { + "epoch": 0.7275808936825886, + "grad_norm": 2.6787948191167, + "learning_rate": 1.8231193997176632e-06, + "loss": 0.3509, + "step": 33054 + }, + { + "epoch": 0.7276029055690073, + "grad_norm": 1.4075500849719262, + "learning_rate": 1.8228441511718447e-06, + "loss": 0.2917, + "step": 33055 + }, + { + "epoch": 0.7276249174554259, + "grad_norm": 3.5515168211470494, + "learning_rate": 1.822568918773751e-06, + "loss": 0.254, + "step": 33056 + }, + { + "epoch": 0.7276469293418446, + "grad_norm": 1.638950767289522, + "learning_rate": 1.8222937025247773e-06, + "loss": 0.2838, + "step": 33057 + }, + { + "epoch": 0.7276689412282633, + "grad_norm": 1.6948303915355911, + "learning_rate": 1.8220185024263243e-06, + "loss": 0.2228, + "step": 33058 + }, + { + "epoch": 0.7276909531146819, + "grad_norm": 1.2407304268595485, + "learning_rate": 1.8217433184797928e-06, + "loss": 0.3393, + "step": 33059 + }, + { + "epoch": 0.7277129650011006, + "grad_norm": 1.4196585122467822, + "learning_rate": 1.8214681506865773e-06, + "loss": 0.2282, + "step": 33060 + }, + { + "epoch": 0.7277349768875193, + "grad_norm": 0.6116577685129545, + "learning_rate": 1.8211929990480803e-06, + "loss": 0.4346, + "step": 33061 + }, + { + "epoch": 0.7277569887739379, + "grad_norm": 2.1320172110165507, + "learning_rate": 1.820917863565696e-06, + "loss": 0.297, + "step": 33062 + }, + { + "epoch": 0.7277790006603566, + "grad_norm": 1.2685990249986867, + "learning_rate": 1.8206427442408253e-06, + "loss": 0.2201, + "step": 33063 + }, + { + "epoch": 0.7278010125467752, + "grad_norm": 1.726276096564178, + "learning_rate": 1.8203676410748684e-06, + "loss": 0.2943, + "step": 33064 + }, + { + "epoch": 0.7278230244331939, + "grad_norm": 1.3205616511345608, + "learning_rate": 1.8200925540692187e-06, + "loss": 0.227, + "step": 33065 + }, + { + "epoch": 0.7278450363196126, + "grad_norm": 1.4171811602839803, + "learning_rate": 1.8198174832252791e-06, + "loss": 0.2773, + "step": 33066 + }, + { + "epoch": 0.7278670482060312, + "grad_norm": 1.5538350867058952, + "learning_rate": 1.8195424285444457e-06, + "loss": 0.2577, + "step": 33067 + }, + { + "epoch": 0.7278890600924499, + "grad_norm": 1.7529617971890896, + "learning_rate": 1.8192673900281143e-06, + "loss": 0.3325, + "step": 33068 + }, + { + "epoch": 0.7279110719788686, + "grad_norm": 1.5067713106745, + "learning_rate": 1.8189923676776844e-06, + "loss": 0.2986, + "step": 33069 + }, + { + "epoch": 0.7279330838652872, + "grad_norm": 2.3820924070118457, + "learning_rate": 1.8187173614945542e-06, + "loss": 0.2963, + "step": 33070 + }, + { + "epoch": 0.7279550957517059, + "grad_norm": 1.611073337875079, + "learning_rate": 1.818442371480123e-06, + "loss": 0.3633, + "step": 33071 + }, + { + "epoch": 0.7279771076381246, + "grad_norm": 1.6842145666935724, + "learning_rate": 1.8181673976357865e-06, + "loss": 0.3226, + "step": 33072 + }, + { + "epoch": 0.7279991195245432, + "grad_norm": 1.5717591193833866, + "learning_rate": 1.8178924399629411e-06, + "loss": 0.2388, + "step": 33073 + }, + { + "epoch": 0.7280211314109619, + "grad_norm": 2.00067098845662, + "learning_rate": 1.8176174984629858e-06, + "loss": 0.2724, + "step": 33074 + }, + { + "epoch": 0.7280431432973806, + "grad_norm": 1.3138197840166566, + "learning_rate": 1.8173425731373173e-06, + "loss": 0.299, + "step": 33075 + }, + { + "epoch": 0.7280651551837992, + "grad_norm": 2.893541724775054, + "learning_rate": 1.8170676639873347e-06, + "loss": 0.2834, + "step": 33076 + }, + { + "epoch": 0.7280871670702179, + "grad_norm": 1.3893258376263549, + "learning_rate": 1.8167927710144346e-06, + "loss": 0.3464, + "step": 33077 + }, + { + "epoch": 0.7281091789566366, + "grad_norm": 1.3013434828458057, + "learning_rate": 1.8165178942200113e-06, + "loss": 0.2368, + "step": 33078 + }, + { + "epoch": 0.7281311908430552, + "grad_norm": 1.592470391447731, + "learning_rate": 1.8162430336054638e-06, + "loss": 0.2929, + "step": 33079 + }, + { + "epoch": 0.7281532027294739, + "grad_norm": 1.7475507854210495, + "learning_rate": 1.8159681891721897e-06, + "loss": 0.3252, + "step": 33080 + }, + { + "epoch": 0.7281752146158926, + "grad_norm": 1.8604512686778782, + "learning_rate": 1.8156933609215866e-06, + "loss": 0.3401, + "step": 33081 + }, + { + "epoch": 0.7281972265023112, + "grad_norm": 1.510235549844643, + "learning_rate": 1.8154185488550503e-06, + "loss": 0.2925, + "step": 33082 + }, + { + "epoch": 0.7282192383887299, + "grad_norm": 1.4116808222595345, + "learning_rate": 1.8151437529739758e-06, + "loss": 0.3561, + "step": 33083 + }, + { + "epoch": 0.7282412502751486, + "grad_norm": 1.8702948868337785, + "learning_rate": 1.8148689732797608e-06, + "loss": 0.3109, + "step": 33084 + }, + { + "epoch": 0.7282632621615672, + "grad_norm": 2.1307337280072436, + "learning_rate": 1.814594209773804e-06, + "loss": 0.2814, + "step": 33085 + }, + { + "epoch": 0.7282852740479859, + "grad_norm": 1.8878405013872979, + "learning_rate": 1.8143194624574988e-06, + "loss": 0.2793, + "step": 33086 + }, + { + "epoch": 0.7283072859344045, + "grad_norm": 1.6878919882083627, + "learning_rate": 1.8140447313322445e-06, + "loss": 0.2333, + "step": 33087 + }, + { + "epoch": 0.7283292978208232, + "grad_norm": 1.4069245456656445, + "learning_rate": 1.8137700163994338e-06, + "loss": 0.2394, + "step": 33088 + }, + { + "epoch": 0.7283513097072419, + "grad_norm": 1.6528328999932869, + "learning_rate": 1.8134953176604652e-06, + "loss": 0.3707, + "step": 33089 + }, + { + "epoch": 0.7283733215936605, + "grad_norm": 1.4393195231633926, + "learning_rate": 1.8132206351167369e-06, + "loss": 0.253, + "step": 33090 + }, + { + "epoch": 0.7283953334800792, + "grad_norm": 1.9308975832078623, + "learning_rate": 1.8129459687696405e-06, + "loss": 0.3415, + "step": 33091 + }, + { + "epoch": 0.7284173453664979, + "grad_norm": 0.629501612499326, + "learning_rate": 1.8126713186205757e-06, + "loss": 0.4823, + "step": 33092 + }, + { + "epoch": 0.7284393572529165, + "grad_norm": 1.5476013504075157, + "learning_rate": 1.812396684670935e-06, + "loss": 0.3123, + "step": 33093 + }, + { + "epoch": 0.7284613691393352, + "grad_norm": 0.6287266910043027, + "learning_rate": 1.8121220669221167e-06, + "loss": 0.4508, + "step": 33094 + }, + { + "epoch": 0.7284833810257539, + "grad_norm": 2.139200952291099, + "learning_rate": 1.8118474653755169e-06, + "loss": 0.2964, + "step": 33095 + }, + { + "epoch": 0.7285053929121725, + "grad_norm": 0.5970872874985158, + "learning_rate": 1.8115728800325289e-06, + "loss": 0.48, + "step": 33096 + }, + { + "epoch": 0.7285274047985912, + "grad_norm": 1.4754902082521102, + "learning_rate": 1.8112983108945514e-06, + "loss": 0.3597, + "step": 33097 + }, + { + "epoch": 0.7285494166850099, + "grad_norm": 0.6031748365468109, + "learning_rate": 1.8110237579629763e-06, + "loss": 0.4406, + "step": 33098 + }, + { + "epoch": 0.7285714285714285, + "grad_norm": 0.6136951916486809, + "learning_rate": 1.8107492212392025e-06, + "loss": 0.424, + "step": 33099 + }, + { + "epoch": 0.7285934404578472, + "grad_norm": 1.6092016690893762, + "learning_rate": 1.810474700724622e-06, + "loss": 0.306, + "step": 33100 + }, + { + "epoch": 0.728615452344266, + "grad_norm": 3.0840075079898983, + "learning_rate": 1.8102001964206312e-06, + "loss": 0.2984, + "step": 33101 + }, + { + "epoch": 0.7286374642306845, + "grad_norm": 2.42649800718441, + "learning_rate": 1.8099257083286275e-06, + "loss": 0.2812, + "step": 33102 + }, + { + "epoch": 0.7286594761171032, + "grad_norm": 1.8596995865174, + "learning_rate": 1.809651236450003e-06, + "loss": 0.3063, + "step": 33103 + }, + { + "epoch": 0.728681488003522, + "grad_norm": 1.8250970046204702, + "learning_rate": 1.809376780786155e-06, + "loss": 0.3533, + "step": 33104 + }, + { + "epoch": 0.7287034998899405, + "grad_norm": 1.97463079202981, + "learning_rate": 1.8091023413384755e-06, + "loss": 0.3084, + "step": 33105 + }, + { + "epoch": 0.7287255117763592, + "grad_norm": 1.570695502331127, + "learning_rate": 1.8088279181083618e-06, + "loss": 0.249, + "step": 33106 + }, + { + "epoch": 0.728747523662778, + "grad_norm": 1.4520218180381037, + "learning_rate": 1.808553511097209e-06, + "loss": 0.3622, + "step": 33107 + }, + { + "epoch": 0.7287695355491965, + "grad_norm": 1.5280877333938394, + "learning_rate": 1.808279120306411e-06, + "loss": 0.2252, + "step": 33108 + }, + { + "epoch": 0.7287915474356153, + "grad_norm": 2.786394162793967, + "learning_rate": 1.80800474573736e-06, + "loss": 0.3228, + "step": 33109 + }, + { + "epoch": 0.7288135593220338, + "grad_norm": 2.1988773592286446, + "learning_rate": 1.8077303873914525e-06, + "loss": 0.3101, + "step": 33110 + }, + { + "epoch": 0.7288355712084525, + "grad_norm": 1.5293985573040587, + "learning_rate": 1.8074560452700834e-06, + "loss": 0.3278, + "step": 33111 + }, + { + "epoch": 0.7288575830948713, + "grad_norm": 0.5859240713457543, + "learning_rate": 1.8071817193746483e-06, + "loss": 0.4453, + "step": 33112 + }, + { + "epoch": 0.7288795949812898, + "grad_norm": 2.0320287767990997, + "learning_rate": 1.8069074097065393e-06, + "loss": 0.4089, + "step": 33113 + }, + { + "epoch": 0.7289016068677086, + "grad_norm": 1.3957497229382594, + "learning_rate": 1.8066331162671496e-06, + "loss": 0.25, + "step": 33114 + }, + { + "epoch": 0.7289236187541273, + "grad_norm": 1.5199910496244977, + "learning_rate": 1.8063588390578746e-06, + "loss": 0.3533, + "step": 33115 + }, + { + "epoch": 0.7289456306405458, + "grad_norm": 1.456422343045629, + "learning_rate": 1.8060845780801085e-06, + "loss": 0.2596, + "step": 33116 + }, + { + "epoch": 0.7289676425269646, + "grad_norm": 1.8743160457164691, + "learning_rate": 1.8058103333352468e-06, + "loss": 0.3019, + "step": 33117 + }, + { + "epoch": 0.7289896544133833, + "grad_norm": 1.9695002846070588, + "learning_rate": 1.8055361048246817e-06, + "loss": 0.3333, + "step": 33118 + }, + { + "epoch": 0.7290116662998019, + "grad_norm": 1.663584350229793, + "learning_rate": 1.805261892549805e-06, + "loss": 0.2903, + "step": 33119 + }, + { + "epoch": 0.7290336781862206, + "grad_norm": 1.9667112535364482, + "learning_rate": 1.8049876965120123e-06, + "loss": 0.3314, + "step": 33120 + }, + { + "epoch": 0.7290556900726393, + "grad_norm": 4.7804789851694895, + "learning_rate": 1.8047135167126973e-06, + "loss": 0.2887, + "step": 33121 + }, + { + "epoch": 0.7290777019590579, + "grad_norm": 1.6124620028179342, + "learning_rate": 1.8044393531532545e-06, + "loss": 0.267, + "step": 33122 + }, + { + "epoch": 0.7290997138454766, + "grad_norm": 0.6123651393842869, + "learning_rate": 1.8041652058350768e-06, + "loss": 0.4643, + "step": 33123 + }, + { + "epoch": 0.7291217257318953, + "grad_norm": 0.6328384790334823, + "learning_rate": 1.8038910747595546e-06, + "loss": 0.4486, + "step": 33124 + }, + { + "epoch": 0.7291437376183139, + "grad_norm": 1.520012205898537, + "learning_rate": 1.803616959928084e-06, + "loss": 0.2713, + "step": 33125 + }, + { + "epoch": 0.7291657495047326, + "grad_norm": 1.5012954300341708, + "learning_rate": 1.803342861342059e-06, + "loss": 0.2923, + "step": 33126 + }, + { + "epoch": 0.7291877613911513, + "grad_norm": 2.163520876362154, + "learning_rate": 1.80306877900287e-06, + "loss": 0.2411, + "step": 33127 + }, + { + "epoch": 0.7292097732775699, + "grad_norm": 23.753269513937205, + "learning_rate": 1.8027947129119127e-06, + "loss": 0.3144, + "step": 33128 + }, + { + "epoch": 0.7292317851639886, + "grad_norm": 1.4965950819917277, + "learning_rate": 1.8025206630705771e-06, + "loss": 0.3495, + "step": 33129 + }, + { + "epoch": 0.7292537970504073, + "grad_norm": 2.548500269579777, + "learning_rate": 1.8022466294802577e-06, + "loss": 0.2917, + "step": 33130 + }, + { + "epoch": 0.7292758089368259, + "grad_norm": 1.3464639541763346, + "learning_rate": 1.8019726121423486e-06, + "loss": 0.3096, + "step": 33131 + }, + { + "epoch": 0.7292978208232446, + "grad_norm": 2.4893197153647555, + "learning_rate": 1.8016986110582396e-06, + "loss": 0.3097, + "step": 33132 + }, + { + "epoch": 0.7293198327096632, + "grad_norm": 4.046187092586115, + "learning_rate": 1.8014246262293261e-06, + "loss": 0.3325, + "step": 33133 + }, + { + "epoch": 0.7293418445960819, + "grad_norm": 2.0464096736189012, + "learning_rate": 1.8011506576569982e-06, + "loss": 0.3321, + "step": 33134 + }, + { + "epoch": 0.7293638564825006, + "grad_norm": 1.5154978342632228, + "learning_rate": 1.8008767053426501e-06, + "loss": 0.3449, + "step": 33135 + }, + { + "epoch": 0.7293858683689192, + "grad_norm": 0.6320871114919555, + "learning_rate": 1.8006027692876727e-06, + "loss": 0.4615, + "step": 33136 + }, + { + "epoch": 0.7294078802553379, + "grad_norm": 2.127239534431345, + "learning_rate": 1.8003288494934584e-06, + "loss": 0.3357, + "step": 33137 + }, + { + "epoch": 0.7294298921417566, + "grad_norm": 0.6484268154186583, + "learning_rate": 1.800054945961402e-06, + "loss": 0.4752, + "step": 33138 + }, + { + "epoch": 0.7294519040281752, + "grad_norm": 0.6094783316801375, + "learning_rate": 1.7997810586928916e-06, + "loss": 0.4576, + "step": 33139 + }, + { + "epoch": 0.7294739159145939, + "grad_norm": 1.6125768682348662, + "learning_rate": 1.7995071876893233e-06, + "loss": 0.3061, + "step": 33140 + }, + { + "epoch": 0.7294959278010126, + "grad_norm": 1.4983097446240865, + "learning_rate": 1.799233332952085e-06, + "loss": 0.3171, + "step": 33141 + }, + { + "epoch": 0.7295179396874312, + "grad_norm": 1.357876998095965, + "learning_rate": 1.7989594944825706e-06, + "loss": 0.2814, + "step": 33142 + }, + { + "epoch": 0.7295399515738499, + "grad_norm": 1.4502502181422487, + "learning_rate": 1.7986856722821733e-06, + "loss": 0.3217, + "step": 33143 + }, + { + "epoch": 0.7295619634602686, + "grad_norm": 2.0164842529788842, + "learning_rate": 1.7984118663522832e-06, + "loss": 0.3113, + "step": 33144 + }, + { + "epoch": 0.7295839753466872, + "grad_norm": 1.4637849526927678, + "learning_rate": 1.7981380766942903e-06, + "loss": 0.2393, + "step": 33145 + }, + { + "epoch": 0.7296059872331059, + "grad_norm": 2.2894487687284126, + "learning_rate": 1.7978643033095878e-06, + "loss": 0.3242, + "step": 33146 + }, + { + "epoch": 0.7296279991195246, + "grad_norm": 1.6068990968370018, + "learning_rate": 1.7975905461995669e-06, + "loss": 0.3108, + "step": 33147 + }, + { + "epoch": 0.7296500110059432, + "grad_norm": 2.8393341952886852, + "learning_rate": 1.797316805365621e-06, + "loss": 0.3167, + "step": 33148 + }, + { + "epoch": 0.7296720228923619, + "grad_norm": 1.9563676850480793, + "learning_rate": 1.7970430808091393e-06, + "loss": 0.2065, + "step": 33149 + }, + { + "epoch": 0.7296940347787806, + "grad_norm": 1.8130616012062408, + "learning_rate": 1.796769372531512e-06, + "loss": 0.3373, + "step": 33150 + }, + { + "epoch": 0.7297160466651992, + "grad_norm": 1.3597074736073562, + "learning_rate": 1.796495680534131e-06, + "loss": 0.221, + "step": 33151 + }, + { + "epoch": 0.7297380585516179, + "grad_norm": 1.6825878367779612, + "learning_rate": 1.7962220048183882e-06, + "loss": 0.2653, + "step": 33152 + }, + { + "epoch": 0.7297600704380366, + "grad_norm": 0.6136547740265097, + "learning_rate": 1.7959483453856757e-06, + "loss": 0.4561, + "step": 33153 + }, + { + "epoch": 0.7297820823244552, + "grad_norm": 4.001110573805522, + "learning_rate": 1.7956747022373826e-06, + "loss": 0.2798, + "step": 33154 + }, + { + "epoch": 0.7298040942108739, + "grad_norm": 1.352054427499533, + "learning_rate": 1.795401075374898e-06, + "loss": 0.3158, + "step": 33155 + }, + { + "epoch": 0.7298261060972926, + "grad_norm": 2.0558164678409803, + "learning_rate": 1.7951274647996148e-06, + "loss": 0.3267, + "step": 33156 + }, + { + "epoch": 0.7298481179837112, + "grad_norm": 1.6480733270680301, + "learning_rate": 1.794853870512923e-06, + "loss": 0.3653, + "step": 33157 + }, + { + "epoch": 0.7298701298701299, + "grad_norm": 2.5323051845790854, + "learning_rate": 1.7945802925162147e-06, + "loss": 0.3234, + "step": 33158 + }, + { + "epoch": 0.7298921417565485, + "grad_norm": 2.210302300205369, + "learning_rate": 1.7943067308108796e-06, + "loss": 0.24, + "step": 33159 + }, + { + "epoch": 0.7299141536429672, + "grad_norm": 1.5337178939633205, + "learning_rate": 1.7940331853983057e-06, + "loss": 0.234, + "step": 33160 + }, + { + "epoch": 0.7299361655293859, + "grad_norm": 1.2961326955146137, + "learning_rate": 1.7937596562798847e-06, + "loss": 0.2411, + "step": 33161 + }, + { + "epoch": 0.7299581774158045, + "grad_norm": 1.4840763787529632, + "learning_rate": 1.7934861434570078e-06, + "loss": 0.3924, + "step": 33162 + }, + { + "epoch": 0.7299801893022232, + "grad_norm": 5.364598683373201, + "learning_rate": 1.7932126469310656e-06, + "loss": 0.3526, + "step": 33163 + }, + { + "epoch": 0.7300022011886419, + "grad_norm": 1.8056860927785632, + "learning_rate": 1.7929391667034469e-06, + "loss": 0.3309, + "step": 33164 + }, + { + "epoch": 0.7300242130750605, + "grad_norm": 0.6113200052954947, + "learning_rate": 1.7926657027755407e-06, + "loss": 0.449, + "step": 33165 + }, + { + "epoch": 0.7300462249614792, + "grad_norm": 1.6463941680928067, + "learning_rate": 1.7923922551487372e-06, + "loss": 0.3743, + "step": 33166 + }, + { + "epoch": 0.7300682368478979, + "grad_norm": 1.347621244189481, + "learning_rate": 1.7921188238244292e-06, + "loss": 0.2893, + "step": 33167 + }, + { + "epoch": 0.7300902487343165, + "grad_norm": 1.5379051962315489, + "learning_rate": 1.7918454088040017e-06, + "loss": 0.3009, + "step": 33168 + }, + { + "epoch": 0.7301122606207352, + "grad_norm": 0.5964259318069698, + "learning_rate": 1.791572010088849e-06, + "loss": 0.4372, + "step": 33169 + }, + { + "epoch": 0.7301342725071539, + "grad_norm": 1.6142862441504209, + "learning_rate": 1.7912986276803562e-06, + "loss": 0.3738, + "step": 33170 + }, + { + "epoch": 0.7301562843935725, + "grad_norm": 1.8648390068976293, + "learning_rate": 1.7910252615799168e-06, + "loss": 0.3954, + "step": 33171 + }, + { + "epoch": 0.7301782962799912, + "grad_norm": 1.7739093308768847, + "learning_rate": 1.7907519117889165e-06, + "loss": 0.3247, + "step": 33172 + }, + { + "epoch": 0.7302003081664099, + "grad_norm": 0.6242956769218141, + "learning_rate": 1.7904785783087463e-06, + "loss": 0.4744, + "step": 33173 + }, + { + "epoch": 0.7302223200528285, + "grad_norm": 1.6069699109833957, + "learning_rate": 1.7902052611407978e-06, + "loss": 0.2706, + "step": 33174 + }, + { + "epoch": 0.7302443319392472, + "grad_norm": 1.7546541909930045, + "learning_rate": 1.789931960286455e-06, + "loss": 0.2601, + "step": 33175 + }, + { + "epoch": 0.7302663438256659, + "grad_norm": 1.212476105536327, + "learning_rate": 1.7896586757471124e-06, + "loss": 0.2246, + "step": 33176 + }, + { + "epoch": 0.7302883557120845, + "grad_norm": 12.149344229793101, + "learning_rate": 1.789385407524154e-06, + "loss": 0.3267, + "step": 33177 + }, + { + "epoch": 0.7303103675985032, + "grad_norm": 0.6404908496460318, + "learning_rate": 1.7891121556189711e-06, + "loss": 0.4372, + "step": 33178 + }, + { + "epoch": 0.7303323794849219, + "grad_norm": 7.4455213484471505, + "learning_rate": 1.7888389200329537e-06, + "loss": 0.2678, + "step": 33179 + }, + { + "epoch": 0.7303543913713405, + "grad_norm": 0.6412299909738977, + "learning_rate": 1.7885657007674879e-06, + "loss": 0.4736, + "step": 33180 + }, + { + "epoch": 0.7303764032577592, + "grad_norm": 1.9289673599159813, + "learning_rate": 1.7882924978239652e-06, + "loss": 0.3282, + "step": 33181 + }, + { + "epoch": 0.7303984151441778, + "grad_norm": 1.5795574768204845, + "learning_rate": 1.7880193112037702e-06, + "loss": 0.2966, + "step": 33182 + }, + { + "epoch": 0.7304204270305965, + "grad_norm": 1.5791769776034783, + "learning_rate": 1.7877461409082947e-06, + "loss": 0.2708, + "step": 33183 + }, + { + "epoch": 0.7304424389170152, + "grad_norm": 0.6511620459226231, + "learning_rate": 1.7874729869389268e-06, + "loss": 0.4777, + "step": 33184 + }, + { + "epoch": 0.7304644508034338, + "grad_norm": 1.9520470180671499, + "learning_rate": 1.7871998492970545e-06, + "loss": 0.2505, + "step": 33185 + }, + { + "epoch": 0.7304864626898525, + "grad_norm": 1.8393372194897466, + "learning_rate": 1.7869267279840635e-06, + "loss": 0.2673, + "step": 33186 + }, + { + "epoch": 0.7305084745762712, + "grad_norm": 1.953568003544879, + "learning_rate": 1.7866536230013442e-06, + "loss": 0.3255, + "step": 33187 + }, + { + "epoch": 0.7305304864626898, + "grad_norm": 1.194776998284199, + "learning_rate": 1.7863805343502838e-06, + "loss": 0.213, + "step": 33188 + }, + { + "epoch": 0.7305524983491085, + "grad_norm": 1.5327152526303807, + "learning_rate": 1.7861074620322733e-06, + "loss": 0.2409, + "step": 33189 + }, + { + "epoch": 0.7305745102355272, + "grad_norm": 0.6253443016287324, + "learning_rate": 1.7858344060486976e-06, + "loss": 0.4689, + "step": 33190 + }, + { + "epoch": 0.7305965221219458, + "grad_norm": 1.3123475163092808, + "learning_rate": 1.7855613664009436e-06, + "loss": 0.3, + "step": 33191 + }, + { + "epoch": 0.7306185340083645, + "grad_norm": 63.987492994201176, + "learning_rate": 1.7852883430904005e-06, + "loss": 0.3474, + "step": 33192 + }, + { + "epoch": 0.7306405458947832, + "grad_norm": 1.5559564759867064, + "learning_rate": 1.7850153361184552e-06, + "loss": 0.3407, + "step": 33193 + }, + { + "epoch": 0.7306625577812018, + "grad_norm": 1.7986591965421423, + "learning_rate": 1.7847423454864983e-06, + "loss": 0.4124, + "step": 33194 + }, + { + "epoch": 0.7306845696676205, + "grad_norm": 1.554805925874297, + "learning_rate": 1.784469371195915e-06, + "loss": 0.2911, + "step": 33195 + }, + { + "epoch": 0.7307065815540392, + "grad_norm": 1.2846064492925304, + "learning_rate": 1.7841964132480898e-06, + "loss": 0.259, + "step": 33196 + }, + { + "epoch": 0.7307285934404578, + "grad_norm": 1.376976409320704, + "learning_rate": 1.7839234716444137e-06, + "loss": 0.2919, + "step": 33197 + }, + { + "epoch": 0.7307506053268765, + "grad_norm": 1.4759557729390513, + "learning_rate": 1.7836505463862724e-06, + "loss": 0.2751, + "step": 33198 + }, + { + "epoch": 0.7307726172132952, + "grad_norm": 1.6787067560589923, + "learning_rate": 1.7833776374750555e-06, + "loss": 0.337, + "step": 33199 + }, + { + "epoch": 0.7307946290997138, + "grad_norm": 19.121126337318433, + "learning_rate": 1.783104744912148e-06, + "loss": 0.2893, + "step": 33200 + }, + { + "epoch": 0.7308166409861325, + "grad_norm": 1.3697580463171857, + "learning_rate": 1.7828318686989355e-06, + "loss": 0.2922, + "step": 33201 + }, + { + "epoch": 0.7308386528725512, + "grad_norm": 3.2058013477305876, + "learning_rate": 1.7825590088368061e-06, + "loss": 0.3088, + "step": 33202 + }, + { + "epoch": 0.7308606647589698, + "grad_norm": 2.2881145443317568, + "learning_rate": 1.7822861653271467e-06, + "loss": 0.3153, + "step": 33203 + }, + { + "epoch": 0.7308826766453885, + "grad_norm": 1.6144250195323528, + "learning_rate": 1.7820133381713462e-06, + "loss": 0.2497, + "step": 33204 + }, + { + "epoch": 0.7309046885318071, + "grad_norm": 1.551762188582355, + "learning_rate": 1.7817405273707889e-06, + "loss": 0.3649, + "step": 33205 + }, + { + "epoch": 0.7309267004182258, + "grad_norm": 3.607895078269006, + "learning_rate": 1.7814677329268597e-06, + "loss": 0.3042, + "step": 33206 + }, + { + "epoch": 0.7309487123046445, + "grad_norm": 1.5937109568952978, + "learning_rate": 1.781194954840949e-06, + "loss": 0.2611, + "step": 33207 + }, + { + "epoch": 0.7309707241910631, + "grad_norm": 5.321383844605411, + "learning_rate": 1.7809221931144393e-06, + "loss": 0.2584, + "step": 33208 + }, + { + "epoch": 0.7309927360774818, + "grad_norm": 43.24865049190715, + "learning_rate": 1.7806494477487192e-06, + "loss": 0.307, + "step": 33209 + }, + { + "epoch": 0.7310147479639005, + "grad_norm": 2.243745638673412, + "learning_rate": 1.780376718745176e-06, + "loss": 0.275, + "step": 33210 + }, + { + "epoch": 0.7310367598503191, + "grad_norm": 0.6575022048846011, + "learning_rate": 1.7801040061051922e-06, + "loss": 0.4757, + "step": 33211 + }, + { + "epoch": 0.7310587717367378, + "grad_norm": 1.9152721692259627, + "learning_rate": 1.7798313098301578e-06, + "loss": 0.2838, + "step": 33212 + }, + { + "epoch": 0.7310807836231565, + "grad_norm": 1.3145317979196063, + "learning_rate": 1.7795586299214556e-06, + "loss": 0.2267, + "step": 33213 + }, + { + "epoch": 0.7311027955095751, + "grad_norm": 0.6058373784462836, + "learning_rate": 1.7792859663804723e-06, + "loss": 0.4323, + "step": 33214 + }, + { + "epoch": 0.7311248073959938, + "grad_norm": 0.6326472148700967, + "learning_rate": 1.7790133192085962e-06, + "loss": 0.4437, + "step": 33215 + }, + { + "epoch": 0.7311468192824125, + "grad_norm": 1.2529514497285588, + "learning_rate": 1.7787406884072094e-06, + "loss": 0.2435, + "step": 33216 + }, + { + "epoch": 0.7311688311688311, + "grad_norm": 1.9496270009955468, + "learning_rate": 1.778468073977701e-06, + "loss": 0.3189, + "step": 33217 + }, + { + "epoch": 0.7311908430552498, + "grad_norm": 2.0092355420151486, + "learning_rate": 1.7781954759214526e-06, + "loss": 0.3276, + "step": 33218 + }, + { + "epoch": 0.7312128549416685, + "grad_norm": 1.8550129645018623, + "learning_rate": 1.7779228942398518e-06, + "loss": 0.3597, + "step": 33219 + }, + { + "epoch": 0.7312348668280871, + "grad_norm": 1.4704218617208618, + "learning_rate": 1.7776503289342856e-06, + "loss": 0.3539, + "step": 33220 + }, + { + "epoch": 0.7312568787145058, + "grad_norm": 0.6017280174756464, + "learning_rate": 1.7773777800061365e-06, + "loss": 0.4278, + "step": 33221 + }, + { + "epoch": 0.7312788906009245, + "grad_norm": 0.6080942538318654, + "learning_rate": 1.7771052474567917e-06, + "loss": 0.4431, + "step": 33222 + }, + { + "epoch": 0.7313009024873431, + "grad_norm": 1.2730917711050354, + "learning_rate": 1.7768327312876343e-06, + "loss": 0.2861, + "step": 33223 + }, + { + "epoch": 0.7313229143737618, + "grad_norm": 2.9804004407004148, + "learning_rate": 1.7765602315000503e-06, + "loss": 0.2784, + "step": 33224 + }, + { + "epoch": 0.7313449262601806, + "grad_norm": 1.2169846329219944, + "learning_rate": 1.7762877480954266e-06, + "loss": 0.2588, + "step": 33225 + }, + { + "epoch": 0.7313669381465991, + "grad_norm": 1.700664536153035, + "learning_rate": 1.776015281075147e-06, + "loss": 0.2485, + "step": 33226 + }, + { + "epoch": 0.7313889500330178, + "grad_norm": 0.5893495986007876, + "learning_rate": 1.7757428304405932e-06, + "loss": 0.4523, + "step": 33227 + }, + { + "epoch": 0.7314109619194364, + "grad_norm": 1.7350671193319906, + "learning_rate": 1.7754703961931525e-06, + "loss": 0.3793, + "step": 33228 + }, + { + "epoch": 0.7314329738058551, + "grad_norm": 2.0206122132735294, + "learning_rate": 1.77519797833421e-06, + "loss": 0.2998, + "step": 33229 + }, + { + "epoch": 0.7314549856922739, + "grad_norm": 0.6053770869137115, + "learning_rate": 1.7749255768651507e-06, + "loss": 0.4444, + "step": 33230 + }, + { + "epoch": 0.7314769975786924, + "grad_norm": 0.5866389681165237, + "learning_rate": 1.7746531917873583e-06, + "loss": 0.4423, + "step": 33231 + }, + { + "epoch": 0.7314990094651111, + "grad_norm": 0.6295201933412622, + "learning_rate": 1.7743808231022152e-06, + "loss": 0.4568, + "step": 33232 + }, + { + "epoch": 0.7315210213515299, + "grad_norm": 1.8657100270061002, + "learning_rate": 1.7741084708111072e-06, + "loss": 0.3413, + "step": 33233 + }, + { + "epoch": 0.7315430332379484, + "grad_norm": 1.5197824850035209, + "learning_rate": 1.7738361349154193e-06, + "loss": 0.3273, + "step": 33234 + }, + { + "epoch": 0.7315650451243672, + "grad_norm": 1.795248509867856, + "learning_rate": 1.7735638154165363e-06, + "loss": 0.3784, + "step": 33235 + }, + { + "epoch": 0.7315870570107859, + "grad_norm": 1.5719724564898871, + "learning_rate": 1.7732915123158412e-06, + "loss": 0.2846, + "step": 33236 + }, + { + "epoch": 0.7316090688972045, + "grad_norm": 1.5553652919545438, + "learning_rate": 1.7730192256147161e-06, + "loss": 0.279, + "step": 33237 + }, + { + "epoch": 0.7316310807836232, + "grad_norm": 0.6217639103594361, + "learning_rate": 1.7727469553145465e-06, + "loss": 0.4435, + "step": 33238 + }, + { + "epoch": 0.7316530926700419, + "grad_norm": 1.7074140390393, + "learning_rate": 1.7724747014167165e-06, + "loss": 0.2896, + "step": 33239 + }, + { + "epoch": 0.7316751045564605, + "grad_norm": 0.6126036423764605, + "learning_rate": 1.772202463922611e-06, + "loss": 0.4489, + "step": 33240 + }, + { + "epoch": 0.7316971164428792, + "grad_norm": 1.5372822105767132, + "learning_rate": 1.7719302428336127e-06, + "loss": 0.2869, + "step": 33241 + }, + { + "epoch": 0.7317191283292979, + "grad_norm": 1.6864619342307545, + "learning_rate": 1.7716580381511024e-06, + "loss": 0.2618, + "step": 33242 + }, + { + "epoch": 0.7317411402157165, + "grad_norm": 2.0688702712682954, + "learning_rate": 1.7713858498764675e-06, + "loss": 0.321, + "step": 33243 + }, + { + "epoch": 0.7317631521021352, + "grad_norm": 1.4968420804502114, + "learning_rate": 1.7711136780110882e-06, + "loss": 0.275, + "step": 33244 + }, + { + "epoch": 0.7317851639885539, + "grad_norm": 2.4727860741024963, + "learning_rate": 1.7708415225563496e-06, + "loss": 0.2998, + "step": 33245 + }, + { + "epoch": 0.7318071758749725, + "grad_norm": 8.134072988146187, + "learning_rate": 1.7705693835136362e-06, + "loss": 0.3139, + "step": 33246 + }, + { + "epoch": 0.7318291877613912, + "grad_norm": 1.595615756260418, + "learning_rate": 1.7702972608843278e-06, + "loss": 0.2147, + "step": 33247 + }, + { + "epoch": 0.7318511996478099, + "grad_norm": 1.207927610538363, + "learning_rate": 1.7700251546698105e-06, + "loss": 0.2336, + "step": 33248 + }, + { + "epoch": 0.7318732115342285, + "grad_norm": 1.6758792574778074, + "learning_rate": 1.7697530648714645e-06, + "loss": 0.2295, + "step": 33249 + }, + { + "epoch": 0.7318952234206472, + "grad_norm": 2.1927236163108637, + "learning_rate": 1.769480991490674e-06, + "loss": 0.2796, + "step": 33250 + }, + { + "epoch": 0.7319172353070658, + "grad_norm": 1.2499083930284893, + "learning_rate": 1.7692089345288244e-06, + "loss": 0.2473, + "step": 33251 + }, + { + "epoch": 0.7319392471934845, + "grad_norm": 1.9882770455413097, + "learning_rate": 1.7689368939872936e-06, + "loss": 0.3006, + "step": 33252 + }, + { + "epoch": 0.7319612590799032, + "grad_norm": 1.2342569012699476, + "learning_rate": 1.7686648698674685e-06, + "loss": 0.236, + "step": 33253 + }, + { + "epoch": 0.7319832709663218, + "grad_norm": 0.6244946221254164, + "learning_rate": 1.768392862170728e-06, + "loss": 0.4548, + "step": 33254 + }, + { + "epoch": 0.7320052828527405, + "grad_norm": 1.8273612655019826, + "learning_rate": 1.7681208708984565e-06, + "loss": 0.2521, + "step": 33255 + }, + { + "epoch": 0.7320272947391592, + "grad_norm": 1.96588000283834, + "learning_rate": 1.7678488960520374e-06, + "loss": 0.3254, + "step": 33256 + }, + { + "epoch": 0.7320493066255778, + "grad_norm": 1.5195779495665993, + "learning_rate": 1.7675769376328506e-06, + "loss": 0.3878, + "step": 33257 + }, + { + "epoch": 0.7320713185119965, + "grad_norm": 1.5703894192558987, + "learning_rate": 1.7673049956422812e-06, + "loss": 0.3284, + "step": 33258 + }, + { + "epoch": 0.7320933303984152, + "grad_norm": 1.6238562720931642, + "learning_rate": 1.7670330700817074e-06, + "loss": 0.3592, + "step": 33259 + }, + { + "epoch": 0.7321153422848338, + "grad_norm": 2.042276610688939, + "learning_rate": 1.7667611609525143e-06, + "loss": 0.235, + "step": 33260 + }, + { + "epoch": 0.7321373541712525, + "grad_norm": 1.8109311859543287, + "learning_rate": 1.766489268256084e-06, + "loss": 0.2588, + "step": 33261 + }, + { + "epoch": 0.7321593660576712, + "grad_norm": 1.6276503448206816, + "learning_rate": 1.7662173919937958e-06, + "loss": 0.275, + "step": 33262 + }, + { + "epoch": 0.7321813779440898, + "grad_norm": 1.834692056272583, + "learning_rate": 1.7659455321670354e-06, + "loss": 0.3162, + "step": 33263 + }, + { + "epoch": 0.7322033898305085, + "grad_norm": 0.6160226028014495, + "learning_rate": 1.7656736887771802e-06, + "loss": 0.44, + "step": 33264 + }, + { + "epoch": 0.7322254017169272, + "grad_norm": 1.6446449053774008, + "learning_rate": 1.7654018618256135e-06, + "loss": 0.2732, + "step": 33265 + }, + { + "epoch": 0.7322474136033458, + "grad_norm": 1.8372506493578382, + "learning_rate": 1.7651300513137193e-06, + "loss": 0.25, + "step": 33266 + }, + { + "epoch": 0.7322694254897645, + "grad_norm": 2.281745694458909, + "learning_rate": 1.7648582572428769e-06, + "loss": 0.2516, + "step": 33267 + }, + { + "epoch": 0.7322914373761832, + "grad_norm": 1.9476416505133292, + "learning_rate": 1.7645864796144663e-06, + "loss": 0.3687, + "step": 33268 + }, + { + "epoch": 0.7323134492626018, + "grad_norm": 1.461593362027355, + "learning_rate": 1.7643147184298699e-06, + "loss": 0.2872, + "step": 33269 + }, + { + "epoch": 0.7323354611490205, + "grad_norm": 1.5590469932716877, + "learning_rate": 1.764042973690469e-06, + "loss": 0.2858, + "step": 33270 + }, + { + "epoch": 0.7323574730354392, + "grad_norm": 1.502115848531171, + "learning_rate": 1.7637712453976474e-06, + "loss": 0.2652, + "step": 33271 + }, + { + "epoch": 0.7323794849218578, + "grad_norm": 2.2453104844642215, + "learning_rate": 1.7634995335527833e-06, + "loss": 0.2672, + "step": 33272 + }, + { + "epoch": 0.7324014968082765, + "grad_norm": 1.5343360912438453, + "learning_rate": 1.7632278381572564e-06, + "loss": 0.2489, + "step": 33273 + }, + { + "epoch": 0.7324235086946951, + "grad_norm": 1.634833407362459, + "learning_rate": 1.7629561592124494e-06, + "loss": 0.2793, + "step": 33274 + }, + { + "epoch": 0.7324455205811138, + "grad_norm": 1.406878708601021, + "learning_rate": 1.7626844967197431e-06, + "loss": 0.2545, + "step": 33275 + }, + { + "epoch": 0.7324675324675325, + "grad_norm": 1.4120700097382783, + "learning_rate": 1.7624128506805193e-06, + "loss": 0.2307, + "step": 33276 + }, + { + "epoch": 0.7324895443539511, + "grad_norm": 1.4998765510224268, + "learning_rate": 1.7621412210961575e-06, + "loss": 0.3507, + "step": 33277 + }, + { + "epoch": 0.7325115562403698, + "grad_norm": 1.5664541367458706, + "learning_rate": 1.7618696079680369e-06, + "loss": 0.2062, + "step": 33278 + }, + { + "epoch": 0.7325335681267885, + "grad_norm": 1.6450635718223616, + "learning_rate": 1.7615980112975405e-06, + "loss": 0.4037, + "step": 33279 + }, + { + "epoch": 0.7325555800132071, + "grad_norm": 0.6229900943441696, + "learning_rate": 1.761326431086044e-06, + "loss": 0.4793, + "step": 33280 + }, + { + "epoch": 0.7325775918996258, + "grad_norm": 2.5241041641954474, + "learning_rate": 1.7610548673349349e-06, + "loss": 0.269, + "step": 33281 + }, + { + "epoch": 0.7325996037860445, + "grad_norm": 1.2622095198808374, + "learning_rate": 1.7607833200455888e-06, + "loss": 0.1821, + "step": 33282 + }, + { + "epoch": 0.7326216156724631, + "grad_norm": 1.26583585330135, + "learning_rate": 1.7605117892193858e-06, + "loss": 0.229, + "step": 33283 + }, + { + "epoch": 0.7326436275588818, + "grad_norm": 1.6947150477276731, + "learning_rate": 1.7602402748577073e-06, + "loss": 0.3216, + "step": 33284 + }, + { + "epoch": 0.7326656394453005, + "grad_norm": 1.6847876742786494, + "learning_rate": 1.7599687769619316e-06, + "loss": 0.2881, + "step": 33285 + }, + { + "epoch": 0.7326876513317191, + "grad_norm": 1.354967722666188, + "learning_rate": 1.75969729553344e-06, + "loss": 0.2599, + "step": 33286 + }, + { + "epoch": 0.7327096632181378, + "grad_norm": 13.526659551669935, + "learning_rate": 1.7594258305736134e-06, + "loss": 0.2218, + "step": 33287 + }, + { + "epoch": 0.7327316751045565, + "grad_norm": 0.6047285608767872, + "learning_rate": 1.7591543820838287e-06, + "loss": 0.4209, + "step": 33288 + }, + { + "epoch": 0.7327536869909751, + "grad_norm": 1.4864079174229579, + "learning_rate": 1.758882950065468e-06, + "loss": 0.343, + "step": 33289 + }, + { + "epoch": 0.7327756988773938, + "grad_norm": 0.6007930874007158, + "learning_rate": 1.7586115345199085e-06, + "loss": 0.4618, + "step": 33290 + }, + { + "epoch": 0.7327977107638125, + "grad_norm": 0.6238632194865394, + "learning_rate": 1.7583401354485302e-06, + "loss": 0.4714, + "step": 33291 + }, + { + "epoch": 0.7328197226502311, + "grad_norm": 1.326478388614571, + "learning_rate": 1.7580687528527152e-06, + "loss": 0.2476, + "step": 33292 + }, + { + "epoch": 0.7328417345366498, + "grad_norm": 1.5141993315100977, + "learning_rate": 1.7577973867338395e-06, + "loss": 0.2016, + "step": 33293 + }, + { + "epoch": 0.7328637464230685, + "grad_norm": 1.4140928496065732, + "learning_rate": 1.7575260370932846e-06, + "loss": 0.2478, + "step": 33294 + }, + { + "epoch": 0.7328857583094871, + "grad_norm": 1.9266593119953814, + "learning_rate": 1.7572547039324272e-06, + "loss": 0.3224, + "step": 33295 + }, + { + "epoch": 0.7329077701959058, + "grad_norm": 1.4604178502396186, + "learning_rate": 1.7569833872526475e-06, + "loss": 0.2957, + "step": 33296 + }, + { + "epoch": 0.7329297820823245, + "grad_norm": 1.6391939345073718, + "learning_rate": 1.7567120870553267e-06, + "loss": 0.3636, + "step": 33297 + }, + { + "epoch": 0.7329517939687431, + "grad_norm": 1.8523940379941366, + "learning_rate": 1.7564408033418395e-06, + "loss": 0.2923, + "step": 33298 + }, + { + "epoch": 0.7329738058551618, + "grad_norm": 1.3312360880775391, + "learning_rate": 1.756169536113569e-06, + "loss": 0.2718, + "step": 33299 + }, + { + "epoch": 0.7329958177415804, + "grad_norm": 1.8942968998782272, + "learning_rate": 1.7558982853718898e-06, + "loss": 0.2944, + "step": 33300 + }, + { + "epoch": 0.7330178296279991, + "grad_norm": 1.8464606581869079, + "learning_rate": 1.7556270511181822e-06, + "loss": 0.4132, + "step": 33301 + }, + { + "epoch": 0.7330398415144178, + "grad_norm": 2.003098571457305, + "learning_rate": 1.7553558333538268e-06, + "loss": 0.2777, + "step": 33302 + }, + { + "epoch": 0.7330618534008364, + "grad_norm": 2.658078490445098, + "learning_rate": 1.7550846320801984e-06, + "loss": 0.3765, + "step": 33303 + }, + { + "epoch": 0.7330838652872551, + "grad_norm": 1.677317406965363, + "learning_rate": 1.754813447298679e-06, + "loss": 0.3155, + "step": 33304 + }, + { + "epoch": 0.7331058771736738, + "grad_norm": 2.0313591249690273, + "learning_rate": 1.754542279010643e-06, + "loss": 0.2448, + "step": 33305 + }, + { + "epoch": 0.7331278890600924, + "grad_norm": 1.5314237313966081, + "learning_rate": 1.7542711272174711e-06, + "loss": 0.277, + "step": 33306 + }, + { + "epoch": 0.7331499009465111, + "grad_norm": 1.7042478904187508, + "learning_rate": 1.753999991920543e-06, + "loss": 0.251, + "step": 33307 + }, + { + "epoch": 0.7331719128329298, + "grad_norm": 0.6021538149627172, + "learning_rate": 1.7537288731212342e-06, + "loss": 0.4475, + "step": 33308 + }, + { + "epoch": 0.7331939247193484, + "grad_norm": 1.3435531314255964, + "learning_rate": 1.753457770820921e-06, + "loss": 0.2978, + "step": 33309 + }, + { + "epoch": 0.7332159366057671, + "grad_norm": 2.22831605314552, + "learning_rate": 1.7531866850209862e-06, + "loss": 0.2268, + "step": 33310 + }, + { + "epoch": 0.7332379484921858, + "grad_norm": 1.7456036173547012, + "learning_rate": 1.7529156157228005e-06, + "loss": 0.2687, + "step": 33311 + }, + { + "epoch": 0.7332599603786044, + "grad_norm": 1.9225919025457248, + "learning_rate": 1.7526445629277495e-06, + "loss": 0.3365, + "step": 33312 + }, + { + "epoch": 0.7332819722650231, + "grad_norm": 2.8108838939608973, + "learning_rate": 1.7523735266372073e-06, + "loss": 0.2855, + "step": 33313 + }, + { + "epoch": 0.7333039841514418, + "grad_norm": 0.6381698665935288, + "learning_rate": 1.7521025068525494e-06, + "loss": 0.4718, + "step": 33314 + }, + { + "epoch": 0.7333259960378604, + "grad_norm": 2.0705496417689777, + "learning_rate": 1.7518315035751571e-06, + "loss": 0.2509, + "step": 33315 + }, + { + "epoch": 0.7333480079242791, + "grad_norm": 1.424560209280266, + "learning_rate": 1.7515605168064027e-06, + "loss": 0.3161, + "step": 33316 + }, + { + "epoch": 0.7333700198106978, + "grad_norm": 0.6091010304485573, + "learning_rate": 1.7512895465476693e-06, + "loss": 0.4675, + "step": 33317 + }, + { + "epoch": 0.7333920316971164, + "grad_norm": 0.6141097856924498, + "learning_rate": 1.751018592800332e-06, + "loss": 0.4377, + "step": 33318 + }, + { + "epoch": 0.7334140435835351, + "grad_norm": 1.349480016426404, + "learning_rate": 1.7507476555657655e-06, + "loss": 0.2835, + "step": 33319 + }, + { + "epoch": 0.7334360554699538, + "grad_norm": 0.5966925627024108, + "learning_rate": 1.75047673484535e-06, + "loss": 0.4593, + "step": 33320 + }, + { + "epoch": 0.7334580673563724, + "grad_norm": 1.7942715652968124, + "learning_rate": 1.7502058306404596e-06, + "loss": 0.253, + "step": 33321 + }, + { + "epoch": 0.7334800792427911, + "grad_norm": 2.383548249258614, + "learning_rate": 1.7499349429524725e-06, + "loss": 0.3485, + "step": 33322 + }, + { + "epoch": 0.7335020911292097, + "grad_norm": 3.594928746345921, + "learning_rate": 1.7496640717827672e-06, + "loss": 0.3654, + "step": 33323 + }, + { + "epoch": 0.7335241030156284, + "grad_norm": 2.861139928955292, + "learning_rate": 1.7493932171327171e-06, + "loss": 0.2626, + "step": 33324 + }, + { + "epoch": 0.7335461149020471, + "grad_norm": 1.5480499917386568, + "learning_rate": 1.7491223790037027e-06, + "loss": 0.2627, + "step": 33325 + }, + { + "epoch": 0.7335681267884657, + "grad_norm": 2.1409814610065507, + "learning_rate": 1.748851557397096e-06, + "loss": 0.3366, + "step": 33326 + }, + { + "epoch": 0.7335901386748844, + "grad_norm": 3.801286545528464, + "learning_rate": 1.7485807523142763e-06, + "loss": 0.2965, + "step": 33327 + }, + { + "epoch": 0.7336121505613031, + "grad_norm": 1.9332544080593512, + "learning_rate": 1.7483099637566203e-06, + "loss": 0.3361, + "step": 33328 + }, + { + "epoch": 0.7336341624477217, + "grad_norm": 1.6257384948629954, + "learning_rate": 1.7480391917255024e-06, + "loss": 0.3062, + "step": 33329 + }, + { + "epoch": 0.7336561743341404, + "grad_norm": 1.489988188050215, + "learning_rate": 1.7477684362223007e-06, + "loss": 0.3364, + "step": 33330 + }, + { + "epoch": 0.7336781862205591, + "grad_norm": 1.6245340385170635, + "learning_rate": 1.7474976972483892e-06, + "loss": 0.2642, + "step": 33331 + }, + { + "epoch": 0.7337001981069777, + "grad_norm": 1.4874691507684883, + "learning_rate": 1.747226974805145e-06, + "loss": 0.341, + "step": 33332 + }, + { + "epoch": 0.7337222099933964, + "grad_norm": 1.9866668714009217, + "learning_rate": 1.7469562688939456e-06, + "loss": 0.2865, + "step": 33333 + }, + { + "epoch": 0.7337442218798151, + "grad_norm": 3.904361334738041, + "learning_rate": 1.7466855795161641e-06, + "loss": 0.2839, + "step": 33334 + }, + { + "epoch": 0.7337662337662337, + "grad_norm": 0.6232925038856919, + "learning_rate": 1.7464149066731784e-06, + "loss": 0.4697, + "step": 33335 + }, + { + "epoch": 0.7337882456526524, + "grad_norm": 1.6248330129804833, + "learning_rate": 1.7461442503663622e-06, + "loss": 0.3679, + "step": 33336 + }, + { + "epoch": 0.7338102575390711, + "grad_norm": 1.4256246864869102, + "learning_rate": 1.745873610597092e-06, + "loss": 0.302, + "step": 33337 + }, + { + "epoch": 0.7338322694254897, + "grad_norm": 0.6134636279764276, + "learning_rate": 1.7456029873667451e-06, + "loss": 0.4586, + "step": 33338 + }, + { + "epoch": 0.7338542813119084, + "grad_norm": 1.8101579852225602, + "learning_rate": 1.7453323806766942e-06, + "loss": 0.3256, + "step": 33339 + }, + { + "epoch": 0.7338762931983271, + "grad_norm": 1.4538382515566577, + "learning_rate": 1.7450617905283168e-06, + "loss": 0.2376, + "step": 33340 + }, + { + "epoch": 0.7338983050847457, + "grad_norm": 1.3990821191605005, + "learning_rate": 1.7447912169229859e-06, + "loss": 0.2698, + "step": 33341 + }, + { + "epoch": 0.7339203169711644, + "grad_norm": 2.028448990318678, + "learning_rate": 1.7445206598620774e-06, + "loss": 0.3455, + "step": 33342 + }, + { + "epoch": 0.7339423288575831, + "grad_norm": 1.4591939644023921, + "learning_rate": 1.7442501193469686e-06, + "loss": 0.2693, + "step": 33343 + }, + { + "epoch": 0.7339643407440017, + "grad_norm": 1.631413654092689, + "learning_rate": 1.7439795953790334e-06, + "loss": 0.3989, + "step": 33344 + }, + { + "epoch": 0.7339863526304204, + "grad_norm": 1.5054008297496204, + "learning_rate": 1.7437090879596436e-06, + "loss": 0.2701, + "step": 33345 + }, + { + "epoch": 0.734008364516839, + "grad_norm": 1.551170301913244, + "learning_rate": 1.743438597090179e-06, + "loss": 0.2735, + "step": 33346 + }, + { + "epoch": 0.7340303764032577, + "grad_norm": 1.4070127910640815, + "learning_rate": 1.7431681227720081e-06, + "loss": 0.2989, + "step": 33347 + }, + { + "epoch": 0.7340523882896764, + "grad_norm": 1.9488513509060588, + "learning_rate": 1.7428976650065126e-06, + "loss": 0.3117, + "step": 33348 + }, + { + "epoch": 0.734074400176095, + "grad_norm": 1.5273735601394354, + "learning_rate": 1.7426272237950647e-06, + "loss": 0.2901, + "step": 33349 + }, + { + "epoch": 0.7340964120625137, + "grad_norm": 1.9952192415972487, + "learning_rate": 1.7423567991390355e-06, + "loss": 0.2593, + "step": 33350 + }, + { + "epoch": 0.7341184239489325, + "grad_norm": 1.3616363057630125, + "learning_rate": 1.7420863910398045e-06, + "loss": 0.2737, + "step": 33351 + }, + { + "epoch": 0.734140435835351, + "grad_norm": 1.905277326201341, + "learning_rate": 1.7418159994987395e-06, + "loss": 0.4032, + "step": 33352 + }, + { + "epoch": 0.7341624477217698, + "grad_norm": 12.650663397280901, + "learning_rate": 1.7415456245172223e-06, + "loss": 0.225, + "step": 33353 + }, + { + "epoch": 0.7341844596081885, + "grad_norm": 1.6167999559983741, + "learning_rate": 1.7412752660966236e-06, + "loss": 0.2753, + "step": 33354 + }, + { + "epoch": 0.734206471494607, + "grad_norm": 1.3871400063771377, + "learning_rate": 1.741004924238316e-06, + "loss": 0.2277, + "step": 33355 + }, + { + "epoch": 0.7342284833810258, + "grad_norm": 1.8339298309113725, + "learning_rate": 1.7407345989436763e-06, + "loss": 0.3215, + "step": 33356 + }, + { + "epoch": 0.7342504952674445, + "grad_norm": 3.436775851148634, + "learning_rate": 1.740464290214075e-06, + "loss": 0.3153, + "step": 33357 + }, + { + "epoch": 0.734272507153863, + "grad_norm": 1.5393871811307416, + "learning_rate": 1.7401939980508886e-06, + "loss": 0.3285, + "step": 33358 + }, + { + "epoch": 0.7342945190402818, + "grad_norm": 1.342931020625959, + "learning_rate": 1.7399237224554915e-06, + "loss": 0.2052, + "step": 33359 + }, + { + "epoch": 0.7343165309267005, + "grad_norm": 1.5142868850282702, + "learning_rate": 1.739653463429255e-06, + "loss": 0.2509, + "step": 33360 + }, + { + "epoch": 0.734338542813119, + "grad_norm": 2.189419910635468, + "learning_rate": 1.739383220973555e-06, + "loss": 0.2301, + "step": 33361 + }, + { + "epoch": 0.7343605546995378, + "grad_norm": 2.6098299531710767, + "learning_rate": 1.7391129950897618e-06, + "loss": 0.3127, + "step": 33362 + }, + { + "epoch": 0.7343825665859565, + "grad_norm": 2.3055743119849534, + "learning_rate": 1.7388427857792511e-06, + "loss": 0.3025, + "step": 33363 + }, + { + "epoch": 0.7344045784723751, + "grad_norm": 1.7165515361078423, + "learning_rate": 1.7385725930433977e-06, + "loss": 0.304, + "step": 33364 + }, + { + "epoch": 0.7344265903587938, + "grad_norm": 2.0732324886460267, + "learning_rate": 1.738302416883571e-06, + "loss": 0.2904, + "step": 33365 + }, + { + "epoch": 0.7344486022452125, + "grad_norm": 1.9732105608490298, + "learning_rate": 1.7380322573011483e-06, + "loss": 0.2613, + "step": 33366 + }, + { + "epoch": 0.7344706141316311, + "grad_norm": 1.3666375640197113, + "learning_rate": 1.7377621142974982e-06, + "loss": 0.3159, + "step": 33367 + }, + { + "epoch": 0.7344926260180498, + "grad_norm": 1.6769181253475565, + "learning_rate": 1.7374919878739965e-06, + "loss": 0.2504, + "step": 33368 + }, + { + "epoch": 0.7345146379044684, + "grad_norm": 11.050875833803994, + "learning_rate": 1.7372218780320172e-06, + "loss": 0.271, + "step": 33369 + }, + { + "epoch": 0.7345366497908871, + "grad_norm": 1.9927605578566492, + "learning_rate": 1.7369517847729301e-06, + "loss": 0.3889, + "step": 33370 + }, + { + "epoch": 0.7345586616773058, + "grad_norm": 2.163627682694531, + "learning_rate": 1.7366817080981107e-06, + "loss": 0.3118, + "step": 33371 + }, + { + "epoch": 0.7345806735637244, + "grad_norm": 1.4858349425296993, + "learning_rate": 1.736411648008929e-06, + "loss": 0.2252, + "step": 33372 + }, + { + "epoch": 0.7346026854501431, + "grad_norm": 1.521182984554446, + "learning_rate": 1.7361416045067586e-06, + "loss": 0.2535, + "step": 33373 + }, + { + "epoch": 0.7346246973365618, + "grad_norm": 10.200182451324334, + "learning_rate": 1.735871577592974e-06, + "loss": 0.2936, + "step": 33374 + }, + { + "epoch": 0.7346467092229804, + "grad_norm": 1.6409303075091501, + "learning_rate": 1.7356015672689442e-06, + "loss": 0.2529, + "step": 33375 + }, + { + "epoch": 0.7346687211093991, + "grad_norm": 2.3500435815689076, + "learning_rate": 1.7353315735360448e-06, + "loss": 0.3113, + "step": 33376 + }, + { + "epoch": 0.7346907329958178, + "grad_norm": 1.6381582442676577, + "learning_rate": 1.7350615963956452e-06, + "loss": 0.2722, + "step": 33377 + }, + { + "epoch": 0.7347127448822364, + "grad_norm": 1.655108830489148, + "learning_rate": 1.7347916358491185e-06, + "loss": 0.3346, + "step": 33378 + }, + { + "epoch": 0.7347347567686551, + "grad_norm": 1.4594049324682292, + "learning_rate": 1.7345216918978385e-06, + "loss": 0.289, + "step": 33379 + }, + { + "epoch": 0.7347567686550738, + "grad_norm": 1.601680534483586, + "learning_rate": 1.7342517645431745e-06, + "loss": 0.2826, + "step": 33380 + }, + { + "epoch": 0.7347787805414924, + "grad_norm": 1.8950785422464274, + "learning_rate": 1.7339818537865005e-06, + "loss": 0.2698, + "step": 33381 + }, + { + "epoch": 0.7348007924279111, + "grad_norm": 1.5662471791946784, + "learning_rate": 1.733711959629188e-06, + "loss": 0.2663, + "step": 33382 + }, + { + "epoch": 0.7348228043143298, + "grad_norm": 1.8854039129567206, + "learning_rate": 1.733442082072605e-06, + "loss": 0.2623, + "step": 33383 + }, + { + "epoch": 0.7348448162007484, + "grad_norm": 1.367412699508327, + "learning_rate": 1.733172221118129e-06, + "loss": 0.2847, + "step": 33384 + }, + { + "epoch": 0.7348668280871671, + "grad_norm": 0.630939825504447, + "learning_rate": 1.7329023767671293e-06, + "loss": 0.4628, + "step": 33385 + }, + { + "epoch": 0.7348888399735858, + "grad_norm": 1.5144939244320923, + "learning_rate": 1.732632549020975e-06, + "loss": 0.2657, + "step": 33386 + }, + { + "epoch": 0.7349108518600044, + "grad_norm": 2.4845138071684865, + "learning_rate": 1.7323627378810409e-06, + "loss": 0.2629, + "step": 33387 + }, + { + "epoch": 0.7349328637464231, + "grad_norm": 1.3742412141929892, + "learning_rate": 1.7320929433486938e-06, + "loss": 0.2612, + "step": 33388 + }, + { + "epoch": 0.7349548756328418, + "grad_norm": 1.606305743718111, + "learning_rate": 1.7318231654253109e-06, + "loss": 0.2482, + "step": 33389 + }, + { + "epoch": 0.7349768875192604, + "grad_norm": 1.7641910305638413, + "learning_rate": 1.7315534041122605e-06, + "loss": 0.345, + "step": 33390 + }, + { + "epoch": 0.7349988994056791, + "grad_norm": 1.1673634228889547, + "learning_rate": 1.7312836594109117e-06, + "loss": 0.2493, + "step": 33391 + }, + { + "epoch": 0.7350209112920977, + "grad_norm": 1.875925282460488, + "learning_rate": 1.7310139313226394e-06, + "loss": 0.2747, + "step": 33392 + }, + { + "epoch": 0.7350429231785164, + "grad_norm": 0.6251489338425638, + "learning_rate": 1.73074421984881e-06, + "loss": 0.467, + "step": 33393 + }, + { + "epoch": 0.7350649350649351, + "grad_norm": 0.6162333893599107, + "learning_rate": 1.7304745249907968e-06, + "loss": 0.4391, + "step": 33394 + }, + { + "epoch": 0.7350869469513537, + "grad_norm": 1.6591110865348497, + "learning_rate": 1.7302048467499716e-06, + "loss": 0.292, + "step": 33395 + }, + { + "epoch": 0.7351089588377724, + "grad_norm": 1.4269113208283652, + "learning_rate": 1.7299351851277025e-06, + "loss": 0.2874, + "step": 33396 + }, + { + "epoch": 0.7351309707241911, + "grad_norm": 1.2984887213059522, + "learning_rate": 1.729665540125363e-06, + "loss": 0.3111, + "step": 33397 + }, + { + "epoch": 0.7351529826106097, + "grad_norm": 1.3739440248760688, + "learning_rate": 1.7293959117443198e-06, + "loss": 0.2861, + "step": 33398 + }, + { + "epoch": 0.7351749944970284, + "grad_norm": 0.643755182425623, + "learning_rate": 1.729126299985946e-06, + "loss": 0.4661, + "step": 33399 + }, + { + "epoch": 0.7351970063834471, + "grad_norm": 1.9085130992717139, + "learning_rate": 1.7288567048516126e-06, + "loss": 0.2774, + "step": 33400 + }, + { + "epoch": 0.7352190182698657, + "grad_norm": 2.0599975444459107, + "learning_rate": 1.7285871263426862e-06, + "loss": 0.2768, + "step": 33401 + }, + { + "epoch": 0.7352410301562844, + "grad_norm": 2.099388945584618, + "learning_rate": 1.728317564460541e-06, + "loss": 0.3602, + "step": 33402 + }, + { + "epoch": 0.7352630420427031, + "grad_norm": 1.706833567748048, + "learning_rate": 1.7280480192065435e-06, + "loss": 0.3555, + "step": 33403 + }, + { + "epoch": 0.7352850539291217, + "grad_norm": 1.5878129929029212, + "learning_rate": 1.7277784905820655e-06, + "loss": 0.3053, + "step": 33404 + }, + { + "epoch": 0.7353070658155404, + "grad_norm": 1.5081918116277948, + "learning_rate": 1.7275089785884786e-06, + "loss": 0.3098, + "step": 33405 + }, + { + "epoch": 0.7353290777019591, + "grad_norm": 0.6175711148178231, + "learning_rate": 1.7272394832271483e-06, + "loss": 0.4486, + "step": 33406 + }, + { + "epoch": 0.7353510895883777, + "grad_norm": 1.4992087112558818, + "learning_rate": 1.7269700044994487e-06, + "loss": 0.2802, + "step": 33407 + }, + { + "epoch": 0.7353731014747964, + "grad_norm": 0.6114383721250424, + "learning_rate": 1.7267005424067456e-06, + "loss": 0.4375, + "step": 33408 + }, + { + "epoch": 0.7353951133612151, + "grad_norm": 1.5401378091698787, + "learning_rate": 1.7264310969504106e-06, + "loss": 0.3206, + "step": 33409 + }, + { + "epoch": 0.7354171252476337, + "grad_norm": 0.6218987585806305, + "learning_rate": 1.726161668131814e-06, + "loss": 0.4651, + "step": 33410 + }, + { + "epoch": 0.7354391371340524, + "grad_norm": 0.6171462518933126, + "learning_rate": 1.7258922559523221e-06, + "loss": 0.4376, + "step": 33411 + }, + { + "epoch": 0.7354611490204711, + "grad_norm": 0.6204676956531225, + "learning_rate": 1.7256228604133084e-06, + "loss": 0.4396, + "step": 33412 + }, + { + "epoch": 0.7354831609068897, + "grad_norm": 1.7632490485966577, + "learning_rate": 1.7253534815161378e-06, + "loss": 0.2774, + "step": 33413 + }, + { + "epoch": 0.7355051727933084, + "grad_norm": 2.222021728701057, + "learning_rate": 1.7250841192621814e-06, + "loss": 0.3433, + "step": 33414 + }, + { + "epoch": 0.735527184679727, + "grad_norm": 1.688256497747368, + "learning_rate": 1.7248147736528099e-06, + "loss": 0.318, + "step": 33415 + }, + { + "epoch": 0.7355491965661457, + "grad_norm": 1.3005246220351103, + "learning_rate": 1.7245454446893883e-06, + "loss": 0.3118, + "step": 33416 + }, + { + "epoch": 0.7355712084525644, + "grad_norm": 0.5986514861962686, + "learning_rate": 1.72427613237329e-06, + "loss": 0.4331, + "step": 33417 + }, + { + "epoch": 0.735593220338983, + "grad_norm": 1.5099575668641654, + "learning_rate": 1.724006836705881e-06, + "loss": 0.2179, + "step": 33418 + }, + { + "epoch": 0.7356152322254017, + "grad_norm": 1.8587235000407587, + "learning_rate": 1.7237375576885267e-06, + "loss": 0.3519, + "step": 33419 + }, + { + "epoch": 0.7356372441118204, + "grad_norm": 1.8955942405192847, + "learning_rate": 1.7234682953226035e-06, + "loss": 0.4066, + "step": 33420 + }, + { + "epoch": 0.735659255998239, + "grad_norm": 1.8401008211791685, + "learning_rate": 1.7231990496094736e-06, + "loss": 0.2943, + "step": 33421 + }, + { + "epoch": 0.7356812678846577, + "grad_norm": 1.7739237870832387, + "learning_rate": 1.7229298205505096e-06, + "loss": 0.2952, + "step": 33422 + }, + { + "epoch": 0.7357032797710764, + "grad_norm": 1.7796615903727746, + "learning_rate": 1.7226606081470776e-06, + "loss": 0.2794, + "step": 33423 + }, + { + "epoch": 0.735725291657495, + "grad_norm": 1.4027902159570105, + "learning_rate": 1.7223914124005447e-06, + "loss": 0.2643, + "step": 33424 + }, + { + "epoch": 0.7357473035439137, + "grad_norm": 1.6487302215636273, + "learning_rate": 1.7221222333122806e-06, + "loss": 0.2278, + "step": 33425 + }, + { + "epoch": 0.7357693154303324, + "grad_norm": 1.7807090532050331, + "learning_rate": 1.7218530708836544e-06, + "loss": 0.3366, + "step": 33426 + }, + { + "epoch": 0.735791327316751, + "grad_norm": 2.1735241583047897, + "learning_rate": 1.7215839251160316e-06, + "loss": 0.3073, + "step": 33427 + }, + { + "epoch": 0.7358133392031697, + "grad_norm": 1.5794242246283232, + "learning_rate": 1.7213147960107834e-06, + "loss": 0.2933, + "step": 33428 + }, + { + "epoch": 0.7358353510895884, + "grad_norm": 1.757386782206764, + "learning_rate": 1.721045683569274e-06, + "loss": 0.3463, + "step": 33429 + }, + { + "epoch": 0.735857362976007, + "grad_norm": 1.5391247872326184, + "learning_rate": 1.7207765877928728e-06, + "loss": 0.3087, + "step": 33430 + }, + { + "epoch": 0.7358793748624257, + "grad_norm": 1.9396544045728599, + "learning_rate": 1.7205075086829493e-06, + "loss": 0.3517, + "step": 33431 + }, + { + "epoch": 0.7359013867488444, + "grad_norm": 1.3799747074699193, + "learning_rate": 1.7202384462408671e-06, + "loss": 0.2948, + "step": 33432 + }, + { + "epoch": 0.735923398635263, + "grad_norm": 1.4666067807104748, + "learning_rate": 1.7199694004679984e-06, + "loss": 0.3313, + "step": 33433 + }, + { + "epoch": 0.7359454105216817, + "grad_norm": 1.5265247557195223, + "learning_rate": 1.7197003713657063e-06, + "loss": 0.206, + "step": 33434 + }, + { + "epoch": 0.7359674224081004, + "grad_norm": 1.825133787915513, + "learning_rate": 1.7194313589353595e-06, + "loss": 0.2308, + "step": 33435 + }, + { + "epoch": 0.735989434294519, + "grad_norm": 1.7165643337055017, + "learning_rate": 1.7191623631783277e-06, + "loss": 0.3007, + "step": 33436 + }, + { + "epoch": 0.7360114461809377, + "grad_norm": 1.8756531124035183, + "learning_rate": 1.7188933840959749e-06, + "loss": 0.3485, + "step": 33437 + }, + { + "epoch": 0.7360334580673564, + "grad_norm": 1.5367114153261912, + "learning_rate": 1.7186244216896702e-06, + "loss": 0.2886, + "step": 33438 + }, + { + "epoch": 0.736055469953775, + "grad_norm": 1.820629758354084, + "learning_rate": 1.7183554759607786e-06, + "loss": 0.2702, + "step": 33439 + }, + { + "epoch": 0.7360774818401937, + "grad_norm": 1.6657884733372583, + "learning_rate": 1.7180865469106678e-06, + "loss": 0.2231, + "step": 33440 + }, + { + "epoch": 0.7360994937266123, + "grad_norm": 3.802718486762892, + "learning_rate": 1.7178176345407066e-06, + "loss": 0.2872, + "step": 33441 + }, + { + "epoch": 0.736121505613031, + "grad_norm": 1.7326215153581077, + "learning_rate": 1.7175487388522588e-06, + "loss": 0.3675, + "step": 33442 + }, + { + "epoch": 0.7361435174994497, + "grad_norm": 2.089284622584538, + "learning_rate": 1.7172798598466932e-06, + "loss": 0.286, + "step": 33443 + }, + { + "epoch": 0.7361655293858683, + "grad_norm": 1.5389913985592851, + "learning_rate": 1.7170109975253745e-06, + "loss": 0.3113, + "step": 33444 + }, + { + "epoch": 0.736187541272287, + "grad_norm": 1.4253845772299787, + "learning_rate": 1.71674215188967e-06, + "loss": 0.2853, + "step": 33445 + }, + { + "epoch": 0.7362095531587057, + "grad_norm": 2.374761043350972, + "learning_rate": 1.716473322940948e-06, + "loss": 0.2952, + "step": 33446 + }, + { + "epoch": 0.7362315650451243, + "grad_norm": 2.068540986323688, + "learning_rate": 1.7162045106805713e-06, + "loss": 0.3066, + "step": 33447 + }, + { + "epoch": 0.736253576931543, + "grad_norm": 1.7883963118557922, + "learning_rate": 1.7159357151099093e-06, + "loss": 0.3216, + "step": 33448 + }, + { + "epoch": 0.7362755888179617, + "grad_norm": 0.613610728031237, + "learning_rate": 1.7156669362303257e-06, + "loss": 0.4453, + "step": 33449 + }, + { + "epoch": 0.7362976007043803, + "grad_norm": 0.5979928734678366, + "learning_rate": 1.7153981740431869e-06, + "loss": 0.4345, + "step": 33450 + }, + { + "epoch": 0.736319612590799, + "grad_norm": 1.9366272532754965, + "learning_rate": 1.7151294285498615e-06, + "loss": 0.3199, + "step": 33451 + }, + { + "epoch": 0.7363416244772177, + "grad_norm": 2.1754617465550035, + "learning_rate": 1.714860699751712e-06, + "loss": 0.3396, + "step": 33452 + }, + { + "epoch": 0.7363636363636363, + "grad_norm": 1.329015388088092, + "learning_rate": 1.7145919876501072e-06, + "loss": 0.3197, + "step": 33453 + }, + { + "epoch": 0.736385648250055, + "grad_norm": 2.316787127102066, + "learning_rate": 1.7143232922464108e-06, + "loss": 0.3444, + "step": 33454 + }, + { + "epoch": 0.7364076601364737, + "grad_norm": 1.505598649993157, + "learning_rate": 1.7140546135419861e-06, + "loss": 0.1616, + "step": 33455 + }, + { + "epoch": 0.7364296720228923, + "grad_norm": 1.6267655044301685, + "learning_rate": 1.713785951538205e-06, + "loss": 0.2497, + "step": 33456 + }, + { + "epoch": 0.736451683909311, + "grad_norm": 1.4703416548255628, + "learning_rate": 1.7135173062364269e-06, + "loss": 0.3221, + "step": 33457 + }, + { + "epoch": 0.7364736957957297, + "grad_norm": 1.698563648621272, + "learning_rate": 1.7132486776380219e-06, + "loss": 0.3216, + "step": 33458 + }, + { + "epoch": 0.7364957076821483, + "grad_norm": 1.3260714570716872, + "learning_rate": 1.712980065744353e-06, + "loss": 0.2808, + "step": 33459 + }, + { + "epoch": 0.736517719568567, + "grad_norm": 1.4076978745874766, + "learning_rate": 1.7127114705567832e-06, + "loss": 0.3518, + "step": 33460 + }, + { + "epoch": 0.7365397314549857, + "grad_norm": 1.4360781690182947, + "learning_rate": 1.7124428920766805e-06, + "loss": 0.3617, + "step": 33461 + }, + { + "epoch": 0.7365617433414043, + "grad_norm": 1.7913811584037558, + "learning_rate": 1.712174330305409e-06, + "loss": 0.3538, + "step": 33462 + }, + { + "epoch": 0.736583755227823, + "grad_norm": 1.8172452682465794, + "learning_rate": 1.711905785244336e-06, + "loss": 0.2937, + "step": 33463 + }, + { + "epoch": 0.7366057671142416, + "grad_norm": 2.0499119001489117, + "learning_rate": 1.7116372568948241e-06, + "loss": 0.2417, + "step": 33464 + }, + { + "epoch": 0.7366277790006603, + "grad_norm": 0.6257355564126457, + "learning_rate": 1.7113687452582367e-06, + "loss": 0.4616, + "step": 33465 + }, + { + "epoch": 0.736649790887079, + "grad_norm": 1.448570546704207, + "learning_rate": 1.7111002503359397e-06, + "loss": 0.2931, + "step": 33466 + }, + { + "epoch": 0.7366718027734976, + "grad_norm": 1.507939158431026, + "learning_rate": 1.7108317721293005e-06, + "loss": 0.3356, + "step": 33467 + }, + { + "epoch": 0.7366938146599163, + "grad_norm": 1.4259973814088682, + "learning_rate": 1.7105633106396792e-06, + "loss": 0.2672, + "step": 33468 + }, + { + "epoch": 0.736715826546335, + "grad_norm": 1.5969752553147505, + "learning_rate": 1.7102948658684437e-06, + "loss": 0.2382, + "step": 33469 + }, + { + "epoch": 0.7367378384327536, + "grad_norm": 1.492289801489608, + "learning_rate": 1.7100264378169557e-06, + "loss": 0.3052, + "step": 33470 + }, + { + "epoch": 0.7367598503191723, + "grad_norm": 2.3536278030423903, + "learning_rate": 1.7097580264865803e-06, + "loss": 0.3765, + "step": 33471 + }, + { + "epoch": 0.736781862205591, + "grad_norm": 1.7866765175102617, + "learning_rate": 1.7094896318786841e-06, + "loss": 0.3182, + "step": 33472 + }, + { + "epoch": 0.7368038740920096, + "grad_norm": 1.9343795848505392, + "learning_rate": 1.709221253994628e-06, + "loss": 0.361, + "step": 33473 + }, + { + "epoch": 0.7368258859784284, + "grad_norm": 2.245763324669565, + "learning_rate": 1.7089528928357779e-06, + "loss": 0.373, + "step": 33474 + }, + { + "epoch": 0.7368478978648471, + "grad_norm": 0.6306074335332861, + "learning_rate": 1.708684548403496e-06, + "loss": 0.4459, + "step": 33475 + }, + { + "epoch": 0.7368699097512657, + "grad_norm": 1.6185505999761296, + "learning_rate": 1.7084162206991473e-06, + "loss": 0.3181, + "step": 33476 + }, + { + "epoch": 0.7368919216376844, + "grad_norm": 1.969048324551536, + "learning_rate": 1.7081479097240965e-06, + "loss": 0.386, + "step": 33477 + }, + { + "epoch": 0.7369139335241031, + "grad_norm": 1.5336098589968432, + "learning_rate": 1.707879615479705e-06, + "loss": 0.2758, + "step": 33478 + }, + { + "epoch": 0.7369359454105217, + "grad_norm": 1.4827406440070545, + "learning_rate": 1.7076113379673393e-06, + "loss": 0.2923, + "step": 33479 + }, + { + "epoch": 0.7369579572969404, + "grad_norm": 1.9811616206982092, + "learning_rate": 1.7073430771883592e-06, + "loss": 0.3026, + "step": 33480 + }, + { + "epoch": 0.7369799691833591, + "grad_norm": 1.443097874978085, + "learning_rate": 1.7070748331441305e-06, + "loss": 0.2944, + "step": 33481 + }, + { + "epoch": 0.7370019810697777, + "grad_norm": 1.3129203491884192, + "learning_rate": 1.7068066058360177e-06, + "loss": 0.2918, + "step": 33482 + }, + { + "epoch": 0.7370239929561964, + "grad_norm": 1.6914220967526759, + "learning_rate": 1.7065383952653809e-06, + "loss": 0.2708, + "step": 33483 + }, + { + "epoch": 0.7370460048426151, + "grad_norm": 1.6435565154206535, + "learning_rate": 1.7062702014335863e-06, + "loss": 0.3263, + "step": 33484 + }, + { + "epoch": 0.7370680167290337, + "grad_norm": 2.751879652467641, + "learning_rate": 1.7060020243419944e-06, + "loss": 0.2661, + "step": 33485 + }, + { + "epoch": 0.7370900286154524, + "grad_norm": 1.4445211342995572, + "learning_rate": 1.7057338639919691e-06, + "loss": 0.2791, + "step": 33486 + }, + { + "epoch": 0.737112040501871, + "grad_norm": 1.777151642568323, + "learning_rate": 1.7054657203848756e-06, + "loss": 0.2672, + "step": 33487 + }, + { + "epoch": 0.7371340523882897, + "grad_norm": 1.3372679589213146, + "learning_rate": 1.7051975935220726e-06, + "loss": 0.2576, + "step": 33488 + }, + { + "epoch": 0.7371560642747084, + "grad_norm": 3.145297248891455, + "learning_rate": 1.7049294834049267e-06, + "loss": 0.3207, + "step": 33489 + }, + { + "epoch": 0.737178076161127, + "grad_norm": 1.531320912412882, + "learning_rate": 1.7046613900347992e-06, + "loss": 0.239, + "step": 33490 + }, + { + "epoch": 0.7372000880475457, + "grad_norm": 2.0203390631887763, + "learning_rate": 1.7043933134130486e-06, + "loss": 0.3559, + "step": 33491 + }, + { + "epoch": 0.7372220999339644, + "grad_norm": 1.6119355594661446, + "learning_rate": 1.7041252535410441e-06, + "loss": 0.2743, + "step": 33492 + }, + { + "epoch": 0.737244111820383, + "grad_norm": 2.313994687528374, + "learning_rate": 1.703857210420144e-06, + "loss": 0.3173, + "step": 33493 + }, + { + "epoch": 0.7372661237068017, + "grad_norm": 1.7817856946163257, + "learning_rate": 1.7035891840517133e-06, + "loss": 0.3313, + "step": 33494 + }, + { + "epoch": 0.7372881355932204, + "grad_norm": 2.194107971536013, + "learning_rate": 1.703321174437112e-06, + "loss": 0.3118, + "step": 33495 + }, + { + "epoch": 0.737310147479639, + "grad_norm": 1.6322357069004836, + "learning_rate": 1.7030531815777018e-06, + "loss": 0.2357, + "step": 33496 + }, + { + "epoch": 0.7373321593660577, + "grad_norm": 1.8408566815568272, + "learning_rate": 1.7027852054748457e-06, + "loss": 0.3226, + "step": 33497 + }, + { + "epoch": 0.7373541712524764, + "grad_norm": 1.4893233414298914, + "learning_rate": 1.7025172461299055e-06, + "loss": 0.1817, + "step": 33498 + }, + { + "epoch": 0.737376183138895, + "grad_norm": 1.284267359325517, + "learning_rate": 1.702249303544245e-06, + "loss": 0.2592, + "step": 33499 + }, + { + "epoch": 0.7373981950253137, + "grad_norm": 2.3609639373132936, + "learning_rate": 1.7019813777192246e-06, + "loss": 0.2959, + "step": 33500 + }, + { + "epoch": 0.7374202069117324, + "grad_norm": 3.0672193216556036, + "learning_rate": 1.7017134686562041e-06, + "loss": 0.3341, + "step": 33501 + }, + { + "epoch": 0.737442218798151, + "grad_norm": 0.6355555357492372, + "learning_rate": 1.7014455763565467e-06, + "loss": 0.4557, + "step": 33502 + }, + { + "epoch": 0.7374642306845697, + "grad_norm": 2.2906619711010565, + "learning_rate": 1.7011777008216147e-06, + "loss": 0.2271, + "step": 33503 + }, + { + "epoch": 0.7374862425709884, + "grad_norm": 2.103529849915823, + "learning_rate": 1.7009098420527697e-06, + "loss": 0.3343, + "step": 33504 + }, + { + "epoch": 0.737508254457407, + "grad_norm": 5.402134695355897, + "learning_rate": 1.700642000051373e-06, + "loss": 0.2778, + "step": 33505 + }, + { + "epoch": 0.7375302663438257, + "grad_norm": 1.6294817036068172, + "learning_rate": 1.7003741748187835e-06, + "loss": 0.3597, + "step": 33506 + }, + { + "epoch": 0.7375522782302444, + "grad_norm": 1.2687154594456407, + "learning_rate": 1.7001063663563637e-06, + "loss": 0.3054, + "step": 33507 + }, + { + "epoch": 0.737574290116663, + "grad_norm": 1.719754683042305, + "learning_rate": 1.6998385746654771e-06, + "loss": 0.2956, + "step": 33508 + }, + { + "epoch": 0.7375963020030817, + "grad_norm": 1.4796803050980634, + "learning_rate": 1.6995707997474808e-06, + "loss": 0.2438, + "step": 33509 + }, + { + "epoch": 0.7376183138895003, + "grad_norm": 1.4221527706528545, + "learning_rate": 1.6993030416037398e-06, + "loss": 0.2968, + "step": 33510 + }, + { + "epoch": 0.737640325775919, + "grad_norm": 1.5277625869354967, + "learning_rate": 1.6990353002356108e-06, + "loss": 0.3782, + "step": 33511 + }, + { + "epoch": 0.7376623376623377, + "grad_norm": 1.8660209967285972, + "learning_rate": 1.6987675756444571e-06, + "loss": 0.3455, + "step": 33512 + }, + { + "epoch": 0.7376843495487563, + "grad_norm": 1.7174811868940512, + "learning_rate": 1.6984998678316406e-06, + "loss": 0.3327, + "step": 33513 + }, + { + "epoch": 0.737706361435175, + "grad_norm": 1.6824350560553851, + "learning_rate": 1.6982321767985182e-06, + "loss": 0.2895, + "step": 33514 + }, + { + "epoch": 0.7377283733215937, + "grad_norm": 2.3801926230716624, + "learning_rate": 1.6979645025464547e-06, + "loss": 0.2556, + "step": 33515 + }, + { + "epoch": 0.7377503852080123, + "grad_norm": 0.6012342728709529, + "learning_rate": 1.6976968450768061e-06, + "loss": 0.4441, + "step": 33516 + }, + { + "epoch": 0.737772397094431, + "grad_norm": 1.8388124193520916, + "learning_rate": 1.6974292043909357e-06, + "loss": 0.3738, + "step": 33517 + }, + { + "epoch": 0.7377944089808497, + "grad_norm": 8.608153616856676, + "learning_rate": 1.6971615804902042e-06, + "loss": 0.2136, + "step": 33518 + }, + { + "epoch": 0.7378164208672683, + "grad_norm": 1.7762659267704126, + "learning_rate": 1.6968939733759692e-06, + "loss": 0.3578, + "step": 33519 + }, + { + "epoch": 0.737838432753687, + "grad_norm": 2.1252644190592314, + "learning_rate": 1.6966263830495939e-06, + "loss": 0.2359, + "step": 33520 + }, + { + "epoch": 0.7378604446401057, + "grad_norm": 1.22670872634724, + "learning_rate": 1.696358809512435e-06, + "loss": 0.2504, + "step": 33521 + }, + { + "epoch": 0.7378824565265243, + "grad_norm": 3.1345005250020015, + "learning_rate": 1.6960912527658541e-06, + "loss": 0.3202, + "step": 33522 + }, + { + "epoch": 0.737904468412943, + "grad_norm": 2.1518813667618684, + "learning_rate": 1.6958237128112125e-06, + "loss": 0.2528, + "step": 33523 + }, + { + "epoch": 0.7379264802993617, + "grad_norm": 1.8594003275727657, + "learning_rate": 1.6955561896498669e-06, + "loss": 0.3055, + "step": 33524 + }, + { + "epoch": 0.7379484921857803, + "grad_norm": 1.474799683610065, + "learning_rate": 1.6952886832831805e-06, + "loss": 0.2547, + "step": 33525 + }, + { + "epoch": 0.737970504072199, + "grad_norm": 1.9310746125944858, + "learning_rate": 1.6950211937125105e-06, + "loss": 0.3222, + "step": 33526 + }, + { + "epoch": 0.7379925159586177, + "grad_norm": 0.6177552538995997, + "learning_rate": 1.694753720939214e-06, + "loss": 0.452, + "step": 33527 + }, + { + "epoch": 0.7380145278450363, + "grad_norm": 1.3230651911047955, + "learning_rate": 1.6944862649646566e-06, + "loss": 0.2225, + "step": 33528 + }, + { + "epoch": 0.738036539731455, + "grad_norm": 2.593999930667992, + "learning_rate": 1.6942188257901926e-06, + "loss": 0.2494, + "step": 33529 + }, + { + "epoch": 0.7380585516178737, + "grad_norm": 0.6385206346844654, + "learning_rate": 1.6939514034171845e-06, + "loss": 0.4645, + "step": 33530 + }, + { + "epoch": 0.7380805635042923, + "grad_norm": 1.522298277804193, + "learning_rate": 1.6936839978469894e-06, + "loss": 0.2794, + "step": 33531 + }, + { + "epoch": 0.738102575390711, + "grad_norm": 1.2884674075697413, + "learning_rate": 1.6934166090809655e-06, + "loss": 0.2671, + "step": 33532 + }, + { + "epoch": 0.7381245872771296, + "grad_norm": 1.6271986981098883, + "learning_rate": 1.6931492371204738e-06, + "loss": 0.2811, + "step": 33533 + }, + { + "epoch": 0.7381465991635483, + "grad_norm": 1.4316415346341858, + "learning_rate": 1.692881881966872e-06, + "loss": 0.3145, + "step": 33534 + }, + { + "epoch": 0.738168611049967, + "grad_norm": 1.5369469664704647, + "learning_rate": 1.6926145436215213e-06, + "loss": 0.3173, + "step": 33535 + }, + { + "epoch": 0.7381906229363856, + "grad_norm": 1.8751979991497583, + "learning_rate": 1.6923472220857785e-06, + "loss": 0.2637, + "step": 33536 + }, + { + "epoch": 0.7382126348228043, + "grad_norm": 1.9668770526395796, + "learning_rate": 1.692079917361001e-06, + "loss": 0.2617, + "step": 33537 + }, + { + "epoch": 0.738234646709223, + "grad_norm": 1.3283505833834663, + "learning_rate": 1.6918126294485486e-06, + "loss": 0.229, + "step": 33538 + }, + { + "epoch": 0.7382566585956416, + "grad_norm": 1.682968414864924, + "learning_rate": 1.6915453583497797e-06, + "loss": 0.3497, + "step": 33539 + }, + { + "epoch": 0.7382786704820603, + "grad_norm": 1.2213638980774304, + "learning_rate": 1.6912781040660548e-06, + "loss": 0.2084, + "step": 33540 + }, + { + "epoch": 0.738300682368479, + "grad_norm": 1.78644319682012, + "learning_rate": 1.6910108665987301e-06, + "loss": 0.2374, + "step": 33541 + }, + { + "epoch": 0.7383226942548976, + "grad_norm": 2.9372677899868562, + "learning_rate": 1.6907436459491627e-06, + "loss": 0.3035, + "step": 33542 + }, + { + "epoch": 0.7383447061413163, + "grad_norm": 2.104120221815623, + "learning_rate": 1.690476442118712e-06, + "loss": 0.2755, + "step": 33543 + }, + { + "epoch": 0.738366718027735, + "grad_norm": 2.9879036519421214, + "learning_rate": 1.6902092551087374e-06, + "loss": 0.3278, + "step": 33544 + }, + { + "epoch": 0.7383887299141536, + "grad_norm": 1.6242726246915922, + "learning_rate": 1.6899420849205945e-06, + "loss": 0.2111, + "step": 33545 + }, + { + "epoch": 0.7384107418005723, + "grad_norm": 1.746413772957424, + "learning_rate": 1.6896749315556431e-06, + "loss": 0.3109, + "step": 33546 + }, + { + "epoch": 0.738432753686991, + "grad_norm": 0.6295402757371343, + "learning_rate": 1.6894077950152393e-06, + "loss": 0.4415, + "step": 33547 + }, + { + "epoch": 0.7384547655734096, + "grad_norm": 8.962839793802717, + "learning_rate": 1.689140675300741e-06, + "loss": 0.2317, + "step": 33548 + }, + { + "epoch": 0.7384767774598283, + "grad_norm": 1.3895668753438357, + "learning_rate": 1.6888735724135087e-06, + "loss": 0.2636, + "step": 33549 + }, + { + "epoch": 0.738498789346247, + "grad_norm": 1.7877040544906673, + "learning_rate": 1.6886064863548956e-06, + "loss": 0.3064, + "step": 33550 + }, + { + "epoch": 0.7385208012326656, + "grad_norm": 1.6670466378873599, + "learning_rate": 1.688339417126263e-06, + "loss": 0.3514, + "step": 33551 + }, + { + "epoch": 0.7385428131190843, + "grad_norm": 1.8760008954853953, + "learning_rate": 1.688072364728965e-06, + "loss": 0.2866, + "step": 33552 + }, + { + "epoch": 0.738564825005503, + "grad_norm": 1.5831876158171956, + "learning_rate": 1.6878053291643603e-06, + "loss": 0.3408, + "step": 33553 + }, + { + "epoch": 0.7385868368919216, + "grad_norm": 2.141047764861539, + "learning_rate": 1.6875383104338077e-06, + "loss": 0.3165, + "step": 33554 + }, + { + "epoch": 0.7386088487783403, + "grad_norm": 1.5469904250147624, + "learning_rate": 1.6872713085386615e-06, + "loss": 0.2467, + "step": 33555 + }, + { + "epoch": 0.7386308606647589, + "grad_norm": 1.4844574329669604, + "learning_rate": 1.6870043234802819e-06, + "loss": 0.3355, + "step": 33556 + }, + { + "epoch": 0.7386528725511776, + "grad_norm": 1.555213039570648, + "learning_rate": 1.686737355260022e-06, + "loss": 0.2566, + "step": 33557 + }, + { + "epoch": 0.7386748844375963, + "grad_norm": 2.9226877959330753, + "learning_rate": 1.68647040387924e-06, + "loss": 0.3568, + "step": 33558 + }, + { + "epoch": 0.7386968963240149, + "grad_norm": 1.399693994104857, + "learning_rate": 1.6862034693392958e-06, + "loss": 0.1905, + "step": 33559 + }, + { + "epoch": 0.7387189082104336, + "grad_norm": 2.7161819627323536, + "learning_rate": 1.6859365516415416e-06, + "loss": 0.2715, + "step": 33560 + }, + { + "epoch": 0.7387409200968523, + "grad_norm": 2.439016408273768, + "learning_rate": 1.6856696507873377e-06, + "loss": 0.271, + "step": 33561 + }, + { + "epoch": 0.7387629319832709, + "grad_norm": 2.2184934302173396, + "learning_rate": 1.685402766778037e-06, + "loss": 0.2354, + "step": 33562 + }, + { + "epoch": 0.7387849438696896, + "grad_norm": 4.08407119050052, + "learning_rate": 1.685135899614998e-06, + "loss": 0.2221, + "step": 33563 + }, + { + "epoch": 0.7388069557561083, + "grad_norm": 0.6014263323906477, + "learning_rate": 1.6848690492995784e-06, + "loss": 0.448, + "step": 33564 + }, + { + "epoch": 0.7388289676425269, + "grad_norm": 0.6123340820970457, + "learning_rate": 1.6846022158331315e-06, + "loss": 0.4779, + "step": 33565 + }, + { + "epoch": 0.7388509795289456, + "grad_norm": 1.4007148346743143, + "learning_rate": 1.6843353992170159e-06, + "loss": 0.2947, + "step": 33566 + }, + { + "epoch": 0.7388729914153643, + "grad_norm": 0.6174339461704972, + "learning_rate": 1.6840685994525873e-06, + "loss": 0.4357, + "step": 33567 + }, + { + "epoch": 0.7388950033017829, + "grad_norm": 1.7899081077843813, + "learning_rate": 1.683801816541199e-06, + "loss": 0.3042, + "step": 33568 + }, + { + "epoch": 0.7389170151882016, + "grad_norm": 1.5472734349331747, + "learning_rate": 1.683535050484209e-06, + "loss": 0.2175, + "step": 33569 + }, + { + "epoch": 0.7389390270746203, + "grad_norm": 1.828223086330802, + "learning_rate": 1.683268301282973e-06, + "loss": 0.271, + "step": 33570 + }, + { + "epoch": 0.7389610389610389, + "grad_norm": 1.4503522614191648, + "learning_rate": 1.6830015689388486e-06, + "loss": 0.2401, + "step": 33571 + }, + { + "epoch": 0.7389830508474576, + "grad_norm": 1.2693031265554426, + "learning_rate": 1.6827348534531896e-06, + "loss": 0.2266, + "step": 33572 + }, + { + "epoch": 0.7390050627338763, + "grad_norm": 1.794095136995042, + "learning_rate": 1.6824681548273497e-06, + "loss": 0.313, + "step": 33573 + }, + { + "epoch": 0.7390270746202949, + "grad_norm": 1.2211839056967662, + "learning_rate": 1.6822014730626863e-06, + "loss": 0.227, + "step": 33574 + }, + { + "epoch": 0.7390490865067136, + "grad_norm": 1.3877653408128843, + "learning_rate": 1.6819348081605553e-06, + "loss": 0.328, + "step": 33575 + }, + { + "epoch": 0.7390710983931323, + "grad_norm": 0.6270685236372232, + "learning_rate": 1.6816681601223129e-06, + "loss": 0.4672, + "step": 33576 + }, + { + "epoch": 0.7390931102795509, + "grad_norm": 1.8497822083457456, + "learning_rate": 1.681401528949313e-06, + "loss": 0.3399, + "step": 33577 + }, + { + "epoch": 0.7391151221659696, + "grad_norm": 1.7162191045562063, + "learning_rate": 1.6811349146429085e-06, + "loss": 0.2572, + "step": 33578 + }, + { + "epoch": 0.7391371340523883, + "grad_norm": 0.6225990175823272, + "learning_rate": 1.680868317204457e-06, + "loss": 0.4433, + "step": 33579 + }, + { + "epoch": 0.7391591459388069, + "grad_norm": 1.499710407287927, + "learning_rate": 1.6806017366353132e-06, + "loss": 0.2984, + "step": 33580 + }, + { + "epoch": 0.7391811578252256, + "grad_norm": 1.5719454591694748, + "learning_rate": 1.6803351729368334e-06, + "loss": 0.2434, + "step": 33581 + }, + { + "epoch": 0.7392031697116442, + "grad_norm": 0.6401119891871007, + "learning_rate": 1.680068626110371e-06, + "loss": 0.4511, + "step": 33582 + }, + { + "epoch": 0.7392251815980629, + "grad_norm": 3.068243969959423, + "learning_rate": 1.6798020961572785e-06, + "loss": 0.324, + "step": 33583 + }, + { + "epoch": 0.7392471934844816, + "grad_norm": 2.687296853316888, + "learning_rate": 1.6795355830789127e-06, + "loss": 0.3387, + "step": 33584 + }, + { + "epoch": 0.7392692053709002, + "grad_norm": 2.1873538153460808, + "learning_rate": 1.6792690868766304e-06, + "loss": 0.2197, + "step": 33585 + }, + { + "epoch": 0.7392912172573189, + "grad_norm": 1.3269946715536125, + "learning_rate": 1.6790026075517812e-06, + "loss": 0.3475, + "step": 33586 + }, + { + "epoch": 0.7393132291437376, + "grad_norm": 1.2986293235381237, + "learning_rate": 1.6787361451057244e-06, + "loss": 0.2114, + "step": 33587 + }, + { + "epoch": 0.7393352410301562, + "grad_norm": 1.3046304448121309, + "learning_rate": 1.6784696995398092e-06, + "loss": 0.2353, + "step": 33588 + }, + { + "epoch": 0.739357252916575, + "grad_norm": 3.089079853302457, + "learning_rate": 1.678203270855393e-06, + "loss": 0.2166, + "step": 33589 + }, + { + "epoch": 0.7393792648029937, + "grad_norm": 1.5403441428480198, + "learning_rate": 1.6779368590538309e-06, + "loss": 0.2674, + "step": 33590 + }, + { + "epoch": 0.7394012766894122, + "grad_norm": 1.9631383197449084, + "learning_rate": 1.6776704641364738e-06, + "loss": 0.3042, + "step": 33591 + }, + { + "epoch": 0.739423288575831, + "grad_norm": 4.04674328375, + "learning_rate": 1.6774040861046791e-06, + "loss": 0.2409, + "step": 33592 + }, + { + "epoch": 0.7394453004622497, + "grad_norm": 0.6102881346475421, + "learning_rate": 1.6771377249597964e-06, + "loss": 0.4396, + "step": 33593 + }, + { + "epoch": 0.7394673123486682, + "grad_norm": 1.4985362524399477, + "learning_rate": 1.6768713807031823e-06, + "loss": 0.2689, + "step": 33594 + }, + { + "epoch": 0.739489324235087, + "grad_norm": 1.703880649080053, + "learning_rate": 1.6766050533361916e-06, + "loss": 0.3477, + "step": 33595 + }, + { + "epoch": 0.7395113361215057, + "grad_norm": 1.4792030399890532, + "learning_rate": 1.6763387428601745e-06, + "loss": 0.2277, + "step": 33596 + }, + { + "epoch": 0.7395333480079243, + "grad_norm": 2.470220090285474, + "learning_rate": 1.6760724492764885e-06, + "loss": 0.3003, + "step": 33597 + }, + { + "epoch": 0.739555359894343, + "grad_norm": 0.6220354853620973, + "learning_rate": 1.6758061725864827e-06, + "loss": 0.4714, + "step": 33598 + }, + { + "epoch": 0.7395773717807617, + "grad_norm": 0.6141185188012394, + "learning_rate": 1.6755399127915123e-06, + "loss": 0.4529, + "step": 33599 + }, + { + "epoch": 0.7395993836671803, + "grad_norm": 1.6526560352812971, + "learning_rate": 1.6752736698929328e-06, + "loss": 0.3396, + "step": 33600 + }, + { + "epoch": 0.739621395553599, + "grad_norm": 0.602409747020384, + "learning_rate": 1.6750074438920939e-06, + "loss": 0.4589, + "step": 33601 + }, + { + "epoch": 0.7396434074400177, + "grad_norm": 1.3211860530085002, + "learning_rate": 1.6747412347903513e-06, + "loss": 0.1835, + "step": 33602 + }, + { + "epoch": 0.7396654193264363, + "grad_norm": 1.7893913565618296, + "learning_rate": 1.6744750425890554e-06, + "loss": 0.2717, + "step": 33603 + }, + { + "epoch": 0.739687431212855, + "grad_norm": 1.925202544820661, + "learning_rate": 1.674208867289563e-06, + "loss": 0.2599, + "step": 33604 + }, + { + "epoch": 0.7397094430992736, + "grad_norm": 2.0055964173735785, + "learning_rate": 1.673942708893222e-06, + "loss": 0.3657, + "step": 33605 + }, + { + "epoch": 0.7397314549856923, + "grad_norm": 1.3005021655611755, + "learning_rate": 1.6736765674013877e-06, + "loss": 0.3468, + "step": 33606 + }, + { + "epoch": 0.739753466872111, + "grad_norm": 1.9272312184008662, + "learning_rate": 1.6734104428154142e-06, + "loss": 0.3908, + "step": 33607 + }, + { + "epoch": 0.7397754787585296, + "grad_norm": 0.6118310413889938, + "learning_rate": 1.6731443351366528e-06, + "loss": 0.4775, + "step": 33608 + }, + { + "epoch": 0.7397974906449483, + "grad_norm": 1.5881850130508548, + "learning_rate": 1.6728782443664537e-06, + "loss": 0.3435, + "step": 33609 + }, + { + "epoch": 0.739819502531367, + "grad_norm": 1.4648999399104146, + "learning_rate": 1.6726121705061716e-06, + "loss": 0.2615, + "step": 33610 + }, + { + "epoch": 0.7398415144177856, + "grad_norm": 1.95306122402387, + "learning_rate": 1.672346113557159e-06, + "loss": 0.2377, + "step": 33611 + }, + { + "epoch": 0.7398635263042043, + "grad_norm": 1.8609922561541552, + "learning_rate": 1.6720800735207682e-06, + "loss": 0.3475, + "step": 33612 + }, + { + "epoch": 0.739885538190623, + "grad_norm": 0.6194243607380913, + "learning_rate": 1.6718140503983515e-06, + "loss": 0.4661, + "step": 33613 + }, + { + "epoch": 0.7399075500770416, + "grad_norm": 1.6942796326665888, + "learning_rate": 1.6715480441912586e-06, + "loss": 0.3186, + "step": 33614 + }, + { + "epoch": 0.7399295619634603, + "grad_norm": 1.6198166915837975, + "learning_rate": 1.6712820549008423e-06, + "loss": 0.235, + "step": 33615 + }, + { + "epoch": 0.739951573849879, + "grad_norm": 1.7473019875759934, + "learning_rate": 1.671016082528456e-06, + "loss": 0.2887, + "step": 33616 + }, + { + "epoch": 0.7399735857362976, + "grad_norm": 1.6163887566580473, + "learning_rate": 1.6707501270754522e-06, + "loss": 0.297, + "step": 33617 + }, + { + "epoch": 0.7399955976227163, + "grad_norm": 1.3870476517981274, + "learning_rate": 1.670484188543181e-06, + "loss": 0.2471, + "step": 33618 + }, + { + "epoch": 0.740017609509135, + "grad_norm": 2.4690856807926482, + "learning_rate": 1.6702182669329931e-06, + "loss": 0.3455, + "step": 33619 + }, + { + "epoch": 0.7400396213955536, + "grad_norm": 1.9557988482737056, + "learning_rate": 1.6699523622462404e-06, + "loss": 0.334, + "step": 33620 + }, + { + "epoch": 0.7400616332819723, + "grad_norm": 2.449643677405207, + "learning_rate": 1.669686474484275e-06, + "loss": 0.3096, + "step": 33621 + }, + { + "epoch": 0.740083645168391, + "grad_norm": 1.4723767758348592, + "learning_rate": 1.6694206036484507e-06, + "loss": 0.3486, + "step": 33622 + }, + { + "epoch": 0.7401056570548096, + "grad_norm": 1.7360222654176198, + "learning_rate": 1.6691547497401157e-06, + "loss": 0.291, + "step": 33623 + }, + { + "epoch": 0.7401276689412283, + "grad_norm": 1.6948734677241157, + "learning_rate": 1.66888891276062e-06, + "loss": 0.3698, + "step": 33624 + }, + { + "epoch": 0.740149680827647, + "grad_norm": 1.4174865753784667, + "learning_rate": 1.6686230927113172e-06, + "loss": 0.2848, + "step": 33625 + }, + { + "epoch": 0.7401716927140656, + "grad_norm": 1.8799643194105708, + "learning_rate": 1.668357289593559e-06, + "loss": 0.3783, + "step": 33626 + }, + { + "epoch": 0.7401937046004843, + "grad_norm": 1.447582303719183, + "learning_rate": 1.6680915034086936e-06, + "loss": 0.3283, + "step": 33627 + }, + { + "epoch": 0.7402157164869029, + "grad_norm": 1.725930613741195, + "learning_rate": 1.6678257341580745e-06, + "loss": 0.3055, + "step": 33628 + }, + { + "epoch": 0.7402377283733216, + "grad_norm": 4.599810376729685, + "learning_rate": 1.66755998184305e-06, + "loss": 0.3136, + "step": 33629 + }, + { + "epoch": 0.7402597402597403, + "grad_norm": 0.6371941774071668, + "learning_rate": 1.6672942464649715e-06, + "loss": 0.4597, + "step": 33630 + }, + { + "epoch": 0.7402817521461589, + "grad_norm": 2.2237088589117646, + "learning_rate": 1.6670285280251925e-06, + "loss": 0.2143, + "step": 33631 + }, + { + "epoch": 0.7403037640325776, + "grad_norm": 2.4454366627119843, + "learning_rate": 1.6667628265250585e-06, + "loss": 0.2426, + "step": 33632 + }, + { + "epoch": 0.7403257759189963, + "grad_norm": 1.4532826634049851, + "learning_rate": 1.6664971419659248e-06, + "loss": 0.2528, + "step": 33633 + }, + { + "epoch": 0.7403477878054149, + "grad_norm": 2.459771935739656, + "learning_rate": 1.6662314743491375e-06, + "loss": 0.2861, + "step": 33634 + }, + { + "epoch": 0.7403697996918336, + "grad_norm": 1.3144789039059674, + "learning_rate": 1.66596582367605e-06, + "loss": 0.2972, + "step": 33635 + }, + { + "epoch": 0.7403918115782523, + "grad_norm": 0.6002807380033893, + "learning_rate": 1.6657001899480102e-06, + "loss": 0.4441, + "step": 33636 + }, + { + "epoch": 0.7404138234646709, + "grad_norm": 1.5563480279832593, + "learning_rate": 1.665434573166369e-06, + "loss": 0.3595, + "step": 33637 + }, + { + "epoch": 0.7404358353510896, + "grad_norm": 1.3420597796929399, + "learning_rate": 1.6651689733324784e-06, + "loss": 0.256, + "step": 33638 + }, + { + "epoch": 0.7404578472375083, + "grad_norm": 2.2405007357544475, + "learning_rate": 1.6649033904476842e-06, + "loss": 0.2537, + "step": 33639 + }, + { + "epoch": 0.7404798591239269, + "grad_norm": 1.9196032430259615, + "learning_rate": 1.6646378245133404e-06, + "loss": 0.2731, + "step": 33640 + }, + { + "epoch": 0.7405018710103456, + "grad_norm": 2.497352705324106, + "learning_rate": 1.6643722755307934e-06, + "loss": 0.3223, + "step": 33641 + }, + { + "epoch": 0.7405238828967643, + "grad_norm": 1.5614174683919284, + "learning_rate": 1.664106743501394e-06, + "loss": 0.2482, + "step": 33642 + }, + { + "epoch": 0.7405458947831829, + "grad_norm": 1.4932979865330298, + "learning_rate": 1.6638412284264937e-06, + "loss": 0.282, + "step": 33643 + }, + { + "epoch": 0.7405679066696016, + "grad_norm": 2.798382310989612, + "learning_rate": 1.66357573030744e-06, + "loss": 0.3101, + "step": 33644 + }, + { + "epoch": 0.7405899185560203, + "grad_norm": 5.194376249743203, + "learning_rate": 1.663310249145581e-06, + "loss": 0.3742, + "step": 33645 + }, + { + "epoch": 0.7406119304424389, + "grad_norm": 1.8849774724035946, + "learning_rate": 1.6630447849422677e-06, + "loss": 0.3274, + "step": 33646 + }, + { + "epoch": 0.7406339423288576, + "grad_norm": 1.6307562925314447, + "learning_rate": 1.6627793376988489e-06, + "loss": 0.2692, + "step": 33647 + }, + { + "epoch": 0.7406559542152763, + "grad_norm": 3.335505620469998, + "learning_rate": 1.6625139074166757e-06, + "loss": 0.1527, + "step": 33648 + }, + { + "epoch": 0.7406779661016949, + "grad_norm": 1.7424167783556372, + "learning_rate": 1.6622484940970957e-06, + "loss": 0.2925, + "step": 33649 + }, + { + "epoch": 0.7406999779881136, + "grad_norm": 1.658023930806429, + "learning_rate": 1.6619830977414553e-06, + "loss": 0.2477, + "step": 33650 + }, + { + "epoch": 0.7407219898745322, + "grad_norm": 1.6545523885058253, + "learning_rate": 1.6617177183511064e-06, + "loss": 0.3607, + "step": 33651 + }, + { + "epoch": 0.7407440017609509, + "grad_norm": 0.6265504593873716, + "learning_rate": 1.661452355927396e-06, + "loss": 0.4705, + "step": 33652 + }, + { + "epoch": 0.7407660136473696, + "grad_norm": 1.376033969746228, + "learning_rate": 1.661187010471676e-06, + "loss": 0.2847, + "step": 33653 + }, + { + "epoch": 0.7407880255337882, + "grad_norm": 1.9316529059230363, + "learning_rate": 1.6609216819852925e-06, + "loss": 0.2189, + "step": 33654 + }, + { + "epoch": 0.7408100374202069, + "grad_norm": 2.0463379216934374, + "learning_rate": 1.6606563704695927e-06, + "loss": 0.3005, + "step": 33655 + }, + { + "epoch": 0.7408320493066256, + "grad_norm": 2.261879914852534, + "learning_rate": 1.6603910759259268e-06, + "loss": 0.2909, + "step": 33656 + }, + { + "epoch": 0.7408540611930442, + "grad_norm": 2.2756295959189967, + "learning_rate": 1.6601257983556429e-06, + "loss": 0.2932, + "step": 33657 + }, + { + "epoch": 0.7408760730794629, + "grad_norm": 1.5957803643619908, + "learning_rate": 1.6598605377600906e-06, + "loss": 0.2822, + "step": 33658 + }, + { + "epoch": 0.7408980849658816, + "grad_norm": 1.8991257448916516, + "learning_rate": 1.6595952941406175e-06, + "loss": 0.3234, + "step": 33659 + }, + { + "epoch": 0.7409200968523002, + "grad_norm": 0.6244170429502681, + "learning_rate": 1.6593300674985685e-06, + "loss": 0.5033, + "step": 33660 + }, + { + "epoch": 0.7409421087387189, + "grad_norm": 1.2511589288702765, + "learning_rate": 1.6590648578352948e-06, + "loss": 0.2412, + "step": 33661 + }, + { + "epoch": 0.7409641206251376, + "grad_norm": 1.866785119678714, + "learning_rate": 1.6587996651521433e-06, + "loss": 0.3113, + "step": 33662 + }, + { + "epoch": 0.7409861325115562, + "grad_norm": 1.7402690252609012, + "learning_rate": 1.6585344894504635e-06, + "loss": 0.3402, + "step": 33663 + }, + { + "epoch": 0.7410081443979749, + "grad_norm": 1.5078045657130101, + "learning_rate": 1.6582693307316023e-06, + "loss": 0.239, + "step": 33664 + }, + { + "epoch": 0.7410301562843936, + "grad_norm": 1.4904768160388555, + "learning_rate": 1.6580041889969046e-06, + "loss": 0.207, + "step": 33665 + }, + { + "epoch": 0.7410521681708122, + "grad_norm": 1.4111364312789099, + "learning_rate": 1.6577390642477204e-06, + "loss": 0.3082, + "step": 33666 + }, + { + "epoch": 0.7410741800572309, + "grad_norm": 1.4224648657509036, + "learning_rate": 1.6574739564853986e-06, + "loss": 0.2534, + "step": 33667 + }, + { + "epoch": 0.7410961919436496, + "grad_norm": 2.181352834098221, + "learning_rate": 1.6572088657112833e-06, + "loss": 0.2527, + "step": 33668 + }, + { + "epoch": 0.7411182038300682, + "grad_norm": 1.6851753950566506, + "learning_rate": 1.6569437919267255e-06, + "loss": 0.3322, + "step": 33669 + }, + { + "epoch": 0.7411402157164869, + "grad_norm": 1.8382806996356036, + "learning_rate": 1.6566787351330687e-06, + "loss": 0.3356, + "step": 33670 + }, + { + "epoch": 0.7411622276029056, + "grad_norm": 1.5579279666676158, + "learning_rate": 1.656413695331663e-06, + "loss": 0.2781, + "step": 33671 + }, + { + "epoch": 0.7411842394893242, + "grad_norm": 1.867260215601183, + "learning_rate": 1.6561486725238525e-06, + "loss": 0.4133, + "step": 33672 + }, + { + "epoch": 0.7412062513757429, + "grad_norm": 1.6065565150467858, + "learning_rate": 1.6558836667109863e-06, + "loss": 0.3383, + "step": 33673 + }, + { + "epoch": 0.7412282632621615, + "grad_norm": 1.8117846790215801, + "learning_rate": 1.6556186778944121e-06, + "loss": 0.3603, + "step": 33674 + }, + { + "epoch": 0.7412502751485802, + "grad_norm": 2.9993130081482264, + "learning_rate": 1.6553537060754737e-06, + "loss": 0.2347, + "step": 33675 + }, + { + "epoch": 0.7412722870349989, + "grad_norm": 3.6030338309447196, + "learning_rate": 1.6550887512555215e-06, + "loss": 0.2519, + "step": 33676 + }, + { + "epoch": 0.7412942989214175, + "grad_norm": 1.2283937676184187, + "learning_rate": 1.6548238134358979e-06, + "loss": 0.2696, + "step": 33677 + }, + { + "epoch": 0.7413163108078362, + "grad_norm": 0.6405093795902647, + "learning_rate": 1.6545588926179524e-06, + "loss": 0.4433, + "step": 33678 + }, + { + "epoch": 0.7413383226942549, + "grad_norm": 0.6163238285454324, + "learning_rate": 1.654293988803032e-06, + "loss": 0.4316, + "step": 33679 + }, + { + "epoch": 0.7413603345806735, + "grad_norm": 1.6886886152239533, + "learning_rate": 1.6540291019924803e-06, + "loss": 0.3302, + "step": 33680 + }, + { + "epoch": 0.7413823464670922, + "grad_norm": 1.7940898925598165, + "learning_rate": 1.6537642321876462e-06, + "loss": 0.2547, + "step": 33681 + }, + { + "epoch": 0.7414043583535109, + "grad_norm": 2.8772879461142646, + "learning_rate": 1.6534993793898735e-06, + "loss": 0.3421, + "step": 33682 + }, + { + "epoch": 0.7414263702399295, + "grad_norm": 14.359839409218536, + "learning_rate": 1.6532345436005099e-06, + "loss": 0.2719, + "step": 33683 + }, + { + "epoch": 0.7414483821263482, + "grad_norm": 1.260350653456437, + "learning_rate": 1.652969724820902e-06, + "loss": 0.2474, + "step": 33684 + }, + { + "epoch": 0.7414703940127669, + "grad_norm": 1.2812694471418606, + "learning_rate": 1.652704923052395e-06, + "loss": 0.1573, + "step": 33685 + }, + { + "epoch": 0.7414924058991855, + "grad_norm": 1.4162101176128026, + "learning_rate": 1.652440138296333e-06, + "loss": 0.3601, + "step": 33686 + }, + { + "epoch": 0.7415144177856042, + "grad_norm": 5.485289629600084, + "learning_rate": 1.6521753705540634e-06, + "loss": 0.2376, + "step": 33687 + }, + { + "epoch": 0.7415364296720229, + "grad_norm": 2.2025720692584736, + "learning_rate": 1.6519106198269313e-06, + "loss": 0.3564, + "step": 33688 + }, + { + "epoch": 0.7415584415584415, + "grad_norm": 4.333213679288895, + "learning_rate": 1.651645886116285e-06, + "loss": 0.3157, + "step": 33689 + }, + { + "epoch": 0.7415804534448602, + "grad_norm": 1.5207775994613733, + "learning_rate": 1.651381169423467e-06, + "loss": 0.26, + "step": 33690 + }, + { + "epoch": 0.7416024653312789, + "grad_norm": 1.5522924670185683, + "learning_rate": 1.6511164697498222e-06, + "loss": 0.2875, + "step": 33691 + }, + { + "epoch": 0.7416244772176975, + "grad_norm": 0.6461762072135268, + "learning_rate": 1.6508517870966967e-06, + "loss": 0.4733, + "step": 33692 + }, + { + "epoch": 0.7416464891041162, + "grad_norm": 1.6577117282856693, + "learning_rate": 1.6505871214654367e-06, + "loss": 0.332, + "step": 33693 + }, + { + "epoch": 0.7416685009905349, + "grad_norm": 0.6184999485555054, + "learning_rate": 1.6503224728573887e-06, + "loss": 0.4599, + "step": 33694 + }, + { + "epoch": 0.7416905128769535, + "grad_norm": 0.5746949672091001, + "learning_rate": 1.6500578412738948e-06, + "loss": 0.436, + "step": 33695 + }, + { + "epoch": 0.7417125247633722, + "grad_norm": 4.093097420743116, + "learning_rate": 1.6497932267163002e-06, + "loss": 0.2946, + "step": 33696 + }, + { + "epoch": 0.7417345366497908, + "grad_norm": 2.1553199443190825, + "learning_rate": 1.6495286291859509e-06, + "loss": 0.3003, + "step": 33697 + }, + { + "epoch": 0.7417565485362095, + "grad_norm": 2.3843166956831485, + "learning_rate": 1.649264048684191e-06, + "loss": 0.2458, + "step": 33698 + }, + { + "epoch": 0.7417785604226282, + "grad_norm": 1.3689301054562966, + "learning_rate": 1.6489994852123676e-06, + "loss": 0.2517, + "step": 33699 + }, + { + "epoch": 0.7418005723090468, + "grad_norm": 1.3617590305380718, + "learning_rate": 1.6487349387718233e-06, + "loss": 0.2573, + "step": 33700 + }, + { + "epoch": 0.7418225841954655, + "grad_norm": 3.708133866648882, + "learning_rate": 1.648470409363901e-06, + "loss": 0.3148, + "step": 33701 + }, + { + "epoch": 0.7418445960818842, + "grad_norm": 1.7138606861182266, + "learning_rate": 1.6482058969899466e-06, + "loss": 0.3613, + "step": 33702 + }, + { + "epoch": 0.7418666079683028, + "grad_norm": 1.5768259354962266, + "learning_rate": 1.647941401651305e-06, + "loss": 0.1941, + "step": 33703 + }, + { + "epoch": 0.7418886198547215, + "grad_norm": 2.4389555438118196, + "learning_rate": 1.6476769233493223e-06, + "loss": 0.3289, + "step": 33704 + }, + { + "epoch": 0.7419106317411402, + "grad_norm": 1.4575731974494224, + "learning_rate": 1.6474124620853405e-06, + "loss": 0.2318, + "step": 33705 + }, + { + "epoch": 0.7419326436275588, + "grad_norm": 1.5501510906819695, + "learning_rate": 1.6471480178607018e-06, + "loss": 0.2661, + "step": 33706 + }, + { + "epoch": 0.7419546555139775, + "grad_norm": 30.808712186666956, + "learning_rate": 1.6468835906767544e-06, + "loss": 0.2246, + "step": 33707 + }, + { + "epoch": 0.7419766674003963, + "grad_norm": 1.6244546400271307, + "learning_rate": 1.6466191805348381e-06, + "loss": 0.3448, + "step": 33708 + }, + { + "epoch": 0.7419986792868148, + "grad_norm": 1.1846368947507797, + "learning_rate": 1.6463547874362989e-06, + "loss": 0.2298, + "step": 33709 + }, + { + "epoch": 0.7420206911732335, + "grad_norm": 0.614417525063528, + "learning_rate": 1.6460904113824822e-06, + "loss": 0.4303, + "step": 33710 + }, + { + "epoch": 0.7420427030596523, + "grad_norm": 1.9092671831773456, + "learning_rate": 1.6458260523747283e-06, + "loss": 0.2754, + "step": 33711 + }, + { + "epoch": 0.7420647149460708, + "grad_norm": 1.4492078449600097, + "learning_rate": 1.645561710414384e-06, + "loss": 0.3199, + "step": 33712 + }, + { + "epoch": 0.7420867268324896, + "grad_norm": 1.882669001126487, + "learning_rate": 1.6452973855027897e-06, + "loss": 0.2763, + "step": 33713 + }, + { + "epoch": 0.7421087387189083, + "grad_norm": 1.428993037950299, + "learning_rate": 1.6450330776412903e-06, + "loss": 0.2269, + "step": 33714 + }, + { + "epoch": 0.7421307506053268, + "grad_norm": 1.5556404475106236, + "learning_rate": 1.6447687868312307e-06, + "loss": 0.2109, + "step": 33715 + }, + { + "epoch": 0.7421527624917456, + "grad_norm": 1.758523876524879, + "learning_rate": 1.6445045130739507e-06, + "loss": 0.2171, + "step": 33716 + }, + { + "epoch": 0.7421747743781643, + "grad_norm": 1.8592128408021589, + "learning_rate": 1.6442402563707971e-06, + "loss": 0.2724, + "step": 33717 + }, + { + "epoch": 0.7421967862645829, + "grad_norm": 1.3316291553599058, + "learning_rate": 1.64397601672311e-06, + "loss": 0.2637, + "step": 33718 + }, + { + "epoch": 0.7422187981510016, + "grad_norm": 0.6349390345484929, + "learning_rate": 1.6437117941322333e-06, + "loss": 0.4419, + "step": 33719 + }, + { + "epoch": 0.7422408100374203, + "grad_norm": 1.759691464661685, + "learning_rate": 1.6434475885995116e-06, + "loss": 0.2907, + "step": 33720 + }, + { + "epoch": 0.7422628219238389, + "grad_norm": 1.669438664061071, + "learning_rate": 1.6431834001262853e-06, + "loss": 0.2334, + "step": 33721 + }, + { + "epoch": 0.7422848338102576, + "grad_norm": 0.6187802295841188, + "learning_rate": 1.6429192287138995e-06, + "loss": 0.4578, + "step": 33722 + }, + { + "epoch": 0.7423068456966762, + "grad_norm": 1.3879127318113844, + "learning_rate": 1.6426550743636938e-06, + "loss": 0.3493, + "step": 33723 + }, + { + "epoch": 0.7423288575830949, + "grad_norm": 1.816342701987023, + "learning_rate": 1.642390937077012e-06, + "loss": 0.2974, + "step": 33724 + }, + { + "epoch": 0.7423508694695136, + "grad_norm": 1.4463878828060985, + "learning_rate": 1.642126816855199e-06, + "loss": 0.284, + "step": 33725 + }, + { + "epoch": 0.7423728813559322, + "grad_norm": 2.277102524394512, + "learning_rate": 1.6418627136995957e-06, + "loss": 0.3063, + "step": 33726 + }, + { + "epoch": 0.7423948932423509, + "grad_norm": 1.5091219659427537, + "learning_rate": 1.6415986276115414e-06, + "loss": 0.3459, + "step": 33727 + }, + { + "epoch": 0.7424169051287696, + "grad_norm": 3.6159079202013484, + "learning_rate": 1.6413345585923812e-06, + "loss": 0.3419, + "step": 33728 + }, + { + "epoch": 0.7424389170151882, + "grad_norm": 0.6374197751639051, + "learning_rate": 1.6410705066434568e-06, + "loss": 0.4684, + "step": 33729 + }, + { + "epoch": 0.7424609289016069, + "grad_norm": 1.530097641326307, + "learning_rate": 1.6408064717661115e-06, + "loss": 0.2833, + "step": 33730 + }, + { + "epoch": 0.7424829407880256, + "grad_norm": 1.462970161600667, + "learning_rate": 1.640542453961686e-06, + "loss": 0.2899, + "step": 33731 + }, + { + "epoch": 0.7425049526744442, + "grad_norm": 3.188805103608633, + "learning_rate": 1.6402784532315209e-06, + "loss": 0.3384, + "step": 33732 + }, + { + "epoch": 0.7425269645608629, + "grad_norm": 3.395863350450244, + "learning_rate": 1.6400144695769587e-06, + "loss": 0.3638, + "step": 33733 + }, + { + "epoch": 0.7425489764472816, + "grad_norm": 2.0085752313977725, + "learning_rate": 1.6397505029993421e-06, + "loss": 0.2706, + "step": 33734 + }, + { + "epoch": 0.7425709883337002, + "grad_norm": 1.4441644190455047, + "learning_rate": 1.6394865535000133e-06, + "loss": 0.2386, + "step": 33735 + }, + { + "epoch": 0.7425930002201189, + "grad_norm": 1.9954701752934265, + "learning_rate": 1.639222621080313e-06, + "loss": 0.2937, + "step": 33736 + }, + { + "epoch": 0.7426150121065376, + "grad_norm": 1.6649824841495948, + "learning_rate": 1.6389587057415806e-06, + "loss": 0.3171, + "step": 33737 + }, + { + "epoch": 0.7426370239929562, + "grad_norm": 1.6973219167920743, + "learning_rate": 1.6386948074851589e-06, + "loss": 0.2949, + "step": 33738 + }, + { + "epoch": 0.7426590358793749, + "grad_norm": 3.9406126667374917, + "learning_rate": 1.6384309263123894e-06, + "loss": 0.3158, + "step": 33739 + }, + { + "epoch": 0.7426810477657936, + "grad_norm": 1.5580948842374045, + "learning_rate": 1.638167062224615e-06, + "loss": 0.201, + "step": 33740 + }, + { + "epoch": 0.7427030596522122, + "grad_norm": 1.735430934802103, + "learning_rate": 1.637903215223175e-06, + "loss": 0.2774, + "step": 33741 + }, + { + "epoch": 0.7427250715386309, + "grad_norm": 1.3695549620954357, + "learning_rate": 1.6376393853094087e-06, + "loss": 0.3192, + "step": 33742 + }, + { + "epoch": 0.7427470834250496, + "grad_norm": 2.0427384573689635, + "learning_rate": 1.63737557248466e-06, + "loss": 0.3704, + "step": 33743 + }, + { + "epoch": 0.7427690953114682, + "grad_norm": 2.408277266258116, + "learning_rate": 1.6371117767502676e-06, + "loss": 0.2839, + "step": 33744 + }, + { + "epoch": 0.7427911071978869, + "grad_norm": 1.3481616899755446, + "learning_rate": 1.6368479981075725e-06, + "loss": 0.2143, + "step": 33745 + }, + { + "epoch": 0.7428131190843055, + "grad_norm": 1.1729069131215357, + "learning_rate": 1.6365842365579176e-06, + "loss": 0.2387, + "step": 33746 + }, + { + "epoch": 0.7428351309707242, + "grad_norm": 1.9431529095809421, + "learning_rate": 1.63632049210264e-06, + "loss": 0.2291, + "step": 33747 + }, + { + "epoch": 0.7428571428571429, + "grad_norm": 1.8693803492756063, + "learning_rate": 1.6360567647430835e-06, + "loss": 0.3716, + "step": 33748 + }, + { + "epoch": 0.7428791547435615, + "grad_norm": 0.6000077390679446, + "learning_rate": 1.6357930544805856e-06, + "loss": 0.4333, + "step": 33749 + }, + { + "epoch": 0.7429011666299802, + "grad_norm": 1.867166373607348, + "learning_rate": 1.6355293613164874e-06, + "loss": 0.2795, + "step": 33750 + }, + { + "epoch": 0.7429231785163989, + "grad_norm": 0.6193234163370438, + "learning_rate": 1.635265685252132e-06, + "loss": 0.4379, + "step": 33751 + }, + { + "epoch": 0.7429451904028175, + "grad_norm": 1.1978883974173602, + "learning_rate": 1.635002026288855e-06, + "loss": 0.326, + "step": 33752 + }, + { + "epoch": 0.7429672022892362, + "grad_norm": 1.7086357953655376, + "learning_rate": 1.6347383844279995e-06, + "loss": 0.3433, + "step": 33753 + }, + { + "epoch": 0.7429892141756549, + "grad_norm": 4.345203913903784, + "learning_rate": 1.6344747596709037e-06, + "loss": 0.2379, + "step": 33754 + }, + { + "epoch": 0.7430112260620735, + "grad_norm": 2.049345131637292, + "learning_rate": 1.634211152018908e-06, + "loss": 0.3185, + "step": 33755 + }, + { + "epoch": 0.7430332379484922, + "grad_norm": 1.8233285754993602, + "learning_rate": 1.6339475614733542e-06, + "loss": 0.2396, + "step": 33756 + }, + { + "epoch": 0.7430552498349109, + "grad_norm": 1.6477328601817425, + "learning_rate": 1.6336839880355782e-06, + "loss": 0.2589, + "step": 33757 + }, + { + "epoch": 0.7430772617213295, + "grad_norm": 1.6570995710245862, + "learning_rate": 1.6334204317069235e-06, + "loss": 0.3467, + "step": 33758 + }, + { + "epoch": 0.7430992736077482, + "grad_norm": 1.3350320505856732, + "learning_rate": 1.633156892488726e-06, + "loss": 0.2835, + "step": 33759 + }, + { + "epoch": 0.7431212854941669, + "grad_norm": 0.5941493900799997, + "learning_rate": 1.632893370382327e-06, + "loss": 0.4371, + "step": 33760 + }, + { + "epoch": 0.7431432973805855, + "grad_norm": 1.717537992990316, + "learning_rate": 1.632629865389067e-06, + "loss": 0.2625, + "step": 33761 + }, + { + "epoch": 0.7431653092670042, + "grad_norm": 8.070096533691874, + "learning_rate": 1.6323663775102822e-06, + "loss": 0.3143, + "step": 33762 + }, + { + "epoch": 0.7431873211534229, + "grad_norm": 1.5068503173950767, + "learning_rate": 1.6321029067473153e-06, + "loss": 0.2774, + "step": 33763 + }, + { + "epoch": 0.7432093330398415, + "grad_norm": 1.3991071243186792, + "learning_rate": 1.631839453101502e-06, + "loss": 0.2287, + "step": 33764 + }, + { + "epoch": 0.7432313449262602, + "grad_norm": 1.3644428027572635, + "learning_rate": 1.6315760165741829e-06, + "loss": 0.2838, + "step": 33765 + }, + { + "epoch": 0.7432533568126789, + "grad_norm": 1.78271849008579, + "learning_rate": 1.6313125971666982e-06, + "loss": 0.2664, + "step": 33766 + }, + { + "epoch": 0.7432753686990975, + "grad_norm": 1.4713195362995715, + "learning_rate": 1.6310491948803858e-06, + "loss": 0.2221, + "step": 33767 + }, + { + "epoch": 0.7432973805855162, + "grad_norm": 1.5044859483377426, + "learning_rate": 1.6307858097165818e-06, + "loss": 0.3337, + "step": 33768 + }, + { + "epoch": 0.7433193924719348, + "grad_norm": 1.5520987878463, + "learning_rate": 1.6305224416766274e-06, + "loss": 0.2625, + "step": 33769 + }, + { + "epoch": 0.7433414043583535, + "grad_norm": 1.503150163292322, + "learning_rate": 1.6302590907618603e-06, + "loss": 0.344, + "step": 33770 + }, + { + "epoch": 0.7433634162447722, + "grad_norm": 2.068477322397525, + "learning_rate": 1.629995756973622e-06, + "loss": 0.2632, + "step": 33771 + }, + { + "epoch": 0.7433854281311908, + "grad_norm": 1.2515207129139991, + "learning_rate": 1.6297324403132476e-06, + "loss": 0.2306, + "step": 33772 + }, + { + "epoch": 0.7434074400176095, + "grad_norm": 0.6167907121067694, + "learning_rate": 1.6294691407820745e-06, + "loss": 0.459, + "step": 33773 + }, + { + "epoch": 0.7434294519040282, + "grad_norm": 1.3206741463679805, + "learning_rate": 1.629205858381443e-06, + "loss": 0.2071, + "step": 33774 + }, + { + "epoch": 0.7434514637904468, + "grad_norm": 2.5734811988826753, + "learning_rate": 1.6289425931126901e-06, + "loss": 0.3111, + "step": 33775 + }, + { + "epoch": 0.7434734756768655, + "grad_norm": 1.4978465565931418, + "learning_rate": 1.6286793449771565e-06, + "loss": 0.3281, + "step": 33776 + }, + { + "epoch": 0.7434954875632842, + "grad_norm": 0.5974788768554858, + "learning_rate": 1.6284161139761778e-06, + "loss": 0.4648, + "step": 33777 + }, + { + "epoch": 0.7435174994497028, + "grad_norm": 1.6853402087408047, + "learning_rate": 1.6281529001110907e-06, + "loss": 0.3535, + "step": 33778 + }, + { + "epoch": 0.7435395113361215, + "grad_norm": 1.4571119471089162, + "learning_rate": 1.6278897033832364e-06, + "loss": 0.2728, + "step": 33779 + }, + { + "epoch": 0.7435615232225402, + "grad_norm": 2.047573353369535, + "learning_rate": 1.6276265237939471e-06, + "loss": 0.2954, + "step": 33780 + }, + { + "epoch": 0.7435835351089588, + "grad_norm": 3.201916270888378, + "learning_rate": 1.6273633613445671e-06, + "loss": 0.382, + "step": 33781 + }, + { + "epoch": 0.7436055469953775, + "grad_norm": 1.632277625561532, + "learning_rate": 1.6271002160364313e-06, + "loss": 0.2952, + "step": 33782 + }, + { + "epoch": 0.7436275588817962, + "grad_norm": 1.9518245573924062, + "learning_rate": 1.6268370878708744e-06, + "loss": 0.2691, + "step": 33783 + }, + { + "epoch": 0.7436495707682148, + "grad_norm": 1.6759389250009964, + "learning_rate": 1.6265739768492372e-06, + "loss": 0.3189, + "step": 33784 + }, + { + "epoch": 0.7436715826546335, + "grad_norm": 1.473495446134809, + "learning_rate": 1.6263108829728547e-06, + "loss": 0.2826, + "step": 33785 + }, + { + "epoch": 0.7436935945410522, + "grad_norm": 2.9497332813343213, + "learning_rate": 1.6260478062430645e-06, + "loss": 0.3602, + "step": 33786 + }, + { + "epoch": 0.7437156064274708, + "grad_norm": 0.6360262314762899, + "learning_rate": 1.6257847466612065e-06, + "loss": 0.4628, + "step": 33787 + }, + { + "epoch": 0.7437376183138895, + "grad_norm": 1.9180098205981604, + "learning_rate": 1.625521704228613e-06, + "loss": 0.3081, + "step": 33788 + }, + { + "epoch": 0.7437596302003082, + "grad_norm": 1.576035866931313, + "learning_rate": 1.6252586789466251e-06, + "loss": 0.3001, + "step": 33789 + }, + { + "epoch": 0.7437816420867268, + "grad_norm": 1.651011360457781, + "learning_rate": 1.6249956708165765e-06, + "loss": 0.3653, + "step": 33790 + }, + { + "epoch": 0.7438036539731455, + "grad_norm": 1.7744798699233209, + "learning_rate": 1.6247326798398045e-06, + "loss": 0.2902, + "step": 33791 + }, + { + "epoch": 0.7438256658595641, + "grad_norm": 1.8423230242282744, + "learning_rate": 1.6244697060176484e-06, + "loss": 0.2525, + "step": 33792 + }, + { + "epoch": 0.7438476777459828, + "grad_norm": 0.6156846377560168, + "learning_rate": 1.6242067493514412e-06, + "loss": 0.4253, + "step": 33793 + }, + { + "epoch": 0.7438696896324015, + "grad_norm": 0.6253927545733994, + "learning_rate": 1.6239438098425225e-06, + "loss": 0.4655, + "step": 33794 + }, + { + "epoch": 0.7438917015188201, + "grad_norm": 4.355325928185904, + "learning_rate": 1.6236808874922261e-06, + "loss": 0.343, + "step": 33795 + }, + { + "epoch": 0.7439137134052388, + "grad_norm": 0.6189952065726774, + "learning_rate": 1.6234179823018887e-06, + "loss": 0.465, + "step": 33796 + }, + { + "epoch": 0.7439357252916575, + "grad_norm": 1.3643786898297792, + "learning_rate": 1.6231550942728485e-06, + "loss": 0.2456, + "step": 33797 + }, + { + "epoch": 0.7439577371780761, + "grad_norm": 1.5172208977840946, + "learning_rate": 1.6228922234064392e-06, + "loss": 0.2236, + "step": 33798 + }, + { + "epoch": 0.7439797490644948, + "grad_norm": 1.516271930343324, + "learning_rate": 1.622629369703999e-06, + "loss": 0.2856, + "step": 33799 + }, + { + "epoch": 0.7440017609509135, + "grad_norm": 1.5999543480900207, + "learning_rate": 1.6223665331668614e-06, + "loss": 0.2725, + "step": 33800 + }, + { + "epoch": 0.7440237728373321, + "grad_norm": 2.506302265089955, + "learning_rate": 1.622103713796363e-06, + "loss": 0.3453, + "step": 33801 + }, + { + "epoch": 0.7440457847237508, + "grad_norm": 1.752881461336936, + "learning_rate": 1.6218409115938422e-06, + "loss": 0.2992, + "step": 33802 + }, + { + "epoch": 0.7440677966101695, + "grad_norm": 0.6193923714423771, + "learning_rate": 1.6215781265606311e-06, + "loss": 0.4363, + "step": 33803 + }, + { + "epoch": 0.7440898084965881, + "grad_norm": 1.294049308655611, + "learning_rate": 1.6213153586980685e-06, + "loss": 0.2881, + "step": 33804 + }, + { + "epoch": 0.7441118203830068, + "grad_norm": 1.5435288495481234, + "learning_rate": 1.6210526080074861e-06, + "loss": 0.2979, + "step": 33805 + }, + { + "epoch": 0.7441338322694255, + "grad_norm": 1.8973803236647486, + "learning_rate": 1.6207898744902213e-06, + "loss": 0.3115, + "step": 33806 + }, + { + "epoch": 0.7441558441558441, + "grad_norm": 0.6267209050051882, + "learning_rate": 1.6205271581476118e-06, + "loss": 0.4468, + "step": 33807 + }, + { + "epoch": 0.7441778560422628, + "grad_norm": 2.123255122663284, + "learning_rate": 1.6202644589809902e-06, + "loss": 0.2338, + "step": 33808 + }, + { + "epoch": 0.7441998679286815, + "grad_norm": 0.6239568500701858, + "learning_rate": 1.6200017769916903e-06, + "loss": 0.471, + "step": 33809 + }, + { + "epoch": 0.7442218798151001, + "grad_norm": 1.6949150218623976, + "learning_rate": 1.6197391121810486e-06, + "loss": 0.2876, + "step": 33810 + }, + { + "epoch": 0.7442438917015188, + "grad_norm": 1.6864805001304366, + "learning_rate": 1.6194764645504008e-06, + "loss": 0.2721, + "step": 33811 + }, + { + "epoch": 0.7442659035879375, + "grad_norm": 2.1779451037668425, + "learning_rate": 1.6192138341010833e-06, + "loss": 0.3322, + "step": 33812 + }, + { + "epoch": 0.7442879154743561, + "grad_norm": 1.7299985022785416, + "learning_rate": 1.6189512208344282e-06, + "loss": 0.2459, + "step": 33813 + }, + { + "epoch": 0.7443099273607748, + "grad_norm": 2.104605036653318, + "learning_rate": 1.6186886247517698e-06, + "loss": 0.2459, + "step": 33814 + }, + { + "epoch": 0.7443319392471934, + "grad_norm": 1.163533613929947, + "learning_rate": 1.6184260458544453e-06, + "loss": 0.2619, + "step": 33815 + }, + { + "epoch": 0.7443539511336121, + "grad_norm": 1.8889458339632896, + "learning_rate": 1.6181634841437848e-06, + "loss": 0.2899, + "step": 33816 + }, + { + "epoch": 0.7443759630200308, + "grad_norm": 0.6164121973063726, + "learning_rate": 1.6179009396211292e-06, + "loss": 0.4598, + "step": 33817 + }, + { + "epoch": 0.7443979749064494, + "grad_norm": 0.6192501637457131, + "learning_rate": 1.6176384122878097e-06, + "loss": 0.4797, + "step": 33818 + }, + { + "epoch": 0.7444199867928681, + "grad_norm": 0.6169055645440412, + "learning_rate": 1.6173759021451586e-06, + "loss": 0.4543, + "step": 33819 + }, + { + "epoch": 0.7444419986792868, + "grad_norm": 1.754301601349994, + "learning_rate": 1.6171134091945128e-06, + "loss": 0.2894, + "step": 33820 + }, + { + "epoch": 0.7444640105657054, + "grad_norm": 1.7368471878673375, + "learning_rate": 1.6168509334372047e-06, + "loss": 0.3119, + "step": 33821 + }, + { + "epoch": 0.7444860224521241, + "grad_norm": 2.070904001309, + "learning_rate": 1.6165884748745685e-06, + "loss": 0.2616, + "step": 33822 + }, + { + "epoch": 0.7445080343385428, + "grad_norm": 1.4164192142569678, + "learning_rate": 1.6163260335079406e-06, + "loss": 0.3187, + "step": 33823 + }, + { + "epoch": 0.7445300462249614, + "grad_norm": 0.607540126041423, + "learning_rate": 1.616063609338651e-06, + "loss": 0.4409, + "step": 33824 + }, + { + "epoch": 0.7445520581113801, + "grad_norm": 1.3114291230485258, + "learning_rate": 1.6158012023680375e-06, + "loss": 0.2359, + "step": 33825 + }, + { + "epoch": 0.7445740699977988, + "grad_norm": 0.6052431142773942, + "learning_rate": 1.6155388125974297e-06, + "loss": 0.4468, + "step": 33826 + }, + { + "epoch": 0.7445960818842174, + "grad_norm": 1.5272161576534693, + "learning_rate": 1.6152764400281633e-06, + "loss": 0.2528, + "step": 33827 + }, + { + "epoch": 0.7446180937706361, + "grad_norm": 10.815163119367265, + "learning_rate": 1.6150140846615737e-06, + "loss": 0.2951, + "step": 33828 + }, + { + "epoch": 0.7446401056570549, + "grad_norm": 1.4438076761553436, + "learning_rate": 1.6147517464989903e-06, + "loss": 0.2317, + "step": 33829 + }, + { + "epoch": 0.7446621175434734, + "grad_norm": 1.533557772813071, + "learning_rate": 1.6144894255417498e-06, + "loss": 0.2306, + "step": 33830 + }, + { + "epoch": 0.7446841294298921, + "grad_norm": 1.715830074898841, + "learning_rate": 1.6142271217911825e-06, + "loss": 0.3083, + "step": 33831 + }, + { + "epoch": 0.7447061413163109, + "grad_norm": 1.7424642733837008, + "learning_rate": 1.6139648352486232e-06, + "loss": 0.2454, + "step": 33832 + }, + { + "epoch": 0.7447281532027294, + "grad_norm": 1.8707366657034923, + "learning_rate": 1.6137025659154065e-06, + "loss": 0.254, + "step": 33833 + }, + { + "epoch": 0.7447501650891482, + "grad_norm": 2.7178478787544655, + "learning_rate": 1.6134403137928617e-06, + "loss": 0.3209, + "step": 33834 + }, + { + "epoch": 0.7447721769755669, + "grad_norm": 1.2930760509134054, + "learning_rate": 1.613178078882326e-06, + "loss": 0.2356, + "step": 33835 + }, + { + "epoch": 0.7447941888619855, + "grad_norm": 1.8786116390707865, + "learning_rate": 1.6129158611851276e-06, + "loss": 0.251, + "step": 33836 + }, + { + "epoch": 0.7448162007484042, + "grad_norm": 1.3844997531497227, + "learning_rate": 1.6126536607026016e-06, + "loss": 0.2577, + "step": 33837 + }, + { + "epoch": 0.7448382126348227, + "grad_norm": 1.703848757850832, + "learning_rate": 1.6123914774360821e-06, + "loss": 0.3127, + "step": 33838 + }, + { + "epoch": 0.7448602245212415, + "grad_norm": 1.535797567826756, + "learning_rate": 1.6121293113868985e-06, + "loss": 0.3054, + "step": 33839 + }, + { + "epoch": 0.7448822364076602, + "grad_norm": 0.6041398005272681, + "learning_rate": 1.6118671625563864e-06, + "loss": 0.4327, + "step": 33840 + }, + { + "epoch": 0.7449042482940788, + "grad_norm": 1.899950141013881, + "learning_rate": 1.6116050309458747e-06, + "loss": 0.3371, + "step": 33841 + }, + { + "epoch": 0.7449262601804975, + "grad_norm": 1.8348444096954708, + "learning_rate": 1.611342916556698e-06, + "loss": 0.2814, + "step": 33842 + }, + { + "epoch": 0.7449482720669162, + "grad_norm": 1.6568259947786483, + "learning_rate": 1.6110808193901888e-06, + "loss": 0.2859, + "step": 33843 + }, + { + "epoch": 0.7449702839533348, + "grad_norm": 1.373902705073635, + "learning_rate": 1.6108187394476794e-06, + "loss": 0.2413, + "step": 33844 + }, + { + "epoch": 0.7449922958397535, + "grad_norm": 1.781708882604008, + "learning_rate": 1.6105566767304986e-06, + "loss": 0.2163, + "step": 33845 + }, + { + "epoch": 0.7450143077261722, + "grad_norm": 0.6095090147664269, + "learning_rate": 1.610294631239982e-06, + "loss": 0.4596, + "step": 33846 + }, + { + "epoch": 0.7450363196125908, + "grad_norm": 2.230539275977839, + "learning_rate": 1.6100326029774571e-06, + "loss": 0.3336, + "step": 33847 + }, + { + "epoch": 0.7450583314990095, + "grad_norm": 1.3845072509521934, + "learning_rate": 1.6097705919442613e-06, + "loss": 0.3085, + "step": 33848 + }, + { + "epoch": 0.7450803433854282, + "grad_norm": 1.312146188600398, + "learning_rate": 1.6095085981417235e-06, + "loss": 0.3099, + "step": 33849 + }, + { + "epoch": 0.7451023552718468, + "grad_norm": 1.8271304398068127, + "learning_rate": 1.6092466215711738e-06, + "loss": 0.2452, + "step": 33850 + }, + { + "epoch": 0.7451243671582655, + "grad_norm": 1.4419055130708789, + "learning_rate": 1.6089846622339467e-06, + "loss": 0.2177, + "step": 33851 + }, + { + "epoch": 0.7451463790446842, + "grad_norm": 2.739351776025421, + "learning_rate": 1.6087227201313682e-06, + "loss": 0.3152, + "step": 33852 + }, + { + "epoch": 0.7451683909311028, + "grad_norm": 1.3241961053511144, + "learning_rate": 1.6084607952647773e-06, + "loss": 0.2132, + "step": 33853 + }, + { + "epoch": 0.7451904028175215, + "grad_norm": 0.6351829227288148, + "learning_rate": 1.6081988876355009e-06, + "loss": 0.4513, + "step": 33854 + }, + { + "epoch": 0.7452124147039402, + "grad_norm": 1.973057796783898, + "learning_rate": 1.6079369972448693e-06, + "loss": 0.2489, + "step": 33855 + }, + { + "epoch": 0.7452344265903588, + "grad_norm": 1.585437012986392, + "learning_rate": 1.6076751240942162e-06, + "loss": 0.2884, + "step": 33856 + }, + { + "epoch": 0.7452564384767775, + "grad_norm": 2.1925471466467195, + "learning_rate": 1.6074132681848697e-06, + "loss": 0.2927, + "step": 33857 + }, + { + "epoch": 0.7452784503631962, + "grad_norm": 0.6328560506773416, + "learning_rate": 1.6071514295181617e-06, + "loss": 0.4726, + "step": 33858 + }, + { + "epoch": 0.7453004622496148, + "grad_norm": 1.7403841354882708, + "learning_rate": 1.6068896080954255e-06, + "loss": 0.2831, + "step": 33859 + }, + { + "epoch": 0.7453224741360335, + "grad_norm": 1.2377317685285962, + "learning_rate": 1.6066278039179878e-06, + "loss": 0.2476, + "step": 33860 + }, + { + "epoch": 0.7453444860224522, + "grad_norm": 2.874803610401563, + "learning_rate": 1.6063660169871825e-06, + "loss": 0.2595, + "step": 33861 + }, + { + "epoch": 0.7453664979088708, + "grad_norm": 1.8518020732239027, + "learning_rate": 1.606104247304337e-06, + "loss": 0.2467, + "step": 33862 + }, + { + "epoch": 0.7453885097952895, + "grad_norm": 2.0312558252019377, + "learning_rate": 1.605842494870784e-06, + "loss": 0.2627, + "step": 33863 + }, + { + "epoch": 0.7454105216817081, + "grad_norm": 1.2950210719729232, + "learning_rate": 1.6055807596878548e-06, + "loss": 0.2411, + "step": 33864 + }, + { + "epoch": 0.7454325335681268, + "grad_norm": 1.9767825248960196, + "learning_rate": 1.6053190417568759e-06, + "loss": 0.3669, + "step": 33865 + }, + { + "epoch": 0.7454545454545455, + "grad_norm": 1.7634753255400473, + "learning_rate": 1.605057341079182e-06, + "loss": 0.2993, + "step": 33866 + }, + { + "epoch": 0.7454765573409641, + "grad_norm": 1.3321860799590528, + "learning_rate": 1.6047956576560991e-06, + "loss": 0.2793, + "step": 33867 + }, + { + "epoch": 0.7454985692273828, + "grad_norm": 2.775877530962491, + "learning_rate": 1.6045339914889585e-06, + "loss": 0.3343, + "step": 33868 + }, + { + "epoch": 0.7455205811138015, + "grad_norm": 1.24105902627006, + "learning_rate": 1.604272342579093e-06, + "loss": 0.2872, + "step": 33869 + }, + { + "epoch": 0.7455425930002201, + "grad_norm": 1.4852818085867405, + "learning_rate": 1.604010710927828e-06, + "loss": 0.2889, + "step": 33870 + }, + { + "epoch": 0.7455646048866388, + "grad_norm": 1.4912207646881015, + "learning_rate": 1.6037490965364965e-06, + "loss": 0.2996, + "step": 33871 + }, + { + "epoch": 0.7455866167730575, + "grad_norm": 1.6146548570446504, + "learning_rate": 1.6034874994064253e-06, + "loss": 0.3388, + "step": 33872 + }, + { + "epoch": 0.7456086286594761, + "grad_norm": 2.178649694065295, + "learning_rate": 1.603225919538946e-06, + "loss": 0.2564, + "step": 33873 + }, + { + "epoch": 0.7456306405458948, + "grad_norm": 0.5922750904400892, + "learning_rate": 1.6029643569353893e-06, + "loss": 0.4355, + "step": 33874 + }, + { + "epoch": 0.7456526524323135, + "grad_norm": 1.7624988319748636, + "learning_rate": 1.6027028115970806e-06, + "loss": 0.2384, + "step": 33875 + }, + { + "epoch": 0.7456746643187321, + "grad_norm": 1.451973351214039, + "learning_rate": 1.6024412835253534e-06, + "loss": 0.2391, + "step": 33876 + }, + { + "epoch": 0.7456966762051508, + "grad_norm": 2.1278468626076883, + "learning_rate": 1.6021797727215331e-06, + "loss": 0.3341, + "step": 33877 + }, + { + "epoch": 0.7457186880915695, + "grad_norm": 1.5197909463399695, + "learning_rate": 1.601918279186951e-06, + "loss": 0.2698, + "step": 33878 + }, + { + "epoch": 0.7457406999779881, + "grad_norm": 1.7216280369292878, + "learning_rate": 1.6016568029229373e-06, + "loss": 0.2435, + "step": 33879 + }, + { + "epoch": 0.7457627118644068, + "grad_norm": 1.4276659372705187, + "learning_rate": 1.6013953439308178e-06, + "loss": 0.2418, + "step": 33880 + }, + { + "epoch": 0.7457847237508255, + "grad_norm": 1.3225977376262001, + "learning_rate": 1.6011339022119244e-06, + "loss": 0.2755, + "step": 33881 + }, + { + "epoch": 0.7458067356372441, + "grad_norm": 0.6129762133637984, + "learning_rate": 1.6008724777675844e-06, + "loss": 0.4681, + "step": 33882 + }, + { + "epoch": 0.7458287475236628, + "grad_norm": 0.6045207566026766, + "learning_rate": 1.6006110705991235e-06, + "loss": 0.4576, + "step": 33883 + }, + { + "epoch": 0.7458507594100815, + "grad_norm": 1.301904254083383, + "learning_rate": 1.6003496807078766e-06, + "loss": 0.3084, + "step": 33884 + }, + { + "epoch": 0.7458727712965001, + "grad_norm": 1.9293712285895068, + "learning_rate": 1.6000883080951691e-06, + "loss": 0.2486, + "step": 33885 + }, + { + "epoch": 0.7458947831829188, + "grad_norm": 1.5686729344538097, + "learning_rate": 1.5998269527623272e-06, + "loss": 0.2981, + "step": 33886 + }, + { + "epoch": 0.7459167950693374, + "grad_norm": 1.9759925557769105, + "learning_rate": 1.599565614710683e-06, + "loss": 0.2606, + "step": 33887 + }, + { + "epoch": 0.7459388069557561, + "grad_norm": 1.75031185151333, + "learning_rate": 1.5993042939415598e-06, + "loss": 0.2849, + "step": 33888 + }, + { + "epoch": 0.7459608188421748, + "grad_norm": 1.547506231321053, + "learning_rate": 1.5990429904562921e-06, + "loss": 0.2558, + "step": 33889 + }, + { + "epoch": 0.7459828307285934, + "grad_norm": 1.9402371857318672, + "learning_rate": 1.5987817042562048e-06, + "loss": 0.3224, + "step": 33890 + }, + { + "epoch": 0.7460048426150121, + "grad_norm": 1.3514928465115648, + "learning_rate": 1.5985204353426243e-06, + "loss": 0.208, + "step": 33891 + }, + { + "epoch": 0.7460268545014308, + "grad_norm": 1.2888639045709511, + "learning_rate": 1.5982591837168815e-06, + "loss": 0.2414, + "step": 33892 + }, + { + "epoch": 0.7460488663878494, + "grad_norm": 1.2174844898149553, + "learning_rate": 1.5979979493803016e-06, + "loss": 0.2469, + "step": 33893 + }, + { + "epoch": 0.7460708782742681, + "grad_norm": 1.397453886437032, + "learning_rate": 1.597736732334213e-06, + "loss": 0.3528, + "step": 33894 + }, + { + "epoch": 0.7460928901606868, + "grad_norm": 1.668236522323664, + "learning_rate": 1.5974755325799458e-06, + "loss": 0.3074, + "step": 33895 + }, + { + "epoch": 0.7461149020471054, + "grad_norm": 1.4543895786741643, + "learning_rate": 1.5972143501188236e-06, + "loss": 0.2479, + "step": 33896 + }, + { + "epoch": 0.7461369139335241, + "grad_norm": 2.647482157605836, + "learning_rate": 1.5969531849521774e-06, + "loss": 0.3424, + "step": 33897 + }, + { + "epoch": 0.7461589258199428, + "grad_norm": 1.4314619444328391, + "learning_rate": 1.596692037081331e-06, + "loss": 0.2646, + "step": 33898 + }, + { + "epoch": 0.7461809377063614, + "grad_norm": 1.3976974152693382, + "learning_rate": 1.5964309065076139e-06, + "loss": 0.2825, + "step": 33899 + }, + { + "epoch": 0.7462029495927801, + "grad_norm": 0.6228278639692472, + "learning_rate": 1.5961697932323543e-06, + "loss": 0.4257, + "step": 33900 + }, + { + "epoch": 0.7462249614791988, + "grad_norm": 1.3529101578112295, + "learning_rate": 1.595908697256877e-06, + "loss": 0.2103, + "step": 33901 + }, + { + "epoch": 0.7462469733656174, + "grad_norm": 1.5052011116097765, + "learning_rate": 1.5956476185825109e-06, + "loss": 0.2912, + "step": 33902 + }, + { + "epoch": 0.7462689852520361, + "grad_norm": 2.4958743334975737, + "learning_rate": 1.595386557210581e-06, + "loss": 0.3434, + "step": 33903 + }, + { + "epoch": 0.7462909971384548, + "grad_norm": 1.5806485332487814, + "learning_rate": 1.5951255131424147e-06, + "loss": 0.3272, + "step": 33904 + }, + { + "epoch": 0.7463130090248734, + "grad_norm": 1.826446130294637, + "learning_rate": 1.5948644863793406e-06, + "loss": 0.3518, + "step": 33905 + }, + { + "epoch": 0.7463350209112921, + "grad_norm": 1.93402639070898, + "learning_rate": 1.5946034769226826e-06, + "loss": 0.2889, + "step": 33906 + }, + { + "epoch": 0.7463570327977108, + "grad_norm": 1.3497540212121475, + "learning_rate": 1.5943424847737704e-06, + "loss": 0.3054, + "step": 33907 + }, + { + "epoch": 0.7463790446841294, + "grad_norm": 1.696902028550082, + "learning_rate": 1.594081509933927e-06, + "loss": 0.1936, + "step": 33908 + }, + { + "epoch": 0.7464010565705481, + "grad_norm": 1.3343230806457493, + "learning_rate": 1.5938205524044797e-06, + "loss": 0.2456, + "step": 33909 + }, + { + "epoch": 0.7464230684569667, + "grad_norm": 2.9783867214345117, + "learning_rate": 1.593559612186758e-06, + "loss": 0.3048, + "step": 33910 + }, + { + "epoch": 0.7464450803433854, + "grad_norm": 0.622362651954556, + "learning_rate": 1.5932986892820834e-06, + "loss": 0.4658, + "step": 33911 + }, + { + "epoch": 0.7464670922298041, + "grad_norm": 0.6119858359824588, + "learning_rate": 1.5930377836917866e-06, + "loss": 0.4538, + "step": 33912 + }, + { + "epoch": 0.7464891041162227, + "grad_norm": 1.6202873401345983, + "learning_rate": 1.592776895417189e-06, + "loss": 0.2391, + "step": 33913 + }, + { + "epoch": 0.7465111160026414, + "grad_norm": 0.6216523087528064, + "learning_rate": 1.5925160244596195e-06, + "loss": 0.4344, + "step": 33914 + }, + { + "epoch": 0.7465331278890601, + "grad_norm": 1.3002481389717002, + "learning_rate": 1.5922551708204048e-06, + "loss": 0.2626, + "step": 33915 + }, + { + "epoch": 0.7465551397754787, + "grad_norm": 1.8309084435772718, + "learning_rate": 1.5919943345008677e-06, + "loss": 0.3949, + "step": 33916 + }, + { + "epoch": 0.7465771516618974, + "grad_norm": 2.1915799812007757, + "learning_rate": 1.5917335155023368e-06, + "loss": 0.3039, + "step": 33917 + }, + { + "epoch": 0.7465991635483161, + "grad_norm": 1.5106327724875785, + "learning_rate": 1.591472713826136e-06, + "loss": 0.3404, + "step": 33918 + }, + { + "epoch": 0.7466211754347347, + "grad_norm": 0.6044933807314831, + "learning_rate": 1.5912119294735885e-06, + "loss": 0.4445, + "step": 33919 + }, + { + "epoch": 0.7466431873211534, + "grad_norm": 2.367165249491323, + "learning_rate": 1.590951162446025e-06, + "loss": 0.2741, + "step": 33920 + }, + { + "epoch": 0.7466651992075721, + "grad_norm": 1.2337026177962003, + "learning_rate": 1.590690412744767e-06, + "loss": 0.2826, + "step": 33921 + }, + { + "epoch": 0.7466872110939907, + "grad_norm": 1.23105248331483, + "learning_rate": 1.5904296803711422e-06, + "loss": 0.2475, + "step": 33922 + }, + { + "epoch": 0.7467092229804094, + "grad_norm": 1.229025686933072, + "learning_rate": 1.5901689653264747e-06, + "loss": 0.2898, + "step": 33923 + }, + { + "epoch": 0.7467312348668281, + "grad_norm": 0.6432565104569657, + "learning_rate": 1.5899082676120864e-06, + "loss": 0.498, + "step": 33924 + }, + { + "epoch": 0.7467532467532467, + "grad_norm": 0.6262764144096952, + "learning_rate": 1.5896475872293076e-06, + "loss": 0.4679, + "step": 33925 + }, + { + "epoch": 0.7467752586396654, + "grad_norm": 0.5960616070887196, + "learning_rate": 1.589386924179462e-06, + "loss": 0.4266, + "step": 33926 + }, + { + "epoch": 0.7467972705260841, + "grad_norm": 1.4109703198118284, + "learning_rate": 1.5891262784638706e-06, + "loss": 0.293, + "step": 33927 + }, + { + "epoch": 0.7468192824125027, + "grad_norm": 3.042659969271211, + "learning_rate": 1.5888656500838629e-06, + "loss": 0.2899, + "step": 33928 + }, + { + "epoch": 0.7468412942989214, + "grad_norm": 1.6338551312516256, + "learning_rate": 1.5886050390407593e-06, + "loss": 0.2531, + "step": 33929 + }, + { + "epoch": 0.7468633061853401, + "grad_norm": 2.340799322857954, + "learning_rate": 1.5883444453358865e-06, + "loss": 0.305, + "step": 33930 + }, + { + "epoch": 0.7468853180717587, + "grad_norm": 1.6435236881877433, + "learning_rate": 1.5880838689705708e-06, + "loss": 0.2954, + "step": 33931 + }, + { + "epoch": 0.7469073299581774, + "grad_norm": 0.6297210056034349, + "learning_rate": 1.5878233099461333e-06, + "loss": 0.4666, + "step": 33932 + }, + { + "epoch": 0.746929341844596, + "grad_norm": 1.3961734287735885, + "learning_rate": 1.5875627682639006e-06, + "loss": 0.2264, + "step": 33933 + }, + { + "epoch": 0.7469513537310147, + "grad_norm": 2.31357670918979, + "learning_rate": 1.5873022439251944e-06, + "loss": 0.3231, + "step": 33934 + }, + { + "epoch": 0.7469733656174334, + "grad_norm": 0.6046884086260836, + "learning_rate": 1.5870417369313407e-06, + "loss": 0.4623, + "step": 33935 + }, + { + "epoch": 0.746995377503852, + "grad_norm": 3.0024478830874353, + "learning_rate": 1.5867812472836642e-06, + "loss": 0.281, + "step": 33936 + }, + { + "epoch": 0.7470173893902707, + "grad_norm": 1.3163165717499614, + "learning_rate": 1.5865207749834865e-06, + "loss": 0.2059, + "step": 33937 + }, + { + "epoch": 0.7470394012766894, + "grad_norm": 1.3153328531364068, + "learning_rate": 1.5862603200321341e-06, + "loss": 0.2682, + "step": 33938 + }, + { + "epoch": 0.747061413163108, + "grad_norm": 1.6782268907951337, + "learning_rate": 1.5859998824309275e-06, + "loss": 0.2892, + "step": 33939 + }, + { + "epoch": 0.7470834250495267, + "grad_norm": 1.8095609265571706, + "learning_rate": 1.5857394621811928e-06, + "loss": 0.239, + "step": 33940 + }, + { + "epoch": 0.7471054369359454, + "grad_norm": 4.298853470886822, + "learning_rate": 1.585479059284254e-06, + "loss": 0.3391, + "step": 33941 + }, + { + "epoch": 0.747127448822364, + "grad_norm": 1.5645308019748916, + "learning_rate": 1.585218673741432e-06, + "loss": 0.3162, + "step": 33942 + }, + { + "epoch": 0.7471494607087827, + "grad_norm": 1.4828756278992938, + "learning_rate": 1.5849583055540535e-06, + "loss": 0.302, + "step": 33943 + }, + { + "epoch": 0.7471714725952014, + "grad_norm": 0.618852509136738, + "learning_rate": 1.5846979547234386e-06, + "loss": 0.4313, + "step": 33944 + }, + { + "epoch": 0.74719348448162, + "grad_norm": 1.5140503562999952, + "learning_rate": 1.5844376212509115e-06, + "loss": 0.2777, + "step": 33945 + }, + { + "epoch": 0.7472154963680387, + "grad_norm": 0.621559182466908, + "learning_rate": 1.5841773051377978e-06, + "loss": 0.4545, + "step": 33946 + }, + { + "epoch": 0.7472375082544574, + "grad_norm": 1.9042070256351267, + "learning_rate": 1.583917006385417e-06, + "loss": 0.2859, + "step": 33947 + }, + { + "epoch": 0.747259520140876, + "grad_norm": 2.2538084568917953, + "learning_rate": 1.583656724995095e-06, + "loss": 0.1948, + "step": 33948 + }, + { + "epoch": 0.7472815320272947, + "grad_norm": 20.77453083549015, + "learning_rate": 1.583396460968152e-06, + "loss": 0.2999, + "step": 33949 + }, + { + "epoch": 0.7473035439137135, + "grad_norm": 1.3269308984483037, + "learning_rate": 1.5831362143059115e-06, + "loss": 0.2189, + "step": 33950 + }, + { + "epoch": 0.747325555800132, + "grad_norm": 2.7303971768273345, + "learning_rate": 1.5828759850096987e-06, + "loss": 0.3503, + "step": 33951 + }, + { + "epoch": 0.7473475676865508, + "grad_norm": 3.886115847791795, + "learning_rate": 1.5826157730808322e-06, + "loss": 0.2984, + "step": 33952 + }, + { + "epoch": 0.7473695795729695, + "grad_norm": 0.5945077240320783, + "learning_rate": 1.5823555785206384e-06, + "loss": 0.4638, + "step": 33953 + }, + { + "epoch": 0.747391591459388, + "grad_norm": 1.7048907421555972, + "learning_rate": 1.582095401330438e-06, + "loss": 0.2599, + "step": 33954 + }, + { + "epoch": 0.7474136033458068, + "grad_norm": 0.6151068494933941, + "learning_rate": 1.5818352415115495e-06, + "loss": 0.4519, + "step": 33955 + }, + { + "epoch": 0.7474356152322253, + "grad_norm": 0.6050259626189135, + "learning_rate": 1.5815750990653029e-06, + "loss": 0.4376, + "step": 33956 + }, + { + "epoch": 0.747457627118644, + "grad_norm": 1.2857208048009992, + "learning_rate": 1.581314973993014e-06, + "loss": 0.3214, + "step": 33957 + }, + { + "epoch": 0.7474796390050628, + "grad_norm": 1.4635366988721803, + "learning_rate": 1.581054866296009e-06, + "loss": 0.2649, + "step": 33958 + }, + { + "epoch": 0.7475016508914814, + "grad_norm": 1.492361517759896, + "learning_rate": 1.580794775975608e-06, + "loss": 0.3343, + "step": 33959 + }, + { + "epoch": 0.7475236627779, + "grad_norm": 1.8629637158138448, + "learning_rate": 1.5805347030331297e-06, + "loss": 0.2936, + "step": 33960 + }, + { + "epoch": 0.7475456746643188, + "grad_norm": 1.6102004572496849, + "learning_rate": 1.5802746474699027e-06, + "loss": 0.2688, + "step": 33961 + }, + { + "epoch": 0.7475676865507374, + "grad_norm": 0.6033504847926712, + "learning_rate": 1.5800146092872431e-06, + "loss": 0.4361, + "step": 33962 + }, + { + "epoch": 0.7475896984371561, + "grad_norm": 1.6473043525145283, + "learning_rate": 1.5797545884864761e-06, + "loss": 0.3551, + "step": 33963 + }, + { + "epoch": 0.7476117103235748, + "grad_norm": 2.06543892771046, + "learning_rate": 1.5794945850689225e-06, + "loss": 0.3214, + "step": 33964 + }, + { + "epoch": 0.7476337222099934, + "grad_norm": 1.7675458204756798, + "learning_rate": 1.579234599035901e-06, + "loss": 0.3106, + "step": 33965 + }, + { + "epoch": 0.7476557340964121, + "grad_norm": 1.3862859531855405, + "learning_rate": 1.5789746303887349e-06, + "loss": 0.2731, + "step": 33966 + }, + { + "epoch": 0.7476777459828308, + "grad_norm": 1.7977470364318808, + "learning_rate": 1.5787146791287471e-06, + "loss": 0.4485, + "step": 33967 + }, + { + "epoch": 0.7476997578692494, + "grad_norm": 1.886476219588831, + "learning_rate": 1.5784547452572558e-06, + "loss": 0.2252, + "step": 33968 + }, + { + "epoch": 0.7477217697556681, + "grad_norm": 1.4253303793964525, + "learning_rate": 1.5781948287755856e-06, + "loss": 0.2498, + "step": 33969 + }, + { + "epoch": 0.7477437816420868, + "grad_norm": 1.8504246212756965, + "learning_rate": 1.5779349296850533e-06, + "loss": 0.2543, + "step": 33970 + }, + { + "epoch": 0.7477657935285054, + "grad_norm": 2.159658214015021, + "learning_rate": 1.5776750479869818e-06, + "loss": 0.3686, + "step": 33971 + }, + { + "epoch": 0.7477878054149241, + "grad_norm": 0.6108162319639429, + "learning_rate": 1.577415183682694e-06, + "loss": 0.4255, + "step": 33972 + }, + { + "epoch": 0.7478098173013428, + "grad_norm": 1.3818010119515214, + "learning_rate": 1.577155336773507e-06, + "loss": 0.1946, + "step": 33973 + }, + { + "epoch": 0.7478318291877614, + "grad_norm": 1.1574273236028132, + "learning_rate": 1.5768955072607451e-06, + "loss": 0.2201, + "step": 33974 + }, + { + "epoch": 0.7478538410741801, + "grad_norm": 3.045775940095139, + "learning_rate": 1.576635695145725e-06, + "loss": 0.2903, + "step": 33975 + }, + { + "epoch": 0.7478758529605988, + "grad_norm": 1.835595965573373, + "learning_rate": 1.5763759004297692e-06, + "loss": 0.2977, + "step": 33976 + }, + { + "epoch": 0.7478978648470174, + "grad_norm": 1.3245696615419609, + "learning_rate": 1.5761161231141997e-06, + "loss": 0.2476, + "step": 33977 + }, + { + "epoch": 0.7479198767334361, + "grad_norm": 0.6124064193765596, + "learning_rate": 1.5758563632003337e-06, + "loss": 0.4486, + "step": 33978 + }, + { + "epoch": 0.7479418886198547, + "grad_norm": 1.64307564865322, + "learning_rate": 1.5755966206894946e-06, + "loss": 0.3199, + "step": 33979 + }, + { + "epoch": 0.7479639005062734, + "grad_norm": 0.6175352530761643, + "learning_rate": 1.5753368955829989e-06, + "loss": 0.4493, + "step": 33980 + }, + { + "epoch": 0.7479859123926921, + "grad_norm": 2.7865987019038303, + "learning_rate": 1.5750771878821686e-06, + "loss": 0.3019, + "step": 33981 + }, + { + "epoch": 0.7480079242791107, + "grad_norm": 2.297470590946983, + "learning_rate": 1.5748174975883251e-06, + "loss": 0.3089, + "step": 33982 + }, + { + "epoch": 0.7480299361655294, + "grad_norm": 1.8954010247487139, + "learning_rate": 1.5745578247027854e-06, + "loss": 0.2816, + "step": 33983 + }, + { + "epoch": 0.7480519480519481, + "grad_norm": 4.060171581459068, + "learning_rate": 1.5742981692268722e-06, + "loss": 0.3818, + "step": 33984 + }, + { + "epoch": 0.7480739599383667, + "grad_norm": 4.016357762812684, + "learning_rate": 1.5740385311619017e-06, + "loss": 0.2912, + "step": 33985 + }, + { + "epoch": 0.7480959718247854, + "grad_norm": 1.6020723103538637, + "learning_rate": 1.5737789105091956e-06, + "loss": 0.229, + "step": 33986 + }, + { + "epoch": 0.7481179837112041, + "grad_norm": 3.1477591379216805, + "learning_rate": 1.5735193072700749e-06, + "loss": 0.3079, + "step": 33987 + }, + { + "epoch": 0.7481399955976227, + "grad_norm": 0.6032922659488509, + "learning_rate": 1.5732597214458556e-06, + "loss": 0.4461, + "step": 33988 + }, + { + "epoch": 0.7481620074840414, + "grad_norm": 1.6316595685553863, + "learning_rate": 1.5730001530378607e-06, + "loss": 0.3789, + "step": 33989 + }, + { + "epoch": 0.7481840193704601, + "grad_norm": 2.1101989840706405, + "learning_rate": 1.5727406020474073e-06, + "loss": 0.3454, + "step": 33990 + }, + { + "epoch": 0.7482060312568787, + "grad_norm": 1.6091129627237075, + "learning_rate": 1.5724810684758114e-06, + "loss": 0.2776, + "step": 33991 + }, + { + "epoch": 0.7482280431432974, + "grad_norm": 1.9974615835022647, + "learning_rate": 1.5722215523243988e-06, + "loss": 0.2565, + "step": 33992 + }, + { + "epoch": 0.7482500550297161, + "grad_norm": 1.8860084083856492, + "learning_rate": 1.5719620535944835e-06, + "loss": 0.3286, + "step": 33993 + }, + { + "epoch": 0.7482720669161347, + "grad_norm": 1.2193978417249465, + "learning_rate": 1.5717025722873874e-06, + "loss": 0.2927, + "step": 33994 + }, + { + "epoch": 0.7482940788025534, + "grad_norm": 1.481254616654994, + "learning_rate": 1.571443108404428e-06, + "loss": 0.37, + "step": 33995 + }, + { + "epoch": 0.7483160906889721, + "grad_norm": 1.3157079142365946, + "learning_rate": 1.5711836619469223e-06, + "loss": 0.2894, + "step": 33996 + }, + { + "epoch": 0.7483381025753907, + "grad_norm": 2.0886509147071233, + "learning_rate": 1.5709242329161905e-06, + "loss": 0.2673, + "step": 33997 + }, + { + "epoch": 0.7483601144618094, + "grad_norm": 3.7615838778569777, + "learning_rate": 1.570664821313551e-06, + "loss": 0.3202, + "step": 33998 + }, + { + "epoch": 0.7483821263482281, + "grad_norm": 0.5993528346727985, + "learning_rate": 1.5704054271403247e-06, + "loss": 0.4497, + "step": 33999 + }, + { + "epoch": 0.7484041382346467, + "grad_norm": 4.0946277044592225, + "learning_rate": 1.5701460503978272e-06, + "loss": 0.2725, + "step": 34000 + }, + { + "epoch": 0.7484261501210654, + "grad_norm": 1.5965462481483979, + "learning_rate": 1.5698866910873756e-06, + "loss": 0.2942, + "step": 34001 + }, + { + "epoch": 0.7484481620074841, + "grad_norm": 1.5278802565449536, + "learning_rate": 1.5696273492102892e-06, + "loss": 0.2789, + "step": 34002 + }, + { + "epoch": 0.7484701738939027, + "grad_norm": 1.3872396322327367, + "learning_rate": 1.5693680247678873e-06, + "loss": 0.278, + "step": 34003 + }, + { + "epoch": 0.7484921857803214, + "grad_norm": 1.914599135419359, + "learning_rate": 1.5691087177614889e-06, + "loss": 0.2468, + "step": 34004 + }, + { + "epoch": 0.74851419766674, + "grad_norm": 0.6179214901644821, + "learning_rate": 1.56884942819241e-06, + "loss": 0.4584, + "step": 34005 + }, + { + "epoch": 0.7485362095531587, + "grad_norm": 1.8404771738149936, + "learning_rate": 1.5685901560619666e-06, + "loss": 0.3408, + "step": 34006 + }, + { + "epoch": 0.7485582214395774, + "grad_norm": 1.392459678770574, + "learning_rate": 1.5683309013714793e-06, + "loss": 0.2617, + "step": 34007 + }, + { + "epoch": 0.748580233325996, + "grad_norm": 1.9760058668084177, + "learning_rate": 1.568071664122266e-06, + "loss": 0.2521, + "step": 34008 + }, + { + "epoch": 0.7486022452124147, + "grad_norm": 1.6379158407119336, + "learning_rate": 1.5678124443156416e-06, + "loss": 0.2204, + "step": 34009 + }, + { + "epoch": 0.7486242570988334, + "grad_norm": 1.4066665470330368, + "learning_rate": 1.5675532419529266e-06, + "loss": 0.3074, + "step": 34010 + }, + { + "epoch": 0.748646268985252, + "grad_norm": 1.547593966409446, + "learning_rate": 1.5672940570354356e-06, + "loss": 0.3183, + "step": 34011 + }, + { + "epoch": 0.7486682808716707, + "grad_norm": 1.6369664119911504, + "learning_rate": 1.5670348895644871e-06, + "loss": 0.3175, + "step": 34012 + }, + { + "epoch": 0.7486902927580894, + "grad_norm": 8.309669074014868, + "learning_rate": 1.5667757395414001e-06, + "loss": 0.283, + "step": 34013 + }, + { + "epoch": 0.748712304644508, + "grad_norm": 0.610089832429893, + "learning_rate": 1.5665166069674887e-06, + "loss": 0.479, + "step": 34014 + }, + { + "epoch": 0.7487343165309267, + "grad_norm": 1.3218142396739492, + "learning_rate": 1.5662574918440726e-06, + "loss": 0.3099, + "step": 34015 + }, + { + "epoch": 0.7487563284173454, + "grad_norm": 2.451467933615069, + "learning_rate": 1.565998394172466e-06, + "loss": 0.2767, + "step": 34016 + }, + { + "epoch": 0.748778340303764, + "grad_norm": 1.4024632693937569, + "learning_rate": 1.5657393139539866e-06, + "loss": 0.3112, + "step": 34017 + }, + { + "epoch": 0.7488003521901827, + "grad_norm": 1.5879904558930225, + "learning_rate": 1.5654802511899543e-06, + "loss": 0.3268, + "step": 34018 + }, + { + "epoch": 0.7488223640766014, + "grad_norm": 1.572135751635971, + "learning_rate": 1.5652212058816806e-06, + "loss": 0.3433, + "step": 34019 + }, + { + "epoch": 0.74884437596302, + "grad_norm": 0.5947443933007048, + "learning_rate": 1.564962178030487e-06, + "loss": 0.4573, + "step": 34020 + }, + { + "epoch": 0.7488663878494387, + "grad_norm": 1.7653693351393789, + "learning_rate": 1.5647031676376862e-06, + "loss": 0.2999, + "step": 34021 + }, + { + "epoch": 0.7488883997358574, + "grad_norm": 1.6368534178984133, + "learning_rate": 1.5644441747045952e-06, + "loss": 0.2502, + "step": 34022 + }, + { + "epoch": 0.748910411622276, + "grad_norm": 1.7797443722840471, + "learning_rate": 1.5641851992325336e-06, + "loss": 0.2806, + "step": 34023 + }, + { + "epoch": 0.7489324235086947, + "grad_norm": 1.2805668446955158, + "learning_rate": 1.5639262412228135e-06, + "loss": 0.243, + "step": 34024 + }, + { + "epoch": 0.7489544353951134, + "grad_norm": 2.472217037946019, + "learning_rate": 1.5636673006767538e-06, + "loss": 0.3237, + "step": 34025 + }, + { + "epoch": 0.748976447281532, + "grad_norm": 19.614521591734967, + "learning_rate": 1.5634083775956703e-06, + "loss": 0.2286, + "step": 34026 + }, + { + "epoch": 0.7489984591679507, + "grad_norm": 1.628527639496063, + "learning_rate": 1.5631494719808749e-06, + "loss": 0.2627, + "step": 34027 + }, + { + "epoch": 0.7490204710543693, + "grad_norm": 1.4043050259951948, + "learning_rate": 1.5628905838336894e-06, + "loss": 0.2836, + "step": 34028 + }, + { + "epoch": 0.749042482940788, + "grad_norm": 1.4532202460406214, + "learning_rate": 1.562631713155426e-06, + "loss": 0.3101, + "step": 34029 + }, + { + "epoch": 0.7490644948272067, + "grad_norm": 0.6025593745368529, + "learning_rate": 1.5623728599474025e-06, + "loss": 0.4388, + "step": 34030 + }, + { + "epoch": 0.7490865067136253, + "grad_norm": 3.1377941428513627, + "learning_rate": 1.5621140242109333e-06, + "loss": 0.3527, + "step": 34031 + }, + { + "epoch": 0.749108518600044, + "grad_norm": 0.6229424362044309, + "learning_rate": 1.5618552059473325e-06, + "loss": 0.4568, + "step": 34032 + }, + { + "epoch": 0.7491305304864627, + "grad_norm": 1.6455259071307564, + "learning_rate": 1.5615964051579168e-06, + "loss": 0.3484, + "step": 34033 + }, + { + "epoch": 0.7491525423728813, + "grad_norm": 1.577141079396255, + "learning_rate": 1.5613376218440025e-06, + "loss": 0.2648, + "step": 34034 + }, + { + "epoch": 0.7491745542593, + "grad_norm": 1.6363025255789003, + "learning_rate": 1.5610788560069045e-06, + "loss": 0.2684, + "step": 34035 + }, + { + "epoch": 0.7491965661457187, + "grad_norm": 1.3417762528990418, + "learning_rate": 1.5608201076479384e-06, + "loss": 0.2518, + "step": 34036 + }, + { + "epoch": 0.7492185780321373, + "grad_norm": 2.0402228180429436, + "learning_rate": 1.5605613767684169e-06, + "loss": 0.3219, + "step": 34037 + }, + { + "epoch": 0.749240589918556, + "grad_norm": 2.069154696804115, + "learning_rate": 1.5603026633696561e-06, + "loss": 0.3235, + "step": 34038 + }, + { + "epoch": 0.7492626018049747, + "grad_norm": 2.0687942322685524, + "learning_rate": 1.5600439674529716e-06, + "loss": 0.3511, + "step": 34039 + }, + { + "epoch": 0.7492846136913933, + "grad_norm": 0.6127239781980617, + "learning_rate": 1.559785289019679e-06, + "loss": 0.4459, + "step": 34040 + }, + { + "epoch": 0.749306625577812, + "grad_norm": 2.145522366926153, + "learning_rate": 1.5595266280710924e-06, + "loss": 0.2692, + "step": 34041 + }, + { + "epoch": 0.7493286374642307, + "grad_norm": 2.0230288667628873, + "learning_rate": 1.5592679846085245e-06, + "loss": 0.2906, + "step": 34042 + }, + { + "epoch": 0.7493506493506493, + "grad_norm": 1.579623323672696, + "learning_rate": 1.559009358633291e-06, + "loss": 0.256, + "step": 34043 + }, + { + "epoch": 0.749372661237068, + "grad_norm": 0.6309984660550708, + "learning_rate": 1.5587507501467087e-06, + "loss": 0.449, + "step": 34044 + }, + { + "epoch": 0.7493946731234867, + "grad_norm": 1.627440122241496, + "learning_rate": 1.558492159150088e-06, + "loss": 0.286, + "step": 34045 + }, + { + "epoch": 0.7494166850099053, + "grad_norm": 1.2790423160049909, + "learning_rate": 1.5582335856447468e-06, + "loss": 0.2415, + "step": 34046 + }, + { + "epoch": 0.749438696896324, + "grad_norm": 1.9979115118706114, + "learning_rate": 1.5579750296319957e-06, + "loss": 0.3326, + "step": 34047 + }, + { + "epoch": 0.7494607087827427, + "grad_norm": 0.6189790899761628, + "learning_rate": 1.557716491113151e-06, + "loss": 0.4617, + "step": 34048 + }, + { + "epoch": 0.7494827206691613, + "grad_norm": 1.4765403503397176, + "learning_rate": 1.5574579700895277e-06, + "loss": 0.2772, + "step": 34049 + }, + { + "epoch": 0.74950473255558, + "grad_norm": 1.6138078364543675, + "learning_rate": 1.5571994665624367e-06, + "loss": 0.3262, + "step": 34050 + }, + { + "epoch": 0.7495267444419986, + "grad_norm": 0.590921846672655, + "learning_rate": 1.5569409805331959e-06, + "loss": 0.4431, + "step": 34051 + }, + { + "epoch": 0.7495487563284173, + "grad_norm": 2.0895777208604036, + "learning_rate": 1.5566825120031143e-06, + "loss": 0.3368, + "step": 34052 + }, + { + "epoch": 0.749570768214836, + "grad_norm": 1.2355486858868534, + "learning_rate": 1.5564240609735082e-06, + "loss": 0.2409, + "step": 34053 + }, + { + "epoch": 0.7495927801012546, + "grad_norm": 1.6149117895712768, + "learning_rate": 1.5561656274456926e-06, + "loss": 0.2801, + "step": 34054 + }, + { + "epoch": 0.7496147919876733, + "grad_norm": 3.8717673353222666, + "learning_rate": 1.5559072114209777e-06, + "loss": 0.2663, + "step": 34055 + }, + { + "epoch": 0.749636803874092, + "grad_norm": 1.7888896945246944, + "learning_rate": 1.5556488129006802e-06, + "loss": 0.3904, + "step": 34056 + }, + { + "epoch": 0.7496588157605106, + "grad_norm": 0.6205822610626066, + "learning_rate": 1.5553904318861096e-06, + "loss": 0.4661, + "step": 34057 + }, + { + "epoch": 0.7496808276469293, + "grad_norm": 0.6103051481028425, + "learning_rate": 1.5551320683785814e-06, + "loss": 0.4519, + "step": 34058 + }, + { + "epoch": 0.749702839533348, + "grad_norm": 2.073036840488775, + "learning_rate": 1.5548737223794103e-06, + "loss": 0.3145, + "step": 34059 + }, + { + "epoch": 0.7497248514197666, + "grad_norm": 0.6074650505331636, + "learning_rate": 1.5546153938899055e-06, + "loss": 0.4411, + "step": 34060 + }, + { + "epoch": 0.7497468633061853, + "grad_norm": 0.612212423993728, + "learning_rate": 1.554357082911384e-06, + "loss": 0.4631, + "step": 34061 + }, + { + "epoch": 0.749768875192604, + "grad_norm": 0.6145788043507409, + "learning_rate": 1.5540987894451548e-06, + "loss": 0.4684, + "step": 34062 + }, + { + "epoch": 0.7497908870790226, + "grad_norm": 1.5162031637256437, + "learning_rate": 1.5538405134925322e-06, + "loss": 0.3533, + "step": 34063 + }, + { + "epoch": 0.7498128989654413, + "grad_norm": 1.4221746737853873, + "learning_rate": 1.5535822550548313e-06, + "loss": 0.2733, + "step": 34064 + }, + { + "epoch": 0.74983491085186, + "grad_norm": 1.5975303612911362, + "learning_rate": 1.5533240141333606e-06, + "loss": 0.3295, + "step": 34065 + }, + { + "epoch": 0.7498569227382786, + "grad_norm": 3.138645851163138, + "learning_rate": 1.5530657907294362e-06, + "loss": 0.2437, + "step": 34066 + }, + { + "epoch": 0.7498789346246973, + "grad_norm": 2.413929985966876, + "learning_rate": 1.5528075848443686e-06, + "loss": 0.3101, + "step": 34067 + }, + { + "epoch": 0.749900946511116, + "grad_norm": 0.6108054227142714, + "learning_rate": 1.5525493964794685e-06, + "loss": 0.4527, + "step": 34068 + }, + { + "epoch": 0.7499229583975346, + "grad_norm": 0.6096903247935853, + "learning_rate": 1.5522912256360506e-06, + "loss": 0.4478, + "step": 34069 + }, + { + "epoch": 0.7499449702839533, + "grad_norm": 0.6069539156596245, + "learning_rate": 1.5520330723154258e-06, + "loss": 0.4568, + "step": 34070 + }, + { + "epoch": 0.749966982170372, + "grad_norm": 1.7820311917508307, + "learning_rate": 1.551774936518909e-06, + "loss": 0.2808, + "step": 34071 + }, + { + "epoch": 0.7499889940567906, + "grad_norm": 2.0363440913077735, + "learning_rate": 1.5515168182478096e-06, + "loss": 0.3917, + "step": 34072 + }, + { + "epoch": 0.7500110059432094, + "grad_norm": 3.598955281680569, + "learning_rate": 1.5512587175034382e-06, + "loss": 0.2844, + "step": 34073 + }, + { + "epoch": 0.750033017829628, + "grad_norm": 1.5587972421209193, + "learning_rate": 1.5510006342871082e-06, + "loss": 0.3548, + "step": 34074 + }, + { + "epoch": 0.7500550297160467, + "grad_norm": 1.619204945814757, + "learning_rate": 1.5507425686001316e-06, + "loss": 0.2727, + "step": 34075 + }, + { + "epoch": 0.7500770416024654, + "grad_norm": 1.5551039197466343, + "learning_rate": 1.5504845204438207e-06, + "loss": 0.2372, + "step": 34076 + }, + { + "epoch": 0.750099053488884, + "grad_norm": 1.1640629097259385, + "learning_rate": 1.5502264898194862e-06, + "loss": 0.292, + "step": 34077 + }, + { + "epoch": 0.7501210653753027, + "grad_norm": 2.6448570980614816, + "learning_rate": 1.549968476728438e-06, + "loss": 0.2817, + "step": 34078 + }, + { + "epoch": 0.7501430772617214, + "grad_norm": 1.4846730973989783, + "learning_rate": 1.5497104811719887e-06, + "loss": 0.3176, + "step": 34079 + }, + { + "epoch": 0.75016508914814, + "grad_norm": 1.5914656748446825, + "learning_rate": 1.5494525031514496e-06, + "loss": 0.2486, + "step": 34080 + }, + { + "epoch": 0.7501871010345587, + "grad_norm": 1.540237072780581, + "learning_rate": 1.5491945426681333e-06, + "loss": 0.2892, + "step": 34081 + }, + { + "epoch": 0.7502091129209774, + "grad_norm": 2.1173663017989974, + "learning_rate": 1.5489365997233492e-06, + "loss": 0.2908, + "step": 34082 + }, + { + "epoch": 0.750231124807396, + "grad_norm": 1.697285865186186, + "learning_rate": 1.5486786743184073e-06, + "loss": 0.3814, + "step": 34083 + }, + { + "epoch": 0.7502531366938147, + "grad_norm": 1.0987362033282517, + "learning_rate": 1.5484207664546198e-06, + "loss": 0.2611, + "step": 34084 + }, + { + "epoch": 0.7502751485802334, + "grad_norm": 1.6142393931690522, + "learning_rate": 1.5481628761332985e-06, + "loss": 0.2569, + "step": 34085 + }, + { + "epoch": 0.750297160466652, + "grad_norm": 1.6733496256278046, + "learning_rate": 1.5479050033557518e-06, + "loss": 0.2401, + "step": 34086 + }, + { + "epoch": 0.7503191723530707, + "grad_norm": 3.140795902160338, + "learning_rate": 1.547647148123293e-06, + "loss": 0.2677, + "step": 34087 + }, + { + "epoch": 0.7503411842394894, + "grad_norm": 1.349296393024346, + "learning_rate": 1.5473893104372295e-06, + "loss": 0.2593, + "step": 34088 + }, + { + "epoch": 0.750363196125908, + "grad_norm": 0.6270070615636952, + "learning_rate": 1.5471314902988737e-06, + "loss": 0.4576, + "step": 34089 + }, + { + "epoch": 0.7503852080123267, + "grad_norm": 2.175892105337705, + "learning_rate": 1.5468736877095376e-06, + "loss": 0.3479, + "step": 34090 + }, + { + "epoch": 0.7504072198987454, + "grad_norm": 2.711246656583083, + "learning_rate": 1.5466159026705274e-06, + "loss": 0.2062, + "step": 34091 + }, + { + "epoch": 0.750429231785164, + "grad_norm": 2.06597864824401, + "learning_rate": 1.546358135183157e-06, + "loss": 0.2824, + "step": 34092 + }, + { + "epoch": 0.7504512436715827, + "grad_norm": 2.898705038420633, + "learning_rate": 1.5461003852487333e-06, + "loss": 0.2683, + "step": 34093 + }, + { + "epoch": 0.7504732555580014, + "grad_norm": 1.253614795481313, + "learning_rate": 1.5458426528685684e-06, + "loss": 0.2634, + "step": 34094 + }, + { + "epoch": 0.75049526744442, + "grad_norm": 1.3963798843413997, + "learning_rate": 1.545584938043973e-06, + "loss": 0.3212, + "step": 34095 + }, + { + "epoch": 0.7505172793308387, + "grad_norm": 1.5808522036780202, + "learning_rate": 1.5453272407762542e-06, + "loss": 0.2697, + "step": 34096 + }, + { + "epoch": 0.7505392912172573, + "grad_norm": 1.509030890950468, + "learning_rate": 1.5450695610667244e-06, + "loss": 0.2837, + "step": 34097 + }, + { + "epoch": 0.750561303103676, + "grad_norm": 1.5853210580654735, + "learning_rate": 1.544811898916691e-06, + "loss": 0.3464, + "step": 34098 + }, + { + "epoch": 0.7505833149900947, + "grad_norm": 1.6181694007567753, + "learning_rate": 1.5445542543274645e-06, + "loss": 0.3805, + "step": 34099 + }, + { + "epoch": 0.7506053268765133, + "grad_norm": 1.4214384761412708, + "learning_rate": 1.5442966273003557e-06, + "loss": 0.2635, + "step": 34100 + }, + { + "epoch": 0.750627338762932, + "grad_norm": 2.2452425527358764, + "learning_rate": 1.544039017836671e-06, + "loss": 0.3055, + "step": 34101 + }, + { + "epoch": 0.7506493506493507, + "grad_norm": 1.385093019381353, + "learning_rate": 1.5437814259377237e-06, + "loss": 0.365, + "step": 34102 + }, + { + "epoch": 0.7506713625357693, + "grad_norm": 1.7718080308793376, + "learning_rate": 1.5435238516048185e-06, + "loss": 0.3452, + "step": 34103 + }, + { + "epoch": 0.750693374422188, + "grad_norm": 1.887098821220876, + "learning_rate": 1.5432662948392684e-06, + "loss": 0.3371, + "step": 34104 + }, + { + "epoch": 0.7507153863086067, + "grad_norm": 1.5336444319448617, + "learning_rate": 1.5430087556423796e-06, + "loss": 0.2888, + "step": 34105 + }, + { + "epoch": 0.7507373981950253, + "grad_norm": 3.4190648675590567, + "learning_rate": 1.5427512340154615e-06, + "loss": 0.3584, + "step": 34106 + }, + { + "epoch": 0.750759410081444, + "grad_norm": 29.872979547469825, + "learning_rate": 1.5424937299598258e-06, + "loss": 0.2697, + "step": 34107 + }, + { + "epoch": 0.7507814219678627, + "grad_norm": 2.577926643083182, + "learning_rate": 1.542236243476779e-06, + "loss": 0.3636, + "step": 34108 + }, + { + "epoch": 0.7508034338542813, + "grad_norm": 1.4694968834491906, + "learning_rate": 1.541978774567628e-06, + "loss": 0.2336, + "step": 34109 + }, + { + "epoch": 0.7508254457407, + "grad_norm": 1.9163919927425914, + "learning_rate": 1.541721323233683e-06, + "loss": 0.264, + "step": 34110 + }, + { + "epoch": 0.7508474576271187, + "grad_norm": 1.9614666064523918, + "learning_rate": 1.541463889476253e-06, + "loss": 0.2452, + "step": 34111 + }, + { + "epoch": 0.7508694695135373, + "grad_norm": 2.0335418678778536, + "learning_rate": 1.5412064732966475e-06, + "loss": 0.2427, + "step": 34112 + }, + { + "epoch": 0.750891481399956, + "grad_norm": 1.3178761410060986, + "learning_rate": 1.5409490746961737e-06, + "loss": 0.2869, + "step": 34113 + }, + { + "epoch": 0.7509134932863747, + "grad_norm": 1.7185119648368707, + "learning_rate": 1.5406916936761374e-06, + "loss": 0.2763, + "step": 34114 + }, + { + "epoch": 0.7509355051727933, + "grad_norm": 0.6229186024945762, + "learning_rate": 1.5404343302378489e-06, + "loss": 0.4591, + "step": 34115 + }, + { + "epoch": 0.750957517059212, + "grad_norm": 0.5899558818589776, + "learning_rate": 1.5401769843826165e-06, + "loss": 0.4473, + "step": 34116 + }, + { + "epoch": 0.7509795289456307, + "grad_norm": 1.8005540334164651, + "learning_rate": 1.539919656111749e-06, + "loss": 0.293, + "step": 34117 + }, + { + "epoch": 0.7510015408320493, + "grad_norm": 1.2320493376425532, + "learning_rate": 1.539662345426553e-06, + "loss": 0.3025, + "step": 34118 + }, + { + "epoch": 0.751023552718468, + "grad_norm": 2.2284773183360835, + "learning_rate": 1.539405052328335e-06, + "loss": 0.3345, + "step": 34119 + }, + { + "epoch": 0.7510455646048866, + "grad_norm": 3.195843314208797, + "learning_rate": 1.5391477768184038e-06, + "loss": 0.3422, + "step": 34120 + }, + { + "epoch": 0.7510675764913053, + "grad_norm": 1.7570346002609887, + "learning_rate": 1.5388905188980674e-06, + "loss": 0.2713, + "step": 34121 + }, + { + "epoch": 0.751089588377724, + "grad_norm": 1.9071420989234908, + "learning_rate": 1.5386332785686342e-06, + "loss": 0.3165, + "step": 34122 + }, + { + "epoch": 0.7511116002641426, + "grad_norm": 1.4529050628004756, + "learning_rate": 1.538376055831411e-06, + "loss": 0.2013, + "step": 34123 + }, + { + "epoch": 0.7511336121505613, + "grad_norm": 1.3787161409982494, + "learning_rate": 1.5381188506877032e-06, + "loss": 0.3351, + "step": 34124 + }, + { + "epoch": 0.75115562403698, + "grad_norm": 3.07205177569964, + "learning_rate": 1.5378616631388188e-06, + "loss": 0.3027, + "step": 34125 + }, + { + "epoch": 0.7511776359233986, + "grad_norm": 0.6323324068656011, + "learning_rate": 1.5376044931860673e-06, + "loss": 0.4617, + "step": 34126 + }, + { + "epoch": 0.7511996478098173, + "grad_norm": 3.4606181271061334, + "learning_rate": 1.5373473408307526e-06, + "loss": 0.3151, + "step": 34127 + }, + { + "epoch": 0.751221659696236, + "grad_norm": 1.4106007724529745, + "learning_rate": 1.537090206074185e-06, + "loss": 0.2359, + "step": 34128 + }, + { + "epoch": 0.7512436715826546, + "grad_norm": 1.8683403227838062, + "learning_rate": 1.5368330889176675e-06, + "loss": 0.3262, + "step": 34129 + }, + { + "epoch": 0.7512656834690733, + "grad_norm": 0.6396446143712965, + "learning_rate": 1.5365759893625088e-06, + "loss": 0.4625, + "step": 34130 + }, + { + "epoch": 0.751287695355492, + "grad_norm": 1.6729373864886945, + "learning_rate": 1.5363189074100176e-06, + "loss": 0.2842, + "step": 34131 + }, + { + "epoch": 0.7513097072419106, + "grad_norm": 3.7425414050117145, + "learning_rate": 1.5360618430614965e-06, + "loss": 0.3324, + "step": 34132 + }, + { + "epoch": 0.7513317191283293, + "grad_norm": 1.5025580191587498, + "learning_rate": 1.535804796318256e-06, + "loss": 0.308, + "step": 34133 + }, + { + "epoch": 0.751353731014748, + "grad_norm": 1.5762170843208578, + "learning_rate": 1.5355477671815993e-06, + "loss": 0.2602, + "step": 34134 + }, + { + "epoch": 0.7513757429011666, + "grad_norm": 2.095294954127348, + "learning_rate": 1.5352907556528335e-06, + "loss": 0.3361, + "step": 34135 + }, + { + "epoch": 0.7513977547875853, + "grad_norm": 2.8454493552378644, + "learning_rate": 1.5350337617332672e-06, + "loss": 0.3577, + "step": 34136 + }, + { + "epoch": 0.751419766674004, + "grad_norm": 1.5125843165287876, + "learning_rate": 1.5347767854242029e-06, + "loss": 0.3377, + "step": 34137 + }, + { + "epoch": 0.7514417785604226, + "grad_norm": 1.3896019044254686, + "learning_rate": 1.5345198267269506e-06, + "loss": 0.2539, + "step": 34138 + }, + { + "epoch": 0.7514637904468413, + "grad_norm": 2.131636389018279, + "learning_rate": 1.534262885642812e-06, + "loss": 0.2872, + "step": 34139 + }, + { + "epoch": 0.75148580233326, + "grad_norm": 1.7808691717853273, + "learning_rate": 1.5340059621730974e-06, + "loss": 0.2906, + "step": 34140 + }, + { + "epoch": 0.7515078142196786, + "grad_norm": 1.923845058939839, + "learning_rate": 1.5337490563191081e-06, + "loss": 0.3549, + "step": 34141 + }, + { + "epoch": 0.7515298261060973, + "grad_norm": 30.693847019892125, + "learning_rate": 1.5334921680821524e-06, + "loss": 0.2334, + "step": 34142 + }, + { + "epoch": 0.7515518379925159, + "grad_norm": 1.5021632617249352, + "learning_rate": 1.5332352974635373e-06, + "loss": 0.2472, + "step": 34143 + }, + { + "epoch": 0.7515738498789346, + "grad_norm": 1.8176246565593661, + "learning_rate": 1.5329784444645668e-06, + "loss": 0.2613, + "step": 34144 + }, + { + "epoch": 0.7515958617653533, + "grad_norm": 1.5735489931063467, + "learning_rate": 1.5327216090865444e-06, + "loss": 0.2856, + "step": 34145 + }, + { + "epoch": 0.7516178736517719, + "grad_norm": 2.735215522601264, + "learning_rate": 1.5324647913307772e-06, + "loss": 0.3251, + "step": 34146 + }, + { + "epoch": 0.7516398855381906, + "grad_norm": 1.9916280870835756, + "learning_rate": 1.5322079911985704e-06, + "loss": 0.2574, + "step": 34147 + }, + { + "epoch": 0.7516618974246093, + "grad_norm": 1.953210869795087, + "learning_rate": 1.531951208691231e-06, + "loss": 0.2961, + "step": 34148 + }, + { + "epoch": 0.7516839093110279, + "grad_norm": 2.4817983821576264, + "learning_rate": 1.5316944438100622e-06, + "loss": 0.2679, + "step": 34149 + }, + { + "epoch": 0.7517059211974466, + "grad_norm": 1.5883949283822396, + "learning_rate": 1.5314376965563677e-06, + "loss": 0.241, + "step": 34150 + }, + { + "epoch": 0.7517279330838653, + "grad_norm": 2.043381483936195, + "learning_rate": 1.5311809669314536e-06, + "loss": 0.2996, + "step": 34151 + }, + { + "epoch": 0.7517499449702839, + "grad_norm": 1.5602925807185264, + "learning_rate": 1.5309242549366254e-06, + "loss": 0.3056, + "step": 34152 + }, + { + "epoch": 0.7517719568567026, + "grad_norm": 1.3241364957318689, + "learning_rate": 1.5306675605731885e-06, + "loss": 0.2336, + "step": 34153 + }, + { + "epoch": 0.7517939687431213, + "grad_norm": 1.8191871925438095, + "learning_rate": 1.5304108838424469e-06, + "loss": 0.297, + "step": 34154 + }, + { + "epoch": 0.7518159806295399, + "grad_norm": 1.261226432304983, + "learning_rate": 1.530154224745703e-06, + "loss": 0.1541, + "step": 34155 + }, + { + "epoch": 0.7518379925159586, + "grad_norm": 1.865770019861322, + "learning_rate": 1.5298975832842621e-06, + "loss": 0.2726, + "step": 34156 + }, + { + "epoch": 0.7518600044023773, + "grad_norm": 1.812546749218797, + "learning_rate": 1.5296409594594302e-06, + "loss": 0.3579, + "step": 34157 + }, + { + "epoch": 0.7518820162887959, + "grad_norm": 1.1545266816880124, + "learning_rate": 1.5293843532725123e-06, + "loss": 0.3013, + "step": 34158 + }, + { + "epoch": 0.7519040281752146, + "grad_norm": 1.9549837966592485, + "learning_rate": 1.529127764724811e-06, + "loss": 0.2506, + "step": 34159 + }, + { + "epoch": 0.7519260400616333, + "grad_norm": 1.4741250826438468, + "learning_rate": 1.5288711938176287e-06, + "loss": 0.2522, + "step": 34160 + }, + { + "epoch": 0.7519480519480519, + "grad_norm": 1.2435110841650305, + "learning_rate": 1.5286146405522716e-06, + "loss": 0.2733, + "step": 34161 + }, + { + "epoch": 0.7519700638344706, + "grad_norm": 0.6042039606804168, + "learning_rate": 1.5283581049300427e-06, + "loss": 0.443, + "step": 34162 + }, + { + "epoch": 0.7519920757208893, + "grad_norm": 2.185829532730387, + "learning_rate": 1.528101586952248e-06, + "loss": 0.3155, + "step": 34163 + }, + { + "epoch": 0.7520140876073079, + "grad_norm": 1.6356816337609466, + "learning_rate": 1.52784508662019e-06, + "loss": 0.3651, + "step": 34164 + }, + { + "epoch": 0.7520360994937266, + "grad_norm": 0.6164791762864682, + "learning_rate": 1.5275886039351695e-06, + "loss": 0.4373, + "step": 34165 + }, + { + "epoch": 0.7520581113801453, + "grad_norm": 1.9518377845884274, + "learning_rate": 1.527332138898493e-06, + "loss": 0.2723, + "step": 34166 + }, + { + "epoch": 0.7520801232665639, + "grad_norm": 2.308795075924132, + "learning_rate": 1.527075691511465e-06, + "loss": 0.3344, + "step": 34167 + }, + { + "epoch": 0.7521021351529826, + "grad_norm": 2.45250438298906, + "learning_rate": 1.5268192617753852e-06, + "loss": 0.2718, + "step": 34168 + }, + { + "epoch": 0.7521241470394012, + "grad_norm": 1.3292347432850302, + "learning_rate": 1.5265628496915613e-06, + "loss": 0.2761, + "step": 34169 + }, + { + "epoch": 0.7521461589258199, + "grad_norm": 3.242564459378329, + "learning_rate": 1.5263064552612922e-06, + "loss": 0.2867, + "step": 34170 + }, + { + "epoch": 0.7521681708122386, + "grad_norm": 0.622016693266836, + "learning_rate": 1.526050078485883e-06, + "loss": 0.4361, + "step": 34171 + }, + { + "epoch": 0.7521901826986572, + "grad_norm": 1.7911302097141273, + "learning_rate": 1.5257937193666383e-06, + "loss": 0.3163, + "step": 34172 + }, + { + "epoch": 0.7522121945850759, + "grad_norm": 1.788949283759827, + "learning_rate": 1.5255373779048582e-06, + "loss": 0.4487, + "step": 34173 + }, + { + "epoch": 0.7522342064714946, + "grad_norm": 0.6100830950726113, + "learning_rate": 1.5252810541018482e-06, + "loss": 0.4425, + "step": 34174 + }, + { + "epoch": 0.7522562183579132, + "grad_norm": 1.4661931710054612, + "learning_rate": 1.525024747958908e-06, + "loss": 0.2676, + "step": 34175 + }, + { + "epoch": 0.7522782302443319, + "grad_norm": 1.2937663285916268, + "learning_rate": 1.5247684594773431e-06, + "loss": 0.2512, + "step": 34176 + }, + { + "epoch": 0.7523002421307506, + "grad_norm": 1.6202391609743025, + "learning_rate": 1.5245121886584541e-06, + "loss": 0.3174, + "step": 34177 + }, + { + "epoch": 0.7523222540171692, + "grad_norm": 4.435382226317197, + "learning_rate": 1.524255935503544e-06, + "loss": 0.3051, + "step": 34178 + }, + { + "epoch": 0.7523442659035879, + "grad_norm": 1.9382452176305975, + "learning_rate": 1.5239997000139167e-06, + "loss": 0.2904, + "step": 34179 + }, + { + "epoch": 0.7523662777900066, + "grad_norm": 1.34949271536389, + "learning_rate": 1.523743482190872e-06, + "loss": 0.2441, + "step": 34180 + }, + { + "epoch": 0.7523882896764252, + "grad_norm": 0.6212159329257759, + "learning_rate": 1.5234872820357149e-06, + "loss": 0.4545, + "step": 34181 + }, + { + "epoch": 0.7524103015628439, + "grad_norm": 1.4670973312470676, + "learning_rate": 1.5232310995497446e-06, + "loss": 0.2365, + "step": 34182 + }, + { + "epoch": 0.7524323134492626, + "grad_norm": 2.4918502020075626, + "learning_rate": 1.5229749347342643e-06, + "loss": 0.31, + "step": 34183 + }, + { + "epoch": 0.7524543253356812, + "grad_norm": 1.8954310558760379, + "learning_rate": 1.5227187875905775e-06, + "loss": 0.3007, + "step": 34184 + }, + { + "epoch": 0.7524763372221, + "grad_norm": 1.8427779337064927, + "learning_rate": 1.5224626581199853e-06, + "loss": 0.3317, + "step": 34185 + }, + { + "epoch": 0.7524983491085186, + "grad_norm": 1.9858098965568067, + "learning_rate": 1.5222065463237866e-06, + "loss": 0.2514, + "step": 34186 + }, + { + "epoch": 0.7525203609949372, + "grad_norm": 1.788687355842327, + "learning_rate": 1.5219504522032858e-06, + "loss": 0.3097, + "step": 34187 + }, + { + "epoch": 0.752542372881356, + "grad_norm": 10.298347660465502, + "learning_rate": 1.5216943757597841e-06, + "loss": 0.2926, + "step": 34188 + }, + { + "epoch": 0.7525643847677747, + "grad_norm": 1.2392945390895014, + "learning_rate": 1.5214383169945846e-06, + "loss": 0.2506, + "step": 34189 + }, + { + "epoch": 0.7525863966541932, + "grad_norm": 1.69935570204118, + "learning_rate": 1.5211822759089872e-06, + "loss": 0.2994, + "step": 34190 + }, + { + "epoch": 0.752608408540612, + "grad_norm": 1.620671702197782, + "learning_rate": 1.520926252504291e-06, + "loss": 0.3029, + "step": 34191 + }, + { + "epoch": 0.7526304204270305, + "grad_norm": 0.614176564895617, + "learning_rate": 1.5206702467817997e-06, + "loss": 0.4365, + "step": 34192 + }, + { + "epoch": 0.7526524323134492, + "grad_norm": 3.765586410598047, + "learning_rate": 1.520414258742814e-06, + "loss": 0.272, + "step": 34193 + }, + { + "epoch": 0.752674444199868, + "grad_norm": 2.2538217800740004, + "learning_rate": 1.5201582883886363e-06, + "loss": 0.3268, + "step": 34194 + }, + { + "epoch": 0.7526964560862865, + "grad_norm": 1.7408741857965553, + "learning_rate": 1.5199023357205662e-06, + "loss": 0.31, + "step": 34195 + }, + { + "epoch": 0.7527184679727053, + "grad_norm": 2.0582189171108074, + "learning_rate": 1.5196464007399032e-06, + "loss": 0.2742, + "step": 34196 + }, + { + "epoch": 0.752740479859124, + "grad_norm": 2.0649618639956615, + "learning_rate": 1.5193904834479495e-06, + "loss": 0.2976, + "step": 34197 + }, + { + "epoch": 0.7527624917455426, + "grad_norm": 0.6112018850608218, + "learning_rate": 1.5191345838460053e-06, + "loss": 0.4585, + "step": 34198 + }, + { + "epoch": 0.7527845036319613, + "grad_norm": 1.500200150358852, + "learning_rate": 1.518878701935374e-06, + "loss": 0.3486, + "step": 34199 + }, + { + "epoch": 0.75280651551838, + "grad_norm": 7.666117506787114, + "learning_rate": 1.5186228377173534e-06, + "loss": 0.2535, + "step": 34200 + }, + { + "epoch": 0.7528285274047986, + "grad_norm": 1.4585618430395568, + "learning_rate": 1.5183669911932425e-06, + "loss": 0.2786, + "step": 34201 + }, + { + "epoch": 0.7528505392912173, + "grad_norm": 1.3623474446797414, + "learning_rate": 1.5181111623643436e-06, + "loss": 0.2962, + "step": 34202 + }, + { + "epoch": 0.752872551177636, + "grad_norm": 1.22841339921729, + "learning_rate": 1.5178553512319566e-06, + "loss": 0.2756, + "step": 34203 + }, + { + "epoch": 0.7528945630640546, + "grad_norm": 2.0955292116772473, + "learning_rate": 1.5175995577973833e-06, + "loss": 0.2548, + "step": 34204 + }, + { + "epoch": 0.7529165749504733, + "grad_norm": 1.7342515243155945, + "learning_rate": 1.517343782061923e-06, + "loss": 0.3423, + "step": 34205 + }, + { + "epoch": 0.752938586836892, + "grad_norm": 94.9729107026176, + "learning_rate": 1.5170880240268726e-06, + "loss": 0.2284, + "step": 34206 + }, + { + "epoch": 0.7529605987233106, + "grad_norm": 5.530590575864072, + "learning_rate": 1.516832283693536e-06, + "loss": 0.2917, + "step": 34207 + }, + { + "epoch": 0.7529826106097293, + "grad_norm": 1.7581526274476547, + "learning_rate": 1.5165765610632094e-06, + "loss": 0.3363, + "step": 34208 + }, + { + "epoch": 0.753004622496148, + "grad_norm": 1.8590589724403743, + "learning_rate": 1.516320856137195e-06, + "loss": 0.2407, + "step": 34209 + }, + { + "epoch": 0.7530266343825666, + "grad_norm": 1.9071457696085898, + "learning_rate": 1.5160651689167931e-06, + "loss": 0.321, + "step": 34210 + }, + { + "epoch": 0.7530486462689853, + "grad_norm": 1.6286013513214508, + "learning_rate": 1.5158094994033001e-06, + "loss": 0.2979, + "step": 34211 + }, + { + "epoch": 0.753070658155404, + "grad_norm": 1.490110596952626, + "learning_rate": 1.5155538475980192e-06, + "loss": 0.2246, + "step": 34212 + }, + { + "epoch": 0.7530926700418226, + "grad_norm": 1.768435990854734, + "learning_rate": 1.515298213502246e-06, + "loss": 0.2209, + "step": 34213 + }, + { + "epoch": 0.7531146819282413, + "grad_norm": 1.7677619111085403, + "learning_rate": 1.5150425971172812e-06, + "loss": 0.3241, + "step": 34214 + }, + { + "epoch": 0.7531366938146599, + "grad_norm": 3.0832325045884934, + "learning_rate": 1.5147869984444258e-06, + "loss": 0.2548, + "step": 34215 + }, + { + "epoch": 0.7531587057010786, + "grad_norm": 1.6174447075619687, + "learning_rate": 1.5145314174849761e-06, + "loss": 0.3048, + "step": 34216 + }, + { + "epoch": 0.7531807175874973, + "grad_norm": 0.6198823583370778, + "learning_rate": 1.5142758542402335e-06, + "loss": 0.4497, + "step": 34217 + }, + { + "epoch": 0.7532027294739159, + "grad_norm": 1.4467092807274973, + "learning_rate": 1.5140203087114946e-06, + "loss": 0.351, + "step": 34218 + }, + { + "epoch": 0.7532247413603346, + "grad_norm": 1.5060761400945555, + "learning_rate": 1.5137647809000588e-06, + "loss": 0.2422, + "step": 34219 + }, + { + "epoch": 0.7532467532467533, + "grad_norm": 3.0231127829636537, + "learning_rate": 1.5135092708072274e-06, + "loss": 0.329, + "step": 34220 + }, + { + "epoch": 0.7532687651331719, + "grad_norm": 1.5407490051187933, + "learning_rate": 1.513253778434295e-06, + "loss": 0.2709, + "step": 34221 + }, + { + "epoch": 0.7532907770195906, + "grad_norm": 1.5300987938380919, + "learning_rate": 1.5129983037825634e-06, + "loss": 0.2383, + "step": 34222 + }, + { + "epoch": 0.7533127889060093, + "grad_norm": 1.3674830865820675, + "learning_rate": 1.5127428468533284e-06, + "loss": 0.2987, + "step": 34223 + }, + { + "epoch": 0.7533348007924279, + "grad_norm": 0.6114044885774632, + "learning_rate": 1.5124874076478897e-06, + "loss": 0.4537, + "step": 34224 + }, + { + "epoch": 0.7533568126788466, + "grad_norm": 2.6091547670882296, + "learning_rate": 1.5122319861675472e-06, + "loss": 0.273, + "step": 34225 + }, + { + "epoch": 0.7533788245652653, + "grad_norm": 1.8453524473005518, + "learning_rate": 1.5119765824135967e-06, + "loss": 0.307, + "step": 34226 + }, + { + "epoch": 0.7534008364516839, + "grad_norm": 1.6454235260597483, + "learning_rate": 1.5117211963873358e-06, + "loss": 0.3175, + "step": 34227 + }, + { + "epoch": 0.7534228483381026, + "grad_norm": 3.1819758158232543, + "learning_rate": 1.5114658280900635e-06, + "loss": 0.3117, + "step": 34228 + }, + { + "epoch": 0.7534448602245213, + "grad_norm": 2.822279212797301, + "learning_rate": 1.511210477523078e-06, + "loss": 0.3369, + "step": 34229 + }, + { + "epoch": 0.7534668721109399, + "grad_norm": 0.6173742753255854, + "learning_rate": 1.510955144687678e-06, + "loss": 0.4472, + "step": 34230 + }, + { + "epoch": 0.7534888839973586, + "grad_norm": 2.1999410275218416, + "learning_rate": 1.5106998295851605e-06, + "loss": 0.3142, + "step": 34231 + }, + { + "epoch": 0.7535108958837773, + "grad_norm": 1.9923266803780209, + "learning_rate": 1.5104445322168209e-06, + "loss": 0.2981, + "step": 34232 + }, + { + "epoch": 0.7535329077701959, + "grad_norm": 2.2551067387718144, + "learning_rate": 1.5101892525839584e-06, + "loss": 0.3129, + "step": 34233 + }, + { + "epoch": 0.7535549196566146, + "grad_norm": 2.0184115053641856, + "learning_rate": 1.5099339906878712e-06, + "loss": 0.2794, + "step": 34234 + }, + { + "epoch": 0.7535769315430333, + "grad_norm": 1.831963887489386, + "learning_rate": 1.5096787465298573e-06, + "loss": 0.2592, + "step": 34235 + }, + { + "epoch": 0.7535989434294519, + "grad_norm": 4.7640936885975815, + "learning_rate": 1.5094235201112128e-06, + "loss": 0.2238, + "step": 34236 + }, + { + "epoch": 0.7536209553158706, + "grad_norm": 1.957405762069155, + "learning_rate": 1.5091683114332334e-06, + "loss": 0.2702, + "step": 34237 + }, + { + "epoch": 0.7536429672022892, + "grad_norm": 0.6209736072096973, + "learning_rate": 1.5089131204972175e-06, + "loss": 0.4511, + "step": 34238 + }, + { + "epoch": 0.7536649790887079, + "grad_norm": 2.495145013724212, + "learning_rate": 1.5086579473044616e-06, + "loss": 0.3684, + "step": 34239 + }, + { + "epoch": 0.7536869909751266, + "grad_norm": 1.4150874126516588, + "learning_rate": 1.5084027918562655e-06, + "loss": 0.3934, + "step": 34240 + }, + { + "epoch": 0.7537090028615452, + "grad_norm": 0.6304006160520909, + "learning_rate": 1.5081476541539235e-06, + "loss": 0.4672, + "step": 34241 + }, + { + "epoch": 0.7537310147479639, + "grad_norm": 2.9243421818266704, + "learning_rate": 1.507892534198731e-06, + "loss": 0.2656, + "step": 34242 + }, + { + "epoch": 0.7537530266343826, + "grad_norm": 1.7320301820395525, + "learning_rate": 1.5076374319919873e-06, + "loss": 0.3298, + "step": 34243 + }, + { + "epoch": 0.7537750385208012, + "grad_norm": 9.06497462435675, + "learning_rate": 1.5073823475349868e-06, + "loss": 0.2266, + "step": 34244 + }, + { + "epoch": 0.7537970504072199, + "grad_norm": 1.657380121542747, + "learning_rate": 1.5071272808290265e-06, + "loss": 0.1851, + "step": 34245 + }, + { + "epoch": 0.7538190622936386, + "grad_norm": 1.543631685717358, + "learning_rate": 1.5068722318754053e-06, + "loss": 0.2306, + "step": 34246 + }, + { + "epoch": 0.7538410741800572, + "grad_norm": 1.405436070635454, + "learning_rate": 1.5066172006754154e-06, + "loss": 0.2308, + "step": 34247 + }, + { + "epoch": 0.7538630860664759, + "grad_norm": 1.8758254720186205, + "learning_rate": 1.5063621872303568e-06, + "loss": 0.3099, + "step": 34248 + }, + { + "epoch": 0.7538850979528946, + "grad_norm": 1.3452457294302358, + "learning_rate": 1.5061071915415221e-06, + "loss": 0.3077, + "step": 34249 + }, + { + "epoch": 0.7539071098393132, + "grad_norm": 1.228648428215686, + "learning_rate": 1.5058522136102093e-06, + "loss": 0.1854, + "step": 34250 + }, + { + "epoch": 0.7539291217257319, + "grad_norm": 1.5049491338510423, + "learning_rate": 1.505597253437715e-06, + "loss": 0.285, + "step": 34251 + }, + { + "epoch": 0.7539511336121506, + "grad_norm": 2.5396759483278166, + "learning_rate": 1.505342311025333e-06, + "loss": 0.2901, + "step": 34252 + }, + { + "epoch": 0.7539731454985692, + "grad_norm": 1.324374136804738, + "learning_rate": 1.505087386374361e-06, + "loss": 0.2009, + "step": 34253 + }, + { + "epoch": 0.7539951573849879, + "grad_norm": 1.8969937143936282, + "learning_rate": 1.5048324794860925e-06, + "loss": 0.2881, + "step": 34254 + }, + { + "epoch": 0.7540171692714066, + "grad_norm": 1.5714139447528166, + "learning_rate": 1.5045775903618243e-06, + "loss": 0.2077, + "step": 34255 + }, + { + "epoch": 0.7540391811578252, + "grad_norm": 1.2021870010723092, + "learning_rate": 1.504322719002853e-06, + "loss": 0.2776, + "step": 34256 + }, + { + "epoch": 0.7540611930442439, + "grad_norm": 2.9722531628966413, + "learning_rate": 1.5040678654104712e-06, + "loss": 0.2881, + "step": 34257 + }, + { + "epoch": 0.7540832049306626, + "grad_norm": 1.8703728217306061, + "learning_rate": 1.5038130295859777e-06, + "loss": 0.2514, + "step": 34258 + }, + { + "epoch": 0.7541052168170812, + "grad_norm": 0.6231685148093374, + "learning_rate": 1.503558211530664e-06, + "loss": 0.4366, + "step": 34259 + }, + { + "epoch": 0.7541272287034999, + "grad_norm": 1.4422080397836428, + "learning_rate": 1.5033034112458266e-06, + "loss": 0.2301, + "step": 34260 + }, + { + "epoch": 0.7541492405899185, + "grad_norm": 11.836762751367301, + "learning_rate": 1.5030486287327628e-06, + "loss": 0.2025, + "step": 34261 + }, + { + "epoch": 0.7541712524763372, + "grad_norm": 1.960999167803404, + "learning_rate": 1.5027938639927642e-06, + "loss": 0.282, + "step": 34262 + }, + { + "epoch": 0.7541932643627559, + "grad_norm": 1.6154046437636866, + "learning_rate": 1.502539117027128e-06, + "loss": 0.3012, + "step": 34263 + }, + { + "epoch": 0.7542152762491745, + "grad_norm": 1.6565417678810297, + "learning_rate": 1.5022843878371468e-06, + "loss": 0.2565, + "step": 34264 + }, + { + "epoch": 0.7542372881355932, + "grad_norm": 1.9215302209930862, + "learning_rate": 1.502029676424116e-06, + "loss": 0.2201, + "step": 34265 + }, + { + "epoch": 0.7542593000220119, + "grad_norm": 2.845618911985619, + "learning_rate": 1.5017749827893318e-06, + "loss": 0.3315, + "step": 34266 + }, + { + "epoch": 0.7542813119084305, + "grad_norm": 1.5141796080742964, + "learning_rate": 1.5015203069340884e-06, + "loss": 0.2888, + "step": 34267 + }, + { + "epoch": 0.7543033237948492, + "grad_norm": 1.1654166731572075, + "learning_rate": 1.5012656488596766e-06, + "loss": 0.1882, + "step": 34268 + }, + { + "epoch": 0.7543253356812679, + "grad_norm": 1.3527641149593526, + "learning_rate": 1.501011008567394e-06, + "loss": 0.2711, + "step": 34269 + }, + { + "epoch": 0.7543473475676865, + "grad_norm": 1.5585543385720488, + "learning_rate": 1.5007563860585334e-06, + "loss": 0.3309, + "step": 34270 + }, + { + "epoch": 0.7543693594541052, + "grad_norm": 3.586000848620486, + "learning_rate": 1.5005017813343913e-06, + "loss": 0.2751, + "step": 34271 + }, + { + "epoch": 0.7543913713405239, + "grad_norm": 1.437867853437467, + "learning_rate": 1.5002471943962604e-06, + "loss": 0.2995, + "step": 34272 + }, + { + "epoch": 0.7544133832269425, + "grad_norm": 1.7356220395122697, + "learning_rate": 1.499992625245432e-06, + "loss": 0.3745, + "step": 34273 + }, + { + "epoch": 0.7544353951133612, + "grad_norm": 0.6179176507968871, + "learning_rate": 1.499738073883203e-06, + "loss": 0.4546, + "step": 34274 + }, + { + "epoch": 0.7544574069997799, + "grad_norm": 0.5951224544351191, + "learning_rate": 1.4994835403108653e-06, + "loss": 0.4291, + "step": 34275 + }, + { + "epoch": 0.7544794188861985, + "grad_norm": 1.7233324072340235, + "learning_rate": 1.4992290245297154e-06, + "loss": 0.2332, + "step": 34276 + }, + { + "epoch": 0.7545014307726172, + "grad_norm": 28.45014097841581, + "learning_rate": 1.4989745265410456e-06, + "loss": 0.3506, + "step": 34277 + }, + { + "epoch": 0.7545234426590359, + "grad_norm": 1.399827761943064, + "learning_rate": 1.4987200463461466e-06, + "loss": 0.2885, + "step": 34278 + }, + { + "epoch": 0.7545454545454545, + "grad_norm": 2.14709605307621, + "learning_rate": 1.4984655839463153e-06, + "loss": 0.3176, + "step": 34279 + }, + { + "epoch": 0.7545674664318732, + "grad_norm": 1.7354517501492226, + "learning_rate": 1.4982111393428406e-06, + "loss": 0.2694, + "step": 34280 + }, + { + "epoch": 0.7545894783182919, + "grad_norm": 2.770687134738617, + "learning_rate": 1.497956712537022e-06, + "loss": 0.2326, + "step": 34281 + }, + { + "epoch": 0.7546114902047105, + "grad_norm": 1.8395093135196994, + "learning_rate": 1.4977023035301497e-06, + "loss": 0.2759, + "step": 34282 + }, + { + "epoch": 0.7546335020911292, + "grad_norm": 0.6105730579662845, + "learning_rate": 1.497447912323514e-06, + "loss": 0.4485, + "step": 34283 + }, + { + "epoch": 0.7546555139775478, + "grad_norm": 1.8937089516587993, + "learning_rate": 1.4971935389184127e-06, + "loss": 0.2823, + "step": 34284 + }, + { + "epoch": 0.7546775258639665, + "grad_norm": 0.5986771963418748, + "learning_rate": 1.496939183316134e-06, + "loss": 0.4332, + "step": 34285 + }, + { + "epoch": 0.7546995377503852, + "grad_norm": 1.6827089729162965, + "learning_rate": 1.4966848455179727e-06, + "loss": 0.3002, + "step": 34286 + }, + { + "epoch": 0.7547215496368038, + "grad_norm": 1.6360955280610645, + "learning_rate": 1.4964305255252238e-06, + "loss": 0.3253, + "step": 34287 + }, + { + "epoch": 0.7547435615232225, + "grad_norm": 0.6044565170420861, + "learning_rate": 1.4961762233391753e-06, + "loss": 0.443, + "step": 34288 + }, + { + "epoch": 0.7547655734096412, + "grad_norm": 1.7724250005722915, + "learning_rate": 1.4959219389611235e-06, + "loss": 0.3566, + "step": 34289 + }, + { + "epoch": 0.7547875852960598, + "grad_norm": 1.331049649603534, + "learning_rate": 1.495667672392358e-06, + "loss": 0.3027, + "step": 34290 + }, + { + "epoch": 0.7548095971824785, + "grad_norm": 1.7400266825107422, + "learning_rate": 1.4954134236341723e-06, + "loss": 0.3678, + "step": 34291 + }, + { + "epoch": 0.7548316090688972, + "grad_norm": 1.1492363869995428, + "learning_rate": 1.4951591926878605e-06, + "loss": 0.165, + "step": 34292 + }, + { + "epoch": 0.7548536209553158, + "grad_norm": 1.4612698693395896, + "learning_rate": 1.4949049795547111e-06, + "loss": 0.1769, + "step": 34293 + }, + { + "epoch": 0.7548756328417345, + "grad_norm": 1.545235262936186, + "learning_rate": 1.4946507842360192e-06, + "loss": 0.3392, + "step": 34294 + }, + { + "epoch": 0.7548976447281532, + "grad_norm": 1.3073588348259821, + "learning_rate": 1.4943966067330744e-06, + "loss": 0.2919, + "step": 34295 + }, + { + "epoch": 0.7549196566145718, + "grad_norm": 1.123417434888888, + "learning_rate": 1.4941424470471693e-06, + "loss": 0.2337, + "step": 34296 + }, + { + "epoch": 0.7549416685009905, + "grad_norm": 1.7627934258891589, + "learning_rate": 1.4938883051795977e-06, + "loss": 0.2816, + "step": 34297 + }, + { + "epoch": 0.7549636803874092, + "grad_norm": 0.6252492281234792, + "learning_rate": 1.4936341811316473e-06, + "loss": 0.4639, + "step": 34298 + }, + { + "epoch": 0.7549856922738278, + "grad_norm": 1.5765977320150557, + "learning_rate": 1.493380074904614e-06, + "loss": 0.2861, + "step": 34299 + }, + { + "epoch": 0.7550077041602465, + "grad_norm": 1.352650868672716, + "learning_rate": 1.4931259864997854e-06, + "loss": 0.3553, + "step": 34300 + }, + { + "epoch": 0.7550297160466652, + "grad_norm": 1.619514512564285, + "learning_rate": 1.4928719159184552e-06, + "loss": 0.3425, + "step": 34301 + }, + { + "epoch": 0.7550517279330838, + "grad_norm": 1.3607858901607017, + "learning_rate": 1.4926178631619154e-06, + "loss": 0.2171, + "step": 34302 + }, + { + "epoch": 0.7550737398195025, + "grad_norm": 1.4141759061009156, + "learning_rate": 1.492363828231454e-06, + "loss": 0.2933, + "step": 34303 + }, + { + "epoch": 0.7550957517059212, + "grad_norm": 1.397252949187344, + "learning_rate": 1.4921098111283661e-06, + "loss": 0.2569, + "step": 34304 + }, + { + "epoch": 0.7551177635923398, + "grad_norm": 0.613772966068669, + "learning_rate": 1.491855811853939e-06, + "loss": 0.4656, + "step": 34305 + }, + { + "epoch": 0.7551397754787585, + "grad_norm": 0.6291293577442465, + "learning_rate": 1.4916018304094647e-06, + "loss": 0.492, + "step": 34306 + }, + { + "epoch": 0.7551617873651773, + "grad_norm": 1.6436951006989853, + "learning_rate": 1.4913478667962372e-06, + "loss": 0.3231, + "step": 34307 + }, + { + "epoch": 0.7551837992515958, + "grad_norm": 1.6180095626773747, + "learning_rate": 1.4910939210155444e-06, + "loss": 0.3245, + "step": 34308 + }, + { + "epoch": 0.7552058111380145, + "grad_norm": 1.3291184912865637, + "learning_rate": 1.4908399930686756e-06, + "loss": 0.3301, + "step": 34309 + }, + { + "epoch": 0.7552278230244331, + "grad_norm": 2.999538669273415, + "learning_rate": 1.490586082956923e-06, + "loss": 0.2446, + "step": 34310 + }, + { + "epoch": 0.7552498349108518, + "grad_norm": 1.1487302055098527, + "learning_rate": 1.4903321906815776e-06, + "loss": 0.2473, + "step": 34311 + }, + { + "epoch": 0.7552718467972706, + "grad_norm": 0.6436857576327338, + "learning_rate": 1.490078316243931e-06, + "loss": 0.4635, + "step": 34312 + }, + { + "epoch": 0.7552938586836891, + "grad_norm": 12.685773507892817, + "learning_rate": 1.489824459645271e-06, + "loss": 0.294, + "step": 34313 + }, + { + "epoch": 0.7553158705701079, + "grad_norm": 1.352075485250207, + "learning_rate": 1.4895706208868876e-06, + "loss": 0.3265, + "step": 34314 + }, + { + "epoch": 0.7553378824565266, + "grad_norm": 1.780871184660175, + "learning_rate": 1.4893167999700735e-06, + "loss": 0.2351, + "step": 34315 + }, + { + "epoch": 0.7553598943429451, + "grad_norm": 1.613671713442184, + "learning_rate": 1.4890629968961146e-06, + "loss": 0.2352, + "step": 34316 + }, + { + "epoch": 0.7553819062293639, + "grad_norm": 1.3271794336227862, + "learning_rate": 1.488809211666306e-06, + "loss": 0.239, + "step": 34317 + }, + { + "epoch": 0.7554039181157826, + "grad_norm": 1.6058640486509315, + "learning_rate": 1.4885554442819355e-06, + "loss": 0.2617, + "step": 34318 + }, + { + "epoch": 0.7554259300022012, + "grad_norm": 1.5279282559421117, + "learning_rate": 1.4883016947442908e-06, + "loss": 0.2774, + "step": 34319 + }, + { + "epoch": 0.7554479418886199, + "grad_norm": 1.6940682940569596, + "learning_rate": 1.4880479630546641e-06, + "loss": 0.2446, + "step": 34320 + }, + { + "epoch": 0.7554699537750386, + "grad_norm": 2.0777394712525106, + "learning_rate": 1.4877942492143431e-06, + "loss": 0.3044, + "step": 34321 + }, + { + "epoch": 0.7554919656614572, + "grad_norm": 1.4786639114287183, + "learning_rate": 1.4875405532246179e-06, + "loss": 0.2146, + "step": 34322 + }, + { + "epoch": 0.7555139775478759, + "grad_norm": 1.5191976601435342, + "learning_rate": 1.4872868750867797e-06, + "loss": 0.2261, + "step": 34323 + }, + { + "epoch": 0.7555359894342946, + "grad_norm": 1.5939698082561058, + "learning_rate": 1.487033214802115e-06, + "loss": 0.2786, + "step": 34324 + }, + { + "epoch": 0.7555580013207132, + "grad_norm": 1.7729090299538708, + "learning_rate": 1.4867795723719158e-06, + "loss": 0.3072, + "step": 34325 + }, + { + "epoch": 0.7555800132071319, + "grad_norm": 1.2783184288244898, + "learning_rate": 1.486525947797468e-06, + "loss": 0.2713, + "step": 34326 + }, + { + "epoch": 0.7556020250935506, + "grad_norm": 2.5833389226657038, + "learning_rate": 1.4862723410800628e-06, + "loss": 0.3066, + "step": 34327 + }, + { + "epoch": 0.7556240369799692, + "grad_norm": 1.6516071190787387, + "learning_rate": 1.4860187522209901e-06, + "loss": 0.3117, + "step": 34328 + }, + { + "epoch": 0.7556460488663879, + "grad_norm": 0.6002137623989403, + "learning_rate": 1.4857651812215356e-06, + "loss": 0.4354, + "step": 34329 + }, + { + "epoch": 0.7556680607528066, + "grad_norm": 1.8812949275683426, + "learning_rate": 1.4855116280829917e-06, + "loss": 0.3199, + "step": 34330 + }, + { + "epoch": 0.7556900726392252, + "grad_norm": 8.824525435142581, + "learning_rate": 1.4852580928066435e-06, + "loss": 0.2617, + "step": 34331 + }, + { + "epoch": 0.7557120845256439, + "grad_norm": 2.1834611724183364, + "learning_rate": 1.4850045753937813e-06, + "loss": 0.264, + "step": 34332 + }, + { + "epoch": 0.7557340964120625, + "grad_norm": 2.075431848956201, + "learning_rate": 1.484751075845695e-06, + "loss": 0.3003, + "step": 34333 + }, + { + "epoch": 0.7557561082984812, + "grad_norm": 2.0098024680935604, + "learning_rate": 1.4844975941636702e-06, + "loss": 0.2988, + "step": 34334 + }, + { + "epoch": 0.7557781201848999, + "grad_norm": 1.470840511829242, + "learning_rate": 1.484244130348998e-06, + "loss": 0.2636, + "step": 34335 + }, + { + "epoch": 0.7558001320713185, + "grad_norm": 0.616832673092993, + "learning_rate": 1.4839906844029638e-06, + "loss": 0.4536, + "step": 34336 + }, + { + "epoch": 0.7558221439577372, + "grad_norm": 1.3053851702706578, + "learning_rate": 1.483737256326857e-06, + "loss": 0.2602, + "step": 34337 + }, + { + "epoch": 0.7558441558441559, + "grad_norm": 1.5129396916066853, + "learning_rate": 1.4834838461219676e-06, + "loss": 0.2437, + "step": 34338 + }, + { + "epoch": 0.7558661677305745, + "grad_norm": 2.0031569112597034, + "learning_rate": 1.4832304537895798e-06, + "loss": 0.3866, + "step": 34339 + }, + { + "epoch": 0.7558881796169932, + "grad_norm": 1.4690093347070878, + "learning_rate": 1.482977079330985e-06, + "loss": 0.2682, + "step": 34340 + }, + { + "epoch": 0.7559101915034119, + "grad_norm": 1.7009473530888957, + "learning_rate": 1.4827237227474677e-06, + "loss": 0.3184, + "step": 34341 + }, + { + "epoch": 0.7559322033898305, + "grad_norm": 1.5896455727063985, + "learning_rate": 1.4824703840403176e-06, + "loss": 0.2692, + "step": 34342 + }, + { + "epoch": 0.7559542152762492, + "grad_norm": 1.6282820331586128, + "learning_rate": 1.4822170632108228e-06, + "loss": 0.3676, + "step": 34343 + }, + { + "epoch": 0.7559762271626679, + "grad_norm": 2.144116707570824, + "learning_rate": 1.4819637602602704e-06, + "loss": 0.3127, + "step": 34344 + }, + { + "epoch": 0.7559982390490865, + "grad_norm": 1.8170742743670403, + "learning_rate": 1.4817104751899458e-06, + "loss": 0.2861, + "step": 34345 + }, + { + "epoch": 0.7560202509355052, + "grad_norm": 3.0181551695375064, + "learning_rate": 1.4814572080011375e-06, + "loss": 0.2093, + "step": 34346 + }, + { + "epoch": 0.7560422628219239, + "grad_norm": 2.2568541909043325, + "learning_rate": 1.4812039586951327e-06, + "loss": 0.333, + "step": 34347 + }, + { + "epoch": 0.7560642747083425, + "grad_norm": 1.5495570423786096, + "learning_rate": 1.4809507272732204e-06, + "loss": 0.2864, + "step": 34348 + }, + { + "epoch": 0.7560862865947612, + "grad_norm": 1.6197709729444512, + "learning_rate": 1.4806975137366864e-06, + "loss": 0.3296, + "step": 34349 + }, + { + "epoch": 0.7561082984811799, + "grad_norm": 1.6336417133518832, + "learning_rate": 1.4804443180868155e-06, + "loss": 0.3018, + "step": 34350 + }, + { + "epoch": 0.7561303103675985, + "grad_norm": 1.4971931886731615, + "learning_rate": 1.480191140324898e-06, + "loss": 0.3338, + "step": 34351 + }, + { + "epoch": 0.7561523222540172, + "grad_norm": 1.8784843152539217, + "learning_rate": 1.4799379804522162e-06, + "loss": 0.3063, + "step": 34352 + }, + { + "epoch": 0.7561743341404359, + "grad_norm": 2.2256554669756623, + "learning_rate": 1.4796848384700625e-06, + "loss": 0.2813, + "step": 34353 + }, + { + "epoch": 0.7561963460268545, + "grad_norm": 1.3856341919262356, + "learning_rate": 1.4794317143797198e-06, + "loss": 0.2572, + "step": 34354 + }, + { + "epoch": 0.7562183579132732, + "grad_norm": 0.6125394307722604, + "learning_rate": 1.479178608182475e-06, + "loss": 0.441, + "step": 34355 + }, + { + "epoch": 0.7562403697996918, + "grad_norm": 1.636078684199495, + "learning_rate": 1.4789255198796154e-06, + "loss": 0.4078, + "step": 34356 + }, + { + "epoch": 0.7562623816861105, + "grad_norm": 1.8139064569983816, + "learning_rate": 1.4786724494724259e-06, + "loss": 0.2613, + "step": 34357 + }, + { + "epoch": 0.7562843935725292, + "grad_norm": 1.9009782875816674, + "learning_rate": 1.478419396962193e-06, + "loss": 0.2936, + "step": 34358 + }, + { + "epoch": 0.7563064054589478, + "grad_norm": 1.6642328741983319, + "learning_rate": 1.4781663623502052e-06, + "loss": 0.3401, + "step": 34359 + }, + { + "epoch": 0.7563284173453665, + "grad_norm": 2.1000423215318373, + "learning_rate": 1.4779133456377454e-06, + "loss": 0.2233, + "step": 34360 + }, + { + "epoch": 0.7563504292317852, + "grad_norm": 1.3523068208618927, + "learning_rate": 1.477660346826102e-06, + "loss": 0.2491, + "step": 34361 + }, + { + "epoch": 0.7563724411182038, + "grad_norm": 1.5683435779033068, + "learning_rate": 1.4774073659165582e-06, + "loss": 0.3313, + "step": 34362 + }, + { + "epoch": 0.7563944530046225, + "grad_norm": 1.6737475975256868, + "learning_rate": 1.4771544029104018e-06, + "loss": 0.3272, + "step": 34363 + }, + { + "epoch": 0.7564164648910412, + "grad_norm": 2.3405872739458733, + "learning_rate": 1.476901457808919e-06, + "loss": 0.2766, + "step": 34364 + }, + { + "epoch": 0.7564384767774598, + "grad_norm": 0.6107808324769061, + "learning_rate": 1.4766485306133931e-06, + "loss": 0.4348, + "step": 34365 + }, + { + "epoch": 0.7564604886638785, + "grad_norm": 2.2797630300514315, + "learning_rate": 1.4763956213251125e-06, + "loss": 0.2683, + "step": 34366 + }, + { + "epoch": 0.7564825005502972, + "grad_norm": 1.2814543254648856, + "learning_rate": 1.4761427299453591e-06, + "loss": 0.2725, + "step": 34367 + }, + { + "epoch": 0.7565045124367158, + "grad_norm": 2.016182052770154, + "learning_rate": 1.4758898564754199e-06, + "loss": 0.2396, + "step": 34368 + }, + { + "epoch": 0.7565265243231345, + "grad_norm": 0.6277718595137893, + "learning_rate": 1.475637000916582e-06, + "loss": 0.4391, + "step": 34369 + }, + { + "epoch": 0.7565485362095532, + "grad_norm": 1.7141674959529598, + "learning_rate": 1.4753841632701272e-06, + "loss": 0.3285, + "step": 34370 + }, + { + "epoch": 0.7565705480959718, + "grad_norm": 1.4610208524282882, + "learning_rate": 1.4751313435373437e-06, + "loss": 0.2915, + "step": 34371 + }, + { + "epoch": 0.7565925599823905, + "grad_norm": 0.6005924852933309, + "learning_rate": 1.4748785417195132e-06, + "loss": 0.4456, + "step": 34372 + }, + { + "epoch": 0.7566145718688092, + "grad_norm": 4.031470724766375, + "learning_rate": 1.4746257578179219e-06, + "loss": 0.2986, + "step": 34373 + }, + { + "epoch": 0.7566365837552278, + "grad_norm": 1.4476936166882364, + "learning_rate": 1.4743729918338567e-06, + "loss": 0.2745, + "step": 34374 + }, + { + "epoch": 0.7566585956416465, + "grad_norm": 2.2862041518913028, + "learning_rate": 1.4741202437685992e-06, + "loss": 0.2626, + "step": 34375 + }, + { + "epoch": 0.7566806075280652, + "grad_norm": 1.5463413219638922, + "learning_rate": 1.4738675136234366e-06, + "loss": 0.2745, + "step": 34376 + }, + { + "epoch": 0.7567026194144838, + "grad_norm": 1.359917294409477, + "learning_rate": 1.4736148013996498e-06, + "loss": 0.2715, + "step": 34377 + }, + { + "epoch": 0.7567246313009025, + "grad_norm": 2.168930990237453, + "learning_rate": 1.473362107098526e-06, + "loss": 0.2569, + "step": 34378 + }, + { + "epoch": 0.7567466431873211, + "grad_norm": 1.7135541318557752, + "learning_rate": 1.4731094307213507e-06, + "loss": 0.274, + "step": 34379 + }, + { + "epoch": 0.7567686550737398, + "grad_norm": 1.6746658899519613, + "learning_rate": 1.4728567722694043e-06, + "loss": 0.2727, + "step": 34380 + }, + { + "epoch": 0.7567906669601585, + "grad_norm": 1.6432440892407172, + "learning_rate": 1.4726041317439744e-06, + "loss": 0.2392, + "step": 34381 + }, + { + "epoch": 0.7568126788465771, + "grad_norm": 1.6156816054976544, + "learning_rate": 1.4723515091463426e-06, + "loss": 0.3635, + "step": 34382 + }, + { + "epoch": 0.7568346907329958, + "grad_norm": 2.9657913147569466, + "learning_rate": 1.4720989044777934e-06, + "loss": 0.2949, + "step": 34383 + }, + { + "epoch": 0.7568567026194145, + "grad_norm": 1.351242504599635, + "learning_rate": 1.4718463177396125e-06, + "loss": 0.2211, + "step": 34384 + }, + { + "epoch": 0.7568787145058331, + "grad_norm": 1.9391912262462345, + "learning_rate": 1.4715937489330822e-06, + "loss": 0.2491, + "step": 34385 + }, + { + "epoch": 0.7569007263922518, + "grad_norm": 0.6305924259019257, + "learning_rate": 1.471341198059485e-06, + "loss": 0.459, + "step": 34386 + }, + { + "epoch": 0.7569227382786705, + "grad_norm": 1.6527475692165383, + "learning_rate": 1.4710886651201068e-06, + "loss": 0.3109, + "step": 34387 + }, + { + "epoch": 0.7569447501650891, + "grad_norm": 2.421338506316534, + "learning_rate": 1.4708361501162271e-06, + "loss": 0.3385, + "step": 34388 + }, + { + "epoch": 0.7569667620515078, + "grad_norm": 0.5966276769091208, + "learning_rate": 1.4705836530491352e-06, + "loss": 0.4515, + "step": 34389 + }, + { + "epoch": 0.7569887739379265, + "grad_norm": 2.1563083728699732, + "learning_rate": 1.470331173920111e-06, + "loss": 0.3113, + "step": 34390 + }, + { + "epoch": 0.7570107858243451, + "grad_norm": 2.754252819725451, + "learning_rate": 1.4700787127304367e-06, + "loss": 0.3763, + "step": 34391 + }, + { + "epoch": 0.7570327977107638, + "grad_norm": 0.6202156666817501, + "learning_rate": 1.4698262694813987e-06, + "loss": 0.4633, + "step": 34392 + }, + { + "epoch": 0.7570548095971825, + "grad_norm": 1.8104791637963367, + "learning_rate": 1.4695738441742763e-06, + "loss": 0.2765, + "step": 34393 + }, + { + "epoch": 0.7570768214836011, + "grad_norm": 1.8332145923916063, + "learning_rate": 1.4693214368103542e-06, + "loss": 0.3531, + "step": 34394 + }, + { + "epoch": 0.7570988333700198, + "grad_norm": 2.629651387027281, + "learning_rate": 1.4690690473909164e-06, + "loss": 0.264, + "step": 34395 + }, + { + "epoch": 0.7571208452564385, + "grad_norm": 0.6121265016522353, + "learning_rate": 1.4688166759172434e-06, + "loss": 0.4524, + "step": 34396 + }, + { + "epoch": 0.7571428571428571, + "grad_norm": 1.7342636284586967, + "learning_rate": 1.4685643223906209e-06, + "loss": 0.2876, + "step": 34397 + }, + { + "epoch": 0.7571648690292758, + "grad_norm": 1.3993851192149576, + "learning_rate": 1.4683119868123275e-06, + "loss": 0.304, + "step": 34398 + }, + { + "epoch": 0.7571868809156945, + "grad_norm": 0.6114038082538519, + "learning_rate": 1.4680596691836479e-06, + "loss": 0.4383, + "step": 34399 + }, + { + "epoch": 0.7572088928021131, + "grad_norm": 2.197208643620697, + "learning_rate": 1.4678073695058664e-06, + "loss": 0.2949, + "step": 34400 + }, + { + "epoch": 0.7572309046885318, + "grad_norm": 2.0707306102896808, + "learning_rate": 1.4675550877802609e-06, + "loss": 0.2908, + "step": 34401 + }, + { + "epoch": 0.7572529165749504, + "grad_norm": 1.3751547528801418, + "learning_rate": 1.4673028240081184e-06, + "loss": 0.31, + "step": 34402 + }, + { + "epoch": 0.7572749284613691, + "grad_norm": 1.2460471931443076, + "learning_rate": 1.4670505781907162e-06, + "loss": 0.2707, + "step": 34403 + }, + { + "epoch": 0.7572969403477878, + "grad_norm": 2.096471615180883, + "learning_rate": 1.4667983503293387e-06, + "loss": 0.3196, + "step": 34404 + }, + { + "epoch": 0.7573189522342064, + "grad_norm": 1.2171116238938657, + "learning_rate": 1.4665461404252702e-06, + "loss": 0.1685, + "step": 34405 + }, + { + "epoch": 0.7573409641206251, + "grad_norm": 0.6176423249088686, + "learning_rate": 1.4662939484797877e-06, + "loss": 0.4779, + "step": 34406 + }, + { + "epoch": 0.7573629760070438, + "grad_norm": 1.3832479801795516, + "learning_rate": 1.4660417744941774e-06, + "loss": 0.2014, + "step": 34407 + }, + { + "epoch": 0.7573849878934624, + "grad_norm": 2.1271254578048353, + "learning_rate": 1.4657896184697173e-06, + "loss": 0.2349, + "step": 34408 + }, + { + "epoch": 0.7574069997798811, + "grad_norm": 1.3091425765799096, + "learning_rate": 1.4655374804076911e-06, + "loss": 0.2832, + "step": 34409 + }, + { + "epoch": 0.7574290116662998, + "grad_norm": 1.4555110133054154, + "learning_rate": 1.465285360309381e-06, + "loss": 0.2577, + "step": 34410 + }, + { + "epoch": 0.7574510235527184, + "grad_norm": 2.4516457637493803, + "learning_rate": 1.4650332581760656e-06, + "loss": 0.2947, + "step": 34411 + }, + { + "epoch": 0.7574730354391371, + "grad_norm": 2.2813344464592067, + "learning_rate": 1.4647811740090296e-06, + "loss": 0.2092, + "step": 34412 + }, + { + "epoch": 0.7574950473255558, + "grad_norm": 1.7373191026204167, + "learning_rate": 1.4645291078095508e-06, + "loss": 0.2464, + "step": 34413 + }, + { + "epoch": 0.7575170592119744, + "grad_norm": 0.6222952232954503, + "learning_rate": 1.4642770595789119e-06, + "loss": 0.4676, + "step": 34414 + }, + { + "epoch": 0.7575390710983931, + "grad_norm": 1.7243553580150397, + "learning_rate": 1.4640250293183956e-06, + "loss": 0.2956, + "step": 34415 + }, + { + "epoch": 0.7575610829848118, + "grad_norm": 1.6619507880234008, + "learning_rate": 1.4637730170292797e-06, + "loss": 0.2428, + "step": 34416 + }, + { + "epoch": 0.7575830948712304, + "grad_norm": 1.9341735859905116, + "learning_rate": 1.4635210227128482e-06, + "loss": 0.345, + "step": 34417 + }, + { + "epoch": 0.7576051067576491, + "grad_norm": 1.7482746722663733, + "learning_rate": 1.4632690463703797e-06, + "loss": 0.3242, + "step": 34418 + }, + { + "epoch": 0.7576271186440678, + "grad_norm": 4.441544678348888, + "learning_rate": 1.4630170880031524e-06, + "loss": 0.2746, + "step": 34419 + }, + { + "epoch": 0.7576491305304864, + "grad_norm": 1.5188157745056978, + "learning_rate": 1.4627651476124532e-06, + "loss": 0.3173, + "step": 34420 + }, + { + "epoch": 0.7576711424169051, + "grad_norm": 1.9027113117129972, + "learning_rate": 1.4625132251995571e-06, + "loss": 0.3007, + "step": 34421 + }, + { + "epoch": 0.7576931543033238, + "grad_norm": 2.3061530082137645, + "learning_rate": 1.4622613207657487e-06, + "loss": 0.2572, + "step": 34422 + }, + { + "epoch": 0.7577151661897424, + "grad_norm": 1.5003103899750028, + "learning_rate": 1.462009434312306e-06, + "loss": 0.3128, + "step": 34423 + }, + { + "epoch": 0.7577371780761611, + "grad_norm": 3.131421653266354, + "learning_rate": 1.461757565840506e-06, + "loss": 0.2691, + "step": 34424 + }, + { + "epoch": 0.7577591899625797, + "grad_norm": 1.5022926841284459, + "learning_rate": 1.4615057153516353e-06, + "loss": 0.2564, + "step": 34425 + }, + { + "epoch": 0.7577812018489984, + "grad_norm": 1.668430380287531, + "learning_rate": 1.4612538828469708e-06, + "loss": 0.2319, + "step": 34426 + }, + { + "epoch": 0.7578032137354171, + "grad_norm": 2.7114802699619633, + "learning_rate": 1.461002068327791e-06, + "loss": 0.4054, + "step": 34427 + }, + { + "epoch": 0.7578252256218357, + "grad_norm": 2.013117186672715, + "learning_rate": 1.460750271795378e-06, + "loss": 0.2903, + "step": 34428 + }, + { + "epoch": 0.7578472375082544, + "grad_norm": 0.6238746615910479, + "learning_rate": 1.4604984932510097e-06, + "loss": 0.4441, + "step": 34429 + }, + { + "epoch": 0.7578692493946732, + "grad_norm": 1.4740339734835457, + "learning_rate": 1.4602467326959669e-06, + "loss": 0.3088, + "step": 34430 + }, + { + "epoch": 0.7578912612810917, + "grad_norm": 1.2586179000185533, + "learning_rate": 1.4599949901315303e-06, + "loss": 0.2552, + "step": 34431 + }, + { + "epoch": 0.7579132731675104, + "grad_norm": 1.851473071060981, + "learning_rate": 1.4597432655589766e-06, + "loss": 0.3164, + "step": 34432 + }, + { + "epoch": 0.7579352850539292, + "grad_norm": 1.8382528360209587, + "learning_rate": 1.4594915589795877e-06, + "loss": 0.2696, + "step": 34433 + }, + { + "epoch": 0.7579572969403477, + "grad_norm": 2.2296393319158296, + "learning_rate": 1.4592398703946408e-06, + "loss": 0.361, + "step": 34434 + }, + { + "epoch": 0.7579793088267665, + "grad_norm": 2.1853547404599847, + "learning_rate": 1.458988199805416e-06, + "loss": 0.2567, + "step": 34435 + }, + { + "epoch": 0.7580013207131852, + "grad_norm": 0.6056220385217657, + "learning_rate": 1.4587365472131938e-06, + "loss": 0.4502, + "step": 34436 + }, + { + "epoch": 0.7580233325996037, + "grad_norm": 1.696059132556339, + "learning_rate": 1.4584849126192508e-06, + "loss": 0.3472, + "step": 34437 + }, + { + "epoch": 0.7580453444860225, + "grad_norm": 1.4944810412705014, + "learning_rate": 1.458233296024868e-06, + "loss": 0.2685, + "step": 34438 + }, + { + "epoch": 0.7580673563724412, + "grad_norm": 1.534584005817916, + "learning_rate": 1.4579816974313221e-06, + "loss": 0.2691, + "step": 34439 + }, + { + "epoch": 0.7580893682588598, + "grad_norm": 1.8964058598893145, + "learning_rate": 1.4577301168398933e-06, + "loss": 0.3826, + "step": 34440 + }, + { + "epoch": 0.7581113801452785, + "grad_norm": 2.0672098498929, + "learning_rate": 1.457478554251861e-06, + "loss": 0.3304, + "step": 34441 + }, + { + "epoch": 0.7581333920316972, + "grad_norm": 1.599717823493257, + "learning_rate": 1.4572270096685009e-06, + "loss": 0.3228, + "step": 34442 + }, + { + "epoch": 0.7581554039181158, + "grad_norm": 1.7587309322864249, + "learning_rate": 1.4569754830910953e-06, + "loss": 0.2384, + "step": 34443 + }, + { + "epoch": 0.7581774158045345, + "grad_norm": 1.7577699105232738, + "learning_rate": 1.4567239745209188e-06, + "loss": 0.3033, + "step": 34444 + }, + { + "epoch": 0.7581994276909532, + "grad_norm": 1.5921561425914803, + "learning_rate": 1.4564724839592514e-06, + "loss": 0.2726, + "step": 34445 + }, + { + "epoch": 0.7582214395773718, + "grad_norm": 1.5390534888262788, + "learning_rate": 1.4562210114073733e-06, + "loss": 0.303, + "step": 34446 + }, + { + "epoch": 0.7582434514637905, + "grad_norm": 0.6114465279344722, + "learning_rate": 1.4559695568665583e-06, + "loss": 0.4435, + "step": 34447 + }, + { + "epoch": 0.7582654633502092, + "grad_norm": 1.7055430096745683, + "learning_rate": 1.4557181203380887e-06, + "loss": 0.279, + "step": 34448 + }, + { + "epoch": 0.7582874752366278, + "grad_norm": 1.4123621020195065, + "learning_rate": 1.4554667018232388e-06, + "loss": 0.2683, + "step": 34449 + }, + { + "epoch": 0.7583094871230465, + "grad_norm": 1.8742859106688277, + "learning_rate": 1.4552153013232879e-06, + "loss": 0.3449, + "step": 34450 + }, + { + "epoch": 0.7583314990094651, + "grad_norm": 1.4215053725991198, + "learning_rate": 1.4549639188395153e-06, + "loss": 0.3272, + "step": 34451 + }, + { + "epoch": 0.7583535108958838, + "grad_norm": 1.6282783445743427, + "learning_rate": 1.454712554373196e-06, + "loss": 0.35, + "step": 34452 + }, + { + "epoch": 0.7583755227823025, + "grad_norm": 0.5988859293641332, + "learning_rate": 1.4544612079256105e-06, + "loss": 0.4339, + "step": 34453 + }, + { + "epoch": 0.7583975346687211, + "grad_norm": 1.1310497892278233, + "learning_rate": 1.4542098794980341e-06, + "loss": 0.2251, + "step": 34454 + }, + { + "epoch": 0.7584195465551398, + "grad_norm": 2.8674383870384665, + "learning_rate": 1.4539585690917412e-06, + "loss": 0.2968, + "step": 34455 + }, + { + "epoch": 0.7584415584415585, + "grad_norm": 4.0178311222360445, + "learning_rate": 1.4537072767080157e-06, + "loss": 0.3109, + "step": 34456 + }, + { + "epoch": 0.7584635703279771, + "grad_norm": 2.7267731904683727, + "learning_rate": 1.4534560023481303e-06, + "loss": 0.2845, + "step": 34457 + }, + { + "epoch": 0.7584855822143958, + "grad_norm": 2.3423520725776448, + "learning_rate": 1.4532047460133647e-06, + "loss": 0.3274, + "step": 34458 + }, + { + "epoch": 0.7585075941008145, + "grad_norm": 3.0909810477120083, + "learning_rate": 1.4529535077049945e-06, + "loss": 0.3154, + "step": 34459 + }, + { + "epoch": 0.7585296059872331, + "grad_norm": 1.5224986122074669, + "learning_rate": 1.452702287424293e-06, + "loss": 0.2494, + "step": 34460 + }, + { + "epoch": 0.7585516178736518, + "grad_norm": 1.7904607308071436, + "learning_rate": 1.4524510851725442e-06, + "loss": 0.299, + "step": 34461 + }, + { + "epoch": 0.7585736297600705, + "grad_norm": 1.7935974656962608, + "learning_rate": 1.4521999009510197e-06, + "loss": 0.2777, + "step": 34462 + }, + { + "epoch": 0.7585956416464891, + "grad_norm": 1.734344761147521, + "learning_rate": 1.451948734760999e-06, + "loss": 0.351, + "step": 34463 + }, + { + "epoch": 0.7586176535329078, + "grad_norm": 3.848258812333849, + "learning_rate": 1.4516975866037574e-06, + "loss": 0.3213, + "step": 34464 + }, + { + "epoch": 0.7586396654193265, + "grad_norm": 1.7150388191641963, + "learning_rate": 1.4514464564805696e-06, + "loss": 0.3573, + "step": 34465 + }, + { + "epoch": 0.7586616773057451, + "grad_norm": 1.3929016360848705, + "learning_rate": 1.451195344392714e-06, + "loss": 0.276, + "step": 34466 + }, + { + "epoch": 0.7586836891921638, + "grad_norm": 2.0095287438021687, + "learning_rate": 1.4509442503414677e-06, + "loss": 0.2597, + "step": 34467 + }, + { + "epoch": 0.7587057010785825, + "grad_norm": 1.4452293513833678, + "learning_rate": 1.4506931743281045e-06, + "loss": 0.2215, + "step": 34468 + }, + { + "epoch": 0.7587277129650011, + "grad_norm": 0.6115730159740712, + "learning_rate": 1.4504421163539022e-06, + "loss": 0.4389, + "step": 34469 + }, + { + "epoch": 0.7587497248514198, + "grad_norm": 1.8218792777971364, + "learning_rate": 1.4501910764201355e-06, + "loss": 0.2806, + "step": 34470 + }, + { + "epoch": 0.7587717367378385, + "grad_norm": 4.992916252579804, + "learning_rate": 1.4499400545280811e-06, + "loss": 0.302, + "step": 34471 + }, + { + "epoch": 0.7587937486242571, + "grad_norm": 2.411858963443796, + "learning_rate": 1.4496890506790157e-06, + "loss": 0.3446, + "step": 34472 + }, + { + "epoch": 0.7588157605106758, + "grad_norm": 2.1004314685729737, + "learning_rate": 1.4494380648742128e-06, + "loss": 0.3249, + "step": 34473 + }, + { + "epoch": 0.7588377723970944, + "grad_norm": 2.658723697510724, + "learning_rate": 1.4491870971149508e-06, + "loss": 0.3313, + "step": 34474 + }, + { + "epoch": 0.7588597842835131, + "grad_norm": 2.5758356951088666, + "learning_rate": 1.4489361474025022e-06, + "loss": 0.2391, + "step": 34475 + }, + { + "epoch": 0.7588817961699318, + "grad_norm": 2.375798698541921, + "learning_rate": 1.4486852157381437e-06, + "loss": 0.3193, + "step": 34476 + }, + { + "epoch": 0.7589038080563504, + "grad_norm": 1.7020932426432418, + "learning_rate": 1.4484343021231523e-06, + "loss": 0.3605, + "step": 34477 + }, + { + "epoch": 0.7589258199427691, + "grad_norm": 1.2944314459710227, + "learning_rate": 1.4481834065588007e-06, + "loss": 0.3014, + "step": 34478 + }, + { + "epoch": 0.7589478318291878, + "grad_norm": 1.4326559128054417, + "learning_rate": 1.4479325290463665e-06, + "loss": 0.2978, + "step": 34479 + }, + { + "epoch": 0.7589698437156064, + "grad_norm": 1.8295399127695067, + "learning_rate": 1.4476816695871222e-06, + "loss": 0.2699, + "step": 34480 + }, + { + "epoch": 0.7589918556020251, + "grad_norm": 1.4678682079162244, + "learning_rate": 1.4474308281823435e-06, + "loss": 0.2753, + "step": 34481 + }, + { + "epoch": 0.7590138674884438, + "grad_norm": 0.6088571698907467, + "learning_rate": 1.4471800048333074e-06, + "loss": 0.4461, + "step": 34482 + }, + { + "epoch": 0.7590358793748624, + "grad_norm": 0.635324259950545, + "learning_rate": 1.446929199541286e-06, + "loss": 0.4711, + "step": 34483 + }, + { + "epoch": 0.7590578912612811, + "grad_norm": 3.89533445031784, + "learning_rate": 1.446678412307556e-06, + "loss": 0.2988, + "step": 34484 + }, + { + "epoch": 0.7590799031476998, + "grad_norm": 1.6073366909159483, + "learning_rate": 1.4464276431333902e-06, + "loss": 0.3399, + "step": 34485 + }, + { + "epoch": 0.7591019150341184, + "grad_norm": 1.6165726833360532, + "learning_rate": 1.4461768920200641e-06, + "loss": 0.2485, + "step": 34486 + }, + { + "epoch": 0.7591239269205371, + "grad_norm": 1.5265645736892341, + "learning_rate": 1.4459261589688535e-06, + "loss": 0.2733, + "step": 34487 + }, + { + "epoch": 0.7591459388069558, + "grad_norm": 1.3891673155774575, + "learning_rate": 1.4456754439810295e-06, + "loss": 0.3544, + "step": 34488 + }, + { + "epoch": 0.7591679506933744, + "grad_norm": 1.4175856974747603, + "learning_rate": 1.4454247470578698e-06, + "loss": 0.268, + "step": 34489 + }, + { + "epoch": 0.7591899625797931, + "grad_norm": 1.922208812275295, + "learning_rate": 1.4451740682006472e-06, + "loss": 0.2864, + "step": 34490 + }, + { + "epoch": 0.7592119744662118, + "grad_norm": 1.8889216995731455, + "learning_rate": 1.4449234074106323e-06, + "loss": 0.304, + "step": 34491 + }, + { + "epoch": 0.7592339863526304, + "grad_norm": 2.1872933561459575, + "learning_rate": 1.4446727646891051e-06, + "loss": 0.2929, + "step": 34492 + }, + { + "epoch": 0.7592559982390491, + "grad_norm": 1.638610148437106, + "learning_rate": 1.4444221400373355e-06, + "loss": 0.3298, + "step": 34493 + }, + { + "epoch": 0.7592780101254678, + "grad_norm": 1.0600700711916329, + "learning_rate": 1.4441715334565993e-06, + "loss": 0.1696, + "step": 34494 + }, + { + "epoch": 0.7593000220118864, + "grad_norm": 0.6257037190776573, + "learning_rate": 1.44392094494817e-06, + "loss": 0.4541, + "step": 34495 + }, + { + "epoch": 0.7593220338983051, + "grad_norm": 1.1644511416734782, + "learning_rate": 1.4436703745133167e-06, + "loss": 0.2101, + "step": 34496 + }, + { + "epoch": 0.7593440457847237, + "grad_norm": 1.4390775206550086, + "learning_rate": 1.4434198221533207e-06, + "loss": 0.291, + "step": 34497 + }, + { + "epoch": 0.7593660576711424, + "grad_norm": 2.0142848548090004, + "learning_rate": 1.4431692878694491e-06, + "loss": 0.2734, + "step": 34498 + }, + { + "epoch": 0.7593880695575611, + "grad_norm": 2.1328425387035175, + "learning_rate": 1.44291877166298e-06, + "loss": 0.2397, + "step": 34499 + }, + { + "epoch": 0.7594100814439797, + "grad_norm": 1.9529678545467362, + "learning_rate": 1.4426682735351833e-06, + "loss": 0.3376, + "step": 34500 + }, + { + "epoch": 0.7594320933303984, + "grad_norm": 1.7054587701727633, + "learning_rate": 1.4424177934873317e-06, + "loss": 0.3373, + "step": 34501 + }, + { + "epoch": 0.7594541052168171, + "grad_norm": 1.4028116305835492, + "learning_rate": 1.4421673315207003e-06, + "loss": 0.3054, + "step": 34502 + }, + { + "epoch": 0.7594761171032357, + "grad_norm": 1.997771391339801, + "learning_rate": 1.4419168876365608e-06, + "loss": 0.2981, + "step": 34503 + }, + { + "epoch": 0.7594981289896544, + "grad_norm": 1.380781257584778, + "learning_rate": 1.441666461836188e-06, + "loss": 0.3383, + "step": 34504 + }, + { + "epoch": 0.7595201408760731, + "grad_norm": 1.553399242988609, + "learning_rate": 1.4414160541208533e-06, + "loss": 0.3434, + "step": 34505 + }, + { + "epoch": 0.7595421527624917, + "grad_norm": 1.4414229119561282, + "learning_rate": 1.4411656644918282e-06, + "loss": 0.243, + "step": 34506 + }, + { + "epoch": 0.7595641646489104, + "grad_norm": 0.6054404133472502, + "learning_rate": 1.440915292950386e-06, + "loss": 0.4307, + "step": 34507 + }, + { + "epoch": 0.7595861765353291, + "grad_norm": 1.56341428388772, + "learning_rate": 1.4406649394978017e-06, + "loss": 0.3038, + "step": 34508 + }, + { + "epoch": 0.7596081884217477, + "grad_norm": 1.555938689498295, + "learning_rate": 1.440414604135344e-06, + "loss": 0.3277, + "step": 34509 + }, + { + "epoch": 0.7596302003081664, + "grad_norm": 1.7517238973440368, + "learning_rate": 1.4401642868642885e-06, + "loss": 0.3556, + "step": 34510 + }, + { + "epoch": 0.7596522121945851, + "grad_norm": 1.2096833752125018, + "learning_rate": 1.4399139876859042e-06, + "loss": 0.1968, + "step": 34511 + }, + { + "epoch": 0.7596742240810037, + "grad_norm": 0.6143074117183266, + "learning_rate": 1.4396637066014652e-06, + "loss": 0.4581, + "step": 34512 + }, + { + "epoch": 0.7596962359674224, + "grad_norm": 1.440500578539065, + "learning_rate": 1.4394134436122442e-06, + "loss": 0.2845, + "step": 34513 + }, + { + "epoch": 0.7597182478538411, + "grad_norm": 0.6568453123759195, + "learning_rate": 1.439163198719511e-06, + "loss": 0.4871, + "step": 34514 + }, + { + "epoch": 0.7597402597402597, + "grad_norm": 1.7140435565286087, + "learning_rate": 1.4389129719245399e-06, + "loss": 0.2798, + "step": 34515 + }, + { + "epoch": 0.7597622716266784, + "grad_norm": 1.5535850149304018, + "learning_rate": 1.4386627632285998e-06, + "loss": 0.2373, + "step": 34516 + }, + { + "epoch": 0.7597842835130971, + "grad_norm": 2.0931000811314706, + "learning_rate": 1.438412572632964e-06, + "loss": 0.2467, + "step": 34517 + }, + { + "epoch": 0.7598062953995157, + "grad_norm": 1.3454562321358834, + "learning_rate": 1.4381624001389056e-06, + "loss": 0.3341, + "step": 34518 + }, + { + "epoch": 0.7598283072859344, + "grad_norm": 1.6585074653203264, + "learning_rate": 1.437912245747693e-06, + "loss": 0.3477, + "step": 34519 + }, + { + "epoch": 0.759850319172353, + "grad_norm": 3.424685000167625, + "learning_rate": 1.4376621094606007e-06, + "loss": 0.2633, + "step": 34520 + }, + { + "epoch": 0.7598723310587717, + "grad_norm": 1.9012187989515383, + "learning_rate": 1.4374119912788964e-06, + "loss": 0.2728, + "step": 34521 + }, + { + "epoch": 0.7598943429451904, + "grad_norm": 0.5863825064012017, + "learning_rate": 1.4371618912038538e-06, + "loss": 0.4361, + "step": 34522 + }, + { + "epoch": 0.759916354831609, + "grad_norm": 2.5996655483568265, + "learning_rate": 1.4369118092367452e-06, + "loss": 0.3156, + "step": 34523 + }, + { + "epoch": 0.7599383667180277, + "grad_norm": 1.2821735808168455, + "learning_rate": 1.4366617453788378e-06, + "loss": 0.2477, + "step": 34524 + }, + { + "epoch": 0.7599603786044464, + "grad_norm": 1.5155575409629622, + "learning_rate": 1.436411699631407e-06, + "loss": 0.2699, + "step": 34525 + }, + { + "epoch": 0.759982390490865, + "grad_norm": 1.3949424595358926, + "learning_rate": 1.436161671995721e-06, + "loss": 0.2705, + "step": 34526 + }, + { + "epoch": 0.7600044023772837, + "grad_norm": 2.7334144278160504, + "learning_rate": 1.4359116624730474e-06, + "loss": 0.3272, + "step": 34527 + }, + { + "epoch": 0.7600264142637024, + "grad_norm": 1.7736455270371607, + "learning_rate": 1.4356616710646642e-06, + "loss": 0.2868, + "step": 34528 + }, + { + "epoch": 0.760048426150121, + "grad_norm": 2.67133133951067, + "learning_rate": 1.4354116977718364e-06, + "loss": 0.2521, + "step": 34529 + }, + { + "epoch": 0.7600704380365397, + "grad_norm": 1.6764864590658979, + "learning_rate": 1.4351617425958375e-06, + "loss": 0.3368, + "step": 34530 + }, + { + "epoch": 0.7600924499229584, + "grad_norm": 3.710666147556868, + "learning_rate": 1.4349118055379364e-06, + "loss": 0.2603, + "step": 34531 + }, + { + "epoch": 0.760114461809377, + "grad_norm": 1.710457093355036, + "learning_rate": 1.4346618865994023e-06, + "loss": 0.2866, + "step": 34532 + }, + { + "epoch": 0.7601364736957957, + "grad_norm": 1.7413094686468458, + "learning_rate": 1.4344119857815064e-06, + "loss": 0.1938, + "step": 34533 + }, + { + "epoch": 0.7601584855822144, + "grad_norm": 2.024471999532808, + "learning_rate": 1.4341621030855197e-06, + "loss": 0.2871, + "step": 34534 + }, + { + "epoch": 0.760180497468633, + "grad_norm": 2.337890311563204, + "learning_rate": 1.4339122385127125e-06, + "loss": 0.3613, + "step": 34535 + }, + { + "epoch": 0.7602025093550517, + "grad_norm": 2.0194369860894623, + "learning_rate": 1.4336623920643545e-06, + "loss": 0.3721, + "step": 34536 + }, + { + "epoch": 0.7602245212414704, + "grad_norm": 1.2762618561374584, + "learning_rate": 1.433412563741713e-06, + "loss": 0.2313, + "step": 34537 + }, + { + "epoch": 0.760246533127889, + "grad_norm": 0.6161545376300971, + "learning_rate": 1.433162753546059e-06, + "loss": 0.4565, + "step": 34538 + }, + { + "epoch": 0.7602685450143077, + "grad_norm": 2.671688130359347, + "learning_rate": 1.4329129614786636e-06, + "loss": 0.3393, + "step": 34539 + }, + { + "epoch": 0.7602905569007264, + "grad_norm": 1.6794855330178922, + "learning_rate": 1.4326631875407965e-06, + "loss": 0.225, + "step": 34540 + }, + { + "epoch": 0.760312568787145, + "grad_norm": 3.3416554312348623, + "learning_rate": 1.4324134317337264e-06, + "loss": 0.4022, + "step": 34541 + }, + { + "epoch": 0.7603345806735637, + "grad_norm": 0.6353565563613854, + "learning_rate": 1.4321636940587208e-06, + "loss": 0.4738, + "step": 34542 + }, + { + "epoch": 0.7603565925599823, + "grad_norm": 2.162244540806374, + "learning_rate": 1.4319139745170508e-06, + "loss": 0.2804, + "step": 34543 + }, + { + "epoch": 0.760378604446401, + "grad_norm": 1.613655922520334, + "learning_rate": 1.4316642731099867e-06, + "loss": 0.268, + "step": 34544 + }, + { + "epoch": 0.7604006163328197, + "grad_norm": 1.2631870854898684, + "learning_rate": 1.431414589838795e-06, + "loss": 0.2429, + "step": 34545 + }, + { + "epoch": 0.7604226282192383, + "grad_norm": 1.446535609832321, + "learning_rate": 1.431164924704747e-06, + "loss": 0.2173, + "step": 34546 + }, + { + "epoch": 0.760444640105657, + "grad_norm": 1.5625253286458851, + "learning_rate": 1.4309152777091095e-06, + "loss": 0.2613, + "step": 34547 + }, + { + "epoch": 0.7604666519920757, + "grad_norm": 1.5342666951783472, + "learning_rate": 1.4306656488531524e-06, + "loss": 0.3105, + "step": 34548 + }, + { + "epoch": 0.7604886638784943, + "grad_norm": 1.926948132607622, + "learning_rate": 1.430416038138146e-06, + "loss": 0.2993, + "step": 34549 + }, + { + "epoch": 0.760510675764913, + "grad_norm": 1.5224823327971428, + "learning_rate": 1.4301664455653558e-06, + "loss": 0.3329, + "step": 34550 + }, + { + "epoch": 0.7605326876513318, + "grad_norm": 1.967200336172771, + "learning_rate": 1.4299168711360534e-06, + "loss": 0.2904, + "step": 34551 + }, + { + "epoch": 0.7605546995377503, + "grad_norm": 1.595594246079284, + "learning_rate": 1.4296673148515039e-06, + "loss": 0.2948, + "step": 34552 + }, + { + "epoch": 0.760576711424169, + "grad_norm": 1.9747040070673136, + "learning_rate": 1.429417776712978e-06, + "loss": 0.3421, + "step": 34553 + }, + { + "epoch": 0.7605987233105878, + "grad_norm": 9.995917738059697, + "learning_rate": 1.4291682567217446e-06, + "loss": 0.2701, + "step": 34554 + }, + { + "epoch": 0.7606207351970063, + "grad_norm": 1.30596417807251, + "learning_rate": 1.4289187548790695e-06, + "loss": 0.2181, + "step": 34555 + }, + { + "epoch": 0.760642747083425, + "grad_norm": 2.9291648409434323, + "learning_rate": 1.4286692711862238e-06, + "loss": 0.2195, + "step": 34556 + }, + { + "epoch": 0.7606647589698438, + "grad_norm": 2.1849126989337613, + "learning_rate": 1.4284198056444721e-06, + "loss": 0.2887, + "step": 34557 + }, + { + "epoch": 0.7606867708562624, + "grad_norm": 1.622305888510888, + "learning_rate": 1.4281703582550837e-06, + "loss": 0.3443, + "step": 34558 + }, + { + "epoch": 0.7607087827426811, + "grad_norm": 1.7327863347245693, + "learning_rate": 1.427920929019328e-06, + "loss": 0.3031, + "step": 34559 + }, + { + "epoch": 0.7607307946290998, + "grad_norm": 1.6925107587852233, + "learning_rate": 1.4276715179384698e-06, + "loss": 0.3046, + "step": 34560 + }, + { + "epoch": 0.7607528065155184, + "grad_norm": 0.619681272803276, + "learning_rate": 1.4274221250137804e-06, + "loss": 0.4631, + "step": 34561 + }, + { + "epoch": 0.7607748184019371, + "grad_norm": 0.6207642251382897, + "learning_rate": 1.4271727502465233e-06, + "loss": 0.4436, + "step": 34562 + }, + { + "epoch": 0.7607968302883558, + "grad_norm": 1.9348129579573716, + "learning_rate": 1.4269233936379673e-06, + "loss": 0.3744, + "step": 34563 + }, + { + "epoch": 0.7608188421747744, + "grad_norm": 1.7044517260688774, + "learning_rate": 1.4266740551893826e-06, + "loss": 0.2974, + "step": 34564 + }, + { + "epoch": 0.7608408540611931, + "grad_norm": 1.4290474627391443, + "learning_rate": 1.4264247349020322e-06, + "loss": 0.2234, + "step": 34565 + }, + { + "epoch": 0.7608628659476117, + "grad_norm": 1.3097373988892815, + "learning_rate": 1.4261754327771864e-06, + "loss": 0.2839, + "step": 34566 + }, + { + "epoch": 0.7608848778340304, + "grad_norm": 0.6167253338710867, + "learning_rate": 1.4259261488161113e-06, + "loss": 0.4533, + "step": 34567 + }, + { + "epoch": 0.7609068897204491, + "grad_norm": 0.6280024427053534, + "learning_rate": 1.4256768830200724e-06, + "loss": 0.4466, + "step": 34568 + }, + { + "epoch": 0.7609289016068677, + "grad_norm": 0.5997138076626503, + "learning_rate": 1.4254276353903373e-06, + "loss": 0.4378, + "step": 34569 + }, + { + "epoch": 0.7609509134932864, + "grad_norm": 2.5271618677404377, + "learning_rate": 1.425178405928173e-06, + "loss": 0.3016, + "step": 34570 + }, + { + "epoch": 0.7609729253797051, + "grad_norm": 1.392561167135782, + "learning_rate": 1.4249291946348488e-06, + "loss": 0.309, + "step": 34571 + }, + { + "epoch": 0.7609949372661237, + "grad_norm": 2.334064929910286, + "learning_rate": 1.4246800015116285e-06, + "loss": 0.3962, + "step": 34572 + }, + { + "epoch": 0.7610169491525424, + "grad_norm": 1.6119434408013515, + "learning_rate": 1.4244308265597773e-06, + "loss": 0.3417, + "step": 34573 + }, + { + "epoch": 0.7610389610389611, + "grad_norm": 1.6854893831785538, + "learning_rate": 1.4241816697805639e-06, + "loss": 0.2272, + "step": 34574 + }, + { + "epoch": 0.7610609729253797, + "grad_norm": 1.4793753266317144, + "learning_rate": 1.4239325311752539e-06, + "loss": 0.3046, + "step": 34575 + }, + { + "epoch": 0.7610829848117984, + "grad_norm": 1.6180097406886147, + "learning_rate": 1.4236834107451154e-06, + "loss": 0.2305, + "step": 34576 + }, + { + "epoch": 0.7611049966982171, + "grad_norm": 2.581329868399728, + "learning_rate": 1.4234343084914127e-06, + "loss": 0.3219, + "step": 34577 + }, + { + "epoch": 0.7611270085846357, + "grad_norm": 1.6296857714480528, + "learning_rate": 1.4231852244154103e-06, + "loss": 0.2388, + "step": 34578 + }, + { + "epoch": 0.7611490204710544, + "grad_norm": 1.5230225478990422, + "learning_rate": 1.422936158518376e-06, + "loss": 0.2647, + "step": 34579 + }, + { + "epoch": 0.7611710323574731, + "grad_norm": 1.7328430203875387, + "learning_rate": 1.4226871108015754e-06, + "loss": 0.3116, + "step": 34580 + }, + { + "epoch": 0.7611930442438917, + "grad_norm": 1.7628798293825765, + "learning_rate": 1.422438081266277e-06, + "loss": 0.3211, + "step": 34581 + }, + { + "epoch": 0.7612150561303104, + "grad_norm": 1.45365547743205, + "learning_rate": 1.4221890699137431e-06, + "loss": 0.3445, + "step": 34582 + }, + { + "epoch": 0.7612370680167291, + "grad_norm": 1.4104606586956172, + "learning_rate": 1.4219400767452384e-06, + "loss": 0.2393, + "step": 34583 + }, + { + "epoch": 0.7612590799031477, + "grad_norm": 1.7706688301515079, + "learning_rate": 1.4216911017620305e-06, + "loss": 0.3626, + "step": 34584 + }, + { + "epoch": 0.7612810917895664, + "grad_norm": 1.826977246645452, + "learning_rate": 1.4214421449653853e-06, + "loss": 0.292, + "step": 34585 + }, + { + "epoch": 0.7613031036759851, + "grad_norm": 2.367971958418195, + "learning_rate": 1.4211932063565665e-06, + "loss": 0.3207, + "step": 34586 + }, + { + "epoch": 0.7613251155624037, + "grad_norm": 1.7449554027825003, + "learning_rate": 1.4209442859368411e-06, + "loss": 0.3078, + "step": 34587 + }, + { + "epoch": 0.7613471274488224, + "grad_norm": 2.352314617613835, + "learning_rate": 1.4206953837074717e-06, + "loss": 0.2863, + "step": 34588 + }, + { + "epoch": 0.7613691393352411, + "grad_norm": 1.5297661084181513, + "learning_rate": 1.4204464996697248e-06, + "loss": 0.3014, + "step": 34589 + }, + { + "epoch": 0.7613911512216597, + "grad_norm": 1.7685420736795887, + "learning_rate": 1.4201976338248663e-06, + "loss": 0.2856, + "step": 34590 + }, + { + "epoch": 0.7614131631080784, + "grad_norm": 0.6262218703310128, + "learning_rate": 1.4199487861741596e-06, + "loss": 0.4564, + "step": 34591 + }, + { + "epoch": 0.761435174994497, + "grad_norm": 0.631171481103483, + "learning_rate": 1.4196999567188707e-06, + "loss": 0.4665, + "step": 34592 + }, + { + "epoch": 0.7614571868809157, + "grad_norm": 1.6486678738202336, + "learning_rate": 1.4194511454602623e-06, + "loss": 0.2762, + "step": 34593 + }, + { + "epoch": 0.7614791987673344, + "grad_norm": 2.386049562649357, + "learning_rate": 1.4192023523995996e-06, + "loss": 0.4022, + "step": 34594 + }, + { + "epoch": 0.761501210653753, + "grad_norm": 0.6176163273987952, + "learning_rate": 1.4189535775381496e-06, + "loss": 0.448, + "step": 34595 + }, + { + "epoch": 0.7615232225401717, + "grad_norm": 0.606869866075813, + "learning_rate": 1.4187048208771731e-06, + "loss": 0.4451, + "step": 34596 + }, + { + "epoch": 0.7615452344265904, + "grad_norm": 0.6094584155802616, + "learning_rate": 1.4184560824179373e-06, + "loss": 0.4563, + "step": 34597 + }, + { + "epoch": 0.761567246313009, + "grad_norm": 0.58893012509749, + "learning_rate": 1.4182073621617037e-06, + "loss": 0.4308, + "step": 34598 + }, + { + "epoch": 0.7615892581994277, + "grad_norm": 1.4715649708181553, + "learning_rate": 1.4179586601097384e-06, + "loss": 0.2939, + "step": 34599 + }, + { + "epoch": 0.7616112700858464, + "grad_norm": 0.6311160669342326, + "learning_rate": 1.4177099762633057e-06, + "loss": 0.4622, + "step": 34600 + }, + { + "epoch": 0.761633281972265, + "grad_norm": 1.253585705783669, + "learning_rate": 1.4174613106236674e-06, + "loss": 0.2253, + "step": 34601 + }, + { + "epoch": 0.7616552938586837, + "grad_norm": 0.6224028184488369, + "learning_rate": 1.4172126631920895e-06, + "loss": 0.4751, + "step": 34602 + }, + { + "epoch": 0.7616773057451024, + "grad_norm": 1.5416911129667221, + "learning_rate": 1.4169640339698343e-06, + "loss": 0.267, + "step": 34603 + }, + { + "epoch": 0.761699317631521, + "grad_norm": 1.5135597135745031, + "learning_rate": 1.416715422958167e-06, + "loss": 0.2291, + "step": 34604 + }, + { + "epoch": 0.7617213295179397, + "grad_norm": 0.617570123540337, + "learning_rate": 1.4164668301583484e-06, + "loss": 0.4612, + "step": 34605 + }, + { + "epoch": 0.7617433414043584, + "grad_norm": 1.7267476575689007, + "learning_rate": 1.4162182555716436e-06, + "loss": 0.245, + "step": 34606 + }, + { + "epoch": 0.761765353290777, + "grad_norm": 3.48965640397711, + "learning_rate": 1.4159696991993176e-06, + "loss": 0.2555, + "step": 34607 + }, + { + "epoch": 0.7617873651771957, + "grad_norm": 1.7210628165738047, + "learning_rate": 1.4157211610426325e-06, + "loss": 0.2609, + "step": 34608 + }, + { + "epoch": 0.7618093770636144, + "grad_norm": 1.577237661115097, + "learning_rate": 1.4154726411028497e-06, + "loss": 0.2795, + "step": 34609 + }, + { + "epoch": 0.761831388950033, + "grad_norm": 1.8114032926621373, + "learning_rate": 1.4152241393812332e-06, + "loss": 0.248, + "step": 34610 + }, + { + "epoch": 0.7618534008364517, + "grad_norm": 0.6265749058979823, + "learning_rate": 1.4149756558790467e-06, + "loss": 0.458, + "step": 34611 + }, + { + "epoch": 0.7618754127228704, + "grad_norm": 3.651454739430043, + "learning_rate": 1.414727190597554e-06, + "loss": 0.3396, + "step": 34612 + }, + { + "epoch": 0.761897424609289, + "grad_norm": 1.833303505237465, + "learning_rate": 1.414478743538017e-06, + "loss": 0.3967, + "step": 34613 + }, + { + "epoch": 0.7619194364957077, + "grad_norm": 0.6113150591282231, + "learning_rate": 1.4142303147016973e-06, + "loss": 0.4542, + "step": 34614 + }, + { + "epoch": 0.7619414483821263, + "grad_norm": 0.6106524674847686, + "learning_rate": 1.4139819040898578e-06, + "loss": 0.4513, + "step": 34615 + }, + { + "epoch": 0.761963460268545, + "grad_norm": 0.6125002129370196, + "learning_rate": 1.4137335117037616e-06, + "loss": 0.4568, + "step": 34616 + }, + { + "epoch": 0.7619854721549637, + "grad_norm": 1.1956128006351388, + "learning_rate": 1.413485137544673e-06, + "loss": 0.3218, + "step": 34617 + }, + { + "epoch": 0.7620074840413823, + "grad_norm": 1.6976795996313676, + "learning_rate": 1.4132367816138526e-06, + "loss": 0.2397, + "step": 34618 + }, + { + "epoch": 0.762029495927801, + "grad_norm": 1.8283209523686565, + "learning_rate": 1.4129884439125607e-06, + "loss": 0.2839, + "step": 34619 + }, + { + "epoch": 0.7620515078142197, + "grad_norm": 1.365803567938908, + "learning_rate": 1.412740124442062e-06, + "loss": 0.2051, + "step": 34620 + }, + { + "epoch": 0.7620735197006383, + "grad_norm": 0.6337545348844863, + "learning_rate": 1.4124918232036177e-06, + "loss": 0.4712, + "step": 34621 + }, + { + "epoch": 0.762095531587057, + "grad_norm": 0.6267351627129543, + "learning_rate": 1.4122435401984913e-06, + "loss": 0.4515, + "step": 34622 + }, + { + "epoch": 0.7621175434734757, + "grad_norm": 1.4249357257316253, + "learning_rate": 1.4119952754279436e-06, + "loss": 0.3037, + "step": 34623 + }, + { + "epoch": 0.7621395553598943, + "grad_norm": 0.6077612411583665, + "learning_rate": 1.4117470288932345e-06, + "loss": 0.4202, + "step": 34624 + }, + { + "epoch": 0.762161567246313, + "grad_norm": 1.5932311322740653, + "learning_rate": 1.4114988005956277e-06, + "loss": 0.3042, + "step": 34625 + }, + { + "epoch": 0.7621835791327317, + "grad_norm": 1.920911035336278, + "learning_rate": 1.411250590536386e-06, + "loss": 0.2776, + "step": 34626 + }, + { + "epoch": 0.7622055910191503, + "grad_norm": 2.942689062714689, + "learning_rate": 1.4110023987167677e-06, + "loss": 0.2832, + "step": 34627 + }, + { + "epoch": 0.762227602905569, + "grad_norm": 2.404566699913981, + "learning_rate": 1.410754225138038e-06, + "loss": 0.2255, + "step": 34628 + }, + { + "epoch": 0.7622496147919877, + "grad_norm": 1.7414317292393213, + "learning_rate": 1.4105060698014539e-06, + "loss": 0.3013, + "step": 34629 + }, + { + "epoch": 0.7622716266784063, + "grad_norm": 0.643647211542347, + "learning_rate": 1.4102579327082794e-06, + "loss": 0.4693, + "step": 34630 + }, + { + "epoch": 0.762293638564825, + "grad_norm": 1.8932558364420633, + "learning_rate": 1.410009813859776e-06, + "loss": 0.3141, + "step": 34631 + }, + { + "epoch": 0.7623156504512437, + "grad_norm": 1.4329002511511115, + "learning_rate": 1.4097617132572034e-06, + "loss": 0.2694, + "step": 34632 + }, + { + "epoch": 0.7623376623376623, + "grad_norm": 1.5036612453333877, + "learning_rate": 1.4095136309018242e-06, + "loss": 0.2783, + "step": 34633 + }, + { + "epoch": 0.762359674224081, + "grad_norm": 1.3539019770436211, + "learning_rate": 1.4092655667948962e-06, + "loss": 0.3295, + "step": 34634 + }, + { + "epoch": 0.7623816861104997, + "grad_norm": 1.3197347407855458, + "learning_rate": 1.4090175209376828e-06, + "loss": 0.2719, + "step": 34635 + }, + { + "epoch": 0.7624036979969183, + "grad_norm": 4.028580683445002, + "learning_rate": 1.408769493331445e-06, + "loss": 0.3153, + "step": 34636 + }, + { + "epoch": 0.762425709883337, + "grad_norm": 0.5977138924786854, + "learning_rate": 1.408521483977441e-06, + "loss": 0.4599, + "step": 34637 + }, + { + "epoch": 0.7624477217697556, + "grad_norm": 1.9561677698500772, + "learning_rate": 1.4082734928769342e-06, + "loss": 0.2836, + "step": 34638 + }, + { + "epoch": 0.7624697336561743, + "grad_norm": 2.4054466949273383, + "learning_rate": 1.4080255200311815e-06, + "loss": 0.2526, + "step": 34639 + }, + { + "epoch": 0.762491745542593, + "grad_norm": 0.6144050994681702, + "learning_rate": 1.4077775654414472e-06, + "loss": 0.4655, + "step": 34640 + }, + { + "epoch": 0.7625137574290116, + "grad_norm": 0.6457215759174627, + "learning_rate": 1.407529629108988e-06, + "loss": 0.4695, + "step": 34641 + }, + { + "epoch": 0.7625357693154303, + "grad_norm": 1.5447609999735679, + "learning_rate": 1.4072817110350656e-06, + "loss": 0.2262, + "step": 34642 + }, + { + "epoch": 0.762557781201849, + "grad_norm": 2.9168138297170856, + "learning_rate": 1.407033811220941e-06, + "loss": 0.3335, + "step": 34643 + }, + { + "epoch": 0.7625797930882676, + "grad_norm": 0.6130912204496382, + "learning_rate": 1.4067859296678731e-06, + "loss": 0.4637, + "step": 34644 + }, + { + "epoch": 0.7626018049746863, + "grad_norm": 2.0602958017291764, + "learning_rate": 1.4065380663771205e-06, + "loss": 0.2883, + "step": 34645 + }, + { + "epoch": 0.762623816861105, + "grad_norm": 0.624005749459946, + "learning_rate": 1.406290221349944e-06, + "loss": 0.4608, + "step": 34646 + }, + { + "epoch": 0.7626458287475236, + "grad_norm": 2.118923742319596, + "learning_rate": 1.406042394587604e-06, + "loss": 0.2583, + "step": 34647 + }, + { + "epoch": 0.7626678406339423, + "grad_norm": 1.8699588086642385, + "learning_rate": 1.4057945860913601e-06, + "loss": 0.2621, + "step": 34648 + }, + { + "epoch": 0.762689852520361, + "grad_norm": 1.959086906163737, + "learning_rate": 1.405546795862472e-06, + "loss": 0.3339, + "step": 34649 + }, + { + "epoch": 0.7627118644067796, + "grad_norm": 3.4448647561729686, + "learning_rate": 1.4052990239021963e-06, + "loss": 0.2525, + "step": 34650 + }, + { + "epoch": 0.7627338762931983, + "grad_norm": 2.6317735175066375, + "learning_rate": 1.4050512702117942e-06, + "loss": 0.3463, + "step": 34651 + }, + { + "epoch": 0.762755888179617, + "grad_norm": 2.265594402478734, + "learning_rate": 1.4048035347925249e-06, + "loss": 0.3013, + "step": 34652 + }, + { + "epoch": 0.7627779000660356, + "grad_norm": 2.1144443318430266, + "learning_rate": 1.4045558176456492e-06, + "loss": 0.3497, + "step": 34653 + }, + { + "epoch": 0.7627999119524543, + "grad_norm": 0.6377368675076185, + "learning_rate": 1.4043081187724244e-06, + "loss": 0.4475, + "step": 34654 + }, + { + "epoch": 0.762821923838873, + "grad_norm": 1.9683496580980924, + "learning_rate": 1.404060438174108e-06, + "loss": 0.2535, + "step": 34655 + }, + { + "epoch": 0.7628439357252916, + "grad_norm": 2.071861464262556, + "learning_rate": 1.4038127758519604e-06, + "loss": 0.2285, + "step": 34656 + }, + { + "epoch": 0.7628659476117103, + "grad_norm": 1.6861089417246455, + "learning_rate": 1.4035651318072402e-06, + "loss": 0.3507, + "step": 34657 + }, + { + "epoch": 0.762887959498129, + "grad_norm": 1.2765477503207436, + "learning_rate": 1.4033175060412075e-06, + "loss": 0.2669, + "step": 34658 + }, + { + "epoch": 0.7629099713845476, + "grad_norm": 1.3848865789035287, + "learning_rate": 1.403069898555119e-06, + "loss": 0.1742, + "step": 34659 + }, + { + "epoch": 0.7629319832709663, + "grad_norm": 1.6413680315528638, + "learning_rate": 1.4028223093502325e-06, + "loss": 0.2455, + "step": 34660 + }, + { + "epoch": 0.7629539951573849, + "grad_norm": 2.432194539143317, + "learning_rate": 1.402574738427807e-06, + "loss": 0.2233, + "step": 34661 + }, + { + "epoch": 0.7629760070438036, + "grad_norm": 1.4250969350464648, + "learning_rate": 1.402327185789102e-06, + "loss": 0.3022, + "step": 34662 + }, + { + "epoch": 0.7629980189302223, + "grad_norm": 3.8888209226009307, + "learning_rate": 1.4020796514353756e-06, + "loss": 0.2825, + "step": 34663 + }, + { + "epoch": 0.7630200308166409, + "grad_norm": 1.5551077288996218, + "learning_rate": 1.4018321353678853e-06, + "loss": 0.3056, + "step": 34664 + }, + { + "epoch": 0.7630420427030596, + "grad_norm": 1.6726819060607947, + "learning_rate": 1.4015846375878878e-06, + "loss": 0.2426, + "step": 34665 + }, + { + "epoch": 0.7630640545894783, + "grad_norm": 1.9562084888652496, + "learning_rate": 1.4013371580966417e-06, + "loss": 0.2876, + "step": 34666 + }, + { + "epoch": 0.7630860664758969, + "grad_norm": 0.6364683668217032, + "learning_rate": 1.401089696895407e-06, + "loss": 0.4607, + "step": 34667 + }, + { + "epoch": 0.7631080783623156, + "grad_norm": 1.7535107992947319, + "learning_rate": 1.4008422539854378e-06, + "loss": 0.2464, + "step": 34668 + }, + { + "epoch": 0.7631300902487343, + "grad_norm": 1.6696929342960973, + "learning_rate": 1.400594829367995e-06, + "loss": 0.3326, + "step": 34669 + }, + { + "epoch": 0.7631521021351529, + "grad_norm": 1.5920527561496272, + "learning_rate": 1.4003474230443332e-06, + "loss": 0.2611, + "step": 34670 + }, + { + "epoch": 0.7631741140215716, + "grad_norm": 1.626050086303433, + "learning_rate": 1.4001000350157112e-06, + "loss": 0.3121, + "step": 34671 + }, + { + "epoch": 0.7631961259079904, + "grad_norm": 1.7870840447004168, + "learning_rate": 1.3998526652833883e-06, + "loss": 0.3155, + "step": 34672 + }, + { + "epoch": 0.763218137794409, + "grad_norm": 1.6629167005293775, + "learning_rate": 1.3996053138486182e-06, + "loss": 0.3527, + "step": 34673 + }, + { + "epoch": 0.7632401496808277, + "grad_norm": 1.5508400067131856, + "learning_rate": 1.399357980712661e-06, + "loss": 0.342, + "step": 34674 + }, + { + "epoch": 0.7632621615672464, + "grad_norm": 2.7467496292329074, + "learning_rate": 1.3991106658767711e-06, + "loss": 0.3245, + "step": 34675 + }, + { + "epoch": 0.763284173453665, + "grad_norm": 1.2523156355617773, + "learning_rate": 1.3988633693422088e-06, + "loss": 0.246, + "step": 34676 + }, + { + "epoch": 0.7633061853400837, + "grad_norm": 2.1142927224879866, + "learning_rate": 1.398616091110227e-06, + "loss": 0.2813, + "step": 34677 + }, + { + "epoch": 0.7633281972265024, + "grad_norm": 1.8569781798424785, + "learning_rate": 1.3983688311820847e-06, + "loss": 0.3103, + "step": 34678 + }, + { + "epoch": 0.763350209112921, + "grad_norm": 2.435323415392495, + "learning_rate": 1.3981215895590395e-06, + "loss": 0.2654, + "step": 34679 + }, + { + "epoch": 0.7633722209993397, + "grad_norm": 1.9505570383526751, + "learning_rate": 1.3978743662423456e-06, + "loss": 0.2306, + "step": 34680 + }, + { + "epoch": 0.7633942328857584, + "grad_norm": 0.6151645961867838, + "learning_rate": 1.3976271612332626e-06, + "loss": 0.4533, + "step": 34681 + }, + { + "epoch": 0.763416244772177, + "grad_norm": 0.5928499288896388, + "learning_rate": 1.397379974533043e-06, + "loss": 0.4462, + "step": 34682 + }, + { + "epoch": 0.7634382566585957, + "grad_norm": 1.7011943026215166, + "learning_rate": 1.3971328061429456e-06, + "loss": 0.3681, + "step": 34683 + }, + { + "epoch": 0.7634602685450143, + "grad_norm": 1.5632291197711323, + "learning_rate": 1.3968856560642275e-06, + "loss": 0.3161, + "step": 34684 + }, + { + "epoch": 0.763482280431433, + "grad_norm": 2.455457887208677, + "learning_rate": 1.3966385242981434e-06, + "loss": 0.3839, + "step": 34685 + }, + { + "epoch": 0.7635042923178517, + "grad_norm": 2.5673190832705495, + "learning_rate": 1.3963914108459486e-06, + "loss": 0.3539, + "step": 34686 + }, + { + "epoch": 0.7635263042042703, + "grad_norm": 1.4585301362117888, + "learning_rate": 1.3961443157088994e-06, + "loss": 0.2642, + "step": 34687 + }, + { + "epoch": 0.763548316090689, + "grad_norm": 1.3406653680683087, + "learning_rate": 1.3958972388882519e-06, + "loss": 0.2395, + "step": 34688 + }, + { + "epoch": 0.7635703279771077, + "grad_norm": 1.9649104034166553, + "learning_rate": 1.395650180385264e-06, + "loss": 0.3048, + "step": 34689 + }, + { + "epoch": 0.7635923398635263, + "grad_norm": 2.4506944330489993, + "learning_rate": 1.3954031402011898e-06, + "loss": 0.2871, + "step": 34690 + }, + { + "epoch": 0.763614351749945, + "grad_norm": 1.6240487234466585, + "learning_rate": 1.3951561183372824e-06, + "loss": 0.2467, + "step": 34691 + }, + { + "epoch": 0.7636363636363637, + "grad_norm": 0.6187982187163216, + "learning_rate": 1.3949091147947996e-06, + "loss": 0.4615, + "step": 34692 + }, + { + "epoch": 0.7636583755227823, + "grad_norm": 0.5933829483425384, + "learning_rate": 1.3946621295749962e-06, + "loss": 0.4711, + "step": 34693 + }, + { + "epoch": 0.763680387409201, + "grad_norm": 1.6311913265723155, + "learning_rate": 1.3944151626791301e-06, + "loss": 0.2666, + "step": 34694 + }, + { + "epoch": 0.7637023992956197, + "grad_norm": 7.016682641216695, + "learning_rate": 1.394168214108454e-06, + "loss": 0.2794, + "step": 34695 + }, + { + "epoch": 0.7637244111820383, + "grad_norm": 1.5939362688690732, + "learning_rate": 1.393921283864222e-06, + "loss": 0.2577, + "step": 34696 + }, + { + "epoch": 0.763746423068457, + "grad_norm": 0.6306478415761594, + "learning_rate": 1.3936743719476897e-06, + "loss": 0.4476, + "step": 34697 + }, + { + "epoch": 0.7637684349548757, + "grad_norm": 1.7725551472782841, + "learning_rate": 1.3934274783601131e-06, + "loss": 0.3008, + "step": 34698 + }, + { + "epoch": 0.7637904468412943, + "grad_norm": 1.498100614716328, + "learning_rate": 1.3931806031027484e-06, + "loss": 0.2902, + "step": 34699 + }, + { + "epoch": 0.763812458727713, + "grad_norm": 1.277318426383211, + "learning_rate": 1.3929337461768483e-06, + "loss": 0.211, + "step": 34700 + }, + { + "epoch": 0.7638344706141317, + "grad_norm": 1.1495573259986132, + "learning_rate": 1.3926869075836663e-06, + "loss": 0.2304, + "step": 34701 + }, + { + "epoch": 0.7638564825005503, + "grad_norm": 1.6433205842938023, + "learning_rate": 1.3924400873244588e-06, + "loss": 0.2669, + "step": 34702 + }, + { + "epoch": 0.763878494386969, + "grad_norm": 1.9612777501852368, + "learning_rate": 1.3921932854004787e-06, + "loss": 0.3089, + "step": 34703 + }, + { + "epoch": 0.7639005062733877, + "grad_norm": 1.462562624352769, + "learning_rate": 1.3919465018129835e-06, + "loss": 0.2096, + "step": 34704 + }, + { + "epoch": 0.7639225181598063, + "grad_norm": 0.6212009322115886, + "learning_rate": 1.3916997365632257e-06, + "loss": 0.4559, + "step": 34705 + }, + { + "epoch": 0.763944530046225, + "grad_norm": 1.3150450862610457, + "learning_rate": 1.3914529896524576e-06, + "loss": 0.2623, + "step": 34706 + }, + { + "epoch": 0.7639665419326436, + "grad_norm": 1.649827669053503, + "learning_rate": 1.3912062610819344e-06, + "loss": 0.3325, + "step": 34707 + }, + { + "epoch": 0.7639885538190623, + "grad_norm": 2.388255379096332, + "learning_rate": 1.3909595508529128e-06, + "loss": 0.2114, + "step": 34708 + }, + { + "epoch": 0.764010565705481, + "grad_norm": 1.3333698517123103, + "learning_rate": 1.3907128589666424e-06, + "loss": 0.2657, + "step": 34709 + }, + { + "epoch": 0.7640325775918996, + "grad_norm": 3.063649940434928, + "learning_rate": 1.3904661854243806e-06, + "loss": 0.3788, + "step": 34710 + }, + { + "epoch": 0.7640545894783183, + "grad_norm": 1.5856191343120003, + "learning_rate": 1.390219530227378e-06, + "loss": 0.2807, + "step": 34711 + }, + { + "epoch": 0.764076601364737, + "grad_norm": 2.060062637574363, + "learning_rate": 1.3899728933768914e-06, + "loss": 0.3278, + "step": 34712 + }, + { + "epoch": 0.7640986132511556, + "grad_norm": 1.4479790649145048, + "learning_rate": 1.3897262748741709e-06, + "loss": 0.3671, + "step": 34713 + }, + { + "epoch": 0.7641206251375743, + "grad_norm": 0.6230337267680915, + "learning_rate": 1.389479674720472e-06, + "loss": 0.4396, + "step": 34714 + }, + { + "epoch": 0.764142637023993, + "grad_norm": 1.6225510135169847, + "learning_rate": 1.3892330929170488e-06, + "loss": 0.3153, + "step": 34715 + }, + { + "epoch": 0.7641646489104116, + "grad_norm": 0.6260052596685776, + "learning_rate": 1.3889865294651527e-06, + "loss": 0.4541, + "step": 34716 + }, + { + "epoch": 0.7641866607968303, + "grad_norm": 1.5817078010053542, + "learning_rate": 1.3887399843660386e-06, + "loss": 0.3642, + "step": 34717 + }, + { + "epoch": 0.764208672683249, + "grad_norm": 1.213980995138162, + "learning_rate": 1.3884934576209575e-06, + "loss": 0.2981, + "step": 34718 + }, + { + "epoch": 0.7642306845696676, + "grad_norm": 1.5874965485471326, + "learning_rate": 1.388246949231163e-06, + "loss": 0.2459, + "step": 34719 + }, + { + "epoch": 0.7642526964560863, + "grad_norm": 1.5086200282288844, + "learning_rate": 1.3880004591979096e-06, + "loss": 0.2423, + "step": 34720 + }, + { + "epoch": 0.764274708342505, + "grad_norm": 1.419556300288118, + "learning_rate": 1.3877539875224477e-06, + "loss": 0.377, + "step": 34721 + }, + { + "epoch": 0.7642967202289236, + "grad_norm": 2.054207994671425, + "learning_rate": 1.387507534206033e-06, + "loss": 0.2636, + "step": 34722 + }, + { + "epoch": 0.7643187321153423, + "grad_norm": 0.6154521828521069, + "learning_rate": 1.3872610992499146e-06, + "loss": 0.4463, + "step": 34723 + }, + { + "epoch": 0.764340744001761, + "grad_norm": 1.690870454722676, + "learning_rate": 1.387014682655346e-06, + "loss": 0.3642, + "step": 34724 + }, + { + "epoch": 0.7643627558881796, + "grad_norm": 0.6170431697448745, + "learning_rate": 1.3867682844235825e-06, + "loss": 0.4345, + "step": 34725 + }, + { + "epoch": 0.7643847677745983, + "grad_norm": 1.3692601840133805, + "learning_rate": 1.3865219045558737e-06, + "loss": 0.2576, + "step": 34726 + }, + { + "epoch": 0.764406779661017, + "grad_norm": 2.4775360500654053, + "learning_rate": 1.3862755430534707e-06, + "loss": 0.3111, + "step": 34727 + }, + { + "epoch": 0.7644287915474356, + "grad_norm": 0.6229454484891874, + "learning_rate": 1.3860291999176273e-06, + "loss": 0.454, + "step": 34728 + }, + { + "epoch": 0.7644508034338543, + "grad_norm": 2.5635189409204173, + "learning_rate": 1.3857828751495955e-06, + "loss": 0.2812, + "step": 34729 + }, + { + "epoch": 0.764472815320273, + "grad_norm": 1.3703025971104892, + "learning_rate": 1.3855365687506283e-06, + "loss": 0.3149, + "step": 34730 + }, + { + "epoch": 0.7644948272066916, + "grad_norm": 1.6275394286375513, + "learning_rate": 1.385290280721977e-06, + "loss": 0.2308, + "step": 34731 + }, + { + "epoch": 0.7645168390931103, + "grad_norm": 1.7550415999104159, + "learning_rate": 1.3850440110648905e-06, + "loss": 0.329, + "step": 34732 + }, + { + "epoch": 0.7645388509795289, + "grad_norm": 1.9609336362521033, + "learning_rate": 1.384797759780623e-06, + "loss": 0.3308, + "step": 34733 + }, + { + "epoch": 0.7645608628659476, + "grad_norm": 4.427873081865791, + "learning_rate": 1.3845515268704256e-06, + "loss": 0.2347, + "step": 34734 + }, + { + "epoch": 0.7645828747523663, + "grad_norm": 1.6407842289217085, + "learning_rate": 1.3843053123355515e-06, + "loss": 0.3507, + "step": 34735 + }, + { + "epoch": 0.7646048866387849, + "grad_norm": 1.5065804025097203, + "learning_rate": 1.3840591161772498e-06, + "loss": 0.2305, + "step": 34736 + }, + { + "epoch": 0.7646268985252036, + "grad_norm": 1.7053600330471597, + "learning_rate": 1.383812938396772e-06, + "loss": 0.3053, + "step": 34737 + }, + { + "epoch": 0.7646489104116223, + "grad_norm": 1.4764252921514607, + "learning_rate": 1.383566778995369e-06, + "loss": 0.2517, + "step": 34738 + }, + { + "epoch": 0.7646709222980409, + "grad_norm": 0.604472057228728, + "learning_rate": 1.3833206379742924e-06, + "loss": 0.4092, + "step": 34739 + }, + { + "epoch": 0.7646929341844596, + "grad_norm": 3.60834521764518, + "learning_rate": 1.3830745153347958e-06, + "loss": 0.2628, + "step": 34740 + }, + { + "epoch": 0.7647149460708783, + "grad_norm": 2.1124552726468546, + "learning_rate": 1.382828411078127e-06, + "loss": 0.2939, + "step": 34741 + }, + { + "epoch": 0.7647369579572969, + "grad_norm": 1.2800699671801998, + "learning_rate": 1.382582325205536e-06, + "loss": 0.276, + "step": 34742 + }, + { + "epoch": 0.7647589698437156, + "grad_norm": 2.4400284501988687, + "learning_rate": 1.3823362577182753e-06, + "loss": 0.247, + "step": 34743 + }, + { + "epoch": 0.7647809817301343, + "grad_norm": 1.3037807714273206, + "learning_rate": 1.3820902086175968e-06, + "loss": 0.3261, + "step": 34744 + }, + { + "epoch": 0.7648029936165529, + "grad_norm": 1.7857050894699962, + "learning_rate": 1.3818441779047476e-06, + "loss": 0.2751, + "step": 34745 + }, + { + "epoch": 0.7648250055029716, + "grad_norm": 0.6240817890892549, + "learning_rate": 1.3815981655809819e-06, + "loss": 0.4505, + "step": 34746 + }, + { + "epoch": 0.7648470173893903, + "grad_norm": 2.916490172679197, + "learning_rate": 1.3813521716475464e-06, + "loss": 0.3315, + "step": 34747 + }, + { + "epoch": 0.7648690292758089, + "grad_norm": 3.230639239406206, + "learning_rate": 1.381106196105695e-06, + "loss": 0.34, + "step": 34748 + }, + { + "epoch": 0.7648910411622276, + "grad_norm": 2.2039335036169643, + "learning_rate": 1.3808602389566743e-06, + "loss": 0.2858, + "step": 34749 + }, + { + "epoch": 0.7649130530486463, + "grad_norm": 1.4449107434629789, + "learning_rate": 1.3806143002017363e-06, + "loss": 0.2272, + "step": 34750 + }, + { + "epoch": 0.7649350649350649, + "grad_norm": 2.4074885124744196, + "learning_rate": 1.380368379842132e-06, + "loss": 0.3579, + "step": 34751 + }, + { + "epoch": 0.7649570768214836, + "grad_norm": 4.1414247474456065, + "learning_rate": 1.3801224778791084e-06, + "loss": 0.2478, + "step": 34752 + }, + { + "epoch": 0.7649790887079023, + "grad_norm": 1.8827673921692838, + "learning_rate": 1.3798765943139182e-06, + "loss": 0.301, + "step": 34753 + }, + { + "epoch": 0.7650011005943209, + "grad_norm": 0.6158305220066744, + "learning_rate": 1.3796307291478084e-06, + "loss": 0.4345, + "step": 34754 + }, + { + "epoch": 0.7650231124807396, + "grad_norm": 3.5774782997744135, + "learning_rate": 1.3793848823820306e-06, + "loss": 0.2425, + "step": 34755 + }, + { + "epoch": 0.7650451243671582, + "grad_norm": 1.5969647413961927, + "learning_rate": 1.3791390540178346e-06, + "loss": 0.2854, + "step": 34756 + }, + { + "epoch": 0.7650671362535769, + "grad_norm": 0.6283581672957562, + "learning_rate": 1.3788932440564673e-06, + "loss": 0.4507, + "step": 34757 + }, + { + "epoch": 0.7650891481399956, + "grad_norm": 1.4153644462867079, + "learning_rate": 1.378647452499181e-06, + "loss": 0.3591, + "step": 34758 + }, + { + "epoch": 0.7651111600264142, + "grad_norm": 0.6082609519995068, + "learning_rate": 1.3784016793472222e-06, + "loss": 0.4682, + "step": 34759 + }, + { + "epoch": 0.7651331719128329, + "grad_norm": 3.1325297718436262, + "learning_rate": 1.378155924601841e-06, + "loss": 0.2427, + "step": 34760 + }, + { + "epoch": 0.7651551837992516, + "grad_norm": 1.358357736359061, + "learning_rate": 1.3779101882642892e-06, + "loss": 0.2241, + "step": 34761 + }, + { + "epoch": 0.7651771956856702, + "grad_norm": 1.2938769843522602, + "learning_rate": 1.3776644703358111e-06, + "loss": 0.2052, + "step": 34762 + }, + { + "epoch": 0.7651992075720889, + "grad_norm": 1.7110914421493286, + "learning_rate": 1.3774187708176596e-06, + "loss": 0.2386, + "step": 34763 + }, + { + "epoch": 0.7652212194585076, + "grad_norm": 2.119226948492909, + "learning_rate": 1.3771730897110797e-06, + "loss": 0.3237, + "step": 34764 + }, + { + "epoch": 0.7652432313449262, + "grad_norm": 1.5880065255910552, + "learning_rate": 1.3769274270173222e-06, + "loss": 0.3661, + "step": 34765 + }, + { + "epoch": 0.7652652432313449, + "grad_norm": 1.6641093697460343, + "learning_rate": 1.3766817827376372e-06, + "loss": 0.3519, + "step": 34766 + }, + { + "epoch": 0.7652872551177636, + "grad_norm": 4.388013607541909, + "learning_rate": 1.376436156873271e-06, + "loss": 0.2908, + "step": 34767 + }, + { + "epoch": 0.7653092670041822, + "grad_norm": 1.357040712383888, + "learning_rate": 1.3761905494254712e-06, + "loss": 0.3437, + "step": 34768 + }, + { + "epoch": 0.7653312788906009, + "grad_norm": 1.4311995833296312, + "learning_rate": 1.3759449603954872e-06, + "loss": 0.2635, + "step": 34769 + }, + { + "epoch": 0.7653532907770196, + "grad_norm": 0.6103282455697664, + "learning_rate": 1.3756993897845673e-06, + "loss": 0.4321, + "step": 34770 + }, + { + "epoch": 0.7653753026634382, + "grad_norm": 1.2820167689031583, + "learning_rate": 1.3754538375939613e-06, + "loss": 0.2755, + "step": 34771 + }, + { + "epoch": 0.7653973145498569, + "grad_norm": 1.5799217834542467, + "learning_rate": 1.3752083038249154e-06, + "loss": 0.3498, + "step": 34772 + }, + { + "epoch": 0.7654193264362756, + "grad_norm": 1.8583686096427225, + "learning_rate": 1.3749627884786753e-06, + "loss": 0.3079, + "step": 34773 + }, + { + "epoch": 0.7654413383226942, + "grad_norm": 0.6165903703699008, + "learning_rate": 1.374717291556492e-06, + "loss": 0.4356, + "step": 34774 + }, + { + "epoch": 0.7654633502091129, + "grad_norm": 0.5925845857646632, + "learning_rate": 1.3744718130596123e-06, + "loss": 0.4489, + "step": 34775 + }, + { + "epoch": 0.7654853620955316, + "grad_norm": 1.453165435232122, + "learning_rate": 1.3742263529892851e-06, + "loss": 0.2957, + "step": 34776 + }, + { + "epoch": 0.7655073739819502, + "grad_norm": 0.6246078908989032, + "learning_rate": 1.3739809113467566e-06, + "loss": 0.4511, + "step": 34777 + }, + { + "epoch": 0.7655293858683689, + "grad_norm": 1.4310572190134463, + "learning_rate": 1.373735488133273e-06, + "loss": 0.2544, + "step": 34778 + }, + { + "epoch": 0.7655513977547875, + "grad_norm": 1.7476870562836642, + "learning_rate": 1.3734900833500846e-06, + "loss": 0.3099, + "step": 34779 + }, + { + "epoch": 0.7655734096412062, + "grad_norm": 1.2483890014613208, + "learning_rate": 1.3732446969984341e-06, + "loss": 0.2392, + "step": 34780 + }, + { + "epoch": 0.7655954215276249, + "grad_norm": 1.4061226793712618, + "learning_rate": 1.3729993290795752e-06, + "loss": 0.3182, + "step": 34781 + }, + { + "epoch": 0.7656174334140435, + "grad_norm": 1.803413434072057, + "learning_rate": 1.3727539795947504e-06, + "loss": 0.2889, + "step": 34782 + }, + { + "epoch": 0.7656394453004622, + "grad_norm": 1.6328673299909835, + "learning_rate": 1.3725086485452065e-06, + "loss": 0.3251, + "step": 34783 + }, + { + "epoch": 0.765661457186881, + "grad_norm": 1.2910183806087885, + "learning_rate": 1.3722633359321935e-06, + "loss": 0.2616, + "step": 34784 + }, + { + "epoch": 0.7656834690732995, + "grad_norm": 1.6499971350402278, + "learning_rate": 1.372018041756954e-06, + "loss": 0.2235, + "step": 34785 + }, + { + "epoch": 0.7657054809597182, + "grad_norm": 2.2793369938552526, + "learning_rate": 1.3717727660207375e-06, + "loss": 0.2756, + "step": 34786 + }, + { + "epoch": 0.765727492846137, + "grad_norm": 1.5120481018050684, + "learning_rate": 1.3715275087247914e-06, + "loss": 0.3483, + "step": 34787 + }, + { + "epoch": 0.7657495047325555, + "grad_norm": 1.3584013072921994, + "learning_rate": 1.3712822698703592e-06, + "loss": 0.2327, + "step": 34788 + }, + { + "epoch": 0.7657715166189742, + "grad_norm": 1.7496368637135309, + "learning_rate": 1.3710370494586905e-06, + "loss": 0.3178, + "step": 34789 + }, + { + "epoch": 0.765793528505393, + "grad_norm": 1.9731178413401063, + "learning_rate": 1.3707918474910286e-06, + "loss": 0.3671, + "step": 34790 + }, + { + "epoch": 0.7658155403918115, + "grad_norm": 1.493618020709999, + "learning_rate": 1.370546663968621e-06, + "loss": 0.2722, + "step": 34791 + }, + { + "epoch": 0.7658375522782302, + "grad_norm": 1.4821366295023393, + "learning_rate": 1.370301498892716e-06, + "loss": 0.2859, + "step": 34792 + }, + { + "epoch": 0.765859564164649, + "grad_norm": 3.191067352297813, + "learning_rate": 1.370056352264556e-06, + "loss": 0.3248, + "step": 34793 + }, + { + "epoch": 0.7658815760510675, + "grad_norm": 1.4378341347881383, + "learning_rate": 1.36981122408539e-06, + "loss": 0.2484, + "step": 34794 + }, + { + "epoch": 0.7659035879374863, + "grad_norm": 1.3878632093874645, + "learning_rate": 1.369566114356461e-06, + "loss": 0.3295, + "step": 34795 + }, + { + "epoch": 0.765925599823905, + "grad_norm": 1.502665329265103, + "learning_rate": 1.369321023079016e-06, + "loss": 0.3228, + "step": 34796 + }, + { + "epoch": 0.7659476117103236, + "grad_norm": 4.8121252408757265, + "learning_rate": 1.3690759502543027e-06, + "loss": 0.2603, + "step": 34797 + }, + { + "epoch": 0.7659696235967423, + "grad_norm": 1.0946035608327531, + "learning_rate": 1.3688308958835632e-06, + "loss": 0.198, + "step": 34798 + }, + { + "epoch": 0.765991635483161, + "grad_norm": 0.624637108372289, + "learning_rate": 1.3685858599680462e-06, + "loss": 0.4429, + "step": 34799 + }, + { + "epoch": 0.7660136473695796, + "grad_norm": 1.9456064407424174, + "learning_rate": 1.368340842508994e-06, + "loss": 0.3628, + "step": 34800 + }, + { + "epoch": 0.7660356592559983, + "grad_norm": 2.063210929829969, + "learning_rate": 1.3680958435076537e-06, + "loss": 0.3416, + "step": 34801 + }, + { + "epoch": 0.7660576711424169, + "grad_norm": 1.4509896112721195, + "learning_rate": 1.3678508629652715e-06, + "loss": 0.3812, + "step": 34802 + }, + { + "epoch": 0.7660796830288356, + "grad_norm": 2.2947750407854053, + "learning_rate": 1.3676059008830895e-06, + "loss": 0.2754, + "step": 34803 + }, + { + "epoch": 0.7661016949152543, + "grad_norm": 3.1568281982488284, + "learning_rate": 1.367360957262356e-06, + "loss": 0.3429, + "step": 34804 + }, + { + "epoch": 0.7661237068016729, + "grad_norm": 1.5704509907137707, + "learning_rate": 1.3671160321043125e-06, + "loss": 0.2473, + "step": 34805 + }, + { + "epoch": 0.7661457186880916, + "grad_norm": 4.153837277325816, + "learning_rate": 1.3668711254102064e-06, + "loss": 0.2972, + "step": 34806 + }, + { + "epoch": 0.7661677305745103, + "grad_norm": 1.2682817257619712, + "learning_rate": 1.3666262371812828e-06, + "loss": 0.2627, + "step": 34807 + }, + { + "epoch": 0.7661897424609289, + "grad_norm": 1.808507505840195, + "learning_rate": 1.3663813674187848e-06, + "loss": 0.2619, + "step": 34808 + }, + { + "epoch": 0.7662117543473476, + "grad_norm": 1.6363236540095871, + "learning_rate": 1.3661365161239565e-06, + "loss": 0.2817, + "step": 34809 + }, + { + "epoch": 0.7662337662337663, + "grad_norm": 1.8063430024839506, + "learning_rate": 1.3658916832980429e-06, + "loss": 0.2663, + "step": 34810 + }, + { + "epoch": 0.7662557781201849, + "grad_norm": 2.109689015005071, + "learning_rate": 1.3656468689422887e-06, + "loss": 0.281, + "step": 34811 + }, + { + "epoch": 0.7662777900066036, + "grad_norm": 1.8816560787086352, + "learning_rate": 1.3654020730579393e-06, + "loss": 0.3431, + "step": 34812 + }, + { + "epoch": 0.7662998018930223, + "grad_norm": 1.1597145748294928, + "learning_rate": 1.3651572956462377e-06, + "loss": 0.2798, + "step": 34813 + }, + { + "epoch": 0.7663218137794409, + "grad_norm": 1.6173620461495304, + "learning_rate": 1.3649125367084266e-06, + "loss": 0.2749, + "step": 34814 + }, + { + "epoch": 0.7663438256658596, + "grad_norm": 0.6555509511681928, + "learning_rate": 1.3646677962457527e-06, + "loss": 0.4561, + "step": 34815 + }, + { + "epoch": 0.7663658375522783, + "grad_norm": 1.8444866734281955, + "learning_rate": 1.364423074259455e-06, + "loss": 0.2749, + "step": 34816 + }, + { + "epoch": 0.7663878494386969, + "grad_norm": 2.4954359510006414, + "learning_rate": 1.3641783707507845e-06, + "loss": 0.3077, + "step": 34817 + }, + { + "epoch": 0.7664098613251156, + "grad_norm": 1.305357920575019, + "learning_rate": 1.3639336857209806e-06, + "loss": 0.2704, + "step": 34818 + }, + { + "epoch": 0.7664318732115343, + "grad_norm": 1.5242437491300378, + "learning_rate": 1.3636890191712864e-06, + "loss": 0.2588, + "step": 34819 + }, + { + "epoch": 0.7664538850979529, + "grad_norm": 1.8001477529920435, + "learning_rate": 1.3634443711029476e-06, + "loss": 0.2476, + "step": 34820 + }, + { + "epoch": 0.7664758969843716, + "grad_norm": 1.4560067086835236, + "learning_rate": 1.3631997415172038e-06, + "loss": 0.2738, + "step": 34821 + }, + { + "epoch": 0.7664979088707903, + "grad_norm": 1.7958200649775637, + "learning_rate": 1.3629551304153032e-06, + "loss": 0.2989, + "step": 34822 + }, + { + "epoch": 0.7665199207572089, + "grad_norm": 1.9688101056796317, + "learning_rate": 1.362710537798488e-06, + "loss": 0.2553, + "step": 34823 + }, + { + "epoch": 0.7665419326436276, + "grad_norm": 1.5425234962160521, + "learning_rate": 1.3624659636679977e-06, + "loss": 0.3031, + "step": 34824 + }, + { + "epoch": 0.7665639445300462, + "grad_norm": 1.5188095723814257, + "learning_rate": 1.3622214080250796e-06, + "loss": 0.3722, + "step": 34825 + }, + { + "epoch": 0.7665859564164649, + "grad_norm": 1.518442223194321, + "learning_rate": 1.361976870870973e-06, + "loss": 0.3068, + "step": 34826 + }, + { + "epoch": 0.7666079683028836, + "grad_norm": 1.5448189074407497, + "learning_rate": 1.3617323522069226e-06, + "loss": 0.2801, + "step": 34827 + }, + { + "epoch": 0.7666299801893022, + "grad_norm": 2.048942200449239, + "learning_rate": 1.361487852034173e-06, + "loss": 0.2506, + "step": 34828 + }, + { + "epoch": 0.7666519920757209, + "grad_norm": 2.332212479290288, + "learning_rate": 1.361243370353963e-06, + "loss": 0.4013, + "step": 34829 + }, + { + "epoch": 0.7666740039621396, + "grad_norm": 1.4830106243753325, + "learning_rate": 1.3609989071675388e-06, + "loss": 0.2299, + "step": 34830 + }, + { + "epoch": 0.7666960158485582, + "grad_norm": 2.2631447482392844, + "learning_rate": 1.3607544624761398e-06, + "loss": 0.2839, + "step": 34831 + }, + { + "epoch": 0.7667180277349769, + "grad_norm": 1.5863907159825603, + "learning_rate": 1.3605100362810097e-06, + "loss": 0.2721, + "step": 34832 + }, + { + "epoch": 0.7667400396213956, + "grad_norm": 1.80127801654202, + "learning_rate": 1.3602656285833926e-06, + "loss": 0.2641, + "step": 34833 + }, + { + "epoch": 0.7667620515078142, + "grad_norm": 2.36993065920504, + "learning_rate": 1.3600212393845269e-06, + "loss": 0.3989, + "step": 34834 + }, + { + "epoch": 0.7667840633942329, + "grad_norm": 2.3695046694730415, + "learning_rate": 1.3597768686856588e-06, + "loss": 0.3565, + "step": 34835 + }, + { + "epoch": 0.7668060752806516, + "grad_norm": 1.546857657446213, + "learning_rate": 1.3595325164880264e-06, + "loss": 0.2741, + "step": 34836 + }, + { + "epoch": 0.7668280871670702, + "grad_norm": 2.022248646174339, + "learning_rate": 1.3592881827928734e-06, + "loss": 0.202, + "step": 34837 + }, + { + "epoch": 0.7668500990534889, + "grad_norm": 1.664467912873982, + "learning_rate": 1.3590438676014434e-06, + "loss": 0.252, + "step": 34838 + }, + { + "epoch": 0.7668721109399076, + "grad_norm": 1.7965778229921308, + "learning_rate": 1.3587995709149749e-06, + "loss": 0.2635, + "step": 34839 + }, + { + "epoch": 0.7668941228263262, + "grad_norm": 2.2002989731987155, + "learning_rate": 1.3585552927347124e-06, + "loss": 0.3032, + "step": 34840 + }, + { + "epoch": 0.7669161347127449, + "grad_norm": 1.6447379300406915, + "learning_rate": 1.3583110330618943e-06, + "loss": 0.2752, + "step": 34841 + }, + { + "epoch": 0.7669381465991636, + "grad_norm": 1.9331351683168796, + "learning_rate": 1.3580667918977642e-06, + "loss": 0.2992, + "step": 34842 + }, + { + "epoch": 0.7669601584855822, + "grad_norm": 0.6127071026650472, + "learning_rate": 1.3578225692435643e-06, + "loss": 0.4461, + "step": 34843 + }, + { + "epoch": 0.7669821703720009, + "grad_norm": 2.184066349313767, + "learning_rate": 1.3575783651005342e-06, + "loss": 0.378, + "step": 34844 + }, + { + "epoch": 0.7670041822584196, + "grad_norm": 0.6254647814539908, + "learning_rate": 1.3573341794699147e-06, + "loss": 0.4436, + "step": 34845 + }, + { + "epoch": 0.7670261941448382, + "grad_norm": 3.1196371008695745, + "learning_rate": 1.3570900123529469e-06, + "loss": 0.2814, + "step": 34846 + }, + { + "epoch": 0.7670482060312569, + "grad_norm": 1.621970742438429, + "learning_rate": 1.3568458637508724e-06, + "loss": 0.312, + "step": 34847 + }, + { + "epoch": 0.7670702179176755, + "grad_norm": 6.429818300861396, + "learning_rate": 1.3566017336649339e-06, + "loss": 0.3094, + "step": 34848 + }, + { + "epoch": 0.7670922298040942, + "grad_norm": 1.7348245229296801, + "learning_rate": 1.3563576220963704e-06, + "loss": 0.3265, + "step": 34849 + }, + { + "epoch": 0.7671142416905129, + "grad_norm": 0.6427715369426976, + "learning_rate": 1.356113529046421e-06, + "loss": 0.4752, + "step": 34850 + }, + { + "epoch": 0.7671362535769315, + "grad_norm": 1.4600389794733843, + "learning_rate": 1.3558694545163299e-06, + "loss": 0.2849, + "step": 34851 + }, + { + "epoch": 0.7671582654633502, + "grad_norm": 3.3065092808897525, + "learning_rate": 1.3556253985073314e-06, + "loss": 0.2735, + "step": 34852 + }, + { + "epoch": 0.7671802773497689, + "grad_norm": 1.4334416858392998, + "learning_rate": 1.355381361020674e-06, + "loss": 0.2809, + "step": 34853 + }, + { + "epoch": 0.7672022892361875, + "grad_norm": 1.5459778001725648, + "learning_rate": 1.3551373420575937e-06, + "loss": 0.3147, + "step": 34854 + }, + { + "epoch": 0.7672243011226062, + "grad_norm": 0.6028501084255978, + "learning_rate": 1.35489334161933e-06, + "loss": 0.452, + "step": 34855 + }, + { + "epoch": 0.7672463130090249, + "grad_norm": 1.3728958458180314, + "learning_rate": 1.3546493597071253e-06, + "loss": 0.3058, + "step": 34856 + }, + { + "epoch": 0.7672683248954435, + "grad_norm": 0.6487434685428369, + "learning_rate": 1.3544053963222153e-06, + "loss": 0.4592, + "step": 34857 + }, + { + "epoch": 0.7672903367818622, + "grad_norm": 1.8353840464166484, + "learning_rate": 1.3541614514658458e-06, + "loss": 0.3707, + "step": 34858 + }, + { + "epoch": 0.7673123486682809, + "grad_norm": 1.655846080503188, + "learning_rate": 1.3539175251392546e-06, + "loss": 0.2977, + "step": 34859 + }, + { + "epoch": 0.7673343605546995, + "grad_norm": 3.406098942977063, + "learning_rate": 1.3536736173436787e-06, + "loss": 0.31, + "step": 34860 + }, + { + "epoch": 0.7673563724411182, + "grad_norm": 4.078853315768016, + "learning_rate": 1.3534297280803615e-06, + "loss": 0.3417, + "step": 34861 + }, + { + "epoch": 0.7673783843275369, + "grad_norm": 2.9285491870405607, + "learning_rate": 1.3531858573505391e-06, + "loss": 0.3134, + "step": 34862 + }, + { + "epoch": 0.7674003962139555, + "grad_norm": 1.3838626355223858, + "learning_rate": 1.3529420051554532e-06, + "loss": 0.3021, + "step": 34863 + }, + { + "epoch": 0.7674224081003742, + "grad_norm": 1.65190104421617, + "learning_rate": 1.352698171496344e-06, + "loss": 0.2003, + "step": 34864 + }, + { + "epoch": 0.7674444199867929, + "grad_norm": 1.9514290289970357, + "learning_rate": 1.3524543563744474e-06, + "loss": 0.3731, + "step": 34865 + }, + { + "epoch": 0.7674664318732115, + "grad_norm": 1.9221239956940737, + "learning_rate": 1.3522105597910068e-06, + "loss": 0.3625, + "step": 34866 + }, + { + "epoch": 0.7674884437596302, + "grad_norm": 0.6177260468252913, + "learning_rate": 1.351966781747257e-06, + "loss": 0.4656, + "step": 34867 + }, + { + "epoch": 0.7675104556460489, + "grad_norm": 1.4560960480435108, + "learning_rate": 1.3517230222444395e-06, + "loss": 0.3407, + "step": 34868 + }, + { + "epoch": 0.7675324675324675, + "grad_norm": 1.8338302649231526, + "learning_rate": 1.351479281283794e-06, + "loss": 0.3683, + "step": 34869 + }, + { + "epoch": 0.7675544794188862, + "grad_norm": 1.8846818315799483, + "learning_rate": 1.3512355588665572e-06, + "loss": 0.2489, + "step": 34870 + }, + { + "epoch": 0.7675764913053049, + "grad_norm": 1.963195541748299, + "learning_rate": 1.3509918549939694e-06, + "loss": 0.333, + "step": 34871 + }, + { + "epoch": 0.7675985031917235, + "grad_norm": 3.441297994094991, + "learning_rate": 1.3507481696672675e-06, + "loss": 0.3311, + "step": 34872 + }, + { + "epoch": 0.7676205150781422, + "grad_norm": 1.6016040353093468, + "learning_rate": 1.3505045028876906e-06, + "loss": 0.2399, + "step": 34873 + }, + { + "epoch": 0.7676425269645608, + "grad_norm": 1.5672822985269184, + "learning_rate": 1.3502608546564794e-06, + "loss": 0.2063, + "step": 34874 + }, + { + "epoch": 0.7676645388509795, + "grad_norm": 1.4523947773678678, + "learning_rate": 1.3500172249748689e-06, + "loss": 0.3287, + "step": 34875 + }, + { + "epoch": 0.7676865507373982, + "grad_norm": 3.8479966815521993, + "learning_rate": 1.3497736138441004e-06, + "loss": 0.3732, + "step": 34876 + }, + { + "epoch": 0.7677085626238168, + "grad_norm": 2.5258419643458536, + "learning_rate": 1.3495300212654084e-06, + "loss": 0.2584, + "step": 34877 + }, + { + "epoch": 0.7677305745102355, + "grad_norm": 1.5572986380431484, + "learning_rate": 1.349286447240033e-06, + "loss": 0.3345, + "step": 34878 + }, + { + "epoch": 0.7677525863966542, + "grad_norm": 1.3597529986875163, + "learning_rate": 1.3490428917692139e-06, + "loss": 0.2979, + "step": 34879 + }, + { + "epoch": 0.7677745982830728, + "grad_norm": 1.266282075943151, + "learning_rate": 1.348799354854186e-06, + "loss": 0.2614, + "step": 34880 + }, + { + "epoch": 0.7677966101694915, + "grad_norm": 1.7771434167767746, + "learning_rate": 1.3485558364961893e-06, + "loss": 0.2175, + "step": 34881 + }, + { + "epoch": 0.7678186220559102, + "grad_norm": 1.9750130468089062, + "learning_rate": 1.348312336696459e-06, + "loss": 0.35, + "step": 34882 + }, + { + "epoch": 0.7678406339423288, + "grad_norm": 1.271255465229348, + "learning_rate": 1.3480688554562342e-06, + "loss": 0.2548, + "step": 34883 + }, + { + "epoch": 0.7678626458287475, + "grad_norm": 1.5751866062404982, + "learning_rate": 1.3478253927767538e-06, + "loss": 0.2227, + "step": 34884 + }, + { + "epoch": 0.7678846577151662, + "grad_norm": 0.6295651334444382, + "learning_rate": 1.347581948659254e-06, + "loss": 0.4535, + "step": 34885 + }, + { + "epoch": 0.7679066696015848, + "grad_norm": 1.4833400304151236, + "learning_rate": 1.3473385231049695e-06, + "loss": 0.2466, + "step": 34886 + }, + { + "epoch": 0.7679286814880035, + "grad_norm": 3.250634940739265, + "learning_rate": 1.347095116115142e-06, + "loss": 0.3608, + "step": 34887 + }, + { + "epoch": 0.7679506933744222, + "grad_norm": 1.9270173055135422, + "learning_rate": 1.3468517276910032e-06, + "loss": 0.2903, + "step": 34888 + }, + { + "epoch": 0.7679727052608408, + "grad_norm": 1.8766755362896272, + "learning_rate": 1.3466083578337958e-06, + "loss": 0.2628, + "step": 34889 + }, + { + "epoch": 0.7679947171472595, + "grad_norm": 1.5079881668769646, + "learning_rate": 1.3463650065447547e-06, + "loss": 0.2441, + "step": 34890 + }, + { + "epoch": 0.7680167290336782, + "grad_norm": 1.5337151047662239, + "learning_rate": 1.3461216738251143e-06, + "loss": 0.2695, + "step": 34891 + }, + { + "epoch": 0.7680387409200968, + "grad_norm": 1.8319382584092159, + "learning_rate": 1.3458783596761149e-06, + "loss": 0.233, + "step": 34892 + }, + { + "epoch": 0.7680607528065155, + "grad_norm": 1.7937938266469409, + "learning_rate": 1.3456350640989902e-06, + "loss": 0.3276, + "step": 34893 + }, + { + "epoch": 0.7680827646929342, + "grad_norm": 1.7341423231093485, + "learning_rate": 1.3453917870949774e-06, + "loss": 0.3268, + "step": 34894 + }, + { + "epoch": 0.7681047765793528, + "grad_norm": 1.7324990890577403, + "learning_rate": 1.3451485286653155e-06, + "loss": 0.2885, + "step": 34895 + }, + { + "epoch": 0.7681267884657715, + "grad_norm": 1.8691876171665402, + "learning_rate": 1.3449052888112375e-06, + "loss": 0.3902, + "step": 34896 + }, + { + "epoch": 0.7681488003521901, + "grad_norm": 3.5946482637019628, + "learning_rate": 1.3446620675339822e-06, + "loss": 0.3717, + "step": 34897 + }, + { + "epoch": 0.7681708122386088, + "grad_norm": 2.157811542028716, + "learning_rate": 1.3444188648347834e-06, + "loss": 0.2182, + "step": 34898 + }, + { + "epoch": 0.7681928241250275, + "grad_norm": 0.6434100269136606, + "learning_rate": 1.344175680714878e-06, + "loss": 0.46, + "step": 34899 + }, + { + "epoch": 0.7682148360114461, + "grad_norm": 2.3158792120664122, + "learning_rate": 1.343932515175504e-06, + "loss": 0.358, + "step": 34900 + }, + { + "epoch": 0.7682368478978648, + "grad_norm": 1.4727142636729258, + "learning_rate": 1.3436893682178943e-06, + "loss": 0.2539, + "step": 34901 + }, + { + "epoch": 0.7682588597842835, + "grad_norm": 1.7828315296677855, + "learning_rate": 1.3434462398432874e-06, + "loss": 0.3172, + "step": 34902 + }, + { + "epoch": 0.7682808716707021, + "grad_norm": 1.333590798544579, + "learning_rate": 1.3432031300529159e-06, + "loss": 0.2682, + "step": 34903 + }, + { + "epoch": 0.7683028835571208, + "grad_norm": 1.1357836381437887, + "learning_rate": 1.3429600388480174e-06, + "loss": 0.2788, + "step": 34904 + }, + { + "epoch": 0.7683248954435395, + "grad_norm": 1.6555693425555589, + "learning_rate": 1.3427169662298278e-06, + "loss": 0.2737, + "step": 34905 + }, + { + "epoch": 0.7683469073299581, + "grad_norm": 1.4768406077457266, + "learning_rate": 1.3424739121995812e-06, + "loss": 0.2642, + "step": 34906 + }, + { + "epoch": 0.7683689192163768, + "grad_norm": 0.6239701295961313, + "learning_rate": 1.3422308767585146e-06, + "loss": 0.4453, + "step": 34907 + }, + { + "epoch": 0.7683909311027955, + "grad_norm": 1.6975476245243786, + "learning_rate": 1.3419878599078606e-06, + "loss": 0.3203, + "step": 34908 + }, + { + "epoch": 0.7684129429892141, + "grad_norm": 2.560221937243874, + "learning_rate": 1.341744861648856e-06, + "loss": 0.2236, + "step": 34909 + }, + { + "epoch": 0.7684349548756328, + "grad_norm": 1.9487097331991927, + "learning_rate": 1.3415018819827369e-06, + "loss": 0.2337, + "step": 34910 + }, + { + "epoch": 0.7684569667620516, + "grad_norm": 2.570749080664802, + "learning_rate": 1.3412589209107352e-06, + "loss": 0.2908, + "step": 34911 + }, + { + "epoch": 0.7684789786484701, + "grad_norm": 1.4309626731109035, + "learning_rate": 1.3410159784340892e-06, + "loss": 0.2981, + "step": 34912 + }, + { + "epoch": 0.7685009905348889, + "grad_norm": 1.4270588564682132, + "learning_rate": 1.340773054554031e-06, + "loss": 0.3605, + "step": 34913 + }, + { + "epoch": 0.7685230024213076, + "grad_norm": 1.7907450015380257, + "learning_rate": 1.3405301492717954e-06, + "loss": 0.2292, + "step": 34914 + }, + { + "epoch": 0.7685450143077261, + "grad_norm": 0.6191487891325985, + "learning_rate": 1.3402872625886198e-06, + "loss": 0.43, + "step": 34915 + }, + { + "epoch": 0.7685670261941449, + "grad_norm": 1.4386615354414112, + "learning_rate": 1.3400443945057352e-06, + "loss": 0.2926, + "step": 34916 + }, + { + "epoch": 0.7685890380805636, + "grad_norm": 2.1351919417958793, + "learning_rate": 1.3398015450243785e-06, + "loss": 0.236, + "step": 34917 + }, + { + "epoch": 0.7686110499669822, + "grad_norm": 2.9637835492970264, + "learning_rate": 1.3395587141457812e-06, + "loss": 0.3506, + "step": 34918 + }, + { + "epoch": 0.7686330618534009, + "grad_norm": 1.4848965304596005, + "learning_rate": 1.3393159018711793e-06, + "loss": 0.3002, + "step": 34919 + }, + { + "epoch": 0.7686550737398194, + "grad_norm": 1.4976511446467111, + "learning_rate": 1.3390731082018082e-06, + "loss": 0.3147, + "step": 34920 + }, + { + "epoch": 0.7686770856262382, + "grad_norm": 1.8084424388651485, + "learning_rate": 1.338830333138899e-06, + "loss": 0.2734, + "step": 34921 + }, + { + "epoch": 0.7686990975126569, + "grad_norm": 1.7120532207641803, + "learning_rate": 1.3385875766836886e-06, + "loss": 0.2805, + "step": 34922 + }, + { + "epoch": 0.7687211093990755, + "grad_norm": 1.7701824143395473, + "learning_rate": 1.3383448388374094e-06, + "loss": 0.3138, + "step": 34923 + }, + { + "epoch": 0.7687431212854942, + "grad_norm": 0.6353136185403337, + "learning_rate": 1.338102119601291e-06, + "loss": 0.4691, + "step": 34924 + }, + { + "epoch": 0.7687651331719129, + "grad_norm": 1.801705716075436, + "learning_rate": 1.3378594189765742e-06, + "loss": 0.3177, + "step": 34925 + }, + { + "epoch": 0.7687871450583315, + "grad_norm": 1.8303382999610798, + "learning_rate": 1.3376167369644894e-06, + "loss": 0.296, + "step": 34926 + }, + { + "epoch": 0.7688091569447502, + "grad_norm": 1.460038354785944, + "learning_rate": 1.3373740735662682e-06, + "loss": 0.2434, + "step": 34927 + }, + { + "epoch": 0.7688311688311689, + "grad_norm": 1.9950741503841238, + "learning_rate": 1.3371314287831471e-06, + "loss": 0.3049, + "step": 34928 + }, + { + "epoch": 0.7688531807175875, + "grad_norm": 1.3084355442947118, + "learning_rate": 1.3368888026163568e-06, + "loss": 0.2622, + "step": 34929 + }, + { + "epoch": 0.7688751926040062, + "grad_norm": 3.1759931100842547, + "learning_rate": 1.3366461950671305e-06, + "loss": 0.253, + "step": 34930 + }, + { + "epoch": 0.7688972044904249, + "grad_norm": 1.4053072131164035, + "learning_rate": 1.3364036061367041e-06, + "loss": 0.2838, + "step": 34931 + }, + { + "epoch": 0.7689192163768435, + "grad_norm": 0.5940885609491815, + "learning_rate": 1.3361610358263071e-06, + "loss": 0.4295, + "step": 34932 + }, + { + "epoch": 0.7689412282632622, + "grad_norm": 0.6227621059575897, + "learning_rate": 1.3359184841371752e-06, + "loss": 0.469, + "step": 34933 + }, + { + "epoch": 0.7689632401496809, + "grad_norm": 1.3545759444663357, + "learning_rate": 1.335675951070538e-06, + "loss": 0.3485, + "step": 34934 + }, + { + "epoch": 0.7689852520360995, + "grad_norm": 1.3224685500976163, + "learning_rate": 1.3354334366276306e-06, + "loss": 0.2501, + "step": 34935 + }, + { + "epoch": 0.7690072639225182, + "grad_norm": 1.2179510695956313, + "learning_rate": 1.3351909408096859e-06, + "loss": 0.316, + "step": 34936 + }, + { + "epoch": 0.7690292758089369, + "grad_norm": 4.110602099858704, + "learning_rate": 1.3349484636179339e-06, + "loss": 0.2852, + "step": 34937 + }, + { + "epoch": 0.7690512876953555, + "grad_norm": 0.6103517582391221, + "learning_rate": 1.3347060050536104e-06, + "loss": 0.4305, + "step": 34938 + }, + { + "epoch": 0.7690732995817742, + "grad_norm": 1.4210936746819194, + "learning_rate": 1.3344635651179438e-06, + "loss": 0.3088, + "step": 34939 + }, + { + "epoch": 0.7690953114681929, + "grad_norm": 1.4706487756248303, + "learning_rate": 1.3342211438121682e-06, + "loss": 0.2907, + "step": 34940 + }, + { + "epoch": 0.7691173233546115, + "grad_norm": 1.8504980434646452, + "learning_rate": 1.3339787411375176e-06, + "loss": 0.2874, + "step": 34941 + }, + { + "epoch": 0.7691393352410302, + "grad_norm": 2.2227573546598447, + "learning_rate": 1.3337363570952206e-06, + "loss": 0.192, + "step": 34942 + }, + { + "epoch": 0.7691613471274488, + "grad_norm": 1.7004325679462173, + "learning_rate": 1.3334939916865119e-06, + "loss": 0.2637, + "step": 34943 + }, + { + "epoch": 0.7691833590138675, + "grad_norm": 1.7831489836529304, + "learning_rate": 1.3332516449126203e-06, + "loss": 0.3031, + "step": 34944 + }, + { + "epoch": 0.7692053709002862, + "grad_norm": 1.897612866977709, + "learning_rate": 1.3330093167747799e-06, + "loss": 0.23, + "step": 34945 + }, + { + "epoch": 0.7692273827867048, + "grad_norm": 1.8009062085653043, + "learning_rate": 1.3327670072742227e-06, + "loss": 0.3744, + "step": 34946 + }, + { + "epoch": 0.7692493946731235, + "grad_norm": 1.5359447366015173, + "learning_rate": 1.332524716412178e-06, + "loss": 0.2712, + "step": 34947 + }, + { + "epoch": 0.7692714065595422, + "grad_norm": 1.8344856965294873, + "learning_rate": 1.3322824441898796e-06, + "loss": 0.3234, + "step": 34948 + }, + { + "epoch": 0.7692934184459608, + "grad_norm": 1.4440844759264, + "learning_rate": 1.3320401906085568e-06, + "loss": 0.2618, + "step": 34949 + }, + { + "epoch": 0.7693154303323795, + "grad_norm": 1.7844388228359458, + "learning_rate": 1.331797955669441e-06, + "loss": 0.2956, + "step": 34950 + }, + { + "epoch": 0.7693374422187982, + "grad_norm": 2.0388561605838524, + "learning_rate": 1.331555739373766e-06, + "loss": 0.2823, + "step": 34951 + }, + { + "epoch": 0.7693594541052168, + "grad_norm": 1.4566634097872126, + "learning_rate": 1.331313541722759e-06, + "loss": 0.2852, + "step": 34952 + }, + { + "epoch": 0.7693814659916355, + "grad_norm": 1.63226187737063, + "learning_rate": 1.3310713627176547e-06, + "loss": 0.2492, + "step": 34953 + }, + { + "epoch": 0.7694034778780542, + "grad_norm": 1.7920755613825134, + "learning_rate": 1.3308292023596808e-06, + "loss": 0.2443, + "step": 34954 + }, + { + "epoch": 0.7694254897644728, + "grad_norm": 1.9408508435899425, + "learning_rate": 1.3305870606500687e-06, + "loss": 0.397, + "step": 34955 + }, + { + "epoch": 0.7694475016508915, + "grad_norm": 1.6007789093092524, + "learning_rate": 1.3303449375900519e-06, + "loss": 0.2755, + "step": 34956 + }, + { + "epoch": 0.7694695135373102, + "grad_norm": 1.7319418096187444, + "learning_rate": 1.3301028331808568e-06, + "loss": 0.3802, + "step": 34957 + }, + { + "epoch": 0.7694915254237288, + "grad_norm": 0.6258924105522728, + "learning_rate": 1.329860747423718e-06, + "loss": 0.4566, + "step": 34958 + }, + { + "epoch": 0.7695135373101475, + "grad_norm": 0.6000894022310398, + "learning_rate": 1.3296186803198635e-06, + "loss": 0.4277, + "step": 34959 + }, + { + "epoch": 0.7695355491965662, + "grad_norm": 1.6233961819337408, + "learning_rate": 1.329376631870521e-06, + "loss": 0.3654, + "step": 34960 + }, + { + "epoch": 0.7695575610829848, + "grad_norm": 1.9124984663681077, + "learning_rate": 1.3291346020769263e-06, + "loss": 0.2424, + "step": 34961 + }, + { + "epoch": 0.7695795729694035, + "grad_norm": 2.138736601426069, + "learning_rate": 1.3288925909403056e-06, + "loss": 0.2083, + "step": 34962 + }, + { + "epoch": 0.7696015848558222, + "grad_norm": 2.5940447405837554, + "learning_rate": 1.328650598461892e-06, + "loss": 0.3691, + "step": 34963 + }, + { + "epoch": 0.7696235967422408, + "grad_norm": 0.5958049389717021, + "learning_rate": 1.3284086246429128e-06, + "loss": 0.4453, + "step": 34964 + }, + { + "epoch": 0.7696456086286595, + "grad_norm": 1.3162178148214212, + "learning_rate": 1.3281666694845974e-06, + "loss": 0.2977, + "step": 34965 + }, + { + "epoch": 0.7696676205150781, + "grad_norm": 1.5301034693050235, + "learning_rate": 1.3279247329881768e-06, + "loss": 0.3052, + "step": 34966 + }, + { + "epoch": 0.7696896324014968, + "grad_norm": 1.7418093731389965, + "learning_rate": 1.3276828151548815e-06, + "loss": 0.2769, + "step": 34967 + }, + { + "epoch": 0.7697116442879155, + "grad_norm": 1.8284127473412082, + "learning_rate": 1.3274409159859386e-06, + "loss": 0.2691, + "step": 34968 + }, + { + "epoch": 0.7697336561743341, + "grad_norm": 1.6926253248258485, + "learning_rate": 1.327199035482581e-06, + "loss": 0.2602, + "step": 34969 + }, + { + "epoch": 0.7697556680607528, + "grad_norm": 1.7209768832692343, + "learning_rate": 1.326957173646034e-06, + "loss": 0.2727, + "step": 34970 + }, + { + "epoch": 0.7697776799471715, + "grad_norm": 2.150057395320461, + "learning_rate": 1.3267153304775293e-06, + "loss": 0.2878, + "step": 34971 + }, + { + "epoch": 0.7697996918335901, + "grad_norm": 0.5970135972277736, + "learning_rate": 1.3264735059782969e-06, + "loss": 0.4502, + "step": 34972 + }, + { + "epoch": 0.7698217037200088, + "grad_norm": 1.6472282641751133, + "learning_rate": 1.3262317001495628e-06, + "loss": 0.3098, + "step": 34973 + }, + { + "epoch": 0.7698437156064275, + "grad_norm": 1.6790169876617869, + "learning_rate": 1.3259899129925597e-06, + "loss": 0.316, + "step": 34974 + }, + { + "epoch": 0.7698657274928461, + "grad_norm": 1.431620280516808, + "learning_rate": 1.325748144508513e-06, + "loss": 0.2717, + "step": 34975 + }, + { + "epoch": 0.7698877393792648, + "grad_norm": 1.4969579508693962, + "learning_rate": 1.3255063946986528e-06, + "loss": 0.2996, + "step": 34976 + }, + { + "epoch": 0.7699097512656835, + "grad_norm": 1.4096012797071436, + "learning_rate": 1.3252646635642098e-06, + "loss": 0.3136, + "step": 34977 + }, + { + "epoch": 0.7699317631521021, + "grad_norm": 2.446738238139349, + "learning_rate": 1.3250229511064089e-06, + "loss": 0.2477, + "step": 34978 + }, + { + "epoch": 0.7699537750385208, + "grad_norm": 1.4627954422378933, + "learning_rate": 1.3247812573264817e-06, + "loss": 0.335, + "step": 34979 + }, + { + "epoch": 0.7699757869249395, + "grad_norm": 2.202211668680273, + "learning_rate": 1.324539582225654e-06, + "loss": 0.2836, + "step": 34980 + }, + { + "epoch": 0.7699977988113581, + "grad_norm": 1.692915412793102, + "learning_rate": 1.3242979258051558e-06, + "loss": 0.2997, + "step": 34981 + }, + { + "epoch": 0.7700198106977768, + "grad_norm": 1.341257319633429, + "learning_rate": 1.3240562880662162e-06, + "loss": 0.3083, + "step": 34982 + }, + { + "epoch": 0.7700418225841955, + "grad_norm": 2.3312887585990563, + "learning_rate": 1.3238146690100605e-06, + "loss": 0.4047, + "step": 34983 + }, + { + "epoch": 0.7700638344706141, + "grad_norm": 0.6177826265745492, + "learning_rate": 1.32357306863792e-06, + "loss": 0.4598, + "step": 34984 + }, + { + "epoch": 0.7700858463570328, + "grad_norm": 0.6199441924221665, + "learning_rate": 1.3233314869510193e-06, + "loss": 0.4227, + "step": 34985 + }, + { + "epoch": 0.7701078582434515, + "grad_norm": 1.4598230647899912, + "learning_rate": 1.3230899239505879e-06, + "loss": 0.294, + "step": 34986 + }, + { + "epoch": 0.7701298701298701, + "grad_norm": 0.6024759274601851, + "learning_rate": 1.3228483796378549e-06, + "loss": 0.4288, + "step": 34987 + }, + { + "epoch": 0.7701518820162888, + "grad_norm": 4.194194552254521, + "learning_rate": 1.3226068540140452e-06, + "loss": 0.2308, + "step": 34988 + }, + { + "epoch": 0.7701738939027074, + "grad_norm": 1.7391730589383307, + "learning_rate": 1.3223653470803887e-06, + "loss": 0.3575, + "step": 34989 + }, + { + "epoch": 0.7701959057891261, + "grad_norm": 1.509331191781378, + "learning_rate": 1.322123858838112e-06, + "loss": 0.2553, + "step": 34990 + }, + { + "epoch": 0.7702179176755448, + "grad_norm": 0.6043751050988619, + "learning_rate": 1.3218823892884392e-06, + "loss": 0.4524, + "step": 34991 + }, + { + "epoch": 0.7702399295619634, + "grad_norm": 2.4392144523606087, + "learning_rate": 1.3216409384326035e-06, + "loss": 0.2653, + "step": 34992 + }, + { + "epoch": 0.7702619414483821, + "grad_norm": 0.605050869465773, + "learning_rate": 1.3213995062718278e-06, + "loss": 0.457, + "step": 34993 + }, + { + "epoch": 0.7702839533348008, + "grad_norm": 0.609432576538169, + "learning_rate": 1.321158092807342e-06, + "loss": 0.4357, + "step": 34994 + }, + { + "epoch": 0.7703059652212194, + "grad_norm": 1.694023983466479, + "learning_rate": 1.3209166980403714e-06, + "loss": 0.3064, + "step": 34995 + }, + { + "epoch": 0.7703279771076381, + "grad_norm": 2.0681798757849035, + "learning_rate": 1.3206753219721402e-06, + "loss": 0.3302, + "step": 34996 + }, + { + "epoch": 0.7703499889940568, + "grad_norm": 1.5757448505748504, + "learning_rate": 1.320433964603881e-06, + "loss": 0.3303, + "step": 34997 + }, + { + "epoch": 0.7703720008804754, + "grad_norm": 1.8697535237519227, + "learning_rate": 1.3201926259368158e-06, + "loss": 0.2896, + "step": 34998 + }, + { + "epoch": 0.7703940127668941, + "grad_norm": 1.528340212711495, + "learning_rate": 1.3199513059721748e-06, + "loss": 0.2544, + "step": 34999 + }, + { + "epoch": 0.7704160246533128, + "grad_norm": 1.6522464644120518, + "learning_rate": 1.3197100047111828e-06, + "loss": 0.2808, + "step": 35000 + }, + { + "epoch": 0.7704380365397314, + "grad_norm": 2.3761130574377174, + "learning_rate": 1.3194687221550639e-06, + "loss": 0.329, + "step": 35001 + }, + { + "epoch": 0.7704600484261501, + "grad_norm": 4.307929267845835, + "learning_rate": 1.319227458305047e-06, + "loss": 0.3515, + "step": 35002 + }, + { + "epoch": 0.7704820603125688, + "grad_norm": 4.4211786018592525, + "learning_rate": 1.318986213162357e-06, + "loss": 0.295, + "step": 35003 + }, + { + "epoch": 0.7705040721989874, + "grad_norm": 1.3864777265374857, + "learning_rate": 1.3187449867282231e-06, + "loss": 0.3059, + "step": 35004 + }, + { + "epoch": 0.7705260840854061, + "grad_norm": 2.0705826659504534, + "learning_rate": 1.318503779003869e-06, + "loss": 0.2596, + "step": 35005 + }, + { + "epoch": 0.7705480959718248, + "grad_norm": 1.3624618403688307, + "learning_rate": 1.3182625899905193e-06, + "loss": 0.3129, + "step": 35006 + }, + { + "epoch": 0.7705701078582434, + "grad_norm": 1.5206022165223765, + "learning_rate": 1.318021419689401e-06, + "loss": 0.3466, + "step": 35007 + }, + { + "epoch": 0.7705921197446621, + "grad_norm": 0.6339821541119486, + "learning_rate": 1.3177802681017415e-06, + "loss": 0.4759, + "step": 35008 + }, + { + "epoch": 0.7706141316310808, + "grad_norm": 1.9369898649192192, + "learning_rate": 1.3175391352287636e-06, + "loss": 0.287, + "step": 35009 + }, + { + "epoch": 0.7706361435174994, + "grad_norm": 0.601314899351541, + "learning_rate": 1.3172980210716956e-06, + "loss": 0.4379, + "step": 35010 + }, + { + "epoch": 0.7706581554039181, + "grad_norm": 1.2623996784591698, + "learning_rate": 1.3170569256317606e-06, + "loss": 0.2612, + "step": 35011 + }, + { + "epoch": 0.7706801672903368, + "grad_norm": 2.4114422181317035, + "learning_rate": 1.3168158489101845e-06, + "loss": 0.2776, + "step": 35012 + }, + { + "epoch": 0.7707021791767554, + "grad_norm": 1.5211816514601209, + "learning_rate": 1.3165747909081945e-06, + "loss": 0.2283, + "step": 35013 + }, + { + "epoch": 0.7707241910631741, + "grad_norm": 1.650455799911984, + "learning_rate": 1.3163337516270125e-06, + "loss": 0.2929, + "step": 35014 + }, + { + "epoch": 0.7707462029495927, + "grad_norm": 1.7094559706147108, + "learning_rate": 1.3160927310678672e-06, + "loss": 0.2494, + "step": 35015 + }, + { + "epoch": 0.7707682148360114, + "grad_norm": 2.049747054620274, + "learning_rate": 1.31585172923198e-06, + "loss": 0.3866, + "step": 35016 + }, + { + "epoch": 0.7707902267224301, + "grad_norm": 0.6002260927135628, + "learning_rate": 1.315610746120578e-06, + "loss": 0.4586, + "step": 35017 + }, + { + "epoch": 0.7708122386088487, + "grad_norm": 0.6296972624263277, + "learning_rate": 1.3153697817348865e-06, + "loss": 0.4497, + "step": 35018 + }, + { + "epoch": 0.7708342504952674, + "grad_norm": 1.7820862004377316, + "learning_rate": 1.3151288360761278e-06, + "loss": 0.322, + "step": 35019 + }, + { + "epoch": 0.7708562623816861, + "grad_norm": 1.5812358929428225, + "learning_rate": 1.3148879091455291e-06, + "loss": 0.3817, + "step": 35020 + }, + { + "epoch": 0.7708782742681047, + "grad_norm": 4.4251801253017815, + "learning_rate": 1.3146470009443125e-06, + "loss": 0.3133, + "step": 35021 + }, + { + "epoch": 0.7709002861545234, + "grad_norm": 0.6178924192561007, + "learning_rate": 1.3144061114737033e-06, + "loss": 0.4348, + "step": 35022 + }, + { + "epoch": 0.7709222980409421, + "grad_norm": 1.5336106988464246, + "learning_rate": 1.314165240734928e-06, + "loss": 0.3803, + "step": 35023 + }, + { + "epoch": 0.7709443099273607, + "grad_norm": 1.4984452218306439, + "learning_rate": 1.3139243887292068e-06, + "loss": 0.2451, + "step": 35024 + }, + { + "epoch": 0.7709663218137794, + "grad_norm": 0.6175591026492988, + "learning_rate": 1.3136835554577681e-06, + "loss": 0.456, + "step": 35025 + }, + { + "epoch": 0.7709883337001981, + "grad_norm": 0.6174165932031442, + "learning_rate": 1.3134427409218332e-06, + "loss": 0.4542, + "step": 35026 + }, + { + "epoch": 0.7710103455866167, + "grad_norm": 1.713927355406799, + "learning_rate": 1.3132019451226236e-06, + "loss": 0.2735, + "step": 35027 + }, + { + "epoch": 0.7710323574730354, + "grad_norm": 1.8723944333876223, + "learning_rate": 1.3129611680613692e-06, + "loss": 0.2989, + "step": 35028 + }, + { + "epoch": 0.7710543693594542, + "grad_norm": 2.164548231599317, + "learning_rate": 1.3127204097392886e-06, + "loss": 0.3176, + "step": 35029 + }, + { + "epoch": 0.7710763812458727, + "grad_norm": 1.339579631125977, + "learning_rate": 1.3124796701576098e-06, + "loss": 0.2225, + "step": 35030 + }, + { + "epoch": 0.7710983931322914, + "grad_norm": 1.1808169820911674, + "learning_rate": 1.312238949317553e-06, + "loss": 0.2127, + "step": 35031 + }, + { + "epoch": 0.7711204050187102, + "grad_norm": 0.6100851609246687, + "learning_rate": 1.3119982472203408e-06, + "loss": 0.443, + "step": 35032 + }, + { + "epoch": 0.7711424169051287, + "grad_norm": 1.7693156248755564, + "learning_rate": 1.3117575638672004e-06, + "loss": 0.3775, + "step": 35033 + }, + { + "epoch": 0.7711644287915475, + "grad_norm": 0.6382788573522103, + "learning_rate": 1.3115168992593518e-06, + "loss": 0.4566, + "step": 35034 + }, + { + "epoch": 0.7711864406779662, + "grad_norm": 1.7496637734570113, + "learning_rate": 1.3112762533980212e-06, + "loss": 0.2362, + "step": 35035 + }, + { + "epoch": 0.7712084525643847, + "grad_norm": 3.9991624491937827, + "learning_rate": 1.3110356262844297e-06, + "loss": 0.3337, + "step": 35036 + }, + { + "epoch": 0.7712304644508035, + "grad_norm": 1.5677292197536985, + "learning_rate": 1.3107950179197993e-06, + "loss": 0.3606, + "step": 35037 + }, + { + "epoch": 0.771252476337222, + "grad_norm": 0.6234958002171493, + "learning_rate": 1.310554428305354e-06, + "loss": 0.466, + "step": 35038 + }, + { + "epoch": 0.7712744882236408, + "grad_norm": 1.538465058931867, + "learning_rate": 1.3103138574423163e-06, + "loss": 0.3495, + "step": 35039 + }, + { + "epoch": 0.7712965001100595, + "grad_norm": 1.9042191067628569, + "learning_rate": 1.310073305331911e-06, + "loss": 0.3026, + "step": 35040 + }, + { + "epoch": 0.771318511996478, + "grad_norm": 1.6558941337004822, + "learning_rate": 1.3098327719753583e-06, + "loss": 0.3579, + "step": 35041 + }, + { + "epoch": 0.7713405238828968, + "grad_norm": 1.2656557738670944, + "learning_rate": 1.3095922573738807e-06, + "loss": 0.2694, + "step": 35042 + }, + { + "epoch": 0.7713625357693155, + "grad_norm": 1.3415669133895414, + "learning_rate": 1.3093517615287005e-06, + "loss": 0.2269, + "step": 35043 + }, + { + "epoch": 0.771384547655734, + "grad_norm": 0.6108993873543188, + "learning_rate": 1.3091112844410426e-06, + "loss": 0.428, + "step": 35044 + }, + { + "epoch": 0.7714065595421528, + "grad_norm": 0.6087349312739441, + "learning_rate": 1.3088708261121257e-06, + "loss": 0.4601, + "step": 35045 + }, + { + "epoch": 0.7714285714285715, + "grad_norm": 1.5711864203728716, + "learning_rate": 1.3086303865431755e-06, + "loss": 0.3458, + "step": 35046 + }, + { + "epoch": 0.7714505833149901, + "grad_norm": 4.7648457517989735, + "learning_rate": 1.3083899657354098e-06, + "loss": 0.3073, + "step": 35047 + }, + { + "epoch": 0.7714725952014088, + "grad_norm": 1.4546007989479426, + "learning_rate": 1.308149563690053e-06, + "loss": 0.2946, + "step": 35048 + }, + { + "epoch": 0.7714946070878275, + "grad_norm": 1.8045524030554205, + "learning_rate": 1.3079091804083288e-06, + "loss": 0.2251, + "step": 35049 + }, + { + "epoch": 0.7715166189742461, + "grad_norm": 1.5117659228934737, + "learning_rate": 1.3076688158914552e-06, + "loss": 0.293, + "step": 35050 + }, + { + "epoch": 0.7715386308606648, + "grad_norm": 1.5864227268938313, + "learning_rate": 1.3074284701406565e-06, + "loss": 0.2826, + "step": 35051 + }, + { + "epoch": 0.7715606427470835, + "grad_norm": 1.4115947852715505, + "learning_rate": 1.307188143157152e-06, + "loss": 0.2078, + "step": 35052 + }, + { + "epoch": 0.7715826546335021, + "grad_norm": 2.3404909119792427, + "learning_rate": 1.3069478349421643e-06, + "loss": 0.3478, + "step": 35053 + }, + { + "epoch": 0.7716046665199208, + "grad_norm": 2.2099553189623293, + "learning_rate": 1.3067075454969168e-06, + "loss": 0.3074, + "step": 35054 + }, + { + "epoch": 0.7716266784063395, + "grad_norm": 2.7623396278055723, + "learning_rate": 1.306467274822627e-06, + "loss": 0.3495, + "step": 35055 + }, + { + "epoch": 0.7716486902927581, + "grad_norm": 4.905201365073902, + "learning_rate": 1.3062270229205193e-06, + "loss": 0.3121, + "step": 35056 + }, + { + "epoch": 0.7716707021791768, + "grad_norm": 1.4255721123941734, + "learning_rate": 1.3059867897918116e-06, + "loss": 0.3336, + "step": 35057 + }, + { + "epoch": 0.7716927140655955, + "grad_norm": 4.4235427020261255, + "learning_rate": 1.3057465754377268e-06, + "loss": 0.2488, + "step": 35058 + }, + { + "epoch": 0.7717147259520141, + "grad_norm": 1.747877313585348, + "learning_rate": 1.3055063798594875e-06, + "loss": 0.3298, + "step": 35059 + }, + { + "epoch": 0.7717367378384328, + "grad_norm": 2.3713026264022994, + "learning_rate": 1.3052662030583107e-06, + "loss": 0.286, + "step": 35060 + }, + { + "epoch": 0.7717587497248514, + "grad_norm": 0.6110728675156443, + "learning_rate": 1.3050260450354207e-06, + "loss": 0.4429, + "step": 35061 + }, + { + "epoch": 0.7717807616112701, + "grad_norm": 0.6439031048771386, + "learning_rate": 1.3047859057920349e-06, + "loss": 0.4418, + "step": 35062 + }, + { + "epoch": 0.7718027734976888, + "grad_norm": 1.4620811854408522, + "learning_rate": 1.3045457853293746e-06, + "loss": 0.2308, + "step": 35063 + }, + { + "epoch": 0.7718247853841074, + "grad_norm": 1.6691581488026197, + "learning_rate": 1.3043056836486628e-06, + "loss": 0.344, + "step": 35064 + }, + { + "epoch": 0.7718467972705261, + "grad_norm": 1.556997965720627, + "learning_rate": 1.304065600751116e-06, + "loss": 0.2877, + "step": 35065 + }, + { + "epoch": 0.7718688091569448, + "grad_norm": 2.5039521917782666, + "learning_rate": 1.3038255366379581e-06, + "loss": 0.2625, + "step": 35066 + }, + { + "epoch": 0.7718908210433634, + "grad_norm": 2.483739768873125, + "learning_rate": 1.3035854913104074e-06, + "loss": 0.3556, + "step": 35067 + }, + { + "epoch": 0.7719128329297821, + "grad_norm": 0.6208338914759843, + "learning_rate": 1.3033454647696809e-06, + "loss": 0.4619, + "step": 35068 + }, + { + "epoch": 0.7719348448162008, + "grad_norm": 0.592335086105508, + "learning_rate": 1.3031054570170043e-06, + "loss": 0.4319, + "step": 35069 + }, + { + "epoch": 0.7719568567026194, + "grad_norm": 0.5933806650267849, + "learning_rate": 1.3028654680535934e-06, + "loss": 0.4628, + "step": 35070 + }, + { + "epoch": 0.7719788685890381, + "grad_norm": 1.4842264669409133, + "learning_rate": 1.3026254978806702e-06, + "loss": 0.2525, + "step": 35071 + }, + { + "epoch": 0.7720008804754568, + "grad_norm": 0.6165823227144761, + "learning_rate": 1.3023855464994533e-06, + "loss": 0.4602, + "step": 35072 + }, + { + "epoch": 0.7720228923618754, + "grad_norm": 1.6437084209869355, + "learning_rate": 1.3021456139111616e-06, + "loss": 0.2841, + "step": 35073 + }, + { + "epoch": 0.7720449042482941, + "grad_norm": 0.6376783674841242, + "learning_rate": 1.3019057001170143e-06, + "loss": 0.4437, + "step": 35074 + }, + { + "epoch": 0.7720669161347128, + "grad_norm": 1.3007771003630288, + "learning_rate": 1.3016658051182317e-06, + "loss": 0.2852, + "step": 35075 + }, + { + "epoch": 0.7720889280211314, + "grad_norm": 8.866945333208392, + "learning_rate": 1.3014259289160348e-06, + "loss": 0.2973, + "step": 35076 + }, + { + "epoch": 0.7721109399075501, + "grad_norm": 1.691072128040711, + "learning_rate": 1.3011860715116408e-06, + "loss": 0.2332, + "step": 35077 + }, + { + "epoch": 0.7721329517939688, + "grad_norm": 1.6392848908940354, + "learning_rate": 1.3009462329062671e-06, + "loss": 0.2675, + "step": 35078 + }, + { + "epoch": 0.7721549636803874, + "grad_norm": 1.5723826254875366, + "learning_rate": 1.3007064131011355e-06, + "loss": 0.2755, + "step": 35079 + }, + { + "epoch": 0.7721769755668061, + "grad_norm": 0.6051286194159746, + "learning_rate": 1.3004666120974629e-06, + "loss": 0.4458, + "step": 35080 + }, + { + "epoch": 0.7721989874532248, + "grad_norm": 0.6125166856966499, + "learning_rate": 1.3002268298964703e-06, + "loss": 0.4425, + "step": 35081 + }, + { + "epoch": 0.7722209993396434, + "grad_norm": 1.8518216633690145, + "learning_rate": 1.2999870664993758e-06, + "loss": 0.3818, + "step": 35082 + }, + { + "epoch": 0.7722430112260621, + "grad_norm": 1.5680648484211606, + "learning_rate": 1.2997473219073957e-06, + "loss": 0.2155, + "step": 35083 + }, + { + "epoch": 0.7722650231124807, + "grad_norm": 1.5909504771029337, + "learning_rate": 1.2995075961217496e-06, + "loss": 0.2566, + "step": 35084 + }, + { + "epoch": 0.7722870349988994, + "grad_norm": 1.465913324640832, + "learning_rate": 1.2992678891436584e-06, + "loss": 0.341, + "step": 35085 + }, + { + "epoch": 0.7723090468853181, + "grad_norm": 1.5267519699525045, + "learning_rate": 1.2990282009743365e-06, + "loss": 0.2917, + "step": 35086 + }, + { + "epoch": 0.7723310587717367, + "grad_norm": 1.617402327766514, + "learning_rate": 1.2987885316150056e-06, + "loss": 0.2679, + "step": 35087 + }, + { + "epoch": 0.7723530706581554, + "grad_norm": 1.523290810437592, + "learning_rate": 1.2985488810668806e-06, + "loss": 0.2725, + "step": 35088 + }, + { + "epoch": 0.7723750825445741, + "grad_norm": 0.6245025788092639, + "learning_rate": 1.2983092493311811e-06, + "loss": 0.453, + "step": 35089 + }, + { + "epoch": 0.7723970944309927, + "grad_norm": 1.7305875208856787, + "learning_rate": 1.2980696364091267e-06, + "loss": 0.3326, + "step": 35090 + }, + { + "epoch": 0.7724191063174114, + "grad_norm": 1.6305179702834272, + "learning_rate": 1.297830042301932e-06, + "loss": 0.234, + "step": 35091 + }, + { + "epoch": 0.7724411182038301, + "grad_norm": 1.892469732368095, + "learning_rate": 1.2975904670108176e-06, + "loss": 0.3041, + "step": 35092 + }, + { + "epoch": 0.7724631300902487, + "grad_norm": 0.6278096803189063, + "learning_rate": 1.297350910536998e-06, + "loss": 0.4483, + "step": 35093 + }, + { + "epoch": 0.7724851419766674, + "grad_norm": 0.6203660864643226, + "learning_rate": 1.2971113728816925e-06, + "loss": 0.4545, + "step": 35094 + }, + { + "epoch": 0.7725071538630861, + "grad_norm": 2.0599038434894745, + "learning_rate": 1.2968718540461205e-06, + "loss": 0.3291, + "step": 35095 + }, + { + "epoch": 0.7725291657495047, + "grad_norm": 1.2958622871694516, + "learning_rate": 1.2966323540314957e-06, + "loss": 0.1819, + "step": 35096 + }, + { + "epoch": 0.7725511776359234, + "grad_norm": 1.4131753506774924, + "learning_rate": 1.2963928728390384e-06, + "loss": 0.2903, + "step": 35097 + }, + { + "epoch": 0.7725731895223421, + "grad_norm": 1.4678347141189698, + "learning_rate": 1.296153410469963e-06, + "loss": 0.2703, + "step": 35098 + }, + { + "epoch": 0.7725952014087607, + "grad_norm": 1.529997718435119, + "learning_rate": 1.2959139669254877e-06, + "loss": 0.283, + "step": 35099 + }, + { + "epoch": 0.7726172132951794, + "grad_norm": 1.8540939971821924, + "learning_rate": 1.2956745422068312e-06, + "loss": 0.3024, + "step": 35100 + }, + { + "epoch": 0.7726392251815981, + "grad_norm": 1.9470631695911893, + "learning_rate": 1.295435136315208e-06, + "loss": 0.2862, + "step": 35101 + }, + { + "epoch": 0.7726612370680167, + "grad_norm": 0.6427870423570636, + "learning_rate": 1.2951957492518363e-06, + "loss": 0.472, + "step": 35102 + }, + { + "epoch": 0.7726832489544354, + "grad_norm": 0.6120196365606225, + "learning_rate": 1.294956381017931e-06, + "loss": 0.4429, + "step": 35103 + }, + { + "epoch": 0.7727052608408541, + "grad_norm": 1.4211116074321068, + "learning_rate": 1.2947170316147118e-06, + "loss": 0.2389, + "step": 35104 + }, + { + "epoch": 0.7727272727272727, + "grad_norm": 1.5726640933169298, + "learning_rate": 1.2944777010433911e-06, + "loss": 0.2447, + "step": 35105 + }, + { + "epoch": 0.7727492846136914, + "grad_norm": 1.848212024221467, + "learning_rate": 1.294238389305188e-06, + "loss": 0.2876, + "step": 35106 + }, + { + "epoch": 0.77277129650011, + "grad_norm": 1.7053735440375557, + "learning_rate": 1.2939990964013195e-06, + "loss": 0.2501, + "step": 35107 + }, + { + "epoch": 0.7727933083865287, + "grad_norm": 0.6056384744217179, + "learning_rate": 1.2937598223330006e-06, + "loss": 0.451, + "step": 35108 + }, + { + "epoch": 0.7728153202729474, + "grad_norm": 1.6363221833486683, + "learning_rate": 1.2935205671014455e-06, + "loss": 0.3692, + "step": 35109 + }, + { + "epoch": 0.772837332159366, + "grad_norm": 1.8817099544627913, + "learning_rate": 1.2932813307078722e-06, + "loss": 0.3466, + "step": 35110 + }, + { + "epoch": 0.7728593440457847, + "grad_norm": 1.593540588456731, + "learning_rate": 1.2930421131534971e-06, + "loss": 0.2191, + "step": 35111 + }, + { + "epoch": 0.7728813559322034, + "grad_norm": 0.6040479107975459, + "learning_rate": 1.292802914439536e-06, + "loss": 0.4446, + "step": 35112 + }, + { + "epoch": 0.772903367818622, + "grad_norm": 0.6244940392310423, + "learning_rate": 1.2925637345672038e-06, + "loss": 0.4469, + "step": 35113 + }, + { + "epoch": 0.7729253797050407, + "grad_norm": 5.752737562660468, + "learning_rate": 1.2923245735377156e-06, + "loss": 0.3103, + "step": 35114 + }, + { + "epoch": 0.7729473915914594, + "grad_norm": 1.6232552703653687, + "learning_rate": 1.2920854313522868e-06, + "loss": 0.2651, + "step": 35115 + }, + { + "epoch": 0.772969403477878, + "grad_norm": 1.4458375540860278, + "learning_rate": 1.291846308012134e-06, + "loss": 0.3618, + "step": 35116 + }, + { + "epoch": 0.7729914153642967, + "grad_norm": 0.6137582385732216, + "learning_rate": 1.2916072035184736e-06, + "loss": 0.435, + "step": 35117 + }, + { + "epoch": 0.7730134272507154, + "grad_norm": 1.2765057291615356, + "learning_rate": 1.291368117872519e-06, + "loss": 0.2978, + "step": 35118 + }, + { + "epoch": 0.773035439137134, + "grad_norm": 0.6241404199667377, + "learning_rate": 1.2911290510754849e-06, + "loss": 0.4466, + "step": 35119 + }, + { + "epoch": 0.7730574510235527, + "grad_norm": 4.539224221615539, + "learning_rate": 1.2908900031285864e-06, + "loss": 0.3446, + "step": 35120 + }, + { + "epoch": 0.7730794629099714, + "grad_norm": 2.6699132372538745, + "learning_rate": 1.2906509740330398e-06, + "loss": 0.3622, + "step": 35121 + }, + { + "epoch": 0.77310147479639, + "grad_norm": 0.6133932490340206, + "learning_rate": 1.290411963790061e-06, + "loss": 0.4502, + "step": 35122 + }, + { + "epoch": 0.7731234866828087, + "grad_norm": 1.580654694984198, + "learning_rate": 1.2901729724008627e-06, + "loss": 0.3042, + "step": 35123 + }, + { + "epoch": 0.7731454985692274, + "grad_norm": 1.2960864111830694, + "learning_rate": 1.2899339998666582e-06, + "loss": 0.2747, + "step": 35124 + }, + { + "epoch": 0.773167510455646, + "grad_norm": 1.502425452668144, + "learning_rate": 1.2896950461886642e-06, + "loss": 0.3199, + "step": 35125 + }, + { + "epoch": 0.7731895223420647, + "grad_norm": 1.5940978738065517, + "learning_rate": 1.2894561113680964e-06, + "loss": 0.3336, + "step": 35126 + }, + { + "epoch": 0.7732115342284834, + "grad_norm": 1.4570270831686274, + "learning_rate": 1.2892171954061662e-06, + "loss": 0.2586, + "step": 35127 + }, + { + "epoch": 0.773233546114902, + "grad_norm": 1.1880278334406316, + "learning_rate": 1.2889782983040905e-06, + "loss": 0.2594, + "step": 35128 + }, + { + "epoch": 0.7732555580013207, + "grad_norm": 1.8175772708657083, + "learning_rate": 1.288739420063081e-06, + "loss": 0.3192, + "step": 35129 + }, + { + "epoch": 0.7732775698877393, + "grad_norm": 1.0995255840503064, + "learning_rate": 1.2885005606843532e-06, + "loss": 0.2013, + "step": 35130 + }, + { + "epoch": 0.773299581774158, + "grad_norm": 1.6325313208261267, + "learning_rate": 1.2882617201691223e-06, + "loss": 0.2767, + "step": 35131 + }, + { + "epoch": 0.7733215936605767, + "grad_norm": 1.638636266791123, + "learning_rate": 1.288022898518599e-06, + "loss": 0.3603, + "step": 35132 + }, + { + "epoch": 0.7733436055469953, + "grad_norm": 1.509210671848565, + "learning_rate": 1.2877840957340004e-06, + "loss": 0.3044, + "step": 35133 + }, + { + "epoch": 0.773365617433414, + "grad_norm": 1.8908294605146596, + "learning_rate": 1.2875453118165375e-06, + "loss": 0.2577, + "step": 35134 + }, + { + "epoch": 0.7733876293198327, + "grad_norm": 1.8627392101273699, + "learning_rate": 1.2873065467674252e-06, + "loss": 0.2865, + "step": 35135 + }, + { + "epoch": 0.7734096412062513, + "grad_norm": 1.333930560744394, + "learning_rate": 1.287067800587878e-06, + "loss": 0.305, + "step": 35136 + }, + { + "epoch": 0.77343165309267, + "grad_norm": 1.407316055026075, + "learning_rate": 1.2868290732791073e-06, + "loss": 0.2518, + "step": 35137 + }, + { + "epoch": 0.7734536649790887, + "grad_norm": 1.8201519545231177, + "learning_rate": 1.2865903648423284e-06, + "loss": 0.2694, + "step": 35138 + }, + { + "epoch": 0.7734756768655073, + "grad_norm": 1.8668189863097873, + "learning_rate": 1.286351675278752e-06, + "loss": 0.2402, + "step": 35139 + }, + { + "epoch": 0.773497688751926, + "grad_norm": 1.606162174799024, + "learning_rate": 1.2861130045895942e-06, + "loss": 0.2757, + "step": 35140 + }, + { + "epoch": 0.7735197006383447, + "grad_norm": 1.1636247077600619, + "learning_rate": 1.2858743527760653e-06, + "loss": 0.2865, + "step": 35141 + }, + { + "epoch": 0.7735417125247633, + "grad_norm": 1.8552208745058125, + "learning_rate": 1.2856357198393793e-06, + "loss": 0.2986, + "step": 35142 + }, + { + "epoch": 0.773563724411182, + "grad_norm": 1.6252064336298266, + "learning_rate": 1.2853971057807513e-06, + "loss": 0.2974, + "step": 35143 + }, + { + "epoch": 0.7735857362976007, + "grad_norm": 0.6097165658162281, + "learning_rate": 1.2851585106013913e-06, + "loss": 0.4403, + "step": 35144 + }, + { + "epoch": 0.7736077481840193, + "grad_norm": 0.5941268453434146, + "learning_rate": 1.2849199343025115e-06, + "loss": 0.4309, + "step": 35145 + }, + { + "epoch": 0.773629760070438, + "grad_norm": 1.4917929801484602, + "learning_rate": 1.284681376885325e-06, + "loss": 0.2761, + "step": 35146 + }, + { + "epoch": 0.7736517719568567, + "grad_norm": 1.6265790752374094, + "learning_rate": 1.2844428383510455e-06, + "loss": 0.3414, + "step": 35147 + }, + { + "epoch": 0.7736737838432753, + "grad_norm": 1.4261388055835644, + "learning_rate": 1.2842043187008856e-06, + "loss": 0.2748, + "step": 35148 + }, + { + "epoch": 0.773695795729694, + "grad_norm": 1.8768961409786908, + "learning_rate": 1.283965817936057e-06, + "loss": 0.2549, + "step": 35149 + }, + { + "epoch": 0.7737178076161128, + "grad_norm": 1.6100108675014355, + "learning_rate": 1.2837273360577701e-06, + "loss": 0.3177, + "step": 35150 + }, + { + "epoch": 0.7737398195025313, + "grad_norm": 1.9280752152845748, + "learning_rate": 1.2834888730672384e-06, + "loss": 0.3477, + "step": 35151 + }, + { + "epoch": 0.77376183138895, + "grad_norm": 1.4372477919458084, + "learning_rate": 1.2832504289656733e-06, + "loss": 0.3271, + "step": 35152 + }, + { + "epoch": 0.7737838432753688, + "grad_norm": 7.043690947149269, + "learning_rate": 1.2830120037542893e-06, + "loss": 0.247, + "step": 35153 + }, + { + "epoch": 0.7738058551617873, + "grad_norm": 1.7843788065450112, + "learning_rate": 1.2827735974342959e-06, + "loss": 0.2696, + "step": 35154 + }, + { + "epoch": 0.773827867048206, + "grad_norm": 1.7466357673251696, + "learning_rate": 1.2825352100069038e-06, + "loss": 0.3516, + "step": 35155 + }, + { + "epoch": 0.7738498789346246, + "grad_norm": 1.7746697046870035, + "learning_rate": 1.2822968414733255e-06, + "loss": 0.3212, + "step": 35156 + }, + { + "epoch": 0.7738718908210434, + "grad_norm": 0.6053678894272474, + "learning_rate": 1.2820584918347728e-06, + "loss": 0.4422, + "step": 35157 + }, + { + "epoch": 0.7738939027074621, + "grad_norm": 0.5888254838349694, + "learning_rate": 1.2818201610924587e-06, + "loss": 0.4125, + "step": 35158 + }, + { + "epoch": 0.7739159145938806, + "grad_norm": 1.5050739777725146, + "learning_rate": 1.2815818492475929e-06, + "loss": 0.2734, + "step": 35159 + }, + { + "epoch": 0.7739379264802994, + "grad_norm": 5.31643884544175, + "learning_rate": 1.2813435563013849e-06, + "loss": 0.2674, + "step": 35160 + }, + { + "epoch": 0.7739599383667181, + "grad_norm": 2.3560283946861436, + "learning_rate": 1.281105282255048e-06, + "loss": 0.2482, + "step": 35161 + }, + { + "epoch": 0.7739819502531367, + "grad_norm": 1.5902005818326004, + "learning_rate": 1.2808670271097922e-06, + "loss": 0.3367, + "step": 35162 + }, + { + "epoch": 0.7740039621395554, + "grad_norm": 1.7289495885735326, + "learning_rate": 1.2806287908668301e-06, + "loss": 0.3104, + "step": 35163 + }, + { + "epoch": 0.7740259740259741, + "grad_norm": 5.6016457421295565, + "learning_rate": 1.2803905735273725e-06, + "loss": 0.3436, + "step": 35164 + }, + { + "epoch": 0.7740479859123927, + "grad_norm": 1.8674556798277335, + "learning_rate": 1.2801523750926264e-06, + "loss": 0.2928, + "step": 35165 + }, + { + "epoch": 0.7740699977988114, + "grad_norm": 1.2735640073699055, + "learning_rate": 1.2799141955638056e-06, + "loss": 0.2302, + "step": 35166 + }, + { + "epoch": 0.7740920096852301, + "grad_norm": 0.6075299966545605, + "learning_rate": 1.2796760349421216e-06, + "loss": 0.457, + "step": 35167 + }, + { + "epoch": 0.7741140215716487, + "grad_norm": 2.8834158597427244, + "learning_rate": 1.279437893228782e-06, + "loss": 0.3399, + "step": 35168 + }, + { + "epoch": 0.7741360334580674, + "grad_norm": 0.623510190084911, + "learning_rate": 1.2791997704249992e-06, + "loss": 0.4621, + "step": 35169 + }, + { + "epoch": 0.7741580453444861, + "grad_norm": 2.1018763889381016, + "learning_rate": 1.278961666531982e-06, + "loss": 0.3911, + "step": 35170 + }, + { + "epoch": 0.7741800572309047, + "grad_norm": 1.8074527201499977, + "learning_rate": 1.2787235815509407e-06, + "loss": 0.284, + "step": 35171 + }, + { + "epoch": 0.7742020691173234, + "grad_norm": 2.2530565873563995, + "learning_rate": 1.2784855154830877e-06, + "loss": 0.2808, + "step": 35172 + }, + { + "epoch": 0.7742240810037421, + "grad_norm": 1.891107172869388, + "learning_rate": 1.2782474683296291e-06, + "loss": 0.2475, + "step": 35173 + }, + { + "epoch": 0.7742460928901607, + "grad_norm": 1.516521696280651, + "learning_rate": 1.2780094400917792e-06, + "loss": 0.2985, + "step": 35174 + }, + { + "epoch": 0.7742681047765794, + "grad_norm": 1.5934566445155194, + "learning_rate": 1.2777714307707434e-06, + "loss": 0.2111, + "step": 35175 + }, + { + "epoch": 0.7742901166629981, + "grad_norm": 2.138543371306701, + "learning_rate": 1.2775334403677347e-06, + "loss": 0.2411, + "step": 35176 + }, + { + "epoch": 0.7743121285494167, + "grad_norm": 8.613272278926384, + "learning_rate": 1.27729546888396e-06, + "loss": 0.2259, + "step": 35177 + }, + { + "epoch": 0.7743341404358354, + "grad_norm": 1.5378333409703246, + "learning_rate": 1.27705751632063e-06, + "loss": 0.2919, + "step": 35178 + }, + { + "epoch": 0.774356152322254, + "grad_norm": 1.4702685137487912, + "learning_rate": 1.2768195826789559e-06, + "loss": 0.3153, + "step": 35179 + }, + { + "epoch": 0.7743781642086727, + "grad_norm": 0.6550834327832881, + "learning_rate": 1.2765816679601435e-06, + "loss": 0.4721, + "step": 35180 + }, + { + "epoch": 0.7744001760950914, + "grad_norm": 1.1613907527182343, + "learning_rate": 1.2763437721654053e-06, + "loss": 0.2212, + "step": 35181 + }, + { + "epoch": 0.77442218798151, + "grad_norm": 1.7323615946007345, + "learning_rate": 1.2761058952959476e-06, + "loss": 0.2765, + "step": 35182 + }, + { + "epoch": 0.7744441998679287, + "grad_norm": 1.8384346841531596, + "learning_rate": 1.2758680373529804e-06, + "loss": 0.2868, + "step": 35183 + }, + { + "epoch": 0.7744662117543474, + "grad_norm": 1.2822663793963394, + "learning_rate": 1.2756301983377145e-06, + "loss": 0.2872, + "step": 35184 + }, + { + "epoch": 0.774488223640766, + "grad_norm": 1.6333263957267605, + "learning_rate": 1.2753923782513566e-06, + "loss": 0.2508, + "step": 35185 + }, + { + "epoch": 0.7745102355271847, + "grad_norm": 1.780845250879742, + "learning_rate": 1.2751545770951152e-06, + "loss": 0.2126, + "step": 35186 + }, + { + "epoch": 0.7745322474136034, + "grad_norm": 1.3768887430320313, + "learning_rate": 1.274916794870199e-06, + "loss": 0.2789, + "step": 35187 + }, + { + "epoch": 0.774554259300022, + "grad_norm": 2.1277164180871178, + "learning_rate": 1.2746790315778168e-06, + "loss": 0.361, + "step": 35188 + }, + { + "epoch": 0.7745762711864407, + "grad_norm": 1.374161593478328, + "learning_rate": 1.274441287219179e-06, + "loss": 0.2188, + "step": 35189 + }, + { + "epoch": 0.7745982830728594, + "grad_norm": 2.6979543646399273, + "learning_rate": 1.2742035617954928e-06, + "loss": 0.3458, + "step": 35190 + }, + { + "epoch": 0.774620294959278, + "grad_norm": 3.6931607924323755, + "learning_rate": 1.2739658553079637e-06, + "loss": 0.3576, + "step": 35191 + }, + { + "epoch": 0.7746423068456967, + "grad_norm": 2.000171191658425, + "learning_rate": 1.2737281677578022e-06, + "loss": 0.3857, + "step": 35192 + }, + { + "epoch": 0.7746643187321154, + "grad_norm": 2.409928631809178, + "learning_rate": 1.2734904991462166e-06, + "loss": 0.3545, + "step": 35193 + }, + { + "epoch": 0.774686330618534, + "grad_norm": 2.375832388381501, + "learning_rate": 1.273252849474415e-06, + "loss": 0.361, + "step": 35194 + }, + { + "epoch": 0.7747083425049527, + "grad_norm": 2.902677337230935, + "learning_rate": 1.2730152187436052e-06, + "loss": 0.3296, + "step": 35195 + }, + { + "epoch": 0.7747303543913714, + "grad_norm": 1.407385306178456, + "learning_rate": 1.2727776069549924e-06, + "loss": 0.2269, + "step": 35196 + }, + { + "epoch": 0.77475236627779, + "grad_norm": 1.7284215810417487, + "learning_rate": 1.2725400141097865e-06, + "loss": 0.277, + "step": 35197 + }, + { + "epoch": 0.7747743781642087, + "grad_norm": 2.302649627852705, + "learning_rate": 1.272302440209195e-06, + "loss": 0.2152, + "step": 35198 + }, + { + "epoch": 0.7747963900506274, + "grad_norm": 1.5722986415040876, + "learning_rate": 1.2720648852544259e-06, + "loss": 0.2429, + "step": 35199 + }, + { + "epoch": 0.774818401937046, + "grad_norm": 1.3079654991260163, + "learning_rate": 1.271827349246686e-06, + "loss": 0.3024, + "step": 35200 + }, + { + "epoch": 0.7748404138234647, + "grad_norm": 4.222085399331384, + "learning_rate": 1.2715898321871811e-06, + "loss": 0.3476, + "step": 35201 + }, + { + "epoch": 0.7748624257098833, + "grad_norm": 1.4657353065171776, + "learning_rate": 1.2713523340771195e-06, + "loss": 0.2854, + "step": 35202 + }, + { + "epoch": 0.774884437596302, + "grad_norm": 1.9250999323844298, + "learning_rate": 1.2711148549177082e-06, + "loss": 0.212, + "step": 35203 + }, + { + "epoch": 0.7749064494827207, + "grad_norm": 2.3306073241007765, + "learning_rate": 1.270877394710156e-06, + "loss": 0.3478, + "step": 35204 + }, + { + "epoch": 0.7749284613691393, + "grad_norm": 0.6399610684334966, + "learning_rate": 1.2706399534556679e-06, + "loss": 0.4571, + "step": 35205 + }, + { + "epoch": 0.774950473255558, + "grad_norm": 1.2766810046449044, + "learning_rate": 1.2704025311554497e-06, + "loss": 0.286, + "step": 35206 + }, + { + "epoch": 0.7749724851419767, + "grad_norm": 1.4610009971390105, + "learning_rate": 1.2701651278107092e-06, + "loss": 0.2933, + "step": 35207 + }, + { + "epoch": 0.7749944970283953, + "grad_norm": 1.7196986732681907, + "learning_rate": 1.2699277434226543e-06, + "loss": 0.3489, + "step": 35208 + }, + { + "epoch": 0.775016508914814, + "grad_norm": 1.8278786565029082, + "learning_rate": 1.269690377992489e-06, + "loss": 0.3757, + "step": 35209 + }, + { + "epoch": 0.7750385208012327, + "grad_norm": 1.525748436163706, + "learning_rate": 1.2694530315214225e-06, + "loss": 0.3542, + "step": 35210 + }, + { + "epoch": 0.7750605326876513, + "grad_norm": 2.2440831249145132, + "learning_rate": 1.2692157040106579e-06, + "loss": 0.3587, + "step": 35211 + }, + { + "epoch": 0.77508254457407, + "grad_norm": 1.8535020736476782, + "learning_rate": 1.2689783954614044e-06, + "loss": 0.3081, + "step": 35212 + }, + { + "epoch": 0.7751045564604887, + "grad_norm": 1.916951580204462, + "learning_rate": 1.2687411058748656e-06, + "loss": 0.2677, + "step": 35213 + }, + { + "epoch": 0.7751265683469073, + "grad_norm": 1.348094680107706, + "learning_rate": 1.2685038352522488e-06, + "loss": 0.2931, + "step": 35214 + }, + { + "epoch": 0.775148580233326, + "grad_norm": 1.4178467155842458, + "learning_rate": 1.2682665835947611e-06, + "loss": 0.2575, + "step": 35215 + }, + { + "epoch": 0.7751705921197447, + "grad_norm": 1.72854619467019, + "learning_rate": 1.2680293509036058e-06, + "loss": 0.2943, + "step": 35216 + }, + { + "epoch": 0.7751926040061633, + "grad_norm": 1.6589457755459753, + "learning_rate": 1.2677921371799906e-06, + "loss": 0.3338, + "step": 35217 + }, + { + "epoch": 0.775214615892582, + "grad_norm": 1.3863234959004473, + "learning_rate": 1.2675549424251192e-06, + "loss": 0.2596, + "step": 35218 + }, + { + "epoch": 0.7752366277790007, + "grad_norm": 11.687032940271429, + "learning_rate": 1.2673177666401987e-06, + "loss": 0.2948, + "step": 35219 + }, + { + "epoch": 0.7752586396654193, + "grad_norm": 1.4771881481380946, + "learning_rate": 1.2670806098264354e-06, + "loss": 0.2422, + "step": 35220 + }, + { + "epoch": 0.775280651551838, + "grad_norm": 2.116108514028525, + "learning_rate": 1.266843471985032e-06, + "loss": 0.3061, + "step": 35221 + }, + { + "epoch": 0.7753026634382567, + "grad_norm": 1.6041021826689934, + "learning_rate": 1.2666063531171963e-06, + "loss": 0.2776, + "step": 35222 + }, + { + "epoch": 0.7753246753246753, + "grad_norm": 1.5177309510313048, + "learning_rate": 1.266369253224131e-06, + "loss": 0.2517, + "step": 35223 + }, + { + "epoch": 0.775346687211094, + "grad_norm": 1.3384872628081002, + "learning_rate": 1.2661321723070425e-06, + "loss": 0.299, + "step": 35224 + }, + { + "epoch": 0.7753686990975126, + "grad_norm": 0.6202165253702108, + "learning_rate": 1.2658951103671368e-06, + "loss": 0.4467, + "step": 35225 + }, + { + "epoch": 0.7753907109839313, + "grad_norm": 2.2480103970792737, + "learning_rate": 1.2656580674056179e-06, + "loss": 0.2759, + "step": 35226 + }, + { + "epoch": 0.77541272287035, + "grad_norm": 1.9818313736168556, + "learning_rate": 1.2654210434236886e-06, + "loss": 0.2946, + "step": 35227 + }, + { + "epoch": 0.7754347347567686, + "grad_norm": 1.5215184290262733, + "learning_rate": 1.265184038422555e-06, + "loss": 0.2503, + "step": 35228 + }, + { + "epoch": 0.7754567466431873, + "grad_norm": 1.445170671875213, + "learning_rate": 1.2649470524034225e-06, + "loss": 0.2719, + "step": 35229 + }, + { + "epoch": 0.775478758529606, + "grad_norm": 1.9505429848131295, + "learning_rate": 1.264710085367496e-06, + "loss": 0.2794, + "step": 35230 + }, + { + "epoch": 0.7755007704160246, + "grad_norm": 3.521668166852584, + "learning_rate": 1.264473137315979e-06, + "loss": 0.3704, + "step": 35231 + }, + { + "epoch": 0.7755227823024433, + "grad_norm": 2.115236615044596, + "learning_rate": 1.2642362082500741e-06, + "loss": 0.2441, + "step": 35232 + }, + { + "epoch": 0.775544794188862, + "grad_norm": 1.7438023442572534, + "learning_rate": 1.263999298170987e-06, + "loss": 0.3686, + "step": 35233 + }, + { + "epoch": 0.7755668060752806, + "grad_norm": 1.253893688014078, + "learning_rate": 1.2637624070799216e-06, + "loss": 0.2771, + "step": 35234 + }, + { + "epoch": 0.7755888179616993, + "grad_norm": 0.620916229336193, + "learning_rate": 1.2635255349780835e-06, + "loss": 0.4641, + "step": 35235 + }, + { + "epoch": 0.775610829848118, + "grad_norm": 0.5973856437699029, + "learning_rate": 1.2632886818666756e-06, + "loss": 0.4605, + "step": 35236 + }, + { + "epoch": 0.7756328417345366, + "grad_norm": 1.4826499618037436, + "learning_rate": 1.2630518477468994e-06, + "loss": 0.2697, + "step": 35237 + }, + { + "epoch": 0.7756548536209553, + "grad_norm": 1.9758707247169518, + "learning_rate": 1.2628150326199607e-06, + "loss": 0.3237, + "step": 35238 + }, + { + "epoch": 0.775676865507374, + "grad_norm": 1.4077593478666734, + "learning_rate": 1.2625782364870626e-06, + "loss": 0.3006, + "step": 35239 + }, + { + "epoch": 0.7756988773937926, + "grad_norm": 1.8659314223431631, + "learning_rate": 1.2623414593494099e-06, + "loss": 0.2847, + "step": 35240 + }, + { + "epoch": 0.7757208892802113, + "grad_norm": 7.011372787942094, + "learning_rate": 1.2621047012082055e-06, + "loss": 0.2488, + "step": 35241 + }, + { + "epoch": 0.77574290116663, + "grad_norm": 1.3372456865820133, + "learning_rate": 1.2618679620646506e-06, + "loss": 0.2371, + "step": 35242 + }, + { + "epoch": 0.7757649130530486, + "grad_norm": 1.1660747428177918, + "learning_rate": 1.26163124191995e-06, + "loss": 0.2489, + "step": 35243 + }, + { + "epoch": 0.7757869249394673, + "grad_norm": 6.675743151162732, + "learning_rate": 1.261394540775308e-06, + "loss": 0.2824, + "step": 35244 + }, + { + "epoch": 0.775808936825886, + "grad_norm": 0.6192366061384326, + "learning_rate": 1.2611578586319257e-06, + "loss": 0.4616, + "step": 35245 + }, + { + "epoch": 0.7758309487123046, + "grad_norm": 1.4857848521690122, + "learning_rate": 1.2609211954910072e-06, + "loss": 0.2658, + "step": 35246 + }, + { + "epoch": 0.7758529605987233, + "grad_norm": 1.4774554866067826, + "learning_rate": 1.260684551353754e-06, + "loss": 0.2817, + "step": 35247 + }, + { + "epoch": 0.7758749724851419, + "grad_norm": 2.141030344047183, + "learning_rate": 1.2604479262213708e-06, + "loss": 0.269, + "step": 35248 + }, + { + "epoch": 0.7758969843715606, + "grad_norm": 1.4739147403892483, + "learning_rate": 1.2602113200950583e-06, + "loss": 0.2583, + "step": 35249 + }, + { + "epoch": 0.7759189962579793, + "grad_norm": 1.5407038199855276, + "learning_rate": 1.2599747329760192e-06, + "loss": 0.282, + "step": 35250 + }, + { + "epoch": 0.7759410081443979, + "grad_norm": 1.563364022618674, + "learning_rate": 1.2597381648654584e-06, + "loss": 0.2362, + "step": 35251 + }, + { + "epoch": 0.7759630200308166, + "grad_norm": 1.0506684809101448, + "learning_rate": 1.2595016157645745e-06, + "loss": 0.1702, + "step": 35252 + }, + { + "epoch": 0.7759850319172353, + "grad_norm": 1.2627260022258684, + "learning_rate": 1.2592650856745737e-06, + "loss": 0.2355, + "step": 35253 + }, + { + "epoch": 0.7760070438036539, + "grad_norm": 3.9189279495279226, + "learning_rate": 1.2590285745966547e-06, + "loss": 0.2885, + "step": 35254 + }, + { + "epoch": 0.7760290556900726, + "grad_norm": 1.7559156828360072, + "learning_rate": 1.258792082532021e-06, + "loss": 0.2442, + "step": 35255 + }, + { + "epoch": 0.7760510675764913, + "grad_norm": 0.6318917977522244, + "learning_rate": 1.2585556094818757e-06, + "loss": 0.4512, + "step": 35256 + }, + { + "epoch": 0.7760730794629099, + "grad_norm": 1.5229647601079805, + "learning_rate": 1.2583191554474183e-06, + "loss": 0.2814, + "step": 35257 + }, + { + "epoch": 0.7760950913493286, + "grad_norm": 2.2223342895648597, + "learning_rate": 1.258082720429853e-06, + "loss": 0.3901, + "step": 35258 + }, + { + "epoch": 0.7761171032357473, + "grad_norm": 1.9122354466431195, + "learning_rate": 1.257846304430379e-06, + "loss": 0.2729, + "step": 35259 + }, + { + "epoch": 0.7761391151221659, + "grad_norm": 1.6283597446385352, + "learning_rate": 1.2576099074501992e-06, + "loss": 0.2899, + "step": 35260 + }, + { + "epoch": 0.7761611270085846, + "grad_norm": 3.228504870669025, + "learning_rate": 1.257373529490516e-06, + "loss": 0.2196, + "step": 35261 + }, + { + "epoch": 0.7761831388950033, + "grad_norm": 1.1918487348091908, + "learning_rate": 1.2571371705525286e-06, + "loss": 0.2291, + "step": 35262 + }, + { + "epoch": 0.7762051507814219, + "grad_norm": 1.7215677684716655, + "learning_rate": 1.2569008306374415e-06, + "loss": 0.2796, + "step": 35263 + }, + { + "epoch": 0.7762271626678406, + "grad_norm": 2.6193369991765336, + "learning_rate": 1.2566645097464515e-06, + "loss": 0.336, + "step": 35264 + }, + { + "epoch": 0.7762491745542593, + "grad_norm": 1.514319695810658, + "learning_rate": 1.2564282078807627e-06, + "loss": 0.2755, + "step": 35265 + }, + { + "epoch": 0.7762711864406779, + "grad_norm": 1.8671954982330892, + "learning_rate": 1.2561919250415766e-06, + "loss": 0.2554, + "step": 35266 + }, + { + "epoch": 0.7762931983270966, + "grad_norm": 2.0756897500728946, + "learning_rate": 1.2559556612300928e-06, + "loss": 0.307, + "step": 35267 + }, + { + "epoch": 0.7763152102135153, + "grad_norm": 1.3400454224167295, + "learning_rate": 1.2557194164475106e-06, + "loss": 0.2023, + "step": 35268 + }, + { + "epoch": 0.7763372220999339, + "grad_norm": 0.6138677915277148, + "learning_rate": 1.2554831906950326e-06, + "loss": 0.4194, + "step": 35269 + }, + { + "epoch": 0.7763592339863526, + "grad_norm": 2.433891220832397, + "learning_rate": 1.2552469839738595e-06, + "loss": 0.3401, + "step": 35270 + }, + { + "epoch": 0.7763812458727712, + "grad_norm": 0.647501992556207, + "learning_rate": 1.2550107962851926e-06, + "loss": 0.4559, + "step": 35271 + }, + { + "epoch": 0.77640325775919, + "grad_norm": 1.880559737893545, + "learning_rate": 1.2547746276302308e-06, + "loss": 0.3454, + "step": 35272 + }, + { + "epoch": 0.7764252696456087, + "grad_norm": 1.7180911286716414, + "learning_rate": 1.2545384780101738e-06, + "loss": 0.303, + "step": 35273 + }, + { + "epoch": 0.7764472815320272, + "grad_norm": 2.7847383257922007, + "learning_rate": 1.2543023474262223e-06, + "loss": 0.271, + "step": 35274 + }, + { + "epoch": 0.776469293418446, + "grad_norm": 0.6029410505973342, + "learning_rate": 1.2540662358795774e-06, + "loss": 0.4447, + "step": 35275 + }, + { + "epoch": 0.7764913053048647, + "grad_norm": 1.7303027903970696, + "learning_rate": 1.2538301433714396e-06, + "loss": 0.2532, + "step": 35276 + }, + { + "epoch": 0.7765133171912832, + "grad_norm": 3.3387195898587283, + "learning_rate": 1.253594069903008e-06, + "loss": 0.2392, + "step": 35277 + }, + { + "epoch": 0.776535329077702, + "grad_norm": 0.6031551199901761, + "learning_rate": 1.2533580154754809e-06, + "loss": 0.4445, + "step": 35278 + }, + { + "epoch": 0.7765573409641207, + "grad_norm": 1.3802022023470755, + "learning_rate": 1.2531219800900595e-06, + "loss": 0.1971, + "step": 35279 + }, + { + "epoch": 0.7765793528505393, + "grad_norm": 0.6117035812843512, + "learning_rate": 1.2528859637479436e-06, + "loss": 0.4536, + "step": 35280 + }, + { + "epoch": 0.776601364736958, + "grad_norm": 2.0847377503448103, + "learning_rate": 1.2526499664503334e-06, + "loss": 0.3001, + "step": 35281 + }, + { + "epoch": 0.7766233766233767, + "grad_norm": 2.906156266316768, + "learning_rate": 1.2524139881984276e-06, + "loss": 0.3173, + "step": 35282 + }, + { + "epoch": 0.7766453885097953, + "grad_norm": 1.709475849408856, + "learning_rate": 1.2521780289934237e-06, + "loss": 0.2974, + "step": 35283 + }, + { + "epoch": 0.776667400396214, + "grad_norm": 2.0512878903179717, + "learning_rate": 1.2519420888365247e-06, + "loss": 0.2801, + "step": 35284 + }, + { + "epoch": 0.7766894122826327, + "grad_norm": 0.6199885198153936, + "learning_rate": 1.2517061677289261e-06, + "loss": 0.4738, + "step": 35285 + }, + { + "epoch": 0.7767114241690513, + "grad_norm": 1.4960626783011566, + "learning_rate": 1.251470265671828e-06, + "loss": 0.2547, + "step": 35286 + }, + { + "epoch": 0.77673343605547, + "grad_norm": 1.3909870420179842, + "learning_rate": 1.251234382666432e-06, + "loss": 0.2817, + "step": 35287 + }, + { + "epoch": 0.7767554479418887, + "grad_norm": 1.482682194508343, + "learning_rate": 1.2509985187139328e-06, + "loss": 0.2817, + "step": 35288 + }, + { + "epoch": 0.7767774598283073, + "grad_norm": 1.7815768832808072, + "learning_rate": 1.2507626738155332e-06, + "loss": 0.3343, + "step": 35289 + }, + { + "epoch": 0.776799471714726, + "grad_norm": 1.6046181478865178, + "learning_rate": 1.2505268479724281e-06, + "loss": 0.2768, + "step": 35290 + }, + { + "epoch": 0.7768214836011447, + "grad_norm": 1.4863745781068562, + "learning_rate": 1.2502910411858182e-06, + "loss": 0.29, + "step": 35291 + }, + { + "epoch": 0.7768434954875633, + "grad_norm": 0.6196981842775258, + "learning_rate": 1.2500552534569028e-06, + "loss": 0.4451, + "step": 35292 + }, + { + "epoch": 0.776865507373982, + "grad_norm": 1.6024723628678865, + "learning_rate": 1.2498194847868778e-06, + "loss": 0.3487, + "step": 35293 + }, + { + "epoch": 0.7768875192604006, + "grad_norm": 2.3567010262403807, + "learning_rate": 1.2495837351769442e-06, + "loss": 0.3111, + "step": 35294 + }, + { + "epoch": 0.7769095311468193, + "grad_norm": 1.9611167781067962, + "learning_rate": 1.2493480046282974e-06, + "loss": 0.2108, + "step": 35295 + }, + { + "epoch": 0.776931543033238, + "grad_norm": 0.617901649897332, + "learning_rate": 1.2491122931421368e-06, + "loss": 0.455, + "step": 35296 + }, + { + "epoch": 0.7769535549196566, + "grad_norm": 0.6032577321383377, + "learning_rate": 1.2488766007196617e-06, + "loss": 0.4558, + "step": 35297 + }, + { + "epoch": 0.7769755668060753, + "grad_norm": 2.1605738719845706, + "learning_rate": 1.2486409273620675e-06, + "loss": 0.273, + "step": 35298 + }, + { + "epoch": 0.776997578692494, + "grad_norm": 1.3694122851049155, + "learning_rate": 1.2484052730705548e-06, + "loss": 0.2989, + "step": 35299 + }, + { + "epoch": 0.7770195905789126, + "grad_norm": 2.4311710484887246, + "learning_rate": 1.2481696378463187e-06, + "loss": 0.1959, + "step": 35300 + }, + { + "epoch": 0.7770416024653313, + "grad_norm": 2.579035566585246, + "learning_rate": 1.2479340216905572e-06, + "loss": 0.3911, + "step": 35301 + }, + { + "epoch": 0.77706361435175, + "grad_norm": 1.651768912276805, + "learning_rate": 1.2476984246044704e-06, + "loss": 0.2006, + "step": 35302 + }, + { + "epoch": 0.7770856262381686, + "grad_norm": 2.14202567632634, + "learning_rate": 1.2474628465892535e-06, + "loss": 0.2563, + "step": 35303 + }, + { + "epoch": 0.7771076381245873, + "grad_norm": 2.2112208201804338, + "learning_rate": 1.2472272876461028e-06, + "loss": 0.2863, + "step": 35304 + }, + { + "epoch": 0.777129650011006, + "grad_norm": 1.5402567728788383, + "learning_rate": 1.2469917477762167e-06, + "loss": 0.2463, + "step": 35305 + }, + { + "epoch": 0.7771516618974246, + "grad_norm": 2.10739077450183, + "learning_rate": 1.2467562269807925e-06, + "loss": 0.2992, + "step": 35306 + }, + { + "epoch": 0.7771736737838433, + "grad_norm": 1.3483324156857033, + "learning_rate": 1.2465207252610284e-06, + "loss": 0.2858, + "step": 35307 + }, + { + "epoch": 0.777195685670262, + "grad_norm": 1.9947265310364943, + "learning_rate": 1.2462852426181198e-06, + "loss": 0.2921, + "step": 35308 + }, + { + "epoch": 0.7772176975566806, + "grad_norm": 0.6279852326616067, + "learning_rate": 1.2460497790532634e-06, + "loss": 0.4465, + "step": 35309 + }, + { + "epoch": 0.7772397094430993, + "grad_norm": 8.042826304873358, + "learning_rate": 1.2458143345676553e-06, + "loss": 0.3325, + "step": 35310 + }, + { + "epoch": 0.777261721329518, + "grad_norm": 1.1134018668866426, + "learning_rate": 1.2455789091624938e-06, + "loss": 0.2469, + "step": 35311 + }, + { + "epoch": 0.7772837332159366, + "grad_norm": 1.8827653830857616, + "learning_rate": 1.245343502838976e-06, + "loss": 0.306, + "step": 35312 + }, + { + "epoch": 0.7773057451023553, + "grad_norm": 3.2406065404925912, + "learning_rate": 1.245108115598297e-06, + "loss": 0.2364, + "step": 35313 + }, + { + "epoch": 0.777327756988774, + "grad_norm": 1.6085952956755543, + "learning_rate": 1.2448727474416522e-06, + "loss": 0.313, + "step": 35314 + }, + { + "epoch": 0.7773497688751926, + "grad_norm": 1.913986635823419, + "learning_rate": 1.2446373983702398e-06, + "loss": 0.2691, + "step": 35315 + }, + { + "epoch": 0.7773717807616113, + "grad_norm": 1.5672403789100864, + "learning_rate": 1.2444020683852526e-06, + "loss": 0.2805, + "step": 35316 + }, + { + "epoch": 0.77739379264803, + "grad_norm": 1.2974391182863836, + "learning_rate": 1.244166757487892e-06, + "loss": 0.2627, + "step": 35317 + }, + { + "epoch": 0.7774158045344486, + "grad_norm": 2.5933545480383384, + "learning_rate": 1.243931465679351e-06, + "loss": 0.223, + "step": 35318 + }, + { + "epoch": 0.7774378164208673, + "grad_norm": 1.6007928185063443, + "learning_rate": 1.2436961929608244e-06, + "loss": 0.3412, + "step": 35319 + }, + { + "epoch": 0.7774598283072859, + "grad_norm": 1.7809633348762217, + "learning_rate": 1.24346093933351e-06, + "loss": 0.3246, + "step": 35320 + }, + { + "epoch": 0.7774818401937046, + "grad_norm": 1.5276960956478711, + "learning_rate": 1.2432257047986001e-06, + "loss": 0.2328, + "step": 35321 + }, + { + "epoch": 0.7775038520801233, + "grad_norm": 1.6575121923830602, + "learning_rate": 1.2429904893572958e-06, + "loss": 0.3025, + "step": 35322 + }, + { + "epoch": 0.7775258639665419, + "grad_norm": 5.142217668268333, + "learning_rate": 1.242755293010789e-06, + "loss": 0.2709, + "step": 35323 + }, + { + "epoch": 0.7775478758529606, + "grad_norm": 0.6248672179520303, + "learning_rate": 1.2425201157602745e-06, + "loss": 0.4183, + "step": 35324 + }, + { + "epoch": 0.7775698877393793, + "grad_norm": 2.125405069456768, + "learning_rate": 1.24228495760695e-06, + "loss": 0.4202, + "step": 35325 + }, + { + "epoch": 0.7775918996257979, + "grad_norm": 1.6026334786346415, + "learning_rate": 1.2420498185520074e-06, + "loss": 0.2878, + "step": 35326 + }, + { + "epoch": 0.7776139115122166, + "grad_norm": 1.6516834589486502, + "learning_rate": 1.2418146985966446e-06, + "loss": 0.3026, + "step": 35327 + }, + { + "epoch": 0.7776359233986353, + "grad_norm": 1.5795004597033118, + "learning_rate": 1.2415795977420564e-06, + "loss": 0.2534, + "step": 35328 + }, + { + "epoch": 0.7776579352850539, + "grad_norm": 2.2522394051885297, + "learning_rate": 1.241344515989436e-06, + "loss": 0.2645, + "step": 35329 + }, + { + "epoch": 0.7776799471714726, + "grad_norm": 1.4389245199977623, + "learning_rate": 1.2411094533399803e-06, + "loss": 0.338, + "step": 35330 + }, + { + "epoch": 0.7777019590578913, + "grad_norm": 2.5108671122806903, + "learning_rate": 1.2408744097948816e-06, + "loss": 0.383, + "step": 35331 + }, + { + "epoch": 0.7777239709443099, + "grad_norm": 2.5478053257272224, + "learning_rate": 1.2406393853553361e-06, + "loss": 0.3579, + "step": 35332 + }, + { + "epoch": 0.7777459828307286, + "grad_norm": 1.6178898006885063, + "learning_rate": 1.2404043800225391e-06, + "loss": 0.2301, + "step": 35333 + }, + { + "epoch": 0.7777679947171473, + "grad_norm": 0.6015463561225287, + "learning_rate": 1.240169393797682e-06, + "loss": 0.4431, + "step": 35334 + }, + { + "epoch": 0.7777900066035659, + "grad_norm": 2.226915808090319, + "learning_rate": 1.2399344266819629e-06, + "loss": 0.3398, + "step": 35335 + }, + { + "epoch": 0.7778120184899846, + "grad_norm": 1.7632719518934505, + "learning_rate": 1.2396994786765726e-06, + "loss": 0.2791, + "step": 35336 + }, + { + "epoch": 0.7778340303764033, + "grad_norm": 0.6289510601800774, + "learning_rate": 1.2394645497827067e-06, + "loss": 0.4718, + "step": 35337 + }, + { + "epoch": 0.7778560422628219, + "grad_norm": 1.3693814589044147, + "learning_rate": 1.2392296400015607e-06, + "loss": 0.2715, + "step": 35338 + }, + { + "epoch": 0.7778780541492406, + "grad_norm": 1.98933927466098, + "learning_rate": 1.2389947493343252e-06, + "loss": 0.2985, + "step": 35339 + }, + { + "epoch": 0.7779000660356593, + "grad_norm": 1.2730221632230485, + "learning_rate": 1.2387598777821974e-06, + "loss": 0.2857, + "step": 35340 + }, + { + "epoch": 0.7779220779220779, + "grad_norm": 0.6283341687128529, + "learning_rate": 1.2385250253463682e-06, + "loss": 0.4441, + "step": 35341 + }, + { + "epoch": 0.7779440898084966, + "grad_norm": 1.666818712060287, + "learning_rate": 1.2382901920280326e-06, + "loss": 0.2521, + "step": 35342 + }, + { + "epoch": 0.7779661016949152, + "grad_norm": 0.619584900082083, + "learning_rate": 1.2380553778283854e-06, + "loss": 0.4471, + "step": 35343 + }, + { + "epoch": 0.7779881135813339, + "grad_norm": 1.623130009166348, + "learning_rate": 1.2378205827486185e-06, + "loss": 0.2312, + "step": 35344 + }, + { + "epoch": 0.7780101254677526, + "grad_norm": 2.533103521550457, + "learning_rate": 1.2375858067899239e-06, + "loss": 0.3015, + "step": 35345 + }, + { + "epoch": 0.7780321373541712, + "grad_norm": 1.870647353691929, + "learning_rate": 1.2373510499534963e-06, + "loss": 0.3216, + "step": 35346 + }, + { + "epoch": 0.7780541492405899, + "grad_norm": 1.7648306213919809, + "learning_rate": 1.2371163122405288e-06, + "loss": 0.2554, + "step": 35347 + }, + { + "epoch": 0.7780761611270086, + "grad_norm": 1.6954693924535238, + "learning_rate": 1.2368815936522166e-06, + "loss": 0.2472, + "step": 35348 + }, + { + "epoch": 0.7780981730134272, + "grad_norm": 1.3965545853878174, + "learning_rate": 1.2366468941897498e-06, + "loss": 0.3929, + "step": 35349 + }, + { + "epoch": 0.7781201848998459, + "grad_norm": 2.867231598028731, + "learning_rate": 1.2364122138543205e-06, + "loss": 0.3055, + "step": 35350 + }, + { + "epoch": 0.7781421967862646, + "grad_norm": 1.5965245190219892, + "learning_rate": 1.2361775526471247e-06, + "loss": 0.3581, + "step": 35351 + }, + { + "epoch": 0.7781642086726832, + "grad_norm": 0.6186387172776449, + "learning_rate": 1.2359429105693505e-06, + "loss": 0.4407, + "step": 35352 + }, + { + "epoch": 0.7781862205591019, + "grad_norm": 1.6078836294312946, + "learning_rate": 1.235708287622196e-06, + "loss": 0.303, + "step": 35353 + }, + { + "epoch": 0.7782082324455206, + "grad_norm": 1.6827021245912024, + "learning_rate": 1.235473683806851e-06, + "loss": 0.3498, + "step": 35354 + }, + { + "epoch": 0.7782302443319392, + "grad_norm": 1.3970378330414297, + "learning_rate": 1.2352390991245061e-06, + "loss": 0.2327, + "step": 35355 + }, + { + "epoch": 0.7782522562183579, + "grad_norm": 1.417089514152872, + "learning_rate": 1.2350045335763566e-06, + "loss": 0.3887, + "step": 35356 + }, + { + "epoch": 0.7782742681047766, + "grad_norm": 1.4612066754225705, + "learning_rate": 1.234769987163591e-06, + "loss": 0.2834, + "step": 35357 + }, + { + "epoch": 0.7782962799911952, + "grad_norm": 1.2913496095792507, + "learning_rate": 1.234535459887406e-06, + "loss": 0.2646, + "step": 35358 + }, + { + "epoch": 0.7783182918776139, + "grad_norm": 3.569739950637886, + "learning_rate": 1.2343009517489912e-06, + "loss": 0.2441, + "step": 35359 + }, + { + "epoch": 0.7783403037640326, + "grad_norm": 1.990574303205609, + "learning_rate": 1.2340664627495374e-06, + "loss": 0.3183, + "step": 35360 + }, + { + "epoch": 0.7783623156504512, + "grad_norm": 0.5880538859253559, + "learning_rate": 1.2338319928902388e-06, + "loss": 0.4449, + "step": 35361 + }, + { + "epoch": 0.7783843275368699, + "grad_norm": 0.6412120817265149, + "learning_rate": 1.233597542172284e-06, + "loss": 0.4674, + "step": 35362 + }, + { + "epoch": 0.7784063394232886, + "grad_norm": 1.411727447126264, + "learning_rate": 1.2333631105968668e-06, + "loss": 0.4128, + "step": 35363 + }, + { + "epoch": 0.7784283513097072, + "grad_norm": 2.153583748982422, + "learning_rate": 1.2331286981651797e-06, + "loss": 0.2693, + "step": 35364 + }, + { + "epoch": 0.7784503631961259, + "grad_norm": 0.599962010798191, + "learning_rate": 1.2328943048784108e-06, + "loss": 0.4524, + "step": 35365 + }, + { + "epoch": 0.7784723750825445, + "grad_norm": 1.200495283596953, + "learning_rate": 1.232659930737755e-06, + "loss": 0.2075, + "step": 35366 + }, + { + "epoch": 0.7784943869689632, + "grad_norm": 1.5598080642422727, + "learning_rate": 1.2324255757443998e-06, + "loss": 0.2761, + "step": 35367 + }, + { + "epoch": 0.7785163988553819, + "grad_norm": 0.6269732119637174, + "learning_rate": 1.232191239899539e-06, + "loss": 0.4658, + "step": 35368 + }, + { + "epoch": 0.7785384107418005, + "grad_norm": 2.6692554316024806, + "learning_rate": 1.2319569232043638e-06, + "loss": 0.2596, + "step": 35369 + }, + { + "epoch": 0.7785604226282192, + "grad_norm": 2.1812249643182673, + "learning_rate": 1.2317226256600623e-06, + "loss": 0.3209, + "step": 35370 + }, + { + "epoch": 0.7785824345146379, + "grad_norm": 1.531489733544334, + "learning_rate": 1.231488347267829e-06, + "loss": 0.2863, + "step": 35371 + }, + { + "epoch": 0.7786044464010565, + "grad_norm": 1.3325868050563865, + "learning_rate": 1.2312540880288515e-06, + "loss": 0.3305, + "step": 35372 + }, + { + "epoch": 0.7786264582874752, + "grad_norm": 0.6237635523056133, + "learning_rate": 1.2310198479443213e-06, + "loss": 0.4481, + "step": 35373 + }, + { + "epoch": 0.7786484701738939, + "grad_norm": 1.8927260618873618, + "learning_rate": 1.2307856270154306e-06, + "loss": 0.3513, + "step": 35374 + }, + { + "epoch": 0.7786704820603125, + "grad_norm": 5.401674812856306, + "learning_rate": 1.230551425243367e-06, + "loss": 0.3073, + "step": 35375 + }, + { + "epoch": 0.7786924939467312, + "grad_norm": 1.526799745559196, + "learning_rate": 1.2303172426293243e-06, + "loss": 0.3045, + "step": 35376 + }, + { + "epoch": 0.7787145058331499, + "grad_norm": 1.653881777304471, + "learning_rate": 1.2300830791744884e-06, + "loss": 0.2618, + "step": 35377 + }, + { + "epoch": 0.7787365177195685, + "grad_norm": 4.984478974049143, + "learning_rate": 1.2298489348800523e-06, + "loss": 0.4231, + "step": 35378 + }, + { + "epoch": 0.7787585296059872, + "grad_norm": 0.5923595264378834, + "learning_rate": 1.2296148097472066e-06, + "loss": 0.4382, + "step": 35379 + }, + { + "epoch": 0.7787805414924059, + "grad_norm": 1.930884211060254, + "learning_rate": 1.2293807037771388e-06, + "loss": 0.3775, + "step": 35380 + }, + { + "epoch": 0.7788025533788245, + "grad_norm": 1.6833650003517937, + "learning_rate": 1.2291466169710415e-06, + "loss": 0.2798, + "step": 35381 + }, + { + "epoch": 0.7788245652652432, + "grad_norm": 1.3968129687028459, + "learning_rate": 1.2289125493301013e-06, + "loss": 0.2624, + "step": 35382 + }, + { + "epoch": 0.778846577151662, + "grad_norm": 2.0139385956123297, + "learning_rate": 1.2286785008555098e-06, + "loss": 0.3192, + "step": 35383 + }, + { + "epoch": 0.7788685890380805, + "grad_norm": 1.635734400477213, + "learning_rate": 1.2284444715484572e-06, + "loss": 0.3346, + "step": 35384 + }, + { + "epoch": 0.7788906009244992, + "grad_norm": 1.3561653552609592, + "learning_rate": 1.2282104614101326e-06, + "loss": 0.2618, + "step": 35385 + }, + { + "epoch": 0.778912612810918, + "grad_norm": 1.4837814604787687, + "learning_rate": 1.227976470441723e-06, + "loss": 0.2659, + "step": 35386 + }, + { + "epoch": 0.7789346246973365, + "grad_norm": 1.7680302939142656, + "learning_rate": 1.2277424986444202e-06, + "loss": 0.3506, + "step": 35387 + }, + { + "epoch": 0.7789566365837552, + "grad_norm": 1.4626888757215963, + "learning_rate": 1.2275085460194104e-06, + "loss": 0.3031, + "step": 35388 + }, + { + "epoch": 0.7789786484701738, + "grad_norm": 1.6393954479748205, + "learning_rate": 1.227274612567887e-06, + "loss": 0.2632, + "step": 35389 + }, + { + "epoch": 0.7790006603565925, + "grad_norm": 1.9739120159562011, + "learning_rate": 1.227040698291037e-06, + "loss": 0.2821, + "step": 35390 + }, + { + "epoch": 0.7790226722430112, + "grad_norm": 1.373877328845299, + "learning_rate": 1.2268068031900476e-06, + "loss": 0.3002, + "step": 35391 + }, + { + "epoch": 0.7790446841294298, + "grad_norm": 0.6299403534001292, + "learning_rate": 1.2265729272661104e-06, + "loss": 0.456, + "step": 35392 + }, + { + "epoch": 0.7790666960158485, + "grad_norm": 1.457112229937725, + "learning_rate": 1.2263390705204093e-06, + "loss": 0.254, + "step": 35393 + }, + { + "epoch": 0.7790887079022673, + "grad_norm": 1.5455661944334285, + "learning_rate": 1.2261052329541395e-06, + "loss": 0.3352, + "step": 35394 + }, + { + "epoch": 0.7791107197886858, + "grad_norm": 1.4844821069240521, + "learning_rate": 1.2258714145684858e-06, + "loss": 0.2952, + "step": 35395 + }, + { + "epoch": 0.7791327316751046, + "grad_norm": 0.627566782738431, + "learning_rate": 1.2256376153646353e-06, + "loss": 0.4604, + "step": 35396 + }, + { + "epoch": 0.7791547435615233, + "grad_norm": 1.4309319629416126, + "learning_rate": 1.2254038353437797e-06, + "loss": 0.2681, + "step": 35397 + }, + { + "epoch": 0.7791767554479418, + "grad_norm": 1.160543139608634, + "learning_rate": 1.2251700745071038e-06, + "loss": 0.2617, + "step": 35398 + }, + { + "epoch": 0.7791987673343606, + "grad_norm": 9.327418267902623, + "learning_rate": 1.2249363328557972e-06, + "loss": 0.3237, + "step": 35399 + }, + { + "epoch": 0.7792207792207793, + "grad_norm": 2.6518037542605937, + "learning_rate": 1.2247026103910491e-06, + "loss": 0.343, + "step": 35400 + }, + { + "epoch": 0.7792427911071979, + "grad_norm": 1.916990716696725, + "learning_rate": 1.2244689071140452e-06, + "loss": 0.3019, + "step": 35401 + }, + { + "epoch": 0.7792648029936166, + "grad_norm": 8.49414105038358, + "learning_rate": 1.2242352230259752e-06, + "loss": 0.2947, + "step": 35402 + }, + { + "epoch": 0.7792868148800353, + "grad_norm": 1.5026212486038326, + "learning_rate": 1.2240015581280246e-06, + "loss": 0.247, + "step": 35403 + }, + { + "epoch": 0.7793088267664539, + "grad_norm": 1.2560713985228433, + "learning_rate": 1.2237679124213818e-06, + "loss": 0.2216, + "step": 35404 + }, + { + "epoch": 0.7793308386528726, + "grad_norm": 1.3295335907566384, + "learning_rate": 1.2235342859072369e-06, + "loss": 0.3295, + "step": 35405 + }, + { + "epoch": 0.7793528505392913, + "grad_norm": 1.8509161197372477, + "learning_rate": 1.2233006785867735e-06, + "loss": 0.2684, + "step": 35406 + }, + { + "epoch": 0.7793748624257099, + "grad_norm": 1.7786042815635341, + "learning_rate": 1.2230670904611813e-06, + "loss": 0.3877, + "step": 35407 + }, + { + "epoch": 0.7793968743121286, + "grad_norm": 1.4604149032809983, + "learning_rate": 1.222833521531646e-06, + "loss": 0.3102, + "step": 35408 + }, + { + "epoch": 0.7794188861985473, + "grad_norm": 1.7518224160258073, + "learning_rate": 1.2225999717993553e-06, + "loss": 0.3377, + "step": 35409 + }, + { + "epoch": 0.7794408980849659, + "grad_norm": 2.619503366110015, + "learning_rate": 1.2223664412654972e-06, + "loss": 0.3106, + "step": 35410 + }, + { + "epoch": 0.7794629099713846, + "grad_norm": 2.458620033560933, + "learning_rate": 1.2221329299312569e-06, + "loss": 0.3026, + "step": 35411 + }, + { + "epoch": 0.7794849218578032, + "grad_norm": 0.6232015247505414, + "learning_rate": 1.2218994377978234e-06, + "loss": 0.4541, + "step": 35412 + }, + { + "epoch": 0.7795069337442219, + "grad_norm": 1.3528689894907815, + "learning_rate": 1.2216659648663803e-06, + "loss": 0.2714, + "step": 35413 + }, + { + "epoch": 0.7795289456306406, + "grad_norm": 1.4977994072560084, + "learning_rate": 1.2214325111381159e-06, + "loss": 0.2817, + "step": 35414 + }, + { + "epoch": 0.7795509575170592, + "grad_norm": 1.7016068109059044, + "learning_rate": 1.2211990766142184e-06, + "loss": 0.3337, + "step": 35415 + }, + { + "epoch": 0.7795729694034779, + "grad_norm": 1.6851753900868232, + "learning_rate": 1.2209656612958708e-06, + "loss": 0.3355, + "step": 35416 + }, + { + "epoch": 0.7795949812898966, + "grad_norm": 0.6198339687020723, + "learning_rate": 1.2207322651842634e-06, + "loss": 0.4473, + "step": 35417 + }, + { + "epoch": 0.7796169931763152, + "grad_norm": 1.652680539576411, + "learning_rate": 1.220498888280578e-06, + "loss": 0.2159, + "step": 35418 + }, + { + "epoch": 0.7796390050627339, + "grad_norm": 2.036924242548081, + "learning_rate": 1.2202655305860033e-06, + "loss": 0.4352, + "step": 35419 + }, + { + "epoch": 0.7796610169491526, + "grad_norm": 0.6163665218112365, + "learning_rate": 1.2200321921017266e-06, + "loss": 0.4396, + "step": 35420 + }, + { + "epoch": 0.7796830288355712, + "grad_norm": 4.085202969908622, + "learning_rate": 1.2197988728289307e-06, + "loss": 0.3426, + "step": 35421 + }, + { + "epoch": 0.7797050407219899, + "grad_norm": 2.2748794271156343, + "learning_rate": 1.2195655727688044e-06, + "loss": 0.2679, + "step": 35422 + }, + { + "epoch": 0.7797270526084086, + "grad_norm": 2.373954979446388, + "learning_rate": 1.2193322919225314e-06, + "loss": 0.2606, + "step": 35423 + }, + { + "epoch": 0.7797490644948272, + "grad_norm": 1.7502792728736292, + "learning_rate": 1.2190990302912959e-06, + "loss": 0.3211, + "step": 35424 + }, + { + "epoch": 0.7797710763812459, + "grad_norm": 1.6928367645464852, + "learning_rate": 1.2188657878762877e-06, + "loss": 0.2694, + "step": 35425 + }, + { + "epoch": 0.7797930882676646, + "grad_norm": 2.0393255395488654, + "learning_rate": 1.21863256467869e-06, + "loss": 0.2788, + "step": 35426 + }, + { + "epoch": 0.7798151001540832, + "grad_norm": 1.4632415201251707, + "learning_rate": 1.2183993606996874e-06, + "loss": 0.2583, + "step": 35427 + }, + { + "epoch": 0.7798371120405019, + "grad_norm": 1.8475262071770007, + "learning_rate": 1.2181661759404667e-06, + "loss": 0.3452, + "step": 35428 + }, + { + "epoch": 0.7798591239269206, + "grad_norm": 1.6830842560164232, + "learning_rate": 1.2179330104022097e-06, + "loss": 0.2679, + "step": 35429 + }, + { + "epoch": 0.7798811358133392, + "grad_norm": 1.5033615020961468, + "learning_rate": 1.2176998640861066e-06, + "loss": 0.3193, + "step": 35430 + }, + { + "epoch": 0.7799031476997579, + "grad_norm": 1.6523529269963437, + "learning_rate": 1.2174667369933397e-06, + "loss": 0.2741, + "step": 35431 + }, + { + "epoch": 0.7799251595861766, + "grad_norm": 1.5760332947222353, + "learning_rate": 1.2172336291250925e-06, + "loss": 0.2396, + "step": 35432 + }, + { + "epoch": 0.7799471714725952, + "grad_norm": 2.0616690608430273, + "learning_rate": 1.217000540482553e-06, + "loss": 0.2558, + "step": 35433 + }, + { + "epoch": 0.7799691833590139, + "grad_norm": 2.295943523225123, + "learning_rate": 1.216767471066902e-06, + "loss": 0.3174, + "step": 35434 + }, + { + "epoch": 0.7799911952454325, + "grad_norm": 1.3313716270254363, + "learning_rate": 1.2165344208793268e-06, + "loss": 0.2782, + "step": 35435 + }, + { + "epoch": 0.7800132071318512, + "grad_norm": 1.5911608913446398, + "learning_rate": 1.216301389921012e-06, + "loss": 0.2761, + "step": 35436 + }, + { + "epoch": 0.7800352190182699, + "grad_norm": 2.0551227449340543, + "learning_rate": 1.2160683781931398e-06, + "loss": 0.3078, + "step": 35437 + }, + { + "epoch": 0.7800572309046885, + "grad_norm": 1.718870746968596, + "learning_rate": 1.2158353856968974e-06, + "loss": 0.2659, + "step": 35438 + }, + { + "epoch": 0.7800792427911072, + "grad_norm": 1.347520864741806, + "learning_rate": 1.2156024124334654e-06, + "loss": 0.2741, + "step": 35439 + }, + { + "epoch": 0.7801012546775259, + "grad_norm": 1.6426433168114212, + "learning_rate": 1.2153694584040305e-06, + "loss": 0.3374, + "step": 35440 + }, + { + "epoch": 0.7801232665639445, + "grad_norm": 1.6210262162117457, + "learning_rate": 1.2151365236097767e-06, + "loss": 0.3058, + "step": 35441 + }, + { + "epoch": 0.7801452784503632, + "grad_norm": 1.4723183046358546, + "learning_rate": 1.214903608051886e-06, + "loss": 0.2825, + "step": 35442 + }, + { + "epoch": 0.7801672903367819, + "grad_norm": 2.5000720031202555, + "learning_rate": 1.2146707117315455e-06, + "loss": 0.3354, + "step": 35443 + }, + { + "epoch": 0.7801893022232005, + "grad_norm": 0.6166485745622173, + "learning_rate": 1.2144378346499347e-06, + "loss": 0.4652, + "step": 35444 + }, + { + "epoch": 0.7802113141096192, + "grad_norm": 1.4889741539494288, + "learning_rate": 1.2142049768082393e-06, + "loss": 0.2644, + "step": 35445 + }, + { + "epoch": 0.7802333259960379, + "grad_norm": 1.6996727695878846, + "learning_rate": 1.213972138207644e-06, + "loss": 0.2634, + "step": 35446 + }, + { + "epoch": 0.7802553378824565, + "grad_norm": 1.4273102114142586, + "learning_rate": 1.2137393188493297e-06, + "loss": 0.2408, + "step": 35447 + }, + { + "epoch": 0.7802773497688752, + "grad_norm": 1.6902934265831149, + "learning_rate": 1.2135065187344824e-06, + "loss": 0.273, + "step": 35448 + }, + { + "epoch": 0.7802993616552939, + "grad_norm": 1.4300421309499418, + "learning_rate": 1.2132737378642824e-06, + "loss": 0.2995, + "step": 35449 + }, + { + "epoch": 0.7803213735417125, + "grad_norm": 1.8217377519620293, + "learning_rate": 1.2130409762399137e-06, + "loss": 0.3222, + "step": 35450 + }, + { + "epoch": 0.7803433854281312, + "grad_norm": 1.813457065784005, + "learning_rate": 1.2128082338625618e-06, + "loss": 0.3183, + "step": 35451 + }, + { + "epoch": 0.7803653973145499, + "grad_norm": 1.563798088761508, + "learning_rate": 1.212575510733406e-06, + "loss": 0.2805, + "step": 35452 + }, + { + "epoch": 0.7803874092009685, + "grad_norm": 2.889692135483112, + "learning_rate": 1.2123428068536318e-06, + "loss": 0.3222, + "step": 35453 + }, + { + "epoch": 0.7804094210873872, + "grad_norm": 1.5719492598156093, + "learning_rate": 1.2121101222244198e-06, + "loss": 0.2592, + "step": 35454 + }, + { + "epoch": 0.7804314329738059, + "grad_norm": 1.512368652217027, + "learning_rate": 1.2118774568469533e-06, + "loss": 0.2411, + "step": 35455 + }, + { + "epoch": 0.7804534448602245, + "grad_norm": 1.7754544458949288, + "learning_rate": 1.2116448107224165e-06, + "loss": 0.3428, + "step": 35456 + }, + { + "epoch": 0.7804754567466432, + "grad_norm": 2.4796997764014184, + "learning_rate": 1.2114121838519894e-06, + "loss": 0.3725, + "step": 35457 + }, + { + "epoch": 0.7804974686330619, + "grad_norm": 2.005738431978751, + "learning_rate": 1.2111795762368566e-06, + "loss": 0.3487, + "step": 35458 + }, + { + "epoch": 0.7805194805194805, + "grad_norm": 2.052125494378085, + "learning_rate": 1.2109469878781987e-06, + "loss": 0.3085, + "step": 35459 + }, + { + "epoch": 0.7805414924058992, + "grad_norm": 0.6258851403193938, + "learning_rate": 1.2107144187771958e-06, + "loss": 0.477, + "step": 35460 + }, + { + "epoch": 0.7805635042923178, + "grad_norm": 1.7291841157526138, + "learning_rate": 1.210481868935035e-06, + "loss": 0.2898, + "step": 35461 + }, + { + "epoch": 0.7805855161787365, + "grad_norm": 0.6180653018794381, + "learning_rate": 1.210249338352894e-06, + "loss": 0.4432, + "step": 35462 + }, + { + "epoch": 0.7806075280651552, + "grad_norm": 1.9681974277069427, + "learning_rate": 1.2100168270319573e-06, + "loss": 0.2732, + "step": 35463 + }, + { + "epoch": 0.7806295399515738, + "grad_norm": 0.6432488528894559, + "learning_rate": 1.2097843349734056e-06, + "loss": 0.4527, + "step": 35464 + }, + { + "epoch": 0.7806515518379925, + "grad_norm": 1.6820005946692782, + "learning_rate": 1.2095518621784191e-06, + "loss": 0.3394, + "step": 35465 + }, + { + "epoch": 0.7806735637244112, + "grad_norm": 1.5761112382370197, + "learning_rate": 1.2093194086481808e-06, + "loss": 0.2311, + "step": 35466 + }, + { + "epoch": 0.7806955756108298, + "grad_norm": 1.5409138556009838, + "learning_rate": 1.2090869743838734e-06, + "loss": 0.2344, + "step": 35467 + }, + { + "epoch": 0.7807175874972485, + "grad_norm": 0.6167403904844952, + "learning_rate": 1.2088545593866752e-06, + "loss": 0.4604, + "step": 35468 + }, + { + "epoch": 0.7807395993836672, + "grad_norm": 1.2475322259722486, + "learning_rate": 1.2086221636577706e-06, + "loss": 0.2512, + "step": 35469 + }, + { + "epoch": 0.7807616112700858, + "grad_norm": 1.3431298045127649, + "learning_rate": 1.2083897871983375e-06, + "loss": 0.1839, + "step": 35470 + }, + { + "epoch": 0.7807836231565045, + "grad_norm": 1.4495176765269366, + "learning_rate": 1.2081574300095582e-06, + "loss": 0.2654, + "step": 35471 + }, + { + "epoch": 0.7808056350429232, + "grad_norm": 1.5580483504885763, + "learning_rate": 1.2079250920926161e-06, + "loss": 0.3184, + "step": 35472 + }, + { + "epoch": 0.7808276469293418, + "grad_norm": 1.8183277804921931, + "learning_rate": 1.2076927734486886e-06, + "loss": 0.2815, + "step": 35473 + }, + { + "epoch": 0.7808496588157605, + "grad_norm": 0.6040798804387822, + "learning_rate": 1.2074604740789592e-06, + "loss": 0.4342, + "step": 35474 + }, + { + "epoch": 0.7808716707021792, + "grad_norm": 1.568433102898662, + "learning_rate": 1.2072281939846059e-06, + "loss": 0.2539, + "step": 35475 + }, + { + "epoch": 0.7808936825885978, + "grad_norm": 1.654736865967395, + "learning_rate": 1.2069959331668102e-06, + "loss": 0.3205, + "step": 35476 + }, + { + "epoch": 0.7809156944750165, + "grad_norm": 1.7173372514726692, + "learning_rate": 1.2067636916267543e-06, + "loss": 0.2351, + "step": 35477 + }, + { + "epoch": 0.7809377063614352, + "grad_norm": 1.3025485150934255, + "learning_rate": 1.206531469365616e-06, + "loss": 0.2608, + "step": 35478 + }, + { + "epoch": 0.7809597182478538, + "grad_norm": 2.46464530299925, + "learning_rate": 1.206299266384578e-06, + "loss": 0.3807, + "step": 35479 + }, + { + "epoch": 0.7809817301342725, + "grad_norm": 1.5932319203050622, + "learning_rate": 1.2060670826848175e-06, + "loss": 0.3229, + "step": 35480 + }, + { + "epoch": 0.7810037420206912, + "grad_norm": 4.445995451733636, + "learning_rate": 1.2058349182675167e-06, + "loss": 0.2821, + "step": 35481 + }, + { + "epoch": 0.7810257539071098, + "grad_norm": 1.967411039645366, + "learning_rate": 1.205602773133856e-06, + "loss": 0.2903, + "step": 35482 + }, + { + "epoch": 0.7810477657935285, + "grad_norm": 1.809360600442422, + "learning_rate": 1.2053706472850135e-06, + "loss": 0.2403, + "step": 35483 + }, + { + "epoch": 0.7810697776799471, + "grad_norm": 1.746864537319629, + "learning_rate": 1.2051385407221705e-06, + "loss": 0.2127, + "step": 35484 + }, + { + "epoch": 0.7810917895663658, + "grad_norm": 1.6522339827144739, + "learning_rate": 1.2049064534465055e-06, + "loss": 0.3093, + "step": 35485 + }, + { + "epoch": 0.7811138014527845, + "grad_norm": 1.4122045757174924, + "learning_rate": 1.204674385459198e-06, + "loss": 0.3426, + "step": 35486 + }, + { + "epoch": 0.7811358133392031, + "grad_norm": 1.6251852007890197, + "learning_rate": 1.2044423367614294e-06, + "loss": 0.3075, + "step": 35487 + }, + { + "epoch": 0.7811578252256218, + "grad_norm": 0.6110515541179932, + "learning_rate": 1.2042103073543764e-06, + "loss": 0.4382, + "step": 35488 + }, + { + "epoch": 0.7811798371120405, + "grad_norm": 3.7128088877065295, + "learning_rate": 1.2039782972392211e-06, + "loss": 0.2738, + "step": 35489 + }, + { + "epoch": 0.7812018489984591, + "grad_norm": 0.6264921043680283, + "learning_rate": 1.2037463064171394e-06, + "loss": 0.4483, + "step": 35490 + }, + { + "epoch": 0.7812238608848778, + "grad_norm": 1.65782657838542, + "learning_rate": 1.2035143348893125e-06, + "loss": 0.3022, + "step": 35491 + }, + { + "epoch": 0.7812458727712965, + "grad_norm": 1.6333084100483237, + "learning_rate": 1.2032823826569207e-06, + "loss": 0.2822, + "step": 35492 + }, + { + "epoch": 0.7812678846577151, + "grad_norm": 0.6226130957466217, + "learning_rate": 1.2030504497211392e-06, + "loss": 0.4478, + "step": 35493 + }, + { + "epoch": 0.7812898965441338, + "grad_norm": 1.351174797504681, + "learning_rate": 1.2028185360831507e-06, + "loss": 0.3108, + "step": 35494 + }, + { + "epoch": 0.7813119084305525, + "grad_norm": 1.3496392601188485, + "learning_rate": 1.2025866417441319e-06, + "loss": 0.2752, + "step": 35495 + }, + { + "epoch": 0.7813339203169711, + "grad_norm": 1.5840251501260685, + "learning_rate": 1.2023547667052582e-06, + "loss": 0.2711, + "step": 35496 + }, + { + "epoch": 0.7813559322033898, + "grad_norm": 0.6260053048109547, + "learning_rate": 1.2021229109677146e-06, + "loss": 0.4401, + "step": 35497 + }, + { + "epoch": 0.7813779440898085, + "grad_norm": 0.6228511949930271, + "learning_rate": 1.201891074532675e-06, + "loss": 0.4698, + "step": 35498 + }, + { + "epoch": 0.7813999559762271, + "grad_norm": 0.6271504221618346, + "learning_rate": 1.2016592574013197e-06, + "loss": 0.4556, + "step": 35499 + }, + { + "epoch": 0.7814219678626458, + "grad_norm": 1.601592648819067, + "learning_rate": 1.201427459574827e-06, + "loss": 0.2992, + "step": 35500 + }, + { + "epoch": 0.7814439797490645, + "grad_norm": 1.9800161637623772, + "learning_rate": 1.2011956810543722e-06, + "loss": 0.3032, + "step": 35501 + }, + { + "epoch": 0.7814659916354831, + "grad_norm": 1.7913367547661345, + "learning_rate": 1.2009639218411356e-06, + "loss": 0.3339, + "step": 35502 + }, + { + "epoch": 0.7814880035219018, + "grad_norm": 1.3354019441978968, + "learning_rate": 1.200732181936296e-06, + "loss": 0.3358, + "step": 35503 + }, + { + "epoch": 0.7815100154083205, + "grad_norm": 2.04425222131193, + "learning_rate": 1.2005004613410287e-06, + "loss": 0.2812, + "step": 35504 + }, + { + "epoch": 0.7815320272947391, + "grad_norm": 4.668116522683587, + "learning_rate": 1.2002687600565138e-06, + "loss": 0.334, + "step": 35505 + }, + { + "epoch": 0.7815540391811578, + "grad_norm": 7.024201234960678, + "learning_rate": 1.2000370780839265e-06, + "loss": 0.284, + "step": 35506 + }, + { + "epoch": 0.7815760510675764, + "grad_norm": 41.16797853389083, + "learning_rate": 1.1998054154244454e-06, + "loss": 0.2741, + "step": 35507 + }, + { + "epoch": 0.7815980629539951, + "grad_norm": 1.6817021310546014, + "learning_rate": 1.19957377207925e-06, + "loss": 0.252, + "step": 35508 + }, + { + "epoch": 0.7816200748404138, + "grad_norm": 2.471560842673245, + "learning_rate": 1.1993421480495143e-06, + "loss": 0.3078, + "step": 35509 + }, + { + "epoch": 0.7816420867268324, + "grad_norm": 1.7181303947510578, + "learning_rate": 1.199110543336418e-06, + "loss": 0.2886, + "step": 35510 + }, + { + "epoch": 0.7816640986132511, + "grad_norm": 0.612967959237124, + "learning_rate": 1.1988789579411358e-06, + "loss": 0.4508, + "step": 35511 + }, + { + "epoch": 0.7816861104996699, + "grad_norm": 3.4497066309868427, + "learning_rate": 1.1986473918648462e-06, + "loss": 0.2539, + "step": 35512 + }, + { + "epoch": 0.7817081223860884, + "grad_norm": 1.452903342929818, + "learning_rate": 1.1984158451087275e-06, + "loss": 0.306, + "step": 35513 + }, + { + "epoch": 0.7817301342725071, + "grad_norm": 1.6218540336465146, + "learning_rate": 1.1981843176739532e-06, + "loss": 0.3708, + "step": 35514 + }, + { + "epoch": 0.7817521461589259, + "grad_norm": 2.033555964855303, + "learning_rate": 1.1979528095617038e-06, + "loss": 0.2847, + "step": 35515 + }, + { + "epoch": 0.7817741580453444, + "grad_norm": 1.40171925431269, + "learning_rate": 1.1977213207731524e-06, + "loss": 0.3081, + "step": 35516 + }, + { + "epoch": 0.7817961699317632, + "grad_norm": 1.2812815988617687, + "learning_rate": 1.1974898513094767e-06, + "loss": 0.2956, + "step": 35517 + }, + { + "epoch": 0.7818181818181819, + "grad_norm": 4.657846153300192, + "learning_rate": 1.1972584011718557e-06, + "loss": 0.3415, + "step": 35518 + }, + { + "epoch": 0.7818401937046005, + "grad_norm": 1.6101178676751033, + "learning_rate": 1.1970269703614617e-06, + "loss": 0.304, + "step": 35519 + }, + { + "epoch": 0.7818622055910192, + "grad_norm": 1.7965814055230005, + "learning_rate": 1.1967955588794744e-06, + "loss": 0.3122, + "step": 35520 + }, + { + "epoch": 0.7818842174774379, + "grad_norm": 1.4801132212441896, + "learning_rate": 1.1965641667270668e-06, + "loss": 0.3148, + "step": 35521 + }, + { + "epoch": 0.7819062293638565, + "grad_norm": 1.7845136639316421, + "learning_rate": 1.1963327939054165e-06, + "loss": 0.2323, + "step": 35522 + }, + { + "epoch": 0.7819282412502752, + "grad_norm": 1.3265859515442309, + "learning_rate": 1.1961014404157007e-06, + "loss": 0.2798, + "step": 35523 + }, + { + "epoch": 0.7819502531366939, + "grad_norm": 1.7507691967241525, + "learning_rate": 1.1958701062590928e-06, + "loss": 0.2482, + "step": 35524 + }, + { + "epoch": 0.7819722650231125, + "grad_norm": 2.1084986266898533, + "learning_rate": 1.195638791436771e-06, + "loss": 0.3733, + "step": 35525 + }, + { + "epoch": 0.7819942769095312, + "grad_norm": 1.5569648019196562, + "learning_rate": 1.1954074959499095e-06, + "loss": 0.3755, + "step": 35526 + }, + { + "epoch": 0.7820162887959499, + "grad_norm": 1.2673585444984865, + "learning_rate": 1.1951762197996813e-06, + "loss": 0.2536, + "step": 35527 + }, + { + "epoch": 0.7820383006823685, + "grad_norm": 3.2363952144346095, + "learning_rate": 1.1949449629872678e-06, + "loss": 0.3073, + "step": 35528 + }, + { + "epoch": 0.7820603125687872, + "grad_norm": 1.8797002014464512, + "learning_rate": 1.194713725513839e-06, + "loss": 0.2718, + "step": 35529 + }, + { + "epoch": 0.7820823244552058, + "grad_norm": 0.6153769741582713, + "learning_rate": 1.1944825073805744e-06, + "loss": 0.4533, + "step": 35530 + }, + { + "epoch": 0.7821043363416245, + "grad_norm": 1.9314506980078454, + "learning_rate": 1.1942513085886465e-06, + "loss": 0.2887, + "step": 35531 + }, + { + "epoch": 0.7821263482280432, + "grad_norm": 5.40020452217396, + "learning_rate": 1.1940201291392278e-06, + "loss": 0.3269, + "step": 35532 + }, + { + "epoch": 0.7821483601144618, + "grad_norm": 2.045420988034736, + "learning_rate": 1.1937889690334997e-06, + "loss": 0.2732, + "step": 35533 + }, + { + "epoch": 0.7821703720008805, + "grad_norm": 1.7841402829823998, + "learning_rate": 1.1935578282726318e-06, + "loss": 0.265, + "step": 35534 + }, + { + "epoch": 0.7821923838872992, + "grad_norm": 1.6603682217631188, + "learning_rate": 1.1933267068578025e-06, + "loss": 0.3115, + "step": 35535 + }, + { + "epoch": 0.7822143957737178, + "grad_norm": 1.743945780520195, + "learning_rate": 1.1930956047901848e-06, + "loss": 0.3254, + "step": 35536 + }, + { + "epoch": 0.7822364076601365, + "grad_norm": 1.570634094006006, + "learning_rate": 1.1928645220709512e-06, + "loss": 0.2333, + "step": 35537 + }, + { + "epoch": 0.7822584195465552, + "grad_norm": 1.741107185768304, + "learning_rate": 1.1926334587012783e-06, + "loss": 0.2327, + "step": 35538 + }, + { + "epoch": 0.7822804314329738, + "grad_norm": 1.7438201081126423, + "learning_rate": 1.192402414682341e-06, + "loss": 0.2133, + "step": 35539 + }, + { + "epoch": 0.7823024433193925, + "grad_norm": 2.058918435519642, + "learning_rate": 1.1921713900153137e-06, + "loss": 0.2758, + "step": 35540 + }, + { + "epoch": 0.7823244552058112, + "grad_norm": 1.5873182102717731, + "learning_rate": 1.1919403847013695e-06, + "loss": 0.3012, + "step": 35541 + }, + { + "epoch": 0.7823464670922298, + "grad_norm": 1.679086150340698, + "learning_rate": 1.1917093987416817e-06, + "loss": 0.3217, + "step": 35542 + }, + { + "epoch": 0.7823684789786485, + "grad_norm": 1.145074601875078, + "learning_rate": 1.191478432137425e-06, + "loss": 0.226, + "step": 35543 + }, + { + "epoch": 0.7823904908650672, + "grad_norm": 1.701960783320234, + "learning_rate": 1.1912474848897753e-06, + "loss": 0.3261, + "step": 35544 + }, + { + "epoch": 0.7824125027514858, + "grad_norm": 1.9996906853998186, + "learning_rate": 1.1910165569999027e-06, + "loss": 0.2876, + "step": 35545 + }, + { + "epoch": 0.7824345146379045, + "grad_norm": 1.736165436210731, + "learning_rate": 1.1907856484689851e-06, + "loss": 0.2969, + "step": 35546 + }, + { + "epoch": 0.7824565265243232, + "grad_norm": 0.6125507364554341, + "learning_rate": 1.1905547592981914e-06, + "loss": 0.4508, + "step": 35547 + }, + { + "epoch": 0.7824785384107418, + "grad_norm": 0.6049016158242597, + "learning_rate": 1.1903238894886982e-06, + "loss": 0.4376, + "step": 35548 + }, + { + "epoch": 0.7825005502971605, + "grad_norm": 1.7973416396105404, + "learning_rate": 1.1900930390416792e-06, + "loss": 0.2298, + "step": 35549 + }, + { + "epoch": 0.7825225621835792, + "grad_norm": 1.936418018234741, + "learning_rate": 1.189862207958305e-06, + "loss": 0.3637, + "step": 35550 + }, + { + "epoch": 0.7825445740699978, + "grad_norm": 1.3453503965635445, + "learning_rate": 1.189631396239752e-06, + "loss": 0.2654, + "step": 35551 + }, + { + "epoch": 0.7825665859564165, + "grad_norm": 1.2576197536926799, + "learning_rate": 1.1894006038871907e-06, + "loss": 0.2224, + "step": 35552 + }, + { + "epoch": 0.7825885978428351, + "grad_norm": 3.9693018391451997, + "learning_rate": 1.1891698309017951e-06, + "loss": 0.321, + "step": 35553 + }, + { + "epoch": 0.7826106097292538, + "grad_norm": 1.656343090615889, + "learning_rate": 1.1889390772847392e-06, + "loss": 0.2521, + "step": 35554 + }, + { + "epoch": 0.7826326216156725, + "grad_norm": 0.5990964928039061, + "learning_rate": 1.1887083430371932e-06, + "loss": 0.4436, + "step": 35555 + }, + { + "epoch": 0.7826546335020911, + "grad_norm": 1.4262541914528455, + "learning_rate": 1.1884776281603328e-06, + "loss": 0.3359, + "step": 35556 + }, + { + "epoch": 0.7826766453885098, + "grad_norm": 1.9331994482802877, + "learning_rate": 1.1882469326553281e-06, + "loss": 0.3436, + "step": 35557 + }, + { + "epoch": 0.7826986572749285, + "grad_norm": 1.6641787111818602, + "learning_rate": 1.1880162565233522e-06, + "loss": 0.3514, + "step": 35558 + }, + { + "epoch": 0.7827206691613471, + "grad_norm": 1.8751024620331211, + "learning_rate": 1.1877855997655791e-06, + "loss": 0.3832, + "step": 35559 + }, + { + "epoch": 0.7827426810477658, + "grad_norm": 1.4079759885550043, + "learning_rate": 1.1875549623831784e-06, + "loss": 0.2594, + "step": 35560 + }, + { + "epoch": 0.7827646929341845, + "grad_norm": 1.4471547666880336, + "learning_rate": 1.1873243443773257e-06, + "loss": 0.271, + "step": 35561 + }, + { + "epoch": 0.7827867048206031, + "grad_norm": 0.6240759187689451, + "learning_rate": 1.1870937457491893e-06, + "loss": 0.4412, + "step": 35562 + }, + { + "epoch": 0.7828087167070218, + "grad_norm": 5.708856400592146, + "learning_rate": 1.1868631664999436e-06, + "loss": 0.25, + "step": 35563 + }, + { + "epoch": 0.7828307285934405, + "grad_norm": 0.6133935907684553, + "learning_rate": 1.186632606630761e-06, + "loss": 0.4623, + "step": 35564 + }, + { + "epoch": 0.7828527404798591, + "grad_norm": 1.3140387447802095, + "learning_rate": 1.186402066142811e-06, + "loss": 0.2996, + "step": 35565 + }, + { + "epoch": 0.7828747523662778, + "grad_norm": 1.4223693377519528, + "learning_rate": 1.186171545037268e-06, + "loss": 0.2807, + "step": 35566 + }, + { + "epoch": 0.7828967642526965, + "grad_norm": 1.5138681817114548, + "learning_rate": 1.1859410433153024e-06, + "loss": 0.3825, + "step": 35567 + }, + { + "epoch": 0.7829187761391151, + "grad_norm": 1.6199346736039617, + "learning_rate": 1.1857105609780822e-06, + "loss": 0.2772, + "step": 35568 + }, + { + "epoch": 0.7829407880255338, + "grad_norm": 1.3374109721903948, + "learning_rate": 1.1854800980267854e-06, + "loss": 0.2609, + "step": 35569 + }, + { + "epoch": 0.7829627999119525, + "grad_norm": 3.0837830938902138, + "learning_rate": 1.1852496544625785e-06, + "loss": 0.292, + "step": 35570 + }, + { + "epoch": 0.7829848117983711, + "grad_norm": 2.6481858005735326, + "learning_rate": 1.185019230286636e-06, + "loss": 0.2387, + "step": 35571 + }, + { + "epoch": 0.7830068236847898, + "grad_norm": 1.7924183428872502, + "learning_rate": 1.1847888255001267e-06, + "loss": 0.3157, + "step": 35572 + }, + { + "epoch": 0.7830288355712085, + "grad_norm": 1.4166359043692218, + "learning_rate": 1.1845584401042215e-06, + "loss": 0.2402, + "step": 35573 + }, + { + "epoch": 0.7830508474576271, + "grad_norm": 1.712085987036455, + "learning_rate": 1.1843280741000918e-06, + "loss": 0.3042, + "step": 35574 + }, + { + "epoch": 0.7830728593440458, + "grad_norm": 1.2283620544424163, + "learning_rate": 1.1840977274889082e-06, + "loss": 0.2923, + "step": 35575 + }, + { + "epoch": 0.7830948712304644, + "grad_norm": 1.8023776823480417, + "learning_rate": 1.1838674002718437e-06, + "loss": 0.2812, + "step": 35576 + }, + { + "epoch": 0.7831168831168831, + "grad_norm": 1.4950681088625795, + "learning_rate": 1.1836370924500672e-06, + "loss": 0.3415, + "step": 35577 + }, + { + "epoch": 0.7831388950033018, + "grad_norm": 1.1977186641225515, + "learning_rate": 1.1834068040247476e-06, + "loss": 0.2137, + "step": 35578 + }, + { + "epoch": 0.7831609068897204, + "grad_norm": 1.9513677857265275, + "learning_rate": 1.183176534997057e-06, + "loss": 0.3134, + "step": 35579 + }, + { + "epoch": 0.7831829187761391, + "grad_norm": 0.5907111543089106, + "learning_rate": 1.1829462853681656e-06, + "loss": 0.4335, + "step": 35580 + }, + { + "epoch": 0.7832049306625578, + "grad_norm": 2.1192972172865017, + "learning_rate": 1.1827160551392446e-06, + "loss": 0.3805, + "step": 35581 + }, + { + "epoch": 0.7832269425489764, + "grad_norm": 1.6723708628933842, + "learning_rate": 1.182485844311464e-06, + "loss": 0.2901, + "step": 35582 + }, + { + "epoch": 0.7832489544353951, + "grad_norm": 0.6208800649316704, + "learning_rate": 1.1822556528859907e-06, + "loss": 0.4565, + "step": 35583 + }, + { + "epoch": 0.7832709663218138, + "grad_norm": 1.566528909854538, + "learning_rate": 1.1820254808639975e-06, + "loss": 0.2615, + "step": 35584 + }, + { + "epoch": 0.7832929782082324, + "grad_norm": 2.0957935490725577, + "learning_rate": 1.1817953282466548e-06, + "loss": 0.3546, + "step": 35585 + }, + { + "epoch": 0.7833149900946511, + "grad_norm": 1.5190614072924218, + "learning_rate": 1.1815651950351297e-06, + "loss": 0.2782, + "step": 35586 + }, + { + "epoch": 0.7833370019810698, + "grad_norm": 1.2580357733947996, + "learning_rate": 1.181335081230595e-06, + "loss": 0.2173, + "step": 35587 + }, + { + "epoch": 0.7833590138674884, + "grad_norm": 0.6119011662739569, + "learning_rate": 1.1811049868342172e-06, + "loss": 0.4477, + "step": 35588 + }, + { + "epoch": 0.7833810257539071, + "grad_norm": 1.6677246218011037, + "learning_rate": 1.1808749118471673e-06, + "loss": 0.3349, + "step": 35589 + }, + { + "epoch": 0.7834030376403258, + "grad_norm": 1.746648881216816, + "learning_rate": 1.1806448562706159e-06, + "loss": 0.2361, + "step": 35590 + }, + { + "epoch": 0.7834250495267444, + "grad_norm": 4.044914703659802, + "learning_rate": 1.1804148201057287e-06, + "loss": 0.3153, + "step": 35591 + }, + { + "epoch": 0.7834470614131631, + "grad_norm": 2.892514299726111, + "learning_rate": 1.180184803353679e-06, + "loss": 0.2799, + "step": 35592 + }, + { + "epoch": 0.7834690732995818, + "grad_norm": 6.901454230045959, + "learning_rate": 1.1799548060156323e-06, + "loss": 0.3373, + "step": 35593 + }, + { + "epoch": 0.7834910851860004, + "grad_norm": 0.6172756659656268, + "learning_rate": 1.1797248280927588e-06, + "loss": 0.4278, + "step": 35594 + }, + { + "epoch": 0.7835130970724191, + "grad_norm": 2.0728551118882548, + "learning_rate": 1.1794948695862296e-06, + "loss": 0.3177, + "step": 35595 + }, + { + "epoch": 0.7835351089588378, + "grad_norm": 1.6945974653510067, + "learning_rate": 1.1792649304972093e-06, + "loss": 0.2075, + "step": 35596 + }, + { + "epoch": 0.7835571208452564, + "grad_norm": 1.692662607066568, + "learning_rate": 1.1790350108268706e-06, + "loss": 0.2665, + "step": 35597 + }, + { + "epoch": 0.7835791327316751, + "grad_norm": 0.6632795653202131, + "learning_rate": 1.1788051105763788e-06, + "loss": 0.4689, + "step": 35598 + }, + { + "epoch": 0.7836011446180938, + "grad_norm": 2.202610650106892, + "learning_rate": 1.1785752297469038e-06, + "loss": 0.321, + "step": 35599 + }, + { + "epoch": 0.7836231565045124, + "grad_norm": 1.4170795773591491, + "learning_rate": 1.1783453683396152e-06, + "loss": 0.2851, + "step": 35600 + }, + { + "epoch": 0.7836451683909311, + "grad_norm": 1.4079100487645821, + "learning_rate": 1.1781155263556782e-06, + "loss": 0.2905, + "step": 35601 + }, + { + "epoch": 0.7836671802773497, + "grad_norm": 1.8791114444643011, + "learning_rate": 1.1778857037962642e-06, + "loss": 0.3169, + "step": 35602 + }, + { + "epoch": 0.7836891921637684, + "grad_norm": 1.7531470618694676, + "learning_rate": 1.17765590066254e-06, + "loss": 0.2103, + "step": 35603 + }, + { + "epoch": 0.7837112040501871, + "grad_norm": 2.204507611640993, + "learning_rate": 1.1774261169556704e-06, + "loss": 0.3973, + "step": 35604 + }, + { + "epoch": 0.7837332159366057, + "grad_norm": 1.282344146237546, + "learning_rate": 1.1771963526768293e-06, + "loss": 0.2748, + "step": 35605 + }, + { + "epoch": 0.7837552278230244, + "grad_norm": 1.3953849327038794, + "learning_rate": 1.1769666078271796e-06, + "loss": 0.2786, + "step": 35606 + }, + { + "epoch": 0.7837772397094431, + "grad_norm": 0.6275158523264057, + "learning_rate": 1.1767368824078924e-06, + "loss": 0.4276, + "step": 35607 + }, + { + "epoch": 0.7837992515958617, + "grad_norm": 1.5267944228917163, + "learning_rate": 1.176507176420133e-06, + "loss": 0.1954, + "step": 35608 + }, + { + "epoch": 0.7838212634822804, + "grad_norm": 2.559052323757463, + "learning_rate": 1.1762774898650687e-06, + "loss": 0.2639, + "step": 35609 + }, + { + "epoch": 0.7838432753686991, + "grad_norm": 2.0069342497706053, + "learning_rate": 1.1760478227438672e-06, + "loss": 0.2746, + "step": 35610 + }, + { + "epoch": 0.7838652872551177, + "grad_norm": 3.4710242611858053, + "learning_rate": 1.1758181750576963e-06, + "loss": 0.188, + "step": 35611 + }, + { + "epoch": 0.7838872991415364, + "grad_norm": 1.7327937630738266, + "learning_rate": 1.1755885468077245e-06, + "loss": 0.2848, + "step": 35612 + }, + { + "epoch": 0.7839093110279551, + "grad_norm": 2.1644478983069444, + "learning_rate": 1.1753589379951174e-06, + "loss": 0.329, + "step": 35613 + }, + { + "epoch": 0.7839313229143737, + "grad_norm": 1.5325088136686347, + "learning_rate": 1.1751293486210409e-06, + "loss": 0.2434, + "step": 35614 + }, + { + "epoch": 0.7839533348007924, + "grad_norm": 1.7055608534386966, + "learning_rate": 1.1748997786866627e-06, + "loss": 0.2604, + "step": 35615 + }, + { + "epoch": 0.7839753466872111, + "grad_norm": 0.635157034605723, + "learning_rate": 1.17467022819315e-06, + "loss": 0.467, + "step": 35616 + }, + { + "epoch": 0.7839973585736297, + "grad_norm": 2.745801895068076, + "learning_rate": 1.1744406971416706e-06, + "loss": 0.3681, + "step": 35617 + }, + { + "epoch": 0.7840193704600484, + "grad_norm": 1.7409272036611758, + "learning_rate": 1.17421118553339e-06, + "loss": 0.3554, + "step": 35618 + }, + { + "epoch": 0.7840413823464671, + "grad_norm": 1.5933072889226096, + "learning_rate": 1.1739816933694731e-06, + "loss": 0.2647, + "step": 35619 + }, + { + "epoch": 0.7840633942328857, + "grad_norm": 1.4963891862849783, + "learning_rate": 1.1737522206510875e-06, + "loss": 0.2679, + "step": 35620 + }, + { + "epoch": 0.7840854061193044, + "grad_norm": 5.1511140891751115, + "learning_rate": 1.1735227673793998e-06, + "loss": 0.2269, + "step": 35621 + }, + { + "epoch": 0.7841074180057231, + "grad_norm": 0.6148709638084056, + "learning_rate": 1.1732933335555775e-06, + "loss": 0.4468, + "step": 35622 + }, + { + "epoch": 0.7841294298921417, + "grad_norm": 1.9168214507016341, + "learning_rate": 1.1730639191807851e-06, + "loss": 0.2392, + "step": 35623 + }, + { + "epoch": 0.7841514417785604, + "grad_norm": 1.9283137381666096, + "learning_rate": 1.1728345242561873e-06, + "loss": 0.3043, + "step": 35624 + }, + { + "epoch": 0.784173453664979, + "grad_norm": 1.5743400400011922, + "learning_rate": 1.1726051487829516e-06, + "loss": 0.3169, + "step": 35625 + }, + { + "epoch": 0.7841954655513977, + "grad_norm": 0.6291359041330034, + "learning_rate": 1.1723757927622448e-06, + "loss": 0.455, + "step": 35626 + }, + { + "epoch": 0.7842174774378164, + "grad_norm": 1.4800499687167872, + "learning_rate": 1.17214645619523e-06, + "loss": 0.2407, + "step": 35627 + }, + { + "epoch": 0.784239489324235, + "grad_norm": 1.7612647514924837, + "learning_rate": 1.171917139083076e-06, + "loss": 0.2665, + "step": 35628 + }, + { + "epoch": 0.7842615012106537, + "grad_norm": 2.2820106071669257, + "learning_rate": 1.1716878414269444e-06, + "loss": 0.3321, + "step": 35629 + }, + { + "epoch": 0.7842835130970724, + "grad_norm": 1.4100052792781004, + "learning_rate": 1.171458563228003e-06, + "loss": 0.2893, + "step": 35630 + }, + { + "epoch": 0.784305524983491, + "grad_norm": 1.328656396433905, + "learning_rate": 1.1712293044874184e-06, + "loss": 0.2209, + "step": 35631 + }, + { + "epoch": 0.7843275368699097, + "grad_norm": 1.291132649714274, + "learning_rate": 1.1710000652063525e-06, + "loss": 0.261, + "step": 35632 + }, + { + "epoch": 0.7843495487563285, + "grad_norm": 2.080417236595368, + "learning_rate": 1.1707708453859733e-06, + "loss": 0.2604, + "step": 35633 + }, + { + "epoch": 0.784371560642747, + "grad_norm": 1.6989637732980827, + "learning_rate": 1.1705416450274433e-06, + "loss": 0.3052, + "step": 35634 + }, + { + "epoch": 0.7843935725291658, + "grad_norm": 2.0758495636148537, + "learning_rate": 1.1703124641319285e-06, + "loss": 0.2806, + "step": 35635 + }, + { + "epoch": 0.7844155844155845, + "grad_norm": 1.5454418945277788, + "learning_rate": 1.1700833027005958e-06, + "loss": 0.3455, + "step": 35636 + }, + { + "epoch": 0.784437596302003, + "grad_norm": 2.212404050667598, + "learning_rate": 1.1698541607346058e-06, + "loss": 0.273, + "step": 35637 + }, + { + "epoch": 0.7844596081884218, + "grad_norm": 2.075689342818325, + "learning_rate": 1.169625038235127e-06, + "loss": 0.2762, + "step": 35638 + }, + { + "epoch": 0.7844816200748405, + "grad_norm": 1.6217589378500568, + "learning_rate": 1.1693959352033207e-06, + "loss": 0.2642, + "step": 35639 + }, + { + "epoch": 0.784503631961259, + "grad_norm": 1.5337143986421542, + "learning_rate": 1.1691668516403532e-06, + "loss": 0.18, + "step": 35640 + }, + { + "epoch": 0.7845256438476778, + "grad_norm": 1.4672201040996844, + "learning_rate": 1.168937787547389e-06, + "loss": 0.2638, + "step": 35641 + }, + { + "epoch": 0.7845476557340965, + "grad_norm": 1.8269418520414256, + "learning_rate": 1.168708742925591e-06, + "loss": 0.2818, + "step": 35642 + }, + { + "epoch": 0.784569667620515, + "grad_norm": 1.5375718169026553, + "learning_rate": 1.1684797177761248e-06, + "loss": 0.2036, + "step": 35643 + }, + { + "epoch": 0.7845916795069338, + "grad_norm": 1.7954224480185792, + "learning_rate": 1.168250712100154e-06, + "loss": 0.3016, + "step": 35644 + }, + { + "epoch": 0.7846136913933525, + "grad_norm": 1.589286754927612, + "learning_rate": 1.1680217258988402e-06, + "loss": 0.2285, + "step": 35645 + }, + { + "epoch": 0.7846357032797711, + "grad_norm": 1.6233777026480671, + "learning_rate": 1.1677927591733495e-06, + "loss": 0.3012, + "step": 35646 + }, + { + "epoch": 0.7846577151661898, + "grad_norm": 1.16079348567419, + "learning_rate": 1.167563811924845e-06, + "loss": 0.1799, + "step": 35647 + }, + { + "epoch": 0.7846797270526084, + "grad_norm": 2.762889461804778, + "learning_rate": 1.1673348841544918e-06, + "loss": 0.2801, + "step": 35648 + }, + { + "epoch": 0.7847017389390271, + "grad_norm": 2.45526649936144, + "learning_rate": 1.1671059758634529e-06, + "loss": 0.2633, + "step": 35649 + }, + { + "epoch": 0.7847237508254458, + "grad_norm": 0.6200526347757714, + "learning_rate": 1.166877087052889e-06, + "loss": 0.4295, + "step": 35650 + }, + { + "epoch": 0.7847457627118644, + "grad_norm": 2.3053851418275277, + "learning_rate": 1.1666482177239653e-06, + "loss": 0.3174, + "step": 35651 + }, + { + "epoch": 0.7847677745982831, + "grad_norm": 1.7427067583055167, + "learning_rate": 1.1664193678778446e-06, + "loss": 0.2934, + "step": 35652 + }, + { + "epoch": 0.7847897864847018, + "grad_norm": 1.8768016676132935, + "learning_rate": 1.166190537515693e-06, + "loss": 0.2995, + "step": 35653 + }, + { + "epoch": 0.7848117983711204, + "grad_norm": 3.623532141002203, + "learning_rate": 1.1659617266386703e-06, + "loss": 0.2069, + "step": 35654 + }, + { + "epoch": 0.7848338102575391, + "grad_norm": 0.6149186774099128, + "learning_rate": 1.1657329352479385e-06, + "loss": 0.4669, + "step": 35655 + }, + { + "epoch": 0.7848558221439578, + "grad_norm": 1.9168327531276175, + "learning_rate": 1.1655041633446629e-06, + "loss": 0.2189, + "step": 35656 + }, + { + "epoch": 0.7848778340303764, + "grad_norm": 0.5991112230958813, + "learning_rate": 1.1652754109300045e-06, + "loss": 0.4414, + "step": 35657 + }, + { + "epoch": 0.7848998459167951, + "grad_norm": 0.6379280362076803, + "learning_rate": 1.165046678005129e-06, + "loss": 0.442, + "step": 35658 + }, + { + "epoch": 0.7849218578032138, + "grad_norm": 4.555994761299074, + "learning_rate": 1.164817964571196e-06, + "loss": 0.3652, + "step": 35659 + }, + { + "epoch": 0.7849438696896324, + "grad_norm": 1.5627008083036633, + "learning_rate": 1.1645892706293676e-06, + "loss": 0.305, + "step": 35660 + }, + { + "epoch": 0.7849658815760511, + "grad_norm": 2.694041133489439, + "learning_rate": 1.1643605961808075e-06, + "loss": 0.2541, + "step": 35661 + }, + { + "epoch": 0.7849878934624698, + "grad_norm": 1.6645434479993786, + "learning_rate": 1.1641319412266765e-06, + "loss": 0.2455, + "step": 35662 + }, + { + "epoch": 0.7850099053488884, + "grad_norm": 1.8444283832352368, + "learning_rate": 1.1639033057681403e-06, + "loss": 0.2951, + "step": 35663 + }, + { + "epoch": 0.7850319172353071, + "grad_norm": 2.006188459186314, + "learning_rate": 1.1636746898063584e-06, + "loss": 0.2857, + "step": 35664 + }, + { + "epoch": 0.7850539291217258, + "grad_norm": 6.004154013491275, + "learning_rate": 1.1634460933424908e-06, + "loss": 0.2835, + "step": 35665 + }, + { + "epoch": 0.7850759410081444, + "grad_norm": 0.6073613814742883, + "learning_rate": 1.1632175163777015e-06, + "loss": 0.4543, + "step": 35666 + }, + { + "epoch": 0.7850979528945631, + "grad_norm": 1.4331226305130902, + "learning_rate": 1.1629889589131537e-06, + "loss": 0.2506, + "step": 35667 + }, + { + "epoch": 0.7851199647809818, + "grad_norm": 1.8093917965262059, + "learning_rate": 1.1627604209500054e-06, + "loss": 0.3164, + "step": 35668 + }, + { + "epoch": 0.7851419766674004, + "grad_norm": 2.3753517373466306, + "learning_rate": 1.1625319024894222e-06, + "loss": 0.3506, + "step": 35669 + }, + { + "epoch": 0.7851639885538191, + "grad_norm": 1.8556153193782379, + "learning_rate": 1.1623034035325614e-06, + "loss": 0.2118, + "step": 35670 + }, + { + "epoch": 0.7851860004402377, + "grad_norm": 0.6051674508885088, + "learning_rate": 1.1620749240805867e-06, + "loss": 0.4236, + "step": 35671 + }, + { + "epoch": 0.7852080123266564, + "grad_norm": 2.7102813308758638, + "learning_rate": 1.1618464641346605e-06, + "loss": 0.3497, + "step": 35672 + }, + { + "epoch": 0.7852300242130751, + "grad_norm": 1.5243257459055743, + "learning_rate": 1.1616180236959408e-06, + "loss": 0.27, + "step": 35673 + }, + { + "epoch": 0.7852520360994937, + "grad_norm": 1.4185170511982712, + "learning_rate": 1.1613896027655918e-06, + "loss": 0.2099, + "step": 35674 + }, + { + "epoch": 0.7852740479859124, + "grad_norm": 2.5107066095396426, + "learning_rate": 1.1611612013447716e-06, + "loss": 0.3134, + "step": 35675 + }, + { + "epoch": 0.7852960598723311, + "grad_norm": 1.5684004138903482, + "learning_rate": 1.1609328194346436e-06, + "loss": 0.2746, + "step": 35676 + }, + { + "epoch": 0.7853180717587497, + "grad_norm": 2.1785224559530274, + "learning_rate": 1.1607044570363656e-06, + "loss": 0.285, + "step": 35677 + }, + { + "epoch": 0.7853400836451684, + "grad_norm": 1.8130636926369321, + "learning_rate": 1.1604761141511006e-06, + "loss": 0.3255, + "step": 35678 + }, + { + "epoch": 0.7853620955315871, + "grad_norm": 0.6243265875656181, + "learning_rate": 1.1602477907800092e-06, + "loss": 0.4656, + "step": 35679 + }, + { + "epoch": 0.7853841074180057, + "grad_norm": 1.6730556227902011, + "learning_rate": 1.1600194869242505e-06, + "loss": 0.3532, + "step": 35680 + }, + { + "epoch": 0.7854061193044244, + "grad_norm": 1.788311389131458, + "learning_rate": 1.159791202584986e-06, + "loss": 0.4147, + "step": 35681 + }, + { + "epoch": 0.7854281311908431, + "grad_norm": 1.3874249861035837, + "learning_rate": 1.1595629377633743e-06, + "loss": 0.2855, + "step": 35682 + }, + { + "epoch": 0.7854501430772617, + "grad_norm": 1.6104907735271055, + "learning_rate": 1.159334692460577e-06, + "loss": 0.2697, + "step": 35683 + }, + { + "epoch": 0.7854721549636804, + "grad_norm": 1.4707053340590968, + "learning_rate": 1.1591064666777547e-06, + "loss": 0.2164, + "step": 35684 + }, + { + "epoch": 0.7854941668500991, + "grad_norm": 1.852144056208584, + "learning_rate": 1.1588782604160663e-06, + "loss": 0.2539, + "step": 35685 + }, + { + "epoch": 0.7855161787365177, + "grad_norm": 2.3230466285047977, + "learning_rate": 1.158650073676671e-06, + "loss": 0.2866, + "step": 35686 + }, + { + "epoch": 0.7855381906229364, + "grad_norm": 2.9949845353114273, + "learning_rate": 1.1584219064607288e-06, + "loss": 0.3655, + "step": 35687 + }, + { + "epoch": 0.7855602025093551, + "grad_norm": 1.2962590071096796, + "learning_rate": 1.1581937587693998e-06, + "loss": 0.2776, + "step": 35688 + }, + { + "epoch": 0.7855822143957737, + "grad_norm": 1.460346524954969, + "learning_rate": 1.1579656306038456e-06, + "loss": 0.2753, + "step": 35689 + }, + { + "epoch": 0.7856042262821924, + "grad_norm": 2.5554426799043033, + "learning_rate": 1.157737521965223e-06, + "loss": 0.3213, + "step": 35690 + }, + { + "epoch": 0.7856262381686111, + "grad_norm": 1.3538713588714708, + "learning_rate": 1.157509432854691e-06, + "loss": 0.208, + "step": 35691 + }, + { + "epoch": 0.7856482500550297, + "grad_norm": 2.2750467359936164, + "learning_rate": 1.1572813632734097e-06, + "loss": 0.2868, + "step": 35692 + }, + { + "epoch": 0.7856702619414484, + "grad_norm": 1.2700586804199927, + "learning_rate": 1.1570533132225387e-06, + "loss": 0.2886, + "step": 35693 + }, + { + "epoch": 0.785692273827867, + "grad_norm": 0.6453239060676387, + "learning_rate": 1.1568252827032378e-06, + "loss": 0.4699, + "step": 35694 + }, + { + "epoch": 0.7857142857142857, + "grad_norm": 1.8383819518274624, + "learning_rate": 1.156597271716665e-06, + "loss": 0.3665, + "step": 35695 + }, + { + "epoch": 0.7857362976007044, + "grad_norm": 0.6108548182824366, + "learning_rate": 1.1563692802639776e-06, + "loss": 0.44, + "step": 35696 + }, + { + "epoch": 0.785758309487123, + "grad_norm": 0.6074760696070266, + "learning_rate": 1.1561413083463358e-06, + "loss": 0.42, + "step": 35697 + }, + { + "epoch": 0.7857803213735417, + "grad_norm": 1.5145931065696412, + "learning_rate": 1.1559133559648983e-06, + "loss": 0.3403, + "step": 35698 + }, + { + "epoch": 0.7858023332599604, + "grad_norm": 1.446994836851264, + "learning_rate": 1.1556854231208254e-06, + "loss": 0.2559, + "step": 35699 + }, + { + "epoch": 0.785824345146379, + "grad_norm": 1.8832869648827635, + "learning_rate": 1.1554575098152738e-06, + "loss": 0.3027, + "step": 35700 + }, + { + "epoch": 0.7858463570327977, + "grad_norm": 1.699416653699641, + "learning_rate": 1.1552296160493998e-06, + "loss": 0.3549, + "step": 35701 + }, + { + "epoch": 0.7858683689192164, + "grad_norm": 1.398721738839647, + "learning_rate": 1.1550017418243642e-06, + "loss": 0.2605, + "step": 35702 + }, + { + "epoch": 0.785890380805635, + "grad_norm": 1.2894900422706974, + "learning_rate": 1.1547738871413262e-06, + "loss": 0.3554, + "step": 35703 + }, + { + "epoch": 0.7859123926920537, + "grad_norm": 1.8414045885912809, + "learning_rate": 1.1545460520014407e-06, + "loss": 0.3306, + "step": 35704 + }, + { + "epoch": 0.7859344045784724, + "grad_norm": 1.8200384855940839, + "learning_rate": 1.1543182364058685e-06, + "loss": 0.3001, + "step": 35705 + }, + { + "epoch": 0.785956416464891, + "grad_norm": 0.6040696032993169, + "learning_rate": 1.1540904403557652e-06, + "loss": 0.4336, + "step": 35706 + }, + { + "epoch": 0.7859784283513097, + "grad_norm": 1.684781625773546, + "learning_rate": 1.1538626638522893e-06, + "loss": 0.2434, + "step": 35707 + }, + { + "epoch": 0.7860004402377284, + "grad_norm": 0.6034744873067458, + "learning_rate": 1.1536349068966e-06, + "loss": 0.4228, + "step": 35708 + }, + { + "epoch": 0.786022452124147, + "grad_norm": 2.54827969632783, + "learning_rate": 1.153407169489853e-06, + "loss": 0.3289, + "step": 35709 + }, + { + "epoch": 0.7860444640105657, + "grad_norm": 4.571800918281622, + "learning_rate": 1.1531794516332072e-06, + "loss": 0.3269, + "step": 35710 + }, + { + "epoch": 0.7860664758969844, + "grad_norm": 1.1737583723516707, + "learning_rate": 1.1529517533278179e-06, + "loss": 0.2903, + "step": 35711 + }, + { + "epoch": 0.786088487783403, + "grad_norm": 1.9605421047624896, + "learning_rate": 1.1527240745748447e-06, + "loss": 0.3283, + "step": 35712 + }, + { + "epoch": 0.7861104996698217, + "grad_norm": 0.6292464771218024, + "learning_rate": 1.1524964153754425e-06, + "loss": 0.4457, + "step": 35713 + }, + { + "epoch": 0.7861325115562404, + "grad_norm": 1.6825187235034849, + "learning_rate": 1.15226877573077e-06, + "loss": 0.301, + "step": 35714 + }, + { + "epoch": 0.786154523442659, + "grad_norm": 1.9807623101816043, + "learning_rate": 1.1520411556419847e-06, + "loss": 0.3508, + "step": 35715 + }, + { + "epoch": 0.7861765353290777, + "grad_norm": 2.2205260843261154, + "learning_rate": 1.151813555110241e-06, + "loss": 0.3487, + "step": 35716 + }, + { + "epoch": 0.7861985472154963, + "grad_norm": 1.7047240364291816, + "learning_rate": 1.1515859741366987e-06, + "loss": 0.2522, + "step": 35717 + }, + { + "epoch": 0.786220559101915, + "grad_norm": 1.5378755893764757, + "learning_rate": 1.1513584127225113e-06, + "loss": 0.3392, + "step": 35718 + }, + { + "epoch": 0.7862425709883337, + "grad_norm": 1.517372192426567, + "learning_rate": 1.1511308708688373e-06, + "loss": 0.2406, + "step": 35719 + }, + { + "epoch": 0.7862645828747523, + "grad_norm": 1.665515870913801, + "learning_rate": 1.1509033485768345e-06, + "loss": 0.2393, + "step": 35720 + }, + { + "epoch": 0.786286594761171, + "grad_norm": 2.428892288079618, + "learning_rate": 1.1506758458476557e-06, + "loss": 0.2562, + "step": 35721 + }, + { + "epoch": 0.7863086066475897, + "grad_norm": 2.294186994644932, + "learning_rate": 1.150448362682461e-06, + "loss": 0.3087, + "step": 35722 + }, + { + "epoch": 0.7863306185340083, + "grad_norm": 1.9518639488930345, + "learning_rate": 1.1502208990824027e-06, + "loss": 0.2614, + "step": 35723 + }, + { + "epoch": 0.786352630420427, + "grad_norm": 1.7751405997914294, + "learning_rate": 1.149993455048639e-06, + "loss": 0.2839, + "step": 35724 + }, + { + "epoch": 0.7863746423068457, + "grad_norm": 1.3469777418060622, + "learning_rate": 1.1497660305823271e-06, + "loss": 0.2792, + "step": 35725 + }, + { + "epoch": 0.7863966541932643, + "grad_norm": 1.6650310395091854, + "learning_rate": 1.149538625684622e-06, + "loss": 0.272, + "step": 35726 + }, + { + "epoch": 0.786418666079683, + "grad_norm": 2.4323826841051157, + "learning_rate": 1.149311240356677e-06, + "loss": 0.3123, + "step": 35727 + }, + { + "epoch": 0.7864406779661017, + "grad_norm": 1.4038494055013289, + "learning_rate": 1.1490838745996497e-06, + "loss": 0.231, + "step": 35728 + }, + { + "epoch": 0.7864626898525203, + "grad_norm": 4.115595657846209, + "learning_rate": 1.1488565284146957e-06, + "loss": 0.258, + "step": 35729 + }, + { + "epoch": 0.786484701738939, + "grad_norm": 1.6253347227124302, + "learning_rate": 1.148629201802972e-06, + "loss": 0.2565, + "step": 35730 + }, + { + "epoch": 0.7865067136253577, + "grad_norm": 1.9927310515771897, + "learning_rate": 1.1484018947656328e-06, + "loss": 0.2687, + "step": 35731 + }, + { + "epoch": 0.7865287255117763, + "grad_norm": 1.6031864050757203, + "learning_rate": 1.1481746073038313e-06, + "loss": 0.2702, + "step": 35732 + }, + { + "epoch": 0.786550737398195, + "grad_norm": 1.8646158930490206, + "learning_rate": 1.1479473394187245e-06, + "loss": 0.2874, + "step": 35733 + }, + { + "epoch": 0.7865727492846137, + "grad_norm": 1.5138197431122757, + "learning_rate": 1.1477200911114672e-06, + "loss": 0.246, + "step": 35734 + }, + { + "epoch": 0.7865947611710323, + "grad_norm": 1.3041194863351302, + "learning_rate": 1.147492862383216e-06, + "loss": 0.194, + "step": 35735 + }, + { + "epoch": 0.786616773057451, + "grad_norm": 1.4441364417082816, + "learning_rate": 1.1472656532351245e-06, + "loss": 0.2941, + "step": 35736 + }, + { + "epoch": 0.7866387849438697, + "grad_norm": 1.9388845708077338, + "learning_rate": 1.1470384636683457e-06, + "loss": 0.2748, + "step": 35737 + }, + { + "epoch": 0.7866607968302883, + "grad_norm": 2.3042726656728894, + "learning_rate": 1.1468112936840365e-06, + "loss": 0.3174, + "step": 35738 + }, + { + "epoch": 0.786682808716707, + "grad_norm": 1.7183390821203879, + "learning_rate": 1.1465841432833502e-06, + "loss": 0.3025, + "step": 35739 + }, + { + "epoch": 0.7867048206031257, + "grad_norm": 1.6826003504485685, + "learning_rate": 1.1463570124674439e-06, + "loss": 0.2999, + "step": 35740 + }, + { + "epoch": 0.7867268324895443, + "grad_norm": 0.6395934500373941, + "learning_rate": 1.1461299012374694e-06, + "loss": 0.4435, + "step": 35741 + }, + { + "epoch": 0.786748844375963, + "grad_norm": 1.6012260692527538, + "learning_rate": 1.1459028095945807e-06, + "loss": 0.2695, + "step": 35742 + }, + { + "epoch": 0.7867708562623816, + "grad_norm": 3.3010391262298, + "learning_rate": 1.1456757375399325e-06, + "loss": 0.29, + "step": 35743 + }, + { + "epoch": 0.7867928681488003, + "grad_norm": 1.7173230576031984, + "learning_rate": 1.145448685074681e-06, + "loss": 0.2718, + "step": 35744 + }, + { + "epoch": 0.786814880035219, + "grad_norm": 0.6040520997425546, + "learning_rate": 1.1452216521999771e-06, + "loss": 0.4532, + "step": 35745 + }, + { + "epoch": 0.7868368919216376, + "grad_norm": 6.924660727724703, + "learning_rate": 1.1449946389169774e-06, + "loss": 0.2988, + "step": 35746 + }, + { + "epoch": 0.7868589038080563, + "grad_norm": 1.4291610687268408, + "learning_rate": 1.1447676452268325e-06, + "loss": 0.2767, + "step": 35747 + }, + { + "epoch": 0.786880915694475, + "grad_norm": 1.797589772725659, + "learning_rate": 1.1445406711307e-06, + "loss": 0.2858, + "step": 35748 + }, + { + "epoch": 0.7869029275808936, + "grad_norm": 6.39992597044805, + "learning_rate": 1.1443137166297296e-06, + "loss": 0.2524, + "step": 35749 + }, + { + "epoch": 0.7869249394673123, + "grad_norm": 1.274559097402074, + "learning_rate": 1.1440867817250767e-06, + "loss": 0.2782, + "step": 35750 + }, + { + "epoch": 0.786946951353731, + "grad_norm": 1.6457520368694432, + "learning_rate": 1.1438598664178956e-06, + "loss": 0.3058, + "step": 35751 + }, + { + "epoch": 0.7869689632401496, + "grad_norm": 2.5215309384472384, + "learning_rate": 1.1436329707093375e-06, + "loss": 0.2446, + "step": 35752 + }, + { + "epoch": 0.7869909751265683, + "grad_norm": 1.555336740056681, + "learning_rate": 1.1434060946005582e-06, + "loss": 0.3144, + "step": 35753 + }, + { + "epoch": 0.787012987012987, + "grad_norm": 2.578169111633656, + "learning_rate": 1.1431792380927075e-06, + "loss": 0.3269, + "step": 35754 + }, + { + "epoch": 0.7870349988994056, + "grad_norm": 0.5947520377236609, + "learning_rate": 1.1429524011869397e-06, + "loss": 0.4515, + "step": 35755 + }, + { + "epoch": 0.7870570107858244, + "grad_norm": 1.9417509266558584, + "learning_rate": 1.1427255838844104e-06, + "loss": 0.2915, + "step": 35756 + }, + { + "epoch": 0.7870790226722431, + "grad_norm": 1.4824255206517403, + "learning_rate": 1.142498786186268e-06, + "loss": 0.2933, + "step": 35757 + }, + { + "epoch": 0.7871010345586616, + "grad_norm": 1.1891383002104094, + "learning_rate": 1.1422720080936689e-06, + "loss": 0.2142, + "step": 35758 + }, + { + "epoch": 0.7871230464450804, + "grad_norm": 1.3466002827335093, + "learning_rate": 1.1420452496077628e-06, + "loss": 0.2884, + "step": 35759 + }, + { + "epoch": 0.7871450583314991, + "grad_norm": 0.6278615065905846, + "learning_rate": 1.141818510729703e-06, + "loss": 0.4644, + "step": 35760 + }, + { + "epoch": 0.7871670702179177, + "grad_norm": 0.6089302634661646, + "learning_rate": 1.1415917914606444e-06, + "loss": 0.4632, + "step": 35761 + }, + { + "epoch": 0.7871890821043364, + "grad_norm": 1.7116211930850422, + "learning_rate": 1.1413650918017354e-06, + "loss": 0.3163, + "step": 35762 + }, + { + "epoch": 0.7872110939907551, + "grad_norm": 0.6065774644324644, + "learning_rate": 1.1411384117541313e-06, + "loss": 0.4253, + "step": 35763 + }, + { + "epoch": 0.7872331058771737, + "grad_norm": 1.5958632628676603, + "learning_rate": 1.1409117513189815e-06, + "loss": 0.2515, + "step": 35764 + }, + { + "epoch": 0.7872551177635924, + "grad_norm": 1.569565551596923, + "learning_rate": 1.1406851104974397e-06, + "loss": 0.2644, + "step": 35765 + }, + { + "epoch": 0.787277129650011, + "grad_norm": 1.2258154633329659, + "learning_rate": 1.1404584892906583e-06, + "loss": 0.3071, + "step": 35766 + }, + { + "epoch": 0.7872991415364297, + "grad_norm": 0.6185169316627386, + "learning_rate": 1.1402318876997892e-06, + "loss": 0.4572, + "step": 35767 + }, + { + "epoch": 0.7873211534228484, + "grad_norm": 1.6972402677414058, + "learning_rate": 1.1400053057259807e-06, + "loss": 0.2719, + "step": 35768 + }, + { + "epoch": 0.787343165309267, + "grad_norm": 0.5984696716697921, + "learning_rate": 1.1397787433703878e-06, + "loss": 0.4251, + "step": 35769 + }, + { + "epoch": 0.7873651771956857, + "grad_norm": 7.4620716580813395, + "learning_rate": 1.1395522006341603e-06, + "loss": 0.2742, + "step": 35770 + }, + { + "epoch": 0.7873871890821044, + "grad_norm": 1.7148640747938908, + "learning_rate": 1.1393256775184518e-06, + "loss": 0.299, + "step": 35771 + }, + { + "epoch": 0.787409200968523, + "grad_norm": 2.202851643187022, + "learning_rate": 1.1390991740244123e-06, + "loss": 0.2791, + "step": 35772 + }, + { + "epoch": 0.7874312128549417, + "grad_norm": 3.389780610072158, + "learning_rate": 1.1388726901531917e-06, + "loss": 0.2855, + "step": 35773 + }, + { + "epoch": 0.7874532247413604, + "grad_norm": 1.4868525829357555, + "learning_rate": 1.1386462259059416e-06, + "loss": 0.2366, + "step": 35774 + }, + { + "epoch": 0.787475236627779, + "grad_norm": 0.6017852406481151, + "learning_rate": 1.1384197812838143e-06, + "loss": 0.4142, + "step": 35775 + }, + { + "epoch": 0.7874972485141977, + "grad_norm": 1.7201891236762523, + "learning_rate": 1.138193356287961e-06, + "loss": 0.2897, + "step": 35776 + }, + { + "epoch": 0.7875192604006164, + "grad_norm": 1.6070770070410116, + "learning_rate": 1.1379669509195313e-06, + "loss": 0.2947, + "step": 35777 + }, + { + "epoch": 0.787541272287035, + "grad_norm": 2.0690607590556422, + "learning_rate": 1.1377405651796748e-06, + "loss": 0.3003, + "step": 35778 + }, + { + "epoch": 0.7875632841734537, + "grad_norm": 2.1274260754224366, + "learning_rate": 1.1375141990695432e-06, + "loss": 0.2627, + "step": 35779 + }, + { + "epoch": 0.7875852960598724, + "grad_norm": 2.1610074245320603, + "learning_rate": 1.1372878525902876e-06, + "loss": 0.2261, + "step": 35780 + }, + { + "epoch": 0.787607307946291, + "grad_norm": 1.7931569140006367, + "learning_rate": 1.137061525743059e-06, + "loss": 0.3043, + "step": 35781 + }, + { + "epoch": 0.7876293198327097, + "grad_norm": 1.8436834211155733, + "learning_rate": 1.1368352185290065e-06, + "loss": 0.2859, + "step": 35782 + }, + { + "epoch": 0.7876513317191284, + "grad_norm": 1.2929395964068948, + "learning_rate": 1.1366089309492794e-06, + "loss": 0.2584, + "step": 35783 + }, + { + "epoch": 0.787673343605547, + "grad_norm": 2.2104496581387876, + "learning_rate": 1.13638266300503e-06, + "loss": 0.2962, + "step": 35784 + }, + { + "epoch": 0.7876953554919657, + "grad_norm": 1.137577357622182, + "learning_rate": 1.1361564146974064e-06, + "loss": 0.1874, + "step": 35785 + }, + { + "epoch": 0.7877173673783844, + "grad_norm": 5.820150959646468, + "learning_rate": 1.1359301860275584e-06, + "loss": 0.2857, + "step": 35786 + }, + { + "epoch": 0.787739379264803, + "grad_norm": 1.8385878394633115, + "learning_rate": 1.1357039769966382e-06, + "loss": 0.2955, + "step": 35787 + }, + { + "epoch": 0.7877613911512217, + "grad_norm": 1.641334742404587, + "learning_rate": 1.1354777876057926e-06, + "loss": 0.2894, + "step": 35788 + }, + { + "epoch": 0.7877834030376403, + "grad_norm": 2.1750922051439687, + "learning_rate": 1.135251617856174e-06, + "loss": 0.3091, + "step": 35789 + }, + { + "epoch": 0.787805414924059, + "grad_norm": 5.903061398514755, + "learning_rate": 1.1350254677489293e-06, + "loss": 0.3335, + "step": 35790 + }, + { + "epoch": 0.7878274268104777, + "grad_norm": 1.9479870626947269, + "learning_rate": 1.1347993372852084e-06, + "loss": 0.2954, + "step": 35791 + }, + { + "epoch": 0.7878494386968963, + "grad_norm": 1.9774941634612846, + "learning_rate": 1.1345732264661624e-06, + "loss": 0.3013, + "step": 35792 + }, + { + "epoch": 0.787871450583315, + "grad_norm": 1.531265725935699, + "learning_rate": 1.1343471352929386e-06, + "loss": 0.3352, + "step": 35793 + }, + { + "epoch": 0.7878934624697337, + "grad_norm": 1.3440410433015264, + "learning_rate": 1.1341210637666878e-06, + "loss": 0.2516, + "step": 35794 + }, + { + "epoch": 0.7879154743561523, + "grad_norm": 2.0989017886978076, + "learning_rate": 1.1338950118885567e-06, + "loss": 0.3621, + "step": 35795 + }, + { + "epoch": 0.787937486242571, + "grad_norm": 0.6315965378541096, + "learning_rate": 1.1336689796596956e-06, + "loss": 0.4583, + "step": 35796 + }, + { + "epoch": 0.7879594981289897, + "grad_norm": 1.3908320544920123, + "learning_rate": 1.133442967081254e-06, + "loss": 0.3014, + "step": 35797 + }, + { + "epoch": 0.7879815100154083, + "grad_norm": 1.5212881397971676, + "learning_rate": 1.1332169741543791e-06, + "loss": 0.2464, + "step": 35798 + }, + { + "epoch": 0.788003521901827, + "grad_norm": 0.6234408336132717, + "learning_rate": 1.1329910008802213e-06, + "loss": 0.4419, + "step": 35799 + }, + { + "epoch": 0.7880255337882457, + "grad_norm": 0.6170507846614397, + "learning_rate": 1.1327650472599266e-06, + "loss": 0.4254, + "step": 35800 + }, + { + "epoch": 0.7880475456746643, + "grad_norm": 0.6134753738495745, + "learning_rate": 1.1325391132946445e-06, + "loss": 0.4432, + "step": 35801 + }, + { + "epoch": 0.788069557561083, + "grad_norm": 0.589438563603906, + "learning_rate": 1.1323131989855251e-06, + "loss": 0.4549, + "step": 35802 + }, + { + "epoch": 0.7880915694475017, + "grad_norm": 1.4909707396261522, + "learning_rate": 1.1320873043337156e-06, + "loss": 0.2954, + "step": 35803 + }, + { + "epoch": 0.7881135813339203, + "grad_norm": 2.65971316953697, + "learning_rate": 1.1318614293403612e-06, + "loss": 0.3407, + "step": 35804 + }, + { + "epoch": 0.788135593220339, + "grad_norm": 2.7815854291063467, + "learning_rate": 1.1316355740066132e-06, + "loss": 0.2732, + "step": 35805 + }, + { + "epoch": 0.7881576051067577, + "grad_norm": 0.615552304089664, + "learning_rate": 1.1314097383336182e-06, + "loss": 0.4508, + "step": 35806 + }, + { + "epoch": 0.7881796169931763, + "grad_norm": 1.2128196545033194, + "learning_rate": 1.1311839223225251e-06, + "loss": 0.2685, + "step": 35807 + }, + { + "epoch": 0.788201628879595, + "grad_norm": 1.7398741419145956, + "learning_rate": 1.1309581259744816e-06, + "loss": 0.3853, + "step": 35808 + }, + { + "epoch": 0.7882236407660137, + "grad_norm": 5.3451599133019085, + "learning_rate": 1.130732349290633e-06, + "loss": 0.2295, + "step": 35809 + }, + { + "epoch": 0.7882456526524323, + "grad_norm": 0.6227526722351393, + "learning_rate": 1.1305065922721282e-06, + "loss": 0.4421, + "step": 35810 + }, + { + "epoch": 0.788267664538851, + "grad_norm": 1.4351151415046757, + "learning_rate": 1.130280854920115e-06, + "loss": 0.2202, + "step": 35811 + }, + { + "epoch": 0.7882896764252696, + "grad_norm": 1.7083320002728628, + "learning_rate": 1.1300551372357415e-06, + "loss": 0.2685, + "step": 35812 + }, + { + "epoch": 0.7883116883116883, + "grad_norm": 1.8206474563713446, + "learning_rate": 1.1298294392201542e-06, + "loss": 0.3374, + "step": 35813 + }, + { + "epoch": 0.788333700198107, + "grad_norm": 1.6826745189527839, + "learning_rate": 1.129603760874498e-06, + "loss": 0.283, + "step": 35814 + }, + { + "epoch": 0.7883557120845256, + "grad_norm": 2.0119523768110503, + "learning_rate": 1.1293781021999217e-06, + "loss": 0.3206, + "step": 35815 + }, + { + "epoch": 0.7883777239709443, + "grad_norm": 1.4261433524795843, + "learning_rate": 1.1291524631975725e-06, + "loss": 0.3002, + "step": 35816 + }, + { + "epoch": 0.788399735857363, + "grad_norm": 1.733703101944813, + "learning_rate": 1.1289268438685986e-06, + "loss": 0.4249, + "step": 35817 + }, + { + "epoch": 0.7884217477437816, + "grad_norm": 1.6948870920234989, + "learning_rate": 1.128701244214145e-06, + "loss": 0.3463, + "step": 35818 + }, + { + "epoch": 0.7884437596302003, + "grad_norm": 0.6237819800545642, + "learning_rate": 1.1284756642353563e-06, + "loss": 0.4683, + "step": 35819 + }, + { + "epoch": 0.788465771516619, + "grad_norm": 1.7162413850176812, + "learning_rate": 1.1282501039333827e-06, + "loss": 0.2817, + "step": 35820 + }, + { + "epoch": 0.7884877834030376, + "grad_norm": 1.861044834004334, + "learning_rate": 1.1280245633093662e-06, + "loss": 0.3286, + "step": 35821 + }, + { + "epoch": 0.7885097952894563, + "grad_norm": 19.998611632268283, + "learning_rate": 1.1277990423644585e-06, + "loss": 0.2028, + "step": 35822 + }, + { + "epoch": 0.788531807175875, + "grad_norm": 0.6248073138856037, + "learning_rate": 1.1275735410998034e-06, + "loss": 0.4604, + "step": 35823 + }, + { + "epoch": 0.7885538190622936, + "grad_norm": 1.7833691974224883, + "learning_rate": 1.1273480595165454e-06, + "loss": 0.3346, + "step": 35824 + }, + { + "epoch": 0.7885758309487123, + "grad_norm": 0.6252066135538065, + "learning_rate": 1.1271225976158329e-06, + "loss": 0.4349, + "step": 35825 + }, + { + "epoch": 0.788597842835131, + "grad_norm": 2.31916654520576, + "learning_rate": 1.1268971553988095e-06, + "loss": 0.2878, + "step": 35826 + }, + { + "epoch": 0.7886198547215496, + "grad_norm": 1.7407654825439243, + "learning_rate": 1.1266717328666222e-06, + "loss": 0.3068, + "step": 35827 + }, + { + "epoch": 0.7886418666079683, + "grad_norm": 1.9145950579591904, + "learning_rate": 1.1264463300204176e-06, + "loss": 0.29, + "step": 35828 + }, + { + "epoch": 0.788663878494387, + "grad_norm": 1.7691337696482785, + "learning_rate": 1.1262209468613394e-06, + "loss": 0.3039, + "step": 35829 + }, + { + "epoch": 0.7886858903808056, + "grad_norm": 1.3988495702440973, + "learning_rate": 1.1259955833905356e-06, + "loss": 0.3035, + "step": 35830 + }, + { + "epoch": 0.7887079022672243, + "grad_norm": 0.6185123604134991, + "learning_rate": 1.1257702396091486e-06, + "loss": 0.4675, + "step": 35831 + }, + { + "epoch": 0.788729914153643, + "grad_norm": 2.034532221810134, + "learning_rate": 1.1255449155183251e-06, + "loss": 0.2653, + "step": 35832 + }, + { + "epoch": 0.7887519260400616, + "grad_norm": 1.8726466829702981, + "learning_rate": 1.1253196111192122e-06, + "loss": 0.2967, + "step": 35833 + }, + { + "epoch": 0.7887739379264803, + "grad_norm": 2.081713186210974, + "learning_rate": 1.125094326412951e-06, + "loss": 0.3082, + "step": 35834 + }, + { + "epoch": 0.7887959498128989, + "grad_norm": 1.4636118155647015, + "learning_rate": 1.1248690614006908e-06, + "loss": 0.236, + "step": 35835 + }, + { + "epoch": 0.7888179616993176, + "grad_norm": 1.5538636368467398, + "learning_rate": 1.1246438160835726e-06, + "loss": 0.2938, + "step": 35836 + }, + { + "epoch": 0.7888399735857363, + "grad_norm": 1.7297717540968247, + "learning_rate": 1.1244185904627435e-06, + "loss": 0.331, + "step": 35837 + }, + { + "epoch": 0.7888619854721549, + "grad_norm": 1.4203543933232583, + "learning_rate": 1.1241933845393488e-06, + "loss": 0.2349, + "step": 35838 + }, + { + "epoch": 0.7888839973585736, + "grad_norm": 2.840008423988593, + "learning_rate": 1.12396819831453e-06, + "loss": 0.3009, + "step": 35839 + }, + { + "epoch": 0.7889060092449923, + "grad_norm": 1.3447262445412187, + "learning_rate": 1.123743031789436e-06, + "loss": 0.2476, + "step": 35840 + }, + { + "epoch": 0.7889280211314109, + "grad_norm": 1.6039368926771882, + "learning_rate": 1.1235178849652067e-06, + "loss": 0.1853, + "step": 35841 + }, + { + "epoch": 0.7889500330178296, + "grad_norm": 2.494084007946232, + "learning_rate": 1.1232927578429886e-06, + "loss": 0.3601, + "step": 35842 + }, + { + "epoch": 0.7889720449042483, + "grad_norm": 2.306037438403056, + "learning_rate": 1.1230676504239273e-06, + "loss": 0.3131, + "step": 35843 + }, + { + "epoch": 0.7889940567906669, + "grad_norm": 1.827577374410765, + "learning_rate": 1.1228425627091649e-06, + "loss": 0.3243, + "step": 35844 + }, + { + "epoch": 0.7890160686770856, + "grad_norm": 2.4097648326904255, + "learning_rate": 1.1226174946998447e-06, + "loss": 0.2288, + "step": 35845 + }, + { + "epoch": 0.7890380805635043, + "grad_norm": 1.8130105259610934, + "learning_rate": 1.1223924463971119e-06, + "loss": 0.2782, + "step": 35846 + }, + { + "epoch": 0.7890600924499229, + "grad_norm": 1.909546037670942, + "learning_rate": 1.1221674178021102e-06, + "loss": 0.3204, + "step": 35847 + }, + { + "epoch": 0.7890821043363416, + "grad_norm": 3.046970408131901, + "learning_rate": 1.1219424089159847e-06, + "loss": 0.2572, + "step": 35848 + }, + { + "epoch": 0.7891041162227603, + "grad_norm": 0.6394128802722041, + "learning_rate": 1.1217174197398773e-06, + "loss": 0.4784, + "step": 35849 + }, + { + "epoch": 0.7891261281091789, + "grad_norm": 1.5957999383117534, + "learning_rate": 1.1214924502749303e-06, + "loss": 0.327, + "step": 35850 + }, + { + "epoch": 0.7891481399955976, + "grad_norm": 1.5563559331342618, + "learning_rate": 1.121267500522289e-06, + "loss": 0.2473, + "step": 35851 + }, + { + "epoch": 0.7891701518820163, + "grad_norm": 0.6249340609863405, + "learning_rate": 1.1210425704830957e-06, + "loss": 0.4454, + "step": 35852 + }, + { + "epoch": 0.7891921637684349, + "grad_norm": 1.6212840544338833, + "learning_rate": 1.1208176601584958e-06, + "loss": 0.3157, + "step": 35853 + }, + { + "epoch": 0.7892141756548536, + "grad_norm": 0.6172365351028979, + "learning_rate": 1.1205927695496304e-06, + "loss": 0.4551, + "step": 35854 + }, + { + "epoch": 0.7892361875412723, + "grad_norm": 1.3445812748391577, + "learning_rate": 1.1203678986576416e-06, + "loss": 0.2525, + "step": 35855 + }, + { + "epoch": 0.7892581994276909, + "grad_norm": 1.3170910307899324, + "learning_rate": 1.120143047483675e-06, + "loss": 0.235, + "step": 35856 + }, + { + "epoch": 0.7892802113141096, + "grad_norm": 2.99841861391425, + "learning_rate": 1.119918216028869e-06, + "loss": 0.2438, + "step": 35857 + }, + { + "epoch": 0.7893022232005282, + "grad_norm": 1.9867374366115413, + "learning_rate": 1.119693404294372e-06, + "loss": 0.362, + "step": 35858 + }, + { + "epoch": 0.7893242350869469, + "grad_norm": 1.7515578450371807, + "learning_rate": 1.1194686122813237e-06, + "loss": 0.2957, + "step": 35859 + }, + { + "epoch": 0.7893462469733656, + "grad_norm": 1.9414328374110723, + "learning_rate": 1.1192438399908656e-06, + "loss": 0.2169, + "step": 35860 + }, + { + "epoch": 0.7893682588597842, + "grad_norm": 1.82440129449995, + "learning_rate": 1.1190190874241425e-06, + "loss": 0.2836, + "step": 35861 + }, + { + "epoch": 0.7893902707462029, + "grad_norm": 1.9982366168716517, + "learning_rate": 1.1187943545822937e-06, + "loss": 0.3551, + "step": 35862 + }, + { + "epoch": 0.7894122826326216, + "grad_norm": 1.7278335322595708, + "learning_rate": 1.1185696414664632e-06, + "loss": 0.3138, + "step": 35863 + }, + { + "epoch": 0.7894342945190402, + "grad_norm": 1.8502692042791995, + "learning_rate": 1.1183449480777942e-06, + "loss": 0.3435, + "step": 35864 + }, + { + "epoch": 0.7894563064054589, + "grad_norm": 6.743648846492171, + "learning_rate": 1.1181202744174263e-06, + "loss": 0.2089, + "step": 35865 + }, + { + "epoch": 0.7894783182918776, + "grad_norm": 2.347223667815567, + "learning_rate": 1.1178956204865039e-06, + "loss": 0.2474, + "step": 35866 + }, + { + "epoch": 0.7895003301782962, + "grad_norm": 2.306941389246804, + "learning_rate": 1.1176709862861657e-06, + "loss": 0.3234, + "step": 35867 + }, + { + "epoch": 0.7895223420647149, + "grad_norm": 0.6220737283330181, + "learning_rate": 1.1174463718175554e-06, + "loss": 0.4741, + "step": 35868 + }, + { + "epoch": 0.7895443539511336, + "grad_norm": 1.9925197613158339, + "learning_rate": 1.1172217770818155e-06, + "loss": 0.2816, + "step": 35869 + }, + { + "epoch": 0.7895663658375522, + "grad_norm": 1.438595237925668, + "learning_rate": 1.1169972020800851e-06, + "loss": 0.2378, + "step": 35870 + }, + { + "epoch": 0.789588377723971, + "grad_norm": 3.8911533959572013, + "learning_rate": 1.1167726468135082e-06, + "loss": 0.2623, + "step": 35871 + }, + { + "epoch": 0.7896103896103897, + "grad_norm": 2.1451036416725247, + "learning_rate": 1.1165481112832232e-06, + "loss": 0.3259, + "step": 35872 + }, + { + "epoch": 0.7896324014968082, + "grad_norm": 2.2092716822335965, + "learning_rate": 1.116323595490373e-06, + "loss": 0.2727, + "step": 35873 + }, + { + "epoch": 0.789654413383227, + "grad_norm": 1.6449107240101917, + "learning_rate": 1.1160990994360994e-06, + "loss": 0.3669, + "step": 35874 + }, + { + "epoch": 0.7896764252696457, + "grad_norm": 1.9571703291626206, + "learning_rate": 1.1158746231215411e-06, + "loss": 0.3085, + "step": 35875 + }, + { + "epoch": 0.7896984371560642, + "grad_norm": 1.646728899820966, + "learning_rate": 1.1156501665478419e-06, + "loss": 0.2531, + "step": 35876 + }, + { + "epoch": 0.789720449042483, + "grad_norm": 1.6194164822865689, + "learning_rate": 1.1154257297161397e-06, + "loss": 0.2656, + "step": 35877 + }, + { + "epoch": 0.7897424609289017, + "grad_norm": 0.6253644895779115, + "learning_rate": 1.1152013126275763e-06, + "loss": 0.4618, + "step": 35878 + }, + { + "epoch": 0.7897644728153203, + "grad_norm": 1.684710515803769, + "learning_rate": 1.114976915283294e-06, + "loss": 0.2997, + "step": 35879 + }, + { + "epoch": 0.789786484701739, + "grad_norm": 1.6372986019370808, + "learning_rate": 1.1147525376844304e-06, + "loss": 0.2754, + "step": 35880 + }, + { + "epoch": 0.7898084965881577, + "grad_norm": 2.174646688246362, + "learning_rate": 1.114528179832129e-06, + "loss": 0.2949, + "step": 35881 + }, + { + "epoch": 0.7898305084745763, + "grad_norm": 23.106145978323312, + "learning_rate": 1.1143038417275265e-06, + "loss": 0.3837, + "step": 35882 + }, + { + "epoch": 0.789852520360995, + "grad_norm": 1.3882910253044687, + "learning_rate": 1.1140795233717649e-06, + "loss": 0.3427, + "step": 35883 + }, + { + "epoch": 0.7898745322474136, + "grad_norm": 1.818595018800889, + "learning_rate": 1.1138552247659857e-06, + "loss": 0.4034, + "step": 35884 + }, + { + "epoch": 0.7898965441338323, + "grad_norm": 1.4121170572866124, + "learning_rate": 1.1136309459113282e-06, + "loss": 0.3227, + "step": 35885 + }, + { + "epoch": 0.789918556020251, + "grad_norm": 1.6275418583432244, + "learning_rate": 1.1134066868089293e-06, + "loss": 0.3382, + "step": 35886 + }, + { + "epoch": 0.7899405679066696, + "grad_norm": 1.6741010967030938, + "learning_rate": 1.113182447459933e-06, + "loss": 0.3057, + "step": 35887 + }, + { + "epoch": 0.7899625797930883, + "grad_norm": 2.1512715009170433, + "learning_rate": 1.1129582278654743e-06, + "loss": 0.3001, + "step": 35888 + }, + { + "epoch": 0.789984591679507, + "grad_norm": 1.725112147540037, + "learning_rate": 1.112734028026698e-06, + "loss": 0.2978, + "step": 35889 + }, + { + "epoch": 0.7900066035659256, + "grad_norm": 0.6170888621732051, + "learning_rate": 1.1125098479447416e-06, + "loss": 0.4634, + "step": 35890 + }, + { + "epoch": 0.7900286154523443, + "grad_norm": 1.4856816067318082, + "learning_rate": 1.1122856876207422e-06, + "loss": 0.2693, + "step": 35891 + }, + { + "epoch": 0.790050627338763, + "grad_norm": 0.5925765594716197, + "learning_rate": 1.1120615470558426e-06, + "loss": 0.4211, + "step": 35892 + }, + { + "epoch": 0.7900726392251816, + "grad_norm": 3.6306036585892674, + "learning_rate": 1.1118374262511778e-06, + "loss": 0.2709, + "step": 35893 + }, + { + "epoch": 0.7900946511116003, + "grad_norm": 1.769324796291631, + "learning_rate": 1.1116133252078913e-06, + "loss": 0.2562, + "step": 35894 + }, + { + "epoch": 0.790116662998019, + "grad_norm": 1.5715077966237996, + "learning_rate": 1.1113892439271207e-06, + "loss": 0.3204, + "step": 35895 + }, + { + "epoch": 0.7901386748844376, + "grad_norm": 1.7981176714942533, + "learning_rate": 1.1111651824100029e-06, + "loss": 0.3528, + "step": 35896 + }, + { + "epoch": 0.7901606867708563, + "grad_norm": 1.7566229458731095, + "learning_rate": 1.1109411406576791e-06, + "loss": 0.3734, + "step": 35897 + }, + { + "epoch": 0.790182698657275, + "grad_norm": 1.9267220270194003, + "learning_rate": 1.1107171186712862e-06, + "loss": 0.2492, + "step": 35898 + }, + { + "epoch": 0.7902047105436936, + "grad_norm": 1.3086002474379623, + "learning_rate": 1.1104931164519632e-06, + "loss": 0.3556, + "step": 35899 + }, + { + "epoch": 0.7902267224301123, + "grad_norm": 2.9685070393015858, + "learning_rate": 1.1102691340008509e-06, + "loss": 0.3275, + "step": 35900 + }, + { + "epoch": 0.790248734316531, + "grad_norm": 0.6337283647984788, + "learning_rate": 1.1100451713190835e-06, + "loss": 0.4756, + "step": 35901 + }, + { + "epoch": 0.7902707462029496, + "grad_norm": 0.5953051645184129, + "learning_rate": 1.1098212284078037e-06, + "loss": 0.4435, + "step": 35902 + }, + { + "epoch": 0.7902927580893683, + "grad_norm": 1.3895429253467297, + "learning_rate": 1.1095973052681452e-06, + "loss": 0.236, + "step": 35903 + }, + { + "epoch": 0.790314769975787, + "grad_norm": 3.212527557033842, + "learning_rate": 1.1093734019012493e-06, + "loss": 0.3125, + "step": 35904 + }, + { + "epoch": 0.7903367818622056, + "grad_norm": 3.2588092015176744, + "learning_rate": 1.109149518308254e-06, + "loss": 0.2795, + "step": 35905 + }, + { + "epoch": 0.7903587937486243, + "grad_norm": 0.6012975818786384, + "learning_rate": 1.1089256544902948e-06, + "loss": 0.4496, + "step": 35906 + }, + { + "epoch": 0.7903808056350429, + "grad_norm": 1.4886202932915238, + "learning_rate": 1.108701810448512e-06, + "loss": 0.3576, + "step": 35907 + }, + { + "epoch": 0.7904028175214616, + "grad_norm": 1.6752703671517484, + "learning_rate": 1.108477986184041e-06, + "loss": 0.2649, + "step": 35908 + }, + { + "epoch": 0.7904248294078803, + "grad_norm": 0.6195682644721207, + "learning_rate": 1.1082541816980208e-06, + "loss": 0.465, + "step": 35909 + }, + { + "epoch": 0.7904468412942989, + "grad_norm": 1.6224479381968613, + "learning_rate": 1.10803039699159e-06, + "loss": 0.2429, + "step": 35910 + }, + { + "epoch": 0.7904688531807176, + "grad_norm": 1.7140468617471531, + "learning_rate": 1.1078066320658832e-06, + "loss": 0.2147, + "step": 35911 + }, + { + "epoch": 0.7904908650671363, + "grad_norm": 2.5264120188539945, + "learning_rate": 1.10758288692204e-06, + "loss": 0.2672, + "step": 35912 + }, + { + "epoch": 0.7905128769535549, + "grad_norm": 1.8602333860425682, + "learning_rate": 1.1073591615611957e-06, + "loss": 0.3567, + "step": 35913 + }, + { + "epoch": 0.7905348888399736, + "grad_norm": 1.9929643357528346, + "learning_rate": 1.1071354559844877e-06, + "loss": 0.298, + "step": 35914 + }, + { + "epoch": 0.7905569007263923, + "grad_norm": 2.4632148657983977, + "learning_rate": 1.1069117701930554e-06, + "loss": 0.3697, + "step": 35915 + }, + { + "epoch": 0.7905789126128109, + "grad_norm": 2.502469355980423, + "learning_rate": 1.1066881041880328e-06, + "loss": 0.3032, + "step": 35916 + }, + { + "epoch": 0.7906009244992296, + "grad_norm": 1.8753068954060093, + "learning_rate": 1.1064644579705586e-06, + "loss": 0.3085, + "step": 35917 + }, + { + "epoch": 0.7906229363856483, + "grad_norm": 1.6661476555990242, + "learning_rate": 1.1062408315417672e-06, + "loss": 0.1757, + "step": 35918 + }, + { + "epoch": 0.7906449482720669, + "grad_norm": 2.0570480259871062, + "learning_rate": 1.1060172249027968e-06, + "loss": 0.2749, + "step": 35919 + }, + { + "epoch": 0.7906669601584856, + "grad_norm": 1.2331235238677298, + "learning_rate": 1.105793638054785e-06, + "loss": 0.2437, + "step": 35920 + }, + { + "epoch": 0.7906889720449043, + "grad_norm": 1.5111618215814, + "learning_rate": 1.1055700709988654e-06, + "loss": 0.254, + "step": 35921 + }, + { + "epoch": 0.7907109839313229, + "grad_norm": 0.5977868525861877, + "learning_rate": 1.1053465237361772e-06, + "loss": 0.4263, + "step": 35922 + }, + { + "epoch": 0.7907329958177416, + "grad_norm": 0.6284154506886483, + "learning_rate": 1.1051229962678545e-06, + "loss": 0.4599, + "step": 35923 + }, + { + "epoch": 0.7907550077041603, + "grad_norm": 1.659160043981887, + "learning_rate": 1.1048994885950316e-06, + "loss": 0.2263, + "step": 35924 + }, + { + "epoch": 0.7907770195905789, + "grad_norm": 1.7774569190051182, + "learning_rate": 1.1046760007188495e-06, + "loss": 0.3043, + "step": 35925 + }, + { + "epoch": 0.7907990314769976, + "grad_norm": 1.8372455155836225, + "learning_rate": 1.104452532640441e-06, + "loss": 0.2482, + "step": 35926 + }, + { + "epoch": 0.7908210433634163, + "grad_norm": 2.0735554618586587, + "learning_rate": 1.104229084360941e-06, + "loss": 0.3715, + "step": 35927 + }, + { + "epoch": 0.7908430552498349, + "grad_norm": 2.1062057504048517, + "learning_rate": 1.1040056558814877e-06, + "loss": 0.266, + "step": 35928 + }, + { + "epoch": 0.7908650671362536, + "grad_norm": 1.3069156306810623, + "learning_rate": 1.1037822472032127e-06, + "loss": 0.2744, + "step": 35929 + }, + { + "epoch": 0.7908870790226722, + "grad_norm": 1.5583108063189595, + "learning_rate": 1.1035588583272566e-06, + "loss": 0.3112, + "step": 35930 + }, + { + "epoch": 0.7909090909090909, + "grad_norm": 2.2970718988357017, + "learning_rate": 1.1033354892547525e-06, + "loss": 0.2344, + "step": 35931 + }, + { + "epoch": 0.7909311027955096, + "grad_norm": 1.369615098044617, + "learning_rate": 1.1031121399868338e-06, + "loss": 0.2405, + "step": 35932 + }, + { + "epoch": 0.7909531146819282, + "grad_norm": 1.672235363508251, + "learning_rate": 1.1028888105246388e-06, + "loss": 0.3236, + "step": 35933 + }, + { + "epoch": 0.7909751265683469, + "grad_norm": 0.6073096663557095, + "learning_rate": 1.1026655008693e-06, + "loss": 0.4444, + "step": 35934 + }, + { + "epoch": 0.7909971384547656, + "grad_norm": 0.6245453622710329, + "learning_rate": 1.102442211021953e-06, + "loss": 0.4565, + "step": 35935 + }, + { + "epoch": 0.7910191503411842, + "grad_norm": 0.624574344248579, + "learning_rate": 1.1022189409837342e-06, + "loss": 0.4553, + "step": 35936 + }, + { + "epoch": 0.7910411622276029, + "grad_norm": 3.166015560625895, + "learning_rate": 1.101995690755776e-06, + "loss": 0.248, + "step": 35937 + }, + { + "epoch": 0.7910631741140216, + "grad_norm": 4.348880145903913, + "learning_rate": 1.101772460339216e-06, + "loss": 0.2975, + "step": 35938 + }, + { + "epoch": 0.7910851860004402, + "grad_norm": 1.5904958103480487, + "learning_rate": 1.1015492497351854e-06, + "loss": 0.2633, + "step": 35939 + }, + { + "epoch": 0.7911071978868589, + "grad_norm": 1.3976528702508464, + "learning_rate": 1.10132605894482e-06, + "loss": 0.262, + "step": 35940 + }, + { + "epoch": 0.7911292097732776, + "grad_norm": 1.5178954856106246, + "learning_rate": 1.101102887969256e-06, + "loss": 0.3009, + "step": 35941 + }, + { + "epoch": 0.7911512216596962, + "grad_norm": 2.923120569854917, + "learning_rate": 1.1008797368096247e-06, + "loss": 0.3024, + "step": 35942 + }, + { + "epoch": 0.7911732335461149, + "grad_norm": 1.2936214549380802, + "learning_rate": 1.100656605467063e-06, + "loss": 0.2758, + "step": 35943 + }, + { + "epoch": 0.7911952454325336, + "grad_norm": 1.905671298310894, + "learning_rate": 1.1004334939427019e-06, + "loss": 0.3453, + "step": 35944 + }, + { + "epoch": 0.7912172573189522, + "grad_norm": 1.6398692771426209, + "learning_rate": 1.1002104022376776e-06, + "loss": 0.2931, + "step": 35945 + }, + { + "epoch": 0.7912392692053709, + "grad_norm": 2.376968723929096, + "learning_rate": 1.0999873303531245e-06, + "loss": 0.2923, + "step": 35946 + }, + { + "epoch": 0.7912612810917896, + "grad_norm": 1.7941843299495812, + "learning_rate": 1.099764278290174e-06, + "loss": 0.2494, + "step": 35947 + }, + { + "epoch": 0.7912832929782082, + "grad_norm": 1.561933909268609, + "learning_rate": 1.0995412460499622e-06, + "loss": 0.2746, + "step": 35948 + }, + { + "epoch": 0.7913053048646269, + "grad_norm": 1.5824533510420657, + "learning_rate": 1.09931823363362e-06, + "loss": 0.2468, + "step": 35949 + }, + { + "epoch": 0.7913273167510456, + "grad_norm": 1.618417899890451, + "learning_rate": 1.0990952410422822e-06, + "loss": 0.2817, + "step": 35950 + }, + { + "epoch": 0.7913493286374642, + "grad_norm": 2.233766230334066, + "learning_rate": 1.0988722682770841e-06, + "loss": 0.3071, + "step": 35951 + }, + { + "epoch": 0.7913713405238829, + "grad_norm": 1.627971325124716, + "learning_rate": 1.0986493153391554e-06, + "loss": 0.3219, + "step": 35952 + }, + { + "epoch": 0.7913933524103015, + "grad_norm": 0.6352424042696465, + "learning_rate": 1.0984263822296327e-06, + "loss": 0.4656, + "step": 35953 + }, + { + "epoch": 0.7914153642967202, + "grad_norm": 1.5493931049793135, + "learning_rate": 1.098203468949645e-06, + "loss": 0.2742, + "step": 35954 + }, + { + "epoch": 0.7914373761831389, + "grad_norm": 1.759283155679335, + "learning_rate": 1.097980575500328e-06, + "loss": 0.2814, + "step": 35955 + }, + { + "epoch": 0.7914593880695575, + "grad_norm": 1.9031071236423753, + "learning_rate": 1.0977577018828155e-06, + "loss": 0.3704, + "step": 35956 + }, + { + "epoch": 0.7914813999559762, + "grad_norm": 5.032748275363319, + "learning_rate": 1.0975348480982379e-06, + "loss": 0.3599, + "step": 35957 + }, + { + "epoch": 0.7915034118423949, + "grad_norm": 0.6084307220588484, + "learning_rate": 1.0973120141477296e-06, + "loss": 0.4623, + "step": 35958 + }, + { + "epoch": 0.7915254237288135, + "grad_norm": 2.605732310512802, + "learning_rate": 1.0970892000324224e-06, + "loss": 0.293, + "step": 35959 + }, + { + "epoch": 0.7915474356152322, + "grad_norm": 0.6372688612345507, + "learning_rate": 1.0968664057534456e-06, + "loss": 0.4329, + "step": 35960 + }, + { + "epoch": 0.7915694475016509, + "grad_norm": 2.962937602209794, + "learning_rate": 1.0966436313119377e-06, + "loss": 0.2694, + "step": 35961 + }, + { + "epoch": 0.7915914593880695, + "grad_norm": 1.3527004370388325, + "learning_rate": 1.0964208767090262e-06, + "loss": 0.2936, + "step": 35962 + }, + { + "epoch": 0.7916134712744882, + "grad_norm": 0.6419400513692991, + "learning_rate": 1.0961981419458462e-06, + "loss": 0.4575, + "step": 35963 + }, + { + "epoch": 0.7916354831609069, + "grad_norm": 1.8278294623695648, + "learning_rate": 1.0959754270235285e-06, + "loss": 0.304, + "step": 35964 + }, + { + "epoch": 0.7916574950473255, + "grad_norm": 0.6223812974909749, + "learning_rate": 1.095752731943202e-06, + "loss": 0.463, + "step": 35965 + }, + { + "epoch": 0.7916795069337442, + "grad_norm": 2.527779684551032, + "learning_rate": 1.0955300567060046e-06, + "loss": 0.3352, + "step": 35966 + }, + { + "epoch": 0.7917015188201629, + "grad_norm": 1.5246469115028518, + "learning_rate": 1.0953074013130648e-06, + "loss": 0.3483, + "step": 35967 + }, + { + "epoch": 0.7917235307065815, + "grad_norm": 1.9091916228908923, + "learning_rate": 1.0950847657655128e-06, + "loss": 0.3475, + "step": 35968 + }, + { + "epoch": 0.7917455425930002, + "grad_norm": 1.4674752755577443, + "learning_rate": 1.094862150064483e-06, + "loss": 0.2529, + "step": 35969 + }, + { + "epoch": 0.7917675544794189, + "grad_norm": 1.3874463358315023, + "learning_rate": 1.0946395542111043e-06, + "loss": 0.3088, + "step": 35970 + }, + { + "epoch": 0.7917895663658375, + "grad_norm": 1.1582009615544075, + "learning_rate": 1.094416978206509e-06, + "loss": 0.2213, + "step": 35971 + }, + { + "epoch": 0.7918115782522562, + "grad_norm": 1.8393363472753312, + "learning_rate": 1.0941944220518303e-06, + "loss": 0.3161, + "step": 35972 + }, + { + "epoch": 0.7918335901386749, + "grad_norm": 1.2998540935386846, + "learning_rate": 1.0939718857481956e-06, + "loss": 0.2413, + "step": 35973 + }, + { + "epoch": 0.7918556020250935, + "grad_norm": 1.3963303961097564, + "learning_rate": 1.0937493692967398e-06, + "loss": 0.2885, + "step": 35974 + }, + { + "epoch": 0.7918776139115122, + "grad_norm": 2.179530986027296, + "learning_rate": 1.0935268726985903e-06, + "loss": 0.2827, + "step": 35975 + }, + { + "epoch": 0.7918996257979308, + "grad_norm": 1.4710152762441637, + "learning_rate": 1.0933043959548794e-06, + "loss": 0.2445, + "step": 35976 + }, + { + "epoch": 0.7919216376843495, + "grad_norm": 2.622923412543956, + "learning_rate": 1.0930819390667396e-06, + "loss": 0.2586, + "step": 35977 + }, + { + "epoch": 0.7919436495707682, + "grad_norm": 0.6496331397267878, + "learning_rate": 1.0928595020352988e-06, + "loss": 0.473, + "step": 35978 + }, + { + "epoch": 0.7919656614571868, + "grad_norm": 1.6514853845816255, + "learning_rate": 1.0926370848616902e-06, + "loss": 0.3042, + "step": 35979 + }, + { + "epoch": 0.7919876733436055, + "grad_norm": 0.6274300602842566, + "learning_rate": 1.0924146875470405e-06, + "loss": 0.4583, + "step": 35980 + }, + { + "epoch": 0.7920096852300242, + "grad_norm": 1.573154244275758, + "learning_rate": 1.0921923100924824e-06, + "loss": 0.289, + "step": 35981 + }, + { + "epoch": 0.7920316971164428, + "grad_norm": 0.6138442440335681, + "learning_rate": 1.0919699524991478e-06, + "loss": 0.4542, + "step": 35982 + }, + { + "epoch": 0.7920537090028615, + "grad_norm": 3.9585561923019172, + "learning_rate": 1.0917476147681632e-06, + "loss": 0.2519, + "step": 35983 + }, + { + "epoch": 0.7920757208892802, + "grad_norm": 1.554783178003884, + "learning_rate": 1.0915252969006617e-06, + "loss": 0.2238, + "step": 35984 + }, + { + "epoch": 0.7920977327756988, + "grad_norm": 1.4483972949611792, + "learning_rate": 1.0913029988977708e-06, + "loss": 0.3112, + "step": 35985 + }, + { + "epoch": 0.7921197446621175, + "grad_norm": 1.4227419110151958, + "learning_rate": 1.0910807207606212e-06, + "loss": 0.3225, + "step": 35986 + }, + { + "epoch": 0.7921417565485362, + "grad_norm": 3.1005177647090245, + "learning_rate": 1.090858462490344e-06, + "loss": 0.28, + "step": 35987 + }, + { + "epoch": 0.7921637684349548, + "grad_norm": 0.6090897733901457, + "learning_rate": 1.0906362240880663e-06, + "loss": 0.4597, + "step": 35988 + }, + { + "epoch": 0.7921857803213735, + "grad_norm": 1.8770480639668556, + "learning_rate": 1.0904140055549201e-06, + "loss": 0.334, + "step": 35989 + }, + { + "epoch": 0.7922077922077922, + "grad_norm": 0.5917635323014849, + "learning_rate": 1.0901918068920325e-06, + "loss": 0.4371, + "step": 35990 + }, + { + "epoch": 0.7922298040942108, + "grad_norm": 1.5795022325873669, + "learning_rate": 1.089969628100534e-06, + "loss": 0.2816, + "step": 35991 + }, + { + "epoch": 0.7922518159806295, + "grad_norm": 1.2759554171802734, + "learning_rate": 1.0897474691815552e-06, + "loss": 0.292, + "step": 35992 + }, + { + "epoch": 0.7922738278670483, + "grad_norm": 1.6358516480416587, + "learning_rate": 1.0895253301362225e-06, + "loss": 0.3361, + "step": 35993 + }, + { + "epoch": 0.7922958397534668, + "grad_norm": 0.6171370057128492, + "learning_rate": 1.0893032109656676e-06, + "loss": 0.4263, + "step": 35994 + }, + { + "epoch": 0.7923178516398856, + "grad_norm": 2.704698575845838, + "learning_rate": 1.0890811116710177e-06, + "loss": 0.345, + "step": 35995 + }, + { + "epoch": 0.7923398635263043, + "grad_norm": 1.535570945929088, + "learning_rate": 1.0888590322533993e-06, + "loss": 0.3252, + "step": 35996 + }, + { + "epoch": 0.7923618754127228, + "grad_norm": 0.5972328774752882, + "learning_rate": 1.0886369727139462e-06, + "loss": 0.4365, + "step": 35997 + }, + { + "epoch": 0.7923838872991416, + "grad_norm": 0.6221590696529437, + "learning_rate": 1.0884149330537835e-06, + "loss": 0.447, + "step": 35998 + }, + { + "epoch": 0.7924058991855601, + "grad_norm": 1.442127487550424, + "learning_rate": 1.088192913274042e-06, + "loss": 0.335, + "step": 35999 + }, + { + "epoch": 0.7924279110719789, + "grad_norm": 2.061010159022367, + "learning_rate": 1.0879709133758493e-06, + "loss": 0.2578, + "step": 36000 + }, + { + "epoch": 0.7924499229583976, + "grad_norm": 1.5910519874879645, + "learning_rate": 1.087748933360331e-06, + "loss": 0.28, + "step": 36001 + }, + { + "epoch": 0.7924719348448162, + "grad_norm": 1.3582659645819448, + "learning_rate": 1.087526973228618e-06, + "loss": 0.2342, + "step": 36002 + }, + { + "epoch": 0.7924939467312349, + "grad_norm": 1.5603330380695475, + "learning_rate": 1.087305032981839e-06, + "loss": 0.2921, + "step": 36003 + }, + { + "epoch": 0.7925159586176536, + "grad_norm": 4.529495501400964, + "learning_rate": 1.08708311262112e-06, + "loss": 0.3219, + "step": 36004 + }, + { + "epoch": 0.7925379705040722, + "grad_norm": 1.6331110131180537, + "learning_rate": 1.0868612121475909e-06, + "loss": 0.3223, + "step": 36005 + }, + { + "epoch": 0.7925599823904909, + "grad_norm": 3.532094138841712, + "learning_rate": 1.0866393315623775e-06, + "loss": 0.2735, + "step": 36006 + }, + { + "epoch": 0.7925819942769096, + "grad_norm": 2.12729986039189, + "learning_rate": 1.0864174708666076e-06, + "loss": 0.221, + "step": 36007 + }, + { + "epoch": 0.7926040061633282, + "grad_norm": 0.6026694542444451, + "learning_rate": 1.086195630061412e-06, + "loss": 0.4305, + "step": 36008 + }, + { + "epoch": 0.7926260180497469, + "grad_norm": 0.6088934914368058, + "learning_rate": 1.0859738091479138e-06, + "loss": 0.4374, + "step": 36009 + }, + { + "epoch": 0.7926480299361656, + "grad_norm": 1.8529154391715288, + "learning_rate": 1.0857520081272439e-06, + "loss": 0.2877, + "step": 36010 + }, + { + "epoch": 0.7926700418225842, + "grad_norm": 0.609736005257441, + "learning_rate": 1.0855302270005263e-06, + "loss": 0.4469, + "step": 36011 + }, + { + "epoch": 0.7926920537090029, + "grad_norm": 1.8513043936722968, + "learning_rate": 1.0853084657688907e-06, + "loss": 0.2489, + "step": 36012 + }, + { + "epoch": 0.7927140655954216, + "grad_norm": 0.6162428606869975, + "learning_rate": 1.0850867244334646e-06, + "loss": 0.4269, + "step": 36013 + }, + { + "epoch": 0.7927360774818402, + "grad_norm": 0.5883321251848804, + "learning_rate": 1.0848650029953722e-06, + "loss": 0.4192, + "step": 36014 + }, + { + "epoch": 0.7927580893682589, + "grad_norm": 1.4037238500206015, + "learning_rate": 1.084643301455744e-06, + "loss": 0.3096, + "step": 36015 + }, + { + "epoch": 0.7927801012546776, + "grad_norm": 0.6227271820992623, + "learning_rate": 1.0844216198157026e-06, + "loss": 0.4652, + "step": 36016 + }, + { + "epoch": 0.7928021131410962, + "grad_norm": 1.5847280934149355, + "learning_rate": 1.0841999580763778e-06, + "loss": 0.3641, + "step": 36017 + }, + { + "epoch": 0.7928241250275149, + "grad_norm": 1.423423803427505, + "learning_rate": 1.0839783162388961e-06, + "loss": 0.2901, + "step": 36018 + }, + { + "epoch": 0.7928461369139336, + "grad_norm": 1.5080023383196794, + "learning_rate": 1.083756694304382e-06, + "loss": 0.3274, + "step": 36019 + }, + { + "epoch": 0.7928681488003522, + "grad_norm": 0.6479505830922443, + "learning_rate": 1.0835350922739647e-06, + "loss": 0.4444, + "step": 36020 + }, + { + "epoch": 0.7928901606867709, + "grad_norm": 1.734835094708742, + "learning_rate": 1.0833135101487673e-06, + "loss": 0.2742, + "step": 36021 + }, + { + "epoch": 0.7929121725731896, + "grad_norm": 1.2424501239434937, + "learning_rate": 1.0830919479299174e-06, + "loss": 0.1669, + "step": 36022 + }, + { + "epoch": 0.7929341844596082, + "grad_norm": 2.2829916941041515, + "learning_rate": 1.082870405618543e-06, + "loss": 0.2981, + "step": 36023 + }, + { + "epoch": 0.7929561963460269, + "grad_norm": 1.2150287138717102, + "learning_rate": 1.0826488832157666e-06, + "loss": 0.292, + "step": 36024 + }, + { + "epoch": 0.7929782082324455, + "grad_norm": 0.6275312432920522, + "learning_rate": 1.0824273807227171e-06, + "loss": 0.4407, + "step": 36025 + }, + { + "epoch": 0.7930002201188642, + "grad_norm": 0.6190098395458684, + "learning_rate": 1.0822058981405175e-06, + "loss": 0.4493, + "step": 36026 + }, + { + "epoch": 0.7930222320052829, + "grad_norm": 2.2101345878152006, + "learning_rate": 1.0819844354702953e-06, + "loss": 0.3351, + "step": 36027 + }, + { + "epoch": 0.7930442438917015, + "grad_norm": 2.578168877600796, + "learning_rate": 1.0817629927131767e-06, + "loss": 0.3348, + "step": 36028 + }, + { + "epoch": 0.7930662557781202, + "grad_norm": 1.2765437079930393, + "learning_rate": 1.0815415698702847e-06, + "loss": 0.2575, + "step": 36029 + }, + { + "epoch": 0.7930882676645389, + "grad_norm": 0.6381169312853849, + "learning_rate": 1.0813201669427475e-06, + "loss": 0.4715, + "step": 36030 + }, + { + "epoch": 0.7931102795509575, + "grad_norm": 1.2568927621008876, + "learning_rate": 1.0810987839316895e-06, + "loss": 0.2164, + "step": 36031 + }, + { + "epoch": 0.7931322914373762, + "grad_norm": 0.6207083924051673, + "learning_rate": 1.0808774208382322e-06, + "loss": 0.4543, + "step": 36032 + }, + { + "epoch": 0.7931543033237949, + "grad_norm": 1.590907213174647, + "learning_rate": 1.0806560776635066e-06, + "loss": 0.2707, + "step": 36033 + }, + { + "epoch": 0.7931763152102135, + "grad_norm": 1.9906984943008985, + "learning_rate": 1.0804347544086336e-06, + "loss": 0.2631, + "step": 36034 + }, + { + "epoch": 0.7931983270966322, + "grad_norm": 1.722219450592724, + "learning_rate": 1.0802134510747403e-06, + "loss": 0.2582, + "step": 36035 + }, + { + "epoch": 0.7932203389830509, + "grad_norm": 0.6161679567610355, + "learning_rate": 1.079992167662951e-06, + "loss": 0.4538, + "step": 36036 + }, + { + "epoch": 0.7932423508694695, + "grad_norm": 0.6106754300610018, + "learning_rate": 1.0797709041743888e-06, + "loss": 0.4286, + "step": 36037 + }, + { + "epoch": 0.7932643627558882, + "grad_norm": 0.6223947083967605, + "learning_rate": 1.0795496606101785e-06, + "loss": 0.431, + "step": 36038 + }, + { + "epoch": 0.7932863746423069, + "grad_norm": 6.1142468937463725, + "learning_rate": 1.0793284369714462e-06, + "loss": 0.3344, + "step": 36039 + }, + { + "epoch": 0.7933083865287255, + "grad_norm": 1.6060088930142644, + "learning_rate": 1.079107233259316e-06, + "loss": 0.2968, + "step": 36040 + }, + { + "epoch": 0.7933303984151442, + "grad_norm": 2.1906388554676552, + "learning_rate": 1.0788860494749126e-06, + "loss": 0.2733, + "step": 36041 + }, + { + "epoch": 0.7933524103015629, + "grad_norm": 2.2784060538751048, + "learning_rate": 1.0786648856193572e-06, + "loss": 0.3372, + "step": 36042 + }, + { + "epoch": 0.7933744221879815, + "grad_norm": 1.7634695465076256, + "learning_rate": 1.0784437416937753e-06, + "loss": 0.3048, + "step": 36043 + }, + { + "epoch": 0.7933964340744002, + "grad_norm": 1.3852749628117638, + "learning_rate": 1.0782226176992938e-06, + "loss": 0.3188, + "step": 36044 + }, + { + "epoch": 0.7934184459608189, + "grad_norm": 2.293613132950566, + "learning_rate": 1.078001513637032e-06, + "loss": 0.3292, + "step": 36045 + }, + { + "epoch": 0.7934404578472375, + "grad_norm": 1.5843706630541903, + "learning_rate": 1.0777804295081174e-06, + "loss": 0.2964, + "step": 36046 + }, + { + "epoch": 0.7934624697336562, + "grad_norm": 1.6058037304784782, + "learning_rate": 1.0775593653136706e-06, + "loss": 0.2396, + "step": 36047 + }, + { + "epoch": 0.7934844816200748, + "grad_norm": 1.4730615537712752, + "learning_rate": 1.0773383210548166e-06, + "loss": 0.2423, + "step": 36048 + }, + { + "epoch": 0.7935064935064935, + "grad_norm": 1.8353161625588508, + "learning_rate": 1.0771172967326804e-06, + "loss": 0.2535, + "step": 36049 + }, + { + "epoch": 0.7935285053929122, + "grad_norm": 1.8044664131412391, + "learning_rate": 1.0768962923483822e-06, + "loss": 0.3074, + "step": 36050 + }, + { + "epoch": 0.7935505172793308, + "grad_norm": 1.3004276479317591, + "learning_rate": 1.0766753079030478e-06, + "loss": 0.2868, + "step": 36051 + }, + { + "epoch": 0.7935725291657495, + "grad_norm": 1.755762544930124, + "learning_rate": 1.0764543433977987e-06, + "loss": 0.327, + "step": 36052 + }, + { + "epoch": 0.7935945410521682, + "grad_norm": 2.353350859350303, + "learning_rate": 1.076233398833758e-06, + "loss": 0.2781, + "step": 36053 + }, + { + "epoch": 0.7936165529385868, + "grad_norm": 3.466310534784717, + "learning_rate": 1.076012474212051e-06, + "loss": 0.3252, + "step": 36054 + }, + { + "epoch": 0.7936385648250055, + "grad_norm": 0.6031462510681038, + "learning_rate": 1.0757915695337967e-06, + "loss": 0.4356, + "step": 36055 + }, + { + "epoch": 0.7936605767114242, + "grad_norm": 4.278214583225155, + "learning_rate": 1.0755706848001218e-06, + "loss": 0.3499, + "step": 36056 + }, + { + "epoch": 0.7936825885978428, + "grad_norm": 1.6374900144883306, + "learning_rate": 1.0753498200121454e-06, + "loss": 0.2444, + "step": 36057 + }, + { + "epoch": 0.7937046004842615, + "grad_norm": 1.2774761270256512, + "learning_rate": 1.075128975170992e-06, + "loss": 0.3178, + "step": 36058 + }, + { + "epoch": 0.7937266123706802, + "grad_norm": 0.5879776202516578, + "learning_rate": 1.0749081502777847e-06, + "loss": 0.4197, + "step": 36059 + }, + { + "epoch": 0.7937486242570988, + "grad_norm": 1.7284932480187551, + "learning_rate": 1.0746873453336437e-06, + "loss": 0.3265, + "step": 36060 + }, + { + "epoch": 0.7937706361435175, + "grad_norm": 1.8060737554966664, + "learning_rate": 1.0744665603396937e-06, + "loss": 0.2722, + "step": 36061 + }, + { + "epoch": 0.7937926480299362, + "grad_norm": 0.6164332149064129, + "learning_rate": 1.0742457952970543e-06, + "loss": 0.4441, + "step": 36062 + }, + { + "epoch": 0.7938146599163548, + "grad_norm": 1.4579754422884044, + "learning_rate": 1.0740250502068482e-06, + "loss": 0.2884, + "step": 36063 + }, + { + "epoch": 0.7938366718027735, + "grad_norm": 1.700921055271991, + "learning_rate": 1.0738043250701997e-06, + "loss": 0.2912, + "step": 36064 + }, + { + "epoch": 0.7938586836891922, + "grad_norm": 0.5958419111671666, + "learning_rate": 1.073583619888227e-06, + "loss": 0.4347, + "step": 36065 + }, + { + "epoch": 0.7938806955756108, + "grad_norm": 1.7888519235366702, + "learning_rate": 1.073362934662055e-06, + "loss": 0.2887, + "step": 36066 + }, + { + "epoch": 0.7939027074620295, + "grad_norm": 2.5988835533104955, + "learning_rate": 1.073142269392804e-06, + "loss": 0.2894, + "step": 36067 + }, + { + "epoch": 0.7939247193484482, + "grad_norm": 2.354738987223162, + "learning_rate": 1.0729216240815926e-06, + "loss": 0.3046, + "step": 36068 + }, + { + "epoch": 0.7939467312348668, + "grad_norm": 2.265388566365093, + "learning_rate": 1.0727009987295477e-06, + "loss": 0.2777, + "step": 36069 + }, + { + "epoch": 0.7939687431212855, + "grad_norm": 11.05360713442488, + "learning_rate": 1.072480393337787e-06, + "loss": 0.2366, + "step": 36070 + }, + { + "epoch": 0.7939907550077041, + "grad_norm": 1.610883816522175, + "learning_rate": 1.0722598079074337e-06, + "loss": 0.3203, + "step": 36071 + }, + { + "epoch": 0.7940127668941228, + "grad_norm": 1.9831445670259327, + "learning_rate": 1.0720392424396075e-06, + "loss": 0.3179, + "step": 36072 + }, + { + "epoch": 0.7940347787805415, + "grad_norm": 1.2995612021684755, + "learning_rate": 1.0718186969354288e-06, + "loss": 0.2845, + "step": 36073 + }, + { + "epoch": 0.7940567906669601, + "grad_norm": 1.5387625254485802, + "learning_rate": 1.07159817139602e-06, + "loss": 0.3132, + "step": 36074 + }, + { + "epoch": 0.7940788025533788, + "grad_norm": 1.4585614774612865, + "learning_rate": 1.0713776658225005e-06, + "loss": 0.2987, + "step": 36075 + }, + { + "epoch": 0.7941008144397975, + "grad_norm": 1.3496044279520019, + "learning_rate": 1.0711571802159937e-06, + "loss": 0.2474, + "step": 36076 + }, + { + "epoch": 0.7941228263262161, + "grad_norm": 9.14777563105017, + "learning_rate": 1.0709367145776188e-06, + "loss": 0.2749, + "step": 36077 + }, + { + "epoch": 0.7941448382126348, + "grad_norm": 1.6712113473540113, + "learning_rate": 1.0707162689084944e-06, + "loss": 0.3247, + "step": 36078 + }, + { + "epoch": 0.7941668500990535, + "grad_norm": 1.8237967142206155, + "learning_rate": 1.0704958432097417e-06, + "loss": 0.3485, + "step": 36079 + }, + { + "epoch": 0.7941888619854721, + "grad_norm": 1.3955034156078094, + "learning_rate": 1.0702754374824826e-06, + "loss": 0.1575, + "step": 36080 + }, + { + "epoch": 0.7942108738718908, + "grad_norm": 3.096331186518572, + "learning_rate": 1.0700550517278374e-06, + "loss": 0.3008, + "step": 36081 + }, + { + "epoch": 0.7942328857583095, + "grad_norm": 1.5216053147400832, + "learning_rate": 1.0698346859469255e-06, + "loss": 0.3092, + "step": 36082 + }, + { + "epoch": 0.7942548976447281, + "grad_norm": 1.4471883290276375, + "learning_rate": 1.0696143401408648e-06, + "loss": 0.2889, + "step": 36083 + }, + { + "epoch": 0.7942769095311468, + "grad_norm": 2.6696080732505685, + "learning_rate": 1.0693940143107772e-06, + "loss": 0.3092, + "step": 36084 + }, + { + "epoch": 0.7942989214175655, + "grad_norm": 0.641280659574015, + "learning_rate": 1.069173708457784e-06, + "loss": 0.4423, + "step": 36085 + }, + { + "epoch": 0.7943209333039841, + "grad_norm": 1.8043303345388275, + "learning_rate": 1.0689534225830016e-06, + "loss": 0.3238, + "step": 36086 + }, + { + "epoch": 0.7943429451904028, + "grad_norm": 1.9063126681377145, + "learning_rate": 1.0687331566875526e-06, + "loss": 0.293, + "step": 36087 + }, + { + "epoch": 0.7943649570768215, + "grad_norm": 1.4719667780914683, + "learning_rate": 1.0685129107725534e-06, + "loss": 0.2774, + "step": 36088 + }, + { + "epoch": 0.7943869689632401, + "grad_norm": 8.575336289144822, + "learning_rate": 1.0682926848391256e-06, + "loss": 0.3263, + "step": 36089 + }, + { + "epoch": 0.7944089808496588, + "grad_norm": 2.037505694083407, + "learning_rate": 1.068072478888389e-06, + "loss": 0.3666, + "step": 36090 + }, + { + "epoch": 0.7944309927360775, + "grad_norm": 0.5963296403554957, + "learning_rate": 1.0678522929214608e-06, + "loss": 0.4056, + "step": 36091 + }, + { + "epoch": 0.7944530046224961, + "grad_norm": 0.6182585697775557, + "learning_rate": 1.0676321269394618e-06, + "loss": 0.4415, + "step": 36092 + }, + { + "epoch": 0.7944750165089148, + "grad_norm": 1.7236615634216004, + "learning_rate": 1.0674119809435091e-06, + "loss": 0.2809, + "step": 36093 + }, + { + "epoch": 0.7944970283953334, + "grad_norm": 0.6098800533960338, + "learning_rate": 1.0671918549347227e-06, + "loss": 0.4297, + "step": 36094 + }, + { + "epoch": 0.7945190402817521, + "grad_norm": 1.9114985346695967, + "learning_rate": 1.066971748914223e-06, + "loss": 0.294, + "step": 36095 + }, + { + "epoch": 0.7945410521681708, + "grad_norm": 1.969309736038688, + "learning_rate": 1.0667516628831253e-06, + "loss": 0.3561, + "step": 36096 + }, + { + "epoch": 0.7945630640545894, + "grad_norm": 1.7503349431660458, + "learning_rate": 1.0665315968425516e-06, + "loss": 0.2526, + "step": 36097 + }, + { + "epoch": 0.7945850759410081, + "grad_norm": 0.6083922753097503, + "learning_rate": 1.0663115507936172e-06, + "loss": 0.4369, + "step": 36098 + }, + { + "epoch": 0.7946070878274268, + "grad_norm": 1.6005228791191004, + "learning_rate": 1.0660915247374421e-06, + "loss": 0.3415, + "step": 36099 + }, + { + "epoch": 0.7946290997138454, + "grad_norm": 0.6347287604557758, + "learning_rate": 1.0658715186751457e-06, + "loss": 0.4877, + "step": 36100 + }, + { + "epoch": 0.7946511116002641, + "grad_norm": 0.613438398825476, + "learning_rate": 1.0656515326078437e-06, + "loss": 0.4487, + "step": 36101 + }, + { + "epoch": 0.7946731234866828, + "grad_norm": 1.6622045781404984, + "learning_rate": 1.0654315665366565e-06, + "loss": 0.3293, + "step": 36102 + }, + { + "epoch": 0.7946951353731014, + "grad_norm": 1.6288801063509182, + "learning_rate": 1.065211620462701e-06, + "loss": 0.29, + "step": 36103 + }, + { + "epoch": 0.7947171472595201, + "grad_norm": 1.4588319574798472, + "learning_rate": 1.0649916943870918e-06, + "loss": 0.3005, + "step": 36104 + }, + { + "epoch": 0.7947391591459388, + "grad_norm": 0.6161473298784786, + "learning_rate": 1.0647717883109537e-06, + "loss": 0.4683, + "step": 36105 + }, + { + "epoch": 0.7947611710323574, + "grad_norm": 2.1558966352604365, + "learning_rate": 1.0645519022353985e-06, + "loss": 0.3668, + "step": 36106 + }, + { + "epoch": 0.7947831829187761, + "grad_norm": 0.627557712205817, + "learning_rate": 1.0643320361615473e-06, + "loss": 0.4605, + "step": 36107 + }, + { + "epoch": 0.7948051948051948, + "grad_norm": 1.8023999238778183, + "learning_rate": 1.0641121900905161e-06, + "loss": 0.2449, + "step": 36108 + }, + { + "epoch": 0.7948272066916134, + "grad_norm": 2.8618327042982563, + "learning_rate": 1.0638923640234206e-06, + "loss": 0.2757, + "step": 36109 + }, + { + "epoch": 0.7948492185780321, + "grad_norm": 1.7291198663618095, + "learning_rate": 1.0636725579613794e-06, + "loss": 0.4016, + "step": 36110 + }, + { + "epoch": 0.7948712304644509, + "grad_norm": 0.6279608324986823, + "learning_rate": 1.0634527719055104e-06, + "loss": 0.4558, + "step": 36111 + }, + { + "epoch": 0.7948932423508694, + "grad_norm": 1.7854999529103752, + "learning_rate": 1.0632330058569312e-06, + "loss": 0.2947, + "step": 36112 + }, + { + "epoch": 0.7949152542372881, + "grad_norm": 1.788448997080572, + "learning_rate": 1.063013259816758e-06, + "loss": 0.3017, + "step": 36113 + }, + { + "epoch": 0.7949372661237069, + "grad_norm": 0.5960527157012215, + "learning_rate": 1.0627935337861062e-06, + "loss": 0.4475, + "step": 36114 + }, + { + "epoch": 0.7949592780101254, + "grad_norm": 2.3013730977567146, + "learning_rate": 1.0625738277660935e-06, + "loss": 0.3014, + "step": 36115 + }, + { + "epoch": 0.7949812898965442, + "grad_norm": 2.871510325000487, + "learning_rate": 1.0623541417578365e-06, + "loss": 0.2862, + "step": 36116 + }, + { + "epoch": 0.7950033017829627, + "grad_norm": 2.1164566744603115, + "learning_rate": 1.0621344757624535e-06, + "loss": 0.2741, + "step": 36117 + }, + { + "epoch": 0.7950253136693815, + "grad_norm": 1.884742764000557, + "learning_rate": 1.0619148297810595e-06, + "loss": 0.2245, + "step": 36118 + }, + { + "epoch": 0.7950473255558002, + "grad_norm": 2.012158247023052, + "learning_rate": 1.0616952038147693e-06, + "loss": 0.2983, + "step": 36119 + }, + { + "epoch": 0.7950693374422187, + "grad_norm": 1.6689831874752163, + "learning_rate": 1.0614755978647006e-06, + "loss": 0.2938, + "step": 36120 + }, + { + "epoch": 0.7950913493286375, + "grad_norm": 1.5834893912988086, + "learning_rate": 1.06125601193197e-06, + "loss": 0.3306, + "step": 36121 + }, + { + "epoch": 0.7951133612150562, + "grad_norm": 1.7566201908465564, + "learning_rate": 1.0610364460176936e-06, + "loss": 0.4112, + "step": 36122 + }, + { + "epoch": 0.7951353731014748, + "grad_norm": 1.8847086701774443, + "learning_rate": 1.0608169001229872e-06, + "loss": 0.4152, + "step": 36123 + }, + { + "epoch": 0.7951573849878935, + "grad_norm": 1.4586565483889815, + "learning_rate": 1.060597374248965e-06, + "loss": 0.2392, + "step": 36124 + }, + { + "epoch": 0.7951793968743122, + "grad_norm": 1.743943357002045, + "learning_rate": 1.0603778683967441e-06, + "loss": 0.354, + "step": 36125 + }, + { + "epoch": 0.7952014087607308, + "grad_norm": 2.1668993959129637, + "learning_rate": 1.0601583825674412e-06, + "loss": 0.2243, + "step": 36126 + }, + { + "epoch": 0.7952234206471495, + "grad_norm": 1.8104456572007317, + "learning_rate": 1.0599389167621693e-06, + "loss": 0.368, + "step": 36127 + }, + { + "epoch": 0.7952454325335682, + "grad_norm": 1.5495247189955605, + "learning_rate": 1.0597194709820463e-06, + "loss": 0.2515, + "step": 36128 + }, + { + "epoch": 0.7952674444199868, + "grad_norm": 1.9091706473480508, + "learning_rate": 1.059500045228185e-06, + "loss": 0.2752, + "step": 36129 + }, + { + "epoch": 0.7952894563064055, + "grad_norm": 4.3655825897092315, + "learning_rate": 1.0592806395017024e-06, + "loss": 0.2784, + "step": 36130 + }, + { + "epoch": 0.7953114681928242, + "grad_norm": 5.864763226395141, + "learning_rate": 1.0590612538037143e-06, + "loss": 0.3668, + "step": 36131 + }, + { + "epoch": 0.7953334800792428, + "grad_norm": 1.294714844891834, + "learning_rate": 1.058841888135333e-06, + "loss": 0.2614, + "step": 36132 + }, + { + "epoch": 0.7953554919656615, + "grad_norm": 1.7751906547405716, + "learning_rate": 1.0586225424976765e-06, + "loss": 0.2974, + "step": 36133 + }, + { + "epoch": 0.7953775038520802, + "grad_norm": 1.7513025847500714, + "learning_rate": 1.0584032168918574e-06, + "loss": 0.2421, + "step": 36134 + }, + { + "epoch": 0.7953995157384988, + "grad_norm": 1.8958784253794556, + "learning_rate": 1.0581839113189901e-06, + "loss": 0.2188, + "step": 36135 + }, + { + "epoch": 0.7954215276249175, + "grad_norm": 1.6270144391678485, + "learning_rate": 1.0579646257801923e-06, + "loss": 0.3089, + "step": 36136 + }, + { + "epoch": 0.7954435395113362, + "grad_norm": 1.5654398228888986, + "learning_rate": 1.0577453602765752e-06, + "loss": 0.308, + "step": 36137 + }, + { + "epoch": 0.7954655513977548, + "grad_norm": 1.8521055348061939, + "learning_rate": 1.0575261148092552e-06, + "loss": 0.3095, + "step": 36138 + }, + { + "epoch": 0.7954875632841735, + "grad_norm": 1.6086782672300688, + "learning_rate": 1.0573068893793453e-06, + "loss": 0.2404, + "step": 36139 + }, + { + "epoch": 0.7955095751705921, + "grad_norm": 1.4840334423233619, + "learning_rate": 1.0570876839879597e-06, + "loss": 0.2753, + "step": 36140 + }, + { + "epoch": 0.7955315870570108, + "grad_norm": 4.646375945663697, + "learning_rate": 1.0568684986362148e-06, + "loss": 0.2362, + "step": 36141 + }, + { + "epoch": 0.7955535989434295, + "grad_norm": 1.4219978357588783, + "learning_rate": 1.0566493333252214e-06, + "loss": 0.3285, + "step": 36142 + }, + { + "epoch": 0.7955756108298481, + "grad_norm": 1.6767177856096493, + "learning_rate": 1.0564301880560962e-06, + "loss": 0.2612, + "step": 36143 + }, + { + "epoch": 0.7955976227162668, + "grad_norm": 1.6827459115207386, + "learning_rate": 1.0562110628299515e-06, + "loss": 0.3465, + "step": 36144 + }, + { + "epoch": 0.7956196346026855, + "grad_norm": 1.7908856521862784, + "learning_rate": 1.0559919576479e-06, + "loss": 0.3029, + "step": 36145 + }, + { + "epoch": 0.7956416464891041, + "grad_norm": 2.069217822846805, + "learning_rate": 1.0557728725110567e-06, + "loss": 0.2673, + "step": 36146 + }, + { + "epoch": 0.7956636583755228, + "grad_norm": 3.2924920756415132, + "learning_rate": 1.055553807420534e-06, + "loss": 0.3482, + "step": 36147 + }, + { + "epoch": 0.7956856702619415, + "grad_norm": 1.5868520973316123, + "learning_rate": 1.0553347623774484e-06, + "loss": 0.3052, + "step": 36148 + }, + { + "epoch": 0.7957076821483601, + "grad_norm": 1.5786389173742394, + "learning_rate": 1.0551157373829102e-06, + "loss": 0.2594, + "step": 36149 + }, + { + "epoch": 0.7957296940347788, + "grad_norm": 5.653681673625923, + "learning_rate": 1.0548967324380328e-06, + "loss": 0.2576, + "step": 36150 + }, + { + "epoch": 0.7957517059211975, + "grad_norm": 1.8686911168008635, + "learning_rate": 1.054677747543929e-06, + "loss": 0.2336, + "step": 36151 + }, + { + "epoch": 0.7957737178076161, + "grad_norm": 1.5044111049232127, + "learning_rate": 1.0544587827017128e-06, + "loss": 0.2339, + "step": 36152 + }, + { + "epoch": 0.7957957296940348, + "grad_norm": 1.9925590462003777, + "learning_rate": 1.054239837912498e-06, + "loss": 0.2389, + "step": 36153 + }, + { + "epoch": 0.7958177415804535, + "grad_norm": 1.8911331633062083, + "learning_rate": 1.0540209131773966e-06, + "loss": 0.2565, + "step": 36154 + }, + { + "epoch": 0.7958397534668721, + "grad_norm": 1.743246563911292, + "learning_rate": 1.0538020084975193e-06, + "loss": 0.2537, + "step": 36155 + }, + { + "epoch": 0.7958617653532908, + "grad_norm": 5.321101327843579, + "learning_rate": 1.0535831238739803e-06, + "loss": 0.2779, + "step": 36156 + }, + { + "epoch": 0.7958837772397095, + "grad_norm": 5.554988587170588, + "learning_rate": 1.0533642593078918e-06, + "loss": 0.3599, + "step": 36157 + }, + { + "epoch": 0.7959057891261281, + "grad_norm": 1.6013764698565145, + "learning_rate": 1.053145414800368e-06, + "loss": 0.313, + "step": 36158 + }, + { + "epoch": 0.7959278010125468, + "grad_norm": 1.9370033791842058, + "learning_rate": 1.0529265903525192e-06, + "loss": 0.2246, + "step": 36159 + }, + { + "epoch": 0.7959498128989655, + "grad_norm": 2.069474644778665, + "learning_rate": 1.0527077859654567e-06, + "loss": 0.372, + "step": 36160 + }, + { + "epoch": 0.7959718247853841, + "grad_norm": 4.273645143824785, + "learning_rate": 1.052489001640294e-06, + "loss": 0.2303, + "step": 36161 + }, + { + "epoch": 0.7959938366718028, + "grad_norm": 1.4331648177579812, + "learning_rate": 1.052270237378143e-06, + "loss": 0.2524, + "step": 36162 + }, + { + "epoch": 0.7960158485582215, + "grad_norm": 2.0260484991146903, + "learning_rate": 1.0520514931801162e-06, + "loss": 0.3064, + "step": 36163 + }, + { + "epoch": 0.7960378604446401, + "grad_norm": 1.4654353066767298, + "learning_rate": 1.0518327690473246e-06, + "loss": 0.267, + "step": 36164 + }, + { + "epoch": 0.7960598723310588, + "grad_norm": 1.3265706014847316, + "learning_rate": 1.0516140649808788e-06, + "loss": 0.2765, + "step": 36165 + }, + { + "epoch": 0.7960818842174774, + "grad_norm": 1.5272127891751517, + "learning_rate": 1.0513953809818912e-06, + "loss": 0.2323, + "step": 36166 + }, + { + "epoch": 0.7961038961038961, + "grad_norm": 1.8301686948487572, + "learning_rate": 1.0511767170514747e-06, + "loss": 0.323, + "step": 36167 + }, + { + "epoch": 0.7961259079903148, + "grad_norm": 2.9080051850779434, + "learning_rate": 1.050958073190738e-06, + "loss": 0.3177, + "step": 36168 + }, + { + "epoch": 0.7961479198767334, + "grad_norm": 1.506303736674464, + "learning_rate": 1.050739449400795e-06, + "loss": 0.214, + "step": 36169 + }, + { + "epoch": 0.7961699317631521, + "grad_norm": 1.4739578675489269, + "learning_rate": 1.0505208456827544e-06, + "loss": 0.229, + "step": 36170 + }, + { + "epoch": 0.7961919436495708, + "grad_norm": 2.2375695028240936, + "learning_rate": 1.0503022620377284e-06, + "loss": 0.2674, + "step": 36171 + }, + { + "epoch": 0.7962139555359894, + "grad_norm": 2.2477794965235507, + "learning_rate": 1.050083698466829e-06, + "loss": 0.259, + "step": 36172 + }, + { + "epoch": 0.7962359674224081, + "grad_norm": 5.775610382198101, + "learning_rate": 1.0498651549711647e-06, + "loss": 0.3985, + "step": 36173 + }, + { + "epoch": 0.7962579793088268, + "grad_norm": 1.4192242844539336, + "learning_rate": 1.0496466315518489e-06, + "loss": 0.2875, + "step": 36174 + }, + { + "epoch": 0.7962799911952454, + "grad_norm": 1.6834950097163446, + "learning_rate": 1.0494281282099895e-06, + "loss": 0.2761, + "step": 36175 + }, + { + "epoch": 0.7963020030816641, + "grad_norm": 2.029475507622783, + "learning_rate": 1.0492096449466981e-06, + "loss": 0.3085, + "step": 36176 + }, + { + "epoch": 0.7963240149680828, + "grad_norm": 0.6323456586659434, + "learning_rate": 1.048991181763087e-06, + "loss": 0.4642, + "step": 36177 + }, + { + "epoch": 0.7963460268545014, + "grad_norm": 0.6348271840942589, + "learning_rate": 1.0487727386602637e-06, + "loss": 0.4311, + "step": 36178 + }, + { + "epoch": 0.7963680387409201, + "grad_norm": 3.0655223873019297, + "learning_rate": 1.0485543156393407e-06, + "loss": 0.2699, + "step": 36179 + }, + { + "epoch": 0.7963900506273388, + "grad_norm": 2.431354226318806, + "learning_rate": 1.0483359127014259e-06, + "loss": 0.2542, + "step": 36180 + }, + { + "epoch": 0.7964120625137574, + "grad_norm": 2.77439687099803, + "learning_rate": 1.0481175298476315e-06, + "loss": 0.2739, + "step": 36181 + }, + { + "epoch": 0.7964340744001761, + "grad_norm": 1.6978725180741918, + "learning_rate": 1.047899167079065e-06, + "loss": 0.2929, + "step": 36182 + }, + { + "epoch": 0.7964560862865948, + "grad_norm": 3.8893605794958472, + "learning_rate": 1.0476808243968384e-06, + "loss": 0.2905, + "step": 36183 + }, + { + "epoch": 0.7964780981730134, + "grad_norm": 1.3753872506261255, + "learning_rate": 1.0474625018020612e-06, + "loss": 0.2608, + "step": 36184 + }, + { + "epoch": 0.7965001100594321, + "grad_norm": 1.609554791120855, + "learning_rate": 1.0472441992958426e-06, + "loss": 0.2958, + "step": 36185 + }, + { + "epoch": 0.7965221219458508, + "grad_norm": 0.5999745254543191, + "learning_rate": 1.0470259168792906e-06, + "loss": 0.4396, + "step": 36186 + }, + { + "epoch": 0.7965441338322694, + "grad_norm": 1.2845390059037487, + "learning_rate": 1.0468076545535167e-06, + "loss": 0.2003, + "step": 36187 + }, + { + "epoch": 0.7965661457186881, + "grad_norm": 2.0500011377102645, + "learning_rate": 1.0465894123196285e-06, + "loss": 0.2577, + "step": 36188 + }, + { + "epoch": 0.7965881576051067, + "grad_norm": 1.7028707820625364, + "learning_rate": 1.0463711901787383e-06, + "loss": 0.3521, + "step": 36189 + }, + { + "epoch": 0.7966101694915254, + "grad_norm": 1.822629890701222, + "learning_rate": 1.0461529881319527e-06, + "loss": 0.3277, + "step": 36190 + }, + { + "epoch": 0.7966321813779441, + "grad_norm": 1.9091076490740564, + "learning_rate": 1.0459348061803793e-06, + "loss": 0.2906, + "step": 36191 + }, + { + "epoch": 0.7966541932643627, + "grad_norm": 1.428844076327494, + "learning_rate": 1.0457166443251294e-06, + "loss": 0.2011, + "step": 36192 + }, + { + "epoch": 0.7966762051507814, + "grad_norm": 1.7350900880008444, + "learning_rate": 1.045498502567311e-06, + "loss": 0.2792, + "step": 36193 + }, + { + "epoch": 0.7966982170372001, + "grad_norm": 9.064818345903104, + "learning_rate": 1.045280380908034e-06, + "loss": 0.3402, + "step": 36194 + }, + { + "epoch": 0.7967202289236187, + "grad_norm": 2.0002403093928907, + "learning_rate": 1.0450622793484067e-06, + "loss": 0.3304, + "step": 36195 + }, + { + "epoch": 0.7967422408100374, + "grad_norm": 2.5498339421518894, + "learning_rate": 1.0448441978895346e-06, + "loss": 0.2633, + "step": 36196 + }, + { + "epoch": 0.7967642526964561, + "grad_norm": 1.4394924365533077, + "learning_rate": 1.0446261365325288e-06, + "loss": 0.2674, + "step": 36197 + }, + { + "epoch": 0.7967862645828747, + "grad_norm": 3.0058342793483184, + "learning_rate": 1.0444080952784974e-06, + "loss": 0.3446, + "step": 36198 + }, + { + "epoch": 0.7968082764692934, + "grad_norm": 0.6278791012879831, + "learning_rate": 1.0441900741285494e-06, + "loss": 0.4495, + "step": 36199 + }, + { + "epoch": 0.7968302883557121, + "grad_norm": 1.4139944665938065, + "learning_rate": 1.0439720730837916e-06, + "loss": 0.2513, + "step": 36200 + }, + { + "epoch": 0.7968523002421307, + "grad_norm": 1.1861545435840246, + "learning_rate": 1.043754092145331e-06, + "loss": 0.1764, + "step": 36201 + }, + { + "epoch": 0.7968743121285494, + "grad_norm": 0.6316445978690893, + "learning_rate": 1.043536131314276e-06, + "loss": 0.4782, + "step": 36202 + }, + { + "epoch": 0.7968963240149681, + "grad_norm": 2.371176567512334, + "learning_rate": 1.0433181905917372e-06, + "loss": 0.3218, + "step": 36203 + }, + { + "epoch": 0.7969183359013867, + "grad_norm": 1.77980612523152, + "learning_rate": 1.0431002699788184e-06, + "loss": 0.3075, + "step": 36204 + }, + { + "epoch": 0.7969403477878054, + "grad_norm": 0.6206583421182176, + "learning_rate": 1.0428823694766304e-06, + "loss": 0.4718, + "step": 36205 + }, + { + "epoch": 0.7969623596742241, + "grad_norm": 1.5772836533760934, + "learning_rate": 1.0426644890862775e-06, + "loss": 0.239, + "step": 36206 + }, + { + "epoch": 0.7969843715606427, + "grad_norm": 0.5955624232698106, + "learning_rate": 1.0424466288088692e-06, + "loss": 0.46, + "step": 36207 + }, + { + "epoch": 0.7970063834470614, + "grad_norm": 0.588794517108855, + "learning_rate": 1.042228788645513e-06, + "loss": 0.4276, + "step": 36208 + }, + { + "epoch": 0.7970283953334801, + "grad_norm": 1.977451112793159, + "learning_rate": 1.0420109685973146e-06, + "loss": 0.3289, + "step": 36209 + }, + { + "epoch": 0.7970504072198987, + "grad_norm": 1.5811065204867123, + "learning_rate": 1.0417931686653826e-06, + "loss": 0.3274, + "step": 36210 + }, + { + "epoch": 0.7970724191063174, + "grad_norm": 1.9445778376692118, + "learning_rate": 1.0415753888508217e-06, + "loss": 0.2261, + "step": 36211 + }, + { + "epoch": 0.797094430992736, + "grad_norm": 1.5896682916048244, + "learning_rate": 1.0413576291547417e-06, + "loss": 0.2371, + "step": 36212 + }, + { + "epoch": 0.7971164428791547, + "grad_norm": 1.364589942980507, + "learning_rate": 1.0411398895782466e-06, + "loss": 0.2762, + "step": 36213 + }, + { + "epoch": 0.7971384547655734, + "grad_norm": 0.6239407379022647, + "learning_rate": 1.0409221701224441e-06, + "loss": 0.4642, + "step": 36214 + }, + { + "epoch": 0.797160466651992, + "grad_norm": 2.8520672222386847, + "learning_rate": 1.0407044707884424e-06, + "loss": 0.295, + "step": 36215 + }, + { + "epoch": 0.7971824785384107, + "grad_norm": 0.623900678710061, + "learning_rate": 1.0404867915773454e-06, + "loss": 0.4474, + "step": 36216 + }, + { + "epoch": 0.7972044904248294, + "grad_norm": 5.154537449596879, + "learning_rate": 1.0402691324902615e-06, + "loss": 0.3499, + "step": 36217 + }, + { + "epoch": 0.797226502311248, + "grad_norm": 1.881655788970445, + "learning_rate": 1.0400514935282946e-06, + "loss": 0.4201, + "step": 36218 + }, + { + "epoch": 0.7972485141976667, + "grad_norm": 1.8493322236769942, + "learning_rate": 1.0398338746925524e-06, + "loss": 0.2723, + "step": 36219 + }, + { + "epoch": 0.7972705260840854, + "grad_norm": 0.6222325830485775, + "learning_rate": 1.0396162759841416e-06, + "loss": 0.4454, + "step": 36220 + }, + { + "epoch": 0.797292537970504, + "grad_norm": 1.3771077205744569, + "learning_rate": 1.0393986974041664e-06, + "loss": 0.2998, + "step": 36221 + }, + { + "epoch": 0.7973145498569227, + "grad_norm": 1.53450538858615, + "learning_rate": 1.0391811389537349e-06, + "loss": 0.345, + "step": 36222 + }, + { + "epoch": 0.7973365617433414, + "grad_norm": 1.7339823868153046, + "learning_rate": 1.0389636006339498e-06, + "loss": 0.2739, + "step": 36223 + }, + { + "epoch": 0.79735857362976, + "grad_norm": 1.371413343230514, + "learning_rate": 1.0387460824459184e-06, + "loss": 0.3097, + "step": 36224 + }, + { + "epoch": 0.7973805855161787, + "grad_norm": 28.622203598864147, + "learning_rate": 1.0385285843907478e-06, + "loss": 0.3502, + "step": 36225 + }, + { + "epoch": 0.7974025974025974, + "grad_norm": 1.4418866641347274, + "learning_rate": 1.0383111064695412e-06, + "loss": 0.2846, + "step": 36226 + }, + { + "epoch": 0.797424609289016, + "grad_norm": 1.6399120250010668, + "learning_rate": 1.0380936486834037e-06, + "loss": 0.2957, + "step": 36227 + }, + { + "epoch": 0.7974466211754347, + "grad_norm": 1.5198918569445752, + "learning_rate": 1.0378762110334411e-06, + "loss": 0.1994, + "step": 36228 + }, + { + "epoch": 0.7974686330618534, + "grad_norm": 0.5981449286572263, + "learning_rate": 1.0376587935207588e-06, + "loss": 0.4384, + "step": 36229 + }, + { + "epoch": 0.797490644948272, + "grad_norm": 2.710647174546372, + "learning_rate": 1.037441396146463e-06, + "loss": 0.2017, + "step": 36230 + }, + { + "epoch": 0.7975126568346907, + "grad_norm": 1.9651839151251074, + "learning_rate": 1.0372240189116578e-06, + "loss": 0.2737, + "step": 36231 + }, + { + "epoch": 0.7975346687211095, + "grad_norm": 0.6116979538699018, + "learning_rate": 1.0370066618174458e-06, + "loss": 0.4732, + "step": 36232 + }, + { + "epoch": 0.797556680607528, + "grad_norm": 1.3161529550926705, + "learning_rate": 1.0367893248649335e-06, + "loss": 0.3348, + "step": 36233 + }, + { + "epoch": 0.7975786924939468, + "grad_norm": 7.4447355071012735, + "learning_rate": 1.036572008055226e-06, + "loss": 0.2775, + "step": 36234 + }, + { + "epoch": 0.7976007043803653, + "grad_norm": 1.5413362902030114, + "learning_rate": 1.036354711389428e-06, + "loss": 0.2469, + "step": 36235 + }, + { + "epoch": 0.797622716266784, + "grad_norm": 1.5149698408997192, + "learning_rate": 1.0361374348686437e-06, + "loss": 0.2628, + "step": 36236 + }, + { + "epoch": 0.7976447281532028, + "grad_norm": 1.947085722975701, + "learning_rate": 1.0359201784939755e-06, + "loss": 0.2785, + "step": 36237 + }, + { + "epoch": 0.7976667400396213, + "grad_norm": 1.7926897791398848, + "learning_rate": 1.0357029422665282e-06, + "loss": 0.2659, + "step": 36238 + }, + { + "epoch": 0.79768875192604, + "grad_norm": 1.8072234964403122, + "learning_rate": 1.0354857261874075e-06, + "loss": 0.2395, + "step": 36239 + }, + { + "epoch": 0.7977107638124588, + "grad_norm": 0.641994532878541, + "learning_rate": 1.0352685302577177e-06, + "loss": 0.448, + "step": 36240 + }, + { + "epoch": 0.7977327756988773, + "grad_norm": 2.0024261072371203, + "learning_rate": 1.0350513544785617e-06, + "loss": 0.3545, + "step": 36241 + }, + { + "epoch": 0.797754787585296, + "grad_norm": 1.9549516027596094, + "learning_rate": 1.0348341988510413e-06, + "loss": 0.3218, + "step": 36242 + }, + { + "epoch": 0.7977767994717148, + "grad_norm": 2.4823452897124114, + "learning_rate": 1.0346170633762626e-06, + "loss": 0.3329, + "step": 36243 + }, + { + "epoch": 0.7977988113581334, + "grad_norm": 1.85255052050861, + "learning_rate": 1.0343999480553297e-06, + "loss": 0.2627, + "step": 36244 + }, + { + "epoch": 0.7978208232445521, + "grad_norm": 2.5646784546758803, + "learning_rate": 1.0341828528893433e-06, + "loss": 0.3, + "step": 36245 + }, + { + "epoch": 0.7978428351309708, + "grad_norm": 2.245441737508792, + "learning_rate": 1.0339657778794104e-06, + "loss": 0.2428, + "step": 36246 + }, + { + "epoch": 0.7978648470173894, + "grad_norm": 2.2977374213420343, + "learning_rate": 1.0337487230266307e-06, + "loss": 0.1971, + "step": 36247 + }, + { + "epoch": 0.7978868589038081, + "grad_norm": 2.0226851348894543, + "learning_rate": 1.03353168833211e-06, + "loss": 0.358, + "step": 36248 + }, + { + "epoch": 0.7979088707902268, + "grad_norm": 1.6423754272659867, + "learning_rate": 1.0333146737969497e-06, + "loss": 0.3741, + "step": 36249 + }, + { + "epoch": 0.7979308826766454, + "grad_norm": 1.7054795341559659, + "learning_rate": 1.0330976794222525e-06, + "loss": 0.1867, + "step": 36250 + }, + { + "epoch": 0.7979528945630641, + "grad_norm": 1.7239218820565285, + "learning_rate": 1.0328807052091245e-06, + "loss": 0.3488, + "step": 36251 + }, + { + "epoch": 0.7979749064494828, + "grad_norm": 1.8150148204877898, + "learning_rate": 1.032663751158664e-06, + "loss": 0.3784, + "step": 36252 + }, + { + "epoch": 0.7979969183359014, + "grad_norm": 10.891753586623253, + "learning_rate": 1.0324468172719776e-06, + "loss": 0.3066, + "step": 36253 + }, + { + "epoch": 0.7980189302223201, + "grad_norm": 1.7565988580786602, + "learning_rate": 1.0322299035501642e-06, + "loss": 0.2336, + "step": 36254 + }, + { + "epoch": 0.7980409421087388, + "grad_norm": 3.4846478442237374, + "learning_rate": 1.0320130099943287e-06, + "loss": 0.2756, + "step": 36255 + }, + { + "epoch": 0.7980629539951574, + "grad_norm": 1.2789539724007517, + "learning_rate": 1.031796136605574e-06, + "loss": 0.1937, + "step": 36256 + }, + { + "epoch": 0.7980849658815761, + "grad_norm": 1.578156826937793, + "learning_rate": 1.0315792833849997e-06, + "loss": 0.257, + "step": 36257 + }, + { + "epoch": 0.7981069777679947, + "grad_norm": 1.4015658019280333, + "learning_rate": 1.031362450333711e-06, + "loss": 0.3369, + "step": 36258 + }, + { + "epoch": 0.7981289896544134, + "grad_norm": 1.4880513752518707, + "learning_rate": 1.0311456374528073e-06, + "loss": 0.2895, + "step": 36259 + }, + { + "epoch": 0.7981510015408321, + "grad_norm": 1.6488599991499588, + "learning_rate": 1.0309288447433912e-06, + "loss": 0.2918, + "step": 36260 + }, + { + "epoch": 0.7981730134272507, + "grad_norm": 4.078598488077781, + "learning_rate": 1.0307120722065667e-06, + "loss": 0.2291, + "step": 36261 + }, + { + "epoch": 0.7981950253136694, + "grad_norm": 2.063931945949428, + "learning_rate": 1.030495319843432e-06, + "loss": 0.3395, + "step": 36262 + }, + { + "epoch": 0.7982170372000881, + "grad_norm": 10.286570793467442, + "learning_rate": 1.030278587655092e-06, + "loss": 0.3632, + "step": 36263 + }, + { + "epoch": 0.7982390490865067, + "grad_norm": 1.5418203686932672, + "learning_rate": 1.0300618756426456e-06, + "loss": 0.2857, + "step": 36264 + }, + { + "epoch": 0.7982610609729254, + "grad_norm": 1.6301414238049268, + "learning_rate": 1.0298451838071948e-06, + "loss": 0.2514, + "step": 36265 + }, + { + "epoch": 0.7982830728593441, + "grad_norm": 1.9836290028125694, + "learning_rate": 1.0296285121498435e-06, + "loss": 0.2391, + "step": 36266 + }, + { + "epoch": 0.7983050847457627, + "grad_norm": 1.7788670781047793, + "learning_rate": 1.0294118606716906e-06, + "loss": 0.2942, + "step": 36267 + }, + { + "epoch": 0.7983270966321814, + "grad_norm": 1.4650564513732476, + "learning_rate": 1.0291952293738367e-06, + "loss": 0.2385, + "step": 36268 + }, + { + "epoch": 0.7983491085186001, + "grad_norm": 1.863855069705415, + "learning_rate": 1.0289786182573836e-06, + "loss": 0.3059, + "step": 36269 + }, + { + "epoch": 0.7983711204050187, + "grad_norm": 1.4560409027653731, + "learning_rate": 1.0287620273234317e-06, + "loss": 0.2126, + "step": 36270 + }, + { + "epoch": 0.7983931322914374, + "grad_norm": 1.573339030788964, + "learning_rate": 1.028545456573084e-06, + "loss": 0.3182, + "step": 36271 + }, + { + "epoch": 0.7984151441778561, + "grad_norm": 3.272810520348723, + "learning_rate": 1.0283289060074397e-06, + "loss": 0.328, + "step": 36272 + }, + { + "epoch": 0.7984371560642747, + "grad_norm": 1.731907731994155, + "learning_rate": 1.0281123756275979e-06, + "loss": 0.3078, + "step": 36273 + }, + { + "epoch": 0.7984591679506934, + "grad_norm": 2.3741040220420464, + "learning_rate": 1.0278958654346605e-06, + "loss": 0.3306, + "step": 36274 + }, + { + "epoch": 0.7984811798371121, + "grad_norm": 1.791905593359605, + "learning_rate": 1.0276793754297281e-06, + "loss": 0.3316, + "step": 36275 + }, + { + "epoch": 0.7985031917235307, + "grad_norm": 1.408430800878039, + "learning_rate": 1.0274629056139023e-06, + "loss": 0.301, + "step": 36276 + }, + { + "epoch": 0.7985252036099494, + "grad_norm": 0.5793474076800154, + "learning_rate": 1.0272464559882816e-06, + "loss": 0.4239, + "step": 36277 + }, + { + "epoch": 0.7985472154963681, + "grad_norm": 3.180608238597486, + "learning_rate": 1.0270300265539646e-06, + "loss": 0.2655, + "step": 36278 + }, + { + "epoch": 0.7985692273827867, + "grad_norm": 2.348247431168133, + "learning_rate": 1.0268136173120535e-06, + "loss": 0.2147, + "step": 36279 + }, + { + "epoch": 0.7985912392692054, + "grad_norm": 0.6083442252636283, + "learning_rate": 1.0265972282636476e-06, + "loss": 0.4576, + "step": 36280 + }, + { + "epoch": 0.798613251155624, + "grad_norm": 1.2643028843493647, + "learning_rate": 1.0263808594098479e-06, + "loss": 0.2241, + "step": 36281 + }, + { + "epoch": 0.7986352630420427, + "grad_norm": 1.477832925314783, + "learning_rate": 1.0261645107517532e-06, + "loss": 0.2889, + "step": 36282 + }, + { + "epoch": 0.7986572749284614, + "grad_norm": 2.1080326573312758, + "learning_rate": 1.0259481822904616e-06, + "loss": 0.3866, + "step": 36283 + }, + { + "epoch": 0.79867928681488, + "grad_norm": 1.6715767386555778, + "learning_rate": 1.025731874027075e-06, + "loss": 0.2828, + "step": 36284 + }, + { + "epoch": 0.7987012987012987, + "grad_norm": 2.7538263228490965, + "learning_rate": 1.0255155859626896e-06, + "loss": 0.2695, + "step": 36285 + }, + { + "epoch": 0.7987233105877174, + "grad_norm": 0.6125476789141978, + "learning_rate": 1.0252993180984077e-06, + "loss": 0.4658, + "step": 36286 + }, + { + "epoch": 0.798745322474136, + "grad_norm": 1.409304288851476, + "learning_rate": 1.0250830704353283e-06, + "loss": 0.2049, + "step": 36287 + }, + { + "epoch": 0.7987673343605547, + "grad_norm": 1.5560345461633975, + "learning_rate": 1.0248668429745484e-06, + "loss": 0.3042, + "step": 36288 + }, + { + "epoch": 0.7987893462469734, + "grad_norm": 0.6106568859742197, + "learning_rate": 1.0246506357171693e-06, + "loss": 0.4508, + "step": 36289 + }, + { + "epoch": 0.798811358133392, + "grad_norm": 1.3731903845283833, + "learning_rate": 1.0244344486642872e-06, + "loss": 0.2986, + "step": 36290 + }, + { + "epoch": 0.7988333700198107, + "grad_norm": 2.4670091103428273, + "learning_rate": 1.0242182818170027e-06, + "loss": 0.2694, + "step": 36291 + }, + { + "epoch": 0.7988553819062294, + "grad_norm": 2.6319076044937124, + "learning_rate": 1.0240021351764157e-06, + "loss": 0.3988, + "step": 36292 + }, + { + "epoch": 0.798877393792648, + "grad_norm": 2.925599215314494, + "learning_rate": 1.0237860087436214e-06, + "loss": 0.3028, + "step": 36293 + }, + { + "epoch": 0.7988994056790667, + "grad_norm": 1.6873840249934016, + "learning_rate": 1.0235699025197215e-06, + "loss": 0.2608, + "step": 36294 + }, + { + "epoch": 0.7989214175654854, + "grad_norm": 2.0022795820691135, + "learning_rate": 1.0233538165058116e-06, + "loss": 0.3399, + "step": 36295 + }, + { + "epoch": 0.798943429451904, + "grad_norm": 1.4711254828082097, + "learning_rate": 1.0231377507029915e-06, + "loss": 0.2339, + "step": 36296 + }, + { + "epoch": 0.7989654413383227, + "grad_norm": 0.6266482119718495, + "learning_rate": 1.0229217051123602e-06, + "loss": 0.4517, + "step": 36297 + }, + { + "epoch": 0.7989874532247414, + "grad_norm": 2.222842864895868, + "learning_rate": 1.0227056797350132e-06, + "loss": 0.3228, + "step": 36298 + }, + { + "epoch": 0.79900946511116, + "grad_norm": 1.859642511058009, + "learning_rate": 1.0224896745720513e-06, + "loss": 0.3299, + "step": 36299 + }, + { + "epoch": 0.7990314769975787, + "grad_norm": 2.264937184351912, + "learning_rate": 1.0222736896245694e-06, + "loss": 0.3581, + "step": 36300 + }, + { + "epoch": 0.7990534888839974, + "grad_norm": 4.332959318419312, + "learning_rate": 1.0220577248936669e-06, + "loss": 0.2906, + "step": 36301 + }, + { + "epoch": 0.799075500770416, + "grad_norm": 1.5394046668796664, + "learning_rate": 1.0218417803804426e-06, + "loss": 0.3794, + "step": 36302 + }, + { + "epoch": 0.7990975126568347, + "grad_norm": 0.6077138283302618, + "learning_rate": 1.0216258560859926e-06, + "loss": 0.4319, + "step": 36303 + }, + { + "epoch": 0.7991195245432534, + "grad_norm": 2.02046413421721, + "learning_rate": 1.0214099520114129e-06, + "loss": 0.3131, + "step": 36304 + }, + { + "epoch": 0.799141536429672, + "grad_norm": 1.5014107867383337, + "learning_rate": 1.0211940681578026e-06, + "loss": 0.2728, + "step": 36305 + }, + { + "epoch": 0.7991635483160907, + "grad_norm": 1.4170667936408392, + "learning_rate": 1.0209782045262584e-06, + "loss": 0.2998, + "step": 36306 + }, + { + "epoch": 0.7991855602025093, + "grad_norm": 0.6311899939806068, + "learning_rate": 1.020762361117879e-06, + "loss": 0.4794, + "step": 36307 + }, + { + "epoch": 0.799207572088928, + "grad_norm": 2.2761942952991685, + "learning_rate": 1.0205465379337604e-06, + "loss": 0.3126, + "step": 36308 + }, + { + "epoch": 0.7992295839753467, + "grad_norm": 1.6894833122772381, + "learning_rate": 1.0203307349749975e-06, + "loss": 0.2391, + "step": 36309 + }, + { + "epoch": 0.7992515958617653, + "grad_norm": 4.037113853735033, + "learning_rate": 1.020114952242689e-06, + "loss": 0.2429, + "step": 36310 + }, + { + "epoch": 0.799273607748184, + "grad_norm": 0.6154273459346125, + "learning_rate": 1.0198991897379312e-06, + "loss": 0.4429, + "step": 36311 + }, + { + "epoch": 0.7992956196346027, + "grad_norm": 2.2108008015382548, + "learning_rate": 1.0196834474618228e-06, + "loss": 0.2984, + "step": 36312 + }, + { + "epoch": 0.7993176315210213, + "grad_norm": 1.8902353620129935, + "learning_rate": 1.0194677254154579e-06, + "loss": 0.3384, + "step": 36313 + }, + { + "epoch": 0.79933964340744, + "grad_norm": 4.197676258215536, + "learning_rate": 1.0192520235999321e-06, + "loss": 0.3293, + "step": 36314 + }, + { + "epoch": 0.7993616552938587, + "grad_norm": 0.5990597882925376, + "learning_rate": 1.0190363420163429e-06, + "loss": 0.4632, + "step": 36315 + }, + { + "epoch": 0.7993836671802773, + "grad_norm": 3.3279687280600743, + "learning_rate": 1.0188206806657869e-06, + "loss": 0.3481, + "step": 36316 + }, + { + "epoch": 0.799405679066696, + "grad_norm": 1.8094533459912543, + "learning_rate": 1.0186050395493608e-06, + "loss": 0.3264, + "step": 36317 + }, + { + "epoch": 0.7994276909531147, + "grad_norm": 1.4781208419023721, + "learning_rate": 1.0183894186681597e-06, + "loss": 0.3463, + "step": 36318 + }, + { + "epoch": 0.7994497028395333, + "grad_norm": 0.6208110740349574, + "learning_rate": 1.018173818023278e-06, + "loss": 0.4653, + "step": 36319 + }, + { + "epoch": 0.799471714725952, + "grad_norm": 3.3119455645718627, + "learning_rate": 1.0179582376158148e-06, + "loss": 0.3005, + "step": 36320 + }, + { + "epoch": 0.7994937266123707, + "grad_norm": 0.6110877082518038, + "learning_rate": 1.0177426774468606e-06, + "loss": 0.4566, + "step": 36321 + }, + { + "epoch": 0.7995157384987893, + "grad_norm": 1.7331170451639193, + "learning_rate": 1.017527137517517e-06, + "loss": 0.3039, + "step": 36322 + }, + { + "epoch": 0.799537750385208, + "grad_norm": 0.6202336764271856, + "learning_rate": 1.0173116178288773e-06, + "loss": 0.4653, + "step": 36323 + }, + { + "epoch": 0.7995597622716267, + "grad_norm": 1.6664815794852852, + "learning_rate": 1.0170961183820344e-06, + "loss": 0.27, + "step": 36324 + }, + { + "epoch": 0.7995817741580453, + "grad_norm": 1.719150769978057, + "learning_rate": 1.0168806391780867e-06, + "loss": 0.3141, + "step": 36325 + }, + { + "epoch": 0.799603786044464, + "grad_norm": 2.5161513175128567, + "learning_rate": 1.0166651802181271e-06, + "loss": 0.4377, + "step": 36326 + }, + { + "epoch": 0.7996257979308827, + "grad_norm": 0.6115787775033772, + "learning_rate": 1.0164497415032515e-06, + "loss": 0.4585, + "step": 36327 + }, + { + "epoch": 0.7996478098173013, + "grad_norm": 1.7844065908768028, + "learning_rate": 1.0162343230345562e-06, + "loss": 0.276, + "step": 36328 + }, + { + "epoch": 0.79966982170372, + "grad_norm": 2.9813541153323504, + "learning_rate": 1.0160189248131336e-06, + "loss": 0.3905, + "step": 36329 + }, + { + "epoch": 0.7996918335901386, + "grad_norm": 2.781725716581697, + "learning_rate": 1.0158035468400812e-06, + "loss": 0.2984, + "step": 36330 + }, + { + "epoch": 0.7997138454765573, + "grad_norm": 2.390379980856194, + "learning_rate": 1.0155881891164909e-06, + "loss": 0.2533, + "step": 36331 + }, + { + "epoch": 0.799735857362976, + "grad_norm": 1.6063247299040568, + "learning_rate": 1.0153728516434585e-06, + "loss": 0.2425, + "step": 36332 + }, + { + "epoch": 0.7997578692493946, + "grad_norm": 1.459540290159985, + "learning_rate": 1.0151575344220798e-06, + "loss": 0.2262, + "step": 36333 + }, + { + "epoch": 0.7997798811358133, + "grad_norm": 4.142578577485258, + "learning_rate": 1.0149422374534468e-06, + "loss": 0.3387, + "step": 36334 + }, + { + "epoch": 0.799801893022232, + "grad_norm": 1.8836298304250834, + "learning_rate": 1.0147269607386557e-06, + "loss": 0.3207, + "step": 36335 + }, + { + "epoch": 0.7998239049086506, + "grad_norm": 10.225556114611798, + "learning_rate": 1.0145117042787982e-06, + "loss": 0.3991, + "step": 36336 + }, + { + "epoch": 0.7998459167950693, + "grad_norm": 1.997508445105181, + "learning_rate": 1.0142964680749701e-06, + "loss": 0.2362, + "step": 36337 + }, + { + "epoch": 0.799867928681488, + "grad_norm": 3.863387552481135, + "learning_rate": 1.0140812521282661e-06, + "loss": 0.3827, + "step": 36338 + }, + { + "epoch": 0.7998899405679066, + "grad_norm": 2.530140259000754, + "learning_rate": 1.0138660564397772e-06, + "loss": 0.2967, + "step": 36339 + }, + { + "epoch": 0.7999119524543253, + "grad_norm": 1.5502133744540212, + "learning_rate": 1.013650881010601e-06, + "loss": 0.2803, + "step": 36340 + }, + { + "epoch": 0.799933964340744, + "grad_norm": 0.6149015568585451, + "learning_rate": 1.0134357258418275e-06, + "loss": 0.443, + "step": 36341 + }, + { + "epoch": 0.7999559762271626, + "grad_norm": 5.311484780898034, + "learning_rate": 1.0132205909345516e-06, + "loss": 0.3306, + "step": 36342 + }, + { + "epoch": 0.7999779881135813, + "grad_norm": 1.6049986020246003, + "learning_rate": 1.0130054762898678e-06, + "loss": 0.2143, + "step": 36343 + }, + { + "epoch": 0.8, + "grad_norm": 2.1615335555920954, + "learning_rate": 1.0127903819088691e-06, + "loss": 0.3862, + "step": 36344 + }, + { + "epoch": 0.8000220118864186, + "grad_norm": 0.6365177349474438, + "learning_rate": 1.012575307792646e-06, + "loss": 0.476, + "step": 36345 + }, + { + "epoch": 0.8000440237728373, + "grad_norm": 1.820039145686414, + "learning_rate": 1.0123602539422938e-06, + "loss": 0.2714, + "step": 36346 + }, + { + "epoch": 0.800066035659256, + "grad_norm": 1.59540982139624, + "learning_rate": 1.0121452203589056e-06, + "loss": 0.3471, + "step": 36347 + }, + { + "epoch": 0.8000880475456746, + "grad_norm": 2.351874979126651, + "learning_rate": 1.0119302070435754e-06, + "loss": 0.2914, + "step": 36348 + }, + { + "epoch": 0.8001100594320933, + "grad_norm": 1.6441853240877793, + "learning_rate": 1.011715213997394e-06, + "loss": 0.3745, + "step": 36349 + }, + { + "epoch": 0.800132071318512, + "grad_norm": 1.7048405075227886, + "learning_rate": 1.0115002412214541e-06, + "loss": 0.404, + "step": 36350 + }, + { + "epoch": 0.8001540832049306, + "grad_norm": 1.5675536479514252, + "learning_rate": 1.0112852887168484e-06, + "loss": 0.2663, + "step": 36351 + }, + { + "epoch": 0.8001760950913493, + "grad_norm": 1.564793989908383, + "learning_rate": 1.0110703564846703e-06, + "loss": 0.3829, + "step": 36352 + }, + { + "epoch": 0.8001981069777679, + "grad_norm": 0.6218695905013845, + "learning_rate": 1.0108554445260126e-06, + "loss": 0.4361, + "step": 36353 + }, + { + "epoch": 0.8002201188641866, + "grad_norm": 0.6349770981116512, + "learning_rate": 1.010640552841967e-06, + "loss": 0.4767, + "step": 36354 + }, + { + "epoch": 0.8002421307506054, + "grad_norm": 1.7795394000742417, + "learning_rate": 1.0104256814336243e-06, + "loss": 0.2532, + "step": 36355 + }, + { + "epoch": 0.800264142637024, + "grad_norm": 2.3529235124204435, + "learning_rate": 1.010210830302079e-06, + "loss": 0.3072, + "step": 36356 + }, + { + "epoch": 0.8002861545234426, + "grad_norm": 1.5347893308527403, + "learning_rate": 1.0099959994484187e-06, + "loss": 0.2822, + "step": 36357 + }, + { + "epoch": 0.8003081664098614, + "grad_norm": 1.6399683787609936, + "learning_rate": 1.0097811888737414e-06, + "loss": 0.2777, + "step": 36358 + }, + { + "epoch": 0.80033017829628, + "grad_norm": 0.6249719340171553, + "learning_rate": 1.0095663985791354e-06, + "loss": 0.4519, + "step": 36359 + }, + { + "epoch": 0.8003521901826987, + "grad_norm": 2.38180766392675, + "learning_rate": 1.0093516285656917e-06, + "loss": 0.2401, + "step": 36360 + }, + { + "epoch": 0.8003742020691174, + "grad_norm": 1.3886918832406014, + "learning_rate": 1.009136878834504e-06, + "loss": 0.2606, + "step": 36361 + }, + { + "epoch": 0.800396213955536, + "grad_norm": 1.893300562451031, + "learning_rate": 1.0089221493866614e-06, + "loss": 0.3496, + "step": 36362 + }, + { + "epoch": 0.8004182258419547, + "grad_norm": 2.4102331767905407, + "learning_rate": 1.0087074402232566e-06, + "loss": 0.2922, + "step": 36363 + }, + { + "epoch": 0.8004402377283734, + "grad_norm": 1.479830759378328, + "learning_rate": 1.0084927513453817e-06, + "loss": 0.1899, + "step": 36364 + }, + { + "epoch": 0.800462249614792, + "grad_norm": 0.6121789564852104, + "learning_rate": 1.008278082754126e-06, + "loss": 0.4534, + "step": 36365 + }, + { + "epoch": 0.8004842615012107, + "grad_norm": 1.6219963629891248, + "learning_rate": 1.008063434450583e-06, + "loss": 0.2554, + "step": 36366 + }, + { + "epoch": 0.8005062733876294, + "grad_norm": 1.4916690391721898, + "learning_rate": 1.0078488064358405e-06, + "loss": 0.3573, + "step": 36367 + }, + { + "epoch": 0.800528285274048, + "grad_norm": 2.1924936988270494, + "learning_rate": 1.0076341987109906e-06, + "loss": 0.2929, + "step": 36368 + }, + { + "epoch": 0.8005502971604667, + "grad_norm": 1.6126935033451986, + "learning_rate": 1.0074196112771263e-06, + "loss": 0.2592, + "step": 36369 + }, + { + "epoch": 0.8005723090468854, + "grad_norm": 1.5052334206946152, + "learning_rate": 1.0072050441353342e-06, + "loss": 0.2592, + "step": 36370 + }, + { + "epoch": 0.800594320933304, + "grad_norm": 3.818873913069042, + "learning_rate": 1.006990497286709e-06, + "loss": 0.2777, + "step": 36371 + }, + { + "epoch": 0.8006163328197227, + "grad_norm": 2.648280386274242, + "learning_rate": 1.0067759707323372e-06, + "loss": 0.2685, + "step": 36372 + }, + { + "epoch": 0.8006383447061414, + "grad_norm": 1.6493566316041681, + "learning_rate": 1.0065614644733108e-06, + "loss": 0.2717, + "step": 36373 + }, + { + "epoch": 0.80066035659256, + "grad_norm": 1.7131178642115705, + "learning_rate": 1.0063469785107216e-06, + "loss": 0.2686, + "step": 36374 + }, + { + "epoch": 0.8006823684789787, + "grad_norm": 1.354109617840798, + "learning_rate": 1.0061325128456573e-06, + "loss": 0.3505, + "step": 36375 + }, + { + "epoch": 0.8007043803653973, + "grad_norm": 1.3000752991733453, + "learning_rate": 1.0059180674792101e-06, + "loss": 0.247, + "step": 36376 + }, + { + "epoch": 0.800726392251816, + "grad_norm": 1.3767337793497072, + "learning_rate": 1.0057036424124672e-06, + "loss": 0.3148, + "step": 36377 + }, + { + "epoch": 0.8007484041382347, + "grad_norm": 2.511668831952474, + "learning_rate": 1.0054892376465197e-06, + "loss": 0.2637, + "step": 36378 + }, + { + "epoch": 0.8007704160246533, + "grad_norm": 1.3155700979189946, + "learning_rate": 1.005274853182459e-06, + "loss": 0.3006, + "step": 36379 + }, + { + "epoch": 0.800792427911072, + "grad_norm": 1.4874583831068373, + "learning_rate": 1.0050604890213722e-06, + "loss": 0.2982, + "step": 36380 + }, + { + "epoch": 0.8008144397974907, + "grad_norm": 0.6011247175656628, + "learning_rate": 1.004846145164351e-06, + "loss": 0.4423, + "step": 36381 + }, + { + "epoch": 0.8008364516839093, + "grad_norm": 1.845759414678812, + "learning_rate": 1.0046318216124817e-06, + "loss": 0.2505, + "step": 36382 + }, + { + "epoch": 0.800858463570328, + "grad_norm": 1.6469046758877774, + "learning_rate": 1.0044175183668558e-06, + "loss": 0.3278, + "step": 36383 + }, + { + "epoch": 0.8008804754567467, + "grad_norm": 1.4351132529373773, + "learning_rate": 1.0042032354285642e-06, + "loss": 0.3261, + "step": 36384 + }, + { + "epoch": 0.8009024873431653, + "grad_norm": 0.5963819943214486, + "learning_rate": 1.0039889727986928e-06, + "loss": 0.4461, + "step": 36385 + }, + { + "epoch": 0.800924499229584, + "grad_norm": 1.8309082118365971, + "learning_rate": 1.0037747304783308e-06, + "loss": 0.2659, + "step": 36386 + }, + { + "epoch": 0.8009465111160027, + "grad_norm": 1.6324177452117332, + "learning_rate": 1.0035605084685684e-06, + "loss": 0.2749, + "step": 36387 + }, + { + "epoch": 0.8009685230024213, + "grad_norm": 1.3265355979129845, + "learning_rate": 1.0033463067704934e-06, + "loss": 0.2891, + "step": 36388 + }, + { + "epoch": 0.80099053488884, + "grad_norm": 1.642271935565867, + "learning_rate": 1.0031321253851966e-06, + "loss": 0.2401, + "step": 36389 + }, + { + "epoch": 0.8010125467752587, + "grad_norm": 1.7768953070167257, + "learning_rate": 1.0029179643137648e-06, + "loss": 0.2954, + "step": 36390 + }, + { + "epoch": 0.8010345586616773, + "grad_norm": 0.6144697027908251, + "learning_rate": 1.0027038235572856e-06, + "loss": 0.4491, + "step": 36391 + }, + { + "epoch": 0.801056570548096, + "grad_norm": 1.5291004504690047, + "learning_rate": 1.0024897031168496e-06, + "loss": 0.2162, + "step": 36392 + }, + { + "epoch": 0.8010785824345147, + "grad_norm": 1.6130869304780426, + "learning_rate": 1.0022756029935416e-06, + "loss": 0.3521, + "step": 36393 + }, + { + "epoch": 0.8011005943209333, + "grad_norm": 3.726619294854069, + "learning_rate": 1.002061523188454e-06, + "loss": 0.2436, + "step": 36394 + }, + { + "epoch": 0.801122606207352, + "grad_norm": 1.31504227366946, + "learning_rate": 1.0018474637026731e-06, + "loss": 0.2754, + "step": 36395 + }, + { + "epoch": 0.8011446180937707, + "grad_norm": 0.631118398487372, + "learning_rate": 1.0016334245372855e-06, + "loss": 0.4365, + "step": 36396 + }, + { + "epoch": 0.8011666299801893, + "grad_norm": 2.1361991243642984, + "learning_rate": 1.0014194056933812e-06, + "loss": 0.3748, + "step": 36397 + }, + { + "epoch": 0.801188641866608, + "grad_norm": 1.4292705070586706, + "learning_rate": 1.0012054071720456e-06, + "loss": 0.3144, + "step": 36398 + }, + { + "epoch": 0.8012106537530266, + "grad_norm": 3.0997805859330154, + "learning_rate": 1.0009914289743677e-06, + "loss": 0.2365, + "step": 36399 + }, + { + "epoch": 0.8012326656394453, + "grad_norm": 1.6477756294944619, + "learning_rate": 1.0007774711014368e-06, + "loss": 0.384, + "step": 36400 + }, + { + "epoch": 0.801254677525864, + "grad_norm": 1.6893285936175197, + "learning_rate": 1.0005635335543369e-06, + "loss": 0.3231, + "step": 36401 + }, + { + "epoch": 0.8012766894122826, + "grad_norm": 2.1207203634352214, + "learning_rate": 1.0003496163341574e-06, + "loss": 0.2851, + "step": 36402 + }, + { + "epoch": 0.8012987012987013, + "grad_norm": 2.974906364155327, + "learning_rate": 1.0001357194419848e-06, + "loss": 0.3565, + "step": 36403 + }, + { + "epoch": 0.80132071318512, + "grad_norm": 1.4739014296810697, + "learning_rate": 9.99921842878906e-07, + "loss": 0.2574, + "step": 36404 + }, + { + "epoch": 0.8013427250715386, + "grad_norm": 1.7205255910048165, + "learning_rate": 9.997079866460097e-07, + "loss": 0.2668, + "step": 36405 + }, + { + "epoch": 0.8013647369579573, + "grad_norm": 2.0346006808605437, + "learning_rate": 9.994941507443807e-07, + "loss": 0.3318, + "step": 36406 + }, + { + "epoch": 0.801386748844376, + "grad_norm": 1.8946469987644934, + "learning_rate": 9.992803351751074e-07, + "loss": 0.2718, + "step": 36407 + }, + { + "epoch": 0.8014087607307946, + "grad_norm": 1.453677124832125, + "learning_rate": 9.990665399392752e-07, + "loss": 0.3697, + "step": 36408 + }, + { + "epoch": 0.8014307726172133, + "grad_norm": 1.6238701828415671, + "learning_rate": 9.98852765037971e-07, + "loss": 0.251, + "step": 36409 + }, + { + "epoch": 0.801452784503632, + "grad_norm": 1.5521207572557314, + "learning_rate": 9.986390104722826e-07, + "loss": 0.2533, + "step": 36410 + }, + { + "epoch": 0.8014747963900506, + "grad_norm": 2.3354788193289036, + "learning_rate": 9.984252762432944e-07, + "loss": 0.2794, + "step": 36411 + }, + { + "epoch": 0.8014968082764693, + "grad_norm": 1.3388119741547126, + "learning_rate": 9.98211562352095e-07, + "loss": 0.1907, + "step": 36412 + }, + { + "epoch": 0.801518820162888, + "grad_norm": 1.5597574719054752, + "learning_rate": 9.979978687997678e-07, + "loss": 0.3403, + "step": 36413 + }, + { + "epoch": 0.8015408320493066, + "grad_norm": 2.1950592618345017, + "learning_rate": 9.977841955874008e-07, + "loss": 0.2683, + "step": 36414 + }, + { + "epoch": 0.8015628439357253, + "grad_norm": 1.6784684480106378, + "learning_rate": 9.975705427160802e-07, + "loss": 0.2767, + "step": 36415 + }, + { + "epoch": 0.801584855822144, + "grad_norm": 1.4885763066578301, + "learning_rate": 9.973569101868901e-07, + "loss": 0.2429, + "step": 36416 + }, + { + "epoch": 0.8016068677085626, + "grad_norm": 1.7502887755650358, + "learning_rate": 9.971432980009193e-07, + "loss": 0.278, + "step": 36417 + }, + { + "epoch": 0.8016288795949813, + "grad_norm": 1.753398491846309, + "learning_rate": 9.969297061592497e-07, + "loss": 0.2028, + "step": 36418 + }, + { + "epoch": 0.8016508914814, + "grad_norm": 1.761476103793016, + "learning_rate": 9.967161346629689e-07, + "loss": 0.3364, + "step": 36419 + }, + { + "epoch": 0.8016729033678186, + "grad_norm": 1.9412791358855894, + "learning_rate": 9.965025835131637e-07, + "loss": 0.3338, + "step": 36420 + }, + { + "epoch": 0.8016949152542373, + "grad_norm": 7.552446116649118, + "learning_rate": 9.962890527109164e-07, + "loss": 0.2796, + "step": 36421 + }, + { + "epoch": 0.8017169271406559, + "grad_norm": 0.6450433336562985, + "learning_rate": 9.960755422573154e-07, + "loss": 0.454, + "step": 36422 + }, + { + "epoch": 0.8017389390270746, + "grad_norm": 2.188281810849042, + "learning_rate": 9.958620521534428e-07, + "loss": 0.3028, + "step": 36423 + }, + { + "epoch": 0.8017609509134933, + "grad_norm": 2.2391545181674557, + "learning_rate": 9.956485824003854e-07, + "loss": 0.3391, + "step": 36424 + }, + { + "epoch": 0.8017829627999119, + "grad_norm": 1.777530216058112, + "learning_rate": 9.95435132999229e-07, + "loss": 0.3206, + "step": 36425 + }, + { + "epoch": 0.8018049746863306, + "grad_norm": 2.510135832867506, + "learning_rate": 9.952217039510576e-07, + "loss": 0.2432, + "step": 36426 + }, + { + "epoch": 0.8018269865727493, + "grad_norm": 3.7144996863233226, + "learning_rate": 9.95008295256954e-07, + "loss": 0.2989, + "step": 36427 + }, + { + "epoch": 0.8018489984591679, + "grad_norm": 1.5790808932828053, + "learning_rate": 9.947949069180058e-07, + "loss": 0.2816, + "step": 36428 + }, + { + "epoch": 0.8018710103455866, + "grad_norm": 0.6366754033754317, + "learning_rate": 9.945815389352941e-07, + "loss": 0.44, + "step": 36429 + }, + { + "epoch": 0.8018930222320053, + "grad_norm": 2.466622293224853, + "learning_rate": 9.943681913099078e-07, + "loss": 0.3232, + "step": 36430 + }, + { + "epoch": 0.8019150341184239, + "grad_norm": 1.614926709721497, + "learning_rate": 9.941548640429288e-07, + "loss": 0.372, + "step": 36431 + }, + { + "epoch": 0.8019370460048426, + "grad_norm": 1.898602504983002, + "learning_rate": 9.939415571354406e-07, + "loss": 0.2223, + "step": 36432 + }, + { + "epoch": 0.8019590578912613, + "grad_norm": 1.4703612854789798, + "learning_rate": 9.937282705885286e-07, + "loss": 0.3073, + "step": 36433 + }, + { + "epoch": 0.8019810697776799, + "grad_norm": 1.991732558734322, + "learning_rate": 9.935150044032759e-07, + "loss": 0.2853, + "step": 36434 + }, + { + "epoch": 0.8020030816640986, + "grad_norm": 1.5949264913379753, + "learning_rate": 9.933017585807663e-07, + "loss": 0.2995, + "step": 36435 + }, + { + "epoch": 0.8020250935505173, + "grad_norm": 1.4620083614417432, + "learning_rate": 9.930885331220857e-07, + "loss": 0.2869, + "step": 36436 + }, + { + "epoch": 0.8020471054369359, + "grad_norm": 0.6058918966241825, + "learning_rate": 9.928753280283154e-07, + "loss": 0.4537, + "step": 36437 + }, + { + "epoch": 0.8020691173233546, + "grad_norm": 2.1541497117807467, + "learning_rate": 9.926621433005407e-07, + "loss": 0.4043, + "step": 36438 + }, + { + "epoch": 0.8020911292097733, + "grad_norm": 1.4311913395649816, + "learning_rate": 9.924489789398434e-07, + "loss": 0.234, + "step": 36439 + }, + { + "epoch": 0.8021131410961919, + "grad_norm": 1.980785192928831, + "learning_rate": 9.922358349473076e-07, + "loss": 0.2791, + "step": 36440 + }, + { + "epoch": 0.8021351529826106, + "grad_norm": 2.727207742087056, + "learning_rate": 9.920227113240184e-07, + "loss": 0.2762, + "step": 36441 + }, + { + "epoch": 0.8021571648690293, + "grad_norm": 1.6541603512835827, + "learning_rate": 9.918096080710554e-07, + "loss": 0.3076, + "step": 36442 + }, + { + "epoch": 0.8021791767554479, + "grad_norm": 2.7177592505485135, + "learning_rate": 9.915965251895054e-07, + "loss": 0.38, + "step": 36443 + }, + { + "epoch": 0.8022011886418666, + "grad_norm": 13.479975761449495, + "learning_rate": 9.913834626804486e-07, + "loss": 0.2685, + "step": 36444 + }, + { + "epoch": 0.8022232005282853, + "grad_norm": 1.595392291378351, + "learning_rate": 9.911704205449685e-07, + "loss": 0.2509, + "step": 36445 + }, + { + "epoch": 0.8022452124147039, + "grad_norm": 7.651731519635836, + "learning_rate": 9.909573987841498e-07, + "loss": 0.3328, + "step": 36446 + }, + { + "epoch": 0.8022672243011226, + "grad_norm": 1.7161863276134015, + "learning_rate": 9.90744397399072e-07, + "loss": 0.4239, + "step": 36447 + }, + { + "epoch": 0.8022892361875412, + "grad_norm": 2.214316832211175, + "learning_rate": 9.905314163908213e-07, + "loss": 0.2594, + "step": 36448 + }, + { + "epoch": 0.8023112480739599, + "grad_norm": 2.523584896799119, + "learning_rate": 9.903184557604768e-07, + "loss": 0.3139, + "step": 36449 + }, + { + "epoch": 0.8023332599603786, + "grad_norm": 1.761493285529658, + "learning_rate": 9.901055155091221e-07, + "loss": 0.2622, + "step": 36450 + }, + { + "epoch": 0.8023552718467972, + "grad_norm": 1.8615240147020078, + "learning_rate": 9.898925956378408e-07, + "loss": 0.3698, + "step": 36451 + }, + { + "epoch": 0.8023772837332159, + "grad_norm": 2.877305882843077, + "learning_rate": 9.896796961477129e-07, + "loss": 0.2342, + "step": 36452 + }, + { + "epoch": 0.8023992956196346, + "grad_norm": 1.8178577495042096, + "learning_rate": 9.894668170398224e-07, + "loss": 0.3324, + "step": 36453 + }, + { + "epoch": 0.8024213075060532, + "grad_norm": 1.7590957370164193, + "learning_rate": 9.892539583152494e-07, + "loss": 0.2441, + "step": 36454 + }, + { + "epoch": 0.8024433193924719, + "grad_norm": 1.8418458185355993, + "learning_rate": 9.890411199750761e-07, + "loss": 0.3011, + "step": 36455 + }, + { + "epoch": 0.8024653312788906, + "grad_norm": 1.2993744095169362, + "learning_rate": 9.888283020203865e-07, + "loss": 0.2189, + "step": 36456 + }, + { + "epoch": 0.8024873431653092, + "grad_norm": 2.3125174224507736, + "learning_rate": 9.88615504452259e-07, + "loss": 0.2917, + "step": 36457 + }, + { + "epoch": 0.8025093550517279, + "grad_norm": 1.792650169958599, + "learning_rate": 9.884027272717777e-07, + "loss": 0.2462, + "step": 36458 + }, + { + "epoch": 0.8025313669381466, + "grad_norm": 2.012007288609838, + "learning_rate": 9.881899704800235e-07, + "loss": 0.291, + "step": 36459 + }, + { + "epoch": 0.8025533788245652, + "grad_norm": 0.6048493665168246, + "learning_rate": 9.87977234078074e-07, + "loss": 0.4457, + "step": 36460 + }, + { + "epoch": 0.8025753907109839, + "grad_norm": 1.3815650583901815, + "learning_rate": 9.87764518067017e-07, + "loss": 0.2918, + "step": 36461 + }, + { + "epoch": 0.8025974025974026, + "grad_norm": 2.0981448729624974, + "learning_rate": 9.875518224479286e-07, + "loss": 0.2665, + "step": 36462 + }, + { + "epoch": 0.8026194144838212, + "grad_norm": 1.5015323888148129, + "learning_rate": 9.873391472218924e-07, + "loss": 0.2692, + "step": 36463 + }, + { + "epoch": 0.8026414263702399, + "grad_norm": 0.6104760287219755, + "learning_rate": 9.871264923899888e-07, + "loss": 0.4417, + "step": 36464 + }, + { + "epoch": 0.8026634382566586, + "grad_norm": 1.3226664094453369, + "learning_rate": 9.86913857953295e-07, + "loss": 0.2135, + "step": 36465 + }, + { + "epoch": 0.8026854501430772, + "grad_norm": 2.1330251983121267, + "learning_rate": 9.867012439128976e-07, + "loss": 0.2866, + "step": 36466 + }, + { + "epoch": 0.8027074620294959, + "grad_norm": 1.5384692606681467, + "learning_rate": 9.86488650269875e-07, + "loss": 0.363, + "step": 36467 + }, + { + "epoch": 0.8027294739159146, + "grad_norm": 1.9808021556723343, + "learning_rate": 9.862760770253055e-07, + "loss": 0.3097, + "step": 36468 + }, + { + "epoch": 0.8027514858023332, + "grad_norm": 1.5769423877178859, + "learning_rate": 9.860635241802723e-07, + "loss": 0.2517, + "step": 36469 + }, + { + "epoch": 0.802773497688752, + "grad_norm": 1.6395094591360457, + "learning_rate": 9.858509917358533e-07, + "loss": 0.2704, + "step": 36470 + }, + { + "epoch": 0.8027955095751705, + "grad_norm": 3.283730406260808, + "learning_rate": 9.8563847969313e-07, + "loss": 0.2262, + "step": 36471 + }, + { + "epoch": 0.8028175214615892, + "grad_norm": 2.0503705212903083, + "learning_rate": 9.854259880531837e-07, + "loss": 0.2925, + "step": 36472 + }, + { + "epoch": 0.802839533348008, + "grad_norm": 1.3599473055692335, + "learning_rate": 9.852135168170918e-07, + "loss": 0.2918, + "step": 36473 + }, + { + "epoch": 0.8028615452344265, + "grad_norm": 1.5354540427795584, + "learning_rate": 9.850010659859367e-07, + "loss": 0.2948, + "step": 36474 + }, + { + "epoch": 0.8028835571208452, + "grad_norm": 1.881496367686981, + "learning_rate": 9.847886355607956e-07, + "loss": 0.3792, + "step": 36475 + }, + { + "epoch": 0.802905569007264, + "grad_norm": 1.5841393751237405, + "learning_rate": 9.845762255427494e-07, + "loss": 0.2796, + "step": 36476 + }, + { + "epoch": 0.8029275808936825, + "grad_norm": 0.6297235968259554, + "learning_rate": 9.843638359328794e-07, + "loss": 0.4552, + "step": 36477 + }, + { + "epoch": 0.8029495927801013, + "grad_norm": 1.5783944446909197, + "learning_rate": 9.84151466732262e-07, + "loss": 0.3096, + "step": 36478 + }, + { + "epoch": 0.80297160466652, + "grad_norm": 1.859305150865521, + "learning_rate": 9.83939117941979e-07, + "loss": 0.2807, + "step": 36479 + }, + { + "epoch": 0.8029936165529385, + "grad_norm": 1.6744603327634309, + "learning_rate": 9.837267895631076e-07, + "loss": 0.2393, + "step": 36480 + }, + { + "epoch": 0.8030156284393573, + "grad_norm": 1.9093880834059689, + "learning_rate": 9.835144815967278e-07, + "loss": 0.3147, + "step": 36481 + }, + { + "epoch": 0.803037640325776, + "grad_norm": 1.3150013907080447, + "learning_rate": 9.833021940439202e-07, + "loss": 0.2408, + "step": 36482 + }, + { + "epoch": 0.8030596522121946, + "grad_norm": 1.8647926515863997, + "learning_rate": 9.83089926905761e-07, + "loss": 0.3117, + "step": 36483 + }, + { + "epoch": 0.8030816640986133, + "grad_norm": 6.942308401421043, + "learning_rate": 9.828776801833317e-07, + "loss": 0.269, + "step": 36484 + }, + { + "epoch": 0.803103675985032, + "grad_norm": 1.4759939552280548, + "learning_rate": 9.826654538777081e-07, + "loss": 0.3018, + "step": 36485 + }, + { + "epoch": 0.8031256878714506, + "grad_norm": 1.6173486677829956, + "learning_rate": 9.824532479899706e-07, + "loss": 0.3029, + "step": 36486 + }, + { + "epoch": 0.8031476997578693, + "grad_norm": 1.7493336660388281, + "learning_rate": 9.822410625211993e-07, + "loss": 0.3539, + "step": 36487 + }, + { + "epoch": 0.803169711644288, + "grad_norm": 1.2446302014935349, + "learning_rate": 9.820288974724695e-07, + "loss": 0.2784, + "step": 36488 + }, + { + "epoch": 0.8031917235307066, + "grad_norm": 1.5827067984369672, + "learning_rate": 9.818167528448619e-07, + "loss": 0.2736, + "step": 36489 + }, + { + "epoch": 0.8032137354171253, + "grad_norm": 1.5172345763311743, + "learning_rate": 9.816046286394526e-07, + "loss": 0.3129, + "step": 36490 + }, + { + "epoch": 0.803235747303544, + "grad_norm": 1.4472876042707312, + "learning_rate": 9.813925248573208e-07, + "loss": 0.2107, + "step": 36491 + }, + { + "epoch": 0.8032577591899626, + "grad_norm": 1.4957287416868674, + "learning_rate": 9.811804414995456e-07, + "loss": 0.3122, + "step": 36492 + }, + { + "epoch": 0.8032797710763813, + "grad_norm": 0.6061131381773951, + "learning_rate": 9.809683785672026e-07, + "loss": 0.4169, + "step": 36493 + }, + { + "epoch": 0.8033017829627999, + "grad_norm": 0.62477517697255, + "learning_rate": 9.807563360613726e-07, + "loss": 0.4643, + "step": 36494 + }, + { + "epoch": 0.8033237948492186, + "grad_norm": 0.6316045867550238, + "learning_rate": 9.805443139831311e-07, + "loss": 0.4735, + "step": 36495 + }, + { + "epoch": 0.8033458067356373, + "grad_norm": 0.6102184848344373, + "learning_rate": 9.803323123335534e-07, + "loss": 0.4581, + "step": 36496 + }, + { + "epoch": 0.8033678186220559, + "grad_norm": 1.6329378384621769, + "learning_rate": 9.801203311137225e-07, + "loss": 0.2728, + "step": 36497 + }, + { + "epoch": 0.8033898305084746, + "grad_norm": 1.8655901650766824, + "learning_rate": 9.79908370324712e-07, + "loss": 0.2858, + "step": 36498 + }, + { + "epoch": 0.8034118423948933, + "grad_norm": 2.125622397594967, + "learning_rate": 9.796964299676009e-07, + "loss": 0.3071, + "step": 36499 + }, + { + "epoch": 0.8034338542813119, + "grad_norm": 0.6198239581333308, + "learning_rate": 9.794845100434664e-07, + "loss": 0.4464, + "step": 36500 + }, + { + "epoch": 0.8034558661677306, + "grad_norm": 1.7317513230391328, + "learning_rate": 9.792726105533817e-07, + "loss": 0.2538, + "step": 36501 + }, + { + "epoch": 0.8034778780541493, + "grad_norm": 1.3620763653741945, + "learning_rate": 9.790607314984296e-07, + "loss": 0.2786, + "step": 36502 + }, + { + "epoch": 0.8034998899405679, + "grad_norm": 5.162682624687926, + "learning_rate": 9.78848872879684e-07, + "loss": 0.2656, + "step": 36503 + }, + { + "epoch": 0.8035219018269866, + "grad_norm": 2.6814739875619735, + "learning_rate": 9.78637034698221e-07, + "loss": 0.2904, + "step": 36504 + }, + { + "epoch": 0.8035439137134053, + "grad_norm": 1.4709348507758886, + "learning_rate": 9.78425216955119e-07, + "loss": 0.334, + "step": 36505 + }, + { + "epoch": 0.8035659255998239, + "grad_norm": 1.9627429766681086, + "learning_rate": 9.782134196514531e-07, + "loss": 0.2769, + "step": 36506 + }, + { + "epoch": 0.8035879374862426, + "grad_norm": 1.9625331939907584, + "learning_rate": 9.780016427883e-07, + "loss": 0.2428, + "step": 36507 + }, + { + "epoch": 0.8036099493726613, + "grad_norm": 1.6789732592238964, + "learning_rate": 9.777898863667372e-07, + "loss": 0.3102, + "step": 36508 + }, + { + "epoch": 0.8036319612590799, + "grad_norm": 2.677859897026679, + "learning_rate": 9.775781503878396e-07, + "loss": 0.287, + "step": 36509 + }, + { + "epoch": 0.8036539731454986, + "grad_norm": 1.7430254625783885, + "learning_rate": 9.773664348526851e-07, + "loss": 0.3045, + "step": 36510 + }, + { + "epoch": 0.8036759850319173, + "grad_norm": 0.6111285390760124, + "learning_rate": 9.771547397623472e-07, + "loss": 0.4566, + "step": 36511 + }, + { + "epoch": 0.8036979969183359, + "grad_norm": 50.97490556200234, + "learning_rate": 9.76943065117903e-07, + "loss": 0.3331, + "step": 36512 + }, + { + "epoch": 0.8037200088047546, + "grad_norm": 1.798999207655329, + "learning_rate": 9.767314109204295e-07, + "loss": 0.283, + "step": 36513 + }, + { + "epoch": 0.8037420206911733, + "grad_norm": 1.7979360968318907, + "learning_rate": 9.765197771710006e-07, + "loss": 0.2605, + "step": 36514 + }, + { + "epoch": 0.8037640325775919, + "grad_norm": 1.427074595803414, + "learning_rate": 9.763081638706935e-07, + "loss": 0.2408, + "step": 36515 + }, + { + "epoch": 0.8037860444640106, + "grad_norm": 3.2532308597885553, + "learning_rate": 9.760965710205823e-07, + "loss": 0.231, + "step": 36516 + }, + { + "epoch": 0.8038080563504292, + "grad_norm": 1.7823217325178327, + "learning_rate": 9.758849986217421e-07, + "loss": 0.3253, + "step": 36517 + }, + { + "epoch": 0.8038300682368479, + "grad_norm": 19.102751208798953, + "learning_rate": 9.756734466752515e-07, + "loss": 0.3341, + "step": 36518 + }, + { + "epoch": 0.8038520801232666, + "grad_norm": 1.4990797592553566, + "learning_rate": 9.754619151821816e-07, + "loss": 0.2565, + "step": 36519 + }, + { + "epoch": 0.8038740920096852, + "grad_norm": 0.6278885654680958, + "learning_rate": 9.7525040414361e-07, + "loss": 0.4783, + "step": 36520 + }, + { + "epoch": 0.8038961038961039, + "grad_norm": 1.9363838102073514, + "learning_rate": 9.750389135606102e-07, + "loss": 0.2446, + "step": 36521 + }, + { + "epoch": 0.8039181157825226, + "grad_norm": 2.1559431913343285, + "learning_rate": 9.74827443434258e-07, + "loss": 0.348, + "step": 36522 + }, + { + "epoch": 0.8039401276689412, + "grad_norm": 5.3340942079269285, + "learning_rate": 9.746159937656291e-07, + "loss": 0.2582, + "step": 36523 + }, + { + "epoch": 0.8039621395553599, + "grad_norm": 6.177392787256503, + "learning_rate": 9.744045645557959e-07, + "loss": 0.3052, + "step": 36524 + }, + { + "epoch": 0.8039841514417786, + "grad_norm": 0.6420381884357417, + "learning_rate": 9.74193155805835e-07, + "loss": 0.4705, + "step": 36525 + }, + { + "epoch": 0.8040061633281972, + "grad_norm": 4.750568460995776, + "learning_rate": 9.73981767516819e-07, + "loss": 0.3222, + "step": 36526 + }, + { + "epoch": 0.8040281752146159, + "grad_norm": 1.4061433278155808, + "learning_rate": 9.737703996898234e-07, + "loss": 0.2494, + "step": 36527 + }, + { + "epoch": 0.8040501871010346, + "grad_norm": 1.5552105333578687, + "learning_rate": 9.735590523259237e-07, + "loss": 0.2772, + "step": 36528 + }, + { + "epoch": 0.8040721989874532, + "grad_norm": 0.5897419644992005, + "learning_rate": 9.733477254261915e-07, + "loss": 0.4345, + "step": 36529 + }, + { + "epoch": 0.8040942108738719, + "grad_norm": 1.6396782634052918, + "learning_rate": 9.73136418991703e-07, + "loss": 0.3124, + "step": 36530 + }, + { + "epoch": 0.8041162227602906, + "grad_norm": 1.6418853892686955, + "learning_rate": 9.729251330235317e-07, + "loss": 0.299, + "step": 36531 + }, + { + "epoch": 0.8041382346467092, + "grad_norm": 1.8827559678949821, + "learning_rate": 9.727138675227481e-07, + "loss": 0.3027, + "step": 36532 + }, + { + "epoch": 0.8041602465331279, + "grad_norm": 1.577241052469912, + "learning_rate": 9.725026224904316e-07, + "loss": 0.3359, + "step": 36533 + }, + { + "epoch": 0.8041822584195466, + "grad_norm": 2.9898681741149673, + "learning_rate": 9.72291397927652e-07, + "loss": 0.3247, + "step": 36534 + }, + { + "epoch": 0.8042042703059652, + "grad_norm": 1.7272563039061237, + "learning_rate": 9.720801938354852e-07, + "loss": 0.281, + "step": 36535 + }, + { + "epoch": 0.8042262821923839, + "grad_norm": 2.0534767977081247, + "learning_rate": 9.71869010215003e-07, + "loss": 0.3058, + "step": 36536 + }, + { + "epoch": 0.8042482940788026, + "grad_norm": 4.156628824570928, + "learning_rate": 9.716578470672767e-07, + "loss": 0.2797, + "step": 36537 + }, + { + "epoch": 0.8042703059652212, + "grad_norm": 1.7977315059683778, + "learning_rate": 9.714467043933845e-07, + "loss": 0.3602, + "step": 36538 + }, + { + "epoch": 0.8042923178516399, + "grad_norm": 3.5237257271756697, + "learning_rate": 9.712355821943954e-07, + "loss": 0.3261, + "step": 36539 + }, + { + "epoch": 0.8043143297380585, + "grad_norm": 1.9944212562817054, + "learning_rate": 9.710244804713852e-07, + "loss": 0.2313, + "step": 36540 + }, + { + "epoch": 0.8043363416244772, + "grad_norm": 1.5696551786674369, + "learning_rate": 9.708133992254258e-07, + "loss": 0.2976, + "step": 36541 + }, + { + "epoch": 0.8043583535108959, + "grad_norm": 0.6266803825872926, + "learning_rate": 9.706023384575885e-07, + "loss": 0.4461, + "step": 36542 + }, + { + "epoch": 0.8043803653973145, + "grad_norm": 1.5566866003488684, + "learning_rate": 9.703912981689473e-07, + "loss": 0.3237, + "step": 36543 + }, + { + "epoch": 0.8044023772837332, + "grad_norm": 0.6239305478941293, + "learning_rate": 9.701802783605763e-07, + "loss": 0.4561, + "step": 36544 + }, + { + "epoch": 0.8044243891701519, + "grad_norm": 0.6189688549445138, + "learning_rate": 9.699692790335447e-07, + "loss": 0.4404, + "step": 36545 + }, + { + "epoch": 0.8044464010565705, + "grad_norm": 0.6048547196419516, + "learning_rate": 9.697583001889283e-07, + "loss": 0.4546, + "step": 36546 + }, + { + "epoch": 0.8044684129429892, + "grad_norm": 1.7503555207329737, + "learning_rate": 9.695473418277963e-07, + "loss": 0.2834, + "step": 36547 + }, + { + "epoch": 0.8044904248294079, + "grad_norm": 1.8140879150301197, + "learning_rate": 9.693364039512227e-07, + "loss": 0.2936, + "step": 36548 + }, + { + "epoch": 0.8045124367158265, + "grad_norm": 0.6508295626741974, + "learning_rate": 9.6912548656028e-07, + "loss": 0.4663, + "step": 36549 + }, + { + "epoch": 0.8045344486022452, + "grad_norm": 1.9434214086054464, + "learning_rate": 9.68914589656038e-07, + "loss": 0.2902, + "step": 36550 + }, + { + "epoch": 0.8045564604886639, + "grad_norm": 1.4779528671490705, + "learning_rate": 9.687037132395716e-07, + "loss": 0.3156, + "step": 36551 + }, + { + "epoch": 0.8045784723750825, + "grad_norm": 1.7666675286627769, + "learning_rate": 9.684928573119495e-07, + "loss": 0.2934, + "step": 36552 + }, + { + "epoch": 0.8046004842615012, + "grad_norm": 1.4041517070973883, + "learning_rate": 9.68282021874245e-07, + "loss": 0.2553, + "step": 36553 + }, + { + "epoch": 0.8046224961479199, + "grad_norm": 1.407866282237473, + "learning_rate": 9.680712069275306e-07, + "loss": 0.3235, + "step": 36554 + }, + { + "epoch": 0.8046445080343385, + "grad_norm": 1.334908080131439, + "learning_rate": 9.678604124728757e-07, + "loss": 0.227, + "step": 36555 + }, + { + "epoch": 0.8046665199207572, + "grad_norm": 0.6302838475144393, + "learning_rate": 9.676496385113533e-07, + "loss": 0.4357, + "step": 36556 + }, + { + "epoch": 0.8046885318071759, + "grad_norm": 1.3677409411198806, + "learning_rate": 9.67438885044033e-07, + "loss": 0.2247, + "step": 36557 + }, + { + "epoch": 0.8047105436935945, + "grad_norm": 1.834652242308952, + "learning_rate": 9.672281520719867e-07, + "loss": 0.2956, + "step": 36558 + }, + { + "epoch": 0.8047325555800132, + "grad_norm": 1.8251896147899322, + "learning_rate": 9.670174395962867e-07, + "loss": 0.2647, + "step": 36559 + }, + { + "epoch": 0.8047545674664319, + "grad_norm": 1.4120865331231722, + "learning_rate": 9.66806747618002e-07, + "loss": 0.3125, + "step": 36560 + }, + { + "epoch": 0.8047765793528505, + "grad_norm": 1.6037564162210733, + "learning_rate": 9.665960761382054e-07, + "loss": 0.2994, + "step": 36561 + }, + { + "epoch": 0.8047985912392692, + "grad_norm": 0.6378855588706016, + "learning_rate": 9.663854251579646e-07, + "loss": 0.4767, + "step": 36562 + }, + { + "epoch": 0.8048206031256878, + "grad_norm": 2.030768745114486, + "learning_rate": 9.66174794678353e-07, + "loss": 0.3199, + "step": 36563 + }, + { + "epoch": 0.8048426150121065, + "grad_norm": 1.9161530734028926, + "learning_rate": 9.659641847004408e-07, + "loss": 0.3361, + "step": 36564 + }, + { + "epoch": 0.8048646268985252, + "grad_norm": 2.1020332006171247, + "learning_rate": 9.657535952252973e-07, + "loss": 0.3294, + "step": 36565 + }, + { + "epoch": 0.8048866387849438, + "grad_norm": 2.3320242247308136, + "learning_rate": 9.65543026253994e-07, + "loss": 0.316, + "step": 36566 + }, + { + "epoch": 0.8049086506713625, + "grad_norm": 1.7865377144149424, + "learning_rate": 9.653324777876005e-07, + "loss": 0.2437, + "step": 36567 + }, + { + "epoch": 0.8049306625577812, + "grad_norm": 1.3881173071084947, + "learning_rate": 9.651219498271846e-07, + "loss": 0.1905, + "step": 36568 + }, + { + "epoch": 0.8049526744441998, + "grad_norm": 1.5540624012803383, + "learning_rate": 9.64911442373821e-07, + "loss": 0.2869, + "step": 36569 + }, + { + "epoch": 0.8049746863306185, + "grad_norm": 0.6026615100020676, + "learning_rate": 9.647009554285758e-07, + "loss": 0.4639, + "step": 36570 + }, + { + "epoch": 0.8049966982170372, + "grad_norm": 1.6933781305841535, + "learning_rate": 9.644904889925217e-07, + "loss": 0.3087, + "step": 36571 + }, + { + "epoch": 0.8050187101034558, + "grad_norm": 2.1696054193806247, + "learning_rate": 9.642800430667264e-07, + "loss": 0.3333, + "step": 36572 + }, + { + "epoch": 0.8050407219898745, + "grad_norm": 1.567679738757991, + "learning_rate": 9.64069617652259e-07, + "loss": 0.27, + "step": 36573 + }, + { + "epoch": 0.8050627338762932, + "grad_norm": 2.1799808662169333, + "learning_rate": 9.638592127501894e-07, + "loss": 0.3374, + "step": 36574 + }, + { + "epoch": 0.8050847457627118, + "grad_norm": 1.388180878251625, + "learning_rate": 9.63648828361588e-07, + "loss": 0.3156, + "step": 36575 + }, + { + "epoch": 0.8051067576491305, + "grad_norm": 1.8923208436925603, + "learning_rate": 9.634384644875244e-07, + "loss": 0.3171, + "step": 36576 + }, + { + "epoch": 0.8051287695355492, + "grad_norm": 1.7365761254138325, + "learning_rate": 9.632281211290673e-07, + "loss": 0.3186, + "step": 36577 + }, + { + "epoch": 0.8051507814219678, + "grad_norm": 1.7364902442457852, + "learning_rate": 9.63017798287284e-07, + "loss": 0.2222, + "step": 36578 + }, + { + "epoch": 0.8051727933083865, + "grad_norm": 1.7705597346845345, + "learning_rate": 9.628074959632445e-07, + "loss": 0.2744, + "step": 36579 + }, + { + "epoch": 0.8051948051948052, + "grad_norm": 3.2325663442312647, + "learning_rate": 9.625972141580182e-07, + "loss": 0.3324, + "step": 36580 + }, + { + "epoch": 0.8052168170812238, + "grad_norm": 3.4419910344213798, + "learning_rate": 9.623869528726743e-07, + "loss": 0.3342, + "step": 36581 + }, + { + "epoch": 0.8052388289676425, + "grad_norm": 3.459380038182655, + "learning_rate": 9.621767121082814e-07, + "loss": 0.3438, + "step": 36582 + }, + { + "epoch": 0.8052608408540612, + "grad_norm": 1.497863406122154, + "learning_rate": 9.619664918659054e-07, + "loss": 0.2799, + "step": 36583 + }, + { + "epoch": 0.8052828527404798, + "grad_norm": 3.7478189437166565, + "learning_rate": 9.617562921466173e-07, + "loss": 0.2642, + "step": 36584 + }, + { + "epoch": 0.8053048646268985, + "grad_norm": 1.7612060438075132, + "learning_rate": 9.615461129514859e-07, + "loss": 0.3116, + "step": 36585 + }, + { + "epoch": 0.8053268765133171, + "grad_norm": 1.9683290890077132, + "learning_rate": 9.61335954281577e-07, + "loss": 0.2368, + "step": 36586 + }, + { + "epoch": 0.8053488883997358, + "grad_norm": 1.659376316684871, + "learning_rate": 9.611258161379616e-07, + "loss": 0.299, + "step": 36587 + }, + { + "epoch": 0.8053709002861545, + "grad_norm": 1.5853064562169847, + "learning_rate": 9.60915698521705e-07, + "loss": 0.3799, + "step": 36588 + }, + { + "epoch": 0.8053929121725731, + "grad_norm": 1.7012205425952398, + "learning_rate": 9.607056014338762e-07, + "loss": 0.2691, + "step": 36589 + }, + { + "epoch": 0.8054149240589918, + "grad_norm": 19.217605523228155, + "learning_rate": 9.604955248755437e-07, + "loss": 0.4376, + "step": 36590 + }, + { + "epoch": 0.8054369359454105, + "grad_norm": 1.897287128485311, + "learning_rate": 9.602854688477743e-07, + "loss": 0.3091, + "step": 36591 + }, + { + "epoch": 0.8054589478318291, + "grad_norm": 1.9097437877004007, + "learning_rate": 9.600754333516365e-07, + "loss": 0.3237, + "step": 36592 + }, + { + "epoch": 0.8054809597182478, + "grad_norm": 1.5665310781055033, + "learning_rate": 9.598654183881967e-07, + "loss": 0.2878, + "step": 36593 + }, + { + "epoch": 0.8055029716046666, + "grad_norm": 2.2486046915529494, + "learning_rate": 9.59655423958522e-07, + "loss": 0.2333, + "step": 36594 + }, + { + "epoch": 0.8055249834910851, + "grad_norm": 1.973149601278571, + "learning_rate": 9.594454500636818e-07, + "loss": 0.2696, + "step": 36595 + }, + { + "epoch": 0.8055469953775038, + "grad_norm": 2.1814380021719706, + "learning_rate": 9.592354967047412e-07, + "loss": 0.3146, + "step": 36596 + }, + { + "epoch": 0.8055690072639226, + "grad_norm": 2.4357776661253965, + "learning_rate": 9.59025563882769e-07, + "loss": 0.4016, + "step": 36597 + }, + { + "epoch": 0.8055910191503411, + "grad_norm": 1.5111231307367468, + "learning_rate": 9.5881565159883e-07, + "loss": 0.3011, + "step": 36598 + }, + { + "epoch": 0.8056130310367599, + "grad_norm": 1.7813695931063662, + "learning_rate": 9.586057598539922e-07, + "loss": 0.2883, + "step": 36599 + }, + { + "epoch": 0.8056350429231786, + "grad_norm": 1.416162084720872, + "learning_rate": 9.58395888649324e-07, + "loss": 0.2614, + "step": 36600 + }, + { + "epoch": 0.8056570548095972, + "grad_norm": 0.6075748882996724, + "learning_rate": 9.581860379858888e-07, + "loss": 0.4618, + "step": 36601 + }, + { + "epoch": 0.8056790666960159, + "grad_norm": 1.1328908107718587, + "learning_rate": 9.579762078647564e-07, + "loss": 0.3077, + "step": 36602 + }, + { + "epoch": 0.8057010785824346, + "grad_norm": 1.64449485251197, + "learning_rate": 9.577663982869917e-07, + "loss": 0.2977, + "step": 36603 + }, + { + "epoch": 0.8057230904688532, + "grad_norm": 0.6103361926798175, + "learning_rate": 9.575566092536586e-07, + "loss": 0.4645, + "step": 36604 + }, + { + "epoch": 0.8057451023552719, + "grad_norm": 1.607990737460274, + "learning_rate": 9.573468407658281e-07, + "loss": 0.2684, + "step": 36605 + }, + { + "epoch": 0.8057671142416906, + "grad_norm": 4.353862178179919, + "learning_rate": 9.571370928245626e-07, + "loss": 0.2851, + "step": 36606 + }, + { + "epoch": 0.8057891261281092, + "grad_norm": 1.770363123596006, + "learning_rate": 9.569273654309314e-07, + "loss": 0.3184, + "step": 36607 + }, + { + "epoch": 0.8058111380145279, + "grad_norm": 1.5726190598749197, + "learning_rate": 9.567176585859983e-07, + "loss": 0.3478, + "step": 36608 + }, + { + "epoch": 0.8058331499009466, + "grad_norm": 1.4171362599606447, + "learning_rate": 9.565079722908283e-07, + "loss": 0.206, + "step": 36609 + }, + { + "epoch": 0.8058551617873652, + "grad_norm": 0.6246142211235444, + "learning_rate": 9.562983065464875e-07, + "loss": 0.4661, + "step": 36610 + }, + { + "epoch": 0.8058771736737839, + "grad_norm": 0.5884087047998525, + "learning_rate": 9.560886613540431e-07, + "loss": 0.448, + "step": 36611 + }, + { + "epoch": 0.8058991855602025, + "grad_norm": 0.6383417400708262, + "learning_rate": 9.558790367145604e-07, + "loss": 0.4507, + "step": 36612 + }, + { + "epoch": 0.8059211974466212, + "grad_norm": 0.6537912142287329, + "learning_rate": 9.55669432629105e-07, + "loss": 0.4681, + "step": 36613 + }, + { + "epoch": 0.8059432093330399, + "grad_norm": 1.9285038838337252, + "learning_rate": 9.554598490987394e-07, + "loss": 0.331, + "step": 36614 + }, + { + "epoch": 0.8059652212194585, + "grad_norm": 2.6843958729680004, + "learning_rate": 9.55250286124531e-07, + "loss": 0.2959, + "step": 36615 + }, + { + "epoch": 0.8059872331058772, + "grad_norm": 1.9098162992879415, + "learning_rate": 9.550407437075443e-07, + "loss": 0.2343, + "step": 36616 + }, + { + "epoch": 0.8060092449922959, + "grad_norm": 2.263689271464872, + "learning_rate": 9.548312218488465e-07, + "loss": 0.3867, + "step": 36617 + }, + { + "epoch": 0.8060312568787145, + "grad_norm": 2.1422111428782844, + "learning_rate": 9.546217205495e-07, + "loss": 0.3437, + "step": 36618 + }, + { + "epoch": 0.8060532687651332, + "grad_norm": 1.6473767157644734, + "learning_rate": 9.544122398105694e-07, + "loss": 0.2302, + "step": 36619 + }, + { + "epoch": 0.8060752806515519, + "grad_norm": 1.8317487265247134, + "learning_rate": 9.5420277963312e-07, + "loss": 0.397, + "step": 36620 + }, + { + "epoch": 0.8060972925379705, + "grad_norm": 1.7642156185932145, + "learning_rate": 9.539933400182166e-07, + "loss": 0.3317, + "step": 36621 + }, + { + "epoch": 0.8061193044243892, + "grad_norm": 1.4002182579816875, + "learning_rate": 9.537839209669247e-07, + "loss": 0.3555, + "step": 36622 + }, + { + "epoch": 0.8061413163108079, + "grad_norm": 4.378407513046219, + "learning_rate": 9.535745224803078e-07, + "loss": 0.267, + "step": 36623 + }, + { + "epoch": 0.8061633281972265, + "grad_norm": 0.6211594977807758, + "learning_rate": 9.533651445594289e-07, + "loss": 0.4516, + "step": 36624 + }, + { + "epoch": 0.8061853400836452, + "grad_norm": 2.139262474279236, + "learning_rate": 9.531557872053526e-07, + "loss": 0.3473, + "step": 36625 + }, + { + "epoch": 0.8062073519700639, + "grad_norm": 1.4651138671705777, + "learning_rate": 9.529464504191454e-07, + "loss": 0.2946, + "step": 36626 + }, + { + "epoch": 0.8062293638564825, + "grad_norm": 1.853574860893281, + "learning_rate": 9.527371342018677e-07, + "loss": 0.3407, + "step": 36627 + }, + { + "epoch": 0.8062513757429012, + "grad_norm": 0.6075960571565306, + "learning_rate": 9.525278385545866e-07, + "loss": 0.4531, + "step": 36628 + }, + { + "epoch": 0.8062733876293199, + "grad_norm": 9.760376664212158, + "learning_rate": 9.523185634783627e-07, + "loss": 0.3654, + "step": 36629 + }, + { + "epoch": 0.8062953995157385, + "grad_norm": 1.9242730485421529, + "learning_rate": 9.521093089742612e-07, + "loss": 0.3378, + "step": 36630 + }, + { + "epoch": 0.8063174114021572, + "grad_norm": 0.6511335938864491, + "learning_rate": 9.519000750433471e-07, + "loss": 0.4428, + "step": 36631 + }, + { + "epoch": 0.8063394232885759, + "grad_norm": 1.3002001756661492, + "learning_rate": 9.516908616866804e-07, + "loss": 0.2356, + "step": 36632 + }, + { + "epoch": 0.8063614351749945, + "grad_norm": 1.3139632974175248, + "learning_rate": 9.514816689053285e-07, + "loss": 0.3392, + "step": 36633 + }, + { + "epoch": 0.8063834470614132, + "grad_norm": 1.96863625111738, + "learning_rate": 9.512724967003506e-07, + "loss": 0.3299, + "step": 36634 + }, + { + "epoch": 0.8064054589478318, + "grad_norm": 3.7787181546272857, + "learning_rate": 9.510633450728118e-07, + "loss": 0.2217, + "step": 36635 + }, + { + "epoch": 0.8064274708342505, + "grad_norm": 31.10049129985622, + "learning_rate": 9.508542140237764e-07, + "loss": 0.2528, + "step": 36636 + }, + { + "epoch": 0.8064494827206692, + "grad_norm": 1.4459434218349172, + "learning_rate": 9.506451035543046e-07, + "loss": 0.3055, + "step": 36637 + }, + { + "epoch": 0.8064714946070878, + "grad_norm": 1.545257365229429, + "learning_rate": 9.504360136654611e-07, + "loss": 0.3298, + "step": 36638 + }, + { + "epoch": 0.8064935064935065, + "grad_norm": 1.4028869978093266, + "learning_rate": 9.502269443583074e-07, + "loss": 0.3807, + "step": 36639 + }, + { + "epoch": 0.8065155183799252, + "grad_norm": 1.5968183047735762, + "learning_rate": 9.500178956339062e-07, + "loss": 0.2243, + "step": 36640 + }, + { + "epoch": 0.8065375302663438, + "grad_norm": 2.6784651549700205, + "learning_rate": 9.49808867493322e-07, + "loss": 0.2854, + "step": 36641 + }, + { + "epoch": 0.8065595421527625, + "grad_norm": 0.6129833957175582, + "learning_rate": 9.495998599376138e-07, + "loss": 0.4436, + "step": 36642 + }, + { + "epoch": 0.8065815540391812, + "grad_norm": 2.103537661232779, + "learning_rate": 9.493908729678475e-07, + "loss": 0.2226, + "step": 36643 + }, + { + "epoch": 0.8066035659255998, + "grad_norm": 1.804856397179035, + "learning_rate": 9.491819065850827e-07, + "loss": 0.3026, + "step": 36644 + }, + { + "epoch": 0.8066255778120185, + "grad_norm": 2.063825190855664, + "learning_rate": 9.489729607903814e-07, + "loss": 0.2191, + "step": 36645 + }, + { + "epoch": 0.8066475896984372, + "grad_norm": 1.868617710312753, + "learning_rate": 9.487640355848066e-07, + "loss": 0.2608, + "step": 36646 + }, + { + "epoch": 0.8066696015848558, + "grad_norm": 1.5542162050764243, + "learning_rate": 9.485551309694191e-07, + "loss": 0.2268, + "step": 36647 + }, + { + "epoch": 0.8066916134712745, + "grad_norm": 0.5964429611765357, + "learning_rate": 9.483462469452831e-07, + "loss": 0.4457, + "step": 36648 + }, + { + "epoch": 0.8067136253576932, + "grad_norm": 1.8312477306774064, + "learning_rate": 9.481373835134588e-07, + "loss": 0.307, + "step": 36649 + }, + { + "epoch": 0.8067356372441118, + "grad_norm": 0.6151644747466493, + "learning_rate": 9.479285406750061e-07, + "loss": 0.4678, + "step": 36650 + }, + { + "epoch": 0.8067576491305305, + "grad_norm": 1.66148484220886, + "learning_rate": 9.477197184309877e-07, + "loss": 0.2553, + "step": 36651 + }, + { + "epoch": 0.8067796610169492, + "grad_norm": 0.627161707807131, + "learning_rate": 9.475109167824647e-07, + "loss": 0.471, + "step": 36652 + }, + { + "epoch": 0.8068016729033678, + "grad_norm": 2.530923581655251, + "learning_rate": 9.473021357305006e-07, + "loss": 0.2765, + "step": 36653 + }, + { + "epoch": 0.8068236847897865, + "grad_norm": 1.7067790326689023, + "learning_rate": 9.47093375276154e-07, + "loss": 0.2471, + "step": 36654 + }, + { + "epoch": 0.8068456966762052, + "grad_norm": 1.5906780026269083, + "learning_rate": 9.468846354204858e-07, + "loss": 0.2543, + "step": 36655 + }, + { + "epoch": 0.8068677085626238, + "grad_norm": 2.203134616954864, + "learning_rate": 9.466759161645572e-07, + "loss": 0.3632, + "step": 36656 + }, + { + "epoch": 0.8068897204490425, + "grad_norm": 0.6008748413564317, + "learning_rate": 9.464672175094297e-07, + "loss": 0.4312, + "step": 36657 + }, + { + "epoch": 0.8069117323354611, + "grad_norm": 1.9890765182056598, + "learning_rate": 9.462585394561651e-07, + "loss": 0.2904, + "step": 36658 + }, + { + "epoch": 0.8069337442218798, + "grad_norm": 6.255599450800168, + "learning_rate": 9.460498820058222e-07, + "loss": 0.2733, + "step": 36659 + }, + { + "epoch": 0.8069557561082985, + "grad_norm": 0.6155979427144859, + "learning_rate": 9.458412451594612e-07, + "loss": 0.4328, + "step": 36660 + }, + { + "epoch": 0.8069777679947171, + "grad_norm": 1.5750271378898437, + "learning_rate": 9.456326289181428e-07, + "loss": 0.3095, + "step": 36661 + }, + { + "epoch": 0.8069997798811358, + "grad_norm": 1.6438607591733239, + "learning_rate": 9.454240332829273e-07, + "loss": 0.2334, + "step": 36662 + }, + { + "epoch": 0.8070217917675545, + "grad_norm": 2.406468138764452, + "learning_rate": 9.452154582548772e-07, + "loss": 0.309, + "step": 36663 + }, + { + "epoch": 0.8070438036539731, + "grad_norm": 1.5982854299536142, + "learning_rate": 9.450069038350507e-07, + "loss": 0.395, + "step": 36664 + }, + { + "epoch": 0.8070658155403918, + "grad_norm": 1.9304677119002454, + "learning_rate": 9.447983700245061e-07, + "loss": 0.3262, + "step": 36665 + }, + { + "epoch": 0.8070878274268105, + "grad_norm": 0.6355908258148844, + "learning_rate": 9.445898568243045e-07, + "loss": 0.4385, + "step": 36666 + }, + { + "epoch": 0.8071098393132291, + "grad_norm": 1.592858416910317, + "learning_rate": 9.443813642355076e-07, + "loss": 0.2545, + "step": 36667 + }, + { + "epoch": 0.8071318511996478, + "grad_norm": 0.6008290093393004, + "learning_rate": 9.441728922591725e-07, + "loss": 0.4376, + "step": 36668 + }, + { + "epoch": 0.8071538630860665, + "grad_norm": 2.1416870046789542, + "learning_rate": 9.439644408963605e-07, + "loss": 0.3172, + "step": 36669 + }, + { + "epoch": 0.8071758749724851, + "grad_norm": 1.7754199814010951, + "learning_rate": 9.437560101481297e-07, + "loss": 0.3042, + "step": 36670 + }, + { + "epoch": 0.8071978868589038, + "grad_norm": 1.5190551708293716, + "learning_rate": 9.435476000155392e-07, + "loss": 0.2357, + "step": 36671 + }, + { + "epoch": 0.8072198987453225, + "grad_norm": 3.2431192443242884, + "learning_rate": 9.433392104996508e-07, + "loss": 0.3252, + "step": 36672 + }, + { + "epoch": 0.8072419106317411, + "grad_norm": 1.6432245023003027, + "learning_rate": 9.4313084160152e-07, + "loss": 0.2872, + "step": 36673 + }, + { + "epoch": 0.8072639225181598, + "grad_norm": 1.229664082288832, + "learning_rate": 9.429224933222097e-07, + "loss": 0.2695, + "step": 36674 + }, + { + "epoch": 0.8072859344045785, + "grad_norm": 1.5868974375145333, + "learning_rate": 9.427141656627747e-07, + "loss": 0.3352, + "step": 36675 + }, + { + "epoch": 0.8073079462909971, + "grad_norm": 1.4947144572326059, + "learning_rate": 9.425058586242769e-07, + "loss": 0.2497, + "step": 36676 + }, + { + "epoch": 0.8073299581774158, + "grad_norm": 1.89141362996674, + "learning_rate": 9.422975722077743e-07, + "loss": 0.2769, + "step": 36677 + }, + { + "epoch": 0.8073519700638345, + "grad_norm": 0.6208063628895304, + "learning_rate": 9.420893064143244e-07, + "loss": 0.4634, + "step": 36678 + }, + { + "epoch": 0.8073739819502531, + "grad_norm": 1.5731971693316038, + "learning_rate": 9.41881061244988e-07, + "loss": 0.2849, + "step": 36679 + }, + { + "epoch": 0.8073959938366718, + "grad_norm": 1.4742626000407175, + "learning_rate": 9.416728367008204e-07, + "loss": 0.2128, + "step": 36680 + }, + { + "epoch": 0.8074180057230904, + "grad_norm": 3.0569756009126543, + "learning_rate": 9.41464632782883e-07, + "loss": 0.4133, + "step": 36681 + }, + { + "epoch": 0.8074400176095091, + "grad_norm": 2.057205351478432, + "learning_rate": 9.41256449492231e-07, + "loss": 0.2924, + "step": 36682 + }, + { + "epoch": 0.8074620294959278, + "grad_norm": 1.4214189943172968, + "learning_rate": 9.410482868299243e-07, + "loss": 0.2427, + "step": 36683 + }, + { + "epoch": 0.8074840413823464, + "grad_norm": 0.6201509183758471, + "learning_rate": 9.408401447970217e-07, + "loss": 0.4358, + "step": 36684 + }, + { + "epoch": 0.8075060532687651, + "grad_norm": 1.5327789733219144, + "learning_rate": 9.406320233945798e-07, + "loss": 0.3745, + "step": 36685 + }, + { + "epoch": 0.8075280651551838, + "grad_norm": 1.7557778872138656, + "learning_rate": 9.404239226236555e-07, + "loss": 0.3621, + "step": 36686 + }, + { + "epoch": 0.8075500770416024, + "grad_norm": 1.6443275573963085, + "learning_rate": 9.40215842485307e-07, + "loss": 0.3029, + "step": 36687 + }, + { + "epoch": 0.8075720889280211, + "grad_norm": 1.3559335536655537, + "learning_rate": 9.400077829805926e-07, + "loss": 0.3183, + "step": 36688 + }, + { + "epoch": 0.8075941008144398, + "grad_norm": 2.6888901131241134, + "learning_rate": 9.397997441105705e-07, + "loss": 0.336, + "step": 36689 + }, + { + "epoch": 0.8076161127008584, + "grad_norm": 3.1116214639222237, + "learning_rate": 9.395917258762976e-07, + "loss": 0.359, + "step": 36690 + }, + { + "epoch": 0.8076381245872771, + "grad_norm": 1.7177693787851995, + "learning_rate": 9.393837282788287e-07, + "loss": 0.2565, + "step": 36691 + }, + { + "epoch": 0.8076601364736958, + "grad_norm": 1.6800966836793982, + "learning_rate": 9.391757513192229e-07, + "loss": 0.141, + "step": 36692 + }, + { + "epoch": 0.8076821483601144, + "grad_norm": 1.499773953053446, + "learning_rate": 9.389677949985371e-07, + "loss": 0.2204, + "step": 36693 + }, + { + "epoch": 0.8077041602465331, + "grad_norm": 1.5731719294854756, + "learning_rate": 9.387598593178293e-07, + "loss": 0.3081, + "step": 36694 + }, + { + "epoch": 0.8077261721329518, + "grad_norm": 1.965818480932175, + "learning_rate": 9.385519442781554e-07, + "loss": 0.2873, + "step": 36695 + }, + { + "epoch": 0.8077481840193704, + "grad_norm": 11.843373605900007, + "learning_rate": 9.383440498805712e-07, + "loss": 0.3404, + "step": 36696 + }, + { + "epoch": 0.8077701959057891, + "grad_norm": 2.3803531228942445, + "learning_rate": 9.381361761261337e-07, + "loss": 0.2701, + "step": 36697 + }, + { + "epoch": 0.8077922077922078, + "grad_norm": 1.5120484527929443, + "learning_rate": 9.379283230158997e-07, + "loss": 0.2916, + "step": 36698 + }, + { + "epoch": 0.8078142196786264, + "grad_norm": 0.5983138657218792, + "learning_rate": 9.377204905509269e-07, + "loss": 0.4353, + "step": 36699 + }, + { + "epoch": 0.8078362315650451, + "grad_norm": 2.363156215841571, + "learning_rate": 9.375126787322708e-07, + "loss": 0.2916, + "step": 36700 + }, + { + "epoch": 0.8078582434514638, + "grad_norm": 0.5987115742016117, + "learning_rate": 9.373048875609852e-07, + "loss": 0.4273, + "step": 36701 + }, + { + "epoch": 0.8078802553378824, + "grad_norm": 1.9413176546155768, + "learning_rate": 9.370971170381288e-07, + "loss": 0.2838, + "step": 36702 + }, + { + "epoch": 0.8079022672243011, + "grad_norm": 1.464634946176484, + "learning_rate": 9.368893671647583e-07, + "loss": 0.313, + "step": 36703 + }, + { + "epoch": 0.8079242791107197, + "grad_norm": 1.6173230047120084, + "learning_rate": 9.366816379419263e-07, + "loss": 0.2993, + "step": 36704 + }, + { + "epoch": 0.8079462909971384, + "grad_norm": 0.5989634335023578, + "learning_rate": 9.364739293706926e-07, + "loss": 0.4215, + "step": 36705 + }, + { + "epoch": 0.8079683028835571, + "grad_norm": 1.4556637768915588, + "learning_rate": 9.36266241452109e-07, + "loss": 0.3207, + "step": 36706 + }, + { + "epoch": 0.8079903147699757, + "grad_norm": 1.8667654812512133, + "learning_rate": 9.36058574187233e-07, + "loss": 0.3125, + "step": 36707 + }, + { + "epoch": 0.8080123266563944, + "grad_norm": 0.6070059053402943, + "learning_rate": 9.35850927577121e-07, + "loss": 0.4508, + "step": 36708 + }, + { + "epoch": 0.8080343385428131, + "grad_norm": 1.953564745056642, + "learning_rate": 9.356433016228261e-07, + "loss": 0.2753, + "step": 36709 + }, + { + "epoch": 0.8080563504292317, + "grad_norm": 0.6242939451907168, + "learning_rate": 9.354356963254058e-07, + "loss": 0.4644, + "step": 36710 + }, + { + "epoch": 0.8080783623156504, + "grad_norm": 2.1183654032200185, + "learning_rate": 9.35228111685913e-07, + "loss": 0.2982, + "step": 36711 + }, + { + "epoch": 0.8081003742020691, + "grad_norm": 1.327035315726115, + "learning_rate": 9.350205477054036e-07, + "loss": 0.2586, + "step": 36712 + }, + { + "epoch": 0.8081223860884877, + "grad_norm": 2.4414081522450424, + "learning_rate": 9.348130043849346e-07, + "loss": 0.2932, + "step": 36713 + }, + { + "epoch": 0.8081443979749064, + "grad_norm": 1.6872951276055788, + "learning_rate": 9.346054817255573e-07, + "loss": 0.2726, + "step": 36714 + }, + { + "epoch": 0.8081664098613252, + "grad_norm": 0.6342359267054298, + "learning_rate": 9.343979797283298e-07, + "loss": 0.4566, + "step": 36715 + }, + { + "epoch": 0.8081884217477437, + "grad_norm": 1.9031683280497147, + "learning_rate": 9.341904983943039e-07, + "loss": 0.2214, + "step": 36716 + }, + { + "epoch": 0.8082104336341625, + "grad_norm": 1.8973477910361403, + "learning_rate": 9.339830377245363e-07, + "loss": 0.2587, + "step": 36717 + }, + { + "epoch": 0.8082324455205812, + "grad_norm": 2.3766364450499746, + "learning_rate": 9.337755977200791e-07, + "loss": 0.3291, + "step": 36718 + }, + { + "epoch": 0.8082544574069997, + "grad_norm": 1.5826713545461688, + "learning_rate": 9.335681783819877e-07, + "loss": 0.2691, + "step": 36719 + }, + { + "epoch": 0.8082764692934185, + "grad_norm": 0.5839422260607339, + "learning_rate": 9.333607797113181e-07, + "loss": 0.4512, + "step": 36720 + }, + { + "epoch": 0.8082984811798372, + "grad_norm": 2.714240592756197, + "learning_rate": 9.331534017091215e-07, + "loss": 0.3139, + "step": 36721 + }, + { + "epoch": 0.8083204930662558, + "grad_norm": 1.3146091097619643, + "learning_rate": 9.329460443764543e-07, + "loss": 0.2647, + "step": 36722 + }, + { + "epoch": 0.8083425049526745, + "grad_norm": 1.6785831863500673, + "learning_rate": 9.327387077143679e-07, + "loss": 0.3124, + "step": 36723 + }, + { + "epoch": 0.8083645168390932, + "grad_norm": 1.2704193566752222, + "learning_rate": 9.325313917239182e-07, + "loss": 0.2838, + "step": 36724 + }, + { + "epoch": 0.8083865287255118, + "grad_norm": 1.3326710967996038, + "learning_rate": 9.323240964061587e-07, + "loss": 0.2842, + "step": 36725 + }, + { + "epoch": 0.8084085406119305, + "grad_norm": 1.447063845734521, + "learning_rate": 9.321168217621423e-07, + "loss": 0.3564, + "step": 36726 + }, + { + "epoch": 0.808430552498349, + "grad_norm": 1.3579477588470847, + "learning_rate": 9.319095677929213e-07, + "loss": 0.2856, + "step": 36727 + }, + { + "epoch": 0.8084525643847678, + "grad_norm": 0.6188949168475085, + "learning_rate": 9.317023344995507e-07, + "loss": 0.4391, + "step": 36728 + }, + { + "epoch": 0.8084745762711865, + "grad_norm": 1.9401508616030785, + "learning_rate": 9.314951218830831e-07, + "loss": 0.2679, + "step": 36729 + }, + { + "epoch": 0.8084965881576051, + "grad_norm": 1.4829102260068479, + "learning_rate": 9.312879299445726e-07, + "loss": 0.3166, + "step": 36730 + }, + { + "epoch": 0.8085186000440238, + "grad_norm": 1.4211729513043208, + "learning_rate": 9.310807586850724e-07, + "loss": 0.2918, + "step": 36731 + }, + { + "epoch": 0.8085406119304425, + "grad_norm": 0.6075136958259838, + "learning_rate": 9.30873608105633e-07, + "loss": 0.4508, + "step": 36732 + }, + { + "epoch": 0.8085626238168611, + "grad_norm": 2.3812204538754593, + "learning_rate": 9.306664782073089e-07, + "loss": 0.3113, + "step": 36733 + }, + { + "epoch": 0.8085846357032798, + "grad_norm": 1.3987487723264644, + "learning_rate": 9.304593689911523e-07, + "loss": 0.3403, + "step": 36734 + }, + { + "epoch": 0.8086066475896985, + "grad_norm": 3.238335833817788, + "learning_rate": 9.302522804582182e-07, + "loss": 0.3305, + "step": 36735 + }, + { + "epoch": 0.8086286594761171, + "grad_norm": 1.7051729495653867, + "learning_rate": 9.300452126095566e-07, + "loss": 0.2657, + "step": 36736 + }, + { + "epoch": 0.8086506713625358, + "grad_norm": 1.632162992221775, + "learning_rate": 9.298381654462196e-07, + "loss": 0.3994, + "step": 36737 + }, + { + "epoch": 0.8086726832489545, + "grad_norm": 1.715686629092784, + "learning_rate": 9.296311389692597e-07, + "loss": 0.3301, + "step": 36738 + }, + { + "epoch": 0.8086946951353731, + "grad_norm": 1.94880350992462, + "learning_rate": 9.294241331797305e-07, + "loss": 0.3229, + "step": 36739 + }, + { + "epoch": 0.8087167070217918, + "grad_norm": 2.2100802948547806, + "learning_rate": 9.292171480786844e-07, + "loss": 0.2736, + "step": 36740 + }, + { + "epoch": 0.8087387189082105, + "grad_norm": 3.525879817675182, + "learning_rate": 9.290101836671722e-07, + "loss": 0.3375, + "step": 36741 + }, + { + "epoch": 0.8087607307946291, + "grad_norm": 1.456098968988348, + "learning_rate": 9.288032399462454e-07, + "loss": 0.3021, + "step": 36742 + }, + { + "epoch": 0.8087827426810478, + "grad_norm": 1.6617677898211158, + "learning_rate": 9.285963169169554e-07, + "loss": 0.2786, + "step": 36743 + }, + { + "epoch": 0.8088047545674665, + "grad_norm": 1.599879506079327, + "learning_rate": 9.283894145803568e-07, + "loss": 0.3391, + "step": 36744 + }, + { + "epoch": 0.8088267664538851, + "grad_norm": 1.8568129655979155, + "learning_rate": 9.281825329374977e-07, + "loss": 0.3041, + "step": 36745 + }, + { + "epoch": 0.8088487783403038, + "grad_norm": 2.6455092341874074, + "learning_rate": 9.279756719894323e-07, + "loss": 0.3253, + "step": 36746 + }, + { + "epoch": 0.8088707902267225, + "grad_norm": 1.7724121138121836, + "learning_rate": 9.277688317372091e-07, + "loss": 0.2559, + "step": 36747 + }, + { + "epoch": 0.8088928021131411, + "grad_norm": 1.3381749527089448, + "learning_rate": 9.275620121818812e-07, + "loss": 0.2986, + "step": 36748 + }, + { + "epoch": 0.8089148139995598, + "grad_norm": 1.6078178850655236, + "learning_rate": 9.273552133245011e-07, + "loss": 0.3228, + "step": 36749 + }, + { + "epoch": 0.8089368258859785, + "grad_norm": 0.6312913311996924, + "learning_rate": 9.271484351661164e-07, + "loss": 0.4603, + "step": 36750 + }, + { + "epoch": 0.8089588377723971, + "grad_norm": 1.27136675236398, + "learning_rate": 9.269416777077816e-07, + "loss": 0.1787, + "step": 36751 + }, + { + "epoch": 0.8089808496588158, + "grad_norm": 1.4890322962902396, + "learning_rate": 9.267349409505444e-07, + "loss": 0.2828, + "step": 36752 + }, + { + "epoch": 0.8090028615452344, + "grad_norm": 1.2576733155986528, + "learning_rate": 9.265282248954583e-07, + "loss": 0.216, + "step": 36753 + }, + { + "epoch": 0.8090248734316531, + "grad_norm": 0.6272108893915368, + "learning_rate": 9.263215295435712e-07, + "loss": 0.4488, + "step": 36754 + }, + { + "epoch": 0.8090468853180718, + "grad_norm": 2.251560012210915, + "learning_rate": 9.261148548959353e-07, + "loss": 0.256, + "step": 36755 + }, + { + "epoch": 0.8090688972044904, + "grad_norm": 1.8957065725001323, + "learning_rate": 9.259082009536019e-07, + "loss": 0.2245, + "step": 36756 + }, + { + "epoch": 0.8090909090909091, + "grad_norm": 3.340043956293993, + "learning_rate": 9.257015677176184e-07, + "loss": 0.2958, + "step": 36757 + }, + { + "epoch": 0.8091129209773278, + "grad_norm": 1.5862232040795141, + "learning_rate": 9.254949551890386e-07, + "loss": 0.2884, + "step": 36758 + }, + { + "epoch": 0.8091349328637464, + "grad_norm": 1.6017838138285418, + "learning_rate": 9.252883633689092e-07, + "loss": 0.3339, + "step": 36759 + }, + { + "epoch": 0.8091569447501651, + "grad_norm": 1.5644955997911572, + "learning_rate": 9.250817922582817e-07, + "loss": 0.2794, + "step": 36760 + }, + { + "epoch": 0.8091789566365838, + "grad_norm": 6.275501656725492, + "learning_rate": 9.248752418582069e-07, + "loss": 0.2619, + "step": 36761 + }, + { + "epoch": 0.8092009685230024, + "grad_norm": 1.5215677215886594, + "learning_rate": 9.246687121697328e-07, + "loss": 0.2693, + "step": 36762 + }, + { + "epoch": 0.8092229804094211, + "grad_norm": 0.6129950323270089, + "learning_rate": 9.24462203193911e-07, + "loss": 0.469, + "step": 36763 + }, + { + "epoch": 0.8092449922958398, + "grad_norm": 1.8744441311944688, + "learning_rate": 9.242557149317893e-07, + "loss": 0.2016, + "step": 36764 + }, + { + "epoch": 0.8092670041822584, + "grad_norm": 1.9237984814798412, + "learning_rate": 9.240492473844176e-07, + "loss": 0.2939, + "step": 36765 + }, + { + "epoch": 0.8092890160686771, + "grad_norm": 2.220063314053793, + "learning_rate": 9.238428005528466e-07, + "loss": 0.3545, + "step": 36766 + }, + { + "epoch": 0.8093110279550958, + "grad_norm": 2.8041297756607277, + "learning_rate": 9.236363744381244e-07, + "loss": 0.2897, + "step": 36767 + }, + { + "epoch": 0.8093330398415144, + "grad_norm": 1.4221724914710618, + "learning_rate": 9.234299690412995e-07, + "loss": 0.2611, + "step": 36768 + }, + { + "epoch": 0.8093550517279331, + "grad_norm": 1.5394218151460946, + "learning_rate": 9.232235843634213e-07, + "loss": 0.3219, + "step": 36769 + }, + { + "epoch": 0.8093770636143518, + "grad_norm": 0.603243932068824, + "learning_rate": 9.230172204055394e-07, + "loss": 0.4092, + "step": 36770 + }, + { + "epoch": 0.8093990755007704, + "grad_norm": 1.3927510093939293, + "learning_rate": 9.228108771687033e-07, + "loss": 0.2825, + "step": 36771 + }, + { + "epoch": 0.8094210873871891, + "grad_norm": 1.3311228275675089, + "learning_rate": 9.226045546539609e-07, + "loss": 0.2726, + "step": 36772 + }, + { + "epoch": 0.8094430992736078, + "grad_norm": 1.7731545017311623, + "learning_rate": 9.223982528623593e-07, + "loss": 0.3585, + "step": 36773 + }, + { + "epoch": 0.8094651111600264, + "grad_norm": 2.0242270711385206, + "learning_rate": 9.22191971794949e-07, + "loss": 0.2991, + "step": 36774 + }, + { + "epoch": 0.8094871230464451, + "grad_norm": 0.6221200813388517, + "learning_rate": 9.219857114527775e-07, + "loss": 0.4382, + "step": 36775 + }, + { + "epoch": 0.8095091349328637, + "grad_norm": 1.7196036885987966, + "learning_rate": 9.21779471836895e-07, + "loss": 0.2682, + "step": 36776 + }, + { + "epoch": 0.8095311468192824, + "grad_norm": 0.6044445761165661, + "learning_rate": 9.215732529483479e-07, + "loss": 0.4624, + "step": 36777 + }, + { + "epoch": 0.8095531587057011, + "grad_norm": 2.4244567889441893, + "learning_rate": 9.213670547881831e-07, + "loss": 0.1969, + "step": 36778 + }, + { + "epoch": 0.8095751705921197, + "grad_norm": 2.7194174102512654, + "learning_rate": 9.211608773574504e-07, + "loss": 0.4376, + "step": 36779 + }, + { + "epoch": 0.8095971824785384, + "grad_norm": 2.261036268447344, + "learning_rate": 9.209547206571972e-07, + "loss": 0.2618, + "step": 36780 + }, + { + "epoch": 0.8096191943649571, + "grad_norm": 1.4371547243298468, + "learning_rate": 9.207485846884723e-07, + "loss": 0.2539, + "step": 36781 + }, + { + "epoch": 0.8096412062513757, + "grad_norm": 1.5814434262998132, + "learning_rate": 9.205424694523229e-07, + "loss": 0.2194, + "step": 36782 + }, + { + "epoch": 0.8096632181377944, + "grad_norm": 1.406781892109542, + "learning_rate": 9.203363749497951e-07, + "loss": 0.2615, + "step": 36783 + }, + { + "epoch": 0.8096852300242131, + "grad_norm": 2.729040102528437, + "learning_rate": 9.201303011819384e-07, + "loss": 0.3061, + "step": 36784 + }, + { + "epoch": 0.8097072419106317, + "grad_norm": 1.6115468482702868, + "learning_rate": 9.199242481497977e-07, + "loss": 0.2562, + "step": 36785 + }, + { + "epoch": 0.8097292537970504, + "grad_norm": 1.7073539117101426, + "learning_rate": 9.197182158544221e-07, + "loss": 0.3239, + "step": 36786 + }, + { + "epoch": 0.8097512656834691, + "grad_norm": 2.058399123739674, + "learning_rate": 9.195122042968596e-07, + "loss": 0.3094, + "step": 36787 + }, + { + "epoch": 0.8097732775698877, + "grad_norm": 1.5031525286837213, + "learning_rate": 9.19306213478155e-07, + "loss": 0.3135, + "step": 36788 + }, + { + "epoch": 0.8097952894563064, + "grad_norm": 2.8002760876959756, + "learning_rate": 9.191002433993568e-07, + "loss": 0.2475, + "step": 36789 + }, + { + "epoch": 0.8098173013427251, + "grad_norm": 1.4566281131847534, + "learning_rate": 9.188942940615108e-07, + "loss": 0.2864, + "step": 36790 + }, + { + "epoch": 0.8098393132291437, + "grad_norm": 1.6491922282211091, + "learning_rate": 9.186883654656637e-07, + "loss": 0.3117, + "step": 36791 + }, + { + "epoch": 0.8098613251155624, + "grad_norm": 1.354751120791385, + "learning_rate": 9.184824576128643e-07, + "loss": 0.2476, + "step": 36792 + }, + { + "epoch": 0.8098833370019811, + "grad_norm": 1.4858731126752387, + "learning_rate": 9.182765705041558e-07, + "loss": 0.2608, + "step": 36793 + }, + { + "epoch": 0.8099053488883997, + "grad_norm": 0.6300822238319714, + "learning_rate": 9.180707041405879e-07, + "loss": 0.4344, + "step": 36794 + }, + { + "epoch": 0.8099273607748184, + "grad_norm": 0.5979639624146208, + "learning_rate": 9.178648585232036e-07, + "loss": 0.4182, + "step": 36795 + }, + { + "epoch": 0.8099493726612371, + "grad_norm": 3.317029993426891, + "learning_rate": 9.176590336530511e-07, + "loss": 0.3201, + "step": 36796 + }, + { + "epoch": 0.8099713845476557, + "grad_norm": 1.6207604466552141, + "learning_rate": 9.174532295311772e-07, + "loss": 0.3197, + "step": 36797 + }, + { + "epoch": 0.8099933964340744, + "grad_norm": 2.328631228855755, + "learning_rate": 9.172474461586256e-07, + "loss": 0.3116, + "step": 36798 + }, + { + "epoch": 0.810015408320493, + "grad_norm": 2.3228082438896993, + "learning_rate": 9.170416835364449e-07, + "loss": 0.3684, + "step": 36799 + }, + { + "epoch": 0.8100374202069117, + "grad_norm": 3.3432548898842374, + "learning_rate": 9.16835941665678e-07, + "loss": 0.3122, + "step": 36800 + }, + { + "epoch": 0.8100594320933304, + "grad_norm": 0.6001326652242518, + "learning_rate": 9.16630220547372e-07, + "loss": 0.4226, + "step": 36801 + }, + { + "epoch": 0.810081443979749, + "grad_norm": 2.3771606361721127, + "learning_rate": 9.164245201825739e-07, + "loss": 0.3683, + "step": 36802 + }, + { + "epoch": 0.8101034558661677, + "grad_norm": 1.5362162825161287, + "learning_rate": 9.162188405723277e-07, + "loss": 0.3591, + "step": 36803 + }, + { + "epoch": 0.8101254677525864, + "grad_norm": 1.7694494919366028, + "learning_rate": 9.160131817176776e-07, + "loss": 0.3084, + "step": 36804 + }, + { + "epoch": 0.810147479639005, + "grad_norm": 1.7836951681564663, + "learning_rate": 9.158075436196695e-07, + "loss": 0.3477, + "step": 36805 + }, + { + "epoch": 0.8101694915254237, + "grad_norm": 2.5165123452729095, + "learning_rate": 9.156019262793497e-07, + "loss": 0.3581, + "step": 36806 + }, + { + "epoch": 0.8101915034118424, + "grad_norm": 1.4563285301443671, + "learning_rate": 9.153963296977636e-07, + "loss": 0.2803, + "step": 36807 + }, + { + "epoch": 0.810213515298261, + "grad_norm": 1.997160807870328, + "learning_rate": 9.151907538759546e-07, + "loss": 0.3721, + "step": 36808 + }, + { + "epoch": 0.8102355271846797, + "grad_norm": 2.138322058508847, + "learning_rate": 9.149851988149677e-07, + "loss": 0.3112, + "step": 36809 + }, + { + "epoch": 0.8102575390710984, + "grad_norm": 1.8180654102902647, + "learning_rate": 9.147796645158474e-07, + "loss": 0.2497, + "step": 36810 + }, + { + "epoch": 0.810279550957517, + "grad_norm": 2.0901653370363933, + "learning_rate": 9.145741509796391e-07, + "loss": 0.2461, + "step": 36811 + }, + { + "epoch": 0.8103015628439357, + "grad_norm": 1.256760284800502, + "learning_rate": 9.143686582073885e-07, + "loss": 0.2731, + "step": 36812 + }, + { + "epoch": 0.8103235747303544, + "grad_norm": 1.5780804189772464, + "learning_rate": 9.141631862001383e-07, + "loss": 0.3227, + "step": 36813 + }, + { + "epoch": 0.810345586616773, + "grad_norm": 0.6247605777510657, + "learning_rate": 9.139577349589324e-07, + "loss": 0.4478, + "step": 36814 + }, + { + "epoch": 0.8103675985031917, + "grad_norm": 0.6156616865300368, + "learning_rate": 9.137523044848151e-07, + "loss": 0.4388, + "step": 36815 + }, + { + "epoch": 0.8103896103896104, + "grad_norm": 0.6243199647444693, + "learning_rate": 9.135468947788312e-07, + "loss": 0.4417, + "step": 36816 + }, + { + "epoch": 0.810411622276029, + "grad_norm": 1.598004312113793, + "learning_rate": 9.133415058420258e-07, + "loss": 0.2196, + "step": 36817 + }, + { + "epoch": 0.8104336341624477, + "grad_norm": 1.4989905384771551, + "learning_rate": 9.131361376754422e-07, + "loss": 0.3167, + "step": 36818 + }, + { + "epoch": 0.8104556460488664, + "grad_norm": 1.7452274672585883, + "learning_rate": 9.129307902801216e-07, + "loss": 0.25, + "step": 36819 + }, + { + "epoch": 0.810477657935285, + "grad_norm": 0.6381523571042896, + "learning_rate": 9.127254636571114e-07, + "loss": 0.4371, + "step": 36820 + }, + { + "epoch": 0.8104996698217037, + "grad_norm": 2.376211359241781, + "learning_rate": 9.125201578074505e-07, + "loss": 0.3028, + "step": 36821 + }, + { + "epoch": 0.8105216817081223, + "grad_norm": 0.6160590751914078, + "learning_rate": 9.123148727321879e-07, + "loss": 0.4551, + "step": 36822 + }, + { + "epoch": 0.810543693594541, + "grad_norm": 0.6219015447176723, + "learning_rate": 9.121096084323639e-07, + "loss": 0.4373, + "step": 36823 + }, + { + "epoch": 0.8105657054809597, + "grad_norm": 1.6817047940505712, + "learning_rate": 9.119043649090214e-07, + "loss": 0.3415, + "step": 36824 + }, + { + "epoch": 0.8105877173673783, + "grad_norm": 2.2132654884740104, + "learning_rate": 9.116991421632054e-07, + "loss": 0.3352, + "step": 36825 + }, + { + "epoch": 0.810609729253797, + "grad_norm": 7.080379902864089, + "learning_rate": 9.11493940195956e-07, + "loss": 0.269, + "step": 36826 + }, + { + "epoch": 0.8106317411402157, + "grad_norm": 7.801911073843248, + "learning_rate": 9.112887590083186e-07, + "loss": 0.4, + "step": 36827 + }, + { + "epoch": 0.8106537530266343, + "grad_norm": 0.6244423205067209, + "learning_rate": 9.110835986013361e-07, + "loss": 0.4367, + "step": 36828 + }, + { + "epoch": 0.810675764913053, + "grad_norm": 1.868282699333576, + "learning_rate": 9.108784589760494e-07, + "loss": 0.2107, + "step": 36829 + }, + { + "epoch": 0.8106977767994717, + "grad_norm": 1.539116685844438, + "learning_rate": 9.106733401335038e-07, + "loss": 0.3506, + "step": 36830 + }, + { + "epoch": 0.8107197886858903, + "grad_norm": 1.5199970190572738, + "learning_rate": 9.104682420747385e-07, + "loss": 0.3252, + "step": 36831 + }, + { + "epoch": 0.810741800572309, + "grad_norm": 1.5900084447701903, + "learning_rate": 9.102631648007975e-07, + "loss": 0.2179, + "step": 36832 + }, + { + "epoch": 0.8107638124587278, + "grad_norm": 4.830197515928977, + "learning_rate": 9.100581083127247e-07, + "loss": 0.2713, + "step": 36833 + }, + { + "epoch": 0.8107858243451463, + "grad_norm": 1.695016694806774, + "learning_rate": 9.098530726115596e-07, + "loss": 0.2805, + "step": 36834 + }, + { + "epoch": 0.810807836231565, + "grad_norm": 1.3560162788372216, + "learning_rate": 9.096480576983463e-07, + "loss": 0.2365, + "step": 36835 + }, + { + "epoch": 0.8108298481179838, + "grad_norm": 1.42900602187519, + "learning_rate": 9.094430635741247e-07, + "loss": 0.2589, + "step": 36836 + }, + { + "epoch": 0.8108518600044023, + "grad_norm": 1.9324076356751148, + "learning_rate": 9.092380902399378e-07, + "loss": 0.329, + "step": 36837 + }, + { + "epoch": 0.810873871890821, + "grad_norm": 1.70036809725883, + "learning_rate": 9.090331376968293e-07, + "loss": 0.2444, + "step": 36838 + }, + { + "epoch": 0.8108958837772398, + "grad_norm": 1.6773877039605236, + "learning_rate": 9.088282059458376e-07, + "loss": 0.2617, + "step": 36839 + }, + { + "epoch": 0.8109178956636584, + "grad_norm": 1.638274582423852, + "learning_rate": 9.086232949880064e-07, + "loss": 0.2588, + "step": 36840 + }, + { + "epoch": 0.810939907550077, + "grad_norm": 2.137815176697644, + "learning_rate": 9.084184048243755e-07, + "loss": 0.282, + "step": 36841 + }, + { + "epoch": 0.8109619194364958, + "grad_norm": 3.153139226818156, + "learning_rate": 9.082135354559873e-07, + "loss": 0.2771, + "step": 36842 + }, + { + "epoch": 0.8109839313229144, + "grad_norm": 0.6123959665087799, + "learning_rate": 9.080086868838838e-07, + "loss": 0.4443, + "step": 36843 + }, + { + "epoch": 0.8110059432093331, + "grad_norm": 18.228910384323154, + "learning_rate": 9.078038591091054e-07, + "loss": 0.3966, + "step": 36844 + }, + { + "epoch": 0.8110279550957517, + "grad_norm": 2.8883932950250766, + "learning_rate": 9.075990521326922e-07, + "loss": 0.3747, + "step": 36845 + }, + { + "epoch": 0.8110499669821704, + "grad_norm": 1.3830137407562555, + "learning_rate": 9.073942659556851e-07, + "loss": 0.2191, + "step": 36846 + }, + { + "epoch": 0.8110719788685891, + "grad_norm": 1.205829359864858, + "learning_rate": 9.07189500579126e-07, + "loss": 0.2221, + "step": 36847 + }, + { + "epoch": 0.8110939907550077, + "grad_norm": 1.215958944180209, + "learning_rate": 9.069847560040568e-07, + "loss": 0.2792, + "step": 36848 + }, + { + "epoch": 0.8111160026414264, + "grad_norm": 0.6326502737033753, + "learning_rate": 9.067800322315163e-07, + "loss": 0.4507, + "step": 36849 + }, + { + "epoch": 0.8111380145278451, + "grad_norm": 0.6318917799981583, + "learning_rate": 9.06575329262544e-07, + "loss": 0.4458, + "step": 36850 + }, + { + "epoch": 0.8111600264142637, + "grad_norm": 0.618592681214268, + "learning_rate": 9.06370647098182e-07, + "loss": 0.4629, + "step": 36851 + }, + { + "epoch": 0.8111820383006824, + "grad_norm": 1.34172561106181, + "learning_rate": 9.061659857394695e-07, + "loss": 0.234, + "step": 36852 + }, + { + "epoch": 0.8112040501871011, + "grad_norm": 1.764699902416281, + "learning_rate": 9.059613451874494e-07, + "loss": 0.3126, + "step": 36853 + }, + { + "epoch": 0.8112260620735197, + "grad_norm": 2.977521089923947, + "learning_rate": 9.057567254431593e-07, + "loss": 0.232, + "step": 36854 + }, + { + "epoch": 0.8112480739599384, + "grad_norm": 1.7843861050898495, + "learning_rate": 9.055521265076383e-07, + "loss": 0.3441, + "step": 36855 + }, + { + "epoch": 0.8112700858463571, + "grad_norm": 1.6855167514493734, + "learning_rate": 9.053475483819291e-07, + "loss": 0.2672, + "step": 36856 + }, + { + "epoch": 0.8112920977327757, + "grad_norm": 1.7722243601718055, + "learning_rate": 9.051429910670672e-07, + "loss": 0.3694, + "step": 36857 + }, + { + "epoch": 0.8113141096191944, + "grad_norm": 1.4420260651154035, + "learning_rate": 9.049384545640976e-07, + "loss": 0.243, + "step": 36858 + }, + { + "epoch": 0.8113361215056131, + "grad_norm": 1.654437323551438, + "learning_rate": 9.047339388740573e-07, + "loss": 0.3555, + "step": 36859 + }, + { + "epoch": 0.8113581333920317, + "grad_norm": 1.8862255855925316, + "learning_rate": 9.04529443997984e-07, + "loss": 0.3042, + "step": 36860 + }, + { + "epoch": 0.8113801452784504, + "grad_norm": 2.359224862118788, + "learning_rate": 9.0432496993692e-07, + "loss": 0.3509, + "step": 36861 + }, + { + "epoch": 0.8114021571648691, + "grad_norm": 1.2408070191985359, + "learning_rate": 9.041205166919009e-07, + "loss": 0.1908, + "step": 36862 + }, + { + "epoch": 0.8114241690512877, + "grad_norm": 1.8345309266644634, + "learning_rate": 9.039160842639711e-07, + "loss": 0.2541, + "step": 36863 + }, + { + "epoch": 0.8114461809377064, + "grad_norm": 1.5423598688870277, + "learning_rate": 9.037116726541661e-07, + "loss": 0.3064, + "step": 36864 + }, + { + "epoch": 0.8114681928241251, + "grad_norm": 1.6236538262875078, + "learning_rate": 9.035072818635243e-07, + "loss": 0.2679, + "step": 36865 + }, + { + "epoch": 0.8114902047105437, + "grad_norm": 2.3421651824385825, + "learning_rate": 9.033029118930869e-07, + "loss": 0.3353, + "step": 36866 + }, + { + "epoch": 0.8115122165969624, + "grad_norm": 0.6071986009980663, + "learning_rate": 9.030985627438898e-07, + "loss": 0.4601, + "step": 36867 + }, + { + "epoch": 0.811534228483381, + "grad_norm": 1.4512860057580017, + "learning_rate": 9.028942344169734e-07, + "loss": 0.2577, + "step": 36868 + }, + { + "epoch": 0.8115562403697997, + "grad_norm": 1.2980001679882334, + "learning_rate": 9.026899269133765e-07, + "loss": 0.2572, + "step": 36869 + }, + { + "epoch": 0.8115782522562184, + "grad_norm": 1.8242540092035018, + "learning_rate": 9.024856402341365e-07, + "loss": 0.2259, + "step": 36870 + }, + { + "epoch": 0.811600264142637, + "grad_norm": 1.426049056093026, + "learning_rate": 9.022813743802927e-07, + "loss": 0.3213, + "step": 36871 + }, + { + "epoch": 0.8116222760290557, + "grad_norm": 1.343987842455788, + "learning_rate": 9.020771293528813e-07, + "loss": 0.263, + "step": 36872 + }, + { + "epoch": 0.8116442879154744, + "grad_norm": 1.4832997931900533, + "learning_rate": 9.018729051529418e-07, + "loss": 0.3023, + "step": 36873 + }, + { + "epoch": 0.811666299801893, + "grad_norm": 1.6831721267688944, + "learning_rate": 9.016687017815134e-07, + "loss": 0.3027, + "step": 36874 + }, + { + "epoch": 0.8116883116883117, + "grad_norm": 1.6499865317161937, + "learning_rate": 9.014645192396309e-07, + "loss": 0.3323, + "step": 36875 + }, + { + "epoch": 0.8117103235747304, + "grad_norm": 1.8007219757100945, + "learning_rate": 9.012603575283351e-07, + "loss": 0.252, + "step": 36876 + }, + { + "epoch": 0.811732335461149, + "grad_norm": 1.4600182671867472, + "learning_rate": 9.010562166486614e-07, + "loss": 0.2955, + "step": 36877 + }, + { + "epoch": 0.8117543473475677, + "grad_norm": 0.5998073844624028, + "learning_rate": 9.008520966016476e-07, + "loss": 0.4309, + "step": 36878 + }, + { + "epoch": 0.8117763592339864, + "grad_norm": 1.1737099987581445, + "learning_rate": 9.006479973883336e-07, + "loss": 0.172, + "step": 36879 + }, + { + "epoch": 0.811798371120405, + "grad_norm": 1.3202975725221833, + "learning_rate": 9.00443919009753e-07, + "loss": 0.2132, + "step": 36880 + }, + { + "epoch": 0.8118203830068237, + "grad_norm": 1.732200873837527, + "learning_rate": 9.002398614669467e-07, + "loss": 0.3084, + "step": 36881 + }, + { + "epoch": 0.8118423948932424, + "grad_norm": 1.2457191563353454, + "learning_rate": 9.00035824760948e-07, + "loss": 0.2595, + "step": 36882 + }, + { + "epoch": 0.811864406779661, + "grad_norm": 1.5851677572215637, + "learning_rate": 8.998318088927966e-07, + "loss": 0.321, + "step": 36883 + }, + { + "epoch": 0.8118864186660797, + "grad_norm": 1.3768088695376945, + "learning_rate": 8.996278138635295e-07, + "loss": 0.2587, + "step": 36884 + }, + { + "epoch": 0.8119084305524984, + "grad_norm": 1.4872448534531968, + "learning_rate": 8.994238396741827e-07, + "loss": 0.3368, + "step": 36885 + }, + { + "epoch": 0.811930442438917, + "grad_norm": 2.2970986998156615, + "learning_rate": 8.992198863257917e-07, + "loss": 0.2274, + "step": 36886 + }, + { + "epoch": 0.8119524543253357, + "grad_norm": 0.5837043292884415, + "learning_rate": 8.990159538193943e-07, + "loss": 0.4389, + "step": 36887 + }, + { + "epoch": 0.8119744662117544, + "grad_norm": 1.4884538876211504, + "learning_rate": 8.988120421560265e-07, + "loss": 0.2155, + "step": 36888 + }, + { + "epoch": 0.811996478098173, + "grad_norm": 2.6609249811523954, + "learning_rate": 8.986081513367267e-07, + "loss": 0.3644, + "step": 36889 + }, + { + "epoch": 0.8120184899845917, + "grad_norm": 1.8054901997662782, + "learning_rate": 8.984042813625294e-07, + "loss": 0.2341, + "step": 36890 + }, + { + "epoch": 0.8120405018710104, + "grad_norm": 2.3763947519159125, + "learning_rate": 8.982004322344701e-07, + "loss": 0.291, + "step": 36891 + }, + { + "epoch": 0.812062513757429, + "grad_norm": 3.1638887101153665, + "learning_rate": 8.979966039535865e-07, + "loss": 0.2228, + "step": 36892 + }, + { + "epoch": 0.8120845256438477, + "grad_norm": 0.6126304085054219, + "learning_rate": 8.97792796520911e-07, + "loss": 0.4418, + "step": 36893 + }, + { + "epoch": 0.8121065375302663, + "grad_norm": 2.244341799002371, + "learning_rate": 8.975890099374851e-07, + "loss": 0.3118, + "step": 36894 + }, + { + "epoch": 0.812128549416685, + "grad_norm": 1.6954899138121615, + "learning_rate": 8.973852442043412e-07, + "loss": 0.303, + "step": 36895 + }, + { + "epoch": 0.8121505613031037, + "grad_norm": 0.6476676302195314, + "learning_rate": 8.971814993225142e-07, + "loss": 0.4752, + "step": 36896 + }, + { + "epoch": 0.8121725731895223, + "grad_norm": 2.430727980275141, + "learning_rate": 8.96977775293042e-07, + "loss": 0.2317, + "step": 36897 + }, + { + "epoch": 0.812194585075941, + "grad_norm": 0.6109021977991234, + "learning_rate": 8.96774072116956e-07, + "loss": 0.4821, + "step": 36898 + }, + { + "epoch": 0.8122165969623597, + "grad_norm": 1.4288317825726844, + "learning_rate": 8.965703897952971e-07, + "loss": 0.3735, + "step": 36899 + }, + { + "epoch": 0.8122386088487783, + "grad_norm": 1.7783551362106167, + "learning_rate": 8.963667283290977e-07, + "loss": 0.3267, + "step": 36900 + }, + { + "epoch": 0.812260620735197, + "grad_norm": 1.6576165164480603, + "learning_rate": 8.961630877193917e-07, + "loss": 0.2794, + "step": 36901 + }, + { + "epoch": 0.8122826326216157, + "grad_norm": 1.8455023246409044, + "learning_rate": 8.959594679672162e-07, + "loss": 0.2979, + "step": 36902 + }, + { + "epoch": 0.8123046445080343, + "grad_norm": 1.78123105231138, + "learning_rate": 8.957558690736041e-07, + "loss": 0.2705, + "step": 36903 + }, + { + "epoch": 0.812326656394453, + "grad_norm": 2.401783059392063, + "learning_rate": 8.955522910395908e-07, + "loss": 0.277, + "step": 36904 + }, + { + "epoch": 0.8123486682808717, + "grad_norm": 0.6352660372131455, + "learning_rate": 8.953487338662132e-07, + "loss": 0.4627, + "step": 36905 + }, + { + "epoch": 0.8123706801672903, + "grad_norm": 1.307458931227972, + "learning_rate": 8.951451975545022e-07, + "loss": 0.2856, + "step": 36906 + }, + { + "epoch": 0.812392692053709, + "grad_norm": 1.814361834579335, + "learning_rate": 8.949416821054957e-07, + "loss": 0.3468, + "step": 36907 + }, + { + "epoch": 0.8124147039401277, + "grad_norm": 1.8671814324290332, + "learning_rate": 8.947381875202254e-07, + "loss": 0.3314, + "step": 36908 + }, + { + "epoch": 0.8124367158265463, + "grad_norm": 1.661705727270323, + "learning_rate": 8.945347137997263e-07, + "loss": 0.2143, + "step": 36909 + }, + { + "epoch": 0.812458727712965, + "grad_norm": 1.7042912951485483, + "learning_rate": 8.943312609450344e-07, + "loss": 0.3158, + "step": 36910 + }, + { + "epoch": 0.8124807395993837, + "grad_norm": 1.8551648486079821, + "learning_rate": 8.941278289571809e-07, + "loss": 0.2883, + "step": 36911 + }, + { + "epoch": 0.8125027514858023, + "grad_norm": 1.8339790380164334, + "learning_rate": 8.93924417837202e-07, + "loss": 0.2803, + "step": 36912 + }, + { + "epoch": 0.812524763372221, + "grad_norm": 1.8660656286486919, + "learning_rate": 8.937210275861297e-07, + "loss": 0.34, + "step": 36913 + }, + { + "epoch": 0.8125467752586397, + "grad_norm": 1.9753874845143076, + "learning_rate": 8.935176582049982e-07, + "loss": 0.3504, + "step": 36914 + }, + { + "epoch": 0.8125687871450583, + "grad_norm": 1.7782239667736275, + "learning_rate": 8.933143096948433e-07, + "loss": 0.2535, + "step": 36915 + }, + { + "epoch": 0.812590799031477, + "grad_norm": 1.770409802706466, + "learning_rate": 8.931109820566952e-07, + "loss": 0.2804, + "step": 36916 + }, + { + "epoch": 0.8126128109178956, + "grad_norm": 1.3415105935401337, + "learning_rate": 8.9290767529159e-07, + "loss": 0.1938, + "step": 36917 + }, + { + "epoch": 0.8126348228043143, + "grad_norm": 1.537665261195311, + "learning_rate": 8.927043894005583e-07, + "loss": 0.3059, + "step": 36918 + }, + { + "epoch": 0.812656834690733, + "grad_norm": 1.6719068091921927, + "learning_rate": 8.925011243846354e-07, + "loss": 0.3398, + "step": 36919 + }, + { + "epoch": 0.8126788465771516, + "grad_norm": 2.0223767269483948, + "learning_rate": 8.922978802448545e-07, + "loss": 0.2551, + "step": 36920 + }, + { + "epoch": 0.8127008584635703, + "grad_norm": 3.5333697336911154, + "learning_rate": 8.920946569822469e-07, + "loss": 0.3085, + "step": 36921 + }, + { + "epoch": 0.812722870349989, + "grad_norm": 0.6097245257626862, + "learning_rate": 8.918914545978474e-07, + "loss": 0.4502, + "step": 36922 + }, + { + "epoch": 0.8127448822364076, + "grad_norm": 1.9367843300151262, + "learning_rate": 8.916882730926868e-07, + "loss": 0.2763, + "step": 36923 + }, + { + "epoch": 0.8127668941228263, + "grad_norm": 1.4269806694566867, + "learning_rate": 8.914851124677987e-07, + "loss": 0.2554, + "step": 36924 + }, + { + "epoch": 0.812788906009245, + "grad_norm": 1.5779443774331259, + "learning_rate": 8.912819727242173e-07, + "loss": 0.2865, + "step": 36925 + }, + { + "epoch": 0.8128109178956636, + "grad_norm": 2.4043892467291226, + "learning_rate": 8.910788538629728e-07, + "loss": 0.3197, + "step": 36926 + }, + { + "epoch": 0.8128329297820823, + "grad_norm": 1.6724087643014516, + "learning_rate": 8.908757558850972e-07, + "loss": 0.3093, + "step": 36927 + }, + { + "epoch": 0.812854941668501, + "grad_norm": 3.1001648613392705, + "learning_rate": 8.906726787916253e-07, + "loss": 0.3082, + "step": 36928 + }, + { + "epoch": 0.8128769535549196, + "grad_norm": 2.061478250172449, + "learning_rate": 8.904696225835846e-07, + "loss": 0.2631, + "step": 36929 + }, + { + "epoch": 0.8128989654413383, + "grad_norm": 1.2640730430838083, + "learning_rate": 8.902665872620131e-07, + "loss": 0.2665, + "step": 36930 + }, + { + "epoch": 0.812920977327757, + "grad_norm": 2.734954639022737, + "learning_rate": 8.900635728279394e-07, + "loss": 0.3587, + "step": 36931 + }, + { + "epoch": 0.8129429892141756, + "grad_norm": 0.5989482385687601, + "learning_rate": 8.898605792823949e-07, + "loss": 0.4129, + "step": 36932 + }, + { + "epoch": 0.8129650011005943, + "grad_norm": 2.538402228911638, + "learning_rate": 8.896576066264135e-07, + "loss": 0.4237, + "step": 36933 + }, + { + "epoch": 0.812987012987013, + "grad_norm": 2.1402458923449608, + "learning_rate": 8.894546548610233e-07, + "loss": 0.3485, + "step": 36934 + }, + { + "epoch": 0.8130090248734316, + "grad_norm": 0.6188822730611921, + "learning_rate": 8.892517239872589e-07, + "loss": 0.4697, + "step": 36935 + }, + { + "epoch": 0.8130310367598503, + "grad_norm": 1.4122771058926873, + "learning_rate": 8.890488140061515e-07, + "loss": 0.3902, + "step": 36936 + }, + { + "epoch": 0.813053048646269, + "grad_norm": 1.7901107451834537, + "learning_rate": 8.888459249187297e-07, + "loss": 0.3133, + "step": 36937 + }, + { + "epoch": 0.8130750605326876, + "grad_norm": 1.4481214132824345, + "learning_rate": 8.886430567260284e-07, + "loss": 0.2478, + "step": 36938 + }, + { + "epoch": 0.8130970724191063, + "grad_norm": 1.3216721865227947, + "learning_rate": 8.884402094290757e-07, + "loss": 0.3027, + "step": 36939 + }, + { + "epoch": 0.8131190843055249, + "grad_norm": 0.5901458962868544, + "learning_rate": 8.882373830289037e-07, + "loss": 0.439, + "step": 36940 + }, + { + "epoch": 0.8131410961919436, + "grad_norm": 0.5952538443260507, + "learning_rate": 8.880345775265442e-07, + "loss": 0.4023, + "step": 36941 + }, + { + "epoch": 0.8131631080783623, + "grad_norm": 6.277765149429661, + "learning_rate": 8.878317929230257e-07, + "loss": 0.335, + "step": 36942 + }, + { + "epoch": 0.8131851199647809, + "grad_norm": 1.4870608019877032, + "learning_rate": 8.876290292193813e-07, + "loss": 0.2131, + "step": 36943 + }, + { + "epoch": 0.8132071318511996, + "grad_norm": 0.6050664452285501, + "learning_rate": 8.874262864166394e-07, + "loss": 0.4353, + "step": 36944 + }, + { + "epoch": 0.8132291437376183, + "grad_norm": 1.6303308799542933, + "learning_rate": 8.872235645158311e-07, + "loss": 0.2774, + "step": 36945 + }, + { + "epoch": 0.8132511556240369, + "grad_norm": 1.6916132713865735, + "learning_rate": 8.870208635179883e-07, + "loss": 0.3413, + "step": 36946 + }, + { + "epoch": 0.8132731675104556, + "grad_norm": 1.8963103757533248, + "learning_rate": 8.868181834241385e-07, + "loss": 0.3048, + "step": 36947 + }, + { + "epoch": 0.8132951793968743, + "grad_norm": 3.4137413142207076, + "learning_rate": 8.866155242353147e-07, + "loss": 0.3173, + "step": 36948 + }, + { + "epoch": 0.8133171912832929, + "grad_norm": 2.0740490956070494, + "learning_rate": 8.864128859525445e-07, + "loss": 0.3002, + "step": 36949 + }, + { + "epoch": 0.8133392031697116, + "grad_norm": 0.6226375973925985, + "learning_rate": 8.862102685768581e-07, + "loss": 0.4476, + "step": 36950 + }, + { + "epoch": 0.8133612150561303, + "grad_norm": 1.7798491689961529, + "learning_rate": 8.860076721092875e-07, + "loss": 0.2678, + "step": 36951 + }, + { + "epoch": 0.8133832269425489, + "grad_norm": 1.5826872966592944, + "learning_rate": 8.858050965508596e-07, + "loss": 0.2658, + "step": 36952 + }, + { + "epoch": 0.8134052388289676, + "grad_norm": 2.1707121862052885, + "learning_rate": 8.856025419026065e-07, + "loss": 0.2961, + "step": 36953 + }, + { + "epoch": 0.8134272507153864, + "grad_norm": 1.6203026053202767, + "learning_rate": 8.854000081655544e-07, + "loss": 0.2704, + "step": 36954 + }, + { + "epoch": 0.813449262601805, + "grad_norm": 2.830562599186009, + "learning_rate": 8.851974953407356e-07, + "loss": 0.2316, + "step": 36955 + }, + { + "epoch": 0.8134712744882237, + "grad_norm": 1.9291878231907473, + "learning_rate": 8.849950034291788e-07, + "loss": 0.3415, + "step": 36956 + }, + { + "epoch": 0.8134932863746424, + "grad_norm": 1.33968370736979, + "learning_rate": 8.847925324319124e-07, + "loss": 0.2935, + "step": 36957 + }, + { + "epoch": 0.813515298261061, + "grad_norm": 1.5818665401025507, + "learning_rate": 8.845900823499659e-07, + "loss": 0.3188, + "step": 36958 + }, + { + "epoch": 0.8135373101474797, + "grad_norm": 0.6248301910841845, + "learning_rate": 8.843876531843676e-07, + "loss": 0.4511, + "step": 36959 + }, + { + "epoch": 0.8135593220338984, + "grad_norm": 2.151202874206924, + "learning_rate": 8.841852449361472e-07, + "loss": 0.25, + "step": 36960 + }, + { + "epoch": 0.813581333920317, + "grad_norm": 1.7666730148617105, + "learning_rate": 8.839828576063336e-07, + "loss": 0.2702, + "step": 36961 + }, + { + "epoch": 0.8136033458067357, + "grad_norm": 1.4819582862235159, + "learning_rate": 8.837804911959541e-07, + "loss": 0.3259, + "step": 36962 + }, + { + "epoch": 0.8136253576931542, + "grad_norm": 2.3131821626025104, + "learning_rate": 8.835781457060388e-07, + "loss": 0.3161, + "step": 36963 + }, + { + "epoch": 0.813647369579573, + "grad_norm": 1.6683793861188594, + "learning_rate": 8.833758211376159e-07, + "loss": 0.2678, + "step": 36964 + }, + { + "epoch": 0.8136693814659917, + "grad_norm": 1.643851175838465, + "learning_rate": 8.8317351749171e-07, + "loss": 0.2749, + "step": 36965 + }, + { + "epoch": 0.8136913933524103, + "grad_norm": 1.474834703199821, + "learning_rate": 8.829712347693553e-07, + "loss": 0.2573, + "step": 36966 + }, + { + "epoch": 0.813713405238829, + "grad_norm": 2.654320768732527, + "learning_rate": 8.827689729715771e-07, + "loss": 0.3109, + "step": 36967 + }, + { + "epoch": 0.8137354171252477, + "grad_norm": 2.0847609852781344, + "learning_rate": 8.825667320994024e-07, + "loss": 0.273, + "step": 36968 + }, + { + "epoch": 0.8137574290116663, + "grad_norm": 0.5963824338672279, + "learning_rate": 8.823645121538604e-07, + "loss": 0.4375, + "step": 36969 + }, + { + "epoch": 0.813779440898085, + "grad_norm": 1.4996563535735898, + "learning_rate": 8.82162313135978e-07, + "loss": 0.3595, + "step": 36970 + }, + { + "epoch": 0.8138014527845037, + "grad_norm": 1.4542211420513649, + "learning_rate": 8.819601350467832e-07, + "loss": 0.3257, + "step": 36971 + }, + { + "epoch": 0.8138234646709223, + "grad_norm": 0.6142573235809035, + "learning_rate": 8.817579778873048e-07, + "loss": 0.4314, + "step": 36972 + }, + { + "epoch": 0.813845476557341, + "grad_norm": 3.086996005718224, + "learning_rate": 8.815558416585679e-07, + "loss": 0.2805, + "step": 36973 + }, + { + "epoch": 0.8138674884437597, + "grad_norm": 2.058689875307551, + "learning_rate": 8.813537263616018e-07, + "loss": 0.3443, + "step": 36974 + }, + { + "epoch": 0.8138895003301783, + "grad_norm": 1.9268909160087304, + "learning_rate": 8.811516319974322e-07, + "loss": 0.2838, + "step": 36975 + }, + { + "epoch": 0.813911512216597, + "grad_norm": 1.489465644629932, + "learning_rate": 8.809495585670874e-07, + "loss": 0.2717, + "step": 36976 + }, + { + "epoch": 0.8139335241030157, + "grad_norm": 0.601382842558654, + "learning_rate": 8.807475060715953e-07, + "loss": 0.4688, + "step": 36977 + }, + { + "epoch": 0.8139555359894343, + "grad_norm": 1.495221200913148, + "learning_rate": 8.805454745119801e-07, + "loss": 0.2212, + "step": 36978 + }, + { + "epoch": 0.813977547875853, + "grad_norm": 1.4496071280788112, + "learning_rate": 8.803434638892711e-07, + "loss": 0.3193, + "step": 36979 + }, + { + "epoch": 0.8139995597622717, + "grad_norm": 1.6356475750191601, + "learning_rate": 8.801414742044934e-07, + "loss": 0.378, + "step": 36980 + }, + { + "epoch": 0.8140215716486903, + "grad_norm": 1.5428047334954849, + "learning_rate": 8.79939505458674e-07, + "loss": 0.2391, + "step": 36981 + }, + { + "epoch": 0.814043583535109, + "grad_norm": 1.5096544981060667, + "learning_rate": 8.797375576528405e-07, + "loss": 0.2558, + "step": 36982 + }, + { + "epoch": 0.8140655954215277, + "grad_norm": 1.4721579173227284, + "learning_rate": 8.795356307880176e-07, + "loss": 0.3218, + "step": 36983 + }, + { + "epoch": 0.8140876073079463, + "grad_norm": 0.6178576372341359, + "learning_rate": 8.793337248652334e-07, + "loss": 0.4499, + "step": 36984 + }, + { + "epoch": 0.814109619194365, + "grad_norm": 0.6154479822762003, + "learning_rate": 8.79131839885512e-07, + "loss": 0.4446, + "step": 36985 + }, + { + "epoch": 0.8141316310807836, + "grad_norm": 0.6288556449334746, + "learning_rate": 8.78929975849881e-07, + "loss": 0.4276, + "step": 36986 + }, + { + "epoch": 0.8141536429672023, + "grad_norm": 2.0001123275999046, + "learning_rate": 8.787281327593661e-07, + "loss": 0.2727, + "step": 36987 + }, + { + "epoch": 0.814175654853621, + "grad_norm": 1.5618404924344318, + "learning_rate": 8.785263106149927e-07, + "loss": 0.2286, + "step": 36988 + }, + { + "epoch": 0.8141976667400396, + "grad_norm": 1.592163678934862, + "learning_rate": 8.783245094177878e-07, + "loss": 0.4257, + "step": 36989 + }, + { + "epoch": 0.8142196786264583, + "grad_norm": 1.3025044986236536, + "learning_rate": 8.781227291687744e-07, + "loss": 0.2741, + "step": 36990 + }, + { + "epoch": 0.814241690512877, + "grad_norm": 1.7193844514694707, + "learning_rate": 8.7792096986898e-07, + "loss": 0.2714, + "step": 36991 + }, + { + "epoch": 0.8142637023992956, + "grad_norm": 1.8146484927750228, + "learning_rate": 8.777192315194311e-07, + "loss": 0.3624, + "step": 36992 + }, + { + "epoch": 0.8142857142857143, + "grad_norm": 1.603504906476313, + "learning_rate": 8.7751751412115e-07, + "loss": 0.3076, + "step": 36993 + }, + { + "epoch": 0.814307726172133, + "grad_norm": 2.097961342522611, + "learning_rate": 8.77315817675165e-07, + "loss": 0.3313, + "step": 36994 + }, + { + "epoch": 0.8143297380585516, + "grad_norm": 2.7779328405757178, + "learning_rate": 8.771141421825002e-07, + "loss": 0.2915, + "step": 36995 + }, + { + "epoch": 0.8143517499449703, + "grad_norm": 0.6057799857017482, + "learning_rate": 8.76912487644177e-07, + "loss": 0.4403, + "step": 36996 + }, + { + "epoch": 0.814373761831389, + "grad_norm": 2.8987046444929248, + "learning_rate": 8.767108540612268e-07, + "loss": 0.3851, + "step": 36997 + }, + { + "epoch": 0.8143957737178076, + "grad_norm": 2.138619634085016, + "learning_rate": 8.765092414346693e-07, + "loss": 0.3074, + "step": 36998 + }, + { + "epoch": 0.8144177856042263, + "grad_norm": 1.3729679866509448, + "learning_rate": 8.763076497655321e-07, + "loss": 0.265, + "step": 36999 + }, + { + "epoch": 0.814439797490645, + "grad_norm": 2.4536095603629295, + "learning_rate": 8.761060790548387e-07, + "loss": 0.3318, + "step": 37000 + }, + { + "epoch": 0.8144618093770636, + "grad_norm": 2.0317099361905693, + "learning_rate": 8.759045293036111e-07, + "loss": 0.3391, + "step": 37001 + }, + { + "epoch": 0.8144838212634823, + "grad_norm": 1.429649017683632, + "learning_rate": 8.757030005128786e-07, + "loss": 0.3137, + "step": 37002 + }, + { + "epoch": 0.814505833149901, + "grad_norm": 0.6535138651400397, + "learning_rate": 8.75501492683663e-07, + "loss": 0.4563, + "step": 37003 + }, + { + "epoch": 0.8145278450363196, + "grad_norm": 1.47027510164792, + "learning_rate": 8.753000058169874e-07, + "loss": 0.2685, + "step": 37004 + }, + { + "epoch": 0.8145498569227383, + "grad_norm": 1.5842194854459892, + "learning_rate": 8.750985399138778e-07, + "loss": 0.3023, + "step": 37005 + }, + { + "epoch": 0.814571868809157, + "grad_norm": 1.9544911404032181, + "learning_rate": 8.74897094975356e-07, + "loss": 0.2699, + "step": 37006 + }, + { + "epoch": 0.8145938806955756, + "grad_norm": 0.6379303784545336, + "learning_rate": 8.746956710024468e-07, + "loss": 0.4638, + "step": 37007 + }, + { + "epoch": 0.8146158925819943, + "grad_norm": 9.402390952036297, + "learning_rate": 8.744942679961754e-07, + "loss": 0.2386, + "step": 37008 + }, + { + "epoch": 0.8146379044684129, + "grad_norm": 1.3654808072593914, + "learning_rate": 8.742928859575628e-07, + "loss": 0.2036, + "step": 37009 + }, + { + "epoch": 0.8146599163548316, + "grad_norm": 1.7625891682693486, + "learning_rate": 8.740915248876358e-07, + "loss": 0.2286, + "step": 37010 + }, + { + "epoch": 0.8146819282412503, + "grad_norm": 1.6947043387409064, + "learning_rate": 8.738901847874138e-07, + "loss": 0.2663, + "step": 37011 + }, + { + "epoch": 0.8147039401276689, + "grad_norm": 1.5063278634706387, + "learning_rate": 8.736888656579228e-07, + "loss": 0.2829, + "step": 37012 + }, + { + "epoch": 0.8147259520140876, + "grad_norm": 1.050964034135527, + "learning_rate": 8.734875675001864e-07, + "loss": 0.2299, + "step": 37013 + }, + { + "epoch": 0.8147479639005063, + "grad_norm": 2.725196147113993, + "learning_rate": 8.73286290315225e-07, + "loss": 0.2769, + "step": 37014 + }, + { + "epoch": 0.8147699757869249, + "grad_norm": 1.7232191630040763, + "learning_rate": 8.730850341040653e-07, + "loss": 0.2692, + "step": 37015 + }, + { + "epoch": 0.8147919876733436, + "grad_norm": 1.8399434979972837, + "learning_rate": 8.728837988677263e-07, + "loss": 0.3127, + "step": 37016 + }, + { + "epoch": 0.8148139995597623, + "grad_norm": 1.498447160662966, + "learning_rate": 8.726825846072329e-07, + "loss": 0.2809, + "step": 37017 + }, + { + "epoch": 0.8148360114461809, + "grad_norm": 3.403486476122763, + "learning_rate": 8.724813913236079e-07, + "loss": 0.3679, + "step": 37018 + }, + { + "epoch": 0.8148580233325996, + "grad_norm": 2.472826082329064, + "learning_rate": 8.722802190178731e-07, + "loss": 0.1977, + "step": 37019 + }, + { + "epoch": 0.8148800352190183, + "grad_norm": 1.565878085045018, + "learning_rate": 8.720790676910523e-07, + "loss": 0.3061, + "step": 37020 + }, + { + "epoch": 0.8149020471054369, + "grad_norm": 1.7259526019931382, + "learning_rate": 8.718779373441649e-07, + "loss": 0.2622, + "step": 37021 + }, + { + "epoch": 0.8149240589918556, + "grad_norm": 1.4374392278312838, + "learning_rate": 8.716768279782356e-07, + "loss": 0.2578, + "step": 37022 + }, + { + "epoch": 0.8149460708782743, + "grad_norm": 1.784360067015226, + "learning_rate": 8.714757395942874e-07, + "loss": 0.2755, + "step": 37023 + }, + { + "epoch": 0.8149680827646929, + "grad_norm": 4.991441677555351, + "learning_rate": 8.71274672193339e-07, + "loss": 0.3439, + "step": 37024 + }, + { + "epoch": 0.8149900946511116, + "grad_norm": 2.177557960731204, + "learning_rate": 8.710736257764158e-07, + "loss": 0.383, + "step": 37025 + }, + { + "epoch": 0.8150121065375303, + "grad_norm": 1.5556999353916383, + "learning_rate": 8.708726003445367e-07, + "loss": 0.2367, + "step": 37026 + }, + { + "epoch": 0.8150341184239489, + "grad_norm": 0.6220527133971153, + "learning_rate": 8.706715958987244e-07, + "loss": 0.4468, + "step": 37027 + }, + { + "epoch": 0.8150561303103676, + "grad_norm": 1.737304034768055, + "learning_rate": 8.704706124400026e-07, + "loss": 0.33, + "step": 37028 + }, + { + "epoch": 0.8150781421967863, + "grad_norm": 0.6115972848098157, + "learning_rate": 8.702696499693892e-07, + "loss": 0.4325, + "step": 37029 + }, + { + "epoch": 0.8151001540832049, + "grad_norm": 2.042992191827557, + "learning_rate": 8.700687084879089e-07, + "loss": 0.322, + "step": 37030 + }, + { + "epoch": 0.8151221659696236, + "grad_norm": 1.4992345605078647, + "learning_rate": 8.698677879965811e-07, + "loss": 0.299, + "step": 37031 + }, + { + "epoch": 0.8151441778560423, + "grad_norm": 1.6632472372198335, + "learning_rate": 8.696668884964254e-07, + "loss": 0.3249, + "step": 37032 + }, + { + "epoch": 0.8151661897424609, + "grad_norm": 0.6162596800837231, + "learning_rate": 8.694660099884666e-07, + "loss": 0.4409, + "step": 37033 + }, + { + "epoch": 0.8151882016288796, + "grad_norm": 2.1682084717476564, + "learning_rate": 8.692651524737228e-07, + "loss": 0.2711, + "step": 37034 + }, + { + "epoch": 0.8152102135152982, + "grad_norm": 1.6620388775621768, + "learning_rate": 8.690643159532169e-07, + "loss": 0.2293, + "step": 37035 + }, + { + "epoch": 0.8152322254017169, + "grad_norm": 1.4423494278913935, + "learning_rate": 8.688635004279689e-07, + "loss": 0.3344, + "step": 37036 + }, + { + "epoch": 0.8152542372881356, + "grad_norm": 1.6915226848334621, + "learning_rate": 8.686627058989966e-07, + "loss": 0.2542, + "step": 37037 + }, + { + "epoch": 0.8152762491745542, + "grad_norm": 1.4985525691220671, + "learning_rate": 8.684619323673255e-07, + "loss": 0.2695, + "step": 37038 + }, + { + "epoch": 0.8152982610609729, + "grad_norm": 2.1092884394882367, + "learning_rate": 8.682611798339724e-07, + "loss": 0.2659, + "step": 37039 + }, + { + "epoch": 0.8153202729473916, + "grad_norm": 1.6010881406836612, + "learning_rate": 8.6806044829996e-07, + "loss": 0.2832, + "step": 37040 + }, + { + "epoch": 0.8153422848338102, + "grad_norm": 2.0488759794151576, + "learning_rate": 8.678597377663073e-07, + "loss": 0.294, + "step": 37041 + }, + { + "epoch": 0.8153642967202289, + "grad_norm": 1.9109075738987653, + "learning_rate": 8.67659048234033e-07, + "loss": 0.2877, + "step": 37042 + }, + { + "epoch": 0.8153863086066476, + "grad_norm": 2.012850734074387, + "learning_rate": 8.674583797041586e-07, + "loss": 0.4127, + "step": 37043 + }, + { + "epoch": 0.8154083204930662, + "grad_norm": 1.804431735468354, + "learning_rate": 8.672577321777053e-07, + "loss": 0.3433, + "step": 37044 + }, + { + "epoch": 0.8154303323794849, + "grad_norm": 0.6249094448273348, + "learning_rate": 8.670571056556898e-07, + "loss": 0.4753, + "step": 37045 + }, + { + "epoch": 0.8154523442659036, + "grad_norm": 1.8015428404509977, + "learning_rate": 8.668565001391349e-07, + "loss": 0.2721, + "step": 37046 + }, + { + "epoch": 0.8154743561523222, + "grad_norm": 1.802699405065522, + "learning_rate": 8.666559156290577e-07, + "loss": 0.2294, + "step": 37047 + }, + { + "epoch": 0.8154963680387409, + "grad_norm": 1.4866732775656053, + "learning_rate": 8.664553521264784e-07, + "loss": 0.2277, + "step": 37048 + }, + { + "epoch": 0.8155183799251596, + "grad_norm": 1.544270590881031, + "learning_rate": 8.66254809632418e-07, + "loss": 0.2721, + "step": 37049 + }, + { + "epoch": 0.8155403918115782, + "grad_norm": 2.4637004527562887, + "learning_rate": 8.660542881478928e-07, + "loss": 0.3317, + "step": 37050 + }, + { + "epoch": 0.8155624036979969, + "grad_norm": 1.8807175520748187, + "learning_rate": 8.658537876739254e-07, + "loss": 0.2961, + "step": 37051 + }, + { + "epoch": 0.8155844155844156, + "grad_norm": 0.6242106859970166, + "learning_rate": 8.65653308211531e-07, + "loss": 0.4585, + "step": 37052 + }, + { + "epoch": 0.8156064274708342, + "grad_norm": 1.6731292580731412, + "learning_rate": 8.65452849761731e-07, + "loss": 0.3177, + "step": 37053 + }, + { + "epoch": 0.8156284393572529, + "grad_norm": 1.8080343871971647, + "learning_rate": 8.652524123255446e-07, + "loss": 0.3693, + "step": 37054 + }, + { + "epoch": 0.8156504512436716, + "grad_norm": 1.3437250860668555, + "learning_rate": 8.650519959039882e-07, + "loss": 0.2123, + "step": 37055 + }, + { + "epoch": 0.8156724631300902, + "grad_norm": 4.071006272934947, + "learning_rate": 8.648516004980828e-07, + "loss": 0.2681, + "step": 37056 + }, + { + "epoch": 0.8156944750165089, + "grad_norm": 1.4154867486425293, + "learning_rate": 8.646512261088453e-07, + "loss": 0.3231, + "step": 37057 + }, + { + "epoch": 0.8157164869029275, + "grad_norm": 1.4159848793489402, + "learning_rate": 8.644508727372947e-07, + "loss": 0.327, + "step": 37058 + }, + { + "epoch": 0.8157384987893462, + "grad_norm": 0.5965534132964716, + "learning_rate": 8.642505403844498e-07, + "loss": 0.434, + "step": 37059 + }, + { + "epoch": 0.8157605106757649, + "grad_norm": 1.8386519278448865, + "learning_rate": 8.640502290513275e-07, + "loss": 0.3013, + "step": 37060 + }, + { + "epoch": 0.8157825225621835, + "grad_norm": 2.0605226720548675, + "learning_rate": 8.638499387389481e-07, + "loss": 0.4354, + "step": 37061 + }, + { + "epoch": 0.8158045344486022, + "grad_norm": 1.5155094581270168, + "learning_rate": 8.636496694483265e-07, + "loss": 0.3008, + "step": 37062 + }, + { + "epoch": 0.8158265463350209, + "grad_norm": 15.454522673458628, + "learning_rate": 8.634494211804823e-07, + "loss": 0.3281, + "step": 37063 + }, + { + "epoch": 0.8158485582214395, + "grad_norm": 1.6578381459948908, + "learning_rate": 8.63249193936434e-07, + "loss": 0.2284, + "step": 37064 + }, + { + "epoch": 0.8158705701078582, + "grad_norm": 0.6193420088713154, + "learning_rate": 8.630489877171977e-07, + "loss": 0.4675, + "step": 37065 + }, + { + "epoch": 0.8158925819942769, + "grad_norm": 1.5753305329539238, + "learning_rate": 8.628488025237925e-07, + "loss": 0.3075, + "step": 37066 + }, + { + "epoch": 0.8159145938806955, + "grad_norm": 1.6915269110699906, + "learning_rate": 8.626486383572347e-07, + "loss": 0.2737, + "step": 37067 + }, + { + "epoch": 0.8159366057671142, + "grad_norm": 0.6124915194816563, + "learning_rate": 8.624484952185397e-07, + "loss": 0.4615, + "step": 37068 + }, + { + "epoch": 0.815958617653533, + "grad_norm": 2.4379124549023534, + "learning_rate": 8.622483731087289e-07, + "loss": 0.2654, + "step": 37069 + }, + { + "epoch": 0.8159806295399515, + "grad_norm": 1.5292245783828406, + "learning_rate": 8.620482720288164e-07, + "loss": 0.2569, + "step": 37070 + }, + { + "epoch": 0.8160026414263702, + "grad_norm": 1.651458386928309, + "learning_rate": 8.618481919798216e-07, + "loss": 0.3608, + "step": 37071 + }, + { + "epoch": 0.816024653312789, + "grad_norm": 1.9144981866255866, + "learning_rate": 8.616481329627596e-07, + "loss": 0.3552, + "step": 37072 + }, + { + "epoch": 0.8160466651992075, + "grad_norm": 1.3997201689843768, + "learning_rate": 8.614480949786452e-07, + "loss": 0.3422, + "step": 37073 + }, + { + "epoch": 0.8160686770856262, + "grad_norm": 0.6316104253452756, + "learning_rate": 8.612480780284998e-07, + "loss": 0.4536, + "step": 37074 + }, + { + "epoch": 0.816090688972045, + "grad_norm": 1.6065057720069054, + "learning_rate": 8.610480821133365e-07, + "loss": 0.2966, + "step": 37075 + }, + { + "epoch": 0.8161127008584635, + "grad_norm": 2.107591990625284, + "learning_rate": 8.60848107234174e-07, + "loss": 0.3032, + "step": 37076 + }, + { + "epoch": 0.8161347127448823, + "grad_norm": 1.6015094222922877, + "learning_rate": 8.606481533920274e-07, + "loss": 0.2577, + "step": 37077 + }, + { + "epoch": 0.816156724631301, + "grad_norm": 2.8563563800176333, + "learning_rate": 8.604482205879122e-07, + "loss": 0.2597, + "step": 37078 + }, + { + "epoch": 0.8161787365177195, + "grad_norm": 1.3382595219049556, + "learning_rate": 8.602483088228452e-07, + "loss": 0.2706, + "step": 37079 + }, + { + "epoch": 0.8162007484041383, + "grad_norm": 1.5674051726903229, + "learning_rate": 8.60048418097843e-07, + "loss": 0.2992, + "step": 37080 + }, + { + "epoch": 0.8162227602905568, + "grad_norm": 1.572544333491584, + "learning_rate": 8.598485484139224e-07, + "loss": 0.2486, + "step": 37081 + }, + { + "epoch": 0.8162447721769756, + "grad_norm": 2.010094312483317, + "learning_rate": 8.596486997720977e-07, + "loss": 0.2564, + "step": 37082 + }, + { + "epoch": 0.8162667840633943, + "grad_norm": 2.113230072035892, + "learning_rate": 8.594488721733834e-07, + "loss": 0.2392, + "step": 37083 + }, + { + "epoch": 0.8162887959498129, + "grad_norm": 2.053990588818316, + "learning_rate": 8.592490656187973e-07, + "loss": 0.2947, + "step": 37084 + }, + { + "epoch": 0.8163108078362316, + "grad_norm": 1.5227400866706606, + "learning_rate": 8.590492801093553e-07, + "loss": 0.2744, + "step": 37085 + }, + { + "epoch": 0.8163328197226503, + "grad_norm": 2.093767074554851, + "learning_rate": 8.588495156460708e-07, + "loss": 0.2826, + "step": 37086 + }, + { + "epoch": 0.8163548316090689, + "grad_norm": 0.6034251932070914, + "learning_rate": 8.586497722299608e-07, + "loss": 0.4539, + "step": 37087 + }, + { + "epoch": 0.8163768434954876, + "grad_norm": 14.776712624874975, + "learning_rate": 8.584500498620391e-07, + "loss": 0.2582, + "step": 37088 + }, + { + "epoch": 0.8163988553819063, + "grad_norm": 1.7499965032640683, + "learning_rate": 8.582503485433213e-07, + "loss": 0.3329, + "step": 37089 + }, + { + "epoch": 0.8164208672683249, + "grad_norm": 0.6169701802435209, + "learning_rate": 8.580506682748235e-07, + "loss": 0.4422, + "step": 37090 + }, + { + "epoch": 0.8164428791547436, + "grad_norm": 1.7199529484635443, + "learning_rate": 8.578510090575581e-07, + "loss": 0.304, + "step": 37091 + }, + { + "epoch": 0.8164648910411623, + "grad_norm": 1.4846409276904777, + "learning_rate": 8.576513708925432e-07, + "loss": 0.3083, + "step": 37092 + }, + { + "epoch": 0.8164869029275809, + "grad_norm": 1.5071950756837595, + "learning_rate": 8.574517537807897e-07, + "loss": 0.3526, + "step": 37093 + }, + { + "epoch": 0.8165089148139996, + "grad_norm": 1.3755987110677284, + "learning_rate": 8.572521577233145e-07, + "loss": 0.2479, + "step": 37094 + }, + { + "epoch": 0.8165309267004183, + "grad_norm": 1.5445268682521716, + "learning_rate": 8.570525827211324e-07, + "loss": 0.3411, + "step": 37095 + }, + { + "epoch": 0.8165529385868369, + "grad_norm": 1.7329360610343825, + "learning_rate": 8.568530287752558e-07, + "loss": 0.2895, + "step": 37096 + }, + { + "epoch": 0.8165749504732556, + "grad_norm": 1.4307646140663928, + "learning_rate": 8.566534958867013e-07, + "loss": 0.2562, + "step": 37097 + }, + { + "epoch": 0.8165969623596743, + "grad_norm": 2.1073130357058045, + "learning_rate": 8.564539840564806e-07, + "loss": 0.2323, + "step": 37098 + }, + { + "epoch": 0.8166189742460929, + "grad_norm": 1.6698533590261875, + "learning_rate": 8.562544932856088e-07, + "loss": 0.2915, + "step": 37099 + }, + { + "epoch": 0.8166409861325116, + "grad_norm": 4.0517489139344445, + "learning_rate": 8.560550235751008e-07, + "loss": 0.3358, + "step": 37100 + }, + { + "epoch": 0.8166629980189303, + "grad_norm": 1.5898867202601723, + "learning_rate": 8.558555749259684e-07, + "loss": 0.2698, + "step": 37101 + }, + { + "epoch": 0.8166850099053489, + "grad_norm": 2.025129897946003, + "learning_rate": 8.556561473392272e-07, + "loss": 0.2823, + "step": 37102 + }, + { + "epoch": 0.8167070217917676, + "grad_norm": 1.5458929173414682, + "learning_rate": 8.554567408158904e-07, + "loss": 0.294, + "step": 37103 + }, + { + "epoch": 0.8167290336781862, + "grad_norm": 9.640966109195203, + "learning_rate": 8.552573553569681e-07, + "loss": 0.24, + "step": 37104 + }, + { + "epoch": 0.8167510455646049, + "grad_norm": 2.1134893032762596, + "learning_rate": 8.550579909634793e-07, + "loss": 0.2973, + "step": 37105 + }, + { + "epoch": 0.8167730574510236, + "grad_norm": 1.2074774944507176, + "learning_rate": 8.548586476364329e-07, + "loss": 0.2475, + "step": 37106 + }, + { + "epoch": 0.8167950693374422, + "grad_norm": 0.6320234627103275, + "learning_rate": 8.546593253768454e-07, + "loss": 0.4435, + "step": 37107 + }, + { + "epoch": 0.8168170812238609, + "grad_norm": 1.4987736129039295, + "learning_rate": 8.544600241857276e-07, + "loss": 0.2907, + "step": 37108 + }, + { + "epoch": 0.8168390931102796, + "grad_norm": 0.6201332424799691, + "learning_rate": 8.542607440640905e-07, + "loss": 0.4452, + "step": 37109 + }, + { + "epoch": 0.8168611049966982, + "grad_norm": 0.6393914497635693, + "learning_rate": 8.540614850129519e-07, + "loss": 0.4416, + "step": 37110 + }, + { + "epoch": 0.8168831168831169, + "grad_norm": 1.7949440205510205, + "learning_rate": 8.538622470333207e-07, + "loss": 0.252, + "step": 37111 + }, + { + "epoch": 0.8169051287695356, + "grad_norm": 1.4867330887261363, + "learning_rate": 8.536630301262122e-07, + "loss": 0.2358, + "step": 37112 + }, + { + "epoch": 0.8169271406559542, + "grad_norm": 1.61658976722805, + "learning_rate": 8.534638342926371e-07, + "loss": 0.3618, + "step": 37113 + }, + { + "epoch": 0.8169491525423729, + "grad_norm": 1.3044707166478444, + "learning_rate": 8.532646595336075e-07, + "loss": 0.1857, + "step": 37114 + }, + { + "epoch": 0.8169711644287916, + "grad_norm": 1.9399316749574784, + "learning_rate": 8.530655058501363e-07, + "loss": 0.2487, + "step": 37115 + }, + { + "epoch": 0.8169931763152102, + "grad_norm": 2.954371368084724, + "learning_rate": 8.528663732432357e-07, + "loss": 0.2776, + "step": 37116 + }, + { + "epoch": 0.8170151882016289, + "grad_norm": 1.673790525962095, + "learning_rate": 8.526672617139187e-07, + "loss": 0.3052, + "step": 37117 + }, + { + "epoch": 0.8170372000880476, + "grad_norm": 1.5145362989298519, + "learning_rate": 8.524681712631971e-07, + "loss": 0.2724, + "step": 37118 + }, + { + "epoch": 0.8170592119744662, + "grad_norm": 1.2566279448350588, + "learning_rate": 8.522691018920803e-07, + "loss": 0.3456, + "step": 37119 + }, + { + "epoch": 0.8170812238608849, + "grad_norm": 1.5211097859831189, + "learning_rate": 8.520700536015819e-07, + "loss": 0.2675, + "step": 37120 + }, + { + "epoch": 0.8171032357473036, + "grad_norm": 1.9488444474393205, + "learning_rate": 8.518710263927138e-07, + "loss": 0.2395, + "step": 37121 + }, + { + "epoch": 0.8171252476337222, + "grad_norm": 2.281003622839564, + "learning_rate": 8.516720202664885e-07, + "loss": 0.3174, + "step": 37122 + }, + { + "epoch": 0.8171472595201409, + "grad_norm": 1.5457686184776098, + "learning_rate": 8.514730352239165e-07, + "loss": 0.2172, + "step": 37123 + }, + { + "epoch": 0.8171692714065596, + "grad_norm": 2.4859964204508045, + "learning_rate": 8.512740712660067e-07, + "loss": 0.3369, + "step": 37124 + }, + { + "epoch": 0.8171912832929782, + "grad_norm": 1.952519231929161, + "learning_rate": 8.510751283937729e-07, + "loss": 0.3835, + "step": 37125 + }, + { + "epoch": 0.8172132951793969, + "grad_norm": 3.000195984528742, + "learning_rate": 8.508762066082271e-07, + "loss": 0.3249, + "step": 37126 + }, + { + "epoch": 0.8172353070658155, + "grad_norm": 1.9033392246521268, + "learning_rate": 8.506773059103779e-07, + "loss": 0.3039, + "step": 37127 + }, + { + "epoch": 0.8172573189522342, + "grad_norm": 1.6403779724556617, + "learning_rate": 8.504784263012383e-07, + "loss": 0.25, + "step": 37128 + }, + { + "epoch": 0.8172793308386529, + "grad_norm": 1.804923049590196, + "learning_rate": 8.502795677818171e-07, + "loss": 0.3523, + "step": 37129 + }, + { + "epoch": 0.8173013427250715, + "grad_norm": 1.8202870435993095, + "learning_rate": 8.500807303531256e-07, + "loss": 0.2603, + "step": 37130 + }, + { + "epoch": 0.8173233546114902, + "grad_norm": 2.0947448815653837, + "learning_rate": 8.498819140161763e-07, + "loss": 0.2929, + "step": 37131 + }, + { + "epoch": 0.8173453664979089, + "grad_norm": 1.4961511780785102, + "learning_rate": 8.496831187719768e-07, + "loss": 0.2986, + "step": 37132 + }, + { + "epoch": 0.8173673783843275, + "grad_norm": 1.5907677325387028, + "learning_rate": 8.494843446215395e-07, + "loss": 0.3099, + "step": 37133 + }, + { + "epoch": 0.8173893902707462, + "grad_norm": 2.5215050933137593, + "learning_rate": 8.492855915658732e-07, + "loss": 0.2881, + "step": 37134 + }, + { + "epoch": 0.8174114021571649, + "grad_norm": 1.6500677497462704, + "learning_rate": 8.490868596059887e-07, + "loss": 0.2918, + "step": 37135 + }, + { + "epoch": 0.8174334140435835, + "grad_norm": 1.44474563196609, + "learning_rate": 8.488881487428974e-07, + "loss": 0.2549, + "step": 37136 + }, + { + "epoch": 0.8174554259300022, + "grad_norm": 1.5167262648422768, + "learning_rate": 8.486894589776068e-07, + "loss": 0.2617, + "step": 37137 + }, + { + "epoch": 0.8174774378164209, + "grad_norm": 1.963328751238077, + "learning_rate": 8.484907903111289e-07, + "loss": 0.281, + "step": 37138 + }, + { + "epoch": 0.8174994497028395, + "grad_norm": 1.6385432388406, + "learning_rate": 8.482921427444712e-07, + "loss": 0.3237, + "step": 37139 + }, + { + "epoch": 0.8175214615892582, + "grad_norm": 0.6078048440328498, + "learning_rate": 8.480935162786447e-07, + "loss": 0.4486, + "step": 37140 + }, + { + "epoch": 0.8175434734756769, + "grad_norm": 1.9996796148693603, + "learning_rate": 8.4789491091466e-07, + "loss": 0.2458, + "step": 37141 + }, + { + "epoch": 0.8175654853620955, + "grad_norm": 1.7673087134890024, + "learning_rate": 8.476963266535243e-07, + "loss": 0.2937, + "step": 37142 + }, + { + "epoch": 0.8175874972485142, + "grad_norm": 0.6250215537066686, + "learning_rate": 8.474977634962489e-07, + "loss": 0.4582, + "step": 37143 + }, + { + "epoch": 0.8176095091349329, + "grad_norm": 0.6449326776954406, + "learning_rate": 8.472992214438419e-07, + "loss": 0.4544, + "step": 37144 + }, + { + "epoch": 0.8176315210213515, + "grad_norm": 0.6179282385575556, + "learning_rate": 8.471007004973108e-07, + "loss": 0.4571, + "step": 37145 + }, + { + "epoch": 0.8176535329077702, + "grad_norm": 1.7548374986137536, + "learning_rate": 8.469022006576666e-07, + "loss": 0.3447, + "step": 37146 + }, + { + "epoch": 0.8176755447941889, + "grad_norm": 1.5743988439510799, + "learning_rate": 8.467037219259178e-07, + "loss": 0.3689, + "step": 37147 + }, + { + "epoch": 0.8176975566806075, + "grad_norm": 1.510477331727244, + "learning_rate": 8.465052643030741e-07, + "loss": 0.299, + "step": 37148 + }, + { + "epoch": 0.8177195685670262, + "grad_norm": 1.4092136229517522, + "learning_rate": 8.463068277901432e-07, + "loss": 0.2268, + "step": 37149 + }, + { + "epoch": 0.8177415804534448, + "grad_norm": 1.6248289668611806, + "learning_rate": 8.461084123881325e-07, + "loss": 0.323, + "step": 37150 + }, + { + "epoch": 0.8177635923398635, + "grad_norm": 0.6147197722581835, + "learning_rate": 8.459100180980512e-07, + "loss": 0.4473, + "step": 37151 + }, + { + "epoch": 0.8177856042262822, + "grad_norm": 0.5921482136897291, + "learning_rate": 8.45711644920908e-07, + "loss": 0.4371, + "step": 37152 + }, + { + "epoch": 0.8178076161127008, + "grad_norm": 0.616349782024089, + "learning_rate": 8.455132928577126e-07, + "loss": 0.4333, + "step": 37153 + }, + { + "epoch": 0.8178296279991195, + "grad_norm": 0.6161129193497487, + "learning_rate": 8.453149619094719e-07, + "loss": 0.4523, + "step": 37154 + }, + { + "epoch": 0.8178516398855382, + "grad_norm": 13.332113919750444, + "learning_rate": 8.451166520771919e-07, + "loss": 0.3017, + "step": 37155 + }, + { + "epoch": 0.8178736517719568, + "grad_norm": 1.7311619660493898, + "learning_rate": 8.449183633618824e-07, + "loss": 0.2763, + "step": 37156 + }, + { + "epoch": 0.8178956636583755, + "grad_norm": 2.7051491325756523, + "learning_rate": 8.447200957645508e-07, + "loss": 0.3037, + "step": 37157 + }, + { + "epoch": 0.8179176755447942, + "grad_norm": 3.2606913404543016, + "learning_rate": 8.445218492862062e-07, + "loss": 0.2845, + "step": 37158 + }, + { + "epoch": 0.8179396874312128, + "grad_norm": 1.537115474228126, + "learning_rate": 8.443236239278552e-07, + "loss": 0.3889, + "step": 37159 + }, + { + "epoch": 0.8179616993176315, + "grad_norm": 1.453244464553695, + "learning_rate": 8.441254196905036e-07, + "loss": 0.3142, + "step": 37160 + }, + { + "epoch": 0.8179837112040502, + "grad_norm": 0.6086817841036656, + "learning_rate": 8.439272365751599e-07, + "loss": 0.4598, + "step": 37161 + }, + { + "epoch": 0.8180057230904688, + "grad_norm": 1.2693501520014197, + "learning_rate": 8.437290745828319e-07, + "loss": 0.1631, + "step": 37162 + }, + { + "epoch": 0.8180277349768875, + "grad_norm": 1.1633038283289208, + "learning_rate": 8.435309337145281e-07, + "loss": 0.1904, + "step": 37163 + }, + { + "epoch": 0.8180497468633062, + "grad_norm": 1.7681509726697506, + "learning_rate": 8.433328139712538e-07, + "loss": 0.2583, + "step": 37164 + }, + { + "epoch": 0.8180717587497248, + "grad_norm": 1.6425516412459746, + "learning_rate": 8.431347153540143e-07, + "loss": 0.2076, + "step": 37165 + }, + { + "epoch": 0.8180937706361435, + "grad_norm": 1.6345241274258875, + "learning_rate": 8.429366378638187e-07, + "loss": 0.2601, + "step": 37166 + }, + { + "epoch": 0.8181157825225622, + "grad_norm": 1.6773273667482844, + "learning_rate": 8.427385815016742e-07, + "loss": 0.2886, + "step": 37167 + }, + { + "epoch": 0.8181377944089808, + "grad_norm": 1.398724862888236, + "learning_rate": 8.425405462685854e-07, + "loss": 0.2477, + "step": 37168 + }, + { + "epoch": 0.8181598062953995, + "grad_norm": 5.404765333218402, + "learning_rate": 8.423425321655609e-07, + "loss": 0.3427, + "step": 37169 + }, + { + "epoch": 0.8181818181818182, + "grad_norm": 1.758018984144711, + "learning_rate": 8.421445391936051e-07, + "loss": 0.3392, + "step": 37170 + }, + { + "epoch": 0.8182038300682368, + "grad_norm": 1.7642762080271277, + "learning_rate": 8.419465673537248e-07, + "loss": 0.2793, + "step": 37171 + }, + { + "epoch": 0.8182258419546555, + "grad_norm": 0.6057083289523572, + "learning_rate": 8.41748616646928e-07, + "loss": 0.4394, + "step": 37172 + }, + { + "epoch": 0.8182478538410742, + "grad_norm": 0.6047744768947773, + "learning_rate": 8.415506870742179e-07, + "loss": 0.4354, + "step": 37173 + }, + { + "epoch": 0.8182698657274928, + "grad_norm": 1.5144075673253934, + "learning_rate": 8.413527786366033e-07, + "loss": 0.2756, + "step": 37174 + }, + { + "epoch": 0.8182918776139115, + "grad_norm": 1.5910967932498405, + "learning_rate": 8.411548913350875e-07, + "loss": 0.2349, + "step": 37175 + }, + { + "epoch": 0.8183138895003301, + "grad_norm": 1.8507886254511225, + "learning_rate": 8.409570251706778e-07, + "loss": 0.1933, + "step": 37176 + }, + { + "epoch": 0.8183359013867488, + "grad_norm": 2.443118054856528, + "learning_rate": 8.407591801443804e-07, + "loss": 0.4196, + "step": 37177 + }, + { + "epoch": 0.8183579132731675, + "grad_norm": 1.7996071796529012, + "learning_rate": 8.405613562571984e-07, + "loss": 0.2988, + "step": 37178 + }, + { + "epoch": 0.8183799251595861, + "grad_norm": 1.2553782536946354, + "learning_rate": 8.403635535101401e-07, + "loss": 0.2332, + "step": 37179 + }, + { + "epoch": 0.8184019370460048, + "grad_norm": 1.4303084910704877, + "learning_rate": 8.401657719042084e-07, + "loss": 0.2107, + "step": 37180 + }, + { + "epoch": 0.8184239489324235, + "grad_norm": 2.1798499189021214, + "learning_rate": 8.399680114404108e-07, + "loss": 0.2534, + "step": 37181 + }, + { + "epoch": 0.8184459608188421, + "grad_norm": 3.14866846769759, + "learning_rate": 8.397702721197499e-07, + "loss": 0.3157, + "step": 37182 + }, + { + "epoch": 0.8184679727052608, + "grad_norm": 0.5914849931591726, + "learning_rate": 8.39572553943232e-07, + "loss": 0.4324, + "step": 37183 + }, + { + "epoch": 0.8184899845916795, + "grad_norm": 0.6173598990461835, + "learning_rate": 8.393748569118637e-07, + "loss": 0.4329, + "step": 37184 + }, + { + "epoch": 0.8185119964780981, + "grad_norm": 4.920664183059786, + "learning_rate": 8.391771810266475e-07, + "loss": 0.2025, + "step": 37185 + }, + { + "epoch": 0.8185340083645168, + "grad_norm": 0.6145217152806279, + "learning_rate": 8.389795262885875e-07, + "loss": 0.4457, + "step": 37186 + }, + { + "epoch": 0.8185560202509355, + "grad_norm": 1.7185907946025705, + "learning_rate": 8.387818926986896e-07, + "loss": 0.2657, + "step": 37187 + }, + { + "epoch": 0.8185780321373541, + "grad_norm": 1.6354656203445994, + "learning_rate": 8.385842802579575e-07, + "loss": 0.2447, + "step": 37188 + }, + { + "epoch": 0.8186000440237728, + "grad_norm": 73.909174118565, + "learning_rate": 8.38386688967398e-07, + "loss": 0.2915, + "step": 37189 + }, + { + "epoch": 0.8186220559101915, + "grad_norm": 2.216809005175548, + "learning_rate": 8.381891188280133e-07, + "loss": 0.2749, + "step": 37190 + }, + { + "epoch": 0.8186440677966101, + "grad_norm": 2.0746787113815737, + "learning_rate": 8.379915698408064e-07, + "loss": 0.2327, + "step": 37191 + }, + { + "epoch": 0.8186660796830288, + "grad_norm": 1.9670318668660274, + "learning_rate": 8.377940420067827e-07, + "loss": 0.3647, + "step": 37192 + }, + { + "epoch": 0.8186880915694476, + "grad_norm": 1.5762070909598676, + "learning_rate": 8.375965353269461e-07, + "loss": 0.2822, + "step": 37193 + }, + { + "epoch": 0.8187101034558661, + "grad_norm": 2.7135388361990254, + "learning_rate": 8.373990498023016e-07, + "loss": 0.2446, + "step": 37194 + }, + { + "epoch": 0.8187321153422848, + "grad_norm": 1.6868279747423751, + "learning_rate": 8.372015854338517e-07, + "loss": 0.2468, + "step": 37195 + }, + { + "epoch": 0.8187541272287036, + "grad_norm": 1.777781865326977, + "learning_rate": 8.370041422225989e-07, + "loss": 0.2387, + "step": 37196 + }, + { + "epoch": 0.8187761391151221, + "grad_norm": 1.73033804425369, + "learning_rate": 8.368067201695473e-07, + "loss": 0.2945, + "step": 37197 + }, + { + "epoch": 0.8187981510015409, + "grad_norm": 1.830090798278275, + "learning_rate": 8.366093192757013e-07, + "loss": 0.3511, + "step": 37198 + }, + { + "epoch": 0.8188201628879594, + "grad_norm": 1.6192368167343865, + "learning_rate": 8.36411939542065e-07, + "loss": 0.2982, + "step": 37199 + }, + { + "epoch": 0.8188421747743782, + "grad_norm": 2.742590058548084, + "learning_rate": 8.362145809696398e-07, + "loss": 0.3215, + "step": 37200 + }, + { + "epoch": 0.8188641866607969, + "grad_norm": 2.0461213728251835, + "learning_rate": 8.360172435594283e-07, + "loss": 0.28, + "step": 37201 + }, + { + "epoch": 0.8188861985472154, + "grad_norm": 1.7390928062851956, + "learning_rate": 8.35819927312434e-07, + "loss": 0.3239, + "step": 37202 + }, + { + "epoch": 0.8189082104336342, + "grad_norm": 3.4079679747066516, + "learning_rate": 8.356226322296618e-07, + "loss": 0.3247, + "step": 37203 + }, + { + "epoch": 0.8189302223200529, + "grad_norm": 1.4971182731823562, + "learning_rate": 8.354253583121108e-07, + "loss": 0.4036, + "step": 37204 + }, + { + "epoch": 0.8189522342064715, + "grad_norm": 0.6158274427603313, + "learning_rate": 8.352281055607874e-07, + "loss": 0.4496, + "step": 37205 + }, + { + "epoch": 0.8189742460928902, + "grad_norm": 0.6315583253997813, + "learning_rate": 8.350308739766911e-07, + "loss": 0.4428, + "step": 37206 + }, + { + "epoch": 0.8189962579793089, + "grad_norm": 1.2422758271860752, + "learning_rate": 8.34833663560825e-07, + "loss": 0.3293, + "step": 37207 + }, + { + "epoch": 0.8190182698657275, + "grad_norm": 2.055450030947583, + "learning_rate": 8.346364743141938e-07, + "loss": 0.2674, + "step": 37208 + }, + { + "epoch": 0.8190402817521462, + "grad_norm": 1.464745569661744, + "learning_rate": 8.344393062377964e-07, + "loss": 0.2417, + "step": 37209 + }, + { + "epoch": 0.8190622936385649, + "grad_norm": 1.9523223949673925, + "learning_rate": 8.342421593326377e-07, + "loss": 0.2445, + "step": 37210 + }, + { + "epoch": 0.8190843055249835, + "grad_norm": 2.4770678192885036, + "learning_rate": 8.340450335997169e-07, + "loss": 0.3698, + "step": 37211 + }, + { + "epoch": 0.8191063174114022, + "grad_norm": 1.7503370439517976, + "learning_rate": 8.33847929040037e-07, + "loss": 0.2093, + "step": 37212 + }, + { + "epoch": 0.8191283292978209, + "grad_norm": 1.5034219737733343, + "learning_rate": 8.336508456546016e-07, + "loss": 0.3615, + "step": 37213 + }, + { + "epoch": 0.8191503411842395, + "grad_norm": 1.737428714941103, + "learning_rate": 8.334537834444101e-07, + "loss": 0.2905, + "step": 37214 + }, + { + "epoch": 0.8191723530706582, + "grad_norm": 4.934730895192252, + "learning_rate": 8.332567424104654e-07, + "loss": 0.3008, + "step": 37215 + }, + { + "epoch": 0.8191943649570769, + "grad_norm": 1.7093591777564316, + "learning_rate": 8.330597225537673e-07, + "loss": 0.2506, + "step": 37216 + }, + { + "epoch": 0.8192163768434955, + "grad_norm": 1.6362375742607707, + "learning_rate": 8.328627238753195e-07, + "loss": 0.3013, + "step": 37217 + }, + { + "epoch": 0.8192383887299142, + "grad_norm": 4.770796301000673, + "learning_rate": 8.326657463761207e-07, + "loss": 0.3774, + "step": 37218 + }, + { + "epoch": 0.8192604006163329, + "grad_norm": 1.4418598152385151, + "learning_rate": 8.324687900571732e-07, + "loss": 0.2763, + "step": 37219 + }, + { + "epoch": 0.8192824125027515, + "grad_norm": 1.5395346061083137, + "learning_rate": 8.322718549194792e-07, + "loss": 0.3676, + "step": 37220 + }, + { + "epoch": 0.8193044243891702, + "grad_norm": 0.626783928723763, + "learning_rate": 8.320749409640372e-07, + "loss": 0.4404, + "step": 37221 + }, + { + "epoch": 0.8193264362755888, + "grad_norm": 0.6146386619764829, + "learning_rate": 8.318780481918509e-07, + "loss": 0.4558, + "step": 37222 + }, + { + "epoch": 0.8193484481620075, + "grad_norm": 1.4779273912044015, + "learning_rate": 8.316811766039179e-07, + "loss": 0.323, + "step": 37223 + }, + { + "epoch": 0.8193704600484262, + "grad_norm": 1.541684093953885, + "learning_rate": 8.314843262012401e-07, + "loss": 0.3446, + "step": 37224 + }, + { + "epoch": 0.8193924719348448, + "grad_norm": 0.6317528275008292, + "learning_rate": 8.312874969848201e-07, + "loss": 0.446, + "step": 37225 + }, + { + "epoch": 0.8194144838212635, + "grad_norm": 3.1794789309047404, + "learning_rate": 8.310906889556558e-07, + "loss": 0.3159, + "step": 37226 + }, + { + "epoch": 0.8194364957076822, + "grad_norm": 18.41967218600491, + "learning_rate": 8.308939021147467e-07, + "loss": 0.3664, + "step": 37227 + }, + { + "epoch": 0.8194585075941008, + "grad_norm": 2.037818390885902, + "learning_rate": 8.306971364630944e-07, + "loss": 0.2685, + "step": 37228 + }, + { + "epoch": 0.8194805194805195, + "grad_norm": 1.6518466098929208, + "learning_rate": 8.305003920016991e-07, + "loss": 0.2752, + "step": 37229 + }, + { + "epoch": 0.8195025313669382, + "grad_norm": 1.7836693663859673, + "learning_rate": 8.303036687315613e-07, + "loss": 0.342, + "step": 37230 + }, + { + "epoch": 0.8195245432533568, + "grad_norm": 31.790408409933345, + "learning_rate": 8.301069666536799e-07, + "loss": 0.2881, + "step": 37231 + }, + { + "epoch": 0.8195465551397755, + "grad_norm": 2.09171628686673, + "learning_rate": 8.299102857690538e-07, + "loss": 0.2407, + "step": 37232 + }, + { + "epoch": 0.8195685670261942, + "grad_norm": 1.337026546670774, + "learning_rate": 8.297136260786837e-07, + "loss": 0.2877, + "step": 37233 + }, + { + "epoch": 0.8195905789126128, + "grad_norm": 1.4534305664644167, + "learning_rate": 8.295169875835685e-07, + "loss": 0.2461, + "step": 37234 + }, + { + "epoch": 0.8196125907990315, + "grad_norm": 4.9291179047509734, + "learning_rate": 8.293203702847091e-07, + "loss": 0.3017, + "step": 37235 + }, + { + "epoch": 0.8196346026854502, + "grad_norm": 0.5933811743259305, + "learning_rate": 8.291237741831043e-07, + "loss": 0.4181, + "step": 37236 + }, + { + "epoch": 0.8196566145718688, + "grad_norm": 1.2440383690601595, + "learning_rate": 8.289271992797515e-07, + "loss": 0.3071, + "step": 37237 + }, + { + "epoch": 0.8196786264582875, + "grad_norm": 1.5276453838667616, + "learning_rate": 8.287306455756506e-07, + "loss": 0.365, + "step": 37238 + }, + { + "epoch": 0.8197006383447062, + "grad_norm": 2.062681209626306, + "learning_rate": 8.285341130718017e-07, + "loss": 0.317, + "step": 37239 + }, + { + "epoch": 0.8197226502311248, + "grad_norm": 2.1182513089057817, + "learning_rate": 8.283376017692035e-07, + "loss": 0.2857, + "step": 37240 + }, + { + "epoch": 0.8197446621175435, + "grad_norm": 2.1386458410573947, + "learning_rate": 8.281411116688543e-07, + "loss": 0.3443, + "step": 37241 + }, + { + "epoch": 0.8197666740039622, + "grad_norm": 1.5621634590282225, + "learning_rate": 8.279446427717519e-07, + "loss": 0.2797, + "step": 37242 + }, + { + "epoch": 0.8197886858903808, + "grad_norm": 0.6106783153696406, + "learning_rate": 8.277481950788951e-07, + "loss": 0.4367, + "step": 37243 + }, + { + "epoch": 0.8198106977767995, + "grad_norm": 1.313112597545889, + "learning_rate": 8.275517685912842e-07, + "loss": 0.2058, + "step": 37244 + }, + { + "epoch": 0.8198327096632181, + "grad_norm": 1.2910537039672132, + "learning_rate": 8.273553633099151e-07, + "loss": 0.2797, + "step": 37245 + }, + { + "epoch": 0.8198547215496368, + "grad_norm": 1.4087793325970275, + "learning_rate": 8.271589792357887e-07, + "loss": 0.3314, + "step": 37246 + }, + { + "epoch": 0.8198767334360555, + "grad_norm": 1.617538243749241, + "learning_rate": 8.269626163698996e-07, + "loss": 0.2084, + "step": 37247 + }, + { + "epoch": 0.8198987453224741, + "grad_norm": 2.2228258203267766, + "learning_rate": 8.267662747132482e-07, + "loss": 0.3121, + "step": 37248 + }, + { + "epoch": 0.8199207572088928, + "grad_norm": 1.5205731614374987, + "learning_rate": 8.265699542668327e-07, + "loss": 0.3311, + "step": 37249 + }, + { + "epoch": 0.8199427690953115, + "grad_norm": 0.6153823114468535, + "learning_rate": 8.263736550316493e-07, + "loss": 0.4215, + "step": 37250 + }, + { + "epoch": 0.8199647809817301, + "grad_norm": 1.7037525779561886, + "learning_rate": 8.261773770086972e-07, + "loss": 0.2877, + "step": 37251 + }, + { + "epoch": 0.8199867928681488, + "grad_norm": 2.4176208200794704, + "learning_rate": 8.25981120198972e-07, + "loss": 0.179, + "step": 37252 + }, + { + "epoch": 0.8200088047545675, + "grad_norm": 2.3896983457814986, + "learning_rate": 8.257848846034743e-07, + "loss": 0.375, + "step": 37253 + }, + { + "epoch": 0.8200308166409861, + "grad_norm": 1.57577593068926, + "learning_rate": 8.25588670223198e-07, + "loss": 0.3129, + "step": 37254 + }, + { + "epoch": 0.8200528285274048, + "grad_norm": 2.2335188842289058, + "learning_rate": 8.253924770591421e-07, + "loss": 0.3288, + "step": 37255 + }, + { + "epoch": 0.8200748404138235, + "grad_norm": 1.414576232544821, + "learning_rate": 8.251963051123041e-07, + "loss": 0.2101, + "step": 37256 + }, + { + "epoch": 0.8200968523002421, + "grad_norm": 1.7758191042906701, + "learning_rate": 8.2500015438368e-07, + "loss": 0.3391, + "step": 37257 + }, + { + "epoch": 0.8201188641866608, + "grad_norm": 1.7936120971648353, + "learning_rate": 8.248040248742678e-07, + "loss": 0.3378, + "step": 37258 + }, + { + "epoch": 0.8201408760730795, + "grad_norm": 1.1289382960066658, + "learning_rate": 8.246079165850628e-07, + "loss": 0.2271, + "step": 37259 + }, + { + "epoch": 0.8201628879594981, + "grad_norm": 1.384348559965752, + "learning_rate": 8.244118295170633e-07, + "loss": 0.2075, + "step": 37260 + }, + { + "epoch": 0.8201848998459168, + "grad_norm": 3.321541621352103, + "learning_rate": 8.242157636712656e-07, + "loss": 0.2263, + "step": 37261 + }, + { + "epoch": 0.8202069117323355, + "grad_norm": 1.8212777738236603, + "learning_rate": 8.240197190486648e-07, + "loss": 0.3682, + "step": 37262 + }, + { + "epoch": 0.8202289236187541, + "grad_norm": 0.6030098969145815, + "learning_rate": 8.238236956502599e-07, + "loss": 0.4283, + "step": 37263 + }, + { + "epoch": 0.8202509355051728, + "grad_norm": 1.471118285249392, + "learning_rate": 8.236276934770443e-07, + "loss": 0.2572, + "step": 37264 + }, + { + "epoch": 0.8202729473915915, + "grad_norm": 2.54594898446673, + "learning_rate": 8.234317125300151e-07, + "loss": 0.2786, + "step": 37265 + }, + { + "epoch": 0.8202949592780101, + "grad_norm": 1.5141838968802677, + "learning_rate": 8.232357528101703e-07, + "loss": 0.2522, + "step": 37266 + }, + { + "epoch": 0.8203169711644288, + "grad_norm": 2.4501495537202778, + "learning_rate": 8.230398143185042e-07, + "loss": 0.3135, + "step": 37267 + }, + { + "epoch": 0.8203389830508474, + "grad_norm": 1.780110711458124, + "learning_rate": 8.228438970560115e-07, + "loss": 0.3276, + "step": 37268 + }, + { + "epoch": 0.8203609949372661, + "grad_norm": 1.9362031286647239, + "learning_rate": 8.226480010236892e-07, + "loss": 0.1817, + "step": 37269 + }, + { + "epoch": 0.8203830068236848, + "grad_norm": 1.397760312874497, + "learning_rate": 8.224521262225326e-07, + "loss": 0.1912, + "step": 37270 + }, + { + "epoch": 0.8204050187101034, + "grad_norm": 2.5890757137739726, + "learning_rate": 8.222562726535393e-07, + "loss": 0.2614, + "step": 37271 + }, + { + "epoch": 0.8204270305965221, + "grad_norm": 2.252358314717844, + "learning_rate": 8.220604403177029e-07, + "loss": 0.2673, + "step": 37272 + }, + { + "epoch": 0.8204490424829408, + "grad_norm": 1.5650897100792363, + "learning_rate": 8.21864629216017e-07, + "loss": 0.2181, + "step": 37273 + }, + { + "epoch": 0.8204710543693594, + "grad_norm": 1.6221655434186757, + "learning_rate": 8.216688393494793e-07, + "loss": 0.2842, + "step": 37274 + }, + { + "epoch": 0.8204930662557781, + "grad_norm": 2.7301352738889393, + "learning_rate": 8.214730707190832e-07, + "loss": 0.3244, + "step": 37275 + }, + { + "epoch": 0.8205150781421968, + "grad_norm": 1.6902449386728748, + "learning_rate": 8.212773233258265e-07, + "loss": 0.3487, + "step": 37276 + }, + { + "epoch": 0.8205370900286154, + "grad_norm": 1.5312514796953538, + "learning_rate": 8.21081597170702e-07, + "loss": 0.2839, + "step": 37277 + }, + { + "epoch": 0.8205591019150341, + "grad_norm": 0.6232794808152019, + "learning_rate": 8.208858922547036e-07, + "loss": 0.4389, + "step": 37278 + }, + { + "epoch": 0.8205811138014528, + "grad_norm": 2.344414147142306, + "learning_rate": 8.20690208578827e-07, + "loss": 0.2559, + "step": 37279 + }, + { + "epoch": 0.8206031256878714, + "grad_norm": 1.3728860866277588, + "learning_rate": 8.204945461440667e-07, + "loss": 0.2247, + "step": 37280 + }, + { + "epoch": 0.8206251375742901, + "grad_norm": 4.0556861206802495, + "learning_rate": 8.202989049514187e-07, + "loss": 0.2832, + "step": 37281 + }, + { + "epoch": 0.8206471494607088, + "grad_norm": 2.231064587198985, + "learning_rate": 8.201032850018759e-07, + "loss": 0.2288, + "step": 37282 + }, + { + "epoch": 0.8206691613471274, + "grad_norm": 1.935098326669811, + "learning_rate": 8.199076862964306e-07, + "loss": 0.3884, + "step": 37283 + }, + { + "epoch": 0.8206911732335461, + "grad_norm": 3.104811004510427, + "learning_rate": 8.197121088360793e-07, + "loss": 0.2852, + "step": 37284 + }, + { + "epoch": 0.8207131851199648, + "grad_norm": 1.7466975785843044, + "learning_rate": 8.195165526218168e-07, + "loss": 0.3252, + "step": 37285 + }, + { + "epoch": 0.8207351970063834, + "grad_norm": 1.3714128105553862, + "learning_rate": 8.193210176546345e-07, + "loss": 0.3138, + "step": 37286 + }, + { + "epoch": 0.8207572088928021, + "grad_norm": 1.7267088751889745, + "learning_rate": 8.191255039355284e-07, + "loss": 0.358, + "step": 37287 + }, + { + "epoch": 0.8207792207792208, + "grad_norm": 0.6303183137085887, + "learning_rate": 8.189300114654903e-07, + "loss": 0.4491, + "step": 37288 + }, + { + "epoch": 0.8208012326656394, + "grad_norm": 1.511033093491677, + "learning_rate": 8.187345402455155e-07, + "loss": 0.2936, + "step": 37289 + }, + { + "epoch": 0.8208232445520581, + "grad_norm": 1.7266937162436438, + "learning_rate": 8.185390902765961e-07, + "loss": 0.3156, + "step": 37290 + }, + { + "epoch": 0.8208452564384767, + "grad_norm": 2.1670615590051896, + "learning_rate": 8.183436615597251e-07, + "loss": 0.3001, + "step": 37291 + }, + { + "epoch": 0.8208672683248954, + "grad_norm": 2.787827776685618, + "learning_rate": 8.181482540958985e-07, + "loss": 0.2648, + "step": 37292 + }, + { + "epoch": 0.8208892802113141, + "grad_norm": 1.9793169658498146, + "learning_rate": 8.179528678861065e-07, + "loss": 0.285, + "step": 37293 + }, + { + "epoch": 0.8209112920977327, + "grad_norm": 1.5609708927732933, + "learning_rate": 8.177575029313445e-07, + "loss": 0.2086, + "step": 37294 + }, + { + "epoch": 0.8209333039841514, + "grad_norm": 0.6268064710461085, + "learning_rate": 8.175621592326027e-07, + "loss": 0.4603, + "step": 37295 + }, + { + "epoch": 0.8209553158705701, + "grad_norm": 0.6065985617529155, + "learning_rate": 8.173668367908755e-07, + "loss": 0.4501, + "step": 37296 + }, + { + "epoch": 0.8209773277569887, + "grad_norm": 1.6279199099478028, + "learning_rate": 8.171715356071569e-07, + "loss": 0.3263, + "step": 37297 + }, + { + "epoch": 0.8209993396434074, + "grad_norm": 1.6684684564568493, + "learning_rate": 8.169762556824368e-07, + "loss": 0.2797, + "step": 37298 + }, + { + "epoch": 0.8210213515298261, + "grad_norm": 0.6145006488107918, + "learning_rate": 8.167809970177105e-07, + "loss": 0.461, + "step": 37299 + }, + { + "epoch": 0.8210433634162447, + "grad_norm": 1.7546904081375778, + "learning_rate": 8.16585759613967e-07, + "loss": 0.2587, + "step": 37300 + }, + { + "epoch": 0.8210653753026634, + "grad_norm": 1.3364440496423582, + "learning_rate": 8.163905434722014e-07, + "loss": 0.2016, + "step": 37301 + }, + { + "epoch": 0.8210873871890821, + "grad_norm": 1.4777122727559577, + "learning_rate": 8.161953485934054e-07, + "loss": 0.2645, + "step": 37302 + }, + { + "epoch": 0.8211093990755007, + "grad_norm": 3.1810619672246094, + "learning_rate": 8.160001749785712e-07, + "loss": 0.3298, + "step": 37303 + }, + { + "epoch": 0.8211314109619194, + "grad_norm": 1.327698024739493, + "learning_rate": 8.15805022628689e-07, + "loss": 0.2325, + "step": 37304 + }, + { + "epoch": 0.8211534228483381, + "grad_norm": 3.4730867945572417, + "learning_rate": 8.156098915447519e-07, + "loss": 0.223, + "step": 37305 + }, + { + "epoch": 0.8211754347347567, + "grad_norm": 1.492281149901444, + "learning_rate": 8.154147817277513e-07, + "loss": 0.3177, + "step": 37306 + }, + { + "epoch": 0.8211974466211754, + "grad_norm": 3.157290870700994, + "learning_rate": 8.15219693178681e-07, + "loss": 0.2548, + "step": 37307 + }, + { + "epoch": 0.8212194585075941, + "grad_norm": 1.6418615791249693, + "learning_rate": 8.1502462589853e-07, + "loss": 0.3237, + "step": 37308 + }, + { + "epoch": 0.8212414703940127, + "grad_norm": 1.4790759018690556, + "learning_rate": 8.148295798882893e-07, + "loss": 0.2729, + "step": 37309 + }, + { + "epoch": 0.8212634822804314, + "grad_norm": 0.6279328955093324, + "learning_rate": 8.146345551489515e-07, + "loss": 0.4306, + "step": 37310 + }, + { + "epoch": 0.8212854941668501, + "grad_norm": 0.6203705467550492, + "learning_rate": 8.144395516815079e-07, + "loss": 0.4295, + "step": 37311 + }, + { + "epoch": 0.8213075060532687, + "grad_norm": 1.4602022481103072, + "learning_rate": 8.1424456948695e-07, + "loss": 0.3249, + "step": 37312 + }, + { + "epoch": 0.8213295179396874, + "grad_norm": 1.5216399771912996, + "learning_rate": 8.140496085662685e-07, + "loss": 0.3252, + "step": 37313 + }, + { + "epoch": 0.8213515298261062, + "grad_norm": 0.6295981452778423, + "learning_rate": 8.138546689204524e-07, + "loss": 0.4529, + "step": 37314 + }, + { + "epoch": 0.8213735417125247, + "grad_norm": 1.7424584498016569, + "learning_rate": 8.13659750550494e-07, + "loss": 0.2835, + "step": 37315 + }, + { + "epoch": 0.8213955535989435, + "grad_norm": 1.382826322783436, + "learning_rate": 8.134648534573841e-07, + "loss": 0.2685, + "step": 37316 + }, + { + "epoch": 0.821417565485362, + "grad_norm": 3.9648988389150004, + "learning_rate": 8.132699776421143e-07, + "loss": 0.3019, + "step": 37317 + }, + { + "epoch": 0.8214395773717807, + "grad_norm": 5.002314531412131, + "learning_rate": 8.130751231056733e-07, + "loss": 0.3588, + "step": 37318 + }, + { + "epoch": 0.8214615892581995, + "grad_norm": 4.983551531627822, + "learning_rate": 8.128802898490512e-07, + "loss": 0.2377, + "step": 37319 + }, + { + "epoch": 0.821483601144618, + "grad_norm": 3.086002222409336, + "learning_rate": 8.12685477873239e-07, + "loss": 0.2733, + "step": 37320 + }, + { + "epoch": 0.8215056130310368, + "grad_norm": 0.6547987511720281, + "learning_rate": 8.124906871792265e-07, + "loss": 0.476, + "step": 37321 + }, + { + "epoch": 0.8215276249174555, + "grad_norm": 2.063186769852793, + "learning_rate": 8.122959177680051e-07, + "loss": 0.3521, + "step": 37322 + }, + { + "epoch": 0.821549636803874, + "grad_norm": 1.3224027638390585, + "learning_rate": 8.121011696405639e-07, + "loss": 0.259, + "step": 37323 + }, + { + "epoch": 0.8215716486902928, + "grad_norm": 1.645386066862371, + "learning_rate": 8.119064427978913e-07, + "loss": 0.3027, + "step": 37324 + }, + { + "epoch": 0.8215936605767115, + "grad_norm": 2.060415222887689, + "learning_rate": 8.117117372409794e-07, + "loss": 0.2414, + "step": 37325 + }, + { + "epoch": 0.82161567246313, + "grad_norm": 3.104730732537821, + "learning_rate": 8.11517052970815e-07, + "loss": 0.2763, + "step": 37326 + }, + { + "epoch": 0.8216376843495488, + "grad_norm": 1.2692759313840485, + "learning_rate": 8.113223899883887e-07, + "loss": 0.3393, + "step": 37327 + }, + { + "epoch": 0.8216596962359675, + "grad_norm": 0.6044300008721701, + "learning_rate": 8.111277482946922e-07, + "loss": 0.4189, + "step": 37328 + }, + { + "epoch": 0.8216817081223861, + "grad_norm": 1.9434886488176863, + "learning_rate": 8.109331278907112e-07, + "loss": 0.3396, + "step": 37329 + }, + { + "epoch": 0.8217037200088048, + "grad_norm": 7.806479175750887, + "learning_rate": 8.107385287774371e-07, + "loss": 0.3627, + "step": 37330 + }, + { + "epoch": 0.8217257318952235, + "grad_norm": 2.176930054855303, + "learning_rate": 8.105439509558577e-07, + "loss": 0.2959, + "step": 37331 + }, + { + "epoch": 0.8217477437816421, + "grad_norm": 2.853325759536767, + "learning_rate": 8.103493944269624e-07, + "loss": 0.3362, + "step": 37332 + }, + { + "epoch": 0.8217697556680608, + "grad_norm": 1.4650768548592397, + "learning_rate": 8.101548591917413e-07, + "loss": 0.2802, + "step": 37333 + }, + { + "epoch": 0.8217917675544795, + "grad_norm": 7.286569506365342, + "learning_rate": 8.099603452511806e-07, + "loss": 0.3098, + "step": 37334 + }, + { + "epoch": 0.8218137794408981, + "grad_norm": 0.6153580700962658, + "learning_rate": 8.097658526062718e-07, + "loss": 0.439, + "step": 37335 + }, + { + "epoch": 0.8218357913273168, + "grad_norm": 1.539260550964277, + "learning_rate": 8.095713812580003e-07, + "loss": 0.3049, + "step": 37336 + }, + { + "epoch": 0.8218578032137355, + "grad_norm": 1.8662190324694297, + "learning_rate": 8.093769312073563e-07, + "loss": 0.3178, + "step": 37337 + }, + { + "epoch": 0.8218798151001541, + "grad_norm": 0.6127148503754376, + "learning_rate": 8.091825024553285e-07, + "loss": 0.4399, + "step": 37338 + }, + { + "epoch": 0.8219018269865728, + "grad_norm": 2.6875457876539994, + "learning_rate": 8.089880950029033e-07, + "loss": 0.2501, + "step": 37339 + }, + { + "epoch": 0.8219238388729914, + "grad_norm": 2.7792981039656315, + "learning_rate": 8.087937088510716e-07, + "loss": 0.3394, + "step": 37340 + }, + { + "epoch": 0.8219458507594101, + "grad_norm": 2.3661788264744037, + "learning_rate": 8.085993440008177e-07, + "loss": 0.3365, + "step": 37341 + }, + { + "epoch": 0.8219678626458288, + "grad_norm": 1.8566639726133547, + "learning_rate": 8.084050004531319e-07, + "loss": 0.3445, + "step": 37342 + }, + { + "epoch": 0.8219898745322474, + "grad_norm": 0.6509552813536861, + "learning_rate": 8.082106782090021e-07, + "loss": 0.463, + "step": 37343 + }, + { + "epoch": 0.8220118864186661, + "grad_norm": 2.4698104136936925, + "learning_rate": 8.080163772694155e-07, + "loss": 0.349, + "step": 37344 + }, + { + "epoch": 0.8220338983050848, + "grad_norm": 1.7559960944465438, + "learning_rate": 8.078220976353585e-07, + "loss": 0.3556, + "step": 37345 + }, + { + "epoch": 0.8220559101915034, + "grad_norm": 2.1535905053445923, + "learning_rate": 8.076278393078185e-07, + "loss": 0.2152, + "step": 37346 + }, + { + "epoch": 0.8220779220779221, + "grad_norm": 2.343799508579798, + "learning_rate": 8.074336022877843e-07, + "loss": 0.3085, + "step": 37347 + }, + { + "epoch": 0.8220999339643408, + "grad_norm": 1.9867892708683528, + "learning_rate": 8.072393865762434e-07, + "loss": 0.2695, + "step": 37348 + }, + { + "epoch": 0.8221219458507594, + "grad_norm": 2.4796171302266927, + "learning_rate": 8.070451921741823e-07, + "loss": 0.2137, + "step": 37349 + }, + { + "epoch": 0.8221439577371781, + "grad_norm": 0.615885313644505, + "learning_rate": 8.068510190825862e-07, + "loss": 0.4475, + "step": 37350 + }, + { + "epoch": 0.8221659696235968, + "grad_norm": 0.61173549555078, + "learning_rate": 8.066568673024439e-07, + "loss": 0.4464, + "step": 37351 + }, + { + "epoch": 0.8221879815100154, + "grad_norm": 1.6631348866245228, + "learning_rate": 8.064627368347411e-07, + "loss": 0.2712, + "step": 37352 + }, + { + "epoch": 0.8222099933964341, + "grad_norm": 2.3149865336237077, + "learning_rate": 8.06268627680466e-07, + "loss": 0.2737, + "step": 37353 + }, + { + "epoch": 0.8222320052828528, + "grad_norm": 1.5459624244588632, + "learning_rate": 8.06074539840605e-07, + "loss": 0.2735, + "step": 37354 + }, + { + "epoch": 0.8222540171692714, + "grad_norm": 2.568467293771409, + "learning_rate": 8.058804733161418e-07, + "loss": 0.3822, + "step": 37355 + }, + { + "epoch": 0.8222760290556901, + "grad_norm": 1.7058693544358141, + "learning_rate": 8.056864281080662e-07, + "loss": 0.2895, + "step": 37356 + }, + { + "epoch": 0.8222980409421088, + "grad_norm": 1.8057381738774274, + "learning_rate": 8.054924042173607e-07, + "loss": 0.3608, + "step": 37357 + }, + { + "epoch": 0.8223200528285274, + "grad_norm": 0.6062485284774435, + "learning_rate": 8.052984016450154e-07, + "loss": 0.4468, + "step": 37358 + }, + { + "epoch": 0.8223420647149461, + "grad_norm": 1.77751436444279, + "learning_rate": 8.05104420392015e-07, + "loss": 0.2883, + "step": 37359 + }, + { + "epoch": 0.8223640766013648, + "grad_norm": 1.4462640261651036, + "learning_rate": 8.049104604593438e-07, + "loss": 0.2538, + "step": 37360 + }, + { + "epoch": 0.8223860884877834, + "grad_norm": 1.6871658878017375, + "learning_rate": 8.047165218479897e-07, + "loss": 0.3262, + "step": 37361 + }, + { + "epoch": 0.8224081003742021, + "grad_norm": 1.6751138417290616, + "learning_rate": 8.045226045589349e-07, + "loss": 0.32, + "step": 37362 + }, + { + "epoch": 0.8224301122606207, + "grad_norm": 1.4755292041417332, + "learning_rate": 8.0432870859317e-07, + "loss": 0.331, + "step": 37363 + }, + { + "epoch": 0.8224521241470394, + "grad_norm": 1.9080057240661012, + "learning_rate": 8.041348339516775e-07, + "loss": 0.3704, + "step": 37364 + }, + { + "epoch": 0.8224741360334581, + "grad_norm": 8.722587985096862, + "learning_rate": 8.039409806354426e-07, + "loss": 0.3948, + "step": 37365 + }, + { + "epoch": 0.8224961479198767, + "grad_norm": 1.61726236807115, + "learning_rate": 8.037471486454518e-07, + "loss": 0.2966, + "step": 37366 + }, + { + "epoch": 0.8225181598062954, + "grad_norm": 1.9430204960360313, + "learning_rate": 8.035533379826887e-07, + "loss": 0.2929, + "step": 37367 + }, + { + "epoch": 0.8225401716927141, + "grad_norm": 0.6087330687301281, + "learning_rate": 8.033595486481394e-07, + "loss": 0.4346, + "step": 37368 + }, + { + "epoch": 0.8225621835791327, + "grad_norm": 1.9297347186555533, + "learning_rate": 8.031657806427895e-07, + "loss": 0.3401, + "step": 37369 + }, + { + "epoch": 0.8225841954655514, + "grad_norm": 1.233465097853549, + "learning_rate": 8.029720339676217e-07, + "loss": 0.2661, + "step": 37370 + }, + { + "epoch": 0.8226062073519701, + "grad_norm": 1.4899424389315314, + "learning_rate": 8.027783086236235e-07, + "loss": 0.2478, + "step": 37371 + }, + { + "epoch": 0.8226282192383887, + "grad_norm": 1.5295805138372804, + "learning_rate": 8.025846046117763e-07, + "loss": 0.4137, + "step": 37372 + }, + { + "epoch": 0.8226502311248074, + "grad_norm": 1.5438381588625703, + "learning_rate": 8.023909219330666e-07, + "loss": 0.274, + "step": 37373 + }, + { + "epoch": 0.8226722430112261, + "grad_norm": 1.8814615898137133, + "learning_rate": 8.021972605884792e-07, + "loss": 0.3385, + "step": 37374 + }, + { + "epoch": 0.8226942548976447, + "grad_norm": 1.303117793246621, + "learning_rate": 8.020036205789972e-07, + "loss": 0.2481, + "step": 37375 + }, + { + "epoch": 0.8227162667840634, + "grad_norm": 2.7017079337491112, + "learning_rate": 8.018100019056058e-07, + "loss": 0.2639, + "step": 37376 + }, + { + "epoch": 0.8227382786704821, + "grad_norm": 2.3807275362214972, + "learning_rate": 8.01616404569287e-07, + "loss": 0.2929, + "step": 37377 + }, + { + "epoch": 0.8227602905569007, + "grad_norm": 1.5352388779599422, + "learning_rate": 8.014228285710263e-07, + "loss": 0.3008, + "step": 37378 + }, + { + "epoch": 0.8227823024433194, + "grad_norm": 1.3995118371167827, + "learning_rate": 8.012292739118088e-07, + "loss": 0.2933, + "step": 37379 + }, + { + "epoch": 0.8228043143297381, + "grad_norm": 1.6750503712435878, + "learning_rate": 8.010357405926156e-07, + "loss": 0.2344, + "step": 37380 + }, + { + "epoch": 0.8228263262161567, + "grad_norm": 2.4842481491732933, + "learning_rate": 8.008422286144324e-07, + "loss": 0.3469, + "step": 37381 + }, + { + "epoch": 0.8228483381025754, + "grad_norm": 0.6088147396625113, + "learning_rate": 8.006487379782413e-07, + "loss": 0.4276, + "step": 37382 + }, + { + "epoch": 0.8228703499889941, + "grad_norm": 1.6335813270634905, + "learning_rate": 8.004552686850259e-07, + "loss": 0.2874, + "step": 37383 + }, + { + "epoch": 0.8228923618754127, + "grad_norm": 2.3789964573839946, + "learning_rate": 8.00261820735771e-07, + "loss": 0.4013, + "step": 37384 + }, + { + "epoch": 0.8229143737618314, + "grad_norm": 0.5908198108377878, + "learning_rate": 8.000683941314591e-07, + "loss": 0.4279, + "step": 37385 + }, + { + "epoch": 0.82293638564825, + "grad_norm": 1.6035878734478126, + "learning_rate": 7.998749888730711e-07, + "loss": 0.2973, + "step": 37386 + }, + { + "epoch": 0.8229583975346687, + "grad_norm": 1.4159554776251042, + "learning_rate": 7.996816049615918e-07, + "loss": 0.2175, + "step": 37387 + }, + { + "epoch": 0.8229804094210874, + "grad_norm": 2.033571854793247, + "learning_rate": 7.994882423980044e-07, + "loss": 0.2918, + "step": 37388 + }, + { + "epoch": 0.823002421307506, + "grad_norm": 0.612174944242873, + "learning_rate": 7.992949011832917e-07, + "loss": 0.4172, + "step": 37389 + }, + { + "epoch": 0.8230244331939247, + "grad_norm": 1.3165611934918464, + "learning_rate": 7.991015813184367e-07, + "loss": 0.2441, + "step": 37390 + }, + { + "epoch": 0.8230464450803434, + "grad_norm": 1.4826538308867236, + "learning_rate": 7.989082828044192e-07, + "loss": 0.3429, + "step": 37391 + }, + { + "epoch": 0.823068456966762, + "grad_norm": 1.3683591528432388, + "learning_rate": 7.987150056422254e-07, + "loss": 0.2142, + "step": 37392 + }, + { + "epoch": 0.8230904688531807, + "grad_norm": 1.7198842652425905, + "learning_rate": 7.985217498328329e-07, + "loss": 0.3108, + "step": 37393 + }, + { + "epoch": 0.8231124807395994, + "grad_norm": 2.194676293722125, + "learning_rate": 7.983285153772297e-07, + "loss": 0.2841, + "step": 37394 + }, + { + "epoch": 0.823134492626018, + "grad_norm": 2.0201272631692917, + "learning_rate": 7.981353022763943e-07, + "loss": 0.328, + "step": 37395 + }, + { + "epoch": 0.8231565045124367, + "grad_norm": 2.1635212707130913, + "learning_rate": 7.979421105313084e-07, + "loss": 0.4214, + "step": 37396 + }, + { + "epoch": 0.8231785163988554, + "grad_norm": 1.976381497796686, + "learning_rate": 7.977489401429562e-07, + "loss": 0.3112, + "step": 37397 + }, + { + "epoch": 0.823200528285274, + "grad_norm": 2.0929663451096268, + "learning_rate": 7.975557911123161e-07, + "loss": 0.2847, + "step": 37398 + }, + { + "epoch": 0.8232225401716927, + "grad_norm": 1.5396840871028306, + "learning_rate": 7.973626634403737e-07, + "loss": 0.2764, + "step": 37399 + }, + { + "epoch": 0.8232445520581114, + "grad_norm": 1.5759370890289264, + "learning_rate": 7.971695571281085e-07, + "loss": 0.2007, + "step": 37400 + }, + { + "epoch": 0.82326656394453, + "grad_norm": 7.005696886789293, + "learning_rate": 7.969764721765016e-07, + "loss": 0.3113, + "step": 37401 + }, + { + "epoch": 0.8232885758309487, + "grad_norm": 2.0288723420070656, + "learning_rate": 7.967834085865356e-07, + "loss": 0.2631, + "step": 37402 + }, + { + "epoch": 0.8233105877173674, + "grad_norm": 1.8058123101615575, + "learning_rate": 7.965903663591895e-07, + "loss": 0.3086, + "step": 37403 + }, + { + "epoch": 0.823332599603786, + "grad_norm": 1.7025595854836744, + "learning_rate": 7.963973454954465e-07, + "loss": 0.2509, + "step": 37404 + }, + { + "epoch": 0.8233546114902047, + "grad_norm": 1.8441446078052488, + "learning_rate": 7.962043459962882e-07, + "loss": 0.3336, + "step": 37405 + }, + { + "epoch": 0.8233766233766234, + "grad_norm": 1.5817553883499247, + "learning_rate": 7.960113678626929e-07, + "loss": 0.255, + "step": 37406 + }, + { + "epoch": 0.823398635263042, + "grad_norm": 2.718003086048857, + "learning_rate": 7.95818411095644e-07, + "loss": 0.2917, + "step": 37407 + }, + { + "epoch": 0.8234206471494607, + "grad_norm": 1.7273446823519258, + "learning_rate": 7.956254756961202e-07, + "loss": 0.2276, + "step": 37408 + }, + { + "epoch": 0.8234426590358793, + "grad_norm": 1.7696521721827707, + "learning_rate": 7.954325616651032e-07, + "loss": 0.3143, + "step": 37409 + }, + { + "epoch": 0.823464670922298, + "grad_norm": 6.627354672725854, + "learning_rate": 7.952396690035736e-07, + "loss": 0.315, + "step": 37410 + }, + { + "epoch": 0.8234866828087167, + "grad_norm": 0.6093449465104033, + "learning_rate": 7.950467977125109e-07, + "loss": 0.4399, + "step": 37411 + }, + { + "epoch": 0.8235086946951353, + "grad_norm": 2.3948750437889346, + "learning_rate": 7.948539477928962e-07, + "loss": 0.3359, + "step": 37412 + }, + { + "epoch": 0.823530706581554, + "grad_norm": 3.3197296167637016, + "learning_rate": 7.946611192457088e-07, + "loss": 0.3226, + "step": 37413 + }, + { + "epoch": 0.8235527184679727, + "grad_norm": 0.6075508041549761, + "learning_rate": 7.944683120719293e-07, + "loss": 0.4478, + "step": 37414 + }, + { + "epoch": 0.8235747303543913, + "grad_norm": 1.7564490671322959, + "learning_rate": 7.942755262725383e-07, + "loss": 0.329, + "step": 37415 + }, + { + "epoch": 0.82359674224081, + "grad_norm": 2.1490301552924023, + "learning_rate": 7.940827618485137e-07, + "loss": 0.2776, + "step": 37416 + }, + { + "epoch": 0.8236187541272287, + "grad_norm": 1.7404829431644997, + "learning_rate": 7.93890018800838e-07, + "loss": 0.3064, + "step": 37417 + }, + { + "epoch": 0.8236407660136473, + "grad_norm": 1.6161187579687233, + "learning_rate": 7.936972971304874e-07, + "loss": 0.3176, + "step": 37418 + }, + { + "epoch": 0.823662777900066, + "grad_norm": 0.6029998142300901, + "learning_rate": 7.935045968384436e-07, + "loss": 0.4126, + "step": 37419 + }, + { + "epoch": 0.8236847897864847, + "grad_norm": 1.796875117421317, + "learning_rate": 7.933119179256871e-07, + "loss": 0.2602, + "step": 37420 + }, + { + "epoch": 0.8237068016729033, + "grad_norm": 2.013815696423229, + "learning_rate": 7.931192603931936e-07, + "loss": 0.3472, + "step": 37421 + }, + { + "epoch": 0.823728813559322, + "grad_norm": 1.8528618517346371, + "learning_rate": 7.929266242419465e-07, + "loss": 0.3675, + "step": 37422 + }, + { + "epoch": 0.8237508254457407, + "grad_norm": 0.6352011068931642, + "learning_rate": 7.927340094729208e-07, + "loss": 0.467, + "step": 37423 + }, + { + "epoch": 0.8237728373321593, + "grad_norm": 1.7721141264654863, + "learning_rate": 7.925414160870976e-07, + "loss": 0.2286, + "step": 37424 + }, + { + "epoch": 0.823794849218578, + "grad_norm": 1.4649826113567237, + "learning_rate": 7.923488440854566e-07, + "loss": 0.2488, + "step": 37425 + }, + { + "epoch": 0.8238168611049967, + "grad_norm": 0.621110348991827, + "learning_rate": 7.921562934689759e-07, + "loss": 0.4201, + "step": 37426 + }, + { + "epoch": 0.8238388729914153, + "grad_norm": 1.573450628510336, + "learning_rate": 7.919637642386319e-07, + "loss": 0.2639, + "step": 37427 + }, + { + "epoch": 0.823860884877834, + "grad_norm": 0.6016840503264453, + "learning_rate": 7.917712563954066e-07, + "loss": 0.4182, + "step": 37428 + }, + { + "epoch": 0.8238828967642527, + "grad_norm": 1.8728130008708515, + "learning_rate": 7.915787699402738e-07, + "loss": 0.3098, + "step": 37429 + }, + { + "epoch": 0.8239049086506713, + "grad_norm": 2.097167250431183, + "learning_rate": 7.913863048742171e-07, + "loss": 0.2983, + "step": 37430 + }, + { + "epoch": 0.82392692053709, + "grad_norm": 2.0341728408585062, + "learning_rate": 7.911938611982128e-07, + "loss": 0.2562, + "step": 37431 + }, + { + "epoch": 0.8239489324235086, + "grad_norm": 2.094355699745606, + "learning_rate": 7.91001438913237e-07, + "loss": 0.2999, + "step": 37432 + }, + { + "epoch": 0.8239709443099273, + "grad_norm": 1.6660664177328517, + "learning_rate": 7.908090380202699e-07, + "loss": 0.3206, + "step": 37433 + }, + { + "epoch": 0.823992956196346, + "grad_norm": 1.8306781812897268, + "learning_rate": 7.906166585202868e-07, + "loss": 0.2857, + "step": 37434 + }, + { + "epoch": 0.8240149680827646, + "grad_norm": 1.5299219241565991, + "learning_rate": 7.904243004142698e-07, + "loss": 0.2949, + "step": 37435 + }, + { + "epoch": 0.8240369799691833, + "grad_norm": 2.4101649926901048, + "learning_rate": 7.902319637031936e-07, + "loss": 0.4022, + "step": 37436 + }, + { + "epoch": 0.824058991855602, + "grad_norm": 1.9129222380010902, + "learning_rate": 7.900396483880351e-07, + "loss": 0.2998, + "step": 37437 + }, + { + "epoch": 0.8240810037420206, + "grad_norm": 1.4452493457142503, + "learning_rate": 7.898473544697738e-07, + "loss": 0.2512, + "step": 37438 + }, + { + "epoch": 0.8241030156284394, + "grad_norm": 1.7686478029662482, + "learning_rate": 7.896550819493854e-07, + "loss": 0.2863, + "step": 37439 + }, + { + "epoch": 0.824125027514858, + "grad_norm": 1.5158845145576723, + "learning_rate": 7.894628308278474e-07, + "loss": 0.259, + "step": 37440 + }, + { + "epoch": 0.8241470394012766, + "grad_norm": 0.6280458046601238, + "learning_rate": 7.892706011061385e-07, + "loss": 0.4688, + "step": 37441 + }, + { + "epoch": 0.8241690512876954, + "grad_norm": 0.6104065225835401, + "learning_rate": 7.890783927852331e-07, + "loss": 0.4647, + "step": 37442 + }, + { + "epoch": 0.8241910631741141, + "grad_norm": 1.9611335880065033, + "learning_rate": 7.888862058661107e-07, + "loss": 0.2762, + "step": 37443 + }, + { + "epoch": 0.8242130750605327, + "grad_norm": 1.3977964246067085, + "learning_rate": 7.886940403497456e-07, + "loss": 0.3202, + "step": 37444 + }, + { + "epoch": 0.8242350869469514, + "grad_norm": 1.3205047435202057, + "learning_rate": 7.885018962371155e-07, + "loss": 0.2339, + "step": 37445 + }, + { + "epoch": 0.8242570988333701, + "grad_norm": 2.050590477506255, + "learning_rate": 7.883097735291984e-07, + "loss": 0.3136, + "step": 37446 + }, + { + "epoch": 0.8242791107197887, + "grad_norm": 1.6651650248112169, + "learning_rate": 7.881176722269685e-07, + "loss": 0.2608, + "step": 37447 + }, + { + "epoch": 0.8243011226062074, + "grad_norm": 0.607599977804097, + "learning_rate": 7.879255923314044e-07, + "loss": 0.4361, + "step": 37448 + }, + { + "epoch": 0.8243231344926261, + "grad_norm": 1.5402095364234805, + "learning_rate": 7.877335338434799e-07, + "loss": 0.247, + "step": 37449 + }, + { + "epoch": 0.8243451463790447, + "grad_norm": 1.6884805196517338, + "learning_rate": 7.87541496764172e-07, + "loss": 0.3356, + "step": 37450 + }, + { + "epoch": 0.8243671582654634, + "grad_norm": 1.5354854452939581, + "learning_rate": 7.873494810944581e-07, + "loss": 0.2692, + "step": 37451 + }, + { + "epoch": 0.8243891701518821, + "grad_norm": 1.9126373914421178, + "learning_rate": 7.871574868353127e-07, + "loss": 0.3113, + "step": 37452 + }, + { + "epoch": 0.8244111820383007, + "grad_norm": 1.542240471925239, + "learning_rate": 7.869655139877125e-07, + "loss": 0.2911, + "step": 37453 + }, + { + "epoch": 0.8244331939247194, + "grad_norm": 1.498371174683, + "learning_rate": 7.867735625526312e-07, + "loss": 0.3117, + "step": 37454 + }, + { + "epoch": 0.8244552058111381, + "grad_norm": 2.1734001969931622, + "learning_rate": 7.86581632531046e-07, + "loss": 0.2341, + "step": 37455 + }, + { + "epoch": 0.8244772176975567, + "grad_norm": 1.5770553745618618, + "learning_rate": 7.863897239239338e-07, + "loss": 0.4043, + "step": 37456 + }, + { + "epoch": 0.8244992295839754, + "grad_norm": 1.8953169165324286, + "learning_rate": 7.861978367322664e-07, + "loss": 0.3711, + "step": 37457 + }, + { + "epoch": 0.824521241470394, + "grad_norm": 1.5383843780955713, + "learning_rate": 7.860059709570229e-07, + "loss": 0.2958, + "step": 37458 + }, + { + "epoch": 0.8245432533568127, + "grad_norm": 2.198437588177807, + "learning_rate": 7.858141265991748e-07, + "loss": 0.4045, + "step": 37459 + }, + { + "epoch": 0.8245652652432314, + "grad_norm": 1.626762205657146, + "learning_rate": 7.856223036596994e-07, + "loss": 0.2591, + "step": 37460 + }, + { + "epoch": 0.82458727712965, + "grad_norm": 2.8973585537374316, + "learning_rate": 7.854305021395719e-07, + "loss": 0.3111, + "step": 37461 + }, + { + "epoch": 0.8246092890160687, + "grad_norm": 0.6250966197133384, + "learning_rate": 7.852387220397656e-07, + "loss": 0.4627, + "step": 37462 + }, + { + "epoch": 0.8246313009024874, + "grad_norm": 0.6004363031144235, + "learning_rate": 7.850469633612567e-07, + "loss": 0.4217, + "step": 37463 + }, + { + "epoch": 0.824653312788906, + "grad_norm": 1.3489705119663382, + "learning_rate": 7.848552261050191e-07, + "loss": 0.2609, + "step": 37464 + }, + { + "epoch": 0.8246753246753247, + "grad_norm": 4.2758196215847555, + "learning_rate": 7.846635102720251e-07, + "loss": 0.3371, + "step": 37465 + }, + { + "epoch": 0.8246973365617434, + "grad_norm": 0.6236991515960261, + "learning_rate": 7.844718158632536e-07, + "loss": 0.459, + "step": 37466 + }, + { + "epoch": 0.824719348448162, + "grad_norm": 1.54277860464938, + "learning_rate": 7.842801428796759e-07, + "loss": 0.2771, + "step": 37467 + }, + { + "epoch": 0.8247413603345807, + "grad_norm": 1.7618485447871932, + "learning_rate": 7.840884913222663e-07, + "loss": 0.3202, + "step": 37468 + }, + { + "epoch": 0.8247633722209994, + "grad_norm": 0.603420775088782, + "learning_rate": 7.838968611920006e-07, + "loss": 0.4384, + "step": 37469 + }, + { + "epoch": 0.824785384107418, + "grad_norm": 0.6124786588312184, + "learning_rate": 7.837052524898497e-07, + "loss": 0.4632, + "step": 37470 + }, + { + "epoch": 0.8248073959938367, + "grad_norm": 2.230599188260485, + "learning_rate": 7.835136652167896e-07, + "loss": 0.3223, + "step": 37471 + }, + { + "epoch": 0.8248294078802554, + "grad_norm": 1.7772919501863635, + "learning_rate": 7.83322099373795e-07, + "loss": 0.2839, + "step": 37472 + }, + { + "epoch": 0.824851419766674, + "grad_norm": 2.2382138934818196, + "learning_rate": 7.831305549618368e-07, + "loss": 0.2884, + "step": 37473 + }, + { + "epoch": 0.8248734316530927, + "grad_norm": 2.073522390739229, + "learning_rate": 7.829390319818908e-07, + "loss": 0.2929, + "step": 37474 + }, + { + "epoch": 0.8248954435395114, + "grad_norm": 1.8088239279766887, + "learning_rate": 7.827475304349285e-07, + "loss": 0.3321, + "step": 37475 + }, + { + "epoch": 0.82491745542593, + "grad_norm": 1.4887919678609316, + "learning_rate": 7.825560503219242e-07, + "loss": 0.3039, + "step": 37476 + }, + { + "epoch": 0.8249394673123487, + "grad_norm": 1.4061634043392497, + "learning_rate": 7.823645916438527e-07, + "loss": 0.2648, + "step": 37477 + }, + { + "epoch": 0.8249614791987674, + "grad_norm": 1.4402303287791112, + "learning_rate": 7.821731544016836e-07, + "loss": 0.3786, + "step": 37478 + }, + { + "epoch": 0.824983491085186, + "grad_norm": 0.5989463263423802, + "learning_rate": 7.819817385963935e-07, + "loss": 0.4415, + "step": 37479 + }, + { + "epoch": 0.8250055029716047, + "grad_norm": 0.642082285620754, + "learning_rate": 7.81790344228952e-07, + "loss": 0.4986, + "step": 37480 + }, + { + "epoch": 0.8250275148580233, + "grad_norm": 1.4448357778338634, + "learning_rate": 7.81598971300333e-07, + "loss": 0.2024, + "step": 37481 + }, + { + "epoch": 0.825049526744442, + "grad_norm": 1.7891020337217356, + "learning_rate": 7.814076198115112e-07, + "loss": 0.2843, + "step": 37482 + }, + { + "epoch": 0.8250715386308607, + "grad_norm": 1.4999927097609267, + "learning_rate": 7.812162897634562e-07, + "loss": 0.2986, + "step": 37483 + }, + { + "epoch": 0.8250935505172793, + "grad_norm": 0.589282099798189, + "learning_rate": 7.810249811571424e-07, + "loss": 0.4159, + "step": 37484 + }, + { + "epoch": 0.825115562403698, + "grad_norm": 1.6367625000428083, + "learning_rate": 7.808336939935407e-07, + "loss": 0.2347, + "step": 37485 + }, + { + "epoch": 0.8251375742901167, + "grad_norm": 0.6271586055468158, + "learning_rate": 7.806424282736236e-07, + "loss": 0.4414, + "step": 37486 + }, + { + "epoch": 0.8251595861765353, + "grad_norm": 1.3682592837341019, + "learning_rate": 7.804511839983653e-07, + "loss": 0.2858, + "step": 37487 + }, + { + "epoch": 0.825181598062954, + "grad_norm": 1.862748057394744, + "learning_rate": 7.802599611687345e-07, + "loss": 0.3261, + "step": 37488 + }, + { + "epoch": 0.8252036099493727, + "grad_norm": 1.5537546487145564, + "learning_rate": 7.800687597857053e-07, + "loss": 0.3046, + "step": 37489 + }, + { + "epoch": 0.8252256218357913, + "grad_norm": 2.215586945518436, + "learning_rate": 7.798775798502484e-07, + "loss": 0.2873, + "step": 37490 + }, + { + "epoch": 0.82524763372221, + "grad_norm": 1.923010114785637, + "learning_rate": 7.79686421363336e-07, + "loss": 0.2921, + "step": 37491 + }, + { + "epoch": 0.8252696456086287, + "grad_norm": 0.5906446097230129, + "learning_rate": 7.794952843259401e-07, + "loss": 0.4315, + "step": 37492 + }, + { + "epoch": 0.8252916574950473, + "grad_norm": 0.615499807444305, + "learning_rate": 7.793041687390307e-07, + "loss": 0.423, + "step": 37493 + }, + { + "epoch": 0.825313669381466, + "grad_norm": 1.7300907838825597, + "learning_rate": 7.791130746035807e-07, + "loss": 0.2725, + "step": 37494 + }, + { + "epoch": 0.8253356812678847, + "grad_norm": 1.487031499563239, + "learning_rate": 7.789220019205601e-07, + "loss": 0.2599, + "step": 37495 + }, + { + "epoch": 0.8253576931543033, + "grad_norm": 1.51241111312608, + "learning_rate": 7.787309506909402e-07, + "loss": 0.3233, + "step": 37496 + }, + { + "epoch": 0.825379705040722, + "grad_norm": 1.8419050709831093, + "learning_rate": 7.785399209156936e-07, + "loss": 0.341, + "step": 37497 + }, + { + "epoch": 0.8254017169271407, + "grad_norm": 0.613100859378162, + "learning_rate": 7.783489125957883e-07, + "loss": 0.456, + "step": 37498 + }, + { + "epoch": 0.8254237288135593, + "grad_norm": 1.9080967880903115, + "learning_rate": 7.781579257321986e-07, + "loss": 0.3917, + "step": 37499 + }, + { + "epoch": 0.825445740699978, + "grad_norm": 1.8118076402445988, + "learning_rate": 7.779669603258933e-07, + "loss": 0.2996, + "step": 37500 + }, + { + "epoch": 0.8254677525863967, + "grad_norm": 1.5403671204571636, + "learning_rate": 7.777760163778402e-07, + "loss": 0.2347, + "step": 37501 + }, + { + "epoch": 0.8254897644728153, + "grad_norm": 1.458236052132401, + "learning_rate": 7.775850938890151e-07, + "loss": 0.2254, + "step": 37502 + }, + { + "epoch": 0.825511776359234, + "grad_norm": 1.2936417869136125, + "learning_rate": 7.773941928603856e-07, + "loss": 0.2732, + "step": 37503 + }, + { + "epoch": 0.8255337882456526, + "grad_norm": 0.6254760625135439, + "learning_rate": 7.772033132929214e-07, + "loss": 0.445, + "step": 37504 + }, + { + "epoch": 0.8255558001320713, + "grad_norm": 1.6012703351448418, + "learning_rate": 7.770124551875941e-07, + "loss": 0.3232, + "step": 37505 + }, + { + "epoch": 0.82557781201849, + "grad_norm": 1.5726109220591467, + "learning_rate": 7.76821618545372e-07, + "loss": 0.3729, + "step": 37506 + }, + { + "epoch": 0.8255998239049086, + "grad_norm": 0.5932760757290144, + "learning_rate": 7.76630803367226e-07, + "loss": 0.427, + "step": 37507 + }, + { + "epoch": 0.8256218357913273, + "grad_norm": 1.5901066975710159, + "learning_rate": 7.764400096541275e-07, + "loss": 0.3978, + "step": 37508 + }, + { + "epoch": 0.825643847677746, + "grad_norm": 2.25093916143229, + "learning_rate": 7.762492374070429e-07, + "loss": 0.3806, + "step": 37509 + }, + { + "epoch": 0.8256658595641646, + "grad_norm": 1.2870156918965936, + "learning_rate": 7.760584866269449e-07, + "loss": 0.2807, + "step": 37510 + }, + { + "epoch": 0.8256878714505833, + "grad_norm": 3.388595854963356, + "learning_rate": 7.758677573148004e-07, + "loss": 0.23, + "step": 37511 + }, + { + "epoch": 0.825709883337002, + "grad_norm": 1.5696777910359623, + "learning_rate": 7.756770494715804e-07, + "loss": 0.3335, + "step": 37512 + }, + { + "epoch": 0.8257318952234206, + "grad_norm": 1.5847980723956003, + "learning_rate": 7.754863630982546e-07, + "loss": 0.3084, + "step": 37513 + }, + { + "epoch": 0.8257539071098393, + "grad_norm": 1.8346760216688867, + "learning_rate": 7.752956981957898e-07, + "loss": 0.3026, + "step": 37514 + }, + { + "epoch": 0.825775918996258, + "grad_norm": 0.6194741193789641, + "learning_rate": 7.751050547651578e-07, + "loss": 0.4565, + "step": 37515 + }, + { + "epoch": 0.8257979308826766, + "grad_norm": 1.7276577627124017, + "learning_rate": 7.749144328073255e-07, + "loss": 0.2593, + "step": 37516 + }, + { + "epoch": 0.8258199427690953, + "grad_norm": 2.2909025797647753, + "learning_rate": 7.747238323232625e-07, + "loss": 0.2466, + "step": 37517 + }, + { + "epoch": 0.825841954655514, + "grad_norm": 2.1124570857583467, + "learning_rate": 7.745332533139382e-07, + "loss": 0.2856, + "step": 37518 + }, + { + "epoch": 0.8258639665419326, + "grad_norm": 1.4565737028568162, + "learning_rate": 7.743426957803196e-07, + "loss": 0.2441, + "step": 37519 + }, + { + "epoch": 0.8258859784283513, + "grad_norm": 4.506222832867244, + "learning_rate": 7.741521597233775e-07, + "loss": 0.2309, + "step": 37520 + }, + { + "epoch": 0.82590799031477, + "grad_norm": 1.4759378788835689, + "learning_rate": 7.739616451440774e-07, + "loss": 0.2208, + "step": 37521 + }, + { + "epoch": 0.8259300022011886, + "grad_norm": 2.1340249495113794, + "learning_rate": 7.737711520433893e-07, + "loss": 0.252, + "step": 37522 + }, + { + "epoch": 0.8259520140876073, + "grad_norm": 3.425734162988206, + "learning_rate": 7.735806804222822e-07, + "loss": 0.2166, + "step": 37523 + }, + { + "epoch": 0.825974025974026, + "grad_norm": 1.6101504495227184, + "learning_rate": 7.73390230281722e-07, + "loss": 0.3478, + "step": 37524 + }, + { + "epoch": 0.8259960378604446, + "grad_norm": 1.7120270032228142, + "learning_rate": 7.731998016226788e-07, + "loss": 0.3797, + "step": 37525 + }, + { + "epoch": 0.8260180497468633, + "grad_norm": 0.6480299654462947, + "learning_rate": 7.730093944461181e-07, + "loss": 0.4766, + "step": 37526 + }, + { + "epoch": 0.8260400616332819, + "grad_norm": 1.3475907109910457, + "learning_rate": 7.728190087530091e-07, + "loss": 0.2556, + "step": 37527 + }, + { + "epoch": 0.8260620735197006, + "grad_norm": 1.8062333585019374, + "learning_rate": 7.72628644544321e-07, + "loss": 0.3785, + "step": 37528 + }, + { + "epoch": 0.8260840854061193, + "grad_norm": 1.6101944480538468, + "learning_rate": 7.72438301821018e-07, + "loss": 0.379, + "step": 37529 + }, + { + "epoch": 0.8261060972925379, + "grad_norm": 1.6322221517349398, + "learning_rate": 7.7224798058407e-07, + "loss": 0.3512, + "step": 37530 + }, + { + "epoch": 0.8261281091789566, + "grad_norm": 1.9009014614352089, + "learning_rate": 7.720576808344426e-07, + "loss": 0.3371, + "step": 37531 + }, + { + "epoch": 0.8261501210653753, + "grad_norm": 1.8111290104750304, + "learning_rate": 7.718674025731032e-07, + "loss": 0.2504, + "step": 37532 + }, + { + "epoch": 0.8261721329517939, + "grad_norm": 1.5185170726078692, + "learning_rate": 7.716771458010213e-07, + "loss": 0.2452, + "step": 37533 + }, + { + "epoch": 0.8261941448382126, + "grad_norm": 1.7027069827670438, + "learning_rate": 7.714869105191608e-07, + "loss": 0.2658, + "step": 37534 + }, + { + "epoch": 0.8262161567246313, + "grad_norm": 1.9466004747075778, + "learning_rate": 7.712966967284907e-07, + "loss": 0.3643, + "step": 37535 + }, + { + "epoch": 0.8262381686110499, + "grad_norm": 1.5017314631597363, + "learning_rate": 7.71106504429977e-07, + "loss": 0.3305, + "step": 37536 + }, + { + "epoch": 0.8262601804974686, + "grad_norm": 0.6525453973093468, + "learning_rate": 7.709163336245834e-07, + "loss": 0.4637, + "step": 37537 + }, + { + "epoch": 0.8262821923838873, + "grad_norm": 1.7735559506488285, + "learning_rate": 7.707261843132824e-07, + "loss": 0.3092, + "step": 37538 + }, + { + "epoch": 0.8263042042703059, + "grad_norm": 1.80753753316161, + "learning_rate": 7.705360564970354e-07, + "loss": 0.2852, + "step": 37539 + }, + { + "epoch": 0.8263262161567246, + "grad_norm": 1.4005974713597886, + "learning_rate": 7.703459501768118e-07, + "loss": 0.2933, + "step": 37540 + }, + { + "epoch": 0.8263482280431433, + "grad_norm": 2.002215117943076, + "learning_rate": 7.701558653535768e-07, + "loss": 0.2685, + "step": 37541 + }, + { + "epoch": 0.8263702399295619, + "grad_norm": 0.6312418128453842, + "learning_rate": 7.699658020282946e-07, + "loss": 0.4575, + "step": 37542 + }, + { + "epoch": 0.8263922518159806, + "grad_norm": 1.5593401399271125, + "learning_rate": 7.697757602019335e-07, + "loss": 0.2873, + "step": 37543 + }, + { + "epoch": 0.8264142637023993, + "grad_norm": 0.6174918503724582, + "learning_rate": 7.695857398754597e-07, + "loss": 0.4756, + "step": 37544 + }, + { + "epoch": 0.8264362755888179, + "grad_norm": 1.3761544608189091, + "learning_rate": 7.693957410498365e-07, + "loss": 0.2776, + "step": 37545 + }, + { + "epoch": 0.8264582874752366, + "grad_norm": 0.6362728674307441, + "learning_rate": 7.692057637260325e-07, + "loss": 0.4563, + "step": 37546 + }, + { + "epoch": 0.8264802993616553, + "grad_norm": 1.9658501155133852, + "learning_rate": 7.690158079050103e-07, + "loss": 0.2197, + "step": 37547 + }, + { + "epoch": 0.8265023112480739, + "grad_norm": 1.739812285123008, + "learning_rate": 7.688258735877374e-07, + "loss": 0.2527, + "step": 37548 + }, + { + "epoch": 0.8265243231344926, + "grad_norm": 2.047621387935419, + "learning_rate": 7.686359607751792e-07, + "loss": 0.342, + "step": 37549 + }, + { + "epoch": 0.8265463350209112, + "grad_norm": 1.966217263935806, + "learning_rate": 7.684460694682994e-07, + "loss": 0.2806, + "step": 37550 + }, + { + "epoch": 0.8265683469073299, + "grad_norm": 0.6070288859449119, + "learning_rate": 7.682561996680654e-07, + "loss": 0.4594, + "step": 37551 + }, + { + "epoch": 0.8265903587937486, + "grad_norm": 0.6033379922257177, + "learning_rate": 7.680663513754394e-07, + "loss": 0.4427, + "step": 37552 + }, + { + "epoch": 0.8266123706801672, + "grad_norm": 1.4388285156542076, + "learning_rate": 7.678765245913871e-07, + "loss": 0.3042, + "step": 37553 + }, + { + "epoch": 0.826634382566586, + "grad_norm": 0.6489102935548486, + "learning_rate": 7.676867193168757e-07, + "loss": 0.4479, + "step": 37554 + }, + { + "epoch": 0.8266563944530047, + "grad_norm": 1.5226113531222787, + "learning_rate": 7.674969355528671e-07, + "loss": 0.295, + "step": 37555 + }, + { + "epoch": 0.8266784063394232, + "grad_norm": 1.8739928303211884, + "learning_rate": 7.673071733003273e-07, + "loss": 0.3028, + "step": 37556 + }, + { + "epoch": 0.826700418225842, + "grad_norm": 0.5866486473359549, + "learning_rate": 7.671174325602198e-07, + "loss": 0.4143, + "step": 37557 + }, + { + "epoch": 0.8267224301122607, + "grad_norm": 1.9935124046247017, + "learning_rate": 7.66927713333509e-07, + "loss": 0.337, + "step": 37558 + }, + { + "epoch": 0.8267444419986792, + "grad_norm": 2.0117706323407063, + "learning_rate": 7.667380156211606e-07, + "loss": 0.2453, + "step": 37559 + }, + { + "epoch": 0.826766453885098, + "grad_norm": 1.8922739288673034, + "learning_rate": 7.665483394241368e-07, + "loss": 0.3139, + "step": 37560 + }, + { + "epoch": 0.8267884657715167, + "grad_norm": 1.7888020098716002, + "learning_rate": 7.66358684743404e-07, + "loss": 0.2918, + "step": 37561 + }, + { + "epoch": 0.8268104776579352, + "grad_norm": 1.4188892682889553, + "learning_rate": 7.66169051579923e-07, + "loss": 0.2535, + "step": 37562 + }, + { + "epoch": 0.826832489544354, + "grad_norm": 1.5608478246594715, + "learning_rate": 7.659794399346593e-07, + "loss": 0.3105, + "step": 37563 + }, + { + "epoch": 0.8268545014307727, + "grad_norm": 3.1699982151542767, + "learning_rate": 7.657898498085775e-07, + "loss": 0.2135, + "step": 37564 + }, + { + "epoch": 0.8268765133171913, + "grad_norm": 1.747770776685389, + "learning_rate": 7.656002812026392e-07, + "loss": 0.3315, + "step": 37565 + }, + { + "epoch": 0.82689852520361, + "grad_norm": 1.7463573347394767, + "learning_rate": 7.654107341178102e-07, + "loss": 0.3496, + "step": 37566 + }, + { + "epoch": 0.8269205370900287, + "grad_norm": 5.634806076337512, + "learning_rate": 7.652212085550526e-07, + "loss": 0.2398, + "step": 37567 + }, + { + "epoch": 0.8269425489764473, + "grad_norm": 1.8771086623097462, + "learning_rate": 7.650317045153271e-07, + "loss": 0.2853, + "step": 37568 + }, + { + "epoch": 0.826964560862866, + "grad_norm": 1.8640374443687644, + "learning_rate": 7.648422219996016e-07, + "loss": 0.268, + "step": 37569 + }, + { + "epoch": 0.8269865727492847, + "grad_norm": 1.5325941345024052, + "learning_rate": 7.646527610088355e-07, + "loss": 0.3325, + "step": 37570 + }, + { + "epoch": 0.8270085846357033, + "grad_norm": 1.9320691115009838, + "learning_rate": 7.644633215439945e-07, + "loss": 0.3028, + "step": 37571 + }, + { + "epoch": 0.827030596522122, + "grad_norm": 0.6120404222912973, + "learning_rate": 7.642739036060398e-07, + "loss": 0.4596, + "step": 37572 + }, + { + "epoch": 0.8270526084085406, + "grad_norm": 2.747775493049173, + "learning_rate": 7.640845071959324e-07, + "loss": 0.3323, + "step": 37573 + }, + { + "epoch": 0.8270746202949593, + "grad_norm": 1.7914315907631821, + "learning_rate": 7.638951323146388e-07, + "loss": 0.2859, + "step": 37574 + }, + { + "epoch": 0.827096632181378, + "grad_norm": 1.7369507169102767, + "learning_rate": 7.637057789631181e-07, + "loss": 0.2463, + "step": 37575 + }, + { + "epoch": 0.8271186440677966, + "grad_norm": 0.6330626077384973, + "learning_rate": 7.635164471423357e-07, + "loss": 0.4552, + "step": 37576 + }, + { + "epoch": 0.8271406559542153, + "grad_norm": 2.835802757943656, + "learning_rate": 7.633271368532519e-07, + "loss": 0.1768, + "step": 37577 + }, + { + "epoch": 0.827162667840634, + "grad_norm": 3.9599423819581916, + "learning_rate": 7.631378480968283e-07, + "loss": 0.2964, + "step": 37578 + }, + { + "epoch": 0.8271846797270526, + "grad_norm": 0.6127156922120418, + "learning_rate": 7.629485808740278e-07, + "loss": 0.4518, + "step": 37579 + }, + { + "epoch": 0.8272066916134713, + "grad_norm": 1.5606985288747948, + "learning_rate": 7.627593351858126e-07, + "loss": 0.3087, + "step": 37580 + }, + { + "epoch": 0.82722870349989, + "grad_norm": 1.8286771625086884, + "learning_rate": 7.625701110331452e-07, + "loss": 0.4012, + "step": 37581 + }, + { + "epoch": 0.8272507153863086, + "grad_norm": 0.596162205320804, + "learning_rate": 7.623809084169865e-07, + "loss": 0.4193, + "step": 37582 + }, + { + "epoch": 0.8272727272727273, + "grad_norm": 1.2364899538982528, + "learning_rate": 7.621917273382967e-07, + "loss": 0.217, + "step": 37583 + }, + { + "epoch": 0.827294739159146, + "grad_norm": 0.6305262795285036, + "learning_rate": 7.620025677980386e-07, + "loss": 0.4676, + "step": 37584 + }, + { + "epoch": 0.8273167510455646, + "grad_norm": 0.6229273659823535, + "learning_rate": 7.618134297971752e-07, + "loss": 0.4703, + "step": 37585 + }, + { + "epoch": 0.8273387629319833, + "grad_norm": 1.7833898926771552, + "learning_rate": 7.616243133366646e-07, + "loss": 0.2533, + "step": 37586 + }, + { + "epoch": 0.827360774818402, + "grad_norm": 1.7616365115065462, + "learning_rate": 7.614352184174712e-07, + "loss": 0.2511, + "step": 37587 + }, + { + "epoch": 0.8273827867048206, + "grad_norm": 1.6197916982016736, + "learning_rate": 7.612461450405534e-07, + "loss": 0.226, + "step": 37588 + }, + { + "epoch": 0.8274047985912393, + "grad_norm": 2.4728170532178146, + "learning_rate": 7.610570932068729e-07, + "loss": 0.288, + "step": 37589 + }, + { + "epoch": 0.827426810477658, + "grad_norm": 2.2301285573288365, + "learning_rate": 7.60868062917392e-07, + "loss": 0.3012, + "step": 37590 + }, + { + "epoch": 0.8274488223640766, + "grad_norm": 1.888066728879105, + "learning_rate": 7.606790541730696e-07, + "loss": 0.3202, + "step": 37591 + }, + { + "epoch": 0.8274708342504953, + "grad_norm": 2.563298359544743, + "learning_rate": 7.604900669748677e-07, + "loss": 0.2511, + "step": 37592 + }, + { + "epoch": 0.827492846136914, + "grad_norm": 0.6072255281473776, + "learning_rate": 7.603011013237449e-07, + "loss": 0.4528, + "step": 37593 + }, + { + "epoch": 0.8275148580233326, + "grad_norm": 1.5268437847713499, + "learning_rate": 7.601121572206632e-07, + "loss": 0.258, + "step": 37594 + }, + { + "epoch": 0.8275368699097513, + "grad_norm": 3.8180244828071848, + "learning_rate": 7.599232346665841e-07, + "loss": 0.3092, + "step": 37595 + }, + { + "epoch": 0.82755888179617, + "grad_norm": 0.6196517275832846, + "learning_rate": 7.597343336624646e-07, + "loss": 0.4243, + "step": 37596 + }, + { + "epoch": 0.8275808936825886, + "grad_norm": 2.3053075122778357, + "learning_rate": 7.595454542092672e-07, + "loss": 0.244, + "step": 37597 + }, + { + "epoch": 0.8276029055690073, + "grad_norm": 1.4621471938032058, + "learning_rate": 7.593565963079508e-07, + "loss": 0.2311, + "step": 37598 + }, + { + "epoch": 0.8276249174554259, + "grad_norm": 1.9753864188339674, + "learning_rate": 7.591677599594749e-07, + "loss": 0.3212, + "step": 37599 + }, + { + "epoch": 0.8276469293418446, + "grad_norm": 2.5264589079153867, + "learning_rate": 7.589789451648016e-07, + "loss": 0.3164, + "step": 37600 + }, + { + "epoch": 0.8276689412282633, + "grad_norm": 0.6049859837626472, + "learning_rate": 7.587901519248869e-07, + "loss": 0.4515, + "step": 37601 + }, + { + "epoch": 0.8276909531146819, + "grad_norm": 2.3672028414585604, + "learning_rate": 7.586013802406944e-07, + "loss": 0.2887, + "step": 37602 + }, + { + "epoch": 0.8277129650011006, + "grad_norm": 3.0430202144993648, + "learning_rate": 7.584126301131805e-07, + "loss": 0.2988, + "step": 37603 + }, + { + "epoch": 0.8277349768875193, + "grad_norm": 3.6273719437630203, + "learning_rate": 7.582239015433035e-07, + "loss": 0.3512, + "step": 37604 + }, + { + "epoch": 0.8277569887739379, + "grad_norm": 2.5272734841140125, + "learning_rate": 7.580351945320269e-07, + "loss": 0.3612, + "step": 37605 + }, + { + "epoch": 0.8277790006603566, + "grad_norm": 1.674526075791358, + "learning_rate": 7.578465090803067e-07, + "loss": 0.252, + "step": 37606 + }, + { + "epoch": 0.8278010125467753, + "grad_norm": 1.844144789060214, + "learning_rate": 7.576578451891031e-07, + "loss": 0.2296, + "step": 37607 + }, + { + "epoch": 0.8278230244331939, + "grad_norm": 1.6960853596120655, + "learning_rate": 7.574692028593745e-07, + "loss": 0.2795, + "step": 37608 + }, + { + "epoch": 0.8278450363196126, + "grad_norm": 1.7464794965518549, + "learning_rate": 7.572805820920776e-07, + "loss": 0.2763, + "step": 37609 + }, + { + "epoch": 0.8278670482060313, + "grad_norm": 2.09061800898824, + "learning_rate": 7.570919828881751e-07, + "loss": 0.3322, + "step": 37610 + }, + { + "epoch": 0.8278890600924499, + "grad_norm": 3.8919434272336506, + "learning_rate": 7.569034052486224e-07, + "loss": 0.2953, + "step": 37611 + }, + { + "epoch": 0.8279110719788686, + "grad_norm": 1.8049734880721149, + "learning_rate": 7.567148491743803e-07, + "loss": 0.3159, + "step": 37612 + }, + { + "epoch": 0.8279330838652873, + "grad_norm": 1.499650141078596, + "learning_rate": 7.56526314666406e-07, + "loss": 0.282, + "step": 37613 + }, + { + "epoch": 0.8279550957517059, + "grad_norm": 1.6587557900113847, + "learning_rate": 7.563378017256567e-07, + "loss": 0.2636, + "step": 37614 + }, + { + "epoch": 0.8279771076381246, + "grad_norm": 0.6057230190440435, + "learning_rate": 7.56149310353091e-07, + "loss": 0.4427, + "step": 37615 + }, + { + "epoch": 0.8279991195245433, + "grad_norm": 1.4420522712544812, + "learning_rate": 7.559608405496677e-07, + "loss": 0.2526, + "step": 37616 + }, + { + "epoch": 0.8280211314109619, + "grad_norm": 2.408358767372479, + "learning_rate": 7.557723923163451e-07, + "loss": 0.3375, + "step": 37617 + }, + { + "epoch": 0.8280431432973806, + "grad_norm": 0.6024885254644322, + "learning_rate": 7.555839656540803e-07, + "loss": 0.437, + "step": 37618 + }, + { + "epoch": 0.8280651551837993, + "grad_norm": 1.511548417929273, + "learning_rate": 7.553955605638297e-07, + "loss": 0.2455, + "step": 37619 + }, + { + "epoch": 0.8280871670702179, + "grad_norm": 1.8143367300741096, + "learning_rate": 7.552071770465518e-07, + "loss": 0.2735, + "step": 37620 + }, + { + "epoch": 0.8281091789566366, + "grad_norm": 1.470787933312806, + "learning_rate": 7.550188151032045e-07, + "loss": 0.2578, + "step": 37621 + }, + { + "epoch": 0.8281311908430552, + "grad_norm": 1.62855968300849, + "learning_rate": 7.548304747347457e-07, + "loss": 0.2686, + "step": 37622 + }, + { + "epoch": 0.8281532027294739, + "grad_norm": 1.3140188359314182, + "learning_rate": 7.546421559421324e-07, + "loss": 0.2576, + "step": 37623 + }, + { + "epoch": 0.8281752146158926, + "grad_norm": 1.6316408289223179, + "learning_rate": 7.544538587263195e-07, + "loss": 0.292, + "step": 37624 + }, + { + "epoch": 0.8281972265023112, + "grad_norm": 1.7763093519467068, + "learning_rate": 7.542655830882661e-07, + "loss": 0.359, + "step": 37625 + }, + { + "epoch": 0.8282192383887299, + "grad_norm": 1.6952040103461332, + "learning_rate": 7.540773290289294e-07, + "loss": 0.2444, + "step": 37626 + }, + { + "epoch": 0.8282412502751486, + "grad_norm": 1.6243051177885222, + "learning_rate": 7.538890965492646e-07, + "loss": 0.3095, + "step": 37627 + }, + { + "epoch": 0.8282632621615672, + "grad_norm": 1.3505467962915936, + "learning_rate": 7.537008856502304e-07, + "loss": 0.2419, + "step": 37628 + }, + { + "epoch": 0.8282852740479859, + "grad_norm": 2.9672218060184434, + "learning_rate": 7.535126963327805e-07, + "loss": 0.2632, + "step": 37629 + }, + { + "epoch": 0.8283072859344046, + "grad_norm": 1.4746834795282064, + "learning_rate": 7.533245285978735e-07, + "loss": 0.2751, + "step": 37630 + }, + { + "epoch": 0.8283292978208232, + "grad_norm": 1.7113456214336615, + "learning_rate": 7.531363824464666e-07, + "loss": 0.3233, + "step": 37631 + }, + { + "epoch": 0.8283513097072419, + "grad_norm": 1.4763453039607002, + "learning_rate": 7.529482578795133e-07, + "loss": 0.3433, + "step": 37632 + }, + { + "epoch": 0.8283733215936606, + "grad_norm": 0.6078413162243785, + "learning_rate": 7.527601548979723e-07, + "loss": 0.4392, + "step": 37633 + }, + { + "epoch": 0.8283953334800792, + "grad_norm": 2.5294628503331222, + "learning_rate": 7.52572073502798e-07, + "loss": 0.2166, + "step": 37634 + }, + { + "epoch": 0.8284173453664979, + "grad_norm": 1.6668190949679453, + "learning_rate": 7.523840136949462e-07, + "loss": 0.2934, + "step": 37635 + }, + { + "epoch": 0.8284393572529166, + "grad_norm": 2.3131967217867238, + "learning_rate": 7.521959754753744e-07, + "loss": 0.3144, + "step": 37636 + }, + { + "epoch": 0.8284613691393352, + "grad_norm": 2.2294403408292642, + "learning_rate": 7.520079588450363e-07, + "loss": 0.232, + "step": 37637 + }, + { + "epoch": 0.8284833810257539, + "grad_norm": 14.102667561149064, + "learning_rate": 7.518199638048901e-07, + "loss": 0.3199, + "step": 37638 + }, + { + "epoch": 0.8285053929121726, + "grad_norm": 3.2988997739200006, + "learning_rate": 7.516319903558877e-07, + "loss": 0.335, + "step": 37639 + }, + { + "epoch": 0.8285274047985912, + "grad_norm": 1.609491848181322, + "learning_rate": 7.514440384989863e-07, + "loss": 0.2842, + "step": 37640 + }, + { + "epoch": 0.8285494166850099, + "grad_norm": 1.3840444048569287, + "learning_rate": 7.512561082351427e-07, + "loss": 0.2544, + "step": 37641 + }, + { + "epoch": 0.8285714285714286, + "grad_norm": 1.8194337959557756, + "learning_rate": 7.510681995653096e-07, + "loss": 0.402, + "step": 37642 + }, + { + "epoch": 0.8285934404578472, + "grad_norm": 1.2366797144005934, + "learning_rate": 7.50880312490444e-07, + "loss": 0.2535, + "step": 37643 + }, + { + "epoch": 0.8286154523442659, + "grad_norm": 1.885034509069004, + "learning_rate": 7.506924470114995e-07, + "loss": 0.2844, + "step": 37644 + }, + { + "epoch": 0.8286374642306845, + "grad_norm": 1.6034109718596454, + "learning_rate": 7.505046031294289e-07, + "loss": 0.2828, + "step": 37645 + }, + { + "epoch": 0.8286594761171032, + "grad_norm": 0.6056720945874079, + "learning_rate": 7.503167808451916e-07, + "loss": 0.4415, + "step": 37646 + }, + { + "epoch": 0.8286814880035219, + "grad_norm": 1.6509390168134355, + "learning_rate": 7.501289801597389e-07, + "loss": 0.3067, + "step": 37647 + }, + { + "epoch": 0.8287034998899405, + "grad_norm": 1.9619056471963352, + "learning_rate": 7.499412010740276e-07, + "loss": 0.305, + "step": 37648 + }, + { + "epoch": 0.8287255117763592, + "grad_norm": 0.6375509384893914, + "learning_rate": 7.497534435890103e-07, + "loss": 0.4606, + "step": 37649 + }, + { + "epoch": 0.8287475236627779, + "grad_norm": 1.843991357903714, + "learning_rate": 7.495657077056406e-07, + "loss": 0.3247, + "step": 37650 + }, + { + "epoch": 0.8287695355491965, + "grad_norm": 2.035518581014306, + "learning_rate": 7.493779934248735e-07, + "loss": 0.3306, + "step": 37651 + }, + { + "epoch": 0.8287915474356152, + "grad_norm": 1.3452336107648775, + "learning_rate": 7.491903007476636e-07, + "loss": 0.2147, + "step": 37652 + }, + { + "epoch": 0.8288135593220339, + "grad_norm": 1.4768784932974357, + "learning_rate": 7.490026296749659e-07, + "loss": 0.2911, + "step": 37653 + }, + { + "epoch": 0.8288355712084525, + "grad_norm": 4.396414418323015, + "learning_rate": 7.488149802077321e-07, + "loss": 0.3032, + "step": 37654 + }, + { + "epoch": 0.8288575830948712, + "grad_norm": 1.548321643787216, + "learning_rate": 7.486273523469162e-07, + "loss": 0.2666, + "step": 37655 + }, + { + "epoch": 0.8288795949812899, + "grad_norm": 1.4747417387840638, + "learning_rate": 7.484397460934716e-07, + "loss": 0.2558, + "step": 37656 + }, + { + "epoch": 0.8289016068677085, + "grad_norm": 3.542349093676252, + "learning_rate": 7.482521614483529e-07, + "loss": 0.3269, + "step": 37657 + }, + { + "epoch": 0.8289236187541272, + "grad_norm": 0.6238043501765898, + "learning_rate": 7.48064598412514e-07, + "loss": 0.4554, + "step": 37658 + }, + { + "epoch": 0.8289456306405459, + "grad_norm": 1.574579197332436, + "learning_rate": 7.478770569869076e-07, + "loss": 0.249, + "step": 37659 + }, + { + "epoch": 0.8289676425269645, + "grad_norm": 1.5587493323530508, + "learning_rate": 7.476895371724851e-07, + "loss": 0.3915, + "step": 37660 + }, + { + "epoch": 0.8289896544133832, + "grad_norm": 1.638290370690337, + "learning_rate": 7.475020389702009e-07, + "loss": 0.2537, + "step": 37661 + }, + { + "epoch": 0.8290116662998019, + "grad_norm": 0.6168447071013861, + "learning_rate": 7.473145623810079e-07, + "loss": 0.4492, + "step": 37662 + }, + { + "epoch": 0.8290336781862205, + "grad_norm": 3.196315719110474, + "learning_rate": 7.471271074058601e-07, + "loss": 0.2405, + "step": 37663 + }, + { + "epoch": 0.8290556900726392, + "grad_norm": 9.151646766305019, + "learning_rate": 7.469396740457096e-07, + "loss": 0.325, + "step": 37664 + }, + { + "epoch": 0.8290777019590579, + "grad_norm": 2.4279876185823857, + "learning_rate": 7.467522623015072e-07, + "loss": 0.28, + "step": 37665 + }, + { + "epoch": 0.8290997138454765, + "grad_norm": 1.5046606181028703, + "learning_rate": 7.46564872174207e-07, + "loss": 0.291, + "step": 37666 + }, + { + "epoch": 0.8291217257318952, + "grad_norm": 2.6076327188540014, + "learning_rate": 7.463775036647619e-07, + "loss": 0.2692, + "step": 37667 + }, + { + "epoch": 0.8291437376183138, + "grad_norm": 1.6957606422913136, + "learning_rate": 7.461901567741231e-07, + "loss": 0.2582, + "step": 37668 + }, + { + "epoch": 0.8291657495047325, + "grad_norm": 1.7751805799056295, + "learning_rate": 7.460028315032436e-07, + "loss": 0.2477, + "step": 37669 + }, + { + "epoch": 0.8291877613911512, + "grad_norm": 1.917744809261708, + "learning_rate": 7.458155278530743e-07, + "loss": 0.2639, + "step": 37670 + }, + { + "epoch": 0.8292097732775698, + "grad_norm": 0.6360180676668477, + "learning_rate": 7.456282458245678e-07, + "loss": 0.4561, + "step": 37671 + }, + { + "epoch": 0.8292317851639885, + "grad_norm": 2.0875662811789018, + "learning_rate": 7.454409854186773e-07, + "loss": 0.2591, + "step": 37672 + }, + { + "epoch": 0.8292537970504072, + "grad_norm": 1.3818363194120042, + "learning_rate": 7.452537466363518e-07, + "loss": 0.2715, + "step": 37673 + }, + { + "epoch": 0.8292758089368258, + "grad_norm": 1.8614503204258595, + "learning_rate": 7.450665294785459e-07, + "loss": 0.2948, + "step": 37674 + }, + { + "epoch": 0.8292978208232445, + "grad_norm": 1.8626109315727521, + "learning_rate": 7.448793339462085e-07, + "loss": 0.2754, + "step": 37675 + }, + { + "epoch": 0.8293198327096633, + "grad_norm": 2.0969784114016035, + "learning_rate": 7.446921600402917e-07, + "loss": 0.2445, + "step": 37676 + }, + { + "epoch": 0.8293418445960818, + "grad_norm": 0.6237790947517086, + "learning_rate": 7.445050077617488e-07, + "loss": 0.4559, + "step": 37677 + }, + { + "epoch": 0.8293638564825005, + "grad_norm": 0.5962176279899953, + "learning_rate": 7.443178771115284e-07, + "loss": 0.4145, + "step": 37678 + }, + { + "epoch": 0.8293858683689193, + "grad_norm": 1.9291224032515923, + "learning_rate": 7.441307680905835e-07, + "loss": 0.3465, + "step": 37679 + }, + { + "epoch": 0.8294078802553378, + "grad_norm": 1.7475753688045166, + "learning_rate": 7.439436806998634e-07, + "loss": 0.3567, + "step": 37680 + }, + { + "epoch": 0.8294298921417566, + "grad_norm": 1.6026330585328252, + "learning_rate": 7.437566149403209e-07, + "loss": 0.2769, + "step": 37681 + }, + { + "epoch": 0.8294519040281753, + "grad_norm": 0.6321122393397152, + "learning_rate": 7.435695708129043e-07, + "loss": 0.4708, + "step": 37682 + }, + { + "epoch": 0.8294739159145939, + "grad_norm": 1.49694872144208, + "learning_rate": 7.433825483185652e-07, + "loss": 0.2424, + "step": 37683 + }, + { + "epoch": 0.8294959278010126, + "grad_norm": 1.7003260614478914, + "learning_rate": 7.431955474582563e-07, + "loss": 0.3336, + "step": 37684 + }, + { + "epoch": 0.8295179396874313, + "grad_norm": 1.6316688628091642, + "learning_rate": 7.430085682329258e-07, + "loss": 0.3342, + "step": 37685 + }, + { + "epoch": 0.8295399515738499, + "grad_norm": 3.0919630096222424, + "learning_rate": 7.428216106435238e-07, + "loss": 0.3117, + "step": 37686 + }, + { + "epoch": 0.8295619634602686, + "grad_norm": 2.4738239888614078, + "learning_rate": 7.426346746910007e-07, + "loss": 0.31, + "step": 37687 + }, + { + "epoch": 0.8295839753466873, + "grad_norm": 0.6105490956877238, + "learning_rate": 7.42447760376307e-07, + "loss": 0.4451, + "step": 37688 + }, + { + "epoch": 0.8296059872331059, + "grad_norm": 1.4851606729946067, + "learning_rate": 7.42260867700394e-07, + "loss": 0.2535, + "step": 37689 + }, + { + "epoch": 0.8296279991195246, + "grad_norm": 1.7470401124174042, + "learning_rate": 7.420739966642098e-07, + "loss": 0.2834, + "step": 37690 + }, + { + "epoch": 0.8296500110059432, + "grad_norm": 1.5878720505501256, + "learning_rate": 7.41887147268704e-07, + "loss": 0.2123, + "step": 37691 + }, + { + "epoch": 0.8296720228923619, + "grad_norm": 2.160125432964102, + "learning_rate": 7.41700319514827e-07, + "loss": 0.2592, + "step": 37692 + }, + { + "epoch": 0.8296940347787806, + "grad_norm": 1.5325784784461702, + "learning_rate": 7.415135134035279e-07, + "loss": 0.3491, + "step": 37693 + }, + { + "epoch": 0.8297160466651992, + "grad_norm": 1.943190296148038, + "learning_rate": 7.413267289357578e-07, + "loss": 0.3108, + "step": 37694 + }, + { + "epoch": 0.8297380585516179, + "grad_norm": 1.9716990899522016, + "learning_rate": 7.411399661124646e-07, + "loss": 0.2339, + "step": 37695 + }, + { + "epoch": 0.8297600704380366, + "grad_norm": 3.0146234341049345, + "learning_rate": 7.409532249345963e-07, + "loss": 0.3282, + "step": 37696 + }, + { + "epoch": 0.8297820823244552, + "grad_norm": 1.6355421988790961, + "learning_rate": 7.407665054031038e-07, + "loss": 0.3377, + "step": 37697 + }, + { + "epoch": 0.8298040942108739, + "grad_norm": 1.9285273960946256, + "learning_rate": 7.405798075189352e-07, + "loss": 0.2143, + "step": 37698 + }, + { + "epoch": 0.8298261060972926, + "grad_norm": 2.0183983141701547, + "learning_rate": 7.403931312830409e-07, + "loss": 0.3265, + "step": 37699 + }, + { + "epoch": 0.8298481179837112, + "grad_norm": 5.98608875384367, + "learning_rate": 7.402064766963685e-07, + "loss": 0.2032, + "step": 37700 + }, + { + "epoch": 0.8298701298701299, + "grad_norm": 10.228102681121946, + "learning_rate": 7.400198437598654e-07, + "loss": 0.217, + "step": 37701 + }, + { + "epoch": 0.8298921417565486, + "grad_norm": 5.276143209383088, + "learning_rate": 7.398332324744823e-07, + "loss": 0.2535, + "step": 37702 + }, + { + "epoch": 0.8299141536429672, + "grad_norm": 1.7415295165491582, + "learning_rate": 7.396466428411669e-07, + "loss": 0.275, + "step": 37703 + }, + { + "epoch": 0.8299361655293859, + "grad_norm": 1.816455248119835, + "learning_rate": 7.394600748608671e-07, + "loss": 0.2343, + "step": 37704 + }, + { + "epoch": 0.8299581774158046, + "grad_norm": 1.2552797818279786, + "learning_rate": 7.392735285345321e-07, + "loss": 0.2721, + "step": 37705 + }, + { + "epoch": 0.8299801893022232, + "grad_norm": 0.6248742067464856, + "learning_rate": 7.390870038631082e-07, + "loss": 0.4493, + "step": 37706 + }, + { + "epoch": 0.8300022011886419, + "grad_norm": 1.5201602875979328, + "learning_rate": 7.389005008475447e-07, + "loss": 0.2712, + "step": 37707 + }, + { + "epoch": 0.8300242130750606, + "grad_norm": 1.4112585886073785, + "learning_rate": 7.38714019488791e-07, + "loss": 0.2471, + "step": 37708 + }, + { + "epoch": 0.8300462249614792, + "grad_norm": 0.6065414713263318, + "learning_rate": 7.385275597877912e-07, + "loss": 0.4244, + "step": 37709 + }, + { + "epoch": 0.8300682368478979, + "grad_norm": 1.801732278343985, + "learning_rate": 7.383411217454966e-07, + "loss": 0.3189, + "step": 37710 + }, + { + "epoch": 0.8300902487343166, + "grad_norm": 0.6302015766724758, + "learning_rate": 7.381547053628524e-07, + "loss": 0.4604, + "step": 37711 + }, + { + "epoch": 0.8301122606207352, + "grad_norm": 1.7515406545633556, + "learning_rate": 7.379683106408065e-07, + "loss": 0.2357, + "step": 37712 + }, + { + "epoch": 0.8301342725071539, + "grad_norm": 1.2571992268618988, + "learning_rate": 7.377819375803074e-07, + "loss": 0.2836, + "step": 37713 + }, + { + "epoch": 0.8301562843935725, + "grad_norm": 1.5075321407885247, + "learning_rate": 7.375955861823009e-07, + "loss": 0.2545, + "step": 37714 + }, + { + "epoch": 0.8301782962799912, + "grad_norm": 5.520156565738891, + "learning_rate": 7.374092564477353e-07, + "loss": 0.2799, + "step": 37715 + }, + { + "epoch": 0.8302003081664099, + "grad_norm": 0.5886339024049427, + "learning_rate": 7.372229483775562e-07, + "loss": 0.4262, + "step": 37716 + }, + { + "epoch": 0.8302223200528285, + "grad_norm": 2.2058911363117066, + "learning_rate": 7.370366619727126e-07, + "loss": 0.33, + "step": 37717 + }, + { + "epoch": 0.8302443319392472, + "grad_norm": 4.660420650350935, + "learning_rate": 7.368503972341484e-07, + "loss": 0.3008, + "step": 37718 + }, + { + "epoch": 0.8302663438256659, + "grad_norm": 0.6143397695077625, + "learning_rate": 7.366641541628122e-07, + "loss": 0.463, + "step": 37719 + }, + { + "epoch": 0.8302883557120845, + "grad_norm": 2.772042389922913, + "learning_rate": 7.364779327596516e-07, + "loss": 0.2074, + "step": 37720 + }, + { + "epoch": 0.8303103675985032, + "grad_norm": 1.8255240783724114, + "learning_rate": 7.362917330256103e-07, + "loss": 0.3267, + "step": 37721 + }, + { + "epoch": 0.8303323794849219, + "grad_norm": 3.8209053234662096, + "learning_rate": 7.361055549616375e-07, + "loss": 0.2004, + "step": 37722 + }, + { + "epoch": 0.8303543913713405, + "grad_norm": 1.5192191269210487, + "learning_rate": 7.359193985686763e-07, + "loss": 0.2378, + "step": 37723 + }, + { + "epoch": 0.8303764032577592, + "grad_norm": 1.3861316535804784, + "learning_rate": 7.35733263847675e-07, + "loss": 0.2444, + "step": 37724 + }, + { + "epoch": 0.8303984151441779, + "grad_norm": 1.812220275156266, + "learning_rate": 7.355471507995804e-07, + "loss": 0.2885, + "step": 37725 + }, + { + "epoch": 0.8304204270305965, + "grad_norm": 1.3706962962379237, + "learning_rate": 7.353610594253368e-07, + "loss": 0.1906, + "step": 37726 + }, + { + "epoch": 0.8304424389170152, + "grad_norm": 1.4822458128853555, + "learning_rate": 7.351749897258897e-07, + "loss": 0.2871, + "step": 37727 + }, + { + "epoch": 0.8304644508034339, + "grad_norm": 1.4387154775861923, + "learning_rate": 7.349889417021855e-07, + "loss": 0.2464, + "step": 37728 + }, + { + "epoch": 0.8304864626898525, + "grad_norm": 0.6006984604927941, + "learning_rate": 7.348029153551694e-07, + "loss": 0.4432, + "step": 37729 + }, + { + "epoch": 0.8305084745762712, + "grad_norm": 2.0849875127047044, + "learning_rate": 7.346169106857881e-07, + "loss": 0.2535, + "step": 37730 + }, + { + "epoch": 0.8305304864626899, + "grad_norm": 0.6330637082955486, + "learning_rate": 7.344309276949868e-07, + "loss": 0.4537, + "step": 37731 + }, + { + "epoch": 0.8305524983491085, + "grad_norm": 2.5388448636952132, + "learning_rate": 7.342449663837081e-07, + "loss": 0.2729, + "step": 37732 + }, + { + "epoch": 0.8305745102355272, + "grad_norm": 0.6422227557853615, + "learning_rate": 7.340590267528996e-07, + "loss": 0.451, + "step": 37733 + }, + { + "epoch": 0.8305965221219459, + "grad_norm": 1.5733976497717483, + "learning_rate": 7.338731088035056e-07, + "loss": 0.2483, + "step": 37734 + }, + { + "epoch": 0.8306185340083645, + "grad_norm": 0.6031129465282096, + "learning_rate": 7.336872125364725e-07, + "loss": 0.42, + "step": 37735 + }, + { + "epoch": 0.8306405458947832, + "grad_norm": 1.6686084076005914, + "learning_rate": 7.33501337952744e-07, + "loss": 0.2518, + "step": 37736 + }, + { + "epoch": 0.8306625577812018, + "grad_norm": 1.3586176327691841, + "learning_rate": 7.333154850532631e-07, + "loss": 0.3422, + "step": 37737 + }, + { + "epoch": 0.8306845696676205, + "grad_norm": 19.76349902196182, + "learning_rate": 7.33129653838976e-07, + "loss": 0.2332, + "step": 37738 + }, + { + "epoch": 0.8307065815540392, + "grad_norm": 1.5503751677513227, + "learning_rate": 7.329438443108267e-07, + "loss": 0.3467, + "step": 37739 + }, + { + "epoch": 0.8307285934404578, + "grad_norm": 2.5991390416361564, + "learning_rate": 7.327580564697617e-07, + "loss": 0.2513, + "step": 37740 + }, + { + "epoch": 0.8307506053268765, + "grad_norm": 0.587397532944699, + "learning_rate": 7.325722903167231e-07, + "loss": 0.4318, + "step": 37741 + }, + { + "epoch": 0.8307726172132952, + "grad_norm": 1.7349775372526948, + "learning_rate": 7.323865458526547e-07, + "loss": 0.2804, + "step": 37742 + }, + { + "epoch": 0.8307946290997138, + "grad_norm": 0.6273017796368991, + "learning_rate": 7.322008230785004e-07, + "loss": 0.4436, + "step": 37743 + }, + { + "epoch": 0.8308166409861325, + "grad_norm": 0.6121406093169988, + "learning_rate": 7.320151219952071e-07, + "loss": 0.4379, + "step": 37744 + }, + { + "epoch": 0.8308386528725512, + "grad_norm": 1.9995113616113716, + "learning_rate": 7.318294426037153e-07, + "loss": 0.2885, + "step": 37745 + }, + { + "epoch": 0.8308606647589698, + "grad_norm": 2.3886405811178504, + "learning_rate": 7.316437849049707e-07, + "loss": 0.2344, + "step": 37746 + }, + { + "epoch": 0.8308826766453885, + "grad_norm": 1.6660747197793269, + "learning_rate": 7.31458148899915e-07, + "loss": 0.3046, + "step": 37747 + }, + { + "epoch": 0.8309046885318072, + "grad_norm": 1.688772146882743, + "learning_rate": 7.312725345894928e-07, + "loss": 0.3361, + "step": 37748 + }, + { + "epoch": 0.8309267004182258, + "grad_norm": 3.510022336916522, + "learning_rate": 7.310869419746491e-07, + "loss": 0.2927, + "step": 37749 + }, + { + "epoch": 0.8309487123046445, + "grad_norm": 2.618468159143515, + "learning_rate": 7.309013710563239e-07, + "loss": 0.3397, + "step": 37750 + }, + { + "epoch": 0.8309707241910632, + "grad_norm": 0.6141090823066434, + "learning_rate": 7.307158218354632e-07, + "loss": 0.4339, + "step": 37751 + }, + { + "epoch": 0.8309927360774818, + "grad_norm": 0.6064013876287491, + "learning_rate": 7.305302943130077e-07, + "loss": 0.4203, + "step": 37752 + }, + { + "epoch": 0.8310147479639005, + "grad_norm": 2.0766640832435623, + "learning_rate": 7.303447884899029e-07, + "loss": 0.2824, + "step": 37753 + }, + { + "epoch": 0.8310367598503192, + "grad_norm": 0.6027978835819424, + "learning_rate": 7.30159304367089e-07, + "loss": 0.4466, + "step": 37754 + }, + { + "epoch": 0.8310587717367378, + "grad_norm": 1.963296474820601, + "learning_rate": 7.299738419455099e-07, + "loss": 0.3209, + "step": 37755 + }, + { + "epoch": 0.8310807836231565, + "grad_norm": 1.5490749457358823, + "learning_rate": 7.297884012261091e-07, + "loss": 0.2505, + "step": 37756 + }, + { + "epoch": 0.8311027955095752, + "grad_norm": 0.6099155565890353, + "learning_rate": 7.296029822098272e-07, + "loss": 0.4223, + "step": 37757 + }, + { + "epoch": 0.8311248073959938, + "grad_norm": 1.7777629133425972, + "learning_rate": 7.294175848976088e-07, + "loss": 0.287, + "step": 37758 + }, + { + "epoch": 0.8311468192824125, + "grad_norm": 1.83163841343827, + "learning_rate": 7.292322092903942e-07, + "loss": 0.3576, + "step": 37759 + }, + { + "epoch": 0.8311688311688312, + "grad_norm": 1.6520302685743342, + "learning_rate": 7.290468553891255e-07, + "loss": 0.2943, + "step": 37760 + }, + { + "epoch": 0.8311908430552498, + "grad_norm": 1.4874945624047413, + "learning_rate": 7.288615231947472e-07, + "loss": 0.2767, + "step": 37761 + }, + { + "epoch": 0.8312128549416685, + "grad_norm": 1.7293959610175054, + "learning_rate": 7.286762127081986e-07, + "loss": 0.3489, + "step": 37762 + }, + { + "epoch": 0.8312348668280871, + "grad_norm": 2.688329899959413, + "learning_rate": 7.284909239304233e-07, + "loss": 0.2985, + "step": 37763 + }, + { + "epoch": 0.8312568787145058, + "grad_norm": 3.006473161579606, + "learning_rate": 7.283056568623614e-07, + "loss": 0.4296, + "step": 37764 + }, + { + "epoch": 0.8312788906009245, + "grad_norm": 1.4166617943439213, + "learning_rate": 7.281204115049556e-07, + "loss": 0.312, + "step": 37765 + }, + { + "epoch": 0.8313009024873431, + "grad_norm": 2.037967006629146, + "learning_rate": 7.279351878591479e-07, + "loss": 0.2886, + "step": 37766 + }, + { + "epoch": 0.8313229143737618, + "grad_norm": 10.70275265375722, + "learning_rate": 7.277499859258791e-07, + "loss": 0.3068, + "step": 37767 + }, + { + "epoch": 0.8313449262601805, + "grad_norm": 0.6161855750889416, + "learning_rate": 7.275648057060892e-07, + "loss": 0.4322, + "step": 37768 + }, + { + "epoch": 0.8313669381465991, + "grad_norm": 1.6578371808889596, + "learning_rate": 7.273796472007205e-07, + "loss": 0.2959, + "step": 37769 + }, + { + "epoch": 0.8313889500330178, + "grad_norm": 1.3303131723602923, + "learning_rate": 7.271945104107142e-07, + "loss": 0.2826, + "step": 37770 + }, + { + "epoch": 0.8314109619194365, + "grad_norm": 1.84688805848266, + "learning_rate": 7.270093953370123e-07, + "loss": 0.2566, + "step": 37771 + }, + { + "epoch": 0.8314329738058551, + "grad_norm": 8.185811956738865, + "learning_rate": 7.268243019805543e-07, + "loss": 0.3079, + "step": 37772 + }, + { + "epoch": 0.8314549856922738, + "grad_norm": 1.8381327935692695, + "learning_rate": 7.266392303422798e-07, + "loss": 0.2955, + "step": 37773 + }, + { + "epoch": 0.8314769975786925, + "grad_norm": 2.114190185545867, + "learning_rate": 7.264541804231313e-07, + "loss": 0.3856, + "step": 37774 + }, + { + "epoch": 0.8314990094651111, + "grad_norm": 1.5036920630400168, + "learning_rate": 7.262691522240484e-07, + "loss": 0.3157, + "step": 37775 + }, + { + "epoch": 0.8315210213515298, + "grad_norm": 2.1578707323815096, + "learning_rate": 7.260841457459727e-07, + "loss": 0.3254, + "step": 37776 + }, + { + "epoch": 0.8315430332379485, + "grad_norm": 2.23235279541995, + "learning_rate": 7.258991609898436e-07, + "loss": 0.37, + "step": 37777 + }, + { + "epoch": 0.8315650451243671, + "grad_norm": 1.8614064835567834, + "learning_rate": 7.257141979566002e-07, + "loss": 0.2233, + "step": 37778 + }, + { + "epoch": 0.8315870570107858, + "grad_norm": 1.9629896956170994, + "learning_rate": 7.255292566471839e-07, + "loss": 0.3855, + "step": 37779 + }, + { + "epoch": 0.8316090688972045, + "grad_norm": 1.1356382347411407, + "learning_rate": 7.253443370625341e-07, + "loss": 0.2043, + "step": 37780 + }, + { + "epoch": 0.8316310807836231, + "grad_norm": 4.4844833127766215, + "learning_rate": 7.251594392035921e-07, + "loss": 0.3076, + "step": 37781 + }, + { + "epoch": 0.8316530926700418, + "grad_norm": 0.6236173435508376, + "learning_rate": 7.249745630712968e-07, + "loss": 0.4603, + "step": 37782 + }, + { + "epoch": 0.8316751045564605, + "grad_norm": 0.6035152684838987, + "learning_rate": 7.247897086665856e-07, + "loss": 0.4223, + "step": 37783 + }, + { + "epoch": 0.8316971164428791, + "grad_norm": 1.7915078916211564, + "learning_rate": 7.246048759904006e-07, + "loss": 0.3267, + "step": 37784 + }, + { + "epoch": 0.8317191283292978, + "grad_norm": 1.7872153329314306, + "learning_rate": 7.244200650436816e-07, + "loss": 0.2742, + "step": 37785 + }, + { + "epoch": 0.8317411402157164, + "grad_norm": 1.6485709498450984, + "learning_rate": 7.242352758273652e-07, + "loss": 0.3695, + "step": 37786 + }, + { + "epoch": 0.8317631521021351, + "grad_norm": 1.8930066540427812, + "learning_rate": 7.240505083423938e-07, + "loss": 0.2252, + "step": 37787 + }, + { + "epoch": 0.8317851639885538, + "grad_norm": 2.004017968012229, + "learning_rate": 7.238657625897033e-07, + "loss": 0.2119, + "step": 37788 + }, + { + "epoch": 0.8318071758749724, + "grad_norm": 1.6102352928570334, + "learning_rate": 7.236810385702353e-07, + "loss": 0.2567, + "step": 37789 + }, + { + "epoch": 0.8318291877613911, + "grad_norm": 1.3573846744360942, + "learning_rate": 7.234963362849267e-07, + "loss": 0.2292, + "step": 37790 + }, + { + "epoch": 0.8318511996478098, + "grad_norm": 2.0082443448252247, + "learning_rate": 7.23311655734717e-07, + "loss": 0.3357, + "step": 37791 + }, + { + "epoch": 0.8318732115342284, + "grad_norm": 0.6231556253384253, + "learning_rate": 7.231269969205462e-07, + "loss": 0.4686, + "step": 37792 + }, + { + "epoch": 0.8318952234206471, + "grad_norm": 0.6086185892407293, + "learning_rate": 7.229423598433505e-07, + "loss": 0.4531, + "step": 37793 + }, + { + "epoch": 0.8319172353070658, + "grad_norm": 1.9525837304524014, + "learning_rate": 7.227577445040701e-07, + "loss": 0.3123, + "step": 37794 + }, + { + "epoch": 0.8319392471934844, + "grad_norm": 0.6271211357749058, + "learning_rate": 7.225731509036415e-07, + "loss": 0.4398, + "step": 37795 + }, + { + "epoch": 0.8319612590799031, + "grad_norm": 2.1371594040619715, + "learning_rate": 7.223885790430041e-07, + "loss": 0.2234, + "step": 37796 + }, + { + "epoch": 0.8319832709663219, + "grad_norm": 1.690442382892701, + "learning_rate": 7.22204028923097e-07, + "loss": 0.3346, + "step": 37797 + }, + { + "epoch": 0.8320052828527404, + "grad_norm": 7.641758045180486, + "learning_rate": 7.220195005448555e-07, + "loss": 0.2651, + "step": 37798 + }, + { + "epoch": 0.8320272947391592, + "grad_norm": 0.6089399230396935, + "learning_rate": 7.218349939092201e-07, + "loss": 0.4176, + "step": 37799 + }, + { + "epoch": 0.8320493066255779, + "grad_norm": 4.28349684314524, + "learning_rate": 7.216505090171266e-07, + "loss": 0.3319, + "step": 37800 + }, + { + "epoch": 0.8320713185119964, + "grad_norm": 2.813335825463004, + "learning_rate": 7.214660458695133e-07, + "loss": 0.2669, + "step": 37801 + }, + { + "epoch": 0.8320933303984152, + "grad_norm": 2.2782717417590703, + "learning_rate": 7.212816044673187e-07, + "loss": 0.2928, + "step": 37802 + }, + { + "epoch": 0.8321153422848339, + "grad_norm": 0.6251998962167968, + "learning_rate": 7.210971848114795e-07, + "loss": 0.4403, + "step": 37803 + }, + { + "epoch": 0.8321373541712525, + "grad_norm": 2.377454890941534, + "learning_rate": 7.20912786902932e-07, + "loss": 0.3998, + "step": 37804 + }, + { + "epoch": 0.8321593660576712, + "grad_norm": 1.6704239088571178, + "learning_rate": 7.207284107426138e-07, + "loss": 0.2343, + "step": 37805 + }, + { + "epoch": 0.8321813779440899, + "grad_norm": 1.6211204950428544, + "learning_rate": 7.205440563314625e-07, + "loss": 0.2554, + "step": 37806 + }, + { + "epoch": 0.8322033898305085, + "grad_norm": 1.7629747166423648, + "learning_rate": 7.203597236704163e-07, + "loss": 0.3057, + "step": 37807 + }, + { + "epoch": 0.8322254017169272, + "grad_norm": 2.512888877787028, + "learning_rate": 7.20175412760411e-07, + "loss": 0.3423, + "step": 37808 + }, + { + "epoch": 0.8322474136033458, + "grad_norm": 0.6132513700817845, + "learning_rate": 7.199911236023815e-07, + "loss": 0.4544, + "step": 37809 + }, + { + "epoch": 0.8322694254897645, + "grad_norm": 1.8581178661074604, + "learning_rate": 7.19806856197266e-07, + "loss": 0.3118, + "step": 37810 + }, + { + "epoch": 0.8322914373761832, + "grad_norm": 1.7977209514587131, + "learning_rate": 7.196226105460014e-07, + "loss": 0.2752, + "step": 37811 + }, + { + "epoch": 0.8323134492626018, + "grad_norm": 1.460484937117677, + "learning_rate": 7.194383866495253e-07, + "loss": 0.3094, + "step": 37812 + }, + { + "epoch": 0.8323354611490205, + "grad_norm": 1.6692952595390773, + "learning_rate": 7.192541845087719e-07, + "loss": 0.2829, + "step": 37813 + }, + { + "epoch": 0.8323574730354392, + "grad_norm": 1.698367410193519, + "learning_rate": 7.190700041246767e-07, + "loss": 0.2805, + "step": 37814 + }, + { + "epoch": 0.8323794849218578, + "grad_norm": 4.1507987416060805, + "learning_rate": 7.188858454981773e-07, + "loss": 0.1787, + "step": 37815 + }, + { + "epoch": 0.8324014968082765, + "grad_norm": 2.755273192579132, + "learning_rate": 7.187017086302095e-07, + "loss": 0.2419, + "step": 37816 + }, + { + "epoch": 0.8324235086946952, + "grad_norm": 2.2815620697935293, + "learning_rate": 7.185175935217103e-07, + "loss": 0.2968, + "step": 37817 + }, + { + "epoch": 0.8324455205811138, + "grad_norm": 42.22740858966951, + "learning_rate": 7.183335001736147e-07, + "loss": 0.1927, + "step": 37818 + }, + { + "epoch": 0.8324675324675325, + "grad_norm": 2.4164134360748606, + "learning_rate": 7.181494285868562e-07, + "loss": 0.3763, + "step": 37819 + }, + { + "epoch": 0.8324895443539512, + "grad_norm": 1.3749507383221822, + "learning_rate": 7.179653787623725e-07, + "loss": 0.296, + "step": 37820 + }, + { + "epoch": 0.8325115562403698, + "grad_norm": 2.1889472702115085, + "learning_rate": 7.177813507010983e-07, + "loss": 0.3763, + "step": 37821 + }, + { + "epoch": 0.8325335681267885, + "grad_norm": 1.6601450736283092, + "learning_rate": 7.175973444039703e-07, + "loss": 0.235, + "step": 37822 + }, + { + "epoch": 0.8325555800132072, + "grad_norm": 0.7083924642801085, + "learning_rate": 7.174133598719223e-07, + "loss": 0.4375, + "step": 37823 + }, + { + "epoch": 0.8325775918996258, + "grad_norm": 5.0396989178733165, + "learning_rate": 7.172293971058891e-07, + "loss": 0.3571, + "step": 37824 + }, + { + "epoch": 0.8325996037860445, + "grad_norm": 5.892648109758295, + "learning_rate": 7.170454561068074e-07, + "loss": 0.2838, + "step": 37825 + }, + { + "epoch": 0.8326216156724632, + "grad_norm": 1.5029945218676812, + "learning_rate": 7.168615368756093e-07, + "loss": 0.2157, + "step": 37826 + }, + { + "epoch": 0.8326436275588818, + "grad_norm": 1.6455358204907307, + "learning_rate": 7.166776394132318e-07, + "loss": 0.2995, + "step": 37827 + }, + { + "epoch": 0.8326656394453005, + "grad_norm": 2.0527456606922665, + "learning_rate": 7.164937637206099e-07, + "loss": 0.2396, + "step": 37828 + }, + { + "epoch": 0.8326876513317192, + "grad_norm": 1.7182705398458855, + "learning_rate": 7.163099097986759e-07, + "loss": 0.3718, + "step": 37829 + }, + { + "epoch": 0.8327096632181378, + "grad_norm": 1.6438615423315546, + "learning_rate": 7.161260776483669e-07, + "loss": 0.3104, + "step": 37830 + }, + { + "epoch": 0.8327316751045565, + "grad_norm": 0.6169297051999875, + "learning_rate": 7.159422672706145e-07, + "loss": 0.4498, + "step": 37831 + }, + { + "epoch": 0.8327536869909751, + "grad_norm": 0.6088577849878627, + "learning_rate": 7.157584786663541e-07, + "loss": 0.4529, + "step": 37832 + }, + { + "epoch": 0.8327756988773938, + "grad_norm": 0.6385069728930578, + "learning_rate": 7.155747118365214e-07, + "loss": 0.4708, + "step": 37833 + }, + { + "epoch": 0.8327977107638125, + "grad_norm": 1.8282265790548242, + "learning_rate": 7.153909667820474e-07, + "loss": 0.3579, + "step": 37834 + }, + { + "epoch": 0.8328197226502311, + "grad_norm": 1.8983067737607644, + "learning_rate": 7.152072435038693e-07, + "loss": 0.3108, + "step": 37835 + }, + { + "epoch": 0.8328417345366498, + "grad_norm": 1.6621104921602063, + "learning_rate": 7.150235420029173e-07, + "loss": 0.2671, + "step": 37836 + }, + { + "epoch": 0.8328637464230685, + "grad_norm": 2.2018229824598556, + "learning_rate": 7.148398622801272e-07, + "loss": 0.3255, + "step": 37837 + }, + { + "epoch": 0.8328857583094871, + "grad_norm": 1.6524893532869565, + "learning_rate": 7.14656204336433e-07, + "loss": 0.2384, + "step": 37838 + }, + { + "epoch": 0.8329077701959058, + "grad_norm": 1.6892542351352404, + "learning_rate": 7.14472568172766e-07, + "loss": 0.2023, + "step": 37839 + }, + { + "epoch": 0.8329297820823245, + "grad_norm": 0.6003320428065548, + "learning_rate": 7.142889537900622e-07, + "loss": 0.4475, + "step": 37840 + }, + { + "epoch": 0.8329517939687431, + "grad_norm": 1.7104306303618078, + "learning_rate": 7.141053611892523e-07, + "loss": 0.2496, + "step": 37841 + }, + { + "epoch": 0.8329738058551618, + "grad_norm": 1.5704907447693148, + "learning_rate": 7.139217903712709e-07, + "loss": 0.2831, + "step": 37842 + }, + { + "epoch": 0.8329958177415805, + "grad_norm": 1.96981228718893, + "learning_rate": 7.137382413370508e-07, + "loss": 0.2774, + "step": 37843 + }, + { + "epoch": 0.8330178296279991, + "grad_norm": 1.6789248230052531, + "learning_rate": 7.135547140875254e-07, + "loss": 0.2652, + "step": 37844 + }, + { + "epoch": 0.8330398415144178, + "grad_norm": 1.3671467616242619, + "learning_rate": 7.133712086236255e-07, + "loss": 0.2663, + "step": 37845 + }, + { + "epoch": 0.8330618534008365, + "grad_norm": 2.790513441738366, + "learning_rate": 7.131877249462848e-07, + "loss": 0.283, + "step": 37846 + }, + { + "epoch": 0.8330838652872551, + "grad_norm": 1.3416282700517648, + "learning_rate": 7.130042630564365e-07, + "loss": 0.2811, + "step": 37847 + }, + { + "epoch": 0.8331058771736738, + "grad_norm": 1.62401815007883, + "learning_rate": 7.128208229550133e-07, + "loss": 0.267, + "step": 37848 + }, + { + "epoch": 0.8331278890600925, + "grad_norm": 1.4590862285897077, + "learning_rate": 7.126374046429468e-07, + "loss": 0.2903, + "step": 37849 + }, + { + "epoch": 0.8331499009465111, + "grad_norm": 1.6816211598390767, + "learning_rate": 7.124540081211684e-07, + "loss": 0.3183, + "step": 37850 + }, + { + "epoch": 0.8331719128329298, + "grad_norm": 1.7148745482116206, + "learning_rate": 7.12270633390611e-07, + "loss": 0.2641, + "step": 37851 + }, + { + "epoch": 0.8331939247193485, + "grad_norm": 2.397480011264448, + "learning_rate": 7.120872804522061e-07, + "loss": 0.2747, + "step": 37852 + }, + { + "epoch": 0.8332159366057671, + "grad_norm": 1.3250908538950958, + "learning_rate": 7.11903949306888e-07, + "loss": 0.3063, + "step": 37853 + }, + { + "epoch": 0.8332379484921858, + "grad_norm": 1.8466891801112955, + "learning_rate": 7.11720639955586e-07, + "loss": 0.2804, + "step": 37854 + }, + { + "epoch": 0.8332599603786044, + "grad_norm": 0.6049985768652493, + "learning_rate": 7.115373523992314e-07, + "loss": 0.4248, + "step": 37855 + }, + { + "epoch": 0.8332819722650231, + "grad_norm": 1.9677290988765461, + "learning_rate": 7.113540866387564e-07, + "loss": 0.393, + "step": 37856 + }, + { + "epoch": 0.8333039841514418, + "grad_norm": 2.4084072240086334, + "learning_rate": 7.11170842675093e-07, + "loss": 0.2987, + "step": 37857 + }, + { + "epoch": 0.8333259960378604, + "grad_norm": 1.7032701688218461, + "learning_rate": 7.109876205091737e-07, + "loss": 0.3308, + "step": 37858 + }, + { + "epoch": 0.8333480079242791, + "grad_norm": 2.2401543839317526, + "learning_rate": 7.108044201419279e-07, + "loss": 0.3442, + "step": 37859 + }, + { + "epoch": 0.8333700198106978, + "grad_norm": 0.614917889983723, + "learning_rate": 7.106212415742858e-07, + "loss": 0.4617, + "step": 37860 + }, + { + "epoch": 0.8333920316971164, + "grad_norm": 1.4463276403051049, + "learning_rate": 7.104380848071812e-07, + "loss": 0.2511, + "step": 37861 + }, + { + "epoch": 0.8334140435835351, + "grad_norm": 2.3962191864681794, + "learning_rate": 7.102549498415406e-07, + "loss": 0.3502, + "step": 37862 + }, + { + "epoch": 0.8334360554699538, + "grad_norm": 1.8799793027809881, + "learning_rate": 7.100718366783005e-07, + "loss": 0.3042, + "step": 37863 + }, + { + "epoch": 0.8334580673563724, + "grad_norm": 0.6321982104608603, + "learning_rate": 7.098887453183884e-07, + "loss": 0.4533, + "step": 37864 + }, + { + "epoch": 0.8334800792427911, + "grad_norm": 1.5616438899183898, + "learning_rate": 7.097056757627341e-07, + "loss": 0.2508, + "step": 37865 + }, + { + "epoch": 0.8335020911292098, + "grad_norm": 2.3669649994935567, + "learning_rate": 7.095226280122702e-07, + "loss": 0.2897, + "step": 37866 + }, + { + "epoch": 0.8335241030156284, + "grad_norm": 2.7783564324105434, + "learning_rate": 7.093396020679244e-07, + "loss": 0.3693, + "step": 37867 + }, + { + "epoch": 0.8335461149020471, + "grad_norm": 0.6146450941890905, + "learning_rate": 7.091565979306292e-07, + "loss": 0.4554, + "step": 37868 + }, + { + "epoch": 0.8335681267884658, + "grad_norm": 1.6190305384403008, + "learning_rate": 7.089736156013144e-07, + "loss": 0.282, + "step": 37869 + }, + { + "epoch": 0.8335901386748844, + "grad_norm": 1.7577293081121632, + "learning_rate": 7.087906550809087e-07, + "loss": 0.2922, + "step": 37870 + }, + { + "epoch": 0.8336121505613031, + "grad_norm": 0.617775990876572, + "learning_rate": 7.086077163703437e-07, + "loss": 0.4591, + "step": 37871 + }, + { + "epoch": 0.8336341624477218, + "grad_norm": 1.6787295995834937, + "learning_rate": 7.084247994705473e-07, + "loss": 0.3394, + "step": 37872 + }, + { + "epoch": 0.8336561743341404, + "grad_norm": 0.612548056109152, + "learning_rate": 7.082419043824506e-07, + "loss": 0.46, + "step": 37873 + }, + { + "epoch": 0.8336781862205591, + "grad_norm": 1.9573088887284855, + "learning_rate": 7.080590311069835e-07, + "loss": 0.3164, + "step": 37874 + }, + { + "epoch": 0.8337001981069778, + "grad_norm": 11.308436481778717, + "learning_rate": 7.078761796450734e-07, + "loss": 0.3194, + "step": 37875 + }, + { + "epoch": 0.8337222099933964, + "grad_norm": 1.8611517881389825, + "learning_rate": 7.076933499976518e-07, + "loss": 0.2628, + "step": 37876 + }, + { + "epoch": 0.8337442218798151, + "grad_norm": 2.905829037568337, + "learning_rate": 7.075105421656464e-07, + "loss": 0.3733, + "step": 37877 + }, + { + "epoch": 0.8337662337662337, + "grad_norm": 1.8121331939992484, + "learning_rate": 7.07327756149987e-07, + "loss": 0.3272, + "step": 37878 + }, + { + "epoch": 0.8337882456526524, + "grad_norm": 3.2982343996858354, + "learning_rate": 7.071449919516032e-07, + "loss": 0.2887, + "step": 37879 + }, + { + "epoch": 0.8338102575390711, + "grad_norm": 1.5107362710391896, + "learning_rate": 7.069622495714223e-07, + "loss": 0.2506, + "step": 37880 + }, + { + "epoch": 0.8338322694254897, + "grad_norm": 0.607805225695727, + "learning_rate": 7.067795290103751e-07, + "loss": 0.4337, + "step": 37881 + }, + { + "epoch": 0.8338542813119084, + "grad_norm": 3.8058693506522827, + "learning_rate": 7.065968302693882e-07, + "loss": 0.2133, + "step": 37882 + }, + { + "epoch": 0.8338762931983271, + "grad_norm": 0.6334436705461175, + "learning_rate": 7.064141533493906e-07, + "loss": 0.4428, + "step": 37883 + }, + { + "epoch": 0.8338983050847457, + "grad_norm": 1.6520757052527304, + "learning_rate": 7.06231498251313e-07, + "loss": 0.2581, + "step": 37884 + }, + { + "epoch": 0.8339203169711644, + "grad_norm": 1.4513990829778152, + "learning_rate": 7.060488649760816e-07, + "loss": 0.3791, + "step": 37885 + }, + { + "epoch": 0.8339423288575831, + "grad_norm": 0.6078798004210242, + "learning_rate": 7.058662535246241e-07, + "loss": 0.4446, + "step": 37886 + }, + { + "epoch": 0.8339643407440017, + "grad_norm": 1.6170297430404639, + "learning_rate": 7.056836638978698e-07, + "loss": 0.2778, + "step": 37887 + }, + { + "epoch": 0.8339863526304204, + "grad_norm": 2.408192736053973, + "learning_rate": 7.055010960967462e-07, + "loss": 0.2774, + "step": 37888 + }, + { + "epoch": 0.8340083645168391, + "grad_norm": 2.012666546435694, + "learning_rate": 7.053185501221826e-07, + "loss": 0.385, + "step": 37889 + }, + { + "epoch": 0.8340303764032577, + "grad_norm": 3.3726508528887664, + "learning_rate": 7.051360259751056e-07, + "loss": 0.2994, + "step": 37890 + }, + { + "epoch": 0.8340523882896764, + "grad_norm": 1.5378397070260506, + "learning_rate": 7.049535236564414e-07, + "loss": 0.2416, + "step": 37891 + }, + { + "epoch": 0.8340744001760951, + "grad_norm": 3.1857816910853254, + "learning_rate": 7.047710431671195e-07, + "loss": 0.2057, + "step": 37892 + }, + { + "epoch": 0.8340964120625137, + "grad_norm": 0.6189036142391783, + "learning_rate": 7.045885845080663e-07, + "loss": 0.4632, + "step": 37893 + }, + { + "epoch": 0.8341184239489324, + "grad_norm": 1.6184284773903628, + "learning_rate": 7.044061476802117e-07, + "loss": 0.2538, + "step": 37894 + }, + { + "epoch": 0.8341404358353511, + "grad_norm": 1.4477091470707986, + "learning_rate": 7.042237326844803e-07, + "loss": 0.3758, + "step": 37895 + }, + { + "epoch": 0.8341624477217697, + "grad_norm": 3.6408482719852584, + "learning_rate": 7.040413395217987e-07, + "loss": 0.3109, + "step": 37896 + }, + { + "epoch": 0.8341844596081884, + "grad_norm": 2.3998185449439133, + "learning_rate": 7.038589681930963e-07, + "loss": 0.2971, + "step": 37897 + }, + { + "epoch": 0.8342064714946071, + "grad_norm": 0.6264379308459159, + "learning_rate": 7.036766186992966e-07, + "loss": 0.4435, + "step": 37898 + }, + { + "epoch": 0.8342284833810257, + "grad_norm": 1.4165011058526336, + "learning_rate": 7.034942910413311e-07, + "loss": 0.3182, + "step": 37899 + }, + { + "epoch": 0.8342504952674444, + "grad_norm": 2.786422362456692, + "learning_rate": 7.033119852201237e-07, + "loss": 0.2646, + "step": 37900 + }, + { + "epoch": 0.8342725071538631, + "grad_norm": 1.650036933414281, + "learning_rate": 7.031297012365995e-07, + "loss": 0.2615, + "step": 37901 + }, + { + "epoch": 0.8342945190402817, + "grad_norm": 0.6258703094539075, + "learning_rate": 7.029474390916885e-07, + "loss": 0.4511, + "step": 37902 + }, + { + "epoch": 0.8343165309267004, + "grad_norm": 1.8433466334455584, + "learning_rate": 7.027651987863137e-07, + "loss": 0.3418, + "step": 37903 + }, + { + "epoch": 0.834338542813119, + "grad_norm": 1.4986609378748115, + "learning_rate": 7.025829803214029e-07, + "loss": 0.2731, + "step": 37904 + }, + { + "epoch": 0.8343605546995377, + "grad_norm": 0.5881803936432617, + "learning_rate": 7.02400783697883e-07, + "loss": 0.4397, + "step": 37905 + }, + { + "epoch": 0.8343825665859564, + "grad_norm": 1.550097488596798, + "learning_rate": 7.022186089166783e-07, + "loss": 0.2561, + "step": 37906 + }, + { + "epoch": 0.834404578472375, + "grad_norm": 1.7834828189319565, + "learning_rate": 7.020364559787163e-07, + "loss": 0.297, + "step": 37907 + }, + { + "epoch": 0.8344265903587937, + "grad_norm": 1.8433524853066439, + "learning_rate": 7.018543248849207e-07, + "loss": 0.3294, + "step": 37908 + }, + { + "epoch": 0.8344486022452124, + "grad_norm": 1.38578216813666, + "learning_rate": 7.016722156362188e-07, + "loss": 0.2217, + "step": 37909 + }, + { + "epoch": 0.834470614131631, + "grad_norm": 1.1612053554411543, + "learning_rate": 7.014901282335368e-07, + "loss": 0.2241, + "step": 37910 + }, + { + "epoch": 0.8344926260180497, + "grad_norm": 1.585904954124759, + "learning_rate": 7.01308062677798e-07, + "loss": 0.2524, + "step": 37911 + }, + { + "epoch": 0.8345146379044684, + "grad_norm": 0.5948099096846673, + "learning_rate": 7.011260189699304e-07, + "loss": 0.4433, + "step": 37912 + }, + { + "epoch": 0.834536649790887, + "grad_norm": 1.930940062260006, + "learning_rate": 7.009439971108561e-07, + "loss": 0.2786, + "step": 37913 + }, + { + "epoch": 0.8345586616773057, + "grad_norm": 1.6143281931103397, + "learning_rate": 7.007619971015023e-07, + "loss": 0.2636, + "step": 37914 + }, + { + "epoch": 0.8345806735637245, + "grad_norm": 1.7523784033928091, + "learning_rate": 7.005800189427941e-07, + "loss": 0.4962, + "step": 37915 + }, + { + "epoch": 0.834602685450143, + "grad_norm": 1.4685814462685338, + "learning_rate": 7.003980626356549e-07, + "loss": 0.2564, + "step": 37916 + }, + { + "epoch": 0.8346246973365617, + "grad_norm": 1.683276054343472, + "learning_rate": 7.002161281810116e-07, + "loss": 0.3128, + "step": 37917 + }, + { + "epoch": 0.8346467092229805, + "grad_norm": 1.741105791652074, + "learning_rate": 7.000342155797867e-07, + "loss": 0.2635, + "step": 37918 + }, + { + "epoch": 0.834668721109399, + "grad_norm": 0.5911692195470175, + "learning_rate": 6.998523248329058e-07, + "loss": 0.4218, + "step": 37919 + }, + { + "epoch": 0.8346907329958178, + "grad_norm": 1.4781781806311003, + "learning_rate": 6.996704559412942e-07, + "loss": 0.2873, + "step": 37920 + }, + { + "epoch": 0.8347127448822365, + "grad_norm": 1.6861264974752308, + "learning_rate": 6.994886089058744e-07, + "loss": 0.3203, + "step": 37921 + }, + { + "epoch": 0.834734756768655, + "grad_norm": 1.9869696096495861, + "learning_rate": 6.99306783727573e-07, + "loss": 0.3858, + "step": 37922 + }, + { + "epoch": 0.8347567686550738, + "grad_norm": 1.7038926947271311, + "learning_rate": 6.991249804073108e-07, + "loss": 0.2415, + "step": 37923 + }, + { + "epoch": 0.8347787805414925, + "grad_norm": 2.1632146355934694, + "learning_rate": 6.989431989460143e-07, + "loss": 0.3471, + "step": 37924 + }, + { + "epoch": 0.834800792427911, + "grad_norm": 2.1282766805534794, + "learning_rate": 6.987614393446074e-07, + "loss": 0.2942, + "step": 37925 + }, + { + "epoch": 0.8348228043143298, + "grad_norm": 2.671314045193448, + "learning_rate": 6.985797016040136e-07, + "loss": 0.2726, + "step": 37926 + }, + { + "epoch": 0.8348448162007484, + "grad_norm": 2.2862088907915155, + "learning_rate": 6.983979857251554e-07, + "loss": 0.3006, + "step": 37927 + }, + { + "epoch": 0.8348668280871671, + "grad_norm": 2.0590740488406447, + "learning_rate": 6.98216291708958e-07, + "loss": 0.3708, + "step": 37928 + }, + { + "epoch": 0.8348888399735858, + "grad_norm": 2.1244054991362513, + "learning_rate": 6.980346195563414e-07, + "loss": 0.3807, + "step": 37929 + }, + { + "epoch": 0.8349108518600044, + "grad_norm": 1.8320799595001311, + "learning_rate": 6.978529692682346e-07, + "loss": 0.2165, + "step": 37930 + }, + { + "epoch": 0.8349328637464231, + "grad_norm": 0.9934528397537359, + "learning_rate": 6.976713408455576e-07, + "loss": 0.1943, + "step": 37931 + }, + { + "epoch": 0.8349548756328418, + "grad_norm": 1.6327700271667667, + "learning_rate": 6.974897342892322e-07, + "loss": 0.3319, + "step": 37932 + }, + { + "epoch": 0.8349768875192604, + "grad_norm": 1.4219784795850425, + "learning_rate": 6.973081496001844e-07, + "loss": 0.2793, + "step": 37933 + }, + { + "epoch": 0.8349988994056791, + "grad_norm": 1.4239032440535482, + "learning_rate": 6.971265867793337e-07, + "loss": 0.298, + "step": 37934 + }, + { + "epoch": 0.8350209112920978, + "grad_norm": 0.6287931951171647, + "learning_rate": 6.969450458276066e-07, + "loss": 0.45, + "step": 37935 + }, + { + "epoch": 0.8350429231785164, + "grad_norm": 0.6334782226023359, + "learning_rate": 6.967635267459239e-07, + "loss": 0.4477, + "step": 37936 + }, + { + "epoch": 0.8350649350649351, + "grad_norm": 2.668212147090157, + "learning_rate": 6.965820295352077e-07, + "loss": 0.3941, + "step": 37937 + }, + { + "epoch": 0.8350869469513538, + "grad_norm": 1.7167209820332991, + "learning_rate": 6.964005541963814e-07, + "loss": 0.1962, + "step": 37938 + }, + { + "epoch": 0.8351089588377724, + "grad_norm": 0.6234237782847946, + "learning_rate": 6.962191007303665e-07, + "loss": 0.4602, + "step": 37939 + }, + { + "epoch": 0.8351309707241911, + "grad_norm": 1.8651336407304409, + "learning_rate": 6.960376691380855e-07, + "loss": 0.2992, + "step": 37940 + }, + { + "epoch": 0.8351529826106098, + "grad_norm": 1.4435265716107244, + "learning_rate": 6.958562594204616e-07, + "loss": 0.2738, + "step": 37941 + }, + { + "epoch": 0.8351749944970284, + "grad_norm": 2.1591307091621763, + "learning_rate": 6.956748715784145e-07, + "loss": 0.3545, + "step": 37942 + }, + { + "epoch": 0.8351970063834471, + "grad_norm": 1.550173880697123, + "learning_rate": 6.954935056128692e-07, + "loss": 0.286, + "step": 37943 + }, + { + "epoch": 0.8352190182698658, + "grad_norm": 6.696910890592819, + "learning_rate": 6.953121615247443e-07, + "loss": 0.3031, + "step": 37944 + }, + { + "epoch": 0.8352410301562844, + "grad_norm": 2.4758915219612123, + "learning_rate": 6.95130839314963e-07, + "loss": 0.2827, + "step": 37945 + }, + { + "epoch": 0.8352630420427031, + "grad_norm": 1.4363443105462423, + "learning_rate": 6.94949538984448e-07, + "loss": 0.3631, + "step": 37946 + }, + { + "epoch": 0.8352850539291218, + "grad_norm": 1.7321518315900006, + "learning_rate": 6.947682605341183e-07, + "loss": 0.2563, + "step": 37947 + }, + { + "epoch": 0.8353070658155404, + "grad_norm": 1.9622146734370287, + "learning_rate": 6.945870039648977e-07, + "loss": 0.3303, + "step": 37948 + }, + { + "epoch": 0.8353290777019591, + "grad_norm": 1.888714955150354, + "learning_rate": 6.944057692777051e-07, + "loss": 0.3492, + "step": 37949 + }, + { + "epoch": 0.8353510895883777, + "grad_norm": 1.6015107936489605, + "learning_rate": 6.942245564734629e-07, + "loss": 0.2413, + "step": 37950 + }, + { + "epoch": 0.8353731014747964, + "grad_norm": 2.0734395885749786, + "learning_rate": 6.940433655530926e-07, + "loss": 0.2417, + "step": 37951 + }, + { + "epoch": 0.8353951133612151, + "grad_norm": 1.2621237461033648, + "learning_rate": 6.938621965175135e-07, + "loss": 0.2265, + "step": 37952 + }, + { + "epoch": 0.8354171252476337, + "grad_norm": 1.3740045471377245, + "learning_rate": 6.936810493676488e-07, + "loss": 0.2497, + "step": 37953 + }, + { + "epoch": 0.8354391371340524, + "grad_norm": 1.710781655192146, + "learning_rate": 6.93499924104416e-07, + "loss": 0.285, + "step": 37954 + }, + { + "epoch": 0.8354611490204711, + "grad_norm": 5.943335422550424, + "learning_rate": 6.933188207287378e-07, + "loss": 0.3292, + "step": 37955 + }, + { + "epoch": 0.8354831609068897, + "grad_norm": 2.2509990482385653, + "learning_rate": 6.93137739241535e-07, + "loss": 0.3001, + "step": 37956 + }, + { + "epoch": 0.8355051727933084, + "grad_norm": 1.4091450093639086, + "learning_rate": 6.929566796437265e-07, + "loss": 0.3349, + "step": 37957 + }, + { + "epoch": 0.8355271846797271, + "grad_norm": 2.0690237825665148, + "learning_rate": 6.927756419362336e-07, + "loss": 0.2633, + "step": 37958 + }, + { + "epoch": 0.8355491965661457, + "grad_norm": 2.149097532301401, + "learning_rate": 6.92594626119975e-07, + "loss": 0.2854, + "step": 37959 + }, + { + "epoch": 0.8355712084525644, + "grad_norm": 2.047470833377808, + "learning_rate": 6.924136321958719e-07, + "loss": 0.3521, + "step": 37960 + }, + { + "epoch": 0.8355932203389831, + "grad_norm": 1.961934715741748, + "learning_rate": 6.922326601648449e-07, + "loss": 0.2752, + "step": 37961 + }, + { + "epoch": 0.8356152322254017, + "grad_norm": 1.6224945998238312, + "learning_rate": 6.920517100278118e-07, + "loss": 0.3574, + "step": 37962 + }, + { + "epoch": 0.8356372441118204, + "grad_norm": 3.174187905184407, + "learning_rate": 6.918707817856945e-07, + "loss": 0.2582, + "step": 37963 + }, + { + "epoch": 0.8356592559982391, + "grad_norm": 1.9653148740429724, + "learning_rate": 6.916898754394114e-07, + "loss": 0.2617, + "step": 37964 + }, + { + "epoch": 0.8356812678846577, + "grad_norm": 0.6088534259885806, + "learning_rate": 6.915089909898792e-07, + "loss": 0.4439, + "step": 37965 + }, + { + "epoch": 0.8357032797710764, + "grad_norm": 1.8659633274271945, + "learning_rate": 6.91328128438023e-07, + "loss": 0.2918, + "step": 37966 + }, + { + "epoch": 0.8357252916574951, + "grad_norm": 1.8293243187683441, + "learning_rate": 6.911472877847581e-07, + "loss": 0.1998, + "step": 37967 + }, + { + "epoch": 0.8357473035439137, + "grad_norm": 1.8042776409784806, + "learning_rate": 6.909664690310042e-07, + "loss": 0.3312, + "step": 37968 + }, + { + "epoch": 0.8357693154303324, + "grad_norm": 2.230139024729026, + "learning_rate": 6.907856721776812e-07, + "loss": 0.293, + "step": 37969 + }, + { + "epoch": 0.8357913273167511, + "grad_norm": 0.6216258232919423, + "learning_rate": 6.906048972257051e-07, + "loss": 0.4623, + "step": 37970 + }, + { + "epoch": 0.8358133392031697, + "grad_norm": 1.8864074329589133, + "learning_rate": 6.904241441759996e-07, + "loss": 0.3405, + "step": 37971 + }, + { + "epoch": 0.8358353510895884, + "grad_norm": 1.6920756031461215, + "learning_rate": 6.902434130294799e-07, + "loss": 0.2877, + "step": 37972 + }, + { + "epoch": 0.835857362976007, + "grad_norm": 1.5676498532101621, + "learning_rate": 6.90062703787065e-07, + "loss": 0.2892, + "step": 37973 + }, + { + "epoch": 0.8358793748624257, + "grad_norm": 1.3974381560309455, + "learning_rate": 6.89882016449675e-07, + "loss": 0.3349, + "step": 37974 + }, + { + "epoch": 0.8359013867488444, + "grad_norm": 1.6028409841343383, + "learning_rate": 6.897013510182255e-07, + "loss": 0.2879, + "step": 37975 + }, + { + "epoch": 0.835923398635263, + "grad_norm": 1.3724292259252828, + "learning_rate": 6.89520707493636e-07, + "loss": 0.3168, + "step": 37976 + }, + { + "epoch": 0.8359454105216817, + "grad_norm": 3.198507003367095, + "learning_rate": 6.893400858768262e-07, + "loss": 0.4005, + "step": 37977 + }, + { + "epoch": 0.8359674224081004, + "grad_norm": 1.8303062056946258, + "learning_rate": 6.891594861687112e-07, + "loss": 0.2768, + "step": 37978 + }, + { + "epoch": 0.835989434294519, + "grad_norm": 1.91601388293638, + "learning_rate": 6.88978908370212e-07, + "loss": 0.2942, + "step": 37979 + }, + { + "epoch": 0.8360114461809377, + "grad_norm": 1.7258886885949596, + "learning_rate": 6.887983524822433e-07, + "loss": 0.2689, + "step": 37980 + }, + { + "epoch": 0.8360334580673564, + "grad_norm": 1.9351280861163211, + "learning_rate": 6.886178185057241e-07, + "loss": 0.4125, + "step": 37981 + }, + { + "epoch": 0.836055469953775, + "grad_norm": 1.6884066535416227, + "learning_rate": 6.884373064415734e-07, + "loss": 0.2477, + "step": 37982 + }, + { + "epoch": 0.8360774818401937, + "grad_norm": 1.9922472647443474, + "learning_rate": 6.88256816290706e-07, + "loss": 0.3169, + "step": 37983 + }, + { + "epoch": 0.8360994937266124, + "grad_norm": 1.8239785373381392, + "learning_rate": 6.88076348054042e-07, + "loss": 0.3135, + "step": 37984 + }, + { + "epoch": 0.836121505613031, + "grad_norm": 1.6824392502701797, + "learning_rate": 6.87895901732496e-07, + "loss": 0.3606, + "step": 37985 + }, + { + "epoch": 0.8361435174994497, + "grad_norm": 2.024962717742759, + "learning_rate": 6.877154773269862e-07, + "loss": 0.2584, + "step": 37986 + }, + { + "epoch": 0.8361655293858684, + "grad_norm": 1.694905464424518, + "learning_rate": 6.875350748384307e-07, + "loss": 0.2513, + "step": 37987 + }, + { + "epoch": 0.836187541272287, + "grad_norm": 2.090354205288286, + "learning_rate": 6.873546942677445e-07, + "loss": 0.1832, + "step": 37988 + }, + { + "epoch": 0.8362095531587057, + "grad_norm": 1.5834970855432335, + "learning_rate": 6.871743356158461e-07, + "loss": 0.2579, + "step": 37989 + }, + { + "epoch": 0.8362315650451244, + "grad_norm": 1.1768001784695712, + "learning_rate": 6.869939988836505e-07, + "loss": 0.3175, + "step": 37990 + }, + { + "epoch": 0.836253576931543, + "grad_norm": 0.6250307806547152, + "learning_rate": 6.868136840720746e-07, + "loss": 0.4454, + "step": 37991 + }, + { + "epoch": 0.8362755888179617, + "grad_norm": 1.8438772375141452, + "learning_rate": 6.866333911820366e-07, + "loss": 0.2478, + "step": 37992 + }, + { + "epoch": 0.8362976007043804, + "grad_norm": 1.6811831958698067, + "learning_rate": 6.864531202144508e-07, + "loss": 0.3276, + "step": 37993 + }, + { + "epoch": 0.836319612590799, + "grad_norm": 1.200719192160001, + "learning_rate": 6.862728711702349e-07, + "loss": 0.3017, + "step": 37994 + }, + { + "epoch": 0.8363416244772177, + "grad_norm": 1.4721979368829288, + "learning_rate": 6.860926440503029e-07, + "loss": 0.2028, + "step": 37995 + }, + { + "epoch": 0.8363636363636363, + "grad_norm": 0.6135343352078102, + "learning_rate": 6.859124388555727e-07, + "loss": 0.4486, + "step": 37996 + }, + { + "epoch": 0.836385648250055, + "grad_norm": 1.8515310253060562, + "learning_rate": 6.857322555869605e-07, + "loss": 0.3602, + "step": 37997 + }, + { + "epoch": 0.8364076601364737, + "grad_norm": 1.297338689349803, + "learning_rate": 6.855520942453797e-07, + "loss": 0.2379, + "step": 37998 + }, + { + "epoch": 0.8364296720228923, + "grad_norm": 1.7822424856697658, + "learning_rate": 6.853719548317489e-07, + "loss": 0.2763, + "step": 37999 + }, + { + "epoch": 0.836451683909311, + "grad_norm": 1.4982885865803321, + "learning_rate": 6.851918373469823e-07, + "loss": 0.2579, + "step": 38000 + }, + { + "epoch": 0.8364736957957297, + "grad_norm": 1.844958330102164, + "learning_rate": 6.850117417919927e-07, + "loss": 0.2653, + "step": 38001 + }, + { + "epoch": 0.8364957076821483, + "grad_norm": 1.4112296405600466, + "learning_rate": 6.848316681677003e-07, + "loss": 0.2294, + "step": 38002 + }, + { + "epoch": 0.836517719568567, + "grad_norm": 1.9329492670612016, + "learning_rate": 6.846516164750183e-07, + "loss": 0.332, + "step": 38003 + }, + { + "epoch": 0.8365397314549857, + "grad_norm": 1.8130587631444481, + "learning_rate": 6.844715867148599e-07, + "loss": 0.2743, + "step": 38004 + }, + { + "epoch": 0.8365617433414043, + "grad_norm": 0.6379801826154702, + "learning_rate": 6.842915788881433e-07, + "loss": 0.4573, + "step": 38005 + }, + { + "epoch": 0.836583755227823, + "grad_norm": 1.8066864003177723, + "learning_rate": 6.841115929957792e-07, + "loss": 0.3472, + "step": 38006 + }, + { + "epoch": 0.8366057671142417, + "grad_norm": 4.571019615984287, + "learning_rate": 6.839316290386872e-07, + "loss": 0.3157, + "step": 38007 + }, + { + "epoch": 0.8366277790006603, + "grad_norm": 0.6024859488582663, + "learning_rate": 6.837516870177796e-07, + "loss": 0.4462, + "step": 38008 + }, + { + "epoch": 0.836649790887079, + "grad_norm": 0.6072717842727849, + "learning_rate": 6.835717669339703e-07, + "loss": 0.418, + "step": 38009 + }, + { + "epoch": 0.8366718027734977, + "grad_norm": 1.4843104800017974, + "learning_rate": 6.833918687881752e-07, + "loss": 0.253, + "step": 38010 + }, + { + "epoch": 0.8366938146599163, + "grad_norm": 1.6414864215270037, + "learning_rate": 6.832119925813069e-07, + "loss": 0.2522, + "step": 38011 + }, + { + "epoch": 0.836715826546335, + "grad_norm": 1.4858914542929798, + "learning_rate": 6.830321383142802e-07, + "loss": 0.217, + "step": 38012 + }, + { + "epoch": 0.8367378384327537, + "grad_norm": 1.8532748940322148, + "learning_rate": 6.82852305988011e-07, + "loss": 0.2861, + "step": 38013 + }, + { + "epoch": 0.8367598503191723, + "grad_norm": 1.386325279923115, + "learning_rate": 6.826724956034108e-07, + "loss": 0.295, + "step": 38014 + }, + { + "epoch": 0.836781862205591, + "grad_norm": 2.6576753710812606, + "learning_rate": 6.824927071613957e-07, + "loss": 0.2846, + "step": 38015 + }, + { + "epoch": 0.8368038740920097, + "grad_norm": 0.6161481317917757, + "learning_rate": 6.823129406628776e-07, + "loss": 0.4528, + "step": 38016 + }, + { + "epoch": 0.8368258859784283, + "grad_norm": 2.2184905856614505, + "learning_rate": 6.821331961087702e-07, + "loss": 0.3078, + "step": 38017 + }, + { + "epoch": 0.836847897864847, + "grad_norm": 1.4201436917578565, + "learning_rate": 6.819534734999888e-07, + "loss": 0.2618, + "step": 38018 + }, + { + "epoch": 0.8368699097512656, + "grad_norm": 1.7200778544331454, + "learning_rate": 6.817737728374452e-07, + "loss": 0.3389, + "step": 38019 + }, + { + "epoch": 0.8368919216376843, + "grad_norm": 1.452432506587964, + "learning_rate": 6.815940941220539e-07, + "loss": 0.2473, + "step": 38020 + }, + { + "epoch": 0.836913933524103, + "grad_norm": 1.6813899583361607, + "learning_rate": 6.814144373547266e-07, + "loss": 0.2456, + "step": 38021 + }, + { + "epoch": 0.8369359454105216, + "grad_norm": 1.1996043806787795, + "learning_rate": 6.812348025363774e-07, + "loss": 0.2001, + "step": 38022 + }, + { + "epoch": 0.8369579572969403, + "grad_norm": 1.5362508242808564, + "learning_rate": 6.810551896679202e-07, + "loss": 0.2504, + "step": 38023 + }, + { + "epoch": 0.836979969183359, + "grad_norm": 1.5369678027884892, + "learning_rate": 6.808755987502658e-07, + "loss": 0.2749, + "step": 38024 + }, + { + "epoch": 0.8370019810697776, + "grad_norm": 1.288373459073516, + "learning_rate": 6.80696029784329e-07, + "loss": 0.1785, + "step": 38025 + }, + { + "epoch": 0.8370239929561963, + "grad_norm": 4.664678157980782, + "learning_rate": 6.805164827710204e-07, + "loss": 0.2951, + "step": 38026 + }, + { + "epoch": 0.837046004842615, + "grad_norm": 2.66555103970849, + "learning_rate": 6.803369577112534e-07, + "loss": 0.3333, + "step": 38027 + }, + { + "epoch": 0.8370680167290336, + "grad_norm": 4.903139588857659, + "learning_rate": 6.801574546059414e-07, + "loss": 0.2971, + "step": 38028 + }, + { + "epoch": 0.8370900286154523, + "grad_norm": 1.6796144480850794, + "learning_rate": 6.799779734559953e-07, + "loss": 0.3758, + "step": 38029 + }, + { + "epoch": 0.837112040501871, + "grad_norm": 0.6321087467761625, + "learning_rate": 6.797985142623287e-07, + "loss": 0.4595, + "step": 38030 + }, + { + "epoch": 0.8371340523882896, + "grad_norm": 1.2814725834643692, + "learning_rate": 6.796190770258521e-07, + "loss": 0.2497, + "step": 38031 + }, + { + "epoch": 0.8371560642747083, + "grad_norm": 1.885194200025407, + "learning_rate": 6.794396617474785e-07, + "loss": 0.2082, + "step": 38032 + }, + { + "epoch": 0.837178076161127, + "grad_norm": 2.3401419233503775, + "learning_rate": 6.792602684281202e-07, + "loss": 0.384, + "step": 38033 + }, + { + "epoch": 0.8372000880475456, + "grad_norm": 0.6215075656230674, + "learning_rate": 6.790808970686874e-07, + "loss": 0.4436, + "step": 38034 + }, + { + "epoch": 0.8372220999339643, + "grad_norm": 1.4850468946744266, + "learning_rate": 6.789015476700939e-07, + "loss": 0.2066, + "step": 38035 + }, + { + "epoch": 0.837244111820383, + "grad_norm": 1.6581912916938528, + "learning_rate": 6.7872222023325e-07, + "loss": 0.2656, + "step": 38036 + }, + { + "epoch": 0.8372661237068016, + "grad_norm": 2.7013930201465475, + "learning_rate": 6.785429147590644e-07, + "loss": 0.2732, + "step": 38037 + }, + { + "epoch": 0.8372881355932204, + "grad_norm": 2.619228438744995, + "learning_rate": 6.783636312484538e-07, + "loss": 0.2728, + "step": 38038 + }, + { + "epoch": 0.837310147479639, + "grad_norm": 1.4632599830109183, + "learning_rate": 6.781843697023255e-07, + "loss": 0.2913, + "step": 38039 + }, + { + "epoch": 0.8373321593660576, + "grad_norm": 2.227479772172627, + "learning_rate": 6.780051301215928e-07, + "loss": 0.2586, + "step": 38040 + }, + { + "epoch": 0.8373541712524764, + "grad_norm": 1.634902475826912, + "learning_rate": 6.778259125071662e-07, + "loss": 0.2031, + "step": 38041 + }, + { + "epoch": 0.8373761831388951, + "grad_norm": 1.592071673930759, + "learning_rate": 6.776467168599544e-07, + "loss": 0.2666, + "step": 38042 + }, + { + "epoch": 0.8373981950253137, + "grad_norm": 0.6046549614156829, + "learning_rate": 6.774675431808703e-07, + "loss": 0.4462, + "step": 38043 + }, + { + "epoch": 0.8374202069117324, + "grad_norm": 2.3467937561959764, + "learning_rate": 6.772883914708245e-07, + "loss": 0.2177, + "step": 38044 + }, + { + "epoch": 0.837442218798151, + "grad_norm": 1.9387450028563467, + "learning_rate": 6.771092617307268e-07, + "loss": 0.31, + "step": 38045 + }, + { + "epoch": 0.8374642306845697, + "grad_norm": 2.2565768855409645, + "learning_rate": 6.769301539614881e-07, + "loss": 0.2476, + "step": 38046 + }, + { + "epoch": 0.8374862425709884, + "grad_norm": 1.7198963728158778, + "learning_rate": 6.767510681640183e-07, + "loss": 0.2948, + "step": 38047 + }, + { + "epoch": 0.837508254457407, + "grad_norm": 1.2806220157779427, + "learning_rate": 6.765720043392271e-07, + "loss": 0.2091, + "step": 38048 + }, + { + "epoch": 0.8375302663438257, + "grad_norm": 1.5718350987592447, + "learning_rate": 6.763929624880267e-07, + "loss": 0.2857, + "step": 38049 + }, + { + "epoch": 0.8375522782302444, + "grad_norm": 1.8739288526410256, + "learning_rate": 6.76213942611324e-07, + "loss": 0.3479, + "step": 38050 + }, + { + "epoch": 0.837574290116663, + "grad_norm": 1.430298852827593, + "learning_rate": 6.760349447100322e-07, + "loss": 0.2481, + "step": 38051 + }, + { + "epoch": 0.8375963020030817, + "grad_norm": 0.6257497168122681, + "learning_rate": 6.75855968785058e-07, + "loss": 0.4593, + "step": 38052 + }, + { + "epoch": 0.8376183138895004, + "grad_norm": 1.6553339466716233, + "learning_rate": 6.756770148373126e-07, + "loss": 0.4053, + "step": 38053 + }, + { + "epoch": 0.837640325775919, + "grad_norm": 1.3967665900348227, + "learning_rate": 6.754980828677065e-07, + "loss": 0.2737, + "step": 38054 + }, + { + "epoch": 0.8376623376623377, + "grad_norm": 3.0424021426224765, + "learning_rate": 6.753191728771469e-07, + "loss": 0.2361, + "step": 38055 + }, + { + "epoch": 0.8376843495487564, + "grad_norm": 1.7133901974175028, + "learning_rate": 6.751402848665445e-07, + "loss": 0.3192, + "step": 38056 + }, + { + "epoch": 0.837706361435175, + "grad_norm": 1.787717573036565, + "learning_rate": 6.749614188368076e-07, + "loss": 0.2312, + "step": 38057 + }, + { + "epoch": 0.8377283733215937, + "grad_norm": 2.441647889132869, + "learning_rate": 6.747825747888459e-07, + "loss": 0.381, + "step": 38058 + }, + { + "epoch": 0.8377503852080124, + "grad_norm": 1.524803612694726, + "learning_rate": 6.746037527235688e-07, + "loss": 0.2365, + "step": 38059 + }, + { + "epoch": 0.837772397094431, + "grad_norm": 0.6242884875127609, + "learning_rate": 6.744249526418839e-07, + "loss": 0.4423, + "step": 38060 + }, + { + "epoch": 0.8377944089808497, + "grad_norm": 2.364524656105928, + "learning_rate": 6.742461745447015e-07, + "loss": 0.355, + "step": 38061 + }, + { + "epoch": 0.8378164208672684, + "grad_norm": 1.4609085920688232, + "learning_rate": 6.740674184329288e-07, + "loss": 0.2596, + "step": 38062 + }, + { + "epoch": 0.837838432753687, + "grad_norm": 2.217437283822472, + "learning_rate": 6.738886843074744e-07, + "loss": 0.3158, + "step": 38063 + }, + { + "epoch": 0.8378604446401057, + "grad_norm": 6.3041679551922485, + "learning_rate": 6.737099721692486e-07, + "loss": 0.315, + "step": 38064 + }, + { + "epoch": 0.8378824565265244, + "grad_norm": 2.9962430704893555, + "learning_rate": 6.73531282019157e-07, + "loss": 0.3121, + "step": 38065 + }, + { + "epoch": 0.837904468412943, + "grad_norm": 1.4539079062227354, + "learning_rate": 6.733526138581104e-07, + "loss": 0.2987, + "step": 38066 + }, + { + "epoch": 0.8379264802993617, + "grad_norm": 1.323432299894686, + "learning_rate": 6.731739676870141e-07, + "loss": 0.3193, + "step": 38067 + }, + { + "epoch": 0.8379484921857803, + "grad_norm": 1.7910569849480809, + "learning_rate": 6.729953435067777e-07, + "loss": 0.2259, + "step": 38068 + }, + { + "epoch": 0.837970504072199, + "grad_norm": 0.6196578382809479, + "learning_rate": 6.728167413183101e-07, + "loss": 0.4862, + "step": 38069 + }, + { + "epoch": 0.8379925159586177, + "grad_norm": 0.5900738872771618, + "learning_rate": 6.726381611225163e-07, + "loss": 0.4311, + "step": 38070 + }, + { + "epoch": 0.8380145278450363, + "grad_norm": 0.6206013890371075, + "learning_rate": 6.724596029203073e-07, + "loss": 0.4279, + "step": 38071 + }, + { + "epoch": 0.838036539731455, + "grad_norm": 1.367069807037139, + "learning_rate": 6.722810667125879e-07, + "loss": 0.2242, + "step": 38072 + }, + { + "epoch": 0.8380585516178737, + "grad_norm": 1.619561024038263, + "learning_rate": 6.721025525002645e-07, + "loss": 0.2757, + "step": 38073 + }, + { + "epoch": 0.8380805635042923, + "grad_norm": 2.5874762676917897, + "learning_rate": 6.719240602842486e-07, + "loss": 0.3559, + "step": 38074 + }, + { + "epoch": 0.838102575390711, + "grad_norm": 0.6176300533478108, + "learning_rate": 6.717455900654435e-07, + "loss": 0.4092, + "step": 38075 + }, + { + "epoch": 0.8381245872771297, + "grad_norm": 0.6044143070295764, + "learning_rate": 6.715671418447595e-07, + "loss": 0.4108, + "step": 38076 + }, + { + "epoch": 0.8381465991635483, + "grad_norm": 0.6216958641318842, + "learning_rate": 6.713887156231013e-07, + "loss": 0.4379, + "step": 38077 + }, + { + "epoch": 0.838168611049967, + "grad_norm": 1.5011507573523655, + "learning_rate": 6.712103114013752e-07, + "loss": 0.2832, + "step": 38078 + }, + { + "epoch": 0.8381906229363857, + "grad_norm": 1.8514671275005017, + "learning_rate": 6.71031929180489e-07, + "loss": 0.3606, + "step": 38079 + }, + { + "epoch": 0.8382126348228043, + "grad_norm": 1.96350948579735, + "learning_rate": 6.708535689613494e-07, + "loss": 0.2843, + "step": 38080 + }, + { + "epoch": 0.838234646709223, + "grad_norm": 0.6088395240073853, + "learning_rate": 6.706752307448639e-07, + "loss": 0.4443, + "step": 38081 + }, + { + "epoch": 0.8382566585956417, + "grad_norm": 0.6221652898775089, + "learning_rate": 6.704969145319379e-07, + "loss": 0.446, + "step": 38082 + }, + { + "epoch": 0.8382786704820603, + "grad_norm": 1.7052601881139844, + "learning_rate": 6.703186203234762e-07, + "loss": 0.264, + "step": 38083 + }, + { + "epoch": 0.838300682368479, + "grad_norm": 1.7469102305782764, + "learning_rate": 6.701403481203867e-07, + "loss": 0.3519, + "step": 38084 + }, + { + "epoch": 0.8383226942548977, + "grad_norm": 16.440688349001288, + "learning_rate": 6.69962097923576e-07, + "loss": 0.3512, + "step": 38085 + }, + { + "epoch": 0.8383447061413163, + "grad_norm": 1.5474731305309977, + "learning_rate": 6.697838697339482e-07, + "loss": 0.2763, + "step": 38086 + }, + { + "epoch": 0.838366718027735, + "grad_norm": 1.8815865731786208, + "learning_rate": 6.696056635524107e-07, + "loss": 0.3317, + "step": 38087 + }, + { + "epoch": 0.8383887299141537, + "grad_norm": 1.3500761006562239, + "learning_rate": 6.694274793798683e-07, + "loss": 0.237, + "step": 38088 + }, + { + "epoch": 0.8384107418005723, + "grad_norm": 1.8286856892884666, + "learning_rate": 6.692493172172265e-07, + "loss": 0.3758, + "step": 38089 + }, + { + "epoch": 0.838432753686991, + "grad_norm": 2.1211868711139656, + "learning_rate": 6.690711770653923e-07, + "loss": 0.2747, + "step": 38090 + }, + { + "epoch": 0.8384547655734096, + "grad_norm": 1.892133036860186, + "learning_rate": 6.68893058925269e-07, + "loss": 0.2607, + "step": 38091 + }, + { + "epoch": 0.8384767774598283, + "grad_norm": 0.6091514928368524, + "learning_rate": 6.68714962797764e-07, + "loss": 0.4405, + "step": 38092 + }, + { + "epoch": 0.838498789346247, + "grad_norm": 1.4049570273150098, + "learning_rate": 6.685368886837806e-07, + "loss": 0.2739, + "step": 38093 + }, + { + "epoch": 0.8385208012326656, + "grad_norm": 1.7400272775545933, + "learning_rate": 6.683588365842241e-07, + "loss": 0.3407, + "step": 38094 + }, + { + "epoch": 0.8385428131190843, + "grad_norm": 1.7865782704183113, + "learning_rate": 6.681808065000017e-07, + "loss": 0.3049, + "step": 38095 + }, + { + "epoch": 0.838564825005503, + "grad_norm": 1.4395426137722447, + "learning_rate": 6.680027984320147e-07, + "loss": 0.3445, + "step": 38096 + }, + { + "epoch": 0.8385868368919216, + "grad_norm": 2.1528913989387175, + "learning_rate": 6.67824812381171e-07, + "loss": 0.3386, + "step": 38097 + }, + { + "epoch": 0.8386088487783403, + "grad_norm": 2.2353446567234654, + "learning_rate": 6.676468483483728e-07, + "loss": 0.1863, + "step": 38098 + }, + { + "epoch": 0.838630860664759, + "grad_norm": 1.671315201877208, + "learning_rate": 6.674689063345252e-07, + "loss": 0.2954, + "step": 38099 + }, + { + "epoch": 0.8386528725511776, + "grad_norm": 0.599094061845716, + "learning_rate": 6.672909863405346e-07, + "loss": 0.4305, + "step": 38100 + }, + { + "epoch": 0.8386748844375963, + "grad_norm": 1.8900138142530685, + "learning_rate": 6.671130883673022e-07, + "loss": 0.3, + "step": 38101 + }, + { + "epoch": 0.838696896324015, + "grad_norm": 2.7230215339564117, + "learning_rate": 6.669352124157353e-07, + "loss": 0.1989, + "step": 38102 + }, + { + "epoch": 0.8387189082104336, + "grad_norm": 2.222459615640907, + "learning_rate": 6.667573584867343e-07, + "loss": 0.3128, + "step": 38103 + }, + { + "epoch": 0.8387409200968523, + "grad_norm": 0.6217345207566928, + "learning_rate": 6.665795265812058e-07, + "loss": 0.4609, + "step": 38104 + }, + { + "epoch": 0.838762931983271, + "grad_norm": 1.4732495674942063, + "learning_rate": 6.664017167000536e-07, + "loss": 0.2572, + "step": 38105 + }, + { + "epoch": 0.8387849438696896, + "grad_norm": 1.4563871606139094, + "learning_rate": 6.662239288441797e-07, + "loss": 0.2674, + "step": 38106 + }, + { + "epoch": 0.8388069557561083, + "grad_norm": 1.2024894764770404, + "learning_rate": 6.660461630144899e-07, + "loss": 0.2395, + "step": 38107 + }, + { + "epoch": 0.838828967642527, + "grad_norm": 1.788304964600229, + "learning_rate": 6.658684192118864e-07, + "loss": 0.3029, + "step": 38108 + }, + { + "epoch": 0.8388509795289456, + "grad_norm": 0.6201830174034533, + "learning_rate": 6.656906974372701e-07, + "loss": 0.4604, + "step": 38109 + }, + { + "epoch": 0.8388729914153643, + "grad_norm": 0.6175691237022685, + "learning_rate": 6.655129976915497e-07, + "loss": 0.4269, + "step": 38110 + }, + { + "epoch": 0.838895003301783, + "grad_norm": 1.5399972776032775, + "learning_rate": 6.653353199756241e-07, + "loss": 0.2705, + "step": 38111 + }, + { + "epoch": 0.8389170151882016, + "grad_norm": 7.850656724807876, + "learning_rate": 6.65157664290399e-07, + "loss": 0.2824, + "step": 38112 + }, + { + "epoch": 0.8389390270746203, + "grad_norm": 1.3855088450879514, + "learning_rate": 6.649800306367754e-07, + "loss": 0.3844, + "step": 38113 + }, + { + "epoch": 0.8389610389610389, + "grad_norm": 1.6327870756464768, + "learning_rate": 6.648024190156565e-07, + "loss": 0.2805, + "step": 38114 + }, + { + "epoch": 0.8389830508474576, + "grad_norm": 2.0889162848693714, + "learning_rate": 6.646248294279445e-07, + "loss": 0.3198, + "step": 38115 + }, + { + "epoch": 0.8390050627338763, + "grad_norm": 1.6656489532663836, + "learning_rate": 6.644472618745435e-07, + "loss": 0.2791, + "step": 38116 + }, + { + "epoch": 0.8390270746202949, + "grad_norm": 1.8009302532725446, + "learning_rate": 6.642697163563555e-07, + "loss": 0.3664, + "step": 38117 + }, + { + "epoch": 0.8390490865067136, + "grad_norm": 1.5346482169367015, + "learning_rate": 6.640921928742832e-07, + "loss": 0.2542, + "step": 38118 + }, + { + "epoch": 0.8390710983931323, + "grad_norm": 1.5836451341773585, + "learning_rate": 6.639146914292271e-07, + "loss": 0.2708, + "step": 38119 + }, + { + "epoch": 0.8390931102795509, + "grad_norm": 4.310634229533151, + "learning_rate": 6.637372120220903e-07, + "loss": 0.2967, + "step": 38120 + }, + { + "epoch": 0.8391151221659696, + "grad_norm": 1.325298835828126, + "learning_rate": 6.635597546537753e-07, + "loss": 0.2395, + "step": 38121 + }, + { + "epoch": 0.8391371340523883, + "grad_norm": 1.4337057822736181, + "learning_rate": 6.633823193251843e-07, + "loss": 0.3171, + "step": 38122 + }, + { + "epoch": 0.8391591459388069, + "grad_norm": 0.6064616258102443, + "learning_rate": 6.632049060372192e-07, + "loss": 0.4525, + "step": 38123 + }, + { + "epoch": 0.8391811578252256, + "grad_norm": 1.5891578811457099, + "learning_rate": 6.630275147907794e-07, + "loss": 0.3037, + "step": 38124 + }, + { + "epoch": 0.8392031697116443, + "grad_norm": 0.6031992834927667, + "learning_rate": 6.628501455867681e-07, + "loss": 0.4299, + "step": 38125 + }, + { + "epoch": 0.8392251815980629, + "grad_norm": 5.995826819373511, + "learning_rate": 6.626727984260883e-07, + "loss": 0.3111, + "step": 38126 + }, + { + "epoch": 0.8392471934844816, + "grad_norm": 2.1652325863619986, + "learning_rate": 6.624954733096384e-07, + "loss": 0.3284, + "step": 38127 + }, + { + "epoch": 0.8392692053709003, + "grad_norm": 1.20233928979405, + "learning_rate": 6.623181702383224e-07, + "loss": 0.3358, + "step": 38128 + }, + { + "epoch": 0.8392912172573189, + "grad_norm": 1.6707745621547565, + "learning_rate": 6.621408892130388e-07, + "loss": 0.2838, + "step": 38129 + }, + { + "epoch": 0.8393132291437376, + "grad_norm": 2.1355448314855434, + "learning_rate": 6.619636302346899e-07, + "loss": 0.2824, + "step": 38130 + }, + { + "epoch": 0.8393352410301563, + "grad_norm": 1.7284058966806892, + "learning_rate": 6.617863933041779e-07, + "loss": 0.363, + "step": 38131 + }, + { + "epoch": 0.8393572529165749, + "grad_norm": 1.3835069053095879, + "learning_rate": 6.616091784224016e-07, + "loss": 0.3, + "step": 38132 + }, + { + "epoch": 0.8393792648029936, + "grad_norm": 1.6634455621656454, + "learning_rate": 6.614319855902629e-07, + "loss": 0.2936, + "step": 38133 + }, + { + "epoch": 0.8394012766894123, + "grad_norm": 2.5149001160397804, + "learning_rate": 6.612548148086612e-07, + "loss": 0.3526, + "step": 38134 + }, + { + "epoch": 0.8394232885758309, + "grad_norm": 1.6304933066131502, + "learning_rate": 6.610776660784979e-07, + "loss": 0.292, + "step": 38135 + }, + { + "epoch": 0.8394453004622496, + "grad_norm": 1.5144660092082427, + "learning_rate": 6.609005394006734e-07, + "loss": 0.2057, + "step": 38136 + }, + { + "epoch": 0.8394673123486682, + "grad_norm": 2.6092851460990785, + "learning_rate": 6.607234347760871e-07, + "loss": 0.275, + "step": 38137 + }, + { + "epoch": 0.8394893242350869, + "grad_norm": 1.8702022717853068, + "learning_rate": 6.60546352205641e-07, + "loss": 0.3546, + "step": 38138 + }, + { + "epoch": 0.8395113361215056, + "grad_norm": 2.34994449534782, + "learning_rate": 6.603692916902322e-07, + "loss": 0.339, + "step": 38139 + }, + { + "epoch": 0.8395333480079242, + "grad_norm": 1.9642142118385308, + "learning_rate": 6.601922532307625e-07, + "loss": 0.3143, + "step": 38140 + }, + { + "epoch": 0.8395553598943429, + "grad_norm": 1.4031417128711587, + "learning_rate": 6.600152368281321e-07, + "loss": 0.2349, + "step": 38141 + }, + { + "epoch": 0.8395773717807616, + "grad_norm": 1.7052656101629788, + "learning_rate": 6.598382424832394e-07, + "loss": 0.2948, + "step": 38142 + }, + { + "epoch": 0.8395993836671802, + "grad_norm": 0.6309358006919562, + "learning_rate": 6.596612701969852e-07, + "loss": 0.4502, + "step": 38143 + }, + { + "epoch": 0.8396213955535989, + "grad_norm": 2.2732709445815127, + "learning_rate": 6.594843199702683e-07, + "loss": 0.3316, + "step": 38144 + }, + { + "epoch": 0.8396434074400176, + "grad_norm": 1.4763793474353026, + "learning_rate": 6.593073918039861e-07, + "loss": 0.2951, + "step": 38145 + }, + { + "epoch": 0.8396654193264362, + "grad_norm": 1.876667430131597, + "learning_rate": 6.591304856990416e-07, + "loss": 0.2886, + "step": 38146 + }, + { + "epoch": 0.8396874312128549, + "grad_norm": 1.7060864730187475, + "learning_rate": 6.589536016563314e-07, + "loss": 0.2935, + "step": 38147 + }, + { + "epoch": 0.8397094430992736, + "grad_norm": 2.3764421687832438, + "learning_rate": 6.587767396767558e-07, + "loss": 0.2477, + "step": 38148 + }, + { + "epoch": 0.8397314549856922, + "grad_norm": 1.4912112034696134, + "learning_rate": 6.58599899761213e-07, + "loss": 0.2618, + "step": 38149 + }, + { + "epoch": 0.8397534668721109, + "grad_norm": 3.9509383312097643, + "learning_rate": 6.584230819106014e-07, + "loss": 0.331, + "step": 38150 + }, + { + "epoch": 0.8397754787585296, + "grad_norm": 1.7596268236297852, + "learning_rate": 6.582462861258193e-07, + "loss": 0.3035, + "step": 38151 + }, + { + "epoch": 0.8397974906449482, + "grad_norm": 1.8564913827295608, + "learning_rate": 6.580695124077663e-07, + "loss": 0.2874, + "step": 38152 + }, + { + "epoch": 0.839819502531367, + "grad_norm": 1.3263784070205853, + "learning_rate": 6.578927607573421e-07, + "loss": 0.2703, + "step": 38153 + }, + { + "epoch": 0.8398415144177857, + "grad_norm": 1.9138493771240586, + "learning_rate": 6.577160311754433e-07, + "loss": 0.3589, + "step": 38154 + }, + { + "epoch": 0.8398635263042042, + "grad_norm": 1.3539542651771928, + "learning_rate": 6.575393236629668e-07, + "loss": 0.2776, + "step": 38155 + }, + { + "epoch": 0.839885538190623, + "grad_norm": 3.020338878298175, + "learning_rate": 6.573626382208126e-07, + "loss": 0.2796, + "step": 38156 + }, + { + "epoch": 0.8399075500770417, + "grad_norm": 0.6196756879021489, + "learning_rate": 6.571859748498782e-07, + "loss": 0.4857, + "step": 38157 + }, + { + "epoch": 0.8399295619634602, + "grad_norm": 0.6187413330036358, + "learning_rate": 6.57009333551063e-07, + "loss": 0.4627, + "step": 38158 + }, + { + "epoch": 0.839951573849879, + "grad_norm": 0.6285763794545673, + "learning_rate": 6.568327143252634e-07, + "loss": 0.4722, + "step": 38159 + }, + { + "epoch": 0.8399735857362975, + "grad_norm": 1.644756361472899, + "learning_rate": 6.566561171733759e-07, + "loss": 0.2771, + "step": 38160 + }, + { + "epoch": 0.8399955976227163, + "grad_norm": 1.4833630475019348, + "learning_rate": 6.564795420962989e-07, + "loss": 0.2245, + "step": 38161 + }, + { + "epoch": 0.840017609509135, + "grad_norm": 1.4698650348841469, + "learning_rate": 6.563029890949302e-07, + "loss": 0.2466, + "step": 38162 + }, + { + "epoch": 0.8400396213955535, + "grad_norm": 1.4628877848067283, + "learning_rate": 6.561264581701682e-07, + "loss": 0.3491, + "step": 38163 + }, + { + "epoch": 0.8400616332819723, + "grad_norm": 1.7200965544839908, + "learning_rate": 6.559499493229094e-07, + "loss": 0.3312, + "step": 38164 + }, + { + "epoch": 0.840083645168391, + "grad_norm": 1.7558412562238432, + "learning_rate": 6.55773462554049e-07, + "loss": 0.2438, + "step": 38165 + }, + { + "epoch": 0.8401056570548096, + "grad_norm": 0.6056086346092918, + "learning_rate": 6.555969978644855e-07, + "loss": 0.4428, + "step": 38166 + }, + { + "epoch": 0.8401276689412283, + "grad_norm": 1.8809874321096, + "learning_rate": 6.554205552551168e-07, + "loss": 0.3987, + "step": 38167 + }, + { + "epoch": 0.840149680827647, + "grad_norm": 1.8329830654673764, + "learning_rate": 6.552441347268374e-07, + "loss": 0.326, + "step": 38168 + }, + { + "epoch": 0.8401716927140656, + "grad_norm": 2.499736167142063, + "learning_rate": 6.550677362805464e-07, + "loss": 0.3298, + "step": 38169 + }, + { + "epoch": 0.8401937046004843, + "grad_norm": 0.6243090076925344, + "learning_rate": 6.548913599171375e-07, + "loss": 0.4714, + "step": 38170 + }, + { + "epoch": 0.840215716486903, + "grad_norm": 5.491248969934004, + "learning_rate": 6.54715005637509e-07, + "loss": 0.261, + "step": 38171 + }, + { + "epoch": 0.8402377283733216, + "grad_norm": 4.245767330704047, + "learning_rate": 6.545386734425579e-07, + "loss": 0.2049, + "step": 38172 + }, + { + "epoch": 0.8402597402597403, + "grad_norm": 1.5487892443726665, + "learning_rate": 6.543623633331786e-07, + "loss": 0.3564, + "step": 38173 + }, + { + "epoch": 0.840281752146159, + "grad_norm": 1.9985980725111665, + "learning_rate": 6.541860753102686e-07, + "loss": 0.297, + "step": 38174 + }, + { + "epoch": 0.8403037640325776, + "grad_norm": 1.7016477679950144, + "learning_rate": 6.540098093747227e-07, + "loss": 0.3336, + "step": 38175 + }, + { + "epoch": 0.8403257759189963, + "grad_norm": 0.6217151020494912, + "learning_rate": 6.538335655274369e-07, + "loss": 0.4523, + "step": 38176 + }, + { + "epoch": 0.840347787805415, + "grad_norm": 1.4092597469283061, + "learning_rate": 6.536573437693089e-07, + "loss": 0.2619, + "step": 38177 + }, + { + "epoch": 0.8403697996918336, + "grad_norm": 0.6191393715709442, + "learning_rate": 6.534811441012312e-07, + "loss": 0.4435, + "step": 38178 + }, + { + "epoch": 0.8403918115782523, + "grad_norm": 1.7347418084888941, + "learning_rate": 6.53304966524102e-07, + "loss": 0.1962, + "step": 38179 + }, + { + "epoch": 0.840413823464671, + "grad_norm": 2.0277297279379187, + "learning_rate": 6.531288110388151e-07, + "loss": 0.289, + "step": 38180 + }, + { + "epoch": 0.8404358353510896, + "grad_norm": 8.240491029303477, + "learning_rate": 6.529526776462658e-07, + "loss": 0.2273, + "step": 38181 + }, + { + "epoch": 0.8404578472375083, + "grad_norm": 0.6281767768348399, + "learning_rate": 6.527765663473512e-07, + "loss": 0.4469, + "step": 38182 + }, + { + "epoch": 0.840479859123927, + "grad_norm": 3.8566345394570622, + "learning_rate": 6.52600477142964e-07, + "loss": 0.3059, + "step": 38183 + }, + { + "epoch": 0.8405018710103456, + "grad_norm": 9.704486524327736, + "learning_rate": 6.52424410034001e-07, + "loss": 0.3799, + "step": 38184 + }, + { + "epoch": 0.8405238828967643, + "grad_norm": 1.6787833932896004, + "learning_rate": 6.522483650213562e-07, + "loss": 0.3323, + "step": 38185 + }, + { + "epoch": 0.8405458947831829, + "grad_norm": 2.4761201182030623, + "learning_rate": 6.520723421059233e-07, + "loss": 0.3042, + "step": 38186 + }, + { + "epoch": 0.8405679066696016, + "grad_norm": 1.7579052482333755, + "learning_rate": 6.518963412885981e-07, + "loss": 0.247, + "step": 38187 + }, + { + "epoch": 0.8405899185560203, + "grad_norm": 8.761303554534146, + "learning_rate": 6.51720362570275e-07, + "loss": 0.2847, + "step": 38188 + }, + { + "epoch": 0.8406119304424389, + "grad_norm": 2.2324283344952622, + "learning_rate": 6.515444059518494e-07, + "loss": 0.3189, + "step": 38189 + }, + { + "epoch": 0.8406339423288576, + "grad_norm": 1.5472731337924968, + "learning_rate": 6.51368471434215e-07, + "loss": 0.3058, + "step": 38190 + }, + { + "epoch": 0.8406559542152763, + "grad_norm": 1.6101991278559025, + "learning_rate": 6.511925590182638e-07, + "loss": 0.2979, + "step": 38191 + }, + { + "epoch": 0.8406779661016949, + "grad_norm": 2.867923709532799, + "learning_rate": 6.510166687048925e-07, + "loss": 0.2905, + "step": 38192 + }, + { + "epoch": 0.8406999779881136, + "grad_norm": 2.206490328965688, + "learning_rate": 6.508408004949934e-07, + "loss": 0.3011, + "step": 38193 + }, + { + "epoch": 0.8407219898745323, + "grad_norm": 1.7483567240021807, + "learning_rate": 6.506649543894622e-07, + "loss": 0.2662, + "step": 38194 + }, + { + "epoch": 0.8407440017609509, + "grad_norm": 0.6117856847596114, + "learning_rate": 6.50489130389192e-07, + "loss": 0.4545, + "step": 38195 + }, + { + "epoch": 0.8407660136473696, + "grad_norm": 0.6255324227572239, + "learning_rate": 6.503133284950747e-07, + "loss": 0.4322, + "step": 38196 + }, + { + "epoch": 0.8407880255337883, + "grad_norm": 0.6285900908293532, + "learning_rate": 6.501375487080053e-07, + "loss": 0.462, + "step": 38197 + }, + { + "epoch": 0.8408100374202069, + "grad_norm": 2.0130411727147055, + "learning_rate": 6.49961791028877e-07, + "loss": 0.1999, + "step": 38198 + }, + { + "epoch": 0.8408320493066256, + "grad_norm": 0.5889117749635058, + "learning_rate": 6.497860554585839e-07, + "loss": 0.4377, + "step": 38199 + }, + { + "epoch": 0.8408540611930443, + "grad_norm": 1.5468713261165579, + "learning_rate": 6.496103419980182e-07, + "loss": 0.227, + "step": 38200 + }, + { + "epoch": 0.8408760730794629, + "grad_norm": 1.6115273803997947, + "learning_rate": 6.49434650648072e-07, + "loss": 0.2907, + "step": 38201 + }, + { + "epoch": 0.8408980849658816, + "grad_norm": 2.1460446503271897, + "learning_rate": 6.492589814096395e-07, + "loss": 0.3094, + "step": 38202 + }, + { + "epoch": 0.8409200968523003, + "grad_norm": 2.312393265340325, + "learning_rate": 6.490833342836144e-07, + "loss": 0.2466, + "step": 38203 + }, + { + "epoch": 0.8409421087387189, + "grad_norm": 1.6928158602840313, + "learning_rate": 6.489077092708873e-07, + "loss": 0.2975, + "step": 38204 + }, + { + "epoch": 0.8409641206251376, + "grad_norm": 1.8489655090042523, + "learning_rate": 6.487321063723529e-07, + "loss": 0.2439, + "step": 38205 + }, + { + "epoch": 0.8409861325115563, + "grad_norm": 0.6118899598769661, + "learning_rate": 6.485565255889014e-07, + "loss": 0.4363, + "step": 38206 + }, + { + "epoch": 0.8410081443979749, + "grad_norm": 1.8093185584442686, + "learning_rate": 6.483809669214269e-07, + "loss": 0.1825, + "step": 38207 + }, + { + "epoch": 0.8410301562843936, + "grad_norm": 1.8367112224294357, + "learning_rate": 6.482054303708218e-07, + "loss": 0.266, + "step": 38208 + }, + { + "epoch": 0.8410521681708122, + "grad_norm": 2.0631150310977637, + "learning_rate": 6.480299159379766e-07, + "loss": 0.2637, + "step": 38209 + }, + { + "epoch": 0.8410741800572309, + "grad_norm": 2.460423338054206, + "learning_rate": 6.47854423623786e-07, + "loss": 0.253, + "step": 38210 + }, + { + "epoch": 0.8410961919436496, + "grad_norm": 1.8056903920574299, + "learning_rate": 6.476789534291389e-07, + "loss": 0.2846, + "step": 38211 + }, + { + "epoch": 0.8411182038300682, + "grad_norm": 1.4493149508048346, + "learning_rate": 6.475035053549284e-07, + "loss": 0.2946, + "step": 38212 + }, + { + "epoch": 0.8411402157164869, + "grad_norm": 1.2707557180769695, + "learning_rate": 6.473280794020476e-07, + "loss": 0.2417, + "step": 38213 + }, + { + "epoch": 0.8411622276029056, + "grad_norm": 1.616717812054844, + "learning_rate": 6.471526755713859e-07, + "loss": 0.3056, + "step": 38214 + }, + { + "epoch": 0.8411842394893242, + "grad_norm": 1.5223066336807403, + "learning_rate": 6.469772938638363e-07, + "loss": 0.2821, + "step": 38215 + }, + { + "epoch": 0.8412062513757429, + "grad_norm": 1.9509985863232107, + "learning_rate": 6.468019342802889e-07, + "loss": 0.3157, + "step": 38216 + }, + { + "epoch": 0.8412282632621616, + "grad_norm": 1.8933612000059534, + "learning_rate": 6.466265968216356e-07, + "loss": 0.2251, + "step": 38217 + }, + { + "epoch": 0.8412502751485802, + "grad_norm": 1.8880540729613409, + "learning_rate": 6.464512814887686e-07, + "loss": 0.3116, + "step": 38218 + }, + { + "epoch": 0.8412722870349989, + "grad_norm": 1.4830664185050055, + "learning_rate": 6.46275988282577e-07, + "loss": 0.3455, + "step": 38219 + }, + { + "epoch": 0.8412942989214176, + "grad_norm": 0.6271503898040484, + "learning_rate": 6.461007172039535e-07, + "loss": 0.4569, + "step": 38220 + }, + { + "epoch": 0.8413163108078362, + "grad_norm": 1.7800758797170977, + "learning_rate": 6.459254682537875e-07, + "loss": 0.2339, + "step": 38221 + }, + { + "epoch": 0.8413383226942549, + "grad_norm": 1.8526834647690573, + "learning_rate": 6.457502414329708e-07, + "loss": 0.283, + "step": 38222 + }, + { + "epoch": 0.8413603345806736, + "grad_norm": 2.3555074171342905, + "learning_rate": 6.455750367423924e-07, + "loss": 0.3363, + "step": 38223 + }, + { + "epoch": 0.8413823464670922, + "grad_norm": 0.6278441033742501, + "learning_rate": 6.453998541829443e-07, + "loss": 0.4472, + "step": 38224 + }, + { + "epoch": 0.8414043583535109, + "grad_norm": 2.060824120763633, + "learning_rate": 6.452246937555167e-07, + "loss": 0.307, + "step": 38225 + }, + { + "epoch": 0.8414263702399296, + "grad_norm": 1.4264782283933188, + "learning_rate": 6.450495554609997e-07, + "loss": 0.2248, + "step": 38226 + }, + { + "epoch": 0.8414483821263482, + "grad_norm": 1.7487403084462447, + "learning_rate": 6.448744393002826e-07, + "loss": 0.373, + "step": 38227 + }, + { + "epoch": 0.8414703940127669, + "grad_norm": 1.5330364349868073, + "learning_rate": 6.446993452742556e-07, + "loss": 0.3093, + "step": 38228 + }, + { + "epoch": 0.8414924058991856, + "grad_norm": 0.6148403252470367, + "learning_rate": 6.445242733838091e-07, + "loss": 0.4621, + "step": 38229 + }, + { + "epoch": 0.8415144177856042, + "grad_norm": 2.1066685078937692, + "learning_rate": 6.443492236298338e-07, + "loss": 0.2454, + "step": 38230 + }, + { + "epoch": 0.8415364296720229, + "grad_norm": 1.5231187996692674, + "learning_rate": 6.441741960132186e-07, + "loss": 0.2889, + "step": 38231 + }, + { + "epoch": 0.8415584415584415, + "grad_norm": 1.454126528909177, + "learning_rate": 6.43999190534852e-07, + "loss": 0.374, + "step": 38232 + }, + { + "epoch": 0.8415804534448602, + "grad_norm": 1.3079322002272753, + "learning_rate": 6.438242071956235e-07, + "loss": 0.2388, + "step": 38233 + }, + { + "epoch": 0.8416024653312789, + "grad_norm": 1.3965351236151593, + "learning_rate": 6.436492459964244e-07, + "loss": 0.2605, + "step": 38234 + }, + { + "epoch": 0.8416244772176975, + "grad_norm": 0.6221264472133229, + "learning_rate": 6.434743069381432e-07, + "loss": 0.4516, + "step": 38235 + }, + { + "epoch": 0.8416464891041162, + "grad_norm": 1.755468505595765, + "learning_rate": 6.432993900216689e-07, + "loss": 0.3158, + "step": 38236 + }, + { + "epoch": 0.8416685009905349, + "grad_norm": 1.3670218033723196, + "learning_rate": 6.431244952478888e-07, + "loss": 0.362, + "step": 38237 + }, + { + "epoch": 0.8416905128769535, + "grad_norm": 0.6225176733237329, + "learning_rate": 6.42949622617694e-07, + "loss": 0.4339, + "step": 38238 + }, + { + "epoch": 0.8417125247633722, + "grad_norm": 1.4501025004966956, + "learning_rate": 6.427747721319716e-07, + "loss": 0.3288, + "step": 38239 + }, + { + "epoch": 0.8417345366497909, + "grad_norm": 3.286480795808849, + "learning_rate": 6.425999437916131e-07, + "loss": 0.2833, + "step": 38240 + }, + { + "epoch": 0.8417565485362095, + "grad_norm": 4.686966567842962, + "learning_rate": 6.424251375975044e-07, + "loss": 0.28, + "step": 38241 + }, + { + "epoch": 0.8417785604226282, + "grad_norm": 1.8976507523861834, + "learning_rate": 6.422503535505342e-07, + "loss": 0.3728, + "step": 38242 + }, + { + "epoch": 0.8418005723090469, + "grad_norm": 4.708089511583108, + "learning_rate": 6.420755916515908e-07, + "loss": 0.3092, + "step": 38243 + }, + { + "epoch": 0.8418225841954655, + "grad_norm": 1.2689009812830738, + "learning_rate": 6.419008519015646e-07, + "loss": 0.2314, + "step": 38244 + }, + { + "epoch": 0.8418445960818842, + "grad_norm": 3.0192890462546322, + "learning_rate": 6.417261343013403e-07, + "loss": 0.2982, + "step": 38245 + }, + { + "epoch": 0.8418666079683029, + "grad_norm": 1.6188576475858556, + "learning_rate": 6.41551438851809e-07, + "loss": 0.2041, + "step": 38246 + }, + { + "epoch": 0.8418886198547215, + "grad_norm": 2.814037672113601, + "learning_rate": 6.413767655538566e-07, + "loss": 0.3297, + "step": 38247 + }, + { + "epoch": 0.8419106317411402, + "grad_norm": 1.5372520209539107, + "learning_rate": 6.412021144083707e-07, + "loss": 0.3161, + "step": 38248 + }, + { + "epoch": 0.8419326436275589, + "grad_norm": 1.4285389020770591, + "learning_rate": 6.410274854162413e-07, + "loss": 0.2651, + "step": 38249 + }, + { + "epoch": 0.8419546555139775, + "grad_norm": 1.546099820924929, + "learning_rate": 6.408528785783535e-07, + "loss": 0.2727, + "step": 38250 + }, + { + "epoch": 0.8419766674003962, + "grad_norm": 0.6520103652896787, + "learning_rate": 6.406782938955963e-07, + "loss": 0.4224, + "step": 38251 + }, + { + "epoch": 0.8419986792868149, + "grad_norm": 1.5608593774370247, + "learning_rate": 6.40503731368855e-07, + "loss": 0.3576, + "step": 38252 + }, + { + "epoch": 0.8420206911732335, + "grad_norm": 1.3244189482471127, + "learning_rate": 6.403291909990195e-07, + "loss": 0.3102, + "step": 38253 + }, + { + "epoch": 0.8420427030596522, + "grad_norm": 2.3405914198675237, + "learning_rate": 6.401546727869745e-07, + "loss": 0.3711, + "step": 38254 + }, + { + "epoch": 0.8420647149460708, + "grad_norm": 1.850543014178436, + "learning_rate": 6.399801767336078e-07, + "loss": 0.3222, + "step": 38255 + }, + { + "epoch": 0.8420867268324895, + "grad_norm": 0.6178143393024068, + "learning_rate": 6.39805702839808e-07, + "loss": 0.4573, + "step": 38256 + }, + { + "epoch": 0.8421087387189082, + "grad_norm": 30.654279974130628, + "learning_rate": 6.39631251106459e-07, + "loss": 0.2945, + "step": 38257 + }, + { + "epoch": 0.8421307506053268, + "grad_norm": 2.2275565396421397, + "learning_rate": 6.394568215344494e-07, + "loss": 0.2975, + "step": 38258 + }, + { + "epoch": 0.8421527624917455, + "grad_norm": 1.6351355007862816, + "learning_rate": 6.392824141246645e-07, + "loss": 0.2564, + "step": 38259 + }, + { + "epoch": 0.8421747743781642, + "grad_norm": 4.9523947460977125, + "learning_rate": 6.391080288779911e-07, + "loss": 0.3823, + "step": 38260 + }, + { + "epoch": 0.8421967862645828, + "grad_norm": 0.6070783383183099, + "learning_rate": 6.389336657953171e-07, + "loss": 0.4325, + "step": 38261 + }, + { + "epoch": 0.8422187981510015, + "grad_norm": 1.5510260600738812, + "learning_rate": 6.387593248775259e-07, + "loss": 0.3528, + "step": 38262 + }, + { + "epoch": 0.8422408100374202, + "grad_norm": 0.6480713224024874, + "learning_rate": 6.385850061255061e-07, + "loss": 0.4591, + "step": 38263 + }, + { + "epoch": 0.8422628219238388, + "grad_norm": 1.3213457251173295, + "learning_rate": 6.384107095401415e-07, + "loss": 0.2092, + "step": 38264 + }, + { + "epoch": 0.8422848338102575, + "grad_norm": 17.32113069082579, + "learning_rate": 6.382364351223191e-07, + "loss": 0.2836, + "step": 38265 + }, + { + "epoch": 0.8423068456966762, + "grad_norm": 2.031153482427522, + "learning_rate": 6.380621828729256e-07, + "loss": 0.3166, + "step": 38266 + }, + { + "epoch": 0.8423288575830948, + "grad_norm": 1.7593006834281069, + "learning_rate": 6.378879527928456e-07, + "loss": 0.2349, + "step": 38267 + }, + { + "epoch": 0.8423508694695135, + "grad_norm": 2.1024130551260862, + "learning_rate": 6.377137448829635e-07, + "loss": 0.257, + "step": 38268 + }, + { + "epoch": 0.8423728813559322, + "grad_norm": 1.5015496038123306, + "learning_rate": 6.375395591441652e-07, + "loss": 0.3327, + "step": 38269 + }, + { + "epoch": 0.8423948932423508, + "grad_norm": 2.3976447901236293, + "learning_rate": 6.373653955773373e-07, + "loss": 0.2899, + "step": 38270 + }, + { + "epoch": 0.8424169051287695, + "grad_norm": 1.7531634648787098, + "learning_rate": 6.371912541833652e-07, + "loss": 0.335, + "step": 38271 + }, + { + "epoch": 0.8424389170151882, + "grad_norm": 2.3941022951711397, + "learning_rate": 6.370171349631327e-07, + "loss": 0.2823, + "step": 38272 + }, + { + "epoch": 0.8424609289016068, + "grad_norm": 1.9629565768195447, + "learning_rate": 6.368430379175239e-07, + "loss": 0.2835, + "step": 38273 + }, + { + "epoch": 0.8424829407880255, + "grad_norm": 1.8715413912096344, + "learning_rate": 6.366689630474254e-07, + "loss": 0.2781, + "step": 38274 + }, + { + "epoch": 0.8425049526744443, + "grad_norm": 0.6403854286624557, + "learning_rate": 6.36494910353721e-07, + "loss": 0.4437, + "step": 38275 + }, + { + "epoch": 0.8425269645608628, + "grad_norm": 6.257153370882419, + "learning_rate": 6.363208798372966e-07, + "loss": 0.3547, + "step": 38276 + }, + { + "epoch": 0.8425489764472816, + "grad_norm": 2.4744483286058494, + "learning_rate": 6.361468714990359e-07, + "loss": 0.3288, + "step": 38277 + }, + { + "epoch": 0.8425709883337001, + "grad_norm": 2.1824005963330246, + "learning_rate": 6.359728853398222e-07, + "loss": 0.3154, + "step": 38278 + }, + { + "epoch": 0.8425930002201188, + "grad_norm": 2.2410439528705406, + "learning_rate": 6.357989213605403e-07, + "loss": 0.2387, + "step": 38279 + }, + { + "epoch": 0.8426150121065376, + "grad_norm": 1.4566731460178612, + "learning_rate": 6.356249795620751e-07, + "loss": 0.276, + "step": 38280 + }, + { + "epoch": 0.8426370239929561, + "grad_norm": 0.5978557492491786, + "learning_rate": 6.354510599453112e-07, + "loss": 0.442, + "step": 38281 + }, + { + "epoch": 0.8426590358793749, + "grad_norm": 1.5439805748284046, + "learning_rate": 6.35277162511132e-07, + "loss": 0.2754, + "step": 38282 + }, + { + "epoch": 0.8426810477657936, + "grad_norm": 2.6494902861668304, + "learning_rate": 6.351032872604196e-07, + "loss": 0.3334, + "step": 38283 + }, + { + "epoch": 0.8427030596522121, + "grad_norm": 1.429399038552041, + "learning_rate": 6.349294341940593e-07, + "loss": 0.2279, + "step": 38284 + }, + { + "epoch": 0.8427250715386309, + "grad_norm": 1.6180223905975832, + "learning_rate": 6.347556033129354e-07, + "loss": 0.297, + "step": 38285 + }, + { + "epoch": 0.8427470834250496, + "grad_norm": 1.3183384707200063, + "learning_rate": 6.345817946179294e-07, + "loss": 0.264, + "step": 38286 + }, + { + "epoch": 0.8427690953114682, + "grad_norm": 0.5987831200949735, + "learning_rate": 6.344080081099268e-07, + "loss": 0.4281, + "step": 38287 + }, + { + "epoch": 0.8427911071978869, + "grad_norm": 1.9698744364640162, + "learning_rate": 6.342342437898086e-07, + "loss": 0.24, + "step": 38288 + }, + { + "epoch": 0.8428131190843056, + "grad_norm": 2.0052994952559975, + "learning_rate": 6.340605016584605e-07, + "loss": 0.2795, + "step": 38289 + }, + { + "epoch": 0.8428351309707242, + "grad_norm": 1.502336048166831, + "learning_rate": 6.338867817167632e-07, + "loss": 0.26, + "step": 38290 + }, + { + "epoch": 0.8428571428571429, + "grad_norm": 1.6776906608259567, + "learning_rate": 6.337130839656003e-07, + "loss": 0.2978, + "step": 38291 + }, + { + "epoch": 0.8428791547435616, + "grad_norm": 1.4078056567348873, + "learning_rate": 6.33539408405856e-07, + "loss": 0.2147, + "step": 38292 + }, + { + "epoch": 0.8429011666299802, + "grad_norm": 1.376544880213853, + "learning_rate": 6.33365755038411e-07, + "loss": 0.2772, + "step": 38293 + }, + { + "epoch": 0.8429231785163989, + "grad_norm": 1.6345854553691166, + "learning_rate": 6.331921238641497e-07, + "loss": 0.2639, + "step": 38294 + }, + { + "epoch": 0.8429451904028176, + "grad_norm": 1.376761572965949, + "learning_rate": 6.330185148839529e-07, + "loss": 0.2928, + "step": 38295 + }, + { + "epoch": 0.8429672022892362, + "grad_norm": 1.54208019751732, + "learning_rate": 6.328449280987037e-07, + "loss": 0.1483, + "step": 38296 + }, + { + "epoch": 0.8429892141756549, + "grad_norm": 1.6539676088408717, + "learning_rate": 6.326713635092857e-07, + "loss": 0.2291, + "step": 38297 + }, + { + "epoch": 0.8430112260620736, + "grad_norm": 3.0518328433861637, + "learning_rate": 6.324978211165783e-07, + "loss": 0.3006, + "step": 38298 + }, + { + "epoch": 0.8430332379484922, + "grad_norm": 2.0225956798937026, + "learning_rate": 6.323243009214658e-07, + "loss": 0.2186, + "step": 38299 + }, + { + "epoch": 0.8430552498349109, + "grad_norm": 0.622938752228088, + "learning_rate": 6.321508029248286e-07, + "loss": 0.4632, + "step": 38300 + }, + { + "epoch": 0.8430772617213295, + "grad_norm": 7.45339111362407, + "learning_rate": 6.319773271275486e-07, + "loss": 0.3347, + "step": 38301 + }, + { + "epoch": 0.8430992736077482, + "grad_norm": 0.6085686263173796, + "learning_rate": 6.318038735305093e-07, + "loss": 0.4596, + "step": 38302 + }, + { + "epoch": 0.8431212854941669, + "grad_norm": 1.4399857208693865, + "learning_rate": 6.316304421345914e-07, + "loss": 0.3187, + "step": 38303 + }, + { + "epoch": 0.8431432973805855, + "grad_norm": 1.9741497041387965, + "learning_rate": 6.314570329406744e-07, + "loss": 0.3173, + "step": 38304 + }, + { + "epoch": 0.8431653092670042, + "grad_norm": 1.844362342596712, + "learning_rate": 6.312836459496413e-07, + "loss": 0.3023, + "step": 38305 + }, + { + "epoch": 0.8431873211534229, + "grad_norm": 1.7328387373675977, + "learning_rate": 6.311102811623732e-07, + "loss": 0.3217, + "step": 38306 + }, + { + "epoch": 0.8432093330398415, + "grad_norm": 1.4897521623038206, + "learning_rate": 6.309369385797526e-07, + "loss": 0.2196, + "step": 38307 + }, + { + "epoch": 0.8432313449262602, + "grad_norm": 1.7942292742023667, + "learning_rate": 6.307636182026589e-07, + "loss": 0.3322, + "step": 38308 + }, + { + "epoch": 0.8432533568126789, + "grad_norm": 1.7194110857538725, + "learning_rate": 6.305903200319724e-07, + "loss": 0.3609, + "step": 38309 + }, + { + "epoch": 0.8432753686990975, + "grad_norm": 0.6356123891809208, + "learning_rate": 6.304170440685742e-07, + "loss": 0.4495, + "step": 38310 + }, + { + "epoch": 0.8432973805855162, + "grad_norm": 1.657611668139006, + "learning_rate": 6.302437903133457e-07, + "loss": 0.2463, + "step": 38311 + }, + { + "epoch": 0.8433193924719349, + "grad_norm": 1.727196124831654, + "learning_rate": 6.300705587671685e-07, + "loss": 0.2613, + "step": 38312 + }, + { + "epoch": 0.8433414043583535, + "grad_norm": 14.768825741594947, + "learning_rate": 6.298973494309212e-07, + "loss": 0.2567, + "step": 38313 + }, + { + "epoch": 0.8433634162447722, + "grad_norm": 1.9673922363948895, + "learning_rate": 6.297241623054845e-07, + "loss": 0.3432, + "step": 38314 + }, + { + "epoch": 0.8433854281311909, + "grad_norm": 1.9013303351341082, + "learning_rate": 6.29550997391738e-07, + "loss": 0.3431, + "step": 38315 + }, + { + "epoch": 0.8434074400176095, + "grad_norm": 0.6214064067072265, + "learning_rate": 6.293778546905632e-07, + "loss": 0.4352, + "step": 38316 + }, + { + "epoch": 0.8434294519040282, + "grad_norm": 1.965402911734256, + "learning_rate": 6.292047342028401e-07, + "loss": 0.2667, + "step": 38317 + }, + { + "epoch": 0.8434514637904469, + "grad_norm": 1.7628512516834967, + "learning_rate": 6.290316359294479e-07, + "loss": 0.2906, + "step": 38318 + }, + { + "epoch": 0.8434734756768655, + "grad_norm": 1.7895815631838803, + "learning_rate": 6.28858559871266e-07, + "loss": 0.3137, + "step": 38319 + }, + { + "epoch": 0.8434954875632842, + "grad_norm": 1.769829071674602, + "learning_rate": 6.286855060291741e-07, + "loss": 0.3251, + "step": 38320 + }, + { + "epoch": 0.8435174994497029, + "grad_norm": 1.478082643880327, + "learning_rate": 6.285124744040521e-07, + "loss": 0.2869, + "step": 38321 + }, + { + "epoch": 0.8435395113361215, + "grad_norm": 2.7297444696924478, + "learning_rate": 6.283394649967805e-07, + "loss": 0.2244, + "step": 38322 + }, + { + "epoch": 0.8435615232225402, + "grad_norm": 0.6251612554013826, + "learning_rate": 6.281664778082375e-07, + "loss": 0.4399, + "step": 38323 + }, + { + "epoch": 0.8435835351089589, + "grad_norm": 3.376691442194877, + "learning_rate": 6.279935128393011e-07, + "loss": 0.2759, + "step": 38324 + }, + { + "epoch": 0.8436055469953775, + "grad_norm": 1.319880469734309, + "learning_rate": 6.27820570090853e-07, + "loss": 0.2211, + "step": 38325 + }, + { + "epoch": 0.8436275588817962, + "grad_norm": 1.4050298108711972, + "learning_rate": 6.276476495637701e-07, + "loss": 0.2203, + "step": 38326 + }, + { + "epoch": 0.8436495707682148, + "grad_norm": 3.714762781599934, + "learning_rate": 6.274747512589313e-07, + "loss": 0.302, + "step": 38327 + }, + { + "epoch": 0.8436715826546335, + "grad_norm": 1.286576798622573, + "learning_rate": 6.273018751772175e-07, + "loss": 0.2236, + "step": 38328 + }, + { + "epoch": 0.8436935945410522, + "grad_norm": 1.5524841474443225, + "learning_rate": 6.271290213195047e-07, + "loss": 0.298, + "step": 38329 + }, + { + "epoch": 0.8437156064274708, + "grad_norm": 1.5100574875121828, + "learning_rate": 6.269561896866733e-07, + "loss": 0.2893, + "step": 38330 + }, + { + "epoch": 0.8437376183138895, + "grad_norm": 1.9302310334796462, + "learning_rate": 6.267833802796003e-07, + "loss": 0.3167, + "step": 38331 + }, + { + "epoch": 0.8437596302003082, + "grad_norm": 2.1841028571107852, + "learning_rate": 6.266105930991645e-07, + "loss": 0.2646, + "step": 38332 + }, + { + "epoch": 0.8437816420867268, + "grad_norm": 2.8655800117027628, + "learning_rate": 6.264378281462458e-07, + "loss": 0.3039, + "step": 38333 + }, + { + "epoch": 0.8438036539731455, + "grad_norm": 1.99347245057101, + "learning_rate": 6.26265085421719e-07, + "loss": 0.2745, + "step": 38334 + }, + { + "epoch": 0.8438256658595642, + "grad_norm": 0.6087215192296075, + "learning_rate": 6.26092364926465e-07, + "loss": 0.4508, + "step": 38335 + }, + { + "epoch": 0.8438476777459828, + "grad_norm": 4.1675121142162235, + "learning_rate": 6.259196666613593e-07, + "loss": 0.3745, + "step": 38336 + }, + { + "epoch": 0.8438696896324015, + "grad_norm": 1.754945640569506, + "learning_rate": 6.257469906272811e-07, + "loss": 0.3772, + "step": 38337 + }, + { + "epoch": 0.8438917015188202, + "grad_norm": 1.5114521020227012, + "learning_rate": 6.255743368251083e-07, + "loss": 0.3104, + "step": 38338 + }, + { + "epoch": 0.8439137134052388, + "grad_norm": 1.7306409771906046, + "learning_rate": 6.25401705255717e-07, + "loss": 0.3081, + "step": 38339 + }, + { + "epoch": 0.8439357252916575, + "grad_norm": 1.5394901087129331, + "learning_rate": 6.252290959199864e-07, + "loss": 0.2594, + "step": 38340 + }, + { + "epoch": 0.8439577371780762, + "grad_norm": 1.7816587968899436, + "learning_rate": 6.250565088187915e-07, + "loss": 0.332, + "step": 38341 + }, + { + "epoch": 0.8439797490644948, + "grad_norm": 2.495262229771202, + "learning_rate": 6.248839439530108e-07, + "loss": 0.1879, + "step": 38342 + }, + { + "epoch": 0.8440017609509135, + "grad_norm": 2.1298103878791714, + "learning_rate": 6.247114013235222e-07, + "loss": 0.2773, + "step": 38343 + }, + { + "epoch": 0.8440237728373322, + "grad_norm": 1.4642848957122307, + "learning_rate": 6.24538880931202e-07, + "loss": 0.3264, + "step": 38344 + }, + { + "epoch": 0.8440457847237508, + "grad_norm": 2.210283085077701, + "learning_rate": 6.243663827769253e-07, + "loss": 0.2627, + "step": 38345 + }, + { + "epoch": 0.8440677966101695, + "grad_norm": 1.544575125858363, + "learning_rate": 6.241939068615705e-07, + "loss": 0.2818, + "step": 38346 + }, + { + "epoch": 0.8440898084965882, + "grad_norm": 1.4582066645695944, + "learning_rate": 6.240214531860139e-07, + "loss": 0.2951, + "step": 38347 + }, + { + "epoch": 0.8441118203830068, + "grad_norm": 0.6221133762686623, + "learning_rate": 6.238490217511328e-07, + "loss": 0.4213, + "step": 38348 + }, + { + "epoch": 0.8441338322694255, + "grad_norm": 1.4270197399890232, + "learning_rate": 6.236766125578031e-07, + "loss": 0.3123, + "step": 38349 + }, + { + "epoch": 0.8441558441558441, + "grad_norm": 1.9911954088092203, + "learning_rate": 6.235042256068997e-07, + "loss": 0.2718, + "step": 38350 + }, + { + "epoch": 0.8441778560422628, + "grad_norm": 2.4932059877713786, + "learning_rate": 6.233318608992999e-07, + "loss": 0.312, + "step": 38351 + }, + { + "epoch": 0.8441998679286815, + "grad_norm": 0.6255233142302734, + "learning_rate": 6.231595184358791e-07, + "loss": 0.455, + "step": 38352 + }, + { + "epoch": 0.8442218798151001, + "grad_norm": 0.6356756807210617, + "learning_rate": 6.22987198217515e-07, + "loss": 0.456, + "step": 38353 + }, + { + "epoch": 0.8442438917015188, + "grad_norm": 1.6865683605300135, + "learning_rate": 6.228149002450823e-07, + "loss": 0.2874, + "step": 38354 + }, + { + "epoch": 0.8442659035879375, + "grad_norm": 1.3195021441660484, + "learning_rate": 6.226426245194556e-07, + "loss": 0.2809, + "step": 38355 + }, + { + "epoch": 0.8442879154743561, + "grad_norm": 1.6516911612742204, + "learning_rate": 6.224703710415115e-07, + "loss": 0.2327, + "step": 38356 + }, + { + "epoch": 0.8443099273607748, + "grad_norm": 2.033838156210295, + "learning_rate": 6.222981398121253e-07, + "loss": 0.262, + "step": 38357 + }, + { + "epoch": 0.8443319392471935, + "grad_norm": 2.241066961524233, + "learning_rate": 6.221259308321731e-07, + "loss": 0.4354, + "step": 38358 + }, + { + "epoch": 0.8443539511336121, + "grad_norm": 0.6048496476076317, + "learning_rate": 6.219537441025303e-07, + "loss": 0.462, + "step": 38359 + }, + { + "epoch": 0.8443759630200308, + "grad_norm": 2.0311050876756536, + "learning_rate": 6.217815796240695e-07, + "loss": 0.2778, + "step": 38360 + }, + { + "epoch": 0.8443979749064495, + "grad_norm": 1.4520694793156217, + "learning_rate": 6.216094373976689e-07, + "loss": 0.296, + "step": 38361 + }, + { + "epoch": 0.8444199867928681, + "grad_norm": 1.4406263947670697, + "learning_rate": 6.214373174241994e-07, + "loss": 0.2374, + "step": 38362 + }, + { + "epoch": 0.8444419986792868, + "grad_norm": 1.8827295571264622, + "learning_rate": 6.21265219704541e-07, + "loss": 0.2892, + "step": 38363 + }, + { + "epoch": 0.8444640105657055, + "grad_norm": 1.788000116300403, + "learning_rate": 6.210931442395651e-07, + "loss": 0.2131, + "step": 38364 + }, + { + "epoch": 0.8444860224521241, + "grad_norm": 1.4441784187983047, + "learning_rate": 6.209210910301467e-07, + "loss": 0.2676, + "step": 38365 + }, + { + "epoch": 0.8445080343385428, + "grad_norm": 1.5855284948893387, + "learning_rate": 6.207490600771609e-07, + "loss": 0.2911, + "step": 38366 + }, + { + "epoch": 0.8445300462249615, + "grad_norm": 5.405421662135261, + "learning_rate": 6.205770513814807e-07, + "loss": 0.2918, + "step": 38367 + }, + { + "epoch": 0.8445520581113801, + "grad_norm": 1.8050740508732934, + "learning_rate": 6.204050649439808e-07, + "loss": 0.2791, + "step": 38368 + }, + { + "epoch": 0.8445740699977988, + "grad_norm": 1.44060338948501, + "learning_rate": 6.202331007655371e-07, + "loss": 0.2927, + "step": 38369 + }, + { + "epoch": 0.8445960818842175, + "grad_norm": 1.8048452224187543, + "learning_rate": 6.200611588470212e-07, + "loss": 0.3427, + "step": 38370 + }, + { + "epoch": 0.8446180937706361, + "grad_norm": 0.6177551357626575, + "learning_rate": 6.198892391893091e-07, + "loss": 0.4362, + "step": 38371 + }, + { + "epoch": 0.8446401056570548, + "grad_norm": 0.613529101389657, + "learning_rate": 6.197173417932723e-07, + "loss": 0.4292, + "step": 38372 + }, + { + "epoch": 0.8446621175434734, + "grad_norm": 0.6144731854870323, + "learning_rate": 6.195454666597855e-07, + "loss": 0.4252, + "step": 38373 + }, + { + "epoch": 0.8446841294298921, + "grad_norm": 1.508644115430714, + "learning_rate": 6.193736137897238e-07, + "loss": 0.2481, + "step": 38374 + }, + { + "epoch": 0.8447061413163108, + "grad_norm": 1.6750253823178132, + "learning_rate": 6.19201783183958e-07, + "loss": 0.3523, + "step": 38375 + }, + { + "epoch": 0.8447281532027294, + "grad_norm": 5.851294423218309, + "learning_rate": 6.190299748433637e-07, + "loss": 0.2259, + "step": 38376 + }, + { + "epoch": 0.8447501650891481, + "grad_norm": 0.6154010640617918, + "learning_rate": 6.188581887688116e-07, + "loss": 0.4383, + "step": 38377 + }, + { + "epoch": 0.8447721769755668, + "grad_norm": 1.6349884852050516, + "learning_rate": 6.186864249611762e-07, + "loss": 0.2867, + "step": 38378 + }, + { + "epoch": 0.8447941888619854, + "grad_norm": 1.4878753186209939, + "learning_rate": 6.185146834213324e-07, + "loss": 0.2759, + "step": 38379 + }, + { + "epoch": 0.8448162007484041, + "grad_norm": 2.8376901378278747, + "learning_rate": 6.183429641501493e-07, + "loss": 0.3078, + "step": 38380 + }, + { + "epoch": 0.8448382126348228, + "grad_norm": 0.6320442640770758, + "learning_rate": 6.181712671485029e-07, + "loss": 0.4737, + "step": 38381 + }, + { + "epoch": 0.8448602245212414, + "grad_norm": 1.6973172695065801, + "learning_rate": 6.179995924172633e-07, + "loss": 0.3222, + "step": 38382 + }, + { + "epoch": 0.8448822364076601, + "grad_norm": 1.7213519297203783, + "learning_rate": 6.178279399573045e-07, + "loss": 0.2293, + "step": 38383 + }, + { + "epoch": 0.8449042482940788, + "grad_norm": 1.7384229414827814, + "learning_rate": 6.176563097694993e-07, + "loss": 0.318, + "step": 38384 + }, + { + "epoch": 0.8449262601804974, + "grad_norm": 1.3670619659398682, + "learning_rate": 6.174847018547198e-07, + "loss": 0.2766, + "step": 38385 + }, + { + "epoch": 0.8449482720669161, + "grad_norm": 1.605768579762498, + "learning_rate": 6.173131162138363e-07, + "loss": 0.3237, + "step": 38386 + }, + { + "epoch": 0.8449702839533348, + "grad_norm": 1.7409259167969304, + "learning_rate": 6.171415528477221e-07, + "loss": 0.2756, + "step": 38387 + }, + { + "epoch": 0.8449922958397534, + "grad_norm": 2.195659710057109, + "learning_rate": 6.169700117572497e-07, + "loss": 0.3548, + "step": 38388 + }, + { + "epoch": 0.8450143077261721, + "grad_norm": 5.376831569310257, + "learning_rate": 6.167984929432913e-07, + "loss": 0.3466, + "step": 38389 + }, + { + "epoch": 0.8450363196125908, + "grad_norm": 1.7792866402523422, + "learning_rate": 6.166269964067184e-07, + "loss": 0.3002, + "step": 38390 + }, + { + "epoch": 0.8450583314990094, + "grad_norm": 1.8328976379385722, + "learning_rate": 6.164555221484009e-07, + "loss": 0.283, + "step": 38391 + }, + { + "epoch": 0.8450803433854281, + "grad_norm": 8.169845415932363, + "learning_rate": 6.162840701692113e-07, + "loss": 0.2305, + "step": 38392 + }, + { + "epoch": 0.8451023552718469, + "grad_norm": 2.0783673545975394, + "learning_rate": 6.161126404700213e-07, + "loss": 0.2381, + "step": 38393 + }, + { + "epoch": 0.8451243671582654, + "grad_norm": 0.6139209820282221, + "learning_rate": 6.159412330517034e-07, + "loss": 0.4225, + "step": 38394 + }, + { + "epoch": 0.8451463790446841, + "grad_norm": 1.476008374209137, + "learning_rate": 6.157698479151275e-07, + "loss": 0.3137, + "step": 38395 + }, + { + "epoch": 0.8451683909311027, + "grad_norm": 1.9073870301416922, + "learning_rate": 6.155984850611635e-07, + "loss": 0.2938, + "step": 38396 + }, + { + "epoch": 0.8451904028175214, + "grad_norm": 1.8198349271201437, + "learning_rate": 6.154271444906846e-07, + "loss": 0.3236, + "step": 38397 + }, + { + "epoch": 0.8452124147039402, + "grad_norm": 1.5463717397094967, + "learning_rate": 6.152558262045583e-07, + "loss": 0.29, + "step": 38398 + }, + { + "epoch": 0.8452344265903587, + "grad_norm": 6.140265734585001, + "learning_rate": 6.150845302036596e-07, + "loss": 0.2499, + "step": 38399 + }, + { + "epoch": 0.8452564384767774, + "grad_norm": 1.9419661297098612, + "learning_rate": 6.149132564888572e-07, + "loss": 0.2339, + "step": 38400 + }, + { + "epoch": 0.8452784503631962, + "grad_norm": 1.9627195788776035, + "learning_rate": 6.147420050610203e-07, + "loss": 0.3183, + "step": 38401 + }, + { + "epoch": 0.8453004622496147, + "grad_norm": 1.545907780335286, + "learning_rate": 6.145707759210213e-07, + "loss": 0.3224, + "step": 38402 + }, + { + "epoch": 0.8453224741360335, + "grad_norm": 0.6271974853979595, + "learning_rate": 6.143995690697285e-07, + "loss": 0.4567, + "step": 38403 + }, + { + "epoch": 0.8453444860224522, + "grad_norm": 1.7936893945785526, + "learning_rate": 6.142283845080132e-07, + "loss": 0.2653, + "step": 38404 + }, + { + "epoch": 0.8453664979088708, + "grad_norm": 1.5349504157350684, + "learning_rate": 6.140572222367469e-07, + "loss": 0.2943, + "step": 38405 + }, + { + "epoch": 0.8453885097952895, + "grad_norm": 3.2334332808487343, + "learning_rate": 6.138860822567965e-07, + "loss": 0.3021, + "step": 38406 + }, + { + "epoch": 0.8454105216817082, + "grad_norm": 0.6099550847300467, + "learning_rate": 6.137149645690343e-07, + "loss": 0.4542, + "step": 38407 + }, + { + "epoch": 0.8454325335681268, + "grad_norm": 1.7505468221476692, + "learning_rate": 6.135438691743284e-07, + "loss": 0.3604, + "step": 38408 + }, + { + "epoch": 0.8454545454545455, + "grad_norm": 3.727134364191264, + "learning_rate": 6.133727960735486e-07, + "loss": 0.2156, + "step": 38409 + }, + { + "epoch": 0.8454765573409642, + "grad_norm": 1.6275120284849922, + "learning_rate": 6.132017452675654e-07, + "loss": 0.2271, + "step": 38410 + }, + { + "epoch": 0.8454985692273828, + "grad_norm": 2.0878085983919017, + "learning_rate": 6.130307167572469e-07, + "loss": 0.3088, + "step": 38411 + }, + { + "epoch": 0.8455205811138015, + "grad_norm": 2.281373827855333, + "learning_rate": 6.128597105434641e-07, + "loss": 0.2537, + "step": 38412 + }, + { + "epoch": 0.8455425930002202, + "grad_norm": 1.8330420270236516, + "learning_rate": 6.126887266270836e-07, + "loss": 0.3307, + "step": 38413 + }, + { + "epoch": 0.8455646048866388, + "grad_norm": 0.6066329566092837, + "learning_rate": 6.125177650089758e-07, + "loss": 0.4597, + "step": 38414 + }, + { + "epoch": 0.8455866167730575, + "grad_norm": 1.5572591195657832, + "learning_rate": 6.123468256900111e-07, + "loss": 0.2377, + "step": 38415 + }, + { + "epoch": 0.8456086286594762, + "grad_norm": 1.6782038028498294, + "learning_rate": 6.121759086710555e-07, + "loss": 0.2727, + "step": 38416 + }, + { + "epoch": 0.8456306405458948, + "grad_norm": 1.5058378347737578, + "learning_rate": 6.120050139529798e-07, + "loss": 0.3096, + "step": 38417 + }, + { + "epoch": 0.8456526524323135, + "grad_norm": 0.6181353300610525, + "learning_rate": 6.118341415366507e-07, + "loss": 0.4592, + "step": 38418 + }, + { + "epoch": 0.8456746643187321, + "grad_norm": 3.6450079957262114, + "learning_rate": 6.116632914229381e-07, + "loss": 0.3432, + "step": 38419 + }, + { + "epoch": 0.8456966762051508, + "grad_norm": 2.227403855713296, + "learning_rate": 6.114924636127107e-07, + "loss": 0.2718, + "step": 38420 + }, + { + "epoch": 0.8457186880915695, + "grad_norm": 1.3532982670175766, + "learning_rate": 6.113216581068343e-07, + "loss": 0.1805, + "step": 38421 + }, + { + "epoch": 0.8457406999779881, + "grad_norm": 1.8243298064905575, + "learning_rate": 6.111508749061807e-07, + "loss": 0.2082, + "step": 38422 + }, + { + "epoch": 0.8457627118644068, + "grad_norm": 1.3455700278179927, + "learning_rate": 6.109801140116139e-07, + "loss": 0.2557, + "step": 38423 + }, + { + "epoch": 0.8457847237508255, + "grad_norm": 0.5847244711207802, + "learning_rate": 6.108093754240041e-07, + "loss": 0.4076, + "step": 38424 + }, + { + "epoch": 0.8458067356372441, + "grad_norm": 1.689513001479759, + "learning_rate": 6.106386591442198e-07, + "loss": 0.3576, + "step": 38425 + }, + { + "epoch": 0.8458287475236628, + "grad_norm": 2.6824682284567505, + "learning_rate": 6.104679651731277e-07, + "loss": 0.2544, + "step": 38426 + }, + { + "epoch": 0.8458507594100815, + "grad_norm": 1.6087215267213282, + "learning_rate": 6.102972935115936e-07, + "loss": 0.3005, + "step": 38427 + }, + { + "epoch": 0.8458727712965001, + "grad_norm": 1.8862708907637251, + "learning_rate": 6.10126644160487e-07, + "loss": 0.2406, + "step": 38428 + }, + { + "epoch": 0.8458947831829188, + "grad_norm": 1.6726763787575267, + "learning_rate": 6.099560171206748e-07, + "loss": 0.2632, + "step": 38429 + }, + { + "epoch": 0.8459167950693375, + "grad_norm": 2.2607623959235768, + "learning_rate": 6.097854123930247e-07, + "loss": 0.2637, + "step": 38430 + }, + { + "epoch": 0.8459388069557561, + "grad_norm": 1.7276544505833302, + "learning_rate": 6.096148299784033e-07, + "loss": 0.2797, + "step": 38431 + }, + { + "epoch": 0.8459608188421748, + "grad_norm": 1.5693994830012263, + "learning_rate": 6.09444269877677e-07, + "loss": 0.2477, + "step": 38432 + }, + { + "epoch": 0.8459828307285935, + "grad_norm": 0.6200795315084162, + "learning_rate": 6.092737320917136e-07, + "loss": 0.4387, + "step": 38433 + }, + { + "epoch": 0.8460048426150121, + "grad_norm": 0.6364620870872927, + "learning_rate": 6.091032166213778e-07, + "loss": 0.4535, + "step": 38434 + }, + { + "epoch": 0.8460268545014308, + "grad_norm": 2.4393853102015304, + "learning_rate": 6.089327234675397e-07, + "loss": 0.3242, + "step": 38435 + }, + { + "epoch": 0.8460488663878495, + "grad_norm": 1.5509481448058993, + "learning_rate": 6.087622526310638e-07, + "loss": 0.2832, + "step": 38436 + }, + { + "epoch": 0.8460708782742681, + "grad_norm": 1.7765066569943104, + "learning_rate": 6.085918041128158e-07, + "loss": 0.2539, + "step": 38437 + }, + { + "epoch": 0.8460928901606868, + "grad_norm": 1.6250727171444748, + "learning_rate": 6.084213779136639e-07, + "loss": 0.3479, + "step": 38438 + }, + { + "epoch": 0.8461149020471055, + "grad_norm": 1.5650772059470117, + "learning_rate": 6.082509740344722e-07, + "loss": 0.2307, + "step": 38439 + }, + { + "epoch": 0.8461369139335241, + "grad_norm": 2.1283297474133653, + "learning_rate": 6.08080592476108e-07, + "loss": 0.3069, + "step": 38440 + }, + { + "epoch": 0.8461589258199428, + "grad_norm": 0.5954264131469186, + "learning_rate": 6.079102332394377e-07, + "loss": 0.421, + "step": 38441 + }, + { + "epoch": 0.8461809377063614, + "grad_norm": 1.5181878457926203, + "learning_rate": 6.077398963253262e-07, + "loss": 0.2553, + "step": 38442 + }, + { + "epoch": 0.8462029495927801, + "grad_norm": 1.8415996062789817, + "learning_rate": 6.075695817346399e-07, + "loss": 0.2776, + "step": 38443 + }, + { + "epoch": 0.8462249614791988, + "grad_norm": 0.6180623200434251, + "learning_rate": 6.073992894682435e-07, + "loss": 0.4315, + "step": 38444 + }, + { + "epoch": 0.8462469733656174, + "grad_norm": 1.62967406983052, + "learning_rate": 6.072290195270025e-07, + "loss": 0.243, + "step": 38445 + }, + { + "epoch": 0.8462689852520361, + "grad_norm": 1.7162124761918354, + "learning_rate": 6.070587719117843e-07, + "loss": 0.1902, + "step": 38446 + }, + { + "epoch": 0.8462909971384548, + "grad_norm": 1.8327223857391166, + "learning_rate": 6.068885466234514e-07, + "loss": 0.2641, + "step": 38447 + }, + { + "epoch": 0.8463130090248734, + "grad_norm": 2.6449711462571606, + "learning_rate": 6.067183436628716e-07, + "loss": 0.4029, + "step": 38448 + }, + { + "epoch": 0.8463350209112921, + "grad_norm": 0.5931906466993015, + "learning_rate": 6.065481630309078e-07, + "loss": 0.4319, + "step": 38449 + }, + { + "epoch": 0.8463570327977108, + "grad_norm": 2.3371242192227206, + "learning_rate": 6.063780047284256e-07, + "loss": 0.3189, + "step": 38450 + }, + { + "epoch": 0.8463790446841294, + "grad_norm": 1.677234380211192, + "learning_rate": 6.062078687562906e-07, + "loss": 0.2219, + "step": 38451 + }, + { + "epoch": 0.8464010565705481, + "grad_norm": 0.6160985332873462, + "learning_rate": 6.060377551153663e-07, + "loss": 0.4373, + "step": 38452 + }, + { + "epoch": 0.8464230684569668, + "grad_norm": 2.5668295060026844, + "learning_rate": 6.058676638065186e-07, + "loss": 0.3301, + "step": 38453 + }, + { + "epoch": 0.8464450803433854, + "grad_norm": 0.6086894564975027, + "learning_rate": 6.056975948306109e-07, + "loss": 0.4323, + "step": 38454 + }, + { + "epoch": 0.8464670922298041, + "grad_norm": 1.5389083630147833, + "learning_rate": 6.055275481885075e-07, + "loss": 0.3184, + "step": 38455 + }, + { + "epoch": 0.8464891041162228, + "grad_norm": 3.8811464895874423, + "learning_rate": 6.053575238810739e-07, + "loss": 0.421, + "step": 38456 + }, + { + "epoch": 0.8465111160026414, + "grad_norm": 2.090299091247387, + "learning_rate": 6.051875219091724e-07, + "loss": 0.2861, + "step": 38457 + }, + { + "epoch": 0.8465331278890601, + "grad_norm": 3.0591124363804676, + "learning_rate": 6.050175422736698e-07, + "loss": 0.2674, + "step": 38458 + }, + { + "epoch": 0.8465551397754788, + "grad_norm": 2.787701651252005, + "learning_rate": 6.048475849754265e-07, + "loss": 0.2822, + "step": 38459 + }, + { + "epoch": 0.8465771516618974, + "grad_norm": 1.8505130771523044, + "learning_rate": 6.046776500153085e-07, + "loss": 0.2674, + "step": 38460 + }, + { + "epoch": 0.8465991635483161, + "grad_norm": 2.1098988714157643, + "learning_rate": 6.045077373941799e-07, + "loss": 0.2595, + "step": 38461 + }, + { + "epoch": 0.8466211754347348, + "grad_norm": 0.6057365836203754, + "learning_rate": 6.043378471129024e-07, + "loss": 0.4534, + "step": 38462 + }, + { + "epoch": 0.8466431873211534, + "grad_norm": 2.1833224514747553, + "learning_rate": 6.041679791723415e-07, + "loss": 0.2997, + "step": 38463 + }, + { + "epoch": 0.8466651992075721, + "grad_norm": 2.1128302393101586, + "learning_rate": 6.039981335733591e-07, + "loss": 0.2506, + "step": 38464 + }, + { + "epoch": 0.8466872110939908, + "grad_norm": 0.6164263212302832, + "learning_rate": 6.038283103168175e-07, + "loss": 0.4257, + "step": 38465 + }, + { + "epoch": 0.8467092229804094, + "grad_norm": 1.7029177412777579, + "learning_rate": 6.036585094035829e-07, + "loss": 0.3458, + "step": 38466 + }, + { + "epoch": 0.8467312348668281, + "grad_norm": 1.1624763858252722, + "learning_rate": 6.034887308345167e-07, + "loss": 0.3225, + "step": 38467 + }, + { + "epoch": 0.8467532467532467, + "grad_norm": 1.3473401523428894, + "learning_rate": 6.033189746104806e-07, + "loss": 0.2745, + "step": 38468 + }, + { + "epoch": 0.8467752586396654, + "grad_norm": 1.5010389240339737, + "learning_rate": 6.031492407323392e-07, + "loss": 0.2412, + "step": 38469 + }, + { + "epoch": 0.8467972705260841, + "grad_norm": 1.8885391557821207, + "learning_rate": 6.029795292009527e-07, + "loss": 0.2837, + "step": 38470 + }, + { + "epoch": 0.8468192824125027, + "grad_norm": 0.6216422785182063, + "learning_rate": 6.028098400171872e-07, + "loss": 0.4571, + "step": 38471 + }, + { + "epoch": 0.8468412942989214, + "grad_norm": 0.6247793406646267, + "learning_rate": 6.026401731819037e-07, + "loss": 0.4758, + "step": 38472 + }, + { + "epoch": 0.8468633061853401, + "grad_norm": 1.9004729312720678, + "learning_rate": 6.024705286959626e-07, + "loss": 0.2115, + "step": 38473 + }, + { + "epoch": 0.8468853180717587, + "grad_norm": 6.06990129233614, + "learning_rate": 6.023009065602287e-07, + "loss": 0.3069, + "step": 38474 + }, + { + "epoch": 0.8469073299581774, + "grad_norm": 1.419313064469819, + "learning_rate": 6.021313067755624e-07, + "loss": 0.3034, + "step": 38475 + }, + { + "epoch": 0.8469293418445961, + "grad_norm": 0.6202904357908507, + "learning_rate": 6.019617293428259e-07, + "loss": 0.45, + "step": 38476 + }, + { + "epoch": 0.8469513537310147, + "grad_norm": 1.4616896630850544, + "learning_rate": 6.017921742628829e-07, + "loss": 0.2818, + "step": 38477 + }, + { + "epoch": 0.8469733656174334, + "grad_norm": 1.9304363680355539, + "learning_rate": 6.016226415365923e-07, + "loss": 0.3391, + "step": 38478 + }, + { + "epoch": 0.8469953775038521, + "grad_norm": 1.9787763136066692, + "learning_rate": 6.014531311648181e-07, + "loss": 0.2905, + "step": 38479 + }, + { + "epoch": 0.8470173893902707, + "grad_norm": 5.608816192737192, + "learning_rate": 6.012836431484203e-07, + "loss": 0.3009, + "step": 38480 + }, + { + "epoch": 0.8470394012766894, + "grad_norm": 1.3489021510535857, + "learning_rate": 6.011141774882606e-07, + "loss": 0.295, + "step": 38481 + }, + { + "epoch": 0.8470614131631081, + "grad_norm": 1.6086863763969645, + "learning_rate": 6.009447341852021e-07, + "loss": 0.3239, + "step": 38482 + }, + { + "epoch": 0.8470834250495267, + "grad_norm": 1.7732117970441004, + "learning_rate": 6.007753132401029e-07, + "loss": 0.2744, + "step": 38483 + }, + { + "epoch": 0.8471054369359454, + "grad_norm": 0.6196642735881128, + "learning_rate": 6.006059146538268e-07, + "loss": 0.4626, + "step": 38484 + }, + { + "epoch": 0.8471274488223641, + "grad_norm": 1.5827536953519497, + "learning_rate": 6.004365384272326e-07, + "loss": 0.2972, + "step": 38485 + }, + { + "epoch": 0.8471494607087827, + "grad_norm": 0.6176403221158726, + "learning_rate": 6.002671845611824e-07, + "loss": 0.467, + "step": 38486 + }, + { + "epoch": 0.8471714725952014, + "grad_norm": 1.495423896827912, + "learning_rate": 6.000978530565377e-07, + "loss": 0.2605, + "step": 38487 + }, + { + "epoch": 0.8471934844816201, + "grad_norm": 1.4748207157560635, + "learning_rate": 5.999285439141567e-07, + "loss": 0.2721, + "step": 38488 + }, + { + "epoch": 0.8472154963680387, + "grad_norm": 3.306442366623387, + "learning_rate": 5.997592571349026e-07, + "loss": 0.2545, + "step": 38489 + }, + { + "epoch": 0.8472375082544574, + "grad_norm": 1.6310301754820824, + "learning_rate": 5.995899927196335e-07, + "loss": 0.2063, + "step": 38490 + }, + { + "epoch": 0.847259520140876, + "grad_norm": 1.4825706690726554, + "learning_rate": 5.994207506692112e-07, + "loss": 0.3803, + "step": 38491 + }, + { + "epoch": 0.8472815320272947, + "grad_norm": 1.9310979057058528, + "learning_rate": 5.99251530984496e-07, + "loss": 0.3331, + "step": 38492 + }, + { + "epoch": 0.8473035439137134, + "grad_norm": 7.234013551903742, + "learning_rate": 5.99082333666346e-07, + "loss": 0.339, + "step": 38493 + }, + { + "epoch": 0.847325555800132, + "grad_norm": 1.5594079549677233, + "learning_rate": 5.989131587156238e-07, + "loss": 0.2648, + "step": 38494 + }, + { + "epoch": 0.8473475676865507, + "grad_norm": 1.7935767883726939, + "learning_rate": 5.987440061331868e-07, + "loss": 0.3279, + "step": 38495 + }, + { + "epoch": 0.8473695795729694, + "grad_norm": 1.7189674339248215, + "learning_rate": 5.985748759198956e-07, + "loss": 0.2994, + "step": 38496 + }, + { + "epoch": 0.847391591459388, + "grad_norm": 2.129078290269919, + "learning_rate": 5.984057680766109e-07, + "loss": 0.3607, + "step": 38497 + }, + { + "epoch": 0.8474136033458067, + "grad_norm": 1.7726968399718728, + "learning_rate": 5.982366826041902e-07, + "loss": 0.3327, + "step": 38498 + }, + { + "epoch": 0.8474356152322254, + "grad_norm": 1.2925225120510626, + "learning_rate": 5.980676195034957e-07, + "loss": 0.2805, + "step": 38499 + }, + { + "epoch": 0.847457627118644, + "grad_norm": 1.8362259671929384, + "learning_rate": 5.978985787753839e-07, + "loss": 0.3161, + "step": 38500 + }, + { + "epoch": 0.8474796390050627, + "grad_norm": 1.1923916363503786, + "learning_rate": 5.977295604207134e-07, + "loss": 0.246, + "step": 38501 + }, + { + "epoch": 0.8475016508914814, + "grad_norm": 1.5538443626262886, + "learning_rate": 5.975605644403465e-07, + "loss": 0.3272, + "step": 38502 + }, + { + "epoch": 0.8475236627779, + "grad_norm": 1.5801479891671881, + "learning_rate": 5.973915908351408e-07, + "loss": 0.2542, + "step": 38503 + }, + { + "epoch": 0.8475456746643187, + "grad_norm": 1.3768670671025915, + "learning_rate": 5.972226396059533e-07, + "loss": 0.33, + "step": 38504 + }, + { + "epoch": 0.8475676865507374, + "grad_norm": 1.8105923637380974, + "learning_rate": 5.97053710753645e-07, + "loss": 0.2686, + "step": 38505 + }, + { + "epoch": 0.847589698437156, + "grad_norm": 1.522516680329637, + "learning_rate": 5.96884804279072e-07, + "loss": 0.2527, + "step": 38506 + }, + { + "epoch": 0.8476117103235747, + "grad_norm": 1.4069904286724912, + "learning_rate": 5.967159201830958e-07, + "loss": 0.2867, + "step": 38507 + }, + { + "epoch": 0.8476337222099934, + "grad_norm": 2.320430574073696, + "learning_rate": 5.965470584665733e-07, + "loss": 0.2557, + "step": 38508 + }, + { + "epoch": 0.847655734096412, + "grad_norm": 1.8404056551507013, + "learning_rate": 5.963782191303619e-07, + "loss": 0.3812, + "step": 38509 + }, + { + "epoch": 0.8476777459828307, + "grad_norm": 1.7809225993150959, + "learning_rate": 5.962094021753212e-07, + "loss": 0.3279, + "step": 38510 + }, + { + "epoch": 0.8476997578692494, + "grad_norm": 1.5091817851212972, + "learning_rate": 5.960406076023074e-07, + "loss": 0.2471, + "step": 38511 + }, + { + "epoch": 0.847721769755668, + "grad_norm": 1.5365143811109907, + "learning_rate": 5.958718354121796e-07, + "loss": 0.3305, + "step": 38512 + }, + { + "epoch": 0.8477437816420867, + "grad_norm": 2.887500093040931, + "learning_rate": 5.95703085605796e-07, + "loss": 0.2411, + "step": 38513 + }, + { + "epoch": 0.8477657935285053, + "grad_norm": 0.6232259650055151, + "learning_rate": 5.955343581840134e-07, + "loss": 0.434, + "step": 38514 + }, + { + "epoch": 0.847787805414924, + "grad_norm": 3.51430833555464, + "learning_rate": 5.953656531476903e-07, + "loss": 0.3448, + "step": 38515 + }, + { + "epoch": 0.8478098173013427, + "grad_norm": 1.8481245033025329, + "learning_rate": 5.951969704976823e-07, + "loss": 0.3332, + "step": 38516 + }, + { + "epoch": 0.8478318291877613, + "grad_norm": 1.6067850043330099, + "learning_rate": 5.95028310234848e-07, + "loss": 0.2347, + "step": 38517 + }, + { + "epoch": 0.84785384107418, + "grad_norm": 1.6497997670094988, + "learning_rate": 5.948596723600458e-07, + "loss": 0.2319, + "step": 38518 + }, + { + "epoch": 0.8478758529605988, + "grad_norm": 1.8309167336286518, + "learning_rate": 5.946910568741299e-07, + "loss": 0.3129, + "step": 38519 + }, + { + "epoch": 0.8478978648470173, + "grad_norm": 3.7954222706315863, + "learning_rate": 5.945224637779606e-07, + "loss": 0.3277, + "step": 38520 + }, + { + "epoch": 0.847919876733436, + "grad_norm": 1.5266900364493008, + "learning_rate": 5.943538930723914e-07, + "loss": 0.2376, + "step": 38521 + }, + { + "epoch": 0.8479418886198548, + "grad_norm": 0.6298276460860834, + "learning_rate": 5.941853447582807e-07, + "loss": 0.4895, + "step": 38522 + }, + { + "epoch": 0.8479639005062733, + "grad_norm": 1.9436381748388518, + "learning_rate": 5.940168188364864e-07, + "loss": 0.264, + "step": 38523 + }, + { + "epoch": 0.847985912392692, + "grad_norm": 1.6721338067436131, + "learning_rate": 5.938483153078628e-07, + "loss": 0.3744, + "step": 38524 + }, + { + "epoch": 0.8480079242791108, + "grad_norm": 1.4125051802906694, + "learning_rate": 5.936798341732686e-07, + "loss": 0.2893, + "step": 38525 + }, + { + "epoch": 0.8480299361655294, + "grad_norm": 2.246729959816717, + "learning_rate": 5.93511375433557e-07, + "loss": 0.2123, + "step": 38526 + }, + { + "epoch": 0.8480519480519481, + "grad_norm": 1.5740094961504256, + "learning_rate": 5.933429390895861e-07, + "loss": 0.2875, + "step": 38527 + }, + { + "epoch": 0.8480739599383668, + "grad_norm": 2.2864670129019973, + "learning_rate": 5.931745251422132e-07, + "loss": 0.2811, + "step": 38528 + }, + { + "epoch": 0.8480959718247854, + "grad_norm": 1.6988384674109092, + "learning_rate": 5.930061335922915e-07, + "loss": 0.2916, + "step": 38529 + }, + { + "epoch": 0.8481179837112041, + "grad_norm": 1.4006289673031156, + "learning_rate": 5.928377644406796e-07, + "loss": 0.2706, + "step": 38530 + }, + { + "epoch": 0.8481399955976228, + "grad_norm": 0.6524957981214933, + "learning_rate": 5.926694176882309e-07, + "loss": 0.4609, + "step": 38531 + }, + { + "epoch": 0.8481620074840414, + "grad_norm": 0.597602636685434, + "learning_rate": 5.925010933358011e-07, + "loss": 0.4374, + "step": 38532 + }, + { + "epoch": 0.8481840193704601, + "grad_norm": 1.6444583297374933, + "learning_rate": 5.92332791384248e-07, + "loss": 0.2647, + "step": 38533 + }, + { + "epoch": 0.8482060312568788, + "grad_norm": 1.8839179808146682, + "learning_rate": 5.921645118344249e-07, + "loss": 0.2891, + "step": 38534 + }, + { + "epoch": 0.8482280431432974, + "grad_norm": 1.234185460279393, + "learning_rate": 5.919962546871883e-07, + "loss": 0.2346, + "step": 38535 + }, + { + "epoch": 0.8482500550297161, + "grad_norm": 1.3045098071827501, + "learning_rate": 5.918280199433929e-07, + "loss": 0.257, + "step": 38536 + }, + { + "epoch": 0.8482720669161347, + "grad_norm": 1.6854467396215562, + "learning_rate": 5.916598076038915e-07, + "loss": 0.3277, + "step": 38537 + }, + { + "epoch": 0.8482940788025534, + "grad_norm": 1.5763671412882163, + "learning_rate": 5.914916176695434e-07, + "loss": 0.2595, + "step": 38538 + }, + { + "epoch": 0.8483160906889721, + "grad_norm": 0.6251834506794426, + "learning_rate": 5.913234501411996e-07, + "loss": 0.4339, + "step": 38539 + }, + { + "epoch": 0.8483381025753907, + "grad_norm": 1.8822860444187306, + "learning_rate": 5.911553050197178e-07, + "loss": 0.3483, + "step": 38540 + }, + { + "epoch": 0.8483601144618094, + "grad_norm": 0.6176717334508487, + "learning_rate": 5.90987182305951e-07, + "loss": 0.4593, + "step": 38541 + }, + { + "epoch": 0.8483821263482281, + "grad_norm": 1.7118195518827866, + "learning_rate": 5.90819082000752e-07, + "loss": 0.2589, + "step": 38542 + }, + { + "epoch": 0.8484041382346467, + "grad_norm": 1.7516684088387906, + "learning_rate": 5.90651004104979e-07, + "loss": 0.3089, + "step": 38543 + }, + { + "epoch": 0.8484261501210654, + "grad_norm": 0.5972425618563902, + "learning_rate": 5.904829486194841e-07, + "loss": 0.4558, + "step": 38544 + }, + { + "epoch": 0.8484481620074841, + "grad_norm": 1.9823143194211765, + "learning_rate": 5.903149155451204e-07, + "loss": 0.3283, + "step": 38545 + }, + { + "epoch": 0.8484701738939027, + "grad_norm": 1.740476990940271, + "learning_rate": 5.90146904882744e-07, + "loss": 0.281, + "step": 38546 + }, + { + "epoch": 0.8484921857803214, + "grad_norm": 1.7504863947543607, + "learning_rate": 5.89978916633207e-07, + "loss": 0.2949, + "step": 38547 + }, + { + "epoch": 0.8485141976667401, + "grad_norm": 0.6152225828150826, + "learning_rate": 5.898109507973643e-07, + "loss": 0.447, + "step": 38548 + }, + { + "epoch": 0.8485362095531587, + "grad_norm": 1.712449141705048, + "learning_rate": 5.8964300737607e-07, + "loss": 0.2654, + "step": 38549 + }, + { + "epoch": 0.8485582214395774, + "grad_norm": 1.7853899386214596, + "learning_rate": 5.894750863701765e-07, + "loss": 0.4141, + "step": 38550 + }, + { + "epoch": 0.8485802333259961, + "grad_norm": 1.8175476083590578, + "learning_rate": 5.893071877805385e-07, + "loss": 0.314, + "step": 38551 + }, + { + "epoch": 0.8486022452124147, + "grad_norm": 1.360999144784041, + "learning_rate": 5.891393116080074e-07, + "loss": 0.2329, + "step": 38552 + }, + { + "epoch": 0.8486242570988334, + "grad_norm": 3.484695070033263, + "learning_rate": 5.889714578534378e-07, + "loss": 0.239, + "step": 38553 + }, + { + "epoch": 0.8486462689852521, + "grad_norm": 1.7542015690341073, + "learning_rate": 5.888036265176834e-07, + "loss": 0.3667, + "step": 38554 + }, + { + "epoch": 0.8486682808716707, + "grad_norm": 1.4782353321663162, + "learning_rate": 5.886358176015955e-07, + "loss": 0.2962, + "step": 38555 + }, + { + "epoch": 0.8486902927580894, + "grad_norm": 2.37638494569587, + "learning_rate": 5.884680311060293e-07, + "loss": 0.3053, + "step": 38556 + }, + { + "epoch": 0.8487123046445081, + "grad_norm": 2.648962448862383, + "learning_rate": 5.883002670318344e-07, + "loss": 0.2997, + "step": 38557 + }, + { + "epoch": 0.8487343165309267, + "grad_norm": 1.9903461683007284, + "learning_rate": 5.881325253798659e-07, + "loss": 0.285, + "step": 38558 + }, + { + "epoch": 0.8487563284173454, + "grad_norm": 1.7403549358902906, + "learning_rate": 5.879648061509763e-07, + "loss": 0.3025, + "step": 38559 + }, + { + "epoch": 0.848778340303764, + "grad_norm": 2.6438138014413615, + "learning_rate": 5.877971093460167e-07, + "loss": 0.222, + "step": 38560 + }, + { + "epoch": 0.8488003521901827, + "grad_norm": 0.6386555344495856, + "learning_rate": 5.876294349658407e-07, + "loss": 0.4644, + "step": 38561 + }, + { + "epoch": 0.8488223640766014, + "grad_norm": 1.3155794297631958, + "learning_rate": 5.874617830112994e-07, + "loss": 0.2749, + "step": 38562 + }, + { + "epoch": 0.84884437596302, + "grad_norm": 1.4232599160754702, + "learning_rate": 5.872941534832449e-07, + "loss": 0.3013, + "step": 38563 + }, + { + "epoch": 0.8488663878494387, + "grad_norm": 2.1450843088228884, + "learning_rate": 5.87126546382531e-07, + "loss": 0.3112, + "step": 38564 + }, + { + "epoch": 0.8488883997358574, + "grad_norm": 1.6648914923788793, + "learning_rate": 5.869589617100069e-07, + "loss": 0.2661, + "step": 38565 + }, + { + "epoch": 0.848910411622276, + "grad_norm": 1.308814987543401, + "learning_rate": 5.867913994665264e-07, + "loss": 0.2805, + "step": 38566 + }, + { + "epoch": 0.8489324235086947, + "grad_norm": 1.5551567940449793, + "learning_rate": 5.8662385965294e-07, + "loss": 0.3762, + "step": 38567 + }, + { + "epoch": 0.8489544353951134, + "grad_norm": 1.6246395273270962, + "learning_rate": 5.864563422700992e-07, + "loss": 0.302, + "step": 38568 + }, + { + "epoch": 0.848976447281532, + "grad_norm": 18.181179374907973, + "learning_rate": 5.862888473188566e-07, + "loss": 0.3699, + "step": 38569 + }, + { + "epoch": 0.8489984591679507, + "grad_norm": 1.620455659534115, + "learning_rate": 5.861213748000616e-07, + "loss": 0.2453, + "step": 38570 + }, + { + "epoch": 0.8490204710543694, + "grad_norm": 1.8779774301531538, + "learning_rate": 5.859539247145679e-07, + "loss": 0.3738, + "step": 38571 + }, + { + "epoch": 0.849042482940788, + "grad_norm": 0.609632030608215, + "learning_rate": 5.857864970632249e-07, + "loss": 0.4449, + "step": 38572 + }, + { + "epoch": 0.8490644948272067, + "grad_norm": 1.6360894556958558, + "learning_rate": 5.856190918468813e-07, + "loss": 0.2241, + "step": 38573 + }, + { + "epoch": 0.8490865067136254, + "grad_norm": 1.7664960700029955, + "learning_rate": 5.854517090663931e-07, + "loss": 0.2897, + "step": 38574 + }, + { + "epoch": 0.849108518600044, + "grad_norm": 17.228815119671868, + "learning_rate": 5.852843487226068e-07, + "loss": 0.3592, + "step": 38575 + }, + { + "epoch": 0.8491305304864627, + "grad_norm": 2.0028976709116195, + "learning_rate": 5.851170108163751e-07, + "loss": 0.2903, + "step": 38576 + }, + { + "epoch": 0.8491525423728814, + "grad_norm": 2.5685666423369797, + "learning_rate": 5.849496953485484e-07, + "loss": 0.2893, + "step": 38577 + }, + { + "epoch": 0.8491745542593, + "grad_norm": 0.6060630459995149, + "learning_rate": 5.847824023199744e-07, + "loss": 0.4345, + "step": 38578 + }, + { + "epoch": 0.8491965661457187, + "grad_norm": 1.8170509838547242, + "learning_rate": 5.846151317315074e-07, + "loss": 0.3541, + "step": 38579 + }, + { + "epoch": 0.8492185780321374, + "grad_norm": 4.113529077155861, + "learning_rate": 5.844478835839945e-07, + "loss": 0.3325, + "step": 38580 + }, + { + "epoch": 0.849240589918556, + "grad_norm": 1.8535518967885338, + "learning_rate": 5.842806578782883e-07, + "loss": 0.2663, + "step": 38581 + }, + { + "epoch": 0.8492626018049747, + "grad_norm": 1.3767353572138883, + "learning_rate": 5.841134546152366e-07, + "loss": 0.2533, + "step": 38582 + }, + { + "epoch": 0.8492846136913933, + "grad_norm": 1.829521034413316, + "learning_rate": 5.839462737956897e-07, + "loss": 0.2972, + "step": 38583 + }, + { + "epoch": 0.849306625577812, + "grad_norm": 1.3464530984816128, + "learning_rate": 5.837791154204969e-07, + "loss": 0.3087, + "step": 38584 + }, + { + "epoch": 0.8493286374642307, + "grad_norm": 1.6010563428529374, + "learning_rate": 5.836119794905093e-07, + "loss": 0.231, + "step": 38585 + }, + { + "epoch": 0.8493506493506493, + "grad_norm": 1.6799630032675055, + "learning_rate": 5.834448660065745e-07, + "loss": 0.3313, + "step": 38586 + }, + { + "epoch": 0.849372661237068, + "grad_norm": 2.274230744638882, + "learning_rate": 5.832777749695434e-07, + "loss": 0.257, + "step": 38587 + }, + { + "epoch": 0.8493946731234867, + "grad_norm": 1.3436679730800491, + "learning_rate": 5.831107063802637e-07, + "loss": 0.2296, + "step": 38588 + }, + { + "epoch": 0.8494166850099053, + "grad_norm": 2.993262818621114, + "learning_rate": 5.829436602395861e-07, + "loss": 0.3657, + "step": 38589 + }, + { + "epoch": 0.849438696896324, + "grad_norm": 3.2169011703837014, + "learning_rate": 5.82776636548359e-07, + "loss": 0.3973, + "step": 38590 + }, + { + "epoch": 0.8494607087827427, + "grad_norm": 1.8600325083377534, + "learning_rate": 5.826096353074307e-07, + "loss": 0.3305, + "step": 38591 + }, + { + "epoch": 0.8494827206691613, + "grad_norm": 1.3166301127963325, + "learning_rate": 5.824426565176511e-07, + "loss": 0.3014, + "step": 38592 + }, + { + "epoch": 0.84950473255558, + "grad_norm": 0.5983572801542449, + "learning_rate": 5.822757001798673e-07, + "loss": 0.4658, + "step": 38593 + }, + { + "epoch": 0.8495267444419987, + "grad_norm": 0.6206075917674623, + "learning_rate": 5.821087662949288e-07, + "loss": 0.4425, + "step": 38594 + }, + { + "epoch": 0.8495487563284173, + "grad_norm": 2.476861987629492, + "learning_rate": 5.819418548636852e-07, + "loss": 0.3577, + "step": 38595 + }, + { + "epoch": 0.849570768214836, + "grad_norm": 1.4817207464828162, + "learning_rate": 5.81774965886982e-07, + "loss": 0.2273, + "step": 38596 + }, + { + "epoch": 0.8495927801012547, + "grad_norm": 2.0214752177459765, + "learning_rate": 5.816080993656708e-07, + "loss": 0.3322, + "step": 38597 + }, + { + "epoch": 0.8496147919876733, + "grad_norm": 2.229959709496035, + "learning_rate": 5.814412553005966e-07, + "loss": 0.2802, + "step": 38598 + }, + { + "epoch": 0.849636803874092, + "grad_norm": 1.8949476565862944, + "learning_rate": 5.812744336926085e-07, + "loss": 0.3787, + "step": 38599 + }, + { + "epoch": 0.8496588157605107, + "grad_norm": 1.7255946560601836, + "learning_rate": 5.811076345425554e-07, + "loss": 0.2248, + "step": 38600 + }, + { + "epoch": 0.8496808276469293, + "grad_norm": 2.0627355976801542, + "learning_rate": 5.809408578512831e-07, + "loss": 0.4118, + "step": 38601 + }, + { + "epoch": 0.849702839533348, + "grad_norm": 2.2386973334112907, + "learning_rate": 5.807741036196419e-07, + "loss": 0.2836, + "step": 38602 + }, + { + "epoch": 0.8497248514197667, + "grad_norm": 1.767806124789047, + "learning_rate": 5.806073718484761e-07, + "loss": 0.2728, + "step": 38603 + }, + { + "epoch": 0.8497468633061853, + "grad_norm": 1.6179452660124318, + "learning_rate": 5.804406625386356e-07, + "loss": 0.3088, + "step": 38604 + }, + { + "epoch": 0.849768875192604, + "grad_norm": 1.6444291429849252, + "learning_rate": 5.80273975690967e-07, + "loss": 0.2583, + "step": 38605 + }, + { + "epoch": 0.8497908870790227, + "grad_norm": 0.5873285730456234, + "learning_rate": 5.801073113063166e-07, + "loss": 0.442, + "step": 38606 + }, + { + "epoch": 0.8498128989654413, + "grad_norm": 1.8401241159519401, + "learning_rate": 5.799406693855331e-07, + "loss": 0.2907, + "step": 38607 + }, + { + "epoch": 0.84983491085186, + "grad_norm": 2.7971153633444743, + "learning_rate": 5.797740499294629e-07, + "loss": 0.3381, + "step": 38608 + }, + { + "epoch": 0.8498569227382786, + "grad_norm": 2.033914269098587, + "learning_rate": 5.796074529389501e-07, + "loss": 0.259, + "step": 38609 + }, + { + "epoch": 0.8498789346246973, + "grad_norm": 2.6886159076984795, + "learning_rate": 5.794408784148459e-07, + "loss": 0.2402, + "step": 38610 + }, + { + "epoch": 0.849900946511116, + "grad_norm": 0.6069426964080517, + "learning_rate": 5.792743263579937e-07, + "loss": 0.4188, + "step": 38611 + }, + { + "epoch": 0.8499229583975346, + "grad_norm": 1.722636941050779, + "learning_rate": 5.791077967692421e-07, + "loss": 0.3024, + "step": 38612 + }, + { + "epoch": 0.8499449702839533, + "grad_norm": 1.7430402560200464, + "learning_rate": 5.789412896494362e-07, + "loss": 0.3272, + "step": 38613 + }, + { + "epoch": 0.849966982170372, + "grad_norm": 0.6213632859979729, + "learning_rate": 5.787748049994218e-07, + "loss": 0.4297, + "step": 38614 + }, + { + "epoch": 0.8499889940567906, + "grad_norm": 0.6091489665234736, + "learning_rate": 5.786083428200457e-07, + "loss": 0.4289, + "step": 38615 + }, + { + "epoch": 0.8500110059432093, + "grad_norm": 3.0784252577783535, + "learning_rate": 5.78441903112154e-07, + "loss": 0.273, + "step": 38616 + }, + { + "epoch": 0.850033017829628, + "grad_norm": 1.622036488897942, + "learning_rate": 5.782754858765938e-07, + "loss": 0.2357, + "step": 38617 + }, + { + "epoch": 0.8500550297160466, + "grad_norm": 1.5123725403812924, + "learning_rate": 5.781090911142095e-07, + "loss": 0.3181, + "step": 38618 + }, + { + "epoch": 0.8500770416024653, + "grad_norm": 8.070716384665445, + "learning_rate": 5.779427188258458e-07, + "loss": 0.2377, + "step": 38619 + }, + { + "epoch": 0.850099053488884, + "grad_norm": 0.6127895890039193, + "learning_rate": 5.7777636901235e-07, + "loss": 0.4572, + "step": 38620 + }, + { + "epoch": 0.8501210653753026, + "grad_norm": 0.585319365885064, + "learning_rate": 5.776100416745666e-07, + "loss": 0.4396, + "step": 38621 + }, + { + "epoch": 0.8501430772617213, + "grad_norm": 1.4205778835389777, + "learning_rate": 5.774437368133429e-07, + "loss": 0.2911, + "step": 38622 + }, + { + "epoch": 0.85016508914814, + "grad_norm": 1.6697070956899014, + "learning_rate": 5.772774544295223e-07, + "loss": 0.3029, + "step": 38623 + }, + { + "epoch": 0.8501871010345586, + "grad_norm": 0.6324035072687241, + "learning_rate": 5.771111945239494e-07, + "loss": 0.4593, + "step": 38624 + }, + { + "epoch": 0.8502091129209773, + "grad_norm": 1.3705016622918667, + "learning_rate": 5.7694495709747e-07, + "loss": 0.255, + "step": 38625 + }, + { + "epoch": 0.850231124807396, + "grad_norm": 1.7060544230940515, + "learning_rate": 5.767787421509303e-07, + "loss": 0.2767, + "step": 38626 + }, + { + "epoch": 0.8502531366938146, + "grad_norm": 1.8191545455086795, + "learning_rate": 5.766125496851722e-07, + "loss": 0.3621, + "step": 38627 + }, + { + "epoch": 0.8502751485802333, + "grad_norm": 1.5179058808543722, + "learning_rate": 5.764463797010433e-07, + "loss": 0.3117, + "step": 38628 + }, + { + "epoch": 0.850297160466652, + "grad_norm": 1.8667357184809399, + "learning_rate": 5.762802321993865e-07, + "loss": 0.3275, + "step": 38629 + }, + { + "epoch": 0.8503191723530706, + "grad_norm": 1.5493159477875178, + "learning_rate": 5.761141071810461e-07, + "loss": 0.2704, + "step": 38630 + }, + { + "epoch": 0.8503411842394893, + "grad_norm": 4.844912751975518, + "learning_rate": 5.759480046468674e-07, + "loss": 0.3556, + "step": 38631 + }, + { + "epoch": 0.8503631961259079, + "grad_norm": 0.6393903941102315, + "learning_rate": 5.757819245976937e-07, + "loss": 0.4434, + "step": 38632 + }, + { + "epoch": 0.8503852080123266, + "grad_norm": 1.499636083574045, + "learning_rate": 5.756158670343709e-07, + "loss": 0.2682, + "step": 38633 + }, + { + "epoch": 0.8504072198987453, + "grad_norm": 1.868454717908463, + "learning_rate": 5.754498319577401e-07, + "loss": 0.3073, + "step": 38634 + }, + { + "epoch": 0.8504292317851639, + "grad_norm": 0.5984714302664368, + "learning_rate": 5.752838193686466e-07, + "loss": 0.4524, + "step": 38635 + }, + { + "epoch": 0.8504512436715826, + "grad_norm": 1.6704397879879456, + "learning_rate": 5.751178292679349e-07, + "loss": 0.272, + "step": 38636 + }, + { + "epoch": 0.8504732555580014, + "grad_norm": 2.644729906849523, + "learning_rate": 5.749518616564475e-07, + "loss": 0.2999, + "step": 38637 + }, + { + "epoch": 0.85049526744442, + "grad_norm": 2.1105679526627648, + "learning_rate": 5.747859165350295e-07, + "loss": 0.2889, + "step": 38638 + }, + { + "epoch": 0.8505172793308386, + "grad_norm": 0.6260691166914154, + "learning_rate": 5.746199939045216e-07, + "loss": 0.474, + "step": 38639 + }, + { + "epoch": 0.8505392912172574, + "grad_norm": 1.8058766786616345, + "learning_rate": 5.744540937657689e-07, + "loss": 0.3001, + "step": 38640 + }, + { + "epoch": 0.850561303103676, + "grad_norm": 1.9960475445382597, + "learning_rate": 5.742882161196156e-07, + "loss": 0.3213, + "step": 38641 + }, + { + "epoch": 0.8505833149900947, + "grad_norm": 1.8607118494055228, + "learning_rate": 5.741223609669022e-07, + "loss": 0.2826, + "step": 38642 + }, + { + "epoch": 0.8506053268765134, + "grad_norm": 0.6229102108148669, + "learning_rate": 5.739565283084741e-07, + "loss": 0.4578, + "step": 38643 + }, + { + "epoch": 0.850627338762932, + "grad_norm": 1.5944838916194706, + "learning_rate": 5.73790718145173e-07, + "loss": 0.2979, + "step": 38644 + }, + { + "epoch": 0.8506493506493507, + "grad_norm": 2.233964465518438, + "learning_rate": 5.736249304778397e-07, + "loss": 0.3072, + "step": 38645 + }, + { + "epoch": 0.8506713625357694, + "grad_norm": 2.3471605567628893, + "learning_rate": 5.734591653073202e-07, + "loss": 0.3096, + "step": 38646 + }, + { + "epoch": 0.850693374422188, + "grad_norm": 2.5795754579471493, + "learning_rate": 5.732934226344549e-07, + "loss": 0.3578, + "step": 38647 + }, + { + "epoch": 0.8507153863086067, + "grad_norm": 1.8195133999173696, + "learning_rate": 5.731277024600878e-07, + "loss": 0.3261, + "step": 38648 + }, + { + "epoch": 0.8507373981950254, + "grad_norm": 2.1176731665537565, + "learning_rate": 5.729620047850604e-07, + "loss": 0.2809, + "step": 38649 + }, + { + "epoch": 0.850759410081444, + "grad_norm": 0.6008939290264121, + "learning_rate": 5.727963296102135e-07, + "loss": 0.4537, + "step": 38650 + }, + { + "epoch": 0.8507814219678627, + "grad_norm": 1.9449393930612353, + "learning_rate": 5.726306769363899e-07, + "loss": 0.2768, + "step": 38651 + }, + { + "epoch": 0.8508034338542814, + "grad_norm": 1.5620212503177728, + "learning_rate": 5.724650467644322e-07, + "loss": 0.3359, + "step": 38652 + }, + { + "epoch": 0.8508254457407, + "grad_norm": 1.9451737873850623, + "learning_rate": 5.722994390951825e-07, + "loss": 0.3192, + "step": 38653 + }, + { + "epoch": 0.8508474576271187, + "grad_norm": 1.5437501848627888, + "learning_rate": 5.721338539294824e-07, + "loss": 0.2437, + "step": 38654 + }, + { + "epoch": 0.8508694695135373, + "grad_norm": 1.9614734275311294, + "learning_rate": 5.719682912681712e-07, + "loss": 0.2954, + "step": 38655 + }, + { + "epoch": 0.850891481399956, + "grad_norm": 0.6055489612761192, + "learning_rate": 5.718027511120927e-07, + "loss": 0.4204, + "step": 38656 + }, + { + "epoch": 0.8509134932863747, + "grad_norm": 1.526851480583611, + "learning_rate": 5.716372334620873e-07, + "loss": 0.3268, + "step": 38657 + }, + { + "epoch": 0.8509355051727933, + "grad_norm": 0.6162091117676507, + "learning_rate": 5.714717383189977e-07, + "loss": 0.4583, + "step": 38658 + }, + { + "epoch": 0.850957517059212, + "grad_norm": 1.9519094206992371, + "learning_rate": 5.713062656836643e-07, + "loss": 0.3739, + "step": 38659 + }, + { + "epoch": 0.8509795289456307, + "grad_norm": 1.447179184685525, + "learning_rate": 5.711408155569264e-07, + "loss": 0.2863, + "step": 38660 + }, + { + "epoch": 0.8510015408320493, + "grad_norm": 1.587879450591868, + "learning_rate": 5.709753879396262e-07, + "loss": 0.2385, + "step": 38661 + }, + { + "epoch": 0.851023552718468, + "grad_norm": 2.668958716224554, + "learning_rate": 5.708099828326052e-07, + "loss": 0.3198, + "step": 38662 + }, + { + "epoch": 0.8510455646048867, + "grad_norm": 2.0726151010169214, + "learning_rate": 5.706446002367027e-07, + "loss": 0.2936, + "step": 38663 + }, + { + "epoch": 0.8510675764913053, + "grad_norm": 0.6247034556111744, + "learning_rate": 5.704792401527609e-07, + "loss": 0.4635, + "step": 38664 + }, + { + "epoch": 0.851089588377724, + "grad_norm": 1.900226012123771, + "learning_rate": 5.703139025816179e-07, + "loss": 0.3435, + "step": 38665 + }, + { + "epoch": 0.8511116002641427, + "grad_norm": 1.6941502766893473, + "learning_rate": 5.701485875241158e-07, + "loss": 0.2184, + "step": 38666 + }, + { + "epoch": 0.8511336121505613, + "grad_norm": 1.5079292846588666, + "learning_rate": 5.699832949810952e-07, + "loss": 0.28, + "step": 38667 + }, + { + "epoch": 0.85115562403698, + "grad_norm": 1.4208203559398682, + "learning_rate": 5.698180249533941e-07, + "loss": 0.2672, + "step": 38668 + }, + { + "epoch": 0.8511776359233987, + "grad_norm": 2.119239947321411, + "learning_rate": 5.696527774418548e-07, + "loss": 0.3038, + "step": 38669 + }, + { + "epoch": 0.8511996478098173, + "grad_norm": 0.6112316018594043, + "learning_rate": 5.694875524473153e-07, + "loss": 0.4632, + "step": 38670 + }, + { + "epoch": 0.851221659696236, + "grad_norm": 2.1137835855141285, + "learning_rate": 5.693223499706158e-07, + "loss": 0.3269, + "step": 38671 + }, + { + "epoch": 0.8512436715826547, + "grad_norm": 0.601625823123132, + "learning_rate": 5.691571700125976e-07, + "loss": 0.4256, + "step": 38672 + }, + { + "epoch": 0.8512656834690733, + "grad_norm": 0.6372976076881768, + "learning_rate": 5.689920125740972e-07, + "loss": 0.4786, + "step": 38673 + }, + { + "epoch": 0.851287695355492, + "grad_norm": 1.3056934375340952, + "learning_rate": 5.688268776559575e-07, + "loss": 0.2527, + "step": 38674 + }, + { + "epoch": 0.8513097072419107, + "grad_norm": 2.180764371057861, + "learning_rate": 5.686617652590143e-07, + "loss": 0.2437, + "step": 38675 + }, + { + "epoch": 0.8513317191283293, + "grad_norm": 3.9431304999428956, + "learning_rate": 5.684966753841093e-07, + "loss": 0.3286, + "step": 38676 + }, + { + "epoch": 0.851353731014748, + "grad_norm": 2.2936999332120807, + "learning_rate": 5.683316080320811e-07, + "loss": 0.2814, + "step": 38677 + }, + { + "epoch": 0.8513757429011666, + "grad_norm": 1.7987145280898469, + "learning_rate": 5.68166563203767e-07, + "loss": 0.1823, + "step": 38678 + }, + { + "epoch": 0.8513977547875853, + "grad_norm": 2.8680882261781044, + "learning_rate": 5.680015409000084e-07, + "loss": 0.3451, + "step": 38679 + }, + { + "epoch": 0.851419766674004, + "grad_norm": 1.7288890845060976, + "learning_rate": 5.678365411216419e-07, + "loss": 0.3249, + "step": 38680 + }, + { + "epoch": 0.8514417785604226, + "grad_norm": 2.2523643063529235, + "learning_rate": 5.676715638695063e-07, + "loss": 0.3257, + "step": 38681 + }, + { + "epoch": 0.8514637904468413, + "grad_norm": 2.0825727774020786, + "learning_rate": 5.67506609144442e-07, + "loss": 0.2707, + "step": 38682 + }, + { + "epoch": 0.85148580233326, + "grad_norm": 1.7452462907968012, + "learning_rate": 5.673416769472856e-07, + "loss": 0.2369, + "step": 38683 + }, + { + "epoch": 0.8515078142196786, + "grad_norm": 0.6028158329699611, + "learning_rate": 5.671767672788763e-07, + "loss": 0.4391, + "step": 38684 + }, + { + "epoch": 0.8515298261060973, + "grad_norm": 0.61987759695747, + "learning_rate": 5.670118801400515e-07, + "loss": 0.4324, + "step": 38685 + }, + { + "epoch": 0.851551837992516, + "grad_norm": 1.301239731378841, + "learning_rate": 5.668470155316491e-07, + "loss": 0.2616, + "step": 38686 + }, + { + "epoch": 0.8515738498789346, + "grad_norm": 1.8324268590200348, + "learning_rate": 5.666821734545069e-07, + "loss": 0.3296, + "step": 38687 + }, + { + "epoch": 0.8515958617653533, + "grad_norm": 0.6199741545296834, + "learning_rate": 5.665173539094637e-07, + "loss": 0.4488, + "step": 38688 + }, + { + "epoch": 0.851617873651772, + "grad_norm": 1.5715367548863048, + "learning_rate": 5.663525568973571e-07, + "loss": 0.2601, + "step": 38689 + }, + { + "epoch": 0.8516398855381906, + "grad_norm": 1.6108587616492187, + "learning_rate": 5.661877824190249e-07, + "loss": 0.235, + "step": 38690 + }, + { + "epoch": 0.8516618974246093, + "grad_norm": 1.791518868722288, + "learning_rate": 5.660230304753022e-07, + "loss": 0.3448, + "step": 38691 + }, + { + "epoch": 0.851683909311028, + "grad_norm": 1.7020200957728828, + "learning_rate": 5.658583010670288e-07, + "loss": 0.2922, + "step": 38692 + }, + { + "epoch": 0.8517059211974466, + "grad_norm": 1.3592546018129747, + "learning_rate": 5.656935941950409e-07, + "loss": 0.2672, + "step": 38693 + }, + { + "epoch": 0.8517279330838653, + "grad_norm": 2.1275815955387203, + "learning_rate": 5.655289098601769e-07, + "loss": 0.3025, + "step": 38694 + }, + { + "epoch": 0.851749944970284, + "grad_norm": 1.677827156837936, + "learning_rate": 5.653642480632727e-07, + "loss": 0.3027, + "step": 38695 + }, + { + "epoch": 0.8517719568567026, + "grad_norm": 1.6185461788639908, + "learning_rate": 5.651996088051642e-07, + "loss": 0.3333, + "step": 38696 + }, + { + "epoch": 0.8517939687431213, + "grad_norm": 0.5873389533914284, + "learning_rate": 5.650349920866893e-07, + "loss": 0.4309, + "step": 38697 + }, + { + "epoch": 0.85181598062954, + "grad_norm": 2.500396300448609, + "learning_rate": 5.64870397908685e-07, + "loss": 0.3329, + "step": 38698 + }, + { + "epoch": 0.8518379925159586, + "grad_norm": 2.594281266512907, + "learning_rate": 5.647058262719885e-07, + "loss": 0.3485, + "step": 38699 + }, + { + "epoch": 0.8518600044023773, + "grad_norm": 2.4821901946941134, + "learning_rate": 5.645412771774344e-07, + "loss": 0.3613, + "step": 38700 + }, + { + "epoch": 0.8518820162887959, + "grad_norm": 4.301709481104592, + "learning_rate": 5.643767506258596e-07, + "loss": 0.2495, + "step": 38701 + }, + { + "epoch": 0.8519040281752146, + "grad_norm": 1.598160975650319, + "learning_rate": 5.642122466181005e-07, + "loss": 0.3087, + "step": 38702 + }, + { + "epoch": 0.8519260400616333, + "grad_norm": 0.6181465238485698, + "learning_rate": 5.640477651549935e-07, + "loss": 0.4481, + "step": 38703 + }, + { + "epoch": 0.8519480519480519, + "grad_norm": 0.6357240078090693, + "learning_rate": 5.638833062373739e-07, + "loss": 0.4434, + "step": 38704 + }, + { + "epoch": 0.8519700638344706, + "grad_norm": 0.5898548650220402, + "learning_rate": 5.637188698660784e-07, + "loss": 0.4631, + "step": 38705 + }, + { + "epoch": 0.8519920757208893, + "grad_norm": 5.555709015167375, + "learning_rate": 5.635544560419415e-07, + "loss": 0.3672, + "step": 38706 + }, + { + "epoch": 0.8520140876073079, + "grad_norm": 2.7587882964016766, + "learning_rate": 5.633900647657997e-07, + "loss": 0.4204, + "step": 38707 + }, + { + "epoch": 0.8520360994937266, + "grad_norm": 1.603876758059413, + "learning_rate": 5.632256960384891e-07, + "loss": 0.2748, + "step": 38708 + }, + { + "epoch": 0.8520581113801453, + "grad_norm": 1.3717122997397906, + "learning_rate": 5.630613498608433e-07, + "loss": 0.2589, + "step": 38709 + }, + { + "epoch": 0.8520801232665639, + "grad_norm": 1.7843545140227555, + "learning_rate": 5.628970262337003e-07, + "loss": 0.2965, + "step": 38710 + }, + { + "epoch": 0.8521021351529826, + "grad_norm": 1.4344775934796632, + "learning_rate": 5.627327251578918e-07, + "loss": 0.271, + "step": 38711 + }, + { + "epoch": 0.8521241470394013, + "grad_norm": 2.1955618308242664, + "learning_rate": 5.625684466342551e-07, + "loss": 0.3166, + "step": 38712 + }, + { + "epoch": 0.8521461589258199, + "grad_norm": 1.7587245807772958, + "learning_rate": 5.624041906636257e-07, + "loss": 0.2793, + "step": 38713 + }, + { + "epoch": 0.8521681708122386, + "grad_norm": 1.52376677316738, + "learning_rate": 5.622399572468362e-07, + "loss": 0.2637, + "step": 38714 + }, + { + "epoch": 0.8521901826986573, + "grad_norm": 1.28371467867418, + "learning_rate": 5.620757463847237e-07, + "loss": 0.2529, + "step": 38715 + }, + { + "epoch": 0.8522121945850759, + "grad_norm": 1.3915994195988992, + "learning_rate": 5.61911558078121e-07, + "loss": 0.3756, + "step": 38716 + }, + { + "epoch": 0.8522342064714946, + "grad_norm": 1.7105848039811435, + "learning_rate": 5.617473923278627e-07, + "loss": 0.2635, + "step": 38717 + }, + { + "epoch": 0.8522562183579133, + "grad_norm": 0.5973697386639626, + "learning_rate": 5.615832491347845e-07, + "loss": 0.4481, + "step": 38718 + }, + { + "epoch": 0.8522782302443319, + "grad_norm": 1.358641501266985, + "learning_rate": 5.614191284997194e-07, + "loss": 0.2875, + "step": 38719 + }, + { + "epoch": 0.8523002421307506, + "grad_norm": 1.6680933393934014, + "learning_rate": 5.612550304235026e-07, + "loss": 0.3142, + "step": 38720 + }, + { + "epoch": 0.8523222540171693, + "grad_norm": 1.675259083025707, + "learning_rate": 5.610909549069665e-07, + "loss": 0.2931, + "step": 38721 + }, + { + "epoch": 0.8523442659035879, + "grad_norm": 6.347795361116757, + "learning_rate": 5.609269019509472e-07, + "loss": 0.2168, + "step": 38722 + }, + { + "epoch": 0.8523662777900066, + "grad_norm": 1.6723264462604066, + "learning_rate": 5.607628715562757e-07, + "loss": 0.2344, + "step": 38723 + }, + { + "epoch": 0.8523882896764252, + "grad_norm": 1.8578052208141747, + "learning_rate": 5.605988637237875e-07, + "loss": 0.3343, + "step": 38724 + }, + { + "epoch": 0.8524103015628439, + "grad_norm": 1.4010495201637578, + "learning_rate": 5.604348784543173e-07, + "loss": 0.2328, + "step": 38725 + }, + { + "epoch": 0.8524323134492626, + "grad_norm": 1.6447591421790386, + "learning_rate": 5.602709157486969e-07, + "loss": 0.3097, + "step": 38726 + }, + { + "epoch": 0.8524543253356812, + "grad_norm": 1.9123189010116528, + "learning_rate": 5.601069756077588e-07, + "loss": 0.3092, + "step": 38727 + }, + { + "epoch": 0.8524763372220999, + "grad_norm": 2.009763086830486, + "learning_rate": 5.599430580323373e-07, + "loss": 0.3261, + "step": 38728 + }, + { + "epoch": 0.8524983491085186, + "grad_norm": 1.7459082074303276, + "learning_rate": 5.597791630232657e-07, + "loss": 0.3897, + "step": 38729 + }, + { + "epoch": 0.8525203609949372, + "grad_norm": 0.6097954833376453, + "learning_rate": 5.596152905813774e-07, + "loss": 0.4516, + "step": 38730 + }, + { + "epoch": 0.8525423728813559, + "grad_norm": 3.662555118982407, + "learning_rate": 5.594514407075053e-07, + "loss": 0.2532, + "step": 38731 + }, + { + "epoch": 0.8525643847677746, + "grad_norm": 1.9032023096601616, + "learning_rate": 5.592876134024799e-07, + "loss": 0.2386, + "step": 38732 + }, + { + "epoch": 0.8525863966541932, + "grad_norm": 2.0313494035495796, + "learning_rate": 5.591238086671358e-07, + "loss": 0.2395, + "step": 38733 + }, + { + "epoch": 0.8526084085406119, + "grad_norm": 1.665198494587379, + "learning_rate": 5.589600265023048e-07, + "loss": 0.2546, + "step": 38734 + }, + { + "epoch": 0.8526304204270306, + "grad_norm": 2.1230264676327044, + "learning_rate": 5.58796266908821e-07, + "loss": 0.2984, + "step": 38735 + }, + { + "epoch": 0.8526524323134492, + "grad_norm": 1.443169553155496, + "learning_rate": 5.586325298875151e-07, + "loss": 0.2414, + "step": 38736 + }, + { + "epoch": 0.8526744441998679, + "grad_norm": 2.109592347948174, + "learning_rate": 5.584688154392193e-07, + "loss": 0.2996, + "step": 38737 + }, + { + "epoch": 0.8526964560862866, + "grad_norm": 0.6302002214981941, + "learning_rate": 5.583051235647652e-07, + "loss": 0.4572, + "step": 38738 + }, + { + "epoch": 0.8527184679727052, + "grad_norm": 2.335286096828045, + "learning_rate": 5.58141454264986e-07, + "loss": 0.2314, + "step": 38739 + }, + { + "epoch": 0.8527404798591239, + "grad_norm": 1.8448077159967562, + "learning_rate": 5.579778075407138e-07, + "loss": 0.2525, + "step": 38740 + }, + { + "epoch": 0.8527624917455426, + "grad_norm": 3.232959490543221, + "learning_rate": 5.578141833927797e-07, + "loss": 0.2399, + "step": 38741 + }, + { + "epoch": 0.8527845036319612, + "grad_norm": 2.5723996769970494, + "learning_rate": 5.57650581822014e-07, + "loss": 0.2475, + "step": 38742 + }, + { + "epoch": 0.8528065155183799, + "grad_norm": 3.8456051546207872, + "learning_rate": 5.574870028292495e-07, + "loss": 0.3097, + "step": 38743 + }, + { + "epoch": 0.8528285274047986, + "grad_norm": 2.5717892044777115, + "learning_rate": 5.573234464153188e-07, + "loss": 0.2687, + "step": 38744 + }, + { + "epoch": 0.8528505392912172, + "grad_norm": 0.6319974286748179, + "learning_rate": 5.571599125810506e-07, + "loss": 0.4529, + "step": 38745 + }, + { + "epoch": 0.8528725511776359, + "grad_norm": 1.4351280977268919, + "learning_rate": 5.569964013272777e-07, + "loss": 0.2525, + "step": 38746 + }, + { + "epoch": 0.8528945630640546, + "grad_norm": 0.628386222256444, + "learning_rate": 5.568329126548305e-07, + "loss": 0.4814, + "step": 38747 + }, + { + "epoch": 0.8529165749504732, + "grad_norm": 1.5907452857695925, + "learning_rate": 5.566694465645401e-07, + "loss": 0.3255, + "step": 38748 + }, + { + "epoch": 0.8529385868368919, + "grad_norm": 1.7721173999236222, + "learning_rate": 5.565060030572378e-07, + "loss": 0.3334, + "step": 38749 + }, + { + "epoch": 0.8529605987233105, + "grad_norm": 2.9037575328136764, + "learning_rate": 5.563425821337531e-07, + "loss": 0.2591, + "step": 38750 + }, + { + "epoch": 0.8529826106097292, + "grad_norm": 1.4188208137239615, + "learning_rate": 5.561791837949188e-07, + "loss": 0.2495, + "step": 38751 + }, + { + "epoch": 0.853004622496148, + "grad_norm": 1.9919062277365431, + "learning_rate": 5.560158080415623e-07, + "loss": 0.3061, + "step": 38752 + }, + { + "epoch": 0.8530266343825665, + "grad_norm": 1.2034787329817422, + "learning_rate": 5.558524548745159e-07, + "loss": 0.2656, + "step": 38753 + }, + { + "epoch": 0.8530486462689852, + "grad_norm": 2.7932319971504698, + "learning_rate": 5.5568912429461e-07, + "loss": 0.2859, + "step": 38754 + }, + { + "epoch": 0.853070658155404, + "grad_norm": 2.7569058268348443, + "learning_rate": 5.555258163026738e-07, + "loss": 0.2325, + "step": 38755 + }, + { + "epoch": 0.8530926700418225, + "grad_norm": 1.9451653039766021, + "learning_rate": 5.55362530899538e-07, + "loss": 0.4253, + "step": 38756 + }, + { + "epoch": 0.8531146819282412, + "grad_norm": 1.6089466023130283, + "learning_rate": 5.551992680860319e-07, + "loss": 0.2327, + "step": 38757 + }, + { + "epoch": 0.85313669381466, + "grad_norm": 1.1476999332224902, + "learning_rate": 5.550360278629863e-07, + "loss": 0.1893, + "step": 38758 + }, + { + "epoch": 0.8531587057010785, + "grad_norm": 34.14079534823747, + "learning_rate": 5.54872810231229e-07, + "loss": 0.3547, + "step": 38759 + }, + { + "epoch": 0.8531807175874973, + "grad_norm": 1.960137484346552, + "learning_rate": 5.547096151915909e-07, + "loss": 0.2659, + "step": 38760 + }, + { + "epoch": 0.853202729473916, + "grad_norm": 1.519821878915583, + "learning_rate": 5.545464427449016e-07, + "loss": 0.2589, + "step": 38761 + }, + { + "epoch": 0.8532247413603345, + "grad_norm": 1.606772659478541, + "learning_rate": 5.543832928919907e-07, + "loss": 0.3124, + "step": 38762 + }, + { + "epoch": 0.8532467532467533, + "grad_norm": 1.984312356466291, + "learning_rate": 5.542201656336854e-07, + "loss": 0.3844, + "step": 38763 + }, + { + "epoch": 0.853268765133172, + "grad_norm": 1.4029587495633393, + "learning_rate": 5.540570609708162e-07, + "loss": 0.2148, + "step": 38764 + }, + { + "epoch": 0.8532907770195906, + "grad_norm": 1.411090538220584, + "learning_rate": 5.538939789042113e-07, + "loss": 0.2409, + "step": 38765 + }, + { + "epoch": 0.8533127889060093, + "grad_norm": 0.6433869388730191, + "learning_rate": 5.537309194347024e-07, + "loss": 0.4671, + "step": 38766 + }, + { + "epoch": 0.853334800792428, + "grad_norm": 1.343739099718622, + "learning_rate": 5.535678825631152e-07, + "loss": 0.2989, + "step": 38767 + }, + { + "epoch": 0.8533568126788466, + "grad_norm": 1.7571443036552978, + "learning_rate": 5.534048682902781e-07, + "loss": 0.2059, + "step": 38768 + }, + { + "epoch": 0.8533788245652653, + "grad_norm": 1.739704201158915, + "learning_rate": 5.532418766170211e-07, + "loss": 0.3371, + "step": 38769 + }, + { + "epoch": 0.853400836451684, + "grad_norm": 1.815643066124495, + "learning_rate": 5.530789075441723e-07, + "loss": 0.3064, + "step": 38770 + }, + { + "epoch": 0.8534228483381026, + "grad_norm": 3.804821401495162, + "learning_rate": 5.52915961072561e-07, + "loss": 0.3279, + "step": 38771 + }, + { + "epoch": 0.8534448602245213, + "grad_norm": 1.8615931614471697, + "learning_rate": 5.527530372030138e-07, + "loss": 0.2773, + "step": 38772 + }, + { + "epoch": 0.8534668721109399, + "grad_norm": 2.4165469361588237, + "learning_rate": 5.525901359363583e-07, + "loss": 0.2657, + "step": 38773 + }, + { + "epoch": 0.8534888839973586, + "grad_norm": 1.2830492214557057, + "learning_rate": 5.52427257273423e-07, + "loss": 0.2389, + "step": 38774 + }, + { + "epoch": 0.8535108958837773, + "grad_norm": 2.8264805467527827, + "learning_rate": 5.522644012150363e-07, + "loss": 0.316, + "step": 38775 + }, + { + "epoch": 0.8535329077701959, + "grad_norm": 1.4552320310745581, + "learning_rate": 5.521015677620273e-07, + "loss": 0.2641, + "step": 38776 + }, + { + "epoch": 0.8535549196566146, + "grad_norm": 1.8315017447429536, + "learning_rate": 5.519387569152213e-07, + "loss": 0.2778, + "step": 38777 + }, + { + "epoch": 0.8535769315430333, + "grad_norm": 1.8983545560798658, + "learning_rate": 5.517759686754454e-07, + "loss": 0.3381, + "step": 38778 + }, + { + "epoch": 0.8535989434294519, + "grad_norm": 1.4628028032421403, + "learning_rate": 5.516132030435278e-07, + "loss": 0.3063, + "step": 38779 + }, + { + "epoch": 0.8536209553158706, + "grad_norm": 1.7825876314220244, + "learning_rate": 5.514504600202963e-07, + "loss": 0.2579, + "step": 38780 + }, + { + "epoch": 0.8536429672022893, + "grad_norm": 1.6335436655048612, + "learning_rate": 5.512877396065792e-07, + "loss": 0.2918, + "step": 38781 + }, + { + "epoch": 0.8536649790887079, + "grad_norm": 1.6666178137136145, + "learning_rate": 5.51125041803201e-07, + "loss": 0.2484, + "step": 38782 + }, + { + "epoch": 0.8536869909751266, + "grad_norm": 1.7479023092207606, + "learning_rate": 5.509623666109893e-07, + "loss": 0.35, + "step": 38783 + }, + { + "epoch": 0.8537090028615453, + "grad_norm": 1.8311613517521617, + "learning_rate": 5.507997140307708e-07, + "loss": 0.2675, + "step": 38784 + }, + { + "epoch": 0.8537310147479639, + "grad_norm": 1.9664822639316042, + "learning_rate": 5.506370840633734e-07, + "loss": 0.3075, + "step": 38785 + }, + { + "epoch": 0.8537530266343826, + "grad_norm": 1.3661840643510421, + "learning_rate": 5.504744767096221e-07, + "loss": 0.2861, + "step": 38786 + }, + { + "epoch": 0.8537750385208013, + "grad_norm": 1.6373608577592662, + "learning_rate": 5.503118919703454e-07, + "loss": 0.3025, + "step": 38787 + }, + { + "epoch": 0.8537970504072199, + "grad_norm": 0.6236270149631672, + "learning_rate": 5.501493298463667e-07, + "loss": 0.4381, + "step": 38788 + }, + { + "epoch": 0.8538190622936386, + "grad_norm": 1.365034610348863, + "learning_rate": 5.499867903385136e-07, + "loss": 0.2928, + "step": 38789 + }, + { + "epoch": 0.8538410741800573, + "grad_norm": 2.1928672208038753, + "learning_rate": 5.498242734476139e-07, + "loss": 0.2286, + "step": 38790 + }, + { + "epoch": 0.8538630860664759, + "grad_norm": 2.0406543374267203, + "learning_rate": 5.496617791744913e-07, + "loss": 0.2692, + "step": 38791 + }, + { + "epoch": 0.8538850979528946, + "grad_norm": 2.9708145583178904, + "learning_rate": 5.494993075199728e-07, + "loss": 0.2929, + "step": 38792 + }, + { + "epoch": 0.8539071098393133, + "grad_norm": 1.482694539908449, + "learning_rate": 5.493368584848835e-07, + "loss": 0.2554, + "step": 38793 + }, + { + "epoch": 0.8539291217257319, + "grad_norm": 1.6278679644705345, + "learning_rate": 5.491744320700498e-07, + "loss": 0.2855, + "step": 38794 + }, + { + "epoch": 0.8539511336121506, + "grad_norm": 1.5601898980498217, + "learning_rate": 5.490120282762957e-07, + "loss": 0.2654, + "step": 38795 + }, + { + "epoch": 0.8539731454985692, + "grad_norm": 1.977083040799513, + "learning_rate": 5.48849647104448e-07, + "loss": 0.2795, + "step": 38796 + }, + { + "epoch": 0.8539951573849879, + "grad_norm": 0.6011538882163474, + "learning_rate": 5.486872885553324e-07, + "loss": 0.4318, + "step": 38797 + }, + { + "epoch": 0.8540171692714066, + "grad_norm": 1.4611763753316422, + "learning_rate": 5.485249526297726e-07, + "loss": 0.3217, + "step": 38798 + }, + { + "epoch": 0.8540391811578252, + "grad_norm": 1.5808198355610048, + "learning_rate": 5.483626393285951e-07, + "loss": 0.2895, + "step": 38799 + }, + { + "epoch": 0.8540611930442439, + "grad_norm": 2.4576594540289656, + "learning_rate": 5.482003486526233e-07, + "loss": 0.2935, + "step": 38800 + }, + { + "epoch": 0.8540832049306626, + "grad_norm": 3.916184345019063, + "learning_rate": 5.48038080602683e-07, + "loss": 0.2992, + "step": 38801 + }, + { + "epoch": 0.8541052168170812, + "grad_norm": 0.6055674381378879, + "learning_rate": 5.478758351795999e-07, + "loss": 0.425, + "step": 38802 + }, + { + "epoch": 0.8541272287034999, + "grad_norm": 2.604631209667639, + "learning_rate": 5.47713612384197e-07, + "loss": 0.2771, + "step": 38803 + }, + { + "epoch": 0.8541492405899186, + "grad_norm": 1.960798050607041, + "learning_rate": 5.475514122172987e-07, + "loss": 0.2576, + "step": 38804 + }, + { + "epoch": 0.8541712524763372, + "grad_norm": 1.4765315342640306, + "learning_rate": 5.473892346797299e-07, + "loss": 0.2688, + "step": 38805 + }, + { + "epoch": 0.8541932643627559, + "grad_norm": 1.2750540533558707, + "learning_rate": 5.47227079772315e-07, + "loss": 0.2312, + "step": 38806 + }, + { + "epoch": 0.8542152762491746, + "grad_norm": 3.7037954363624253, + "learning_rate": 5.47064947495879e-07, + "loss": 0.3335, + "step": 38807 + }, + { + "epoch": 0.8542372881355932, + "grad_norm": 6.083671748921455, + "learning_rate": 5.469028378512447e-07, + "loss": 0.2834, + "step": 38808 + }, + { + "epoch": 0.8542593000220119, + "grad_norm": 0.6246611536404078, + "learning_rate": 5.467407508392358e-07, + "loss": 0.4495, + "step": 38809 + }, + { + "epoch": 0.8542813119084306, + "grad_norm": 1.5365579827273037, + "learning_rate": 5.465786864606765e-07, + "loss": 0.1918, + "step": 38810 + }, + { + "epoch": 0.8543033237948492, + "grad_norm": 10.207288652889309, + "learning_rate": 5.464166447163904e-07, + "loss": 0.3517, + "step": 38811 + }, + { + "epoch": 0.8543253356812679, + "grad_norm": 0.6121362766447876, + "learning_rate": 5.462546256072027e-07, + "loss": 0.4446, + "step": 38812 + }, + { + "epoch": 0.8543473475676866, + "grad_norm": 1.808519017986051, + "learning_rate": 5.460926291339346e-07, + "loss": 0.2427, + "step": 38813 + }, + { + "epoch": 0.8543693594541052, + "grad_norm": 1.5782903294770212, + "learning_rate": 5.459306552974103e-07, + "loss": 0.2656, + "step": 38814 + }, + { + "epoch": 0.8543913713405239, + "grad_norm": 1.9772668900572072, + "learning_rate": 5.457687040984521e-07, + "loss": 0.2716, + "step": 38815 + }, + { + "epoch": 0.8544133832269426, + "grad_norm": 2.2136238826193297, + "learning_rate": 5.456067755378847e-07, + "loss": 0.3154, + "step": 38816 + }, + { + "epoch": 0.8544353951133612, + "grad_norm": 2.5831018813549402, + "learning_rate": 5.454448696165304e-07, + "loss": 0.3054, + "step": 38817 + }, + { + "epoch": 0.8544574069997799, + "grad_norm": 1.9154723114096037, + "learning_rate": 5.452829863352127e-07, + "loss": 0.2483, + "step": 38818 + }, + { + "epoch": 0.8544794188861985, + "grad_norm": 2.1039576902165513, + "learning_rate": 5.451211256947531e-07, + "loss": 0.2825, + "step": 38819 + }, + { + "epoch": 0.8545014307726172, + "grad_norm": 1.5204818869920047, + "learning_rate": 5.449592876959741e-07, + "loss": 0.3405, + "step": 38820 + }, + { + "epoch": 0.8545234426590359, + "grad_norm": 1.307814551475883, + "learning_rate": 5.447974723396998e-07, + "loss": 0.2773, + "step": 38821 + }, + { + "epoch": 0.8545454545454545, + "grad_norm": 0.5918020732735938, + "learning_rate": 5.446356796267522e-07, + "loss": 0.434, + "step": 38822 + }, + { + "epoch": 0.8545674664318732, + "grad_norm": 1.513771160663177, + "learning_rate": 5.444739095579533e-07, + "loss": 0.3155, + "step": 38823 + }, + { + "epoch": 0.8545894783182919, + "grad_norm": 1.6299649955958226, + "learning_rate": 5.443121621341246e-07, + "loss": 0.2671, + "step": 38824 + }, + { + "epoch": 0.8546114902047105, + "grad_norm": 1.7772886035136557, + "learning_rate": 5.441504373560891e-07, + "loss": 0.3047, + "step": 38825 + }, + { + "epoch": 0.8546335020911292, + "grad_norm": 3.1506369034772455, + "learning_rate": 5.439887352246681e-07, + "loss": 0.3532, + "step": 38826 + }, + { + "epoch": 0.8546555139775479, + "grad_norm": 3.1481998577978634, + "learning_rate": 5.438270557406833e-07, + "loss": 0.3378, + "step": 38827 + }, + { + "epoch": 0.8546775258639665, + "grad_norm": 1.7789837177758232, + "learning_rate": 5.436653989049584e-07, + "loss": 0.3975, + "step": 38828 + }, + { + "epoch": 0.8546995377503852, + "grad_norm": 1.517780074454438, + "learning_rate": 5.435037647183123e-07, + "loss": 0.3116, + "step": 38829 + }, + { + "epoch": 0.8547215496368039, + "grad_norm": 2.165852313325381, + "learning_rate": 5.433421531815686e-07, + "loss": 0.3087, + "step": 38830 + }, + { + "epoch": 0.8547435615232225, + "grad_norm": 2.0364175072558135, + "learning_rate": 5.431805642955468e-07, + "loss": 0.3181, + "step": 38831 + }, + { + "epoch": 0.8547655734096412, + "grad_norm": 0.6243190279970718, + "learning_rate": 5.430189980610695e-07, + "loss": 0.4471, + "step": 38832 + }, + { + "epoch": 0.8547875852960599, + "grad_norm": 1.4999959180189235, + "learning_rate": 5.428574544789583e-07, + "loss": 0.2949, + "step": 38833 + }, + { + "epoch": 0.8548095971824785, + "grad_norm": 1.9715569565341993, + "learning_rate": 5.42695933550032e-07, + "loss": 0.2667, + "step": 38834 + }, + { + "epoch": 0.8548316090688972, + "grad_norm": 0.6279617664889398, + "learning_rate": 5.425344352751144e-07, + "loss": 0.4582, + "step": 38835 + }, + { + "epoch": 0.8548536209553159, + "grad_norm": 1.9675169327867186, + "learning_rate": 5.423729596550243e-07, + "loss": 0.2745, + "step": 38836 + }, + { + "epoch": 0.8548756328417345, + "grad_norm": 1.557858527191469, + "learning_rate": 5.422115066905826e-07, + "loss": 0.3788, + "step": 38837 + }, + { + "epoch": 0.8548976447281532, + "grad_norm": 1.3649284787805966, + "learning_rate": 5.420500763826108e-07, + "loss": 0.3109, + "step": 38838 + }, + { + "epoch": 0.8549196566145719, + "grad_norm": 1.437504538406982, + "learning_rate": 5.418886687319286e-07, + "loss": 0.2906, + "step": 38839 + }, + { + "epoch": 0.8549416685009905, + "grad_norm": 1.3914830335583763, + "learning_rate": 5.417272837393573e-07, + "loss": 0.2578, + "step": 38840 + }, + { + "epoch": 0.8549636803874092, + "grad_norm": 0.5980404224167015, + "learning_rate": 5.415659214057151e-07, + "loss": 0.4299, + "step": 38841 + }, + { + "epoch": 0.8549856922738278, + "grad_norm": 2.2682806031494285, + "learning_rate": 5.414045817318237e-07, + "loss": 0.2914, + "step": 38842 + }, + { + "epoch": 0.8550077041602465, + "grad_norm": 2.152215731735536, + "learning_rate": 5.412432647185035e-07, + "loss": 0.389, + "step": 38843 + }, + { + "epoch": 0.8550297160466652, + "grad_norm": 1.760652089424314, + "learning_rate": 5.410819703665738e-07, + "loss": 0.3647, + "step": 38844 + }, + { + "epoch": 0.8550517279330838, + "grad_norm": 0.6288048423856535, + "learning_rate": 5.409206986768533e-07, + "loss": 0.4446, + "step": 38845 + }, + { + "epoch": 0.8550737398195025, + "grad_norm": 2.437939868927194, + "learning_rate": 5.407594496501628e-07, + "loss": 0.2348, + "step": 38846 + }, + { + "epoch": 0.8550957517059212, + "grad_norm": 1.7538143623822906, + "learning_rate": 5.405982232873214e-07, + "loss": 0.3147, + "step": 38847 + }, + { + "epoch": 0.8551177635923398, + "grad_norm": 1.9311094212400484, + "learning_rate": 5.4043701958915e-07, + "loss": 0.3248, + "step": 38848 + }, + { + "epoch": 0.8551397754787585, + "grad_norm": 1.891163963464129, + "learning_rate": 5.402758385564666e-07, + "loss": 0.3212, + "step": 38849 + }, + { + "epoch": 0.8551617873651772, + "grad_norm": 4.995829276376971, + "learning_rate": 5.401146801900897e-07, + "loss": 0.3225, + "step": 38850 + }, + { + "epoch": 0.8551837992515958, + "grad_norm": 5.584665889797567, + "learning_rate": 5.399535444908388e-07, + "loss": 0.3767, + "step": 38851 + }, + { + "epoch": 0.8552058111380145, + "grad_norm": 2.5261971172500752, + "learning_rate": 5.397924314595332e-07, + "loss": 0.2545, + "step": 38852 + }, + { + "epoch": 0.8552278230244332, + "grad_norm": 2.9220763943530628, + "learning_rate": 5.396313410969928e-07, + "loss": 0.3071, + "step": 38853 + }, + { + "epoch": 0.8552498349108518, + "grad_norm": 2.5203062108246814, + "learning_rate": 5.394702734040358e-07, + "loss": 0.2904, + "step": 38854 + }, + { + "epoch": 0.8552718467972705, + "grad_norm": 1.7398590610835227, + "learning_rate": 5.393092283814788e-07, + "loss": 0.2385, + "step": 38855 + }, + { + "epoch": 0.8552938586836892, + "grad_norm": 1.605231485344802, + "learning_rate": 5.391482060301417e-07, + "loss": 0.2785, + "step": 38856 + }, + { + "epoch": 0.8553158705701078, + "grad_norm": 1.602608856390461, + "learning_rate": 5.389872063508434e-07, + "loss": 0.2856, + "step": 38857 + }, + { + "epoch": 0.8553378824565265, + "grad_norm": 0.6317320491366988, + "learning_rate": 5.38826229344403e-07, + "loss": 0.4328, + "step": 38858 + }, + { + "epoch": 0.8553598943429452, + "grad_norm": 2.0399618778540427, + "learning_rate": 5.386652750116367e-07, + "loss": 0.2509, + "step": 38859 + }, + { + "epoch": 0.8553819062293638, + "grad_norm": 1.6244819851852323, + "learning_rate": 5.385043433533627e-07, + "loss": 0.3048, + "step": 38860 + }, + { + "epoch": 0.8554039181157825, + "grad_norm": 2.0870845527074833, + "learning_rate": 5.383434343704008e-07, + "loss": 0.2554, + "step": 38861 + }, + { + "epoch": 0.8554259300022012, + "grad_norm": 1.8929256795464127, + "learning_rate": 5.381825480635661e-07, + "loss": 0.314, + "step": 38862 + }, + { + "epoch": 0.8554479418886198, + "grad_norm": 2.6299083977569375, + "learning_rate": 5.380216844336778e-07, + "loss": 0.2849, + "step": 38863 + }, + { + "epoch": 0.8554699537750385, + "grad_norm": 1.3996347961361872, + "learning_rate": 5.378608434815547e-07, + "loss": 0.3124, + "step": 38864 + }, + { + "epoch": 0.8554919656614571, + "grad_norm": 0.6261922981258599, + "learning_rate": 5.377000252080116e-07, + "loss": 0.425, + "step": 38865 + }, + { + "epoch": 0.8555139775478758, + "grad_norm": 1.6430786484272881, + "learning_rate": 5.375392296138682e-07, + "loss": 0.4185, + "step": 38866 + }, + { + "epoch": 0.8555359894342945, + "grad_norm": 0.614538489242045, + "learning_rate": 5.373784566999402e-07, + "loss": 0.4506, + "step": 38867 + }, + { + "epoch": 0.8555580013207131, + "grad_norm": 0.6165526567680591, + "learning_rate": 5.372177064670447e-07, + "loss": 0.4351, + "step": 38868 + }, + { + "epoch": 0.8555800132071318, + "grad_norm": 1.4653377515712938, + "learning_rate": 5.370569789160002e-07, + "loss": 0.2522, + "step": 38869 + }, + { + "epoch": 0.8556020250935505, + "grad_norm": 4.157119887292788, + "learning_rate": 5.368962740476219e-07, + "loss": 0.3432, + "step": 38870 + }, + { + "epoch": 0.8556240369799691, + "grad_norm": 2.8348774051249848, + "learning_rate": 5.367355918627287e-07, + "loss": 0.2234, + "step": 38871 + }, + { + "epoch": 0.8556460488663878, + "grad_norm": 2.394235245847422, + "learning_rate": 5.365749323621344e-07, + "loss": 0.2822, + "step": 38872 + }, + { + "epoch": 0.8556680607528065, + "grad_norm": 1.7339750585708806, + "learning_rate": 5.364142955466572e-07, + "loss": 0.2654, + "step": 38873 + }, + { + "epoch": 0.8556900726392251, + "grad_norm": 1.861932149272113, + "learning_rate": 5.362536814171143e-07, + "loss": 0.3118, + "step": 38874 + }, + { + "epoch": 0.8557120845256438, + "grad_norm": 1.5178072930536661, + "learning_rate": 5.360930899743206e-07, + "loss": 0.2512, + "step": 38875 + }, + { + "epoch": 0.8557340964120626, + "grad_norm": 1.5289618023313756, + "learning_rate": 5.35932521219093e-07, + "loss": 0.3255, + "step": 38876 + }, + { + "epoch": 0.8557561082984811, + "grad_norm": 0.6163550464965671, + "learning_rate": 5.357719751522467e-07, + "loss": 0.4582, + "step": 38877 + }, + { + "epoch": 0.8557781201848998, + "grad_norm": 1.5766974515147771, + "learning_rate": 5.356114517745986e-07, + "loss": 0.2332, + "step": 38878 + }, + { + "epoch": 0.8558001320713186, + "grad_norm": 1.2630653670901495, + "learning_rate": 5.354509510869649e-07, + "loss": 0.29, + "step": 38879 + }, + { + "epoch": 0.8558221439577371, + "grad_norm": 1.5574852552606182, + "learning_rate": 5.352904730901598e-07, + "loss": 0.3327, + "step": 38880 + }, + { + "epoch": 0.8558441558441559, + "grad_norm": 2.122581318476851, + "learning_rate": 5.35130017785001e-07, + "loss": 0.2901, + "step": 38881 + }, + { + "epoch": 0.8558661677305746, + "grad_norm": 1.5748857702867403, + "learning_rate": 5.349695851723014e-07, + "loss": 0.2615, + "step": 38882 + }, + { + "epoch": 0.8558881796169931, + "grad_norm": 0.650349414737973, + "learning_rate": 5.348091752528784e-07, + "loss": 0.4553, + "step": 38883 + }, + { + "epoch": 0.8559101915034119, + "grad_norm": 1.570547117257086, + "learning_rate": 5.346487880275475e-07, + "loss": 0.2704, + "step": 38884 + }, + { + "epoch": 0.8559322033898306, + "grad_norm": 1.6772199552095846, + "learning_rate": 5.344884234971226e-07, + "loss": 0.3274, + "step": 38885 + }, + { + "epoch": 0.8559542152762492, + "grad_norm": 1.8290399214008966, + "learning_rate": 5.343280816624185e-07, + "loss": 0.2928, + "step": 38886 + }, + { + "epoch": 0.8559762271626679, + "grad_norm": 1.7466718523002038, + "learning_rate": 5.341677625242514e-07, + "loss": 0.2236, + "step": 38887 + }, + { + "epoch": 0.8559982390490865, + "grad_norm": 2.8234400254540555, + "learning_rate": 5.340074660834349e-07, + "loss": 0.2754, + "step": 38888 + }, + { + "epoch": 0.8560202509355052, + "grad_norm": 2.344183838232165, + "learning_rate": 5.338471923407857e-07, + "loss": 0.3619, + "step": 38889 + }, + { + "epoch": 0.8560422628219239, + "grad_norm": 0.6102988442543402, + "learning_rate": 5.336869412971163e-07, + "loss": 0.4569, + "step": 38890 + }, + { + "epoch": 0.8560642747083425, + "grad_norm": 1.369225801742715, + "learning_rate": 5.335267129532417e-07, + "loss": 0.3359, + "step": 38891 + }, + { + "epoch": 0.8560862865947612, + "grad_norm": 1.9090583363787326, + "learning_rate": 5.333665073099764e-07, + "loss": 0.3136, + "step": 38892 + }, + { + "epoch": 0.8561082984811799, + "grad_norm": 1.9843296032889384, + "learning_rate": 5.332063243681341e-07, + "loss": 0.2772, + "step": 38893 + }, + { + "epoch": 0.8561303103675985, + "grad_norm": 1.8023174564909776, + "learning_rate": 5.330461641285312e-07, + "loss": 0.3151, + "step": 38894 + }, + { + "epoch": 0.8561523222540172, + "grad_norm": 1.6330668350317668, + "learning_rate": 5.328860265919794e-07, + "loss": 0.2754, + "step": 38895 + }, + { + "epoch": 0.8561743341404359, + "grad_norm": 1.4701062149845399, + "learning_rate": 5.327259117592926e-07, + "loss": 0.2935, + "step": 38896 + }, + { + "epoch": 0.8561963460268545, + "grad_norm": 3.855374286942287, + "learning_rate": 5.32565819631286e-07, + "loss": 0.2456, + "step": 38897 + }, + { + "epoch": 0.8562183579132732, + "grad_norm": 1.8183726667743976, + "learning_rate": 5.324057502087704e-07, + "loss": 0.2511, + "step": 38898 + }, + { + "epoch": 0.8562403697996919, + "grad_norm": 1.4764261927649727, + "learning_rate": 5.322457034925638e-07, + "loss": 0.3292, + "step": 38899 + }, + { + "epoch": 0.8562623816861105, + "grad_norm": 1.7211712299080517, + "learning_rate": 5.320856794834767e-07, + "loss": 0.3188, + "step": 38900 + }, + { + "epoch": 0.8562843935725292, + "grad_norm": 1.531096766191776, + "learning_rate": 5.319256781823217e-07, + "loss": 0.2536, + "step": 38901 + }, + { + "epoch": 0.8563064054589479, + "grad_norm": 2.830920742596016, + "learning_rate": 5.317656995899151e-07, + "loss": 0.2931, + "step": 38902 + }, + { + "epoch": 0.8563284173453665, + "grad_norm": 2.0842789314470784, + "learning_rate": 5.316057437070666e-07, + "loss": 0.2879, + "step": 38903 + }, + { + "epoch": 0.8563504292317852, + "grad_norm": 1.9052091594822116, + "learning_rate": 5.314458105345904e-07, + "loss": 0.2655, + "step": 38904 + }, + { + "epoch": 0.8563724411182039, + "grad_norm": 2.206011766651662, + "learning_rate": 5.312859000733012e-07, + "loss": 0.2164, + "step": 38905 + }, + { + "epoch": 0.8563944530046225, + "grad_norm": 1.357655147566233, + "learning_rate": 5.311260123240086e-07, + "loss": 0.2516, + "step": 38906 + }, + { + "epoch": 0.8564164648910412, + "grad_norm": 1.9167961115944345, + "learning_rate": 5.309661472875283e-07, + "loss": 0.3737, + "step": 38907 + }, + { + "epoch": 0.8564384767774599, + "grad_norm": 1.7261310241559336, + "learning_rate": 5.3080630496467e-07, + "loss": 0.2908, + "step": 38908 + }, + { + "epoch": 0.8564604886638785, + "grad_norm": 1.9255593760445102, + "learning_rate": 5.30646485356247e-07, + "loss": 0.2377, + "step": 38909 + }, + { + "epoch": 0.8564825005502972, + "grad_norm": 1.5797839060068803, + "learning_rate": 5.304866884630738e-07, + "loss": 0.3271, + "step": 38910 + }, + { + "epoch": 0.8565045124367159, + "grad_norm": 1.7377054110901942, + "learning_rate": 5.303269142859591e-07, + "loss": 0.2553, + "step": 38911 + }, + { + "epoch": 0.8565265243231345, + "grad_norm": 0.601411258609944, + "learning_rate": 5.301671628257177e-07, + "loss": 0.4581, + "step": 38912 + }, + { + "epoch": 0.8565485362095532, + "grad_norm": 2.047407642228063, + "learning_rate": 5.300074340831596e-07, + "loss": 0.3658, + "step": 38913 + }, + { + "epoch": 0.8565705480959718, + "grad_norm": 13.245520471890353, + "learning_rate": 5.298477280590969e-07, + "loss": 0.3195, + "step": 38914 + }, + { + "epoch": 0.8565925599823905, + "grad_norm": 1.9809944367473988, + "learning_rate": 5.296880447543434e-07, + "loss": 0.3352, + "step": 38915 + }, + { + "epoch": 0.8566145718688092, + "grad_norm": 3.825611265233498, + "learning_rate": 5.295283841697074e-07, + "loss": 0.2986, + "step": 38916 + }, + { + "epoch": 0.8566365837552278, + "grad_norm": 2.527578722454096, + "learning_rate": 5.293687463060038e-07, + "loss": 0.3392, + "step": 38917 + }, + { + "epoch": 0.8566585956416465, + "grad_norm": 2.0195209355596506, + "learning_rate": 5.292091311640407e-07, + "loss": 0.2931, + "step": 38918 + }, + { + "epoch": 0.8566806075280652, + "grad_norm": 1.8303843303012086, + "learning_rate": 5.29049538744631e-07, + "loss": 0.2427, + "step": 38919 + }, + { + "epoch": 0.8567026194144838, + "grad_norm": 1.8902628638310037, + "learning_rate": 5.288899690485866e-07, + "loss": 0.3607, + "step": 38920 + }, + { + "epoch": 0.8567246313009025, + "grad_norm": 3.047414638857481, + "learning_rate": 5.287304220767165e-07, + "loss": 0.2779, + "step": 38921 + }, + { + "epoch": 0.8567466431873212, + "grad_norm": 1.8599952786106504, + "learning_rate": 5.285708978298332e-07, + "loss": 0.2982, + "step": 38922 + }, + { + "epoch": 0.8567686550737398, + "grad_norm": 4.1915753322251765, + "learning_rate": 5.284113963087467e-07, + "loss": 0.2726, + "step": 38923 + }, + { + "epoch": 0.8567906669601585, + "grad_norm": 0.6209995513967702, + "learning_rate": 5.282519175142675e-07, + "loss": 0.4559, + "step": 38924 + }, + { + "epoch": 0.8568126788465772, + "grad_norm": 2.0545349309156236, + "learning_rate": 5.280924614472072e-07, + "loss": 0.2931, + "step": 38925 + }, + { + "epoch": 0.8568346907329958, + "grad_norm": 1.6475301875170196, + "learning_rate": 5.279330281083761e-07, + "loss": 0.3098, + "step": 38926 + }, + { + "epoch": 0.8568567026194145, + "grad_norm": 1.8486406867916683, + "learning_rate": 5.277736174985826e-07, + "loss": 0.3115, + "step": 38927 + }, + { + "epoch": 0.8568787145058332, + "grad_norm": 1.5727777498422506, + "learning_rate": 5.276142296186381e-07, + "loss": 0.2, + "step": 38928 + }, + { + "epoch": 0.8569007263922518, + "grad_norm": 0.6318530003417689, + "learning_rate": 5.274548644693533e-07, + "loss": 0.4643, + "step": 38929 + }, + { + "epoch": 0.8569227382786705, + "grad_norm": 1.7534339290740693, + "learning_rate": 5.272955220515385e-07, + "loss": 0.2821, + "step": 38930 + }, + { + "epoch": 0.8569447501650892, + "grad_norm": 1.3901541731211662, + "learning_rate": 5.271362023660027e-07, + "loss": 0.279, + "step": 38931 + }, + { + "epoch": 0.8569667620515078, + "grad_norm": 1.6721129604328615, + "learning_rate": 5.269769054135548e-07, + "loss": 0.3042, + "step": 38932 + }, + { + "epoch": 0.8569887739379265, + "grad_norm": 1.6030978884338116, + "learning_rate": 5.268176311950062e-07, + "loss": 0.2824, + "step": 38933 + }, + { + "epoch": 0.8570107858243452, + "grad_norm": 2.4351882149437345, + "learning_rate": 5.266583797111635e-07, + "loss": 0.3038, + "step": 38934 + }, + { + "epoch": 0.8570327977107638, + "grad_norm": 1.5610346716363295, + "learning_rate": 5.264991509628398e-07, + "loss": 0.2847, + "step": 38935 + }, + { + "epoch": 0.8570548095971825, + "grad_norm": 1.6633835884917685, + "learning_rate": 5.263399449508428e-07, + "loss": 0.2413, + "step": 38936 + }, + { + "epoch": 0.8570768214836011, + "grad_norm": 1.8478468979638614, + "learning_rate": 5.261807616759806e-07, + "loss": 0.2675, + "step": 38937 + }, + { + "epoch": 0.8570988333700198, + "grad_norm": 1.4135402524616285, + "learning_rate": 5.260216011390645e-07, + "loss": 0.2658, + "step": 38938 + }, + { + "epoch": 0.8571208452564385, + "grad_norm": 2.4165657861386585, + "learning_rate": 5.258624633409004e-07, + "loss": 0.3382, + "step": 38939 + }, + { + "epoch": 0.8571428571428571, + "grad_norm": 1.9084187503911965, + "learning_rate": 5.257033482822993e-07, + "loss": 0.2958, + "step": 38940 + }, + { + "epoch": 0.8571648690292758, + "grad_norm": 1.6967050688588792, + "learning_rate": 5.2554425596407e-07, + "loss": 0.3135, + "step": 38941 + }, + { + "epoch": 0.8571868809156945, + "grad_norm": 2.3133935296347525, + "learning_rate": 5.253851863870197e-07, + "loss": 0.318, + "step": 38942 + }, + { + "epoch": 0.8572088928021131, + "grad_norm": 1.439924275800393, + "learning_rate": 5.252261395519587e-07, + "loss": 0.2204, + "step": 38943 + }, + { + "epoch": 0.8572309046885318, + "grad_norm": 1.2330811754062219, + "learning_rate": 5.250671154596931e-07, + "loss": 0.3269, + "step": 38944 + }, + { + "epoch": 0.8572529165749505, + "grad_norm": 1.2932420476908275, + "learning_rate": 5.249081141110324e-07, + "loss": 0.3697, + "step": 38945 + }, + { + "epoch": 0.8572749284613691, + "grad_norm": 16.794508042732474, + "learning_rate": 5.247491355067852e-07, + "loss": 0.3103, + "step": 38946 + }, + { + "epoch": 0.8572969403477878, + "grad_norm": 6.1192094423935455, + "learning_rate": 5.245901796477582e-07, + "loss": 0.2825, + "step": 38947 + }, + { + "epoch": 0.8573189522342065, + "grad_norm": 1.7947148820143746, + "learning_rate": 5.244312465347612e-07, + "loss": 0.2306, + "step": 38948 + }, + { + "epoch": 0.8573409641206251, + "grad_norm": 1.3971514558672897, + "learning_rate": 5.242723361685997e-07, + "loss": 0.1721, + "step": 38949 + }, + { + "epoch": 0.8573629760070438, + "grad_norm": 1.504079029551004, + "learning_rate": 5.241134485500821e-07, + "loss": 0.3536, + "step": 38950 + }, + { + "epoch": 0.8573849878934625, + "grad_norm": 0.6486269064880926, + "learning_rate": 5.239545836800175e-07, + "loss": 0.4299, + "step": 38951 + }, + { + "epoch": 0.8574069997798811, + "grad_norm": 1.1581749345469763, + "learning_rate": 5.23795741559211e-07, + "loss": 0.2564, + "step": 38952 + }, + { + "epoch": 0.8574290116662998, + "grad_norm": 1.603778555820147, + "learning_rate": 5.236369221884724e-07, + "loss": 0.2201, + "step": 38953 + }, + { + "epoch": 0.8574510235527185, + "grad_norm": 1.7587159925341567, + "learning_rate": 5.234781255686061e-07, + "loss": 0.3694, + "step": 38954 + }, + { + "epoch": 0.8574730354391371, + "grad_norm": 1.553204540400047, + "learning_rate": 5.233193517004204e-07, + "loss": 0.2229, + "step": 38955 + }, + { + "epoch": 0.8574950473255558, + "grad_norm": 1.7606867479709154, + "learning_rate": 5.231606005847234e-07, + "loss": 0.3154, + "step": 38956 + }, + { + "epoch": 0.8575170592119745, + "grad_norm": 1.8771538758470243, + "learning_rate": 5.230018722223201e-07, + "loss": 0.2615, + "step": 38957 + }, + { + "epoch": 0.8575390710983931, + "grad_norm": 1.7211376055444276, + "learning_rate": 5.228431666140194e-07, + "loss": 0.3291, + "step": 38958 + }, + { + "epoch": 0.8575610829848118, + "grad_norm": 0.6252599843706367, + "learning_rate": 5.226844837606248e-07, + "loss": 0.4753, + "step": 38959 + }, + { + "epoch": 0.8575830948712304, + "grad_norm": 0.6594613722823083, + "learning_rate": 5.225258236629454e-07, + "loss": 0.4558, + "step": 38960 + }, + { + "epoch": 0.8576051067576491, + "grad_norm": 5.026765885118635, + "learning_rate": 5.223671863217872e-07, + "loss": 0.4047, + "step": 38961 + }, + { + "epoch": 0.8576271186440678, + "grad_norm": 2.3667721466616527, + "learning_rate": 5.222085717379566e-07, + "loss": 0.3048, + "step": 38962 + }, + { + "epoch": 0.8576491305304864, + "grad_norm": 1.4598915251041351, + "learning_rate": 5.220499799122581e-07, + "loss": 0.2679, + "step": 38963 + }, + { + "epoch": 0.8576711424169051, + "grad_norm": 2.0270273815015374, + "learning_rate": 5.218914108454981e-07, + "loss": 0.2629, + "step": 38964 + }, + { + "epoch": 0.8576931543033238, + "grad_norm": 1.5379882913480396, + "learning_rate": 5.21732864538484e-07, + "loss": 0.2569, + "step": 38965 + }, + { + "epoch": 0.8577151661897424, + "grad_norm": 0.6355468377014858, + "learning_rate": 5.215743409920215e-07, + "loss": 0.4749, + "step": 38966 + }, + { + "epoch": 0.8577371780761611, + "grad_norm": 1.394610493758079, + "learning_rate": 5.214158402069158e-07, + "loss": 0.2413, + "step": 38967 + }, + { + "epoch": 0.8577591899625798, + "grad_norm": 0.6396204822158215, + "learning_rate": 5.212573621839711e-07, + "loss": 0.4719, + "step": 38968 + }, + { + "epoch": 0.8577812018489984, + "grad_norm": 1.9112460430577205, + "learning_rate": 5.210989069239952e-07, + "loss": 0.346, + "step": 38969 + }, + { + "epoch": 0.8578032137354171, + "grad_norm": 1.408352199630096, + "learning_rate": 5.209404744277907e-07, + "loss": 0.2414, + "step": 38970 + }, + { + "epoch": 0.8578252256218358, + "grad_norm": 1.7886388257008097, + "learning_rate": 5.207820646961659e-07, + "loss": 0.2571, + "step": 38971 + }, + { + "epoch": 0.8578472375082544, + "grad_norm": 1.434454780959295, + "learning_rate": 5.206236777299245e-07, + "loss": 0.2432, + "step": 38972 + }, + { + "epoch": 0.8578692493946731, + "grad_norm": 1.6539470433412542, + "learning_rate": 5.204653135298703e-07, + "loss": 0.3001, + "step": 38973 + }, + { + "epoch": 0.8578912612810918, + "grad_norm": 4.034798697947598, + "learning_rate": 5.203069720968107e-07, + "loss": 0.3449, + "step": 38974 + }, + { + "epoch": 0.8579132731675104, + "grad_norm": 0.6043713982155082, + "learning_rate": 5.201486534315475e-07, + "loss": 0.4529, + "step": 38975 + }, + { + "epoch": 0.8579352850539291, + "grad_norm": 1.5921927644007892, + "learning_rate": 5.199903575348875e-07, + "loss": 0.2339, + "step": 38976 + }, + { + "epoch": 0.8579572969403478, + "grad_norm": 2.0114807431579416, + "learning_rate": 5.198320844076355e-07, + "loss": 0.3546, + "step": 38977 + }, + { + "epoch": 0.8579793088267664, + "grad_norm": 1.5281604250633944, + "learning_rate": 5.196738340505942e-07, + "loss": 0.298, + "step": 38978 + }, + { + "epoch": 0.8580013207131851, + "grad_norm": 1.6813095641403364, + "learning_rate": 5.19515606464569e-07, + "loss": 0.3298, + "step": 38979 + }, + { + "epoch": 0.8580233325996038, + "grad_norm": 1.496318090809886, + "learning_rate": 5.193574016503639e-07, + "loss": 0.1863, + "step": 38980 + }, + { + "epoch": 0.8580453444860224, + "grad_norm": 1.460859893352839, + "learning_rate": 5.191992196087819e-07, + "loss": 0.1873, + "step": 38981 + }, + { + "epoch": 0.8580673563724411, + "grad_norm": 1.9849555229366171, + "learning_rate": 5.190410603406299e-07, + "loss": 0.2005, + "step": 38982 + }, + { + "epoch": 0.8580893682588597, + "grad_norm": 6.30269516450648, + "learning_rate": 5.18882923846708e-07, + "loss": 0.2969, + "step": 38983 + }, + { + "epoch": 0.8581113801452784, + "grad_norm": 0.6097590467357844, + "learning_rate": 5.187248101278231e-07, + "loss": 0.43, + "step": 38984 + }, + { + "epoch": 0.8581333920316971, + "grad_norm": 1.4329733956588555, + "learning_rate": 5.185667191847765e-07, + "loss": 0.2728, + "step": 38985 + }, + { + "epoch": 0.8581554039181157, + "grad_norm": 1.7375281861438285, + "learning_rate": 5.184086510183723e-07, + "loss": 0.2999, + "step": 38986 + }, + { + "epoch": 0.8581774158045344, + "grad_norm": 2.3010531650377697, + "learning_rate": 5.182506056294151e-07, + "loss": 0.3308, + "step": 38987 + }, + { + "epoch": 0.8581994276909531, + "grad_norm": 0.6146726508615947, + "learning_rate": 5.180925830187067e-07, + "loss": 0.4284, + "step": 38988 + }, + { + "epoch": 0.8582214395773717, + "grad_norm": 3.276048040945927, + "learning_rate": 5.179345831870514e-07, + "loss": 0.3022, + "step": 38989 + }, + { + "epoch": 0.8582434514637904, + "grad_norm": 1.9645615432788335, + "learning_rate": 5.177766061352512e-07, + "loss": 0.2449, + "step": 38990 + }, + { + "epoch": 0.8582654633502091, + "grad_norm": 1.251209102980599, + "learning_rate": 5.176186518641085e-07, + "loss": 0.2742, + "step": 38991 + }, + { + "epoch": 0.8582874752366277, + "grad_norm": 2.1375141261063, + "learning_rate": 5.174607203744286e-07, + "loss": 0.2983, + "step": 38992 + }, + { + "epoch": 0.8583094871230464, + "grad_norm": 2.0631352634015747, + "learning_rate": 5.173028116670115e-07, + "loss": 0.3273, + "step": 38993 + }, + { + "epoch": 0.8583314990094651, + "grad_norm": 1.3270696914819033, + "learning_rate": 5.17144925742662e-07, + "loss": 0.2015, + "step": 38994 + }, + { + "epoch": 0.8583535108958837, + "grad_norm": 5.908943389215391, + "learning_rate": 5.1698706260218e-07, + "loss": 0.2987, + "step": 38995 + }, + { + "epoch": 0.8583755227823024, + "grad_norm": 1.5634461288601285, + "learning_rate": 5.168292222463695e-07, + "loss": 0.2539, + "step": 38996 + }, + { + "epoch": 0.8583975346687212, + "grad_norm": 1.63344748954713, + "learning_rate": 5.166714046760335e-07, + "loss": 0.2888, + "step": 38997 + }, + { + "epoch": 0.8584195465551397, + "grad_norm": 2.003338726150691, + "learning_rate": 5.16513609891972e-07, + "loss": 0.3378, + "step": 38998 + }, + { + "epoch": 0.8584415584415584, + "grad_norm": 0.6229584361392383, + "learning_rate": 5.16355837894989e-07, + "loss": 0.4502, + "step": 38999 + }, + { + "epoch": 0.8584635703279772, + "grad_norm": 1.7306248329916716, + "learning_rate": 5.161980886858842e-07, + "loss": 0.2606, + "step": 39000 + }, + { + "epoch": 0.8584855822143957, + "grad_norm": 2.5611862664523204, + "learning_rate": 5.160403622654609e-07, + "loss": 0.2841, + "step": 39001 + }, + { + "epoch": 0.8585075941008145, + "grad_norm": 1.50129424580834, + "learning_rate": 5.158826586345212e-07, + "loss": 0.26, + "step": 39002 + }, + { + "epoch": 0.8585296059872332, + "grad_norm": 1.7547333809882164, + "learning_rate": 5.157249777938655e-07, + "loss": 0.2895, + "step": 39003 + }, + { + "epoch": 0.8585516178736518, + "grad_norm": 11.13698747465751, + "learning_rate": 5.155673197442951e-07, + "loss": 0.2706, + "step": 39004 + }, + { + "epoch": 0.8585736297600705, + "grad_norm": 9.333003314470073, + "learning_rate": 5.154096844866124e-07, + "loss": 0.339, + "step": 39005 + }, + { + "epoch": 0.858595641646489, + "grad_norm": 6.200498354310613, + "learning_rate": 5.152520720216159e-07, + "loss": 0.2896, + "step": 39006 + }, + { + "epoch": 0.8586176535329078, + "grad_norm": 1.4348552277541422, + "learning_rate": 5.150944823501103e-07, + "loss": 0.3037, + "step": 39007 + }, + { + "epoch": 0.8586396654193265, + "grad_norm": 1.5479288913435842, + "learning_rate": 5.149369154728956e-07, + "loss": 0.3249, + "step": 39008 + }, + { + "epoch": 0.858661677305745, + "grad_norm": 1.2476084930091176, + "learning_rate": 5.1477937139077e-07, + "loss": 0.2554, + "step": 39009 + }, + { + "epoch": 0.8586836891921638, + "grad_norm": 1.4903419868592915, + "learning_rate": 5.146218501045375e-07, + "loss": 0.2681, + "step": 39010 + }, + { + "epoch": 0.8587057010785825, + "grad_norm": 1.7363206582262551, + "learning_rate": 5.144643516149966e-07, + "loss": 0.2573, + "step": 39011 + }, + { + "epoch": 0.858727712965001, + "grad_norm": 1.5067381590943907, + "learning_rate": 5.14306875922948e-07, + "loss": 0.3239, + "step": 39012 + }, + { + "epoch": 0.8587497248514198, + "grad_norm": 1.2915864662894243, + "learning_rate": 5.141494230291938e-07, + "loss": 0.2827, + "step": 39013 + }, + { + "epoch": 0.8587717367378385, + "grad_norm": 0.6082472964775084, + "learning_rate": 5.139919929345321e-07, + "loss": 0.4177, + "step": 39014 + }, + { + "epoch": 0.8587937486242571, + "grad_norm": 2.356245343911697, + "learning_rate": 5.138345856397648e-07, + "loss": 0.2242, + "step": 39015 + }, + { + "epoch": 0.8588157605106758, + "grad_norm": 1.818104453737339, + "learning_rate": 5.136772011456897e-07, + "loss": 0.3445, + "step": 39016 + }, + { + "epoch": 0.8588377723970945, + "grad_norm": 1.6985552638016652, + "learning_rate": 5.135198394531082e-07, + "loss": 0.2892, + "step": 39017 + }, + { + "epoch": 0.8588597842835131, + "grad_norm": 1.615808984596709, + "learning_rate": 5.133625005628213e-07, + "loss": 0.3009, + "step": 39018 + }, + { + "epoch": 0.8588817961699318, + "grad_norm": 2.437288644202799, + "learning_rate": 5.132051844756258e-07, + "loss": 0.3531, + "step": 39019 + }, + { + "epoch": 0.8589038080563505, + "grad_norm": 2.003905084864564, + "learning_rate": 5.130478911923237e-07, + "loss": 0.2878, + "step": 39020 + }, + { + "epoch": 0.8589258199427691, + "grad_norm": 1.810894194567265, + "learning_rate": 5.128906207137125e-07, + "loss": 0.289, + "step": 39021 + }, + { + "epoch": 0.8589478318291878, + "grad_norm": 3.3359575200987033, + "learning_rate": 5.127333730405926e-07, + "loss": 0.3156, + "step": 39022 + }, + { + "epoch": 0.8589698437156065, + "grad_norm": 2.8849901143575356, + "learning_rate": 5.12576148173764e-07, + "loss": 0.2634, + "step": 39023 + }, + { + "epoch": 0.8589918556020251, + "grad_norm": 1.6232293369088087, + "learning_rate": 5.124189461140238e-07, + "loss": 0.3562, + "step": 39024 + }, + { + "epoch": 0.8590138674884438, + "grad_norm": 1.5569169306326773, + "learning_rate": 5.122617668621732e-07, + "loss": 0.2259, + "step": 39025 + }, + { + "epoch": 0.8590358793748625, + "grad_norm": 0.6316197228126529, + "learning_rate": 5.121046104190091e-07, + "loss": 0.4334, + "step": 39026 + }, + { + "epoch": 0.8590578912612811, + "grad_norm": 4.765332981820453, + "learning_rate": 5.119474767853305e-07, + "loss": 0.2772, + "step": 39027 + }, + { + "epoch": 0.8590799031476998, + "grad_norm": 1.7124801461489343, + "learning_rate": 5.117903659619377e-07, + "loss": 0.3243, + "step": 39028 + }, + { + "epoch": 0.8591019150341184, + "grad_norm": 1.7143799871488896, + "learning_rate": 5.116332779496275e-07, + "loss": 0.2235, + "step": 39029 + }, + { + "epoch": 0.8591239269205371, + "grad_norm": 1.6459214954190473, + "learning_rate": 5.114762127491996e-07, + "loss": 0.2432, + "step": 39030 + }, + { + "epoch": 0.8591459388069558, + "grad_norm": 1.7169578330885416, + "learning_rate": 5.113191703614501e-07, + "loss": 0.3105, + "step": 39031 + }, + { + "epoch": 0.8591679506933744, + "grad_norm": 1.357494665263639, + "learning_rate": 5.11162150787179e-07, + "loss": 0.2471, + "step": 39032 + }, + { + "epoch": 0.8591899625797931, + "grad_norm": 2.0288010653386928, + "learning_rate": 5.110051540271849e-07, + "loss": 0.2515, + "step": 39033 + }, + { + "epoch": 0.8592119744662118, + "grad_norm": 1.688398030313192, + "learning_rate": 5.108481800822634e-07, + "loss": 0.3336, + "step": 39034 + }, + { + "epoch": 0.8592339863526304, + "grad_norm": 1.8492886964852826, + "learning_rate": 5.106912289532151e-07, + "loss": 0.3522, + "step": 39035 + }, + { + "epoch": 0.8592559982390491, + "grad_norm": 1.875090117023856, + "learning_rate": 5.105343006408365e-07, + "loss": 0.3669, + "step": 39036 + }, + { + "epoch": 0.8592780101254678, + "grad_norm": 1.528073901232951, + "learning_rate": 5.103773951459224e-07, + "loss": 0.3752, + "step": 39037 + }, + { + "epoch": 0.8593000220118864, + "grad_norm": 1.4788553776968298, + "learning_rate": 5.10220512469275e-07, + "loss": 0.2177, + "step": 39038 + }, + { + "epoch": 0.8593220338983051, + "grad_norm": 0.6176286436210668, + "learning_rate": 5.100636526116881e-07, + "loss": 0.4421, + "step": 39039 + }, + { + "epoch": 0.8593440457847238, + "grad_norm": 0.6041873988671406, + "learning_rate": 5.099068155739617e-07, + "loss": 0.4304, + "step": 39040 + }, + { + "epoch": 0.8593660576711424, + "grad_norm": 2.04925067578499, + "learning_rate": 5.097500013568918e-07, + "loss": 0.274, + "step": 39041 + }, + { + "epoch": 0.8593880695575611, + "grad_norm": 1.914827732911697, + "learning_rate": 5.095932099612727e-07, + "loss": 0.2881, + "step": 39042 + }, + { + "epoch": 0.8594100814439798, + "grad_norm": 1.658668939410924, + "learning_rate": 5.094364413879055e-07, + "loss": 0.2032, + "step": 39043 + }, + { + "epoch": 0.8594320933303984, + "grad_norm": 2.1155365794279035, + "learning_rate": 5.092796956375851e-07, + "loss": 0.3476, + "step": 39044 + }, + { + "epoch": 0.8594541052168171, + "grad_norm": 0.6116565201683029, + "learning_rate": 5.091229727111075e-07, + "loss": 0.4223, + "step": 39045 + }, + { + "epoch": 0.8594761171032358, + "grad_norm": 1.8819648109330513, + "learning_rate": 5.089662726092709e-07, + "loss": 0.2041, + "step": 39046 + }, + { + "epoch": 0.8594981289896544, + "grad_norm": 1.392762761477312, + "learning_rate": 5.088095953328697e-07, + "loss": 0.2941, + "step": 39047 + }, + { + "epoch": 0.8595201408760731, + "grad_norm": 0.6226945469210803, + "learning_rate": 5.086529408827018e-07, + "loss": 0.4291, + "step": 39048 + }, + { + "epoch": 0.8595421527624918, + "grad_norm": 1.6002030073434683, + "learning_rate": 5.084963092595629e-07, + "loss": 0.2967, + "step": 39049 + }, + { + "epoch": 0.8595641646489104, + "grad_norm": 0.6164276904652626, + "learning_rate": 5.083397004642488e-07, + "loss": 0.4553, + "step": 39050 + }, + { + "epoch": 0.8595861765353291, + "grad_norm": 1.5657930360271408, + "learning_rate": 5.081831144975563e-07, + "loss": 0.213, + "step": 39051 + }, + { + "epoch": 0.8596081884217478, + "grad_norm": 1.4817253867603897, + "learning_rate": 5.080265513602794e-07, + "loss": 0.2286, + "step": 39052 + }, + { + "epoch": 0.8596302003081664, + "grad_norm": 1.3133552375155584, + "learning_rate": 5.078700110532154e-07, + "loss": 0.2922, + "step": 39053 + }, + { + "epoch": 0.8596522121945851, + "grad_norm": 1.647253337064531, + "learning_rate": 5.077134935771605e-07, + "loss": 0.3339, + "step": 39054 + }, + { + "epoch": 0.8596742240810037, + "grad_norm": 4.025185366579195, + "learning_rate": 5.075569989329082e-07, + "loss": 0.3059, + "step": 39055 + }, + { + "epoch": 0.8596962359674224, + "grad_norm": 1.7909243796100582, + "learning_rate": 5.074005271212562e-07, + "loss": 0.3217, + "step": 39056 + }, + { + "epoch": 0.8597182478538411, + "grad_norm": 2.5363774053445756, + "learning_rate": 5.072440781429972e-07, + "loss": 0.3482, + "step": 39057 + }, + { + "epoch": 0.8597402597402597, + "grad_norm": 1.9562597782098992, + "learning_rate": 5.070876519989276e-07, + "loss": 0.2507, + "step": 39058 + }, + { + "epoch": 0.8597622716266784, + "grad_norm": 1.44939786606156, + "learning_rate": 5.069312486898437e-07, + "loss": 0.2595, + "step": 39059 + }, + { + "epoch": 0.8597842835130971, + "grad_norm": 1.905582809976986, + "learning_rate": 5.067748682165375e-07, + "loss": 0.3561, + "step": 39060 + }, + { + "epoch": 0.8598062953995157, + "grad_norm": 2.067559007065608, + "learning_rate": 5.066185105798071e-07, + "loss": 0.3616, + "step": 39061 + }, + { + "epoch": 0.8598283072859344, + "grad_norm": 1.7583304888748212, + "learning_rate": 5.064621757804445e-07, + "loss": 0.2973, + "step": 39062 + }, + { + "epoch": 0.8598503191723531, + "grad_norm": 0.628506362793448, + "learning_rate": 5.063058638192448e-07, + "loss": 0.4619, + "step": 39063 + }, + { + "epoch": 0.8598723310587717, + "grad_norm": 1.5587821036279994, + "learning_rate": 5.061495746970046e-07, + "loss": 0.305, + "step": 39064 + }, + { + "epoch": 0.8598943429451904, + "grad_norm": 1.2569201418880709, + "learning_rate": 5.059933084145152e-07, + "loss": 0.2713, + "step": 39065 + }, + { + "epoch": 0.8599163548316091, + "grad_norm": 1.5937805819377489, + "learning_rate": 5.05837064972573e-07, + "loss": 0.3193, + "step": 39066 + }, + { + "epoch": 0.8599383667180277, + "grad_norm": 2.1151417944990287, + "learning_rate": 5.056808443719708e-07, + "loss": 0.2731, + "step": 39067 + }, + { + "epoch": 0.8599603786044464, + "grad_norm": 1.7952613658007432, + "learning_rate": 5.055246466135028e-07, + "loss": 0.3147, + "step": 39068 + }, + { + "epoch": 0.8599823904908651, + "grad_norm": 1.3176908012516846, + "learning_rate": 5.05368471697964e-07, + "loss": 0.2295, + "step": 39069 + }, + { + "epoch": 0.8600044023772837, + "grad_norm": 2.962900030832504, + "learning_rate": 5.052123196261466e-07, + "loss": 0.2446, + "step": 39070 + }, + { + "epoch": 0.8600264142637024, + "grad_norm": 1.84554464484288, + "learning_rate": 5.050561903988454e-07, + "loss": 0.2692, + "step": 39071 + }, + { + "epoch": 0.8600484261501211, + "grad_norm": 2.212927421584103, + "learning_rate": 5.049000840168539e-07, + "loss": 0.3798, + "step": 39072 + }, + { + "epoch": 0.8600704380365397, + "grad_norm": 0.6256900809256052, + "learning_rate": 5.047440004809628e-07, + "loss": 0.4581, + "step": 39073 + }, + { + "epoch": 0.8600924499229584, + "grad_norm": 1.7483619735720526, + "learning_rate": 5.045879397919701e-07, + "loss": 0.3218, + "step": 39074 + }, + { + "epoch": 0.8601144618093771, + "grad_norm": 1.898815539188033, + "learning_rate": 5.04431901950665e-07, + "loss": 0.3402, + "step": 39075 + }, + { + "epoch": 0.8601364736957957, + "grad_norm": 1.583305098993444, + "learning_rate": 5.042758869578429e-07, + "loss": 0.3067, + "step": 39076 + }, + { + "epoch": 0.8601584855822144, + "grad_norm": 1.9527718781765386, + "learning_rate": 5.041198948142966e-07, + "loss": 0.2609, + "step": 39077 + }, + { + "epoch": 0.860180497468633, + "grad_norm": 2.0471317017719186, + "learning_rate": 5.039639255208156e-07, + "loss": 0.2842, + "step": 39078 + }, + { + "epoch": 0.8602025093550517, + "grad_norm": 1.8336099365588239, + "learning_rate": 5.038079790781974e-07, + "loss": 0.2357, + "step": 39079 + }, + { + "epoch": 0.8602245212414704, + "grad_norm": 0.6130696471031695, + "learning_rate": 5.036520554872315e-07, + "loss": 0.4394, + "step": 39080 + }, + { + "epoch": 0.860246533127889, + "grad_norm": 1.3803021874618095, + "learning_rate": 5.034961547487122e-07, + "loss": 0.253, + "step": 39081 + }, + { + "epoch": 0.8602685450143077, + "grad_norm": 1.6749446942518178, + "learning_rate": 5.03340276863431e-07, + "loss": 0.2909, + "step": 39082 + }, + { + "epoch": 0.8602905569007264, + "grad_norm": 1.5365530533728702, + "learning_rate": 5.031844218321791e-07, + "loss": 0.3358, + "step": 39083 + }, + { + "epoch": 0.860312568787145, + "grad_norm": 1.8429436711049334, + "learning_rate": 5.030285896557497e-07, + "loss": 0.302, + "step": 39084 + }, + { + "epoch": 0.8603345806735637, + "grad_norm": 1.739260393585852, + "learning_rate": 5.028727803349354e-07, + "loss": 0.2597, + "step": 39085 + }, + { + "epoch": 0.8603565925599824, + "grad_norm": 1.3827522954523135, + "learning_rate": 5.027169938705268e-07, + "loss": 0.2758, + "step": 39086 + }, + { + "epoch": 0.860378604446401, + "grad_norm": 2.0637280184925624, + "learning_rate": 5.025612302633165e-07, + "loss": 0.3497, + "step": 39087 + }, + { + "epoch": 0.8604006163328197, + "grad_norm": 3.3955968585582865, + "learning_rate": 5.024054895140956e-07, + "loss": 0.2771, + "step": 39088 + }, + { + "epoch": 0.8604226282192384, + "grad_norm": 1.9331072713183437, + "learning_rate": 5.022497716236563e-07, + "loss": 0.2918, + "step": 39089 + }, + { + "epoch": 0.860444640105657, + "grad_norm": 2.1726806832847694, + "learning_rate": 5.020940765927901e-07, + "loss": 0.3206, + "step": 39090 + }, + { + "epoch": 0.8604666519920757, + "grad_norm": 2.5323966620063914, + "learning_rate": 5.019384044222869e-07, + "loss": 0.2276, + "step": 39091 + }, + { + "epoch": 0.8604886638784944, + "grad_norm": 1.7090484691155874, + "learning_rate": 5.0178275511294e-07, + "loss": 0.2562, + "step": 39092 + }, + { + "epoch": 0.860510675764913, + "grad_norm": 2.669827686826133, + "learning_rate": 5.016271286655389e-07, + "loss": 0.2193, + "step": 39093 + }, + { + "epoch": 0.8605326876513317, + "grad_norm": 1.3128841404487253, + "learning_rate": 5.014715250808744e-07, + "loss": 0.2706, + "step": 39094 + }, + { + "epoch": 0.8605546995377504, + "grad_norm": 1.4914566662022255, + "learning_rate": 5.013159443597393e-07, + "loss": 0.2866, + "step": 39095 + }, + { + "epoch": 0.860576711424169, + "grad_norm": 1.2444697313377524, + "learning_rate": 5.011603865029225e-07, + "loss": 0.2284, + "step": 39096 + }, + { + "epoch": 0.8605987233105877, + "grad_norm": 1.9544543939849277, + "learning_rate": 5.010048515112154e-07, + "loss": 0.2369, + "step": 39097 + }, + { + "epoch": 0.8606207351970064, + "grad_norm": 1.3506691226639997, + "learning_rate": 5.008493393854075e-07, + "loss": 0.2498, + "step": 39098 + }, + { + "epoch": 0.860642747083425, + "grad_norm": 1.4578782033347564, + "learning_rate": 5.006938501262898e-07, + "loss": 0.2806, + "step": 39099 + }, + { + "epoch": 0.8606647589698437, + "grad_norm": 1.7571251618837036, + "learning_rate": 5.005383837346539e-07, + "loss": 0.3383, + "step": 39100 + }, + { + "epoch": 0.8606867708562623, + "grad_norm": 1.2533485332134564, + "learning_rate": 5.003829402112881e-07, + "loss": 0.2168, + "step": 39101 + }, + { + "epoch": 0.860708782742681, + "grad_norm": 4.830536875667069, + "learning_rate": 5.002275195569833e-07, + "loss": 0.197, + "step": 39102 + }, + { + "epoch": 0.8607307946290997, + "grad_norm": 1.7372900041980461, + "learning_rate": 5.000721217725291e-07, + "loss": 0.3681, + "step": 39103 + }, + { + "epoch": 0.8607528065155183, + "grad_norm": 5.286129242003826, + "learning_rate": 4.999167468587146e-07, + "loss": 0.2244, + "step": 39104 + }, + { + "epoch": 0.860774818401937, + "grad_norm": 1.4824899715514133, + "learning_rate": 4.997613948163316e-07, + "loss": 0.3113, + "step": 39105 + }, + { + "epoch": 0.8607968302883557, + "grad_norm": 1.7590496210610884, + "learning_rate": 4.996060656461677e-07, + "loss": 0.1639, + "step": 39106 + }, + { + "epoch": 0.8608188421747743, + "grad_norm": 2.2093844631889192, + "learning_rate": 4.994507593490133e-07, + "loss": 0.3055, + "step": 39107 + }, + { + "epoch": 0.860840854061193, + "grad_norm": 1.4824312138048563, + "learning_rate": 4.99295475925658e-07, + "loss": 0.2272, + "step": 39108 + }, + { + "epoch": 0.8608628659476117, + "grad_norm": 1.588956120685605, + "learning_rate": 4.991402153768882e-07, + "loss": 0.3546, + "step": 39109 + }, + { + "epoch": 0.8608848778340303, + "grad_norm": 2.5983673653280044, + "learning_rate": 4.989849777034972e-07, + "loss": 0.3202, + "step": 39110 + }, + { + "epoch": 0.860906889720449, + "grad_norm": 1.5744123778034078, + "learning_rate": 4.98829762906271e-07, + "loss": 0.3127, + "step": 39111 + }, + { + "epoch": 0.8609289016068677, + "grad_norm": 0.6112386488758859, + "learning_rate": 4.986745709860008e-07, + "loss": 0.4437, + "step": 39112 + }, + { + "epoch": 0.8609509134932863, + "grad_norm": 1.5234932078617451, + "learning_rate": 4.985194019434742e-07, + "loss": 0.2337, + "step": 39113 + }, + { + "epoch": 0.860972925379705, + "grad_norm": 3.89472804764624, + "learning_rate": 4.983642557794776e-07, + "loss": 0.3968, + "step": 39114 + }, + { + "epoch": 0.8609949372661237, + "grad_norm": 1.934035670265208, + "learning_rate": 4.982091324948035e-07, + "loss": 0.2664, + "step": 39115 + }, + { + "epoch": 0.8610169491525423, + "grad_norm": 1.468647053352291, + "learning_rate": 4.980540320902371e-07, + "loss": 0.3706, + "step": 39116 + }, + { + "epoch": 0.861038961038961, + "grad_norm": 1.8309532779165676, + "learning_rate": 4.978989545665697e-07, + "loss": 0.2971, + "step": 39117 + }, + { + "epoch": 0.8610609729253798, + "grad_norm": 0.6300532378217594, + "learning_rate": 4.977438999245876e-07, + "loss": 0.4399, + "step": 39118 + }, + { + "epoch": 0.8610829848117983, + "grad_norm": 2.0715415574120475, + "learning_rate": 4.975888681650781e-07, + "loss": 0.2813, + "step": 39119 + }, + { + "epoch": 0.861104996698217, + "grad_norm": 1.8935895176606756, + "learning_rate": 4.974338592888301e-07, + "loss": 0.2995, + "step": 39120 + }, + { + "epoch": 0.8611270085846358, + "grad_norm": 2.759078155582722, + "learning_rate": 4.97278873296631e-07, + "loss": 0.2965, + "step": 39121 + }, + { + "epoch": 0.8611490204710543, + "grad_norm": 1.8909097345277788, + "learning_rate": 4.971239101892699e-07, + "loss": 0.294, + "step": 39122 + }, + { + "epoch": 0.861171032357473, + "grad_norm": 1.8025862513292852, + "learning_rate": 4.969689699675339e-07, + "loss": 0.2617, + "step": 39123 + }, + { + "epoch": 0.8611930442438916, + "grad_norm": 2.3735097468311865, + "learning_rate": 4.968140526322091e-07, + "loss": 0.265, + "step": 39124 + }, + { + "epoch": 0.8612150561303104, + "grad_norm": 1.5569733065930353, + "learning_rate": 4.966591581840835e-07, + "loss": 0.2033, + "step": 39125 + }, + { + "epoch": 0.8612370680167291, + "grad_norm": 26.743047361027678, + "learning_rate": 4.965042866239456e-07, + "loss": 0.2714, + "step": 39126 + }, + { + "epoch": 0.8612590799031477, + "grad_norm": 0.6090579562749552, + "learning_rate": 4.963494379525808e-07, + "loss": 0.4445, + "step": 39127 + }, + { + "epoch": 0.8612810917895664, + "grad_norm": 1.3197108441802496, + "learning_rate": 4.961946121707772e-07, + "loss": 0.2528, + "step": 39128 + }, + { + "epoch": 0.8613031036759851, + "grad_norm": 2.436654760623628, + "learning_rate": 4.960398092793212e-07, + "loss": 0.369, + "step": 39129 + }, + { + "epoch": 0.8613251155624037, + "grad_norm": 1.4614591806075519, + "learning_rate": 4.958850292789991e-07, + "loss": 0.2725, + "step": 39130 + }, + { + "epoch": 0.8613471274488224, + "grad_norm": 122.62812746160458, + "learning_rate": 4.957302721705992e-07, + "loss": 0.3686, + "step": 39131 + }, + { + "epoch": 0.8613691393352411, + "grad_norm": 1.6787676890015621, + "learning_rate": 4.955755379549059e-07, + "loss": 0.2257, + "step": 39132 + }, + { + "epoch": 0.8613911512216597, + "grad_norm": 1.761993101741713, + "learning_rate": 4.954208266327082e-07, + "loss": 0.2667, + "step": 39133 + }, + { + "epoch": 0.8614131631080784, + "grad_norm": 1.7667219117348834, + "learning_rate": 4.952661382047896e-07, + "loss": 0.2229, + "step": 39134 + }, + { + "epoch": 0.8614351749944971, + "grad_norm": 1.6855132282090224, + "learning_rate": 4.95111472671938e-07, + "loss": 0.3019, + "step": 39135 + }, + { + "epoch": 0.8614571868809157, + "grad_norm": 0.6237941264676153, + "learning_rate": 4.9495683003494e-07, + "loss": 0.4589, + "step": 39136 + }, + { + "epoch": 0.8614791987673344, + "grad_norm": 0.6071871123661148, + "learning_rate": 4.948022102945798e-07, + "loss": 0.4564, + "step": 39137 + }, + { + "epoch": 0.8615012106537531, + "grad_norm": 2.100461818390678, + "learning_rate": 4.946476134516448e-07, + "loss": 0.2766, + "step": 39138 + }, + { + "epoch": 0.8615232225401717, + "grad_norm": 2.628473984095668, + "learning_rate": 4.944930395069192e-07, + "loss": 0.255, + "step": 39139 + }, + { + "epoch": 0.8615452344265904, + "grad_norm": 1.4928034454180423, + "learning_rate": 4.943384884611896e-07, + "loss": 0.1853, + "step": 39140 + }, + { + "epoch": 0.8615672463130091, + "grad_norm": 1.4503531553370121, + "learning_rate": 4.941839603152426e-07, + "loss": 0.2742, + "step": 39141 + }, + { + "epoch": 0.8615892581994277, + "grad_norm": 1.8765691185454794, + "learning_rate": 4.94029455069861e-07, + "loss": 0.2141, + "step": 39142 + }, + { + "epoch": 0.8616112700858464, + "grad_norm": 1.4707584946171168, + "learning_rate": 4.938749727258324e-07, + "loss": 0.2546, + "step": 39143 + }, + { + "epoch": 0.8616332819722651, + "grad_norm": 1.3755630353352115, + "learning_rate": 4.937205132839412e-07, + "loss": 0.3664, + "step": 39144 + }, + { + "epoch": 0.8616552938586837, + "grad_norm": 1.4456191950045785, + "learning_rate": 4.935660767449701e-07, + "loss": 0.2628, + "step": 39145 + }, + { + "epoch": 0.8616773057451024, + "grad_norm": 1.7627197037699958, + "learning_rate": 4.93411663109708e-07, + "loss": 0.2627, + "step": 39146 + }, + { + "epoch": 0.861699317631521, + "grad_norm": 0.6015362523937913, + "learning_rate": 4.932572723789369e-07, + "loss": 0.4391, + "step": 39147 + }, + { + "epoch": 0.8617213295179397, + "grad_norm": 1.7132061706686366, + "learning_rate": 4.931029045534435e-07, + "loss": 0.3064, + "step": 39148 + }, + { + "epoch": 0.8617433414043584, + "grad_norm": 0.6077325431512444, + "learning_rate": 4.929485596340111e-07, + "loss": 0.4505, + "step": 39149 + }, + { + "epoch": 0.861765353290777, + "grad_norm": 1.4213410414512415, + "learning_rate": 4.927942376214234e-07, + "loss": 0.2557, + "step": 39150 + }, + { + "epoch": 0.8617873651771957, + "grad_norm": 2.252690110718226, + "learning_rate": 4.926399385164659e-07, + "loss": 0.2942, + "step": 39151 + }, + { + "epoch": 0.8618093770636144, + "grad_norm": 1.2789605461728453, + "learning_rate": 4.92485662319922e-07, + "loss": 0.2413, + "step": 39152 + }, + { + "epoch": 0.861831388950033, + "grad_norm": 1.4156263358710326, + "learning_rate": 4.923314090325781e-07, + "loss": 0.262, + "step": 39153 + }, + { + "epoch": 0.8618534008364517, + "grad_norm": 1.788443928315695, + "learning_rate": 4.92177178655216e-07, + "loss": 0.3549, + "step": 39154 + }, + { + "epoch": 0.8618754127228704, + "grad_norm": 1.8443653884137792, + "learning_rate": 4.920229711886193e-07, + "loss": 0.263, + "step": 39155 + }, + { + "epoch": 0.861897424609289, + "grad_norm": 1.7998218131821262, + "learning_rate": 4.918687866335719e-07, + "loss": 0.3008, + "step": 39156 + }, + { + "epoch": 0.8619194364957077, + "grad_norm": 2.1665833367238156, + "learning_rate": 4.917146249908583e-07, + "loss": 0.2978, + "step": 39157 + }, + { + "epoch": 0.8619414483821264, + "grad_norm": 1.7775286145738467, + "learning_rate": 4.915604862612633e-07, + "loss": 0.269, + "step": 39158 + }, + { + "epoch": 0.861963460268545, + "grad_norm": 0.6279173904738993, + "learning_rate": 4.914063704455679e-07, + "loss": 0.4354, + "step": 39159 + }, + { + "epoch": 0.8619854721549637, + "grad_norm": 2.09546803440297, + "learning_rate": 4.912522775445555e-07, + "loss": 0.347, + "step": 39160 + }, + { + "epoch": 0.8620074840413824, + "grad_norm": 10.517328527035056, + "learning_rate": 4.910982075590104e-07, + "loss": 0.309, + "step": 39161 + }, + { + "epoch": 0.862029495927801, + "grad_norm": 2.426397406927346, + "learning_rate": 4.909441604897158e-07, + "loss": 0.2647, + "step": 39162 + }, + { + "epoch": 0.8620515078142197, + "grad_norm": 2.184304338274622, + "learning_rate": 4.907901363374529e-07, + "loss": 0.3012, + "step": 39163 + }, + { + "epoch": 0.8620735197006384, + "grad_norm": 0.6002582654408567, + "learning_rate": 4.906361351030071e-07, + "loss": 0.4515, + "step": 39164 + }, + { + "epoch": 0.862095531587057, + "grad_norm": 2.2446864613375856, + "learning_rate": 4.904821567871587e-07, + "loss": 0.3957, + "step": 39165 + }, + { + "epoch": 0.8621175434734757, + "grad_norm": 1.9561878093371055, + "learning_rate": 4.903282013906913e-07, + "loss": 0.2904, + "step": 39166 + }, + { + "epoch": 0.8621395553598944, + "grad_norm": 1.7665669419328516, + "learning_rate": 4.901742689143879e-07, + "loss": 0.3223, + "step": 39167 + }, + { + "epoch": 0.862161567246313, + "grad_norm": 1.777438077164923, + "learning_rate": 4.900203593590297e-07, + "loss": 0.3384, + "step": 39168 + }, + { + "epoch": 0.8621835791327317, + "grad_norm": 6.196691802016361, + "learning_rate": 4.898664727254005e-07, + "loss": 0.3028, + "step": 39169 + }, + { + "epoch": 0.8622055910191503, + "grad_norm": 1.69549461379226, + "learning_rate": 4.897126090142801e-07, + "loss": 0.248, + "step": 39170 + }, + { + "epoch": 0.862227602905569, + "grad_norm": 1.658422069868814, + "learning_rate": 4.895587682264524e-07, + "loss": 0.3026, + "step": 39171 + }, + { + "epoch": 0.8622496147919877, + "grad_norm": 2.595022492298288, + "learning_rate": 4.894049503626991e-07, + "loss": 0.2255, + "step": 39172 + }, + { + "epoch": 0.8622716266784063, + "grad_norm": 0.6322491128819401, + "learning_rate": 4.89251155423801e-07, + "loss": 0.4712, + "step": 39173 + }, + { + "epoch": 0.862293638564825, + "grad_norm": 0.6133352992429865, + "learning_rate": 4.890973834105417e-07, + "loss": 0.4537, + "step": 39174 + }, + { + "epoch": 0.8623156504512437, + "grad_norm": 1.843921171547854, + "learning_rate": 4.889436343237003e-07, + "loss": 0.3022, + "step": 39175 + }, + { + "epoch": 0.8623376623376623, + "grad_norm": 1.505247026503814, + "learning_rate": 4.887899081640591e-07, + "loss": 0.3072, + "step": 39176 + }, + { + "epoch": 0.862359674224081, + "grad_norm": 25.713732522700404, + "learning_rate": 4.886362049324006e-07, + "loss": 0.3208, + "step": 39177 + }, + { + "epoch": 0.8623816861104997, + "grad_norm": 2.1767845265748553, + "learning_rate": 4.884825246295039e-07, + "loss": 0.2491, + "step": 39178 + }, + { + "epoch": 0.8624036979969183, + "grad_norm": 2.0368922374563705, + "learning_rate": 4.883288672561526e-07, + "loss": 0.3412, + "step": 39179 + }, + { + "epoch": 0.862425709883337, + "grad_norm": 1.7291968561672835, + "learning_rate": 4.88175232813125e-07, + "loss": 0.3105, + "step": 39180 + }, + { + "epoch": 0.8624477217697557, + "grad_norm": 2.109248629786953, + "learning_rate": 4.880216213012034e-07, + "loss": 0.2847, + "step": 39181 + }, + { + "epoch": 0.8624697336561743, + "grad_norm": 0.6147444507555967, + "learning_rate": 4.878680327211694e-07, + "loss": 0.4512, + "step": 39182 + }, + { + "epoch": 0.862491745542593, + "grad_norm": 1.670665660338974, + "learning_rate": 4.87714467073801e-07, + "loss": 0.2693, + "step": 39183 + }, + { + "epoch": 0.8625137574290117, + "grad_norm": 0.611951484433609, + "learning_rate": 4.875609243598817e-07, + "loss": 0.4434, + "step": 39184 + }, + { + "epoch": 0.8625357693154303, + "grad_norm": 2.038887846289754, + "learning_rate": 4.874074045801902e-07, + "loss": 0.3733, + "step": 39185 + }, + { + "epoch": 0.862557781201849, + "grad_norm": 2.294081899177377, + "learning_rate": 4.87253907735506e-07, + "loss": 0.3302, + "step": 39186 + }, + { + "epoch": 0.8625797930882677, + "grad_norm": 2.189793998399405, + "learning_rate": 4.8710043382661e-07, + "loss": 0.2211, + "step": 39187 + }, + { + "epoch": 0.8626018049746863, + "grad_norm": 1.5056184978642957, + "learning_rate": 4.869469828542827e-07, + "loss": 0.2451, + "step": 39188 + }, + { + "epoch": 0.862623816861105, + "grad_norm": 2.139846369580775, + "learning_rate": 4.867935548193042e-07, + "loss": 0.2717, + "step": 39189 + }, + { + "epoch": 0.8626458287475237, + "grad_norm": 0.6054495266688582, + "learning_rate": 4.866401497224543e-07, + "loss": 0.4516, + "step": 39190 + }, + { + "epoch": 0.8626678406339423, + "grad_norm": 0.5910227971151034, + "learning_rate": 4.864867675645113e-07, + "loss": 0.4174, + "step": 39191 + }, + { + "epoch": 0.862689852520361, + "grad_norm": 4.127121272450634, + "learning_rate": 4.863334083462551e-07, + "loss": 0.3184, + "step": 39192 + }, + { + "epoch": 0.8627118644067797, + "grad_norm": 1.4893349961748192, + "learning_rate": 4.861800720684662e-07, + "loss": 0.2599, + "step": 39193 + }, + { + "epoch": 0.8627338762931983, + "grad_norm": 1.9111977410735803, + "learning_rate": 4.86026758731924e-07, + "loss": 0.2849, + "step": 39194 + }, + { + "epoch": 0.862755888179617, + "grad_norm": 1.6627029449736896, + "learning_rate": 4.858734683374072e-07, + "loss": 0.2799, + "step": 39195 + }, + { + "epoch": 0.8627779000660356, + "grad_norm": 0.6278229586827939, + "learning_rate": 4.857202008856937e-07, + "loss": 0.4552, + "step": 39196 + }, + { + "epoch": 0.8627999119524543, + "grad_norm": 1.6734122956851287, + "learning_rate": 4.855669563775639e-07, + "loss": 0.3436, + "step": 39197 + }, + { + "epoch": 0.862821923838873, + "grad_norm": 2.681632744133246, + "learning_rate": 4.85413734813796e-07, + "loss": 0.3186, + "step": 39198 + }, + { + "epoch": 0.8628439357252916, + "grad_norm": 1.7892991051910643, + "learning_rate": 4.852605361951701e-07, + "loss": 0.2603, + "step": 39199 + }, + { + "epoch": 0.8628659476117103, + "grad_norm": 0.6128919745888427, + "learning_rate": 4.851073605224637e-07, + "loss": 0.4446, + "step": 39200 + }, + { + "epoch": 0.862887959498129, + "grad_norm": 0.6109580766751392, + "learning_rate": 4.849542077964547e-07, + "loss": 0.4548, + "step": 39201 + }, + { + "epoch": 0.8629099713845476, + "grad_norm": 5.431105591664478, + "learning_rate": 4.848010780179225e-07, + "loss": 0.2983, + "step": 39202 + }, + { + "epoch": 0.8629319832709663, + "grad_norm": 1.916668312453638, + "learning_rate": 4.84647971187645e-07, + "loss": 0.3477, + "step": 39203 + }, + { + "epoch": 0.862953995157385, + "grad_norm": 1.4299248191433604, + "learning_rate": 4.844948873064004e-07, + "loss": 0.239, + "step": 39204 + }, + { + "epoch": 0.8629760070438036, + "grad_norm": 2.2146575874751515, + "learning_rate": 4.843418263749672e-07, + "loss": 0.2925, + "step": 39205 + }, + { + "epoch": 0.8629980189302223, + "grad_norm": 1.5248606751869458, + "learning_rate": 4.84188788394122e-07, + "loss": 0.2714, + "step": 39206 + }, + { + "epoch": 0.863020030816641, + "grad_norm": 1.607584455912929, + "learning_rate": 4.840357733646433e-07, + "loss": 0.2546, + "step": 39207 + }, + { + "epoch": 0.8630420427030596, + "grad_norm": 1.588090962063689, + "learning_rate": 4.838827812873103e-07, + "loss": 0.2556, + "step": 39208 + }, + { + "epoch": 0.8630640545894783, + "grad_norm": 3.766542186892093, + "learning_rate": 4.837298121628986e-07, + "loss": 0.3319, + "step": 39209 + }, + { + "epoch": 0.863086066475897, + "grad_norm": 1.5348057120581589, + "learning_rate": 4.835768659921864e-07, + "loss": 0.2518, + "step": 39210 + }, + { + "epoch": 0.8631080783623156, + "grad_norm": 3.818941190438795, + "learning_rate": 4.834239427759507e-07, + "loss": 0.3232, + "step": 39211 + }, + { + "epoch": 0.8631300902487343, + "grad_norm": 9.758176807239565, + "learning_rate": 4.832710425149684e-07, + "loss": 0.3486, + "step": 39212 + }, + { + "epoch": 0.863152102135153, + "grad_norm": 1.8638207507274558, + "learning_rate": 4.831181652100187e-07, + "loss": 0.2851, + "step": 39213 + }, + { + "epoch": 0.8631741140215716, + "grad_norm": 1.9391490312755517, + "learning_rate": 4.82965310861876e-07, + "loss": 0.3268, + "step": 39214 + }, + { + "epoch": 0.8631961259079903, + "grad_norm": 0.6301817925657895, + "learning_rate": 4.82812479471319e-07, + "loss": 0.4662, + "step": 39215 + }, + { + "epoch": 0.863218137794409, + "grad_norm": 2.636908473784351, + "learning_rate": 4.826596710391224e-07, + "loss": 0.2162, + "step": 39216 + }, + { + "epoch": 0.8632401496808276, + "grad_norm": 2.1217282016597996, + "learning_rate": 4.825068855660647e-07, + "loss": 0.2896, + "step": 39217 + }, + { + "epoch": 0.8632621615672463, + "grad_norm": 0.6144552295718689, + "learning_rate": 4.823541230529228e-07, + "loss": 0.4429, + "step": 39218 + }, + { + "epoch": 0.8632841734536649, + "grad_norm": 1.8609835252367046, + "learning_rate": 4.822013835004707e-07, + "loss": 0.3463, + "step": 39219 + }, + { + "epoch": 0.8633061853400836, + "grad_norm": 2.021766154938739, + "learning_rate": 4.820486669094876e-07, + "loss": 0.2348, + "step": 39220 + }, + { + "epoch": 0.8633281972265023, + "grad_norm": 1.7646260671852958, + "learning_rate": 4.818959732807472e-07, + "loss": 0.2847, + "step": 39221 + }, + { + "epoch": 0.8633502091129209, + "grad_norm": 1.9996632959605705, + "learning_rate": 4.817433026150276e-07, + "loss": 0.3334, + "step": 39222 + }, + { + "epoch": 0.8633722209993396, + "grad_norm": 1.276400309207738, + "learning_rate": 4.815906549131027e-07, + "loss": 0.2406, + "step": 39223 + }, + { + "epoch": 0.8633942328857583, + "grad_norm": 1.6732093846593925, + "learning_rate": 4.814380301757493e-07, + "loss": 0.3084, + "step": 39224 + }, + { + "epoch": 0.8634162447721769, + "grad_norm": 1.5887873254304636, + "learning_rate": 4.812854284037443e-07, + "loss": 0.2593, + "step": 39225 + }, + { + "epoch": 0.8634382566585956, + "grad_norm": 1.9243138302209746, + "learning_rate": 4.811328495978618e-07, + "loss": 0.2847, + "step": 39226 + }, + { + "epoch": 0.8634602685450143, + "grad_norm": 0.6074868462653825, + "learning_rate": 4.809802937588771e-07, + "loss": 0.4621, + "step": 39227 + }, + { + "epoch": 0.8634822804314329, + "grad_norm": 1.793703326680053, + "learning_rate": 4.808277608875661e-07, + "loss": 0.2613, + "step": 39228 + }, + { + "epoch": 0.8635042923178516, + "grad_norm": 0.63675107589444, + "learning_rate": 4.806752509847035e-07, + "loss": 0.4459, + "step": 39229 + }, + { + "epoch": 0.8635263042042703, + "grad_norm": 3.669829563905401, + "learning_rate": 4.805227640510662e-07, + "loss": 0.282, + "step": 39230 + }, + { + "epoch": 0.8635483160906889, + "grad_norm": 1.9780397616940613, + "learning_rate": 4.803703000874282e-07, + "loss": 0.3876, + "step": 39231 + }, + { + "epoch": 0.8635703279771076, + "grad_norm": 1.2725893654417155, + "learning_rate": 4.802178590945627e-07, + "loss": 0.1991, + "step": 39232 + }, + { + "epoch": 0.8635923398635263, + "grad_norm": 3.3144166351001294, + "learning_rate": 4.800654410732464e-07, + "loss": 0.2821, + "step": 39233 + }, + { + "epoch": 0.8636143517499449, + "grad_norm": 1.6836542701559882, + "learning_rate": 4.799130460242535e-07, + "loss": 0.2777, + "step": 39234 + }, + { + "epoch": 0.8636363636363636, + "grad_norm": 1.7321774842993858, + "learning_rate": 4.797606739483591e-07, + "loss": 0.3125, + "step": 39235 + }, + { + "epoch": 0.8636583755227824, + "grad_norm": 1.8211432878164249, + "learning_rate": 4.796083248463374e-07, + "loss": 0.208, + "step": 39236 + }, + { + "epoch": 0.863680387409201, + "grad_norm": 1.9672958159972522, + "learning_rate": 4.794559987189607e-07, + "loss": 0.2637, + "step": 39237 + }, + { + "epoch": 0.8637023992956196, + "grad_norm": 0.5856676510529787, + "learning_rate": 4.793036955670055e-07, + "loss": 0.4335, + "step": 39238 + }, + { + "epoch": 0.8637244111820384, + "grad_norm": 1.313131404779192, + "learning_rate": 4.791514153912452e-07, + "loss": 0.2632, + "step": 39239 + }, + { + "epoch": 0.863746423068457, + "grad_norm": 1.9146705098602483, + "learning_rate": 4.789991581924541e-07, + "loss": 0.3334, + "step": 39240 + }, + { + "epoch": 0.8637684349548757, + "grad_norm": 1.3972295333489009, + "learning_rate": 4.788469239714061e-07, + "loss": 0.235, + "step": 39241 + }, + { + "epoch": 0.8637904468412942, + "grad_norm": 1.2770067383249253, + "learning_rate": 4.786947127288738e-07, + "loss": 0.233, + "step": 39242 + }, + { + "epoch": 0.863812458727713, + "grad_norm": 0.6004980566983381, + "learning_rate": 4.785425244656311e-07, + "loss": 0.4397, + "step": 39243 + }, + { + "epoch": 0.8638344706141317, + "grad_norm": 0.6144282177808693, + "learning_rate": 4.783903591824535e-07, + "loss": 0.4304, + "step": 39244 + }, + { + "epoch": 0.8638564825005502, + "grad_norm": 0.6038271006764606, + "learning_rate": 4.782382168801114e-07, + "loss": 0.4483, + "step": 39245 + }, + { + "epoch": 0.863878494386969, + "grad_norm": 3.034898396494793, + "learning_rate": 4.780860975593804e-07, + "loss": 0.3239, + "step": 39246 + }, + { + "epoch": 0.8639005062733877, + "grad_norm": 1.486413078298279, + "learning_rate": 4.779340012210316e-07, + "loss": 0.2768, + "step": 39247 + }, + { + "epoch": 0.8639225181598063, + "grad_norm": 0.602240833522145, + "learning_rate": 4.777819278658391e-07, + "loss": 0.4537, + "step": 39248 + }, + { + "epoch": 0.863944530046225, + "grad_norm": 2.579200115670045, + "learning_rate": 4.77629877494577e-07, + "loss": 0.2707, + "step": 39249 + }, + { + "epoch": 0.8639665419326437, + "grad_norm": 1.9267024457283, + "learning_rate": 4.774778501080158e-07, + "loss": 0.3249, + "step": 39250 + }, + { + "epoch": 0.8639885538190623, + "grad_norm": 0.6314492709796911, + "learning_rate": 4.773258457069302e-07, + "loss": 0.4789, + "step": 39251 + }, + { + "epoch": 0.864010565705481, + "grad_norm": 2.0518218152268157, + "learning_rate": 4.771738642920915e-07, + "loss": 0.3177, + "step": 39252 + }, + { + "epoch": 0.8640325775918997, + "grad_norm": 1.3462536351503442, + "learning_rate": 4.770219058642717e-07, + "loss": 0.247, + "step": 39253 + }, + { + "epoch": 0.8640545894783183, + "grad_norm": 0.6004954340348931, + "learning_rate": 4.768699704242452e-07, + "loss": 0.4542, + "step": 39254 + }, + { + "epoch": 0.864076601364737, + "grad_norm": 1.618595710185061, + "learning_rate": 4.767180579727815e-07, + "loss": 0.3119, + "step": 39255 + }, + { + "epoch": 0.8640986132511557, + "grad_norm": 0.6063498989680288, + "learning_rate": 4.765661685106554e-07, + "loss": 0.457, + "step": 39256 + }, + { + "epoch": 0.8641206251375743, + "grad_norm": 1.374090868572185, + "learning_rate": 4.7641430203863635e-07, + "loss": 0.2846, + "step": 39257 + }, + { + "epoch": 0.864142637023993, + "grad_norm": 4.199316381295486, + "learning_rate": 4.7626245855749874e-07, + "loss": 0.3292, + "step": 39258 + }, + { + "epoch": 0.8641646489104117, + "grad_norm": 2.2354537904668437, + "learning_rate": 4.761106380680114e-07, + "loss": 0.2755, + "step": 39259 + }, + { + "epoch": 0.8641866607968303, + "grad_norm": 1.9360780889075602, + "learning_rate": 4.759588405709481e-07, + "loss": 0.2442, + "step": 39260 + }, + { + "epoch": 0.864208672683249, + "grad_norm": 1.672858094665197, + "learning_rate": 4.758070660670805e-07, + "loss": 0.2366, + "step": 39261 + }, + { + "epoch": 0.8642306845696677, + "grad_norm": 3.0362639993133493, + "learning_rate": 4.756553145571791e-07, + "loss": 0.248, + "step": 39262 + }, + { + "epoch": 0.8642526964560863, + "grad_norm": 1.4777646424722755, + "learning_rate": 4.755035860420143e-07, + "loss": 0.1764, + "step": 39263 + }, + { + "epoch": 0.864274708342505, + "grad_norm": 2.3251242612615153, + "learning_rate": 4.7535188052235836e-07, + "loss": 0.2981, + "step": 39264 + }, + { + "epoch": 0.8642967202289236, + "grad_norm": 0.6155596579262104, + "learning_rate": 4.7520019799898234e-07, + "loss": 0.4467, + "step": 39265 + }, + { + "epoch": 0.8643187321153423, + "grad_norm": 0.59810181626247, + "learning_rate": 4.750485384726577e-07, + "loss": 0.4323, + "step": 39266 + }, + { + "epoch": 0.864340744001761, + "grad_norm": 1.4975731069850777, + "learning_rate": 4.7489690194415515e-07, + "loss": 0.2553, + "step": 39267 + }, + { + "epoch": 0.8643627558881796, + "grad_norm": 0.6221762030859423, + "learning_rate": 4.747452884142434e-07, + "loss": 0.4539, + "step": 39268 + }, + { + "epoch": 0.8643847677745983, + "grad_norm": 1.6941538862984642, + "learning_rate": 4.745936978836946e-07, + "loss": 0.2906, + "step": 39269 + }, + { + "epoch": 0.864406779661017, + "grad_norm": 1.8347575263532228, + "learning_rate": 4.744421303532787e-07, + "loss": 0.2151, + "step": 39270 + }, + { + "epoch": 0.8644287915474356, + "grad_norm": 5.5878499333053515, + "learning_rate": 4.742905858237673e-07, + "loss": 0.2657, + "step": 39271 + }, + { + "epoch": 0.8644508034338543, + "grad_norm": 1.5179199037106879, + "learning_rate": 4.7413906429592984e-07, + "loss": 0.2476, + "step": 39272 + }, + { + "epoch": 0.864472815320273, + "grad_norm": 3.1558963965489233, + "learning_rate": 4.7398756577053563e-07, + "loss": 0.2642, + "step": 39273 + }, + { + "epoch": 0.8644948272066916, + "grad_norm": 1.5715442839914815, + "learning_rate": 4.7383609024835463e-07, + "loss": 0.2028, + "step": 39274 + }, + { + "epoch": 0.8645168390931103, + "grad_norm": 2.3204889312145975, + "learning_rate": 4.7368463773015794e-07, + "loss": 0.2796, + "step": 39275 + }, + { + "epoch": 0.864538850979529, + "grad_norm": 1.4137242803928174, + "learning_rate": 4.7353320821671547e-07, + "loss": 0.2511, + "step": 39276 + }, + { + "epoch": 0.8645608628659476, + "grad_norm": 1.5581329369896437, + "learning_rate": 4.73381801708796e-07, + "loss": 0.3787, + "step": 39277 + }, + { + "epoch": 0.8645828747523663, + "grad_norm": 1.8241148424647502, + "learning_rate": 4.7323041820716784e-07, + "loss": 0.2508, + "step": 39278 + }, + { + "epoch": 0.864604886638785, + "grad_norm": 1.5130000374918828, + "learning_rate": 4.730790577126021e-07, + "loss": 0.3186, + "step": 39279 + }, + { + "epoch": 0.8646268985252036, + "grad_norm": 1.5727068212537791, + "learning_rate": 4.729277202258675e-07, + "loss": 0.3069, + "step": 39280 + }, + { + "epoch": 0.8646489104116223, + "grad_norm": 1.3019756380303498, + "learning_rate": 4.7277640574773465e-07, + "loss": 0.2075, + "step": 39281 + }, + { + "epoch": 0.864670922298041, + "grad_norm": 1.4899084489770515, + "learning_rate": 4.726251142789712e-07, + "loss": 0.2896, + "step": 39282 + }, + { + "epoch": 0.8646929341844596, + "grad_norm": 1.3621393010906584, + "learning_rate": 4.724738458203448e-07, + "loss": 0.204, + "step": 39283 + }, + { + "epoch": 0.8647149460708783, + "grad_norm": 2.5927114997628338, + "learning_rate": 4.723226003726261e-07, + "loss": 0.3089, + "step": 39284 + }, + { + "epoch": 0.864736957957297, + "grad_norm": 1.7311814993176995, + "learning_rate": 4.721713779365844e-07, + "loss": 0.2786, + "step": 39285 + }, + { + "epoch": 0.8647589698437156, + "grad_norm": 2.8352725066900297, + "learning_rate": 4.720201785129858e-07, + "loss": 0.3524, + "step": 39286 + }, + { + "epoch": 0.8647809817301343, + "grad_norm": 2.528181816824152, + "learning_rate": 4.7186900210260124e-07, + "loss": 0.2597, + "step": 39287 + }, + { + "epoch": 0.8648029936165529, + "grad_norm": 2.324899830592366, + "learning_rate": 4.7171784870619743e-07, + "loss": 0.2467, + "step": 39288 + }, + { + "epoch": 0.8648250055029716, + "grad_norm": 1.4811821626268125, + "learning_rate": 4.715667183245426e-07, + "loss": 0.3768, + "step": 39289 + }, + { + "epoch": 0.8648470173893903, + "grad_norm": 1.8183737357120169, + "learning_rate": 4.714156109584067e-07, + "loss": 0.3351, + "step": 39290 + }, + { + "epoch": 0.8648690292758089, + "grad_norm": 1.9068870281663972, + "learning_rate": 4.712645266085558e-07, + "loss": 0.3555, + "step": 39291 + }, + { + "epoch": 0.8648910411622276, + "grad_norm": 0.6349928326256127, + "learning_rate": 4.711134652757593e-07, + "loss": 0.4517, + "step": 39292 + }, + { + "epoch": 0.8649130530486463, + "grad_norm": 1.5065679710038755, + "learning_rate": 4.7096242696078263e-07, + "loss": 0.2984, + "step": 39293 + }, + { + "epoch": 0.8649350649350649, + "grad_norm": 1.5521151648272191, + "learning_rate": 4.708114116643964e-07, + "loss": 0.3074, + "step": 39294 + }, + { + "epoch": 0.8649570768214836, + "grad_norm": 1.694236342387087, + "learning_rate": 4.706604193873654e-07, + "loss": 0.2809, + "step": 39295 + }, + { + "epoch": 0.8649790887079023, + "grad_norm": 0.6354954950005411, + "learning_rate": 4.7050945013045854e-07, + "loss": 0.4484, + "step": 39296 + }, + { + "epoch": 0.8650011005943209, + "grad_norm": 1.929243660476792, + "learning_rate": 4.703585038944436e-07, + "loss": 0.2793, + "step": 39297 + }, + { + "epoch": 0.8650231124807396, + "grad_norm": 1.501064535805789, + "learning_rate": 4.7020758068008655e-07, + "loss": 0.3071, + "step": 39298 + }, + { + "epoch": 0.8650451243671583, + "grad_norm": 2.1531827764395985, + "learning_rate": 4.7005668048815514e-07, + "loss": 0.2444, + "step": 39299 + }, + { + "epoch": 0.8650671362535769, + "grad_norm": 1.8500512259251596, + "learning_rate": 4.6990580331941594e-07, + "loss": 0.3099, + "step": 39300 + }, + { + "epoch": 0.8650891481399956, + "grad_norm": 1.6530182243834552, + "learning_rate": 4.6975494917463505e-07, + "loss": 0.2991, + "step": 39301 + }, + { + "epoch": 0.8651111600264143, + "grad_norm": 1.9657315982409744, + "learning_rate": 4.6960411805458183e-07, + "loss": 0.2715, + "step": 39302 + }, + { + "epoch": 0.8651331719128329, + "grad_norm": 1.3543193881736375, + "learning_rate": 4.694533099600207e-07, + "loss": 0.2384, + "step": 39303 + }, + { + "epoch": 0.8651551837992516, + "grad_norm": 1.7629624196395661, + "learning_rate": 4.693025248917177e-07, + "loss": 0.3158, + "step": 39304 + }, + { + "epoch": 0.8651771956856703, + "grad_norm": 1.2794014016714608, + "learning_rate": 4.6915176285043937e-07, + "loss": 0.2633, + "step": 39305 + }, + { + "epoch": 0.8651992075720889, + "grad_norm": 1.250371473589773, + "learning_rate": 4.6900102383695356e-07, + "loss": 0.2527, + "step": 39306 + }, + { + "epoch": 0.8652212194585076, + "grad_norm": 1.7939036469129173, + "learning_rate": 4.688503078520257e-07, + "loss": 0.2578, + "step": 39307 + }, + { + "epoch": 0.8652432313449263, + "grad_norm": 1.4355327243476812, + "learning_rate": 4.686996148964218e-07, + "loss": 0.3207, + "step": 39308 + }, + { + "epoch": 0.8652652432313449, + "grad_norm": 1.4476199758564285, + "learning_rate": 4.685489449709063e-07, + "loss": 0.3321, + "step": 39309 + }, + { + "epoch": 0.8652872551177636, + "grad_norm": 1.3012452642284316, + "learning_rate": 4.6839829807624636e-07, + "loss": 0.2866, + "step": 39310 + }, + { + "epoch": 0.8653092670041822, + "grad_norm": 2.340424736861906, + "learning_rate": 4.682476742132075e-07, + "loss": 0.3282, + "step": 39311 + }, + { + "epoch": 0.8653312788906009, + "grad_norm": 3.369417355059053, + "learning_rate": 4.680970733825563e-07, + "loss": 0.2756, + "step": 39312 + }, + { + "epoch": 0.8653532907770196, + "grad_norm": 2.136324156120601, + "learning_rate": 4.6794649558505656e-07, + "loss": 0.2905, + "step": 39313 + }, + { + "epoch": 0.8653753026634382, + "grad_norm": 2.231700914967563, + "learning_rate": 4.677959408214733e-07, + "loss": 0.2315, + "step": 39314 + }, + { + "epoch": 0.8653973145498569, + "grad_norm": 1.8513453659566157, + "learning_rate": 4.676454090925725e-07, + "loss": 0.418, + "step": 39315 + }, + { + "epoch": 0.8654193264362756, + "grad_norm": 1.7875066510106399, + "learning_rate": 4.6749490039911917e-07, + "loss": 0.2855, + "step": 39316 + }, + { + "epoch": 0.8654413383226942, + "grad_norm": 1.6928357986938347, + "learning_rate": 4.673444147418793e-07, + "loss": 0.2669, + "step": 39317 + }, + { + "epoch": 0.8654633502091129, + "grad_norm": 1.6343460574103212, + "learning_rate": 4.6719395212161734e-07, + "loss": 0.255, + "step": 39318 + }, + { + "epoch": 0.8654853620955316, + "grad_norm": 1.5808708821183488, + "learning_rate": 4.670435125390954e-07, + "loss": 0.272, + "step": 39319 + }, + { + "epoch": 0.8655073739819502, + "grad_norm": 5.361175371941015, + "learning_rate": 4.668930959950807e-07, + "loss": 0.3409, + "step": 39320 + }, + { + "epoch": 0.8655293858683689, + "grad_norm": 1.6895975749319887, + "learning_rate": 4.6674270249033707e-07, + "loss": 0.2777, + "step": 39321 + }, + { + "epoch": 0.8655513977547876, + "grad_norm": 1.7025089764485668, + "learning_rate": 4.6659233202562993e-07, + "loss": 0.2693, + "step": 39322 + }, + { + "epoch": 0.8655734096412062, + "grad_norm": 0.6056987894669557, + "learning_rate": 4.664419846017221e-07, + "loss": 0.4267, + "step": 39323 + }, + { + "epoch": 0.8655954215276249, + "grad_norm": 1.342574999517821, + "learning_rate": 4.662916602193779e-07, + "loss": 0.2461, + "step": 39324 + }, + { + "epoch": 0.8656174334140436, + "grad_norm": 2.141926840270867, + "learning_rate": 4.661413588793606e-07, + "loss": 0.3631, + "step": 39325 + }, + { + "epoch": 0.8656394453004622, + "grad_norm": 1.439586155665397, + "learning_rate": 4.659910805824369e-07, + "loss": 0.2949, + "step": 39326 + }, + { + "epoch": 0.8656614571868809, + "grad_norm": 1.6768629006339373, + "learning_rate": 4.6584082532936725e-07, + "loss": 0.3246, + "step": 39327 + }, + { + "epoch": 0.8656834690732996, + "grad_norm": 1.6974415645600438, + "learning_rate": 4.6569059312091826e-07, + "loss": 0.3048, + "step": 39328 + }, + { + "epoch": 0.8657054809597182, + "grad_norm": 1.7300613048175706, + "learning_rate": 4.655403839578504e-07, + "loss": 0.2635, + "step": 39329 + }, + { + "epoch": 0.8657274928461369, + "grad_norm": 2.0681296275090983, + "learning_rate": 4.6539019784093033e-07, + "loss": 0.208, + "step": 39330 + }, + { + "epoch": 0.8657495047325556, + "grad_norm": 6.169445414665986, + "learning_rate": 4.652400347709185e-07, + "loss": 0.3693, + "step": 39331 + }, + { + "epoch": 0.8657715166189742, + "grad_norm": 3.0201826281266446, + "learning_rate": 4.6508989474857937e-07, + "loss": 0.3035, + "step": 39332 + }, + { + "epoch": 0.8657935285053929, + "grad_norm": 17.72034035507218, + "learning_rate": 4.6493977777467724e-07, + "loss": 0.3043, + "step": 39333 + }, + { + "epoch": 0.8658155403918116, + "grad_norm": 2.1431054577169664, + "learning_rate": 4.647896838499727e-07, + "loss": 0.3158, + "step": 39334 + }, + { + "epoch": 0.8658375522782302, + "grad_norm": 1.7631798778722414, + "learning_rate": 4.646396129752306e-07, + "loss": 0.2112, + "step": 39335 + }, + { + "epoch": 0.8658595641646489, + "grad_norm": 1.6909463401325213, + "learning_rate": 4.644895651512121e-07, + "loss": 0.2436, + "step": 39336 + }, + { + "epoch": 0.8658815760510675, + "grad_norm": 3.9454307167670333, + "learning_rate": 4.643395403786799e-07, + "loss": 0.3125, + "step": 39337 + }, + { + "epoch": 0.8659035879374862, + "grad_norm": 1.952219794618107, + "learning_rate": 4.6418953865839886e-07, + "loss": 0.3125, + "step": 39338 + }, + { + "epoch": 0.8659255998239049, + "grad_norm": 1.6615802877075427, + "learning_rate": 4.640395599911285e-07, + "loss": 0.3083, + "step": 39339 + }, + { + "epoch": 0.8659476117103235, + "grad_norm": 1.3667585400276858, + "learning_rate": 4.6388960437763254e-07, + "loss": 0.2146, + "step": 39340 + }, + { + "epoch": 0.8659696235967422, + "grad_norm": 1.713145687678156, + "learning_rate": 4.637396718186726e-07, + "loss": 0.27, + "step": 39341 + }, + { + "epoch": 0.8659916354831609, + "grad_norm": 2.20940109860319, + "learning_rate": 4.635897623150104e-07, + "loss": 0.2872, + "step": 39342 + }, + { + "epoch": 0.8660136473695795, + "grad_norm": 1.7882398346732458, + "learning_rate": 4.634398758674097e-07, + "loss": 0.3133, + "step": 39343 + }, + { + "epoch": 0.8660356592559982, + "grad_norm": 1.6480922445727222, + "learning_rate": 4.632900124766304e-07, + "loss": 0.3259, + "step": 39344 + }, + { + "epoch": 0.8660576711424169, + "grad_norm": 1.6495254964933967, + "learning_rate": 4.631401721434342e-07, + "loss": 0.2643, + "step": 39345 + }, + { + "epoch": 0.8660796830288355, + "grad_norm": 1.5960733598545631, + "learning_rate": 4.629903548685827e-07, + "loss": 0.2746, + "step": 39346 + }, + { + "epoch": 0.8661016949152542, + "grad_norm": 1.909420456685646, + "learning_rate": 4.62840560652838e-07, + "loss": 0.2633, + "step": 39347 + }, + { + "epoch": 0.8661237068016729, + "grad_norm": 3.2326927947180306, + "learning_rate": 4.6269078949696236e-07, + "loss": 0.2806, + "step": 39348 + }, + { + "epoch": 0.8661457186880915, + "grad_norm": 0.656187718829018, + "learning_rate": 4.625410414017156e-07, + "loss": 0.5001, + "step": 39349 + }, + { + "epoch": 0.8661677305745102, + "grad_norm": 4.430009710120009, + "learning_rate": 4.6239131636785784e-07, + "loss": 0.1949, + "step": 39350 + }, + { + "epoch": 0.866189742460929, + "grad_norm": 2.272847655028245, + "learning_rate": 4.6224161439615166e-07, + "loss": 0.367, + "step": 39351 + }, + { + "epoch": 0.8662117543473475, + "grad_norm": 0.6180143061845523, + "learning_rate": 4.620919354873571e-07, + "loss": 0.4338, + "step": 39352 + }, + { + "epoch": 0.8662337662337662, + "grad_norm": 0.6397954490511532, + "learning_rate": 4.619422796422368e-07, + "loss": 0.4609, + "step": 39353 + }, + { + "epoch": 0.866255778120185, + "grad_norm": 1.7410924184595193, + "learning_rate": 4.617926468615491e-07, + "loss": 0.2744, + "step": 39354 + }, + { + "epoch": 0.8662777900066035, + "grad_norm": 0.6332767126115505, + "learning_rate": 4.6164303714605506e-07, + "loss": 0.4491, + "step": 39355 + }, + { + "epoch": 0.8662998018930222, + "grad_norm": 1.6220456409800232, + "learning_rate": 4.614934504965146e-07, + "loss": 0.2362, + "step": 39356 + }, + { + "epoch": 0.866321813779441, + "grad_norm": 0.6063027223678161, + "learning_rate": 4.6134388691368935e-07, + "loss": 0.4281, + "step": 39357 + }, + { + "epoch": 0.8663438256658595, + "grad_norm": 1.77683511028507, + "learning_rate": 4.6119434639833927e-07, + "loss": 0.2896, + "step": 39358 + }, + { + "epoch": 0.8663658375522783, + "grad_norm": 1.8731690339437423, + "learning_rate": 4.610448289512237e-07, + "loss": 0.3441, + "step": 39359 + }, + { + "epoch": 0.8663878494386968, + "grad_norm": 1.7475742950119635, + "learning_rate": 4.608953345731021e-07, + "loss": 0.2983, + "step": 39360 + }, + { + "epoch": 0.8664098613251155, + "grad_norm": 1.3115311368049554, + "learning_rate": 4.6074586326473604e-07, + "loss": 0.3046, + "step": 39361 + }, + { + "epoch": 0.8664318732115343, + "grad_norm": 3.1534965380141147, + "learning_rate": 4.6059641502688266e-07, + "loss": 0.3699, + "step": 39362 + }, + { + "epoch": 0.8664538850979528, + "grad_norm": 0.6067249477486235, + "learning_rate": 4.604469898603037e-07, + "loss": 0.44, + "step": 39363 + }, + { + "epoch": 0.8664758969843716, + "grad_norm": 1.73060564299193, + "learning_rate": 4.6029758776575785e-07, + "loss": 0.2916, + "step": 39364 + }, + { + "epoch": 0.8664979088707903, + "grad_norm": 1.6663999534322504, + "learning_rate": 4.60148208744004e-07, + "loss": 0.3137, + "step": 39365 + }, + { + "epoch": 0.8665199207572089, + "grad_norm": 1.709463339705338, + "learning_rate": 4.599988527958027e-07, + "loss": 0.3045, + "step": 39366 + }, + { + "epoch": 0.8665419326436276, + "grad_norm": 1.9291594144608866, + "learning_rate": 4.5984951992191106e-07, + "loss": 0.2253, + "step": 39367 + }, + { + "epoch": 0.8665639445300463, + "grad_norm": 0.6466805334601744, + "learning_rate": 4.5970021012308907e-07, + "loss": 0.4641, + "step": 39368 + }, + { + "epoch": 0.8665859564164649, + "grad_norm": 1.8128877888106847, + "learning_rate": 4.5955092340009667e-07, + "loss": 0.3336, + "step": 39369 + }, + { + "epoch": 0.8666079683028836, + "grad_norm": 2.892842732552502, + "learning_rate": 4.59401659753691e-07, + "loss": 0.2689, + "step": 39370 + }, + { + "epoch": 0.8666299801893023, + "grad_norm": 1.4248531621143325, + "learning_rate": 4.592524191846315e-07, + "loss": 0.2975, + "step": 39371 + }, + { + "epoch": 0.8666519920757209, + "grad_norm": 1.6133583079404383, + "learning_rate": 4.591032016936758e-07, + "loss": 0.226, + "step": 39372 + }, + { + "epoch": 0.8666740039621396, + "grad_norm": 1.572929494130145, + "learning_rate": 4.589540072815829e-07, + "loss": 0.2603, + "step": 39373 + }, + { + "epoch": 0.8666960158485583, + "grad_norm": 1.6533943285404873, + "learning_rate": 4.588048359491121e-07, + "loss": 0.326, + "step": 39374 + }, + { + "epoch": 0.8667180277349769, + "grad_norm": 0.6126822225367323, + "learning_rate": 4.5865568769701995e-07, + "loss": 0.4648, + "step": 39375 + }, + { + "epoch": 0.8667400396213956, + "grad_norm": 1.4954330520887675, + "learning_rate": 4.585065625260654e-07, + "loss": 0.3108, + "step": 39376 + }, + { + "epoch": 0.8667620515078143, + "grad_norm": 0.584038306959351, + "learning_rate": 4.5835746043700556e-07, + "loss": 0.4148, + "step": 39377 + }, + { + "epoch": 0.8667840633942329, + "grad_norm": 4.217807774076647, + "learning_rate": 4.5820838143059876e-07, + "loss": 0.2878, + "step": 39378 + }, + { + "epoch": 0.8668060752806516, + "grad_norm": 3.7107420357710743, + "learning_rate": 4.5805932550760323e-07, + "loss": 0.2948, + "step": 39379 + }, + { + "epoch": 0.8668280871670703, + "grad_norm": 5.161962504365243, + "learning_rate": 4.579102926687751e-07, + "loss": 0.3439, + "step": 39380 + }, + { + "epoch": 0.8668500990534889, + "grad_norm": 2.3318303751365432, + "learning_rate": 4.5776128291487365e-07, + "loss": 0.3423, + "step": 39381 + }, + { + "epoch": 0.8668721109399076, + "grad_norm": 1.544091554907107, + "learning_rate": 4.576122962466545e-07, + "loss": 0.2991, + "step": 39382 + }, + { + "epoch": 0.8668941228263262, + "grad_norm": 0.6276956258285056, + "learning_rate": 4.5746333266487585e-07, + "loss": 0.4662, + "step": 39383 + }, + { + "epoch": 0.8669161347127449, + "grad_norm": 1.7102238100966785, + "learning_rate": 4.573143921702949e-07, + "loss": 0.2175, + "step": 39384 + }, + { + "epoch": 0.8669381465991636, + "grad_norm": 2.0193456878482867, + "learning_rate": 4.571654747636689e-07, + "loss": 0.4302, + "step": 39385 + }, + { + "epoch": 0.8669601584855822, + "grad_norm": 0.6222746381039934, + "learning_rate": 4.570165804457527e-07, + "loss": 0.4676, + "step": 39386 + }, + { + "epoch": 0.8669821703720009, + "grad_norm": 0.601604340323627, + "learning_rate": 4.568677092173046e-07, + "loss": 0.4299, + "step": 39387 + }, + { + "epoch": 0.8670041822584196, + "grad_norm": 1.3277155611468014, + "learning_rate": 4.5671886107908125e-07, + "loss": 0.2377, + "step": 39388 + }, + { + "epoch": 0.8670261941448382, + "grad_norm": 2.018234759316889, + "learning_rate": 4.565700360318398e-07, + "loss": 0.2696, + "step": 39389 + }, + { + "epoch": 0.8670482060312569, + "grad_norm": 2.235248911606761, + "learning_rate": 4.5642123407633576e-07, + "loss": 0.2894, + "step": 39390 + }, + { + "epoch": 0.8670702179176756, + "grad_norm": 0.6200787462295515, + "learning_rate": 4.562724552133252e-07, + "loss": 0.4419, + "step": 39391 + }, + { + "epoch": 0.8670922298040942, + "grad_norm": 1.4593706245512643, + "learning_rate": 4.561236994435636e-07, + "loss": 0.263, + "step": 39392 + }, + { + "epoch": 0.8671142416905129, + "grad_norm": 1.791331719289245, + "learning_rate": 4.559749667678087e-07, + "loss": 0.2934, + "step": 39393 + }, + { + "epoch": 0.8671362535769316, + "grad_norm": 2.476076702562847, + "learning_rate": 4.558262571868166e-07, + "loss": 0.2459, + "step": 39394 + }, + { + "epoch": 0.8671582654633502, + "grad_norm": 1.75887711096455, + "learning_rate": 4.556775707013417e-07, + "loss": 0.3846, + "step": 39395 + }, + { + "epoch": 0.8671802773497689, + "grad_norm": 0.6052868475271795, + "learning_rate": 4.5552890731213996e-07, + "loss": 0.4267, + "step": 39396 + }, + { + "epoch": 0.8672022892361876, + "grad_norm": 1.4824583295297113, + "learning_rate": 4.5538026701996754e-07, + "loss": 0.2753, + "step": 39397 + }, + { + "epoch": 0.8672243011226062, + "grad_norm": 2.095260345389467, + "learning_rate": 4.552316498255782e-07, + "loss": 0.3264, + "step": 39398 + }, + { + "epoch": 0.8672463130090249, + "grad_norm": 0.6107311211334839, + "learning_rate": 4.550830557297303e-07, + "loss": 0.4407, + "step": 39399 + }, + { + "epoch": 0.8672683248954436, + "grad_norm": 0.6278126630087721, + "learning_rate": 4.5493448473317713e-07, + "loss": 0.4589, + "step": 39400 + }, + { + "epoch": 0.8672903367818622, + "grad_norm": 0.6152763451780456, + "learning_rate": 4.547859368366736e-07, + "loss": 0.4536, + "step": 39401 + }, + { + "epoch": 0.8673123486682809, + "grad_norm": 1.7042097492594837, + "learning_rate": 4.546374120409758e-07, + "loss": 0.2103, + "step": 39402 + }, + { + "epoch": 0.8673343605546996, + "grad_norm": 1.3119058934261352, + "learning_rate": 4.5448891034683697e-07, + "loss": 0.2477, + "step": 39403 + }, + { + "epoch": 0.8673563724411182, + "grad_norm": 2.2994271772888553, + "learning_rate": 4.543404317550126e-07, + "loss": 0.3131, + "step": 39404 + }, + { + "epoch": 0.8673783843275369, + "grad_norm": 0.6279526500943897, + "learning_rate": 4.5419197626625887e-07, + "loss": 0.4471, + "step": 39405 + }, + { + "epoch": 0.8674003962139555, + "grad_norm": 1.577412051650003, + "learning_rate": 4.540435438813284e-07, + "loss": 0.2966, + "step": 39406 + }, + { + "epoch": 0.8674224081003742, + "grad_norm": 1.6382993535333843, + "learning_rate": 4.538951346009768e-07, + "loss": 0.2776, + "step": 39407 + }, + { + "epoch": 0.8674444199867929, + "grad_norm": 0.6372182595147116, + "learning_rate": 4.537467484259567e-07, + "loss": 0.4764, + "step": 39408 + }, + { + "epoch": 0.8674664318732115, + "grad_norm": 1.594421607249192, + "learning_rate": 4.535983853570236e-07, + "loss": 0.2886, + "step": 39409 + }, + { + "epoch": 0.8674884437596302, + "grad_norm": 1.9947925852448514, + "learning_rate": 4.53450045394932e-07, + "loss": 0.3098, + "step": 39410 + }, + { + "epoch": 0.8675104556460489, + "grad_norm": 0.602263586851274, + "learning_rate": 4.5330172854043395e-07, + "loss": 0.4389, + "step": 39411 + }, + { + "epoch": 0.8675324675324675, + "grad_norm": 0.6270589065130331, + "learning_rate": 4.5315343479428563e-07, + "loss": 0.4366, + "step": 39412 + }, + { + "epoch": 0.8675544794188862, + "grad_norm": 1.4091605203681292, + "learning_rate": 4.5300516415723863e-07, + "loss": 0.2192, + "step": 39413 + }, + { + "epoch": 0.8675764913053049, + "grad_norm": 2.234859163831991, + "learning_rate": 4.528569166300473e-07, + "loss": 0.3225, + "step": 39414 + }, + { + "epoch": 0.8675985031917235, + "grad_norm": 1.5020867182401176, + "learning_rate": 4.52708692213466e-07, + "loss": 0.311, + "step": 39415 + }, + { + "epoch": 0.8676205150781422, + "grad_norm": 2.2993412311848602, + "learning_rate": 4.525604909082465e-07, + "loss": 0.3747, + "step": 39416 + }, + { + "epoch": 0.8676425269645609, + "grad_norm": 2.1167415656676702, + "learning_rate": 4.5241231271514365e-07, + "loss": 0.2251, + "step": 39417 + }, + { + "epoch": 0.8676645388509795, + "grad_norm": 0.6151102775126466, + "learning_rate": 4.522641576349085e-07, + "loss": 0.4319, + "step": 39418 + }, + { + "epoch": 0.8676865507373982, + "grad_norm": 1.4939345699790345, + "learning_rate": 4.521160256682955e-07, + "loss": 0.2703, + "step": 39419 + }, + { + "epoch": 0.8677085626238169, + "grad_norm": 2.86416431780992, + "learning_rate": 4.5196791681605844e-07, + "loss": 0.3313, + "step": 39420 + }, + { + "epoch": 0.8677305745102355, + "grad_norm": 2.4217115950439223, + "learning_rate": 4.5181983107894787e-07, + "loss": 0.2761, + "step": 39421 + }, + { + "epoch": 0.8677525863966542, + "grad_norm": 1.5471399242522061, + "learning_rate": 4.516717684577182e-07, + "loss": 0.288, + "step": 39422 + }, + { + "epoch": 0.8677745982830729, + "grad_norm": 1.5100044167405635, + "learning_rate": 4.5152372895312047e-07, + "loss": 0.2915, + "step": 39423 + }, + { + "epoch": 0.8677966101694915, + "grad_norm": 1.7833060239932843, + "learning_rate": 4.513757125659074e-07, + "loss": 0.2728, + "step": 39424 + }, + { + "epoch": 0.8678186220559102, + "grad_norm": 4.950540401820926, + "learning_rate": 4.512277192968334e-07, + "loss": 0.3253, + "step": 39425 + }, + { + "epoch": 0.8678406339423289, + "grad_norm": 1.3885539064346757, + "learning_rate": 4.5107974914664843e-07, + "loss": 0.2012, + "step": 39426 + }, + { + "epoch": 0.8678626458287475, + "grad_norm": 1.6977748613790602, + "learning_rate": 4.509318021161041e-07, + "loss": 0.2417, + "step": 39427 + }, + { + "epoch": 0.8678846577151662, + "grad_norm": 1.887751246661546, + "learning_rate": 4.5078387820595315e-07, + "loss": 0.328, + "step": 39428 + }, + { + "epoch": 0.8679066696015848, + "grad_norm": 0.5854991572331075, + "learning_rate": 4.5063597741694777e-07, + "loss": 0.403, + "step": 39429 + }, + { + "epoch": 0.8679286814880035, + "grad_norm": 1.9923742446032553, + "learning_rate": 4.504880997498406e-07, + "loss": 0.2997, + "step": 39430 + }, + { + "epoch": 0.8679506933744222, + "grad_norm": 1.7128983209303918, + "learning_rate": 4.5034024520538176e-07, + "loss": 0.2336, + "step": 39431 + }, + { + "epoch": 0.8679727052608408, + "grad_norm": 2.326237929994646, + "learning_rate": 4.5019241378432223e-07, + "loss": 0.2758, + "step": 39432 + }, + { + "epoch": 0.8679947171472595, + "grad_norm": 1.429748889950838, + "learning_rate": 4.500446054874147e-07, + "loss": 0.2619, + "step": 39433 + }, + { + "epoch": 0.8680167290336782, + "grad_norm": 1.972878606589132, + "learning_rate": 4.498968203154086e-07, + "loss": 0.3249, + "step": 39434 + }, + { + "epoch": 0.8680387409200968, + "grad_norm": 2.0808762018980183, + "learning_rate": 4.4974905826905725e-07, + "loss": 0.3266, + "step": 39435 + }, + { + "epoch": 0.8680607528065155, + "grad_norm": 1.6583123285508192, + "learning_rate": 4.496013193491111e-07, + "loss": 0.2693, + "step": 39436 + }, + { + "epoch": 0.8680827646929342, + "grad_norm": 1.3917187989926691, + "learning_rate": 4.4945360355632016e-07, + "loss": 0.2621, + "step": 39437 + }, + { + "epoch": 0.8681047765793528, + "grad_norm": 1.3311359803096807, + "learning_rate": 4.4930591089143603e-07, + "loss": 0.2971, + "step": 39438 + }, + { + "epoch": 0.8681267884657715, + "grad_norm": 2.309675390129447, + "learning_rate": 4.4915824135520703e-07, + "loss": 0.3377, + "step": 39439 + }, + { + "epoch": 0.8681488003521902, + "grad_norm": 1.9539312383878897, + "learning_rate": 4.490105949483875e-07, + "loss": 0.2786, + "step": 39440 + }, + { + "epoch": 0.8681708122386088, + "grad_norm": 0.6316697231794902, + "learning_rate": 4.4886297167172576e-07, + "loss": 0.4563, + "step": 39441 + }, + { + "epoch": 0.8681928241250275, + "grad_norm": 1.6345132734563992, + "learning_rate": 4.487153715259718e-07, + "loss": 0.3389, + "step": 39442 + }, + { + "epoch": 0.8682148360114462, + "grad_norm": 1.9249604625852286, + "learning_rate": 4.485677945118766e-07, + "loss": 0.3238, + "step": 39443 + }, + { + "epoch": 0.8682368478978648, + "grad_norm": 4.176973559935096, + "learning_rate": 4.484202406301891e-07, + "loss": 0.3343, + "step": 39444 + }, + { + "epoch": 0.8682588597842835, + "grad_norm": 0.6403644717372828, + "learning_rate": 4.482727098816603e-07, + "loss": 0.4427, + "step": 39445 + }, + { + "epoch": 0.8682808716707022, + "grad_norm": 2.9789119551880407, + "learning_rate": 4.4812520226704015e-07, + "loss": 0.2932, + "step": 39446 + }, + { + "epoch": 0.8683028835571208, + "grad_norm": 1.7637973543940502, + "learning_rate": 4.4797771778707757e-07, + "loss": 0.2297, + "step": 39447 + }, + { + "epoch": 0.8683248954435395, + "grad_norm": 2.847194925245185, + "learning_rate": 4.47830256442523e-07, + "loss": 0.2962, + "step": 39448 + }, + { + "epoch": 0.8683469073299582, + "grad_norm": 1.6840170300644743, + "learning_rate": 4.476828182341247e-07, + "loss": 0.2959, + "step": 39449 + }, + { + "epoch": 0.8683689192163768, + "grad_norm": 2.012034482645813, + "learning_rate": 4.4753540316263276e-07, + "loss": 0.3134, + "step": 39450 + }, + { + "epoch": 0.8683909311027955, + "grad_norm": 1.7440118410342422, + "learning_rate": 4.4738801122879707e-07, + "loss": 0.2842, + "step": 39451 + }, + { + "epoch": 0.8684129429892141, + "grad_norm": 3.0845986488240564, + "learning_rate": 4.4724064243336527e-07, + "loss": 0.3157, + "step": 39452 + }, + { + "epoch": 0.8684349548756328, + "grad_norm": 1.6347715657224873, + "learning_rate": 4.47093296777088e-07, + "loss": 0.2729, + "step": 39453 + }, + { + "epoch": 0.8684569667620515, + "grad_norm": 1.7746041880437167, + "learning_rate": 4.4694597426071184e-07, + "loss": 0.2971, + "step": 39454 + }, + { + "epoch": 0.8684789786484701, + "grad_norm": 1.7203631231227015, + "learning_rate": 4.4679867488498787e-07, + "loss": 0.3255, + "step": 39455 + }, + { + "epoch": 0.8685009905348888, + "grad_norm": 1.7816328243597588, + "learning_rate": 4.4665139865066375e-07, + "loss": 0.2593, + "step": 39456 + }, + { + "epoch": 0.8685230024213075, + "grad_norm": 0.5872414986785308, + "learning_rate": 4.465041455584879e-07, + "loss": 0.4289, + "step": 39457 + }, + { + "epoch": 0.8685450143077261, + "grad_norm": 1.6178575910289794, + "learning_rate": 4.4635691560920956e-07, + "loss": 0.2794, + "step": 39458 + }, + { + "epoch": 0.8685670261941448, + "grad_norm": 2.0724564845873057, + "learning_rate": 4.462097088035755e-07, + "loss": 0.2847, + "step": 39459 + }, + { + "epoch": 0.8685890380805635, + "grad_norm": 3.627773270889628, + "learning_rate": 4.460625251423345e-07, + "loss": 0.3125, + "step": 39460 + }, + { + "epoch": 0.8686110499669821, + "grad_norm": 0.6278159698315545, + "learning_rate": 4.4591536462623595e-07, + "loss": 0.4592, + "step": 39461 + }, + { + "epoch": 0.8686330618534008, + "grad_norm": 1.998069333339373, + "learning_rate": 4.4576822725602707e-07, + "loss": 0.3232, + "step": 39462 + }, + { + "epoch": 0.8686550737398195, + "grad_norm": 1.6041689621618902, + "learning_rate": 4.456211130324539e-07, + "loss": 0.3111, + "step": 39463 + }, + { + "epoch": 0.8686770856262381, + "grad_norm": 1.6905295837427996, + "learning_rate": 4.454740219562659e-07, + "loss": 0.2081, + "step": 39464 + }, + { + "epoch": 0.8686990975126568, + "grad_norm": 1.6533494168284364, + "learning_rate": 4.453269540282096e-07, + "loss": 0.2852, + "step": 39465 + }, + { + "epoch": 0.8687211093990755, + "grad_norm": 1.1962689182431405, + "learning_rate": 4.45179909249035e-07, + "loss": 0.2617, + "step": 39466 + }, + { + "epoch": 0.8687431212854941, + "grad_norm": 1.8785761139398485, + "learning_rate": 4.45032887619487e-07, + "loss": 0.2571, + "step": 39467 + }, + { + "epoch": 0.8687651331719128, + "grad_norm": 2.125539505931273, + "learning_rate": 4.448858891403124e-07, + "loss": 0.2585, + "step": 39468 + }, + { + "epoch": 0.8687871450583315, + "grad_norm": 2.7907057015734336, + "learning_rate": 4.44738913812261e-07, + "loss": 0.2965, + "step": 39469 + }, + { + "epoch": 0.8688091569447501, + "grad_norm": 0.6295679834356772, + "learning_rate": 4.4459196163607553e-07, + "loss": 0.4622, + "step": 39470 + }, + { + "epoch": 0.8688311688311688, + "grad_norm": 1.7517646244942375, + "learning_rate": 4.444450326125077e-07, + "loss": 0.2472, + "step": 39471 + }, + { + "epoch": 0.8688531807175875, + "grad_norm": 1.9949589602447098, + "learning_rate": 4.442981267423019e-07, + "loss": 0.3105, + "step": 39472 + }, + { + "epoch": 0.8688751926040061, + "grad_norm": 1.8599581728932362, + "learning_rate": 4.441512440262041e-07, + "loss": 0.3045, + "step": 39473 + }, + { + "epoch": 0.8688972044904248, + "grad_norm": 2.3302722412584247, + "learning_rate": 4.440043844649627e-07, + "loss": 0.3373, + "step": 39474 + }, + { + "epoch": 0.8689192163768436, + "grad_norm": 1.6015128250899753, + "learning_rate": 4.43857548059321e-07, + "loss": 0.253, + "step": 39475 + }, + { + "epoch": 0.8689412282632621, + "grad_norm": 0.6265603421230701, + "learning_rate": 4.437107348100295e-07, + "loss": 0.4617, + "step": 39476 + }, + { + "epoch": 0.8689632401496808, + "grad_norm": 2.278007144961272, + "learning_rate": 4.4356394471783135e-07, + "loss": 0.3132, + "step": 39477 + }, + { + "epoch": 0.8689852520360994, + "grad_norm": 1.6500275512986184, + "learning_rate": 4.4341717778347335e-07, + "loss": 0.2788, + "step": 39478 + }, + { + "epoch": 0.8690072639225181, + "grad_norm": 1.6706909352470012, + "learning_rate": 4.432704340077021e-07, + "loss": 0.2473, + "step": 39479 + }, + { + "epoch": 0.8690292758089369, + "grad_norm": 2.0702744903360926, + "learning_rate": 4.431237133912625e-07, + "loss": 0.3065, + "step": 39480 + }, + { + "epoch": 0.8690512876953554, + "grad_norm": 0.6160451638671569, + "learning_rate": 4.4297701593490005e-07, + "loss": 0.4461, + "step": 39481 + }, + { + "epoch": 0.8690732995817742, + "grad_norm": 2.1059427637239603, + "learning_rate": 4.42830341639362e-07, + "loss": 0.3207, + "step": 39482 + }, + { + "epoch": 0.8690953114681929, + "grad_norm": 0.5879609704936298, + "learning_rate": 4.426836905053916e-07, + "loss": 0.4277, + "step": 39483 + }, + { + "epoch": 0.8691173233546114, + "grad_norm": 1.636969201803577, + "learning_rate": 4.425370625337366e-07, + "loss": 0.3148, + "step": 39484 + }, + { + "epoch": 0.8691393352410302, + "grad_norm": 1.8955479260025923, + "learning_rate": 4.4239045772513977e-07, + "loss": 0.3253, + "step": 39485 + }, + { + "epoch": 0.8691613471274489, + "grad_norm": 1.8855385967356022, + "learning_rate": 4.422438760803471e-07, + "loss": 0.2214, + "step": 39486 + }, + { + "epoch": 0.8691833590138675, + "grad_norm": 2.155764490668174, + "learning_rate": 4.4209731760010535e-07, + "loss": 0.3364, + "step": 39487 + }, + { + "epoch": 0.8692053709002862, + "grad_norm": 1.5425090424942525, + "learning_rate": 4.4195078228515654e-07, + "loss": 0.3119, + "step": 39488 + }, + { + "epoch": 0.8692273827867049, + "grad_norm": 1.6328761567934404, + "learning_rate": 4.418042701362479e-07, + "loss": 0.2681, + "step": 39489 + }, + { + "epoch": 0.8692493946731235, + "grad_norm": 0.5904583289165776, + "learning_rate": 4.416577811541223e-07, + "loss": 0.4304, + "step": 39490 + }, + { + "epoch": 0.8692714065595422, + "grad_norm": 2.0527072874160432, + "learning_rate": 4.415113153395245e-07, + "loss": 0.3766, + "step": 39491 + }, + { + "epoch": 0.8692934184459609, + "grad_norm": 2.2210292595764054, + "learning_rate": 4.413648726932007e-07, + "loss": 0.2561, + "step": 39492 + }, + { + "epoch": 0.8693154303323795, + "grad_norm": 2.1263879196631033, + "learning_rate": 4.4121845321589253e-07, + "loss": 0.3133, + "step": 39493 + }, + { + "epoch": 0.8693374422187982, + "grad_norm": 0.5888776570243418, + "learning_rate": 4.4107205690834655e-07, + "loss": 0.4352, + "step": 39494 + }, + { + "epoch": 0.8693594541052169, + "grad_norm": 1.5100207403346684, + "learning_rate": 4.4092568377130495e-07, + "loss": 0.3081, + "step": 39495 + }, + { + "epoch": 0.8693814659916355, + "grad_norm": 1.510396371169337, + "learning_rate": 4.407793338055122e-07, + "loss": 0.3199, + "step": 39496 + }, + { + "epoch": 0.8694034778780542, + "grad_norm": 1.756601019358914, + "learning_rate": 4.406330070117132e-07, + "loss": 0.3058, + "step": 39497 + }, + { + "epoch": 0.8694254897644729, + "grad_norm": 1.464492454860116, + "learning_rate": 4.404867033906501e-07, + "loss": 0.2279, + "step": 39498 + }, + { + "epoch": 0.8694475016508915, + "grad_norm": 0.5628823748156342, + "learning_rate": 4.40340422943068e-07, + "loss": 0.4185, + "step": 39499 + }, + { + "epoch": 0.8694695135373102, + "grad_norm": 1.9008241154199599, + "learning_rate": 4.401941656697084e-07, + "loss": 0.285, + "step": 39500 + }, + { + "epoch": 0.8694915254237288, + "grad_norm": 4.307773807557373, + "learning_rate": 4.4004793157131577e-07, + "loss": 0.2042, + "step": 39501 + }, + { + "epoch": 0.8695135373101475, + "grad_norm": 0.6169856080219829, + "learning_rate": 4.3990172064863444e-07, + "loss": 0.4498, + "step": 39502 + }, + { + "epoch": 0.8695355491965662, + "grad_norm": 1.950403732998576, + "learning_rate": 4.397555329024067e-07, + "loss": 0.2724, + "step": 39503 + }, + { + "epoch": 0.8695575610829848, + "grad_norm": 1.4135401843279785, + "learning_rate": 4.396093683333741e-07, + "loss": 0.2965, + "step": 39504 + }, + { + "epoch": 0.8695795729694035, + "grad_norm": 1.512001168207848, + "learning_rate": 4.394632269422816e-07, + "loss": 0.2872, + "step": 39505 + }, + { + "epoch": 0.8696015848558222, + "grad_norm": 1.5149024208017783, + "learning_rate": 4.393171087298692e-07, + "loss": 0.289, + "step": 39506 + }, + { + "epoch": 0.8696235967422408, + "grad_norm": 2.5764029463603992, + "learning_rate": 4.3917101369688297e-07, + "loss": 0.2484, + "step": 39507 + }, + { + "epoch": 0.8696456086286595, + "grad_norm": 1.645307194304424, + "learning_rate": 4.390249418440645e-07, + "loss": 0.2616, + "step": 39508 + }, + { + "epoch": 0.8696676205150782, + "grad_norm": 1.4776024903134894, + "learning_rate": 4.388788931721538e-07, + "loss": 0.2842, + "step": 39509 + }, + { + "epoch": 0.8696896324014968, + "grad_norm": 1.4409287456706685, + "learning_rate": 4.387328676818964e-07, + "loss": 0.2835, + "step": 39510 + }, + { + "epoch": 0.8697116442879155, + "grad_norm": 0.5699308478267009, + "learning_rate": 4.385868653740316e-07, + "loss": 0.4343, + "step": 39511 + }, + { + "epoch": 0.8697336561743342, + "grad_norm": 1.2974908321637075, + "learning_rate": 4.384408862493034e-07, + "loss": 0.2429, + "step": 39512 + }, + { + "epoch": 0.8697556680607528, + "grad_norm": 1.4930766033012899, + "learning_rate": 4.382949303084533e-07, + "loss": 0.2631, + "step": 39513 + }, + { + "epoch": 0.8697776799471715, + "grad_norm": 0.6135677968814289, + "learning_rate": 4.381489975522224e-07, + "loss": 0.4391, + "step": 39514 + }, + { + "epoch": 0.8697996918335902, + "grad_norm": 1.533347693304553, + "learning_rate": 4.3800308798135406e-07, + "loss": 0.2437, + "step": 39515 + }, + { + "epoch": 0.8698217037200088, + "grad_norm": 1.606678311975468, + "learning_rate": 4.3785720159658763e-07, + "loss": 0.3178, + "step": 39516 + }, + { + "epoch": 0.8698437156064275, + "grad_norm": 2.405614253063114, + "learning_rate": 4.377113383986653e-07, + "loss": 0.294, + "step": 39517 + }, + { + "epoch": 0.8698657274928462, + "grad_norm": 5.29297645045435, + "learning_rate": 4.375654983883304e-07, + "loss": 0.2879, + "step": 39518 + }, + { + "epoch": 0.8698877393792648, + "grad_norm": 4.203536472573006, + "learning_rate": 4.374196815663212e-07, + "loss": 0.2421, + "step": 39519 + }, + { + "epoch": 0.8699097512656835, + "grad_norm": 1.6771769331839463, + "learning_rate": 4.37273887933381e-07, + "loss": 0.3299, + "step": 39520 + }, + { + "epoch": 0.8699317631521022, + "grad_norm": 0.5867671976361457, + "learning_rate": 4.371281174902492e-07, + "loss": 0.4251, + "step": 39521 + }, + { + "epoch": 0.8699537750385208, + "grad_norm": 1.9482312184014703, + "learning_rate": 4.3698237023766745e-07, + "loss": 0.2581, + "step": 39522 + }, + { + "epoch": 0.8699757869249395, + "grad_norm": 1.9905335067357965, + "learning_rate": 4.3683664617637735e-07, + "loss": 0.3878, + "step": 39523 + }, + { + "epoch": 0.8699977988113581, + "grad_norm": 1.7221436274196535, + "learning_rate": 4.366909453071172e-07, + "loss": 0.2571, + "step": 39524 + }, + { + "epoch": 0.8700198106977768, + "grad_norm": 1.7004251369873662, + "learning_rate": 4.3654526763063033e-07, + "loss": 0.2694, + "step": 39525 + }, + { + "epoch": 0.8700418225841955, + "grad_norm": 1.398077931718314, + "learning_rate": 4.36399613147655e-07, + "loss": 0.2474, + "step": 39526 + }, + { + "epoch": 0.8700638344706141, + "grad_norm": 2.9483149312953287, + "learning_rate": 4.362539818589323e-07, + "loss": 0.3062, + "step": 39527 + }, + { + "epoch": 0.8700858463570328, + "grad_norm": 1.2685061098301273, + "learning_rate": 4.3610837376520277e-07, + "loss": 0.3191, + "step": 39528 + }, + { + "epoch": 0.8701078582434515, + "grad_norm": 1.4951148828412466, + "learning_rate": 4.359627888672058e-07, + "loss": 0.2874, + "step": 39529 + }, + { + "epoch": 0.8701298701298701, + "grad_norm": 1.5505143836789042, + "learning_rate": 4.3581722716568186e-07, + "loss": 0.3023, + "step": 39530 + }, + { + "epoch": 0.8701518820162888, + "grad_norm": 1.5925263254642488, + "learning_rate": 4.356716886613699e-07, + "loss": 0.2028, + "step": 39531 + }, + { + "epoch": 0.8701738939027075, + "grad_norm": 1.5324905324359333, + "learning_rate": 4.355261733550098e-07, + "loss": 0.2779, + "step": 39532 + }, + { + "epoch": 0.8701959057891261, + "grad_norm": 2.501299558229689, + "learning_rate": 4.353806812473427e-07, + "loss": 0.3313, + "step": 39533 + }, + { + "epoch": 0.8702179176755448, + "grad_norm": 1.4248353283524067, + "learning_rate": 4.352352123391057e-07, + "loss": 0.2552, + "step": 39534 + }, + { + "epoch": 0.8702399295619635, + "grad_norm": 1.5047061366441263, + "learning_rate": 4.350897666310405e-07, + "loss": 0.2807, + "step": 39535 + }, + { + "epoch": 0.8702619414483821, + "grad_norm": 1.8871539878784251, + "learning_rate": 4.349443441238843e-07, + "loss": 0.2637, + "step": 39536 + }, + { + "epoch": 0.8702839533348008, + "grad_norm": 1.6684649978442312, + "learning_rate": 4.347989448183765e-07, + "loss": 0.2543, + "step": 39537 + }, + { + "epoch": 0.8703059652212195, + "grad_norm": 0.6046149715934881, + "learning_rate": 4.346535687152581e-07, + "loss": 0.4275, + "step": 39538 + }, + { + "epoch": 0.8703279771076381, + "grad_norm": 1.4450486387198351, + "learning_rate": 4.3450821581526527e-07, + "loss": 0.2374, + "step": 39539 + }, + { + "epoch": 0.8703499889940568, + "grad_norm": 2.1093219387158593, + "learning_rate": 4.34362886119139e-07, + "loss": 0.3466, + "step": 39540 + }, + { + "epoch": 0.8703720008804755, + "grad_norm": 1.6894107712704713, + "learning_rate": 4.3421757962761657e-07, + "loss": 0.2545, + "step": 39541 + }, + { + "epoch": 0.8703940127668941, + "grad_norm": 3.093172563703176, + "learning_rate": 4.340722963414351e-07, + "loss": 0.3922, + "step": 39542 + }, + { + "epoch": 0.8704160246533128, + "grad_norm": 1.438455479313142, + "learning_rate": 4.339270362613368e-07, + "loss": 0.3161, + "step": 39543 + }, + { + "epoch": 0.8704380365397315, + "grad_norm": 1.4590156910948255, + "learning_rate": 4.337817993880572e-07, + "loss": 0.2945, + "step": 39544 + }, + { + "epoch": 0.8704600484261501, + "grad_norm": 2.225254351388405, + "learning_rate": 4.3363658572233457e-07, + "loss": 0.3497, + "step": 39545 + }, + { + "epoch": 0.8704820603125688, + "grad_norm": 1.380211967942452, + "learning_rate": 4.3349139526490837e-07, + "loss": 0.3084, + "step": 39546 + }, + { + "epoch": 0.8705040721989874, + "grad_norm": 1.7819693263719925, + "learning_rate": 4.3334622801651463e-07, + "loss": 0.3698, + "step": 39547 + }, + { + "epoch": 0.8705260840854061, + "grad_norm": 1.5315197743190532, + "learning_rate": 4.332010839778922e-07, + "loss": 0.3097, + "step": 39548 + }, + { + "epoch": 0.8705480959718248, + "grad_norm": 1.3767831347213135, + "learning_rate": 4.3305596314977994e-07, + "loss": 0.2547, + "step": 39549 + }, + { + "epoch": 0.8705701078582434, + "grad_norm": 1.7211167147933055, + "learning_rate": 4.3291086553291283e-07, + "loss": 0.341, + "step": 39550 + }, + { + "epoch": 0.8705921197446621, + "grad_norm": 0.5891468010422125, + "learning_rate": 4.3276579112803084e-07, + "loss": 0.4372, + "step": 39551 + }, + { + "epoch": 0.8706141316310808, + "grad_norm": 1.5156722899515989, + "learning_rate": 4.3262073993586897e-07, + "loss": 0.3951, + "step": 39552 + }, + { + "epoch": 0.8706361435174994, + "grad_norm": 1.3494220934234749, + "learning_rate": 4.32475711957166e-07, + "loss": 0.2886, + "step": 39553 + }, + { + "epoch": 0.8706581554039181, + "grad_norm": 0.6078224106048925, + "learning_rate": 4.323307071926591e-07, + "loss": 0.4328, + "step": 39554 + }, + { + "epoch": 0.8706801672903368, + "grad_norm": 0.6268155920572901, + "learning_rate": 4.3218572564308394e-07, + "loss": 0.4424, + "step": 39555 + }, + { + "epoch": 0.8707021791767554, + "grad_norm": 1.660532342142149, + "learning_rate": 4.3204076730917923e-07, + "loss": 0.2169, + "step": 39556 + }, + { + "epoch": 0.8707241910631741, + "grad_norm": 2.5593318684510136, + "learning_rate": 4.3189583219168e-07, + "loss": 0.3293, + "step": 39557 + }, + { + "epoch": 0.8707462029495928, + "grad_norm": 2.014010954620711, + "learning_rate": 4.3175092029132336e-07, + "loss": 0.2985, + "step": 39558 + }, + { + "epoch": 0.8707682148360114, + "grad_norm": 2.1000457043067717, + "learning_rate": 4.3160603160884716e-07, + "loss": 0.3342, + "step": 39559 + }, + { + "epoch": 0.8707902267224301, + "grad_norm": 0.6383298147175982, + "learning_rate": 4.314611661449858e-07, + "loss": 0.4673, + "step": 39560 + }, + { + "epoch": 0.8708122386088488, + "grad_norm": 1.923860118783867, + "learning_rate": 4.3131632390047694e-07, + "loss": 0.3004, + "step": 39561 + }, + { + "epoch": 0.8708342504952674, + "grad_norm": 0.6199995070652676, + "learning_rate": 4.3117150487605565e-07, + "loss": 0.4359, + "step": 39562 + }, + { + "epoch": 0.8708562623816861, + "grad_norm": 2.086664882292095, + "learning_rate": 4.3102670907245794e-07, + "loss": 0.3302, + "step": 39563 + }, + { + "epoch": 0.8708782742681048, + "grad_norm": 1.5563540561993845, + "learning_rate": 4.308819364904221e-07, + "loss": 0.2837, + "step": 39564 + }, + { + "epoch": 0.8709002861545234, + "grad_norm": 1.2791250689434452, + "learning_rate": 4.3073718713068037e-07, + "loss": 0.2669, + "step": 39565 + }, + { + "epoch": 0.8709222980409421, + "grad_norm": 1.8575166985619995, + "learning_rate": 4.305924609939716e-07, + "loss": 0.314, + "step": 39566 + }, + { + "epoch": 0.8709443099273608, + "grad_norm": 1.8754414877606762, + "learning_rate": 4.30447758081029e-07, + "loss": 0.2056, + "step": 39567 + }, + { + "epoch": 0.8709663218137794, + "grad_norm": 2.0734351771631454, + "learning_rate": 4.303030783925888e-07, + "loss": 0.3276, + "step": 39568 + }, + { + "epoch": 0.8709883337001981, + "grad_norm": 0.5893179093901618, + "learning_rate": 4.301584219293881e-07, + "loss": 0.4334, + "step": 39569 + }, + { + "epoch": 0.8710103455866167, + "grad_norm": 2.8488884796409244, + "learning_rate": 4.300137886921585e-07, + "loss": 0.2467, + "step": 39570 + }, + { + "epoch": 0.8710323574730354, + "grad_norm": 1.91159849499802, + "learning_rate": 4.298691786816389e-07, + "loss": 0.3221, + "step": 39571 + }, + { + "epoch": 0.8710543693594541, + "grad_norm": 1.3533408162517608, + "learning_rate": 4.2972459189856097e-07, + "loss": 0.2568, + "step": 39572 + }, + { + "epoch": 0.8710763812458727, + "grad_norm": 1.8173360390113475, + "learning_rate": 4.295800283436613e-07, + "loss": 0.3334, + "step": 39573 + }, + { + "epoch": 0.8710983931322914, + "grad_norm": 1.6572971876464928, + "learning_rate": 4.294354880176754e-07, + "loss": 0.3107, + "step": 39574 + }, + { + "epoch": 0.8711204050187101, + "grad_norm": 1.7652231743171407, + "learning_rate": 4.292909709213361e-07, + "loss": 0.2953, + "step": 39575 + }, + { + "epoch": 0.8711424169051287, + "grad_norm": 1.705959458058289, + "learning_rate": 4.2914647705537936e-07, + "loss": 0.2939, + "step": 39576 + }, + { + "epoch": 0.8711644287915474, + "grad_norm": 3.3576103817392453, + "learning_rate": 4.290020064205386e-07, + "loss": 0.2485, + "step": 39577 + }, + { + "epoch": 0.8711864406779661, + "grad_norm": 1.5043389060659689, + "learning_rate": 4.28857559017547e-07, + "loss": 0.2886, + "step": 39578 + }, + { + "epoch": 0.8712084525643847, + "grad_norm": 1.7364766351272123, + "learning_rate": 4.287131348471418e-07, + "loss": 0.3201, + "step": 39579 + }, + { + "epoch": 0.8712304644508034, + "grad_norm": 0.60191094405088, + "learning_rate": 4.2856873391005415e-07, + "loss": 0.4167, + "step": 39580 + }, + { + "epoch": 0.8712524763372221, + "grad_norm": 1.7911297089833291, + "learning_rate": 4.2842435620702003e-07, + "loss": 0.3003, + "step": 39581 + }, + { + "epoch": 0.8712744882236407, + "grad_norm": 1.6029672641274588, + "learning_rate": 4.2828000173877284e-07, + "loss": 0.2818, + "step": 39582 + }, + { + "epoch": 0.8712965001100594, + "grad_norm": 1.9103680247812078, + "learning_rate": 4.281356705060441e-07, + "loss": 0.2982, + "step": 39583 + }, + { + "epoch": 0.8713185119964781, + "grad_norm": 1.6844734916010702, + "learning_rate": 4.279913625095694e-07, + "loss": 0.2494, + "step": 39584 + }, + { + "epoch": 0.8713405238828967, + "grad_norm": 1.8912939573069134, + "learning_rate": 4.278470777500826e-07, + "loss": 0.2975, + "step": 39585 + }, + { + "epoch": 0.8713625357693154, + "grad_norm": 1.5881390706531722, + "learning_rate": 4.2770281622831534e-07, + "loss": 0.2465, + "step": 39586 + }, + { + "epoch": 0.8713845476557341, + "grad_norm": 1.9826678549186336, + "learning_rate": 4.2755857794500255e-07, + "loss": 0.2374, + "step": 39587 + }, + { + "epoch": 0.8714065595421527, + "grad_norm": 1.7644256221266683, + "learning_rate": 4.274143629008759e-07, + "loss": 0.2099, + "step": 39588 + }, + { + "epoch": 0.8714285714285714, + "grad_norm": 1.4707788015003864, + "learning_rate": 4.2727017109666814e-07, + "loss": 0.3102, + "step": 39589 + }, + { + "epoch": 0.8714505833149901, + "grad_norm": 1.3154632818426202, + "learning_rate": 4.271260025331142e-07, + "loss": 0.2452, + "step": 39590 + }, + { + "epoch": 0.8714725952014087, + "grad_norm": 2.530349461435946, + "learning_rate": 4.2698185721094465e-07, + "loss": 0.2381, + "step": 39591 + }, + { + "epoch": 0.8714946070878274, + "grad_norm": 1.7073086356175826, + "learning_rate": 4.2683773513089334e-07, + "loss": 0.2889, + "step": 39592 + }, + { + "epoch": 0.871516618974246, + "grad_norm": 1.6720297882263033, + "learning_rate": 4.2669363629369187e-07, + "loss": 0.2951, + "step": 39593 + }, + { + "epoch": 0.8715386308606647, + "grad_norm": 1.4634505243969147, + "learning_rate": 4.26549560700073e-07, + "loss": 0.2758, + "step": 39594 + }, + { + "epoch": 0.8715606427470834, + "grad_norm": 1.2472470375586637, + "learning_rate": 4.264055083507701e-07, + "loss": 0.2391, + "step": 39595 + }, + { + "epoch": 0.871582654633502, + "grad_norm": 1.5512652491922179, + "learning_rate": 4.262614792465131e-07, + "loss": 0.3541, + "step": 39596 + }, + { + "epoch": 0.8716046665199207, + "grad_norm": 1.3657456326547242, + "learning_rate": 4.2611747338803633e-07, + "loss": 0.2857, + "step": 39597 + }, + { + "epoch": 0.8716266784063395, + "grad_norm": 1.6452300970128986, + "learning_rate": 4.2597349077606985e-07, + "loss": 0.2553, + "step": 39598 + }, + { + "epoch": 0.871648690292758, + "grad_norm": 2.132393263613259, + "learning_rate": 4.2582953141134587e-07, + "loss": 0.2796, + "step": 39599 + }, + { + "epoch": 0.8716707021791767, + "grad_norm": 1.4859351060918145, + "learning_rate": 4.2568559529459706e-07, + "loss": 0.2379, + "step": 39600 + }, + { + "epoch": 0.8716927140655955, + "grad_norm": 1.6275842764214374, + "learning_rate": 4.2554168242655404e-07, + "loss": 0.2816, + "step": 39601 + }, + { + "epoch": 0.871714725952014, + "grad_norm": 1.9088083251798549, + "learning_rate": 4.253977928079489e-07, + "loss": 0.2822, + "step": 39602 + }, + { + "epoch": 0.8717367378384328, + "grad_norm": 1.7993305612013468, + "learning_rate": 4.2525392643951113e-07, + "loss": 0.3264, + "step": 39603 + }, + { + "epoch": 0.8717587497248515, + "grad_norm": 0.6183248493825807, + "learning_rate": 4.25110083321974e-07, + "loss": 0.4652, + "step": 39604 + }, + { + "epoch": 0.87178076161127, + "grad_norm": 1.5461086605784315, + "learning_rate": 4.249662634560686e-07, + "loss": 0.3741, + "step": 39605 + }, + { + "epoch": 0.8718027734976888, + "grad_norm": 1.3958586171428318, + "learning_rate": 4.2482246684252436e-07, + "loss": 0.22, + "step": 39606 + }, + { + "epoch": 0.8718247853841075, + "grad_norm": 1.804721184059587, + "learning_rate": 4.246786934820735e-07, + "loss": 0.2293, + "step": 39607 + }, + { + "epoch": 0.871846797270526, + "grad_norm": 2.0382762884945866, + "learning_rate": 4.2453494337544644e-07, + "loss": 0.2484, + "step": 39608 + }, + { + "epoch": 0.8718688091569448, + "grad_norm": 2.081124788901952, + "learning_rate": 4.2439121652337157e-07, + "loss": 0.3845, + "step": 39609 + }, + { + "epoch": 0.8718908210433635, + "grad_norm": 2.9990278924678417, + "learning_rate": 4.242475129265833e-07, + "loss": 0.2743, + "step": 39610 + }, + { + "epoch": 0.8719128329297821, + "grad_norm": 1.7650664691245448, + "learning_rate": 4.241038325858088e-07, + "loss": 0.2181, + "step": 39611 + }, + { + "epoch": 0.8719348448162008, + "grad_norm": 1.571219695104059, + "learning_rate": 4.2396017550178083e-07, + "loss": 0.2737, + "step": 39612 + }, + { + "epoch": 0.8719568567026195, + "grad_norm": 1.946301548997243, + "learning_rate": 4.2381654167522824e-07, + "loss": 0.2424, + "step": 39613 + }, + { + "epoch": 0.8719788685890381, + "grad_norm": 0.6398513840273237, + "learning_rate": 4.236729311068788e-07, + "loss": 0.4517, + "step": 39614 + }, + { + "epoch": 0.8720008804754568, + "grad_norm": 1.5518015592109826, + "learning_rate": 4.235293437974669e-07, + "loss": 0.2753, + "step": 39615 + }, + { + "epoch": 0.8720228923618755, + "grad_norm": 0.5988856304238814, + "learning_rate": 4.233857797477187e-07, + "loss": 0.417, + "step": 39616 + }, + { + "epoch": 0.8720449042482941, + "grad_norm": 1.9344885723498269, + "learning_rate": 4.2324223895836626e-07, + "loss": 0.2406, + "step": 39617 + }, + { + "epoch": 0.8720669161347128, + "grad_norm": 2.036689017849262, + "learning_rate": 4.23098721430138e-07, + "loss": 0.3043, + "step": 39618 + }, + { + "epoch": 0.8720889280211314, + "grad_norm": 1.5085955460782032, + "learning_rate": 4.2295522716376267e-07, + "loss": 0.2848, + "step": 39619 + }, + { + "epoch": 0.8721109399075501, + "grad_norm": 1.624575685003086, + "learning_rate": 4.228117561599698e-07, + "loss": 0.2236, + "step": 39620 + }, + { + "epoch": 0.8721329517939688, + "grad_norm": 0.5962102484316253, + "learning_rate": 4.226683084194899e-07, + "loss": 0.4294, + "step": 39621 + }, + { + "epoch": 0.8721549636803874, + "grad_norm": 1.5363111401354292, + "learning_rate": 4.2252488394305126e-07, + "loss": 0.2102, + "step": 39622 + }, + { + "epoch": 0.8721769755668061, + "grad_norm": 1.6772701266526526, + "learning_rate": 4.2238148273138323e-07, + "loss": 0.2924, + "step": 39623 + }, + { + "epoch": 0.8721989874532248, + "grad_norm": 1.271880930169195, + "learning_rate": 4.2223810478521363e-07, + "loss": 0.2039, + "step": 39624 + }, + { + "epoch": 0.8722209993396434, + "grad_norm": 0.6297522660613829, + "learning_rate": 4.220947501052708e-07, + "loss": 0.4672, + "step": 39625 + }, + { + "epoch": 0.8722430112260621, + "grad_norm": 2.3221193443543253, + "learning_rate": 4.2195141869228574e-07, + "loss": 0.4075, + "step": 39626 + }, + { + "epoch": 0.8722650231124808, + "grad_norm": 4.299041090301504, + "learning_rate": 4.2180811054698456e-07, + "loss": 0.3254, + "step": 39627 + }, + { + "epoch": 0.8722870349988994, + "grad_norm": 3.9789450329012865, + "learning_rate": 4.2166482567009725e-07, + "loss": 0.3558, + "step": 39628 + }, + { + "epoch": 0.8723090468853181, + "grad_norm": 1.8243491786095052, + "learning_rate": 4.2152156406235045e-07, + "loss": 0.3465, + "step": 39629 + }, + { + "epoch": 0.8723310587717368, + "grad_norm": 2.1109843491426634, + "learning_rate": 4.21378325724473e-07, + "loss": 0.2883, + "step": 39630 + }, + { + "epoch": 0.8723530706581554, + "grad_norm": 1.742354298681985, + "learning_rate": 4.2123511065719436e-07, + "loss": 0.2571, + "step": 39631 + }, + { + "epoch": 0.8723750825445741, + "grad_norm": 1.5812996142127733, + "learning_rate": 4.210919188612395e-07, + "loss": 0.2501, + "step": 39632 + }, + { + "epoch": 0.8723970944309928, + "grad_norm": 1.5638732490646325, + "learning_rate": 4.209487503373394e-07, + "loss": 0.2601, + "step": 39633 + }, + { + "epoch": 0.8724191063174114, + "grad_norm": 2.109739422792522, + "learning_rate": 4.208056050862186e-07, + "loss": 0.3016, + "step": 39634 + }, + { + "epoch": 0.8724411182038301, + "grad_norm": 0.6384611294883684, + "learning_rate": 4.206624831086059e-07, + "loss": 0.4586, + "step": 39635 + }, + { + "epoch": 0.8724631300902488, + "grad_norm": 1.4377860871391028, + "learning_rate": 4.205193844052302e-07, + "loss": 0.2415, + "step": 39636 + }, + { + "epoch": 0.8724851419766674, + "grad_norm": 2.0895138019677746, + "learning_rate": 4.2037630897681646e-07, + "loss": 0.3449, + "step": 39637 + }, + { + "epoch": 0.8725071538630861, + "grad_norm": 1.4000001740175212, + "learning_rate": 4.202332568240941e-07, + "loss": 0.3129, + "step": 39638 + }, + { + "epoch": 0.8725291657495048, + "grad_norm": 1.7761113767296037, + "learning_rate": 4.2009022794778807e-07, + "loss": 0.3403, + "step": 39639 + }, + { + "epoch": 0.8725511776359234, + "grad_norm": 2.175251452322581, + "learning_rate": 4.199472223486256e-07, + "loss": 0.2921, + "step": 39640 + }, + { + "epoch": 0.8725731895223421, + "grad_norm": 1.443048171484449, + "learning_rate": 4.1980424002733557e-07, + "loss": 0.2413, + "step": 39641 + }, + { + "epoch": 0.8725952014087607, + "grad_norm": 2.025642270812549, + "learning_rate": 4.196612809846423e-07, + "loss": 0.302, + "step": 39642 + }, + { + "epoch": 0.8726172132951794, + "grad_norm": 1.674402691917759, + "learning_rate": 4.1951834522127367e-07, + "loss": 0.2998, + "step": 39643 + }, + { + "epoch": 0.8726392251815981, + "grad_norm": 3.5539721810029006, + "learning_rate": 4.193754327379562e-07, + "loss": 0.2336, + "step": 39644 + }, + { + "epoch": 0.8726612370680167, + "grad_norm": 0.6139113097550772, + "learning_rate": 4.1923254353541386e-07, + "loss": 0.4156, + "step": 39645 + }, + { + "epoch": 0.8726832489544354, + "grad_norm": 7.268291337483012, + "learning_rate": 4.190896776143766e-07, + "loss": 0.3073, + "step": 39646 + }, + { + "epoch": 0.8727052608408541, + "grad_norm": 1.9344551559289822, + "learning_rate": 4.1894683497556764e-07, + "loss": 0.3152, + "step": 39647 + }, + { + "epoch": 0.8727272727272727, + "grad_norm": 0.6104531556105347, + "learning_rate": 4.188040156197154e-07, + "loss": 0.423, + "step": 39648 + }, + { + "epoch": 0.8727492846136914, + "grad_norm": 3.226312061806503, + "learning_rate": 4.186612195475442e-07, + "loss": 0.2975, + "step": 39649 + }, + { + "epoch": 0.8727712965001101, + "grad_norm": 1.5131157993091768, + "learning_rate": 4.185184467597786e-07, + "loss": 0.2802, + "step": 39650 + }, + { + "epoch": 0.8727933083865287, + "grad_norm": 0.6540303704925978, + "learning_rate": 4.183756972571468e-07, + "loss": 0.4407, + "step": 39651 + }, + { + "epoch": 0.8728153202729474, + "grad_norm": 1.6931428392719095, + "learning_rate": 4.1823297104037265e-07, + "loss": 0.2276, + "step": 39652 + }, + { + "epoch": 0.8728373321593661, + "grad_norm": 2.4192764546381462, + "learning_rate": 4.180902681101828e-07, + "loss": 0.2951, + "step": 39653 + }, + { + "epoch": 0.8728593440457847, + "grad_norm": 1.722648683987702, + "learning_rate": 4.1794758846730177e-07, + "loss": 0.1962, + "step": 39654 + }, + { + "epoch": 0.8728813559322034, + "grad_norm": 2.188733005536955, + "learning_rate": 4.178049321124544e-07, + "loss": 0.3215, + "step": 39655 + }, + { + "epoch": 0.8729033678186221, + "grad_norm": 1.5919187302810636, + "learning_rate": 4.1766229904636523e-07, + "loss": 0.2472, + "step": 39656 + }, + { + "epoch": 0.8729253797050407, + "grad_norm": 1.8812089462770338, + "learning_rate": 4.175196892697608e-07, + "loss": 0.3101, + "step": 39657 + }, + { + "epoch": 0.8729473915914594, + "grad_norm": 5.622249789337569, + "learning_rate": 4.1737710278336563e-07, + "loss": 0.3172, + "step": 39658 + }, + { + "epoch": 0.8729694034778781, + "grad_norm": 1.6829758885819481, + "learning_rate": 4.1723453958790406e-07, + "loss": 0.2191, + "step": 39659 + }, + { + "epoch": 0.8729914153642967, + "grad_norm": 1.8137634901326458, + "learning_rate": 4.170919996840994e-07, + "loss": 0.2225, + "step": 39660 + }, + { + "epoch": 0.8730134272507154, + "grad_norm": 1.8862293638799186, + "learning_rate": 4.1694948307267724e-07, + "loss": 0.2521, + "step": 39661 + }, + { + "epoch": 0.8730354391371341, + "grad_norm": 6.2890546196605674, + "learning_rate": 4.1680698975436306e-07, + "loss": 0.3372, + "step": 39662 + }, + { + "epoch": 0.8730574510235527, + "grad_norm": 7.405341931926355, + "learning_rate": 4.166645197298791e-07, + "loss": 0.312, + "step": 39663 + }, + { + "epoch": 0.8730794629099714, + "grad_norm": 1.5725252065863176, + "learning_rate": 4.1652207299995085e-07, + "loss": 0.3237, + "step": 39664 + }, + { + "epoch": 0.87310147479639, + "grad_norm": 0.610419209371732, + "learning_rate": 4.163796495653011e-07, + "loss": 0.4404, + "step": 39665 + }, + { + "epoch": 0.8731234866828087, + "grad_norm": 1.9523319103826497, + "learning_rate": 4.162372494266542e-07, + "loss": 0.352, + "step": 39666 + }, + { + "epoch": 0.8731454985692274, + "grad_norm": 1.6530583659472717, + "learning_rate": 4.160948725847347e-07, + "loss": 0.3343, + "step": 39667 + }, + { + "epoch": 0.873167510455646, + "grad_norm": 1.6377564767067396, + "learning_rate": 4.1595251904026524e-07, + "loss": 0.2795, + "step": 39668 + }, + { + "epoch": 0.8731895223420647, + "grad_norm": 2.7441000183850717, + "learning_rate": 4.1581018879396973e-07, + "loss": 0.3075, + "step": 39669 + }, + { + "epoch": 0.8732115342284834, + "grad_norm": 0.6214304725296486, + "learning_rate": 4.156678818465709e-07, + "loss": 0.4436, + "step": 39670 + }, + { + "epoch": 0.873233546114902, + "grad_norm": 1.6676116881178884, + "learning_rate": 4.155255981987927e-07, + "loss": 0.2406, + "step": 39671 + }, + { + "epoch": 0.8732555580013207, + "grad_norm": 0.6194950233417036, + "learning_rate": 4.153833378513589e-07, + "loss": 0.4209, + "step": 39672 + }, + { + "epoch": 0.8732775698877394, + "grad_norm": 1.8611827452731073, + "learning_rate": 4.1524110080499113e-07, + "loss": 0.2535, + "step": 39673 + }, + { + "epoch": 0.873299581774158, + "grad_norm": 4.689916866601418, + "learning_rate": 4.1509888706041335e-07, + "loss": 0.2478, + "step": 39674 + }, + { + "epoch": 0.8733215936605767, + "grad_norm": 1.702618675204702, + "learning_rate": 4.149566966183477e-07, + "loss": 0.2635, + "step": 39675 + }, + { + "epoch": 0.8733436055469954, + "grad_norm": 1.4141155629665714, + "learning_rate": 4.148145294795164e-07, + "loss": 0.2477, + "step": 39676 + }, + { + "epoch": 0.873365617433414, + "grad_norm": 2.464322313117672, + "learning_rate": 4.1467238564464386e-07, + "loss": 0.2806, + "step": 39677 + }, + { + "epoch": 0.8733876293198327, + "grad_norm": 2.19376899283891, + "learning_rate": 4.145302651144506e-07, + "loss": 0.298, + "step": 39678 + }, + { + "epoch": 0.8734096412062514, + "grad_norm": 1.8818618663033944, + "learning_rate": 4.1438816788966053e-07, + "loss": 0.2884, + "step": 39679 + }, + { + "epoch": 0.87343165309267, + "grad_norm": 0.6115329454015929, + "learning_rate": 4.142460939709941e-07, + "loss": 0.4557, + "step": 39680 + }, + { + "epoch": 0.8734536649790887, + "grad_norm": 1.7687845557516702, + "learning_rate": 4.141040433591748e-07, + "loss": 0.3294, + "step": 39681 + }, + { + "epoch": 0.8734756768655074, + "grad_norm": 2.904920482329832, + "learning_rate": 4.1396201605492457e-07, + "loss": 0.2654, + "step": 39682 + }, + { + "epoch": 0.873497688751926, + "grad_norm": 1.8406659002257693, + "learning_rate": 4.1382001205896363e-07, + "loss": 0.3517, + "step": 39683 + }, + { + "epoch": 0.8735197006383447, + "grad_norm": 1.7450031626416687, + "learning_rate": 4.136780313720162e-07, + "loss": 0.3212, + "step": 39684 + }, + { + "epoch": 0.8735417125247634, + "grad_norm": 4.137092522081935, + "learning_rate": 4.1353607399480243e-07, + "loss": 0.1853, + "step": 39685 + }, + { + "epoch": 0.873563724411182, + "grad_norm": 1.930907101175349, + "learning_rate": 4.1339413992804223e-07, + "loss": 0.2965, + "step": 39686 + }, + { + "epoch": 0.8735857362976007, + "grad_norm": 2.226855118719461, + "learning_rate": 4.132522291724605e-07, + "loss": 0.2726, + "step": 39687 + }, + { + "epoch": 0.8736077481840193, + "grad_norm": 1.5633616153114265, + "learning_rate": 4.1311034172877576e-07, + "loss": 0.2748, + "step": 39688 + }, + { + "epoch": 0.873629760070438, + "grad_norm": 2.2344677922492107, + "learning_rate": 4.129684775977111e-07, + "loss": 0.315, + "step": 39689 + }, + { + "epoch": 0.8736517719568567, + "grad_norm": 1.4510367604073597, + "learning_rate": 4.12826636779986e-07, + "loss": 0.2725, + "step": 39690 + }, + { + "epoch": 0.8736737838432753, + "grad_norm": 1.7209724475818784, + "learning_rate": 4.1268481927632165e-07, + "loss": 0.3541, + "step": 39691 + }, + { + "epoch": 0.873695795729694, + "grad_norm": 3.2802237012793865, + "learning_rate": 4.125430250874385e-07, + "loss": 0.2876, + "step": 39692 + }, + { + "epoch": 0.8737178076161127, + "grad_norm": 1.4769974656713745, + "learning_rate": 4.124012542140582e-07, + "loss": 0.3125, + "step": 39693 + }, + { + "epoch": 0.8737398195025313, + "grad_norm": 2.0046220591291433, + "learning_rate": 4.122595066569013e-07, + "loss": 0.3173, + "step": 39694 + }, + { + "epoch": 0.87376183138895, + "grad_norm": 0.6262377191441255, + "learning_rate": 4.121177824166883e-07, + "loss": 0.4592, + "step": 39695 + }, + { + "epoch": 0.8737838432753687, + "grad_norm": 2.0776541361669683, + "learning_rate": 4.1197608149413757e-07, + "loss": 0.2453, + "step": 39696 + }, + { + "epoch": 0.8738058551617873, + "grad_norm": 0.6043792721278992, + "learning_rate": 4.118344038899713e-07, + "loss": 0.4151, + "step": 39697 + }, + { + "epoch": 0.873827867048206, + "grad_norm": 1.5928259609219, + "learning_rate": 4.1169274960490834e-07, + "loss": 0.2417, + "step": 39698 + }, + { + "epoch": 0.8738498789346247, + "grad_norm": 2.098839210211908, + "learning_rate": 4.1155111863967037e-07, + "loss": 0.267, + "step": 39699 + }, + { + "epoch": 0.8738718908210433, + "grad_norm": 1.3102987588943464, + "learning_rate": 4.1140951099497617e-07, + "loss": 0.3128, + "step": 39700 + }, + { + "epoch": 0.873893902707462, + "grad_norm": 0.6353505874815711, + "learning_rate": 4.112679266715447e-07, + "loss": 0.4596, + "step": 39701 + }, + { + "epoch": 0.8739159145938807, + "grad_norm": 1.9973853723943378, + "learning_rate": 4.111263656700959e-07, + "loss": 0.3333, + "step": 39702 + }, + { + "epoch": 0.8739379264802993, + "grad_norm": 2.4083147691231086, + "learning_rate": 4.109848279913509e-07, + "loss": 0.2862, + "step": 39703 + }, + { + "epoch": 0.873959938366718, + "grad_norm": 2.221044149176155, + "learning_rate": 4.108433136360268e-07, + "loss": 0.2693, + "step": 39704 + }, + { + "epoch": 0.8739819502531367, + "grad_norm": 2.4046705073214554, + "learning_rate": 4.1070182260484424e-07, + "loss": 0.2967, + "step": 39705 + }, + { + "epoch": 0.8740039621395553, + "grad_norm": 2.8593850400830916, + "learning_rate": 4.1056035489852154e-07, + "loss": 0.3045, + "step": 39706 + }, + { + "epoch": 0.874025974025974, + "grad_norm": 1.9004735011786056, + "learning_rate": 4.104189105177775e-07, + "loss": 0.249, + "step": 39707 + }, + { + "epoch": 0.8740479859123927, + "grad_norm": 1.3836027541627007, + "learning_rate": 4.1027748946333326e-07, + "loss": 0.241, + "step": 39708 + }, + { + "epoch": 0.8740699977988113, + "grad_norm": 1.1547075180957722, + "learning_rate": 4.101360917359043e-07, + "loss": 0.229, + "step": 39709 + }, + { + "epoch": 0.87409200968523, + "grad_norm": 1.9471940405134878, + "learning_rate": 4.099947173362118e-07, + "loss": 0.301, + "step": 39710 + }, + { + "epoch": 0.8741140215716486, + "grad_norm": 1.7393026565694043, + "learning_rate": 4.0985336626497283e-07, + "loss": 0.2788, + "step": 39711 + }, + { + "epoch": 0.8741360334580673, + "grad_norm": 1.6412764905657704, + "learning_rate": 4.0971203852290585e-07, + "loss": 0.3078, + "step": 39712 + }, + { + "epoch": 0.874158045344486, + "grad_norm": 1.598855900098497, + "learning_rate": 4.0957073411073076e-07, + "loss": 0.2504, + "step": 39713 + }, + { + "epoch": 0.8741800572309046, + "grad_norm": 1.5353528154278377, + "learning_rate": 4.094294530291637e-07, + "loss": 0.2982, + "step": 39714 + }, + { + "epoch": 0.8742020691173233, + "grad_norm": 1.091939991994222, + "learning_rate": 4.092881952789246e-07, + "loss": 0.1775, + "step": 39715 + }, + { + "epoch": 0.874224081003742, + "grad_norm": 0.6150855669908519, + "learning_rate": 4.091469608607296e-07, + "loss": 0.4479, + "step": 39716 + }, + { + "epoch": 0.8742460928901606, + "grad_norm": 3.67138820576806, + "learning_rate": 4.0900574977529695e-07, + "loss": 0.2817, + "step": 39717 + }, + { + "epoch": 0.8742681047765793, + "grad_norm": 1.338070476506739, + "learning_rate": 4.0886456202334614e-07, + "loss": 0.2182, + "step": 39718 + }, + { + "epoch": 0.874290116662998, + "grad_norm": 1.5979808724639517, + "learning_rate": 4.087233976055921e-07, + "loss": 0.3421, + "step": 39719 + }, + { + "epoch": 0.8743121285494166, + "grad_norm": 1.672393956761665, + "learning_rate": 4.0858225652275427e-07, + "loss": 0.3128, + "step": 39720 + }, + { + "epoch": 0.8743341404358353, + "grad_norm": 1.5616933903109576, + "learning_rate": 4.084411387755488e-07, + "loss": 0.2117, + "step": 39721 + }, + { + "epoch": 0.874356152322254, + "grad_norm": 1.531402865742304, + "learning_rate": 4.083000443646939e-07, + "loss": 0.2879, + "step": 39722 + }, + { + "epoch": 0.8743781642086726, + "grad_norm": 0.6169371125804777, + "learning_rate": 4.0815897329090526e-07, + "loss": 0.461, + "step": 39723 + }, + { + "epoch": 0.8744001760950914, + "grad_norm": 1.736007709340118, + "learning_rate": 4.080179255549005e-07, + "loss": 0.2537, + "step": 39724 + }, + { + "epoch": 0.8744221879815101, + "grad_norm": 1.376746242090118, + "learning_rate": 4.07876901157398e-07, + "loss": 0.2735, + "step": 39725 + }, + { + "epoch": 0.8744441998679287, + "grad_norm": 1.3302992586506037, + "learning_rate": 4.0773590009911326e-07, + "loss": 0.2471, + "step": 39726 + }, + { + "epoch": 0.8744662117543474, + "grad_norm": 1.8142033904850021, + "learning_rate": 4.075949223807618e-07, + "loss": 0.2617, + "step": 39727 + }, + { + "epoch": 0.8744882236407661, + "grad_norm": 3.0352389104166226, + "learning_rate": 4.0745396800306145e-07, + "loss": 0.3848, + "step": 39728 + }, + { + "epoch": 0.8745102355271847, + "grad_norm": 3.0616020131954045, + "learning_rate": 4.0731303696672777e-07, + "loss": 0.2687, + "step": 39729 + }, + { + "epoch": 0.8745322474136034, + "grad_norm": 0.6066025767597563, + "learning_rate": 4.0717212927247896e-07, + "loss": 0.4155, + "step": 39730 + }, + { + "epoch": 0.8745542593000221, + "grad_norm": 0.6082619856209243, + "learning_rate": 4.070312449210301e-07, + "loss": 0.4224, + "step": 39731 + }, + { + "epoch": 0.8745762711864407, + "grad_norm": 1.6948440091401158, + "learning_rate": 4.0689038391309556e-07, + "loss": 0.197, + "step": 39732 + }, + { + "epoch": 0.8745982830728594, + "grad_norm": 1.790076980102952, + "learning_rate": 4.0674954624939314e-07, + "loss": 0.2991, + "step": 39733 + }, + { + "epoch": 0.874620294959278, + "grad_norm": 1.483767128463099, + "learning_rate": 4.0660873193063776e-07, + "loss": 0.2918, + "step": 39734 + }, + { + "epoch": 0.8746423068456967, + "grad_norm": 2.0866852170301904, + "learning_rate": 4.064679409575467e-07, + "loss": 0.305, + "step": 39735 + }, + { + "epoch": 0.8746643187321154, + "grad_norm": 1.6540006089024637, + "learning_rate": 4.0632717333083437e-07, + "loss": 0.2843, + "step": 39736 + }, + { + "epoch": 0.874686330618534, + "grad_norm": 1.9530325495940912, + "learning_rate": 4.061864290512152e-07, + "loss": 0.404, + "step": 39737 + }, + { + "epoch": 0.8747083425049527, + "grad_norm": 1.4810560632338328, + "learning_rate": 4.0604570811940524e-07, + "loss": 0.2998, + "step": 39738 + }, + { + "epoch": 0.8747303543913714, + "grad_norm": 1.6519078676672514, + "learning_rate": 4.059050105361201e-07, + "loss": 0.2635, + "step": 39739 + }, + { + "epoch": 0.87475236627779, + "grad_norm": 1.570876804111758, + "learning_rate": 4.057643363020758e-07, + "loss": 0.3134, + "step": 39740 + }, + { + "epoch": 0.8747743781642087, + "grad_norm": 0.616534697562151, + "learning_rate": 4.056236854179862e-07, + "loss": 0.4373, + "step": 39741 + }, + { + "epoch": 0.8747963900506274, + "grad_norm": 1.945736620162406, + "learning_rate": 4.054830578845653e-07, + "loss": 0.2792, + "step": 39742 + }, + { + "epoch": 0.874818401937046, + "grad_norm": 1.5087060122892317, + "learning_rate": 4.0534245370252855e-07, + "loss": 0.28, + "step": 39743 + }, + { + "epoch": 0.8748404138234647, + "grad_norm": 2.076983293469812, + "learning_rate": 4.052018728725915e-07, + "loss": 0.2761, + "step": 39744 + }, + { + "epoch": 0.8748624257098834, + "grad_norm": 1.3238678300073954, + "learning_rate": 4.050613153954669e-07, + "loss": 0.2089, + "step": 39745 + }, + { + "epoch": 0.874884437596302, + "grad_norm": 1.6921089727680785, + "learning_rate": 4.049207812718714e-07, + "loss": 0.3166, + "step": 39746 + }, + { + "epoch": 0.8749064494827207, + "grad_norm": 0.6209299743990988, + "learning_rate": 4.0478027050251667e-07, + "loss": 0.4614, + "step": 39747 + }, + { + "epoch": 0.8749284613691394, + "grad_norm": 1.3283719338683369, + "learning_rate": 4.046397830881188e-07, + "loss": 0.267, + "step": 39748 + }, + { + "epoch": 0.874950473255558, + "grad_norm": 1.4722882508466706, + "learning_rate": 4.0449931902939164e-07, + "loss": 0.2226, + "step": 39749 + }, + { + "epoch": 0.8749724851419767, + "grad_norm": 1.4889019463611637, + "learning_rate": 4.043588783270475e-07, + "loss": 0.3192, + "step": 39750 + }, + { + "epoch": 0.8749944970283954, + "grad_norm": 0.6148951880822825, + "learning_rate": 4.0421846098180227e-07, + "loss": 0.4696, + "step": 39751 + }, + { + "epoch": 0.875016508914814, + "grad_norm": 2.668393312358615, + "learning_rate": 4.040780669943678e-07, + "loss": 0.2442, + "step": 39752 + }, + { + "epoch": 0.8750385208012327, + "grad_norm": 2.4416449258443107, + "learning_rate": 4.039376963654584e-07, + "loss": 0.2501, + "step": 39753 + }, + { + "epoch": 0.8750605326876514, + "grad_norm": 2.318089150756942, + "learning_rate": 4.0379734909578807e-07, + "loss": 0.2575, + "step": 39754 + }, + { + "epoch": 0.87508254457407, + "grad_norm": 1.5098846697880621, + "learning_rate": 4.0365702518606885e-07, + "loss": 0.2694, + "step": 39755 + }, + { + "epoch": 0.8751045564604887, + "grad_norm": 1.828179705682672, + "learning_rate": 4.0351672463701587e-07, + "loss": 0.2524, + "step": 39756 + }, + { + "epoch": 0.8751265683469074, + "grad_norm": 1.7287345315623042, + "learning_rate": 4.033764474493401e-07, + "loss": 0.2542, + "step": 39757 + }, + { + "epoch": 0.875148580233326, + "grad_norm": 1.774841487286367, + "learning_rate": 4.032361936237561e-07, + "loss": 0.2347, + "step": 39758 + }, + { + "epoch": 0.8751705921197447, + "grad_norm": 1.7043618747221403, + "learning_rate": 4.030959631609749e-07, + "loss": 0.3044, + "step": 39759 + }, + { + "epoch": 0.8751926040061633, + "grad_norm": 1.3907075149116297, + "learning_rate": 4.0295575606171033e-07, + "loss": 0.267, + "step": 39760 + }, + { + "epoch": 0.875214615892582, + "grad_norm": 1.6498419698864137, + "learning_rate": 4.0281557232667634e-07, + "loss": 0.3266, + "step": 39761 + }, + { + "epoch": 0.8752366277790007, + "grad_norm": 2.3927349098291892, + "learning_rate": 4.0267541195658346e-07, + "loss": 0.3642, + "step": 39762 + }, + { + "epoch": 0.8752586396654193, + "grad_norm": 2.148727656072215, + "learning_rate": 4.025352749521438e-07, + "loss": 0.2757, + "step": 39763 + }, + { + "epoch": 0.875280651551838, + "grad_norm": 1.7817027045802436, + "learning_rate": 4.023951613140703e-07, + "loss": 0.2307, + "step": 39764 + }, + { + "epoch": 0.8753026634382567, + "grad_norm": 1.3365052554000478, + "learning_rate": 4.022550710430756e-07, + "loss": 0.2827, + "step": 39765 + }, + { + "epoch": 0.8753246753246753, + "grad_norm": 2.83347818406361, + "learning_rate": 4.021150041398719e-07, + "loss": 0.3116, + "step": 39766 + }, + { + "epoch": 0.875346687211094, + "grad_norm": 1.8263653994603768, + "learning_rate": 4.019749606051704e-07, + "loss": 0.2507, + "step": 39767 + }, + { + "epoch": 0.8753686990975127, + "grad_norm": 1.9883055863293106, + "learning_rate": 4.0183494043968206e-07, + "loss": 0.3616, + "step": 39768 + }, + { + "epoch": 0.8753907109839313, + "grad_norm": 1.4259563237704567, + "learning_rate": 4.016949436441198e-07, + "loss": 0.271, + "step": 39769 + }, + { + "epoch": 0.87541272287035, + "grad_norm": 1.5525003063686638, + "learning_rate": 4.015549702191945e-07, + "loss": 0.2713, + "step": 39770 + }, + { + "epoch": 0.8754347347567687, + "grad_norm": 1.936605308056522, + "learning_rate": 4.014150201656186e-07, + "loss": 0.2994, + "step": 39771 + }, + { + "epoch": 0.8754567466431873, + "grad_norm": 2.0068932240728117, + "learning_rate": 4.0127509348410256e-07, + "loss": 0.3136, + "step": 39772 + }, + { + "epoch": 0.875478758529606, + "grad_norm": 2.0185762023132217, + "learning_rate": 4.011351901753563e-07, + "loss": 0.3329, + "step": 39773 + }, + { + "epoch": 0.8755007704160247, + "grad_norm": 4.482437717498538, + "learning_rate": 4.0099531024009264e-07, + "loss": 0.3543, + "step": 39774 + }, + { + "epoch": 0.8755227823024433, + "grad_norm": 1.7028761240580406, + "learning_rate": 4.0085545367902214e-07, + "loss": 0.272, + "step": 39775 + }, + { + "epoch": 0.875544794188862, + "grad_norm": 1.5079994438873865, + "learning_rate": 4.007156204928564e-07, + "loss": 0.3571, + "step": 39776 + }, + { + "epoch": 0.8755668060752807, + "grad_norm": 3.6631740946779097, + "learning_rate": 4.005758106823049e-07, + "loss": 0.294, + "step": 39777 + }, + { + "epoch": 0.8755888179616993, + "grad_norm": 2.8236022703921293, + "learning_rate": 4.004360242480776e-07, + "loss": 0.3532, + "step": 39778 + }, + { + "epoch": 0.875610829848118, + "grad_norm": 2.2705367716630005, + "learning_rate": 4.002962611908862e-07, + "loss": 0.3634, + "step": 39779 + }, + { + "epoch": 0.8756328417345367, + "grad_norm": 1.45321335705536, + "learning_rate": 4.001565215114406e-07, + "loss": 0.2838, + "step": 39780 + }, + { + "epoch": 0.8756548536209553, + "grad_norm": 1.404460944375771, + "learning_rate": 4.0001680521045196e-07, + "loss": 0.3711, + "step": 39781 + }, + { + "epoch": 0.875676865507374, + "grad_norm": 1.809395805595025, + "learning_rate": 3.9987711228862966e-07, + "loss": 0.3049, + "step": 39782 + }, + { + "epoch": 0.8756988773937926, + "grad_norm": 1.4588596009685, + "learning_rate": 3.9973744274668267e-07, + "loss": 0.2803, + "step": 39783 + }, + { + "epoch": 0.8757208892802113, + "grad_norm": 0.6362904910545775, + "learning_rate": 3.995977965853215e-07, + "loss": 0.4663, + "step": 39784 + }, + { + "epoch": 0.87574290116663, + "grad_norm": 1.5950921361449844, + "learning_rate": 3.9945817380525775e-07, + "loss": 0.3056, + "step": 39785 + }, + { + "epoch": 0.8757649130530486, + "grad_norm": 2.061388073238567, + "learning_rate": 3.9931857440719814e-07, + "loss": 0.348, + "step": 39786 + }, + { + "epoch": 0.8757869249394673, + "grad_norm": 1.3678753714037346, + "learning_rate": 3.9917899839185425e-07, + "loss": 0.2787, + "step": 39787 + }, + { + "epoch": 0.875808936825886, + "grad_norm": 0.6154728209292298, + "learning_rate": 3.9903944575993393e-07, + "loss": 0.427, + "step": 39788 + }, + { + "epoch": 0.8758309487123046, + "grad_norm": 1.8664134362770217, + "learning_rate": 3.988999165121471e-07, + "loss": 0.3725, + "step": 39789 + }, + { + "epoch": 0.8758529605987233, + "grad_norm": 0.6154204094092761, + "learning_rate": 3.9876041064920434e-07, + "loss": 0.447, + "step": 39790 + }, + { + "epoch": 0.875874972485142, + "grad_norm": 2.224040956542679, + "learning_rate": 3.986209281718123e-07, + "loss": 0.332, + "step": 39791 + }, + { + "epoch": 0.8758969843715606, + "grad_norm": 1.5073273107326355, + "learning_rate": 3.984814690806815e-07, + "loss": 0.3426, + "step": 39792 + }, + { + "epoch": 0.8759189962579793, + "grad_norm": 0.6275022700976802, + "learning_rate": 3.9834203337651977e-07, + "loss": 0.4649, + "step": 39793 + }, + { + "epoch": 0.875941008144398, + "grad_norm": 1.7079888668836818, + "learning_rate": 3.9820262106003704e-07, + "loss": 0.2381, + "step": 39794 + }, + { + "epoch": 0.8759630200308166, + "grad_norm": 1.4155403562619027, + "learning_rate": 3.9806323213193997e-07, + "loss": 0.3302, + "step": 39795 + }, + { + "epoch": 0.8759850319172353, + "grad_norm": 1.688274138371742, + "learning_rate": 3.9792386659293805e-07, + "loss": 0.2231, + "step": 39796 + }, + { + "epoch": 0.876007043803654, + "grad_norm": 2.0421029307801946, + "learning_rate": 3.977845244437406e-07, + "loss": 0.3696, + "step": 39797 + }, + { + "epoch": 0.8760290556900726, + "grad_norm": 1.5909725176086837, + "learning_rate": 3.976452056850544e-07, + "loss": 0.3138, + "step": 39798 + }, + { + "epoch": 0.8760510675764913, + "grad_norm": 1.8924772818319167, + "learning_rate": 3.9750591031758823e-07, + "loss": 0.2764, + "step": 39799 + }, + { + "epoch": 0.87607307946291, + "grad_norm": 1.7743565659913971, + "learning_rate": 3.9736663834204937e-07, + "loss": 0.2774, + "step": 39800 + }, + { + "epoch": 0.8760950913493286, + "grad_norm": 1.575132029102696, + "learning_rate": 3.972273897591455e-07, + "loss": 0.2882, + "step": 39801 + }, + { + "epoch": 0.8761171032357473, + "grad_norm": 1.2304484351922935, + "learning_rate": 3.970881645695862e-07, + "loss": 0.2858, + "step": 39802 + }, + { + "epoch": 0.876139115122166, + "grad_norm": 85.81191604823023, + "learning_rate": 3.96948962774078e-07, + "loss": 0.2361, + "step": 39803 + }, + { + "epoch": 0.8761611270085846, + "grad_norm": 3.1703098232902684, + "learning_rate": 3.9680978437332705e-07, + "loss": 0.302, + "step": 39804 + }, + { + "epoch": 0.8761831388950033, + "grad_norm": 2.0779760501244646, + "learning_rate": 3.9667062936804166e-07, + "loss": 0.285, + "step": 39805 + }, + { + "epoch": 0.8762051507814219, + "grad_norm": 2.1214825996608035, + "learning_rate": 3.965314977589296e-07, + "loss": 0.2744, + "step": 39806 + }, + { + "epoch": 0.8762271626678406, + "grad_norm": 0.6296985492692845, + "learning_rate": 3.963923895466987e-07, + "loss": 0.46, + "step": 39807 + }, + { + "epoch": 0.8762491745542593, + "grad_norm": 1.741152611773931, + "learning_rate": 3.9625330473205493e-07, + "loss": 0.2739, + "step": 39808 + }, + { + "epoch": 0.8762711864406779, + "grad_norm": 1.9058732398474396, + "learning_rate": 3.961142433157039e-07, + "loss": 0.3316, + "step": 39809 + }, + { + "epoch": 0.8762931983270966, + "grad_norm": 0.6109869477916802, + "learning_rate": 3.9597520529835396e-07, + "loss": 0.4528, + "step": 39810 + }, + { + "epoch": 0.8763152102135153, + "grad_norm": 1.7175805965715822, + "learning_rate": 3.958361906807112e-07, + "loss": 0.2612, + "step": 39811 + }, + { + "epoch": 0.8763372220999339, + "grad_norm": 4.503029490886495, + "learning_rate": 3.956971994634834e-07, + "loss": 0.3527, + "step": 39812 + }, + { + "epoch": 0.8763592339863526, + "grad_norm": 1.7370878974046933, + "learning_rate": 3.955582316473761e-07, + "loss": 0.278, + "step": 39813 + }, + { + "epoch": 0.8763812458727713, + "grad_norm": 1.7006435636463313, + "learning_rate": 3.9541928723309477e-07, + "loss": 0.2489, + "step": 39814 + }, + { + "epoch": 0.8764032577591899, + "grad_norm": 1.7725958041270102, + "learning_rate": 3.9528036622134615e-07, + "loss": 0.3016, + "step": 39815 + }, + { + "epoch": 0.8764252696456086, + "grad_norm": 1.7725975668588152, + "learning_rate": 3.951414686128363e-07, + "loss": 0.2998, + "step": 39816 + }, + { + "epoch": 0.8764472815320273, + "grad_norm": 1.4825049839233229, + "learning_rate": 3.950025944082725e-07, + "loss": 0.2654, + "step": 39817 + }, + { + "epoch": 0.8764692934184459, + "grad_norm": 2.4432314606251904, + "learning_rate": 3.9486374360835966e-07, + "loss": 0.2921, + "step": 39818 + }, + { + "epoch": 0.8764913053048646, + "grad_norm": 2.3768482505712054, + "learning_rate": 3.947249162138017e-07, + "loss": 0.2933, + "step": 39819 + }, + { + "epoch": 0.8765133171912833, + "grad_norm": 4.564586699970044, + "learning_rate": 3.945861122253064e-07, + "loss": 0.2265, + "step": 39820 + }, + { + "epoch": 0.8765353290777019, + "grad_norm": 1.805524345783922, + "learning_rate": 3.9444733164357817e-07, + "loss": 0.294, + "step": 39821 + }, + { + "epoch": 0.8765573409641206, + "grad_norm": 1.7253996324414214, + "learning_rate": 3.943085744693231e-07, + "loss": 0.3557, + "step": 39822 + }, + { + "epoch": 0.8765793528505393, + "grad_norm": 1.6875122081204736, + "learning_rate": 3.941698407032468e-07, + "loss": 0.3321, + "step": 39823 + }, + { + "epoch": 0.8766013647369579, + "grad_norm": 0.6198097181026435, + "learning_rate": 3.9403113034605257e-07, + "loss": 0.4506, + "step": 39824 + }, + { + "epoch": 0.8766233766233766, + "grad_norm": 1.5573929548745948, + "learning_rate": 3.938924433984465e-07, + "loss": 0.2205, + "step": 39825 + }, + { + "epoch": 0.8766453885097953, + "grad_norm": 1.601084010752009, + "learning_rate": 3.9375377986113407e-07, + "loss": 0.2889, + "step": 39826 + }, + { + "epoch": 0.8766674003962139, + "grad_norm": 1.9988760351007862, + "learning_rate": 3.936151397348187e-07, + "loss": 0.2968, + "step": 39827 + }, + { + "epoch": 0.8766894122826326, + "grad_norm": 0.6458512226826212, + "learning_rate": 3.9347652302020643e-07, + "loss": 0.4759, + "step": 39828 + }, + { + "epoch": 0.8767114241690512, + "grad_norm": 2.096543444185139, + "learning_rate": 3.9333792971800014e-07, + "loss": 0.2989, + "step": 39829 + }, + { + "epoch": 0.8767334360554699, + "grad_norm": 0.6353714157596041, + "learning_rate": 3.9319935982890634e-07, + "loss": 0.4644, + "step": 39830 + }, + { + "epoch": 0.8767554479418886, + "grad_norm": 4.442814514031564, + "learning_rate": 3.9306081335362676e-07, + "loss": 0.2996, + "step": 39831 + }, + { + "epoch": 0.8767774598283072, + "grad_norm": 1.9297870685862453, + "learning_rate": 3.92922290292867e-07, + "loss": 0.196, + "step": 39832 + }, + { + "epoch": 0.8767994717147259, + "grad_norm": 1.7595290264481611, + "learning_rate": 3.9278379064733194e-07, + "loss": 0.3216, + "step": 39833 + }, + { + "epoch": 0.8768214836011446, + "grad_norm": 2.1208829662815587, + "learning_rate": 3.926453144177239e-07, + "loss": 0.3478, + "step": 39834 + }, + { + "epoch": 0.8768434954875632, + "grad_norm": 1.2527249167095753, + "learning_rate": 3.925068616047478e-07, + "loss": 0.2744, + "step": 39835 + }, + { + "epoch": 0.876865507373982, + "grad_norm": 1.8174809840940367, + "learning_rate": 3.923684322091059e-07, + "loss": 0.3073, + "step": 39836 + }, + { + "epoch": 0.8768875192604006, + "grad_norm": 1.8314370872040515, + "learning_rate": 3.922300262315032e-07, + "loss": 0.2935, + "step": 39837 + }, + { + "epoch": 0.8769095311468192, + "grad_norm": 1.6060085499360968, + "learning_rate": 3.92091643672643e-07, + "loss": 0.311, + "step": 39838 + }, + { + "epoch": 0.876931543033238, + "grad_norm": 1.9577553302949466, + "learning_rate": 3.919532845332275e-07, + "loss": 0.3568, + "step": 39839 + }, + { + "epoch": 0.8769535549196567, + "grad_norm": 1.3734169319110938, + "learning_rate": 3.9181494881396167e-07, + "loss": 0.1811, + "step": 39840 + }, + { + "epoch": 0.8769755668060752, + "grad_norm": 2.0463115564808865, + "learning_rate": 3.9167663651554665e-07, + "loss": 0.3392, + "step": 39841 + }, + { + "epoch": 0.876997578692494, + "grad_norm": 1.738916898168589, + "learning_rate": 3.9153834763868636e-07, + "loss": 0.2345, + "step": 39842 + }, + { + "epoch": 0.8770195905789127, + "grad_norm": 1.796610628048433, + "learning_rate": 3.914000821840841e-07, + "loss": 0.217, + "step": 39843 + }, + { + "epoch": 0.8770416024653312, + "grad_norm": 2.420420782517841, + "learning_rate": 3.912618401524426e-07, + "loss": 0.3063, + "step": 39844 + }, + { + "epoch": 0.87706361435175, + "grad_norm": 1.60584713818736, + "learning_rate": 3.91123621544463e-07, + "loss": 0.199, + "step": 39845 + }, + { + "epoch": 0.8770856262381687, + "grad_norm": 1.4984801228716158, + "learning_rate": 3.909854263608487e-07, + "loss": 0.2807, + "step": 39846 + }, + { + "epoch": 0.8771076381245873, + "grad_norm": 1.8968603349968647, + "learning_rate": 3.9084725460230176e-07, + "loss": 0.3513, + "step": 39847 + }, + { + "epoch": 0.877129650011006, + "grad_norm": 1.7380001928949382, + "learning_rate": 3.9070910626952617e-07, + "loss": 0.3402, + "step": 39848 + }, + { + "epoch": 0.8771516618974247, + "grad_norm": 1.6225641586296589, + "learning_rate": 3.90570981363223e-07, + "loss": 0.2894, + "step": 39849 + }, + { + "epoch": 0.8771736737838433, + "grad_norm": 1.4188063889203433, + "learning_rate": 3.904328798840923e-07, + "loss": 0.1871, + "step": 39850 + }, + { + "epoch": 0.877195685670262, + "grad_norm": 2.037275975653224, + "learning_rate": 3.902948018328384e-07, + "loss": 0.2922, + "step": 39851 + }, + { + "epoch": 0.8772176975566806, + "grad_norm": 1.6497897577528517, + "learning_rate": 3.9015674721016195e-07, + "loss": 0.3126, + "step": 39852 + }, + { + "epoch": 0.8772397094430993, + "grad_norm": 2.415345667661596, + "learning_rate": 3.900187160167657e-07, + "loss": 0.3345, + "step": 39853 + }, + { + "epoch": 0.877261721329518, + "grad_norm": 1.5978684137682193, + "learning_rate": 3.8988070825335065e-07, + "loss": 0.2162, + "step": 39854 + }, + { + "epoch": 0.8772837332159366, + "grad_norm": 1.81797367065844, + "learning_rate": 3.8974272392061695e-07, + "loss": 0.3023, + "step": 39855 + }, + { + "epoch": 0.8773057451023553, + "grad_norm": 1.662601523933163, + "learning_rate": 3.896047630192673e-07, + "loss": 0.289, + "step": 39856 + }, + { + "epoch": 0.877327756988774, + "grad_norm": 1.6861742608022945, + "learning_rate": 3.8946682555000224e-07, + "loss": 0.2937, + "step": 39857 + }, + { + "epoch": 0.8773497688751926, + "grad_norm": 1.4723964450609563, + "learning_rate": 3.89328911513524e-07, + "loss": 0.2664, + "step": 39858 + }, + { + "epoch": 0.8773717807616113, + "grad_norm": 1.8976026712491254, + "learning_rate": 3.891910209105326e-07, + "loss": 0.3936, + "step": 39859 + }, + { + "epoch": 0.87739379264803, + "grad_norm": 8.439117286497842, + "learning_rate": 3.89053153741728e-07, + "loss": 0.3074, + "step": 39860 + }, + { + "epoch": 0.8774158045344486, + "grad_norm": 1.6180741463964532, + "learning_rate": 3.889153100078119e-07, + "loss": 0.3192, + "step": 39861 + }, + { + "epoch": 0.8774378164208673, + "grad_norm": 1.4323041662340052, + "learning_rate": 3.8877748970948593e-07, + "loss": 0.3011, + "step": 39862 + }, + { + "epoch": 0.877459828307286, + "grad_norm": 0.6123794014021865, + "learning_rate": 3.886396928474484e-07, + "loss": 0.446, + "step": 39863 + }, + { + "epoch": 0.8774818401937046, + "grad_norm": 1.9032466972042184, + "learning_rate": 3.8850191942240156e-07, + "loss": 0.3419, + "step": 39864 + }, + { + "epoch": 0.8775038520801233, + "grad_norm": 2.103000893645608, + "learning_rate": 3.8836416943504375e-07, + "loss": 0.309, + "step": 39865 + }, + { + "epoch": 0.877525863966542, + "grad_norm": 0.6284179094053474, + "learning_rate": 3.8822644288607713e-07, + "loss": 0.4595, + "step": 39866 + }, + { + "epoch": 0.8775478758529606, + "grad_norm": 2.236848517011897, + "learning_rate": 3.8808873977619956e-07, + "loss": 0.3152, + "step": 39867 + }, + { + "epoch": 0.8775698877393793, + "grad_norm": 1.850138935994073, + "learning_rate": 3.8795106010611206e-07, + "loss": 0.2539, + "step": 39868 + }, + { + "epoch": 0.877591899625798, + "grad_norm": 0.6221279732361974, + "learning_rate": 3.8781340387651524e-07, + "loss": 0.4461, + "step": 39869 + }, + { + "epoch": 0.8776139115122166, + "grad_norm": 2.6343758713504535, + "learning_rate": 3.8767577108810683e-07, + "loss": 0.2745, + "step": 39870 + }, + { + "epoch": 0.8776359233986353, + "grad_norm": 1.6342197836555112, + "learning_rate": 3.87538161741588e-07, + "loss": 0.2901, + "step": 39871 + }, + { + "epoch": 0.877657935285054, + "grad_norm": 1.470848167323941, + "learning_rate": 3.87400575837657e-07, + "loss": 0.3244, + "step": 39872 + }, + { + "epoch": 0.8776799471714726, + "grad_norm": 2.235890993000311, + "learning_rate": 3.872630133770133e-07, + "loss": 0.2318, + "step": 39873 + }, + { + "epoch": 0.8777019590578913, + "grad_norm": 1.6150107298970053, + "learning_rate": 3.87125474360357e-07, + "loss": 0.2813, + "step": 39874 + }, + { + "epoch": 0.8777239709443099, + "grad_norm": 1.6015182955255007, + "learning_rate": 3.869879587883857e-07, + "loss": 0.2878, + "step": 39875 + }, + { + "epoch": 0.8777459828307286, + "grad_norm": 3.440600496030488, + "learning_rate": 3.868504666618e-07, + "loss": 0.2915, + "step": 39876 + }, + { + "epoch": 0.8777679947171473, + "grad_norm": 1.7396604972735885, + "learning_rate": 3.8671299798129727e-07, + "loss": 0.2526, + "step": 39877 + }, + { + "epoch": 0.8777900066035659, + "grad_norm": 1.946610102904222, + "learning_rate": 3.865755527475762e-07, + "loss": 0.3408, + "step": 39878 + }, + { + "epoch": 0.8778120184899846, + "grad_norm": 0.6007208872263412, + "learning_rate": 3.8643813096133687e-07, + "loss": 0.4387, + "step": 39879 + }, + { + "epoch": 0.8778340303764033, + "grad_norm": 1.7003447247963202, + "learning_rate": 3.863007326232754e-07, + "loss": 0.3222, + "step": 39880 + }, + { + "epoch": 0.8778560422628219, + "grad_norm": 10.468156092726908, + "learning_rate": 3.8616335773409287e-07, + "loss": 0.3127, + "step": 39881 + }, + { + "epoch": 0.8778780541492406, + "grad_norm": 80.02618038636784, + "learning_rate": 3.860260062944848e-07, + "loss": 0.3527, + "step": 39882 + }, + { + "epoch": 0.8779000660356593, + "grad_norm": 1.4807629717000836, + "learning_rate": 3.858886783051502e-07, + "loss": 0.2466, + "step": 39883 + }, + { + "epoch": 0.8779220779220779, + "grad_norm": 0.607225350361325, + "learning_rate": 3.857513737667884e-07, + "loss": 0.4222, + "step": 39884 + }, + { + "epoch": 0.8779440898084966, + "grad_norm": 0.599551586406142, + "learning_rate": 3.856140926800961e-07, + "loss": 0.4416, + "step": 39885 + }, + { + "epoch": 0.8779661016949153, + "grad_norm": 2.2186424106605895, + "learning_rate": 3.8547683504577047e-07, + "loss": 0.3287, + "step": 39886 + }, + { + "epoch": 0.8779881135813339, + "grad_norm": 1.7922443553675529, + "learning_rate": 3.8533960086450937e-07, + "loss": 0.3756, + "step": 39887 + }, + { + "epoch": 0.8780101254677526, + "grad_norm": 1.4257144996886908, + "learning_rate": 3.8520239013701057e-07, + "loss": 0.2396, + "step": 39888 + }, + { + "epoch": 0.8780321373541713, + "grad_norm": 1.571237576522046, + "learning_rate": 3.850652028639723e-07, + "loss": 0.2735, + "step": 39889 + }, + { + "epoch": 0.8780541492405899, + "grad_norm": 1.7696626060740606, + "learning_rate": 3.849280390460913e-07, + "loss": 0.2585, + "step": 39890 + }, + { + "epoch": 0.8780761611270086, + "grad_norm": 1.532238865625247, + "learning_rate": 3.8479089868406314e-07, + "loss": 0.3137, + "step": 39891 + }, + { + "epoch": 0.8780981730134273, + "grad_norm": 3.1230333232213368, + "learning_rate": 3.846537817785867e-07, + "loss": 0.3418, + "step": 39892 + }, + { + "epoch": 0.8781201848998459, + "grad_norm": 0.6311776513711086, + "learning_rate": 3.8451668833035747e-07, + "loss": 0.4484, + "step": 39893 + }, + { + "epoch": 0.8781421967862646, + "grad_norm": 1.6081571182200052, + "learning_rate": 3.843796183400744e-07, + "loss": 0.2421, + "step": 39894 + }, + { + "epoch": 0.8781642086726833, + "grad_norm": 2.05856433150833, + "learning_rate": 3.842425718084325e-07, + "loss": 0.276, + "step": 39895 + }, + { + "epoch": 0.8781862205591019, + "grad_norm": 1.5429898841504568, + "learning_rate": 3.841055487361278e-07, + "loss": 0.3388, + "step": 39896 + }, + { + "epoch": 0.8782082324455206, + "grad_norm": 3.0995051116352195, + "learning_rate": 3.839685491238576e-07, + "loss": 0.2697, + "step": 39897 + }, + { + "epoch": 0.8782302443319393, + "grad_norm": 0.616817749509725, + "learning_rate": 3.83831572972318e-07, + "loss": 0.442, + "step": 39898 + }, + { + "epoch": 0.8782522562183579, + "grad_norm": 1.66398610733761, + "learning_rate": 3.836946202822056e-07, + "loss": 0.2491, + "step": 39899 + }, + { + "epoch": 0.8782742681047766, + "grad_norm": 1.4523551711396785, + "learning_rate": 3.835576910542166e-07, + "loss": 0.2135, + "step": 39900 + }, + { + "epoch": 0.8782962799911952, + "grad_norm": 3.663610671300706, + "learning_rate": 3.8342078528904535e-07, + "loss": 0.2434, + "step": 39901 + }, + { + "epoch": 0.8783182918776139, + "grad_norm": 2.231010040882304, + "learning_rate": 3.8328390298738916e-07, + "loss": 0.3033, + "step": 39902 + }, + { + "epoch": 0.8783403037640326, + "grad_norm": 9.059866642314264, + "learning_rate": 3.8314704414994295e-07, + "loss": 0.3057, + "step": 39903 + }, + { + "epoch": 0.8783623156504512, + "grad_norm": 1.7982830818030584, + "learning_rate": 3.8301020877740234e-07, + "loss": 0.2628, + "step": 39904 + }, + { + "epoch": 0.8783843275368699, + "grad_norm": 1.2250163045886426, + "learning_rate": 3.8287339687046455e-07, + "loss": 0.2505, + "step": 39905 + }, + { + "epoch": 0.8784063394232886, + "grad_norm": 1.5913153048140363, + "learning_rate": 3.8273660842982176e-07, + "loss": 0.329, + "step": 39906 + }, + { + "epoch": 0.8784283513097072, + "grad_norm": 2.0211621471307613, + "learning_rate": 3.825998434561723e-07, + "loss": 0.2758, + "step": 39907 + }, + { + "epoch": 0.8784503631961259, + "grad_norm": 1.3154320580943482, + "learning_rate": 3.82463101950209e-07, + "loss": 0.2554, + "step": 39908 + }, + { + "epoch": 0.8784723750825446, + "grad_norm": 0.588581080905554, + "learning_rate": 3.8232638391262736e-07, + "loss": 0.4317, + "step": 39909 + }, + { + "epoch": 0.8784943869689632, + "grad_norm": 2.0006972619076144, + "learning_rate": 3.821896893441235e-07, + "loss": 0.2633, + "step": 39910 + }, + { + "epoch": 0.8785163988553819, + "grad_norm": 5.813190942153282, + "learning_rate": 3.8205301824539074e-07, + "loss": 0.3102, + "step": 39911 + }, + { + "epoch": 0.8785384107418006, + "grad_norm": 0.6171613379020755, + "learning_rate": 3.8191637061712474e-07, + "loss": 0.4376, + "step": 39912 + }, + { + "epoch": 0.8785604226282192, + "grad_norm": 0.6251403019728828, + "learning_rate": 3.8177974646001867e-07, + "loss": 0.4522, + "step": 39913 + }, + { + "epoch": 0.8785824345146379, + "grad_norm": 1.8513867655051877, + "learning_rate": 3.816431457747677e-07, + "loss": 0.2671, + "step": 39914 + }, + { + "epoch": 0.8786044464010566, + "grad_norm": 1.4878038950823407, + "learning_rate": 3.815065685620667e-07, + "loss": 0.27, + "step": 39915 + }, + { + "epoch": 0.8786264582874752, + "grad_norm": 1.8424421065557341, + "learning_rate": 3.8137001482260905e-07, + "loss": 0.2224, + "step": 39916 + }, + { + "epoch": 0.8786484701738939, + "grad_norm": 9.098117580038474, + "learning_rate": 3.812334845570892e-07, + "loss": 0.2789, + "step": 39917 + }, + { + "epoch": 0.8786704820603126, + "grad_norm": 1.5277737606139719, + "learning_rate": 3.8109697776620047e-07, + "loss": 0.2423, + "step": 39918 + }, + { + "epoch": 0.8786924939467312, + "grad_norm": 1.6733400543198953, + "learning_rate": 3.8096049445063623e-07, + "loss": 0.2874, + "step": 39919 + }, + { + "epoch": 0.8787145058331499, + "grad_norm": 0.6160280751801561, + "learning_rate": 3.808240346110925e-07, + "loss": 0.4349, + "step": 39920 + }, + { + "epoch": 0.8787365177195686, + "grad_norm": 1.5907440245305222, + "learning_rate": 3.8068759824826e-07, + "loss": 0.2777, + "step": 39921 + }, + { + "epoch": 0.8787585296059872, + "grad_norm": 1.2885019141663028, + "learning_rate": 3.805511853628341e-07, + "loss": 0.2752, + "step": 39922 + }, + { + "epoch": 0.8787805414924059, + "grad_norm": 1.8795132803135586, + "learning_rate": 3.8041479595550714e-07, + "loss": 0.3738, + "step": 39923 + }, + { + "epoch": 0.8788025533788245, + "grad_norm": 1.9733505396617277, + "learning_rate": 3.802784300269724e-07, + "loss": 0.2322, + "step": 39924 + }, + { + "epoch": 0.8788245652652432, + "grad_norm": 2.619712363425748, + "learning_rate": 3.801420875779238e-07, + "loss": 0.3257, + "step": 39925 + }, + { + "epoch": 0.8788465771516619, + "grad_norm": 0.5996567906455332, + "learning_rate": 3.8000576860905356e-07, + "loss": 0.4369, + "step": 39926 + }, + { + "epoch": 0.8788685890380805, + "grad_norm": 4.212712926399029, + "learning_rate": 3.7986947312105336e-07, + "loss": 0.2684, + "step": 39927 + }, + { + "epoch": 0.8788906009244992, + "grad_norm": 1.8977953096280067, + "learning_rate": 3.7973320111461763e-07, + "loss": 0.2582, + "step": 39928 + }, + { + "epoch": 0.8789126128109179, + "grad_norm": 1.7886866272867252, + "learning_rate": 3.795969525904386e-07, + "loss": 0.2973, + "step": 39929 + }, + { + "epoch": 0.8789346246973365, + "grad_norm": 1.3880825568602984, + "learning_rate": 3.7946072754920957e-07, + "loss": 0.3005, + "step": 39930 + }, + { + "epoch": 0.8789566365837552, + "grad_norm": 1.770127420902624, + "learning_rate": 3.793245259916212e-07, + "loss": 0.3171, + "step": 39931 + }, + { + "epoch": 0.8789786484701739, + "grad_norm": 2.3311256551135355, + "learning_rate": 3.7918834791836613e-07, + "loss": 0.2318, + "step": 39932 + }, + { + "epoch": 0.8790006603565925, + "grad_norm": 1.4234320430084901, + "learning_rate": 3.7905219333013777e-07, + "loss": 0.2952, + "step": 39933 + }, + { + "epoch": 0.8790226722430112, + "grad_norm": 0.6394989960116846, + "learning_rate": 3.7891606222762556e-07, + "loss": 0.4652, + "step": 39934 + }, + { + "epoch": 0.8790446841294299, + "grad_norm": 1.7511005511611029, + "learning_rate": 3.7877995461152396e-07, + "loss": 0.1773, + "step": 39935 + }, + { + "epoch": 0.8790666960158485, + "grad_norm": 1.6476249898031643, + "learning_rate": 3.786438704825246e-07, + "loss": 0.2903, + "step": 39936 + }, + { + "epoch": 0.8790887079022672, + "grad_norm": 2.330645851585684, + "learning_rate": 3.7850780984131697e-07, + "loss": 0.2613, + "step": 39937 + }, + { + "epoch": 0.8791107197886859, + "grad_norm": 1.781243124926764, + "learning_rate": 3.7837177268859495e-07, + "loss": 0.4503, + "step": 39938 + }, + { + "epoch": 0.8791327316751045, + "grad_norm": 1.9573978524145974, + "learning_rate": 3.782357590250468e-07, + "loss": 0.259, + "step": 39939 + }, + { + "epoch": 0.8791547435615232, + "grad_norm": 1.8049066678609322, + "learning_rate": 3.780997688513677e-07, + "loss": 0.2696, + "step": 39940 + }, + { + "epoch": 0.8791767554479419, + "grad_norm": 0.6248880017436866, + "learning_rate": 3.779638021682475e-07, + "loss": 0.4506, + "step": 39941 + }, + { + "epoch": 0.8791987673343605, + "grad_norm": 2.5675064152201887, + "learning_rate": 3.778278589763756e-07, + "loss": 0.3009, + "step": 39942 + }, + { + "epoch": 0.8792207792207792, + "grad_norm": 2.084861637072901, + "learning_rate": 3.7769193927644443e-07, + "loss": 0.3358, + "step": 39943 + }, + { + "epoch": 0.8792427911071979, + "grad_norm": 0.6008268125050601, + "learning_rate": 3.775560430691444e-07, + "loss": 0.4574, + "step": 39944 + }, + { + "epoch": 0.8792648029936165, + "grad_norm": 1.4500568768583062, + "learning_rate": 3.7742017035516556e-07, + "loss": 0.2904, + "step": 39945 + }, + { + "epoch": 0.8792868148800352, + "grad_norm": 1.7647436216118144, + "learning_rate": 3.7728432113520007e-07, + "loss": 0.2958, + "step": 39946 + }, + { + "epoch": 0.8793088267664538, + "grad_norm": 1.8775072718480588, + "learning_rate": 3.771484954099369e-07, + "loss": 0.335, + "step": 39947 + }, + { + "epoch": 0.8793308386528725, + "grad_norm": 0.6331169613850632, + "learning_rate": 3.770126931800677e-07, + "loss": 0.4464, + "step": 39948 + }, + { + "epoch": 0.8793528505392912, + "grad_norm": 1.713608022101761, + "learning_rate": 3.768769144462814e-07, + "loss": 0.3495, + "step": 39949 + }, + { + "epoch": 0.8793748624257098, + "grad_norm": 1.3575533817090568, + "learning_rate": 3.767411592092679e-07, + "loss": 0.2595, + "step": 39950 + }, + { + "epoch": 0.8793968743121285, + "grad_norm": 0.6151625802814258, + "learning_rate": 3.7660542746971896e-07, + "loss": 0.4608, + "step": 39951 + }, + { + "epoch": 0.8794188861985472, + "grad_norm": 2.082627661225397, + "learning_rate": 3.764697192283229e-07, + "loss": 0.3155, + "step": 39952 + }, + { + "epoch": 0.8794408980849658, + "grad_norm": 0.6255082384423131, + "learning_rate": 3.763340344857702e-07, + "loss": 0.4713, + "step": 39953 + }, + { + "epoch": 0.8794629099713845, + "grad_norm": 1.774455023586636, + "learning_rate": 3.7619837324274934e-07, + "loss": 0.2484, + "step": 39954 + }, + { + "epoch": 0.8794849218578032, + "grad_norm": 1.6476716718177482, + "learning_rate": 3.760627354999508e-07, + "loss": 0.249, + "step": 39955 + }, + { + "epoch": 0.8795069337442218, + "grad_norm": 0.6086469756925826, + "learning_rate": 3.759271212580645e-07, + "loss": 0.4438, + "step": 39956 + }, + { + "epoch": 0.8795289456306405, + "grad_norm": 0.6215666091296114, + "learning_rate": 3.7579153051777837e-07, + "loss": 0.4539, + "step": 39957 + }, + { + "epoch": 0.8795509575170593, + "grad_norm": 0.6158744277851964, + "learning_rate": 3.756559632797824e-07, + "loss": 0.4559, + "step": 39958 + }, + { + "epoch": 0.8795729694034778, + "grad_norm": 1.8753161906100273, + "learning_rate": 3.755204195447654e-07, + "loss": 0.2519, + "step": 39959 + }, + { + "epoch": 0.8795949812898965, + "grad_norm": 1.5448242441007, + "learning_rate": 3.7538489931341516e-07, + "loss": 0.3596, + "step": 39960 + }, + { + "epoch": 0.8796169931763153, + "grad_norm": 1.522537259343658, + "learning_rate": 3.752494025864234e-07, + "loss": 0.2922, + "step": 39961 + }, + { + "epoch": 0.8796390050627338, + "grad_norm": 1.6243327766683133, + "learning_rate": 3.751139293644762e-07, + "loss": 0.3777, + "step": 39962 + }, + { + "epoch": 0.8796610169491526, + "grad_norm": 1.565404845945351, + "learning_rate": 3.749784796482619e-07, + "loss": 0.2807, + "step": 39963 + }, + { + "epoch": 0.8796830288355713, + "grad_norm": 0.6208375606277517, + "learning_rate": 3.7484305343847004e-07, + "loss": 0.4451, + "step": 39964 + }, + { + "epoch": 0.8797050407219899, + "grad_norm": 0.6222906637563748, + "learning_rate": 3.747076507357883e-07, + "loss": 0.4683, + "step": 39965 + }, + { + "epoch": 0.8797270526084086, + "grad_norm": 0.6025794635180132, + "learning_rate": 3.745722715409067e-07, + "loss": 0.4286, + "step": 39966 + }, + { + "epoch": 0.8797490644948273, + "grad_norm": 1.8070558894232451, + "learning_rate": 3.7443691585451134e-07, + "loss": 0.3529, + "step": 39967 + }, + { + "epoch": 0.8797710763812459, + "grad_norm": 2.7267186990906285, + "learning_rate": 3.7430158367729007e-07, + "loss": 0.2806, + "step": 39968 + }, + { + "epoch": 0.8797930882676646, + "grad_norm": 0.6199029767119576, + "learning_rate": 3.7416627500993173e-07, + "loss": 0.4576, + "step": 39969 + }, + { + "epoch": 0.8798151001540832, + "grad_norm": 2.743815133761253, + "learning_rate": 3.740309898531219e-07, + "loss": 0.3145, + "step": 39970 + }, + { + "epoch": 0.8798371120405019, + "grad_norm": 2.5165297887621705, + "learning_rate": 3.738957282075517e-07, + "loss": 0.2149, + "step": 39971 + }, + { + "epoch": 0.8798591239269206, + "grad_norm": 1.9915197121776913, + "learning_rate": 3.737604900739067e-07, + "loss": 0.2118, + "step": 39972 + }, + { + "epoch": 0.8798811358133392, + "grad_norm": 1.3985337371420654, + "learning_rate": 3.7362527545287353e-07, + "loss": 0.2915, + "step": 39973 + }, + { + "epoch": 0.8799031476997579, + "grad_norm": 2.2429047451689095, + "learning_rate": 3.7349008434514113e-07, + "loss": 0.3755, + "step": 39974 + }, + { + "epoch": 0.8799251595861766, + "grad_norm": 2.945153368232951, + "learning_rate": 3.7335491675139333e-07, + "loss": 0.2351, + "step": 39975 + }, + { + "epoch": 0.8799471714725952, + "grad_norm": 1.3885725067958932, + "learning_rate": 3.7321977267232136e-07, + "loss": 0.2873, + "step": 39976 + }, + { + "epoch": 0.8799691833590139, + "grad_norm": 2.2428613514685405, + "learning_rate": 3.730846521086101e-07, + "loss": 0.3637, + "step": 39977 + }, + { + "epoch": 0.8799911952454326, + "grad_norm": 1.7466002230265691, + "learning_rate": 3.7294955506094576e-07, + "loss": 0.2369, + "step": 39978 + }, + { + "epoch": 0.8800132071318512, + "grad_norm": 1.664208090868954, + "learning_rate": 3.728144815300161e-07, + "loss": 0.2074, + "step": 39979 + }, + { + "epoch": 0.8800352190182699, + "grad_norm": 1.9912766494153606, + "learning_rate": 3.726794315165061e-07, + "loss": 0.3797, + "step": 39980 + }, + { + "epoch": 0.8800572309046886, + "grad_norm": 5.791087513392679, + "learning_rate": 3.72544405021103e-07, + "loss": 0.3079, + "step": 39981 + }, + { + "epoch": 0.8800792427911072, + "grad_norm": 1.4377140158793629, + "learning_rate": 3.7240940204449404e-07, + "loss": 0.3146, + "step": 39982 + }, + { + "epoch": 0.8801012546775259, + "grad_norm": 2.559809871121079, + "learning_rate": 3.722744225873637e-07, + "loss": 0.3553, + "step": 39983 + }, + { + "epoch": 0.8801232665639446, + "grad_norm": 0.6205839371811567, + "learning_rate": 3.7213946665039967e-07, + "loss": 0.4464, + "step": 39984 + }, + { + "epoch": 0.8801452784503632, + "grad_norm": 2.398788060258593, + "learning_rate": 3.7200453423428596e-07, + "loss": 0.3041, + "step": 39985 + }, + { + "epoch": 0.8801672903367819, + "grad_norm": 0.6126103012585962, + "learning_rate": 3.7186962533970973e-07, + "loss": 0.4264, + "step": 39986 + }, + { + "epoch": 0.8801893022232006, + "grad_norm": 1.6712115981470488, + "learning_rate": 3.7173473996735656e-07, + "loss": 0.3694, + "step": 39987 + }, + { + "epoch": 0.8802113141096192, + "grad_norm": 1.8814257052051935, + "learning_rate": 3.7159987811791143e-07, + "loss": 0.2217, + "step": 39988 + }, + { + "epoch": 0.8802333259960379, + "grad_norm": 2.3317185524208317, + "learning_rate": 3.714650397920605e-07, + "loss": 0.2179, + "step": 39989 + }, + { + "epoch": 0.8802553378824566, + "grad_norm": 0.6471642161005646, + "learning_rate": 3.7133022499048766e-07, + "loss": 0.4581, + "step": 39990 + }, + { + "epoch": 0.8802773497688752, + "grad_norm": 2.6496236455509603, + "learning_rate": 3.71195433713879e-07, + "loss": 0.354, + "step": 39991 + }, + { + "epoch": 0.8802993616552939, + "grad_norm": 1.990128849450165, + "learning_rate": 3.710606659629212e-07, + "loss": 0.3304, + "step": 39992 + }, + { + "epoch": 0.8803213735417125, + "grad_norm": 1.4376550239279196, + "learning_rate": 3.70925921738296e-07, + "loss": 0.3326, + "step": 39993 + }, + { + "epoch": 0.8803433854281312, + "grad_norm": 0.6164417143237736, + "learning_rate": 3.707912010406911e-07, + "loss": 0.4322, + "step": 39994 + }, + { + "epoch": 0.8803653973145499, + "grad_norm": 0.6117686949337305, + "learning_rate": 3.7065650387078934e-07, + "loss": 0.4674, + "step": 39995 + }, + { + "epoch": 0.8803874092009685, + "grad_norm": 1.8376726430019774, + "learning_rate": 3.705218302292757e-07, + "loss": 0.3487, + "step": 39996 + }, + { + "epoch": 0.8804094210873872, + "grad_norm": 0.6008056303859283, + "learning_rate": 3.703871801168357e-07, + "loss": 0.4294, + "step": 39997 + }, + { + "epoch": 0.8804314329738059, + "grad_norm": 0.6181433311914035, + "learning_rate": 3.7025255353415223e-07, + "loss": 0.4422, + "step": 39998 + }, + { + "epoch": 0.8804534448602245, + "grad_norm": 0.636583540412837, + "learning_rate": 3.7011795048191136e-07, + "loss": 0.4468, + "step": 39999 + }, + { + "epoch": 0.8804754567466432, + "grad_norm": 2.630435892651278, + "learning_rate": 3.6998337096079474e-07, + "loss": 0.3687, + "step": 40000 + }, + { + "epoch": 0.8804974686330619, + "grad_norm": 0.6296382048675759, + "learning_rate": 3.6984881497148795e-07, + "loss": 0.4835, + "step": 40001 + }, + { + "epoch": 0.8805194805194805, + "grad_norm": 1.8169915802238583, + "learning_rate": 3.697142825146749e-07, + "loss": 0.2453, + "step": 40002 + }, + { + "epoch": 0.8805414924058992, + "grad_norm": 0.6215250028832371, + "learning_rate": 3.6957977359103945e-07, + "loss": 0.43, + "step": 40003 + }, + { + "epoch": 0.8805635042923179, + "grad_norm": 4.3953253788980025, + "learning_rate": 3.694452882012645e-07, + "loss": 0.2413, + "step": 40004 + }, + { + "epoch": 0.8805855161787365, + "grad_norm": 1.8366005677656283, + "learning_rate": 3.6931082634603376e-07, + "loss": 0.2846, + "step": 40005 + }, + { + "epoch": 0.8806075280651552, + "grad_norm": 2.187095737737109, + "learning_rate": 3.6917638802602964e-07, + "loss": 0.2673, + "step": 40006 + }, + { + "epoch": 0.8806295399515739, + "grad_norm": 1.540258857133351, + "learning_rate": 3.690419732419381e-07, + "loss": 0.3149, + "step": 40007 + }, + { + "epoch": 0.8806515518379925, + "grad_norm": 1.442570436901822, + "learning_rate": 3.6890758199444045e-07, + "loss": 0.22, + "step": 40008 + }, + { + "epoch": 0.8806735637244112, + "grad_norm": 0.6267368096748488, + "learning_rate": 3.687732142842193e-07, + "loss": 0.4741, + "step": 40009 + }, + { + "epoch": 0.8806955756108299, + "grad_norm": 1.654349166432212, + "learning_rate": 3.686388701119592e-07, + "loss": 0.367, + "step": 40010 + }, + { + "epoch": 0.8807175874972485, + "grad_norm": 3.7880143587018558, + "learning_rate": 3.685045494783407e-07, + "loss": 0.2345, + "step": 40011 + }, + { + "epoch": 0.8807395993836672, + "grad_norm": 2.7391830823684864, + "learning_rate": 3.683702523840488e-07, + "loss": 0.276, + "step": 40012 + }, + { + "epoch": 0.8807616112700859, + "grad_norm": 4.0779418457461, + "learning_rate": 3.6823597882976516e-07, + "loss": 0.2672, + "step": 40013 + }, + { + "epoch": 0.8807836231565045, + "grad_norm": 2.065230112530102, + "learning_rate": 3.6810172881617145e-07, + "loss": 0.3013, + "step": 40014 + }, + { + "epoch": 0.8808056350429232, + "grad_norm": 1.962883462967035, + "learning_rate": 3.679675023439516e-07, + "loss": 0.2746, + "step": 40015 + }, + { + "epoch": 0.8808276469293418, + "grad_norm": 1.8991681580998234, + "learning_rate": 3.6783329941378564e-07, + "loss": 0.2796, + "step": 40016 + }, + { + "epoch": 0.8808496588157605, + "grad_norm": 3.6734069294490723, + "learning_rate": 3.676991200263569e-07, + "loss": 0.2625, + "step": 40017 + }, + { + "epoch": 0.8808716707021792, + "grad_norm": 1.5524323290582955, + "learning_rate": 3.6756496418234867e-07, + "loss": 0.2874, + "step": 40018 + }, + { + "epoch": 0.8808936825885978, + "grad_norm": 1.543352025085574, + "learning_rate": 3.674308318824399e-07, + "loss": 0.2388, + "step": 40019 + }, + { + "epoch": 0.8809156944750165, + "grad_norm": 1.4630303259063104, + "learning_rate": 3.67296723127315e-07, + "loss": 0.2897, + "step": 40020 + }, + { + "epoch": 0.8809377063614352, + "grad_norm": 0.6239994817155916, + "learning_rate": 3.671626379176535e-07, + "loss": 0.4435, + "step": 40021 + }, + { + "epoch": 0.8809597182478538, + "grad_norm": 2.0651110945941893, + "learning_rate": 3.6702857625413815e-07, + "loss": 0.3152, + "step": 40022 + }, + { + "epoch": 0.8809817301342725, + "grad_norm": 1.5005890903723846, + "learning_rate": 3.6689453813745e-07, + "loss": 0.2682, + "step": 40023 + }, + { + "epoch": 0.8810037420206912, + "grad_norm": 1.611589239480484, + "learning_rate": 3.667605235682697e-07, + "loss": 0.3479, + "step": 40024 + }, + { + "epoch": 0.8810257539071098, + "grad_norm": 0.6139813009537872, + "learning_rate": 3.6662653254728007e-07, + "loss": 0.4506, + "step": 40025 + }, + { + "epoch": 0.8810477657935285, + "grad_norm": 1.5381990679753381, + "learning_rate": 3.6649256507515997e-07, + "loss": 0.3181, + "step": 40026 + }, + { + "epoch": 0.8810697776799472, + "grad_norm": 2.0223264024752137, + "learning_rate": 3.66358621152591e-07, + "loss": 0.3533, + "step": 40027 + }, + { + "epoch": 0.8810917895663658, + "grad_norm": 1.816060031365471, + "learning_rate": 3.662247007802555e-07, + "loss": 0.2977, + "step": 40028 + }, + { + "epoch": 0.8811138014527845, + "grad_norm": 1.7995002713359407, + "learning_rate": 3.660908039588318e-07, + "loss": 0.2617, + "step": 40029 + }, + { + "epoch": 0.8811358133392032, + "grad_norm": 1.4798894412749592, + "learning_rate": 3.6595693068900207e-07, + "loss": 0.264, + "step": 40030 + }, + { + "epoch": 0.8811578252256218, + "grad_norm": 0.6001639710693257, + "learning_rate": 3.6582308097144523e-07, + "loss": 0.4249, + "step": 40031 + }, + { + "epoch": 0.8811798371120405, + "grad_norm": 1.473891764024779, + "learning_rate": 3.656892548068419e-07, + "loss": 0.2383, + "step": 40032 + }, + { + "epoch": 0.8812018489984592, + "grad_norm": 1.9667087995516759, + "learning_rate": 3.6555545219587427e-07, + "loss": 0.3403, + "step": 40033 + }, + { + "epoch": 0.8812238608848778, + "grad_norm": 4.0667875283339505, + "learning_rate": 3.6542167313922017e-07, + "loss": 0.2982, + "step": 40034 + }, + { + "epoch": 0.8812458727712965, + "grad_norm": 1.9754390087107656, + "learning_rate": 3.6528791763756067e-07, + "loss": 0.3066, + "step": 40035 + }, + { + "epoch": 0.8812678846577152, + "grad_norm": 4.217269482722668, + "learning_rate": 3.6515418569157413e-07, + "loss": 0.3002, + "step": 40036 + }, + { + "epoch": 0.8812898965441338, + "grad_norm": 0.6077081425127941, + "learning_rate": 3.650204773019411e-07, + "loss": 0.4599, + "step": 40037 + }, + { + "epoch": 0.8813119084305525, + "grad_norm": 2.340969153874248, + "learning_rate": 3.648867924693428e-07, + "loss": 0.3147, + "step": 40038 + }, + { + "epoch": 0.8813339203169712, + "grad_norm": 1.8619933445484433, + "learning_rate": 3.6475313119445575e-07, + "loss": 0.3042, + "step": 40039 + }, + { + "epoch": 0.8813559322033898, + "grad_norm": 0.5958395007098481, + "learning_rate": 3.646194934779618e-07, + "loss": 0.4493, + "step": 40040 + }, + { + "epoch": 0.8813779440898085, + "grad_norm": 1.693668235290989, + "learning_rate": 3.644858793205386e-07, + "loss": 0.3524, + "step": 40041 + }, + { + "epoch": 0.8813999559762271, + "grad_norm": 0.620882388727137, + "learning_rate": 3.643522887228645e-07, + "loss": 0.4573, + "step": 40042 + }, + { + "epoch": 0.8814219678626458, + "grad_norm": 0.6094295589006987, + "learning_rate": 3.642187216856208e-07, + "loss": 0.4466, + "step": 40043 + }, + { + "epoch": 0.8814439797490645, + "grad_norm": 1.3493769786574215, + "learning_rate": 3.6408517820948565e-07, + "loss": 0.2532, + "step": 40044 + }, + { + "epoch": 0.8814659916354831, + "grad_norm": 2.224668389619281, + "learning_rate": 3.6395165829513635e-07, + "loss": 0.2782, + "step": 40045 + }, + { + "epoch": 0.8814880035219018, + "grad_norm": 1.5414205040415956, + "learning_rate": 3.638181619432529e-07, + "loss": 0.2835, + "step": 40046 + }, + { + "epoch": 0.8815100154083205, + "grad_norm": 2.964416929165514, + "learning_rate": 3.6368468915451316e-07, + "loss": 0.3049, + "step": 40047 + }, + { + "epoch": 0.8815320272947391, + "grad_norm": 1.7354852846435176, + "learning_rate": 3.6355123992959483e-07, + "loss": 0.3079, + "step": 40048 + }, + { + "epoch": 0.8815540391811578, + "grad_norm": 2.176026062980532, + "learning_rate": 3.6341781426917857e-07, + "loss": 0.282, + "step": 40049 + }, + { + "epoch": 0.8815760510675765, + "grad_norm": 1.306969910899136, + "learning_rate": 3.6328441217393983e-07, + "loss": 0.3156, + "step": 40050 + }, + { + "epoch": 0.8815980629539951, + "grad_norm": 6.375312835952347, + "learning_rate": 3.6315103364455873e-07, + "loss": 0.2556, + "step": 40051 + }, + { + "epoch": 0.8816200748404138, + "grad_norm": 0.5982522305638364, + "learning_rate": 3.6301767868171136e-07, + "loss": 0.4364, + "step": 40052 + }, + { + "epoch": 0.8816420867268325, + "grad_norm": 1.3468891526288491, + "learning_rate": 3.6288434728607603e-07, + "loss": 0.3556, + "step": 40053 + }, + { + "epoch": 0.8816640986132511, + "grad_norm": 1.755508733532735, + "learning_rate": 3.627510394583317e-07, + "loss": 0.2857, + "step": 40054 + }, + { + "epoch": 0.8816861104996698, + "grad_norm": 3.283376502488481, + "learning_rate": 3.6261775519915387e-07, + "loss": 0.2569, + "step": 40055 + }, + { + "epoch": 0.8817081223860885, + "grad_norm": 0.6135328880670963, + "learning_rate": 3.6248449450922207e-07, + "loss": 0.4612, + "step": 40056 + }, + { + "epoch": 0.8817301342725071, + "grad_norm": 1.3964009689390111, + "learning_rate": 3.623512573892113e-07, + "loss": 0.1679, + "step": 40057 + }, + { + "epoch": 0.8817521461589258, + "grad_norm": 8.323051328725198, + "learning_rate": 3.6221804383979983e-07, + "loss": 0.2338, + "step": 40058 + }, + { + "epoch": 0.8817741580453445, + "grad_norm": 1.286886928957506, + "learning_rate": 3.6208485386166613e-07, + "loss": 0.2371, + "step": 40059 + }, + { + "epoch": 0.8817961699317631, + "grad_norm": 0.5893142472145273, + "learning_rate": 3.6195168745548457e-07, + "loss": 0.4349, + "step": 40060 + }, + { + "epoch": 0.8818181818181818, + "grad_norm": 1.2835870069109634, + "learning_rate": 3.618185446219336e-07, + "loss": 0.2634, + "step": 40061 + }, + { + "epoch": 0.8818401937046005, + "grad_norm": 1.7501925086864134, + "learning_rate": 3.6168542536168914e-07, + "loss": 0.2968, + "step": 40062 + }, + { + "epoch": 0.8818622055910191, + "grad_norm": 0.6237868153776926, + "learning_rate": 3.615523296754281e-07, + "loss": 0.4539, + "step": 40063 + }, + { + "epoch": 0.8818842174774378, + "grad_norm": 1.786772326698064, + "learning_rate": 3.6141925756382756e-07, + "loss": 0.2509, + "step": 40064 + }, + { + "epoch": 0.8819062293638564, + "grad_norm": 0.6474098140833282, + "learning_rate": 3.612862090275626e-07, + "loss": 0.4525, + "step": 40065 + }, + { + "epoch": 0.8819282412502751, + "grad_norm": 2.8769121621984923, + "learning_rate": 3.6115318406731105e-07, + "loss": 0.3469, + "step": 40066 + }, + { + "epoch": 0.8819502531366938, + "grad_norm": 1.6852110151116624, + "learning_rate": 3.610201826837467e-07, + "loss": 0.2478, + "step": 40067 + }, + { + "epoch": 0.8819722650231124, + "grad_norm": 12.27006402692472, + "learning_rate": 3.608872048775469e-07, + "loss": 0.3063, + "step": 40068 + }, + { + "epoch": 0.8819942769095311, + "grad_norm": 1.3809785166314725, + "learning_rate": 3.607542506493883e-07, + "loss": 0.2511, + "step": 40069 + }, + { + "epoch": 0.8820162887959498, + "grad_norm": 1.4503677030225455, + "learning_rate": 3.606213199999453e-07, + "loss": 0.2796, + "step": 40070 + }, + { + "epoch": 0.8820383006823684, + "grad_norm": 1.7691259177443708, + "learning_rate": 3.6048841292989413e-07, + "loss": 0.3408, + "step": 40071 + }, + { + "epoch": 0.8820603125687871, + "grad_norm": 4.0646992178376955, + "learning_rate": 3.603555294399097e-07, + "loss": 0.2957, + "step": 40072 + }, + { + "epoch": 0.8820823244552058, + "grad_norm": 0.6040963586863656, + "learning_rate": 3.602226695306682e-07, + "loss": 0.4646, + "step": 40073 + }, + { + "epoch": 0.8821043363416244, + "grad_norm": 1.435304516020368, + "learning_rate": 3.600898332028446e-07, + "loss": 0.2879, + "step": 40074 + }, + { + "epoch": 0.8821263482280431, + "grad_norm": 1.8837546065147517, + "learning_rate": 3.5995702045711344e-07, + "loss": 0.3575, + "step": 40075 + }, + { + "epoch": 0.8821483601144618, + "grad_norm": 1.4235645808699946, + "learning_rate": 3.5982423129415077e-07, + "loss": 0.2138, + "step": 40076 + }, + { + "epoch": 0.8821703720008804, + "grad_norm": 1.433300745932593, + "learning_rate": 3.5969146571463167e-07, + "loss": 0.3272, + "step": 40077 + }, + { + "epoch": 0.8821923838872991, + "grad_norm": 1.4044169436745084, + "learning_rate": 3.5955872371922774e-07, + "loss": 0.2229, + "step": 40078 + }, + { + "epoch": 0.8822143957737179, + "grad_norm": 1.5478620211813443, + "learning_rate": 3.5942600530861793e-07, + "loss": 0.34, + "step": 40079 + }, + { + "epoch": 0.8822364076601364, + "grad_norm": 2.1721071305421447, + "learning_rate": 3.5929331048347447e-07, + "loss": 0.246, + "step": 40080 + }, + { + "epoch": 0.8822584195465552, + "grad_norm": 2.2406724520599357, + "learning_rate": 3.5916063924447243e-07, + "loss": 0.2929, + "step": 40081 + }, + { + "epoch": 0.8822804314329739, + "grad_norm": 1.2503151974921203, + "learning_rate": 3.5902799159228617e-07, + "loss": 0.2881, + "step": 40082 + }, + { + "epoch": 0.8823024433193924, + "grad_norm": 1.8378866932333215, + "learning_rate": 3.5889536752758913e-07, + "loss": 0.284, + "step": 40083 + }, + { + "epoch": 0.8823244552058112, + "grad_norm": 1.2216291269389923, + "learning_rate": 3.5876276705105517e-07, + "loss": 0.2235, + "step": 40084 + }, + { + "epoch": 0.8823464670922299, + "grad_norm": 0.5819054522204914, + "learning_rate": 3.5863019016336045e-07, + "loss": 0.4179, + "step": 40085 + }, + { + "epoch": 0.8823684789786485, + "grad_norm": 0.6191975517277035, + "learning_rate": 3.584976368651755e-07, + "loss": 0.4365, + "step": 40086 + }, + { + "epoch": 0.8823904908650672, + "grad_norm": 1.4942743696646557, + "learning_rate": 3.58365107157177e-07, + "loss": 0.3133, + "step": 40087 + }, + { + "epoch": 0.8824125027514857, + "grad_norm": 1.5149401500740127, + "learning_rate": 3.5823260104003666e-07, + "loss": 0.3153, + "step": 40088 + }, + { + "epoch": 0.8824345146379045, + "grad_norm": 1.7703621069563265, + "learning_rate": 3.581001185144284e-07, + "loss": 0.2942, + "step": 40089 + }, + { + "epoch": 0.8824565265243232, + "grad_norm": 1.3737284452867708, + "learning_rate": 3.5796765958102606e-07, + "loss": 0.2856, + "step": 40090 + }, + { + "epoch": 0.8824785384107418, + "grad_norm": 1.4797661821075119, + "learning_rate": 3.57835224240502e-07, + "loss": 0.2823, + "step": 40091 + }, + { + "epoch": 0.8825005502971605, + "grad_norm": 2.9170070542826774, + "learning_rate": 3.577028124935311e-07, + "loss": 0.2841, + "step": 40092 + }, + { + "epoch": 0.8825225621835792, + "grad_norm": 1.7359401296526862, + "learning_rate": 3.575704243407835e-07, + "loss": 0.3045, + "step": 40093 + }, + { + "epoch": 0.8825445740699978, + "grad_norm": 2.871796083841612, + "learning_rate": 3.5743805978293357e-07, + "loss": 0.3078, + "step": 40094 + }, + { + "epoch": 0.8825665859564165, + "grad_norm": 1.488312929779593, + "learning_rate": 3.573057188206552e-07, + "loss": 0.3843, + "step": 40095 + }, + { + "epoch": 0.8825885978428352, + "grad_norm": 1.8813208055391624, + "learning_rate": 3.571734014546191e-07, + "loss": 0.2595, + "step": 40096 + }, + { + "epoch": 0.8826106097292538, + "grad_norm": 1.722976721609315, + "learning_rate": 3.57041107685499e-07, + "loss": 0.24, + "step": 40097 + }, + { + "epoch": 0.8826326216156725, + "grad_norm": 1.6566292488168084, + "learning_rate": 3.569088375139662e-07, + "loss": 0.2897, + "step": 40098 + }, + { + "epoch": 0.8826546335020912, + "grad_norm": 1.5618272942617915, + "learning_rate": 3.5677659094069395e-07, + "loss": 0.2844, + "step": 40099 + }, + { + "epoch": 0.8826766453885098, + "grad_norm": 0.596991727720078, + "learning_rate": 3.566443679663545e-07, + "loss": 0.4596, + "step": 40100 + }, + { + "epoch": 0.8826986572749285, + "grad_norm": 2.1434291816664537, + "learning_rate": 3.565121685916184e-07, + "loss": 0.33, + "step": 40101 + }, + { + "epoch": 0.8827206691613472, + "grad_norm": 2.626042997319641, + "learning_rate": 3.563799928171596e-07, + "loss": 0.3338, + "step": 40102 + }, + { + "epoch": 0.8827426810477658, + "grad_norm": 1.9162370907121857, + "learning_rate": 3.562478406436476e-07, + "loss": 0.3505, + "step": 40103 + }, + { + "epoch": 0.8827646929341845, + "grad_norm": 1.8563521814994322, + "learning_rate": 3.5611571207175565e-07, + "loss": 0.2487, + "step": 40104 + }, + { + "epoch": 0.8827867048206032, + "grad_norm": 0.6066663930508411, + "learning_rate": 3.559836071021555e-07, + "loss": 0.4464, + "step": 40105 + }, + { + "epoch": 0.8828087167070218, + "grad_norm": 6.335585459285359, + "learning_rate": 3.5585152573551664e-07, + "loss": 0.29, + "step": 40106 + }, + { + "epoch": 0.8828307285934405, + "grad_norm": 1.3268491001947063, + "learning_rate": 3.5571946797251345e-07, + "loss": 0.2151, + "step": 40107 + }, + { + "epoch": 0.8828527404798592, + "grad_norm": 1.9304694706474599, + "learning_rate": 3.5558743381381377e-07, + "loss": 0.2659, + "step": 40108 + }, + { + "epoch": 0.8828747523662778, + "grad_norm": 1.5842840389301687, + "learning_rate": 3.554554232600904e-07, + "loss": 0.249, + "step": 40109 + }, + { + "epoch": 0.8828967642526965, + "grad_norm": 2.7448579902692725, + "learning_rate": 3.5532343631201507e-07, + "loss": 0.3564, + "step": 40110 + }, + { + "epoch": 0.8829187761391151, + "grad_norm": 0.6165681632547593, + "learning_rate": 3.551914729702566e-07, + "loss": 0.4615, + "step": 40111 + }, + { + "epoch": 0.8829407880255338, + "grad_norm": 3.6111506843644734, + "learning_rate": 3.550595332354878e-07, + "loss": 0.2692, + "step": 40112 + }, + { + "epoch": 0.8829627999119525, + "grad_norm": 2.546513724614901, + "learning_rate": 3.549276171083782e-07, + "loss": 0.3093, + "step": 40113 + }, + { + "epoch": 0.8829848117983711, + "grad_norm": 2.138324075288288, + "learning_rate": 3.547957245895961e-07, + "loss": 0.2404, + "step": 40114 + }, + { + "epoch": 0.8830068236847898, + "grad_norm": 1.536510520052598, + "learning_rate": 3.5466385567981597e-07, + "loss": 0.2741, + "step": 40115 + }, + { + "epoch": 0.8830288355712085, + "grad_norm": 0.6015360815630245, + "learning_rate": 3.545320103797051e-07, + "loss": 0.4422, + "step": 40116 + }, + { + "epoch": 0.8830508474576271, + "grad_norm": 1.8755170093466607, + "learning_rate": 3.544001886899351e-07, + "loss": 0.2275, + "step": 40117 + }, + { + "epoch": 0.8830728593440458, + "grad_norm": 1.2580799203394475, + "learning_rate": 3.5426839061117545e-07, + "loss": 0.2812, + "step": 40118 + }, + { + "epoch": 0.8830948712304645, + "grad_norm": 0.6112361249940417, + "learning_rate": 3.541366161440951e-07, + "loss": 0.4201, + "step": 40119 + }, + { + "epoch": 0.8831168831168831, + "grad_norm": 0.586539505107142, + "learning_rate": 3.540048652893652e-07, + "loss": 0.4422, + "step": 40120 + }, + { + "epoch": 0.8831388950033018, + "grad_norm": 2.6115937062351837, + "learning_rate": 3.53873138047654e-07, + "loss": 0.2488, + "step": 40121 + }, + { + "epoch": 0.8831609068897205, + "grad_norm": 2.006621437298568, + "learning_rate": 3.537414344196327e-07, + "loss": 0.3463, + "step": 40122 + }, + { + "epoch": 0.8831829187761391, + "grad_norm": 1.6511788416553812, + "learning_rate": 3.5360975440596967e-07, + "loss": 0.3206, + "step": 40123 + }, + { + "epoch": 0.8832049306625578, + "grad_norm": 1.7552883537140151, + "learning_rate": 3.5347809800733324e-07, + "loss": 0.2616, + "step": 40124 + }, + { + "epoch": 0.8832269425489765, + "grad_norm": 2.322490062409961, + "learning_rate": 3.5334646522439397e-07, + "loss": 0.2963, + "step": 40125 + }, + { + "epoch": 0.8832489544353951, + "grad_norm": 1.5788369195566363, + "learning_rate": 3.5321485605782134e-07, + "loss": 0.3057, + "step": 40126 + }, + { + "epoch": 0.8832709663218138, + "grad_norm": 1.8956487489959852, + "learning_rate": 3.5308327050828206e-07, + "loss": 0.2936, + "step": 40127 + }, + { + "epoch": 0.8832929782082325, + "grad_norm": 1.610381965987796, + "learning_rate": 3.529517085764472e-07, + "loss": 0.331, + "step": 40128 + }, + { + "epoch": 0.8833149900946511, + "grad_norm": 1.9978791195408951, + "learning_rate": 3.5282017026298355e-07, + "loss": 0.2791, + "step": 40129 + }, + { + "epoch": 0.8833370019810698, + "grad_norm": 1.4558180187910117, + "learning_rate": 3.5268865556856045e-07, + "loss": 0.2575, + "step": 40130 + }, + { + "epoch": 0.8833590138674885, + "grad_norm": 0.6062292158979034, + "learning_rate": 3.525571644938469e-07, + "loss": 0.4414, + "step": 40131 + }, + { + "epoch": 0.8833810257539071, + "grad_norm": 12.126818323528251, + "learning_rate": 3.524256970395101e-07, + "loss": 0.3863, + "step": 40132 + }, + { + "epoch": 0.8834030376403258, + "grad_norm": 0.6274955198793964, + "learning_rate": 3.5229425320622013e-07, + "loss": 0.4602, + "step": 40133 + }, + { + "epoch": 0.8834250495267444, + "grad_norm": 1.7363860149511798, + "learning_rate": 3.5216283299464194e-07, + "loss": 0.292, + "step": 40134 + }, + { + "epoch": 0.8834470614131631, + "grad_norm": 1.5871779572156515, + "learning_rate": 3.520314364054461e-07, + "loss": 0.2486, + "step": 40135 + }, + { + "epoch": 0.8834690732995818, + "grad_norm": 1.3640084277856643, + "learning_rate": 3.5190006343929996e-07, + "loss": 0.2459, + "step": 40136 + }, + { + "epoch": 0.8834910851860004, + "grad_norm": 1.7646938923912503, + "learning_rate": 3.517687140968701e-07, + "loss": 0.3565, + "step": 40137 + }, + { + "epoch": 0.8835130970724191, + "grad_norm": 0.6043749813392327, + "learning_rate": 3.516373883788254e-07, + "loss": 0.4444, + "step": 40138 + }, + { + "epoch": 0.8835351089588378, + "grad_norm": 1.9031303925752963, + "learning_rate": 3.5150608628583163e-07, + "loss": 0.3013, + "step": 40139 + }, + { + "epoch": 0.8835571208452564, + "grad_norm": 2.132061413133252, + "learning_rate": 3.5137480781855805e-07, + "loss": 0.2897, + "step": 40140 + }, + { + "epoch": 0.8835791327316751, + "grad_norm": 6.2659956238531445, + "learning_rate": 3.5124355297767146e-07, + "loss": 0.3321, + "step": 40141 + }, + { + "epoch": 0.8836011446180938, + "grad_norm": 0.5969425560332023, + "learning_rate": 3.5111232176383735e-07, + "loss": 0.4101, + "step": 40142 + }, + { + "epoch": 0.8836231565045124, + "grad_norm": 2.298483056044917, + "learning_rate": 3.509811141777253e-07, + "loss": 0.2946, + "step": 40143 + }, + { + "epoch": 0.8836451683909311, + "grad_norm": 4.64823103080727, + "learning_rate": 3.5084993022000024e-07, + "loss": 0.2671, + "step": 40144 + }, + { + "epoch": 0.8836671802773498, + "grad_norm": 1.821959298289559, + "learning_rate": 3.507187698913278e-07, + "loss": 0.2256, + "step": 40145 + }, + { + "epoch": 0.8836891921637684, + "grad_norm": 1.4403151729685815, + "learning_rate": 3.5058763319237855e-07, + "loss": 0.2358, + "step": 40146 + }, + { + "epoch": 0.8837112040501871, + "grad_norm": 1.8771435482384542, + "learning_rate": 3.5045652012381526e-07, + "loss": 0.3368, + "step": 40147 + }, + { + "epoch": 0.8837332159366058, + "grad_norm": 1.4197039574905217, + "learning_rate": 3.503254306863063e-07, + "loss": 0.2434, + "step": 40148 + }, + { + "epoch": 0.8837552278230244, + "grad_norm": 1.6126344230897447, + "learning_rate": 3.5019436488051784e-07, + "loss": 0.3127, + "step": 40149 + }, + { + "epoch": 0.8837772397094431, + "grad_norm": 1.7313862584998343, + "learning_rate": 3.500633227071132e-07, + "loss": 0.2549, + "step": 40150 + }, + { + "epoch": 0.8837992515958618, + "grad_norm": 0.6343795427128197, + "learning_rate": 3.4993230416676295e-07, + "loss": 0.4473, + "step": 40151 + }, + { + "epoch": 0.8838212634822804, + "grad_norm": 2.093931753565363, + "learning_rate": 3.498013092601299e-07, + "loss": 0.3265, + "step": 40152 + }, + { + "epoch": 0.8838432753686991, + "grad_norm": 0.6115855406508997, + "learning_rate": 3.4967033798788075e-07, + "loss": 0.4389, + "step": 40153 + }, + { + "epoch": 0.8838652872551178, + "grad_norm": 2.3999934681238253, + "learning_rate": 3.4953939035068165e-07, + "loss": 0.4377, + "step": 40154 + }, + { + "epoch": 0.8838872991415364, + "grad_norm": 1.64785349881729, + "learning_rate": 3.49408466349197e-07, + "loss": 0.2429, + "step": 40155 + }, + { + "epoch": 0.8839093110279551, + "grad_norm": 1.6942301450156114, + "learning_rate": 3.492775659840919e-07, + "loss": 0.3994, + "step": 40156 + }, + { + "epoch": 0.8839313229143737, + "grad_norm": 1.5353131151188788, + "learning_rate": 3.49146689256033e-07, + "loss": 0.2732, + "step": 40157 + }, + { + "epoch": 0.8839533348007924, + "grad_norm": 1.534534187555531, + "learning_rate": 3.490158361656859e-07, + "loss": 0.3901, + "step": 40158 + }, + { + "epoch": 0.8839753466872111, + "grad_norm": 1.7410215461938252, + "learning_rate": 3.488850067137145e-07, + "loss": 0.2747, + "step": 40159 + }, + { + "epoch": 0.8839973585736297, + "grad_norm": 1.470693250268599, + "learning_rate": 3.4875420090078326e-07, + "loss": 0.1969, + "step": 40160 + }, + { + "epoch": 0.8840193704600484, + "grad_norm": 0.610712711255026, + "learning_rate": 3.486234187275572e-07, + "loss": 0.4435, + "step": 40161 + }, + { + "epoch": 0.8840413823464671, + "grad_norm": 2.4169623291093316, + "learning_rate": 3.4849266019470305e-07, + "loss": 0.2389, + "step": 40162 + }, + { + "epoch": 0.8840633942328857, + "grad_norm": 0.636200912001249, + "learning_rate": 3.483619253028825e-07, + "loss": 0.453, + "step": 40163 + }, + { + "epoch": 0.8840854061193044, + "grad_norm": 1.7481082755093198, + "learning_rate": 3.4823121405276216e-07, + "loss": 0.2717, + "step": 40164 + }, + { + "epoch": 0.8841074180057231, + "grad_norm": 1.9444385148223153, + "learning_rate": 3.481005264450049e-07, + "loss": 0.3184, + "step": 40165 + }, + { + "epoch": 0.8841294298921417, + "grad_norm": 0.6212258694108711, + "learning_rate": 3.4796986248027577e-07, + "loss": 0.4574, + "step": 40166 + }, + { + "epoch": 0.8841514417785604, + "grad_norm": 1.568443682505191, + "learning_rate": 3.4783922215923913e-07, + "loss": 0.3228, + "step": 40167 + }, + { + "epoch": 0.8841734536649791, + "grad_norm": 1.8501735080147959, + "learning_rate": 3.477086054825579e-07, + "loss": 0.3016, + "step": 40168 + }, + { + "epoch": 0.8841954655513977, + "grad_norm": 2.5182762509669625, + "learning_rate": 3.4757801245089706e-07, + "loss": 0.2863, + "step": 40169 + }, + { + "epoch": 0.8842174774378164, + "grad_norm": 0.6156692902422967, + "learning_rate": 3.474474430649194e-07, + "loss": 0.4465, + "step": 40170 + }, + { + "epoch": 0.8842394893242351, + "grad_norm": 1.7723510971201724, + "learning_rate": 3.4731689732528885e-07, + "loss": 0.2963, + "step": 40171 + }, + { + "epoch": 0.8842615012106537, + "grad_norm": 5.32333614384174, + "learning_rate": 3.471863752326693e-07, + "loss": 0.2835, + "step": 40172 + }, + { + "epoch": 0.8842835130970724, + "grad_norm": 1.6508187792114344, + "learning_rate": 3.4705587678772355e-07, + "loss": 0.3033, + "step": 40173 + }, + { + "epoch": 0.8843055249834911, + "grad_norm": 0.6136657828363643, + "learning_rate": 3.4692540199111557e-07, + "loss": 0.4452, + "step": 40174 + }, + { + "epoch": 0.8843275368699097, + "grad_norm": 3.034064250155515, + "learning_rate": 3.4679495084350755e-07, + "loss": 0.273, + "step": 40175 + }, + { + "epoch": 0.8843495487563284, + "grad_norm": 0.6183449925624847, + "learning_rate": 3.4666452334556234e-07, + "loss": 0.4582, + "step": 40176 + }, + { + "epoch": 0.8843715606427471, + "grad_norm": 1.6682263899564929, + "learning_rate": 3.465341194979449e-07, + "loss": 0.2887, + "step": 40177 + }, + { + "epoch": 0.8843935725291657, + "grad_norm": 1.8124235678436873, + "learning_rate": 3.4640373930131534e-07, + "loss": 0.2994, + "step": 40178 + }, + { + "epoch": 0.8844155844155844, + "grad_norm": 1.7024493554160396, + "learning_rate": 3.462733827563386e-07, + "loss": 0.348, + "step": 40179 + }, + { + "epoch": 0.884437596302003, + "grad_norm": 1.5329881933865934, + "learning_rate": 3.461430498636753e-07, + "loss": 0.294, + "step": 40180 + }, + { + "epoch": 0.8844596081884217, + "grad_norm": 2.8024846947153166, + "learning_rate": 3.4601274062398884e-07, + "loss": 0.2797, + "step": 40181 + }, + { + "epoch": 0.8844816200748404, + "grad_norm": 1.5723033866388765, + "learning_rate": 3.458824550379419e-07, + "loss": 0.2607, + "step": 40182 + }, + { + "epoch": 0.884503631961259, + "grad_norm": 1.569388519095316, + "learning_rate": 3.457521931061958e-07, + "loss": 0.2403, + "step": 40183 + }, + { + "epoch": 0.8845256438476777, + "grad_norm": 1.9375042980717108, + "learning_rate": 3.4562195482941317e-07, + "loss": 0.2856, + "step": 40184 + }, + { + "epoch": 0.8845476557340964, + "grad_norm": 0.6198467983846702, + "learning_rate": 3.454917402082558e-07, + "loss": 0.4371, + "step": 40185 + }, + { + "epoch": 0.884569667620515, + "grad_norm": 4.233573511822731, + "learning_rate": 3.453615492433837e-07, + "loss": 0.2354, + "step": 40186 + }, + { + "epoch": 0.8845916795069337, + "grad_norm": 1.2659522508024903, + "learning_rate": 3.452313819354619e-07, + "loss": 0.2423, + "step": 40187 + }, + { + "epoch": 0.8846136913933524, + "grad_norm": 3.074710085442196, + "learning_rate": 3.4510123828514986e-07, + "loss": 0.2072, + "step": 40188 + }, + { + "epoch": 0.884635703279771, + "grad_norm": 2.217711933225084, + "learning_rate": 3.449711182931098e-07, + "loss": 0.1884, + "step": 40189 + }, + { + "epoch": 0.8846577151661897, + "grad_norm": 0.6275813976878589, + "learning_rate": 3.448410219600029e-07, + "loss": 0.4443, + "step": 40190 + }, + { + "epoch": 0.8846797270526084, + "grad_norm": 1.7129729437687402, + "learning_rate": 3.4471094928648975e-07, + "loss": 0.4036, + "step": 40191 + }, + { + "epoch": 0.884701738939027, + "grad_norm": 1.5613054193084919, + "learning_rate": 3.4458090027323146e-07, + "loss": 0.3198, + "step": 40192 + }, + { + "epoch": 0.8847237508254457, + "grad_norm": 1.6970865452730142, + "learning_rate": 3.444508749208891e-07, + "loss": 0.3451, + "step": 40193 + }, + { + "epoch": 0.8847457627118644, + "grad_norm": 1.6687102049028342, + "learning_rate": 3.443208732301251e-07, + "loss": 0.2402, + "step": 40194 + }, + { + "epoch": 0.884767774598283, + "grad_norm": 2.027163717056325, + "learning_rate": 3.4419089520159874e-07, + "loss": 0.2603, + "step": 40195 + }, + { + "epoch": 0.8847897864847017, + "grad_norm": 1.6564451550690866, + "learning_rate": 3.440609408359702e-07, + "loss": 0.3034, + "step": 40196 + }, + { + "epoch": 0.8848117983711205, + "grad_norm": 1.495677154642047, + "learning_rate": 3.439310101339005e-07, + "loss": 0.3166, + "step": 40197 + }, + { + "epoch": 0.884833810257539, + "grad_norm": 2.078885468771716, + "learning_rate": 3.4380110309604975e-07, + "loss": 0.2917, + "step": 40198 + }, + { + "epoch": 0.8848558221439577, + "grad_norm": 1.7597229963961192, + "learning_rate": 3.436712197230796e-07, + "loss": 0.2821, + "step": 40199 + }, + { + "epoch": 0.8848778340303765, + "grad_norm": 1.6002705144516831, + "learning_rate": 3.4354136001564896e-07, + "loss": 0.3095, + "step": 40200 + }, + { + "epoch": 0.884899845916795, + "grad_norm": 2.345330089517629, + "learning_rate": 3.4341152397441733e-07, + "loss": 0.2536, + "step": 40201 + }, + { + "epoch": 0.8849218578032138, + "grad_norm": 2.375488868432421, + "learning_rate": 3.432817116000453e-07, + "loss": 0.3037, + "step": 40202 + }, + { + "epoch": 0.8849438696896325, + "grad_norm": 1.993772927801843, + "learning_rate": 3.431519228931934e-07, + "loss": 0.2873, + "step": 40203 + }, + { + "epoch": 0.884965881576051, + "grad_norm": 1.8942662080694428, + "learning_rate": 3.4302215785451955e-07, + "loss": 0.322, + "step": 40204 + }, + { + "epoch": 0.8849878934624698, + "grad_norm": 2.3126907011614906, + "learning_rate": 3.4289241648468474e-07, + "loss": 0.3512, + "step": 40205 + }, + { + "epoch": 0.8850099053488883, + "grad_norm": 1.8783350948987532, + "learning_rate": 3.4276269878434687e-07, + "loss": 0.2649, + "step": 40206 + }, + { + "epoch": 0.885031917235307, + "grad_norm": 1.2831475319168284, + "learning_rate": 3.426330047541665e-07, + "loss": 0.2657, + "step": 40207 + }, + { + "epoch": 0.8850539291217258, + "grad_norm": 1.6169537889761798, + "learning_rate": 3.425033343948031e-07, + "loss": 0.2719, + "step": 40208 + }, + { + "epoch": 0.8850759410081444, + "grad_norm": 7.0392038827969525, + "learning_rate": 3.423736877069145e-07, + "loss": 0.2674, + "step": 40209 + }, + { + "epoch": 0.8850979528945631, + "grad_norm": 0.6016537007664057, + "learning_rate": 3.4224406469116067e-07, + "loss": 0.4523, + "step": 40210 + }, + { + "epoch": 0.8851199647809818, + "grad_norm": 2.1438689365249517, + "learning_rate": 3.4211446534819945e-07, + "loss": 0.3048, + "step": 40211 + }, + { + "epoch": 0.8851419766674004, + "grad_norm": 2.341061097855287, + "learning_rate": 3.419848896786898e-07, + "loss": 0.2989, + "step": 40212 + }, + { + "epoch": 0.8851639885538191, + "grad_norm": 1.468623070445313, + "learning_rate": 3.4185533768329115e-07, + "loss": 0.2556, + "step": 40213 + }, + { + "epoch": 0.8851860004402378, + "grad_norm": 1.4283138888870708, + "learning_rate": 3.417258093626608e-07, + "loss": 0.2871, + "step": 40214 + }, + { + "epoch": 0.8852080123266564, + "grad_norm": 1.4826067430852488, + "learning_rate": 3.4159630471745806e-07, + "loss": 0.2917, + "step": 40215 + }, + { + "epoch": 0.8852300242130751, + "grad_norm": 0.6131110717002104, + "learning_rate": 3.4146682374833986e-07, + "loss": 0.4385, + "step": 40216 + }, + { + "epoch": 0.8852520360994938, + "grad_norm": 1.5909008995047889, + "learning_rate": 3.413373664559649e-07, + "loss": 0.348, + "step": 40217 + }, + { + "epoch": 0.8852740479859124, + "grad_norm": 5.0748655559296845, + "learning_rate": 3.412079328409923e-07, + "loss": 0.3165, + "step": 40218 + }, + { + "epoch": 0.8852960598723311, + "grad_norm": 1.8894923771313605, + "learning_rate": 3.4107852290407804e-07, + "loss": 0.2434, + "step": 40219 + }, + { + "epoch": 0.8853180717587498, + "grad_norm": 1.4245505046043487, + "learning_rate": 3.4094913664588113e-07, + "loss": 0.3031, + "step": 40220 + }, + { + "epoch": 0.8853400836451684, + "grad_norm": 1.84878352101996, + "learning_rate": 3.4081977406705825e-07, + "loss": 0.3435, + "step": 40221 + }, + { + "epoch": 0.8853620955315871, + "grad_norm": 1.6930899826397021, + "learning_rate": 3.406904351682672e-07, + "loss": 0.2768, + "step": 40222 + }, + { + "epoch": 0.8853841074180058, + "grad_norm": 4.885726698153813, + "learning_rate": 3.4056111995016626e-07, + "loss": 0.3511, + "step": 40223 + }, + { + "epoch": 0.8854061193044244, + "grad_norm": 1.7401204498046559, + "learning_rate": 3.4043182841341116e-07, + "loss": 0.3259, + "step": 40224 + }, + { + "epoch": 0.8854281311908431, + "grad_norm": 1.507209644583596, + "learning_rate": 3.4030256055866076e-07, + "loss": 0.2987, + "step": 40225 + }, + { + "epoch": 0.8854501430772618, + "grad_norm": 0.6043974535310302, + "learning_rate": 3.401733163865706e-07, + "loss": 0.4441, + "step": 40226 + }, + { + "epoch": 0.8854721549636804, + "grad_norm": 0.6232313372189093, + "learning_rate": 3.4004409589779743e-07, + "loss": 0.4216, + "step": 40227 + }, + { + "epoch": 0.8854941668500991, + "grad_norm": 1.4059701813141334, + "learning_rate": 3.3991489909299846e-07, + "loss": 0.2951, + "step": 40228 + }, + { + "epoch": 0.8855161787365177, + "grad_norm": 1.8928158609522, + "learning_rate": 3.397857259728304e-07, + "loss": 0.3573, + "step": 40229 + }, + { + "epoch": 0.8855381906229364, + "grad_norm": 1.5519421665112092, + "learning_rate": 3.396565765379506e-07, + "loss": 0.3419, + "step": 40230 + }, + { + "epoch": 0.8855602025093551, + "grad_norm": 1.913565071182722, + "learning_rate": 3.395274507890145e-07, + "loss": 0.2693, + "step": 40231 + }, + { + "epoch": 0.8855822143957737, + "grad_norm": 3.923503812806077, + "learning_rate": 3.393983487266783e-07, + "loss": 0.3039, + "step": 40232 + }, + { + "epoch": 0.8856042262821924, + "grad_norm": 4.131702859295858, + "learning_rate": 3.392692703515976e-07, + "loss": 0.2277, + "step": 40233 + }, + { + "epoch": 0.8856262381686111, + "grad_norm": 2.1091095950758616, + "learning_rate": 3.391402156644297e-07, + "loss": 0.3166, + "step": 40234 + }, + { + "epoch": 0.8856482500550297, + "grad_norm": 1.6614451937123356, + "learning_rate": 3.390111846658306e-07, + "loss": 0.2248, + "step": 40235 + }, + { + "epoch": 0.8856702619414484, + "grad_norm": 6.2683859480515896, + "learning_rate": 3.38882177356456e-07, + "loss": 0.3724, + "step": 40236 + }, + { + "epoch": 0.8856922738278671, + "grad_norm": 2.0489013783533445, + "learning_rate": 3.387531937369598e-07, + "loss": 0.2831, + "step": 40237 + }, + { + "epoch": 0.8857142857142857, + "grad_norm": 2.7777705078272783, + "learning_rate": 3.3862423380799923e-07, + "loss": 0.3309, + "step": 40238 + }, + { + "epoch": 0.8857362976007044, + "grad_norm": 1.3215584652783376, + "learning_rate": 3.3849529757022883e-07, + "loss": 0.2357, + "step": 40239 + }, + { + "epoch": 0.8857583094871231, + "grad_norm": 2.1837891252450756, + "learning_rate": 3.3836638502430576e-07, + "loss": 0.284, + "step": 40240 + }, + { + "epoch": 0.8857803213735417, + "grad_norm": 0.608428796951193, + "learning_rate": 3.382374961708834e-07, + "loss": 0.432, + "step": 40241 + }, + { + "epoch": 0.8858023332599604, + "grad_norm": 1.7611408192738265, + "learning_rate": 3.381086310106169e-07, + "loss": 0.2153, + "step": 40242 + }, + { + "epoch": 0.8858243451463791, + "grad_norm": 1.6527724011318017, + "learning_rate": 3.379797895441611e-07, + "loss": 0.2537, + "step": 40243 + }, + { + "epoch": 0.8858463570327977, + "grad_norm": 0.6209489748446306, + "learning_rate": 3.378509717721723e-07, + "loss": 0.4702, + "step": 40244 + }, + { + "epoch": 0.8858683689192164, + "grad_norm": 2.1810396913712844, + "learning_rate": 3.377221776953038e-07, + "loss": 0.3028, + "step": 40245 + }, + { + "epoch": 0.8858903808056351, + "grad_norm": 1.3976728192317578, + "learning_rate": 3.375934073142112e-07, + "loss": 0.2575, + "step": 40246 + }, + { + "epoch": 0.8859123926920537, + "grad_norm": 0.6439742719901173, + "learning_rate": 3.374646606295473e-07, + "loss": 0.4524, + "step": 40247 + }, + { + "epoch": 0.8859344045784724, + "grad_norm": 0.6131748388120447, + "learning_rate": 3.373359376419677e-07, + "loss": 0.4604, + "step": 40248 + }, + { + "epoch": 0.8859564164648911, + "grad_norm": 1.7570738388708953, + "learning_rate": 3.3720723835212745e-07, + "loss": 0.3183, + "step": 40249 + }, + { + "epoch": 0.8859784283513097, + "grad_norm": 1.4853135547639433, + "learning_rate": 3.3707856276067876e-07, + "loss": 0.2178, + "step": 40250 + }, + { + "epoch": 0.8860004402377284, + "grad_norm": 0.6119427363357207, + "learning_rate": 3.3694991086827775e-07, + "loss": 0.4413, + "step": 40251 + }, + { + "epoch": 0.886022452124147, + "grad_norm": 2.194141212488339, + "learning_rate": 3.368212826755757e-07, + "loss": 0.2987, + "step": 40252 + }, + { + "epoch": 0.8860444640105657, + "grad_norm": 1.482345052043571, + "learning_rate": 3.3669267818322806e-07, + "loss": 0.3562, + "step": 40253 + }, + { + "epoch": 0.8860664758969844, + "grad_norm": 0.6156732905171847, + "learning_rate": 3.3656409739188877e-07, + "loss": 0.4412, + "step": 40254 + }, + { + "epoch": 0.886088487783403, + "grad_norm": 2.69364924319873, + "learning_rate": 3.364355403022096e-07, + "loss": 0.4003, + "step": 40255 + }, + { + "epoch": 0.8861104996698217, + "grad_norm": 3.058762333664016, + "learning_rate": 3.363070069148461e-07, + "loss": 0.2159, + "step": 40256 + }, + { + "epoch": 0.8861325115562404, + "grad_norm": 1.7822623622365898, + "learning_rate": 3.361784972304499e-07, + "loss": 0.3194, + "step": 40257 + }, + { + "epoch": 0.886154523442659, + "grad_norm": 4.005809828806015, + "learning_rate": 3.3605001124967454e-07, + "loss": 0.2909, + "step": 40258 + }, + { + "epoch": 0.8861765353290777, + "grad_norm": 0.633699649214872, + "learning_rate": 3.3592154897317374e-07, + "loss": 0.4317, + "step": 40259 + }, + { + "epoch": 0.8861985472154964, + "grad_norm": 2.318626834330959, + "learning_rate": 3.3579311040159936e-07, + "loss": 0.2868, + "step": 40260 + }, + { + "epoch": 0.886220559101915, + "grad_norm": 1.771429090961191, + "learning_rate": 3.356646955356052e-07, + "loss": 0.2674, + "step": 40261 + }, + { + "epoch": 0.8862425709883337, + "grad_norm": 1.465555830542988, + "learning_rate": 3.355363043758442e-07, + "loss": 0.2676, + "step": 40262 + }, + { + "epoch": 0.8862645828747524, + "grad_norm": 1.8201315898471604, + "learning_rate": 3.3540793692296626e-07, + "loss": 0.3434, + "step": 40263 + }, + { + "epoch": 0.886286594761171, + "grad_norm": 2.2098588951856035, + "learning_rate": 3.352795931776265e-07, + "loss": 0.3144, + "step": 40264 + }, + { + "epoch": 0.8863086066475897, + "grad_norm": 1.6538916404588286, + "learning_rate": 3.35151273140476e-07, + "loss": 0.2929, + "step": 40265 + }, + { + "epoch": 0.8863306185340084, + "grad_norm": 1.8872985692012145, + "learning_rate": 3.350229768121677e-07, + "loss": 0.2425, + "step": 40266 + }, + { + "epoch": 0.886352630420427, + "grad_norm": 1.5664832553625865, + "learning_rate": 3.3489470419335366e-07, + "loss": 0.2999, + "step": 40267 + }, + { + "epoch": 0.8863746423068457, + "grad_norm": 1.5826621241030379, + "learning_rate": 3.347664552846847e-07, + "loss": 0.2721, + "step": 40268 + }, + { + "epoch": 0.8863966541932644, + "grad_norm": 1.6897925197946855, + "learning_rate": 3.346382300868134e-07, + "loss": 0.2964, + "step": 40269 + }, + { + "epoch": 0.886418666079683, + "grad_norm": 1.5166633426114609, + "learning_rate": 3.345100286003911e-07, + "loss": 0.2847, + "step": 40270 + }, + { + "epoch": 0.8864406779661017, + "grad_norm": 1.6824820394326012, + "learning_rate": 3.34381850826071e-07, + "loss": 0.2397, + "step": 40271 + }, + { + "epoch": 0.8864626898525204, + "grad_norm": 1.9098924046275532, + "learning_rate": 3.342536967645027e-07, + "loss": 0.2367, + "step": 40272 + }, + { + "epoch": 0.886484701738939, + "grad_norm": 2.089594766032096, + "learning_rate": 3.3412556641633787e-07, + "loss": 0.3119, + "step": 40273 + }, + { + "epoch": 0.8865067136253577, + "grad_norm": 1.5550462122781317, + "learning_rate": 3.339974597822276e-07, + "loss": 0.2564, + "step": 40274 + }, + { + "epoch": 0.8865287255117763, + "grad_norm": 1.4827076114019915, + "learning_rate": 3.3386937686282315e-07, + "loss": 0.2447, + "step": 40275 + }, + { + "epoch": 0.886550737398195, + "grad_norm": 1.492242651874762, + "learning_rate": 3.3374131765877673e-07, + "loss": 0.2413, + "step": 40276 + }, + { + "epoch": 0.8865727492846137, + "grad_norm": 1.5204220973973746, + "learning_rate": 3.336132821707383e-07, + "loss": 0.2357, + "step": 40277 + }, + { + "epoch": 0.8865947611710323, + "grad_norm": 1.8534214397813624, + "learning_rate": 3.334852703993574e-07, + "loss": 0.301, + "step": 40278 + }, + { + "epoch": 0.886616773057451, + "grad_norm": 1.3329337806718327, + "learning_rate": 3.3335728234528573e-07, + "loss": 0.2439, + "step": 40279 + }, + { + "epoch": 0.8866387849438697, + "grad_norm": 1.9194962429668043, + "learning_rate": 3.332293180091739e-07, + "loss": 0.3401, + "step": 40280 + }, + { + "epoch": 0.8866607968302883, + "grad_norm": 2.0046350183339863, + "learning_rate": 3.33101377391673e-07, + "loss": 0.3017, + "step": 40281 + }, + { + "epoch": 0.886682808716707, + "grad_norm": 1.7740289330734327, + "learning_rate": 3.3297346049343206e-07, + "loss": 0.2466, + "step": 40282 + }, + { + "epoch": 0.8867048206031257, + "grad_norm": 0.6344421731999145, + "learning_rate": 3.32845567315101e-07, + "loss": 0.4395, + "step": 40283 + }, + { + "epoch": 0.8867268324895443, + "grad_norm": 0.6234102553955455, + "learning_rate": 3.327176978573299e-07, + "loss": 0.4472, + "step": 40284 + }, + { + "epoch": 0.886748844375963, + "grad_norm": 1.4218022864883564, + "learning_rate": 3.325898521207704e-07, + "loss": 0.2854, + "step": 40285 + }, + { + "epoch": 0.8867708562623817, + "grad_norm": 1.7653439207426742, + "learning_rate": 3.324620301060699e-07, + "loss": 0.2469, + "step": 40286 + }, + { + "epoch": 0.8867928681488003, + "grad_norm": 1.6248350129869935, + "learning_rate": 3.3233423181388003e-07, + "loss": 0.3382, + "step": 40287 + }, + { + "epoch": 0.886814880035219, + "grad_norm": 1.6097048473756193, + "learning_rate": 3.322064572448486e-07, + "loss": 0.2384, + "step": 40288 + }, + { + "epoch": 0.8868368919216377, + "grad_norm": 1.391554319655543, + "learning_rate": 3.3207870639962613e-07, + "loss": 0.3079, + "step": 40289 + }, + { + "epoch": 0.8868589038080563, + "grad_norm": 1.2522858555715142, + "learning_rate": 3.319509792788622e-07, + "loss": 0.1672, + "step": 40290 + }, + { + "epoch": 0.886880915694475, + "grad_norm": 1.619642910314214, + "learning_rate": 3.3182327588320463e-07, + "loss": 0.2789, + "step": 40291 + }, + { + "epoch": 0.8869029275808937, + "grad_norm": 1.747517057096312, + "learning_rate": 3.3169559621330396e-07, + "loss": 0.268, + "step": 40292 + }, + { + "epoch": 0.8869249394673123, + "grad_norm": 1.7989889565145059, + "learning_rate": 3.31567940269808e-07, + "loss": 0.3083, + "step": 40293 + }, + { + "epoch": 0.886946951353731, + "grad_norm": 1.5524316668772367, + "learning_rate": 3.3144030805336627e-07, + "loss": 0.2991, + "step": 40294 + }, + { + "epoch": 0.8869689632401497, + "grad_norm": 0.6124307629169339, + "learning_rate": 3.313126995646265e-07, + "loss": 0.4362, + "step": 40295 + }, + { + "epoch": 0.8869909751265683, + "grad_norm": 1.7921959555785705, + "learning_rate": 3.3118511480423777e-07, + "loss": 0.4419, + "step": 40296 + }, + { + "epoch": 0.887012987012987, + "grad_norm": 2.304026193926649, + "learning_rate": 3.3105755377284945e-07, + "loss": 0.4078, + "step": 40297 + }, + { + "epoch": 0.8870349988994056, + "grad_norm": 1.507207622570007, + "learning_rate": 3.309300164711082e-07, + "loss": 0.2787, + "step": 40298 + }, + { + "epoch": 0.8870570107858243, + "grad_norm": 1.8365331785436487, + "learning_rate": 3.308025028996642e-07, + "loss": 0.3181, + "step": 40299 + }, + { + "epoch": 0.887079022672243, + "grad_norm": 2.234748367903728, + "learning_rate": 3.306750130591635e-07, + "loss": 0.2975, + "step": 40300 + }, + { + "epoch": 0.8871010345586616, + "grad_norm": 1.3189231193860351, + "learning_rate": 3.305475469502545e-07, + "loss": 0.2451, + "step": 40301 + }, + { + "epoch": 0.8871230464450803, + "grad_norm": 1.4267755898982402, + "learning_rate": 3.3042010457358665e-07, + "loss": 0.2255, + "step": 40302 + }, + { + "epoch": 0.887145058331499, + "grad_norm": 0.59204966990585, + "learning_rate": 3.302926859298067e-07, + "loss": 0.4144, + "step": 40303 + }, + { + "epoch": 0.8871670702179176, + "grad_norm": 1.5520250483441522, + "learning_rate": 3.301652910195613e-07, + "loss": 0.2956, + "step": 40304 + }, + { + "epoch": 0.8871890821043363, + "grad_norm": 2.125131903860973, + "learning_rate": 3.300379198434983e-07, + "loss": 0.309, + "step": 40305 + }, + { + "epoch": 0.887211093990755, + "grad_norm": 1.2693805560007818, + "learning_rate": 3.2991057240226556e-07, + "loss": 0.2487, + "step": 40306 + }, + { + "epoch": 0.8872331058771736, + "grad_norm": 1.3107859100013155, + "learning_rate": 3.2978324869651134e-07, + "loss": 0.2404, + "step": 40307 + }, + { + "epoch": 0.8872551177635923, + "grad_norm": 0.6175632034633963, + "learning_rate": 3.2965594872688134e-07, + "loss": 0.4327, + "step": 40308 + }, + { + "epoch": 0.887277129650011, + "grad_norm": 0.6039223368936018, + "learning_rate": 3.295286724940222e-07, + "loss": 0.4408, + "step": 40309 + }, + { + "epoch": 0.8872991415364296, + "grad_norm": 1.6846405641046007, + "learning_rate": 3.294014199985812e-07, + "loss": 0.323, + "step": 40310 + }, + { + "epoch": 0.8873211534228483, + "grad_norm": 1.9126603340789063, + "learning_rate": 3.2927419124120506e-07, + "loss": 0.349, + "step": 40311 + }, + { + "epoch": 0.887343165309267, + "grad_norm": 2.068294551967451, + "learning_rate": 3.2914698622254216e-07, + "loss": 0.2489, + "step": 40312 + }, + { + "epoch": 0.8873651771956856, + "grad_norm": 2.6126112632185756, + "learning_rate": 3.2901980494323693e-07, + "loss": 0.3024, + "step": 40313 + }, + { + "epoch": 0.8873871890821043, + "grad_norm": 1.4900866196676033, + "learning_rate": 3.2889264740393614e-07, + "loss": 0.2849, + "step": 40314 + }, + { + "epoch": 0.887409200968523, + "grad_norm": 1.5358193669178897, + "learning_rate": 3.2876551360528585e-07, + "loss": 0.1936, + "step": 40315 + }, + { + "epoch": 0.8874312128549416, + "grad_norm": 1.450306789116736, + "learning_rate": 3.2863840354793284e-07, + "loss": 0.2867, + "step": 40316 + }, + { + "epoch": 0.8874532247413603, + "grad_norm": 3.1139260406972036, + "learning_rate": 3.285113172325238e-07, + "loss": 0.3486, + "step": 40317 + }, + { + "epoch": 0.887475236627779, + "grad_norm": 1.3988430707762038, + "learning_rate": 3.283842546597038e-07, + "loss": 0.2535, + "step": 40318 + }, + { + "epoch": 0.8874972485141976, + "grad_norm": 2.111907065290288, + "learning_rate": 3.282572158301178e-07, + "loss": 0.2531, + "step": 40319 + }, + { + "epoch": 0.8875192604006163, + "grad_norm": 1.5984744986146173, + "learning_rate": 3.2813020074441205e-07, + "loss": 0.2945, + "step": 40320 + }, + { + "epoch": 0.8875412722870349, + "grad_norm": 3.1509153466328317, + "learning_rate": 3.280032094032326e-07, + "loss": 0.2301, + "step": 40321 + }, + { + "epoch": 0.8875632841734536, + "grad_norm": 2.287374999743796, + "learning_rate": 3.278762418072251e-07, + "loss": 0.2048, + "step": 40322 + }, + { + "epoch": 0.8875852960598724, + "grad_norm": 1.9660583244837058, + "learning_rate": 3.2774929795703457e-07, + "loss": 0.2902, + "step": 40323 + }, + { + "epoch": 0.887607307946291, + "grad_norm": 2.181569068375025, + "learning_rate": 3.276223778533055e-07, + "loss": 0.2876, + "step": 40324 + }, + { + "epoch": 0.8876293198327097, + "grad_norm": 1.745472958896535, + "learning_rate": 3.27495481496683e-07, + "loss": 0.3623, + "step": 40325 + }, + { + "epoch": 0.8876513317191284, + "grad_norm": 0.6209140758560636, + "learning_rate": 3.2736860888781365e-07, + "loss": 0.4552, + "step": 40326 + }, + { + "epoch": 0.887673343605547, + "grad_norm": 6.063423664937588, + "learning_rate": 3.272417600273398e-07, + "loss": 0.3091, + "step": 40327 + }, + { + "epoch": 0.8876953554919657, + "grad_norm": 1.91479415911121, + "learning_rate": 3.271149349159086e-07, + "loss": 0.2676, + "step": 40328 + }, + { + "epoch": 0.8877173673783844, + "grad_norm": 2.210894800857737, + "learning_rate": 3.2698813355416246e-07, + "loss": 0.2605, + "step": 40329 + }, + { + "epoch": 0.887739379264803, + "grad_norm": 1.7568290199996865, + "learning_rate": 3.268613559427475e-07, + "loss": 0.2771, + "step": 40330 + }, + { + "epoch": 0.8877613911512217, + "grad_norm": 0.6094222601308829, + "learning_rate": 3.26734602082307e-07, + "loss": 0.4473, + "step": 40331 + }, + { + "epoch": 0.8877834030376404, + "grad_norm": 1.4225514223430642, + "learning_rate": 3.2660787197348553e-07, + "loss": 0.2561, + "step": 40332 + }, + { + "epoch": 0.887805414924059, + "grad_norm": 1.9763092854760025, + "learning_rate": 3.2648116561692754e-07, + "loss": 0.2997, + "step": 40333 + }, + { + "epoch": 0.8878274268104777, + "grad_norm": 1.8215376869112918, + "learning_rate": 3.263544830132764e-07, + "loss": 0.27, + "step": 40334 + }, + { + "epoch": 0.8878494386968964, + "grad_norm": 1.7669503207718418, + "learning_rate": 3.262278241631772e-07, + "loss": 0.2929, + "step": 40335 + }, + { + "epoch": 0.887871450583315, + "grad_norm": 1.9973443653879925, + "learning_rate": 3.2610118906727163e-07, + "loss": 0.2529, + "step": 40336 + }, + { + "epoch": 0.8878934624697337, + "grad_norm": 1.9386722327650552, + "learning_rate": 3.259745777262047e-07, + "loss": 0.3156, + "step": 40337 + }, + { + "epoch": 0.8879154743561524, + "grad_norm": 3.6704958024162457, + "learning_rate": 3.258479901406203e-07, + "loss": 0.2845, + "step": 40338 + }, + { + "epoch": 0.887937486242571, + "grad_norm": 1.664892365701233, + "learning_rate": 3.257214263111602e-07, + "loss": 0.3032, + "step": 40339 + }, + { + "epoch": 0.8879594981289897, + "grad_norm": 1.6467858373858777, + "learning_rate": 3.2559488623846946e-07, + "loss": 0.3481, + "step": 40340 + }, + { + "epoch": 0.8879815100154084, + "grad_norm": 1.8347448241186903, + "learning_rate": 3.2546836992318976e-07, + "loss": 0.1959, + "step": 40341 + }, + { + "epoch": 0.888003521901827, + "grad_norm": 6.032120550033399, + "learning_rate": 3.2534187736596446e-07, + "loss": 0.2493, + "step": 40342 + }, + { + "epoch": 0.8880255337882457, + "grad_norm": 1.9940057008061047, + "learning_rate": 3.252154085674375e-07, + "loss": 0.3263, + "step": 40343 + }, + { + "epoch": 0.8880475456746644, + "grad_norm": 1.48065437796106, + "learning_rate": 3.2508896352825115e-07, + "loss": 0.3449, + "step": 40344 + }, + { + "epoch": 0.888069557561083, + "grad_norm": 0.6174956010171437, + "learning_rate": 3.2496254224904656e-07, + "loss": 0.4228, + "step": 40345 + }, + { + "epoch": 0.8880915694475017, + "grad_norm": 1.633958848075257, + "learning_rate": 3.2483614473046764e-07, + "loss": 0.2764, + "step": 40346 + }, + { + "epoch": 0.8881135813339203, + "grad_norm": 2.7102862427705294, + "learning_rate": 3.247097709731567e-07, + "loss": 0.3135, + "step": 40347 + }, + { + "epoch": 0.888135593220339, + "grad_norm": 1.5407378913065182, + "learning_rate": 3.245834209777565e-07, + "loss": 0.2364, + "step": 40348 + }, + { + "epoch": 0.8881576051067577, + "grad_norm": 1.7545437608477628, + "learning_rate": 3.244570947449088e-07, + "loss": 0.2591, + "step": 40349 + }, + { + "epoch": 0.8881796169931763, + "grad_norm": 1.8623178425811513, + "learning_rate": 3.2433079227525475e-07, + "loss": 0.3537, + "step": 40350 + }, + { + "epoch": 0.888201628879595, + "grad_norm": 1.6323946500347448, + "learning_rate": 3.242045135694372e-07, + "loss": 0.2505, + "step": 40351 + }, + { + "epoch": 0.8882236407660137, + "grad_norm": 1.9074996481336, + "learning_rate": 3.2407825862809717e-07, + "loss": 0.2201, + "step": 40352 + }, + { + "epoch": 0.8882456526524323, + "grad_norm": 2.06995244641699, + "learning_rate": 3.2395202745187813e-07, + "loss": 0.3125, + "step": 40353 + }, + { + "epoch": 0.888267664538851, + "grad_norm": 1.3794676809176492, + "learning_rate": 3.238258200414207e-07, + "loss": 0.2624, + "step": 40354 + }, + { + "epoch": 0.8882896764252697, + "grad_norm": 4.258848694444142, + "learning_rate": 3.236996363973649e-07, + "loss": 0.3646, + "step": 40355 + }, + { + "epoch": 0.8883116883116883, + "grad_norm": 1.4875746853782763, + "learning_rate": 3.2357347652035356e-07, + "loss": 0.2161, + "step": 40356 + }, + { + "epoch": 0.888333700198107, + "grad_norm": 1.67407125480409, + "learning_rate": 3.234473404110272e-07, + "loss": 0.2824, + "step": 40357 + }, + { + "epoch": 0.8883557120845257, + "grad_norm": 2.3189720697612124, + "learning_rate": 3.233212280700282e-07, + "loss": 0.424, + "step": 40358 + }, + { + "epoch": 0.8883777239709443, + "grad_norm": 0.6158881602218462, + "learning_rate": 3.2319513949799654e-07, + "loss": 0.4753, + "step": 40359 + }, + { + "epoch": 0.888399735857363, + "grad_norm": 3.7377385184580225, + "learning_rate": 3.2306907469557226e-07, + "loss": 0.2577, + "step": 40360 + }, + { + "epoch": 0.8884217477437817, + "grad_norm": 0.613748787552409, + "learning_rate": 3.229430336633971e-07, + "loss": 0.4721, + "step": 40361 + }, + { + "epoch": 0.8884437596302003, + "grad_norm": 1.6482817953131825, + "learning_rate": 3.2281701640211217e-07, + "loss": 0.3194, + "step": 40362 + }, + { + "epoch": 0.888465771516619, + "grad_norm": 0.6230110050834781, + "learning_rate": 3.2269102291235643e-07, + "loss": 0.4341, + "step": 40363 + }, + { + "epoch": 0.8884877834030377, + "grad_norm": 1.5471653757229145, + "learning_rate": 3.2256505319477217e-07, + "loss": 0.2957, + "step": 40364 + }, + { + "epoch": 0.8885097952894563, + "grad_norm": 2.2334656630835137, + "learning_rate": 3.224391072499972e-07, + "loss": 0.328, + "step": 40365 + }, + { + "epoch": 0.888531807175875, + "grad_norm": 1.6204524331927004, + "learning_rate": 3.2231318507867426e-07, + "loss": 0.2467, + "step": 40366 + }, + { + "epoch": 0.8885538190622937, + "grad_norm": 1.8489365089724343, + "learning_rate": 3.221872866814407e-07, + "loss": 0.2907, + "step": 40367 + }, + { + "epoch": 0.8885758309487123, + "grad_norm": 1.4833711091407265, + "learning_rate": 3.220614120589377e-07, + "loss": 0.243, + "step": 40368 + }, + { + "epoch": 0.888597842835131, + "grad_norm": 2.367477159678945, + "learning_rate": 3.2193556121180633e-07, + "loss": 0.3861, + "step": 40369 + }, + { + "epoch": 0.8886198547215496, + "grad_norm": 1.7105256630749999, + "learning_rate": 3.2180973414068395e-07, + "loss": 0.2328, + "step": 40370 + }, + { + "epoch": 0.8886418666079683, + "grad_norm": 0.5980085909199697, + "learning_rate": 3.216839308462116e-07, + "loss": 0.456, + "step": 40371 + }, + { + "epoch": 0.888663878494387, + "grad_norm": 1.3681264817906278, + "learning_rate": 3.2155815132902724e-07, + "loss": 0.2216, + "step": 40372 + }, + { + "epoch": 0.8886858903808056, + "grad_norm": 1.5915506084886892, + "learning_rate": 3.2143239558977135e-07, + "loss": 0.2364, + "step": 40373 + }, + { + "epoch": 0.8887079022672243, + "grad_norm": 2.136311344456209, + "learning_rate": 3.2130666362908293e-07, + "loss": 0.2868, + "step": 40374 + }, + { + "epoch": 0.888729914153643, + "grad_norm": 1.5603238864522553, + "learning_rate": 3.2118095544760084e-07, + "loss": 0.2686, + "step": 40375 + }, + { + "epoch": 0.8887519260400616, + "grad_norm": 2.1085867561707152, + "learning_rate": 3.210552710459641e-07, + "loss": 0.2846, + "step": 40376 + }, + { + "epoch": 0.8887739379264803, + "grad_norm": 1.6992438064644182, + "learning_rate": 3.2092961042481055e-07, + "loss": 0.2788, + "step": 40377 + }, + { + "epoch": 0.888795949812899, + "grad_norm": 1.3361218388380511, + "learning_rate": 3.208039735847801e-07, + "loss": 0.2007, + "step": 40378 + }, + { + "epoch": 0.8888179616993176, + "grad_norm": 0.6236090996227293, + "learning_rate": 3.2067836052651125e-07, + "loss": 0.4499, + "step": 40379 + }, + { + "epoch": 0.8888399735857363, + "grad_norm": 0.6122153244724663, + "learning_rate": 3.205527712506412e-07, + "loss": 0.4432, + "step": 40380 + }, + { + "epoch": 0.888861985472155, + "grad_norm": 1.5983084539499635, + "learning_rate": 3.2042720575781003e-07, + "loss": 0.3575, + "step": 40381 + }, + { + "epoch": 0.8888839973585736, + "grad_norm": 1.5836685601333105, + "learning_rate": 3.203016640486545e-07, + "loss": 0.3383, + "step": 40382 + }, + { + "epoch": 0.8889060092449923, + "grad_norm": 1.3376664492389034, + "learning_rate": 3.2017614612381234e-07, + "loss": 0.2339, + "step": 40383 + }, + { + "epoch": 0.888928021131411, + "grad_norm": 0.6159804648840894, + "learning_rate": 3.200506519839236e-07, + "loss": 0.4687, + "step": 40384 + }, + { + "epoch": 0.8889500330178296, + "grad_norm": 1.7881236267378655, + "learning_rate": 3.199251816296245e-07, + "loss": 0.2763, + "step": 40385 + }, + { + "epoch": 0.8889720449042483, + "grad_norm": 1.877724000127615, + "learning_rate": 3.197997350615523e-07, + "loss": 0.3446, + "step": 40386 + }, + { + "epoch": 0.888994056790667, + "grad_norm": 3.9877024101251353, + "learning_rate": 3.196743122803453e-07, + "loss": 0.3552, + "step": 40387 + }, + { + "epoch": 0.8890160686770856, + "grad_norm": 1.7388255928981409, + "learning_rate": 3.195489132866414e-07, + "loss": 0.2756, + "step": 40388 + }, + { + "epoch": 0.8890380805635043, + "grad_norm": 1.6903693114443419, + "learning_rate": 3.194235380810773e-07, + "loss": 0.2561, + "step": 40389 + }, + { + "epoch": 0.889060092449923, + "grad_norm": 2.111556222978731, + "learning_rate": 3.1929818666429135e-07, + "loss": 0.2859, + "step": 40390 + }, + { + "epoch": 0.8890821043363416, + "grad_norm": 1.7899768107929477, + "learning_rate": 3.191728590369181e-07, + "loss": 0.3955, + "step": 40391 + }, + { + "epoch": 0.8891041162227603, + "grad_norm": 1.5056754257212719, + "learning_rate": 3.190475551995964e-07, + "loss": 0.2097, + "step": 40392 + }, + { + "epoch": 0.8891261281091789, + "grad_norm": 0.5950935782667871, + "learning_rate": 3.1892227515296303e-07, + "loss": 0.4447, + "step": 40393 + }, + { + "epoch": 0.8891481399955976, + "grad_norm": 1.8418808889767073, + "learning_rate": 3.187970188976547e-07, + "loss": 0.3637, + "step": 40394 + }, + { + "epoch": 0.8891701518820163, + "grad_norm": 1.6031430827080724, + "learning_rate": 3.186717864343081e-07, + "loss": 0.2632, + "step": 40395 + }, + { + "epoch": 0.8891921637684349, + "grad_norm": 2.385111247393987, + "learning_rate": 3.185465777635588e-07, + "loss": 0.3505, + "step": 40396 + }, + { + "epoch": 0.8892141756548536, + "grad_norm": 0.6044280570951432, + "learning_rate": 3.1842139288604356e-07, + "loss": 0.4446, + "step": 40397 + }, + { + "epoch": 0.8892361875412723, + "grad_norm": 1.6718188605470525, + "learning_rate": 3.1829623180239857e-07, + "loss": 0.2586, + "step": 40398 + }, + { + "epoch": 0.8892581994276909, + "grad_norm": 1.7310789341953383, + "learning_rate": 3.1817109451326157e-07, + "loss": 0.2698, + "step": 40399 + }, + { + "epoch": 0.8892802113141096, + "grad_norm": 1.9829639843765376, + "learning_rate": 3.180459810192671e-07, + "loss": 0.3442, + "step": 40400 + }, + { + "epoch": 0.8893022232005283, + "grad_norm": 1.79948977099245, + "learning_rate": 3.179208913210502e-07, + "loss": 0.2719, + "step": 40401 + }, + { + "epoch": 0.8893242350869469, + "grad_norm": 2.1684065028635398, + "learning_rate": 3.1779582541924816e-07, + "loss": 0.2592, + "step": 40402 + }, + { + "epoch": 0.8893462469733656, + "grad_norm": 1.6146273188929094, + "learning_rate": 3.1767078331449544e-07, + "loss": 0.2412, + "step": 40403 + }, + { + "epoch": 0.8893682588597843, + "grad_norm": 1.477379992847914, + "learning_rate": 3.175457650074287e-07, + "loss": 0.2643, + "step": 40404 + }, + { + "epoch": 0.8893902707462029, + "grad_norm": 0.6184967827255441, + "learning_rate": 3.1742077049868316e-07, + "loss": 0.4614, + "step": 40405 + }, + { + "epoch": 0.8894122826326216, + "grad_norm": 1.3174951877149053, + "learning_rate": 3.172957997888926e-07, + "loss": 0.274, + "step": 40406 + }, + { + "epoch": 0.8894342945190403, + "grad_norm": 1.9698293672532992, + "learning_rate": 3.171708528786949e-07, + "loss": 0.2615, + "step": 40407 + }, + { + "epoch": 0.8894563064054589, + "grad_norm": 1.464177447996925, + "learning_rate": 3.1704592976872226e-07, + "loss": 0.2664, + "step": 40408 + }, + { + "epoch": 0.8894783182918776, + "grad_norm": 6.288374325074279, + "learning_rate": 3.1692103045961097e-07, + "loss": 0.234, + "step": 40409 + }, + { + "epoch": 0.8895003301782963, + "grad_norm": 1.6532048129516868, + "learning_rate": 3.167961549519966e-07, + "loss": 0.2751, + "step": 40410 + }, + { + "epoch": 0.8895223420647149, + "grad_norm": 1.5832690012158732, + "learning_rate": 3.1667130324651184e-07, + "loss": 0.2929, + "step": 40411 + }, + { + "epoch": 0.8895443539511336, + "grad_norm": 1.8847144193482208, + "learning_rate": 3.1654647534379303e-07, + "loss": 0.2897, + "step": 40412 + }, + { + "epoch": 0.8895663658375523, + "grad_norm": 1.4673829400266851, + "learning_rate": 3.1642167124447345e-07, + "loss": 0.3102, + "step": 40413 + }, + { + "epoch": 0.8895883777239709, + "grad_norm": 2.415572192108429, + "learning_rate": 3.162968909491881e-07, + "loss": 0.2989, + "step": 40414 + }, + { + "epoch": 0.8896103896103896, + "grad_norm": 0.637755531107547, + "learning_rate": 3.161721344585711e-07, + "loss": 0.4364, + "step": 40415 + }, + { + "epoch": 0.8896324014968082, + "grad_norm": 2.0364578019694126, + "learning_rate": 3.160474017732562e-07, + "loss": 0.3112, + "step": 40416 + }, + { + "epoch": 0.8896544133832269, + "grad_norm": 1.4402260799522795, + "learning_rate": 3.15922692893878e-07, + "loss": 0.2907, + "step": 40417 + }, + { + "epoch": 0.8896764252696456, + "grad_norm": 1.5995482914984367, + "learning_rate": 3.157980078210693e-07, + "loss": 0.3288, + "step": 40418 + }, + { + "epoch": 0.8896984371560642, + "grad_norm": 1.261533548377664, + "learning_rate": 3.15673346555464e-07, + "loss": 0.2088, + "step": 40419 + }, + { + "epoch": 0.8897204490424829, + "grad_norm": 1.4576498276545764, + "learning_rate": 3.1554870909769664e-07, + "loss": 0.3103, + "step": 40420 + }, + { + "epoch": 0.8897424609289016, + "grad_norm": 0.6248951205347698, + "learning_rate": 3.154240954483995e-07, + "loss": 0.4583, + "step": 40421 + }, + { + "epoch": 0.8897644728153202, + "grad_norm": 1.7337834901823472, + "learning_rate": 3.1529950560820754e-07, + "loss": 0.2644, + "step": 40422 + }, + { + "epoch": 0.8897864847017389, + "grad_norm": 1.340793040917635, + "learning_rate": 3.15174939577752e-07, + "loss": 0.2227, + "step": 40423 + }, + { + "epoch": 0.8898084965881576, + "grad_norm": 1.817465244613184, + "learning_rate": 3.150503973576668e-07, + "loss": 0.3255, + "step": 40424 + }, + { + "epoch": 0.8898305084745762, + "grad_norm": 1.8024064378056568, + "learning_rate": 3.1492587894858596e-07, + "loss": 0.2527, + "step": 40425 + }, + { + "epoch": 0.8898525203609949, + "grad_norm": 2.0537120605678134, + "learning_rate": 3.1480138435114103e-07, + "loss": 0.2967, + "step": 40426 + }, + { + "epoch": 0.8898745322474136, + "grad_norm": 0.5925491672285846, + "learning_rate": 3.146769135659644e-07, + "loss": 0.4189, + "step": 40427 + }, + { + "epoch": 0.8898965441338322, + "grad_norm": 1.7088415304863391, + "learning_rate": 3.1455246659368996e-07, + "loss": 0.2896, + "step": 40428 + }, + { + "epoch": 0.8899185560202509, + "grad_norm": 1.6357353162056583, + "learning_rate": 3.144280434349489e-07, + "loss": 0.3585, + "step": 40429 + }, + { + "epoch": 0.8899405679066696, + "grad_norm": 1.992988371082142, + "learning_rate": 3.143036440903757e-07, + "loss": 0.2359, + "step": 40430 + }, + { + "epoch": 0.8899625797930882, + "grad_norm": 2.9942402430719843, + "learning_rate": 3.14179268560601e-07, + "loss": 0.2281, + "step": 40431 + }, + { + "epoch": 0.8899845916795069, + "grad_norm": 1.6868796348909982, + "learning_rate": 3.1405491684625646e-07, + "loss": 0.2391, + "step": 40432 + }, + { + "epoch": 0.8900066035659256, + "grad_norm": 1.7648444090847863, + "learning_rate": 3.139305889479749e-07, + "loss": 0.3138, + "step": 40433 + }, + { + "epoch": 0.8900286154523442, + "grad_norm": 1.3959384793453815, + "learning_rate": 3.1380628486638807e-07, + "loss": 0.2538, + "step": 40434 + }, + { + "epoch": 0.890050627338763, + "grad_norm": 3.9881584065068845, + "learning_rate": 3.1368200460212883e-07, + "loss": 0.2249, + "step": 40435 + }, + { + "epoch": 0.8900726392251816, + "grad_norm": 1.584930389904755, + "learning_rate": 3.1355774815582716e-07, + "loss": 0.2508, + "step": 40436 + }, + { + "epoch": 0.8900946511116002, + "grad_norm": 1.2413775612231492, + "learning_rate": 3.1343351552811484e-07, + "loss": 0.23, + "step": 40437 + }, + { + "epoch": 0.890116662998019, + "grad_norm": 1.6940576780119596, + "learning_rate": 3.1330930671962465e-07, + "loss": 0.2541, + "step": 40438 + }, + { + "epoch": 0.8901386748844375, + "grad_norm": 0.5992821280542268, + "learning_rate": 3.1318512173098504e-07, + "loss": 0.4507, + "step": 40439 + }, + { + "epoch": 0.8901606867708562, + "grad_norm": 0.6255322807483495, + "learning_rate": 3.1306096056283096e-07, + "loss": 0.4419, + "step": 40440 + }, + { + "epoch": 0.890182698657275, + "grad_norm": 1.3954978546905124, + "learning_rate": 3.1293682321579087e-07, + "loss": 0.2956, + "step": 40441 + }, + { + "epoch": 0.8902047105436935, + "grad_norm": 2.641969582701925, + "learning_rate": 3.1281270969049595e-07, + "loss": 0.2831, + "step": 40442 + }, + { + "epoch": 0.8902267224301122, + "grad_norm": 1.2973612234742649, + "learning_rate": 3.1268861998757837e-07, + "loss": 0.2925, + "step": 40443 + }, + { + "epoch": 0.890248734316531, + "grad_norm": 1.7550663400873436, + "learning_rate": 3.125645541076666e-07, + "loss": 0.238, + "step": 40444 + }, + { + "epoch": 0.8902707462029495, + "grad_norm": 2.0667986766513744, + "learning_rate": 3.1244051205139294e-07, + "loss": 0.3577, + "step": 40445 + }, + { + "epoch": 0.8902927580893683, + "grad_norm": 1.7554336209326131, + "learning_rate": 3.1231649381938734e-07, + "loss": 0.2194, + "step": 40446 + }, + { + "epoch": 0.890314769975787, + "grad_norm": 4.456445506856721, + "learning_rate": 3.121924994122799e-07, + "loss": 0.2789, + "step": 40447 + }, + { + "epoch": 0.8903367818622056, + "grad_norm": 3.44745164178593, + "learning_rate": 3.120685288307013e-07, + "loss": 0.3142, + "step": 40448 + }, + { + "epoch": 0.8903587937486243, + "grad_norm": 2.954124906897243, + "learning_rate": 3.1194458207528087e-07, + "loss": 0.3247, + "step": 40449 + }, + { + "epoch": 0.890380805635043, + "grad_norm": 2.1790620188821204, + "learning_rate": 3.1182065914664936e-07, + "loss": 0.2593, + "step": 40450 + }, + { + "epoch": 0.8904028175214616, + "grad_norm": 2.056167983568199, + "learning_rate": 3.1169676004543624e-07, + "loss": 0.3321, + "step": 40451 + }, + { + "epoch": 0.8904248294078803, + "grad_norm": 2.112950225612876, + "learning_rate": 3.1157288477227097e-07, + "loss": 0.2234, + "step": 40452 + }, + { + "epoch": 0.890446841294299, + "grad_norm": 2.115894051225356, + "learning_rate": 3.114490333277842e-07, + "loss": 0.3157, + "step": 40453 + }, + { + "epoch": 0.8904688531807176, + "grad_norm": 1.8507939469670178, + "learning_rate": 3.1132520571260373e-07, + "loss": 0.2495, + "step": 40454 + }, + { + "epoch": 0.8904908650671363, + "grad_norm": 1.6583838731738347, + "learning_rate": 3.112014019273596e-07, + "loss": 0.297, + "step": 40455 + }, + { + "epoch": 0.890512876953555, + "grad_norm": 1.8252450994928637, + "learning_rate": 3.1107762197268187e-07, + "loss": 0.2745, + "step": 40456 + }, + { + "epoch": 0.8905348888399736, + "grad_norm": 2.3838555980623215, + "learning_rate": 3.109538658491984e-07, + "loss": 0.4562, + "step": 40457 + }, + { + "epoch": 0.8905569007263923, + "grad_norm": 0.6208109389986258, + "learning_rate": 3.108301335575398e-07, + "loss": 0.44, + "step": 40458 + }, + { + "epoch": 0.890578912612811, + "grad_norm": 1.9837920855931177, + "learning_rate": 3.107064250983327e-07, + "loss": 0.3444, + "step": 40459 + }, + { + "epoch": 0.8906009244992296, + "grad_norm": 1.5160959290851475, + "learning_rate": 3.1058274047220727e-07, + "loss": 0.273, + "step": 40460 + }, + { + "epoch": 0.8906229363856483, + "grad_norm": 1.778756829093137, + "learning_rate": 3.1045907967979236e-07, + "loss": 0.1718, + "step": 40461 + }, + { + "epoch": 0.8906449482720669, + "grad_norm": 1.7538050031777481, + "learning_rate": 3.1033544272171643e-07, + "loss": 0.263, + "step": 40462 + }, + { + "epoch": 0.8906669601584856, + "grad_norm": 0.6027017964902784, + "learning_rate": 3.1021182959860675e-07, + "loss": 0.4269, + "step": 40463 + }, + { + "epoch": 0.8906889720449043, + "grad_norm": 2.0005556503106665, + "learning_rate": 3.1008824031109166e-07, + "loss": 0.2745, + "step": 40464 + }, + { + "epoch": 0.8907109839313229, + "grad_norm": 0.6008388942250636, + "learning_rate": 3.099646748598001e-07, + "loss": 0.4262, + "step": 40465 + }, + { + "epoch": 0.8907329958177416, + "grad_norm": 0.5847032676599984, + "learning_rate": 3.0984113324536105e-07, + "loss": 0.4416, + "step": 40466 + }, + { + "epoch": 0.8907550077041603, + "grad_norm": 1.610410017114572, + "learning_rate": 3.097176154684006e-07, + "loss": 0.3121, + "step": 40467 + }, + { + "epoch": 0.8907770195905789, + "grad_norm": 1.5126892330225714, + "learning_rate": 3.095941215295467e-07, + "loss": 0.2648, + "step": 40468 + }, + { + "epoch": 0.8907990314769976, + "grad_norm": 1.4347069680603197, + "learning_rate": 3.0947065142942766e-07, + "loss": 0.3059, + "step": 40469 + }, + { + "epoch": 0.8908210433634163, + "grad_norm": 1.7998070307951013, + "learning_rate": 3.093472051686708e-07, + "loss": 0.3229, + "step": 40470 + }, + { + "epoch": 0.8908430552498349, + "grad_norm": 1.6959593597370177, + "learning_rate": 3.0922378274790385e-07, + "loss": 0.2268, + "step": 40471 + }, + { + "epoch": 0.8908650671362536, + "grad_norm": 1.5660630126769692, + "learning_rate": 3.0910038416775424e-07, + "loss": 0.2944, + "step": 40472 + }, + { + "epoch": 0.8908870790226723, + "grad_norm": 1.67860664866877, + "learning_rate": 3.0897700942884744e-07, + "loss": 0.3641, + "step": 40473 + }, + { + "epoch": 0.8909090909090909, + "grad_norm": 2.04483480497008, + "learning_rate": 3.08853658531813e-07, + "loss": 0.3493, + "step": 40474 + }, + { + "epoch": 0.8909311027955096, + "grad_norm": 1.5779255584367835, + "learning_rate": 3.087303314772749e-07, + "loss": 0.2866, + "step": 40475 + }, + { + "epoch": 0.8909531146819283, + "grad_norm": 1.5996007387908548, + "learning_rate": 3.0860702826586265e-07, + "loss": 0.3335, + "step": 40476 + }, + { + "epoch": 0.8909751265683469, + "grad_norm": 1.3688563235524116, + "learning_rate": 3.084837488982023e-07, + "loss": 0.2999, + "step": 40477 + }, + { + "epoch": 0.8909971384547656, + "grad_norm": 1.7827264491976857, + "learning_rate": 3.083604933749196e-07, + "loss": 0.324, + "step": 40478 + }, + { + "epoch": 0.8910191503411843, + "grad_norm": 1.4714403940674436, + "learning_rate": 3.082372616966417e-07, + "loss": 0.3622, + "step": 40479 + }, + { + "epoch": 0.8910411622276029, + "grad_norm": 1.5208817235985153, + "learning_rate": 3.081140538639943e-07, + "loss": 0.29, + "step": 40480 + }, + { + "epoch": 0.8910631741140216, + "grad_norm": 1.4712503099234424, + "learning_rate": 3.079908698776035e-07, + "loss": 0.2653, + "step": 40481 + }, + { + "epoch": 0.8910851860004403, + "grad_norm": 2.080080195303949, + "learning_rate": 3.0786770973809664e-07, + "loss": 0.3046, + "step": 40482 + }, + { + "epoch": 0.8911071978868589, + "grad_norm": 0.6165575444204244, + "learning_rate": 3.077445734460982e-07, + "loss": 0.4434, + "step": 40483 + }, + { + "epoch": 0.8911292097732776, + "grad_norm": 1.8035389318037651, + "learning_rate": 3.076214610022349e-07, + "loss": 0.3335, + "step": 40484 + }, + { + "epoch": 0.8911512216596963, + "grad_norm": 1.5688470652568889, + "learning_rate": 3.074983724071318e-07, + "loss": 0.2748, + "step": 40485 + }, + { + "epoch": 0.8911732335461149, + "grad_norm": 2.096521881579467, + "learning_rate": 3.07375307661415e-07, + "loss": 0.2483, + "step": 40486 + }, + { + "epoch": 0.8911952454325336, + "grad_norm": 1.434801997898824, + "learning_rate": 3.0725226676571076e-07, + "loss": 0.3084, + "step": 40487 + }, + { + "epoch": 0.8912172573189522, + "grad_norm": 1.4701195370320188, + "learning_rate": 3.071292497206424e-07, + "loss": 0.2937, + "step": 40488 + }, + { + "epoch": 0.8912392692053709, + "grad_norm": 1.295722529911404, + "learning_rate": 3.0700625652683725e-07, + "loss": 0.2866, + "step": 40489 + }, + { + "epoch": 0.8912612810917896, + "grad_norm": 2.1479596706952195, + "learning_rate": 3.0688328718491864e-07, + "loss": 0.2898, + "step": 40490 + }, + { + "epoch": 0.8912832929782082, + "grad_norm": 3.4408109044923982, + "learning_rate": 3.067603416955123e-07, + "loss": 0.2432, + "step": 40491 + }, + { + "epoch": 0.8913053048646269, + "grad_norm": 2.229023242884827, + "learning_rate": 3.0663742005924366e-07, + "loss": 0.3258, + "step": 40492 + }, + { + "epoch": 0.8913273167510456, + "grad_norm": 1.3639130405971092, + "learning_rate": 3.0651452227673684e-07, + "loss": 0.2598, + "step": 40493 + }, + { + "epoch": 0.8913493286374642, + "grad_norm": 1.2650241757507317, + "learning_rate": 3.0639164834861677e-07, + "loss": 0.252, + "step": 40494 + }, + { + "epoch": 0.8913713405238829, + "grad_norm": 2.561569907145969, + "learning_rate": 3.062687982755069e-07, + "loss": 0.2281, + "step": 40495 + }, + { + "epoch": 0.8913933524103016, + "grad_norm": 0.6145790434611252, + "learning_rate": 3.0614597205803285e-07, + "loss": 0.4542, + "step": 40496 + }, + { + "epoch": 0.8914153642967202, + "grad_norm": 1.5937468094884457, + "learning_rate": 3.060231696968191e-07, + "loss": 0.3139, + "step": 40497 + }, + { + "epoch": 0.8914373761831389, + "grad_norm": 1.886132844300879, + "learning_rate": 3.059003911924885e-07, + "loss": 0.326, + "step": 40498 + }, + { + "epoch": 0.8914593880695576, + "grad_norm": 1.3072075856067198, + "learning_rate": 3.057776365456666e-07, + "loss": 0.2831, + "step": 40499 + }, + { + "epoch": 0.8914813999559762, + "grad_norm": 0.6003923717816076, + "learning_rate": 3.0565490575697523e-07, + "loss": 0.4244, + "step": 40500 + }, + { + "epoch": 0.8915034118423949, + "grad_norm": 1.5087351137425862, + "learning_rate": 3.055321988270399e-07, + "loss": 0.2106, + "step": 40501 + }, + { + "epoch": 0.8915254237288136, + "grad_norm": 16.33718116416689, + "learning_rate": 3.0540951575648403e-07, + "loss": 0.3757, + "step": 40502 + }, + { + "epoch": 0.8915474356152322, + "grad_norm": 1.4655222456450459, + "learning_rate": 3.0528685654593103e-07, + "loss": 0.3023, + "step": 40503 + }, + { + "epoch": 0.8915694475016509, + "grad_norm": 1.3671555511022468, + "learning_rate": 3.051642211960032e-07, + "loss": 0.3133, + "step": 40504 + }, + { + "epoch": 0.8915914593880696, + "grad_norm": 0.6111387176674098, + "learning_rate": 3.0504160970732555e-07, + "loss": 0.4307, + "step": 40505 + }, + { + "epoch": 0.8916134712744882, + "grad_norm": 1.592360441847973, + "learning_rate": 3.049190220805193e-07, + "loss": 0.2426, + "step": 40506 + }, + { + "epoch": 0.8916354831609069, + "grad_norm": 2.224277279719443, + "learning_rate": 3.0479645831620994e-07, + "loss": 0.3288, + "step": 40507 + }, + { + "epoch": 0.8916574950473256, + "grad_norm": 1.1170510016206392, + "learning_rate": 3.0467391841501937e-07, + "loss": 0.1756, + "step": 40508 + }, + { + "epoch": 0.8916795069337442, + "grad_norm": 2.106632960729853, + "learning_rate": 3.0455140237756866e-07, + "loss": 0.4081, + "step": 40509 + }, + { + "epoch": 0.8917015188201629, + "grad_norm": 1.479861592488287, + "learning_rate": 3.044289102044834e-07, + "loss": 0.2839, + "step": 40510 + }, + { + "epoch": 0.8917235307065815, + "grad_norm": 1.3172285999182947, + "learning_rate": 3.043064418963826e-07, + "loss": 0.3035, + "step": 40511 + }, + { + "epoch": 0.8917455425930002, + "grad_norm": 1.5093570454835368, + "learning_rate": 3.0418399745389294e-07, + "loss": 0.226, + "step": 40512 + }, + { + "epoch": 0.8917675544794189, + "grad_norm": 1.5253016776414052, + "learning_rate": 3.0406157687763396e-07, + "loss": 0.279, + "step": 40513 + }, + { + "epoch": 0.8917895663658375, + "grad_norm": 0.632037531457022, + "learning_rate": 3.039391801682284e-07, + "loss": 0.476, + "step": 40514 + }, + { + "epoch": 0.8918115782522562, + "grad_norm": 1.915673713345327, + "learning_rate": 3.038168073262987e-07, + "loss": 0.2388, + "step": 40515 + }, + { + "epoch": 0.8918335901386749, + "grad_norm": 0.615346695802229, + "learning_rate": 3.0369445835246593e-07, + "loss": 0.4584, + "step": 40516 + }, + { + "epoch": 0.8918556020250935, + "grad_norm": 2.080125183559841, + "learning_rate": 3.0357213324735233e-07, + "loss": 0.3326, + "step": 40517 + }, + { + "epoch": 0.8918776139115122, + "grad_norm": 2.843580802083289, + "learning_rate": 3.0344983201158085e-07, + "loss": 0.273, + "step": 40518 + }, + { + "epoch": 0.8918996257979309, + "grad_norm": 2.8099312181497607, + "learning_rate": 3.033275546457715e-07, + "loss": 0.2345, + "step": 40519 + }, + { + "epoch": 0.8919216376843495, + "grad_norm": 1.7325666309580108, + "learning_rate": 3.032053011505465e-07, + "loss": 0.2975, + "step": 40520 + }, + { + "epoch": 0.8919436495707682, + "grad_norm": 1.7207960711110617, + "learning_rate": 3.030830715265265e-07, + "loss": 0.2966, + "step": 40521 + }, + { + "epoch": 0.8919656614571869, + "grad_norm": 1.5439755795282828, + "learning_rate": 3.0296086577433324e-07, + "loss": 0.252, + "step": 40522 + }, + { + "epoch": 0.8919876733436055, + "grad_norm": 1.8117692632464604, + "learning_rate": 3.0283868389458846e-07, + "loss": 0.1958, + "step": 40523 + }, + { + "epoch": 0.8920096852300242, + "grad_norm": 1.281889628799472, + "learning_rate": 3.0271652588791165e-07, + "loss": 0.2562, + "step": 40524 + }, + { + "epoch": 0.8920316971164429, + "grad_norm": 1.404918094975218, + "learning_rate": 3.025943917549257e-07, + "loss": 0.2462, + "step": 40525 + }, + { + "epoch": 0.8920537090028615, + "grad_norm": 2.3351397945847996, + "learning_rate": 3.0247228149624886e-07, + "loss": 0.3677, + "step": 40526 + }, + { + "epoch": 0.8920757208892802, + "grad_norm": 1.6115740361400008, + "learning_rate": 3.023501951125035e-07, + "loss": 0.233, + "step": 40527 + }, + { + "epoch": 0.8920977327756989, + "grad_norm": 1.6502201364216695, + "learning_rate": 3.022281326043103e-07, + "loss": 0.3031, + "step": 40528 + }, + { + "epoch": 0.8921197446621175, + "grad_norm": 0.6201845557844249, + "learning_rate": 3.0210609397228806e-07, + "loss": 0.4621, + "step": 40529 + }, + { + "epoch": 0.8921417565485362, + "grad_norm": 3.004221993884775, + "learning_rate": 3.0198407921705863e-07, + "loss": 0.3292, + "step": 40530 + }, + { + "epoch": 0.8921637684349549, + "grad_norm": 2.7737163323502614, + "learning_rate": 3.018620883392409e-07, + "loss": 0.3174, + "step": 40531 + }, + { + "epoch": 0.8921857803213735, + "grad_norm": 1.6164498019535607, + "learning_rate": 3.017401213394555e-07, + "loss": 0.306, + "step": 40532 + }, + { + "epoch": 0.8922077922077922, + "grad_norm": 0.6347123983131691, + "learning_rate": 3.016181782183231e-07, + "loss": 0.4561, + "step": 40533 + }, + { + "epoch": 0.8922298040942108, + "grad_norm": 2.740121275888175, + "learning_rate": 3.0149625897646194e-07, + "loss": 0.3309, + "step": 40534 + }, + { + "epoch": 0.8922518159806295, + "grad_norm": 1.5002393051209775, + "learning_rate": 3.013743636144928e-07, + "loss": 0.356, + "step": 40535 + }, + { + "epoch": 0.8922738278670482, + "grad_norm": 1.9256988488377178, + "learning_rate": 3.01252492133034e-07, + "loss": 0.3965, + "step": 40536 + }, + { + "epoch": 0.8922958397534668, + "grad_norm": 2.224793746064279, + "learning_rate": 3.011306445327061e-07, + "loss": 0.2544, + "step": 40537 + }, + { + "epoch": 0.8923178516398855, + "grad_norm": 1.924901020409626, + "learning_rate": 3.0100882081412876e-07, + "loss": 0.2927, + "step": 40538 + }, + { + "epoch": 0.8923398635263042, + "grad_norm": 1.5281261083060225, + "learning_rate": 3.008870209779197e-07, + "loss": 0.2858, + "step": 40539 + }, + { + "epoch": 0.8923618754127228, + "grad_norm": 2.6278919604446447, + "learning_rate": 3.0076524502469897e-07, + "loss": 0.2515, + "step": 40540 + }, + { + "epoch": 0.8923838872991415, + "grad_norm": 2.0467647020796838, + "learning_rate": 3.0064349295508554e-07, + "loss": 0.3071, + "step": 40541 + }, + { + "epoch": 0.8924058991855602, + "grad_norm": 1.5471705545250605, + "learning_rate": 3.005217647696962e-07, + "loss": 0.2766, + "step": 40542 + }, + { + "epoch": 0.8924279110719788, + "grad_norm": 0.6067019126731507, + "learning_rate": 3.004000604691526e-07, + "loss": 0.4324, + "step": 40543 + }, + { + "epoch": 0.8924499229583975, + "grad_norm": 0.6393538135264115, + "learning_rate": 3.002783800540726e-07, + "loss": 0.456, + "step": 40544 + }, + { + "epoch": 0.8924719348448162, + "grad_norm": 1.611794414587328, + "learning_rate": 3.0015672352507295e-07, + "loss": 0.3013, + "step": 40545 + }, + { + "epoch": 0.8924939467312348, + "grad_norm": 0.6094921854879928, + "learning_rate": 3.000350908827737e-07, + "loss": 0.4462, + "step": 40546 + }, + { + "epoch": 0.8925159586176535, + "grad_norm": 1.7271430909013643, + "learning_rate": 2.9991348212779045e-07, + "loss": 0.2466, + "step": 40547 + }, + { + "epoch": 0.8925379705040722, + "grad_norm": 1.9663718867278093, + "learning_rate": 2.9979189726074497e-07, + "loss": 0.2726, + "step": 40548 + }, + { + "epoch": 0.8925599823904908, + "grad_norm": 1.4521620076491188, + "learning_rate": 2.9967033628225397e-07, + "loss": 0.2387, + "step": 40549 + }, + { + "epoch": 0.8925819942769095, + "grad_norm": 1.882275588810602, + "learning_rate": 2.9954879919293355e-07, + "loss": 0.3794, + "step": 40550 + }, + { + "epoch": 0.8926040061633282, + "grad_norm": 1.7153410367236555, + "learning_rate": 2.9942728599340334e-07, + "loss": 0.2846, + "step": 40551 + }, + { + "epoch": 0.8926260180497468, + "grad_norm": 1.8514925697888924, + "learning_rate": 2.9930579668427937e-07, + "loss": 0.294, + "step": 40552 + }, + { + "epoch": 0.8926480299361655, + "grad_norm": 1.7927219778999495, + "learning_rate": 2.991843312661796e-07, + "loss": 0.2868, + "step": 40553 + }, + { + "epoch": 0.8926700418225842, + "grad_norm": 1.6698928119440408, + "learning_rate": 2.9906288973972296e-07, + "loss": 0.3172, + "step": 40554 + }, + { + "epoch": 0.8926920537090028, + "grad_norm": 2.147494079582523, + "learning_rate": 2.9894147210552447e-07, + "loss": 0.3143, + "step": 40555 + }, + { + "epoch": 0.8927140655954215, + "grad_norm": 1.9142085548330738, + "learning_rate": 2.9882007836420314e-07, + "loss": 0.3135, + "step": 40556 + }, + { + "epoch": 0.8927360774818401, + "grad_norm": 1.6798057406245797, + "learning_rate": 2.9869870851637396e-07, + "loss": 0.2504, + "step": 40557 + }, + { + "epoch": 0.8927580893682588, + "grad_norm": 1.952574230638054, + "learning_rate": 2.985773625626542e-07, + "loss": 0.3329, + "step": 40558 + }, + { + "epoch": 0.8927801012546775, + "grad_norm": 2.195315395178069, + "learning_rate": 2.9845604050366296e-07, + "loss": 0.3469, + "step": 40559 + }, + { + "epoch": 0.8928021131410961, + "grad_norm": 6.280182056475272, + "learning_rate": 2.9833474234001347e-07, + "loss": 0.3327, + "step": 40560 + }, + { + "epoch": 0.8928241250275148, + "grad_norm": 2.1072163216946875, + "learning_rate": 2.9821346807232534e-07, + "loss": 0.2205, + "step": 40561 + }, + { + "epoch": 0.8928461369139336, + "grad_norm": 2.0405442570501844, + "learning_rate": 2.9809221770121245e-07, + "loss": 0.272, + "step": 40562 + }, + { + "epoch": 0.8928681488003521, + "grad_norm": 2.6533298246363626, + "learning_rate": 2.9797099122729156e-07, + "loss": 0.2469, + "step": 40563 + }, + { + "epoch": 0.8928901606867709, + "grad_norm": 0.6284606700457138, + "learning_rate": 2.978497886511805e-07, + "loss": 0.4573, + "step": 40564 + }, + { + "epoch": 0.8929121725731896, + "grad_norm": 1.8442690306654261, + "learning_rate": 2.9772860997349273e-07, + "loss": 0.3912, + "step": 40565 + }, + { + "epoch": 0.8929341844596081, + "grad_norm": 1.3124478720124655, + "learning_rate": 2.976074551948466e-07, + "loss": 0.2365, + "step": 40566 + }, + { + "epoch": 0.8929561963460269, + "grad_norm": 2.0616426788403044, + "learning_rate": 2.97486324315856e-07, + "loss": 0.3246, + "step": 40567 + }, + { + "epoch": 0.8929782082324456, + "grad_norm": 1.8737235326866768, + "learning_rate": 2.973652173371372e-07, + "loss": 0.2809, + "step": 40568 + }, + { + "epoch": 0.8930002201188642, + "grad_norm": 1.5629775808144704, + "learning_rate": 2.9724413425930685e-07, + "loss": 0.3331, + "step": 40569 + }, + { + "epoch": 0.8930222320052829, + "grad_norm": 1.6400600498096463, + "learning_rate": 2.971230750829779e-07, + "loss": 0.2574, + "step": 40570 + }, + { + "epoch": 0.8930442438917016, + "grad_norm": 1.9854837608071654, + "learning_rate": 2.970020398087681e-07, + "loss": 0.3164, + "step": 40571 + }, + { + "epoch": 0.8930662557781202, + "grad_norm": 1.4059945664612898, + "learning_rate": 2.9688102843729084e-07, + "loss": 0.3383, + "step": 40572 + }, + { + "epoch": 0.8930882676645389, + "grad_norm": 1.6148712082639212, + "learning_rate": 2.9676004096916176e-07, + "loss": 0.2389, + "step": 40573 + }, + { + "epoch": 0.8931102795509576, + "grad_norm": 1.4307904576268073, + "learning_rate": 2.9663907740499655e-07, + "loss": 0.2551, + "step": 40574 + }, + { + "epoch": 0.8931322914373762, + "grad_norm": 2.763833907911703, + "learning_rate": 2.9651813774540903e-07, + "loss": 0.2691, + "step": 40575 + }, + { + "epoch": 0.8931543033237949, + "grad_norm": 2.4643460962420045, + "learning_rate": 2.9639722199101437e-07, + "loss": 0.2788, + "step": 40576 + }, + { + "epoch": 0.8931763152102136, + "grad_norm": 1.8900355711806698, + "learning_rate": 2.962763301424271e-07, + "loss": 0.2522, + "step": 40577 + }, + { + "epoch": 0.8931983270966322, + "grad_norm": 1.612524835348938, + "learning_rate": 2.961554622002599e-07, + "loss": 0.2431, + "step": 40578 + }, + { + "epoch": 0.8932203389830509, + "grad_norm": 1.6260434444161826, + "learning_rate": 2.9603461816512966e-07, + "loss": 0.1822, + "step": 40579 + }, + { + "epoch": 0.8932423508694695, + "grad_norm": 1.4228170236669433, + "learning_rate": 2.9591379803764973e-07, + "loss": 0.2867, + "step": 40580 + }, + { + "epoch": 0.8932643627558882, + "grad_norm": 0.6334821456604638, + "learning_rate": 2.9579300181843406e-07, + "loss": 0.4506, + "step": 40581 + }, + { + "epoch": 0.8932863746423069, + "grad_norm": 3.9836094291358823, + "learning_rate": 2.956722295080966e-07, + "loss": 0.2653, + "step": 40582 + }, + { + "epoch": 0.8933083865287255, + "grad_norm": 0.6000111014364492, + "learning_rate": 2.955514811072496e-07, + "loss": 0.4352, + "step": 40583 + }, + { + "epoch": 0.8933303984151442, + "grad_norm": 1.7697219413167973, + "learning_rate": 2.954307566165093e-07, + "loss": 0.2651, + "step": 40584 + }, + { + "epoch": 0.8933524103015629, + "grad_norm": 2.035358505660187, + "learning_rate": 2.9531005603648854e-07, + "loss": 0.2665, + "step": 40585 + }, + { + "epoch": 0.8933744221879815, + "grad_norm": 3.9597403940534095, + "learning_rate": 2.9518937936780014e-07, + "loss": 0.3268, + "step": 40586 + }, + { + "epoch": 0.8933964340744002, + "grad_norm": 2.4721156188823525, + "learning_rate": 2.9506872661105803e-07, + "loss": 0.2905, + "step": 40587 + }, + { + "epoch": 0.8934184459608189, + "grad_norm": 0.6121428210567192, + "learning_rate": 2.9494809776687394e-07, + "loss": 0.4474, + "step": 40588 + }, + { + "epoch": 0.8934404578472375, + "grad_norm": 2.7376726961293794, + "learning_rate": 2.9482749283586297e-07, + "loss": 0.2677, + "step": 40589 + }, + { + "epoch": 0.8934624697336562, + "grad_norm": 1.533653923299162, + "learning_rate": 2.947069118186374e-07, + "loss": 0.3031, + "step": 40590 + }, + { + "epoch": 0.8934844816200749, + "grad_norm": 0.6541772247863848, + "learning_rate": 2.9458635471580955e-07, + "loss": 0.4484, + "step": 40591 + }, + { + "epoch": 0.8935064935064935, + "grad_norm": 1.6588227261811148, + "learning_rate": 2.9446582152799274e-07, + "loss": 0.278, + "step": 40592 + }, + { + "epoch": 0.8935285053929122, + "grad_norm": 1.3834560612687998, + "learning_rate": 2.943453122557993e-07, + "loss": 0.307, + "step": 40593 + }, + { + "epoch": 0.8935505172793309, + "grad_norm": 1.6997365299409795, + "learning_rate": 2.94224826899841e-07, + "loss": 0.3028, + "step": 40594 + }, + { + "epoch": 0.8935725291657495, + "grad_norm": 1.7179660565979689, + "learning_rate": 2.9410436546073227e-07, + "loss": 0.3232, + "step": 40595 + }, + { + "epoch": 0.8935945410521682, + "grad_norm": 1.7885989621849618, + "learning_rate": 2.9398392793908324e-07, + "loss": 0.2812, + "step": 40596 + }, + { + "epoch": 0.8936165529385869, + "grad_norm": 1.8385719323871592, + "learning_rate": 2.938635143355073e-07, + "loss": 0.2713, + "step": 40597 + }, + { + "epoch": 0.8936385648250055, + "grad_norm": 2.230705151301537, + "learning_rate": 2.9374312465061563e-07, + "loss": 0.3076, + "step": 40598 + }, + { + "epoch": 0.8936605767114242, + "grad_norm": 1.8874245810930597, + "learning_rate": 2.936227588850199e-07, + "loss": 0.3084, + "step": 40599 + }, + { + "epoch": 0.8936825885978429, + "grad_norm": 1.5432724465349799, + "learning_rate": 2.9350241703933357e-07, + "loss": 0.2797, + "step": 40600 + }, + { + "epoch": 0.8937046004842615, + "grad_norm": 1.9265138859712225, + "learning_rate": 2.9338209911416616e-07, + "loss": 0.3404, + "step": 40601 + }, + { + "epoch": 0.8937266123706802, + "grad_norm": 1.4397502176921972, + "learning_rate": 2.93261805110131e-07, + "loss": 0.262, + "step": 40602 + }, + { + "epoch": 0.8937486242570988, + "grad_norm": 2.18867051833942, + "learning_rate": 2.9314153502783827e-07, + "loss": 0.3295, + "step": 40603 + }, + { + "epoch": 0.8937706361435175, + "grad_norm": 1.8184679378919677, + "learning_rate": 2.9302128886789905e-07, + "loss": 0.275, + "step": 40604 + }, + { + "epoch": 0.8937926480299362, + "grad_norm": 1.3740290771143597, + "learning_rate": 2.9290106663092566e-07, + "loss": 0.2879, + "step": 40605 + }, + { + "epoch": 0.8938146599163548, + "grad_norm": 3.2681542510000305, + "learning_rate": 2.9278086831752817e-07, + "loss": 0.3228, + "step": 40606 + }, + { + "epoch": 0.8938366718027735, + "grad_norm": 0.626897081050037, + "learning_rate": 2.926606939283183e-07, + "loss": 0.4417, + "step": 40607 + }, + { + "epoch": 0.8938586836891922, + "grad_norm": 1.409197045322876, + "learning_rate": 2.925405434639056e-07, + "loss": 0.3016, + "step": 40608 + }, + { + "epoch": 0.8938806955756108, + "grad_norm": 0.6074996469335348, + "learning_rate": 2.924204169249012e-07, + "loss": 0.4504, + "step": 40609 + }, + { + "epoch": 0.8939027074620295, + "grad_norm": 0.6102590153094462, + "learning_rate": 2.9230031431191683e-07, + "loss": 0.4316, + "step": 40610 + }, + { + "epoch": 0.8939247193484482, + "grad_norm": 0.6264423476167956, + "learning_rate": 2.9218023562556154e-07, + "loss": 0.4339, + "step": 40611 + }, + { + "epoch": 0.8939467312348668, + "grad_norm": 1.4934559410708605, + "learning_rate": 2.9206018086644636e-07, + "loss": 0.3372, + "step": 40612 + }, + { + "epoch": 0.8939687431212855, + "grad_norm": 1.5604719623122998, + "learning_rate": 2.9194015003518095e-07, + "loss": 0.3647, + "step": 40613 + }, + { + "epoch": 0.8939907550077042, + "grad_norm": 2.068753029582967, + "learning_rate": 2.918201431323747e-07, + "loss": 0.3137, + "step": 40614 + }, + { + "epoch": 0.8940127668941228, + "grad_norm": 1.5035262309169508, + "learning_rate": 2.917001601586394e-07, + "loss": 0.3263, + "step": 40615 + }, + { + "epoch": 0.8940347787805415, + "grad_norm": 2.0061458310100155, + "learning_rate": 2.9158020111458297e-07, + "loss": 0.2664, + "step": 40616 + }, + { + "epoch": 0.8940567906669602, + "grad_norm": 1.4983548600130256, + "learning_rate": 2.91460266000817e-07, + "loss": 0.2522, + "step": 40617 + }, + { + "epoch": 0.8940788025533788, + "grad_norm": 1.698201806457437, + "learning_rate": 2.9134035481795e-07, + "loss": 0.2406, + "step": 40618 + }, + { + "epoch": 0.8941008144397975, + "grad_norm": 1.9681601388928813, + "learning_rate": 2.9122046756659086e-07, + "loss": 0.2623, + "step": 40619 + }, + { + "epoch": 0.8941228263262162, + "grad_norm": 2.1103447061719582, + "learning_rate": 2.911006042473491e-07, + "loss": 0.2712, + "step": 40620 + }, + { + "epoch": 0.8941448382126348, + "grad_norm": 5.329253373309704, + "learning_rate": 2.909807648608348e-07, + "loss": 0.2647, + "step": 40621 + }, + { + "epoch": 0.8941668500990535, + "grad_norm": 1.8689870168004163, + "learning_rate": 2.9086094940765695e-07, + "loss": 0.41, + "step": 40622 + }, + { + "epoch": 0.8941888619854722, + "grad_norm": 1.5213842780207858, + "learning_rate": 2.9074115788842394e-07, + "loss": 0.2691, + "step": 40623 + }, + { + "epoch": 0.8942108738718908, + "grad_norm": 1.6883340396339153, + "learning_rate": 2.9062139030374417e-07, + "loss": 0.2445, + "step": 40624 + }, + { + "epoch": 0.8942328857583095, + "grad_norm": 1.4952364897895276, + "learning_rate": 2.905016466542271e-07, + "loss": 0.304, + "step": 40625 + }, + { + "epoch": 0.8942548976447282, + "grad_norm": 1.6414106842071126, + "learning_rate": 2.9038192694048173e-07, + "loss": 0.2611, + "step": 40626 + }, + { + "epoch": 0.8942769095311468, + "grad_norm": 1.316840592624066, + "learning_rate": 2.9026223116311536e-07, + "loss": 0.2429, + "step": 40627 + }, + { + "epoch": 0.8942989214175655, + "grad_norm": 1.4736845883852208, + "learning_rate": 2.901425593227375e-07, + "loss": 0.2532, + "step": 40628 + }, + { + "epoch": 0.8943209333039841, + "grad_norm": 1.7599937376777863, + "learning_rate": 2.900229114199554e-07, + "loss": 0.2395, + "step": 40629 + }, + { + "epoch": 0.8943429451904028, + "grad_norm": 1.5695223672215035, + "learning_rate": 2.89903287455377e-07, + "loss": 0.2995, + "step": 40630 + }, + { + "epoch": 0.8943649570768215, + "grad_norm": 1.8433698754334924, + "learning_rate": 2.897836874296123e-07, + "loss": 0.2497, + "step": 40631 + }, + { + "epoch": 0.8943869689632401, + "grad_norm": 1.8090054755387708, + "learning_rate": 2.8966411134326635e-07, + "loss": 0.2487, + "step": 40632 + }, + { + "epoch": 0.8944089808496588, + "grad_norm": 1.795654097535168, + "learning_rate": 2.8954455919694925e-07, + "loss": 0.2915, + "step": 40633 + }, + { + "epoch": 0.8944309927360775, + "grad_norm": 1.702598143555876, + "learning_rate": 2.894250309912666e-07, + "loss": 0.2951, + "step": 40634 + }, + { + "epoch": 0.8944530046224961, + "grad_norm": 1.2237676130629285, + "learning_rate": 2.8930552672682744e-07, + "loss": 0.2789, + "step": 40635 + }, + { + "epoch": 0.8944750165089148, + "grad_norm": 1.5324775111741626, + "learning_rate": 2.8918604640423897e-07, + "loss": 0.2439, + "step": 40636 + }, + { + "epoch": 0.8944970283953335, + "grad_norm": 1.4732214925831009, + "learning_rate": 2.890665900241074e-07, + "loss": 0.2439, + "step": 40637 + }, + { + "epoch": 0.8945190402817521, + "grad_norm": 1.3684059442483922, + "learning_rate": 2.889471575870417e-07, + "loss": 0.2804, + "step": 40638 + }, + { + "epoch": 0.8945410521681708, + "grad_norm": 1.7988797465432118, + "learning_rate": 2.8882774909364644e-07, + "loss": 0.3354, + "step": 40639 + }, + { + "epoch": 0.8945630640545895, + "grad_norm": 0.633279341613557, + "learning_rate": 2.887083645445299e-07, + "loss": 0.4641, + "step": 40640 + }, + { + "epoch": 0.8945850759410081, + "grad_norm": 2.0772075930889877, + "learning_rate": 2.8858900394029996e-07, + "loss": 0.1629, + "step": 40641 + }, + { + "epoch": 0.8946070878274268, + "grad_norm": 1.4684347107557034, + "learning_rate": 2.884696672815607e-07, + "loss": 0.2779, + "step": 40642 + }, + { + "epoch": 0.8946290997138455, + "grad_norm": 1.276976251433999, + "learning_rate": 2.8835035456892093e-07, + "loss": 0.2018, + "step": 40643 + }, + { + "epoch": 0.8946511116002641, + "grad_norm": 1.4352523555494199, + "learning_rate": 2.882310658029858e-07, + "loss": 0.2251, + "step": 40644 + }, + { + "epoch": 0.8946731234866828, + "grad_norm": 1.5307046157507456, + "learning_rate": 2.881118009843614e-07, + "loss": 0.2442, + "step": 40645 + }, + { + "epoch": 0.8946951353731015, + "grad_norm": 14.541578554736557, + "learning_rate": 2.879925601136552e-07, + "loss": 0.2999, + "step": 40646 + }, + { + "epoch": 0.8947171472595201, + "grad_norm": 0.6191617201943264, + "learning_rate": 2.8787334319147153e-07, + "loss": 0.4659, + "step": 40647 + }, + { + "epoch": 0.8947391591459388, + "grad_norm": 1.9993267901371996, + "learning_rate": 2.8775415021841834e-07, + "loss": 0.4008, + "step": 40648 + }, + { + "epoch": 0.8947611710323575, + "grad_norm": 1.8914761272812717, + "learning_rate": 2.876349811951001e-07, + "loss": 0.3138, + "step": 40649 + }, + { + "epoch": 0.8947831829187761, + "grad_norm": 1.2881290901832019, + "learning_rate": 2.875158361221214e-07, + "loss": 0.2302, + "step": 40650 + }, + { + "epoch": 0.8948051948051948, + "grad_norm": 2.0256881522977537, + "learning_rate": 2.8739671500009004e-07, + "loss": 0.2829, + "step": 40651 + }, + { + "epoch": 0.8948272066916134, + "grad_norm": 1.8711267301694072, + "learning_rate": 2.8727761782960996e-07, + "loss": 0.1891, + "step": 40652 + }, + { + "epoch": 0.8948492185780321, + "grad_norm": 1.3387423270870227, + "learning_rate": 2.871585446112879e-07, + "loss": 0.2846, + "step": 40653 + }, + { + "epoch": 0.8948712304644508, + "grad_norm": 1.5564071452776607, + "learning_rate": 2.8703949534572786e-07, + "loss": 0.3218, + "step": 40654 + }, + { + "epoch": 0.8948932423508694, + "grad_norm": 1.7824603953770881, + "learning_rate": 2.869204700335343e-07, + "loss": 0.3416, + "step": 40655 + }, + { + "epoch": 0.8949152542372881, + "grad_norm": 1.1926823936892694, + "learning_rate": 2.8680146867531346e-07, + "loss": 0.2417, + "step": 40656 + }, + { + "epoch": 0.8949372661237068, + "grad_norm": 0.6098954192209721, + "learning_rate": 2.8668249127166925e-07, + "loss": 0.443, + "step": 40657 + }, + { + "epoch": 0.8949592780101254, + "grad_norm": 3.083046411879993, + "learning_rate": 2.865635378232079e-07, + "loss": 0.2186, + "step": 40658 + }, + { + "epoch": 0.8949812898965441, + "grad_norm": 2.5292618488184107, + "learning_rate": 2.864446083305328e-07, + "loss": 0.2525, + "step": 40659 + }, + { + "epoch": 0.8950033017829628, + "grad_norm": 2.0392561176918544, + "learning_rate": 2.863257027942473e-07, + "loss": 0.2881, + "step": 40660 + }, + { + "epoch": 0.8950253136693814, + "grad_norm": 0.6255905940461874, + "learning_rate": 2.8620682121495767e-07, + "loss": 0.441, + "step": 40661 + }, + { + "epoch": 0.8950473255558001, + "grad_norm": 0.6262275747918058, + "learning_rate": 2.8608796359326787e-07, + "loss": 0.4422, + "step": 40662 + }, + { + "epoch": 0.8950693374422188, + "grad_norm": 1.5358984233812647, + "learning_rate": 2.8596912992978063e-07, + "loss": 0.2993, + "step": 40663 + }, + { + "epoch": 0.8950913493286374, + "grad_norm": 1.536687681230319, + "learning_rate": 2.8585032022510115e-07, + "loss": 0.345, + "step": 40664 + }, + { + "epoch": 0.8951133612150561, + "grad_norm": 5.155328901033487, + "learning_rate": 2.857315344798328e-07, + "loss": 0.2551, + "step": 40665 + }, + { + "epoch": 0.8951353731014748, + "grad_norm": 1.7297891848089952, + "learning_rate": 2.85612772694579e-07, + "loss": 0.2271, + "step": 40666 + }, + { + "epoch": 0.8951573849878934, + "grad_norm": 2.2452590738776026, + "learning_rate": 2.854940348699448e-07, + "loss": 0.4051, + "step": 40667 + }, + { + "epoch": 0.8951793968743121, + "grad_norm": 2.164054729999326, + "learning_rate": 2.853753210065319e-07, + "loss": 0.3043, + "step": 40668 + }, + { + "epoch": 0.8952014087607308, + "grad_norm": 1.515426989458559, + "learning_rate": 2.8525663110494485e-07, + "loss": 0.3316, + "step": 40669 + }, + { + "epoch": 0.8952234206471494, + "grad_norm": 1.2544605644338735, + "learning_rate": 2.8513796516578595e-07, + "loss": 0.1984, + "step": 40670 + }, + { + "epoch": 0.8952454325335681, + "grad_norm": 1.667685807750684, + "learning_rate": 2.8501932318965864e-07, + "loss": 0.2742, + "step": 40671 + }, + { + "epoch": 0.8952674444199868, + "grad_norm": 1.7435488782261421, + "learning_rate": 2.849007051771674e-07, + "loss": 0.3537, + "step": 40672 + }, + { + "epoch": 0.8952894563064054, + "grad_norm": 0.6256282376015458, + "learning_rate": 2.847821111289123e-07, + "loss": 0.4275, + "step": 40673 + }, + { + "epoch": 0.8953114681928241, + "grad_norm": 2.6362565869489525, + "learning_rate": 2.846635410454984e-07, + "loss": 0.2363, + "step": 40674 + }, + { + "epoch": 0.8953334800792427, + "grad_norm": 1.855592915844472, + "learning_rate": 2.8454499492752695e-07, + "loss": 0.3, + "step": 40675 + }, + { + "epoch": 0.8953554919656614, + "grad_norm": 1.7306001832859115, + "learning_rate": 2.844264727756008e-07, + "loss": 0.2045, + "step": 40676 + }, + { + "epoch": 0.8953775038520801, + "grad_norm": 1.4639428314867937, + "learning_rate": 2.843079745903238e-07, + "loss": 0.2609, + "step": 40677 + }, + { + "epoch": 0.8953995157384987, + "grad_norm": 1.174492783526607, + "learning_rate": 2.8418950037229554e-07, + "loss": 0.2619, + "step": 40678 + }, + { + "epoch": 0.8954215276249174, + "grad_norm": 2.553616045872241, + "learning_rate": 2.840710501221205e-07, + "loss": 0.3489, + "step": 40679 + }, + { + "epoch": 0.8954435395113362, + "grad_norm": 1.766306675087083, + "learning_rate": 2.839526238403989e-07, + "loss": 0.3724, + "step": 40680 + }, + { + "epoch": 0.8954655513977547, + "grad_norm": 1.3715986203025996, + "learning_rate": 2.83834221527734e-07, + "loss": 0.3011, + "step": 40681 + }, + { + "epoch": 0.8954875632841734, + "grad_norm": 0.5970336688582296, + "learning_rate": 2.83715843184727e-07, + "loss": 0.4282, + "step": 40682 + }, + { + "epoch": 0.8955095751705922, + "grad_norm": 1.7573510415041715, + "learning_rate": 2.8359748881197914e-07, + "loss": 0.3058, + "step": 40683 + }, + { + "epoch": 0.8955315870570107, + "grad_norm": 1.525560908704033, + "learning_rate": 2.8347915841009377e-07, + "loss": 0.2494, + "step": 40684 + }, + { + "epoch": 0.8955535989434295, + "grad_norm": 1.4726102686728888, + "learning_rate": 2.8336085197967045e-07, + "loss": 0.2947, + "step": 40685 + }, + { + "epoch": 0.8955756108298482, + "grad_norm": 2.4001474253812867, + "learning_rate": 2.832425695213092e-07, + "loss": 0.4099, + "step": 40686 + }, + { + "epoch": 0.8955976227162668, + "grad_norm": 2.3288106986324437, + "learning_rate": 2.8312431103561464e-07, + "loss": 0.2791, + "step": 40687 + }, + { + "epoch": 0.8956196346026855, + "grad_norm": 2.0204886311356143, + "learning_rate": 2.8300607652318567e-07, + "loss": 0.2706, + "step": 40688 + }, + { + "epoch": 0.8956416464891042, + "grad_norm": 0.5898084955554159, + "learning_rate": 2.82887865984624e-07, + "loss": 0.4295, + "step": 40689 + }, + { + "epoch": 0.8956636583755228, + "grad_norm": 0.6156478189573382, + "learning_rate": 2.8276967942053035e-07, + "loss": 0.4531, + "step": 40690 + }, + { + "epoch": 0.8956856702619415, + "grad_norm": 1.4593393036062723, + "learning_rate": 2.8265151683150413e-07, + "loss": 0.2783, + "step": 40691 + }, + { + "epoch": 0.8957076821483602, + "grad_norm": 1.7274120311068544, + "learning_rate": 2.825333782181472e-07, + "loss": 0.3125, + "step": 40692 + }, + { + "epoch": 0.8957296940347788, + "grad_norm": 2.385166113982714, + "learning_rate": 2.8241526358105955e-07, + "loss": 0.2827, + "step": 40693 + }, + { + "epoch": 0.8957517059211975, + "grad_norm": 0.5995998704894739, + "learning_rate": 2.8229717292084233e-07, + "loss": 0.4598, + "step": 40694 + }, + { + "epoch": 0.8957737178076162, + "grad_norm": 1.8236306880647306, + "learning_rate": 2.821791062380952e-07, + "loss": 0.2585, + "step": 40695 + }, + { + "epoch": 0.8957957296940348, + "grad_norm": 1.3256654903084626, + "learning_rate": 2.82061063533417e-07, + "loss": 0.2899, + "step": 40696 + }, + { + "epoch": 0.8958177415804535, + "grad_norm": 2.0338742063304607, + "learning_rate": 2.81943044807409e-07, + "loss": 0.3145, + "step": 40697 + }, + { + "epoch": 0.8958397534668721, + "grad_norm": 2.8059094536874456, + "learning_rate": 2.818250500606706e-07, + "loss": 0.2523, + "step": 40698 + }, + { + "epoch": 0.8958617653532908, + "grad_norm": 0.6297939700083662, + "learning_rate": 2.817070792938026e-07, + "loss": 0.4311, + "step": 40699 + }, + { + "epoch": 0.8958837772397095, + "grad_norm": 0.6324942426817657, + "learning_rate": 2.815891325074038e-07, + "loss": 0.4421, + "step": 40700 + }, + { + "epoch": 0.8959057891261281, + "grad_norm": 1.8211328433483507, + "learning_rate": 2.814712097020722e-07, + "loss": 0.2809, + "step": 40701 + }, + { + "epoch": 0.8959278010125468, + "grad_norm": 8.33366862591459, + "learning_rate": 2.813533108784089e-07, + "loss": 0.3137, + "step": 40702 + }, + { + "epoch": 0.8959498128989655, + "grad_norm": 2.4614043773371934, + "learning_rate": 2.8123543603701344e-07, + "loss": 0.2851, + "step": 40703 + }, + { + "epoch": 0.8959718247853841, + "grad_norm": 4.1315286117356935, + "learning_rate": 2.8111758517848307e-07, + "loss": 0.2301, + "step": 40704 + }, + { + "epoch": 0.8959938366718028, + "grad_norm": 0.5952813296297988, + "learning_rate": 2.8099975830341854e-07, + "loss": 0.4036, + "step": 40705 + }, + { + "epoch": 0.8960158485582215, + "grad_norm": 2.124729540882329, + "learning_rate": 2.808819554124176e-07, + "loss": 0.1911, + "step": 40706 + }, + { + "epoch": 0.8960378604446401, + "grad_norm": 1.6065778129411166, + "learning_rate": 2.8076417650607867e-07, + "loss": 0.3725, + "step": 40707 + }, + { + "epoch": 0.8960598723310588, + "grad_norm": 1.8097389916619828, + "learning_rate": 2.8064642158500246e-07, + "loss": 0.3122, + "step": 40708 + }, + { + "epoch": 0.8960818842174775, + "grad_norm": 1.5865178871875314, + "learning_rate": 2.8052869064978507e-07, + "loss": 0.2531, + "step": 40709 + }, + { + "epoch": 0.8961038961038961, + "grad_norm": 1.5233844821276292, + "learning_rate": 2.8041098370102657e-07, + "loss": 0.1827, + "step": 40710 + }, + { + "epoch": 0.8961259079903148, + "grad_norm": 1.4991060675716796, + "learning_rate": 2.8029330073932327e-07, + "loss": 0.2279, + "step": 40711 + }, + { + "epoch": 0.8961479198767335, + "grad_norm": 2.2758677719600566, + "learning_rate": 2.8017564176527513e-07, + "loss": 0.2678, + "step": 40712 + }, + { + "epoch": 0.8961699317631521, + "grad_norm": 0.6170851847476376, + "learning_rate": 2.8005800677948e-07, + "loss": 0.4396, + "step": 40713 + }, + { + "epoch": 0.8961919436495708, + "grad_norm": 1.2744172706708263, + "learning_rate": 2.799403957825342e-07, + "loss": 0.2415, + "step": 40714 + }, + { + "epoch": 0.8962139555359895, + "grad_norm": 1.8710044153075323, + "learning_rate": 2.7982280877503763e-07, + "loss": 0.3245, + "step": 40715 + }, + { + "epoch": 0.8962359674224081, + "grad_norm": 1.4650775049929943, + "learning_rate": 2.797052457575855e-07, + "loss": 0.3094, + "step": 40716 + }, + { + "epoch": 0.8962579793088268, + "grad_norm": 1.6599160206259191, + "learning_rate": 2.7958770673077727e-07, + "loss": 0.2593, + "step": 40717 + }, + { + "epoch": 0.8962799911952455, + "grad_norm": 1.4982891650080987, + "learning_rate": 2.794701916952103e-07, + "loss": 0.264, + "step": 40718 + }, + { + "epoch": 0.8963020030816641, + "grad_norm": 1.9114583215523455, + "learning_rate": 2.793527006514801e-07, + "loss": 0.3686, + "step": 40719 + }, + { + "epoch": 0.8963240149680828, + "grad_norm": 3.7907950585510735, + "learning_rate": 2.7923523360018576e-07, + "loss": 0.2684, + "step": 40720 + }, + { + "epoch": 0.8963460268545014, + "grad_norm": 1.899815586307679, + "learning_rate": 2.791177905419234e-07, + "loss": 0.2787, + "step": 40721 + }, + { + "epoch": 0.8963680387409201, + "grad_norm": 1.6260777242909654, + "learning_rate": 2.790003714772893e-07, + "loss": 0.2327, + "step": 40722 + }, + { + "epoch": 0.8963900506273388, + "grad_norm": 0.6086754124637681, + "learning_rate": 2.7888297640688177e-07, + "loss": 0.4586, + "step": 40723 + }, + { + "epoch": 0.8964120625137574, + "grad_norm": 1.4228256664711054, + "learning_rate": 2.787656053312965e-07, + "loss": 0.3309, + "step": 40724 + }, + { + "epoch": 0.8964340744001761, + "grad_norm": 4.028026317163112, + "learning_rate": 2.786482582511302e-07, + "loss": 0.3037, + "step": 40725 + }, + { + "epoch": 0.8964560862865948, + "grad_norm": 1.1429983016221803, + "learning_rate": 2.785309351669796e-07, + "loss": 0.1942, + "step": 40726 + }, + { + "epoch": 0.8964780981730134, + "grad_norm": 1.1592858592178568, + "learning_rate": 2.7841363607943927e-07, + "loss": 0.1951, + "step": 40727 + }, + { + "epoch": 0.8965001100594321, + "grad_norm": 1.7216637699835213, + "learning_rate": 2.782963609891076e-07, + "loss": 0.2955, + "step": 40728 + }, + { + "epoch": 0.8965221219458508, + "grad_norm": 1.452941214309641, + "learning_rate": 2.7817910989657915e-07, + "loss": 0.307, + "step": 40729 + }, + { + "epoch": 0.8965441338322694, + "grad_norm": 1.706311165527035, + "learning_rate": 2.7806188280245116e-07, + "loss": 0.2096, + "step": 40730 + }, + { + "epoch": 0.8965661457186881, + "grad_norm": 1.6043449873537838, + "learning_rate": 2.779446797073193e-07, + "loss": 0.3077, + "step": 40731 + }, + { + "epoch": 0.8965881576051068, + "grad_norm": 0.6282421997308072, + "learning_rate": 2.7782750061177757e-07, + "loss": 0.4489, + "step": 40732 + }, + { + "epoch": 0.8966101694915254, + "grad_norm": 1.6547252870708493, + "learning_rate": 2.7771034551642264e-07, + "loss": 0.2487, + "step": 40733 + }, + { + "epoch": 0.8966321813779441, + "grad_norm": 1.9750489788971737, + "learning_rate": 2.7759321442185015e-07, + "loss": 0.2213, + "step": 40734 + }, + { + "epoch": 0.8966541932643628, + "grad_norm": 2.120275289919755, + "learning_rate": 2.774761073286553e-07, + "loss": 0.3181, + "step": 40735 + }, + { + "epoch": 0.8966762051507814, + "grad_norm": 0.6207597955822517, + "learning_rate": 2.7735902423743355e-07, + "loss": 0.4471, + "step": 40736 + }, + { + "epoch": 0.8966982170372001, + "grad_norm": 2.236589947520104, + "learning_rate": 2.77241965148779e-07, + "loss": 0.3735, + "step": 40737 + }, + { + "epoch": 0.8967202289236188, + "grad_norm": 2.1483173776582656, + "learning_rate": 2.771249300632872e-07, + "loss": 0.3256, + "step": 40738 + }, + { + "epoch": 0.8967422408100374, + "grad_norm": 0.6170156613196589, + "learning_rate": 2.7700791898155277e-07, + "loss": 0.4579, + "step": 40739 + }, + { + "epoch": 0.8967642526964561, + "grad_norm": 1.7457260727937953, + "learning_rate": 2.768909319041718e-07, + "loss": 0.2589, + "step": 40740 + }, + { + "epoch": 0.8967862645828748, + "grad_norm": 1.647903855979021, + "learning_rate": 2.7677396883173723e-07, + "loss": 0.2687, + "step": 40741 + }, + { + "epoch": 0.8968082764692934, + "grad_norm": 1.3399474410313625, + "learning_rate": 2.7665702976484355e-07, + "loss": 0.2426, + "step": 40742 + }, + { + "epoch": 0.8968302883557121, + "grad_norm": 1.9470305537429078, + "learning_rate": 2.7654011470408524e-07, + "loss": 0.4004, + "step": 40743 + }, + { + "epoch": 0.8968523002421307, + "grad_norm": 1.7245683622218073, + "learning_rate": 2.764232236500575e-07, + "loss": 0.3192, + "step": 40744 + }, + { + "epoch": 0.8968743121285494, + "grad_norm": 1.344366978566451, + "learning_rate": 2.7630635660335304e-07, + "loss": 0.2613, + "step": 40745 + }, + { + "epoch": 0.8968963240149681, + "grad_norm": 1.5619598692046002, + "learning_rate": 2.76189513564567e-07, + "loss": 0.2964, + "step": 40746 + }, + { + "epoch": 0.8969183359013867, + "grad_norm": 1.8057885487068293, + "learning_rate": 2.7607269453429233e-07, + "loss": 0.3309, + "step": 40747 + }, + { + "epoch": 0.8969403477878054, + "grad_norm": 0.5972366658549653, + "learning_rate": 2.759558995131234e-07, + "loss": 0.4306, + "step": 40748 + }, + { + "epoch": 0.8969623596742241, + "grad_norm": 1.3362840131535028, + "learning_rate": 2.7583912850165375e-07, + "loss": 0.2036, + "step": 40749 + }, + { + "epoch": 0.8969843715606427, + "grad_norm": 1.654119653015017, + "learning_rate": 2.7572238150047613e-07, + "loss": 0.317, + "step": 40750 + }, + { + "epoch": 0.8970063834470614, + "grad_norm": 0.6267378358459542, + "learning_rate": 2.756056585101852e-07, + "loss": 0.4499, + "step": 40751 + }, + { + "epoch": 0.8970283953334801, + "grad_norm": 11.200836824959582, + "learning_rate": 2.754889595313731e-07, + "loss": 0.399, + "step": 40752 + }, + { + "epoch": 0.8970504072198987, + "grad_norm": 1.7756584755304592, + "learning_rate": 2.7537228456463283e-07, + "loss": 0.3083, + "step": 40753 + }, + { + "epoch": 0.8970724191063174, + "grad_norm": 0.5985847264061083, + "learning_rate": 2.7525563361055885e-07, + "loss": 0.4398, + "step": 40754 + }, + { + "epoch": 0.8970944309927361, + "grad_norm": 2.6633791972473966, + "learning_rate": 2.751390066697418e-07, + "loss": 0.3323, + "step": 40755 + }, + { + "epoch": 0.8971164428791547, + "grad_norm": 1.69430766628622, + "learning_rate": 2.7502240374277676e-07, + "loss": 0.3465, + "step": 40756 + }, + { + "epoch": 0.8971384547655734, + "grad_norm": 2.587007431749223, + "learning_rate": 2.74905824830255e-07, + "loss": 0.2868, + "step": 40757 + }, + { + "epoch": 0.8971604666519921, + "grad_norm": 1.5741024666564247, + "learning_rate": 2.7478926993276865e-07, + "loss": 0.3182, + "step": 40758 + }, + { + "epoch": 0.8971824785384107, + "grad_norm": 1.724509225914707, + "learning_rate": 2.746727390509113e-07, + "loss": 0.259, + "step": 40759 + }, + { + "epoch": 0.8972044904248294, + "grad_norm": 1.4121035736275767, + "learning_rate": 2.745562321852746e-07, + "loss": 0.1828, + "step": 40760 + }, + { + "epoch": 0.8972265023112481, + "grad_norm": 0.6138625014175629, + "learning_rate": 2.744397493364509e-07, + "loss": 0.4499, + "step": 40761 + }, + { + "epoch": 0.8972485141976667, + "grad_norm": 1.436028382499377, + "learning_rate": 2.743232905050325e-07, + "loss": 0.2676, + "step": 40762 + }, + { + "epoch": 0.8972705260840854, + "grad_norm": 0.5902500654634492, + "learning_rate": 2.7420685569160996e-07, + "loss": 0.4314, + "step": 40763 + }, + { + "epoch": 0.8972925379705041, + "grad_norm": 2.451752554237165, + "learning_rate": 2.740904448967757e-07, + "loss": 0.3383, + "step": 40764 + }, + { + "epoch": 0.8973145498569227, + "grad_norm": 0.6078891007315642, + "learning_rate": 2.7397405812112197e-07, + "loss": 0.4436, + "step": 40765 + }, + { + "epoch": 0.8973365617433414, + "grad_norm": 2.1367372887826064, + "learning_rate": 2.7385769536523997e-07, + "loss": 0.3192, + "step": 40766 + }, + { + "epoch": 0.8973585736297601, + "grad_norm": 1.9699428102586662, + "learning_rate": 2.73741356629722e-07, + "loss": 0.2751, + "step": 40767 + }, + { + "epoch": 0.8973805855161787, + "grad_norm": 2.580483565841064, + "learning_rate": 2.7362504191515706e-07, + "loss": 0.2953, + "step": 40768 + }, + { + "epoch": 0.8974025974025974, + "grad_norm": 1.6965434212366464, + "learning_rate": 2.735087512221379e-07, + "loss": 0.2756, + "step": 40769 + }, + { + "epoch": 0.897424609289016, + "grad_norm": 1.6051437218295679, + "learning_rate": 2.7339248455125534e-07, + "loss": 0.2739, + "step": 40770 + }, + { + "epoch": 0.8974466211754347, + "grad_norm": 1.6073562064734093, + "learning_rate": 2.7327624190310096e-07, + "loss": 0.3449, + "step": 40771 + }, + { + "epoch": 0.8974686330618534, + "grad_norm": 1.9651580461502587, + "learning_rate": 2.7316002327826496e-07, + "loss": 0.3173, + "step": 40772 + }, + { + "epoch": 0.897490644948272, + "grad_norm": 0.6195730768931966, + "learning_rate": 2.7304382867733734e-07, + "loss": 0.4477, + "step": 40773 + }, + { + "epoch": 0.8975126568346907, + "grad_norm": 0.6243906285969125, + "learning_rate": 2.729276581009094e-07, + "loss": 0.4648, + "step": 40774 + }, + { + "epoch": 0.8975346687211094, + "grad_norm": 2.565453168731445, + "learning_rate": 2.7281151154957106e-07, + "loss": 0.2767, + "step": 40775 + }, + { + "epoch": 0.897556680607528, + "grad_norm": 1.4245939768932612, + "learning_rate": 2.726953890239137e-07, + "loss": 0.3193, + "step": 40776 + }, + { + "epoch": 0.8975786924939467, + "grad_norm": 1.2448771261720784, + "learning_rate": 2.7257929052452726e-07, + "loss": 0.215, + "step": 40777 + }, + { + "epoch": 0.8976007043803654, + "grad_norm": 1.8204287748849644, + "learning_rate": 2.724632160520002e-07, + "loss": 0.2153, + "step": 40778 + }, + { + "epoch": 0.897622716266784, + "grad_norm": 1.9070651027323324, + "learning_rate": 2.7234716560692374e-07, + "loss": 0.3948, + "step": 40779 + }, + { + "epoch": 0.8976447281532027, + "grad_norm": 1.9839345349002913, + "learning_rate": 2.7223113918988795e-07, + "loss": 0.2889, + "step": 40780 + }, + { + "epoch": 0.8976667400396214, + "grad_norm": 6.2033037508883035, + "learning_rate": 2.721151368014824e-07, + "loss": 0.2582, + "step": 40781 + }, + { + "epoch": 0.89768875192604, + "grad_norm": 0.6196131197641324, + "learning_rate": 2.7199915844229653e-07, + "loss": 0.4551, + "step": 40782 + }, + { + "epoch": 0.8977107638124587, + "grad_norm": 1.9710447739421344, + "learning_rate": 2.718832041129188e-07, + "loss": 0.2974, + "step": 40783 + }, + { + "epoch": 0.8977327756988774, + "grad_norm": 1.9759506191566112, + "learning_rate": 2.7176727381393986e-07, + "loss": 0.3365, + "step": 40784 + }, + { + "epoch": 0.897754787585296, + "grad_norm": 2.0163279802180916, + "learning_rate": 2.716513675459487e-07, + "loss": 0.2236, + "step": 40785 + }, + { + "epoch": 0.8977767994717147, + "grad_norm": 3.669010279348793, + "learning_rate": 2.7153548530953367e-07, + "loss": 0.3787, + "step": 40786 + }, + { + "epoch": 0.8977988113581334, + "grad_norm": 1.6811585956001793, + "learning_rate": 2.7141962710528546e-07, + "loss": 0.295, + "step": 40787 + }, + { + "epoch": 0.897820823244552, + "grad_norm": 1.6608307093668184, + "learning_rate": 2.713037929337903e-07, + "loss": 0.2214, + "step": 40788 + }, + { + "epoch": 0.8978428351309707, + "grad_norm": 0.6138098737101462, + "learning_rate": 2.711879827956387e-07, + "loss": 0.4416, + "step": 40789 + }, + { + "epoch": 0.8978648470173894, + "grad_norm": 6.1887544765318845, + "learning_rate": 2.7107219669141926e-07, + "loss": 0.3258, + "step": 40790 + }, + { + "epoch": 0.897886858903808, + "grad_norm": 2.2699146199225417, + "learning_rate": 2.709564346217197e-07, + "loss": 0.2897, + "step": 40791 + }, + { + "epoch": 0.8979088707902267, + "grad_norm": 0.614687571821627, + "learning_rate": 2.708406965871291e-07, + "loss": 0.4434, + "step": 40792 + }, + { + "epoch": 0.8979308826766453, + "grad_norm": 1.4283452725957118, + "learning_rate": 2.707249825882352e-07, + "loss": 0.2181, + "step": 40793 + }, + { + "epoch": 0.897952894563064, + "grad_norm": 2.459242302351676, + "learning_rate": 2.70609292625626e-07, + "loss": 0.3478, + "step": 40794 + }, + { + "epoch": 0.8979749064494827, + "grad_norm": 0.6177030564918132, + "learning_rate": 2.704936266998903e-07, + "loss": 0.4645, + "step": 40795 + }, + { + "epoch": 0.8979969183359013, + "grad_norm": 2.039696502391675, + "learning_rate": 2.703779848116145e-07, + "loss": 0.2779, + "step": 40796 + }, + { + "epoch": 0.89801893022232, + "grad_norm": 0.6153931226579316, + "learning_rate": 2.7026236696138853e-07, + "loss": 0.4482, + "step": 40797 + }, + { + "epoch": 0.8980409421087387, + "grad_norm": 2.951435863932786, + "learning_rate": 2.701467731497975e-07, + "loss": 0.236, + "step": 40798 + }, + { + "epoch": 0.8980629539951573, + "grad_norm": 2.4750936992130854, + "learning_rate": 2.7003120337743106e-07, + "loss": 0.2058, + "step": 40799 + }, + { + "epoch": 0.898084965881576, + "grad_norm": 1.9449097179598567, + "learning_rate": 2.6991565764487524e-07, + "loss": 0.2703, + "step": 40800 + }, + { + "epoch": 0.8981069777679948, + "grad_norm": 1.54750017943072, + "learning_rate": 2.6980013595271747e-07, + "loss": 0.2609, + "step": 40801 + }, + { + "epoch": 0.8981289896544133, + "grad_norm": 4.981867605165228, + "learning_rate": 2.696846383015461e-07, + "loss": 0.2063, + "step": 40802 + }, + { + "epoch": 0.898151001540832, + "grad_norm": 1.4033302951770963, + "learning_rate": 2.695691646919468e-07, + "loss": 0.2497, + "step": 40803 + }, + { + "epoch": 0.8981730134272508, + "grad_norm": 1.5491805814993354, + "learning_rate": 2.694537151245064e-07, + "loss": 0.2531, + "step": 40804 + }, + { + "epoch": 0.8981950253136693, + "grad_norm": 1.9668384101144043, + "learning_rate": 2.6933828959981144e-07, + "loss": 0.3379, + "step": 40805 + }, + { + "epoch": 0.898217037200088, + "grad_norm": 1.6548157949998767, + "learning_rate": 2.6922288811845e-07, + "loss": 0.255, + "step": 40806 + }, + { + "epoch": 0.8982390490865068, + "grad_norm": 1.276142550870654, + "learning_rate": 2.6910751068100816e-07, + "loss": 0.2402, + "step": 40807 + }, + { + "epoch": 0.8982610609729254, + "grad_norm": 2.1407635511877707, + "learning_rate": 2.6899215728807216e-07, + "loss": 0.2661, + "step": 40808 + }, + { + "epoch": 0.8982830728593441, + "grad_norm": 1.5352733466786401, + "learning_rate": 2.6887682794022763e-07, + "loss": 0.2812, + "step": 40809 + }, + { + "epoch": 0.8983050847457628, + "grad_norm": 1.7072261101996695, + "learning_rate": 2.6876152263806076e-07, + "loss": 0.2614, + "step": 40810 + }, + { + "epoch": 0.8983270966321814, + "grad_norm": 0.6522817060746289, + "learning_rate": 2.686462413821583e-07, + "loss": 0.4614, + "step": 40811 + }, + { + "epoch": 0.8983491085186001, + "grad_norm": 2.039667573524233, + "learning_rate": 2.6853098417310654e-07, + "loss": 0.2951, + "step": 40812 + }, + { + "epoch": 0.8983711204050188, + "grad_norm": 1.4496513729117388, + "learning_rate": 2.6841575101149043e-07, + "loss": 0.2873, + "step": 40813 + }, + { + "epoch": 0.8983931322914374, + "grad_norm": 1.7471103056997253, + "learning_rate": 2.683005418978951e-07, + "loss": 0.2715, + "step": 40814 + }, + { + "epoch": 0.8984151441778561, + "grad_norm": 22.381920550598714, + "learning_rate": 2.6818535683290683e-07, + "loss": 0.3178, + "step": 40815 + }, + { + "epoch": 0.8984371560642747, + "grad_norm": 0.6217115749727586, + "learning_rate": 2.6807019581711115e-07, + "loss": 0.4614, + "step": 40816 + }, + { + "epoch": 0.8984591679506934, + "grad_norm": 0.6389866655131493, + "learning_rate": 2.679550588510943e-07, + "loss": 0.4687, + "step": 40817 + }, + { + "epoch": 0.8984811798371121, + "grad_norm": 1.6913844016745139, + "learning_rate": 2.6783994593543974e-07, + "loss": 0.3059, + "step": 40818 + }, + { + "epoch": 0.8985031917235307, + "grad_norm": 1.6123237371648815, + "learning_rate": 2.6772485707073303e-07, + "loss": 0.2563, + "step": 40819 + }, + { + "epoch": 0.8985252036099494, + "grad_norm": 1.9603761260619768, + "learning_rate": 2.676097922575588e-07, + "loss": 0.2814, + "step": 40820 + }, + { + "epoch": 0.8985472154963681, + "grad_norm": 1.724256491361038, + "learning_rate": 2.674947514965026e-07, + "loss": 0.3015, + "step": 40821 + }, + { + "epoch": 0.8985692273827867, + "grad_norm": 2.0422977775740265, + "learning_rate": 2.6737973478814906e-07, + "loss": 0.3553, + "step": 40822 + }, + { + "epoch": 0.8985912392692054, + "grad_norm": 3.967574334291288, + "learning_rate": 2.672647421330832e-07, + "loss": 0.3325, + "step": 40823 + }, + { + "epoch": 0.8986132511556241, + "grad_norm": 1.2975559381251807, + "learning_rate": 2.6714977353188786e-07, + "loss": 0.2547, + "step": 40824 + }, + { + "epoch": 0.8986352630420427, + "grad_norm": 1.5523357733545873, + "learning_rate": 2.6703482898514767e-07, + "loss": 0.1963, + "step": 40825 + }, + { + "epoch": 0.8986572749284614, + "grad_norm": 1.761536767526462, + "learning_rate": 2.669199084934487e-07, + "loss": 0.2355, + "step": 40826 + }, + { + "epoch": 0.8986792868148801, + "grad_norm": 0.6298478439314987, + "learning_rate": 2.668050120573734e-07, + "loss": 0.455, + "step": 40827 + }, + { + "epoch": 0.8987012987012987, + "grad_norm": 1.487269481579343, + "learning_rate": 2.6669013967750623e-07, + "loss": 0.301, + "step": 40828 + }, + { + "epoch": 0.8987233105877174, + "grad_norm": 1.8051272908106428, + "learning_rate": 2.6657529135443004e-07, + "loss": 0.3768, + "step": 40829 + }, + { + "epoch": 0.8987453224741361, + "grad_norm": 1.984392015213302, + "learning_rate": 2.6646046708873053e-07, + "loss": 0.2623, + "step": 40830 + }, + { + "epoch": 0.8987673343605547, + "grad_norm": 1.6925029412294295, + "learning_rate": 2.663456668809894e-07, + "loss": 0.3175, + "step": 40831 + }, + { + "epoch": 0.8987893462469734, + "grad_norm": 1.5044252362871773, + "learning_rate": 2.6623089073179064e-07, + "loss": 0.2596, + "step": 40832 + }, + { + "epoch": 0.8988113581333921, + "grad_norm": 1.3593529040386494, + "learning_rate": 2.661161386417188e-07, + "loss": 0.1756, + "step": 40833 + }, + { + "epoch": 0.8988333700198107, + "grad_norm": 1.3590065916626797, + "learning_rate": 2.660014106113551e-07, + "loss": 0.2982, + "step": 40834 + }, + { + "epoch": 0.8988553819062294, + "grad_norm": 2.5032204879917836, + "learning_rate": 2.658867066412846e-07, + "loss": 0.2622, + "step": 40835 + }, + { + "epoch": 0.8988773937926481, + "grad_norm": 2.248362553415272, + "learning_rate": 2.657720267320885e-07, + "loss": 0.291, + "step": 40836 + }, + { + "epoch": 0.8988994056790667, + "grad_norm": 1.845297735105416, + "learning_rate": 2.6565737088435073e-07, + "loss": 0.3153, + "step": 40837 + }, + { + "epoch": 0.8989214175654854, + "grad_norm": 1.3808803450882996, + "learning_rate": 2.6554273909865426e-07, + "loss": 0.2558, + "step": 40838 + }, + { + "epoch": 0.898943429451904, + "grad_norm": 1.8579907945115732, + "learning_rate": 2.6542813137558077e-07, + "loss": 0.3179, + "step": 40839 + }, + { + "epoch": 0.8989654413383227, + "grad_norm": 0.5895403897360361, + "learning_rate": 2.653135477157132e-07, + "loss": 0.4093, + "step": 40840 + }, + { + "epoch": 0.8989874532247414, + "grad_norm": 1.5568120380712043, + "learning_rate": 2.651989881196337e-07, + "loss": 0.3135, + "step": 40841 + }, + { + "epoch": 0.89900946511116, + "grad_norm": 0.6064846674047052, + "learning_rate": 2.650844525879248e-07, + "loss": 0.4325, + "step": 40842 + }, + { + "epoch": 0.8990314769975787, + "grad_norm": 1.6338693715450148, + "learning_rate": 2.6496994112116925e-07, + "loss": 0.2693, + "step": 40843 + }, + { + "epoch": 0.8990534888839974, + "grad_norm": 1.639264456039913, + "learning_rate": 2.6485545371994823e-07, + "loss": 0.2545, + "step": 40844 + }, + { + "epoch": 0.899075500770416, + "grad_norm": 1.7917551613148883, + "learning_rate": 2.6474099038484303e-07, + "loss": 0.3714, + "step": 40845 + }, + { + "epoch": 0.8990975126568347, + "grad_norm": 3.707591049308334, + "learning_rate": 2.646265511164359e-07, + "loss": 0.3946, + "step": 40846 + }, + { + "epoch": 0.8991195245432534, + "grad_norm": 2.0014909098540588, + "learning_rate": 2.645121359153091e-07, + "loss": 0.3014, + "step": 40847 + }, + { + "epoch": 0.899141536429672, + "grad_norm": 1.8626205116122703, + "learning_rate": 2.6439774478204395e-07, + "loss": 0.3299, + "step": 40848 + }, + { + "epoch": 0.8991635483160907, + "grad_norm": 3.249486205488526, + "learning_rate": 2.6428337771722213e-07, + "loss": 0.2281, + "step": 40849 + }, + { + "epoch": 0.8991855602025094, + "grad_norm": 1.4374314648793765, + "learning_rate": 2.641690347214232e-07, + "loss": 0.35, + "step": 40850 + }, + { + "epoch": 0.899207572088928, + "grad_norm": 1.969153457135656, + "learning_rate": 2.6405471579522946e-07, + "loss": 0.2011, + "step": 40851 + }, + { + "epoch": 0.8992295839753467, + "grad_norm": 1.8930729833191113, + "learning_rate": 2.639404209392221e-07, + "loss": 0.2675, + "step": 40852 + }, + { + "epoch": 0.8992515958617654, + "grad_norm": 1.6081782271341292, + "learning_rate": 2.638261501539829e-07, + "loss": 0.2678, + "step": 40853 + }, + { + "epoch": 0.899273607748184, + "grad_norm": 1.8585717459515605, + "learning_rate": 2.6371190344009147e-07, + "loss": 0.2851, + "step": 40854 + }, + { + "epoch": 0.8992956196346027, + "grad_norm": 1.9790940141856959, + "learning_rate": 2.635976807981272e-07, + "loss": 0.3055, + "step": 40855 + }, + { + "epoch": 0.8993176315210214, + "grad_norm": 2.066103336660937, + "learning_rate": 2.63483482228673e-07, + "loss": 0.2771, + "step": 40856 + }, + { + "epoch": 0.89933964340744, + "grad_norm": 2.2874253023399955, + "learning_rate": 2.633693077323074e-07, + "loss": 0.367, + "step": 40857 + }, + { + "epoch": 0.8993616552938587, + "grad_norm": 1.885548456676859, + "learning_rate": 2.6325515730961315e-07, + "loss": 0.2592, + "step": 40858 + }, + { + "epoch": 0.8993836671802774, + "grad_norm": 1.7488907672672624, + "learning_rate": 2.6314103096116814e-07, + "loss": 0.2888, + "step": 40859 + }, + { + "epoch": 0.899405679066696, + "grad_norm": 0.6430052858652677, + "learning_rate": 2.6302692868755255e-07, + "loss": 0.4478, + "step": 40860 + }, + { + "epoch": 0.8994276909531147, + "grad_norm": 0.6287631047189245, + "learning_rate": 2.629128504893469e-07, + "loss": 0.4513, + "step": 40861 + }, + { + "epoch": 0.8994497028395333, + "grad_norm": 2.6130123592021626, + "learning_rate": 2.627987963671319e-07, + "loss": 0.3267, + "step": 40862 + }, + { + "epoch": 0.899471714725952, + "grad_norm": 2.2693289041198015, + "learning_rate": 2.626847663214854e-07, + "loss": 0.3997, + "step": 40863 + }, + { + "epoch": 0.8994937266123707, + "grad_norm": 0.6124529662726886, + "learning_rate": 2.625707603529881e-07, + "loss": 0.4206, + "step": 40864 + }, + { + "epoch": 0.8995157384987893, + "grad_norm": 1.4574386482652364, + "learning_rate": 2.6245677846221894e-07, + "loss": 0.2523, + "step": 40865 + }, + { + "epoch": 0.899537750385208, + "grad_norm": 2.7130547113913894, + "learning_rate": 2.6234282064975804e-07, + "loss": 0.3074, + "step": 40866 + }, + { + "epoch": 0.8995597622716267, + "grad_norm": 1.3067275135743712, + "learning_rate": 2.622288869161832e-07, + "loss": 0.2276, + "step": 40867 + }, + { + "epoch": 0.8995817741580453, + "grad_norm": 1.68061672127685, + "learning_rate": 2.621149772620746e-07, + "loss": 0.3431, + "step": 40868 + }, + { + "epoch": 0.899603786044464, + "grad_norm": 2.459898416744085, + "learning_rate": 2.620010916880111e-07, + "loss": 0.2771, + "step": 40869 + }, + { + "epoch": 0.8996257979308827, + "grad_norm": 1.696839312216347, + "learning_rate": 2.6188723019457076e-07, + "loss": 0.2911, + "step": 40870 + }, + { + "epoch": 0.8996478098173013, + "grad_norm": 1.2080513697107698, + "learning_rate": 2.617733927823335e-07, + "loss": 0.2843, + "step": 40871 + }, + { + "epoch": 0.89966982170372, + "grad_norm": 1.4485482766509705, + "learning_rate": 2.616595794518761e-07, + "loss": 0.3131, + "step": 40872 + }, + { + "epoch": 0.8996918335901387, + "grad_norm": 1.8922252722273467, + "learning_rate": 2.615457902037788e-07, + "loss": 0.3188, + "step": 40873 + }, + { + "epoch": 0.8997138454765573, + "grad_norm": 0.6088137345918732, + "learning_rate": 2.614320250386193e-07, + "loss": 0.4436, + "step": 40874 + }, + { + "epoch": 0.899735857362976, + "grad_norm": 1.3021540158976164, + "learning_rate": 2.61318283956975e-07, + "loss": 0.2634, + "step": 40875 + }, + { + "epoch": 0.8997578692493947, + "grad_norm": 1.4895630711659607, + "learning_rate": 2.61204566959426e-07, + "loss": 0.3008, + "step": 40876 + }, + { + "epoch": 0.8997798811358133, + "grad_norm": 1.5180548478446678, + "learning_rate": 2.6109087404654786e-07, + "loss": 0.2448, + "step": 40877 + }, + { + "epoch": 0.899801893022232, + "grad_norm": 1.5504704406466188, + "learning_rate": 2.6097720521891913e-07, + "loss": 0.2943, + "step": 40878 + }, + { + "epoch": 0.8998239049086507, + "grad_norm": 1.495749132053476, + "learning_rate": 2.608635604771192e-07, + "loss": 0.2947, + "step": 40879 + }, + { + "epoch": 0.8998459167950693, + "grad_norm": 1.7173038115342485, + "learning_rate": 2.607499398217234e-07, + "loss": 0.3542, + "step": 40880 + }, + { + "epoch": 0.899867928681488, + "grad_norm": 0.6318653907067502, + "learning_rate": 2.60636343253311e-07, + "loss": 0.4628, + "step": 40881 + }, + { + "epoch": 0.8998899405679067, + "grad_norm": 1.6027207980228857, + "learning_rate": 2.605227707724578e-07, + "loss": 0.3395, + "step": 40882 + }, + { + "epoch": 0.8999119524543253, + "grad_norm": 1.5055608730912944, + "learning_rate": 2.604092223797416e-07, + "loss": 0.3175, + "step": 40883 + }, + { + "epoch": 0.899933964340744, + "grad_norm": 1.5715354798864043, + "learning_rate": 2.602956980757404e-07, + "loss": 0.2871, + "step": 40884 + }, + { + "epoch": 0.8999559762271626, + "grad_norm": 4.494533427385035, + "learning_rate": 2.601821978610303e-07, + "loss": 0.3256, + "step": 40885 + }, + { + "epoch": 0.8999779881135813, + "grad_norm": 0.6356261291974347, + "learning_rate": 2.6006872173618804e-07, + "loss": 0.4891, + "step": 40886 + }, + { + "epoch": 0.9, + "grad_norm": 2.1264105384616454, + "learning_rate": 2.599552697017899e-07, + "loss": 0.2528, + "step": 40887 + }, + { + "epoch": 0.9000220118864186, + "grad_norm": 1.493830705477604, + "learning_rate": 2.5984184175841364e-07, + "loss": 0.3427, + "step": 40888 + }, + { + "epoch": 0.9000440237728373, + "grad_norm": 1.6824292782061945, + "learning_rate": 2.5972843790663617e-07, + "loss": 0.2767, + "step": 40889 + }, + { + "epoch": 0.900066035659256, + "grad_norm": 2.193731597657221, + "learning_rate": 2.596150581470325e-07, + "loss": 0.3388, + "step": 40890 + }, + { + "epoch": 0.9000880475456746, + "grad_norm": 1.554362432097831, + "learning_rate": 2.5950170248017937e-07, + "loss": 0.2978, + "step": 40891 + }, + { + "epoch": 0.9001100594320933, + "grad_norm": 29.209887068383463, + "learning_rate": 2.593883709066525e-07, + "loss": 0.301, + "step": 40892 + }, + { + "epoch": 0.900132071318512, + "grad_norm": 1.266281521864803, + "learning_rate": 2.59275063427028e-07, + "loss": 0.2207, + "step": 40893 + }, + { + "epoch": 0.9001540832049306, + "grad_norm": 1.564939470708329, + "learning_rate": 2.591617800418833e-07, + "loss": 0.2779, + "step": 40894 + }, + { + "epoch": 0.9001760950913493, + "grad_norm": 2.140727809358449, + "learning_rate": 2.590485207517923e-07, + "loss": 0.2404, + "step": 40895 + }, + { + "epoch": 0.900198106977768, + "grad_norm": 1.6789868415506564, + "learning_rate": 2.589352855573313e-07, + "loss": 0.2375, + "step": 40896 + }, + { + "epoch": 0.9002201188641866, + "grad_norm": 2.1113860259066826, + "learning_rate": 2.588220744590753e-07, + "loss": 0.2968, + "step": 40897 + }, + { + "epoch": 0.9002421307506053, + "grad_norm": 0.6292013602933375, + "learning_rate": 2.587088874575999e-07, + "loss": 0.437, + "step": 40898 + }, + { + "epoch": 0.900264142637024, + "grad_norm": 1.6079924491030557, + "learning_rate": 2.5859572455348203e-07, + "loss": 0.238, + "step": 40899 + }, + { + "epoch": 0.9002861545234426, + "grad_norm": 1.8455478663208014, + "learning_rate": 2.58482585747295e-07, + "loss": 0.274, + "step": 40900 + }, + { + "epoch": 0.9003081664098613, + "grad_norm": 4.117229444800993, + "learning_rate": 2.583694710396134e-07, + "loss": 0.2927, + "step": 40901 + }, + { + "epoch": 0.90033017829628, + "grad_norm": 42.68721752702711, + "learning_rate": 2.5825638043101397e-07, + "loss": 0.3027, + "step": 40902 + }, + { + "epoch": 0.9003521901826986, + "grad_norm": 1.6985335161583184, + "learning_rate": 2.581433139220696e-07, + "loss": 0.3018, + "step": 40903 + }, + { + "epoch": 0.9003742020691173, + "grad_norm": 0.615054750235532, + "learning_rate": 2.5803027151335593e-07, + "loss": 0.4267, + "step": 40904 + }, + { + "epoch": 0.900396213955536, + "grad_norm": 2.343326835000882, + "learning_rate": 2.57917253205448e-07, + "loss": 0.3443, + "step": 40905 + }, + { + "epoch": 0.9004182258419546, + "grad_norm": 2.0887628455471186, + "learning_rate": 2.578042589989188e-07, + "loss": 0.2769, + "step": 40906 + }, + { + "epoch": 0.9004402377283733, + "grad_norm": 1.563541723979417, + "learning_rate": 2.576912888943445e-07, + "loss": 0.2049, + "step": 40907 + }, + { + "epoch": 0.900462249614792, + "grad_norm": 1.973154159189482, + "learning_rate": 2.575783428922968e-07, + "loss": 0.358, + "step": 40908 + }, + { + "epoch": 0.9004842615012106, + "grad_norm": 1.3842817235504805, + "learning_rate": 2.574654209933519e-07, + "loss": 0.2595, + "step": 40909 + }, + { + "epoch": 0.9005062733876293, + "grad_norm": 1.4942331959083188, + "learning_rate": 2.5735252319808337e-07, + "loss": 0.2377, + "step": 40910 + }, + { + "epoch": 0.9005282852740479, + "grad_norm": 2.5957818045887904, + "learning_rate": 2.5723964950706393e-07, + "loss": 0.2435, + "step": 40911 + }, + { + "epoch": 0.9005502971604666, + "grad_norm": 1.5001828692923391, + "learning_rate": 2.5712679992086877e-07, + "loss": 0.3208, + "step": 40912 + }, + { + "epoch": 0.9005723090468853, + "grad_norm": 2.2196669756921, + "learning_rate": 2.570139744400696e-07, + "loss": 0.3044, + "step": 40913 + }, + { + "epoch": 0.9005943209333039, + "grad_norm": 1.6111067535488657, + "learning_rate": 2.5690117306524096e-07, + "loss": 0.2398, + "step": 40914 + }, + { + "epoch": 0.9006163328197226, + "grad_norm": 2.2598883941104124, + "learning_rate": 2.567883957969569e-07, + "loss": 0.2974, + "step": 40915 + }, + { + "epoch": 0.9006383447061413, + "grad_norm": 1.5357543520380177, + "learning_rate": 2.566756426357886e-07, + "loss": 0.2765, + "step": 40916 + }, + { + "epoch": 0.9006603565925599, + "grad_norm": 1.9067627836947092, + "learning_rate": 2.565629135823111e-07, + "loss": 0.2503, + "step": 40917 + }, + { + "epoch": 0.9006823684789786, + "grad_norm": 1.957909490936079, + "learning_rate": 2.5645020863709626e-07, + "loss": 0.2007, + "step": 40918 + }, + { + "epoch": 0.9007043803653974, + "grad_norm": 1.915697112052992, + "learning_rate": 2.563375278007169e-07, + "loss": 0.2596, + "step": 40919 + }, + { + "epoch": 0.9007263922518159, + "grad_norm": 3.3295931399541367, + "learning_rate": 2.562248710737464e-07, + "loss": 0.3037, + "step": 40920 + }, + { + "epoch": 0.9007484041382346, + "grad_norm": 0.6158433941248072, + "learning_rate": 2.561122384567566e-07, + "loss": 0.4481, + "step": 40921 + }, + { + "epoch": 0.9007704160246534, + "grad_norm": 0.6160346823784546, + "learning_rate": 2.559996299503209e-07, + "loss": 0.4506, + "step": 40922 + }, + { + "epoch": 0.900792427911072, + "grad_norm": 1.546693702381742, + "learning_rate": 2.5588704555501e-07, + "loss": 0.203, + "step": 40923 + }, + { + "epoch": 0.9008144397974907, + "grad_norm": 2.150517273176043, + "learning_rate": 2.557744852713973e-07, + "loss": 0.2413, + "step": 40924 + }, + { + "epoch": 0.9008364516839094, + "grad_norm": 0.5934972104181573, + "learning_rate": 2.5566194910005506e-07, + "loss": 0.441, + "step": 40925 + }, + { + "epoch": 0.900858463570328, + "grad_norm": 1.5223336900573863, + "learning_rate": 2.555494370415551e-07, + "loss": 0.1992, + "step": 40926 + }, + { + "epoch": 0.9008804754567467, + "grad_norm": 1.4444120152887354, + "learning_rate": 2.554369490964681e-07, + "loss": 0.2531, + "step": 40927 + }, + { + "epoch": 0.9009024873431654, + "grad_norm": 1.762847705553362, + "learning_rate": 2.5532448526536637e-07, + "loss": 0.2061, + "step": 40928 + }, + { + "epoch": 0.900924499229584, + "grad_norm": 1.7623465639290015, + "learning_rate": 2.5521204554882216e-07, + "loss": 0.3118, + "step": 40929 + }, + { + "epoch": 0.9009465111160027, + "grad_norm": 1.6556595803041054, + "learning_rate": 2.5509962994740734e-07, + "loss": 0.292, + "step": 40930 + }, + { + "epoch": 0.9009685230024214, + "grad_norm": 2.356071633083104, + "learning_rate": 2.549872384616925e-07, + "loss": 0.2356, + "step": 40931 + }, + { + "epoch": 0.90099053488884, + "grad_norm": 1.808556950137543, + "learning_rate": 2.548748710922483e-07, + "loss": 0.3052, + "step": 40932 + }, + { + "epoch": 0.9010125467752587, + "grad_norm": 1.5110602901187333, + "learning_rate": 2.5476252783964597e-07, + "loss": 0.3209, + "step": 40933 + }, + { + "epoch": 0.9010345586616773, + "grad_norm": 0.608671703099787, + "learning_rate": 2.546502087044572e-07, + "loss": 0.4402, + "step": 40934 + }, + { + "epoch": 0.901056570548096, + "grad_norm": 1.870347088965036, + "learning_rate": 2.545379136872528e-07, + "loss": 0.2493, + "step": 40935 + }, + { + "epoch": 0.9010785824345147, + "grad_norm": 2.117115895957412, + "learning_rate": 2.544256427886038e-07, + "loss": 0.3682, + "step": 40936 + }, + { + "epoch": 0.9011005943209333, + "grad_norm": 2.170667287433685, + "learning_rate": 2.5431339600907934e-07, + "loss": 0.3478, + "step": 40937 + }, + { + "epoch": 0.901122606207352, + "grad_norm": 3.913566868843597, + "learning_rate": 2.5420117334925165e-07, + "loss": 0.3106, + "step": 40938 + }, + { + "epoch": 0.9011446180937707, + "grad_norm": 1.9337955372899382, + "learning_rate": 2.5408897480968865e-07, + "loss": 0.2365, + "step": 40939 + }, + { + "epoch": 0.9011666299801893, + "grad_norm": 0.6163184079997002, + "learning_rate": 2.539768003909637e-07, + "loss": 0.4507, + "step": 40940 + }, + { + "epoch": 0.901188641866608, + "grad_norm": 1.4402616376010746, + "learning_rate": 2.5386465009364537e-07, + "loss": 0.2176, + "step": 40941 + }, + { + "epoch": 0.9012106537530267, + "grad_norm": 1.8283042344983882, + "learning_rate": 2.537525239183031e-07, + "loss": 0.2748, + "step": 40942 + }, + { + "epoch": 0.9012326656394453, + "grad_norm": 1.846959454429494, + "learning_rate": 2.536404218655081e-07, + "loss": 0.3542, + "step": 40943 + }, + { + "epoch": 0.901254677525864, + "grad_norm": 1.854126646613834, + "learning_rate": 2.5352834393582883e-07, + "loss": 0.3257, + "step": 40944 + }, + { + "epoch": 0.9012766894122827, + "grad_norm": 0.6230268245889508, + "learning_rate": 2.5341629012983537e-07, + "loss": 0.4337, + "step": 40945 + }, + { + "epoch": 0.9012987012987013, + "grad_norm": 1.5941074349781719, + "learning_rate": 2.533042604480984e-07, + "loss": 0.3195, + "step": 40946 + }, + { + "epoch": 0.90132071318512, + "grad_norm": 1.7089587471752263, + "learning_rate": 2.5319225489118526e-07, + "loss": 0.2728, + "step": 40947 + }, + { + "epoch": 0.9013427250715387, + "grad_norm": 2.3420012405849358, + "learning_rate": 2.530802734596671e-07, + "loss": 0.2277, + "step": 40948 + }, + { + "epoch": 0.9013647369579573, + "grad_norm": 1.717721465185469, + "learning_rate": 2.5296831615411135e-07, + "loss": 0.2954, + "step": 40949 + }, + { + "epoch": 0.901386748844376, + "grad_norm": 1.61340189795771, + "learning_rate": 2.528563829750874e-07, + "loss": 0.3318, + "step": 40950 + }, + { + "epoch": 0.9014087607307947, + "grad_norm": 3.0966740292524135, + "learning_rate": 2.52744473923166e-07, + "loss": 0.403, + "step": 40951 + }, + { + "epoch": 0.9014307726172133, + "grad_norm": 0.6129799637220608, + "learning_rate": 2.526325889989134e-07, + "loss": 0.4396, + "step": 40952 + }, + { + "epoch": 0.901452784503632, + "grad_norm": 1.5624403941621006, + "learning_rate": 2.525207282029002e-07, + "loss": 0.2983, + "step": 40953 + }, + { + "epoch": 0.9014747963900507, + "grad_norm": 0.6142015853025053, + "learning_rate": 2.5240889153569315e-07, + "loss": 0.4467, + "step": 40954 + }, + { + "epoch": 0.9014968082764693, + "grad_norm": 1.9260027099221806, + "learning_rate": 2.522970789978618e-07, + "loss": 0.3176, + "step": 40955 + }, + { + "epoch": 0.901518820162888, + "grad_norm": 2.9311942458388125, + "learning_rate": 2.5218529058997523e-07, + "loss": 0.2689, + "step": 40956 + }, + { + "epoch": 0.9015408320493066, + "grad_norm": 1.9165635011565623, + "learning_rate": 2.520735263125995e-07, + "loss": 0.2616, + "step": 40957 + }, + { + "epoch": 0.9015628439357253, + "grad_norm": 0.6229665511443203, + "learning_rate": 2.5196178616630487e-07, + "loss": 0.4401, + "step": 40958 + }, + { + "epoch": 0.901584855822144, + "grad_norm": 1.587818282863848, + "learning_rate": 2.5185007015165687e-07, + "loss": 0.2892, + "step": 40959 + }, + { + "epoch": 0.9016068677085626, + "grad_norm": 0.6072301917434604, + "learning_rate": 2.5173837826922455e-07, + "loss": 0.4531, + "step": 40960 + }, + { + "epoch": 0.9016288795949813, + "grad_norm": 0.6119923578074857, + "learning_rate": 2.5162671051957685e-07, + "loss": 0.4492, + "step": 40961 + }, + { + "epoch": 0.9016508914814, + "grad_norm": 0.5938437412561096, + "learning_rate": 2.5151506690327955e-07, + "loss": 0.3935, + "step": 40962 + }, + { + "epoch": 0.9016729033678186, + "grad_norm": 1.750586342209251, + "learning_rate": 2.514034474209004e-07, + "loss": 0.2904, + "step": 40963 + }, + { + "epoch": 0.9016949152542373, + "grad_norm": 1.8347903568734767, + "learning_rate": 2.512918520730062e-07, + "loss": 0.1668, + "step": 40964 + }, + { + "epoch": 0.901716927140656, + "grad_norm": 1.9637195234261904, + "learning_rate": 2.511802808601649e-07, + "loss": 0.3037, + "step": 40965 + }, + { + "epoch": 0.9017389390270746, + "grad_norm": 1.506964503767747, + "learning_rate": 2.5106873378294484e-07, + "loss": 0.2985, + "step": 40966 + }, + { + "epoch": 0.9017609509134933, + "grad_norm": 2.743758137590537, + "learning_rate": 2.509572108419106e-07, + "loss": 0.3377, + "step": 40967 + }, + { + "epoch": 0.901782962799912, + "grad_norm": 1.8169179295319215, + "learning_rate": 2.508457120376301e-07, + "loss": 0.3043, + "step": 40968 + }, + { + "epoch": 0.9018049746863306, + "grad_norm": 1.9627600965000307, + "learning_rate": 2.507342373706695e-07, + "loss": 0.3726, + "step": 40969 + }, + { + "epoch": 0.9018269865727493, + "grad_norm": 1.9560303150844218, + "learning_rate": 2.506227868415956e-07, + "loss": 0.2899, + "step": 40970 + }, + { + "epoch": 0.901848998459168, + "grad_norm": 3.1389558958339596, + "learning_rate": 2.505113604509757e-07, + "loss": 0.3071, + "step": 40971 + }, + { + "epoch": 0.9018710103455866, + "grad_norm": 2.086250371496357, + "learning_rate": 2.5039995819937543e-07, + "loss": 0.3416, + "step": 40972 + }, + { + "epoch": 0.9018930222320053, + "grad_norm": 1.4721168191852103, + "learning_rate": 2.502885800873606e-07, + "loss": 0.2584, + "step": 40973 + }, + { + "epoch": 0.901915034118424, + "grad_norm": 1.3075025704672127, + "learning_rate": 2.501772261154978e-07, + "loss": 0.2938, + "step": 40974 + }, + { + "epoch": 0.9019370460048426, + "grad_norm": 0.6294039066582323, + "learning_rate": 2.500658962843516e-07, + "loss": 0.442, + "step": 40975 + }, + { + "epoch": 0.9019590578912613, + "grad_norm": 3.656599655096197, + "learning_rate": 2.4995459059449056e-07, + "loss": 0.2948, + "step": 40976 + }, + { + "epoch": 0.90198106977768, + "grad_norm": 6.695834533760078, + "learning_rate": 2.4984330904647856e-07, + "loss": 0.2168, + "step": 40977 + }, + { + "epoch": 0.9020030816640986, + "grad_norm": 1.1880883066561967, + "learning_rate": 2.497320516408813e-07, + "loss": 0.255, + "step": 40978 + }, + { + "epoch": 0.9020250935505173, + "grad_norm": 1.5403662842510046, + "learning_rate": 2.49620818378265e-07, + "loss": 0.2298, + "step": 40979 + }, + { + "epoch": 0.9020471054369359, + "grad_norm": 1.7129032170147107, + "learning_rate": 2.4950960925919364e-07, + "loss": 0.3143, + "step": 40980 + }, + { + "epoch": 0.9020691173233546, + "grad_norm": 2.340453854107273, + "learning_rate": 2.4939842428423286e-07, + "loss": 0.2132, + "step": 40981 + }, + { + "epoch": 0.9020911292097733, + "grad_norm": 2.9095784914924634, + "learning_rate": 2.4928726345394947e-07, + "loss": 0.3372, + "step": 40982 + }, + { + "epoch": 0.9021131410961919, + "grad_norm": 0.6151820337197891, + "learning_rate": 2.491761267689058e-07, + "loss": 0.4363, + "step": 40983 + }, + { + "epoch": 0.9021351529826106, + "grad_norm": 2.0994022781365813, + "learning_rate": 2.4906501422966913e-07, + "loss": 0.3263, + "step": 40984 + }, + { + "epoch": 0.9021571648690293, + "grad_norm": 1.5677055752793008, + "learning_rate": 2.4895392583680176e-07, + "loss": 0.2058, + "step": 40985 + }, + { + "epoch": 0.9021791767554479, + "grad_norm": 2.291920477261914, + "learning_rate": 2.4884286159087e-07, + "loss": 0.3347, + "step": 40986 + }, + { + "epoch": 0.9022011886418666, + "grad_norm": 2.2199723345036046, + "learning_rate": 2.4873182149243893e-07, + "loss": 0.3391, + "step": 40987 + }, + { + "epoch": 0.9022232005282853, + "grad_norm": 1.786197994425107, + "learning_rate": 2.4862080554207077e-07, + "loss": 0.3216, + "step": 40988 + }, + { + "epoch": 0.9022452124147039, + "grad_norm": 1.820211823395045, + "learning_rate": 2.4850981374033136e-07, + "loss": 0.2713, + "step": 40989 + }, + { + "epoch": 0.9022672243011226, + "grad_norm": 0.6271753842016213, + "learning_rate": 2.48398846087784e-07, + "loss": 0.4621, + "step": 40990 + }, + { + "epoch": 0.9022892361875413, + "grad_norm": 1.6021063066707288, + "learning_rate": 2.482879025849927e-07, + "loss": 0.2891, + "step": 40991 + }, + { + "epoch": 0.9023112480739599, + "grad_norm": 2.029431155589512, + "learning_rate": 2.481769832325226e-07, + "loss": 0.2811, + "step": 40992 + }, + { + "epoch": 0.9023332599603786, + "grad_norm": 7.41228924171831, + "learning_rate": 2.4806608803093546e-07, + "loss": 0.1887, + "step": 40993 + }, + { + "epoch": 0.9023552718467973, + "grad_norm": 1.7925247857727564, + "learning_rate": 2.479552169807964e-07, + "loss": 0.2733, + "step": 40994 + }, + { + "epoch": 0.9023772837332159, + "grad_norm": 0.6128745591320255, + "learning_rate": 2.4784437008266824e-07, + "loss": 0.4549, + "step": 40995 + }, + { + "epoch": 0.9023992956196346, + "grad_norm": 4.751056538267333, + "learning_rate": 2.47733547337114e-07, + "loss": 0.2388, + "step": 40996 + }, + { + "epoch": 0.9024213075060533, + "grad_norm": 1.2684699152328525, + "learning_rate": 2.476227487446986e-07, + "loss": 0.2485, + "step": 40997 + }, + { + "epoch": 0.9024433193924719, + "grad_norm": 1.4657159400981252, + "learning_rate": 2.4751197430598285e-07, + "loss": 0.263, + "step": 40998 + }, + { + "epoch": 0.9024653312788906, + "grad_norm": 2.9307599052501496, + "learning_rate": 2.474012240215318e-07, + "loss": 0.2884, + "step": 40999 + }, + { + "epoch": 0.9024873431653093, + "grad_norm": 1.7329491549894644, + "learning_rate": 2.472904978919066e-07, + "loss": 0.4475, + "step": 41000 + }, + { + "epoch": 0.9025093550517279, + "grad_norm": 1.788076871390806, + "learning_rate": 2.471797959176714e-07, + "loss": 0.263, + "step": 41001 + }, + { + "epoch": 0.9025313669381466, + "grad_norm": 1.6210454142029038, + "learning_rate": 2.470691180993884e-07, + "loss": 0.2573, + "step": 41002 + }, + { + "epoch": 0.9025533788245652, + "grad_norm": 1.5100364678070708, + "learning_rate": 2.469584644376205e-07, + "loss": 0.3042, + "step": 41003 + }, + { + "epoch": 0.9025753907109839, + "grad_norm": 3.6642972534042966, + "learning_rate": 2.4684783493292844e-07, + "loss": 0.3476, + "step": 41004 + }, + { + "epoch": 0.9025974025974026, + "grad_norm": 2.240127593321889, + "learning_rate": 2.4673722958587607e-07, + "loss": 0.3033, + "step": 41005 + }, + { + "epoch": 0.9026194144838212, + "grad_norm": 1.9598927183612256, + "learning_rate": 2.4662664839702477e-07, + "loss": 0.3158, + "step": 41006 + }, + { + "epoch": 0.9026414263702399, + "grad_norm": 1.5019919006984348, + "learning_rate": 2.4651609136693787e-07, + "loss": 0.2987, + "step": 41007 + }, + { + "epoch": 0.9026634382566586, + "grad_norm": 2.658759052416077, + "learning_rate": 2.464055584961761e-07, + "loss": 0.2436, + "step": 41008 + }, + { + "epoch": 0.9026854501430772, + "grad_norm": 1.4582912361788438, + "learning_rate": 2.462950497853006e-07, + "loss": 0.3085, + "step": 41009 + }, + { + "epoch": 0.9027074620294959, + "grad_norm": 2.0163758490402266, + "learning_rate": 2.4618456523487487e-07, + "loss": 0.3363, + "step": 41010 + }, + { + "epoch": 0.9027294739159146, + "grad_norm": 1.7391393289405281, + "learning_rate": 2.460741048454579e-07, + "loss": 0.2592, + "step": 41011 + }, + { + "epoch": 0.9027514858023332, + "grad_norm": 1.67208083794873, + "learning_rate": 2.459636686176142e-07, + "loss": 0.2352, + "step": 41012 + }, + { + "epoch": 0.9027734976887519, + "grad_norm": 1.711441355267099, + "learning_rate": 2.458532565519034e-07, + "loss": 0.2532, + "step": 41013 + }, + { + "epoch": 0.9027955095751706, + "grad_norm": 0.6139986040073914, + "learning_rate": 2.4574286864888606e-07, + "loss": 0.4614, + "step": 41014 + }, + { + "epoch": 0.9028175214615892, + "grad_norm": 2.508804225943969, + "learning_rate": 2.456325049091246e-07, + "loss": 0.3238, + "step": 41015 + }, + { + "epoch": 0.9028395333480079, + "grad_norm": 1.9261755786763999, + "learning_rate": 2.455221653331785e-07, + "loss": 0.332, + "step": 41016 + }, + { + "epoch": 0.9028615452344266, + "grad_norm": 1.930419181722319, + "learning_rate": 2.454118499216096e-07, + "loss": 0.3084, + "step": 41017 + }, + { + "epoch": 0.9028835571208452, + "grad_norm": 0.6056055645275615, + "learning_rate": 2.453015586749785e-07, + "loss": 0.4475, + "step": 41018 + }, + { + "epoch": 0.9029055690072639, + "grad_norm": 1.5131368699276397, + "learning_rate": 2.4519129159384533e-07, + "loss": 0.2474, + "step": 41019 + }, + { + "epoch": 0.9029275808936826, + "grad_norm": 1.396342817224465, + "learning_rate": 2.450810486787708e-07, + "loss": 0.2526, + "step": 41020 + }, + { + "epoch": 0.9029495927801012, + "grad_norm": 1.7198228021486104, + "learning_rate": 2.4497082993031497e-07, + "loss": 0.3055, + "step": 41021 + }, + { + "epoch": 0.9029716046665199, + "grad_norm": 1.7684535377290553, + "learning_rate": 2.4486063534903794e-07, + "loss": 0.2889, + "step": 41022 + }, + { + "epoch": 0.9029936165529386, + "grad_norm": 1.8112382710301838, + "learning_rate": 2.4475046493550094e-07, + "loss": 0.3378, + "step": 41023 + }, + { + "epoch": 0.9030156284393572, + "grad_norm": 1.7985098592632365, + "learning_rate": 2.446403186902624e-07, + "loss": 0.2562, + "step": 41024 + }, + { + "epoch": 0.9030376403257759, + "grad_norm": 2.3017702053607967, + "learning_rate": 2.44530196613883e-07, + "loss": 0.3384, + "step": 41025 + }, + { + "epoch": 0.9030596522121945, + "grad_norm": 1.597821333147426, + "learning_rate": 2.444200987069212e-07, + "loss": 0.2532, + "step": 41026 + }, + { + "epoch": 0.9030816640986132, + "grad_norm": 0.6220739410627649, + "learning_rate": 2.4431002496993815e-07, + "loss": 0.4172, + "step": 41027 + }, + { + "epoch": 0.9031036759850319, + "grad_norm": 1.9943464353069296, + "learning_rate": 2.441999754034929e-07, + "loss": 0.2865, + "step": 41028 + }, + { + "epoch": 0.9031256878714505, + "grad_norm": 0.6164899523291406, + "learning_rate": 2.4408995000814395e-07, + "loss": 0.4593, + "step": 41029 + }, + { + "epoch": 0.9031476997578692, + "grad_norm": 4.12514891219223, + "learning_rate": 2.4397994878445184e-07, + "loss": 0.3569, + "step": 41030 + }, + { + "epoch": 0.9031697116442879, + "grad_norm": 3.245771498451629, + "learning_rate": 2.4386997173297453e-07, + "loss": 0.2751, + "step": 41031 + }, + { + "epoch": 0.9031917235307065, + "grad_norm": 1.458791069850569, + "learning_rate": 2.43760018854271e-07, + "loss": 0.2991, + "step": 41032 + }, + { + "epoch": 0.9032137354171252, + "grad_norm": 1.633033292078807, + "learning_rate": 2.436500901489008e-07, + "loss": 0.3649, + "step": 41033 + }, + { + "epoch": 0.903235747303544, + "grad_norm": 1.833066746990601, + "learning_rate": 2.4354018561742185e-07, + "loss": 0.2826, + "step": 41034 + }, + { + "epoch": 0.9032577591899625, + "grad_norm": 2.9471876032353435, + "learning_rate": 2.434303052603937e-07, + "loss": 0.3259, + "step": 41035 + }, + { + "epoch": 0.9032797710763812, + "grad_norm": 0.6509860377554748, + "learning_rate": 2.433204490783736e-07, + "loss": 0.4608, + "step": 41036 + }, + { + "epoch": 0.9033017829628, + "grad_norm": 2.1757182017166627, + "learning_rate": 2.432106170719206e-07, + "loss": 0.2864, + "step": 41037 + }, + { + "epoch": 0.9033237948492185, + "grad_norm": 0.6179439285057376, + "learning_rate": 2.431008092415932e-07, + "loss": 0.4517, + "step": 41038 + }, + { + "epoch": 0.9033458067356372, + "grad_norm": 1.2731119839921725, + "learning_rate": 2.4299102558794864e-07, + "loss": 0.2082, + "step": 41039 + }, + { + "epoch": 0.903367818622056, + "grad_norm": 1.5893354850179833, + "learning_rate": 2.428812661115465e-07, + "loss": 0.2178, + "step": 41040 + }, + { + "epoch": 0.9033898305084745, + "grad_norm": 2.292840988224578, + "learning_rate": 2.4277153081294303e-07, + "loss": 0.2686, + "step": 41041 + }, + { + "epoch": 0.9034118423948932, + "grad_norm": 1.5294301776911392, + "learning_rate": 2.42661819692695e-07, + "loss": 0.2648, + "step": 41042 + }, + { + "epoch": 0.903433854281312, + "grad_norm": 1.7161751657867768, + "learning_rate": 2.4255213275136303e-07, + "loss": 0.3202, + "step": 41043 + }, + { + "epoch": 0.9034558661677305, + "grad_norm": 1.1927437623308272, + "learning_rate": 2.4244246998950283e-07, + "loss": 0.2486, + "step": 41044 + }, + { + "epoch": 0.9034778780541493, + "grad_norm": 2.0187469371987374, + "learning_rate": 2.4233283140767116e-07, + "loss": 0.3292, + "step": 41045 + }, + { + "epoch": 0.903499889940568, + "grad_norm": 4.128422300518166, + "learning_rate": 2.4222321700642705e-07, + "loss": 0.3499, + "step": 41046 + }, + { + "epoch": 0.9035219018269866, + "grad_norm": 0.6192088182630583, + "learning_rate": 2.42113626786325e-07, + "loss": 0.4268, + "step": 41047 + }, + { + "epoch": 0.9035439137134053, + "grad_norm": 2.1912984570853684, + "learning_rate": 2.420040607479246e-07, + "loss": 0.2918, + "step": 41048 + }, + { + "epoch": 0.903565925599824, + "grad_norm": 0.6106916817658032, + "learning_rate": 2.4189451889178205e-07, + "loss": 0.4398, + "step": 41049 + }, + { + "epoch": 0.9035879374862426, + "grad_norm": 0.6089279328690277, + "learning_rate": 2.4178500121845305e-07, + "loss": 0.4521, + "step": 41050 + }, + { + "epoch": 0.9036099493726613, + "grad_norm": 0.6018413796753651, + "learning_rate": 2.4167550772849544e-07, + "loss": 0.4577, + "step": 41051 + }, + { + "epoch": 0.9036319612590799, + "grad_norm": 0.614913149392281, + "learning_rate": 2.41566038422465e-07, + "loss": 0.4471, + "step": 41052 + }, + { + "epoch": 0.9036539731454986, + "grad_norm": 2.013712014403499, + "learning_rate": 2.4145659330091774e-07, + "loss": 0.2286, + "step": 41053 + }, + { + "epoch": 0.9036759850319173, + "grad_norm": 1.74233237403227, + "learning_rate": 2.413471723644112e-07, + "loss": 0.3042, + "step": 41054 + }, + { + "epoch": 0.9036979969183359, + "grad_norm": 1.571395669809669, + "learning_rate": 2.412377756135004e-07, + "loss": 0.2976, + "step": 41055 + }, + { + "epoch": 0.9037200088047546, + "grad_norm": 1.4562457283577424, + "learning_rate": 2.4112840304874205e-07, + "loss": 0.2619, + "step": 41056 + }, + { + "epoch": 0.9037420206911733, + "grad_norm": 2.3557473932202786, + "learning_rate": 2.410190546706909e-07, + "loss": 0.3497, + "step": 41057 + }, + { + "epoch": 0.9037640325775919, + "grad_norm": 2.346719336242539, + "learning_rate": 2.409097304799035e-07, + "loss": 0.2531, + "step": 41058 + }, + { + "epoch": 0.9037860444640106, + "grad_norm": 1.9168719843111837, + "learning_rate": 2.4080043047693626e-07, + "loss": 0.4152, + "step": 41059 + }, + { + "epoch": 0.9038080563504293, + "grad_norm": 3.5522270785275514, + "learning_rate": 2.4069115466234305e-07, + "loss": 0.2914, + "step": 41060 + }, + { + "epoch": 0.9038300682368479, + "grad_norm": 1.3001876551998583, + "learning_rate": 2.4058190303668126e-07, + "loss": 0.2755, + "step": 41061 + }, + { + "epoch": 0.9038520801232666, + "grad_norm": 1.8796455635775788, + "learning_rate": 2.404726756005038e-07, + "loss": 0.278, + "step": 41062 + }, + { + "epoch": 0.9038740920096853, + "grad_norm": 0.6266127879724124, + "learning_rate": 2.403634723543674e-07, + "loss": 0.4643, + "step": 41063 + }, + { + "epoch": 0.9038961038961039, + "grad_norm": 1.728436275656756, + "learning_rate": 2.402542932988272e-07, + "loss": 0.3662, + "step": 41064 + }, + { + "epoch": 0.9039181157825226, + "grad_norm": 1.2988880451363278, + "learning_rate": 2.4014513843443667e-07, + "loss": 0.245, + "step": 41065 + }, + { + "epoch": 0.9039401276689413, + "grad_norm": 1.713001097141579, + "learning_rate": 2.400360077617525e-07, + "loss": 0.2679, + "step": 41066 + }, + { + "epoch": 0.9039621395553599, + "grad_norm": 1.5004660715478202, + "learning_rate": 2.399269012813277e-07, + "loss": 0.231, + "step": 41067 + }, + { + "epoch": 0.9039841514417786, + "grad_norm": 2.637972327296351, + "learning_rate": 2.3981781899371725e-07, + "loss": 0.2568, + "step": 41068 + }, + { + "epoch": 0.9040061633281973, + "grad_norm": 1.465532743450969, + "learning_rate": 2.397087608994764e-07, + "loss": 0.272, + "step": 41069 + }, + { + "epoch": 0.9040281752146159, + "grad_norm": 0.6047971556297272, + "learning_rate": 2.3959972699915844e-07, + "loss": 0.4402, + "step": 41070 + }, + { + "epoch": 0.9040501871010346, + "grad_norm": 0.6104643858831134, + "learning_rate": 2.394907172933181e-07, + "loss": 0.4356, + "step": 41071 + }, + { + "epoch": 0.9040721989874533, + "grad_norm": 0.6038714170981381, + "learning_rate": 2.3938173178250867e-07, + "loss": 0.4253, + "step": 41072 + }, + { + "epoch": 0.9040942108738719, + "grad_norm": 0.5959114785736426, + "learning_rate": 2.392727704672848e-07, + "loss": 0.4318, + "step": 41073 + }, + { + "epoch": 0.9041162227602906, + "grad_norm": 0.620581705239046, + "learning_rate": 2.391638333482005e-07, + "loss": 0.4642, + "step": 41074 + }, + { + "epoch": 0.9041382346467092, + "grad_norm": 2.1427404207939773, + "learning_rate": 2.3905492042580804e-07, + "loss": 0.2238, + "step": 41075 + }, + { + "epoch": 0.9041602465331279, + "grad_norm": 2.8062482187463025, + "learning_rate": 2.389460317006631e-07, + "loss": 0.2936, + "step": 41076 + }, + { + "epoch": 0.9041822584195466, + "grad_norm": 1.4936105443950802, + "learning_rate": 2.3883716717331753e-07, + "loss": 0.2589, + "step": 41077 + }, + { + "epoch": 0.9042042703059652, + "grad_norm": 2.0114402202718007, + "learning_rate": 2.387283268443241e-07, + "loss": 0.2262, + "step": 41078 + }, + { + "epoch": 0.9042262821923839, + "grad_norm": 0.6267343539425774, + "learning_rate": 2.3861951071423806e-07, + "loss": 0.4699, + "step": 41079 + }, + { + "epoch": 0.9042482940788026, + "grad_norm": 1.7612775681479023, + "learning_rate": 2.3851071878361055e-07, + "loss": 0.222, + "step": 41080 + }, + { + "epoch": 0.9042703059652212, + "grad_norm": 0.6215907129762516, + "learning_rate": 2.3840195105299557e-07, + "loss": 0.4392, + "step": 41081 + }, + { + "epoch": 0.9042923178516399, + "grad_norm": 1.6722786506290028, + "learning_rate": 2.3829320752294606e-07, + "loss": 0.3311, + "step": 41082 + }, + { + "epoch": 0.9043143297380586, + "grad_norm": 2.4811837411214523, + "learning_rate": 2.381844881940132e-07, + "loss": 0.2514, + "step": 41083 + }, + { + "epoch": 0.9043363416244772, + "grad_norm": 0.6187235924709573, + "learning_rate": 2.3807579306675155e-07, + "loss": 0.4682, + "step": 41084 + }, + { + "epoch": 0.9043583535108959, + "grad_norm": 1.7183382094069406, + "learning_rate": 2.3796712214171237e-07, + "loss": 0.2296, + "step": 41085 + }, + { + "epoch": 0.9043803653973146, + "grad_norm": 2.287402494973166, + "learning_rate": 2.3785847541944795e-07, + "loss": 0.2988, + "step": 41086 + }, + { + "epoch": 0.9044023772837332, + "grad_norm": 0.6219358691664378, + "learning_rate": 2.3774985290051123e-07, + "loss": 0.4335, + "step": 41087 + }, + { + "epoch": 0.9044243891701519, + "grad_norm": 1.7738653008841347, + "learning_rate": 2.3764125458545284e-07, + "loss": 0.3159, + "step": 41088 + }, + { + "epoch": 0.9044464010565706, + "grad_norm": 1.4086485206581887, + "learning_rate": 2.3753268047482625e-07, + "loss": 0.2097, + "step": 41089 + }, + { + "epoch": 0.9044684129429892, + "grad_norm": 1.6788563350830599, + "learning_rate": 2.3742413056918268e-07, + "loss": 0.2826, + "step": 41090 + }, + { + "epoch": 0.9044904248294079, + "grad_norm": 0.6206555250226822, + "learning_rate": 2.373156048690739e-07, + "loss": 0.434, + "step": 41091 + }, + { + "epoch": 0.9045124367158266, + "grad_norm": 3.3164978220836616, + "learning_rate": 2.372071033750517e-07, + "loss": 0.3224, + "step": 41092 + }, + { + "epoch": 0.9045344486022452, + "grad_norm": 3.809480742993363, + "learning_rate": 2.370986260876662e-07, + "loss": 0.3826, + "step": 41093 + }, + { + "epoch": 0.9045564604886639, + "grad_norm": 1.6866175654786817, + "learning_rate": 2.369901730074703e-07, + "loss": 0.2901, + "step": 41094 + }, + { + "epoch": 0.9045784723750826, + "grad_norm": 1.5124250458672868, + "learning_rate": 2.3688174413501574e-07, + "loss": 0.2524, + "step": 41095 + }, + { + "epoch": 0.9046004842615012, + "grad_norm": 2.1771947936774554, + "learning_rate": 2.3677333947085101e-07, + "loss": 0.2896, + "step": 41096 + }, + { + "epoch": 0.9046224961479199, + "grad_norm": 0.6050312047359894, + "learning_rate": 2.366649590155301e-07, + "loss": 0.4478, + "step": 41097 + }, + { + "epoch": 0.9046445080343385, + "grad_norm": 2.2844713689241054, + "learning_rate": 2.3655660276960146e-07, + "loss": 0.3147, + "step": 41098 + }, + { + "epoch": 0.9046665199207572, + "grad_norm": 2.0047847326890422, + "learning_rate": 2.3644827073361632e-07, + "loss": 0.3194, + "step": 41099 + }, + { + "epoch": 0.9046885318071759, + "grad_norm": 1.7284958647269788, + "learning_rate": 2.3633996290812645e-07, + "loss": 0.2976, + "step": 41100 + }, + { + "epoch": 0.9047105436935945, + "grad_norm": 1.6914067985865913, + "learning_rate": 2.362316792936814e-07, + "loss": 0.2674, + "step": 41101 + }, + { + "epoch": 0.9047325555800132, + "grad_norm": 1.804379087232106, + "learning_rate": 2.3612341989083187e-07, + "loss": 0.2675, + "step": 41102 + }, + { + "epoch": 0.9047545674664319, + "grad_norm": 0.6216133613864869, + "learning_rate": 2.360151847001274e-07, + "loss": 0.4683, + "step": 41103 + }, + { + "epoch": 0.9047765793528505, + "grad_norm": 0.6377895900281615, + "learning_rate": 2.3590697372211812e-07, + "loss": 0.4542, + "step": 41104 + }, + { + "epoch": 0.9047985912392692, + "grad_norm": 1.3657682073249526, + "learning_rate": 2.357987869573558e-07, + "loss": 0.2777, + "step": 41105 + }, + { + "epoch": 0.9048206031256879, + "grad_norm": 1.6176769432496247, + "learning_rate": 2.356906244063878e-07, + "loss": 0.2626, + "step": 41106 + }, + { + "epoch": 0.9048426150121065, + "grad_norm": 2.634734044927357, + "learning_rate": 2.3558248606976586e-07, + "loss": 0.2693, + "step": 41107 + }, + { + "epoch": 0.9048646268985252, + "grad_norm": 2.021239311935131, + "learning_rate": 2.354743719480379e-07, + "loss": 0.2868, + "step": 41108 + }, + { + "epoch": 0.9048866387849439, + "grad_norm": 0.5982388280820088, + "learning_rate": 2.3536628204175405e-07, + "loss": 0.4487, + "step": 41109 + }, + { + "epoch": 0.9049086506713625, + "grad_norm": 2.6596891891354306, + "learning_rate": 2.3525821635146496e-07, + "loss": 0.1715, + "step": 41110 + }, + { + "epoch": 0.9049306625577812, + "grad_norm": 0.6139948165222364, + "learning_rate": 2.3515017487771797e-07, + "loss": 0.4632, + "step": 41111 + }, + { + "epoch": 0.9049526744441999, + "grad_norm": 1.8667322662641104, + "learning_rate": 2.3504215762106376e-07, + "loss": 0.249, + "step": 41112 + }, + { + "epoch": 0.9049746863306185, + "grad_norm": 1.3383254105821272, + "learning_rate": 2.349341645820502e-07, + "loss": 0.3407, + "step": 41113 + }, + { + "epoch": 0.9049966982170372, + "grad_norm": 0.5996296180835989, + "learning_rate": 2.3482619576122523e-07, + "loss": 0.4323, + "step": 41114 + }, + { + "epoch": 0.9050187101034559, + "grad_norm": 3.961682867841517, + "learning_rate": 2.3471825115914006e-07, + "loss": 0.3009, + "step": 41115 + }, + { + "epoch": 0.9050407219898745, + "grad_norm": 1.7546794262848546, + "learning_rate": 2.34610330776342e-07, + "loss": 0.3009, + "step": 41116 + }, + { + "epoch": 0.9050627338762932, + "grad_norm": 1.683165604565699, + "learning_rate": 2.3450243461337952e-07, + "loss": 0.2447, + "step": 41117 + }, + { + "epoch": 0.9050847457627119, + "grad_norm": 1.4373630615759725, + "learning_rate": 2.3439456267080163e-07, + "loss": 0.2248, + "step": 41118 + }, + { + "epoch": 0.9051067576491305, + "grad_norm": 1.3424133569306287, + "learning_rate": 2.3428671494915455e-07, + "loss": 0.2599, + "step": 41119 + }, + { + "epoch": 0.9051287695355492, + "grad_norm": 1.5616684265194993, + "learning_rate": 2.3417889144898953e-07, + "loss": 0.2561, + "step": 41120 + }, + { + "epoch": 0.9051507814219678, + "grad_norm": 1.6679278856236885, + "learning_rate": 2.3407109217085223e-07, + "loss": 0.332, + "step": 41121 + }, + { + "epoch": 0.9051727933083865, + "grad_norm": 0.6356200646234277, + "learning_rate": 2.339633171152922e-07, + "loss": 0.4543, + "step": 41122 + }, + { + "epoch": 0.9051948051948052, + "grad_norm": 2.4770430865606725, + "learning_rate": 2.338555662828562e-07, + "loss": 0.2938, + "step": 41123 + }, + { + "epoch": 0.9052168170812238, + "grad_norm": 0.5878907693855167, + "learning_rate": 2.337478396740911e-07, + "loss": 0.4174, + "step": 41124 + }, + { + "epoch": 0.9052388289676425, + "grad_norm": 2.021397208981032, + "learning_rate": 2.3364013728954526e-07, + "loss": 0.2678, + "step": 41125 + }, + { + "epoch": 0.9052608408540612, + "grad_norm": 1.6863140997710067, + "learning_rate": 2.335324591297672e-07, + "loss": 0.2378, + "step": 41126 + }, + { + "epoch": 0.9052828527404798, + "grad_norm": 1.568932397990287, + "learning_rate": 2.3342480519530197e-07, + "loss": 0.2898, + "step": 41127 + }, + { + "epoch": 0.9053048646268985, + "grad_norm": 1.5226380926033773, + "learning_rate": 2.3331717548669864e-07, + "loss": 0.3163, + "step": 41128 + }, + { + "epoch": 0.9053268765133172, + "grad_norm": 1.6018862745177964, + "learning_rate": 2.3320957000450284e-07, + "loss": 0.279, + "step": 41129 + }, + { + "epoch": 0.9053488883997358, + "grad_norm": 0.6339997791075342, + "learning_rate": 2.3310198874926192e-07, + "loss": 0.4494, + "step": 41130 + }, + { + "epoch": 0.9053709002861545, + "grad_norm": 4.415518882998492, + "learning_rate": 2.3299443172152326e-07, + "loss": 0.2605, + "step": 41131 + }, + { + "epoch": 0.9053929121725732, + "grad_norm": 2.3372502699159203, + "learning_rate": 2.3288689892183248e-07, + "loss": 0.2909, + "step": 41132 + }, + { + "epoch": 0.9054149240589918, + "grad_norm": 1.9247723610312422, + "learning_rate": 2.3277939035073748e-07, + "loss": 0.2352, + "step": 41133 + }, + { + "epoch": 0.9054369359454105, + "grad_norm": 1.7187648433359497, + "learning_rate": 2.3267190600878342e-07, + "loss": 0.331, + "step": 41134 + }, + { + "epoch": 0.9054589478318292, + "grad_norm": 1.6958973988626853, + "learning_rate": 2.325644458965165e-07, + "loss": 0.3045, + "step": 41135 + }, + { + "epoch": 0.9054809597182478, + "grad_norm": 1.763164241437252, + "learning_rate": 2.3245701001448462e-07, + "loss": 0.2227, + "step": 41136 + }, + { + "epoch": 0.9055029716046665, + "grad_norm": 3.0309050323783078, + "learning_rate": 2.323495983632318e-07, + "loss": 0.2022, + "step": 41137 + }, + { + "epoch": 0.9055249834910852, + "grad_norm": 2.0390873346077556, + "learning_rate": 2.3224221094330535e-07, + "loss": 0.3327, + "step": 41138 + }, + { + "epoch": 0.9055469953775038, + "grad_norm": 1.4245158550538652, + "learning_rate": 2.321348477552493e-07, + "loss": 0.2288, + "step": 41139 + }, + { + "epoch": 0.9055690072639225, + "grad_norm": 1.7438157886100059, + "learning_rate": 2.32027508799611e-07, + "loss": 0.3077, + "step": 41140 + }, + { + "epoch": 0.9055910191503412, + "grad_norm": 2.307058489329004, + "learning_rate": 2.319201940769361e-07, + "loss": 0.2863, + "step": 41141 + }, + { + "epoch": 0.9056130310367598, + "grad_norm": 3.2746203126611757, + "learning_rate": 2.3181290358776863e-07, + "loss": 0.2711, + "step": 41142 + }, + { + "epoch": 0.9056350429231785, + "grad_norm": 1.7873754218396207, + "learning_rate": 2.3170563733265539e-07, + "loss": 0.3057, + "step": 41143 + }, + { + "epoch": 0.9056570548095971, + "grad_norm": 1.7906366880552427, + "learning_rate": 2.3159839531214033e-07, + "loss": 0.288, + "step": 41144 + }, + { + "epoch": 0.9056790666960158, + "grad_norm": 3.2287130320664774, + "learning_rate": 2.3149117752676865e-07, + "loss": 0.2986, + "step": 41145 + }, + { + "epoch": 0.9057010785824345, + "grad_norm": 1.585588910420821, + "learning_rate": 2.313839839770865e-07, + "loss": 0.2237, + "step": 41146 + }, + { + "epoch": 0.9057230904688531, + "grad_norm": 2.0585106141617886, + "learning_rate": 2.3127681466363737e-07, + "loss": 0.3316, + "step": 41147 + }, + { + "epoch": 0.9057451023552718, + "grad_norm": 1.7832367708667234, + "learning_rate": 2.311696695869664e-07, + "loss": 0.3095, + "step": 41148 + }, + { + "epoch": 0.9057671142416905, + "grad_norm": 1.8966955357058113, + "learning_rate": 2.3106254874761869e-07, + "loss": 0.3295, + "step": 41149 + }, + { + "epoch": 0.9057891261281091, + "grad_norm": 1.4636588398921924, + "learning_rate": 2.3095545214613657e-07, + "loss": 0.3055, + "step": 41150 + }, + { + "epoch": 0.9058111380145278, + "grad_norm": 2.3969294234433116, + "learning_rate": 2.308483797830674e-07, + "loss": 0.3028, + "step": 41151 + }, + { + "epoch": 0.9058331499009465, + "grad_norm": 2.2227432680426387, + "learning_rate": 2.3074133165895296e-07, + "loss": 0.2504, + "step": 41152 + }, + { + "epoch": 0.9058551617873651, + "grad_norm": 7.491821687436099, + "learning_rate": 2.3063430777433894e-07, + "loss": 0.323, + "step": 41153 + }, + { + "epoch": 0.9058771736737838, + "grad_norm": 1.7620527124536909, + "learning_rate": 2.305273081297682e-07, + "loss": 0.3132, + "step": 41154 + }, + { + "epoch": 0.9058991855602025, + "grad_norm": 3.70304100919904, + "learning_rate": 2.3042033272578424e-07, + "loss": 0.3041, + "step": 41155 + }, + { + "epoch": 0.9059211974466211, + "grad_norm": 1.539660308176579, + "learning_rate": 2.3031338156293215e-07, + "loss": 0.2967, + "step": 41156 + }, + { + "epoch": 0.9059432093330398, + "grad_norm": 1.8445327910980451, + "learning_rate": 2.3020645464175485e-07, + "loss": 0.248, + "step": 41157 + }, + { + "epoch": 0.9059652212194585, + "grad_norm": 1.413852103657814, + "learning_rate": 2.3009955196279576e-07, + "loss": 0.2674, + "step": 41158 + }, + { + "epoch": 0.9059872331058771, + "grad_norm": 1.4349282826304424, + "learning_rate": 2.2999267352659836e-07, + "loss": 0.3006, + "step": 41159 + }, + { + "epoch": 0.9060092449922958, + "grad_norm": 2.0006528784354667, + "learning_rate": 2.2988581933370557e-07, + "loss": 0.3035, + "step": 41160 + }, + { + "epoch": 0.9060312568787146, + "grad_norm": 1.7371084003309418, + "learning_rate": 2.297789893846597e-07, + "loss": 0.2605, + "step": 41161 + }, + { + "epoch": 0.9060532687651331, + "grad_norm": 1.5497517810371222, + "learning_rate": 2.2967218368000588e-07, + "loss": 0.2373, + "step": 41162 + }, + { + "epoch": 0.9060752806515519, + "grad_norm": 2.254484880867875, + "learning_rate": 2.295654022202848e-07, + "loss": 0.2722, + "step": 41163 + }, + { + "epoch": 0.9060972925379706, + "grad_norm": 1.6167244188630598, + "learning_rate": 2.2945864500604043e-07, + "loss": 0.3284, + "step": 41164 + }, + { + "epoch": 0.9061193044243891, + "grad_norm": 1.4083253899510695, + "learning_rate": 2.293519120378146e-07, + "loss": 0.2372, + "step": 41165 + }, + { + "epoch": 0.9061413163108079, + "grad_norm": 1.732789178691538, + "learning_rate": 2.2924520331615018e-07, + "loss": 0.2431, + "step": 41166 + }, + { + "epoch": 0.9061633281972264, + "grad_norm": 1.519311126267846, + "learning_rate": 2.291385188415901e-07, + "loss": 0.346, + "step": 41167 + }, + { + "epoch": 0.9061853400836452, + "grad_norm": 1.4114110126002155, + "learning_rate": 2.2903185861467502e-07, + "loss": 0.2457, + "step": 41168 + }, + { + "epoch": 0.9062073519700639, + "grad_norm": 1.747430378865533, + "learning_rate": 2.289252226359484e-07, + "loss": 0.3645, + "step": 41169 + }, + { + "epoch": 0.9062293638564825, + "grad_norm": 1.5397148463270192, + "learning_rate": 2.2881861090595147e-07, + "loss": 0.2354, + "step": 41170 + }, + { + "epoch": 0.9062513757429012, + "grad_norm": 1.933068395752858, + "learning_rate": 2.2871202342522657e-07, + "loss": 0.317, + "step": 41171 + }, + { + "epoch": 0.9062733876293199, + "grad_norm": 1.6600048754011947, + "learning_rate": 2.286054601943155e-07, + "loss": 0.2882, + "step": 41172 + }, + { + "epoch": 0.9062953995157385, + "grad_norm": 2.073868370824627, + "learning_rate": 2.2849892121375895e-07, + "loss": 0.292, + "step": 41173 + }, + { + "epoch": 0.9063174114021572, + "grad_norm": 1.6115552923148728, + "learning_rate": 2.2839240648409978e-07, + "loss": 0.3119, + "step": 41174 + }, + { + "epoch": 0.9063394232885759, + "grad_norm": 2.223739361474615, + "learning_rate": 2.2828591600587814e-07, + "loss": 0.2559, + "step": 41175 + }, + { + "epoch": 0.9063614351749945, + "grad_norm": 1.4218052884145844, + "learning_rate": 2.2817944977963523e-07, + "loss": 0.3697, + "step": 41176 + }, + { + "epoch": 0.9063834470614132, + "grad_norm": 1.8951842190968025, + "learning_rate": 2.2807300780591345e-07, + "loss": 0.2726, + "step": 41177 + }, + { + "epoch": 0.9064054589478319, + "grad_norm": 1.763041383534089, + "learning_rate": 2.279665900852529e-07, + "loss": 0.308, + "step": 41178 + }, + { + "epoch": 0.9064274708342505, + "grad_norm": 1.53541208458962, + "learning_rate": 2.278601966181948e-07, + "loss": 0.3324, + "step": 41179 + }, + { + "epoch": 0.9064494827206692, + "grad_norm": 2.084062872248273, + "learning_rate": 2.2775382740527874e-07, + "loss": 0.2429, + "step": 41180 + }, + { + "epoch": 0.9064714946070879, + "grad_norm": 1.4529030023020952, + "learning_rate": 2.276474824470465e-07, + "loss": 0.3481, + "step": 41181 + }, + { + "epoch": 0.9064935064935065, + "grad_norm": 2.418970228501981, + "learning_rate": 2.2754116174403873e-07, + "loss": 0.3354, + "step": 41182 + }, + { + "epoch": 0.9065155183799252, + "grad_norm": 1.5785729662653705, + "learning_rate": 2.2743486529679448e-07, + "loss": 0.3566, + "step": 41183 + }, + { + "epoch": 0.9065375302663439, + "grad_norm": 1.5165475802151365, + "learning_rate": 2.2732859310585608e-07, + "loss": 0.273, + "step": 41184 + }, + { + "epoch": 0.9065595421527625, + "grad_norm": 0.6083377197776099, + "learning_rate": 2.27222345171762e-07, + "loss": 0.4401, + "step": 41185 + }, + { + "epoch": 0.9065815540391812, + "grad_norm": 1.9610314145073235, + "learning_rate": 2.271161214950518e-07, + "loss": 0.2734, + "step": 41186 + }, + { + "epoch": 0.9066035659255999, + "grad_norm": 1.7074565270286453, + "learning_rate": 2.2700992207626727e-07, + "loss": 0.3757, + "step": 41187 + }, + { + "epoch": 0.9066255778120185, + "grad_norm": 0.6017343232513457, + "learning_rate": 2.269037469159463e-07, + "loss": 0.4083, + "step": 41188 + }, + { + "epoch": 0.9066475896984372, + "grad_norm": 1.5171270763802696, + "learning_rate": 2.2679759601463015e-07, + "loss": 0.3118, + "step": 41189 + }, + { + "epoch": 0.9066696015848559, + "grad_norm": 2.0106584926556503, + "learning_rate": 2.266914693728578e-07, + "loss": 0.2669, + "step": 41190 + }, + { + "epoch": 0.9066916134712745, + "grad_norm": 1.274924126833024, + "learning_rate": 2.2658536699116774e-07, + "loss": 0.3064, + "step": 41191 + }, + { + "epoch": 0.9067136253576932, + "grad_norm": 1.6795103719629794, + "learning_rate": 2.2647928887009952e-07, + "loss": 0.3576, + "step": 41192 + }, + { + "epoch": 0.9067356372441118, + "grad_norm": 1.2540563752401792, + "learning_rate": 2.263732350101927e-07, + "loss": 0.2712, + "step": 41193 + }, + { + "epoch": 0.9067576491305305, + "grad_norm": 1.6135944905408255, + "learning_rate": 2.2626720541198687e-07, + "loss": 0.2623, + "step": 41194 + }, + { + "epoch": 0.9067796610169492, + "grad_norm": 0.627637959461502, + "learning_rate": 2.2616120007601993e-07, + "loss": 0.462, + "step": 41195 + }, + { + "epoch": 0.9068016729033678, + "grad_norm": 3.917946732325804, + "learning_rate": 2.2605521900283089e-07, + "loss": 0.2683, + "step": 41196 + }, + { + "epoch": 0.9068236847897865, + "grad_norm": 0.6177817881864681, + "learning_rate": 2.2594926219295766e-07, + "loss": 0.4687, + "step": 41197 + }, + { + "epoch": 0.9068456966762052, + "grad_norm": 0.6329613438239762, + "learning_rate": 2.2584332964694033e-07, + "loss": 0.4418, + "step": 41198 + }, + { + "epoch": 0.9068677085626238, + "grad_norm": 1.8411573738847118, + "learning_rate": 2.2573742136531684e-07, + "loss": 0.4213, + "step": 41199 + }, + { + "epoch": 0.9068897204490425, + "grad_norm": 1.452308170757048, + "learning_rate": 2.2563153734862508e-07, + "loss": 0.355, + "step": 41200 + }, + { + "epoch": 0.9069117323354612, + "grad_norm": 2.8563138122311833, + "learning_rate": 2.2552567759740296e-07, + "loss": 0.2785, + "step": 41201 + }, + { + "epoch": 0.9069337442218798, + "grad_norm": 1.7095232204548372, + "learning_rate": 2.2541984211218893e-07, + "loss": 0.3755, + "step": 41202 + }, + { + "epoch": 0.9069557561082985, + "grad_norm": 0.6076435341724219, + "learning_rate": 2.2531403089352087e-07, + "loss": 0.4481, + "step": 41203 + }, + { + "epoch": 0.9069777679947172, + "grad_norm": 0.6155695841846731, + "learning_rate": 2.2520824394193618e-07, + "loss": 0.4669, + "step": 41204 + }, + { + "epoch": 0.9069997798811358, + "grad_norm": 1.7384189614026964, + "learning_rate": 2.251024812579733e-07, + "loss": 0.2095, + "step": 41205 + }, + { + "epoch": 0.9070217917675545, + "grad_norm": 2.6456495764341246, + "learning_rate": 2.24996742842169e-07, + "loss": 0.2984, + "step": 41206 + }, + { + "epoch": 0.9070438036539732, + "grad_norm": 1.3260555792141997, + "learning_rate": 2.2489102869506062e-07, + "loss": 0.3004, + "step": 41207 + }, + { + "epoch": 0.9070658155403918, + "grad_norm": 1.4249597616967051, + "learning_rate": 2.2478533881718612e-07, + "loss": 0.2218, + "step": 41208 + }, + { + "epoch": 0.9070878274268105, + "grad_norm": 1.874175147430713, + "learning_rate": 2.2467967320908225e-07, + "loss": 0.3002, + "step": 41209 + }, + { + "epoch": 0.9071098393132292, + "grad_norm": 1.8642818833319157, + "learning_rate": 2.2457403187128635e-07, + "loss": 0.393, + "step": 41210 + }, + { + "epoch": 0.9071318511996478, + "grad_norm": 1.3725588468097956, + "learning_rate": 2.2446841480433467e-07, + "loss": 0.2425, + "step": 41211 + }, + { + "epoch": 0.9071538630860665, + "grad_norm": 1.3637739353825304, + "learning_rate": 2.243628220087646e-07, + "loss": 0.3241, + "step": 41212 + }, + { + "epoch": 0.9071758749724852, + "grad_norm": 1.5535935392189142, + "learning_rate": 2.242572534851134e-07, + "loss": 0.3009, + "step": 41213 + }, + { + "epoch": 0.9071978868589038, + "grad_norm": 1.432279406650046, + "learning_rate": 2.241517092339157e-07, + "loss": 0.3002, + "step": 41214 + }, + { + "epoch": 0.9072198987453225, + "grad_norm": 1.3135589698191215, + "learning_rate": 2.2404618925571053e-07, + "loss": 0.2397, + "step": 41215 + }, + { + "epoch": 0.9072419106317411, + "grad_norm": 2.043524403413254, + "learning_rate": 2.239406935510313e-07, + "loss": 0.248, + "step": 41216 + }, + { + "epoch": 0.9072639225181598, + "grad_norm": 2.027952616318113, + "learning_rate": 2.2383522212041598e-07, + "loss": 0.3467, + "step": 41217 + }, + { + "epoch": 0.9072859344045785, + "grad_norm": 0.6405438936077563, + "learning_rate": 2.2372977496440128e-07, + "loss": 0.4909, + "step": 41218 + }, + { + "epoch": 0.9073079462909971, + "grad_norm": 1.293559093852933, + "learning_rate": 2.2362435208352128e-07, + "loss": 0.1966, + "step": 41219 + }, + { + "epoch": 0.9073299581774158, + "grad_norm": 1.7955833550141116, + "learning_rate": 2.2351895347831275e-07, + "loss": 0.3225, + "step": 41220 + }, + { + "epoch": 0.9073519700638345, + "grad_norm": 2.908437794306515, + "learning_rate": 2.234135791493114e-07, + "loss": 0.2765, + "step": 41221 + }, + { + "epoch": 0.9073739819502531, + "grad_norm": 1.8415787360190676, + "learning_rate": 2.233082290970523e-07, + "loss": 0.3621, + "step": 41222 + }, + { + "epoch": 0.9073959938366718, + "grad_norm": 3.3056012311277096, + "learning_rate": 2.2320290332207174e-07, + "loss": 0.2475, + "step": 41223 + }, + { + "epoch": 0.9074180057230905, + "grad_norm": 0.6163859235945551, + "learning_rate": 2.2309760182490425e-07, + "loss": 0.4365, + "step": 41224 + }, + { + "epoch": 0.9074400176095091, + "grad_norm": 1.608155283384227, + "learning_rate": 2.2299232460608556e-07, + "loss": 0.3187, + "step": 41225 + }, + { + "epoch": 0.9074620294959278, + "grad_norm": 1.6545241366825465, + "learning_rate": 2.2288707166615075e-07, + "loss": 0.2859, + "step": 41226 + }, + { + "epoch": 0.9074840413823465, + "grad_norm": 1.5190771757164052, + "learning_rate": 2.2278184300563388e-07, + "loss": 0.2792, + "step": 41227 + }, + { + "epoch": 0.9075060532687651, + "grad_norm": 0.6281672802699387, + "learning_rate": 2.2267663862507004e-07, + "loss": 0.4594, + "step": 41228 + }, + { + "epoch": 0.9075280651551838, + "grad_norm": 1.2425863130988022, + "learning_rate": 2.2257145852499495e-07, + "loss": 0.2691, + "step": 41229 + }, + { + "epoch": 0.9075500770416025, + "grad_norm": 1.7426192823318196, + "learning_rate": 2.224663027059426e-07, + "loss": 0.3415, + "step": 41230 + }, + { + "epoch": 0.9075720889280211, + "grad_norm": 1.8312472983741015, + "learning_rate": 2.2236117116844756e-07, + "loss": 0.329, + "step": 41231 + }, + { + "epoch": 0.9075941008144398, + "grad_norm": 9.448206202160074, + "learning_rate": 2.2225606391304277e-07, + "loss": 0.263, + "step": 41232 + }, + { + "epoch": 0.9076161127008585, + "grad_norm": 1.8470225103700202, + "learning_rate": 2.2215098094026443e-07, + "loss": 0.2797, + "step": 41233 + }, + { + "epoch": 0.9076381245872771, + "grad_norm": 1.7615574277807555, + "learning_rate": 2.2204592225064492e-07, + "loss": 0.318, + "step": 41234 + }, + { + "epoch": 0.9076601364736958, + "grad_norm": 0.6364432914224003, + "learning_rate": 2.2194088784472044e-07, + "loss": 0.4756, + "step": 41235 + }, + { + "epoch": 0.9076821483601145, + "grad_norm": 2.016734432314569, + "learning_rate": 2.2183587772302283e-07, + "loss": 0.3328, + "step": 41236 + }, + { + "epoch": 0.9077041602465331, + "grad_norm": 2.323288878335551, + "learning_rate": 2.217308918860861e-07, + "loss": 0.3133, + "step": 41237 + }, + { + "epoch": 0.9077261721329518, + "grad_norm": 2.445116789108379, + "learning_rate": 2.2162593033444425e-07, + "loss": 0.2542, + "step": 41238 + }, + { + "epoch": 0.9077481840193704, + "grad_norm": 0.6184740312770893, + "learning_rate": 2.2152099306863073e-07, + "loss": 0.4468, + "step": 41239 + }, + { + "epoch": 0.9077701959057891, + "grad_norm": 1.5758696002410035, + "learning_rate": 2.2141608008917902e-07, + "loss": 0.2485, + "step": 41240 + }, + { + "epoch": 0.9077922077922078, + "grad_norm": 2.4610511956420473, + "learning_rate": 2.2131119139662262e-07, + "loss": 0.2426, + "step": 41241 + }, + { + "epoch": 0.9078142196786264, + "grad_norm": 1.76677594716208, + "learning_rate": 2.212063269914927e-07, + "loss": 0.2594, + "step": 41242 + }, + { + "epoch": 0.9078362315650451, + "grad_norm": 1.6082109471836703, + "learning_rate": 2.2110148687432443e-07, + "loss": 0.2014, + "step": 41243 + }, + { + "epoch": 0.9078582434514638, + "grad_norm": 1.803728545867148, + "learning_rate": 2.2099667104565015e-07, + "loss": 0.3118, + "step": 41244 + }, + { + "epoch": 0.9078802553378824, + "grad_norm": 1.6987868513509203, + "learning_rate": 2.2089187950600166e-07, + "loss": 0.2945, + "step": 41245 + }, + { + "epoch": 0.9079022672243011, + "grad_norm": 1.8290848491142646, + "learning_rate": 2.2078711225591244e-07, + "loss": 0.3348, + "step": 41246 + }, + { + "epoch": 0.9079242791107198, + "grad_norm": 2.2145647911160995, + "learning_rate": 2.2068236929591425e-07, + "loss": 0.2784, + "step": 41247 + }, + { + "epoch": 0.9079462909971384, + "grad_norm": 2.0247948581106474, + "learning_rate": 2.2057765062654e-07, + "loss": 0.2245, + "step": 41248 + }, + { + "epoch": 0.9079683028835571, + "grad_norm": 0.6160151686384032, + "learning_rate": 2.2047295624832265e-07, + "loss": 0.4482, + "step": 41249 + }, + { + "epoch": 0.9079903147699758, + "grad_norm": 0.6130066888285259, + "learning_rate": 2.2036828616179283e-07, + "loss": 0.4472, + "step": 41250 + }, + { + "epoch": 0.9080123266563944, + "grad_norm": 1.6060825967708179, + "learning_rate": 2.2026364036748348e-07, + "loss": 0.2663, + "step": 41251 + }, + { + "epoch": 0.9080343385428131, + "grad_norm": 3.3495313239866307, + "learning_rate": 2.2015901886592582e-07, + "loss": 0.3623, + "step": 41252 + }, + { + "epoch": 0.9080563504292318, + "grad_norm": 2.782463628967782, + "learning_rate": 2.2005442165765166e-07, + "loss": 0.1992, + "step": 41253 + }, + { + "epoch": 0.9080783623156504, + "grad_norm": 1.7394760320601508, + "learning_rate": 2.1994984874319336e-07, + "loss": 0.3559, + "step": 41254 + }, + { + "epoch": 0.9081003742020691, + "grad_norm": 1.4538408107549532, + "learning_rate": 2.198453001230816e-07, + "loss": 0.2611, + "step": 41255 + }, + { + "epoch": 0.9081223860884878, + "grad_norm": 1.9991872337818777, + "learning_rate": 2.197407757978487e-07, + "loss": 0.2358, + "step": 41256 + }, + { + "epoch": 0.9081443979749064, + "grad_norm": 1.5715466863476604, + "learning_rate": 2.1963627576802427e-07, + "loss": 0.2977, + "step": 41257 + }, + { + "epoch": 0.9081664098613251, + "grad_norm": 1.3829866779776123, + "learning_rate": 2.1953180003414064e-07, + "loss": 0.2509, + "step": 41258 + }, + { + "epoch": 0.9081884217477438, + "grad_norm": 1.6155522734829377, + "learning_rate": 2.194273485967291e-07, + "loss": 0.2856, + "step": 41259 + }, + { + "epoch": 0.9082104336341624, + "grad_norm": 2.9916619487119807, + "learning_rate": 2.1932292145631916e-07, + "loss": 0.2577, + "step": 41260 + }, + { + "epoch": 0.9082324455205811, + "grad_norm": 2.5963802989653306, + "learning_rate": 2.1921851861344323e-07, + "loss": 0.2182, + "step": 41261 + }, + { + "epoch": 0.9082544574069997, + "grad_norm": 1.5322634293376025, + "learning_rate": 2.1911414006863086e-07, + "loss": 0.2406, + "step": 41262 + }, + { + "epoch": 0.9082764692934184, + "grad_norm": 1.7621995520209648, + "learning_rate": 2.1900978582241218e-07, + "loss": 0.3365, + "step": 41263 + }, + { + "epoch": 0.9082984811798371, + "grad_norm": 1.7885473817100017, + "learning_rate": 2.1890545587531842e-07, + "loss": 0.2186, + "step": 41264 + }, + { + "epoch": 0.9083204930662557, + "grad_norm": 1.8475026512601942, + "learning_rate": 2.1880115022787917e-07, + "loss": 0.3438, + "step": 41265 + }, + { + "epoch": 0.9083425049526744, + "grad_norm": 3.5498649138584057, + "learning_rate": 2.1869686888062568e-07, + "loss": 0.3047, + "step": 41266 + }, + { + "epoch": 0.9083645168390931, + "grad_norm": 1.8070080443811425, + "learning_rate": 2.1859261183408753e-07, + "loss": 0.3272, + "step": 41267 + }, + { + "epoch": 0.9083865287255117, + "grad_norm": 0.5979694711001462, + "learning_rate": 2.184883790887937e-07, + "loss": 0.4176, + "step": 41268 + }, + { + "epoch": 0.9084085406119304, + "grad_norm": 0.5916216331190086, + "learning_rate": 2.1838417064527384e-07, + "loss": 0.4246, + "step": 41269 + }, + { + "epoch": 0.9084305524983491, + "grad_norm": 1.3533073745210504, + "learning_rate": 2.1827998650405914e-07, + "loss": 0.2259, + "step": 41270 + }, + { + "epoch": 0.9084525643847677, + "grad_norm": 3.0416522182355985, + "learning_rate": 2.181758266656786e-07, + "loss": 0.2721, + "step": 41271 + }, + { + "epoch": 0.9084745762711864, + "grad_norm": 3.836785238245555, + "learning_rate": 2.1807169113066072e-07, + "loss": 0.2979, + "step": 41272 + }, + { + "epoch": 0.9084965881576051, + "grad_norm": 2.55626463228239, + "learning_rate": 2.1796757989953508e-07, + "loss": 0.285, + "step": 41273 + }, + { + "epoch": 0.9085186000440237, + "grad_norm": 1.1102396198690307, + "learning_rate": 2.1786349297283127e-07, + "loss": 0.2279, + "step": 41274 + }, + { + "epoch": 0.9085406119304424, + "grad_norm": 1.4336524565949087, + "learning_rate": 2.1775943035107772e-07, + "loss": 0.2527, + "step": 41275 + }, + { + "epoch": 0.9085626238168611, + "grad_norm": 2.197628866648345, + "learning_rate": 2.1765539203480458e-07, + "loss": 0.3624, + "step": 41276 + }, + { + "epoch": 0.9085846357032797, + "grad_norm": 7.702530961019106, + "learning_rate": 2.1755137802453974e-07, + "loss": 0.3824, + "step": 41277 + }, + { + "epoch": 0.9086066475896984, + "grad_norm": 1.8615474125604183, + "learning_rate": 2.1744738832081114e-07, + "loss": 0.2127, + "step": 41278 + }, + { + "epoch": 0.9086286594761172, + "grad_norm": 0.5957422337654259, + "learning_rate": 2.173434229241478e-07, + "loss": 0.4349, + "step": 41279 + }, + { + "epoch": 0.9086506713625357, + "grad_norm": 1.6063977174679405, + "learning_rate": 2.172394818350787e-07, + "loss": 0.2783, + "step": 41280 + }, + { + "epoch": 0.9086726832489544, + "grad_norm": 2.217632593047783, + "learning_rate": 2.1713556505413181e-07, + "loss": 0.2816, + "step": 41281 + }, + { + "epoch": 0.9086946951353732, + "grad_norm": 2.5301603660016156, + "learning_rate": 2.1703167258183554e-07, + "loss": 0.2917, + "step": 41282 + }, + { + "epoch": 0.9087167070217917, + "grad_norm": 1.7464991488603558, + "learning_rate": 2.1692780441871676e-07, + "loss": 0.2443, + "step": 41283 + }, + { + "epoch": 0.9087387189082105, + "grad_norm": 3.433158477169338, + "learning_rate": 2.168239605653044e-07, + "loss": 0.3065, + "step": 41284 + }, + { + "epoch": 0.908760730794629, + "grad_norm": 2.114474326444937, + "learning_rate": 2.1672014102212647e-07, + "loss": 0.3016, + "step": 41285 + }, + { + "epoch": 0.9087827426810478, + "grad_norm": 2.1364665808987313, + "learning_rate": 2.166163457897097e-07, + "loss": 0.3885, + "step": 41286 + }, + { + "epoch": 0.9088047545674665, + "grad_norm": 0.6106101300276033, + "learning_rate": 2.1651257486858312e-07, + "loss": 0.4144, + "step": 41287 + }, + { + "epoch": 0.908826766453885, + "grad_norm": 1.7806701151994526, + "learning_rate": 2.164088282592719e-07, + "loss": 0.3175, + "step": 41288 + }, + { + "epoch": 0.9088487783403038, + "grad_norm": 1.8612610114357258, + "learning_rate": 2.1630510596230448e-07, + "loss": 0.333, + "step": 41289 + }, + { + "epoch": 0.9088707902267225, + "grad_norm": 1.7106337470391302, + "learning_rate": 2.1620140797820933e-07, + "loss": 0.32, + "step": 41290 + }, + { + "epoch": 0.908892802113141, + "grad_norm": 2.187036533977993, + "learning_rate": 2.1609773430751102e-07, + "loss": 0.3189, + "step": 41291 + }, + { + "epoch": 0.9089148139995598, + "grad_norm": 1.9827021281747728, + "learning_rate": 2.159940849507386e-07, + "loss": 0.2274, + "step": 41292 + }, + { + "epoch": 0.9089368258859785, + "grad_norm": 1.2866449479113102, + "learning_rate": 2.1589045990841717e-07, + "loss": 0.2903, + "step": 41293 + }, + { + "epoch": 0.908958837772397, + "grad_norm": 2.884108540503994, + "learning_rate": 2.1578685918107467e-07, + "loss": 0.3167, + "step": 41294 + }, + { + "epoch": 0.9089808496588158, + "grad_norm": 1.7466489901271038, + "learning_rate": 2.1568328276923734e-07, + "loss": 0.262, + "step": 41295 + }, + { + "epoch": 0.9090028615452345, + "grad_norm": 3.5185754828341467, + "learning_rate": 2.1557973067343085e-07, + "loss": 0.4007, + "step": 41296 + }, + { + "epoch": 0.9090248734316531, + "grad_norm": 1.8907534986269192, + "learning_rate": 2.1547620289418315e-07, + "loss": 0.3345, + "step": 41297 + }, + { + "epoch": 0.9090468853180718, + "grad_norm": 0.613765177744152, + "learning_rate": 2.1537269943201822e-07, + "loss": 0.4469, + "step": 41298 + }, + { + "epoch": 0.9090688972044905, + "grad_norm": 2.4233925507530434, + "learning_rate": 2.15269220287464e-07, + "loss": 0.3794, + "step": 41299 + }, + { + "epoch": 0.9090909090909091, + "grad_norm": 1.799694066548395, + "learning_rate": 2.1516576546104507e-07, + "loss": 0.3732, + "step": 41300 + }, + { + "epoch": 0.9091129209773278, + "grad_norm": 1.7060500485885537, + "learning_rate": 2.1506233495328766e-07, + "loss": 0.2848, + "step": 41301 + }, + { + "epoch": 0.9091349328637465, + "grad_norm": 2.325021775559113, + "learning_rate": 2.1495892876471857e-07, + "loss": 0.3725, + "step": 41302 + }, + { + "epoch": 0.9091569447501651, + "grad_norm": 1.7605131065704334, + "learning_rate": 2.1485554689586185e-07, + "loss": 0.2786, + "step": 41303 + }, + { + "epoch": 0.9091789566365838, + "grad_norm": 2.0520061086265837, + "learning_rate": 2.1475218934724318e-07, + "loss": 0.3106, + "step": 41304 + }, + { + "epoch": 0.9092009685230025, + "grad_norm": 1.5911491730628424, + "learning_rate": 2.1464885611938825e-07, + "loss": 0.335, + "step": 41305 + }, + { + "epoch": 0.9092229804094211, + "grad_norm": 1.874261746828865, + "learning_rate": 2.1454554721282218e-07, + "loss": 0.3475, + "step": 41306 + }, + { + "epoch": 0.9092449922958398, + "grad_norm": 0.6217314287059442, + "learning_rate": 2.1444226262807067e-07, + "loss": 0.4466, + "step": 41307 + }, + { + "epoch": 0.9092670041822584, + "grad_norm": 0.6126892734183925, + "learning_rate": 2.1433900236565775e-07, + "loss": 0.4312, + "step": 41308 + }, + { + "epoch": 0.9092890160686771, + "grad_norm": 1.4005744048641875, + "learning_rate": 2.1423576642610798e-07, + "loss": 0.3085, + "step": 41309 + }, + { + "epoch": 0.9093110279550958, + "grad_norm": 1.4851164356701116, + "learning_rate": 2.1413255480994654e-07, + "loss": 0.3169, + "step": 41310 + }, + { + "epoch": 0.9093330398415144, + "grad_norm": 1.9230469420901055, + "learning_rate": 2.1402936751769799e-07, + "loss": 0.3151, + "step": 41311 + }, + { + "epoch": 0.9093550517279331, + "grad_norm": 1.3376998972241658, + "learning_rate": 2.1392620454988744e-07, + "loss": 0.3667, + "step": 41312 + }, + { + "epoch": 0.9093770636143518, + "grad_norm": 2.879147557128497, + "learning_rate": 2.138230659070384e-07, + "loss": 0.3724, + "step": 41313 + }, + { + "epoch": 0.9093990755007704, + "grad_norm": 2.066187331348714, + "learning_rate": 2.1371995158967484e-07, + "loss": 0.2114, + "step": 41314 + }, + { + "epoch": 0.9094210873871891, + "grad_norm": 1.595981780600093, + "learning_rate": 2.1361686159832141e-07, + "loss": 0.3201, + "step": 41315 + }, + { + "epoch": 0.9094430992736078, + "grad_norm": 2.293747592323078, + "learning_rate": 2.1351379593350152e-07, + "loss": 0.2776, + "step": 41316 + }, + { + "epoch": 0.9094651111600264, + "grad_norm": 1.2115907766523646, + "learning_rate": 2.1341075459574033e-07, + "loss": 0.1704, + "step": 41317 + }, + { + "epoch": 0.9094871230464451, + "grad_norm": 2.2129795925346922, + "learning_rate": 2.1330773758556022e-07, + "loss": 0.3073, + "step": 41318 + }, + { + "epoch": 0.9095091349328638, + "grad_norm": 1.4439307418235283, + "learning_rate": 2.1320474490348463e-07, + "loss": 0.2563, + "step": 41319 + }, + { + "epoch": 0.9095311468192824, + "grad_norm": 1.514370466480452, + "learning_rate": 2.131017765500376e-07, + "loss": 0.2881, + "step": 41320 + }, + { + "epoch": 0.9095531587057011, + "grad_norm": 1.4879113478838524, + "learning_rate": 2.1299883252574206e-07, + "loss": 0.237, + "step": 41321 + }, + { + "epoch": 0.9095751705921198, + "grad_norm": 2.3720769452906683, + "learning_rate": 2.12895912831122e-07, + "loss": 0.3007, + "step": 41322 + }, + { + "epoch": 0.9095971824785384, + "grad_norm": 1.72325286311528, + "learning_rate": 2.1279301746670033e-07, + "loss": 0.2551, + "step": 41323 + }, + { + "epoch": 0.9096191943649571, + "grad_norm": 1.2222060072395413, + "learning_rate": 2.1269014643299946e-07, + "loss": 0.2827, + "step": 41324 + }, + { + "epoch": 0.9096412062513758, + "grad_norm": 0.6335207217057077, + "learning_rate": 2.1258729973054172e-07, + "loss": 0.4531, + "step": 41325 + }, + { + "epoch": 0.9096632181377944, + "grad_norm": 0.6134795561686189, + "learning_rate": 2.1248447735985167e-07, + "loss": 0.4456, + "step": 41326 + }, + { + "epoch": 0.9096852300242131, + "grad_norm": 1.5925397384809619, + "learning_rate": 2.1238167932145004e-07, + "loss": 0.2881, + "step": 41327 + }, + { + "epoch": 0.9097072419106318, + "grad_norm": 0.6410536078405515, + "learning_rate": 2.1227890561586084e-07, + "loss": 0.4738, + "step": 41328 + }, + { + "epoch": 0.9097292537970504, + "grad_norm": 2.0095833277343944, + "learning_rate": 2.1217615624360476e-07, + "loss": 0.3709, + "step": 41329 + }, + { + "epoch": 0.9097512656834691, + "grad_norm": 1.7480975049577858, + "learning_rate": 2.1207343120520474e-07, + "loss": 0.3485, + "step": 41330 + }, + { + "epoch": 0.9097732775698877, + "grad_norm": 2.0968354691225786, + "learning_rate": 2.1197073050118422e-07, + "loss": 0.2564, + "step": 41331 + }, + { + "epoch": 0.9097952894563064, + "grad_norm": 2.11832765379158, + "learning_rate": 2.118680541320628e-07, + "loss": 0.2977, + "step": 41332 + }, + { + "epoch": 0.9098173013427251, + "grad_norm": 7.666814408520589, + "learning_rate": 2.117654020983645e-07, + "loss": 0.333, + "step": 41333 + }, + { + "epoch": 0.9098393132291437, + "grad_norm": 2.5438886987013847, + "learning_rate": 2.116627744006089e-07, + "loss": 0.293, + "step": 41334 + }, + { + "epoch": 0.9098613251155624, + "grad_norm": 1.8439297636147285, + "learning_rate": 2.1156017103932002e-07, + "loss": 0.3002, + "step": 41335 + }, + { + "epoch": 0.9098833370019811, + "grad_norm": 19.751744961016012, + "learning_rate": 2.1145759201501693e-07, + "loss": 0.3418, + "step": 41336 + }, + { + "epoch": 0.9099053488883997, + "grad_norm": 0.6175067275485456, + "learning_rate": 2.113550373282225e-07, + "loss": 0.4485, + "step": 41337 + }, + { + "epoch": 0.9099273607748184, + "grad_norm": 8.717975771309712, + "learning_rate": 2.1125250697945797e-07, + "loss": 0.2656, + "step": 41338 + }, + { + "epoch": 0.9099493726612371, + "grad_norm": 2.048487256002692, + "learning_rate": 2.1115000096924297e-07, + "loss": 0.2592, + "step": 41339 + }, + { + "epoch": 0.9099713845476557, + "grad_norm": 0.6087167417963923, + "learning_rate": 2.1104751929810096e-07, + "loss": 0.4408, + "step": 41340 + }, + { + "epoch": 0.9099933964340744, + "grad_norm": 1.5902411068896134, + "learning_rate": 2.109450619665504e-07, + "loss": 0.2479, + "step": 41341 + }, + { + "epoch": 0.9100154083204931, + "grad_norm": 1.6000868301640967, + "learning_rate": 2.108426289751131e-07, + "loss": 0.3003, + "step": 41342 + }, + { + "epoch": 0.9100374202069117, + "grad_norm": 2.2832855158147676, + "learning_rate": 2.107402203243103e-07, + "loss": 0.25, + "step": 41343 + }, + { + "epoch": 0.9100594320933304, + "grad_norm": 1.4882043151672704, + "learning_rate": 2.1063783601466158e-07, + "loss": 0.2112, + "step": 41344 + }, + { + "epoch": 0.9100814439797491, + "grad_norm": 1.3505882549925385, + "learning_rate": 2.1053547604668656e-07, + "loss": 0.2527, + "step": 41345 + }, + { + "epoch": 0.9101034558661677, + "grad_norm": 0.59367016908692, + "learning_rate": 2.1043314042090702e-07, + "loss": 0.4492, + "step": 41346 + }, + { + "epoch": 0.9101254677525864, + "grad_norm": 0.6069586753339555, + "learning_rate": 2.1033082913784198e-07, + "loss": 0.4345, + "step": 41347 + }, + { + "epoch": 0.9101474796390051, + "grad_norm": 1.3842230302057377, + "learning_rate": 2.102285421980127e-07, + "loss": 0.2586, + "step": 41348 + }, + { + "epoch": 0.9101694915254237, + "grad_norm": 0.6340440051965993, + "learning_rate": 2.101262796019382e-07, + "loss": 0.4602, + "step": 41349 + }, + { + "epoch": 0.9101915034118424, + "grad_norm": 1.401902637075458, + "learning_rate": 2.100240413501381e-07, + "loss": 0.2896, + "step": 41350 + }, + { + "epoch": 0.9102135152982611, + "grad_norm": 1.8346916649916092, + "learning_rate": 2.0992182744313193e-07, + "loss": 0.2655, + "step": 41351 + }, + { + "epoch": 0.9102355271846797, + "grad_norm": 1.693757710430037, + "learning_rate": 2.0981963788143934e-07, + "loss": 0.2823, + "step": 41352 + }, + { + "epoch": 0.9102575390710984, + "grad_norm": 1.5833892398544704, + "learning_rate": 2.097174726655804e-07, + "loss": 0.2426, + "step": 41353 + }, + { + "epoch": 0.9102795509575171, + "grad_norm": 1.6816825597984009, + "learning_rate": 2.0961533179607417e-07, + "loss": 0.3122, + "step": 41354 + }, + { + "epoch": 0.9103015628439357, + "grad_norm": 1.6905028615130262, + "learning_rate": 2.0951321527343858e-07, + "loss": 0.2457, + "step": 41355 + }, + { + "epoch": 0.9103235747303544, + "grad_norm": 3.7372921973056963, + "learning_rate": 2.0941112309819324e-07, + "loss": 0.2919, + "step": 41356 + }, + { + "epoch": 0.910345586616773, + "grad_norm": 1.5594039682461616, + "learning_rate": 2.0930905527085766e-07, + "loss": 0.2523, + "step": 41357 + }, + { + "epoch": 0.9103675985031917, + "grad_norm": 1.59448564677781, + "learning_rate": 2.0920701179195036e-07, + "loss": 0.3574, + "step": 41358 + }, + { + "epoch": 0.9103896103896104, + "grad_norm": 2.544916801720394, + "learning_rate": 2.091049926619898e-07, + "loss": 0.307, + "step": 41359 + }, + { + "epoch": 0.910411622276029, + "grad_norm": 1.7563518440918418, + "learning_rate": 2.0900299788149448e-07, + "loss": 0.2903, + "step": 41360 + }, + { + "epoch": 0.9104336341624477, + "grad_norm": 0.6328239060490828, + "learning_rate": 2.0890102745098228e-07, + "loss": 0.4549, + "step": 41361 + }, + { + "epoch": 0.9104556460488664, + "grad_norm": 1.5120000304932695, + "learning_rate": 2.0879908137097226e-07, + "loss": 0.2173, + "step": 41362 + }, + { + "epoch": 0.910477657935285, + "grad_norm": 2.3261898225558824, + "learning_rate": 2.0869715964198178e-07, + "loss": 0.2175, + "step": 41363 + }, + { + "epoch": 0.9104996698217037, + "grad_norm": 1.6122781949593277, + "learning_rate": 2.0859526226453042e-07, + "loss": 0.3166, + "step": 41364 + }, + { + "epoch": 0.9105216817081224, + "grad_norm": 1.5019455374030517, + "learning_rate": 2.0849338923913386e-07, + "loss": 0.2564, + "step": 41365 + }, + { + "epoch": 0.910543693594541, + "grad_norm": 1.4500189362309575, + "learning_rate": 2.0839154056631117e-07, + "loss": 0.2061, + "step": 41366 + }, + { + "epoch": 0.9105657054809597, + "grad_norm": 1.4724454801356117, + "learning_rate": 2.0828971624658023e-07, + "loss": 0.2295, + "step": 41367 + }, + { + "epoch": 0.9105877173673784, + "grad_norm": 1.5118418240756195, + "learning_rate": 2.0818791628045787e-07, + "loss": 0.2718, + "step": 41368 + }, + { + "epoch": 0.910609729253797, + "grad_norm": 2.0103034021291526, + "learning_rate": 2.08086140668462e-07, + "loss": 0.2775, + "step": 41369 + }, + { + "epoch": 0.9106317411402157, + "grad_norm": 0.6312244276391845, + "learning_rate": 2.0798438941110943e-07, + "loss": 0.4421, + "step": 41370 + }, + { + "epoch": 0.9106537530266344, + "grad_norm": 1.750550683381666, + "learning_rate": 2.0788266250891754e-07, + "loss": 0.2384, + "step": 41371 + }, + { + "epoch": 0.910675764913053, + "grad_norm": 1.5201071299023523, + "learning_rate": 2.077809599624031e-07, + "loss": 0.2408, + "step": 41372 + }, + { + "epoch": 0.9106977767994717, + "grad_norm": 1.3177194847464397, + "learning_rate": 2.0767928177208298e-07, + "loss": 0.2022, + "step": 41373 + }, + { + "epoch": 0.9107197886858904, + "grad_norm": 1.8489629074357752, + "learning_rate": 2.0757762793847503e-07, + "loss": 0.3588, + "step": 41374 + }, + { + "epoch": 0.910741800572309, + "grad_norm": 2.176753901909109, + "learning_rate": 2.0747599846209387e-07, + "loss": 0.3066, + "step": 41375 + }, + { + "epoch": 0.9107638124587277, + "grad_norm": 1.6810084157735565, + "learning_rate": 2.0737439334345856e-07, + "loss": 0.3666, + "step": 41376 + }, + { + "epoch": 0.9107858243451464, + "grad_norm": 2.118080649526562, + "learning_rate": 2.0727281258308307e-07, + "loss": 0.2447, + "step": 41377 + }, + { + "epoch": 0.910807836231565, + "grad_norm": 0.6131207038093592, + "learning_rate": 2.0717125618148483e-07, + "loss": 0.453, + "step": 41378 + }, + { + "epoch": 0.9108298481179837, + "grad_norm": 1.9832529366233296, + "learning_rate": 2.0706972413918058e-07, + "loss": 0.3068, + "step": 41379 + }, + { + "epoch": 0.9108518600044023, + "grad_norm": 2.673417095655481, + "learning_rate": 2.069682164566855e-07, + "loss": 0.2798, + "step": 41380 + }, + { + "epoch": 0.910873871890821, + "grad_norm": 1.4376312279965644, + "learning_rate": 2.0686673313451588e-07, + "loss": 0.2402, + "step": 41381 + }, + { + "epoch": 0.9108958837772397, + "grad_norm": 0.6235231323772278, + "learning_rate": 2.0676527417318682e-07, + "loss": 0.4647, + "step": 41382 + }, + { + "epoch": 0.9109178956636583, + "grad_norm": 0.6121733545493449, + "learning_rate": 2.06663839573214e-07, + "loss": 0.4181, + "step": 41383 + }, + { + "epoch": 0.910939907550077, + "grad_norm": 2.3249485583891065, + "learning_rate": 2.0656242933511484e-07, + "loss": 0.3723, + "step": 41384 + }, + { + "epoch": 0.9109619194364957, + "grad_norm": 1.574263637634563, + "learning_rate": 2.0646104345940275e-07, + "loss": 0.2546, + "step": 41385 + }, + { + "epoch": 0.9109839313229143, + "grad_norm": 1.9743431828465792, + "learning_rate": 2.063596819465935e-07, + "loss": 0.3025, + "step": 41386 + }, + { + "epoch": 0.911005943209333, + "grad_norm": 1.6046788448933886, + "learning_rate": 2.0625834479720163e-07, + "loss": 0.3214, + "step": 41387 + }, + { + "epoch": 0.9110279550957517, + "grad_norm": 1.5614268414817232, + "learning_rate": 2.0615703201174342e-07, + "loss": 0.281, + "step": 41388 + }, + { + "epoch": 0.9110499669821703, + "grad_norm": 0.6257411106105811, + "learning_rate": 2.0605574359073398e-07, + "loss": 0.4302, + "step": 41389 + }, + { + "epoch": 0.911071978868589, + "grad_norm": 2.2425247303003752, + "learning_rate": 2.059544795346874e-07, + "loss": 0.2391, + "step": 41390 + }, + { + "epoch": 0.9110939907550077, + "grad_norm": 1.2220868596293555, + "learning_rate": 2.0585323984411765e-07, + "loss": 0.2339, + "step": 41391 + }, + { + "epoch": 0.9111160026414263, + "grad_norm": 1.5106399125753682, + "learning_rate": 2.057520245195399e-07, + "loss": 0.3304, + "step": 41392 + }, + { + "epoch": 0.911138014527845, + "grad_norm": 0.5937499269877223, + "learning_rate": 2.0565083356146876e-07, + "loss": 0.4163, + "step": 41393 + }, + { + "epoch": 0.9111600264142637, + "grad_norm": 2.1439050675355027, + "learning_rate": 2.0554966697041877e-07, + "loss": 0.3817, + "step": 41394 + }, + { + "epoch": 0.9111820383006823, + "grad_norm": 1.7014071584700203, + "learning_rate": 2.0544852474690401e-07, + "loss": 0.343, + "step": 41395 + }, + { + "epoch": 0.911204050187101, + "grad_norm": 0.6205807657598537, + "learning_rate": 2.0534740689143794e-07, + "loss": 0.4367, + "step": 41396 + }, + { + "epoch": 0.9112260620735197, + "grad_norm": 1.5758350390811555, + "learning_rate": 2.052463134045346e-07, + "loss": 0.2612, + "step": 41397 + }, + { + "epoch": 0.9112480739599383, + "grad_norm": 2.265289900411978, + "learning_rate": 2.0514524428670801e-07, + "loss": 0.2796, + "step": 41398 + }, + { + "epoch": 0.911270085846357, + "grad_norm": 3.119470493289523, + "learning_rate": 2.050441995384722e-07, + "loss": 0.3097, + "step": 41399 + }, + { + "epoch": 0.9112920977327758, + "grad_norm": 1.8969343452894312, + "learning_rate": 2.049431791603407e-07, + "loss": 0.3284, + "step": 41400 + }, + { + "epoch": 0.9113141096191943, + "grad_norm": 1.4109074546530258, + "learning_rate": 2.048421831528258e-07, + "loss": 0.2424, + "step": 41401 + }, + { + "epoch": 0.911336121505613, + "grad_norm": 1.8080414552282702, + "learning_rate": 2.047412115164421e-07, + "loss": 0.2788, + "step": 41402 + }, + { + "epoch": 0.9113581333920316, + "grad_norm": 2.3135423430339572, + "learning_rate": 2.0464026425170202e-07, + "loss": 0.2263, + "step": 41403 + }, + { + "epoch": 0.9113801452784503, + "grad_norm": 3.8554792873803407, + "learning_rate": 2.0453934135911902e-07, + "loss": 0.3425, + "step": 41404 + }, + { + "epoch": 0.911402157164869, + "grad_norm": 1.3224727464955286, + "learning_rate": 2.04438442839206e-07, + "loss": 0.2834, + "step": 41405 + }, + { + "epoch": 0.9114241690512876, + "grad_norm": 0.6229381197221285, + "learning_rate": 2.0433756869247533e-07, + "loss": 0.4274, + "step": 41406 + }, + { + "epoch": 0.9114461809377064, + "grad_norm": 1.3824012323690942, + "learning_rate": 2.0423671891944107e-07, + "loss": 0.2586, + "step": 41407 + }, + { + "epoch": 0.9114681928241251, + "grad_norm": 1.51074791090724, + "learning_rate": 2.041358935206139e-07, + "loss": 0.3072, + "step": 41408 + }, + { + "epoch": 0.9114902047105437, + "grad_norm": 1.7716629585619563, + "learning_rate": 2.0403509249650676e-07, + "loss": 0.4002, + "step": 41409 + }, + { + "epoch": 0.9115122165969624, + "grad_norm": 1.66865762680033, + "learning_rate": 2.039343158476337e-07, + "loss": 0.2406, + "step": 41410 + }, + { + "epoch": 0.9115342284833811, + "grad_norm": 1.403390207289142, + "learning_rate": 2.0383356357450423e-07, + "loss": 0.2299, + "step": 41411 + }, + { + "epoch": 0.9115562403697997, + "grad_norm": 1.3640900313303916, + "learning_rate": 2.0373283567763247e-07, + "loss": 0.2079, + "step": 41412 + }, + { + "epoch": 0.9115782522562184, + "grad_norm": 1.742122485560349, + "learning_rate": 2.0363213215752963e-07, + "loss": 0.2148, + "step": 41413 + }, + { + "epoch": 0.9116002641426371, + "grad_norm": 1.4796340739388145, + "learning_rate": 2.03531453014707e-07, + "loss": 0.2995, + "step": 41414 + }, + { + "epoch": 0.9116222760290557, + "grad_norm": 2.183908179192689, + "learning_rate": 2.034307982496775e-07, + "loss": 0.2849, + "step": 41415 + }, + { + "epoch": 0.9116442879154744, + "grad_norm": 2.6442966656605242, + "learning_rate": 2.0333016786295124e-07, + "loss": 0.2923, + "step": 41416 + }, + { + "epoch": 0.9116662998018931, + "grad_norm": 0.6330918756471893, + "learning_rate": 2.0322956185504116e-07, + "loss": 0.4775, + "step": 41417 + }, + { + "epoch": 0.9116883116883117, + "grad_norm": 1.7433106608846625, + "learning_rate": 2.0312898022645743e-07, + "loss": 0.2923, + "step": 41418 + }, + { + "epoch": 0.9117103235747304, + "grad_norm": 1.6812546440344214, + "learning_rate": 2.0302842297771132e-07, + "loss": 0.1873, + "step": 41419 + }, + { + "epoch": 0.9117323354611491, + "grad_norm": 1.4200531177711255, + "learning_rate": 2.0292789010931513e-07, + "loss": 0.2212, + "step": 41420 + }, + { + "epoch": 0.9117543473475677, + "grad_norm": 0.6244223554281394, + "learning_rate": 2.028273816217785e-07, + "loss": 0.4546, + "step": 41421 + }, + { + "epoch": 0.9117763592339864, + "grad_norm": 0.6216433859863407, + "learning_rate": 2.0272689751561326e-07, + "loss": 0.4509, + "step": 41422 + }, + { + "epoch": 0.9117983711204051, + "grad_norm": 1.903574280124234, + "learning_rate": 2.0262643779132896e-07, + "loss": 0.3956, + "step": 41423 + }, + { + "epoch": 0.9118203830068237, + "grad_norm": 2.9958385670086938, + "learning_rate": 2.0252600244943688e-07, + "loss": 0.2909, + "step": 41424 + }, + { + "epoch": 0.9118423948932424, + "grad_norm": 1.3775806286168608, + "learning_rate": 2.024255914904477e-07, + "loss": 0.3147, + "step": 41425 + }, + { + "epoch": 0.911864406779661, + "grad_norm": 2.2153024681670086, + "learning_rate": 2.023252049148716e-07, + "loss": 0.3355, + "step": 41426 + }, + { + "epoch": 0.9118864186660797, + "grad_norm": 1.801775074409803, + "learning_rate": 2.0222484272321762e-07, + "loss": 0.4003, + "step": 41427 + }, + { + "epoch": 0.9119084305524984, + "grad_norm": 1.517418876494367, + "learning_rate": 2.0212450491599754e-07, + "loss": 0.2573, + "step": 41428 + }, + { + "epoch": 0.911930442438917, + "grad_norm": 0.6265553165538444, + "learning_rate": 2.0202419149372043e-07, + "loss": 0.4673, + "step": 41429 + }, + { + "epoch": 0.9119524543253357, + "grad_norm": 0.6292330383011362, + "learning_rate": 2.0192390245689642e-07, + "loss": 0.452, + "step": 41430 + }, + { + "epoch": 0.9119744662117544, + "grad_norm": 2.395539486887646, + "learning_rate": 2.0182363780603565e-07, + "loss": 0.315, + "step": 41431 + }, + { + "epoch": 0.911996478098173, + "grad_norm": 1.6670238350056772, + "learning_rate": 2.017233975416466e-07, + "loss": 0.3008, + "step": 41432 + }, + { + "epoch": 0.9120184899845917, + "grad_norm": 3.2041041706109676, + "learning_rate": 2.0162318166423945e-07, + "loss": 0.3123, + "step": 41433 + }, + { + "epoch": 0.9120405018710104, + "grad_norm": 1.5992095691755068, + "learning_rate": 2.015229901743232e-07, + "loss": 0.4105, + "step": 41434 + }, + { + "epoch": 0.912062513757429, + "grad_norm": 2.658326629778508, + "learning_rate": 2.01422823072408e-07, + "loss": 0.2946, + "step": 41435 + }, + { + "epoch": 0.9120845256438477, + "grad_norm": 2.487430279997675, + "learning_rate": 2.0132268035900182e-07, + "loss": 0.2456, + "step": 41436 + }, + { + "epoch": 0.9121065375302664, + "grad_norm": 1.7205522890730123, + "learning_rate": 2.0122256203461366e-07, + "loss": 0.2547, + "step": 41437 + }, + { + "epoch": 0.912128549416685, + "grad_norm": 9.630169815871836, + "learning_rate": 2.0112246809975366e-07, + "loss": 0.3143, + "step": 41438 + }, + { + "epoch": 0.9121505613031037, + "grad_norm": 1.6765803792698462, + "learning_rate": 2.010223985549281e-07, + "loss": 0.2598, + "step": 41439 + }, + { + "epoch": 0.9121725731895224, + "grad_norm": 2.4660428018762928, + "learning_rate": 2.009223534006488e-07, + "loss": 0.291, + "step": 41440 + }, + { + "epoch": 0.912194585075941, + "grad_norm": 1.6392471084907971, + "learning_rate": 2.00822332637422e-07, + "loss": 0.3928, + "step": 41441 + }, + { + "epoch": 0.9122165969623597, + "grad_norm": 1.9066035264433185, + "learning_rate": 2.007223362657562e-07, + "loss": 0.3205, + "step": 41442 + }, + { + "epoch": 0.9122386088487784, + "grad_norm": 1.5962792680409283, + "learning_rate": 2.0062236428616045e-07, + "loss": 0.3121, + "step": 41443 + }, + { + "epoch": 0.912260620735197, + "grad_norm": 1.6006581146832637, + "learning_rate": 2.0052241669914208e-07, + "loss": 0.2901, + "step": 41444 + }, + { + "epoch": 0.9122826326216157, + "grad_norm": 2.0384413305605444, + "learning_rate": 2.0042249350520958e-07, + "loss": 0.3113, + "step": 41445 + }, + { + "epoch": 0.9123046445080344, + "grad_norm": 1.8069020326059668, + "learning_rate": 2.0032259470487093e-07, + "loss": 0.3094, + "step": 41446 + }, + { + "epoch": 0.912326656394453, + "grad_norm": 0.6156127676527458, + "learning_rate": 2.0022272029863287e-07, + "loss": 0.4447, + "step": 41447 + }, + { + "epoch": 0.9123486682808717, + "grad_norm": 3.2516598050072387, + "learning_rate": 2.001228702870045e-07, + "loss": 0.2849, + "step": 41448 + }, + { + "epoch": 0.9123706801672903, + "grad_norm": 0.628251005707567, + "learning_rate": 2.0002304467049206e-07, + "loss": 0.4396, + "step": 41449 + }, + { + "epoch": 0.912392692053709, + "grad_norm": 1.1649923919231397, + "learning_rate": 1.999232434496029e-07, + "loss": 0.2602, + "step": 41450 + }, + { + "epoch": 0.9124147039401277, + "grad_norm": 2.32027044528877, + "learning_rate": 1.998234666248461e-07, + "loss": 0.2956, + "step": 41451 + }, + { + "epoch": 0.9124367158265463, + "grad_norm": 1.641808278928906, + "learning_rate": 1.997237141967262e-07, + "loss": 0.3094, + "step": 41452 + }, + { + "epoch": 0.912458727712965, + "grad_norm": 1.445838770632063, + "learning_rate": 1.9962398616575174e-07, + "loss": 0.3451, + "step": 41453 + }, + { + "epoch": 0.9124807395993837, + "grad_norm": 1.9908752379611716, + "learning_rate": 1.9952428253242895e-07, + "loss": 0.3016, + "step": 41454 + }, + { + "epoch": 0.9125027514858023, + "grad_norm": 1.864315178965512, + "learning_rate": 1.9942460329726464e-07, + "loss": 0.2687, + "step": 41455 + }, + { + "epoch": 0.912524763372221, + "grad_norm": 2.6498635938318156, + "learning_rate": 1.9932494846076678e-07, + "loss": 0.2727, + "step": 41456 + }, + { + "epoch": 0.9125467752586397, + "grad_norm": 1.965771521385648, + "learning_rate": 1.9922531802343936e-07, + "loss": 0.2992, + "step": 41457 + }, + { + "epoch": 0.9125687871450583, + "grad_norm": 0.6033591944784815, + "learning_rate": 1.9912571198579088e-07, + "loss": 0.4303, + "step": 41458 + }, + { + "epoch": 0.912590799031477, + "grad_norm": 1.6267489846024752, + "learning_rate": 1.9902613034832652e-07, + "loss": 0.2793, + "step": 41459 + }, + { + "epoch": 0.9126128109178957, + "grad_norm": 1.3577433640473495, + "learning_rate": 1.989265731115525e-07, + "loss": 0.2801, + "step": 41460 + }, + { + "epoch": 0.9126348228043143, + "grad_norm": 5.9882893901306185, + "learning_rate": 1.9882704027597567e-07, + "loss": 0.3273, + "step": 41461 + }, + { + "epoch": 0.912656834690733, + "grad_norm": 1.8856647935031319, + "learning_rate": 1.9872753184210115e-07, + "loss": 0.3454, + "step": 41462 + }, + { + "epoch": 0.9126788465771517, + "grad_norm": 1.7848309931699922, + "learning_rate": 1.9862804781043354e-07, + "loss": 0.2547, + "step": 41463 + }, + { + "epoch": 0.9127008584635703, + "grad_norm": 1.714677718095665, + "learning_rate": 1.9852858818148025e-07, + "loss": 0.2774, + "step": 41464 + }, + { + "epoch": 0.912722870349989, + "grad_norm": 1.9810349887890542, + "learning_rate": 1.9842915295574638e-07, + "loss": 0.3801, + "step": 41465 + }, + { + "epoch": 0.9127448822364077, + "grad_norm": 1.6418979219940575, + "learning_rate": 1.9832974213373713e-07, + "loss": 0.3112, + "step": 41466 + }, + { + "epoch": 0.9127668941228263, + "grad_norm": 2.583141971003323, + "learning_rate": 1.9823035571595816e-07, + "loss": 0.3852, + "step": 41467 + }, + { + "epoch": 0.912788906009245, + "grad_norm": 0.6155825645773152, + "learning_rate": 1.9813099370291356e-07, + "loss": 0.4174, + "step": 41468 + }, + { + "epoch": 0.9128109178956637, + "grad_norm": 0.6169998607094697, + "learning_rate": 1.98031656095109e-07, + "loss": 0.4386, + "step": 41469 + }, + { + "epoch": 0.9128329297820823, + "grad_norm": 5.127661993789567, + "learning_rate": 1.9793234289304907e-07, + "loss": 0.2774, + "step": 41470 + }, + { + "epoch": 0.912854941668501, + "grad_norm": 0.6249231916428715, + "learning_rate": 1.9783305409723952e-07, + "loss": 0.4514, + "step": 41471 + }, + { + "epoch": 0.9128769535549196, + "grad_norm": 2.1495608501959196, + "learning_rate": 1.977337897081838e-07, + "loss": 0.3387, + "step": 41472 + }, + { + "epoch": 0.9128989654413383, + "grad_norm": 1.5633574121054097, + "learning_rate": 1.976345497263865e-07, + "loss": 0.3354, + "step": 41473 + }, + { + "epoch": 0.912920977327757, + "grad_norm": 2.5022480459350067, + "learning_rate": 1.9753533415235338e-07, + "loss": 0.288, + "step": 41474 + }, + { + "epoch": 0.9129429892141756, + "grad_norm": 1.5163181250478974, + "learning_rate": 1.9743614298658566e-07, + "loss": 0.2479, + "step": 41475 + }, + { + "epoch": 0.9129650011005943, + "grad_norm": 1.709009805352896, + "learning_rate": 1.9733697622959125e-07, + "loss": 0.263, + "step": 41476 + }, + { + "epoch": 0.912987012987013, + "grad_norm": 0.6238138546712939, + "learning_rate": 1.97237833881872e-07, + "loss": 0.43, + "step": 41477 + }, + { + "epoch": 0.9130090248734316, + "grad_norm": 1.2636156439401192, + "learning_rate": 1.9713871594393196e-07, + "loss": 0.2916, + "step": 41478 + }, + { + "epoch": 0.9130310367598503, + "grad_norm": 1.6466246487278169, + "learning_rate": 1.9703962241627516e-07, + "loss": 0.2918, + "step": 41479 + }, + { + "epoch": 0.913053048646269, + "grad_norm": 1.7434237632501415, + "learning_rate": 1.9694055329940453e-07, + "loss": 0.2659, + "step": 41480 + }, + { + "epoch": 0.9130750605326876, + "grad_norm": 1.4667776719944086, + "learning_rate": 1.9684150859382523e-07, + "loss": 0.2275, + "step": 41481 + }, + { + "epoch": 0.9130970724191063, + "grad_norm": 1.6714957565292168, + "learning_rate": 1.967424883000396e-07, + "loss": 0.2946, + "step": 41482 + }, + { + "epoch": 0.913119084305525, + "grad_norm": 2.1413871969565466, + "learning_rate": 1.9664349241855062e-07, + "loss": 0.3276, + "step": 41483 + }, + { + "epoch": 0.9131410961919436, + "grad_norm": 0.6087129116280898, + "learning_rate": 1.9654452094986176e-07, + "loss": 0.4342, + "step": 41484 + }, + { + "epoch": 0.9131631080783623, + "grad_norm": 1.7524098519448552, + "learning_rate": 1.9644557389447595e-07, + "loss": 0.2079, + "step": 41485 + }, + { + "epoch": 0.913185119964781, + "grad_norm": 2.01879742248964, + "learning_rate": 1.963466512528961e-07, + "loss": 0.3188, + "step": 41486 + }, + { + "epoch": 0.9132071318511996, + "grad_norm": 2.294443720711687, + "learning_rate": 1.9624775302562571e-07, + "loss": 0.3842, + "step": 41487 + }, + { + "epoch": 0.9132291437376183, + "grad_norm": 0.6231467882973436, + "learning_rate": 1.9614887921316606e-07, + "loss": 0.455, + "step": 41488 + }, + { + "epoch": 0.913251155624037, + "grad_norm": 1.7782159225586491, + "learning_rate": 1.9605002981602118e-07, + "loss": 0.3547, + "step": 41489 + }, + { + "epoch": 0.9132731675104556, + "grad_norm": 2.0692819264296816, + "learning_rate": 1.9595120483469176e-07, + "loss": 0.2675, + "step": 41490 + }, + { + "epoch": 0.9132951793968743, + "grad_norm": 1.5818315213404865, + "learning_rate": 1.9585240426968077e-07, + "loss": 0.2683, + "step": 41491 + }, + { + "epoch": 0.913317191283293, + "grad_norm": 1.9696178836904916, + "learning_rate": 1.9575362812149168e-07, + "loss": 0.2831, + "step": 41492 + }, + { + "epoch": 0.9133392031697116, + "grad_norm": 2.851633154293899, + "learning_rate": 1.9565487639062464e-07, + "loss": 0.3297, + "step": 41493 + }, + { + "epoch": 0.9133612150561303, + "grad_norm": 1.7510515881945319, + "learning_rate": 1.9555614907758313e-07, + "loss": 0.3472, + "step": 41494 + }, + { + "epoch": 0.913383226942549, + "grad_norm": 1.4839079207461796, + "learning_rate": 1.9545744618286732e-07, + "loss": 0.3357, + "step": 41495 + }, + { + "epoch": 0.9134052388289676, + "grad_norm": 2.0854460000869706, + "learning_rate": 1.953587677069796e-07, + "loss": 0.3062, + "step": 41496 + }, + { + "epoch": 0.9134272507153863, + "grad_norm": 1.3775975718663682, + "learning_rate": 1.952601136504223e-07, + "loss": 0.2521, + "step": 41497 + }, + { + "epoch": 0.9134492626018049, + "grad_norm": 2.75999238908692, + "learning_rate": 1.9516148401369506e-07, + "loss": 0.2989, + "step": 41498 + }, + { + "epoch": 0.9134712744882236, + "grad_norm": 2.7299729560571353, + "learning_rate": 1.9506287879730135e-07, + "loss": 0.2991, + "step": 41499 + }, + { + "epoch": 0.9134932863746423, + "grad_norm": 1.2888028063775103, + "learning_rate": 1.9496429800174022e-07, + "loss": 0.2238, + "step": 41500 + }, + { + "epoch": 0.9135152982610609, + "grad_norm": 1.7243274381354812, + "learning_rate": 1.9486574162751349e-07, + "loss": 0.3812, + "step": 41501 + }, + { + "epoch": 0.9135373101474796, + "grad_norm": 1.6611368570487128, + "learning_rate": 1.9476720967512298e-07, + "loss": 0.2267, + "step": 41502 + }, + { + "epoch": 0.9135593220338983, + "grad_norm": 1.7089439795632353, + "learning_rate": 1.9466870214506883e-07, + "loss": 0.2742, + "step": 41503 + }, + { + "epoch": 0.9135813339203169, + "grad_norm": 2.477357898872799, + "learning_rate": 1.9457021903785067e-07, + "loss": 0.3262, + "step": 41504 + }, + { + "epoch": 0.9136033458067356, + "grad_norm": 1.6318807491668588, + "learning_rate": 1.944717603539703e-07, + "loss": 0.3538, + "step": 41505 + }, + { + "epoch": 0.9136253576931543, + "grad_norm": 3.942985381674342, + "learning_rate": 1.943733260939279e-07, + "loss": 0.2231, + "step": 41506 + }, + { + "epoch": 0.9136473695795729, + "grad_norm": 1.929865118285286, + "learning_rate": 1.9427491625822358e-07, + "loss": 0.3183, + "step": 41507 + }, + { + "epoch": 0.9136693814659916, + "grad_norm": 0.6344026627591104, + "learning_rate": 1.941765308473581e-07, + "loss": 0.4825, + "step": 41508 + }, + { + "epoch": 0.9136913933524103, + "grad_norm": 1.8370757166100367, + "learning_rate": 1.9407816986182993e-07, + "loss": 0.2869, + "step": 41509 + }, + { + "epoch": 0.9137134052388289, + "grad_norm": 1.857889409357036, + "learning_rate": 1.9397983330214088e-07, + "loss": 0.2906, + "step": 41510 + }, + { + "epoch": 0.9137354171252476, + "grad_norm": 0.6120603401220298, + "learning_rate": 1.938815211687889e-07, + "loss": 0.437, + "step": 41511 + }, + { + "epoch": 0.9137574290116663, + "grad_norm": 0.6406458288538467, + "learning_rate": 1.9378323346227523e-07, + "loss": 0.4614, + "step": 41512 + }, + { + "epoch": 0.9137794408980849, + "grad_norm": 0.6169497774513051, + "learning_rate": 1.936849701830995e-07, + "loss": 0.4398, + "step": 41513 + }, + { + "epoch": 0.9138014527845036, + "grad_norm": 1.7264459244175248, + "learning_rate": 1.935867313317591e-07, + "loss": 0.2121, + "step": 41514 + }, + { + "epoch": 0.9138234646709223, + "grad_norm": 3.5118142117780966, + "learning_rate": 1.934885169087558e-07, + "loss": 0.3239, + "step": 41515 + }, + { + "epoch": 0.9138454765573409, + "grad_norm": 1.918843624650579, + "learning_rate": 1.9339032691458648e-07, + "loss": 0.253, + "step": 41516 + }, + { + "epoch": 0.9138674884437596, + "grad_norm": 1.4148666788978725, + "learning_rate": 1.9329216134975182e-07, + "loss": 0.2792, + "step": 41517 + }, + { + "epoch": 0.9138895003301784, + "grad_norm": 1.5175335271129624, + "learning_rate": 1.9319402021475085e-07, + "loss": 0.2178, + "step": 41518 + }, + { + "epoch": 0.9139115122165969, + "grad_norm": 0.6313081778359964, + "learning_rate": 1.93095903510081e-07, + "loss": 0.4461, + "step": 41519 + }, + { + "epoch": 0.9139335241030156, + "grad_norm": 1.8328624261758726, + "learning_rate": 1.9299781123624238e-07, + "loss": 0.3128, + "step": 41520 + }, + { + "epoch": 0.9139555359894342, + "grad_norm": 1.5011165247559322, + "learning_rate": 1.9289974339373186e-07, + "loss": 0.3134, + "step": 41521 + }, + { + "epoch": 0.913977547875853, + "grad_norm": 2.3857978798272397, + "learning_rate": 1.9280169998304955e-07, + "loss": 0.2678, + "step": 41522 + }, + { + "epoch": 0.9139995597622717, + "grad_norm": 2.5270832376954155, + "learning_rate": 1.9270368100469284e-07, + "loss": 0.3926, + "step": 41523 + }, + { + "epoch": 0.9140215716486902, + "grad_norm": 2.0180788640387313, + "learning_rate": 1.9260568645916023e-07, + "loss": 0.3137, + "step": 41524 + }, + { + "epoch": 0.914043583535109, + "grad_norm": 1.678148392926276, + "learning_rate": 1.925077163469502e-07, + "loss": 0.2721, + "step": 41525 + }, + { + "epoch": 0.9140655954215277, + "grad_norm": 1.9032716574819044, + "learning_rate": 1.9240977066855903e-07, + "loss": 0.3136, + "step": 41526 + }, + { + "epoch": 0.9140876073079462, + "grad_norm": 1.795981420804235, + "learning_rate": 1.9231184942448634e-07, + "loss": 0.2353, + "step": 41527 + }, + { + "epoch": 0.914109619194365, + "grad_norm": 2.280736822244951, + "learning_rate": 1.9221395261522945e-07, + "loss": 0.3276, + "step": 41528 + }, + { + "epoch": 0.9141316310807837, + "grad_norm": 1.4994953081325832, + "learning_rate": 1.921160802412847e-07, + "loss": 0.3155, + "step": 41529 + }, + { + "epoch": 0.9141536429672023, + "grad_norm": 1.4800015783407734, + "learning_rate": 1.9201823230315165e-07, + "loss": 0.2243, + "step": 41530 + }, + { + "epoch": 0.914175654853621, + "grad_norm": 1.858557130824644, + "learning_rate": 1.9192040880132545e-07, + "loss": 0.188, + "step": 41531 + }, + { + "epoch": 0.9141976667400397, + "grad_norm": 4.522593033007907, + "learning_rate": 1.9182260973630406e-07, + "loss": 0.35, + "step": 41532 + }, + { + "epoch": 0.9142196786264583, + "grad_norm": 1.8629215593236268, + "learning_rate": 1.917248351085854e-07, + "loss": 0.2666, + "step": 41533 + }, + { + "epoch": 0.914241690512877, + "grad_norm": 1.614528574741515, + "learning_rate": 1.9162708491866466e-07, + "loss": 0.3038, + "step": 41534 + }, + { + "epoch": 0.9142637023992957, + "grad_norm": 3.611197468077171, + "learning_rate": 1.9152935916704085e-07, + "loss": 0.2916, + "step": 41535 + }, + { + "epoch": 0.9142857142857143, + "grad_norm": 2.357576307040246, + "learning_rate": 1.9143165785420858e-07, + "loss": 0.3248, + "step": 41536 + }, + { + "epoch": 0.914307726172133, + "grad_norm": 3.2440146672016352, + "learning_rate": 1.9133398098066525e-07, + "loss": 0.2885, + "step": 41537 + }, + { + "epoch": 0.9143297380585517, + "grad_norm": 1.998368971768314, + "learning_rate": 1.9123632854690821e-07, + "loss": 0.334, + "step": 41538 + }, + { + "epoch": 0.9143517499449703, + "grad_norm": 1.7480068781008862, + "learning_rate": 1.9113870055343208e-07, + "loss": 0.2387, + "step": 41539 + }, + { + "epoch": 0.914373761831389, + "grad_norm": 1.3587267350653651, + "learning_rate": 1.910410970007348e-07, + "loss": 0.3312, + "step": 41540 + }, + { + "epoch": 0.9143957737178077, + "grad_norm": 1.6551683887128632, + "learning_rate": 1.9094351788931041e-07, + "loss": 0.3012, + "step": 41541 + }, + { + "epoch": 0.9144177856042263, + "grad_norm": 0.6126441044228651, + "learning_rate": 1.9084596321965633e-07, + "loss": 0.4462, + "step": 41542 + }, + { + "epoch": 0.914439797490645, + "grad_norm": 1.4892348414475165, + "learning_rate": 1.9074843299226875e-07, + "loss": 0.2496, + "step": 41543 + }, + { + "epoch": 0.9144618093770636, + "grad_norm": 0.6387971141582196, + "learning_rate": 1.9065092720764232e-07, + "loss": 0.427, + "step": 41544 + }, + { + "epoch": 0.9144838212634823, + "grad_norm": 0.6056805461760377, + "learning_rate": 1.9055344586627221e-07, + "loss": 0.4279, + "step": 41545 + }, + { + "epoch": 0.914505833149901, + "grad_norm": 1.6630920492363925, + "learning_rate": 1.9045598896865524e-07, + "loss": 0.2167, + "step": 41546 + }, + { + "epoch": 0.9145278450363196, + "grad_norm": 1.6761933787502137, + "learning_rate": 1.9035855651528545e-07, + "loss": 0.3509, + "step": 41547 + }, + { + "epoch": 0.9145498569227383, + "grad_norm": 1.615591009151577, + "learning_rate": 1.902611485066591e-07, + "loss": 0.2139, + "step": 41548 + }, + { + "epoch": 0.914571868809157, + "grad_norm": 1.5096895432823194, + "learning_rate": 1.9016376494327082e-07, + "loss": 0.272, + "step": 41549 + }, + { + "epoch": 0.9145938806955756, + "grad_norm": 2.286143357918085, + "learning_rate": 1.900664058256152e-07, + "loss": 0.2346, + "step": 41550 + }, + { + "epoch": 0.9146158925819943, + "grad_norm": 0.6090906507352251, + "learning_rate": 1.8996907115418794e-07, + "loss": 0.4379, + "step": 41551 + }, + { + "epoch": 0.914637904468413, + "grad_norm": 1.5446544640459283, + "learning_rate": 1.8987176092948255e-07, + "loss": 0.2543, + "step": 41552 + }, + { + "epoch": 0.9146599163548316, + "grad_norm": 1.6061593126834834, + "learning_rate": 1.8977447515199476e-07, + "loss": 0.3808, + "step": 41553 + }, + { + "epoch": 0.9146819282412503, + "grad_norm": 2.4863055798564804, + "learning_rate": 1.896772138222186e-07, + "loss": 0.2494, + "step": 41554 + }, + { + "epoch": 0.914703940127669, + "grad_norm": 1.5760352742802999, + "learning_rate": 1.8957997694064756e-07, + "loss": 0.2535, + "step": 41555 + }, + { + "epoch": 0.9147259520140876, + "grad_norm": 1.6856834584848976, + "learning_rate": 1.8948276450777737e-07, + "loss": 0.284, + "step": 41556 + }, + { + "epoch": 0.9147479639005063, + "grad_norm": 1.6896507799372966, + "learning_rate": 1.8938557652410094e-07, + "loss": 0.3248, + "step": 41557 + }, + { + "epoch": 0.914769975786925, + "grad_norm": 1.8986744568939617, + "learning_rate": 1.8928841299011236e-07, + "loss": 0.3651, + "step": 41558 + }, + { + "epoch": 0.9147919876733436, + "grad_norm": 1.6801220833249373, + "learning_rate": 1.891912739063062e-07, + "loss": 0.2981, + "step": 41559 + }, + { + "epoch": 0.9148139995597623, + "grad_norm": 1.8866317227753775, + "learning_rate": 1.8909415927317544e-07, + "loss": 0.259, + "step": 41560 + }, + { + "epoch": 0.914836011446181, + "grad_norm": 2.376066051545813, + "learning_rate": 1.8899706909121406e-07, + "loss": 0.2823, + "step": 41561 + }, + { + "epoch": 0.9148580233325996, + "grad_norm": 2.073361108574526, + "learning_rate": 1.8890000336091506e-07, + "loss": 0.2498, + "step": 41562 + }, + { + "epoch": 0.9148800352190183, + "grad_norm": 2.972924405752068, + "learning_rate": 1.8880296208277192e-07, + "loss": 0.3679, + "step": 41563 + }, + { + "epoch": 0.914902047105437, + "grad_norm": 1.4135361601228738, + "learning_rate": 1.8870594525727814e-07, + "loss": 0.2293, + "step": 41564 + }, + { + "epoch": 0.9149240589918556, + "grad_norm": 0.6023596716860239, + "learning_rate": 1.8860895288492664e-07, + "loss": 0.4444, + "step": 41565 + }, + { + "epoch": 0.9149460708782743, + "grad_norm": 2.6610288380273377, + "learning_rate": 1.8851198496621092e-07, + "loss": 0.2932, + "step": 41566 + }, + { + "epoch": 0.9149680827646929, + "grad_norm": 1.5665711856570654, + "learning_rate": 1.884150415016228e-07, + "loss": 0.2595, + "step": 41567 + }, + { + "epoch": 0.9149900946511116, + "grad_norm": 2.4286630826797064, + "learning_rate": 1.8831812249165527e-07, + "loss": 0.3194, + "step": 41568 + }, + { + "epoch": 0.9150121065375303, + "grad_norm": 0.6364043487019183, + "learning_rate": 1.8822122793680174e-07, + "loss": 0.4397, + "step": 41569 + }, + { + "epoch": 0.9150341184239489, + "grad_norm": 1.4316575268872291, + "learning_rate": 1.88124357837553e-07, + "loss": 0.2924, + "step": 41570 + }, + { + "epoch": 0.9150561303103676, + "grad_norm": 1.7308572151534014, + "learning_rate": 1.8802751219440361e-07, + "loss": 0.3173, + "step": 41571 + }, + { + "epoch": 0.9150781421967863, + "grad_norm": 1.4708796938612612, + "learning_rate": 1.8793069100784377e-07, + "loss": 0.3071, + "step": 41572 + }, + { + "epoch": 0.9151001540832049, + "grad_norm": 0.6276602080602518, + "learning_rate": 1.878338942783664e-07, + "loss": 0.4342, + "step": 41573 + }, + { + "epoch": 0.9151221659696236, + "grad_norm": 1.7296059242291872, + "learning_rate": 1.8773712200646442e-07, + "loss": 0.3529, + "step": 41574 + }, + { + "epoch": 0.9151441778560423, + "grad_norm": 1.688413168381264, + "learning_rate": 1.8764037419262748e-07, + "loss": 0.2716, + "step": 41575 + }, + { + "epoch": 0.9151661897424609, + "grad_norm": 1.5094602362182172, + "learning_rate": 1.875436508373496e-07, + "loss": 0.2276, + "step": 41576 + }, + { + "epoch": 0.9151882016288796, + "grad_norm": 1.7162147563004373, + "learning_rate": 1.874469519411204e-07, + "loss": 0.2141, + "step": 41577 + }, + { + "epoch": 0.9152102135152983, + "grad_norm": 3.1267124469285204, + "learning_rate": 1.873502775044328e-07, + "loss": 0.2474, + "step": 41578 + }, + { + "epoch": 0.9152322254017169, + "grad_norm": 1.9396303123797154, + "learning_rate": 1.8725362752777753e-07, + "loss": 0.342, + "step": 41579 + }, + { + "epoch": 0.9152542372881356, + "grad_norm": 1.934545028757769, + "learning_rate": 1.8715700201164532e-07, + "loss": 0.2785, + "step": 41580 + }, + { + "epoch": 0.9152762491745543, + "grad_norm": 1.2398480183920706, + "learning_rate": 1.870604009565291e-07, + "loss": 0.1882, + "step": 41581 + }, + { + "epoch": 0.9152982610609729, + "grad_norm": 2.4491038245565107, + "learning_rate": 1.8696382436291794e-07, + "loss": 0.3659, + "step": 41582 + }, + { + "epoch": 0.9153202729473916, + "grad_norm": 2.2750930956510333, + "learning_rate": 1.8686727223130197e-07, + "loss": 0.2808, + "step": 41583 + }, + { + "epoch": 0.9153422848338103, + "grad_norm": 1.6268812246701119, + "learning_rate": 1.867707445621747e-07, + "loss": 0.2071, + "step": 41584 + }, + { + "epoch": 0.9153642967202289, + "grad_norm": 1.9917627760871854, + "learning_rate": 1.8667424135602573e-07, + "loss": 0.2852, + "step": 41585 + }, + { + "epoch": 0.9153863086066476, + "grad_norm": 2.146917849400901, + "learning_rate": 1.865777626133436e-07, + "loss": 0.2459, + "step": 41586 + }, + { + "epoch": 0.9154083204930663, + "grad_norm": 2.0410125355181914, + "learning_rate": 1.8648130833462118e-07, + "loss": 0.3716, + "step": 41587 + }, + { + "epoch": 0.9154303323794849, + "grad_norm": 1.585755457984842, + "learning_rate": 1.8638487852034702e-07, + "loss": 0.2415, + "step": 41588 + }, + { + "epoch": 0.9154523442659036, + "grad_norm": 1.2572418379059416, + "learning_rate": 1.8628847317101185e-07, + "loss": 0.2026, + "step": 41589 + }, + { + "epoch": 0.9154743561523222, + "grad_norm": 1.618195168436592, + "learning_rate": 1.8619209228710634e-07, + "loss": 0.3, + "step": 41590 + }, + { + "epoch": 0.9154963680387409, + "grad_norm": 1.6636807999927818, + "learning_rate": 1.86095735869119e-07, + "loss": 0.2659, + "step": 41591 + }, + { + "epoch": 0.9155183799251596, + "grad_norm": 2.008928061518948, + "learning_rate": 1.8599940391754055e-07, + "loss": 0.2399, + "step": 41592 + }, + { + "epoch": 0.9155403918115782, + "grad_norm": 0.619235711837742, + "learning_rate": 1.859030964328601e-07, + "loss": 0.4502, + "step": 41593 + }, + { + "epoch": 0.9155624036979969, + "grad_norm": 1.473756800861123, + "learning_rate": 1.8580681341556662e-07, + "loss": 0.164, + "step": 41594 + }, + { + "epoch": 0.9155844155844156, + "grad_norm": 1.769118799633616, + "learning_rate": 1.8571055486615142e-07, + "loss": 0.3185, + "step": 41595 + }, + { + "epoch": 0.9156064274708342, + "grad_norm": 10.734567276227738, + "learning_rate": 1.8561432078510132e-07, + "loss": 0.2507, + "step": 41596 + }, + { + "epoch": 0.9156284393572529, + "grad_norm": 2.1340452533569825, + "learning_rate": 1.8551811117290707e-07, + "loss": 0.2164, + "step": 41597 + }, + { + "epoch": 0.9156504512436716, + "grad_norm": 0.6111891081632733, + "learning_rate": 1.854219260300566e-07, + "loss": 0.4653, + "step": 41598 + }, + { + "epoch": 0.9156724631300902, + "grad_norm": 0.5922560948015262, + "learning_rate": 1.853257653570395e-07, + "loss": 0.425, + "step": 41599 + }, + { + "epoch": 0.9156944750165089, + "grad_norm": 1.6251988737099172, + "learning_rate": 1.8522962915434428e-07, + "loss": 0.3317, + "step": 41600 + }, + { + "epoch": 0.9157164869029276, + "grad_norm": 2.2167980149728517, + "learning_rate": 1.8513351742245945e-07, + "loss": 0.3193, + "step": 41601 + }, + { + "epoch": 0.9157384987893462, + "grad_norm": 1.4335735100850027, + "learning_rate": 1.8503743016187402e-07, + "loss": 0.2223, + "step": 41602 + }, + { + "epoch": 0.9157605106757649, + "grad_norm": 1.58181114289948, + "learning_rate": 1.8494136737307545e-07, + "loss": 0.3225, + "step": 41603 + }, + { + "epoch": 0.9157825225621836, + "grad_norm": 1.5563416570641921, + "learning_rate": 1.8484532905655217e-07, + "loss": 0.2721, + "step": 41604 + }, + { + "epoch": 0.9158045344486022, + "grad_norm": 4.342703395592125, + "learning_rate": 1.8474931521279327e-07, + "loss": 0.3143, + "step": 41605 + }, + { + "epoch": 0.9158265463350209, + "grad_norm": 1.427428202593563, + "learning_rate": 1.8465332584228556e-07, + "loss": 0.2699, + "step": 41606 + }, + { + "epoch": 0.9158485582214396, + "grad_norm": 1.6740577468855595, + "learning_rate": 1.8455736094551812e-07, + "loss": 0.2821, + "step": 41607 + }, + { + "epoch": 0.9158705701078582, + "grad_norm": 1.657974438839753, + "learning_rate": 1.8446142052297666e-07, + "loss": 0.2875, + "step": 41608 + }, + { + "epoch": 0.9158925819942769, + "grad_norm": 1.727475105562275, + "learning_rate": 1.843655045751508e-07, + "loss": 0.2947, + "step": 41609 + }, + { + "epoch": 0.9159145938806956, + "grad_norm": 1.9013981461685332, + "learning_rate": 1.8426961310252732e-07, + "loss": 0.2998, + "step": 41610 + }, + { + "epoch": 0.9159366057671142, + "grad_norm": 1.8308837928602208, + "learning_rate": 1.8417374610559314e-07, + "loss": 0.2484, + "step": 41611 + }, + { + "epoch": 0.9159586176535329, + "grad_norm": 0.6213757375671853, + "learning_rate": 1.8407790358483669e-07, + "loss": 0.4579, + "step": 41612 + }, + { + "epoch": 0.9159806295399515, + "grad_norm": 1.2241697984402349, + "learning_rate": 1.839820855407437e-07, + "loss": 0.2479, + "step": 41613 + }, + { + "epoch": 0.9160026414263702, + "grad_norm": 1.7331061938966061, + "learning_rate": 1.8388629197380104e-07, + "loss": 0.278, + "step": 41614 + }, + { + "epoch": 0.9160246533127889, + "grad_norm": 1.7316164692438685, + "learning_rate": 1.8379052288449716e-07, + "loss": 0.2486, + "step": 41615 + }, + { + "epoch": 0.9160466651992075, + "grad_norm": 1.7070687281602048, + "learning_rate": 1.8369477827331783e-07, + "loss": 0.2235, + "step": 41616 + }, + { + "epoch": 0.9160686770856262, + "grad_norm": 1.3015250544468004, + "learning_rate": 1.8359905814074985e-07, + "loss": 0.2463, + "step": 41617 + }, + { + "epoch": 0.9160906889720449, + "grad_norm": 0.6387441419957668, + "learning_rate": 1.8350336248728007e-07, + "loss": 0.4187, + "step": 41618 + }, + { + "epoch": 0.9161127008584635, + "grad_norm": 1.458804058929746, + "learning_rate": 1.834076913133931e-07, + "loss": 0.2479, + "step": 41619 + }, + { + "epoch": 0.9161347127448822, + "grad_norm": 2.2198875335325163, + "learning_rate": 1.833120446195774e-07, + "loss": 0.2765, + "step": 41620 + }, + { + "epoch": 0.9161567246313009, + "grad_norm": 2.1986476758601246, + "learning_rate": 1.832164224063182e-07, + "loss": 0.2441, + "step": 41621 + }, + { + "epoch": 0.9161787365177195, + "grad_norm": 2.0445586506723763, + "learning_rate": 1.8312082467410176e-07, + "loss": 0.2948, + "step": 41622 + }, + { + "epoch": 0.9162007484041382, + "grad_norm": 0.598257333365776, + "learning_rate": 1.830252514234132e-07, + "loss": 0.4287, + "step": 41623 + }, + { + "epoch": 0.9162227602905569, + "grad_norm": 1.3422760952014867, + "learning_rate": 1.8292970265473885e-07, + "loss": 0.2246, + "step": 41624 + }, + { + "epoch": 0.9162447721769755, + "grad_norm": 1.755505415940975, + "learning_rate": 1.8283417836856388e-07, + "loss": 0.2433, + "step": 41625 + }, + { + "epoch": 0.9162667840633942, + "grad_norm": 3.8896319262009045, + "learning_rate": 1.827386785653751e-07, + "loss": 0.3067, + "step": 41626 + }, + { + "epoch": 0.9162887959498129, + "grad_norm": 0.6152026692801925, + "learning_rate": 1.82643203245656e-07, + "loss": 0.4506, + "step": 41627 + }, + { + "epoch": 0.9163108078362315, + "grad_norm": 1.512266483003021, + "learning_rate": 1.8254775240989288e-07, + "loss": 0.2154, + "step": 41628 + }, + { + "epoch": 0.9163328197226502, + "grad_norm": 0.6062483135185428, + "learning_rate": 1.824523260585709e-07, + "loss": 0.438, + "step": 41629 + }, + { + "epoch": 0.9163548316090689, + "grad_norm": 1.6315077812159555, + "learning_rate": 1.8235692419217466e-07, + "loss": 0.3223, + "step": 41630 + }, + { + "epoch": 0.9163768434954875, + "grad_norm": 1.5413164520833575, + "learning_rate": 1.8226154681118992e-07, + "loss": 0.3342, + "step": 41631 + }, + { + "epoch": 0.9163988553819062, + "grad_norm": 1.6942287802560534, + "learning_rate": 1.8216619391609957e-07, + "loss": 0.2782, + "step": 41632 + }, + { + "epoch": 0.916420867268325, + "grad_norm": 1.854172592698701, + "learning_rate": 1.820708655073905e-07, + "loss": 0.3183, + "step": 41633 + }, + { + "epoch": 0.9164428791547435, + "grad_norm": 2.4329997802142924, + "learning_rate": 1.8197556158554563e-07, + "loss": 0.3946, + "step": 41634 + }, + { + "epoch": 0.9164648910411622, + "grad_norm": 1.7583706939728503, + "learning_rate": 1.8188028215105014e-07, + "loss": 0.2442, + "step": 41635 + }, + { + "epoch": 0.916486902927581, + "grad_norm": 1.858615391509764, + "learning_rate": 1.8178502720438807e-07, + "loss": 0.2129, + "step": 41636 + }, + { + "epoch": 0.9165089148139995, + "grad_norm": 1.7275357046606363, + "learning_rate": 1.816897967460429e-07, + "loss": 0.2431, + "step": 41637 + }, + { + "epoch": 0.9165309267004182, + "grad_norm": 1.7240759540184325, + "learning_rate": 1.8159459077649988e-07, + "loss": 0.3157, + "step": 41638 + }, + { + "epoch": 0.9165529385868368, + "grad_norm": 0.6049845049252233, + "learning_rate": 1.814994092962419e-07, + "loss": 0.4453, + "step": 41639 + }, + { + "epoch": 0.9165749504732555, + "grad_norm": 1.9270516545823437, + "learning_rate": 1.8140425230575297e-07, + "loss": 0.2992, + "step": 41640 + }, + { + "epoch": 0.9165969623596742, + "grad_norm": 2.047403471675198, + "learning_rate": 1.8130911980551725e-07, + "loss": 0.3067, + "step": 41641 + }, + { + "epoch": 0.9166189742460928, + "grad_norm": 2.8577923038584934, + "learning_rate": 1.8121401179601706e-07, + "loss": 0.3657, + "step": 41642 + }, + { + "epoch": 0.9166409861325115, + "grad_norm": 0.6271132308515918, + "learning_rate": 1.8111892827773702e-07, + "loss": 0.4737, + "step": 41643 + }, + { + "epoch": 0.9166629980189303, + "grad_norm": 1.4261041293487402, + "learning_rate": 1.8102386925115956e-07, + "loss": 0.2575, + "step": 41644 + }, + { + "epoch": 0.9166850099053488, + "grad_norm": 2.127282204829326, + "learning_rate": 1.8092883471676815e-07, + "loss": 0.3464, + "step": 41645 + }, + { + "epoch": 0.9167070217917676, + "grad_norm": 2.2489704913784823, + "learning_rate": 1.8083382467504628e-07, + "loss": 0.2965, + "step": 41646 + }, + { + "epoch": 0.9167290336781863, + "grad_norm": 2.6497842641666907, + "learning_rate": 1.8073883912647527e-07, + "loss": 0.2457, + "step": 41647 + }, + { + "epoch": 0.9167510455646048, + "grad_norm": 1.8656631808042092, + "learning_rate": 1.8064387807154026e-07, + "loss": 0.2826, + "step": 41648 + }, + { + "epoch": 0.9167730574510236, + "grad_norm": 1.9804506576121284, + "learning_rate": 1.80548941510722e-07, + "loss": 0.2677, + "step": 41649 + }, + { + "epoch": 0.9167950693374423, + "grad_norm": 3.963276485162509, + "learning_rate": 1.8045402944450286e-07, + "loss": 0.2899, + "step": 41650 + }, + { + "epoch": 0.9168170812238609, + "grad_norm": 0.6107309369636765, + "learning_rate": 1.803591418733669e-07, + "loss": 0.4476, + "step": 41651 + }, + { + "epoch": 0.9168390931102796, + "grad_norm": 0.625919790706795, + "learning_rate": 1.802642787977943e-07, + "loss": 0.4563, + "step": 41652 + }, + { + "epoch": 0.9168611049966983, + "grad_norm": 2.7875525693656242, + "learning_rate": 1.8016944021826966e-07, + "loss": 0.2985, + "step": 41653 + }, + { + "epoch": 0.9168831168831169, + "grad_norm": 2.410727531403239, + "learning_rate": 1.8007462613527314e-07, + "loss": 0.2198, + "step": 41654 + }, + { + "epoch": 0.9169051287695356, + "grad_norm": 1.6494031053770748, + "learning_rate": 1.7997983654928608e-07, + "loss": 0.2585, + "step": 41655 + }, + { + "epoch": 0.9169271406559543, + "grad_norm": 2.143984884680502, + "learning_rate": 1.7988507146079248e-07, + "loss": 0.2738, + "step": 41656 + }, + { + "epoch": 0.9169491525423729, + "grad_norm": 2.4032465850518525, + "learning_rate": 1.7979033087027254e-07, + "loss": 0.2383, + "step": 41657 + }, + { + "epoch": 0.9169711644287916, + "grad_norm": 1.3773771473838707, + "learning_rate": 1.7969561477820808e-07, + "loss": 0.3141, + "step": 41658 + }, + { + "epoch": 0.9169931763152103, + "grad_norm": 0.6310897665725816, + "learning_rate": 1.7960092318508038e-07, + "loss": 0.4866, + "step": 41659 + }, + { + "epoch": 0.9170151882016289, + "grad_norm": 1.8104980066195515, + "learning_rate": 1.7950625609137074e-07, + "loss": 0.2233, + "step": 41660 + }, + { + "epoch": 0.9170372000880476, + "grad_norm": 0.6269708797526695, + "learning_rate": 1.7941161349755988e-07, + "loss": 0.4503, + "step": 41661 + }, + { + "epoch": 0.9170592119744662, + "grad_norm": 1.6399272538265972, + "learning_rate": 1.793169954041296e-07, + "loss": 0.2516, + "step": 41662 + }, + { + "epoch": 0.9170812238608849, + "grad_norm": 2.5854565713265996, + "learning_rate": 1.7922240181156015e-07, + "loss": 0.2478, + "step": 41663 + }, + { + "epoch": 0.9171032357473036, + "grad_norm": 1.4070324639182135, + "learning_rate": 1.791278327203333e-07, + "loss": 0.1831, + "step": 41664 + }, + { + "epoch": 0.9171252476337222, + "grad_norm": 0.6133076988673152, + "learning_rate": 1.7903328813092813e-07, + "loss": 0.4734, + "step": 41665 + }, + { + "epoch": 0.9171472595201409, + "grad_norm": 1.9037736145741129, + "learning_rate": 1.7893876804382593e-07, + "loss": 0.3036, + "step": 41666 + }, + { + "epoch": 0.9171692714065596, + "grad_norm": 0.6036349216591675, + "learning_rate": 1.7884427245950798e-07, + "loss": 0.4446, + "step": 41667 + }, + { + "epoch": 0.9171912832929782, + "grad_norm": 1.799091310984041, + "learning_rate": 1.7874980137845332e-07, + "loss": 0.2851, + "step": 41668 + }, + { + "epoch": 0.9172132951793969, + "grad_norm": 1.5279902072780998, + "learning_rate": 1.786553548011427e-07, + "loss": 0.2133, + "step": 41669 + }, + { + "epoch": 0.9172353070658156, + "grad_norm": 1.7939887491787119, + "learning_rate": 1.7856093272805576e-07, + "loss": 0.3215, + "step": 41670 + }, + { + "epoch": 0.9172573189522342, + "grad_norm": 1.7339273923248426, + "learning_rate": 1.7846653515967205e-07, + "loss": 0.2321, + "step": 41671 + }, + { + "epoch": 0.9172793308386529, + "grad_norm": 1.6763223522687247, + "learning_rate": 1.783721620964729e-07, + "loss": 0.2518, + "step": 41672 + }, + { + "epoch": 0.9173013427250716, + "grad_norm": 2.064974105825411, + "learning_rate": 1.7827781353893624e-07, + "loss": 0.289, + "step": 41673 + }, + { + "epoch": 0.9173233546114902, + "grad_norm": 2.641542523777691, + "learning_rate": 1.7818348948754282e-07, + "loss": 0.2914, + "step": 41674 + }, + { + "epoch": 0.9173453664979089, + "grad_norm": 1.5565993185398037, + "learning_rate": 1.7808918994277115e-07, + "loss": 0.2828, + "step": 41675 + }, + { + "epoch": 0.9173673783843276, + "grad_norm": 1.4122879443233813, + "learning_rate": 1.7799491490510024e-07, + "loss": 0.3268, + "step": 41676 + }, + { + "epoch": 0.9173893902707462, + "grad_norm": 1.756613858879323, + "learning_rate": 1.7790066437501087e-07, + "loss": 0.4107, + "step": 41677 + }, + { + "epoch": 0.9174114021571649, + "grad_norm": 1.4959900950770442, + "learning_rate": 1.7780643835298039e-07, + "loss": 0.3575, + "step": 41678 + }, + { + "epoch": 0.9174334140435836, + "grad_norm": 1.851839058501021, + "learning_rate": 1.77712236839489e-07, + "loss": 0.2541, + "step": 41679 + }, + { + "epoch": 0.9174554259300022, + "grad_norm": 2.365309932583915, + "learning_rate": 1.7761805983501413e-07, + "loss": 0.2562, + "step": 41680 + }, + { + "epoch": 0.9174774378164209, + "grad_norm": 0.6046846648576577, + "learning_rate": 1.7752390734003532e-07, + "loss": 0.4338, + "step": 41681 + }, + { + "epoch": 0.9174994497028396, + "grad_norm": 0.6325092305446242, + "learning_rate": 1.7742977935503115e-07, + "loss": 0.4405, + "step": 41682 + }, + { + "epoch": 0.9175214615892582, + "grad_norm": 1.7733819477021242, + "learning_rate": 1.773356758804795e-07, + "loss": 0.2102, + "step": 41683 + }, + { + "epoch": 0.9175434734756769, + "grad_norm": 0.6038011806563526, + "learning_rate": 1.772415969168595e-07, + "loss": 0.4438, + "step": 41684 + }, + { + "epoch": 0.9175654853620955, + "grad_norm": 2.486012339536393, + "learning_rate": 1.771475424646485e-07, + "loss": 0.3106, + "step": 41685 + }, + { + "epoch": 0.9175874972485142, + "grad_norm": 9.727265227817151, + "learning_rate": 1.7705351252432334e-07, + "loss": 0.2892, + "step": 41686 + }, + { + "epoch": 0.9176095091349329, + "grad_norm": 1.564814613882576, + "learning_rate": 1.7695950709636477e-07, + "loss": 0.2832, + "step": 41687 + }, + { + "epoch": 0.9176315210213515, + "grad_norm": 1.742074513703781, + "learning_rate": 1.7686552618124853e-07, + "loss": 0.2622, + "step": 41688 + }, + { + "epoch": 0.9176535329077702, + "grad_norm": 0.6332828369801495, + "learning_rate": 1.7677156977945366e-07, + "loss": 0.4465, + "step": 41689 + }, + { + "epoch": 0.9176755447941889, + "grad_norm": 1.5786857214900478, + "learning_rate": 1.76677637891457e-07, + "loss": 0.325, + "step": 41690 + }, + { + "epoch": 0.9176975566806075, + "grad_norm": 1.4368853784215834, + "learning_rate": 1.7658373051773427e-07, + "loss": 0.2375, + "step": 41691 + }, + { + "epoch": 0.9177195685670262, + "grad_norm": 1.9479139570655517, + "learning_rate": 1.7648984765876565e-07, + "loss": 0.268, + "step": 41692 + }, + { + "epoch": 0.9177415804534449, + "grad_norm": 0.589513222885403, + "learning_rate": 1.7639598931502632e-07, + "loss": 0.4414, + "step": 41693 + }, + { + "epoch": 0.9177635923398635, + "grad_norm": 2.1392832026534156, + "learning_rate": 1.7630215548699425e-07, + "loss": 0.262, + "step": 41694 + }, + { + "epoch": 0.9177856042262822, + "grad_norm": 1.6567963194914215, + "learning_rate": 1.7620834617514682e-07, + "loss": 0.2666, + "step": 41695 + }, + { + "epoch": 0.9178076161127009, + "grad_norm": 1.5119138437353745, + "learning_rate": 1.7611456137995863e-07, + "loss": 0.3026, + "step": 41696 + }, + { + "epoch": 0.9178296279991195, + "grad_norm": 2.226810354416156, + "learning_rate": 1.7602080110190823e-07, + "loss": 0.3218, + "step": 41697 + }, + { + "epoch": 0.9178516398855382, + "grad_norm": 2.2712203096235917, + "learning_rate": 1.7592706534147186e-07, + "loss": 0.249, + "step": 41698 + }, + { + "epoch": 0.9178736517719569, + "grad_norm": 1.5087379100656377, + "learning_rate": 1.7583335409912583e-07, + "loss": 0.226, + "step": 41699 + }, + { + "epoch": 0.9178956636583755, + "grad_norm": 3.30084755915292, + "learning_rate": 1.7573966737534642e-07, + "loss": 0.2998, + "step": 41700 + }, + { + "epoch": 0.9179176755447942, + "grad_norm": 1.8421703966196055, + "learning_rate": 1.7564600517060938e-07, + "loss": 0.3587, + "step": 41701 + }, + { + "epoch": 0.9179396874312129, + "grad_norm": 0.6083534796309582, + "learning_rate": 1.7555236748539095e-07, + "loss": 0.4291, + "step": 41702 + }, + { + "epoch": 0.9179616993176315, + "grad_norm": 0.6069678922241025, + "learning_rate": 1.7545875432016802e-07, + "loss": 0.4446, + "step": 41703 + }, + { + "epoch": 0.9179837112040502, + "grad_norm": 1.7309130624354059, + "learning_rate": 1.7536516567541462e-07, + "loss": 0.2947, + "step": 41704 + }, + { + "epoch": 0.9180057230904689, + "grad_norm": 0.5847665261499922, + "learning_rate": 1.7527160155160762e-07, + "loss": 0.449, + "step": 41705 + }, + { + "epoch": 0.9180277349768875, + "grad_norm": 1.4255209768604284, + "learning_rate": 1.7517806194922214e-07, + "loss": 0.3251, + "step": 41706 + }, + { + "epoch": 0.9180497468633062, + "grad_norm": 1.2066806925236726, + "learning_rate": 1.7508454686873343e-07, + "loss": 0.1755, + "step": 41707 + }, + { + "epoch": 0.9180717587497248, + "grad_norm": 2.6538642277342963, + "learning_rate": 1.7499105631061775e-07, + "loss": 0.2515, + "step": 41708 + }, + { + "epoch": 0.9180937706361435, + "grad_norm": 1.851814482435071, + "learning_rate": 1.7489759027534858e-07, + "loss": 0.2811, + "step": 41709 + }, + { + "epoch": 0.9181157825225622, + "grad_norm": 3.127878400459061, + "learning_rate": 1.7480414876340278e-07, + "loss": 0.4346, + "step": 41710 + }, + { + "epoch": 0.9181377944089808, + "grad_norm": 1.780866600005957, + "learning_rate": 1.7471073177525388e-07, + "loss": 0.3493, + "step": 41711 + }, + { + "epoch": 0.9181598062953995, + "grad_norm": 1.7088485461199288, + "learning_rate": 1.7461733931137703e-07, + "loss": 0.3435, + "step": 41712 + }, + { + "epoch": 0.9181818181818182, + "grad_norm": 2.064648080262577, + "learning_rate": 1.745239713722474e-07, + "loss": 0.2927, + "step": 41713 + }, + { + "epoch": 0.9182038300682368, + "grad_norm": 3.408234622293118, + "learning_rate": 1.744306279583391e-07, + "loss": 0.3379, + "step": 41714 + }, + { + "epoch": 0.9182258419546555, + "grad_norm": 2.1957061716058477, + "learning_rate": 1.743373090701267e-07, + "loss": 0.2898, + "step": 41715 + }, + { + "epoch": 0.9182478538410742, + "grad_norm": 2.3374180530437605, + "learning_rate": 1.7424401470808427e-07, + "loss": 0.2843, + "step": 41716 + }, + { + "epoch": 0.9182698657274928, + "grad_norm": 2.137778095750195, + "learning_rate": 1.741507448726859e-07, + "loss": 0.295, + "step": 41717 + }, + { + "epoch": 0.9182918776139115, + "grad_norm": 2.0229177926734017, + "learning_rate": 1.740574995644062e-07, + "loss": 0.3318, + "step": 41718 + }, + { + "epoch": 0.9183138895003302, + "grad_norm": 1.9290179018779394, + "learning_rate": 1.7396427878371867e-07, + "loss": 0.2881, + "step": 41719 + }, + { + "epoch": 0.9183359013867488, + "grad_norm": 0.6140653168460786, + "learning_rate": 1.738710825310974e-07, + "loss": 0.4447, + "step": 41720 + }, + { + "epoch": 0.9183579132731675, + "grad_norm": 0.6211050658320404, + "learning_rate": 1.7377791080701533e-07, + "loss": 0.4609, + "step": 41721 + }, + { + "epoch": 0.9183799251595862, + "grad_norm": 1.7232767158893991, + "learning_rate": 1.7368476361194652e-07, + "loss": 0.3117, + "step": 41722 + }, + { + "epoch": 0.9184019370460048, + "grad_norm": 1.9266050070500145, + "learning_rate": 1.735916409463645e-07, + "loss": 0.2532, + "step": 41723 + }, + { + "epoch": 0.9184239489324235, + "grad_norm": 1.8881246215708152, + "learning_rate": 1.7349854281074218e-07, + "loss": 0.2496, + "step": 41724 + }, + { + "epoch": 0.9184459608188422, + "grad_norm": 2.667155534617034, + "learning_rate": 1.7340546920555368e-07, + "loss": 0.2888, + "step": 41725 + }, + { + "epoch": 0.9184679727052608, + "grad_norm": 0.6204990551503421, + "learning_rate": 1.733124201312708e-07, + "loss": 0.4592, + "step": 41726 + }, + { + "epoch": 0.9184899845916795, + "grad_norm": 1.3999250834144248, + "learning_rate": 1.7321939558836654e-07, + "loss": 0.2933, + "step": 41727 + }, + { + "epoch": 0.9185119964780982, + "grad_norm": 0.6123044112464497, + "learning_rate": 1.7312639557731438e-07, + "loss": 0.4457, + "step": 41728 + }, + { + "epoch": 0.9185340083645168, + "grad_norm": 1.7349850121104538, + "learning_rate": 1.730334200985867e-07, + "loss": 0.348, + "step": 41729 + }, + { + "epoch": 0.9185560202509355, + "grad_norm": 2.3137829368703704, + "learning_rate": 1.729404691526565e-07, + "loss": 0.3169, + "step": 41730 + }, + { + "epoch": 0.9185780321373541, + "grad_norm": 1.4391204851989354, + "learning_rate": 1.728475427399956e-07, + "loss": 0.3191, + "step": 41731 + }, + { + "epoch": 0.9186000440237728, + "grad_norm": 1.5548789639182785, + "learning_rate": 1.7275464086107641e-07, + "loss": 0.3062, + "step": 41732 + }, + { + "epoch": 0.9186220559101915, + "grad_norm": 1.5259626112749867, + "learning_rate": 1.7266176351637077e-07, + "loss": 0.2725, + "step": 41733 + }, + { + "epoch": 0.9186440677966101, + "grad_norm": 0.6190169396390774, + "learning_rate": 1.7256891070635108e-07, + "loss": 0.4418, + "step": 41734 + }, + { + "epoch": 0.9186660796830288, + "grad_norm": 1.9525763692480245, + "learning_rate": 1.7247608243148972e-07, + "loss": 0.3083, + "step": 41735 + }, + { + "epoch": 0.9186880915694475, + "grad_norm": 0.6187874892584548, + "learning_rate": 1.7238327869225802e-07, + "loss": 0.4418, + "step": 41736 + }, + { + "epoch": 0.9187101034558661, + "grad_norm": 1.6148932937563867, + "learning_rate": 1.722904994891278e-07, + "loss": 0.288, + "step": 41737 + }, + { + "epoch": 0.9187321153422848, + "grad_norm": 0.6294221749425963, + "learning_rate": 1.7219774482256978e-07, + "loss": 0.4236, + "step": 41738 + }, + { + "epoch": 0.9187541272287035, + "grad_norm": 2.3794669495478895, + "learning_rate": 1.721050146930564e-07, + "loss": 0.2835, + "step": 41739 + }, + { + "epoch": 0.9187761391151221, + "grad_norm": 1.9436172117674304, + "learning_rate": 1.7201230910105892e-07, + "loss": 0.3034, + "step": 41740 + }, + { + "epoch": 0.9187981510015408, + "grad_norm": 2.0882787865140955, + "learning_rate": 1.7191962804704865e-07, + "loss": 0.286, + "step": 41741 + }, + { + "epoch": 0.9188201628879595, + "grad_norm": 3.5622838382526076, + "learning_rate": 1.718269715314952e-07, + "loss": 0.244, + "step": 41742 + }, + { + "epoch": 0.9188421747743781, + "grad_norm": 1.7805498593834932, + "learning_rate": 1.7173433955487096e-07, + "loss": 0.2771, + "step": 41743 + }, + { + "epoch": 0.9188641866607968, + "grad_norm": 1.306043701346449, + "learning_rate": 1.7164173211764613e-07, + "loss": 0.1952, + "step": 41744 + }, + { + "epoch": 0.9188861985472155, + "grad_norm": 1.9529993567722996, + "learning_rate": 1.715491492202914e-07, + "loss": 0.268, + "step": 41745 + }, + { + "epoch": 0.9189082104336341, + "grad_norm": 5.333064620597289, + "learning_rate": 1.7145659086327816e-07, + "loss": 0.2989, + "step": 41746 + }, + { + "epoch": 0.9189302223200528, + "grad_norm": 1.9887861042558803, + "learning_rate": 1.7136405704707538e-07, + "loss": 0.2715, + "step": 41747 + }, + { + "epoch": 0.9189522342064715, + "grad_norm": 1.7474931126090585, + "learning_rate": 1.7127154777215439e-07, + "loss": 0.2792, + "step": 41748 + }, + { + "epoch": 0.9189742460928901, + "grad_norm": 2.1144070181070145, + "learning_rate": 1.7117906303898534e-07, + "loss": 0.3234, + "step": 41749 + }, + { + "epoch": 0.9189962579793088, + "grad_norm": 0.6057653038902591, + "learning_rate": 1.7108660284803736e-07, + "loss": 0.4625, + "step": 41750 + }, + { + "epoch": 0.9190182698657275, + "grad_norm": 1.3810525640958093, + "learning_rate": 1.7099416719978167e-07, + "loss": 0.3102, + "step": 41751 + }, + { + "epoch": 0.9190402817521461, + "grad_norm": 0.6045688166562825, + "learning_rate": 1.7090175609468684e-07, + "loss": 0.455, + "step": 41752 + }, + { + "epoch": 0.9190622936385648, + "grad_norm": 2.3233095173629956, + "learning_rate": 1.708093695332236e-07, + "loss": 0.2857, + "step": 41753 + }, + { + "epoch": 0.9190843055249834, + "grad_norm": 1.6234000245188571, + "learning_rate": 1.7071700751586096e-07, + "loss": 0.3525, + "step": 41754 + }, + { + "epoch": 0.9191063174114021, + "grad_norm": 2.3156910591662547, + "learning_rate": 1.7062467004306805e-07, + "loss": 0.2938, + "step": 41755 + }, + { + "epoch": 0.9191283292978208, + "grad_norm": 1.7365275986402495, + "learning_rate": 1.7053235711531502e-07, + "loss": 0.2675, + "step": 41756 + }, + { + "epoch": 0.9191503411842394, + "grad_norm": 2.4213168012250943, + "learning_rate": 1.7044006873306984e-07, + "loss": 0.2475, + "step": 41757 + }, + { + "epoch": 0.9191723530706581, + "grad_norm": 1.800027795127306, + "learning_rate": 1.7034780489680269e-07, + "loss": 0.2439, + "step": 41758 + }, + { + "epoch": 0.9191943649570768, + "grad_norm": 1.8360723458835795, + "learning_rate": 1.7025556560698264e-07, + "loss": 0.3103, + "step": 41759 + }, + { + "epoch": 0.9192163768434954, + "grad_norm": 0.649655396369728, + "learning_rate": 1.701633508640771e-07, + "loss": 0.4705, + "step": 41760 + }, + { + "epoch": 0.9192383887299141, + "grad_norm": 1.4022515923680976, + "learning_rate": 1.7007116066855624e-07, + "loss": 0.3087, + "step": 41761 + }, + { + "epoch": 0.9192604006163329, + "grad_norm": 4.0172877333755235, + "learning_rate": 1.69978995020888e-07, + "loss": 0.2592, + "step": 41762 + }, + { + "epoch": 0.9192824125027514, + "grad_norm": 1.5062165865994457, + "learning_rate": 1.698868539215398e-07, + "loss": 0.2339, + "step": 41763 + }, + { + "epoch": 0.9193044243891701, + "grad_norm": 1.6887956398833162, + "learning_rate": 1.6979473737098128e-07, + "loss": 0.3524, + "step": 41764 + }, + { + "epoch": 0.9193264362755889, + "grad_norm": 0.6299950048112329, + "learning_rate": 1.6970264536968039e-07, + "loss": 0.4716, + "step": 41765 + }, + { + "epoch": 0.9193484481620074, + "grad_norm": 1.9027347400676426, + "learning_rate": 1.696105779181051e-07, + "loss": 0.2985, + "step": 41766 + }, + { + "epoch": 0.9193704600484262, + "grad_norm": 2.491397953660988, + "learning_rate": 1.6951853501672388e-07, + "loss": 0.3068, + "step": 41767 + }, + { + "epoch": 0.9193924719348449, + "grad_norm": 2.560963218702147, + "learning_rate": 1.6942651666600251e-07, + "loss": 0.3235, + "step": 41768 + }, + { + "epoch": 0.9194144838212635, + "grad_norm": 2.3669241725621526, + "learning_rate": 1.693345228664106e-07, + "loss": 0.2426, + "step": 41769 + }, + { + "epoch": 0.9194364957076822, + "grad_norm": 2.1238914070931187, + "learning_rate": 1.6924255361841557e-07, + "loss": 0.2034, + "step": 41770 + }, + { + "epoch": 0.9194585075941009, + "grad_norm": 2.104473597849504, + "learning_rate": 1.6915060892248424e-07, + "loss": 0.2897, + "step": 41771 + }, + { + "epoch": 0.9194805194805195, + "grad_norm": 0.6217281153722221, + "learning_rate": 1.690586887790846e-07, + "loss": 0.4379, + "step": 41772 + }, + { + "epoch": 0.9195025313669382, + "grad_norm": 1.756023735272273, + "learning_rate": 1.689667931886829e-07, + "loss": 0.3009, + "step": 41773 + }, + { + "epoch": 0.9195245432533569, + "grad_norm": 1.4800097071579759, + "learning_rate": 1.688749221517466e-07, + "loss": 0.2381, + "step": 41774 + }, + { + "epoch": 0.9195465551397755, + "grad_norm": 1.7383798830138255, + "learning_rate": 1.6878307566874253e-07, + "loss": 0.2448, + "step": 41775 + }, + { + "epoch": 0.9195685670261942, + "grad_norm": 2.2142460261012196, + "learning_rate": 1.6869125374013805e-07, + "loss": 0.3356, + "step": 41776 + }, + { + "epoch": 0.9195905789126129, + "grad_norm": 1.2101780041409702, + "learning_rate": 1.685994563663995e-07, + "loss": 0.2281, + "step": 41777 + }, + { + "epoch": 0.9196125907990315, + "grad_norm": 0.6222537693682756, + "learning_rate": 1.6850768354799318e-07, + "loss": 0.4618, + "step": 41778 + }, + { + "epoch": 0.9196346026854502, + "grad_norm": 2.0966488253190945, + "learning_rate": 1.6841593528538535e-07, + "loss": 0.3606, + "step": 41779 + }, + { + "epoch": 0.9196566145718688, + "grad_norm": 2.358692079931427, + "learning_rate": 1.6832421157904288e-07, + "loss": 0.301, + "step": 41780 + }, + { + "epoch": 0.9196786264582875, + "grad_norm": 3.795343885855057, + "learning_rate": 1.6823251242943263e-07, + "loss": 0.2426, + "step": 41781 + }, + { + "epoch": 0.9197006383447062, + "grad_norm": 0.6099696854265025, + "learning_rate": 1.6814083783701918e-07, + "loss": 0.4435, + "step": 41782 + }, + { + "epoch": 0.9197226502311248, + "grad_norm": 1.599099571537945, + "learning_rate": 1.6804918780226886e-07, + "loss": 0.2404, + "step": 41783 + }, + { + "epoch": 0.9197446621175435, + "grad_norm": 1.6363642596812453, + "learning_rate": 1.6795756232564743e-07, + "loss": 0.2438, + "step": 41784 + }, + { + "epoch": 0.9197666740039622, + "grad_norm": 1.5644324940126093, + "learning_rate": 1.6786596140762168e-07, + "loss": 0.3665, + "step": 41785 + }, + { + "epoch": 0.9197886858903808, + "grad_norm": 1.869409156071017, + "learning_rate": 1.6777438504865518e-07, + "loss": 0.3385, + "step": 41786 + }, + { + "epoch": 0.9198106977767995, + "grad_norm": 0.6193729847592905, + "learning_rate": 1.6768283324921587e-07, + "loss": 0.4594, + "step": 41787 + }, + { + "epoch": 0.9198327096632182, + "grad_norm": 1.4983568109698486, + "learning_rate": 1.675913060097667e-07, + "loss": 0.2576, + "step": 41788 + }, + { + "epoch": 0.9198547215496368, + "grad_norm": 2.1416062301392267, + "learning_rate": 1.6749980333077342e-07, + "loss": 0.3082, + "step": 41789 + }, + { + "epoch": 0.9198767334360555, + "grad_norm": 0.6325340299138915, + "learning_rate": 1.6740832521270233e-07, + "loss": 0.453, + "step": 41790 + }, + { + "epoch": 0.9198987453224742, + "grad_norm": 2.3590446930083404, + "learning_rate": 1.673168716560175e-07, + "loss": 0.2535, + "step": 41791 + }, + { + "epoch": 0.9199207572088928, + "grad_norm": 1.5995353467898028, + "learning_rate": 1.6722544266118357e-07, + "loss": 0.265, + "step": 41792 + }, + { + "epoch": 0.9199427690953115, + "grad_norm": 2.7460663446193543, + "learning_rate": 1.6713403822866515e-07, + "loss": 0.2839, + "step": 41793 + }, + { + "epoch": 0.9199647809817302, + "grad_norm": 1.455428490854503, + "learning_rate": 1.670426583589274e-07, + "loss": 0.2666, + "step": 41794 + }, + { + "epoch": 0.9199867928681488, + "grad_norm": 1.7551964861595004, + "learning_rate": 1.6695130305243445e-07, + "loss": 0.371, + "step": 41795 + }, + { + "epoch": 0.9200088047545675, + "grad_norm": 2.438768784179452, + "learning_rate": 1.6685997230965034e-07, + "loss": 0.2875, + "step": 41796 + }, + { + "epoch": 0.9200308166409862, + "grad_norm": 1.7893586722173418, + "learning_rate": 1.6676866613104027e-07, + "loss": 0.3393, + "step": 41797 + }, + { + "epoch": 0.9200528285274048, + "grad_norm": 3.0804110645122194, + "learning_rate": 1.666773845170666e-07, + "loss": 0.2705, + "step": 41798 + }, + { + "epoch": 0.9200748404138235, + "grad_norm": 2.1943710917272825, + "learning_rate": 1.6658612746819457e-07, + "loss": 0.3462, + "step": 41799 + }, + { + "epoch": 0.9200968523002422, + "grad_norm": 1.3691840606043653, + "learning_rate": 1.6649489498488713e-07, + "loss": 0.2105, + "step": 41800 + }, + { + "epoch": 0.9201188641866608, + "grad_norm": 1.4972311445848978, + "learning_rate": 1.664036870676089e-07, + "loss": 0.2499, + "step": 41801 + }, + { + "epoch": 0.9201408760730795, + "grad_norm": 0.6364965473181674, + "learning_rate": 1.6631250371682285e-07, + "loss": 0.4507, + "step": 41802 + }, + { + "epoch": 0.9201628879594981, + "grad_norm": 1.4719236210948978, + "learning_rate": 1.6622134493299302e-07, + "loss": 0.2537, + "step": 41803 + }, + { + "epoch": 0.9201848998459168, + "grad_norm": 2.20197380478568, + "learning_rate": 1.6613021071658187e-07, + "loss": 0.2706, + "step": 41804 + }, + { + "epoch": 0.9202069117323355, + "grad_norm": 1.993743883023478, + "learning_rate": 1.660391010680523e-07, + "loss": 0.2766, + "step": 41805 + }, + { + "epoch": 0.9202289236187541, + "grad_norm": 2.395500568598658, + "learning_rate": 1.6594801598786848e-07, + "loss": 0.3319, + "step": 41806 + }, + { + "epoch": 0.9202509355051728, + "grad_norm": 0.6004744449161032, + "learning_rate": 1.6585695547649327e-07, + "loss": 0.4328, + "step": 41807 + }, + { + "epoch": 0.9202729473915915, + "grad_norm": 1.8822358786314743, + "learning_rate": 1.6576591953438914e-07, + "loss": 0.3238, + "step": 41808 + }, + { + "epoch": 0.9202949592780101, + "grad_norm": 1.6182524051743388, + "learning_rate": 1.6567490816201847e-07, + "loss": 0.228, + "step": 41809 + }, + { + "epoch": 0.9203169711644288, + "grad_norm": 3.328517091396398, + "learning_rate": 1.655839213598437e-07, + "loss": 0.2576, + "step": 41810 + }, + { + "epoch": 0.9203389830508475, + "grad_norm": 1.4482813052574008, + "learning_rate": 1.6549295912832774e-07, + "loss": 0.2599, + "step": 41811 + }, + { + "epoch": 0.9203609949372661, + "grad_norm": 2.549795435376302, + "learning_rate": 1.654020214679336e-07, + "loss": 0.317, + "step": 41812 + }, + { + "epoch": 0.9203830068236848, + "grad_norm": 4.0652078180121425, + "learning_rate": 1.6531110837912256e-07, + "loss": 0.2687, + "step": 41813 + }, + { + "epoch": 0.9204050187101035, + "grad_norm": 14.633733335171062, + "learning_rate": 1.6522021986235593e-07, + "loss": 0.2819, + "step": 41814 + }, + { + "epoch": 0.9204270305965221, + "grad_norm": 0.6167647422212665, + "learning_rate": 1.6512935591809664e-07, + "loss": 0.4216, + "step": 41815 + }, + { + "epoch": 0.9204490424829408, + "grad_norm": 2.2440679969254997, + "learning_rate": 1.6503851654680658e-07, + "loss": 0.2524, + "step": 41816 + }, + { + "epoch": 0.9204710543693595, + "grad_norm": 1.313045293673578, + "learning_rate": 1.6494770174894815e-07, + "loss": 0.2374, + "step": 41817 + }, + { + "epoch": 0.9204930662557781, + "grad_norm": 1.5071072451211815, + "learning_rate": 1.6485691152498152e-07, + "loss": 0.3048, + "step": 41818 + }, + { + "epoch": 0.9205150781421968, + "grad_norm": 2.058464993481607, + "learning_rate": 1.6476614587536798e-07, + "loss": 0.3096, + "step": 41819 + }, + { + "epoch": 0.9205370900286155, + "grad_norm": 0.6006258389598842, + "learning_rate": 1.6467540480056943e-07, + "loss": 0.4485, + "step": 41820 + }, + { + "epoch": 0.9205591019150341, + "grad_norm": 3.7407516000951464, + "learning_rate": 1.6458468830104712e-07, + "loss": 0.2853, + "step": 41821 + }, + { + "epoch": 0.9205811138014528, + "grad_norm": 1.6627331379961001, + "learning_rate": 1.6449399637726295e-07, + "loss": 0.2815, + "step": 41822 + }, + { + "epoch": 0.9206031256878715, + "grad_norm": 1.614934020459806, + "learning_rate": 1.6440332902967648e-07, + "loss": 0.2093, + "step": 41823 + }, + { + "epoch": 0.9206251375742901, + "grad_norm": 2.1558473003148824, + "learning_rate": 1.6431268625874908e-07, + "loss": 0.2584, + "step": 41824 + }, + { + "epoch": 0.9206471494607088, + "grad_norm": 2.7804218800475695, + "learning_rate": 1.6422206806494089e-07, + "loss": 0.2183, + "step": 41825 + }, + { + "epoch": 0.9206691613471274, + "grad_norm": 1.5041645587951944, + "learning_rate": 1.641314744487138e-07, + "loss": 0.2371, + "step": 41826 + }, + { + "epoch": 0.9206911732335461, + "grad_norm": 0.6327724388996933, + "learning_rate": 1.6404090541052686e-07, + "loss": 0.4537, + "step": 41827 + }, + { + "epoch": 0.9207131851199648, + "grad_norm": 1.6998841985492998, + "learning_rate": 1.6395036095084138e-07, + "loss": 0.3503, + "step": 41828 + }, + { + "epoch": 0.9207351970063834, + "grad_norm": 4.583638731557331, + "learning_rate": 1.6385984107011643e-07, + "loss": 0.2327, + "step": 41829 + }, + { + "epoch": 0.9207572088928021, + "grad_norm": 1.6384220554450921, + "learning_rate": 1.6376934576881276e-07, + "loss": 0.3157, + "step": 41830 + }, + { + "epoch": 0.9207792207792208, + "grad_norm": 2.2311292942161685, + "learning_rate": 1.636788750473911e-07, + "loss": 0.3373, + "step": 41831 + }, + { + "epoch": 0.9208012326656394, + "grad_norm": 2.019725806609958, + "learning_rate": 1.6358842890630944e-07, + "loss": 0.3514, + "step": 41832 + }, + { + "epoch": 0.9208232445520581, + "grad_norm": 2.089147318367316, + "learning_rate": 1.634980073460296e-07, + "loss": 0.2833, + "step": 41833 + }, + { + "epoch": 0.9208452564384768, + "grad_norm": 1.3834573079175128, + "learning_rate": 1.6340761036700902e-07, + "loss": 0.2298, + "step": 41834 + }, + { + "epoch": 0.9208672683248954, + "grad_norm": 1.2811891771865154, + "learning_rate": 1.6331723796970899e-07, + "loss": 0.2834, + "step": 41835 + }, + { + "epoch": 0.9208892802113141, + "grad_norm": 0.6049311292724465, + "learning_rate": 1.632268901545875e-07, + "loss": 0.458, + "step": 41836 + }, + { + "epoch": 0.9209112920977328, + "grad_norm": 1.9599782997729633, + "learning_rate": 1.631365669221041e-07, + "loss": 0.2567, + "step": 41837 + }, + { + "epoch": 0.9209333039841514, + "grad_norm": 1.5574064232323237, + "learning_rate": 1.6304626827271852e-07, + "loss": 0.2164, + "step": 41838 + }, + { + "epoch": 0.9209553158705701, + "grad_norm": 1.438674811730413, + "learning_rate": 1.6295599420688811e-07, + "loss": 0.2972, + "step": 41839 + }, + { + "epoch": 0.9209773277569888, + "grad_norm": 1.5661229914657933, + "learning_rate": 1.6286574472507423e-07, + "loss": 0.2335, + "step": 41840 + }, + { + "epoch": 0.9209993396434074, + "grad_norm": 2.7742808508091, + "learning_rate": 1.6277551982773254e-07, + "loss": 0.2111, + "step": 41841 + }, + { + "epoch": 0.9210213515298261, + "grad_norm": 2.0018243598260375, + "learning_rate": 1.6268531951532386e-07, + "loss": 0.2278, + "step": 41842 + }, + { + "epoch": 0.9210433634162448, + "grad_norm": 1.5948188695023304, + "learning_rate": 1.6259514378830609e-07, + "loss": 0.3228, + "step": 41843 + }, + { + "epoch": 0.9210653753026634, + "grad_norm": 2.3444804222456423, + "learning_rate": 1.6250499264713726e-07, + "loss": 0.3958, + "step": 41844 + }, + { + "epoch": 0.9210873871890821, + "grad_norm": 1.5101900007088065, + "learning_rate": 1.624148660922753e-07, + "loss": 0.2823, + "step": 41845 + }, + { + "epoch": 0.9211093990755008, + "grad_norm": 1.8616252268810654, + "learning_rate": 1.6232476412417875e-07, + "loss": 0.2429, + "step": 41846 + }, + { + "epoch": 0.9211314109619194, + "grad_norm": 1.752566002131773, + "learning_rate": 1.62234686743305e-07, + "loss": 0.292, + "step": 41847 + }, + { + "epoch": 0.9211534228483381, + "grad_norm": 1.5790163464345754, + "learning_rate": 1.6214463395011314e-07, + "loss": 0.2869, + "step": 41848 + }, + { + "epoch": 0.9211754347347567, + "grad_norm": 0.6105012867321139, + "learning_rate": 1.6205460574506005e-07, + "loss": 0.4304, + "step": 41849 + }, + { + "epoch": 0.9211974466211754, + "grad_norm": 1.8462353284003303, + "learning_rate": 1.6196460212860254e-07, + "loss": 0.3239, + "step": 41850 + }, + { + "epoch": 0.9212194585075941, + "grad_norm": 1.848445985554582, + "learning_rate": 1.6187462310119862e-07, + "loss": 0.2748, + "step": 41851 + }, + { + "epoch": 0.9212414703940127, + "grad_norm": 1.7292478608282342, + "learning_rate": 1.6178466866330565e-07, + "loss": 0.243, + "step": 41852 + }, + { + "epoch": 0.9212634822804314, + "grad_norm": 1.555800011160586, + "learning_rate": 1.6169473881538168e-07, + "loss": 0.2409, + "step": 41853 + }, + { + "epoch": 0.9212854941668501, + "grad_norm": 1.8981771504432567, + "learning_rate": 1.6160483355788292e-07, + "loss": 0.2919, + "step": 41854 + }, + { + "epoch": 0.9213075060532687, + "grad_norm": 2.470400998323194, + "learning_rate": 1.6151495289126573e-07, + "loss": 0.262, + "step": 41855 + }, + { + "epoch": 0.9213295179396874, + "grad_norm": 1.5609183340084443, + "learning_rate": 1.6142509681598806e-07, + "loss": 0.3084, + "step": 41856 + }, + { + "epoch": 0.9213515298261061, + "grad_norm": 2.510380113002165, + "learning_rate": 1.6133526533250566e-07, + "loss": 0.3804, + "step": 41857 + }, + { + "epoch": 0.9213735417125247, + "grad_norm": 1.6297440479572234, + "learning_rate": 1.6124545844127593e-07, + "loss": 0.3735, + "step": 41858 + }, + { + "epoch": 0.9213955535989434, + "grad_norm": 1.6707574962618423, + "learning_rate": 1.6115567614275518e-07, + "loss": 0.2936, + "step": 41859 + }, + { + "epoch": 0.9214175654853621, + "grad_norm": 1.5530858940213668, + "learning_rate": 1.6106591843739917e-07, + "loss": 0.2265, + "step": 41860 + }, + { + "epoch": 0.9214395773717807, + "grad_norm": 1.5503377249460533, + "learning_rate": 1.6097618532566416e-07, + "loss": 0.2561, + "step": 41861 + }, + { + "epoch": 0.9214615892581994, + "grad_norm": 1.7556265756231946, + "learning_rate": 1.6088647680800705e-07, + "loss": 0.3667, + "step": 41862 + }, + { + "epoch": 0.9214836011446181, + "grad_norm": 1.5142921700609737, + "learning_rate": 1.6079679288488248e-07, + "loss": 0.2323, + "step": 41863 + }, + { + "epoch": 0.9215056130310367, + "grad_norm": 0.60506596747102, + "learning_rate": 1.6070713355674728e-07, + "loss": 0.4241, + "step": 41864 + }, + { + "epoch": 0.9215276249174554, + "grad_norm": 4.292705307647479, + "learning_rate": 1.6061749882405664e-07, + "loss": 0.1953, + "step": 41865 + }, + { + "epoch": 0.9215496368038741, + "grad_norm": 2.334673240824749, + "learning_rate": 1.6052788868726632e-07, + "loss": 0.3016, + "step": 41866 + }, + { + "epoch": 0.9215716486902927, + "grad_norm": 0.6295018415828347, + "learning_rate": 1.6043830314683206e-07, + "loss": 0.4726, + "step": 41867 + }, + { + "epoch": 0.9215936605767114, + "grad_norm": 1.701755260820907, + "learning_rate": 1.6034874220320852e-07, + "loss": 0.2634, + "step": 41868 + }, + { + "epoch": 0.9216156724631301, + "grad_norm": 2.5696754629815906, + "learning_rate": 1.6025920585685195e-07, + "loss": 0.2475, + "step": 41869 + }, + { + "epoch": 0.9216376843495487, + "grad_norm": 1.4589567099165375, + "learning_rate": 1.6016969410821538e-07, + "loss": 0.2737, + "step": 41870 + }, + { + "epoch": 0.9216596962359674, + "grad_norm": 1.6518522746366646, + "learning_rate": 1.6008020695775616e-07, + "loss": 0.2996, + "step": 41871 + }, + { + "epoch": 0.921681708122386, + "grad_norm": 2.9646795539482187, + "learning_rate": 1.5999074440592732e-07, + "loss": 0.3006, + "step": 41872 + }, + { + "epoch": 0.9217037200088047, + "grad_norm": 1.4451612840322523, + "learning_rate": 1.599013064531846e-07, + "loss": 0.32, + "step": 41873 + }, + { + "epoch": 0.9217257318952234, + "grad_norm": 2.284497466394719, + "learning_rate": 1.59811893099982e-07, + "loss": 0.3028, + "step": 41874 + }, + { + "epoch": 0.921747743781642, + "grad_norm": 0.5962427420277047, + "learning_rate": 1.5972250434677426e-07, + "loss": 0.4542, + "step": 41875 + }, + { + "epoch": 0.9217697556680607, + "grad_norm": 1.8114111091561353, + "learning_rate": 1.5963314019401543e-07, + "loss": 0.3505, + "step": 41876 + }, + { + "epoch": 0.9217917675544794, + "grad_norm": 1.6827411525164917, + "learning_rate": 1.5954380064216014e-07, + "loss": 0.278, + "step": 41877 + }, + { + "epoch": 0.921813779440898, + "grad_norm": 1.431659505734372, + "learning_rate": 1.5945448569166133e-07, + "loss": 0.2728, + "step": 41878 + }, + { + "epoch": 0.9218357913273167, + "grad_norm": 2.447804065682395, + "learning_rate": 1.5936519534297477e-07, + "loss": 0.2871, + "step": 41879 + }, + { + "epoch": 0.9218578032137354, + "grad_norm": 0.6196688774382381, + "learning_rate": 1.5927592959655293e-07, + "loss": 0.4521, + "step": 41880 + }, + { + "epoch": 0.921879815100154, + "grad_norm": 1.6940012539584537, + "learning_rate": 1.591866884528498e-07, + "loss": 0.3456, + "step": 41881 + }, + { + "epoch": 0.9219018269865727, + "grad_norm": 1.9894757480459546, + "learning_rate": 1.5909747191231894e-07, + "loss": 0.3288, + "step": 41882 + }, + { + "epoch": 0.9219238388729915, + "grad_norm": 0.6239620425207609, + "learning_rate": 1.5900827997541336e-07, + "loss": 0.4518, + "step": 41883 + }, + { + "epoch": 0.92194585075941, + "grad_norm": 1.6887245262299937, + "learning_rate": 1.5891911264258818e-07, + "loss": 0.2549, + "step": 41884 + }, + { + "epoch": 0.9219678626458288, + "grad_norm": 0.6149006422862265, + "learning_rate": 1.5882996991429477e-07, + "loss": 0.4546, + "step": 41885 + }, + { + "epoch": 0.9219898745322475, + "grad_norm": 1.7400967875570992, + "learning_rate": 1.587408517909861e-07, + "loss": 0.3106, + "step": 41886 + }, + { + "epoch": 0.922011886418666, + "grad_norm": 2.970443415685658, + "learning_rate": 1.5865175827311564e-07, + "loss": 0.3546, + "step": 41887 + }, + { + "epoch": 0.9220338983050848, + "grad_norm": 1.990933517007661, + "learning_rate": 1.5856268936113583e-07, + "loss": 0.2738, + "step": 41888 + }, + { + "epoch": 0.9220559101915035, + "grad_norm": 1.4395888940025083, + "learning_rate": 1.5847364505550078e-07, + "loss": 0.2482, + "step": 41889 + }, + { + "epoch": 0.922077922077922, + "grad_norm": 1.3994860118202046, + "learning_rate": 1.5838462535666232e-07, + "loss": 0.271, + "step": 41890 + }, + { + "epoch": 0.9220999339643408, + "grad_norm": 0.6047525489927152, + "learning_rate": 1.5829563026507177e-07, + "loss": 0.4454, + "step": 41891 + }, + { + "epoch": 0.9221219458507595, + "grad_norm": 1.5375265096763726, + "learning_rate": 1.582066597811821e-07, + "loss": 0.3487, + "step": 41892 + }, + { + "epoch": 0.922143957737178, + "grad_norm": 1.7355985267515153, + "learning_rate": 1.5811771390544628e-07, + "loss": 0.3009, + "step": 41893 + }, + { + "epoch": 0.9221659696235968, + "grad_norm": 2.555410499565271, + "learning_rate": 1.5802879263831562e-07, + "loss": 0.265, + "step": 41894 + }, + { + "epoch": 0.9221879815100154, + "grad_norm": 2.6310571895898036, + "learning_rate": 1.579398959802425e-07, + "loss": 0.3109, + "step": 41895 + }, + { + "epoch": 0.9222099933964341, + "grad_norm": 1.4279584973695905, + "learning_rate": 1.5785102393167828e-07, + "loss": 0.2577, + "step": 41896 + }, + { + "epoch": 0.9222320052828528, + "grad_norm": 1.79580731892206, + "learning_rate": 1.5776217649307425e-07, + "loss": 0.2943, + "step": 41897 + }, + { + "epoch": 0.9222540171692714, + "grad_norm": 2.2015310615140313, + "learning_rate": 1.576733536648828e-07, + "loss": 0.2391, + "step": 41898 + }, + { + "epoch": 0.9222760290556901, + "grad_norm": 0.610976181720403, + "learning_rate": 1.575845554475558e-07, + "loss": 0.4391, + "step": 41899 + }, + { + "epoch": 0.9222980409421088, + "grad_norm": 0.6386112742386076, + "learning_rate": 1.5749578184154346e-07, + "loss": 0.4594, + "step": 41900 + }, + { + "epoch": 0.9223200528285274, + "grad_norm": 1.574360172647619, + "learning_rate": 1.5740703284729707e-07, + "loss": 0.2882, + "step": 41901 + }, + { + "epoch": 0.9223420647149461, + "grad_norm": 1.6588095795918762, + "learning_rate": 1.5731830846526796e-07, + "loss": 0.2585, + "step": 41902 + }, + { + "epoch": 0.9223640766013648, + "grad_norm": 0.6200831152081276, + "learning_rate": 1.5722960869590797e-07, + "loss": 0.4464, + "step": 41903 + }, + { + "epoch": 0.9223860884877834, + "grad_norm": 1.9002223389485835, + "learning_rate": 1.571409335396662e-07, + "loss": 0.2724, + "step": 41904 + }, + { + "epoch": 0.9224081003742021, + "grad_norm": 1.6625105408712997, + "learning_rate": 1.5705228299699448e-07, + "loss": 0.1651, + "step": 41905 + }, + { + "epoch": 0.9224301122606208, + "grad_norm": 1.4547591503840125, + "learning_rate": 1.5696365706834304e-07, + "loss": 0.2676, + "step": 41906 + }, + { + "epoch": 0.9224521241470394, + "grad_norm": 1.4605798579534115, + "learning_rate": 1.568750557541626e-07, + "loss": 0.2849, + "step": 41907 + }, + { + "epoch": 0.9224741360334581, + "grad_norm": 1.29246986498575, + "learning_rate": 1.567864790549023e-07, + "loss": 0.2561, + "step": 41908 + }, + { + "epoch": 0.9224961479198768, + "grad_norm": 1.544947817839243, + "learning_rate": 1.5669792697101394e-07, + "loss": 0.2535, + "step": 41909 + }, + { + "epoch": 0.9225181598062954, + "grad_norm": 1.4468167756049828, + "learning_rate": 1.566093995029472e-07, + "loss": 0.2427, + "step": 41910 + }, + { + "epoch": 0.9225401716927141, + "grad_norm": 1.7303289982545955, + "learning_rate": 1.5652089665115112e-07, + "loss": 0.2172, + "step": 41911 + }, + { + "epoch": 0.9225621835791328, + "grad_norm": 2.260636252308009, + "learning_rate": 1.5643241841607647e-07, + "loss": 0.3201, + "step": 41912 + }, + { + "epoch": 0.9225841954655514, + "grad_norm": 0.6150630061643072, + "learning_rate": 1.5634396479817181e-07, + "loss": 0.4356, + "step": 41913 + }, + { + "epoch": 0.9226062073519701, + "grad_norm": 2.0477972317933215, + "learning_rate": 1.5625553579788788e-07, + "loss": 0.3787, + "step": 41914 + }, + { + "epoch": 0.9226282192383888, + "grad_norm": 2.937832909314344, + "learning_rate": 1.561671314156743e-07, + "loss": 0.2999, + "step": 41915 + }, + { + "epoch": 0.9226502311248074, + "grad_norm": 1.4433505087520482, + "learning_rate": 1.5607875165197905e-07, + "loss": 0.2382, + "step": 41916 + }, + { + "epoch": 0.9226722430112261, + "grad_norm": 1.645880580958258, + "learning_rate": 1.559903965072529e-07, + "loss": 0.2546, + "step": 41917 + }, + { + "epoch": 0.9226942548976448, + "grad_norm": 1.6382482497316184, + "learning_rate": 1.5590206598194323e-07, + "loss": 0.2319, + "step": 41918 + }, + { + "epoch": 0.9227162667840634, + "grad_norm": 1.6516637797414118, + "learning_rate": 1.5581376007649973e-07, + "loss": 0.2496, + "step": 41919 + }, + { + "epoch": 0.9227382786704821, + "grad_norm": 1.9319496345165559, + "learning_rate": 1.5572547879137146e-07, + "loss": 0.2849, + "step": 41920 + }, + { + "epoch": 0.9227602905569007, + "grad_norm": 1.6116350613674393, + "learning_rate": 1.5563722212700694e-07, + "loss": 0.2229, + "step": 41921 + }, + { + "epoch": 0.9227823024433194, + "grad_norm": 2.2388448779906316, + "learning_rate": 1.5554899008385526e-07, + "loss": 0.2591, + "step": 41922 + }, + { + "epoch": 0.9228043143297381, + "grad_norm": 1.9345915528206576, + "learning_rate": 1.554607826623633e-07, + "loss": 0.3099, + "step": 41923 + }, + { + "epoch": 0.9228263262161567, + "grad_norm": 0.6016470054654477, + "learning_rate": 1.553725998629807e-07, + "loss": 0.445, + "step": 41924 + }, + { + "epoch": 0.9228483381025754, + "grad_norm": 1.7378192364747216, + "learning_rate": 1.5528444168615597e-07, + "loss": 0.2908, + "step": 41925 + }, + { + "epoch": 0.9228703499889941, + "grad_norm": 1.7942502156997535, + "learning_rate": 1.55196308132336e-07, + "loss": 0.2585, + "step": 41926 + }, + { + "epoch": 0.9228923618754127, + "grad_norm": 1.3689969194547327, + "learning_rate": 1.5510819920196928e-07, + "loss": 0.2453, + "step": 41927 + }, + { + "epoch": 0.9229143737618314, + "grad_norm": 0.6273977239737946, + "learning_rate": 1.5502011489550328e-07, + "loss": 0.4472, + "step": 41928 + }, + { + "epoch": 0.9229363856482501, + "grad_norm": 1.4066843537976115, + "learning_rate": 1.5493205521338596e-07, + "loss": 0.3182, + "step": 41929 + }, + { + "epoch": 0.9229583975346687, + "grad_norm": 1.867500260479226, + "learning_rate": 1.5484402015606526e-07, + "loss": 0.2797, + "step": 41930 + }, + { + "epoch": 0.9229804094210874, + "grad_norm": 3.260046903539521, + "learning_rate": 1.5475600972398862e-07, + "loss": 0.3629, + "step": 41931 + }, + { + "epoch": 0.9230024213075061, + "grad_norm": 1.2787581728578508, + "learning_rate": 1.5466802391760238e-07, + "loss": 0.2588, + "step": 41932 + }, + { + "epoch": 0.9230244331939247, + "grad_norm": 2.337718543683649, + "learning_rate": 1.5458006273735392e-07, + "loss": 0.3219, + "step": 41933 + }, + { + "epoch": 0.9230464450803434, + "grad_norm": 0.6288204040541898, + "learning_rate": 1.5449212618369126e-07, + "loss": 0.4778, + "step": 41934 + }, + { + "epoch": 0.9230684569667621, + "grad_norm": 1.5833166661015197, + "learning_rate": 1.544042142570612e-07, + "loss": 0.2694, + "step": 41935 + }, + { + "epoch": 0.9230904688531807, + "grad_norm": 1.983594042182094, + "learning_rate": 1.543163269579101e-07, + "loss": 0.3137, + "step": 41936 + }, + { + "epoch": 0.9231124807395994, + "grad_norm": 1.9117005759437875, + "learning_rate": 1.5422846428668369e-07, + "loss": 0.3449, + "step": 41937 + }, + { + "epoch": 0.9231344926260181, + "grad_norm": 1.5718108356136504, + "learning_rate": 1.5414062624382997e-07, + "loss": 0.2562, + "step": 41938 + }, + { + "epoch": 0.9231565045124367, + "grad_norm": 1.6217609169305065, + "learning_rate": 1.5405281282979468e-07, + "loss": 0.287, + "step": 41939 + }, + { + "epoch": 0.9231785163988554, + "grad_norm": 0.6476008720681637, + "learning_rate": 1.5396502404502467e-07, + "loss": 0.4641, + "step": 41940 + }, + { + "epoch": 0.9232005282852741, + "grad_norm": 0.6274696086204045, + "learning_rate": 1.5387725988996628e-07, + "loss": 0.4486, + "step": 41941 + }, + { + "epoch": 0.9232225401716927, + "grad_norm": 1.773631030820158, + "learning_rate": 1.5378952036506467e-07, + "loss": 0.3807, + "step": 41942 + }, + { + "epoch": 0.9232445520581114, + "grad_norm": 2.2739180697868875, + "learning_rate": 1.5370180547076618e-07, + "loss": 0.2784, + "step": 41943 + }, + { + "epoch": 0.92326656394453, + "grad_norm": 3.6791407053233858, + "learning_rate": 1.53614115207516e-07, + "loss": 0.2109, + "step": 41944 + }, + { + "epoch": 0.9232885758309487, + "grad_norm": 2.9352795565569116, + "learning_rate": 1.5352644957576102e-07, + "loss": 0.3447, + "step": 41945 + }, + { + "epoch": 0.9233105877173674, + "grad_norm": 0.644112448336114, + "learning_rate": 1.534388085759464e-07, + "loss": 0.4368, + "step": 41946 + }, + { + "epoch": 0.923332599603786, + "grad_norm": 0.610220493694745, + "learning_rate": 1.533511922085168e-07, + "loss": 0.4434, + "step": 41947 + }, + { + "epoch": 0.9233546114902047, + "grad_norm": 1.5168237689887007, + "learning_rate": 1.5326360047391908e-07, + "loss": 0.2237, + "step": 41948 + }, + { + "epoch": 0.9233766233766234, + "grad_norm": 1.7891878606678702, + "learning_rate": 1.5317603337259622e-07, + "loss": 0.2931, + "step": 41949 + }, + { + "epoch": 0.923398635263042, + "grad_norm": 1.9464660242402017, + "learning_rate": 1.5308849090499513e-07, + "loss": 0.398, + "step": 41950 + }, + { + "epoch": 0.9234206471494607, + "grad_norm": 1.5515940742850802, + "learning_rate": 1.5300097307156038e-07, + "loss": 0.278, + "step": 41951 + }, + { + "epoch": 0.9234426590358794, + "grad_norm": 1.761325360063671, + "learning_rate": 1.529134798727361e-07, + "loss": 0.2773, + "step": 41952 + }, + { + "epoch": 0.923464670922298, + "grad_norm": 1.693899946928075, + "learning_rate": 1.5282601130896745e-07, + "loss": 0.3052, + "step": 41953 + }, + { + "epoch": 0.9234866828087167, + "grad_norm": 0.6104016318340706, + "learning_rate": 1.5273856738069914e-07, + "loss": 0.4461, + "step": 41954 + }, + { + "epoch": 0.9235086946951354, + "grad_norm": 1.5205128484181865, + "learning_rate": 1.5265114808837467e-07, + "loss": 0.2549, + "step": 41955 + }, + { + "epoch": 0.923530706581554, + "grad_norm": 2.093180416405618, + "learning_rate": 1.525637534324398e-07, + "loss": 0.299, + "step": 41956 + }, + { + "epoch": 0.9235527184679727, + "grad_norm": 1.591381842918454, + "learning_rate": 1.5247638341333747e-07, + "loss": 0.2301, + "step": 41957 + }, + { + "epoch": 0.9235747303543914, + "grad_norm": 4.31956816562462, + "learning_rate": 1.5238903803151294e-07, + "loss": 0.2305, + "step": 41958 + }, + { + "epoch": 0.92359674224081, + "grad_norm": 2.610780243972022, + "learning_rate": 1.5230171728740862e-07, + "loss": 0.1479, + "step": 41959 + }, + { + "epoch": 0.9236187541272287, + "grad_norm": 1.456542973814593, + "learning_rate": 1.5221442118146911e-07, + "loss": 0.317, + "step": 41960 + }, + { + "epoch": 0.9236407660136474, + "grad_norm": 1.4417373861744738, + "learning_rate": 1.5212714971413857e-07, + "loss": 0.1896, + "step": 41961 + }, + { + "epoch": 0.923662777900066, + "grad_norm": 1.744445379606445, + "learning_rate": 1.5203990288586045e-07, + "loss": 0.2673, + "step": 41962 + }, + { + "epoch": 0.9236847897864847, + "grad_norm": 1.5498761338032152, + "learning_rate": 1.5195268069707724e-07, + "loss": 0.2547, + "step": 41963 + }, + { + "epoch": 0.9237068016729034, + "grad_norm": 5.412803000999991, + "learning_rate": 1.5186548314823246e-07, + "loss": 0.3157, + "step": 41964 + }, + { + "epoch": 0.923728813559322, + "grad_norm": 1.6099832147976219, + "learning_rate": 1.5177831023976963e-07, + "loss": 0.2632, + "step": 41965 + }, + { + "epoch": 0.9237508254457407, + "grad_norm": 1.5006844500344012, + "learning_rate": 1.5169116197213228e-07, + "loss": 0.2496, + "step": 41966 + }, + { + "epoch": 0.9237728373321593, + "grad_norm": 1.5449030235864651, + "learning_rate": 1.5160403834576286e-07, + "loss": 0.3664, + "step": 41967 + }, + { + "epoch": 0.923794849218578, + "grad_norm": 2.12065183315999, + "learning_rate": 1.515169393611038e-07, + "loss": 0.3905, + "step": 41968 + }, + { + "epoch": 0.9238168611049967, + "grad_norm": 2.038844065424539, + "learning_rate": 1.5142986501859802e-07, + "loss": 0.3109, + "step": 41969 + }, + { + "epoch": 0.9238388729914153, + "grad_norm": 2.230860344130723, + "learning_rate": 1.5134281531868855e-07, + "loss": 0.3906, + "step": 41970 + }, + { + "epoch": 0.923860884877834, + "grad_norm": 1.7502256431402137, + "learning_rate": 1.512557902618178e-07, + "loss": 0.2979, + "step": 41971 + }, + { + "epoch": 0.9238828967642527, + "grad_norm": 1.4423430982901138, + "learning_rate": 1.5116878984842764e-07, + "loss": 0.244, + "step": 41972 + }, + { + "epoch": 0.9239049086506713, + "grad_norm": 1.604097771562497, + "learning_rate": 1.510818140789594e-07, + "loss": 0.232, + "step": 41973 + }, + { + "epoch": 0.92392692053709, + "grad_norm": 3.2338007478031012, + "learning_rate": 1.5099486295385713e-07, + "loss": 0.2479, + "step": 41974 + }, + { + "epoch": 0.9239489324235087, + "grad_norm": 1.3103924497015575, + "learning_rate": 1.509079364735605e-07, + "loss": 0.2136, + "step": 41975 + }, + { + "epoch": 0.9239709443099273, + "grad_norm": 2.309006067208585, + "learning_rate": 1.5082103463851362e-07, + "loss": 0.2664, + "step": 41976 + }, + { + "epoch": 0.923992956196346, + "grad_norm": 2.371388756955016, + "learning_rate": 1.5073415744915666e-07, + "loss": 0.381, + "step": 41977 + }, + { + "epoch": 0.9240149680827647, + "grad_norm": 1.5818864561489012, + "learning_rate": 1.506473049059315e-07, + "loss": 0.2878, + "step": 41978 + }, + { + "epoch": 0.9240369799691833, + "grad_norm": 2.0465192756290964, + "learning_rate": 1.5056047700928002e-07, + "loss": 0.2345, + "step": 41979 + }, + { + "epoch": 0.924058991855602, + "grad_norm": 2.2445483701763553, + "learning_rate": 1.5047367375964184e-07, + "loss": 0.1801, + "step": 41980 + }, + { + "epoch": 0.9240810037420207, + "grad_norm": 1.5315825059741806, + "learning_rate": 1.503868951574605e-07, + "loss": 0.2603, + "step": 41981 + }, + { + "epoch": 0.9241030156284393, + "grad_norm": 0.6061130492181145, + "learning_rate": 1.5030014120317626e-07, + "loss": 0.4372, + "step": 41982 + }, + { + "epoch": 0.924125027514858, + "grad_norm": 7.717245408689677, + "learning_rate": 1.502134118972287e-07, + "loss": 0.2173, + "step": 41983 + }, + { + "epoch": 0.9241470394012767, + "grad_norm": 1.867197614510881, + "learning_rate": 1.5012670724006028e-07, + "loss": 0.3506, + "step": 41984 + }, + { + "epoch": 0.9241690512876953, + "grad_norm": 1.5093240241370145, + "learning_rate": 1.5004002723211064e-07, + "loss": 0.3984, + "step": 41985 + }, + { + "epoch": 0.924191063174114, + "grad_norm": 2.1430142820654394, + "learning_rate": 1.499533718738211e-07, + "loss": 0.3288, + "step": 41986 + }, + { + "epoch": 0.9242130750605327, + "grad_norm": 1.564747973828074, + "learning_rate": 1.4986674116563183e-07, + "loss": 0.2088, + "step": 41987 + }, + { + "epoch": 0.9242350869469513, + "grad_norm": 1.9283147933599236, + "learning_rate": 1.497801351079825e-07, + "loss": 0.3714, + "step": 41988 + }, + { + "epoch": 0.92425709883337, + "grad_norm": 2.1563551864326844, + "learning_rate": 1.4969355370131444e-07, + "loss": 0.3272, + "step": 41989 + }, + { + "epoch": 0.9242791107197886, + "grad_norm": 2.0869270587390907, + "learning_rate": 1.4960699694606617e-07, + "loss": 0.231, + "step": 41990 + }, + { + "epoch": 0.9243011226062073, + "grad_norm": 2.1143126352805504, + "learning_rate": 1.4952046484267845e-07, + "loss": 0.2946, + "step": 41991 + }, + { + "epoch": 0.924323134492626, + "grad_norm": 1.7556223222724945, + "learning_rate": 1.4943395739159206e-07, + "loss": 0.2653, + "step": 41992 + }, + { + "epoch": 0.9243451463790446, + "grad_norm": 1.341217935298447, + "learning_rate": 1.4934747459324495e-07, + "loss": 0.2839, + "step": 41993 + }, + { + "epoch": 0.9243671582654633, + "grad_norm": 2.171885431837904, + "learning_rate": 1.492610164480779e-07, + "loss": 0.2896, + "step": 41994 + }, + { + "epoch": 0.924389170151882, + "grad_norm": 0.6258209717797399, + "learning_rate": 1.4917458295652943e-07, + "loss": 0.4624, + "step": 41995 + }, + { + "epoch": 0.9244111820383006, + "grad_norm": 0.6029916524909466, + "learning_rate": 1.4908817411903975e-07, + "loss": 0.4346, + "step": 41996 + }, + { + "epoch": 0.9244331939247193, + "grad_norm": 1.8120360598938783, + "learning_rate": 1.490017899360474e-07, + "loss": 0.2813, + "step": 41997 + }, + { + "epoch": 0.924455205811138, + "grad_norm": 1.5778333544883787, + "learning_rate": 1.4891543040799095e-07, + "loss": 0.2847, + "step": 41998 + }, + { + "epoch": 0.9244772176975566, + "grad_norm": 1.466007858701707, + "learning_rate": 1.488290955353111e-07, + "loss": 0.3266, + "step": 41999 + }, + { + "epoch": 0.9244992295839753, + "grad_norm": 2.322248700840949, + "learning_rate": 1.4874278531844422e-07, + "loss": 0.2697, + "step": 42000 + }, + { + "epoch": 0.924521241470394, + "grad_norm": 2.839634316845161, + "learning_rate": 1.4865649975783047e-07, + "loss": 0.247, + "step": 42001 + }, + { + "epoch": 0.9245432533568126, + "grad_norm": 1.5586535770358265, + "learning_rate": 1.4857023885390898e-07, + "loss": 0.216, + "step": 42002 + }, + { + "epoch": 0.9245652652432313, + "grad_norm": 1.5071862480530802, + "learning_rate": 1.4848400260711714e-07, + "loss": 0.2311, + "step": 42003 + }, + { + "epoch": 0.92458727712965, + "grad_norm": 2.1985355007742773, + "learning_rate": 1.4839779101789297e-07, + "loss": 0.299, + "step": 42004 + }, + { + "epoch": 0.9246092890160686, + "grad_norm": 2.065937100587188, + "learning_rate": 1.48311604086675e-07, + "loss": 0.3263, + "step": 42005 + }, + { + "epoch": 0.9246313009024874, + "grad_norm": 1.6652317441987827, + "learning_rate": 1.482254418139012e-07, + "loss": 0.2769, + "step": 42006 + }, + { + "epoch": 0.9246533127889061, + "grad_norm": 1.819388449882622, + "learning_rate": 1.4813930420001065e-07, + "loss": 0.2537, + "step": 42007 + }, + { + "epoch": 0.9246753246753247, + "grad_norm": 0.6079782237352783, + "learning_rate": 1.4805319124543972e-07, + "loss": 0.4524, + "step": 42008 + }, + { + "epoch": 0.9246973365617434, + "grad_norm": 1.8911477532709267, + "learning_rate": 1.4796710295062634e-07, + "loss": 0.3485, + "step": 42009 + }, + { + "epoch": 0.9247193484481621, + "grad_norm": 1.5431768405183013, + "learning_rate": 1.478810393160085e-07, + "loss": 0.2386, + "step": 42010 + }, + { + "epoch": 0.9247413603345807, + "grad_norm": 3.41897587698149, + "learning_rate": 1.4779500034202199e-07, + "loss": 0.3547, + "step": 42011 + }, + { + "epoch": 0.9247633722209994, + "grad_norm": 0.6368369166404237, + "learning_rate": 1.4770898602910698e-07, + "loss": 0.4492, + "step": 42012 + }, + { + "epoch": 0.924785384107418, + "grad_norm": 1.7175546827583084, + "learning_rate": 1.4762299637769872e-07, + "loss": 0.272, + "step": 42013 + }, + { + "epoch": 0.9248073959938367, + "grad_norm": 1.7348939864567585, + "learning_rate": 1.475370313882346e-07, + "loss": 0.2722, + "step": 42014 + }, + { + "epoch": 0.9248294078802554, + "grad_norm": 0.6064870483966537, + "learning_rate": 1.4745109106115152e-07, + "loss": 0.4354, + "step": 42015 + }, + { + "epoch": 0.924851419766674, + "grad_norm": 1.418866570057622, + "learning_rate": 1.4736517539688578e-07, + "loss": 0.3138, + "step": 42016 + }, + { + "epoch": 0.9248734316530927, + "grad_norm": 1.480225960192156, + "learning_rate": 1.472792843958748e-07, + "loss": 0.2556, + "step": 42017 + }, + { + "epoch": 0.9248954435395114, + "grad_norm": 1.7465308520754501, + "learning_rate": 1.4719341805855546e-07, + "loss": 0.2269, + "step": 42018 + }, + { + "epoch": 0.92491745542593, + "grad_norm": 1.7696647594272699, + "learning_rate": 1.47107576385363e-07, + "loss": 0.354, + "step": 42019 + }, + { + "epoch": 0.9249394673123487, + "grad_norm": 2.8176127098084924, + "learning_rate": 1.470217593767348e-07, + "loss": 0.3403, + "step": 42020 + }, + { + "epoch": 0.9249614791987674, + "grad_norm": 2.268242686322064, + "learning_rate": 1.469359670331061e-07, + "loss": 0.2705, + "step": 42021 + }, + { + "epoch": 0.924983491085186, + "grad_norm": 0.6138948489541136, + "learning_rate": 1.4685019935491319e-07, + "loss": 0.452, + "step": 42022 + }, + { + "epoch": 0.9250055029716047, + "grad_norm": 1.8673146224074455, + "learning_rate": 1.467644563425924e-07, + "loss": 0.2835, + "step": 42023 + }, + { + "epoch": 0.9250275148580234, + "grad_norm": 1.8942137078618648, + "learning_rate": 1.4667873799657895e-07, + "loss": 0.3008, + "step": 42024 + }, + { + "epoch": 0.925049526744442, + "grad_norm": 38.09722311018435, + "learning_rate": 1.465930443173097e-07, + "loss": 0.2666, + "step": 42025 + }, + { + "epoch": 0.9250715386308607, + "grad_norm": 2.300945798593609, + "learning_rate": 1.465073753052182e-07, + "loss": 0.266, + "step": 42026 + }, + { + "epoch": 0.9250935505172794, + "grad_norm": 1.6715674178118964, + "learning_rate": 1.464217309607413e-07, + "loss": 0.2791, + "step": 42027 + }, + { + "epoch": 0.925115562403698, + "grad_norm": 1.4078958915217208, + "learning_rate": 1.4633611128431423e-07, + "loss": 0.3092, + "step": 42028 + }, + { + "epoch": 0.9251375742901167, + "grad_norm": 1.9909520895896546, + "learning_rate": 1.462505162763711e-07, + "loss": 0.3638, + "step": 42029 + }, + { + "epoch": 0.9251595861765354, + "grad_norm": 1.4040928562068, + "learning_rate": 1.4616494593734876e-07, + "loss": 0.2499, + "step": 42030 + }, + { + "epoch": 0.925181598062954, + "grad_norm": 1.3091906347948021, + "learning_rate": 1.460794002676802e-07, + "loss": 0.271, + "step": 42031 + }, + { + "epoch": 0.9252036099493727, + "grad_norm": 2.1116778517441364, + "learning_rate": 1.4599387926780118e-07, + "loss": 0.2834, + "step": 42032 + }, + { + "epoch": 0.9252256218357914, + "grad_norm": 0.6078867025361036, + "learning_rate": 1.4590838293814634e-07, + "loss": 0.4485, + "step": 42033 + }, + { + "epoch": 0.92524763372221, + "grad_norm": 0.5908468350070296, + "learning_rate": 1.4582291127914984e-07, + "loss": 0.4413, + "step": 42034 + }, + { + "epoch": 0.9252696456086287, + "grad_norm": 2.272799604004909, + "learning_rate": 1.457374642912468e-07, + "loss": 0.2844, + "step": 42035 + }, + { + "epoch": 0.9252916574950473, + "grad_norm": 1.915559132185251, + "learning_rate": 1.4565204197487082e-07, + "loss": 0.2822, + "step": 42036 + }, + { + "epoch": 0.925313669381466, + "grad_norm": 0.6324758919687273, + "learning_rate": 1.4556664433045598e-07, + "loss": 0.4308, + "step": 42037 + }, + { + "epoch": 0.9253356812678847, + "grad_norm": 1.8875153644131877, + "learning_rate": 1.4548127135843693e-07, + "loss": 0.3209, + "step": 42038 + }, + { + "epoch": 0.9253576931543033, + "grad_norm": 1.9968009636041175, + "learning_rate": 1.4539592305924664e-07, + "loss": 0.264, + "step": 42039 + }, + { + "epoch": 0.925379705040722, + "grad_norm": 0.6357088234452158, + "learning_rate": 1.4531059943332037e-07, + "loss": 0.4322, + "step": 42040 + }, + { + "epoch": 0.9254017169271407, + "grad_norm": 2.1200457875345995, + "learning_rate": 1.452253004810905e-07, + "loss": 0.2745, + "step": 42041 + }, + { + "epoch": 0.9254237288135593, + "grad_norm": 1.6625214709984786, + "learning_rate": 1.451400262029906e-07, + "loss": 0.2195, + "step": 42042 + }, + { + "epoch": 0.925445740699978, + "grad_norm": 1.2552418118362938, + "learning_rate": 1.4505477659945478e-07, + "loss": 0.2067, + "step": 42043 + }, + { + "epoch": 0.9254677525863967, + "grad_norm": 1.7163553724137561, + "learning_rate": 1.4496955167091653e-07, + "loss": 0.3141, + "step": 42044 + }, + { + "epoch": 0.9254897644728153, + "grad_norm": 1.489324352505724, + "learning_rate": 1.4488435141780775e-07, + "loss": 0.2707, + "step": 42045 + }, + { + "epoch": 0.925511776359234, + "grad_norm": 3.2311171111317245, + "learning_rate": 1.44799175840562e-07, + "loss": 0.3478, + "step": 42046 + }, + { + "epoch": 0.9255337882456527, + "grad_norm": 1.4177987535936425, + "learning_rate": 1.447140249396123e-07, + "loss": 0.2667, + "step": 42047 + }, + { + "epoch": 0.9255558001320713, + "grad_norm": 2.381283142833939, + "learning_rate": 1.4462889871539155e-07, + "loss": 0.2878, + "step": 42048 + }, + { + "epoch": 0.92557781201849, + "grad_norm": 1.4687278877518226, + "learning_rate": 1.4454379716833277e-07, + "loss": 0.2408, + "step": 42049 + }, + { + "epoch": 0.9255998239049087, + "grad_norm": 2.107825227969522, + "learning_rate": 1.4445872029886732e-07, + "loss": 0.2811, + "step": 42050 + }, + { + "epoch": 0.9256218357913273, + "grad_norm": 2.261139302564576, + "learning_rate": 1.4437366810742925e-07, + "loss": 0.2429, + "step": 42051 + }, + { + "epoch": 0.925643847677746, + "grad_norm": 2.8320686014609198, + "learning_rate": 1.4428864059444825e-07, + "loss": 0.2608, + "step": 42052 + }, + { + "epoch": 0.9256658595641647, + "grad_norm": 1.7141512093816837, + "learning_rate": 1.442036377603595e-07, + "loss": 0.3217, + "step": 42053 + }, + { + "epoch": 0.9256878714505833, + "grad_norm": 1.9703275147335284, + "learning_rate": 1.4411865960559325e-07, + "loss": 0.204, + "step": 42054 + }, + { + "epoch": 0.925709883337002, + "grad_norm": 1.7280793653099336, + "learning_rate": 1.4403370613058132e-07, + "loss": 0.3355, + "step": 42055 + }, + { + "epoch": 0.9257318952234207, + "grad_norm": 1.3723253421053556, + "learning_rate": 1.4394877733575674e-07, + "loss": 0.2166, + "step": 42056 + }, + { + "epoch": 0.9257539071098393, + "grad_norm": 2.0799266981210063, + "learning_rate": 1.438638732215497e-07, + "loss": 0.3499, + "step": 42057 + }, + { + "epoch": 0.925775918996258, + "grad_norm": 2.0797874274184003, + "learning_rate": 1.437789937883921e-07, + "loss": 0.2816, + "step": 42058 + }, + { + "epoch": 0.9257979308826767, + "grad_norm": 3.1336908011883087, + "learning_rate": 1.4369413903671635e-07, + "loss": 0.3013, + "step": 42059 + }, + { + "epoch": 0.9258199427690953, + "grad_norm": 1.8953717349840125, + "learning_rate": 1.436093089669527e-07, + "loss": 0.3417, + "step": 42060 + }, + { + "epoch": 0.925841954655514, + "grad_norm": 2.454257171970915, + "learning_rate": 1.435245035795324e-07, + "loss": 0.2102, + "step": 42061 + }, + { + "epoch": 0.9258639665419326, + "grad_norm": 2.051029181795612, + "learning_rate": 1.4343972287488682e-07, + "loss": 0.3636, + "step": 42062 + }, + { + "epoch": 0.9258859784283513, + "grad_norm": 0.6490064283294379, + "learning_rate": 1.433549668534462e-07, + "loss": 0.4478, + "step": 42063 + }, + { + "epoch": 0.92590799031477, + "grad_norm": 1.9335373987492455, + "learning_rate": 1.4327023551564234e-07, + "loss": 0.3052, + "step": 42064 + }, + { + "epoch": 0.9259300022011886, + "grad_norm": 0.6187733312288561, + "learning_rate": 1.4318552886190494e-07, + "loss": 0.4301, + "step": 42065 + }, + { + "epoch": 0.9259520140876073, + "grad_norm": 2.0356362346396937, + "learning_rate": 1.4310084689266534e-07, + "loss": 0.3243, + "step": 42066 + }, + { + "epoch": 0.925974025974026, + "grad_norm": 0.585239622856331, + "learning_rate": 1.4301618960835263e-07, + "loss": 0.4205, + "step": 42067 + }, + { + "epoch": 0.9259960378604446, + "grad_norm": 1.873724117291573, + "learning_rate": 1.429315570093981e-07, + "loss": 0.3378, + "step": 42068 + }, + { + "epoch": 0.9260180497468633, + "grad_norm": 1.4700164819780472, + "learning_rate": 1.4284694909623253e-07, + "loss": 0.2212, + "step": 42069 + }, + { + "epoch": 0.926040061633282, + "grad_norm": 1.56098797910051, + "learning_rate": 1.427623658692845e-07, + "loss": 0.2368, + "step": 42070 + }, + { + "epoch": 0.9260620735197006, + "grad_norm": 2.310343870465518, + "learning_rate": 1.4267780732898473e-07, + "loss": 0.2744, + "step": 42071 + }, + { + "epoch": 0.9260840854061193, + "grad_norm": 1.4386972076547357, + "learning_rate": 1.4259327347576236e-07, + "loss": 0.262, + "step": 42072 + }, + { + "epoch": 0.926106097292538, + "grad_norm": 0.59888055090717, + "learning_rate": 1.4250876431004756e-07, + "loss": 0.4138, + "step": 42073 + }, + { + "epoch": 0.9261281091789566, + "grad_norm": 2.310495041452422, + "learning_rate": 1.4242427983227058e-07, + "loss": 0.2828, + "step": 42074 + }, + { + "epoch": 0.9261501210653753, + "grad_norm": 1.6187816866220377, + "learning_rate": 1.4233982004285883e-07, + "loss": 0.2271, + "step": 42075 + }, + { + "epoch": 0.926172132951794, + "grad_norm": 3.3468830105103664, + "learning_rate": 1.4225538494224367e-07, + "loss": 0.3148, + "step": 42076 + }, + { + "epoch": 0.9261941448382126, + "grad_norm": 1.8023489754718043, + "learning_rate": 1.4217097453085248e-07, + "loss": 0.292, + "step": 42077 + }, + { + "epoch": 0.9262161567246313, + "grad_norm": 1.760128038345942, + "learning_rate": 1.4208658880911497e-07, + "loss": 0.2949, + "step": 42078 + }, + { + "epoch": 0.92623816861105, + "grad_norm": 1.870136629490156, + "learning_rate": 1.4200222777746076e-07, + "loss": 0.3509, + "step": 42079 + }, + { + "epoch": 0.9262601804974686, + "grad_norm": 1.3440173579032657, + "learning_rate": 1.419178914363173e-07, + "loss": 0.2534, + "step": 42080 + }, + { + "epoch": 0.9262821923838873, + "grad_norm": 1.825278510720639, + "learning_rate": 1.4183357978611424e-07, + "loss": 0.3088, + "step": 42081 + }, + { + "epoch": 0.926304204270306, + "grad_norm": 2.2912091682412368, + "learning_rate": 1.4174929282728012e-07, + "loss": 0.2891, + "step": 42082 + }, + { + "epoch": 0.9263262161567246, + "grad_norm": 2.1967616439887174, + "learning_rate": 1.4166503056024183e-07, + "loss": 0.2057, + "step": 42083 + }, + { + "epoch": 0.9263482280431433, + "grad_norm": 3.011101374952901, + "learning_rate": 1.415807929854296e-07, + "loss": 0.2725, + "step": 42084 + }, + { + "epoch": 0.9263702399295619, + "grad_norm": 1.3684146927870962, + "learning_rate": 1.4149658010327084e-07, + "loss": 0.2228, + "step": 42085 + }, + { + "epoch": 0.9263922518159806, + "grad_norm": 0.603060164083987, + "learning_rate": 1.4141239191419298e-07, + "loss": 0.4298, + "step": 42086 + }, + { + "epoch": 0.9264142637023993, + "grad_norm": 3.8896206492024183, + "learning_rate": 1.4132822841862458e-07, + "loss": 0.3421, + "step": 42087 + }, + { + "epoch": 0.9264362755888179, + "grad_norm": 0.6280340104996068, + "learning_rate": 1.412440896169931e-07, + "loss": 0.4434, + "step": 42088 + }, + { + "epoch": 0.9264582874752366, + "grad_norm": 1.5670374582375304, + "learning_rate": 1.4115997550972594e-07, + "loss": 0.283, + "step": 42089 + }, + { + "epoch": 0.9264802993616553, + "grad_norm": 3.952293141321721, + "learning_rate": 1.410758860972511e-07, + "loss": 0.2736, + "step": 42090 + }, + { + "epoch": 0.9265023112480739, + "grad_norm": 1.4818913147846255, + "learning_rate": 1.4099182137999545e-07, + "loss": 0.286, + "step": 42091 + }, + { + "epoch": 0.9265243231344926, + "grad_norm": 6.215573217521199, + "learning_rate": 1.40907781358387e-07, + "loss": 0.3006, + "step": 42092 + }, + { + "epoch": 0.9265463350209113, + "grad_norm": 4.1214312225612435, + "learning_rate": 1.4082376603285208e-07, + "loss": 0.2564, + "step": 42093 + }, + { + "epoch": 0.9265683469073299, + "grad_norm": 1.5587116101369067, + "learning_rate": 1.4073977540381812e-07, + "loss": 0.2587, + "step": 42094 + }, + { + "epoch": 0.9265903587937486, + "grad_norm": 10.969429790565568, + "learning_rate": 1.4065580947171197e-07, + "loss": 0.2017, + "step": 42095 + }, + { + "epoch": 0.9266123706801673, + "grad_norm": 1.6334739130707807, + "learning_rate": 1.4057186823696056e-07, + "loss": 0.2178, + "step": 42096 + }, + { + "epoch": 0.9266343825665859, + "grad_norm": 1.669763910921298, + "learning_rate": 1.4048795169999019e-07, + "loss": 0.2173, + "step": 42097 + }, + { + "epoch": 0.9266563944530046, + "grad_norm": 3.0258473316783636, + "learning_rate": 1.4040405986122717e-07, + "loss": 0.3015, + "step": 42098 + }, + { + "epoch": 0.9266784063394233, + "grad_norm": 2.6880786412299416, + "learning_rate": 1.4032019272109787e-07, + "loss": 0.3831, + "step": 42099 + }, + { + "epoch": 0.9267004182258419, + "grad_norm": 0.6328530934259361, + "learning_rate": 1.4023635028002969e-07, + "loss": 0.4662, + "step": 42100 + }, + { + "epoch": 0.9267224301122606, + "grad_norm": 1.4431394052683968, + "learning_rate": 1.401525325384473e-07, + "loss": 0.2476, + "step": 42101 + }, + { + "epoch": 0.9267444419986793, + "grad_norm": 2.806954479814612, + "learning_rate": 1.4006873949677759e-07, + "loss": 0.1919, + "step": 42102 + }, + { + "epoch": 0.9267664538850979, + "grad_norm": 3.5240878016479753, + "learning_rate": 1.3998497115544573e-07, + "loss": 0.2815, + "step": 42103 + }, + { + "epoch": 0.9267884657715166, + "grad_norm": 2.016277394077198, + "learning_rate": 1.3990122751487812e-07, + "loss": 0.3077, + "step": 42104 + }, + { + "epoch": 0.9268104776579353, + "grad_norm": 2.4638515234371114, + "learning_rate": 1.3981750857550047e-07, + "loss": 0.3182, + "step": 42105 + }, + { + "epoch": 0.9268324895443539, + "grad_norm": 2.138420143311547, + "learning_rate": 1.3973381433773746e-07, + "loss": 0.3138, + "step": 42106 + }, + { + "epoch": 0.9268545014307726, + "grad_norm": 2.212997232358248, + "learning_rate": 1.3965014480201545e-07, + "loss": 0.3134, + "step": 42107 + }, + { + "epoch": 0.9268765133171912, + "grad_norm": 1.698665806509268, + "learning_rate": 1.395664999687585e-07, + "loss": 0.317, + "step": 42108 + }, + { + "epoch": 0.9268985252036099, + "grad_norm": 1.5506024746002876, + "learning_rate": 1.3948287983839238e-07, + "loss": 0.2787, + "step": 42109 + }, + { + "epoch": 0.9269205370900286, + "grad_norm": 1.3627845588398724, + "learning_rate": 1.3939928441134288e-07, + "loss": 0.2506, + "step": 42110 + }, + { + "epoch": 0.9269425489764472, + "grad_norm": 4.116619933803781, + "learning_rate": 1.3931571368803353e-07, + "loss": 0.3405, + "step": 42111 + }, + { + "epoch": 0.9269645608628659, + "grad_norm": 1.4614800099034868, + "learning_rate": 1.3923216766889015e-07, + "loss": 0.2135, + "step": 42112 + }, + { + "epoch": 0.9269865727492846, + "grad_norm": 1.4230210270882457, + "learning_rate": 1.3914864635433623e-07, + "loss": 0.2402, + "step": 42113 + }, + { + "epoch": 0.9270085846357032, + "grad_norm": 2.77656363858083, + "learning_rate": 1.3906514974479702e-07, + "loss": 0.2871, + "step": 42114 + }, + { + "epoch": 0.9270305965221219, + "grad_norm": 1.6712445621159642, + "learning_rate": 1.389816778406977e-07, + "loss": 0.2392, + "step": 42115 + }, + { + "epoch": 0.9270526084085406, + "grad_norm": 0.6180760868151585, + "learning_rate": 1.3889823064246021e-07, + "loss": 0.4517, + "step": 42116 + }, + { + "epoch": 0.9270746202949592, + "grad_norm": 11.97114189298748, + "learning_rate": 1.3881480815051085e-07, + "loss": 0.3157, + "step": 42117 + }, + { + "epoch": 0.9270966321813779, + "grad_norm": 2.507830379566449, + "learning_rate": 1.3873141036527314e-07, + "loss": 0.2659, + "step": 42118 + }, + { + "epoch": 0.9271186440677966, + "grad_norm": 9.060615832947443, + "learning_rate": 1.38648037287169e-07, + "loss": 0.2739, + "step": 42119 + }, + { + "epoch": 0.9271406559542152, + "grad_norm": 2.0405341944688864, + "learning_rate": 1.3856468891662534e-07, + "loss": 0.279, + "step": 42120 + }, + { + "epoch": 0.927162667840634, + "grad_norm": 1.946332269270779, + "learning_rate": 1.3848136525406397e-07, + "loss": 0.2696, + "step": 42121 + }, + { + "epoch": 0.9271846797270527, + "grad_norm": 1.9193891202504485, + "learning_rate": 1.3839806629990792e-07, + "loss": 0.2645, + "step": 42122 + }, + { + "epoch": 0.9272066916134712, + "grad_norm": 0.6501298735838744, + "learning_rate": 1.3831479205458187e-07, + "loss": 0.4731, + "step": 42123 + }, + { + "epoch": 0.92722870349989, + "grad_norm": 1.6506821140620027, + "learning_rate": 1.3823154251850767e-07, + "loss": 0.2501, + "step": 42124 + }, + { + "epoch": 0.9272507153863087, + "grad_norm": 0.6205105299284497, + "learning_rate": 1.3814831769210947e-07, + "loss": 0.4471, + "step": 42125 + }, + { + "epoch": 0.9272727272727272, + "grad_norm": 1.3925765493318296, + "learning_rate": 1.3806511757581076e-07, + "loss": 0.2171, + "step": 42126 + }, + { + "epoch": 0.927294739159146, + "grad_norm": 1.8719998569629208, + "learning_rate": 1.3798194217003236e-07, + "loss": 0.301, + "step": 42127 + }, + { + "epoch": 0.9273167510455647, + "grad_norm": 1.921322802518885, + "learning_rate": 1.3789879147519948e-07, + "loss": 0.2975, + "step": 42128 + }, + { + "epoch": 0.9273387629319833, + "grad_norm": 1.197543889965581, + "learning_rate": 1.3781566549173232e-07, + "loss": 0.2834, + "step": 42129 + }, + { + "epoch": 0.927360774818402, + "grad_norm": 1.9279291993343892, + "learning_rate": 1.37732564220055e-07, + "loss": 0.3055, + "step": 42130 + }, + { + "epoch": 0.9273827867048205, + "grad_norm": 0.6086217600554862, + "learning_rate": 1.3764948766058994e-07, + "loss": 0.4458, + "step": 42131 + }, + { + "epoch": 0.9274047985912393, + "grad_norm": 0.6336770930669552, + "learning_rate": 1.3756643581375794e-07, + "loss": 0.4538, + "step": 42132 + }, + { + "epoch": 0.927426810477658, + "grad_norm": 2.3727310785111397, + "learning_rate": 1.374834086799831e-07, + "loss": 0.2586, + "step": 42133 + }, + { + "epoch": 0.9274488223640766, + "grad_norm": 2.0245398737751046, + "learning_rate": 1.3740040625968565e-07, + "loss": 0.356, + "step": 42134 + }, + { + "epoch": 0.9274708342504953, + "grad_norm": 1.9966229274002498, + "learning_rate": 1.37317428553288e-07, + "loss": 0.275, + "step": 42135 + }, + { + "epoch": 0.927492846136914, + "grad_norm": 1.7751374818451575, + "learning_rate": 1.3723447556121205e-07, + "loss": 0.2971, + "step": 42136 + }, + { + "epoch": 0.9275148580233326, + "grad_norm": 0.6138031847077167, + "learning_rate": 1.3715154728387968e-07, + "loss": 0.4358, + "step": 42137 + }, + { + "epoch": 0.9275368699097513, + "grad_norm": 2.3730172718954634, + "learning_rate": 1.3706864372171225e-07, + "loss": 0.2799, + "step": 42138 + }, + { + "epoch": 0.92755888179617, + "grad_norm": 4.259567773664954, + "learning_rate": 1.3698576487512993e-07, + "loss": 0.2144, + "step": 42139 + }, + { + "epoch": 0.9275808936825886, + "grad_norm": 0.628144493817403, + "learning_rate": 1.369029107445552e-07, + "loss": 0.4481, + "step": 42140 + }, + { + "epoch": 0.9276029055690073, + "grad_norm": 2.567720650946255, + "learning_rate": 1.368200813304099e-07, + "loss": 0.2275, + "step": 42141 + }, + { + "epoch": 0.927624917455426, + "grad_norm": 2.830380685885591, + "learning_rate": 1.3673727663311264e-07, + "loss": 0.4101, + "step": 42142 + }, + { + "epoch": 0.9276469293418446, + "grad_norm": 1.2267524211045326, + "learning_rate": 1.3665449665308638e-07, + "loss": 0.2977, + "step": 42143 + }, + { + "epoch": 0.9276689412282633, + "grad_norm": 4.628634914696962, + "learning_rate": 1.365717413907508e-07, + "loss": 0.3031, + "step": 42144 + }, + { + "epoch": 0.927690953114682, + "grad_norm": 1.770565900229636, + "learning_rate": 1.3648901084652667e-07, + "loss": 0.3135, + "step": 42145 + }, + { + "epoch": 0.9277129650011006, + "grad_norm": 1.7241240667216484, + "learning_rate": 1.364063050208353e-07, + "loss": 0.3226, + "step": 42146 + }, + { + "epoch": 0.9277349768875193, + "grad_norm": 1.2119219532645973, + "learning_rate": 1.363236239140958e-07, + "loss": 0.2172, + "step": 42147 + }, + { + "epoch": 0.927756988773938, + "grad_norm": 1.5589021393050173, + "learning_rate": 1.36240967526729e-07, + "loss": 0.2174, + "step": 42148 + }, + { + "epoch": 0.9277790006603566, + "grad_norm": 1.3921195209094488, + "learning_rate": 1.361583358591545e-07, + "loss": 0.2112, + "step": 42149 + }, + { + "epoch": 0.9278010125467753, + "grad_norm": 1.425447305028194, + "learning_rate": 1.3607572891179312e-07, + "loss": 0.2668, + "step": 42150 + }, + { + "epoch": 0.927823024433194, + "grad_norm": 1.755941975935336, + "learning_rate": 1.3599314668506446e-07, + "loss": 0.2819, + "step": 42151 + }, + { + "epoch": 0.9278450363196126, + "grad_norm": 0.6172496900109379, + "learning_rate": 1.3591058917938715e-07, + "loss": 0.4588, + "step": 42152 + }, + { + "epoch": 0.9278670482060313, + "grad_norm": 0.6106237046439211, + "learning_rate": 1.358280563951825e-07, + "loss": 0.4387, + "step": 42153 + }, + { + "epoch": 0.9278890600924499, + "grad_norm": 2.3065536331612573, + "learning_rate": 1.3574554833286958e-07, + "loss": 0.355, + "step": 42154 + }, + { + "epoch": 0.9279110719788686, + "grad_norm": 2.2909073688463715, + "learning_rate": 1.3566306499286585e-07, + "loss": 0.3476, + "step": 42155 + }, + { + "epoch": 0.9279330838652873, + "grad_norm": 0.6179550391259117, + "learning_rate": 1.3558060637559267e-07, + "loss": 0.4574, + "step": 42156 + }, + { + "epoch": 0.9279550957517059, + "grad_norm": 1.8476882783502158, + "learning_rate": 1.3549817248146856e-07, + "loss": 0.3075, + "step": 42157 + }, + { + "epoch": 0.9279771076381246, + "grad_norm": 1.519899355375037, + "learning_rate": 1.3541576331091267e-07, + "loss": 0.2993, + "step": 42158 + }, + { + "epoch": 0.9279991195245433, + "grad_norm": 1.8054452122229034, + "learning_rate": 1.353333788643435e-07, + "loss": 0.2628, + "step": 42159 + }, + { + "epoch": 0.9280211314109619, + "grad_norm": 1.308308200008918, + "learning_rate": 1.3525101914217907e-07, + "loss": 0.2665, + "step": 42160 + }, + { + "epoch": 0.9280431432973806, + "grad_norm": 3.057181096664445, + "learning_rate": 1.351686841448391e-07, + "loss": 0.2605, + "step": 42161 + }, + { + "epoch": 0.9280651551837993, + "grad_norm": 1.6228915280162242, + "learning_rate": 1.3508637387274148e-07, + "loss": 0.2873, + "step": 42162 + }, + { + "epoch": 0.9280871670702179, + "grad_norm": 1.8982645290253608, + "learning_rate": 1.3500408832630484e-07, + "loss": 0.2562, + "step": 42163 + }, + { + "epoch": 0.9281091789566366, + "grad_norm": 1.4722243114123916, + "learning_rate": 1.3492182750594774e-07, + "loss": 0.3463, + "step": 42164 + }, + { + "epoch": 0.9281311908430553, + "grad_norm": 0.649917980668789, + "learning_rate": 1.3483959141208702e-07, + "loss": 0.4647, + "step": 42165 + }, + { + "epoch": 0.9281532027294739, + "grad_norm": 1.5964793150946213, + "learning_rate": 1.3475738004514182e-07, + "loss": 0.3279, + "step": 42166 + }, + { + "epoch": 0.9281752146158926, + "grad_norm": 1.9184081606077454, + "learning_rate": 1.346751934055296e-07, + "loss": 0.31, + "step": 42167 + }, + { + "epoch": 0.9281972265023113, + "grad_norm": 1.87220528667606, + "learning_rate": 1.3459303149366777e-07, + "loss": 0.2292, + "step": 42168 + }, + { + "epoch": 0.9282192383887299, + "grad_norm": 1.8856690646776235, + "learning_rate": 1.3451089430997488e-07, + "loss": 0.2793, + "step": 42169 + }, + { + "epoch": 0.9282412502751486, + "grad_norm": 1.5690365718801, + "learning_rate": 1.3442878185486675e-07, + "loss": 0.3004, + "step": 42170 + }, + { + "epoch": 0.9282632621615673, + "grad_norm": 6.342785699605364, + "learning_rate": 1.3434669412876188e-07, + "loss": 0.3235, + "step": 42171 + }, + { + "epoch": 0.9282852740479859, + "grad_norm": 2.2895463929586457, + "learning_rate": 1.3426463113207777e-07, + "loss": 0.3284, + "step": 42172 + }, + { + "epoch": 0.9283072859344046, + "grad_norm": 1.688029083809243, + "learning_rate": 1.3418259286523072e-07, + "loss": 0.2476, + "step": 42173 + }, + { + "epoch": 0.9283292978208233, + "grad_norm": 1.3964027742158496, + "learning_rate": 1.3410057932863817e-07, + "loss": 0.2658, + "step": 42174 + }, + { + "epoch": 0.9283513097072419, + "grad_norm": 1.4427784482307848, + "learning_rate": 1.3401859052271594e-07, + "loss": 0.2378, + "step": 42175 + }, + { + "epoch": 0.9283733215936606, + "grad_norm": 1.600380415662829, + "learning_rate": 1.3393662644788197e-07, + "loss": 0.3783, + "step": 42176 + }, + { + "epoch": 0.9283953334800792, + "grad_norm": 1.5871491690818542, + "learning_rate": 1.3385468710455317e-07, + "loss": 0.2162, + "step": 42177 + }, + { + "epoch": 0.9284173453664979, + "grad_norm": 1.780729216268689, + "learning_rate": 1.3377277249314423e-07, + "loss": 0.353, + "step": 42178 + }, + { + "epoch": 0.9284393572529166, + "grad_norm": 1.8850134222554957, + "learning_rate": 1.3369088261407316e-07, + "loss": 0.2627, + "step": 42179 + }, + { + "epoch": 0.9284613691393352, + "grad_norm": 1.5010606459980451, + "learning_rate": 1.3360901746775457e-07, + "loss": 0.3061, + "step": 42180 + }, + { + "epoch": 0.9284833810257539, + "grad_norm": 2.0041526663101648, + "learning_rate": 1.3352717705460595e-07, + "loss": 0.3333, + "step": 42181 + }, + { + "epoch": 0.9285053929121726, + "grad_norm": 1.7392004760660922, + "learning_rate": 1.334453613750425e-07, + "loss": 0.2636, + "step": 42182 + }, + { + "epoch": 0.9285274047985912, + "grad_norm": 0.6107084487561075, + "learning_rate": 1.3336357042948057e-07, + "loss": 0.4564, + "step": 42183 + }, + { + "epoch": 0.9285494166850099, + "grad_norm": 1.3172478700295438, + "learning_rate": 1.3328180421833536e-07, + "loss": 0.1648, + "step": 42184 + }, + { + "epoch": 0.9285714285714286, + "grad_norm": 1.6239961092976818, + "learning_rate": 1.3320006274202324e-07, + "loss": 0.3517, + "step": 42185 + }, + { + "epoch": 0.9285934404578472, + "grad_norm": 1.9269707529002762, + "learning_rate": 1.3311834600095776e-07, + "loss": 0.2767, + "step": 42186 + }, + { + "epoch": 0.9286154523442659, + "grad_norm": 1.5868168667454243, + "learning_rate": 1.3303665399555688e-07, + "loss": 0.253, + "step": 42187 + }, + { + "epoch": 0.9286374642306846, + "grad_norm": 2.793284309534143, + "learning_rate": 1.3295498672623363e-07, + "loss": 0.3681, + "step": 42188 + }, + { + "epoch": 0.9286594761171032, + "grad_norm": 1.396074369416955, + "learning_rate": 1.3287334419340436e-07, + "loss": 0.2686, + "step": 42189 + }, + { + "epoch": 0.9286814880035219, + "grad_norm": 1.6841267792727626, + "learning_rate": 1.327917263974837e-07, + "loss": 0.2463, + "step": 42190 + }, + { + "epoch": 0.9287034998899406, + "grad_norm": 1.6006551160663391, + "learning_rate": 1.3271013333888526e-07, + "loss": 0.2657, + "step": 42191 + }, + { + "epoch": 0.9287255117763592, + "grad_norm": 1.7871825654794955, + "learning_rate": 1.3262856501802534e-07, + "loss": 0.2905, + "step": 42192 + }, + { + "epoch": 0.9287475236627779, + "grad_norm": 0.6093038464205212, + "learning_rate": 1.3254702143531805e-07, + "loss": 0.4405, + "step": 42193 + }, + { + "epoch": 0.9287695355491966, + "grad_norm": 0.6010764828307996, + "learning_rate": 1.3246550259117808e-07, + "loss": 0.4434, + "step": 42194 + }, + { + "epoch": 0.9287915474356152, + "grad_norm": 1.7192267402570525, + "learning_rate": 1.3238400848601952e-07, + "loss": 0.2644, + "step": 42195 + }, + { + "epoch": 0.9288135593220339, + "grad_norm": 1.5851821807558673, + "learning_rate": 1.3230253912025593e-07, + "loss": 0.2798, + "step": 42196 + }, + { + "epoch": 0.9288355712084526, + "grad_norm": 2.855041482021281, + "learning_rate": 1.3222109449430198e-07, + "loss": 0.2367, + "step": 42197 + }, + { + "epoch": 0.9288575830948712, + "grad_norm": 3.748735177341872, + "learning_rate": 1.3213967460857125e-07, + "loss": 0.2755, + "step": 42198 + }, + { + "epoch": 0.9288795949812899, + "grad_norm": 1.4784923933457388, + "learning_rate": 1.3205827946347839e-07, + "loss": 0.2404, + "step": 42199 + }, + { + "epoch": 0.9289016068677086, + "grad_norm": 3.0302646914656997, + "learning_rate": 1.3197690905943639e-07, + "loss": 0.3394, + "step": 42200 + }, + { + "epoch": 0.9289236187541272, + "grad_norm": 1.8609463071743468, + "learning_rate": 1.3189556339685882e-07, + "loss": 0.3239, + "step": 42201 + }, + { + "epoch": 0.9289456306405459, + "grad_norm": 1.6620662749457562, + "learning_rate": 1.318142424761587e-07, + "loss": 0.2665, + "step": 42202 + }, + { + "epoch": 0.9289676425269645, + "grad_norm": 1.954614359321257, + "learning_rate": 1.3173294629775068e-07, + "loss": 0.2554, + "step": 42203 + }, + { + "epoch": 0.9289896544133832, + "grad_norm": 0.6198779125808623, + "learning_rate": 1.3165167486204666e-07, + "loss": 0.4239, + "step": 42204 + }, + { + "epoch": 0.9290116662998019, + "grad_norm": 1.7567960032026728, + "learning_rate": 1.3157042816946074e-07, + "loss": 0.3052, + "step": 42205 + }, + { + "epoch": 0.9290336781862205, + "grad_norm": 2.4027589474965456, + "learning_rate": 1.3148920622040483e-07, + "loss": 0.3083, + "step": 42206 + }, + { + "epoch": 0.9290556900726392, + "grad_norm": 1.5391658133240136, + "learning_rate": 1.3140800901529195e-07, + "loss": 0.2944, + "step": 42207 + }, + { + "epoch": 0.9290777019590579, + "grad_norm": 2.2353616803665863, + "learning_rate": 1.313268365545356e-07, + "loss": 0.2476, + "step": 42208 + }, + { + "epoch": 0.9290997138454765, + "grad_norm": 1.5692476835606202, + "learning_rate": 1.3124568883854715e-07, + "loss": 0.2999, + "step": 42209 + }, + { + "epoch": 0.9291217257318952, + "grad_norm": 2.174093107645986, + "learning_rate": 1.3116456586774018e-07, + "loss": 0.2261, + "step": 42210 + }, + { + "epoch": 0.9291437376183139, + "grad_norm": 1.8169606639799072, + "learning_rate": 1.3108346764252598e-07, + "loss": 0.2402, + "step": 42211 + }, + { + "epoch": 0.9291657495047325, + "grad_norm": 1.6868681489778914, + "learning_rate": 1.310023941633176e-07, + "loss": 0.1863, + "step": 42212 + }, + { + "epoch": 0.9291877613911512, + "grad_norm": 1.5995490583702017, + "learning_rate": 1.3092134543052693e-07, + "loss": 0.2072, + "step": 42213 + }, + { + "epoch": 0.9292097732775699, + "grad_norm": 1.579701938190956, + "learning_rate": 1.308403214445647e-07, + "loss": 0.3337, + "step": 42214 + }, + { + "epoch": 0.9292317851639885, + "grad_norm": 2.16663828716545, + "learning_rate": 1.3075932220584453e-07, + "loss": 0.3351, + "step": 42215 + }, + { + "epoch": 0.9292537970504072, + "grad_norm": 1.5625961293266768, + "learning_rate": 1.306783477147766e-07, + "loss": 0.3107, + "step": 42216 + }, + { + "epoch": 0.9292758089368259, + "grad_norm": 5.168080685050044, + "learning_rate": 1.305973979717734e-07, + "loss": 0.25, + "step": 42217 + }, + { + "epoch": 0.9292978208232445, + "grad_norm": 1.9916440651278737, + "learning_rate": 1.3051647297724623e-07, + "loss": 0.2791, + "step": 42218 + }, + { + "epoch": 0.9293198327096632, + "grad_norm": 2.712640630764035, + "learning_rate": 1.304355727316059e-07, + "loss": 0.3519, + "step": 42219 + }, + { + "epoch": 0.9293418445960819, + "grad_norm": 2.412556527440342, + "learning_rate": 1.303546972352643e-07, + "loss": 0.2929, + "step": 42220 + }, + { + "epoch": 0.9293638564825005, + "grad_norm": 1.895735510408469, + "learning_rate": 1.3027384648863217e-07, + "loss": 0.3132, + "step": 42221 + }, + { + "epoch": 0.9293858683689192, + "grad_norm": 1.5922821799608022, + "learning_rate": 1.3019302049211868e-07, + "loss": 0.2666, + "step": 42222 + }, + { + "epoch": 0.9294078802553379, + "grad_norm": 0.6111128789084079, + "learning_rate": 1.3011221924613792e-07, + "loss": 0.4332, + "step": 42223 + }, + { + "epoch": 0.9294298921417565, + "grad_norm": 0.6252919720551287, + "learning_rate": 1.3003144275109847e-07, + "loss": 0.4651, + "step": 42224 + }, + { + "epoch": 0.9294519040281752, + "grad_norm": 2.1128291767808154, + "learning_rate": 1.299506910074111e-07, + "loss": 0.3701, + "step": 42225 + }, + { + "epoch": 0.9294739159145938, + "grad_norm": 0.6197126888283744, + "learning_rate": 1.2986996401548714e-07, + "loss": 0.4575, + "step": 42226 + }, + { + "epoch": 0.9294959278010125, + "grad_norm": 1.5721872332013243, + "learning_rate": 1.2978926177573458e-07, + "loss": 0.2391, + "step": 42227 + }, + { + "epoch": 0.9295179396874312, + "grad_norm": 2.149884638448829, + "learning_rate": 1.2970858428856648e-07, + "loss": 0.2744, + "step": 42228 + }, + { + "epoch": 0.9295399515738498, + "grad_norm": 0.606969131569089, + "learning_rate": 1.2962793155439134e-07, + "loss": 0.4314, + "step": 42229 + }, + { + "epoch": 0.9295619634602685, + "grad_norm": 0.6267326725732894, + "learning_rate": 1.2954730357361945e-07, + "loss": 0.4624, + "step": 42230 + }, + { + "epoch": 0.9295839753466872, + "grad_norm": 1.2288083220319, + "learning_rate": 1.294667003466604e-07, + "loss": 0.2094, + "step": 42231 + }, + { + "epoch": 0.9296059872331058, + "grad_norm": 2.9585022011629256, + "learning_rate": 1.2938612187392342e-07, + "loss": 0.2113, + "step": 42232 + }, + { + "epoch": 0.9296279991195245, + "grad_norm": 1.7027361428638528, + "learning_rate": 1.2930556815581862e-07, + "loss": 0.278, + "step": 42233 + }, + { + "epoch": 0.9296500110059432, + "grad_norm": 2.03666377300782, + "learning_rate": 1.292250391927552e-07, + "loss": 0.2308, + "step": 42234 + }, + { + "epoch": 0.9296720228923618, + "grad_norm": 3.850102656763784, + "learning_rate": 1.2914453498514335e-07, + "loss": 0.2931, + "step": 42235 + }, + { + "epoch": 0.9296940347787805, + "grad_norm": 3.4030675963391195, + "learning_rate": 1.2906405553339108e-07, + "loss": 0.2573, + "step": 42236 + }, + { + "epoch": 0.9297160466651992, + "grad_norm": 0.6123110858955516, + "learning_rate": 1.289836008379075e-07, + "loss": 0.4377, + "step": 42237 + }, + { + "epoch": 0.9297380585516178, + "grad_norm": 1.4538140833076811, + "learning_rate": 1.2890317089910176e-07, + "loss": 0.2356, + "step": 42238 + }, + { + "epoch": 0.9297600704380365, + "grad_norm": 0.6177591757621002, + "learning_rate": 1.2882276571738294e-07, + "loss": 0.4667, + "step": 42239 + }, + { + "epoch": 0.9297820823244553, + "grad_norm": 1.7451487016379665, + "learning_rate": 1.2874238529315963e-07, + "loss": 0.3211, + "step": 42240 + }, + { + "epoch": 0.9298040942108738, + "grad_norm": 1.5775486735344386, + "learning_rate": 1.2866202962683981e-07, + "loss": 0.2962, + "step": 42241 + }, + { + "epoch": 0.9298261060972925, + "grad_norm": 1.680648089021718, + "learning_rate": 1.2858169871883207e-07, + "loss": 0.2577, + "step": 42242 + }, + { + "epoch": 0.9298481179837113, + "grad_norm": 2.271543935705863, + "learning_rate": 1.2850139256954497e-07, + "loss": 0.3459, + "step": 42243 + }, + { + "epoch": 0.9298701298701298, + "grad_norm": 2.642513188443991, + "learning_rate": 1.2842111117938706e-07, + "loss": 0.244, + "step": 42244 + }, + { + "epoch": 0.9298921417565486, + "grad_norm": 16.333489576193603, + "learning_rate": 1.2834085454876522e-07, + "loss": 0.3625, + "step": 42245 + }, + { + "epoch": 0.9299141536429673, + "grad_norm": 1.635296547893655, + "learning_rate": 1.282606226780886e-07, + "loss": 0.2642, + "step": 42246 + }, + { + "epoch": 0.9299361655293858, + "grad_norm": 1.3719472340377221, + "learning_rate": 1.281804155677635e-07, + "loss": 0.2372, + "step": 42247 + }, + { + "epoch": 0.9299581774158046, + "grad_norm": 1.7516556493725253, + "learning_rate": 1.2810023321819853e-07, + "loss": 0.2656, + "step": 42248 + }, + { + "epoch": 0.9299801893022231, + "grad_norm": 2.224551529872583, + "learning_rate": 1.2802007562980223e-07, + "loss": 0.4481, + "step": 42249 + }, + { + "epoch": 0.9300022011886419, + "grad_norm": 1.4065730240207914, + "learning_rate": 1.2793994280297984e-07, + "loss": 0.259, + "step": 42250 + }, + { + "epoch": 0.9300242130750606, + "grad_norm": 0.6158592106376525, + "learning_rate": 1.278598347381399e-07, + "loss": 0.4562, + "step": 42251 + }, + { + "epoch": 0.9300462249614792, + "grad_norm": 1.858011637762577, + "learning_rate": 1.2777975143568934e-07, + "loss": 0.3535, + "step": 42252 + }, + { + "epoch": 0.9300682368478979, + "grad_norm": 2.1998308758212244, + "learning_rate": 1.27699692896035e-07, + "loss": 0.3116, + "step": 42253 + }, + { + "epoch": 0.9300902487343166, + "grad_norm": 1.5090506519364213, + "learning_rate": 1.2761965911958385e-07, + "loss": 0.2438, + "step": 42254 + }, + { + "epoch": 0.9301122606207352, + "grad_norm": 1.3019424117105478, + "learning_rate": 1.2753965010674274e-07, + "loss": 0.2719, + "step": 42255 + }, + { + "epoch": 0.9301342725071539, + "grad_norm": 2.7711962384501163, + "learning_rate": 1.2745966585791913e-07, + "loss": 0.3375, + "step": 42256 + }, + { + "epoch": 0.9301562843935726, + "grad_norm": 1.8869888607422156, + "learning_rate": 1.2737970637351772e-07, + "loss": 0.2075, + "step": 42257 + }, + { + "epoch": 0.9301782962799912, + "grad_norm": 3.0685315448510155, + "learning_rate": 1.272997716539459e-07, + "loss": 0.2949, + "step": 42258 + }, + { + "epoch": 0.9302003081664099, + "grad_norm": 1.319756161138404, + "learning_rate": 1.2721986169961064e-07, + "loss": 0.2776, + "step": 42259 + }, + { + "epoch": 0.9302223200528286, + "grad_norm": 0.5910509435790473, + "learning_rate": 1.2713997651091715e-07, + "loss": 0.4407, + "step": 42260 + }, + { + "epoch": 0.9302443319392472, + "grad_norm": 1.9254548867294154, + "learning_rate": 1.2706011608827174e-07, + "loss": 0.3234, + "step": 42261 + }, + { + "epoch": 0.9302663438256659, + "grad_norm": 1.592243005513412, + "learning_rate": 1.2698028043208021e-07, + "loss": 0.2342, + "step": 42262 + }, + { + "epoch": 0.9302883557120846, + "grad_norm": 5.77399025414408, + "learning_rate": 1.2690046954274725e-07, + "loss": 0.3201, + "step": 42263 + }, + { + "epoch": 0.9303103675985032, + "grad_norm": 0.6193897003769308, + "learning_rate": 1.2682068342068088e-07, + "loss": 0.4437, + "step": 42264 + }, + { + "epoch": 0.9303323794849219, + "grad_norm": 0.6256132396936416, + "learning_rate": 1.2674092206628463e-07, + "loss": 0.4529, + "step": 42265 + }, + { + "epoch": 0.9303543913713406, + "grad_norm": 2.2112452038475494, + "learning_rate": 1.2666118547996487e-07, + "loss": 0.2438, + "step": 42266 + }, + { + "epoch": 0.9303764032577592, + "grad_norm": 1.6386537082296588, + "learning_rate": 1.265814736621268e-07, + "loss": 0.3266, + "step": 42267 + }, + { + "epoch": 0.9303984151441779, + "grad_norm": 3.2277243227594195, + "learning_rate": 1.2650178661317458e-07, + "loss": 0.3148, + "step": 42268 + }, + { + "epoch": 0.9304204270305966, + "grad_norm": 1.4354335727560616, + "learning_rate": 1.2642212433351397e-07, + "loss": 0.2647, + "step": 42269 + }, + { + "epoch": 0.9304424389170152, + "grad_norm": 1.6526308786818997, + "learning_rate": 1.2634248682355023e-07, + "loss": 0.3359, + "step": 42270 + }, + { + "epoch": 0.9304644508034339, + "grad_norm": 1.4288986223996039, + "learning_rate": 1.2626287408368743e-07, + "loss": 0.3318, + "step": 42271 + }, + { + "epoch": 0.9304864626898525, + "grad_norm": 2.5338679990452158, + "learning_rate": 1.2618328611433083e-07, + "loss": 0.2707, + "step": 42272 + }, + { + "epoch": 0.9305084745762712, + "grad_norm": 1.275032325420538, + "learning_rate": 1.2610372291588402e-07, + "loss": 0.1814, + "step": 42273 + }, + { + "epoch": 0.9305304864626899, + "grad_norm": 1.9143994737323846, + "learning_rate": 1.260241844887522e-07, + "loss": 0.308, + "step": 42274 + }, + { + "epoch": 0.9305524983491085, + "grad_norm": 0.5957870918040576, + "learning_rate": 1.2594467083333894e-07, + "loss": 0.4189, + "step": 42275 + }, + { + "epoch": 0.9305745102355272, + "grad_norm": 3.3607943194628254, + "learning_rate": 1.2586518195004948e-07, + "loss": 0.3221, + "step": 42276 + }, + { + "epoch": 0.9305965221219459, + "grad_norm": 1.6153096470699124, + "learning_rate": 1.2578571783928685e-07, + "loss": 0.3312, + "step": 42277 + }, + { + "epoch": 0.9306185340083645, + "grad_norm": 1.6161673546940156, + "learning_rate": 1.257062785014551e-07, + "loss": 0.2334, + "step": 42278 + }, + { + "epoch": 0.9306405458947832, + "grad_norm": 1.7732752587211182, + "learning_rate": 1.2562686393695788e-07, + "loss": 0.3132, + "step": 42279 + }, + { + "epoch": 0.9306625577812019, + "grad_norm": 2.8365343212564817, + "learning_rate": 1.255474741461987e-07, + "loss": 0.2868, + "step": 42280 + }, + { + "epoch": 0.9306845696676205, + "grad_norm": 1.9076319236759023, + "learning_rate": 1.2546810912958174e-07, + "loss": 0.3378, + "step": 42281 + }, + { + "epoch": 0.9307065815540392, + "grad_norm": 1.6022517751395515, + "learning_rate": 1.253887688875105e-07, + "loss": 0.2709, + "step": 42282 + }, + { + "epoch": 0.9307285934404579, + "grad_norm": 0.6073834387070016, + "learning_rate": 1.2530945342038693e-07, + "loss": 0.4548, + "step": 42283 + }, + { + "epoch": 0.9307506053268765, + "grad_norm": 0.6403990439368308, + "learning_rate": 1.2523016272861453e-07, + "loss": 0.479, + "step": 42284 + }, + { + "epoch": 0.9307726172132952, + "grad_norm": 1.7539258822450252, + "learning_rate": 1.251508968125975e-07, + "loss": 0.2844, + "step": 42285 + }, + { + "epoch": 0.9307946290997139, + "grad_norm": 1.8291533834614115, + "learning_rate": 1.2507165567273717e-07, + "loss": 0.2546, + "step": 42286 + }, + { + "epoch": 0.9308166409861325, + "grad_norm": 0.6026354674336154, + "learning_rate": 1.2499243930943762e-07, + "loss": 0.4156, + "step": 42287 + }, + { + "epoch": 0.9308386528725512, + "grad_norm": 2.2594649929826183, + "learning_rate": 1.2491324772310022e-07, + "loss": 0.312, + "step": 42288 + }, + { + "epoch": 0.9308606647589699, + "grad_norm": 0.6133848990268442, + "learning_rate": 1.2483408091412795e-07, + "loss": 0.4357, + "step": 42289 + }, + { + "epoch": 0.9308826766453885, + "grad_norm": 3.078789602227933, + "learning_rate": 1.247549388829239e-07, + "loss": 0.2494, + "step": 42290 + }, + { + "epoch": 0.9309046885318072, + "grad_norm": 2.7723812605743277, + "learning_rate": 1.2467582162988934e-07, + "loss": 0.364, + "step": 42291 + }, + { + "epoch": 0.9309267004182259, + "grad_norm": 0.6586379236088853, + "learning_rate": 1.245967291554273e-07, + "loss": 0.4695, + "step": 42292 + }, + { + "epoch": 0.9309487123046445, + "grad_norm": 1.934198647625074, + "learning_rate": 1.2451766145993803e-07, + "loss": 0.4023, + "step": 42293 + }, + { + "epoch": 0.9309707241910632, + "grad_norm": 1.8685563568887162, + "learning_rate": 1.244386185438251e-07, + "loss": 0.3595, + "step": 42294 + }, + { + "epoch": 0.9309927360774818, + "grad_norm": 0.6459137161544, + "learning_rate": 1.2435960040748986e-07, + "loss": 0.4733, + "step": 42295 + }, + { + "epoch": 0.9310147479639005, + "grad_norm": 0.6163367019918812, + "learning_rate": 1.2428060705133361e-07, + "loss": 0.4581, + "step": 42296 + }, + { + "epoch": 0.9310367598503192, + "grad_norm": 1.396951219625323, + "learning_rate": 1.242016384757583e-07, + "loss": 0.2629, + "step": 42297 + }, + { + "epoch": 0.9310587717367378, + "grad_norm": 1.8488804381093153, + "learning_rate": 1.241226946811641e-07, + "loss": 0.3296, + "step": 42298 + }, + { + "epoch": 0.9310807836231565, + "grad_norm": 2.1920241745400526, + "learning_rate": 1.2404377566795412e-07, + "loss": 0.3963, + "step": 42299 + }, + { + "epoch": 0.9311027955095752, + "grad_norm": 1.772214994248369, + "learning_rate": 1.239648814365274e-07, + "loss": 0.2504, + "step": 42300 + }, + { + "epoch": 0.9311248073959938, + "grad_norm": 1.773741942028708, + "learning_rate": 1.238860119872859e-07, + "loss": 0.2599, + "step": 42301 + }, + { + "epoch": 0.9311468192824125, + "grad_norm": 1.571548886724107, + "learning_rate": 1.2380716732063147e-07, + "loss": 0.2681, + "step": 42302 + }, + { + "epoch": 0.9311688311688312, + "grad_norm": 1.6003771449224098, + "learning_rate": 1.2372834743696328e-07, + "loss": 0.2941, + "step": 42303 + }, + { + "epoch": 0.9311908430552498, + "grad_norm": 7.15567572918559, + "learning_rate": 1.2364955233668264e-07, + "loss": 0.288, + "step": 42304 + }, + { + "epoch": 0.9312128549416685, + "grad_norm": 1.7021349166463655, + "learning_rate": 1.2357078202018923e-07, + "loss": 0.323, + "step": 42305 + }, + { + "epoch": 0.9312348668280872, + "grad_norm": 1.6457759181635394, + "learning_rate": 1.2349203648788444e-07, + "loss": 0.2804, + "step": 42306 + }, + { + "epoch": 0.9312568787145058, + "grad_norm": 7.546843646768752, + "learning_rate": 1.2341331574016846e-07, + "loss": 0.2407, + "step": 42307 + }, + { + "epoch": 0.9312788906009245, + "grad_norm": 2.0301793191652, + "learning_rate": 1.2333461977744155e-07, + "loss": 0.2992, + "step": 42308 + }, + { + "epoch": 0.9313009024873432, + "grad_norm": 2.476391273486657, + "learning_rate": 1.2325594860010225e-07, + "loss": 0.2874, + "step": 42309 + }, + { + "epoch": 0.9313229143737618, + "grad_norm": 1.5974067231096523, + "learning_rate": 1.2317730220855084e-07, + "loss": 0.2253, + "step": 42310 + }, + { + "epoch": 0.9313449262601805, + "grad_norm": 1.8415796921838252, + "learning_rate": 1.2309868060318808e-07, + "loss": 0.248, + "step": 42311 + }, + { + "epoch": 0.9313669381465992, + "grad_norm": 1.947916323540548, + "learning_rate": 1.2302008378441366e-07, + "loss": 0.3046, + "step": 42312 + }, + { + "epoch": 0.9313889500330178, + "grad_norm": 2.489730038494463, + "learning_rate": 1.2294151175262614e-07, + "loss": 0.3841, + "step": 42313 + }, + { + "epoch": 0.9314109619194365, + "grad_norm": 1.5734446467250587, + "learning_rate": 1.228629645082241e-07, + "loss": 0.2782, + "step": 42314 + }, + { + "epoch": 0.9314329738058552, + "grad_norm": 1.691667454342868, + "learning_rate": 1.2278444205160833e-07, + "loss": 0.2339, + "step": 42315 + }, + { + "epoch": 0.9314549856922738, + "grad_norm": 1.7257211612948513, + "learning_rate": 1.2270594438317684e-07, + "loss": 0.2681, + "step": 42316 + }, + { + "epoch": 0.9314769975786925, + "grad_norm": 2.938513502024921, + "learning_rate": 1.2262747150332987e-07, + "loss": 0.2903, + "step": 42317 + }, + { + "epoch": 0.9314990094651111, + "grad_norm": 1.5480423237053824, + "learning_rate": 1.22549023412466e-07, + "loss": 0.3541, + "step": 42318 + }, + { + "epoch": 0.9315210213515298, + "grad_norm": 1.686337583343427, + "learning_rate": 1.224706001109821e-07, + "loss": 0.309, + "step": 42319 + }, + { + "epoch": 0.9315430332379485, + "grad_norm": 2.3063001075595033, + "learning_rate": 1.2239220159927845e-07, + "loss": 0.3332, + "step": 42320 + }, + { + "epoch": 0.9315650451243671, + "grad_norm": 1.628537075309642, + "learning_rate": 1.223138278777536e-07, + "loss": 0.2652, + "step": 42321 + }, + { + "epoch": 0.9315870570107858, + "grad_norm": 1.7419710176062522, + "learning_rate": 1.2223547894680442e-07, + "loss": 0.1971, + "step": 42322 + }, + { + "epoch": 0.9316090688972045, + "grad_norm": 1.6172820525759255, + "learning_rate": 1.221571548068312e-07, + "loss": 0.2971, + "step": 42323 + }, + { + "epoch": 0.9316310807836231, + "grad_norm": 1.4924343163841383, + "learning_rate": 1.2207885545822973e-07, + "loss": 0.288, + "step": 42324 + }, + { + "epoch": 0.9316530926700418, + "grad_norm": 0.607871521317983, + "learning_rate": 1.2200058090139964e-07, + "loss": 0.4189, + "step": 42325 + }, + { + "epoch": 0.9316751045564605, + "grad_norm": 1.8145691235837345, + "learning_rate": 1.2192233113673847e-07, + "loss": 0.2835, + "step": 42326 + }, + { + "epoch": 0.9316971164428791, + "grad_norm": 1.591813555943305, + "learning_rate": 1.218441061646436e-07, + "loss": 0.2598, + "step": 42327 + }, + { + "epoch": 0.9317191283292978, + "grad_norm": 2.087069146985392, + "learning_rate": 1.2176590598551307e-07, + "loss": 0.2696, + "step": 42328 + }, + { + "epoch": 0.9317411402157165, + "grad_norm": 0.6919613093879322, + "learning_rate": 1.2168773059974326e-07, + "loss": 0.4431, + "step": 42329 + }, + { + "epoch": 0.9317631521021351, + "grad_norm": 1.6892337793984677, + "learning_rate": 1.2160958000773214e-07, + "loss": 0.2535, + "step": 42330 + }, + { + "epoch": 0.9317851639885538, + "grad_norm": 9.462748021255658, + "learning_rate": 1.2153145420987778e-07, + "loss": 0.2761, + "step": 42331 + }, + { + "epoch": 0.9318071758749725, + "grad_norm": 0.5971007947874688, + "learning_rate": 1.2145335320657535e-07, + "loss": 0.42, + "step": 42332 + }, + { + "epoch": 0.9318291877613911, + "grad_norm": 1.6812360967699962, + "learning_rate": 1.2137527699822405e-07, + "loss": 0.3148, + "step": 42333 + }, + { + "epoch": 0.9318511996478098, + "grad_norm": 1.956984558319285, + "learning_rate": 1.212972255852185e-07, + "loss": 0.2307, + "step": 42334 + }, + { + "epoch": 0.9318732115342285, + "grad_norm": 2.1173314473262597, + "learning_rate": 1.212191989679573e-07, + "loss": 0.2434, + "step": 42335 + }, + { + "epoch": 0.9318952234206471, + "grad_norm": 1.913913388516529, + "learning_rate": 1.2114119714683513e-07, + "loss": 0.2756, + "step": 42336 + }, + { + "epoch": 0.9319172353070658, + "grad_norm": 2.3133975155722575, + "learning_rate": 1.2106322012225002e-07, + "loss": 0.3262, + "step": 42337 + }, + { + "epoch": 0.9319392471934845, + "grad_norm": 1.96446533265455, + "learning_rate": 1.2098526789459775e-07, + "loss": 0.2863, + "step": 42338 + }, + { + "epoch": 0.9319612590799031, + "grad_norm": 0.6043786004173367, + "learning_rate": 1.209073404642741e-07, + "loss": 0.4379, + "step": 42339 + }, + { + "epoch": 0.9319832709663218, + "grad_norm": 2.6862807028129994, + "learning_rate": 1.20829437831676e-07, + "loss": 0.307, + "step": 42340 + }, + { + "epoch": 0.9320052828527405, + "grad_norm": 1.9425794165591226, + "learning_rate": 1.2075155999719866e-07, + "loss": 0.23, + "step": 42341 + }, + { + "epoch": 0.9320272947391591, + "grad_norm": 1.75538847535914, + "learning_rate": 1.2067370696123793e-07, + "loss": 0.3666, + "step": 42342 + }, + { + "epoch": 0.9320493066255778, + "grad_norm": 1.4122558727140666, + "learning_rate": 1.205958787241901e-07, + "loss": 0.252, + "step": 42343 + }, + { + "epoch": 0.9320713185119964, + "grad_norm": 2.110587658605595, + "learning_rate": 1.2051807528644987e-07, + "loss": 0.3397, + "step": 42344 + }, + { + "epoch": 0.9320933303984151, + "grad_norm": 1.767428085205705, + "learning_rate": 1.2044029664841307e-07, + "loss": 0.2196, + "step": 42345 + }, + { + "epoch": 0.9321153422848338, + "grad_norm": 17.029010485943694, + "learning_rate": 1.2036254281047488e-07, + "loss": 0.3376, + "step": 42346 + }, + { + "epoch": 0.9321373541712524, + "grad_norm": 1.5871794202576885, + "learning_rate": 1.202848137730306e-07, + "loss": 0.2116, + "step": 42347 + }, + { + "epoch": 0.9321593660576711, + "grad_norm": 1.8628552066709034, + "learning_rate": 1.2020710953647595e-07, + "loss": 0.291, + "step": 42348 + }, + { + "epoch": 0.9321813779440898, + "grad_norm": 0.6211243040704069, + "learning_rate": 1.2012943010120514e-07, + "loss": 0.4444, + "step": 42349 + }, + { + "epoch": 0.9322033898305084, + "grad_norm": 1.6577103348487252, + "learning_rate": 1.2005177546761282e-07, + "loss": 0.2693, + "step": 42350 + }, + { + "epoch": 0.9322254017169271, + "grad_norm": 1.6022417604848438, + "learning_rate": 1.199741456360931e-07, + "loss": 0.272, + "step": 42351 + }, + { + "epoch": 0.9322474136033458, + "grad_norm": 1.4315392246616696, + "learning_rate": 1.1989654060704236e-07, + "loss": 0.2378, + "step": 42352 + }, + { + "epoch": 0.9322694254897644, + "grad_norm": 3.472288148805635, + "learning_rate": 1.198189603808536e-07, + "loss": 0.3259, + "step": 42353 + }, + { + "epoch": 0.9322914373761831, + "grad_norm": 0.5945411954401556, + "learning_rate": 1.1974140495792208e-07, + "loss": 0.4083, + "step": 42354 + }, + { + "epoch": 0.9323134492626018, + "grad_norm": 0.6085089898357183, + "learning_rate": 1.196638743386408e-07, + "loss": 0.4457, + "step": 42355 + }, + { + "epoch": 0.9323354611490204, + "grad_norm": 1.8306499127211442, + "learning_rate": 1.1958636852340387e-07, + "loss": 0.2763, + "step": 42356 + }, + { + "epoch": 0.9323574730354391, + "grad_norm": 1.7676397163570907, + "learning_rate": 1.1950888751260658e-07, + "loss": 0.3367, + "step": 42357 + }, + { + "epoch": 0.9323794849218578, + "grad_norm": 0.64460810211145, + "learning_rate": 1.1943143130664192e-07, + "loss": 0.4533, + "step": 42358 + }, + { + "epoch": 0.9324014968082764, + "grad_norm": 1.7313692494074464, + "learning_rate": 1.1935399990590345e-07, + "loss": 0.235, + "step": 42359 + }, + { + "epoch": 0.9324235086946951, + "grad_norm": 1.5568108667052094, + "learning_rate": 1.1927659331078422e-07, + "loss": 0.3701, + "step": 42360 + }, + { + "epoch": 0.9324455205811139, + "grad_norm": 0.6156468812130883, + "learning_rate": 1.1919921152167835e-07, + "loss": 0.4339, + "step": 42361 + }, + { + "epoch": 0.9324675324675324, + "grad_norm": 1.5215478888046507, + "learning_rate": 1.1912185453897995e-07, + "loss": 0.2908, + "step": 42362 + }, + { + "epoch": 0.9324895443539511, + "grad_norm": 1.5203606686051079, + "learning_rate": 1.190445223630804e-07, + "loss": 0.2963, + "step": 42363 + }, + { + "epoch": 0.9325115562403699, + "grad_norm": 1.4932775541263201, + "learning_rate": 1.189672149943738e-07, + "loss": 0.2788, + "step": 42364 + }, + { + "epoch": 0.9325335681267884, + "grad_norm": 2.3688800055078785, + "learning_rate": 1.1888993243325209e-07, + "loss": 0.303, + "step": 42365 + }, + { + "epoch": 0.9325555800132072, + "grad_norm": 2.10774296214719, + "learning_rate": 1.1881267468010937e-07, + "loss": 0.2769, + "step": 42366 + }, + { + "epoch": 0.9325775918996257, + "grad_norm": 1.2566017198391188, + "learning_rate": 1.1873544173533814e-07, + "loss": 0.193, + "step": 42367 + }, + { + "epoch": 0.9325996037860445, + "grad_norm": 0.6151848173993784, + "learning_rate": 1.186582335993297e-07, + "loss": 0.4641, + "step": 42368 + }, + { + "epoch": 0.9326216156724632, + "grad_norm": 0.6247253784711778, + "learning_rate": 1.1858105027247768e-07, + "loss": 0.4457, + "step": 42369 + }, + { + "epoch": 0.9326436275588817, + "grad_norm": 1.3799582338244367, + "learning_rate": 1.1850389175517396e-07, + "loss": 0.2937, + "step": 42370 + }, + { + "epoch": 0.9326656394453005, + "grad_norm": 1.8096765733488596, + "learning_rate": 1.1842675804781101e-07, + "loss": 0.3041, + "step": 42371 + }, + { + "epoch": 0.9326876513317192, + "grad_norm": 4.8291080951723915, + "learning_rate": 1.1834964915077963e-07, + "loss": 0.2902, + "step": 42372 + }, + { + "epoch": 0.9327096632181378, + "grad_norm": 4.023480574070884, + "learning_rate": 1.1827256506447283e-07, + "loss": 0.2181, + "step": 42373 + }, + { + "epoch": 0.9327316751045565, + "grad_norm": 1.4791795299400725, + "learning_rate": 1.1819550578928252e-07, + "loss": 0.3266, + "step": 42374 + }, + { + "epoch": 0.9327536869909752, + "grad_norm": 3.4943446594936276, + "learning_rate": 1.1811847132560006e-07, + "loss": 0.2536, + "step": 42375 + }, + { + "epoch": 0.9327756988773938, + "grad_norm": 5.293863813373252, + "learning_rate": 1.1804146167381681e-07, + "loss": 0.3535, + "step": 42376 + }, + { + "epoch": 0.9327977107638125, + "grad_norm": 5.368374181391703, + "learning_rate": 1.1796447683432411e-07, + "loss": 0.2869, + "step": 42377 + }, + { + "epoch": 0.9328197226502312, + "grad_norm": 2.872596095736359, + "learning_rate": 1.1788751680751331e-07, + "loss": 0.3252, + "step": 42378 + }, + { + "epoch": 0.9328417345366498, + "grad_norm": 1.8355807219462148, + "learning_rate": 1.1781058159377578e-07, + "loss": 0.2434, + "step": 42379 + }, + { + "epoch": 0.9328637464230685, + "grad_norm": 1.502265754276132, + "learning_rate": 1.1773367119350232e-07, + "loss": 0.2578, + "step": 42380 + }, + { + "epoch": 0.9328857583094872, + "grad_norm": 1.890265982475079, + "learning_rate": 1.1765678560708427e-07, + "loss": 0.2993, + "step": 42381 + }, + { + "epoch": 0.9329077701959058, + "grad_norm": 0.617508674533172, + "learning_rate": 1.1757992483491132e-07, + "loss": 0.4393, + "step": 42382 + }, + { + "epoch": 0.9329297820823245, + "grad_norm": 2.0851778679791684, + "learning_rate": 1.1750308887737538e-07, + "loss": 0.2615, + "step": 42383 + }, + { + "epoch": 0.9329517939687432, + "grad_norm": 1.6400000499097203, + "learning_rate": 1.1742627773486614e-07, + "loss": 0.2727, + "step": 42384 + }, + { + "epoch": 0.9329738058551618, + "grad_norm": 0.6132209956985102, + "learning_rate": 1.1734949140777496e-07, + "loss": 0.4551, + "step": 42385 + }, + { + "epoch": 0.9329958177415805, + "grad_norm": 2.056143187863048, + "learning_rate": 1.1727272989649041e-07, + "loss": 0.2693, + "step": 42386 + }, + { + "epoch": 0.9330178296279992, + "grad_norm": 2.2226639336527887, + "learning_rate": 1.1719599320140384e-07, + "loss": 0.3826, + "step": 42387 + }, + { + "epoch": 0.9330398415144178, + "grad_norm": 0.6074928121456723, + "learning_rate": 1.1711928132290495e-07, + "loss": 0.4274, + "step": 42388 + }, + { + "epoch": 0.9330618534008365, + "grad_norm": 1.791050140291274, + "learning_rate": 1.1704259426138399e-07, + "loss": 0.2992, + "step": 42389 + }, + { + "epoch": 0.9330838652872551, + "grad_norm": 3.326470570045662, + "learning_rate": 1.169659320172306e-07, + "loss": 0.2903, + "step": 42390 + }, + { + "epoch": 0.9331058771736738, + "grad_norm": 0.6154644250614919, + "learning_rate": 1.1688929459083398e-07, + "loss": 0.4385, + "step": 42391 + }, + { + "epoch": 0.9331278890600925, + "grad_norm": 1.4941343795465192, + "learning_rate": 1.1681268198258377e-07, + "loss": 0.2633, + "step": 42392 + }, + { + "epoch": 0.9331499009465111, + "grad_norm": 1.6657173677128116, + "learning_rate": 1.1673609419286913e-07, + "loss": 0.3196, + "step": 42393 + }, + { + "epoch": 0.9331719128329298, + "grad_norm": 1.7890257903398175, + "learning_rate": 1.1665953122208029e-07, + "loss": 0.3404, + "step": 42394 + }, + { + "epoch": 0.9331939247193485, + "grad_norm": 1.9642259504385426, + "learning_rate": 1.1658299307060583e-07, + "loss": 0.2392, + "step": 42395 + }, + { + "epoch": 0.9332159366057671, + "grad_norm": 2.359665609028288, + "learning_rate": 1.1650647973883434e-07, + "loss": 0.2761, + "step": 42396 + }, + { + "epoch": 0.9332379484921858, + "grad_norm": 2.12088975908721, + "learning_rate": 1.1642999122715492e-07, + "loss": 0.3182, + "step": 42397 + }, + { + "epoch": 0.9332599603786045, + "grad_norm": 5.404387981049737, + "learning_rate": 1.163535275359562e-07, + "loss": 0.2626, + "step": 42398 + }, + { + "epoch": 0.9332819722650231, + "grad_norm": 1.4803094273468484, + "learning_rate": 1.1627708866562781e-07, + "loss": 0.305, + "step": 42399 + }, + { + "epoch": 0.9333039841514418, + "grad_norm": 2.012595879845346, + "learning_rate": 1.1620067461655726e-07, + "loss": 0.3146, + "step": 42400 + }, + { + "epoch": 0.9333259960378605, + "grad_norm": 1.4907665699391968, + "learning_rate": 1.1612428538913256e-07, + "loss": 0.2465, + "step": 42401 + }, + { + "epoch": 0.9333480079242791, + "grad_norm": 2.540087195802922, + "learning_rate": 1.1604792098374284e-07, + "loss": 0.333, + "step": 42402 + }, + { + "epoch": 0.9333700198106978, + "grad_norm": 0.6326970565585026, + "learning_rate": 1.1597158140077613e-07, + "loss": 0.4507, + "step": 42403 + }, + { + "epoch": 0.9333920316971165, + "grad_norm": 1.7209506560565957, + "learning_rate": 1.1589526664061989e-07, + "loss": 0.247, + "step": 42404 + }, + { + "epoch": 0.9334140435835351, + "grad_norm": 2.0515597841087203, + "learning_rate": 1.1581897670366271e-07, + "loss": 0.3336, + "step": 42405 + }, + { + "epoch": 0.9334360554699538, + "grad_norm": 0.6373852027338813, + "learning_rate": 1.1574271159029094e-07, + "loss": 0.4504, + "step": 42406 + }, + { + "epoch": 0.9334580673563725, + "grad_norm": 0.611230482352853, + "learning_rate": 1.1566647130089425e-07, + "loss": 0.4324, + "step": 42407 + }, + { + "epoch": 0.9334800792427911, + "grad_norm": 1.8891809302742792, + "learning_rate": 1.1559025583585793e-07, + "loss": 0.2505, + "step": 42408 + }, + { + "epoch": 0.9335020911292098, + "grad_norm": 0.6117742417691816, + "learning_rate": 1.1551406519557106e-07, + "loss": 0.4167, + "step": 42409 + }, + { + "epoch": 0.9335241030156285, + "grad_norm": 2.2088710523989117, + "learning_rate": 1.154378993804206e-07, + "loss": 0.3246, + "step": 42410 + }, + { + "epoch": 0.9335461149020471, + "grad_norm": 1.5844029095688852, + "learning_rate": 1.1536175839079233e-07, + "loss": 0.3213, + "step": 42411 + }, + { + "epoch": 0.9335681267884658, + "grad_norm": 0.5976551500863778, + "learning_rate": 1.1528564222707539e-07, + "loss": 0.4235, + "step": 42412 + }, + { + "epoch": 0.9335901386748844, + "grad_norm": 1.5417774455414457, + "learning_rate": 1.1520955088965446e-07, + "loss": 0.3223, + "step": 42413 + }, + { + "epoch": 0.9336121505613031, + "grad_norm": 4.465787878869539, + "learning_rate": 1.1513348437891758e-07, + "loss": 0.3279, + "step": 42414 + }, + { + "epoch": 0.9336341624477218, + "grad_norm": 3.0962281438354102, + "learning_rate": 1.1505744269525166e-07, + "loss": 0.3442, + "step": 42415 + }, + { + "epoch": 0.9336561743341404, + "grad_norm": 1.4549393631507468, + "learning_rate": 1.1498142583904137e-07, + "loss": 0.2491, + "step": 42416 + }, + { + "epoch": 0.9336781862205591, + "grad_norm": 2.7953453706458156, + "learning_rate": 1.1490543381067532e-07, + "loss": 0.314, + "step": 42417 + }, + { + "epoch": 0.9337001981069778, + "grad_norm": 1.761573430967779, + "learning_rate": 1.1482946661053817e-07, + "loss": 0.2551, + "step": 42418 + }, + { + "epoch": 0.9337222099933964, + "grad_norm": 2.014779317178323, + "learning_rate": 1.1475352423901632e-07, + "loss": 0.2744, + "step": 42419 + }, + { + "epoch": 0.9337442218798151, + "grad_norm": 1.0535466172460204, + "learning_rate": 1.1467760669649664e-07, + "loss": 0.1653, + "step": 42420 + }, + { + "epoch": 0.9337662337662338, + "grad_norm": 0.6239098921715803, + "learning_rate": 1.146017139833644e-07, + "loss": 0.4481, + "step": 42421 + }, + { + "epoch": 0.9337882456526524, + "grad_norm": 1.5805354978308668, + "learning_rate": 1.1452584610000428e-07, + "loss": 0.3131, + "step": 42422 + }, + { + "epoch": 0.9338102575390711, + "grad_norm": 2.2208140347078933, + "learning_rate": 1.1445000304680321e-07, + "loss": 0.2351, + "step": 42423 + }, + { + "epoch": 0.9338322694254898, + "grad_norm": 1.8288054416253527, + "learning_rate": 1.1437418482414586e-07, + "loss": 0.2814, + "step": 42424 + }, + { + "epoch": 0.9338542813119084, + "grad_norm": 2.4808649979913273, + "learning_rate": 1.1429839143241862e-07, + "loss": 0.3404, + "step": 42425 + }, + { + "epoch": 0.9338762931983271, + "grad_norm": 0.6376268983182622, + "learning_rate": 1.1422262287200613e-07, + "loss": 0.4648, + "step": 42426 + }, + { + "epoch": 0.9338983050847458, + "grad_norm": 2.120730477365909, + "learning_rate": 1.1414687914329258e-07, + "loss": 0.2941, + "step": 42427 + }, + { + "epoch": 0.9339203169711644, + "grad_norm": 1.5134932586524932, + "learning_rate": 1.1407116024666376e-07, + "loss": 0.2201, + "step": 42428 + }, + { + "epoch": 0.9339423288575831, + "grad_norm": 2.0004701441777653, + "learning_rate": 1.139954661825049e-07, + "loss": 0.362, + "step": 42429 + }, + { + "epoch": 0.9339643407440018, + "grad_norm": 1.6675858953102938, + "learning_rate": 1.139197969512007e-07, + "loss": 0.3203, + "step": 42430 + }, + { + "epoch": 0.9339863526304204, + "grad_norm": 1.5124605147409367, + "learning_rate": 1.1384415255313474e-07, + "loss": 0.2662, + "step": 42431 + }, + { + "epoch": 0.9340083645168391, + "grad_norm": 1.8173355314014636, + "learning_rate": 1.1376853298869229e-07, + "loss": 0.2841, + "step": 42432 + }, + { + "epoch": 0.9340303764032578, + "grad_norm": 1.7212471773799631, + "learning_rate": 1.1369293825825744e-07, + "loss": 0.3441, + "step": 42433 + }, + { + "epoch": 0.9340523882896764, + "grad_norm": 2.0588397123728046, + "learning_rate": 1.1361736836221382e-07, + "loss": 0.3333, + "step": 42434 + }, + { + "epoch": 0.9340744001760951, + "grad_norm": 1.969504140944995, + "learning_rate": 1.1354182330094721e-07, + "loss": 0.3116, + "step": 42435 + }, + { + "epoch": 0.9340964120625137, + "grad_norm": 1.5189509848770368, + "learning_rate": 1.1346630307484008e-07, + "loss": 0.3115, + "step": 42436 + }, + { + "epoch": 0.9341184239489324, + "grad_norm": 0.6206778538187626, + "learning_rate": 1.1339080768427657e-07, + "loss": 0.4456, + "step": 42437 + }, + { + "epoch": 0.9341404358353511, + "grad_norm": 2.037688118249841, + "learning_rate": 1.1331533712964027e-07, + "loss": 0.3096, + "step": 42438 + }, + { + "epoch": 0.9341624477217697, + "grad_norm": 14.025489035854072, + "learning_rate": 1.1323989141131475e-07, + "loss": 0.3021, + "step": 42439 + }, + { + "epoch": 0.9341844596081884, + "grad_norm": 0.6231159295597671, + "learning_rate": 1.1316447052968471e-07, + "loss": 0.435, + "step": 42440 + }, + { + "epoch": 0.9342064714946071, + "grad_norm": 1.555418815120766, + "learning_rate": 1.1308907448513206e-07, + "loss": 0.2816, + "step": 42441 + }, + { + "epoch": 0.9342284833810257, + "grad_norm": 2.3129997606917048, + "learning_rate": 1.1301370327803984e-07, + "loss": 0.2703, + "step": 42442 + }, + { + "epoch": 0.9342504952674444, + "grad_norm": 1.4683119149127628, + "learning_rate": 1.1293835690879218e-07, + "loss": 0.2602, + "step": 42443 + }, + { + "epoch": 0.9342725071538631, + "grad_norm": 2.026350250663857, + "learning_rate": 1.12863035377771e-07, + "loss": 0.3542, + "step": 42444 + }, + { + "epoch": 0.9342945190402817, + "grad_norm": 0.6119823945528085, + "learning_rate": 1.127877386853593e-07, + "loss": 0.4457, + "step": 42445 + }, + { + "epoch": 0.9343165309267004, + "grad_norm": 1.8149841366506432, + "learning_rate": 1.127124668319407e-07, + "loss": 0.2446, + "step": 42446 + }, + { + "epoch": 0.9343385428131191, + "grad_norm": 2.2788349364326432, + "learning_rate": 1.126372198178971e-07, + "loss": 0.2691, + "step": 42447 + }, + { + "epoch": 0.9343605546995377, + "grad_norm": 1.996440734635592, + "learning_rate": 1.1256199764361097e-07, + "loss": 0.3073, + "step": 42448 + }, + { + "epoch": 0.9343825665859564, + "grad_norm": 1.617655031476928, + "learning_rate": 1.1248680030946424e-07, + "loss": 0.266, + "step": 42449 + }, + { + "epoch": 0.9344045784723751, + "grad_norm": 1.521310867984638, + "learning_rate": 1.1241162781583992e-07, + "loss": 0.2687, + "step": 42450 + }, + { + "epoch": 0.9344265903587937, + "grad_norm": 1.6620767575868425, + "learning_rate": 1.1233648016311938e-07, + "loss": 0.2862, + "step": 42451 + }, + { + "epoch": 0.9344486022452124, + "grad_norm": 1.5494340101744033, + "learning_rate": 1.1226135735168453e-07, + "loss": 0.2364, + "step": 42452 + }, + { + "epoch": 0.9344706141316311, + "grad_norm": 15.206732647683237, + "learning_rate": 1.121862593819184e-07, + "loss": 0.2784, + "step": 42453 + }, + { + "epoch": 0.9344926260180497, + "grad_norm": 1.7462221779096354, + "learning_rate": 1.121111862542007e-07, + "loss": 0.2413, + "step": 42454 + }, + { + "epoch": 0.9345146379044684, + "grad_norm": 1.8115789665205972, + "learning_rate": 1.1203613796891444e-07, + "loss": 0.3647, + "step": 42455 + }, + { + "epoch": 0.9345366497908871, + "grad_norm": 1.6840848729786289, + "learning_rate": 1.1196111452644098e-07, + "loss": 0.2507, + "step": 42456 + }, + { + "epoch": 0.9345586616773057, + "grad_norm": 1.5643084560567264, + "learning_rate": 1.1188611592716059e-07, + "loss": 0.2824, + "step": 42457 + }, + { + "epoch": 0.9345806735637244, + "grad_norm": 4.869348275595483, + "learning_rate": 1.1181114217145573e-07, + "loss": 0.29, + "step": 42458 + }, + { + "epoch": 0.934602685450143, + "grad_norm": 1.6323014867832553, + "learning_rate": 1.1173619325970609e-07, + "loss": 0.2784, + "step": 42459 + }, + { + "epoch": 0.9346246973365617, + "grad_norm": 2.2315331169961596, + "learning_rate": 1.1166126919229358e-07, + "loss": 0.2672, + "step": 42460 + }, + { + "epoch": 0.9346467092229804, + "grad_norm": 3.9386111951061835, + "learning_rate": 1.115863699695996e-07, + "loss": 0.3218, + "step": 42461 + }, + { + "epoch": 0.934668721109399, + "grad_norm": 2.229662879497774, + "learning_rate": 1.1151149559200325e-07, + "loss": 0.3026, + "step": 42462 + }, + { + "epoch": 0.9346907329958177, + "grad_norm": 0.612367408326623, + "learning_rate": 1.1143664605988591e-07, + "loss": 0.4575, + "step": 42463 + }, + { + "epoch": 0.9347127448822364, + "grad_norm": 1.645696151962292, + "learning_rate": 1.1136182137362784e-07, + "loss": 0.2651, + "step": 42464 + }, + { + "epoch": 0.934734756768655, + "grad_norm": 1.6927072180302818, + "learning_rate": 1.1128702153360927e-07, + "loss": 0.2198, + "step": 42465 + }, + { + "epoch": 0.9347567686550737, + "grad_norm": 1.9372018765702372, + "learning_rate": 1.1121224654021101e-07, + "loss": 0.2917, + "step": 42466 + }, + { + "epoch": 0.9347787805414924, + "grad_norm": 0.634216943586641, + "learning_rate": 1.1113749639381222e-07, + "loss": 0.462, + "step": 42467 + }, + { + "epoch": 0.934800792427911, + "grad_norm": 1.531431129751311, + "learning_rate": 1.1106277109479313e-07, + "loss": 0.3277, + "step": 42468 + }, + { + "epoch": 0.9348228043143297, + "grad_norm": 1.4506146437266474, + "learning_rate": 1.1098807064353346e-07, + "loss": 0.2564, + "step": 42469 + }, + { + "epoch": 0.9348448162007484, + "grad_norm": 2.054456719416799, + "learning_rate": 1.1091339504041288e-07, + "loss": 0.3117, + "step": 42470 + }, + { + "epoch": 0.934866828087167, + "grad_norm": 0.6163329901644461, + "learning_rate": 1.1083874428581166e-07, + "loss": 0.4553, + "step": 42471 + }, + { + "epoch": 0.9348888399735857, + "grad_norm": 1.887050456520338, + "learning_rate": 1.1076411838010837e-07, + "loss": 0.3027, + "step": 42472 + }, + { + "epoch": 0.9349108518600044, + "grad_norm": 0.5908555657755481, + "learning_rate": 1.1068951732368215e-07, + "loss": 0.4275, + "step": 42473 + }, + { + "epoch": 0.934932863746423, + "grad_norm": 3.0467635894213396, + "learning_rate": 1.1061494111691217e-07, + "loss": 0.2791, + "step": 42474 + }, + { + "epoch": 0.9349548756328417, + "grad_norm": 1.6563373241033938, + "learning_rate": 1.1054038976017811e-07, + "loss": 0.1904, + "step": 42475 + }, + { + "epoch": 0.9349768875192604, + "grad_norm": 1.5243868724344527, + "learning_rate": 1.1046586325385911e-07, + "loss": 0.2362, + "step": 42476 + }, + { + "epoch": 0.934998899405679, + "grad_norm": 1.8902513231212008, + "learning_rate": 1.1039136159833319e-07, + "loss": 0.3542, + "step": 42477 + }, + { + "epoch": 0.9350209112920977, + "grad_norm": 1.5605152870661427, + "learning_rate": 1.103168847939784e-07, + "loss": 0.3776, + "step": 42478 + }, + { + "epoch": 0.9350429231785164, + "grad_norm": 2.054712654623339, + "learning_rate": 1.1024243284117497e-07, + "loss": 0.276, + "step": 42479 + }, + { + "epoch": 0.935064935064935, + "grad_norm": 2.027798999260247, + "learning_rate": 1.101680057402993e-07, + "loss": 0.2667, + "step": 42480 + }, + { + "epoch": 0.9350869469513537, + "grad_norm": 2.184424637796361, + "learning_rate": 1.1009360349173215e-07, + "loss": 0.3182, + "step": 42481 + }, + { + "epoch": 0.9351089588377723, + "grad_norm": 1.3120517459123895, + "learning_rate": 1.1001922609584992e-07, + "loss": 0.2671, + "step": 42482 + }, + { + "epoch": 0.935130970724191, + "grad_norm": 1.8692473797728046, + "learning_rate": 1.0994487355303007e-07, + "loss": 0.273, + "step": 42483 + }, + { + "epoch": 0.9351529826106098, + "grad_norm": 2.4722275319728997, + "learning_rate": 1.0987054586365231e-07, + "loss": 0.3041, + "step": 42484 + }, + { + "epoch": 0.9351749944970283, + "grad_norm": 2.0304789845693283, + "learning_rate": 1.0979624302809299e-07, + "loss": 0.272, + "step": 42485 + }, + { + "epoch": 0.935197006383447, + "grad_norm": 1.690507209656147, + "learning_rate": 1.0972196504673016e-07, + "loss": 0.3586, + "step": 42486 + }, + { + "epoch": 0.9352190182698658, + "grad_norm": 1.6098021173367207, + "learning_rate": 1.096477119199424e-07, + "loss": 0.2315, + "step": 42487 + }, + { + "epoch": 0.9352410301562843, + "grad_norm": 1.9647155567518464, + "learning_rate": 1.0957348364810493e-07, + "loss": 0.2657, + "step": 42488 + }, + { + "epoch": 0.935263042042703, + "grad_norm": 1.9994389289036136, + "learning_rate": 1.0949928023159694e-07, + "loss": 0.2758, + "step": 42489 + }, + { + "epoch": 0.9352850539291218, + "grad_norm": 0.6167551304901382, + "learning_rate": 1.0942510167079478e-07, + "loss": 0.4372, + "step": 42490 + }, + { + "epoch": 0.9353070658155404, + "grad_norm": 2.7564312982954, + "learning_rate": 1.0935094796607482e-07, + "loss": 0.2735, + "step": 42491 + }, + { + "epoch": 0.935329077701959, + "grad_norm": 1.6720216771320109, + "learning_rate": 1.0927681911781562e-07, + "loss": 0.2171, + "step": 42492 + }, + { + "epoch": 0.9353510895883778, + "grad_norm": 1.7062711302548272, + "learning_rate": 1.0920271512639247e-07, + "loss": 0.2508, + "step": 42493 + }, + { + "epoch": 0.9353731014747964, + "grad_norm": 1.4031237565676502, + "learning_rate": 1.0912863599218281e-07, + "loss": 0.32, + "step": 42494 + }, + { + "epoch": 0.9353951133612151, + "grad_norm": 1.4360062760918602, + "learning_rate": 1.0905458171556194e-07, + "loss": 0.2848, + "step": 42495 + }, + { + "epoch": 0.9354171252476338, + "grad_norm": 1.7470012111638478, + "learning_rate": 1.0898055229690784e-07, + "loss": 0.2297, + "step": 42496 + }, + { + "epoch": 0.9354391371340524, + "grad_norm": 0.6239256194933057, + "learning_rate": 1.0890654773659581e-07, + "loss": 0.4485, + "step": 42497 + }, + { + "epoch": 0.9354611490204711, + "grad_norm": 0.5993665801387931, + "learning_rate": 1.088325680350022e-07, + "loss": 0.4357, + "step": 42498 + }, + { + "epoch": 0.9354831609068898, + "grad_norm": 1.7628119809708713, + "learning_rate": 1.0875861319250337e-07, + "loss": 0.3543, + "step": 42499 + }, + { + "epoch": 0.9355051727933084, + "grad_norm": 2.2878458573581746, + "learning_rate": 1.0868468320947401e-07, + "loss": 0.2349, + "step": 42500 + }, + { + "epoch": 0.9355271846797271, + "grad_norm": 1.8633562706221107, + "learning_rate": 1.0861077808629105e-07, + "loss": 0.2299, + "step": 42501 + }, + { + "epoch": 0.9355491965661458, + "grad_norm": 0.6098703815284041, + "learning_rate": 1.0853689782333032e-07, + "loss": 0.4253, + "step": 42502 + }, + { + "epoch": 0.9355712084525644, + "grad_norm": 2.0073322194627696, + "learning_rate": 1.084630424209665e-07, + "loss": 0.3157, + "step": 42503 + }, + { + "epoch": 0.9355932203389831, + "grad_norm": 0.597725541679554, + "learning_rate": 1.0838921187957486e-07, + "loss": 0.4315, + "step": 42504 + }, + { + "epoch": 0.9356152322254018, + "grad_norm": 1.6607359586296468, + "learning_rate": 1.0831540619953063e-07, + "loss": 0.2638, + "step": 42505 + }, + { + "epoch": 0.9356372441118204, + "grad_norm": 0.62852581711646, + "learning_rate": 1.0824162538120964e-07, + "loss": 0.4659, + "step": 42506 + }, + { + "epoch": 0.9356592559982391, + "grad_norm": 2.655841849996431, + "learning_rate": 1.0816786942498659e-07, + "loss": 0.3076, + "step": 42507 + }, + { + "epoch": 0.9356812678846577, + "grad_norm": 1.6039123194999514, + "learning_rate": 1.0809413833123616e-07, + "loss": 0.2527, + "step": 42508 + }, + { + "epoch": 0.9357032797710764, + "grad_norm": 0.595929632161325, + "learning_rate": 1.0802043210033308e-07, + "loss": 0.4673, + "step": 42509 + }, + { + "epoch": 0.9357252916574951, + "grad_norm": 1.857473563980351, + "learning_rate": 1.079467507326526e-07, + "loss": 0.3451, + "step": 42510 + }, + { + "epoch": 0.9357473035439137, + "grad_norm": 1.903233366184429, + "learning_rate": 1.0787309422856774e-07, + "loss": 0.2527, + "step": 42511 + }, + { + "epoch": 0.9357693154303324, + "grad_norm": 0.6261754586873535, + "learning_rate": 1.0779946258845431e-07, + "loss": 0.4346, + "step": 42512 + }, + { + "epoch": 0.9357913273167511, + "grad_norm": 1.6744696974399969, + "learning_rate": 1.077258558126859e-07, + "loss": 0.3015, + "step": 42513 + }, + { + "epoch": 0.9358133392031697, + "grad_norm": 5.954344866459093, + "learning_rate": 1.0765227390163669e-07, + "loss": 0.3486, + "step": 42514 + }, + { + "epoch": 0.9358353510895884, + "grad_norm": 1.6275114934201313, + "learning_rate": 1.0757871685568133e-07, + "loss": 0.2792, + "step": 42515 + }, + { + "epoch": 0.9358573629760071, + "grad_norm": 1.7657070857116508, + "learning_rate": 1.0750518467519122e-07, + "loss": 0.238, + "step": 42516 + }, + { + "epoch": 0.9358793748624257, + "grad_norm": 1.2106502850316818, + "learning_rate": 1.0743167736054383e-07, + "loss": 0.2727, + "step": 42517 + }, + { + "epoch": 0.9359013867488444, + "grad_norm": 1.548510328691677, + "learning_rate": 1.0735819491210998e-07, + "loss": 0.3178, + "step": 42518 + }, + { + "epoch": 0.9359233986352631, + "grad_norm": 1.3215680467183297, + "learning_rate": 1.072847373302638e-07, + "loss": 0.2114, + "step": 42519 + }, + { + "epoch": 0.9359454105216817, + "grad_norm": 4.231564042494648, + "learning_rate": 1.0721130461537943e-07, + "loss": 0.2958, + "step": 42520 + }, + { + "epoch": 0.9359674224081004, + "grad_norm": 1.4276623997727318, + "learning_rate": 1.0713789676782882e-07, + "loss": 0.2367, + "step": 42521 + }, + { + "epoch": 0.9359894342945191, + "grad_norm": 1.4855641081055222, + "learning_rate": 1.0706451378798555e-07, + "loss": 0.2507, + "step": 42522 + }, + { + "epoch": 0.9360114461809377, + "grad_norm": 1.8651838226321757, + "learning_rate": 1.069911556762232e-07, + "loss": 0.2386, + "step": 42523 + }, + { + "epoch": 0.9360334580673564, + "grad_norm": 1.977037340661198, + "learning_rate": 1.069178224329137e-07, + "loss": 0.2255, + "step": 42524 + }, + { + "epoch": 0.9360554699537751, + "grad_norm": 0.6236014029066094, + "learning_rate": 1.0684451405843122e-07, + "loss": 0.4559, + "step": 42525 + }, + { + "epoch": 0.9360774818401937, + "grad_norm": 1.4323218176185983, + "learning_rate": 1.0677123055314598e-07, + "loss": 0.247, + "step": 42526 + }, + { + "epoch": 0.9360994937266124, + "grad_norm": 2.1552322656594347, + "learning_rate": 1.066979719174327e-07, + "loss": 0.292, + "step": 42527 + }, + { + "epoch": 0.9361215056130311, + "grad_norm": 1.871510576804918, + "learning_rate": 1.0662473815166274e-07, + "loss": 0.2647, + "step": 42528 + }, + { + "epoch": 0.9361435174994497, + "grad_norm": 9.399918866268182, + "learning_rate": 1.0655152925620804e-07, + "loss": 0.2829, + "step": 42529 + }, + { + "epoch": 0.9361655293858684, + "grad_norm": 1.3602605340529248, + "learning_rate": 1.0647834523144108e-07, + "loss": 0.2382, + "step": 42530 + }, + { + "epoch": 0.936187541272287, + "grad_norm": 1.6939567546606054, + "learning_rate": 1.0640518607773375e-07, + "loss": 0.4073, + "step": 42531 + }, + { + "epoch": 0.9362095531587057, + "grad_norm": 1.5046663094153327, + "learning_rate": 1.0633205179545747e-07, + "loss": 0.2652, + "step": 42532 + }, + { + "epoch": 0.9362315650451244, + "grad_norm": 1.7766454354049677, + "learning_rate": 1.0625894238498524e-07, + "loss": 0.3154, + "step": 42533 + }, + { + "epoch": 0.936253576931543, + "grad_norm": 1.6351385295823604, + "learning_rate": 1.0618585784668678e-07, + "loss": 0.3397, + "step": 42534 + }, + { + "epoch": 0.9362755888179617, + "grad_norm": 1.8827061371498177, + "learning_rate": 1.0611279818093567e-07, + "loss": 0.243, + "step": 42535 + }, + { + "epoch": 0.9362976007043804, + "grad_norm": 1.6070417636066114, + "learning_rate": 1.0603976338810108e-07, + "loss": 0.2894, + "step": 42536 + }, + { + "epoch": 0.936319612590799, + "grad_norm": 1.7811740243453988, + "learning_rate": 1.0596675346855545e-07, + "loss": 0.2814, + "step": 42537 + }, + { + "epoch": 0.9363416244772177, + "grad_norm": 1.8212407885011699, + "learning_rate": 1.0589376842266963e-07, + "loss": 0.3427, + "step": 42538 + }, + { + "epoch": 0.9363636363636364, + "grad_norm": 1.3601654263052896, + "learning_rate": 1.0582080825081442e-07, + "loss": 0.2285, + "step": 42539 + }, + { + "epoch": 0.936385648250055, + "grad_norm": 0.6279738694729938, + "learning_rate": 1.0574787295336064e-07, + "loss": 0.4665, + "step": 42540 + }, + { + "epoch": 0.9364076601364737, + "grad_norm": 2.760982978783503, + "learning_rate": 1.056749625306791e-07, + "loss": 0.3524, + "step": 42541 + }, + { + "epoch": 0.9364296720228924, + "grad_norm": 0.6252193442555759, + "learning_rate": 1.0560207698314063e-07, + "loss": 0.4504, + "step": 42542 + }, + { + "epoch": 0.936451683909311, + "grad_norm": 2.061832836185637, + "learning_rate": 1.0552921631111546e-07, + "loss": 0.3075, + "step": 42543 + }, + { + "epoch": 0.9364736957957297, + "grad_norm": 1.812641008233854, + "learning_rate": 1.0545638051497331e-07, + "loss": 0.3619, + "step": 42544 + }, + { + "epoch": 0.9364957076821484, + "grad_norm": 1.9418490819420442, + "learning_rate": 1.05383569595085e-07, + "loss": 0.288, + "step": 42545 + }, + { + "epoch": 0.936517719568567, + "grad_norm": 2.060135590778795, + "learning_rate": 1.0531078355182078e-07, + "loss": 0.3266, + "step": 42546 + }, + { + "epoch": 0.9365397314549857, + "grad_norm": 0.6569292779158555, + "learning_rate": 1.0523802238554926e-07, + "loss": 0.4848, + "step": 42547 + }, + { + "epoch": 0.9365617433414044, + "grad_norm": 1.612720175188928, + "learning_rate": 1.0516528609664234e-07, + "loss": 0.2914, + "step": 42548 + }, + { + "epoch": 0.936583755227823, + "grad_norm": 1.7477553083469028, + "learning_rate": 1.0509257468546863e-07, + "loss": 0.3504, + "step": 42549 + }, + { + "epoch": 0.9366057671142417, + "grad_norm": 1.8061078086509874, + "learning_rate": 1.0501988815239728e-07, + "loss": 0.3021, + "step": 42550 + }, + { + "epoch": 0.9366277790006604, + "grad_norm": 8.227958359265331, + "learning_rate": 1.0494722649779799e-07, + "loss": 0.3125, + "step": 42551 + }, + { + "epoch": 0.936649790887079, + "grad_norm": 5.974654328161342, + "learning_rate": 1.0487458972203935e-07, + "loss": 0.2852, + "step": 42552 + }, + { + "epoch": 0.9366718027734977, + "grad_norm": 1.6960083325572464, + "learning_rate": 1.0480197782549272e-07, + "loss": 0.3572, + "step": 42553 + }, + { + "epoch": 0.9366938146599163, + "grad_norm": 1.375560253609313, + "learning_rate": 1.0472939080852562e-07, + "loss": 0.2523, + "step": 42554 + }, + { + "epoch": 0.936715826546335, + "grad_norm": 1.4141053208879912, + "learning_rate": 1.0465682867150662e-07, + "loss": 0.2891, + "step": 42555 + }, + { + "epoch": 0.9367378384327537, + "grad_norm": 1.4294597604556727, + "learning_rate": 1.0458429141480486e-07, + "loss": 0.3354, + "step": 42556 + }, + { + "epoch": 0.9367598503191723, + "grad_norm": 2.2675803467935958, + "learning_rate": 1.0451177903878951e-07, + "loss": 0.3656, + "step": 42557 + }, + { + "epoch": 0.936781862205591, + "grad_norm": 1.5609584566723602, + "learning_rate": 1.0443929154382859e-07, + "loss": 0.308, + "step": 42558 + }, + { + "epoch": 0.9368038740920097, + "grad_norm": 2.5500631184784814, + "learning_rate": 1.0436682893029071e-07, + "loss": 0.2946, + "step": 42559 + }, + { + "epoch": 0.9368258859784283, + "grad_norm": 1.7391724795232133, + "learning_rate": 1.042943911985439e-07, + "loss": 0.2394, + "step": 42560 + }, + { + "epoch": 0.936847897864847, + "grad_norm": 0.6083997853210863, + "learning_rate": 1.0422197834895675e-07, + "loss": 0.4492, + "step": 42561 + }, + { + "epoch": 0.9368699097512657, + "grad_norm": 0.6149320078711195, + "learning_rate": 1.0414959038189731e-07, + "loss": 0.4536, + "step": 42562 + }, + { + "epoch": 0.9368919216376843, + "grad_norm": 1.6314071261611833, + "learning_rate": 1.0407722729773251e-07, + "loss": 0.3165, + "step": 42563 + }, + { + "epoch": 0.936913933524103, + "grad_norm": 2.0675256818363974, + "learning_rate": 1.0400488909683204e-07, + "loss": 0.3188, + "step": 42564 + }, + { + "epoch": 0.9369359454105217, + "grad_norm": 1.9758784383801677, + "learning_rate": 1.0393257577956118e-07, + "loss": 0.3442, + "step": 42565 + }, + { + "epoch": 0.9369579572969403, + "grad_norm": 2.3168355457691967, + "learning_rate": 1.0386028734628961e-07, + "loss": 0.1998, + "step": 42566 + }, + { + "epoch": 0.936979969183359, + "grad_norm": 1.6193228555346757, + "learning_rate": 1.0378802379738318e-07, + "loss": 0.2626, + "step": 42567 + }, + { + "epoch": 0.9370019810697777, + "grad_norm": 1.7318236200152277, + "learning_rate": 1.037157851332099e-07, + "loss": 0.2363, + "step": 42568 + }, + { + "epoch": 0.9370239929561963, + "grad_norm": 1.965341389326331, + "learning_rate": 1.0364357135413728e-07, + "loss": 0.2827, + "step": 42569 + }, + { + "epoch": 0.937046004842615, + "grad_norm": 1.5647952281925295, + "learning_rate": 1.0357138246053167e-07, + "loss": 0.3183, + "step": 42570 + }, + { + "epoch": 0.9370680167290337, + "grad_norm": 2.085628808499059, + "learning_rate": 1.0349921845276001e-07, + "loss": 0.3252, + "step": 42571 + }, + { + "epoch": 0.9370900286154523, + "grad_norm": 3.0225553926864275, + "learning_rate": 1.0342707933118923e-07, + "loss": 0.273, + "step": 42572 + }, + { + "epoch": 0.937112040501871, + "grad_norm": 2.6078423253702345, + "learning_rate": 1.0335496509618569e-07, + "loss": 0.3186, + "step": 42573 + }, + { + "epoch": 0.9371340523882897, + "grad_norm": 2.445910405394295, + "learning_rate": 1.0328287574811691e-07, + "loss": 0.3481, + "step": 42574 + }, + { + "epoch": 0.9371560642747083, + "grad_norm": 1.772051014968765, + "learning_rate": 1.0321081128734756e-07, + "loss": 0.2416, + "step": 42575 + }, + { + "epoch": 0.937178076161127, + "grad_norm": 1.5252337687941822, + "learning_rate": 1.031387717142457e-07, + "loss": 0.2782, + "step": 42576 + }, + { + "epoch": 0.9372000880475456, + "grad_norm": 1.6410488340376956, + "learning_rate": 1.0306675702917656e-07, + "loss": 0.2523, + "step": 42577 + }, + { + "epoch": 0.9372220999339643, + "grad_norm": 1.6750401351546067, + "learning_rate": 1.0299476723250545e-07, + "loss": 0.2641, + "step": 42578 + }, + { + "epoch": 0.937244111820383, + "grad_norm": 0.6348885127065418, + "learning_rate": 1.0292280232460039e-07, + "loss": 0.4627, + "step": 42579 + }, + { + "epoch": 0.9372661237068016, + "grad_norm": 1.732823161850584, + "learning_rate": 1.0285086230582497e-07, + "loss": 0.2886, + "step": 42580 + }, + { + "epoch": 0.9372881355932203, + "grad_norm": 1.8381500871353553, + "learning_rate": 1.0277894717654613e-07, + "loss": 0.2712, + "step": 42581 + }, + { + "epoch": 0.937310147479639, + "grad_norm": 1.5602877285452066, + "learning_rate": 1.0270705693712857e-07, + "loss": 0.2843, + "step": 42582 + }, + { + "epoch": 0.9373321593660576, + "grad_norm": 1.7973195558734727, + "learning_rate": 1.0263519158793756e-07, + "loss": 0.3406, + "step": 42583 + }, + { + "epoch": 0.9373541712524763, + "grad_norm": 2.0530926907659994, + "learning_rate": 1.0256335112933946e-07, + "loss": 0.3048, + "step": 42584 + }, + { + "epoch": 0.937376183138895, + "grad_norm": 2.581284245190428, + "learning_rate": 1.0249153556169845e-07, + "loss": 0.3004, + "step": 42585 + }, + { + "epoch": 0.9373981950253136, + "grad_norm": 1.5782594997947783, + "learning_rate": 1.0241974488537975e-07, + "loss": 0.3777, + "step": 42586 + }, + { + "epoch": 0.9374202069117323, + "grad_norm": 1.8736666147918375, + "learning_rate": 1.0234797910074867e-07, + "loss": 0.4201, + "step": 42587 + }, + { + "epoch": 0.937442218798151, + "grad_norm": 1.6586183657075386, + "learning_rate": 1.0227623820816879e-07, + "loss": 0.2952, + "step": 42588 + }, + { + "epoch": 0.9374642306845696, + "grad_norm": 1.8119895784049167, + "learning_rate": 1.0220452220800647e-07, + "loss": 0.3362, + "step": 42589 + }, + { + "epoch": 0.9374862425709883, + "grad_norm": 0.6068824635771997, + "learning_rate": 1.0213283110062478e-07, + "loss": 0.4449, + "step": 42590 + }, + { + "epoch": 0.937508254457407, + "grad_norm": 1.5381058042790519, + "learning_rate": 1.020611648863884e-07, + "loss": 0.2331, + "step": 42591 + }, + { + "epoch": 0.9375302663438256, + "grad_norm": 2.1226536098426205, + "learning_rate": 1.0198952356566261e-07, + "loss": 0.3514, + "step": 42592 + }, + { + "epoch": 0.9375522782302443, + "grad_norm": 2.151426765983141, + "learning_rate": 1.0191790713880933e-07, + "loss": 0.2622, + "step": 42593 + }, + { + "epoch": 0.937574290116663, + "grad_norm": 1.7461167688460195, + "learning_rate": 1.018463156061944e-07, + "loss": 0.2644, + "step": 42594 + }, + { + "epoch": 0.9375963020030816, + "grad_norm": 1.52249581329358, + "learning_rate": 1.017747489681814e-07, + "loss": 0.2519, + "step": 42595 + }, + { + "epoch": 0.9376183138895003, + "grad_norm": 1.7540706090144422, + "learning_rate": 1.0170320722513338e-07, + "loss": 0.2733, + "step": 42596 + }, + { + "epoch": 0.937640325775919, + "grad_norm": 2.947599809172061, + "learning_rate": 1.0163169037741504e-07, + "loss": 0.2732, + "step": 42597 + }, + { + "epoch": 0.9376623376623376, + "grad_norm": 1.4317651176962776, + "learning_rate": 1.0156019842538833e-07, + "loss": 0.299, + "step": 42598 + }, + { + "epoch": 0.9376843495487563, + "grad_norm": 0.5989075643461006, + "learning_rate": 1.0148873136941794e-07, + "loss": 0.4352, + "step": 42599 + }, + { + "epoch": 0.9377063614351749, + "grad_norm": 1.6581574526236937, + "learning_rate": 1.0141728920986638e-07, + "loss": 0.2843, + "step": 42600 + }, + { + "epoch": 0.9377283733215936, + "grad_norm": 3.067613724566065, + "learning_rate": 1.0134587194709722e-07, + "loss": 0.2881, + "step": 42601 + }, + { + "epoch": 0.9377503852080123, + "grad_norm": 1.8046103555660906, + "learning_rate": 1.0127447958147352e-07, + "loss": 0.2343, + "step": 42602 + }, + { + "epoch": 0.9377723970944309, + "grad_norm": 1.7404081356939287, + "learning_rate": 1.0120311211335721e-07, + "loss": 0.3034, + "step": 42603 + }, + { + "epoch": 0.9377944089808496, + "grad_norm": 0.6243993555937588, + "learning_rate": 1.0113176954311132e-07, + "loss": 0.457, + "step": 42604 + }, + { + "epoch": 0.9378164208672684, + "grad_norm": 2.249223016729344, + "learning_rate": 1.0106045187109947e-07, + "loss": 0.2426, + "step": 42605 + }, + { + "epoch": 0.937838432753687, + "grad_norm": 1.8920261361534454, + "learning_rate": 1.0098915909768303e-07, + "loss": 0.2814, + "step": 42606 + }, + { + "epoch": 0.9378604446401057, + "grad_norm": 2.8542624836197508, + "learning_rate": 1.0091789122322504e-07, + "loss": 0.2991, + "step": 42607 + }, + { + "epoch": 0.9378824565265244, + "grad_norm": 0.6161485188782667, + "learning_rate": 1.0084664824808688e-07, + "loss": 0.4514, + "step": 42608 + }, + { + "epoch": 0.937904468412943, + "grad_norm": 0.6116917244127009, + "learning_rate": 1.0077543017263103e-07, + "loss": 0.4546, + "step": 42609 + }, + { + "epoch": 0.9379264802993617, + "grad_norm": 2.5439530544364333, + "learning_rate": 1.0070423699722054e-07, + "loss": 0.3339, + "step": 42610 + }, + { + "epoch": 0.9379484921857804, + "grad_norm": 1.718224199788581, + "learning_rate": 1.0063306872221512e-07, + "loss": 0.242, + "step": 42611 + }, + { + "epoch": 0.937970504072199, + "grad_norm": 1.5759695952897643, + "learning_rate": 1.0056192534797837e-07, + "loss": 0.2071, + "step": 42612 + }, + { + "epoch": 0.9379925159586177, + "grad_norm": 1.8207229450866016, + "learning_rate": 1.0049080687487112e-07, + "loss": 0.257, + "step": 42613 + }, + { + "epoch": 0.9380145278450364, + "grad_norm": 1.9109517702686183, + "learning_rate": 1.0041971330325418e-07, + "loss": 0.2086, + "step": 42614 + }, + { + "epoch": 0.938036539731455, + "grad_norm": 1.9082431332038654, + "learning_rate": 1.0034864463349004e-07, + "loss": 0.2657, + "step": 42615 + }, + { + "epoch": 0.9380585516178737, + "grad_norm": 1.1653050009392039, + "learning_rate": 1.0027760086593951e-07, + "loss": 0.1355, + "step": 42616 + }, + { + "epoch": 0.9380805635042924, + "grad_norm": 2.3451946763472016, + "learning_rate": 1.0020658200096344e-07, + "loss": 0.2828, + "step": 42617 + }, + { + "epoch": 0.938102575390711, + "grad_norm": 1.7420682786920751, + "learning_rate": 1.0013558803892321e-07, + "loss": 0.3429, + "step": 42618 + }, + { + "epoch": 0.9381245872771297, + "grad_norm": 1.7892676427688146, + "learning_rate": 1.0006461898017794e-07, + "loss": 0.3582, + "step": 42619 + }, + { + "epoch": 0.9381465991635484, + "grad_norm": 1.873488068740852, + "learning_rate": 9.999367482509126e-08, + "loss": 0.2677, + "step": 42620 + }, + { + "epoch": 0.938168611049967, + "grad_norm": 2.192188591264125, + "learning_rate": 9.992275557402176e-08, + "loss": 0.312, + "step": 42621 + }, + { + "epoch": 0.9381906229363857, + "grad_norm": 1.5087124482150422, + "learning_rate": 9.98518612273297e-08, + "loss": 0.2747, + "step": 42622 + }, + { + "epoch": 0.9382126348228043, + "grad_norm": 1.38348157038693, + "learning_rate": 9.978099178537704e-08, + "loss": 0.2753, + "step": 42623 + }, + { + "epoch": 0.938234646709223, + "grad_norm": 1.6362557047495485, + "learning_rate": 9.97101472485218e-08, + "loss": 0.2593, + "step": 42624 + }, + { + "epoch": 0.9382566585956417, + "grad_norm": 1.6261838255409298, + "learning_rate": 9.963932761712591e-08, + "loss": 0.2032, + "step": 42625 + }, + { + "epoch": 0.9382786704820603, + "grad_norm": 1.9039220125132075, + "learning_rate": 9.956853289154856e-08, + "loss": 0.2984, + "step": 42626 + }, + { + "epoch": 0.938300682368479, + "grad_norm": 2.0671643706536766, + "learning_rate": 9.949776307214943e-08, + "loss": 0.3483, + "step": 42627 + }, + { + "epoch": 0.9383226942548977, + "grad_norm": 2.046611537203746, + "learning_rate": 9.942701815928879e-08, + "loss": 0.2966, + "step": 42628 + }, + { + "epoch": 0.9383447061413163, + "grad_norm": 0.6019825404543158, + "learning_rate": 9.93562981533247e-08, + "loss": 0.4401, + "step": 42629 + }, + { + "epoch": 0.938366718027735, + "grad_norm": 1.4485055039432169, + "learning_rate": 9.928560305461855e-08, + "loss": 0.3525, + "step": 42630 + }, + { + "epoch": 0.9383887299141537, + "grad_norm": 1.681584867422611, + "learning_rate": 9.921493286352834e-08, + "loss": 0.2767, + "step": 42631 + }, + { + "epoch": 0.9384107418005723, + "grad_norm": 1.76185928144746, + "learning_rate": 9.914428758041383e-08, + "loss": 0.3099, + "step": 42632 + }, + { + "epoch": 0.938432753686991, + "grad_norm": 1.6014139760409791, + "learning_rate": 9.907366720563416e-08, + "loss": 0.2532, + "step": 42633 + }, + { + "epoch": 0.9384547655734097, + "grad_norm": 0.6743192930849473, + "learning_rate": 9.900307173954737e-08, + "loss": 0.4503, + "step": 42634 + }, + { + "epoch": 0.9384767774598283, + "grad_norm": 0.6194432724081808, + "learning_rate": 9.893250118251319e-08, + "loss": 0.4504, + "step": 42635 + }, + { + "epoch": 0.938498789346247, + "grad_norm": 1.4475684061724667, + "learning_rate": 9.88619555348902e-08, + "loss": 0.3016, + "step": 42636 + }, + { + "epoch": 0.9385208012326657, + "grad_norm": 2.8201520280562264, + "learning_rate": 9.87914347970359e-08, + "loss": 0.272, + "step": 42637 + }, + { + "epoch": 0.9385428131190843, + "grad_norm": 1.6642171573754312, + "learning_rate": 9.872093896931056e-08, + "loss": 0.3076, + "step": 42638 + }, + { + "epoch": 0.938564825005503, + "grad_norm": 0.6274743913257766, + "learning_rate": 9.865046805207113e-08, + "loss": 0.4455, + "step": 42639 + }, + { + "epoch": 0.9385868368919217, + "grad_norm": 1.495624523034488, + "learning_rate": 9.858002204567563e-08, + "loss": 0.3638, + "step": 42640 + }, + { + "epoch": 0.9386088487783403, + "grad_norm": 1.726855524190406, + "learning_rate": 9.850960095048323e-08, + "loss": 0.2568, + "step": 42641 + }, + { + "epoch": 0.938630860664759, + "grad_norm": 1.9588400443081329, + "learning_rate": 9.843920476685032e-08, + "loss": 0.2974, + "step": 42642 + }, + { + "epoch": 0.9386528725511777, + "grad_norm": 0.5976062499162482, + "learning_rate": 9.836883349513659e-08, + "loss": 0.4387, + "step": 42643 + }, + { + "epoch": 0.9386748844375963, + "grad_norm": 2.5707265347838035, + "learning_rate": 9.82984871356979e-08, + "loss": 0.2435, + "step": 42644 + }, + { + "epoch": 0.938696896324015, + "grad_norm": 2.228997115154648, + "learning_rate": 9.822816568889227e-08, + "loss": 0.3129, + "step": 42645 + }, + { + "epoch": 0.9387189082104337, + "grad_norm": 2.1541239069312033, + "learning_rate": 9.81578691550783e-08, + "loss": 0.2619, + "step": 42646 + }, + { + "epoch": 0.9387409200968523, + "grad_norm": 2.947335544010402, + "learning_rate": 9.808759753461128e-08, + "loss": 0.2748, + "step": 42647 + }, + { + "epoch": 0.938762931983271, + "grad_norm": 0.6158622248682516, + "learning_rate": 9.801735082785036e-08, + "loss": 0.4305, + "step": 42648 + }, + { + "epoch": 0.9387849438696896, + "grad_norm": 6.60382649075575, + "learning_rate": 9.79471290351508e-08, + "loss": 0.3023, + "step": 42649 + }, + { + "epoch": 0.9388069557561083, + "grad_norm": 2.439350812007127, + "learning_rate": 9.787693215687066e-08, + "loss": 0.2663, + "step": 42650 + }, + { + "epoch": 0.938828967642527, + "grad_norm": 1.6650250074760233, + "learning_rate": 9.780676019336632e-08, + "loss": 0.2999, + "step": 42651 + }, + { + "epoch": 0.9388509795289456, + "grad_norm": 3.3940760826108316, + "learning_rate": 9.773661314499471e-08, + "loss": 0.3823, + "step": 42652 + }, + { + "epoch": 0.9388729914153643, + "grad_norm": 1.366250887275851, + "learning_rate": 9.766649101211223e-08, + "loss": 0.2218, + "step": 42653 + }, + { + "epoch": 0.938895003301783, + "grad_norm": 2.2079999152445007, + "learning_rate": 9.759639379507468e-08, + "loss": 0.3398, + "step": 42654 + }, + { + "epoch": 0.9389170151882016, + "grad_norm": 0.5936300424482678, + "learning_rate": 9.752632149423847e-08, + "loss": 0.433, + "step": 42655 + }, + { + "epoch": 0.9389390270746203, + "grad_norm": 1.6316526318432587, + "learning_rate": 9.745627410996106e-08, + "loss": 0.3148, + "step": 42656 + }, + { + "epoch": 0.938961038961039, + "grad_norm": 1.4216448894562201, + "learning_rate": 9.738625164259663e-08, + "loss": 0.2474, + "step": 42657 + }, + { + "epoch": 0.9389830508474576, + "grad_norm": 1.5375795059921769, + "learning_rate": 9.731625409250212e-08, + "loss": 0.2947, + "step": 42658 + }, + { + "epoch": 0.9390050627338763, + "grad_norm": 1.7187390225194197, + "learning_rate": 9.724628146003334e-08, + "loss": 0.3588, + "step": 42659 + }, + { + "epoch": 0.939027074620295, + "grad_norm": 1.9693505757813123, + "learning_rate": 9.717633374554558e-08, + "loss": 0.3018, + "step": 42660 + }, + { + "epoch": 0.9390490865067136, + "grad_norm": 1.8725719863129766, + "learning_rate": 9.71064109493941e-08, + "loss": 0.2618, + "step": 42661 + }, + { + "epoch": 0.9390710983931323, + "grad_norm": 1.5694684315062672, + "learning_rate": 9.703651307193474e-08, + "loss": 0.1539, + "step": 42662 + }, + { + "epoch": 0.939093110279551, + "grad_norm": 1.4845052235787302, + "learning_rate": 9.696664011352275e-08, + "loss": 0.2611, + "step": 42663 + }, + { + "epoch": 0.9391151221659696, + "grad_norm": 1.3520158180860609, + "learning_rate": 9.689679207451286e-08, + "loss": 0.2716, + "step": 42664 + }, + { + "epoch": 0.9391371340523883, + "grad_norm": 2.667568954960009, + "learning_rate": 9.682696895526033e-08, + "loss": 0.2796, + "step": 42665 + }, + { + "epoch": 0.939159145938807, + "grad_norm": 2.5619940752292667, + "learning_rate": 9.67571707561199e-08, + "loss": 0.2675, + "step": 42666 + }, + { + "epoch": 0.9391811578252256, + "grad_norm": 0.6585503483347301, + "learning_rate": 9.668739747744737e-08, + "loss": 0.4686, + "step": 42667 + }, + { + "epoch": 0.9392031697116443, + "grad_norm": 1.3871955120496617, + "learning_rate": 9.661764911959526e-08, + "loss": 0.257, + "step": 42668 + }, + { + "epoch": 0.939225181598063, + "grad_norm": 1.4563109158652277, + "learning_rate": 9.654792568291993e-08, + "loss": 0.258, + "step": 42669 + }, + { + "epoch": 0.9392471934844816, + "grad_norm": 1.9128285360072455, + "learning_rate": 9.647822716777444e-08, + "loss": 0.3648, + "step": 42670 + }, + { + "epoch": 0.9392692053709003, + "grad_norm": 2.0351719088470994, + "learning_rate": 9.640855357451407e-08, + "loss": 0.2542, + "step": 42671 + }, + { + "epoch": 0.9392912172573189, + "grad_norm": 1.6436624052700848, + "learning_rate": 9.63389049034924e-08, + "loss": 0.349, + "step": 42672 + }, + { + "epoch": 0.9393132291437376, + "grad_norm": 1.5681527939259605, + "learning_rate": 9.626928115506362e-08, + "loss": 0.2282, + "step": 42673 + }, + { + "epoch": 0.9393352410301563, + "grad_norm": 3.76913686280149, + "learning_rate": 9.619968232958133e-08, + "loss": 0.3305, + "step": 42674 + }, + { + "epoch": 0.9393572529165749, + "grad_norm": 1.4308801319328874, + "learning_rate": 9.613010842739967e-08, + "loss": 0.2995, + "step": 42675 + }, + { + "epoch": 0.9393792648029936, + "grad_norm": 3.9331818805609338, + "learning_rate": 9.606055944887172e-08, + "loss": 0.3578, + "step": 42676 + }, + { + "epoch": 0.9394012766894123, + "grad_norm": 1.855535769448059, + "learning_rate": 9.599103539435106e-08, + "loss": 0.2694, + "step": 42677 + }, + { + "epoch": 0.9394232885758309, + "grad_norm": 1.4917518178560158, + "learning_rate": 9.592153626419131e-08, + "loss": 0.3184, + "step": 42678 + }, + { + "epoch": 0.9394453004622496, + "grad_norm": 2.457879640350879, + "learning_rate": 9.585206205874609e-08, + "loss": 0.4046, + "step": 42679 + }, + { + "epoch": 0.9394673123486683, + "grad_norm": 0.6276142567128089, + "learning_rate": 9.578261277836731e-08, + "loss": 0.4684, + "step": 42680 + }, + { + "epoch": 0.9394893242350869, + "grad_norm": 0.6272764791949824, + "learning_rate": 9.571318842340915e-08, + "loss": 0.4198, + "step": 42681 + }, + { + "epoch": 0.9395113361215056, + "grad_norm": 1.9134279198219222, + "learning_rate": 9.564378899422355e-08, + "loss": 0.3305, + "step": 42682 + }, + { + "epoch": 0.9395333480079243, + "grad_norm": 0.6224728085609798, + "learning_rate": 9.557441449116411e-08, + "loss": 0.4384, + "step": 42683 + }, + { + "epoch": 0.9395553598943429, + "grad_norm": 1.8341344031646036, + "learning_rate": 9.550506491458278e-08, + "loss": 0.2703, + "step": 42684 + }, + { + "epoch": 0.9395773717807616, + "grad_norm": 3.5492901198589983, + "learning_rate": 9.543574026483149e-08, + "loss": 0.3555, + "step": 42685 + }, + { + "epoch": 0.9395993836671803, + "grad_norm": 0.6079484991010324, + "learning_rate": 9.536644054226385e-08, + "loss": 0.429, + "step": 42686 + }, + { + "epoch": 0.9396213955535989, + "grad_norm": 4.108347669509302, + "learning_rate": 9.52971657472318e-08, + "loss": 0.3775, + "step": 42687 + }, + { + "epoch": 0.9396434074400176, + "grad_norm": 0.6105658068492856, + "learning_rate": 9.522791588008618e-08, + "loss": 0.4197, + "step": 42688 + }, + { + "epoch": 0.9396654193264363, + "grad_norm": 1.5120398673450577, + "learning_rate": 9.515869094118113e-08, + "loss": 0.265, + "step": 42689 + }, + { + "epoch": 0.9396874312128549, + "grad_norm": 17.27377137447791, + "learning_rate": 9.50894909308664e-08, + "loss": 0.3849, + "step": 42690 + }, + { + "epoch": 0.9397094430992736, + "grad_norm": 1.8049651423609407, + "learning_rate": 9.502031584949444e-08, + "loss": 0.213, + "step": 42691 + }, + { + "epoch": 0.9397314549856923, + "grad_norm": 2.8391274727821387, + "learning_rate": 9.495116569741725e-08, + "loss": 0.2929, + "step": 42692 + }, + { + "epoch": 0.9397534668721109, + "grad_norm": 1.332483215892118, + "learning_rate": 9.488204047498561e-08, + "loss": 0.245, + "step": 42693 + }, + { + "epoch": 0.9397754787585296, + "grad_norm": 0.6176328982652448, + "learning_rate": 9.481294018255149e-08, + "loss": 0.4475, + "step": 42694 + }, + { + "epoch": 0.9397974906449482, + "grad_norm": 1.4333020687546951, + "learning_rate": 9.474386482046626e-08, + "loss": 0.2991, + "step": 42695 + }, + { + "epoch": 0.9398195025313669, + "grad_norm": 1.6534436385209739, + "learning_rate": 9.467481438907967e-08, + "loss": 0.3142, + "step": 42696 + }, + { + "epoch": 0.9398415144177856, + "grad_norm": 1.5857432120822168, + "learning_rate": 9.460578888874361e-08, + "loss": 0.2588, + "step": 42697 + }, + { + "epoch": 0.9398635263042042, + "grad_norm": 1.7200333646776544, + "learning_rate": 9.453678831980894e-08, + "loss": 0.266, + "step": 42698 + }, + { + "epoch": 0.9398855381906229, + "grad_norm": 0.6150620083701291, + "learning_rate": 9.44678126826265e-08, + "loss": 0.4512, + "step": 42699 + }, + { + "epoch": 0.9399075500770416, + "grad_norm": 0.6111885830398778, + "learning_rate": 9.439886197754655e-08, + "loss": 0.4393, + "step": 42700 + }, + { + "epoch": 0.9399295619634602, + "grad_norm": 2.64755985533031, + "learning_rate": 9.432993620491938e-08, + "loss": 0.3592, + "step": 42701 + }, + { + "epoch": 0.9399515738498789, + "grad_norm": 1.4723638820018652, + "learning_rate": 9.426103536509523e-08, + "loss": 0.3173, + "step": 42702 + }, + { + "epoch": 0.9399735857362976, + "grad_norm": 5.60091429803701, + "learning_rate": 9.419215945842497e-08, + "loss": 0.3141, + "step": 42703 + }, + { + "epoch": 0.9399955976227162, + "grad_norm": 1.9362226815919386, + "learning_rate": 9.412330848525774e-08, + "loss": 0.3176, + "step": 42704 + }, + { + "epoch": 0.9400176095091349, + "grad_norm": 1.7010038127788185, + "learning_rate": 9.405448244594439e-08, + "loss": 0.3275, + "step": 42705 + }, + { + "epoch": 0.9400396213955536, + "grad_norm": 1.5091778024687823, + "learning_rate": 9.398568134083352e-08, + "loss": 0.269, + "step": 42706 + }, + { + "epoch": 0.9400616332819722, + "grad_norm": 1.7196868745600857, + "learning_rate": 9.391690517027597e-08, + "loss": 0.2958, + "step": 42707 + }, + { + "epoch": 0.9400836451683909, + "grad_norm": 1.3757219963511633, + "learning_rate": 9.384815393462143e-08, + "loss": 0.2185, + "step": 42708 + }, + { + "epoch": 0.9401056570548096, + "grad_norm": 2.0364972773863412, + "learning_rate": 9.377942763421799e-08, + "loss": 0.395, + "step": 42709 + }, + { + "epoch": 0.9401276689412282, + "grad_norm": 1.7781631730948282, + "learning_rate": 9.371072626941647e-08, + "loss": 0.3735, + "step": 42710 + }, + { + "epoch": 0.9401496808276469, + "grad_norm": 11.208163358484954, + "learning_rate": 9.364204984056435e-08, + "loss": 0.2656, + "step": 42711 + }, + { + "epoch": 0.9401716927140656, + "grad_norm": 2.810697300137521, + "learning_rate": 9.357339834801249e-08, + "loss": 0.3226, + "step": 42712 + }, + { + "epoch": 0.9401937046004842, + "grad_norm": 2.062781787555838, + "learning_rate": 9.350477179210837e-08, + "loss": 0.303, + "step": 42713 + }, + { + "epoch": 0.9402157164869029, + "grad_norm": 0.6227316053897207, + "learning_rate": 9.343617017320172e-08, + "loss": 0.4436, + "step": 42714 + }, + { + "epoch": 0.9402377283733216, + "grad_norm": 1.672800162174571, + "learning_rate": 9.336759349164115e-08, + "loss": 0.2574, + "step": 42715 + }, + { + "epoch": 0.9402597402597402, + "grad_norm": 2.558875718938501, + "learning_rate": 9.329904174777416e-08, + "loss": 0.2777, + "step": 42716 + }, + { + "epoch": 0.940281752146159, + "grad_norm": 4.3743723084438955, + "learning_rate": 9.32305149419499e-08, + "loss": 0.3158, + "step": 42717 + }, + { + "epoch": 0.9403037640325775, + "grad_norm": 1.5741175703037487, + "learning_rate": 9.316201307451756e-08, + "loss": 0.229, + "step": 42718 + }, + { + "epoch": 0.9403257759189962, + "grad_norm": 1.5519363587009907, + "learning_rate": 9.309353614582351e-08, + "loss": 0.3151, + "step": 42719 + }, + { + "epoch": 0.940347787805415, + "grad_norm": 1.612921119000029, + "learning_rate": 9.302508415621691e-08, + "loss": 0.3131, + "step": 42720 + }, + { + "epoch": 0.9403697996918335, + "grad_norm": 1.5718570934545923, + "learning_rate": 9.295665710604584e-08, + "loss": 0.3147, + "step": 42721 + }, + { + "epoch": 0.9403918115782522, + "grad_norm": 0.6389956594313306, + "learning_rate": 9.288825499565668e-08, + "loss": 0.4507, + "step": 42722 + }, + { + "epoch": 0.940413823464671, + "grad_norm": 2.192259281110741, + "learning_rate": 9.281987782539858e-08, + "loss": 0.2336, + "step": 42723 + }, + { + "epoch": 0.9404358353510895, + "grad_norm": 1.4892857254257734, + "learning_rate": 9.275152559561851e-08, + "loss": 0.2527, + "step": 42724 + }, + { + "epoch": 0.9404578472375082, + "grad_norm": 2.2471738954545506, + "learning_rate": 9.268319830666395e-08, + "loss": 0.3223, + "step": 42725 + }, + { + "epoch": 0.940479859123927, + "grad_norm": 1.8931944259573856, + "learning_rate": 9.26148959588824e-08, + "loss": 0.2838, + "step": 42726 + }, + { + "epoch": 0.9405018710103455, + "grad_norm": 1.7308652162512168, + "learning_rate": 9.25466185526197e-08, + "loss": 0.27, + "step": 42727 + }, + { + "epoch": 0.9405238828967643, + "grad_norm": 2.08768858168692, + "learning_rate": 9.247836608822503e-08, + "loss": 0.2768, + "step": 42728 + }, + { + "epoch": 0.940545894783183, + "grad_norm": 1.5903707170670087, + "learning_rate": 9.241013856604308e-08, + "loss": 0.2288, + "step": 42729 + }, + { + "epoch": 0.9405679066696016, + "grad_norm": 1.912883987480718, + "learning_rate": 9.23419359864225e-08, + "loss": 0.2873, + "step": 42730 + }, + { + "epoch": 0.9405899185560203, + "grad_norm": 0.5952291542322066, + "learning_rate": 9.227375834970909e-08, + "loss": 0.43, + "step": 42731 + }, + { + "epoch": 0.940611930442439, + "grad_norm": 1.6257406444994555, + "learning_rate": 9.22056056562487e-08, + "loss": 0.2841, + "step": 42732 + }, + { + "epoch": 0.9406339423288576, + "grad_norm": 2.2192727396994245, + "learning_rate": 9.213747790638883e-08, + "loss": 0.3768, + "step": 42733 + }, + { + "epoch": 0.9406559542152763, + "grad_norm": 2.357820068490855, + "learning_rate": 9.206937510047476e-08, + "loss": 0.3214, + "step": 42734 + }, + { + "epoch": 0.940677966101695, + "grad_norm": 1.9590054991887322, + "learning_rate": 9.200129723885343e-08, + "loss": 0.2749, + "step": 42735 + }, + { + "epoch": 0.9406999779881136, + "grad_norm": 0.6089417047907651, + "learning_rate": 9.193324432187123e-08, + "loss": 0.4418, + "step": 42736 + }, + { + "epoch": 0.9407219898745323, + "grad_norm": 1.538515663961121, + "learning_rate": 9.186521634987233e-08, + "loss": 0.2934, + "step": 42737 + }, + { + "epoch": 0.940744001760951, + "grad_norm": 1.7199442029318026, + "learning_rate": 9.179721332320369e-08, + "loss": 0.3012, + "step": 42738 + }, + { + "epoch": 0.9407660136473696, + "grad_norm": 2.376244289902256, + "learning_rate": 9.172923524221056e-08, + "loss": 0.3085, + "step": 42739 + }, + { + "epoch": 0.9407880255337883, + "grad_norm": 1.8623835430492115, + "learning_rate": 9.166128210723934e-08, + "loss": 0.2793, + "step": 42740 + }, + { + "epoch": 0.9408100374202069, + "grad_norm": 1.723112776581435, + "learning_rate": 9.159335391863422e-08, + "loss": 0.3224, + "step": 42741 + }, + { + "epoch": 0.9408320493066256, + "grad_norm": 1.5404267144802959, + "learning_rate": 9.152545067674101e-08, + "loss": 0.2598, + "step": 42742 + }, + { + "epoch": 0.9408540611930443, + "grad_norm": 1.6269484574641793, + "learning_rate": 9.145757238190445e-08, + "loss": 0.2533, + "step": 42743 + }, + { + "epoch": 0.9408760730794629, + "grad_norm": 0.586059798165962, + "learning_rate": 9.138971903447036e-08, + "loss": 0.4167, + "step": 42744 + }, + { + "epoch": 0.9408980849658816, + "grad_norm": 2.0684820372612993, + "learning_rate": 9.132189063478236e-08, + "loss": 0.3057, + "step": 42745 + }, + { + "epoch": 0.9409200968523003, + "grad_norm": 1.8441218167362927, + "learning_rate": 9.12540871831863e-08, + "loss": 0.2264, + "step": 42746 + }, + { + "epoch": 0.9409421087387189, + "grad_norm": 1.6215426724358706, + "learning_rate": 9.118630868002576e-08, + "loss": 0.2739, + "step": 42747 + }, + { + "epoch": 0.9409641206251376, + "grad_norm": 0.6239827554467543, + "learning_rate": 9.111855512564605e-08, + "loss": 0.4681, + "step": 42748 + }, + { + "epoch": 0.9409861325115563, + "grad_norm": 3.372804668260632, + "learning_rate": 9.105082652039187e-08, + "loss": 0.3584, + "step": 42749 + }, + { + "epoch": 0.9410081443979749, + "grad_norm": 2.1188761403307583, + "learning_rate": 9.098312286460575e-08, + "loss": 0.305, + "step": 42750 + }, + { + "epoch": 0.9410301562843936, + "grad_norm": 4.299072723858269, + "learning_rate": 9.091544415863406e-08, + "loss": 0.2654, + "step": 42751 + }, + { + "epoch": 0.9410521681708123, + "grad_norm": 2.646533236200404, + "learning_rate": 9.084779040281877e-08, + "loss": 0.2579, + "step": 42752 + }, + { + "epoch": 0.9410741800572309, + "grad_norm": 1.709163577826719, + "learning_rate": 9.078016159750458e-08, + "loss": 0.3018, + "step": 42753 + }, + { + "epoch": 0.9410961919436496, + "grad_norm": 1.2782305396614928, + "learning_rate": 9.071255774303567e-08, + "loss": 0.3495, + "step": 42754 + }, + { + "epoch": 0.9411182038300683, + "grad_norm": 7.004657548390953, + "learning_rate": 9.064497883975454e-08, + "loss": 0.3364, + "step": 42755 + }, + { + "epoch": 0.9411402157164869, + "grad_norm": 1.887393890727593, + "learning_rate": 9.057742488800591e-08, + "loss": 0.267, + "step": 42756 + }, + { + "epoch": 0.9411622276029056, + "grad_norm": 3.560338795802261, + "learning_rate": 9.050989588813175e-08, + "loss": 0.2319, + "step": 42757 + }, + { + "epoch": 0.9411842394893243, + "grad_norm": 1.7062744933778993, + "learning_rate": 9.044239184047621e-08, + "loss": 0.3975, + "step": 42758 + }, + { + "epoch": 0.9412062513757429, + "grad_norm": 1.1807795093620679, + "learning_rate": 9.03749127453829e-08, + "loss": 0.2106, + "step": 42759 + }, + { + "epoch": 0.9412282632621616, + "grad_norm": 2.484389741812817, + "learning_rate": 9.030745860319323e-08, + "loss": 0.3224, + "step": 42760 + }, + { + "epoch": 0.9412502751485803, + "grad_norm": 1.9734517189696308, + "learning_rate": 9.024002941425137e-08, + "loss": 0.2906, + "step": 42761 + }, + { + "epoch": 0.9412722870349989, + "grad_norm": 1.5775236198842653, + "learning_rate": 9.017262517889925e-08, + "loss": 0.2441, + "step": 42762 + }, + { + "epoch": 0.9412942989214176, + "grad_norm": 1.7890335175874392, + "learning_rate": 9.010524589747937e-08, + "loss": 0.329, + "step": 42763 + }, + { + "epoch": 0.9413163108078362, + "grad_norm": 1.8156329012730716, + "learning_rate": 9.003789157033483e-08, + "loss": 0.2722, + "step": 42764 + }, + { + "epoch": 0.9413383226942549, + "grad_norm": 2.0856423305594816, + "learning_rate": 8.997056219780753e-08, + "loss": 0.2548, + "step": 42765 + }, + { + "epoch": 0.9413603345806736, + "grad_norm": 2.1153418012656178, + "learning_rate": 8.990325778024e-08, + "loss": 0.3541, + "step": 42766 + }, + { + "epoch": 0.9413823464670922, + "grad_norm": 0.5987987319210794, + "learning_rate": 8.98359783179742e-08, + "loss": 0.4362, + "step": 42767 + }, + { + "epoch": 0.9414043583535109, + "grad_norm": 1.6899114495097856, + "learning_rate": 8.976872381135149e-08, + "loss": 0.3597, + "step": 42768 + }, + { + "epoch": 0.9414263702399296, + "grad_norm": 7.053192810859457, + "learning_rate": 8.970149426071384e-08, + "loss": 0.2331, + "step": 42769 + }, + { + "epoch": 0.9414483821263482, + "grad_norm": 1.6931964236641885, + "learning_rate": 8.963428966640375e-08, + "loss": 0.2606, + "step": 42770 + }, + { + "epoch": 0.9414703940127669, + "grad_norm": 2.1961359596364924, + "learning_rate": 8.956711002876262e-08, + "loss": 0.3288, + "step": 42771 + }, + { + "epoch": 0.9414924058991856, + "grad_norm": 1.6894012171537192, + "learning_rate": 8.949995534813128e-08, + "loss": 0.2547, + "step": 42772 + }, + { + "epoch": 0.9415144177856042, + "grad_norm": 1.3517784230024745, + "learning_rate": 8.943282562485111e-08, + "loss": 0.2107, + "step": 42773 + }, + { + "epoch": 0.9415364296720229, + "grad_norm": 2.5609713560842606, + "learning_rate": 8.936572085926354e-08, + "loss": 0.3188, + "step": 42774 + }, + { + "epoch": 0.9415584415584416, + "grad_norm": 1.6159330849094795, + "learning_rate": 8.929864105170938e-08, + "loss": 0.2828, + "step": 42775 + }, + { + "epoch": 0.9415804534448602, + "grad_norm": 0.625892609545581, + "learning_rate": 8.923158620253003e-08, + "loss": 0.4764, + "step": 42776 + }, + { + "epoch": 0.9416024653312789, + "grad_norm": 1.3971883044310551, + "learning_rate": 8.916455631206634e-08, + "loss": 0.3357, + "step": 42777 + }, + { + "epoch": 0.9416244772176976, + "grad_norm": 1.6845895615127422, + "learning_rate": 8.9097551380658e-08, + "loss": 0.277, + "step": 42778 + }, + { + "epoch": 0.9416464891041162, + "grad_norm": 2.38396359654634, + "learning_rate": 8.903057140864646e-08, + "loss": 0.3138, + "step": 42779 + }, + { + "epoch": 0.9416685009905349, + "grad_norm": 4.635744756811015, + "learning_rate": 8.896361639637196e-08, + "loss": 0.2681, + "step": 42780 + }, + { + "epoch": 0.9416905128769536, + "grad_norm": 1.7633711114717023, + "learning_rate": 8.889668634417537e-08, + "loss": 0.3016, + "step": 42781 + }, + { + "epoch": 0.9417125247633722, + "grad_norm": 1.9507837483499038, + "learning_rate": 8.882978125239583e-08, + "loss": 0.3242, + "step": 42782 + }, + { + "epoch": 0.9417345366497909, + "grad_norm": 1.843923225645044, + "learning_rate": 8.87629011213731e-08, + "loss": 0.2886, + "step": 42783 + }, + { + "epoch": 0.9417565485362096, + "grad_norm": 1.8314071581534839, + "learning_rate": 8.869604595144854e-08, + "loss": 0.2895, + "step": 42784 + }, + { + "epoch": 0.9417785604226282, + "grad_norm": 1.6830408591736734, + "learning_rate": 8.862921574296079e-08, + "loss": 0.2603, + "step": 42785 + }, + { + "epoch": 0.9418005723090469, + "grad_norm": 1.6675331875866177, + "learning_rate": 8.856241049625013e-08, + "loss": 0.2248, + "step": 42786 + }, + { + "epoch": 0.9418225841954656, + "grad_norm": 1.8565812824519505, + "learning_rate": 8.849563021165575e-08, + "loss": 0.2749, + "step": 42787 + }, + { + "epoch": 0.9418445960818842, + "grad_norm": 1.475572653748167, + "learning_rate": 8.842887488951735e-08, + "loss": 0.2626, + "step": 42788 + }, + { + "epoch": 0.9418666079683029, + "grad_norm": 1.5714888511045861, + "learning_rate": 8.836214453017355e-08, + "loss": 0.2888, + "step": 42789 + }, + { + "epoch": 0.9418886198547215, + "grad_norm": 2.143847681369775, + "learning_rate": 8.829543913396465e-08, + "loss": 0.2794, + "step": 42790 + }, + { + "epoch": 0.9419106317411402, + "grad_norm": 0.6238105122657778, + "learning_rate": 8.822875870122872e-08, + "loss": 0.4611, + "step": 42791 + }, + { + "epoch": 0.9419326436275589, + "grad_norm": 1.4587389362162737, + "learning_rate": 8.81621032323049e-08, + "loss": 0.2567, + "step": 42792 + }, + { + "epoch": 0.9419546555139775, + "grad_norm": 1.7525984366478853, + "learning_rate": 8.809547272753183e-08, + "loss": 0.2517, + "step": 42793 + }, + { + "epoch": 0.9419766674003962, + "grad_norm": 2.0868319890370386, + "learning_rate": 8.802886718724813e-08, + "loss": 0.2626, + "step": 42794 + }, + { + "epoch": 0.9419986792868149, + "grad_norm": 7.309649247560679, + "learning_rate": 8.796228661179351e-08, + "loss": 0.2242, + "step": 42795 + }, + { + "epoch": 0.9420206911732335, + "grad_norm": 1.5494637970112979, + "learning_rate": 8.78957310015044e-08, + "loss": 0.2587, + "step": 42796 + }, + { + "epoch": 0.9420427030596522, + "grad_norm": 1.846574963213995, + "learning_rate": 8.782920035672048e-08, + "loss": 0.3018, + "step": 42797 + }, + { + "epoch": 0.9420647149460709, + "grad_norm": 0.6288277977719852, + "learning_rate": 8.776269467777932e-08, + "loss": 0.4553, + "step": 42798 + }, + { + "epoch": 0.9420867268324895, + "grad_norm": 1.9643738502289179, + "learning_rate": 8.769621396501893e-08, + "loss": 0.2661, + "step": 42799 + }, + { + "epoch": 0.9421087387189082, + "grad_norm": 1.605025931948522, + "learning_rate": 8.762975821877795e-08, + "loss": 0.2441, + "step": 42800 + }, + { + "epoch": 0.9421307506053269, + "grad_norm": 1.6087131998029973, + "learning_rate": 8.756332743939278e-08, + "loss": 0.246, + "step": 42801 + }, + { + "epoch": 0.9421527624917455, + "grad_norm": 2.2401504235089353, + "learning_rate": 8.749692162720258e-08, + "loss": 0.2788, + "step": 42802 + }, + { + "epoch": 0.9421747743781642, + "grad_norm": 3.7197274070390636, + "learning_rate": 8.743054078254376e-08, + "loss": 0.3126, + "step": 42803 + }, + { + "epoch": 0.9421967862645829, + "grad_norm": 1.456095807128212, + "learning_rate": 8.736418490575383e-08, + "loss": 0.2758, + "step": 42804 + }, + { + "epoch": 0.9422187981510015, + "grad_norm": 1.62020504792857, + "learning_rate": 8.729785399717028e-08, + "loss": 0.2831, + "step": 42805 + }, + { + "epoch": 0.9422408100374202, + "grad_norm": 1.765840019314294, + "learning_rate": 8.723154805712952e-08, + "loss": 0.3143, + "step": 42806 + }, + { + "epoch": 0.9422628219238389, + "grad_norm": 1.8161195815569118, + "learning_rate": 8.716526708597017e-08, + "loss": 0.2503, + "step": 42807 + }, + { + "epoch": 0.9422848338102575, + "grad_norm": 1.3202975220105944, + "learning_rate": 8.709901108402807e-08, + "loss": 0.2627, + "step": 42808 + }, + { + "epoch": 0.9423068456966762, + "grad_norm": 2.8090531645171515, + "learning_rate": 8.703278005163962e-08, + "loss": 0.3874, + "step": 42809 + }, + { + "epoch": 0.9423288575830949, + "grad_norm": 2.0781990494325786, + "learning_rate": 8.696657398914177e-08, + "loss": 0.2864, + "step": 42810 + }, + { + "epoch": 0.9423508694695135, + "grad_norm": 1.7836506296637482, + "learning_rate": 8.69003928968709e-08, + "loss": 0.2917, + "step": 42811 + }, + { + "epoch": 0.9423728813559322, + "grad_norm": 4.578318151137025, + "learning_rate": 8.6834236775164e-08, + "loss": 0.3144, + "step": 42812 + }, + { + "epoch": 0.9423948932423508, + "grad_norm": 1.9558004428074804, + "learning_rate": 8.676810562435689e-08, + "loss": 0.2065, + "step": 42813 + }, + { + "epoch": 0.9424169051287695, + "grad_norm": 2.166447674159089, + "learning_rate": 8.670199944478486e-08, + "loss": 0.2829, + "step": 42814 + }, + { + "epoch": 0.9424389170151882, + "grad_norm": 1.6481815170302003, + "learning_rate": 8.663591823678542e-08, + "loss": 0.3793, + "step": 42815 + }, + { + "epoch": 0.9424609289016068, + "grad_norm": 1.6980899877908762, + "learning_rate": 8.656986200069329e-08, + "loss": 0.3374, + "step": 42816 + }, + { + "epoch": 0.9424829407880255, + "grad_norm": 2.141468551149537, + "learning_rate": 8.650383073684488e-08, + "loss": 0.2583, + "step": 42817 + }, + { + "epoch": 0.9425049526744442, + "grad_norm": 1.4951377414701867, + "learning_rate": 8.643782444557547e-08, + "loss": 0.2973, + "step": 42818 + }, + { + "epoch": 0.9425269645608628, + "grad_norm": 1.7922143331037352, + "learning_rate": 8.637184312722035e-08, + "loss": 0.3135, + "step": 42819 + }, + { + "epoch": 0.9425489764472815, + "grad_norm": 2.7163268171905406, + "learning_rate": 8.630588678211482e-08, + "loss": 0.2745, + "step": 42820 + }, + { + "epoch": 0.9425709883337002, + "grad_norm": 3.4374058177605336, + "learning_rate": 8.623995541059471e-08, + "loss": 0.2086, + "step": 42821 + }, + { + "epoch": 0.9425930002201188, + "grad_norm": 3.6466802317385767, + "learning_rate": 8.617404901299475e-08, + "loss": 0.3652, + "step": 42822 + }, + { + "epoch": 0.9426150121065375, + "grad_norm": 2.4001535042137387, + "learning_rate": 8.610816758964968e-08, + "loss": 0.299, + "step": 42823 + }, + { + "epoch": 0.9426370239929562, + "grad_norm": 0.6113848124288039, + "learning_rate": 8.604231114089479e-08, + "loss": 0.4576, + "step": 42824 + }, + { + "epoch": 0.9426590358793748, + "grad_norm": 0.5942706686382299, + "learning_rate": 8.597647966706423e-08, + "loss": 0.4244, + "step": 42825 + }, + { + "epoch": 0.9426810477657935, + "grad_norm": 1.7929520042528695, + "learning_rate": 8.591067316849333e-08, + "loss": 0.2767, + "step": 42826 + }, + { + "epoch": 0.9427030596522122, + "grad_norm": 3.3051529360235223, + "learning_rate": 8.584489164551568e-08, + "loss": 0.2657, + "step": 42827 + }, + { + "epoch": 0.9427250715386308, + "grad_norm": 1.5902387037725125, + "learning_rate": 8.577913509846658e-08, + "loss": 0.213, + "step": 42828 + }, + { + "epoch": 0.9427470834250495, + "grad_norm": 0.6097069930841302, + "learning_rate": 8.571340352767965e-08, + "loss": 0.4403, + "step": 42829 + }, + { + "epoch": 0.9427690953114682, + "grad_norm": 1.3893905730850935, + "learning_rate": 8.564769693348851e-08, + "loss": 0.2645, + "step": 42830 + }, + { + "epoch": 0.9427911071978868, + "grad_norm": 2.7041540800815365, + "learning_rate": 8.558201531622845e-08, + "loss": 0.3867, + "step": 42831 + }, + { + "epoch": 0.9428131190843055, + "grad_norm": 2.12191181252663, + "learning_rate": 8.551635867623142e-08, + "loss": 0.2721, + "step": 42832 + }, + { + "epoch": 0.9428351309707242, + "grad_norm": 2.2401689450655464, + "learning_rate": 8.545072701383328e-08, + "loss": 0.3004, + "step": 42833 + }, + { + "epoch": 0.9428571428571428, + "grad_norm": 1.6901069010064917, + "learning_rate": 8.538512032936596e-08, + "loss": 0.3371, + "step": 42834 + }, + { + "epoch": 0.9428791547435615, + "grad_norm": 0.6349674176612069, + "learning_rate": 8.531953862316311e-08, + "loss": 0.4523, + "step": 42835 + }, + { + "epoch": 0.9429011666299801, + "grad_norm": 1.5352316959375971, + "learning_rate": 8.52539818955589e-08, + "loss": 0.1826, + "step": 42836 + }, + { + "epoch": 0.9429231785163988, + "grad_norm": 0.5991301920135342, + "learning_rate": 8.518845014688581e-08, + "loss": 0.443, + "step": 42837 + }, + { + "epoch": 0.9429451904028175, + "grad_norm": 0.5701379980272758, + "learning_rate": 8.512294337747751e-08, + "loss": 0.4363, + "step": 42838 + }, + { + "epoch": 0.9429672022892361, + "grad_norm": 8.929438909805086, + "learning_rate": 8.505746158766592e-08, + "loss": 0.2618, + "step": 42839 + }, + { + "epoch": 0.9429892141756548, + "grad_norm": 1.9510760201090005, + "learning_rate": 8.499200477778468e-08, + "loss": 0.3028, + "step": 42840 + }, + { + "epoch": 0.9430112260620735, + "grad_norm": 1.3672244902071597, + "learning_rate": 8.492657294816576e-08, + "loss": 0.2167, + "step": 42841 + }, + { + "epoch": 0.9430332379484921, + "grad_norm": 1.7628516716847296, + "learning_rate": 8.486116609914275e-08, + "loss": 0.286, + "step": 42842 + }, + { + "epoch": 0.9430552498349108, + "grad_norm": 0.6310365939843958, + "learning_rate": 8.479578423104762e-08, + "loss": 0.4697, + "step": 42843 + }, + { + "epoch": 0.9430772617213296, + "grad_norm": 0.628405055929144, + "learning_rate": 8.473042734421233e-08, + "loss": 0.4721, + "step": 42844 + }, + { + "epoch": 0.9430992736077481, + "grad_norm": 0.6238916987887894, + "learning_rate": 8.466509543896884e-08, + "loss": 0.4509, + "step": 42845 + }, + { + "epoch": 0.9431212854941668, + "grad_norm": 1.5168455455067333, + "learning_rate": 8.459978851565021e-08, + "loss": 0.2615, + "step": 42846 + }, + { + "epoch": 0.9431432973805856, + "grad_norm": 1.5943121096034285, + "learning_rate": 8.453450657458728e-08, + "loss": 0.2814, + "step": 42847 + }, + { + "epoch": 0.9431653092670041, + "grad_norm": 1.9104759646428295, + "learning_rate": 8.446924961611313e-08, + "loss": 0.3028, + "step": 42848 + }, + { + "epoch": 0.9431873211534229, + "grad_norm": 0.6208728900686906, + "learning_rate": 8.440401764055805e-08, + "loss": 0.4556, + "step": 42849 + }, + { + "epoch": 0.9432093330398416, + "grad_norm": 1.558858430301784, + "learning_rate": 8.433881064825455e-08, + "loss": 0.3462, + "step": 42850 + }, + { + "epoch": 0.9432313449262602, + "grad_norm": 1.9790423533282986, + "learning_rate": 8.427362863953292e-08, + "loss": 0.2784, + "step": 42851 + }, + { + "epoch": 0.9432533568126789, + "grad_norm": 1.7713481380587695, + "learning_rate": 8.420847161472567e-08, + "loss": 0.3517, + "step": 42852 + }, + { + "epoch": 0.9432753686990976, + "grad_norm": 2.121403439598854, + "learning_rate": 8.414333957416365e-08, + "loss": 0.3267, + "step": 42853 + }, + { + "epoch": 0.9432973805855162, + "grad_norm": 1.6993152628020072, + "learning_rate": 8.40782325181777e-08, + "loss": 0.2512, + "step": 42854 + }, + { + "epoch": 0.9433193924719349, + "grad_norm": 1.3721086596817618, + "learning_rate": 8.401315044709868e-08, + "loss": 0.2851, + "step": 42855 + }, + { + "epoch": 0.9433414043583536, + "grad_norm": 1.8386839919344815, + "learning_rate": 8.394809336125686e-08, + "loss": 0.3041, + "step": 42856 + }, + { + "epoch": 0.9433634162447722, + "grad_norm": 0.6237821294945189, + "learning_rate": 8.388306126098367e-08, + "loss": 0.4382, + "step": 42857 + }, + { + "epoch": 0.9433854281311909, + "grad_norm": 0.6343294339173604, + "learning_rate": 8.381805414660993e-08, + "loss": 0.4568, + "step": 42858 + }, + { + "epoch": 0.9434074400176095, + "grad_norm": 0.6213375724817263, + "learning_rate": 8.375307201846539e-08, + "loss": 0.463, + "step": 42859 + }, + { + "epoch": 0.9434294519040282, + "grad_norm": 3.4649112474373376, + "learning_rate": 8.368811487688033e-08, + "loss": 0.2785, + "step": 42860 + }, + { + "epoch": 0.9434514637904469, + "grad_norm": 1.8059007296745424, + "learning_rate": 8.36231827221845e-08, + "loss": 0.299, + "step": 42861 + }, + { + "epoch": 0.9434734756768655, + "grad_norm": 1.8433053696789967, + "learning_rate": 8.35582755547093e-08, + "loss": 0.2407, + "step": 42862 + }, + { + "epoch": 0.9434954875632842, + "grad_norm": 0.6307186609675435, + "learning_rate": 8.349339337478335e-08, + "loss": 0.4547, + "step": 42863 + }, + { + "epoch": 0.9435174994497029, + "grad_norm": 1.9045750934889725, + "learning_rate": 8.342853618273694e-08, + "loss": 0.3242, + "step": 42864 + }, + { + "epoch": 0.9435395113361215, + "grad_norm": 1.3266830425779086, + "learning_rate": 8.336370397889926e-08, + "loss": 0.2756, + "step": 42865 + }, + { + "epoch": 0.9435615232225402, + "grad_norm": 2.7212640338255483, + "learning_rate": 8.329889676360003e-08, + "loss": 0.2511, + "step": 42866 + }, + { + "epoch": 0.9435835351089589, + "grad_norm": 1.4724011509600476, + "learning_rate": 8.3234114537169e-08, + "loss": 0.2549, + "step": 42867 + }, + { + "epoch": 0.9436055469953775, + "grad_norm": 2.08712674031004, + "learning_rate": 8.316935729993481e-08, + "loss": 0.3537, + "step": 42868 + }, + { + "epoch": 0.9436275588817962, + "grad_norm": 1.9459160173878447, + "learning_rate": 8.310462505222771e-08, + "loss": 0.3276, + "step": 42869 + }, + { + "epoch": 0.9436495707682149, + "grad_norm": 2.566050327812104, + "learning_rate": 8.303991779437471e-08, + "loss": 0.2965, + "step": 42870 + }, + { + "epoch": 0.9436715826546335, + "grad_norm": 7.015480226096987, + "learning_rate": 8.297523552670717e-08, + "loss": 0.3109, + "step": 42871 + }, + { + "epoch": 0.9436935945410522, + "grad_norm": 1.9951911185348008, + "learning_rate": 8.291057824955151e-08, + "loss": 0.2222, + "step": 42872 + }, + { + "epoch": 0.9437156064274709, + "grad_norm": 1.7677969233353166, + "learning_rate": 8.284594596323748e-08, + "loss": 0.3152, + "step": 42873 + }, + { + "epoch": 0.9437376183138895, + "grad_norm": 1.7440786640698143, + "learning_rate": 8.278133866809423e-08, + "loss": 0.2721, + "step": 42874 + }, + { + "epoch": 0.9437596302003082, + "grad_norm": 1.758835635975574, + "learning_rate": 8.271675636444876e-08, + "loss": 0.3138, + "step": 42875 + }, + { + "epoch": 0.9437816420867269, + "grad_norm": 1.625910139639332, + "learning_rate": 8.265219905263022e-08, + "loss": 0.2669, + "step": 42876 + }, + { + "epoch": 0.9438036539731455, + "grad_norm": 1.9114839561237378, + "learning_rate": 8.258766673296615e-08, + "loss": 0.2593, + "step": 42877 + }, + { + "epoch": 0.9438256658595642, + "grad_norm": 1.4636182935774749, + "learning_rate": 8.252315940578459e-08, + "loss": 0.1839, + "step": 42878 + }, + { + "epoch": 0.9438476777459829, + "grad_norm": 1.930738156689573, + "learning_rate": 8.24586770714142e-08, + "loss": 0.2781, + "step": 42879 + }, + { + "epoch": 0.9438696896324015, + "grad_norm": 1.900625224552628, + "learning_rate": 8.239421973018135e-08, + "loss": 0.3585, + "step": 42880 + }, + { + "epoch": 0.9438917015188202, + "grad_norm": 2.504803158629445, + "learning_rate": 8.232978738241471e-08, + "loss": 0.3249, + "step": 42881 + }, + { + "epoch": 0.9439137134052388, + "grad_norm": 1.803333730401955, + "learning_rate": 8.226538002844176e-08, + "loss": 0.2949, + "step": 42882 + }, + { + "epoch": 0.9439357252916575, + "grad_norm": 4.89037240061954, + "learning_rate": 8.220099766858891e-08, + "loss": 0.3118, + "step": 42883 + }, + { + "epoch": 0.9439577371780762, + "grad_norm": 2.5556606895066896, + "learning_rate": 8.21366403031848e-08, + "loss": 0.3327, + "step": 42884 + }, + { + "epoch": 0.9439797490644948, + "grad_norm": 2.0327160290059068, + "learning_rate": 8.207230793255527e-08, + "loss": 0.352, + "step": 42885 + }, + { + "epoch": 0.9440017609509135, + "grad_norm": 2.572263833871241, + "learning_rate": 8.200800055702729e-08, + "loss": 0.3065, + "step": 42886 + }, + { + "epoch": 0.9440237728373322, + "grad_norm": 0.6248945694303137, + "learning_rate": 8.194371817692837e-08, + "loss": 0.4599, + "step": 42887 + }, + { + "epoch": 0.9440457847237508, + "grad_norm": 2.070571882242569, + "learning_rate": 8.187946079258546e-08, + "loss": 0.2658, + "step": 42888 + }, + { + "epoch": 0.9440677966101695, + "grad_norm": 0.6177966811380935, + "learning_rate": 8.181522840432443e-08, + "loss": 0.4176, + "step": 42889 + }, + { + "epoch": 0.9440898084965882, + "grad_norm": 1.4217829110099474, + "learning_rate": 8.175102101247223e-08, + "loss": 0.2729, + "step": 42890 + }, + { + "epoch": 0.9441118203830068, + "grad_norm": 1.9589360104017004, + "learning_rate": 8.168683861735416e-08, + "loss": 0.3193, + "step": 42891 + }, + { + "epoch": 0.9441338322694255, + "grad_norm": 1.9198977568652087, + "learning_rate": 8.162268121929773e-08, + "loss": 0.3022, + "step": 42892 + }, + { + "epoch": 0.9441558441558442, + "grad_norm": 1.5124273649296869, + "learning_rate": 8.155854881862824e-08, + "loss": 0.2905, + "step": 42893 + }, + { + "epoch": 0.9441778560422628, + "grad_norm": 1.751086225738813, + "learning_rate": 8.149444141567209e-08, + "loss": 0.2296, + "step": 42894 + }, + { + "epoch": 0.9441998679286815, + "grad_norm": 1.537624470714665, + "learning_rate": 8.143035901075569e-08, + "loss": 0.2846, + "step": 42895 + }, + { + "epoch": 0.9442218798151002, + "grad_norm": 1.747067097935828, + "learning_rate": 8.136630160420267e-08, + "loss": 0.2494, + "step": 42896 + }, + { + "epoch": 0.9442438917015188, + "grad_norm": 1.590544598383588, + "learning_rate": 8.130226919634055e-08, + "loss": 0.3179, + "step": 42897 + }, + { + "epoch": 0.9442659035879375, + "grad_norm": 0.6029556302063316, + "learning_rate": 8.12382617874935e-08, + "loss": 0.4464, + "step": 42898 + }, + { + "epoch": 0.9442879154743562, + "grad_norm": 1.873949052962681, + "learning_rate": 8.117427937798794e-08, + "loss": 0.3697, + "step": 42899 + }, + { + "epoch": 0.9443099273607748, + "grad_norm": 1.8718257762001926, + "learning_rate": 8.11103219681486e-08, + "loss": 0.2135, + "step": 42900 + }, + { + "epoch": 0.9443319392471935, + "grad_norm": 1.8120898868705628, + "learning_rate": 8.104638955830024e-08, + "loss": 0.2969, + "step": 42901 + }, + { + "epoch": 0.9443539511336122, + "grad_norm": 1.6606684261502958, + "learning_rate": 8.098248214876813e-08, + "loss": 0.3095, + "step": 42902 + }, + { + "epoch": 0.9443759630200308, + "grad_norm": 2.189624725979371, + "learning_rate": 8.091859973987759e-08, + "loss": 0.237, + "step": 42903 + }, + { + "epoch": 0.9443979749064495, + "grad_norm": 3.002818187636357, + "learning_rate": 8.085474233195168e-08, + "loss": 0.2978, + "step": 42904 + }, + { + "epoch": 0.9444199867928681, + "grad_norm": 1.531160110706492, + "learning_rate": 8.079090992531679e-08, + "loss": 0.2615, + "step": 42905 + }, + { + "epoch": 0.9444419986792868, + "grad_norm": 3.034279056936618, + "learning_rate": 8.072710252029603e-08, + "loss": 0.2467, + "step": 42906 + }, + { + "epoch": 0.9444640105657055, + "grad_norm": 1.851602131433351, + "learning_rate": 8.066332011721412e-08, + "loss": 0.2316, + "step": 42907 + }, + { + "epoch": 0.9444860224521241, + "grad_norm": 1.560392626446329, + "learning_rate": 8.059956271639579e-08, + "loss": 0.2229, + "step": 42908 + }, + { + "epoch": 0.9445080343385428, + "grad_norm": 1.7605165295422076, + "learning_rate": 8.053583031816415e-08, + "loss": 0.2201, + "step": 42909 + }, + { + "epoch": 0.9445300462249615, + "grad_norm": 4.358645901603941, + "learning_rate": 8.04721229228439e-08, + "loss": 0.3673, + "step": 42910 + }, + { + "epoch": 0.9445520581113801, + "grad_norm": 8.85051200722807, + "learning_rate": 8.040844053075813e-08, + "loss": 0.3268, + "step": 42911 + }, + { + "epoch": 0.9445740699977988, + "grad_norm": 3.1361533130601242, + "learning_rate": 8.034478314223104e-08, + "loss": 0.2619, + "step": 42912 + }, + { + "epoch": 0.9445960818842175, + "grad_norm": 3.142227751318463, + "learning_rate": 8.028115075758569e-08, + "loss": 0.2692, + "step": 42913 + }, + { + "epoch": 0.9446180937706361, + "grad_norm": 1.9677877437787739, + "learning_rate": 8.021754337714627e-08, + "loss": 0.3155, + "step": 42914 + }, + { + "epoch": 0.9446401056570548, + "grad_norm": 0.6215215025713167, + "learning_rate": 8.015396100123529e-08, + "loss": 0.4477, + "step": 42915 + }, + { + "epoch": 0.9446621175434735, + "grad_norm": 2.4599112972923227, + "learning_rate": 8.00904036301764e-08, + "loss": 0.188, + "step": 42916 + }, + { + "epoch": 0.9446841294298921, + "grad_norm": 0.621163662681127, + "learning_rate": 8.002687126429265e-08, + "loss": 0.4465, + "step": 42917 + }, + { + "epoch": 0.9447061413163108, + "grad_norm": 1.5941718691619817, + "learning_rate": 7.996336390390603e-08, + "loss": 0.2661, + "step": 42918 + }, + { + "epoch": 0.9447281532027295, + "grad_norm": 1.7933393097566441, + "learning_rate": 7.989988154934015e-08, + "loss": 0.3334, + "step": 42919 + }, + { + "epoch": 0.9447501650891481, + "grad_norm": 1.83148065435342, + "learning_rate": 7.983642420091809e-08, + "loss": 0.2434, + "step": 42920 + }, + { + "epoch": 0.9447721769755668, + "grad_norm": 2.710516120382999, + "learning_rate": 7.977299185896181e-08, + "loss": 0.2984, + "step": 42921 + }, + { + "epoch": 0.9447941888619855, + "grad_norm": 3.430478260732677, + "learning_rate": 7.970958452379274e-08, + "loss": 0.1965, + "step": 42922 + }, + { + "epoch": 0.9448162007484041, + "grad_norm": 1.641986589468636, + "learning_rate": 7.964620219573449e-08, + "loss": 0.1985, + "step": 42923 + }, + { + "epoch": 0.9448382126348228, + "grad_norm": 1.3216571510227901, + "learning_rate": 7.958284487510847e-08, + "loss": 0.2507, + "step": 42924 + }, + { + "epoch": 0.9448602245212415, + "grad_norm": 1.4578816584212566, + "learning_rate": 7.951951256223778e-08, + "loss": 0.4049, + "step": 42925 + }, + { + "epoch": 0.9448822364076601, + "grad_norm": 0.6105575829952351, + "learning_rate": 7.945620525744325e-08, + "loss": 0.468, + "step": 42926 + }, + { + "epoch": 0.9449042482940788, + "grad_norm": 2.121331208959688, + "learning_rate": 7.939292296104683e-08, + "loss": 0.2808, + "step": 42927 + }, + { + "epoch": 0.9449262601804975, + "grad_norm": 1.4678956954584537, + "learning_rate": 7.932966567336998e-08, + "loss": 0.2189, + "step": 42928 + }, + { + "epoch": 0.9449482720669161, + "grad_norm": 1.5056654416619422, + "learning_rate": 7.926643339473461e-08, + "loss": 0.3239, + "step": 42929 + }, + { + "epoch": 0.9449702839533348, + "grad_norm": 1.9476060527415775, + "learning_rate": 7.920322612546216e-08, + "loss": 0.2692, + "step": 42930 + }, + { + "epoch": 0.9449922958397534, + "grad_norm": 1.9262411339988779, + "learning_rate": 7.914004386587348e-08, + "loss": 0.2546, + "step": 42931 + }, + { + "epoch": 0.9450143077261721, + "grad_norm": 1.5345410425901063, + "learning_rate": 7.907688661628999e-08, + "loss": 0.3965, + "step": 42932 + }, + { + "epoch": 0.9450363196125908, + "grad_norm": 1.511348784513872, + "learning_rate": 7.901375437703196e-08, + "loss": 0.2548, + "step": 42933 + }, + { + "epoch": 0.9450583314990094, + "grad_norm": 1.654232854665422, + "learning_rate": 7.895064714842138e-08, + "loss": 0.3063, + "step": 42934 + }, + { + "epoch": 0.9450803433854281, + "grad_norm": 0.6179608389460761, + "learning_rate": 7.888756493077854e-08, + "loss": 0.4459, + "step": 42935 + }, + { + "epoch": 0.9451023552718468, + "grad_norm": 2.177804190585244, + "learning_rate": 7.88245077244243e-08, + "loss": 0.3197, + "step": 42936 + }, + { + "epoch": 0.9451243671582654, + "grad_norm": 2.179800596364719, + "learning_rate": 7.876147552967783e-08, + "loss": 0.2964, + "step": 42937 + }, + { + "epoch": 0.9451463790446841, + "grad_norm": 0.6254459714458056, + "learning_rate": 7.869846834686113e-08, + "loss": 0.4275, + "step": 42938 + }, + { + "epoch": 0.9451683909311028, + "grad_norm": 2.2459092719844054, + "learning_rate": 7.863548617629336e-08, + "loss": 0.2639, + "step": 42939 + }, + { + "epoch": 0.9451904028175214, + "grad_norm": 1.6469245631718203, + "learning_rate": 7.857252901829538e-08, + "loss": 0.233, + "step": 42940 + }, + { + "epoch": 0.9452124147039401, + "grad_norm": 2.1533392245136054, + "learning_rate": 7.850959687318694e-08, + "loss": 0.4021, + "step": 42941 + }, + { + "epoch": 0.9452344265903588, + "grad_norm": 1.4819133988277564, + "learning_rate": 7.844668974128778e-08, + "loss": 0.3234, + "step": 42942 + }, + { + "epoch": 0.9452564384767774, + "grad_norm": 2.0461760682771, + "learning_rate": 7.838380762291763e-08, + "loss": 0.3317, + "step": 42943 + }, + { + "epoch": 0.9452784503631961, + "grad_norm": 2.1396489275714545, + "learning_rate": 7.832095051839573e-08, + "loss": 0.25, + "step": 42944 + }, + { + "epoch": 0.9453004622496148, + "grad_norm": 1.5109613448936612, + "learning_rate": 7.825811842804176e-08, + "loss": 0.3302, + "step": 42945 + }, + { + "epoch": 0.9453224741360334, + "grad_norm": 1.2841249632774894, + "learning_rate": 7.81953113521755e-08, + "loss": 0.2352, + "step": 42946 + }, + { + "epoch": 0.9453444860224521, + "grad_norm": 2.585309236527875, + "learning_rate": 7.813252929111559e-08, + "loss": 0.355, + "step": 42947 + }, + { + "epoch": 0.9453664979088708, + "grad_norm": 2.125635994314212, + "learning_rate": 7.806977224518175e-08, + "loss": 0.3354, + "step": 42948 + }, + { + "epoch": 0.9453885097952894, + "grad_norm": 1.7489520860356547, + "learning_rate": 7.800704021469263e-08, + "loss": 0.3025, + "step": 42949 + }, + { + "epoch": 0.9454105216817081, + "grad_norm": 1.4142438097441774, + "learning_rate": 7.794433319996631e-08, + "loss": 0.2649, + "step": 42950 + }, + { + "epoch": 0.9454325335681268, + "grad_norm": 0.6277514028501531, + "learning_rate": 7.788165120132307e-08, + "loss": 0.4602, + "step": 42951 + }, + { + "epoch": 0.9454545454545454, + "grad_norm": 1.577594539553305, + "learning_rate": 7.78189942190799e-08, + "loss": 0.3284, + "step": 42952 + }, + { + "epoch": 0.9454765573409641, + "grad_norm": 1.7108646995028538, + "learning_rate": 7.775636225355653e-08, + "loss": 0.2695, + "step": 42953 + }, + { + "epoch": 0.9454985692273827, + "grad_norm": 1.5493472322289563, + "learning_rate": 7.76937553050705e-08, + "loss": 0.2223, + "step": 42954 + }, + { + "epoch": 0.9455205811138014, + "grad_norm": 1.696035798100249, + "learning_rate": 7.763117337393988e-08, + "loss": 0.2272, + "step": 42955 + }, + { + "epoch": 0.9455425930002201, + "grad_norm": 0.6367383987780675, + "learning_rate": 7.75686164604833e-08, + "loss": 0.4464, + "step": 42956 + }, + { + "epoch": 0.9455646048866387, + "grad_norm": 1.4832353251122539, + "learning_rate": 7.750608456501885e-08, + "loss": 0.2853, + "step": 42957 + }, + { + "epoch": 0.9455866167730574, + "grad_norm": 1.5151483740152183, + "learning_rate": 7.744357768786348e-08, + "loss": 0.2732, + "step": 42958 + }, + { + "epoch": 0.9456086286594761, + "grad_norm": 2.4867199885793494, + "learning_rate": 7.738109582933529e-08, + "loss": 0.309, + "step": 42959 + }, + { + "epoch": 0.9456306405458947, + "grad_norm": 0.5911139292460766, + "learning_rate": 7.731863898975234e-08, + "loss": 0.4498, + "step": 42960 + }, + { + "epoch": 0.9456526524323134, + "grad_norm": 2.0578142521095777, + "learning_rate": 7.725620716943161e-08, + "loss": 0.358, + "step": 42961 + }, + { + "epoch": 0.9456746643187321, + "grad_norm": 1.3928542346382904, + "learning_rate": 7.719380036869062e-08, + "loss": 0.2737, + "step": 42962 + }, + { + "epoch": 0.9456966762051507, + "grad_norm": 6.258227715806189, + "learning_rate": 7.713141858784579e-08, + "loss": 0.3094, + "step": 42963 + }, + { + "epoch": 0.9457186880915694, + "grad_norm": 1.8896519516914727, + "learning_rate": 7.706906182721464e-08, + "loss": 0.2673, + "step": 42964 + }, + { + "epoch": 0.9457406999779882, + "grad_norm": 2.044449347975988, + "learning_rate": 7.700673008711467e-08, + "loss": 0.2892, + "step": 42965 + }, + { + "epoch": 0.9457627118644067, + "grad_norm": 1.7708219980745192, + "learning_rate": 7.694442336786234e-08, + "loss": 0.2628, + "step": 42966 + }, + { + "epoch": 0.9457847237508255, + "grad_norm": 1.6504675744033113, + "learning_rate": 7.688214166977404e-08, + "loss": 0.2944, + "step": 42967 + }, + { + "epoch": 0.9458067356372442, + "grad_norm": 1.4838576608783558, + "learning_rate": 7.681988499316617e-08, + "loss": 0.2464, + "step": 42968 + }, + { + "epoch": 0.9458287475236627, + "grad_norm": 0.5949519146703902, + "learning_rate": 7.675765333835572e-08, + "loss": 0.4299, + "step": 42969 + }, + { + "epoch": 0.9458507594100815, + "grad_norm": 1.5519105545085494, + "learning_rate": 7.669544670565853e-08, + "loss": 0.2747, + "step": 42970 + }, + { + "epoch": 0.9458727712965002, + "grad_norm": 0.612993900230688, + "learning_rate": 7.663326509539104e-08, + "loss": 0.4419, + "step": 42971 + }, + { + "epoch": 0.9458947831829188, + "grad_norm": 1.6240263162044015, + "learning_rate": 7.657110850786964e-08, + "loss": 0.326, + "step": 42972 + }, + { + "epoch": 0.9459167950693375, + "grad_norm": 0.6094283284364519, + "learning_rate": 7.650897694340909e-08, + "loss": 0.4358, + "step": 42973 + }, + { + "epoch": 0.9459388069557562, + "grad_norm": 1.7716787418902686, + "learning_rate": 7.644687040232634e-08, + "loss": 0.2842, + "step": 42974 + }, + { + "epoch": 0.9459608188421748, + "grad_norm": 1.2135785943372412, + "learning_rate": 7.638478888493617e-08, + "loss": 0.2077, + "step": 42975 + }, + { + "epoch": 0.9459828307285935, + "grad_norm": 1.5243132479792025, + "learning_rate": 7.632273239155441e-08, + "loss": 0.3295, + "step": 42976 + }, + { + "epoch": 0.946004842615012, + "grad_norm": 1.9420230014522688, + "learning_rate": 7.626070092249749e-08, + "loss": 0.2884, + "step": 42977 + }, + { + "epoch": 0.9460268545014308, + "grad_norm": 1.8484244698195027, + "learning_rate": 7.619869447807848e-08, + "loss": 0.2087, + "step": 42978 + }, + { + "epoch": 0.9460488663878495, + "grad_norm": 1.6072719132861293, + "learning_rate": 7.613671305861492e-08, + "loss": 0.2157, + "step": 42979 + }, + { + "epoch": 0.9460708782742681, + "grad_norm": 1.7991601631040852, + "learning_rate": 7.607475666441933e-08, + "loss": 0.2799, + "step": 42980 + }, + { + "epoch": 0.9460928901606868, + "grad_norm": 2.2041103168976393, + "learning_rate": 7.601282529580922e-08, + "loss": 0.2808, + "step": 42981 + }, + { + "epoch": 0.9461149020471055, + "grad_norm": 1.5370864121976522, + "learning_rate": 7.59509189530977e-08, + "loss": 0.375, + "step": 42982 + }, + { + "epoch": 0.9461369139335241, + "grad_norm": 1.4712559843154345, + "learning_rate": 7.58890376365995e-08, + "loss": 0.3083, + "step": 42983 + }, + { + "epoch": 0.9461589258199428, + "grad_norm": 1.7085765619945728, + "learning_rate": 7.582718134662992e-08, + "loss": 0.2981, + "step": 42984 + }, + { + "epoch": 0.9461809377063615, + "grad_norm": 3.478718637327468, + "learning_rate": 7.576535008350262e-08, + "loss": 0.3124, + "step": 42985 + }, + { + "epoch": 0.9462029495927801, + "grad_norm": 1.2483928842316447, + "learning_rate": 7.570354384753175e-08, + "loss": 0.2753, + "step": 42986 + }, + { + "epoch": 0.9462249614791988, + "grad_norm": 4.36184642834619, + "learning_rate": 7.564176263903212e-08, + "loss": 0.3567, + "step": 42987 + }, + { + "epoch": 0.9462469733656175, + "grad_norm": 1.3037238270601212, + "learning_rate": 7.558000645831731e-08, + "loss": 0.2104, + "step": 42988 + }, + { + "epoch": 0.9462689852520361, + "grad_norm": 1.6058194573111504, + "learning_rate": 7.551827530570156e-08, + "loss": 0.319, + "step": 42989 + }, + { + "epoch": 0.9462909971384548, + "grad_norm": 0.6022342866933366, + "learning_rate": 7.545656918149735e-08, + "loss": 0.4298, + "step": 42990 + }, + { + "epoch": 0.9463130090248735, + "grad_norm": 0.6132284725648414, + "learning_rate": 7.539488808602002e-08, + "loss": 0.4425, + "step": 42991 + }, + { + "epoch": 0.9463350209112921, + "grad_norm": 1.5656286347531048, + "learning_rate": 7.533323201958265e-08, + "loss": 0.2569, + "step": 42992 + }, + { + "epoch": 0.9463570327977108, + "grad_norm": 1.7403964792924747, + "learning_rate": 7.52716009824972e-08, + "loss": 0.3418, + "step": 42993 + }, + { + "epoch": 0.9463790446841295, + "grad_norm": 2.8558320502267063, + "learning_rate": 7.520999497507897e-08, + "loss": 0.3094, + "step": 42994 + }, + { + "epoch": 0.9464010565705481, + "grad_norm": 2.2860903965195565, + "learning_rate": 7.51484139976394e-08, + "loss": 0.3528, + "step": 42995 + }, + { + "epoch": 0.9464230684569668, + "grad_norm": 1.341412151786225, + "learning_rate": 7.508685805049266e-08, + "loss": 0.2308, + "step": 42996 + }, + { + "epoch": 0.9464450803433855, + "grad_norm": 2.1392135772775664, + "learning_rate": 7.502532713395072e-08, + "loss": 0.299, + "step": 42997 + }, + { + "epoch": 0.9464670922298041, + "grad_norm": 3.6087521106558795, + "learning_rate": 7.496382124832668e-08, + "loss": 0.3371, + "step": 42998 + }, + { + "epoch": 0.9464891041162228, + "grad_norm": 1.57277981022421, + "learning_rate": 7.490234039393363e-08, + "loss": 0.2978, + "step": 42999 + }, + { + "epoch": 0.9465111160026414, + "grad_norm": 2.281136404184355, + "learning_rate": 7.484088457108296e-08, + "loss": 0.2864, + "step": 43000 + }, + { + "epoch": 0.9465331278890601, + "grad_norm": 1.4037634418215132, + "learning_rate": 7.477945378008722e-08, + "loss": 0.2376, + "step": 43001 + }, + { + "epoch": 0.9465551397754788, + "grad_norm": 1.368314432313723, + "learning_rate": 7.471804802126003e-08, + "loss": 0.2782, + "step": 43002 + }, + { + "epoch": 0.9465771516618974, + "grad_norm": 0.6387266165911636, + "learning_rate": 7.465666729491228e-08, + "loss": 0.4751, + "step": 43003 + }, + { + "epoch": 0.9465991635483161, + "grad_norm": 1.8837481335430792, + "learning_rate": 7.459531160135536e-08, + "loss": 0.2988, + "step": 43004 + }, + { + "epoch": 0.9466211754347348, + "grad_norm": 1.9830008048972076, + "learning_rate": 7.453398094090236e-08, + "loss": 0.3553, + "step": 43005 + }, + { + "epoch": 0.9466431873211534, + "grad_norm": 2.1979844325130204, + "learning_rate": 7.447267531386415e-08, + "loss": 0.3605, + "step": 43006 + }, + { + "epoch": 0.9466651992075721, + "grad_norm": 3.0712821830077246, + "learning_rate": 7.441139472055325e-08, + "loss": 0.3595, + "step": 43007 + }, + { + "epoch": 0.9466872110939908, + "grad_norm": 1.545652451942443, + "learning_rate": 7.435013916128053e-08, + "loss": 0.2383, + "step": 43008 + }, + { + "epoch": 0.9467092229804094, + "grad_norm": 1.3281461450382697, + "learning_rate": 7.428890863635684e-08, + "loss": 0.2419, + "step": 43009 + }, + { + "epoch": 0.9467312348668281, + "grad_norm": 3.484595123280476, + "learning_rate": 7.422770314609362e-08, + "loss": 0.2029, + "step": 43010 + }, + { + "epoch": 0.9467532467532468, + "grad_norm": 2.398369794030652, + "learning_rate": 7.416652269080226e-08, + "loss": 0.314, + "step": 43011 + }, + { + "epoch": 0.9467752586396654, + "grad_norm": 2.0764192869492857, + "learning_rate": 7.41053672707942e-08, + "loss": 0.3392, + "step": 43012 + }, + { + "epoch": 0.9467972705260841, + "grad_norm": 1.4220125913913624, + "learning_rate": 7.404423688637974e-08, + "loss": 0.25, + "step": 43013 + }, + { + "epoch": 0.9468192824125028, + "grad_norm": 0.6195194405079537, + "learning_rate": 7.398313153786863e-08, + "loss": 0.4643, + "step": 43014 + }, + { + "epoch": 0.9468412942989214, + "grad_norm": 2.394960312401084, + "learning_rate": 7.392205122557339e-08, + "loss": 0.2957, + "step": 43015 + }, + { + "epoch": 0.9468633061853401, + "grad_norm": 1.4710657984708404, + "learning_rate": 7.386099594980212e-08, + "loss": 0.25, + "step": 43016 + }, + { + "epoch": 0.9468853180717588, + "grad_norm": 1.5696761977572882, + "learning_rate": 7.379996571086734e-08, + "loss": 0.2496, + "step": 43017 + }, + { + "epoch": 0.9469073299581774, + "grad_norm": 0.6041822401601676, + "learning_rate": 7.373896050907825e-08, + "loss": 0.4411, + "step": 43018 + }, + { + "epoch": 0.9469293418445961, + "grad_norm": 8.220926142610677, + "learning_rate": 7.36779803447446e-08, + "loss": 0.2282, + "step": 43019 + }, + { + "epoch": 0.9469513537310148, + "grad_norm": 1.7698051026843495, + "learning_rate": 7.361702521817671e-08, + "loss": 0.2745, + "step": 43020 + }, + { + "epoch": 0.9469733656174334, + "grad_norm": 1.3946577373844518, + "learning_rate": 7.355609512968432e-08, + "loss": 0.2433, + "step": 43021 + }, + { + "epoch": 0.9469953775038521, + "grad_norm": 1.6823116047310362, + "learning_rate": 7.349519007957717e-08, + "loss": 0.2848, + "step": 43022 + }, + { + "epoch": 0.9470173893902707, + "grad_norm": 1.7209534916254607, + "learning_rate": 7.343431006816504e-08, + "loss": 0.2612, + "step": 43023 + }, + { + "epoch": 0.9470394012766894, + "grad_norm": 1.4955558532333602, + "learning_rate": 7.337345509575655e-08, + "loss": 0.2605, + "step": 43024 + }, + { + "epoch": 0.9470614131631081, + "grad_norm": 1.8894545934523395, + "learning_rate": 7.331262516266146e-08, + "loss": 0.2874, + "step": 43025 + }, + { + "epoch": 0.9470834250495267, + "grad_norm": 1.8397308475991476, + "learning_rate": 7.325182026918897e-08, + "loss": 0.4517, + "step": 43026 + }, + { + "epoch": 0.9471054369359454, + "grad_norm": 1.9860389678808685, + "learning_rate": 7.319104041564829e-08, + "loss": 0.2513, + "step": 43027 + }, + { + "epoch": 0.9471274488223641, + "grad_norm": 2.2215338145460595, + "learning_rate": 7.313028560234859e-08, + "loss": 0.3026, + "step": 43028 + }, + { + "epoch": 0.9471494607087827, + "grad_norm": 2.4366732490287752, + "learning_rate": 7.306955582959796e-08, + "loss": 0.3879, + "step": 43029 + }, + { + "epoch": 0.9471714725952014, + "grad_norm": 1.622683611807534, + "learning_rate": 7.300885109770506e-08, + "loss": 0.4132, + "step": 43030 + }, + { + "epoch": 0.9471934844816201, + "grad_norm": 1.7697725300936213, + "learning_rate": 7.294817140697908e-08, + "loss": 0.3028, + "step": 43031 + }, + { + "epoch": 0.9472154963680387, + "grad_norm": 1.873627652626888, + "learning_rate": 7.288751675772754e-08, + "loss": 0.3095, + "step": 43032 + }, + { + "epoch": 0.9472375082544574, + "grad_norm": 0.6216313559373171, + "learning_rate": 7.282688715025966e-08, + "loss": 0.461, + "step": 43033 + }, + { + "epoch": 0.9472595201408761, + "grad_norm": 1.914112456181279, + "learning_rate": 7.276628258488239e-08, + "loss": 0.2547, + "step": 43034 + }, + { + "epoch": 0.9472815320272947, + "grad_norm": 1.5372439978407635, + "learning_rate": 7.270570306190549e-08, + "loss": 0.2625, + "step": 43035 + }, + { + "epoch": 0.9473035439137134, + "grad_norm": 0.6207951115890353, + "learning_rate": 7.264514858163485e-08, + "loss": 0.4395, + "step": 43036 + }, + { + "epoch": 0.9473255558001321, + "grad_norm": 3.278852211952795, + "learning_rate": 7.258461914437964e-08, + "loss": 0.2516, + "step": 43037 + }, + { + "epoch": 0.9473475676865507, + "grad_norm": 1.237424777766012, + "learning_rate": 7.25241147504474e-08, + "loss": 0.2468, + "step": 43038 + }, + { + "epoch": 0.9473695795729694, + "grad_norm": 1.8112371425885314, + "learning_rate": 7.246363540014512e-08, + "loss": 0.3485, + "step": 43039 + }, + { + "epoch": 0.9473915914593881, + "grad_norm": 1.4076634384086555, + "learning_rate": 7.240318109378031e-08, + "loss": 0.3015, + "step": 43040 + }, + { + "epoch": 0.9474136033458067, + "grad_norm": 3.4214552228903328, + "learning_rate": 7.23427518316605e-08, + "loss": 0.2197, + "step": 43041 + }, + { + "epoch": 0.9474356152322254, + "grad_norm": 0.6013653561222845, + "learning_rate": 7.228234761409214e-08, + "loss": 0.4216, + "step": 43042 + }, + { + "epoch": 0.9474576271186441, + "grad_norm": 1.892247725366884, + "learning_rate": 7.22219684413833e-08, + "loss": 0.2288, + "step": 43043 + }, + { + "epoch": 0.9474796390050627, + "grad_norm": 1.7215982309062432, + "learning_rate": 7.21616143138404e-08, + "loss": 0.3293, + "step": 43044 + }, + { + "epoch": 0.9475016508914814, + "grad_norm": 1.8480652543849068, + "learning_rate": 7.210128523176929e-08, + "loss": 0.2862, + "step": 43045 + }, + { + "epoch": 0.9475236627779, + "grad_norm": 2.177824887897823, + "learning_rate": 7.204098119547754e-08, + "loss": 0.2942, + "step": 43046 + }, + { + "epoch": 0.9475456746643187, + "grad_norm": 2.5584235993316007, + "learning_rate": 7.198070220527153e-08, + "loss": 0.2108, + "step": 43047 + }, + { + "epoch": 0.9475676865507374, + "grad_norm": 2.414692292902645, + "learning_rate": 7.192044826145772e-08, + "loss": 0.2636, + "step": 43048 + }, + { + "epoch": 0.947589698437156, + "grad_norm": 2.765273160945189, + "learning_rate": 7.186021936434251e-08, + "loss": 0.3449, + "step": 43049 + }, + { + "epoch": 0.9476117103235747, + "grad_norm": 1.9596528507864492, + "learning_rate": 7.180001551423121e-08, + "loss": 0.3478, + "step": 43050 + }, + { + "epoch": 0.9476337222099934, + "grad_norm": 1.6061683755589342, + "learning_rate": 7.173983671143026e-08, + "loss": 0.3209, + "step": 43051 + }, + { + "epoch": 0.947655734096412, + "grad_norm": 5.984311999903456, + "learning_rate": 7.167968295624494e-08, + "loss": 0.2855, + "step": 43052 + }, + { + "epoch": 0.9476777459828307, + "grad_norm": 1.6570627183370688, + "learning_rate": 7.161955424898282e-08, + "loss": 0.3133, + "step": 43053 + }, + { + "epoch": 0.9476997578692494, + "grad_norm": 1.5942950890845458, + "learning_rate": 7.15594505899475e-08, + "loss": 0.2703, + "step": 43054 + }, + { + "epoch": 0.947721769755668, + "grad_norm": 0.6146429264968882, + "learning_rate": 7.149937197944546e-08, + "loss": 0.4339, + "step": 43055 + }, + { + "epoch": 0.9477437816420867, + "grad_norm": 1.9418134083289922, + "learning_rate": 7.143931841778195e-08, + "loss": 0.2548, + "step": 43056 + }, + { + "epoch": 0.9477657935285054, + "grad_norm": 1.2634780080211154, + "learning_rate": 7.137928990526122e-08, + "loss": 0.1862, + "step": 43057 + }, + { + "epoch": 0.947787805414924, + "grad_norm": 1.9584611801294398, + "learning_rate": 7.131928644218966e-08, + "loss": 0.3039, + "step": 43058 + }, + { + "epoch": 0.9478098173013427, + "grad_norm": 1.5317031649883466, + "learning_rate": 7.125930802887204e-08, + "loss": 0.3182, + "step": 43059 + }, + { + "epoch": 0.9478318291877614, + "grad_norm": 1.628295324029968, + "learning_rate": 7.119935466561256e-08, + "loss": 0.3272, + "step": 43060 + }, + { + "epoch": 0.94785384107418, + "grad_norm": 1.6289227046193846, + "learning_rate": 7.113942635271653e-08, + "loss": 0.2223, + "step": 43061 + }, + { + "epoch": 0.9478758529605987, + "grad_norm": 0.6237849127392824, + "learning_rate": 7.10795230904876e-08, + "loss": 0.4464, + "step": 43062 + }, + { + "epoch": 0.9478978648470174, + "grad_norm": 1.9611668890625207, + "learning_rate": 7.101964487923164e-08, + "loss": 0.2964, + "step": 43063 + }, + { + "epoch": 0.947919876733436, + "grad_norm": 1.5563149111789265, + "learning_rate": 7.095979171925172e-08, + "loss": 0.2589, + "step": 43064 + }, + { + "epoch": 0.9479418886198547, + "grad_norm": 0.6315405570848438, + "learning_rate": 7.089996361085261e-08, + "loss": 0.4431, + "step": 43065 + }, + { + "epoch": 0.9479639005062734, + "grad_norm": 2.285446405705959, + "learning_rate": 7.084016055433906e-08, + "loss": 0.2398, + "step": 43066 + }, + { + "epoch": 0.947985912392692, + "grad_norm": 0.6326429405209191, + "learning_rate": 7.078038255001363e-08, + "loss": 0.4526, + "step": 43067 + }, + { + "epoch": 0.9480079242791107, + "grad_norm": 3.5312796850405017, + "learning_rate": 7.07206295981805e-08, + "loss": 0.2945, + "step": 43068 + }, + { + "epoch": 0.9480299361655294, + "grad_norm": 0.5983395118425523, + "learning_rate": 7.066090169914441e-08, + "loss": 0.4457, + "step": 43069 + }, + { + "epoch": 0.948051948051948, + "grad_norm": 2.2971880039881003, + "learning_rate": 7.060119885320737e-08, + "loss": 0.2981, + "step": 43070 + }, + { + "epoch": 0.9480739599383667, + "grad_norm": 1.3485718346564974, + "learning_rate": 7.054152106067413e-08, + "loss": 0.248, + "step": 43071 + }, + { + "epoch": 0.9480959718247853, + "grad_norm": 1.1814378263369607, + "learning_rate": 7.048186832184722e-08, + "loss": 0.2038, + "step": 43072 + }, + { + "epoch": 0.948117983711204, + "grad_norm": 3.998552381597727, + "learning_rate": 7.04222406370303e-08, + "loss": 0.3128, + "step": 43073 + }, + { + "epoch": 0.9481399955976227, + "grad_norm": 1.8526314101264854, + "learning_rate": 7.036263800652643e-08, + "loss": 0.4158, + "step": 43074 + }, + { + "epoch": 0.9481620074840413, + "grad_norm": 1.7518172119872206, + "learning_rate": 7.030306043063761e-08, + "loss": 0.2774, + "step": 43075 + }, + { + "epoch": 0.94818401937046, + "grad_norm": 1.4591030542789756, + "learning_rate": 7.024350790966806e-08, + "loss": 0.304, + "step": 43076 + }, + { + "epoch": 0.9482060312568787, + "grad_norm": 2.462752416266086, + "learning_rate": 7.018398044391916e-08, + "loss": 0.293, + "step": 43077 + }, + { + "epoch": 0.9482280431432973, + "grad_norm": 1.2594754828074057, + "learning_rate": 7.01244780336946e-08, + "loss": 0.2238, + "step": 43078 + }, + { + "epoch": 0.948250055029716, + "grad_norm": 1.3960619613393592, + "learning_rate": 7.006500067929578e-08, + "loss": 0.2704, + "step": 43079 + }, + { + "epoch": 0.9482720669161347, + "grad_norm": 5.550548887920955, + "learning_rate": 7.000554838102581e-08, + "loss": 0.2208, + "step": 43080 + }, + { + "epoch": 0.9482940788025533, + "grad_norm": 2.0283379923904854, + "learning_rate": 6.994612113918608e-08, + "loss": 0.247, + "step": 43081 + }, + { + "epoch": 0.948316090688972, + "grad_norm": 3.393662517076189, + "learning_rate": 6.988671895407972e-08, + "loss": 0.264, + "step": 43082 + }, + { + "epoch": 0.9483381025753908, + "grad_norm": 1.7423617821352122, + "learning_rate": 6.982734182600648e-08, + "loss": 0.4142, + "step": 43083 + }, + { + "epoch": 0.9483601144618093, + "grad_norm": 1.8534941143003154, + "learning_rate": 6.976798975527111e-08, + "loss": 0.3238, + "step": 43084 + }, + { + "epoch": 0.948382126348228, + "grad_norm": 1.9406984133015692, + "learning_rate": 6.970866274217336e-08, + "loss": 0.3286, + "step": 43085 + }, + { + "epoch": 0.9484041382346468, + "grad_norm": 3.1042362146458697, + "learning_rate": 6.964936078701468e-08, + "loss": 0.3601, + "step": 43086 + }, + { + "epoch": 0.9484261501210653, + "grad_norm": 1.620818885372606, + "learning_rate": 6.959008389009758e-08, + "loss": 0.2618, + "step": 43087 + }, + { + "epoch": 0.948448162007484, + "grad_norm": 2.779175967277512, + "learning_rate": 6.953083205172185e-08, + "loss": 0.3349, + "step": 43088 + }, + { + "epoch": 0.9484701738939028, + "grad_norm": 1.7333419719102043, + "learning_rate": 6.947160527219054e-08, + "loss": 0.3152, + "step": 43089 + }, + { + "epoch": 0.9484921857803214, + "grad_norm": 0.6181780847852668, + "learning_rate": 6.94124035518029e-08, + "loss": 0.4429, + "step": 43090 + }, + { + "epoch": 0.94851419766674, + "grad_norm": 1.6371972768613081, + "learning_rate": 6.935322689086088e-08, + "loss": 0.3833, + "step": 43091 + }, + { + "epoch": 0.9485362095531588, + "grad_norm": 8.995022181666084, + "learning_rate": 6.929407528966481e-08, + "loss": 0.3129, + "step": 43092 + }, + { + "epoch": 0.9485582214395774, + "grad_norm": 1.2846158288629046, + "learning_rate": 6.9234948748515e-08, + "loss": 0.3393, + "step": 43093 + }, + { + "epoch": 0.9485802333259961, + "grad_norm": 0.6261275423013214, + "learning_rate": 6.917584726771287e-08, + "loss": 0.4593, + "step": 43094 + }, + { + "epoch": 0.9486022452124147, + "grad_norm": 3.7711195855559487, + "learning_rate": 6.91167708475582e-08, + "loss": 0.2651, + "step": 43095 + }, + { + "epoch": 0.9486242570988334, + "grad_norm": 1.990416296069054, + "learning_rate": 6.905771948835127e-08, + "loss": 0.3065, + "step": 43096 + }, + { + "epoch": 0.9486462689852521, + "grad_norm": 2.0009637105764844, + "learning_rate": 6.899869319039243e-08, + "loss": 0.3654, + "step": 43097 + }, + { + "epoch": 0.9486682808716707, + "grad_norm": 2.308149739451726, + "learning_rate": 6.893969195398087e-08, + "loss": 0.3264, + "step": 43098 + }, + { + "epoch": 0.9486902927580894, + "grad_norm": 1.5305517418568364, + "learning_rate": 6.888071577941746e-08, + "loss": 0.2642, + "step": 43099 + }, + { + "epoch": 0.9487123046445081, + "grad_norm": 1.7931873497391717, + "learning_rate": 6.882176466700197e-08, + "loss": 0.2501, + "step": 43100 + }, + { + "epoch": 0.9487343165309267, + "grad_norm": 2.2430419933847294, + "learning_rate": 6.876283861703359e-08, + "loss": 0.4163, + "step": 43101 + }, + { + "epoch": 0.9487563284173454, + "grad_norm": 0.6066572454234577, + "learning_rate": 6.870393762981153e-08, + "loss": 0.4505, + "step": 43102 + }, + { + "epoch": 0.9487783403037641, + "grad_norm": 2.3312667822354856, + "learning_rate": 6.864506170563557e-08, + "loss": 0.3047, + "step": 43103 + }, + { + "epoch": 0.9488003521901827, + "grad_norm": 1.3764787682104964, + "learning_rate": 6.858621084480488e-08, + "loss": 0.2758, + "step": 43104 + }, + { + "epoch": 0.9488223640766014, + "grad_norm": 2.2271341586542652, + "learning_rate": 6.852738504761869e-08, + "loss": 0.3434, + "step": 43105 + }, + { + "epoch": 0.9488443759630201, + "grad_norm": 1.8100477404171047, + "learning_rate": 6.846858431437565e-08, + "loss": 0.3645, + "step": 43106 + }, + { + "epoch": 0.9488663878494387, + "grad_norm": 1.7990011510995645, + "learning_rate": 6.840980864537494e-08, + "loss": 0.2771, + "step": 43107 + }, + { + "epoch": 0.9488883997358574, + "grad_norm": 0.5903496265479329, + "learning_rate": 6.835105804091468e-08, + "loss": 0.4412, + "step": 43108 + }, + { + "epoch": 0.9489104116222761, + "grad_norm": 1.5321383244231246, + "learning_rate": 6.829233250129408e-08, + "loss": 0.2395, + "step": 43109 + }, + { + "epoch": 0.9489324235086947, + "grad_norm": 1.8752564585740537, + "learning_rate": 6.823363202681177e-08, + "loss": 0.2723, + "step": 43110 + }, + { + "epoch": 0.9489544353951134, + "grad_norm": 1.4140610473472075, + "learning_rate": 6.817495661776585e-08, + "loss": 0.2757, + "step": 43111 + }, + { + "epoch": 0.9489764472815321, + "grad_norm": 1.5796304486049544, + "learning_rate": 6.811630627445442e-08, + "loss": 0.2524, + "step": 43112 + }, + { + "epoch": 0.9489984591679507, + "grad_norm": 1.9068499913966412, + "learning_rate": 6.805768099717502e-08, + "loss": 0.2649, + "step": 43113 + }, + { + "epoch": 0.9490204710543694, + "grad_norm": 0.6193115063359098, + "learning_rate": 6.79990807862263e-08, + "loss": 0.4443, + "step": 43114 + }, + { + "epoch": 0.9490424829407881, + "grad_norm": 4.212281860990343, + "learning_rate": 6.79405056419069e-08, + "loss": 0.2702, + "step": 43115 + }, + { + "epoch": 0.9490644948272067, + "grad_norm": 1.4760699558749755, + "learning_rate": 6.788195556451271e-08, + "loss": 0.2782, + "step": 43116 + }, + { + "epoch": 0.9490865067136254, + "grad_norm": 3.6995178539664804, + "learning_rate": 6.782343055434238e-08, + "loss": 0.3197, + "step": 43117 + }, + { + "epoch": 0.949108518600044, + "grad_norm": 1.7339949706989972, + "learning_rate": 6.776493061169343e-08, + "loss": 0.3689, + "step": 43118 + }, + { + "epoch": 0.9491305304864627, + "grad_norm": 0.6172685596192424, + "learning_rate": 6.77064557368623e-08, + "loss": 0.4755, + "step": 43119 + }, + { + "epoch": 0.9491525423728814, + "grad_norm": 1.3700753052329606, + "learning_rate": 6.764800593014765e-08, + "loss": 0.215, + "step": 43120 + }, + { + "epoch": 0.9491745542593, + "grad_norm": 3.02286978292653, + "learning_rate": 6.758958119184588e-08, + "loss": 0.3914, + "step": 43121 + }, + { + "epoch": 0.9491965661457187, + "grad_norm": 2.2108864893045292, + "learning_rate": 6.753118152225291e-08, + "loss": 0.3124, + "step": 43122 + }, + { + "epoch": 0.9492185780321374, + "grad_norm": 1.8801560173979903, + "learning_rate": 6.747280692166736e-08, + "loss": 0.202, + "step": 43123 + }, + { + "epoch": 0.949240589918556, + "grad_norm": 1.559283575081612, + "learning_rate": 6.741445739038343e-08, + "loss": 0.2213, + "step": 43124 + }, + { + "epoch": 0.9492626018049747, + "grad_norm": 2.268236560245481, + "learning_rate": 6.735613292870036e-08, + "loss": 0.2189, + "step": 43125 + }, + { + "epoch": 0.9492846136913934, + "grad_norm": 2.2580924469420336, + "learning_rate": 6.729783353691399e-08, + "loss": 0.3178, + "step": 43126 + }, + { + "epoch": 0.949306625577812, + "grad_norm": 1.7408234717587163, + "learning_rate": 6.723955921531911e-08, + "loss": 0.3109, + "step": 43127 + }, + { + "epoch": 0.9493286374642307, + "grad_norm": 1.523345276135961, + "learning_rate": 6.718130996421324e-08, + "loss": 0.3584, + "step": 43128 + }, + { + "epoch": 0.9493506493506494, + "grad_norm": 2.5575819002475604, + "learning_rate": 6.712308578389171e-08, + "loss": 0.2886, + "step": 43129 + }, + { + "epoch": 0.949372661237068, + "grad_norm": 0.6100955876287808, + "learning_rate": 6.706488667465094e-08, + "loss": 0.4498, + "step": 43130 + }, + { + "epoch": 0.9493946731234867, + "grad_norm": 1.5867037209794024, + "learning_rate": 6.700671263678682e-08, + "loss": 0.3563, + "step": 43131 + }, + { + "epoch": 0.9494166850099054, + "grad_norm": 5.240697555786852, + "learning_rate": 6.694856367059466e-08, + "loss": 0.3027, + "step": 43132 + }, + { + "epoch": 0.949438696896324, + "grad_norm": 1.6413140708665221, + "learning_rate": 6.689043977636977e-08, + "loss": 0.2837, + "step": 43133 + }, + { + "epoch": 0.9494607087827427, + "grad_norm": 1.782959595650452, + "learning_rate": 6.683234095440805e-08, + "loss": 0.3525, + "step": 43134 + }, + { + "epoch": 0.9494827206691614, + "grad_norm": 2.034427317683453, + "learning_rate": 6.677426720500479e-08, + "loss": 0.2537, + "step": 43135 + }, + { + "epoch": 0.94950473255558, + "grad_norm": 2.056935027664817, + "learning_rate": 6.671621852845533e-08, + "loss": 0.2996, + "step": 43136 + }, + { + "epoch": 0.9495267444419987, + "grad_norm": 2.272562351271808, + "learning_rate": 6.665819492505333e-08, + "loss": 0.2971, + "step": 43137 + }, + { + "epoch": 0.9495487563284174, + "grad_norm": 2.3337508450618043, + "learning_rate": 6.660019639509574e-08, + "loss": 0.2968, + "step": 43138 + }, + { + "epoch": 0.949570768214836, + "grad_norm": 2.2028465781535735, + "learning_rate": 6.654222293887568e-08, + "loss": 0.2854, + "step": 43139 + }, + { + "epoch": 0.9495927801012547, + "grad_norm": 1.5582646510539098, + "learning_rate": 6.648427455668849e-08, + "loss": 0.2456, + "step": 43140 + }, + { + "epoch": 0.9496147919876733, + "grad_norm": 0.636446608655873, + "learning_rate": 6.64263512488289e-08, + "loss": 0.4564, + "step": 43141 + }, + { + "epoch": 0.949636803874092, + "grad_norm": 1.3718576151359616, + "learning_rate": 6.636845301559114e-08, + "loss": 0.2904, + "step": 43142 + }, + { + "epoch": 0.9496588157605107, + "grad_norm": 1.6699042299294227, + "learning_rate": 6.63105798572694e-08, + "loss": 0.3061, + "step": 43143 + }, + { + "epoch": 0.9496808276469293, + "grad_norm": 1.5728650520301608, + "learning_rate": 6.625273177415736e-08, + "loss": 0.3468, + "step": 43144 + }, + { + "epoch": 0.949702839533348, + "grad_norm": 1.5772282686529848, + "learning_rate": 6.619490876654922e-08, + "loss": 0.2973, + "step": 43145 + }, + { + "epoch": 0.9497248514197667, + "grad_norm": 1.2908035898223462, + "learning_rate": 6.613711083473973e-08, + "loss": 0.2592, + "step": 43146 + }, + { + "epoch": 0.9497468633061853, + "grad_norm": 1.6088683215478516, + "learning_rate": 6.607933797902144e-08, + "loss": 0.3294, + "step": 43147 + }, + { + "epoch": 0.949768875192604, + "grad_norm": 2.545481334855923, + "learning_rate": 6.602159019968912e-08, + "loss": 0.3063, + "step": 43148 + }, + { + "epoch": 0.9497908870790227, + "grad_norm": 1.446356614749584, + "learning_rate": 6.596386749703532e-08, + "loss": 0.2966, + "step": 43149 + }, + { + "epoch": 0.9498128989654413, + "grad_norm": 2.4115983314818275, + "learning_rate": 6.590616987135367e-08, + "loss": 0.2628, + "step": 43150 + }, + { + "epoch": 0.94983491085186, + "grad_norm": 0.5937393246330198, + "learning_rate": 6.584849732293786e-08, + "loss": 0.4316, + "step": 43151 + }, + { + "epoch": 0.9498569227382787, + "grad_norm": 1.6290994982839653, + "learning_rate": 6.57908498520804e-08, + "loss": 0.3605, + "step": 43152 + }, + { + "epoch": 0.9498789346246973, + "grad_norm": 0.6008327468949975, + "learning_rate": 6.573322745907496e-08, + "loss": 0.428, + "step": 43153 + }, + { + "epoch": 0.949900946511116, + "grad_norm": 1.33266539311569, + "learning_rate": 6.567563014421408e-08, + "loss": 0.2372, + "step": 43154 + }, + { + "epoch": 0.9499229583975347, + "grad_norm": 0.6127185881905968, + "learning_rate": 6.561805790778974e-08, + "loss": 0.4531, + "step": 43155 + }, + { + "epoch": 0.9499449702839533, + "grad_norm": 1.7265845375491207, + "learning_rate": 6.556051075009562e-08, + "loss": 0.2487, + "step": 43156 + }, + { + "epoch": 0.949966982170372, + "grad_norm": 7.9991463023211695, + "learning_rate": 6.550298867142368e-08, + "loss": 0.2715, + "step": 43157 + }, + { + "epoch": 0.9499889940567907, + "grad_norm": 2.0645002857103476, + "learning_rate": 6.544549167206704e-08, + "loss": 0.3132, + "step": 43158 + }, + { + "epoch": 0.9500110059432093, + "grad_norm": 1.5061580560388868, + "learning_rate": 6.538801975231656e-08, + "loss": 0.2528, + "step": 43159 + }, + { + "epoch": 0.950033017829628, + "grad_norm": 2.1829564359894893, + "learning_rate": 6.533057291246481e-08, + "loss": 0.3078, + "step": 43160 + }, + { + "epoch": 0.9500550297160467, + "grad_norm": 3.3127056091958553, + "learning_rate": 6.527315115280486e-08, + "loss": 0.3335, + "step": 43161 + }, + { + "epoch": 0.9500770416024653, + "grad_norm": 2.2777205724136405, + "learning_rate": 6.52157544736276e-08, + "loss": 0.2513, + "step": 43162 + }, + { + "epoch": 0.950099053488884, + "grad_norm": 1.3871180214265, + "learning_rate": 6.515838287522447e-08, + "loss": 0.2792, + "step": 43163 + }, + { + "epoch": 0.9501210653753026, + "grad_norm": 1.37223636710313, + "learning_rate": 6.510103635788745e-08, + "loss": 0.1828, + "step": 43164 + }, + { + "epoch": 0.9501430772617213, + "grad_norm": 0.6229174956707234, + "learning_rate": 6.504371492190798e-08, + "loss": 0.4431, + "step": 43165 + }, + { + "epoch": 0.95016508914814, + "grad_norm": 0.5992112209106941, + "learning_rate": 6.498641856757749e-08, + "loss": 0.4396, + "step": 43166 + }, + { + "epoch": 0.9501871010345586, + "grad_norm": 0.633058787651997, + "learning_rate": 6.49291472951874e-08, + "loss": 0.4323, + "step": 43167 + }, + { + "epoch": 0.9502091129209773, + "grad_norm": 0.6034671233423028, + "learning_rate": 6.487190110502806e-08, + "loss": 0.4486, + "step": 43168 + }, + { + "epoch": 0.950231124807396, + "grad_norm": 0.6295040259166931, + "learning_rate": 6.481467999739088e-08, + "loss": 0.437, + "step": 43169 + }, + { + "epoch": 0.9502531366938146, + "grad_norm": 1.941833519227745, + "learning_rate": 6.475748397256676e-08, + "loss": 0.2764, + "step": 43170 + }, + { + "epoch": 0.9502751485802333, + "grad_norm": 2.4086508833468074, + "learning_rate": 6.4700313030846e-08, + "loss": 0.2802, + "step": 43171 + }, + { + "epoch": 0.950297160466652, + "grad_norm": 1.5339400018290261, + "learning_rate": 6.464316717252006e-08, + "loss": 0.3284, + "step": 43172 + }, + { + "epoch": 0.9503191723530706, + "grad_norm": 2.3010183674194673, + "learning_rate": 6.458604639787814e-08, + "loss": 0.3667, + "step": 43173 + }, + { + "epoch": 0.9503411842394893, + "grad_norm": 2.0120739953645983, + "learning_rate": 6.452895070721166e-08, + "loss": 0.2726, + "step": 43174 + }, + { + "epoch": 0.950363196125908, + "grad_norm": 2.535001017300522, + "learning_rate": 6.447188010080985e-08, + "loss": 0.3636, + "step": 43175 + }, + { + "epoch": 0.9503852080123266, + "grad_norm": 0.6117979376435345, + "learning_rate": 6.441483457896358e-08, + "loss": 0.4415, + "step": 43176 + }, + { + "epoch": 0.9504072198987453, + "grad_norm": 1.5376248477068748, + "learning_rate": 6.435781414196263e-08, + "loss": 0.2539, + "step": 43177 + }, + { + "epoch": 0.950429231785164, + "grad_norm": 1.8191622615574115, + "learning_rate": 6.43008187900962e-08, + "loss": 0.3158, + "step": 43178 + }, + { + "epoch": 0.9504512436715826, + "grad_norm": 0.6163126438471355, + "learning_rate": 6.424384852365518e-08, + "loss": 0.4236, + "step": 43179 + }, + { + "epoch": 0.9504732555580013, + "grad_norm": 1.8238671593979676, + "learning_rate": 6.41869033429271e-08, + "loss": 0.3139, + "step": 43180 + }, + { + "epoch": 0.95049526744442, + "grad_norm": 1.8283415781846462, + "learning_rate": 6.412998324820341e-08, + "loss": 0.3681, + "step": 43181 + }, + { + "epoch": 0.9505172793308386, + "grad_norm": 2.554323618724786, + "learning_rate": 6.407308823977276e-08, + "loss": 0.2935, + "step": 43182 + }, + { + "epoch": 0.9505392912172573, + "grad_norm": 1.8387291153704628, + "learning_rate": 6.401621831792382e-08, + "loss": 0.3148, + "step": 43183 + }, + { + "epoch": 0.950561303103676, + "grad_norm": 1.8459944102229273, + "learning_rate": 6.395937348294579e-08, + "loss": 0.2604, + "step": 43184 + }, + { + "epoch": 0.9505833149900946, + "grad_norm": 1.1611371119079492, + "learning_rate": 6.390255373512788e-08, + "loss": 0.1909, + "step": 43185 + }, + { + "epoch": 0.9506053268765133, + "grad_norm": 2.1563911965312528, + "learning_rate": 6.384575907475876e-08, + "loss": 0.2946, + "step": 43186 + }, + { + "epoch": 0.9506273387629319, + "grad_norm": 1.361642468692422, + "learning_rate": 6.378898950212765e-08, + "loss": 0.2761, + "step": 43187 + }, + { + "epoch": 0.9506493506493506, + "grad_norm": 1.8545879610949905, + "learning_rate": 6.373224501752206e-08, + "loss": 0.2406, + "step": 43188 + }, + { + "epoch": 0.9506713625357693, + "grad_norm": 1.6301493060359196, + "learning_rate": 6.36755256212307e-08, + "loss": 0.3183, + "step": 43189 + }, + { + "epoch": 0.9506933744221879, + "grad_norm": 1.7051956153220007, + "learning_rate": 6.361883131354218e-08, + "loss": 0.3826, + "step": 43190 + }, + { + "epoch": 0.9507153863086066, + "grad_norm": 1.4296345433959459, + "learning_rate": 6.356216209474408e-08, + "loss": 0.2395, + "step": 43191 + }, + { + "epoch": 0.9507373981950253, + "grad_norm": 2.2811359420159394, + "learning_rate": 6.350551796512506e-08, + "loss": 0.3462, + "step": 43192 + }, + { + "epoch": 0.9507594100814439, + "grad_norm": 2.08732569313554, + "learning_rate": 6.344889892497263e-08, + "loss": 0.347, + "step": 43193 + }, + { + "epoch": 0.9507814219678626, + "grad_norm": 1.6350547095112435, + "learning_rate": 6.339230497457494e-08, + "loss": 0.2739, + "step": 43194 + }, + { + "epoch": 0.9508034338542813, + "grad_norm": 0.6136896108599821, + "learning_rate": 6.333573611421951e-08, + "loss": 0.4472, + "step": 43195 + }, + { + "epoch": 0.9508254457406999, + "grad_norm": 2.001145389055776, + "learning_rate": 6.327919234419334e-08, + "loss": 0.2749, + "step": 43196 + }, + { + "epoch": 0.9508474576271186, + "grad_norm": 2.0822476575089874, + "learning_rate": 6.322267366478396e-08, + "loss": 0.2298, + "step": 43197 + }, + { + "epoch": 0.9508694695135373, + "grad_norm": 2.4046613933351626, + "learning_rate": 6.316618007627895e-08, + "loss": 0.3162, + "step": 43198 + }, + { + "epoch": 0.9508914813999559, + "grad_norm": 1.6058019882488304, + "learning_rate": 6.310971157896528e-08, + "loss": 0.3056, + "step": 43199 + }, + { + "epoch": 0.9509134932863746, + "grad_norm": 0.6160310724925349, + "learning_rate": 6.305326817313051e-08, + "loss": 0.4259, + "step": 43200 + }, + { + "epoch": 0.9509355051727933, + "grad_norm": 1.366170266945715, + "learning_rate": 6.299684985906051e-08, + "loss": 0.2462, + "step": 43201 + }, + { + "epoch": 0.9509575170592119, + "grad_norm": 1.5691147653731063, + "learning_rate": 6.294045663704229e-08, + "loss": 0.2195, + "step": 43202 + }, + { + "epoch": 0.9509795289456306, + "grad_norm": 1.513897024063294, + "learning_rate": 6.288408850736283e-08, + "loss": 0.262, + "step": 43203 + }, + { + "epoch": 0.9510015408320494, + "grad_norm": 1.7639410253072758, + "learning_rate": 6.282774547030856e-08, + "loss": 0.305, + "step": 43204 + }, + { + "epoch": 0.951023552718468, + "grad_norm": 1.625669779510363, + "learning_rate": 6.277142752616594e-08, + "loss": 0.3385, + "step": 43205 + }, + { + "epoch": 0.9510455646048867, + "grad_norm": 1.5975534894257768, + "learning_rate": 6.271513467522028e-08, + "loss": 0.2643, + "step": 43206 + }, + { + "epoch": 0.9510675764913054, + "grad_norm": 0.6289027706801994, + "learning_rate": 6.265886691775914e-08, + "loss": 0.4334, + "step": 43207 + }, + { + "epoch": 0.951089588377724, + "grad_norm": 1.619033957597926, + "learning_rate": 6.260262425406726e-08, + "loss": 0.3502, + "step": 43208 + }, + { + "epoch": 0.9511116002641427, + "grad_norm": 1.6693331053553673, + "learning_rate": 6.254640668443113e-08, + "loss": 0.2794, + "step": 43209 + }, + { + "epoch": 0.9511336121505614, + "grad_norm": 1.7673012624447664, + "learning_rate": 6.249021420913658e-08, + "loss": 0.3091, + "step": 43210 + }, + { + "epoch": 0.95115562403698, + "grad_norm": 1.9198331856953623, + "learning_rate": 6.243404682846843e-08, + "loss": 0.2258, + "step": 43211 + }, + { + "epoch": 0.9511776359233987, + "grad_norm": 5.722180941199845, + "learning_rate": 6.237790454271308e-08, + "loss": 0.4017, + "step": 43212 + }, + { + "epoch": 0.9511996478098173, + "grad_norm": 2.427390807219058, + "learning_rate": 6.232178735215533e-08, + "loss": 0.296, + "step": 43213 + }, + { + "epoch": 0.951221659696236, + "grad_norm": 1.8765757830430982, + "learning_rate": 6.226569525708048e-08, + "loss": 0.2956, + "step": 43214 + }, + { + "epoch": 0.9512436715826547, + "grad_norm": 1.8862120510991438, + "learning_rate": 6.220962825777388e-08, + "loss": 0.244, + "step": 43215 + }, + { + "epoch": 0.9512656834690733, + "grad_norm": 1.5150953945121954, + "learning_rate": 6.215358635452029e-08, + "loss": 0.302, + "step": 43216 + }, + { + "epoch": 0.951287695355492, + "grad_norm": 2.4701387522606426, + "learning_rate": 6.209756954760393e-08, + "loss": 0.3356, + "step": 43217 + }, + { + "epoch": 0.9513097072419107, + "grad_norm": 3.113766667930834, + "learning_rate": 6.204157783731068e-08, + "loss": 0.2698, + "step": 43218 + }, + { + "epoch": 0.9513317191283293, + "grad_norm": 1.8960008612324775, + "learning_rate": 6.198561122392421e-08, + "loss": 0.3201, + "step": 43219 + }, + { + "epoch": 0.951353731014748, + "grad_norm": 1.5618673034375585, + "learning_rate": 6.192966970772985e-08, + "loss": 0.339, + "step": 43220 + }, + { + "epoch": 0.9513757429011667, + "grad_norm": 1.589824130219561, + "learning_rate": 6.187375328901069e-08, + "loss": 0.3624, + "step": 43221 + }, + { + "epoch": 0.9513977547875853, + "grad_norm": 1.425842981199371, + "learning_rate": 6.181786196805151e-08, + "loss": 0.2682, + "step": 43222 + }, + { + "epoch": 0.951419766674004, + "grad_norm": 3.0363114384537555, + "learning_rate": 6.176199574513708e-08, + "loss": 0.3154, + "step": 43223 + }, + { + "epoch": 0.9514417785604227, + "grad_norm": 1.9658093130979573, + "learning_rate": 6.170615462054997e-08, + "loss": 0.2855, + "step": 43224 + }, + { + "epoch": 0.9514637904468413, + "grad_norm": 2.1403240316570336, + "learning_rate": 6.165033859457547e-08, + "loss": 0.2987, + "step": 43225 + }, + { + "epoch": 0.95148580233326, + "grad_norm": 2.1251220889948788, + "learning_rate": 6.159454766749618e-08, + "loss": 0.3147, + "step": 43226 + }, + { + "epoch": 0.9515078142196787, + "grad_norm": 1.3751191533966876, + "learning_rate": 6.153878183959571e-08, + "loss": 0.2921, + "step": 43227 + }, + { + "epoch": 0.9515298261060973, + "grad_norm": 2.5591507365446073, + "learning_rate": 6.148304111115777e-08, + "loss": 0.3386, + "step": 43228 + }, + { + "epoch": 0.951551837992516, + "grad_norm": 0.6084415277166672, + "learning_rate": 6.1427325482466e-08, + "loss": 0.4261, + "step": 43229 + }, + { + "epoch": 0.9515738498789347, + "grad_norm": 2.0407895837437056, + "learning_rate": 6.13716349538035e-08, + "loss": 0.2516, + "step": 43230 + }, + { + "epoch": 0.9515958617653533, + "grad_norm": 1.478537487789829, + "learning_rate": 6.131596952545282e-08, + "loss": 0.2519, + "step": 43231 + }, + { + "epoch": 0.951617873651772, + "grad_norm": 1.4381181016009643, + "learning_rate": 6.126032919769653e-08, + "loss": 0.3138, + "step": 43232 + }, + { + "epoch": 0.9516398855381907, + "grad_norm": 0.598439473230492, + "learning_rate": 6.120471397081829e-08, + "loss": 0.4198, + "step": 43233 + }, + { + "epoch": 0.9516618974246093, + "grad_norm": 2.465274823024728, + "learning_rate": 6.114912384510064e-08, + "loss": 0.3042, + "step": 43234 + }, + { + "epoch": 0.951683909311028, + "grad_norm": 4.705812132606106, + "learning_rate": 6.109355882082612e-08, + "loss": 0.3149, + "step": 43235 + }, + { + "epoch": 0.9517059211974466, + "grad_norm": 2.4836410260499924, + "learning_rate": 6.103801889827677e-08, + "loss": 0.3273, + "step": 43236 + }, + { + "epoch": 0.9517279330838653, + "grad_norm": 2.070855478601703, + "learning_rate": 6.09825040777351e-08, + "loss": 0.3467, + "step": 43237 + }, + { + "epoch": 0.951749944970284, + "grad_norm": 2.128673100118615, + "learning_rate": 6.092701435948257e-08, + "loss": 0.233, + "step": 43238 + }, + { + "epoch": 0.9517719568567026, + "grad_norm": 0.5995329345471805, + "learning_rate": 6.087154974380227e-08, + "loss": 0.4167, + "step": 43239 + }, + { + "epoch": 0.9517939687431213, + "grad_norm": 1.5183397529900249, + "learning_rate": 6.081611023097622e-08, + "loss": 0.2704, + "step": 43240 + }, + { + "epoch": 0.95181598062954, + "grad_norm": 1.5897724361524717, + "learning_rate": 6.076069582128529e-08, + "loss": 0.3276, + "step": 43241 + }, + { + "epoch": 0.9518379925159586, + "grad_norm": 2.165916998532708, + "learning_rate": 6.070530651501095e-08, + "loss": 0.2805, + "step": 43242 + }, + { + "epoch": 0.9518600044023773, + "grad_norm": 10.513362854647477, + "learning_rate": 6.06499423124357e-08, + "loss": 0.2141, + "step": 43243 + }, + { + "epoch": 0.951882016288796, + "grad_norm": 2.458799987272344, + "learning_rate": 6.059460321384048e-08, + "loss": 0.3209, + "step": 43244 + }, + { + "epoch": 0.9519040281752146, + "grad_norm": 2.845475147061233, + "learning_rate": 6.053928921950614e-08, + "loss": 0.2835, + "step": 43245 + }, + { + "epoch": 0.9519260400616333, + "grad_norm": 2.99374068728487, + "learning_rate": 6.048400032971468e-08, + "loss": 0.2462, + "step": 43246 + }, + { + "epoch": 0.951948051948052, + "grad_norm": 2.5515318039193784, + "learning_rate": 6.042873654474646e-08, + "loss": 0.3102, + "step": 43247 + }, + { + "epoch": 0.9519700638344706, + "grad_norm": 1.635413646345883, + "learning_rate": 6.037349786488234e-08, + "loss": 0.322, + "step": 43248 + }, + { + "epoch": 0.9519920757208893, + "grad_norm": 0.6291486214205207, + "learning_rate": 6.03182842904032e-08, + "loss": 0.4408, + "step": 43249 + }, + { + "epoch": 0.952014087607308, + "grad_norm": 1.763137699081051, + "learning_rate": 6.026309582158996e-08, + "loss": 0.3473, + "step": 43250 + }, + { + "epoch": 0.9520360994937266, + "grad_norm": 1.5591122777147544, + "learning_rate": 6.020793245872292e-08, + "loss": 0.2991, + "step": 43251 + }, + { + "epoch": 0.9520581113801453, + "grad_norm": 0.6139577581348339, + "learning_rate": 6.015279420208187e-08, + "loss": 0.445, + "step": 43252 + }, + { + "epoch": 0.952080123266564, + "grad_norm": 2.286282722111597, + "learning_rate": 6.009768105194769e-08, + "loss": 0.2358, + "step": 43253 + }, + { + "epoch": 0.9521021351529826, + "grad_norm": 1.7563134000380725, + "learning_rate": 6.004259300860072e-08, + "loss": 0.3099, + "step": 43254 + }, + { + "epoch": 0.9521241470394013, + "grad_norm": 1.7248090211777707, + "learning_rate": 5.998753007232016e-08, + "loss": 0.3679, + "step": 43255 + }, + { + "epoch": 0.95214615892582, + "grad_norm": 1.3445665902013475, + "learning_rate": 5.993249224338693e-08, + "loss": 0.1931, + "step": 43256 + }, + { + "epoch": 0.9521681708122386, + "grad_norm": 2.188409594580418, + "learning_rate": 5.987747952207967e-08, + "loss": 0.353, + "step": 43257 + }, + { + "epoch": 0.9521901826986573, + "grad_norm": 2.4520515016612285, + "learning_rate": 5.982249190867816e-08, + "loss": 0.235, + "step": 43258 + }, + { + "epoch": 0.9522121945850759, + "grad_norm": 2.1674468293484517, + "learning_rate": 5.976752940346275e-08, + "loss": 0.323, + "step": 43259 + }, + { + "epoch": 0.9522342064714946, + "grad_norm": 0.620147649067992, + "learning_rate": 5.971259200671153e-08, + "loss": 0.4518, + "step": 43260 + }, + { + "epoch": 0.9522562183579133, + "grad_norm": 1.8544461424591971, + "learning_rate": 5.965767971870484e-08, + "loss": 0.2527, + "step": 43261 + }, + { + "epoch": 0.9522782302443319, + "grad_norm": 1.8080870845034183, + "learning_rate": 5.960279253972134e-08, + "loss": 0.3083, + "step": 43262 + }, + { + "epoch": 0.9523002421307506, + "grad_norm": 1.6665147684295618, + "learning_rate": 5.954793047003915e-08, + "loss": 0.2525, + "step": 43263 + }, + { + "epoch": 0.9523222540171693, + "grad_norm": 0.6189344329021369, + "learning_rate": 5.949309350993859e-08, + "loss": 0.4527, + "step": 43264 + }, + { + "epoch": 0.9523442659035879, + "grad_norm": 2.4917722435971763, + "learning_rate": 5.943828165969723e-08, + "loss": 0.424, + "step": 43265 + }, + { + "epoch": 0.9523662777900066, + "grad_norm": 1.5002621892213748, + "learning_rate": 5.9383494919594834e-08, + "loss": 0.2307, + "step": 43266 + }, + { + "epoch": 0.9523882896764253, + "grad_norm": 0.6218873065711722, + "learning_rate": 5.93287332899084e-08, + "loss": 0.4284, + "step": 43267 + }, + { + "epoch": 0.9524103015628439, + "grad_norm": 1.535260896995461, + "learning_rate": 5.927399677091716e-08, + "loss": 0.2726, + "step": 43268 + }, + { + "epoch": 0.9524323134492626, + "grad_norm": 1.769818865878759, + "learning_rate": 5.921928536289867e-08, + "loss": 0.2888, + "step": 43269 + }, + { + "epoch": 0.9524543253356813, + "grad_norm": 3.3992740961710557, + "learning_rate": 5.916459906613159e-08, + "loss": 0.2837, + "step": 43270 + }, + { + "epoch": 0.9524763372220999, + "grad_norm": 2.014458079385053, + "learning_rate": 5.910993788089403e-08, + "loss": 0.3392, + "step": 43271 + }, + { + "epoch": 0.9524983491085186, + "grad_norm": 2.363849183187576, + "learning_rate": 5.905530180746355e-08, + "loss": 0.3426, + "step": 43272 + }, + { + "epoch": 0.9525203609949373, + "grad_norm": 1.9160558871239524, + "learning_rate": 5.90006908461177e-08, + "loss": 0.3214, + "step": 43273 + }, + { + "epoch": 0.9525423728813559, + "grad_norm": 3.841988041585182, + "learning_rate": 5.894610499713349e-08, + "loss": 0.2726, + "step": 43274 + }, + { + "epoch": 0.9525643847677746, + "grad_norm": 1.9864967670504587, + "learning_rate": 5.889154426078958e-08, + "loss": 0.2628, + "step": 43275 + }, + { + "epoch": 0.9525863966541933, + "grad_norm": 1.465628372462684, + "learning_rate": 5.883700863736242e-08, + "loss": 0.2734, + "step": 43276 + }, + { + "epoch": 0.9526084085406119, + "grad_norm": 0.6215908050003164, + "learning_rate": 5.878249812712955e-08, + "loss": 0.446, + "step": 43277 + }, + { + "epoch": 0.9526304204270306, + "grad_norm": 1.9887919076858898, + "learning_rate": 5.872801273036743e-08, + "loss": 0.2788, + "step": 43278 + }, + { + "epoch": 0.9526524323134493, + "grad_norm": 1.788134917872, + "learning_rate": 5.867355244735362e-08, + "loss": 0.2935, + "step": 43279 + }, + { + "epoch": 0.9526744441998679, + "grad_norm": 2.1705779851241496, + "learning_rate": 5.861911727836456e-08, + "loss": 0.2857, + "step": 43280 + }, + { + "epoch": 0.9526964560862866, + "grad_norm": 2.775869744224707, + "learning_rate": 5.8564707223677795e-08, + "loss": 0.3558, + "step": 43281 + }, + { + "epoch": 0.9527184679727052, + "grad_norm": 0.6232498014502549, + "learning_rate": 5.8510322283568674e-08, + "loss": 0.4658, + "step": 43282 + }, + { + "epoch": 0.9527404798591239, + "grad_norm": 2.245535004714262, + "learning_rate": 5.84559624583142e-08, + "loss": 0.2445, + "step": 43283 + }, + { + "epoch": 0.9527624917455426, + "grad_norm": 2.4679384141808143, + "learning_rate": 5.840162774819025e-08, + "loss": 0.3201, + "step": 43284 + }, + { + "epoch": 0.9527845036319612, + "grad_norm": 0.632304344499411, + "learning_rate": 5.834731815347327e-08, + "loss": 0.4632, + "step": 43285 + }, + { + "epoch": 0.9528065155183799, + "grad_norm": 1.6148734254769697, + "learning_rate": 5.829303367443917e-08, + "loss": 0.1969, + "step": 43286 + }, + { + "epoch": 0.9528285274047986, + "grad_norm": 1.7045579522701972, + "learning_rate": 5.823877431136438e-08, + "loss": 0.3923, + "step": 43287 + }, + { + "epoch": 0.9528505392912172, + "grad_norm": 1.6545317872994785, + "learning_rate": 5.8184540064523675e-08, + "loss": 0.263, + "step": 43288 + }, + { + "epoch": 0.9528725511776359, + "grad_norm": 2.0995163348911956, + "learning_rate": 5.8130330934193515e-08, + "loss": 0.3688, + "step": 43289 + }, + { + "epoch": 0.9528945630640546, + "grad_norm": 2.3278666544968996, + "learning_rate": 5.8076146920649226e-08, + "loss": 0.2336, + "step": 43290 + }, + { + "epoch": 0.9529165749504732, + "grad_norm": 1.635394996445608, + "learning_rate": 5.8021988024166146e-08, + "loss": 0.3169, + "step": 43291 + }, + { + "epoch": 0.9529385868368919, + "grad_norm": 2.658122816326397, + "learning_rate": 5.796785424501905e-08, + "loss": 0.2961, + "step": 43292 + }, + { + "epoch": 0.9529605987233106, + "grad_norm": 0.6084454973374857, + "learning_rate": 5.791374558348384e-08, + "loss": 0.4575, + "step": 43293 + }, + { + "epoch": 0.9529826106097292, + "grad_norm": 1.5753697440887338, + "learning_rate": 5.7859662039835286e-08, + "loss": 0.2344, + "step": 43294 + }, + { + "epoch": 0.9530046224961479, + "grad_norm": 2.2781369461754775, + "learning_rate": 5.780560361434817e-08, + "loss": 0.3149, + "step": 43295 + }, + { + "epoch": 0.9530266343825666, + "grad_norm": 1.7124953473828175, + "learning_rate": 5.7751570307297275e-08, + "loss": 0.3146, + "step": 43296 + }, + { + "epoch": 0.9530486462689852, + "grad_norm": 1.8310364696662194, + "learning_rate": 5.7697562118957384e-08, + "loss": 0.2677, + "step": 43297 + }, + { + "epoch": 0.9530706581554039, + "grad_norm": 0.6042545006479826, + "learning_rate": 5.764357904960216e-08, + "loss": 0.4129, + "step": 43298 + }, + { + "epoch": 0.9530926700418226, + "grad_norm": 2.103338081945774, + "learning_rate": 5.7589621099506946e-08, + "loss": 0.2154, + "step": 43299 + }, + { + "epoch": 0.9531146819282412, + "grad_norm": 1.3758313154654036, + "learning_rate": 5.753568826894596e-08, + "loss": 0.2472, + "step": 43300 + }, + { + "epoch": 0.9531366938146599, + "grad_norm": 1.5440323181987208, + "learning_rate": 5.748178055819287e-08, + "loss": 0.3086, + "step": 43301 + }, + { + "epoch": 0.9531587057010786, + "grad_norm": 2.16702739957727, + "learning_rate": 5.742789796752191e-08, + "loss": 0.2921, + "step": 43302 + }, + { + "epoch": 0.9531807175874972, + "grad_norm": 1.4629568862049602, + "learning_rate": 5.737404049720674e-08, + "loss": 0.262, + "step": 43303 + }, + { + "epoch": 0.9532027294739159, + "grad_norm": 1.35505933185188, + "learning_rate": 5.732020814752104e-08, + "loss": 0.2548, + "step": 43304 + }, + { + "epoch": 0.9532247413603345, + "grad_norm": 1.638208283618866, + "learning_rate": 5.7266400918738475e-08, + "loss": 0.3177, + "step": 43305 + }, + { + "epoch": 0.9532467532467532, + "grad_norm": 2.099826582677287, + "learning_rate": 5.721261881113216e-08, + "loss": 0.2424, + "step": 43306 + }, + { + "epoch": 0.9532687651331719, + "grad_norm": 1.6057499259539463, + "learning_rate": 5.715886182497687e-08, + "loss": 0.2831, + "step": 43307 + }, + { + "epoch": 0.9532907770195905, + "grad_norm": 1.676535832988396, + "learning_rate": 5.7105129960544626e-08, + "loss": 0.3562, + "step": 43308 + }, + { + "epoch": 0.9533127889060092, + "grad_norm": 2.248463228056836, + "learning_rate": 5.705142321810797e-08, + "loss": 0.3684, + "step": 43309 + }, + { + "epoch": 0.9533348007924279, + "grad_norm": 2.251089263100961, + "learning_rate": 5.6997741597941136e-08, + "loss": 0.4027, + "step": 43310 + }, + { + "epoch": 0.9533568126788465, + "grad_norm": 1.5360631804026872, + "learning_rate": 5.6944085100316125e-08, + "loss": 0.2332, + "step": 43311 + }, + { + "epoch": 0.9533788245652652, + "grad_norm": 1.5303485260812209, + "learning_rate": 5.689045372550606e-08, + "loss": 0.2273, + "step": 43312 + }, + { + "epoch": 0.9534008364516839, + "grad_norm": 1.425480626225615, + "learning_rate": 5.683684747378349e-08, + "loss": 0.2564, + "step": 43313 + }, + { + "epoch": 0.9534228483381025, + "grad_norm": 1.5300739734532112, + "learning_rate": 5.6783266345420974e-08, + "loss": 0.2543, + "step": 43314 + }, + { + "epoch": 0.9534448602245212, + "grad_norm": 1.5862474755552916, + "learning_rate": 5.672971034068997e-08, + "loss": 0.4169, + "step": 43315 + }, + { + "epoch": 0.95346687211094, + "grad_norm": 1.5369290316683863, + "learning_rate": 5.667617945986359e-08, + "loss": 0.2064, + "step": 43316 + }, + { + "epoch": 0.9534888839973585, + "grad_norm": 4.2681666770171285, + "learning_rate": 5.662267370321439e-08, + "loss": 0.3204, + "step": 43317 + }, + { + "epoch": 0.9535108958837772, + "grad_norm": 0.5995883997531333, + "learning_rate": 5.6569193071012715e-08, + "loss": 0.4382, + "step": 43318 + }, + { + "epoch": 0.953532907770196, + "grad_norm": 1.7617625658487632, + "learning_rate": 5.651573756353168e-08, + "loss": 0.3097, + "step": 43319 + }, + { + "epoch": 0.9535549196566145, + "grad_norm": 2.0281430106208775, + "learning_rate": 5.646230718104218e-08, + "loss": 0.2646, + "step": 43320 + }, + { + "epoch": 0.9535769315430332, + "grad_norm": 2.411559017745305, + "learning_rate": 5.64089019238162e-08, + "loss": 0.2895, + "step": 43321 + }, + { + "epoch": 0.953598943429452, + "grad_norm": 1.466043042698942, + "learning_rate": 5.635552179212522e-08, + "loss": 0.2965, + "step": 43322 + }, + { + "epoch": 0.9536209553158705, + "grad_norm": 2.2539797271974042, + "learning_rate": 5.6302166786240675e-08, + "loss": 0.3455, + "step": 43323 + }, + { + "epoch": 0.9536429672022892, + "grad_norm": 1.7066441180270147, + "learning_rate": 5.62488369064329e-08, + "loss": 0.3384, + "step": 43324 + }, + { + "epoch": 0.953664979088708, + "grad_norm": 1.3390107502088902, + "learning_rate": 5.619553215297391e-08, + "loss": 0.2386, + "step": 43325 + }, + { + "epoch": 0.9536869909751265, + "grad_norm": 1.9340520359824689, + "learning_rate": 5.6142252526134034e-08, + "loss": 0.25, + "step": 43326 + }, + { + "epoch": 0.9537090028615453, + "grad_norm": 1.6772439336985772, + "learning_rate": 5.608899802618417e-08, + "loss": 0.2976, + "step": 43327 + }, + { + "epoch": 0.9537310147479638, + "grad_norm": 2.973996328117363, + "learning_rate": 5.603576865339522e-08, + "loss": 0.2729, + "step": 43328 + }, + { + "epoch": 0.9537530266343826, + "grad_norm": 2.716467055071345, + "learning_rate": 5.598256440803751e-08, + "loss": 0.2991, + "step": 43329 + }, + { + "epoch": 0.9537750385208013, + "grad_norm": 1.560335553548507, + "learning_rate": 5.592938529038139e-08, + "loss": 0.3174, + "step": 43330 + }, + { + "epoch": 0.9537970504072198, + "grad_norm": 5.057108747335469, + "learning_rate": 5.587623130069775e-08, + "loss": 0.3217, + "step": 43331 + }, + { + "epoch": 0.9538190622936386, + "grad_norm": 1.7722166260635959, + "learning_rate": 5.582310243925582e-08, + "loss": 0.3013, + "step": 43332 + }, + { + "epoch": 0.9538410741800573, + "grad_norm": 3.5215497266103215, + "learning_rate": 5.5769998706325934e-08, + "loss": 0.2861, + "step": 43333 + }, + { + "epoch": 0.9538630860664759, + "grad_norm": 0.6125040031130954, + "learning_rate": 5.571692010217844e-08, + "loss": 0.4322, + "step": 43334 + }, + { + "epoch": 0.9538850979528946, + "grad_norm": 1.5088073718397694, + "learning_rate": 5.5663866627082563e-08, + "loss": 0.255, + "step": 43335 + }, + { + "epoch": 0.9539071098393133, + "grad_norm": 1.7952605993602517, + "learning_rate": 5.561083828130809e-08, + "loss": 0.3026, + "step": 43336 + }, + { + "epoch": 0.9539291217257319, + "grad_norm": 1.4257061952329992, + "learning_rate": 5.55578350651248e-08, + "loss": 0.248, + "step": 43337 + }, + { + "epoch": 0.9539511336121506, + "grad_norm": 2.317992186331226, + "learning_rate": 5.5504856978801926e-08, + "loss": 0.2799, + "step": 43338 + }, + { + "epoch": 0.9539731454985693, + "grad_norm": 1.333333689013901, + "learning_rate": 5.545190402260814e-08, + "loss": 0.2669, + "step": 43339 + }, + { + "epoch": 0.9539951573849879, + "grad_norm": 0.5922860291202676, + "learning_rate": 5.539897619681378e-08, + "loss": 0.4289, + "step": 43340 + }, + { + "epoch": 0.9540171692714066, + "grad_norm": 1.5343463951911107, + "learning_rate": 5.534607350168697e-08, + "loss": 0.2237, + "step": 43341 + }, + { + "epoch": 0.9540391811578253, + "grad_norm": 0.6085446571175656, + "learning_rate": 5.529319593749638e-08, + "loss": 0.4438, + "step": 43342 + }, + { + "epoch": 0.9540611930442439, + "grad_norm": 2.1908437512164625, + "learning_rate": 5.524034350451179e-08, + "loss": 0.3207, + "step": 43343 + }, + { + "epoch": 0.9540832049306626, + "grad_norm": 3.2046825212482246, + "learning_rate": 5.5187516203001334e-08, + "loss": 0.271, + "step": 43344 + }, + { + "epoch": 0.9541052168170813, + "grad_norm": 1.346527938000532, + "learning_rate": 5.513471403323256e-08, + "loss": 0.2452, + "step": 43345 + }, + { + "epoch": 0.9541272287034999, + "grad_norm": 1.6871341651166805, + "learning_rate": 5.5081936995474706e-08, + "loss": 0.2942, + "step": 43346 + }, + { + "epoch": 0.9541492405899186, + "grad_norm": 7.805742863697445, + "learning_rate": 5.502918508999644e-08, + "loss": 0.2946, + "step": 43347 + }, + { + "epoch": 0.9541712524763373, + "grad_norm": 3.1568071860288183, + "learning_rate": 5.4976458317065326e-08, + "loss": 0.2654, + "step": 43348 + }, + { + "epoch": 0.9541932643627559, + "grad_norm": 2.8683438642775116, + "learning_rate": 5.492375667694949e-08, + "loss": 0.3181, + "step": 43349 + }, + { + "epoch": 0.9542152762491746, + "grad_norm": 11.136452959809, + "learning_rate": 5.487108016991649e-08, + "loss": 0.281, + "step": 43350 + }, + { + "epoch": 0.9542372881355933, + "grad_norm": 2.244334949379841, + "learning_rate": 5.481842879623389e-08, + "loss": 0.3444, + "step": 43351 + }, + { + "epoch": 0.9542593000220119, + "grad_norm": 1.4599194329677057, + "learning_rate": 5.476580255616981e-08, + "loss": 0.2084, + "step": 43352 + }, + { + "epoch": 0.9542813119084306, + "grad_norm": 2.1581339338104057, + "learning_rate": 5.4713201449991815e-08, + "loss": 0.2133, + "step": 43353 + }, + { + "epoch": 0.9543033237948492, + "grad_norm": 1.394026311986912, + "learning_rate": 5.466062547796747e-08, + "loss": 0.2317, + "step": 43354 + }, + { + "epoch": 0.9543253356812679, + "grad_norm": 1.4987123739580444, + "learning_rate": 5.4608074640362664e-08, + "loss": 0.2912, + "step": 43355 + }, + { + "epoch": 0.9543473475676866, + "grad_norm": 1.5034633874059156, + "learning_rate": 5.4555548937445526e-08, + "loss": 0.2866, + "step": 43356 + }, + { + "epoch": 0.9543693594541052, + "grad_norm": 1.900668056090736, + "learning_rate": 5.4503048369482505e-08, + "loss": 0.2438, + "step": 43357 + }, + { + "epoch": 0.9543913713405239, + "grad_norm": 2.141952371571485, + "learning_rate": 5.445057293674172e-08, + "loss": 0.2566, + "step": 43358 + }, + { + "epoch": 0.9544133832269426, + "grad_norm": 4.472867564173945, + "learning_rate": 5.4398122639488515e-08, + "loss": 0.2901, + "step": 43359 + }, + { + "epoch": 0.9544353951133612, + "grad_norm": 2.183818552463282, + "learning_rate": 5.434569747798934e-08, + "loss": 0.2131, + "step": 43360 + }, + { + "epoch": 0.9544574069997799, + "grad_norm": 0.597420166979609, + "learning_rate": 5.429329745251122e-08, + "loss": 0.4089, + "step": 43361 + }, + { + "epoch": 0.9544794188861986, + "grad_norm": 2.269662005012821, + "learning_rate": 5.424092256332114e-08, + "loss": 0.3229, + "step": 43362 + }, + { + "epoch": 0.9545014307726172, + "grad_norm": 1.9568150444851387, + "learning_rate": 5.4188572810683904e-08, + "loss": 0.3563, + "step": 43363 + }, + { + "epoch": 0.9545234426590359, + "grad_norm": 2.4110162941955933, + "learning_rate": 5.4136248194866516e-08, + "loss": 0.2995, + "step": 43364 + }, + { + "epoch": 0.9545454545454546, + "grad_norm": 0.6346681263133193, + "learning_rate": 5.4083948716134315e-08, + "loss": 0.4378, + "step": 43365 + }, + { + "epoch": 0.9545674664318732, + "grad_norm": 1.86540859219532, + "learning_rate": 5.403167437475321e-08, + "loss": 0.2754, + "step": 43366 + }, + { + "epoch": 0.9545894783182919, + "grad_norm": 1.7604967112488235, + "learning_rate": 5.397942517099019e-08, + "loss": 0.2368, + "step": 43367 + }, + { + "epoch": 0.9546114902047106, + "grad_norm": 1.9128520758558152, + "learning_rate": 5.392720110510896e-08, + "loss": 0.2774, + "step": 43368 + }, + { + "epoch": 0.9546335020911292, + "grad_norm": 1.8056474952078325, + "learning_rate": 5.387500217737596e-08, + "loss": 0.3114, + "step": 43369 + }, + { + "epoch": 0.9546555139775479, + "grad_norm": 2.3132115220524936, + "learning_rate": 5.382282838805597e-08, + "loss": 0.3428, + "step": 43370 + }, + { + "epoch": 0.9546775258639666, + "grad_norm": 2.551705639616188, + "learning_rate": 5.377067973741434e-08, + "loss": 0.2603, + "step": 43371 + }, + { + "epoch": 0.9546995377503852, + "grad_norm": 2.08161919429158, + "learning_rate": 5.3718556225716423e-08, + "loss": 0.2784, + "step": 43372 + }, + { + "epoch": 0.9547215496368039, + "grad_norm": 0.6293765885607485, + "learning_rate": 5.366645785322644e-08, + "loss": 0.43, + "step": 43373 + }, + { + "epoch": 0.9547435615232226, + "grad_norm": 1.7959876447798426, + "learning_rate": 5.361438462021029e-08, + "loss": 0.3097, + "step": 43374 + }, + { + "epoch": 0.9547655734096412, + "grad_norm": 4.607565509017608, + "learning_rate": 5.356233652693166e-08, + "loss": 0.3199, + "step": 43375 + }, + { + "epoch": 0.9547875852960599, + "grad_norm": 1.5731836245791624, + "learning_rate": 5.351031357365588e-08, + "loss": 0.3343, + "step": 43376 + }, + { + "epoch": 0.9548095971824785, + "grad_norm": 1.9765171105565222, + "learning_rate": 5.345831576064664e-08, + "loss": 0.3493, + "step": 43377 + }, + { + "epoch": 0.9548316090688972, + "grad_norm": 1.9686158818639226, + "learning_rate": 5.3406343088168164e-08, + "loss": 0.3009, + "step": 43378 + }, + { + "epoch": 0.9548536209553159, + "grad_norm": 1.8136215965472484, + "learning_rate": 5.335439555648525e-08, + "loss": 0.2681, + "step": 43379 + }, + { + "epoch": 0.9548756328417345, + "grad_norm": 2.471305100644126, + "learning_rate": 5.3302473165861014e-08, + "loss": 0.2836, + "step": 43380 + }, + { + "epoch": 0.9548976447281532, + "grad_norm": 1.7398034093552925, + "learning_rate": 5.325057591656024e-08, + "loss": 0.3906, + "step": 43381 + }, + { + "epoch": 0.9549196566145719, + "grad_norm": 1.5807076176415353, + "learning_rate": 5.319870380884662e-08, + "loss": 0.2821, + "step": 43382 + }, + { + "epoch": 0.9549416685009905, + "grad_norm": 1.790031250309207, + "learning_rate": 5.314685684298271e-08, + "loss": 0.285, + "step": 43383 + }, + { + "epoch": 0.9549636803874092, + "grad_norm": 2.0500569036202214, + "learning_rate": 5.3095035019233856e-08, + "loss": 0.3797, + "step": 43384 + }, + { + "epoch": 0.9549856922738279, + "grad_norm": 1.3547678588605576, + "learning_rate": 5.304323833786207e-08, + "loss": 0.3136, + "step": 43385 + }, + { + "epoch": 0.9550077041602465, + "grad_norm": 4.585954520336805, + "learning_rate": 5.299146679913103e-08, + "loss": 0.2073, + "step": 43386 + }, + { + "epoch": 0.9550297160466652, + "grad_norm": 6.160757483226394, + "learning_rate": 5.293972040330331e-08, + "loss": 0.2421, + "step": 43387 + }, + { + "epoch": 0.9550517279330839, + "grad_norm": 2.030255095657418, + "learning_rate": 5.288799915064258e-08, + "loss": 0.3761, + "step": 43388 + }, + { + "epoch": 0.9550737398195025, + "grad_norm": 1.835316540473039, + "learning_rate": 5.2836303041411406e-08, + "loss": 0.2585, + "step": 43389 + }, + { + "epoch": 0.9550957517059212, + "grad_norm": 2.05910481608612, + "learning_rate": 5.2784632075873474e-08, + "loss": 0.2709, + "step": 43390 + }, + { + "epoch": 0.9551177635923399, + "grad_norm": 1.9631798535482858, + "learning_rate": 5.273298625428969e-08, + "loss": 0.3021, + "step": 43391 + }, + { + "epoch": 0.9551397754787585, + "grad_norm": 1.4194779679407767, + "learning_rate": 5.268136557692316e-08, + "loss": 0.2751, + "step": 43392 + }, + { + "epoch": 0.9551617873651772, + "grad_norm": 0.609791023334639, + "learning_rate": 5.262977004403702e-08, + "loss": 0.4304, + "step": 43393 + }, + { + "epoch": 0.9551837992515959, + "grad_norm": 2.2329316451631205, + "learning_rate": 5.257819965589328e-08, + "loss": 0.303, + "step": 43394 + }, + { + "epoch": 0.9552058111380145, + "grad_norm": 1.851425962528184, + "learning_rate": 5.25266544127534e-08, + "loss": 0.3097, + "step": 43395 + }, + { + "epoch": 0.9552278230244332, + "grad_norm": 1.8200753394156393, + "learning_rate": 5.2475134314879384e-08, + "loss": 0.2338, + "step": 43396 + }, + { + "epoch": 0.9552498349108519, + "grad_norm": 2.2777624429661802, + "learning_rate": 5.242363936253381e-08, + "loss": 0.3522, + "step": 43397 + }, + { + "epoch": 0.9552718467972705, + "grad_norm": 2.811408243419961, + "learning_rate": 5.237216955597757e-08, + "loss": 0.4081, + "step": 43398 + }, + { + "epoch": 0.9552938586836892, + "grad_norm": 1.684577319322928, + "learning_rate": 5.232072489547324e-08, + "loss": 0.3796, + "step": 43399 + }, + { + "epoch": 0.9553158705701078, + "grad_norm": 2.326384813334469, + "learning_rate": 5.226930538128172e-08, + "loss": 0.2912, + "step": 43400 + }, + { + "epoch": 0.9553378824565265, + "grad_norm": 1.4939073069624678, + "learning_rate": 5.221791101366391e-08, + "loss": 0.2395, + "step": 43401 + }, + { + "epoch": 0.9553598943429452, + "grad_norm": 1.9860015341862747, + "learning_rate": 5.216654179288127e-08, + "loss": 0.2411, + "step": 43402 + }, + { + "epoch": 0.9553819062293638, + "grad_norm": 4.824580411562922, + "learning_rate": 5.2115197719195824e-08, + "loss": 0.2956, + "step": 43403 + }, + { + "epoch": 0.9554039181157825, + "grad_norm": 1.5249508034709296, + "learning_rate": 5.206387879286734e-08, + "loss": 0.3183, + "step": 43404 + }, + { + "epoch": 0.9554259300022012, + "grad_norm": 1.5766349267263668, + "learning_rate": 5.2012585014156734e-08, + "loss": 0.2989, + "step": 43405 + }, + { + "epoch": 0.9554479418886198, + "grad_norm": 0.600370768982675, + "learning_rate": 5.196131638332547e-08, + "loss": 0.4313, + "step": 43406 + }, + { + "epoch": 0.9554699537750385, + "grad_norm": 1.4161880378534233, + "learning_rate": 5.1910072900633876e-08, + "loss": 0.2391, + "step": 43407 + }, + { + "epoch": 0.9554919656614572, + "grad_norm": 1.8792929184415463, + "learning_rate": 5.185885456634177e-08, + "loss": 0.3171, + "step": 43408 + }, + { + "epoch": 0.9555139775478758, + "grad_norm": 1.7173158187690654, + "learning_rate": 5.180766138071003e-08, + "loss": 0.2801, + "step": 43409 + }, + { + "epoch": 0.9555359894342945, + "grad_norm": 0.6069048972770827, + "learning_rate": 5.175649334399846e-08, + "loss": 0.4563, + "step": 43410 + }, + { + "epoch": 0.9555580013207132, + "grad_norm": 2.6227791057420764, + "learning_rate": 5.170535045646741e-08, + "loss": 0.3387, + "step": 43411 + }, + { + "epoch": 0.9555800132071318, + "grad_norm": 2.0507773227830963, + "learning_rate": 5.165423271837722e-08, + "loss": 0.3289, + "step": 43412 + }, + { + "epoch": 0.9556020250935505, + "grad_norm": 2.101646147344216, + "learning_rate": 5.1603140129987126e-08, + "loss": 0.2904, + "step": 43413 + }, + { + "epoch": 0.9556240369799692, + "grad_norm": 1.3873257562280439, + "learning_rate": 5.155207269155638e-08, + "loss": 0.2758, + "step": 43414 + }, + { + "epoch": 0.9556460488663878, + "grad_norm": 1.9804346573484726, + "learning_rate": 5.150103040334586e-08, + "loss": 0.3889, + "step": 43415 + }, + { + "epoch": 0.9556680607528065, + "grad_norm": 2.239026135680521, + "learning_rate": 5.145001326561372e-08, + "loss": 0.2345, + "step": 43416 + }, + { + "epoch": 0.9556900726392252, + "grad_norm": 0.6028807254186939, + "learning_rate": 5.1399021278619735e-08, + "loss": 0.4362, + "step": 43417 + }, + { + "epoch": 0.9557120845256438, + "grad_norm": 2.176521502902814, + "learning_rate": 5.134805444262314e-08, + "loss": 0.3339, + "step": 43418 + }, + { + "epoch": 0.9557340964120625, + "grad_norm": 1.8650146904719926, + "learning_rate": 5.1297112757882626e-08, + "loss": 0.3369, + "step": 43419 + }, + { + "epoch": 0.9557561082984812, + "grad_norm": 2.141188072323775, + "learning_rate": 5.1246196224657984e-08, + "loss": 0.3299, + "step": 43420 + }, + { + "epoch": 0.9557781201848998, + "grad_norm": 1.876151452905849, + "learning_rate": 5.1195304843207337e-08, + "loss": 0.2899, + "step": 43421 + }, + { + "epoch": 0.9558001320713185, + "grad_norm": 0.619847252952471, + "learning_rate": 5.114443861378882e-08, + "loss": 0.4519, + "step": 43422 + }, + { + "epoch": 0.9558221439577371, + "grad_norm": 1.3973647911590115, + "learning_rate": 5.1093597536661654e-08, + "loss": 0.2972, + "step": 43423 + }, + { + "epoch": 0.9558441558441558, + "grad_norm": 1.8700983902240402, + "learning_rate": 5.104278161208398e-08, + "loss": 0.2353, + "step": 43424 + }, + { + "epoch": 0.9558661677305745, + "grad_norm": 0.5999986042504662, + "learning_rate": 5.099199084031503e-08, + "loss": 0.4401, + "step": 43425 + }, + { + "epoch": 0.9558881796169931, + "grad_norm": 2.818023181619849, + "learning_rate": 5.094122522161127e-08, + "loss": 0.2316, + "step": 43426 + }, + { + "epoch": 0.9559101915034118, + "grad_norm": 4.4356331640909055, + "learning_rate": 5.0890484756231925e-08, + "loss": 0.2462, + "step": 43427 + }, + { + "epoch": 0.9559322033898305, + "grad_norm": 1.8425693313723817, + "learning_rate": 5.083976944443403e-08, + "loss": 0.2957, + "step": 43428 + }, + { + "epoch": 0.9559542152762491, + "grad_norm": 1.6939104729902512, + "learning_rate": 5.078907928647625e-08, + "loss": 0.2735, + "step": 43429 + }, + { + "epoch": 0.9559762271626678, + "grad_norm": 2.1073944205357704, + "learning_rate": 5.0738414282615614e-08, + "loss": 0.2535, + "step": 43430 + }, + { + "epoch": 0.9559982390490865, + "grad_norm": 0.6150066577954227, + "learning_rate": 5.068777443311024e-08, + "loss": 0.4194, + "step": 43431 + }, + { + "epoch": 0.9560202509355051, + "grad_norm": 1.2861342455813782, + "learning_rate": 5.063715973821659e-08, + "loss": 0.2675, + "step": 43432 + }, + { + "epoch": 0.9560422628219238, + "grad_norm": 1.9063530346914237, + "learning_rate": 5.058657019819224e-08, + "loss": 0.2838, + "step": 43433 + }, + { + "epoch": 0.9560642747083425, + "grad_norm": 2.4573145434694275, + "learning_rate": 5.053600581329421e-08, + "loss": 0.3411, + "step": 43434 + }, + { + "epoch": 0.9560862865947611, + "grad_norm": 1.9418784960327564, + "learning_rate": 5.048546658378062e-08, + "loss": 0.3078, + "step": 43435 + }, + { + "epoch": 0.9561082984811798, + "grad_norm": 1.8839117282970883, + "learning_rate": 5.043495250990682e-08, + "loss": 0.2744, + "step": 43436 + }, + { + "epoch": 0.9561303103675985, + "grad_norm": 5.136529459381626, + "learning_rate": 5.038446359193039e-08, + "loss": 0.3124, + "step": 43437 + }, + { + "epoch": 0.9561523222540171, + "grad_norm": 1.80313881796498, + "learning_rate": 5.0333999830107226e-08, + "loss": 0.3154, + "step": 43438 + }, + { + "epoch": 0.9561743341404358, + "grad_norm": 0.6174908810663419, + "learning_rate": 5.028356122469491e-08, + "loss": 0.4601, + "step": 43439 + }, + { + "epoch": 0.9561963460268545, + "grad_norm": 1.749130900551528, + "learning_rate": 5.0233147775948785e-08, + "loss": 0.3242, + "step": 43440 + }, + { + "epoch": 0.9562183579132731, + "grad_norm": 1.4657466050924843, + "learning_rate": 5.0182759484125876e-08, + "loss": 0.2796, + "step": 43441 + }, + { + "epoch": 0.9562403697996918, + "grad_norm": 2.8635442931368496, + "learning_rate": 5.013239634948153e-08, + "loss": 0.269, + "step": 43442 + }, + { + "epoch": 0.9562623816861106, + "grad_norm": 2.2675053994191883, + "learning_rate": 5.00820583722722e-08, + "loss": 0.2952, + "step": 43443 + }, + { + "epoch": 0.9562843935725291, + "grad_norm": 1.8777504672381913, + "learning_rate": 5.003174555275325e-08, + "loss": 0.3229, + "step": 43444 + }, + { + "epoch": 0.9563064054589479, + "grad_norm": 1.6443416389926173, + "learning_rate": 4.998145789118114e-08, + "loss": 0.2508, + "step": 43445 + }, + { + "epoch": 0.9563284173453664, + "grad_norm": 0.6173209055254991, + "learning_rate": 4.993119538781066e-08, + "loss": 0.4475, + "step": 43446 + }, + { + "epoch": 0.9563504292317851, + "grad_norm": 4.612046874551745, + "learning_rate": 4.988095804289772e-08, + "loss": 0.2722, + "step": 43447 + }, + { + "epoch": 0.9563724411182039, + "grad_norm": 2.4398132501517065, + "learning_rate": 4.983074585669823e-08, + "loss": 0.2858, + "step": 43448 + }, + { + "epoch": 0.9563944530046224, + "grad_norm": 1.7451981922229927, + "learning_rate": 4.9780558829465867e-08, + "loss": 0.2817, + "step": 43449 + }, + { + "epoch": 0.9564164648910412, + "grad_norm": 2.6873875138628476, + "learning_rate": 4.973039696145654e-08, + "loss": 0.2582, + "step": 43450 + }, + { + "epoch": 0.9564384767774599, + "grad_norm": 1.9684954690347336, + "learning_rate": 4.96802602529256e-08, + "loss": 0.3447, + "step": 43451 + }, + { + "epoch": 0.9564604886638784, + "grad_norm": 1.5164533439429488, + "learning_rate": 4.96301487041273e-08, + "loss": 0.2489, + "step": 43452 + }, + { + "epoch": 0.9564825005502972, + "grad_norm": 7.391847287765178, + "learning_rate": 4.958006231531698e-08, + "loss": 0.254, + "step": 43453 + }, + { + "epoch": 0.9565045124367159, + "grad_norm": 0.6229341190014728, + "learning_rate": 4.953000108674833e-08, + "loss": 0.4523, + "step": 43454 + }, + { + "epoch": 0.9565265243231345, + "grad_norm": 1.763788822438048, + "learning_rate": 4.947996501867613e-08, + "loss": 0.3318, + "step": 43455 + }, + { + "epoch": 0.9565485362095532, + "grad_norm": 0.6188556142529196, + "learning_rate": 4.9429954111355205e-08, + "loss": 0.4392, + "step": 43456 + }, + { + "epoch": 0.9565705480959719, + "grad_norm": 2.193298682241491, + "learning_rate": 4.937996836503867e-08, + "loss": 0.3343, + "step": 43457 + }, + { + "epoch": 0.9565925599823905, + "grad_norm": 1.6336967771908686, + "learning_rate": 4.9330007779981314e-08, + "loss": 0.2395, + "step": 43458 + }, + { + "epoch": 0.9566145718688092, + "grad_norm": 0.6202624809024491, + "learning_rate": 4.928007235643739e-08, + "loss": 0.4754, + "step": 43459 + }, + { + "epoch": 0.9566365837552279, + "grad_norm": 2.192111770534057, + "learning_rate": 4.9230162094659476e-08, + "loss": 0.3083, + "step": 43460 + }, + { + "epoch": 0.9566585956416465, + "grad_norm": 1.8417565436772623, + "learning_rate": 4.918027699490236e-08, + "loss": 0.2987, + "step": 43461 + }, + { + "epoch": 0.9566806075280652, + "grad_norm": 1.9201283223635763, + "learning_rate": 4.913041705741972e-08, + "loss": 0.2416, + "step": 43462 + }, + { + "epoch": 0.9567026194144839, + "grad_norm": 1.682898872179413, + "learning_rate": 4.90805822824636e-08, + "loss": 0.2756, + "step": 43463 + }, + { + "epoch": 0.9567246313009025, + "grad_norm": 2.247478454321449, + "learning_rate": 4.903077267028822e-08, + "loss": 0.2365, + "step": 43464 + }, + { + "epoch": 0.9567466431873212, + "grad_norm": 1.6485998353128886, + "learning_rate": 4.898098822114672e-08, + "loss": 0.2621, + "step": 43465 + }, + { + "epoch": 0.9567686550737399, + "grad_norm": 1.4044907960326716, + "learning_rate": 4.8931228935292784e-08, + "loss": 0.2836, + "step": 43466 + }, + { + "epoch": 0.9567906669601585, + "grad_norm": 1.4149227148287797, + "learning_rate": 4.888149481297788e-08, + "loss": 0.239, + "step": 43467 + }, + { + "epoch": 0.9568126788465772, + "grad_norm": 2.4795090718985384, + "learning_rate": 4.883178585445569e-08, + "loss": 0.3466, + "step": 43468 + }, + { + "epoch": 0.9568346907329958, + "grad_norm": 1.7498888441029223, + "learning_rate": 4.878210205997824e-08, + "loss": 0.2981, + "step": 43469 + }, + { + "epoch": 0.9568567026194145, + "grad_norm": 1.5609120744997838, + "learning_rate": 4.873244342979866e-08, + "loss": 0.2963, + "step": 43470 + }, + { + "epoch": 0.9568787145058332, + "grad_norm": 1.8817403940353437, + "learning_rate": 4.868280996416952e-08, + "loss": 0.2692, + "step": 43471 + }, + { + "epoch": 0.9569007263922518, + "grad_norm": 3.9126492494233296, + "learning_rate": 4.8633201663342845e-08, + "loss": 0.2964, + "step": 43472 + }, + { + "epoch": 0.9569227382786705, + "grad_norm": 1.4892558796146225, + "learning_rate": 4.8583618527570095e-08, + "loss": 0.1842, + "step": 43473 + }, + { + "epoch": 0.9569447501650892, + "grad_norm": 0.6032207899972449, + "learning_rate": 4.8534060557103855e-08, + "loss": 0.4483, + "step": 43474 + }, + { + "epoch": 0.9569667620515078, + "grad_norm": 1.5688526125597868, + "learning_rate": 4.848452775219559e-08, + "loss": 0.2441, + "step": 43475 + }, + { + "epoch": 0.9569887739379265, + "grad_norm": 0.632640360554775, + "learning_rate": 4.843502011309786e-08, + "loss": 0.4531, + "step": 43476 + }, + { + "epoch": 0.9570107858243452, + "grad_norm": 1.8756382656692872, + "learning_rate": 4.838553764006215e-08, + "loss": 0.2621, + "step": 43477 + }, + { + "epoch": 0.9570327977107638, + "grad_norm": 1.4949583353257654, + "learning_rate": 4.833608033333881e-08, + "loss": 0.309, + "step": 43478 + }, + { + "epoch": 0.9570548095971825, + "grad_norm": 2.277468034995903, + "learning_rate": 4.8286648193180966e-08, + "loss": 0.3377, + "step": 43479 + }, + { + "epoch": 0.9570768214836012, + "grad_norm": 1.62856734053128, + "learning_rate": 4.8237241219837864e-08, + "loss": 0.2717, + "step": 43480 + }, + { + "epoch": 0.9570988333700198, + "grad_norm": 1.537159711722481, + "learning_rate": 4.8187859413562075e-08, + "loss": 0.1825, + "step": 43481 + }, + { + "epoch": 0.9571208452564385, + "grad_norm": 1.6561819869588819, + "learning_rate": 4.813850277460452e-08, + "loss": 0.3697, + "step": 43482 + }, + { + "epoch": 0.9571428571428572, + "grad_norm": 2.0336570967537733, + "learning_rate": 4.8089171303215e-08, + "loss": 0.245, + "step": 43483 + }, + { + "epoch": 0.9571648690292758, + "grad_norm": 1.8189296676247626, + "learning_rate": 4.803986499964552e-08, + "loss": 0.3232, + "step": 43484 + }, + { + "epoch": 0.9571868809156945, + "grad_norm": 1.8426237640020295, + "learning_rate": 4.7990583864145903e-08, + "loss": 0.2943, + "step": 43485 + }, + { + "epoch": 0.9572088928021132, + "grad_norm": 1.6945427191634765, + "learning_rate": 4.794132789696648e-08, + "loss": 0.3058, + "step": 43486 + }, + { + "epoch": 0.9572309046885318, + "grad_norm": 0.6051896360026359, + "learning_rate": 4.7892097098358734e-08, + "loss": 0.4288, + "step": 43487 + }, + { + "epoch": 0.9572529165749505, + "grad_norm": 1.843128713771741, + "learning_rate": 4.784289146857135e-08, + "loss": 0.341, + "step": 43488 + }, + { + "epoch": 0.9572749284613692, + "grad_norm": 1.5819457326035598, + "learning_rate": 4.779371100785524e-08, + "loss": 0.2335, + "step": 43489 + }, + { + "epoch": 0.9572969403477878, + "grad_norm": 0.6242176195999719, + "learning_rate": 4.7744555716460193e-08, + "loss": 0.4349, + "step": 43490 + }, + { + "epoch": 0.9573189522342065, + "grad_norm": 1.3713027078130913, + "learning_rate": 4.7695425594636026e-08, + "loss": 0.2442, + "step": 43491 + }, + { + "epoch": 0.9573409641206252, + "grad_norm": 1.469023679969604, + "learning_rate": 4.7646320642632526e-08, + "loss": 0.2836, + "step": 43492 + }, + { + "epoch": 0.9573629760070438, + "grad_norm": 0.6339322496570142, + "learning_rate": 4.75972408606995e-08, + "loss": 0.4413, + "step": 43493 + }, + { + "epoch": 0.9573849878934625, + "grad_norm": 0.6111008104205183, + "learning_rate": 4.7548186249086194e-08, + "loss": 0.4416, + "step": 43494 + }, + { + "epoch": 0.9574069997798811, + "grad_norm": 0.6036853732438245, + "learning_rate": 4.749915680804129e-08, + "loss": 0.4408, + "step": 43495 + }, + { + "epoch": 0.9574290116662998, + "grad_norm": 2.878166392104915, + "learning_rate": 4.74501525378146e-08, + "loss": 0.201, + "step": 43496 + }, + { + "epoch": 0.9574510235527185, + "grad_norm": 1.4059114339473413, + "learning_rate": 4.740117343865591e-08, + "loss": 0.2981, + "step": 43497 + }, + { + "epoch": 0.9574730354391371, + "grad_norm": 1.291002103961197, + "learning_rate": 4.7352219510812814e-08, + "loss": 0.2378, + "step": 43498 + }, + { + "epoch": 0.9574950473255558, + "grad_norm": 1.776179997549569, + "learning_rate": 4.730329075453455e-08, + "loss": 0.3029, + "step": 43499 + }, + { + "epoch": 0.9575170592119745, + "grad_norm": 2.3719982301201408, + "learning_rate": 4.725438717007036e-08, + "loss": 0.2912, + "step": 43500 + }, + { + "epoch": 0.9575390710983931, + "grad_norm": 1.3134561996603409, + "learning_rate": 4.720550875766783e-08, + "loss": 0.2281, + "step": 43501 + }, + { + "epoch": 0.9575610829848118, + "grad_norm": 1.9523488521343326, + "learning_rate": 4.71566555175762e-08, + "loss": 0.3168, + "step": 43502 + }, + { + "epoch": 0.9575830948712305, + "grad_norm": 1.7267931201548623, + "learning_rate": 4.7107827450044164e-08, + "loss": 0.2833, + "step": 43503 + }, + { + "epoch": 0.9576051067576491, + "grad_norm": 2.762043555881728, + "learning_rate": 4.705902455531819e-08, + "loss": 0.2621, + "step": 43504 + }, + { + "epoch": 0.9576271186440678, + "grad_norm": 1.5303419342244675, + "learning_rate": 4.701024683364752e-08, + "loss": 0.3074, + "step": 43505 + }, + { + "epoch": 0.9576491305304865, + "grad_norm": 1.6783838995558724, + "learning_rate": 4.696149428527974e-08, + "loss": 0.2945, + "step": 43506 + }, + { + "epoch": 0.9576711424169051, + "grad_norm": 1.6288100292954073, + "learning_rate": 4.6912766910463535e-08, + "loss": 0.2931, + "step": 43507 + }, + { + "epoch": 0.9576931543033238, + "grad_norm": 2.3010236236488595, + "learning_rate": 4.686406470944538e-08, + "loss": 0.2604, + "step": 43508 + }, + { + "epoch": 0.9577151661897425, + "grad_norm": 1.817697033093289, + "learning_rate": 4.681538768247285e-08, + "loss": 0.2856, + "step": 43509 + }, + { + "epoch": 0.9577371780761611, + "grad_norm": 0.6100165798425156, + "learning_rate": 4.6766735829794076e-08, + "loss": 0.4326, + "step": 43510 + }, + { + "epoch": 0.9577591899625798, + "grad_norm": 1.5902198159305052, + "learning_rate": 4.671810915165553e-08, + "loss": 0.3553, + "step": 43511 + }, + { + "epoch": 0.9577812018489985, + "grad_norm": 2.033561072549874, + "learning_rate": 4.666950764830536e-08, + "loss": 0.2695, + "step": 43512 + }, + { + "epoch": 0.9578032137354171, + "grad_norm": 2.1775139629843387, + "learning_rate": 4.6620931319990015e-08, + "loss": 0.323, + "step": 43513 + }, + { + "epoch": 0.9578252256218358, + "grad_norm": 1.5404070765388946, + "learning_rate": 4.6572380166955975e-08, + "loss": 0.2817, + "step": 43514 + }, + { + "epoch": 0.9578472375082545, + "grad_norm": 1.5057067741956494, + "learning_rate": 4.652385418945027e-08, + "loss": 0.3292, + "step": 43515 + }, + { + "epoch": 0.9578692493946731, + "grad_norm": 1.8779595203254178, + "learning_rate": 4.647535338771936e-08, + "loss": 0.4204, + "step": 43516 + }, + { + "epoch": 0.9578912612810918, + "grad_norm": 1.9139265037110196, + "learning_rate": 4.6426877762010824e-08, + "loss": 0.3641, + "step": 43517 + }, + { + "epoch": 0.9579132731675104, + "grad_norm": 0.5750121971008729, + "learning_rate": 4.6378427312570583e-08, + "loss": 0.4396, + "step": 43518 + }, + { + "epoch": 0.9579352850539291, + "grad_norm": 1.8466232757970606, + "learning_rate": 4.633000203964344e-08, + "loss": 0.2577, + "step": 43519 + }, + { + "epoch": 0.9579572969403478, + "grad_norm": 1.432744023624986, + "learning_rate": 4.628160194347753e-08, + "loss": 0.2966, + "step": 43520 + }, + { + "epoch": 0.9579793088267664, + "grad_norm": 1.341709449835605, + "learning_rate": 4.6233227024317096e-08, + "loss": 0.2979, + "step": 43521 + }, + { + "epoch": 0.9580013207131851, + "grad_norm": 2.3937868040319197, + "learning_rate": 4.6184877282409166e-08, + "loss": 0.2191, + "step": 43522 + }, + { + "epoch": 0.9580233325996038, + "grad_norm": 3.40039597527205, + "learning_rate": 4.613655271799966e-08, + "loss": 0.2763, + "step": 43523 + }, + { + "epoch": 0.9580453444860224, + "grad_norm": 1.8097456166792452, + "learning_rate": 4.6088253331333375e-08, + "loss": 0.3696, + "step": 43524 + }, + { + "epoch": 0.9580673563724411, + "grad_norm": 1.6595481087770259, + "learning_rate": 4.603997912265623e-08, + "loss": 0.2545, + "step": 43525 + }, + { + "epoch": 0.9580893682588598, + "grad_norm": 0.6079996162797515, + "learning_rate": 4.599173009221303e-08, + "loss": 0.4212, + "step": 43526 + }, + { + "epoch": 0.9581113801452784, + "grad_norm": 3.543391082732633, + "learning_rate": 4.594350624024912e-08, + "loss": 0.2143, + "step": 43527 + }, + { + "epoch": 0.9581333920316971, + "grad_norm": 0.5914946150966248, + "learning_rate": 4.589530756701099e-08, + "loss": 0.4224, + "step": 43528 + }, + { + "epoch": 0.9581554039181158, + "grad_norm": 1.6568014551707977, + "learning_rate": 4.584713407274177e-08, + "loss": 0.3231, + "step": 43529 + }, + { + "epoch": 0.9581774158045344, + "grad_norm": 1.5519549824054364, + "learning_rate": 4.579898575768737e-08, + "loss": 0.2387, + "step": 43530 + }, + { + "epoch": 0.9581994276909531, + "grad_norm": 1.710477331322143, + "learning_rate": 4.5750862622091496e-08, + "loss": 0.2317, + "step": 43531 + }, + { + "epoch": 0.9582214395773718, + "grad_norm": 0.5886830246960095, + "learning_rate": 4.570276466619949e-08, + "loss": 0.4099, + "step": 43532 + }, + { + "epoch": 0.9582434514637904, + "grad_norm": 1.4962877291035437, + "learning_rate": 4.565469189025673e-08, + "loss": 0.3227, + "step": 43533 + }, + { + "epoch": 0.9582654633502091, + "grad_norm": 1.5427610215290928, + "learning_rate": 4.5606644294505235e-08, + "loss": 0.2985, + "step": 43534 + }, + { + "epoch": 0.9582874752366278, + "grad_norm": 4.505892577724937, + "learning_rate": 4.5558621879191466e-08, + "loss": 0.2608, + "step": 43535 + }, + { + "epoch": 0.9583094871230464, + "grad_norm": 1.5759003587919982, + "learning_rate": 4.551062464455802e-08, + "loss": 0.2746, + "step": 43536 + }, + { + "epoch": 0.9583314990094651, + "grad_norm": 1.659428971365666, + "learning_rate": 4.5462652590849146e-08, + "loss": 0.3045, + "step": 43537 + }, + { + "epoch": 0.9583535108958838, + "grad_norm": 13.455406064493026, + "learning_rate": 4.541470571830964e-08, + "loss": 0.3088, + "step": 43538 + }, + { + "epoch": 0.9583755227823024, + "grad_norm": 0.627248090088793, + "learning_rate": 4.5366784027182084e-08, + "loss": 0.4704, + "step": 43539 + }, + { + "epoch": 0.9583975346687211, + "grad_norm": 0.6067143051224856, + "learning_rate": 4.531888751771019e-08, + "loss": 0.4231, + "step": 43540 + }, + { + "epoch": 0.9584195465551397, + "grad_norm": 0.6053597238089466, + "learning_rate": 4.527101619013763e-08, + "loss": 0.442, + "step": 43541 + }, + { + "epoch": 0.9584415584415584, + "grad_norm": 1.659871030002172, + "learning_rate": 4.522317004470755e-08, + "loss": 0.335, + "step": 43542 + }, + { + "epoch": 0.9584635703279771, + "grad_norm": 2.2296375031183864, + "learning_rate": 4.51753490816631e-08, + "loss": 0.2832, + "step": 43543 + }, + { + "epoch": 0.9584855822143957, + "grad_norm": 1.4030984786645937, + "learning_rate": 4.512755330124796e-08, + "loss": 0.3254, + "step": 43544 + }, + { + "epoch": 0.9585075941008144, + "grad_norm": 2.979872818374204, + "learning_rate": 4.507978270370417e-08, + "loss": 0.2525, + "step": 43545 + }, + { + "epoch": 0.9585296059872331, + "grad_norm": 1.8426619831987303, + "learning_rate": 4.50320372892743e-08, + "loss": 0.245, + "step": 43546 + }, + { + "epoch": 0.9585516178736517, + "grad_norm": 1.7730549682778873, + "learning_rate": 4.498431705820261e-08, + "loss": 0.328, + "step": 43547 + }, + { + "epoch": 0.9585736297600704, + "grad_norm": 0.6621712903147449, + "learning_rate": 4.4936622010730014e-08, + "loss": 0.4145, + "step": 43548 + }, + { + "epoch": 0.9585956416464891, + "grad_norm": 2.869868171669714, + "learning_rate": 4.4888952147100205e-08, + "loss": 0.2855, + "step": 43549 + }, + { + "epoch": 0.9586176535329077, + "grad_norm": 0.6137640753556948, + "learning_rate": 4.4841307467554107e-08, + "loss": 0.4621, + "step": 43550 + }, + { + "epoch": 0.9586396654193264, + "grad_norm": 18.54320038136374, + "learning_rate": 4.4793687972334856e-08, + "loss": 0.2702, + "step": 43551 + }, + { + "epoch": 0.9586616773057451, + "grad_norm": 5.531454538963307, + "learning_rate": 4.474609366168337e-08, + "loss": 0.2537, + "step": 43552 + }, + { + "epoch": 0.9586836891921637, + "grad_norm": 1.698490911833471, + "learning_rate": 4.4698524535843335e-08, + "loss": 0.2884, + "step": 43553 + }, + { + "epoch": 0.9587057010785824, + "grad_norm": 1.7076628079809593, + "learning_rate": 4.4650980595055125e-08, + "loss": 0.2842, + "step": 43554 + }, + { + "epoch": 0.9587277129650011, + "grad_norm": 1.5278053536633416, + "learning_rate": 4.460346183956077e-08, + "loss": 0.3135, + "step": 43555 + }, + { + "epoch": 0.9587497248514197, + "grad_norm": 2.1528342954325272, + "learning_rate": 4.455596826960229e-08, + "loss": 0.249, + "step": 43556 + }, + { + "epoch": 0.9587717367378384, + "grad_norm": 1.8757480025489082, + "learning_rate": 4.450849988542005e-08, + "loss": 0.2609, + "step": 43557 + }, + { + "epoch": 0.9587937486242571, + "grad_norm": 2.0218119825873937, + "learning_rate": 4.4461056687255536e-08, + "loss": 0.2745, + "step": 43558 + }, + { + "epoch": 0.9588157605106757, + "grad_norm": 1.8547972361309748, + "learning_rate": 4.441363867535076e-08, + "loss": 0.2564, + "step": 43559 + }, + { + "epoch": 0.9588377723970944, + "grad_norm": 1.4840405207773888, + "learning_rate": 4.436624584994609e-08, + "loss": 0.3316, + "step": 43560 + }, + { + "epoch": 0.9588597842835132, + "grad_norm": 1.6035061120878817, + "learning_rate": 4.4318878211282445e-08, + "loss": 0.2301, + "step": 43561 + }, + { + "epoch": 0.9588817961699317, + "grad_norm": 3.3431296176465883, + "learning_rate": 4.427153575960019e-08, + "loss": 0.2561, + "step": 43562 + }, + { + "epoch": 0.9589038080563504, + "grad_norm": 1.2338958442456651, + "learning_rate": 4.4224218495140246e-08, + "loss": 0.2445, + "step": 43563 + }, + { + "epoch": 0.958925819942769, + "grad_norm": 1.7481068057498477, + "learning_rate": 4.417692641814353e-08, + "loss": 0.2961, + "step": 43564 + }, + { + "epoch": 0.9589478318291877, + "grad_norm": 3.2288439489398506, + "learning_rate": 4.4129659528849844e-08, + "loss": 0.384, + "step": 43565 + }, + { + "epoch": 0.9589698437156065, + "grad_norm": 1.6760517712827125, + "learning_rate": 4.4082417827499555e-08, + "loss": 0.3272, + "step": 43566 + }, + { + "epoch": 0.958991855602025, + "grad_norm": 0.6176800029531263, + "learning_rate": 4.403520131433303e-08, + "loss": 0.4387, + "step": 43567 + }, + { + "epoch": 0.9590138674884437, + "grad_norm": 1.522159267364957, + "learning_rate": 4.398800998959008e-08, + "loss": 0.2466, + "step": 43568 + }, + { + "epoch": 0.9590358793748625, + "grad_norm": 1.6624547578797206, + "learning_rate": 4.3940843853510495e-08, + "loss": 0.2424, + "step": 43569 + }, + { + "epoch": 0.959057891261281, + "grad_norm": 1.786393795538872, + "learning_rate": 4.3893702906333545e-08, + "loss": 0.2942, + "step": 43570 + }, + { + "epoch": 0.9590799031476998, + "grad_norm": 1.3606314328426599, + "learning_rate": 4.3846587148299593e-08, + "loss": 0.2436, + "step": 43571 + }, + { + "epoch": 0.9591019150341185, + "grad_norm": 1.684854978041771, + "learning_rate": 4.379949657964788e-08, + "loss": 0.2031, + "step": 43572 + }, + { + "epoch": 0.959123926920537, + "grad_norm": 1.3620816308414996, + "learning_rate": 4.375243120061767e-08, + "loss": 0.2711, + "step": 43573 + }, + { + "epoch": 0.9591459388069558, + "grad_norm": 0.6355223679657911, + "learning_rate": 4.3705391011448215e-08, + "loss": 0.4431, + "step": 43574 + }, + { + "epoch": 0.9591679506933745, + "grad_norm": 0.5785459241588351, + "learning_rate": 4.365837601237821e-08, + "loss": 0.4214, + "step": 43575 + }, + { + "epoch": 0.959189962579793, + "grad_norm": 0.6213604876621294, + "learning_rate": 4.3611386203646914e-08, + "loss": 0.4573, + "step": 43576 + }, + { + "epoch": 0.9592119744662118, + "grad_norm": 2.7659941407929325, + "learning_rate": 4.356442158549357e-08, + "loss": 0.2579, + "step": 43577 + }, + { + "epoch": 0.9592339863526305, + "grad_norm": 1.8603935692764597, + "learning_rate": 4.351748215815577e-08, + "loss": 0.3595, + "step": 43578 + }, + { + "epoch": 0.9592559982390491, + "grad_norm": 1.671690053384858, + "learning_rate": 4.347056792187388e-08, + "loss": 0.2757, + "step": 43579 + }, + { + "epoch": 0.9592780101254678, + "grad_norm": 0.6286203622507046, + "learning_rate": 4.342367887688437e-08, + "loss": 0.4423, + "step": 43580 + }, + { + "epoch": 0.9593000220118865, + "grad_norm": 1.7849449665614499, + "learning_rate": 4.33768150234265e-08, + "loss": 0.2194, + "step": 43581 + }, + { + "epoch": 0.9593220338983051, + "grad_norm": 1.5134709869972405, + "learning_rate": 4.3329976361738414e-08, + "loss": 0.2904, + "step": 43582 + }, + { + "epoch": 0.9593440457847238, + "grad_norm": 1.7876947801785126, + "learning_rate": 4.328316289205825e-08, + "loss": 0.3616, + "step": 43583 + }, + { + "epoch": 0.9593660576711425, + "grad_norm": 1.648644814914922, + "learning_rate": 4.323637461462416e-08, + "loss": 0.2693, + "step": 43584 + }, + { + "epoch": 0.9593880695575611, + "grad_norm": 2.465182332899641, + "learning_rate": 4.318961152967316e-08, + "loss": 0.2571, + "step": 43585 + }, + { + "epoch": 0.9594100814439798, + "grad_norm": 2.0798384042294242, + "learning_rate": 4.3142873637443406e-08, + "loss": 0.3801, + "step": 43586 + }, + { + "epoch": 0.9594320933303984, + "grad_norm": 1.4802227392873255, + "learning_rate": 4.3096160938172484e-08, + "loss": 0.2472, + "step": 43587 + }, + { + "epoch": 0.9594541052168171, + "grad_norm": 0.6243465251826094, + "learning_rate": 4.304947343209742e-08, + "loss": 0.4425, + "step": 43588 + }, + { + "epoch": 0.9594761171032358, + "grad_norm": 2.038353302541828, + "learning_rate": 4.300281111945637e-08, + "loss": 0.3101, + "step": 43589 + }, + { + "epoch": 0.9594981289896544, + "grad_norm": 1.945439608381696, + "learning_rate": 4.2956174000485793e-08, + "loss": 0.2472, + "step": 43590 + }, + { + "epoch": 0.9595201408760731, + "grad_norm": 1.7222189011743363, + "learning_rate": 4.290956207542219e-08, + "loss": 0.3625, + "step": 43591 + }, + { + "epoch": 0.9595421527624918, + "grad_norm": 1.393266087834371, + "learning_rate": 4.286297534450368e-08, + "loss": 0.3154, + "step": 43592 + }, + { + "epoch": 0.9595641646489104, + "grad_norm": 1.3517676539550403, + "learning_rate": 4.281641380796619e-08, + "loss": 0.2096, + "step": 43593 + }, + { + "epoch": 0.9595861765353291, + "grad_norm": 1.7771903681860068, + "learning_rate": 4.2769877466046774e-08, + "loss": 0.2693, + "step": 43594 + }, + { + "epoch": 0.9596081884217478, + "grad_norm": 2.0162811459417775, + "learning_rate": 4.2723366318981883e-08, + "loss": 0.3076, + "step": 43595 + }, + { + "epoch": 0.9596302003081664, + "grad_norm": 0.5904917293920554, + "learning_rate": 4.2676880367008014e-08, + "loss": 0.4206, + "step": 43596 + }, + { + "epoch": 0.9596522121945851, + "grad_norm": 3.130621523215153, + "learning_rate": 4.263041961036107e-08, + "loss": 0.2534, + "step": 43597 + }, + { + "epoch": 0.9596742240810038, + "grad_norm": 2.0393595002097564, + "learning_rate": 4.2583984049276996e-08, + "loss": 0.3151, + "step": 43598 + }, + { + "epoch": 0.9596962359674224, + "grad_norm": 1.5071859966136767, + "learning_rate": 4.25375736839928e-08, + "loss": 0.2491, + "step": 43599 + }, + { + "epoch": 0.9597182478538411, + "grad_norm": 90.95846959726532, + "learning_rate": 4.249118851474332e-08, + "loss": 0.3638, + "step": 43600 + }, + { + "epoch": 0.9597402597402598, + "grad_norm": 1.684676052831448, + "learning_rate": 4.244482854176502e-08, + "loss": 0.3224, + "step": 43601 + }, + { + "epoch": 0.9597622716266784, + "grad_norm": 0.5836183259071589, + "learning_rate": 4.239849376529326e-08, + "loss": 0.4071, + "step": 43602 + }, + { + "epoch": 0.9597842835130971, + "grad_norm": 1.6601494275187694, + "learning_rate": 4.235218418556342e-08, + "loss": 0.3179, + "step": 43603 + }, + { + "epoch": 0.9598062953995158, + "grad_norm": 1.6518995416473423, + "learning_rate": 4.230589980281086e-08, + "loss": 0.3119, + "step": 43604 + }, + { + "epoch": 0.9598283072859344, + "grad_norm": 6.22495532655572, + "learning_rate": 4.2259640617270946e-08, + "loss": 0.276, + "step": 43605 + }, + { + "epoch": 0.9598503191723531, + "grad_norm": 1.9342106818499851, + "learning_rate": 4.22134066291785e-08, + "loss": 0.2856, + "step": 43606 + }, + { + "epoch": 0.9598723310587718, + "grad_norm": 1.9535136786550873, + "learning_rate": 4.2167197838769436e-08, + "loss": 0.2581, + "step": 43607 + }, + { + "epoch": 0.9598943429451904, + "grad_norm": 1.5417974558449352, + "learning_rate": 4.212101424627746e-08, + "loss": 0.2769, + "step": 43608 + }, + { + "epoch": 0.9599163548316091, + "grad_norm": 1.6414998587962135, + "learning_rate": 4.207485585193738e-08, + "loss": 0.2733, + "step": 43609 + }, + { + "epoch": 0.9599383667180277, + "grad_norm": 1.4896810169562684, + "learning_rate": 4.202872265598512e-08, + "loss": 0.2503, + "step": 43610 + }, + { + "epoch": 0.9599603786044464, + "grad_norm": 1.972563624433516, + "learning_rate": 4.198261465865328e-08, + "loss": 0.32, + "step": 43611 + }, + { + "epoch": 0.9599823904908651, + "grad_norm": 2.293644962376409, + "learning_rate": 4.1936531860177767e-08, + "loss": 0.2674, + "step": 43612 + }, + { + "epoch": 0.9600044023772837, + "grad_norm": 0.6066803277724528, + "learning_rate": 4.189047426079229e-08, + "loss": 0.4511, + "step": 43613 + }, + { + "epoch": 0.9600264142637024, + "grad_norm": 1.4405025126980298, + "learning_rate": 4.184444186073e-08, + "loss": 0.2241, + "step": 43614 + }, + { + "epoch": 0.9600484261501211, + "grad_norm": 2.116385107244023, + "learning_rate": 4.1798434660226264e-08, + "loss": 0.3333, + "step": 43615 + }, + { + "epoch": 0.9600704380365397, + "grad_norm": 1.7267946282743944, + "learning_rate": 4.175245265951422e-08, + "loss": 0.2534, + "step": 43616 + }, + { + "epoch": 0.9600924499229584, + "grad_norm": 5.8870433886793725, + "learning_rate": 4.1706495858827575e-08, + "loss": 0.2368, + "step": 43617 + }, + { + "epoch": 0.9601144618093771, + "grad_norm": 2.8000721284370633, + "learning_rate": 4.166056425840004e-08, + "loss": 0.3569, + "step": 43618 + }, + { + "epoch": 0.9601364736957957, + "grad_norm": 1.6303959410346165, + "learning_rate": 4.161465785846475e-08, + "loss": 0.2757, + "step": 43619 + }, + { + "epoch": 0.9601584855822144, + "grad_norm": 1.755438845243288, + "learning_rate": 4.156877665925596e-08, + "loss": 0.2363, + "step": 43620 + }, + { + "epoch": 0.9601804974686331, + "grad_norm": 3.6575671405375734, + "learning_rate": 4.1522920661005713e-08, + "loss": 0.3994, + "step": 43621 + }, + { + "epoch": 0.9602025093550517, + "grad_norm": 1.602247131487576, + "learning_rate": 4.147708986394716e-08, + "loss": 0.2526, + "step": 43622 + }, + { + "epoch": 0.9602245212414704, + "grad_norm": 1.6470492141967974, + "learning_rate": 4.1431284268314e-08, + "loss": 0.278, + "step": 43623 + }, + { + "epoch": 0.9602465331278891, + "grad_norm": 0.5945527911437216, + "learning_rate": 4.138550387433826e-08, + "loss": 0.4185, + "step": 43624 + }, + { + "epoch": 0.9602685450143077, + "grad_norm": 1.940370333018974, + "learning_rate": 4.1339748682253656e-08, + "loss": 0.2593, + "step": 43625 + }, + { + "epoch": 0.9602905569007264, + "grad_norm": 2.205652419948112, + "learning_rate": 4.129401869229166e-08, + "loss": 0.3338, + "step": 43626 + }, + { + "epoch": 0.9603125687871451, + "grad_norm": 1.4401192605295503, + "learning_rate": 4.124831390468542e-08, + "loss": 0.2604, + "step": 43627 + }, + { + "epoch": 0.9603345806735637, + "grad_norm": 1.693804143532018, + "learning_rate": 4.120263431966643e-08, + "loss": 0.2052, + "step": 43628 + }, + { + "epoch": 0.9603565925599824, + "grad_norm": 2.1171581059702453, + "learning_rate": 4.115697993746781e-08, + "loss": 0.2976, + "step": 43629 + }, + { + "epoch": 0.9603786044464011, + "grad_norm": 1.6982479571274858, + "learning_rate": 4.111135075832051e-08, + "loss": 0.2658, + "step": 43630 + }, + { + "epoch": 0.9604006163328197, + "grad_norm": 1.7016702593261246, + "learning_rate": 4.1065746782457674e-08, + "loss": 0.2915, + "step": 43631 + }, + { + "epoch": 0.9604226282192384, + "grad_norm": 1.4112184977473814, + "learning_rate": 4.102016801011022e-08, + "loss": 0.2566, + "step": 43632 + }, + { + "epoch": 0.9604446401056571, + "grad_norm": 0.6041142820650047, + "learning_rate": 4.0974614441510184e-08, + "loss": 0.4266, + "step": 43633 + }, + { + "epoch": 0.9604666519920757, + "grad_norm": 1.7337069299975887, + "learning_rate": 4.0929086076888506e-08, + "loss": 0.2754, + "step": 43634 + }, + { + "epoch": 0.9604886638784944, + "grad_norm": 1.9920611676470856, + "learning_rate": 4.0883582916477205e-08, + "loss": 0.299, + "step": 43635 + }, + { + "epoch": 0.960510675764913, + "grad_norm": 0.6199876280591714, + "learning_rate": 4.083810496050777e-08, + "loss": 0.4337, + "step": 43636 + }, + { + "epoch": 0.9605326876513317, + "grad_norm": 1.8719356016947153, + "learning_rate": 4.079265220921058e-08, + "loss": 0.3341, + "step": 43637 + }, + { + "epoch": 0.9605546995377504, + "grad_norm": 1.6492135830847179, + "learning_rate": 4.074722466281711e-08, + "loss": 0.3433, + "step": 43638 + }, + { + "epoch": 0.960576711424169, + "grad_norm": 5.146090788376735, + "learning_rate": 4.070182232155773e-08, + "loss": 0.2794, + "step": 43639 + }, + { + "epoch": 0.9605987233105877, + "grad_norm": 1.4605116534223226, + "learning_rate": 4.0656445185663914e-08, + "loss": 0.2204, + "step": 43640 + }, + { + "epoch": 0.9606207351970064, + "grad_norm": 1.6351583458904249, + "learning_rate": 4.061109325536605e-08, + "loss": 0.2461, + "step": 43641 + }, + { + "epoch": 0.960642747083425, + "grad_norm": 0.629319327812024, + "learning_rate": 4.05657665308945e-08, + "loss": 0.459, + "step": 43642 + }, + { + "epoch": 0.9606647589698437, + "grad_norm": 1.8261717498116459, + "learning_rate": 4.0520465012479635e-08, + "loss": 0.3815, + "step": 43643 + }, + { + "epoch": 0.9606867708562624, + "grad_norm": 2.388913206210233, + "learning_rate": 4.047518870035183e-08, + "loss": 0.2489, + "step": 43644 + }, + { + "epoch": 0.960708782742681, + "grad_norm": 1.9346973901789732, + "learning_rate": 4.04299375947409e-08, + "loss": 0.3314, + "step": 43645 + }, + { + "epoch": 0.9607307946290997, + "grad_norm": 1.6859969789600509, + "learning_rate": 4.0384711695877765e-08, + "loss": 0.319, + "step": 43646 + }, + { + "epoch": 0.9607528065155184, + "grad_norm": 0.6137615838083722, + "learning_rate": 4.0339511003991135e-08, + "loss": 0.4202, + "step": 43647 + }, + { + "epoch": 0.960774818401937, + "grad_norm": 1.3378418183521086, + "learning_rate": 4.029433551931083e-08, + "loss": 0.2658, + "step": 43648 + }, + { + "epoch": 0.9607968302883557, + "grad_norm": 1.9015941447295595, + "learning_rate": 4.024918524206722e-08, + "loss": 0.3009, + "step": 43649 + }, + { + "epoch": 0.9608188421747744, + "grad_norm": 1.423934726770379, + "learning_rate": 4.0204060172489015e-08, + "loss": 0.2691, + "step": 43650 + }, + { + "epoch": 0.960840854061193, + "grad_norm": 0.5998162600881356, + "learning_rate": 4.0158960310806573e-08, + "loss": 0.4361, + "step": 43651 + }, + { + "epoch": 0.9608628659476117, + "grad_norm": 0.6176165115875227, + "learning_rate": 4.01138856572475e-08, + "loss": 0.4352, + "step": 43652 + }, + { + "epoch": 0.9608848778340304, + "grad_norm": 1.509461059188018, + "learning_rate": 4.0068836212042714e-08, + "loss": 0.278, + "step": 43653 + }, + { + "epoch": 0.960906889720449, + "grad_norm": 1.4869471596593338, + "learning_rate": 4.0023811975419826e-08, + "loss": 0.3264, + "step": 43654 + }, + { + "epoch": 0.9609289016068677, + "grad_norm": 1.7854499419190493, + "learning_rate": 3.997881294760808e-08, + "loss": 0.293, + "step": 43655 + }, + { + "epoch": 0.9609509134932864, + "grad_norm": 2.9021774450303472, + "learning_rate": 3.9933839128836746e-08, + "loss": 0.3407, + "step": 43656 + }, + { + "epoch": 0.960972925379705, + "grad_norm": 1.6501091348612482, + "learning_rate": 3.988889051933342e-08, + "loss": 0.2341, + "step": 43657 + }, + { + "epoch": 0.9609949372661237, + "grad_norm": 1.7507535189292978, + "learning_rate": 3.984396711932736e-08, + "loss": 0.2804, + "step": 43658 + }, + { + "epoch": 0.9610169491525423, + "grad_norm": 0.6240128240131496, + "learning_rate": 3.979906892904617e-08, + "loss": 0.4485, + "step": 43659 + }, + { + "epoch": 0.961038961038961, + "grad_norm": 1.5153918438404899, + "learning_rate": 3.975419594871799e-08, + "loss": 0.2626, + "step": 43660 + }, + { + "epoch": 0.9610609729253797, + "grad_norm": 1.8562211691444663, + "learning_rate": 3.970934817857208e-08, + "loss": 0.2643, + "step": 43661 + }, + { + "epoch": 0.9610829848117983, + "grad_norm": 1.7842189266637465, + "learning_rate": 3.9664525618836047e-08, + "loss": 0.2762, + "step": 43662 + }, + { + "epoch": 0.961104996698217, + "grad_norm": 0.621133447931356, + "learning_rate": 3.961972826973637e-08, + "loss": 0.4567, + "step": 43663 + }, + { + "epoch": 0.9611270085846357, + "grad_norm": 1.8454953527999571, + "learning_rate": 3.957495613150231e-08, + "loss": 0.3381, + "step": 43664 + }, + { + "epoch": 0.9611490204710543, + "grad_norm": 1.4532786582430115, + "learning_rate": 3.9530209204359795e-08, + "loss": 0.2178, + "step": 43665 + }, + { + "epoch": 0.961171032357473, + "grad_norm": 4.869950655428587, + "learning_rate": 3.9485487488537534e-08, + "loss": 0.2056, + "step": 43666 + }, + { + "epoch": 0.9611930442438917, + "grad_norm": 1.6099204580543236, + "learning_rate": 3.944079098426312e-08, + "loss": 0.2582, + "step": 43667 + }, + { + "epoch": 0.9612150561303103, + "grad_norm": 1.8125021519437936, + "learning_rate": 3.939611969176249e-08, + "loss": 0.2563, + "step": 43668 + }, + { + "epoch": 0.961237068016729, + "grad_norm": 2.4437421111523374, + "learning_rate": 3.935147361126379e-08, + "loss": 0.286, + "step": 43669 + }, + { + "epoch": 0.9612590799031477, + "grad_norm": 1.5422112702428605, + "learning_rate": 3.930685274299295e-08, + "loss": 0.4102, + "step": 43670 + }, + { + "epoch": 0.9612810917895663, + "grad_norm": 2.900912427732378, + "learning_rate": 3.926225708717701e-08, + "loss": 0.2999, + "step": 43671 + }, + { + "epoch": 0.961303103675985, + "grad_norm": 1.6833978387567397, + "learning_rate": 3.921768664404302e-08, + "loss": 0.1987, + "step": 43672 + }, + { + "epoch": 0.9613251155624037, + "grad_norm": 0.6151610680588364, + "learning_rate": 3.917314141381745e-08, + "loss": 0.4537, + "step": 43673 + }, + { + "epoch": 0.9613471274488223, + "grad_norm": 3.4043653827476184, + "learning_rate": 3.9128621396726794e-08, + "loss": 0.3248, + "step": 43674 + }, + { + "epoch": 0.961369139335241, + "grad_norm": 1.9419347977588737, + "learning_rate": 3.908412659299643e-08, + "loss": 0.2889, + "step": 43675 + }, + { + "epoch": 0.9613911512216597, + "grad_norm": 1.703535307678525, + "learning_rate": 3.903965700285395e-08, + "loss": 0.354, + "step": 43676 + }, + { + "epoch": 0.9614131631080783, + "grad_norm": 6.38668744871532, + "learning_rate": 3.899521262652417e-08, + "loss": 0.3084, + "step": 43677 + }, + { + "epoch": 0.961435174994497, + "grad_norm": 1.5438858908227182, + "learning_rate": 3.895079346423303e-08, + "loss": 0.22, + "step": 43678 + }, + { + "epoch": 0.9614571868809157, + "grad_norm": 0.6069745414211385, + "learning_rate": 3.890639951620756e-08, + "loss": 0.4496, + "step": 43679 + }, + { + "epoch": 0.9614791987673343, + "grad_norm": 1.8678918984684938, + "learning_rate": 3.886203078267148e-08, + "loss": 0.204, + "step": 43680 + }, + { + "epoch": 0.961501210653753, + "grad_norm": 1.857855620040246, + "learning_rate": 3.881768726385182e-08, + "loss": 0.253, + "step": 43681 + }, + { + "epoch": 0.9615232225401716, + "grad_norm": 1.7108279747177, + "learning_rate": 3.877336895997341e-08, + "loss": 0.2875, + "step": 43682 + }, + { + "epoch": 0.9615452344265903, + "grad_norm": 1.8682623697353624, + "learning_rate": 3.8729075871261626e-08, + "loss": 0.2863, + "step": 43683 + }, + { + "epoch": 0.961567246313009, + "grad_norm": 2.5224505194815485, + "learning_rate": 3.868480799794128e-08, + "loss": 0.2196, + "step": 43684 + }, + { + "epoch": 0.9615892581994276, + "grad_norm": 1.9761494590314506, + "learning_rate": 3.864056534023775e-08, + "loss": 0.3358, + "step": 43685 + }, + { + "epoch": 0.9616112700858463, + "grad_norm": 1.902628326723193, + "learning_rate": 3.85963478983753e-08, + "loss": 0.2679, + "step": 43686 + }, + { + "epoch": 0.961633281972265, + "grad_norm": 2.9211746951593267, + "learning_rate": 3.8552155672579305e-08, + "loss": 0.3272, + "step": 43687 + }, + { + "epoch": 0.9616552938586836, + "grad_norm": 1.9408270438577582, + "learning_rate": 3.850798866307403e-08, + "loss": 0.2839, + "step": 43688 + }, + { + "epoch": 0.9616773057451024, + "grad_norm": 2.3076655468310747, + "learning_rate": 3.846384687008431e-08, + "loss": 0.2323, + "step": 43689 + }, + { + "epoch": 0.9616993176315211, + "grad_norm": 1.8750032123400973, + "learning_rate": 3.841973029383439e-08, + "loss": 0.3182, + "step": 43690 + }, + { + "epoch": 0.9617213295179396, + "grad_norm": 1.768439589579095, + "learning_rate": 3.8375638934547435e-08, + "loss": 0.25, + "step": 43691 + }, + { + "epoch": 0.9617433414043584, + "grad_norm": 2.280866894928773, + "learning_rate": 3.833157279244881e-08, + "loss": 0.3738, + "step": 43692 + }, + { + "epoch": 0.9617653532907771, + "grad_norm": 0.6210844784060654, + "learning_rate": 3.828753186776224e-08, + "loss": 0.4332, + "step": 43693 + }, + { + "epoch": 0.9617873651771957, + "grad_norm": 1.6862435204936095, + "learning_rate": 3.824351616071198e-08, + "loss": 0.2819, + "step": 43694 + }, + { + "epoch": 0.9618093770636144, + "grad_norm": 1.4012486985354475, + "learning_rate": 3.819952567152063e-08, + "loss": 0.2561, + "step": 43695 + }, + { + "epoch": 0.9618313889500331, + "grad_norm": 2.289725755574424, + "learning_rate": 3.815556040041191e-08, + "loss": 0.2823, + "step": 43696 + }, + { + "epoch": 0.9618534008364517, + "grad_norm": 1.801269019927158, + "learning_rate": 3.811162034761007e-08, + "loss": 0.2804, + "step": 43697 + }, + { + "epoch": 0.9618754127228704, + "grad_norm": 1.8054787704656174, + "learning_rate": 3.806770551333827e-08, + "loss": 0.1755, + "step": 43698 + }, + { + "epoch": 0.9618974246092891, + "grad_norm": 2.225486456739443, + "learning_rate": 3.802381589781912e-08, + "loss": 0.3034, + "step": 43699 + }, + { + "epoch": 0.9619194364957077, + "grad_norm": 1.3856379540218347, + "learning_rate": 3.7979951501276316e-08, + "loss": 0.2788, + "step": 43700 + }, + { + "epoch": 0.9619414483821264, + "grad_norm": 0.6349868294238491, + "learning_rate": 3.7936112323932463e-08, + "loss": 0.4385, + "step": 43701 + }, + { + "epoch": 0.9619634602685451, + "grad_norm": 0.6133891064101982, + "learning_rate": 3.789229836601016e-08, + "loss": 0.4556, + "step": 43702 + }, + { + "epoch": 0.9619854721549637, + "grad_norm": 1.8615749953294476, + "learning_rate": 3.7848509627732566e-08, + "loss": 0.2754, + "step": 43703 + }, + { + "epoch": 0.9620074840413824, + "grad_norm": 2.1995834334431756, + "learning_rate": 3.7804746109321723e-08, + "loss": 0.3097, + "step": 43704 + }, + { + "epoch": 0.962029495927801, + "grad_norm": 1.4337599683497841, + "learning_rate": 3.776100781100078e-08, + "loss": 0.2966, + "step": 43705 + }, + { + "epoch": 0.9620515078142197, + "grad_norm": 1.6748563318856724, + "learning_rate": 3.7717294732991793e-08, + "loss": 0.2877, + "step": 43706 + }, + { + "epoch": 0.9620735197006384, + "grad_norm": 1.8131411248510882, + "learning_rate": 3.767360687551625e-08, + "loss": 0.299, + "step": 43707 + }, + { + "epoch": 0.962095531587057, + "grad_norm": 1.9318624316893829, + "learning_rate": 3.7629944238796735e-08, + "loss": 0.2593, + "step": 43708 + }, + { + "epoch": 0.9621175434734757, + "grad_norm": 1.4418805908610233, + "learning_rate": 3.7586306823055305e-08, + "loss": 0.2946, + "step": 43709 + }, + { + "epoch": 0.9621395553598944, + "grad_norm": 1.4459229747876483, + "learning_rate": 3.754269462851345e-08, + "loss": 0.2108, + "step": 43710 + }, + { + "epoch": 0.962161567246313, + "grad_norm": 1.9422052668924301, + "learning_rate": 3.749910765539322e-08, + "loss": 0.2963, + "step": 43711 + }, + { + "epoch": 0.9621835791327317, + "grad_norm": 1.9614915126693615, + "learning_rate": 3.7455545903915536e-08, + "loss": 0.2118, + "step": 43712 + }, + { + "epoch": 0.9622055910191504, + "grad_norm": 1.7061425405371902, + "learning_rate": 3.741200937430245e-08, + "loss": 0.2581, + "step": 43713 + }, + { + "epoch": 0.962227602905569, + "grad_norm": 1.8307729539763573, + "learning_rate": 3.736849806677434e-08, + "loss": 0.3616, + "step": 43714 + }, + { + "epoch": 0.9622496147919877, + "grad_norm": 1.7008980001312737, + "learning_rate": 3.73250119815538e-08, + "loss": 0.3275, + "step": 43715 + }, + { + "epoch": 0.9622716266784064, + "grad_norm": 5.315816500731192, + "learning_rate": 3.728155111886012e-08, + "loss": 0.2869, + "step": 43716 + }, + { + "epoch": 0.962293638564825, + "grad_norm": 1.3575247953474725, + "learning_rate": 3.723811547891531e-08, + "loss": 0.3531, + "step": 43717 + }, + { + "epoch": 0.9623156504512437, + "grad_norm": 1.5288256987288082, + "learning_rate": 3.7194705061939786e-08, + "loss": 0.1943, + "step": 43718 + }, + { + "epoch": 0.9623376623376624, + "grad_norm": 2.3154644479281914, + "learning_rate": 3.7151319868154454e-08, + "loss": 0.3455, + "step": 43719 + }, + { + "epoch": 0.962359674224081, + "grad_norm": 2.515490086210051, + "learning_rate": 3.710795989777971e-08, + "loss": 0.3654, + "step": 43720 + }, + { + "epoch": 0.9623816861104997, + "grad_norm": 2.1559125931668452, + "learning_rate": 3.706462515103537e-08, + "loss": 0.2917, + "step": 43721 + }, + { + "epoch": 0.9624036979969184, + "grad_norm": 1.73756025962773, + "learning_rate": 3.702131562814237e-08, + "loss": 0.362, + "step": 43722 + }, + { + "epoch": 0.962425709883337, + "grad_norm": 2.1477941368295017, + "learning_rate": 3.697803132932054e-08, + "loss": 0.319, + "step": 43723 + }, + { + "epoch": 0.9624477217697557, + "grad_norm": 1.4241939132496055, + "learning_rate": 3.693477225479025e-08, + "loss": 0.264, + "step": 43724 + }, + { + "epoch": 0.9624697336561744, + "grad_norm": 1.5948132029437023, + "learning_rate": 3.6891538404770776e-08, + "loss": 0.2828, + "step": 43725 + }, + { + "epoch": 0.962491745542593, + "grad_norm": 1.357687521830723, + "learning_rate": 3.6848329779482496e-08, + "loss": 0.2966, + "step": 43726 + }, + { + "epoch": 0.9625137574290117, + "grad_norm": 1.6589091651725056, + "learning_rate": 3.6805146379143565e-08, + "loss": 0.3124, + "step": 43727 + }, + { + "epoch": 0.9625357693154303, + "grad_norm": 0.6230349118998959, + "learning_rate": 3.676198820397547e-08, + "loss": 0.4405, + "step": 43728 + }, + { + "epoch": 0.962557781201849, + "grad_norm": 1.3007186671232487, + "learning_rate": 3.6718855254196384e-08, + "loss": 0.2523, + "step": 43729 + }, + { + "epoch": 0.9625797930882677, + "grad_norm": 1.8539529529383194, + "learning_rate": 3.6675747530025565e-08, + "loss": 0.2605, + "step": 43730 + }, + { + "epoch": 0.9626018049746863, + "grad_norm": 4.733755304023059, + "learning_rate": 3.6632665031682834e-08, + "loss": 0.2396, + "step": 43731 + }, + { + "epoch": 0.962623816861105, + "grad_norm": 1.5577509839304327, + "learning_rate": 3.6589607759385805e-08, + "loss": 0.3106, + "step": 43732 + }, + { + "epoch": 0.9626458287475237, + "grad_norm": 0.6401864972855864, + "learning_rate": 3.6546575713354846e-08, + "loss": 0.4448, + "step": 43733 + }, + { + "epoch": 0.9626678406339423, + "grad_norm": 1.7216830200383175, + "learning_rate": 3.650356889380757e-08, + "loss": 0.2821, + "step": 43734 + }, + { + "epoch": 0.962689852520361, + "grad_norm": 0.5959292458521981, + "learning_rate": 3.646058730096325e-08, + "loss": 0.4495, + "step": 43735 + }, + { + "epoch": 0.9627118644067797, + "grad_norm": 1.8945958971510437, + "learning_rate": 3.6417630935040025e-08, + "loss": 0.2394, + "step": 43736 + }, + { + "epoch": 0.9627338762931983, + "grad_norm": 0.5936827624200958, + "learning_rate": 3.6374699796256075e-08, + "loss": 0.4323, + "step": 43737 + }, + { + "epoch": 0.962755888179617, + "grad_norm": 3.7560741337947836, + "learning_rate": 3.633179388482955e-08, + "loss": 0.2537, + "step": 43738 + }, + { + "epoch": 0.9627779000660357, + "grad_norm": 2.1122656211590636, + "learning_rate": 3.62889132009786e-08, + "loss": 0.3612, + "step": 43739 + }, + { + "epoch": 0.9627999119524543, + "grad_norm": 1.8300488827607613, + "learning_rate": 3.624605774492196e-08, + "loss": 0.358, + "step": 43740 + }, + { + "epoch": 0.962821923838873, + "grad_norm": 1.4989320531269754, + "learning_rate": 3.620322751687611e-08, + "loss": 0.2896, + "step": 43741 + }, + { + "epoch": 0.9628439357252917, + "grad_norm": 0.610375420862411, + "learning_rate": 3.616042251705976e-08, + "loss": 0.4545, + "step": 43742 + }, + { + "epoch": 0.9628659476117103, + "grad_norm": 1.6214541750436136, + "learning_rate": 3.611764274568941e-08, + "loss": 0.2321, + "step": 43743 + }, + { + "epoch": 0.962887959498129, + "grad_norm": 1.256153597941519, + "learning_rate": 3.607488820298377e-08, + "loss": 0.1972, + "step": 43744 + }, + { + "epoch": 0.9629099713845477, + "grad_norm": 1.6363604090249515, + "learning_rate": 3.603215888915934e-08, + "loss": 0.2818, + "step": 43745 + }, + { + "epoch": 0.9629319832709663, + "grad_norm": 1.5710480968886136, + "learning_rate": 3.598945480443372e-08, + "loss": 0.2118, + "step": 43746 + }, + { + "epoch": 0.962953995157385, + "grad_norm": 0.638311641952483, + "learning_rate": 3.5946775949023406e-08, + "loss": 0.4634, + "step": 43747 + }, + { + "epoch": 0.9629760070438037, + "grad_norm": 2.003290680825852, + "learning_rate": 3.5904122323145995e-08, + "loss": 0.3087, + "step": 43748 + }, + { + "epoch": 0.9629980189302223, + "grad_norm": 2.1348169622350928, + "learning_rate": 3.586149392701743e-08, + "loss": 0.2394, + "step": 43749 + }, + { + "epoch": 0.963020030816641, + "grad_norm": 2.0061680474968284, + "learning_rate": 3.5818890760854766e-08, + "loss": 0.327, + "step": 43750 + }, + { + "epoch": 0.9630420427030596, + "grad_norm": 1.367037844124007, + "learning_rate": 3.577631282487504e-08, + "loss": 0.2587, + "step": 43751 + }, + { + "epoch": 0.9630640545894783, + "grad_norm": 0.6029424654568405, + "learning_rate": 3.573376011929419e-08, + "loss": 0.4449, + "step": 43752 + }, + { + "epoch": 0.963086066475897, + "grad_norm": 1.4395704155059144, + "learning_rate": 3.569123264432817e-08, + "loss": 0.3342, + "step": 43753 + }, + { + "epoch": 0.9631080783623156, + "grad_norm": 0.6218239210891964, + "learning_rate": 3.564873040019345e-08, + "loss": 0.4474, + "step": 43754 + }, + { + "epoch": 0.9631300902487343, + "grad_norm": 0.5910814052412638, + "learning_rate": 3.5606253387105994e-08, + "loss": 0.4543, + "step": 43755 + }, + { + "epoch": 0.963152102135153, + "grad_norm": 1.7637920545629318, + "learning_rate": 3.556380160528228e-08, + "loss": 0.3096, + "step": 43756 + }, + { + "epoch": 0.9631741140215716, + "grad_norm": 3.6656041283466725, + "learning_rate": 3.552137505493658e-08, + "loss": 0.3042, + "step": 43757 + }, + { + "epoch": 0.9631961259079903, + "grad_norm": 6.77556433255556, + "learning_rate": 3.547897373628595e-08, + "loss": 0.2724, + "step": 43758 + }, + { + "epoch": 0.963218137794409, + "grad_norm": 1.8686170549588696, + "learning_rate": 3.543659764954577e-08, + "loss": 0.2996, + "step": 43759 + }, + { + "epoch": 0.9632401496808276, + "grad_norm": 1.6072694372134129, + "learning_rate": 3.5394246794930864e-08, + "loss": 0.3171, + "step": 43760 + }, + { + "epoch": 0.9632621615672463, + "grad_norm": 1.6383971434625428, + "learning_rate": 3.5351921172656624e-08, + "loss": 0.3143, + "step": 43761 + }, + { + "epoch": 0.963284173453665, + "grad_norm": 1.7818696126858478, + "learning_rate": 3.530962078293843e-08, + "loss": 0.3075, + "step": 43762 + }, + { + "epoch": 0.9633061853400836, + "grad_norm": 1.5682192140509525, + "learning_rate": 3.526734562599054e-08, + "loss": 0.2333, + "step": 43763 + }, + { + "epoch": 0.9633281972265023, + "grad_norm": 2.007732567057231, + "learning_rate": 3.5225095702028364e-08, + "loss": 0.3257, + "step": 43764 + }, + { + "epoch": 0.963350209112921, + "grad_norm": 1.4033545216038512, + "learning_rate": 3.5182871011266716e-08, + "loss": 0.3132, + "step": 43765 + }, + { + "epoch": 0.9633722209993396, + "grad_norm": 1.9292213963939024, + "learning_rate": 3.5140671553920426e-08, + "loss": 0.2966, + "step": 43766 + }, + { + "epoch": 0.9633942328857583, + "grad_norm": 1.4659505023853208, + "learning_rate": 3.509849733020376e-08, + "loss": 0.3363, + "step": 43767 + }, + { + "epoch": 0.963416244772177, + "grad_norm": 1.7164784490798148, + "learning_rate": 3.505634834033045e-08, + "loss": 0.2619, + "step": 43768 + }, + { + "epoch": 0.9634382566585956, + "grad_norm": 2.485870988674194, + "learning_rate": 3.501422458451531e-08, + "loss": 0.3298, + "step": 43769 + }, + { + "epoch": 0.9634602685450143, + "grad_norm": 2.099810167482288, + "learning_rate": 3.4972126062972066e-08, + "loss": 0.2814, + "step": 43770 + }, + { + "epoch": 0.963482280431433, + "grad_norm": 1.4711255468632114, + "learning_rate": 3.4930052775915544e-08, + "loss": 0.2172, + "step": 43771 + }, + { + "epoch": 0.9635042923178516, + "grad_norm": 2.492531659185622, + "learning_rate": 3.488800472355891e-08, + "loss": 0.2538, + "step": 43772 + }, + { + "epoch": 0.9635263042042703, + "grad_norm": 1.53782060742521, + "learning_rate": 3.4845981906115325e-08, + "loss": 0.2178, + "step": 43773 + }, + { + "epoch": 0.9635483160906889, + "grad_norm": 1.5513814902473086, + "learning_rate": 3.480398432379906e-08, + "loss": 0.3236, + "step": 43774 + }, + { + "epoch": 0.9635703279771076, + "grad_norm": 1.9688737502134728, + "learning_rate": 3.476201197682383e-08, + "loss": 0.2662, + "step": 43775 + }, + { + "epoch": 0.9635923398635263, + "grad_norm": 1.6720857312538029, + "learning_rate": 3.4720064865402805e-08, + "loss": 0.2526, + "step": 43776 + }, + { + "epoch": 0.9636143517499449, + "grad_norm": 0.5993166541855327, + "learning_rate": 3.4678142989749145e-08, + "loss": 0.4204, + "step": 43777 + }, + { + "epoch": 0.9636363636363636, + "grad_norm": 1.2985798110194302, + "learning_rate": 3.46362463500749e-08, + "loss": 0.3147, + "step": 43778 + }, + { + "epoch": 0.9636583755227823, + "grad_norm": 0.6308262142070133, + "learning_rate": 3.4594374946594343e-08, + "loss": 0.4499, + "step": 43779 + }, + { + "epoch": 0.9636803874092009, + "grad_norm": 0.5951769712223314, + "learning_rate": 3.455252877951953e-08, + "loss": 0.4611, + "step": 43780 + }, + { + "epoch": 0.9637023992956196, + "grad_norm": 4.192058261398392, + "learning_rate": 3.4510707849064185e-08, + "loss": 0.1966, + "step": 43781 + }, + { + "epoch": 0.9637244111820383, + "grad_norm": 0.6038117834852735, + "learning_rate": 3.4468912155439235e-08, + "loss": 0.4302, + "step": 43782 + }, + { + "epoch": 0.9637464230684569, + "grad_norm": 1.63755562553265, + "learning_rate": 3.442714169885841e-08, + "loss": 0.2857, + "step": 43783 + }, + { + "epoch": 0.9637684349548756, + "grad_norm": 1.780035346979921, + "learning_rate": 3.4385396479533205e-08, + "loss": 0.2695, + "step": 43784 + }, + { + "epoch": 0.9637904468412943, + "grad_norm": 1.6230664036109332, + "learning_rate": 3.434367649767678e-08, + "loss": 0.3126, + "step": 43785 + }, + { + "epoch": 0.9638124587277129, + "grad_norm": 2.0842551322794405, + "learning_rate": 3.430198175350008e-08, + "loss": 0.2912, + "step": 43786 + }, + { + "epoch": 0.9638344706141316, + "grad_norm": 1.8741678474236905, + "learning_rate": 3.426031224721571e-08, + "loss": 0.4167, + "step": 43787 + }, + { + "epoch": 0.9638564825005503, + "grad_norm": 1.7139053884359425, + "learning_rate": 3.4218667979034615e-08, + "loss": 0.3343, + "step": 43788 + }, + { + "epoch": 0.9638784943869689, + "grad_norm": 1.7632708243305477, + "learning_rate": 3.41770489491694e-08, + "loss": 0.2613, + "step": 43789 + }, + { + "epoch": 0.9639005062733876, + "grad_norm": 1.4673667196320512, + "learning_rate": 3.4135455157831564e-08, + "loss": 0.2249, + "step": 43790 + }, + { + "epoch": 0.9639225181598063, + "grad_norm": 1.3682250540331051, + "learning_rate": 3.409388660523205e-08, + "loss": 0.3138, + "step": 43791 + }, + { + "epoch": 0.9639445300462249, + "grad_norm": 2.0961105705647993, + "learning_rate": 3.4052343291582357e-08, + "loss": 0.2971, + "step": 43792 + }, + { + "epoch": 0.9639665419326436, + "grad_norm": 1.360263145940407, + "learning_rate": 3.401082521709287e-08, + "loss": 0.2795, + "step": 43793 + }, + { + "epoch": 0.9639885538190623, + "grad_norm": 2.767266234355468, + "learning_rate": 3.396933238197564e-08, + "loss": 0.2255, + "step": 43794 + }, + { + "epoch": 0.9640105657054809, + "grad_norm": 1.700595086155996, + "learning_rate": 3.392786478644161e-08, + "loss": 0.276, + "step": 43795 + }, + { + "epoch": 0.9640325775918996, + "grad_norm": 0.6319539075248618, + "learning_rate": 3.3886422430700615e-08, + "loss": 0.4438, + "step": 43796 + }, + { + "epoch": 0.9640545894783183, + "grad_norm": 1.376629412246248, + "learning_rate": 3.3845005314964154e-08, + "loss": 0.2137, + "step": 43797 + }, + { + "epoch": 0.9640766013647369, + "grad_norm": 2.9180880443450747, + "learning_rate": 3.3803613439441496e-08, + "loss": 0.3151, + "step": 43798 + }, + { + "epoch": 0.9640986132511556, + "grad_norm": 1.6288645738098995, + "learning_rate": 3.37622468043447e-08, + "loss": 0.2895, + "step": 43799 + }, + { + "epoch": 0.9641206251375742, + "grad_norm": 1.3133311508384604, + "learning_rate": 3.372090540988304e-08, + "loss": 0.3134, + "step": 43800 + }, + { + "epoch": 0.9641426370239929, + "grad_norm": 3.4766230005314647, + "learning_rate": 3.3679589256266356e-08, + "loss": 0.2011, + "step": 43801 + }, + { + "epoch": 0.9641646489104116, + "grad_norm": 1.280200327889462, + "learning_rate": 3.363829834370558e-08, + "loss": 0.3065, + "step": 43802 + }, + { + "epoch": 0.9641866607968302, + "grad_norm": 1.607791213883814, + "learning_rate": 3.359703267240999e-08, + "loss": 0.2364, + "step": 43803 + }, + { + "epoch": 0.964208672683249, + "grad_norm": 0.6200373200594573, + "learning_rate": 3.355579224258887e-08, + "loss": 0.4087, + "step": 43804 + }, + { + "epoch": 0.9642306845696677, + "grad_norm": 2.847285775958686, + "learning_rate": 3.35145770544526e-08, + "loss": 0.2823, + "step": 43805 + }, + { + "epoch": 0.9642526964560862, + "grad_norm": 2.188247743287345, + "learning_rate": 3.347338710821046e-08, + "loss": 0.3751, + "step": 43806 + }, + { + "epoch": 0.964274708342505, + "grad_norm": 2.1662673434106354, + "learning_rate": 3.3432222404071736e-08, + "loss": 0.2451, + "step": 43807 + }, + { + "epoch": 0.9642967202289237, + "grad_norm": 1.492781431194264, + "learning_rate": 3.339108294224513e-08, + "loss": 0.284, + "step": 43808 + }, + { + "epoch": 0.9643187321153422, + "grad_norm": 2.0314304967685843, + "learning_rate": 3.33499687229405e-08, + "loss": 0.2693, + "step": 43809 + }, + { + "epoch": 0.964340744001761, + "grad_norm": 1.3565609709107738, + "learning_rate": 3.330887974636654e-08, + "loss": 0.3267, + "step": 43810 + }, + { + "epoch": 0.9643627558881797, + "grad_norm": 2.1556772511681124, + "learning_rate": 3.326781601273199e-08, + "loss": 0.2686, + "step": 43811 + }, + { + "epoch": 0.9643847677745983, + "grad_norm": 1.6638909095476226, + "learning_rate": 3.322677752224612e-08, + "loss": 0.343, + "step": 43812 + }, + { + "epoch": 0.964406779661017, + "grad_norm": 0.6025047662521638, + "learning_rate": 3.318576427511655e-08, + "loss": 0.4448, + "step": 43813 + }, + { + "epoch": 0.9644287915474357, + "grad_norm": 1.371097559594165, + "learning_rate": 3.314477627155199e-08, + "loss": 0.2935, + "step": 43814 + }, + { + "epoch": 0.9644508034338543, + "grad_norm": 1.794175364049653, + "learning_rate": 3.310381351176117e-08, + "loss": 0.2952, + "step": 43815 + }, + { + "epoch": 0.964472815320273, + "grad_norm": 1.83299654473181, + "learning_rate": 3.306287599595226e-08, + "loss": 0.2646, + "step": 43816 + }, + { + "epoch": 0.9644948272066917, + "grad_norm": 1.5770685302808705, + "learning_rate": 3.302196372433286e-08, + "loss": 0.374, + "step": 43817 + }, + { + "epoch": 0.9645168390931103, + "grad_norm": 1.8400517931646196, + "learning_rate": 3.298107669711115e-08, + "loss": 0.3949, + "step": 43818 + }, + { + "epoch": 0.964538850979529, + "grad_norm": 1.822287497478269, + "learning_rate": 3.294021491449528e-08, + "loss": 0.2994, + "step": 43819 + }, + { + "epoch": 0.9645608628659477, + "grad_norm": 1.944157892016858, + "learning_rate": 3.2899378376692324e-08, + "loss": 0.3068, + "step": 43820 + }, + { + "epoch": 0.9645828747523663, + "grad_norm": 3.1540734813148785, + "learning_rate": 3.285856708391044e-08, + "loss": 0.3449, + "step": 43821 + }, + { + "epoch": 0.964604886638785, + "grad_norm": 0.6131136465603292, + "learning_rate": 3.281778103635613e-08, + "loss": 0.4399, + "step": 43822 + }, + { + "epoch": 0.9646268985252036, + "grad_norm": 0.6277492395492694, + "learning_rate": 3.2777020234238123e-08, + "loss": 0.4725, + "step": 43823 + }, + { + "epoch": 0.9646489104116223, + "grad_norm": 3.628402414562652, + "learning_rate": 3.27362846777618e-08, + "loss": 0.3481, + "step": 43824 + }, + { + "epoch": 0.964670922298041, + "grad_norm": 3.050899404031536, + "learning_rate": 3.269557436713588e-08, + "loss": 0.2916, + "step": 43825 + }, + { + "epoch": 0.9646929341844596, + "grad_norm": 2.3519171378909887, + "learning_rate": 3.265488930256633e-08, + "loss": 0.2077, + "step": 43826 + }, + { + "epoch": 0.9647149460708783, + "grad_norm": 1.769214555878805, + "learning_rate": 3.261422948425963e-08, + "loss": 0.3342, + "step": 43827 + }, + { + "epoch": 0.964736957957297, + "grad_norm": 4.3837664281620725, + "learning_rate": 3.257359491242396e-08, + "loss": 0.3025, + "step": 43828 + }, + { + "epoch": 0.9647589698437156, + "grad_norm": 2.3923038842004987, + "learning_rate": 3.253298558726359e-08, + "loss": 0.2821, + "step": 43829 + }, + { + "epoch": 0.9647809817301343, + "grad_norm": 2.0724065245508405, + "learning_rate": 3.2492401508986693e-08, + "loss": 0.247, + "step": 43830 + }, + { + "epoch": 0.964802993616553, + "grad_norm": 1.5942261979273589, + "learning_rate": 3.245184267779922e-08, + "loss": 0.3285, + "step": 43831 + }, + { + "epoch": 0.9648250055029716, + "grad_norm": 2.791015788876953, + "learning_rate": 3.241130909390711e-08, + "loss": 0.2351, + "step": 43832 + }, + { + "epoch": 0.9648470173893903, + "grad_norm": 2.9680864947214083, + "learning_rate": 3.237080075751631e-08, + "loss": 0.2402, + "step": 43833 + }, + { + "epoch": 0.964869029275809, + "grad_norm": 1.79964461556077, + "learning_rate": 3.2330317668832214e-08, + "loss": 0.2226, + "step": 43834 + }, + { + "epoch": 0.9648910411622276, + "grad_norm": 1.6643247021259704, + "learning_rate": 3.228985982806132e-08, + "loss": 0.2203, + "step": 43835 + }, + { + "epoch": 0.9649130530486463, + "grad_norm": 2.065185181489589, + "learning_rate": 3.2249427235409025e-08, + "loss": 0.2349, + "step": 43836 + }, + { + "epoch": 0.964935064935065, + "grad_norm": 1.5830734897734051, + "learning_rate": 3.220901989108127e-08, + "loss": 0.23, + "step": 43837 + }, + { + "epoch": 0.9649570768214836, + "grad_norm": 1.698375021132253, + "learning_rate": 3.216863779528234e-08, + "loss": 0.2881, + "step": 43838 + }, + { + "epoch": 0.9649790887079023, + "grad_norm": 2.1069310307482256, + "learning_rate": 3.212828094821874e-08, + "loss": 0.3076, + "step": 43839 + }, + { + "epoch": 0.965001100594321, + "grad_norm": 1.5470573376652614, + "learning_rate": 3.208794935009474e-08, + "loss": 0.3219, + "step": 43840 + }, + { + "epoch": 0.9650231124807396, + "grad_norm": 2.4913498764612263, + "learning_rate": 3.2047643001115184e-08, + "loss": 0.2709, + "step": 43841 + }, + { + "epoch": 0.9650451243671583, + "grad_norm": 0.6228549174661369, + "learning_rate": 3.2007361901485455e-08, + "loss": 0.4454, + "step": 43842 + }, + { + "epoch": 0.965067136253577, + "grad_norm": 1.9348131980887453, + "learning_rate": 3.196710605140985e-08, + "loss": 0.2728, + "step": 43843 + }, + { + "epoch": 0.9650891481399956, + "grad_norm": 1.578134356134629, + "learning_rate": 3.1926875451093744e-08, + "loss": 0.2536, + "step": 43844 + }, + { + "epoch": 0.9651111600264143, + "grad_norm": 2.027785592347462, + "learning_rate": 3.188667010074087e-08, + "loss": 0.2683, + "step": 43845 + }, + { + "epoch": 0.9651331719128329, + "grad_norm": 0.620575509752701, + "learning_rate": 3.1846490000555505e-08, + "loss": 0.4545, + "step": 43846 + }, + { + "epoch": 0.9651551837992516, + "grad_norm": 2.0520531054927242, + "learning_rate": 3.1806335150742494e-08, + "loss": 0.2972, + "step": 43847 + }, + { + "epoch": 0.9651771956856703, + "grad_norm": 1.8379601741763745, + "learning_rate": 3.176620555150556e-08, + "loss": 0.3947, + "step": 43848 + }, + { + "epoch": 0.9651992075720889, + "grad_norm": 1.5348765274811191, + "learning_rate": 3.172610120304842e-08, + "loss": 0.3153, + "step": 43849 + }, + { + "epoch": 0.9652212194585076, + "grad_norm": 1.525497992697022, + "learning_rate": 3.168602210557536e-08, + "loss": 0.2979, + "step": 43850 + }, + { + "epoch": 0.9652432313449263, + "grad_norm": 1.7138008107906562, + "learning_rate": 3.1645968259289564e-08, + "loss": 0.2929, + "step": 43851 + }, + { + "epoch": 0.9652652432313449, + "grad_norm": 0.6073401226664541, + "learning_rate": 3.160593966439473e-08, + "loss": 0.4362, + "step": 43852 + }, + { + "epoch": 0.9652872551177636, + "grad_norm": 1.5655599429591363, + "learning_rate": 3.156593632109517e-08, + "loss": 0.3343, + "step": 43853 + }, + { + "epoch": 0.9653092670041823, + "grad_norm": 1.7486503616607985, + "learning_rate": 3.1525958229593475e-08, + "loss": 0.3431, + "step": 43854 + }, + { + "epoch": 0.9653312788906009, + "grad_norm": 0.6221144570432385, + "learning_rate": 3.1486005390092275e-08, + "loss": 0.4426, + "step": 43855 + }, + { + "epoch": 0.9653532907770196, + "grad_norm": 4.124450024588224, + "learning_rate": 3.144607780279529e-08, + "loss": 0.297, + "step": 43856 + }, + { + "epoch": 0.9653753026634383, + "grad_norm": 1.6918967957855124, + "learning_rate": 3.1406175467905695e-08, + "loss": 0.3076, + "step": 43857 + }, + { + "epoch": 0.9653973145498569, + "grad_norm": 1.7726808596966457, + "learning_rate": 3.13662983856261e-08, + "loss": 0.2497, + "step": 43858 + }, + { + "epoch": 0.9654193264362756, + "grad_norm": 4.929738169444717, + "learning_rate": 3.132644655615857e-08, + "loss": 0.2829, + "step": 43859 + }, + { + "epoch": 0.9654413383226943, + "grad_norm": 1.6796407473434454, + "learning_rate": 3.128661997970628e-08, + "loss": 0.3703, + "step": 43860 + }, + { + "epoch": 0.9654633502091129, + "grad_norm": 1.6661116652407781, + "learning_rate": 3.124681865647183e-08, + "loss": 0.2863, + "step": 43861 + }, + { + "epoch": 0.9654853620955316, + "grad_norm": 1.3611076088638956, + "learning_rate": 3.120704258665674e-08, + "loss": 0.2709, + "step": 43862 + }, + { + "epoch": 0.9655073739819503, + "grad_norm": 2.1825831098653645, + "learning_rate": 3.116729177046418e-08, + "loss": 0.285, + "step": 43863 + }, + { + "epoch": 0.9655293858683689, + "grad_norm": 2.2266112455396536, + "learning_rate": 3.11275662080951e-08, + "loss": 0.3533, + "step": 43864 + }, + { + "epoch": 0.9655513977547876, + "grad_norm": 2.4068971052395955, + "learning_rate": 3.10878658997521e-08, + "loss": 0.2973, + "step": 43865 + }, + { + "epoch": 0.9655734096412063, + "grad_norm": 2.0875060116003965, + "learning_rate": 3.104819084563671e-08, + "loss": 0.3385, + "step": 43866 + }, + { + "epoch": 0.9655954215276249, + "grad_norm": 1.804262267410078, + "learning_rate": 3.100854104595041e-08, + "loss": 0.2598, + "step": 43867 + }, + { + "epoch": 0.9656174334140436, + "grad_norm": 0.6007182563850346, + "learning_rate": 3.0968916500895286e-08, + "loss": 0.4311, + "step": 43868 + }, + { + "epoch": 0.9656394453004622, + "grad_norm": 0.6238130261951892, + "learning_rate": 3.092931721067227e-08, + "loss": 0.4687, + "step": 43869 + }, + { + "epoch": 0.9656614571868809, + "grad_norm": 1.3047586664998376, + "learning_rate": 3.0889743175482325e-08, + "loss": 0.208, + "step": 43870 + }, + { + "epoch": 0.9656834690732996, + "grad_norm": 1.4451306787659737, + "learning_rate": 3.08501943955275e-08, + "loss": 0.2596, + "step": 43871 + }, + { + "epoch": 0.9657054809597182, + "grad_norm": 1.537514557632684, + "learning_rate": 3.08106708710082e-08, + "loss": 0.258, + "step": 43872 + }, + { + "epoch": 0.9657274928461369, + "grad_norm": 1.6126382171021227, + "learning_rate": 3.077117260212537e-08, + "loss": 0.2812, + "step": 43873 + }, + { + "epoch": 0.9657495047325556, + "grad_norm": 1.5307224060261677, + "learning_rate": 3.0731699589079956e-08, + "loss": 0.2789, + "step": 43874 + }, + { + "epoch": 0.9657715166189742, + "grad_norm": 2.6470056796014942, + "learning_rate": 3.069225183207181e-08, + "loss": 0.3559, + "step": 43875 + }, + { + "epoch": 0.9657935285053929, + "grad_norm": 1.7307824013799513, + "learning_rate": 3.0652829331302424e-08, + "loss": 0.2942, + "step": 43876 + }, + { + "epoch": 0.9658155403918116, + "grad_norm": 1.5484465416339894, + "learning_rate": 3.061343208697165e-08, + "loss": 0.2016, + "step": 43877 + }, + { + "epoch": 0.9658375522782302, + "grad_norm": 1.7059206392718127, + "learning_rate": 3.057406009927988e-08, + "loss": 0.3057, + "step": 43878 + }, + { + "epoch": 0.9658595641646489, + "grad_norm": 2.7348722545613504, + "learning_rate": 3.05347133684275e-08, + "loss": 0.2555, + "step": 43879 + }, + { + "epoch": 0.9658815760510676, + "grad_norm": 1.9333478324762057, + "learning_rate": 3.049539189461381e-08, + "loss": 0.2204, + "step": 43880 + }, + { + "epoch": 0.9659035879374862, + "grad_norm": 0.6207267931877626, + "learning_rate": 3.04560956780392e-08, + "loss": 0.4434, + "step": 43881 + }, + { + "epoch": 0.9659255998239049, + "grad_norm": 1.7157782749389503, + "learning_rate": 3.0416824718902945e-08, + "loss": 0.2887, + "step": 43882 + }, + { + "epoch": 0.9659476117103236, + "grad_norm": 0.6106889772762869, + "learning_rate": 3.037757901740435e-08, + "loss": 0.4272, + "step": 43883 + }, + { + "epoch": 0.9659696235967422, + "grad_norm": 2.3682680754028578, + "learning_rate": 3.033835857374434e-08, + "loss": 0.341, + "step": 43884 + }, + { + "epoch": 0.9659916354831609, + "grad_norm": 1.54269372476928, + "learning_rate": 3.029916338812111e-08, + "loss": 0.2252, + "step": 43885 + }, + { + "epoch": 0.9660136473695796, + "grad_norm": 2.1926392385186158, + "learning_rate": 3.025999346073394e-08, + "loss": 0.3441, + "step": 43886 + }, + { + "epoch": 0.9660356592559982, + "grad_norm": 0.6085275129376891, + "learning_rate": 3.022084879178211e-08, + "loss": 0.4417, + "step": 43887 + }, + { + "epoch": 0.9660576711424169, + "grad_norm": 0.5959723483909588, + "learning_rate": 3.018172938146435e-08, + "loss": 0.4232, + "step": 43888 + }, + { + "epoch": 0.9660796830288356, + "grad_norm": 2.0960932653527364, + "learning_rate": 3.0142635229979954e-08, + "loss": 0.242, + "step": 43889 + }, + { + "epoch": 0.9661016949152542, + "grad_norm": 1.4649183834820438, + "learning_rate": 3.010356633752709e-08, + "loss": 0.329, + "step": 43890 + }, + { + "epoch": 0.9661237068016729, + "grad_norm": 0.6062019254752201, + "learning_rate": 3.0064522704304486e-08, + "loss": 0.4656, + "step": 43891 + }, + { + "epoch": 0.9661457186880915, + "grad_norm": 2.1760781831416396, + "learning_rate": 3.0025504330510325e-08, + "loss": 0.313, + "step": 43892 + }, + { + "epoch": 0.9661677305745102, + "grad_norm": 1.994102631220702, + "learning_rate": 2.9986511216343883e-08, + "loss": 0.3872, + "step": 43893 + }, + { + "epoch": 0.9661897424609289, + "grad_norm": 0.6217776996792637, + "learning_rate": 2.994754336200223e-08, + "loss": 0.4157, + "step": 43894 + }, + { + "epoch": 0.9662117543473475, + "grad_norm": 0.6045461242373396, + "learning_rate": 2.990860076768409e-08, + "loss": 0.4364, + "step": 43895 + }, + { + "epoch": 0.9662337662337662, + "grad_norm": 2.002432903219003, + "learning_rate": 2.986968343358709e-08, + "loss": 0.2165, + "step": 43896 + }, + { + "epoch": 0.9662557781201849, + "grad_norm": 1.447369318793617, + "learning_rate": 2.983079135990941e-08, + "loss": 0.2589, + "step": 43897 + }, + { + "epoch": 0.9662777900066035, + "grad_norm": 1.9390768669962837, + "learning_rate": 2.979192454684754e-08, + "loss": 0.2404, + "step": 43898 + }, + { + "epoch": 0.9662998018930222, + "grad_norm": 1.8687629927953662, + "learning_rate": 2.975308299460078e-08, + "loss": 0.2439, + "step": 43899 + }, + { + "epoch": 0.9663218137794409, + "grad_norm": 1.5632837534705704, + "learning_rate": 2.9714266703365634e-08, + "loss": 0.2459, + "step": 43900 + }, + { + "epoch": 0.9663438256658595, + "grad_norm": 1.7459020052739533, + "learning_rate": 2.9675475673338616e-08, + "loss": 0.3061, + "step": 43901 + }, + { + "epoch": 0.9663658375522782, + "grad_norm": 1.4930732960323225, + "learning_rate": 2.963670990471845e-08, + "loss": 0.2683, + "step": 43902 + }, + { + "epoch": 0.9663878494386969, + "grad_norm": 2.2042361398057078, + "learning_rate": 2.9597969397701098e-08, + "loss": 0.2953, + "step": 43903 + }, + { + "epoch": 0.9664098613251155, + "grad_norm": 2.0128171279814584, + "learning_rate": 2.9559254152483617e-08, + "loss": 0.3411, + "step": 43904 + }, + { + "epoch": 0.9664318732115342, + "grad_norm": 2.37882333579296, + "learning_rate": 2.9520564169263077e-08, + "loss": 0.3179, + "step": 43905 + }, + { + "epoch": 0.9664538850979529, + "grad_norm": 1.59120591475707, + "learning_rate": 2.948189944823543e-08, + "loss": 0.2894, + "step": 43906 + }, + { + "epoch": 0.9664758969843715, + "grad_norm": 0.6208148629000987, + "learning_rate": 2.944325998959774e-08, + "loss": 0.4535, + "step": 43907 + }, + { + "epoch": 0.9664979088707902, + "grad_norm": 1.791339811367171, + "learning_rate": 2.9404645793547072e-08, + "loss": 0.2279, + "step": 43908 + }, + { + "epoch": 0.9665199207572089, + "grad_norm": 0.6008367165801801, + "learning_rate": 2.936605686027827e-08, + "loss": 0.4403, + "step": 43909 + }, + { + "epoch": 0.9665419326436275, + "grad_norm": 2.114760359634141, + "learning_rate": 2.9327493189988398e-08, + "loss": 0.2839, + "step": 43910 + }, + { + "epoch": 0.9665639445300462, + "grad_norm": 1.6048342400468592, + "learning_rate": 2.9288954782873415e-08, + "loss": 0.2745, + "step": 43911 + }, + { + "epoch": 0.9665859564164649, + "grad_norm": 2.1598236644138553, + "learning_rate": 2.9250441639128712e-08, + "loss": 0.3129, + "step": 43912 + }, + { + "epoch": 0.9666079683028835, + "grad_norm": 5.6017514185391075, + "learning_rate": 2.921195375895025e-08, + "loss": 0.3003, + "step": 43913 + }, + { + "epoch": 0.9666299801893022, + "grad_norm": 1.8040543477131885, + "learning_rate": 2.9173491142533427e-08, + "loss": 0.287, + "step": 43914 + }, + { + "epoch": 0.9666519920757208, + "grad_norm": 1.6969853577028502, + "learning_rate": 2.9135053790074196e-08, + "loss": 0.2596, + "step": 43915 + }, + { + "epoch": 0.9666740039621395, + "grad_norm": 1.9203257798108784, + "learning_rate": 2.909664170176796e-08, + "loss": 0.3233, + "step": 43916 + }, + { + "epoch": 0.9666960158485582, + "grad_norm": 3.1268290612121743, + "learning_rate": 2.9058254877809555e-08, + "loss": 0.298, + "step": 43917 + }, + { + "epoch": 0.9667180277349768, + "grad_norm": 2.0686684454563156, + "learning_rate": 2.9019893318393832e-08, + "loss": 0.2905, + "step": 43918 + }, + { + "epoch": 0.9667400396213955, + "grad_norm": 1.663836094244637, + "learning_rate": 2.898155702371619e-08, + "loss": 0.3351, + "step": 43919 + }, + { + "epoch": 0.9667620515078142, + "grad_norm": 2.642963547582706, + "learning_rate": 2.8943245993972026e-08, + "loss": 0.2643, + "step": 43920 + }, + { + "epoch": 0.9667840633942328, + "grad_norm": 1.8533495581263133, + "learning_rate": 2.8904960229355072e-08, + "loss": 0.3361, + "step": 43921 + }, + { + "epoch": 0.9668060752806515, + "grad_norm": 0.6056096330227876, + "learning_rate": 2.886669973006018e-08, + "loss": 0.4351, + "step": 43922 + }, + { + "epoch": 0.9668280871670702, + "grad_norm": 0.6257431420083336, + "learning_rate": 2.8828464496281627e-08, + "loss": 0.4451, + "step": 43923 + }, + { + "epoch": 0.9668500990534888, + "grad_norm": 1.5457205843729696, + "learning_rate": 2.879025452821427e-08, + "loss": 0.2801, + "step": 43924 + }, + { + "epoch": 0.9668721109399075, + "grad_norm": 0.6443177168218586, + "learning_rate": 2.8752069826052388e-08, + "loss": 0.4635, + "step": 43925 + }, + { + "epoch": 0.9668941228263263, + "grad_norm": 0.5991176594773734, + "learning_rate": 2.8713910389989718e-08, + "loss": 0.4317, + "step": 43926 + }, + { + "epoch": 0.9669161347127448, + "grad_norm": 1.7178692077904578, + "learning_rate": 2.8675776220219996e-08, + "loss": 0.3574, + "step": 43927 + }, + { + "epoch": 0.9669381465991636, + "grad_norm": 1.9011681394685325, + "learning_rate": 2.8637667316936956e-08, + "loss": 0.2845, + "step": 43928 + }, + { + "epoch": 0.9669601584855823, + "grad_norm": 1.4505902741946657, + "learning_rate": 2.8599583680334885e-08, + "loss": 0.2811, + "step": 43929 + }, + { + "epoch": 0.9669821703720008, + "grad_norm": 1.3543490441303618, + "learning_rate": 2.8561525310606964e-08, + "loss": 0.2023, + "step": 43930 + }, + { + "epoch": 0.9670041822584196, + "grad_norm": 0.5984612893300731, + "learning_rate": 2.8523492207946924e-08, + "loss": 0.441, + "step": 43931 + }, + { + "epoch": 0.9670261941448383, + "grad_norm": 0.613700040906429, + "learning_rate": 2.848548437254739e-08, + "loss": 0.4251, + "step": 43932 + }, + { + "epoch": 0.9670482060312569, + "grad_norm": 1.7633280137645875, + "learning_rate": 2.8447501804602096e-08, + "loss": 0.3814, + "step": 43933 + }, + { + "epoch": 0.9670702179176756, + "grad_norm": 1.8434345185845578, + "learning_rate": 2.840954450430422e-08, + "loss": 0.2406, + "step": 43934 + }, + { + "epoch": 0.9670922298040943, + "grad_norm": 1.5841201922126449, + "learning_rate": 2.8371612471846386e-08, + "loss": 0.2808, + "step": 43935 + }, + { + "epoch": 0.9671142416905129, + "grad_norm": 4.1301506176929985, + "learning_rate": 2.8333705707421777e-08, + "loss": 0.3012, + "step": 43936 + }, + { + "epoch": 0.9671362535769316, + "grad_norm": 3.5723365571094363, + "learning_rate": 2.82958242112219e-08, + "loss": 0.2802, + "step": 43937 + }, + { + "epoch": 0.9671582654633503, + "grad_norm": 1.8587528163913227, + "learning_rate": 2.8257967983440493e-08, + "loss": 0.3166, + "step": 43938 + }, + { + "epoch": 0.9671802773497689, + "grad_norm": 0.6081227391356944, + "learning_rate": 2.8220137024269623e-08, + "loss": 0.4231, + "step": 43939 + }, + { + "epoch": 0.9672022892361876, + "grad_norm": 1.6548984486872478, + "learning_rate": 2.8182331333901357e-08, + "loss": 0.2677, + "step": 43940 + }, + { + "epoch": 0.9672243011226062, + "grad_norm": 1.5232285032752282, + "learning_rate": 2.8144550912528324e-08, + "loss": 0.2836, + "step": 43941 + }, + { + "epoch": 0.9672463130090249, + "grad_norm": 2.5082135546851583, + "learning_rate": 2.8106795760341477e-08, + "loss": 0.2556, + "step": 43942 + }, + { + "epoch": 0.9672683248954436, + "grad_norm": 1.6529644914821886, + "learning_rate": 2.8069065877533997e-08, + "loss": 0.309, + "step": 43943 + }, + { + "epoch": 0.9672903367818622, + "grad_norm": 1.640201631555718, + "learning_rate": 2.8031361264296842e-08, + "loss": 0.2821, + "step": 43944 + }, + { + "epoch": 0.9673123486682809, + "grad_norm": 2.075173353827226, + "learning_rate": 2.799368192082208e-08, + "loss": 0.2741, + "step": 43945 + }, + { + "epoch": 0.9673343605546996, + "grad_norm": 1.7990733582017235, + "learning_rate": 2.795602784730067e-08, + "loss": 0.2398, + "step": 43946 + }, + { + "epoch": 0.9673563724411182, + "grad_norm": 0.5977962591864466, + "learning_rate": 2.791839904392468e-08, + "loss": 0.4402, + "step": 43947 + }, + { + "epoch": 0.9673783843275369, + "grad_norm": 1.816485093193962, + "learning_rate": 2.7880795510885074e-08, + "loss": 0.2183, + "step": 43948 + }, + { + "epoch": 0.9674003962139556, + "grad_norm": 1.945546182916659, + "learning_rate": 2.7843217248372245e-08, + "loss": 0.2791, + "step": 43949 + }, + { + "epoch": 0.9674224081003742, + "grad_norm": 1.7468606711554648, + "learning_rate": 2.780566425657827e-08, + "loss": 0.2938, + "step": 43950 + }, + { + "epoch": 0.9674444199867929, + "grad_norm": 1.6165230441193401, + "learning_rate": 2.7768136535693545e-08, + "loss": 0.262, + "step": 43951 + }, + { + "epoch": 0.9674664318732116, + "grad_norm": 1.6163187249377162, + "learning_rate": 2.773063408590848e-08, + "loss": 0.2964, + "step": 43952 + }, + { + "epoch": 0.9674884437596302, + "grad_norm": 2.564551963231038, + "learning_rate": 2.7693156907414586e-08, + "loss": 0.3095, + "step": 43953 + }, + { + "epoch": 0.9675104556460489, + "grad_norm": 2.029892199030637, + "learning_rate": 2.765570500040171e-08, + "loss": 0.2441, + "step": 43954 + }, + { + "epoch": 0.9675324675324676, + "grad_norm": 1.4822900081124966, + "learning_rate": 2.7618278365059705e-08, + "loss": 0.2219, + "step": 43955 + }, + { + "epoch": 0.9675544794188862, + "grad_norm": 3.2245200119483473, + "learning_rate": 2.7580877001580075e-08, + "loss": 0.3913, + "step": 43956 + }, + { + "epoch": 0.9675764913053049, + "grad_norm": 1.6325555283164983, + "learning_rate": 2.7543500910151567e-08, + "loss": 0.2308, + "step": 43957 + }, + { + "epoch": 0.9675985031917236, + "grad_norm": 1.698695542920274, + "learning_rate": 2.750615009096569e-08, + "loss": 0.2561, + "step": 43958 + }, + { + "epoch": 0.9676205150781422, + "grad_norm": 1.5457709153454864, + "learning_rate": 2.7468824544210625e-08, + "loss": 0.2996, + "step": 43959 + }, + { + "epoch": 0.9676425269645609, + "grad_norm": 1.7883224625549194, + "learning_rate": 2.743152427007678e-08, + "loss": 0.2601, + "step": 43960 + }, + { + "epoch": 0.9676645388509796, + "grad_norm": 2.138445918670337, + "learning_rate": 2.7394249268754002e-08, + "loss": 0.2637, + "step": 43961 + }, + { + "epoch": 0.9676865507373982, + "grad_norm": 1.9906796839881258, + "learning_rate": 2.735699954043103e-08, + "loss": 0.3403, + "step": 43962 + }, + { + "epoch": 0.9677085626238169, + "grad_norm": 2.1729739046655614, + "learning_rate": 2.7319775085298263e-08, + "loss": 0.3267, + "step": 43963 + }, + { + "epoch": 0.9677305745102355, + "grad_norm": 1.9410343242214598, + "learning_rate": 2.7282575903543886e-08, + "loss": 0.2608, + "step": 43964 + }, + { + "epoch": 0.9677525863966542, + "grad_norm": 6.979580742116842, + "learning_rate": 2.724540199535719e-08, + "loss": 0.281, + "step": 43965 + }, + { + "epoch": 0.9677745982830729, + "grad_norm": 1.8004698182706262, + "learning_rate": 2.7208253360928027e-08, + "loss": 0.3324, + "step": 43966 + }, + { + "epoch": 0.9677966101694915, + "grad_norm": 2.8818923041819406, + "learning_rate": 2.7171130000444025e-08, + "loss": 0.2932, + "step": 43967 + }, + { + "epoch": 0.9678186220559102, + "grad_norm": 0.6198740235816046, + "learning_rate": 2.7134031914093915e-08, + "loss": 0.4563, + "step": 43968 + }, + { + "epoch": 0.9678406339423289, + "grad_norm": 2.3836769248708487, + "learning_rate": 2.7096959102066998e-08, + "loss": 0.2172, + "step": 43969 + }, + { + "epoch": 0.9678626458287475, + "grad_norm": 1.9146079577008155, + "learning_rate": 2.70599115645509e-08, + "loss": 0.3386, + "step": 43970 + }, + { + "epoch": 0.9678846577151662, + "grad_norm": 1.8227911713032252, + "learning_rate": 2.702288930173491e-08, + "loss": 0.2167, + "step": 43971 + }, + { + "epoch": 0.9679066696015849, + "grad_norm": 1.845404431530609, + "learning_rate": 2.698589231380666e-08, + "loss": 0.2572, + "step": 43972 + }, + { + "epoch": 0.9679286814880035, + "grad_norm": 1.5800618656072538, + "learning_rate": 2.6948920600953775e-08, + "loss": 0.2801, + "step": 43973 + }, + { + "epoch": 0.9679506933744222, + "grad_norm": 1.9871870930024125, + "learning_rate": 2.691197416336444e-08, + "loss": 0.2973, + "step": 43974 + }, + { + "epoch": 0.9679727052608409, + "grad_norm": 1.7854749025529775, + "learning_rate": 2.687505300122628e-08, + "loss": 0.2782, + "step": 43975 + }, + { + "epoch": 0.9679947171472595, + "grad_norm": 1.4324034764859725, + "learning_rate": 2.6838157114728037e-08, + "loss": 0.3344, + "step": 43976 + }, + { + "epoch": 0.9680167290336782, + "grad_norm": 1.4248463509198153, + "learning_rate": 2.6801286504055666e-08, + "loss": 0.2682, + "step": 43977 + }, + { + "epoch": 0.9680387409200969, + "grad_norm": 1.8466364186696922, + "learning_rate": 2.676444116939736e-08, + "loss": 0.241, + "step": 43978 + }, + { + "epoch": 0.9680607528065155, + "grad_norm": 1.6932944494308875, + "learning_rate": 2.6727621110940737e-08, + "loss": 0.31, + "step": 43979 + }, + { + "epoch": 0.9680827646929342, + "grad_norm": 2.6392984763020126, + "learning_rate": 2.6690826328871766e-08, + "loss": 0.2833, + "step": 43980 + }, + { + "epoch": 0.9681047765793529, + "grad_norm": 0.6107011511521662, + "learning_rate": 2.6654056823378625e-08, + "loss": 0.4179, + "step": 43981 + }, + { + "epoch": 0.9681267884657715, + "grad_norm": 0.635484761535271, + "learning_rate": 2.6617312594647837e-08, + "loss": 0.4533, + "step": 43982 + }, + { + "epoch": 0.9681488003521902, + "grad_norm": 2.139271862776227, + "learning_rate": 2.6580593642865917e-08, + "loss": 0.2564, + "step": 43983 + }, + { + "epoch": 0.9681708122386089, + "grad_norm": 1.8612257427111134, + "learning_rate": 2.6543899968219933e-08, + "loss": 0.2548, + "step": 43984 + }, + { + "epoch": 0.9681928241250275, + "grad_norm": 2.016155061858102, + "learning_rate": 2.6507231570895298e-08, + "loss": 0.2833, + "step": 43985 + }, + { + "epoch": 0.9682148360114462, + "grad_norm": 1.5135202733647033, + "learning_rate": 2.6470588451079637e-08, + "loss": 0.2257, + "step": 43986 + }, + { + "epoch": 0.9682368478978648, + "grad_norm": 1.7560766326365251, + "learning_rate": 2.6433970608958913e-08, + "loss": 0.271, + "step": 43987 + }, + { + "epoch": 0.9682588597842835, + "grad_norm": 1.886414915477139, + "learning_rate": 2.6397378044718535e-08, + "loss": 0.2926, + "step": 43988 + }, + { + "epoch": 0.9682808716707022, + "grad_norm": 1.6231686685558153, + "learning_rate": 2.6360810758545575e-08, + "loss": 0.3091, + "step": 43989 + }, + { + "epoch": 0.9683028835571208, + "grad_norm": 2.288595683062323, + "learning_rate": 2.6324268750624882e-08, + "loss": 0.2038, + "step": 43990 + }, + { + "epoch": 0.9683248954435395, + "grad_norm": 0.6131382407840181, + "learning_rate": 2.628775202114242e-08, + "loss": 0.4461, + "step": 43991 + }, + { + "epoch": 0.9683469073299582, + "grad_norm": 1.6764448846965114, + "learning_rate": 2.6251260570284154e-08, + "loss": 0.2861, + "step": 43992 + }, + { + "epoch": 0.9683689192163768, + "grad_norm": 2.0621928582927525, + "learning_rate": 2.621479439823549e-08, + "loss": 0.3314, + "step": 43993 + }, + { + "epoch": 0.9683909311027955, + "grad_norm": 1.94427245437921, + "learning_rate": 2.6178353505181274e-08, + "loss": 0.311, + "step": 43994 + }, + { + "epoch": 0.9684129429892142, + "grad_norm": 2.2432868309461393, + "learning_rate": 2.614193789130748e-08, + "loss": 0.3338, + "step": 43995 + }, + { + "epoch": 0.9684349548756328, + "grad_norm": 1.5995225301855225, + "learning_rate": 2.6105547556798395e-08, + "loss": 0.3015, + "step": 43996 + }, + { + "epoch": 0.9684569667620515, + "grad_norm": 2.70161381427815, + "learning_rate": 2.6069182501839428e-08, + "loss": 0.3865, + "step": 43997 + }, + { + "epoch": 0.9684789786484702, + "grad_norm": 0.6339359195916828, + "learning_rate": 2.6032842726615436e-08, + "loss": 0.4277, + "step": 43998 + }, + { + "epoch": 0.9685009905348888, + "grad_norm": 2.082920180349143, + "learning_rate": 2.5996528231310715e-08, + "loss": 0.3689, + "step": 43999 + }, + { + "epoch": 0.9685230024213075, + "grad_norm": 1.4596843587364208, + "learning_rate": 2.5960239016110112e-08, + "loss": 0.1833, + "step": 44000 + }, + { + "epoch": 0.9685450143077262, + "grad_norm": 1.456952547473698, + "learning_rate": 2.5923975081198484e-08, + "loss": 0.2666, + "step": 44001 + }, + { + "epoch": 0.9685670261941448, + "grad_norm": 1.4249755569735116, + "learning_rate": 2.588773642675957e-08, + "loss": 0.2552, + "step": 44002 + }, + { + "epoch": 0.9685890380805635, + "grad_norm": 1.4126803464521291, + "learning_rate": 2.5851523052977668e-08, + "loss": 0.2057, + "step": 44003 + }, + { + "epoch": 0.9686110499669822, + "grad_norm": 1.653403939564122, + "learning_rate": 2.581533496003652e-08, + "loss": 0.2834, + "step": 44004 + }, + { + "epoch": 0.9686330618534008, + "grad_norm": 1.9109431994878292, + "learning_rate": 2.5779172148120424e-08, + "loss": 0.2487, + "step": 44005 + }, + { + "epoch": 0.9686550737398195, + "grad_norm": 22.91376320805323, + "learning_rate": 2.574303461741312e-08, + "loss": 0.2648, + "step": 44006 + }, + { + "epoch": 0.9686770856262382, + "grad_norm": 1.6060423590551904, + "learning_rate": 2.5706922368098908e-08, + "loss": 0.3211, + "step": 44007 + }, + { + "epoch": 0.9686990975126568, + "grad_norm": 2.5421312645284813, + "learning_rate": 2.5670835400360415e-08, + "loss": 0.2934, + "step": 44008 + }, + { + "epoch": 0.9687211093990755, + "grad_norm": 2.3978654972801157, + "learning_rate": 2.563477371438139e-08, + "loss": 0.3679, + "step": 44009 + }, + { + "epoch": 0.9687431212854941, + "grad_norm": 1.6137213424044061, + "learning_rate": 2.5598737310344456e-08, + "loss": 0.2764, + "step": 44010 + }, + { + "epoch": 0.9687651331719128, + "grad_norm": 2.312426928569742, + "learning_rate": 2.5562726188433918e-08, + "loss": 0.3976, + "step": 44011 + }, + { + "epoch": 0.9687871450583315, + "grad_norm": 1.6704107696094714, + "learning_rate": 2.55267403488324e-08, + "loss": 0.3131, + "step": 44012 + }, + { + "epoch": 0.9688091569447501, + "grad_norm": 1.5332139143652725, + "learning_rate": 2.549077979172254e-08, + "loss": 0.2852, + "step": 44013 + }, + { + "epoch": 0.9688311688311688, + "grad_norm": 9.384244326861795, + "learning_rate": 2.545484451728697e-08, + "loss": 0.3119, + "step": 44014 + }, + { + "epoch": 0.9688531807175875, + "grad_norm": 0.6344746935417874, + "learning_rate": 2.541893452570887e-08, + "loss": 0.4522, + "step": 44015 + }, + { + "epoch": 0.9688751926040061, + "grad_norm": 1.7386151178031952, + "learning_rate": 2.5383049817170323e-08, + "loss": 0.3061, + "step": 44016 + }, + { + "epoch": 0.9688972044904248, + "grad_norm": 1.8039191486817643, + "learning_rate": 2.5347190391853958e-08, + "loss": 0.2693, + "step": 44017 + }, + { + "epoch": 0.9689192163768435, + "grad_norm": 1.4621931271989392, + "learning_rate": 2.5311356249941855e-08, + "loss": 0.3014, + "step": 44018 + }, + { + "epoch": 0.9689412282632621, + "grad_norm": 1.421129224631089, + "learning_rate": 2.5275547391616083e-08, + "loss": 0.2716, + "step": 44019 + }, + { + "epoch": 0.9689632401496808, + "grad_norm": 1.9177537580160695, + "learning_rate": 2.523976381705928e-08, + "loss": 0.2828, + "step": 44020 + }, + { + "epoch": 0.9689852520360995, + "grad_norm": 2.5581050521994473, + "learning_rate": 2.5204005526452414e-08, + "loss": 0.3149, + "step": 44021 + }, + { + "epoch": 0.9690072639225181, + "grad_norm": 1.6472421221340985, + "learning_rate": 2.5168272519977553e-08, + "loss": 0.2898, + "step": 44022 + }, + { + "epoch": 0.9690292758089368, + "grad_norm": 1.7760344092751112, + "learning_rate": 2.5132564797816784e-08, + "loss": 0.3341, + "step": 44023 + }, + { + "epoch": 0.9690512876953555, + "grad_norm": 1.7905003578765155, + "learning_rate": 2.509688236015051e-08, + "loss": 0.307, + "step": 44024 + }, + { + "epoch": 0.9690732995817741, + "grad_norm": 0.6169610241516512, + "learning_rate": 2.5061225207161367e-08, + "loss": 0.4458, + "step": 44025 + }, + { + "epoch": 0.9690953114681928, + "grad_norm": 1.4665208957683467, + "learning_rate": 2.5025593339029762e-08, + "loss": 0.2378, + "step": 44026 + }, + { + "epoch": 0.9691173233546115, + "grad_norm": 1.773220028072376, + "learning_rate": 2.4989986755936668e-08, + "loss": 0.2719, + "step": 44027 + }, + { + "epoch": 0.9691393352410301, + "grad_norm": 0.6270321035626666, + "learning_rate": 2.4954405458063602e-08, + "loss": 0.4646, + "step": 44028 + }, + { + "epoch": 0.9691613471274488, + "grad_norm": 2.4384493816233506, + "learning_rate": 2.491884944559153e-08, + "loss": 0.2983, + "step": 44029 + }, + { + "epoch": 0.9691833590138675, + "grad_norm": 3.284730132500857, + "learning_rate": 2.488331871870031e-08, + "loss": 0.3359, + "step": 44030 + }, + { + "epoch": 0.9692053709002861, + "grad_norm": 2.0227522994113842, + "learning_rate": 2.484781327757091e-08, + "loss": 0.3154, + "step": 44031 + }, + { + "epoch": 0.9692273827867048, + "grad_norm": 1.513896005314106, + "learning_rate": 2.481233312238429e-08, + "loss": 0.2361, + "step": 44032 + }, + { + "epoch": 0.9692493946731234, + "grad_norm": 3.5564476430394714, + "learning_rate": 2.4776878253320312e-08, + "loss": 0.2497, + "step": 44033 + }, + { + "epoch": 0.9692714065595421, + "grad_norm": 2.0868974579973028, + "learning_rate": 2.4741448670559387e-08, + "loss": 0.2826, + "step": 44034 + }, + { + "epoch": 0.9692934184459608, + "grad_norm": 2.240340737944931, + "learning_rate": 2.4706044374281367e-08, + "loss": 0.2399, + "step": 44035 + }, + { + "epoch": 0.9693154303323794, + "grad_norm": 0.6320648126258523, + "learning_rate": 2.4670665364666112e-08, + "loss": 0.4384, + "step": 44036 + }, + { + "epoch": 0.9693374422187981, + "grad_norm": 2.6739841645085525, + "learning_rate": 2.4635311641893477e-08, + "loss": 0.2681, + "step": 44037 + }, + { + "epoch": 0.9693594541052168, + "grad_norm": 2.5142636999297183, + "learning_rate": 2.4599983206143874e-08, + "loss": 0.2583, + "step": 44038 + }, + { + "epoch": 0.9693814659916354, + "grad_norm": 2.210307564995864, + "learning_rate": 2.456468005759549e-08, + "loss": 0.2673, + "step": 44039 + }, + { + "epoch": 0.9694034778780541, + "grad_norm": 2.3073623202253213, + "learning_rate": 2.4529402196429296e-08, + "loss": 0.3264, + "step": 44040 + }, + { + "epoch": 0.9694254897644728, + "grad_norm": 2.729112831793459, + "learning_rate": 2.4494149622822928e-08, + "loss": 0.2631, + "step": 44041 + }, + { + "epoch": 0.9694475016508914, + "grad_norm": 1.4057545397734366, + "learning_rate": 2.4458922336956793e-08, + "loss": 0.2446, + "step": 44042 + }, + { + "epoch": 0.9694695135373101, + "grad_norm": 1.6165235906405369, + "learning_rate": 2.4423720339010194e-08, + "loss": 0.2936, + "step": 44043 + }, + { + "epoch": 0.9694915254237289, + "grad_norm": 1.9719992806558087, + "learning_rate": 2.4388543629160765e-08, + "loss": 0.2232, + "step": 44044 + }, + { + "epoch": 0.9695135373101474, + "grad_norm": 4.041048265400013, + "learning_rate": 2.4353392207587813e-08, + "loss": 0.3393, + "step": 44045 + }, + { + "epoch": 0.9695355491965661, + "grad_norm": 1.6337501028150758, + "learning_rate": 2.4318266074470632e-08, + "loss": 0.2354, + "step": 44046 + }, + { + "epoch": 0.9695575610829849, + "grad_norm": 2.88634589281851, + "learning_rate": 2.4283165229986862e-08, + "loss": 0.311, + "step": 44047 + }, + { + "epoch": 0.9695795729694034, + "grad_norm": 0.6173800862228017, + "learning_rate": 2.4248089674315245e-08, + "loss": 0.4508, + "step": 44048 + }, + { + "epoch": 0.9696015848558222, + "grad_norm": 1.8638523687860724, + "learning_rate": 2.4213039407634532e-08, + "loss": 0.3147, + "step": 44049 + }, + { + "epoch": 0.9696235967422409, + "grad_norm": 1.737346959151741, + "learning_rate": 2.41780144301218e-08, + "loss": 0.2678, + "step": 44050 + }, + { + "epoch": 0.9696456086286595, + "grad_norm": 1.524097707581524, + "learning_rate": 2.414301474195635e-08, + "loss": 0.2587, + "step": 44051 + }, + { + "epoch": 0.9696676205150782, + "grad_norm": 2.36304060388301, + "learning_rate": 2.410804034331471e-08, + "loss": 0.2385, + "step": 44052 + }, + { + "epoch": 0.9696896324014969, + "grad_norm": 0.5933476903727755, + "learning_rate": 2.4073091234375624e-08, + "loss": 0.4268, + "step": 44053 + }, + { + "epoch": 0.9697116442879155, + "grad_norm": 1.244556184791679, + "learning_rate": 2.4038167415316725e-08, + "loss": 0.2348, + "step": 44054 + }, + { + "epoch": 0.9697336561743342, + "grad_norm": 1.7557851933224866, + "learning_rate": 2.4003268886314546e-08, + "loss": 0.3218, + "step": 44055 + }, + { + "epoch": 0.9697556680607528, + "grad_norm": 2.1038582889471678, + "learning_rate": 2.3968395647547828e-08, + "loss": 0.2664, + "step": 44056 + }, + { + "epoch": 0.9697776799471715, + "grad_norm": 0.5996321286394787, + "learning_rate": 2.393354769919254e-08, + "loss": 0.4403, + "step": 44057 + }, + { + "epoch": 0.9697996918335902, + "grad_norm": 0.5921617003157248, + "learning_rate": 2.389872504142632e-08, + "loss": 0.4398, + "step": 44058 + }, + { + "epoch": 0.9698217037200088, + "grad_norm": 1.8202128507868434, + "learning_rate": 2.38639276744268e-08, + "loss": 0.1895, + "step": 44059 + }, + { + "epoch": 0.9698437156064275, + "grad_norm": 0.6426862695153519, + "learning_rate": 2.382915559836996e-08, + "loss": 0.4634, + "step": 44060 + }, + { + "epoch": 0.9698657274928462, + "grad_norm": 1.422553465757152, + "learning_rate": 2.3794408813432314e-08, + "loss": 0.2507, + "step": 44061 + }, + { + "epoch": 0.9698877393792648, + "grad_norm": 2.4131770279134126, + "learning_rate": 2.37596873197915e-08, + "loss": 0.3445, + "step": 44062 + }, + { + "epoch": 0.9699097512656835, + "grad_norm": 2.08603476337044, + "learning_rate": 2.3724991117622942e-08, + "loss": 0.2854, + "step": 44063 + }, + { + "epoch": 0.9699317631521022, + "grad_norm": 1.8891749999003804, + "learning_rate": 2.3690320207103713e-08, + "loss": 0.2774, + "step": 44064 + }, + { + "epoch": 0.9699537750385208, + "grad_norm": 1.7439354644841576, + "learning_rate": 2.3655674588409783e-08, + "loss": 0.3057, + "step": 44065 + }, + { + "epoch": 0.9699757869249395, + "grad_norm": 1.65750045212283, + "learning_rate": 2.3621054261717124e-08, + "loss": 0.2888, + "step": 44066 + }, + { + "epoch": 0.9699977988113582, + "grad_norm": 0.6403104998545385, + "learning_rate": 2.3586459227202262e-08, + "loss": 0.4369, + "step": 44067 + }, + { + "epoch": 0.9700198106977768, + "grad_norm": 1.6430881148004528, + "learning_rate": 2.355188948504006e-08, + "loss": 0.3262, + "step": 44068 + }, + { + "epoch": 0.9700418225841955, + "grad_norm": 1.8471450939427083, + "learning_rate": 2.3517345035407035e-08, + "loss": 0.3563, + "step": 44069 + }, + { + "epoch": 0.9700638344706142, + "grad_norm": 0.6182866968616488, + "learning_rate": 2.3482825878478056e-08, + "loss": 0.4486, + "step": 44070 + }, + { + "epoch": 0.9700858463570328, + "grad_norm": 2.434105783731318, + "learning_rate": 2.3448332014429086e-08, + "loss": 0.2805, + "step": 44071 + }, + { + "epoch": 0.9701078582434515, + "grad_norm": 1.4823205156517698, + "learning_rate": 2.3413863443434992e-08, + "loss": 0.248, + "step": 44072 + }, + { + "epoch": 0.9701298701298702, + "grad_norm": 1.959999876388519, + "learning_rate": 2.337942016567174e-08, + "loss": 0.2954, + "step": 44073 + }, + { + "epoch": 0.9701518820162888, + "grad_norm": 1.5358614900798004, + "learning_rate": 2.334500218131419e-08, + "loss": 0.2869, + "step": 44074 + }, + { + "epoch": 0.9701738939027075, + "grad_norm": 1.870971302595874, + "learning_rate": 2.3310609490536097e-08, + "loss": 0.4064, + "step": 44075 + }, + { + "epoch": 0.9701959057891262, + "grad_norm": 2.058692700651096, + "learning_rate": 2.327624209351398e-08, + "loss": 0.3444, + "step": 44076 + }, + { + "epoch": 0.9702179176755448, + "grad_norm": 1.7264179842000893, + "learning_rate": 2.3241899990421035e-08, + "loss": 0.2796, + "step": 44077 + }, + { + "epoch": 0.9702399295619635, + "grad_norm": 0.6208563875177877, + "learning_rate": 2.3207583181432682e-08, + "loss": 0.4436, + "step": 44078 + }, + { + "epoch": 0.9702619414483822, + "grad_norm": 2.1592906434108077, + "learning_rate": 2.317329166672322e-08, + "loss": 0.2513, + "step": 44079 + }, + { + "epoch": 0.9702839533348008, + "grad_norm": 2.1098461008709455, + "learning_rate": 2.3139025446466955e-08, + "loss": 0.2989, + "step": 44080 + }, + { + "epoch": 0.9703059652212195, + "grad_norm": 1.868726364878552, + "learning_rate": 2.3104784520837642e-08, + "loss": 0.3218, + "step": 44081 + }, + { + "epoch": 0.9703279771076381, + "grad_norm": 1.426273548471791, + "learning_rate": 2.3070568890009582e-08, + "loss": 0.2487, + "step": 44082 + }, + { + "epoch": 0.9703499889940568, + "grad_norm": 2.09623566365538, + "learning_rate": 2.3036378554156526e-08, + "loss": 0.274, + "step": 44083 + }, + { + "epoch": 0.9703720008804755, + "grad_norm": 2.4224243159219676, + "learning_rate": 2.3002213513452776e-08, + "loss": 0.3014, + "step": 44084 + }, + { + "epoch": 0.9703940127668941, + "grad_norm": 1.550877057007334, + "learning_rate": 2.296807376807153e-08, + "loss": 0.3421, + "step": 44085 + }, + { + "epoch": 0.9704160246533128, + "grad_norm": 1.9047844458577154, + "learning_rate": 2.2933959318185983e-08, + "loss": 0.2971, + "step": 44086 + }, + { + "epoch": 0.9704380365397315, + "grad_norm": 1.5538118262445173, + "learning_rate": 2.289987016397044e-08, + "loss": 0.2691, + "step": 44087 + }, + { + "epoch": 0.9704600484261501, + "grad_norm": 1.4776144147858634, + "learning_rate": 2.2865806305596982e-08, + "loss": 0.1792, + "step": 44088 + }, + { + "epoch": 0.9704820603125688, + "grad_norm": 1.8123350657134143, + "learning_rate": 2.283176774323992e-08, + "loss": 0.3141, + "step": 44089 + }, + { + "epoch": 0.9705040721989875, + "grad_norm": 1.8469566059098117, + "learning_rate": 2.279775447707133e-08, + "loss": 0.3097, + "step": 44090 + }, + { + "epoch": 0.9705260840854061, + "grad_norm": 1.9176038317063402, + "learning_rate": 2.2763766507264417e-08, + "loss": 0.2382, + "step": 44091 + }, + { + "epoch": 0.9705480959718248, + "grad_norm": 2.366485936391193, + "learning_rate": 2.272980383399237e-08, + "loss": 0.3447, + "step": 44092 + }, + { + "epoch": 0.9705701078582435, + "grad_norm": 1.5312920218041008, + "learning_rate": 2.2695866457426717e-08, + "loss": 0.313, + "step": 44093 + }, + { + "epoch": 0.9705921197446621, + "grad_norm": 2.269319765624091, + "learning_rate": 2.2661954377741212e-08, + "loss": 0.3469, + "step": 44094 + }, + { + "epoch": 0.9706141316310808, + "grad_norm": 2.6950856604621216, + "learning_rate": 2.2628067595107383e-08, + "loss": 0.3154, + "step": 44095 + }, + { + "epoch": 0.9706361435174995, + "grad_norm": 1.5791920840943745, + "learning_rate": 2.2594206109697314e-08, + "loss": 0.2933, + "step": 44096 + }, + { + "epoch": 0.9706581554039181, + "grad_norm": 1.7560464042592276, + "learning_rate": 2.25603699216842e-08, + "loss": 0.3332, + "step": 44097 + }, + { + "epoch": 0.9706801672903368, + "grad_norm": 2.5379848256358812, + "learning_rate": 2.2526559031239014e-08, + "loss": 0.2798, + "step": 44098 + }, + { + "epoch": 0.9707021791767555, + "grad_norm": 1.2600027617025458, + "learning_rate": 2.2492773438533288e-08, + "loss": 0.2669, + "step": 44099 + }, + { + "epoch": 0.9707241910631741, + "grad_norm": 1.7082528055476882, + "learning_rate": 2.2459013143740214e-08, + "loss": 0.2552, + "step": 44100 + }, + { + "epoch": 0.9707462029495928, + "grad_norm": 3.133328577585117, + "learning_rate": 2.242527814702966e-08, + "loss": 0.2934, + "step": 44101 + }, + { + "epoch": 0.9707682148360115, + "grad_norm": 2.008578859788581, + "learning_rate": 2.239156844857482e-08, + "loss": 0.3388, + "step": 44102 + }, + { + "epoch": 0.9707902267224301, + "grad_norm": 1.798118326544512, + "learning_rate": 2.2357884048545553e-08, + "loss": 0.2948, + "step": 44103 + }, + { + "epoch": 0.9708122386088488, + "grad_norm": 0.6453026037412296, + "learning_rate": 2.2324224947113394e-08, + "loss": 0.4632, + "step": 44104 + }, + { + "epoch": 0.9708342504952674, + "grad_norm": 1.9267969665917755, + "learning_rate": 2.2290591144449868e-08, + "loss": 0.33, + "step": 44105 + }, + { + "epoch": 0.9708562623816861, + "grad_norm": 1.8565181126583459, + "learning_rate": 2.2256982640725956e-08, + "loss": 0.3125, + "step": 44106 + }, + { + "epoch": 0.9708782742681048, + "grad_norm": 0.6081637517253051, + "learning_rate": 2.222339943611207e-08, + "loss": 0.4475, + "step": 44107 + }, + { + "epoch": 0.9709002861545234, + "grad_norm": 1.4867817341526173, + "learning_rate": 2.2189841530778633e-08, + "loss": 0.2474, + "step": 44108 + }, + { + "epoch": 0.9709222980409421, + "grad_norm": 1.6913078419322376, + "learning_rate": 2.215630892489662e-08, + "loss": 0.2463, + "step": 44109 + }, + { + "epoch": 0.9709443099273608, + "grad_norm": 0.5914325525216757, + "learning_rate": 2.2122801618637003e-08, + "loss": 0.4218, + "step": 44110 + }, + { + "epoch": 0.9709663218137794, + "grad_norm": 1.4798660202860805, + "learning_rate": 2.208931961216909e-08, + "loss": 0.3135, + "step": 44111 + }, + { + "epoch": 0.9709883337001981, + "grad_norm": 2.348855694328215, + "learning_rate": 2.2055862905663305e-08, + "loss": 0.2775, + "step": 44112 + }, + { + "epoch": 0.9710103455866168, + "grad_norm": 2.048995538283299, + "learning_rate": 2.2022431499290064e-08, + "loss": 0.2083, + "step": 44113 + }, + { + "epoch": 0.9710323574730354, + "grad_norm": 2.111168190594274, + "learning_rate": 2.198902539321923e-08, + "loss": 0.3437, + "step": 44114 + }, + { + "epoch": 0.9710543693594541, + "grad_norm": 1.561090716370884, + "learning_rate": 2.1955644587620118e-08, + "loss": 0.25, + "step": 44115 + }, + { + "epoch": 0.9710763812458728, + "grad_norm": 1.9404986915462512, + "learning_rate": 2.192228908266314e-08, + "loss": 0.2509, + "step": 44116 + }, + { + "epoch": 0.9710983931322914, + "grad_norm": 1.8171858764300224, + "learning_rate": 2.1888958878517054e-08, + "loss": 0.2423, + "step": 44117 + }, + { + "epoch": 0.9711204050187101, + "grad_norm": 2.0028837021975896, + "learning_rate": 2.185565397535172e-08, + "loss": 0.2911, + "step": 44118 + }, + { + "epoch": 0.9711424169051288, + "grad_norm": 2.072825058811493, + "learning_rate": 2.1822374373335898e-08, + "loss": 0.3361, + "step": 44119 + }, + { + "epoch": 0.9711644287915474, + "grad_norm": 1.4363021881066973, + "learning_rate": 2.1789120072639447e-08, + "loss": 0.2319, + "step": 44120 + }, + { + "epoch": 0.9711864406779661, + "grad_norm": 2.481818897102076, + "learning_rate": 2.1755891073431124e-08, + "loss": 0.3683, + "step": 44121 + }, + { + "epoch": 0.9712084525643848, + "grad_norm": 1.8898030010803608, + "learning_rate": 2.1722687375879127e-08, + "loss": 0.2641, + "step": 44122 + }, + { + "epoch": 0.9712304644508034, + "grad_norm": 2.1347004176168483, + "learning_rate": 2.1689508980153317e-08, + "loss": 0.3058, + "step": 44123 + }, + { + "epoch": 0.9712524763372221, + "grad_norm": 1.2524034996320133, + "learning_rate": 2.1656355886421342e-08, + "loss": 0.2219, + "step": 44124 + }, + { + "epoch": 0.9712744882236408, + "grad_norm": 2.3789929551175395, + "learning_rate": 2.1623228094853067e-08, + "loss": 0.3067, + "step": 44125 + }, + { + "epoch": 0.9712965001100594, + "grad_norm": 1.46631278235499, + "learning_rate": 2.1590125605615574e-08, + "loss": 0.2838, + "step": 44126 + }, + { + "epoch": 0.9713185119964781, + "grad_norm": 1.5449839259617955, + "learning_rate": 2.155704841887707e-08, + "loss": 0.2705, + "step": 44127 + }, + { + "epoch": 0.9713405238828967, + "grad_norm": 0.6081909011883614, + "learning_rate": 2.1523996534806303e-08, + "loss": 0.4458, + "step": 44128 + }, + { + "epoch": 0.9713625357693154, + "grad_norm": 0.6316727313406164, + "learning_rate": 2.149096995357147e-08, + "loss": 0.4532, + "step": 44129 + }, + { + "epoch": 0.9713845476557341, + "grad_norm": 0.619448247345888, + "learning_rate": 2.1457968675339113e-08, + "loss": 0.4552, + "step": 44130 + }, + { + "epoch": 0.9714065595421527, + "grad_norm": 1.4403600858611132, + "learning_rate": 2.1424992700278536e-08, + "loss": 0.1848, + "step": 44131 + }, + { + "epoch": 0.9714285714285714, + "grad_norm": 2.015433700023619, + "learning_rate": 2.1392042028556826e-08, + "loss": 0.3157, + "step": 44132 + }, + { + "epoch": 0.9714505833149901, + "grad_norm": 2.6815752810188225, + "learning_rate": 2.1359116660341072e-08, + "loss": 0.3767, + "step": 44133 + }, + { + "epoch": 0.9714725952014087, + "grad_norm": 0.6318921357338646, + "learning_rate": 2.1326216595798367e-08, + "loss": 0.452, + "step": 44134 + }, + { + "epoch": 0.9714946070878274, + "grad_norm": 2.3027246617271393, + "learning_rate": 2.1293341835096903e-08, + "loss": 0.3801, + "step": 44135 + }, + { + "epoch": 0.9715166189742461, + "grad_norm": 2.0789120053181325, + "learning_rate": 2.126049237840322e-08, + "loss": 0.2234, + "step": 44136 + }, + { + "epoch": 0.9715386308606647, + "grad_norm": 0.6139884605286797, + "learning_rate": 2.1227668225883848e-08, + "loss": 0.4486, + "step": 44137 + }, + { + "epoch": 0.9715606427470834, + "grad_norm": 2.005124386587482, + "learning_rate": 2.1194869377706427e-08, + "loss": 0.345, + "step": 44138 + }, + { + "epoch": 0.9715826546335021, + "grad_norm": 4.320828845363436, + "learning_rate": 2.1162095834036946e-08, + "loss": 0.2464, + "step": 44139 + }, + { + "epoch": 0.9716046665199207, + "grad_norm": 1.5788688799092565, + "learning_rate": 2.1129347595042483e-08, + "loss": 0.2346, + "step": 44140 + }, + { + "epoch": 0.9716266784063394, + "grad_norm": 1.799078609352138, + "learning_rate": 2.1096624660889575e-08, + "loss": 0.3363, + "step": 44141 + }, + { + "epoch": 0.9716486902927581, + "grad_norm": 1.7734164416522165, + "learning_rate": 2.1063927031744203e-08, + "loss": 0.2437, + "step": 44142 + }, + { + "epoch": 0.9716707021791767, + "grad_norm": 1.493703803503409, + "learning_rate": 2.1031254707772895e-08, + "loss": 0.3126, + "step": 44143 + }, + { + "epoch": 0.9716927140655954, + "grad_norm": 3.5932666691871646, + "learning_rate": 2.0998607689140525e-08, + "loss": 0.2169, + "step": 44144 + }, + { + "epoch": 0.9717147259520141, + "grad_norm": 1.9530871444267546, + "learning_rate": 2.0965985976014736e-08, + "loss": 0.2436, + "step": 44145 + }, + { + "epoch": 0.9717367378384327, + "grad_norm": 1.9167743264941337, + "learning_rate": 2.093338956856039e-08, + "loss": 0.3044, + "step": 44146 + }, + { + "epoch": 0.9717587497248514, + "grad_norm": 0.6222794970780718, + "learning_rate": 2.0900818466942918e-08, + "loss": 0.4434, + "step": 44147 + }, + { + "epoch": 0.9717807616112701, + "grad_norm": 1.7291940033844226, + "learning_rate": 2.086827267132885e-08, + "loss": 0.3136, + "step": 44148 + }, + { + "epoch": 0.9718027734976887, + "grad_norm": 0.6267194068662997, + "learning_rate": 2.0835752181882497e-08, + "loss": 0.4541, + "step": 44149 + }, + { + "epoch": 0.9718247853841074, + "grad_norm": 0.6242892185825454, + "learning_rate": 2.0803256998769285e-08, + "loss": 0.4575, + "step": 44150 + }, + { + "epoch": 0.971846797270526, + "grad_norm": 2.379055626253851, + "learning_rate": 2.077078712215519e-08, + "loss": 0.2262, + "step": 44151 + }, + { + "epoch": 0.9718688091569447, + "grad_norm": 2.1693374342312386, + "learning_rate": 2.0738342552204528e-08, + "loss": 0.2789, + "step": 44152 + }, + { + "epoch": 0.9718908210433634, + "grad_norm": 1.483963169545612, + "learning_rate": 2.070592328908272e-08, + "loss": 0.2662, + "step": 44153 + }, + { + "epoch": 0.971912832929782, + "grad_norm": 6.434037925131838, + "learning_rate": 2.067352933295408e-08, + "loss": 0.3217, + "step": 44154 + }, + { + "epoch": 0.9719348448162007, + "grad_norm": 1.5576088916673076, + "learning_rate": 2.0641160683983473e-08, + "loss": 0.2519, + "step": 44155 + }, + { + "epoch": 0.9719568567026194, + "grad_norm": 1.6991451653298189, + "learning_rate": 2.060881734233522e-08, + "loss": 0.2708, + "step": 44156 + }, + { + "epoch": 0.971978868589038, + "grad_norm": 2.5094499780855575, + "learning_rate": 2.057649930817418e-08, + "loss": 0.3166, + "step": 44157 + }, + { + "epoch": 0.9720008804754567, + "grad_norm": 1.880172891739645, + "learning_rate": 2.0544206581664117e-08, + "loss": 0.2963, + "step": 44158 + }, + { + "epoch": 0.9720228923618754, + "grad_norm": 1.8608218486200752, + "learning_rate": 2.051193916296934e-08, + "loss": 0.2847, + "step": 44159 + }, + { + "epoch": 0.972044904248294, + "grad_norm": 1.5978059660893353, + "learning_rate": 2.0479697052253057e-08, + "loss": 0.2419, + "step": 44160 + }, + { + "epoch": 0.9720669161347127, + "grad_norm": 1.5320981011573231, + "learning_rate": 2.0447480249680685e-08, + "loss": 0.3478, + "step": 44161 + }, + { + "epoch": 0.9720889280211314, + "grad_norm": 1.4390713225288934, + "learning_rate": 2.0415288755414876e-08, + "loss": 0.2676, + "step": 44162 + }, + { + "epoch": 0.97211093990755, + "grad_norm": 1.6240132444701725, + "learning_rate": 2.0383122569619384e-08, + "loss": 0.2604, + "step": 44163 + }, + { + "epoch": 0.9721329517939687, + "grad_norm": 0.6071566738044939, + "learning_rate": 2.0350981692458525e-08, + "loss": 0.4496, + "step": 44164 + }, + { + "epoch": 0.9721549636803875, + "grad_norm": 0.6100461279284456, + "learning_rate": 2.0318866124094395e-08, + "loss": 0.4323, + "step": 44165 + }, + { + "epoch": 0.972176975566806, + "grad_norm": 1.3209035165499956, + "learning_rate": 2.0286775864690745e-08, + "loss": 0.244, + "step": 44166 + }, + { + "epoch": 0.9721989874532247, + "grad_norm": 0.6172316905595013, + "learning_rate": 2.025471091441078e-08, + "loss": 0.4393, + "step": 44167 + }, + { + "epoch": 0.9722209993396435, + "grad_norm": 1.4957576162602495, + "learning_rate": 2.022267127341715e-08, + "loss": 0.3058, + "step": 44168 + }, + { + "epoch": 0.972243011226062, + "grad_norm": 1.4266793878976847, + "learning_rate": 2.019065694187361e-08, + "loss": 0.2804, + "step": 44169 + }, + { + "epoch": 0.9722650231124808, + "grad_norm": 1.6070468958246846, + "learning_rate": 2.01586679199417e-08, + "loss": 0.2504, + "step": 44170 + }, + { + "epoch": 0.9722870349988995, + "grad_norm": 1.9601672582842997, + "learning_rate": 2.012670420778462e-08, + "loss": 0.2567, + "step": 44171 + }, + { + "epoch": 0.972309046885318, + "grad_norm": 1.810070023770795, + "learning_rate": 2.0094765805564464e-08, + "loss": 0.2223, + "step": 44172 + }, + { + "epoch": 0.9723310587717368, + "grad_norm": 2.3895468967548656, + "learning_rate": 2.0062852713444435e-08, + "loss": 0.2752, + "step": 44173 + }, + { + "epoch": 0.9723530706581553, + "grad_norm": 1.5717435881092758, + "learning_rate": 2.0030964931585517e-08, + "loss": 0.3117, + "step": 44174 + }, + { + "epoch": 0.972375082544574, + "grad_norm": 2.047532935018292, + "learning_rate": 1.999910246015091e-08, + "loss": 0.2978, + "step": 44175 + }, + { + "epoch": 0.9723970944309928, + "grad_norm": 1.54811753088129, + "learning_rate": 1.9967265299301598e-08, + "loss": 0.2391, + "step": 44176 + }, + { + "epoch": 0.9724191063174114, + "grad_norm": 1.6935699633315415, + "learning_rate": 1.9935453449199672e-08, + "loss": 0.3025, + "step": 44177 + }, + { + "epoch": 0.9724411182038301, + "grad_norm": 1.6654156130683373, + "learning_rate": 1.9903666910007223e-08, + "loss": 0.2925, + "step": 44178 + }, + { + "epoch": 0.9724631300902488, + "grad_norm": 1.5338657837644896, + "learning_rate": 1.9871905681885796e-08, + "loss": 0.2534, + "step": 44179 + }, + { + "epoch": 0.9724851419766674, + "grad_norm": 0.618575624085683, + "learning_rate": 1.984016976499581e-08, + "loss": 0.4402, + "step": 44180 + }, + { + "epoch": 0.9725071538630861, + "grad_norm": 1.5348508921745447, + "learning_rate": 1.9808459159499916e-08, + "loss": 0.3372, + "step": 44181 + }, + { + "epoch": 0.9725291657495048, + "grad_norm": 1.630469031948324, + "learning_rate": 1.977677386555854e-08, + "loss": 0.282, + "step": 44182 + }, + { + "epoch": 0.9725511776359234, + "grad_norm": 1.6275746662329909, + "learning_rate": 1.974511388333267e-08, + "loss": 0.2599, + "step": 44183 + }, + { + "epoch": 0.9725731895223421, + "grad_norm": 1.5559988693342892, + "learning_rate": 1.9713479212983278e-08, + "loss": 0.2964, + "step": 44184 + }, + { + "epoch": 0.9725952014087608, + "grad_norm": 1.3939148580141325, + "learning_rate": 1.9681869854671353e-08, + "loss": 0.2446, + "step": 44185 + }, + { + "epoch": 0.9726172132951794, + "grad_norm": 2.1558233532471323, + "learning_rate": 1.9650285808557877e-08, + "loss": 0.2387, + "step": 44186 + }, + { + "epoch": 0.9726392251815981, + "grad_norm": 1.328701910305738, + "learning_rate": 1.9618727074802724e-08, + "loss": 0.1878, + "step": 44187 + }, + { + "epoch": 0.9726612370680168, + "grad_norm": 1.2018691953789953, + "learning_rate": 1.9587193653566318e-08, + "loss": 0.2327, + "step": 44188 + }, + { + "epoch": 0.9726832489544354, + "grad_norm": 1.9532637554079002, + "learning_rate": 1.9555685545009085e-08, + "loss": 0.2126, + "step": 44189 + }, + { + "epoch": 0.9727052608408541, + "grad_norm": 2.10254759140528, + "learning_rate": 1.9524202749291455e-08, + "loss": 0.3069, + "step": 44190 + }, + { + "epoch": 0.9727272727272728, + "grad_norm": 0.6063510497735095, + "learning_rate": 1.9492745266572743e-08, + "loss": 0.4572, + "step": 44191 + }, + { + "epoch": 0.9727492846136914, + "grad_norm": 2.029984104946381, + "learning_rate": 1.946131309701338e-08, + "loss": 0.2929, + "step": 44192 + }, + { + "epoch": 0.9727712965001101, + "grad_norm": 1.557046269160332, + "learning_rate": 1.9429906240773234e-08, + "loss": 0.3138, + "step": 44193 + }, + { + "epoch": 0.9727933083865288, + "grad_norm": 1.9480481320359924, + "learning_rate": 1.9398524698011623e-08, + "loss": 0.3787, + "step": 44194 + }, + { + "epoch": 0.9728153202729474, + "grad_norm": 2.0493232893190694, + "learning_rate": 1.936716846888842e-08, + "loss": 0.2718, + "step": 44195 + }, + { + "epoch": 0.9728373321593661, + "grad_norm": 1.4379314606260316, + "learning_rate": 1.9335837553562388e-08, + "loss": 0.2504, + "step": 44196 + }, + { + "epoch": 0.9728593440457847, + "grad_norm": 1.4268908600313364, + "learning_rate": 1.930453195219284e-08, + "loss": 0.2431, + "step": 44197 + }, + { + "epoch": 0.9728813559322034, + "grad_norm": 1.47751015948118, + "learning_rate": 1.927325166493965e-08, + "loss": 0.2636, + "step": 44198 + }, + { + "epoch": 0.9729033678186221, + "grad_norm": 2.1858751798049365, + "learning_rate": 1.924199669196103e-08, + "loss": 0.2473, + "step": 44199 + }, + { + "epoch": 0.9729253797050407, + "grad_norm": 0.6415059679395121, + "learning_rate": 1.9210767033416288e-08, + "loss": 0.4656, + "step": 44200 + }, + { + "epoch": 0.9729473915914594, + "grad_norm": 1.3641554292293094, + "learning_rate": 1.9179562689463084e-08, + "loss": 0.2178, + "step": 44201 + }, + { + "epoch": 0.9729694034778781, + "grad_norm": 1.4832544339188085, + "learning_rate": 1.914838366026184e-08, + "loss": 0.3218, + "step": 44202 + }, + { + "epoch": 0.9729914153642967, + "grad_norm": 2.2024085252403913, + "learning_rate": 1.911722994596965e-08, + "loss": 0.3071, + "step": 44203 + }, + { + "epoch": 0.9730134272507154, + "grad_norm": 1.8662263704615967, + "learning_rate": 1.9086101546744728e-08, + "loss": 0.2944, + "step": 44204 + }, + { + "epoch": 0.9730354391371341, + "grad_norm": 1.558947353117579, + "learning_rate": 1.9054998462746388e-08, + "loss": 0.2863, + "step": 44205 + }, + { + "epoch": 0.9730574510235527, + "grad_norm": 1.677313438741721, + "learning_rate": 1.902392069413228e-08, + "loss": 0.2097, + "step": 44206 + }, + { + "epoch": 0.9730794629099714, + "grad_norm": 0.6010621487271747, + "learning_rate": 1.8992868241059503e-08, + "loss": 0.4485, + "step": 44207 + }, + { + "epoch": 0.9731014747963901, + "grad_norm": 2.2938875864385535, + "learning_rate": 1.896184110368737e-08, + "loss": 0.3735, + "step": 44208 + }, + { + "epoch": 0.9731234866828087, + "grad_norm": 1.5492097036244359, + "learning_rate": 1.8930839282171875e-08, + "loss": 0.2593, + "step": 44209 + }, + { + "epoch": 0.9731454985692274, + "grad_norm": 1.7197852400585665, + "learning_rate": 1.889986277667233e-08, + "loss": 0.3522, + "step": 44210 + }, + { + "epoch": 0.9731675104556461, + "grad_norm": 1.5504026718588853, + "learning_rate": 1.886891158734472e-08, + "loss": 0.2967, + "step": 44211 + }, + { + "epoch": 0.9731895223420647, + "grad_norm": 4.297744782864734, + "learning_rate": 1.8837985714347253e-08, + "loss": 0.2056, + "step": 44212 + }, + { + "epoch": 0.9732115342284834, + "grad_norm": 1.3829499601663142, + "learning_rate": 1.880708515783647e-08, + "loss": 0.2425, + "step": 44213 + }, + { + "epoch": 0.9732335461149021, + "grad_norm": 1.5602234356281663, + "learning_rate": 1.8776209917970024e-08, + "loss": 0.2052, + "step": 44214 + }, + { + "epoch": 0.9732555580013207, + "grad_norm": 1.198691430783511, + "learning_rate": 1.8745359994904454e-08, + "loss": 0.1993, + "step": 44215 + }, + { + "epoch": 0.9732775698877394, + "grad_norm": 0.5999478682572231, + "learning_rate": 1.871453538879686e-08, + "loss": 0.4381, + "step": 44216 + }, + { + "epoch": 0.9732995817741581, + "grad_norm": 1.8128246335287836, + "learning_rate": 1.8683736099803783e-08, + "loss": 0.3576, + "step": 44217 + }, + { + "epoch": 0.9733215936605767, + "grad_norm": 1.959543759837714, + "learning_rate": 1.8652962128081208e-08, + "loss": 0.2383, + "step": 44218 + }, + { + "epoch": 0.9733436055469954, + "grad_norm": 1.8360262246027343, + "learning_rate": 1.8622213473786233e-08, + "loss": 0.2855, + "step": 44219 + }, + { + "epoch": 0.9733656174334141, + "grad_norm": 1.4346295575047847, + "learning_rate": 1.8591490137075396e-08, + "loss": 0.2258, + "step": 44220 + }, + { + "epoch": 0.9733876293198327, + "grad_norm": 2.377782587002763, + "learning_rate": 1.8560792118103577e-08, + "loss": 0.3453, + "step": 44221 + }, + { + "epoch": 0.9734096412062514, + "grad_norm": 2.145224311817843, + "learning_rate": 1.853011941702787e-08, + "loss": 0.338, + "step": 44222 + }, + { + "epoch": 0.97343165309267, + "grad_norm": 0.6262755187267469, + "learning_rate": 1.849947203400426e-08, + "loss": 0.4715, + "step": 44223 + }, + { + "epoch": 0.9734536649790887, + "grad_norm": 1.9848435087218745, + "learning_rate": 1.8468849969188184e-08, + "loss": 0.2609, + "step": 44224 + }, + { + "epoch": 0.9734756768655074, + "grad_norm": 1.8907041328008176, + "learning_rate": 1.843825322273507e-08, + "loss": 0.2591, + "step": 44225 + }, + { + "epoch": 0.973497688751926, + "grad_norm": 1.6797716971534173, + "learning_rate": 1.8407681794800346e-08, + "loss": 0.3224, + "step": 44226 + }, + { + "epoch": 0.9735197006383447, + "grad_norm": 1.5898910824769557, + "learning_rate": 1.837713568553945e-08, + "loss": 0.2821, + "step": 44227 + }, + { + "epoch": 0.9735417125247634, + "grad_norm": 2.9888245680031735, + "learning_rate": 1.8346614895108362e-08, + "loss": 0.3199, + "step": 44228 + }, + { + "epoch": 0.973563724411182, + "grad_norm": 0.6265573260691147, + "learning_rate": 1.8316119423661404e-08, + "loss": 0.4407, + "step": 44229 + }, + { + "epoch": 0.9735857362976007, + "grad_norm": 2.446484172723229, + "learning_rate": 1.8285649271354012e-08, + "loss": 0.2159, + "step": 44230 + }, + { + "epoch": 0.9736077481840194, + "grad_norm": 1.9835523328929014, + "learning_rate": 1.8255204438340502e-08, + "loss": 0.2634, + "step": 44231 + }, + { + "epoch": 0.973629760070438, + "grad_norm": 2.522618306139346, + "learning_rate": 1.8224784924775752e-08, + "loss": 0.2691, + "step": 44232 + }, + { + "epoch": 0.9736517719568567, + "grad_norm": 0.6020787916191752, + "learning_rate": 1.8194390730815192e-08, + "loss": 0.4244, + "step": 44233 + }, + { + "epoch": 0.9736737838432754, + "grad_norm": 1.561358653903282, + "learning_rate": 1.8164021856612592e-08, + "loss": 0.2635, + "step": 44234 + }, + { + "epoch": 0.973695795729694, + "grad_norm": 1.6147530538398265, + "learning_rate": 1.8133678302322267e-08, + "loss": 0.3315, + "step": 44235 + }, + { + "epoch": 0.9737178076161127, + "grad_norm": 0.6145368567685011, + "learning_rate": 1.810336006809854e-08, + "loss": 0.4317, + "step": 44236 + }, + { + "epoch": 0.9737398195025314, + "grad_norm": 1.9665297615397355, + "learning_rate": 1.8073067154095735e-08, + "loss": 0.3186, + "step": 44237 + }, + { + "epoch": 0.97376183138895, + "grad_norm": 1.710159936362291, + "learning_rate": 1.804279956046706e-08, + "loss": 0.3365, + "step": 44238 + }, + { + "epoch": 0.9737838432753687, + "grad_norm": 1.4728280928205437, + "learning_rate": 1.8012557287367394e-08, + "loss": 0.2618, + "step": 44239 + }, + { + "epoch": 0.9738058551617874, + "grad_norm": 0.6285230035103618, + "learning_rate": 1.7982340334949942e-08, + "loss": 0.4503, + "step": 44240 + }, + { + "epoch": 0.973827867048206, + "grad_norm": 2.36820449976467, + "learning_rate": 1.7952148703368478e-08, + "loss": 0.2755, + "step": 44241 + }, + { + "epoch": 0.9738498789346247, + "grad_norm": 1.659803784702391, + "learning_rate": 1.792198239277565e-08, + "loss": 0.3091, + "step": 44242 + }, + { + "epoch": 0.9738718908210434, + "grad_norm": 0.5861651713903536, + "learning_rate": 1.7891841403325784e-08, + "loss": 0.4285, + "step": 44243 + }, + { + "epoch": 0.973893902707462, + "grad_norm": 2.228290277217935, + "learning_rate": 1.7861725735171533e-08, + "loss": 0.276, + "step": 44244 + }, + { + "epoch": 0.9739159145938807, + "grad_norm": 2.329910157087057, + "learning_rate": 1.7831635388466108e-08, + "loss": 0.2853, + "step": 44245 + }, + { + "epoch": 0.9739379264802993, + "grad_norm": 1.689908494862688, + "learning_rate": 1.7801570363362164e-08, + "loss": 0.2729, + "step": 44246 + }, + { + "epoch": 0.973959938366718, + "grad_norm": 1.9088916413689838, + "learning_rate": 1.7771530660012914e-08, + "loss": 0.3321, + "step": 44247 + }, + { + "epoch": 0.9739819502531367, + "grad_norm": 1.7001135033187944, + "learning_rate": 1.7741516278571013e-08, + "loss": 0.352, + "step": 44248 + }, + { + "epoch": 0.9740039621395553, + "grad_norm": 2.0806946375540205, + "learning_rate": 1.7711527219189116e-08, + "loss": 0.3673, + "step": 44249 + }, + { + "epoch": 0.974025974025974, + "grad_norm": 2.100030296643282, + "learning_rate": 1.768156348201877e-08, + "loss": 0.3617, + "step": 44250 + }, + { + "epoch": 0.9740479859123927, + "grad_norm": 0.6114179527497201, + "learning_rate": 1.7651625067213184e-08, + "loss": 0.4526, + "step": 44251 + }, + { + "epoch": 0.9740699977988113, + "grad_norm": 0.6198816164397548, + "learning_rate": 1.762171197492446e-08, + "loss": 0.4339, + "step": 44252 + }, + { + "epoch": 0.97409200968523, + "grad_norm": 2.232181550068616, + "learning_rate": 1.7591824205304143e-08, + "loss": 0.2995, + "step": 44253 + }, + { + "epoch": 0.9741140215716487, + "grad_norm": 1.4608467481814302, + "learning_rate": 1.7561961758504332e-08, + "loss": 0.2482, + "step": 44254 + }, + { + "epoch": 0.9741360334580673, + "grad_norm": 1.597506798689593, + "learning_rate": 1.7532124634677126e-08, + "loss": 0.3386, + "step": 44255 + }, + { + "epoch": 0.974158045344486, + "grad_norm": 0.6059516636989479, + "learning_rate": 1.7502312833973523e-08, + "loss": 0.4501, + "step": 44256 + }, + { + "epoch": 0.9741800572309047, + "grad_norm": 1.990762015015346, + "learning_rate": 1.7472526356545617e-08, + "loss": 0.2437, + "step": 44257 + }, + { + "epoch": 0.9742020691173233, + "grad_norm": 0.6298205053401575, + "learning_rate": 1.7442765202544954e-08, + "loss": 0.4613, + "step": 44258 + }, + { + "epoch": 0.974224081003742, + "grad_norm": 2.8953632933312425, + "learning_rate": 1.741302937212197e-08, + "loss": 0.3037, + "step": 44259 + }, + { + "epoch": 0.9742460928901607, + "grad_norm": 0.6323622580631396, + "learning_rate": 1.7383318865428213e-08, + "loss": 0.4539, + "step": 44260 + }, + { + "epoch": 0.9742681047765793, + "grad_norm": 6.426107675699444, + "learning_rate": 1.7353633682615223e-08, + "loss": 0.3087, + "step": 44261 + }, + { + "epoch": 0.974290116662998, + "grad_norm": 2.029822068019033, + "learning_rate": 1.732397382383344e-08, + "loss": 0.2832, + "step": 44262 + }, + { + "epoch": 0.9743121285494167, + "grad_norm": 2.3058335187628964, + "learning_rate": 1.7294339289232742e-08, + "loss": 0.3319, + "step": 44263 + }, + { + "epoch": 0.9743341404358353, + "grad_norm": 1.7906589916864573, + "learning_rate": 1.726473007896523e-08, + "loss": 0.3188, + "step": 44264 + }, + { + "epoch": 0.974356152322254, + "grad_norm": 1.6553008362182282, + "learning_rate": 1.723514619318023e-08, + "loss": 0.2465, + "step": 44265 + }, + { + "epoch": 0.9743781642086727, + "grad_norm": 1.5315524900444624, + "learning_rate": 1.7205587632029287e-08, + "loss": 0.25, + "step": 44266 + }, + { + "epoch": 0.9744001760950913, + "grad_norm": 1.8703683553125017, + "learning_rate": 1.7176054395661724e-08, + "loss": 0.3658, + "step": 44267 + }, + { + "epoch": 0.97442218798151, + "grad_norm": 1.4957056349962126, + "learning_rate": 1.7146546484227423e-08, + "loss": 0.288, + "step": 44268 + }, + { + "epoch": 0.9744441998679286, + "grad_norm": 9.401200560963318, + "learning_rate": 1.7117063897877374e-08, + "loss": 0.2562, + "step": 44269 + }, + { + "epoch": 0.9744662117543473, + "grad_norm": 1.5903966529535927, + "learning_rate": 1.7087606636760344e-08, + "loss": 0.2824, + "step": 44270 + }, + { + "epoch": 0.974488223640766, + "grad_norm": 2.8598815104193385, + "learning_rate": 1.7058174701026777e-08, + "loss": 0.2374, + "step": 44271 + }, + { + "epoch": 0.9745102355271846, + "grad_norm": 3.428980469822175, + "learning_rate": 1.7028768090825988e-08, + "loss": 0.2697, + "step": 44272 + }, + { + "epoch": 0.9745322474136033, + "grad_norm": 1.667718028482963, + "learning_rate": 1.6999386806307305e-08, + "loss": 0.2956, + "step": 44273 + }, + { + "epoch": 0.974554259300022, + "grad_norm": 1.9863682850220734, + "learning_rate": 1.6970030847620056e-08, + "loss": 0.3428, + "step": 44274 + }, + { + "epoch": 0.9745762711864406, + "grad_norm": 1.9209717730790232, + "learning_rate": 1.694070021491412e-08, + "loss": 0.2254, + "step": 44275 + }, + { + "epoch": 0.9745982830728593, + "grad_norm": 2.2394549518864504, + "learning_rate": 1.691139490833771e-08, + "loss": 0.2854, + "step": 44276 + }, + { + "epoch": 0.974620294959278, + "grad_norm": 0.5861390464242757, + "learning_rate": 1.6882114928040705e-08, + "loss": 0.4348, + "step": 44277 + }, + { + "epoch": 0.9746423068456966, + "grad_norm": 1.525143988917429, + "learning_rate": 1.6852860274170768e-08, + "loss": 0.307, + "step": 44278 + }, + { + "epoch": 0.9746643187321153, + "grad_norm": 2.932917196010307, + "learning_rate": 1.6823630946876668e-08, + "loss": 0.2852, + "step": 44279 + }, + { + "epoch": 0.974686330618534, + "grad_norm": 4.552872337532928, + "learning_rate": 1.6794426946308284e-08, + "loss": 0.3563, + "step": 44280 + }, + { + "epoch": 0.9747083425049526, + "grad_norm": 1.3414942255814626, + "learning_rate": 1.676524827261272e-08, + "loss": 0.2877, + "step": 44281 + }, + { + "epoch": 0.9747303543913713, + "grad_norm": 0.6260337369774386, + "learning_rate": 1.6736094925939306e-08, + "loss": 0.4583, + "step": 44282 + }, + { + "epoch": 0.97475236627779, + "grad_norm": 2.067129535386614, + "learning_rate": 1.670696690643514e-08, + "loss": 0.2545, + "step": 44283 + }, + { + "epoch": 0.9747743781642086, + "grad_norm": 1.8449605035771777, + "learning_rate": 1.6677864214248995e-08, + "loss": 0.2558, + "step": 44284 + }, + { + "epoch": 0.9747963900506273, + "grad_norm": 1.8981021789766614, + "learning_rate": 1.664878684952853e-08, + "loss": 0.323, + "step": 44285 + }, + { + "epoch": 0.974818401937046, + "grad_norm": 3.173848439228495, + "learning_rate": 1.6619734812421408e-08, + "loss": 0.324, + "step": 44286 + }, + { + "epoch": 0.9748404138234646, + "grad_norm": 3.2804473963063647, + "learning_rate": 1.6590708103075283e-08, + "loss": 0.2941, + "step": 44287 + }, + { + "epoch": 0.9748624257098834, + "grad_norm": 1.7556019579576556, + "learning_rate": 1.656170672163837e-08, + "loss": 0.3397, + "step": 44288 + }, + { + "epoch": 0.9748844375963021, + "grad_norm": 1.6981411489983553, + "learning_rate": 1.6532730668257224e-08, + "loss": 0.3595, + "step": 44289 + }, + { + "epoch": 0.9749064494827206, + "grad_norm": 1.6297285248510178, + "learning_rate": 1.65037799430795e-08, + "loss": 0.2742, + "step": 44290 + }, + { + "epoch": 0.9749284613691394, + "grad_norm": 2.0329070419000366, + "learning_rate": 1.6474854546252305e-08, + "loss": 0.3273, + "step": 44291 + }, + { + "epoch": 0.974950473255558, + "grad_norm": 1.4736892187252981, + "learning_rate": 1.6445954477922742e-08, + "loss": 0.2368, + "step": 44292 + }, + { + "epoch": 0.9749724851419767, + "grad_norm": 1.4987548451048043, + "learning_rate": 1.641707973823736e-08, + "loss": 0.3043, + "step": 44293 + }, + { + "epoch": 0.9749944970283954, + "grad_norm": 1.6342586045991445, + "learning_rate": 1.6388230327342713e-08, + "loss": 0.2654, + "step": 44294 + }, + { + "epoch": 0.975016508914814, + "grad_norm": 1.4944955029321345, + "learning_rate": 1.6359406245386455e-08, + "loss": 0.2364, + "step": 44295 + }, + { + "epoch": 0.9750385208012327, + "grad_norm": 3.2411349502641196, + "learning_rate": 1.6330607492514027e-08, + "loss": 0.3606, + "step": 44296 + }, + { + "epoch": 0.9750605326876514, + "grad_norm": 1.6005052088804947, + "learning_rate": 1.6301834068872536e-08, + "loss": 0.3201, + "step": 44297 + }, + { + "epoch": 0.97508254457407, + "grad_norm": 0.6265583687964582, + "learning_rate": 1.627308597460797e-08, + "loss": 0.449, + "step": 44298 + }, + { + "epoch": 0.9751045564604887, + "grad_norm": 0.5845378755815278, + "learning_rate": 1.6244363209866332e-08, + "loss": 0.4349, + "step": 44299 + }, + { + "epoch": 0.9751265683469074, + "grad_norm": 7.775885149171079, + "learning_rate": 1.621566577479361e-08, + "loss": 0.2295, + "step": 44300 + }, + { + "epoch": 0.975148580233326, + "grad_norm": 1.6714372761714191, + "learning_rate": 1.61869936695358e-08, + "loss": 0.3436, + "step": 44301 + }, + { + "epoch": 0.9751705921197447, + "grad_norm": 1.752049871853931, + "learning_rate": 1.615834689423834e-08, + "loss": 0.2878, + "step": 44302 + }, + { + "epoch": 0.9751926040061634, + "grad_norm": 1.8025027385611014, + "learning_rate": 1.612972544904723e-08, + "loss": 0.2972, + "step": 44303 + }, + { + "epoch": 0.975214615892582, + "grad_norm": 1.9293519842944526, + "learning_rate": 1.6101129334107346e-08, + "loss": 0.3543, + "step": 44304 + }, + { + "epoch": 0.9752366277790007, + "grad_norm": 1.947762845399836, + "learning_rate": 1.6072558549564686e-08, + "loss": 0.267, + "step": 44305 + }, + { + "epoch": 0.9752586396654194, + "grad_norm": 1.634678854248451, + "learning_rate": 1.604401309556414e-08, + "loss": 0.3342, + "step": 44306 + }, + { + "epoch": 0.975280651551838, + "grad_norm": 1.5746050155705447, + "learning_rate": 1.6015492972251134e-08, + "loss": 0.3109, + "step": 44307 + }, + { + "epoch": 0.9753026634382567, + "grad_norm": 1.6472526882926222, + "learning_rate": 1.5986998179770007e-08, + "loss": 0.2938, + "step": 44308 + }, + { + "epoch": 0.9753246753246754, + "grad_norm": 2.0871602778135716, + "learning_rate": 1.595852871826564e-08, + "loss": 0.3197, + "step": 44309 + }, + { + "epoch": 0.975346687211094, + "grad_norm": 4.370605268188052, + "learning_rate": 1.5930084587883476e-08, + "loss": 0.2686, + "step": 44310 + }, + { + "epoch": 0.9753686990975127, + "grad_norm": 0.6185953965982834, + "learning_rate": 1.5901665788767283e-08, + "loss": 0.4791, + "step": 44311 + }, + { + "epoch": 0.9753907109839314, + "grad_norm": 1.6919015847955916, + "learning_rate": 1.5873272321061393e-08, + "loss": 0.2822, + "step": 44312 + }, + { + "epoch": 0.97541272287035, + "grad_norm": 3.514502123964952, + "learning_rate": 1.5844904184911246e-08, + "loss": 0.2921, + "step": 44313 + }, + { + "epoch": 0.9754347347567687, + "grad_norm": 1.7202972601302147, + "learning_rate": 1.581656138046006e-08, + "loss": 0.1944, + "step": 44314 + }, + { + "epoch": 0.9754567466431873, + "grad_norm": 3.2210103303829554, + "learning_rate": 1.5788243907852163e-08, + "loss": 0.2868, + "step": 44315 + }, + { + "epoch": 0.975478758529606, + "grad_norm": 0.6294184676050526, + "learning_rate": 1.5759951767230775e-08, + "loss": 0.4617, + "step": 44316 + }, + { + "epoch": 0.9755007704160247, + "grad_norm": 1.9471491098567162, + "learning_rate": 1.5731684958741333e-08, + "loss": 0.2997, + "step": 44317 + }, + { + "epoch": 0.9755227823024433, + "grad_norm": 1.5341972497193268, + "learning_rate": 1.5703443482525948e-08, + "loss": 0.2643, + "step": 44318 + }, + { + "epoch": 0.975544794188862, + "grad_norm": 1.5755557637203301, + "learning_rate": 1.5675227338728948e-08, + "loss": 0.2993, + "step": 44319 + }, + { + "epoch": 0.9755668060752807, + "grad_norm": 1.817551933022793, + "learning_rate": 1.5647036527493552e-08, + "loss": 0.3086, + "step": 44320 + }, + { + "epoch": 0.9755888179616993, + "grad_norm": 2.523071680624176, + "learning_rate": 1.561887104896298e-08, + "loss": 0.2965, + "step": 44321 + }, + { + "epoch": 0.975610829848118, + "grad_norm": 1.6556366771608422, + "learning_rate": 1.5590730903279893e-08, + "loss": 0.3151, + "step": 44322 + }, + { + "epoch": 0.9756328417345367, + "grad_norm": 1.569522024502225, + "learning_rate": 1.5562616090588624e-08, + "loss": 0.3056, + "step": 44323 + }, + { + "epoch": 0.9756548536209553, + "grad_norm": 1.531178677349078, + "learning_rate": 1.553452661103072e-08, + "loss": 0.2925, + "step": 44324 + }, + { + "epoch": 0.975676865507374, + "grad_norm": 2.401239223921956, + "learning_rate": 1.5506462464749404e-08, + "loss": 0.2668, + "step": 44325 + }, + { + "epoch": 0.9756988773937927, + "grad_norm": 1.6568202116667796, + "learning_rate": 1.5478423651887896e-08, + "loss": 0.3252, + "step": 44326 + }, + { + "epoch": 0.9757208892802113, + "grad_norm": 5.513819302963096, + "learning_rate": 1.545041017258775e-08, + "loss": 0.311, + "step": 44327 + }, + { + "epoch": 0.97574290116663, + "grad_norm": 2.7592037262054494, + "learning_rate": 1.542242202699218e-08, + "loss": 0.3292, + "step": 44328 + }, + { + "epoch": 0.9757649130530487, + "grad_norm": 0.6202931795091111, + "learning_rate": 1.539445921524274e-08, + "loss": 0.4595, + "step": 44329 + }, + { + "epoch": 0.9757869249394673, + "grad_norm": 1.6501938821903395, + "learning_rate": 1.5366521737481545e-08, + "loss": 0.3267, + "step": 44330 + }, + { + "epoch": 0.975808936825886, + "grad_norm": 1.8335019333681168, + "learning_rate": 1.533860959385125e-08, + "loss": 0.3524, + "step": 44331 + }, + { + "epoch": 0.9758309487123047, + "grad_norm": 3.2650484838185263, + "learning_rate": 1.5310722784493416e-08, + "loss": 0.3166, + "step": 44332 + }, + { + "epoch": 0.9758529605987233, + "grad_norm": 1.4927626666167986, + "learning_rate": 1.5282861309550147e-08, + "loss": 0.3078, + "step": 44333 + }, + { + "epoch": 0.975874972485142, + "grad_norm": 2.380450497492746, + "learning_rate": 1.5255025169161884e-08, + "loss": 0.3039, + "step": 44334 + }, + { + "epoch": 0.9758969843715607, + "grad_norm": 1.894919879317413, + "learning_rate": 1.5227214363471297e-08, + "loss": 0.3028, + "step": 44335 + }, + { + "epoch": 0.9759189962579793, + "grad_norm": 2.903783934656664, + "learning_rate": 1.5199428892618828e-08, + "loss": 0.261, + "step": 44336 + }, + { + "epoch": 0.975941008144398, + "grad_norm": 1.7719600869609644, + "learning_rate": 1.5171668756746582e-08, + "loss": 0.3413, + "step": 44337 + }, + { + "epoch": 0.9759630200308166, + "grad_norm": 1.4844370966338634, + "learning_rate": 1.5143933955995006e-08, + "loss": 0.1779, + "step": 44338 + }, + { + "epoch": 0.9759850319172353, + "grad_norm": 1.3801020306317786, + "learning_rate": 1.5116224490505093e-08, + "loss": 0.2087, + "step": 44339 + }, + { + "epoch": 0.976007043803654, + "grad_norm": 2.4797547828737714, + "learning_rate": 1.50885403604184e-08, + "loss": 0.3061, + "step": 44340 + }, + { + "epoch": 0.9760290556900726, + "grad_norm": 1.6421901559211334, + "learning_rate": 1.506088156587482e-08, + "loss": 0.2014, + "step": 44341 + }, + { + "epoch": 0.9760510675764913, + "grad_norm": 1.5149224227876106, + "learning_rate": 1.5033248107014787e-08, + "loss": 0.2232, + "step": 44342 + }, + { + "epoch": 0.97607307946291, + "grad_norm": 1.6206213168869599, + "learning_rate": 1.5005639983979857e-08, + "loss": 0.2171, + "step": 44343 + }, + { + "epoch": 0.9760950913493286, + "grad_norm": 1.7424206273498337, + "learning_rate": 1.4978057196909367e-08, + "loss": 0.2625, + "step": 44344 + }, + { + "epoch": 0.9761171032357473, + "grad_norm": 2.5401172456031924, + "learning_rate": 1.4950499745943203e-08, + "loss": 0.3104, + "step": 44345 + }, + { + "epoch": 0.976139115122166, + "grad_norm": 1.9928474752079548, + "learning_rate": 1.4922967631222363e-08, + "loss": 0.3327, + "step": 44346 + }, + { + "epoch": 0.9761611270085846, + "grad_norm": 1.5852513467390066, + "learning_rate": 1.4895460852886734e-08, + "loss": 0.3319, + "step": 44347 + }, + { + "epoch": 0.9761831388950033, + "grad_norm": 1.418463620094265, + "learning_rate": 1.486797941107565e-08, + "loss": 0.2439, + "step": 44348 + }, + { + "epoch": 0.976205150781422, + "grad_norm": 2.7802494607763624, + "learning_rate": 1.4840523305929e-08, + "loss": 0.3072, + "step": 44349 + }, + { + "epoch": 0.9762271626678406, + "grad_norm": 13.28698172039533, + "learning_rate": 1.481309253758556e-08, + "loss": 0.3204, + "step": 44350 + }, + { + "epoch": 0.9762491745542593, + "grad_norm": 1.8002709693974301, + "learning_rate": 1.478568710618633e-08, + "loss": 0.2941, + "step": 44351 + }, + { + "epoch": 0.976271186440678, + "grad_norm": 0.6217383029764524, + "learning_rate": 1.4758307011868977e-08, + "loss": 0.4533, + "step": 44352 + }, + { + "epoch": 0.9762931983270966, + "grad_norm": 1.7689606799380695, + "learning_rate": 1.4730952254773944e-08, + "loss": 0.2952, + "step": 44353 + }, + { + "epoch": 0.9763152102135153, + "grad_norm": 1.7493830964287602, + "learning_rate": 1.4703622835039455e-08, + "loss": 0.2535, + "step": 44354 + }, + { + "epoch": 0.976337222099934, + "grad_norm": 1.945149107810073, + "learning_rate": 1.467631875280484e-08, + "loss": 0.3296, + "step": 44355 + }, + { + "epoch": 0.9763592339863526, + "grad_norm": 0.6103839512761845, + "learning_rate": 1.4649040008208326e-08, + "loss": 0.4556, + "step": 44356 + }, + { + "epoch": 0.9763812458727713, + "grad_norm": 2.3865952150496863, + "learning_rate": 1.4621786601389243e-08, + "loss": 0.3061, + "step": 44357 + }, + { + "epoch": 0.97640325775919, + "grad_norm": 2.1620522369661264, + "learning_rate": 1.4594558532485814e-08, + "loss": 0.3804, + "step": 44358 + }, + { + "epoch": 0.9764252696456086, + "grad_norm": 2.2425672754122226, + "learning_rate": 1.4567355801636262e-08, + "loss": 0.261, + "step": 44359 + }, + { + "epoch": 0.9764472815320273, + "grad_norm": 1.5573562376392127, + "learning_rate": 1.4540178408978811e-08, + "loss": 0.3016, + "step": 44360 + }, + { + "epoch": 0.976469293418446, + "grad_norm": 1.9981921014344912, + "learning_rate": 1.4513026354651683e-08, + "loss": 0.266, + "step": 44361 + }, + { + "epoch": 0.9764913053048646, + "grad_norm": 1.7671430832204338, + "learning_rate": 1.4485899638793655e-08, + "loss": 0.2509, + "step": 44362 + }, + { + "epoch": 0.9765133171912833, + "grad_norm": 2.240107074201444, + "learning_rate": 1.4458798261541285e-08, + "loss": 0.3163, + "step": 44363 + }, + { + "epoch": 0.9765353290777019, + "grad_norm": 0.6280425617679668, + "learning_rate": 1.4431722223032795e-08, + "loss": 0.4445, + "step": 44364 + }, + { + "epoch": 0.9765573409641206, + "grad_norm": 2.0969060771008183, + "learning_rate": 1.440467152340641e-08, + "loss": 0.3264, + "step": 44365 + }, + { + "epoch": 0.9765793528505393, + "grad_norm": 1.7604974286505055, + "learning_rate": 1.4377646162798686e-08, + "loss": 0.236, + "step": 44366 + }, + { + "epoch": 0.9766013647369579, + "grad_norm": 1.6016036070147144, + "learning_rate": 1.4350646141347846e-08, + "loss": 0.3012, + "step": 44367 + }, + { + "epoch": 0.9766233766233766, + "grad_norm": 1.4017587116778405, + "learning_rate": 1.4323671459190447e-08, + "loss": 0.2564, + "step": 44368 + }, + { + "epoch": 0.9766453885097953, + "grad_norm": 1.4417284137666118, + "learning_rate": 1.4296722116463602e-08, + "loss": 0.3138, + "step": 44369 + }, + { + "epoch": 0.9766674003962139, + "grad_norm": 1.7019903945904076, + "learning_rate": 1.4269798113304422e-08, + "loss": 0.2928, + "step": 44370 + }, + { + "epoch": 0.9766894122826326, + "grad_norm": 2.027305478073056, + "learning_rate": 1.4242899449850577e-08, + "loss": 0.2919, + "step": 44371 + }, + { + "epoch": 0.9767114241690513, + "grad_norm": 3.251750797451051, + "learning_rate": 1.4216026126237514e-08, + "loss": 0.3755, + "step": 44372 + }, + { + "epoch": 0.9767334360554699, + "grad_norm": 0.6193939681041006, + "learning_rate": 1.4189178142602344e-08, + "loss": 0.4494, + "step": 44373 + }, + { + "epoch": 0.9767554479418886, + "grad_norm": 1.8794376369375867, + "learning_rate": 1.4162355499081627e-08, + "loss": 0.2557, + "step": 44374 + }, + { + "epoch": 0.9767774598283073, + "grad_norm": 0.6305114492576139, + "learning_rate": 1.413555819581136e-08, + "loss": 0.4684, + "step": 44375 + }, + { + "epoch": 0.9767994717147259, + "grad_norm": 1.7071139045998516, + "learning_rate": 1.4108786232928107e-08, + "loss": 0.2398, + "step": 44376 + }, + { + "epoch": 0.9768214836011446, + "grad_norm": 1.4953479109366032, + "learning_rate": 1.4082039610567865e-08, + "loss": 0.2306, + "step": 44377 + }, + { + "epoch": 0.9768434954875633, + "grad_norm": 1.9227165751653263, + "learning_rate": 1.4055318328866085e-08, + "loss": 0.2391, + "step": 44378 + }, + { + "epoch": 0.9768655073739819, + "grad_norm": 4.030842457464166, + "learning_rate": 1.4028622387959322e-08, + "loss": 0.3077, + "step": 44379 + }, + { + "epoch": 0.9768875192604006, + "grad_norm": 0.6207582601367264, + "learning_rate": 1.400195178798247e-08, + "loss": 0.4425, + "step": 44380 + }, + { + "epoch": 0.9769095311468193, + "grad_norm": 1.5865038521003767, + "learning_rate": 1.3975306529072086e-08, + "loss": 0.2513, + "step": 44381 + }, + { + "epoch": 0.9769315430332379, + "grad_norm": 1.6357117671383503, + "learning_rate": 1.394868661136306e-08, + "loss": 0.3804, + "step": 44382 + }, + { + "epoch": 0.9769535549196566, + "grad_norm": 2.174980469990374, + "learning_rate": 1.3922092034990286e-08, + "loss": 0.2785, + "step": 44383 + }, + { + "epoch": 0.9769755668060753, + "grad_norm": 2.006457653173134, + "learning_rate": 1.3895522800089212e-08, + "loss": 0.2904, + "step": 44384 + }, + { + "epoch": 0.9769975786924939, + "grad_norm": 2.6330627580360173, + "learning_rate": 1.3868978906795284e-08, + "loss": 0.3654, + "step": 44385 + }, + { + "epoch": 0.9770195905789126, + "grad_norm": 1.7469446125285595, + "learning_rate": 1.3842460355242837e-08, + "loss": 0.2416, + "step": 44386 + }, + { + "epoch": 0.9770416024653312, + "grad_norm": 2.7998833323674854, + "learning_rate": 1.3815967145566767e-08, + "loss": 0.2943, + "step": 44387 + }, + { + "epoch": 0.9770636143517499, + "grad_norm": 2.223118499142971, + "learning_rate": 1.3789499277901963e-08, + "loss": 0.2672, + "step": 44388 + }, + { + "epoch": 0.9770856262381686, + "grad_norm": 1.9875407761608757, + "learning_rate": 1.3763056752383319e-08, + "loss": 0.2516, + "step": 44389 + }, + { + "epoch": 0.9771076381245872, + "grad_norm": 1.8521370527390522, + "learning_rate": 1.3736639569144616e-08, + "loss": 0.2349, + "step": 44390 + }, + { + "epoch": 0.9771296500110059, + "grad_norm": 1.9299131572421582, + "learning_rate": 1.3710247728319636e-08, + "loss": 0.2827, + "step": 44391 + }, + { + "epoch": 0.9771516618974246, + "grad_norm": 1.3997588537494454, + "learning_rate": 1.3683881230043271e-08, + "loss": 0.1804, + "step": 44392 + }, + { + "epoch": 0.9771736737838432, + "grad_norm": 2.0280578502238558, + "learning_rate": 1.365754007444986e-08, + "loss": 0.4023, + "step": 44393 + }, + { + "epoch": 0.9771956856702619, + "grad_norm": 2.070945255650398, + "learning_rate": 1.3631224261672626e-08, + "loss": 0.3055, + "step": 44394 + }, + { + "epoch": 0.9772176975566806, + "grad_norm": 2.631078181963751, + "learning_rate": 1.3604933791845355e-08, + "loss": 0.3021, + "step": 44395 + }, + { + "epoch": 0.9772397094430992, + "grad_norm": 1.864380602770161, + "learning_rate": 1.3578668665101825e-08, + "loss": 0.2194, + "step": 44396 + }, + { + "epoch": 0.9772617213295179, + "grad_norm": 2.094921134773249, + "learning_rate": 1.3552428881575818e-08, + "loss": 0.3363, + "step": 44397 + }, + { + "epoch": 0.9772837332159366, + "grad_norm": 1.9538294871918709, + "learning_rate": 1.352621444140001e-08, + "loss": 0.3422, + "step": 44398 + }, + { + "epoch": 0.9773057451023552, + "grad_norm": 1.9353023898452129, + "learning_rate": 1.3500025344708179e-08, + "loss": 0.2598, + "step": 44399 + }, + { + "epoch": 0.9773277569887739, + "grad_norm": 3.8909505815761, + "learning_rate": 1.3473861591632998e-08, + "loss": 0.317, + "step": 44400 + }, + { + "epoch": 0.9773497688751926, + "grad_norm": 3.142256882015186, + "learning_rate": 1.3447723182307692e-08, + "loss": 0.3259, + "step": 44401 + }, + { + "epoch": 0.9773717807616112, + "grad_norm": 1.485524747742627, + "learning_rate": 1.342161011686549e-08, + "loss": 0.3073, + "step": 44402 + }, + { + "epoch": 0.97739379264803, + "grad_norm": 2.169210041192089, + "learning_rate": 1.3395522395438509e-08, + "loss": 0.224, + "step": 44403 + }, + { + "epoch": 0.9774158045344487, + "grad_norm": 1.7408082367233706, + "learning_rate": 1.3369460018159974e-08, + "loss": 0.3127, + "step": 44404 + }, + { + "epoch": 0.9774378164208672, + "grad_norm": 1.7172723240070706, + "learning_rate": 1.3343422985161447e-08, + "loss": 0.3077, + "step": 44405 + }, + { + "epoch": 0.977459828307286, + "grad_norm": 0.6194192173404397, + "learning_rate": 1.33174112965756e-08, + "loss": 0.4381, + "step": 44406 + }, + { + "epoch": 0.9774818401937047, + "grad_norm": 1.7959909642850196, + "learning_rate": 1.3291424952535104e-08, + "loss": 0.2221, + "step": 44407 + }, + { + "epoch": 0.9775038520801232, + "grad_norm": 5.727980104475494, + "learning_rate": 1.3265463953172075e-08, + "loss": 0.2852, + "step": 44408 + }, + { + "epoch": 0.977525863966542, + "grad_norm": 2.318952056000476, + "learning_rate": 1.3239528298617521e-08, + "loss": 0.201, + "step": 44409 + }, + { + "epoch": 0.9775478758529605, + "grad_norm": 1.6010350989895354, + "learning_rate": 1.3213617989004113e-08, + "loss": 0.3422, + "step": 44410 + }, + { + "epoch": 0.9775698877393793, + "grad_norm": 0.6352781904451518, + "learning_rate": 1.3187733024463411e-08, + "loss": 0.4513, + "step": 44411 + }, + { + "epoch": 0.977591899625798, + "grad_norm": 1.6580603987371405, + "learning_rate": 1.3161873405126424e-08, + "loss": 0.2652, + "step": 44412 + }, + { + "epoch": 0.9776139115122165, + "grad_norm": 2.3697579827015103, + "learning_rate": 1.3136039131125266e-08, + "loss": 0.3266, + "step": 44413 + }, + { + "epoch": 0.9776359233986353, + "grad_norm": 1.8765949498460044, + "learning_rate": 1.3110230202590945e-08, + "loss": 0.3003, + "step": 44414 + }, + { + "epoch": 0.977657935285054, + "grad_norm": 0.6127638656913473, + "learning_rate": 1.3084446619654466e-08, + "loss": 0.454, + "step": 44415 + }, + { + "epoch": 0.9776799471714726, + "grad_norm": 2.033445520822203, + "learning_rate": 1.3058688382447392e-08, + "loss": 0.2752, + "step": 44416 + }, + { + "epoch": 0.9777019590578913, + "grad_norm": 1.6075531156047498, + "learning_rate": 1.3032955491100174e-08, + "loss": 0.2855, + "step": 44417 + }, + { + "epoch": 0.97772397094431, + "grad_norm": 1.5375547782501287, + "learning_rate": 1.3007247945743817e-08, + "loss": 0.2647, + "step": 44418 + }, + { + "epoch": 0.9777459828307286, + "grad_norm": 1.9571833066275615, + "learning_rate": 1.2981565746508217e-08, + "loss": 0.2638, + "step": 44419 + }, + { + "epoch": 0.9777679947171473, + "grad_norm": 1.2500607646768502, + "learning_rate": 1.2955908893525493e-08, + "loss": 0.2054, + "step": 44420 + }, + { + "epoch": 0.977790006603566, + "grad_norm": 1.8768291188257953, + "learning_rate": 1.2930277386924982e-08, + "loss": 0.2915, + "step": 44421 + }, + { + "epoch": 0.9778120184899846, + "grad_norm": 1.6158153081118751, + "learning_rate": 1.2904671226836584e-08, + "loss": 0.354, + "step": 44422 + }, + { + "epoch": 0.9778340303764033, + "grad_norm": 1.6597201865970916, + "learning_rate": 1.2879090413391304e-08, + "loss": 0.2809, + "step": 44423 + }, + { + "epoch": 0.977856042262822, + "grad_norm": 2.372388640900193, + "learning_rate": 1.2853534946718482e-08, + "loss": 0.3655, + "step": 44424 + }, + { + "epoch": 0.9778780541492406, + "grad_norm": 1.1524809675007677, + "learning_rate": 1.282800482694857e-08, + "loss": 0.1808, + "step": 44425 + }, + { + "epoch": 0.9779000660356593, + "grad_norm": 1.5423832895960692, + "learning_rate": 1.280250005421091e-08, + "loss": 0.2434, + "step": 44426 + }, + { + "epoch": 0.977922077922078, + "grad_norm": 1.685079997615234, + "learning_rate": 1.2777020628635395e-08, + "loss": 0.2611, + "step": 44427 + }, + { + "epoch": 0.9779440898084966, + "grad_norm": 1.8726472254284952, + "learning_rate": 1.2751566550351369e-08, + "loss": 0.2933, + "step": 44428 + }, + { + "epoch": 0.9779661016949153, + "grad_norm": 4.3480500478746675, + "learning_rate": 1.2726137819488726e-08, + "loss": 0.3088, + "step": 44429 + }, + { + "epoch": 0.977988113581334, + "grad_norm": 1.5177229546344897, + "learning_rate": 1.2700734436175699e-08, + "loss": 0.192, + "step": 44430 + }, + { + "epoch": 0.9780101254677526, + "grad_norm": 1.9113962154291682, + "learning_rate": 1.2675356400542183e-08, + "loss": 0.282, + "step": 44431 + }, + { + "epoch": 0.9780321373541713, + "grad_norm": 2.1104658972009647, + "learning_rate": 1.2650003712716963e-08, + "loss": 0.3383, + "step": 44432 + }, + { + "epoch": 0.9780541492405899, + "grad_norm": 2.036359606848692, + "learning_rate": 1.262467637282827e-08, + "loss": 0.2969, + "step": 44433 + }, + { + "epoch": 0.9780761611270086, + "grad_norm": 1.8716506000590332, + "learning_rate": 1.2599374381006002e-08, + "loss": 0.2573, + "step": 44434 + }, + { + "epoch": 0.9780981730134273, + "grad_norm": 2.5683170495360224, + "learning_rate": 1.2574097737378388e-08, + "loss": 0.3175, + "step": 44435 + }, + { + "epoch": 0.9781201848998459, + "grad_norm": 1.790321475404961, + "learning_rate": 1.2548846442073104e-08, + "loss": 0.2739, + "step": 44436 + }, + { + "epoch": 0.9781421967862646, + "grad_norm": 1.7111827589341604, + "learning_rate": 1.2523620495218936e-08, + "loss": 0.3451, + "step": 44437 + }, + { + "epoch": 0.9781642086726833, + "grad_norm": 1.709472299727591, + "learning_rate": 1.249841989694467e-08, + "loss": 0.3127, + "step": 44438 + }, + { + "epoch": 0.9781862205591019, + "grad_norm": 1.9150115100802014, + "learning_rate": 1.2473244647377425e-08, + "loss": 0.2174, + "step": 44439 + }, + { + "epoch": 0.9782082324455206, + "grad_norm": 1.656664350884964, + "learning_rate": 1.2448094746646543e-08, + "loss": 0.2853, + "step": 44440 + }, + { + "epoch": 0.9782302443319393, + "grad_norm": 2.047527794570629, + "learning_rate": 1.2422970194878037e-08, + "loss": 0.3042, + "step": 44441 + }, + { + "epoch": 0.9782522562183579, + "grad_norm": 0.5975296084415989, + "learning_rate": 1.2397870992201243e-08, + "loss": 0.4345, + "step": 44442 + }, + { + "epoch": 0.9782742681047766, + "grad_norm": 2.0190184242284217, + "learning_rate": 1.2372797138742176e-08, + "loss": 0.284, + "step": 44443 + }, + { + "epoch": 0.9782962799911953, + "grad_norm": 0.620164704733416, + "learning_rate": 1.2347748634629619e-08, + "loss": 0.4554, + "step": 44444 + }, + { + "epoch": 0.9783182918776139, + "grad_norm": 1.836111601629047, + "learning_rate": 1.2322725479990138e-08, + "loss": 0.2678, + "step": 44445 + }, + { + "epoch": 0.9783403037640326, + "grad_norm": 1.8968116632266692, + "learning_rate": 1.2297727674951409e-08, + "loss": 0.2518, + "step": 44446 + }, + { + "epoch": 0.9783623156504513, + "grad_norm": 2.0397092754288333, + "learning_rate": 1.2272755219639997e-08, + "loss": 0.2663, + "step": 44447 + }, + { + "epoch": 0.9783843275368699, + "grad_norm": 1.6761690706573344, + "learning_rate": 1.224780811418358e-08, + "loss": 0.2755, + "step": 44448 + }, + { + "epoch": 0.9784063394232886, + "grad_norm": 1.8306823056560455, + "learning_rate": 1.2222886358707609e-08, + "loss": 0.2737, + "step": 44449 + }, + { + "epoch": 0.9784283513097073, + "grad_norm": 1.503636587367104, + "learning_rate": 1.2197989953339762e-08, + "loss": 0.2557, + "step": 44450 + }, + { + "epoch": 0.9784503631961259, + "grad_norm": 2.7115671525400193, + "learning_rate": 1.2173118898206604e-08, + "loss": 0.274, + "step": 44451 + }, + { + "epoch": 0.9784723750825446, + "grad_norm": 0.6267825955865645, + "learning_rate": 1.2148273193434146e-08, + "loss": 0.4527, + "step": 44452 + }, + { + "epoch": 0.9784943869689633, + "grad_norm": 2.3362767389628143, + "learning_rate": 1.2123452839148952e-08, + "loss": 0.2434, + "step": 44453 + }, + { + "epoch": 0.9785163988553819, + "grad_norm": 1.8707961913343878, + "learning_rate": 1.2098657835477035e-08, + "loss": 0.3205, + "step": 44454 + }, + { + "epoch": 0.9785384107418006, + "grad_norm": 3.404152368170316, + "learning_rate": 1.2073888182543847e-08, + "loss": 0.2322, + "step": 44455 + }, + { + "epoch": 0.9785604226282192, + "grad_norm": 0.6084869475843181, + "learning_rate": 1.204914388047651e-08, + "loss": 0.4405, + "step": 44456 + }, + { + "epoch": 0.9785824345146379, + "grad_norm": 1.533241724779068, + "learning_rate": 1.202442492939937e-08, + "loss": 0.2372, + "step": 44457 + }, + { + "epoch": 0.9786044464010566, + "grad_norm": 1.4812795079824144, + "learning_rate": 1.1999731329439545e-08, + "loss": 0.302, + "step": 44458 + }, + { + "epoch": 0.9786264582874752, + "grad_norm": 1.8723602630771032, + "learning_rate": 1.1975063080721383e-08, + "loss": 0.3315, + "step": 44459 + }, + { + "epoch": 0.9786484701738939, + "grad_norm": 0.6243183224410528, + "learning_rate": 1.1950420183370891e-08, + "loss": 0.4437, + "step": 44460 + }, + { + "epoch": 0.9786704820603126, + "grad_norm": 0.6011782101647108, + "learning_rate": 1.1925802637512974e-08, + "loss": 0.4246, + "step": 44461 + }, + { + "epoch": 0.9786924939467312, + "grad_norm": 0.6214725443927439, + "learning_rate": 1.1901210443273081e-08, + "loss": 0.4327, + "step": 44462 + }, + { + "epoch": 0.9787145058331499, + "grad_norm": 1.8366197043458676, + "learning_rate": 1.1876643600775561e-08, + "loss": 0.2559, + "step": 44463 + }, + { + "epoch": 0.9787365177195686, + "grad_norm": 2.221316942738218, + "learning_rate": 1.1852102110145868e-08, + "loss": 0.2695, + "step": 44464 + }, + { + "epoch": 0.9787585296059872, + "grad_norm": 1.6430876671691663, + "learning_rate": 1.1827585971508904e-08, + "loss": 0.3131, + "step": 44465 + }, + { + "epoch": 0.9787805414924059, + "grad_norm": 0.6182613349850731, + "learning_rate": 1.1803095184988456e-08, + "loss": 0.4291, + "step": 44466 + }, + { + "epoch": 0.9788025533788246, + "grad_norm": 2.011448979464595, + "learning_rate": 1.177862975070998e-08, + "loss": 0.2779, + "step": 44467 + }, + { + "epoch": 0.9788245652652432, + "grad_norm": 1.6108125788672778, + "learning_rate": 1.1754189668797266e-08, + "loss": 0.2975, + "step": 44468 + }, + { + "epoch": 0.9788465771516619, + "grad_norm": 1.7164463289681433, + "learning_rate": 1.1729774939374106e-08, + "loss": 0.2405, + "step": 44469 + }, + { + "epoch": 0.9788685890380806, + "grad_norm": 2.18344315036699, + "learning_rate": 1.1705385562565397e-08, + "loss": 0.3008, + "step": 44470 + }, + { + "epoch": 0.9788906009244992, + "grad_norm": 2.1081772357332507, + "learning_rate": 1.1681021538494374e-08, + "loss": 0.2515, + "step": 44471 + }, + { + "epoch": 0.9789126128109179, + "grad_norm": 2.4529317027287636, + "learning_rate": 1.165668286728594e-08, + "loss": 0.3321, + "step": 44472 + }, + { + "epoch": 0.9789346246973366, + "grad_norm": 2.2447919810185706, + "learning_rate": 1.1632369549062772e-08, + "loss": 0.2292, + "step": 44473 + }, + { + "epoch": 0.9789566365837552, + "grad_norm": 1.7967217648168994, + "learning_rate": 1.1608081583948661e-08, + "loss": 0.3466, + "step": 44474 + }, + { + "epoch": 0.9789786484701739, + "grad_norm": 1.8671071894864444, + "learning_rate": 1.158381897206684e-08, + "loss": 0.2441, + "step": 44475 + }, + { + "epoch": 0.9790006603565926, + "grad_norm": 1.8651452183563413, + "learning_rate": 1.1559581713541656e-08, + "loss": 0.3146, + "step": 44476 + }, + { + "epoch": 0.9790226722430112, + "grad_norm": 1.5787534155671665, + "learning_rate": 1.1535369808495233e-08, + "loss": 0.2828, + "step": 44477 + }, + { + "epoch": 0.9790446841294299, + "grad_norm": 1.9343393153073167, + "learning_rate": 1.1511183257050806e-08, + "loss": 0.3325, + "step": 44478 + }, + { + "epoch": 0.9790666960158485, + "grad_norm": 1.4104749140680555, + "learning_rate": 1.148702205933161e-08, + "loss": 0.3293, + "step": 44479 + }, + { + "epoch": 0.9790887079022672, + "grad_norm": 1.8525221539002754, + "learning_rate": 1.1462886215460323e-08, + "loss": 0.327, + "step": 44480 + }, + { + "epoch": 0.9791107197886859, + "grad_norm": 2.4476489486034083, + "learning_rate": 1.143877572556018e-08, + "loss": 0.2905, + "step": 44481 + }, + { + "epoch": 0.9791327316751045, + "grad_norm": 1.58294571863901, + "learning_rate": 1.1414690589752197e-08, + "loss": 0.234, + "step": 44482 + }, + { + "epoch": 0.9791547435615232, + "grad_norm": 1.776436159062956, + "learning_rate": 1.1390630808160164e-08, + "loss": 0.2868, + "step": 44483 + }, + { + "epoch": 0.9791767554479419, + "grad_norm": 2.3055151421486024, + "learning_rate": 1.1366596380906204e-08, + "loss": 0.2651, + "step": 44484 + }, + { + "epoch": 0.9791987673343605, + "grad_norm": 1.3854227637458636, + "learning_rate": 1.1342587308111886e-08, + "loss": 0.2285, + "step": 44485 + }, + { + "epoch": 0.9792207792207792, + "grad_norm": 1.9802770066490276, + "learning_rate": 1.1318603589899335e-08, + "loss": 0.4048, + "step": 44486 + }, + { + "epoch": 0.9792427911071979, + "grad_norm": 2.136569087226775, + "learning_rate": 1.1294645226391233e-08, + "loss": 0.2637, + "step": 44487 + }, + { + "epoch": 0.9792648029936165, + "grad_norm": 2.211046812956483, + "learning_rate": 1.1270712217708036e-08, + "loss": 0.2066, + "step": 44488 + }, + { + "epoch": 0.9792868148800352, + "grad_norm": 2.1124837182427645, + "learning_rate": 1.1246804563972425e-08, + "loss": 0.2893, + "step": 44489 + }, + { + "epoch": 0.9793088267664539, + "grad_norm": 2.3109880728654857, + "learning_rate": 1.122292226530597e-08, + "loss": 0.2698, + "step": 44490 + }, + { + "epoch": 0.9793308386528725, + "grad_norm": 1.6295407237068247, + "learning_rate": 1.1199065321829128e-08, + "loss": 0.282, + "step": 44491 + }, + { + "epoch": 0.9793528505392912, + "grad_norm": 2.8839144296060155, + "learning_rate": 1.1175233733664027e-08, + "loss": 0.3842, + "step": 44492 + }, + { + "epoch": 0.9793748624257099, + "grad_norm": 1.4209893001710612, + "learning_rate": 1.1151427500931123e-08, + "loss": 0.302, + "step": 44493 + }, + { + "epoch": 0.9793968743121285, + "grad_norm": 0.6365527113865522, + "learning_rate": 1.1127646623752541e-08, + "loss": 0.4685, + "step": 44494 + }, + { + "epoch": 0.9794188861985472, + "grad_norm": 2.2075508141865914, + "learning_rate": 1.1103891102247633e-08, + "loss": 0.3672, + "step": 44495 + }, + { + "epoch": 0.9794408980849659, + "grad_norm": 2.100560153852443, + "learning_rate": 1.1080160936537965e-08, + "loss": 0.3341, + "step": 44496 + }, + { + "epoch": 0.9794629099713845, + "grad_norm": 1.4038615515310002, + "learning_rate": 1.1056456126743998e-08, + "loss": 0.2544, + "step": 44497 + }, + { + "epoch": 0.9794849218578032, + "grad_norm": 1.6163683353556244, + "learning_rate": 1.1032776672986745e-08, + "loss": 0.2215, + "step": 44498 + }, + { + "epoch": 0.9795069337442219, + "grad_norm": 1.3550836501382926, + "learning_rate": 1.1009122575385556e-08, + "loss": 0.2731, + "step": 44499 + }, + { + "epoch": 0.9795289456306405, + "grad_norm": 1.8384157338978064, + "learning_rate": 1.0985493834061445e-08, + "loss": 0.2867, + "step": 44500 + }, + { + "epoch": 0.9795509575170592, + "grad_norm": 2.8928970212961884, + "learning_rate": 1.0961890449133761e-08, + "loss": 0.2932, + "step": 44501 + }, + { + "epoch": 0.9795729694034779, + "grad_norm": 2.7648403862273594, + "learning_rate": 1.0938312420723518e-08, + "loss": 0.3052, + "step": 44502 + }, + { + "epoch": 0.9795949812898965, + "grad_norm": 0.6220037896734837, + "learning_rate": 1.0914759748949511e-08, + "loss": 0.4268, + "step": 44503 + }, + { + "epoch": 0.9796169931763152, + "grad_norm": 2.1289829638650763, + "learning_rate": 1.0891232433932197e-08, + "loss": 0.2539, + "step": 44504 + }, + { + "epoch": 0.9796390050627338, + "grad_norm": 1.7982980308445202, + "learning_rate": 1.0867730475790373e-08, + "loss": 0.2997, + "step": 44505 + }, + { + "epoch": 0.9796610169491525, + "grad_norm": 2.1401909993456445, + "learning_rate": 1.0844253874644495e-08, + "loss": 0.3445, + "step": 44506 + }, + { + "epoch": 0.9796830288355712, + "grad_norm": 1.885766414338091, + "learning_rate": 1.0820802630613359e-08, + "loss": 0.2744, + "step": 44507 + }, + { + "epoch": 0.9797050407219898, + "grad_norm": 0.5951276351534259, + "learning_rate": 1.0797376743815757e-08, + "loss": 0.4535, + "step": 44508 + }, + { + "epoch": 0.9797270526084085, + "grad_norm": 0.6360434554732718, + "learning_rate": 1.077397621437104e-08, + "loss": 0.4319, + "step": 44509 + }, + { + "epoch": 0.9797490644948272, + "grad_norm": 1.3702382069915664, + "learning_rate": 1.0750601042398557e-08, + "loss": 0.2765, + "step": 44510 + }, + { + "epoch": 0.9797710763812458, + "grad_norm": 1.6529604488498302, + "learning_rate": 1.0727251228016544e-08, + "loss": 0.2927, + "step": 44511 + }, + { + "epoch": 0.9797930882676645, + "grad_norm": 2.986089768619982, + "learning_rate": 1.0703926771344354e-08, + "loss": 0.2908, + "step": 44512 + }, + { + "epoch": 0.9798151001540832, + "grad_norm": 1.8105575884893574, + "learning_rate": 1.0680627672499666e-08, + "loss": 0.3317, + "step": 44513 + }, + { + "epoch": 0.9798371120405018, + "grad_norm": 2.8618202455590462, + "learning_rate": 1.0657353931601833e-08, + "loss": 0.3858, + "step": 44514 + }, + { + "epoch": 0.9798591239269205, + "grad_norm": 0.6272822296653301, + "learning_rate": 1.0634105548767982e-08, + "loss": 0.4334, + "step": 44515 + }, + { + "epoch": 0.9798811358133392, + "grad_norm": 2.0603518110712575, + "learning_rate": 1.061088252411746e-08, + "loss": 0.248, + "step": 44516 + }, + { + "epoch": 0.9799031476997578, + "grad_norm": 2.096594554565082, + "learning_rate": 1.05876848577674e-08, + "loss": 0.2971, + "step": 44517 + }, + { + "epoch": 0.9799251595861765, + "grad_norm": 1.4456511033721684, + "learning_rate": 1.0564512549836592e-08, + "loss": 0.2617, + "step": 44518 + }, + { + "epoch": 0.9799471714725952, + "grad_norm": 1.7082482617793584, + "learning_rate": 1.0541365600441611e-08, + "loss": 0.291, + "step": 44519 + }, + { + "epoch": 0.9799691833590138, + "grad_norm": 2.3080922783626163, + "learning_rate": 1.051824400970125e-08, + "loss": 0.2614, + "step": 44520 + }, + { + "epoch": 0.9799911952454325, + "grad_norm": 1.9778511411511648, + "learning_rate": 1.0495147777732084e-08, + "loss": 0.2989, + "step": 44521 + }, + { + "epoch": 0.9800132071318512, + "grad_norm": 1.568704497034369, + "learning_rate": 1.047207690465235e-08, + "loss": 0.354, + "step": 44522 + }, + { + "epoch": 0.9800352190182698, + "grad_norm": 1.425838486029851, + "learning_rate": 1.0449031390578623e-08, + "loss": 0.2644, + "step": 44523 + }, + { + "epoch": 0.9800572309046885, + "grad_norm": 0.6183962760019324, + "learning_rate": 1.0426011235628585e-08, + "loss": 0.4463, + "step": 44524 + }, + { + "epoch": 0.9800792427911073, + "grad_norm": 1.8748414488452911, + "learning_rate": 1.0403016439918812e-08, + "loss": 0.3091, + "step": 44525 + }, + { + "epoch": 0.9801012546775258, + "grad_norm": 1.6479012666956327, + "learning_rate": 1.0380047003566428e-08, + "loss": 0.2495, + "step": 44526 + }, + { + "epoch": 0.9801232665639446, + "grad_norm": 0.6263727695232696, + "learning_rate": 1.0357102926688012e-08, + "loss": 0.4526, + "step": 44527 + }, + { + "epoch": 0.9801452784503631, + "grad_norm": 0.6096237278704377, + "learning_rate": 1.0334184209400133e-08, + "loss": 0.4461, + "step": 44528 + }, + { + "epoch": 0.9801672903367818, + "grad_norm": 1.5024081913849288, + "learning_rate": 1.0311290851819921e-08, + "loss": 0.3047, + "step": 44529 + }, + { + "epoch": 0.9801893022232006, + "grad_norm": 2.2446463630785978, + "learning_rate": 1.028842285406284e-08, + "loss": 0.2556, + "step": 44530 + }, + { + "epoch": 0.9802113141096191, + "grad_norm": 1.7916485793875694, + "learning_rate": 1.0265580216244908e-08, + "loss": 0.2727, + "step": 44531 + }, + { + "epoch": 0.9802333259960379, + "grad_norm": 1.8871005404208616, + "learning_rate": 1.0242762938483253e-08, + "loss": 0.3002, + "step": 44532 + }, + { + "epoch": 0.9802553378824566, + "grad_norm": 1.4049401787345726, + "learning_rate": 1.0219971020893893e-08, + "loss": 0.2693, + "step": 44533 + }, + { + "epoch": 0.9802773497688752, + "grad_norm": 1.900234279738768, + "learning_rate": 1.0197204463591181e-08, + "loss": 0.2705, + "step": 44534 + }, + { + "epoch": 0.9802993616552939, + "grad_norm": 2.2525671781166032, + "learning_rate": 1.0174463266692803e-08, + "loss": 0.3113, + "step": 44535 + }, + { + "epoch": 0.9803213735417126, + "grad_norm": 1.8758434244635063, + "learning_rate": 1.0151747430312552e-08, + "loss": 0.3055, + "step": 44536 + }, + { + "epoch": 0.9803433854281312, + "grad_norm": 1.3409381535102796, + "learning_rate": 1.0129056954567006e-08, + "loss": 0.2778, + "step": 44537 + }, + { + "epoch": 0.9803653973145499, + "grad_norm": 1.299135867175289, + "learning_rate": 1.0106391839571072e-08, + "loss": 0.1921, + "step": 44538 + }, + { + "epoch": 0.9803874092009686, + "grad_norm": 2.0026414162345483, + "learning_rate": 1.0083752085439657e-08, + "loss": 0.3316, + "step": 44539 + }, + { + "epoch": 0.9804094210873872, + "grad_norm": 1.6667481973661444, + "learning_rate": 1.006113769228878e-08, + "loss": 0.1908, + "step": 44540 + }, + { + "epoch": 0.9804314329738059, + "grad_norm": 3.5232738902796217, + "learning_rate": 1.003854866023224e-08, + "loss": 0.334, + "step": 44541 + }, + { + "epoch": 0.9804534448602246, + "grad_norm": 2.3723134893656055, + "learning_rate": 1.0015984989385496e-08, + "loss": 0.3223, + "step": 44542 + }, + { + "epoch": 0.9804754567466432, + "grad_norm": 1.9704568892328018, + "learning_rate": 9.993446679863462e-09, + "loss": 0.2404, + "step": 44543 + }, + { + "epoch": 0.9804974686330619, + "grad_norm": 1.4231026160458626, + "learning_rate": 9.970933731780485e-09, + "loss": 0.3223, + "step": 44544 + }, + { + "epoch": 0.9805194805194806, + "grad_norm": 0.6059650192900585, + "learning_rate": 9.948446145250369e-09, + "loss": 0.4383, + "step": 44545 + }, + { + "epoch": 0.9805414924058992, + "grad_norm": 2.109282959833516, + "learning_rate": 9.925983920387462e-09, + "loss": 0.3257, + "step": 44546 + }, + { + "epoch": 0.9805635042923179, + "grad_norm": 1.3920333076544535, + "learning_rate": 9.903547057307228e-09, + "loss": 0.2567, + "step": 44547 + }, + { + "epoch": 0.9805855161787366, + "grad_norm": 2.4034984367554473, + "learning_rate": 9.881135556122356e-09, + "loss": 0.3352, + "step": 44548 + }, + { + "epoch": 0.9806075280651552, + "grad_norm": 1.7317037980738348, + "learning_rate": 9.858749416947199e-09, + "loss": 0.198, + "step": 44549 + }, + { + "epoch": 0.9806295399515739, + "grad_norm": 2.1027560875006883, + "learning_rate": 9.836388639895556e-09, + "loss": 0.2667, + "step": 44550 + }, + { + "epoch": 0.9806515518379925, + "grad_norm": 1.3171900141334403, + "learning_rate": 9.814053225081221e-09, + "loss": 0.2725, + "step": 44551 + }, + { + "epoch": 0.9806735637244112, + "grad_norm": 0.613754255233249, + "learning_rate": 9.79174317261744e-09, + "loss": 0.457, + "step": 44552 + }, + { + "epoch": 0.9806955756108299, + "grad_norm": 2.384320016139632, + "learning_rate": 9.76945848261801e-09, + "loss": 0.3021, + "step": 44553 + }, + { + "epoch": 0.9807175874972485, + "grad_norm": 0.6211824947670063, + "learning_rate": 9.747199155195618e-09, + "loss": 0.4537, + "step": 44554 + }, + { + "epoch": 0.9807395993836672, + "grad_norm": 0.6110351649508643, + "learning_rate": 9.724965190463509e-09, + "loss": 0.4389, + "step": 44555 + }, + { + "epoch": 0.9807616112700859, + "grad_norm": 1.8210400348935125, + "learning_rate": 9.702756588534922e-09, + "loss": 0.2922, + "step": 44556 + }, + { + "epoch": 0.9807836231565045, + "grad_norm": 1.6345862064180905, + "learning_rate": 9.680573349522548e-09, + "loss": 0.2608, + "step": 44557 + }, + { + "epoch": 0.9808056350429232, + "grad_norm": 3.4718464087366643, + "learning_rate": 9.658415473539628e-09, + "loss": 0.2776, + "step": 44558 + }, + { + "epoch": 0.9808276469293419, + "grad_norm": 2.0129785221315575, + "learning_rate": 9.636282960698295e-09, + "loss": 0.3589, + "step": 44559 + }, + { + "epoch": 0.9808496588157605, + "grad_norm": 1.9901817083262987, + "learning_rate": 9.614175811111237e-09, + "loss": 0.3656, + "step": 44560 + }, + { + "epoch": 0.9808716707021792, + "grad_norm": 0.6198590638150658, + "learning_rate": 9.592094024890585e-09, + "loss": 0.433, + "step": 44561 + }, + { + "epoch": 0.9808936825885979, + "grad_norm": 2.1679511926283874, + "learning_rate": 9.570037602148474e-09, + "loss": 0.2879, + "step": 44562 + }, + { + "epoch": 0.9809156944750165, + "grad_norm": 1.6594772888279208, + "learning_rate": 9.54800654299759e-09, + "loss": 0.3231, + "step": 44563 + }, + { + "epoch": 0.9809377063614352, + "grad_norm": 2.6497163915893416, + "learning_rate": 9.52600084754951e-09, + "loss": 0.2412, + "step": 44564 + }, + { + "epoch": 0.9809597182478539, + "grad_norm": 2.0222398807023283, + "learning_rate": 9.504020515915812e-09, + "loss": 0.3048, + "step": 44565 + }, + { + "epoch": 0.9809817301342725, + "grad_norm": 2.441786592390436, + "learning_rate": 9.48206554820863e-09, + "loss": 0.3275, + "step": 44566 + }, + { + "epoch": 0.9810037420206912, + "grad_norm": 1.6249653390638952, + "learning_rate": 9.460135944539539e-09, + "loss": 0.2746, + "step": 44567 + }, + { + "epoch": 0.9810257539071099, + "grad_norm": 1.5755149919961222, + "learning_rate": 9.438231705019563e-09, + "loss": 0.2619, + "step": 44568 + }, + { + "epoch": 0.9810477657935285, + "grad_norm": 1.6823837905235166, + "learning_rate": 9.416352829760833e-09, + "loss": 0.2467, + "step": 44569 + }, + { + "epoch": 0.9810697776799472, + "grad_norm": 3.1438699569753963, + "learning_rate": 9.394499318873818e-09, + "loss": 0.2786, + "step": 44570 + }, + { + "epoch": 0.9810917895663659, + "grad_norm": 1.7415792395122478, + "learning_rate": 9.372671172469538e-09, + "loss": 0.2727, + "step": 44571 + }, + { + "epoch": 0.9811138014527845, + "grad_norm": 0.6280354911233613, + "learning_rate": 9.350868390659572e-09, + "loss": 0.4674, + "step": 44572 + }, + { + "epoch": 0.9811358133392032, + "grad_norm": 1.7926562316893524, + "learning_rate": 9.329090973554389e-09, + "loss": 0.2927, + "step": 44573 + }, + { + "epoch": 0.9811578252256218, + "grad_norm": 1.6189756960613222, + "learning_rate": 9.307338921264453e-09, + "loss": 0.2742, + "step": 44574 + }, + { + "epoch": 0.9811798371120405, + "grad_norm": 1.5416553463050686, + "learning_rate": 9.285612233900786e-09, + "loss": 0.229, + "step": 44575 + }, + { + "epoch": 0.9812018489984592, + "grad_norm": 2.1559260024562157, + "learning_rate": 9.263910911573303e-09, + "loss": 0.3299, + "step": 44576 + }, + { + "epoch": 0.9812238608848778, + "grad_norm": 2.559288439524954, + "learning_rate": 9.242234954393026e-09, + "loss": 0.3731, + "step": 44577 + }, + { + "epoch": 0.9812458727712965, + "grad_norm": 1.89318488452731, + "learning_rate": 9.220584362468754e-09, + "loss": 0.2771, + "step": 44578 + }, + { + "epoch": 0.9812678846577152, + "grad_norm": 0.6262070387727814, + "learning_rate": 9.198959135912067e-09, + "loss": 0.4683, + "step": 44579 + }, + { + "epoch": 0.9812898965441338, + "grad_norm": 2.176905458050161, + "learning_rate": 9.17735927483232e-09, + "loss": 0.2897, + "step": 44580 + }, + { + "epoch": 0.9813119084305525, + "grad_norm": 2.0923081677927744, + "learning_rate": 9.155784779338872e-09, + "loss": 0.3509, + "step": 44581 + }, + { + "epoch": 0.9813339203169712, + "grad_norm": 2.535891557180978, + "learning_rate": 9.13423564954219e-09, + "loss": 0.2561, + "step": 44582 + }, + { + "epoch": 0.9813559322033898, + "grad_norm": 2.270160916244109, + "learning_rate": 9.112711885551073e-09, + "loss": 0.3788, + "step": 44583 + }, + { + "epoch": 0.9813779440898085, + "grad_norm": 3.57164134148412, + "learning_rate": 9.091213487474881e-09, + "loss": 0.2894, + "step": 44584 + }, + { + "epoch": 0.9813999559762272, + "grad_norm": 2.7510991154544833, + "learning_rate": 9.069740455423526e-09, + "loss": 0.3842, + "step": 44585 + }, + { + "epoch": 0.9814219678626458, + "grad_norm": 2.2611346395227434, + "learning_rate": 9.048292789505254e-09, + "loss": 0.2946, + "step": 44586 + }, + { + "epoch": 0.9814439797490645, + "grad_norm": 1.319754323282287, + "learning_rate": 9.026870489829975e-09, + "loss": 0.2576, + "step": 44587 + }, + { + "epoch": 0.9814659916354832, + "grad_norm": 4.295137768590908, + "learning_rate": 9.005473556505939e-09, + "loss": 0.2879, + "step": 44588 + }, + { + "epoch": 0.9814880035219018, + "grad_norm": 3.269655660729122, + "learning_rate": 8.984101989642502e-09, + "loss": 0.2682, + "step": 44589 + }, + { + "epoch": 0.9815100154083205, + "grad_norm": 2.5890406613485086, + "learning_rate": 8.96275578934791e-09, + "loss": 0.3492, + "step": 44590 + }, + { + "epoch": 0.9815320272947392, + "grad_norm": 0.6241201574457688, + "learning_rate": 8.94143495573041e-09, + "loss": 0.4325, + "step": 44591 + }, + { + "epoch": 0.9815540391811578, + "grad_norm": 1.4646150184500257, + "learning_rate": 8.920139488898804e-09, + "loss": 0.2716, + "step": 44592 + }, + { + "epoch": 0.9815760510675765, + "grad_norm": 1.8008556784330059, + "learning_rate": 8.898869388960784e-09, + "loss": 0.2238, + "step": 44593 + }, + { + "epoch": 0.9815980629539952, + "grad_norm": 1.6750946046312838, + "learning_rate": 8.877624656025153e-09, + "loss": 0.3072, + "step": 44594 + }, + { + "epoch": 0.9816200748404138, + "grad_norm": 2.8599720763313954, + "learning_rate": 8.856405290199043e-09, + "loss": 0.3073, + "step": 44595 + }, + { + "epoch": 0.9816420867268325, + "grad_norm": 0.6250668340513972, + "learning_rate": 8.835211291591261e-09, + "loss": 0.4394, + "step": 44596 + }, + { + "epoch": 0.9816640986132511, + "grad_norm": 2.278692219051098, + "learning_rate": 8.814042660308942e-09, + "loss": 0.2728, + "step": 44597 + }, + { + "epoch": 0.9816861104996698, + "grad_norm": 1.6510331708402866, + "learning_rate": 8.792899396459775e-09, + "loss": 0.2868, + "step": 44598 + }, + { + "epoch": 0.9817081223860885, + "grad_norm": 1.8023153229349727, + "learning_rate": 8.771781500151456e-09, + "loss": 0.3053, + "step": 44599 + }, + { + "epoch": 0.9817301342725071, + "grad_norm": 1.330215016570493, + "learning_rate": 8.750688971490562e-09, + "loss": 0.2533, + "step": 44600 + }, + { + "epoch": 0.9817521461589258, + "grad_norm": 0.6058013930646233, + "learning_rate": 8.729621810585342e-09, + "loss": 0.4507, + "step": 44601 + }, + { + "epoch": 0.9817741580453445, + "grad_norm": 2.5143597052967093, + "learning_rate": 8.708580017542379e-09, + "loss": 0.3411, + "step": 44602 + }, + { + "epoch": 0.9817961699317631, + "grad_norm": 0.5975433970025684, + "learning_rate": 8.68756359246825e-09, + "loss": 0.4497, + "step": 44603 + }, + { + "epoch": 0.9818181818181818, + "grad_norm": 2.156146707625871, + "learning_rate": 8.666572535470097e-09, + "loss": 0.3241, + "step": 44604 + }, + { + "epoch": 0.9818401937046005, + "grad_norm": 0.616522327844975, + "learning_rate": 8.645606846654497e-09, + "loss": 0.4384, + "step": 44605 + }, + { + "epoch": 0.9818622055910191, + "grad_norm": 1.7986075493644083, + "learning_rate": 8.624666526128588e-09, + "loss": 0.3212, + "step": 44606 + }, + { + "epoch": 0.9818842174774378, + "grad_norm": 7.43937092479328, + "learning_rate": 8.603751573998399e-09, + "loss": 0.3037, + "step": 44607 + }, + { + "epoch": 0.9819062293638565, + "grad_norm": 3.347580429236334, + "learning_rate": 8.582861990369395e-09, + "loss": 0.2719, + "step": 44608 + }, + { + "epoch": 0.9819282412502751, + "grad_norm": 1.7785002580519662, + "learning_rate": 8.561997775349273e-09, + "loss": 0.2474, + "step": 44609 + }, + { + "epoch": 0.9819502531366938, + "grad_norm": 1.743207718666987, + "learning_rate": 8.541158929042948e-09, + "loss": 0.2629, + "step": 44610 + }, + { + "epoch": 0.9819722650231125, + "grad_norm": 1.9496566753184223, + "learning_rate": 8.520345451557e-09, + "loss": 0.3359, + "step": 44611 + }, + { + "epoch": 0.9819942769095311, + "grad_norm": 1.7282359569086387, + "learning_rate": 8.4995573429969e-09, + "loss": 0.3682, + "step": 44612 + }, + { + "epoch": 0.9820162887959498, + "grad_norm": 1.4287320701146522, + "learning_rate": 8.478794603468122e-09, + "loss": 0.2509, + "step": 44613 + }, + { + "epoch": 0.9820383006823685, + "grad_norm": 2.6926186753215897, + "learning_rate": 8.458057233076134e-09, + "loss": 0.3307, + "step": 44614 + }, + { + "epoch": 0.9820603125687871, + "grad_norm": 1.480753630411019, + "learning_rate": 8.437345231926964e-09, + "loss": 0.2786, + "step": 44615 + }, + { + "epoch": 0.9820823244552058, + "grad_norm": 1.292446082364808, + "learning_rate": 8.416658600125527e-09, + "loss": 0.1732, + "step": 44616 + }, + { + "epoch": 0.9821043363416245, + "grad_norm": 3.1492705411159316, + "learning_rate": 8.39599733777674e-09, + "loss": 0.3311, + "step": 44617 + }, + { + "epoch": 0.9821263482280431, + "grad_norm": 0.6232992478294068, + "learning_rate": 8.375361444986074e-09, + "loss": 0.4635, + "step": 44618 + }, + { + "epoch": 0.9821483601144618, + "grad_norm": 1.6942397178577597, + "learning_rate": 8.35475092185789e-09, + "loss": 0.297, + "step": 44619 + }, + { + "epoch": 0.9821703720008804, + "grad_norm": 3.5338571899524625, + "learning_rate": 8.334165768497105e-09, + "loss": 0.2582, + "step": 44620 + }, + { + "epoch": 0.9821923838872991, + "grad_norm": 1.440801173130138, + "learning_rate": 8.313605985009187e-09, + "loss": 0.238, + "step": 44621 + }, + { + "epoch": 0.9822143957737178, + "grad_norm": 1.907225986367113, + "learning_rate": 8.293071571496835e-09, + "loss": 0.2598, + "step": 44622 + }, + { + "epoch": 0.9822364076601364, + "grad_norm": 1.8374768607874366, + "learning_rate": 8.272562528066075e-09, + "loss": 0.3845, + "step": 44623 + }, + { + "epoch": 0.9822584195465551, + "grad_norm": 1.438229739644176, + "learning_rate": 8.252078854820156e-09, + "loss": 0.2357, + "step": 44624 + }, + { + "epoch": 0.9822804314329738, + "grad_norm": 1.740458011625921, + "learning_rate": 8.231620551863995e-09, + "loss": 0.2084, + "step": 44625 + }, + { + "epoch": 0.9823024433193924, + "grad_norm": 0.6042428797877246, + "learning_rate": 8.211187619300842e-09, + "loss": 0.4414, + "step": 44626 + }, + { + "epoch": 0.9823244552058111, + "grad_norm": 0.6194736291697372, + "learning_rate": 8.19078005723506e-09, + "loss": 0.4547, + "step": 44627 + }, + { + "epoch": 0.9823464670922298, + "grad_norm": 0.643472326992435, + "learning_rate": 8.170397865769897e-09, + "loss": 0.4524, + "step": 44628 + }, + { + "epoch": 0.9823684789786484, + "grad_norm": 1.9404489686005575, + "learning_rate": 8.15004104500916e-09, + "loss": 0.312, + "step": 44629 + }, + { + "epoch": 0.9823904908650671, + "grad_norm": 3.3731961164100164, + "learning_rate": 8.129709595056657e-09, + "loss": 0.2902, + "step": 44630 + }, + { + "epoch": 0.9824125027514858, + "grad_norm": 2.0782707814396892, + "learning_rate": 8.109403516015634e-09, + "loss": 0.359, + "step": 44631 + }, + { + "epoch": 0.9824345146379044, + "grad_norm": 2.4357546626751625, + "learning_rate": 8.089122807988237e-09, + "loss": 0.2344, + "step": 44632 + }, + { + "epoch": 0.9824565265243231, + "grad_norm": 2.03605325654561, + "learning_rate": 8.068867471078823e-09, + "loss": 0.2826, + "step": 44633 + }, + { + "epoch": 0.9824785384107418, + "grad_norm": 1.7433633340228967, + "learning_rate": 8.04863750539009e-09, + "loss": 0.293, + "step": 44634 + }, + { + "epoch": 0.9825005502971604, + "grad_norm": 1.760022183923393, + "learning_rate": 8.028432911024175e-09, + "loss": 0.3091, + "step": 44635 + }, + { + "epoch": 0.9825225621835791, + "grad_norm": 1.412256413599139, + "learning_rate": 8.008253688084888e-09, + "loss": 0.3001, + "step": 44636 + }, + { + "epoch": 0.9825445740699978, + "grad_norm": 2.22492970022881, + "learning_rate": 7.988099836673258e-09, + "loss": 0.3255, + "step": 44637 + }, + { + "epoch": 0.9825665859564164, + "grad_norm": 1.1993575707529025, + "learning_rate": 7.967971356893644e-09, + "loss": 0.1818, + "step": 44638 + }, + { + "epoch": 0.9825885978428351, + "grad_norm": 1.7632952842318856, + "learning_rate": 7.947868248846524e-09, + "loss": 0.3654, + "step": 44639 + }, + { + "epoch": 0.9826106097292538, + "grad_norm": 0.6214315319451861, + "learning_rate": 7.927790512635147e-09, + "loss": 0.4406, + "step": 44640 + }, + { + "epoch": 0.9826326216156724, + "grad_norm": 1.1691124040150183, + "learning_rate": 7.9077381483611e-09, + "loss": 0.1828, + "step": 44641 + }, + { + "epoch": 0.9826546335020911, + "grad_norm": 0.6151891060178258, + "learning_rate": 7.88771115612652e-09, + "loss": 0.4558, + "step": 44642 + }, + { + "epoch": 0.9826766453885099, + "grad_norm": 1.8976918829367808, + "learning_rate": 7.867709536033552e-09, + "loss": 0.3372, + "step": 44643 + }, + { + "epoch": 0.9826986572749284, + "grad_norm": 2.091604722828779, + "learning_rate": 7.847733288182669e-09, + "loss": 0.2561, + "step": 44644 + }, + { + "epoch": 0.9827206691613471, + "grad_norm": 3.6938833234708297, + "learning_rate": 7.827782412676565e-09, + "loss": 0.3207, + "step": 44645 + }, + { + "epoch": 0.9827426810477657, + "grad_norm": 5.4994043435941995, + "learning_rate": 7.807856909616274e-09, + "loss": 0.2602, + "step": 44646 + }, + { + "epoch": 0.9827646929341844, + "grad_norm": 2.0771307936598213, + "learning_rate": 7.787956779102823e-09, + "loss": 0.2888, + "step": 44647 + }, + { + "epoch": 0.9827867048206032, + "grad_norm": 1.4234908450930839, + "learning_rate": 7.7680820212378e-09, + "loss": 0.2487, + "step": 44648 + }, + { + "epoch": 0.9828087167070217, + "grad_norm": 2.9245939113761747, + "learning_rate": 7.748232636122232e-09, + "loss": 0.3069, + "step": 44649 + }, + { + "epoch": 0.9828307285934405, + "grad_norm": 1.3102778627301208, + "learning_rate": 7.728408623856044e-09, + "loss": 0.2662, + "step": 44650 + }, + { + "epoch": 0.9828527404798592, + "grad_norm": 0.608515124981019, + "learning_rate": 7.70860998454137e-09, + "loss": 0.4266, + "step": 44651 + }, + { + "epoch": 0.9828747523662777, + "grad_norm": 1.7563843698169186, + "learning_rate": 7.688836718278137e-09, + "loss": 0.2355, + "step": 44652 + }, + { + "epoch": 0.9828967642526965, + "grad_norm": 1.7817764971237413, + "learning_rate": 7.66908882516626e-09, + "loss": 0.3663, + "step": 44653 + }, + { + "epoch": 0.9829187761391152, + "grad_norm": 2.213728878669397, + "learning_rate": 7.649366305307327e-09, + "loss": 0.2191, + "step": 44654 + }, + { + "epoch": 0.9829407880255338, + "grad_norm": 3.736673311525149, + "learning_rate": 7.629669158800701e-09, + "loss": 0.254, + "step": 44655 + }, + { + "epoch": 0.9829627999119525, + "grad_norm": 2.6502279577995322, + "learning_rate": 7.609997385746858e-09, + "loss": 0.3065, + "step": 44656 + }, + { + "epoch": 0.9829848117983712, + "grad_norm": 0.6116474047512473, + "learning_rate": 7.59035098624572e-09, + "loss": 0.4259, + "step": 44657 + }, + { + "epoch": 0.9830068236847898, + "grad_norm": 2.0811007145697302, + "learning_rate": 7.57072996039665e-09, + "loss": 0.2639, + "step": 44658 + }, + { + "epoch": 0.9830288355712085, + "grad_norm": 1.4728962383485036, + "learning_rate": 7.551134308300123e-09, + "loss": 0.2579, + "step": 44659 + }, + { + "epoch": 0.9830508474576272, + "grad_norm": 5.283844889821032, + "learning_rate": 7.531564030055504e-09, + "loss": 0.3055, + "step": 44660 + }, + { + "epoch": 0.9830728593440458, + "grad_norm": 0.6168784791947457, + "learning_rate": 7.51201912576216e-09, + "loss": 0.4613, + "step": 44661 + }, + { + "epoch": 0.9830948712304645, + "grad_norm": 1.901286368420962, + "learning_rate": 7.492499595519453e-09, + "loss": 0.271, + "step": 44662 + }, + { + "epoch": 0.9831168831168832, + "grad_norm": 2.8069753165101234, + "learning_rate": 7.47300543942675e-09, + "loss": 0.2078, + "step": 44663 + }, + { + "epoch": 0.9831388950033018, + "grad_norm": 1.8722485113876493, + "learning_rate": 7.453536657582861e-09, + "loss": 0.2717, + "step": 44664 + }, + { + "epoch": 0.9831609068897205, + "grad_norm": 1.5310030933859144, + "learning_rate": 7.434093250086594e-09, + "loss": 0.3127, + "step": 44665 + }, + { + "epoch": 0.9831829187761392, + "grad_norm": 1.317356811246047, + "learning_rate": 7.4146752170367595e-09, + "loss": 0.2343, + "step": 44666 + }, + { + "epoch": 0.9832049306625578, + "grad_norm": 1.5375569662627475, + "learning_rate": 7.395282558532724e-09, + "loss": 0.2493, + "step": 44667 + }, + { + "epoch": 0.9832269425489765, + "grad_norm": 3.1364786587919795, + "learning_rate": 7.37591527467274e-09, + "loss": 0.2402, + "step": 44668 + }, + { + "epoch": 0.9832489544353951, + "grad_norm": 1.395614932934858, + "learning_rate": 7.356573365555064e-09, + "loss": 0.2849, + "step": 44669 + }, + { + "epoch": 0.9832709663218138, + "grad_norm": 1.960953773831515, + "learning_rate": 7.33725683127795e-09, + "loss": 0.2451, + "step": 44670 + }, + { + "epoch": 0.9832929782082325, + "grad_norm": 4.945742694594735, + "learning_rate": 7.317965671939653e-09, + "loss": 0.3408, + "step": 44671 + }, + { + "epoch": 0.9833149900946511, + "grad_norm": 0.6356604511213156, + "learning_rate": 7.298699887638428e-09, + "loss": 0.4617, + "step": 44672 + }, + { + "epoch": 0.9833370019810698, + "grad_norm": 1.5600272206004109, + "learning_rate": 7.279459478471973e-09, + "loss": 0.2674, + "step": 44673 + }, + { + "epoch": 0.9833590138674885, + "grad_norm": 1.5481447199665763, + "learning_rate": 7.2602444445379894e-09, + "loss": 0.2627, + "step": 44674 + }, + { + "epoch": 0.9833810257539071, + "grad_norm": 2.0230846842548305, + "learning_rate": 7.241054785934176e-09, + "loss": 0.3203, + "step": 44675 + }, + { + "epoch": 0.9834030376403258, + "grad_norm": 2.2326586083065196, + "learning_rate": 7.221890502758233e-09, + "loss": 0.2538, + "step": 44676 + }, + { + "epoch": 0.9834250495267445, + "grad_norm": 1.6821272235590177, + "learning_rate": 7.20275159510786e-09, + "loss": 0.3357, + "step": 44677 + }, + { + "epoch": 0.9834470614131631, + "grad_norm": 2.1346048537269247, + "learning_rate": 7.183638063079646e-09, + "loss": 0.2769, + "step": 44678 + }, + { + "epoch": 0.9834690732995818, + "grad_norm": 1.723451836448962, + "learning_rate": 7.164549906770734e-09, + "loss": 0.235, + "step": 44679 + }, + { + "epoch": 0.9834910851860005, + "grad_norm": 1.4270047101248826, + "learning_rate": 7.145487126278827e-09, + "loss": 0.2136, + "step": 44680 + }, + { + "epoch": 0.9835130970724191, + "grad_norm": 1.4525348829133151, + "learning_rate": 7.1264497217005125e-09, + "loss": 0.2433, + "step": 44681 + }, + { + "epoch": 0.9835351089588378, + "grad_norm": 2.0710726967782027, + "learning_rate": 7.107437693132379e-09, + "loss": 0.278, + "step": 44682 + }, + { + "epoch": 0.9835571208452565, + "grad_norm": 2.6231557285271063, + "learning_rate": 7.088451040671018e-09, + "loss": 0.3011, + "step": 44683 + }, + { + "epoch": 0.9835791327316751, + "grad_norm": 0.6211405293924268, + "learning_rate": 7.069489764413018e-09, + "loss": 0.4466, + "step": 44684 + }, + { + "epoch": 0.9836011446180938, + "grad_norm": 0.6469200717611471, + "learning_rate": 7.050553864454968e-09, + "loss": 0.4587, + "step": 44685 + }, + { + "epoch": 0.9836231565045125, + "grad_norm": 1.4851344148108385, + "learning_rate": 7.031643340892902e-09, + "loss": 0.2669, + "step": 44686 + }, + { + "epoch": 0.9836451683909311, + "grad_norm": 1.637377651784895, + "learning_rate": 7.012758193822855e-09, + "loss": 0.2851, + "step": 44687 + }, + { + "epoch": 0.9836671802773498, + "grad_norm": 1.731214101122846, + "learning_rate": 6.993898423340862e-09, + "loss": 0.2921, + "step": 44688 + }, + { + "epoch": 0.9836891921637685, + "grad_norm": 1.7526536263628298, + "learning_rate": 6.975064029542955e-09, + "loss": 0.2386, + "step": 44689 + }, + { + "epoch": 0.9837112040501871, + "grad_norm": 1.7951599207760274, + "learning_rate": 6.956255012524615e-09, + "loss": 0.2966, + "step": 44690 + }, + { + "epoch": 0.9837332159366058, + "grad_norm": 1.9986313072878934, + "learning_rate": 6.937471372381321e-09, + "loss": 0.2677, + "step": 44691 + }, + { + "epoch": 0.9837552278230244, + "grad_norm": 1.4803468504051231, + "learning_rate": 6.918713109209108e-09, + "loss": 0.2173, + "step": 44692 + }, + { + "epoch": 0.9837772397094431, + "grad_norm": 0.5979497288076171, + "learning_rate": 6.899980223102898e-09, + "loss": 0.4188, + "step": 44693 + }, + { + "epoch": 0.9837992515958618, + "grad_norm": 1.5889721629745155, + "learning_rate": 6.881272714158171e-09, + "loss": 0.2963, + "step": 44694 + }, + { + "epoch": 0.9838212634822804, + "grad_norm": 1.55000317457101, + "learning_rate": 6.8625905824692976e-09, + "loss": 0.3127, + "step": 44695 + }, + { + "epoch": 0.9838432753686991, + "grad_norm": 2.541671663988087, + "learning_rate": 6.8439338281323096e-09, + "loss": 0.2301, + "step": 44696 + }, + { + "epoch": 0.9838652872551178, + "grad_norm": 1.8025577631003182, + "learning_rate": 6.825302451241023e-09, + "loss": 0.263, + "step": 44697 + }, + { + "epoch": 0.9838872991415364, + "grad_norm": 0.6135848187777028, + "learning_rate": 6.806696451890915e-09, + "loss": 0.4282, + "step": 44698 + }, + { + "epoch": 0.9839093110279551, + "grad_norm": 1.761611059575152, + "learning_rate": 6.788115830175801e-09, + "loss": 0.2875, + "step": 44699 + }, + { + "epoch": 0.9839313229143738, + "grad_norm": 6.367833659824085, + "learning_rate": 6.7695605861911595e-09, + "loss": 0.3127, + "step": 44700 + }, + { + "epoch": 0.9839533348007924, + "grad_norm": 2.205899257801386, + "learning_rate": 6.751030720029694e-09, + "loss": 0.3014, + "step": 44701 + }, + { + "epoch": 0.9839753466872111, + "grad_norm": 2.540801415081388, + "learning_rate": 6.732526231787439e-09, + "loss": 0.2663, + "step": 44702 + }, + { + "epoch": 0.9839973585736298, + "grad_norm": 1.5628043341636189, + "learning_rate": 6.714047121557099e-09, + "loss": 0.2345, + "step": 44703 + }, + { + "epoch": 0.9840193704600484, + "grad_norm": 0.6046275833194422, + "learning_rate": 6.695593389433042e-09, + "loss": 0.4393, + "step": 44704 + }, + { + "epoch": 0.9840413823464671, + "grad_norm": 1.5709707800901778, + "learning_rate": 6.677165035509081e-09, + "loss": 0.3444, + "step": 44705 + }, + { + "epoch": 0.9840633942328858, + "grad_norm": 1.8610709900214433, + "learning_rate": 6.658762059879032e-09, + "loss": 0.2394, + "step": 44706 + }, + { + "epoch": 0.9840854061193044, + "grad_norm": 1.5283885544179858, + "learning_rate": 6.640384462636151e-09, + "loss": 0.4601, + "step": 44707 + }, + { + "epoch": 0.9841074180057231, + "grad_norm": 3.1942077326350375, + "learning_rate": 6.622032243873699e-09, + "loss": 0.3537, + "step": 44708 + }, + { + "epoch": 0.9841294298921418, + "grad_norm": 2.406790515030761, + "learning_rate": 6.6037054036854895e-09, + "loss": 0.2927, + "step": 44709 + }, + { + "epoch": 0.9841514417785604, + "grad_norm": 1.8657275627330518, + "learning_rate": 6.585403942164226e-09, + "loss": 0.3359, + "step": 44710 + }, + { + "epoch": 0.9841734536649791, + "grad_norm": 1.5262401314658112, + "learning_rate": 6.567127859403166e-09, + "loss": 0.3236, + "step": 44711 + }, + { + "epoch": 0.9841954655513978, + "grad_norm": 1.3074480675599491, + "learning_rate": 6.548877155495015e-09, + "loss": 0.2486, + "step": 44712 + }, + { + "epoch": 0.9842174774378164, + "grad_norm": 0.6150718356867387, + "learning_rate": 6.53065183053303e-09, + "loss": 0.444, + "step": 44713 + }, + { + "epoch": 0.9842394893242351, + "grad_norm": 3.161172614803558, + "learning_rate": 6.512451884608806e-09, + "loss": 0.3347, + "step": 44714 + }, + { + "epoch": 0.9842615012106537, + "grad_norm": 1.604024162878437, + "learning_rate": 6.4942773178156e-09, + "loss": 0.2854, + "step": 44715 + }, + { + "epoch": 0.9842835130970724, + "grad_norm": 1.7934725164212602, + "learning_rate": 6.476128130245563e-09, + "loss": 0.277, + "step": 44716 + }, + { + "epoch": 0.9843055249834911, + "grad_norm": 2.0279097612662556, + "learning_rate": 6.458004321991396e-09, + "loss": 0.3296, + "step": 44717 + }, + { + "epoch": 0.9843275368699097, + "grad_norm": 0.616902331155394, + "learning_rate": 6.439905893144693e-09, + "loss": 0.4309, + "step": 44718 + }, + { + "epoch": 0.9843495487563284, + "grad_norm": 1.6367913192708692, + "learning_rate": 6.421832843797604e-09, + "loss": 0.3034, + "step": 44719 + }, + { + "epoch": 0.9843715606427471, + "grad_norm": 3.527300825141279, + "learning_rate": 6.4037851740417216e-09, + "loss": 0.3555, + "step": 44720 + }, + { + "epoch": 0.9843935725291657, + "grad_norm": 1.694512395876483, + "learning_rate": 6.385762883969193e-09, + "loss": 0.2519, + "step": 44721 + }, + { + "epoch": 0.9844155844155844, + "grad_norm": 4.984656185328726, + "learning_rate": 6.367765973671058e-09, + "loss": 0.2589, + "step": 44722 + }, + { + "epoch": 0.9844375963020031, + "grad_norm": 3.2243785170081987, + "learning_rate": 6.349794443240021e-09, + "loss": 0.3849, + "step": 44723 + }, + { + "epoch": 0.9844596081884217, + "grad_norm": 1.903731181763598, + "learning_rate": 6.331848292765452e-09, + "loss": 0.3155, + "step": 44724 + }, + { + "epoch": 0.9844816200748404, + "grad_norm": 3.916053005254999, + "learning_rate": 6.313927522340613e-09, + "loss": 0.3149, + "step": 44725 + }, + { + "epoch": 0.9845036319612591, + "grad_norm": 2.8412664615709913, + "learning_rate": 6.296032132055429e-09, + "loss": 0.3169, + "step": 44726 + }, + { + "epoch": 0.9845256438476777, + "grad_norm": 1.65992765865953, + "learning_rate": 6.278162122000387e-09, + "loss": 0.2899, + "step": 44727 + }, + { + "epoch": 0.9845476557340964, + "grad_norm": 1.6996141823743132, + "learning_rate": 6.260317492267631e-09, + "loss": 0.292, + "step": 44728 + }, + { + "epoch": 0.9845696676205151, + "grad_norm": 1.9209153599662914, + "learning_rate": 6.242498242947092e-09, + "loss": 0.3544, + "step": 44729 + }, + { + "epoch": 0.9845916795069337, + "grad_norm": 1.2694171217562478, + "learning_rate": 6.2247043741298084e-09, + "loss": 0.3223, + "step": 44730 + }, + { + "epoch": 0.9846136913933524, + "grad_norm": 4.45905501330226, + "learning_rate": 6.206935885905152e-09, + "loss": 0.3715, + "step": 44731 + }, + { + "epoch": 0.9846357032797711, + "grad_norm": 2.7762368603803407, + "learning_rate": 6.189192778364717e-09, + "loss": 0.325, + "step": 44732 + }, + { + "epoch": 0.9846577151661897, + "grad_norm": 0.5923518959347743, + "learning_rate": 6.171475051597875e-09, + "loss": 0.4394, + "step": 44733 + }, + { + "epoch": 0.9846797270526084, + "grad_norm": 1.7359894994558178, + "learning_rate": 6.153782705694556e-09, + "loss": 0.241, + "step": 44734 + }, + { + "epoch": 0.9847017389390271, + "grad_norm": 1.7898942297946856, + "learning_rate": 6.136115740745796e-09, + "loss": 0.3567, + "step": 44735 + }, + { + "epoch": 0.9847237508254457, + "grad_norm": 2.186843887311145, + "learning_rate": 6.118474156840415e-09, + "loss": 0.3293, + "step": 44736 + }, + { + "epoch": 0.9847457627118644, + "grad_norm": 2.2059904015043825, + "learning_rate": 6.1008579540677845e-09, + "loss": 0.2814, + "step": 44737 + }, + { + "epoch": 0.984767774598283, + "grad_norm": 1.255567825850199, + "learning_rate": 6.083267132518389e-09, + "loss": 0.1958, + "step": 44738 + }, + { + "epoch": 0.9847897864847017, + "grad_norm": 2.6964035667491024, + "learning_rate": 6.065701692281045e-09, + "loss": 0.3051, + "step": 44739 + }, + { + "epoch": 0.9848117983711204, + "grad_norm": 2.697922508364444, + "learning_rate": 6.0481616334451266e-09, + "loss": 0.303, + "step": 44740 + }, + { + "epoch": 0.984833810257539, + "grad_norm": 1.4335879080284177, + "learning_rate": 6.030646956099451e-09, + "loss": 0.2576, + "step": 44741 + }, + { + "epoch": 0.9848558221439577, + "grad_norm": 1.8313181871229018, + "learning_rate": 6.0131576603339455e-09, + "loss": 0.2747, + "step": 44742 + }, + { + "epoch": 0.9848778340303764, + "grad_norm": 1.5944200501548838, + "learning_rate": 5.9956937462368745e-09, + "loss": 0.3549, + "step": 44743 + }, + { + "epoch": 0.984899845916795, + "grad_norm": 1.9536580664404328, + "learning_rate": 5.9782552138970554e-09, + "loss": 0.2467, + "step": 44744 + }, + { + "epoch": 0.9849218578032137, + "grad_norm": 0.6170340114457532, + "learning_rate": 5.96084206340275e-09, + "loss": 0.4503, + "step": 44745 + }, + { + "epoch": 0.9849438696896324, + "grad_norm": 1.3688609700010215, + "learning_rate": 5.943454294843332e-09, + "loss": 0.1909, + "step": 44746 + }, + { + "epoch": 0.984965881576051, + "grad_norm": 2.8039055553617915, + "learning_rate": 5.926091908306508e-09, + "loss": 0.2585, + "step": 44747 + }, + { + "epoch": 0.9849878934624697, + "grad_norm": 1.7358435488139576, + "learning_rate": 5.908754903880542e-09, + "loss": 0.2185, + "step": 44748 + }, + { + "epoch": 0.9850099053488884, + "grad_norm": 1.3018755559822284, + "learning_rate": 5.891443281654252e-09, + "loss": 0.2975, + "step": 44749 + }, + { + "epoch": 0.985031917235307, + "grad_norm": 2.4780033238987227, + "learning_rate": 5.87415704171479e-09, + "loss": 0.2832, + "step": 44750 + }, + { + "epoch": 0.9850539291217257, + "grad_norm": 2.1586280206153403, + "learning_rate": 5.856896184149863e-09, + "loss": 0.275, + "step": 44751 + }, + { + "epoch": 0.9850759410081444, + "grad_norm": 2.5074489883192297, + "learning_rate": 5.8396607090482896e-09, + "loss": 0.2729, + "step": 44752 + }, + { + "epoch": 0.985097952894563, + "grad_norm": 0.6210935001315548, + "learning_rate": 5.822450616496667e-09, + "loss": 0.4415, + "step": 44753 + }, + { + "epoch": 0.9851199647809817, + "grad_norm": 0.6213485816587467, + "learning_rate": 5.8052659065827026e-09, + "loss": 0.4595, + "step": 44754 + }, + { + "epoch": 0.9851419766674004, + "grad_norm": 2.1881409819159394, + "learning_rate": 5.788106579394104e-09, + "loss": 0.233, + "step": 44755 + }, + { + "epoch": 0.985163988553819, + "grad_norm": 1.7985165758600892, + "learning_rate": 5.770972635018024e-09, + "loss": 0.2555, + "step": 44756 + }, + { + "epoch": 0.9851860004402377, + "grad_norm": 2.075463500875722, + "learning_rate": 5.75386407354106e-09, + "loss": 0.2886, + "step": 44757 + }, + { + "epoch": 0.9852080123266564, + "grad_norm": 1.4246824198058632, + "learning_rate": 5.736780895050365e-09, + "loss": 0.2561, + "step": 44758 + }, + { + "epoch": 0.985230024213075, + "grad_norm": 1.508270484414656, + "learning_rate": 5.71972309963309e-09, + "loss": 0.286, + "step": 44759 + }, + { + "epoch": 0.9852520360994937, + "grad_norm": 0.604340774608391, + "learning_rate": 5.702690687375279e-09, + "loss": 0.4397, + "step": 44760 + }, + { + "epoch": 0.9852740479859123, + "grad_norm": 3.0353921600060514, + "learning_rate": 5.685683658364638e-09, + "loss": 0.5247, + "step": 44761 + }, + { + "epoch": 0.985296059872331, + "grad_norm": 1.7562071643385226, + "learning_rate": 5.6687020126861004e-09, + "loss": 0.2204, + "step": 44762 + }, + { + "epoch": 0.9853180717587497, + "grad_norm": 1.8389968049191368, + "learning_rate": 5.651745750426818e-09, + "loss": 0.3135, + "step": 44763 + }, + { + "epoch": 0.9853400836451683, + "grad_norm": 1.7875468450173921, + "learning_rate": 5.634814871673389e-09, + "loss": 0.2873, + "step": 44764 + }, + { + "epoch": 0.985362095531587, + "grad_norm": 1.5674943543030062, + "learning_rate": 5.617909376511299e-09, + "loss": 0.2732, + "step": 44765 + }, + { + "epoch": 0.9853841074180058, + "grad_norm": 2.0027570172348863, + "learning_rate": 5.601029265026592e-09, + "loss": 0.2365, + "step": 44766 + }, + { + "epoch": 0.9854061193044243, + "grad_norm": 2.4996317335397404, + "learning_rate": 5.584174537304754e-09, + "loss": 0.3207, + "step": 44767 + }, + { + "epoch": 0.985428131190843, + "grad_norm": 0.6237204919680124, + "learning_rate": 5.567345193431828e-09, + "loss": 0.4241, + "step": 44768 + }, + { + "epoch": 0.9854501430772618, + "grad_norm": 1.863567616364294, + "learning_rate": 5.5505412334933e-09, + "loss": 0.3376, + "step": 44769 + }, + { + "epoch": 0.9854721549636803, + "grad_norm": 1.6827294338602459, + "learning_rate": 5.533762657574105e-09, + "loss": 0.2401, + "step": 44770 + }, + { + "epoch": 0.985494166850099, + "grad_norm": 2.357074921270681, + "learning_rate": 5.517009465760281e-09, + "loss": 0.4542, + "step": 44771 + }, + { + "epoch": 0.9855161787365178, + "grad_norm": 1.3278064141039467, + "learning_rate": 5.500281658136763e-09, + "loss": 0.3202, + "step": 44772 + }, + { + "epoch": 0.9855381906229363, + "grad_norm": 1.2010936324032662, + "learning_rate": 5.483579234788483e-09, + "loss": 0.2282, + "step": 44773 + }, + { + "epoch": 0.985560202509355, + "grad_norm": 1.396095501268675, + "learning_rate": 5.466902195799817e-09, + "loss": 0.2058, + "step": 44774 + }, + { + "epoch": 0.9855822143957738, + "grad_norm": 0.6167140398257855, + "learning_rate": 5.450250541256252e-09, + "loss": 0.4383, + "step": 44775 + }, + { + "epoch": 0.9856042262821924, + "grad_norm": 1.449588297568947, + "learning_rate": 5.433624271242166e-09, + "loss": 0.2562, + "step": 44776 + }, + { + "epoch": 0.9856262381686111, + "grad_norm": 0.618388650505701, + "learning_rate": 5.417023385841935e-09, + "loss": 0.4238, + "step": 44777 + }, + { + "epoch": 0.9856482500550298, + "grad_norm": 1.5003880779417562, + "learning_rate": 5.400447885139937e-09, + "loss": 0.3299, + "step": 44778 + }, + { + "epoch": 0.9856702619414484, + "grad_norm": 1.5654071996510361, + "learning_rate": 5.3838977692205475e-09, + "loss": 0.2698, + "step": 44779 + }, + { + "epoch": 0.9856922738278671, + "grad_norm": 1.9276467122415197, + "learning_rate": 5.367373038168144e-09, + "loss": 0.278, + "step": 44780 + }, + { + "epoch": 0.9857142857142858, + "grad_norm": 1.520037873898464, + "learning_rate": 5.350873692065994e-09, + "loss": 0.2314, + "step": 44781 + }, + { + "epoch": 0.9857362976007044, + "grad_norm": 0.6277269747309543, + "learning_rate": 5.334399730998474e-09, + "loss": 0.4414, + "step": 44782 + }, + { + "epoch": 0.9857583094871231, + "grad_norm": 2.473986288266175, + "learning_rate": 5.317951155049406e-09, + "loss": 0.368, + "step": 44783 + }, + { + "epoch": 0.9857803213735418, + "grad_norm": 2.137149359867979, + "learning_rate": 5.3015279643020555e-09, + "loss": 0.2519, + "step": 44784 + }, + { + "epoch": 0.9858023332599604, + "grad_norm": 1.6110437083671836, + "learning_rate": 5.285130158840246e-09, + "loss": 0.2527, + "step": 44785 + }, + { + "epoch": 0.9858243451463791, + "grad_norm": 0.6429362107176363, + "learning_rate": 5.268757738747243e-09, + "loss": 0.4512, + "step": 44786 + }, + { + "epoch": 0.9858463570327977, + "grad_norm": 1.6515494679325966, + "learning_rate": 5.25241070410576e-09, + "loss": 0.3187, + "step": 44787 + }, + { + "epoch": 0.9858683689192164, + "grad_norm": 0.6115488915339047, + "learning_rate": 5.23608905499906e-09, + "loss": 0.4446, + "step": 44788 + }, + { + "epoch": 0.9858903808056351, + "grad_norm": 0.6125587386673573, + "learning_rate": 5.2197927915109695e-09, + "loss": 0.4488, + "step": 44789 + }, + { + "epoch": 0.9859123926920537, + "grad_norm": 1.998952747519687, + "learning_rate": 5.2035219137230864e-09, + "loss": 0.2816, + "step": 44790 + }, + { + "epoch": 0.9859344045784724, + "grad_norm": 1.4397452714219188, + "learning_rate": 5.187276421718679e-09, + "loss": 0.2737, + "step": 44791 + }, + { + "epoch": 0.9859564164648911, + "grad_norm": 1.8828439308664053, + "learning_rate": 5.171056315580458e-09, + "loss": 0.3306, + "step": 44792 + }, + { + "epoch": 0.9859784283513097, + "grad_norm": 1.663014631355596, + "learning_rate": 5.154861595390581e-09, + "loss": 0.205, + "step": 44793 + }, + { + "epoch": 0.9860004402377284, + "grad_norm": 1.7465872413424932, + "learning_rate": 5.1386922612317595e-09, + "loss": 0.2942, + "step": 44794 + }, + { + "epoch": 0.9860224521241471, + "grad_norm": 3.46380859741846, + "learning_rate": 5.122548313185594e-09, + "loss": 0.2978, + "step": 44795 + }, + { + "epoch": 0.9860444640105657, + "grad_norm": 1.7886456227671654, + "learning_rate": 5.106429751334241e-09, + "loss": 0.3258, + "step": 44796 + }, + { + "epoch": 0.9860664758969844, + "grad_norm": 1.787176051609472, + "learning_rate": 5.090336575759858e-09, + "loss": 0.288, + "step": 44797 + }, + { + "epoch": 0.9860884877834031, + "grad_norm": 1.3010215683276505, + "learning_rate": 5.0742687865446005e-09, + "loss": 0.2194, + "step": 44798 + }, + { + "epoch": 0.9861104996698217, + "grad_norm": 1.7282272975554758, + "learning_rate": 5.05822638376896e-09, + "loss": 0.2008, + "step": 44799 + }, + { + "epoch": 0.9861325115562404, + "grad_norm": 2.808773524836348, + "learning_rate": 5.042209367516204e-09, + "loss": 0.269, + "step": 44800 + }, + { + "epoch": 0.9861545234426591, + "grad_norm": 1.7501506124361963, + "learning_rate": 5.026217737865713e-09, + "loss": 0.2469, + "step": 44801 + }, + { + "epoch": 0.9861765353290777, + "grad_norm": 1.4769298237337325, + "learning_rate": 5.010251494900753e-09, + "loss": 0.2262, + "step": 44802 + }, + { + "epoch": 0.9861985472154964, + "grad_norm": 5.0775473964332605, + "learning_rate": 4.994310638701261e-09, + "loss": 0.323, + "step": 44803 + }, + { + "epoch": 0.9862205591019151, + "grad_norm": 1.1758765548656358, + "learning_rate": 4.9783951693477274e-09, + "loss": 0.2237, + "step": 44804 + }, + { + "epoch": 0.9862425709883337, + "grad_norm": 2.283157752094431, + "learning_rate": 4.962505086922864e-09, + "loss": 0.3408, + "step": 44805 + }, + { + "epoch": 0.9862645828747524, + "grad_norm": 1.6123019157941136, + "learning_rate": 4.946640391506052e-09, + "loss": 0.2911, + "step": 44806 + }, + { + "epoch": 0.9862865947611711, + "grad_norm": 2.926965566476933, + "learning_rate": 4.930801083177783e-09, + "loss": 0.2592, + "step": 44807 + }, + { + "epoch": 0.9863086066475897, + "grad_norm": 4.544935939032148, + "learning_rate": 4.914987162019658e-09, + "loss": 0.4295, + "step": 44808 + }, + { + "epoch": 0.9863306185340084, + "grad_norm": 1.8463712015623992, + "learning_rate": 4.8991986281116125e-09, + "loss": 0.2466, + "step": 44809 + }, + { + "epoch": 0.986352630420427, + "grad_norm": 1.5170796889730953, + "learning_rate": 4.883435481533028e-09, + "loss": 0.267, + "step": 44810 + }, + { + "epoch": 0.9863746423068457, + "grad_norm": 1.9971326185134508, + "learning_rate": 4.86769772236495e-09, + "loss": 0.3031, + "step": 44811 + }, + { + "epoch": 0.9863966541932644, + "grad_norm": 4.378770657342369, + "learning_rate": 4.851985350687316e-09, + "loss": 0.323, + "step": 44812 + }, + { + "epoch": 0.986418666079683, + "grad_norm": 0.6205044012174247, + "learning_rate": 4.836298366579506e-09, + "loss": 0.4562, + "step": 44813 + }, + { + "epoch": 0.9864406779661017, + "grad_norm": 1.9722171105256852, + "learning_rate": 4.820636770122011e-09, + "loss": 0.2168, + "step": 44814 + }, + { + "epoch": 0.9864626898525204, + "grad_norm": 1.7080709476774245, + "learning_rate": 4.805000561393658e-09, + "loss": 0.3701, + "step": 44815 + }, + { + "epoch": 0.986484701738939, + "grad_norm": 2.120599076301544, + "learning_rate": 4.789389740473827e-09, + "loss": 0.3052, + "step": 44816 + }, + { + "epoch": 0.9865067136253577, + "grad_norm": 1.987531948698244, + "learning_rate": 4.773804307443008e-09, + "loss": 0.3172, + "step": 44817 + }, + { + "epoch": 0.9865287255117764, + "grad_norm": 1.84111744993108, + "learning_rate": 4.75824426237892e-09, + "loss": 0.2409, + "step": 44818 + }, + { + "epoch": 0.986550737398195, + "grad_norm": 1.4472484651285413, + "learning_rate": 4.74270960536205e-09, + "loss": 0.274, + "step": 44819 + }, + { + "epoch": 0.9865727492846137, + "grad_norm": 0.6082247181060978, + "learning_rate": 4.727200336470117e-09, + "loss": 0.478, + "step": 44820 + }, + { + "epoch": 0.9865947611710324, + "grad_norm": 1.7531080020004681, + "learning_rate": 4.7117164557830555e-09, + "loss": 0.2468, + "step": 44821 + }, + { + "epoch": 0.986616773057451, + "grad_norm": 2.105767082231065, + "learning_rate": 4.696257963378581e-09, + "loss": 0.2864, + "step": 44822 + }, + { + "epoch": 0.9866387849438697, + "grad_norm": 1.4047667298118598, + "learning_rate": 4.68082485933552e-09, + "loss": 0.2648, + "step": 44823 + }, + { + "epoch": 0.9866607968302884, + "grad_norm": 1.6106378518667048, + "learning_rate": 4.665417143732698e-09, + "loss": 0.2858, + "step": 44824 + }, + { + "epoch": 0.986682808716707, + "grad_norm": 2.374386228469242, + "learning_rate": 4.650034816648385e-09, + "loss": 0.2768, + "step": 44825 + }, + { + "epoch": 0.9867048206031257, + "grad_norm": 1.5629869100161322, + "learning_rate": 4.634677878160298e-09, + "loss": 0.22, + "step": 44826 + }, + { + "epoch": 0.9867268324895444, + "grad_norm": 5.783599505768112, + "learning_rate": 4.619346328347263e-09, + "loss": 0.2506, + "step": 44827 + }, + { + "epoch": 0.986748844375963, + "grad_norm": 0.6150665443844784, + "learning_rate": 4.6040401672864384e-09, + "loss": 0.4603, + "step": 44828 + }, + { + "epoch": 0.9867708562623817, + "grad_norm": 2.16633069492632, + "learning_rate": 4.588759395055542e-09, + "loss": 0.3016, + "step": 44829 + }, + { + "epoch": 0.9867928681488004, + "grad_norm": 1.4819118273732206, + "learning_rate": 4.573504011732843e-09, + "loss": 0.2802, + "step": 44830 + }, + { + "epoch": 0.986814880035219, + "grad_norm": 21.00244724513947, + "learning_rate": 4.558274017396058e-09, + "loss": 0.2874, + "step": 44831 + }, + { + "epoch": 0.9868368919216377, + "grad_norm": 0.6109575633775468, + "learning_rate": 4.543069412121237e-09, + "loss": 0.4397, + "step": 44832 + }, + { + "epoch": 0.9868589038080563, + "grad_norm": 2.92689262431435, + "learning_rate": 4.527890195987206e-09, + "loss": 0.3262, + "step": 44833 + }, + { + "epoch": 0.986880915694475, + "grad_norm": 1.3763895321992081, + "learning_rate": 4.5127363690705696e-09, + "loss": 0.2394, + "step": 44834 + }, + { + "epoch": 0.9869029275808937, + "grad_norm": 1.6237279514910472, + "learning_rate": 4.4976079314479336e-09, + "loss": 0.2391, + "step": 44835 + }, + { + "epoch": 0.9869249394673123, + "grad_norm": 5.299959385810171, + "learning_rate": 4.482504883196459e-09, + "loss": 0.246, + "step": 44836 + }, + { + "epoch": 0.986946951353731, + "grad_norm": 2.490571094492529, + "learning_rate": 4.467427224393306e-09, + "loss": 0.251, + "step": 44837 + }, + { + "epoch": 0.9869689632401497, + "grad_norm": 1.8569178958868215, + "learning_rate": 4.452374955114524e-09, + "loss": 0.3397, + "step": 44838 + }, + { + "epoch": 0.9869909751265683, + "grad_norm": 1.8559172908231119, + "learning_rate": 4.437348075437276e-09, + "loss": 0.274, + "step": 44839 + }, + { + "epoch": 0.987012987012987, + "grad_norm": 1.423023366674422, + "learning_rate": 4.422346585437054e-09, + "loss": 0.3034, + "step": 44840 + }, + { + "epoch": 0.9870349988994057, + "grad_norm": 0.6207327201066489, + "learning_rate": 4.407370485190465e-09, + "loss": 0.4425, + "step": 44841 + }, + { + "epoch": 0.9870570107858243, + "grad_norm": 1.650693475913136, + "learning_rate": 4.392419774774115e-09, + "loss": 0.3369, + "step": 44842 + }, + { + "epoch": 0.987079022672243, + "grad_norm": 1.4425277701112131, + "learning_rate": 4.377494454262943e-09, + "loss": 0.2337, + "step": 44843 + }, + { + "epoch": 0.9871010345586617, + "grad_norm": 2.3588613778398275, + "learning_rate": 4.362594523734109e-09, + "loss": 0.2382, + "step": 44844 + }, + { + "epoch": 0.9871230464450803, + "grad_norm": 1.96491819257492, + "learning_rate": 4.347719983262555e-09, + "loss": 0.346, + "step": 44845 + }, + { + "epoch": 0.987145058331499, + "grad_norm": 2.101667100883566, + "learning_rate": 4.3328708329237744e-09, + "loss": 0.2564, + "step": 44846 + }, + { + "epoch": 0.9871670702179177, + "grad_norm": 1.9122312801335362, + "learning_rate": 4.318047072793818e-09, + "loss": 0.2354, + "step": 44847 + }, + { + "epoch": 0.9871890821043363, + "grad_norm": 6.358839342746536, + "learning_rate": 4.303248702947071e-09, + "loss": 0.2447, + "step": 44848 + }, + { + "epoch": 0.987211093990755, + "grad_norm": 2.1436828130811407, + "learning_rate": 4.2884757234601395e-09, + "loss": 0.3215, + "step": 44849 + }, + { + "epoch": 0.9872331058771737, + "grad_norm": 1.5595783833712051, + "learning_rate": 4.2737281344068515e-09, + "loss": 0.2818, + "step": 44850 + }, + { + "epoch": 0.9872551177635923, + "grad_norm": 0.6147980868325094, + "learning_rate": 4.259005935862703e-09, + "loss": 0.4507, + "step": 44851 + }, + { + "epoch": 0.987277129650011, + "grad_norm": 1.474169294856421, + "learning_rate": 4.24430912790208e-09, + "loss": 0.2546, + "step": 44852 + }, + { + "epoch": 0.9872991415364297, + "grad_norm": 1.4954546817971708, + "learning_rate": 4.229637710601031e-09, + "loss": 0.1993, + "step": 44853 + }, + { + "epoch": 0.9873211534228483, + "grad_norm": 3.1656281784662035, + "learning_rate": 4.2149916840322765e-09, + "loss": 0.2282, + "step": 44854 + }, + { + "epoch": 0.987343165309267, + "grad_norm": 1.5900281076226093, + "learning_rate": 4.200371048271312e-09, + "loss": 0.2276, + "step": 44855 + }, + { + "epoch": 0.9873651771956856, + "grad_norm": 1.9213373344405968, + "learning_rate": 4.185775803392522e-09, + "loss": 0.3273, + "step": 44856 + }, + { + "epoch": 0.9873871890821043, + "grad_norm": 1.4231587681865898, + "learning_rate": 4.171205949469182e-09, + "loss": 0.3315, + "step": 44857 + }, + { + "epoch": 0.987409200968523, + "grad_norm": 2.008928152218993, + "learning_rate": 4.156661486576785e-09, + "loss": 0.3296, + "step": 44858 + }, + { + "epoch": 0.9874312128549416, + "grad_norm": 1.680482136683133, + "learning_rate": 4.142142414788053e-09, + "loss": 0.2769, + "step": 44859 + }, + { + "epoch": 0.9874532247413603, + "grad_norm": 0.6323657154331078, + "learning_rate": 4.1276487341773695e-09, + "loss": 0.4527, + "step": 44860 + }, + { + "epoch": 0.987475236627779, + "grad_norm": 1.6749694494796796, + "learning_rate": 4.11318044481801e-09, + "loss": 0.1908, + "step": 44861 + }, + { + "epoch": 0.9874972485141976, + "grad_norm": 1.89428536124291, + "learning_rate": 4.098737546783249e-09, + "loss": 0.3381, + "step": 44862 + }, + { + "epoch": 0.9875192604006163, + "grad_norm": 2.4119983471675406, + "learning_rate": 4.084320040147471e-09, + "loss": 0.245, + "step": 44863 + }, + { + "epoch": 0.987541272287035, + "grad_norm": 1.7434182864919157, + "learning_rate": 4.069927924983397e-09, + "loss": 0.3523, + "step": 44864 + }, + { + "epoch": 0.9875632841734536, + "grad_norm": 1.649930092578049, + "learning_rate": 4.0555612013642996e-09, + "loss": 0.2685, + "step": 44865 + }, + { + "epoch": 0.9875852960598723, + "grad_norm": 1.8474904603828515, + "learning_rate": 4.0412198693629e-09, + "loss": 0.3051, + "step": 44866 + }, + { + "epoch": 0.987607307946291, + "grad_norm": 1.8956820413949835, + "learning_rate": 4.026903929052473e-09, + "loss": 0.322, + "step": 44867 + }, + { + "epoch": 0.9876293198327096, + "grad_norm": 1.7794099407616666, + "learning_rate": 4.012613380505181e-09, + "loss": 0.3612, + "step": 44868 + }, + { + "epoch": 0.9876513317191283, + "grad_norm": 1.6047063538779518, + "learning_rate": 3.998348223794302e-09, + "loss": 0.2872, + "step": 44869 + }, + { + "epoch": 0.987673343605547, + "grad_norm": 1.476102554105853, + "learning_rate": 3.984108458991998e-09, + "loss": 0.2576, + "step": 44870 + }, + { + "epoch": 0.9876953554919656, + "grad_norm": 1.994101204948247, + "learning_rate": 3.969894086170989e-09, + "loss": 0.3048, + "step": 44871 + }, + { + "epoch": 0.9877173673783843, + "grad_norm": 0.6140357748479589, + "learning_rate": 3.955705105402885e-09, + "loss": 0.4444, + "step": 44872 + }, + { + "epoch": 0.987739379264803, + "grad_norm": 0.5956746927821083, + "learning_rate": 3.941541516760406e-09, + "loss": 0.4444, + "step": 44873 + }, + { + "epoch": 0.9877613911512216, + "grad_norm": 1.6179711715176612, + "learning_rate": 3.927403320315714e-09, + "loss": 0.3131, + "step": 44874 + }, + { + "epoch": 0.9877834030376403, + "grad_norm": 1.667065456327562, + "learning_rate": 3.913290516139867e-09, + "loss": 0.3398, + "step": 44875 + }, + { + "epoch": 0.987805414924059, + "grad_norm": 1.6925414546540922, + "learning_rate": 3.899203104305027e-09, + "loss": 0.3564, + "step": 44876 + }, + { + "epoch": 0.9878274268104776, + "grad_norm": 0.603542361426545, + "learning_rate": 3.8851410848828045e-09, + "loss": 0.4295, + "step": 44877 + }, + { + "epoch": 0.9878494386968963, + "grad_norm": 2.341664662863132, + "learning_rate": 3.871104457944252e-09, + "loss": 0.3001, + "step": 44878 + }, + { + "epoch": 0.9878714505833149, + "grad_norm": 1.898873692087537, + "learning_rate": 3.857093223561537e-09, + "loss": 0.2596, + "step": 44879 + }, + { + "epoch": 0.9878934624697336, + "grad_norm": 1.8355885233933862, + "learning_rate": 3.843107381805155e-09, + "loss": 0.3684, + "step": 44880 + }, + { + "epoch": 0.9879154743561523, + "grad_norm": 3.883211242281952, + "learning_rate": 3.829146932746719e-09, + "loss": 0.2855, + "step": 44881 + }, + { + "epoch": 0.9879374862425709, + "grad_norm": 2.406362607220647, + "learning_rate": 3.815211876456726e-09, + "loss": 0.3244, + "step": 44882 + }, + { + "epoch": 0.9879594981289896, + "grad_norm": 1.3147809254760205, + "learning_rate": 3.8013022130062306e-09, + "loss": 0.1928, + "step": 44883 + }, + { + "epoch": 0.9879815100154083, + "grad_norm": 1.411478219594276, + "learning_rate": 3.787417942465732e-09, + "loss": 0.2467, + "step": 44884 + }, + { + "epoch": 0.9880035219018269, + "grad_norm": 1.525745258036248, + "learning_rate": 3.773559064906285e-09, + "loss": 0.2495, + "step": 44885 + }, + { + "epoch": 0.9880255337882456, + "grad_norm": 1.530324847262175, + "learning_rate": 3.7597255803978325e-09, + "loss": 0.2644, + "step": 44886 + }, + { + "epoch": 0.9880475456746644, + "grad_norm": 2.6265335550854036, + "learning_rate": 3.745917489010875e-09, + "loss": 0.2879, + "step": 44887 + }, + { + "epoch": 0.988069557561083, + "grad_norm": 0.6046600020431565, + "learning_rate": 3.732134790815356e-09, + "loss": 0.4519, + "step": 44888 + }, + { + "epoch": 0.9880915694475016, + "grad_norm": 2.2648546419306452, + "learning_rate": 3.7183774858817745e-09, + "loss": 0.3514, + "step": 44889 + }, + { + "epoch": 0.9881135813339204, + "grad_norm": 1.6510859607766402, + "learning_rate": 3.7046455742795193e-09, + "loss": 0.2216, + "step": 44890 + }, + { + "epoch": 0.988135593220339, + "grad_norm": 0.6095016581496314, + "learning_rate": 3.6909390560790903e-09, + "loss": 0.4361, + "step": 44891 + }, + { + "epoch": 0.9881576051067577, + "grad_norm": 1.7064516478396792, + "learning_rate": 3.6772579313493204e-09, + "loss": 0.2567, + "step": 44892 + }, + { + "epoch": 0.9881796169931764, + "grad_norm": 1.8090007051021948, + "learning_rate": 3.6636022001601545e-09, + "loss": 0.2902, + "step": 44893 + }, + { + "epoch": 0.988201628879595, + "grad_norm": 2.105557652748768, + "learning_rate": 3.649971862580981e-09, + "loss": 0.3054, + "step": 44894 + }, + { + "epoch": 0.9882236407660137, + "grad_norm": 1.6678120929250566, + "learning_rate": 3.636366918681744e-09, + "loss": 0.2796, + "step": 44895 + }, + { + "epoch": 0.9882456526524324, + "grad_norm": 0.6284632801517513, + "learning_rate": 3.622787368530167e-09, + "loss": 0.4564, + "step": 44896 + }, + { + "epoch": 0.988267664538851, + "grad_norm": 1.6852995098641121, + "learning_rate": 3.6092332121961947e-09, + "loss": 0.2779, + "step": 44897 + }, + { + "epoch": 0.9882896764252697, + "grad_norm": 3.2080589453759156, + "learning_rate": 3.5957044497486604e-09, + "loss": 0.366, + "step": 44898 + }, + { + "epoch": 0.9883116883116884, + "grad_norm": 3.371772744283906, + "learning_rate": 3.5822010812563978e-09, + "loss": 0.2484, + "step": 44899 + }, + { + "epoch": 0.988333700198107, + "grad_norm": 2.350643754992872, + "learning_rate": 3.568723106787686e-09, + "loss": 0.2967, + "step": 44900 + }, + { + "epoch": 0.9883557120845257, + "grad_norm": 2.2376830098706835, + "learning_rate": 3.5552705264108033e-09, + "loss": 0.266, + "step": 44901 + }, + { + "epoch": 0.9883777239709443, + "grad_norm": 1.8541196228370556, + "learning_rate": 3.541843340195139e-09, + "loss": 0.2241, + "step": 44902 + }, + { + "epoch": 0.988399735857363, + "grad_norm": 1.5819515118824965, + "learning_rate": 3.5284415482078614e-09, + "loss": 0.1964, + "step": 44903 + }, + { + "epoch": 0.9884217477437817, + "grad_norm": 0.6214141632390823, + "learning_rate": 3.5150651505178048e-09, + "loss": 0.4311, + "step": 44904 + }, + { + "epoch": 0.9884437596302003, + "grad_norm": 2.8571165373839804, + "learning_rate": 3.5017141471926917e-09, + "loss": 0.3459, + "step": 44905 + }, + { + "epoch": 0.988465771516619, + "grad_norm": 2.3237166184282594, + "learning_rate": 3.488388538300247e-09, + "loss": 0.3401, + "step": 44906 + }, + { + "epoch": 0.9884877834030377, + "grad_norm": 2.514191648756793, + "learning_rate": 3.4750883239081933e-09, + "loss": 0.3281, + "step": 44907 + }, + { + "epoch": 0.9885097952894563, + "grad_norm": 1.7599189336702417, + "learning_rate": 3.4618135040842547e-09, + "loss": 0.3329, + "step": 44908 + }, + { + "epoch": 0.988531807175875, + "grad_norm": 1.6252747348588943, + "learning_rate": 3.448564078896155e-09, + "loss": 0.2573, + "step": 44909 + }, + { + "epoch": 0.9885538190622937, + "grad_norm": 2.2656722249671692, + "learning_rate": 3.435340048411062e-09, + "loss": 0.3568, + "step": 44910 + }, + { + "epoch": 0.9885758309487123, + "grad_norm": 2.8326800232687557, + "learning_rate": 3.42214141269559e-09, + "loss": 0.3426, + "step": 44911 + }, + { + "epoch": 0.988597842835131, + "grad_norm": 2.069414363607158, + "learning_rate": 3.408968171818017e-09, + "loss": 0.2702, + "step": 44912 + }, + { + "epoch": 0.9886198547215497, + "grad_norm": 1.836733401436703, + "learning_rate": 3.395820325843846e-09, + "loss": 0.3105, + "step": 44913 + }, + { + "epoch": 0.9886418666079683, + "grad_norm": 1.6514969632603733, + "learning_rate": 3.3826978748413563e-09, + "loss": 0.2986, + "step": 44914 + }, + { + "epoch": 0.988663878494387, + "grad_norm": 1.4889512408461318, + "learning_rate": 3.369600818876051e-09, + "loss": 0.2196, + "step": 44915 + }, + { + "epoch": 0.9886858903808057, + "grad_norm": 0.6030778077599378, + "learning_rate": 3.3565291580150984e-09, + "loss": 0.4449, + "step": 44916 + }, + { + "epoch": 0.9887079022672243, + "grad_norm": 0.582514319067846, + "learning_rate": 3.3434828923251116e-09, + "loss": 0.4674, + "step": 44917 + }, + { + "epoch": 0.988729914153643, + "grad_norm": 1.7805319909212372, + "learning_rate": 3.3304620218721496e-09, + "loss": 0.2783, + "step": 44918 + }, + { + "epoch": 0.9887519260400617, + "grad_norm": 0.6179294354765046, + "learning_rate": 3.3174665467217148e-09, + "loss": 0.4414, + "step": 44919 + }, + { + "epoch": 0.9887739379264803, + "grad_norm": 1.93779297945214, + "learning_rate": 3.304496466940976e-09, + "loss": 0.2679, + "step": 44920 + }, + { + "epoch": 0.988795949812899, + "grad_norm": 1.8219729038589902, + "learning_rate": 3.2915517825954367e-09, + "loss": 0.2273, + "step": 44921 + }, + { + "epoch": 0.9888179616993177, + "grad_norm": 1.7364832421913923, + "learning_rate": 3.2786324937506e-09, + "loss": 0.2998, + "step": 44922 + }, + { + "epoch": 0.9888399735857363, + "grad_norm": 1.6898475288100947, + "learning_rate": 3.2657386004725235e-09, + "loss": 0.2185, + "step": 44923 + }, + { + "epoch": 0.988861985472155, + "grad_norm": 0.6073961892433756, + "learning_rate": 3.252870102826711e-09, + "loss": 0.4485, + "step": 44924 + }, + { + "epoch": 0.9888839973585736, + "grad_norm": 0.6021705244534863, + "learning_rate": 3.24002700087811e-09, + "loss": 0.4404, + "step": 44925 + }, + { + "epoch": 0.9889060092449923, + "grad_norm": 2.340546240363018, + "learning_rate": 3.227209294692224e-09, + "loss": 0.357, + "step": 44926 + }, + { + "epoch": 0.988928021131411, + "grad_norm": 1.7251105108577116, + "learning_rate": 3.2144169843345562e-09, + "loss": 0.3499, + "step": 44927 + }, + { + "epoch": 0.9889500330178296, + "grad_norm": 0.6162968375305157, + "learning_rate": 3.2016500698694995e-09, + "loss": 0.4225, + "step": 44928 + }, + { + "epoch": 0.9889720449042483, + "grad_norm": 2.3518086502783992, + "learning_rate": 3.1889085513625574e-09, + "loss": 0.295, + "step": 44929 + }, + { + "epoch": 0.988994056790667, + "grad_norm": 0.6034538340070705, + "learning_rate": 3.176192428878677e-09, + "loss": 0.4359, + "step": 44930 + }, + { + "epoch": 0.9890160686770856, + "grad_norm": 0.6238020261883607, + "learning_rate": 3.163501702481697e-09, + "loss": 0.4597, + "step": 44931 + }, + { + "epoch": 0.9890380805635043, + "grad_norm": 1.1331686992075831, + "learning_rate": 3.1508363722360102e-09, + "loss": 0.2182, + "step": 44932 + }, + { + "epoch": 0.989060092449923, + "grad_norm": 0.6422636804855463, + "learning_rate": 3.1381964382065645e-09, + "loss": 0.4517, + "step": 44933 + }, + { + "epoch": 0.9890821043363416, + "grad_norm": 1.208223576338071, + "learning_rate": 3.1255819004577527e-09, + "loss": 0.2348, + "step": 44934 + }, + { + "epoch": 0.9891041162227603, + "grad_norm": 0.6180736710768792, + "learning_rate": 3.1129927590534125e-09, + "loss": 0.4267, + "step": 44935 + }, + { + "epoch": 0.989126128109179, + "grad_norm": 0.5935447471194362, + "learning_rate": 3.1004290140573823e-09, + "loss": 0.4339, + "step": 44936 + }, + { + "epoch": 0.9891481399955976, + "grad_norm": 0.601498656163601, + "learning_rate": 3.0878906655334994e-09, + "loss": 0.4225, + "step": 44937 + }, + { + "epoch": 0.9891701518820163, + "grad_norm": 1.4243738901313374, + "learning_rate": 3.075377713546157e-09, + "loss": 0.2871, + "step": 44938 + }, + { + "epoch": 0.989192163768435, + "grad_norm": 2.248595634325574, + "learning_rate": 3.062890158158083e-09, + "loss": 0.3072, + "step": 44939 + }, + { + "epoch": 0.9892141756548536, + "grad_norm": 1.6464336370088808, + "learning_rate": 3.0504279994331145e-09, + "loss": 0.2799, + "step": 44940 + }, + { + "epoch": 0.9892361875412723, + "grad_norm": 1.78755919832683, + "learning_rate": 3.0379912374345344e-09, + "loss": 0.2884, + "step": 44941 + }, + { + "epoch": 0.989258199427691, + "grad_norm": 2.212616918431086, + "learning_rate": 3.0255798722256256e-09, + "loss": 0.3804, + "step": 44942 + }, + { + "epoch": 0.9892802113141096, + "grad_norm": 2.4126160278687743, + "learning_rate": 3.0131939038691162e-09, + "loss": 0.3372, + "step": 44943 + }, + { + "epoch": 0.9893022232005283, + "grad_norm": 1.6121737346810094, + "learning_rate": 3.000833332428843e-09, + "loss": 0.3021, + "step": 44944 + }, + { + "epoch": 0.989324235086947, + "grad_norm": 2.2788527948542683, + "learning_rate": 2.9884981579669793e-09, + "loss": 0.262, + "step": 44945 + }, + { + "epoch": 0.9893462469733656, + "grad_norm": 0.6250888647209294, + "learning_rate": 2.976188380545697e-09, + "loss": 0.4625, + "step": 44946 + }, + { + "epoch": 0.9893682588597843, + "grad_norm": 1.6547011197981223, + "learning_rate": 2.9639040002288342e-09, + "loss": 0.307, + "step": 44947 + }, + { + "epoch": 0.989390270746203, + "grad_norm": 2.0893393633742683, + "learning_rate": 2.9516450170780085e-09, + "loss": 0.2827, + "step": 44948 + }, + { + "epoch": 0.9894122826326216, + "grad_norm": 2.2217588036071327, + "learning_rate": 2.9394114311553925e-09, + "loss": 0.3491, + "step": 44949 + }, + { + "epoch": 0.9894342945190403, + "grad_norm": 2.0652856363360916, + "learning_rate": 2.927203242523713e-09, + "loss": 0.3209, + "step": 44950 + }, + { + "epoch": 0.9894563064054589, + "grad_norm": 0.6147565626440941, + "learning_rate": 2.915020451244588e-09, + "loss": 0.433, + "step": 44951 + }, + { + "epoch": 0.9894783182918776, + "grad_norm": 0.6455911194250742, + "learning_rate": 2.9028630573801896e-09, + "loss": 0.4442, + "step": 44952 + }, + { + "epoch": 0.9895003301782963, + "grad_norm": 0.6283742163508061, + "learning_rate": 2.890731060992136e-09, + "loss": 0.4264, + "step": 44953 + }, + { + "epoch": 0.9895223420647149, + "grad_norm": 1.831823987844624, + "learning_rate": 2.8786244621420435e-09, + "loss": 0.3132, + "step": 44954 + }, + { + "epoch": 0.9895443539511336, + "grad_norm": 5.037536663763637, + "learning_rate": 2.86654326089153e-09, + "loss": 0.2827, + "step": 44955 + }, + { + "epoch": 0.9895663658375523, + "grad_norm": 1.6108828301728335, + "learning_rate": 2.854487457302768e-09, + "loss": 0.2333, + "step": 44956 + }, + { + "epoch": 0.9895883777239709, + "grad_norm": 2.141362420851111, + "learning_rate": 2.8424570514357097e-09, + "loss": 0.249, + "step": 44957 + }, + { + "epoch": 0.9896103896103896, + "grad_norm": 1.9931876837768645, + "learning_rate": 2.8304520433519723e-09, + "loss": 0.2826, + "step": 44958 + }, + { + "epoch": 0.9896324014968083, + "grad_norm": 1.3524468105451928, + "learning_rate": 2.8184724331126177e-09, + "loss": 0.2635, + "step": 44959 + }, + { + "epoch": 0.9896544133832269, + "grad_norm": 1.450148691973942, + "learning_rate": 2.806518220778709e-09, + "loss": 0.281, + "step": 44960 + }, + { + "epoch": 0.9896764252696456, + "grad_norm": 3.5127921758421126, + "learning_rate": 2.7945894064113075e-09, + "loss": 0.2941, + "step": 44961 + }, + { + "epoch": 0.9896984371560643, + "grad_norm": 1.5834106624935624, + "learning_rate": 2.7826859900698113e-09, + "loss": 0.3184, + "step": 44962 + }, + { + "epoch": 0.9897204490424829, + "grad_norm": 1.932717582598362, + "learning_rate": 2.7708079718163916e-09, + "loss": 0.2671, + "step": 44963 + }, + { + "epoch": 0.9897424609289016, + "grad_norm": 2.5772210369985706, + "learning_rate": 2.758955351709891e-09, + "loss": 0.1909, + "step": 44964 + }, + { + "epoch": 0.9897644728153203, + "grad_norm": 3.1662483821298792, + "learning_rate": 2.7471281298113717e-09, + "loss": 0.2814, + "step": 44965 + }, + { + "epoch": 0.9897864847017389, + "grad_norm": 1.7822224177518173, + "learning_rate": 2.7353263061807857e-09, + "loss": 0.3566, + "step": 44966 + }, + { + "epoch": 0.9898084965881576, + "grad_norm": 1.5301712534708793, + "learning_rate": 2.72354988087864e-09, + "loss": 0.2717, + "step": 44967 + }, + { + "epoch": 0.9898305084745763, + "grad_norm": 2.4126533067047, + "learning_rate": 2.7117988539637764e-09, + "loss": 0.2385, + "step": 44968 + }, + { + "epoch": 0.9898525203609949, + "grad_norm": 1.6799914950762338, + "learning_rate": 2.700073225496147e-09, + "loss": 0.324, + "step": 44969 + }, + { + "epoch": 0.9898745322474136, + "grad_norm": 1.6243467735803916, + "learning_rate": 2.688372995535704e-09, + "loss": 0.2608, + "step": 44970 + }, + { + "epoch": 0.9898965441338323, + "grad_norm": 0.60394243069678, + "learning_rate": 2.6766981641423995e-09, + "loss": 0.4607, + "step": 44971 + }, + { + "epoch": 0.9899185560202509, + "grad_norm": 2.0710376309010776, + "learning_rate": 2.6650487313745198e-09, + "loss": 0.3531, + "step": 44972 + }, + { + "epoch": 0.9899405679066696, + "grad_norm": 0.6191811835451775, + "learning_rate": 2.6534246972920174e-09, + "loss": 0.4428, + "step": 44973 + }, + { + "epoch": 0.9899625797930882, + "grad_norm": 1.7296698362843554, + "learning_rate": 2.641826061953179e-09, + "loss": 0.2681, + "step": 44974 + }, + { + "epoch": 0.9899845916795069, + "grad_norm": 1.4775473643450667, + "learning_rate": 2.630252825417956e-09, + "loss": 0.2745, + "step": 44975 + }, + { + "epoch": 0.9900066035659256, + "grad_norm": 2.9753768567127263, + "learning_rate": 2.6187049877440805e-09, + "loss": 0.279, + "step": 44976 + }, + { + "epoch": 0.9900286154523442, + "grad_norm": 1.713150222500324, + "learning_rate": 2.6071825489915048e-09, + "loss": 0.2729, + "step": 44977 + }, + { + "epoch": 0.9900506273387629, + "grad_norm": 1.677779899919122, + "learning_rate": 2.59568550921796e-09, + "loss": 0.2866, + "step": 44978 + }, + { + "epoch": 0.9900726392251816, + "grad_norm": 1.84116056763687, + "learning_rate": 2.584213868481733e-09, + "loss": 0.3486, + "step": 44979 + }, + { + "epoch": 0.9900946511116002, + "grad_norm": 2.125018296301798, + "learning_rate": 2.5727676268416656e-09, + "loss": 0.2959, + "step": 44980 + }, + { + "epoch": 0.9901166629980189, + "grad_norm": 1.746621350286898, + "learning_rate": 2.5613467843560445e-09, + "loss": 0.3484, + "step": 44981 + }, + { + "epoch": 0.9901386748844376, + "grad_norm": 2.065960307280817, + "learning_rate": 2.5499513410820466e-09, + "loss": 0.2407, + "step": 44982 + }, + { + "epoch": 0.9901606867708562, + "grad_norm": 2.757916936264795, + "learning_rate": 2.5385812970785128e-09, + "loss": 0.3476, + "step": 44983 + }, + { + "epoch": 0.9901826986572749, + "grad_norm": 1.3556580599346337, + "learning_rate": 2.5272366524026203e-09, + "loss": 0.2517, + "step": 44984 + }, + { + "epoch": 0.9902047105436936, + "grad_norm": 1.596965384321747, + "learning_rate": 2.5159174071121007e-09, + "loss": 0.2674, + "step": 44985 + }, + { + "epoch": 0.9902267224301122, + "grad_norm": 0.6143188265167, + "learning_rate": 2.5046235612646852e-09, + "loss": 0.4589, + "step": 44986 + }, + { + "epoch": 0.9902487343165309, + "grad_norm": 2.5161449051627636, + "learning_rate": 2.493355114918106e-09, + "loss": 0.3599, + "step": 44987 + }, + { + "epoch": 0.9902707462029496, + "grad_norm": 2.221395137014655, + "learning_rate": 2.4821120681289834e-09, + "loss": 0.2872, + "step": 44988 + }, + { + "epoch": 0.9902927580893682, + "grad_norm": 1.4170761525780946, + "learning_rate": 2.470894420954495e-09, + "loss": 0.2119, + "step": 44989 + }, + { + "epoch": 0.9903147699757869, + "grad_norm": 1.7583495140922178, + "learning_rate": 2.459702173452372e-09, + "loss": 0.3857, + "step": 44990 + }, + { + "epoch": 0.9903367818622056, + "grad_norm": 1.4197428643774321, + "learning_rate": 2.4485353256786803e-09, + "loss": 0.2265, + "step": 44991 + }, + { + "epoch": 0.9903587937486242, + "grad_norm": 1.8027621719923412, + "learning_rate": 2.437393877690597e-09, + "loss": 0.4119, + "step": 44992 + }, + { + "epoch": 0.9903808056350429, + "grad_norm": 1.840153511537004, + "learning_rate": 2.426277829544188e-09, + "loss": 0.3554, + "step": 44993 + }, + { + "epoch": 0.9904028175214616, + "grad_norm": 1.7353741910189064, + "learning_rate": 2.415187181297185e-09, + "loss": 0.2601, + "step": 44994 + }, + { + "epoch": 0.9904248294078802, + "grad_norm": 2.2103841673810543, + "learning_rate": 2.404121933004544e-09, + "loss": 0.335, + "step": 44995 + }, + { + "epoch": 0.9904468412942989, + "grad_norm": 2.228057920387285, + "learning_rate": 2.3930820847234415e-09, + "loss": 0.3403, + "step": 44996 + }, + { + "epoch": 0.9904688531807175, + "grad_norm": 1.4655736615879202, + "learning_rate": 2.382067636509944e-09, + "loss": 0.2438, + "step": 44997 + }, + { + "epoch": 0.9904908650671362, + "grad_norm": 2.218594996284009, + "learning_rate": 2.371078588419562e-09, + "loss": 0.2668, + "step": 44998 + }, + { + "epoch": 0.9905128769535549, + "grad_norm": 2.325947437065025, + "learning_rate": 2.360114940508362e-09, + "loss": 0.3562, + "step": 44999 + }, + { + "epoch": 0.9905348888399735, + "grad_norm": 2.0588528860841455, + "learning_rate": 2.349176692831856e-09, + "loss": 0.2982, + "step": 45000 + }, + { + "epoch": 0.9905569007263922, + "grad_norm": 2.377422559352269, + "learning_rate": 2.3382638454461095e-09, + "loss": 0.314, + "step": 45001 + }, + { + "epoch": 0.990578912612811, + "grad_norm": 1.6440671921467669, + "learning_rate": 2.3273763984060784e-09, + "loss": 0.2713, + "step": 45002 + }, + { + "epoch": 0.9906009244992295, + "grad_norm": 1.499093803382969, + "learning_rate": 2.316514351767829e-09, + "loss": 0.2488, + "step": 45003 + }, + { + "epoch": 0.9906229363856482, + "grad_norm": 1.6181411412792543, + "learning_rate": 2.305677705585763e-09, + "loss": 0.2807, + "step": 45004 + }, + { + "epoch": 0.990644948272067, + "grad_norm": 1.7064616658336687, + "learning_rate": 2.294866459914835e-09, + "loss": 0.3604, + "step": 45005 + }, + { + "epoch": 0.9906669601584855, + "grad_norm": 1.5052871921201667, + "learning_rate": 2.2840806148111127e-09, + "loss": 0.2333, + "step": 45006 + }, + { + "epoch": 0.9906889720449042, + "grad_norm": 2.1704633141393157, + "learning_rate": 2.2733201703284415e-09, + "loss": 0.2697, + "step": 45007 + }, + { + "epoch": 0.990710983931323, + "grad_norm": 1.689847681525505, + "learning_rate": 2.262585126521777e-09, + "loss": 0.3554, + "step": 45008 + }, + { + "epoch": 0.9907329958177415, + "grad_norm": 1.308982231396875, + "learning_rate": 2.2518754834460754e-09, + "loss": 0.2469, + "step": 45009 + }, + { + "epoch": 0.9907550077041603, + "grad_norm": 4.793880855318524, + "learning_rate": 2.241191241154628e-09, + "loss": 0.2276, + "step": 45010 + }, + { + "epoch": 0.990777019590579, + "grad_norm": 0.6272950846201317, + "learning_rate": 2.230532399702945e-09, + "loss": 0.4312, + "step": 45011 + }, + { + "epoch": 0.9907990314769975, + "grad_norm": 2.393729482579479, + "learning_rate": 2.2198989591448728e-09, + "loss": 0.276, + "step": 45012 + }, + { + "epoch": 0.9908210433634163, + "grad_norm": 1.642959420786595, + "learning_rate": 2.209290919534812e-09, + "loss": 0.259, + "step": 45013 + }, + { + "epoch": 0.990843055249835, + "grad_norm": 2.226191523175612, + "learning_rate": 2.198708280925499e-09, + "loss": 0.3171, + "step": 45014 + }, + { + "epoch": 0.9908650671362536, + "grad_norm": 1.4275173228947045, + "learning_rate": 2.1881510433713336e-09, + "loss": 0.3007, + "step": 45015 + }, + { + "epoch": 0.9908870790226723, + "grad_norm": 1.5608288142108075, + "learning_rate": 2.1776192069267177e-09, + "loss": 0.2513, + "step": 45016 + }, + { + "epoch": 0.990909090909091, + "grad_norm": 1.907710227978751, + "learning_rate": 2.1671127716438313e-09, + "loss": 0.309, + "step": 45017 + }, + { + "epoch": 0.9909311027955096, + "grad_norm": 0.6290884123914839, + "learning_rate": 2.156631737577075e-09, + "loss": 0.4474, + "step": 45018 + }, + { + "epoch": 0.9909531146819283, + "grad_norm": 2.0742440874730392, + "learning_rate": 2.1461761047797405e-09, + "loss": 0.2507, + "step": 45019 + }, + { + "epoch": 0.9909751265683469, + "grad_norm": 1.6888069354734487, + "learning_rate": 2.1357458733045624e-09, + "loss": 0.2523, + "step": 45020 + }, + { + "epoch": 0.9909971384547656, + "grad_norm": 1.3649328621501116, + "learning_rate": 2.1253410432042765e-09, + "loss": 0.3661, + "step": 45021 + }, + { + "epoch": 0.9910191503411843, + "grad_norm": 1.5542046366502265, + "learning_rate": 2.1149616145321737e-09, + "loss": 0.205, + "step": 45022 + }, + { + "epoch": 0.9910411622276029, + "grad_norm": 0.6187799071688191, + "learning_rate": 2.10460758734099e-09, + "loss": 0.4477, + "step": 45023 + }, + { + "epoch": 0.9910631741140216, + "grad_norm": 4.133032917990254, + "learning_rate": 2.09427896168346e-09, + "loss": 0.2962, + "step": 45024 + }, + { + "epoch": 0.9910851860004403, + "grad_norm": 1.9569771832793839, + "learning_rate": 2.0839757376117655e-09, + "loss": 0.2778, + "step": 45025 + }, + { + "epoch": 0.9911071978868589, + "grad_norm": 1.6398006330117068, + "learning_rate": 2.0736979151786407e-09, + "loss": 0.2848, + "step": 45026 + }, + { + "epoch": 0.9911292097732776, + "grad_norm": 0.6217331222957336, + "learning_rate": 2.0634454944357117e-09, + "loss": 0.4664, + "step": 45027 + }, + { + "epoch": 0.9911512216596963, + "grad_norm": 0.6306663113800655, + "learning_rate": 2.0532184754362695e-09, + "loss": 0.4574, + "step": 45028 + }, + { + "epoch": 0.9911732335461149, + "grad_norm": 1.787249246712249, + "learning_rate": 2.0430168582308285e-09, + "loss": 0.2214, + "step": 45029 + }, + { + "epoch": 0.9911952454325336, + "grad_norm": 1.5643675919572495, + "learning_rate": 2.0328406428721247e-09, + "loss": 0.2968, + "step": 45030 + }, + { + "epoch": 0.9912172573189523, + "grad_norm": 1.7091086725515752, + "learning_rate": 2.0226898294117836e-09, + "loss": 0.2797, + "step": 45031 + }, + { + "epoch": 0.9912392692053709, + "grad_norm": 4.610006429375349, + "learning_rate": 2.012564417900875e-09, + "loss": 0.2179, + "step": 45032 + }, + { + "epoch": 0.9912612810917896, + "grad_norm": 1.8939091793250606, + "learning_rate": 2.002464408392135e-09, + "loss": 0.189, + "step": 45033 + }, + { + "epoch": 0.9912832929782083, + "grad_norm": 1.6383198664160366, + "learning_rate": 1.9923898009349684e-09, + "loss": 0.2701, + "step": 45034 + }, + { + "epoch": 0.9913053048646269, + "grad_norm": 2.7160205530274495, + "learning_rate": 1.982340595582666e-09, + "loss": 0.278, + "step": 45035 + }, + { + "epoch": 0.9913273167510456, + "grad_norm": 1.9562735266172544, + "learning_rate": 1.972316792384632e-09, + "loss": 0.2998, + "step": 45036 + }, + { + "epoch": 0.9913493286374643, + "grad_norm": 2.4343680194935593, + "learning_rate": 1.9623183913924928e-09, + "loss": 0.223, + "step": 45037 + }, + { + "epoch": 0.9913713405238829, + "grad_norm": 2.2120934559932324, + "learning_rate": 1.952345392657318e-09, + "loss": 0.2477, + "step": 45038 + }, + { + "epoch": 0.9913933524103016, + "grad_norm": 1.5280465374705454, + "learning_rate": 1.9423977962296226e-09, + "loss": 0.2455, + "step": 45039 + }, + { + "epoch": 0.9914153642967203, + "grad_norm": 1.6311548280751396, + "learning_rate": 1.932475602159367e-09, + "loss": 0.2662, + "step": 45040 + }, + { + "epoch": 0.9914373761831389, + "grad_norm": 3.3076598716583616, + "learning_rate": 1.9225788104976216e-09, + "loss": 0.3699, + "step": 45041 + }, + { + "epoch": 0.9914593880695576, + "grad_norm": 1.622969319080282, + "learning_rate": 1.9127074212949014e-09, + "loss": 0.3162, + "step": 45042 + }, + { + "epoch": 0.9914813999559762, + "grad_norm": 1.7870618403942051, + "learning_rate": 1.9028614346006112e-09, + "loss": 0.2287, + "step": 45043 + }, + { + "epoch": 0.9915034118423949, + "grad_norm": 2.4292064172501093, + "learning_rate": 1.8930408504652665e-09, + "loss": 0.2609, + "step": 45044 + }, + { + "epoch": 0.9915254237288136, + "grad_norm": 3.660350447404485, + "learning_rate": 1.8832456689388266e-09, + "loss": 0.3168, + "step": 45045 + }, + { + "epoch": 0.9915474356152322, + "grad_norm": 2.364789264953553, + "learning_rate": 1.873475890070697e-09, + "loss": 0.2373, + "step": 45046 + }, + { + "epoch": 0.9915694475016509, + "grad_norm": 1.7112149432195678, + "learning_rate": 1.863731513911393e-09, + "loss": 0.2998, + "step": 45047 + }, + { + "epoch": 0.9915914593880696, + "grad_norm": 1.554157768531767, + "learning_rate": 1.8540125405092091e-09, + "loss": 0.3035, + "step": 45048 + }, + { + "epoch": 0.9916134712744882, + "grad_norm": 2.0267210850976256, + "learning_rate": 1.8443189699146603e-09, + "loss": 0.3344, + "step": 45049 + }, + { + "epoch": 0.9916354831609069, + "grad_norm": 2.428716663779169, + "learning_rate": 1.8346508021760412e-09, + "loss": 0.3906, + "step": 45050 + }, + { + "epoch": 0.9916574950473256, + "grad_norm": 1.8152304047905807, + "learning_rate": 1.825008037343312e-09, + "loss": 0.2735, + "step": 45051 + }, + { + "epoch": 0.9916795069337442, + "grad_norm": 2.099201559526945, + "learning_rate": 1.8153906754647676e-09, + "loss": 0.2624, + "step": 45052 + }, + { + "epoch": 0.9917015188201629, + "grad_norm": 2.3372519218188104, + "learning_rate": 1.8057987165898127e-09, + "loss": 0.262, + "step": 45053 + }, + { + "epoch": 0.9917235307065816, + "grad_norm": 1.7001144148193508, + "learning_rate": 1.7962321607672973e-09, + "loss": 0.3477, + "step": 45054 + }, + { + "epoch": 0.9917455425930002, + "grad_norm": 1.634530974238391, + "learning_rate": 1.7866910080455158e-09, + "loss": 0.1946, + "step": 45055 + }, + { + "epoch": 0.9917675544794189, + "grad_norm": 2.657609176725094, + "learning_rate": 1.7771752584733183e-09, + "loss": 0.3616, + "step": 45056 + }, + { + "epoch": 0.9917895663658376, + "grad_norm": 1.7297705078159178, + "learning_rate": 1.767684912097889e-09, + "loss": 0.2895, + "step": 45057 + }, + { + "epoch": 0.9918115782522562, + "grad_norm": 1.602938452677573, + "learning_rate": 1.758219968969188e-09, + "loss": 0.2327, + "step": 45058 + }, + { + "epoch": 0.9918335901386749, + "grad_norm": 1.7102390316261757, + "learning_rate": 1.7487804291344002e-09, + "loss": 0.2895, + "step": 45059 + }, + { + "epoch": 0.9918556020250936, + "grad_norm": 1.6145648198708846, + "learning_rate": 1.7393662926412646e-09, + "loss": 0.311, + "step": 45060 + }, + { + "epoch": 0.9918776139115122, + "grad_norm": 1.722444659424954, + "learning_rate": 1.7299775595380762e-09, + "loss": 0.2556, + "step": 45061 + }, + { + "epoch": 0.9918996257979309, + "grad_norm": 1.6990532705606836, + "learning_rate": 1.7206142298725748e-09, + "loss": 0.2696, + "step": 45062 + }, + { + "epoch": 0.9919216376843496, + "grad_norm": 1.7124462573268333, + "learning_rate": 1.7112763036919444e-09, + "loss": 0.3325, + "step": 45063 + }, + { + "epoch": 0.9919436495707682, + "grad_norm": 1.8145382444926177, + "learning_rate": 1.701963781043925e-09, + "loss": 0.2888, + "step": 45064 + }, + { + "epoch": 0.9919656614571869, + "grad_norm": 1.6100875616007555, + "learning_rate": 1.6926766619757006e-09, + "loss": 0.3515, + "step": 45065 + }, + { + "epoch": 0.9919876733436055, + "grad_norm": 1.9764644854370408, + "learning_rate": 1.6834149465350114e-09, + "loss": 0.4062, + "step": 45066 + }, + { + "epoch": 0.9920096852300242, + "grad_norm": 2.8536578097617546, + "learning_rate": 1.6741786347684863e-09, + "loss": 0.3137, + "step": 45067 + }, + { + "epoch": 0.9920316971164429, + "grad_norm": 2.2465389270575473, + "learning_rate": 1.6649677267227548e-09, + "loss": 0.3252, + "step": 45068 + }, + { + "epoch": 0.9920537090028615, + "grad_norm": 2.293444018451402, + "learning_rate": 1.6557822224455566e-09, + "loss": 0.2853, + "step": 45069 + }, + { + "epoch": 0.9920757208892802, + "grad_norm": 1.422289935592587, + "learning_rate": 1.6466221219824107e-09, + "loss": 0.337, + "step": 45070 + }, + { + "epoch": 0.9920977327756989, + "grad_norm": 2.9937926336035874, + "learning_rate": 1.6374874253805017e-09, + "loss": 0.3044, + "step": 45071 + }, + { + "epoch": 0.9921197446621175, + "grad_norm": 2.367032719381599, + "learning_rate": 1.628378132686459e-09, + "loss": 0.2721, + "step": 45072 + }, + { + "epoch": 0.9921417565485362, + "grad_norm": 0.6174569821147778, + "learning_rate": 1.6192942439463567e-09, + "loss": 0.4508, + "step": 45073 + }, + { + "epoch": 0.9921637684349549, + "grad_norm": 1.6334295712686162, + "learning_rate": 1.6102357592062689e-09, + "loss": 0.2989, + "step": 45074 + }, + { + "epoch": 0.9921857803213735, + "grad_norm": 4.0667691539787665, + "learning_rate": 1.6012026785122703e-09, + "loss": 0.298, + "step": 45075 + }, + { + "epoch": 0.9922077922077922, + "grad_norm": 1.5497345432235639, + "learning_rate": 1.5921950019104348e-09, + "loss": 0.2592, + "step": 45076 + }, + { + "epoch": 0.9922298040942109, + "grad_norm": 1.5887987024117987, + "learning_rate": 1.5832127294462817e-09, + "loss": 0.2241, + "step": 45077 + }, + { + "epoch": 0.9922518159806295, + "grad_norm": 3.9502678875958357, + "learning_rate": 1.5742558611653303e-09, + "loss": 0.2502, + "step": 45078 + }, + { + "epoch": 0.9922738278670482, + "grad_norm": 1.7147952924738437, + "learning_rate": 1.5653243971136545e-09, + "loss": 0.37, + "step": 45079 + }, + { + "epoch": 0.9922958397534669, + "grad_norm": 2.036268989758489, + "learning_rate": 1.5564183373362184e-09, + "loss": 0.3248, + "step": 45080 + }, + { + "epoch": 0.9923178516398855, + "grad_norm": 2.028441080178414, + "learning_rate": 1.5475376818790965e-09, + "loss": 0.3852, + "step": 45081 + }, + { + "epoch": 0.9923398635263042, + "grad_norm": 2.2631669505525362, + "learning_rate": 1.5386824307861425e-09, + "loss": 0.2616, + "step": 45082 + }, + { + "epoch": 0.9923618754127229, + "grad_norm": 1.4774600342221738, + "learning_rate": 1.5298525841028756e-09, + "loss": 0.3028, + "step": 45083 + }, + { + "epoch": 0.9923838872991415, + "grad_norm": 1.925293123637957, + "learning_rate": 1.5210481418748146e-09, + "loss": 0.276, + "step": 45084 + }, + { + "epoch": 0.9924058991855602, + "grad_norm": 1.7986391804032995, + "learning_rate": 1.5122691041458138e-09, + "loss": 0.273, + "step": 45085 + }, + { + "epoch": 0.9924279110719789, + "grad_norm": 2.1421347844200187, + "learning_rate": 1.5035154709608367e-09, + "loss": 0.2327, + "step": 45086 + }, + { + "epoch": 0.9924499229583975, + "grad_norm": 1.9426944362918876, + "learning_rate": 1.4947872423642928e-09, + "loss": 0.3545, + "step": 45087 + }, + { + "epoch": 0.9924719348448162, + "grad_norm": 1.851830781713342, + "learning_rate": 1.4860844184005906e-09, + "loss": 0.2967, + "step": 45088 + }, + { + "epoch": 0.9924939467312349, + "grad_norm": 0.6144504641577887, + "learning_rate": 1.4774069991141392e-09, + "loss": 0.4466, + "step": 45089 + }, + { + "epoch": 0.9925159586176535, + "grad_norm": 0.6087218432498507, + "learning_rate": 1.4687549845493476e-09, + "loss": 0.4136, + "step": 45090 + }, + { + "epoch": 0.9925379705040722, + "grad_norm": 1.2931724711772195, + "learning_rate": 1.4601283747489593e-09, + "loss": 0.3394, + "step": 45091 + }, + { + "epoch": 0.9925599823904908, + "grad_norm": 1.9017033799100538, + "learning_rate": 1.451527169757938e-09, + "loss": 0.2832, + "step": 45092 + }, + { + "epoch": 0.9925819942769095, + "grad_norm": 1.9126563512066161, + "learning_rate": 1.442951369620138e-09, + "loss": 0.3687, + "step": 45093 + }, + { + "epoch": 0.9926040061633282, + "grad_norm": 1.9997753299294023, + "learning_rate": 1.4344009743783027e-09, + "loss": 0.3139, + "step": 45094 + }, + { + "epoch": 0.9926260180497468, + "grad_norm": 2.078767614745037, + "learning_rate": 1.425875984076286e-09, + "loss": 0.3069, + "step": 45095 + }, + { + "epoch": 0.9926480299361655, + "grad_norm": 1.768944801098093, + "learning_rate": 1.4173763987573863e-09, + "loss": 0.2403, + "step": 45096 + }, + { + "epoch": 0.9926700418225842, + "grad_norm": 1.860603846069273, + "learning_rate": 1.4089022184643474e-09, + "loss": 0.2656, + "step": 45097 + }, + { + "epoch": 0.9926920537090028, + "grad_norm": 1.3414601514350277, + "learning_rate": 1.4004534432410234e-09, + "loss": 0.2417, + "step": 45098 + }, + { + "epoch": 0.9927140655954215, + "grad_norm": 2.1530488386273317, + "learning_rate": 1.3920300731301572e-09, + "loss": 0.2966, + "step": 45099 + }, + { + "epoch": 0.9927360774818402, + "grad_norm": 4.145147535933143, + "learning_rate": 1.383632108174493e-09, + "loss": 0.2928, + "step": 45100 + }, + { + "epoch": 0.9927580893682588, + "grad_norm": 2.6280038221520323, + "learning_rate": 1.3752595484167742e-09, + "loss": 0.3626, + "step": 45101 + }, + { + "epoch": 0.9927801012546775, + "grad_norm": 1.738828916715768, + "learning_rate": 1.366912393899189e-09, + "loss": 0.2435, + "step": 45102 + }, + { + "epoch": 0.9928021131410962, + "grad_norm": 0.619615193250286, + "learning_rate": 1.3585906446644815e-09, + "loss": 0.442, + "step": 45103 + }, + { + "epoch": 0.9928241250275148, + "grad_norm": 1.7726134152976698, + "learning_rate": 1.3502943007548398e-09, + "loss": 0.3128, + "step": 45104 + }, + { + "epoch": 0.9928461369139335, + "grad_norm": 1.8730077067122932, + "learning_rate": 1.3420233622124523e-09, + "loss": 0.2595, + "step": 45105 + }, + { + "epoch": 0.9928681488003522, + "grad_norm": 1.5092532898784765, + "learning_rate": 1.3337778290795078e-09, + "loss": 0.1709, + "step": 45106 + }, + { + "epoch": 0.9928901606867708, + "grad_norm": 10.387404466829745, + "learning_rate": 1.3255577013976395e-09, + "loss": 0.2113, + "step": 45107 + }, + { + "epoch": 0.9929121725731895, + "grad_norm": 1.3258916434591481, + "learning_rate": 1.3173629792084807e-09, + "loss": 0.2542, + "step": 45108 + }, + { + "epoch": 0.9929341844596082, + "grad_norm": 0.614857899950779, + "learning_rate": 1.30919366255422e-09, + "loss": 0.4547, + "step": 45109 + }, + { + "epoch": 0.9929561963460268, + "grad_norm": 1.2119828634936751, + "learning_rate": 1.3010497514764908e-09, + "loss": 0.199, + "step": 45110 + }, + { + "epoch": 0.9929782082324455, + "grad_norm": 1.4669791228934583, + "learning_rate": 1.2929312460158161e-09, + "loss": 0.2643, + "step": 45111 + }, + { + "epoch": 0.9930002201188642, + "grad_norm": 1.908743680508302, + "learning_rate": 1.2848381462138292e-09, + "loss": 0.3257, + "step": 45112 + }, + { + "epoch": 0.9930222320052828, + "grad_norm": 1.9140179676601592, + "learning_rate": 1.2767704521121638e-09, + "loss": 0.3105, + "step": 45113 + }, + { + "epoch": 0.9930442438917015, + "grad_norm": 1.8992694987447962, + "learning_rate": 1.2687281637513428e-09, + "loss": 0.3192, + "step": 45114 + }, + { + "epoch": 0.9930662557781201, + "grad_norm": 1.424629887819095, + "learning_rate": 1.2607112811724441e-09, + "loss": 0.2116, + "step": 45115 + }, + { + "epoch": 0.9930882676645388, + "grad_norm": 2.6233699086680295, + "learning_rate": 1.2527198044159915e-09, + "loss": 0.2902, + "step": 45116 + }, + { + "epoch": 0.9931102795509575, + "grad_norm": 1.6168946089532656, + "learning_rate": 1.244753733522508e-09, + "loss": 0.3114, + "step": 45117 + }, + { + "epoch": 0.9931322914373761, + "grad_norm": 1.2853171320657701, + "learning_rate": 1.2368130685330715e-09, + "loss": 0.2255, + "step": 45118 + }, + { + "epoch": 0.9931543033237948, + "grad_norm": 1.6543815255842422, + "learning_rate": 1.2288978094876503e-09, + "loss": 0.2842, + "step": 45119 + }, + { + "epoch": 0.9931763152102135, + "grad_norm": 2.6591863861732223, + "learning_rate": 1.2210079564262122e-09, + "loss": 0.2549, + "step": 45120 + }, + { + "epoch": 0.9931983270966321, + "grad_norm": 1.6524716511814752, + "learning_rate": 1.2131435093892807e-09, + "loss": 0.2635, + "step": 45121 + }, + { + "epoch": 0.9932203389830508, + "grad_norm": 1.4531651839069581, + "learning_rate": 1.2053044684168235e-09, + "loss": 0.2454, + "step": 45122 + }, + { + "epoch": 0.9932423508694695, + "grad_norm": 1.7104885062747393, + "learning_rate": 1.1974908335482538e-09, + "loss": 0.2504, + "step": 45123 + }, + { + "epoch": 0.9932643627558881, + "grad_norm": 2.0783326708949215, + "learning_rate": 1.1897026048235395e-09, + "loss": 0.3675, + "step": 45124 + }, + { + "epoch": 0.9932863746423068, + "grad_norm": 1.8696557588134222, + "learning_rate": 1.1819397822826483e-09, + "loss": 0.3443, + "step": 45125 + }, + { + "epoch": 0.9933083865287256, + "grad_norm": 2.0161608824705843, + "learning_rate": 1.1742023659644386e-09, + "loss": 0.2957, + "step": 45126 + }, + { + "epoch": 0.9933303984151441, + "grad_norm": 1.5603349621773206, + "learning_rate": 1.1664903559083229e-09, + "loss": 0.3534, + "step": 45127 + }, + { + "epoch": 0.9933524103015628, + "grad_norm": 0.5976669132802676, + "learning_rate": 1.1588037521537144e-09, + "loss": 0.4514, + "step": 45128 + }, + { + "epoch": 0.9933744221879816, + "grad_norm": 0.5979758388846707, + "learning_rate": 1.1511425547394705e-09, + "loss": 0.4136, + "step": 45129 + }, + { + "epoch": 0.9933964340744001, + "grad_norm": 2.2879659516881623, + "learning_rate": 1.1435067637050046e-09, + "loss": 0.2867, + "step": 45130 + }, + { + "epoch": 0.9934184459608189, + "grad_norm": 1.6573466154423924, + "learning_rate": 1.135896379088619e-09, + "loss": 0.2292, + "step": 45131 + }, + { + "epoch": 0.9934404578472376, + "grad_norm": 0.5968378296032141, + "learning_rate": 1.1283114009291718e-09, + "loss": 0.4266, + "step": 45132 + }, + { + "epoch": 0.9934624697336562, + "grad_norm": 0.6234376943605082, + "learning_rate": 1.1207518292655207e-09, + "loss": 0.461, + "step": 45133 + }, + { + "epoch": 0.9934844816200749, + "grad_norm": 2.3602120732092646, + "learning_rate": 1.113217664135413e-09, + "loss": 0.3596, + "step": 45134 + }, + { + "epoch": 0.9935064935064936, + "grad_norm": 1.4834330923370507, + "learning_rate": 1.105708905577707e-09, + "loss": 0.3147, + "step": 45135 + }, + { + "epoch": 0.9935285053929122, + "grad_norm": 1.8444762117338913, + "learning_rate": 1.09822555363015e-09, + "loss": 0.3194, + "step": 45136 + }, + { + "epoch": 0.9935505172793309, + "grad_norm": 1.8219062472588115, + "learning_rate": 1.0907676083310449e-09, + "loss": 0.3247, + "step": 45137 + }, + { + "epoch": 0.9935725291657495, + "grad_norm": 1.6846533800139718, + "learning_rate": 1.0833350697181389e-09, + "loss": 0.2893, + "step": 45138 + }, + { + "epoch": 0.9935945410521682, + "grad_norm": 1.7767614867729453, + "learning_rate": 1.0759279378297349e-09, + "loss": 0.3022, + "step": 45139 + }, + { + "epoch": 0.9936165529385869, + "grad_norm": 1.546214504286705, + "learning_rate": 1.0685462127024705e-09, + "loss": 0.3088, + "step": 45140 + }, + { + "epoch": 0.9936385648250055, + "grad_norm": 3.9934463204128834, + "learning_rate": 1.061189894374648e-09, + "loss": 0.3036, + "step": 45141 + }, + { + "epoch": 0.9936605767114242, + "grad_norm": 1.64603399496661, + "learning_rate": 1.05385898288346e-09, + "loss": 0.2711, + "step": 45142 + }, + { + "epoch": 0.9936825885978429, + "grad_norm": 2.472347625128144, + "learning_rate": 1.0465534782666543e-09, + "loss": 0.3484, + "step": 45143 + }, + { + "epoch": 0.9937046004842615, + "grad_norm": 1.822204856620711, + "learning_rate": 1.0392733805603127e-09, + "loss": 0.2944, + "step": 45144 + }, + { + "epoch": 0.9937266123706802, + "grad_norm": 0.5989833078446991, + "learning_rate": 1.0320186898016283e-09, + "loss": 0.4137, + "step": 45145 + }, + { + "epoch": 0.9937486242570989, + "grad_norm": 0.6310421160137624, + "learning_rate": 1.0247894060283481e-09, + "loss": 0.451, + "step": 45146 + }, + { + "epoch": 0.9937706361435175, + "grad_norm": 1.9643561965962957, + "learning_rate": 1.0175855292765546e-09, + "loss": 0.1888, + "step": 45147 + }, + { + "epoch": 0.9937926480299362, + "grad_norm": 0.6105853109556828, + "learning_rate": 1.0104070595834403e-09, + "loss": 0.4685, + "step": 45148 + }, + { + "epoch": 0.9938146599163549, + "grad_norm": 1.6508400642181735, + "learning_rate": 1.0032539969845324e-09, + "loss": 0.258, + "step": 45149 + }, + { + "epoch": 0.9938366718027735, + "grad_norm": 2.6376456607175114, + "learning_rate": 9.96126341517023e-10, + "loss": 0.3219, + "step": 45150 + }, + { + "epoch": 0.9938586836891922, + "grad_norm": 0.62660753936546, + "learning_rate": 9.890240932164397e-10, + "loss": 0.4487, + "step": 45151 + }, + { + "epoch": 0.9938806955756109, + "grad_norm": 1.697169598313932, + "learning_rate": 9.819472521194196e-10, + "loss": 0.2734, + "step": 45152 + }, + { + "epoch": 0.9939027074620295, + "grad_norm": 0.5932678169774903, + "learning_rate": 9.74895818262045e-10, + "loss": 0.4192, + "step": 45153 + }, + { + "epoch": 0.9939247193484482, + "grad_norm": 1.575233955390057, + "learning_rate": 9.678697916792879e-10, + "loss": 0.3095, + "step": 45154 + }, + { + "epoch": 0.9939467312348669, + "grad_norm": 1.6718482380370263, + "learning_rate": 9.608691724077857e-10, + "loss": 0.3479, + "step": 45155 + }, + { + "epoch": 0.9939687431212855, + "grad_norm": 1.6489653549598426, + "learning_rate": 9.538939604830656e-10, + "loss": 0.2853, + "step": 45156 + }, + { + "epoch": 0.9939907550077042, + "grad_norm": 6.236467006890881, + "learning_rate": 9.469441559400993e-10, + "loss": 0.27, + "step": 45157 + }, + { + "epoch": 0.9940127668941229, + "grad_norm": 1.4600008920224972, + "learning_rate": 9.400197588144144e-10, + "loss": 0.3309, + "step": 45158 + }, + { + "epoch": 0.9940347787805415, + "grad_norm": 1.6799747536719902, + "learning_rate": 9.331207691409828e-10, + "loss": 0.267, + "step": 45159 + }, + { + "epoch": 0.9940567906669602, + "grad_norm": 1.9745870446110405, + "learning_rate": 9.262471869547762e-10, + "loss": 0.3145, + "step": 45160 + }, + { + "epoch": 0.9940788025533788, + "grad_norm": 1.8829318810956321, + "learning_rate": 9.193990122918773e-10, + "loss": 0.2617, + "step": 45161 + }, + { + "epoch": 0.9941008144397975, + "grad_norm": 0.6147186719138236, + "learning_rate": 9.125762451855924e-10, + "loss": 0.4315, + "step": 45162 + }, + { + "epoch": 0.9941228263262162, + "grad_norm": 1.5987936652148158, + "learning_rate": 9.057788856714489e-10, + "loss": 0.3198, + "step": 45163 + }, + { + "epoch": 0.9941448382126348, + "grad_norm": 2.0309757472123326, + "learning_rate": 8.990069337838636e-10, + "loss": 0.2942, + "step": 45164 + }, + { + "epoch": 0.9941668500990535, + "grad_norm": 1.8703225524335114, + "learning_rate": 8.922603895572535e-10, + "loss": 0.2375, + "step": 45165 + }, + { + "epoch": 0.9941888619854722, + "grad_norm": 1.6422408041263232, + "learning_rate": 8.855392530254803e-10, + "loss": 0.2792, + "step": 45166 + }, + { + "epoch": 0.9942108738718908, + "grad_norm": 1.4872898773126366, + "learning_rate": 8.788435242235161e-10, + "loss": 0.2808, + "step": 45167 + }, + { + "epoch": 0.9942328857583095, + "grad_norm": 2.82651764024128, + "learning_rate": 8.721732031846675e-10, + "loss": 0.2613, + "step": 45168 + }, + { + "epoch": 0.9942548976447282, + "grad_norm": 2.7516790393794115, + "learning_rate": 8.655282899433515e-10, + "loss": 0.3182, + "step": 45169 + }, + { + "epoch": 0.9942769095311468, + "grad_norm": 0.602568472909131, + "learning_rate": 8.589087845328747e-10, + "loss": 0.4343, + "step": 45170 + }, + { + "epoch": 0.9942989214175655, + "grad_norm": 0.6079085152591717, + "learning_rate": 8.523146869870991e-10, + "loss": 0.4127, + "step": 45171 + }, + { + "epoch": 0.9943209333039842, + "grad_norm": 2.0243788267633245, + "learning_rate": 8.457459973393311e-10, + "loss": 0.3264, + "step": 45172 + }, + { + "epoch": 0.9943429451904028, + "grad_norm": 0.6243056668553482, + "learning_rate": 8.392027156234328e-10, + "loss": 0.4425, + "step": 45173 + }, + { + "epoch": 0.9943649570768215, + "grad_norm": 3.107774608562057, + "learning_rate": 8.326848418727107e-10, + "loss": 0.3497, + "step": 45174 + }, + { + "epoch": 0.9943869689632402, + "grad_norm": 1.6239894154789922, + "learning_rate": 8.261923761193613e-10, + "loss": 0.318, + "step": 45175 + }, + { + "epoch": 0.9944089808496588, + "grad_norm": 1.4902526721818443, + "learning_rate": 8.197253183972464e-10, + "loss": 0.261, + "step": 45176 + }, + { + "epoch": 0.9944309927360775, + "grad_norm": 1.9686341754217165, + "learning_rate": 8.132836687391177e-10, + "loss": 0.2476, + "step": 45177 + }, + { + "epoch": 0.9944530046224962, + "grad_norm": 1.494816187396001, + "learning_rate": 8.068674271777266e-10, + "loss": 0.3351, + "step": 45178 + }, + { + "epoch": 0.9944750165089148, + "grad_norm": 1.7525356375367727, + "learning_rate": 8.004765937452696e-10, + "loss": 0.3275, + "step": 45179 + }, + { + "epoch": 0.9944970283953335, + "grad_norm": 1.6062494383637906, + "learning_rate": 7.941111684744984e-10, + "loss": 0.3043, + "step": 45180 + }, + { + "epoch": 0.9945190402817522, + "grad_norm": 0.6289393817372086, + "learning_rate": 7.877711513976094e-10, + "loss": 0.4295, + "step": 45181 + }, + { + "epoch": 0.9945410521681708, + "grad_norm": 1.7684502789127234, + "learning_rate": 7.814565425473542e-10, + "loss": 0.297, + "step": 45182 + }, + { + "epoch": 0.9945630640545895, + "grad_norm": 1.8688475295546236, + "learning_rate": 7.75167341955374e-10, + "loss": 0.2625, + "step": 45183 + }, + { + "epoch": 0.9945850759410081, + "grad_norm": 2.013230127934991, + "learning_rate": 7.689035496533104e-10, + "loss": 0.34, + "step": 45184 + }, + { + "epoch": 0.9946070878274268, + "grad_norm": 1.9108750118289672, + "learning_rate": 7.626651656739148e-10, + "loss": 0.3168, + "step": 45185 + }, + { + "epoch": 0.9946290997138455, + "grad_norm": 1.7844793424885994, + "learning_rate": 7.564521900477184e-10, + "loss": 0.2523, + "step": 45186 + }, + { + "epoch": 0.9946511116002641, + "grad_norm": 0.6260497123892345, + "learning_rate": 7.502646228074728e-10, + "loss": 0.4462, + "step": 45187 + }, + { + "epoch": 0.9946731234866828, + "grad_norm": 1.790277951084576, + "learning_rate": 7.441024639842643e-10, + "loss": 0.2871, + "step": 45188 + }, + { + "epoch": 0.9946951353731015, + "grad_norm": 1.296050374115289, + "learning_rate": 7.379657136091789e-10, + "loss": 0.2371, + "step": 45189 + }, + { + "epoch": 0.9947171472595201, + "grad_norm": 2.0683159771406334, + "learning_rate": 7.318543717133031e-10, + "loss": 0.3222, + "step": 45190 + }, + { + "epoch": 0.9947391591459388, + "grad_norm": 2.289020483313191, + "learning_rate": 7.25768438327723e-10, + "loss": 0.2847, + "step": 45191 + }, + { + "epoch": 0.9947611710323575, + "grad_norm": 0.5976327708998613, + "learning_rate": 7.197079134835249e-10, + "loss": 0.449, + "step": 45192 + }, + { + "epoch": 0.9947831829187761, + "grad_norm": 2.1320998373692017, + "learning_rate": 7.136727972117952e-10, + "loss": 0.2523, + "step": 45193 + }, + { + "epoch": 0.9948051948051948, + "grad_norm": 1.7570514973131846, + "learning_rate": 7.076630895430647e-10, + "loss": 0.4013, + "step": 45194 + }, + { + "epoch": 0.9948272066916135, + "grad_norm": 1.378740360110047, + "learning_rate": 7.016787905073096e-10, + "loss": 0.2446, + "step": 45195 + }, + { + "epoch": 0.9948492185780321, + "grad_norm": 2.0123929378525443, + "learning_rate": 6.957199001356163e-10, + "loss": 0.2969, + "step": 45196 + }, + { + "epoch": 0.9948712304644508, + "grad_norm": 0.566012062763516, + "learning_rate": 6.897864184579605e-10, + "loss": 0.4109, + "step": 45197 + }, + { + "epoch": 0.9948932423508695, + "grad_norm": 0.6145364177624765, + "learning_rate": 6.838783455043186e-10, + "loss": 0.4616, + "step": 45198 + }, + { + "epoch": 0.9949152542372881, + "grad_norm": 0.6163935745742665, + "learning_rate": 6.779956813052213e-10, + "loss": 0.451, + "step": 45199 + }, + { + "epoch": 0.9949372661237068, + "grad_norm": 1.5643831379167175, + "learning_rate": 6.721384258906449e-10, + "loss": 0.3766, + "step": 45200 + }, + { + "epoch": 0.9949592780101255, + "grad_norm": 1.5499078320736626, + "learning_rate": 6.663065792894552e-10, + "loss": 0.3331, + "step": 45201 + }, + { + "epoch": 0.9949812898965441, + "grad_norm": 2.2541392249091903, + "learning_rate": 6.605001415321832e-10, + "loss": 0.3194, + "step": 45202 + }, + { + "epoch": 0.9950033017829628, + "grad_norm": 1.4381819630811161, + "learning_rate": 6.547191126476949e-10, + "loss": 0.2725, + "step": 45203 + }, + { + "epoch": 0.9950253136693815, + "grad_norm": 1.558227784912863, + "learning_rate": 6.48963492665966e-10, + "loss": 0.269, + "step": 45204 + }, + { + "epoch": 0.9950473255558001, + "grad_norm": 0.6203338055463765, + "learning_rate": 6.432332816158626e-10, + "loss": 0.4477, + "step": 45205 + }, + { + "epoch": 0.9950693374422188, + "grad_norm": 1.6494714139100324, + "learning_rate": 6.375284795268056e-10, + "loss": 0.3106, + "step": 45206 + }, + { + "epoch": 0.9950913493286374, + "grad_norm": 2.135225151230389, + "learning_rate": 6.318490864276605e-10, + "loss": 0.3, + "step": 45207 + }, + { + "epoch": 0.9951133612150561, + "grad_norm": 0.6197544348016025, + "learning_rate": 6.261951023467383e-10, + "loss": 0.4381, + "step": 45208 + }, + { + "epoch": 0.9951353731014748, + "grad_norm": 2.0239743230153464, + "learning_rate": 6.205665273134598e-10, + "loss": 0.2785, + "step": 45209 + }, + { + "epoch": 0.9951573849878934, + "grad_norm": 1.661606173593899, + "learning_rate": 6.149633613566908e-10, + "loss": 0.3157, + "step": 45210 + }, + { + "epoch": 0.9951793968743121, + "grad_norm": 2.1242831289438264, + "learning_rate": 6.093856045041869e-10, + "loss": 0.2417, + "step": 45211 + }, + { + "epoch": 0.9952014087607308, + "grad_norm": 2.039309110038017, + "learning_rate": 6.038332567842587e-10, + "loss": 0.3212, + "step": 45212 + }, + { + "epoch": 0.9952234206471494, + "grad_norm": 1.6517778602967867, + "learning_rate": 5.983063182257721e-10, + "loss": 0.2907, + "step": 45213 + }, + { + "epoch": 0.9952454325335681, + "grad_norm": 0.6224894218912662, + "learning_rate": 5.928047888564826e-10, + "loss": 0.4671, + "step": 45214 + }, + { + "epoch": 0.9952674444199868, + "grad_norm": 2.230397276559286, + "learning_rate": 5.873286687047009e-10, + "loss": 0.3191, + "step": 45215 + }, + { + "epoch": 0.9952894563064054, + "grad_norm": 7.095334375756164, + "learning_rate": 5.818779577970723e-10, + "loss": 0.3426, + "step": 45216 + }, + { + "epoch": 0.9953114681928241, + "grad_norm": 1.7196399033147773, + "learning_rate": 5.76452656163018e-10, + "loss": 0.2716, + "step": 45217 + }, + { + "epoch": 0.9953334800792428, + "grad_norm": 1.6618439092832251, + "learning_rate": 5.710527638286278e-10, + "loss": 0.3542, + "step": 45218 + }, + { + "epoch": 0.9953554919656614, + "grad_norm": 1.6949659164516717, + "learning_rate": 5.656782808222127e-10, + "loss": 0.3534, + "step": 45219 + }, + { + "epoch": 0.9953775038520801, + "grad_norm": 0.6319454634224978, + "learning_rate": 5.603292071709732e-10, + "loss": 0.4366, + "step": 45220 + }, + { + "epoch": 0.9953995157384988, + "grad_norm": 1.9097501999760997, + "learning_rate": 5.550055429015544e-10, + "loss": 0.2917, + "step": 45221 + }, + { + "epoch": 0.9954215276249174, + "grad_norm": 2.0042985656303682, + "learning_rate": 5.49707288041712e-10, + "loss": 0.2754, + "step": 45222 + }, + { + "epoch": 0.9954435395113361, + "grad_norm": 1.5790921842346133, + "learning_rate": 5.444344426180914e-10, + "loss": 0.2706, + "step": 45223 + }, + { + "epoch": 0.9954655513977548, + "grad_norm": 2.3005522873615663, + "learning_rate": 5.391870066573379e-10, + "loss": 0.3965, + "step": 45224 + }, + { + "epoch": 0.9954875632841734, + "grad_norm": 1.3717857807454803, + "learning_rate": 5.339649801860969e-10, + "loss": 0.2661, + "step": 45225 + }, + { + "epoch": 0.9955095751705921, + "grad_norm": 1.7438020310305753, + "learning_rate": 5.287683632310136e-10, + "loss": 0.3277, + "step": 45226 + }, + { + "epoch": 0.9955315870570108, + "grad_norm": 2.5898815902815873, + "learning_rate": 5.235971558187336e-10, + "loss": 0.3381, + "step": 45227 + }, + { + "epoch": 0.9955535989434294, + "grad_norm": 2.1150759703121293, + "learning_rate": 5.18451357975902e-10, + "loss": 0.2203, + "step": 45228 + }, + { + "epoch": 0.9955756108298481, + "grad_norm": 2.4100382793939183, + "learning_rate": 5.133309697274991e-10, + "loss": 0.3038, + "step": 45229 + }, + { + "epoch": 0.9955976227162668, + "grad_norm": 1.7511597565954728, + "learning_rate": 5.0823599110017e-10, + "loss": 0.3312, + "step": 45230 + }, + { + "epoch": 0.9956196346026854, + "grad_norm": 1.596854279789049, + "learning_rate": 5.03166422120005e-10, + "loss": 0.2666, + "step": 45231 + }, + { + "epoch": 0.9956416464891041, + "grad_norm": 1.9925498131058361, + "learning_rate": 4.981222628125393e-10, + "loss": 0.289, + "step": 45232 + }, + { + "epoch": 0.9956636583755227, + "grad_norm": 1.505868405284666, + "learning_rate": 4.93103513203863e-10, + "loss": 0.193, + "step": 45233 + }, + { + "epoch": 0.9956856702619414, + "grad_norm": 0.6272901020313997, + "learning_rate": 4.881101733184013e-10, + "loss": 0.4606, + "step": 45234 + }, + { + "epoch": 0.9957076821483601, + "grad_norm": 1.5668267337993582, + "learning_rate": 4.831422431827992e-10, + "loss": 0.2147, + "step": 45235 + }, + { + "epoch": 0.9957296940347787, + "grad_norm": 0.5892429980175804, + "learning_rate": 4.781997228214819e-10, + "loss": 0.421, + "step": 45236 + }, + { + "epoch": 0.9957517059211974, + "grad_norm": 1.7371748507986258, + "learning_rate": 4.732826122594291e-10, + "loss": 0.3194, + "step": 45237 + }, + { + "epoch": 0.9957737178076161, + "grad_norm": 1.5237257160749904, + "learning_rate": 4.683909115221763e-10, + "loss": 0.3075, + "step": 45238 + }, + { + "epoch": 0.9957957296940347, + "grad_norm": 0.6053444835645901, + "learning_rate": 4.6352462063470326e-10, + "loss": 0.4241, + "step": 45239 + }, + { + "epoch": 0.9958177415804534, + "grad_norm": 1.6143579764980431, + "learning_rate": 4.58683739621435e-10, + "loss": 0.3138, + "step": 45240 + }, + { + "epoch": 0.9958397534668721, + "grad_norm": 1.414531356098899, + "learning_rate": 4.538682685067963e-10, + "loss": 0.2405, + "step": 45241 + }, + { + "epoch": 0.9958617653532907, + "grad_norm": 1.488348215744275, + "learning_rate": 4.4907820731576733e-10, + "loss": 0.2488, + "step": 45242 + }, + { + "epoch": 0.9958837772397094, + "grad_norm": 2.2156245343260075, + "learning_rate": 4.443135560716627e-10, + "loss": 0.3899, + "step": 45243 + }, + { + "epoch": 0.9959057891261281, + "grad_norm": 0.6329449877373914, + "learning_rate": 4.3957431480001756e-10, + "loss": 0.4573, + "step": 45244 + }, + { + "epoch": 0.9959278010125467, + "grad_norm": 1.9781195738705704, + "learning_rate": 4.3486048352414655e-10, + "loss": 0.2136, + "step": 45245 + }, + { + "epoch": 0.9959498128989654, + "grad_norm": 2.482209241899286, + "learning_rate": 4.3017206226847464e-10, + "loss": 0.375, + "step": 45246 + }, + { + "epoch": 0.9959718247853842, + "grad_norm": 1.685985802634005, + "learning_rate": 4.2550905105631643e-10, + "loss": 0.2059, + "step": 45247 + }, + { + "epoch": 0.9959938366718027, + "grad_norm": 2.265504582796436, + "learning_rate": 4.2087144991154184e-10, + "loss": 0.3101, + "step": 45248 + }, + { + "epoch": 0.9960158485582215, + "grad_norm": 3.7315108993449373, + "learning_rate": 4.1625925885802053e-10, + "loss": 0.2571, + "step": 45249 + }, + { + "epoch": 0.9960378604446402, + "grad_norm": 1.9516988104185116, + "learning_rate": 4.1167247791851214e-10, + "loss": 0.4032, + "step": 45250 + }, + { + "epoch": 0.9960598723310587, + "grad_norm": 1.7320249824003804, + "learning_rate": 4.071111071168865e-10, + "loss": 0.2493, + "step": 45251 + }, + { + "epoch": 0.9960818842174775, + "grad_norm": 1.7622919447124972, + "learning_rate": 4.0257514647645825e-10, + "loss": 0.2817, + "step": 45252 + }, + { + "epoch": 0.9961038961038962, + "grad_norm": 1.258973440696293, + "learning_rate": 3.980645960199869e-10, + "loss": 0.2714, + "step": 45253 + }, + { + "epoch": 0.9961259079903148, + "grad_norm": 4.362961186058614, + "learning_rate": 3.935794557702322e-10, + "loss": 0.2451, + "step": 45254 + }, + { + "epoch": 0.9961479198767335, + "grad_norm": 1.9013355563009862, + "learning_rate": 3.891197257499535e-10, + "loss": 0.282, + "step": 45255 + }, + { + "epoch": 0.996169931763152, + "grad_norm": 2.1044458142249716, + "learning_rate": 3.846854059824656e-10, + "loss": 0.2432, + "step": 45256 + }, + { + "epoch": 0.9961919436495708, + "grad_norm": 0.594022009314769, + "learning_rate": 3.80276496489973e-10, + "loss": 0.4281, + "step": 45257 + }, + { + "epoch": 0.9962139555359895, + "grad_norm": 2.0839524896818964, + "learning_rate": 3.758929972946801e-10, + "loss": 0.2075, + "step": 45258 + }, + { + "epoch": 0.996235967422408, + "grad_norm": 1.571312873284239, + "learning_rate": 3.715349084187914e-10, + "loss": 0.3346, + "step": 45259 + }, + { + "epoch": 0.9962579793088268, + "grad_norm": 2.653065986656621, + "learning_rate": 3.672022298850664e-10, + "loss": 0.2918, + "step": 45260 + }, + { + "epoch": 0.9962799911952455, + "grad_norm": 1.6737068622346623, + "learning_rate": 3.628949617151545e-10, + "loss": 0.2766, + "step": 45261 + }, + { + "epoch": 0.9963020030816641, + "grad_norm": 2.9266115652624176, + "learning_rate": 3.586131039301499e-10, + "loss": 0.3442, + "step": 45262 + }, + { + "epoch": 0.9963240149680828, + "grad_norm": 0.618089137822312, + "learning_rate": 3.543566565533674e-10, + "loss": 0.462, + "step": 45263 + }, + { + "epoch": 0.9963460268545015, + "grad_norm": 2.875633529284278, + "learning_rate": 3.5012561960534597e-10, + "loss": 0.3224, + "step": 45264 + }, + { + "epoch": 0.9963680387409201, + "grad_norm": 1.813240129784026, + "learning_rate": 3.459199931077351e-10, + "loss": 0.329, + "step": 45265 + }, + { + "epoch": 0.9963900506273388, + "grad_norm": 1.2487017258209336, + "learning_rate": 3.41739777082184e-10, + "loss": 0.2916, + "step": 45266 + }, + { + "epoch": 0.9964120625137575, + "grad_norm": 1.5944202530062022, + "learning_rate": 3.37584971549787e-10, + "loss": 0.3599, + "step": 45267 + }, + { + "epoch": 0.9964340744001761, + "grad_norm": 1.802623454474789, + "learning_rate": 3.334555765321934e-10, + "loss": 0.2474, + "step": 45268 + }, + { + "epoch": 0.9964560862865948, + "grad_norm": 2.879307644235833, + "learning_rate": 3.293515920493873e-10, + "loss": 0.251, + "step": 45269 + }, + { + "epoch": 0.9964780981730135, + "grad_norm": 2.470636448560468, + "learning_rate": 3.252730181224628e-10, + "loss": 0.304, + "step": 45270 + }, + { + "epoch": 0.9965001100594321, + "grad_norm": 0.6195045841281963, + "learning_rate": 3.212198547730694e-10, + "loss": 0.4616, + "step": 45271 + }, + { + "epoch": 0.9965221219458508, + "grad_norm": 2.808499671549883, + "learning_rate": 3.171921020206359e-10, + "loss": 0.3566, + "step": 45272 + }, + { + "epoch": 0.9965441338322695, + "grad_norm": 1.609223217367789, + "learning_rate": 3.1318975988681165e-10, + "loss": 0.2872, + "step": 45273 + }, + { + "epoch": 0.9965661457186881, + "grad_norm": 0.6447331233650848, + "learning_rate": 3.092128283904705e-10, + "loss": 0.4603, + "step": 45274 + }, + { + "epoch": 0.9965881576051068, + "grad_norm": 1.6107263474721238, + "learning_rate": 3.0526130755326177e-10, + "loss": 0.3024, + "step": 45275 + }, + { + "epoch": 0.9966101694915255, + "grad_norm": 2.3727864357674786, + "learning_rate": 3.0133519739405923e-10, + "loss": 0.3392, + "step": 45276 + }, + { + "epoch": 0.9966321813779441, + "grad_norm": 0.6053783650908732, + "learning_rate": 2.974344979339572e-10, + "loss": 0.4563, + "step": 45277 + }, + { + "epoch": 0.9966541932643628, + "grad_norm": 5.33478289992028, + "learning_rate": 2.9355920919182936e-10, + "loss": 0.2535, + "step": 45278 + }, + { + "epoch": 0.9966762051507814, + "grad_norm": 1.9261422522601859, + "learning_rate": 2.8970933118821485e-10, + "loss": 0.3381, + "step": 45279 + }, + { + "epoch": 0.9966982170372001, + "grad_norm": 1.4376254841888554, + "learning_rate": 2.8588486394198756e-10, + "loss": 0.2516, + "step": 45280 + }, + { + "epoch": 0.9967202289236188, + "grad_norm": 1.6522734932843042, + "learning_rate": 2.820858074725763e-10, + "loss": 0.2902, + "step": 45281 + }, + { + "epoch": 0.9967422408100374, + "grad_norm": 0.6129238117912101, + "learning_rate": 2.783121617999651e-10, + "loss": 0.4312, + "step": 45282 + }, + { + "epoch": 0.9967642526964561, + "grad_norm": 2.0044534378835004, + "learning_rate": 2.7456392694247267e-10, + "loss": 0.3175, + "step": 45283 + }, + { + "epoch": 0.9967862645828748, + "grad_norm": 1.5533908062737305, + "learning_rate": 2.70841102920083e-10, + "loss": 0.2773, + "step": 45284 + }, + { + "epoch": 0.9968082764692934, + "grad_norm": 1.5755898791293201, + "learning_rate": 2.671436897505597e-10, + "loss": 0.3716, + "step": 45285 + }, + { + "epoch": 0.9968302883557121, + "grad_norm": 2.143305874818788, + "learning_rate": 2.6347168745388674e-10, + "loss": 0.3943, + "step": 45286 + }, + { + "epoch": 0.9968523002421308, + "grad_norm": 3.190939824291906, + "learning_rate": 2.5982509604782767e-10, + "loss": 0.2813, + "step": 45287 + }, + { + "epoch": 0.9968743121285494, + "grad_norm": 1.792710249568952, + "learning_rate": 2.5620391555181143e-10, + "loss": 0.2304, + "step": 45288 + }, + { + "epoch": 0.9968963240149681, + "grad_norm": 1.8353731584138204, + "learning_rate": 2.5260814598360163e-10, + "loss": 0.3282, + "step": 45289 + }, + { + "epoch": 0.9969183359013868, + "grad_norm": 1.7960936572926711, + "learning_rate": 2.490377873609617e-10, + "loss": 0.3123, + "step": 45290 + }, + { + "epoch": 0.9969403477878054, + "grad_norm": 1.966522446037981, + "learning_rate": 2.454928397033207e-10, + "loss": 0.312, + "step": 45291 + }, + { + "epoch": 0.9969623596742241, + "grad_norm": 1.6210205860738236, + "learning_rate": 2.4197330302788704e-10, + "loss": 0.2533, + "step": 45292 + }, + { + "epoch": 0.9969843715606428, + "grad_norm": 1.5209421490730346, + "learning_rate": 2.384791773524242e-10, + "loss": 0.3324, + "step": 45293 + }, + { + "epoch": 0.9970063834470614, + "grad_norm": 0.6229950644817266, + "learning_rate": 2.3501046269525096e-10, + "loss": 0.4393, + "step": 45294 + }, + { + "epoch": 0.9970283953334801, + "grad_norm": 1.601495162594212, + "learning_rate": 2.3156715907357575e-10, + "loss": 0.2659, + "step": 45295 + }, + { + "epoch": 0.9970504072198988, + "grad_norm": 1.4394032649039195, + "learning_rate": 2.281492665051621e-10, + "loss": 0.2022, + "step": 45296 + }, + { + "epoch": 0.9970724191063174, + "grad_norm": 2.191560747404158, + "learning_rate": 2.2475678500721853e-10, + "loss": 0.288, + "step": 45297 + }, + { + "epoch": 0.9970944309927361, + "grad_norm": 1.5810767477989662, + "learning_rate": 2.2138971459695348e-10, + "loss": 0.2925, + "step": 45298 + }, + { + "epoch": 0.9971164428791548, + "grad_norm": 2.5802373892622326, + "learning_rate": 2.1804805529157536e-10, + "loss": 0.3568, + "step": 45299 + }, + { + "epoch": 0.9971384547655734, + "grad_norm": 2.2284151216358743, + "learning_rate": 2.1473180710829266e-10, + "loss": 0.2739, + "step": 45300 + }, + { + "epoch": 0.9971604666519921, + "grad_norm": 1.8182375893820937, + "learning_rate": 2.1144097006320363e-10, + "loss": 0.3106, + "step": 45301 + }, + { + "epoch": 0.9971824785384107, + "grad_norm": 1.7713580195196914, + "learning_rate": 2.0817554417407182e-10, + "loss": 0.2553, + "step": 45302 + }, + { + "epoch": 0.9972044904248294, + "grad_norm": 2.336967831681717, + "learning_rate": 2.0493552945644035e-10, + "loss": 0.2953, + "step": 45303 + }, + { + "epoch": 0.9972265023112481, + "grad_norm": 2.215936742229239, + "learning_rate": 2.0172092592751769e-10, + "loss": 0.33, + "step": 45304 + }, + { + "epoch": 0.9972485141976667, + "grad_norm": 2.442881466414721, + "learning_rate": 1.9853173360340204e-10, + "loss": 0.2819, + "step": 45305 + }, + { + "epoch": 0.9972705260840854, + "grad_norm": 1.9520944660292991, + "learning_rate": 1.953679525007468e-10, + "loss": 0.2844, + "step": 45306 + }, + { + "epoch": 0.9972925379705041, + "grad_norm": 2.1435157290743536, + "learning_rate": 1.9222958263453994e-10, + "loss": 0.2678, + "step": 45307 + }, + { + "epoch": 0.9973145498569227, + "grad_norm": 3.6540440938471446, + "learning_rate": 1.8911662402198993e-10, + "loss": 0.2953, + "step": 45308 + }, + { + "epoch": 0.9973365617433414, + "grad_norm": 1.6500668640369862, + "learning_rate": 1.8602907667752967e-10, + "loss": 0.2648, + "step": 45309 + }, + { + "epoch": 0.9973585736297601, + "grad_norm": 1.5065343423967916, + "learning_rate": 1.829669406183676e-10, + "loss": 0.3319, + "step": 45310 + }, + { + "epoch": 0.9973805855161787, + "grad_norm": 2.7461464712467927, + "learning_rate": 1.7993021585893666e-10, + "loss": 0.2761, + "step": 45311 + }, + { + "epoch": 0.9974025974025974, + "grad_norm": 2.2422681881836315, + "learning_rate": 1.7691890241533506e-10, + "loss": 0.2589, + "step": 45312 + }, + { + "epoch": 0.9974246092890161, + "grad_norm": 1.7244836000142125, + "learning_rate": 1.7393300030255078e-10, + "loss": 0.4028, + "step": 45313 + }, + { + "epoch": 0.9974466211754347, + "grad_norm": 1.5290915387307251, + "learning_rate": 1.7097250953612698e-10, + "loss": 0.2577, + "step": 45314 + }, + { + "epoch": 0.9974686330618534, + "grad_norm": 0.6166024654129602, + "learning_rate": 1.6803743013049655e-10, + "loss": 0.4293, + "step": 45315 + }, + { + "epoch": 0.9974906449482721, + "grad_norm": 1.381692037160623, + "learning_rate": 1.6512776210064753e-10, + "loss": 0.2242, + "step": 45316 + }, + { + "epoch": 0.9975126568346907, + "grad_norm": 1.9330610321330648, + "learning_rate": 1.6224350546212298e-10, + "loss": 0.291, + "step": 45317 + }, + { + "epoch": 0.9975346687211094, + "grad_norm": 2.11929110930502, + "learning_rate": 1.5938466022880071e-10, + "loss": 0.3366, + "step": 45318 + }, + { + "epoch": 0.9975566806075281, + "grad_norm": 1.673635404698849, + "learning_rate": 1.5655122641511368e-10, + "loss": 0.3609, + "step": 45319 + }, + { + "epoch": 0.9975786924939467, + "grad_norm": 3.4069473914025408, + "learning_rate": 1.5374320403660493e-10, + "loss": 0.2995, + "step": 45320 + }, + { + "epoch": 0.9976007043803654, + "grad_norm": 1.6596461578641923, + "learning_rate": 1.5096059310604206e-10, + "loss": 0.211, + "step": 45321 + }, + { + "epoch": 0.9976227162667841, + "grad_norm": 1.9833598134128863, + "learning_rate": 1.4820339363896818e-10, + "loss": 0.3466, + "step": 45322 + }, + { + "epoch": 0.9976447281532027, + "grad_norm": 1.7066236436206554, + "learning_rate": 1.454716056481509e-10, + "loss": 0.2477, + "step": 45323 + }, + { + "epoch": 0.9976667400396214, + "grad_norm": 1.769055439919293, + "learning_rate": 1.4276522914802303e-10, + "loss": 0.4426, + "step": 45324 + }, + { + "epoch": 0.99768875192604, + "grad_norm": 2.031975547353962, + "learning_rate": 1.4008426415301756e-10, + "loss": 0.4112, + "step": 45325 + }, + { + "epoch": 0.9977107638124587, + "grad_norm": 1.8042433194805314, + "learning_rate": 1.374287106753469e-10, + "loss": 0.2448, + "step": 45326 + }, + { + "epoch": 0.9977327756988774, + "grad_norm": 2.190062186861203, + "learning_rate": 1.3479856872944397e-10, + "loss": 0.2858, + "step": 45327 + }, + { + "epoch": 0.997754787585296, + "grad_norm": 5.871521482441362, + "learning_rate": 1.3219383832863143e-10, + "loss": 0.2858, + "step": 45328 + }, + { + "epoch": 0.9977767994717147, + "grad_norm": 0.6148208052108013, + "learning_rate": 1.2961451948623195e-10, + "loss": 0.4474, + "step": 45329 + }, + { + "epoch": 0.9977988113581334, + "grad_norm": 1.6732959321630096, + "learning_rate": 1.2706061221501308e-10, + "loss": 0.2593, + "step": 45330 + }, + { + "epoch": 0.997820823244552, + "grad_norm": 2.6146680537178364, + "learning_rate": 1.2453211652774243e-10, + "loss": 0.3364, + "step": 45331 + }, + { + "epoch": 0.9978428351309707, + "grad_norm": 1.7100193927869103, + "learning_rate": 1.2202903243829778e-10, + "loss": 0.314, + "step": 45332 + }, + { + "epoch": 0.9978648470173894, + "grad_norm": 2.170481505839264, + "learning_rate": 1.1955135995833644e-10, + "loss": 0.2639, + "step": 45333 + }, + { + "epoch": 0.997886858903808, + "grad_norm": 1.622649584141581, + "learning_rate": 1.17099099100626e-10, + "loss": 0.3162, + "step": 45334 + }, + { + "epoch": 0.9979088707902267, + "grad_norm": 2.1779220001448882, + "learning_rate": 1.1467224987793401e-10, + "loss": 0.325, + "step": 45335 + }, + { + "epoch": 0.9979308826766454, + "grad_norm": 0.5852163184037686, + "learning_rate": 1.1227081230247294e-10, + "loss": 0.4474, + "step": 45336 + }, + { + "epoch": 0.997952894563064, + "grad_norm": 1.6005594743879934, + "learning_rate": 1.0989478638645523e-10, + "loss": 0.2691, + "step": 45337 + }, + { + "epoch": 0.9979749064494827, + "grad_norm": 2.3435737813735824, + "learning_rate": 1.0754417214209333e-10, + "loss": 0.3072, + "step": 45338 + }, + { + "epoch": 0.9979969183359014, + "grad_norm": 1.865823516776413, + "learning_rate": 1.052189695810446e-10, + "loss": 0.2784, + "step": 45339 + }, + { + "epoch": 0.99801893022232, + "grad_norm": 2.1031409651419555, + "learning_rate": 1.0291917871552149e-10, + "loss": 0.2532, + "step": 45340 + }, + { + "epoch": 0.9980409421087387, + "grad_norm": 2.078622380806674, + "learning_rate": 1.0064479955662621e-10, + "loss": 0.2363, + "step": 45341 + }, + { + "epoch": 0.9980629539951574, + "grad_norm": 2.0712709037036356, + "learning_rate": 9.839583211657122e-11, + "loss": 0.3041, + "step": 45342 + }, + { + "epoch": 0.998084965881576, + "grad_norm": 2.506732596264718, + "learning_rate": 9.617227640645877e-11, + "loss": 0.2511, + "step": 45343 + }, + { + "epoch": 0.9981069777679947, + "grad_norm": 0.5957863006179965, + "learning_rate": 9.397413243739107e-11, + "loss": 0.4395, + "step": 45344 + }, + { + "epoch": 0.9981289896544134, + "grad_norm": 4.082210533347547, + "learning_rate": 9.180140022102547e-11, + "loss": 0.3272, + "step": 45345 + }, + { + "epoch": 0.998151001540832, + "grad_norm": 1.5433960236307132, + "learning_rate": 8.965407976790908e-11, + "loss": 0.3647, + "step": 45346 + }, + { + "epoch": 0.9981730134272507, + "grad_norm": 0.622494746586416, + "learning_rate": 8.753217108914413e-11, + "loss": 0.4624, + "step": 45347 + }, + { + "epoch": 0.9981950253136693, + "grad_norm": 2.1648986745326884, + "learning_rate": 8.543567419583288e-11, + "loss": 0.2881, + "step": 45348 + }, + { + "epoch": 0.998217037200088, + "grad_norm": 1.561254176857879, + "learning_rate": 8.336458909852241e-11, + "loss": 0.2808, + "step": 45349 + }, + { + "epoch": 0.9982390490865067, + "grad_norm": 2.735721795452959, + "learning_rate": 8.131891580720474e-11, + "loss": 0.3172, + "step": 45350 + }, + { + "epoch": 0.9982610609729253, + "grad_norm": 2.9439607603234554, + "learning_rate": 7.929865433298212e-11, + "loss": 0.312, + "step": 45351 + }, + { + "epoch": 0.998283072859344, + "grad_norm": 1.5489502448314965, + "learning_rate": 7.730380468529142e-11, + "loss": 0.2378, + "step": 45352 + }, + { + "epoch": 0.9983050847457627, + "grad_norm": 1.352875517345543, + "learning_rate": 7.533436687523488e-11, + "loss": 0.2245, + "step": 45353 + }, + { + "epoch": 0.9983270966321813, + "grad_norm": 1.9210074008334244, + "learning_rate": 7.339034091224939e-11, + "loss": 0.2665, + "step": 45354 + }, + { + "epoch": 0.9983491085186, + "grad_norm": 2.3833367090649333, + "learning_rate": 7.147172680632696e-11, + "loss": 0.2218, + "step": 45355 + }, + { + "epoch": 0.9983711204050187, + "grad_norm": 1.3188517349149798, + "learning_rate": 6.95785245669045e-11, + "loss": 0.3065, + "step": 45356 + }, + { + "epoch": 0.9983931322914373, + "grad_norm": 1.846469043154879, + "learning_rate": 6.771073420452912e-11, + "loss": 0.3032, + "step": 45357 + }, + { + "epoch": 0.998415144177856, + "grad_norm": 2.6561794465273385, + "learning_rate": 6.586835572752748e-11, + "loss": 0.4158, + "step": 45358 + }, + { + "epoch": 0.9984371560642747, + "grad_norm": 1.3070523785489674, + "learning_rate": 6.40513891458916e-11, + "loss": 0.2382, + "step": 45359 + }, + { + "epoch": 0.9984591679506933, + "grad_norm": 1.8003707487256002, + "learning_rate": 6.225983446905837e-11, + "loss": 0.3392, + "step": 45360 + }, + { + "epoch": 0.998481179837112, + "grad_norm": 1.868807635082854, + "learning_rate": 6.049369170535447e-11, + "loss": 0.2926, + "step": 45361 + }, + { + "epoch": 0.9985031917235307, + "grad_norm": 2.5286710061068587, + "learning_rate": 5.87529608647719e-11, + "loss": 0.3437, + "step": 45362 + }, + { + "epoch": 0.9985252036099493, + "grad_norm": 5.724312621460305, + "learning_rate": 5.703764195563732e-11, + "loss": 0.2508, + "step": 45363 + }, + { + "epoch": 0.998547215496368, + "grad_norm": 1.7570391031596497, + "learning_rate": 5.5347734986277435e-11, + "loss": 0.3435, + "step": 45364 + }, + { + "epoch": 0.9985692273827868, + "grad_norm": 1.905926586898804, + "learning_rate": 5.368323996612912e-11, + "loss": 0.2565, + "step": 45365 + }, + { + "epoch": 0.9985912392692053, + "grad_norm": 0.6313210157041111, + "learning_rate": 5.204415690296394e-11, + "loss": 0.4398, + "step": 45366 + }, + { + "epoch": 0.998613251155624, + "grad_norm": 0.5974856344305445, + "learning_rate": 5.043048580510857e-11, + "loss": 0.4263, + "step": 45367 + }, + { + "epoch": 0.9986352630420428, + "grad_norm": 1.653239478877367, + "learning_rate": 4.884222668144478e-11, + "loss": 0.2996, + "step": 45368 + }, + { + "epoch": 0.9986572749284613, + "grad_norm": 0.6228528857003618, + "learning_rate": 4.727937953918904e-11, + "loss": 0.4451, + "step": 45369 + }, + { + "epoch": 0.99867928681488, + "grad_norm": 2.451048011339367, + "learning_rate": 4.574194438722312e-11, + "loss": 0.2701, + "step": 45370 + }, + { + "epoch": 0.9987012987012988, + "grad_norm": 1.9487139684964234, + "learning_rate": 4.422992123220837e-11, + "loss": 0.3388, + "step": 45371 + }, + { + "epoch": 0.9987233105877174, + "grad_norm": 0.6201184377346264, + "learning_rate": 4.274331008302657e-11, + "loss": 0.4331, + "step": 45372 + }, + { + "epoch": 0.998745322474136, + "grad_norm": 1.5300899647641397, + "learning_rate": 4.128211094633905e-11, + "loss": 0.1893, + "step": 45373 + }, + { + "epoch": 0.9987673343605546, + "grad_norm": 0.6154346274614159, + "learning_rate": 3.984632382991738e-11, + "loss": 0.4355, + "step": 45374 + }, + { + "epoch": 0.9987893462469734, + "grad_norm": 2.900798646383567, + "learning_rate": 3.8435948741533116e-11, + "loss": 0.3817, + "step": 45375 + }, + { + "epoch": 0.9988113581333921, + "grad_norm": 1.7602702101729168, + "learning_rate": 3.7050985687292484e-11, + "loss": 0.2761, + "step": 45376 + }, + { + "epoch": 0.9988333700198107, + "grad_norm": 1.4928909319972778, + "learning_rate": 3.569143467496705e-11, + "loss": 0.3671, + "step": 45377 + }, + { + "epoch": 0.9988553819062294, + "grad_norm": 1.7078694488818482, + "learning_rate": 3.435729571121815e-11, + "loss": 0.2341, + "step": 45378 + }, + { + "epoch": 0.9988773937926481, + "grad_norm": 2.171968492221599, + "learning_rate": 3.304856880326224e-11, + "loss": 0.2625, + "step": 45379 + }, + { + "epoch": 0.9988994056790667, + "grad_norm": 2.4976506939030183, + "learning_rate": 3.176525395665042e-11, + "loss": 0.2786, + "step": 45380 + }, + { + "epoch": 0.9989214175654854, + "grad_norm": 0.6036354735730549, + "learning_rate": 3.050735117915426e-11, + "loss": 0.4492, + "step": 45381 + }, + { + "epoch": 0.9989434294519041, + "grad_norm": 1.767523572969259, + "learning_rate": 2.9274860476879996e-11, + "loss": 0.2953, + "step": 45382 + }, + { + "epoch": 0.9989654413383227, + "grad_norm": 1.7423077912998666, + "learning_rate": 2.806778185537873e-11, + "loss": 0.3274, + "step": 45383 + }, + { + "epoch": 0.9989874532247414, + "grad_norm": 1.6841438771791406, + "learning_rate": 2.6886115321866914e-11, + "loss": 0.2748, + "step": 45384 + }, + { + "epoch": 0.9990094651111601, + "grad_norm": 1.8949109012030143, + "learning_rate": 2.5729860881340552e-11, + "loss": 0.2393, + "step": 45385 + }, + { + "epoch": 0.9990314769975787, + "grad_norm": 2.0345024588620135, + "learning_rate": 2.4599018540460984e-11, + "loss": 0.2139, + "step": 45386 + }, + { + "epoch": 0.9990534888839974, + "grad_norm": 2.469704476466553, + "learning_rate": 2.3493588304224212e-11, + "loss": 0.3387, + "step": 45387 + }, + { + "epoch": 0.9990755007704161, + "grad_norm": 3.480507237571116, + "learning_rate": 2.2413570178736466e-11, + "loss": 0.2526, + "step": 45388 + }, + { + "epoch": 0.9990975126568347, + "grad_norm": 1.605902668161001, + "learning_rate": 2.1358964169548856e-11, + "loss": 0.4055, + "step": 45389 + }, + { + "epoch": 0.9991195245432534, + "grad_norm": 2.74962628060194, + "learning_rate": 2.0329770281657392e-11, + "loss": 0.3895, + "step": 45390 + }, + { + "epoch": 0.9991415364296721, + "grad_norm": 1.6409754475074534, + "learning_rate": 1.9325988520613183e-11, + "loss": 0.3549, + "step": 45391 + }, + { + "epoch": 0.9991635483160907, + "grad_norm": 1.4285077360744198, + "learning_rate": 1.8347618891412232e-11, + "loss": 0.2953, + "step": 45392 + }, + { + "epoch": 0.9991855602025094, + "grad_norm": 1.872506684093258, + "learning_rate": 1.739466139905055e-11, + "loss": 0.2687, + "step": 45393 + }, + { + "epoch": 0.9992075720889281, + "grad_norm": 3.8474722082939525, + "learning_rate": 1.6467116048524135e-11, + "loss": 0.2854, + "step": 45394 + }, + { + "epoch": 0.9992295839753467, + "grad_norm": 1.5930115460900862, + "learning_rate": 1.556498284427388e-11, + "loss": 0.2898, + "step": 45395 + }, + { + "epoch": 0.9992515958617654, + "grad_norm": 1.118906270199821, + "learning_rate": 1.4688261790740677e-11, + "loss": 0.2198, + "step": 45396 + }, + { + "epoch": 0.999273607748184, + "grad_norm": 1.672209950358603, + "learning_rate": 1.383695289292053e-11, + "loss": 0.2375, + "step": 45397 + }, + { + "epoch": 0.9992956196346027, + "grad_norm": 1.5697052246632632, + "learning_rate": 1.3011056154699219e-11, + "loss": 0.3729, + "step": 45398 + }, + { + "epoch": 0.9993176315210214, + "grad_norm": 1.5439792602368083, + "learning_rate": 1.2210571579962527e-11, + "loss": 0.1871, + "step": 45399 + }, + { + "epoch": 0.99933964340744, + "grad_norm": 1.6818807038631571, + "learning_rate": 1.1435499173706455e-11, + "loss": 0.2845, + "step": 45400 + }, + { + "epoch": 0.9993616552938587, + "grad_norm": 1.5247012509335192, + "learning_rate": 1.0685838939261672e-11, + "loss": 0.297, + "step": 45401 + }, + { + "epoch": 0.9993836671802774, + "grad_norm": 0.62785689765133, + "learning_rate": 9.961590881069071e-12, + "loss": 0.4422, + "step": 45402 + }, + { + "epoch": 0.999405679066696, + "grad_norm": 1.619756348253367, + "learning_rate": 9.262755001904211e-12, + "loss": 0.3563, + "step": 45403 + }, + { + "epoch": 0.9994276909531147, + "grad_norm": 0.6249568423482206, + "learning_rate": 8.589331305652871e-12, + "loss": 0.4375, + "step": 45404 + }, + { + "epoch": 0.9994497028395334, + "grad_norm": 0.6157657624214368, + "learning_rate": 7.94131979564572e-12, + "loss": 0.4243, + "step": 45405 + }, + { + "epoch": 0.999471714725952, + "grad_norm": 2.891965068912596, + "learning_rate": 7.318720475768537e-12, + "loss": 0.2905, + "step": 45406 + }, + { + "epoch": 0.9994937266123707, + "grad_norm": 2.4835813486218905, + "learning_rate": 6.721533348241771e-12, + "loss": 0.3721, + "step": 45407 + }, + { + "epoch": 0.9995157384987894, + "grad_norm": 1.2101416873068778, + "learning_rate": 6.149758416951201e-12, + "loss": 0.2004, + "step": 45408 + }, + { + "epoch": 0.999537750385208, + "grad_norm": 1.6808967420243193, + "learning_rate": 5.603395684672386e-12, + "loss": 0.2565, + "step": 45409 + }, + { + "epoch": 0.9995597622716267, + "grad_norm": 2.138544842513439, + "learning_rate": 5.0824451536257706e-12, + "loss": 0.2693, + "step": 45410 + }, + { + "epoch": 0.9995817741580454, + "grad_norm": 2.295510837235524, + "learning_rate": 4.586906826586912e-12, + "loss": 0.3582, + "step": 45411 + }, + { + "epoch": 0.999603786044464, + "grad_norm": 1.7298210282304605, + "learning_rate": 4.11678070688648e-12, + "loss": 0.3131, + "step": 45412 + }, + { + "epoch": 0.9996257979308827, + "grad_norm": 2.2798595156814905, + "learning_rate": 3.67206679618981e-12, + "loss": 0.3129, + "step": 45413 + }, + { + "epoch": 0.9996478098173014, + "grad_norm": 1.5535683573614643, + "learning_rate": 3.2527650967173476e-12, + "loss": 0.2533, + "step": 45414 + }, + { + "epoch": 0.99966982170372, + "grad_norm": 0.6291366019969814, + "learning_rate": 2.858875610689538e-12, + "loss": 0.455, + "step": 45415 + }, + { + "epoch": 0.9996918335901387, + "grad_norm": 1.719529277254461, + "learning_rate": 2.4903983403268274e-12, + "loss": 0.3016, + "step": 45416 + }, + { + "epoch": 0.9997138454765574, + "grad_norm": 2.4127435676308306, + "learning_rate": 2.147333287294551e-12, + "loss": 0.2727, + "step": 45417 + }, + { + "epoch": 0.999735857362976, + "grad_norm": 2.1076910635235486, + "learning_rate": 1.829680453813154e-12, + "loss": 0.3171, + "step": 45418 + }, + { + "epoch": 0.9997578692493947, + "grad_norm": 0.6092187133328507, + "learning_rate": 1.5374398404377488e-12, + "loss": 0.4625, + "step": 45419 + }, + { + "epoch": 0.9997798811358133, + "grad_norm": 1.5689381814586674, + "learning_rate": 1.2706114499438927e-12, + "loss": 0.235, + "step": 45420 + }, + { + "epoch": 0.999801893022232, + "grad_norm": 1.6737331100162536, + "learning_rate": 1.029195282886697e-12, + "loss": 0.2767, + "step": 45421 + }, + { + "epoch": 0.9998239049086507, + "grad_norm": 1.8603211578136396, + "learning_rate": 8.131913403763847e-13, + "loss": 0.2491, + "step": 45422 + }, + { + "epoch": 0.9998459167950693, + "grad_norm": 3.0215459997447534, + "learning_rate": 6.225996240782906e-13, + "loss": 0.3112, + "step": 45423 + }, + { + "epoch": 0.999867928681488, + "grad_norm": 1.7186446519124707, + "learning_rate": 4.57420134547526e-13, + "loss": 0.2755, + "step": 45424 + }, + { + "epoch": 0.9998899405679067, + "grad_norm": 3.2504839703842356, + "learning_rate": 3.176528723392025e-13, + "loss": 0.2609, + "step": 45425 + }, + { + "epoch": 0.9999119524543253, + "grad_norm": 1.2731352010457608, + "learning_rate": 2.0329783911865464e-13, + "loss": 0.2155, + "step": 45426 + }, + { + "epoch": 0.999933964340744, + "grad_norm": 1.7179912727948858, + "learning_rate": 1.143550348858824e-13, + "loss": 0.2306, + "step": 45427 + }, + { + "epoch": 0.9999559762271627, + "grad_norm": 0.6290774362536197, + "learning_rate": 5.082446019599729e-14, + "loss": 0.4608, + "step": 45428 + }, + { + "epoch": 0.9999779881135813, + "grad_norm": 1.6515911659017182, + "learning_rate": 1.2706115048999323e-14, + "loss": 0.257, + "step": 45429 + }, + { + "epoch": 1.0, + "grad_norm": 0.6539189741645602, + "learning_rate": 0.0, + "loss": 0.4111, + "step": 45430 + }, + { + "epoch": 1.0, + "step": 45430, + "total_flos": 8488113377509376.0, + "train_loss": 0.35205436017599606, + "train_runtime": 225847.8419, + "train_samples_per_second": 25.748, + "train_steps_per_second": 0.201 + } + ], + "logging_steps": 1.0, + "max_steps": 45430, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 5000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 8488113377509376.0, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}